Downloaded 22 times
![Nonlinear Observers for Closed loop Sliding Mode
Control of Quadrotor UAV
Junaid Anwar, Fahad Mumtaz Malik, Muhammad Bilal Khan
College of Electrical and Mechanical Engineering, National University of Sciences and Technology, Pakistan
Abstract—This paper deals with the performance comparison
of Sliding mode observer with super-twisting algorithm (STSMO)
and High gain Observer (HGO) for a remotely controlled quad
rotor UAV. Under the restriction that inertial co-ordinates and
attitude angles are available for measurement while angular
and linear velocities are estimated. This paper is solved in
two steps for each observer. First the observer (HGO and
STSMO) is designed and then in second stage a second order
(2-SMC) technique is being applied on the basis of estimated
states to design controller(for which systems is portioned into
fully and under-actuated sub-systems).Simulations results shows
the performance comparison of both observers under the same
control scheme.
I. INTRODUCTION
More recently, a growing interest in the UAV has been
shown by industry and academia [1]–[7].The vital and poten-
tial use of flying robots for civil as well as military applications
are attracting the industries and the academia community. The
feature of flying in narrow space and vertical takeoff and land-
ing (VTOL) made quadrotor unique relative to other mobile
robots and conventional aircrafts. The quadrotor is an under
actuated system with six outputs and four inputs, they are
owed to carry out the tasks ranging from surveillance to rescue
mission but the challenges behind the control of quadrotor
aerial vehicle like un-stability and highly nonlinear behavior
are the major source of attraction and many control approaches
to deal with quadrotor dynamics have been presented so far
[8]–[14][14-20].
This paper deals with the development of 2-sliding mode
control scheme that can cater for the model uncertainties,
external disturbances and the chattering phenomenon. Non-
availability of states is a major constraint towards accomplish-
ment of any control scheme, using sensor for each state is
also not feasible due to space limitations and high cost of the
sensors. Even with the availability of all states system model
generally shows parametric mismatch with respect to the
real time environment. These model imperfections, un-certain
initialization and sensor errors also degrades the performance
of the controller. The solution for that is to use state observers,
to estimate the states in real time, Luenberger [15] proves to be
good in the state estimation but these model based observers
fail when the system parameters keep on changing with the
time. Least square and recursive least square (RLS) are also
not able to work on highly nonlinear system such as quadrotor.
A high gain observer was first introduced by Khalil and
Esfandiaro [16] for the design of output feedback controllers
and asymptotic convergence. Researchers have contributed in
the development of their idea towards high gain observer [17]–
[20].High gain observers performance degrades in the presence
of external noise and is shown in simulation section of this
paper, due to which we have to look for an observer which is
robust to sensor noise.
The sliding mode observers are widely used because of their
prominent features like finite time convergence, robustness to
sensor noise and un-certain estimation [21], [22].Asymptotic
second order sliding mode observers were also developed but
they require proof of separation principle.High accuracy and
reduction in chattering are the main features of second order
sliding mode compared to the classical first order motion.
Recently a class of second order sliding mode observer is
introduced so called super-twisting observer [23] for second
order mechanical system which include quadrotor too. Super-
twisting observers can reconstruct the states if the perturbation
is of relative degree two, or reconstructs the perturbation itself,
when it is of relative degree one in finite time.
Aim of this paper is to compare the performance of both ob-
servers namely high-gain observer and super-twisting sliding
mode observer under same set of perturbations, uncertainties
and noise, so that each observer can exhibit its characteristics
for the quadrotor system. Real time estimation always require
knowledge of the pros and cons of the observer relative to the
particular system, so that best observer can be deployed for
real time estimation of states. So there is a need to explore
the type of observers which are application specific.
In the following section model of the quadrotor is developed
and presented. In section-III controller design is presented. In
section-IV observers are designed for the quadrotor model. In
section-V numerical simulation is given and finally in section-
V1 the conclusions are given.
II. DYNAMIC MODELING
A quad rotor UAV is a highly nonlinear dynamical system
and its modeling it is not an easy task due to under actua-
tion. It consists of two pairs of rotors which are moving in
opposite directions to provide the collective thrust as shown
in following fig 1. There are four inputs to this system.
The input U1 is the sum of thrusts provided by individual
rotors. The pitch movement is obtained by changing speeds
of rotors 2 and 4. Similarly the roll movement is achieved by
varying speeds of rotor 1 and 3. These two former operations
should be performed while keeping the total thrust constant
1
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-10-320.jpg)
![To make the model more realistic especially in forward
flight we include the hub forces, rolling moments and variable
aerodynamics coefficients [24].
The hub force is the resultant of horizontal forces acting on
all blade elements.
H = CHρA ΩRrad
2
where CH is the hub coefficient, ρ is the air density and A is
the propeller disk area and Rrad is the propeller radius and ρ
is the speed of the respective propeller.Additionally the drag
moment Q is the moment about the rotor shaft caused by the
aerodynamic forces acting on the blades. In fact drag moments
determine the power required to spin the rotors. It is given by
the following equation
Q = CQρA ΩRrad
2
Rrad
where CQ is the drag coefficient.The rolling moment of
a propeller exists in forward flight when advancing blade
is producing more lift than the retreating blade. It is the
integration over the entire rotor of the lift of each section
acting at a given radius and is given by following equation.
Rm = CRm
ρA ΩRrad
2
Rrad
Where CRm is the rolling moment coefficient.Furthermore the
UAVs operating near the ground (approximately at half rotor
diameter) experience thrust augmentation due to better rotor
efficiency. This is related to a reduction of induced airflow
velocity. This is called Ground Effect. The following equation
represents the ground effect near the surface. It is assumed
that ground effect acts on the UAV when the UAV is below a
certain altitude,zo.
Fgr z =
A
z + zcg)2
−
A
zo + zcg)2
0 < z ≤ zo
After incorporating the above effects and the friction terms,
we obtain a more realistic model of the quad rotor UAV which
is given by the following set of equations:-
¨x = cos φ sin θ cos ψ + sin φ sin ψ
U1
m
−
1
m
4
i=1
Hxi − K1
˙x
ms
¨y = cos φ sin θ sin ψ − sin φ sin ψ
U1
m
−
1
m
4
i=1
Hyi − K2
˙y
ms
¨z = −g +
cos φ cos θ
ms
U1 + Fgr z − K3
z
ms
¨φ = ˙θ ˙ψ
Iy − Iz
Ix
−
Jr
Ix
Ωr
˙θ +
l
Ix
U2 −
h
Ix
4
i=1
Hyi+
(−1)i+1
Ix
4
i=1
Rms
xi − lK4
˙φ
Ix
¨θ = ˙φ ˙ψ
Iz − Ix
Iy
−
Jr
Iy
Ωr
˙φ +
l
Iy
U3 −
h
Iy
4
i=1
Hyi+
(−1)i+1
Iy
4
i=1
Rms
yi − lK5
˙θ
Iy
¨ψ = ˙θ ˙φ
Ix − Iy
Iz
+
C
Iz
U4 +
h
Iz
4
i=1
Hyi+
l
Iz
Hx2 − Hx4 + Hy3 − Hy1
4
i=1
Qiyi−
lK6
˙ψ
Iz
(2)
III. CONTROLLER DESIGN
Let X = x, ˙x, y, ˙y, z, ˙z, φ, ˙φ, θ, ˙θ, ψ, ˙ψ,
T
and U =
U1, U2, U3, U4
T
be the state and control input vectors re-
spectively. The equation set (1) with the addition of friction
term and ground effect term for altitude can be written in state
space representation such as:
˙x1 = x2
˙x2 = cos x7 sin x9 cos x11 + sin x7 sin x11
U1
m
− K1
x2
ms
(3)
˙x3 = x4
˙x4 = cos x7 sin x9 sin x11 − sin x7 sin x11
U1
m
− K2
x4
ms
(4)
˙x5 = x6
˙x6 = −g +
cos x7 cos x9
ms
U1 + Fgr z − K3
x6
ms
(5)
3
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-12-320.jpg)
![To make it negative definite choice of input is as:
U3 =
Iy
l
¨θd − ˙φ ˙ψ
Iz − Ix
Iy
−
Jr
Iy
Ωr
˙φ − lK5
˙θ
Iy
+ α3 ˙xd − x
+ α4
˙θd − ˙θ + ¨xd − ¨x + k5sat(S3) + k6S3
where k5, k6 > 0
Similarly in the same way surface for subsystem (4)
and eqref3f comes out to be the linear combination of position
and velocity tracking errors of two states i.e. y and ψ.
S4 = ˙yd − ˙y + α5 yd − y + ˙ψd − ˙ψ + α6 ψd − ψ
where α5, α6 > 0
and in the same way by the Lyapunov analysis of surface S4
the control U4 comes out to be
U4 =
Iz
c
¨ψd − ˙φ ˙θ
Ix − Iy
Iz
− lK6
˙ψ
Iz
+ α5 ˙yd − ˙y
− α6
˙ψd − ˙ψ + ¨yd − ¨y + k7sat(S4) + k8S4
where k7, k8 > 0
U1 is the control input of z, U2 is the control input of roll,U3
is the control input of pitch and U4 is the control input
of yaw while motion in x and y direction is produced by
help of control inputs of roll, pitch and z. By the help of
U1, U2, U3 and U4 the desired trajectories are achieved and
tracking errors are reduced to zero asymptotically, by virtue
of Sliding mode controller.by keeping the roll and pitch angles
to zero controller robustly stabilize the UAV and move it to the
desired position with a desired yaw angle. The control scheme
is developed and implemented independent of observer and is
shown in the block diagram The controller is designed by
keeping in view the mathematical model of the quad rotor as
given in Section-II without any effects except friction and the
ground effect but that U1, U2, U3 and U4 are capable enough to
tackle not only Rolling Moments, Drag moments, Gyroscopic
effects, Hub forces but also retain its performance which is
being shown in the simulations section in this paper.
Fig. 2. Controller designed independent of the observer
IV. OBSERVER DESIGN
Two types of nonlinear observers are implemented for the
quad rotor system with the same control scheme i.e. 2-Sliding
mode Control. Observability is ensured by [25] for each block
of equation from (3)–(8) separately.
A. High Gain Observer
HGO is basically an approximate differentiator. This ob-
server works well for a wide class of nonlinear systems and
leads to recovery of the performance achieved under state
feedback. Implementation of this observer is quite simple
because it needs less computational effort with an additive
advantage of this observer is that its performance doesn’t
degrade with the presence of model uncertainties in the plant.
High gain observer is an asymptotic observer and dynamics of
this observer can be made arbitrarily fast through epsilon and
gains alpha’s. Separation principle theorem doesn’t need to
be proved and high gain observer can be designed separately
from the controller. The HGO is applied to multiple input and
multiple output system as:
˙x = Ax + Bϑ x, u
The HGO, then, is given by
ˆ˙x = Aˆx + Bϑ0 ˆx, u + H y − Cˆx
y = Cˆx
Where H = blockdiag H1, H2, H3, H4 ,
Hi =
∂1
1
ε
∂2
2
ε
i = 1, 2, 3, 4
ε = positive constant, and constant parameter ∂i
j are
obtained from a Hurwitz polynomial, j=1,2
s2
+ ∂1
1 s + ∂2
2 = 0
The HGO for the quadcopter system is designed in blocks
as [26] i.e six HGOs are designed for each block of equation
from (3)–(8) separately. For equation (3) HGO is implemented
as
With A =
0 1
0 0
, B =
0
1
and C =
1
0
and
ϑ0 = cos x7 sin x9 cos x11 + sin x7 sin x11
U1
m
− K1
x2
ms
and H1 is designed as in the aforementioned equation. Simi-
larly for equation (4), (5),.....(8).HGOs are given in the same
way as for equation (3).constant gains are enlisted in the table
1 in simulation section of this paper.
5
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-14-320.jpg)
![B. Sliding mode Observer with super-twisting Algorithm
One of the best sliding mode observer which offers a
finite time reaching time [27], [23] and which can be used
for sliding mode based observation is the super-twisting ob-
server.Separation principle theorem is trivial in this case too
and the Super-twisting sliding mode observer can be designed
separately from the controller. The finite time convergence
property of sliding mode observers is usually suitable in the
scheme of observation and for the purpose of observer-based
controller design for nonlinear systems Super-twisting sliding
mode observer has the form
ˆ˙x1 = ˆx2 + λ|x1 − ˆx1|1/2
sat x1 − ˆx1 (9)
ˆ˙x2 = f x1, ˆx2, u + τsat x1 − ˆx1 (10)
Taking ˜x1 = x1 − ˆx1 and ˜x2 = x2 − ˆx2 we obtain the error
equations as
˙˜x1 = ˜x2 − λ|˜x1|1/2
sat ˜x1
˙˜x2 = F t, x1, x2, ˆx2 − αsat ˜x1
where,
F x1, x2, ˆx2 = f x1, x2, u − f x1, ˆx2, u + ξ x1, x2, y
ξ is used for perturbations.For the bounded states, existence
of a constant is ensured such that
|F x1, x2, ˆx2 |< f+
Observer designed by equation (9) and (10) takes into ac-
count of partial knowledge of system dynamics while setting
parameters λ and τ and hence more accurate. The full order
Super-twisting Sliding mode observer for equation (3) is given
as
ˆ˙x1 = λ1 + ˆx1|x1 − ˆx1|1/2
sat x1 − ˆx1
ˆ˙x2 =
1
ms
cos ˆφ sin ˆθ cos ˆψ + sin ˆφ sin ˆψ U1 − K1
ˆx2
ms
τ1sat x1 − ˆx1
τ1 and λ1 are designed by the help of aforementioned in-
equality as [27]. Similarly for equations (4), (5),....., (8) Super-
twisting sliding mode observers are implemented in the same
way as for equation (3), while gains are given in the table.1
in simulation section of this paper.
V. SIMULATION STUDY
A. Closed loop Simulation with model uncertainties and with-
out noise for HGO and STSMO
Simulation results for observer-based controller of the
quadrotor are shown in the fig.3 and fig.4 for both observers
HGO and STSMO. Now under output feedback, controller
is in conjunction with HGO and STSMO separately and is
using all the states of the observer. The results in the fig.3
shows that both High gain observer and as well as Super-
twisting sliding mode observer recovers the performance of
state feedback after 12 seconds while x state a bit earlier as
compared to other states, both observers are initialized with
same initial conditions, so that performance can be compared
in a proper way and all the observers parameters are listed
in the table 1 the performance of HGO is slightly better than
STSMO in tracking as HGO estimates desired values a bit
earlier as compared to STSMO. The chattering problem is
intelligently avoided in the sliding mode control by using
continuous approximation to the sign function. This makes this
approach applicable in real applications. As the control laws
are developed for set of equations (1) but implemented on
set of equations (2) which include different kind of effects as
mentioned in section-II, similarly the model used for observer
is based on set of equations (1) and observer giving estimate
for set of equations (2) which is quadrotor model with ground
effects, drag moment, rolling moment and pitch moment.
Fig. 3. HGO and STSMO tracking of desired values
Fig. 4. HGO and STSMO tracking of desired values
B. Closed loop Simulation with noise and model Uncertainties
for HGO and STSMO
A constant noise of 0.1 value is added in the output of
the system in each of six states and the results obtained after
simulation for each observer are shown in the fig.5 and fig.6
6
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-15-320.jpg)
![doesnt allow random initialization unless gains are adjusted on
the other hand super-twisting sliding mode observer provides
flexible environment in initialization.
None of these observes is computationally onerous, but super-
twisting sliding mode observer utilizes the knowledge of
system partially [27] as compared to high-gain observer which
just rely on the Hurwitz polynomial and the tuning parameter
ε [26] so this fact also indicates the practical applicability of
super-twisting sliding mode observer in the cases where model
uncertainties are bounded as it gives finite time convergence as
compared to asymptotic convergence in the case of high gain
observer which are favourable in the environment where model
uncertainties are present or parameters are time varying in
those conditions these filters are preferable than super-twisting
sliding mode observer only. This effort will be a good starting
point to explore super-twisting sliding mode observers and to
compare them with other observers of its breed (higher order
sliding mode observers).
REFERENCES
[1] M. Chen and M. Huzmezan, “A simulation model and h (loop shaping
control of a quad rotor unmanned air vehicle.” in Modelling, Simulation,
and Optimization, 2003, pp. 320–325.
[2] S. Bouabdallah, P. Murrieri, and R. Siegwart, “Design and control
of an indoor micro quadrotor,” in Robotics and Automation, 2004.
Proceedings. ICRA’04. 2004 IEEE International Conference on, vol. 5.
IEEE, 2004, pp. 4393–4398.
[3] B. Heriss´e, T. Hamel, R. Mahony, and F.-X. Russotto, “Landing a vtol
unmanned aerial vehicle on a moving platform using optical flow,”
Robotics, IEEE Transactions on, vol. 28, no. 1, pp. 77–89, 2012.
[4] L. Derafa, A. Benallegue, and L. Fridman, “Super twisting control
algorithm for the attitude tracking of a four rotors uav,” Journal of the
Franklin Institute, vol. 349, no. 2, pp. 685–699, 2012.
[5] A. Tayebi and S. McGilvray, “Attitude stabilization of a vtol quadrotor
aircraft,” Control Systems Technology, IEEE Transactions on, vol. 14,
no. 3, pp. 562–571, 2006.
[6] L. Besnard, Y. B. Shtessel, and B. Landrum, “Quadrotor vehicle control
via sliding mode controller driven by sliding mode disturbance observer,”
Journal of the Franklin Institute, vol. 349, no. 2, pp. 658–684, 2012.
[7] M. Bouchoucha, S. Seghour, and M. Tadjine, “Classical and second
order sliding mode control solution to an attitude stabilization of a four
rotors helicopter: From theory to experiment,” in Mechatronics (ICM),
2011 IEEE International Conference on. IEEE, 2011, pp. 162–169.
[8] E. Altu˘g, J. P. Ostrowski, and R. Mahony, “Control of a quadrotor
helicopter using visual feedback,” in Robotics and Automation, 2002.
Proceedings. ICRA’02. IEEE International Conference on, vol. 1. IEEE,
2002, pp. 72–77.
[9] E. Altu˘g, J. P. Ostrowski, and C. J. Taylor, “Quadrotor control using
dual camera visual feedback,” in Robotics and Automation, 2003.
Proceedings. ICRA’03. IEEE International Conference on, vol. 3. IEEE,
2003, pp. 4294–4299.
[10] T. Madani and A. Benallegue, “Control of a quadrotor mini-helicopter
via full state backstepping technique,” in Decision and Control, 2006
45th IEEE Conference on. IEEE, 2006, pp. 1515–1520.
[11] ——, “Backstepping sliding mode control applied to a miniature quadro-
tor flying robot,” in IEEE Industrial Electronics, IECON 2006-32nd
Annual Conference on. IEEE, 2006, pp. 700–705.
[12] P. Castillo, P. Albertos, P. Garcia, and R. Lozano, “Simple real-time
attitude stabilization of a quad-rotor aircraft with bounded signals,” in
Decision and Control, 2006 45th IEEE Conference on. IEEE, 2006,
pp. 1533–1538.
[13] N. Metni, T. Hamel, and F. Derkx, “Visual tracking control of aerial
robotic systems with adaptive depth estimation,” in Decision and Con-
trol, 2005 and 2005 European Control Conference. CDC-ECC’05. 44th
IEEE Conference on. IEEE, 2005, pp. 6078–6084.
[14] A. Benallegue, A. Mokhtari, and L. Fridman, “Feedback linearization
and high order sliding mode observer for a quadrotor uav,” in Variable
Structure Systems, 2006. VSS’06. International Workshop on. IEEE,
2006, pp. 365–372.
[15] D. G. Luenberger, “Observers for multivariable systems,” Automatic
Control, IEEE Transactions on, vol. 11, no. 2, pp. 190–197, 1966.
[16] F. Esfandiari and H. K. Khalil, “Output feedback stabilization of fully
linearizable systems,” International Journal of control, vol. 56, no. 5,
pp. 1007–1037, 1992.
[17] A. N. Atassi and H. K. Khalil, “A separation principle for the stabiliza-
tion of a class of nonlinear systems,” IEEE Transactions on Automatic
Control, vol. 44, no. 9, pp. 1672–1687, 1999.
[18] A. Isidori, “A remark on the problem of semiglobal nonlinear output
regulation,” IEEE transactions on Automatic Control, vol. 42, no. 12,
pp. 1734–1738, 1997.
[19] Z. Lin and A. Saberi, “Robust semiglobal stabilization of minimum-
phase input-output linearizable systems via partial state and output
feedback,” Automatic Control, IEEE Transactions on, vol. 40, no. 6,
pp. 1029–1041, 1995.
[20] W. J. Rugh, Linear system theory. prentice hall Upper Saddle River,
NJ, 1996, vol. 2.
[21] J. Barbot, M. Djemai, and T. Boukhobza, “Sliding mode observers,”
Sliding Mode Control in Engineering, vol. 11, 2002.
[22] C. Edwards, S. K. Spurgeon, and C. P. Tan, “On the development and
application of sliding mode observers,” in Variable Structure Systems:
Towards the 21st Century. Springer, 2002, pp. 253–282.
[23] J. Davila, L. Fridman, and A. Levant, “Second-order sliding-mode
observer for mechanical systems,” IEEE transactions on automatic
control, vol. 50, no. 11, pp. 1785–1789, 2005.
[24] M. Becker, R. C. B. Sampaio, S. Bouabdallah, V. d. Perrot, and
R. Siegwart, “In-flight collision avoidance controller based only on
os4 embedded sensors,” Journal of the Brazilian Society of Mechanical
Sciences and Engineering, vol. 34, no. 3, pp. 294–307, 2012.
[25] I. Khan, A. Bhatti, Q. Khan, and Q. Ahmad, “Sliding mode control
of lateral dynamics of an auv,” in Applied Sciences and Technology
(IBCAST), 2012 9th International Bhurban Conference on. IEEE, 2012,
pp. 27–31.
[26] H. K. Khalil and J. Grizzle, Nonlinear systems. Prentice hall New
Jersey, 1996, vol. 3.
[27] Y. Shtessel, C. Edwards, L. Fridman, and A. Levant, Sliding mode
control and observation. Springer, 2014.
8
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-17-320.jpg)
![Figure 12 sensor response when temperature falls
VI.INDUCTION HEATING
Induction heating has replaced the traditional furnace
methods because of its efficiency, unpolluted and fast heating
process. Electrically conductive materials are used in
induction heating for heating purposes and this requires high
frequency electricity. An IH system requires a source of
alternating current, an induction coil, and the work piece to be
heated. A magnetic field is generated in the coil due to the
alternating current passing through the coil. The AC is
supplied by the inverter. Work piece placed within the coil
will experienced the magnetic field due to which eddy
currents are induced in the work piece that cause non-contact
type of heating between work piece and the induction coil.
Copper tube is used to make induction coil. The tube is
hollow inside with coil diameter about 3 inches and internal
diameter of tube is about 1cm.the coil is given 8 turns. The
impedance of coil depends on cross-sectional area, length and
the number of turns so to increase the coil impedance.
The impedance matching circuit is designed such that it
converts high volt/low current (coming out from inverter) to
low volt/high current (driven requirement of the load). The
choice is made for impedance matching in order to match the
output parameters of inverter with the input parameter of coil.
Parallel capacitor bank is to be connected between the coil and
inverter.
VII. CONCLUSION
The FLC for induction heating has been presented in this
paper. The integrated controller, implemented on DSKC6713
takes the set temperature (required temperature of load) input
from user, reads actual temperature of load and calculates
error and rate of error, then on basis of Fuzzy Logic, it takes
decision and determines the amount of ΔF that needs to be
added/subtracted. The FLC then feeds this ΔF to/from
operating frequency of variable frequency inverter.
The FL is being implemented in international industries but
is new for Pakistani industries. However, Pakistan industries
are doing their controlling on fuzzy logic but it is a joint
venture of PID and FUZZY. This idea, however, introduces
Fuzzy Temperature Controller as a standalone product.
BIBLIOGRAPHY
[1] Chin-Hsing Cheng, “Design of Fuzzy Controller for Induction
Heating Using DSP”, 5th
IEEE Conference on Industrial Electronics
and Applications, 2010
[2] Yunseop Kim, “Fuzzy Logic Temperature Controller”, Physics
344 project, 2001
[3] Datasheet IRF840:
http://www.datasheetcatalog.com/datasheets_pdf/I/R/F/8/IRF840.sht
ml
[4] HV Floating MOS-Gate Driver ICs, Application Note AN-978,
http://www.irf.com/techincal-info/appnotes/an-978.pdf
[5] Zememe Walle Mekonnen, “Digital Signal Processing
Applications using C6713 DSK”, project work
[6] S. Zinn, S.L Semiatin, “Elements of Induction Heating, Design,
Control and application”
12
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-21-320.jpg)
![In this paper, a method to achieve controllability and
stability of quad-copter at certain height is achieved
such that it is stationery with respect to the earth frame
of reference at certain height. Simulation platform
used are MATLAB®
and LabVIEW®
, while detail
study of quad-copter and propeller is conducted in
ANSYS-FLUENT®
. CAD models are modelled in
CATIA®
software. The algorithm is written by
manipulating the non-linear differential equations with
control theory. The algorithm written is verified by
visualizing results, animations and virtual reality
model to completely study the quad-copter behaviour
and response to inputs.
The algorithm is translated into equivalent C language
for Hardware testing. Microcontroller and sensor used
are Arduino Mega 2560 and 6-DoF Razor IMU only.
II. EQUATION OF MOTION
The governing equations for the control of quad-copter
are derived in this section. First of all, translational and
rotational dynamics of quadrotor are explained
followed by simplifications. Bold symbols are used to
denote three-dimensional vectors, while non-boldface
symbol are used for scalars in the paper.
Figure 2: (A) Dynamic model of a quad-copter with
four propellers in Earth frame of reference (B)
Propeller i producing fi thrust with 𝜔𝑖 rpm in z-
direction
A. Dynamics
A quad-copter is a UAV having four rotors and a mass
‘m’. The forces which act on a quadrotor are its weight
and the thrust f produced by four propellers in body
fixed direction z = (0, 0, 1). Similarly, four torques acts
on each propeller and a total drag torque acts on quad-
copter body. The rotation of the body fixed frame with
respect to some inertial frame is described by the
rotation matrix R which is discussed in detail.
Two coordinate systems are considered in Figure 4
[3]:
The inertial frame (E-frame)
The body-fixed frame (B-frame)
Figure 3: Quad-copter Frame of References
These are related through three rotations:
Roll: Rotation of φ around the x-axis;
Pitch: Rotation of θ around the y-axis;
Yaw: Rotation of ψ around the z-axis.
The following assumptions were made in this
approach:
The body-fixed frame origin and center of
mass (COM) of the body of the vehicle are
coincident.
The axes of the B-frame coincide with the
body principal axes of inertia.
Figure 4: Quad-copter configuration frame
system
Given equation describes the relation of Rotation
matrix with roll, pitch, yaw and quad-copter position
in earth frame.
R (,, ) R(x,) R(y,) R (z, ) ( 1 )
The main equation governing the quad-copter
dynamics is:
14
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-23-320.jpg)
![m𝐝̈ = 𝐑𝐳 ∑ fi +4
i=1 m𝐠 ( 2 )
Where d= (d1, d2, d3) are position vectors in inertial
frame of reference and g = (0, 0, -9.81) is gravitational
constant.
As quad-copter is a rigid body with four propellers.
The body inertia is expressed as diagonal matrix
IB
= [
Ixx
B
0 0
0 Iyy
B
0
0 0 Izz
B
] ( 3 )
The propeller is a symmetric body with respect to its
axis of rotation and can be considered as disc for
simplification. The propeller inertia is given by:
IP
= [
Ixx
P
0 0
0 Iyy
P
0
0 0 Izz
P
] ( 4 )
Due to symmetry of propeller Ixx
P
= Iyy
P
.
The angular velocity of a body can be governed using
the differential equation: [2]
τres = τB + ∑ τP + ⟦ωB
× ⟧(LB +4
i=1
∑ IP
ωB
+ ∑ LP
4
i=1
4
i=1 ) ( 5 )
Where τres is the resultant torque acting on quad-
copter body. τB is the body torque
τB = (Ixx
B
ṗ, Iyy
B
q̇, Izz
B
ṙ) ( 6 )
And τP is the torque produced by propeller.
τP = (0, 0, Izz
P
ωi̇ ) ( 7 )
B. Simplification of Assumptions
The effect of all the moments acting upon the body is
denoted by τres on the right hand side of equation (5).
These include moments due to propeller forces and
torques due to motors. The propeller forces are
assumed to act through the center of each propeller. It
is assumed that the center of propeller is at a horizontal
distance l from the body center of mass.
τres = [
(f2 − f4)l + τdx
(f3 − f1)l + τdy
τ1 + τ2 + τ3 + τ4 + τdz
] ( 8 )
With τd = (τdx
, τdy
, τdz
) is the drag torque. It has
been observed that propellers reaction torque has a
linear relation with the thrust force (with
proportionality constant of kτ and sign given by the
sense of rotation).
τi = (−1)i+1
κτfi ( 9 )
And thrust is related to rotor rpm as
fi = κfωi
2
( 10 )
Solving the above equation from 3 to 8 results in [2]
Ixx
B
ṗ = κf(w2
2
− w4
2)l − (Izz
T
− Ixx
T )qr − Izz
P
q(w1 +
w2 + w3 + w4), ( 11 )
Ixx
B
q̇ = κf(w3
2
− w1
2)l + (Izz
T
− Ixx
T )pr + Izz
P
p(w1 +
w2 + w3 + w4), ( 12 )
Izz
B
ṙ = −γr + κτκf(w1
2
− w2
2
+ w3
2
− w4
2
) ( 13 )
III. CONTROLLABILITY
These non-linear equations are linearized to compute
state space matrices A, B, C and D. An LQR controller
was designed with the cost value of 0.5 s2
rad-2
on the
angular rates, 10 on the deviation from the primary
axis, 0 on the extended motor states and 0.75 N-2
on
the inputs.
IV. EXPERIMENTAL PLATFORM
Computational Programming softwares employed
were MATLAB®
/ SIMULINK®
, LabVIEW®
,
ANSYS®
, Arduino IDE®
and CATIA®
.
CAD models of Quad-copter and equivalent propeller
were modelled in CATIA®
and were exported to
ANSYS-FLUENT®
where surface meshing and
computational fluid dynamics (CFD) was done to
study aerodynamic design such that fluid was passed
on to the quad-copter and propeller at different speeds
and direction to check its serviceability.
The propeller produced vibrations which were verified
from CFD analysis. These vibrations were catered
using prop-balancer system.
15
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-24-320.jpg)
![people which will help in making tasks easier and
more efficient. Developed mathematical model was
successfully implemented on hardware.
IX. ACKNOWLEDGMENT
This work has been made possible by the help of my
co-authors which include my project advisor and co-
advisor. This text has been rectified and proof read by
Undergraduate students M. Moghees Shahid and Ali
Mahmood. They are destined for great things.
This project has been sanctioned by College of
Aeronautical Engineering, NUST.
REFERENCES
[1] T. Luukkonen, "Modelling and control of
quad-copter," Independent research project
in applied mathematics, Espoo, 2011.
[2] M. W. Mueller and R. D'Andrea, "Stability
and control of a quadrocopter despite the
complete loss of one, two, or three
propellers," in Robotics and Automation
(ICRA), 2014 IEEE International Conference
on, 2014, pp. 45-52.
[3] I. Gaponov and A. Razinkova, "Quad-copter
design and implementation as a
multidisciplinary engineering course," in
Teaching, Assessment and Learning for
Engineering (TALE), 2012 IEEE
International Conference on, 2012, pp. H2B-
16-H2B-19.
[4] D. Hanafi, M. Qetkeaw, R. Ghazali, M. Than,
W. Utomo and R. Omar, 'Simple GUI
Wireless Controller of Quad-copter',
International Journal of Communications,
Network and System Sciences, vol. 06, no.
01, pp. 52-59, 2013
[5] Y. Cooper, R. Ganesh Ram, V. Kalaichelvi
and V. Bhatia, 'Stabilization and Control of
an Autonomous Quad-copter', AMM, vol.
666, pp. 161-165, 2014.
19
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-28-320.jpg)
![Detection of Fire Hotspots dealt by Emergency
First Responders in Rawalpindi using GIS
applications
Amna Butt and Sheikh Saeed Ahmad
Department of Environmental Sciences
Fatima Jinnah Women University
Rawalpindi, Pakistan
ambutt91@yahoo.com; drsaeed@fjwu.edu.pk
Abstract— During the past, need of efficient Emergency First
Response (EFR) for Rawalpindi, along with all the major cities of
Pakistan has increased tremendously. Therefore, there is a need
to develop an effective management strategy to improve the first
response services. Present study focused on the identification of
past and current service locations for fire incidences and
mapping these locations for hotspot identification. The incidence
data for past five years (2009-13) was collected and Hotspot and
Spatial Autocorrelation analyses were performed on the data to
detect the fire hotspots and their clustering patterns in the city.
The results revealed a slight shift in fire hotspots in 5 years and
also in the clustering pattern which changed from significantly
clustered (2009) to randomly distributed (2013). Hotspot and
spatial distribution maps were generated to indicate the fire
hotspots in the city. These maps can be helpful to prevent the
future incidents by allocating more service stations focusing these
areas for fire mitigation.
Index Terms— Hotspot Analysis, Fire, Kernel Density,
Emergency First Response (EFR), Spatial Autocorrelation
Analysis
I. INTRODUCTION
Coping with fire, caused either by natural or anthropogenic
factors is one of the challenges faced by the modern societies
[1]. Analyzing the city fire risk is therefore highly significant
for development of effective urban fire protection plan and
regulations and facilitates the coordinated development of
social economy [2].
Application of geostatistical tools of GIS can play a
significant role in improvement of local fire emergency
response [3] primarily by facilitating the visualization and
interpretation of nature and previously observed patterns of
such accidents [4][5]. Generating different fire risk maps on the
basis of geostatistical analysis is also imperative to develop
strategies focused on alleviating the future risk [6].
Numerous approaches based on GIS have been developed
and used over the past to provide geostatistical surveillance of
the precedent emergency patterns for development of several
models for fast and apt response delivery [7][8][9][10][11].
A. Study Area
The study area of the present research was Rawalpindi city
(Fig. 1). The city’s administrative boundaries consist of two
tehsils namely Rawal and Potohar tehsil. Currently five service
stations and two key points of Rescue 1122 (otherwise known
as Punjab emergency service) are providing emergency
response services (including fire brigade service) in different
areas of the city. The resources currently available to them for
providing Fire brigade service include 9 fire vehicles, 14
ambulances and personnel of 400 trained rescue providers.
However, no prior study has been conducted in the city focused
on surveillance of fire emergency response. Furthermore, the
existing management strategy for improvement of fire
emergency response is not very effective and no thought has
been given to incorporating GIS expertise in the department for
this purpose.
Fig. 1. Study Area map: Rawalpindi City
20
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-29-320.jpg)
![Therefore, the primary objective of the present research is
to provide a GIS based surveillance system for the Fire
incidents of the past in order to determine the recurrent service
locations for focusing the future response. The study can
therefore be considered as a baseline for future improvement in
the quality and efficiency of fire emergency responses in
Rawalpindi city.
II. MATERIALS AND METHODS
The research methodology that was applied to obtain the
results primarily consisted of four main steps incorporating
data collection, processing, analyzing and visualizing the
results (Fig.1).
Fig. 2. Flowchart of main steps followed in methodology
B. Data Collection and Processing
The data collected for the purpose of this study was divided
into two categories namely Primary and Secondary data. The
main data obtained for the purpose of the study was secondary
data acquired from the headquarters of Rescue 1122,
Rawalpindi in form of caller and victim directory. This data
was collected on unit level i.e. data from all the emergency
units of 1122 at work in Rawalpindi city was acquired,
compiled and then processed for segregation of Fire cases.
Primary data was then collected on the basis of segregated fire
incidents. This data comprised the GPS readings of incident
locations obtained via handheld Oregon 650 GPS for the
reported fire cases.
Both primary and secondary data was processed in
Microsoft excel and then loaded to ArcGIS 10.2 for further
processing and analyzing.
C. Geostatistical Analysis of data
The data was geostatistically analyzed in ArcGIS 10.2
environment for determination of spatial clustering and
identification of hotspots. The geostatistical analysis performed
for this purpose included Global Moran’s I test (Spatial
autocorrelation analysis) and Hotspot Analysis (Getis-Ord
Gi*).
1) Global Moran’s I statistics
Global Moran’s I statistic gives an indication of any
existing correlation among spatial observations and delineates
the characteristics of the global pattern. The pattern maybe
random, dispersed or clustered depending on the spatial
association present in the data [12][13]. For the purpose of
present study spatial autocorrelation among the fire incidents
was calculated on yearly basis by employing different
threshold limits ranging from 500-2000 meters. The range used
for determination of correlation was -1 to 1 and Z-score value
was calculated to assess the statistical significance of the
observed clustering (based on correlation) for each year. The
highest correlation values were then recorded for each year and
subsequently were employed for hotspot identification.
2) Hotspot Analysis: Getis-Ord Gi*
Fire hotspots were identified based on the Getis-Ord Gi*
statistics. For this purpose, the conceptualization of spatial
relationship among different datasets was done by opting
“Zone of indifference”. The threshold limit was set on the basis
of spatial autocorrelation outcomes for each year (exhibiting
highest Z-score value). Thereafter, the identified hotspots for
Fire cases were interpolated using “Inverse Distance
Weighted” or “IDW” for better visualization of results and
hotspot maps for each year were generated.
III. RESULTS
The emergency callout data on building fires, bursting of
gas pipes, cylinder blasts, and gas leakages was cataloged in
the category of Fire emergencies. Different geostatistical
analyses were then performed to determine the pattern of
emergency cases for the study duration. The reported incidence
of FE cases for the time period of 2009-2013 were 671. Out of
which, 583 (86.9%) were males and 88 (13.1%) females. 15%
of the total fire incidents (102 cases) were reported in 2009 and
37% (247 cases), the highest incidence, in 2010. After 2010 the
incidence rate declined progressively from 24% (163 cases) in
2011 to 10% (66 cases) in 2012 and subsequently rose to 14%
(93 cases) in 2013 (Fig. 3).
Fig. 3. Percentage Contribution of Fire Incidents in Rawalpindi during 2009-
2013
The possible spatial autocorrelation of Fire cases estimated
by Moran’s I statistics revealed significant spatial clustering for
the years 2009 and 2011, whereas 2010 showed mild
clustering. 2012 and 2013 on the other hand showed random
patterns. Moran’s I and G-statistic values (Z-score) for all the
years, given in Table 1 disclosed that 2009 had highest (20.01)
while 2013 had lowest (1.10) Z-score.
21
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-30-320.jpg)
![Fig. 5. Spatial Distribution pattern of Fire incidents in Rawalpindi for study
duration
IV. DISCUSSION
In order to facilitate the efficient management of fire
emergencies in an area, improvement of existing response
systems is of high significance [14]. This can be ensured by the
providing surveillance for past occurrences and understanding
of recurring patterns.
Significant Fire hotspots were manifested in both urban and
rural areas of the city and were mainly contained in Northern
portion of the study area. These were the areas having high rise
buildings, gas stations, commercial areas, suburban areas,
highways and residential areas (with heavy load shedding of
gas). As these hotspots were estimated not only for household,
commercial and secondary fires but vehicular fires as well,
various roads were also identified as hotspots. Mostly the Fire
emergencies were observed on the roads that are used by heavy
vehicles as they are more prone to overturning and catching
fire. Corcoran et al. [11] also analyzed the spatial patterns of
fire by employing GIS and obtained similar results. Rao [15]
also reported similar findings and additionally said that the
reason for Fire incidents in the city is exposed and jumbled
cable wires made of substandard material.
However, the results of the study indicated a significant
decline in the intensity of cases for the duration of the study
(Fig. 3). This declining incidence gave an account of lack of
confidence in general public to refer to firefighting
organizations and attempting to solve the matter themselves.
However the accounted figures do not represent the total
number of cases observed in the city, just the incidence that
was dealt by Rescue 1122. Other organizations at work in the
city for fire control and management include Rawalpindi Fire
Brigade Center and Qureshi Fire Control Services.
Akhter [16] explicated the reasons behind the lack of
confidence among public. The study reasoned that there is
disparity in implementation of fire safety standards in the city
as well as very little coordination among different departments
such as traffic, police and fire fighting units. This lack of
coordination along with unavailability of Incident Command
System (ICS) often translates to poor emergency response
despite good skills and training. Dawn [17] conversely reported
that local fire brigade services lack in performance due to
insufficient professional training, availability of resources,
planning and research (both pre and post fire) and nonexistence
of any fire services act for the city. All these factors, along with
lack of awareness among general public regarding fire fighting
profession has a negative impact on fire emergency response.
Hence, for improvement of future fire emergency response,
it is need of hour to understand the previous and existing
patterns, risk factors and causative agents; and ensure effective
enforcement of building and fire safety laws [18].
V. CONCLUSION AND RECOMMENDATIONS
Present study focused on providing a geostatistical
surveillance approach for ensuring future fire safety by
improving the response quality and apt resource allocation for
high risk areas. Based on the outcomes of the research, it is
concluded that there is both spatial and temporal variation in
the occurrence of Fire incidents in the study area. Most of the
Fire hotspots however were located in the Northern portion of
the study area incorporating both urban and rural areas of the
study indicating the need to shift the focus of fire service in this
region of the study area. The study further concludes that there
is a need of incorporation of GIS based surveillance system in
the rescue department to direct the response from the service
stations in a timely manner.
Therefore, the study recommends generating awareness
among people regarding fire hazard and the factors associated
with it, incorporation of GIS expertise in emergency
departments, promotion of GPS enabled cell phones in dispatch
units and fire vehicles, high level of collaboration among
different departments working in the city for fire services
provision to avoid service duplication and publication of GIS
based maps and models designed for response improvement.
ACKNOWLEDGMENT
We are endepted to the department of Rescue 1122,
Rawalpindi for providing the data regarding fire emergencies
and cooperating with us throughout this research.
REFERENCES
[1] M. I. Channa, and K. M. Ahmed, “Emergency Response
Communications and Associated Security Challenges,” Int. J.
Net. Sec. and its Appli., vol. 2 (4), pp. 179, 2012.
23
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-32-320.jpg)
![[2] W. Aiyou, S. Shiliang, L. Runqiu, T. Deming, and T. Xiafang,
“City Fire Risk Analysis based on Coupling Fault Tree Method
and Triangle Fuzzy Theory,” Proc. Engg., vol. 84, pp. 204-212,
2014.
[3] Environmental Systems Research Institute (ESRI), “Improving
Emergency Planning and Response with Geographic
Information Systems,” Redlands, New York: ESRI. Retrieved
from http://www.esri.com/library/whitepapers/pdfs/emergency-
planning-response.pdf, 2005.
[4] S. Erdogan, I. Yilmaz, T. Baybura, and M. Gullu,
“Geographical information systems aided traffic accident
system case study: city of Afyonkarahisar,” Accid. Anal. and
Prev., vol. 40, pp. 174-181, 2008.
[5] M. Kwan, and J. Lee, “Emergency response after 9/11: the
potential of real-time 3D GIS for quick emergency response in
micro-spatial environments,” Comp., Env. and Urban Sys., vol.
29, pp. 93-113, 2005.
[6] C. Yan-yan, L. Dong, and Z. Hui, “Multi-factor Risk Analysis
in a Building Fire by Two Step Cluster,” Proc. Engg., vol. 11,
pp. 658-665, 2011.
[7] M. H. Hussain, M. P. Ward, M. Body, A. Al-Rawahi, A. A.
Wadir, S. Al-Habsi, M. Saqib, M. S. Ahmed, and M. G.
Almaawali, “Spatio-temporal pattern of sylvatic rabies in the
Sultanate of Oman, 2006–2010,” Prev. Vet. Med., vol. 110, pp.
281-289, 2013.
[8] T. Ruya, M. Ning, L. Qianqian, and L. Yijun, “The Evolution
and Application of Network Analysis Methods,” IEEE Int.
Conf. on Sys., Man, and Cyber., pp.2197-2201, 2013, DOI
10.1109/SMC.2013.376.
[9] D. Dai, “Identifying clusters and risk factors of injuries in
pedestrian–vehicle crashes in a GIS environment,” J. Trans.
Geo., vol. 24, pp. 206-214, 2012.
[10] A. Spoerri, M. Egger, and E.V. Elm, “Mortality from road
traffic accidents in Switzerland: Longitudinal and spatial
analyses,” Accid. Anal. and Prev., vol. 43, pp. 40-48, 2011.
[11] J. Corcoran, G. Higgs, C. Brunsdon, A. Ware, and P. Norman,
“The use of spatial analytical techniques to explore patterns of
fire incidence: A South Wales case study,” Comp., Env. and
Urban Sys., vol. 31, pp. 623-647, 2007.
[12] B.N. Boots, and A. Getis, Point Pattern Analysis Newbury
Park. Newbury Park, CA, USA: Sage Publications, 1998.
[13] L. Fang, L. Yan, S. Liang, S. J. D. Vlas, D. Feng, X. Han, W.
Zhao, B. Xu, L. Bian, H. Yang, P. Gong, J. H. Richardus, and
W. Cao, “Spatial analysis of hemorrhagic fever with renal
syndrome in China,” BMC Infect. Dis., vol. 6, pp. 77-88, 2006.
[14] S.R. Morgan, A. M. Chang, M. Alqatari, and J. M. Pines,
“Non–Emergency Department Interventions to Reduce ED
Utilization: A Systematic Review,” Acad. Emer. Med., vol. 20,
pp. 969-985, 2013.
[15] S. Rao, “Rescue 1122 management plan finalized,” The
Nation. Retrieved from http://nation.com.pk/karachi/06-Jan-
2010/Rescue-1122-management-plan-finalised, 2010.
[16] S. Akhter, “Firefighters’ view on Improving Fire Emergency
Response: A Case Study of Rawalpindi,” Int. J. Hum. and Soc.
Sci., vol. 4(1), pp. 143-149, 2014.
[17] Dawn, “Pindi fire brigade squad runs out of steam,” Dawn.
Retrieved from http://www.dawn.com/news/90148/rawalpindi-
pindi-fire-brigade-squad-runs-out-of-steam, 2003.
[18] Pakistan Observer, “1122 Rawalpindi Rescued 1883 Emergency
Victims,” Pakistan Observer. Retrieved from
http://pakobserver.net/detailnews.asp?id=251475, 2014.
24
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-33-320.jpg)
![Prospects of Airborne Wind Energy Systems
in Pakistan
Z. H. Khan
Dept. of Electrical Engineering
Riphah International University
Islamabad, Pakistan
zeeshan.hameed@riphah.edu.pk
Sohaib Khan
Dept. of Aerospace Engineering
Institute of Space Technology
Islamabad, Pakistan
sohaibkham786@gmail.com
Arsalan Khan
Dept. of Control and Simulation
CESAT
Islamabad, Pakistan
arsalan1@mail.nwpu.edu.cn
Abstract— High altitude wind energy is considered as a high
efficiency, low cost solution to sustainable energy solutions
worldwide today due to highly flexible and adaptive designs of
powered kites. Pakistan is facing energy crisis since years due to
inefficient electrical transmission network, increased demand of
electricity, lack of resource planning and increased cost for
furnace oil which can be used to generate electricity. As a clean
energy source, conventional wind energy is a preferable choice
but it has few disadvantages due to large initial investments
involved and dangers to environment due to rotating machinery
which can result in bird hit and on the other hand, generating
acoustic noise which affect populations of people living nearby.
In response to these issues, high altitude energy is found to be
free from many such problems as found with conventional wind
energy systems. It is found suitable specially for addressing
essential energy needs of off-grid consumers for water pumping,
drilling, refrigeration of vaccines and life-saving medicines and
powering up far-off residential sites e.g. communities living off-
grid at Alaska’s northern region having minimal solar energy
during the long winter season when energy demand is greatest
for heating purposes. In this paper, we propose a UAV design
which can possibly be used as an airborne wind energy system
for electricity generation.
Keywords—Altitude wind; renewable energy; powered kites;
wind energy conversion system; distributed energy
I. INTRODUCTION
Energy production for a world’s economy is directly linked
with GDP growth. Renewable energy solutions are a preferable
choice to address the energy demands world-wide due to
environment friendly green energy technology. It has been
estimated that till 2050, 40% of the world energy requirements
will be meet by renewable energy including solar (photo
voltaic (PV) and thermal), wind, bio-mass, bio-fuel etc. [1].
Among various technologies, wind energy is preferred due to
continuous production of electricity (as long as wind is present
as prime mover) as well as high production rate with lesser
requirement of installation area as compared to solar energy
[2]. However, a detailed analysis of wind availability and on-
site surveying is mandatory for an optimal production of
energy throughout the year. The conventional wind energy
conversion systems (WECS) are popular only in those areas
where sufficient wind is available i.e. in the range of 5-8 m/s2
.
This limitation greatly affects the utilization of WECS as an
effective resource of energy as usually far off areas near sea-
shores or mountains fulfills the required energy density criteria
for feasible investments. Also, long way cabling for
transmission of this energy to grid also results in additional
cost and line losses [3]. Social activists and NGOs raise voices
and show their grievances against environmental impact of
these wind farms due to ever increasing bird accidents as of
result of collision with fast moving blade tips.
Fig 1. Airborne energy production connected to grid
This paper describes some key design features of an
airborne wind energy system (AWES) e.g. a large kite which is
flown in air while maintaining its connection with a generator
kept at ground level by a suitable rotating mechanism such as
cable drum which is linked via tether. The aerodynamic forces
acting on the kite causes traction in the cable drum which is in
turn converted into electrical power by the generator as shown
in Fig 1. The obvious advantages of shifting the heavy
mechanical components on ground result in simple design and
maximum power optimization. Due to the fact that the flying
kite operates in periodic cycles alternating between ―traction
phase‖ and ―reel-in phase‖ of the tether, the electrical power
produced is not continuous rather intermittent which is not
25
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-34-320.jpg)
![suitable for on-grid connection where continuous power is
required. However, a continuous power generation system can
be designed either by using multiple, individually controlled
kites or a battery system for buffering the power generation
across the cycles. This solution is guaranteed to generate
higher power with lower cost as compared to conventional
wind energy conversion system (WECS) due to availability of
faster wind speeds at high altitude.
Fig 2. KGS for generating electricity on-board ship
II. HIGH ALTITUDE WIND ENERGY POTENTIAL
Wind energy is a renewable resource which is not only
cheap but also readily available in most parts of the world.
Wind is discontinuous source of energy and the conventional
WECS provides uneven power supply when connected to the
grid. On the other hand, if altitude is raised, a lot more energy
is available as compared to that blowing at 50-200 m above
ground. It has been estimated that the magnitude of wind
energy above 1000 m is twice as much as found at lower
altitudes. In addition, at an altitude of 800 m above ground,
wind power density is sufficiently available to be used for
altitude energy generation all over the globe [4]. As a rule of
thumb, it has been found that a five-fold cost saving with twice
capacity increase can be achieved in shifting technology from
WECS to AWES [5].
III. ADVANTAGES OF AWES OVER WECS
A. Impact on Environment
The airborne wind energy systems have fewer effects on
environment as compared to WECS. The fast moving blades of
classical wind mills injures many birds flying in close
proximity. On the other hand, AWES has no dangerous edges
which can kill birds [6].
B. Cost comparison
Using AWES instead of WECS, an advantage of more than
90% material savings is guaranteed. This is because large
structures require more and more material to with stand heavy
loads due to wind and the rotating machinery [7]. Many tones
of weight loaded on large erected structures comprises of rotor
blades connected to a hub which drives the generator.
Moreover, the maintenance of WECS is also an issue which
requires access to the system, long down times in case of fault
and danger to the life of technician all adding up the cost
compared to AWES where generator and accessories are
installed on ground [8].
C. Mobility of the power station
In comparison with WECS, the airborne energy converter
can be moved anywhere. Example include rural areas, Coastal
belt, Desert, etc. However, if flying at higher altitudes, caution
must be paid to operate AWES in no-fly zones (NFZ) to avoid
any danger for civil aviation.
D. Efficiency
The advantage of increasing altitude as in the case of
AWES is to get more capacity due to persistent wind as
compared to turbulence due to buildings and infrastructure at
WECS heights [9].
Fig 3. Buoyant Airborne Turbine (BAT) for stand-alone energy generation in
Alaska [10]
IV. CLASSIFICATION OF AWES
The airborne wind energy systems are classified based on
their designs, grid connectivity and aerodynamics e.g. heavier
than or lighter than air configuration. Some types are indicated
as follows:
1) Flying tethered airplane and kite
In this type of AWES, a tethered kite or airplane flies in
circular or Lissajous shaped orbit to produce electricity via
traction force applied on a generator on ground or ship as
shown in Fig 2. The kite generator system (KGS) is one of the
most commonly used system due to its simplest design and
easy control system [11].
2) Multiple wing system
This type of altitude wind system has multiple wings to
generate electrical energy. A laddermill is an example of such
system [12]. Multiple wings allow scalability of the concept in
order to generate more energy.
3) Lighter than air systems
These systems are lighter than air and have an on-board
generator to produce electricity as shown in Fig 3. Such
systems are supported by a lighter than air filled balloon to
reach at 100m or above heights where enough wind is
available to drive the generator [13]. Such systems are more
suitable for areas of the world where sunlight is not available
throughout the year.
26
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-35-320.jpg)
![4) Rotor hover craft systems
These systems hover in air via rotating propellers.
Quadrotor type hovercrafts are popular which have the ease of
vertical take-off and landing operations [14]. Due to on-board
generators, the airborne system is heavier as compared to the
crosswind kite generator in which has the heavy machinery on
ground.
V. FLYING WING GENERATOR SYSTEM DESIGN
In order to visualize a practical system, we can use a flying
wing generator (FWG) for our case study. Our goal is to
evaluate optimal design parameters which can be used to
generate maximum wind at different altitudes. By using a
strong flexible tether, energy kites can reach higher altitudes of
100 to 400 meters). This can save up to 90% of the materials of
WECS used conventionally, resulting in reduced per unit
energy cost. As, these systems are more aerodynamic and can
access higher energy density due to stronger winds, each EKS
can generate 50% more energy than their conventional
counterparts [15].
A. Aerodynamics
The wind energy that can be converted to electricity is
given as:
2
3
27
2
D
L
Lw
C
C
CAvP (1)
The relation provides some significant information about
some key design parameters. Equation 1 states that in order to
increase the power, the key parameters to increase include the
wind speed, gliding ration (L/D), the lift coefficient and the
wing surface. As a comparison, energy kites have lower L/D,
price and weight as compared to flying wing. In theory, about
60 kW can be produced per sq. meter of flying kite generator
system. The CFD model is shown in Fig 4.
(a) Side View (b) Top View (c) Front View
Fig 4. FlyPG simulation for aerodynamic analysis
In order to analyze the aerodynamic performance of our
benchmark system FlyPG, different flight conditions are used
for Aerodata generation. The aircraft wing is taken as NACA-
W-4-4410 which has a span of 4 meters, Sref equal to 1.5 m2
,
Cbar = 0.38 m, Root chord = 0.35 m, Tip chord = 0.15 m. For
the vertical tail design, NACA-V-4-0010 is selected with Root
chord = 0.25m and Tip Chord = 0.1 m. The horizontal tail
design is NACA-H-4-0010 based with root chord = 0.25 m
and Tip chord = 0.1 m.
The flight test condition is selected as Mach 0.05 at an
altitude of 200 meters above ground level. Then the angle of
attack (α) in degrees is varied from 0 to 5 degrees, while Xcg
of 0.9 meter is assumed. The aero-data is plotted in Fig 5 as a
function of angle of attack. It is important to take into account
at least 10% drag increase due to tether connecting the kite
with the ground generator.
Fig 5. Aerodata plots for FlyPG @ Mach = 0.05
B. Control and Autopilot
Modern design of FWGs incorporates automatic take
off/landing and flight for robust operation under varying
environmental conditions [16]. A coordinated control
mechanism is also devised by few manufacturers where
communication between the Kite/airplane controller and main
controller at the ground station occurs to track the given
trajectory [17].
Generally, a multi-objective loop controls the dynamics of
the system in flight. First of all, an optimal track point on
circular or Lissajous trajectory needs to be calculated and
tracked during flight [18]. In most cases a navigation loop
forms the outer one which controls the bearing of the flying
system, while an inner attitude controller controls the roll, pitch
and yaw angle as well as respective rates in order to achieve
the required bearing.
Fig 6. A generalized 2-loop control structure for Flying wing/ Energy Kite
As shown in Fig 6, an outer loop controls the bearing
(ρcom) of the flying wing. The error between the commanded
and measured bearing drives the attitude controller in order to
generate roll, pitch and yaw commands for the flying wing
until it reaches the desired navigational coordinates. In this
27
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-36-320.jpg)
![control loop, however, the cross track controller is not
included which is necessary to keep the flying wing on the
desired trajectory by removing the off-track distance [19]. In
practice, stable and robust control is required for fast tracking.
C. Mechanical Subsystem
The mechanical part of a FWG includes wing design after
aero-foil selection. The control actuators (motors) also need
positioning and the embedded controller is designed w.r.t
nominal torque load as well as keeping its maximum limit as
per maximum torque demand. The selection of tether and its
linkage with the ground generator is also important [20].
However, the drag added by the tether length must be added in
the aerodynamic drag analysis for the autonomous controller
design [21].
D. Electrical Subsystem
The power generation system usually comprises of a
permanent magnet synchronous machine (PMSM) which acts
as a generator [22]. In general, the AWECS can function either
as stand-alone (off-grid) or on-grid. However, an obvious
problem for grid based operation requires an additional cost of
storage system or operating multiple systems at the same time
so that the pumping cycle of one kite is out of phase as
compared to the pumping cycle of the other kites [23]. Thus, at
the instant one kite is relaxing; the other will be supplying
energy to the grid. However, this strategy complicates the
design of the AWECS.
E. Communication Subsystem
The communication system is required in order to control
the system and for data acquisition. Wireless networks are a
preferred choice in order to reduce weight and cost of the
airborne system [24, 25]. A suitable network protocol is chosen
as required depending upon the range of communication
required. Ground based sensors and controllers send necessary
data to the air segment. Security features of the communication
protocol must be operational for robust and uninterrupted
control of the flying system [26].
2000 2003 2006 2009 2012 2015 2018
0
50
100
150
200
250
300
350
400
450
500
ElectricalEnergy(MW)
Time
Fig 7. WECS for energy generation and forecast in Pakistan
VI. ENERGY NEEDS IN PAKISTAN
Pakistan is the 6th
largest country of the world with a total
estimated population of 175 million people. However, 24%
population lives below the poverty line in the country. As per
the economic survey, it is among those countries of the world
which have lowest per capita energy consumption. The
installed electricity generation capacity is around 22,500 MW.
While the current available power is 17,000 MW, the peak
demand raises up to 22,000 MW in summer resulting in a
shortfall of 5000 MW.
The economic advisory council of Ministry of Finance
(Govt of Pakistan) foresees an increase in energy demand up to
122.46 M.MTOE in 2022 [27]. In order to satisfy this demand,
at least 12% of this energy is planned to be obtained from
renewable resources in order to relax the huge demand of
furnace oil exports from 30% at present to 20% in 2022 [28].
Table 1 describes the energy demand and forecast which shows
that a total of 122.46 M.MTOE demand is expected in 2022
out of which 14.70 M.MTOE will be obtained from renewable
energy. Looking at the rise in using wind energy as the
renewable source in Pakistan as shown in Fig. 7, it is
emphasized that Altitude wind projects should also be
commenced in the country as a future energy resource
especially for people living in off-grid areas of Pakistan.
In order to enhance sustainable energy trend, Government
has exempted renewable energy equipment from income
tax/withholding tax and sales tax [29]. Also, import duties have
been waived off for import of such machinery and tools to be
used to develop sustainable energy solutions. Thus, already a
favorable environment is available for foreign investors to
come and finance green energy solutions. Fig. 7 shows a
potential rise of WECS usage as a renewable energy source.
However, the feasibility is only limited to areas where
sufficient wind is present. These areas include the coastal belt
near Karachi (Sindh province), Kati-Bander, Makran and
Jhampir in Balochistan [30, 31].
Figure 8 Map of Pakistan's national grid [32]
As noted from the title of the paper, our target is to address
the energy demands of remote areas of the country far-off from
28
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-37-320.jpg)
![the national grid. Fig. 8 describes the map of Pakistan’s
national grid which shows large areas of Balochistan not
covered by it. This is because of the low population density in
this province as well as poverty due to which it is not feasible
to extend the grid in these areas. Villages in these areas can be
powered up by renewable energy resources by spending less
money and requiring only maintenance cost in the long run
[33].
TABLE I. PAKISTAN ENERGY DEMAND FORECAST [27]
Resource
Energy Demand (M.MTOE)
1992 2008 2022
Oil 11.53 19.18 24.49
Gas Indigenous 11.68 29.88 25.72
Gas Import - - 8.57
Hydel 4.49 6.86 24.49
Coal 2.10 5.79 18.37
LPG 0.15 0.44 2.45
Nuclear - 0.75 3.67
Renewables - - 14.70
Total 29.94 62.90 122.46
Most common uses there require lightening solution, water
pumping for drinkable water supply and also to operate tube-
wells for agriculture use etc. These applications require
lightweight, low-cost, minimum energy storage solutions.
These demands can be fulfilled by using AWES whereby the
intermittent energy produced during reel out phase could be
used directly.
VII. CONCLUSION
In this paper, an overview of altitude wind technology is
discussed and a preliminary analysis is done for the suitability
to address the needs of renewable energy in developing
countries with budget constraints. A flying wing generator
design is also presented with some initial CFD simulation
results. As the model of FlyPG is now generated, the next step
is to design a prototype system and to demonstrate the energy
output by flying it. Moreover, as the energy needs in the
country are growing, it is high time to invest in such aerospace
technologies to obtain optimal designs resulting in maximum
electrical output with light weight flying systems.
REFERENCES
[1] M. A. Sheikh, "Energy and renewable energy scenario of Pakistan,"
Renewable and Sustainable Energy Reviews, vol. 14, pp. 354-363, 2010.
[2] J. I. Lewis and R. H. Wiser, "Fostering a renewable energy technology
industry: An international comparison of wind industry policy support
mechanisms," Energy Policy, vol. 35, pp. 1844-1857, 2007.
[3] G. Kingsley, "Wind energy production using kites and ground mounted
power generators," ed: Google Patents, 2003.
[4] M. lppolito, "Altitude Wind Energy Generation (Information Dossier),"
KiteGen Research, Italy2007.
[5] L. Goldstein, "Theoretical analysis of an airborne wind energy
conversion system with a ground generator and fast motion transfer,"
Energy, vol. 55, pp. 987-995, 2013.
[6] M. I. Blanco, "The economics of wind energy," Renewable and
Sustainable Energy Reviews, vol. 13, pp. 1372-1382, 2009.
[7] B. W. Roberts, "Cost and Security of Electricity Generated by High
Altitude Winds," Altitude Energy2012.
[8] M. Brooks, "To make the most of wind power, go fly a kite," New
Scientist, May, vol. 5, 2008.
[9] R. van der Vlugt, J. Peschel, and R. Schmehl, "Design and Experimental
Characterization of a Pumping Kite Power System," in Airborne Wind
Energy, ed: Springer, 2014, pp. 403-425.
[10] C. Vermillion, B. Glass, and A. Rein, "Lighter-Than-Air Wind Energy
Systems," in Airborne Wind Energy, ed: Springer, 2013, pp. 501-514.
[11] M. Ippolito, "Kite Wind Generator," ed, 2006.
[12] B. Lansdorp and P. Williams, "The laddermill-innovative wind energy
from high altitudes in holland and australia," in Wind Power, 2006.
[13] C. Vermillion, T. Grunnagle, and I. Kolmanovsky, "Modeling and
control design for a prototype lighter-than-air wind energy system," in
American Control Conference (ACC), 2012, 2012, pp. 5813-5818.
[14] G. Mortimer. (2014, 29th October). Multicopter based wind turbines.
Available: http://www.suasnews.com/2010/10/2464/multicopter-based-
wind-turbines/
[15] MAKANI. (2013, 20th November). Advanced Wind Energy "ENERGY
KITES". Available: http://www.google.com/makani/
[16] M. Canale, L. Fagiano, and M. Milanese, "High altitude wind energy
generation using controlled power kites," Control Systems Technology,
IEEE Transactions on, vol. 18, pp. 279-293, 2010.
[17] M. Canale, L. Fagiano, M. Ippolito, and M. Milanese, "Control of
tethered airfoils for a new class of wind energy generator," in Decision
and Control, 2006 45th IEEE Conference on, 2006, pp. 4020-4026.
[18] R. Lozano Jr, M. Alamir, J. Dumon, and A. Hably, "Control of a wind
power system based on a tethered wing," Proceedings of second
EGNCA 2012, 2012.
[19] C. Vermillion, B. Glass, and B. Szalai, "Development and Full-Scale
Experimental Validation of a Rapid Prototyping Environment for Plant
and Control Design of Airborne Wind Energy Systems," in ASME 2014
Dynamic Systems and Control Conference, 2014, pp. V002T18A001-
V002T18A001.
[20] R. Marissen, "Dyneema high strength, high modulus polyethylene
fiber," vol. CIS YA100, Dyneema, Ed., ed: Royal DSM N.V., 2008, p.
4.
[21] L. Fagiano, "Control of tethered airfoils for high–altitude wind energy
generation," Politecnico di Torino, 2009.
[22] M. Ahmed, A. Hably, and S. Bacha, "Grid-connected kite generator
system: Electrical variables control with mppt," in IECON 2011-37th
Annual Conference on IEEE Industrial Electronics Society, 2011, pp.
3152-3157.
[23] P. Williams, B. Lansdorp, and W. Ockesl, "Optimal crosswind towing
and power generation with tethered kites," Journal of guidance, control,
and dynamics, vol. 31, pp. 81-93, 2008.
[24] Z. H. Khan, D. G. Catalot, and J.-M. Thiriet, "Hierarchical wireless
network architecture for distributed applications," in Wireless and
Mobile Communications, 2009. ICWMC'09. Fifth International
Conference on, 2009, pp. 70-75.
[25] Z. H. Khan, J. M. Thiriet, and D. Genon-Catalot, "Wireless network
architecture for diagnosis and monitoring applications," in Consumer
Communications and Networking Conference, 2009. CCNC 2009. 6th
IEEE, 2009, pp. 1-2.
[26] J. Wang, M. Shahidehpour, and Z. Li, "Security-constrained unit
commitment with volatile wind power generation," Power Systems,
IEEE Transactions on, vol. 23, pp. 1319-1327, 2008.
[27] E. A. Council, "Integrated Energy Plan (2009-2022), Report of the
Energy Expert Group " Ministry of Finance2009.
[28] M. A. Khan and U. Ahmad, "Energy demand in Pakistan: a disaggregate
analysis," The Pakistan Development Review, pp. 437-455, 2008.
29
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-38-320.jpg)
![[29] A. W. Bhutto and S. Karim, "Energy-poverty alleviation in Pakistan
through use of indigenous energy resources," Energy for Sustainable
Development, vol. 11, pp. 58-67, 2007.
[30] M. Hussain, S. Abbas, M. Ansari, A. Zaffar, and B. Jan, "Wind Speed
Analysis of Some Coastal Areas near Karachi," Pakistan Academy of
Sciences, 2012.
[31] I. Ullah, Q.-u.-Z. Chaudhry, and A. J. Chipperfield, "An evaluation of
wind energy potential at Kati Bandar, Pakistan," Renewable and
Sustainable Energy Reviews, vol. 14, pp. 856-861, 2010.
[32] GENI. (2012). Global Energy Network Institute. Available:
http://www.geni.org/globalenergy/library/national_energy_grid/pakistan
/pakistaninationalelectricitygrid.shtml
[33] M. Amer and T. U. Daim, "Selection of renewable energy technologies
for a developing county: a case of Pakistan," Energy for Sustainable
Development, vol. 15, pp. 420-435, 2011.
30
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-39-320.jpg)
![Figure 10. Pond Capacity
IV. CONCLUSIONS
Proposed water reservoir in rain-fed area would have
a significant favourable effect on intended
beneficiaries. In addition to direct benefits from
changes in agricultural landscape there will be
indirect benefits for the relevant households from
appreciation of their land values, creation of
employment opportunities / reduction in
underemployment and incidence of poverty,
groundwater recharge and improvement in drinking
water supply.
However to ensure realization of anticipated benefits
of this dam, it will be essential to undertake capacity
building of target beneficiaries in terms of financial
resources and knowledge / technology acquisition
regarding high value agricultural commodities /
enterprises.
In fact, the dam site will improve overall water
availability, soil properties, land use and consequent
socio economic status of the inhabitants and thus
enhancing regional development.
Minor possible health considerations will be
countered through inhabitant sensitization and urging
concerned authorities ensuring better healthcare and
regional safeguard measures. The project is thus
highly recommended as Sustainable Development
and could also be replicated to other parts of the
country.
REFERENCES
[1] Minár, J., et al. "Geomorphological information system: idea
and options for practical implementation." Geograficky
Casopis Slovenskej Akademie, Vied57(3), pp. 247, 2005.
[2] Rabus, B., et al. "The shuttle radar topography mission—a
new class of digital elevation models acquired by
spaceborne radar." ISPRS J. Photogrammetry and Remote
Sensing, vol. 57(4), pp. 241-262, 2003.
[3] Reuter, H. I., et al. "An evaluation of void‐filling
interpolation methods for SRTM data." Intl. J. Geo. Info.
Sci., vol. 21(9), pp. 983-1008, 2007.
[4]Akhtar, M. R, "Impact of resource conservation technologies for
sustainability of irrigated agriculture in Punjab-Pakistan."
Journal of Agricultural Research (Pakistan), 2006.
[5] Jasrotia, A., et al. "Water balance approach for rainwater
harvesting using remote sensing and GIS techniques,
Jammu Himalaya, India." Water resources magt., vol23(14),
pp. 3035-3055, 2009.
[6] Khan, A. A. and S. B. Ali, "Effects of erosion on Indus
River bio-diversity in Pakistan." Pak. J. Biol. Sci, vol. 6(12),
pp. 1035-1040, 2003.
[7] Minár, J., et al. "Geomorphological information system: idea
and options for practical implementation." Geograficky
Casopis Slovenskej Akademie Vied, 57(3), pp. 247, 2005.
[8] Munyao, J. N., et al "Use of Satellite Products to Assess
Water Harvesting Potential in Remote Areas of Africa."
Land Degrad. Develop, vol. 22, pp. 359-372, 2010.
[9] Rabus, B., et al. "The shuttle radar topography mission—a
new class of digital elevation models acquired by
spaceborne radar." ISPRS J. Photogrammetry and Remote
Sensing, vol. 57(4), pp.241-262, 2003.
[10] Reuter, H. I., et al. "An evaluation of void‐filling
interpolation methods for SRTM data." Intl. J. Geo. Info.
Sci,21(9), pp. 983-1008, 2007.
[11] Shah, H., et al. "Potential For Investment In Indigenous
Technologies: A Case Of Low Cost Soil And Water
Conservation Structures In Rainfed Pothwar, Pakistan."
Pakistan J. Agric. Res, vol. 25(2), 2012.
[12] Vohland, K. and B. Barry, "A review of in situ rainwater
harvesting (RWH) practices modifying landscape functions
in African drylands,"inAgriculture, Ecosystems &
Environment, vol 131(3), pp. 119-127, 2009.
1120
1140
1160
1180
1200
1220
1240
0 5000 10000 15000
Elevationinft
Pond Capacity in Acre-Ft
TAMMAN DAM SITE
Pond Capacity Curve
37
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-46-320.jpg)
![Urban change detection of Lahore (Pakistan) using the Thematic Mapper
Images of Landsat since 1992-2010
Muhammad Usman Saleem†
Institute of Geology, University of the Punjab.
P.O. Box, 54590 Lahore Pakistan.
Email: osman.geomatics@gmail.com
Abstract: Using the techniques of remote sensing and
geographic information system, this study is an effort to
detect the urbanization in the Lahore (Pakistan) since
1992-2010. Two Landsat TM images from 1992 to 2010
have been used to investigate the urbanization in Lahore.
Using unsupervised classification, we have maked classes
of land cover in each image. Using Erdas imagine 9.2
urban change detection over Lahore has been
investigated. The results of change detection indicate that
urbanization in Shahdera, Kala Shah Kako, Kamoki,
Sheikhopura, Bharia Town and D.H.A regions of Lahore
has increased over 20% since 1992-2010. Band
differencing, Red-Green bands differencing techniques
verify the results of change detection and confirmed these
pockets as intensely urbanized areas since 1992. Maps in
Arc GIS 9.3 show these changes in the images. To
overcome this urbanization this study also guide us about
the Karol,Mohalanwal, Kot Radha Kishan , Kasur etc are
the regions where the new residents can be make.
Keywords: Urban change detection, remote sensing, land
use, Landsat, Urbanization.
I. Introduction
Urbanization is the conversion of rural areas to urban
settlements in the form of expansion of these settlements.
There are a lot of factors which impact the urbanizations.
The major are industrializations, health, education, and
population increase. Lahore is a historical urban city. It
was located under the old wall city but now it has been
expand tremendously. According to 1998 census eighty
two percent (82%) of the total population of the Lahore
district is urban and eighteen percent (18%) is rural [1].
Spatial variations in urbanization are the variations in the
soil type, habitat, weather pattern and the temporal
variations are the variations which are occurred with time.
†Corresponding Author
Contact: +923217579429. Author postal address:
Institute of Geology, University of the Punjab, Lahore
Pakistan
A lot of analytical and statistical models used which was
on based of the urban geometry, spatial relation, social
and economic parameters of a city [2]. In literature there
was three models of urban land use was used (i)
Concentric Zone Model (ii) Sector Model (iii) Multiple
Nuclei.[3] But the Geographic Information System (GIS)
has totally changed the methods to study the urbanization.
Remote sensing with integration of Geographic
information system is now powerful tool to investigate
urban change detection. Traditional method of in field
surveying and mapping required a lot of time span. That
why GIS and remote sensing due to their technical
soundness, are being mostly used for urban change
detection [4]. Integrated study of remote sensing and
geographic information system were used to analyze the
urbanization process of Lahore city using Landsat
Thematic mapper images of 1992-2010. Spatial
information was obtained from the remotely sensed
images of Landsat satellite. While the results shown in the
form of maps are part of geographic information system.
The urban development of Lahore city has been very
rapid during the last decade and this has dramatically
increased the impact on ecosystems. It is therefore
necessary to monitor this expansion. This would be very
helpful for regional development and decision-making
processes. The goal of this study was to detect urban
change in Lahore city using different remote sensing
techniques e,g; Unsupervised classification , urban
change detection, band differencing and red-green
differencing and these techniques reveal that Lahore city
is facing intensely urbanization from 1992 to 2010. The
output is in the form of binary image (called map) which
represents the locations where urban change has occurred.
II. Aim and Objectives
Aim:
To investigate urban change detection of Lahore and
regions around it by using Landsat TM images.
38
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-47-320.jpg)
![Objectives:
(a) To get quality control data of Landsat Thematic
Mapper of year 1990 and 2010.
(b) To investigate and analyze the urban change
detection of Lahore by making the unsupervised
classification.
(c) Validate the areas of urbanization (where
urbanization increase or decrease) by using band
differencing, red green differencing techniques.
(d) Map these areas of urbanization.
III. Study Area
Lahore (310
34’ N, 740
22’ E) is the capital city of
Punjab province of Pakistan. According to census
1998 it is the second most populated city of Pakistan.
It is located near to the border with India in east,
Sheikhupura on the north-west, Kasur district on the
south. Dated back to a thousand of year, it is an urban
city(see figure 1). This city has extreme climate with
four seasons in a year. Summer season begins in
April and lasted till September. May and June are the
hottest months in the summer where the mean
monthly temperature for maximum and minimum are
45.40
C and 29.30
C respectively [4]. Winter season
start from November till March. December is the
coolest month in this season and also coolest month
of the year. The mean monthly maximum and
minimum temperatures for this month are 21.1 0
C
and 7.20
C respectively [5]. The monsoon period starts
from July and lasted till September. Temperature
falls to 480
C in May and June. The average
maximum and minimum temperatures in Lahore are
30.80
C and 17.80
respectively. Being a major develop
city, urbanization is going to rise rapidly. It is the
consequences of the coming of people from the
village to city due to more chance of job and better
infrastructure. Lahore‘s urbanization going to
increase since 1973 which is leading to rise in
temperature [5]. Increase in the number of industries
give rise to major source of rise in temperature. Total
area of the Lahore is 1772 km2
[3]. Total population
was 6.319million in 1998 with population density of
3,566 people per km2
[3].
Figure 1: Location of Lahore in the Map of Pakistan.
IV. Methodology
In this study, two Landsat Thematic Mapper images
of October 1992 and October 2010 were acquired
from the USGS website. These images are already
geometrically rectified. The row number for these
images was 149. Thematic Mapper Images mostly
used for the environmental assessment and
monitoring [6]. USGS website provides TM data in
the separate bands. We have performed Layer
stacking on these images. Lahore as the study area
for research, we have cut subsets of both images.
Both subsets images are not of the same size.
October 2010 image’s subset is more in extending
than October 1992 image (See fig.2 & 3). The band
combination 4, 3, 2 was selected to detect
urbanization [6]. All these processing of data is
performed in Erdas Imagine 9.2 software. To show
results we have made maps in power software of Arc
GIS 9.3
Lahore
39
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-48-320.jpg)
![Figure 8: Band differencing of TM of Lahore (1992-
2010).
Light gray areas in the figure 8 indicate the areas where
urbanization has increased [3].This technique utilized the
differences in brightness value of one pixel to
corresponding pixel of another image. Neutral gray tune
in figure 8 black color represent zeros difference while
white has the maximum difference between the images.
VIII. Red-Green Differencing
This technique is widely used to find the change in the
areas of two images of different times. To detect the
urban change, open the most recent image (2010) in the
red band and old image (1992) in the green band. In the
Erdas Imagine 9.2, we swipe these two images.
Combination of red and green produce yellow color. But
the areas which have changed show reddish color in
image (see figure 9) are the regions where urbanization
has occurred. In remote sensing we cannot use hard
statements about any results. Through this technique we
also verify the areas where the urbanization has occurred
from 1992-2010.
Figure 9: Red-Green Differencing of Landsat Images
1992-2010.
IX. Results and Discussions
This study focused on the important of remote sensing
and GIS techniques in the analyzing of urban change
detection of Lahore and its neighbor cities. This study
shows that digital image processing and GIS can be used
for producing urbanized maps of Lahore. As in remote
sensing we have to rely on the number of remote sensing
techniques which leading us to one conclusion. We have
used the TM data with the interval of 18 years to
investigate the urbanizations. Change detection tell us that
during the study period urban sprawl of Lahore has been
shifted towards Shahdera, Kala Shah Kako , Sheikupura
and Kamoki where the urbanization has been increased by
20% . For validation of these results we have use the band
differencing and red-green differencing techniques which
shows that old city of Lahore is going to expand in the
urban areas and this pattern of urbanization is shifted
towards Bharia Town, Johar Town, and Chung. In these
areas people are going to make urbanization over than
20% from 1992-2010. According to [3] Lahore was
58977 hectares in 1972 and which now (2009) has
increased to 99173 hectares. Our results matching with
the results of previous study of Lahore’s urbanization of >
20% from 1990 to 2009 has occurred. According to 1998
Census, the average family size of the Lahore city was 7.1
Urbanization
42
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-51-320.jpg)
![persons, but during 1998 to 2007, the urbanization rate
has increased against population [5].
X. Conclusions
This effort proved its usefulness in the present scenario of
rising trends of population of Lahore city. With the use of
remote sensing and GIS techniques, we have investigated
new pockets around the Lahore where the citizen makes
more residents and plots. Major areas which are facing
urbanization since 1990 are Kala Shah Kako,
Sheikhopura, Meridke, Bheria Town, Kamoki, Chungm
Raiwind, Shahdera and Johar Towns where urbanizations
from 18 years raised 0-20 % even more than 20%. This
study also guides us Karol,Mohalanwal , Kot Radha
Kishan ,Kasur are the regions where the Government can
plan new resident plots. We suggest that future study on
different land use/ land cover patterns of Lahore. Future
study should be undertaken the loss in the area under
decrease in natural vegetation and agriculture. This study
strongly recommend for Government and urban scholars
that such studies should be revised after nominal time
interval with high resolution satellite imagery like Spot,
QuickBird etc.
ACKNOWLEDGEMENT
Author would like to say thanks of USGS team who
provide him satellite images for research. Also I would
like to say thanks deeply to Dr. Najam Abbas, Secretary
ICASE 2015,Institute of Space Technology, Islamabad
Pakistan for providing his kind contribution, for
publication of this research work.
XI. REFERENCES
[1] District Census Reports for each District
containing General Description of the
District and Broad Analysis of Population
and Housing Data followed by detailed
statistical tables (1998-Census).
[2] Young,X. & Lo, C. P. 'Modelling urban
growth and landscape changes in the Atlanta
metropolitan area', international journal of
geographical science, vol. 17, no. 5, pp.
463,463-488, 2003.
[3] O. Riaz, 'URBAN CHANGE DETECTION
OF LAHORE (PAKISTAN) USING A
TIME SERIES OF SATELLITE IMAGES
SINCE 1972', ASIAN JOURNAL OF
NATURAL & APPLIED SCIENCES., vol. 2,
no. 4, pp. 101,102,103,104, 2013.
[4] Jensen and D. Cowen, 'Remote Sensing of
Urban Suburban Infrastructure and Socio-
economic Attribute', Photogrammetric
Engineering and Remote Sensing, vol. 65,
pp. 611-622, 1999.
V. Mesev, 'The Use of Census Data in
Urban Image Classification.',
Photogrammetric Engineering and Remote
Sensing, vol. 64, no. 5, pp. 431-438, 1998.
C. Lo and X. Yang, 'Drivers of land-use/
land-cover changes and dynamics modeling
for the Atlanta. Georgia Metropolitan Area.',
Photogrammetric Engineering and Remote
Sensing,, vol. 68, no. 10, pp. 1073–1082.,
2002.
[5] ureshi, S. Mahmood, A. S. Almas and R.
Irshad, 'MONITORING
SPATIOTEMPORAL AND MICRO-
LEVEL CLIMATIC VARIATIONS IN
LAHORE AND SUBRUBS USING
SATELLITE IMAGERY AND MULTI-
SOURCE DATA', Journal of Faculty of
Engineering & Technology, vol. 19, pp.
53,55, 2012.
[6] H. XU, X. WANG and G. XIAO, 'A
REMOTE SENSING AND GIS
INTEGRATED STUDY ON
URBANIZATION WITH ITS IMPACT ON
ARABLE LANDS: FUQING CITY,
FUJIAN PROVINCE, CHINA', LAND
DEGRADATION & DEVELOPMENT, vol.
11, p. 303, 2000.
43
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-52-320.jpg)
![A Review of Fundamentals of Hyperspectral
Imaging and its Applications
Asad Abbas, Khurram Khurshid
Electrical Engineering Department
Institute of Space Technology
Islamabad, Pakistan
Email: asad.abbas@outlook.com, khurram.khurshid@ist.edu.pk
Abstract—Over the past two decades, hyperspectral imagery
has emerged as an effective tool for geo-observation using
airborne and satellite based systems. Current sensor technologies
are capable of covering large surfaces of earth with exceptional
spatial, spectral and temporal resolutions. These features allows
the use of hyperspectral imaging in numerous remote sensing
applications requiring estimation of physical parameters of many
complex surfaces and identification of visually similar materials
having fine spectral signatures. Moreover, applications of hyper-
spectral imaging are growing rapidly, it has shown enormous
potential in various fields including biomedical, archeology, food
quality control, military defense and conservation of artistic and
historic objects. This review describes the differences between
hyperspectral and multispectral data, fundamentals of hyper-
spectral imaging and focuses on the modern applications of
hyperspectral imagery in precision agriculture, water resource
management and document imaging.
I. INTRODUCTION
Human eye is only able to see in a limited part of electro-
magnetic spectrum and can distinguish between objects based
on their different spectral responses in that narrow spectral
range [1]. However, multispectral imaging sensors have been
developed that are able to acquire an image in infrared and
visible segments of electromagnetic spectrum. Thus allowing
material identification on the basis of their unique spectral sig-
nature in a wide spectral range. Multispectral imaging exploits
the property that each material has its own unique spectral
signatures. Spectrum of a single pixel in a multispectral image
provides information about its constituents and surface of the
material.
Multispectral imaging technology is being used for envi-
ronment and land observational remote sensing from satel-
lite and airborne systems since late 1960s [2]. Multispectral
imaging systems acquire data in a small number of spectral
bands by using parallel sensor arrays. Most of the multispectral
imaging systems uses three to six spectral bands with large
optical band intervals, ranging from visible to near infrared
regions of electromagnetic spectrum for scene observation.
However, such low number of spectral bands are the limiting
factor for discrimination of various materials. The development
of hyperspectral sensing systems over the past two decades
made it possible to acquire several hundred spectral bands
of observational scene in a single acquisition. The increased
spectral resolution of these ”hyperspectral” images allow detail
examination of land surfaces and different materials present
in the observational scene, which was previously not possible
with low spectral resolution of multispectral imaging scanners.
Hyperspectral imaging (HSI), also called imaging spec-
trometer [3], is a technique in which an object is photographed
using several well defined optical bands in broad spectral
range. It was originally implemented on satellite and airborne
platforms for remote sensing applications but during last
two decades, HSI has been applied to numerous applications
including agricultural and water resources control [4], [5],
military defense, art conservation and archeology [6], [7],
medical diagnosis [8], [9], analyses of crime scene details [10],
[11], document imaging [12], forensic medicine [13], food
quality control [14], [15] and mineralogical mapping of earth
surface [16].
This review details the fundamentals of hyperspectral
imaging, discusses the common hyperspectral remote sensing
terminologies and highlights the modern applications of hy-
perspectral imagery in the field of precision agriculture, water
resource management and document imaging.
II. FUNDAMENTALS OF HYPERSPECTRAL IMAGING
Unlike normal 2-D images taken by regular camera, hy-
perspectral images are characterized by their spatial as well
as spectral resolution. The spatial resolution measures the
geometric relationship of the image pixels to each other.
While the spectral resolution determines the variations within
image pixels as a function of wavelength. Table I shows the
spatial and spectral resolution of the current airborne and
space satellite imaging sensors. Due to spectral resolution as
well as spatial resolutions generally hyperspectral images are
referred as hyperspectral data cubes. Figure 1 shows a typical
hyperspectral data cube as a result of hyperspectral imaging.
A. Spatial Resolution
Spatial resolution can be defined as the smallest discernible
detail in an image [17]. Which can be described as the measure
of smallest object in an image, that can be distinguished as a
separate entity in the image. In practical situations clarity of
the image is dictated by it spatial resolution, not the number of
pixels in an image. Spatial characteristics of an image depends
on the design of imaging sensor in terms of its field of view
and its altitude [18]. A finite patch of the ground is captured
by each detector in a remote imaging sensor. Spatial resolution
is inversely proportional to the patch size. The smaller the size
of the patch, higher the details that can be interpreted from the
observed scene.
44
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-53-320.jpg)
![Fig. 1. Hyperspectral datacube: for any image pixel a complete spectral curve
is observed [12]
TABLE I. CURRENT SPACEBORNE AND AIRBORNE SYSTEMS
PROVIDING OPTICAL DATA FOR LAND MAPPING
Optical Subsystem Spectral Spectral Spatial Resolution Spatial
Bands (m) Range (µm) Resolution Coverage
Spaceborne
Landsat VNIR-TIR 8 0.45-12.50 15-60 Global
MODIS VNIR-TIR 36 0.40-14.40 250-1000 Global
MERIS VNIR 15 0.39-1.040 300 Global
ASTER VNIR-TIR 15 0.52-11.65 15-90 Global
Hyperion VNIR-SWIR 242 0.40-2.500 30 Regional
ALOS VIS 1 0.52-0.77 2.5 Local
Airborne
AVIRIS VNIR 224 0.38-2.500 4-20 Local
HyMap VNIR-SWIR 128 0.45-2.480 2-10 Local
ROSIS VNIR 115 0.42-0.873 2 Local
DAIS-7915 VNIR-TIR 79 0.45-12 3-10 Local
B. Spectral Resolution
Spectral resolution can be defined as the number of spectral
bands and range of electromagnetic spectrum measured by
the sensor. An imaging sensor might respond to a large
frequency range but still have a low spectral resolution if it
acquires small number of spectral bands. On the contrary, if
a sensor is sensitive to small frequency range but captures
large number of spectral bands has high spectral resolution,
due to its ability to distinguish between scene elements having
close or similar spectral signatures [19]. Multispectral images
have a low spectral resolution, thus unable to resolve finer
spectral signatures present in the scene. Hyperspectral imaging
(HSI) sensors acquire images in numerous contiguous and
extremely narrow spectral bands in mid infrared, near infrared
and visible segments of electromagnetic spectrum. This type
of advance imaging system shows tremendous potential for
material identification on the basis of their unique spectral
signatures [2]. Spectrum of a single pixel in a hyperspectral
image can give considerably more information about the
surface of the material than a normal image.
C. Temporal Resolution
In hyperspectral remote sensing, the temporal resolution
depends on the orbital characteristics of the imaging sensor. It
is generally defined as the time needed by the sensor platform
to revisit and obtain data from the exact same location [20].
Temporal resolution is said to be high if the revisiting fre-
quency of the sensor platform for the exact same location is
high and is said to be low if revisiting frequency is low. It is
normally defined in days.
D. Understanding Spectral Signatures
Materials present on the earth’s surface emit, transmit,
reflect and absorb electromagnetic energy from the sun in their
own unique way. Hyperspectral sensors allows us to measure
all types of electromagnetic energy within a specified range
as it interacts with materials, thus allowing us to observe the
distinct features and changes on earth’s surface. Reflectance
is defined as the percentage of amount of energy bouncing
back from material’s surface. It is a ratio of reflected energy to
incident energy as a function of wavelength [18]. Reflectance is
100% if all the light energy of specific wavelength striking the
object is reflected back to the imaging sensor, on the other hand
reflectance is 0% if all the incident light of specific wavelength
is absorbed by the object. In most practical cases reflectance
values are in between these two extremes.
In a specified range of electromagnetic spectrum, the
reflectance values of different materials present on the earth’s
surface such as soil, forest, water, minerals etc can be plotted
and compared. Such plots are labeled as ”spectral signatures”
or spectral response curves” [21]. Figure 2 demonstrates a
general model of spectral signatures of different materials
present on the earth’s surface. Remotely sensed images can
be classified using these spectral signature plots, as each
material present in an observed scene has its own unique
spectral signature. The more the spectral resolution of an
imaging sensor, the more classification information can be
extracted from spectral signatures. Hyperspectral sensors have
high spectral resolution than multispectral sensors and thus
provide the ability to distinguish more subtle differences
in a scene. Hyperspectral imagery has been utilized by
Fig. 2. A generic scheme of HSI mapping of soil, vegetation and water [22]
geologists for mapping the land and water resources [16]. They
have additionally been utilized to map heavy metals and other
hazardous wastes in historic and active mining areas. Figure 3
shows the spectral response curves of dry bare soil, green
vegetation and clean water.
Figure 3 shows that the reflectance curve for bare soil
has less variations as compared to that of vegetation. This is
because of the fact that the factors that affect soil reflectance
vary in a narrow range of electromagnetic spectrum. These
factors include soil texture, presence of minerals such as iron,
surface roughness and moisture content in soil [21]. Spectral
45
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-54-320.jpg)
![Fig. 3. Spectral response curves of soil, vegetation and water [18]
signatures of green vegetation have valleys in the visible
portion of spectrum that indicates the pigmentation of the
plant. Chlorophyll is the primary photosynthetic pigment in
green vegetation [18], it is absorbs strongly in red (670 nm)
and blue (450 nm) regions called the chlorophyll absorption
spectral bands. When a plant is under stress so that the
chlorophyll growth is reduced, in such cases the amount of
reflectance in red (670 nm) regions increases [18]. The spectral
response of water has a distinctive characteristics of absorption
of light in near infrared and beyond. Common factors affecting
spectral response of water are the suspended sediments and
increases in chlorophyll levels. In each cases spectral response
will be shifted accordingly showing the presence of suspended
sediments or algae in water [21].
III. APPLICATIONS OF HYPERSPECTRAL IMAGING
A. Precision Agriculture
Many studies have indicated that the world’s crop
production needs to be doubled by the end 2050 due to
the rapid increase in world’s population [23]. However,
various studies have shown that the crop yields are not longer
increasing at a rate to fulfill the growing population needs
[24], [25]. Recent studies have indicated that increasing
crop yields, rather than using more land for cultivation,
is the most effective path for ensuring food security [26],
[27]. Global poverty and undernourishment can directly be
reduced by increasing crop production, moreover most of
the poor and undernourished population consists of farmers
themselves [28].
Traditionally crop monitoring for disease, water stress,
nutrients and insect attack was carried out by manual visual
inspection from the ground. These methods were limited by
the fact that the visual symptoms often appear at later stages
of disease, thus making it difficult to restore plant health.
Advancement in airborne and ground based hyperspectral
imaging methods has made possible the evaluation of crop
stresses, analyzing soil and vegetation characteristics in a
cost effective manner, thus replacing the traditional scouting
methods.
Drought stress is an important factor affecting crop
yields. Chances of a successful crop can be highly increased
by timely detection of water related stresses. High water
level stresses are noticeable in variations in photosynthetic
pigments. These changes leads to yellowish tint in crops, due
to the increase reflectance of red wavelength. Unlike human
eye, hyperspectral imaging sensors can detect these changes
at earlier stages. Colombo et al. [29] indicated that changes
in leaf equivalent water thickness (EWT) was responsible
for changes in leaf reflectance in the infrared and visible
spectrum. They stated that hyperspectral regression indices
calculated from hyperspectral imaging were powerful tools
for estimation of water content at leaf as well as at landscape
level. Rascher et al. [30] used a portable hyperspectral
imaging system and photochemical reflectance index to
estimate water stress in leaves of tropical tress and observed
the temporal effects of dehydrations on tree leaves. Rossini
et al. [31] found that hyperspectral imaging is useful in
detecting drought stress at farm level with corn. They showed
that irrigation deficits can be accurately mapped well before
drought stress affected the canopy structure.
Deficiencies in nutrients and soil contamination causes
various symptoms that can be assessed by hyperspectral
imaging. Schuerger et al. [32] used hyperspectral imaging to
observe zinc deficiency and toxicity to determine chlorophyll
levels relating to stress symptoms. They indicated that
traditional direct sampling methods are much more costly
than hyperspectral imaging. Dunagen et al. [33] analyzed
mercury levels in mustard plants and found that spectral
signatures were notably related to the contaminant levels.
Osborne et al. [34] showed that biomass, yield under
stress, nitrogen and phosphorous concentrations can be
estimated by using hyperspectral imaging. Mahlein et al. [35]
studied different development stages of diseased suger beet
leaves using hyperspectral imaging. Figure 4 shows spectral
signatures of healthy as well as diseased sugar beet leaves.
This study also showed that hyperspectral imaging has a great
potential for analyzing plant diseases.
Analysis of soil characteristics can play a vital role in
Fig. 4. Spectral signatures comparison of healthy and diseased suger beet
leaves [35]
increasing crop yields. Ben-Dor et al. [36] al successfully
mapped vital characteristics of soil in a field scale experiment
including moisture, soil organic matter and soil sanity. Rossel
and Bratney [37] estimated the organic carbon content in soil
with accuracy.
B. Water Resource Management
Water is one of the most important resource available
on the earth and is vital for survival of humanity. For this
46
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-55-320.jpg)
![reason, managing the water resource efficiently, analyzing
and monitoring the quality of water has attracted a lot of
attention from the researchers [38], [39], [40], [41], [42], [43].
Hyperspectral remote sensing technology has found enormous
applications in water resource management. Accurate estimates
of water resource parameters are possible by analyzing spatial,
spectral and temporal variations in water bodies.
The efficacy of flood detection and monitoring system is
limited by their incapacity to get important information about
water conditions from airborne and ground observatories in a
timely manner. Recent improvements in remote sensing tech-
nology has improved the early flood warning system and vastly
reduced the time of detection and reaction to flood events
to a few hours [44]. US geological survey and NASA are
incorporating space borne observations of rainfall resources,
rivers and land topography into early warning systems with
potential global applications [45]. Glaber and Reinartz [46]
studies the optimal procedure for detection of flooded areas
with remote sensing data. They investigated the erosive impact
of floods, moisture content in flood plain areas, accumula-
tion of sediments. Roux and Dartus [47] explored the flood
hydrographs and estimated the river discharge from remotely
sensed data. They optimized their model to minimize the error
between system response and their proposed model to estimate
the river discharge.
Hyperspectral remote sensing provides efficient and reli-
able information about water quality parameters which contain
biochemical, hydro-physical and biological attributes [48] [49]
[50]. Hyperspectral imaging enable us to measure chlorophyll,
turbidity and chemical oxygen demand and phosphorous in
water resources. Chlorophyll content in water is extensively
studied by hyperspectral remote sensing, which gives and
estimate of algal level and hence water quality. Studies have
been carried out for evaluation of ammonia changes for
wetland [51], classifying different parameters of lakes [52],
estuaries [53] and analyzing algal blooms [54].
Wetland mapping has played a significant role in order
to enhance the quality of our ecosystem [55]. Hyperspectral
imagery has helped in detailed understanding of vegetation
characteristics of ecosystem. Extensive research studies have
been carried out using remote sensing to explore the signif-
icance of acquiring timely data for monitoring and mapping
aquatic vegetation [56], which is said to be an important aspect
in ecosystem reconstruction and restoration.
C. Document Imaging
Traditionally, forensic document experts and paleographers
used chemical solution based methods to study the extrinsic
and intrinsics components of the important historic docu-
ments [57]. This is due to the fact that the inks used on
documents throughout the history were composed of diverse
substances having distinct chemical and physical properties.
Each of these substances have their own unique way for
reacting with different substrates depending upon the reaction
environment. These chemical solution based methods helped
in document analysis. But unfortunately these techniques were
time consuming, sensitive to temperature changes and destruc-
tive in nature i.e. harms to the important documents were
irreversible.
To overcome such limitations, hyperspectral imaging has
emerged as an effective non-destructive tool for improving
readability [58], ink aging and forensic document analysis [59].
Hyperspectral document imaging works on the principle that
each ink present in the document has its own unique spectral
signature. Many mathematical tools that are being used in
hyperspectral remote sensing can be used on hyperspectral
document images for classification, improving legibility of
extremely deteriorated text, ink aging and fraud detection.
Moreover hyperspectral imaging is non-destructive, automated
and environment insensitive tool for document examination.
In hyperspectral document imaging, ink mismatch de-
tection analysis provides important information to forensic
document examiners to determine the authenticity of legal
documents. Forgery, backdating and fraud can be detected
using ink analysis of documents. Recently, we have worked on
a non-destructive automated hyperspectral document imaging
analysis system, which was able successfully discriminates
between different visually similar ink mixtures (Dataset D1 to
D7) taken from hyperspectral imaging database [60]. Figure 5
shows the final segmentation results of our proposed approach.
Fig. 5. Comparisons of Ground Truth Images (Dataset D1 to D7) with final
segmentation results
IV. DISCUSSION AND CONCLUSION
Among remote sensing technologies, the role of hyperspec-
tral imagery in the geo-observation, identification and detection
of materials and estimation of physical parameters cannot be
stated enough. Due to this very reason there are increasing
number of airborne and spaceborne hyperspectral platforms
based applications being researched. Recent advancement in
47
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-56-320.jpg)
![sensor technologies has encouraged researchers to use hyper-
spectral imagery in many modern applications. Many mathe-
matical tools and algorithms are being researched such as data
fusion, hyperspectral unmixing, hyperspectral classification,
anomaly detection and fast computing for efficient utilization
of hyperspectral data. These mathematical tools can be used
on hyperspectral data across many different applications.
In general, this review focuses on the vast extent to which
hyperspectral imaging has been used to increase the crop
yields, managing water resources and its advanced application
in historic and modern document imaging. Promising results
have been found in those areas and also future research is
being carried out for further improvement as well.
REFERENCES
[1] E. H. Land and J. McCann, “Lightness and retinex theory,” JOSA,
vol. 61, no. 1, pp. 1–11, 1971.
[2] D. Landgrebe, “Information extraction principles and methods for
multispectral and hyperspectral image data,” Information processing for
remote sensing, vol. 82, pp. 3–38, 1999.
[3] J. Solomon and B. Rock, “Imaging spectrometry for earth remote
sensing,” Science, vol. 228, no. 4704, pp. 1147–1152, 1985.
[4] M. Govender, K. Chetty, and H. Bulcock, “A review of hyperspectral
remote sensing and its application in vegetation and water resource
studies,” Water Sa, vol. 33, no. 2, 2007.
[5] E. Adam, O. Mutanga, and D. Rugege, “Multispectral and hyperspectral
remote sensing for identification and mapping of wetland vegetation: a
review,” Wetlands Ecology and Management, vol. 18, no. 3, pp. 281–
296, 2010.
[6] C. Fischer and I. Kakoulli, “Multispectral and hyperspectral imaging
technologies in conservation: current research and potential applica-
tions,” Studies in Conservation, vol. 51, no. Supplement-1, pp. 3–16,
2006.
[7] H. Liang, “Advances in multispectral and hyperspectral imaging for
archaeology and art conservation,” Applied Physics A, vol. 106, no. 2,
pp. 309–323, 2012.
[8] O. Carrasco, R. B. Gomez, A. Chainani, and W. E. Roper, “Hyperspec-
tral imaging applied to medical diagnoses and food safety,” in AeroSense
2003. International Society for Optics and Photonics, 2003, pp. 215–
221.
[9] M. A. Afromowitz, J. B. Callis, D. M. Heimbach, L. A. DeSoto, and
M. K. Norton, “Multispectral imaging of burn wounds: a new clinical
instrument for evaluating burn depth,” Biomedical Engineering, IEEE
Transactions on, vol. 35, no. 10, pp. 842–850, 1988.
[10] J. Kuula, I. P¨ol¨onen, H.-H. Puupponen, T. Selander, T. Reinikainen,
T. Kalenius, and H. Saari, “Using vis/nir and ir spectral cameras for
detecting and separating crime scene details,” in SPIE Defense, Security,
and Sensing. International Society for Optics and Photonics, 2012, pp.
83 590P–83 590P.
[11] R. L. Schuler, P. E. Kish, and C. A. Plese, “Preliminary observations
on the ability of hyperspectral imaging to provide detection and visu-
alization of bloodstain patterns on black fabrics,” Journal of forensic
sciences, vol. 57, no. 6, pp. 1562–1569, 2012.
[12] R. Padoan, T. A. Steemers, M. Klein, B. Aalderink, and G. De Bruin,
“Quantitative hyperspectral imaging of historical documents: technique
and applications,” ART Proceedings, 2008.
[13] G. Edelman, E. Gaston, T. Van Leeuwen, P. Cullen, and M. Aalders,
“Hyperspectral imaging for non-contact analysis of forensic traces,”
Forensic science international, vol. 223, no. 1, pp. 28–39, 2012.
[14] A. Gowen, C. O’Donnell, P. Cullen, G. Downey, and J. Frias, “Hyper-
spectral imaging–an emerging process analytical tool for food quality
and safety control,” Trends in Food Science & Technology, vol. 18,
no. 12, pp. 590–598, 2007.
[15] Y.-Z. Feng and D.-W. Sun, “Application of hyperspectral imaging in
food safety inspection and control: a review,” Critical reviews in food
science and nutrition, vol. 52, no. 11, pp. 1039–1058, 2012.
[16] R. N. Clark and G. A. Swayze, “Mapping minerals, amorphous ma-
terials, environmental materials, vegetation, water, ice, and snow, and
other materials: The usgs tricorder algorithm,” in Summaries of the Fifth
Annual JPL Airborne Earth Science Workshop, vol. 1. JPL Publication,
1995, pp. 39–40.
[17] R. C. Gonzalez, Digital image processing. Pearson Education India,
2009.
[18] R. Smith, “Introduction to remote sensing of the environment,” 2001b.
www. microimages. com, 2001.
[19] J. R. Jensen, Introductory digital image processing: A remote sensing
perspective. Pearson College Division, 2005.
[20] J. Thau, “Temporal resolution,” in Encyclopedia of GIS. Springer US,
2008, pp. 1150–1151. [Online]. Available: http://dx.doi.org/10.1007/
978-0-387-35973-1 1376
[21] T. Lillesand, R. W. Kiefer, and J. Chipman, Remote sensing and image
interpretation. John Wiley & Sons, 2014.
[22] G. A. Shaw and H.-h. K. Burke, “Spectral imaging for remote sensing,”
Lincoln Laboratory Journal, vol. 14, no. 1, pp. 3–28, 2003.
[23] D. Tilman, C. Balzer, J. Hill, and B. L. Befort, “Global food demand
and the sustainable intensification of agriculture,” Proceedings of the
National Academy of Sciences, vol. 108, no. 50, pp. 20 260–20 264,
2011.
[24] K. G. Cassman, “Ecological intensification of cereal production sys-
tems: yield potential, soil quality, and precision agriculture,” Proceed-
ings of the National Academy of Sciences, vol. 96, no. 11, pp. 5952–
5959, 1999.
[25] D. K. Ray, N. Ramankutty, N. D. Mueller, P. C. West, and J. A.
Foley, “Recent patterns of crop yield growth and stagnation,” Nature
communications, vol. 3, p. 1293, 2012.
[26] H. C. J. Godfray, J. R. Beddington, I. R. Crute, L. Haddad, D. Lawrence,
J. F. Muir, J. Pretty, S. Robinson, S. M. Thomas, and C. Toulmin, “Food
security: the challenge of feeding 9 billion people,” science, vol. 327,
no. 5967, pp. 812–818, 2010.
[27] J. A. Foley, N. Ramankutty, K. A. Brauman, E. S. Cassidy, J. S. Gerber,
M. Johnston, N. D. Mueller, C. OConnell, D. K. Ray, P. C. West et al.,
“Solutions for a cultivated planet,” Nature, vol. 478, no. 7369, pp. 337–
342, 2011.
[28] P. L. Pingali, “Green revolution: Impacts, limits, and the path ahead,”
Proceedings of the National Academy of Sciences, vol. 109, no. 31, pp.
12 302–12 308, 2012.
[29] R. Colombo, M. Meroni, A. Marchesi, L. Busetto, M. Rossini, C. Gi-
ardino, and C. Panigada, “Estimation of leaf and canopy water content
in poplar plantations by means of hyperspectral indices and inverse
modeling,” Remote Sensing of Environment, vol. 112, no. 4, pp. 1820–
1834, 2008.
[30] U. Rascher, C. J. Nichol, C. Small, and L. Hendricks, “Monitoring
spatio-temporal dynamics of photosynthesis with a portable hyperspec-
tral imaging system,” Photogrammetric Engineering & Remote Sensing,
vol. 73, no. 1, pp. 45–56, 2007.
[31] M. Rossini, F. Fava, S. Cogliati, M. Meroni, A. Marchesi, C. Panigada,
C. Giardino, L. Busetto, M. Migliavacca, S. Amaducci et al., “Assessing
canopy pri from airborne imagery to map water stress in maize,” ISPRS
Journal of Photogrammetry and Remote Sensing, vol. 86, pp. 168–177,
2013.
[32] A. C. Schuerger, G. A. Capelle, J. A. Di Benedetto, C. Mao, C. N.
Thai, M. D. Evans, J. T. Richards, T. A. Blank, and E. C. Stryjewski,
“Comparison of two hyperspectral imaging and two laser-induced
fluorescence instruments for the detection of zinc stress and chlorophyll
concentration in bahia grass (paspalum notatum flugge.),” Remote
sensing of environment, vol. 84, no. 4, pp. 572–588, 2003.
[33] S. C. Dunagan, M. S. Gilmore, and J. C. Varekamp, “Effects of mercury
on visible/near-infrared reflectance spectra of mustard spinach plants
(brassica rapa p.),” Environmental Pollution, vol. 148, no. 1, pp. 301–
311, 2007.
[34] S. Osborne, J. S. Schepers, D. Francis, and M. R. Schlemmer, “De-
tection of phosphorus and nitrogen deficiencies in corn using spectral
radiance measurements,” Agronomy journal, vol. 94, no. 6, pp. 1215–
1221, 2002.
[35] A.-K. Mahlein, U. Steiner, C. Hillnh¨utter, H.-W. Dehne, and E.-C.
Oerke, “Hyperspectral imaging for small-scale analysis of symptoms
48
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-57-320.jpg)
![caused by different sugar beet diseases,” Plant methods, vol. 8, no. 1,
p. 3, 2012.
[36] E. Ben-Dor, K. Patkin, A. Banin, and A. Karnieli, “Mapping of several
soil properties using dais-7915 hyperspectral scanner data-a case study
over clayey soils in israel,” International Journal of Remote Sensing,
vol. 23, no. 6, pp. 1043–1062, 2002.
[37] C. Gomez, R. A. V. Rossel, and A. B. McBratney, “Soil organic
carbon prediction by hyperspectral remote sensing and field vis-nir
spectroscopy: An australian case study,” Geoderma, vol. 146, no. 3,
pp. 403–411, 2008.
[38] F. Zagolski, V. Pinel, J. Romier, D. Alcayde, J. Fontanari, J. Gastellu-
Etchegorry, G. Giordano, G. Marty, E. Mougin, and R. Joffre, “Forest
canopy chemistry with high spectral resolution remote sensing,” Inter-
national Journal of Remote Sensing, vol. 17, no. 6, pp. 1107–1128,
1996.
[39] G. P. Asner, “Biophysical and biochemical sources of variability in
canopy reflectance,” Remote sensing of Environment, vol. 64, no. 3, pp.
234–253, 1998.
[40] K. McGwire, T. Minor, and L. Fenstermaker, “Hyperspectral mixture
modeling for quantifying sparse vegetation cover in arid environments,”
Remote Sensing of Environment, vol. 72, no. 3, pp. 360–374, 2000.
[41] C. Stone, L. Chisholm, and N. Coops, “Spectral reflectance character-
istics of eucalypt foliage damaged by insects,” Australian Journal of
Botany, vol. 49, no. 6, pp. 687–698, 2001.
[42] N. Coops, S. Dury, M.-L. Smith, M. Martin, and S. Ollinger, “Com-
parison of green leaf eucalypt spectra using spectral decomposition,”
Australian journal of botany, vol. 50, no. 5, pp. 567–576, 2002.
[43] E. Underwood, S. Ustin, and D. DiPietro, “Mapping nonnative plants
using hyperspectral imagery,” Remote Sensing of Environment, vol. 86,
no. 2, pp. 150–161, 2003.
[44] F. Ip, J. Dohm, V. Baker, T. Doggett, A. Davies, R. Castano, S. Chien,
B. Cichy, R. Greeley, R. Sherwood et al., “Flood detection and
monitoring with the autonomous sciencecraft experiment onboard eo-1,”
Remote Sensing of Environment, vol. 101, no. 4, pp. 463–481, 2006.
[45] G. Brakenridge, E. Anderson, and S. Nghiem, “Satellite microwave
detection and measurement of river floods,” in AGU Spring Meeting
Abstracts, vol. 1, 2006, p. 05.
[46] C. Gl¨aßer and P. Reinartz, “Multitemporal and multispectral remote
sensing approach for flood detection in the elbe-mulde region 2002,”
Acta hydrochimica et hydrobiologica, vol. 33, no. 5, pp. 395–403, 2005.
[47] H. Roux and D. Dartus, “Use of parameter optimization to estimate a
flood wave: Potential applications to remote sensing of rivers,” Journal
of Hydrology, vol. 328, no. 1, pp. 258–266, 2006.
[48] M. Chawira, T. Dube, and W. Gumindoga, “Remote sensing based
water quality monitoring in chivero and manyame lakes of zimbabwe,”
Physics and Chemistry of the Earth, Parts A/B/C, vol. 66, pp. 38–44,
2013.
[49] T. Kutser, B. Paavel, C. Verpoorter, T. Kauer, and E. Vahtm¨ae, “Remote
sensing of water quality in optically complex lakes,” in Proceedings of
the XXII Congress of the International Society for Photogrammetry and
Remote Sensing, Melbourne, Australia, 2012.
[50] S. Koponen, “Remote sensing of water quality: the development and
use of water processors available in beam,” Finland, Helsinki: Helsinki
University of Technology TKK, Department of Radio Science and
Engineering, pp. 1–58, 2006.
[51] D. R. Tilley, M. Ahmed, J. H. Son, and H. Badrinarayanan, “Hyper-
spectral reflectance of emergent macrophytes as an indicator of water
column ammonia in an oligohaline, subtropical marsh,” Ecological
Engineering, vol. 21, no. 2, pp. 153–163, 2003.
[52] S. Thiemann and H. Kaufmann, “Lake water quality monitoring using
hyperspectral airborne dataa semiempirical multisensor and multitem-
poral approach for the mecklenburg lake district, germany,” Remote
sensing of Environment, vol. 81, no. 2, pp. 228–237, 2002.
[53] J.-M. Froidefond, L. Gardel, D. Guiral, M. Parra, and J.-F. Ternon,
“Spectral remote sensing reflectances of coastal waters in french guiana
under the amazon influence,” Remote Sensing of Environment, vol. 80,
no. 2, pp. 225–232, 2002.
[54] R. P. Stumpf, “Applications of satellite ocean color sensors for monitor-
ing and predicting harmful algal blooms,” Human and Ecological Risk
Assessment: an International Journal, vol. 7, no. 5, pp. 1363–1368,
2001.
[55] K. Schmidt and A. Skidmore, “Spectral discrimination of vegetation
types in a coastal wetland,” Remote sensing of Environment, vol. 85,
no. 1, pp. 92–108, 2003.
[56] L. Yuan and L. Zhang, “Identification of the spectral characteristics of
submerged plant vallisneria spiralis,” Acta Ecologica Sinica, vol. 26,
no. 4, pp. 1005–1010, 2006.
[57] V. Aginsky, “Forensic examinaton of” slightly soluble” ink pigments
using thin-layer chromatography,” Journal of Forensic Sciences, vol. 38,
pp. 1131–1131, 1993.
[58] S. J. Kim, F. Deng, and M. S. Brown, “Visual enhancement of old
documents with hyperspectral imaging,” Pattern Recognition, vol. 44,
no. 7, pp. 1461–1469, 2011.
[59] G. Edelman, E. Gaston, T. Van Leeuwen, P. Cullen, and M. Aalders,
“Hyperspectral imaging for non-contact analysis of forensic traces,”
Forensic science international, vol. 223, no. 1, pp. 28–39, 2012.
[60] A. Abbas, K. Khurshid, and F. Shafait, “Ink Mismatch Detection using
Hyperspectral Document Image Analysis,” Master’s thesis, Institute of
Space Technology, Islamabad, Pakistan, 2015.
49
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-58-320.jpg)
![A Numerical Study on the Impact Resistance of
Braided Composites
X. Lei
Dept. of Mechanical
Eng., Hanyang
University, South
Korea
K. Hayat
Dept. of Mechanical Eng.,
University of Lahore, Lahore,
Pakistan
Email: khazar.hayat@me.uol.edu.pk
M. T. Hussain
Satellite Research
and Development
Center, Lahore,
Pakistan
H. T. Ali
Dept. of Aerospace Eng.,
Queen's Building,
University of Bristol,
BS8 1TR, UK
R. S. Choudhry
Dept. of Mechanical Eng.,
National University of
Science and Technology,
Islamabad, Pakistan
Abstract— Braided fabric reinforced composites are
attractive alternatives to conventional prepreg laminates due to
higher impact resistant performance. For the development of
aerospace structures made of braided composites, the numerical
simulations can play a key role in reducing time and cost.
However, the numerical tools to evaluate the impact response and
to investigate the damage mechanism of these materials, are not
fully developed yet. This work focuses on the numerical
modelling and analysis of the impact induced response of braided
composites under different impact velocities, using micro-
mechanics of failure (MMF) together with progressive damage
models for fibre and matrix constituents. For implementation of
MMF, the meso- and micro-scale models of unit cell have been
utilized. In the meso-scale unit cell of the braided composites, the
tows effective material properties were updated using a micro
unit cell for which the degradation of constituent material
properties was incorporated. The impact induced damage
progression for the tows and pure matrix was then evaluated.
The impact analyses were performed for bi-axial braided
composites at 45 degrees and 25 degrees braiding angles to
demonstrate the feasibility of the MMF-based approach.
Moreover, the impact response of the braided composites with
both thermosetting and thermos-plastic resin system was also
investigated.
Keywords— Braided composites, Impact resistance, Micro-
mechanics of failure, Progressive damage model
I. INTRODUCTION
Recently, a rapid growth of the applications of textile
braided fabric reinforced composites have been observed in
the marine, aerospace and automobile industries etc. due to its
net-shape fabrication and resistance to delamination damage
and impact load in particular [1]. However, the behavior of the
braided composites is not fully understood yet, and there is a
need of establishing a methodology that can predict the
material failure behaviors under different types of loadings,
primarily the impact loading [2].
The purpose of this study is to evaluate the behavior of
braided composites subjected to impact load with
thermoplastic and thermosetting resin systems, based on
various braiding angles and using micromechanics of failure
(MMF) theory [3]. The analysis procedure involves three
scopes: macro, meso and micro, as shown in Fig. 1. A braided
composite can be assumed to be limited to mesoscale. It
consists of tows and pure matrix encapsulating the tows. Thus,
the use of mesoscale unit cell model serves as a bridge for
transformation of stresses and effective material properties
between the macro level and micro level analyses.
II. MICROMECHANICS-BASED PROGRESSIVE DAMAGE MODEL
A. Geometric modeling of braided composites
The development of an idealized model of the braided
composites is necessary to predict their material behaviors.
The fiber bundle arrangements are modelled using the
geometrical parameters (i.e. braiding angle, width and
thickness of axial tow, width and thickness of bias tow, and
gap between bias tows) illustrated in Fig. 2. Moreover, in the
modeling, the undulation of tows plays a pivotal role in
determining the behavior of the braided composites. Further
details on modeling procedure and techniques can be found in
[2].
Fig. 1. Analysis of braided composites.
B. Meso and micro analyses
The meso- and micro-level analyses are bridged to
characterize the material behaviour of the tows of the braided
Fig. 2. Modeling of braided composites.
50
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-59-320.jpg)
![composites. It should be noted that a tow can be approximated
as a curved continuous unidirectional (UD) lamina, and a
micro unit cell, shown in Fig. 3c, can be used to relate the
constituent behaviour to the tow behaviour. Thus, tow stresses
can be transformed into each constituent stresses, and resulting
constituent damages can be used to determine the tow damage.
For the constituent damage analysis, the degradation of the
constituent stiffness properties is carried out, and the tow
effective material properties are subsequently updated through
the micro unit cell.
Fig. 3. Estimation of effective material properties of tows of braided
composites.
The transformation between the tows stresses at meso-
level, denoted by , and the fiber and matrix stresses at
micro-level, denoted by , and , is carried using the stress
amplification factors (SAFs), denoted by and matrices, as
following [3, 4]:
(1)
Eq. (1) can be written in detailed form, as following [4]:
(2)
In a similar manner, the meso strain and temperature
increment can also be transformed to micro strains, as shown
in Eq. 3, using the strain amplification factors which can be
derived from SAFs using the constituent compliance matrix
and the unit cell compliance matrix .
(3)
C. Damage model of micro-constituents
For each constituent, a separate failure criterion is used
due to their distinct mechanical behaviors. Considering the
isotropic nature of the matrix constituent, the von Mises
failure criterion [3, 5] is used, which follows:
(4)
where , , , and represents the first stress invariant,
von Misses equivalent stress, tensile strength and compressive
strengths of matrix constituent, respectively. The failure
criterion described by Eq. 4 matches the condition that the
Stassi’s equivalent stress , reaches the matrix tensile
strength :
(5)
where the symbol represents the ratio of the compressive
strength to the tensile strength of the matrix constituent.
Further, the equivalent stress can be transformed into the
equivalent strain of the matrix constituent, described in Eq. 6,
using the stress-strain relations.
(6)
where , , and represents the first strain invariant, the
von Mises equivalent strain and the Poisson’s ratio of the
matrix constituent.
It should be noted that the equivalent strain, like the
equivalent stress, is a scalar quantity that is computed from the
strain components. Fig. 4a represents a multi-linear stress-
strain model defined using the equivalent stress and the
equivalent strain, to estimate the matrix constituent damage.
The matrix constituent behaves in a linear manner, before the
occurring of any damage. Once damage occurs, the matrix
constituent demonstrates a nonlinear behavior (i.e. hardening
followed by the softening depending on the damage status).
The stiffness of the matrix constituent is degraded by a factor
as following:
(7)
subjected to the condition: . The symbols
and denotes the matrix yield stress and the matrix
yield strain at the beginning, and the symbols and
denotes the matrix yield stress and matrix yield strain at the
end of ith step. The symbol denotes the current equivalent
strain of the matrix at the ith step, and the symbol denotes
the matrix intact stiffness.
For the mesoscale model, the damage in the pure matrix
was modeled using Eq. 8 only, and additional MMF-based
damage modeling using a micro unit cell was used for the
matrix of tows [6]. The periodic boundary conditions were
used for the meso and micro unit cells [2, 7]. The micro unit
cell model, based on the MMF theory and linear material
model, delivered multi-axial micro stress state for the fibre
and the matrix constituents (i.e. assuming perfect bonding for
interface). To avoid computation instability, the maximum
local damage value for fiber and matrix constituent was set to
0.9 [8].
The mesoscale stresses were applied on the micro unit
cell. After performing the finite element analysis, the micro
stresses/strains in the matrix were computed, as shown in Fig.
5. Afterwards, the damage analysis for the matrix was
performed. For the matrix of tow, the localization of damage
51
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-60-320.jpg)
![was avoided by using a volume-based damage
homogenization technique, represented by Eq. 8 shown below:
(8)
where the symbol denotes a positive weighing factor,
and the symbol denotes the matrix overall volume in the
micro unit cell.
After estimation of damage status, the tow effective
material properties in the meso unit cell model were re-
evaluated from the micro unit cell and the matrix material
properties were degraded. Afterwards, the damage
propagation inside the tows and the pure matrix was
evaluated. For the matrix constituent, following constitutive
relation was used to degrade the stiffness:
(9)
where denotes the stiffness of matrix zone in the micro
unit cell.
For the fiber constituent, the maximum stress failure
criterion, which is , was used for the fiber
constituent. The symbol presents the micro longitudinal
stress of the fiber constituent, and symbols and denotes
the longitudinal tensile strength and longitudinal compressive
strength of the fiber constituent, respectively. Fig. 4b
illustrates that when the fiber constituent fails its stiffness is
dramatically degraded. Since, carbon fiber behaves in a brittle
manner, therefore, the highest value of all elemental damage
factors was considered, as following:
(10)
where the symbol denotes the damage factor of jth
element of the fiber constituent. And, the constitutive relation
for fiber constituent was:
(11)
where denotes the stiffness of overall fiber zone in the
micro unit cell.
D. Numerical implementation
Fig. 6 illustrates the flow chart of the algorithm developed
for combining the MMF methodology and progressive damage
model. In the beginning, the total global strain, denoted
by , is computed at the global time , and augmenting the
global strain increment, denoted by , to global strain at
previous time-step n-1. For the tows, the macro stresses of
each tow element, denoted by , are estimated using the
previous effective stiffness properties, denoted by .
Afterwards, the micro stresses for each element of the both
matrix and fiber zones in the micro unit cell, denoted by
and , are computed from the meso stress, denoted by ,
by using SAFs. The constituent failure criteria are then applied
to both matrix and fiber, and damage factor is estimated for
each jth element in the matrix and fiber zones, denoted by
and , respectively, by employing the relevant constituent
damage models. The overall damage factor for both matrix
and fiber zones, denoted by and , are then evaluated
based on their respective damage methods (i.e. the damage
homogenization for the matrix constituent, and the maximum
damage for fiber constituent, as discussed previously).
Based on the status of overall damage factor, the stiffness
properties of the matrix and fiber are degraded, and the tow
effective properties are evaluated for the next time increment.
The numerical implementation was done using Abaqus 6.11-1,
a commercial finite element solver, combined with its user
subroutine VUSDFLD [9].
Fig. 4. (a) Multi-linear model for the matrix constituent damage, and (b)
linear model for the fiber constituent damage.
Fig. 5. Damage estimation of the matrix constituent: (a) applied load; (b)
micro stresses; (c) element damage factor; (d) overall damage factor.
Fig. 6. Algorithm combining the MMF methodology and progressive
damage model.
52
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-61-320.jpg)
![III. MATERIAL PROPERTIES
The stiffness properties of carbon fiber constituent are
listed in TABLE 1 [1]. However, the strength properties of
carbon fibers, shown in TABLE 1, were back-calculated using
micromechanics approach. The stiffness and strength
properties of the matrix constituent, listed in TABLE 2, were
taken from in-house material testing database.
TABLE 1. MATERIAL PROPERTIES OF CARBON FIBERS [1].
Material property Value
Longitudinal modulus: Ef1 (GPa) 276.0
Transverse modulus: Ef2 (GPa) 27.6
In-plane shear modulus: Gf,12 (GPa) 138
Transverse modulus: Gf,23 (GPa) 7.8
Major Poisson’s ratio: vf,12 0.3
Major Poisson’s ratio: vf,23 0.8
Longitudinal tensile strength: Tf (MPa) 3800
Longitudinal compressive strength: Cf (MPa) 2980
TABLE 2. MATERIAL PROPERTIES OF THE MATRIX RESIN.
Material property Thermosetting Thermoplastic
Elastic modulus: Em (GPa) 3.45 2
Elastic Poisson’s ratio: vm 0.35 0.35
Final tensile strength: Tm (MPa) 65 65
TABLE 3. TOWS EFFECTIVE MATERIAL PROPERTIES.
Material property
Tows with
thermosetting
resin
Tows with
thermoplastic
resin
Longitudinal modulus: E1 (GPa) 215.67 215.67
Transverse modulus: E2 (GPa) 13.55 13.55
In-plane shear modulus: G12 (GPa) 9.0 9.0
Transverse modulus: G23 (GPa) 4.29 4.29
Major Poisson’s ratio: v12 0.309 0.309
Major Poisson’s ratio: v23 0.592 0.592
The tows effective properties, listed in TABLE 3, were
computed from micro unit cell, using the stiffness properties
of the matrix and fiber constituents, along with a fiber volume
fraction of 0.78. It should be noted that the total fiber volume
fraction of the meso unit cell was approximately 0.5,
computed in accordance with [2]. The multi-linear damage
models consisting of hardening and softening behaviors,
shown in Fig. 7, for the matrix constituent (i.e. thermoplastic
and thermosetting resin systems) was also established based
on in-house test data.
IV. IMPACT ANALYSIS SETTING
For impact analysis, the whole model of biaxial (BX)
braided fabrics was composed of fiber, pure matrix, matrix in
tows, and a ballistic projectile. For the geometrical modeling
parameters, the tow width and the two thickness used were
3.01 mm and 0.5 mm, respectively. There was a gap of 0.002
mm between the biased tows. The amplitude of the tow
undulation of biaxial braids was set to 0.252 mm. Two kinds
of biaxial braided composites with thermoplastic and
thermosetting resin systems, and each having braided angle of
45 degrees and 25 degrees, hereafter denoted with BX45 and
BX25, respectively, were investigated.
For the ballistic projectile, the radius was 4 mm. The
initial position of the projectile was set above the braided
fabric model and the gap between the bottom of the projectile
and the top surface of the braided models was set to 0.1 mm.
Fixed boundary conditions were applied on the all four sides
of the braided composites, and the top and bottom boundaries
of the braided model were assigned with free boundary
conditions. The ballistic projectile was taken as a rigid body
controlled by a reference point, which was constrained to
move only in z-direction. As illustrated in Fig. 8, the projectile
was given an initial velocity, and the contact properties were
also assigned on the braided model and the projectile by
setting a value of 0.28 for the coefficient of friction. Once the
projectile touches the top surface of the braided model, the
contact behavior will be taken into account. Two initial
velocities of 20 m/s and 50 m/s were applied to the projectile.
At initial velocity of 20 m/s, no penetration occurred and
projectile bounced back after impact. However, at initial
velocity of 50 m/s, the projectile penetrated through the
braided composite.
V. RESULTS AND DISCUSSION
Fig. 9 shows the damage contours for the BX45 braided
composite models, with thermoplastic and thermosetting
resins, at the impact velocity of 50 m/s. For the pure matrix
and the matrix in tows, the damage contour with element
deletion are shown in Fig. 9(a-c) and Fig. 9(b-d), respectively.
Fig. 7. Stress-strain behaviors of the matrix resin systems.
Fig. 8. Impact analysis of various types of biaxial braided
composites: (a) braids at 45 degrees, and (b) braids at 25 degrees.
53
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-62-320.jpg)
![On the contrary, the impact resistance of the braided
composite model with thermoplastic and thermosetting resins,
can be clearly distinguished at the projectile impact velocity of
50 m/s, as shown in Fig. 11(c-d). It should be noted that the
projectile with a higher impact velocity of 50 m/s, penetrates
through the braided composite model, as can be seen from the
velocity time histories. The residual velocity of the projectile
decreases with the passage of time, but remains positive
throughout the recorded time of 0.2 seconds. For both BX45
and BX25 braided composite model with thermoplastic resin,
the residual velocity of the projectile is clearly lower than that
of the braided composite model with thermosetting resin. The
decrease in residual velocity of the projectile means that the
impact energy is absorbed by the braided composite model.
Higher the decrease in the residual velocity of the projectile,
the higher would be the impact energy absorption. Based on
this fact, it can be inferred that the braided composite model
with thermoplastic resin exhibits better impact resistance than
that of the braided composite model with thermosetting resin.
Fig. 12 and Fig. 13 show that the impact damage evolution
for the pure matrix and for the matrix in tows, respectively, of
the BX45 and BX25 braided composite models with
thermoplastic and thermosetting resin systems. For the pure
resin, it can be seen that the highest damage occurs for both
BX45 and BX25 braided composite models with
thermosetting resin (Fig. 12), for the impact velocities of 20
m/s and 50 m/s.
TABLE 4. PERCENTAGE ELEMENT DELETION.
Impact
velocity
(m/s)
Percentage
element
deletion
in (%)
BX45 Braided
composite
BX25 Braided
composite
Thermo-
plastic
resin
Thermo-
setting
resin
Thermo-
plastic
resin
Thermo-
setting
resin
20
Pure matrix 2.38 5.38 3.05 5.29
Tow matrix 0.09 0.14 0.18 0.31
50
Pure matrix 14.0 20.57 13.59 19.36
Tow matrix 1.89 2.04 1.71 1.84
However, in the case of the matrix in tows, the highest
damage occurs for the BX45 and BX25 braided composite
models with thermoplastic resin (Fig. 13), for impact
velocities of 20 m/s and 50 m/s. It can also be clearly seen that
the damage values of the matrix in the tows are higher in
magnitude than that of the pure matrix. Thus, overall damage
status of the braided composite model is governed by the
matrix in tows. Moreover, higher impact damage values also
mean that more absorption of the impact energy and higher
reduction in the residual velocity of the projectile after impact,
as previously illustrated in Fig. 11. Since, damage values of
the matrix in tows for the braided composite model with
thermoplastic resin are higher and well spread (as previous
shown Fig. 9) than that of the braided composite model with
thermosetting resin, therefore, it can be inferred that the
braided composite with thermoplastic resin system can
withstand higher impact loads, thereby, demonstrate better
impact resistance.
Finally, the percentage element deletion for the braided
composite models with both thermoplastic and thermosetting
resin are compared in TABLE 4. The percentage of element
deletion of the pure resin and of the matrix in tows for the
braided composite with thermoplastic resin system is lower
than that for the braided composite with thermosetting resin
system, as it withstand higher impact loads. Since, the braided
composite with thermoplastic resin system possess better
impact resistance, therefore, the small number of elements
reach the critical damage values that once reach results in the
element deletion.
VI. CONCLUSIONS
A numerical study was carried out to evaluate the impact
resistance of the biaxial (BX) braided composites made of
thermoplastic and thermosetting resin systems, and subjected
to a circular ballistic projectile of radius 4 mm. Two impact
conditions, the bouncing back of projectile with impact
velocity of 20 m/s and the full perpetration of projectile at
impact velocity of 50 m/s, were simulated. Based on the
micromechanics of failure (MMF) theory, the impact
resistance of the braided composite models was evaluated in
terms of the damage evolution in the pure matrix and the
matrix in the tows, and the percentage of element deletion.
Results showed that the braided composites with thermoplastic
resin demonstrate better impact resistance, than that of the
braided composite with thermosetting resin, for both 45
degrees and 25 degrees braiding angles and for all impact
conditions. From the numerical simulations, it was difficult to
evaluate the effect of braiding angles on the impact resistance
of the braided composites, and requires further investigation.
Moreover, future work will involve the validation of the
MMF-based impact analysis methodology, presented here,
with the experimental data.
ACKNOWLEDGMENT
The authors are thankful to the their colleagues Professor
Dr. Iqbal Hussain and Associate Professor Dr. Aamir Khan of
Mechanical Department at the University of Lahore, Main
campus, 1-kM Raiwind Road, Lahore, Pakistan, for their
timely help and support.
REFERENCES
[1] Sun, X.S., V.B.C. Tan, and T.E. Tay, Micromechanics-based
progressive failure analysis of fibre-reinforced composites with non-
iterative element-failure method. Computers and Structures, 2011.
89(11-12): p. 1103-1116.
[2] Xu, L., et al., Prediction of material properties of biaxial and triaxial
braided textile composites. Journal of Composite Materials, 2012.
46(18): p. 2255-2270.
[3] Ha, S.K., et al., Micromechanics of Failure for Ultimate Strength
Predictions of Composite Laminates. Journal of Composite Materials,
2010. 44(20): p. 2347-2361.
[4] Jin, K.-K., et al., Distribution of Micro Stresses and Interfacial Tractions
in Unidirectional Composites. Journal of Composite Materials, 2008.
42(18): p. 1825-1849.
[5] Raghava, R., R.M. Caddell, and G.S.Y. Yeh, The macroscopic yield
behaviour of polymers. Journal of Materials Science, 1973. 8(2): p. 225-
232.
55
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-64-320.jpg)
![[6] Fish, J. and Q. Yu, Two-scale damage modeling of brittle composites.
Composites Science and Technology, 2001. 61(15): p. 2215-2222.
[7] Xia, Z., Y. Zhang, and F. Ellyin, A Unified Periodical Boundary
Conditions for Representative Volume Elements of Composites and
Applications. International Journal of Solids and Structures, 2003. 40(8):
p. 1907-1921.
[8] Hashimoto, M., et al., Prediction of tensile strength of discontinuous
carbon fiber/polypropylene composite with fiber orientation distribution.
Composites Part A: Applied Science and Manufacturing, 2012. 43(10):
p. 1791-1799.
[9] Roget, B. and I. Chopra, Wind-tunnel testing of rotor with individually
controlled trailing-edge flaps for vibration reduction. Journal of Aircraft,
2008. 45(3): p. 868-879.
56
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-65-320.jpg)
![Improving Physical and Mechanical Properties of
Medium Density Fibreboard (MDF)
Waheed Gul
Mechanical Engineering Department
CECOS University of IT & Emerging Sciences
Peshawar, Pakistan
waheed@cecos.edu.pk
Afzal Khan, Abdul Shakoor
Mechanical Engineering Department
University of Engineering & Technology
Peshawar, Pakistan
Abstract— Physical and mechanical properties play a vital
role in Medium Density Fiberboard (MDF) Manufacturing. Urea
Formaldehyde (UF) resins is easily soluble in water and act as a
binder with fibers. It has fast rate of reaction but has a poor
physical and mechanical properties. Hence, there is a need to
improve those properties of UF resin. The current research is
based to investigates and analyze the UF resin to improve its
physical and mechanical properties in manufacturing of MDF.
The Melamine resin and basic green (Malachite Green) 4-crystals
were introduced into UF resin, separately. In both cases, a
significant improvement in physical and mechanical properties of
UF resin was observed. Additionally, economic analysis was
carried out for both additives and it was concluded that
Melamine is not only economically feasible but also gives
superior mechanical and physical properties when compared to
Malachite green doped resin. The resulted MDF with Melamine
doped resin resulted in strong mechanical adhesion, and water
resistance Keywords—component; formatting; style; styling;
insert (key words)
I. INTRODUCTION
The Medium Density Fiberboard (MDF) is a wood product
processed under heat and pressure by using wood fibers or
other plant fibers as raw materials and application of phenol or
Urea-formaldehyde resin as binder. Other suitable additives
may be added to improve the property of the board.[1].The
density of MDF in production is generally control between
690 – 750 kg/m3.The performance index of MDF is divided
into three categories, i.e. Physical performance, Mechanical
performance and Biological performance.
By nominal density, MDF may be classified in to Type 60, 70,
80. By applicable Conditions, MDF can be classified in to
interior MDF, interior humidity-proof MDF and exterior
MDF.By visual quality and internal bonding strength, MDF
can be classified into Super Grade, Grade A and Grade B. [1].
A. MDF Production Process Flow
MDF production process consists of different stages, i.e.
materials preparation, Fiber separation, Fiber treatment, Mat
forming, Hot pressing, Board trimming and sanding, as shown
in MDF process flow chart in figure.1.
↓
↓
↓
↓
↓
Figure 1.1: Production Process flow of MDF [2].
B. Malachite Green
Basic green (Malachite green) is an organic compound,
basically green powder with Metallic luster. The molecular
formula of Malachite green is C52H54N4O12.
Properties of Malachite green are:
It is easily soluble in water
It is extremely soluble in Ethanol.
Its solution in water is Blue-green.
Basic green dyes become yellow in concentrated
sulphuric acid.
Figure 1.2: Basic Green 4 (Malachite green)
Fiber Separation
Hot Press
Board Trimming
Sanding
Fiber Treatment
Material Preparation
Greeen Crystals
57
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-66-320.jpg)
![II. LITERATURE REVIEW
Urea Formaldehyde (UF) resins is used as a binder with fibers
in manufacturing of Medium Density Fiberboard (MDF)
because of its fast rate of reaction, easily soluble in water and
also economically feasible. [3].The only problem is that of
poor water resistance property which limits the uses of MDF.
In order to overcome this problem, a melamine formaldehyde
resin and Malachite Green are added with UF resin. If we
compare the market price of both resin, then it can be easily
understood that Melamine formaldehyde resin is almost 4
times expensive than UF resin. So a very small quantity is to
be added for improving water resistance property. The
Melamine resin can easily soluble in both water and UF resin.
[4]. similarly, the market value of Malachite green is pretty
high. Malachite in our case is added to UF resin. The
Malachite is added for the reason to give a unique color and
extra strength to High Density Fiberboard. As we know that it
is easily solvable in water. So the formulation of Malachite
green is carried out in such a way that it dissolves water before
adding it into UF resin. In other words the solution of
Malachite green was added in definite proportion with water
along with caustic.
Malachite green is prepared in two steps: In ist steps
condensation of of benzaldyde and dimethylniline condensate
and leuco Malachite green in obtained. In second step leuco
compound is oxidize to give malachite green. [5].
𝐶6𝐻5𝐶𝐻𝑂+2𝐶6𝐻5𝑁(𝐶𝐻3)2→𝐶6𝐻5𝐶𝐻(𝐶6𝐻4𝑁(𝐶𝐻3)2)
2+𝐻2𝑂 (1)
𝐶6𝐻5𝐶𝐻(𝐶6𝐻4𝑁(𝐶𝐻3)2)2+𝐻𝐶𝑙+12𝑂2→[𝐶6𝐻5𝐶(𝐶6𝐻4𝑁(𝐶𝐻
3)2)2]𝐶𝑙+𝐻2𝑂 (2)
Manganese dioxide is used as oxidizing agent.
Malachite green can be Hydrolysis and give carbonyl formal.
[𝐶6𝐻5𝐶(𝐶6𝐻4𝑁(𝐶𝐻3)2)2]𝐶𝑙+𝐻2𝑂
→𝐶6𝐻5𝐶(𝑂𝐻)(𝐶6𝐻4𝑁(𝐶𝐻3)2)2+𝐻𝐶𝑙 (3)
Table 1: pH values of Malachite Green in Transition
Uses of Malachite Green:
Malachite green is used as a dye. It is used as antibacterial. In
Microscopic analysis, it is used as a-biological stain. In dye
laser it is used as a saturated absorber. It is also used as a PH
indicator between PH 0.2 – 1.8. [5].
III. MATERIALS AND METHODS
UF and Melamine formaldehyde resin and Malachite Urea
Formaldehyde resin have some properties, i.e. PH value, Gel
time, Viscosity, Specific gravity, Solid content etc.
In order to calculate the above properties the following
methods and procedures are used.
A. pH Value Calculation
There are two methods for calculating the PH value of the
Resin.
One way of measuring the PH value is the use of PH paper.
Procedure: The PH strip is immersed in UF resin for a minute.
Then the strip is taken out and compare the pH value with PH
scale (0- 14) numbers. The PH value of UF resin is from 7.5 –
8.5 means that the UF is alkaline.
Figure 3.1(a): A PH strip is immersed in Resin
In second method, a PH meter is used to measure the PH
(acidity or alkalinity) of UF. It consists of a glass electrode and
electronic meter as shown in fig. The electrode is connected to
the electronic meter that measure and display the PH value.
pH IndicatorUF Resin
Beaker
58
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-67-320.jpg)
![Table: 4.9: Comparison of 16mm board properties with standard values
Malachite
Figure 4.1: Comparasion of swelling in thickness of UF,MUF Malachite UF
with standard value
Figure 4.2: Comparison of internal Bonding of UF, MUF Malachite UF with
standard value
V. ECONOMIC ANALYSIS
As we know from formulation that:
Melamine consumption /sheet = 0.19 (kg/sheet)
Price of 1 Kg of Melamine = Rs. 150/Kg
Melamine cost per sheet = Rs = 28.5
Similarly,
Malachite Green consumption / sheet = 0.07 (Kg/sheet)
Price of 1 Kg of Malachite Green = Rs = 700 / Kg
Malachite cost per sheet = Rs = 49
Cost Difference = Rs. 20.5
For a 154 M³ capacity plant, Production/day = 3240 sheets
(16mm)
Cost saving /day = Rs. 66420
So it is clear from economic Analysis that Melamine is
also economically better than Malachite Green.
VI. CONCLUSION
In the light of the above parameters, samples of 16mm
board were manufactured using the predefined parameters.
The boards were then tested for their physical and mechanical
properties. The tested boards’ properties were then compared
with the international standards.
The 16mm boards samples using Malachite UF and MUF
resins were manufactured and properties were compared and it
was found that almost all properties were according to the
standards. Moreover, using MUF, the swelling in thickness
property was improved by 15%. And Mechanical adhesion
(which is called internal bond) was increased by 5 %.
Economic Analysis was also carried out and it was analyzed
that Melamine is economically better resin.
ACKNOWLEDGMENT
The author is highly thankful to Ciel Woodworks Pvt.ltd
Peshawar, Pakistan for giving him the opportunity to perform
experiments in lab and use the factory equipment.
REFERENCES
[1] ksh, A, and Hosseikhani, H, (2010), “Investigation on physical and
mechanical properties of Medium Density Fiberboard (MDF) produced
from hornbeam wood”, Volume 25, Number 1 (32); Page(s) 1 To10.
[2] Zwawi Ibrahim, Astimar Abdul Aziz, Ridzuan Ramli, Wan Hassamudin
Wan Hassan and Nahrul Hayawin Zianal (2011), “Optimum Parameters
for the Production of MDF using 100 % Oil Palm Trunks”, “Malaysian
Palm Oil Board, Ministry of Plantation industries and commodities,
Malaysia.ISSN, ppt.1511-7871.
[3] Awang Bono, Yeo Kiam Beng @ Abdul Noor & Kinabalu, Sabah,
(2001),”Melamine-Urea-Formaldehyde Resin: changes in Physical
Properties and Strength with Composition Molar Ratio”, Borneo Science
10: 11-23.
[4] Jizhi Zhang, Xiaomei, Wang, Shiefeng Zhang, Qiang Gao, Jianzhang Li
,(2013),”Effects of Melamine Addition stage on the Performance and
Curing Behavior of Melamine-Urea-Formaldehyde (MFU)
Resin,”,Bejing Forestry University, 100083, Vol 8, No 4.
[5] H. Zare-Hosseinabadi, M. Faezipour1, A. Jahan-Latibari2
andA.Enayati1, J. Agric. Sci. Techno, (2008),”Properties of Medium
S.N
o
Properties
Value
s
Units
EN
standard
s
Test
Method
1 Moisture content 8 % 7 ± 3 EN322
2 Density 725 kg/m3
720-740 EN323
3 Swelling in thickness 12.8 % ≤ 12 EN317
4 Internal Bonding 0.7 N/mm² ≥ 0.6 EN319
5 Modulus of Elasticity 3150 N/mm² 2500 EN310
6 Modulus of Rupture 31.9 N/mm² ≥ 30 EN310
7 Screw Holding face 1001 N ≥ 1000 ASTM1761
8 Screw Holding edge 739 N ≥ 700 ASTM1761
62
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-71-320.jpg)
![Density Fiberboard Made from Wet and Dry Stored Bagasse”, Vol. 10:
461-470.
[6] Amee K. Patel, Hardik H. Chaudhary, Khushbu S. Patel and Prof. Dr.
Dhrubo Jyoti Sen,(2014),” Colour of Ecofriendly Dyes Used In Holi
Rather Than Triphenylmethane Dyes”, Volume 3, Issue9, 1287-1305.
[7] Nadir Ayrilmis, (2008),”Effect of Comparison wood on Dimensions
Stability of Medium Density Fiberboard”, University of Istanbul Turkey,
42(2),285-293.
[8] Cristiane Inácio de CamposI, Francisco Antonio Rocco LahrII,
(2004),”Production and characterization of MDF using eucalyptus fibers
and castor oil-based polyurethane resin”, vol.7 no.3
[9] Stefan Ganev, (2003),”Linear Expansion and thickness Swell of MDF as
a Function of Panel Density andSorptionState,””,University
Laval,Quebic,Canada,GIK7P4.
[10] H. Zare-Hosseinabadi, M. Faezipour1, A. Jahan-Latibari2
andA.Enayati1, J. Agric. Sci. Techno, (2008),”Properties of Medium Density
Fiberboard Made from Wet and Dry Stored Bagasse”, Vol. 10: 461-470.
[11] Khalid Pervez Bhatti and Muhammad Zuber, (2009),” Synthesis and
Application of Melamine Urea Based Precondestates”, AUTEX
Research Journal, Vol. 9, No4.
[12] Maintains, G and Berns, J, (2001),”Strawboards bonded with urea
formaldehyde resins”, 35th International Particleboard Composite
Materials Symposium Proceedings, pp.137-144, 2001.
[13] http://www.goodrichsugar.com/mdfnew.asp?no=3, (April12, 2014.
63
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-72-320.jpg)
![Finite Element Analysis of Tool Wear in Ultrasonically
Assisted Turning
A. Rehman1
, S.Maqsood1
1
Dept.of Industrial Engineering,
University of Engineering and Technology,
Peshawar,Pakistan
Khattak_rehman@yahoo.com
R.Muhammad2
, N.Ahmed2
2
Dept.of Mechanical Engineering
CECOS University of IT and Emerging Sciences
Peshawar,Pakistan
Abstract—The tremendous increase in use of titanium and its
alloys in different industries, especially in power and aviation
industries demand further investigation due to its exceptional
mechanical and thermal properties. New assisted and hybrid
machining techniques were introduced to obtain better results in
machiningof these alloys. Among many, ultrasonically assisted
turning (UAT) is one of those in which lowervibration energy
impactisdelivered onthe cutting tool in the direction of velocity.
DuringUAT of titanium alloys, momentous improvement has
been noticed in surface roughness smoothness along with
reduction of overall cutting forces, but tool wear behavior is still
unknown. This paper presents a parametric finite element (FE)
analysis for describing tool wear behavior in UAT of (Ti-
15V3Al3Cr3Sn) alloy on the selected cutting conditions. The
model is useful for optimizing of cutting parameters in UAT for
Ti-15V3Al3Cr3Sn. The proposed FE model is also enhances the
recent experimental work on wear behavior relevant to cemented
carbide tool and comparing UAT and conventional turning (CT)
of Ti-15V3Al3Cr3Sn results.
IndexTerms—Tool Wear, Ultrasonically Assisted Turning, Ti
Alloys
I. INTRODUCTION
Machining is an imperative practice in manufacturing and
production industries. Machining put in a substantial portion to
the total cost of the product. Machining cost and production
time are the key apprehensions caused by the frequent abrasion
and replacement of machining tool. Compare to milling,
drilling[1, 2]and grinding processes, turning process is
receiving more importance due to enhanced production rate in
advance industries[3, 4].
Uses of Titanium and its alloys are tremendously increasing
day by day in different industries, especially in power,
petroleum, aviation and biomedical industries[5]. In many
applications, these materials replace steels and aluminum
alloys, which usually results in weight and/or space saving,
increase of system efficiency by rising the service temperature,
and remove the need of protective coatings that should be used
in steels [6]. Increment in the use of these alloys is due to its
outstanding mechanical properties including elevated strength
to density ratio, corrosion resistance, fatigue properties and
strength behavior at moderately high temperatures [7].
Aforesaid materials are therefore fall in a category of difficult-
to-cut materials, because of the fact that these materials are
facing greater challenge to machining processes against the
higher stresses generation and elevated
temperatures[8].Presently, beta titanium is considered as the
most difficult to machine alloys[9].Keeping in view the high
demand of titanium and its alloys machining, researchers have
continuously worked and developed different techniques for
improvement of machinability of these most hard-to-cut
alloys[10]. Such alloys are difficult to machine using
conventional turning(CT) processes and often results in poor
surface finish with associated lack of dimensional accuracy,
built-edge formations, fast tool wear, and undesired chatter
during operations[11].
In recent, new assisted and hybrid machining techniques have
been introduced to get improved machining results of these
hard-to-cut alloys. Among many, ultrasonically assisted
turning (UAT) is one of those in which
lowervibrationalenergy impacts are imposed on the tool in the
direction of velocity[12-14]. This technique has shown
tremendous and noteworthyenhancement in the machinability
of hard-to-cut alloys [15-18]. In UAT, the tool and workpiece
are not in continuous contact that is why net forces are lower
as compared to CT [12], ultimately friction influence are also
lower [13, 14]. The efficiency of UAT and influence of
vibrational parameters on surface roughness is investigated by
Graeviciuteet. Al[15, 16]. However, limited attention has been
given to investigate the tool life in UAT which is the main
concern in manufacturing processes.
Two of the main critical problems are tool wear and tool
failure, which not only raises the production cost but also
affect the product quality. The thorough and continuous
contact between cutting part of tool and workpiece material
generate high cutting forces and temperature in the cutting
region of hard-to-cut alloys. The selection of appropriate tools
for the machining of these alloys using sustainable
conventional machining processes is still not fully understood
and demands for further research to advance the better tool life
and tool wear.
The subject matter of this work is based on the finite element
simulation of UAT of beta titanium alloy to investigate the
tool life under pre-selected cutting conditions. The simulation
results have been compared with CT. This work would help
the industries to estimate tool’s life in UAT.
64
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-73-320.jpg)
![II. 3D MODELLING
A 3D model for both CT and UAT processes is developed in
DEFORM 3D V 6.1.In current model cutting tool of
Fig.1. Tool and workpiece FE mesh
Tungsten Carbide with (AlTiN) coating is assumed to be
elastoplastic with considering a rake angle as 14.6º, nose angle
as55 º, nose radius as0.8 mm and cutting-edge radius as0.8
mm. Tool is then meshed with more than 45,000 elements and
is oriented according to the setup proposed. Meshing of tool
and workpiece can be seen in Fig. 1.
Workpiece of Ti-15333is assumed to be plastic and is fully
constrained on lower and lateral side so that it cannot move.
Workpiece used in the simulation is5mm long, fine meshing
been considered with the minimum element size of 0.025 mm
and the number of elements aremore than 30,000. Localized
window meshing has beenselected for the workpiece meshing
to obtain better results. 0.3 mm depth of cut and 0.1
mm/revfeed rate are initially considered for all simulations.
The current model is based on Langrangian methodin which
the chip is formed by continuous remeshing.
The tool is immovable for the simulation cases of CT, whereas
it vibrates harmonically for the simulations of UAT. The tool
has given an ultrasonic vibration with 20 kHz frequencyand
8µm of amplitude in the direction of cutting velocity.Figure. 2
clearly elaborate theschematic of tool and workpiece
movements.
Fig.2. Schematic of tool and workpiece movements
TABLE.1 Properties of cutting tool
Young’s modulus, (Mpa) 560*10-3
Thermal conductivity, (Wm-1°C-1) 0.0081*T+11.95
Thermal expansion, (mm mm-1°C-1) 9.4*10-6
Heat capacity, (Nmm-2C-1) 0.0003*T+0.57
Poisson's ratio 0.25
A. Cutting Tool
In the present paper, coated carbide insert (AlTiN) is being
used as a tool to machine titanium alloy (Ti-15333). This
coating provides reduction of fraction and adhesion between
chip and cutting tool. In addition to that, AlTiNat high
temperature has good oxidation resistance, wear resistance and
elevatedchemical stability because of the formation of Al2O3
film. Cutting tool mechanical properties are represented in
Table.1. [18]
B. Workpiece Material
Titanium alloy (Ti-15333) used in our simulation belongs
to the family of beta Ti alloys.These alloys fall in category of
difficult to machine titanium alloys because of the significant
precipitation hardening characteristics. Table.2 and Table.
3[19] shows the composition along with the mechanical
properties of Ti -15333 .
TABLE.2 Ti -15333 (%) chemical composition
Ti V Al Cr Sn
76 15 3 3 3
TABLE.3 Ti-15333Material properties
Parameter Unit value
Solution treated and aged
Density, ρ (kg/m3
) 4900
Young’s modulus, E (GPa) 87
Tensile strength, UTS (Mpa) 1200
Thermal conductivity, k (W/Km) 8.08
Hardness (Rockwell B) 95
65
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-74-320.jpg)
![Metamaterials in Aerospace Industry: Recent
Advances and Prospects
Rabia Zafar, Saima Shabbir
Institute of Space Technology, Islamabad,
Pakistan
drsaimashabbir@gmail.com
rabiazafar38@gmail.com
Azeem Zafar
National University of Science and Technology, Islamabad,
Pakistan
azeemgeologist@gmail.com
Abstract— This article offers a review of the existing
metamaterials and their concepts, recent advances, and research
directions. It also describes applications of metamaterials design
in aerospace and defense sector as an emerging technology.
Acoustic metamaterials and metamaterial based antenna present
advantages over conventional noise absorbers and antenna
technology employed in aerospace components. Gain
enhancement of antennas and efficiency of structures are
described in this article with reference to improvements and
advancements made so far in academia. A rational approach to
metamaterials science and engineering can revolutionize
aerospace design and render better prospects for military
applications.
Keywords—Acoustics, Aerospace, Metamaterials
I. INTRODUCTION
For the past few decades research in metamaterials is
gaining an increasing interest [1, 2] due to counterintuitive
electromagnetic properties incorporated into material
structure. Principally, Metamaterials are left handed
materials which exhibit negative refractive index, unusual
diamagnetic properties, with fractional effective
permeability and a very large in-plane effective permittivity
[3, 4]. The permeability can be attributed to the Eddy
current phenomenon induced in conductive inclusions and
permittivity is attributed to an increased value of dielectric
constant.
For a material to be optically transparent, the incident
wavelength will be refracted off centre the superficial
surface. The angle of refraction is smaller than angle of
incident as demonstrated in Fig. 1. The velocity of incident
wavelength is greater in space than the velocity of
transmitted wavelength [5] and object will appear
transparent.
Metamaterials find immense applications in the field of
science and engineering including wireless communication,
bio sensing, seismology, electromagnetic devices and perfect
lenses [6]. Moreover, metamaterial absorbers and emitters,
super lenses and cloaking devices for military camouflage are
under high consideration.
A. Development history
Developmental history of metamaterials began by 20th
century in 1904, when experiments on negative group
velocities were carried out by Pocklington [7] and progressed
through the mid of 20th
century contributed by distinguished
scientists [8, 9]. John Pendry in 1990’s made significant
contributions in the development of metamaterials theory [6].
Metamaterials do not exist naturally with the exception of
metallic element Bismuth. Plasmas and noble metals show this
behavior in infra red or visible spectrum [10].
B. Existing metamaterials
Metamaterials cover a narrow range of wavelength owing
to the fact that they can respond below red region of visible
wavelength. Based upon the medium properties, metamaterials
can be classified as;
• Negative index metamaterials (NIM) which include
double negative (DNG- ε and µ are negative) and
single negative metamaterials (SNG-either ε or µ is
negative),
• Double positive (DPS- ε and µ are positive),
• Epsilon negative (ENG- ε is negative and µ is
positive) and µ negative (MNG ε is positive and µ is
negative). MNGs show magneto optic effect in which
electromagnetic wave propagation takes place in a
quazistatic altered magnetic field [11].
Based on the response to a particular wavelength
metamaterials are further classified as photonic metamaterials,
acoustic metamaterials, seismic metamaterials,
electromagnetic metamaterials, terahertz metamaterials,
tunable metamaterials, elastic metamaterials, chiral
metamaterials and quantum metamaterials.
69
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-78-320.jpg)
![Fig. 1. Ray diagram of a metamaterial and a conventional material
II. MICRO FABRICATION AND APPLICATIONS
Metamaterials industry is emerging as a leading
technology in future for defense and aerospace sectors by
2020. Currently, there are only few methods available for the
fabrication and characterization which involve complex
principles of microelectronics and lithographic deposition.
The characterization of metamaterials is rather expensive,
which poses high need for capital investment in
instrumentation. Due to these limitations there are only a few
metamaterials which show negative effective permeability and
large permittivity and successfully characterized as
metamaterials.
A. Polymer based metamaterials
Metamaterials are composed of polymer matrix filled with,
copper, gold or silver flakes. They are formed by stacking
alternating layers of ENGs/MNGs type of metamaterials
separated by space.
Metamaterials for interference shielding applications were
fabricated by pulsed current electrolysis routes and by
carefully controlling the concentration of copper metal flakes
as inclusions in polymer matrix composites [12].
B. Metamaterial coatings
Metamaterials multipole coatings are being applied by
carefully tailoring the surface properties ranging from the
lowest ordered multipole coatings for microscopic bodies to
higher ordered multipoles for macroscopic bodies. Concealing
a macroscopic body becomes more difficult as domain size
increases and concealing the particles in a closed proximity is
difficult to achieve practically.
C. Metamaterial emitters
In a single band metamaterial a noble metal such as gold or
silver is deposited by electron beam lithography on a silicon
substrate followed by curing the spun coated
benzocyclobutene on metallic top. Dual band absorber is
fabricated by lithographic deposition with only difference that
the dielectric space is occupied by Al2O3. Such emitters find
applications in energy harvesting devices [13].
D. Metamaterials and seismic cloak
The seismic wave guided metamaterials can be made by
stacking sequences of macromolecular polymer based split
ring resonators overlaid with other meta atoms in a
consecutive fashion. Such a metamaterial can be used to
absorb seismic waves generated by earth quake and can
efficiently refract back or absorb seismic waves protecting
building infrastructure.
E. Stealth applications
Implementation of metamaterials in stealth technology
finds its use to disguise moving bodies from radar detection;
either by absorbing or by scattering all the radiations from the
radar detection system.
F. Acoustic cloaking
Acoustic metamaterials are ordered artificial structures
which can absorb and manipulate sonic, infrasonic and
ultrasonic waves at will [14]. Bulk modulus and mass density
counteract with permittivity and permeability. This analogous
property is under high consideration and research is being
done to direct and control sound waves at will by tailoring
these two parameters. 2D acoustic cloaking has been
optimized by placing five arrays of cylindrical lattices along a
common central axis to achieve omnidirectional cloaking
devices.
G. Metamaterials antenna gain
Metamaterials are being extensively studied to find
applications in wireless communication and improved antenna
performance i.e. gain enhancement, miniaturization of size,
directivity and bandwidth broadening. Antenna is dispersive in
nature and acts as an electrically small resonator. In a
conventional antenna surface waves propagate to increase
cross polarization effects, low gain, narrow bandwidth and
frequency range.
The quest continues as how to improve the output of
conventional antenna by incorporating metamaterial and
which particular type of metamaterial can improve one or
more parameters? An important consideration to achieve
antenna gain is to incorporate double negative DNG
metamaterials and single negative metamaterials into antenna
structure to harvest efficient yield and enhanced bandwidth.
SNGs are dispersive in nature [15]. Using this principle
behavior SNGs can be alternated with either MNG layer or
ENG layer to get a suitable combination of optical properties,
enhanced dispersion and miniature antenna size.
Finally, to reduce surface wave propagation it is necessary
to minimize the size of antenna and for this purpose photonic
band gap metamaterials can be employed.
III. Discussion and outlook
The demonstration of recent advancements and analysis of
metamaterials with all possible applications and future
prospects in both academia and industry is interesting. Various
antenna parameters have been studied and improved by
Metamaterial
Conventional
materials
Incident ray
70
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-79-320.jpg)
![different researchers such as gain enhancement, dispersion,
and bandwidth broadening and miniaturizing the size. Studies
revealed that unit cell metamaterials of omega structure can be
combined to form a slab and used to cover antenna surface
and increases directive gain to 11.5 dBi [16].
Metamaterial based broadband antenna were demonstrated
to show -1 dBi antenna gain at frequency of 2.5 GHz. with -1
dBi [17]. Electrically small antenna were demonstrated by
combining layers of ENG metamaterials alternated with
ordered multipole resonant structures to increase radiation
efficiency and high impedance [18].
Metamaterials show exceptional physical properties which
are overwhelming from both theoretical and academic
perspective and definitely find relevance to aerospace and
defense applications.
A. Metamaterial airborne design
In any purposeful review the electromagnetic properties of
metamaterials, which strongly impinge airborne radome
design cannot be neglected. Overall antenna performance is
strongly dependent on the mechanical and electromagnetic
properties of dielectrics. Unfortunately, there are limited
number of materials which compromise over low dielectric
constant and high mechanical strength. Significant research in
the radome design is necessary in both spheres of
metamaterials design and electromagnetism to circumvent
radome, which may pose restrictions in metamaterials airborne
design in near future.
An outlook of current projects shows that the developed
nations are involved in understanding the metamaterial
concepts and their practical applications. For example, BAE
systems has tested metamaterial based invisible vehicles
which can disguise in infra red and microwave frequency
range. Other international research groups are aiming at
developing nano sized plasmonic devices [19].
U.S. Army Research office in collaboration with university
students is developing liquid crystals tested in the range of 1-
110GHz. Other tunable devices made by liquid crystals are in
the process of development such as tunable antenna [20].
IV. FUTURE CHALLENGES
To address the future challenges metamaterials research
must transcend the theoretical investigations and transform
into practice. Practical applications make transitions through
terahertz (THz), infra red and visible region of
electromagnetic spectrum. THz metamaterials can be tailored
by changing meta atoms at nanoscale and open new horizons
of understanding and research in photonics and electronic
band gap materials. Challenges related to efficiently
generating and detecting broadband THz waves has primarily
limited their use. Currently existing metamaterials respond to
a narrow band of EM spectrum below red region of visible
spectrum. This requires the development of metamaterials to
extend their response in visible, micro and radio wave region
of EM spectrum in both theory and practice. Thus
metamaterials design can miniaturize antenna design [22] and
dispersion coupled with improved satellite communication and
can revolutionize the defense system.
V. SUMMARY
In this review we have taken a brief outlook of existing
metamaterials, their fabrication routes, current projects and
ongoing research to enhance performance and applicability of
metamaterial devices around the world. Survey of
metamaterials clearly demonstrated that metamaterials can be
used for improving the performance of conventional antennas
and can revolutionize the communication system.
Additionally, a brief review of the current work and
development in the field of metamaterials is provided. We also
examined the on-going projects on metamaterial to enhance
the performance of antennas and various novel metamaterials.
Various future challenges are also considered.
ACKNOWLEDGMENT
Authors would like to acknowledge Department of
Materials Science and Engineering, Institute of Space
Technology, Islamabad for encouragement and moral support
to write this review paper.
REFERENCES
[1] A. J. Hoffman, L. Alekseyev, S. S. Howard, K. J.
Franz, D. Wasserman, V. A. Podolskiy, E. E.
Narimanov, D. L. Sivco, and C. Gmach, “Negative
refraction in semiconductor metamaterials,” Nat.
Mater, vol. 6, pp. 946-950, October 2007.
[2] N. Fang, D. Xi, J. Xu, M. Ambati, W. Srituravanich,
C. Sun, and X. Zhang, “Ultrasonic metamaterials
with negative modulus,” Nat. Mater, vol. 5, pp. 452-
456, April 2006.
[3] R. A. Shelby, D. R. Smith, S. Shultz, and S. C.
Nemat-Nasser, "Microwave transmission through a
two-dimensional, isotropic, left-handed
metamaterial," Appl. Phys. Lett. vol. 78, pp. 489-491,
January 2001.
[4] D. R. Smith, W. J. Padilla, D. Vier, S. C. Nemat-
Nasser, and S. Schultz, "Composite Medium with
Simultaneously Negative Permeability and
Permittivity". Phys. Rev. Lett, vol. 84, pp. 4184-
4187, May. 2000.
[5] K. Sanderson, “Materials science: Unexpected tricks
of the light,” Nature, vol. 446, pp. 364-365, March
2007.
[6] J. B. Pendry, “Negative refraction makes perfect
lens,” Phys. Rev. Lett., vol. 85, pp. 3966-3969,
October 2000.
[7] H. C. Pocklington, “Growth of a Wave-group when
the Group-velocity is Negative”, Nature, vol. 71, pp.
607-608, April 1905.
[8] L. I. Mandel’shtam, “Group velocity in a crystal
lattice,” Zh. Eksp. Teor. Fiz, vol. 15, pp. 475-478,
April 1945.
[9] V. G. Veselago, “The electrodynamics of substances
with simultaneously negative values of ε and µ,” Sov.
Phys., Usp, vol.10, pp. 509-514, February 1968.
[10]V. A. Podolskiy, L. Alekseev, and E. E. Narimanov,
“Strongly anisotropic media: The THz perspectives
71
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-80-320.jpg)
![of left-handed materials,” J. Mod. Opt, vol. 52, pp.
2343-2349, January 2005.
[11]Z. Qiu, and S. D. Bader, “Surface Magneto-Optic
Kerr Effect. Characterization of Materials,” 2nd ed.,
John Wiley & Sons, 2012, pp.1-10.
[12]P. Los, A. Lukomska, S. Kowalska, R. Jeziorska, P.
Zaprzalski, and J. Krupka, “Metamaterials based on
polymer dispersions of nanoparticles and particles of
copper obtained by cathodic current pulse
electrolysis.” Materials Science-Poland, vol. 29, pp.
35-40, March 2011.
[13]X. Liu, T. Tyler, T. Starr, A. F. Starr, N. M. Jokerst,
and W. J. Padilla, “Taming the Blackbody with
Infrared Metamaterials as Selective Thermal
Emitters,” vol. 107, pp. 045901-4, July 2011.
[14]C. Jo, J. Jeong, B. J. Kwon, K. C. Park, and I. K. Oh,
“Omnidirectional two-dimensional acoustic cloak by
axisymmetric cylindrical lattices,” Wave Motion, vol.
54, , pp. 157-169, April 2015.
[15]G. V. Eleftheriades, and K. G. Balmain, “Negative-
refraction metamaterials: fundamental principles and
applications.” 1st ed., Wiley-IEEE Press. 2005.
[16]A. S. Ahmad, O. Adib, J. J. Mohd, H. B. Noor, A. A.
Robiatun, and F. F. Mohd, “Small patch antenna on
omega structure met material,” Eur. J. Sci. Res. vol.
43, pp. 527-537, June 2010.
[17]M. Palandoken, A. Grede, and H. Henke, “Broadband
Microstrip Antenna with Left-Handed
Metamaterials,” IEEE Trans. Antennas Propag, vol.
57, February 2009.
[18]A. Erentok, and R. W. Ziolkowski, “Metamaterial
inspired Efficient Electrically Small Antennas,” IEEE
Trans. Antennas Propag, vol. 56, pp. 691-707, March
2008.
[19]J. B. Pendry, A. J. Holden, D. J. Robbins, and W. J.
Stewart, “Low frequency plasmons in thin wire
structures,” J. Phys. Condens. Matter, vol. 10, pp.
4785, June 1998.
[20]D. R. Smith, “Research group- Novel,
electromagnetic materials program,” January 2005.
72
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-81-320.jpg)
![Finite Element Simulation of Composite Body Armor
Numan Khan1
, Abdul Shakoor1
1
Dept. of Mechanical Engineering
University of Engineering and Technology,
Peshawar, Pakistan
Write2numan@gmail.com
Riaz Muhammad2
, Naseer Ahmed2
2
Dept. of Mechanical Engineering
CECOS University of IT & Emerging Sciences,
Peshawar, Pakistan.
Vadim V. Silberschmidt3
3
School of Mechanical and Manufacturing Engineering,
Loughborough University,
Loughborough, UK
Abstract—Finite element simulation of composite body armor
for impact loading conditions is presented in this paper. The
7.62 mm armor piercing projectile with the velocity of 800 m/s
is used for impacting on the target material. The numerical
model is developed in ANSYS AUTODYN using Johnson–
Cook material model for projectile and backing steel material.
The Johnson–Holmquist material model is used for ceramic
material. Different simulations are performed for different
configurations of composite materials and the results of
projectile velocity, kinetic energy, internal energy of target
plate and perforation time is obtained. Finite element
simulation of 6 mm ceramic backed by 2 mm steel plate has
shown full penetration of bullet while the other backed by 4, 6
and 8 mm respectively has prevented complete penetration.
The numerical results show that ceramic material is
responsible for breaking projectile tip and absorbing most of
the kinetic energy. The drop of about 66% in projectile kinetic
energy is observed with increase of backing steel by 2 mm.
Finally optimum ceramic/steel material thickness ratio of 1.5
with light weight is recommended for single body armor.
Keywords- Numerical simulation; Ballistic impact;
Projectile; Armor plate; Johnson–Cook material model;
Johnson–Holmquist model
I. INTRODUCTION
Ever since the discovery of metals and alloys, different type
of protection is made against the enemies attack for
arresting high-velocity projectiles impact by mankind [1, 2].
The development of light weight and effective body armor
is the need of the day due to the current law and order
situation in the world and particularly in Pakistan. For the
mobility of solders light weight efficient and cost effective
body armor is necessary to be developed. Most of the
ballistic materials shows nonlinear deformation and
dynamic behavior under impact loading [3]. Material
properties like strain, strain rate, materials hardening and
thermal softening play a vital role in the design process [2].
With the passage of time different researchers have
developed various ballistic resistive materials.
Normally single high strength metallic plates are used for
body armors. However, multi layered plate configuration
are often used to meet the design specifications which
cannot be fulfilled by metallic single plate. Zukas et al. [4]
studied that a thick uniform steel plate was more resistant
to hemispherical projectile than composite multi layered
target of identical thickness due to free slip between plates.
It was also observed that by increasing the number of plates
in multi layered target plate reduces its resistance, which is
much closed to the data obtained from the experiments of
Almohandes et al. [5]. A detailed analysis of ballistic
performance of composite and hybrid armor plates were
performed using ANSYS AUTODYN software by
Yohannes Regassa et al. [6]. Ceramic faced armor plate
was used for observation of critical penetration velocity,
which shows good agreement with experimental results.
Teng et al. [7] has presented the performance of composite
and single layered steel armor under ballistic conditions and
concluded that nose shape, velocity and mass of projectile
are the key parameters effecting body armor performance.
Similarly, the effect of nose shape and hardness on failure
mechanism of the armor is presented in [8]. Studies of H.
Kurtaran et al [9]. has shown full penetration of
semispherical nose shape projectile through 2mm thick steel
plate at velocity of 500m/s. However the backing of front
plate with a 20 mm thick plate of same material and two
plates at different orientation angle has prevented full
penetration. it was further observed that the results of
Johnson-Cook Material model are more accurate than the
plastic kinematic model for evaluation of the fracture
damage and thermal softening [9]. Various analytical
models have been developed by S. Feli [10]. To analyze the
performance of Alumina plate backed by composite
material made of Twaron fibers. During impact the
fragmentation of ceramic conoid from ceramic tile and
decrease in the semi-angle of ceramic decreases with
increase of initial velocity was absorbed. The research for
design of light weight and efficient body armor required
more efforts for further improvement to increase its safety,
performance and cost effectiveness.
The current study demonstrates the finite element
simulation of projectile piercing ceramic steel target armor.
For fragmentation, the ceramic plate is modeled with
Johnson-Holmquist model where brittle failure, strain rate
effect and enormous deformation are considered. While
backing steel is modeled using Johnson-cook model in
which strain rate hardening effects, thermal softening and
materials fracture are reflected [9].
73
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-82-320.jpg)
![II. FINITE ELEMENT MODELING
The performance of composite body armor under impact
loading is investigated by explicit finite element model
using ANSYS AUTODYN [11, 12]. To reduce the
computational cost a small circular element of diameter 60
mm is considered for analysis. Initially the thickness of the
target plate is kept constant to 6 mm whereas actual bullet
dimensions are used for bullet modeling in the current
analysis. For addressing the increase of damage effect of
bullet on target plate with decrees in distance, these
simulations are carried out at minimum possible distance.
Similarly for simulation of the actual phenomenon of AK-
47 bullet with velocity range of 720 to 750 m/s, the
projectile is fired with a velocity of 800 m/s and its
rotational speed is neglected in the current study. Finite
element model for different configurations of ceramic and
steel are modeled as shown in Table 2. The thickness of the
front ceramic material is kept constant to 6 mm while the
4340 steel of different thickness 2, 4, 6 and 8 mm are used
as a backing material. To complete the perforation of the
plate the total time of simulation is considered as 50 µs. the
3D model of target armor and 7.62 mm diameter projectile
are shown in the Fig. 1. In current study the simulations are
carried out on core i3 PC with 2.2 GHz processor. The time
taken by one simulation is about between 8 to 12 hours of
CPU time.
A. Finite element mesh
A 3-D finite element mesh for both target plate and
projectile is generated by using automatic mesh generation
feature in ANSYS workbench. As can be seen form Fig. 2
the mesh of inner region of plate is kept finer than outer
region. In the mesh the maximum aspect ratio is kept below
five. Both the bullet and plate are meshed with hexagonal
elements of different size between 0.5 to 1.5 mm. The total
number of elements for the projectile and armor are 148,000
including, 52,000 elements for projectile and 96,000 for
armor plate.
B. Material model
Based on the material characteristics different material
models are used for different materials in bullet impact
simulation [1] In the current study, Johnson-Holmquist
model represented by equation 1 [13] is used for the
ceramic materials. This model is representing relationship
between the intact and fractured strength, pressure and
volume. The various parameter of the model are presented
in Table.1.
[ ( ) ][ ] (1)
Where A, B, n, C are material constants and is
representing the normalized strength of material.
is the normalized pressure. The damage D can be represented
by equation 2.
∑
( )
(2)
Here the equivalent plastic strain increment and
is the increment in the volumetric compaction strain.
Fig. 1. 3D model of projectile and target armor plate
Fig. 2. The FE mesh details of projectile and target armor plate
Similarly, Johnson-cook material model is used for the steel
part of the armor. The main feature of the Johnson Cook
model is representation[14-16] of the yield strength of
materials on high strain, strain rate, temperature and material
hardening, and is represented by the equation 3 [17]:
( )( )( - ) (3)
Where A, B, C, n and m are constants and is representing
value of effective stress, is effective plastic strain, is
normalized effective plastic strain
The term can be defined as:
(4)
In which and are used for melting and room
temperatures, respectively. Similarly, the damage of element
in Johnson-cook failure model is given by
∑ (5)
Where shows the change of equivalent plastic strain and
is the strain at fracture. The term can be represented by
the equation 6.
[ ( )][ ][ ] ( )
74
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-83-320.jpg)
![Table. 1 Strength and failure data for SiC using Johnson-Holmquist model
[18]
Parameters Units Values
Density (g/cm3
) 3.215
Equation of State Data (Linear)
Bulk Modulus-A1 GPa 220
Strength Data
Modulus of Rigidity GPa 193
Elastic Limit (Hugoniot) GPa 11.7
Intact Strength Constant-S1 GPa 7.1
Intact Strength Constant-P1 GPa 2.5
Intact Strength Constant-S2 GPa 12.2
Intact Strength Constant-P2 GPa 10
Constant of Strain Rate-C None 0.009
Fracture Strength (Max)-Sfmax GPa 1.3
Constant of Failed Strength-σ None 0.4
Failure Data
Tensile Limit-T GPa -0.75
Constant of Damage- Efmax None 1.2
Constant of Damage- P3 GPa 99.75
Constant of Buckling-β None 1
The fracture constants are expressed by D1, D2, D3, D4
and D5. For describing the pressure volume relationships
Mie–Gruneisen equation of state is used with combination of
Johnson-cook model [9]. Constant for Johnson-cook model
and Mie–Gruneisen equation of state model are given in
Table 3.
III. RESULTS AND DISCUSSION
Results of finite element simulation for projectile velocity of
800 m/s and different configurations as shown in Table.2 are
discussed in this section.
The variation in bullet velocity during penetrations for
different configurations is shown in Fig. 3. The config 1, 2
and 3 has maximum residual velocities of about 580 m/s,
230 m/s and 80 m/s respectively after impact while in config
4 projectile has been stopped and velocity is reached to 0 m/s
at 50 µs. this shows that the percent reduction in projectile
velocity is about 60% with the increase of backing material
thickness by 2 mm. this reduction is mainly due to the
impact of projectile on ceramic front material.
Table. 2 Composition and areal density for configuration of ceramic and
4340 steel plate
Configuration Front
SiC plate
Backing steel
plate
Areal density
(kg/m2
)
Config 1 6 mm 2 mm 34.99
Config 2 6 mm 4 mm 50.69
Config 3 6 mm 6 mm 66.39
Config 4 6 mm 8 mm 82.09
Fig. 3. Variation in projectile velocity
In the same way the Fig .4 gives the dissipation of kinetic
energy for config 1 to 4 after impact of projectile. It is
observed that maximum amount of kinetic energy is
dissipated in ceramic fragmentation and the rest of residual
kinetic energy is observed by the backing steel plate. The
config 4 dissipates maximum amount of energy as compared
to others as it goes through ceramic material. The percent
reduction in kinetic energy is about 66 % with increase of
thickness by 2 mm of backing material. This discloses that
the backing steel plate is responsible for absorption of
residual kinetic energy of the projectile and stop penetration.
Fig. 4. Projectile kinetic energy
The variation in internal energy of target armor plate can be
observed form Fig .5 which shows the increase for different
configurations. The Fig. 5a to Fig. 5d reveals that in all cases
the maximum amount of increase in internal energy of
ceramic plate is occurred. It is also observed that the increase
in internal energy of ceramic front plate is about 20 % with
increase of backing plate thickness by 2 mm. Similarly the
maximum increase in average internal energy is observed in
config 4 as shown in the Fig. 5. Thus by increase of backing
75
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-84-320.jpg)
![plate thickness maximum amount of kinetic can be absorbed in target armor plate. .
Table.3 Johnson- Cook model constants for 4340 steel[19]
Density, ρ (g/cm3
) Elasticity modulus, E (MPa)
Strength / Damage constants
A
(MPa)
B
(MPa)
N c m
7.83 210000 910 586 0.26 0.014 1.03
Failure Constants
D1 D2 D3 D4 D5
-0.8 2.1 -0.5 0.002 0.61
Mie–Gruneisen equation of state constants
Γ0 C1 (m/s)
1.6 3574
S1
1.91
Ref.
Temp
(K)
293
Fig. 5. Change Internal Energy for (a) Config 1, (b) Config 2, (c) Config 3, (d) Config 4
76
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-85-320.jpg)
![Similarly Fig. 6. shows the bullet penetration for
configuration 1 to 4. The config 1 shows full penetration of
bullet at velocity of 800 m/s whereas in config 2 the
penetration is stopped moderately. However, the bullet is
fully stopped by increasing the backing material thickness as
shown in config 3 and 4 in Fig. 6. Throughout the
simulations the damage mechanisms are observed in the
armor materials.
Firstly, Silicon carbide material which is very brittle tends to
shatter upon impact. It is also observed in the simulations
that the cause of failure of silicon carbide is tensile stress
wave that passes through the ceramic. Secondly, the tip of
the projectile become blunt upon impact with ceramic
material and adopted the shape like mushroom as shown in
Fig. 7. The distortion in projectile shape is increasing with in
increase in thickness as clear form Fig. 7. Another important
point is the deceleration of projectile upon advancement in
ceramic material. This reveals the fact that ceramic material
is contributing mainly in the breaking of projectile tip and
kinetic energy absorption. Finally the areal density for
different configuration is listed. Areal density can found by
adding the multiplication of the density and thickness values
of ceramic and backing steel material [1]. Upon comparison
with steel body armor of thickness 25.4 mm, about 100 kg
per square meter can be reduced. Thus reducing the total
weight can greatly improve the mobility of soldiers.
Config1Config2Config3Config4
t=0 µs t=10 µs t=20 µs t=30 µs t=40 µs t=50 µs
Fig. 6. Bullet penetration for different configurations at different time
77
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-86-320.jpg)
![Fig. 7. Bullet shape at 50 µs after impact on target armor plate
IV. CONCLUSIONS
In this work light weight ceramic composite body armor
is analyzed for impact loading of AK 47. A composite
material composed of Silicon Carbide and 4340 steel is
suggested and is simulated for various thicknesses. Based on
the simulation results, maximum reduction in kinetic energy
and increase in internal energy with increase of backing plate
thickness, the final optimum thickness ratio of 1.5 for
ceramic/4340 steel configuration is suggested.
It was also concluded that by reducing the thickness of
the backing materials can significantly reduce the weight of
the body armor, improve strength and will make it easy for
the security forces to carry it. As clear form table. 2 the
weight of about 15 kg per square meter can be saved if 2 mm
less thickness is used for backing material.
V. ACKNOWLEDGMENT
The authors would like to gratefully acknowledge Dr.
Rizwan Alim Mufti (Head of the Department) Pakistan
Institute of Engineering and Applied Sciences (PIEAS) for
his help in simulations.
VI. REFERENCES
[1]. Grujicic, M., et al., A computational analysis of the
ballistic performance of light-weight hybrid composite
armors. Applied Surface Science, 2006. 253(2): p. 730-
745.
[2]. Lee, M. and Y. Yoo, Analysis of ceramic/metal armour
systems. International Journal of Impact Engineering,
2001. 25(9): p. 819-829.
[3]. Børvik, T., S. Dey, and A. Clausen, Perforation
resistance of five different high-strength steel plates
subjected to small-arms projectiles. International Journal
of Impact Engineering, 2009. 36(7): p. 948-964.
[4]. Zukas, J.A. and D.R. Scheffler, Impact effects in
multilayered plates. International Journal of Solids and
Structures, 2001. 38(19): p. 3321-3328.
[5]. Almohandes, A.A., M.S. Abdel-Kader, and A.M.
Eleiche, Experimental investigation of the ballistic
resistance of steel-fiberglass reinforced polyester
laminated plates. Composites Part B: Engineering, 1996.
27(5): p. 447-458.
[6]. Regassa, Y., G. Likeleh, and R. Uppala, Modeling and
Simulation of Bullet Resistant Composite Body Armor.
[7]. Teng, X., T. Wierzbicki, and M. Huang, Ballistic
resistance of double-layered armor plates. International
journal of impact engineering, 2008. 35(8): p. 870-884.
[8]. Corran, R., P. Shadbolt, and C. Ruiz, Impact loading of
plates—an experimental investigation. International
Journal of Impact Engineering, 1983. 1(1): p. 3-22.
[9]. Kurtaran, H., M. Buyuk, and A. Eskandarian, Ballistic
impact simulation of GT model vehicle door using finite
element method. Theoretical and Applied Fracture
Mechanics, 2003. 40(2): p. 113-121.
[10]. Feli, S. and M. Asgari, Finite element simulation of
ceramic/composite armor under ballistic impact.
Composites Part B: Engineering, 2011. 42(4): p. 771-
780.
[11]. Fawaz, Z., W. Zheng, and K. Behdinan, Numerical
simulation of normal and oblique ballistic impact on
ceramic composite armours. Composite Structures, 2004.
63(3): p. 387-395.
[12]. Sadanandan, S. and J. Hetherington, Characterisation of
ceramic/steel and ceramic/aluminium armours subjected
to oblique impact. International journal of impact
engineering, 1997. 19(9): p. 811-819.
[13]. Johnson, G.R., Numerical algorithms and material
models for high-velocity impact computations.
International Journal of Impact Engineering, 2011.
38(6): p. 456-472.
[14]. Muhammad, R., et al. Effect of cutting conditions on
temperature generated in drilling process: A FEA
approach. in Advanced Materials Research. 2011. Trans
Tech Publ.
[15]. Muhammad, R., et al. Finite-element analysis of forces
in drilling of Ti-alloys at elevated temperature. in Solid
State Phenomena. 2012. Trans Tech Publ.
[16]. Muhammad, R., et al., 3D modeling of drilling process
of AISI 1010 steel. Journal of Machining and Forming
Technologies ISSN, 2010. 1947: p. 4369.
[17]. Zukas, J.A., High velocity impact dynamics. 1990:
Wiley-Interscience.
[18]. Holmquist, T.J. and G.R. Johnson, Response of silicon
carbide to high velocity impact. Journal of applied
physics, 2002. 91(9): p. 5858-5866.
[19]. Holmquist, T.J., D.W. Templeton, and K.D. Bishnoi,
Constitutive modeling of aluminum nitride for large
strain, high-strain rate, and high-pressure applications.
International Journal of Impact Engineering, 2001.
25(3): p. 211-231.
78
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-87-320.jpg)
![vibration capabilities of the device to provide haptic feedback
to the user. Internal hardware such as accelerometers,
gyroscopes and proximity sensors are used by some
applications to respond to additional user actions, Android
devices boot to the home screen, the primary navigation and
information point on the device, which is similar to the desktop
found on PCs. Android home screens are typically made up of
app icons and widgets; app icons launch the associated
applications, whereas widgets display live, auto-updating
content such as the weather forecast, the user's email inbox, or
a news ticker directly on the home screen. Third-party apps
available on Google Play and other app stores can extensively
re-theme the home screen, and even mimic the look of other
operating systems, such as Windows or IOS.[59] Most
manufacturers, and some wireless carriers, customized the look
and feel of their Android devices to differentiate themselves
from their competitors.
Various types of Application developer suits are used in
market but Software Development Kit (SDK) is the famous
platform to use for application development in android based
operating devices.
App Work Flow
To develop apps for Android, you use a set of tools that are
included in Android Studio. In addition to using the tools from
Android Studio, can also access most of the SDK tools from
the command line. Developing with Android Studio is the
preferred method because it can directly invoke the tools that
need while developing applications.
However we choose to develop with another IDE or a
simple text editor and invoke the tools on the command line or
with scripts. This is a less streamlined way to develop because
you will sometimes have to call command line tools manually,
but you will have access to the same number of features that
you would have in Android Studio. The basic steps for
developing applications (with or without Android Studio) are
shown in Fig-3. The development steps encompass four
development phases, which include:.
A. Environment Setup
During this phase installed and set up development
environment. And also create Android Virtual Devices (AVDs)
and connect hardware devices on which install the
development applications.
B. Project Setup and Development
During this phase we set up and develop Android Studio
project and application modules, which contain all of the
source code and resource files for designed application.
C. Debugging and Testing
During this phase build an application into a debug
able .apk packages that can install and run on the emulator or
an Android-powered device. Android Studio uses a build
system based on Gradle that provides flexibility, customized
build variants, dependency resolution, and much more. If using
another IDE, build an application using Gradle and install it on
a device using adb. Next, with Android Studio debug
application using the Android Debug Monitor and device log
messages (logact) along with the IntelliJ IDEA intelligent
coding features. We can also use a JDWP-compliant debugger
along with the debugging and logging tools that are provided
with the Android SDK. And an Last, test your application
using various Android SDK testing tools. For more
information, see Test your application with the Testing and
Instrumentation framework
Fig-3
D. Publishing
Publishing is the general process that makes your
Android applications available to the users. When you publish
an Android application you perform two main tasks:
Prepare the application for release.
During the preparation step you build a release version of
your application, which users can download and install on their
Android-powered devices.
Usually, release application through an application
marketplace, such as Google Play. However, we can also
release applications by sending them directly to users or by
letting users download them from your own website.
V. PROPOSED APPLICATION
Few years earlier the firms have no choice to manage &
update changes their area maps by timely surveying. The
rescue response on various emergency conditions response
teams are managed by the area an number of rescue stations are
managed regain, City and country because no one knwn the
batter location of the area maps and totally dependent by the
information that provided by the caller, the caller informed the
condition of emergency and inform his location too, the call
receiver/operator note down information and pointed the
position on given map and then lookout the nearest rescue
operation and mobilized them towards effected areas or parties.
The mention procedure required much time to evacuate the
81
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-90-320.jpg)
![effected partied from effected place. Since 1992 the first GNSS
is lunched worldwide by Unites States department of defense
knows as Global Positioning System (GPS), this system is
designed for military requirement but the system provides the
navigation information round the clock globally with low
accuracy for civilian users. Initially the GPS receivers huge in
size and quit expensive, Now days the receivers are available
in reduces in sizes and the decreases the price too. The reduced
size receiver are used/ installed in the smart devices i.e. cell
phones, tablets, wrest watches etc.
as on proposed work the main aim is build up an
android based application, that’s incorporate the GNSS,
android & Cellular Communication to reduced the time for
rescue/ emergency relief operation by fixing the exact position
of the effected place or parties to save lives and also save them
resources.
A. Application Flow
The android application functional block diagram is shown
in the Fig-4, the functional diagram is operational when the app
is installed on the android based smart device either it’s a cell
phone or the tablet/PDA, whatever its brand but must have a
GNSS receiver, now days devices have GPS/ GLONASS
supported receivers.
Start
Direct Launched
Launched by Dialed
Numbers
Enable GNSS
Receiver
Determine Position
with time
Position Fixed
Determine Position
with time
No
YES
Position
Stored
Sending
Source
Generation
.nav
Extention
InternetSMS
Sent to Destination
Server Via Internet
Sent to Destination
Server Via SMS
Displayed Message
“Position Sent”
End Process
Standby mode
Repeat loop
3 times
Disable GNSS
Receiver
Fig-4
After the installation is done on a device either it’s a cell
phone or tablet. The application is standby mode because the
power is important factor in the smart devices.
The application is lunched by the App icon at app manager
screen or by dialed fixed or predefined number patters i.e.
1122, 115 & 114 etc. The application starts and performed its
initial operation that is the enable the GNSS receiver, in this
step the physical GNSS receiver is power on and a popup on
the screen the GPS enables.
In the next block is performed operation i.e. tracked the
presented satellites (4 signals) over the sky, and compute its
position, this process take some time up 30 seconds initial for
position fixation. It performed frequently operations until the
position is fixed. In next step, but it’s performed the parallel
with previous one (position fixing), its desired the information
from the user (depend on the user of application i.e. courier
services, vaccinations, the community health programs or
emergency Search and rescue relief operation) Information
competed with coordinates, now the information is saved at
local storage with automatically generated label (either its date
with time or with number series). The stored information is
may be required to be converted on the required format (i.e.
.nav, .txt or .ext etc extension files or nay other format
according to the requirement) and also added the cell phone
number for easy identification of the position, after suitable
conversion and addition in the file, the information is ready to
send their central information center using the internet or the
SMS, tasks is repeated at three times in four minutes for better
position for rescue operations. Once the task is complete the
user can disconnect the call the application is automatically
disabled the gnss receiver and goes to standby mode.
CONCLUSION
Currently, GNSS resources are available freely without any
cost and GNSS market is growing quickly and has been
expanded in vast areas of applications. Nowadays, GNSS
receivers are installed in wrest watches, phones, tablets,
satellites, aero planes and other communications devices.
Android has an open source code and resource and can easily
be integrated with GNSS. Applications for the local market can
be produced i.e. health, search and rescue will be improved and
enhanced in Pakistan.
ACKNOWLEDGMENT
I personally thank Dr Najam Abbass department of
aeronautics and astronautics institute of space technology
Islamabad, Pakistan for his kind effort for completion of the
research project and express my warm thanks to Mr. Hassan
Ali for his aspiring guidance, and sincerely grateful to them for
sharing their enlightening views on a number of matters related
to the project.
REFERENCES
1] HOANG, V. D., HA, D. T., & PHUONG, X. Q.
(2013). A Vehicle Monitoring and Navigation System Design
based on Android Smartphone (Space, Aeronautical and
Navigational Electronics).
2] Marti, E., Garcia, J., & Molina, J. M. (2014, July).
Navigation capabilities of mid-cost GNSS/INS vs. smartphone:
Analysis and comparison in urban navigation scenarios. In
Information Fusion (FUSION), 2014 17th International
Conference on (pp. 1-7). IEEE.
82
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-91-320.jpg)
![3] Djuric, U., Abolmasov, B., Dragana, P., Marjanovic,
M., & Kuzmic, P. PORTABLE GEOTECHNICS–USING
ANDROID SMARTPHONES AND TABLETS FOR
GEOTECHNICAL FIELD INVESTIGATIONS.
4] developer.android.com
5] http://en.wikipedia.org/wiki/
Android_(operating_system)
83
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-92-320.jpg)
![Geostatistical Analysis on Seismic Data over
North-Western Regions of Pakistan, Afghanistan
and Eastern Regions of Iran & Tajikistan
Malik Abid Hussain Khokhar1
, Asad Ali1
, Sadaf Javed1
& Razia Rani2
1
Department of Space Science, Institute of Space Technology, Islamabad 44000, Pakistan
1
abidmalikgeo@gmail.com, 1
asad06@gmail.com, 1
nomi4all786@gmail.com,
2
Department of Geography, Government College University, Lahore
2
raziaranigeo@gmail.com
Abstract – Earthquake, it is a process which usually take place
anywhere in the world. But most of the time this phenomena
takes place at the place which is weak part of the earth - faults.
Earthquakes occur where faults are but magnitude and depth
remain less at particular places. Factually, it is considered that
earthquake activity is normally related to particular activity of
tectonic processes under the earth. The Seismic geomorphic
structures contain particular basic source mechanisms and with
regard to seismic activity, these bases are considered as a solution
of tectonic regional planes. In this way, different seismic activities
solutions bind their strong dependence with a recognized fault
linage system. Furthermore, these connected seismic activities
give us benefits of attaining an objective of absolute possible
“depth and magnitude” of earthquake. The seism-o-tectonic
model is an essential parameter that covers the assessment of
earthquake activity. The tectonic framework of the said study
region is extended from the fault system on the west to northwest
Himalaya and boundary of Hindu Kush-Pamir terrain to the
seismic belt of the geographical borders of Pakistan, Afghanistan,
Iran and Tajikistan. Sequel to this, a geostatistical analysis are
conducted by utilizing variogram estimations, anisotropy and
kriging methods for seismic depth conversion and magnitude on
the said sub-surfaces and interval velocity fields in a dependable
way.
Index Terms— Geostatistics, Earthquake, Seismicity,
Magnitude, Variogram, Kriging.
I. INTRODUCTION
Earthquake take place in the faults areas where their
intensity increases if there is weak surface prevailed then it
bring about the source of a number of earthquakes as a natural
calamity. Various features of geology in the region are direct
manifestation of subsurface of the earth dynamics that’s why it
is important to understand variety of problem related to
physical endogenic processes aspects of happening deep down
the earth surface. Earthquakes create faulting and folding due
to above mentioned processes and bring about severe natural
disasters in the affected areas. These happened due to primary
and secondary and long waves that determined depth and
intensity of the earthquakes[1]. The area is extended to
Himalayan arc which encompass about 2,500 km from
northwest to southeast and includes from east to west the width
of the belt varies from 250-350 km. The colossal Himalayas
and the Karakoram, representing the biggest concentration of
lithospheric masses, grew south of the Pamir that goes to Iran
from touching southwestern parts of Pakistan and the western
regions of Afghanistan. The study area contains a captivating
geological record of Precambrian to present and terminate both
east and west with spectacular bends. The region covers the
foot hills, lesser Himalayas, Greater Himalayas and Higher
Tajikistan Mountainous Zones. Changes in the Earth's
orography and consequent geomorphic changes are directly
related to the ongoing seismic event. This collisional events
have long been argued to be liable for geological, geochemical
and seismic magnitudes of global extent[2]. The exogenic /
eperogenic processes are still going on with the approximate
rate of a few centimeters per year imperceptibly with continued
erosional and denudation factors of internal movements. The
eroded material from its rugged topography is regularly shed
into different depositional settings within the Himalayas. In
this region, the stress and strain caused due to plate movements
which cause frequent earthquakes enormous loss of human life.
Landslides and irregular thrusting have been also cause of
frequent natural disasters. Deforestation and road construction
have also been witnessed as a reason of landslides. Earthquakes
occurred within the Baluchistan Range (Pakistan) and Iran
ranges with the Mw6.4 which generated a complex and poorly
located seismic sequence. These earthquakes mostly affected
Pishin and Ziarat districts along with the borders areas of Iran,
an area where several active faults have previously been
identified. To create the relationship between these earthquakes
variograms models have been established for geostatistical
analysis”[3]. In the paper, geostatistical analysis (variogram
estimations, anisotropy, kriging and cross validation) on Depth
and Magnitude of the earthquake data on the subject regions
are applied that display spatial structure ranges and directional
ranges.
II. DATA AND STUDY AREA
Borders of Pakistan, Afghanistan, Iran and Tajikistan are
the study area is chosen where earthquakes are frequent due to
weak subsurface. Area is bounded in between 230
.37’N to
410
N and 440
E to 790
E. The points in the map are showing the
number of earthquakes are frequent due to weak subsurface.
Area is bounded in between 230
.37’N to 410
N and 440
E to 790
E. The points in the map are showing the number of
earthquakes that occurred in between 2014 to March 2015.
84
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-93-320.jpg)
![nugget such as 0 1C C C . Nugget effect is minimized to
zero C(0). Owing to this condition, variogram is not enough for
spatial variation analysis & suitable model must be fitted on
it[4]. In the equation number 2, 3, 4 and 5 are known as the
models as mentioned against each. Here, ‘c’ is a Sill and ‘h’ is
lag specified from the distance/ weight derived from the
coordinates and given values at the coordinates.
C. Kriging
Interpolation provide Kriging methods besides Kernel
Density etc. In geostatistical methods, Kriging is also “Best
Linear Unbiased Predictor" (BLUP) which satisfies a certain
criterion and provide minimum variance[5]. The kriging is
based on the weighted moving average[6].
0
0
(6)
var
Z x Z xii
Z It is an estimated iable at the spatial location
x location
Z x observed parameter of Depth and Magnitudei
every predicted form nearest neighbore of spatial
points
weights applied to these neighboring po ts
a
i
t
in
Unity of sum is the weights are constrained so Equation 6 is
an unbiased estimator[7].
IV. RESULTS
Results in the successive pages are over the seismic data of
the study areas. After the application of various models of
variograms as per the appropriate requirement of model of the
variogram, Ordinary Kriging analysis are applied. It is because
of its real implementation and OK (Ordinary Kriging) is
appropriate applicable where spatial sampled data is large. This
method comprehensively evolves on sampled observations
instead of complete population where trend assessment of
spatial identity always re-assessed so that a smooth and
accommodative model may arise. For depth and magnitude
analysis, Ordinary Kriging (OK) remains always very suitable,
but it does not work properly if we have very small sample data
for using in moving neighborhood observations and so it bring
about the reason of further good and useful results[9]. The
results which showing the variograms are in cyclical with the
models of Exponentials, Linear and Gaussian. These were
derived on the basis of lowest standard that bring about the
enormous appropriate cause of validity of findings for any data
basing upon geostatistical methods. Results are restricted to
bifurcated due to large area and large sample size. Initially,
sample size of the data was accumulatively 600 which split to
two parts of 300 each. Data was full of trend which has been
shown in the variograms models and anisotropy plots. These
plots are derives from the R language and ArcGIS was also
incorporated for visual interpretations which has been useful to
predict the nearby areas by known points. By viewing, figures
12, 13, and 14 are predicting depth and magnitudes of
earthquakes by the application of Kriging and Kriging is found
the most suitable predictors because it is showing very accurate
results. For comprehensive geostatistical analysis,
omnidirectional variograms are not taken to compute the
directional trend in the data which require to analyze trend of
anisotropy in the sampled data. So sequel to these indicators
directional variograms are calculated with the angle tolerance
150
from 00
to 1650
. Figure 6 and 7 are interpreting anisotropic
trends as well as directional trend in the data for depth and
magnitude of seismic data on the subject regions, where point-
paired values are evidently displayed with respect to their
directions Finding enormous trend of direction, it is evident
that direction is not uniform to a specific side rather all sides
which depicts that compactness of the lithosphere which is not
too hard to sustain the shocks of frequent earthquakes that’s the
reason for tremendous isotropic trend. Beside these, another
way of determining said trend, there is a method of Rose-
Diagram that is used for computing anisotropy within the map
having 3600
angular direction, under mentioned figures are
related to said directional trend which are showing the
direction of earthquake magnitude and its depth in the subject
regions. With respect to different angels, an unequivocal
variation in the above shown rose diagrams of Pakistan &
Tajikistan Regions and Iran & Afghanistan regions. When the
data of Magnitude values and depth values of occurred
earthquakes of the subject regions are slanted in a side then
both plots show only that certain direction either 00
degree or
900
degree whereas the trends elucidate the level of depth in
different directions specified rocks of weaken lithosphere. For
comprehending the intensity of the depth and magnitude of the
earthquake, figure 6 and 7 exhibit variogram factor parameters
with all possible angles. Numerical figures of the said
variograms are showing directional variogram for relating
parameters of same dataset within anisotropic influence.
Magnitude and depth have different numbers of values in view
of the sill, partial sill and nuggets and range is comparatively
less differ in the whole dataset because these are based on the
Standard Error’s values which are being selected on the basis
of lowest vales and models are also applied on the
appropriations where Depths have 700 at 1500
and sill is 12000
that is the maximum figure in case of Pakistan and Tajikistan
regions and on the same angles Iran and Afghanistan have the
lowest observation at this angle but this regions has more than
6000 sill vales at 1180
and nugget values is 1500. Deviations
proceed to the resultant anisotropic effect in the data values.
Similarly, Magnitude data is of the both regions have different
sill effects, nugget and rang in terms of their direction
significances. Merely, directional variograms are not sufficient
for geostatistical analysis for computations for interpolation
methods like Kriging and Inverse Distance Weighted analysis
but other analysis of variograms are also needed to interpolated
directional ranges and anisotropic corrected variograms of
current data. Above discussed directional ranges are estimated
directional coverage of predicted surface areas along with
appropriate model fittings. Figure 14 is showing Kriging plot
of depth values of earthquake data in all the areas as mentioned
in topic and figure 13 is established using Geostatistical
Analyst Tool of ArcGIS. Figures 12 is serving a clear depiction
of variables Depth and Magnitudes of whole data set with the
86
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-95-320.jpg)
![V. ROSS VALIDATION
Cross-validation, it is also known as the appreciation with
the name of rotation estimation, that is a technique of model
validation to evaluate geostatistical analysis for generalization
of an independent dataset [10]. It is mainly used in settings
where the goal is prediction, and one wants to estimate how
correctly/ perfectly a predictive model to perform in practice
into implementations of cross validation. It is the tool of
computing the validation of correctness and authentication of
the methodology of geostatistical analysis like Kriging and
Kernel Density etc. It is important to mention over here that
the main reasons for using cross-validation instead of using the
conventional validation is that the error (e.g. Root Mean
Square Error) on the dataset. Because the conventional
validation is not a useful estimator of model performance and
thus the error on the observational dataset does not properly
represent the assessment of model performance[10]. This way
to tool is not only the part of geostatistics methods but also it is
a part of Geostatistical Analyst tool in ArcGIS. Sequel to this,
following are few plots which are derived from Geostatistical
Analyst Tool that includes predicted values, error, and standard
error and QQ plot. QQ plot is showing the distribution of
dataset. Figure 15 is the depiction of distribution of predicted
observations and the sample observations with their cross-
validation.
VI. CONCLUSION
On the subject topic of four countries (i.e. Pakistan, Iran,
Afghanistan and Tajikistan), six hundred sample points/
observations were collected and only three parameter are used
which are coordinates, magnitudes and depth of the
earthquakes. Geostatistical analysis are applied on these
parameters. With the help of R language, major geostatistical
analysis (variogram model analysis, anisotropic analysis and
Kriging etc.) are conducted; however ArcGIS is also
incorporated for visual interpretation with the utilization of
geostatistical analyst tools. Taking all the analysis methods on
board, it is concluded that Ordinary Kriging is the valid/ best
estimation method because it presents accurate smooth canvass
of surface and minimum standard error estimation as revealed
in cross validation plots and table number 3 and 4. For
correctness of the application of Kriging cross the validation on
depth and magnitude of earthquake data is carried out as well.
Seismic spatial point data are gigantic source of information to
see the location and concentration of earthquake where we can
estimate disaster damage which help us to layout decision
making efforts for benefiting mankind. Semi-variogram and
other variogram models analysis of seismic data are very
useful for estimating destructions and directions of waves of
earthquakes as showing in the figures 2,3,4, 5 and it is also
significant for direction of earthquakes. The paper has shown
that semi-variogram variables of range, nugget and sill which
can be easily used to pinpoint earthquake spatial deviation and
consequently the location of severely damaged areas as well.
Hence, it is concluded that Northeastern regions of
Afghanistan, Eastern regions of Tajikistan are major
concentrated areas of earthquake and more deepened than the
other regions of Pakistan and Iran. However, as for as
magnitudes of the earthquakes are concerned, eastern and east-
southern parts of Iran are very vulnerable due to weaken parts
of Lithospheric compactness which can give room for more
severe earthquakes with higher intensity of 6.6. Southern parts
of Pakistan with the borders of Iran and western Balochistan
are also vulnerable places for future earthquakes.
ACKNOWLEDGEMENT
The authors are indebted to Dr. Asad Ali who guided them
for writing this paper and he reviewed the paper as a whole.
His precious time from his busy schedule and commendable
cooperation is highly appreciable in conducting this research
work.
REFERENCES
[1] N. L. and T. M. Gary Mavko, “Seismic Methods For
Imaging Physical Properties of the Earth,” pp. 0–3, 1995.
[2] S. Development and I. M. Region, The Tajik Pamirs
Challenges of Sustainable Development in an Isolated
Mountain Region. .
[3] D. L. P. Parasad, Journal of Nepal Geological Society, vol.
45. .
[4] C. E. G. James, R., “Use Of Geostatistics For Accurate
Mapping Of Earthquake Ground Motion,” no. 1988, pp. 31–
40, 1989.
[5] V. De Rubeis, P. Tosi, C. Gasparini, and A. Solipaca,
“Application of Kriging Technique to Seismic Intensity
Data,” vol. 95, no. 2, pp. 540–548, 2005.
[6] M. Franklin, “Solution to Ordinary and Universal Kriging
Equations,” pp. 1–5, 2014.
[7] R. Giraldo, P. Delicado, and J. Mateu, “Geostatistics for
Functional Data : An Ordinary Kriging Approach.”
[8] K. Chang, Introduction To Geographic Information Systems,
Sixth Edit. Printed in Singapore: McGraw-Hill, 2012, p.
418.
[9] P. Abrahamsen, “GEOSTATISTICS FOR SEISMIC
DEPTH CONVERSION 3 . Kriging in the geophysical
literature 4 . Seismic depth conversion 5 . Velocity or depth
prediction 6 . Variogram estimation,” pp. 1–9, 1996.
[10] S. Geisser, Predictive Inference. Chapman and Hall, 1993,
p. 3471.
97
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-106-320.jpg)
![GIS for estimating optimized water demand using
sustainable water resource management for a planned
city
Rabia Tabassum,1,2
, Mudassar Hassan Arsalan2,3
, Arif Inam Osmani3
, Anam Khalid2
1
Science and Humanities Department
National University of Computer and Emerging Sciences (FAST),
Karachi, Pakistan
2
Geoinformaticsand Sustainable Development Research Lab
Institute of Space and Planetary Astrophysics, University of Karachi, Pakistan
3
Osmani & Company (Pvt.) Ltd., Karachi, Pakistan
.
Abstract
Water, a vital natural resource, is the topic of modern researcher and
getting momentum throughout the world due to its scarcity. Many
regions of the world have reached a point at which present water
resources are already being over-used, as indicated by the depletion
of groundwater aquifer and rivers which no longer to reach the sea.
The alternate water supply approach becomes more important to
achieve sustainable urban development and to reduce the energy as
well. The idea of using modern techniques like remote sensing and
geographical information systems (GIS) in water resource
management studies is quite mature. In the last two decades, RS and
GIS have been broadly used for the preparation of different sorts of
thematic layers and integrating them for multiple purposes,
especially for watershed analysis, ground water recharge and water
demand optimizations.
The DHA City Karachi (DCK), the study area, has been envisioned
as the first sustainable city of Pakistan which will probably serve as
a model for the future independent planned cities in the country and
abroad. The domestic water requirement of 54 gpcd has been
considered as a standard of Karachi Water and Sewerage Board for
the metropolitan city and average water demand per unit area is
calculated as 0.09 g/ft2
for DCK.
This study consists of two parts, in the first part water demand
estimation is done for the DCK land use; Residential, commercial,
Mixed use, Amenities, Recreation, Utilities, Transportation,
Agriculture farming and Reserved spaces. Water in gallon per capita
per day (gpcd) and gallon per unit area per day (gal/ft2
) are
considered as water coefficients for the water demand calculation.
DCK water demand on time scale is estimated as per its development
and inhabitation strategy, which is going to develop in three phases;
Short Term (2012-2015), Mid Term (2015-2020) and Long Term
(2020-2030).
Secondly, to estimate optimized water demand for the land use, the
most feasible and easy to deploy strategies and methods like
efficiency practices (behavior practice and Technology practices)
and gray water recycle and reuse is used. By implementing these
water reduction methods on DCK estimated water demand, it is
evident that recycling and reuse of waste water have a positive
impact on ecological sustainability.
Index Terms:
Water demand efficiency, Recycle and reuse of gray water, Water
demand, Alternate water resources, GIS.
I. INTRODUCTION
Water is one of the most vital natural resources that somewhat
different from the other, as it is viewed as a key to prosperity
and wealth. Water has played a crucial role in the growth,
function and location of communities[2]. Egypt, South Africa
and Pakistan are the most water-limited nation among the
twenty five most populous countries. According to the World
Bank data, Pakistan’s river system has a storage capacity of 30
days of water usage, India stores 120 days, while US stores
900 days of river water. In international and national agendas,
sustainable water policies search is at its peak. Approximately
40% of the world’s population will probably experience
shortages in water by 2050. When water could not find in the
desire quality and amount at the right place and time, water
problem may exist [3]. World’s arid and semiarid regions have
motivated the development of the innovative management
measures due to water shortage. At large, the major part of
Pakistan is arid and semi-arid, therefore water management
system improvement is necessary.
The systems to manipulate and present spatial data can be
considered as Geographical Information systems (GIS),
which are capable of performing a range of function such as
data capture, data storage, data cleaning, data finding and
retrieval, statistical and spatial analysis, and data display [4].
GIS applications as efficacious technology in water resources
management for storing, managing, displaying spatial data are
constantly raising. These applications, including supply and
sewer system modeling, surface hydrology and ground water
modeling, storm water and nonpoint source pollution
modeling for urban and agricultural areas. Because of the
spatial nature of the data for design, planning, operation and
maintenance of water and sewer system in urban areas, GIS
technology used for such systems to provide benefit
significantly. The unique characteristics like family size, lot
size, property value, and soil property, which vary from on
different geographical locations, are associated with
consumers that affect demand[5]. In hydrological modeling,
GIS could be useful to provide the information needed for a
watershed runoff analysis. The integrated technique using
98
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-107-320.jpg)
![GIS and physical based distributed hydrological models
considered as an effective technique in the field of watershed
–based water resource management[6].
II. LITERATURE REVIEW
In the last few decades, GIS use started to proliferate in the
field water resources management and planning. Mark R.
Leipnik and Hugo A. Loaiciga [7] discussed eachaspects of
GIS and focused on the other facets of GIS pertinent to the
planning and management of water resources. Geographical
Information Systems (GIS) have become an integral and
beneficial tool in many analyses such as statistical and spatial
analysis in water resources management [4, 8, 9].
Panagopoulos, et al. [10] estimated the future water
requirement based on population growth using multi-criteria
decision analysis in which Analytical Hierarchy Process
(AHP) and GIS techniques were involved. AHP has proved as
a useful tool for planners that assess future water demands and
it can be suitable for planning of a national water management
system.
Udovyk [11] ‘s research revealed that GIS couple with
modeling have potential to make a significant contribution to
integrated water resources management (IWRM) and also
notified the issue involved in using GIS as part of IWRM. GIS
considered as a powerful tool to interpolate and model the
spatial data, enable to overlay and compare the environment
and population data for quantifying the population at risk, and
also communicate these to decision makers and other
stakeholders[11].
A particular aspect of water resource management problems,
needs a specific approach to the development of GIS data
structure. By constructing a conceptual GIS data model to
integrate the physical and logical components of the modeling
problem into an operational framework, Daene C. Mc Kinney
[12] expended the function of GIS that implements a tight
linkage between Water resources management and GIS model
Ahmad [13] introduced a new approach for Simulation of
Water Resources Systems called spatial system dynamics
(SSD which presented for modeling feedback dynamics
processes in time and space. GIS and System dynamics are
couple to develop this model. In the study presented by Allen,
et al. [14], GIS is used for verifying and calibration a
continuous simulation model with an event –driven model;
they realized that GIS is useful in enhancing the modeling
efforts because of increased accuracy, and minimization of
human error with time.
One of the GIS application is future forecasting of water
demand, Shamsi [15] presented this application and
established the criteria for future development with the help of
potential development map, a digital elevation model, a
zoning map, and water ,sewer and road network maps. Same
works could be made for sewerage flows for the rapidly
growing population areas. For calculating and projecting the
average daily sewerage flows from the different land use such
as residential, commercial, and industrial, Kirshen [16] used
data handling properties of GIS.
For the development of a master plan of waste water system,
Giguere [17] used GIS and discovered its efficiency to bring
the relevant data from present databases into geographic
analysis environment, view information and, create report.
The master plan predicts future wastewater flows, assesses
existing treatment and collection system potential, constructs a
sewer rehabilitation program, assesses and identifies
alternatives to expand wastewater facilities, and makes a
phased plan to finance the required improvement. Moutal [18]
presented same work for different country. The automated
mapping system was used to record, update and database
development for planning, design and operations.
Domestic water demand considered as a complex element of
different factors, such as socioeconomic and physical
characteristics, urban planning strategies, infrastructure, and
public water policy. In GIS environment, different thematic
layers are used such as road network distance, distance from
the city center, distance from the coastline, topographic slope,
land use/land cover, Urban Plan, population density and
existing water supply and sewerage system that correspond for
the planning factors.
This study is basically for newly planned city and GIS tools
not only used for the master plan of the city, but also for
estimating water demand on spatial and temporal bases. The
main emphasis of the study is the estimation of optimized
water demand using sustainable water management approach.
III. STUDY AREA
DHA City Karachi (DCK) has been envisioned as “The First
Sustainable City of Pakistan”, which will serve as a model for
the future independent cities in the country and abroad. It is
with the aim to provide its citizens a safe, comfortable,
efficient, and beautiful lifestyle. DCK is strategically located
on the Karachi-Hyderabad Super Highway in District Malir at
the eastern border of Karachi as an independent urban center
in the vicinity of Karachi Metropolis. The city is planned on
an area spanning 11,668 Acres, comprising residential,
commercial and mixed-use elements.
In geographical characteristics topography and distances are
the main factors affecting the potential water sources. DCK is
naturally drained by Abdar and SukkanNullah (small non-
perennial water tributaries of Malir River). Their productive
aquifer is recharged from the same water streams and small
water reservoirs developed on them. It extends from latitudes
240 57 N to 250 2 N and longitudes 670 24’ E to 670 32’
E, and is a sub-tropical location in character.
The study area has a moderate climate and experiences bulk
rainfall in the duration of monsoon season, but the average
rainfall is recorded as only 7inches per annum. In summer,
May and June are the hottest months and temperature often
reaches the 43°C. January is considered as the coldest month
with temperature 5°C. As a part of Karachi district, DCK may
get its share of water from the bulk water supply by Karachi
99
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-108-320.jpg)
![Water and Sewage Board. But the current water budget of KW
&SB’s has been already in stress and demand is more than
supply.
A. DCK land Use Planning
The DHA City Karachi is the study area of this research and
Figure 1 shows its land use. For DHA city Karachi the entire
planning approach focuses on Sustainability Design
Principles, including the concepts of connectivity, efficiency,
renewability, minimization of externalities and carry
capacity[19]. These goals of sustainable design have been
adopted in the overall districts planning approach by
maintaining the ecological integrity of site nullahs, reserved
spaces and natural contours[20]. Master plan of DCK
cumulatively covers approximately 11,668 Acres of land and
comprises of eight major land use types as shown in Figure 2.
IV. METHODOLOGY
The basic aim of this study is the estimation of optimized
water demand by using alternate water resources for DCK in
GIS environment. For this purpose, the master plan of DCK
was converted into GIS format. ArcGIS software was used to
translate files from AutoCAD drawing files to the ArcGIS
shape file. The master plan of DCK was consisting of detailed
parcels/plots, transportation corridors, reserved spaces for
natural drainage, open and green spaces etc. Land use
classification was applied to parcels / plots, according to the
detailed master plan categories on different levels, such as the
major level comprised Residential, commercial, mixed use,
amenities, recreation, utilities, transportation, agricultural and
reserved spaces. The minor / detailed land use maps were also
developed for each individual major land use class according
to DCK planning. The flow chart of land use layer formation
is shown in Figure 3.
The water demand for DCK was calculated in the spatial
domain, on the basis of water in gallon per capita per day
(gpcd) and gallon per unit area per day (gal/ft2
), which was
measured according to the land use types. On a time scale
DCK water demand is estimated as per its development and
inhabitation strategy, which is going to develop in three
phases; Short-term, Midterm and Long-term. Sewage volume
is calculated, which gives the amount of water that can be
reused for the non-potable use. Also by implementing the
water reduction method such as efficiency practices and grey
water recycling, optimized water demand is estimated.
A. Criteria of Water demand estimation
According to United Nation, in 2012 approximately 50% of
the world's population lived in urban areas and this percentage
is expected to swell to 60 % by 2030[21]. Therefore, water
demand in commercial, domestic, and industrial areas are
consequently growing in cities and result the water
scarcity[22]. For newly planned city, water supply authorities
Figure 1: Study Area “DHA City Karachi with its land use
Figure 2: Land use types and % of Area of DCK
Figure 3: Flow chart of land use layers formation
100
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-109-320.jpg)
![are needed to supply water to their consumers with a certain
degree of reliability. Non uniform distribution of water will
cause water scarcity in some area and excess in other areas
which result the obstacle the natural growth of the city.
Therefore, proper criteria setting is necessary to know the
water demand in spatial and temporal domain instead of
computing demand as a whole which is the main objective of
this study.
Municipal water use commonly consists of residential,
industrial, commercial, and public uses, as well as some minor
use for other purposes like system losses, line cleaning and
firefighting. In developing countries public water withdrawal
is only 13-26 gpcd in contrast, many developed countries’
water withdrawal is 79-158gpcd. Municipal water demand is
exactly related to the amount of water withdrawn by
population in town, cities, housing estates, public and,
domestic service enterprises.
The public demand consumes high quality water from the city
water supply system and as its supply also includes water for
industry that provides directly for the requirements of urban
populations. The portion of municipal water demand that is
used in the home is the residential or domestic water demand
and worldwide typical domestic use of water in comparison
with residential plot of 500 square yards is given in Table 1.
This water use is highly dependent on socioeconomic and
environmental conditions and including toilet flush, bathing,
cooking, drinking, watering lawn, and other uses. The average
water demand varies from 25 to 120 gpcd whereas the most
commonly used range is 50-80gpcd for the world urban areas.
In many developed countries residential water use can
constitute over half of total municipal water use and
consumption. Residential water use is also directly linked to
the general public health. Improving the health of the poor is
one of the main goals of water and sanitation projects.
Therefore, household water use is always the top priority in
the municipal water supply[23].
Estimation of water demand based on per-capita consumption
is one of the common approaches. Karachi Water and
Sewerage Board is the main source of water supply for
Karachi city. DCK’s current estimated water demand stands at
45MGD. Initially the average domestic water requirement has
assumed as 54 gallons per capita per day for the DCK which is
a reasonable value in comparison with per capita daily water
demand for the different world cities, e.g. Tokyo –Japan
(77gpcd) New York-USA (161 gpcd) Colombo-Sri Lanka
(60gpcd) and Mumbai- India (43gpcd) [24].
Most of the people who have purchased plots of land for
residential or commercial purposes in the DCK are likely to
belong to the affluent society of Pakistan. The houses or
dwellings constructed by them are considered to be of superior
quality depicting a higher standard of living, which requires a
large amount of per capita water consumption. It is anticipated
that this may necessitate a higher per capita water supply to be
provided for the water supply system in this area. In
comparison to the general average provision normally adopted
by KWSB for their water supply project as for Karachi city,
which is only 36gpcd[25].
Commercial, Industrial, and Institutional (CII) sectors are
considered as the significant contributors to Public water
demand. Historically utilities have relied on water use
coefficient, which is the number of employees for estimating
CII water use, but it is difficult to obtain this information at
affine resolution. To overcome this situation through spatial,
physical and economic property-based information, Miguel A.
Morales [26] developed a methodology to estimate CII water
use. For calculating commercial water demand, a survey has
been conducted in different commercial areas of Karachi.
With the help of a questionnaire, commercial water demand
has been estimated on the basis of water uses for drinking,
cleaning, washroom and other[27]. The result shows that
water demand per employee has been recorded as 2.3 gal per
employee per day, whereas the water demand per unit area
was 0.031gal/ ft2
. Generally, the commercial water demand is
about 10-20% of total water demand. The average water
demand as per planning of the DCK commercial plot is given
in Figure 4.
Table 1: World Wide Typical domestic use of water in
comparison with residential plot.
Types of Water Use
World Average
(%)
Residential
500 sq.yd plot
(%)
Toilet Flush, including its leakage 33 30
Shower and bath 28 25
Wash basin 11 10
Kitchen 9 10
Drinking, cooking (2 - 6)
Dishwashing (3 - 5)
Garbage Disposal (0 - 6)
Laundry and washing machine 16 15
Lawn sprinkling and
miscellaneous 3 10
*Source: Interim Report-1 “feasibility report of DCK alternate water supply”
[1]
Figure 4 : Average Water Demand as per planning of DCK
commercial plot.
101
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-110-320.jpg)
![Water use in public building such as community hall,
Marriage hall, Mess club, culture and art, school etc. as well
as in public services including fire station, street washing and
landscape irrigation are considered as a Public water use.
Usually 5-10% of Municipal water demand use in public
area[24]. As per planning of DCK the average water demand
for different public building are given in Table 2. The public
water demand may become more precise when the actual
detailed building structure is planned.
V. RESULTS
A. Water Demand for DCK
A key input in municipal water services planning and design is
the estimation of present water demand, and the prediction of
future water demand. In distribution of domestic water equally
to all regions with proper pressure, a uniform spatial water
distribution system will be very helpful[28]. As the total
average water use when DCK will be fully developed
expected to reach up to 45MGD and therefore water demand
per unit area is around 0.09g/ft2
. Land use planning is used for
estimation of water demand for DCK. Table 3 shows the
distribution of water requirement for the land use of DCK.
In the whole water requirement, the biggest share is of
residential use (41% of total water demand) which occupies
the major share of land in DCK (37% of land use). However
the water demand per square feet for residential area is quite
low (0.09g/ft2
). The further distribution of water requirement
on various sizes of plot in the DCK residential area is shown
in Figure 5. It shows that approximately 46% water will be
required for R-5 (i.e. plots of 500 square yards).
In the DCK almost 2% area is designated for commercial plots
of various sizes in community centers and in the central
business district. However, some of the largest plots with
higher FAR are planned as the Mall Zone in a commercial
area and has the highest water demand as compared with the
other commercial areas as shown in Figure 6.
Its vertical height (FAR) and horizontal size (plot size) with
density of water users (population) are the major water
demand controlling factors. In the overall DCK the second
highest water requirement will be for commercial plots, which
is almost 16.8% of the total water demand. Although most of
the mixed use plots in community centers (C3, C4 and C5) are
comparatively small in size, however large designated plots in
the CBD and South Zone are the major role players due to
their sized and respective FAR. Further distribution of water
demand for mixed use plots is given in map of Figure 8. It has
Table 2 Average water demand for public water use in DCK
Land Use Average Water Requirement
Community Hall, Marrige
Hall, 0.15 gallon per sqft per day
Mess , Club, Culture & Art 0.15 gallon per sqft per day
Mosque 0.5 gallon per sqft per day
Hospital 250 gallon per bed
DCK Clinics
0.15 gallon per sqft per day of
covered area
Collage/ School(day scholar,
without boarding) 17 gallon per student per day
University 34 gallon per student per day
Glof club 0.02 gallon pers qft per day
Lake View Park 0.02 gallon per sqft per day
Park and Playground 0.01 gallon per sqft per day
Theme Park 0.01 gallon per sqft per day
Nursery 0.02 gallon per sqft per day
Road Landscaping 0.006 gallon per sqft per day
Table 3: Average Water demand of fully developed DCK
S- No Land use Types Water
Demand(MGD)
Water
Demand %
1 Residential 18.6 41.2
2 Commercial 7.6 16.8
3 Mixed use(Res. Cum Com) 7.4 16.4
4 Amenities 3.7 8.3
5 Recreation 0.7 1.7
6 Utilities 5.4 12.0
7 Transportation 1.0 2.3
8 Agricultural and Reserved
Spaces
0.6 1.2
Total 45.0 100.0
Figure 5: Distribution of Water demand (MGD) in
Residential Land use
Figure 6: Distribution of water requirement (MGD) in
Commercial Land use.
102
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-111-320.jpg)
![parameter. Fruits, flowers, flower colors, fragrance, birds’
attraction and water requirement are the main selection
parameter for soft landscaping. In this connection other than
recreational facilities, transpiration corridors are also planned
with soft landscaping. Transportation corridors cover around
34% of DCK land with different row and road categories. The
water requirement for transportation corridors landscaping and
other minor uses will be around 2.3% of total water demand
with highest share at streets. The details are given in map of
Figure 13.
1) Spatial Analysis of Water Demand
According to zonal distribution, DCK comprises of four main
zones; DCK sectors, South Zone, Down Town and Gateway,
as well as miscellaneous area as shown in Figure 14. In zonal
distribution more than half (60%) of water share is required
for residential sectors with a comparatively low water
requirement per unit (0.07g/ft2
). Second largest requirement of
water is in south zone that is almost 15.2 % of the total
demand with a comparatively higher value per unit area
(0.15g/ft2
). The highest water requirement per unit area in City
Gateway and Downtown (i.e. 0.27g/ft2
), that covers around
14% share. Further sectors wise water requirement are given
in Figure 12 and Figure 15 that show the sector wise split of
land use water demand.
2) Temporal Analysis of Water Demand
Many of engineering applications and operational
management of water distribution system are based on the
future water consumption. Therefore, it is necessary to
forecast water demand not only on spatial scale but also with
development time considerations. It is very hard to get a
reliable estimation of future water demand without a clear
understanding of historical and current trends of water use
patterns [29]. Commonly, water demand prediction is divided
into three categories: long - term, medium –term and short-
term periods. Water forecasting in future years used for
distribution system design refers to long-term period. Seasonal
water consumption refers as median-term period, whereas
short –term period concern with daily and hourly water
demand.
Domestic water use by households is a dominant municipal
water requirement in urban areas. Water utilities to make
system plan, design and asset management are important for
municipal water demand, short- and long- term forecasting.
Long term water demand is useful for management and
operation, whereas short term is essential for municipal water
demand short and long term forecasting. Water demand
prediction accuracy achievement is challenging, however ,
because the predicting model must simultaneously consider
multiple factors associated with economic development,
population growth and migration, climate changes, and even
consumer behavioral patterns[30].
The population growth rate of the city is required for
forecasting the domestic water requirements. Future
population prediction is based on the knowledge about the
city, its development of surrounding areas, environments,
potential trade expansion and, infrastructure and
communication network development. As per DCK
development strategy, it is going to develop in three phases;
Short-Term (2012-2015), Mid- Term (2015-2020) and Long-
Term (2020-2030). On the basis of population growth of
Figure 13: Distribution of Water requirement in
Transportation use
Figure 14: Zonal Wise water demand
Figure 15: Sectors wise water demand of DCK
Figure 12: Sector wise split of land use Water
Requirement.
Table 4: Water requirement with respect to DCK development
phases
Development plants Water requirement
after completion
(MG)
Yearly Water
Resource
Development
requirement(MG)
Short-Term (2012-2015) 11.1 3.7
Mid-Term (2015-2020) 20.7 4.14
Long-Term (2020-2030) 11.5 1.15
104
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-113-320.jpg)
![DCK, water demand for these three phases has been estimated
as shown in Table 4 and Figure 16 is represented its spatial
distribution.
By considering development plans it is estimated that for
short- term development approximately 11.1 MG water will
be required that will serve the complete demand of the short-
term planned area. However, its water source development
will be tricky in the beginning plan; just for the short term
development plan average annual water resource development
requirement is around 3.7MG per year. However the actual
use of water will be started with the occupancy and use of land
for the respective purpose. It is estimated that in initial stage
only water is required for construction purpose and only 10 to
15 % of water will be required in DCK Camp site till the end
of short term development plan. It is better to develop around
1.5 to 3 MG water resource by the end of short term
development plan.
B. Reduced water demand
A demand reduction measure serves to reduce water demand,
water losses, peak water demand, and non-essential water
uses. Reducing water consumption in the home, market and at
public places is a simple and easy way to decrease water and
energy bills and reduce household’s impact on the
environment. Conserving scarce water resources helps reduce
the need to dam rivers, reduce waste water produced and
treated at sewage plants, lower energy requirements for
treating and transporting water and waste water, and reduce
greenhouse emission.
1) Strategy to Reduction water demand
For the efficient use of water at usage point is more effective
to save water and reduce overall municipal water demand.
There are many methods and strategy for water reduction, but
most feasible and easy to deploy are; Use of Water efficient
plumbing devices or fixtures and Recycling or re-use of grey
water
a) Use of Water efficient plumbing devices or fixtures
In the last two decades the use of water efficient plumbing
fixture has been introduced throughout the world considering
water scarcity. It has been observed that water efficient
appliances in the home can reduce indoor water consumption
by 35-50%[31]. Typically used such devices and reported
water efficiencies are given in the Table 5.
Error! Reference source not found. gives figure for water
demand expected saving in water demand at DCK pertaining
to use of water efficient devices. It is clear from the above
table that the use of the aforementioned devices can save 44 %
water and the average water requirement will be reduced to 30
gpcd. The use of such devices can either be regulated in
building codes for their increased use. It is also suggested that
new home building should be guided to use these devices
through media campaigns. The land use like amenities and
commercial building give more saving by using water
conservation devices.
b) Recycling and re-use of Grey water
For grey water recycling, the overriding principle is safe usage
of recycling water. Therefore, we have not taken grey water
recycling where its safe usage can be questioned or its
consumption is not required. For plot sizes smaller than
500sqyrds, in any land use, have not been considered because
they do not have any green area and grey water usage for
landscaping is suggested only for large size plots. Health
facilities have not been incorporated for calculation of grey
water because of the hazardous nature of medical waste.
Therefore, grey water volume has been computed for
residential plots R5, R10 and R20, commercial plots (C5, C10,
and C20), mixed use (M5, M10, and M20), as well as planned
amenities. For this purpose, grey water shall be treated sites
and reused for landscaping. Amenities especially educational
institutes, produce a lot of grey water. This phenomenon
Figure 16: Water Demand for DCK development phases
Table 5: Water demand using water efficiency devices
Name of efficient device Frequency of
use(per person)
Daily Water use
without water
conservation device
(gal/person)
Daily Water use with
water saving device
(gal/person)
Daily Water saving
with water saving
device
(gal/person)
Annual Water saving
(gal/person)
Low-flush Toilet (1.6gpf) 5.1 flushes /day 204 8.2 12.2 4453
Low- volume showerhead
(2.5gpm)
5.3 minutes/day 15.9 13.3 2.6 949
Low- volume Faucet
(1.5gpm)
4 minutes/day 12 6 6 2190
Front-loading washing
machine (27gpl)
0.37 loads/day 18.9 10 8.9 3249
Water efficient dishwasher
(7gpl)
0.1 loads/day 1.1 0.7 0.4 146
Total 68.3 38.2 30.1 10987
*Assumes conventional toilet at 4gpf, showerheads at 3gpm,faucet at 3gpm, washing machine at 51gpl, and dishwasher at 11gpl[1]
105
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-114-320.jpg)
![comparatively high water demand is essential. Karachi water
and sewerage supply is the only water resource for DCK
which already supply insufficient water for Karachi city.
Reuse or recycling of waste water and use of efficient devices
serve as efficient and valuable ways to cope with the scarcity
of water and can reduce the water demand. The estimated
waste water for the DCK is 66% of the total water demand,
taken from the land use; residential, commercial, mixed use
and amenities, can reuse after the treated for landscaping and
agriculture outside the houses. For implementing these
alternate water resources, well-structured planning is
necessary. As for reuse of wastewater, black water drain
system should be separated with grey water and forcefully
implementation of water efficient fixture is essential for the
householder. There will be 45MGD water required for the
whole planning of DCK and water demands by using these
alternate resources are estimated. The results of the first
scenario based on reuse of grey water able of only 7.9 %
reduction of water demand. Whereas in the third scenario,
efficient device use show that efficient fixtures can reduce
water demand up to 39%. The second scenario, based on reuse
of grey water with the use of efficient devices, is considered as
an optimized water demand because of the maximum
reduction in water demand about 43%. After successful
deployment of the suggestion without failure to operate and
maintain the system could result in significant reduction in
water demand. Considering the scarcity of water in the area,
we expect that these critical management practices will be
adopted and implemented.
For further study this research can be extended to estimate the
portable and non-portable water use from the reused waste
water. This research provides a sustainable water resource
management system which is the requirement of the better life
style free from water scarcity which is expected to increase in
future.
VII. ACKNOWLEDGEMENT
The authors are grateful to DHA City Karachi, Pakistan and
Osmani & Company Pvt. Ltd. for sharing planning
information and other relevant literature related to the DCK.
Authors are also thankful to the GIS and planning team of
Osmani & Company Pvt. Ltd. for their great support.
VIII. REFERENCES
[1] OCL, "Interim Report -1 , " Feasibility study for alternate
water supply stretegy including rechagre, reuse and small
dams " Osmani Company (PVT) limited and Pakistan
Defences Officers Housing Authority Karachi 2013.
[2] F. Arbu s, . a. . Garc a- ali as, and . art ne -
Espiñeira, "Estimation of residential water demand: a state-
of-the-art review," The Journal of Socio-Economics, vol.
32, pp. 81-102, 2003.
[3] F. A. Ward, "Decision support for water policy: a review of
economic concepts and tools," Water Policy, vol. 9, p. 1,
2007.
[4] O. Udovyk, "GIS for intergrated water resourses
management," Integrated Urban Water Resources
ManagementNATO Security through Science Series., pp.
35-42, 2006.
[5] R. H. a. H. R. Vassilios A. Tsihrintzis, Fuentes., "Use of
Geographic Information Systems (GIS) in Water
Resources: A Review," Water Resources Management, vol.
10, pp. 251-277, 1996.
[6] Z. X. Xu, K. Ito, G. A. Schultz, and a. J. Y. Li, "Integrated
hydrologica modeling and GIS in water resources
management.," Journal of computing in civil engineering,
vol. 15, pp. 217-223, 2001.
[7] K. K. K. Mark R. Leipnik, and and Hugo A. Loaiciga,
"Implementation of GIS for Water resources planning and
management," Journal of Water Resources Planning and
Management, vol. 119, pp. 184-205, 1993.
[8] A. Gemitzi, V. A. Tsihrintzis, O. Christou, and C. Petalas,
"Use of GIS in siting stabilization pond facilities for
domestic wastewater treatment," J Environ Manage, vol.
82, pp. 155-66, Jan 2007.
[9] K. V. Kalivas DP, Karantounias G, " A GIS for the
assessment of the spatio-temporal changes of
the Kotychi lagoon, western Peloponnese, Greece.," Water Resourece
Management, vol. 17, pp. 19-36, 2003.
[10] G. P. Panagopoulos, G. D. Bathrellos, H. D. Skilodimou,
and F. A. Martsouka, "Mapping Urban Water Demands
Using Multi-Criteria Analysis and GIS," Water Resources
Management, vol. 26, pp. 1347-1363, 2012.
[11] O. Udovyk, "GIS for intergrated water resourses
management," Integrated Urban Water Resources
Management, pp. 35-42, 2006.
[12] X. C. Daene C. Mc Kinney, "Linking GIS and water
resources management models:an object-oriented method,"
Environmental Modelling & Software, vol. 17, pp. 413-
425, 2002.
[13] S. a. S. Ahmad, S. , "Spatial System Dynamics: New
Approach for Simulation of Water Resources Systems,"
Journal of Computing in Civil Engineering, vol. 18, pp.
331–340, 2004.
[14] L. Allen, J. Christian-Smith, and M. Palaniappan,
"Overview of Greywater Reuse: The Potential ofGreywater
Systems to Aid Sustainable WaterManagement," Pacific
Institute 654 13th Street, Preservation Park Oakland,
California.2010.
[15] U. M. Shamsi, "GIS, Remote Sensing, and Master Water
Plan: A Case Study," In Computing in Civil Engineering
and Geographic Information Systems Symposium, pp. 695-
702, 1992.
[16] P. Kirshen, Nvule, D., and Corliss, J, "Estimating urban
and suburban sewerage flows withassistance of GIS
technology," in Proc. ASCE Water Resow Plan. Manage.
Conf, Baltimore, MD, 2-6 August,, 1992, pp. 208-212.
[17] P. R. a. B. S. Giguere, "Application of GIS to waste water
system mmmaster planning in Albuquerque," in ASCE
107
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-116-320.jpg)
![National Conference Civil Eng Appl. Remote Sensing GZS,
Washington DC,, 1991, pp. 115-119.
[18] H. P. a. B. outal, D. , "Updating New York city’s sewer
and water main distributionsystems: practical applications
of GIS," in ASCE National Conference Civil Eng Appl.
Remote Sensing GZS, Washington DC,, 1991, pp. 155-164.
[19] OCL, "DHA city Karachi planning report," Osmani
Company (PVT) Limited and Pakistan Defences Officers
Housing Authority Karachi2012.
[20] OCL, "DHA city Karachi ecology Report," Osmani
Company (PVT) Limited and Pakistan Defences Officers
Housing Authority Karachi2012.
[21] U. Nation, "The Millennium Development Goals Report,"
UNITED NATIONS2012.
[22] P. R. Philippe Loubet , Eleonore Loiseau , and V. Bellon-
Maurel, "Life cycle assessments of urban water systems A
comparative analysis of selected peer-reviewed literature,"
Water Reseach, vol. 64, pp. 187-202, 2014.
[23] T. Lu, "Research of domestic water consumption: a field
study in
Harbin, China," Master of Science of Loughborough University,
Water, Engineering and Development Centre Department
of Civil and BuildingEngineering, 2007.
[24] E. M. M. Syed R. Qasim, Guang Zhu, Water Works
Engineering: Planning, Design, and Operation: Prentice-
Hall Of India Pvt. Limited, 2000.
[25] M. Ahmad, "Pakistan Water and Sanitation Operators
Directory," Water and Sanitation program. 2012.
[26] J. P. H. Miguel A. Morales, Kenneth R. Friedman, and
Jacoueline M. Martin., "Estimating commercial,
industrial,and institutional water use on the basis of heated
building area," Journal American Water Works Association
vol. 103, 2011.
[27] R. Tabassum, "Title," unpublished|.
[28] K. H. V. Durga Rao, "Multi-criteria spatial decision
analysis for forecasting urban water requirements: a case
study of Dehradun city, India," Landscape and Urban
Planning, vol. 71, pp. 163-174, 2005.
[29] M. A. Al-Noaimi, "Decelopment of water resources in
Bahrain: A combined approch of supply demand analysis,"
PDh in School of Earth, Ocean, and Environmental
Sciences, University of Plymouth.
[30] C. Qi and N. B. Chang, "System dynamics modeling for
municipal water demand estimation in an urban region
under uncertain economic impacts," J Environ Manage,
vol. 92, pp. 1628-41, Jun 2011.
[31] D. Inman and P. Jeffrey, "A review of residential water
conservation tool performance and influences on
implementation effectiveness," Urban Water Journal, vol.
3, pp. 127-143, 2006.
108
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-117-320.jpg)
![Optimized Threshold Calculation based on received
signal characteristics for Blanking Non-linearity at
OFDM Receivers
Ferheen Ayaz
SUPARCO Institute of Technical Training, HR Dte Gen,
Space and Upper Atmosphere Research Commission,
Karachi, Pakistan
ferheen@msn.com
Zehra Ali
Space Electronics Wing,
Space and Upper Atmosphere Research Commission,
Karachi, Pakistan
Zehra.aly82@gmail.com
Abstract—In satellite communications, OFDM (Orthogonal
Frequency Division Multiplexing) is evolving quickly as a well-
established scheme. OFDM based communication system
performance, despite of its robustness and high efficiency, is
greatly affected by the addition of Impulsive Noise (IN). Additive
impulsive noise deteriorates the transmitted signal and overall
performance results poor when signal to noise ratio (SNR) and
symbol error rate (SER) are considered. Blanking nonlinearity
scheme is among the simplest methods to reduce the adverse
effect of impulsive noise at OFDM receivers. The performance of
blanking non-linearity is based upon threshold selection. In
blanking, the received samples whose magnitude exceeds a
certain fixed threshold are considered to be IN-affected and are
therefore set to zero. The value of threshold plays a very
important role in detecting the samples affected by IN. Setting a
fixed threshold does not always determine the IN-affected
samples truly as the OFDM distribution characteristics vary
throughout the signal. Several theoretical methods to optimize
the threshold level to improve system performance have been
reported. This paper reviews some of the threshold optimization
methods and proposes a new method based on some
characteristic distribution of the received signal including
median and peak. This proposed optimized threshold can be used
practically and results in higher or comparable Signal to
Impulsive Noise ratio than the other methods which are
incapable to be adopted in practical OFDM receivers.
Keywords—OFDM; impulsive noise; blanking non-linearity;
threshold; median; peak
I. INTRODUCTION
Orthogonal Frequency Division Multiplexing (OFDM) is a
rapidly growing modulation scheme and is widely adopted in
many wireless applications [1]. Performance of OFDM
modulation scheme for GNSS (Global Navigation Satellite
Systems) is one of the latest research topics [2]. Other
applications of OFDM include 4G mobile satellite systems [3]
and European Digital Terrestrial Video Broadcasting (DVB-T)
[4]. Despite of many advantages of OFDM communication
systems, the performance degradation in them still persists due
to high-power additive Impulsive Noise (IN) pulses and their
random occurrence in a channel [5]. DVB-T systems are
affected significantly by Impulsive Noise [6]. Performance
degradation in OFDM systems due to impulsive noise result in
high Symbol Error Rate (SER) and low Signal to Impulsive
Noise Ratio (SINR) [7].
A number of schemes to eliminate or reduce impulsive
noise from the received signal have been reported. These
schemes aim to increase the Signal to Impulsive Noise Ratio
(SINR) of the overall system. The simple approach to mitigate
impulsive noise from a received signal is to introduce a
memoryless non-linearity at OFDM receiver before
demodulation. Blanking, clipping and hybrid clipping-
blanking schemes have been devised using this approach.
Among all the non-linearity schemes, blanking is the easiest to
implement as it only converts the high amplitude samples
corrupted by IN to zero [7-8]. The blanking non-linearity
scheme revolves around the decision of whether the received
sample is affected by impulsive noise or not and then
changing it to zero. The decision is taken by a fixed parameter
known as threshold level or simply threshold. The threshold
level indicates if IN has corrupted the received signal sample
or the original sample value is retained at the receiver exactly
as transmitted. If the received sample’s amplitude exceeds the
threshold level, it is considered to be corrupted by IN and
therefore it is multiplied by zero. If the threshold level is
higher than the sample’s amplitude, it is assumed to be the
original sample value unaffected by IN and hence multiplied
by one to provide no change to the sample [9].
Fixing a value of threshold is very important as it defines
the overall SINR of the system. A low value of threshold may
blank the original signal samples which were initially not
corrupted by IN and a high threshold value may pass the IN
affected samples without blanking [10]. Fig. 1 illustrates the
two values of threshold to be used with blanking nonlinearity
on a signal affected by impulsive noise.
Since the performance of blanking non-linearity is greatly
affected by the selection of threshold, the choice of an
optimized threshold is a crucial decision. Optimized threshold
has been calculated in literature by different methods [11-12].
However, theoretical calculations of optimized threshold
109
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-118-320.jpg)
![cannot be used at practical OFDM receivers where probability
of IN occurrence is unknown. In this article, we review the
two threshold optimization techniques for blanking
nonlinearity; first is adaptive threshold optimization method
which maximizes the Signal to Impulsive Noise Ratio (SINR)
[11] and second is dynamic peak based threshold estimation
method [12]. We also propose a new method for calculating
threshold based upon median and peak of the received signal.
The proposed technique of threshold calculation has the
advantage over the other two schemes as it can be used at
practical OFDM receivers without the information of original
signal transmitted, probability of impulsive noise occurrence
and maximum SINR achievable.
0 2 4 6 8 10 12
x10
4
0
2
4
6
8
10
12
OFDM symbols corrupted by IN
Amplitude
OFDM symbols
Impulsive Noise
Threshold 1
(Too low)
Threshold 2
(Too high)
Fig. 1. Two threshold levels considered for blanking IN-affected samples.
The organization of the remaining paper is as follows:
OFDM transmission system with blanking non-linearity
scheme is described in Section II. Section III overviews the
two threshold optimization techniques and proposes Median
Based Threshold (MBT). Section IV includes simulation
results in which MBT is compared with the other threshold
optimizing techniques with respect to Signal to Impulsive
Noise Ratio (SINR) and Symbol Error Rate (SER).
Conclusion is presented in Section V.
II. THE OFDM TRANSMISSION SYSTEM
OFDM transmission is represented in the form of block
diagram as shown in Fig. 2.
Fig. 2. OFDM transmission system with blanking scheme at receiver.
In OFDM transmission, the time-domain form of
transmitted signal s(t) is found by applying inverse Fourier
transform to the frequency-domain signal by using
11 2 . .
( ) exp , 0 t T
S0
N k t
s t S j
k TN k S
(1)
Where S
k denotes frequency-domain signal, N is number
of subcarriers, 1j , and sT is active symbol interval.
Impulsive Noise (IN) is represented as ki and is included to the
transmitted signal using Bernoulli-Gaussian random process
as follows:
, 0 k N-1k k ki b g (2)
Where ‘ 1kb ’ or ‘ 0kb ’ according to Bernoulli process
of sequence. The probability for ‘ 1kb ’ is p and ‘ 0kb ’
1 p . Whereas, kg is additive white Gaussian noise (AWGN)
with variance
22
(1/ 2) [ ].i kE g Hence, the signal-to-IN
ratio (SIR) can be given by; 2
1010log (1/ )iSIR . It is
worthwhile mentioning here that in this paper we represent
theoretical signal to impulsive noise ratio as SIR and practical
signal to impulsive noise ratio as SINR. The time-domain
signal ( kr ) received after addition of IN and AWGN can be
represented as
if 0
,
if 1
k k
k
k k k
s b
r
s i b
(3)
The received signal is passed to the blanking non-linearity
block. The output of blanking non-linearity ky is then fed to
the OFDM demodulator for the further processing and is
represented as
, 0 k N-1k k ky r d (4)
Where kd represents the blanking non-linearity scheme and
is given as
1
0
k
k
k
r T
d
r T
(5)
Where T is the blanking threshold level and
0, 1, ..., 1k N . Optimizing the value of T for increasing
effectiveness of blanking non-linearity is described in the next
section.
III. THRESHOLD OPTIMIZATION
A. Overview of Threshold Optimization Techniques
An optimized threshold is very important for efficient
blanking of Impulsive Noise. Two threshold optimizing
methods are discussed in this section.
110
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-119-320.jpg)
![An Adaptive Threshold (AT) method is reported in [11]. AT
is the threshold which gives maximum signal to impulsive
noise ratio (SINR) according to (6).
max( ( ))AT SINR T (6)
AT for an OFDM signal is found by applying Blanking
nonlinearity to IN affected received symbols at all thresholds
varying from 0 to 15 with an increment of 0.1 and calculating
SINR each time. The threshold which gives maximum SINR
is denoted by AT as shown in Fig. 3.
0 5 10 15
0
5
10
15
20
Threshold
SINR
AT at max(SINR)
Fig. 3. SINR after applying blanking nonlinearity with respect to threshold
and AT, SIR=-15 dB, p=0.01.
Another optimized threshold is known as Dynamic Peak
Based Threshold (DPBT) [12]. In DPBT estimation method
the peak of the OFDM signal is calculated before it is
transmitted through a noisy channel. At OFDM receiver,
blanking nonlinearity scheme is used with threshold equal to
the peak of the original signal as shown in equation (7).
max( ( )), 0 k N-1kDPBT s dB (7)
Where DPBT denotes the threshold in Dynamic Peak
based Threshold Estimation Method.
Both of the optimized thresholds; AT and DPBT cannot be
used practically as it is impossible to find the exact peak of
transmitted signal and maximum SINR at OFDM receiver
[12].
B. The Proposed Median Based Threshold (MBT)
Calculation Method
Due to practical shortcomings of the techniques discussed
above, a new technique is therefore devised to calculate an
optimized threshold, known as MBT, by using the distribution
characteristics of the received signal which are its peak and
median. Fig. 4 illustrates the method to calculate MBT.
Figure 4. Method to calculate Median Based Threshold.
MBT is given as:
max( ) ( )
,0 k N-1k kr median r
MBT
(8)
Where kr is received signal, max(rk) denotes the peak of the
received signal and is a threshold slope to be fixed at the
receiver. is set to achieve high SINR. Its optimum value
lies in the range for 4 to 6. In our simulations, it is kept 4
which was found to be optimized by using hit and trial
method. Setting an optimal value of is a better approach as
compared to selecting a fixed optimal threshold because a
threshold value may lie from 1 to 15 but the range of β is
comparatively lesser and does not vary with different
probabilities of IN occurrence, unlike a fixed threshold.
Performance comparison of the three methods is shown in
the next section.
IV. SIMULATION RESULTS
The OFDM system is simulated using MATLAB with 16-
QAM modulation, SIR=-15dB, number of subcarriers, N =
10^5, and
22
(1/ 2) [ ]i kE g . Probability (p) of presence of
IN is taken as p=0.01, which suggests that 1% of the received
samples are corrupted by Impulsive Noise. The output SINR is
obtained by
2
2
[ ]
[ ]
k
k k
E s
SINR
E y s
(9)
Fig. 5 shows the SINR achieved by three optimized
111
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-120-320.jpg)
![5.5 6 6.5 7 7.5 8 8.5 9
10
-0.4
10
-0.3
SINR(dB)
SER
AT
DPBT
MBT
Fig. 8. SER with respect to SINR after applying blanking nonlinearity using
AT, DPBT and MBT, p=0.01.
16 16.5 17 17.5 18 18.5
10
-4
10
-3
10
-2
SINR (dB)
SER
AT
DPBT
MBT
Fig. 9. SER with respect to SINR after applying blanking nonlinearity using
AT, DPBT and MBT, p=0.03.
V. CONCLUSION
Two methods of optimizing thresholds of blanking
nonlinearity at OFDM receivers are reviewed. Both of the
methods are purely theoretical and infeasible to use in
practical OFDM receivers. One of the methods suggested peak
of transmitted OFDM signal as the optimized threshold but it
only worked for a single probability of impulsive noise
occurrence and did not give high SINR at other probabilities.
A new method of calculating optimized threshold based on
peak and median of the received signal samples is proposed
which can be used practically. This method results in higher
SINR and lower SER than the method which calculates
threshold by determining the original signal peak for different
probabilities of Impulsive Noise occurrence. The proposed
method is sometimes incapable of achieving the maximum
SINR as that achieved by adaptive threshold optimization
method. However, the SINR achieved by this method is
always close to the maximum and acceptable in practical
scenarios.
Acknowledgment
We gratefully acknowledge Training Division, HR Dte
Gen, Pakistan Space and Upper Atmosphere Research
Commission for their kind coordination in publishing the
article.
References
[1] S. H. Han, and J. H. Lee, “An overview of peak-to-average power ratio
reduction techniques for multicarrier transmission,” IEEE Personal
Communications, vol. 12, no. 2, April 2005, pp. 56–65.
[2] X. Liu, M. Liang, Y. Morton, P. Closas, T. Zhang, and Z. Hong,
“Performance evaluation of MSK and OFDM modulations for future
GNSS signals,” GPS Solutions, vol. 18, no. 2, January 2014, pp. 163-
175.
[3] A. Vaneli-Corali, G. E. Corazza, G. K. Karagiannidist, P. T.
Mathiopoulosl, D. S. Michalopoulost, C. Mosquera, S. Papaharalabost,
and S. Sca1ise, “Satellite Communications: Research Trends and Open
Issues,” International Workshop on Satellite and Sapce Communication
(IWSSC’07), IEEE, 2007, pp. 71-75.
[4] S.V. Zhidkov, “Impulsive noise suppression in OFDM-based
communication systems,” Consumer Electronics, IEEE Transactions,
2003, vol. 49, no. 4, pp. 944-948.
[5] M. Katayama, T. Yamazato, & H. Okada, “A mathematical model of
noise in narrowband power line communication systems,” IEEE Journal
on Selected Areas in Communications, vol. 24, no. 7, 2006, pp. 1267-
1276.
[6] Y. Wu, “Performance comparison of ATSC 8-VSB and DVB-T
COFDM transmission systems for digital television terrestrial
broadcasting,” IEEE Transactions on Consumer Electronics, ,vol. 45, no.
3, 1999, pp. 916-924.
[7] S.V. Zhidkov, “Analysis and comparison of several simple impulsive
noise mitigation schemes for OFDM receivers,” IEEE Transactions on
Communications, vol. 56, no. 1, 2008, pp. 5-9.
[8] U. Epple, K. Shibli, and M. Schnell, “Investigation of blanking
nonlinearity in OFDM systems,” Iternational Conference on
Communications (ICC), June 2011, pp. 1-5.
[9] S.V. Zhidkov, “Performance analysis and optimization of OFDM
receiver with blanking nonlinearity in impulsive noise environment,”
IEEE Transactions on Vehicular Technology, ,vol. 55, no. 1, 2006, pp.
234-242.
[10] S.V. Zhidkov, “On the analysis of OFDM receiver with blanking
nonlinearity in impulsive noise channels,” Proceedings of International
Symposium on Intelligent Signal Processing and Communication
Systems (ISPACS 2004), November 2004, pp. 492-496.
[11] U. Epple & M. Schnell, “Adaptive threshold optimization for a blanking
nonlinearity in OFDM receivers,” Global Communications Conference
(GLOBECOM 2012), IEEE, December 2012, pp. 3661-3666.
[12] E. Alsusa & K. M. Rabie, “Dynamic peak-based threshold estimation
method for mitigating impulsive noise in power-line communication
systems,” IEEE Trans. on Power line, vol. 28, no. 4, October 2013, pp.
2201-2208.
This article is sponsored by Pakistan Space and Upper Atmosphere Research Commission
113
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-122-320.jpg)
![ AMTS overawed all the problems that occur from
IMTS and works on 900 MHz
In Norwegh, OLT which worked on 160 MHz VHF
band was the first land mobile telephone network. In
regularity inflection it worked on 160 -162 MHz for
the mobile portion and in base station it is 168-
170MHz.
MTD include high speed internet structure with fixed
wireless facility that deals constantly on internet
access.
III. FIRST GENERATION (1G)
The 1st
genera of mobile interaction was presented in the
1980 and analog transmission was used as communication.
The 1st
mobile system (1979) developed by Nippon Telephone
and Telegraph (NTT) in Tokyo, Japan. Mobile Telephones
(NMT) and Total Access Communication Systems (TACS)
were two most popular system of analogue system [17].
Frequency division multiple access (FDMA) with channel
capacity of 30 KHz and frequency band 824-894 MHz use
incidence variation system for radio communication based is
derived from a technology called Advance Mobile Phone
Service (AMPS) [14, 15].
Fig 1. 1G Mobile Phone [6]
The initial step of mobile communication is NTM. It
is in the result of heavy and increased usage of physical
cellular communication. The variations of NMT are NMT-450
and NMT-900. The figures specifies usage of band rate
correspondingly. NMT-900 contain other Channels as
compared to NMT-450. NMT cells cast-off full duplex
program. Permitting in real-time acceptance and vocal sound
communication [14].
AMPS was a reporter cellular telephone framework typical
creates in 1983 by Bell Labs, America. It was an initial genera
mobile phone program which use isolated frequencies in place
of individually discussion. It was changed commencing more
seasoned arrangements by "back-end" setup of calls usefulness
that permitted bigger telephones quantities to be bolstered
concluded geological region. In 1983, an aggregate of 40MHz
of range in the 800 MHz band was assigned by FCC for
AMPS. In 1989, additional 10 MHz has been charged to
AMPS because of ascent of cell framework limit. Uplink
frequency for AMPS remains 824- 849MHz and downlink
frequency stays at 869-894MHz. Independently AMPS
frequency is 30 KHz wide [16].
Fig 2. Architecture of Advance Mobile Phone Services (AMPS) [6]
IV. SECOND GENERATION (2G)
After first genera similarity mobile statement system 2G
mobile system was announced around 1991.concept of 2G
based on many base stations where each station circulated
consistently over the world to communicating with the users.
To converse more and more users various access procedures
are used i.e. FDMA, TDMA, and CDMA [10]. 2G technology
make procedures of compression decompression procedure
(codec) and family members of this generation are 2G (GSM,
GPRS, EDGE).
Digital signal usage among recipients and tower build the
framework ability in two forms
1) Digital sound data be able to smashed and multiplexed
proficiently in this way consenting more data is delivered into
the equal quantity of radio data transmission.
2) Less radio power is released while using digital
systems. Therefore creation of small size of and altering better
mobiles in similar quantity of space. Two standard was used
for 2nd
genera technology – CDMA and TDMA on center of
multiplexing. In drill, TDMA and CDMA systems share with
FDMA. TDMA expression is utilized to assign frameworks,
which partition the channel into frequency openings, later on
isolate every rate space into various time openings. Also,
CDMA is truly a cross breed of CDMA and FDMA where the
station is initially partitioned into recurrence openings.
115
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-124-320.jpg)
![Fig 3. 2G Mobile Phone [6]
GPRS is a cell phone wireless technology established
amongst prototype, 2G, and predecessor, 3G) [12]. It’s a radio
innovation for GSM systems that upgrades bundle changing
over conventions, set up period for ISP systems, choice to pay
according to the sent data, instead of association period. In
packet transferring method, material to be delivered is
damaged into packets, size of the packets is in kb, after that
these are moved through network among various ends, each
data packet contain address of destination. Data transmission
and network connection is the provision of GPRS.
EDGE is an ordinal cell phone technique that actions as a
bolt-on enrichment to General Packet Radio Service (GPRS)
networks. It is ideal as compared to GSM because it carries
packet switch data and circuit switch data [11]. Black berry
N97 and N95 mobile phones are famous because of EDGE
technology. By using EDGE technology, no additional
hardware and software resources are required for making
EDGE technology work. Any of the extra expense are not
required for developing such technology [13].
Fig 4. GPRS Architecture [6]
V. THIRD GENERATION (3G)
3G is the innovative generations for cellular interaction
facilities, its services area centered on procedural values of
IMT-2000 containing the consistency, accuracy and also speed
(rate in which data is transferred) that is minimum 200kb per
second [11]. Outside of cell phone, more data transfer ratio
permitted 3G links in personal computers, digital gaming
equipment, tablets, and smart phones including all devices that
are able to take benefit by using high and fast quality of
internet usage. Working capacity of devices and functionality
of devices can be increased many times with the usage of fast
and reliable internet connection. Better safety and less threat is
offered while connecting and transferring data to other
wireless devices by using 3G [2, 3]. Three basic technologies
are covered by 3G that is CDMA2000, TD-SCDMA - Time-
division Synchronous Code-division Multiple Access and W-
CDMA (UMTS) Wideband Code Division Multiple Access.
The vital role of 3G knowhow is, it offers high transfer of data
transmission and improved speed. Packet-switching technique
is used in 3G that is considered more realistic and also much
faster as compared to the previous used technologies and
techniques, issue with it is it requires a totally different
groundwork as compared to the 2G systems.
Fig 5. CDMA Architecture [6]
Mobile communication market is increasing on a
very fast pace, usage of mobile phone industry is increasing in
leaps and bounds. Almost one billion customers globally are
benefiting from cellular phone industry, in which two-thirds
are GSM customers. One of the firmest growing wireless
technology is known as CDMA, it is developing its
subscribers base signify cantle around the world. Nowadays
CDMA subscribers are crossing the line of 200 million users.
3G networks enable network workers to deal customers an
extensive variety of latest progressive facilities and also
realizing better and superior connection ability through
healthier spectral efficiency.
Fig 6. 3G Mobile Phone [6]
116
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-125-320.jpg)
![VI. FOURTH GENERATION (4G)
4G technology uses 3G as a inheritor, developed for the
broadband capabilities with the services of 3G to enable the
high data rate streaming i.e.20 Mb/s per customer, Quality of
services, Fast Speed, Faster Response Time, Routing
efficiency, Higher Bandwidth and Lower Power consumption.
Speed for low mobility is 1 gigabit per second (Gbit/s) and
100 megabits per second (Mbit/s) for peak speed. NTT Do Co
Mo was first successor to achieve. Downloading of packet
transmission at a moving speed of almost 20km/h is 1Gbps in
Tokyo, Japan on June 23rd, 2005 [2]. 4G technology is
combination of IP-Based voice, Seamless access,
Personalization, Quality of services. Its main purpose is to
access network anytime, anywhere and anyhow. 4G wireless
communication technology is furthermore discussed as
“MAGIC”. MAGIC word stands for mobile multimedia, any-
where, global mobility solutions over, integrated wireless,
customized services.
Need for this platform is to integrate pervious mobile
technologies entirely which are present that is GSM, GPRS,
IMT-2000, Wi-Fi, Bluetooth,
WCDMA, HSDPA, CDMA2000 and EVDO, provide
correspondence to user expectations for various services.
Mobile WiMAX regularities and Long Term Evolution (LTE)
rules were deployed for 4G but failed to satisfy the IMT-
Advanced requirement, may well though measured as 4G.
Later on LTE Advanced (latest release of LTE) standardized
by the 3GPP (Third Generation Partnership Project) and
802.16m uniformed by IEEE (that is WiMAX). Demand of
IMT-Advanced for the generations of interactive mobile
services by providing a global platform is increased to gain
unified messaging, quicker information admittance, improved
roaming competencies and broadband multimedia [16].
A. LTE Advanced
LTE is not such a new technology but some essential
enhancements in LTE to improve existing LTE network. It
was officially candidate of ITU-T (International
Telecommunication Union- Telecommunication), meeting the
requirements of the IMT-Advances standard and also
standardized by 3GPP. LTE Advanced has 30bps/Hz peak
spectrum downlink value and 15 Bps/Hz peak spectrum uplink
value, can be enhanced if multiple-input multiple-
output (MIMO) (that is antenna arrays) remain used. Both
coordinated multipoint (COMP) and MIMO remained
recognized such as the crucial procedures for LTE [2].
Fig 7. LTE Advance Architecture [17]
VII. FIFTH GENERATION (5G)
5G (Mobile and wireless Network) can make the perfect
wireless real world without any limitation is known as World
Wide Wireless Web (WWWW).
Beyond the 4G/IMT-Advanced standards 5G symbolizes the
next utmost chief level of cellular phone technology genera.
5G is still not considered and used as an official term on
behalf of any specific requirement and design and also not
publically used by telecommunication companies or
standardization such ITU-R (International Telecommunication
Union-Recommendation), 3GPP, WiMaxForum (Worldwide
Interoperability for Microwave Access) [2]. Every new release
of technology enhances the system performance with some
advance capabilities. 5G technology must be something else
than previous technologies, whose purpose is not only to
increase the throughput but also consider the other
requirements.
Improved coverage of geographical zone and high
rate for data transfer at the edge of the mobile phone
Truncated power and battery consumption
Accessibility of numerous data transmission tracks
Nearby 1 Gbps data rate is definitely thinkable
Additional Security
Energy effectiveness and spectral efficiency are good
[17]
117
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-126-320.jpg)
![Fig 8. 5G Mobile Phone [17]
5G mobile technology will be powerful and
demanding in future due to including most powerful features.
Also it will change the concept of using mobile phones with
more bandwidth. Customers certainly not have experience of
such great and fast tracking technology. The architecture of
the fifth generation consists of cellular networks
interoperability (contains of a customer station and number of
autonomous radio access technologies (RAT)) and wireless IP
based model [5]. User terminal has a significant role in this 5G
technology system. Proposed 5G architecture is shown in fig
8. 5G technology deal with camera, large phone memory,
MP3 recording, dialing speed, video player, audio player and
many creative, innovative and imaginative technologies [13].
Fig 9. Functional Architecture for 5G- Network [18]
VIII. SIXTH GENERATION (6G)
The 6G technology will combine the 5G satellite network
and mobile wireless system. Satellite network include
telecommunication satellite, navigational satellite, imaging
satellite, environmental information system. These system
used for data, voice, video broadcasting and internet; weather;
environmental information Collection; and global positional
system (GPS) respectively [2]. At the same time Fifth
generation will focus on user terminals but it is necessary for
terminals to have access of different wireless technologies.
Roaming and handoff will be the big issue of 6G because of
the different network stations. Solving the problem of handoff
and roaming is still a question. But the charges of mobile call
will be relatively high [8].
IX. UPCOMING GENERATIONS
Upcoming technologies will provide the real image of
wireless world with less boundaries. Artificial intelligence and
ubiquitous network will provide universal computing from
anywhere, anytime, and any device. User can instantly linked
to several technologies and perfectly move between them.
X. COMPARISON
We compare all the technologies of Mobile Technology in
term of data bandwidth, standards, technology, multiplexing,
switching see table 1. [6]
XI. CONCLUSION
Wireless communication technologies has captured the
market since its first start. This study provides the better
understanding of the previous and present technologies along
with their performances, portals, advantage and disadvantages
of one generation to other and also predict incoming future
technologies.
.
REFERENCES
[1] F. R. Y. Chengchao Liang, "Wireless Virtulization for Next Generation
Mobile Cellular Networks," in IEEE Wireless Communications,
February 2015.
[2] P. Sharma, "Evolution of Mobile Wireless Communication Networks-
1G to 5G as well as Future Prospective of Next Generation
Communication Network," in International Journal of Computer
Science and Mobile Computing, August 2013.
[3] E. M. I. A. E. U. M. A. Engr. Muhammad Farooq, "Future Generations
of Mobile Communication Networks," in Academy of Contemporary
Research Journal, January 2013.
[4] A. G. Roopali Sood, "Digital Society from 1G to 5G: A Comparative
Study," in International Journal of Application or Innovation in
Engineering & Management (IJAIEM), February 2014.
[5] R. Rajan, R. Baweja and R. Kumar, "Review on Different Multiple
Access Technique Used in Wireless Communication," in International
Journal of Research (IJR), November 2014.
[6] P. S.Venkata Krishna Kumar, "A Study of Wireless Mobile
Technology," in International Journal of Advanced Research in
Computer Science and Software Engineering, January 2014.
[7] V. G. APANA BHAT, "EVOLUTION OF 4G: A STUDY,"
International Journal of Innovative Research in Computer Science &
Engineering (IJIRCSE), February 2015.
[8] A. S. J. R. Sarmistha Mondal, "A Survey on Evolution of Wireless
Generations 0G to 7G," International Journal of Advance Research in
Science and Engineering- IJARSE,.
[9] A. A. J. I. Nabeel ur Rehman, "3G Mobile Communication Networks,"
EXPLORE SUMMER 2006, 2006.
[10] N. Schmitz, "The Path To 4G Will Take Many Turns," 1 March 2005.
[Online]. Available: http://electronicdesign.com/4g/path-4g-will-take-
many-turns.
[11] A. G. R. S. O. Z. Xichun Li, "The Future of Mobile Wireless
Communication Networks," 2009 International Conference on
Communication Software and Networks, 2009.
[12] A. R. Mishra, Fundamentals of Cellular Network Planning and
Optimisation: 2G/2.5G/3G... Evolution to 4G, John Wiley & Sons
©2004, 2004.
[13] A. F. Molisch, Wireless Communications, Wiley, 2010.
[14] F. ADACHI, "Wireless Past and Future--Evolving Mobile
Communications Systems," IEICE TRANSACTIONS on Fundamentals
of Electronics, Communications and Computer Sciences, 2001.
118
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-127-320.jpg)
![[15] D. Y. L. D. J. S. D. Amit Kumar, "Evolution of Mobile Wireless
Communication Networks: 1G to 4G," International Journal of
Electronics & Communication Technology, December 2010.
[16] A. Pashtan, "wireless terrestrial communication: cellular telephony,"
TELECOMMUNICATION SYSTEM AND TECHNOLOGIES, 2006.
[17] G. M. K. A. Arun Agarwal, "The 5th Generation Mobile Wireless
Networks- Key Concepts, Network Architecture and Challenges,"
American Journal of Electrical and Electronic Engineering,, vol. 3, no.
2, pp. 22-28, February 2015.
[18] [Online].Available:http://www.prismatelecomtesting.com/products/prod
ucts-overview/radio-interface-simulation/uesim/.
TABLE 1. COMPARISON OF MOBILE TECHNOLOGIES
Generation Features 1G 2G 3G 4G 5G
Years 1980s 1990s 2000s 2010s 2020s
Data Bandwidth 2kbps 64kbps 2Mbps 200Mbps 1Gbps
Standards AMPS TDMA, CDMA,GSM,
GPRS
WCDMA Single unified
standards
Single unified standards
Technology Analog cellular Digital Cellular Broadband with
CDMA, IP
Technology
Unified IP & Seamless
combination of
broadband,
LAN,WAN & WLAN
Unified IP & Seamless
combination of
broadband, LAN,WAN
& WLAN,WWWW
Service Mobile technology
(voice)
Digital Voice, SMS,
Higher Capacity
Packetized
Integrated high
quality audio, video
& data
Dynamic information
access, wearable
Devices
Dynamic information
access, wearable
Devices with AI
capabilities
Multiplexing FDMA TDMA,CDMA CDMA CDMA CDMA
Switching Circuit Circuit & Packet Packet All Packet All Packet
119
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-128-320.jpg)
![Intelligent Detection of Distributed Flooding Attack
in Wireless Mesh Network
M.Altaf Khan
Institute of Information Technology
Kohat University of Science & Technology
Kohat, Pakistan
altaf_apc@yahoo.com
Shafi Ullah Khan
Institute of Information Technology
Kohat University of Science & Technology
Kohat, Pakistan
skkust@hotmail.co.uk
Amjad Mehmood
Institute of Information Technology
Kohat University of Science & Technology
Kohat, Pakistan
amjadiitkust@gmail.com
Zeeshan Iqbal
Institute of Information Technology
Kohat University of Science & Technology
Kohat, Pakistan
Zeeshan.zn@gmail.com
Abstract—Multi-hop wireless mesh network (WMN) is a
growing and promising technology that fulfills the requirements
of next generation wireless networking by providing an extensive
assortment of applications that were not assumed to be supported
by a wireless network. In WMN, a Gateway is the only way of
integrating it with other wireless and wired networks. Due to this
critical role, gateways in WMNs are targeted by diverse security
threats. A distributed denial of service (DDoS) attack is one of
the prime security threats to WMN. DDOS attack is an explicit
attempt to prevent the legitimate users from accessing their
service. In a distributed flooding attack, gateway can be easily
exploited by the attackers with the aim to reduce the available
network bandwidth for the valid users. Hence network
performance is reduced and also destruct the services provided to
the end users. Numbers of detection mechanisms have been
provided by researchers in this regard. However, due to
advancements in the attacking tools and techniques, more
intelligent efforts are required in terms of precision and
competence, to protect the WMNs from such attacks. In this
paper, an intelligent mechanism called Distributed Flooding
Attack Detector (DFAD) is proposed with the aim to detect
distributed flooding attacks in WMNs. Network simulator (NS2)
is being used to evaluate the performance of DFAD.
Keywords—Wireless Mesh Network, Distributed Denial of
Service Attack, Mobile Ad-hoc Network, Artificial Neural
Network, Back-Propagation
I. INTRODUCTION
A multi-hop and multi-radio wireless mesh network
(WMN) is a decentralized, dynamically self-organized and
self-configured network. Connections among the nodes in
WMN establish without any central controlling body using
partial or complete mesh topology [1]. Two types of nodes are
existed in a typical WMN i.e. mesh node (MN) and mesh
router (MR). MN may either be static or mobile having one
wireless interface. MRs forms a multi-hop backbone of the
WMN by providing self-configured and self-healed links.
Sometimes, in WMN, a MR may have both wired and wireless
interfaces which enables it to provide functionality of gateway.
Unlike conventional routers, MR exhibits some additional
routing functions to bear mesh networking. Moreover, due to
multi-hop support, mesh routers provide more coverage with
low transmission power than conventional routers. WMNs can
be classified into three main architectures i.e. infrastructure-
based, infrastructure-less and hybrid WMN [2]. Infrastructure-
based WMN is composed of both MNs and MRs.
Infrastructure-less WMN seems to be a pure mobile ad-hoc
network (MANET), because it is consisted of only MNs that
may also carry out routing and self-configuration to form a
peer to peer wireless network. Hybrid WMN is composed of
MNs and MR, where MRs are used to form backbone of WMN
and MNs offer routing abilities in order to achieve an enhanced
coverage and connectivity in the interior WMN. In both
infrastructure-based and hybrid WMNs, through mesh gateway
(MG), a node in WMN can communicates with other existing
networks. Hence, MG plays a vital rule for instigation or
termination of traffic in WMN. So it is mostly targeted by
intruders during different attacks. Once MG is being
compromised, all the provided services of WMN are affected
and hence degrades its performance.
WMNs have made life of its users easier. However, some
serious security threats may be faced by the WMN that may
compromise the confidentiality, integrity or availability of a
WMN [3]. Confidentiality is being compromised through
several attacks like eavesdropping, traffic analysis, corrupt
access point, war driving attack, and brute force attack. Black-
hole, worm-hole, grey-hole etc. are the attacks through which
integrity of WMNs are being compromised. Availability is
compromised by many DOS attacks including TCP SYN-
flooding attack, UDP-flooding attack and smarf attack, where
intruders try to prevent normal services or to make the
network resources unavailable for the legitimate users. When
a DOS attack is implemented by thousands of attackers
simultaneously, with the aim to increase the severity of attack,
it is identified as DDOS attack. An intruder launches such
attack by sending a massive bulk of ineffectual packets
concurrently towards victim, through huge number of
compromised nodes also known as zombies [4]. These
zombies may either be internal or external to the victim’s
network. Distributed flooding attack (DFA) is a one of the
classical approach to launch DDOS attack in WMNs using
120
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-129-320.jpg)
![tools like Trinoo, Tribe Flood Network (TFN), Shaft, Knight
and Trinity [5].
Anomaly based DDoS attack is being detected in [6].
Features of malicious packets are analyzed using radial base
function neural network (RBFNN). A vector of seven features
is being used for activation of RBFNN and classification of
incoming packets flows into classes of attack traffic or normal
flows. Proposed method is evaluated through datasets of
UCLAD. Proposed method classified the incoming traffic into
class of normal or attack traffic, but it cannot classify and
identify the type of attack. A statistical approach, based on
values of numerous features is implemented in [7] to detect
attack traffic. Proposed method used a module for features
extraction and then extracted features are stored in a database
for identification and classification of normal and malicious
packets flows. A novel approach is presented in [8] to detect
flooding and IP spoofing attack. Prevention of flooding attack
is accomplished through considering clock values of
individual node of network. A combined data mining
approach is used in [9] against DDoS attack for detection of
protocol anomaly. In the proposed method for extraction of
features network traffic is utilized and then a classification
algorithm based on data mining is used to cluster incoming
traffic into normal and attack flows.
For detection of such attack numerous techniques are
proposed so far, but the techniques that are based on
intelligent schemes are proved to be more proficient
alternatives in this regard [10,11,12] . Still there is a need to
propose more mechanisms to handle this attack. Our proposed
mechanism is more intelligent and reliable then the
mechanism proposed earlier for detection of DFA in WMNs.
II. PROPOSED MECHANISM
Due to self-configuration and dynamic changes in
topology of WMN, ANN is considered to be the most suitable
technique amongst all other intelligent techniques to detect
distributed flooding attack [5]. Our proposed mechanism i.e.
distributed flooding attack detector (DFAD) is based on ANN
technique and is adaptive in nature.
ANN used the concepts of biological neural networks
found in human brain. With the assistance of this network a
brain involves millions of interconnected neurons to fully
control the human body. Fig.1 depicts the basic structure of a
biological neuron. It illustrates some important components of
a neuron. Actual cell body of the neuron is constructed by
Soma. It is composed of a nucleus and plasma. Information
regarding to innate traits is provided by the nucleolus and
plasma generates material required by neurons. Dendrites and
Axons are used during communication with other neurons.
Through synaptic terminal an Axon spreads out received
signal to other neurons [13].
Following the same concepts, ANN is a network of
interconnected nodes (neurons) that provide required solution
after performing parallel processing. Structure of earliest ANN
called perceptron, is given in the Fig.2.
Using equations (1,2,3), If Sum of the product of binary
inputs (Z1, Z2…. Zn) and their associated weights (W1,
W2…....Wn) is found greater or equivalent to “0” then “f” i.e.
an activation function will turn out “1” as output if not, “0”
will be the output of the perceptron [14].
(1)
( ) (∑ ( )) (2)
( ) { (3)
Two types of structures are existed in ANNs i.e. feed-
forward and feed-back neural networks. Main difference
between both structures is the flow of signals. Feed-forward
neural network allows information flow in just forward
direction while a feed-back neural network allows information
to be flowed in both forward and backward directions[15].
These structures are mainly distributed in three main layers:
input-layer (IL), hidden-layer (HL) and output-layers (OL). IL
receives input for the network and the OL provides the results
of the network. HL offers the neural network with its aptitude
to generalize. Moreover, HL may be composed of more than
one layer. There may be different number of neurons in each
layer. As quantity of layers and nodes in HL increases,
complexity of ANN will be increased. On the other hand,
probability of accurate learning is increased. A feed-forward
ANN is illustrated in Fig.3.
Soma
Dendrites
Axon
Nucleus Synaptic
Terminals
Output (0/1)
X1
X2
X3
Xn
Fig.1. Structure of a biological neuron
Fig.2. Structure of a perceptron
121
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-130-320.jpg)
![Learning of ANN can be performed through either
supervised or un-supervised learning algorithms. In
supervised learning, an output is provided with each input
during training while in unsupervised learning algorithms,
target outputs are not provided with input data during
training. Hence ANNs perform data compression or
clustering to categorize patterns of the same attributes into
same output clusters [16].
In feed-forward ANNs, most commonly used supervised
learning algorithms may include delta rule, perceptron and
back-propagation (BP). For learning a feed-forward ANN,
BP algorithm is most commonly used amongst entire
algorithms.
Delta rule and perceptron algorithms are useful to
resolve merely linear problems, while BP algorithm is able
to provide learning in complicated non-linear problems.
During training, in BP based ANN, signals are transferred in
forward direction i.e. from input to OL. Error between actual
and target output in the OL is calculated in each pass. In BP,
learning process includes forward-pass and backward-pass.
Input data is moved from IL to OL in forward-pass. Most
commonly sigmoid activation function is used in BP
algorithms to generate results from each HL and OL node by
using following equation (4). Output of this sigmoid
function lies between “0” and “1”
( ) (4)
Error (E) between target output (T) and actual output (O)
is calculated at the end of each forward-pass. Gradient
descent technique is used to reduce this error ‘E’ between Oi
with respect to Ti. The value of E, gradient descent and
required changes in the weights of each connection values
are computed through following equations respectively.
∑ ( ) (5)
(6)
( ) (7)
(8)
In equation (6), ‘GrD’ represents value of gradient
descent, its value concludes that either to raise or reduce
value of weights and biases. ‘ ’ denotes momentum and ‘ ’
represents the learning rate. In equation (7) and (8), and
represents the desired change in weights associated with
hidden-to-output layers or input-to-hidden layer connections
and biases respectively in the backward-pass.
In our DFAD, we used multiple layer perceptron (MLP)
having three inputs nodes in the IL, because we aimed to
check the MG after each five seconds. Each input value
represents received number of packets. After trying different
number of nodes and different number of layers in the HL,
finally we got our results with higher detection rates and
lower false rates, using one HL having four nodes in it. As
we are interested to distribute the incoming flows at MG into
three different categories therefore we used three nodes in
the OL to generate one of the desire output i.e. [1,0,0],
[0,1,0], [0,0,1] to represent normal, low distributed flooding
(LDFA) attack and sever distributed flooding attack (HDFA)
flows respectively.
During training of DFAD, a vector of eight (08) input
values is provided. Arrangement of the input vector is
[4,3,7,0,0,1], where [4,3,7] are the amount of packets
received in each second and the remaining values of the input
vector i.e. [0,0,1] represents the target output for the given
input values. We used three different target outputs with
each input vectors i.e. [0, 0, 1], [0, 1, 0] and [1, 0, 0], Where
[0, 0, 1] is imagined as normal flows, [0, 1, 0] is considered
to be an intermediate distributed flooding attack and [1, 0, 0]
is expected to be a high distributed flooding attack flows. A
minute detail of training dataset is given in the Table.1.
Input
Vectors
4, 3, 7, 0, 0, 1, 31, 34, 36, 0, 1, 0, 42, 51, 65,
1, 0, 0, 4, 6, 12, 0, 0, 1, 33, 26, 22, 0, 1, 0,
46, 54, 57,1, 0, 0, 16, 6, 17 0, 0, 1, 37, 35,
46, 0, 1, 0, 54, 58, 56, 1, 0, 0,……..…..
Weights
and
Biases
0.11, 0.21, 0.13, 0.14, 0.25, 0.36, 0.17, 0.8,
0.19,-3.0, -4.0, -2.0, 1.6, 1.8, 1.3, 1.6, 1.2,
1.1, 1.7, 2.1,2.3,-2.3, -7.0, -5.0
Assortment of values for initial weights and biases is an
actually very tough job. Numbers of forward and backward
passes to generate the desired outputs from the network depend
on these values.
In training, maximum 100 epochs are used with each input
vector to minimize the error between target output and actual
output. Dataset used to train the network, is consisted of 5000
of input vectors, having all three types of flows, mentioned
above. After completion of training, updated values of weights
and biases are used by DFAD to distribute the input vectors in
desired category in testing phase.
Algorithm of proposed mechanism is given in Fig.4. and
Fig.5. Main difference between training and testing phase is
that, both passes i.e. forward and backward passes of BP
algorithms are used in training phase while in testing phase
only forward-pass is used. Moreover, in testing phase input
vector is entered in the system without target output values.
Inputs
IL HL OL
Table.1. Training dataset
Fig.3. Feed-forward artificial neural network
122
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-131-320.jpg)
![Dataset used to train the network, is consisted of 5000 of
input vectors, having all three types of flows, mentioned
above. In training, maximum 100 epochs are used with each
input vector to minimize the error between target output and
actual output. After completion of training, updated values of
weights and biases are used by DFAD to distribute the input
vectors in desired category in testing phase.
III. SIMULATION RESULTS
NS2 [17] is considered to be a best choice for simulations of
wired and wireless networks therefore we carried out our
simulations in NS2.34. Effectiveness and accuracy of DFAD
is tested by performing simulations on NS2. Following
parameters are set for our simulations.
=======================
Training:
==================================
1. Initialize all weights and biases with small
random numbers
2. For ∀ Input vectors in the training set
3. Input Current pattern and target output to the
network
4. // Propagated the input forward through the
network:
5. For ∀ node in the Hidden layer
i. Calculate the sum of product of
weights and inputs to the node using
Eq.2
ii. Add the biase of each node to the
calculated sum
iii. Calculate the output using Eq.4 for
each node
6. Next
7. For ∀ node in the Output layer
i. Calculate the sum of product of
weights and inputs to the node using
Eq.2
ii. Add the biase of each node to the
calculated sum
iii. Calculate the output using Eq.4
foreach node
8. Next
9. Calculate sum of error between target and
actual output using Eq.5
10. IF ((maximum number of iterations (epochs)
< 100) && (Error>=0.25))
11. // Propagate the errors backward through the
network
12. For ∀ node in the output layer
i. Calculate Gradient value for node in
the output layer using Eq.6
ii. Update each node's weight and biase
values in the output layer using Eq.7
and Eq.8
13. Next
14. For∀ node in the hidden layer
i. Calculate the Gradient of node's in the
hidden layer using Eq.6
ii. Update each node's weight and biase
value in the hidden layer using Eq.7
and Eq.8
15. Next
16. With updated weights and biases repeat
from Step-5
17. endif
18. Next // select next input vector for training
=====================================
Testing:
=====================================
1. Assign all Updated weights and biases
obtained from training phase
2. Input Current pattern to the network
3. // Propagated the input forward through the
network:
4. For ∀ node in the Hidden layer
i. Calculate the sum of product of
weights and inputs to the
node using Eq.2
ii. Add the biase of each node to the
calculated sum
iii. Calculate the output using Eq.4
for each node
5. Next
6. For ∀ node in the Output layer
i. Calculate the sum of product of
weights and inputs to the
node using Eq.2
ii. Add the biase of each node to the
calculated sum
iii. Calculate the output using Eq.4
for each node
7. Next
8. // Output Layer: Node1’s output: out1,
Node2’s output: out2,Node3’s output: out1;
9. IF(out1>out2&&out1> out3)
10. printf (“Normal flow”);
11. elseif (out2> out1&&out2> out3)
12. printf(“LDFA-Traffic”);
13. else
14. printf(“HDFA-Traffic”);
15. endif
Fig.5. Testing phase
Fig.4. Training phase
123
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-132-320.jpg)
![IV. CONCLUSION AND FUTURE WORK
In WMNs, mesh gateways are mostly susceptible to
DDOS attacks. Numerous techniques are proposed to detect
distributed flooding attack in WMNs, but still there is a need
of an intelligent security mechanism to avoid the influence of
intruders on WMNs. An intelligent mechanism based on
ANN, is being proposed in this paper to classify incoming
flows at gateway into three different categories. Simulations
results verified that DFAD is efficient and precise in
classifying the incoming flows. In future, DFAD will
differentiate flash crowd form attack flows. Moreover, we will
also identify and trace back source of the ongoing distributed
flooding attack.
ACKNOWLEDGEMENT
The authors extend their gratitude to the Research Centre,
Institute of IT, Kohat University of Science & Technology and
the Higher Education Commission of Pakistan, for funding
this research work.
REFERENCES
[1] S. Khan, K. K. Loo, and Z. Din, "Framework for intrusion
detection in IEEE 802.11 wireless mesh," International Arab
Journal of Information Technology, vol. 7, no. 4, pp. 435-439,
2010.
[2] D. Benyamina, A. Hafid, and M Gendreau, "Wireless mesh
networks design — a survey," IEEE Communications Surveys &
Tutorials, vol. 14, no. 2, pp. 299-310, 2011.
[3] F. Xing and W. Wang, "Understanding dynamic denial of
service attack in mobile ad hoc networks," in IEEE Military
communication conference (MILCOM), 2006.
[4] H. Beitollahi and G. Deconinck, "Analyzing well-known
countermeasures against distributed denial of service attacks,"
Computer Communications, vol. 35, no. 11, pp. 1312–1332,
2012.
[5] B B Gupta, Joshi, and Manoj , "Distributed denial of service
prevention echniques," International Journal of Computer and
Electrical Engineering, vol. 2, no. 2, pp. 268-276, 2010.
[6] K Reyhaneh and F Ahmad, "An anomaly-based method for ddos
attacks detection using rbf neural networks," International
Conference on Network and Electronics Engineering IPCST,
vol. 11, pp. 44-48, 2011.
[7] Thwe Thwe Oo and Thandar Phyu, "A statistical approach to
classify and identify DDoS Attacks using UCLA dataset,"
International Journal of Advanced Research in Computer
Engineering & Technology (IJARCET), vol. 2, no. 5, pp. 1766-
1770, 2013.
[8] D. J. Jingle and E. B. Rajsingh, "Defending IP spoofing attack
and TCP SYN flooding attack in next generation multi-hop
wireless networks," International Journal of Information &
Network Security (IJINS), vol. 2, no. 2, pp. 160-166, 2013.
[9] Thwe Thwe Oo and Thandar Phyu, "DDoS detection system
based on a combined data mining approach," 4th International
Conference on Science and Engineering, pp. 315-319, 2013.
[10] G. Wang, J. Hao, J. Ma, and L. Huang, "A new approach to
intrusion detection using artificial neural," Expert Systems with
Applications, vol. 37, no. 9, pp. 6225-6232, 2010.
[11] S. Khan and K. K Loo, "Real-time cross-layer design for large-
scale flood detection and attack traceback," Network Security,
vol. 23, no. 5, pp. 9-16, 2009.
[12] D. Novikov, R. V. Yampolskiy, and L. Reznik, "Artificial
intelligence approaches for intrusion," in IEEE Long Island
Systems, Applications and Technology Conference (LISAT
2006), Long Island, NY, 2006.
[13] E. Gelenbe, "Learning in the recurrent random neural network,"
Neural Computation, vol. 5, no. 1, pp. 154–164, 1993.
[14] S. M. A. Burney, M. S. A. Khan, and Dr. T. A. Jilani, "Feature
deduction and ensemble design of parallel neural networks for
intrusion detection system," International Journal of Computer
Science and Network Security (IJCSNS), vol. 10, no. 10, pp.
259-267, 2010.
[15] G. OKE and G. Loukas, "A denial of service detector based on
maximum likelihood detection and the random neural network,"
The Computer Journal, vol. 50, no. 6, pp. 717-727, 2007.
[16] B. B. Gupta and R. C. Misra,M. Joshi, "ANN based scheme to
predict number of zombies in a DDOS attack," International
Journal of Network Security, vol. 14, no. 2, pp. 61-70, 2012.
[17] The network simulator - ns-2 web page,"http://nsnam.isi.edu/
nsnam/index.php/Main_Page".
Time(Sec)
Packets
HDFA-Pkt.Drop.Rate
LDFA-Pkt.Drop.Rate
Normal-Pkt.Drop.Rate
Fig.10. Packet Drops of UDP flows
125
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-134-320.jpg)
![Touch Panel Based Modern Restaurants Automation
using Zigbee Technology
1,*
Aamir Nawaz,2
Faiz Jalil
Institute of Engineering & Technology
Gomal University
DIKhan, Pakistan
1
engr.aamir09@gu.edu.pk
2
fj_kulachvi@yahoo.com
Abstract
In present scenario, automation is in high demand.
Management of present restaurants and hotels are much
dependent on manpower. This project deals with automation
of restaurants with wireless touch panel based menu systems.
This project reduces manpower required for taking order from
customers. Full menu of all restaurant’s eatable items are
displayed on touch panel for selection. Customers have to put
their orders through touch panels which will be received in
kitchen without any interaction of waiters. Zigbee has been
used for wireless link between touch panels of kitchen and
restaurant tables. PIC microcontroller has been used for
coding of menu display on touch panel and transmitting and
receiving on zigbee modules. Project has been implemented
and tested in software such as Proteus 8. Furthermore,
hardware implementation has been done on PCB layout. By
using embedded system with wireless Technology whole
procedure will built. The order from selected table number go
to server then serve items to selected table customer .We can
speed up the serving and attract customer with non – volatile
memory capable of retaining data for over ten years .By
dumping the code in microcontroller ISP is used
Index Terms—Zigbee, Restaurant automation, Micrcontroller
I. INTRODUCTION
Restaurant business is one of the most profitable
businesses. Therefore, the importance of food serving is of
great significance. Over the years, food and the relative job of
serving has grown so much that need for facilitation and
automation has been increased. industries. Nowadays, when
profit is the prime concern and its measurement has increased
from million to billions, every bit is done to increase profit.
We have made an effort to bring technology into the dining
menu of customers. This research aims to not only improve the
business of restaurants but also to incorporate the essence of
science in dining menu.
Arun in [1], implemented a train anti-collision and level
crossing system based on zigbee technology. He has created
four modules which are train system module, control center
system module, signaling post system module and level
crossing at gate module. He has simulated proposed system in
Proteus electronic simulation package.
Ramasamy [2] has proposed a fire avoiding system on
running train. In his proposed system, train driver will know in
case of fire in any section of train and stop it instantly to take
necessary action.
In [3], Blaho proposed a zigbee wireless network to control
speed and other parameters of train model. He has created m-
files and Simulink environment which can be used for teaching
purposes.
In [4], Vinodhini proposed a restaurant system for two way
ordering system. He has provided table of customer with touch
screen and kitchen with liquid color display (LCD). He has
used ARM cortex M3 for simulation.
In [5], Wang has introduced zigbee technology for
establishing wireless sensor networks (WSN) and its
applications in industry and home appliances.
Bhargave in [6], has introduced a system for restaurant
using android technology. He proposed integration of Order
System Kitchen Order Ticket (KOT), Billing System and
Customer Relationship Management System (CRM).
Implementation of such integration has increased quality and
speed of restaurant service.
Wahab in [7], has presented a new development in
restaurant ordering system. He built up network system
integration for connecting customer table with management
panel. He used microcontroller with VB and RS 232
communication port. Proposed system performed well in
network based system.
The proposed system will have a complete Touch Panel
device in restaurants and hotels that will help to reduce the
human interaction and save time. This project implements a
system that is on customer table by having touch panel menu
ordering system. We have one touch technology by selecting
the items from menu list display in LCD on customer table
The basic concept of the project is to reduce human effort
for the customers and also to deliver them with best facilities.
Wireless communication link between customer’s table and
kitchen has been established for placing order on customer’s
side and verifying it on kitchen side. Furthermore, bill details
126
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-135-320.jpg)
![Figure 4: Kitchen end module operational
V. COMPARISON OF PROPOSED SYSTEM WITH PREVIOUS
SYSTEMS
Table 1 shows comparison results on the basis of different
features considered for comparison. For comparison, Pixel
point [9], LRS[10], Eric software[8] and Smart order system[7]
are considered. As it can be seen that proposed system
outperforms all other systems in comparison.
TABLE 1. COMPARISON OF PROPOSED SYSTEM WITH PREVIOUS SYSTEMS
Features Pixel point LRS Eric soft Smart order system Proposed System
Graphics NO NO NO NO YES
ATM Sweeper NO NO NO NO YES
Prices NO NO NO NO YES
Group Orders NO NO NO YES YES
Status of Ordering NO NO NO NO YES
Wireless Network YES YES YES NO YES
Touch Screen YES NO YES NO YES
VI. SIMULATION AND RESULT EXPLORATION
In restaurant, customer’s table module has a touch screen
display in which menu is shown. When this menu is pressed,
full menu of restaurant eatables are shown on screen to be
selected. Once order is selected and forwarded to kitchen, it
uses zigbee transmitter and receiver to deliver that information
to kitchen. Kitchen module will receive order which will be
displayed on LCD. Kitchen person will confirm order by
pressing confirm button or reject any unavailable eatable items
from menu. Then, customer will get information regarding
unavailability of any items and can re-order. Furthermore,
waiter can be called by pressing “Call waiter” button.
The proposed system can be used with line following
robots or conveyer belts to deliver order till customers. This
will reduce more human effort as there will be no waiter
required to deliver order to tables. Other benefits are:
Provision of facility for handling text electronically using
powerful processors to produce elegant and error free
documents.
In addition to storing the organization’s operational data
on disk drive for back up.
With the installed software, product ordering and delivery
was made easier.
The systematic approaches used during each phase of the
software development provides a clear road map that
would be of immense help to anyone carrying out research
work in this area.
VII. CONCLUSION
This research contributes to recent working environment of
restaurants and can change it from previous manual system to
an automated restaurant with a lot of facilities to customers as
well as restaurant management. Furthermore, it will be helpful
in attracting customers to see and avail such facilities. Zigbee ,
as a wireless link between customer’s table and kitchen can be
replaced by any other wireless technologies for larger
restaurants which supersedes zigbee’s range of communication.
Line following robots or conveyer belts can be used to deliver
eatable items to customers instead of waiters.
REFERENCES
[1] Arun P., Sabarinath G., “Implementation of ZigBee based Train
Anti -Collision And Level Crossing Protection System for
Indian Railways." Int. J. of Lat. Trends in Eng. and Tech., Vol.2,
pp: 12-18, January, 2013.
129
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-138-320.jpg)
![[2] R. P. Ramasamay, M. P. Kumar, S. S. Kumar, R. R. Raman,
“Avoidance of fire accident on running Train using ZigBee
Wireless Sensor Network”, Int. J. of Inform. and Comp. Tech.,
Vol.3, pp: 583-592, 2013.
[3] M. Blaho, J. Ozimy, M. Ernek, M. Foltin, “Zigbee
implementation in Railroad model”.
[4] B. Vinodhini, K. Abinaya, R. Roja, M. Rajeshwari, “Wireless
Two-way Restaurant Ordering System via Touch Screen”,
Vol.3, pp: 01-05, 2014.
[5] W. Wang, G. He, J. Wan, “Research on Zigbee wireless
communication technology”, Int. Conf. on Elect. and Cont. Eng.
(ICECE), 2011.
[6] Bhargave, A., Jadhav, N., Joshi, A., Oke, P., & Lahane, M. S.
(2013). “Digital Ordering System for Restaurant Using
Android”, International Journal of Scientific and Research
Publication, Vol.3, 2013.
[7] Wahab, M. H. A., Kadir, H. A., Ahmad, N., Mutalib, A. A., &
Mohsin, M. F. M. (2008, August).” Implementation of network-
based smart order system”, International Symposium on
Information Technology, 2008. ITSim 2008. (Vol. 1, pp. 1-7).
IEEE.
[8] .[8] M.H.A. Wahab, H.A. Kadir, N. Ahmad, A.A. Mutalib and
M.F.M. Mohsin, “Implementation of network-based smart order
system,” International Symposium on Information Technology
2008
[9] PAR PixelPoint “PixelPoint POS Brochure [Accessed: 11 May
2011]
[10] Advanced Analytical, Inc (October 2004) “LRS Restaurant
Server Pager”, [Accessed: 11 May 2011]
130
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-139-320.jpg)
![Introducing space plant biology to students through
hands-on activities using clinostat
Ferheen Ayaz
SUPARCO Institute of Technical Training, HR Dte Gen,
Space and Upper Atmosphere Research Commission,
Karachi, Pakistan
ferheen@msn.com
Qurat-ul-Ain
SUPARCO Institute of Technical Training, HR Dte Gen,
Space and Upper Atmosphere Research Commission,
Karachi, Pakistan
ain_haider@hotmail.com
Abstract—Space biology is an important topic for research
and education in order to explore future life in space. Plant
growth and development largely depends upon gravitational
acceleration on earth. The growth process of a plant in space is
significantly different from that on earth due to the absence of
gravity, which is referred as zero gravity or microgravity. One-
axis clinostat simulates microgravity environment on earth. On a
horizontal clinostat, an object is rotated perpendicular to the
force of gravity. Although the gravity is present on an object
rotated by clinostat, its continuous reorientation nullifies the net
effect of gravitational acceleration. One-axis/horizontal clinostat
is a useful tool for space education. The simulated microgravity
created by the clinostat, despite its limitations, causes significant
changes in plant growth which may lead to introduce school and
college level students about plant life in space and creating their
interest in space biology. This paper describes some of the
experiments conducted on one axis clinostat, by a group of
teachers and students of SUPARCO Institute of Technical
Training, which proved to be a fruitful activity resulting in
learning of gravitropism basics of plants and seeds. The article is
aimed towards helping the educators in creating hands-on
learning activities for students to have an idea about life in space
by making use of facilities available on earth. The orientation of
roots grown from radish seeds (Raphanus sativus), area and
length of roots and shoots grown from pea seedlings (Pisum
sativum), length and bending angle of shoots of Marigold plant
(Tagetes patula) and growth rate and direction of wheatgrass
(Thinopyrum intermedium) were observed on the clinostat. The
roots of radish seeds did not grow downwards in the direction of
gravity but oriented themselves at random angles on the
clinostat. Pea seedlings also showed different growth rates on
clinostat as compared to their growth on earth. The experiment
on Marigold plant resulted in faster growth in length and larger
bending angle of clinorotated plant shoots as compared with
stationary plant shoots. The growth rate of wheatgrass on earth
and on clinostat was nearly same but their bending angles were
different in both conditions. All the experiments provoked the
minds of students to make inferences on seeds and plants growth
in space. The experiments described in the paper create
inquisitiveness among teachers and students to observe many
other parameters in plant growth under microgravity.
Keywords—clinostat; gravitropism; microgravity
I. INTRODUCTION
Considering the long term goals of space agencies and
seeking habitation of extraterrestrial surfaces; one of the
important concerns is to analyze the development of plants in
space. The reason behind this concern is the dependence of
plants and animals upon each other for their survival and
flourishing of ecosystem [1]. Study of plant growth in space is
different from conventional plant growth on earth mainly
because of the fact that it is greatly influenced by gravitational
acceleration. Recent experiments in space have resulted in
modified growth rates of roots and shoots of different plants
and the major reason of this modification is the absence of
gravity in space [2].
In order to promote space exploration related activities and
establish plant life in space, one of the key initiatives is to
promote space education among young students. Creating
interest of students in space sciences may lead to increase their
curiosity towards exploring life support systems in space and
planetary surfaces. Due to limited facilities of space research
in underdeveloped countries, students are least interested in
space biology. The students are less inclined to study space
sciences because of unavailability of resources and
opportunities to do practical work in this field. The practical
work for science students plays a very important role in their
learning [3]. As it is difficult to create practical setups of
complete space environment on earth, establishing a simulated
microgravity environment can provide excellent practical
opportunity for researchers and students to study life sciences
and plant and animal life in space [4].
Fig. 1. One-axis clinostat to simulate microgravity on earth.
Microgravity for a very short period of time can be created
on earth by elevator, drop tower, rocket and aircraft [5] but
changes in plant growth cannot be observed by these methods
as their growth process is comparatively slower. The efficient
instruments for conducting biology experiments in simulated
131
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-140-320.jpg)
![microgravity environment are 2D and 3D clinostats [6].
Keeping in view the practical needs of space education,
SUPARCO (Space and Upper Atmosphere Research
Commission) Institute of Technical Training acquired 2D
clinostat, as shown in Fig. 1, under ZGIP (Zero Gravity
Instrument Project) initiated by United Nations Human Space
Technology Initiative (HSTI) in 2013. A 2D or one axis
clinostat has one rotational axis which rotates perpendicular to
the direction of gravity vector in order to create the effect of
microgravity and is suitable for studying the behavior of
germinating seeds and analyzing plant growth [7]. Although
horizontal clinostat does not completely create zero gravity
environment on the object under study [8], it can still be used
for educational purposes to demonstrate changes in plant
growth by alteration of gravity.
This article describes some experiments conducted on
seeds and plants using clinostat at SUPARCO Institute of
Technical Training by a group of 2 instructors and 10 students
aged between 17 to 19 years. The objective of describing the
experiments is to promote science teachers and students to
study space biology practically and to provoke the minds of
interested learners to bring up similar but new ideas to
experiment and discover effects of gravity on plants.
The organization of the rest of the paper is as follows:
Section II describes the experiments performed using
clinostat, Section III discusses results of the experiments and
conclusion is mentioned in Section IV.
II. EXPERIMENTS PERFORMED USING CLINOSTAT
A. The radish seeds
The first experiment to study gravitropism on the roots of
radish seeds (Raphanus sativus) was conducted as described in
[7]. Nine samples of radish seeds were placed on each of the
three Petri dishes. The Petri dishes were half filled with agar-
agar solution to aggravate the process of germination.
Humidity, temperature and light conditions were kept same for
each Petri dish. After 48 hours, when the roots started to grow
downward, first Petri dish was placed vertically with roots
along the direction of gravity, second was turned 90o
so that
the roots were held perpendicular to the gravity vector and the
third was attached to clinostat using double sided tape. The
rotation disc of clinostat was held horizontally. The placement
of Perti dishes is illustrated in Fig. 2. The clinostat was set to
rotate with a speed of 10 r.p.m. for 4 hours and photographs of
each Petri dish were captured using a digital camera. After the
experiment, the photographs were observed using Image J
software. The orientation of roots was analyzed in terms of
angle created as the roots bent during the experiment.
Fig. 2 (a). First Petri dish with roots parallel to gravity vector.
Fig. 2 (b). Second Petri dish with roots perpendicular to gravity vector.
Fig. 2 (c). Third Petri dish attached to clinostat.
B. The pea seedlings
This experiment was designed by the working group to
observe not only the roots, but also the shoots grown from pea
seeds (Pisum sativum). Four test tubes were half filled with the
agar-agar solution made by the procedure defined in [7]. A pea
seed was placed inside each test tube such that it lies exactly
in the middle of the tube as shown in Fig. 3. After 7 days, the
shoots and roots of the pea seeds were grown. Two test tubes
were placed horizontally perpendicular to the gravity vector
and the other two test tubes were attached to clinostat. The
132
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-141-320.jpg)
![was 0.118 cm/hour from Day1 to Day2, 0.026 cm/hour
from Day2 to Day3 and 0.024 cm/hour from Day 3 to
Day4. It was almost the same on clinostat. The
calculations came out to be 0.112 cm/hour from Day1
to Day2, 0.027 cm/hour from Day2 to Day3 and 0.026
cm/hour from Day 3 to Day4.
As shown in Fig. 8(b); on earth, the wheatgrass grew
almost straight and angles with respect to the direction
of gravity were in the range of 0 to ±3o
, showing
almost no bending. On clinostat, the angles were
varying in the range to 10o
to 14o
which is the evident
proof of bending.
Summary of the result is presented in Table IV.
TABLE IV. LENGTH AND ANGLE OF WHEATGRASS DRUING 5 DAYS OF
EXPERIMENT
Measurement
of:
Average Length (cm) Average Angle
(degrees)
Day On earth
On
clinostat
On earth
On
clinostat
1 3.69 2.30 0.931 13.73
2 6.54 5.00 -0.69 8.77
3 7.19 5.67 2.81 10.80
4 7.77 6.31 -2.44 8.39
Fig. 8(a). Length of wheatgrass per day on earth and on clinostat.
Fig. 8(b). Angle of wheatgrass per day on earth and on clinostat.
IV. CONCLUSION
Four experiments on different seeds and plants were
conducted on clinostat. The results show that all plants do not
exhibit similar behavior in microgravity. Absence of gravity
speeded up the growth rate in pea seedlings and Marigold
plant but did not influence wheatgrass. The diversity of the
experiments described in this article encourages the students
to try out experiments on other plants and seeds available to
them. Various other parameters like temperature, humidity,
type of soil, amount of water and nutrients provided can also
be modified to discover new effects of gravity and
microgravity on plants.
Acknowledgment
The study is carried out as a part of Zero Gravity
Instrument Project initiated by United Nations Office for
Outer Space Affairs (UNOOSA). We gratefully acknowledge
UNOOSA for donating clinostat, Training Division, HR Dte
Gen, Pakistan Space and Upper Atmosphere Research
Commission for their coordination in publishing the article
and Horticulture Cell, Pakistan Space and Upper Atmosphere
Research Commission for assisting in plant growth.
References
[1] R. Ferl, R. Wheeler, H. G. Levine, and A. L. Paul, “Plants in space.”
Current opinion in plant biology, vol. 5, no. 3, 2002, pp. 258-263.
[2] T. Hoson, “Plant growth and morphogenesis under different gravity
conditions: Relevance to plant life in space,” Life, vol. 4, no. 2, 2014,
pp. 205-216.
[3] V. N. Lunetta, A. Hofstein, and, M. P. Clough, “Learning and teaching
in the school science laboratory: An analysis of research, theory, and
practice,” Handbook of research on science education, 2007, pp. 393–
441.
[4] A. Niu, M. Ochiai, and H. J. Haubold, “The United Nations Human
Space Technology Initiative (HSTI): Science Activities,” IAC-12-
A2.5.11, 2012.
[5] M. J. Rogers, G. L. Vogt, and M. J. Wargo, “Microgravity: A Teacher's
Guide with Activities in Science, Mathematics, and Technology,”
NASA, EG-1997-08-110-HQ, 1997.
[6] M. Cogli, “The fast rotating clinostat: a history of its use in gravitational
biology and a comparison of ground-based and flight experiment
results,” Gravitational and Space Research, vol. 5, no. 2, 2007.
[7] “Programme on Space Applications: Teachers Guide to Plant
Experiments in Microgravity,” United Nations Office for Outer Space
Affairs, 2013.
[8] A.H. Brown, O.A. Dahl, and D.K. Chapman, “"Limitation on the use of
the horizontal clinostat as a gravity compensator," Plant Physiology, vol
58, no. 2, pp. 127-130, 1976.
Sponsored by Pakistan Space and Upper Atmosphere Research
Commission.
135
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-144-320.jpg)
![Helium is the element
with the smallest
molecular size
Used for Leak detection
Has extremely low
solubility in water
Used as a diving gas
Helium-3 is not
radioactive
Used as a heat transfer
medium in fusion reactors
The importance of the noble gas being realized, it is
important to mention that there is a shortage of Helium on
Earth. Currently, Helium supplies are obtained from the
extraction of natural gas. The United States produces about
75% of the world’s Helium and roughly 30% of the world’s
Helium supply comes from the US federal Helium reserve. The
problem is that Helium reserves are being depleted at a rapid
rate because of wide use of the gas. Walter Nelson, director of
helium sourcing for Air Products and Chemicals Inc. predicts
the usable life of the US reserves to last till 2018-2020. This
will have a global impact, especially on healthcare and small-
scale scientific research. MRI machines need Helium to
operate and they will become difficult to maintain with a
shortage of Helium, ultimately making it very difficult to
obtain MRI’s.
Furthermore, Harrison Schmitt, a geologist and astronaut, is
of the opinion that Helium-3 can prove to be a very useful
alternative for deuterium and tritium in nuclear fusion reactors
for power generation, as the reaction involving Helium-3 does
not result in the emission of neutrons, making it clean and non-
polluting.
Since there are not enough supplies on the Earth to provide
the Helium-3 needed to carry out commercial power generation
through fusion, in his book “Return to the Moon” Harrison
Schmitt argues that the moon should be mined for Helium-3.
Since the moon is continuously bombarded with solar wind, it
has a large amount of Helium-3, though in small
concentrations of about 10-20 parts per billion. Schmitt
outlines a plan to for mining helium-3 on the Moon to feed a
commercial industry for generating power, though it will be
very expensive.
Even if many consider Schmitt’s proposal to be unrealistic,
it is a fact that he is not the only one who thinks mining the
moon for Helium-3 is a good idea; The leader of Beijing's
space program, has said generating power via nuclear fusion
using He-3 could "solve energy demand for 10,000 years at
least". So China is seriously considering it.
B. Rare Earth Elements
Rare Earth Elements, or REE for short, are 17 elements in
the periodic table which are widely used in electronics. Their
name tends to be misleading; they are relatively plentiful in the
earth’s crust. [11]
The mining of REE was once dominated by the United
States from 1965 to 1985, until The People’s Republic of
China started dominating the market. In 2010, a “Rare Earths
Trade Dispute” occurred, with China on one side and many
other countries, especially the US, on the other [10]. China had
increased the restrictions on its exports of REE by reducing its
export quota by 40% [2][12]. This resulted in a major increase
in the prices of REE outside the country. Although the Chinese
government argued that the reduction in the quota was for the
benefit of the environment (REE mining is highly polluting),
critics were unconvinced. Due to the importance of the REE in
the military, the question arose in the US of supply
vulnerability of these elements [2][13]. There were also
accusations that China had banned the exports of REE to Japan
after the 2010 Senkaku boat collision incident. The matter of
the Chinese restrictions was brought to the Dispute Settlement
Body of the WTO in 2012, by the United States. China had
joined the WTO in 11 December 2001. When it joined the
WTO, China signed an accession treaty, which did not allow
such restrictions on exports unless the goods in question were
stated, which was not the case here. China defended itself by
arguing that article XX (b) of the GATT allowed exceptions if
the restrictions were put in place for the protection of human,
animal or plant life or health. The majority of the panel was of
the view that the exceptions in article XX (b) did not justify
China’s reduction of export quotas and that these measures
were not necessary for the protection of human, animal or plant
life or health in this case. So the WTO ruled against China in
2014 and the country removed the restrictions in 2015 [10].
There are rocks rich in Potassium, REE, and Phosphorous
along with other elements, referred to as KREEP, on the
moon’s surface. According to the deputy director of the Rock
Mechanics and Explosives Research Centre at the Missouri
University of Science and Technology in Rolla, Leslie
Gertsch, although REE themselves aren’t detectable on the
moon’s surface, KREEP is. Detection of components of
KREEP, such as Thorium, on the lunar surface helps locate
associated REE due to similar geochemical properties that
caused them to crystallize under the same conditions, Gertsch
says. Dale Boucher, director of innovation at the Canada-based
Northern Centre for Advanced Technology Inc., in Sudbury,
Ontario, stated that the feasibility of mining REE on the moon
can only really be judged after exploration intended for this
exact purpose. It is difficult to assess it otherwise, as there are
many factors which determine its feasibility, including cost of
separating REE obtained from ores from each other, cost of
transportation of said REE, cost of refining them etc. (North
America seems to have abundant REE deposits but it is just not
profitable to mine them). But it is still in the realm of
possibilities. If a country does indeed intend to mine REE from
the moon’s surface in the near future, there’s a risk that a
scenario similar to the Rare Earths Trade Dispute might occur
C. Colonization
The United States, Japan, and Russia have shown interest in
the establishment of colonies on the moon [16]. The plans
seemed to have gained more ground on November 13, 2009,
137
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-146-320.jpg)
![when NASA announced the discovery of ice caps at the south
pole of the moon [15].
The reasons given by NASA for the exploration of the moon
are:
1. Human Civilization: Extend human presence to the
Moon to enable eventual settlement
2. Scientific Knowledge: Pursue scientific activities that
address fundamental questions about the history of
Earth, the solar system and the universe; and
therefore, about our place in them
3. Exploration Preparation: Test technologies, systems,
flight operations and exploration techniques to reduce
the risks and increase the productivity of future
missions to Mars and beyond
4. Global Partnerships: Provide a challenging, shared
and peaceful activity that unites nations in pursuit of
common objectives
5. Economic Expansion: Expand Earth's economic
sphere, and conduct lunar activities with benefits to
life on the home planet
6. Public Engagement: Use a lively space exploration
program to engage the public, encourage students and
help develop the high-technology workforce that will
be required to address the challenges of tomorrow
[3][4][5]
The 1st
and 3rd
reasons show their intention to colonize the
moon in the future. In January 2012 Republican candidate for
President of the United States of America, Newt Gingrich,
proposed the establishment of a US moon colony by 2020,
though the plan faced much criticism.[16]
Russia’s intentions seem to be different regarding the 3rd
reason. “The Moon is not an intermediate point in the race; this
process has the beginning, but has no end. We are going to the
moon forever.” stated Russia’s Deputy PM Dmitry Rogozin
who is in charge of space and defense industries. In 2007, the
Soviet Union announced its plan to establish a permanent
moon base by 2025 [14]. The main focus seems to be lunar
tourism [16]. According to a government draft that Izvestia, a
Russian newspaper, claims to have obtained, Russia has a 3-
step plan for manning the moon [14][17]:
1. Sending a robotic craft to the moon, possibly as early
as 2016,
2. Sending manned missions to orbit the moon by 2028,
3. A base will be set up by 2030, using resources from
the moon.
Japan announced its plans of setting up a moon base in
2006. They estimated that the base in question will be
established by 2030. Their main motivation, they claim, is the
development of robotics. [9]
D. Future Missions of Major Parties:
[20]
1. United States (NASA)
a. Exploration Mission-1: An unmanned flight
of the Orion (MPCV), using NASA’s Space
Launch System (SLS). The aim will be to
qualify the Heavy Lift Launch Vehicle (HLV)
and Beyond Earth Orbit (BEO) Orion for
transportation of humans into deep space. The
estimated date of launch is December 17,
2017. The mission is expected to last 7-10
days [18].
b. Exploration Mission-2: The first manned
mission of NASA’s BEO. The crew will
consist of 4 people. The aim will be to test
Orion’s life support systems and to validate
its crew operations. The mission is expected
to be carried out by the year 2021 [19].
2. Russia (ROSCOSMOS)
a. Luna-25: To land on the South Pole of the
moon, and test the lander technology and
communications systems
b. Luna-26: To orbit the moon for detailed
mapping of the moon’s surface and search for
a new landing site for future missions and to
measure the lunar atmosphere.
c. Luna-27: To go to the south pole of the moon
for testing drilling systems and analyzing
contents.
d. Luna-28: To return soil samples to the earth
e. Luna-29: To use a rover to take samples from
different locations on the moon’s surface
[21][22][23][24][25]
3. China
a. Chang’e 4: A backup probe for Chang’e 3
[6], which put which put a lander and the
Jade Rabbit rover on the moon in 2013. The
expected year of launch is 2020. [26][28]
b. Chang’e 5: To return lunar samples [27][29]
c. Chang’e 6: A backup for Chang’e 5. [27]
4. Japan (JAXA)
An attempt at Japan’s first lunar landing, an
unmanned flight, is planned for 2019, though the
budget for the plan is not available yet. [30][31][32]
5. India (ISRO)
Chandrayaan-2: Terrain mapping by the employment
of new technology for better imaging of the lunar
surface.
138
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-147-320.jpg)
![[33][34]
6. European Union (ESA)
ESA plans to send a lander near the moon’s South
Pole in 2018. The purpose of the mission is to probe
the moon’s surface and to test new technologies for
future landings.[7]
7. Private
Google Lunar X Prize: Astrobotic Technology, a
privately held American company, plans on launching
a private launcher in 2015, with the aim of winning the
Google Lunar X Prize, which was created in 2017 to
encourage space entrepreneurs to create a new era of
affordable access to the Moon and beyond. [8]
E. Conclusion
As it can be vividly seen, there are several detailed and
ambitious projects ready to explore the moon and benefit from
its resources. Keeping that in mind it becomes more important
than ever to firmly establish policies right now that are
acceptable to all parties to avoid any conflicts in the future.
ASTEROID MINING
Besides the moon, asteroids also have potential resources
which scientists are looking to explore.
Most asteroids in the solar system are found between Mars
and Jupiter. This region is termed the asteroid belt or main belt.
Asteroids with orbits that bring them within 1.3 AU (121
million miles/195 million kilometres) of the Sun are known as
Earth-approaching or near-Earth asteroids (NEAs). These are
the asteroids which people are looking to mine. Most or all
asteroids found in the main belt appear to be NEAs.
Asteroids are usually categorized according to their albedo
(how reflective they are), composition derived from spectral
features in their reflected sunlight, and inferred similarities to
known meteorite types.
There are 3 main categories:
1. C-type (carbonaceous): More than 75% of known
asteroids belong to this category.
Albedo: 0.03-0.09
Composition: Thought to be similar to the sun’s
composition, except for the absence of hydrogen, helium and
other volatile elements and compounds. Because of their low
temperatures, they are estimated to contain up to 22% water.
Found in: the main belt’s outer region.
2. S-type (silicaceous): About 17% of known asteroids
belong to this category.
Albedo: 0.10-0.22
Composition: metallic iron mixed with iron- and
magnesium-silicates.
Found in: the main belt’s inner region.
3. M-type (metallic):
Albedo: 0.10-0.18
Composition: mainly metallic iron.
Found in: the main belt’s middle region.
[35]
A. Motivations for Asteroid Mining
Based on consumption rates in both developed and
developing countries, it is estimated that terrestrial reserves of
elements needed for modern industry will be exhausted within
50–60 years. Consequently, 44 elements are expected to face
supply limitations in the coming years, including phosphorus,
antimony, zinc, tin, silver, lead, indium, gold, and copper. All
these elements are crucial for industry [36][37][38]. Some of
the applications of a few of these elements are noted in the
following table:
TABLE 2.ELEMENTS FROM ASTEROIDS AND MAJOR APPLICATIONS
Element Applications
Copper . Used in electrical equipment
such as wiring and motors.
. Copper Sulfate is used as an
agricultural poison and as an
algaecide in water purification.
. Fehling’s solution, a solution
containing copper, and other
copper containing solutions are
used in chemistry for qualitative
analysis.
. Copper alloys are used in
making coins.
Zinc . Used to galvanize objects to
prevent rusting.
. Used to produce die-castings.
. Used in alloys such as brass,
nickel silver and aluminum
solder.
. Zinc oxide is used in the
manufacture of paints, rubber,
cosmetics, pharmaceuticals,
plastics, inks, soaps, batteries,
textiles and electrical equipment.
. Zinc sulfide is used in making
luminous paints, fluorescent lights
139
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-148-320.jpg)
![and x-ray screens.
Lead . Used for car batteries, pigments,
ammunition, cable sheathing,
weights for lifting, weight belts
for diving, lead crystal glass,
radiation protection and in some
solders.
. Used to store corrosive liquids.
. Used in architecture, for roofing
and in stained glass windows.
Tin . Used to coat other metals to
prevent corrosion.
. A niobium-tin alloy is used for
superconducting magnets.
. Tin salts sprayed onto glass are
used to produce electrically
conductive coatings.
[39]
B. Mining Considerations
There are three options when considering mining asteroids:
1. Bringing the raw material to earth.
2. Processing the raw material on the asteroid itself and
then bringing the processed material to earth.
3. Transport the asteroid to a safe orbit around the
Moon, Earth or to the ISS. This method is estimated to be the
one which allows the most materials to be used.
[40][41]
C. Proposed Mining Projects
There are four main parties currently proposing projects for
mining asteroids:
1. NASA (US)
An upcoming mission, termed OSIRIS-Rex, scheduled for
launch in 2016, will involve travelling to a near-Earth asteroid
called Bennu and bringing a small (about 60g) sample back to
Earth for study. The destination will be reached in 2018 and the
sample will be transported to the earth in 2023.[49][50]
NASA Institute for Advanced Concepts (NIAC) announced
the Robotic Asteroid Prospector (RAP) project on in
September 2012. The purpose of RAP is to evaluate the
feasibility of asteroid mining in terms of means, methods, and
systems. Their stated strategy is: “Initial prospecting missions
will be robotic, although early industrial mining missions will
require both humans and robots on-site. There will be too many
judgment calls and qualitative decisions on-site for robots
alone. Humans will need the practice of mining in space with
robots to learn how to innovate and to work efficiently,
reliably, and safely.” [42]
2. Planetary Resources (Private, US)
The company announced on April 24, 2012 its plans to
mine NEA. They plan to launch prospective spacecraft for
collecting data about asteroids within 7 years. The company
plans to build new technology which will reduce the price of
space travel so that it can be successful in its mission.
[46][47][48]
4. Deep Space Industries (Private, US)
On January 22, 2013, this private company announced its
plans to launch spacecraft in 2015 for prospecting suitable
asteroids and returning samples to the earth, and to begin
mining NEA by 2023. [43][44][45]
5. Kepler Energy and Space Engineering (KESE)
They aim to build and send a 4-module Automated Mining
System (AMS) to a small asteroid with a simple digging tool to
collect approximately 40 tons of samples by the end of the
decade, using existing guidance, navigation and anchoring
technologies.[51]
D. Conclusion
It becomes clear after looking at the motivations and future
plans of various parties (countries as well as private parties)
that a future marathon for space resources is highly probable.
Unfortunately, the Outer Space and Moon Treaty hinder the
steps made towards the mentioned endeavours as they face
many legal barriers. For now, only the hurdles provided by the
outer space treaty will be discussed, as the moon treaty has not
been signed by major influential parties. Specific parts of the
two treaties will be discussed in more detail later on in the
paper.
Based in part on the Antarctic Treaty, the Outer Space
Treaty only appears to permit research activities on the moon
and other celestial bodies, though this is debatable. “The idea
that you can't claim sovereignty is not necessarily incompatible
with the right to go conduct mining operations. The high seas
are not subject to any sovereignty, but people can go and fish
there.” said international lawyer and space-law expert Timothy
Nelson in an interview with space.com, referring to an article
in the 1967 treaty stating: “Outer space, including the moon
and other celestial bodies, is not subject to national
appropriation by claim of sovereignty”.
Even then, the treaty hinders research, mainly owing to a
lack of incentive. Due to the statement that no particular party
can claim any ownership rights to the resources, there will be a
lack of commercial concern, ultimately leading to a lack of
funding for research, hence lowering the amount of potential
research which can be carried out.
140
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-149-320.jpg)
![HISTORICAL ANALYSIS
The ever-evolving economic landscape presents a unique
challenge in the world of politics. The one of keeping peace
and making sure that the resources of a state are not
illegitimately taken away from it. The solution is an old one;
treaties and organizations to enforce the treaties.
In this section conflicts and resources as a main reason for
conflicts will be discussed, along with how they can be
avoided. This will lead to the importance of treaties in the
international landscape and through the analysis of treaties and
treaty organizations a parallel will be drawn between the
historic treaties and conflict and the current space situation to
show that the current treaties governing space exploration and
resources management are not viable.
A. Wars
History has seen many wars fought over resources. Hitler’s
campaign in 20th century, before the Second World War is a
perfect example of a war fought over resources. When Hitler
came to power in 1933 his whole agenda centered on making
Germany a powerful sovereign country again. There were two
important aspects to this. One was regaining areas with great
economic potential, such as Alsace and Loraine on the western
border; they were highly industrialized zones and had Iron ore
and Coal deposits. Another was to get more land. This was a
part of Lebensraum (literally ‘living space’), a German
ideology proposing expansion to give more space to the
German people. This generally advocated expansion on the
Eastern front. [52]
Another event is the period in history known as the
“Scramble for Africa” between 1881 and 1914. It is the period
of invasion, occupation, colonization and annexation of
African territories by European empires. Africa had a lot of
natural resources, ranging from coal to diamonds, and a huge
untapped potential. Furthermore the African population was
also used as a source of free labor through slavery. As
expected, with such resources available, it was not long before
war broke out, not only the war against the natives to take
control of the area, but also between different European powers
to get the biggest piece of the pie. [53]
The most recent, and one that is still relevant today, is of
the Oil War. Oil war is the term given to all the conflicts over
petroleum resources. Oil has become very important in the
modern world, as primary source of energy its significance is
unprecedented. Petroleum resources alone can run a country’s
economy such as the Middle East states, while on the other
hand import of oil can cripple a country. The result of all this is
that oil became an extremely important resource in the
international landscape, a resource that countries wage wars
over. There have been a number of conflicts over oil [54]. The
Iran-Iraq war (1980-1988), the Gulf War (1990-1991) and the
Afghan War (1978-present) are major examples of this. Some
conflicts that have control of petroleum resources as an ulterior
motive are still present today. The most recent major crisis was
Russia’s annexation of Ukraine, this not only demonstrates
Russia’s interest in controlling the oil pipeline routes, but also
that of United States [55]. And perhaps the most important is
the creation of ISIS. ISIS sits on an oil rich land, and even
though no state can legally buy oil from them, they do have a
steady income through oil sales. This location of ISIS and the
possible effect of their oil sales made it important to other
foreign powers even before they began global terrorism.
[56][57].
As it can be seen, resources are an impetus for war. And as
the previous section elaborated, space has a lot of resources
available and a space race between various parties is highly
probable. Whereas history shows, there is an equally high
probability of conflict as a result of this. And in order to avoid
this we need a comprehensive treaty system.
B. A Failed Treaty
As with anything else, it is important to learn from the
mistakes already made by others. The same goes with treaties.
It is crucial to see where the previous treaties went wrong and
to rectify them. In hope of avoiding future conflicts from
turning violent and it is of immense importance when
considered that now the domain of humans is not just restricted
to the Earth, but also includes the space.
The Treaty of Versailles of 1919 is one of the most
important treaties ever signed. It effectively brought an end of
the First World War, though unknown at the time it was
signed, it was a significant factor in the start of the Second
World War. [58]
The treaty’s ultimate failure was when Germany invaded
Poland in 1939, dragging the whole of Europe into the Second
World War. There were a number of reasons for the failure of
the treaty. One was that the treaty had been made by the
winners of the war and it was very profitable for them i.e. it
was a biased treaty where everyone was not on the table. The
treaty was generally against one country (Germany) and the
actors at the time felt that as long as one rogue state was
controlled, the peace would not be disturbed, but in reality
limitations were needed for all the countries. Another reason
was that there was no way to reinforce the terms of the treaty,
such as the clauses that called for all-around disarmament.
None of the countries wanted to reduce armaments, so no one
was there to enforce the decision either. And whereas countries
like Britain and France had agreed to keep Germany in check,
they were not ready to go to war for it or even impose sections
that would hurt their own economy. Hence it can be seen that
the ultimate weakness of the treaty was that there was no one to
stop any country from breaking it.[52]
Most of these reasons are generic, and very similar reasons
lie behind the failure of almost all treaties. And so when
planning for the future, especially in the case of forming
141
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-150-320.jpg)
![treaties that have universal importance, like as those related to
the space, such weaknesses in treaties of the past should be
avoided. A few of the common shortcomings that must be
avoided are the treaty should not be directed against just a
certain country or to control just the rogue states, but be
equally applied to everyone; furthermore the treaty needs to
absolutely binding, and the powerful states that promise to
direct it, must be legally compelled to do. And most
importantly there should be measures that make it impossible
for states to violate or ignore the treaty.
C. A Failed Organization
An organization created along with Treaty of Versailles
was the League of Nations. It was the predecessor to the United
Nations. It had two main aims, one to maintain international
peace through “collective security” and the second to promote
international cooperation to solve social and economic
problems. [59] The former was the more important function of
the body. The world had just come out of the First World War,
and sought peace and stability. The main idea was to discuss
the problems as they rose, and solve them through an
understanding of all the countries through a unanimous effort.
Though the League looked very good on paper, it was not so
successful in reality, with the Second World War an epitome of
its failure. There are a number of reasons why the organization
failed. Firstly it did not include all the countries, major
exclusions were the US, Germany, and USSR. And there were
serious weaknesses in the covenant, like unanimous decisions
were needed and the League had no force of its own. The result
was that it could advise on how to solve problems, but it could
not force a state. Furthermore it was a very British and French
affair; these were the two most powerful countries at the time
and the league was always under their control. As a result the
League was prone to letting allies of the major states get away
with anything; major examples of this are the Japanese
invasion of Manchuria (1931) and the Italian invasion of
Abyssinia (1935). [52]
Once again it is important to take care of these few
parameters in order to form a comprehensive organization that
would be much more equipped to deal with problems in a
manner that would avoid conflict. The “Regime” is an
organization was proposed by the Moon Treaty to for the use
of the Moon and other celestial bodies, and it is yet to be
formed. Keeping the weaknesses of past organizations in mind,
especially those that led to the failure of the League of Nations,
the framework of the Regime should be such that it covers all
states equally and does not hinder it in its function, yet allows
it to be decisive. All countries should get a say in it, but
decisions should not be unanimous as it would render it
impractical, and similarly it should not give powerful states the
veto power, as it too could hinder the organization.
D. A Successful Treaty
Antarctica is the closest example of space on Earth.
Antarctica, like space, has no native inhabitants. This makes
the Antarctica Treaty System immensely important. Signed in
1959, the treaty established a few ground rules for every state
regarding Antarctica. First off the geographical area of
Antarctica was determined. And it was agreed that the area will
be used for peaceful purposes only. Freedom of scientific
investigation and cooperation was guaranteed. And importantly
it prohibited new territorial sovereignty claims, but did not
dispute nor recognize the old claims. [60]
Further additions were made to the initial treaty, among
them, the most important was the Convention on the
Regulation of Antarctic Mineral Resource Activities (1988),
which made mining illegal in Antarctica even though there coal
deposits and hydrocarbons along with other minerals [61]. And
the Protocol on Environmental Protection to the Antarctic
Treaty (1991) was to insure the protection of the Antarctic
environment and ecosystem.[62]
The Antarctic treaty has been a success so far. It has
managed to peacefully promote Antarctica as a research area
and kept it from being militarized. And it has also successfully
prevented any conflicts over territory by not recognizing or
disputing preexisting claims, and forbidding new claims.
Furthermore the environment of Antarctica has been protected
and it remains the only continent that has not been affected by
the human population.
The Antarctic Treaty has been successful, however that
does not make it a perfect model to work with in terms of a
space treaty. Unlike Antarctica, outer space is not a small
potential contributor to world’s demands for mineral resources.
Moreover, outer space also does not pose any environment
threats to Earth like industrialization of Antarctica could.
Furthermore, research in space might very well soon depend on
extraction and utilization of space resources to fuel it. Yet on
the other hand, the Antarctic Treaty has prevented conflict in
the region in an exemplary way, and the features of the treaty
that guaranteed this should be replicated which will be
discussed latter.
The current treaties are inspired by the Antarctic Treaty but
success cannot be replicated this way since the conditions are
far too different. As a result, the UN must recognize the
difference to move forward to full utilization of outer space’s
potential, yet in such a manner that global peace is not
jeopardized.
FLAWS IN THE TREATIES
A. The Res Communis Controversy
One of the chief controversies is the Res Communis
ideology employed especially in the Moon Treaty 1979 [65].
The phrase “common heritage of mankind” contributes to the
overall vagueness of the treaty and hence fuels the insecurity of
interested nations. Since moon and celestial objects are a
“province of mankind” [66], any form of active military
involvement is forbidden. However, considering the rise of
terrorist organizations like ISIS and rogue nations like North
142
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-151-320.jpg)
![Korea, both increasingly using sophisticated technology; it
seems dubious that space would remain untainted from terrorist
activities in the future.
In the case of USA, however, the reasons for disagreement
were political. The treaty put developed countries in a place
where they or their private companies could be attacked for
using outer space resources by developing countries. This
concerned scientists, entrepreneurs and politicians alike as it
not only meant a decreased interest in outer space by the
government and private groups but also a resulting decline of
space science research because if the moon and celestial bodies
are to remain unusable, then there is little incentive for them to
work.
It is true, however, that it could be done under the watch of
an “international regime” [67] but that too raises the fear of
blackmail since non-space-faring countries are in a majority
and most likely to lobby themselves to obtain economic
advantages. It is expected that the “international regime” would
be similar to “The Enterprise” proposed in Agreement of the
Law of the Sea Convention 1994[68]. The Enterprise, by its
very nature, was to receive a portion of wealth obtained by the
parties exploiting marine resources and distribute it amongst
developing countries. Such a system could prove discouraging
for space resources as there is a much higher capital required
and much greater risk of failure, combining this with low
profits means that space resources may remain untouched for a
long time even after sufficient technology has been developed.
Outer Space Treaty enforced the idea that no state could
claim any region in space [69], however it was not very
specific about private companies and groups. But the Moon
Treaty attempted to make sure that even private groups could
not lay claim to any region in space. As expected, it raised a lot
of opposition from the business community and it became one
of the most controversial points of the treaty.
B. Property and Ownership
It has been agreed that ownership of space and celestial
objects is not possible for countries but there is an apparently
silent approval for private groups. This raises further questions
and controversies.
Firstly, how can the ownership be declared if there is no
national government to maintain and uphold the rights of the
property owner. Such a situation would force the owners to
establish their own system of defense which may lead to an
arms race or recurring physical conflict in space.
As it is later suggested, the establishment of an organization
like “The Enterprise” is critical. Moreover, the UN should be
prepared to accept the eventual necessity of a peace-keeping
force in space to assist the organization in enforcement of its
mandate.
C. Lack of Space Involvement
On the other end of the spectrum, another hurdle in
development of sustainable space policies is a lack of
awareness and interest by some countries stemming from a
lack of space involvement. These countries often have not
developed a significant space involvement and hence deem the
development of such policies to be pointless in the present, as
the delegate of Argentina replied to COPUOS question
regarding definition of Outer Space: “The present magnitude of
space and aviation activities and technologies has not given
rise to a need for the adoption of specific provisions delimiting
outer space [70]”.
Though Argentina has CONAE as its space agency, its
future plans, however, do not include any form of space
exploration, simply satellite management [71], unlike countries
like China, Japan, USA, Russia and India as elaborated in
Section II. The case of Argentina is not singular but it
represents the hesitation of many countries that have too weak
space involvement to be able to make decisions concerning
their role in it. Hence to play it safe they decide not to indulge
in the process of any policy making which they may find even
slightly restrictive in the future.
Consequently, a large number of developing and some
developed countries are not a part of policy making because
their national experts and politicians do not have the
knowledge and understanding of the importance and
significance of space policy making, especially those
pertaining to resources because since Moon and celestial
bodies have been declared a common heritage of mankind it is
important that all countries understand their resulting rights and
limitations
D. Confidentiality and Privacy
The Moon Treaty states “…all space vehicles, equipment,
facilities, stations and installations on the moon shall be open
to other States Parties…” [72]This could give rise to a lot of
possible misuses and controversies such as technology theft,
infringement of privacy and right to confidentiality.
If it is possible for any state to inspect the equipment of any
other state then it is possible that this may be used to replicate
the technology of the first state. It poses many questions about
the right of having intellectual properties in space. Even though
countries and private groups may accept space as a common
heritage but it will be unfair for them to lose the rights to all
their individual research and development in space. Since such
research is extremely expensive, it is very likely that such a
jurisdiction would very severely dent the interest in it.
Parties operating in space and celestial bodies will also
have to disclose the specifics of their activities even though
they may be well within their rights. It too opens up debate on
human rights, among other things. If, for example, a researcher
is studying the health of astronauts on moon, he may have to
disclose data regarding this research, which may contain
personal information of astronauts. This can be taken as a
violation of Universal Declaration of Human Rights [73].
E. Lack of Violation Protocols
Both the treaties fail to outline a mechanism that is to be
followed if the treaties are breached by a country. United
Nations would most likely be unable to undertake any legal or
physical action since the Security Council would need
143
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-152-320.jpg)
![unanimous votes to do so. As a result the treaties appear to
restrict less influential and less powerful countries only, since
the stronger ones can easily get off without any rebuttal as they
can diplomatically procure a veto.
The example of Antarctic Treaty needs to be considered
here [74]. Interestingly enough, Antarctic treaty was the
inspiration for Outer Space treaty and yet such an integral
component was left out. According to Antarctic Treaty, in case
there is a clash of purposes between two parties, the matter can
be taken into International Court of Justice from where the
matter could be resolved. Not only does it safeguard the
integrity of the treaty but it also makes it more trustable, one of
the many reasons of its success.
The two treaties were missing a core element of a
successful international agreement, the violation protocols and
as it was witnessed with the failure of Moon Treaty, other
articles are meaningless without them.
F. Stifling Space Research and Funding
It is true that Outer Space Treaty intends to maximize
research and interest in outer space; however, its actual effects
have been different. As it has been previously established, the
restrictions collectively work to discourage investors. This in
turn leads to a direct decrease in the funding of researchers and
scientists who are researching on space related areas. Apart
from nationally run agencies, few research groups are able to
sustain themselves. Consequently, the research that has been
done is nowhere near the research that could have been done
without the policies.
If some of the most restrictive clauses are removed, there
would be an immediate rise in research, jobs, interest and
awareness regarding space and celestial bodies. Space is an
area that captures the imagination of many young scientists but
they are often unable to find employment or even advanced
education in their area of interest.
SOLUTIONS
A. Establishment of Extraction-Friendly Framework
The current policies need to be modified to encourage both
the private and government space agencies to attempt
extraction of resources from regions beyond Earth. To that end,
several changes need to be made.
Private parties should be allowed to have at least temporary
property rights so their reservations about lack of control can
be met. It makes sense to allow property rights on, for example,
asteroids, because after the extraction is over there will be little
left of it.
Parties should also have a greater degree of privacy and
confidentiality. Specific rules and regulations should be framed
to ensure that the rights to intellectual property and personal
privacy of astronauts are conserved.
Moreover the specifics of a regime like “The Enterprise”
for space should be discussed and decided. Special attention
should be given in making sure that space agencies are not
restricted unfairly by the share they have to provide to such a
regime.
The shares from the extraction should be based on a model
such that small scale extraction is bound to a lesser percentage
and large scale extraction to a greater percentage.
B. Increasing Space Awareness
Established agencies like the NASA should work harder to
develop a sense of understanding and importance of space
resources and hence the importance of policy making
governing it.
Only after the governments and experts of individual
nations realize the significance of such policies and how they
are critical in avoiding future clashes can the UN as a whole
move forward.
It may also be important to get international public support
for the policies to pressurize the governments into making
decisions in the present than letting them hang on for the
future. This requires the use of education and awareness
mechanisms through television, newspapers and public talks.
C. International Space Property Rights
Due to the sheer enormity of space, abundance of celestial
bodies and rapid shifting, it may never be possible to divide
regions of space in a manner that is fair to all. However, the
common heritage ideology may prove extremely impractical in
the future. As an analysis of history shows, such a concept is
extremely naïve when considered alongside the thousands of
wars that people have waged for land and resources.
Space agencies, both private and government, need to have
a tangible incentive for them to risk a great amount of capital
and manpower in attempting to reach for the moon and beyond.
Loss of the right to ownership can discourage space
exploration. They should be entitled to own, use and sell the
resources that they have extracted, to at least some degree.
However, concerns of developing nations should also be
addressed pertaining to property rights. Their current reduced
state of development does not mean that they lose any chance
of being a part of the future.
Taking both sides into consideration, one method that
seems feasible is to employ a procedure whereby regions of
space are provided as property only during the period they are
being actively used by the party to discourage gluttony.
.
D. Developing Violation Protocols
To uphold the good intentions, enforce articles and develop
trust, there needs to be a strong and independent system of
justice which ensures that the rights of less powerful and less
influential countries remain preserved.
To that end, a section of the International Court of Justice
could be formed to deal with space law. However, for its
smooth operation the policies need to be much more detailed.
In fact, it would be wise to draw up an internationally
acceptable constitution governing activities in space and
celestial objects. Not only would it act as a code of conduct for
144
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-153-320.jpg)
![parties interested in space resource extraction but would also
help quickly resolve any minor conflicts that may arise.
E. Using International Diplomacy Positively
Moon Treaty was treated very coldly by the international
community. However, an active interest of a single party such
as the USA could have totally reshaped its outcome as the
decisions of such influential nations are followed by their close
allies as well. In fact, an agreement on an international issue
between USA, China and Russia almost always gains nearly
unanimous global support.
As a result bloc leaders like USA and China should take a
responsive and positive role in the development of space
policies. By considering the interests of their allies, many of
those who do not have an independent space program, they can
help shape a globally acceptable and beneficial treaty.
REFERENCES
[1] [Online] Available: www.islandone.org/Treaties/BH766.html
[2] Leonard David. (2010, October 4). Is Mining Rare Minerals on
the Moon Vital to National Security? [Online]. Available:
http://sservi.nasa.gov/articles/is-mining-rare-minerals-on-the-
moon-vital-to-national-security/
[3] Beyond Earth [Online]. Available:
http://www.nasa.gov/exploration/home/why_moon.html
[4] (December 2006). Lunar Exploration Objectives
(1sted.)[Online]. Available:
http://www.nasa.gov/pdf/163560main_LunarExplorationObjecti
ves.pdf
[5] [Online]. Available: http://www.permanent.com/lunar-mining-
the-moon.html
[6] (2013, December 16). China plans to launch Chang'e 5 in 2017
[Online]. Available:
http://news.xinhuanet.com/english/china/2013-
12/16/c_132971252.htm
[7] (2010, September 16). NEXT STEP FOR ESA’S FIRST
MOON LANDER [Online]. Available:
http://www.esa.int/Our_Activities/Human_Spaceflight/HSF_Dir
ectors_Corner/Next_step_for_ESA_s_first_Moon_lander
[8] [Online]. Available: http://lunar.xprize.org/
[9] (2003, August 3). Japan Plans Moon Base By 2030 [Online].
Available:
http://www.moondaily.com/reports/Japan_Plans_Moon_Base_B
y_2030_999.html
[10] China — Measures Related to the Exportation of Rare Earths,
Tungsten and Molybdenum [Online]. Available:
https://www.wto.org/english/tratop_e/dispu_e/cases_e/ds431_e.
htm
[11] Kieran Mulvaney. (2012, March 14). The Fight Over Rare
Earths: A Primer [Online]. Available:
http://news.discovery.com/earth/the-fight-over-rare-earths-a-
primer-120314.htm
[12] Lesley Stahl. (2015, March 22). Modern life’s devices under
china’s grip?[Online]. Available:
http://www.cbsnews.com/news/rare-earth-elements-china-
monopoly-60-minutes-lesley-stahl/
[13] Leonard David. (2010, October 4). Is Mining Rare Minerals on
the Moon Vital to National Security? [Online]. Available:
http://www.space.com/9250-mining-rare-minerals-moon-vital-
national-security.html
[14] (2014, May 9). Russia will begin Moon colonization in 2030 -
draft space program [Online]. Available:
http://rt.com/news/157800-russia-moon-colonization-plan/
[15] Pat Cahill. (2014, January 20). Mount Holyoke professor Darby
Dyar says a colony on the moon is no fantasy [Online].
Available:
http://www.masslive.com/living/index.ssf/2014/01/mount_holyo
ke_professor_part_of_massive_study_of_outer_space.html
[16] DeAnnaKerley. 15 Ambitious Plans to Colonize the Moon
[Online]. Available: http://mentalfloss.com/article/53588/15-
ambitious-plans-colonize-moon
[17] Anna Dolgov. (2014, May 8). Russia Plans to Colonize Moon
by 2030, Newspaper Reports [Online]. Available:
http://www.themoscowtimes.com/article/499806.html
[18] Chris Bergin. (2012, February 29). Exploration Mission 1: SLS
and Orion mission to the Moon outlined [Online]. Available:
http://www.nasaspaceflight.com/2012/02/exploration-mission-1-
sls-orion-debut-mission-moon-outlined/
[19] Chris Bergin. (2012, October 30). EM-2: Orion crew to spend
up to four days in Lunar Orbit [Online]. Available:
http://www.nasaspaceflight.com/2012/10/em-2-orion-crew-
spend-four-days-lunar-orbit/
[20] Future planetary missions [Online]. Available:
http://spaceexp.tumblr.com/future-missions
[21] Russian Moon missions face three-year delay [Online].
http://www.russianspaceweb.com/spacecraft_planetary_2014.ht
ml
[22] (2014, October 10). SRI: Russia will begin launching automatic
stations to the Moon in 2018 [Online]. Available:
http://ria.ru/space/20141003/1026822990.html#ixzz3eXhKtcXA
http://ria.ru/space/20141003/1026822990.html
[23] Russian Lunar Polar Probes, 2015-2020 [Online]. Available:
http://www.permanent.com/russian-lunar-polar-probes.html
[24] (2014, August 3) Manned Moon Mission to Cost Russia $2.8
Bln – Space Research Institute [Online]. Available:
http://sputniknews.com/russia/20140803/191677434.html#ixzz3
eXiSLeokhttp://sputniknews.com/russia/20140803/191677434.h
tml
[25] (2014, October 10) Russian Luna-25 Mission to Cost Billions:
Space Agency [Online]. Available:
http://sputniknews.com/russia/20141010/193919188.html#ixzz3
eXj6Nr00http://sputniknews.com/russia/20141010/193919188.h
tml
[26] (2015, May 22). China Plans First Ever Landing On The Lunar
Far Side [Online]. Available:
http://www.spacedaily.com/reports/China_Plans_First_Ever_La
nding_on_the_Dark_Side_of_the_Moon_999.html
[27] Leonard David. (2015, March 17). China Outlines New Rockets,
Space Station and Moon Plans [Online]. Available:
http://www.space.com/28809-china-rocket-family-moon-
plans.html
[28] Alexandra Witze. (2014, March 19). China's Moon rover awake
but immobile [Online]. Available:
http://www.nature.com/news/china-s-moon-rover-awake-but-
immobile-1.14906
[29] Bradley Perret. (2013, December 16). China Plans 2017 Lunar
Sample-Return Mission [Online]. Available:
http://aviationweek.com/space/china-plans-2017-lunar-sample-
return-mission
[30] (2015, April 19). JAXA plans Japan’s first moon shot in 2018
[Online]. Available:
http://www.japantimes.co.jp/news/2015/04/19/national/jaxa-
plans-japans-first-moon-shot-in-2018/#.VY5YSruqqkp
[31] Dianne Depra. (2015, April 4). Japan Aiming For Moon
Landing By 2018 According To JAXA [Online]. Available:
http://www.techtimes.com/articles/48266/20150424/japan-
aiming-for-moon-landing-by-2018-according-to-jaxa.htm
[32] (2015, April 22). JAPAN TO ATTEMPT UNMANNED
LUNAR LANDING IN 2018 [Online]. Available:
http://www.spaceflightinsider.com/organizations/jaxa/japans-
jaxa-to-attempt-lunar-landing-in-2018/
145
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-154-320.jpg)
![[33] (2010, Sep 5). ‘We’re launching Chandrayaan-2 for a total
coverage of the moon’ [Online]. Available:
http://timesofindia.indiatimes.com/home/opinion/interviews/We
re-launching-Chandrayaan-2-for-a-total-coverage-of-the-
moon/articleshow/6501413.cms?referral=PM#
[34] India to launch Chandrayaan-2 by 2017 [Online]. Available:
http://www.thehindu.com/sci-tech/science/india-to-launch-
chandrayaanii-by-2017/article5562361.ece?ref=sliderNews
[35] Asteroids [Online]. Available:
http://nssdc.gsfc.nasa.gov/planetary/text/asteroids.txt
[36] Stephen Shaw. (2012, August 21). Posts Tagged ‘C-type
asteroids’ Asteroid Mining [Online]. Available:
http://www.astronomysource.com/tag/c-type-asteroids/
[37] Endangered Elements [Online]. Available:
http://www.acs.org/content/acs/en/greenchemistry/research-
innovation/research-topics/endangered-elements.html
[38] David Cohen. (2007, May 23). Earth's natural wealth: an audit
[Online]. Available:
http://www.sciencearchive.org.au/nova/newscientist/027ns_005.
htm
[39] Periodic Table [Online]. Available: http://www.rsc.org/periodic-
table/
[40] Stephen Harris. (2013, April 8). Your questions answered:
asteroid mining [Online]. Available:
http://www.theengineer.co.uk/aerospace/in-depth/your-
questions-answered-asteroid-mining/1015966.article
[41] (2012, April 2). Asteroid Retrieval Feasibility Study [Online].
Available:
http://kiss.caltech.edu/study/asteroid/asteroid_final_report.pdf
[42] Marc Cohen. Robotic Asteroid Prospector (RAP) Staged from
L-1: Start of the Deep Space Economy [Online]. Available:
http://www.nasa.gov/offices/oct/early_stage_innovation/niac/20
12_phase_I_fellows_cohen.html
[43] Mike wall. (2013, January 22). Asteroid-Mining Project Aims
for Deep-Space Colonies [Online]. Available:
http://www.space.com/19368-asteroid-mining-deep-space-
industries.html
[44] (2013, January 22). Commercial Asteroid Hunters announce
plans for new Robotic Exploration Fleet [Online]. Available:
http://www.prweb.com/releases/2013/1/prweb10346181.htm
[45] Taylor Soper. (2013, January 22). Deep Space Industries
entering asteroid-mining world, creates competition for
Planetary Resources [Online]. Available:
http://www.geekwire.com/2013/deep-space-industries-entering-
asteroidmining-world-creates-competition-planetary-resources/
[46] http://www.planetaryresources.com/technology/#space-based-
observation
[47] (2012, April 24). Company aims to mine resource-rich asteroids
[Online]. Available:
http://lightyears.blogs.cnn.com/2012/04/24/companies-plan-to-
mine-precious-metals-in-space/?hpt=hp_t3
[48] (2012, April 24). Plans for asteroid mining emerge [Online].
Available: http://www.bbc.com/news/science-environment-
17827347
[49] (2013, August 8). New NASA Mission to Help Us Learn How
to Mine Asteroids [Online]. Available:
https://www.nasa.gov/content/goddard/new-nasa-mission-to-
help-us-learn-how-to-mine-asteroids
[50] (2015, June 22). OSIRIS-REx Team Prepares for Next Step in
NASA’s Asteroid Sample Return Mission [Online]. Available:
https://www.nasa.gov/feature/goddard/osiris-rex-team-prepares-
for-next-step-in-nasa-s-asteroid-sample-return-mission
[51] Michael Buet and Robert Frantz. ASTEROID MINING: WHY
AND HOW? [Online]. Available:
http://www.friendsofbest.org/wp-
content/uploads/2015/05/ASTEROID-MINING_WHY-AND-
HOW-Article_03.pdf
[52] Norman Lowe, Mastering Modern World History, 4th edition,
New York: Palgave Macmillan, 2007
[53] Andrina Stiles, "Bismark's Germany" in The Unification f
Germany, 1st edition, London, England: Hodder& Stoughton,
1998, pp. 85-99
[54] John Lowe and Robert Pearce, "Colonial Rivalries, 1870-1914"
in Rivalry and Accord - International Relations 1870-1914, 1st
edition, London, England: Hodder& Stoughton, 2007, pp. 39-69
[55] Michael Klare, The Race for What's Left: The Global Scramble
for the World's Last Resources, 1st edition, New York: Picador,
2013.
[56] Dr. Nafeez Ahmed. (2014, March 6), Ukraine crisis is about
Great Power oil, gas pipeline rivalry [Online]. Available:
http://www.theguardian.com/environment/earth-
insight/2014/mar/06/ukraine-crisis-great-power-oil-gas-rivals-
pipelines
[57] Keith Johnson. (2014, September 24), U.S. Strikes ISIS Oil
Installations [Online]. Available:
http://foreignpolicy.com/2014/09/24/u-s-strikes-isis-oil-
installations/
[58] Treaty of Versailles, Paris, 28 June 1919, League of Nations
Treaty Series, available from
https://en.wikisource.org/wiki/Treaty_of_Versailles
[59] Covanent of League of Nations, available at
https://en.wikisource.org/wiki/Covenant_of_the_League_of_Nat
ions
[60] Antarctic Treaty, Washington, 1 December 1959, available from
http://disarmament.un.org/treaties/t/antarctic/text
[61] Environmental Treaties and Resource Indicators (ENTRI),
Available
From:http://sedac.ciesin.columbia.edu/entri/texts/acrc/cramra.txt
.html
[62] Protocol on Environmental Protection to the Antarctic Treaty,
1991, available from
http://www.antarctica.ac.uk/about_antarctica/geopolitical/treaty/
update_1991.php
[63] Outer Space Treaty, Washington DC, 10 October 1967,
available from
http://www.unoosa.org/oosa/en/ourwork/spacelaw/principles/leg
al-principles.html
[64] Moon Treaty, New York, 18 December 1979, available from
http://www.unoosa.org/oosa/en/ourwork/spacelaw/treaties/moon
-agreement.html
[65] Para 1, Article 11, Agreement Governing the Activities of States
on the Moon and Other Celestial Bodies, United Nations, 1979
[66] Article 1, Treaty On Principles Governing The Activities Of
States In The Exploration And Use Of Outer Space, Including
The Moon And Other Celestial Bodies, United Nations, 1967
[67] Para 5 -7, Article 11, Agreement Governing the Activities of
States on the Moon and Other Celestial Bodies, United Nations,
1979
[68] Part XI, Agreement of the Law of the Sea Convention, United
Nations, 1994
[69] Article 2, Treaty On Principles Governing The Activities Of
States In The Exploration And Use Of Outer Space, Including
The Moon And Other Celestial Bodies, United Nations, 1967
[70] A/AC.105/889/Add.14, Committee on the Peaceful Uses of
Outer Space, United Nations General Assembly
[71] CONAE website : http://www.conae.gov.ar
[72] Para 1, Article 15, Agreement Governing the Activities of States
on the Moon and Other Celestial Bodies, United Nations, 1967
[73] Article 12, Universal Declaration of Human Rights,1948
[74] Para 2 Article 11, The Antarctic Treaty, 1959
146
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-155-320.jpg)
![MATERIAL PROPERTIES
For analysis, material properties of CFRP were
calculated. To have symmetric properties in all the three
direction, Quasiisotropic (0, 90, +45, and 45)Figure 7: Stacking
Sequence sequence was selected as stacking sequence for our
material
Figure 7: Stacking Sequence
To find out the properties of CFRP, “ANALYSIS
AND DPERFORMAANCE OF FIBRE CCOMPOSISTEES”,
written by BHAGWWAN D. AGAARWWAL was concerned,
in CFRP carbon is fiber while epoxy was as matrix, following
calculations were carried out
The expression for elastic modulus is as
𝐸𝑐 = 𝐸𝑓 𝑉𝑓 + 𝐸 𝑚 𝑉𝑚
We have
𝐸𝑓 = 220 𝐺𝑃𝑎
𝐸 𝑚 = 3.54𝐺𝑃𝑎
𝑉𝑓 = 0.5
𝑉𝑚 = 0.5
By putting these values
𝐸𝑐 = 220 ∗ 0.5 + 3.54 ∗ 0.5
𝐸𝑐 = 112 𝐺𝑃𝑎
Now to find out elastic modulus in other two direction
𝐸 𝑇
𝐸 𝑚
=
1 + Ƹƞ𝑉𝑓
1 − Ƹƞ𝑉𝑓
ƞ =
𝐸𝑓
𝐸 𝑚
⁄ − 1
𝐸𝑓
𝐸 𝑚
⁄ + Ƹ
Putting values in above eqn we get
ƞ = .95
𝐸 𝑇 = 17.61 𝐺𝑃𝑎
Similarly 𝐺 𝐿𝑇 becomes
𝐺 𝐿𝑇 = 3.60 𝐺𝑃𝑎
To find out the elastic modulus we will find out the A
matrix for above sequence
For this we need Q matrix for each ply
𝑄 = [
𝑄11 𝑄12 𝑄16
𝑄21 𝑄22 𝑄26
𝑄61 𝑄62 𝑄66
]
𝑄11 =
𝐸𝐿
1 − 𝑣 𝐿𝑇 𝑣 𝑇𝐿
𝑄11 = 134.5 𝐺𝑃𝑎
𝑄22 =
𝐸 𝑇
1 − 𝑣 𝐿𝑇 𝑣 𝑇𝐿
𝑄22 = 17.7 𝐺𝑃𝑎
𝑄12 = 4.619 𝐺𝑃𝑎
𝑄66 = 𝐺 𝐿𝑇 = 3.69 𝐺𝑃𝑎
The Q matrix becomes for 0 deg ply becomes
𝑄 = [
134.5 4.6 0
4.6 17.75 0
0 0 3.6
]
For lamina at any angle the Q matrix becomes
𝑄̅ = [
𝑄̅11 𝑄̅12 𝑄̅16
𝑄̅21 𝑄̅22 𝑄̅26
𝑄̅61 𝑄̅62 𝑄̅66
]
Above values can be calculated from eqns as follows
𝑄̅11 = 𝑄11 𝑐𝑜𝑠4
𝜃 +𝑄22 𝑠𝑖𝑛4
𝜃 + 2(𝑄12
+ 2𝑄66) 𝑠𝑖𝑛2
𝜃 𝑐𝑜𝑠2
𝜃
𝑄̅22 = 𝑄11 𝑠𝑖𝑛4
𝜃 +𝑄22 𝑐𝑜𝑠4
𝜃 + 2(𝑄12
+ 2𝑄66) 𝑠𝑖𝑛2
𝜃 𝑐𝑜𝑠2
𝜃
𝑄̅12 = 𝑄12(𝑐𝑜𝑠4
𝜃 + 𝑠𝑖𝑛4
𝜃) + (𝑄11 + 𝑄12
− 4𝑄66) 𝑠𝑖𝑛2
𝜃 𝑐𝑜𝑠2
𝜃
𝑄̅66 = 𝑄66(𝑐𝑜𝑠4
𝜃 + 𝑠𝑖𝑛4
𝜃) + (𝑄11 + 𝑄22 − 2𝑄12
− 2𝑄66) 𝑠𝑖𝑛2
𝜃 𝑐𝑜𝑠2
𝜃
𝑄̅66 = (𝑄11 − 𝑄12 − 2𝑄66 )𝑐𝑜𝑠3
𝜃 sin 𝜃 + (𝑄22 − 𝑄12
− 2𝑄66) 𝑠𝑖𝑛3
𝜃 cos 𝜃
𝑄̅66 = (𝑄11 − 𝑄12 − 2𝑄66 )𝑠𝑖𝑛3
𝜃 cos 𝜃 + (𝑄22 − 𝑄12
− 2𝑄66) 𝑐𝑜𝑠3
𝜃 sin 𝜃
Q matrix for 45 deg lamina becomes
𝑄̅ = [
44.14 36.76 29.09
36.76 43.96 29.09
29.09 29.09 3.6
]
150
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-159-320.jpg)
![Q matrix for -45 deg lamina becomes
𝑄̅ = ⌈
44.14 36.76 −29.09
36.76 43.96 −29.09
−29.09 −29.09 35.76
⌉
Q matrix for 90 deg lamina becomes
𝑄̅ = [
17.75 4.6 0
4.6 134.5 0
0 0 3.6
]
Now material properties can be found as
𝐸 𝑋 =
𝐴11 𝐴22 − 𝐴12
2
𝐴22 𝑡
𝐸 𝑦 =
𝐴11 𝐴22 − 𝐴12
2
𝐴11 𝑡
𝑣𝑥𝑦 =
𝐴12
𝐴22
𝑣 𝑦𝑥 =
𝐴12
𝐴11
𝐺𝑥𝑦 =
𝐴66
𝑡
So from above expressions we found out the matrix as
follows
Total thickness is “10 mm” & number of lamiae are 20, so the
total thickness of each angled ply becomes 2.5 mm
[𝐴] = 2.5 ∗ [[𝑄̅]0 + [𝑄̅]90 + [𝑄̅]45 + [𝑄̅]−45]
Putting all the related values in the above equation we get the
A matrix as
⌈𝐴⌉ = [
601.25 206.8 0
206.8 600.4 0
0 0 196.8
] 𝐺𝑃𝑎
Putting the values from above in the expression of Ex, Ey, Vxy,
Vxy & Gxy, we get
𝐸𝑥 = 53 𝐺𝑃𝑎
𝐸 𝑦 = 53 𝐺𝑃𝑎
𝐺𝑥𝑦 = 19.68 𝐺𝑃𝑎
𝑣𝑥𝑦 = 0.3449
𝑣𝑥𝑦 = 0.3439
The above calculated values were verified by using
NASTRAN software. Figure 8: Nastran Pattern Results These
values will be embedded in ANSYS Workbench as engineering
data in orthotropic properties. Further computation will be
carried out in the next steps.
Figure 8: Nastran Pattern Results
COMPUTATIONAL ANALYSIS
To check out for the structural integrity of the design, this
CATIA file was converted in “IGES” file. This igs file was
imported in ANSYS workbench. The static structural analysis
was carried out using different conditions. First of all the water
tank, structural analysis of each part is carried out separately.
The design of water alone was analyzed. The water of pressure
151
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-160-320.jpg)
![After running the simulations, following results were
given by ANSYS workbench. Values in the green boxes shows
Von Misses stress at that location.Figure 13: Results from 4g
Loads
Figure 13: Results from 4g Loads
The maximum equivalent stress shows an abrupt rise
due to the acceleration. Similarly the total maximum
deformation was too high. Similar analysis was carried out on
inertial load with respect to 5g acceleration. The following
results were obtained.
When simulations for 5g inertial loads were applied on
the whole structure, the maximum equivalent stress increased
by a big amount. This value is still under the elastic limit range,
so the design will not fail. Figure 14: Results of 5g Loads
Figure 14: Results of 5g Loads
Figure 15: Deformation at 5g Loads
Frrom the results of total deformation the maximum
deformation, the maximum deformation is 0.11176 m which is
too high and strucure will eventually break. So with this design,
we have to keep ouselve within 4g limit. We must try to stay
maximum at 3g. this holds good for all manuvers, translations
and rolls.
CONCLUSION:
With this we can conclude that any cargo helicopter
with weight lifting capacity of 4000 kg or more can be modified
so that it can be used in Aerial firefighting. For that a helitanker
is installed inside the cargo section of the aircraft, which will
by a snorkel from the water reservoirs easily. If we use light
composites like CFRP, it will have enough strength to resist
flight variations up to 3g. The helitanker will have the capacity
to carry 3200 liters of water. Compared with firefighting
helicopters, it will be an efficient helicopter.
REFERENCES
[1] F. Qian, V. Strusevich, I. Gribkovskaia, and Ø.
Halskau, "Minimization of passenger takeoff and
landing risk in offshore helicopter transportation:
Models, approaches and analysis," Omega, vol. 51, pp.
93-106, 2015.
[2] D. G. Nichols Sr, "Electric in-line snorkel pump for
helicopter tanker and method of operation," ed:
Google Patents, 2001.
[3] E. G. Keating, A. R. Morral, C. C. Price, D. Woods,
D. M. Norton, C. Panis, et al., Air Attack Against
Wildfires: Rand Corporation, 2012.
[4] B. D. Agarwal, L. J. Broutman, and K.
Chandrashekhara, Analysis and performance of fiber
composites: John Wiley & Sons, 2006.
153
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-162-320.jpg)
![Fig.5. Reflection parameters using AD450 substrate
Fig 6.Reflection parameters using AD450 substrate
Fig.7.Reflection parameters using FR4 substrate
Fig.8.Reflection parameters using FR4 substrate
As can be seen from the results of above plots, that
performance of Wilkinson power divider is almost same for
both the substrates i.e. low cost FR4 substrate and high cost
and low loss AD450 substrate. Therefore, it can be said that
the design fits well for any substrate and is very useful for the
cheaper substrates without compromising the performance.
V. CONCLUSION
RF transceiver commonly have a Wilkinson power divider
(WPD) or combiner circuit, techniques for designing WPD are
diverse. The paper shows comparison of WPD design on two
substrates having different loss tangent. FR4 being cheap and
easily available is defined as equally comparable to AD450 in
simulations. Next step is the fabrication and testing of these
WPD. The final circuit occupies less space and has good
matching results. With the thickness of 0.5mm, FR4 has
comparable behavior to specialized substrates and can be
considered as a robust and cost effective alternative for
applications in S-band.
REFERENCES
[1] Cripps, S.C., RF Power Amplifiers for Wireless Communications 2nded,
Norwood,MA: Artech House, 2006
[2] David M. Pozar, Microwave Engineering, John Wiley & Sons Inc., 2005
[3]Guillermo Gonzalez,”Microwave Transistor Amplifer Analysis and design,
Prentice Hall
[4] Maas, Stephen A.Nonlinear microwave and RF circuits, “Artech House
microwave library”, 1997.
2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.452.00 2.50
-36
-34
-32
-30
-28
-26
-24
-38
-22
freq, GHz
dB(S(1,1))
dB(S(2,2))
m1
freq=
dB(S(2,1))=-3.272
2.250GHz
m2
freq=
dB(S(3,3))=-36.784
2.250GHz
2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.452.00 2.50
-35
-30
-25
-20
-15
-10
-5
-40
0
freq, GHz
dB(S(2,1))
m1
dB(S(3,3))
m2
m1
freq=
dB(S(2,1))=-3.272
2.250GHz
m2
freq=
dB(S(3,3))=-36.784
2.250GHz
156
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-165-320.jpg)
![OPTIMIZE MANUFACTURING OF
UNIDIRECTIONAL CARBON PREPREGS FOR
SPACE APPLICATIONS
Mateen Tariq Ahmad Bashir, Dr. Sohaib Akbar, Dr. Sajid Mirza
Pakistan Space and Upper Atmosphere Research Commission
Karachi, Pakistan
Mateen_90@hotmail.com
Abstract— High performance composite prepregs are used
globally in aerospace applications for faster and easier
manufacturing. Manufacturing composite parts using
unidirectional carbon prepregs offers flexibility to optimize the
design and layers, resulting in superior properties. Using
prepregs as manufacturing medium, parts can be manufactured
with low void content. Unidirectional prepregs give control over
thickness hence high specific modulus and strength can be
achieved. Fatigue and acoustic performance of composite is
strictly related to type of reinforcement. Unidirectional
reinforcement plies offer good fatigue and acoustic performance
making them suitable candidates for space manufacturing.
Superior properties make them viable for manufacturing of
various parts of Launch vehicles such as payload fairing, motor
casings, fuel tanks and shims for flexible nozzle assembly. This
paper deals with the in-house manufacturing of unidirectional
carbon prepregs. The prepregs were manufactured using carbon
fiber tows and epoxy based prepreg resins. The fiber from tow
was mechanically placed parallel to each other resulting in
unidirectional carbon fiber prepregs. A solvent dip process using
resin bath and roller was used to manufacture prepregs. To stock
the batch of prepreg, cold storage system was utilized. The
process was developed using customized machinery and
equipment. Tests on prepregs showed satisfactory results in
terms of volume and weight fractions. Finally the prepregs were
cured and formed into unidirectional sheets for further tests.
Prepregs were cured in autoclave utilizing vacuum bagging
operation. The cured sheets were subsequently tested which
showed competent results hence asserting the material as suitable
candidate for use in space applications.
Index Terms— Unidirectional Carbon, Prepregs, Polymer
matrix composites.
I. INTRODUCTION
Space applications demand materials with high specific
stiffness and strength so that weight of structure may be
minimized. This need of space industry has lead towards
development of many advance fiber reinforced composite
materials. With the development of light weight composite
structures, many structural challenges can be addressed.
Among various advance composite materials, unidirectional
fiber reinforced composites offer better control on properties
[1]. With the use of unidirectional laminates, desired strength
can be achieved by placement of laminates as be modeled
design. Moreover degree of anisotropic behavior may also be
altered using unidirectional laminates.
Prepreg technology allows fabrication of intricate part with
high quality. Prepregs may be used manually as in hand layup
process of with automated placement machines. For
manufacturing of parts for space applications consistent and
high quality is required which can be achieved by using
prepregs. The desired volume fraction of fiber for prepreg for
space application is about 55-60% [2]. Choice of matrix and
reinforcement depends on the application. Carbon fiber
impregnated with epoxy matrix is used in many applications as
this combination offers superior mechanical properties.
Carbon composites with epoxy matrix have been used
extensively in producing aerospace parts such as landing gear
doors, flaps, aileron and other structural parts. Unidirectional
carbon fiber- epoxy composites offer competitive fatigue
properties in low earth orbits [3]. With superior properties,
unidirectional carbon fiber composites may be used as face
sheets for satellite payload fairing where high strength and high
fatigue life is required. As the space environment offers
vacuum thermal cycling, superior thermos-mechanical
properties of unidirectional carbon fiber composites make them
viable candidate for manufacturing various satellite parts
This paper is distributed in two parts, manufacturing and
testing. The former explains the in-house manufacturing of
carbon fiber prepregs and the subsequent sheet forming,
whereas the later describes the testing results of manufactured
unidirectional carbon fiber sheets.
II. EXPERIMENTAL
A. Materials
Material used for this study were oriented towards high
specific strength. Carbon fiber tow (10K) was used as
reinforcement for prepreg forming. Carbon fiber was chosen
due to its high strength and low coefficient of thermal
expansion. Choice of resin system depends upon the
application hence for high strength application, epoxy based
resin is desired. For this study, Swancor 2552 prepreg resin
was chosen due to its low viscosity and good impregnation
properties.
157
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-166-320.jpg)
![D. Testing of Composite samples
Two type of tests were carried out in this study. Initially,
prepregs were tested for calculating resin content. Four
samples of prepregs before curing were tested to calculate
average value.
Mechanical testing was performed on cured sheet to obtain
strength parameters. Tensile test according to standard ASTM-
D 3039 [4] was performed on a universal testing machine.
III. RESULTS AND DISCUSSION
A. Resin Contents
Resin contents were calculated in terms of total weight.
Tests results indicates average value of 40% resin content. Also
the presence of solvent was calculated as approximately 3%
within the matrix system.
B. Mechanical testing
Tensile test was performed on cured unidirectional carbon
samples. The average results are summarized in table 2.
TABLE II. RESULTS OF MECHANICAL TESTING
Breaking load 23 KN
Ultimate Tensile Strength 1440 MPa
Modulus of elasticity 97 GPa
Elongation 1.4 %
IV. CONCLUSION
Unidirectional carbon fiber prepregs were manufactured
using in-house designed prepreg machine. The prepreg
machine was designed on the concept of filament winding. Dry
fiber from spool was impregnated in a resin bath and then
wrapped around a large mandrel. Semi cured prepreg sheets
were removed from mandrel and stored in low temperature
environment for preservation. The sheets were then withdrawn
and cured in unidirectional laminates to form rigid sheet for
subsequent testing. Test results of prepreg sheet and cured
composite part shows satisfactory results. Prepregs contain
40% resin which is a nominal value. Moreover, cured sheets
offered excellent mechanical properties. Hence composite
structure with light weight and high strength was formed using
unidirectional carbon fiber prepregs.
REFERENCES
[1] Purslow D., “The shear properties of unidirectional carbon fiber
reinforced plastics and their experimental determination,
Aeronautical research council current papers”, 1977 Retrieved
from: naca.central.cranfield.ac.uk/reports/arc/cp/1381.pdf
[2] Larberg Ylva, “Deformability of Unidirectional Prepreg
Materials, Licentiate Thesis Stockholm”, Sweden. 2009
Retrieved from: www.diva-portal.org/smash/get/diva2:224958/
FULLTEXT01.pdf
[3] A. Anvari, “Fatigue Life Prediction Of Unidirectional Carbon
Fiber/Epoxy Composite In Earth Orbit”, Int. J. of Appl. Math
and Mech. 10 (5): 58-85, 2014 Retrieved from:
ijamm.bc.cityu.edu.hk/ijamm/outbox/Y2014V10N5P58C25132
388.pdf
[4] ASTM D 3039, “Standard Test Method for Tensile Properties of
Polymer Matrix Composite Materials”,
159
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-168-320.jpg)
![describe the design and analyze results of EPC module for
new generation communication satellite payload receivers for
a given set of specific space qualified requirements. The
output voltage ripples have been minimized to 25mV is
presented. The input/output voltages, in-rush current
protection, TM and TC characteristics, bus voltage ripples and
other requirements which make its design complex and
significant for RF point of view. Both the output current and
voltage requirements are presented and met successfully. The
design presented is also compliant with the TC requirements.
The under voltage and over current protections are also
implemented to protect satellite in case of short circuits or
lower main bus voltage. All the four major modules of EPC
(main DC-DC converter unit, auxiliary DC-DC converter, TC
circuit and TM circuit) are discussed and analyzed along with
their simulated results.
REFERENCES
[1] Ioana-Monica, Pop-Calimanu, Prutianu Florin, and Popescu Viorel.
"Design and simulation of DC/DC boost converter used for distributed
sensing system based on a multidrop sensor network with RS485
interface", 2012 10th International Symposium on Electronics and
Telecommunications, 2012.
[2] Pequet, E.; Delporte, P.; Fayt, P.; Gak, M.; Canon, T., "ESA qualified
EPC for telecommunication satellites TWTA," Vacuum Electronics
Conference, 2000. Abstracts. International , vol., no., pp.2 pp.,, 2-4 May
2000.
[3] R. E. Sorace, V. S. Reinhardt, and S. A. Vaughn, “High-speed digital-to-
RF converter,” U.S. Patent 5 668 842, Sept. 16, 1997.
[4] Castiaux, J.P.; Bury, P.; Liegeois, B., "Power conditioning units for high
power geostationary satellites," Power Electronics Specialists
Conference, 1997. PESC '97 Record., 28th Annual IEEE, vol.1, no.,
pp.722,733 vol.1, 22-27 Jun 1997.
[5] http://www.space-airbusds.com
[6] https://www.thalesgroup.com
[7] http://www.sstl.co.uk
[8] Garrigos, A.; Carrasco, J.A.; Blanes, J.M.; Sanchis-Kilders, E., "A
power conditioning unit for high power GEO satellites based on the
sequential switching shunt series regulator," Electrotechnical
Conference, 2006. MELECON 2006. IEEE Mediterranean , vol., no.,
pp.1186,1189, 16-19 May 2006.
164
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-173-320.jpg)
![DSP Based Electro-Hydraulic Actuator Control
with Irretraceable Feedback Error
Compensation
Najam-Ud-Din Ahmed, Syed Aleem Azhar, Khadim Hussain
Abstract—This paper describes the interfacing detail of
Electro-Hydraulic actuator with DSP and addresses the
issues which commonly occur during development. The
purpose of designing digital controller on DSP is it’s
easy configurability via software and requirement of
less analog part. Scenarios where a dc-dc converter
used as voltage source may induce erroneous feedback
in the feedback control system due to its dual output
mismatch. And despite a good PID controller real
position of manipulator may have errors which are
irretraceable even by the intelligent controller schemes.
A judicious and intelligent use of available DSP
peripherals like ADCs, DACs etc and minor adjustment
of circuit can effectively increase the real time accuracy
of whole system against irretraceable errors.
Keywords—DSP based PID, PID Controller,
Irretraceable error, Erroneous feedback compensation
I. INTRODUCTION
It is common practice to use DSP or any digital
controller to control analog manipulators and
actuators in mechatronic applications [3]. However,
noise and supply voltage accuracy dependability of
analog systems is always associated with them and
affect the accuracy of the overall system. There are
many advanced as well as conventional control
schemes to increase the performance of the actuator
[2] but these schemes might be less affective if the
feedback is erroneous or we have high tolerances in
the supply voltages. This paper discusses such a
scenario where a dc-dc converter used as voltage
source may induce errors in the feedback system due
to its dual output mismatch. And despite a good
controller real position of manipulator may have
errors which are irretraceable by the intelligent
controller schemes. In addition, use of digital
controller specially a DSP helps in generation of real
time rate for interfacing. Although various techniques
are available for getting precise band-gap reference
voltage [1] but DSP based controller also adds noise
immunity which is always associated with analog
systems.
II. PROBLEM STATEMENT
Two hydraulic pistons are used in a robotic arm
for high speed application prototype. Feedback is
based on linear potentiometer for which dual supply is
required. It is known that 1% error in supply voltage
will produce same error in the feedback which will be
not traceable in potentiometer based feedbacks and
still be added in the real position of actuator.
Output Voltage (Vout) at potentiometer viper is,
(1)
(2)
(3)
(4)
(5)
( )
(6)
This error may be induced due to two factors
1. Due to tolerance in supply
2. Due to mismatch in dual supply
There is always a pinch of tolerance in supply and
it is always expensive to use high grade supply unit or
165
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-174-320.jpg)
![circuit but also the errors associated with use of
additional components in any circuit.
REFERENCES:
[1] LinHaiCui. Design of a High Precision Band-gap
Voltage Reference. Proceedings of the 2011
International Conference on Electronic & Mechanical
Engineering and Information Technology. IEEE, 2011 :
2187~2190.
[2] Sergey Edward Lyshevski and Trevor C Smith.
Tracking Control of Direct-Drive Servos. Proceeding of
the 2011 International Conference on Decision and
Control and European Control Conference(CDC-
ECC).Orlando, FL, USA, IEEE, 2011 : 1602~1607.
[3]Hai-tao Wang, Ze-Zhang and Xiang-yu Liu. Design of
Control System for Brushless DC Motor based on
TMS320F28335.Proceeding of the 2011 Third
International Conference on Measuring Technology and
Mechatronics Automation. IEEE, 2011 : 954~958.
[4]D-module D-sign T manual.
[5]VisualDSP++ 4.0 - C/C++ Compiler and Library
Manual for SHARC Processors.
[6]ADSP-21065L SHARC DSP - User's Manual and
Technical Reference
170
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-179-320.jpg)
![Design for Test Approach using FPGA for BPSK
Modem
Syed Jahanzeb Hussain Pirzada1
, Muhammad Faisal Arain1
, Rahman Mehboob1
1
Space and Upper Atmosphere Research Commission (SUPARCO), Pakistan.
Abstract— Implementation of modem has always been the key
interest for researchers working in the field of communication.
Nowadays, at low frequencies (in kilohertz range) the modulation
and demodulation is performed digitally. As, modem is utilized
for telemetry and telecommand communication between satellite
and ground station. This work provides the design of test bed for
modem. The testing is performed using ECSS based Standards.
This paper explains the design of test bed for testability of Binary
Phase Shift Keying modem implemented on Field Programmable
Gate Arrays. The testing of FPGA based modulator was
performed using signal analyzer. On the other hand the testing of
FPGA based demodulator was complicated and needs to be
design on seperate hardware; as it contains carrier recovery, bit
syncronization and bit recovery. Hence, digital signal processor
and FPGA based co-processing test bed was design and
developed for tesing of demodulator. The experimental results
show the architecture and testing results of the proposed
approach. The proposed methodology describes the problems
encountered during testing and proposed solutions to it. Such as
in demodulation, the bit recovered from demodulation can be
analyzed on oscilloscope but it is cumbersome. Therefore, an
asynchronous bit recovery algorithm is employed with serial
communication for representation and verification.
Keywords- DSP, BPSK, modulator , demodulator.
I. INTRODUCTION
Digital modulation is utilized to convert digital symbols into
waveforms to make it compatible with the characteristics of
transmission channel [3]. In digital modulation, a baseband
signal is changed into band pass signal compatible for
transmission. Similarly, demodulation is utilized in receiving
end to recognize the detected data.
Binary Phase Shift Keying (BPSK) is employed in
transmission and reception of telemetry and telecommand
data in satellites. It is the simplest form of phase shift keying
(PSK). Specially, at low frequencies (in kilohertz range) the
modulation and demodulation is performed digitally.
Generally it consists of two phases one is inphase with the
transmitted carrier and other is out of phase with the
transmitted carrier as represented in Figure 1. The two
conditions are generally represented as logic 0 and logic 1, the
formula associated with it represented in eq-1 and eq-2 [4].
Figure 1: Phase change of BPSK signal.
A Test bed is essential for rigorous testing of the system.
Implementation of digital modulation schemes have become
more and more attractive and simple with the used of digital
signal processors (DSP) and Field Programmable Gate Arrays
(FPGA). Recently, researcher are utilizing the processing
power of FPGA and DSP together to implement systems.
Testing of BPSK modem is very important as all the data
transmitted through the transmission channel depends on the
correctness of implmentation of baseband modualtion.
The remaining paper is organized as follow. Section II
provides the previous work. Section III explains the proposed
methodology. Section IV explains experimental results.
Section V provides conclution of this paper.
II. PREVIOUS WORK
Many researchers have designed and developed BPSK
modulator and demodulator in the Past [2] [5] [7] [8]. The
testing with genralized equipments are usually quoted in many
implementation. Including some results are acquired in the
form of software simulations. In this paper a testbed based
testing mechanism is proposed. Although it is formed using
traditional testing equipment but incombination with custom
solution.
III. PROPOSED METHODOLOGY
Testing of BPSK modem is proposed in this paper. The test
setup is composed of two sets of modems. One is the designed
BPSK modem to be utilized in telemetry and telecommand
module and other is for testing of designed BPSK modem.
For testing of BPSK modulator and demodulator designed on
FPGA to be utilized for telemetry and telecommand. A testing
171
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-180-320.jpg)
![TABLE I
PERFORMANCE CHARACTERISTICS OF BPSK MODULATOR
S.No Modulation Accuracy
Category Results Peak Unit
1 EVM 0.662 1.665 %
2 Magnitude Error 0.561 1.661 %
3 Phase Error 0.20 0.82 %
4 Amplitude Drop 0.01 - dB
5 Origin Offset -73.86 - dB
6 Gain Imbalance 0.00 - dB
7 Quadrature Error 0.00 - deg
8 SNR (MER) 43.58 - dB
9 RHO 0.999 - -
Figure 5: Spectrum representation of BPSK modulator on signal
analyzer.
Testing of demodulator is performed with DSP based BPSK
modulator and FPGA based debugging equipment. Result of
simulation on Code Composer Studios (Software used for
programming TI DSP Kits) are shown in Figure 6, Upper half
of the figure shows input converted to NRZ signal and in the
lower half corresponds to modulated BPSK signal [1]. The
BPSK modulator implemented on Texas Instruments
TMS320C6713 DSP kit has variable frequency and data rate
configurations. Althought in this testing application data rate
and frequency is fixed. The data from BPSK modulator is
provided through test computer serially and it modulates the
data. The modulated data is then input to FPGA based BPSK
demodulator. After the demodulation the data is transfered to
FPGA based Debugger for viewing the retreived output.
V. PROBLEMS AND SOLUTIONS DURING TESTING
Several problems are encountered during the testing of BPSK
modem implemented on FPGA. Firstly, The bit recovered
after demodulation are continous, in long data stream it will be
difficult to monitor it in osciloscope. Therefore, a uart based
debug ging is implemented on FPGA for validation of
demodulator performance. Secondly, Analog to digital
converter has been tested seperately but on integration with
FPGA and DSP based modualator the rescaling was required.
Lastly, there was a transmission rate missmatch between ADC
and uart transmission. So the data from ADC is stored in
memory and then transmitted through uart transmission.
Figure 6: CCS Graph results of NRZ converted data and BPSK
modulated signal.
VI. CONCLUSION
Our experiment showed testbed setup for BPSK modem.
Test setup for testing of BPSK modem using signal analyzer
and DSP/ FPGA test setup was proposed. The problems
encountered are discussed.
ACKNOWLEDGMENT
This work was supported by Space and Upper Atmosphere
Research Commission (SUPARCO). Authors also
acknowledge the efforts of Mr. Muhammad Faisal and Mr.
Razi Iqbal for continuous support and efforts for completion
of this research work.
REFERENCES
[1] G.Susinderrajan, S.Saran Kumar, T.Shankar, K.Sreeram,
P.V.Ramakrishna, "Design of a Low Cost All-Digital PSK/PM Satellite
Telemetry Transceiver", IEEE Electron Device Letters, vol. 20, 569–
571, Nov. 1999.
[2] S. Ruque, I. Ruiz, E. Carrión, "Simulation and implementation of the
BPSK modulation on a FPGA Xilinx Spartan 3 xcs200-4ftp256, using
Simulink and the System Generator blockset for DSP/FPGA", IEEE
Electron Device Letters, vol. 20,569–571, Nov. 1999.
[3] B.Sklar, “DigitalCommunications Fundamentals and Applications”, 2nd
ed. Englewood Cliffs, NJ: Prentice-Hall PTR, 2001.
[4] S. Johanna, Ruque, D.Ruiz, Carlos, “Simulation and implementation of
BPSK system on a FPGA board using system generator blockset for
DSP/FPGA”, School of Electronics and Communication, Technical
university Loja 2005.
[5] C.J. Harsha, D.H. Sandeep, A.S. Mali,"Hardware Implementation of
BPSK system on Virtex2-Pro FPGA using Xilinx System
Generatror".International Refereed Journal of Engineering and Science
(IRJES), vol. 2, pp 18-24, January 2013.
173
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-182-320.jpg)
![[6] J.G. Proakis, Digital Communication, 5th edition, McGraw Hill New
York, 2001.
[7] H. Malik, D.R. Rotake, M. Mahajan,"Design and Implementation of
BPSK Modulator and Demodulator using Vhdl". IOSR Journal of
Electronics and communication Engineering (IOSR-JECE), Vol. 9, pp
98-105, May-Jun 2014.
[8] N. P. Shrirao, A.P. Thakare,"Design of Digital Modulators:BASK,
BPSK and BFSK using VHDL". International Journal of Advanced
Research in Computer Science and Software Engineering (IJARCSSE),
Vol.3 january 2013.
174
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-183-320.jpg)
![Design of a Fuzzy Logic Water Level
Controller
Taimoor Zahid
Institute of Space Technology,
Islamabad, Pakistan
taimoor187@yahoo.com
Nosheen Zafar
COMSATS Institute of Information Technology,
Islamabad, Pakistan
nosh_zaf@hotmail.com
Abstract— This paper analyzes the effectiveness of
water level control using the concepts of Fuzzy Logic.
We have designed a Fuzzy Logic Controller for Water
Tank Level Control by using the linear control valve,
the level transmitter and the fuzzy control rules built in
the Fuzzy Logic Tool-Box and Simulink in MATLAB
and have then compared the control effect with that of a
Proportional–integral–derivative (PID). This design can
be very valuable for the industries where water level
control is extensively used and a slight deviation can
lead to major accidents and huge losses in revenue.
Therefore, it has become necessary to develop an
accurate and cost effective control system. The purpose
of such controller is to sustain a set point at a particular
value and be able to dynamically accept the new values.
Sometimes steady-state error and overshoot can no
longer be taken into account by the PID Controller due
to increase in the complexity, hysteresis, nonlinearity
and coupling of control object. While on the other hand,
with the development of the Intelligent Control Theory,
the application of Fuzzy Controller has become very
popular in practical industrial automation applications
and these problems can be successfully dealt with using
Fuzzy Control.
Keywords— Water Level Control, Fuzzy Logic, PID,
Tank, Mamdani.
I. INTRODUCTION
During the past several years, Fuzzy Control has
emerged as one of the most active and fruitful areas
of research in the application of Fuzzy Set Theory
[1]. Due to its robust, precise and orderly approach
for operating the controller, Fuzzy Logic can be used
in several applications. The field of Control System
Engineering has benefited the most from the success
of Fuzzy Set Theory. The reason behind this
unexpected success is high performance rate and low
cost. Although PID-Controllers are easy to install but
their performance can deteriorate quite fast when
used in non-linear systems [2]. Moreover, the
selection of the parameters of PID is also a difficult
task. If one does not choose the fine parameters, it
will badly impact the working of controller. Fuzzy
Control gives more attention to various parameters,
such as the time of response, the error of steadying
and overshoots [3].
This paper shows the development and
implementation of Fuzzy Logic Controller for Flow
Control Application because Fuzzy Controller can
achieve greater control results where limit of
overshoot is strict [4]. We have taken the level of the
water tank as an object and designed a Fuzzy
Controller in MATLAB. The Simulink models are
generated to see the working of tank. The controllers
are then added in those models to check and
understand the working of the designed controllers.
The control effect is examined and compared with
the effect of PID Controller. After designing the PID
and Fuzzy Controllers, the results are compared. This
approach of simulation by MATLAB allows us to
solve more complicated problems and develop the
measuring states to improve modeling and analysis of
system and to allow more control in the designs [5].
The overall goal of this paper is to show which
controller is better for flow and control applications.
II. THE TANK PROCESS
Input, Process, Feedback and Output are the four
steps of tank process.
1) Input: A square-shaped waveform of
amplitude 0.5 and frequency 0.1 Hz is
generated from the signal generator as given
by (1).
y(t) = Amp * waveform(freq, t) (1)
2) Process: The Tank-Process includes valve,
tank and controller. The water is pumped
into the tank via valve with controllable
degree of opening. Water flows out of the
tank through a hole in the bottom with area
0.05 m2
. The height of the tank is 2m, and
the cross-sectional area of the tank is 1m2
.
175
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-184-320.jpg)
![Technically, the water level in the tank can
be described mathematically using (2).
= √ – (2)
Where h is the water level [m], is the influent
flow rate [m3
/h] and and are parameters
depending on the tank and outflow areas. The
influent flow rate depends on the degree of valve
opening. In simulations, the water level in the tank is
controlled towards some pre-specified value
(reference signal) by letting the controller manipulate
the degree of valve opening. Eq. (2) will be used in
designing the Simulink model for controlling the
water level for the designed tank.
Figure.1 Water Tank System
3) Feedback: The signal is generated from the
signal generator and is further processed
through controller. This output from the
process becomes a feedback. This feedback
goes to a summer named error. It is named
so because it calculates the difference
between actual and desired water level. The
actual value is the one which is coming
through feedback and the desired is the one
which is coming straight through the signal
generator.
4) Output: The output can be generated via:
(a) To workspace: It writes the input to the
specified array or structure in MATLAB’s
main workspace. The “To Workspace”
block takes the signal as input and writes the
data of the signal into the workspace of
MATLAB. Data is not available until the
simulation is paused or stopped. This block
writes the data to an internal buffer during
the simulation and that data is written to the
workspace when a simulation is completed
or paused.
(b) Comparison: This block generates a plot and
that plot gives the comparison between the
reference signal and system output.
(c) S-function: This block provides a direct
access to the additional parameters to be
passed to S-function. The functions
parameters can be stated as variables
separated by commas or as MATLAB
expression. This function displays the
number of I/O ports plus the names of
specific S-functions. In our case, the built-in
file named “tankdemo” is used for
generating the animation. The S-function
used in this application refers to the
“animtank” which is built-in in the
MATLAB directory. ANIMTANK stands
for “Animation of water tank system”.
III. PID-CONTROL OF TANK WATER
LEVEL
A. PID Controller
PID controller refers to a Proportional–integral–
derivative controller. In order to overcome the
shortcoming of self-tuning PID controllers, fuzzy
logic and neural networks are utilized [2]. The
conventional PID is commonly utilized in controlling
the level, but the parameter of those controllers must
be turned by tuning method either in time response or
frequency response to meet their required
performances [6, 7]. The block diagram of PID
Controller with constants is shown in Figure. 2. P
element: e (t) is the proportional to the error at the
instant t, which is the present error. I element:
∫ dt is proportional to the integral of the error
up to the instant t, which can be interpreted as the
accumulation as the past error. D element: is
proportional to the derivative of the error at the
instant t, which can be interpreted as the prediction of
the future error. [8]
Figure.2 Block diagram of PID Controller
176
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-185-320.jpg)
![B. Simulink Model of Water Level Controller
Considering the tank process described above, the
Simulink Model is designed the as shown in the
Figure.4 and is saved with the name of
“tankpid.mdl.” Figure.4 illustrates that the process
includes PID Controller, valve and water tank.
1) Controller: The PID controller is the built-in
block in the Simulink. Its block diagram
representation is shown in the Figure.3. As
described earlier, it has three parts namely
proportional, integral and derivative. The input
is an error signal (difference between the
reference signal and system output) which when
comes into the controller is divided into three
respective parts and output from each of these
parts is then summed up to give a stable output.
The output of this block is the weighted sum of
the input signal, the integral of the input signal
and the derivative of the input signal [9]. In this
case, this controller is acting as a PD-controller
as the parameter which set for the integrator is
zero.
Figure.3 Block Diagram for PID-Controller
2) Valve: The water is pumped into the tank via
valve with controllable degree of opening. 0
means the valve is closed and 1 means that the
valve is fully open. The more the valve is open
(the closer the value is to one), the larger the
inflow to the tank. As illustrated in Figure. 5,
the product of the condition provided in the
limited integrator and inflow rate of the water
gives the rate at which water flows out. One of
the important parts of the valve is limited
integrator. The reason behind using the limited
integrator is to stop the output from surpassing
specific levels. The function of this block is
that, as soon as the output surpasses the limits,
it turns off the integral action for preventing the
integral wind up. The limits can be changed for
the period of simulation.
Figure.5 Simulink Model for Valve
Figure.4 Simulink Model for PID- Controller
3) Tank: The water level in the tank can be
described mathematically using (2) and tank
thus designed is shown Figure 6.
177
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-186-320.jpg)
![The input range is [-1, 1]. It consists of three
membership functions namely high, good
and low. Trimf is used for ‘good’, and
Trapmf for ‘high’ and ‘low’ as a nominal
function. These membership functions are
shown in Figure.9.
Figure.9 Membership function for input variable ‘level’
(b) Input 2: The rate (flow rate) is defined as the
second input. Its range is [-1, 1]. The
membership functions are named as ‘rising’
and ‘falling’. Trapmf is used as a nominal
function.
Figure.10 Membership function input variable ‘rate’
(c) Output: The designed system has only one output
given by the linguistic variable ‘valve’. It basically
represents the position of the valve by accounting for
the change in valve opening. The output space range
is [-1 1]. The membership function for the output is
shown in Figure.11.
Figure.11 Membership function for output variable ‘valve’
2) Rule Base: The input-output relationships
are used to develop the IF-THEN rules for
the FIS [10]. From the two input variables
and one output variable total of four
following rules are developed.
1. If (level is low) then (valve is open_fast)
2. If (level is high) then (valve is close_fast)
Figure.12 Simulink Model of Fuzzy Controller of Water Tank
3. If (level is good) and (rate is rising) then (valve is
close_slow)
4. If (level is good) and (rate is falling) then (valve is
open_slow)
B. Design of the Simulink Model
179
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-188-320.jpg)
![After setting the parameters of the FIS membership
function and making the fuzzy controller; the FIS file
named “tank.fis” is exported to the workspace. The
reason behind the fact that the derivative of the
process variable is taken instead of the derivative of
error is that the derivative mode amplifies the sudden
changes in controller input signal and cause large
variation in the controller output. [11]. Now the tank
with Fuzzy Controller can be simulated by using
model shown in Figure.12 and the output comes out
to be as shown in Figure.13.
Figure.13 Output of Fuzzy Controller
V. CONCLUSION
In plant process control, the measurement and control
of flow are two very important factors. But these two
factors are susceptible to disturbances because of the
different variables affecting the flow rate. Therefore
for optimizing the performance of a plant, the flow
controllers are required to be robust. Fuzzy Logic
Controller is proposed to replace the conventional
PID controller due to its superior applicability and
robustness [12]. The concept of Fuzzy Logic is a
representation of the human thinking and decision
making process. The experimental results show that
Fuzzy Control is better than PID Control not only
because of simplicity of its implementation, better
control performance, robustness and overall stability
but also because it gives a reasonable consideration
to variety of parameters like overshoot and time of
response etc and for this reason, it tunes and develops
the FIS more intuitively as compared to the PID
Controller.
REFERENCES
[1] L.A. Zadeh, “Fuzzy Sets”, Informal Control, Vol. 8, Pp. 338-
353, 1965.
[2] Y. F. Chan, W. Chan and H. T. Mok "Adaptive Neurofuzzy
Network Based PI Controllers with Multi-objective
Functions",Next-Generation Applied Intelligence, Vol. 5579, Pp.
604-613, 2009.
[3] Qi Li, Yanjun Fang, Jizhong Song, Ji Wang, "The Application
of Fuzzy Control in Liquid Level System", 2010 International
Conference on Measuring Technology and Mechatronics
Automation, 03/2010.
[4] Wangnipparnto S. and Tunyasrirut S., "Level Control in
Horizontal Tank by Fuzzy Logic Controller", 2006 SICE-ICASE
International Joint Conference, 10/2006.
[5] Duane Hanselman and Bruce Littlefield, "Book: Mastering
Matlab® 5", ISBN# 0-13-858366-8, Prentice-Hall, 1998.
[6] W. K. Ho, C.C. Hang, and J. H .Zhou, "Performance And Gain
And Phase Margins Of Well-Known Pituning Formulas,"
Accepted For Publication In IEEE Trans. Contr. Syst. Techno,
1995.
[7] J. G .Ziegler and N. B. Nichols, "Optimum Settings for
Automatic Controller," By ASME Trans. Vol. 64, Pp.759-768,
1942.
[8] Aravind, Sekhar R, and B R Vinod, "A generalized method for
improving the performance of controllers by combining PID and
DELTA rule", 2012 Annual IEEE India Conference (INDICON),
2012.
[9] “MATLAB documentation,” Mathworks
(www.mathworks.com )
[10] Pathmanathan E, and R Ibrahim., "Development and
implementation of Fuzzy Logic Controller for Flow Control
Application", 2010 International Conference on Intelligent and
Advanced Systems (ICIAS), 2010.
[11] Silviu Ionita and Emil Sofron, “The Fuzzy Model for Aircraft
Landing Control”, Prentice Hall, 2001
[12] Ronald R. Yager and Dimitar P. Filev, “Essentials Of Fuzzy
Modeling And Control”, Johan Wiley & Sons, 2002.
[13] D. Wu, F. Karray, I. Song, ‘Water level control by Fuzzy
Logic and Neural Networks’, IEEE Conference on Control
Applications, pp.3134-39, 2005.
[14] Namrata Dey, Ria Mandal and M. Monica Subashini, “Design
and Implementation of a Water Level Controller using Fuzzy
Logic”, International Journal of Engineering and Technology
(IJET), Jun-Jul 2013.
[15] J. Yen, "Fuzzy logic-a modern perspective", IEEE
Transactions on Knowledge and Data Engineering, 1999
180
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-189-320.jpg)
![Nuclear reactors have provided electrical power for
some of the U.S. space program’s greatest successes,
including the Apollo lunar landings and the Viking
Landers that searched for life on Mars. RTGs made
possible NASA’s celebrated Voyager explorations of
Jupiter, Saturn, Uranus and Neptune, RTG Power
sources are enabling the Galileo mission to Jupiter, the
international Ulysses mission studying the Sun’s Polar
Regions, and the Cassini mission to Saturn.
Pioneer 10 was launched from Cape Kennedy on 2
March 1972. It was the first interplanetary probe,
successfully navigating the asteroid belt before
Making rendezvous with Jupiter and Saturn The probe
was equipped with an array of instruments for
measuring such phenomena as the solar wind and the
magnetic and radiation fields surrounding Jupiter.
Nuclear reactors being a reliable and competitive source
of energy they have some risk too and as well it also
becomes difficult to do in some condition. The
overcome of these problems are also demonstrated in
this paper.
II. Space power concept:
To remove the waste, to lift a load and as well as for
communication in International space station or on
ground you will need sufficient power to do so for a
long time. Moreover it would be very difficult for to
generate electricity by thermal or chemical means in
microgravity environment [5].
However with the availability of radioactive isotopes
and as well as with the help of nuclear reactors the
Electricity and all the other requirements in ISS would
be easily available without any difficulty for several
years [5]
For the production of electricity heat energy can be
converted by many devices but the precise one is to use
Thermo electric generator which typically works on
Seebak Concept i.e
dV = SAB
.
(Kh - Kc)
Where
dV is Voltage difference, SAB is two dissimilar
metals, & Kh - Kc is temperature gradient.
In this way Nuclear reactors can provide infinite power
for almost any time. However, they are not practicable
for applications below 15 kilo watt (kW).
Figure 1 : regimes of possible space power applicability
182
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-191-320.jpg)
![Radioisotopes are considered to be best used for
continuous supply of low levels (up to 5 kW) of power
or in combinations up to many times this value. For this
reason, especially for long interplanetary missions, the
use of radioisotopes for communications and the
powering of experiments are preferred. The nuclear
process shown in Fig. 2 can either be a critical reactor or
radioisotope fuel source such as plutonium oxide. In
either case the heat can be converted to electricity either
statically through thermoelectric or a thermionic
converter, or dynamically using a turbine generator in
one of several heat cycles (Rankin, Stirling, Brayton)..
The nuclear workhorses for current space missions are
the RTGs and the TEGs powered by radioisotopes in the
Russian Federation that provide electricity through static
(and therefore reliable) conversion at power levels of up
to half a kilowatt, or more by combining modules. And
that energy can be used for to run machinery. [2].
Space Propulsion concept
Nuclear reactor can also be used in rocket propulsion
system. Propulsion system in space is to change the
position Or velocity (v), of space craft under the strong
influence of gravity. [6]
When a space craft moves it generally term is
momentum (mv) so, the basic concept of propulsion is to
change the momentum (mv) of a space craft, and change
in momentum is known as impulse (Imp). So, the basic
aim of propulsion in space is to create impulse, to
measure the impulse is often term is specific impulse
(Isp).[6]
Specific impulse also known the exhaust velocity (ve)
(OR) the impulse per unit weight ( ) but while
discussing the engine in space there is no weight so it
can become impulse per unit mass
The difference arises here only by the acceleration due
to gravity (g).Where its value is 2
on earth.
However the value of g and reaction mass is not
important while the vehicle discussing in space. [4].
The two parameter specific impulse and the mass of the
rocket during launch and then in orbit is measure the
efficiency of propulsion energy.
Where the equation for the measure of specific impulse
as seen as (1)
(1)
Where represent mass of propellant and
represent mass flow.
Here it is seen that mass of propellant is inversely
proportional to specific impulse, So, In chemical
propulsion there is used hydrogen and oxygen which
provides a specific impulse about 4400 m/s with mass
ratio of earth escape of 15.[1]
Figure 2.The conversion of heat energy into kinetic or to
mechanical energy
183
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-192-320.jpg)
![However americium OR hydrogen heated with fission
reactor can achieves twice the impulse with mass ratio
of 3.2. And with different core it can be generate much
more impulse i.e. seven times having mass ratio of only
1.2 [1].
III. Nuclear rocket
Nuclear rocket is the rocket which use nuclear reactors
usually fission concept instead of chemical propulsion to
generate thrust it include nuclear thermal rocket (NTR),
Nuclear electric rocket (NER) or Hybrid (NTR/NER).
[2]
These rocket are propelled by the force of nuclear
explosion, usually liquid hydrogen is heated via nuclear
fission which subsequently can be expanded in nozzle
and thus accelerated at a high ejection velocity (6,000 to
10,000 m/s) to generate thrust. There are also many
design are find for nuclear thermal propulsion i.e. liquid,
solid or gas core rocket. [6]
They use solid, liquid or gas core reactors respectively.
Electro thermal propulsion rocket have been use in
many orbital mission in this type of rocket the propellant
is heated electrically (heat energy of nuclear reactor are
converted into electricity) and then it accelerate the
ionized gas via electrostatic force at supersonic velocity
having range from 1,000 to 5,000 m/s and thrust range
0.01 to 0.5 N. But it has low Specific impulse therefore
it can be use only for interplanetary missions, however,
it can be use for long time instead of chemical rocket [2]
The nuclear rocket has many advantages which
overcome on chemical rocket. Like it can be use in
much complicated mission even it can be travels to the
edge of our solar system or even to our nearest cluster,
nuclear space craft can be also used for to carry heavy
payload or to carry flyby, rover etc to the far planet in
our solar system, because it provides a large specific
impulse as a result faster travel and the probability for
complicated missions can be met, secondly nuclear
reactor has low molecular weight which can increase the
propulsive force per unit of propellant flow and allowing
us for increase the proportion of total weight of payload
[3].
IV. Microgravity effect on core
reactor
In microgravity environment every mass tends to be in
state of motion (try to revolve from heavenly bodies
near them due to centripetal force) So, The big challenge
of core reactors in microgravity environment is to
control the flow of fission. It seems like very difficult
but under certain design it is possible to keep the
reaction self sustain. So the level of energy would be
maintained. In core reactors usually the flow is
turbulent due to the randomness motion of molecule.
And hence it becomes difficult to keep flow within
specified boundaries. For this reason Reynolds number
flow are preferred for microgravity environment.
If it is not under the specified boundaries then the chain
reaction times would be increases due to uncontrolled
condition in the core reactor. And some parts of the core
reactor are exposed and hot spot would be appear on that
as well some parts of the core reactors would be damage
due to high temperature because the distribution of heat
is not possible in such circumstance.
184
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-193-320.jpg)
![Figure 3 Precise core reactors for microgravity environment
1, Reflector moderator 2, Gaseous fissile zone 3,
Working medium flow zone, 4, Fissile material
diminution replenishment 5, working medium Inlet.
To overcome on this problem a special core is design
(shown in figure 3) to sustain fission reactor under
control especially with this core the possibility of
environmental hazard would be reduce. And various
high temperature fuels are possible with this design. in
The fissile material for example Uranium or americium
would be located in the center of cavity enclosed by the
neutron moderator reflector. The working gas flow is
close to the cavity walls and it is heated by the high
temperature plasma radiation which causes the fission
reaction to be self sustained under microgravity
conditions. [8]
V. Radiation hazards of Nuclear
rocket
Expose of a person to radiation does not mean that the
person would get cancer. People are exposing to
radiation on daily basis like on lesser extent from human
activities i.e.-rays. And radiation comes out from natural
environment i.e. in the earth cosmic rays and radon. The
radiation expose is measures in unit of Dose called
millirem. Over the course of year the average person
would expose to a total of 360 millirem to natural
radiation .In nuclear thermal rocket some of the
radioactive element would be release into the
atmosphere The particles which are hazardous to people
would remain high in the atmosphere for long course of
time gradually these particles being spread thinly across
the world. And eventually making their way on the
surface especially on ocean since these materials are
Insoluble once it reaches to the surface most of it would
become trapped by the soil or by the ocean and not pose
a health hazard. Thus most of the release materials
won’t be breath by people. But the small amount of
released material would be breath by the people and as
well this small amount will be also distributed across the
world. So, In this way the amount to be breath will be
much less the person may receive it less than one
millirem in 50 years. This small radiation is negligible
compare to the 15,000 millirem a person will get it from
natural resources over the period of 50 years. [7]
VI. Result and conclusion
In this paper we have emphasize on the importance of
nuclear reactors in space based application, where the
nuclear reactors and radioactive isotopes would provides
infinite heat energy for long durations during journey of
the space craft and that energy can be converted into
other form of energy and the need for power and
propulsion would be met. But in microgravity
environment it becomes difficult to do but however with
the help of gas core reactors and other special core
design would allow us to do so in microgravity
environment too. And the space exploration would be
become more precious and peaceful.
185
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-194-320.jpg)
![VII. Acknowledgment
I would like to thank national space agency SUPARCO
Space awareness society including Mr. Qamar Abbas
and Mr. Tahir hussain and My School principal Mr.
Sami ullah for to guide me as well for their support. I
also like to express my gratitude with SUPARCO
propulsion team for did my conceptual analysis.
I am also grateful to Mr. Rahman u din and Mr. Zulfiqar
Ali for Insatiable support and guide.
References
[1]. Rocket propulsion elements: introduction to
the engineering of rockets by George P. Sutton,
Oscar biblarz___7th
edition.
[2].The role of nuclear power and nuclear
propulsion in peaceful exploration of space by
International atomic energy agency Austria
September 2005.
[3]. Basic of space flight by National Aeronautic
and space administration sec 2, 2002.
[4]. Advance space propulsion for the 21st
Century
by Robert H .Frisbi a report from 9 August 2003 to
15 September 2003[5].Civilian usage of nuclear
reactors in space volume 1, science and global
security, 1989,
[6] Thruster precisely guide by Hiss, M martin and
K, Rachule October 2002.
[7] Space craft power for cassini
(http://www.jpl.nasa.gov/cassini/)
[8] Fission fragments heating for space propulsion
by C. Rubbia
186
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-195-320.jpg)
![Impact of Thermal Aging on Microstructure and
Mechanical Properties of high Sn Content, Sn-Pb
Solders
Muhammad Aamir1
, Riaz Muhammad1
, Naseer
Ahmed1
1
Dept. of Mechanical Engineering
CECOS University of IT & Emerging Sciences,
Peshawar, Pakistan.
aamir@cecos.edu.pk
Khurshid Alam2
2
Department of Mechanical and Industrial Engineering,
Sultan Qaboos University,
Muscat, Oman.
Abstract — The microstructure and mechanical properties of
solder joints in electronic components are evolving when
exposed to thermal aging. The current work is based on the
investigation of high Sn content Sn-Pb solders i.e. 96Sn-04Pb.
Scanning Electron microscopy was carried out to observe the
microstructure examination followed by chemical composition
with elemental mapping using Electron dispersive X-ray.
Mechanical properties were examined using tensile tests. The
effect of thermal aging on microstructure and mechanical
properties has also been studied. The results show that
mechanical properties are highly influenced by its
microstructure after thermal aging. Due to aging effect, the
microstructure becomes coarsen and ultimately the mechanical
properties including yield strength and ultimate tensile
strength degrades but increase in ductility is obtained.
Keywords—96Sn-04Pb, Microstructure examination,
Mechanical properties.
I. INTRODUCTION
The use of the solder alloys as metal joint can be dated back
to thousands of years and advancement in the electronics
industry are continuously in progress. However, the use of
surface mount technology (SMT) in printed circuit boards
(PCB) means that solders performs a leading role to make
electronic connections [1]. Surface mounting of the
components means that during impact loading the joint
transfer whole momentum of a component to the board.
Therefore, mechanical properties of the solder joint are of
great importance. Another important factor is the
accelerated aging due to relatively high temperature under
severe service condition. This also makes the joint in high
fraction of their melting point which results in
microstructural changes. Thermal aging is also responsible
for reduction in strength through grain growth after
microstructural coarsening process [2-6].
The mechanism for the decrease in strength in solder alloys
after aging is associated with coarsening of microstructure.
Coarser microstructure leads to fewer grain boundaries to
block the dislocation movements which ultimately cause
reduction in strength [7].
Solder joints are of great important in the electronic
industry, they are not only used to physically hold
assemblies together but also used for transmission of
electrical signals [8]. When electronic device is in use, the
solder connections faces mechanical stresses.
The on off switching of the system also leads cyclic
mechanical load and ultimately results in stresses in solder
joint [9, 10]. The decrease in the strength and microstructure
coarsening are also the result of high homologous
temperature and ultimately solder alloys will undergo creep
even at room temperature [11]. Pb based solders,
particularly 63Sn-37Pb or near eutectic Sn-Pb, has been
used in electronic industry for a long time. This is due to the
fact that Pb bearing solders have good metallurgical,
mechanical, solder ability, reliability, low cast, physical,
mechanical and metallurgical properties [12]. The use of Pb
in Sn-Pb solder is to reduce the surface tension which
ultimately results in improvement of wettability of the
solder joint [13, 14]. Sn is used as it is ductile in nature and
having good resistance to corrosion. High Sn concentration
also lead to high tensile and shear strength [15].
However, the concern about Pb toxicity and legislation to
ban on Pb bearing electronic products results in moving the
electronic industry towards environmental friendly green
products [16-18].
Therefore, electronic industries and academic institutes
accepted a challenge for characterizing new solders in terms
their mechanical properties, material prosperities and
manufacturing process compatibility with components and
printed circuit boards (PCB) [19].
The current work is based on developing and characterizing
a new solder with high Sn contents. The Pb content has been
limited upto 04% instead of 37% conventional Pb solders.
The reason for limiting Pb upto 04% is to get the properties
of Pb and make the environment as green as possible. The
microstructure including intermetallic particles and
mechanical properties have been examined at room
temperature and after thermal aging at 100°C for 50 hours.
187
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-196-320.jpg)
![II. EXPERIMETAL WORK
Preparation of good sample is necessary for experimental
study for better characterization and mechanical properties.
Samples of 250g ingots of pure metals with composition
(wt. %): 96Sn-04Pb is prepared by putting them in the
crucible and heated in an oven for 30 minutes at room
temperature to get the molten metal. A die composed of two
parts of aluminum and a central steel plate have been used
to get dog-bone specimen after the alloy is then cooled in
air. The final dog bone shape specimen is shown in Fig. 1
[20].
In order to analyze the microstructure, the specimen is cut
into pieces and mounted in bakelite powder using mounting
press for proper handling and avoiding any distortion. The
temperature for using bakelite was is at 160◦
C for 09
minutes with 130 Kg/cm2
. The mounted samples are
initially grinded with different grit sizes of silicon carbide
water proof sand papers. During grinding process, heat is
generated so tap paper is used for lubrication purpose to
avoid any damage to the surface. After sand paper grinding,
the specimens are polished with polycrystalline diamond
suspension with abrasive particle size of 6 μm, 1 μm and
0.25 μm to give extra shine to the surface and then cleaned
with distilled water to remove any residue from the
specimen left during polishing. The specimens are then
etched for 45 seconds using 5% hydrochloric acid and 95%
ethanol. Proper care is necessary for good observation of
grain particles and grain boundaries for examination of
morphology under Scan electron microcopy. Before Scan
electron microscopy (SEM), samples are coated with silver
paste and placed in sputter coater and vacuum evaporator
for gold coating to make the material conductive. The SEM
images have been taken at different location and
magnification followed by Electron dispersive X-ray (EDX)
to confirm the chemical composition of the alloy.
For thermal aging, samples for microstructure examination
and dog bone specimen are exposed to 100°C for 50 hours
in oven. The aging process is necessary to get the better
understanding of performance during extreme operating
conditions. For tensile testing, universal testing machine is
used to determine the mechanical properties. The data has
been taken at room temperature. The universal testing
machine with dog bone specimen is shown in Fig. 2.
Fig. 1 Dog bone specimen [20]
Fig. 2 Universal testing machine with dog-bone shape specimen
III. RESULTS AND DISCUSSION
A. Effect of Thermal Aging on Microstructure
The microstructure of 96Pb-04Pb alloy is described by low
and high magnification of images at different resolutions,
the images are taken for the as casted and after thermal
aging to examine the effect of changes in the morphology.
The “as casted” SEM images are shown in Fig. 3 and
“thermally aged” in Fig. 4. The chemical composition is
confirmed by Electron dispersive x-ray shown in Fig. 5. The
microstructure of 96Sn-04Pb is composed of soft Sn matrix
and hard IMCs of Pb. The black zone plate shape is the Sn
matrix and white color rods shaped are the IMCs of Pb. The
elemental mapping of Sn matrix and Pb particles are shown
in Fig. 6. These IMCs are usually brittle in nature in
comparison with soft Sn matrix and responsible for changes
in mechanical properties. The growth of these IMCs results
in coarsening of the alloy. The size of these IMCs depends
on many parameters including composition of alloy,
environmental conditions under sever temperature during
service and cooling rate. A highly temperature aged alloy
can produce coarse microstructure. By closely observing the
SEM images, it is observed that the Pb white particles
become wider and distributed in Sn matrix after thermal
aging. Since the properties of the solder joints are highly
microstructure dependent which changes reasonably during
its life time so, it is reasonably expected that after thermal
aging the mechanical properties would be worsen. Increase
in the growth of the grain size after thermal aging may also
leads in reducing the strength because larger grains create
fewer obstacles per unit area to the movement of
dislocation.
188
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-197-320.jpg)
![Fig. 7 Mechanical properties of “as casted and “thermally aged” alloy
TABLE I. TENSILE TEST DATA
Specimen Tensile Test
Description
Condition
Alloy Without
Thermal
Aging
Alloy With
Isothermal
aging at 100C
up to 50 hours
Elongation @ Peak (mm) 1.0090 1.4870
Load @ Peak (N) 435.00 324.00
Strain @ Yield (%) 1.2333 1.3946
Strain @ Peak (%) 1.6817 2.0231
Stress @ Yield (N/mm2
) 16.856 13.919
Stress @ Peak (N/mm2
) 19.595 14.595
IV. CONCLUSIONS
It is concluded that from the results and observations that
the properties of the solder joint are highly microstructure
dependent and a strong relationship exists between
microstructural and mechanical properties. It has been
observed that thermal aging leads to coarsen the
microstructure of the alloys and ultimately affects the
mechanical properties. Mechanical properties including
yield strength and ultimate tensile strength decreases but
considerable increase in ductility have been examined. This
also concluded that thermal aging is closely related to
mechanical properties. Therefore, great care must be taken
regarding microstructure evolution and mechanical
properties in designing and developing solders but for
electronics green and environment friendly Pb free solders
must be introduced.
REFERENCES
[1]. Prymark, J., et al., Fundamentals of microsystems
packaging. ed. RR Tummala, Mcgraw-Hill Book
Company, New York, USA, 2001: p. 420.
[2]. Jung, K. and H. Conrad, Microstructure coarsening
during static annealing of 60Sn40Pb solder joints: I
stereology. Journal of Electronic Materials, 2001.
30(10): p. 1294-1302.
[3]. Kang, J. and H. Conrad, Microstructure coarsening
during static annealing of 60Sn40Pb solder joints:
II Eutectic coarsening kinetics. Journal of
Electronic Materials, 2001. 30(10): p. 1303.
[4]. Jung, K. and H. Conrad, Microstructure coarsening
during static annealing of 60Sn40Pb solder joints:
III intermetallic compound growth kientics. Journal
of electronic materials, 2001. 30(10): p. 1308-1312.
[5]. Basaran, C. and Y. Wen, Coarsening in Bga solder
balls: modeling and experimental evaluation.
Journal of Electronic Packaging, 2003. 125(3): p.
426-430.
[6]. Tang, H. and C. Basaran, Influence of
microstructure coarsening on thermomechanical
fatigue behavior of Pb/Sn eutectic solder joints.
International Journal of Damage Mechanics, 2001.
10(3): p. 235-255.
[7]. Ma, H., et al. The influence of elevated temperature
aging on reliability of lead free solder joints. in
Electronic Components and Technology
Conference, 2007. ECTC'07. Proceedings. 57th.
2007. IEEE.
[8]. Efzan, E. and A. Marini, A review of solder
evolution in electronic application. International
Journal of Engineering, 2012. 1(1): p. 2305-8269.
[9]. Frear, D., D. Grivas, and J. Morris, A
microstructural study of the thermal fatigue failures
of 60Sn-40Pb solder joints. Journal of Electronic
Materials, 1988. 17(2): p. 171-180.
[10]. Hacke, P., A. Sprecher, and H. Conrad,
Microstructure coarsening during thermo-
mechanical fatigue of Pb-Sn solder joints. Journal
of Electronic Materials, 1997. 26(7): p. 774-782.
[11]. Wassink, R.K., Soldering in Electronics, 1989.
Electrochemical Publication. 177.
[12]. Sharif, A. and Y. Chan, Dissolution kinetics of BGA
Sn–Pb and Sn–Ag solders with Cu substrates
during reflow. Materials Science and Engineering:
B, 2004. 106(2): p. 126-131.
[13]. Abtew, M. and G. Selvaduray, Lead-free solders in
microelectronics. Materials Science and
Engineering: R: Reports, 2000. 27(5): p. 95-141.
[14]. Zeng, K. and K.-N. Tu, Six cases of reliability study
of Pb-free solder joints in electronic packaging
technology. Materials science and engineering: R:
reports, 2002. 38(2): p. 55-105.
[15]. Needleman, H.L., et al., The long-term effects of
exposure to low doses of lead in childhood: an 11-
year follow-up report. New England journal of
medicine, 1990. 322(2): p. 83-88.
[16]. Noor, E.E.M., et al., Wettability and strength of In–
Bi–Sn lead-free solder alloy on copper substrate.
Journal of Alloys and Compounds, 2010. 507(1): p.
290-296.
[17]. Abadin, H., et al., Toxicological profile for lead.
Atlanta (GA): Agency for Toxic Substances and
Disease Registry, 2007.
190
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-199-320.jpg)
![[18]. Harrison, M., J. Vincent, and H. Steen, Lead-free
reflow soldering for electronics assembly.
Soldering & Surface Mount Technology, 2001.
13(3): p. 21-38.
[19]. Smith III, E.B., Environmental Impacts and
Toxicity of Lead Free Solders. op. cit, 1999: p. 2.
[20]. Sadiq, M., R. Pesci, and M. Cherkaoui, Impact of
thermal aging on the microstructure evolution and
mechanical properties of lanthanum-doped Tin-
Silver-Copper lead-free solders. Journal of
electronic materials, 2013. 42(3): p. 492-501.
191
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-200-320.jpg)
![Effect of aryl diazonium salt fucntionalization on the electrical properties of MWCNTs
and MWCNTs/CF reinforced polymer composite
Madni Shifa, Fawad Tariq, Kiran Ali, Rasheed Ahmed Baloch
Quality Management Directorate General
Pakistan Space and Upper Atmosphere Research Commission (SUPARCO)
75270 Karachi, Pakistan
Email: madnishifa@yahoo.com
Abstract— Multiwall carbon nanotubes (MWCNTs) are kind
of inert and hydrophobic material. These MWCNTs do not
disperse properly and tend to agglomerate in polymer matrices
due to high surface area and strong van der Waal forces
between themselves. Numerous studies have been conducted
in the last few years to resolve the issue of CNTs dispersion
and solubilization in polymer matrix through functionalization
process. However, most of the functionalization treatments
introduce defects and results in deterioration of electrical
properties. In this study, MWCNTs were covalently
functionalized by direct grafting of aryl diazonium salt
(BF4N2-C6H4-NO2) and effect of functionlization on electrical
behavior was assessed under applied electric field in an
electrolytic cell. The MWCNTs were dispersed ultrasonically
in 2-propanol which act as an electrolyte in electrolytic cell.
Voltage was applied on copper electrodes and movement of
MWCNTs inside the electrolyte was carefully observed in-situ
under optical microscope. In case of functionalized MWCNTs
(f-MWCNTs), no effect was observed under applied voltage
up to 100 V whereas un-functionalized MWCNTs (p-
MWCNTs) tend to align themselves in regular fashion
showing conducting behavior. SEM analysis was performed to
see the consequence of functionalization treatment on
morphology of MWCNTs and optical microscopy was carried
out to see the effect of functionalization on the dispersion of
MWCNTs in cured and uncured state of polymer matrix.
MWCNT/CF reinforced polymer composites were fabricated
using hand layup and compression molding technique. DC
electrical conductivity & electromagnetic interference
shielding effectiveness (EMI SE) testing were performed to
examine the effect of MWCNTs functionalization on
fabricated samples. Test results showed that the
functionalization degraded the electrical properties of
MWCNTs.
Keywords— Aryl diazonium salt, Functionalization,
Morphology of MWCNTs, SEM analysis, EMI SE
I. INTRODUCTION
Carbon nanotubes (CNTs) are carbon based advance nano-
materials with unique set of structure and properties.
Commonly two types of CNTs are being used i.e. single wall
carbon nanotubes (SWCNTs) and multi wall carbon nanotubes
(MWCNTs). Outer diameter of SWCNTs is around 1-2 nm,
while the outer shell diameter of MWCNTs ranges from 10-
100 nm. A typical length range of CNTs varies from 10-1000
μm and have aspect ratio of about 1000:1 [1]. CNTs are one of
the most promising materials known to men; their mechanical,
thermal and electrical properties are superb in all aspects [2].
Considering electrical properties of CNTs, electrical resistivity
is 0.4-0.7 Ωm, maximum current density 107-109 Acm-2
and
the band gap of CNTs is 0-0.7 eV depending upon their
nature. Owing to exceptional electrical properties in
conjunction with high aspect ratio and low density (one sixth
time lower than steel), CNTs have gained great interest for
developing lightweight electrically conductive polymer
nanocomposite [3, 4].
CNT/epoxy polymer composites have been widely used in
advance military aircrafts, fairings, missiles, personal
communications and electronic packaging in which CNTs act
as nano-fillers in epoxy matrix [5].High electrical conductivity
and nanometer size of CNTs imply very low level of
percolation threshold in CNT/epoxy nanocomposite [6].
Exceptional electrical properties of CNTs directly related to
high aspect ratio and defect free structure of CNTs. Advance
applications of epoxy polymer composite demands high
electrical conductivity, electromagnetic interference (EMI)
shielding and electrostatic dissipation (ESD) characteristics.
Electromagnetic waves produced from some electronic
component decrease the performance of nearby electronic
system by introducing noise, disturbance, data loss, etc.
Hence, there is an intense need to shield the sensitive
electronic circuits form undesired signals through EMI
shielding material. One basic criteria of EMI shielding
material is that it should be electrically conductive.
CNT/epoxy composite is found to be promising candidate in
this regard [4, 7-12].
However, one key issue of CNT based polymer
nanocomposite is the dispersion of CNTs in polymer matrix.
Dispersion includes separation of CNTs bundles and then
stabilization of CNTs in polymer matrix. As CNTs
agglomerates into bundles due to high surface area and these
bundles exhibits inferior properties as compare to individual
CNT. Another key challenge is the good interfacial bonding
between CNTs and polymer matrix. As CNTs are inert and
atomically smooth material so there is a lack of bonding
between polymer matrix and CNTs [13-15]. Best route to
overcome stated problems is to attach functional groups
through surface modification treatments of the CNTs. There
are numerous approaches to functionalize CNTs depending
upon their end use and the choice of functional group. Among
192
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-201-320.jpg)
![them, modification of CNTs through aryl dizonium salt is
more versatile and easy with variety of functional groups.
However, fucntionalization process induces defects and
decreases the aspect ratio of CNTs thereby degrading the
electrical properties like conductivity, EMI shielding and ESD
of CNT reinforced polymer composite [13, 16-18].
Most of the literature discusses the effect of acidic
functionalization on the electrical properties of CNT based
polymer composite. However, the effect of functionalization
treatment through aryl diazonium salt on electrical properties
of MWCNTs and MWCNT/CF reinforced composite has been
seldom reported. Therefore a detailed study has been
conducted in which MWCNTs were functionalized through
aryl diazonium salt to attach covalent functional group. Effect
of applied electric field on the electrical behavior of f-
MWCNT and p-MWCNTs in electrolyte has been in-situ
observed under light microscope and discussed here in detail.
Moreover, electrical conductivity and EMI shielding
effectiveness of f-MWCNTs/CF reinforced polymer
composite and p-MWCNTs/CF reinforced polymer composite
was also evaluated and compared.
II. MATERIALS AND METHOD
A. Materials
MWCNTs used in this study were purchased from Cheap
Tube Inc, US and were synthesized through CCVD (Catalytic
Chemical Vapor Deposition). Specification of MWCNTs
given by manufacturer: outer diameter 10-20 nm, length 10-30
µm, purity level > 95%, ash content <1.5%, specific surface
area 180-230 m2
/g and bulk density 0.22 g/cm3
. Specifications
details of carbon woven fabric which was used as
reinforcement in epoxy matrix are: areal density 325±5,
thickness 0.5±0.005 mm, breaking strength > 270Kg/5cm in
wrap direction, weave style satin 3, 3K tow and equal no of
threads in weft and warp direction. Thermoset epoxy resin
(Bisphenol-A) containing a reactive diluent was chosen as
matrix because of its long pot life, low viscosity and
multifunctional properties. Cycloaliphatic amine was selected
as hardener and used in the ratio of 10:3.5 by weight with
epoxy resin.
B. Functionalization of MWCNTs
Functionalization of MWCNTs was carried out through a
facile method in which 5-10 mg MWCNTs were dispersed in
acetronitrile through bath sonication. After 30 minutes of
sonication, 0.15-5 equivalent of diazonium salt (BF4N2-C6H4-
NO2) was added and stirring of mixture was carried out for 5
days at room temperature. Then whole mixture was filtered
through 0.45 micron size PTFE filter paper and washed
several time with acetronitrile to remove unreacted dizonum
salt. Functionalized MWCNTs were dried at room temperature
under vacuum for 24 h before use [19]. Functionalization of
MWCNTs was visually confirmed by dispersing 5 mg of
pristine and functionalized MWCNTs in water in different
tubes and kept for 24 h. After 24 h pristine MWCNTs were
settle down at the bottom of tube whereas the functionalized
Fig. 1. Schematic illustration of MWCNTs fucntionalization with 4-
Nitrobenzenediazonium tetrafluoroborate.
Fig. 2. Suspension of F (f-MWCNTs) and P (p-MWCNTs) in water after 24 h.
MWCNTs were remained suspended which confirmed the
functionlization of MWCNTs as shown in Fig. 2 whereas Fig.
1 demonstrates fucntionalization mechanism of MWCNTs.
C. Fabrication of MWCNTs/CF polymer composite
Functionalized MWCNT (0.2 wt% of matrix) was added to
ethanol and sonicated in sonication bath for 30 min to open up
MWCNTs bundles. Then epoxy resin was added to the f-
MWCNTs contained ethanol and further sonicated for 30 min
at 80⁰C to evaporate ethanol and disperse MWCNTs in epoxy
resin. Upon cooling of MWCNT filled epoxy resin, hardener
was added and solution was magnetically stirred for 10 min to
thoroughly mix the hardener in the solution. Finally,
193
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-202-320.jpg)
![Fig. 5. Picture of (a) Electrolytic cell (b) In-situ observation of MWCNTs
under electric field through light microscope.
examined after curing under the microscope to witness the
effect of curing cycle on dispersion state. Similar process was
carried out with p-MWCNTs containing epoxy resin.
D. DC electrical conductivity measurements
DC electrical conductivity of fabricated samples (both f-
MWCNTs/CF and p-MWCNTs/CF) was measured at room
temperature in accordance with ASTM standard D4496-98
and compared with the results of neat CF reinforced polymer
composite. Standard four point probe method was used to
determine the surface and volume conductivity via four Kelvin
pins and LCR meter.
E. EMI shielding effectiveness (EMI SE) test
EMI SE test was carried out in Anechoic chamber using
calibrated Vector Network Analyzer (Rohde & Schwarz) over
the frequency range of 7-14 GHz. Attenuation was measured
by recording S21 parameter (return loss) between source and
receiving antenna inside the anechoic chamber without any
sample. Then test sample was placed between the antennas
(source & receiving) and attenuation was again measured by
noting down the value of S21 parameter. The difference
between S21 values of sample and without sample gives the
value of attenuation (i.e. EMI SE) caused by the sample.
Attenuation is described in terms of shielding effectiveness
(SE), measured in decibel (db) and shows the reflective nature
of the sample. Testing was performed on equally dimensional
samples (150 x 75 x 2 mm) of f-MWCNTS/CF reinforce
polymer composite and p-MWCNTs/CF reinforced polymer
composite.
IV. RESULTS AND DISCUSSION
A. FE-SEM
Figure 6a and 6b shows the high resolution SEM images of
f-MWCNTs and p-MWCNTs respectively. Opening of end
caps and shortening of length was evident in case of f-
MWCNTs as shown in Fig. 6(a) while Fig. 6(b) demonstrates
that p-MWCNTs have long length and damage-free structure.
High aspect ratio and defect free structure of nanotubes is
prime requirement for exhibiting good electrical conductivity.
Morphological study through FE-SEM confirms that the aryl
diazonium salt functionlization has destroyed the physical
structure and decreased the aspect ratio of nanotubes, which in
turn decreased the electrical conductivity of MWCNTs.
B. In-situ observation of MWCNTs under applied electric
field
When DC electric field was applied up to 80 V no
significant effect was observed on MWCNTs under
microscope. However at 100 V fraction of p-MEWCNTs
moved from cathode to anode and network of nanotubes is
formed between electrodes. Under applied electric field
MWCNTs experience force due to polarization effect and
move from cathode to anode get discharged and absorbed on
the anode. Then adsorbed MWCNTs become the source of
adsorption of further nanotubes because of high electric field
at the tip which is connected to anode. In this way, conductive
pathway of p-MWCNTs is formed from anode to cathode
showing the high conducting nature of p-MWCNTs as shown
in Fig. 7 (a). In case of f-MWCNTs most of the nanotubes
under applied electric field attached to the anode without
development of conductive pathway because of imbalance
polarization effect due to structural damage during
fucntionalization process. Similar effect was also observed
with acid functionlized MWCNTs under applied electric field
by other researchers [20]. Transfer of electric current is not
possible between the electrodes without formation of
conductive path suggesting the poor electrical conductivity of
f-MWCNTs. Arrows head show the absence of conductive
pathway in Fig. 7 (b).
C. Optical microscopy
Figure 8 (a, b) shows the optical images of 0.2 wt % f-
MWCNTs and p-MWCNTs filled epoxy matrix in uncured
and cured state. Functionalized MWCNTs were more
uniformly distributed in the matrix and less amount of sub-
micron size aggregates were present as compared to p-
MWCNTs-epoxy matrix in uncured state. f-MWCNTs were
dispersed homogenously in matrix during sonication process
due to the bonding between epoxy and functional group
grafted on MWCNTs. However, in case of p-MWCNTs
homogenous distribution was difficult in matrix because of
195
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-204-320.jpg)
![Fig. 8. Optical images showing the dispersion of (a) f-MWCNTs and (b) p-
MWCNTs in uncured and cured state of polymer matrix.
Fig. 9. Volume and surface dc electrical conductivity of composite samples.
frequency and dependent on the type of MWCNTs addition.
Absorption and multiple reflections play an important role in
shielding against EM waves but dominating mechanism at
lower frequency is reflection which is directly related to
electrical conductivity of the sample [21]. Layered structure of
composite is responsible for the lost of EM waves at the
interface between carbon fabric and the matrix through
multiple reflection phenomena. And some part of EM waves
was absorbed by MWCNTs but mainly EM waves were
attenuated by reflection mechanism. As shown in Fig. 10 that
the p-MWCNTs composite demonstrated overall better EMI
SE which is governed by high electrical conductivity of p-
MWCNTs. For reflection of EM waves, 3D conductive
pathway is required which is better provided by p-MWCNTs
than f-MWCNTs. Maximum SE was observed by p-MWCNTs
sample at frequency range of 7-9 GHz with a peak about 70 db
at a 7.75 GHz frequency. However, in the frequency range of
10-11.5 GHz f-MWCNTs composite performed similar way as
p-MWCNTs sample and maximum SE about 46 db was
observed at 11.4 GHz in case of f-MWCNTs sample.
Minimum attenuation of about 14.8 db was observed at 13.7
GHz by p-MWCNTs sample and in case of f-MWCNTs
minimum attenuation of 9.4 db was attained at the frequency
of 13.5 GHz. From the test results, it can be inferred that p-
MWCNTs composite performed better than f-MWCNTs in
shielding EM waves in the whole tested frequency range with
the maxima at around 70 db.
V. CONCLUSION
Effect of functionalization by aryl diazonium salt on
electrical properties of MWCNTs was assessed through
applied electric field test in electrolytic cell, electrical
conductivity measurement and EMI SE test. Following
conclusions were drawn on the basis of test results:
SEM microscopy showed that the functionalization
process has damaged the physical structure of nanotubes.
It was found from the optical microscopy that the
dispersion of f-MWCNTs in the host matrix was
improved.
Amount of aggregates of MWCNTs increased upon
curing in both cases (p-MWCNTs & f-MWCNTs) but f-
MWCNTs showed more resistance to agglomerate
formation during curing stage.
It was deduced that high aspect ratio and defect-free
structure of MWCNTs is essential for formation of
conductive pathway under applied electric field. Pristine
MWCNTs aligned themselves in a regular fashion under
applied electric field whereas the f-MWCNTs do not
respond appreciably to the applied field.
DC electrical conductivity measurements showed that
the conductivity of f-MWCNTs was decreased as a result
of functionalization. The reason for deterioration in
electrical conductivity was attributed to the introduction
of defects in the MWCNT sidewalls and shortening of
length.
Pristine MWCNTs based composite sample outclassed
the f-MWCNTs composite in EMI SE testing over the
tested frequency range of 7-14 GHz.
3D conductive pathway formation of MWCNTs inside
the matrix is the prime requirement for exhibiting good
electrical properties and this in turn drastically affected
by the choice of functionalization procedure.
197
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-206-320.jpg)
![Fig. 10. Comparison of shielding effectiveness of f-MWCNTs/CF and p-MWCNTs/CF reinforced polymer composite.
ACKNOWLEDGMENT
Authors would like to thank Mr. Ahmed Bilal (Chairman,
SUPARCO) for approval and provision of facilities. The
authors also gratefully acknowledge Mr. Wasim Nawaz for
performing EMI testing and Ms Noureen Owais for
providing assistance in electrical conductivity measurements
of the samples.
REFERENCES
[1] Z.L. Wang, W. H. Manu and Dechang Li, “Structure and growth of CNT
films by Pyrolysis,” Cheml Phys Lett, V. 316(5,6) , pp-349, 2000.
[2] Ajayan PM, Zhou OZ, “Applications of carbon nanotubes. Carbon
Nanotubes,”. pp.391-425, 200.
[3] R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, C. Xu, Y. Hee Lee, S. G.
Ki, A. G. Rinzler, D. T. Colbert, G. E. Scuseria, D. Tomanek, and R, E.
Smalley, Thesis, “Crystalline Ropes of Mettaic Carbon Nanotubes,”
Science Vol. 273. No. 5274, pp. 483-487, 1996.
[4] Prabhakar R, Bandaru “Electrical Properties and Applications of Carbon
Nanotube Structures,” Reviews journal of Nanoscience and
Nanotechnology Vol.7, pp.1–29, 2007.
[5] Christopher Kingston, Richard Zepp, Anthony Andrady, et al. “Review-
Release characteristics of selected carbon nanotube polymer
composites,”CARBON Vol. 68 pp.33–57, 2014.
[6] Sandler JKW, Kirk JE, Kinloch IA, Shaffer MSP, Windle AH, “Ultra-
lowelectrical percolation threshold in carbon-nanotube-epoxy
composites.” Polymer, 2003, Vol. 44, pp.5893-9, 2003.
[7] Baughman R. H, Zakhidov A. A, de Heer, W. A, “Carbon nanotubes-the
route toward applications,” Science, Vol. 297, pp. 787-792, 2002. ISSN
0036-8075
[8] Bauhofer, W., & Kovacs, J. Z, “A review and analysis of electrical
percolation in carbon nanotube polymer composites,” Compos. Sci.
Technol., Vol. 69, pp. 1486-1498,ISSN 0266- 3538
[9] Singh BP, Choudhary V, Saini P, Mathur RB, “Designing of epoxy
composites reinforced with carbon nanotubes grown carbon fiber fabric
for improved electromagnetic interference shielding,” AIP Adv 2012;2:6
[10] Mathur RB, Singh B.P, Tiwari P.K, Gupta T.K, Choudhary V “
Enhancement in the thermomechanical properties of carbon fibre-carbon
nanotubes-epoxy hybrid composites,” International Journal of
Nanotechnology, Vol. 9, pp.1040-1049, 2012.
[11] Huang Y, Li N, Ma Y, Feng D, Li F, He X, et al, “The influence of
single-walled carbon nanotube structure on the electromagnetic
interference shielding efficiency of its epoxy composites,”
Carbon,Vol.45, pp.1614-21, 2007.
[12] Li N, Huang Y, Du F, He XB, Lin X, Gao HJ, et al. “Electromagnetic
interference (EMI) shielding of single- walled carbon nanotube epoxy
composites,” Nano Letters, Vol.6, pp.1141-5, 2006.
[13] Kannan Balasubramanian and Marko Burghard, “Chemically
Functionalized Carbon Nanotubes reviews,” Small, Vol. 1, No. 2, pp.
180–92, 2005.
[14] A.KANAPITSAS, C.TSONOS, “Study of electrical / dielectric and
thermomechanical properties of polymer–carbon nanotubes
nanocomposites,” In Proceedings of the 1st WSEAS International
Conference on NANOTECHNOLOGY.
[15] R.B. Mathu r, Shaila ja Pande, B.P. Sing h, T.L. Dhami, “Electrical and
Mechanical Properties of Multi-Walled Carbon Nanotubes Reinforced
PMMA and PS Composites,” POLYMER COMPOSITES-2008
[16] C.VELASCO-SANTOS, A.L.MARTINEZ-HERNANDEZ and V. M.
CASTANO, “Review Carbon nanotube-polymernanocomposites: The
role of interfaces Composite Interfaces,” Vol. 11, No. 8-9, pp. 567–586,
2005.
[17] Najiba Abdullah Al-Hamdani, “Preparation and Electrical Properties of
Epoxy Resin Reinforced with Functionalized Carbon Nanotubes,” IOSR
Journal of Applied Physics (IOSR-JAP) Volume 6, Issue 4 Ver. I, pp.
54-56, Jul-Aug. 2014
[18] Jeffrey L. Bahr, Jiping Yang, Dmitry V. Kosynkin, Michael J.
Bronikowski, Richard E. Smalley, and James M. Tour, “
Functionalization of Carbon Nanotubes by Electrochemical Reduction of
198
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-207-320.jpg)
![Aryl Diazonium Salts: A Bucky Pap,” J. Am. Chem. Soc,Vol. 123,
6536-6542, 2001.
[19] Anastasia A. Golosova, Christine M. Papadakis and Rainer Jordan
“Chemical functionalization of carbon nanotubes with aryl diazonium
salts” Mater. Res. Soc. Symp. Proc. Vol. 1362 © 2011 Materials
Research Society DOI: 10.1557/opl.2011.1141
[20] Abu Bakar Sulong, Nurhamidi Muhamad, Jaafar Sahari, Rizauddin
Ramli, Baba Md Deros, Joohyuk Park “Electrical Conductivity
Behaviour of Chemical Functionalized MWCNTs Epoxy
Nanocomposites” European Journal of Scientific Research ISSN 1450-
216X Vol.29 No.1, pp.13-21, 2009.
[21] Simon Rea,David Linton, Eddie Orr, Jonathan
McConnell“Electromagnetic Shielding Properties of Carbon Fibre
Composites in Avionic Systems,”Microwave Review, June, 2005
199
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-208-320.jpg)
![Figure 5 Simulated Reflection coefficient (S11) plot
The simulated reflection coefficient is less than -40 dB at
desired frequency while the impedance bandwidth is greater
than 100 MHz. Figure 5 shows the reflection coefficient plot.
IV. MEASUREMENT RESULTS
The proposed antenna is fabricated on wire cut machine in
order to have the slots fabricated with precision along with
maintaining very strict inter element spacing. The fabricated
antenna is shown in figure 6. The measured reflection
coefficient plot is shown in figure 7. Measured reflection
coefficient at desired frequency is -43 dB while impedance
bandwidth is 145 MHz. Antenna radiation pattern measurement
was carried out in Anechoic Chamber Test facility. Measured
azimuth and elevation cuts are shown in figure 8 and 9 the
measured gain at our desired frequency is about 20 dB.
Figure 6 Fabricated slotted waveguide antenna array
Figure 7 Measured Reflection coefficient
Figure 8 Measured Azimuth cut
Figure 9 Measured Elevation Cut
V. CONCLUSION
A resonant C band slotted waveguide antenna is designed,
fabricated and tested as per design requirement. Gain of the
antenna is approximately 20 dB with -43 dB reflection
coefficient value. Impedance bandwidth achieved is 145 MHz
for VSWR < 2.
VI. ACKNOWLEDGMENT
First I would like to thank Almighty for all His blessings
and providing me the strength for the successful completion of
this task. I would also like to thank SUPARCO for providing
me the opportunity and necessary support and guidance from
the team members for accomplishing this task.
VII. REFERENCES
[1] Antenna engineering Handbook Richard Clayton Johnson.
[2] Waveguide slot antenna array by Ronald A Gilbert BAE systems.
[3] W1 GHZ Microwave Antenna Book Paul Wade W1GHZ.
[4] Resonant length calculation and radiation pattern synthesis of
longitudinal slot antenna in rectangular waveguide By M.
Mondal, A. Chakrabarty Progress In Electromagnetics Research
Letters, Vol. 3, 187–195, 2008.
[5] A Design of Slotted Waveguide Antenna Array Operated at X
band Kuo-Lun Hung, His Tseng Chou. Wireless Information
Technology and Systems (ICWITS), 2010 IEEE International
Conference.
202
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-211-320.jpg)
![For Cities like Lahore having major trading points may have
risk of more RTA. To understand this problem above map
illustrate the RTA comparison with commercial area. The map
clearly shows that the majority of accident happened near
commercial areas. Another picture of same issue reflect the
comparison of RTA density with commercial points (Figure XI).
This map demonstrates that red zones for accidents are nearby
commercial areas.
V. CONCLUSION AND RECOMMENDATIONS
This work gives an insight of the present scenario of the
traffic condition in Lahore and shows out the most accident
prone roads. The result shows that total road traffic accidents,
dense population areas, commercial areas are the important
variables to take into the consideration in examining traffic
accidents. Spatial data can be shared with transportation,
emergency services and other government organizations. This
spatial database can be helpful for expert system to make new
traffic plans and rules. It also provides accurate
recommendations to vehicle drivers, the police and to local
monitoring authorities. In order to curtail road traffic accident
on Lahore roads the following recommendations are necessary:
Drivers and bikers should be trained as the mean of effective
dealing with road accidents.
Traffic rules must be deployed by the law enforcement
agencies in the district and country.
Road safety education and seminars should be held by the
transport departments.
Drivers should drive within speed limit and with a speed
unswerving with road condition.
There must have proper parking plazas in all major
commercial zones and markets.
VI. ACKNOWLEDGMENT
This study is sponsored by The Urban Unit (Urban Sector
Planning and Management Service Unit USPMSU). The authors
would like to thanks Mr Niazm-ud-din (GIS Manager at Urban
Unit) for providing data and Mr. Ehsan Saqib (Sr. GIS
Development Specialist), Dr. Ather Ashraf (Sr. GIS Specialist)
for helping in different analysis.
VII. REFERENCES
[1] Bureau of Statistic Bos, “Punjab Development Statistics
2013”,www.bos.gop.pk.
[2] Deepthi Jayan.K, B. Ganeshkumar,”Identification of Accident Hotspot: A
GIS Based Implementation for Kannur District, Kerala” , vol. 1. ISSN
0976-4380.
[3] Dr, A.J.Aderamo, “Spatial Pattern of Road Traffic Accident Casualties in
Nigeria,”, vol. III, ISSN 2039-2117.
[4] Harnam Singh,A.D Aggarwal, “Fatal Road Traffic Accident among
Young Children,” J Indian Acad Forensic Med, 32(4), ISSN 0971-0973.
[5] The Urban Unit, USPMSU Lahore, www.urbanunit.gov.pk
FIGURE IX- RTA DENSITY COMPARISON WRT COMMERCIAL AREA
207
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-216-320.jpg)
![ANALYSIS OF RECENT DROUGHT BASED ON
NDVI AND METEOROLOGICAL DATA
A case study of Tharparkar district, Sindh
Muhammad Arslan1
, Rao Zahid2
, Badar Ghauri3
Department of Remote Sensing and GISc
Institute of Space Technology
Karachi, Pakistan
marslanhafeez100@gmail.com
Abstract— Thar region of Pakistan has often undergone short
and long period of drought. In this study, Satellite Remote
Sensing and meteorological data were used to detect the drought.
Agricultural drought was identified based on the pre and post
monsoon Normalized Difference Vegetation Index (NDVI) with
250meter spatial resolution from MODIS terra satellite from
2008 to 2014. Departure from the mean values of pre and post
monsoon NDVI was categorized to determine the agricultural
drought. Whereas meteorological drought was identified based
on the annual precipitation data for the same period provided by
the Pakistan Meteorological Department (PMD). Pre and post
monsoon mean NDVI for the year 2014 was then compared with
the previous years. Pre monsoon mean NDVI values of 2008 to
2013 was 0.115956, while post monsoon mean NDVI values of
2008 to 2013 were around 0.221801. On the other hand, pre and
post mean NDVI for the year 2014 were observed as 0.127617
and 0.156314 respectively. Mean post-monsoon NDVI value
drastically dropped in the year 2014. Satellite based mean NDVI,
PMD annual rainfall data and field observation showed that a
short term drought occurred in year 2014. This was also
confirmed by the local population during the field visit.
Keywords— Drought, MODIS, NDVI, precipitation.
I. INTRODUCTION
Drought is classified into four classes. Hydrological
drought, meteorological drought, agricultural drought and
socio-economic drought. No universal definition of drought is
still known. Drought is defined as a period of abnormally dry
weather which eventually affects in a change of vegetation of
an area [1]. Drought is also defined as it usually occurs when
the rainfall or runoff is found below a mean truncation level,
derived from the long term rainfall series [2]. For human
society and ecosystem drought is a significant adverse climatic
events. Since 1850, the mean global surface air temperature
has increased by 0.76o
C [3], by the end of the twenty first
century, it is expected that the temperature will increase by 1.5
o
C – 6.4 o
C [4]. Drought may increase under the warm climate
[5]. The frequency and the intensity of drought have increased
in past decades in many regions of the world [6, 7]. This
recurring of drought has negative impact in these areas in term
of annual loss of vegetation and food. Drought monitoring has
gained importance in many countries especially developed
countries due to the serious social, environmental and
economic ramifications [8]. Conventionally, drought
monitoring was based on meteorological station observations
that obviously does not have a continuous spatial coverage for
the monitoring of comprehensive drought. Most of the studies
have used satellite data to monitor hazards including, drought,
flood, earth quake and so on [9, 10]. Satellite data have
frequently been used to monitor and detect the drought
phenomenon [1]. Satellite data provides earth observation for
monitoring the impacts of drought in a precise and effective
way. There exists a time lag between precipitation events and
its response on the vegetation. The time gap between the
occurrence of precipitation and the time precipitation water
reaches the plant roots. It varies from 01 to 12 weeks
depending on vegetation type [11]. There are various studies
that have focused on NDVI and precipitation relationship.
This relationship varies spatially, when the NDVI values
changes with respect to precipitation [12]. Spatial distribution
and natural condition of vegetation are the function of
environmental variables such as precipitation. Strong
relationship is occurred between NDVI and rainfall [11].
Precise and consistent vegetation mapping is essential and
necessary for the management of water resources and
mitigation of drought events.
A. Objectives
to identify the NDVI and meteorological based drought.
to compare 2014 year recent drought with the previous
years
to evaluate the effects of drought in term of socio-
economic condition of the local people
II. MATERIALS AND METHODS
A. Study Area
Tharparkar district of Sindh province consists of deserted
land and its area is 19,638 square kilometers. Administratively
this district is divided into four talukas. These are Diplo,
Nagarparkar, Mithi and Chachro shown in Fig1. It has
208
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-217-320.jpg)
![Figure 5. Annual rainfall provided by PMD (2009-2014)
Figure 6. Means of food for the livestock
Figure 7. Reduction in agriculture
Figure 8. (a and b) Livestock of the area, (c) How local people collect water
(d) Interview from the local people related to drought (e and f) Showing
local people in the picture
IV. CONCLUSION
Drought is a recurring phenomenon and it can be examined
by satellites. Long term satellite data could help the severity
and onset of drought due to its spatial and temporal variability.
Effective and efficient drought mitigation could be achieved
through a well-defined drought monitoring system.
ACKNOWLEDGMENT
This research is part of a project named “Development of
Decision Support System for Drought Monitoring in Sindh”
funded by the Pakistan Higher Education Commission (HEC).
The authors acknowledged the financial grant of HEC.
Thanks are also due to Pakistan Meteorological Department
(PMD) for provision of data. We also thank NASA site for
giving data.
REFERENCES
[1] Kogan, Felix N. "Global drought watch from space." Bulletin of the
American Meteorological Society 78.4 (1997): 621-636.
[2] Dracup, John A., Kil Seong Lee, and Edwin G. Paulson. "On the
statistical characteristics of drought events." Water resources research
16.2 (1980): 289-296.
[3] Jones, P. D., Trenberth, K., Ambenje, P., Bojariu, R., Easterling, D.,
Klein, T., ... & Zhai, P. (2007). Observations: surface and atmospheric
climate change. In Climate Change 2007: The Physical Science Basis.
Contribution of Working Group I to the Fourth Assessment Report of
the Intergovernmental Panel on Climate Change, 235-336.
[4] Solomon, Susan, ed. Climate change 2007-the physical science basis:
Working group I contribution to the fourth assessment report of the
IPCC. Vol. 4. Cambridge University Press, 2007.
[5] Pachauri, Rajendra K., and A. Reisinger. "Synthesis Report." IPCC.
Ginebra (2007).
[6] Hulme, Mike, and Mick Kelly. "Exploring the links between
desertification and climate change." Environment: Science and Policy
for Sustainable Development 35.6 (1993): 4-45.
[7] McCarthy, James J. Climate change 2001: impacts, adaptation, and
vulnerability: contribution of Working Group II to the third assessment
report of the Intergovernmental Panel on Climate Change. Cambridge
University Press, 2001.
[8] M Amin, Owrangi, et al. "Drought monitoring methodology based on
AVHRR images and SPOT vegetation maps." Journal of water resource
and protection 2011 (2011).
[9] Anderson, James Richard. A land use and land cover classification
system for use with remote sensor data. Vol. 964. US Government
Printing Office, 1976.
[10] Peters, Albert J., et al. "Drought monitoring with NDVI-based
standardized vegetation index." Photogrammetric engineering and
remote sensing 68.1 (2002): 71-75.
[11] Chopra, Parual. "Drought risk assessment using remote sensing and GIS:
a case study of Gujarat." ME thesis, Indian Institute of Remote Sensing,
Dehradun & International Institute for Geo-information & Earth
Observation, the Netherlands (2006).
[12] Nicholson, S. E., and T. J. Farrar. "The influence of soil type on the
relationships between NDVI, rainfall, and soil moisture in semiarid
Botswana. I. NDVI response to rainfall." Remote Sensing of
Environment 50.2 (1994): 107-120.
210
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-219-320.jpg)
![Assessment of Urban Growth of Karachi: From A
Tiny Town to A Meta City of the World
Ali Jan Hassan1
, Mudassar H. Arsalan2
, Hira Fatima
Institute of Space and Planetary Astrophysics
University of Karachi
Karachi, Pakistan
1
Alijan.shaaan@gmail.com
2
mharsalan@uok.edu.pk
Abstract— Karachi is the only meta city of Pakistan and capital
of Sindh. It plays a key role in the country’s economy by
contributing 65% of total revenue. However the sustainability of
this city of lights is at risk due to mismanagement and
development without proper long term planning. The increased
urbanization has brought several severe issues such as increased
demographic density, management challenges, resource deficit
and ecological pressures, etc. The sprawl of the city is
uncontrolled spreading of urban developments. Remote sensing
data and GIS are now widely being used for change detection
analysis and applications. Urban sprawl is one of them, which
can be analyzed efficiently and cost effectively by utilizing
satellite data available online in the form of present as well as
historical images in different spatial resolutions.
In this study change detection technique was used to assess the
urban sprawl through the integration of historical maps with
Landsat-5 TM sensor image of 1989 and Landsat-8 OLI images
of 2014. With the aid of these images we monitored, measured
and analyzed the urban sprawl of Karachi. Despite of many
unsustainable conditions in the city, its population growth is
tremendous i.e. approximately 4 % per annum. This growth is
the result of the high natural increase as well as substantial
migrations from other parts of the country. It is assessed that the
living population in the mid 2014 was around 21 million, which
was far greater than of 1998 i.e. 9.2 million. In the last 25 years,
Karachi has extended around 15 percent at the rate of 2 square
kilometers per year to accommodate its increasing population
with considerable high density.
Index Terms— Urban sprawl, Change detection, Sustainability,
Satellite imagery, GIS, RS.
I. INTRODUCTION
Sustainability is not a new concept, it was discussed,
analyzed and even practiced at all levels, from country to
individual. The term ‘Sustainable Urban Developed’ (SUD)
emerged from the Brundtland Report [1], according to this
report, SUD is defined as “Development that meets the needs
of today without compromising the needs of the future.” Many
organizations such as UN-Habitat, WHO, EEA, etc., started to
work on SUD indicators. Indicators for sustainable
development provide valuable information, indicating rather a
development is on the way of sustainability or degradation [2].
Planners, developers and stakeholders use the SUD indicators
to better understand the current development situation as well
as to predict the future development scenario in context of
sustainability[3] Urban Sprawl is one of the main indicators of
SUD, that can determine the increment area of Urban land, as
well as the densification of the city [4]
Urbanization and Urban Sprawl are two different
terminologies. Urbanization is a process of increasing number
of people that migrate from rural to urban areas, results in the
physical growth of urban areas regardless of developments
occur horizontally or vertically [5], whereas Urban Sprawl is
defined as the expansion of urban area in the form of new
development on isolated tracts, separated from other areas by
vacant land [6]. The socioeconomic conditions of any place are
the driving forces for urbanization hence it occurs more
tremendously in areas that are economically well established
[7]. Urbanization is an inevitable process due to economic
development and rapid population growth [8]
In the last part of the 20th century and the beginning of the
21st century human population has increased tremendously. On
average, at every second 4 or 5 children are born, while 1 or 2
people dies, this rate of birth and death means a net gain of 2.5
persons (on average) on the Earth [9]. Unfortunately, this
increase has been occurring in the developing part of the world,
where lack of facilities, poverty, pollution and crime rate are at
its peak [9].
Pakistan is in the list of developing countries and is badly
affected by terrorism, energy crises and law and order situation
in the country [11]. It is an atomic power, bearing world’s
toughest army, having a vast amount of resources of coal, oil
and other minerals [12]. But due to miss-management,
irregularities and irresponsibleness of bureaucracy,
stakeholders and political leaders, the complete potential of the
resources is not being properly utilized [13]. Pakistan is the
world’s sixth most populous country today [14] With an
estimated population of 188 million, and 2.05 percent annual
211
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-220-320.jpg)
![growth rate, Pakistan is expected to become the fifth largely
populated nation by 2050 [14].
Table 1. World’s Most Populous Cities [14]
2014 2015
Country population
(Millions)
Country population
(Millions)
China 1368 India 1628
India 1226 China 1437
USA 320 USA 420
Indonesia 255 Nigeria 299
Brazil 203 Pakistan 295
Pakistan 188 Indonesia 285
A. Study Area
Today Karachi is the largest and most populous city of
Pakistan, lies in the south bounded by the Arabian Sea. It
covers a total area of 3600 square kilometers that includes
Built-up land, Vegetation Coverage, Industrial built-up,
commercial areas and a lot of recreational areas. The Karachi
port serves as a gateway for all the import/export activity by
sea. The industrial and economic activity in Karachi support
Pakistan a 65% of revenue [15]. Karachi appeared as a mega-
city with the beginning of this century with an exponential
growth in its population.
The residents of Karachi are blessed with the sea breeze
that blows throughout the year. Because it is close to the Sea
the temperature remains normal, the mean annual temperature
is in between, the wind normally blows towards North-East.
There are many commercial centers exist in the city, which
enabled the it to develop as an important industrial and
economic center that attract the people not only from Pakistan
but all over the world.
B. Objectives
The Objective of this study was to determine the urban
footprint for the year 1989 and 2014 using remote sensing data
and GIS, to figure out the extension of Built-up land within 25
years (1989-2010).
II. MATERIALS AND METHODS
The simple methodology adopted for this study comprises
of three steps: Data collection, data processing and urban
sprawl analysis. The brief account of these steps is given
below.
A. Data Collection
This study is an application of remote sensing and GIS in
Urban sprawl analysis. As recent as well as historical remote
sensing data is freely available on the internet. Further, using
GIS on this Remote sensing data different analysis can be
done efficiently and cost effectively. For this study Landsat 5
TM sensor clod free image of Karachi (Path-Row 152-42)
captured in November 1989 and Landsat 8 OLI sensor image
captured in January 2014 were acquired from the Landsat
archives available at the USGS website.
B. Data Processing
The pre-processing was carried out for the development of
AOI. Downloaded images of the study area were
geometrically corrected and projected to Universal Transverse
Mercator (UTM) projection system, zone 32. The spheroid
and datum were referenced to WGS1984.
C. Methodological Framework for Urban Sprawl
The change detection technique was adopted to assess the
change in land use and land cover (LULC) of the study area.
Collected and processed Landsat 5 image of 1989 and Landsat
8 image of 2014 were used while the main focus was on built-
Fig.1. Population growth in developing and developed nations, 1750
– 2100 [10]
Fig. 2. Study Area- Karachi, Pakistan
212
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-221-320.jpg)
![up area change. To extract the Urban built up the more
accurately supervised maximum likelihood classification was
done.
The Maximum Likelihood Classification is one of the
supervised classification approaches for pattern recognition in
which the probability of a pixel belonging to each of a
predefined set of classes is calculated, then pixel is assigned to
the class for which the probability is the highest [16]. ML is
based on the Bayesian probability formula:
P(x, w) = P(w|x) P(x) = P(x|w) P(w)
Where x and w are called events, P(x, w) is the probability
of coexistence of events x and w, while P(x) and P(w) are the
prior probabilities of events x and w and P(w|x) is the
conditional probability of event x given event w [16].
By carefully observing the patterns in the images, it was
concluded that the classification should be done for 4 classes
i.e. Urban Land, (including dense, green as well as sparse
Urban areas), Vegetation (including dense and sparse
vegetation as the study area had a vast mangrove forest towards
its south), Water bodies and Open/barren land.
Maximum Likelihood Classification was done on natural
color composite image (Band Combination 3, 2, 1) by training
100 samples of four classes distributed randomly in the image.
Classes were identified by visual interpretation. From this a
thematic map of four classes was generated. The same
procedure was applied to the other image.
1) Accuracy Assessment
To check the level of confidence in the classification results
an error matrix is usually generated. This is some time called
the confusion matrix, in which a matrix presents the
relationship between the classes in the reference maps and
classified images. Estimating classification accuracy enables us
to determine classification performance [16], but there is a lack
of information, because it leads to overall accuracy of the
classes not of individual class. To assess the accuracy of
individual class the concept of Producer’s and User’s accuracy
was used in this paper. A multivariate index, Kappa Coefficient
was also used to assess the difference between actual
agreement and change agreement [16].
III. RESULTS
Due to the highly favorable geographical location, Karachi
has always been a trade route for the merchants of India, but it
gained its importance in the British rule. The development of
perennial irrigation schemes by the British in Punjab and Sindh
mainly accounted for the growth of Karachi along with that the
development of railways, making the quick transportation of
the agricultural material that in turn increased in agricultural
production which was exported through Karachi which made it
the largest exporter of wheat and cotton to India [17].
In 1935, when Sindh separated from Bombay, Karachi
became the capital of Sindh, Government offices and trade
organizations were shifted from Bombay to Karachi. In 1947,
Pakistan got independence and Karachi became first capital
Pakistan.
Table 2 Karachi Population Growth [18]
Year Population Relative
change
No. of year in
between two
consecutive
census
Population
change
(%)
1901 136,297 136,297 10
1911 186,771 323,068 10 37
1921 244,162 430,933 10 30.2
1931 300,779 544,941 10 23.2
1941 435,887 135,108 10 44.9
1951 1,137,667 701,780 10 161
1961 2,044,044 906,377 10 79.67
1972 3,606,746 1,562,702 11 76.5
1981 5,437,984 1,831,238 9 50.8
1998 9,802,134 4,540,422 17 86.29
2001 13,500,000 3,697,866 3 37.7
2005 15,119,000 1,619,000 4 11.9
2014* 20,228,112 5,109,112 9 25.5
From the below graph we can see that the population of
Karachi was 9.2 million in 1998, there is a tremendous increase
of around 12 million in 2014[18].
Fig. 3. Methodological Framework of Urban Sprawl
Fig. 4. The Population Trend of Karachi 1901-2014
213
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-222-320.jpg)
![There was a significant increase of 161% from 1947-1951,
because of the migration of 600,000 refugees from India. The
settlement of those refugees was a challenge for newly built
local government. The refugees started to live on the periphery
and within the city itself, as a result the environment of Karachi
totally changed not only demographically but also ethnically
[17]. The irresponsibilities of higher authorities and lack of
planning and strategy the refugees unwillingly started to live in
squatter settlements and katchi abadis. After 1958, five master
plans have been developed for the development of the city, but
the implementation of those plans has always been an issue
because of political instability within the country.
The Karachi City, administratively is spread over an area of
approximately 3,600 sq. km. More than half of this area
(approximately 2455.5 acres) consists of vacant land. The land-
cover features such as water and vegetation, changes
throughout the year because it depends on the Climatic
conditions at a particular time. The Urban Built-up area has
extended from 343 square km in 1989 to 396 square km in
2014 that is a valuable change of 53 km in 25 years. This
reason behind this rapid increase in population within the city
is the migration from Interior Sindh, Punjab and Khyber
Pakhtoonkhuwa to find better opportunities in the mega city.
There is a 100 square km decrease in open land from which 50
sq. km land is converted into built-up land
The actual built-up land is 396 square km but as per
document figures by the City District Government of Karachi
(CDGK) [19], the built-up land was 1300 square km, it was a
combined built-up and developed open spaces. Developed
open spaces are those housing societies which were initiated in
1970s, 80s, and 90s, however, even after their infrastructure
developed they are vacant and hardly 5% occupied as shown in
table 3.
Table 3 Occupancy Status of New Housing Scheme, [19]
Serial. No. Name of
Scheme
Year of
Notification
Current
Occupancy
Status
(%)
1 Scheme no. 25-
A
1980 5%
2 Scheme no. 33 1971 20%
3 Scheme no. 42 1983 5%
4 Scheme – 43 1986 0%
5 Scheme – 45 1986 5%
From the table below, we can see that there is a 15 %
increase in built-up land and 3% decrease in open land. This
growth is mainly in built-up area and decrease in open land.
Table 4 Area Cover by each class for the years 1989 and 2014 in Sq. km
Serial
No.
Class Area
(1989)
Sq. km
Area
(2014)
Sq. km
Change
Sq. km
Change
(%)
1 Built-up 343 396 53 15
2 Open Land 2556 2455 -100 -3
3 Vegetation 748 849 -100 13
4 Water 1014 972 -41 -4
Table 5 Error Matrix of Land Cover (Karachi - 1989)
Class Built-
up
Open
Land
Veget
ation
Wat
er
Row
Sum
User's
Accura
cy
Prod
ucer's
Accu
racy
Built-
up
14 0 1 0 15 93.33 93.33
Open
Land
1 45 9 0 55 81.82 100
Veget
ation
0 0 10 0 10 100 45.45
Water 0 0 2 18 20 90 100
Colu
mn
Sum
15 45 22 18 100
Overall Accuracy 87% Kappa 0.328
Fig. 8. Land Cover Change of Karachi (1989-2014)
Fig. 9. Land Cover change % of Karachi (1989 - 2014)
214
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-223-320.jpg)
![Table 6 Error Matrix of Land Cover (Karachi - 1989)
Class Built-
up
Open
Land
Veget
ation
Wate
r
Row
Sum
User's
Accurac
y
Produ
cer's
Accur
acy
Built-
up
8 4 0 0 12 66.67 88.89
Open
Land
1 44 5 2 52 84.62 88
Vegeta
tion
0 0 9 0 9 100 64.29
Water 0 2 0 25 27 92.59 92.59
Colum
n Sum
9 50 14 27 100
Overall Accuracy 86% Kappa 0.3563
IV. DISCUSSION AND CONCLUSION
This study reveals that according to CDGK (2007) the
population of the city reaches to around 21 million and if we
consider the whole of Karachi area that is 3,600 km, the
population density is 6000 persons per square kilometer about
25% of its population has increased in last 10 years. However,
the Urban built-up area has extended from 343 square km in
1989 to 396 square km in 2014 that is a change of 53 square
km in 25 years. It shows the average annual growth rate is
slightly above 2 square kilometers per year. If we consider
average population density is 50,000 per square kilometer, the
city can accommodate only 0.1 million people every year.
However, it is not so, as the average population increase is 5
times higher than its accommodating capacity. It indicates the
pressure of population is more towards the existing
neighborhood in the form of densification and slum area's
growth. That results in depletion of resources, degradation in
environmental conditions and lack of recreational areas within
the city center. All these factors produce bad impacts on
social, environmental and economic conditions of the city and
makes difficult to achieve sustainable development.
This study was conducted to assess the conditions of Urban
Sprawl in Karachi that is an effective indicator of SUD,
however many other indicators can also be identified and
assessed, by implementing the techniques of RS and GIS.
Fig. 10. Growth of Karachi 1938-1992 [20]
215
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-224-320.jpg)
![REFERENCES
[1] World Commission, "Report of the World Commission on
Environment and Development: Our Common Future Acronyms
and Note on Terminology Chairman’s Foreword," ed: Oxford
University Press, 1987.
[2] M. Javadian, H. Shamskooshki, and M. Momeni, "Application
of sustainable urban development in environmental suitability
analysis of educational land use by using AHP and GIS in
Tehran," Procedia Engineering, vol. 21, pp. 72-80, 2011.
[3] N. Chrysoulakis, C. Feigenwinter, D. Triantakonstantis, I.
Penyevskiy, A. Tal, E. Parlow, et al., "A Conceptual List of
Indicators for Urban Planning and Management Based on Earth
Observation," ISPRS International Journal of Geo-Information,
vol. 3, pp. 980-1002, 2014.
[4] Sustainable Cities International, Indicators for sustainability-
How cities are monitoring Sustainability, 2012.
[5] UN, "United Nations Conference on Environment &
Development, Rio de Janerio, Brazil, 3 to 14 June 1992," 1992.
[6] J. I. IGBOKWE, I. C. EZEOMEDO, and E. J, "Investigation of
Urban Sprawl Using Remote Sensing and GIS: A case of
Onitsha and its Environments in Southeast, Nigeria,"
International Journal of Remote Sensing & Geoscience (IJRSG),
vol. 2, 2013.
[7] Boundless. (2015, July). The Process of Urbanization.
Available:
https://www.boundless.com/sociology/textbooks/boundless-
sociology-textbook/population-and-urbanization-
17/urbanization-and-the-development-of-cities-123/the-process-
of-urbanization-695-3433/
[8] S. Deep and A. Saklani, "Urban sprawl modeling using cellular
automata," The Egyptian Journal of Remote Sensing and Space
Science, vol. 17, pp. 179-187, 2014.
[9] W. P. Cunningham and M. A. Cunningham, Principles of
environmental science: inquiry & applications: McGraw-Hill,
2011.
[10] D. B. Botkin, E. A. Keller, and D. B. Rosenthal, Environmental
science: Wiley, 2012.
[11] I. Rana, "Foreigners driven away by terrorism, energy crisis," in
The Express Tribune, ed: @etribune, 2011.
[12] H. Naz, Environment of Pakistan, 2006.
[13] I. A. Sodhar. (2011, July). Pakistan Rich in Natural Resources
But Poor in their Management. Available:
http://jworldtimes.com/Article/92011_Pakistan_Rich_in_Natura
l_Resources_But_Poor_in_their_Management
[14] U. S. C. Bureau. (2011, July). International Programs, Country
Rank. Available:
https://www.census.gov/population/international/data/countryra
nk/rank.php
[15] A. Hassan. (2015, July). Ishaq Dar’s budget defies economic
vision: Karachi Chamber of Commerce and Industry. Available:
http://asianetpakistan.com/business-news/general-business-
news/203382/ishaq-dars-budget-defies-economic-vision-
karachi-chamber-of-commerce-and-industry-3/
[16] P. Mather and B. Tso, Classification methods for remotely
sensed data: CRC press, 2009.
[17] A. Hasan and M. Mohib, "The case of Karachi, Pakistan," 2003.
[18] Government of Pakistan, "Population Census Reports."
[19] CDGK, "Karachi Strategic Development Plan 2020," 2007
[20] M. S. Faruqui, Karachi, Physical Situation of Human
Settlements: Master Plan and Environmental Control
Department, Karachi Development Authority, 1982.
217
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-226-320.jpg)
![Identification of Post Disaster Scenario Using
Double Threshold Energy Detection
Shakeel Ahmed Waqas*
, Fahad Shamshad
Electrical Engineering Department, Military College
of Signals, National University of Science and Technology,
Islamabad, Pakistan
*
Shakeel_aw@yahoo.com
Nisar ali
Department of Telecommunication Engineering,
University of Engineering and Technology,
Peshawar, Pakistan
Abstract—The disaster can be naturally or manmade.
Telecommunication cellular system can be completely
destroyed by this disaster. The main priority is to provide
connection to disaster areas before the rescue team are
being arrived and save as much lives as possible. In this
paper, we presented a new algorithm which enable mobile
station to automatic detect the disaster scenario and switch
to emergency mode. We used double threshold energy
detection to reduce the false alarm. The performance of
proposed algorithm is evaluated using Monte Carlo
Simulation in term of probability of detecting post disaster
scenario, probability of miss detection and probability of
false alarm.
Keywords— Post Disaster Communication, Emergency
Mode, Normal Mode, Double Threshold, Energy Detection
I. INTRODUCTION
The telecommunication existing network can be
damaged or destroyed by manmade or natural disaster. In
result, that area will have no connection with national
communication infrastructure. 50% death occurs in two hours
after disaster [1]. The lives can be saved if there are some
mean of communication before the arrival of rescue team. The
survivors need urgent help after the disaster. Therefore, post
disaster communication is required to connect the disaster area
with whole country [1].
The post disaster communication can be achieved by
using Ad-hoc network. Ad-hoc network is used by many
researchers for emergency network. Wireless Emergency
Network can be obtained by multi hop Ad-hoc network. In
Wireless Emergency Network, there is ability of self-healing
[2]. But in disaster scenario, the resources are limited. Ad-hoc
network cannot reuse the available resources and also
unstable. In [1], the authors proposed that using cluster based
Ad-hoc network, the network become stable and resources can
be reuse. Thus, it increase the capacity. In cluster based
network, cluster head will forward the data of cluster
members. In [2], the authors proposed route repairing for
Wireless Emergency Network making it more stable, energy
efficient and reliable. In [3], authors presented cross layer
framework for the post disaster scenario. The authors
presented an adaptive protocol selector which select
appropriate protocol for each layer. Furthermore, the authors
in [4] proposed three algorithms using single threshold based
on cognitive radio, cellular network and artificial neural
network to detect the post disaster scenario. But there is
possibility of false alarm if the mobile is in deep fading area.
The authors in [5], proposed double threshold energy detection
algorithm for cognitive network and proved that false alarm
can be reduced by using double threshold.
In the paper, we presented a new algorithm using
double threshold for detecting post disaster scenario. We used
energy detection technique for the decision making. In this
paper, we presented the performance of our proposed
algorithm using BPSK modulation scheme over AWGN
channel. The performance is analyzed in term of probability of
detection post disaster scenario, probability of misdetection of
post disaster scenario and probability of false alarm. All the
results are evaluated using Monte Carlo simulation in
MATLAB.
II. SYSTEM MODEL
The system investigated in this paper is shown in
figure 1. Our system model consists of one mobile station and
two base stations. The mobile station is denoted by MS and
the two base stations are denoted by BS1 and BS2. MS is
connected with BS1 and BS2 is adjacent base station. The
distance between the MS and BS1 is denoted by d1. Similarly,
the distance between the MS and BS2 is denoted by d2. In our
system model, d1 is 2 km and d2 is 3 km. AWGN channel is
considered between MS, BS1 and BS2. The path loss
exponent is considered as n = 2. All the nodes (MS, BS1 and
BS2) are mounted with single antenna.
It is assumed that MS has already installed software
based identification module. There are two modes on which
219
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-228-320.jpg)
![probability increases. The post disaster scenario can be easily
detected if 0 is high.
The probability of miss detection of post disaster scenario
is shown in figure4. From the result, it is obvious that we will
surely miss the detection of post disaster scenario if the
threshold 1 is kept very low. Keeping the 1 low, then greater
probability of switching to Normal Mode and low probability
of switching to Emergency Mode. As the threshold 1
increases, then we can easily detect the post disaster scenario.
Therefore, the probability of miss detection decreases as the
threshold 1 increases.
The probability of false alarm of our proposed algorithm is
shown in figure5. To calculate false alarm we assume that SNR
Fig4. Probability of miss detection of post disaster
scenario
Fig5. Probability of false alarm
is 0 dB. According to our simulation result, it is stated that if
the thresholds are low then the probability of false alarm
increases. We get 100 % probability of false alarm at 0 dB
threshold.
As the threshold increases, the probability of false alarm
decrease. We get 0% false alarm for threshold equal to or
greater than 10 dB.
V. CONCLUSION
In this paper, we presented double threshold energy
detection algorithm for detecting post disaster scenario. This
algorithm enable mobile station to detect the post disaster
efficiently and take right decision in switching its mode from
normal to emergency or emergency to normal. Decision is
made by considering the energy of signals received from
connected Base station and from adjacent base station. The
performance is presented in term of probability of detection of
post disaster scenario, probability of miss detection of post
disaster scenario and probability of false alarm. In future, we
will work future on identification of post disaster scenario
using double threshold considering the adjacent channels.
REFERENCES
[1] Sonia Majid and Kazi Ahmed, “Cluster-based
Communications System for Immediate Post-disaster
Scenario” Journal of Communication, Vol. 4, No. 5, June,
2009.
[2] Xin Ma, Quan-yi Huang and Yan Kang, “A New Algorithm
for Stable Communication in Wireless Emergency
Network” International Conference on Computer and
Management, pp. 1-4, May 2011, Wuhan, China.
[3] Sonia Majid and Kazi Ahmed, “Cross-layer Framework
for Post-disaster Communications” IEEE International
Conference on Telecommunications and Malaysia
International Conference on Communications, 14-17 May
2007, Penang, Malaysia
[4] Sonia Majid and Kazi Ahmed, “CIP- Cognitive
identification of post-disaster communication” IEEE
conference on Telecommunications, pp. 43-47, June 2012,
Thailand
[5] Jinbo Wu, Tao Luo, Jianfeng Li and Guangxin Yue “A
Cooperative Double-threshold Energy Detection Algorithm
in Cognitive Radio Systems” IEEE conference on Wireless
Communications, Networking and Mobile Computing,
pp.1-4, Beijing, China.
[6] Haitao Wang, Lihua Song, Jianzhou Li, Jiaxin Deng,
“Wireless Self-organizing Emergency Communication
Network based on Network Clustering and Information
Ferry” , 2nd
Internation Conference on consumer
Electronics, Communications and Networks, pp. 3704-
3707, Yiching, China.
[7] Sonia Majid and Kazi Ahmed, “Post-disaster
Communications: A Cognitive Agent Approach” 7th
International Conference on Networking, pp- 23-30 Cancun
Mexico, April 2008.
[8] Marco Di Felice, Luca Bedogni, Angelo Trotta, Luciano
Bononi, Fabio Panzieri, Giuseppe Ruggeri, Gianluca Aloi,
Valeria Loscr, Pasquale Pace, “Smartphones Like Stem
Cells: Cooperation and Evolution for Emergency
Communication in Post-Disaster Scenarios” First IEEE
International Black Sea Conference on Communications
and Networking 2013 (IEEE BlackSeaCom 2013), Batumi,
Georgia.
[9] Jinbo Wu, Tao Luo and Guangxin Yue, “An Energy
Detection Algorithm Based on Double-threshold in
Cognitive Radio Systems” The 1st International Conference
on Information Science and Engineering (ICISE2009), pp.
493-496, Nanjing China.
221
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-230-320.jpg)
![Design and Analysis of Magnetic MEMS
Accelerometer for Inertial Navigation
M. Ameer Umar Malik, Umer Izhar, Umar Shahbaz, Javaid Iqbal
Department of Mechatronics Engineering
College of Electrical and Mechanical Engineering, NUST Islamabad, Pakistan
ameerumar@yahoo.com, umerizhar@yahoo.com
Abstract- Accelerometer is an important sensor of inertial
navigation system. Focus of new developments is a low cost, small
size sensor with high performance that meets the requirements of
navigation. In this paper a multi wafer magnetic MEMS
accelerometer design is proposed for inertial navigation.
Capacitive pick off is used to sense deflection of sensor mass and
permanent magnet & coil is used for rebalancing of deflected
mass of the sensor. Total two sides capacitive area of sensor is 18
mm2
with gap of 8µm. Permanent magnet with thickness of 330
µm and diameter of 700 µm and soft magnet material coating is
used. Out of plane coil with total 64 turns, 8 layers and cross-
section area 20µm x20µm is used. Fabrication process outlines
for both, the magnetic system and coils are proposed separately.
Stress analysis for flexures of sensing pendulum has been carried
out. Close loop frequency response analysis is carried out with a
proposed PID controller.
Key Words- Magnetic MEMS, Accelerometer
I. INTRODUCTION
Accelerometers are the important sensors of inertial
navigation systems. Traditional high accuracy
electromechanical accelerometers with quartz pendulum are
highly accurate with resolution of <1µg and one year
repeatability of ≤ 50 µg with input range of ± 50 g but also are
very expensive [1]. Whereas the best available MEMS devices
with input range of ±10g have resolution of up to 250 µg and
max repeatability of 25mg [2] but cost of MEMS sensors is
very low as compared to traditional electromechanical
accelerometers. Although accuracy of MEMS accelerometers
is yet not comparable to quartz flexure accelerometers but due
to their mass production process these provide very economical
solution particularly for low accuracy navigation applications.
MEMS accelerometers are widely being used in automobiles
but their accuracy is not enough for high accuracy inertial
navigation use. Variety of MEMS accelerometers are being
developed and reported for navigation applications. Most of
these are capacitive devices with electrostatic rebalancing
mechanisms. Recently reported ultra precision MEMS
accelerometer developed by Colibrys is also a capacitive
device working in close loop based on electro static
rebalancing [3-4].
In this paper a magnetic rebalancing mechanism for a
MEMS accelerometer is designed to get navigation grade
performance. Permanent magnets are being used in MEMS for
their repeatable performance. As the techniques of permanent
magnets fabrication in MEMS are being developed their use is
also increasing in high performance devices [5].
II. NAVIGATION GRADE ACCELEROMETERS
High performance accelerometers are required for high
accuracy navigation applications. Different performance
parameters can be specified for inertial navigation grade
accelerometers. In broad terms accelerometer performance can
be specified by two parameters i.e. bias and scale factor [1-6].
Accelerometers for navigation require high level of bias
stability, repeatability, temperature sensitivity and fine
resolution. Requirements of sensor parameters and comparison
for different requirements are discussed in [7]. Some of critical
parameters required for high accuracy navigation application
are given in table-1.
A variety of transduction mechanisms have been used in
MEMS accelerometers. Initially piezo-resistors were used in
MEMS accelerometers [7]. Then other principles like
capacitive, tunneling current, thermal, optical, electromagnetic
and piezoelectric etc are also utilized with their some
advantages and unavoidable disadvantages [8-9]. Capacitive
based sensing gives greater design flexibility than piezoelectric
quartz technology [6]. The capacitive devices are very popular
for their high precision.
TABLE:1 INERTIAL NAVIGATION REQUIREMENTS
Parameters Values
Range ± 25 g
Bias Stability ≤50 µg
Scale Factor Stability ≤ 50 ppm
Resolution ≤ 5µg
Frequency Range DC-200 Hz
Max. Shock in 1msec > 20 g
Temperature Range -40°C to 80 °C
III. PERMANENT MAGNETS IN MEMS
Use of electrostatic force is very common in
microelectronic systems due to its better force scaling at micro
scale level and also technology is highly developed as
compared to electromagnetic systems at micro level.[10]
Scaling differences of electrostatic and electromagnetic forces
222
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-231-320.jpg)
![is discussed in details by Edward P Furlani [5]. Magnetic field
from electromagnet unfavorably scales with diminishing sizing,
whereas the fields from permanent magnet are scale invariant.
Source of magnetic field in electromagnet is current but the
magnetic field in permanent magnet is residual magnetization
which arises at atomic level [5]. For actuation electromagnetic
is more stable for high force and for large gaps [11].
Use of permanent magnets was limited in MEMS due to
difficulties of their fabrication at micro scale level but now it is
increasing as different fabrication technologies are being
reported. Larger actuation distance can be achieved by using
magnets as compared to electrostatic actuation[5,12,13].
Permanent magnet is used in traditional navigation grade
electromechanical accelerometers. In MEMS accelerometer to
get µg resolution and low noise of accelerometer a heavier
proof mass is recommended and to rebalance this proof mass
an electromagnetic system with a permanent magnet can be
used.
IV. PROPOSED DESIGN
Accelerometer can be constructed and operated in open
loop or closed loop configuration. The basic construction of
accelerometer is such that a proof mass is suspended in case
and confined to a zero position by means of a spring, when the
acceleration is applied to the case of sensor, proof mass is
deflected with respect to its null position within the case, this
deflection of proof mass is proportional to specific applied
force along input axis. More accurate output can be obtained
by applying a rebalancing mechanism for deflected proof mass
[14]. This rebalancing mechanism can be either
electromagnetic or electrostatic. In traditional
electromechanical accelerometers the rebalance mechanism is
electromagnetic while in MEMS accelerometers it is
electrostatic.
A permanent magnet based MEMS accelerometer design is
proposed and analyzed. It is a three wafer design in which the
centre wafer with proof mass is sandwiched between two
symmetrical permanent magnet wafers. The bench mark of the
design is to use permanent magnet for rebalancing of proof
mass of MEMS sensor instead of electrostatic rebalancing.
Permanent magnet with volume of 0.7mm3 has been selected
to get bulk fabrication with powder size of 5~ 50 µm and this
can be accommodated in 500 µm thick wafer. Proof mass
includes the rebalance coils and capacitance area for pickoffs.
The designed proof mass has total area of 16 mm2 with
thickness of 500 µm and coils mass is additional. This heavy
proof mass results in a low Brownian noise [15] and greater
stability and repeatability. The proof mass is suspended by two
flexure beams attached to the fixed frame of centre wafer. In
response to the applied input acceleration the proof mass
deflects and change in capacitance is detected. One side
capacitance area for pickoff is 9 mm2
which is large as
compared to in plane MEMS accelerometers, this result in fine
pickoff resolution. Rebalance force is applied by the passing
current through the coils present on the proof mass. This
current is the output of the accelerometer giving information
about the acceleration applied to the system.
This can be modeled as a second order mass damping
system. Proof mass inertia is ‘J’, beams have effective stiffness
‘K’ and damping is ‘D’.
Transfer function is given as
H(s) =
Where ‘a’ is applied acceleration and ‘Θ’ is angular deflection
of proof mass.
Fig.1(a) Centre wafer with proof mass. (b) Three wafer device
For an applied input acceleration the total force depends on
the total proof mass of pendulum part. This input force is to be
balanced by coils. Lorentz force of the coils with current
flowing through is the rebalance force. The Lorentz force on a
conductor of length ‘L’ and with the current ‘i’ in an externally
generated magnetic flux density ‘B’ perpendicular to the wire
is given by F = Li × B. Rebalance force is directly proportional
to the magnetic field available in the working gap of coil. To
get maximum force magnetic field and coil placement need to
be optimized. Few analytical methods to get axial force
between a cylindrical magnet and a thick coil are discussed in
ref [16]. To get maximum magnetic field in working gap of
coil a magnetic loop is designed. Also to get maximum force,
magnetic field direction should be perpendicular to the current
flowing through the coil. Different configurations of magnetic
system with permanent magnet and soft magnet materials are
proposed, performance of each is evaluated to select the best
possible configuration. Magnetic flux density plot of selected
configuration is shown in in figure-2. Field variation in two
dimensions is simulated to get optimized location of coil.
Variation of flux density along z direction is plotted in figure-3.
The resolution of 10 µg is selected as a target. Stiffness
hinges are designed so that the smallest input of 10 µg
produces a deflection of 4.33e-7 rad. This minimum deflection
is used to estimate the overall gain required for pickoff system.
Overall pickoff gain of 2.32e6 is used and gain values of
different portions of pickoff system are calculated. Details are
discussed in closed loop analysis of system. All the parameter
values are calculated for this proposed design.
Stiffness - K= 2.32 g. mm / rad.
Damping - D = 1.16 g-s-mm.
Inertia - J = 0.0000307 g-mm-s2
223
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-232-320.jpg)
![Fig.2 Surface Plot of Magnetic Flux Density
V. FABRICATION PROCESSES
Design of this sensor system consists of three wafers as
shown in figure-4. Three wafers will be manufactured
separately and then bonded together to form a sensor. Two
wafers with permanent magnet are symmetrical. Coil is
fabricated on both sides of the third wafer. This third wafer is
to be fixed between two wafers with permanent magnet.
Fabrication processes for both magnet and proof mass coil are
proposed.
Permanent magnet is the key part in this design and fabrication
of permanent magnets has always been challenging in bulk
batch fabrication of MEMS. Permanent magnets are produced
by metallurgy process of mechanically pressing and heating.
But this process is yet not compatible with MEMS production
processes.
Fig.3 Magnetic Flux density along z direction
In MEMS conventionally permanent magnets are fabricated
by vapor deposition methods like sputtering, evaporation, pulse
laser deposition or by electrochemical deposition. By these
methods a layer of only few micrometers can be deposited.
Unconventional strategies like using magnetic powders of
5~50 micrometer sizes are being adopted to fabricate magnets
of required size at wafer levels. Permanent magnets fabrication
using magnetic powder and fabrication at wafer level is a
choice for this design.
Fig.4 Three Wafers of Magnetic MEMS accelerometer
Details and performance of this method is discussed in
details by Ololade D Oniku in ref [12]. Conventional
deposition of micro magnets with results is described in [13,
17]. Nd-Fe-B powder of size 15~ 50 µm can be used to get the
size required in this design and this can provide remanence Br
of 0.5 ~ 0.7 Tesla. Doctor’s blade technique as discussed in ref
[12] can be easily adopted at wafer fabrication for this design.
For fabrication of required size permanent magnet powder
fabrication is the best suited technique available. Fabrication
steps for complete magnetic system are proposed as
Cavity etching as per dimensions of magnet and
working clearance using DRIE process.
Coating of soft magnet material with thickness of 30 ~
40 µm.
Fill sacrificial material and form a cavity for permanent
magnet powder.
Permanent magnet fabrication in cavity using powder.
Coating of soft magnet material with thickness of 30 ~
40 µm above magnet.
Removal of sacrificial material to form working gap
for coil.
Fig.5 Permanent Magnet Fabrication (a) Cavity etching for soft magnet
coating. (b) Coating of soft magnet material. (c) Cavity forming for
permanent magnet. (d) Permanent magnet fabrication using powder. (e)
Coating of soft magnet material above magnet. (f) Removal of sacrificial
material layer to form working gap.
Fabrication of rebalance coil in multi layers is required on
the centre wafer. Fabrications of spiral coils have been reported
and one such multi layer coil with 36 turns in two layers has
224
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-233-320.jpg)
![been reported by Chong H Ahn, and Mark G Allen [18]. The
rebalance torquer coil required in this design is of 64 turns with
8 layers. The spiral pattern is Cu electroplated in 8 layers with
each layer separated by an insulating layer.
Fabrication steps for coil are proposed as
Pattern planner spiral coil on wafer and electroplate
copper coil layer.
Sputter 1st insulating layer then pattern for 02 copper
links for 2nd layer and electroplate copper links.
Pattern for planner spiral 2nd layer on 1st insulating
layer and electroplate 2nd copper coil layer.
Sputter 2nd insulating layer then pattern for 02 copper
links (different location as compared to step ‘b’) for
3nd layer and electroplate copper links.
Pattern for planer spiral 3rd layer on 2nd insulating
layer and electroplate 3rd copper coil layer.
Repeat this process to complete 8 copper coil layers.
Etching of insulating layer.
This coil is to be fabricated on both sides of wafer. Two
coils are connected in series therefore via connection across
wafer thickness is also made. Capacitive area coating and
connection pads are connected through two flexure hinges.
Fig.6 Coil Fabrication processes (a) Pattern planer spiral coil on wafer and
electroplate copper coil layer. (b) Sputter 1st insulating layer then
pattern for 02 copper links for 2nd layer and electroplate copper links.
(c) Pattern for planer spiral 2nd layer on 1st insulating layer and
electroplate 2nd copper coil layer. (d) Sputter 2nd insulating layer then
pattern for 02 copper links (different location as compare to step ‘b’) for
3nd layer and electroplate copper links. (e) Pattern for planer spiral 3rd
layer on 2nd insulating layer and electroplate 3rd copper coil layer. (f)
Repeat this process to complete 8 copper coil layers. (g) Etching of
insulation layer.
VI. ANALYSIS
A. Structure Stress Analysis
Sensor structure can be subjected to different forces during
manufacturing, storage and operational life. Analysis of
designed structure is carried out to evaluate its survival/
performances in different conditions. Flexures are the critical
parts that are under stress, in free condition when centre part is
not fixed between two wafers the centre part can bend freely
and subjected to maximum stress as shown in figure-7. Input
moment is defined as ‘M’ where ‘P’ is the mass of proof mass
part of centre wafer.
σ = as here are 2 flexure with section modulus ‘S’
σB = = 5.4012 Kg/ mm2
σA = = 6.3832 Kg/ mm2
Also deflection and angle of point ‘B’ in free condition are
calculated.
δB = ( + ) = 0.0111 mm
Θ = ( + ) = 0.0433 rad
Maximum deflection of point ‘C’ in free condition is calculated
as
δC = δB + Θ *2L = 0.2494 mm
But in assembled form this maximum deflection is restricted by
the fixed wafers. Maximum deflection possible in assembled
form is the gap provided for capacitance pickoff. Stress
analysis in the condition when centre wafer is fixed between
two fixed wafers is also carried out.
Fig.7 Static condition when centre wafer is free
This condition is analyzed by using static equilibrium
conditions and general solutions for deflection of statically
determinate beams [19].
RA + RB = nP --- 1
2L RB + M = nPL --- 2
– + ( – ) 2L = 0 --- 3
Moment is calculated by solving three equations and stress
due this moment ‘M’ is calculated σB = M/2S = 0.2381
225
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-234-320.jpg)
![Kg/mm2
. This stress is produced by application of 1g input and
is much less than the stress calculated in free condition. From
this analysis it is concluded that in assembled form the flexures
can bear much higher inputs before failure.
Fig.8 Static equilibrium condition when centre wafer is fixed between two
wafers.
Stress analysis and deflection analysis is performed in FEM
software and the results are verified. Results of FEM analysis
for the free condition are shown in figure-9.
Fig.9 FEM stress analysis for free condition
B. Mechanical Noise
Intrinsic noise of sensor can be due to mechanical and
electronic components. Mechanical–thermal noise due to
Brownian motion is usually analyzed for MEMS sensor and is
expressed total noise equivalent acceleration (TNEA) as
√
Where “kB” is Boltzmann’s constant, “T” is absolute
temperature, “w” is natural frequency and “Q” is quality factor
of flexure. High mass and quality factors lead to low noise.
Noise level of this design is estimates as 4.3µg/√Hz.
C. Closed Loop System Analysis
Resolution, measurement range, bandwidth, non linearity
and noise are the parameters to be analyzed for performance of
sensor. Simulation analysis of the designed sensor is performed
to find out the behavior of sensor for critical performance
parameters. Measurement range is the minimum and maximum
acceleration that sensor can response along its input axis with
certain accuracy. As spring stiffness is designed lowest to get
minimum acceleration detection, in open loop operation the
small distance between the moving capacitance surface and the
fixed surface limits the maximum measurement range.
Performance of open loop accelerometer is very limited
concerning to measurement range, bandwidth and linearity [20].
In close loop operation maximum measurement range is
limited by the feedback force that can be applied to rebalance
the deflected mass.
This design can only be operated in closed loop with high
gain pick off system with negative feedback. To achieve
desired system dynamics PID controller coefficients should be
carefully designed [21]. Stiffness of flexure is very small to
detect smallest input. The deflection produced by smallest
input is to be detected by the pickoff system. Gains of pickoff
blocks are calculated to detect smallest input. Closed loop
analysis of the design is performed in Matlab. Architecture of
close loop with a PID controller HCONT is shown in figure-10.
Pick off gain and other parameter used for control and analysis
are HCA = 10 A/V. HPA = 6880 pF/Arc-min, KPSA =
0,2V/pF with τPSA = 6.37e-6 sec. KTM =201 g-mm/A with
τTM = 3e-5 sec. Steady state sensitivity of sensor depends on
mass of pendulum ‘m’, length of pendulum ‘l’ and KTM of
rebalance torque system. As these parameters remain constant
for a sensor for this system steady state sensitivity if calculated
as Kss = ml / KTM = 0.1005 /201= 0.0005A.
Fig.10 Close loop architecture with output configuration
Fig.11 frequency response plot of closed loop architecture
-350
-300
-250
-200
-150
-100
-50
Magnitude(dB)
10
2
10
3
10
4
10
5
10
6
10
7
-450
-360
-270
-180
-90
0
90
Phase(deg)
Bode Diagram
Gm = 75.5 dB (at 1.66e+004 rad/sec) , Pm = Inf
Frequency (rad/sec)
i
a
pFα
HPA
(pF/arc- min)
Mi
n ΔM
V
-
+
HTM – Coil & Magnet
MTM
ml
HPSA
(V/pF)
HCA
(A/V)
i
Gp
HCONT
226
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-235-320.jpg)
![Frequency response analysis shows a bandwidth of 1.18e+4
rad/sec. Step response shows a rise time of 1.5154e-004 sec
with 27.5 % overshoot for this output configuration as is shown
in figure.12.
Fig.12 Step response of close loop architecture
Output response for a square input signal of 100 Hz is
shown in figure.
Fig.13 Response of close loop architecture
Maximum deflection of proof mass is limited by the
capacitance gap. In close loop operation the deflection of proof
mass must be minimum and less than the capacitance gap to
avoid electrical contact. To control the angular deflection the
close loop analysis is performed by transfer function as per
configuration is shown in figure 13.
Fig.13 Close loop architecture for angular deflection
Deflection for a step input signal of different g is shown in
figure-14. Maximum deflection at input signal of 10 g is less
than 1.5e-3 rad. Minimum deflection of proof mass ensures
reduced non linearity of output.
Fig.14 Deflection of sensor against step input
PID control parameters can be further optimized to achieve
best close loop performance of accelerometer.
VII. CONCLUSION
In this paper a new magnetic MEMS accelerometer is
designed with capacitive pickoff using permanent magnet and
coil for rebalancing of proof mass instead of traditional
electrostatic rebalancing of MEMS accelerometers. Resolution
of 10 µg can be achieved to fulfill navigation requirements.
Heavier proof mass results in reduced noise and also can be
rebalanced by electromagnetic force. Silicon structure provided
good thermal stability and magnets also provide low thermal
sensitivity as in traditional electromechanical accelerometers.
Permanent magnet can be fabricated in the bulk fabrication of
MEMS devices. Fabrication processes of permanent magnet
and coil are also proposed. In future study the device can be
fabricated and different performance parameters can be studied.
REFERENCES
[1] Honeywell, Q-Flex Navigation Accelerometers,
http://www.inertialsensors.com
[2] Colibrys Single axis analog accelerometer
http://www.colibrys.com
[3] Y.Dong, P.Zwahlen, A.M Nguyen,R. Frosio, F. Rudolf,“ Ultra
high precision MEMS accelerometer” Transducers’11, Beijing,
China,June5-9,2011.
[4] P. Zwahlen, Y Dong, A-M. Nguyen, F. Rudolf, J-M Stauffer,
P. Ullah, V. Ragot, “Breakthrough in High Performance Inertial
Navigation Grade Sigma-Delta MEMS Accelerometer”
[5] E.P Furlani, Permanent Magnet and Electromechanical
Devices, Material analysis and Applications Academic press 2001.
[6] Ralph Hopkins, Joseph Miola, Roy Setterlund, Bruce Dow,
William Sawyer, “The Silicon Oscillating Accelerometer: A High-
Performance MEMS Accelerometer for Precision Navigation and
Strategic Guidance Application” Institute of Navigation 61st
Annual Meeting Cambridge, MA, June 27-29, 2005.
[7] Navid Yazdi, Farrokh Ayazi, Khalil Najafi, “Micro-machined
Inertial Sensors” Proceedings of IEEE Vol. 86, No. 8 Aug 1998.
[8] Cenk Acar and Andrei M Shkel, “Experimental evaluation and
comparative analysis of commercial variable-capacitance MEMS
accelerometers”
Step Response
Time (sec)
Amplitude
0 1 2 3 4 5 6
x 10
-3
0
1
2
3
4
5
6
7
8
x 10
-4
System: Gcsys
Peak amplitude: 0.000638
Overshoot (%): 27.5
At time (sec): 0.000434
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
-1
0
1
2
3
4
5
6
x 10
-4
Linear Simulation Results
Time (sec)
Amplitude
0 0.005 0.01 0.015 0.02 0.025
0
0.5
1
1.5
x 10
-3 Step Response- Displacement of pendulum
Time (sec)
Amplitude(rad)
10 g
8 g
6 g
4 g
2 g
1 g
Gp
α
HTM
Coil & Magnet
i
HPA
(pF/arc- min)
HPSA
(V/pF)
HCA
(A/V)
pFVi
HCONT
ml
-
+
ΔM
Min
MTM
a
227
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-236-320.jpg)
![[9] Teodor Lucian Grigorie “The Matlab/Simulink Modeling and
Numerical Simulation of an Analog Capacitive Micro-
Accelerometer. Part1: Open loop ” MEMSTECH’ May 2008
Polyana UKRAINE
[10] Qingkun Zhou, Azhar Iqbal, Pinhas Ben-Tzavi, Dapeng Fan,
“Design, Analysis, and optimization of a magnetic micro
actuators” Proceedings of the ASME 2009, November13-19
Florida ,USA.
[11] Tsung-Shune Chin, “Permanent magnet films for application
in micro-electromechanical systems” Journal of Magnetism and
Magnetic Materials 209 (2000) 75-79.
[12] Ololade D Oniku, Benjamin J Bowers, Sheetal b Shetye,
Naigang Wang and David P Arnold “Permanent magnet
microstructures using dry-pressed magnetic powders” Journal of
Micromechanics and Microengineering published 12-june 2013.
[13] David P. Arnold., Naigang Wang., “Permanent Magnet for
MEMS”. Journal of Electromechanical Systems. Vol.18. No. 6,
December 2009.
[14] D.H.Titterton, J.L. Weston “Strapdown inertial navigation
technology” 1997 Peter Peregrinus Ltd.
[15] Ki-Ho Han , Young-Ho Cho, “ Self-Balanced Navigation-
Grade Capacitive Micro-accelerometer Using Branched Finger
Electrodes and Their Performance for Varying Sense Voltage and
Pressure ” Journal of Microelectromechanical systems, Vol. 12,
No.1, February 2003.
[16] Will Robertson, Ben Cazzolato, Anthony Zander, “Axial
Force Between a Thick Coil and a Cylindrical Permanent Magnet:
Optimizing the Geometry of an Electromagnetic Actuator” IEEE
Transactions on Magnetics Vol. 48 No.9, September 2012.
[17] Paul McGuiness David Jezersek, Spomenka Kobe, “100- µ-
thick Nd-Fe-B magnets for MEMS applications produced via a
low-temperature sintering route.”
[18] Chong H. Ahn, Mark G. Allen. “Micro- machined Planar
Inductors on Silicon Wafers for MEMS Applications”. IEEE
Transactions on Industrial Electronics Vol. 45, No.6 December
1998.
[19] Andrew Pytel, Ferdinand L. Singer, “Strength of Materials”
[20] Teodor Lucian Grigorie “The Matlab/Simulink Modeling and
Numerical Simulation of an Analog Capacitive Micro-
Accelerometer. Part2: Closed loop” MEMSTECH’ May 2008
Polyana UKRAINE
[21] Yin Liang, Liu Xiaowei, Chen Weiping, Zhou Zhiping, “High
resolution interface circuit for closed-loop accelerometer” Journal
of semiconductor Vol.32, No.4 April 2011.
228
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-237-320.jpg)
![Feasibility Assessment of Running JP-8 Fuel in
Diesel Engine
M. Shahan1
, Ali Sarosh2
,S. Javaid3
Department of Aerospace Engineering
National University of Sciences and Technology
Pakistan
1
heavy.cavalry.archer@gmail.com
2
alisarosh@cae.nust.edu.pk
3
saad.javaid@gmail.com
Abstract— Single fuel concept (SFC) demonstrates the need for
a safe and cost effective fuel that can be utilized by light-duty as
well as specialized heavy-duty vehicles. Logistics and pipeline
operations are greatly simplified when SFC is implemented for
vehicles. The study considers the technical feasibility of using
aviation fuel JP-8 in diesel engines. It compares the effects on fuel
injection, combustion, engine performance and emissions when
using JP-8 fuel instead of diesel inside a compression ignition
engine. The performance of both fuels at various operating
conditions is calculated. Besides this, a detailed review of chemical
kinetics of diesel and JP-8 and the effect of thermodynamic
properties on engine performance and emission is also presented.
In addition, combustion analysis of both the fuels is done in which
endothermic enthalpies due to air and water at various fuel-air
ratios are calculated. The results reflect that use of JP-8 in diesel
engine leads to a penalty of torque and fuel consumption primarily
due to lower density and viscosity. The reduced cetane number
causes increase in ignition delay and premixed combustion of JP-
8. Modifications to engine are proposed so as to match the
performance of both the fuels. Analysis show that implementation
of SFC is best suited for supplication in heavy-duty diesel engine
vehicles
Index Terms— JP-8, Enthalpy, Cetane number, Ignition delay,
Pre-mixed combustion
INTRODUCTION
Single fuel policy states that Jet Propellant 8 (JP-8) fuel must
be used in all air and ground battle field vehicles. According to
a research, 38.6% of army logistics supply consists of fuel. JP-8
is considered as a safe fuel for storage as well as during
operations. The US army and NATO will be using JP-8 fuel in
their aircrafts and ground vehicles till 2025[1].Using JP-8 in
ground vehicles offered my advantages. It reduced nozzle
fouling, exhaust emissions, water entrainment and microbial
growth inside the fuel tanks and increased fuel filter replacement
intervals, storage stability and enhanced oil change and filter
replacement intervals[2]. JP-8 is a jet fuel and belongs to the
class of kerosene. The flash point of JP-8 is 100 degrees
Fahrenheit. It is specified by MIL-DTL-83133. It was first used
by the NATO countries having a NATO code F-34. JP-8 is
almost similar to Jet-A1 fuel except that it has three military
additives. Tests were conducted on standard diesel cycle to give
an overall idea of replacing JP-8 fuel with diesel [3].Using JP-8
in ground battle field equipment leads to many advantages.
Logistics are simplified, oil lubricity is increased and exhaust
emissions are reduced. Single fuel concept also carried some
draw backs as well. Wear in fuel injection pumps exhaust valves
increases with the use of JP-8[4].Briefing charts regarding the
properties of JP-8 and other military fuels, different additives
that are used in fuels and their effect on fuel properties has been
presented by Schmitigal, Joel Tebbe and Jill in [5]. Similarly,
world jet fuel specifications have been given in [6]. A brief
review of the previous work done is this regard is given below:
Laboratory evaluation of JP-8 in diesel engine was
conducted as given in [7] in which JP-8 fuel was found
to be satisfactory for use in diesel engine but the
maximum engine power was reduced and leak in fuel
injection lines was observed. Compensation was done
by increasing thermal efficiency.
Engine testing with various blends of JP-8 was
conducted by April Covington with the perception that
in future an EPU will be designed to optimize engine
performance with JP-8[8]
Direct imaging and two color thermometry technique
was applied to verify the emission trends for JP-8 and
fossil diesel fuel in an optically accessible single
cylinder diesel engine. Image analysis focusing was
done to determine the characteristics of combustion.
Results verified that JP-8 emitted less smoke and
increased HC emissions [9].
Tests with petroleum, hydro processed and Fischer
Tropsch diesel and jet fuels were carried out in a direct
injection single cylinder diesel engine. Decrease in
ignition delay was quantified by increasing the derived
cetane number (DCN) of fuel. DCN decreased the
ignition delay[10]
Systematic study of ignition delay for different fuels in
2.44 L heavy duty single cylinder diesel engine was
conducted over a wide range of temperatures and
densities. Results indicated that fuel properties relating
to spray break have a nominal effect on the ignition
delay period [11].
229
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-238-320.jpg)
![ Oxides of nitrogen are harmful to the environment.
During combustion, oxides of nitrogen are formed
depending on the temperature and pressure. Hence
there is a need to reduce nitrogen oxide emissions from
diesel engine. Control on NOx emissions was obtained
by introducing varying quantities of cetane improver
additive in diesel fuel by enhancing the ignition quality
of fuel[12]
Mie-scattering technique is used in [13] to determine
the spray characteristics of fuels which includes spray
angle and spray tip penetration. Spray characteristics
are in turn correlated to combustion and emission
trends of fuels
Tests conducted on S60 engine revealed that JP-8
reduced nitrogen oxide and particulate matter
signatures in the exhaust emission and hence it is
environmentally friendly. Torque and fuel economy
penalty were compensated by increasing the pulse
width of fuel [1].
Tests results carried out on 60 kW direct injection
diesel engine running alternatively on diesel, JP-8 and
JP-8 treated with varying quantities 2-ethylhexyl
nitrate showed that ignition delay decreases with the
addition of cetane improver additives [14]. Sound
analysis regarding maximum heat release rate, engine
efficiency, cylinder pressure and exhaust were also
presented.
Analysis of the usage of blends of biodiesel and JP-8 in
diesel engine as given in [15] revealed that specific fuel
consumption decreases and NOx emissions increase by
increasing the quantity of biodiesel in test fuels. It was
concluded that these fuel blends can be effectively used
inside diesel engine.
METHODOLOGY
The basic aim was the verification of the results whether JP-
8 could be used as a replacement of diesel fuel in diesel engine.
Main objective was to understand the effects on combustion and
performance of engine when replacing diesel with JP-8 in diesel
engine. The approach which was followed is as described:
Analysis of actual diesel and JP-8 fuel properties was
done so that effect of each property on engine
performance can easily be quantified
Heat of combustion analysis of both fuels was
performed to obtain an indirect insight about the
thermodynamic parameters of fuels and power output
of engine when replacing JP-8 with diesel fuel
Dynamometer testing of engine is done to determine
the performance parameters such as torque and power
output
Assessment of fuel economy and strategies for
matching the power output with JP-8 with different
adjustments
FUEL PROPERTIES
Density effects the composition of fuel. Greater density
indicates lower volatility. Density of diesel is greater than JP-
8.It relates to the fuel pressure and dynamic start of fuel injection
in the engine[16]
Cetane number provides a measure of ignition delay of fuel.
This property indicates the quantity of straight chain
hydrocarbons in fuel. Lower cetane number implies lower
proportions of straight chain hydrocarbons and in turn longer
ignition delay[17].
Heating value of a fuel determines its energy content. Higher
heating value results in higher power output of engine [11]. JP-
8 has higher heating value than diesel but its lower density leads
to lower heating value on volume basis.
Viscosity provides a measure of resistance to flow. Working
of fuel injection pump is directly related to viscosity of fuel.
Pump wear and leakage in injection pumps increases due to
lower viscosity. Generation of spray pattern is also governed by
the viscosity of fuel[18].
Volatility of fuel is governed by distillation curve. Heavier
fuel results in incomplete vaporization and combustion and it
causes soot and smoke quantity to be increased[18].
Hydrocarbons having multiple benzene rings are called Poly
Aromatic Hydrocarbons (PAH’s). Diesel fuel contains higher
proportions of aromatics. Aromatics effect density, cetane
number and soot formation characteristics of fuel [17, 19, 20].
Environmental Protection Agency (EPA) requires that sulfur
content in the fuel should be as low as possible. JP-8 has lower
sulfur content as compared to diesel fuel so it is environmentally
friendly as compared to diesel.
CHEMICAL TESTING OF FUELS
Samples of JP-8 and diesel were taken and complete
chemical testing was done so that engine performance trends can
be co-related to chemical properties of fuels. Following results
were obtained from the chemical testing of JP-8.
Chemical testing of JP-8
Table 1 Test tests of JP-8
Test Method Test
Limits
Test
Result
Appearance
Color, Saybolt ASTM D156 Report +29
Composition
Total Acidity, mg
KOH/gm
ASTM
D3242
0.015
max
0.007
Sulfur, Total % mass ASTM
D4294
0.3 max 0.006
Sulfur, Mercaptan, %
mass
ASTM
D3227
0.003
max
0.0003
Volatility
Initial boiling point, °C ASTM D86 Report 146.9
10% vol. Recovery, °C ASTM D86 205 max 166.6
20% vol. Recovery, °C ASTM D86 Report 174.1
230
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-239-320.jpg)
![Diesel because of higher pressure build up and load on
the fuel pump. But the engine heats up earlier because
of greater frictional loss.
Brake Specific Fuel Consumption for JP-8 is higher
than diesel fuel when used in diesel engine due to
higher frictional and mechanical loses and lower cetane
number of JP-8.
Analysis show that implementation of SFC is best
suited for supplication in heavy-duty diesel engine
vehicles
REFERENCES
[1] G. Fernandes, J. Fuschetto, Z. Filipi, D. Assanis, and
H. McKee, "Impact of military JP-8 fuel on heavy-duty diesel
engine performance and emissions," Proceedings of the
Institution of Mechanical Engineers, Part D: Journal of
Automobile Engineering, vol. 221, pp. 957-970, 2007.
[2] D. G. Weir, "Strategic Implications for a Single-Fuel
Concept," DTIC Document1996.
[3] A. Command, "JP-8: the single fuel forward: an
information compendium," Warren, MI, 2001.
[4] A. F. Montemayor, L. L. Stavinoha, S. J. Lestz, and M.
E. LePera, "Potential Benefits from the Use of JP-8 Fuel in
Military Ground Equipment," DTIC Document1989.
[5] J. Schmitigal and J. Tebbe, "JP-8 and other military
fuels," DTIC Document2011.
[6] E. Aviation, "World jet fuel specifications with avgas
supplement," Machelen: ExxonMobil Aviation, 2008.
[7] W. E. Likos, E. C. Owens, and S. J. Lestz, "Laboratory
evaluation of Mil-T-83133 JP-8 fuel in army diesel engines,"
DTIC Document1988.
[8] A. Covington, "The investigation of combustion and
emissions of JP8 fuel in an auxiliary power unit," 2011.
[9] J. Lee, H. Oh, and C. Bae, "Combustion process of JP-
8 and fossil Diesel fuel in a heavy duty diesel engine using two-
color thermometry," Fuel, vol. 102, pp. 264-273, 2012.
[10] S. Gowdagiri, X. M. Cesari, M. Huang, and M. A.
Oehlschlaeger, "A diesel engine study of conventional and
alternative diesel and jet fuels: Ignition and emissions
characteristics," Fuel, vol. 136, pp. 253-260, 2014.
[11] D. A. Rothamer and L. Murphy, "Systematic study of
ignition delay for jet fuels and diesel fuel in a heavy-duty diesel
engine," Proceedings of the Combustion Institute, vol. 34, pp.
3021-3029, 2013.
[12] K. Velmurugan and S. Gowthamn, "Effect of
CETANE Improver Additives on Emissions."
[13] J. Lee and C. Bae, "Application of JP-8 in a heavy duty
diesel engine," Fuel, vol. 90, pp. 1762-1770, 2011.
[14] G. Labeckas, S. Slavinskas, and V. Vilutiene, "Effect
of the Cetane Number Improving Additive on Combustion,
Performance, and Emissions of a DI Diesel Engine Operating on
JP-8 Fuel," Journal of Energy Engineering, 2014.
[15] A. Uyumaz, H. Solmaz, E. Yılmaz, H. Yamık, and S.
Polat, "Experimental examination of the effects of military
aviation fuel JP-8 and biodiesel fuel blends on the engine
performance, exhaust emissions and combustion in a direct
injection engine," Fuel Processing Technology, vol. 128, pp.
158-165, 2014.
[16] D. Kouremenos, D. Hountalas, and A. Kouremenos,
"Experimental investigation of the effect of fuel composition on
the formation of pollutants in direct injection diesel engines,"
SAE Technical Paper 0148-7191, 1999.
[17] Y. Kidoguchi, C. Yang, and K. Miwa, "Effects of fuel
properties on combustion and emission characteristics of a
direct-injection diesel engine," SAE Technical Paper 0148-
7191, 2000.
[18] T. J. Callahan, T. W. Ryan, L. G. Dodge, and J. A.
Schwalb, "Effects of fuel properties on diesel spray
characteristics," SAE Technical Paper1987.
[19] M. Lapuerta, J. Hernandez, R. Ballesteros, and A.
Durán, "Composition and size of diesel particulate emissions
from a commercial European engine tested with present and
future fuels," Proceedings of the Institution of Mechanical
Engineers, Part D: Journal of Automobile Engineering, vol.
217, pp. 907-919, 2003.
[20] T. L. Ullman, K. B. Spreen, and R. L. Mason, "Effects
of cetane number, cetane improver, aromatics, and oxygenates
on 1994 heavy-duty diesel engine emissions," SAE Technical
Paper 0148-7191, 1994.
.
235
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-244-320.jpg)
![1
Artificial Intelligence Robot
Muhammad Usman Saleem Muhammad Shaheer Sajid Muhammad Mohid
Collage of earth and atmospheric science LCAS LCAS
University of the Punjab Lahore Pakistan Lahore Pakistan
Lahore Pakistan shaheersajid15@gmail.com mmoohhiidd111@gmail.com
Osman.geomatics@gmail.com
Abstract: A.I.R or Artificial intelligence robot is a machine
which can be operated in two modes: manual and automatic.
This machine is specially designed to suit the needs for both
domestic and industrial purposes. A.I.R can be control and
operated within 10m of radius. A.I.R can detect obstacles
within distance of 1.5m. Then it will search the path without
obstacles. For automatics mode, it has been programed in such
a way, that we can define a line for its path then it will be
follow this path throughout its journey. Through the
concentration of smoke with in its domains, A.I.R can detect
the location where the smoke occurred. Its can transmit videos
to the receiver station through the camera installed on it. This
video can be investigated on a laptop. Manual control can be
done with the help of receiver.
Keywords: Control system, artificial intelligence, security
robots, and robotics.
I.INTRODUCTION
Now-a-days, Robotics has become one of the essential tools
aiming to make human lives much more compatible and
easier. Over the past years, many different mobile robots have
been created which directly affect the scientific community;
such as crawling and legged robot. This paper deals with a
wheel robot [1]. Similarly, A.I.R has specifications which too
aim for the same outcome. The basic functions of A.I.R
include; obstacle detection, path following, line following and
smoke/fire detectors. Multiple mobile robots have advances on
single mobile robot [2]. Hence we make it a multiple mobile
robot. Alongside, our robot has the ability to be controlled
directly by laptop and also by a remote controller. It may also
be viewed by a light weight wireless camera. At this moment,
it is worth mentioning that A.I.R, though small in size, has
been very efficiently to build, also making it a low cost
autonomous robot. This actually helps large companies to buy
these robots and get access to their numerous advantages and
applications, cheaply.
These functions hold immense importance for what we
hoped to achieve. The use of obstacle detectors and line
following functions actually put up a vast frame of use of the
A.I.R. Examples include, factories, Metro Politian [1] and
even for military purposes. The fire detectors and path
detector functions allow the robot to be used as a fire escaping
gadget, public safety, health care management system
respectively [3]. A.I.R can be operated automatically and
manually. Arduino is used to control A.I.R through laptop.
This robot is 60% recycled.
II. OBECTIVES
To build an artificial intelligence robot for the public safety,
road navigation, health care management system, fire
detection, industrial and military purpose. Primary objectives
were to connect this robot to laptop and get on the spot
information regarding to defense purposes.
III. METHODS AND METHODOLOGY
A. Remote Controller
A.I.R can be operated through remote control .It has a receiver
with its paired transmitters, which work on the principle of
radio waves. A transmitter sends radio waves to the receiver
which are then decoded and further processed (see fig 1).
There are four channels of the receiver which are attached to
the motor deriver circuit that derives the motor (see fig.2 & 6).
The radio circuit is a four data transceiver circuit, 2 controls
for each motor: forward and backward. Hence increasing
control over the robot.
B. Laptop Connection
In order to connect this robot to the laptop we have used
Arduino. The transmitter has a simple 4 channel encoder
circuit (see fig.2) that is connected to a development board
called Arduino (see fig. 3). The Arduino is programmed in
such a way to control it via a computer (see fig.4). Hence the
robot is a computer controlled one.
236
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-245-320.jpg)
![7
ACKNOWLEDGEMENT
We would like to acknowledge the contribution of
Residence Officer, Prof. Dr. Sajid Rashid Ahmad,
Principal of Collage of earth and atmospheric science,
University of the Punjab, Lahore Pakistan for kind
contribution for the plagiarism evaluation for this study.
Also we would like to express our deepest gratitude to
Dr.Najam Abbas, Sectary ICASE 2015, Institute of Space
technology, Islamabad Pakistan for his very kind
contribution in publication of this effort to the public.
REFERENCES
[1] R. Chandra Kumar, M. Saddam Khan, D. Kumar, R.
Birua, S. Mondal and M. Kr.Parai, 'OBSTACLE
AVOIDING ROBOT – A PROMISING
ONE', International Journal of Advanced Research
in Electrical, Electronics and Instrumentation
Engineering, vol. 2, no. 4, p. 1430, 2013.
[2] S. Shan, 'Mobile Robots Based Intelligent Fire
Detection and Escaping System', JOURNAL OF
COMPUTERS, vol. 8, no. 5, p. 1298, 2013.
[3] D. Punetha, N. Kumar and V. Mehta, 'Development
and Applications of Line Following Robot Based
Health Care Management System', International
Journal of Advanced Research in Computer
Engineering & Technology, vol. 2, no. 8, p. 2446,
2013
242
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-251-320.jpg)
![An Intelligent Approach for Edge Detection in Noisy
Images using Fuzzy Logic
Izhar1,2
, Fayyaz3
1
Mechanical Engineering Department,
CECOS University of IT and Emerging Sciences,
Peshawar, Pakistan
2
Institute of Mechatronics Engineering,
University of Engineering and Technology,
Peshawar, Pakistan
1
izhar@cecos.edu.pk
R. Muhammad1
and N. Ahmed1
3
Department of Mechanical Engineering,
Pakistan Institute of Engineering and Applied Sciences,
Islamabad, Pakistan
Abstract—Edge detection has widespread applications in the
field of feature extraction, machine learning, computer vision,
pattern recognition, remote sensing and satellite image analysis.
In satellite images it detects various regions of interest for
different applications. Though it has useful applications, but it is
a complicated task and it becomes more arduous when it comes
to noisy images. This paper reports an intelligent approach for
edge detection in noisy images. The proposed edge detection
approach is based on fuzzy logic. The proposed algorithm
employs a 3 x 3 window mask and a fuzzy inference system. The
values acquired from the window mask are subjected to the fuzzy
rules for edge detection. The proposed technique is successfully
tested on noise free as well as on noisy images. The experimental
results are also compared with the reported established edge
detection techniques such as Laplacian of Gaussian (LOG),
Roberts, Prewitt, Sobel and previously proposed fuzzy based
algorithm. The proposed technique when tested on a grayscale
image having 24 dB ‘salt and pepper’ noise, detected 106 false
edge pixels. However, in comparison the reported edge detection
techniques like Sobel, Prewitt, LOG and previously developed a
fuzzy based technique have detected 3678, 4435, 835 and 2852
false edge pixels, respectively. From the experimental results it is
evident that the proposed edge detection technique provides
better solution to the edge detection problems in noisy images.
Index Terms— Edge Detection, Noisy Images, Fuzzy Logic
I. INTRODUCTION
Edges in an image are contours produced as a result of
abrupt or sudden alteration in any of the (multiple)
characteristics at pixel level. These changes could be observed
due to change in texture, color, shade or light absorption. These
characteristics could further lead in estimating the orientation,
size, depth and surface features in an image [1]. Edge detection
has numerous applications in the field of robotics, medical
image analysis, geographical science, pattern recognition, and
military technology [2]-[7] etc. It is often the case that images
embody high frequency noise or irrelevant data which inhibits
the detection of continuous edge points [8] since edge itself is a
composition of high frequency data. The noise produces false
flags, which often mislead the techniques for an edge.
Several algorithms are employed for the detection of edges
in images [9]-[14]. The motivation behind each technique is to
overcome the limitations in previous methodologies. The
conventional techniques like Canny, Sobel, Robert, Prewitt and
LOG [9]-[14] incorporate the use of spatial differential filters
utilising local gradient. These filters recognise an edge as an
abrupt change of grey scale pixel intensities. These techniques
are well established. These techniques process the data in
relatively short time and are computationally efficient.
However, the conventional techniques are susceptible to affect
generated by noise during edge detection.
Jiange and Bunke [15] proposed an approximation of scan
lines method for edge detection. The technique in comparison
to other segmentation techniques produced considerably good
results. Genming and Bouzong [16] proposed a 5x5 window
mask for the detection of edges based on a fixed threshold
level. However, their limitation was its inadaptability to
regions with varying greyscale due to fixed threshold point.
Latest techniques incorporate the use of artificial immune
system, artificial neural networks, genetic algorithms with
particle swarm optimization and ant colony optimization [17]-
[19].
Fuzzy set theory is another technique that is employed for
edge detection [20]-[21]. The method performs logical and
mathematical reasoning based on approximations rather than
crisp values. The technique therefore significantly reduces the
complexity of problems where fixed values cannot be
predicted. Kaur et al. [22] proposed a fuzzy logic based edge
detection technique employing a 2×2 window mask. Sixteen
fuzzy rules were defined for edge detection in the
algorithm .The results for edge detection was appreciable in
noise free images. However when subjected to noisy image the
technique detected the false edges produced due to noise as
well.
To address the short comings of the reported edge detection
techniques, a fuzzy logic based technique is proposed in this
work. The technique employs a 3×3 window mask guided by
fuzzy rule set for edge detection in noisy images. In our
previous work [23], we have used trapezoidal and gaussian
243
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-252-320.jpg)
![membership functions (MFs), however, in this work we have
preferred Pi and generalized bell shape MF to develop an
effective filter that is convenient to apply to both noise free and
noisy images. Moreover, algorithm developed in this work is
tested on noisy images having different noise levels.
II. PROPOSED METHODOLOGY
In Fig. 1 the conceptual framework of the proposed fuzzy
logic based edge detection is shown. The developed edge
detection technique for noisy images is based on fuzzy logic.
To extract the greyscale values of neighborhood pixels from
the input image, a 3x3 window mask is designed. The
grayscale values of the neighborhood pixels acquired from the
mask are pre-processed before applying to the fuzzy inference
system. A fuzzy inference system is designed that takes these
processed values as input. These values are subsequently
converted into the fuzzy plane. A fuzzy rule base is defined
that determine and show the edge pixels in the output image.
The output of the system is computed based on centroid
method and defuzzification is performed based on Mamdani
implication.
Fig. 1. Conceptual framework of the proposed edge detection approach.
A. Window Mask for Scanning the Image
A 3x3 window mask is designed for scanning the image, in
the proposed approach as shown in Fig. 2. The mask takes the
grayscale values, of eight neighbourhood pixels with the
central pixel, as the out pixel.
Before applying to the fuzzy inference system, the
grayscale values obtained from the mask are pre-processed.
Fig. 3 shows the processed mask, where
– for x = 1, 2, 3,…,8.
B. Fuzzy Membership Functions
In fuzzy inference system, MFs play a key role. In the
fuzzy set fuzziness is measured using MFs as they are the key
constituents of the fuzzy set theory. Based on the nature of
problem the type and shape of the MF should carefully be
selected as they have effects on the fuzzy inference system. In
the developed edge detection technique, for the input data, Pi
MFs are used, because they show reasonably improved results
with minimum error [24]. Moreover for the output data,
244
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-253-320.jpg)
![Table I lists the various terminologies and parameters of
both the input and output fuzzy sets.
TABLE I. PARAMETERS AND TERMINOLOGIES OF INPUT AND OUTPUT FUZZY
SETS
Linguistic Variable Parameter Range MF Type
Fuzzy Input ∆P1
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P2
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P3
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P4
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P5
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P6
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P7
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Input ∆P8
Less [0 0 23 47]
[0 255] Pi MF
More [23 47 255 255]
Fuzzy Output G0
Non-Edge [3 2 10]
[0 255]
Generalized
Bell MFEdge [3 2 248]
D. Fuzzy Knowledge Base
Fuzzy knowledge base or rule base in fuzzy inference
system is a set of linguistic descriptions [25]. Fuzzy rule base
plays a key role in fuzzy inference system as it makes
conclusions related to classifying an input or stabilizing and
adjusting the output. Fuzzy rule base for the proposed edge
detection algorithm consists of the following linguistic
descriptions as listed in Table II.
III. RESULTS AND DISCUSSION
The developed edge detection technique is tested on a number
of grayscale images including noise free and noisy images. In
noise free greyscale images the developed technique has
successfully detected all type of edges as shown in Fig. 6. In
Fig. 6 (a) grayscale image of size 185 x 232 pixels having five
different regions covered by four boundary lines is shown. As
shown in Fig. 6 (c), the proposed technique for edge detection
has detected these four boundary lines (edges) successfully.
Similarly, as shown in Fig. 6 (b) the proposed method has
successfully detected edges in the greyscale (flower) image.
TABLE II. FUZZY KNOWLEDGE BASE FOR THE DEVELOPED EDGE DETECTION
TECHNIQUE
Rules Statement
R1
If ∆G1==More && ∆G2==More && ∆G8==Less then G0 ==
Edge
R2
If ∆G1==More && ∆G4==More && ∆G8==Less then G0 ==
Edge
R3
If ∆G2==More && ∆G3==More && ∆G8==Less then G0 ==
Edge
R4
If ∆G4==More && ∆G6==More && ∆G8==Less then G0 ==
Edge
R5
If ∆G1==More && ∆G2==More && ∆G7==Less then G0 ==
Edge
R6
If ∆G1==More && ∆G4==More && ∆G7==Less then G0 ==
Edge
R7
If ∆G2==More && ∆G3==More && ∆G7==Less then G0 ==
Edge
R8
If ∆G4==More && ∆G6==More && ∆G7==Less then G0 ==
Edge
R9
If ∆G1==More && ∆G2==More && ∆G9==Less then G0 ==
Edge
R10
If ∆G1==More && ∆G4==More && ∆G9==Less then G0 ==
Edge
R11
If ∆G2==More && ∆G3==More && ∆G9==Less then G0 ==
Edge
R12
If ∆G4==More && ∆G6==More && ∆G9==Less then G0 ==
Edge
Fig. 6. Tested images: (a) Image having four different regions, (b) Flower, (c)
Edge detection in image having four different regions, and (d) Edge detection
in flower image.
246
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-255-320.jpg)
![The developed edge detection technique has the advantage
of detecting edges in the noisy images as previously discussed
(in the introduction). This is verified by detecting edges in an
image having 24 dB ‘salt and pepper’ noise. To compute the
noise level in an image through peak signal to noise ratio
(PSNR), first based on equation (3) the mean square error
(MSE) is computed [26]-[27].
∑ ∑[ ] 3
Where and represents the input noise free and noisy
images respectively. While and indicates the total number
of rows and columns of the input images respectively. Finally
the expression for the computation of noise level becomes as in
(4).
[ ] 4
Where denotes the maximum possible intensity value of
the pixel in the input image. The value of for eight bit
unsigned integer data type image is 255.
The developed edge detection technique is applied to a size
500 x 232 pixels image having ‘salt and pepper’ noise at a level
of 24 dB. The simulation results are compared with other
conventional and reported edge detection algorithms as shown
in Fig. 7.
Fig. 7. Comparison of experimental results in noisy image, (a) Original image, (b) Noisy image, (c) Sobel edge detection (d) Prewitt edge detection, (e) LoG edge
detection, (f) Previously developed fuzzy based edge detection technique [22], (g) The proposed method.
From the experimental results it is clear that the proposed
fuzzy based edge detection algorithm has detected a very few
false edge pixels in comparison to the other reported edge
detection techniques. The number of false edge pixels detected
by different reported edge detection techniques, when
subjected to images having various noise levels is shown in
Fig. 8. From Fig. 8 it is evident that the developed edge
detection technique when subject to a noisy image of 500 x 232
size and 24 dB noise level has detected 106 false edge pixels,
while other edge detection techniques for instance, Sobel,
Prewitt, LOG, and previously developed fuzzy logic technique,
247
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-256-320.jpg)
![after fine tuning their parameters have detected 3678, 4435,
835 and 2852 false edge pixels respectively.
Fig. 8. False edge detected pixels in a standard image of 500 x 232 pixels: A
comparision.
IV. CONCLUSION AND FUTURE WORK
This paper proposes and demonstrates a fuzzy logic based
edge detection algorithm for noisy images. The developed
technique employs a 3×3 mask guided by fuzzy rule set for
edge detection in noisy images. The developed technique has
successfully detected all the edge pixels in noise free and
noisy images. The developed algorithm is also compared with
other conventional and previously developed fuzzy logic
based edge detection techniques. The developed edge
detection algorithm when subjected to a 500 x 232 size
grayscale image having 24 dB ‘salt and pepper’ noise has
detected a very few false edge pixels (106), while the reported
edge detection techniques like Sobel, Prewitt, LOG and
previously developed fuzzy logic have detected 3678, 4435,
835 and 2852 respectively. From the experimental results it is
obvious that in case of noisy images the proposed technique
provides better results.
In future an investigation on how to incorporate artificial
immune system with fuzzy logic to develop a hybrid
technique for edge detection is under consideration.
REFERENCES
[1] V. Torre and T. A. Poggio, “On edge detection.,” IEEE Trans. Pattern
Anal. Mach. Intell., vol. 8, no. 2, pp. 147–63, Feb. 1986.
[2] G. O. B. Siliciano, L. Sciavicco, L. Villani, “Modelling, Planning and
Control,” in Robotics, Springer, 2010, pp. 415–418.
[3] Sirikan Chucherd, “Edge detection of medical image processing using
vector field analysis,” in 11th International Joint Conference on
Computer Science and Software Engineering (JCSSE), 2014, pp. 58–63.
[4] H. Lin, P. Du, C. Zhao, and N. Shu, “Edge detection method of remote
sensing images based on mathematical morphology of multi-structure
elements,” Chinese Geogr. Sci., vol. 14, no. 3, pp. 263–268, Sep. 2004.
[5] R. Rulaningtyas and K. Ain, “Edge detection for brain tumor pattern
recognition,” in International Conference on Instrumentation,
Communication, Information Technology, and Biomedical Engineering
2009, 2009, pp. 1–3.
[6] A. A. Goshtasby, 2-D and 3-D Image Registration: for Medical, Remote
Sensing, and Industrial Applications. Wiley, 2005, pp. 34–39.
[7] A. Loft, K. E. Jensen, J. Löfgren, S. Daugaard, and M. M. Petersen,
“PET/MRI for Preoperative Planning in Patients with Soft Tissue
Sarcoma: A Technical Report of Two Patients.,” Case Rep. Med., vol.
2013, p. 791078, Jan. 2013.
[8] M. Setayesh, M. Zhang, and M. Johnston, “Effects of static and dynamic
topologies in Particle Swarm Optimisation for edge detection in noisy
images,” in 2012 IEEE Congress on Evolutionary Computation, 2012,
pp. 1–8.
[9] J. Canny, “A Computational Approach to Edge Detection,” IEEE Trans.
Pattern Anal. Mach. Intell., vol. PAMI-8, no. 6, pp. 679–698, Nov.
1986.
[10] C. Y. H. X. Zhang Jin-Yu, “Edge detection of images based on
improved Sobel operator and genetic algorithms,” in 2009 International
Conference on Image Analysis and Signal Processing, 2009, pp. 31–35.
[11] A. Rosenfeld, “The Max Roberts Operator is a Hueckel-Type Edge
Detector,” IEEE Trans. Pattern Anal. Mach. Intell., vol. PAMI-3, no. 1,
pp. 101–103, Jan. 1981.
[12] R. A. Kirsch, “Computer determination of the constituent structure of
biological images.,” Comput. Biomed. Res., vol. 4, no. 3, pp. 315–28,
Jun. 1971.
[13] L. Yang, X. Wu, D. Zhao, H. Li, and J. Zhai, “An improved Prewitt
algorithm for edge detection based on noised image,” in 2011 4th
International Congress on Image and Signal Processing, 2011, pp. 1197–
1200.
[14] F. Ulupinar and G. Medioni, “Refining edges detected by a LoG
operator,” in Proceedings CVPR ’88: The Computer Society Conference
on Computer Vision and Pattern Recognition, 1988, pp. 202–207.
[15] X. Jiang and H. Bunke, “Edge Detection in Range Images Based on
Scan Line Approximation,” Comput. Vis. Image Underst., vol. 73, no. 2,
pp. 183–199, Feb. 1999.
[16] C. Genming and Y. Baozong, “A new edge detector with thinning and
noise resisting abilities,” J. Electron., vol. 6, no. 4, pp. 314–319, Oct.
1989.
[17] A. Naumenko, V. Lukin, and K. Egiazarian, “SAR-image edge detection
using artificial neural network,” in 2012 International Conference on
Mathematical Methods in Electromagnetic Theory, 2012, pp. 508–512.
[18] O. P. Verma and R. Sharma, “An optimal edge detection using universal
law of gravity and ant colony algorithm,” in 2011 World Congress on
Information and Communication Technologies, 2011, pp. 507–511.
[19] M. Setayesh, M. Zhang, and M. Johnston, “Effects of static and dynamic
topologies in Particle Swarm Optimisation for edge detection in noisy
images,” in 2012 IEEE Congress on Evolutionary Computation, 2012.
[20] C.-C. L. Yau-Hwang , Chang-Shing Lee, “A new fuzzy edge detection
method for image enhancement,” in Proceedings of 6th International
Fuzzy Systems Conference, 1997, vol. 2, pp. 1069–1074.
[21] S. E. El-Khamy, M. Lotfy, and N. El-Yamany, “A modified fuzzy Sobel
edge detector,” in Proceedings of the Seventeenth National Radio
Science Conference. 17th NRSC’2000 (IEEE Cat. No.00EX396), 2000.
[22] K. Kaur, V. Mutenja, and I. S. Gill, “Fuzzy Logic Based Image Edge
Detection Algorithm in MATLAB,” Int. J. Comput. Appl., vol. 1, no.
22, pp. 57–60, Feb. 2010.
[23] S. Mandal, J. Choudhury and S. Chaudhuri, “In Search of Suitable
Fuzzy Membership Function in Prediction of Time Series Data,” IJCSI
International Journal of Computer Science Issues, vol. 9, no.3, 2012.
[24] Izhar, I. U. Haq, K. Shah, T. Khan, K. Azam, S. Anwar, “Fuzzy Logic
Based Edge Detection for Noisy Images” Technical Journal, UET
Taxila, vol. 20 (SII), no. II (S), 2015.
[25] O. Karray, and C. De Silva, Soft computingand intelligent systems
design: theory, tools and applications, Pearson Addison Wesely, Essex
CM20 2JE, England, 2004.
[26] F. Luisier, T. Blu and M. Unser “Image Denoising in Mixed Poisson-
Gaussian Noise,” IEEE Transactions on Image Processing, vol. 20, no.
3, pp. 696-708, 2011.
[27] J. Liu, Z. Huan and H. Huang, “Image restoration under mixed noise
using globally convex segmentation,” Journal of Visual Communication
and Image Representation, vol. 22, no.3, pp. 263-270, 2011.
248
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-257-320.jpg)
![A Survey of Active ITU-R P-Series Propagation
Models
Sundus Najib1
, Ibrar Ali Khan2
, Muhammad Waleed
Aftab1
, Naveed Mufti1
1
Department of Telecommunication Engineering
2
Department of Electrical Engineering
University of Engineering and Technology, Peshawar
Peshawar, Pakistan.
sundus.najib@gmail.com
Rafia Mehreen
Department of Telecommunication Engineering
University of Engineering and Technology, Taxila
Taxila, Pakistan
Abstract— ITU-R (International Telecommunication Union)
P-series (Propagation-series) recommendations encapsulate the
propagation models and data vital for modeling of any radio
communication path, including space and terrestrial paths. For a
new researcher, an understanding of these recommendations
prior to their application in various scenarios is desirable. This
paper presents an introduction to propagation, radio channel
impairments and previous review studies, followed by review of
thirty-one recommendations applicable to numerous
indoor/outdoor propagation scenarios. It is observed that some
outdoor propagation models can be used for partial indoor
scenarios while the only indoor model provided by ITU-R series
can also be used for various outdoor setups. Certain
recommendations are interconnected with other
recommendations, developing a need for their comprehensive
study.
Keywords— ITU-R; Radiowave Propagation; Indoor/Outdoor
propagation models; P-Series.
I. INTRODUCTION
The International Telecommunication Union (ITU)’s
Radiocommunication Sector primarily works for global
harmonization of the use of radio frequency spectrum. In
addition to supervising the international radio spectrum
regulation and satellite orbits resources, ITU-R also develops
Radiocommunication standards and recommendations for the
efficient use of spectrum in Radiocommunication systems [1].
Radio spectrum has a vast range of applications in
considerable operational environments. Three main
mechanisms namely refraction, diffraction and scattering
make it possible for the radio waves to propagate beyond the
horizon in terrestrial communication [2]. Certain factors can
have different effects along the path of transmission, resulting
in the loss of information signal. For communication to
succeed, modeling of the radio spectrum/system to overcome
and minimize these channel impairments is vital [3].
Many modeling techniques for radio channel modeling
have been put forth by researchers from time to time for
effective radio communication. ITU-R encourages research
and investigations into different aspects of
radiocommunication by proposing questions (ITU-R
Questions), constituting Study Groups, seeking input from
Member States and Industry and ultimately converting
different inputs into ITU-R Recommendations. It takes years
of research and hard work for developing and modifying each
recommendation, thus making it valid and reliable. The
Recommendations are further classified on the basis of
‘services’ [1]. There are currently seventy-eight active
recommendations dealing with Radiowave Propagation [4].
Each recommendation is valid for specific conditions and
frequency bands. At the time of writing, there is no known
comparison/review of the ITU-R Recommendations pertaining
to propagation. It is felt that, for a new researcher, it becomes
difficult to understand, sort and pin point the most relevant
ITU-R propagation model for his/her modeling/
scenario/investigation. This review paper is aimed at
providing a comparative summary of the propagation models
contained within the ITU-R P-series Recommendations. It is
hoped that this would help the researchers to get an overview
of the propagation models within the P-series and enable them
to select the most relevant recommendation for modeling their
specific scenario.
The order of remaining paper is as follows: Section-II
explains basic propagation mechanisms, followed by radio
channel impairments. Section III discusses previous review
studies while in Section-IV, ITU-R indoor and outdoor radio
propagation models are discussed. Finally Section-V and
Section-VI puts forward the guidelines and conclusions along
with future work of this review paper.
II. PROPAGATION MECHANISMS AND RADIO CHANNEL
IMPAIRMENTS
A. Propagation Mechanisms
1) Refraction:
In radio propagation, the atmospheric bending of the radio
signal away from the straight path is termed as atmospheric
refraction [2]. A drop in temperature and pressure with
increasing altitudes results in decreasing value of refractive
index with height, contributing to bending of the wave
towards the ground. Atmospheric refraction is calculated in
units of ‘N’ [2]:
Where, is the refractive index.
249
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-258-320.jpg)
![Fig. 1. Refraction Types [2]
In order to follow the earth’s curvature, the change in
refractivity with height, termed as vertical refractivity
gradient, denoted by dN/dh, should be equal to -157 N
units/kilometer [2]. When the gradient exceeds -157 N-
units/kilometer, trapping of signals inside ducts can occur. A
less negative refractivity gradient causes signals to be super-
refracted while gradients more positive than -40 N-
units/kilometer can cause sub-refraction. Fig-1 shows different
refractive types and associated values of gradients.
2) Diffraction:
Diffraction occurs due to the obstructions present in the
path of transmitter and receiver antennas [3]. Objects with
sharp irregularities and impenetrable bodies such as buildings,
vegetation, terrain etc. are the main reason for the diffraction
of radio signal. Three common procedures for diffraction
modeling are Bullington method, Epstein-Peterson method
and Deygout (also known as the ‘principle edge method’) [2].
Bullington method is widely used in ITU-R P-Series
Propagation Models for diffraction calculations which is based
on constructing a comparable knife edge at the junction of
transmitter/receiver horizons.
3) Scattering:
When radio signal encounters objects whose size are of the
order or less than that of propagating wave, it gets scattered
[3]. For long range Very High Frequency (VHF)/Ultra High
Frequency (UHF) propagation, tropospheric scattering plays
an important role, where signals are scattered to the receiver
located beyond the horizon. A limit to the propagation range is
impinged by the joint bulk made by the overlap of
transmitter/receiver antenna patterns which needs to be in the
troposphere [2].
B. Radio Channel Impairments
1) Attenuation:
Attenuation can be defined as the gradual degradation of
signal power during its transmission over a radio path [5]. It is
measured in units of decibels (dB). Attenuated power can be
calculated by:
Where,
Ps= Signal power at transmitter (Watt)
Pd=Signal Power at Receiver (Watt)
2) Free Space Loss (FSL):
FSL is assumed for simplification purposes. A signal can
be visualized as spreading out from a transmitter. As it travels
away from the source, it spreads out in the form of a sphere.
The surface area of the sphere increases as distance from
transmitter increases. Following the law of the conservation of
energy, the signal intensity at a point must decrease as the
surface area of the sphere increases [2].
3) Multipath Effects:
Due to the irregularities and non-uniform behavior of the
radio channel along with the straight path in which the signal
is transmitted, signal may reach the receiver by following
other directions. The resultant signal at the receiver is then the
combined effect of these multipath signals which either
interfere constructively or destructively, depending on the path
differences and antenna heights.
4) Fading:
Fading is a random process that may vary with position,
frequency and time. Fading encountered due to multipath
effects can be categorized as multipath fading while the
shadowing of radio signal from different objects cause
shadowing fading.
5) Signal Penetration into Buildings:
Radio signals are attenuated when they come in contact
with a building/hill etc. Shape of the building, frequency of
operation and materials of the building greatly affect the
penetration phenomena. Normally signals are more attenuated
by metals than plywood/concrete. Aluminum acts as a
reflector medium while gypsum and asbestos have higher
signal absorbing capabilities than other materials [6].
III. PREVIOUS REVIEW STUDIES
Sumit Joshi et al [7] and Govind Sati et al [8] present
reviews of some commonly used empirical, stochastic and
deterministic propagation models, namely FS path loss model,
Okumura model, COST 231 Hata model, Stanford University
Interim (SUI) Model, ECC-33 model (suitable for urban
environments) and COST-231 Walfish-Bertoni model for
small height buildings. Furthermore, Sumit Joshi et al [7]
provided a MATLAB simulated comparison of the models
they tested and generated an error factor to be added with
Bertoni’s model to achieve higher fidelity for scenarios under
their investigation. Pooja Rani et al also reviewed above
mentioned models along with two additional propagation
models i.e. Longley-Rice model and Two-Ray Ground model
[9].
A performance comparison of Longley-Rice Model, ITU-
R P.1546 terrestrial Model and HATA-Davidson Model using
Height Above Average Terrain (HAAT) for Digital Video
Broadcasting Television shows that ITU-R P.1546
250
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-259-320.jpg)
![recommendation is only reliable for distances shorter than 50
kilometers while the Longley-Rice model incorporated with
terrain information can lead to better radio modeling
performance but at the cost of computational efficiency [10].
Empirical and deterministic propagation models for terrestrial
communication are reviewed by Magdy F. Iskander et al [11]
and ray-tracing methods for enhanced deterministic models
are discussed thoroughly. Unlike other studies, Tapan
K.Sarkar et al [12] provide a more comprehensive review of
several statistical and site-specific models that shed light on
various scenarios of small- scale/ large-scale fading. The study
proposed a new numerical approach for prediction of
propagation mechanisms. For indoor environment, an advance
ray tracing method is put forward for locating the rays’ paths
while Finite Difference Time Domain (FDTD) method is used
for analyzing the propagation of radio wave through the walls
of building.
All the above referenced comparative studies focus on
similar indoor and outdoor propagation models. No review has
been conducted for discussing only ITU-R propagation models.
This presents an urge to address these models for convenience
of other researchers.
IV. DISCUSSION OF ITU-R INDOOR/OUTDOOR PROPAGATION
MODELS
At the time of writing, ITU-R Propagation Section has in
force seventy eight recommendations, related to indoor
propagation models, outdoor propagation models, and global
maps along with data, attenuation models, diffraction models,
noise model and list of other factors that should be considered
in modeling of radio paths [4]. Our review study focuses on the
Indoor and outdoor models. In this regard, a total of thirty-one
recommendations have been studied and reviewed in this
paper. The recommendations’ numbers along with area of
concern, frequency of operation and path length applicability
are listed in Table-1.
A. Indoor Propagation Model
Unlike many outdoor models, ITU-R has only published a
single Indoor Propagation model, as part of its P.1238
Recommendation [13]. The model, can however, be applied to
partial outdoor scenarios [14]. This model covers the
mechanisms and factors related to the radio propagation of
indoor signals operating in 900 MHz to 100 GHz frequency
range. The Path loss models are divided in two categories,
namely site-general (little site information needed) models and
site specific (detailed information of building) models. The
models assume that the portable user equipment and the base
station are located inside the building premises. Statistical and
deterministic delay spread models are also discussed. The
impact of polarization and antenna radiation patterns is
discussed for three different cases i.e. LOS case, obstructed
LOS case and cases where portable radio terminals are
randomly oriented resulting in the scatter of transmitted
polarization energy into orthogonal polarizations. Dependence
of signal strength and quality on location of the transmitter and
receiver, effect of building’s structure and its materials’
permittivity, and the effects of motion of objects in room are
also thoroughly discussed in this recommendation.
B. Outdoor Propagation Models:
1) Aeronautical and Space Models
P.528 is an aeronautical and satellite communication based
model, calculating basic transmission loss [15] of aeronautical
and satellite services operating in frequency range of 125-
15500 MHz. It uses an interpolation method to determine
basic transmission loss data, from set of curves that are valid
for ground-satellite, ground-air, air-air, satellite-satellite and
air-satellite links. This transmission loss is predicted on the
basis of distance between antennas, the heights of the antennas
above mean sea level, the frequency, and the time percentage.
P.682 provides guidance on Earth -Space aeronautical mobile
telecommunication systems planning by incorporating models
to be used with P.618 [18]. The models therein take account of
effects generated by signal multipath and scattering from the
surface of the earth [16]. The propagation mechanisms
discussed in P.1409 [17] should be taken into consideration
when planning a radio communication system comprising of
stations at high altitudes and other higher level platforms in
stratosphere, and in the studies of sharing and compatibility.
Recommendations P.531 [19], P.618 [18], P.619 [20], P.679
[21], P.680 [22], P.681 [23] and P.840 [24] are developed for
planning space, satellite and earth-satellite systems. P.531
provides data and methods that help in satellite systems
planning by evaluating the ionosphere’s effects on Earth-space
paths at frequency range from 0.1 GHz to 12 GHz. P.618 also
contributes to the planning of earth-space communication
planning. P.619 puts forward model to be used for calculating
interference between earth and space stations.
Recommendation P.679 uses methods from P.618 and
additional data to model broadcast satellite systems. P.680
also uses methods from P.618 and incorporates other data for
designing Earth-space nautical mobile satellite systems. P.681
deals with modeling of land mobile telecommunication
systems between Earth and space. P.840 incorporates the
models, graphs and global maps needed for the calculation of
attenuation caused fog and clouds in the atmosphere.
2) Terrestrial Models
Total of twenty terrestrial propagation models are
discussed in this section. P.452 [25] addresses the microwave
interference and propagation loss encountered between earth
stations operating at frequency range of 0.1 GHz to 50 GHz.
Recommendation P.525 concerns the free space planning and
provides guidance on the attenuation effects that should be
considered in designing a radio link that assumes free space
environment [26]. For evaluating the diffraction effects on the
received signal strength, P.526 recommends a number of
models that are applicable to various obstacle types and
variety of path geometries such as spherical earth surface and
irregular terrain [27]. P.530 [28] comprehensively describes
the propagation effects that should be tackled when planning a
fixed digital line of sight (LOS) link. ELF, VLF and LF
terrestrial radio waves are confined within the space between
the Earth and the ionosphere in order for them to travel to
greater distances. For this purpose, recommendation P.684
proposes methods for developing services in the frequency
range below 150 KHz [29]. P.842 [30] focuses on the issue of
251
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-260-320.jpg)
![reliability and compatibility in planning and design of High
Frequency (HF) radio communication systems. Considering
the fact that scattering from ionization caused by meteor trails
can provide a convenient way of communication at HF and
VHF, Recommendation P.843 [31] tackles with this issue of
meteor-burst propagation. P.844 provides information that
should be considered when designing radio communication
systems that rely on sharing of frequencies within the 30
MHz-3 GHz Range [32]. Planning of broadcasting services
falling in the LF and MF bands i.e. 150,000 Hz to 1700,000
Hz, having path lengths from 50 km to 12000 km, are
discussed in P.1147 [33]. P.1321 [34] shed light on the
characteristics of LF/MF ground-wave and sky-wave
propagation that might cause an effect on the digital
modulation techniques used in those bands. In designing and
planning of terrestrial land mobile and broadcasting services,
propagation characteristics whose impact cannot be ignored in
the proper modeling of these services are mentioned in P.1406
[35]. Recommendation P.1410 [36] addresses radio
communication systems operating in frequency ranges of
3GHz to 60 GHz (broadband services) and provides guidelines
for line-of-sight and alternative non-LOS propagation
mechanisms. P.1411 [14] is a model for short range
propagation in an outdoor scenario over the frequency
spectrum of 300 KHz to 100 GHz but it can be used as good
modeling technique for cases where some indoor factors are
also involved. A method constituting of interpolation and
extrapolation techniques from field strength curves that are
based on empirical data is presented in recommendation
P.1546 which can be applied to all the polarization types [37].
These curves are plotted as a function of path length, effective
height of the antenna, frequency of operation and percentage
time of the year. This method predicts the radio propagation
behavior in point-to-area scenarios for land based services
operational in the frequency band of 30,000 KHz to 300 MHz.
Recommendation P.1791 [38] provides procedures valid for
frequency ranges from 1 GHz to 10 GHz, to calculate path
loss for Ultra-wideband (UWB) indoor and outdoor
environments for both direct and obstructed paths, and
estimating the power received by a conventional narrow-band
receiver from an ultra-wideband transmitter. The method
presented in P.1812 [39] gives a thorough study of terrain for
point to area services which are land based and operate in the
frequencies between 30,000 KHz to 300 MHz and paths
starting from 0.25 kilometer and going up to 3000 km
distance, with both terminals having maximum height of 3 km
above ground. Radio communication can also be achieved by
using optical and infrared spectra. For taking advantage of this
resource, Recommendation P.1814 provides propagation
estimation procedures for terrestrial free-space optical systems
planning. Attenuation caused by the presence of fog in the
atmosphere, rain, snowfall and attenuation of signal during
normal sunny days is also addressed in this recommendation.
It also covers scintillation and impairments by sunlight [40].
P.1816 [41] is an urban/sub-urban propagation model that
guides on the forecasting of time based profiles and profiles of
spatial elements for broadband land mobile services operating
in the frequencies between 700 MHz to 9000 MHz and
distances 0.05 km-3 km. In 2013, ITU-R published a wide-
range terrestrial propagation model, P.2001 [42], that predicts
the Path Loss due to both enhancement of the radio signal and
occurrence of fading mechanism over the range from zero
percent to hundred percent of an averaged year and is
applicable to the frequency range from 30 MHz to 50 GHz.
Finally, P.2040 deals with the issue of impact of building’s
infrastructure on propagation of the radiowaves i.e. materials’
types used in its construction and the shape of building [43].
V. GUIDELINES FOR RESEARCHERS
We recommend following procedure to be adopted for
selecting a proper recommendation prior to modeling a
scenario.
1. ITU presents two categories of recommendations i.e.
terrestrial and space models, each dealing with separate set of
environmental conditions. Thus foremost, it is important that
scenario should be distinguished whether it is a terrestrial or
space dominant environment.
2. The communication type must also be specified
whether it is point-to-point, point to area or indoor
communication scenario.
3. On the basis of working environment, select a
suitable recommendation that covers the range of operational
frequency and distance. It is of crucial importance to consider
frequency limitation because certain parameters in
recommendations are frequency dependent and if not taken
care of, can produce inaccurate results.
4. Shortlisted recommendations should then be assessed
in terms of scope of that recommendation, whether it tackles
with the desired metrics and purpose of study. For example, if
interference measurement is the required analysis then one
should consider interference modeling recommendations.
5. The final selected model should then be read in detail
and desired input parameters be collected.
6. Some recommendations rely on other
recommendations for calculating certain parameters. A
thorough study of referenced models is also needed.
Although recommendations provide step by step procedures
for calculating each parameter, they require intense
mathematical calculations and a good understanding of
complex mathematical problems.
VI. CONCLUSIONS AND FUTURE WORK
The propagation models and relevant procedures contained
within the ITU-R Propagation Series (P-Series)
Recommendations comprehensively address all the possible
radio propagation environments from indoor to outdoor space,
urban/suburban and aeronautical cases. A detailed
understanding of these recommendations prior to their use in
one’s research is needed. Recommendations rely on each other
for the calculation of certain atmospheric and modeling
parameters. Certain short range outdoor models and the sole
indoor model available, can be used for partial indoor/outdoor
scenarios. Also each recommendation covers some specific
252
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-261-320.jpg)
![environment and frequency of operation, limiting its area of
use.
Researchers are encouraged to benefit from the
comparisons and information contained in this review in
making decisions about most suitable/applicable models for
the scenarios under investigation.
The latest ITU-R Propagation Model, P.2001, presents a
wide-range propagation model for general purpose terrestrial
communication. It covers frequency range from 30 MHz to 50
GHz and distances from 3 km to at least 1000 km. Unlike
most of the previous models, the model predicts path loss
effectively in the range from 0% to 100% of an average year.
The model can be compared with other relevant propagation
models for a study on a specific outdoor scenarios.
TABLE I. ITU INDOOR/OUTDOOR RECOMMENDATIONS’ SPECIFICATIONS [1,44]
S.No. Recommendation
Number
Model type Scope Frequency of
operation
Distance
1 P.452 Terrestrial Interference prediction and path loss estimation 0.1 GHz-50
GHz
Up to radio
horizon and
beyond
2 P.525 Terrestrial FS attenuation calculation ------- -------
3 P.526 Terrestrial Diffraction calculation ------ -------
4 P.528 Aeronautical/ satellite
services
Transmission loss calculation/ path loss 125 MHz-
15500 MHz
0 km- 1800 km
5 P.530 Terrestrial Path loss and diversity enhancement in Digital fixed LOS
links planning
150 MHz- 100
GHz
200 km
6 P.531 Satellite System Ionospheric propagation 0.1 GHz- 12
GHz
-------
7 P.618 Earth-space model Estimation of path loss, amount of diversity gain and cross
polarization discrimination (XPD) in Telecommunication
services
1 GHz- 55 GHz All possible
orbital heights
8 P.619 Earth-space model Interference calculation --------- -------
9 P.679 Satellite Effects of surrounding environment and path loss
prediction in Broadcasting satellite systems
0.5 GHz- 5.1
GHz
All possible
orbital heights
10 P.680 Satellite Fade/fade duration and interference calculation for
Maritime mobile telecommunication
0.8 GHz- 8
GHz
All possible
orbital heights
11 P.681 Satellite Fade and path loss estimation in Land-mobile
telecommunication
0.8 GHz- 20
GHz
All possible
orbital heights
12 P.682 Aeronautical Fade and multipath calculation in Aeronautical mobile
telecommunication systems
1 GHz- 2 GHz
(sea), 1 GHz- 3
GHz (ground)
All possible
orbital heights
13 P.684 Terrestrial ELF, VLF and LF terrestrial radio services 30 kHz- 150
kHz
0 km- 16000 km
14 P.840 Earth-space model Attenuation due to clouds and fog >10 GHz -------
15 P.842 Terrestrial Reliability/compatibility prediction ---------- -------
16 P.843 Terrestrial Transmission loss, meteor flux variations, antenna and
frequency considerations in Meteor-burst communication
30 MHz-100
MHz
100 km- 1000
km
17 P.844 Terrestrial Ionospheric communication 30 MHz-3 GHz -------
18 P.1147 Terrestrial Estimating sky-wave field strength in Broadcast services 150 kHz-1700
kHz
50 km- 12000
km
19 P.1238 Terrestrial Amount of path loss and delay spread in case of Indoor
propagation
900 MHz -
100 GHz
Buildings,
interiors
20 P.1321 Terrestrial Digital modulation ----------- -------
21 P.1406 Terrestrial Land mobile and broadcasting services ------------ -------
22 P.1409 Aeronautical High altitude radio communication systems ---------- -------
23 P.1410 Terrestrial Coverage and reduction in coverage due to rain in
Broadband radio access
3 GHz- 60 GHz 0 km- 5 km
24 P.1411 Terrestrial Path loss and delay spread in Out-door short range
propagation
300MHz- 100
GHz
Less than 1 km
25 P.1546 Terrestrial Field strength measurement in Point-to-area propagation 30 MHz- 3000
MHz
1 km- 1000 km
26 P.1791 Terrestrial UWB path loss for indoor/outdoor cases 1 GHz- 10 GHz -------
27 P.1812 Terrestrial Field strength calculation in Point-to-area propagation 30 MHz- 3000
MHz
Up to radio
horizon and
beyond
28 P.1814 Terrestrial Calculating beam spreading, absorption and scattering loss
and scintillation in FSO radio communication systems
20 THz- 375
THz
No limit
29 P.1816 Terrestrial Urban/suburban planning 0.7 GHz-9 GHz -------
30 P.2001 Terrestrial Path loss due to signal enhancement and fading 30 MHz- 50
GHz
3 km- 1000 km
31 P.2040 Terrestrial Building materials’ properties and structures ---------- -------
253
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-262-320.jpg)
![REFERENCES
[1] “International Telecommunication Union”, (ITU), www.itu.int,
2/2/2015, <http://www.itu.int/en/Pages/default.aspx>
[2] “Propagation Tutorial”, (Mike Willis), 4/2/2015, <http://www.mike-
willis.com/Tutorial/propagation .html >
[3] Andrea Goldsmith, “Path Loss and Shadowing,” in Wireless
Communications, Cambridge, United Kingdom: CUP, 2005.
[4] “Radiowave propagation”, (ITU 2008), Accessed 2/2/2015,<
http://www.itu.int/rec/R-REC-P/en>
[5] L. K. Bandyopadhyay, S. K. Chaulya, and P. K. Mishra, “Mines
Communication Technique,” in Wireless Communication in
Underground Mines: RFID-based Sensor Networking, Springer
Science & Business Media, 2009.
[6] Masri, T., Chew, S.P., Wong, C.P., and Lias, K., “A study of signal
penetration into building materials,” IEEE 1st International
Conference on Computers, Communications, & Signal Processing,
2005. CCSP 2005, Pages: 21-24, Kuala Lumpur, Malaysia.
[7] Sumit Joshi, and Vishal Gupta, “A review on empirical data
collection and analysis of Bertoni’s model at 1.8 GHz,” International
Journal of Computer Applications (IJCA), 2012, Volume 56 -
Number 6, pages: 17-23, October 2012.
[8] Govind Sati, and Sonika Singh, “A review on outdoor propagation
models in radio communication,” International Journal of Computer
Engineering and Science (IJCES), Volume: 4, Issue: 2, pages: 64-68,
October 2014.
[9] Pooja Rani, Vinit Chauhan, Sudhir Kumar, and Dinesh Sharma, “A
review on wireless propagation models,” International Journal of
Engineering and Innovative Technology (IJEIT), Volume: 3, Issue:
11, May 2014.
[10] Stylianos Kasampalis, Pavlos I. Lazaridis, Zaharias D. Zaharis,
Aristotelis Bizopoulos, Spyridon Zettas, and John Cosmas,
“Comparison of Longley-Rice, ITU-R P.1546 and Hata-Davidson
propagation models for DVB-T coverage prediction,” 2014 IEEE
International Symposium on Broadband Multimedia Systems and
Broadcasting (BMSB), pages: 1-4, Beijing, China, June 2014.
[11] Magdy F. Iskander, and Zhengqing Yun, “Propagation Prediction
Models for Wireless Communication Systems,” IEEE Transactions
on Microwave Theory and Techniques, VOL. 50, NO. 3, pages: 662 –
673, March 2002.
[12] Tapan K.Sarkar, Zhong Ji, Kyungjung Kim, Abdellatif Medouri, and
Magdalena Salazar-Palma, “A survey of various propagation models
for mobile communication,” IEEE Antennas and Propagation
Magazine, Volume: 45, Issue: 03, pages: 239 – 307, June 2003.
[13] ITU-R P.1238-5 (2007), “Propagation data and prediction methods
for the planning of indoor radio communication system and radio
local area networks in the frequency range 900 MHz to 100 GHz”,
ITU-R Recommendation
[14] ITU-R P.1411-7 (2013) ,”Propagation data and prediction methods
for the planning of short-range outdoor Radiocommunication systems
and radio local area networks in the frequency range 300 MHz to 100
GHz”, ITU-R Recommendation
[15] ITU-R P.528-3 (2012),”Propagation curves for aeronautical mobile
and radio navigation services using the VHF, UHF and SHF bands”,
ITU-R Recommendation
[16] ITU-R P.682-3 (2012),” Propagation data required for the design of
Earth-space aeronautical mobile telecommunication systems ”, ITU-
R Recommendation
[17] ITU-R P.1409-1(2012),”Propagation data and prediction methods for
systems using high altitude platform stations and other elevated
stations in the stratosphere at frequencies greater than about 1 GHz”,
ITU-R Recommendation
[18] ITU-R P.618-11(2013),”Propagation data and prediction methods
required for the design of Earth-space telecommunication systems”,
ITU-R Recommendation
[19] ITU-R P.531-12(2013),”Ionospheric propagation data and prediction
methods required for the design of satellite services and systems”,
ITU-R Recommendation
[20] ITU-R P.619-1(1992),”Propagation data required for the evaluation
of interference between stations in space and those on the surface of
the Earth”, ITU-R Recommendation
[21] ITU-R P.679-3(2001),”Propagation data required for the design of
broadcasting-satellite systems”, ITU-R Recommendation
[22] ITU-R P.680-3(1999),”Propagation data required for the design of
Earth-space maritime mobile telecommunication systems”, ITU-R
Recommendation
[23] ITU-R P.681-3(1997),” Propagation data required for the design of
Earth-space maritime mobile telecommunication systems”, ITU-R
Recommendation
[24] ITU-R P.840-6(2013),”Attenuation due to clouds and fog”, ITU-R
Recommendation
[25] ITU-R P.452.15(2013),”Prediction procedure for the evaluation of
interference between stations on the surface of the Earth at the
frequencies above about 0.1 GHz”, ITU-R Recommendation
[26] ITU-R P.525-2 (1994),” Calculation of free-space attenuation”, ITU-
R Recommendation
[27] ITU-R P.526-13(2013),”Propagation by diffraction”, ITU-R
Recommendation
[28] ITU-R P.530-15(2013),” Propagation data and prediction methods
required for the design of terrestrial line-of-sight systems”, ITU-R
Recommendation
[29] ITU-R P.684-6(2012),”Prediction of field at the frequencies below
about 150 kHz”, ITU-R Recommendation
[30] ITU-R P.842-5(2013),”Computation of reliability and compatibility
of HF radio systems”, ITU-R Recommendation
[31] ITU-R P.843-1(1997),”Communication by meteor-burst
propagation”, ITU-R Recommendation
[32] ITU-R P.844-1(94),”Ionospheric factors affecting frequency sharing
in the VHF and UHF bands (30 MHz-3 GHz)”, ITU-R
Recommendation
[33] ITU-R P.1147-2(2003),”Prediction of sky-wave field strength at
frequencies between about 150 and 1700 kHz”, ITU-R
Recommendation
[34] ITU-R P.1321(2013),”Propagation factors affecting systems using
digital modulation techniques at LF and MF”, ITU-R
Recommendation
[35] ITU-R P.1406-1(2007),” Propagation effects relating to terrestrial
land mobile and broadcasting services in the VHF and UHF bands”,
ITU-R Recommendation
[36] ITU-R P.1410-5 (2012),”Propagation data and prediction methods
required for the design of terrestrial broadband radio access systems
operating in a frequency range from 3 to 60 GHz”, ITU-R
Recommendation
[37] ITU-R P.1546-5 (2013) ,”Method for point-to-area predictions for
terrestrial services in the frequency range 30 MHz to 3 000 MHz”,
ITU-R Recommendation
[38] ITU-R P.1791-0 (2007),” Propagation prediction methods for
assessment of the impact of ultra-wideband devices “, ITU-R
Recommendation
[39] ITU-R P.1812-3(2013),” A path-specific propagation prediction
method for point-to-area terrestrial services in the VHF and UHF
bands”, ITU-R Recommendation.
[40] ITU-R P.1814-0 (2007) ,” Prediction methods required for the design
of terrestrial free-space optical links”, ITU-R Recommendation
[41] ITU-R P.1816-2 (2013),” The prediction of the time and the spatial
profile for broadband land mobile services using UHF and SHF
bands”, ITU-R Recommendation
[42] ITU-R P. 2001-1 (2013),” A general purpose wide-range terrestrial
propagation model in the frequency range 30 MHz to 50 GHz”, ITU-
R Recommendation
[43] ITU-R P.2040 (2013),” Effects of building materials and structures on
radiowave propagation above about 100 MHz”, ITU-R
Recommendation
[44] ITU-R P.1144-6 (2012), “Guide to the application of the propagation
methods of Radiocommunication Study Group 3”, Recommendation.
254
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-263-320.jpg)
![Issues of Alpha Mechanism in Subsonic Wind Tunnel
Testing for a Hybrid Buoyant Aircraft
Experimental Fluid Dynamics
A.U.Haque*
, W. Asrar, E. Sulaeman, J.S. Mohamed Ali and A.B. Embong
IIUM-LSWT, International Islamic University Malaysia (IIUM)
Kuala Lumpur, Malaysia
*
anwar.haque@live.iium.edu.my
Abstract— In subsonic wind tunnels with large test sections,
scaled down models are attached to the balance with the help of
model support system. Such a system mainly includes a main
strut and an auxiliary strut, also known as pitch rod. Flow
measurements of such a system are required to be subtracted
from the raw data of the wind tunnel. For this purpose, it is
important to find the contribution of model support system
separately. Potential issues and limitations related to such testing
are discussed in the light of technological constraints. Special
focus is given on the streamlined profile of main strut and wind
tunnel testing of a hybrid buoyant aircraft. In this paper we
experimentally simulate a situation in which height of both these
struts are kept constant. Wind tunnels tests are conducted to find
the aerodynamic and stability characteristics at constant angle of
attack as well as yaw angle. An asymmetric pattern is found in
the pitch and yaw stability for a defined array of beta sweep.
Such a support system not contributes towards drag but also in
lift as well as in side force.
Index Terms— Wind Tunnel Testing, Pitch Rod,
Aerodynamic and stability measurements, Data acquisition and
reduction system, Strut Arrangement
I. INTRODUCTION
It is well known that the experimental data obtained from a
wind tunnel is subject to different corrections. Correction due
to model support group is among one of important correction
for which wind tunnel is run for model off and strut on
conditions. Struts are usually geometrically tapered with less
diameter and chord at the tip than that at the root which is
attached with the balance. Moreover, they are of cylindrical or
streamlined shapes. Such struts have more frontal area and
have complex interaction of the flow when scaled down model
is attached at the tip of it. One of the prospective solutions is to
have tapered struts of profile similar to a diamond shaped
airfoil. Diamond shaped struts are usually used in supersonic
[1-3] and hypersonic wind tunnels [4-5] in which there is
requirement to hold the model from the outer profile and no
interaction of the strut with the base of the model. However, to
the authors’ best knowledge, its utilization in subsonic speed
has hardly been seen in the literature. Most of the earlier work
done is to compare the effects of different arrangement of
streamlines, round or cylindrical struts with pitch rod at fixed
position of angle of attack [6-7]. But such a method is not
quick measuring method as the individual has to change the
position of pitch rod at every required angle of attack and
hence forming different “V” which require additional fillets to
be installed for all angular setting of the struts tested
individually [6]. Also, in comparison with the main strut the
dimensions especially the diameter and the thickness of the
pitch rod is usually quite less. One end of the pitch rod is
usually connected with the gear system of alpha mechanism
which allows the rod to swing in a plane perpendicular to the
turntable axis. If the tip of the pitch rod exposed to the free
stream velocity is not physically attached with the main strut
than there might be some effect of flow induced vibration on
the overall results.
The drag of a model mounted only on main strut in a wind
tunnel is always “more than the individual sum of the isolated
strut and isolated body”[8]. Isolated strut are used when model
is either fixed at certain position or the operator manually
change the position with the help of instruments like
inclinometer. Also it is usually used when the geometric
dimensions of the scaled down model quite small to have
auxuilary strut to be position at the aft part of the model.
Examples of few models tested earlier on single strut are
shown in Fig. 1.
It is important to take separate measurement of the strut
arrangement to get a more general picture of the estimated
parameters of main strut with axillary strut for defined range of
side-slip angle and angle of attack. Such a strut arrangement
can cause complex interference flow fields including the
increase in drag and pitching moment. This work aims to
quantify the magnitude of such effects and to give a first and
knowledge of general arrangement of the experimental setup
for fixed height of the pitch rod.
Due to more interference and drag due to strut arranged in
tandem [9] form and further discussion on it or other two or
three strut arrangement is beyond the scope of present work.
In early times, the manufacturing and assembly techniques
for wind tunnel models are not so sophisticated as that of the
present time. Moreover, the size of the test section is also
small. Such technological constrains limits the overall length of
the model and pitch rods has to positioned at some inclined
angle to attached its tip with the base of the model. In order to
255
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-264-320.jpg)
![define any array of input angle of attack; the pitch rod has to
move in a plane perpendicular to the tunnel’s floor.
(a) Delta wing model
(b) Airship Model
(c) Hybrid lifting hull model
Fig. 1. Few models tested in IIUM-LSWT
Now days, through precise machining, scaled down models
in large dimensions can be manufactured by using combination
of metal and composite. For such models, the height of
auxiliary strut increases for negative angle of attack and
decreases for positive angle of attack. In this way there will be
multiple numbers of cases to estimate the contribution of
model support only. One of the optimum solutions to keep
minimum number of blow downs is to keep the height of the
pitch rod equal to that of the main strut. Complete details of
this prospective solution are discussed below with the help of
an example of on-going wind tunnel testing program of a
hybrid buoyant aircraft. Such aircraft are powered by fuel and
has the ability to generate certain amount of aerodynamic lift
for the weight which is not balanced by the aerostatic lift.
II. PROSPECTIVE SOLUTION
In the case of aircraft the operational angle of attack is
usually limited to 16-18 degree on the positive side and -6
degree on the negative side. Hence the cases for experimental
investigation for aerodynamic and stability characteristics
through wind tunnel testing will be more for the positive one.
In order to define a positive angle of attack, the alpha
mechanism pushes the pitch rod in the downward direction. By
doing so, the tail of the aircraft moves towards the tunnel floor
and it is vice versa for the situation where negative angle of
attack is required. In comparison with conventional aircraft,
hybrid buoyant aircraft may need to fly at altitude less than that
of pressure height [10]. Although a lifting fuselage can provide
additional lift and hence high lift to drag ratio for the cruise and
take-off segment of the flight but there may exist a scenario in
which a pilot may need to give negative angle of attack [11]. In
order to simulate such conditions in wind tunnel, there is
always a requirement to do more testing at negative angle of
attack for which the pitch rod has to move up to attain a height
equal to that of the main strut. For data reduction; all the
aerodynamic and stability parameters has to non-
dimensionalized by reference dimensions of the wing. All such
dimensions, obtained from the CAD model (actual as well as
scaled down) are tabulated below in Tab. 1 for quick reference.
A schematic view of the model attached with the diamond
shaped strut and pitch rod attached at its base is shown in Fig.
2. It is important to note that the reference area value is
obtained through an iterative analysis work for which lift from
the fuselage is added into the main wing by using an equivalent
wing with an airfoil as that of the wing [12].
TABLE. 1 REFERENCE DIMENSIONS OF HB AIRCRAFT
Wing
Area Span Chord
m2
m m
Origional 27.5 20 1.5
Scaled down 0.086 1 0.075
It is pertinent to highlight that the scope of present research
work of model support system is not only limited to the
aerospace application like that of hybrid buoyant aircraft,
discussed above. But can also be applied to any body in which
there is requirement to evaluate the effect of defining the body
at some angle to the free stream velocity. Estimation of
256
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-265-320.jpg)
![parachute’s drag [13], bird’s aerodynamics [8,14], car’s
aerodynamics [15-16], study of lift generated wakes [17] are
few examples of it.
Fig. 2. Isometric view of HB aircraft installed in test section
(a) Side view
(b) Front View
Fig. 3. Positive incidence of the model with the help of pitch rod
For such cases, it is simply replacing the terms of reference
area for aerodynamic forces and reference dimension for the
moments. For example the non-dimensional quantity F (where
F is an aerodynamic force) can be expressed in the non-
dimensionalized form, Eq. (1). In this relationship, instead of
taking the actual exposed area of the body, the platform area of
the wing is usually used as S.
SV
F
CF 2
2
1
(1)
Where, CF F,V, S and is coefficient of aerodynamic force,
aerodynamic force, free stream velocity, reference area and
density of air respectively. Similarly, the aerodynamic
moments also can be expressed in the form of non-dimensional
coefficients. Since a moment is the product of a force and a
length it follows that a non-dimensional form for a moment is
SlV
Q
CQ 2
2
1
(2)
Where, Q is any aerodynamic moment about center of gravity
and l is a reference length.
III. EXPERIMENTAL SETUP
In order to find the aerodynamic and static stability
behavior of model support system discussed in the previous
section, tests are conducted in IIUM-LSWT at 40 m/s. IIUM-
LSWT is a closed-loop wind tunnel with a test section of
dimensions 1.5m × 2.3m × 6m and maximum speed of 50 m/s
[18]. A pictorial view of the test section is shown in Fig. 4(a)
[19]. In this tunnel, the balance section is placed on the floor to
have some height for the external balance. Both the struts
(main as well as auxiliary i.e. pitch rod for alpha mechanism)
are attached with the balance. The same can be observed from
Fig. 4(b) in yellow colour. This wind tunnel is currently one of
the largest wind tunnels in Malaysia and has excellent flow.
The limitations of this experimental setup and its
repurcation of the data obtained from DARCS system of the
tunnel are as follows:
a. It is mandatory to have the pitch rod for defining angle of
attack range. Consequently, in order to quantify their
aerodynamic and stability effects; no comparisons are made
between results from diamond shaped strut with and
without the pitch rod strut.
b. No tests are performed with the model attached to the
defined model support system. Therefore the investigation
reported in this paper is limited to the estimation of struts
along contribution and interference arising from
combination of model-strut arrangement is beyond the
scope of presented work.
c. No tests for angle of attack are possible as the model
support system is rigidity fixed with the balance.
IV. EXPERIMENTAL SETUP
In order to evaluate the effectiveness of the proposed
engineering method, Fig. (5); wind tunnel blow downs are
carried out at different velocities. All the results are plotted in
the form of graph to get first-hand knowledge about the general
trends of aerodynamic and stability characteristics of model
support system. First the wind tunnel is run for a series of blow
257
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-266-320.jpg)
![coefficient of yaw and side force are negative, whereas the sign
of pitching moment is positive. For β sweep test; the model
strut assembly remained rested on tunnel turntable and rotated
at -20o
to +20o
range with 50
increments. As more attention is
given for defining the problem description than the flow
conditions. Hence, this technical note does not attempt to
predict the aerodynamic and stability coefficients for beta
sweep at different free stream velocities. A drastic increase in
side force is observed at α = 150
, Fig. 7(b). This figure also
shows a continuous increase in CL for defined range of and a
slight droppage in value of CD for β = ±100
. However, after β =
±100
, CD increases linearly till β = ±200
.
An asymmetric pattern is observed in overall trends of pitch
and yaw moment, Fig. 8(a). Same thing is reflected in
corresponding coefficient plots as well i.e. mC vs and nC vs
plots, respectively. However, general behavior is more
sinusoidal for the plot of nC vs . Fig. 8(c). Value of mC first
increases till for β = 50
but a drastic decrease in its value is
observed till β = 150
, Fig. 8(b). But its magnitude is not
consistent with that of its value at β = 150
. Values of nC for .
±200
are almost constant but a sudden increase in its value is
found at β = 50
. Since the overall behavior of these two
stability coefficients is non-linear. Therefore, it is quite
difficult to comment on the overall slopes of mC and nC .
-5
15
35
55
75
95
115
-20 -15 -10 -5 0 5 10 15 20
Moment(N.m)
(degree)
Pitch Moment
Yaw Moment
(a) Plots of pitch and yaw moment due to sweep
0
0.2
0.4
0.6
0.8
1
1.2
-20 -15 -10 -5 0 5 10 15 20
Cm
β (degree)
(b) mC vs
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0.025
-20 -15 -10 -5 0 5 10 15 20
Cn
β (degree)
(c) nC vs
Fig. 8 Stability characteristics of model support system
Sometimes, the aerodynamic force of the body alone is defined
as the difference between the aerodynamic force of the body
attached with the strut and results of strut alone testing.
However, this is not the true representation as there is always
some effect of the interference due to model support system as
well. For example, the complete analytical relationship of the
drag force is expressed as follows [8]:
ISBm DDDD (1)
Where, Dm, Ds and DI are drag of the body, strut and due to
interference effects respectively. Drag due to tare is usually
automatically subtracted in online calculations by the DARCS
and hence it is not account far in the above equation. DI would
take out any interference effects due to supports and can be
estimated in future by subtracting the model-on wind-on
condition from the model off-wind on condition. This is a
shortcut (not perfect) to avoid lengthy procedure involving
measurements with inverted model [20].
V. CONCLUSION
For measurements of aerodynamic and stability coefficients
of only model support system placed normal to the flow, a
method is used where both the struts are at same height and
attached through a thin wire. For defined range of velocity, all
the aerodynamic and stability coefficients remain almost
constant for α = 00
. Diamond shaped profile of the mains strut
acts as a symmetric airfoil, which result in significant
contribution towards the side force. For side force rest of the
forces and moments, the overall trend is not symmetric.
Specially, an oscillating behaviour is observed in the trend of
nC vs . Such effects of the measured aerodynamic and
stability characteristics caused by main strut and the pitch rod
are of considerable magnitude and should be excluded from the
test results of complete configuration. Although the presented
results of proposed experimental setup furnish some interesting
finding but its interference with the model deserved much more
attention.
259
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-268-320.jpg)
![ACKNOWLEDGMENT
The support of the Ministry of Science, Technology and
Innovation (MOSTI), Malaysia, under the grant 06-01-08-
SF0189 is gratefully acknowledged. Authors are thankful to
Dr. Fadhil Hasim and his team for providing the diamond
shaped strut for the experimental work.
REFERENCES
[1] D., Biermann, & W. H. Herrnstein, The interference between
struts in various combinations. NASA Report No. NASA-L-
468, 1933
[2] E. P., Jones, S. M., Townsend, Y., Guy, & T. E. McLaughlin,
Reduction of the wave drag of a blunt body by means of a
standoff spike. AIAA Aerosp. Sei. Meet., 38th, Reno, NV,
2000
[3] S. Feldman, Some shock tube experiments on the chemical
kinetics of air at high temperatures. Journal of Fluid
Mechanics, 3(03), 225-242, 1957.
[4] K. L., Mikkelsen, & D. F. Long, Refurbishment, Calibration,
and Initial Testing in the ASE Hypersonic Wind Tunnel. In
43rd AIAA Aerospace Sciences Meeting and Exhibit, 2005.
[5] M. H. Bertram, (1950). Investigation of the pressure-ratio
requirements of the Langley 11-inch hypersonic tunnel with a
variable-geometry diffuser, 1950.
[6] D., Biermann, & W. H. Herrnstein, The interference between
struts in various combinations. NASA Report No. NASA-L-
468, 1933
[7] V., Nguyen, Y., Drolet, & Watt, G. (1995, January).
Interference of various support strut configurations in wind
tunnel tests on a model submarine. In 33rd Aerospace
Sciences Meeting and Exhibit (p. 443).
[8] V.A, Tucker, "Measuring aerodynamic interference drag
between a bird body and the mounting strut of a drag
balance." The Journal of Experimental Biology 154.1 (1990):
439-461.
[9] R. C. Pankhurst, (1958). Wind-Tunnel Interference Due to
Model Supports. Aircraft Engineering and Aerospace
Technology, 30(2), 55-55.
[10] A.U., Haque, W. Asrar, A.A, Omar, E., Sulaeman and J.S,
Ali, "Assessment of Engine’s Power Budget for Hydrogen
Powered Hybrid Buoyant Aircraft", Journal of Propulsion
and Power Research (in press), 2015
[11] A.U., Haque, W. Asrar, A.A, Omar, E., Sulaeman and J.S,
Ali, "Aerostatic and Aerodynamic Modules of a Hybrid
Buoyant Aircraft: an Analytical Approach", IIUM
Engineering Journal, IIUM Engineering Journal, Vol 16, No.
1 (2015).
[12] A.U., Haque, W. Asrar, A.A, Omar, E., Sulaeman and J.S,
Ali, “Power-off static stability analysis of a clean
configuration of a hybrid buoyant aircraft”, 7th Ankara
Internatıonal aerospace conference, 11-13 September 2015 -
METU, Ankara Turkey.
[13] J. A., Weiberg, & K. W. Mort, Wind-tunnel tests of a series
of parachutes designed for controllable gliding flight.
National Aeronautics and Space Administration, 1967
[14] G. R.,Spedding, M.Rosén, & A. Hedenström, A family of
vortex wakes generated by a thrush nightingale in free flight
in a wind tunnel over its entire natural range of flight speeds.
Journal of Experimental Biology, 206(14), 2313-2344, 2013
[15] A. Cogotti, Ground effect simulation for full-scale cars in the
pininfarina wind tunnel (No. 950996). SAE Technical
Paper.1995
[16] B. Hetherington, and D. B. Sims-Williams, Support Strut
Interference Effects on Passenger and Racing Car Wind
Tunnel Models. No. 2006-01-0565. SAE Technical Paper,
2006.
[17] J., Rossow, J. N., Sacco, P. A., Askins, L. S., Bisbee and S.
M Smith, Wind-tunnel measurements of hazard posed by lift-
generated wakes. Journal of aircraft, 32(2), 278-284, 1995.
[18] F. Hasim, R. Rusyadi, W. I. Surya, W., Asrar, A. A.Omar, ,
J., Syed Mohamed Ali, & R. Kafafy, The IIUM low speed
wind tunnel. Proceedings of EnCon2008 2nd Engineering
Conference on Sustainable Engineering Infrastructures
Development & Management, December 18-19, 2008,
Kuching, Sarawak, Malaysia
[19] User’s Manual 6-Component External Balance for
International Islamic University Malaysia Wind Tunnel, S &
V Teknik Sdn. Bhd and Aero Engineering, 2015.
[20] J. Barlow, W. Rae, A. Pope, Low Speed Wind Tunnel
Testing, Jon Wiley & Sons, 1999.
260
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-269-320.jpg)
![Fairing Separation Dynamic Analysis Using
Analytical Approach
Muhammad Tanveer Iqbal
Institute of Space Technology
Islamabad, Pakistan
mrtanveeriqbal@gmail.com
Dr. Abdul Majid
Institute of Space Technology
Islamabad, Pakistan
majid363@yahoo.com
Abstract— Fairing is the upper portion of launch vehicle,
usually made in two halves. It is used to protect the satellite from
loadings (thermal, aerodynamic and acoustic) of harsh
environment during ascent. Separation of fairing at particular
altitude is most critical part of launch vehicle flight which
dictates the success or failure of the whole mission.
This paper presents the designing of spring for separation
system of fairing using 1dof dynamic equations. Material
properties of base ring and spring are considered and integrated
in dynamic system of equations. These equations are modeled
and evaluated. Effects on separation time due to variations in
spring stiffness and compressed length are analyzed. After
analysis , best spring parameters are selected on the basis of
fairing separation requirements. Developed analytical method is
also validated through professional fairing separation software.
Keywords—satellite launch vehicle; fairing jettison; heat
shield; dynamic analysis; analytical approach; separation springs
I. INTRODUCTION
Fairing is designed to shield the satellite from acoustic,
aerodynamic and especially thermal loads. To minimize these
loadings inside the fairing, honeycomb and carbon fiber
sandwich structure is frequently employed. Separation of
fairing is typically done at altitude where free molecular heat
flux is less than 1135W/m2
to avoid damages to satellite [1-3].
Bending load and dynamic pressure are not considered for
fairing separation as their utmost values are achieved prior in
the flight [4].
Successful mission is heavily depends on fairing separation.
Many missions were failed just due to no or late separation of
fairing. In the March 4, 2011, Glory climate satellite launched
by Taurus launch vehicle was failed to achieve specific orbital
altitude due to failure of fairing separation and collapsed back
to earth [5].
Two halves of fairing are separated by using pyro-bolts and
gas/spring thrusters. In the given case, spring thrusters are
selected for designing jettison mechanism. Two main
considerations for designing separation mechanism are that the
magnitude of force produced by spring system must be enough
to separate the fairing's halves in given time at very high
acceleration and after separation, there should be no collision
of opened halves with the launch vehicle [6]. Both are directly
related to spring stiffness. If spring stiffness is low, sufficient
force would not be produced to open the fairing halves in given
time. The high spring stiffness will produce the high angular
velocity in fairing halves and cause its collision with launch
vehicle. Therefore, optimized value of stiffness must be
evaluated.
In this paper, along with stiffness, other design parameters
of spring used in fairing jettison mechanism are also evaluated.
For this purpose, analytical equations for rotational dynamics
are derived using laws of motion. Then using Range-Kutta
method, these equations are solved, simulated and analyzed.
Acquired results are validated through professional software
used for fairing separation calculations..
II. FAIRING SEPARATION MECHANISM
Jettison is started by lateral separation mechanism that
unlocks the fairing bottom from launch vehicle. Then pyro-
bolts of longitudinal separation mechanism are exploded that
separate both halves of fairing, followed force is released by
compressed springs which rotate the halves about hinges. After
reaching certain angle, fairing starts free fall rotation about its
center of gravity. Fairing and hinge structures under
consideration are illustrated in Fig. 1(a) and 1(b).
Fig. 1(a): Fairing structure
261
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-270-320.jpg)
![Fig. 1(b): Fairing model along with spring mechanism and hinge structure
III. FAIRING SEPARATION DYNAMICS
Dynamic equations are derived for 1-DOF system. For
analytical derivation, fairing is considered as rigid body.
During separation, hinge point's linear movement is neglected.
Frame moves with constant linear acceleration fixed on launch
vehicle, is taken as reference [7]. Pitch angle of launch vehicle
is taken as zero. Free body diagram of left half of fairing before
fairing separation is illustrated in Fig. 2. Free body diagram of
left half of fairing during separation is shown in Fig. 3.
Fig. 3: Free body diagram of left half of fairing before separation
Fig. 4: Free body diagram of left half of fairing during separation
For left half of fairing, summation of all the moments about
the hinge point is given by:
(1)
After re-arranging (1);
(2)
Where, 'Izz ' is the moment of inertia of left half about
hinge. 'Fs ' is the magnitude of force applied by all springs. 'ds'
is the moment arm of spring force 'Fs '. 'Mfair' is the mass of
fairing half. 'a' is the linear acceleration applied on the fairing
due to instantaneous thrust of launch vehicle, which is
considered as constant during separation. ' ' is the angle of
rotation of fairing half. 'Fs' is calculated using equation given
below.
(3)
Where, 'K' is stiffness of one spring. 'Xtot' is the total
compressed length of a spring. 'Ns' is the number of springs. In
given case, 2 springs are considered in separation mechanism.
'Xdef' is the instantaneous uncompressed length of spring during
separation and it is found using cosine law of triangle given by,
(4)
Equations (1) to (4) are coupled through ' '. These inter-
linked dynamic equations are solved by using Range-Kutta
method. Same is used for dynamic solution of right half of
fairing with trivial modifications.
Theoretically, force exerted by the spring system 'Fs ' on the
base ring and the reaction force 'Fr ' generated by the ring on
the spring system is equal. But in practical, "action is not equal
to reaction" due to inelastic contact of two bodies [8-9]. Spring
impingement on the base ring causes the absorption of energy
in the ring. Amount of energy absorbed depend on the
materials of spring and ring. Thus reaction force produced by
ring will always be less than the spring force. Harder material
absorbs less energy and vice versa. Force loss factor 'e' is
considered to make analytical solution closer to real results.
Therefore 'Fr ' is used instead of 'Fs ' in equation (2).
(5)
Equations (1) and (2) are valid prior to fairing separation
from hinge. After fairing separation, analogous equations are
required to derive for motion of fairing about its center of
gravity.
IV. RESULTS AND DISCUSSION
Equations are implemented and solved in MATLAB.
Spring parameters including spring stiffness, spring material
and compressed length, are evaluated to satisfy given
separation requirements given in table 1.
262
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-271-320.jpg)
![Fig. 4: Reaction Force of Left Half of Fairing
Fairing's opening angle is presented in Fig. 8. At 0.78
sec, fairing achieved rotation of 70o
so it is released from
hinge, satisfying the requirements given in table 1.
Fig. 5: Opening Angle Theta of Left Half of Fairing
Fig. 9 presents the movement of fairing half in body
frame. Red curve shows the cg variation of fairing's half
during separation.
Fig. 6: Fairing's half Movement during Separation
V. CONCLUSION
Results show that designed analytical approach for fairing
separation dynamic analysis is fairly close to professional
software approach. Inclusion of material properties of spring
and base ring made results more comparable. Among various
springs of different stiffness and compressed lengths, best one
(having stiffness of 100,000 N/m and compressed length of
0.08m) is chosen on the basis of given fairing separation
parameters. 3dof and 6dof systems will be considered; modeled
and simulated in future that will provide complete dynamics of
fairing separation along all axes.
ACKNOWLEDGMENT
We would also like to show our gratitude to Dr. Imran
Afzal for providing insight and expertise that greatly assisted
the research.
REFERENCES
[1] Steven J. Isakowitz, Joshua B. Hopkins, Joseph P. Hopkins Jr.,
“International Reference Guide to Space Launch Systems,” 4th ed.,
AIAA, pp. 341, 2004.
[2] “Soyuz User's Manual,” Ariane Space, vol 2 rev. 0 ,March 2012.
[3] “Atlas Launch System Mission Planner's Guide, Atlas V
Addendum(AVMPG),” International Launch Services, pp. 2-7, Dec.
1999.
[4] Antonio Pagano, “Global Launcher Trajectory Optimization for Lunar
Base Settlement,” Delft University of Technology, May 2010.
[5] spacenews.com,‘Orbitals-glory-launch-failure-review-nears-conclusion’,
2011. [Online]. Available: http://spacenews.com/orbitals-glory-launch-
failure-review-nears-conclusion/. [Accessed: Aug. 9, 2015].
[6] Choong-Seok Oh, Byung-Chan Sun, Yong-Kyu Park and Woong-Rae
Roh, “Payload Fairing Separation Analysis Using Constraint Force
Equation,” International Conference on Control, Automation and
Systems, Gyeonggi-do, Korea, Oct. 27-30, 2010.
[7] Raymond H. Schuett, Bruce A. Appleby, Jack D. Martin, “Dynamic
Load Analysis of Space Vehicle System ,Launch and Exit Phase,”
General Dynamics Convair Division, Report No. GDC-DDE66-012,
June 1966.
Fairing's half
separated from
hinge
Path of CG
[9] en.wikipedia.org, 'Inelastic Collision', 2015. [Online]. Available:
https://en.wikipedia.org/wiki/Inelastic_Collision. [Accessed: Aug. 8,
2015].
[8] W. Bauer, G.D. Westfall, “Physics for Scientists and Engineers,” 3rd
ed., McGraw Hill, 2008.
264
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-273-320.jpg)
![Abstract— In this paper, a robust attitude control scheme for
satellite launch vehicle (SLV) is presented that is based on inverse
dynamics control augmented by sliding mode neural network
compensator. In this proposed scheme, the inverse dynamics
control acts as the primary control agent that computes the
control moment required to not only stabilize the launch vehicle
but also to implement the steering commands obtained from
guidance scheme. This control action is derived solely by the
inversion of launch vehicle dynamics; therefore, it is vulnerable to
errors resulting from the uncertainties, simplifying assumptions,
nonlinearities and various local and environmental disturbances
all of which cannot be accurately modeled. This problem is
addressed by the secondary control agent of the proposed control
scheme i.e. a sliding mode neural network compensator. This
nonlinear compensator is independent of the system dynamics and
therefore ensures stability against the un-modeled dynamics and
the disturbances acting on the launch vehicle. Along with this
hybrid control architecture, a guidance scheme is also developed
that reshapes the reference attitude profile on the basis of any
deviation from the desired trajectory of the launch vehicle so that
the desired orbit is reached. Lyapunov stability criterion is
incorporated in the control loop to guarantee system stability and
the efficacy of guidance and control scheme is verified on a six
degree of freedom simulation model developed exclusively in
Matlab/Simulink® using reference data for a four stage launch
vehicle. In these simulations, the launch vehicle is subjected to
various severe combinations of disturbances and parametric
variations and the control system successfully compensates all the
disturbances and efficiently tracks the desired attitude profile and
attains the targeted orbital parameters.
Keywords—Launch Vehicle Control, Dynamic Inversion,
Inverse Dynamic Control, Guidance and Control, Sliding Mode
Neural Network Compensator
NOMENCLATURE
m = Mass of vehicle
α= Angle-of-Attack
β= Side slip angle
δ= Deflection angle
J= Moment of inertia
ω = Angular rate
g = Gravitational acceleration
P = Thrust force
xr = Dist. from nose to nozzle center of gravity
xd = Dist. from nose to vehicle center of pressure
zr = Dist. from central line of vehicle body to parallel
central line of thrust pipe
xz = Dist. from nose to vehicle center of gravity
V = Air stream velocity of vehicle
C = Side force
L= Lift force
D = Drag force
φ= Pitch angle
ψ = Yaw angle
γ = Roll angle
θ = Flight Path Angle
𝜗 = Local flight path angle
σ = Heading angle
I. INTRODUCTION
The access to space in a reliable and cost effective way is a
critical requirement these days. This ability depends largely on
the performance of the guidance and control system of the
launch vehicle. The classical controllers for the ascent phase of
SLV depend heavily on the accuracy of its dynamic modeling
in order to perform as anticipated. Because, it may not satisfy
the stability and performance requirements if the SLV is
outside of its desired trajectory envelop. But, due to the various
uncertainties, nonlinearities and disturbances acting on the
system, the dynamics of SLV cannot be precisely modeled.
Therefore, the controller needs to have the ability to perform
even with inaccurate dynamic model. The controller of a
launch vehicle has to satisfy three often contradictory
requirements that are to stabilize the vehicle, reduce trajectory
deviations by efficiently implementing the steering commands
from guidance scheme and to minimize the angle of attack in
the high dynamic pressure region to avoid structural overload.
During the last few years, a lot of work has been reported in
the field of adaptive and nonlinear control of launch vehicle [1]
[2] [3] [4] [5]. Such techniques have the adaptability to adjust
their output in accordance with the varying dynamics of the
system while retaining the desired performance and stability.
Similarly, various techniques have been devised to enhance the
robustness of the inverse dynamic controller [6] [7] against
modeling uncertainties.
Launch Vehicle Control based on Dynamic
Inversion with Sliding Mode Neural Network
augmentation
Saqib Alam
SUPARCO, Pakistan
e-mail: alam.saqib@gmail.com
Syed Minhaj un Nabi Jafri
SUPARCO, Pakistan
e-mail: jafri_minhaj@yahoo.com
265
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-274-320.jpg)
![This research is aimed at demonstration of the simplicity
and effectiveness of inverse dynamic controller when it is aided
with a sliding mode neural network compensator in developing
the autopilot of SLV. The advantage of using this architecture
is that the primary control produces just enough control
moment that is required for vehicle stabilization and trajectory
following resulting in a simplified and efficient control law.
Whereas all the additional factors are taken care of by the
secondary control, hence ensuring robustness and adaptability.
The weights of the sliding mode neural network are tuned
adaptively in synchronization with the changing environment.
In order to inject a satellite in a desired orbit, an external
guidance loop is also required alongside the inner control loop
to provide the steering commands to the controller on the basis
of trajectory deviations that may occur due to environmental
and local disturbances. A simplified guidance scheme is
proposed here for successful accomplishment of orbital
parameters.
The paper is organized as follows. Section II illustrates the
mathematical modeling of SLV. Section III describes the
design of inverse dynamic control with sliding mode neural
network compensator for attitude control of SLV along with
the guidance loop. The simulation results are presented in
Section IV and Section V provides the concluding remarks.
II. DYNAMIC MODELING OF SLV
To analyze ascent flight trajectory, 6DOF simulation model
is developed in Simulink MATLAB in which the data of four
stage SLV is incorporated. The symbols definition of SLV is
depicted in Fig.1.
Fig. 1 Symbol’s definitions
Table I depicts the four stage data used to simulate the
trajectory of SLV.[1]
TABLE I. LAUNCH VEHICLE DATA
Parameters 1st
stage 2nd
stage 3rd
stage 4th
stage
Liftoff Mass (ton) 31.34 9.31 2.87 0.91
Fuel Mass (ton) 18.76 5.55 1.72 0.54
Thrust (kN) 593 262 94 30
Mass Flow rate
(kg/s)
265 103 37 13
Burn Time (s) 70.7 53.8 46.9 46.9
Stage Dia. (m) 1.3 1.0 0.7 0.5
The launch vehicle experiences various aerodynamic forces
and moments during its flight that act as disturbances and cause
trajectory deviations. In order to simulate these effects,
trajectory of 500 km orbit is considered with injection velocity
of 7612 m/s2
, DATCOM is used to compute aerodynamic
coefficients, mass flow rate and thrust are considered as
constant. The equations of motion are formulated considering
the rigid body dynamics, and are given as follows [2]:
sin cos 4 cos cosmV mg P D (1)
cos 4 sin cos
2 cos
mV mg P
P L
(2)
sin sin 4 sin
cos2 C
mV mg P
P
(3)
2
( )y z y zdx m kr
x
x x x
J Jm qS lPz
J VJ J
(4)
2
( ) ( )
2 ( )
dy m kz x d z
y x z
y y y
R z
y
m qS lJ J C x x
J J VJ
P x x
J
(5)
2
( ) ( )
2 ( )
x y d z dz m k
z y x
z z z
R z
z
J J L x x m qS l
J J VJ
P x x
J
(6)
Where,
III. DESIGN OF IDSNN CONTROL
For Launch vehicle, the proposed control architecture
266
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-275-320.jpg)
![consists of feed-forward term based on Inverse Dynamics of
SLV and feedback term in which sliding mode based neural
network compensator is employed.
A. Problem formulation
In order to apply dynamic inversion to on the attitude
dynamics of launch vehicle, it is assumed that the inertial rates
, ,p q r are equivalent to body rates , , neglecting the
effect of earth rotation. Therefore, the dynamic equations 4-6
can be written as:
(7)
Where, , , , , , , represent aerodynamic
moments, damping moment and control moment respectively.
B. Inverse Dynamic Control
The dynamic inversion methodology to formulate control
input diu requires two steps. The first step is the state and
input transformation to convert nonlinearities of the system
dynamics into its linear time invariant form:
z az bv (8)
The second step is the designing of the control
input v using linear control methodology. To achieve stable
tracking control, three variables of interest i.e.
( , , )diy are chosen as output variables. Transform
system to get the form as shown in Equ.8, Output vector is
differentiated twice to get input vector [8]:
di di di diy M N u (9)
Where,
T
diy ,
0 0 ( ) 0 0
0 0 0 ( ) 0
0 0 0 0 0
diM
0 0
0 0
0 0
diN
and diu
The desired input to the system can be written as:
1
( )di di di diu N M v
(10)
Where,
( )dI dIy v v v v (11)
To design the linear controller to control the output states
( , , )diy , the linear controller is written as:
( ) ( )
( ) ( )
( ) ( )
d d d p d
DI d d d p d
d d d p d
v k k
v v k k
v k k
(12)
C. Feedback term based on Sliding mode Neural Network
In practice, it seems to be impossible to model the exact
dynamics of the system. Since the dynamic inversion control
explicitly utilizes the modeled dynamics, it is assumed
that 0 ,aero
0damp
and 0con
represent aerodynamic, damping and
control moment of the nominal model respectively. The
difference in actual and nominal model is represented by
term for which the moments can be written as:
0 0
0
, ,aero aero aero damp damp damp
con con con
(13)
By substituting these values, Equ. 7 can be written as:
0 0
0 0 0
0 0 0
( , )
( , , )
( , , )
damp con damp con
aero damp con aero damp con
aero damp con aero damp con
f
f
f
(14)
In order to estimate the function ( , )damp con
f , for roll and
( , , )aero damp con
f for pitch and yaw, adaptive Sliding mode
Neural Network compensator (SNN) is proposed [9]. By
employing this, the control deflection computed by the
proposed controller is written as:
ˆ(.)diu f (15)
Radial Basis Neural Network compensator is designed on
the basis of sliding surface and the surface depends on error
and its rate. RBF networks are adaptively used to approximate
the uncertain f.
( )T
i if w h s
267
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-276-320.jpg)
![Where h(s) is the Gaussian function of neural network based
on sliding surface.
2
2
exp , 1,2,...5.
2
i
i
i
s c
h s i
(16)
Fig. 2 SNN structure
The sliding surface is defined as:[8]
*S C e e (17)
Let the nonlinear system is defined as:
( , )x f x x gu dt
We know that
s ce e
ds e ce x x ce (18)
ˆ ( )dx fx gu dt ce ksign s
The control input can be written as:
1 ˆ ( )du x fx dt ce ksign s
g
ˆ( ) ( ) ( )s f x f x dt ksign s
( ) ( )s f x dt ksign s (19)
Where
ˆ ˆ( ) ( ) ( )T T T
f f f w h s w h s w h s
Lyapunov function is defined as:
21 1
2 2
T
L s w w (20)
The derivation is written as:
T
L ss w w
ˆ( ( ) ) ( ( ))T
L w sh s w s dt ksign s (21)
Where ε is approximation error of neural network and γ is a
positive coefficient. The stability Criteria is defined as:
0LL
The adaptive law is selected as
1
ˆ ( )w sh s
Now we can write:
( ( )) ( )L s dt ksign s s dt k s (22)
Since approximation error is sufficiently small and
k dt
We get
0L
D. External Guidance loop:
In order to achieve the desired orbital parameters, an
external guidance loop is incorporated in the autopilot of SLV.
The guidance scheme performs the online reshaping of the
reference attitude profiles in the exo-atmospheric phase. It
computes deviations (Δx, Δy, Δz) from the intended trajectory
and on the basis of these deviations, it formulates the new
reference attitude angles accordingly. The controller in turn
tracks the updated reference profiles to attain the desired orbit.
dx x x (23)
dy y y (24)
dz z z (25)
1
tan
y
x
(26)
1
tan
z
x
(27)
d ref (28)
d ref (29)
The complete block diagram of the proposed control
algorithm with external guidance loop is depicted below:
268
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-277-320.jpg)
![Fig. 3 Guidance & Control architecture
IV. SIMULATION RESULTS
The performance evaluation of the proposed control
architecture was carried out on a six degree of freedom
simulation software developed in MATLAB Simulink. The
simulation result of pitch channel for the first 30 seconds of
flight including the pitch over phase under nominal flight
conditions is shown in Fig. 4.
Fig. 4 Pitch over Attitude profile with Compensator
It can be seen from Fig. 4 that the proposed controller is
efficiently tracking the reference pitch angle profile. The
robustness analysis against parametric variation of the control
scheme is carried out by the incorporation of 30% random
variation in mass properties of the SLV.
Fig. 5 Attitude profile without Compensator with Uncertainty
Fig. 6 Attitude profile with Compensator with Uncertainty
0 5 10 15 20 25 30
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Time [Sec]
Deflection[Deg]
Without compensator
With compensator
Fig. 7 Comparison of Commands with Uncertainty
Fig. 5-7 show the comparison of controller performance
with and without the sliding mode neural network compensator
and their resultant outputs under parametric variations. It can
be seen that the hybrid control structure is immune to modeling
269
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-278-320.jpg)
![uncertainties.
The effectiveness of the proposed guidance scheme with
online trajectory reshaping is depicted in Fig. 8 that shows the
error in orbital attitude with and without guidance under
parametric variations. The guided SLV accurately attains the
desired orbit.
Fig. 8 Injection Altitude with uncertainties
Fig. 9 shows the pitch profile for the whole mission from
lift-off to orbit injection including online attitude reshaping
from guidance scheme to attain the desired orbital parameters.
0 100 200 300 400 500 600
-50
0
50
100
Time [sec]
PitchAngle[deg/sec]
Actual
Reshaping
Reference
Fig. 9 Pitch profile with uncertainties
V. CONCLUSION
An adaptive and robust control strategy for SLV autopilot
is presented in this paper that is based on inverse dynamic
control aided with a sliding mode neural network compensator.
Lyapunov stability criterion is utilized to ensure the stability of
the control algorithm. The simulation results prove that the
proposed controller not only guarantees stability but also
efficiently tacks the reference attitude profile and the steering
commands from the guidance loop even under parametric
variations. The guidance scheme performs the online reshaping
in the exo-atmospheric phase of flight in case of deviations
from the intended trajectory to meet the desired orbital
parameters accurately.
REFERENCES
[1] Uzair Ansari. "Hybrid Genetic Algorithm fuzzy rule based guidance and
control for launch vehicle", 2011 11th International Conference on
Intelligent Systems Design and Applications, 11/2011
[2] Ansari, Uzair, Saqib Alam, and Syed Minhaj un Nabi Jafri. "Trajectory
optimization and adaptive fuzzy based Launch Vehicle attitude control",
2012 20th Mediterranean Conference on Control & Automation (MED),
2012.
[3] Filho, W. C. L., "Application of Adaptive Control Sounding Rocket
Attitude", Proceedings of International Conference on Intelligent
Autonomous Control in Aerospace, Beijing, China, (1995).
[4] Zhihua Qu, “Robust Control of Nonlinear Uncertain Systems” John Wiley
& Sons, Inc. 1998.
[5] Yuri Shtessel, James McDuffie, Mark Jackson, Charles Hall, Don Krupp,
Michael Gallaher, and N. Douglas Hendrix, "Sliding Mode Control of the
X33Vehicle in Launch and Re-entry Modes," AIAA98-4414, Proceedings
of AIAA Guidance, Navigation, and Control Conference, Boston, MA,
August 10-12, 1998
[6] Ito D, Ward DT and Valasek J. Robust dynamic inversion controller
design and analysis for the X-38. In: AIAA conference on guidance,
navigation and control, Canada, 6–9 August 2001, paper no. AIAA-2001-
4380.
[7] David B. Doman and Anhtuan D. Ngo “Dynamic Inversion Based
Adaptive/Reconfigurable Control of the X-33 on Ascent” Journal of
Guidance Control and Dynamics, 2002.
[8] Abhijit Das, Frank L. Lewis and Kamesh Subbarao (2011). Sliding Mode
Approach to Control Quadrotor Using Dynamic Inversion, Challenges
and Paradigms in Applied Robust Control, ISBN: 978-953-307-338-5.
[9] Feng G (1995), “A compensating scheme for robot tracking based on
neural networks”. Robot Auton Syst 15(3):199–206.
270
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-279-320.jpg)
![Design and Development of Low cost Motor Drive
for Hub Wheel based Electric Vehicles
Syed Wasif Ali Shah
SUPARCO-Karachi-Pakistan
wasif_iiee@yahoo.com
Muhammad Asim
SUPARCO-Karachi-Pakistan
m_asim2000@hotmail.com
Abstract—Robotic mobility platforms are serving a wide
range of civil, commercial, domestic and military applications.
This paper describes the design and development of low cost
motor drive, applicable for wide range of permanent magnet DC
(PMDC) motors for electric vehicles and especially for the hub
wheel motors as they exhibit a distinctive frequency behavior.
This controller comprises of 3 units: First, a logic unit providing
control signals; Second, a sensing unit for over current and
batteries protection: third, a high power switching unit for
driving motors. Though applicable to wide range of unmanned
ground vehicles (UGVs) but here it is implemented in a
differential drive autonomous UGV built on hub motor wheels
with a speed of 1.88 m/sec, endurance of 2 hours and payload
capacity of more than 100 kg. Features, performance and
development cost of the drive in comparison with the similar
commercially available drives is also discussed.
Keywords— Mobility platforms; UGV; Hub Wheel; Motor
Drive;
I. INTRODUCTION
Unmanned ground vehicles (UGVs) are the mobility platforms
that have enormous civil, commercial, military, domestic and
urban applications [1][2] ranging from rescue, surveillance,
reconnaissance, mine detection, fire fighting, material
handling, agriculture, handicapped assistance, border security,
special weapons and tactics (SWAT) robot, manipulation,
exploration, automation, transportation, handling in dangerous
environment and many others. UGVs may be based on
different propulsion systems like engine based, electric or
hybrid drive mechanism [3]. Similarly different applications
and features require different traction and locomotion systems
[4] depending upon the required speed, tracks, obstacle
traversing, energy efficiency, load bearing capacity,
mechanical and control complexity and many others. To meet
different applications and fulfill different tasks through UGVs,
varieties of motorized actuation systems are available and
applied; also different mechanical structures and their level of
complexity have different actuation systems [5].DC motors
amongst them are most widely used actuator in electric
vehicles due to their low cost, simple and economical control
and weight to performance ratio [6].Varieties of DC motor
drives are available having different options, ratings and
control interfaces [7][8][9]. The motor drive described here
has the ability to drive wide range of permanent magnet
brushed DC motors with power rating of up to 1.8 KW.
Particularly we have developed a UGV shown in fig: 1 with
the dimensions of 72 cm in length, 91 cm in width and 20 cm
in height, having weight of 90 kg and payload capacity of
more than 100 kg build on the Hub motorized wheels. This
UGV is primarily designed for autonomous transportation
with GPS navigation with maximum speed of 1.88 m/sec. The
UGV has the differential drive mechanism with zero turning
radius. The motor drive has shown the ability to smoothly
control the hub motor’s torque, rpm and direction. The
advantage of hub wheel motor in building the UGV is its
mechanical simplicity and installment, compactness giving
more room to other components, less complex structure of
motion transmission eliminating axels and drivelines, reduced
weight giving more efficiency and endurance and ability of
high torque at low rpm [10]. The implemented hub motor has
the following major parameters given in tab-1.
Table 1: Hub Motor Parameters
Parameter Value
Power 250W
Voltage 24 Vdc
RPM 120
Radius 6 in
Figure 1: The differential drive UGV build with Hub motorized wheels
271
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-280-320.jpg)
![Table 2: Costing of the motor drive.
Component Appromimate
Cost
Drive PCB $10
Drive Components $48
Development $7
Total $65
Table 3: Specifications of the motor drive
Specification value
Supply voltage 13V to 50V
Output Current (continuous) 160A
Output Current (surge) >400
Switching Frequency Upto 16 kHz (with 4 MOSFESTs)
Weight 100 g (without fan)
MOSFETs 16 (4 per leg)
On Resistance .0026 ohm max at 25C
Cooling
Through optional 40 CFM 12 volt
fan
Dimension (mm) 115(L)X80(W)X35(H)
Parameters and features of the drive are given in tab: 3. The
drive has competitive features and performance if compared
with similar commercially available PMDC motor drives
whose prices range from $200-$400 [7][8][9]. The drive is
thoroughly tested with different motors and loading conditions
for different applications. It conforms to the specifications
especially under harsh and open environment of robotic
mobility platforms for long durations. Different features can
also be added in this design with minimal marginal cost which
otherwise incurs heavy cost if procured.
V. CONCLUSION
The paper discussed the design and development of low cost
PMDC brushed motor drive. Particularly implemented in
differential drive UGV with heavy payload capacity, build
with hub motorized wheel showing distinctive frequency
response. The developed controller consists of three units:
logic unit, sensing unit and high switching power unit. The
drive can be driven virtually with any microcontroller and
other signal sources like PC and wirelessly with RC signals. It
senses the current being drawn and the battery voltage levels
and further can be enhanced to monitor other motor
parameters like temperature. Features and development cost of
the drive are also discussed and compared with similar drives
available commercially. The drive is very economical with
easy service and replacement.
VI. REFERENCES
[1] V. Sezer, C. Dikilita, Z. Ercan, H. Heceoglu, A.Buner, A.Apak,
M.Gokasan and A. Mugan “Conversion of a conventional electric
automobile into an unmanned ground vehicle (UGV)” IEEE
international conference on Mechatronics, pp. 564-569, (2012)
[2] A. Bouhraoua, N.Merah, M. Al-Dajani and M. Elshafi “Design and
development of an unmanned ground vehicle for security applications”
7th International symposium on mechatronics and its applications, pp.
1-6, (2010)
[3] LinoGuzzella, Antonio Sciarretta “Vehicle propulsion systems”
Introduction to Modeling and Optimization, 2nded.
[4] L Bruzzone and G. Quaglia “ Review Article: Locomotion systems for
ground mobile robots in unstructured environment” Mech:sci:, 3,49-62,
2012, www.mech-sci.net/3/49/2012/
[5] William Brown “Improving the electromechanical performance of Robot
Arms on Unmanned Ground Vehicles” GRRC Technical report 2010-01,
12/20/2009
[6] Nasser Hashemnia and Behzad Asaei“Comparative study of using
different electric motors in the electric vehicles” proceedings of the 2008
international conference on electrical machines, paper ID 1257.
[7] Roboteq, Inc., “Brushed DC motor controller”, Online:
http://roboteq.com, 2015.
[8] Robot Power. “Robot power products”, Online: http://robotpower.com,
2015.
[9] VEX Robotics. “Victor Motor Controller”. Online:
http://vexrobotics.com, 2015.
[10] Machine Design. “A look at the advantages of hub motors as well as
disadvantages”, Online: http://machinedesgin.com, 2015.
274
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-283-320.jpg)
![Preparation, Structure and Dielectric/Piezoelectric
Properties of BiScO3-PbTiO3-Pb(Mn1/3Nb2/3)O3 High
Temperature Piezoelectric Ceramics
Shahid Karim1
, Jinting Tan1
, Adil Murtaza2
, Shahid Sultan2
, Khalida Kareem2
, Muhammad Nouman Shaikh3
1
Electronic Material research laboratory & International Center of dielectric research, School of Electronics and Information
Engineering, Xi’an Jiaotong University, Xi’an 710049, China
2
School of Sciences, Frontier Institute of Science and Technology, MOE Key Laboratory for Nonequilibrium Synthesis and
Modulation of Condensed Matter, State Key Laboratory for Mechanical Behaviour of Materials, Xi’an Jiaotong University, Xi’an
710049, China
3
Xi’an Jiaotong University, Xi’an 710049, China
Abstract—With the rapid growth of science and technology,
more piezoelectric devices should be used in high-temperature
environment. Commercially using PZT piezoelectric ceramics
have a low curie temperature (<150°C). So the present study is
focused on synthesis and characterization of high Curie
temperature piezoelectric ceramics. As a kind of excellent high
Curie temperature piezoelectric ceramics BiScO3-PbTiO3 (BS-
PT) system at MPB reveals high Curie temperature piezoelectric
properties. Pb(Mn1/3Nb2/3)O3-PbTiO3 (PMnN-PT) system
ceramics have excellent piezoelectric properties at MPB with low
Curie temperature (150°C). 0.1PMnN-BSPT ceramics show the
optimal piezoelectric properties with d33=200pC/N, kp=0.40,
Tc=338°C, Ec=22.75kV/cm and Pr=21.85µC/cm2
. The 0.05PMnN-
BSPT ceramics indicate the optimal piezoelectric properties with
d33=250pC/N, kp=0.54, Tc=373°C, Ec=14.63kV/cm and
Pr=11.51µC/cm2
. Also 0.06PMnN-BSPT system with 64mol.% PT
indicates the optimal piezoelectric properties with d33=210pC/N,
kp=0.49, Tc=328°C, Ec=20.21kV/cm and Pr=8.35µC/cm2
. So the
introduction of PMnN reduces the curie temperature; 10mol.%
of PMnN indicates the “hard” behavior as a function of reducing
the piezoelectric properties; 6mol.% of PMnN indicates the
“soft” behavior with increasing the piezoelectric properties.
Index Terms—Morphotropic phase boundary (MPB), Curie
temperature, Piezoelectric Properties, High-temperature
piezoelectric ceramics, Binary system ceramics
I. INTRODUCTION
Piezoelectric ceramic is a kind of important functional
ceramic material, which can realize the conversion between
mechanical energy and electric energy which are widely used
in manufacturing, micro displacement transducer and
ultrasonic generator etc. With the development of science and
technology, a lot of electrical and electronic equipment which
are used in high temperature piezoelectric devices increased
the demand. Study on the high temperature, with excellent
piezoelectric properties of piezoelectric ceramic materials
have become a hot research topic of functional ceramics
industry. The commercial applications of lead zirconate
titanate (PZT) in piezoelectric ceramics range of Curie
temperature system is 250-380°C, and the temperature is
limited to 150°C or less, which is far less than high
temperature piezoelectric devices. For a long time, a lot of
piezoelectric crystal materials due to the use of high
temperature and excellent piezoelectric properties have
become an important part of high temperature piezoelectric
applications [4]. Therefore high electromechanical power
applications requires such ceramics with better electrical
properties such as the piezoelectric strain constant (d33>200
pC/N), low dielectric loss (tanδ) mechanical quality factor
(Qm>1000), planar electromechanical coupling factor (kp> 0.5)
[1-3]. Lead niobate (PbNb2O6) tungsten bronze structure was
first discovered, it has tetragonal tungsten bronze structure [5].
This system has large anisotropy piezoelectric coefficient,
quality factor, low and high Curie temperature (570°C) is not
easy to depolarization, therefore particularly suitable for the
manufacture of high temperature transducer. But pure lead
niobate piezoelectric activity difference exists, the
electromechanical coupling coefficient is low, and is not easy
to sinter, and its high temperature ferroelectric phase at room
temperature is not stable. There is irregular grain growth and
phase transformation during sintering and cooling process
caused by the large volume change, so it is difficult to prepare
the excellent performance of the pure lead niobate
piezoelectric ceramic materials [6]. Through the substitution
of bismuth layer structure piezoelectric ceramic modification
can improve its piezoelectric activity. Korzunova [5] in the
Bi3TiNbO9 system, using Ti4+
, Ta6+
or W6+
instead of B in the
Nb5+
, using K+
to replace Bi3+
, the piezoelectric constant 5
pC/N is increased to 24 pC/N.
II. EXPERIMENTAL PROCEDURE
Mixture grade (>99.9% purity) powders of Bi2O3, PbO,
Sc2O3, TiO2, MnCO3, and Nb2O5 were used to make xPbTiO3-
(0.9-x)BiScO3-(0.05 to 0.15)Pb(Mn1/3Nb2/3)O3, (x=0.64). The
mixture of MnCO3 and Nb2O5 was ball-milled in a
polyethylene jar for 4h by high purity tetragonal zirconia
polycrystalline (3Y-TZP) ball media (Tosoh Co., Tokyo,
Japan). The mixed slurry was dehydrated and condensed for
calcination and phase formation at 800°C, 850°C and 900°C
for 5h. The Mn1/3Nb2/3O2 phase was confirmed by X-ray
diffractometer (XRD) after calcination, to obtain fine particles
the calcined piece was ball milled again for 4 h. After
preparation of Mn1/3Nb2/3O2 powder, PbO, Bi2O3, Sc2O3, and
275
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-284-320.jpg)
![TiO2 were mixed with it and ball milled to make final
compositions. Pellets (12.7x3mm) were compressed under a
pressure of 100 MPa by using a hydraulic press. Pressureless
sintering was performed in a high purity alumina crucible at
1100°C, 1150°C and 1200°C for 2h. To prevent from reaction
with alumina crucible, the specimens were put on a Pt-foil
inside crucible. After sintering, apparent densities were
measured by the Archimedes method.
Electrode formation was done by sputtering silver on the
lapped surfaces of pallets. By applying dc field of 3.5 kV/mm
for 10 min, electroded specimens were poled at 120°C in
silicone oil. Dielectric properties were acquired by measuring
the capacitance and loss using LCR meter (HP model 4284).
To determine Curie temperature, capacitance measurements
were completed as a function of temperature in an automated
temperature controlled furnace linked with a computer for data
acquirement. The piezoelectric coefficient (d33) was noted from
one-day aged samples by using Berlincourt-d33 meter.
Electromechanical coupling factor (kp) and mechanical quality
factor (Qm) were obtained by using an impedance analyzer
(Model HP4194) built on the IEEE standards [12]. Each
experimental step is given in Fig. 1.
Make the composition of
Mn1/3Nb2/3O2 by the
molecular weight of
MnCO3, and Nb2O5
Calcinate it at
1000'C for 5 hours
Mix the powders of Mn1/3Nb2/
3O2, PbO, Bi2O3, Sc2O3 and
TiO2 and take six compositions of
5%, 6%, 8%, 10%, 12% and 15%
Calcinate each sample
at different temperature
800'C, 850'C and 900'C
respectively
Blend the powders by
mixing with PVA+
Water and take the
specific grain size of
powders to make pallets
Sintered each
sample at different
temperatures
1150'C, and 1200'C
Make the pallets
shape smooth and
then measure the
apparent densities of
each sample
silver was pasted for
the electrode
formation
The electroded
specimens were poled in
silicone oil at 120
o
C by
applying a d.c. field of
3.5 kV/mm for 10 min
piezoelectric coefficient d33
was recorded using a
Berlincourt d33 meter,
dielectric properties were
obtained using an LCR meter
Ball milled,4h
Ball milled,4h
XRD
XRD
Fig. 1. Flow chart of experiment
III. RESULTS AND DISCUSSIONS
A. Sintering behavior and microstructures
Fig. 2, shows Powder XRD for the samples 5, 6, 8, 10, 12
and 15% of PMnN calcined at 800°C, 850°C and 900°C. It is
clearly shown at 8% of PMnN the XRD pattern is smooth.
Also it is studied that all the samples show tetragonal and
rhombohedral phases. Fig. 3, shows sintered XRD for the
samples 5, 6, 8, 10, 12 and 15% PMnN which are sintered at
1150°C. There is clearly shown at all doping percentage of
PMnN the XRD pattern is smooth. Also it is seen from the
angle 44 to 46 at the peak of 200 of all the samples show
tetragonal and rhombohedral phases. Fig. 4, shows sintered
XRD for the samples 5, 6, 8, 10, 12 and 15% PMnN which are
sintered at 1200°C. There is clearly shown at all doping
percentage of PMnN the XRD pattern is smooth. Also it can be
seen from the angle 44 to 46 at 200 peak, all samples show
tetragonal and rhombohedral phases.
Fig. 2. Powder X-ray Diffraction Analysis of Samples
Fig. 3. XRD of the samples of x=5, 6, 8, 10, 12, 15% respectively
sintered at 1150°C
Fig. 4. XRD of the samples of x=5, 6, 8, 10, 12, 15% respectively
sintered at 1200°C
The microstructures of the composites were studied by
Scanning electron microscope (SEM). The SEM micrographs
of the samples (a) 5%; (b) 6%; (c) 8%; (d) 10%; (e) 12% and
(f) 15% of PMnN sintered at 1150°C for 2h are shown in Fig.
5. They showed different microstructures distinctly; grain size
of each sample is (a)3.6; (b)2.8; (c)3; (d)3.2; (e)3.7; and (f)2; in
micrometers respectively and grain boundaries of sample
(c)8% are more clear and significant. The SEM micrographs of
the samples 6% of PMnN sintered at (a) 1100°C; (b) 1150°C
and (c) 1200°C respectively for 2h are shown in Fig. 6. They
showed different microstructures distinctly; grain size of each
276
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-285-320.jpg)
![TABLE I. Dielectric and Piezoelectric Properties with Curie temperature
Systems d33 (pC/N) Tc (°C) kp Qm εr εmax
5% 250 373 0.54 30 1018 12820
6% 210 328 0.49 15.6 761 4874
8% 220 302 0.42 27.06 1149 5104
10% 200 338 0.40 59.60 931 9809
12% 74 286 0.35 14.5 764 3169
15% 160 326 0.50 13.38 927 10248
E. Ferroelectric Properties
Polarization is saturated PS at high electric fields and there
is no more motion of the domain walls. Area of the hysteresis
curve shows heat density generated in the material at one
cycle. It is clearly seen that the 10% of PMnN has high
polarization value as well as electric field as shown in Fig. 15.
If the residual magnetization is high when magnetic
intensity will zero, it means that material have strong value of
magnetization. If the value of residual magnetization is less, it
means it will not retain the magnetization for long time after
removing the magnetic intensity. The results of remanent
polarization and coercive field were given in tab. II.
When the electric field began to decrease because of the
crystal stress domain deviate from the direction of
polarization, the electric field is reduced to zero, still remain in
the polarization direction of ferroelectric domains, thus the
macro reflects the remnant polarization, so that the material is
polarized. When the electric field is reversed, the remnant
polarization disappears, so the coercive field can be regarded
as the minimum field strength of the domain.
Fig. 15. Polarization of all compositions with 5, 6, 8, 10, 12 and 15% of PMnN
TABLE II. Remanent Polarization and Coercive Field
System
Remanent Polarization
(Pr/µC/cm2
)
Coercive Field
(kV/cm)
5% 11.50 14.63
6% 8.36 20.21
8% 10.99 11.03
10% 21.85 22.75
12% 9.42 13.81
15% 9.82 19.78
IV. EFFECTS OF ANNEALING ON DIELECTRIC AND
PIEZOELECTRIC PROPERTIES OF BSPT-PMNN CERAMICS
Microstructure of ceramics sturdily depends on annealing
process and it defines the ceramic’s physical performance.
Density, homogeneity and grains size in the annealing steps are
significant to attain the preferred electrical properties [7]. It
will influence the crystallographic phase, microstructure and
electrical properties of the resultant thin films [8-10]. The
temperatures of pre-annealing and annealing were determined
by the results of the thermal analysis and XRD, respectively.
A. SEM Analysis After Annealing
The microstructure of the compositions after annealed was
investigated by SEM. The SEM graphs of the samples (a) 5%;
(b) 6%; (c) 8%; (d) 10%; (e) 12% and (f) 15% of PMnN
respectively annealed at 800°C for 2h are shown in Fig. 16.
SEM microstructures of the annealed body are not showing any
second phase. They have clearly displayed different
microstructures; grain size of each sample is (a)3.2; (b)2.8;
(c)7.2; (d)4; (e)2.2; and (f)2; in micrometers respectively and
grain boundaries of sample (c)8% are more clear and
significant. They showed various microstructures clearly with
annealed temperature of sample 8% of PMnN, grain size was
increased and grain boundaries of sample (a) and (d) are also
clear and significant. Nevertheless the microstructures of the
every composition are uniform.
B. Piezoelectric Properties After Annealing
The piezoelectric coefficient d33 was noted from one-day
aged samples by a Berlincourt d33 meter (IAAS ZJ-2, Beijing,
China). All the properties displayed highest magnitude of
compositions near MPB boundary that is 0.64PT. As compared
to BS-PT ceramic, there is noticeable decrease in the
magnitude of d33 [11]. Also there is noticeable increase in the
magnitude of d33 after annealing for the samples of BSPT and
other compositions of 5, 6 and 12% of PMnN respectively. At
these compositions the magnitude of d33 are 490, 320, 270,
230, 210, 120 and 165pC/N respectively as shown in Fig. 17.
The electromechanical coupling factor kp was measured
using an impedance analyzer (Model HP 4194) based on the
IEEE standards [12]. Electromechanical resonance properties
determined after annealed at 800°C for 2h from the admittance
spectra at low field ac drive are shown in Fig. 18, as a function
of PMnN composition. Near the MPB boundary, 64% PT the
sample of 10 and 12% of PMnN there is a bit increase in kp
factor, at 5 and 15% there is decrease in kp factor as clearly
shown in Fig. 18.
The factor Qm was calculated by using an impedance
analyzer. As a consequence of the higher amount of coercive
field and Curie temperature of BS-PT, this composition is
predictable as high power piezoelectric material, if we can
increase the value of quality factor Qm. Near the MPB
boundary, 64% PT the sample of 5, 10, 12 and 15% of PMnN
there is clearly increment in quality factor (Qm) as shown in
Fig. 19. So, the replacement of Pb(Mn1/3Nb2/3)O3 is effective to
improve the mechanical quality factor Qm of BS-PT system.
279
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-288-320.jpg)
![TABLE III. Remnant Polarization and Coercive Field after annealing
System
Remnant Polarization
(Pr/µC/cm2
)
Coercive Field
(kV/cm)
5% 23.30 -14.66
6% 24.93 -17.02
10% 25.63 -20.55
12% 17.35 -17.01
15% 21.29 -21.17
BSPT 33.98 -17.33
V. CONCLUSION
BSPT-PMnN dense ceramic samples were prepared by the
columbite pre-production process. XRD results show that, at
the appropriate sintering temperatures, two systems were
obtained with pure perovskite structure. In systems of
0.05PMnN-BSPT, 0.06PMnN-BSPT, 0.08PMnN-BSPT,
0.1PMnN-BSPT, 0.12PMnN-BSPT and 0.15PMnN-BSPT
when the content of PT is 64% is the tetragonal-phase
existence of morphotropic phase boundary region. SEM
images show that the ceramic sample’s grain sizes are in
uniform distribution at the appropriate sintering temperatures,
the grain size is about 1µm. When the content of PMnN is
low, the PMnN-BSPT system shows the characteristics of
typical relaxor ferroelectrics, with the increase of PMnN
content, gradually transition from ferroelectric relaxor to
normal ferroelectrics. Due to low content of PMnN, PMnN-
BSPT system does not show obvious relaxation
characteristics. With the increase of PMnN content, the
remnant polarization Pr, first increased and then decreased.
The coercive field system gradually increases with the
increase of PMnN content. The piezoelectric properties of the
optimal compositions of 0.1PMnN-BS-PT and 0.05PMnN-
BSPT system in the morphotropic phase boundary region are:
d33=250pC/N, kp =0.54 and the Curie temperature of
Tc=373°C. The piezoelectric properties of the optimal
composition of 0.05PMnN-0.95BSPT system in the
morphotropic phase boundary region are better than other
optimal compositions.
After annealing at 800°C for 2h, the resident strain can be
quited and the order degree of lattice may be higher than that
without annealing process. Annealing process may affect the
domain size and domain switch, domain size will be larger
after annealing process. The value of Qm increase evidently
after annealing process.
ACKNOWLEDGMENT
I would like to thank my father and teachers for their kind
and humble guidance from the very start to the end of the paper
writing. I would not have accomplished this without their
constant support and feedback; they made it a thorough
learning and pleasant experience for me. I admire the guidance
and assistance provided by research associates and fellows.
REFERENCES
[1] Ueha, S., “Ultrasonic motors: theory and applications,” Vol. 29.
1993: Oxford University Press, USA.
[2] Chen, H., “The influence of Pb(Mg1/3Nb2/3)O3 content on the
piezoelectric and dielectric properties of PMNN-PZT system
near the MPB. in Applications of Ferroelectrics,” 2002. ISAF
2002. Proceedings of the 13th IEEE International Symposium
on. 2002. IEEE.
[3] DU, H. l., “Effect of sintering temperature and composition on
microstructure and properties of PMS-PZT ceramics,”
Transactions of Nonferrous Metals Society of China, Vol. 16:
pp.165-169, 2006.
[4] Weis, R. and T. Gaylord, “Lithium niobate: summary of
physical properties and crystal structure,” Applied Physics A,
Vol. 37(4), pp. 191-203, 1985.
[5] Korzunova, L. V., and L. A. Shebanov, “New perovskite-like
high temperature ferroelectrics,” Ferroelectrics, vol. 93.1 pp.
111-115, 1989.
[6] B. Frit and J. P. Mercurio, “The crystal chemistry and dielectric
properties of the Aurivillius family of complex bismuth oxides
with perovskite-like layered structures,” Journal of Alloys and
Compounds, vol. 188, pp. 27-35, 1992.
[7] M. Ghasemifard, “Dielectric, piezoelectric and electrical study
of 0.65PMN-0.20PZT-0.15PT relaxor ceramic,” The European
Physical Journal Applied Physics, Vol. 54, No. 2 , pp. 20701-
20707, 2011.
[8] Laughlin, Brian, Jon Ihlefeld, and Jon‐Paul Maria, “Preparation
of sputtered (Bax, Sr1−x)TiO3 thin films directly on copper,”
Journal of the American Ceramic Society, vol. 88.9.pp 2652-
2654 2005.
[9] Malic, Barbara, “Processing and dielectric characterization of
Ba0.3 Sr0.7 TiO3 thin films on alumina substrates,” Journal of the
European Ceramic Society 27, pp. 2945-2948, 2007.
[10] Tahan, Danielle M., Ahmad Safari, and Lisa C. Klein.,
“Preparation and characterization of BaxSr1‐x, TiO3 thin films
by a Sol‐Gel technique,” Journal of the American Ceramic
Society, vol. 79.6, pp.1593-1598, 1996.
[11] Chen, Jun, “Temperature dependence of piezoelectric properties
of high-TC Bi(Mg1/2Ti1/2)O3-PbTiO3,” Journal of Applied
Physics, vol. 106.3 pp.034109-034109, 2009.
[12] IEEE “Standard on piezoelectricity,” IEEE Standard 176-1978
(IEEE, New York, 1978).
281
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-290-320.jpg)
![A Comprehensive Study on QoS for Mobile
Ad Hoc Network
Abdur Rasheed Irfan Nazeer Fayyaz Khalid
Department of Computer Science Department of Computer Science Department of Computer Science
University of Gujrat University of Gujrat University of Gujrat
Gujrat, Pakistan Gujrat, Pakistan Gujrat, Pakistan
abdur.rasheed@live.com irfan6200@yahoo.com fayyazrao@yahoo.com
ABSTACT----A Mobile Ad Hoc Network (MANET) is a
collection of nodes which has no centralized
administration and without the assistant of base station.
All the nodes have equal rights and have their own
resources in MANET. QoS is a challenging issue now-a-
days due to bandwidth constraints and dynamic
topology in MANET. A lot of work done on supporting
on QoS in internet and other networks but most of them
is not suitable for MANET. Most of work is about the
QoS routing, QoS Mac and resource reservation, there
is a little work discuss on QoS models and its routing
protocols. In this article, we review the QoS models and
share some knowledge about QoS routing protocols and
other related work on QoS in MANET. The purpose of
this article is to give the appropriate knowledge about
QoS for MANET at the same platform.
Keywords: QoS, Mobile Ad Hoc Networks
I. Introduction
Wireless network has become a popular network
now-a-days. Wireless communication has been
making our life easy and comfortable. Wireless
communication has been categorized into two
primary models; one is traditional wireless
network in which a fix infrastructure is used to
communicate between one hop to another. In
conventional wireless network, much of the
nodes are mobiles and connected with the help of
fixed backbone nodes using wireless
medium(radio waves). In this type wireless
network a base station is used for
communication. But in some situation (i.e. storm
disaster, battle field) there is no fix infrastructure
is used. Then we used a specific type of network
which is called “Mobile Ad Hoc Network”. A
mobile ad hoc network (MANET) [1][2], is a
dynamic wireless system in which all nodes are
mobile and nodes can move in any direction,
independent of each other. According to [3],
Mobile Ad hoc Networks are class of networks
whose contains characteristics is,
Physical characteristics –As MANET is
a wireless network in which nodes can
create link in any time, bandwidth and
Delay due to the nodes are mobile.
Organization – As mobile nodes are
distributed so resource estimation is not
predefined in MANET
Dynamic Topology – As the nodes have
dynamic topology so it has no fixed
architecture.
MANET is the combination of different mobiles
nodes which does not have any server, router or
access points for communication. It has low cast
wireless network and easy for deployment.
MANET is used in different situations in our life
i.e., Military operations, Rescue search
operations, communication between mobiles
vehicles in smart transportation, mobiles phones,
laptops and smart watches etc. As shown in Fig.1
MANET provides a flexible and seamless access to
internet. Due to in increasing popularity in MANET,
282
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-291-320.jpg)
![QoS has become the main task of research in
MANET. QoS is a challenging issue now-a-days due
to dynamic topology and bandwidth constraints. A lot
of work is done on supporting on QoS for internet
and other networks but most of them are not suitable
for MANET. Due to autonomous wireless network,
MANET has its own protocols.
In this literature, we discuss QoS for MANET
includes QoS models and also discuss some
important protocols in MANET. There are many
researches on QoS on the internet. Due to this, it is
difficult to get appropriate knowledge on QoS in a
specific location. This paper provides a
comprehensive knowledge on QoS models for
MANET and some QoS routing protocols. In first
part, we shall discuss some QoS models which are
specified an architecture in which we discuss some
services that can be used in MANET. In second part,
we discuss some QoS protocols for MANET. These
protocol are divided in the based on route discovery,
metrics and network architecture. Then we gives a
comparison of these protocols on the bases on route
discovery, metrics and network architecture.
II. QoS models and MANETs
QoS model defines the architecture which provides a
platform to different services; which is used in
networks. QoS models use different situations like
dynamic topology etc. QoS model consider the
exiting internet architecture for MANET.
In this section, we explain the different QoS models
especially DiffServ and one of the newly proposed
QoS Model FQMM.
A. IntServ and MANETs
The idea of IntServ[4] is taken from the wired
network architecture and B-ISDN i.e. have the virtual
circuit mechanism. Virtual connection is maintains
through signaling protocol RSVP (Resource
Reservation Protocol)[4].RSVP is very important
protocol in all QoS support routers. IntServ provides
two types of services; one is Guaranteed [5] Service
and the Other is Controlled Load Service [6]. Here
some pros and cons of IntServ mentioned below:
a) Scalability
IntServ provides per-flow granularity so when state
information is increases the number of lows is also
increased. This requires storing a lot storage capacity
and processing is increases, it results a problem that's
called scalability problem of IntServ. The scalability
problem is very less in current MANETs because
there is less number of users, less bandwidth and
small size of network. On the other Hand, in future
the use of MANETs is increasing rapidly and the size
and speed of the MANETs increases so the
Scalability problem is also increased with time so
IntServ is not suitable in MANETs.
b) Signaling
Signaling protocols like Resource Reservation
protocol have three steps; connection maintenance,
connection tear-down and connection establishment.
Corson[7] assumes and then suggest that a larger
number of links is given to the network to control
overhead .For MANETs with link capacities, the
overheads of connection maintenance and dynamic
topology usually a large for connection establishing
so that RSVP signaling protocol is not suitable in
MANETs.
c) Router mechanisms
IntServ is required a large number of routers so that
all router have packet scheduler, classifier, admission
control routine and RSVP that's why it is undesired
able in MANETs.
B. DiffServ and MANETs
Like ATM and IntServ, the tenet of DiffServ [8] is
used a relative-priority scheme to soften the hard
requirements of hard QoS models. DiffServ is
specially designed for the deficiencies of IntServ and
RSVP model like scalability problems in internet
backbone [9]. DiffServ is defined limited classes for
overcome this problem of IntServ. Much kind of
283
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-292-320.jpg)
![services are supported in DiffServ model which are
following [10]:
a) Premium Service
Premium Service is considered to give the end-to-end
bandwidth services, low jitter, low loss and low delay
[11].
b) Assured Service
Assured service is used to provide better reliability
for application. Other, it is more qualitative service
than quantitative and easy to implement.
c) Olympic Service
Three types of services are provided in Olympic
services which are Gold, Silver and Bronze with
decreasing quality [9].
Premium service is not suitable in MANETs because
it implementation in dynamic environment is not
possible. DiffServ is considered a defined solution to
MANETS QoS model because it provides Assured
Service and it is lightweight in inner router which is
suitable to MANETs. Because of this, DiffServ is
designed for wired networks and have some
problems in MANETs which are follows
DiffServ is not sure in to the boundary nodes for
this; it requires a lot of storage capacity.
DiffServ provides a Service Level Agreement
(SLA) which implementation is not possible in
MANETs. SLA is an agreement between
customer and the internet service Provider (ISP)
to receive Differentiated Services. SLA obeys
the partial or whole traffic rules [12]; but in
MANETs, it is not possible to negotiate the
traffic rules.
C. Flexible Quality of Service Model (FQMM)
and MANETs
FQMM is the first QoS model which is preferred in
MANETs. FQMM is proposed in [13] and it tries to
take advantage from IntServ and DiffServ models
both. FQMM is designed for average size MANETs
which comprises less than 50 nodes using at non-
hierarchical topology. However, FQMM is a hybrid
approach because it has combined features from
IntServ and DiffServ. In FQMM infrastructure, it has
types of Nodes like DiffServ which follows:
Ingress Node
Interior Node
Egress Node
Ingress node is a source node which is a mobile node.
Interior node is the intermediate node that sends the
data of other mobile nodes. Egress node is the node
which is received data and it is the destination node.
Note that the role of Ingress node is changing due to
change in its position and network traffic. Resource
determination and allocation is done to the various
mobile nodes by using provisioning in FQMM.
Provisioning is at top level and based on utilization of
the network in present internet. Provisioning is used
to realize ideal per-flow by RSVP in IntServ and is
used in DiffServ as per-class by human agents rather
than RSVP or other. But on the other hand, FQMM
introduced a hybrid scheme as per-flow as well as per
class. Traffic with highest priority is given per-flow
provisioning while other priority is based on per-
classes. Although it is the first efficient model in
MANETs but there is further need to study that how
sessions is entertained as per-flow due to bandwidth
constraints in MANETs.
III. QoS Aware-Routing Protocols
In this section, we discuss some QoS aware routing
protocols. After this, we discuss their functionality
and then see their comparisons with each other with
different aspects. The QoS Aware-Routing Protocols
are follows:
A. Optimized link stat routing Protocol
OLSR [14, 15] is a proactive routing protocols and
the idea of OLSR is given by IETF MANET[3]
working group. OLSR is the further study of Link
stat routing protocols. The idea of OLSR is based on
MPRs (Multi Point Relays).Every MPR have two
hops distance which is the part of communicating
nodes. If a node has no MPR then it cannot transfer
data further. Consider that MPR have larger distance
than a node which is not a MPR then we may miss
some good quality links and that's reason, it is
difficult to achieve a good quality of service in
OLSR.
284
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-293-320.jpg)
![B. Predictive Location Based QoS Routing Protocol
PLBQR [16] is a proactive routing protocol which is
used in MANET.PLBQR is totally based on location
prediction of a node a particular instant in ad hoc
wireless networks. Location prediction is used to find
the geographical location of a node at particular
instant 'tf' when the receiver node picks a packet in
future. The value of 'tf' is calculated by calculating
delay in communication. No resources are reserved
during the propagation from source to destination but
QOS is controlled. This QoS routing protocol is
helped the nodes to update location prediction and
delay during communication.
C. Ticket Based QoS Routing Protocol
Ticket Based QoS Routing protocol [17] is a reactive
protocol which is used in MANETs. The basic idea
of Ticket Based QoS Routing protocols is to use
tickets to find the best path which may delay
constrained or bandwidth constrained. This protocols
is used two types of tickets, one is yellow ticket and
the other is green. The purpose of yellow tickets is to
find the best path with delay/bandwidth constraints.
The other green ticket purpose is to find the low cost
route. The source node has a lot the number of tickets
to establish a QOS aware route. The tickets may be
less or more it depends on the route constraints to
find the best path. The Drawback of this routing
protocol is that the neighbor nodes incorporated to
each other with resource estimation that’s why it
requires more memory. When a route is established
to the destination node acknowledged even resource
reservation is established.
D. Ad Hoc QoS on Demand Routing
The idea of AQOR [18] protocol is to find the best
route which is fulfills the constraints of end to end
delay with specific bandwidth. In route discovery, the
route request is sends to nearby node and attaches the
constraints with the packet and if both nodes satisfy
then this request is rebroadcasted to the next hop with
some expiry time. If the node is not received the
reply in given time then entry will be deleted. If the
nodes do not want to delete path the only
intermediate nodes will reply. Once the route is found
then resource reservation is takes place on the found
route. If the route is deleted the new route request is
sent.
E. Adaptive QoS Routing Algorithm
ADQR [19] is a reactive protocol. Adaptive QoS
Routing algorithm protocol is used to establish a
route between two nodes which have longer life time.
In these protocols, the route breakage and route
creation is bases on signal power. Bandwidth is
considered to the lower layer. Route maintenance
contains two ways. One is pre-routing and the other
is rerouting. Pre-routing is used when the value of the
threshold is greater than signal power and rerouting is
used when the signal power is very low. For example
of this protocol is; if a node want to communicate
with other node then it sends a RREQ packet to
nearby node and then intermediate node attached his
address to this packet and forward to the next. If the
packet is reached to the exact destination then
destination node checked that the signal power.
Signal power ensures that the route is enough or it
disjoints with others .QoS_Reserve packet is sent to
the current route back and QoS Acknowledgment is
sent to the source. [3]
F. Trigger-based Distributed-QoS Routing
Protocol
TDR [20] is a reactive protocol which is a location
based protocol. In this protocol every host node
contains two types of databases. One is for local
neighbor and other is activity based. Hosts are passed
out the mobility and location information and the
neighbor nodes is stored the power level in their local
database. Activity-based data base is stored the every
session routing information and updated in every
session. Route discovery is done through the
neighboring power level and compared this with
threshold time. Route maintenance is done like
ADQR protocol [3].
G. Bandwidth Estimation QoS Routing Protocol
BEQR [21] is a reactive protocol which is inspired by
AODV routing protocol [23].The main objective of
this protocol is to give soft QoS for service with the
bandwidth constraint. In this two schemes is used one
is Feedback scheme and the other is Admission
285
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-294-320.jpg)
![scheme. In Feedback scheme is used for updating
constraints when a node have no bandwidth its own.
Admission Scheme is used for route discovery with
satisfied bandwidth constraints. In this protocol
bandwidth is calculated with ratio of free and busy
times. Resource reservation is not considered. It is
design for soft QoS not hard.
H. Core Extraction Distributed Ad hoc Routing
Protocol
CEDAR [23, 3] was proposed by R. Siva Kumar
[23].CEDAR is a QoS aware routing protocol and it
is a hybrid routing protocol. In this protocol nodes
are selected as Core node and they are responsible for
route maintenance, computation and QoS providing
by using distributed algorithm. Link state information
depends on network state whether it is dynamic or
stable.
Route computation is on-demand based and Cedar
contain following three features which is very
important:
1. Destination location discovery and
establishing the core path to destination
2. Core nodes provides the stable and short
path between source to destination and also
a core path guideline
3. If the path is break then it re-established as
well as typology changes.
CEDAR considers that the resource is reserved.
[3]The route maintenance in CEDAR is done as
source initiated route maintenance or dynamic route
maintenance.
I. INSIGNIA
INSIGIA [24] is a hybrid protocol which is proposed
for adaptive services in ad-hoc networks. Adaptive
services entertains that types of services which
requires minimum QoS guarantee (like minimum
bandwidth) and it varies when resources is sufficient.
INSIGNIA in-band signaling is the very important
component which is entertained fast reservation to
deliver the adaptive services[25].Routing module is
responsible for finding routes between nodes and
sends the packets to the next node(intermediate).In
any time of transmission when topology is changed it
is established a new path. Every data packet has a
optimal QoS field and admission control module is
responsible for resource reservation; it is refreshed
periodically. If an application is used to want
minimum bandwidth then it bandwidth indicator is
"MIN", if the application is used to want maximum
bandwidth then indicator indicates "MAX".
J. Forward Algorithm
FA [26] belongs to the on-demand routing protocols
which is used on demand algorithms with bandwidth
as QoS constraints parameter. FA is the modified
version of AODV [22] is used for route discovery
and some additional information is attached before
sending to the next hope. During route discovery, the
value of local maxima is calculated and forward [27].
FA is used in many other reactive protocols like
TORA [28] and DSR [29] for route discovery. The
Route maintenance is managed by using old routes in
it.
286
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-295-320.jpg)
![IV. Comparison of QoS Aware-Routing Protocols
A comparison of QoS aware-routing protocols is given below:
Routing Protocols QoS Support Redundant Route Route Maintenance Route Breakage
Prediction
Network
Architecture
Route Discovery
OLSR X X X Hierarchical Proactive
PLBQR X X Location
Prediction
Proactive
Ticket Based X Flat Reactive
AQOR X X Flat Reactive
ADQR Flat Reactive
TDR X Location Based Reactive
BEQR X X X Flat Reactive
CEDAR X X Hierarchical Hybrid
INGIGNIA X X Flat Reactive
FA X X Flat Reactive
Conclusion
In first part of this paper, we have described three
QoS Models (IntServ, DiffServ and FQMM).All
three models are suitable for different situations. All
three have some pros and cons; but FQMM is a
flexible QoS model for MANETs. To our knowledge,
FQMM was the first such QoS model proposed for
MANETs. It provides a hybrid provisioning scheme
and a relative and adaptive traffic profile. In second
part of this paper, we described some important QoS
aware-routing protocols in ad hoc wireless networks
which is an active research now-a-days. In this paper,
we discussed some characteristics of the QoS routing
protocols and compare the characteristics of these
protocols in the form of table. The aim of this review
is to give an appropriate knowledge on QoS for
MANETs at the single paper.
References
[1] Loay Abusalah, Ashfaq Khokhar, and Mohsen Guizani “A
Survey of Secure Mobile Ad Hoc Routing Protocols” IEEE
communications surveys & tutorials, VOL. 10, NO. 4, FOURTH
QUARTER, 2008, pp 78-93.
[2] Sunil Taneja and Ashwani Kush” A Survey of Routing
Protocols in Mobile Ad Hoc Networks”, International Journal of
Innovation, Management and Technology, Vol. 1, No. 3, August
2010.
[3] Lei Chen and Wendi B. Heinzelman, “A Survey of Routing
Protocols that Support QoS in Mobile Ad Hoc Networks”, IEEE
Network November/December 2007.
[4] R. Braden, D. Clark, and S. Shenker. Integrated Services in the
Internet architecture - an
Overview. IETF RFC1633, June 1994.
[5] S. Shenker, C. Partridge and R. Guerin “Specification of
Guaranteed Quality of service” RFC 2212 September 1997
[6] J. Wroclawski, “Specification of the Controlled-Load Network
Element Sevice ” RFC 2211, Sept.1997
[7] M.S. Corson. Issues in supporting quality of service in mobile
ad hoc networks. In IFIP 5th Int.Workshop on Quality of Service
(IWQOS'97),May 1997.
[8] S. Blake. An architecture for Di_erentiated Services. IETF
RFC2475, December 1998.
[9] Xipeng Xiao Nad Lionel M. Ni “Internet QoS: a big Picture,”
IEEE Network magazine , March 1999.
[10] Kui Wu and Janelle Harms “Quality of Services in Mobile Ad
Hoc Networks”
[11] K. Nichols, V. Jacobson, and L. Zhang. A two-bit
Di_erentiated Services architecture for the Internet. IETF
RFC2638, July 1999
[12] S. Blake, “An Architecture for Differentiated Services,” IETF
RFC 2475, December 1998.
[13] H. Xiao , W.K.G. Seah, A. Lo and K. C. Chua, “A flexible
Quality of Service Model for Mobile Ad hoc Networks” IEEE
VTC2000-spring . Tokyo, Japan May 2000.
[14] Y. Ge, T. Kunz, and L. Lamont, “Quality of Service Routing
in Ad-Hoc Networks Using OLSR,” Proc. 36th Hawaii Int’l. Conf.
Sys. Sci., Jan. 2003.
[15] P. Jacquet, P. Muhlethaler, A. Qayyum, A. Laouiti, L.
Viennot, T. Clauseen, "Optimized Link State Routing Protocol
draft-ietf-manet-olsr-05.txt", INTERNET-DRAFT, IETF MANET
Working Group.
[16] S.H.Shah and K.Nahrstedt, “Predictive Location- Based QoS
Routing in Mobile Ad Hoc Networks,” Proceedings of IEEE ICC
2002,vol.2, pp.1022-1027, May 2002
287
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-296-320.jpg)
![[17] S. Chen and K. Nahrstedt, “Distributed Quality-of-Service in
Ad Hoc Networks,” IEEE JSAC, vol. 17, no. 8, 1999.
[18] Q. Xue and A. Ganz, “Ad hoc QoS On-demand Routing
(AQOR) in Mobile Ad hoc Networks,” J. Parallel and Distrib.
Comp., vol. 62,no. 2, Feb.2003, pp. 154–65.
[19] Y. Hwang and P. Varshney, “An Adaptive QoS Routing
Protocol with Dispersity for Ad-hoc Networks,” Proc. 36th Hawaii
Int’l.Conf. Sys. Sci., Jan. 2003.
[20] S. De et al., “Trigger-Based Distributed QoS Routing in
Mobile Ad Hoc Networks,” ACM SIGMOBILE Mobile Comp.
and Commun.Rev., vol. 6, no. 3, July 2002, pp. 22–35.
[21] Lei Chen and Wendi B. Heinzelman, “QoS-Aware Routing
Based on Bandwidth Estimation for Mobile Ad Hoc Networks,.
IEEE Journal on Selected Areas in Communications,
23(3):561_572, March 2005.
[22] C. E. Perkins and E. M. Royer. “Ad-hoc On-Demand Distance
Vector Routing”, In Proceedings of the 2nd IEEE Workshop on
Mobile Computing Systems and Applications, pages 90_100, New
Orleans, LA, Feb. 1999
[23] R. Sivakumar, P. Sinha, and V. Bharghavan, “CEDAR: a core
extraction distributed ad hoc routing algorithm” IEEE Journal on
Selected Areas in Communication, Vol. 17, no. 8, pp 1454-65
August1999.
[24] X. Zhang, S.B Le, A. Gahng-Seop and A.T Campbell,
“INSIGNIA: an IP based quality of service framework for mobile
ad hoc networks.”Journal of Parallel and Distributed Computing,
Vol. 60, no. 4,pp. 374-406, April 2000.
[25] C.Siva Ram Murthy,B.S.Manoj, “Ad Hoc Wireless Networks
Architectures and Protocols” Prentice Hall Communications
Engineering and Emerging Technologies Series.
[26] Chenxi Zhu and M. Scott Corson. “QoS routing for mobile ad
hoc networks”. Technical report, Flarion Technologies,
Bedminster, New Jersey 07921, 2002.
[27] Philipp Becker, “QoS Routing Protocols for Mobile Ad-hoc
Networks – A Survey”, TU KL Technical Report 368/08, A
publication by University of Kaiserslautern
[28] V.D. Park and M.S. Corson, “A Highly Adaptive Distributed
Routing Algorithm for Mobile Wireless Networks”, Proceedings of
IEEE INFOCOM 1997, pp.1405-1413, April 1997.
[29] D.B.Johnson and D.A. Maltz, “Dynamic Source Routing in
Ad Hoc Wireless Networks”, Mobile Computing, Kluwer
Academic Publishers,vol.353,spp.153-181,1996.
288
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-297-320.jpg)
![Comparison of Maximum Likelihood Classification
Before and After Applying Weierstrass Transform
Muhammad Shoaib, Zaka Ur Rehman, Muhammad Qasim
Abstract— The aim of this paper is to use Maximum
Likelihood (ML) Classification on multispectral data by means of
qualitative and quantitative approaches. Maximum Likelihood is
a supervised classification algorithm which is based on the
Classical Bayes theorem. It makes use of a discriminant function
to assign pixel to the class with the highest likelihood. Class
means vector and covariance matrix are the key inputs to the
function and can be estimated from training pixels of a particular
class. As Maximum Likelihood need some assumptions before it
has to be applied on the data. In this paper we will compare the
results of Maximum Likelihood Classification (ML) before apply
the Weierstrass Transform and apply Weierstrass Transform
and will see the difference between the accuracy on training
pixels of high resolution Quickbird satellite image. Principle
Component analysis (PCA) is also used for dimension reduction
and also used to check the variation in bands. The results shows
that the separation between mean of the classes in the decision
space is to be the main factor that leads to the high classification
accuracy of Maximum Likelihood (ML) after using Weierstrass
Transform than without using it.
Index Terms— Maximum Likelihood Classificaion (ML),
Weierstrass Transform, Bayes Theorem, Supervised
Classification, Principle Component Analysis (PCA).
I. INTRODUCTION
Satellite images are widely used for various purposes and
with the advancement in Satellite technology every country
wants to solve their problem by using satellite images. Satellite
images can be used in various disciplines such as Metrology,
Environmental sciences, GIS, Geology, Geophysics,
Engineering and Construction, Defense and Intelligence and
the list goes on.
Satellite images contains maximum information about our
planet and main problem is that how can we retrieve the useful
information satellite images. Researchers can have several
problems such how to classify the land cover types, depth of
Aerosol particles in atmosphere, Forest Climates, Flood
predictions and many more. Researcher solve these problems
by analyzing the satellite images.
In this article our main aim is to classify the land cover
types by using Maximum Likelihood Classification algorithm.
First we classify land cover types without using Weierstrass
Transform (WT) and then we compare the results by applying
Weierstrass Transform (WT).
Maximum Likelihood (ML) is a supervised classification
method derived from the Bayes theorem, which states that the a
posteriori distribution P(i|θ), i.e., the probability that a pixel
with feature vector θ belongs to class i, is given by:
where P(θ|i), is the likelihood function, P(i) is the a priori
information, i.e., the probability that class i occurs in the study
area and P(θ) is the probability that is observed, which can be
written as:
where M is the number of classes. P(θ) is often treated as a
normalization constant to ensure ∑P(i|θ) sums to 1. Pixel x is
assigned to class i by the rule:
Each pixel is assigned to the class with the highest likelihood
or labeled as unclassified if the probability values are all
below a threshold set by the user [1].
Maximum Likelihood (ML) assumes that the data within a
given class i obeys a multivariate Gaussian distribution.
Frist we implement Maximum Likelihood (ML) classification
algorithm without checking the assumption of normality and
then we apply Maximum Likelihood (ML) classification after
using Weierstrass Transform on the image data and we have
seen that after applying Weierstrass Transform our data is
normally distributed. The accuracy after applying Weierstrass
Transform (WT) is increased because the distribution of the
data is normal because for ML classification data has to be
normal.
II. STUDY AREA
We use Quickbird satellite image for our analysis.
Quickbird is high-resolution commercial satellite which is
owned by DigitalGlobe and launched in 2001. Quickbird uses
Aerospace's Global Imaging System 2000 (BGIS 2000).The
Quickbird satellite collected panchromatic (black and white) at
61cm resolutin and multispectral imagery at 2.44-(at 450 km)
to 1.63-meter (at 300km) resolution, as orbit altitude is lowered
during the end of mission life [2].
The Quickbird satellite image which we use for our
analysis is sample image of Rome Italy with coordinates
with four bands which are
blue (450-520nm), green (520-600nm), red (630-690 nm) and
NIR (760-900 nm).
289
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-298-320.jpg)
![In the figure 2 we can see that the distribution of each band
is not Gaussian and we know that for ML classification
algorithm the distribution of each has to be Gaussian. In this
case distribution of first three bands are positively skewed
while the histogram of band 4 is looks like Gaussian but not
exactly Gaussian. So as our aim is to check the accuracy of
classification before making data Gaussian and after making it
Gaussian. We use Weierstrass Transform to make the
distribution Gaussian.
B. Weirstrass Transform
As we have seen that distribution of our data is not follow
Gaussian distribution so make that Gaussian we have to use
transformation to make pixel values Gaussian.
Researchers uses various types of transformation to make
the data Gaussian. In this paper we use different type of
transformation called Weirstrass Transform.
The Weirstrass transformation, also known as the Gauss
transform, the Gauss Weirstrass transform and the Hille
transform, is intimately related to the solution of the heat
equation for one- dimensional flow. It is also a special case of
convolution-transform, yet it is considered a singular case of
convolution-transform theory developed by Hirschmann and
Widder [3].
The Weirstrass transformation F(x) of a function f(y)
defined as:
the convolution of f with the Gaussian function .
Instead of F(x) we also write W[f](x). Note that F(x) need not
exist for every real number x, because the defining integral
may fail to converge [3].
When Gaussian filter modifies the input signal by using
Gaussian function; this is known as Weierstrass
transformation.
We can write Gaussian function mathematically in 2D as
follow:
where ρ is the correlation between X and Y and where σ_x>0
and σ_y>0. In this case,
In the bivariate case, the first equivalent condition for
multivariate normality can be made less restrictive: it is
sufficient to verify that countably many distinct linear
combinations of X and Y are normal in order to conclude that
the vector [X Y]′ is bivariate normal.
Gaussian filtering is used to blur images and remove noise
and detail. Gaussian filter uses the above Gaussian function
which is given in equation 1.
After applying Gauss transform or Wierstrass transform we
can compare original false color composite (FCC) image with
transformed false color composite (FCC) image.
(a) original false color composite (FCC) image
(b) Transformed false color composite (FCC) image
Fig. 3.
In figure 3 we can see the clear difference between original
FCC image with transformed FCC image. After applying
transformation we can see that the transformed image is more
blur than original image.
291
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-300-320.jpg)
![In the above figure we can see that band 3(Red) and band
4(NIR) have same information so we have to neglect band 3
or band 4 from our analysis because they have same
information in the image for this purpose we will use
classification on three bands that are Blue, Green and NIR.
D. Training Samples
Training areas were established by choosing one or more
polygons for each class. Pixels fall within the training area
were taken to be the training pixels for a particular class. In
order to select a good training area for a class, the important
properties taken into consideration are its uniformity and how
well they represent the same class throughout the whole image
[4].
We use Stratified random sampling for selecting the
training pixels. A stratified random sample is a population
sample that requires the population to be divided into smaller
groups, called 'strata'. Random samples can be taken from each
stratum, or group.
V. RESULTS
Maximum Likelihood considered to be an most accurate
algorithm when we compare it to classical algorithm such as
parallelepiped classification algorithm. But as we know that
ML classification algorithm assumes that distribution of each
band has to be Gaussian. we had seen that our band's
distribution was not Gaussian so we make distribution
Gaussian by applying Gauss transform or Weirstrass transform.
Now we will check what is the effect on accuracy of
classification classes when the distribution is Gaussian and
when not Gaussian.
A. Classification Before Weirstrass Transform
The outcome of ML classification after assigning the
classes with suitable colors, is shown in Fig.5 (a): Grass
(green), Urban Area (blue), Roads(yellow), Soil (cyan), Trees
(red) and Unclassified (Black). The areas in terms of
percentage and square meters were also computed; the classes
with the largest area is Urban Area. In this case the distribution
of our data is not Gaussian and we can clearly see in the Fig.5.
(a) that some buildings are wrongly classified as Roads
because we are applying ML classification on non normal data.
Class Color Area in % Area in Meters²) Color
Unclassified: [Black] 6.44% (26,783.6400 Meters²)
Trees [Red] 14.20% (59,095.0800 Meters²)
Grass [Green] 13.63% (56,719.0800 Meters²)
Urban Area [Blue] 31.45% (130,868.2800 Meters²)
Roads [Yellow] 30.31% (126,133.9200 Meters²)
Soil [Cyan] 3.97% (16,532.6400 Meters²)
(a)
Class Color Area in % Area in Meters²) Color
Unclassified: [Black] 10.30% (42,851.1600 Meters²)
Trees [Red] 13.75% (57,232.0800 Meters²)
Grass [Green] 10.34% (43,005.2400 Meters²)
Urban Area [Blue] 44.79% (186,373.0800 Meters²)
Roads [Yellow] 19.39% (80,677.4400 Meters²)
Soil [Cyan] 1.44% (5,993.6400 Meters²)
(b)
Fig. 5. (a) Maximum Likelihood classification an original image, (b) Maximum Likelihood classification on transformed image
293
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-302-320.jpg)
![TABLE I. CONFUSION MATRIX OF ORIGINAL CLASSIFIED IMAGE
Class Color Trees Grass Urban Area Roads Soil Color
Unclassified [Black] 0.4 0.05 7.84 2.33 0
Trees [Red] 94.38 8.06 0.94 5.28 0
Grass [Green] 1.64 88.71 2.66 0.03 3.41
Urban Area [Blue] 0 1.06 70.71 1.19 2
Roads [Yellow 3.58 0 4.61 91.17 0
Soil [Cyan] 0 2.11 13.23 0 94.59
Total 100 100 100 100 100
Ground Truth (Percent)
Overall Accuracy = (18318/21394) = 85.62%
Kappa Coefficient = 0.8135
TABLE II. CONFUSION MATRIX OF TRANSFORMED CLASSIFIED IMAGE
Class Color Trees Grass Urban Area Roads Soil Color
Unclassified [Black] 0.85 0.36 4.36 0.52 0
Trees [Red] 98.51 4.72 0.09 0.08 0
Grass [Green] 0.65 94.71 1.92 0 0.3
Urban Area [Blue] 0 0.2 86.5 1.33 0.52
Roads [Yellow] 0 0 3.19 98.07 0
Soil [Cyan] 0 0 3.94 0 99.19
Total 100 100 100 100 100
Ground Truth (Percent)
Overall Accuracy = (20156/21394) = 94.2133%
Kappa Coefficient = 0.9236
Accuracy assessment of the ML classification was
determined by means of a confusion matrix (sometimes called
error matrix), which compares, on a class-by class basis, the
relationship between reference data (ground truth) and the
corresponding results of a classification [1].
The results are given in Table 1. The diagonal elements in
Table 1 represent the percentage of correctly assigned and are
also known as the producer accuracy. Producer accuracy is a
measure of the accuracy of a particular classification scheme
and shows the percentage of a particular ground class that is
correctly classified.
The Kappa coefficient, κ is a second measure of
classification accuracy which incorporates the off-diagonal
elements as well as the diagonal terms to give a more robust
assessment of accuracy than overall accuracy.
The ML classification yielded an overall accuracy of
85.62% and kappa coefficient 0.8135.
B. Classification After Weirstrass Transform
When we apply Gauss transform on data we have already
seen that our band's distribution changes from skew symmetric
to approximately Gaussian.
When we apply ML classification algorithm on
transformed image we have seen that the overall accuracy
increases from 85.62% to 94.21% because pixel values of
image data follows Gaussian distribution and also the value of
Kappa coefficient, κ increases from 0.8135 to 0.9236. In Table
2 we can see the results of classification after applying the
Weirstrass Transform or Gauss transform. So when our data
follows Gaussian distribution we can say that ML classification
produce better.
VI. CONCLUSIONS
In this study, detail analyses of ML classification for
tropical land covers in Rome have been carried out, in which
lead to a number of conclusions. ML classifies the classes that
exist in the study area with a good agreement when distribution
follows Gaussian and when distributions is not Gaussian we
can see that overall accuracy decreases almost 9%. ML
classified the study area into 5 classes, with accuracy 88% (κ=
0.8425). ML classifies pixels based on known properties of
each cover type, but the generated classes may not be
statistically separable.
294
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-303-320.jpg)
![ACKNOWLEDGMENT
The authors would like to thanks the Dr Waqas A. Qazi for
his valuable input and ideas, Muhammad Qasim for helping
with the implementation of Weirstrass transform.
REFERENCES
[1] T.M. Lillesand, R.W. Kiefer and J.W. Chipman, Remote
Sensing and Image Interpretation, John Wiley & Sons,
Hoboken, NJ, USA, 2004.
[2] DigitalGlobe.com. DigitalGlobe. 12 February 2014. Retrieved
19 June 2014.
[3] Ahmed I. Zayed, Handbook of Function and Generalized
Function Transformations, Chapter 18. p.322-324.
[4] J.B. Campbell, Introduction to remote sensing, Taylor &
Francis, London, 2002.
Muhammad Shoaib is Master student of RS and GIS in
the department of Space Science in Institute of Space
Technology Islamabad, Pakistan.
Zaka Ur Rehman is Master student of RS and GIS in the
department of Space Science in Institute of Space Technology
Islamabad, Pakistan.
Muhammad Qasim received the Masters degree in
Mathematics from Comsats Institute of information
technology Abbottabad, Pakistan.
295
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-304-320.jpg)
![Spatio -Temporal Analysis of Shoreline Changes
along Makran Coast Using Remote Sensing and
Geographical Information System
Hira Fatima1
, Mudassar H. Arsalan, Anam Khalid, Kashif Marjan, Mukesh Kumar
Geo-informatics and Sustainable Development Research Lab
Institute of Space and Planetary Astrophysics
University of Karachi
Karachi, Pakistan
1
hirafatima1407@gmail.com
Abstract—Satellite Images of Makran coast captured by
Landsat 5 TM in 1987, 2000 and 2010 were integrated in a GIS
(Geographical Information System) to perform spatio-temporal
analysis of shoreline changes, coastal accretion and erosion, and
coastline length. It was found that Remote sensing data and GIS
are time-saving and cost effective for such analyses. Makran
shoreline has changed noticeably from 1987 to 2010. During
1987-2010, erosion was dominant along western Makran, it was
eroded at the rate of 0.2 sq. km per year. However accretion was
dominant along Eastern Makran coast, it was advanced at the
rate of 0.5 sq. km per year. During the long time span of 23 years
(1987-2010) Gwadar port was eroded 4 sq. km which was quite
higher than that of Jiwani i.e. 2.4 sq. km. However Ormara,
Pasni and Astola Island were relatively stable places.
Index Terms— Shoreline change detection, Erosion - accretion
analysis, Image interpretation, Band ratio index, Satellite images,
Geospatial technologies, GIS, RS, Geo-informatics.
I. INTRODUCTION
A shoreline being the interface between land and water is a
dynamic feature on the earth's surface. It is always changing
due to sediments continually being eroded from one place and
deposited to another. Many factors play a part behind these
changes, they may be natural like waves, winds, tides and
storms as well as human influenced such as dredging, mineral
exploration, vegetation removal, infrastructure development
etc. These natural processes and human activities can change
shoreline on small time scales i.e. days and years or long time
scales such as decades or centuries. Monitoring shoreline
changes, is very crucial in coastal management and planning.
The rate of these changes i.e. erosion and accretion, can be
used to forecast future shoreline trends. Decision makers need
this information to assess the feasibility of a project and
Government organizations for disaster risk management [1-5].
Remote sensing and Geographical Information System are
widely used for shoreline change detection analysis. The
spectral, spatial and temporal resolutions, repetitive coverage,
synoptic view and cost effectiveness of remote sensing data has
made it more favorable than data collected by conventional
techniques. The shoreline change detection analysis is now
widely being done using diversified data such as historical
topographical maps, aerial photographs, Digital Elevation
Models (DEM), and satellite images. Satellite images from
different satellites (such as Landsat 5, Landsat 7, Landsat 8,
IKONOS, Quick bird, IRS etc.) have been used for such
analysis [5-11]. Further, the data management, visualization
and data analysis capabilities of GIS are applied on the data
derived from satellite images to get desirable results [12-14].
The initial and crucial step in the shoreline change
detection is the selection of appropriate feature that can be
treated as shoreline proxy or indicator that can represent the
current position and can act as a reference to monitor the
change in different periods [2, 15]. These proxy shorelines are
typically of two types: a visible and distinguishable line in
coastal imagery known as High Water Line (HWL) or
intersection of a tidal datum with coastal profile (say MHW,
MLW, etc.) [2]. Among all the vertical levels that are used in
the shoreline change detection, HWL is very common as it is
easily interpretable in photos [2]. After identification, shoreline
can be extracted by manual digitization or some automated
techniques in GIS [13, 16]. Further, the shoreline movement,
erosion and accretion rates can be calculated and its future
trends can be predicted [17, 18].
Literature review shows shoreline change detection is now
considered as an important part of sustainable coastal resource
management and environment conservation, hence it is
monitored throughout the World [19]. Table 1 gives a
summary of several shore detection studies done throughout
the world for different parts of coasts such as beaches, river
296
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-305-320.jpg)
![mouths, deltas, lakes, estuaries and even islands, and geospatial
data and techniques used in these case studies.
Over the 100 years, about 70% of world’s sandy coasts
(which constitute about 20% of the world’s coastline) have
been retreating while 20-30% have been stable and less the
10% have been advancing [20]. Coastal erosion is notable
along the sandy beaches of Pakistan [21]. Literature review
shows that very few shoreline change detection studies for
coastal areas of Pakistan have been done [22, 23] and there is a
big research gap in this dimension. This paper aims to bridge
this research gap.
The Pakistan coast consists of Makran coast (the major
portion) and the Sindh coast. Sindh coast is further divided into
Indus deltaic coastline (a network of 17 minor creeks) and the
Karachi coastline that extends from West of Indus delta to the
outfall of the Hub River from where the Makran coast starts
and extends up to the Iranian border [21, 24-26].
The coastal length of Pakistan found in literature is
ambiguous. It varies from 825 to 1200 km in different sources
(see Table 2). Literature review shows that reason behind this
ambiguity is coastline paradox. This phenomena was first
discovered by Lewis Fry Richardson [27]. Coastline does not
have a well-defined length because of its fractal like
characteristics. The coastline has features at very different
scales, hence the length of coastline increases every time when
it is measured at a smaller scale [28]. In short, the length of
coastline depends on the method used to measure it, the scale
and the resolution of data used [28]. The length of coastline
also varies as the shoreline evolves. –with some places say
beaches getting sediments to deposit while others getting
eroded [29]. However, with the advent of geospatial
technologies and high resolution remote sensing data world’s
coastlines can be measured more precisely and this paradox
can be handled to some extent. Not only the length of coastline,
but the area of sea can also be monitored efficiently and cost
effectively using remote sensing data and GIS [30]. In this
study, the length of Makran coast, the study area, in different
years (i.e. 1987, 2000 and 2010) were calculated from Landsat
TM satellite images. However, more precise coastline length
measurements can be made using high resolution data, such as
SPOT [31].
A. Study Area
The coastal area of Pakistan selected for this study is
Makran coast (see Figure 2 and 6). Most of the Makran coast is
undeveloped, with deserted beaches and some fishing villages.
Further, it is characterized by unique landforms such as mud
flats, rocky cliffs, bays, lagoons, deltas, as well as of the
narrow strip of mountains, which is known as Makran ranges.
Makran ranges cover area of about 400 km long and 250 km
wide. These mountains have elevation up to 1500 m above
mean sea level [26, 32]. The Makran coast is blessed with
scenic beauty of several beaches: Somiani, Hingol River,
Ormara, Pasni, Gwadar etc. The main towns and fishing ports
of the study area are Ormara, Pasni and Jiwani. Gwadar deep
sea port is the economic region of Balochistan coast. It is a
major destination in the Pakistan-China economic corridor [32,
33].
Makran coast differs Sindh coast in several ways. Indus is
the only major river of Pakistan and its delta lie at the Sindh
coast, while at Makran coast only small rivers i.e. Hingol, Hab,
Basul and Dasht do exist. The depths along the Sindh coast
change moderately, while the Makran coast is steep. The
average rainfall on the Makran coast is about 10 mm while it is
twice at the Sindh coast [34]. Makran coast is less vulnerable to
sea level rise than Sindh coastal area as the Makran coast is
uplifting about 1-2 mm/year due to the subduction of Indian
oceanic plate [22]. Coastal erosion is one of the typical effect
of sea level rise [35]. A very threatening subduction zone is
located about 100 km away from Makran coast that can cause
tsunami [36, 37]. In 1945 tsunami, Las Bela State of
Balochistan (which is now Western Pakistan) was affected
badly, particularly the Pasni and Ormara towns [38].
Pakistan has mixed semi-diurnal tides, however the tidal
range varies throughout the coastline [22]. The tidal amplitudes
are greater in the Indus deltaic region as the sea water flows
into creeks with high velocity during flood and ebb tides,
however the tidal amplitudes gradually decreases towards west
(Makran coast) [22, 39]. During the Southwest monsoon
season the direction of the prevailing ocean current is
clockwise and counter clockwise during the Northeast
monsoon season in the Arabian Sea [34]. Generally, the
salinity value is approximately 36 ppt [36].
In this study, we examined different bands of Landsat 5
TM, their combination and ratios for shoreline delineation. We
calculated shoreline length for different years and studied
coastline paradox phenomenon. We analyzed the long and
short term shoreline changes and associated erosion and
accretion along the Makran coast using Landsat TM data
acquired in 1987, 2000 and 2010.
II. MATERIALS AND METHODS
In this study, we used Landsat 5 TM data and two commercial
software (i.e. ArcGIS 10.3 and ERDAS Imagine 9.2) for data
management, analysis and visualization. The objectives of
this study were achieved by simple steps shown in the figure
3.Fig. 1. Coastline Paradox Illustration
297
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-306-320.jpg)
![Table 1 Shoreline Change Detection Studies throughout the World
Serial
#
Author Year Study Area Technique used Data Used
1 Shetty
[40]
2015 Mangalore Coast,
Netravathi-Gurupur and
Mulky-pavanje Spits
Visual interpretation and on screen
digitization
Landsat 5, IRS and Landsat 8
2 Natesan
[41]
2015 Tamil Nadu, India Layer Stacking, digitization Landsat MSS, Landsat TM,
Landsat ETM+ and Landsat
OLI
3 Hegde
[42]
2015 Karnataka Coast, India Histogram slicing of infrared band,
unsupervised classification and
digitization
Landsat TM, Landsat ETM+,
Landsat OLI-TIRS
4 Aedla
[16]
2015 Netravati-Gurpur River
mouth
Modified Self–Adaptive Plateau
Histogram Equalization with Mean
Threshold
IRS data
5 Mahapat
[10]
2013 South Gujarat Coast False color composite (bands 2, 3, 4),
visual interpretation and digitization
Landsat MSS, Landsat TM,
Landsat ETM+, IRS
6 Choudharey
[43]
2013 from Karwar to Gokarna -
South West coast of India
Geo referencing, digitization Toposheet and IRS data
7 Murray
[8]
2013 Wotje Atoll, Marshall
Islands
Mosaicking, Digitization Aerial photographs,
IKONOS, QuickBird,
GeoEye-1
8 Rio
[4]
2013 SW Spain Georeferencing, polynomial
correction and digitization
Aerial Photographs
9 Das
[44]
2013 Shankarpur - Mandarmoni
Coast line, West Bengal
Band ratio mode between mid
infrared and green, digitization
Topomap, Landsat TM,
Landsat ETM+ and Google
Earth imagery
10 Azaz
[45]
2012 Wedam_Alsahel area,
Batinah, Oman
Geometric correction, Image
degradation, digitization
IRS, IKONOS
11 Faiboon
[46]
2012 Chalatat Beach in
Songkhla Province,
Thailand
Geometric correction, resampling,
Overlay
Landsat ETM+, Landsat TM,
12 Eludoyin
[1]
2012
Bonny Island, Nigeria
Layer stacking, visual interpretation
and digitization
Landsat TM, Nigersat data
13 Kumaravel
[47]
2012 Cuddalore district, East
coast of Tamil Nadu, India
Georeferencing, layer stacking,
digitization
Toposheets, IRS data
14 Kuleli
[6]
2011 Coastal Ramsar Wetlands
of Turkey
Top of atmosphere segmentation
algorithm, NDWI, Automatic
thresholding and shoreline extraction
Landsat MSS, Landsat TM,
Landsat ETM+
15 Adegoke
[3]
2010 Niger Delta, Nigeria Radiometric and geometric
correction, georeferencing,
mosaicing, Histogram equalization,
Band combination 6,4,2 and
digitization
Landsat TM, Landsat ETM+
data
16 Li
[48]
2010 Pearl River Estuary, China Georeferencing, Mosaicking, Layer
stacking, slicing, classification and
digitization
Topographic map, nautical
map, Landsat TM, Landsat
MSS, Landsat TM+, Spot XS
17 Maiti
[5]
2009 Bay of Bengal, Eastern
India
Georeferencing, gray level
thresholding and segmentation by the
edge enhancement technique of NIR
bands, Digitization
Toposheets, Landsat MSS,
Landsat TM, Landsat ETM+
and ASTER
18 Kumar
[49]
2009 Mangalore Coast, India Georeferencing, layer stacking,
digitization
Toposheets, IRS data
19 Rajamanickkam 2006 Kallar and Vaipar Coast,
Tamilnadu, India,
Edge enhancement, level slicing,
NDVI, digitization
Topo sheets and IRS data
20 Klien
[50]
2006 Position adjacent to Hadera
power station, Israel
Rectification and georeferencing,
Digitization
Aerial Photographs
21 Makota
[13]
2004 Kunduchi Area, Tanzania Visual interpretation and on screen
digitization
Aerial Photographs
22 Frazier
[51] 2000
Channel and flood plain of
Murrumbidgee River
Single band density slicing, multi-
spectral maximum likelihood
algorithm, digitization
Landsat 5 TM data, Aerial
Photographs
298
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-307-320.jpg)
![Table 2 Length of Pakistan Coastline from different sources
Serial
#
Author/
Organization
Year Length
of
Pakistan
Coastline
km
Length
of
Makran
Coastline
km
Length of
Sindh
Coastline
km
1 UNEP [34] 1986 825 ___ ___
2 Quraishee [39]
Tariq [52]
Rashid [23]
1986
2000
2012
990 ___ ___
3 Pakistan
Tourism
Development
Corporation
[53]
2006 1046 ___ ___
4 Jalal [54] 2008 1050 ___ ___
5 Wildlife of
Pakistan [32]
2006 1050 800 250
6 Pike [36]
Hafeez [55]
2014
2007
1050 700 350
7 Einfopedia
[56]
2010 1100 771 329
8 Encyclopedia
Britannica
[25]
2001 ___ ___ 350
9 ESCAP [21] 1996 ___ ___ 370
10 WWF-
Pakistan [24]
2008 ___ 1000 ___
11 Himalayan
Holidays [26]
2002 ___ 754 ___
12 Jalal [33] 2008 1200 ___ ___
A. Data Collection
For this study, we utilized Landsat 5 TM images of the
study area (Path/Row: 156-42, 156, 43, 155-42, 155-43, 154-
42, 154-43, 153- 42 and 153- 43) acquired in 1987, 2000 and
2010. Landsat 5 was a joint effort of NASA and USGS,
launched on 1st March 1984. Landsat 5 was put in a circular,
sun synchronous, near polar orbit at the altitude of 705.3 km.
Landsat 5 carried TM (Thematic Mapper) and MSS (Multi
Spectral Scanner System) sensors. It took approximately 16
days to scan the whole earth's surface. It used to cross equator
at 9:45 a.m. (+/- 15 min). [57]
Landsat 5 was decommissioned on 5th June 2013, due to
several mechanical failures. Landsat 5, in its 29 year, 3 months,
4 day journey orbited 150,000 times, transmitted over 2.5
million images. Several significant events were captured by
Landsat 5, such as 1986’s Chernobyl incident, 2004’s
devastating tsunami in Southeast Asia, etc. [58] Landsat 5 set a
Guinness world record for “Longest operating Earth
observation satellite”. Landsat 5 Thematic mapper sensor data
is available on USGS websites for free. [59]
To assess short term (in a span of 10 and 13 years) and long
term (in a span of 23 years) shoreline change detection, cloud
free and low tide data of Landsat 5 TM were downloaded from
the USGS EarthExplorer website. Further characteristics of
satellite data (Landsat 5 TM) used in this study is summarized
in table 3. Boundaries of water bodies can be delineated more
accurately using fine resolution spatial data for example
IKONOS and SPOT, with spatial resolution of less than 1 m.
Although IKONOS and SPOT are commercial satellites and
their data is not available free of cost. However, all the Landsat
data in USGS archive is available online for free. The summary
of collected data is given in table 4.
B. Data Processing
After data collection, data was processed to make it ready for
our analysis. As the study area was very large (i.e. about 990
km long) the study area was divided into two parts i.e.
Western Makran Coast (Path/Row: 155-42, 155- 43, 156-42
and 156-43) and Eastern Makran Coast (Path/Row: 154-42,
154-43, 153-42 and 153-43) (See figure 6). All the images of
Western and Eastern study area were geometrically corrected,
stacked and mosaicked using ERDAS Imagine 9.2. All the
collected and processed data, and result files of shoreline
change detection and erosion - accretion analysis were stored
in ArcGIS geodatabase. All images and shape files used in this
study were projected to Universal Transverse Mercator
(UTM) projection system, zone 41 and the spheroid and datum
were referenced to WGS1984.
C. Shoreline Identification
Water behaves like a semi-transparent medium for the
electromagnetic radiation, hence when EM radiation strikes
water surface, it gets reflected, transmitted or absorbed in
water. The spectral responses depend on the wavelength of the
radiation as well as the physical and chemical properties of the
water. When water is in the liquid form, it shows greater
reflectance in the visible region between 0.4μm and 0.6μm
however wavelengths beyond 0.7μm are completely absorbed.
Water absorbs most of the energy in NIR and MIR
wavelengths and appears dark in these bands, in contrast to this
land and vegetation show higher reflectance in these
Fig. 2 Some Glimpses of Study Area- Makran Coast, Pakistan
299
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-308-320.jpg)
![wavelengths and appear bright in these bands. This property of
Water is used to differentiate it from land and vegetation in
multi-spectral imagery [7, 9, 62-64]. Figure 5 shows the
spectral profiles of land, water and vegetation drawn in
ERDAS Imagine 9.2 from a Landsat 5 TM image.
Figure 7 shows images of Miani Hor -a part of Makran
Coast- in different bands of the Landsat 5 TM. In first three
bands the contrast between water and other surface features are
not very remarkable. On the contrary, the IR bands (bands 4
and 5) show a sharp contrast between them as water reflects
very less in the IR region of the EMR spectrum. Among all the
bands, band 5 gives the most outstanding contrast between
water and other surface features. However, different band
combinations have their own tendency to differentiate land and
water. Figure 7 shows that Mid IR and Green band ratio is very
suitable for shoreline delineation.
Band ratio is a digital image processing technique in which
DN value of one band is divided by that of any other band. It is
used to enhance the contrast between different features. It also
eliminates the shadowing effects [63]. Band ratio has its
application in several fields such as minerals, soil, vegetation,
water bodies etc. [64]. Three different band ratio images can
be combined to make color composites. This will highlight
certain features in distinctive colors [65].
The initial and crucial step in the shoreline change
detection is the selection of shoreline proxy and then its
identification. Water boundary delineation is one of the
applications of Remote Sensing. In this study, we considered
High Water Line (WHL) as a shoreline proxy. The WHL is
defined as the intersection of land with water surface at an
elevation of high water [66]. High water line has been used in
several studies for shoreline delineation [4, 10, 13, 41, 43, 44].
To identify the high water line (WHL), the mid infrared (band
5) and green (band 2) band ratio index methodology was
adopted.
D. Shoreline Extraction
Band ratio was calculated for every image, then the high
water line was extracted by manual digitization on 1:50,000
scales in ArcGIS 10.3. This scale was kept consistent to in all
digitization work throughout the study to maintain the
consistency. The temporal movement (1987-2010) of shoreline
is shown in figure 8 and 9.
E. Shoreline Length Measurement
The length of shorelines (extracted from band ratio images
of 1987, 2000 and 2010) were measured using calculate
geometry function in ArcGIS 10.3.
F. Shoreline Change/Movement
Shoreline change was analyzed by overlay technique.
Extracted shorelines of 1987, 2000 and 2010 of eastern and
western Makran coast were overlaid on each other to visualize
the shoreline movement.
G. Erosion and Accretion Measurement
Erosion is the transportation process of soil or decomposed
rock material by natural forces (such as storms, flooding,
waves and tides) while accretion is a slight and unnoticeable
accumulation of soil or decomposed rock by natural
phenomenon. Coastal accretion occurs when the alluvion
deposit upon the shore. [68]
Fig. 4. Data Processing
Fig. 3 Materials and Methods
Fig. 5. Spectral Profiles of Land, Water and Vegetation drawn in Erdas
Imagine 9.2 from Landsat 5 TM image
300
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-309-320.jpg)
![4) Accretion - erosion Analysis Results on Local Scale
a) Gwadar
Gwadar (25°07′35″N 62°19′21″E) is a city located on the
southwestern coastline of Pakistan. Gwadar was nominated as
the winter capital of Balochistan in 2011. Gwadar port located
at the mouth of the Persian Gulf. It is a third deep sea port of
Pakistan, inaugurated to increase the port capacity of Pakistan
[69]. The erosion - accretion analysis shows that Gwadar coast
has faced a noticeable erosion problem. During 1987-2000
about 4.5 sq. Km land was eroded at Gwadar port and no
noticeable accretion took place while 1.5 sq. Km land was
eroded and 0.5 sq. Km land was advanced at Western Gwadar
bay. During 2000-2010 about 1.6 sq. Km land was advanced at
Gwadar port while no noticeable erosion took place
meanwhile, no noticeable erosion and accretion took place at
Eastern or Western Gwadar bay. In long time span of 23 years
(1987-2010) about 4 sq. Km land was eroded at Gwadar port
while 1.5 sq. Km land was advanced meanwhile 1.2 sq. Km
land at western and 0.4 sq. Km land at eastern Gwadar bay was
eroded while no noticeable accretion took place (See figure
11).
b) Pasni
Pasni (25°16′N 63°28′E) is a moderate size town and
fishing port located at the Makran coast [70]. During the time
span of 13 years (1987-2000) about 0.2 sq. Km land was
eroded while 4.5 sq. Km land was advanced at Pasni and its
vicinity. During the short time span of 10 years (2000-2010)
about 0.3 sq. Km land was advanced while 5.8 sq. Km land
was eroded at Pasni and its vicinity which is greater than the
accretion took place during 1987-2000. During the long time
Fig. 7. Landsat 5 TM Images of Miani Hor-a part of Makran Coast- in different Spectral bands, band combinations and band ratio using ERDAS Imagine 9.2
302
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-311-320.jpg)
![Table 3 Landsat 5 TM and MSS Sensor Characteristics
Table 4 Summary of Collected Data
Satellite
(Sensor)
Path/Row Year Tide
Phase
Landsat 5
(TM Sensor)
153-42, 153-43
154-42, 154-43
155-42, 155-43
156-42, 156-43
1987
Low2000
2010
Table 5 Makran Coastline Length measured from Landsat 5 TM images
Year Scale Spatial
Resolu-
tion
m
Western
Makran
coast
length
Km
Eastern
Makran
coast
length
Km
Makran
coast
total
length
km
Shoreline
proxy
2010
1:50,000 30
335 634 969
WHL at
low tide
2000 337 661 998
1987 344 667 1011
span of 23 years (1987-2010) about 0.5 sq. Km land was
eroded as well as 0.5 sq. Km land was advanced which showed
a balance in erosion and accretion but at a long time span (See
figure 12).
a) Jiwani
Jiwani (25°03′N 61°44′E) is a town and commercial port. It
is located near the Pak-Iranian border [71]. The accretion -
erosion analysis shows that the coastal area of Jiwani and its
vicinity is under influence of natural forces as well human
induced factors which have been causing coastal erosion.
During 1987-2000 about 5.2 sq. Km land was advanced while
no considerable erosion took place. During 2000-2010 about
4.8 sq. Km land was eroded (near about accretion took place
during 1987-2000) while only 0.8 land was advanced at Jiwani
and vicinity. In 23 year long time span (1987-2010) about 5.2
sq. Km land was advanced while 2.4 sq. Km land was eroded
(See figure 13).
B. Accretion - erosion along Eastern Makran Coast
The western Makran coast accretion - erosion analysis
results are given in table 7. The accretion - erosion analysis
visualization is shown in figure 14.
1) Accretion and Erosion between 1987-2000
The accretion - erosion analysis shows during 1987 to 2000
(i.e. 13 year time span) about 8.5 sq. Km land was eroded
while 12 sq. Km land area was advanced at the east Makran
coast. Hence the total change was 3.5 sq. Km accretion. During
1987-2000 the accretion rate per year was about 0.9 sq. Km,
while the erosion rate was 0.7 sq. Km which was not exactly
same, but close to the accretion rate. Hence, in this short time
span there was a balance in the accretion and erosion to some
extent. In this (1987-2000) thirteen year period the eastern
Makran coast was advanced at a rate of 0.3 sq. Km per year.
2) Accretion and Erosion between 2000-2010
The accretion - erosion analysis shows during the short
time span of 10 years (i.e. 2000- 2010) about 5.5 sq. Km land
was eroded while land was advanced by 16 sq. km at eastern
Makran coast. Hence the total change was 10.5 sq. Km
accretion. During 2000-2010 the erosion rate per year was
about 0.5 sq. Km, while the accretion rate was 1.6 sq. Km,
which was quite higher than erosion rate. Hence the accretion
was far dominant than erosion in this time span and land
advanced at the rate of 1.1 sq. km per year.
3) Accretion and Erosion between 1987-2010
The accretion - erosion analysis shows during the long time
span of 23 years (i.e. 1987- 2010) about 6.8 sq. Km land was
eroded while land was advanced by 17.5 sq. km at eastern
Makran coast. Hence the total change was 10.7 sq. km
erosion. During 1987-2010 the accretion rate per year was
about 0.8 sq. Km, while the erosion rate was 0.3 sq. Km which
was lower than accretion rate. Hence the accretion was quite
dominant than erosion in this 23 year long time span (1987-
2010) and the coastal area of Eastern Makran was advanced at
the rate of 0.5 sq. Km per year.
Satellite Sensors Swath
(Km)
Scene size
(Km)
Resolution Band Spectral
Coverage
(μm)
Temporal Spectral Spatial
(Meters)
Landsat
5
MSS 180 180 x 170 16 days 4 82 Band 4 – Green 0.5 – 0.6
82 Band 5 – Red 0.6 – 0.7
82 Band 6 – Near IR 0.7 – 0.8
82 Band 7 – Near IR 0.8 – 1.1
TM 185 170 x 183 16 days 7 30 Band 1 – Blue 0.45 – 0.52
30 Band 2 – Green 0.52 – 0.6
30 Band 3 – Red 0.63 – 0.69
30 Band 4 – Near IR 0.76 – 0.90
30 Band 5 – Mid IR 1.55 – 1.75
120 Band 6 –Thermal 10.40 – 12.50
30 Band 7 – Mid IR 2.08 – 2.35
303
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-312-320.jpg)
![4) Accretion - erosion Analysis Result on Local Scale
a) Hingol River
Hingol River (25° 23′N 65° 28′E) is located in the Gwadar
district of Balochistan. It is the longest river in Balochistan
(about 560 km long). Unlike the majority of streams in
Balochistan, which flow only in rare rain, Hingol River flows
all year long [72]. The erosion - accretion analysis of mouth of
the Hingol River shows that during 1987-2000 about 1.3 sq.
Km land was eroded there while 1.2 Sq. km land was advanced
in its vicinity. During 2000-2010 about 0.4 sq. Km land was
advanced while only 0.2 Sq. km land was eroded at the mouth
of Hingol River and about 0.6 Sq. km in the vicinity. In long
time span of 23 years (1987-2010) about 0.2 sq. Km land was
eroded while 0.3 sq. Km land was advanced at the mouth of the
Hingol River while in its vicinity about 0.6 Sq. km land
advancement took place and 1.2 sq. km land was eroded. (See
figure 15).
b) Ormara
Ormara (25°12′53″N 64°38′2″E) is an old coastal town
located in the Gwadar district of Balochistan. Ormara has a fish
harbor and a port [73]. The erosion - accretion analysis of
Ormara shows that during 1987-2000 about 0.6 sq. Km land
Fig. 8. Western Makran Coast Temporal Shoreline Movement 1987-2010
Fig. 9. Eastern Makran Coast Temporal Shoreine Movement 1987-2010
304
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-313-320.jpg)
![was eroded while 4 Sq. km land was advanced at Ormara.
During 2000-2010 about 1 sq. Km land was advanced at
Ormara port while only 1.5 Sq. km land was eroded. In long
time span of 23 years (1987-2010) about 0.5 sq. Km land was
eroded at Ormara port while 3.7 sq. Km land was advanced.
(See figure 16).
c) Islands in Study Area
A piece of sub-continental land that is surrounded by water
is known as the Island. Islands are classified into two main
types i.e. continental islands and oceanic islands. They may be
artificial as well [74]. Pakistan has several natural islands off
the Sindh and Balochistan (Makran) coasts. The islands off the
Makran coast are Malan island (a volcanic mud island that
Fig. 10. Western Makran Coast Accretion Erosion 1987-2010
305
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-314-320.jpg)
![appears and disappears), Zalzala island (appeared offshore
Gwadar followed by an earthquake of 7.7 magnitude in
September 2013) and the Astola island [75].
ASTOLA ISLAND
Astola Island (25°7′21″N 63°50′51″E) is Pakistan’s largest
offshore island. It is also known as Haft Talar Island (Island of
seven Hills) with the highest point 246 feet above the sea level
[76]. It is about 6.7 km long and with a maximum width of 2.3
km and 6.7 Sq. km approximated area [76]. However the area
of Astola Island estimated in this study was 2.2 sq. km with
shoreline length 10.2 km.
This study reveals that during short (i.e. 1987-2000 and 2000-
2010) and long time (1987-2010) spans, very nominal
accretion and erosion took place along the shoreline of Astola
Island. During the short time span of 13 years (i.e. 1987-2000)
only 0.13 sq. km land was advanced while 0.02 sq. km land
was eroded. During ten year time span (2000-2010) only 0.04
sq. km land was advanced while 0.01 sq. km land was eroded.
During the long time span of 23 years (i.e. 1987-2010) 0.1 sq.
km land advanced while only 0.04 sq. km land was eroded.
d) Lagoons in Study Area
The lagoon is a shallow water body separated from the
larger water body by reefs and small islands [77]. Lagoons are
classified into coastal and oceanic lagoons. Coastal lagoons are
further classified into three main types i.e. choked, leaky and
restricted [78]. Lagoons are found in a great range of sizes, the
size of a smallest lagoon is about a hectare while the largest
(i.e. Lagoa Patos in Brazil) is about 10,000 km2
. Usually the
length of the lagoon is greater than its width [79].
Lagoons are typical coastal features found parallel to the
coastlines all around the world say Glenrock lagoon in
Australia, Lagoa dos Patos lagoon in Brazil, Apoyo lagoon
natural reserve in Nicaragua, Blue lagoon in Turkey, Venetian
lagoon in Veneto, Miani Hor and Khor Kalmat along Makran
coast, Pakistan [77, 80, 81].
Fig. 11. Gwadar Accretion Erosion 1987-2010
306
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-315-320.jpg)
![KHOR KALMAT
Khor kalmat lagoon (25° 24′N 64° 06′E) is located along
the eastern Makran coast. Basol River flows southward in
the Gwadar District. It drains a deserted Makran region, with
its river mouth at Khor Kalmat lagoon [81].
MIANI HOR
Miani Hor (25° 32′N 66° 12′E) is a swampy lagoon and a
Ramsar wetland, located in Lasbela district [80].
Both Miani Hor and Khor Kalmat were not included in the
accretion - erosion analysis of this study as lagoons are
extremely dynamic ecosystem [78]. Further, they have their
own dynamics different from that of other coastal features [79].
A study in which 7 year beach profile data was analyzed,
showed that the lagoon shorelines were far dynamic than the
ocean shorelines [82]. The flow of water in lagoons is slow and
sluggish and the flushing time depends on the type of lagoon
[78].
C. Accretion - erosion along Makran Coast
The accretion - erosion analysis results of western and
eastern Makran coast were combined to understand the
accretion - erosion scenario of entire Makran coast.
D. Accretion and Erosion between 1987-2000
The accretion - erosion analysis shows during 1987 to 2000
(i.e. 13 year time span) about 17.9 sq. Km land was eroded
while land was advanced by 24.6 sq. Km at Makran coast.
Hence the total change was 6. 7 km accretion. During 1987-
2000 the accretion rate per year was about 1.9 sq. Km, while
the erosion rate was 1.6 sq. Km which was not exactly same,
but close to the accretion rate. Hence, in this short time span
there were balance in the accretion and erosion to some extent.
However, in this (1987-2000) thirteen year period the Makran
coastal area advanced at a rate of 0.5 sq. Km per year.
Fig. 12. Pasni and Vicinity Accretion Erosion 1987-2010
307
Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings](https://image.slidesharecdn.com/conferenceproceedingsicase2015final1-160918062150/85/Conference-proceedings-316-320.jpg)
Uploaded byDr Muhammad Mustansar
Conference proceedings
This document summarizes a study that compares the performance of two state observers - a sliding mode observer with super-twisting algorithm (STSMO) and a high gain observer (HGO) - for estimating unmeasured states of a quadrotor UAV. The paper designs each observer and then applies a second order sliding mode control technique using the estimated states to control the quadrotor. Simulations show the performance of each observer under the same control scheme and perturbations. The study aims to compare the observers' characteristics for state estimation of the quadrotor system to determine the best observer for real-time applications given system uncertainties and noise.
More Related Content
More from Dr Muhammad Mustansar
Conference proceedings
- 2. ICASE - 2015 FourthInternational Conference on Aerospace Science & Engineering September 2-4, 2015 Institute of Space Technology Islamabad Pakistan Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 3. Conference Proceedings Editors Dr. NajamAbbas Naqvi Mr. Raza Butt ISBN 978-1-4673-9123-8 Printed in Pakistan 2016 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 4. i ICASE ORGANIZING COMMITTEE Engr.Imran Rahman (Chairman) Dr. Najam Abbas Naqvi (Secretary) Mr. Zia Sarwar (Treasurer) Dr. Zafar Mohammad Khan Dr. Muddassar Farooq Dr. Badar Munir Ghauri Dr. Qamar-ul-Islam Dr. Abid Ali Khan Dr. Ibrahim Qazi Dr. Farrukh Chishtie Dr. Asif Israr Dr. Salman Ahmed Dr. Mirza Muhammad Naseer Engr. Ishaat Saboor Engr. Khurram Humaiyun Mr. Muhammad Hafeez INTERNATIONAL SCIENTIFIC COMMITTEE Dr. Leonardo Reynari , Italy Dr Dongkai Yang, China Dr. DDGL Dahanayaka , Sri Lanka Dr. Ali Imran , United Kingdom Dr. Muhammad Yusof Ismail , Malaysia Dr. Rakhshan Roohi , Australia Dr. Fawad Inam , United Kingdom Dr. Tahir I. Khan , Canada Dr. Iftikhar Ahmad , Saudi Arabia Dr. Aquib Moin, South Africa NATIONAL SCIENTIFIC COMMITTEE Dr. Badar Munir Ghauri Dr. Muddassar Farooq Dr. Qamar-ul-Islam Dr. Abid Ali Khan Dr. Asif Israr Dr. Ibrahim Qazi Dr. Farrukh Chishtie Dr. Syed Wilayat Hussain Dr. Ahtezaz Qamar Dr. Muhammad Zubair Khan Dr. Umer Iqbal Bhatti Dr. Najam Abbas Naqvi Dr. Aamir Habib Dr. Khurram Khurshid Dr. Moazam Maqsood Dr. Fazeel Mehmood Khan Dr. Waqas Qazi Dr. Rizwan Mughal Dr. Arjumand Zaidi Dr. Abdul Haseeb Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 5. ii EDITORIAL COMMITTEE Dr. NajamAbbas Naqvi Mr. Raza Butt Mr. Waqas Ramzan ICASE SECRETARIAT Dr. Najam Abbas (Secretary ICASE 2015) Mr. Zeeshan Fareed (Marketing and Publicity) Mr. Raza Butt (Media and Public Relations) Mr. Waqas Jilani Joiya (Logistics and Operations) Mr. Muhammad Adeel (Graphic Designing) Mr. Waqas Ramzan (Data Management) Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 6. iii SPONSORS Higher Education Commission(HEC) National ICT R&D Fund COMSTECH Pakistan Atomic Energy Commission (PAEC) Kahuta Research Laboratory (KRL) Pakistan Science Foundation (PSF) Vital Group Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 7. iv PREFACE Institute of SpaceTechnology (IST), Islamabad, Pakistan organized the “Fourth International Conference on Aerospace Science & Engineering” (ICASE) from September 2-4, 2015. This conference is a regular biennial event to provide an international forum in which researchers, engineers, professional and students from all over the world get a chance to interact and discuss the latest themes and trends related with aerospace science and engineering. It provides a platform to share experiences, foster collaborations across industry and academia, and to evaluate emerging aerospace technologies and developments across the globe. The success of the first three conferences in 2009, 2011 and 2013 has earned ICASE a high standing in the domains of high performance aerospace materials, space communication techniques, control and guidance systems, design and construction of space systems and structures. These conferences provided an ideal opportunity for exchange of information amongst scientists, engineers and researchers from all across the globe. ICASE 2015 featured a diverse blend of thematic areas including Aerospace and Avionics, Satellite Design Development and Security, Mechanical Engineering for Aerospace Applications, Aerospace Materials Design and Engineering, Satellite Communication and Image Processing, Global Navigation Satellite Systems, Remote Sensing & Geographic Information Science, Astronomy and Astrophysics, Information Technology and Cyber Security, Space Technology Awareness and Society. A total of 110 papers were presented in the conference while 30 poster presentations were held. There were 20 technical sessions during the conference covering the different themes and track related with aerospace science and engineering. In addition to that, there were 15 panel discussions, tutorials and workshops sessions in connection with conference themes. A galaxy of 30 national and International invited speakers shared their research accomplishments with the academicians, researchers and students from all over Pakistan. The representatives from industry and elite Research and Development organizations also exhibited their industrial and technical paraphernalia during the conference. Extensive deliberations and collaborations were the other significant focuses of ICASE 2015. Key representations included National ICT R&D Fund, AIDL and the National Space Agency of Pakistan, SUPARCO. Main sponsors of ICASE 2015 included Higher Education Commission (HEC), PAEC, KRL, COMSTECH, Pakistan Science Foundation, National ICT R&D Fund and the VITAL group. Prodigious efforts were put in to publish this ICASE 2015 proceedings book. Organizing committee, reviewers, chairs, co-chairs, data processors, proofreaders and the designers contributed their ration remarkably in pooling the valuable research findings in a single document. I am grateful to ICASE 2015 team for their extended efforts in making this conference a great success. My special thanks to our sponsors for their generous financial support in driving the research zeal amongst the researchers, scientists and the engineers’ community. Dr. Najam Abbas Naqvi Secretary ICASE 2015 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 8. v CONTENTS 1 Junaid AnwarA Comparison Study of Advanced state Observers for Quad rotor UAV with Sliding Mode Control 1 2 Bushra Aijaz Fuzzy Temperature Controller for Induction Heating 9 3 Qazi Ejaz ur Rehman Stability and Control Solution of Quad-Copters 13 4 Amna Butt Detection of Fire Hotspots dealt by Emergency First Responders in Rawalpindi using GIS application 20 5 Zeeshan Khan Prospects of Airborne Wind Energy Systems in Pakistan 25 6 Muhammad Amin Integrated use of Potential Rainwater Harvesting Site for Agriculture Using Geo-Spatial Approach 31 7 Muhammad Usman Saleem Urban change detection of Lahore (Pakistan) using the Thematic Mapper Images of Landsat since 1992-2010 38 8 Asad Abbas A Review of Fundamentals of Hyperspectral Imaging and its Applications 44 9 Khazar Hayat A numerical study on the impact resistance of braided composites 50 10 Waheed Gul Improving physical and mechanical properties of medium density fiber board (MDF) 57 11 Abdur Rehman Finite Element Analysis of Tool Wear in Ultrasonically Assisted Turning 64 12 Rabia Zafar Metamaterials in Aerospace Industry: Recent Advances and Prospects 69 13 Engr Numan Khan Finite Element Simulation of Composite Body Armor 73 14 Atiq ur rehman Incorporate GNSS with Android & Improve the Search and Rescue 15 Malik Abid Hussain Geostatistical Analysis on Seismic Data over North-Western Regions of Pakistan, Afghanistan and Eastern Regions of Iran & Tajikistan 16 Rabia Tabassum GIS for estimating Optimized Water Demand Using Sustainable Water Resource ManagmentFor Planned City 17 Ferheen Ayaz Optimized Threshold Calculation based on Received Signal Characteristics for Blanking Non-linearity at OFDM Receivers 18 Farzan Javed Sheikh A Review on Mobile Wireless Communication Networks (0G to upcoming generations) 19 Muhammad Altaf Khan Intelligent Detection of Distributed Flooding Attack in Wireless Mesh Network 21 Ferheen Ayaz Introducing space plant biology to students through hands-on activities using clinostat 22 Faizan Muhammad The Political and Economic Feasibility of Current Space Resource Management Policies 23 M Sohail Shahid Feasibility Study to Install Fire Fighting Equipment on a Cargo Helicopter 24 M. Saad Sohail Design and Optimization of S-band Wilkinson Power Divider for Transceiver Applications 25 Mateen Tariq Optimize Manufacturing of unidirectional carbon prepregs for space Applications 26 Taimoor Zahid Electrical Power Conditing Unit Design for Space Qualified C Band Receiver Geo Satellite Applications 27 Najam ud Din Ahmad DSP based Electro Hydraulic actuator control with irreteraceable feedback error. 79 operation. 84 98 109 114 120 20 Aamir Nawaz Touch panel Based Restaurent Automation Using Zigbee Technology 126 131 136 147 154 157 160 165 Concept 28 Syed Jahanzeb Hussain Pirzada Design for Test Approach using FPGA for BPSK Modem 171 29 Taimoor Zahid Design of a Fuzzy Logic Water Level Controller 175 30 Gohar Ali The use of Nuclear reactor in Space Applications:Propulsion and Power 181 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 9. vi 31 Muhammad AamirImpact of Thermal Aging on Microstructure and Mechanical Properties of high Sn Content, Sn-Pb Solders 32 Madni Shifa Ullah Khan Effect of aryl diazonium salt functionlization on the electrical properties of MWCNTs and MWCNTs/CF reinforced polymer composite 33 Sania Nazir Design of C band Slotted Waveguide Array antenna with high Impedance Bandwidth and improved Reflection Coefficient 34 Muhammad Nauman Hussain Risk Areas Mapping and Identification of Hotspots on the Road- Network of Lahore 36 Ali Jan Hassan Assessment of Urban Growth of Karachi: From A Tiny Town to A Meta City of the World 38 Shakeel Ahmad Waqas Identification of Post Disaster Scenario Using Double Threshold Energy Detection 39 M Ameer Umar Malik Design and Analysis of Magnetic MEMS Accelerometer for Inertial Navigation 47 Syed Wasif Ali Shah Design and Development of Low Cost Motor Drive for Hub Wheel based Electric Vehicles. 48 Shahid Karim Preparation, Structure and Dielectric/Piezoelectric Properties of BiScO3- PbTiO3-Pb(Mn1/3Nb2/3)O3 high temperature piezoelectric ceramics 50 Muhammad Shoaib Comparison of Maximum Likelihood Classification Before and After Applying Weierstrass Transform 51 Hira Fatima Spatio -Temporal Analysis of Shoreline Changes along Makran Coast Using Remote Sensing and Geographical Information System 52 Sadaf Javed Influence Analysis of Minerals on Drinking Water Quality Around River Jhelum 55 Zehra Ali Optimized Threshold Calculation based on Received Signal Characteristics for Blanking Non-linearity at OFDM Receivers 56 Naveed Riaz Measurement & Testing Techniques of Performance Parameters for Electric Servo Actuators 187 192 200 203 35 Muhammad Arslan Analysis of Recent Drought Based on NDVI and Meteorological Data 208 211 37 Abdur Raqeeb Gaziani JUPITER, The Gas GIANT 218 219 222 Fuzzy Logic 40 M Shahan Qamar Feasibility Assessment of Running JP-8 Fuel in Diesel Engine 229 41 Muhammad Usman Saleem Artificial intelligence reboot 236 42 Izhar An Intelligent Approach for Edge Detection in Noisy Images Using 243 Mode Neural Network augmentation. 43 Sundus Najib A Survey of Active ITU-R P-Series Propagation Models 249 44 Anwar Ul Haque An Experimental Study To Evaluate the Effect of Strut and Fairing 255 45 M Tanveer Iqbal Fairing Separation Dynamic Analysis Using Analytical Approach 261 46 Saqib Alam Launch Vehicle Control based on Dynamic Inversion with Sliding 265 271 275 49 Abdur Rasheed A Comprehensive Study On QoS For Mobile Ad Hoc Network 282 289 296 316 53 Iqra Basit Selection of the optimal interpolation method for groundwater quality 325 54 M Tasawer Hussain Thermal Design and Analysis of PNSS-1 Satellite 334 340 345 57 IEEE Publications LIST of 27 IEEE submitted Papers 348 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 10. Nonlinear Observers forClosed loop Sliding Mode Control of Quadrotor UAV Junaid Anwar, Fahad Mumtaz Malik, Muhammad Bilal Khan College of Electrical and Mechanical Engineering, National University of Sciences and Technology, Pakistan Abstract—This paper deals with the performance comparison of Sliding mode observer with super-twisting algorithm (STSMO) and High gain Observer (HGO) for a remotely controlled quad rotor UAV. Under the restriction that inertial co-ordinates and attitude angles are available for measurement while angular and linear velocities are estimated. This paper is solved in two steps for each observer. First the observer (HGO and STSMO) is designed and then in second stage a second order (2-SMC) technique is being applied on the basis of estimated states to design controller(for which systems is portioned into fully and under-actuated sub-systems).Simulations results shows the performance comparison of both observers under the same control scheme. I. INTRODUCTION More recently, a growing interest in the UAV has been shown by industry and academia [1]–[7].The vital and poten- tial use of flying robots for civil as well as military applications are attracting the industries and the academia community. The feature of flying in narrow space and vertical takeoff and land- ing (VTOL) made quadrotor unique relative to other mobile robots and conventional aircrafts. The quadrotor is an under actuated system with six outputs and four inputs, they are owed to carry out the tasks ranging from surveillance to rescue mission but the challenges behind the control of quadrotor aerial vehicle like un-stability and highly nonlinear behavior are the major source of attraction and many control approaches to deal with quadrotor dynamics have been presented so far [8]–[14][14-20]. This paper deals with the development of 2-sliding mode control scheme that can cater for the model uncertainties, external disturbances and the chattering phenomenon. Non- availability of states is a major constraint towards accomplish- ment of any control scheme, using sensor for each state is also not feasible due to space limitations and high cost of the sensors. Even with the availability of all states system model generally shows parametric mismatch with respect to the real time environment. These model imperfections, un-certain initialization and sensor errors also degrades the performance of the controller. The solution for that is to use state observers, to estimate the states in real time, Luenberger [15] proves to be good in the state estimation but these model based observers fail when the system parameters keep on changing with the time. Least square and recursive least square (RLS) are also not able to work on highly nonlinear system such as quadrotor. A high gain observer was first introduced by Khalil and Esfandiaro [16] for the design of output feedback controllers and asymptotic convergence. Researchers have contributed in the development of their idea towards high gain observer [17]– [20].High gain observers performance degrades in the presence of external noise and is shown in simulation section of this paper, due to which we have to look for an observer which is robust to sensor noise. The sliding mode observers are widely used because of their prominent features like finite time convergence, robustness to sensor noise and un-certain estimation [21], [22].Asymptotic second order sliding mode observers were also developed but they require proof of separation principle.High accuracy and reduction in chattering are the main features of second order sliding mode compared to the classical first order motion. Recently a class of second order sliding mode observer is introduced so called super-twisting observer [23] for second order mechanical system which include quadrotor too. Super- twisting observers can reconstruct the states if the perturbation is of relative degree two, or reconstructs the perturbation itself, when it is of relative degree one in finite time. Aim of this paper is to compare the performance of both ob- servers namely high-gain observer and super-twisting sliding mode observer under same set of perturbations, uncertainties and noise, so that each observer can exhibit its characteristics for the quadrotor system. Real time estimation always require knowledge of the pros and cons of the observer relative to the particular system, so that best observer can be deployed for real time estimation of states. So there is a need to explore the type of observers which are application specific. In the following section model of the quadrotor is developed and presented. In section-III controller design is presented. In section-IV observers are designed for the quadrotor model. In section-V numerical simulation is given and finally in section- V1 the conclusions are given. II. DYNAMIC MODELING A quad rotor UAV is a highly nonlinear dynamical system and its modeling it is not an easy task due to under actua- tion. It consists of two pairs of rotors which are moving in opposite directions to provide the collective thrust as shown in following fig 1. There are four inputs to this system. The input U1 is the sum of thrusts provided by individual rotors. The pitch movement is obtained by changing speeds of rotors 2 and 4. Similarly the roll movement is achieved by varying speeds of rotor 1 and 3. These two former operations should be performed while keeping the total thrust constant 1 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 11. otherwise system maylose altitude and crash. The roll and pitch movements are controlled by using inputs U2 and U3 respectively. The yaw movement occurs due to difference between torques of the two pairs of rotors. This movement is stabilized by using input U4. Fig. 1. Quad Rotor UAV Free Body Diagram Let us consider an inertial frame of reference E” and a body fixed frame B” as shown in above figure. The transformation between E and B is provided by a matrix R which is given by the following equation. xn yn zn = cos θ cos ψ − cos φ sin ψ + sin φ sin θ cos ψ sin φ sin ψ + cos φ cos ψ sin θ cos θ sin ψ cos φ cos ψ + sin φ sin θ sin ψ sin θ sin ψ cos φ − cos ψ sin φ sin θ cosθ sin φ cosθ cos φ xb yb zb xn yn zn = R xb yb zb The Newton-Euler formalism is used to present the dynam- ics of quad rotor UAV. The Newtons laws of motion when applied to a rigid body in the presence of external forces and torques are given by following set of equations msI3∗3 O O I ˙V ˙w + w ∗ msV w ∗ Iw = F τ where ms is the mass of the quadrotor, V vector of xyz, w be the vector of φ, θ and ψ,I represents a inertia vector Ix Iy Iz T across x, y and z respectively.τ is torque vector include roll torque,pitch torque and yaw torque and F = 0 0 U1 T .To convert the above equations in inertial frame we use transformation matrix to get the following equations ˙ξ = v ˙v = R F ms ˙R = R ˆw J ˙w = −w ∗ Jw + τ The crude and approximate model of quad rotor UAV from above set of equations can be written as follows ˙ξ = v ˙v = ge3 + Re3 b ms Ω2 i ˙R = R ˆw J ˙w = −w ∗ Iw − Jr w ∗ e3 Ωi + τ where d is some modeling co-efficient, e3 is a vector 0 0 1 T ,r is the rotor co-efficient, ξ is the position vector, R is the transformation matrix, ˆw is the skew-symmetric matrix, Ω is the rotor speed, I is the inertial tensor matrix, Jr is the rotor inertia, while Jp and Jm are the propeller and motor inertia respectively and b is the thrust co-efficient. The torques applied to the quad rotor’s axis is the difference between the torques provided by the rotors on the other axis. τ = lb Ω2 4 − Ω2 2 lb Ω2 3 − Ω2 1 d Ω2 2 + Ω2 4 − Ω2 3 − Ω2 1 The rotor inertia consists of motor inertia, propeller inertia and negligible reversing gearbox inertia and is given by the following equation Jr = Jp − Jmr Now the complete six degrees of freedom model is given by the following system of equations:- ¨x = cos φ sin θ cos ψ + sin φ sin ψ U1 m ¨y = cos φ sin θ sin ψ − sin φ cos ψ U1 m ¨z = −g + cos φ cos θ ms U1 ¨φ = ˙θ ˙ψ Iy − Iz Ix − Jr Ix Ωr ˙θ + l Ix U2 ¨θ = ˙φ ˙ψ Iz − Ix Iy − Jr Iy Ωr ˙φ + l Iy U3 ¨ψ = ˙θ ˙φ Ix − Iy Iz + C Iz U4 (1) where C is the proportional constant.The first term on right hand side of first dynamical equation is the gyroscopic effect caused by the rotation of the rigid body and the second term is due to the propulsion effect. The system inputs are U1, U2, U3 and U4. The inputs are given by the following equations U1 = b Ω2 2 + Ω2 4 − Ω2 3 − Ω2 1 U2 = b Ω2 4 − Ω2 2 U3 = b Ω2 3 − Ω2 1 U4 = d Ω2 2 + Ω2 4 − Ω2 3 − Ω2 1 Ω = d Ω4 + Ω2 − Ω3 − Ω1 2 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 12. To make themodel more realistic especially in forward flight we include the hub forces, rolling moments and variable aerodynamics coefficients [24]. The hub force is the resultant of horizontal forces acting on all blade elements. H = CHρA ΩRrad 2 where CH is the hub coefficient, ρ is the air density and A is the propeller disk area and Rrad is the propeller radius and ρ is the speed of the respective propeller.Additionally the drag moment Q is the moment about the rotor shaft caused by the aerodynamic forces acting on the blades. In fact drag moments determine the power required to spin the rotors. It is given by the following equation Q = CQρA ΩRrad 2 Rrad where CQ is the drag coefficient.The rolling moment of a propeller exists in forward flight when advancing blade is producing more lift than the retreating blade. It is the integration over the entire rotor of the lift of each section acting at a given radius and is given by following equation. Rm = CRm ρA ΩRrad 2 Rrad Where CRm is the rolling moment coefficient.Furthermore the UAVs operating near the ground (approximately at half rotor diameter) experience thrust augmentation due to better rotor efficiency. This is related to a reduction of induced airflow velocity. This is called Ground Effect. The following equation represents the ground effect near the surface. It is assumed that ground effect acts on the UAV when the UAV is below a certain altitude,zo. Fgr z = A z + zcg)2 − A zo + zcg)2 0 < z ≤ zo After incorporating the above effects and the friction terms, we obtain a more realistic model of the quad rotor UAV which is given by the following set of equations:- ¨x = cos φ sin θ cos ψ + sin φ sin ψ U1 m − 1 m 4 i=1 Hxi − K1 ˙x ms ¨y = cos φ sin θ sin ψ − sin φ sin ψ U1 m − 1 m 4 i=1 Hyi − K2 ˙y ms ¨z = −g + cos φ cos θ ms U1 + Fgr z − K3 z ms ¨φ = ˙θ ˙ψ Iy − Iz Ix − Jr Ix Ωr ˙θ + l Ix U2 − h Ix 4 i=1 Hyi+ (−1)i+1 Ix 4 i=1 Rms xi − lK4 ˙φ Ix ¨θ = ˙φ ˙ψ Iz − Ix Iy − Jr Iy Ωr ˙φ + l Iy U3 − h Iy 4 i=1 Hyi+ (−1)i+1 Iy 4 i=1 Rms yi − lK5 ˙θ Iy ¨ψ = ˙θ ˙φ Ix − Iy Iz + C Iz U4 + h Iz 4 i=1 Hyi+ l Iz Hx2 − Hx4 + Hy3 − Hy1 4 i=1 Qiyi− lK6 ˙ψ Iz (2) III. CONTROLLER DESIGN Let X = x, ˙x, y, ˙y, z, ˙z, φ, ˙φ, θ, ˙θ, ψ, ˙ψ, T and U = U1, U2, U3, U4 T be the state and control input vectors re- spectively. The equation set (1) with the addition of friction term and ground effect term for altitude can be written in state space representation such as: ˙x1 = x2 ˙x2 = cos x7 sin x9 cos x11 + sin x7 sin x11 U1 m − K1 x2 ms (3) ˙x3 = x4 ˙x4 = cos x7 sin x9 sin x11 − sin x7 sin x11 U1 m − K2 x4 ms (4) ˙x5 = x6 ˙x6 = −g + cos x7 cos x9 ms U1 + Fgr z − K3 x6 ms (5) 3 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 13. ˙x7 = x8 ˙x8= x10x12 Iy − Iz Ix − Jr Ix Ωrx10 + l Ix U2 − lK4 x8 Ix (6) ˙x9 = x10 ˙x10 = x10x12 Iz − Ix Iy − Jr Iy Ωrx8 + l Iy U3 − lK5 x10 Iy (7) ˙x11 = x12 ˙x12 = x10x8 Ix − Iy Iz + C Iz U4 + lK6 x12 Iz (8) The term represents the ground effect near the surface. It is assumed that the ground effect acts on the UAV when the UAV is below a certain altitude. The goal is to design a second order sliding mode control which is done in two steps. 1) Choice of sliding surface w.r.t tracking error e. 2) Design of Lyapunov function that guarantees negative definiteness so that asymptotic convergence is achieved. Closed loop control system dynamics become insensitive to modeling error, perturbation signals and parameter variation as a by-product of sliding mode control (SMC). Control efforts are calculated by the help of Lyapunov analysis and hence guarantee asymptotic convergence. As quadcopter has 4 inputs while number of variables to be controlled are more than four hence overall it is an under actuated system but we can portioned that system into two parts namely fully actuated part and under-actuated part and then designing the control for each part of the system independently. Therefore control is also portioned into two sub-system. A. Control for fully actuated subsystem Fully-actuated subsystem composed up of ¨z and ¨φ subsys- tems (5) and (6) respectively. Choice of sliding surface for the subsystem (6) comes out from the Lyapunov analysis as: V = e2 φ 2 where eφ = φd − φ ˙V = eφ ˙eφ ˙V = eφ ˙φd − ˙φ So to make ˙V negative definite ˙φ = ˙φd + α1 φd − φ Hence Surface S1 will be S1 = ˙φd − ˙φ + α1 φd − φ Let the Lyapunov function be V = S2 1 2 ˙V = S1 ˙S1 = S1 ¨φd − ¨φ − α1 ˙φd − ˙φ where α1 > 0 = S1 ¨φd − ˙θ ˙ψ Iy − Iz Ix − Jr Ix ˙θΩr − l Ix U2 + K4l ˙φ Ix + α1 ˙φd − ˙φ To make it negative definite choice of input is as: U2 = Ix l ¨φd − ˙θ ˙ψ Iy − Iz Ix − Jr Ix ˙θΩr + K4l ˙φ Ix + α1 ˙φd − ˙φ + k1sat S1 + k2S1 where k1, k2 > 0 Similarly in the same way surface for subsystem (5) comes out to be the linear combination of position and velocity tracking errors of z state. S2 = ˙zd − ˙z + α2 zd − z where α2 > 0 And in the same way by the Lyapunov analysis of surface the control comes out to be U1 = ms cos φ cos θ ¨zd + g + K3 ˙z ms + α2 ˙zd − ˙z + k2sat S2 + k4S2 + Fgr where k3, k4 > 0 Where B. Control for Under-actuated subsystem Under-actuated subsystem composed up of ¨x, ¨y, ¨θ and ¨ψ subsystems. The Choice of sliding surface for the subsys- tem (3) and (7) comes out from the Lyapunov analysis as: V = e2 x 2 + e2 θ 2 where ex = xd − x and eθ = θd − θ ˙V = ex ˙ex + eθ ˙eθ ˙V = ex ˙xd − ˙x + eθ ˙θd − ˙θ So to make ˙V negative definite ˙x = ˙xd + α3 xd − x and ˙θ = ˙θd + α4 θd − θ ˙V < −α3 xd − x 2 − α4 θd − θ 2 where α3, α4 > 0 Hence Surface S3 will be S3 = ˙xd − ˙x + xd − x + ˙θd − ˙θ + θd − θ Let the Lyapunov function be V = S2 3 2 ˙V = S3 ˙S3 = S3 ¨xd − ¨x + α3 ˙xd − ˙x + ¨θd − Iz − Ix Iy ˙φ ˙ψ + Jr Iy ˙φΩr − l Iy U3 + K5 l ˙θ Iy + α4 ˙ θd − ˙θ 4 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 14. To make itnegative definite choice of input is as: U3 = Iy l ¨θd − ˙φ ˙ψ Iz − Ix Iy − Jr Iy Ωr ˙φ − lK5 ˙θ Iy + α3 ˙xd − x + α4 ˙θd − ˙θ + ¨xd − ¨x + k5sat(S3) + k6S3 where k5, k6 > 0 Similarly in the same way surface for subsystem (4) and eqref3f comes out to be the linear combination of position and velocity tracking errors of two states i.e. y and ψ. S4 = ˙yd − ˙y + α5 yd − y + ˙ψd − ˙ψ + α6 ψd − ψ where α5, α6 > 0 and in the same way by the Lyapunov analysis of surface S4 the control U4 comes out to be U4 = Iz c ¨ψd − ˙φ ˙θ Ix − Iy Iz − lK6 ˙ψ Iz + α5 ˙yd − ˙y − α6 ˙ψd − ˙ψ + ¨yd − ¨y + k7sat(S4) + k8S4 where k7, k8 > 0 U1 is the control input of z, U2 is the control input of roll,U3 is the control input of pitch and U4 is the control input of yaw while motion in x and y direction is produced by help of control inputs of roll, pitch and z. By the help of U1, U2, U3 and U4 the desired trajectories are achieved and tracking errors are reduced to zero asymptotically, by virtue of Sliding mode controller.by keeping the roll and pitch angles to zero controller robustly stabilize the UAV and move it to the desired position with a desired yaw angle. The control scheme is developed and implemented independent of observer and is shown in the block diagram The controller is designed by keeping in view the mathematical model of the quad rotor as given in Section-II without any effects except friction and the ground effect but that U1, U2, U3 and U4 are capable enough to tackle not only Rolling Moments, Drag moments, Gyroscopic effects, Hub forces but also retain its performance which is being shown in the simulations section in this paper. Fig. 2. Controller designed independent of the observer IV. OBSERVER DESIGN Two types of nonlinear observers are implemented for the quad rotor system with the same control scheme i.e. 2-Sliding mode Control. Observability is ensured by [25] for each block of equation from (3)–(8) separately. A. High Gain Observer HGO is basically an approximate differentiator. This ob- server works well for a wide class of nonlinear systems and leads to recovery of the performance achieved under state feedback. Implementation of this observer is quite simple because it needs less computational effort with an additive advantage of this observer is that its performance doesn’t degrade with the presence of model uncertainties in the plant. High gain observer is an asymptotic observer and dynamics of this observer can be made arbitrarily fast through epsilon and gains alpha’s. Separation principle theorem doesn’t need to be proved and high gain observer can be designed separately from the controller. The HGO is applied to multiple input and multiple output system as: ˙x = Ax + Bϑ x, u The HGO, then, is given by ˆ˙x = Aˆx + Bϑ0 ˆx, u + H y − Cˆx y = Cˆx Where H = blockdiag H1, H2, H3, H4 , Hi = ∂1 1 ε ∂2 2 ε i = 1, 2, 3, 4 ε = positive constant, and constant parameter ∂i j are obtained from a Hurwitz polynomial, j=1,2 s2 + ∂1 1 s + ∂2 2 = 0 The HGO for the quadcopter system is designed in blocks as [26] i.e six HGOs are designed for each block of equation from (3)–(8) separately. For equation (3) HGO is implemented as With A = 0 1 0 0 , B = 0 1 and C = 1 0 and ϑ0 = cos x7 sin x9 cos x11 + sin x7 sin x11 U1 m − K1 x2 ms and H1 is designed as in the aforementioned equation. Simi- larly for equation (4), (5),.....(8).HGOs are given in the same way as for equation (3).constant gains are enlisted in the table 1 in simulation section of this paper. 5 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 15. B. Sliding modeObserver with super-twisting Algorithm One of the best sliding mode observer which offers a finite time reaching time [27], [23] and which can be used for sliding mode based observation is the super-twisting ob- server.Separation principle theorem is trivial in this case too and the Super-twisting sliding mode observer can be designed separately from the controller. The finite time convergence property of sliding mode observers is usually suitable in the scheme of observation and for the purpose of observer-based controller design for nonlinear systems Super-twisting sliding mode observer has the form ˆ˙x1 = ˆx2 + λ|x1 − ˆx1|1/2 sat x1 − ˆx1 (9) ˆ˙x2 = f x1, ˆx2, u + τsat x1 − ˆx1 (10) Taking ˜x1 = x1 − ˆx1 and ˜x2 = x2 − ˆx2 we obtain the error equations as ˙˜x1 = ˜x2 − λ|˜x1|1/2 sat ˜x1 ˙˜x2 = F t, x1, x2, ˆx2 − αsat ˜x1 where, F x1, x2, ˆx2 = f x1, x2, u − f x1, ˆx2, u + ξ x1, x2, y ξ is used for perturbations.For the bounded states, existence of a constant is ensured such that |F x1, x2, ˆx2 |< f+ Observer designed by equation (9) and (10) takes into ac- count of partial knowledge of system dynamics while setting parameters λ and τ and hence more accurate. The full order Super-twisting Sliding mode observer for equation (3) is given as ˆ˙x1 = λ1 + ˆx1|x1 − ˆx1|1/2 sat x1 − ˆx1 ˆ˙x2 = 1 ms cos ˆφ sin ˆθ cos ˆψ + sin ˆφ sin ˆψ U1 − K1 ˆx2 ms τ1sat x1 − ˆx1 τ1 and λ1 are designed by the help of aforementioned in- equality as [27]. Similarly for equations (4), (5),....., (8) Super- twisting sliding mode observers are implemented in the same way as for equation (3), while gains are given in the table.1 in simulation section of this paper. V. SIMULATION STUDY A. Closed loop Simulation with model uncertainties and with- out noise for HGO and STSMO Simulation results for observer-based controller of the quadrotor are shown in the fig.3 and fig.4 for both observers HGO and STSMO. Now under output feedback, controller is in conjunction with HGO and STSMO separately and is using all the states of the observer. The results in the fig.3 shows that both High gain observer and as well as Super- twisting sliding mode observer recovers the performance of state feedback after 12 seconds while x state a bit earlier as compared to other states, both observers are initialized with same initial conditions, so that performance can be compared in a proper way and all the observers parameters are listed in the table 1 the performance of HGO is slightly better than STSMO in tracking as HGO estimates desired values a bit earlier as compared to STSMO. The chattering problem is intelligently avoided in the sliding mode control by using continuous approximation to the sign function. This makes this approach applicable in real applications. As the control laws are developed for set of equations (1) but implemented on set of equations (2) which include different kind of effects as mentioned in section-II, similarly the model used for observer is based on set of equations (1) and observer giving estimate for set of equations (2) which is quadrotor model with ground effects, drag moment, rolling moment and pitch moment. Fig. 3. HGO and STSMO tracking of desired values Fig. 4. HGO and STSMO tracking of desired values B. Closed loop Simulation with noise and model Uncertainties for HGO and STSMO A constant noise of 0.1 value is added in the output of the system in each of six states and the results obtained after simulation for each observer are shown in the fig.5 and fig.6 6 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 16. which indicated thatin the case of STSMO no effect on the observer’s estimated states while on the other hand HGO estimates deviated by the same amount as disturbance added which opens an era of coupling integral control scheme with HGO to eliminate the steady state error Fig. 5. HGO Estimated values under constant sensor noises Fig. 6. STSMO Estimated values under constant sensor noise C. Effect of observer scheme on Control effort required Control effort is greatly affected by the effects namely drag moment, rolling moments, pitch moments and hub forces which are included in the system model but not included in any of the observer, and by the type of observer used. Simulations results in fig.7 and fig.8 shows that HGO required larger control effort in transient phase as compared to STSMO and fig.9 shows that over all with the conjunction of observer in the closed loop model control effort in transient phase in considerably increased Fig. 7. Control Effort with STSMO *Initial conditions for the quadrotor system are set to zero deliberately for evaluating performance comparison of both observers. Fig. 8. Control Effort with HGO Fig. 9. Control Effort under the state feedback TABLE I THE NOMINAL PARAMETERS AND THE INITIAL CONDITIONS OF THE OBSERVER AND THE SYSTEM FOR THE QUADROTOR MODEL Variable Value Units Initial Condition High Gain Super Twisting Sliding Observer Mode Observer ms 1.1 kg ˆx1(0) 1 1 l 0.21 m ˆx2(0) 2 2 Ix = Iy 1.22 Ns2/rad ˆx3(0) 0.6 0.6 Iz 2.2 Ns2/rad ˆx4(0) -2 -2 lr 0.2 Ns2/rad ˆx5(0) 2 2 K1, K2, K3 0.1 Ns/rad ˆx6(0) 1 1 K4, K5, K6 0.12 Ns/rad ˆx7(0) -1 -1 g 9.81 m/s2 ˆx8(0) 1 1 b 5 Ns2 ˆx9(0) 0.5 0.5 C 1 ˆx10(0) 1 1 k1, k3 0.8 ˆx11(0) 1.3 1.3 k2, k4 2 ˆx12(0) 3 3 k5, k7 0.5 k6, k8 5 α1, α3, α5 2 α2, α4, α6 6 zcg 0.1 4 ∂1, ∂2, ∂3, ∂4 2,1,4,4 ∂5, ∂6, ∂7, ∂8 6,9,10,25 ε 0.9 λ 1 τ 5 ∂9, ∂10, ∂11, ∂12 2,1,6,9 VI. CONCLUSION This paper has presented a comparison study of nonlinear observers, including high gain observer and super-twisting sliding mode observer in conjunction with the 2-Sliding mode controller for the quadrotor system under external disturbances and model uncertainties. The highgain observer can cater for the model uncertainties but not the external disturbance while the super-twisting sliding mode observer not only cater for the model un-certainities but can also performs well under external disturbances (sensor noise). The second important result regarding initialization of high-gain observer is that it 7 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 17. doesnt allow randominitialization unless gains are adjusted on the other hand super-twisting sliding mode observer provides flexible environment in initialization. None of these observes is computationally onerous, but super- twisting sliding mode observer utilizes the knowledge of system partially [27] as compared to high-gain observer which just rely on the Hurwitz polynomial and the tuning parameter ε [26] so this fact also indicates the practical applicability of super-twisting sliding mode observer in the cases where model uncertainties are bounded as it gives finite time convergence as compared to asymptotic convergence in the case of high gain observer which are favourable in the environment where model uncertainties are present or parameters are time varying in those conditions these filters are preferable than super-twisting sliding mode observer only. This effort will be a good starting point to explore super-twisting sliding mode observers and to compare them with other observers of its breed (higher order sliding mode observers). REFERENCES [1] M. Chen and M. Huzmezan, “A simulation model and h (loop shaping control of a quad rotor unmanned air vehicle.” in Modelling, Simulation, and Optimization, 2003, pp. 320–325. [2] S. Bouabdallah, P. Murrieri, and R. Siegwart, “Design and control of an indoor micro quadrotor,” in Robotics and Automation, 2004. Proceedings. ICRA’04. 2004 IEEE International Conference on, vol. 5. IEEE, 2004, pp. 4393–4398. [3] B. Heriss´e, T. Hamel, R. Mahony, and F.-X. Russotto, “Landing a vtol unmanned aerial vehicle on a moving platform using optical flow,” Robotics, IEEE Transactions on, vol. 28, no. 1, pp. 77–89, 2012. [4] L. Derafa, A. Benallegue, and L. Fridman, “Super twisting control algorithm for the attitude tracking of a four rotors uav,” Journal of the Franklin Institute, vol. 349, no. 2, pp. 685–699, 2012. [5] A. Tayebi and S. McGilvray, “Attitude stabilization of a vtol quadrotor aircraft,” Control Systems Technology, IEEE Transactions on, vol. 14, no. 3, pp. 562–571, 2006. [6] L. Besnard, Y. B. Shtessel, and B. Landrum, “Quadrotor vehicle control via sliding mode controller driven by sliding mode disturbance observer,” Journal of the Franklin Institute, vol. 349, no. 2, pp. 658–684, 2012. [7] M. Bouchoucha, S. Seghour, and M. Tadjine, “Classical and second order sliding mode control solution to an attitude stabilization of a four rotors helicopter: From theory to experiment,” in Mechatronics (ICM), 2011 IEEE International Conference on. IEEE, 2011, pp. 162–169. [8] E. Altu˘g, J. P. Ostrowski, and R. Mahony, “Control of a quadrotor helicopter using visual feedback,” in Robotics and Automation, 2002. Proceedings. ICRA’02. IEEE International Conference on, vol. 1. IEEE, 2002, pp. 72–77. [9] E. Altu˘g, J. P. Ostrowski, and C. J. Taylor, “Quadrotor control using dual camera visual feedback,” in Robotics and Automation, 2003. Proceedings. ICRA’03. IEEE International Conference on, vol. 3. IEEE, 2003, pp. 4294–4299. [10] T. Madani and A. Benallegue, “Control of a quadrotor mini-helicopter via full state backstepping technique,” in Decision and Control, 2006 45th IEEE Conference on. IEEE, 2006, pp. 1515–1520. [11] ——, “Backstepping sliding mode control applied to a miniature quadro- tor flying robot,” in IEEE Industrial Electronics, IECON 2006-32nd Annual Conference on. IEEE, 2006, pp. 700–705. [12] P. Castillo, P. Albertos, P. Garcia, and R. Lozano, “Simple real-time attitude stabilization of a quad-rotor aircraft with bounded signals,” in Decision and Control, 2006 45th IEEE Conference on. IEEE, 2006, pp. 1533–1538. [13] N. Metni, T. Hamel, and F. Derkx, “Visual tracking control of aerial robotic systems with adaptive depth estimation,” in Decision and Con- trol, 2005 and 2005 European Control Conference. CDC-ECC’05. 44th IEEE Conference on. IEEE, 2005, pp. 6078–6084. [14] A. Benallegue, A. Mokhtari, and L. Fridman, “Feedback linearization and high order sliding mode observer for a quadrotor uav,” in Variable Structure Systems, 2006. VSS’06. International Workshop on. IEEE, 2006, pp. 365–372. [15] D. G. Luenberger, “Observers for multivariable systems,” Automatic Control, IEEE Transactions on, vol. 11, no. 2, pp. 190–197, 1966. [16] F. Esfandiari and H. K. Khalil, “Output feedback stabilization of fully linearizable systems,” International Journal of control, vol. 56, no. 5, pp. 1007–1037, 1992. [17] A. N. Atassi and H. K. Khalil, “A separation principle for the stabiliza- tion of a class of nonlinear systems,” IEEE Transactions on Automatic Control, vol. 44, no. 9, pp. 1672–1687, 1999. [18] A. Isidori, “A remark on the problem of semiglobal nonlinear output regulation,” IEEE transactions on Automatic Control, vol. 42, no. 12, pp. 1734–1738, 1997. [19] Z. Lin and A. Saberi, “Robust semiglobal stabilization of minimum- phase input-output linearizable systems via partial state and output feedback,” Automatic Control, IEEE Transactions on, vol. 40, no. 6, pp. 1029–1041, 1995. [20] W. J. Rugh, Linear system theory. prentice hall Upper Saddle River, NJ, 1996, vol. 2. [21] J. Barbot, M. Djemai, and T. Boukhobza, “Sliding mode observers,” Sliding Mode Control in Engineering, vol. 11, 2002. [22] C. Edwards, S. K. Spurgeon, and C. P. Tan, “On the development and application of sliding mode observers,” in Variable Structure Systems: Towards the 21st Century. Springer, 2002, pp. 253–282. [23] J. Davila, L. Fridman, and A. Levant, “Second-order sliding-mode observer for mechanical systems,” IEEE transactions on automatic control, vol. 50, no. 11, pp. 1785–1789, 2005. [24] M. Becker, R. C. B. Sampaio, S. Bouabdallah, V. d. Perrot, and R. Siegwart, “In-flight collision avoidance controller based only on os4 embedded sensors,” Journal of the Brazilian Society of Mechanical Sciences and Engineering, vol. 34, no. 3, pp. 294–307, 2012. [25] I. Khan, A. Bhatti, Q. Khan, and Q. Ahmad, “Sliding mode control of lateral dynamics of an auv,” in Applied Sciences and Technology (IBCAST), 2012 9th International Bhurban Conference on. IEEE, 2012, pp. 27–31. [26] H. K. Khalil and J. Grizzle, Nonlinear systems. Prentice hall New Jersey, 1996, vol. 3. [27] Y. Shtessel, C. Edwards, L. Fridman, and A. Levant, Sliding mode control and observation. Springer, 2014. 8 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 18. Fuzzy Temperature Controllerfor Induction Heating Bushra Aijaz Dept. of Electrical Engineering Bahria University Karachi, Pakistan bushra.aijaz@bimcs.edu.pk Rahema Kaleem Dept. of Electronic Engineering NED University of Engg & Tech Karachi, Pakistan rahemakaleem@gmail.com Naeema Saeed Dept. of Electronic Engineering NED University of Engg & Tech Karachi, Pakistan naeemasaeed991@hotmail.com Abstract— this paper focuses on Fuzzy Temperature Controller (FTC) for induction heating, which aims to provide a precise and intelligent temperature controller. Induction heating is a non contact process and so it is safer. The heating all depends on the larger number of Cu and Eddy losses and thus faster. Variable frequency Inverter is used to achieve temperature control for induction heating coils. By controlling the output frequency of inverter, the temperature of the load is controlled. Fuzzy logic is implemented for overall controlling. FTC is modelled on DSP Starter Kit DSKC6713. All programming is written in C- language. The model of this idea has been designed and tested using LabVIEW 8.5. Keywords – Fuzzy Logic, Induction Heating, DSP, Variable Frequency Inverter. I. INTRODUCTION Precise temperature controlling and fast heating processes are integral part of industries. The industries require a new modern technique which can handle temperature controlling with more accuracy and precision as the controllers currently used are slow and inaccurate plus they are unsuitable for non- linear measurements. The block diagram of the system is shown in Figure 1 Project Block Diagram. Figure 1 Project Block Diagram The system consists of DSK DSPC6713 for generating SPWM pulses, isolation circuit inverter, transformer (step-up) and temperature sensing circuit. SPWM pulses are fed to gate drivers and then to inverter. The output is then fed to step up transformer to obtain the desired output level which is then fed to the load for heating. Controlling of temperature is done by controlling the operating frequency of inverter. The rest of the paper is organized as follows: in Section II, we describe the inverter in brief. Fuzzy Logic Controller is described in Section III, SPWM is generated in Section IV, Temperature Sensors are interfaced in section V. Induction Heating is discussed in Section VI and the paper is concluded in Section VII. II. INVERTER This section focuses on DC to AC inverter. The purpose is to efficiently convert a DC power source to a high voltage AC source, which will be required to drive the load. This is achieved by first converting low voltage DC power source to high voltage DC power source and then the HV DC power to AC power source using sinusoidal pulse width modulation technique, the output of which is 220Vac. As the induction coils operate at much higher frequencies so a high frequency inverter is needed. To accomplish this, the converter is designed using a full-wave rectifier. Smoothening capacitors are connected at the output of the full-wave rectifier to convert rectified pulsated output to smooth DC output voltage. This output is then given to H-bridge that gives AC square wave of 220V. The circuit and output is shown in Figure 2. 9 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 19. Figure 2 Designand Simulation of H- Bridge To minimize the power loss and to ensure high switching speeds, N-channel MOSFETs IRF840 are used in H- Bridge. MOSFET gate drivers are used and controlled through SPWM signals coming from the DSPC6713. The values for bootstrapping capacitors and diodes are calculated using Equation 1. The output of H-Bridge gives 220 Vac whose frequency can be varied to more than 200Hz. III. FUZZY LOGIC CONTROLLER Fuzzy logic is doing all sort of controlling. The logic was first proposed by Zadeh in 1965. The Fuzzy logic is a linguistic logic. It is similar to crisp logic but with the difference is that crisp logic has only two levels of decision either 0 or 1 but fuzzy logic can have levels in between them, which makes the logic precise and close to ideal behaviour. The fuzzy structure is based on following three steps; A. Pre-processing As shown above; Error and rate-of-error are the inputs to the fuzzy controller, whereas error is the difference of set temperature and Feedback temperature. The inputs are properly assigned with their membership (µ) functions (after observing temperature rise and fall graph described in section V). Figure 3 and Figure 4 shows the µ function assignments of the two inputs ΔT and ΔT/Δt and Figure 5 shows the µ functions of output ΔF. Figure 3Membership functions for input Fuzzy variable ΔT Figure 4 Membership functions for input Fuzzy variable ΔT/Δt Figure 5 Membership functions for output Fuzzy variable ΔF B. Fuzzy Inference A set of rules is defined for controlling purpose. In this project, Fuzzy logic consists of 5 levels of decision for both the inputs and output. So the rule set comes up with 25 levels. The Table 1demonstrates the set of rules for ΔF. Table 1 Fuzzy Rule Base Delta T Delta (T/t) NB NS Z PS PB NB NB NB NB NS Z NS NB NB NS Z PS Z NB NS Z PS PB PS NS Z PS PB PB PB Z PS PB PB PB C. Defuzzification After evaluating the rule(s), the system comes up with a certain output frequency (ΔF) which is then fed to DSP controller. The ΔF tells the inverter how much (variable) frequency it has to produce for required heating. Figure 6 shows the output ΔF for two inputs ΔT and ΔT/Δt. ΔT ΔT/Δt Fuzzy Controller ΔF 10 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 20. Figure 6 simulationoutput for given input sample values IV.SPWM GENERATION SPWM pulses are constant amplitude pulses with different duty cycles for each period. For SPWM generation reference signal is compared with carrier (triangular wave). A sinusoidal waveform signal is used as a reference signal to shape the output (AC voltage) close to sinusoidal. Figure 7 shows the comparator circuit for SPWM generation. Figure 7 Comparator for SPWM generation For real time frequency variation of SPWM signals, variable frequency sine wave is created. For sampling frequency of 8KHz, frequency of generated sine wave, Equation 2 is used, Where n represents number of points inputted for sine wave generation. With the change in value of ‘n’, the frequency of sinusoidal wave gets changed. Figure 8 shows the output behaviour of Variable frequency sine wave generated. Figure 8 Output of Variable frequency sine wave SPWM signals are generated by comparing triangular wave (1KHz) with variable frequency sinusoidal wave. The amplitude modulation ratio is set to 0.9 and frequency modulation ratio varies as frequency of reference signal varies, determined by Equation 3; Figure 9 shows the successful generation of SPWM wave on CCS graph. Figure 9 SPWM generation on CCS graph V. TEMPERATURE SENSOR INTERFACING Temperature detectors are necessary element in order to carry out the temperature controlling. The output of sensor is connected to DAQ (NI-PXI-6229) to link the sensor with the software environment. This output of DAQ assistant is multiplied with sensor sensitivity. The output of multiplier block is fed to formula block and collector block. The formula block is converting the analog signal into o Centigrade and the collector block produces the mean of collected readings. The collector output is fed to signal conditioning block. The readings finally coming out are being written to measurement file block. The whole circuit is enclosed in a while loop to carry out the reading process continuously as shown in Figure 10. Figure 10 VI window for Temperature Sensor Interfacing Circuit The temperature is forced to bring around 70 to 75o C. The change in temperature is noted down in the “write to measurement file”. This data is plotted and the rise and fall response of temperature is modelled as shown in Figure 11 and Figure 12 below: Figure 11 sensor response when temperature rises V8 TD = TF = .00002775 PW = 5.55e-7 PER = 5.55e-5 V1 = 5v TR = 0.00002775 V2 = -5v U1 OPAMP + - OUT 0 5.000V0 0V R1 1k 0V V2 FREQ = 1khz VAMPL = 4v VOFF = 0v V 0 11 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 21. Figure 12 sensorresponse when temperature falls VI.INDUCTION HEATING Induction heating has replaced the traditional furnace methods because of its efficiency, unpolluted and fast heating process. Electrically conductive materials are used in induction heating for heating purposes and this requires high frequency electricity. An IH system requires a source of alternating current, an induction coil, and the work piece to be heated. A magnetic field is generated in the coil due to the alternating current passing through the coil. The AC is supplied by the inverter. Work piece placed within the coil will experienced the magnetic field due to which eddy currents are induced in the work piece that cause non-contact type of heating between work piece and the induction coil. Copper tube is used to make induction coil. The tube is hollow inside with coil diameter about 3 inches and internal diameter of tube is about 1cm.the coil is given 8 turns. The impedance of coil depends on cross-sectional area, length and the number of turns so to increase the coil impedance. The impedance matching circuit is designed such that it converts high volt/low current (coming out from inverter) to low volt/high current (driven requirement of the load). The choice is made for impedance matching in order to match the output parameters of inverter with the input parameter of coil. Parallel capacitor bank is to be connected between the coil and inverter. VII. CONCLUSION The FLC for induction heating has been presented in this paper. The integrated controller, implemented on DSKC6713 takes the set temperature (required temperature of load) input from user, reads actual temperature of load and calculates error and rate of error, then on basis of Fuzzy Logic, it takes decision and determines the amount of ΔF that needs to be added/subtracted. The FLC then feeds this ΔF to/from operating frequency of variable frequency inverter. The FL is being implemented in international industries but is new for Pakistani industries. However, Pakistan industries are doing their controlling on fuzzy logic but it is a joint venture of PID and FUZZY. This idea, however, introduces Fuzzy Temperature Controller as a standalone product. BIBLIOGRAPHY [1] Chin-Hsing Cheng, “Design of Fuzzy Controller for Induction Heating Using DSP”, 5th IEEE Conference on Industrial Electronics and Applications, 2010 [2] Yunseop Kim, “Fuzzy Logic Temperature Controller”, Physics 344 project, 2001 [3] Datasheet IRF840: http://www.datasheetcatalog.com/datasheets_pdf/I/R/F/8/IRF840.sht ml [4] HV Floating MOS-Gate Driver ICs, Application Note AN-978, http://www.irf.com/techincal-info/appnotes/an-978.pdf [5] Zememe Walle Mekonnen, “Digital Signal Processing Applications using C6713 DSK”, project work [6] S. Zinn, S.L Semiatin, “Elements of Induction Heating, Design, Control and application” 12 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 22. Stability and ControlSolution to Quad-copters Q. Ejaz Ur Rehman1 , S.Akhtar1 , A.Saleem1 Department of Avionics Engineering1 National University of Sciences and Technology Islamabad, Pakistan qejaz@cae.nust.edu.pk Suhail@cae.nust.edu.pk ammar.saleem@cae.nust.edu.pk Abstract ‒ A Quad-copter is a structurally simple and dynamically complex rotorcraft, lifted and propelled by four rotors. It has very small size and is highly maneuverable as compared to conventional helicopters. In this paper, a method to achieve control and stability of a quad-copter is presented. Computational tools employed are MATLAB® /Simulink® , Catia® , LabVIEW® , ANSYS® and Arduino IDE® . A Mathematical model of quad- copter dynamics is developed using set of derived nonlinear equations accompanied by control theory. This nonlinear mathematical model is linearized in Matlab and LabVIEW® . Linear equations are used to design Linear Quadratic Regulator (LQR) controller. Microcontroller and sensor used are Arduino Mega 2560 and 6 DoF1 IMU2 . Stability of quad-copter is validated through experiments and simulations. Keywords: Quad-copter, Stability, Controllability, LQR, Mathematical Modeling, IMU I. INTRODUCTION In recent years, the aeronautical industry has shown a growing interest in UAVs (Unmanned Aerial Vehicles). UAVs are growing in popularity in fields of medicine, engineering, civil and most importantly, military and security. The reduced cost, absence of a trained pilot and small compact size make them viable options for tasks that include inhospitable terrain and remote regions. Quad-copters are the conventional remote control airborne vehicles with four rotors placed at equivalent distance from center of gravity. Quad-copter is elevated and driven by these four rotors only. Quad- copters are structurally simple and unstable unmanned air vehicles (UAVs). Due to its simple structure it is popular UAV nowadays, being used for surveillance, 1 Degree of Freedom 2 Inertial Measurement Unit aerial photography, Bomb search and disposal, vision based pose estimation and Fertilizer/ Pesticide sprayer etc. Unlike most helicopters, Quad-copters use two sets of indistinguishable static level propellers (two clockwise (CW) and two counter-clockwise (CCW)) which are set in an X or + (plus) configuration with X being the preferred configuration as shown in figure 1. These use deviation of RPM to control thrust and torque. Roll, Pitch and Yaw of quad-copter is achieved by altering the rotation rate of one or more rotor discs, thereby changing its torque load and thrust/lift characteristics. Figure 1: Rotors 1 and 3 rotate in one direction, while rotors 2 and 4 rotate in the opposite direction, controlling opposing torques for controllability and stability. The dynamics linked to employing four rotors mounted on edges of a square shape create a highly unstable platform that can only be controlled by embedding complex algorithms onboard. Due to the dynamically unstable nature of rotors, complex control mechanisms are required for a sustained flight. 13 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 23. In this paper,a method to achieve controllability and stability of quad-copter at certain height is achieved such that it is stationery with respect to the earth frame of reference at certain height. Simulation platform used are MATLAB® and LabVIEW® , while detail study of quad-copter and propeller is conducted in ANSYS-FLUENT® . CAD models are modelled in CATIA® software. The algorithm is written by manipulating the non-linear differential equations with control theory. The algorithm written is verified by visualizing results, animations and virtual reality model to completely study the quad-copter behaviour and response to inputs. The algorithm is translated into equivalent C language for Hardware testing. Microcontroller and sensor used are Arduino Mega 2560 and 6-DoF Razor IMU only. II. EQUATION OF MOTION The governing equations for the control of quad-copter are derived in this section. First of all, translational and rotational dynamics of quadrotor are explained followed by simplifications. Bold symbols are used to denote three-dimensional vectors, while non-boldface symbol are used for scalars in the paper. Figure 2: (A) Dynamic model of a quad-copter with four propellers in Earth frame of reference (B) Propeller i producing fi thrust with 𝜔𝑖 rpm in z- direction A. Dynamics A quad-copter is a UAV having four rotors and a mass ‘m’. The forces which act on a quadrotor are its weight and the thrust f produced by four propellers in body fixed direction z = (0, 0, 1). Similarly, four torques acts on each propeller and a total drag torque acts on quad- copter body. The rotation of the body fixed frame with respect to some inertial frame is described by the rotation matrix R which is discussed in detail. Two coordinate systems are considered in Figure 4 [3]: The inertial frame (E-frame) The body-fixed frame (B-frame) Figure 3: Quad-copter Frame of References These are related through three rotations: Roll: Rotation of φ around the x-axis; Pitch: Rotation of θ around the y-axis; Yaw: Rotation of ψ around the z-axis. The following assumptions were made in this approach: The body-fixed frame origin and center of mass (COM) of the body of the vehicle are coincident. The axes of the B-frame coincide with the body principal axes of inertia. Figure 4: Quad-copter configuration frame system Given equation describes the relation of Rotation matrix with roll, pitch, yaw and quad-copter position in earth frame. R (,, ) R(x,) R(y,) R (z, ) ( 1 ) The main equation governing the quad-copter dynamics is: 14 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 24. m𝐝̈ = 𝐑𝐳∑ fi +4 i=1 m𝐠 ( 2 ) Where d= (d1, d2, d3) are position vectors in inertial frame of reference and g = (0, 0, -9.81) is gravitational constant. As quad-copter is a rigid body with four propellers. The body inertia is expressed as diagonal matrix IB = [ Ixx B 0 0 0 Iyy B 0 0 0 Izz B ] ( 3 ) The propeller is a symmetric body with respect to its axis of rotation and can be considered as disc for simplification. The propeller inertia is given by: IP = [ Ixx P 0 0 0 Iyy P 0 0 0 Izz P ] ( 4 ) Due to symmetry of propeller Ixx P = Iyy P . The angular velocity of a body can be governed using the differential equation: [2] τres = τB + ∑ τP + ⟦ωB × ⟧(LB +4 i=1 ∑ IP ωB + ∑ LP 4 i=1 4 i=1 ) ( 5 ) Where τres is the resultant torque acting on quad- copter body. τB is the body torque τB = (Ixx B ṗ, Iyy B q̇, Izz B ṙ) ( 6 ) And τP is the torque produced by propeller. τP = (0, 0, Izz P ωi̇ ) ( 7 ) B. Simplification of Assumptions The effect of all the moments acting upon the body is denoted by τres on the right hand side of equation (5). These include moments due to propeller forces and torques due to motors. The propeller forces are assumed to act through the center of each propeller. It is assumed that the center of propeller is at a horizontal distance l from the body center of mass. τres = [ (f2 − f4)l + τdx (f3 − f1)l + τdy τ1 + τ2 + τ3 + τ4 + τdz ] ( 8 ) With τd = (τdx , τdy , τdz ) is the drag torque. It has been observed that propellers reaction torque has a linear relation with the thrust force (with proportionality constant of kτ and sign given by the sense of rotation). τi = (−1)i+1 κτfi ( 9 ) And thrust is related to rotor rpm as fi = κfωi 2 ( 10 ) Solving the above equation from 3 to 8 results in [2] Ixx B ṗ = κf(w2 2 − w4 2)l − (Izz T − Ixx T )qr − Izz P q(w1 + w2 + w3 + w4), ( 11 ) Ixx B q̇ = κf(w3 2 − w1 2)l + (Izz T − Ixx T )pr + Izz P p(w1 + w2 + w3 + w4), ( 12 ) Izz B ṙ = −γr + κτκf(w1 2 − w2 2 + w3 2 − w4 2 ) ( 13 ) III. CONTROLLABILITY These non-linear equations are linearized to compute state space matrices A, B, C and D. An LQR controller was designed with the cost value of 0.5 s2 rad-2 on the angular rates, 10 on the deviation from the primary axis, 0 on the extended motor states and 0.75 N-2 on the inputs. IV. EXPERIMENTAL PLATFORM Computational Programming softwares employed were MATLAB® / SIMULINK® , LabVIEW® , ANSYS® , Arduino IDE® and CATIA® . CAD models of Quad-copter and equivalent propeller were modelled in CATIA® and were exported to ANSYS-FLUENT® where surface meshing and computational fluid dynamics (CFD) was done to study aerodynamic design such that fluid was passed on to the quad-copter and propeller at different speeds and direction to check its serviceability. The propeller produced vibrations which were verified from CFD analysis. These vibrations were catered using prop-balancer system. 15 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 25. V. MODELLING OFPROPELLER A. Actuator Disk Theory A propeller can be represented as a single disk operating in a stream tube. As the flow passes through the Disk its velocity decreases while pressure increases. The disk is infinitely thin but has an area. Here propeller acts like it is made up of infinite blades. This disk produces pressure jump across it which is equal to thrust per unit area of disk. B. Geometry of Fan In order to model the propeller a thin circular surface of area equal to swept area of propeller was created around the hub. This thin surface was enclosed in a disk shaped volume such that the diameter of disk and thin surface was same. Two fluids, fluid fan 1 and fluid fan 2 were defined on both sides of thin in between thin and disk surface. Figure 7: Thin enclosed in disk Figure 8: Volume Mesh with Quality Figure 9: Vectors through thin The inertia of the quad-copter was measured by suspending the quad-copter about 3 different axis and measuring its period of oscillation. The inertia is given by: I = k (2πT)2 ( 14 ) Where k is the torsion constant and T is the period of oscillation of quad-copter with reference to Figure 5: Actuator Disk in flow stream Figure 6: Thin with fluid fan 16 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 26. equilibrium position. Dueto quad-copter symmetry Ixx P = Iyy P . The propeller inertia was approximately measured by considering propeller as disc and motor as cylinder. As propellers are fixed in zth direction so, Ixx P = Iyy P = 0 The quad-copter’s mass was determined to be 1 Kg. The distance from the center of gravity of the quad- copter to center of propeller was 0.24m. The propeller reaction torque and force constant were estimated as κτ = 0.214272 Nm N and κf = 1.80899e − 5 Ns2 rad2. VI. VOICE CONTROL PANEL A simulation based voice control panel was also developed in LabVIEW® to control quad-copter in 6- DoF which is shown in Figure 8. It operates on the certain commands and performs tasks as per the commands which are embedded in the algorithm. Figure 10: Voice Command Panel in LabVIEW® for quad-copter VII. Mathematical Model A mathematical model was developed in LabVIEW® and MATLAB® using nonlinear equations discussed in section II. Implemented nonlinear equations were linearized to design LQR controller for feedback control. Implemented mathematical model, results and animation is shown: Figure 11: Non-Linear Mathematical Model 17 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 27. Figure 12: SimulationResults From figure 12 it can be seen p, q, r (angular rates (degs/s) about x, y and z axis, respectively) showed variation at the start of quad-copter but came to equilibrium position at t=2, 5, 3 sec respectively. Subsequently roll, pitch and yaw angle also settled after 5 sec of takeoff. While linear velocities U and W about x and z axis rises continuously and linear velocity V about y-axis comes to equilibrium position after 10 sec which depicts the quad-copter has gained height of 10 ft. above ground as justified from X, Y, Z in the figure 10. Figure 13: Throttle and RPM relation with time Figure 13 shows the variation of produced thrust from throttle given at certain time t for each rotor. Figure 14: Animation GUI of Hover Model An animation GUI has also been implemented which shows the behavior of quad-copter and also display the trajectory made by quad-copter to reach height of 10 ft. Figure 15: Virtual Reality Quad-copter Model A quad-copter model has also been made using Matlab® virtual reality toolbox in order to make the simulations near to reality. This virtual model take roll, pitch, yaw and rotor forces as input and depicts the result of the mathematical model. It was stabilizing itself after takeoff, initially showing some small vibrations which can also be visualized from figure 10. VIII. CONCLUSION Quad-copter is a highly unstable UAV, but due to its high maneuverability, it is highly desired for field works. Utilizing nonlinear dynamic equations accompanied with control theory can bring quad- copter to life. These algorithms sufficiently reduce the need of pilot and can be used to build cheap UAVs which can reduce cost to a great extent. Quad-copter are the future robots in this field in particular as well as in other fields in general, being applied alongside 18 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 28. people which willhelp in making tasks easier and more efficient. Developed mathematical model was successfully implemented on hardware. IX. ACKNOWLEDGMENT This work has been made possible by the help of my co-authors which include my project advisor and co- advisor. This text has been rectified and proof read by Undergraduate students M. Moghees Shahid and Ali Mahmood. They are destined for great things. This project has been sanctioned by College of Aeronautical Engineering, NUST. REFERENCES [1] T. Luukkonen, "Modelling and control of quad-copter," Independent research project in applied mathematics, Espoo, 2011. [2] M. W. Mueller and R. D'Andrea, "Stability and control of a quadrocopter despite the complete loss of one, two, or three propellers," in Robotics and Automation (ICRA), 2014 IEEE International Conference on, 2014, pp. 45-52. [3] I. Gaponov and A. Razinkova, "Quad-copter design and implementation as a multidisciplinary engineering course," in Teaching, Assessment and Learning for Engineering (TALE), 2012 IEEE International Conference on, 2012, pp. H2B- 16-H2B-19. [4] D. Hanafi, M. Qetkeaw, R. Ghazali, M. Than, W. Utomo and R. Omar, 'Simple GUI Wireless Controller of Quad-copter', International Journal of Communications, Network and System Sciences, vol. 06, no. 01, pp. 52-59, 2013 [5] Y. Cooper, R. Ganesh Ram, V. Kalaichelvi and V. Bhatia, 'Stabilization and Control of an Autonomous Quad-copter', AMM, vol. 666, pp. 161-165, 2014. 19 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 29. Detection of FireHotspots dealt by Emergency First Responders in Rawalpindi using GIS applications Amna Butt and Sheikh Saeed Ahmad Department of Environmental Sciences Fatima Jinnah Women University Rawalpindi, Pakistan ambutt91@yahoo.com; drsaeed@fjwu.edu.pk Abstract— During the past, need of efficient Emergency First Response (EFR) for Rawalpindi, along with all the major cities of Pakistan has increased tremendously. Therefore, there is a need to develop an effective management strategy to improve the first response services. Present study focused on the identification of past and current service locations for fire incidences and mapping these locations for hotspot identification. The incidence data for past five years (2009-13) was collected and Hotspot and Spatial Autocorrelation analyses were performed on the data to detect the fire hotspots and their clustering patterns in the city. The results revealed a slight shift in fire hotspots in 5 years and also in the clustering pattern which changed from significantly clustered (2009) to randomly distributed (2013). Hotspot and spatial distribution maps were generated to indicate the fire hotspots in the city. These maps can be helpful to prevent the future incidents by allocating more service stations focusing these areas for fire mitigation. Index Terms— Hotspot Analysis, Fire, Kernel Density, Emergency First Response (EFR), Spatial Autocorrelation Analysis I. INTRODUCTION Coping with fire, caused either by natural or anthropogenic factors is one of the challenges faced by the modern societies [1]. Analyzing the city fire risk is therefore highly significant for development of effective urban fire protection plan and regulations and facilitates the coordinated development of social economy [2]. Application of geostatistical tools of GIS can play a significant role in improvement of local fire emergency response [3] primarily by facilitating the visualization and interpretation of nature and previously observed patterns of such accidents [4][5]. Generating different fire risk maps on the basis of geostatistical analysis is also imperative to develop strategies focused on alleviating the future risk [6]. Numerous approaches based on GIS have been developed and used over the past to provide geostatistical surveillance of the precedent emergency patterns for development of several models for fast and apt response delivery [7][8][9][10][11]. A. Study Area The study area of the present research was Rawalpindi city (Fig. 1). The city’s administrative boundaries consist of two tehsils namely Rawal and Potohar tehsil. Currently five service stations and two key points of Rescue 1122 (otherwise known as Punjab emergency service) are providing emergency response services (including fire brigade service) in different areas of the city. The resources currently available to them for providing Fire brigade service include 9 fire vehicles, 14 ambulances and personnel of 400 trained rescue providers. However, no prior study has been conducted in the city focused on surveillance of fire emergency response. Furthermore, the existing management strategy for improvement of fire emergency response is not very effective and no thought has been given to incorporating GIS expertise in the department for this purpose. Fig. 1. Study Area map: Rawalpindi City 20 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 30. Therefore, the primaryobjective of the present research is to provide a GIS based surveillance system for the Fire incidents of the past in order to determine the recurrent service locations for focusing the future response. The study can therefore be considered as a baseline for future improvement in the quality and efficiency of fire emergency responses in Rawalpindi city. II. MATERIALS AND METHODS The research methodology that was applied to obtain the results primarily consisted of four main steps incorporating data collection, processing, analyzing and visualizing the results (Fig.1). Fig. 2. Flowchart of main steps followed in methodology B. Data Collection and Processing The data collected for the purpose of this study was divided into two categories namely Primary and Secondary data. The main data obtained for the purpose of the study was secondary data acquired from the headquarters of Rescue 1122, Rawalpindi in form of caller and victim directory. This data was collected on unit level i.e. data from all the emergency units of 1122 at work in Rawalpindi city was acquired, compiled and then processed for segregation of Fire cases. Primary data was then collected on the basis of segregated fire incidents. This data comprised the GPS readings of incident locations obtained via handheld Oregon 650 GPS for the reported fire cases. Both primary and secondary data was processed in Microsoft excel and then loaded to ArcGIS 10.2 for further processing and analyzing. C. Geostatistical Analysis of data The data was geostatistically analyzed in ArcGIS 10.2 environment for determination of spatial clustering and identification of hotspots. The geostatistical analysis performed for this purpose included Global Moran’s I test (Spatial autocorrelation analysis) and Hotspot Analysis (Getis-Ord Gi*). 1) Global Moran’s I statistics Global Moran’s I statistic gives an indication of any existing correlation among spatial observations and delineates the characteristics of the global pattern. The pattern maybe random, dispersed or clustered depending on the spatial association present in the data [12][13]. For the purpose of present study spatial autocorrelation among the fire incidents was calculated on yearly basis by employing different threshold limits ranging from 500-2000 meters. The range used for determination of correlation was -1 to 1 and Z-score value was calculated to assess the statistical significance of the observed clustering (based on correlation) for each year. The highest correlation values were then recorded for each year and subsequently were employed for hotspot identification. 2) Hotspot Analysis: Getis-Ord Gi* Fire hotspots were identified based on the Getis-Ord Gi* statistics. For this purpose, the conceptualization of spatial relationship among different datasets was done by opting “Zone of indifference”. The threshold limit was set on the basis of spatial autocorrelation outcomes for each year (exhibiting highest Z-score value). Thereafter, the identified hotspots for Fire cases were interpolated using “Inverse Distance Weighted” or “IDW” for better visualization of results and hotspot maps for each year were generated. III. RESULTS The emergency callout data on building fires, bursting of gas pipes, cylinder blasts, and gas leakages was cataloged in the category of Fire emergencies. Different geostatistical analyses were then performed to determine the pattern of emergency cases for the study duration. The reported incidence of FE cases for the time period of 2009-2013 were 671. Out of which, 583 (86.9%) were males and 88 (13.1%) females. 15% of the total fire incidents (102 cases) were reported in 2009 and 37% (247 cases), the highest incidence, in 2010. After 2010 the incidence rate declined progressively from 24% (163 cases) in 2011 to 10% (66 cases) in 2012 and subsequently rose to 14% (93 cases) in 2013 (Fig. 3). Fig. 3. Percentage Contribution of Fire Incidents in Rawalpindi during 2009- 2013 The possible spatial autocorrelation of Fire cases estimated by Moran’s I statistics revealed significant spatial clustering for the years 2009 and 2011, whereas 2010 showed mild clustering. 2012 and 2013 on the other hand showed random patterns. Moran’s I and G-statistic values (Z-score) for all the years, given in Table 1 disclosed that 2009 had highest (20.01) while 2013 had lowest (1.10) Z-score. 21 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 31. Table 1. Globalspatial autocorrelation statistics of Fire emergencies for 2009- 2013 Year Moran's I Z Score P Value Pattern 2009 0.77574 20.0117 0.001000 Clustered 2010 0.01719 2.09997 0.035732 Clustered 2011 0.17427 4.3162 0.000016 Clustered 2012 0.39961 1.51482 0.129818 Random 2013 0.02737 1.10479 0.269251 Random Local Gi* (d) statistic employed to identify the hotspots for Fire incidents in Rawalpindi during 2009-2013 categorized the Z-score outcomes at 5% significance level as either clusters or non-clusters. The identified Fire hotspots covered both urban and rural areas of Rawalpindi city for all the years. The specific hotspot locations for each year are tabulated in Table 2 however. Table 2. Identified Fire Hotspots for the years 2009-2013 Year Identified Hotspot Locations 2009 Asghar Mall Chowk, Banni Chowk, Chaklala Scheme 3, Chandni Chowk, Islamabad Highway, Khanna Pull, Link Road, Murree Road, Muslim Town, Naz Cinema Chowk, Rehmanabad, Sadiqabad Chowk, Transformer Chowk and Waris Khan Stop 2010 A.R.I.D University road, Band Khanna Road, Bilal Hospital road, Dhok Kashmiriyan, Double Road, Faizabad, Ghosia Chowk, Iqbal Town, IJP road, Kattariyan, Khayaban-e-Sirsyed, Kurri road, Rabi Center, Saidpur road Satellite Town, Sixth road, Shamsabad and Sohan Pull 2011 Dhok Mustakeem, Choor Chowk, Golra Morr, Misriyal road, Peerwadhai Morr, Qasimabad, Seham road and Westridge 2012 Askari 11, Faisal Colony, Jhanda Chichi, Military Hospital Road, Peshawar Morr, Pindora Chungi and Shamsabad Stop 2013 Adyala road, Chaklala Scheme 3, Committee Chowk, Khanna Pull, Link road, Raja Bazar, and Rawal road Figure 4 revealed that Fire hotspots for 2009 and 2010 were mostly contained in the Northern region of Rawal Tehsil and shifted towards North-West in 2011. However, during 2012 and 2013 not only the number of hotspots reduced significantly but the spatial distribution pattern also became random. Fig. 4. Mapping of Fire hotspots using Getis-Ord G* statistics during 2009- 2013 Spatial distribution of Fire cases in hotspot locations was also visualized by creating spatial distribution maps (Fig. 5). The highest incidence (based on the number of cases per locale) was observed in 2010 while the lowest was observed in 2012. The distribution pattern revealed that the highest incidence was mainly in urban areas of Rawal tehsil and North- East portion of Potohar tehsil where the reported cases per locality per year were as high as 10. 22 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 32. Fig. 5. SpatialDistribution pattern of Fire incidents in Rawalpindi for study duration IV. DISCUSSION In order to facilitate the efficient management of fire emergencies in an area, improvement of existing response systems is of high significance [14]. This can be ensured by the providing surveillance for past occurrences and understanding of recurring patterns. Significant Fire hotspots were manifested in both urban and rural areas of the city and were mainly contained in Northern portion of the study area. These were the areas having high rise buildings, gas stations, commercial areas, suburban areas, highways and residential areas (with heavy load shedding of gas). As these hotspots were estimated not only for household, commercial and secondary fires but vehicular fires as well, various roads were also identified as hotspots. Mostly the Fire emergencies were observed on the roads that are used by heavy vehicles as they are more prone to overturning and catching fire. Corcoran et al. [11] also analyzed the spatial patterns of fire by employing GIS and obtained similar results. Rao [15] also reported similar findings and additionally said that the reason for Fire incidents in the city is exposed and jumbled cable wires made of substandard material. However, the results of the study indicated a significant decline in the intensity of cases for the duration of the study (Fig. 3). This declining incidence gave an account of lack of confidence in general public to refer to firefighting organizations and attempting to solve the matter themselves. However the accounted figures do not represent the total number of cases observed in the city, just the incidence that was dealt by Rescue 1122. Other organizations at work in the city for fire control and management include Rawalpindi Fire Brigade Center and Qureshi Fire Control Services. Akhter [16] explicated the reasons behind the lack of confidence among public. The study reasoned that there is disparity in implementation of fire safety standards in the city as well as very little coordination among different departments such as traffic, police and fire fighting units. This lack of coordination along with unavailability of Incident Command System (ICS) often translates to poor emergency response despite good skills and training. Dawn [17] conversely reported that local fire brigade services lack in performance due to insufficient professional training, availability of resources, planning and research (both pre and post fire) and nonexistence of any fire services act for the city. All these factors, along with lack of awareness among general public regarding fire fighting profession has a negative impact on fire emergency response. Hence, for improvement of future fire emergency response, it is need of hour to understand the previous and existing patterns, risk factors and causative agents; and ensure effective enforcement of building and fire safety laws [18]. V. CONCLUSION AND RECOMMENDATIONS Present study focused on providing a geostatistical surveillance approach for ensuring future fire safety by improving the response quality and apt resource allocation for high risk areas. Based on the outcomes of the research, it is concluded that there is both spatial and temporal variation in the occurrence of Fire incidents in the study area. Most of the Fire hotspots however were located in the Northern portion of the study area incorporating both urban and rural areas of the study indicating the need to shift the focus of fire service in this region of the study area. The study further concludes that there is a need of incorporation of GIS based surveillance system in the rescue department to direct the response from the service stations in a timely manner. Therefore, the study recommends generating awareness among people regarding fire hazard and the factors associated with it, incorporation of GIS expertise in emergency departments, promotion of GPS enabled cell phones in dispatch units and fire vehicles, high level of collaboration among different departments working in the city for fire services provision to avoid service duplication and publication of GIS based maps and models designed for response improvement. ACKNOWLEDGMENT We are endepted to the department of Rescue 1122, Rawalpindi for providing the data regarding fire emergencies and cooperating with us throughout this research. REFERENCES [1] M. I. Channa, and K. M. Ahmed, “Emergency Response Communications and Associated Security Challenges,” Int. J. Net. Sec. and its Appli., vol. 2 (4), pp. 179, 2012. 23 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 33. [2] W. Aiyou,S. Shiliang, L. Runqiu, T. Deming, and T. Xiafang, “City Fire Risk Analysis based on Coupling Fault Tree Method and Triangle Fuzzy Theory,” Proc. Engg., vol. 84, pp. 204-212, 2014. [3] Environmental Systems Research Institute (ESRI), “Improving Emergency Planning and Response with Geographic Information Systems,” Redlands, New York: ESRI. Retrieved from http://www.esri.com/library/whitepapers/pdfs/emergency- planning-response.pdf, 2005. [4] S. Erdogan, I. Yilmaz, T. Baybura, and M. Gullu, “Geographical information systems aided traffic accident system case study: city of Afyonkarahisar,” Accid. Anal. and Prev., vol. 40, pp. 174-181, 2008. [5] M. Kwan, and J. Lee, “Emergency response after 9/11: the potential of real-time 3D GIS for quick emergency response in micro-spatial environments,” Comp., Env. and Urban Sys., vol. 29, pp. 93-113, 2005. [6] C. Yan-yan, L. Dong, and Z. Hui, “Multi-factor Risk Analysis in a Building Fire by Two Step Cluster,” Proc. Engg., vol. 11, pp. 658-665, 2011. [7] M. H. Hussain, M. P. Ward, M. Body, A. Al-Rawahi, A. A. Wadir, S. Al-Habsi, M. Saqib, M. S. Ahmed, and M. G. Almaawali, “Spatio-temporal pattern of sylvatic rabies in the Sultanate of Oman, 2006–2010,” Prev. Vet. Med., vol. 110, pp. 281-289, 2013. [8] T. Ruya, M. Ning, L. Qianqian, and L. Yijun, “The Evolution and Application of Network Analysis Methods,” IEEE Int. Conf. on Sys., Man, and Cyber., pp.2197-2201, 2013, DOI 10.1109/SMC.2013.376. [9] D. Dai, “Identifying clusters and risk factors of injuries in pedestrian–vehicle crashes in a GIS environment,” J. Trans. Geo., vol. 24, pp. 206-214, 2012. [10] A. Spoerri, M. Egger, and E.V. Elm, “Mortality from road traffic accidents in Switzerland: Longitudinal and spatial analyses,” Accid. Anal. and Prev., vol. 43, pp. 40-48, 2011. [11] J. Corcoran, G. Higgs, C. Brunsdon, A. Ware, and P. Norman, “The use of spatial analytical techniques to explore patterns of fire incidence: A South Wales case study,” Comp., Env. and Urban Sys., vol. 31, pp. 623-647, 2007. [12] B.N. Boots, and A. Getis, Point Pattern Analysis Newbury Park. Newbury Park, CA, USA: Sage Publications, 1998. [13] L. Fang, L. Yan, S. Liang, S. J. D. Vlas, D. Feng, X. Han, W. Zhao, B. Xu, L. Bian, H. Yang, P. Gong, J. H. Richardus, and W. Cao, “Spatial analysis of hemorrhagic fever with renal syndrome in China,” BMC Infect. Dis., vol. 6, pp. 77-88, 2006. [14] S.R. Morgan, A. M. Chang, M. Alqatari, and J. M. Pines, “Non–Emergency Department Interventions to Reduce ED Utilization: A Systematic Review,” Acad. Emer. Med., vol. 20, pp. 969-985, 2013. [15] S. Rao, “Rescue 1122 management plan finalized,” The Nation. Retrieved from http://nation.com.pk/karachi/06-Jan- 2010/Rescue-1122-management-plan-finalised, 2010. [16] S. Akhter, “Firefighters’ view on Improving Fire Emergency Response: A Case Study of Rawalpindi,” Int. J. Hum. and Soc. Sci., vol. 4(1), pp. 143-149, 2014. [17] Dawn, “Pindi fire brigade squad runs out of steam,” Dawn. Retrieved from http://www.dawn.com/news/90148/rawalpindi- pindi-fire-brigade-squad-runs-out-of-steam, 2003. [18] Pakistan Observer, “1122 Rawalpindi Rescued 1883 Emergency Victims,” Pakistan Observer. Retrieved from http://pakobserver.net/detailnews.asp?id=251475, 2014. 24 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 34. Prospects of AirborneWind Energy Systems in Pakistan Z. H. Khan Dept. of Electrical Engineering Riphah International University Islamabad, Pakistan zeeshan.hameed@riphah.edu.pk Sohaib Khan Dept. of Aerospace Engineering Institute of Space Technology Islamabad, Pakistan sohaibkham786@gmail.com Arsalan Khan Dept. of Control and Simulation CESAT Islamabad, Pakistan arsalan1@mail.nwpu.edu.cn Abstract— High altitude wind energy is considered as a high efficiency, low cost solution to sustainable energy solutions worldwide today due to highly flexible and adaptive designs of powered kites. Pakistan is facing energy crisis since years due to inefficient electrical transmission network, increased demand of electricity, lack of resource planning and increased cost for furnace oil which can be used to generate electricity. As a clean energy source, conventional wind energy is a preferable choice but it has few disadvantages due to large initial investments involved and dangers to environment due to rotating machinery which can result in bird hit and on the other hand, generating acoustic noise which affect populations of people living nearby. In response to these issues, high altitude energy is found to be free from many such problems as found with conventional wind energy systems. It is found suitable specially for addressing essential energy needs of off-grid consumers for water pumping, drilling, refrigeration of vaccines and life-saving medicines and powering up far-off residential sites e.g. communities living off- grid at Alaska’s northern region having minimal solar energy during the long winter season when energy demand is greatest for heating purposes. In this paper, we propose a UAV design which can possibly be used as an airborne wind energy system for electricity generation. Keywords—Altitude wind; renewable energy; powered kites; wind energy conversion system; distributed energy I. INTRODUCTION Energy production for a world’s economy is directly linked with GDP growth. Renewable energy solutions are a preferable choice to address the energy demands world-wide due to environment friendly green energy technology. It has been estimated that till 2050, 40% of the world energy requirements will be meet by renewable energy including solar (photo voltaic (PV) and thermal), wind, bio-mass, bio-fuel etc. [1]. Among various technologies, wind energy is preferred due to continuous production of electricity (as long as wind is present as prime mover) as well as high production rate with lesser requirement of installation area as compared to solar energy [2]. However, a detailed analysis of wind availability and on- site surveying is mandatory for an optimal production of energy throughout the year. The conventional wind energy conversion systems (WECS) are popular only in those areas where sufficient wind is available i.e. in the range of 5-8 m/s2 . This limitation greatly affects the utilization of WECS as an effective resource of energy as usually far off areas near sea- shores or mountains fulfills the required energy density criteria for feasible investments. Also, long way cabling for transmission of this energy to grid also results in additional cost and line losses [3]. Social activists and NGOs raise voices and show their grievances against environmental impact of these wind farms due to ever increasing bird accidents as of result of collision with fast moving blade tips. Fig 1. Airborne energy production connected to grid This paper describes some key design features of an airborne wind energy system (AWES) e.g. a large kite which is flown in air while maintaining its connection with a generator kept at ground level by a suitable rotating mechanism such as cable drum which is linked via tether. The aerodynamic forces acting on the kite causes traction in the cable drum which is in turn converted into electrical power by the generator as shown in Fig 1. The obvious advantages of shifting the heavy mechanical components on ground result in simple design and maximum power optimization. Due to the fact that the flying kite operates in periodic cycles alternating between ―traction phase‖ and ―reel-in phase‖ of the tether, the electrical power produced is not continuous rather intermittent which is not 25 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 35. suitable for on-gridconnection where continuous power is required. However, a continuous power generation system can be designed either by using multiple, individually controlled kites or a battery system for buffering the power generation across the cycles. This solution is guaranteed to generate higher power with lower cost as compared to conventional wind energy conversion system (WECS) due to availability of faster wind speeds at high altitude. Fig 2. KGS for generating electricity on-board ship II. HIGH ALTITUDE WIND ENERGY POTENTIAL Wind energy is a renewable resource which is not only cheap but also readily available in most parts of the world. Wind is discontinuous source of energy and the conventional WECS provides uneven power supply when connected to the grid. On the other hand, if altitude is raised, a lot more energy is available as compared to that blowing at 50-200 m above ground. It has been estimated that the magnitude of wind energy above 1000 m is twice as much as found at lower altitudes. In addition, at an altitude of 800 m above ground, wind power density is sufficiently available to be used for altitude energy generation all over the globe [4]. As a rule of thumb, it has been found that a five-fold cost saving with twice capacity increase can be achieved in shifting technology from WECS to AWES [5]. III. ADVANTAGES OF AWES OVER WECS A. Impact on Environment The airborne wind energy systems have fewer effects on environment as compared to WECS. The fast moving blades of classical wind mills injures many birds flying in close proximity. On the other hand, AWES has no dangerous edges which can kill birds [6]. B. Cost comparison Using AWES instead of WECS, an advantage of more than 90% material savings is guaranteed. This is because large structures require more and more material to with stand heavy loads due to wind and the rotating machinery [7]. Many tones of weight loaded on large erected structures comprises of rotor blades connected to a hub which drives the generator. Moreover, the maintenance of WECS is also an issue which requires access to the system, long down times in case of fault and danger to the life of technician all adding up the cost compared to AWES where generator and accessories are installed on ground [8]. C. Mobility of the power station In comparison with WECS, the airborne energy converter can be moved anywhere. Example include rural areas, Coastal belt, Desert, etc. However, if flying at higher altitudes, caution must be paid to operate AWES in no-fly zones (NFZ) to avoid any danger for civil aviation. D. Efficiency The advantage of increasing altitude as in the case of AWES is to get more capacity due to persistent wind as compared to turbulence due to buildings and infrastructure at WECS heights [9]. Fig 3. Buoyant Airborne Turbine (BAT) for stand-alone energy generation in Alaska [10] IV. CLASSIFICATION OF AWES The airborne wind energy systems are classified based on their designs, grid connectivity and aerodynamics e.g. heavier than or lighter than air configuration. Some types are indicated as follows: 1) Flying tethered airplane and kite In this type of AWES, a tethered kite or airplane flies in circular or Lissajous shaped orbit to produce electricity via traction force applied on a generator on ground or ship as shown in Fig 2. The kite generator system (KGS) is one of the most commonly used system due to its simplest design and easy control system [11]. 2) Multiple wing system This type of altitude wind system has multiple wings to generate electrical energy. A laddermill is an example of such system [12]. Multiple wings allow scalability of the concept in order to generate more energy. 3) Lighter than air systems These systems are lighter than air and have an on-board generator to produce electricity as shown in Fig 3. Such systems are supported by a lighter than air filled balloon to reach at 100m or above heights where enough wind is available to drive the generator [13]. Such systems are more suitable for areas of the world where sunlight is not available throughout the year. 26 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 36. 4) Rotor hovercraft systems These systems hover in air via rotating propellers. Quadrotor type hovercrafts are popular which have the ease of vertical take-off and landing operations [14]. Due to on-board generators, the airborne system is heavier as compared to the crosswind kite generator in which has the heavy machinery on ground. V. FLYING WING GENERATOR SYSTEM DESIGN In order to visualize a practical system, we can use a flying wing generator (FWG) for our case study. Our goal is to evaluate optimal design parameters which can be used to generate maximum wind at different altitudes. By using a strong flexible tether, energy kites can reach higher altitudes of 100 to 400 meters). This can save up to 90% of the materials of WECS used conventionally, resulting in reduced per unit energy cost. As, these systems are more aerodynamic and can access higher energy density due to stronger winds, each EKS can generate 50% more energy than their conventional counterparts [15]. A. Aerodynamics The wind energy that can be converted to electricity is given as: 2 3 27 2 D L Lw C C CAvP (1) The relation provides some significant information about some key design parameters. Equation 1 states that in order to increase the power, the key parameters to increase include the wind speed, gliding ration (L/D), the lift coefficient and the wing surface. As a comparison, energy kites have lower L/D, price and weight as compared to flying wing. In theory, about 60 kW can be produced per sq. meter of flying kite generator system. The CFD model is shown in Fig 4. (a) Side View (b) Top View (c) Front View Fig 4. FlyPG simulation for aerodynamic analysis In order to analyze the aerodynamic performance of our benchmark system FlyPG, different flight conditions are used for Aerodata generation. The aircraft wing is taken as NACA- W-4-4410 which has a span of 4 meters, Sref equal to 1.5 m2 , Cbar = 0.38 m, Root chord = 0.35 m, Tip chord = 0.15 m. For the vertical tail design, NACA-V-4-0010 is selected with Root chord = 0.25m and Tip Chord = 0.1 m. The horizontal tail design is NACA-H-4-0010 based with root chord = 0.25 m and Tip chord = 0.1 m. The flight test condition is selected as Mach 0.05 at an altitude of 200 meters above ground level. Then the angle of attack (α) in degrees is varied from 0 to 5 degrees, while Xcg of 0.9 meter is assumed. The aero-data is plotted in Fig 5 as a function of angle of attack. It is important to take into account at least 10% drag increase due to tether connecting the kite with the ground generator. Fig 5. Aerodata plots for FlyPG @ Mach = 0.05 B. Control and Autopilot Modern design of FWGs incorporates automatic take off/landing and flight for robust operation under varying environmental conditions [16]. A coordinated control mechanism is also devised by few manufacturers where communication between the Kite/airplane controller and main controller at the ground station occurs to track the given trajectory [17]. Generally, a multi-objective loop controls the dynamics of the system in flight. First of all, an optimal track point on circular or Lissajous trajectory needs to be calculated and tracked during flight [18]. In most cases a navigation loop forms the outer one which controls the bearing of the flying system, while an inner attitude controller controls the roll, pitch and yaw angle as well as respective rates in order to achieve the required bearing. Fig 6. A generalized 2-loop control structure for Flying wing/ Energy Kite As shown in Fig 6, an outer loop controls the bearing (ρcom) of the flying wing. The error between the commanded and measured bearing drives the attitude controller in order to generate roll, pitch and yaw commands for the flying wing until it reaches the desired navigational coordinates. In this 27 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 37. control loop, however,the cross track controller is not included which is necessary to keep the flying wing on the desired trajectory by removing the off-track distance [19]. In practice, stable and robust control is required for fast tracking. C. Mechanical Subsystem The mechanical part of a FWG includes wing design after aero-foil selection. The control actuators (motors) also need positioning and the embedded controller is designed w.r.t nominal torque load as well as keeping its maximum limit as per maximum torque demand. The selection of tether and its linkage with the ground generator is also important [20]. However, the drag added by the tether length must be added in the aerodynamic drag analysis for the autonomous controller design [21]. D. Electrical Subsystem The power generation system usually comprises of a permanent magnet synchronous machine (PMSM) which acts as a generator [22]. In general, the AWECS can function either as stand-alone (off-grid) or on-grid. However, an obvious problem for grid based operation requires an additional cost of storage system or operating multiple systems at the same time so that the pumping cycle of one kite is out of phase as compared to the pumping cycle of the other kites [23]. Thus, at the instant one kite is relaxing; the other will be supplying energy to the grid. However, this strategy complicates the design of the AWECS. E. Communication Subsystem The communication system is required in order to control the system and for data acquisition. Wireless networks are a preferred choice in order to reduce weight and cost of the airborne system [24, 25]. A suitable network protocol is chosen as required depending upon the range of communication required. Ground based sensors and controllers send necessary data to the air segment. Security features of the communication protocol must be operational for robust and uninterrupted control of the flying system [26]. 2000 2003 2006 2009 2012 2015 2018 0 50 100 150 200 250 300 350 400 450 500 ElectricalEnergy(MW) Time Fig 7. WECS for energy generation and forecast in Pakistan VI. ENERGY NEEDS IN PAKISTAN Pakistan is the 6th largest country of the world with a total estimated population of 175 million people. However, 24% population lives below the poverty line in the country. As per the economic survey, it is among those countries of the world which have lowest per capita energy consumption. The installed electricity generation capacity is around 22,500 MW. While the current available power is 17,000 MW, the peak demand raises up to 22,000 MW in summer resulting in a shortfall of 5000 MW. The economic advisory council of Ministry of Finance (Govt of Pakistan) foresees an increase in energy demand up to 122.46 M.MTOE in 2022 [27]. In order to satisfy this demand, at least 12% of this energy is planned to be obtained from renewable resources in order to relax the huge demand of furnace oil exports from 30% at present to 20% in 2022 [28]. Table 1 describes the energy demand and forecast which shows that a total of 122.46 M.MTOE demand is expected in 2022 out of which 14.70 M.MTOE will be obtained from renewable energy. Looking at the rise in using wind energy as the renewable source in Pakistan as shown in Fig. 7, it is emphasized that Altitude wind projects should also be commenced in the country as a future energy resource especially for people living in off-grid areas of Pakistan. In order to enhance sustainable energy trend, Government has exempted renewable energy equipment from income tax/withholding tax and sales tax [29]. Also, import duties have been waived off for import of such machinery and tools to be used to develop sustainable energy solutions. Thus, already a favorable environment is available for foreign investors to come and finance green energy solutions. Fig. 7 shows a potential rise of WECS usage as a renewable energy source. However, the feasibility is only limited to areas where sufficient wind is present. These areas include the coastal belt near Karachi (Sindh province), Kati-Bander, Makran and Jhampir in Balochistan [30, 31]. Figure 8 Map of Pakistan's national grid [32] As noted from the title of the paper, our target is to address the energy demands of remote areas of the country far-off from 28 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 38. the national grid.Fig. 8 describes the map of Pakistan’s national grid which shows large areas of Balochistan not covered by it. This is because of the low population density in this province as well as poverty due to which it is not feasible to extend the grid in these areas. Villages in these areas can be powered up by renewable energy resources by spending less money and requiring only maintenance cost in the long run [33]. TABLE I. PAKISTAN ENERGY DEMAND FORECAST [27] Resource Energy Demand (M.MTOE) 1992 2008 2022 Oil 11.53 19.18 24.49 Gas Indigenous 11.68 29.88 25.72 Gas Import - - 8.57 Hydel 4.49 6.86 24.49 Coal 2.10 5.79 18.37 LPG 0.15 0.44 2.45 Nuclear - 0.75 3.67 Renewables - - 14.70 Total 29.94 62.90 122.46 Most common uses there require lightening solution, water pumping for drinkable water supply and also to operate tube- wells for agriculture use etc. These applications require lightweight, low-cost, minimum energy storage solutions. These demands can be fulfilled by using AWES whereby the intermittent energy produced during reel out phase could be used directly. VII. CONCLUSION In this paper, an overview of altitude wind technology is discussed and a preliminary analysis is done for the suitability to address the needs of renewable energy in developing countries with budget constraints. A flying wing generator design is also presented with some initial CFD simulation results. As the model of FlyPG is now generated, the next step is to design a prototype system and to demonstrate the energy output by flying it. Moreover, as the energy needs in the country are growing, it is high time to invest in such aerospace technologies to obtain optimal designs resulting in maximum electrical output with light weight flying systems. REFERENCES [1] M. A. Sheikh, "Energy and renewable energy scenario of Pakistan," Renewable and Sustainable Energy Reviews, vol. 14, pp. 354-363, 2010. [2] J. I. Lewis and R. H. Wiser, "Fostering a renewable energy technology industry: An international comparison of wind industry policy support mechanisms," Energy Policy, vol. 35, pp. 1844-1857, 2007. [3] G. Kingsley, "Wind energy production using kites and ground mounted power generators," ed: Google Patents, 2003. [4] M. lppolito, "Altitude Wind Energy Generation (Information Dossier)," KiteGen Research, Italy2007. [5] L. Goldstein, "Theoretical analysis of an airborne wind energy conversion system with a ground generator and fast motion transfer," Energy, vol. 55, pp. 987-995, 2013. [6] M. I. Blanco, "The economics of wind energy," Renewable and Sustainable Energy Reviews, vol. 13, pp. 1372-1382, 2009. [7] B. W. Roberts, "Cost and Security of Electricity Generated by High Altitude Winds," Altitude Energy2012. [8] M. Brooks, "To make the most of wind power, go fly a kite," New Scientist, May, vol. 5, 2008. [9] R. van der Vlugt, J. Peschel, and R. Schmehl, "Design and Experimental Characterization of a Pumping Kite Power System," in Airborne Wind Energy, ed: Springer, 2014, pp. 403-425. [10] C. Vermillion, B. Glass, and A. Rein, "Lighter-Than-Air Wind Energy Systems," in Airborne Wind Energy, ed: Springer, 2013, pp. 501-514. [11] M. Ippolito, "Kite Wind Generator," ed, 2006. [12] B. Lansdorp and P. Williams, "The laddermill-innovative wind energy from high altitudes in holland and australia," in Wind Power, 2006. [13] C. Vermillion, T. Grunnagle, and I. Kolmanovsky, "Modeling and control design for a prototype lighter-than-air wind energy system," in American Control Conference (ACC), 2012, 2012, pp. 5813-5818. [14] G. Mortimer. (2014, 29th October). Multicopter based wind turbines. Available: http://www.suasnews.com/2010/10/2464/multicopter-based- wind-turbines/ [15] MAKANI. (2013, 20th November). Advanced Wind Energy "ENERGY KITES". Available: http://www.google.com/makani/ [16] M. Canale, L. Fagiano, and M. Milanese, "High altitude wind energy generation using controlled power kites," Control Systems Technology, IEEE Transactions on, vol. 18, pp. 279-293, 2010. [17] M. Canale, L. Fagiano, M. Ippolito, and M. Milanese, "Control of tethered airfoils for a new class of wind energy generator," in Decision and Control, 2006 45th IEEE Conference on, 2006, pp. 4020-4026. [18] R. Lozano Jr, M. Alamir, J. Dumon, and A. Hably, "Control of a wind power system based on a tethered wing," Proceedings of second EGNCA 2012, 2012. [19] C. Vermillion, B. Glass, and B. Szalai, "Development and Full-Scale Experimental Validation of a Rapid Prototyping Environment for Plant and Control Design of Airborne Wind Energy Systems," in ASME 2014 Dynamic Systems and Control Conference, 2014, pp. V002T18A001- V002T18A001. [20] R. Marissen, "Dyneema high strength, high modulus polyethylene fiber," vol. CIS YA100, Dyneema, Ed., ed: Royal DSM N.V., 2008, p. 4. [21] L. Fagiano, "Control of tethered airfoils for high–altitude wind energy generation," Politecnico di Torino, 2009. [22] M. Ahmed, A. Hably, and S. Bacha, "Grid-connected kite generator system: Electrical variables control with mppt," in IECON 2011-37th Annual Conference on IEEE Industrial Electronics Society, 2011, pp. 3152-3157. [23] P. Williams, B. Lansdorp, and W. Ockesl, "Optimal crosswind towing and power generation with tethered kites," Journal of guidance, control, and dynamics, vol. 31, pp. 81-93, 2008. [24] Z. H. Khan, D. G. Catalot, and J.-M. Thiriet, "Hierarchical wireless network architecture for distributed applications," in Wireless and Mobile Communications, 2009. ICWMC'09. Fifth International Conference on, 2009, pp. 70-75. [25] Z. H. Khan, J. M. Thiriet, and D. Genon-Catalot, "Wireless network architecture for diagnosis and monitoring applications," in Consumer Communications and Networking Conference, 2009. CCNC 2009. 6th IEEE, 2009, pp. 1-2. [26] J. Wang, M. Shahidehpour, and Z. Li, "Security-constrained unit commitment with volatile wind power generation," Power Systems, IEEE Transactions on, vol. 23, pp. 1319-1327, 2008. [27] E. A. Council, "Integrated Energy Plan (2009-2022), Report of the Energy Expert Group " Ministry of Finance2009. [28] M. A. Khan and U. Ahmad, "Energy demand in Pakistan: a disaggregate analysis," The Pakistan Development Review, pp. 437-455, 2008. 29 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 39. [29] A. W.Bhutto and S. Karim, "Energy-poverty alleviation in Pakistan through use of indigenous energy resources," Energy for Sustainable Development, vol. 11, pp. 58-67, 2007. [30] M. Hussain, S. Abbas, M. Ansari, A. Zaffar, and B. Jan, "Wind Speed Analysis of Some Coastal Areas near Karachi," Pakistan Academy of Sciences, 2012. [31] I. Ullah, Q.-u.-Z. Chaudhry, and A. J. Chipperfield, "An evaluation of wind energy potential at Kati Bandar, Pakistan," Renewable and Sustainable Energy Reviews, vol. 14, pp. 856-861, 2010. [32] GENI. (2012). Global Energy Network Institute. Available: http://www.geni.org/globalenergy/library/national_energy_grid/pakistan /pakistaninationalelectricitygrid.shtml [33] M. Amer and T. U. Daim, "Selection of renewable energy technologies for a developing county: a case of Pakistan," Energy for Sustainable Development, vol. 15, pp. 420-435, 2011. 30 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 40. INTEGRATED USE OFPOTENTIAL RAINWATER HARVESTING SITE FOR AGRICULTURE USING GEO- SPATIAL APPROACH Niaz Ahmad1, M.R Khan2, M.Amin3 , M.Usman4 , Abida Perveen5 1 Professor Dept. of Land and Water Conservation, PMAS Arid Agriculture University Rawalpindi 2 Associate Professor, Dept. of Geo-Informatics, PMAS Arid Agriculture University Rawalpindi 3 Lecturer Dept. of Geo-Informatics, PMAS Arid Agriculture University Rawalpindi 4 Lecturer Dept. of Land and Water Conservation, PMAS Arid Agriculture University Rawalpindi 5 GIS/Hydrology Consultant AAB pvt ltd Islamabad Abstract—In rain fed areas, sustainability of agricultural system is critically dependent uponsustainable agricultural water management and its ensured accessibility. Rainwater harvesting has a greatpotential for storing water for the rising demands for agricultural purpose. The present research utilizes Remote Sensing and GIS applicationsto identify optimal sites for efficient storage of rainwater for agricultural purposes in the hilly areas. The essential problem is the optimal conservation of flows and management of rain water. We present a valuable methodto improve the rainwater availability as irrigated water supply to minimize water deficiency. The main objectives of the study are: firstly to demarcate the potential catchments area and then rain-water storage by proposing reservoir with GIS based estimation of storage capacity. Aspatial data and non-spatial data is also generated to plan various agricultural activities for better utilization of water resources for agricultural purposes. Spatial database consists of administrative boundaries, road network, land use/land cover, Digital Elevation Model (DEM), settlements points, stream network, catchments area, proposed water reservoir site, reservoir area and reservoir capacity. The output of this study can be used forincreasing the agricultural production due to efficient utilization of the rainwater. Keywords—Remote Sensing, GIS, Catchment Area, Rainwater Harvesting, Spatial Database, DEM,GPS I. INTRODUCTION Water is a limiting factor in Pakistan and land is not being used according to its potential. It needs no explanation that there is acute shortage of water both for drinking and irrigation purposes. Pakistan is characterized as arid with 80 percent of the land being arid or semi-arid (Shah, Khan et al. 2012). Due to steep slopes and impervious nature of geological formations, a considerable portion of the runoff goes waste without being utilized for irrigation and drinking purposes. In addition, the erosion of top fertile soil has considerably added to the increased rate of sedimentation in the Indus River and Arabian Sea. Indus river water quality is at severe risk from soil erosion. Indus river is one of the highly polluted rivers (Khan and Ali 2003). If the present erosion trend continues unabated the top soil will be completely lost and will result in vanishing forest with subsequent environmental repercussions. Pakistan rain-fed area lies mostly outside Indus Basin where traditional methods of rainwater irrigation are practiced. The rainwater besides traditional methods is regulated through man-made system like mini dams, small dams, recharging groundwater and abstracting it through tube-wells, karezes, wells etc. The Punjab province is blessed with natural land and water resources(Akhtar 2006). These resources need to be developed for providing much needed dependable perennial water supply for irrigated agricultural development of the area. The development of these resources will also help towards achieving primary objectives of socio-economic uplift and poverty alleviation of the people of the area. Also the fact due to rapid urbanization, infiltration of rainwater into the sub-soil has decreased drastically and reaching groundwater has diminished. So, the rainwater harvesting is essential because surface water is inadequate to meet the ever growing demand and dependence has to be on groundwater. Rainwater harvesting is a deliberate collection and storage of rainwater or it can be said to be an augmentation of groundwater reservoir by man-made structures by utilizing rainwater. The objectives and techniques for rainwater harvesting are highly location specific and techniques for rainwater harvesting are highly location specific and an appropriate technology developed for a particular region cannot be used as such for other areas for geographic, environmental, technical and socio-economic reasons(Jasrotia, Majhi et al. 2009). RWH can be a measure to increase access to water for vulnerable sections of the society in arid and semi-arid parts in countries where resources are scarce and inaccessible(Munyao, Mannaerts et al. 2010). The water harvested can be used for various purposes ranging from domestic, livestock, 31 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 41. agricultural production, industrialand groundwater recharge. A successful implementation of RWH should integrate socio-economic and environmental issues to ensure sustainability and protect fragile ecosystems(Vohland and Barry 2009). RWH may lead to increased food production through minimizing the risk of crop failure during droughts and floods. At watershed level anticipated benefits include recharge to groundwater systems and improvement of environment. The rainwater has to be captured and made to infiltrate at the place where it falls and should not be allowed to generate runoff. But, this cannot be done at all places due to variations in soil type, geomorphology and other related parameters. Remote sensing and GIS are very much helpful in identifying areas which are suitable of rainwater harvesting. Different soil types have different capacity for infiltration. Based on their infiltration rates weights are assigned to them. The geomorphology of the region is also important in identifying suitable areas. There are different hydro-geomorphologic features with different groundwater potentials, based on groundwater holding capacity weights are assigned them too. These weights signify the suitability of various forms of the same layer. Soil type and the Hydro geomorphology are two main criteria involved in identifying suitable areas of recharge. Hence they are given equal weight-age. Then suitability indices for each area are calculated. GIS platform is used for overlaying soil and hydro geomorphology maps and hence suitable areas on basis of given weights are identified. The discussed methodology is suitable for regions with less rainfall. This technique would help in avoiding water scarcity during droughts. Most of the decisions on where to locate the RWH systems are spatial. Different layers of important factors will be needed to be combined to determine their suitability. Normally multi-criteria techniques are used in combining different factors. In multi criteria evaluation, suitability values of each factor and their relative importance weights of different factors need to be established. Suitability levels refer to the degree to which a certain value in a given factor influences the location of a rainwater harvesting technology. For example very steep slopes will not be suitable areas for bench terraces compared to gentle slopes. Some RWH techniques are site-specific and sometimes indigenous knowledge exists in those sites, therefore, there is a need for an in depth study at such locations with the aim of establishing a clear relationship between the respective factors. For example, Makanya watershed in Tanzania has relatively high adoption of various types of RWH technologies. Which make it an ideal site for studying the relationship between the various factors and existing technologies. According to Population Census Organization and Planning Commission, with the increasing population of Pakistan estimated to increase by approximately 50% by the year 2025, the water needed for meeting food and fiber requirements will also increase tremendously. To meet with this increased demand water storages has to be increased so that more area can be brought under irrigated agriculture and crop production is increased. This increase in population will put additional substantive pressure on the scarce water resources of Pakistan which are already deficient in meeting demands of agriculture sector. These additional resources will relieve the pressure of water scarcity to some extent for the growing population. By construction of small water reservoir the local population near to the reservoirs will benefit from irrigated agriculture activity. In this study, we present efficient approach to mitigate the rainwater while conserving this floodwater to provide as irrigated water supply to minimize drought threats. For The uniqueness of this study is development of a methodology for estimation of water reservoir capacity by using remote sensing and GIS tools. The objectives of paper are i. to demarcate the potential catchments area ii. rain-water storage by proposing reservoir with GIS based estimation of storage capacity. A secondary output, a spatial catalog is generated for plan to use its spatial data and non-spatial data for future examine projects. Spatial database consist of administrative boundaries, road network, satellite imagery, DEM, GPS points (GCP), settlements points, stream network, catchments area, proposed water reservoir site, reservoir area and reservoir capacity. The output of this study describe that agricultural water execution would approved efficiently with the use of developed methodology. II. MATERIAL AND METHOD Location and Accessibility Study area is located in Punjab region which is bordered on the west of Khyber Pakhtunkhwa and part of Balochistan Province; on the north by Jammu & Kashmir and part of Khyber Pakhtunkhwa; on the east by Indian part of Punjab and on south by Sindh Province. It is globally located between longitudes 690 30’ to 750 10’ E and latitudes 270 42’ to 340 00’ N. Proposed water reservoir is on SirriNullah, which is tributary of AnkarNullah and joins Soan River and ultimately falls into Indus River. It is adjacent to Tamman village in Tehsil Talagang, District Chakwal. It is about 2.5 miles from Tamman village. The potential site is approachable through TalagangTamman road. From Tamman onwards there is 2 miles metalledand one mile 32 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 42. Kacha road. Thecoordinates of proposed water reservoir are given below:- Northing = 32°58’15.41” Easting = 72°08’40.5” Figure 1. Location Map of Study Area Study area at regional scale consist whole Potohar Plateau of 22293 km2 covering four districts Rawalpindi, Chakwal, Attock and Jhelum of Punjab province of Pakistan. Potential site selection was proposed through Digital Elevation Model (DEM) at 3D module of global mapper software application. Demarcation of catchment area was done in Arc GIS spatial analyst tool by using watershed delineation. To find available rainwater harvesting potential, the region was divided into sub-catchments using a digital elevation model through hydrology tools in Arc GIS. This research was essentially based on field survey and mapping details. The importance accumulation of recent research study is that; it is based on both the use of mapping details and geo- information tools. Study area at local scale is proposed on SirriNullah, which is tributary of Soan River in district Chakwal and is approachable through Talagang-Tamman road. From Tamman onwards there is 2 miles metalled and one mile kacha track. SirriNullah carries runoff produced by rainfall. The area is generally characterized by low to medium relief hills. Highest level in catchment area is 2100 ftamsl while level of the stream bed at dam axis is 1133.6 ft-amsl. Dendritic type drainage pattern is present in catchment area and drainage pattern is considered to be structurally controlled. Stream is non-perennial in nature and it is tributary of Ankarnullah, which is tributary of Soan River. Soan River joins Indus River upstream of Kalabagh. Gradient of nullah in the catchment area is gentle. The catchment area of SirriNullah up to the proposed water reservoir is 101.47 sq.km. The proposed command area has about 6000 acre downstream of water reservoir near village Tamman. Most of the command area is compact and plain. The area has a gradual slope starting from the Dam site. The study area comprises of Potohar plateau and lies in semi-arid to sub-humid zone of climatic region with hot summers and cold winters. The average rainfall at Murree in the north-eastern part, where the major torrents originate, is over 59.06 inches whereas that of Tamman in the south- western part of the area is 11.61 inches. About 60% of the annual rainfall occurs during the Monsoon season and about 40% in the remaining period. Generally, the rainfall intensity goes on decreasing gradually from north-east to south-west. The average annual rainfall covering the plateau is estimated at 26.57 inches a year. Rainfall data of Gujar khan, Jhelum, Kalabagh and Mianwali has been adopted for the proposed dam site being nearest rain gauge station with long term record (1980-2010). Daily rainfall data of the following stations existing in the Potohar region is collected from SWHP, WAPDA, Stations, latitude, longitude; period of record with aerial distance from study area is labeled in table 2. Rainwater Harvesting Mechanism: Typically rainwater harvesting mechanism is depicted in rain water accumulate towards low height grounds. This water normally causes flash floods but it completely depends upon the localized terrain. Before allowing it to enter the main stream, it is preferred to store sufficient amount of water at local level for agricultural activity This study assumes that if terrain is mountainous then ideal solution can be acquired through digital analysis.. This paper clearly suggests that, if we have full information of the elevations then different kinds of water reservoir or Dams hence more sustainable agricultural activity. Dataset and specification: The specification of dataset used in the research work is provided in table1. The procedural particulars of operations and purpose of use are discussed under step-wise methodology. Table 1: Dataset specification Data Type Specification GPS data Point data through ground survey Landsat TM 30m Base Map 1:50,000 GeoEye imagery 0.5m resolution Topography Sheet 1:50,000 Aster GDEM Digital Elevation Model 30m 33 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 43. Hydrological Map 1:50,000 Climaticdata Point base Dataset accuracy and quality: As the key procedure is prepared on ASTER GDEM based Digital Elevation Model (DEM), so data authenticity sketch would guide to ensure the accuracy of output results. The AsterGDEM enquire elevation data on a near- global scale to generate the most complete high resolution digital topographic database of earth (Reuter, Nelson et al. 2007)(Rabus, Eineder et al. 2003). Under The base map of the Chakwal and mountainous area of local level study area provides the information to identify the target areas of interest. The map is of large scale i.e. 1:50,000 however for more accuracy assessment was carried by using GeoEye 0.5m resolution image of Tamman, Chakwal. The satellite image of Landsat TM (30 m resolution) was used for land cover classification. The soil map and hydrological map were used to digitize the ground water potential and to identify land productivity classes respectively at local level study area. Figure 2. Major Landcover Map III. RESULTS AND DISCUSSION Preparation of GIS Database and Watershed Delineation Spatial and non- spatial database is very significant for GIS based analysis consequently it is very important to have an accurate geo-database model to utilize for GIS analysis functionality(Minár, Mentlík et al. 2005). For GIS data layer we used high resolution satellite imagery. For demarcation of Catchment area we used hydrology tool in spatial analyst extension. Hydrology tool were used in order as sink-fill- create flow direction- create- - flow accumulation- Create outlet (pour) points- Delineate Watersheds. After delineation of water- shed raster dataset was converted to vector for further calculations. Figure 3. Watersheds of District Chakwal The catchment area of SirriNullah up to the proposed dam is 101.47sqkm as shown in figure 4. The maximum elevation in the catchment area is 2100 ftamsl while elevation in the river bed at proposed dam site is 1133.60 ftamsl. The length of main channel is 17.4 miles and bed slope of Sirri Figure 4. Catchment area of reservoir Nullah is 1.05%.There is no perennial runoff in the nullah. The stream shows a dendritic stream pattern that is favorable for the surface runoff produced from rainfall. Map showing land cover stream meandering at Tamman dam site. Figure 5.land cover at Tamman dam site (Source: Google Earth) 34 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 44. Rainfall Data Rainfall dataof Kalabagh has been adopted for the proposed dam site being nearest rain gauge station with long term record (1987- 2004). Location of climate station is shown in in table 2 and figure 6. Daily rainfall data of the following stations existing in the Potohar region is collected from SWHP, WAPDA. Stations, easting, northing, period of record with aerial distance from project area is tabulated below. Table2 Climatic Stations Location Station Eastin g Northin g EL.(fta msl) Period of Record Aerial Dist. from Project Area (miles) Gujar Khan 330 15 730 18 1476 1962- 2004 68.35 Jhelum 320 56 730 43 755 1961- 2010 92.31 Kalaba gh 320 57 710 33 702 1987- 2004 31 Mianw ali 320 35 710 31 620 1962- 1979 43.5 Details of Rainfall Analysis Mean annual rainfall at Jhelum & Gujar Khan is higher (>32 inches), while Mianwali and Kalabagh are approximately in between 15-19 inches. Rainfall data of Kalabagh have been adopted for the proposed dam site because it is very near (approximately 31 miles) & rainfall in these areas are approximately same as in the project area. Daily rainfall data for Mianwali for the period of 1962-1979, Gujar Khan for the period of 1980- 1986 and Kalabagh for the period of 1987-2004 have been collected from Surface Water Hydrology Project (SWHP), WAPDA. Mean monthly rainfall at Kalabagh being closest to proposed dam site is computed for the period 1987-2004. Rainfall analysis indicated that mean monthly rainfall for project site varies from 0.26inches to 4.54 inches (Figure 7) while mean annual rainfall for corresponding period is 20.32 inches (Figure 8). It clearly indicates that proposed dam site lies between 15-25 inches rainfall contours which are higher than the rainfall contours at the rainfall station (19 inch). The rainfall factor therefore comes to be 22/19 = 1.15. The rainfalls of these stations are therefore adjusted to derive project rainfall with factor of 1.15. Figure 6. Location of climate station on Google map Figure 7. Mean monthly rainfall (inches) 4.55 0.26 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 RAINFALL(INCHES) MONTHS MEAN MONTHLY RAINFALL FOR TAMMAN DAM 35 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 45. Figure 8. Annualrainfalls (1962-2005) (inches) Monthly temperature data for Kalabagh for period 1986-89 and 1992-2004 have been collected from SWHP. Monthly maximum and minimum temperature at Kalabagh are shown in Figure 9. Data indicates that mean maximum temperature varies from 23 to 47 centigrade, while mean minimum temperature varies from 0.3 to 21.5 centigrade. Figure 9.Mean Maximum and Minimum Temperature at Kalabagh(Degree C) Elevation-Area-Capacity Relationship Gross Storage Capacity Gross storage capacity of 8748.80 acre-ft against El.1212.33 ft Above Mean Sea Level (amsl) is fixed. Live storage capacity 7025.80 acre- ft against average annual inflow of 9061 acre-ft which is above 100% limit. Dead Storage Capacity Dead storage is taken as 1723.00 acre-ft against elevation of 1177.82(amsl). Summary of these parameters are given in Table 3. Table 3. Summary of Reservoir Capacity Capacity (AF) Elevation (ft) Gross Capacity 8748.80 1212.33 Dead Capacity 1723.00 1177.82 Live Capacity 7025.8 0 5 10 15 20 25 30 35 40 RAINFALL(INCHES) YEARS MEAN ANNUAL RAINFALL FOR TAMMAN DAM Mean Annual Month s 36 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 46. Figure 10. PondCapacity IV. CONCLUSIONS Proposed water reservoir in rain-fed area would have a significant favourable effect on intended beneficiaries. In addition to direct benefits from changes in agricultural landscape there will be indirect benefits for the relevant households from appreciation of their land values, creation of employment opportunities / reduction in underemployment and incidence of poverty, groundwater recharge and improvement in drinking water supply. However to ensure realization of anticipated benefits of this dam, it will be essential to undertake capacity building of target beneficiaries in terms of financial resources and knowledge / technology acquisition regarding high value agricultural commodities / enterprises. In fact, the dam site will improve overall water availability, soil properties, land use and consequent socio economic status of the inhabitants and thus enhancing regional development. Minor possible health considerations will be countered through inhabitant sensitization and urging concerned authorities ensuring better healthcare and regional safeguard measures. The project is thus highly recommended as Sustainable Development and could also be replicated to other parts of the country. REFERENCES [1] Minár, J., et al. "Geomorphological information system: idea and options for practical implementation." Geograficky Casopis Slovenskej Akademie, Vied57(3), pp. 247, 2005. [2] Rabus, B., et al. "The shuttle radar topography mission—a new class of digital elevation models acquired by spaceborne radar." ISPRS J. Photogrammetry and Remote Sensing, vol. 57(4), pp. 241-262, 2003. [3] Reuter, H. I., et al. "An evaluation of void‐filling interpolation methods for SRTM data." Intl. J. Geo. Info. Sci., vol. 21(9), pp. 983-1008, 2007. [4]Akhtar, M. R, "Impact of resource conservation technologies for sustainability of irrigated agriculture in Punjab-Pakistan." Journal of Agricultural Research (Pakistan), 2006. [5] Jasrotia, A., et al. "Water balance approach for rainwater harvesting using remote sensing and GIS techniques, Jammu Himalaya, India." Water resources magt., vol23(14), pp. 3035-3055, 2009. [6] Khan, A. A. and S. B. Ali, "Effects of erosion on Indus River bio-diversity in Pakistan." Pak. J. Biol. Sci, vol. 6(12), pp. 1035-1040, 2003. [7] Minár, J., et al. "Geomorphological information system: idea and options for practical implementation." Geograficky Casopis Slovenskej Akademie Vied, 57(3), pp. 247, 2005. [8] Munyao, J. N., et al "Use of Satellite Products to Assess Water Harvesting Potential in Remote Areas of Africa." Land Degrad. Develop, vol. 22, pp. 359-372, 2010. [9] Rabus, B., et al. "The shuttle radar topography mission—a new class of digital elevation models acquired by spaceborne radar." ISPRS J. Photogrammetry and Remote Sensing, vol. 57(4), pp.241-262, 2003. [10] Reuter, H. I., et al. "An evaluation of void‐filling interpolation methods for SRTM data." Intl. J. Geo. Info. Sci,21(9), pp. 983-1008, 2007. [11] Shah, H., et al. "Potential For Investment In Indigenous Technologies: A Case Of Low Cost Soil And Water Conservation Structures In Rainfed Pothwar, Pakistan." Pakistan J. Agric. Res, vol. 25(2), 2012. [12] Vohland, K. and B. Barry, "A review of in situ rainwater harvesting (RWH) practices modifying landscape functions in African drylands,"inAgriculture, Ecosystems & Environment, vol 131(3), pp. 119-127, 2009. 1120 1140 1160 1180 1200 1220 1240 0 5000 10000 15000 Elevationinft Pond Capacity in Acre-Ft TAMMAN DAM SITE Pond Capacity Curve 37 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 47. Urban change detectionof Lahore (Pakistan) using the Thematic Mapper Images of Landsat since 1992-2010 Muhammad Usman Saleem† Institute of Geology, University of the Punjab. P.O. Box, 54590 Lahore Pakistan. Email: osman.geomatics@gmail.com Abstract: Using the techniques of remote sensing and geographic information system, this study is an effort to detect the urbanization in the Lahore (Pakistan) since 1992-2010. Two Landsat TM images from 1992 to 2010 have been used to investigate the urbanization in Lahore. Using unsupervised classification, we have maked classes of land cover in each image. Using Erdas imagine 9.2 urban change detection over Lahore has been investigated. The results of change detection indicate that urbanization in Shahdera, Kala Shah Kako, Kamoki, Sheikhopura, Bharia Town and D.H.A regions of Lahore has increased over 20% since 1992-2010. Band differencing, Red-Green bands differencing techniques verify the results of change detection and confirmed these pockets as intensely urbanized areas since 1992. Maps in Arc GIS 9.3 show these changes in the images. To overcome this urbanization this study also guide us about the Karol,Mohalanwal, Kot Radha Kishan , Kasur etc are the regions where the new residents can be make. Keywords: Urban change detection, remote sensing, land use, Landsat, Urbanization. I. Introduction Urbanization is the conversion of rural areas to urban settlements in the form of expansion of these settlements. There are a lot of factors which impact the urbanizations. The major are industrializations, health, education, and population increase. Lahore is a historical urban city. It was located under the old wall city but now it has been expand tremendously. According to 1998 census eighty two percent (82%) of the total population of the Lahore district is urban and eighteen percent (18%) is rural [1]. Spatial variations in urbanization are the variations in the soil type, habitat, weather pattern and the temporal variations are the variations which are occurred with time. †Corresponding Author Contact: +923217579429. Author postal address: Institute of Geology, University of the Punjab, Lahore Pakistan A lot of analytical and statistical models used which was on based of the urban geometry, spatial relation, social and economic parameters of a city [2]. In literature there was three models of urban land use was used (i) Concentric Zone Model (ii) Sector Model (iii) Multiple Nuclei.[3] But the Geographic Information System (GIS) has totally changed the methods to study the urbanization. Remote sensing with integration of Geographic information system is now powerful tool to investigate urban change detection. Traditional method of in field surveying and mapping required a lot of time span. That why GIS and remote sensing due to their technical soundness, are being mostly used for urban change detection [4]. Integrated study of remote sensing and geographic information system were used to analyze the urbanization process of Lahore city using Landsat Thematic mapper images of 1992-2010. Spatial information was obtained from the remotely sensed images of Landsat satellite. While the results shown in the form of maps are part of geographic information system. The urban development of Lahore city has been very rapid during the last decade and this has dramatically increased the impact on ecosystems. It is therefore necessary to monitor this expansion. This would be very helpful for regional development and decision-making processes. The goal of this study was to detect urban change in Lahore city using different remote sensing techniques e,g; Unsupervised classification , urban change detection, band differencing and red-green differencing and these techniques reveal that Lahore city is facing intensely urbanization from 1992 to 2010. The output is in the form of binary image (called map) which represents the locations where urban change has occurred. II. Aim and Objectives Aim: To investigate urban change detection of Lahore and regions around it by using Landsat TM images. 38 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 48. Objectives: (a) To getquality control data of Landsat Thematic Mapper of year 1990 and 2010. (b) To investigate and analyze the urban change detection of Lahore by making the unsupervised classification. (c) Validate the areas of urbanization (where urbanization increase or decrease) by using band differencing, red green differencing techniques. (d) Map these areas of urbanization. III. Study Area Lahore (310 34’ N, 740 22’ E) is the capital city of Punjab province of Pakistan. According to census 1998 it is the second most populated city of Pakistan. It is located near to the border with India in east, Sheikhupura on the north-west, Kasur district on the south. Dated back to a thousand of year, it is an urban city(see figure 1). This city has extreme climate with four seasons in a year. Summer season begins in April and lasted till September. May and June are the hottest months in the summer where the mean monthly temperature for maximum and minimum are 45.40 C and 29.30 C respectively [4]. Winter season start from November till March. December is the coolest month in this season and also coolest month of the year. The mean monthly maximum and minimum temperatures for this month are 21.1 0 C and 7.20 C respectively [5]. The monsoon period starts from July and lasted till September. Temperature falls to 480 C in May and June. The average maximum and minimum temperatures in Lahore are 30.80 C and 17.80 respectively. Being a major develop city, urbanization is going to rise rapidly. It is the consequences of the coming of people from the village to city due to more chance of job and better infrastructure. Lahore‘s urbanization going to increase since 1973 which is leading to rise in temperature [5]. Increase in the number of industries give rise to major source of rise in temperature. Total area of the Lahore is 1772 km2 [3]. Total population was 6.319million in 1998 with population density of 3,566 people per km2 [3]. Figure 1: Location of Lahore in the Map of Pakistan. IV. Methodology In this study, two Landsat Thematic Mapper images of October 1992 and October 2010 were acquired from the USGS website. These images are already geometrically rectified. The row number for these images was 149. Thematic Mapper Images mostly used for the environmental assessment and monitoring [6]. USGS website provides TM data in the separate bands. We have performed Layer stacking on these images. Lahore as the study area for research, we have cut subsets of both images. Both subsets images are not of the same size. October 2010 image’s subset is more in extending than October 1992 image (See fig.2 & 3). The band combination 4, 3, 2 was selected to detect urbanization [6]. All these processing of data is performed in Erdas Imagine 9.2 software. To show results we have made maps in power software of Arc GIS 9.3 Lahore 39 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 49. Figure 2: Subsetof the Landsat TM image showing Lahore and its neighbors (Oct 1992). Figure 3: Subset of Landsat TM image of Lahore, October 2010. V. Image Classifications Image classification is generally divided into two broad categories, Unsupervised Classification and Supervised Classification. Unsupervised classification is generally performed on software. In which the software search the pixel of equal brightness value to assigned a class of land covers. While in the case of Supervised Classifications, we gives sample class to the software and then software makes the classes of land covers having same brightness values of our sample class. We have done with unsupervised classifications and make four classes in October 1992 and October 2010 TM images. In October 1992 image these classes are Water, Vegetation, Urban Area, Unused Land while in October 2010 image, these are Water, vegetation, Urban Area, Intensely Urbanize area. Each class was labeled and colored for visual assessment. (See figure 4 and 5) Figure 4: Unsupervised Classification of subset Oct, 1992 (blue is water, green is vegetation, red is urban area). VI. Urban Change Detection Urban change detection is the processing of change identification in the states of land covers by observing these changes in the different time periods. With the help of tool change detection in Erdas imagine 9.2, we have find out changes occurred the both images (see the figure 6). As we have mentioned before these two subset images are not of the same size that why borders are in red of figure 6. 40 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 50. Figure 5: Unsupervisedclassification of TM Oct 2010 image (blue is water, yellow is urban area) Table 1: Color scheme for the figure 6 Color Urbanization 1992-2010 Blue 5% decrease Red > 20% increase Yellow 0-20% increase Black Back ground in the image VII. Band Differencing With the help of spatial modeler in Erdas Imagine 9.2, we have made a model to find the difference between two images in band 7 by using nearest neighbor interpolation technique (See figure 7 and 8).But this model is more suitable for the change detection visually. Figure 6: Urban Change Detection of Lahore from 1992 to 2010. Figure 7: Model in the Erdas Imagine 9.2 for the band differencing 41 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 51. Figure 8: Banddifferencing of TM of Lahore (1992- 2010). Light gray areas in the figure 8 indicate the areas where urbanization has increased [3].This technique utilized the differences in brightness value of one pixel to corresponding pixel of another image. Neutral gray tune in figure 8 black color represent zeros difference while white has the maximum difference between the images. VIII. Red-Green Differencing This technique is widely used to find the change in the areas of two images of different times. To detect the urban change, open the most recent image (2010) in the red band and old image (1992) in the green band. In the Erdas Imagine 9.2, we swipe these two images. Combination of red and green produce yellow color. But the areas which have changed show reddish color in image (see figure 9) are the regions where urbanization has occurred. In remote sensing we cannot use hard statements about any results. Through this technique we also verify the areas where the urbanization has occurred from 1992-2010. Figure 9: Red-Green Differencing of Landsat Images 1992-2010. IX. Results and Discussions This study focused on the important of remote sensing and GIS techniques in the analyzing of urban change detection of Lahore and its neighbor cities. This study shows that digital image processing and GIS can be used for producing urbanized maps of Lahore. As in remote sensing we have to rely on the number of remote sensing techniques which leading us to one conclusion. We have used the TM data with the interval of 18 years to investigate the urbanizations. Change detection tell us that during the study period urban sprawl of Lahore has been shifted towards Shahdera, Kala Shah Kako , Sheikupura and Kamoki where the urbanization has been increased by 20% . For validation of these results we have use the band differencing and red-green differencing techniques which shows that old city of Lahore is going to expand in the urban areas and this pattern of urbanization is shifted towards Bharia Town, Johar Town, and Chung. In these areas people are going to make urbanization over than 20% from 1992-2010. According to [3] Lahore was 58977 hectares in 1972 and which now (2009) has increased to 99173 hectares. Our results matching with the results of previous study of Lahore’s urbanization of > 20% from 1990 to 2009 has occurred. According to 1998 Census, the average family size of the Lahore city was 7.1 Urbanization 42 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 52. persons, but during1998 to 2007, the urbanization rate has increased against population [5]. X. Conclusions This effort proved its usefulness in the present scenario of rising trends of population of Lahore city. With the use of remote sensing and GIS techniques, we have investigated new pockets around the Lahore where the citizen makes more residents and plots. Major areas which are facing urbanization since 1990 are Kala Shah Kako, Sheikhopura, Meridke, Bheria Town, Kamoki, Chungm Raiwind, Shahdera and Johar Towns where urbanizations from 18 years raised 0-20 % even more than 20%. This study also guides us Karol,Mohalanwal , Kot Radha Kishan ,Kasur are the regions where the Government can plan new resident plots. We suggest that future study on different land use/ land cover patterns of Lahore. Future study should be undertaken the loss in the area under decrease in natural vegetation and agriculture. This study strongly recommend for Government and urban scholars that such studies should be revised after nominal time interval with high resolution satellite imagery like Spot, QuickBird etc. ACKNOWLEDGEMENT Author would like to say thanks of USGS team who provide him satellite images for research. Also I would like to say thanks deeply to Dr. Najam Abbas, Secretary ICASE 2015,Institute of Space Technology, Islamabad Pakistan for providing his kind contribution, for publication of this research work. XI. REFERENCES [1] District Census Reports for each District containing General Description of the District and Broad Analysis of Population and Housing Data followed by detailed statistical tables (1998-Census). [2] Young,X. & Lo, C. P. 'Modelling urban growth and landscape changes in the Atlanta metropolitan area', international journal of geographical science, vol. 17, no. 5, pp. 463,463-488, 2003. [3] O. Riaz, 'URBAN CHANGE DETECTION OF LAHORE (PAKISTAN) USING A TIME SERIES OF SATELLITE IMAGES SINCE 1972', ASIAN JOURNAL OF NATURAL & APPLIED SCIENCES., vol. 2, no. 4, pp. 101,102,103,104, 2013. [4] Jensen and D. Cowen, 'Remote Sensing of Urban Suburban Infrastructure and Socio- economic Attribute', Photogrammetric Engineering and Remote Sensing, vol. 65, pp. 611-622, 1999. V. Mesev, 'The Use of Census Data in Urban Image Classification.', Photogrammetric Engineering and Remote Sensing, vol. 64, no. 5, pp. 431-438, 1998. C. Lo and X. Yang, 'Drivers of land-use/ land-cover changes and dynamics modeling for the Atlanta. Georgia Metropolitan Area.', Photogrammetric Engineering and Remote Sensing,, vol. 68, no. 10, pp. 1073–1082., 2002. [5] ureshi, S. Mahmood, A. S. Almas and R. Irshad, 'MONITORING SPATIOTEMPORAL AND MICRO- LEVEL CLIMATIC VARIATIONS IN LAHORE AND SUBRUBS USING SATELLITE IMAGERY AND MULTI- SOURCE DATA', Journal of Faculty of Engineering & Technology, vol. 19, pp. 53,55, 2012. [6] H. XU, X. WANG and G. XIAO, 'A REMOTE SENSING AND GIS INTEGRATED STUDY ON URBANIZATION WITH ITS IMPACT ON ARABLE LANDS: FUQING CITY, FUJIAN PROVINCE, CHINA', LAND DEGRADATION & DEVELOPMENT, vol. 11, p. 303, 2000. 43 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 53. A Review ofFundamentals of Hyperspectral Imaging and its Applications Asad Abbas, Khurram Khurshid Electrical Engineering Department Institute of Space Technology Islamabad, Pakistan Email: asad.abbas@outlook.com, khurram.khurshid@ist.edu.pk Abstract—Over the past two decades, hyperspectral imagery has emerged as an effective tool for geo-observation using airborne and satellite based systems. Current sensor technologies are capable of covering large surfaces of earth with exceptional spatial, spectral and temporal resolutions. These features allows the use of hyperspectral imaging in numerous remote sensing applications requiring estimation of physical parameters of many complex surfaces and identification of visually similar materials having fine spectral signatures. Moreover, applications of hyper- spectral imaging are growing rapidly, it has shown enormous potential in various fields including biomedical, archeology, food quality control, military defense and conservation of artistic and historic objects. This review describes the differences between hyperspectral and multispectral data, fundamentals of hyper- spectral imaging and focuses on the modern applications of hyperspectral imagery in precision agriculture, water resource management and document imaging. I. INTRODUCTION Human eye is only able to see in a limited part of electro- magnetic spectrum and can distinguish between objects based on their different spectral responses in that narrow spectral range [1]. However, multispectral imaging sensors have been developed that are able to acquire an image in infrared and visible segments of electromagnetic spectrum. Thus allowing material identification on the basis of their unique spectral sig- nature in a wide spectral range. Multispectral imaging exploits the property that each material has its own unique spectral signatures. Spectrum of a single pixel in a multispectral image provides information about its constituents and surface of the material. Multispectral imaging technology is being used for envi- ronment and land observational remote sensing from satel- lite and airborne systems since late 1960s [2]. Multispectral imaging systems acquire data in a small number of spectral bands by using parallel sensor arrays. Most of the multispectral imaging systems uses three to six spectral bands with large optical band intervals, ranging from visible to near infrared regions of electromagnetic spectrum for scene observation. However, such low number of spectral bands are the limiting factor for discrimination of various materials. The development of hyperspectral sensing systems over the past two decades made it possible to acquire several hundred spectral bands of observational scene in a single acquisition. The increased spectral resolution of these ”hyperspectral” images allow detail examination of land surfaces and different materials present in the observational scene, which was previously not possible with low spectral resolution of multispectral imaging scanners. Hyperspectral imaging (HSI), also called imaging spec- trometer [3], is a technique in which an object is photographed using several well defined optical bands in broad spectral range. It was originally implemented on satellite and airborne platforms for remote sensing applications but during last two decades, HSI has been applied to numerous applications including agricultural and water resources control [4], [5], military defense, art conservation and archeology [6], [7], medical diagnosis [8], [9], analyses of crime scene details [10], [11], document imaging [12], forensic medicine [13], food quality control [14], [15] and mineralogical mapping of earth surface [16]. This review details the fundamentals of hyperspectral imaging, discusses the common hyperspectral remote sensing terminologies and highlights the modern applications of hy- perspectral imagery in the field of precision agriculture, water resource management and document imaging. II. FUNDAMENTALS OF HYPERSPECTRAL IMAGING Unlike normal 2-D images taken by regular camera, hy- perspectral images are characterized by their spatial as well as spectral resolution. The spatial resolution measures the geometric relationship of the image pixels to each other. While the spectral resolution determines the variations within image pixels as a function of wavelength. Table I shows the spatial and spectral resolution of the current airborne and space satellite imaging sensors. Due to spectral resolution as well as spatial resolutions generally hyperspectral images are referred as hyperspectral data cubes. Figure 1 shows a typical hyperspectral data cube as a result of hyperspectral imaging. A. Spatial Resolution Spatial resolution can be defined as the smallest discernible detail in an image [17]. Which can be described as the measure of smallest object in an image, that can be distinguished as a separate entity in the image. In practical situations clarity of the image is dictated by it spatial resolution, not the number of pixels in an image. Spatial characteristics of an image depends on the design of imaging sensor in terms of its field of view and its altitude [18]. A finite patch of the ground is captured by each detector in a remote imaging sensor. Spatial resolution is inversely proportional to the patch size. The smaller the size of the patch, higher the details that can be interpreted from the observed scene. 44 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 54. Fig. 1. Hyperspectraldatacube: for any image pixel a complete spectral curve is observed [12] TABLE I. CURRENT SPACEBORNE AND AIRBORNE SYSTEMS PROVIDING OPTICAL DATA FOR LAND MAPPING Optical Subsystem Spectral Spectral Spatial Resolution Spatial Bands (m) Range (µm) Resolution Coverage Spaceborne Landsat VNIR-TIR 8 0.45-12.50 15-60 Global MODIS VNIR-TIR 36 0.40-14.40 250-1000 Global MERIS VNIR 15 0.39-1.040 300 Global ASTER VNIR-TIR 15 0.52-11.65 15-90 Global Hyperion VNIR-SWIR 242 0.40-2.500 30 Regional ALOS VIS 1 0.52-0.77 2.5 Local Airborne AVIRIS VNIR 224 0.38-2.500 4-20 Local HyMap VNIR-SWIR 128 0.45-2.480 2-10 Local ROSIS VNIR 115 0.42-0.873 2 Local DAIS-7915 VNIR-TIR 79 0.45-12 3-10 Local B. Spectral Resolution Spectral resolution can be defined as the number of spectral bands and range of electromagnetic spectrum measured by the sensor. An imaging sensor might respond to a large frequency range but still have a low spectral resolution if it acquires small number of spectral bands. On the contrary, if a sensor is sensitive to small frequency range but captures large number of spectral bands has high spectral resolution, due to its ability to distinguish between scene elements having close or similar spectral signatures [19]. Multispectral images have a low spectral resolution, thus unable to resolve finer spectral signatures present in the scene. Hyperspectral imaging (HSI) sensors acquire images in numerous contiguous and extremely narrow spectral bands in mid infrared, near infrared and visible segments of electromagnetic spectrum. This type of advance imaging system shows tremendous potential for material identification on the basis of their unique spectral signatures [2]. Spectrum of a single pixel in a hyperspectral image can give considerably more information about the surface of the material than a normal image. C. Temporal Resolution In hyperspectral remote sensing, the temporal resolution depends on the orbital characteristics of the imaging sensor. It is generally defined as the time needed by the sensor platform to revisit and obtain data from the exact same location [20]. Temporal resolution is said to be high if the revisiting fre- quency of the sensor platform for the exact same location is high and is said to be low if revisiting frequency is low. It is normally defined in days. D. Understanding Spectral Signatures Materials present on the earth’s surface emit, transmit, reflect and absorb electromagnetic energy from the sun in their own unique way. Hyperspectral sensors allows us to measure all types of electromagnetic energy within a specified range as it interacts with materials, thus allowing us to observe the distinct features and changes on earth’s surface. Reflectance is defined as the percentage of amount of energy bouncing back from material’s surface. It is a ratio of reflected energy to incident energy as a function of wavelength [18]. Reflectance is 100% if all the light energy of specific wavelength striking the object is reflected back to the imaging sensor, on the other hand reflectance is 0% if all the incident light of specific wavelength is absorbed by the object. In most practical cases reflectance values are in between these two extremes. In a specified range of electromagnetic spectrum, the reflectance values of different materials present on the earth’s surface such as soil, forest, water, minerals etc can be plotted and compared. Such plots are labeled as ”spectral signatures” or spectral response curves” [21]. Figure 2 demonstrates a general model of spectral signatures of different materials present on the earth’s surface. Remotely sensed images can be classified using these spectral signature plots, as each material present in an observed scene has its own unique spectral signature. The more the spectral resolution of an imaging sensor, the more classification information can be extracted from spectral signatures. Hyperspectral sensors have high spectral resolution than multispectral sensors and thus provide the ability to distinguish more subtle differences in a scene. Hyperspectral imagery has been utilized by Fig. 2. A generic scheme of HSI mapping of soil, vegetation and water [22] geologists for mapping the land and water resources [16]. They have additionally been utilized to map heavy metals and other hazardous wastes in historic and active mining areas. Figure 3 shows the spectral response curves of dry bare soil, green vegetation and clean water. Figure 3 shows that the reflectance curve for bare soil has less variations as compared to that of vegetation. This is because of the fact that the factors that affect soil reflectance vary in a narrow range of electromagnetic spectrum. These factors include soil texture, presence of minerals such as iron, surface roughness and moisture content in soil [21]. Spectral 45 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 55. Fig. 3. Spectralresponse curves of soil, vegetation and water [18] signatures of green vegetation have valleys in the visible portion of spectrum that indicates the pigmentation of the plant. Chlorophyll is the primary photosynthetic pigment in green vegetation [18], it is absorbs strongly in red (670 nm) and blue (450 nm) regions called the chlorophyll absorption spectral bands. When a plant is under stress so that the chlorophyll growth is reduced, in such cases the amount of reflectance in red (670 nm) regions increases [18]. The spectral response of water has a distinctive characteristics of absorption of light in near infrared and beyond. Common factors affecting spectral response of water are the suspended sediments and increases in chlorophyll levels. In each cases spectral response will be shifted accordingly showing the presence of suspended sediments or algae in water [21]. III. APPLICATIONS OF HYPERSPECTRAL IMAGING A. Precision Agriculture Many studies have indicated that the world’s crop production needs to be doubled by the end 2050 due to the rapid increase in world’s population [23]. However, various studies have shown that the crop yields are not longer increasing at a rate to fulfill the growing population needs [24], [25]. Recent studies have indicated that increasing crop yields, rather than using more land for cultivation, is the most effective path for ensuring food security [26], [27]. Global poverty and undernourishment can directly be reduced by increasing crop production, moreover most of the poor and undernourished population consists of farmers themselves [28]. Traditionally crop monitoring for disease, water stress, nutrients and insect attack was carried out by manual visual inspection from the ground. These methods were limited by the fact that the visual symptoms often appear at later stages of disease, thus making it difficult to restore plant health. Advancement in airborne and ground based hyperspectral imaging methods has made possible the evaluation of crop stresses, analyzing soil and vegetation characteristics in a cost effective manner, thus replacing the traditional scouting methods. Drought stress is an important factor affecting crop yields. Chances of a successful crop can be highly increased by timely detection of water related stresses. High water level stresses are noticeable in variations in photosynthetic pigments. These changes leads to yellowish tint in crops, due to the increase reflectance of red wavelength. Unlike human eye, hyperspectral imaging sensors can detect these changes at earlier stages. Colombo et al. [29] indicated that changes in leaf equivalent water thickness (EWT) was responsible for changes in leaf reflectance in the infrared and visible spectrum. They stated that hyperspectral regression indices calculated from hyperspectral imaging were powerful tools for estimation of water content at leaf as well as at landscape level. Rascher et al. [30] used a portable hyperspectral imaging system and photochemical reflectance index to estimate water stress in leaves of tropical tress and observed the temporal effects of dehydrations on tree leaves. Rossini et al. [31] found that hyperspectral imaging is useful in detecting drought stress at farm level with corn. They showed that irrigation deficits can be accurately mapped well before drought stress affected the canopy structure. Deficiencies in nutrients and soil contamination causes various symptoms that can be assessed by hyperspectral imaging. Schuerger et al. [32] used hyperspectral imaging to observe zinc deficiency and toxicity to determine chlorophyll levels relating to stress symptoms. They indicated that traditional direct sampling methods are much more costly than hyperspectral imaging. Dunagen et al. [33] analyzed mercury levels in mustard plants and found that spectral signatures were notably related to the contaminant levels. Osborne et al. [34] showed that biomass, yield under stress, nitrogen and phosphorous concentrations can be estimated by using hyperspectral imaging. Mahlein et al. [35] studied different development stages of diseased suger beet leaves using hyperspectral imaging. Figure 4 shows spectral signatures of healthy as well as diseased sugar beet leaves. This study also showed that hyperspectral imaging has a great potential for analyzing plant diseases. Analysis of soil characteristics can play a vital role in Fig. 4. Spectral signatures comparison of healthy and diseased suger beet leaves [35] increasing crop yields. Ben-Dor et al. [36] al successfully mapped vital characteristics of soil in a field scale experiment including moisture, soil organic matter and soil sanity. Rossel and Bratney [37] estimated the organic carbon content in soil with accuracy. B. Water Resource Management Water is one of the most important resource available on the earth and is vital for survival of humanity. For this 46 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 56. reason, managing thewater resource efficiently, analyzing and monitoring the quality of water has attracted a lot of attention from the researchers [38], [39], [40], [41], [42], [43]. Hyperspectral remote sensing technology has found enormous applications in water resource management. Accurate estimates of water resource parameters are possible by analyzing spatial, spectral and temporal variations in water bodies. The efficacy of flood detection and monitoring system is limited by their incapacity to get important information about water conditions from airborne and ground observatories in a timely manner. Recent improvements in remote sensing tech- nology has improved the early flood warning system and vastly reduced the time of detection and reaction to flood events to a few hours [44]. US geological survey and NASA are incorporating space borne observations of rainfall resources, rivers and land topography into early warning systems with potential global applications [45]. Glaber and Reinartz [46] studies the optimal procedure for detection of flooded areas with remote sensing data. They investigated the erosive impact of floods, moisture content in flood plain areas, accumula- tion of sediments. Roux and Dartus [47] explored the flood hydrographs and estimated the river discharge from remotely sensed data. They optimized their model to minimize the error between system response and their proposed model to estimate the river discharge. Hyperspectral remote sensing provides efficient and reli- able information about water quality parameters which contain biochemical, hydro-physical and biological attributes [48] [49] [50]. Hyperspectral imaging enable us to measure chlorophyll, turbidity and chemical oxygen demand and phosphorous in water resources. Chlorophyll content in water is extensively studied by hyperspectral remote sensing, which gives and estimate of algal level and hence water quality. Studies have been carried out for evaluation of ammonia changes for wetland [51], classifying different parameters of lakes [52], estuaries [53] and analyzing algal blooms [54]. Wetland mapping has played a significant role in order to enhance the quality of our ecosystem [55]. Hyperspectral imagery has helped in detailed understanding of vegetation characteristics of ecosystem. Extensive research studies have been carried out using remote sensing to explore the signif- icance of acquiring timely data for monitoring and mapping aquatic vegetation [56], which is said to be an important aspect in ecosystem reconstruction and restoration. C. Document Imaging Traditionally, forensic document experts and paleographers used chemical solution based methods to study the extrinsic and intrinsics components of the important historic docu- ments [57]. This is due to the fact that the inks used on documents throughout the history were composed of diverse substances having distinct chemical and physical properties. Each of these substances have their own unique way for reacting with different substrates depending upon the reaction environment. These chemical solution based methods helped in document analysis. But unfortunately these techniques were time consuming, sensitive to temperature changes and destruc- tive in nature i.e. harms to the important documents were irreversible. To overcome such limitations, hyperspectral imaging has emerged as an effective non-destructive tool for improving readability [58], ink aging and forensic document analysis [59]. Hyperspectral document imaging works on the principle that each ink present in the document has its own unique spectral signature. Many mathematical tools that are being used in hyperspectral remote sensing can be used on hyperspectral document images for classification, improving legibility of extremely deteriorated text, ink aging and fraud detection. Moreover hyperspectral imaging is non-destructive, automated and environment insensitive tool for document examination. In hyperspectral document imaging, ink mismatch de- tection analysis provides important information to forensic document examiners to determine the authenticity of legal documents. Forgery, backdating and fraud can be detected using ink analysis of documents. Recently, we have worked on a non-destructive automated hyperspectral document imaging analysis system, which was able successfully discriminates between different visually similar ink mixtures (Dataset D1 to D7) taken from hyperspectral imaging database [60]. Figure 5 shows the final segmentation results of our proposed approach. Fig. 5. Comparisons of Ground Truth Images (Dataset D1 to D7) with final segmentation results IV. DISCUSSION AND CONCLUSION Among remote sensing technologies, the role of hyperspec- tral imagery in the geo-observation, identification and detection of materials and estimation of physical parameters cannot be stated enough. Due to this very reason there are increasing number of airborne and spaceborne hyperspectral platforms based applications being researched. Recent advancement in 47 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 57. sensor technologies hasencouraged researchers to use hyper- spectral imagery in many modern applications. Many mathe- matical tools and algorithms are being researched such as data fusion, hyperspectral unmixing, hyperspectral classification, anomaly detection and fast computing for efficient utilization of hyperspectral data. These mathematical tools can be used on hyperspectral data across many different applications. In general, this review focuses on the vast extent to which hyperspectral imaging has been used to increase the crop yields, managing water resources and its advanced application in historic and modern document imaging. Promising results have been found in those areas and also future research is being carried out for further improvement as well. REFERENCES [1] E. H. Land and J. McCann, “Lightness and retinex theory,” JOSA, vol. 61, no. 1, pp. 1–11, 1971. [2] D. Landgrebe, “Information extraction principles and methods for multispectral and hyperspectral image data,” Information processing for remote sensing, vol. 82, pp. 3–38, 1999. [3] J. Solomon and B. Rock, “Imaging spectrometry for earth remote sensing,” Science, vol. 228, no. 4704, pp. 1147–1152, 1985. [4] M. Govender, K. Chetty, and H. Bulcock, “A review of hyperspectral remote sensing and its application in vegetation and water resource studies,” Water Sa, vol. 33, no. 2, 2007. [5] E. Adam, O. Mutanga, and D. Rugege, “Multispectral and hyperspectral remote sensing for identification and mapping of wetland vegetation: a review,” Wetlands Ecology and Management, vol. 18, no. 3, pp. 281– 296, 2010. [6] C. Fischer and I. Kakoulli, “Multispectral and hyperspectral imaging technologies in conservation: current research and potential applica- tions,” Studies in Conservation, vol. 51, no. Supplement-1, pp. 3–16, 2006. [7] H. Liang, “Advances in multispectral and hyperspectral imaging for archaeology and art conservation,” Applied Physics A, vol. 106, no. 2, pp. 309–323, 2012. [8] O. Carrasco, R. B. Gomez, A. Chainani, and W. E. Roper, “Hyperspec- tral imaging applied to medical diagnoses and food safety,” in AeroSense 2003. International Society for Optics and Photonics, 2003, pp. 215– 221. [9] M. A. Afromowitz, J. B. Callis, D. M. Heimbach, L. A. DeSoto, and M. K. Norton, “Multispectral imaging of burn wounds: a new clinical instrument for evaluating burn depth,” Biomedical Engineering, IEEE Transactions on, vol. 35, no. 10, pp. 842–850, 1988. [10] J. Kuula, I. P¨ol¨onen, H.-H. Puupponen, T. Selander, T. Reinikainen, T. Kalenius, and H. Saari, “Using vis/nir and ir spectral cameras for detecting and separating crime scene details,” in SPIE Defense, Security, and Sensing. International Society for Optics and Photonics, 2012, pp. 83 590P–83 590P. [11] R. L. Schuler, P. E. Kish, and C. A. Plese, “Preliminary observations on the ability of hyperspectral imaging to provide detection and visu- alization of bloodstain patterns on black fabrics,” Journal of forensic sciences, vol. 57, no. 6, pp. 1562–1569, 2012. [12] R. Padoan, T. A. Steemers, M. Klein, B. Aalderink, and G. De Bruin, “Quantitative hyperspectral imaging of historical documents: technique and applications,” ART Proceedings, 2008. [13] G. Edelman, E. Gaston, T. Van Leeuwen, P. Cullen, and M. Aalders, “Hyperspectral imaging for non-contact analysis of forensic traces,” Forensic science international, vol. 223, no. 1, pp. 28–39, 2012. [14] A. Gowen, C. O’Donnell, P. Cullen, G. Downey, and J. Frias, “Hyper- spectral imaging–an emerging process analytical tool for food quality and safety control,” Trends in Food Science & Technology, vol. 18, no. 12, pp. 590–598, 2007. [15] Y.-Z. Feng and D.-W. Sun, “Application of hyperspectral imaging in food safety inspection and control: a review,” Critical reviews in food science and nutrition, vol. 52, no. 11, pp. 1039–1058, 2012. [16] R. N. Clark and G. A. Swayze, “Mapping minerals, amorphous ma- terials, environmental materials, vegetation, water, ice, and snow, and other materials: The usgs tricorder algorithm,” in Summaries of the Fifth Annual JPL Airborne Earth Science Workshop, vol. 1. JPL Publication, 1995, pp. 39–40. [17] R. C. Gonzalez, Digital image processing. Pearson Education India, 2009. [18] R. Smith, “Introduction to remote sensing of the environment,” 2001b. www. microimages. com, 2001. [19] J. R. Jensen, Introductory digital image processing: A remote sensing perspective. Pearson College Division, 2005. [20] J. Thau, “Temporal resolution,” in Encyclopedia of GIS. Springer US, 2008, pp. 1150–1151. [Online]. Available: http://dx.doi.org/10.1007/ 978-0-387-35973-1 1376 [21] T. Lillesand, R. W. Kiefer, and J. Chipman, Remote sensing and image interpretation. John Wiley & Sons, 2014. [22] G. A. Shaw and H.-h. K. Burke, “Spectral imaging for remote sensing,” Lincoln Laboratory Journal, vol. 14, no. 1, pp. 3–28, 2003. [23] D. Tilman, C. Balzer, J. Hill, and B. L. Befort, “Global food demand and the sustainable intensification of agriculture,” Proceedings of the National Academy of Sciences, vol. 108, no. 50, pp. 20 260–20 264, 2011. [24] K. G. Cassman, “Ecological intensification of cereal production sys- tems: yield potential, soil quality, and precision agriculture,” Proceed- ings of the National Academy of Sciences, vol. 96, no. 11, pp. 5952– 5959, 1999. [25] D. K. Ray, N. Ramankutty, N. D. Mueller, P. C. West, and J. A. Foley, “Recent patterns of crop yield growth and stagnation,” Nature communications, vol. 3, p. 1293, 2012. [26] H. C. J. Godfray, J. R. Beddington, I. R. Crute, L. Haddad, D. Lawrence, J. F. Muir, J. Pretty, S. Robinson, S. M. Thomas, and C. Toulmin, “Food security: the challenge of feeding 9 billion people,” science, vol. 327, no. 5967, pp. 812–818, 2010. [27] J. A. Foley, N. Ramankutty, K. A. Brauman, E. S. Cassidy, J. S. Gerber, M. Johnston, N. D. Mueller, C. OConnell, D. K. Ray, P. C. West et al., “Solutions for a cultivated planet,” Nature, vol. 478, no. 7369, pp. 337– 342, 2011. [28] P. L. Pingali, “Green revolution: Impacts, limits, and the path ahead,” Proceedings of the National Academy of Sciences, vol. 109, no. 31, pp. 12 302–12 308, 2012. [29] R. Colombo, M. Meroni, A. Marchesi, L. Busetto, M. Rossini, C. Gi- ardino, and C. Panigada, “Estimation of leaf and canopy water content in poplar plantations by means of hyperspectral indices and inverse modeling,” Remote Sensing of Environment, vol. 112, no. 4, pp. 1820– 1834, 2008. [30] U. Rascher, C. J. Nichol, C. Small, and L. Hendricks, “Monitoring spatio-temporal dynamics of photosynthesis with a portable hyperspec- tral imaging system,” Photogrammetric Engineering & Remote Sensing, vol. 73, no. 1, pp. 45–56, 2007. [31] M. Rossini, F. Fava, S. Cogliati, M. Meroni, A. Marchesi, C. Panigada, C. Giardino, L. Busetto, M. Migliavacca, S. Amaducci et al., “Assessing canopy pri from airborne imagery to map water stress in maize,” ISPRS Journal of Photogrammetry and Remote Sensing, vol. 86, pp. 168–177, 2013. [32] A. C. Schuerger, G. A. Capelle, J. A. Di Benedetto, C. Mao, C. N. Thai, M. D. Evans, J. T. Richards, T. A. Blank, and E. C. Stryjewski, “Comparison of two hyperspectral imaging and two laser-induced fluorescence instruments for the detection of zinc stress and chlorophyll concentration in bahia grass (paspalum notatum flugge.),” Remote sensing of environment, vol. 84, no. 4, pp. 572–588, 2003. [33] S. C. Dunagan, M. S. Gilmore, and J. C. Varekamp, “Effects of mercury on visible/near-infrared reflectance spectra of mustard spinach plants (brassica rapa p.),” Environmental Pollution, vol. 148, no. 1, pp. 301– 311, 2007. [34] S. Osborne, J. S. Schepers, D. Francis, and M. R. Schlemmer, “De- tection of phosphorus and nitrogen deficiencies in corn using spectral radiance measurements,” Agronomy journal, vol. 94, no. 6, pp. 1215– 1221, 2002. [35] A.-K. Mahlein, U. Steiner, C. Hillnh¨utter, H.-W. Dehne, and E.-C. Oerke, “Hyperspectral imaging for small-scale analysis of symptoms 48 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 58. caused by differentsugar beet diseases,” Plant methods, vol. 8, no. 1, p. 3, 2012. [36] E. Ben-Dor, K. Patkin, A. Banin, and A. Karnieli, “Mapping of several soil properties using dais-7915 hyperspectral scanner data-a case study over clayey soils in israel,” International Journal of Remote Sensing, vol. 23, no. 6, pp. 1043–1062, 2002. [37] C. Gomez, R. A. V. Rossel, and A. B. McBratney, “Soil organic carbon prediction by hyperspectral remote sensing and field vis-nir spectroscopy: An australian case study,” Geoderma, vol. 146, no. 3, pp. 403–411, 2008. [38] F. Zagolski, V. Pinel, J. Romier, D. Alcayde, J. Fontanari, J. Gastellu- Etchegorry, G. Giordano, G. Marty, E. Mougin, and R. Joffre, “Forest canopy chemistry with high spectral resolution remote sensing,” Inter- national Journal of Remote Sensing, vol. 17, no. 6, pp. 1107–1128, 1996. [39] G. P. Asner, “Biophysical and biochemical sources of variability in canopy reflectance,” Remote sensing of Environment, vol. 64, no. 3, pp. 234–253, 1998. [40] K. McGwire, T. Minor, and L. Fenstermaker, “Hyperspectral mixture modeling for quantifying sparse vegetation cover in arid environments,” Remote Sensing of Environment, vol. 72, no. 3, pp. 360–374, 2000. [41] C. Stone, L. Chisholm, and N. Coops, “Spectral reflectance character- istics of eucalypt foliage damaged by insects,” Australian Journal of Botany, vol. 49, no. 6, pp. 687–698, 2001. [42] N. Coops, S. Dury, M.-L. Smith, M. Martin, and S. Ollinger, “Com- parison of green leaf eucalypt spectra using spectral decomposition,” Australian journal of botany, vol. 50, no. 5, pp. 567–576, 2002. [43] E. Underwood, S. Ustin, and D. DiPietro, “Mapping nonnative plants using hyperspectral imagery,” Remote Sensing of Environment, vol. 86, no. 2, pp. 150–161, 2003. [44] F. Ip, J. Dohm, V. Baker, T. Doggett, A. Davies, R. Castano, S. Chien, B. Cichy, R. Greeley, R. Sherwood et al., “Flood detection and monitoring with the autonomous sciencecraft experiment onboard eo-1,” Remote Sensing of Environment, vol. 101, no. 4, pp. 463–481, 2006. [45] G. Brakenridge, E. Anderson, and S. Nghiem, “Satellite microwave detection and measurement of river floods,” in AGU Spring Meeting Abstracts, vol. 1, 2006, p. 05. [46] C. Gl¨aßer and P. Reinartz, “Multitemporal and multispectral remote sensing approach for flood detection in the elbe-mulde region 2002,” Acta hydrochimica et hydrobiologica, vol. 33, no. 5, pp. 395–403, 2005. [47] H. Roux and D. Dartus, “Use of parameter optimization to estimate a flood wave: Potential applications to remote sensing of rivers,” Journal of Hydrology, vol. 328, no. 1, pp. 258–266, 2006. [48] M. Chawira, T. Dube, and W. Gumindoga, “Remote sensing based water quality monitoring in chivero and manyame lakes of zimbabwe,” Physics and Chemistry of the Earth, Parts A/B/C, vol. 66, pp. 38–44, 2013. [49] T. Kutser, B. Paavel, C. Verpoorter, T. Kauer, and E. Vahtm¨ae, “Remote sensing of water quality in optically complex lakes,” in Proceedings of the XXII Congress of the International Society for Photogrammetry and Remote Sensing, Melbourne, Australia, 2012. [50] S. Koponen, “Remote sensing of water quality: the development and use of water processors available in beam,” Finland, Helsinki: Helsinki University of Technology TKK, Department of Radio Science and Engineering, pp. 1–58, 2006. [51] D. R. Tilley, M. Ahmed, J. H. Son, and H. Badrinarayanan, “Hyper- spectral reflectance of emergent macrophytes as an indicator of water column ammonia in an oligohaline, subtropical marsh,” Ecological Engineering, vol. 21, no. 2, pp. 153–163, 2003. [52] S. Thiemann and H. Kaufmann, “Lake water quality monitoring using hyperspectral airborne dataa semiempirical multisensor and multitem- poral approach for the mecklenburg lake district, germany,” Remote sensing of Environment, vol. 81, no. 2, pp. 228–237, 2002. [53] J.-M. Froidefond, L. Gardel, D. Guiral, M. Parra, and J.-F. Ternon, “Spectral remote sensing reflectances of coastal waters in french guiana under the amazon influence,” Remote Sensing of Environment, vol. 80, no. 2, pp. 225–232, 2002. [54] R. P. Stumpf, “Applications of satellite ocean color sensors for monitor- ing and predicting harmful algal blooms,” Human and Ecological Risk Assessment: an International Journal, vol. 7, no. 5, pp. 1363–1368, 2001. [55] K. Schmidt and A. Skidmore, “Spectral discrimination of vegetation types in a coastal wetland,” Remote sensing of Environment, vol. 85, no. 1, pp. 92–108, 2003. [56] L. Yuan and L. Zhang, “Identification of the spectral characteristics of submerged plant vallisneria spiralis,” Acta Ecologica Sinica, vol. 26, no. 4, pp. 1005–1010, 2006. [57] V. Aginsky, “Forensic examinaton of” slightly soluble” ink pigments using thin-layer chromatography,” Journal of Forensic Sciences, vol. 38, pp. 1131–1131, 1993. [58] S. J. Kim, F. Deng, and M. S. Brown, “Visual enhancement of old documents with hyperspectral imaging,” Pattern Recognition, vol. 44, no. 7, pp. 1461–1469, 2011. [59] G. Edelman, E. Gaston, T. Van Leeuwen, P. Cullen, and M. Aalders, “Hyperspectral imaging for non-contact analysis of forensic traces,” Forensic science international, vol. 223, no. 1, pp. 28–39, 2012. [60] A. Abbas, K. Khurshid, and F. Shafait, “Ink Mismatch Detection using Hyperspectral Document Image Analysis,” Master’s thesis, Institute of Space Technology, Islamabad, Pakistan, 2015. 49 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 59. A Numerical Studyon the Impact Resistance of Braided Composites X. Lei Dept. of Mechanical Eng., Hanyang University, South Korea K. Hayat Dept. of Mechanical Eng., University of Lahore, Lahore, Pakistan Email: khazar.hayat@me.uol.edu.pk M. T. Hussain Satellite Research and Development Center, Lahore, Pakistan H. T. Ali Dept. of Aerospace Eng., Queen's Building, University of Bristol, BS8 1TR, UK R. S. Choudhry Dept. of Mechanical Eng., National University of Science and Technology, Islamabad, Pakistan Abstract— Braided fabric reinforced composites are attractive alternatives to conventional prepreg laminates due to higher impact resistant performance. For the development of aerospace structures made of braided composites, the numerical simulations can play a key role in reducing time and cost. However, the numerical tools to evaluate the impact response and to investigate the damage mechanism of these materials, are not fully developed yet. This work focuses on the numerical modelling and analysis of the impact induced response of braided composites under different impact velocities, using micro- mechanics of failure (MMF) together with progressive damage models for fibre and matrix constituents. For implementation of MMF, the meso- and micro-scale models of unit cell have been utilized. In the meso-scale unit cell of the braided composites, the tows effective material properties were updated using a micro unit cell for which the degradation of constituent material properties was incorporated. The impact induced damage progression for the tows and pure matrix was then evaluated. The impact analyses were performed for bi-axial braided composites at 45 degrees and 25 degrees braiding angles to demonstrate the feasibility of the MMF-based approach. Moreover, the impact response of the braided composites with both thermosetting and thermos-plastic resin system was also investigated. Keywords— Braided composites, Impact resistance, Micro- mechanics of failure, Progressive damage model I. INTRODUCTION Recently, a rapid growth of the applications of textile braided fabric reinforced composites have been observed in the marine, aerospace and automobile industries etc. due to its net-shape fabrication and resistance to delamination damage and impact load in particular [1]. However, the behavior of the braided composites is not fully understood yet, and there is a need of establishing a methodology that can predict the material failure behaviors under different types of loadings, primarily the impact loading [2]. The purpose of this study is to evaluate the behavior of braided composites subjected to impact load with thermoplastic and thermosetting resin systems, based on various braiding angles and using micromechanics of failure (MMF) theory [3]. The analysis procedure involves three scopes: macro, meso and micro, as shown in Fig. 1. A braided composite can be assumed to be limited to mesoscale. It consists of tows and pure matrix encapsulating the tows. Thus, the use of mesoscale unit cell model serves as a bridge for transformation of stresses and effective material properties between the macro level and micro level analyses. II. MICROMECHANICS-BASED PROGRESSIVE DAMAGE MODEL A. Geometric modeling of braided composites The development of an idealized model of the braided composites is necessary to predict their material behaviors. The fiber bundle arrangements are modelled using the geometrical parameters (i.e. braiding angle, width and thickness of axial tow, width and thickness of bias tow, and gap between bias tows) illustrated in Fig. 2. Moreover, in the modeling, the undulation of tows plays a pivotal role in determining the behavior of the braided composites. Further details on modeling procedure and techniques can be found in [2]. Fig. 1. Analysis of braided composites. B. Meso and micro analyses The meso- and micro-level analyses are bridged to characterize the material behaviour of the tows of the braided Fig. 2. Modeling of braided composites. 50 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 60. composites. It shouldbe noted that a tow can be approximated as a curved continuous unidirectional (UD) lamina, and a micro unit cell, shown in Fig. 3c, can be used to relate the constituent behaviour to the tow behaviour. Thus, tow stresses can be transformed into each constituent stresses, and resulting constituent damages can be used to determine the tow damage. For the constituent damage analysis, the degradation of the constituent stiffness properties is carried out, and the tow effective material properties are subsequently updated through the micro unit cell. Fig. 3. Estimation of effective material properties of tows of braided composites. The transformation between the tows stresses at meso- level, denoted by , and the fiber and matrix stresses at micro-level, denoted by , and , is carried using the stress amplification factors (SAFs), denoted by and matrices, as following [3, 4]: (1) Eq. (1) can be written in detailed form, as following [4]: (2) In a similar manner, the meso strain and temperature increment can also be transformed to micro strains, as shown in Eq. 3, using the strain amplification factors which can be derived from SAFs using the constituent compliance matrix and the unit cell compliance matrix . (3) C. Damage model of micro-constituents For each constituent, a separate failure criterion is used due to their distinct mechanical behaviors. Considering the isotropic nature of the matrix constituent, the von Mises failure criterion [3, 5] is used, which follows: (4) where , , , and represents the first stress invariant, von Misses equivalent stress, tensile strength and compressive strengths of matrix constituent, respectively. The failure criterion described by Eq. 4 matches the condition that the Stassi’s equivalent stress , reaches the matrix tensile strength : (5) where the symbol represents the ratio of the compressive strength to the tensile strength of the matrix constituent. Further, the equivalent stress can be transformed into the equivalent strain of the matrix constituent, described in Eq. 6, using the stress-strain relations. (6) where , , and represents the first strain invariant, the von Mises equivalent strain and the Poisson’s ratio of the matrix constituent. It should be noted that the equivalent strain, like the equivalent stress, is a scalar quantity that is computed from the strain components. Fig. 4a represents a multi-linear stress- strain model defined using the equivalent stress and the equivalent strain, to estimate the matrix constituent damage. The matrix constituent behaves in a linear manner, before the occurring of any damage. Once damage occurs, the matrix constituent demonstrates a nonlinear behavior (i.e. hardening followed by the softening depending on the damage status). The stiffness of the matrix constituent is degraded by a factor as following: (7) subjected to the condition: . The symbols and denotes the matrix yield stress and the matrix yield strain at the beginning, and the symbols and denotes the matrix yield stress and matrix yield strain at the end of ith step. The symbol denotes the current equivalent strain of the matrix at the ith step, and the symbol denotes the matrix intact stiffness. For the mesoscale model, the damage in the pure matrix was modeled using Eq. 8 only, and additional MMF-based damage modeling using a micro unit cell was used for the matrix of tows [6]. The periodic boundary conditions were used for the meso and micro unit cells [2, 7]. The micro unit cell model, based on the MMF theory and linear material model, delivered multi-axial micro stress state for the fibre and the matrix constituents (i.e. assuming perfect bonding for interface). To avoid computation instability, the maximum local damage value for fiber and matrix constituent was set to 0.9 [8]. The mesoscale stresses were applied on the micro unit cell. After performing the finite element analysis, the micro stresses/strains in the matrix were computed, as shown in Fig. 5. Afterwards, the damage analysis for the matrix was performed. For the matrix of tow, the localization of damage 51 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 61. was avoided byusing a volume-based damage homogenization technique, represented by Eq. 8 shown below: (8) where the symbol denotes a positive weighing factor, and the symbol denotes the matrix overall volume in the micro unit cell. After estimation of damage status, the tow effective material properties in the meso unit cell model were re- evaluated from the micro unit cell and the matrix material properties were degraded. Afterwards, the damage propagation inside the tows and the pure matrix was evaluated. For the matrix constituent, following constitutive relation was used to degrade the stiffness: (9) where denotes the stiffness of matrix zone in the micro unit cell. For the fiber constituent, the maximum stress failure criterion, which is , was used for the fiber constituent. The symbol presents the micro longitudinal stress of the fiber constituent, and symbols and denotes the longitudinal tensile strength and longitudinal compressive strength of the fiber constituent, respectively. Fig. 4b illustrates that when the fiber constituent fails its stiffness is dramatically degraded. Since, carbon fiber behaves in a brittle manner, therefore, the highest value of all elemental damage factors was considered, as following: (10) where the symbol denotes the damage factor of jth element of the fiber constituent. And, the constitutive relation for fiber constituent was: (11) where denotes the stiffness of overall fiber zone in the micro unit cell. D. Numerical implementation Fig. 6 illustrates the flow chart of the algorithm developed for combining the MMF methodology and progressive damage model. In the beginning, the total global strain, denoted by , is computed at the global time , and augmenting the global strain increment, denoted by , to global strain at previous time-step n-1. For the tows, the macro stresses of each tow element, denoted by , are estimated using the previous effective stiffness properties, denoted by . Afterwards, the micro stresses for each element of the both matrix and fiber zones in the micro unit cell, denoted by and , are computed from the meso stress, denoted by , by using SAFs. The constituent failure criteria are then applied to both matrix and fiber, and damage factor is estimated for each jth element in the matrix and fiber zones, denoted by and , respectively, by employing the relevant constituent damage models. The overall damage factor for both matrix and fiber zones, denoted by and , are then evaluated based on their respective damage methods (i.e. the damage homogenization for the matrix constituent, and the maximum damage for fiber constituent, as discussed previously). Based on the status of overall damage factor, the stiffness properties of the matrix and fiber are degraded, and the tow effective properties are evaluated for the next time increment. The numerical implementation was done using Abaqus 6.11-1, a commercial finite element solver, combined with its user subroutine VUSDFLD [9]. Fig. 4. (a) Multi-linear model for the matrix constituent damage, and (b) linear model for the fiber constituent damage. Fig. 5. Damage estimation of the matrix constituent: (a) applied load; (b) micro stresses; (c) element damage factor; (d) overall damage factor. Fig. 6. Algorithm combining the MMF methodology and progressive damage model. 52 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 62. III. MATERIAL PROPERTIES Thestiffness properties of carbon fiber constituent are listed in TABLE 1 [1]. However, the strength properties of carbon fibers, shown in TABLE 1, were back-calculated using micromechanics approach. The stiffness and strength properties of the matrix constituent, listed in TABLE 2, were taken from in-house material testing database. TABLE 1. MATERIAL PROPERTIES OF CARBON FIBERS [1]. Material property Value Longitudinal modulus: Ef1 (GPa) 276.0 Transverse modulus: Ef2 (GPa) 27.6 In-plane shear modulus: Gf,12 (GPa) 138 Transverse modulus: Gf,23 (GPa) 7.8 Major Poisson’s ratio: vf,12 0.3 Major Poisson’s ratio: vf,23 0.8 Longitudinal tensile strength: Tf (MPa) 3800 Longitudinal compressive strength: Cf (MPa) 2980 TABLE 2. MATERIAL PROPERTIES OF THE MATRIX RESIN. Material property Thermosetting Thermoplastic Elastic modulus: Em (GPa) 3.45 2 Elastic Poisson’s ratio: vm 0.35 0.35 Final tensile strength: Tm (MPa) 65 65 TABLE 3. TOWS EFFECTIVE MATERIAL PROPERTIES. Material property Tows with thermosetting resin Tows with thermoplastic resin Longitudinal modulus: E1 (GPa) 215.67 215.67 Transverse modulus: E2 (GPa) 13.55 13.55 In-plane shear modulus: G12 (GPa) 9.0 9.0 Transverse modulus: G23 (GPa) 4.29 4.29 Major Poisson’s ratio: v12 0.309 0.309 Major Poisson’s ratio: v23 0.592 0.592 The tows effective properties, listed in TABLE 3, were computed from micro unit cell, using the stiffness properties of the matrix and fiber constituents, along with a fiber volume fraction of 0.78. It should be noted that the total fiber volume fraction of the meso unit cell was approximately 0.5, computed in accordance with [2]. The multi-linear damage models consisting of hardening and softening behaviors, shown in Fig. 7, for the matrix constituent (i.e. thermoplastic and thermosetting resin systems) was also established based on in-house test data. IV. IMPACT ANALYSIS SETTING For impact analysis, the whole model of biaxial (BX) braided fabrics was composed of fiber, pure matrix, matrix in tows, and a ballistic projectile. For the geometrical modeling parameters, the tow width and the two thickness used were 3.01 mm and 0.5 mm, respectively. There was a gap of 0.002 mm between the biased tows. The amplitude of the tow undulation of biaxial braids was set to 0.252 mm. Two kinds of biaxial braided composites with thermoplastic and thermosetting resin systems, and each having braided angle of 45 degrees and 25 degrees, hereafter denoted with BX45 and BX25, respectively, were investigated. For the ballistic projectile, the radius was 4 mm. The initial position of the projectile was set above the braided fabric model and the gap between the bottom of the projectile and the top surface of the braided models was set to 0.1 mm. Fixed boundary conditions were applied on the all four sides of the braided composites, and the top and bottom boundaries of the braided model were assigned with free boundary conditions. The ballistic projectile was taken as a rigid body controlled by a reference point, which was constrained to move only in z-direction. As illustrated in Fig. 8, the projectile was given an initial velocity, and the contact properties were also assigned on the braided model and the projectile by setting a value of 0.28 for the coefficient of friction. Once the projectile touches the top surface of the braided model, the contact behavior will be taken into account. Two initial velocities of 20 m/s and 50 m/s were applied to the projectile. At initial velocity of 20 m/s, no penetration occurred and projectile bounced back after impact. However, at initial velocity of 50 m/s, the projectile penetrated through the braided composite. V. RESULTS AND DISCUSSION Fig. 9 shows the damage contours for the BX45 braided composite models, with thermoplastic and thermosetting resins, at the impact velocity of 50 m/s. For the pure matrix and the matrix in tows, the damage contour with element deletion are shown in Fig. 9(a-c) and Fig. 9(b-d), respectively. Fig. 7. Stress-strain behaviors of the matrix resin systems. Fig. 8. Impact analysis of various types of biaxial braided composites: (a) braids at 45 degrees, and (b) braids at 25 degrees. 53 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 63. Only the top-sideof the damaged braided composite model, facing the incident projectile, is shown. Similar damage contours, along with element deletion, were also observed for the back-side of the braided composite model, and not shown for brevity. Based on the visual inspection of the damage contours, it can be qualitatively stated that the use of thermoplastic resin for the braided composite results in a better spreading of the impact damage, and lower element deletion. Similar qualitative observations were made for the BX25 braided composites subjected to a projectile with impact velocity of 20 m/s. In order to get a deeper insight into the impact behavior of the braided composite models with thermoplastic and thermosetting resins, the velocity time history of 0.2 seconds of the projectile was recorded. The velocity before the impact, is named initial velocity V and the velocity after impact is named residual velocity V’, as illustrated in Fig. 10. Fig. 11(a-b) shows that for BX45 and BX25 braided composite models, the impact velocity of 20 m/s of the projectile decreases with the passage of time. The impact velocity becomes zero at the time between 0.1-0.15 seconds, and becomes negative afterwards, representing the rebounding of the projectile after impact. At lower impact velocity, the projectile bounces back after hitting the braided composite model. At time 0.2 second, although the residual velocity of the projectile for the braided composite model with thermosetting resin is slightly lower than that of the braided composite model with thermoplastic resin, however, the velocity history curves for both resin systems almost overlap, making it very difficult to differentiate their impact resistance behavior, at lower impact velocity of 20 m/s. Fig. 9. Damaged matrix element deletion for BX45 braided composite at impact velocity of 50m/s: (a) pure thermoplastic matrix, (b) thermoplastic matrix in tows, (c) pure thermosetting matrix, and (d) thermoset resin in tows. Fig. 10. Definition of projectile velocity time histories, before and after impact. Fig. 11. Projectile velocity time histories for: (a) BX45 braided composite and impact velocity of 20 m/s, (b) BX25 braided composite and impact velocity of 20 m/s, (c) BX45 braided composite and impact velocity of 20 m/s, and (d) BX45 braided composite and impact velocity of 50 m/s. Dm Dm Dm Dm (a) (b) (c) (d) Fig. 12. Damage evolution for pure matrix: (a) BX45 braided composite and impact velocity of 20 m/s, (b) BX25 braided composite and impact velocity of 20 m/s, (c) BX45 braided composite and impact velocity of 20 m/s, and (d) BX45 braided composite and impact velocity of 50 m/s. Dm_tDm_t Dm_tDm_t (a) (b) (c) (d) Fig. 13. Damage evolution for the matrix in tows: (a) BX45 braided composite and impact velocity of 20 m/s, (b) BX25 braided composite and impact velocity of 20 m/s, (c) BX45 braided composite and impact velocity of 20 m/s, and (d) BX45 braided composite and impact velocity of 50 m/s. 54 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 64. On the contrary,the impact resistance of the braided composite model with thermoplastic and thermosetting resins, can be clearly distinguished at the projectile impact velocity of 50 m/s, as shown in Fig. 11(c-d). It should be noted that the projectile with a higher impact velocity of 50 m/s, penetrates through the braided composite model, as can be seen from the velocity time histories. The residual velocity of the projectile decreases with the passage of time, but remains positive throughout the recorded time of 0.2 seconds. For both BX45 and BX25 braided composite model with thermoplastic resin, the residual velocity of the projectile is clearly lower than that of the braided composite model with thermosetting resin. The decrease in residual velocity of the projectile means that the impact energy is absorbed by the braided composite model. Higher the decrease in the residual velocity of the projectile, the higher would be the impact energy absorption. Based on this fact, it can be inferred that the braided composite model with thermoplastic resin exhibits better impact resistance than that of the braided composite model with thermosetting resin. Fig. 12 and Fig. 13 show that the impact damage evolution for the pure matrix and for the matrix in tows, respectively, of the BX45 and BX25 braided composite models with thermoplastic and thermosetting resin systems. For the pure resin, it can be seen that the highest damage occurs for both BX45 and BX25 braided composite models with thermosetting resin (Fig. 12), for the impact velocities of 20 m/s and 50 m/s. TABLE 4. PERCENTAGE ELEMENT DELETION. Impact velocity (m/s) Percentage element deletion in (%) BX45 Braided composite BX25 Braided composite Thermo- plastic resin Thermo- setting resin Thermo- plastic resin Thermo- setting resin 20 Pure matrix 2.38 5.38 3.05 5.29 Tow matrix 0.09 0.14 0.18 0.31 50 Pure matrix 14.0 20.57 13.59 19.36 Tow matrix 1.89 2.04 1.71 1.84 However, in the case of the matrix in tows, the highest damage occurs for the BX45 and BX25 braided composite models with thermoplastic resin (Fig. 13), for impact velocities of 20 m/s and 50 m/s. It can also be clearly seen that the damage values of the matrix in the tows are higher in magnitude than that of the pure matrix. Thus, overall damage status of the braided composite model is governed by the matrix in tows. Moreover, higher impact damage values also mean that more absorption of the impact energy and higher reduction in the residual velocity of the projectile after impact, as previously illustrated in Fig. 11. Since, damage values of the matrix in tows for the braided composite model with thermoplastic resin are higher and well spread (as previous shown Fig. 9) than that of the braided composite model with thermosetting resin, therefore, it can be inferred that the braided composite with thermoplastic resin system can withstand higher impact loads, thereby, demonstrate better impact resistance. Finally, the percentage element deletion for the braided composite models with both thermoplastic and thermosetting resin are compared in TABLE 4. The percentage of element deletion of the pure resin and of the matrix in tows for the braided composite with thermoplastic resin system is lower than that for the braided composite with thermosetting resin system, as it withstand higher impact loads. Since, the braided composite with thermoplastic resin system possess better impact resistance, therefore, the small number of elements reach the critical damage values that once reach results in the element deletion. VI. CONCLUSIONS A numerical study was carried out to evaluate the impact resistance of the biaxial (BX) braided composites made of thermoplastic and thermosetting resin systems, and subjected to a circular ballistic projectile of radius 4 mm. Two impact conditions, the bouncing back of projectile with impact velocity of 20 m/s and the full perpetration of projectile at impact velocity of 50 m/s, were simulated. Based on the micromechanics of failure (MMF) theory, the impact resistance of the braided composite models was evaluated in terms of the damage evolution in the pure matrix and the matrix in the tows, and the percentage of element deletion. Results showed that the braided composites with thermoplastic resin demonstrate better impact resistance, than that of the braided composite with thermosetting resin, for both 45 degrees and 25 degrees braiding angles and for all impact conditions. From the numerical simulations, it was difficult to evaluate the effect of braiding angles on the impact resistance of the braided composites, and requires further investigation. Moreover, future work will involve the validation of the MMF-based impact analysis methodology, presented here, with the experimental data. ACKNOWLEDGMENT The authors are thankful to the their colleagues Professor Dr. Iqbal Hussain and Associate Professor Dr. Aamir Khan of Mechanical Department at the University of Lahore, Main campus, 1-kM Raiwind Road, Lahore, Pakistan, for their timely help and support. REFERENCES [1] Sun, X.S., V.B.C. Tan, and T.E. Tay, Micromechanics-based progressive failure analysis of fibre-reinforced composites with non- iterative element-failure method. Computers and Structures, 2011. 89(11-12): p. 1103-1116. [2] Xu, L., et al., Prediction of material properties of biaxial and triaxial braided textile composites. Journal of Composite Materials, 2012. 46(18): p. 2255-2270. [3] Ha, S.K., et al., Micromechanics of Failure for Ultimate Strength Predictions of Composite Laminates. Journal of Composite Materials, 2010. 44(20): p. 2347-2361. [4] Jin, K.-K., et al., Distribution of Micro Stresses and Interfacial Tractions in Unidirectional Composites. Journal of Composite Materials, 2008. 42(18): p. 1825-1849. [5] Raghava, R., R.M. Caddell, and G.S.Y. Yeh, The macroscopic yield behaviour of polymers. Journal of Materials Science, 1973. 8(2): p. 225- 232. 55 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 65. [6] Fish, J.and Q. Yu, Two-scale damage modeling of brittle composites. Composites Science and Technology, 2001. 61(15): p. 2215-2222. [7] Xia, Z., Y. Zhang, and F. Ellyin, A Unified Periodical Boundary Conditions for Representative Volume Elements of Composites and Applications. International Journal of Solids and Structures, 2003. 40(8): p. 1907-1921. [8] Hashimoto, M., et al., Prediction of tensile strength of discontinuous carbon fiber/polypropylene composite with fiber orientation distribution. Composites Part A: Applied Science and Manufacturing, 2012. 43(10): p. 1791-1799. [9] Roget, B. and I. Chopra, Wind-tunnel testing of rotor with individually controlled trailing-edge flaps for vibration reduction. Journal of Aircraft, 2008. 45(3): p. 868-879. 56 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 66. Improving Physical andMechanical Properties of Medium Density Fibreboard (MDF) Waheed Gul Mechanical Engineering Department CECOS University of IT & Emerging Sciences Peshawar, Pakistan waheed@cecos.edu.pk Afzal Khan, Abdul Shakoor Mechanical Engineering Department University of Engineering & Technology Peshawar, Pakistan Abstract— Physical and mechanical properties play a vital role in Medium Density Fiberboard (MDF) Manufacturing. Urea Formaldehyde (UF) resins is easily soluble in water and act as a binder with fibers. It has fast rate of reaction but has a poor physical and mechanical properties. Hence, there is a need to improve those properties of UF resin. The current research is based to investigates and analyze the UF resin to improve its physical and mechanical properties in manufacturing of MDF. The Melamine resin and basic green (Malachite Green) 4-crystals were introduced into UF resin, separately. In both cases, a significant improvement in physical and mechanical properties of UF resin was observed. Additionally, economic analysis was carried out for both additives and it was concluded that Melamine is not only economically feasible but also gives superior mechanical and physical properties when compared to Malachite green doped resin. The resulted MDF with Melamine doped resin resulted in strong mechanical adhesion, and water resistance Keywords—component; formatting; style; styling; insert (key words) I. INTRODUCTION The Medium Density Fiberboard (MDF) is a wood product processed under heat and pressure by using wood fibers or other plant fibers as raw materials and application of phenol or Urea-formaldehyde resin as binder. Other suitable additives may be added to improve the property of the board.[1].The density of MDF in production is generally control between 690 – 750 kg/m3.The performance index of MDF is divided into three categories, i.e. Physical performance, Mechanical performance and Biological performance. By nominal density, MDF may be classified in to Type 60, 70, 80. By applicable Conditions, MDF can be classified in to interior MDF, interior humidity-proof MDF and exterior MDF.By visual quality and internal bonding strength, MDF can be classified into Super Grade, Grade A and Grade B. [1]. A. MDF Production Process Flow MDF production process consists of different stages, i.e. materials preparation, Fiber separation, Fiber treatment, Mat forming, Hot pressing, Board trimming and sanding, as shown in MDF process flow chart in figure.1. ↓ ↓ ↓ ↓ ↓ Figure 1.1: Production Process flow of MDF [2]. B. Malachite Green Basic green (Malachite green) is an organic compound, basically green powder with Metallic luster. The molecular formula of Malachite green is C52H54N4O12. Properties of Malachite green are: It is easily soluble in water It is extremely soluble in Ethanol. Its solution in water is Blue-green. Basic green dyes become yellow in concentrated sulphuric acid. Figure 1.2: Basic Green 4 (Malachite green) Fiber Separation Hot Press Board Trimming Sanding Fiber Treatment Material Preparation Greeen Crystals 57 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 67. II. LITERATURE REVIEW UreaFormaldehyde (UF) resins is used as a binder with fibers in manufacturing of Medium Density Fiberboard (MDF) because of its fast rate of reaction, easily soluble in water and also economically feasible. [3].The only problem is that of poor water resistance property which limits the uses of MDF. In order to overcome this problem, a melamine formaldehyde resin and Malachite Green are added with UF resin. If we compare the market price of both resin, then it can be easily understood that Melamine formaldehyde resin is almost 4 times expensive than UF resin. So a very small quantity is to be added for improving water resistance property. The Melamine resin can easily soluble in both water and UF resin. [4]. similarly, the market value of Malachite green is pretty high. Malachite in our case is added to UF resin. The Malachite is added for the reason to give a unique color and extra strength to High Density Fiberboard. As we know that it is easily solvable in water. So the formulation of Malachite green is carried out in such a way that it dissolves water before adding it into UF resin. In other words the solution of Malachite green was added in definite proportion with water along with caustic. Malachite green is prepared in two steps: In ist steps condensation of of benzaldyde and dimethylniline condensate and leuco Malachite green in obtained. In second step leuco compound is oxidize to give malachite green. [5]. 𝐶6𝐻5𝐶𝐻𝑂+2𝐶6𝐻5𝑁(𝐶𝐻3)2→𝐶6𝐻5𝐶𝐻(𝐶6𝐻4𝑁(𝐶𝐻3)2) 2+𝐻2𝑂 (1) 𝐶6𝐻5𝐶𝐻(𝐶6𝐻4𝑁(𝐶𝐻3)2)2+𝐻𝐶𝑙+12𝑂2→[𝐶6𝐻5𝐶(𝐶6𝐻4𝑁(𝐶𝐻 3)2)2]𝐶𝑙+𝐻2𝑂 (2) Manganese dioxide is used as oxidizing agent. Malachite green can be Hydrolysis and give carbonyl formal. [𝐶6𝐻5𝐶(𝐶6𝐻4𝑁(𝐶𝐻3)2)2]𝐶𝑙+𝐻2𝑂 →𝐶6𝐻5𝐶(𝑂𝐻)(𝐶6𝐻4𝑁(𝐶𝐻3)2)2+𝐻𝐶𝑙 (3) Table 1: pH values of Malachite Green in Transition Uses of Malachite Green: Malachite green is used as a dye. It is used as antibacterial. In Microscopic analysis, it is used as a-biological stain. In dye laser it is used as a saturated absorber. It is also used as a PH indicator between PH 0.2 – 1.8. [5]. III. MATERIALS AND METHODS UF and Melamine formaldehyde resin and Malachite Urea Formaldehyde resin have some properties, i.e. PH value, Gel time, Viscosity, Specific gravity, Solid content etc. In order to calculate the above properties the following methods and procedures are used. A. pH Value Calculation There are two methods for calculating the PH value of the Resin. One way of measuring the PH value is the use of PH paper. Procedure: The PH strip is immersed in UF resin for a minute. Then the strip is taken out and compare the pH value with PH scale (0- 14) numbers. The PH value of UF resin is from 7.5 – 8.5 means that the UF is alkaline. Figure 3.1(a): A PH strip is immersed in Resin In second method, a PH meter is used to measure the PH (acidity or alkalinity) of UF. It consists of a glass electrode and electronic meter as shown in fig. The electrode is connected to the electronic meter that measure and display the PH value. pH IndicatorUF Resin Beaker 58 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 68. Figure 3.1(b): DigitalPH Meter Procedure: Before using the Ph meter, the electrode is washed with distilled water. Then the PH meter is calibrated with Buffer solution of PH 4, 7 or 10. The UF resin or Melamine resin or Malachite added resin,s sample is taken in a beaker and set it on room temperature, i.e. 25 Cº. The glass electrode is then dipped to resin sample. The PH value on the display is allowed to stabilize. After stabilizing the final reading is taken from display indicator. B. Specific Gravity Calculation The specific gravity of UF, Melamine or Malachite added resin is measured by an apparatus called Hydrometer. Hydrometer consists of a graduated tube and a bulb at its end as shown in fig.3.2. Procedure: The resin is taken in a graduated cylinder and adjusted its temperature to 25 Cº. A hydrometer is inserted in the graduated cylinder in such a manner that the bulb of the meter is dipped down. The mercury is raised in the meter and after stabilizing it on graduated cylinder, the reading is taken. Figure 3.2: Hydro Meter C. Gel Time Calculation The gel time of the resin is the time period in which it becomes too viscous to flow. The gel time has a very important role in MDF manufacturing. The gel time measurement is illustrated below. Figure 3.3: Gel Time Measurement Apparatus Procedure: Take 50 ml UF resin in a beaker. Now add 1.4 ml of hardener (Ammonium Chloride) (NH4Cl) to UF resin and mix it. Now take a sample of mix UF resin in a graduated cylinder and put it in water bath at 100 ⁰ C) and on stop watch. Stirrer the resin till it becomes solid and stops the stop watch. Now note down the time from stop watch. D. Viscocity Calculation Viscosity is the resistance of a liquid to flow. The viscosity of UF, MUF or Malachite added resin is generally measure with viscosity cup as shown in fig.5. The viscosity cup has an orifice of 3mm (ISO standard) at the bottom. Procedure: A known volume of UF resin is passed through this orifice and note down the time flow through stop watch. The viscosity of UF resin is between 200 – 300 cps (centi poise). Electrode Flexible Arms Display Screen Beaker Graduatede Cylinder Hydrometer Grip HolderStirrer Beaker Hot Plate 59 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 69. Figure 3.4: ViscosityCup E. Solid Content Calculation The solid content of UF, MUF or Malachite added resin is the remaining percentage after its drying. Procedure: take one gram of UF, Melamine or Malachite added resin and put it in an aluminum foil cup. Now weight it. This will be “Wt”. Now put the aluminum foil cup fill with UF or Melamine resin. in oven at 105 ⁰C) for 2 hours. After 2 hours take out the foil cup and find its weight. This will be W2. Figure 3.5: Solid content Measurement Apparatus Then using the following formula find out the solid content: Solid content (%) = Wt – W1 / Wt x 100. F. Thickness in Swelling Calculation Thickness swelling is the difference of initial specimen thickness to the variation of final thickness after water soak test in percentage. Figure 3.6: MDF pieces for Swelling in Thickness Procedure: Take a piece of MDF specimen 50 mm X 50 mm and measure its thickness, i.e. T1. Now immersed the sample of MDF in cold water for 24 hours. The samples are then taken out from water tray and find its thickness, i.e. T2. The swelling in thickness (TS) is calculated by the following formula: Ts (%) = T1 – T2 / T1 x 100 IV. EXPERIMENTS AND RESULTS A. UF Resin Properties UF resin properties were determined using the above procedures and the properties were then tabulated. Table 4.1: Properties of UF Resin S. No Properties Values 1 Color Transparent/milky white 2 PH 8.5 3 Gel Time 100 sec 4 Solid content 60 % 5 Specific Gravity 1.26 – 1.28 kg/m³ 6 Viscosity 200 – 300 cps B. Melamine Resin Properties Similarly, Melamine resin properties, i.e. color, PH value, Gel time, solid content, specific gravity and viscosity were determined and put in the form of table as illustrated below. Table 4.2: Properties of Melamine Resin S. No Properties Value 1 Color Transparent 2 PH 7.0 3 Gel Time 195 sec 4 Solid content 66 % 5 Specific Gravity 1.2 – 1.24 kg/m³ 6 Viscosity 125 – 150 cps C. Melamine Urea Formaldehyde Resin Properties As earlier discussed that UF has poor water resistance. In order to increase its water resistance properties, 20 gm of Melamine Vescocity Cup Stand Base Aluminum Foil Weighing Scale Display Indicator Pieces of MDF 60 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 70. resin was addedin 1000 gm of UF resin. The mixture of both resin were poured into a beaker and shacked it for 5 minutes. The mixture of resins is then used for determining the following properties. Table 4.3: Properties of Melamine Urea Formaldehyde Resin S. No Properties Value 1 Color Transparent/Milky white 2 PH 6.7 3 Gel Time 115 sec 4 Solid content 60 % 5 Specific Gravity 1.26 – kg/m³ 6 Viscosity 220 cps D. Formulation of Melachite Green Crystal In order to prepare the solution of Malachite green, the following composition of ingredients were made. Table 4.4: Composition of Malachite green solution S. No Ingredients Quantity Unit 1 Water (H2O) 80 Kg 2 Malachite green 6 Kg 3 Caustic (NaoH) 75 gm Samples of the above solution were tests in laboratory for its PH value. The PH thus calculated through Ph meter was 3.2. This value indicates that the solution is highly acidic. But as we know that PH value of UF resin is 8.5. The Malachite green solution was added in UF resin in such a proportion that its PH value maintain between 6 and 6.5. Table 4.5: Composition of Malachite green Urea Formaldehyde Resin S. No Properties Value 1 Color Blue Green 2 PH 6 3 Gel Time 80 sec 4 Solid content 61 % 5 Specific Gravity 1.27 kg/m³ 6 Viscosity 250 cps For comparison and improved water resistance properties, the experiments were performed on 16mm boards. One was manufactured using only 10 % Malachite UF resin and the other was manufactured using 10 % Melamine UF resin. The other parameters were kept constant in both cases. The constant parameters are shown in table below. Table 4.6: Constant parameters for 16 mm board S. No Parameters Value Units 1 Moisture content 9 % 3 Hot press temp 190 ⁰C 4 Pres closing time 35 Sec 5 Thickness of board 16 Mm 6 Total press cycle 320 Sec 7 Wax 0.9 % Table: 4.7 Comparison of 16mm board Properties (UF) Table: 4.8 Comparison of 16mm board properties With standard value using MUF Resin S. No Properties Value s Units EN standard s Test Method 1 Moisture content 8.7 % 7 ± 3 EN322 2 Density 725 kg/m3 720-740 EN323 3 Swelling in thickness 11 % ≤ 12 EN317 4 Internal Bonding 0.73 N/mm² ≥ 0.6 EN319 5 Modulus of Elasticity 3190 N/mm² 2500 EN310 6 Modulus of Rupture 32.5 N/mm² ≥ 30 EN310 7 Screw Holding face 1040 N ≥ 1000 ASTM1761 8 Screw Holding edge 750 N ≥ 700 ASTM1761 S. No Properties Value s Units EN standard s Test Method 1 Moisture content 7.5 % 7 ± 3 EN322 2 Density 730 kg/m3 720-740 EN323 3 Swelling in thickness 15 % ≤ 12 EN317 4 Internal Bonding 0.65 N/mm² ≥ 0.6 EN319 5 Modulus of Elasticity 3100 N/mm² 2500 EN310 6 Modulus of Rupture 31 N/mm² ≥ 30 EN310 7 Screw Holding face 1015 N ≥ 1000 ASTM1761 8 Screw Holding edge 741 N ≥ 700 ASTM1761 61 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 71. Table: 4.9: Comparisonof 16mm board properties with standard values Malachite Figure 4.1: Comparasion of swelling in thickness of UF,MUF Malachite UF with standard value Figure 4.2: Comparison of internal Bonding of UF, MUF Malachite UF with standard value V. ECONOMIC ANALYSIS As we know from formulation that: Melamine consumption /sheet = 0.19 (kg/sheet) Price of 1 Kg of Melamine = Rs. 150/Kg Melamine cost per sheet = Rs = 28.5 Similarly, Malachite Green consumption / sheet = 0.07 (Kg/sheet) Price of 1 Kg of Malachite Green = Rs = 700 / Kg Malachite cost per sheet = Rs = 49 Cost Difference = Rs. 20.5 For a 154 M³ capacity plant, Production/day = 3240 sheets (16mm) Cost saving /day = Rs. 66420 So it is clear from economic Analysis that Melamine is also economically better than Malachite Green. VI. CONCLUSION In the light of the above parameters, samples of 16mm board were manufactured using the predefined parameters. The boards were then tested for their physical and mechanical properties. The tested boards’ properties were then compared with the international standards. The 16mm boards samples using Malachite UF and MUF resins were manufactured and properties were compared and it was found that almost all properties were according to the standards. Moreover, using MUF, the swelling in thickness property was improved by 15%. And Mechanical adhesion (which is called internal bond) was increased by 5 %. Economic Analysis was also carried out and it was analyzed that Melamine is economically better resin. ACKNOWLEDGMENT The author is highly thankful to Ciel Woodworks Pvt.ltd Peshawar, Pakistan for giving him the opportunity to perform experiments in lab and use the factory equipment. REFERENCES [1] ksh, A, and Hosseikhani, H, (2010), “Investigation on physical and mechanical properties of Medium Density Fiberboard (MDF) produced from hornbeam wood”, Volume 25, Number 1 (32); Page(s) 1 To10. [2] Zwawi Ibrahim, Astimar Abdul Aziz, Ridzuan Ramli, Wan Hassamudin Wan Hassan and Nahrul Hayawin Zianal (2011), “Optimum Parameters for the Production of MDF using 100 % Oil Palm Trunks”, “Malaysian Palm Oil Board, Ministry of Plantation industries and commodities, Malaysia.ISSN, ppt.1511-7871. [3] Awang Bono, Yeo Kiam Beng @ Abdul Noor & Kinabalu, Sabah, (2001),”Melamine-Urea-Formaldehyde Resin: changes in Physical Properties and Strength with Composition Molar Ratio”, Borneo Science 10: 11-23. [4] Jizhi Zhang, Xiaomei, Wang, Shiefeng Zhang, Qiang Gao, Jianzhang Li ,(2013),”Effects of Melamine Addition stage on the Performance and Curing Behavior of Melamine-Urea-Formaldehyde (MFU) Resin,”,Bejing Forestry University, 100083, Vol 8, No 4. [5] H. Zare-Hosseinabadi, M. Faezipour1, A. Jahan-Latibari2 andA.Enayati1, J. Agric. Sci. Techno, (2008),”Properties of Medium S.N o Properties Value s Units EN standard s Test Method 1 Moisture content 8 % 7 ± 3 EN322 2 Density 725 kg/m3 720-740 EN323 3 Swelling in thickness 12.8 % ≤ 12 EN317 4 Internal Bonding 0.7 N/mm² ≥ 0.6 EN319 5 Modulus of Elasticity 3150 N/mm² 2500 EN310 6 Modulus of Rupture 31.9 N/mm² ≥ 30 EN310 7 Screw Holding face 1001 N ≥ 1000 ASTM1761 8 Screw Holding edge 739 N ≥ 700 ASTM1761 62 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 72. Density Fiberboard Madefrom Wet and Dry Stored Bagasse”, Vol. 10: 461-470. [6] Amee K. Patel, Hardik H. Chaudhary, Khushbu S. Patel and Prof. Dr. Dhrubo Jyoti Sen,(2014),” Colour of Ecofriendly Dyes Used In Holi Rather Than Triphenylmethane Dyes”, Volume 3, Issue9, 1287-1305. [7] Nadir Ayrilmis, (2008),”Effect of Comparison wood on Dimensions Stability of Medium Density Fiberboard”, University of Istanbul Turkey, 42(2),285-293. [8] Cristiane Inácio de CamposI, Francisco Antonio Rocco LahrII, (2004),”Production and characterization of MDF using eucalyptus fibers and castor oil-based polyurethane resin”, vol.7 no.3 [9] Stefan Ganev, (2003),”Linear Expansion and thickness Swell of MDF as a Function of Panel Density andSorptionState,””,University Laval,Quebic,Canada,GIK7P4. [10] H. Zare-Hosseinabadi, M. Faezipour1, A. Jahan-Latibari2 andA.Enayati1, J. Agric. Sci. Techno, (2008),”Properties of Medium Density Fiberboard Made from Wet and Dry Stored Bagasse”, Vol. 10: 461-470. [11] Khalid Pervez Bhatti and Muhammad Zuber, (2009),” Synthesis and Application of Melamine Urea Based Precondestates”, AUTEX Research Journal, Vol. 9, No4. [12] Maintains, G and Berns, J, (2001),”Strawboards bonded with urea formaldehyde resins”, 35th International Particleboard Composite Materials Symposium Proceedings, pp.137-144, 2001. [13] http://www.goodrichsugar.com/mdfnew.asp?no=3, (April12, 2014. 63 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 73. Finite Element Analysisof Tool Wear in Ultrasonically Assisted Turning A. Rehman1 , S.Maqsood1 1 Dept.of Industrial Engineering, University of Engineering and Technology, Peshawar,Pakistan Khattak_rehman@yahoo.com R.Muhammad2 , N.Ahmed2 2 Dept.of Mechanical Engineering CECOS University of IT and Emerging Sciences Peshawar,Pakistan Abstract—The tremendous increase in use of titanium and its alloys in different industries, especially in power and aviation industries demand further investigation due to its exceptional mechanical and thermal properties. New assisted and hybrid machining techniques were introduced to obtain better results in machiningof these alloys. Among many, ultrasonically assisted turning (UAT) is one of those in which lowervibration energy impactisdelivered onthe cutting tool in the direction of velocity. DuringUAT of titanium alloys, momentous improvement has been noticed in surface roughness smoothness along with reduction of overall cutting forces, but tool wear behavior is still unknown. This paper presents a parametric finite element (FE) analysis for describing tool wear behavior in UAT of (Ti- 15V3Al3Cr3Sn) alloy on the selected cutting conditions. The model is useful for optimizing of cutting parameters in UAT for Ti-15V3Al3Cr3Sn. The proposed FE model is also enhances the recent experimental work on wear behavior relevant to cemented carbide tool and comparing UAT and conventional turning (CT) of Ti-15V3Al3Cr3Sn results. IndexTerms—Tool Wear, Ultrasonically Assisted Turning, Ti Alloys I. INTRODUCTION Machining is an imperative practice in manufacturing and production industries. Machining put in a substantial portion to the total cost of the product. Machining cost and production time are the key apprehensions caused by the frequent abrasion and replacement of machining tool. Compare to milling, drilling[1, 2]and grinding processes, turning process is receiving more importance due to enhanced production rate in advance industries[3, 4]. Uses of Titanium and its alloys are tremendously increasing day by day in different industries, especially in power, petroleum, aviation and biomedical industries[5]. In many applications, these materials replace steels and aluminum alloys, which usually results in weight and/or space saving, increase of system efficiency by rising the service temperature, and remove the need of protective coatings that should be used in steels [6]. Increment in the use of these alloys is due to its outstanding mechanical properties including elevated strength to density ratio, corrosion resistance, fatigue properties and strength behavior at moderately high temperatures [7]. Aforesaid materials are therefore fall in a category of difficult- to-cut materials, because of the fact that these materials are facing greater challenge to machining processes against the higher stresses generation and elevated temperatures[8].Presently, beta titanium is considered as the most difficult to machine alloys[9].Keeping in view the high demand of titanium and its alloys machining, researchers have continuously worked and developed different techniques for improvement of machinability of these most hard-to-cut alloys[10]. Such alloys are difficult to machine using conventional turning(CT) processes and often results in poor surface finish with associated lack of dimensional accuracy, built-edge formations, fast tool wear, and undesired chatter during operations[11]. In recent, new assisted and hybrid machining techniques have been introduced to get improved machining results of these hard-to-cut alloys. Among many, ultrasonically assisted turning (UAT) is one of those in which lowervibrationalenergy impacts are imposed on the tool in the direction of velocity[12-14]. This technique has shown tremendous and noteworthyenhancement in the machinability of hard-to-cut alloys [15-18]. In UAT, the tool and workpiece are not in continuous contact that is why net forces are lower as compared to CT [12], ultimately friction influence are also lower [13, 14]. The efficiency of UAT and influence of vibrational parameters on surface roughness is investigated by Graeviciuteet. Al[15, 16]. However, limited attention has been given to investigate the tool life in UAT which is the main concern in manufacturing processes. Two of the main critical problems are tool wear and tool failure, which not only raises the production cost but also affect the product quality. The thorough and continuous contact between cutting part of tool and workpiece material generate high cutting forces and temperature in the cutting region of hard-to-cut alloys. The selection of appropriate tools for the machining of these alloys using sustainable conventional machining processes is still not fully understood and demands for further research to advance the better tool life and tool wear. The subject matter of this work is based on the finite element simulation of UAT of beta titanium alloy to investigate the tool life under pre-selected cutting conditions. The simulation results have been compared with CT. This work would help the industries to estimate tool’s life in UAT. 64 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 74. II. 3D MODELLING A3D model for both CT and UAT processes is developed in DEFORM 3D V 6.1.In current model cutting tool of Fig.1. Tool and workpiece FE mesh Tungsten Carbide with (AlTiN) coating is assumed to be elastoplastic with considering a rake angle as 14.6º, nose angle as55 º, nose radius as0.8 mm and cutting-edge radius as0.8 mm. Tool is then meshed with more than 45,000 elements and is oriented according to the setup proposed. Meshing of tool and workpiece can be seen in Fig. 1. Workpiece of Ti-15333is assumed to be plastic and is fully constrained on lower and lateral side so that it cannot move. Workpiece used in the simulation is5mm long, fine meshing been considered with the minimum element size of 0.025 mm and the number of elements aremore than 30,000. Localized window meshing has beenselected for the workpiece meshing to obtain better results. 0.3 mm depth of cut and 0.1 mm/revfeed rate are initially considered for all simulations. The current model is based on Langrangian methodin which the chip is formed by continuous remeshing. The tool is immovable for the simulation cases of CT, whereas it vibrates harmonically for the simulations of UAT. The tool has given an ultrasonic vibration with 20 kHz frequencyand 8µm of amplitude in the direction of cutting velocity.Figure. 2 clearly elaborate theschematic of tool and workpiece movements. Fig.2. Schematic of tool and workpiece movements TABLE.1 Properties of cutting tool Young’s modulus, (Mpa) 560*10-3 Thermal conductivity, (Wm-1°C-1) 0.0081*T+11.95 Thermal expansion, (mm mm-1°C-1) 9.4*10-6 Heat capacity, (Nmm-2C-1) 0.0003*T+0.57 Poisson's ratio 0.25 A. Cutting Tool In the present paper, coated carbide insert (AlTiN) is being used as a tool to machine titanium alloy (Ti-15333). This coating provides reduction of fraction and adhesion between chip and cutting tool. In addition to that, AlTiNat high temperature has good oxidation resistance, wear resistance and elevatedchemical stability because of the formation of Al2O3 film. Cutting tool mechanical properties are represented in Table.1. [18] B. Workpiece Material Titanium alloy (Ti-15333) used in our simulation belongs to the family of beta Ti alloys.These alloys fall in category of difficult to machine titanium alloys because of the significant precipitation hardening characteristics. Table.2 and Table. 3[19] shows the composition along with the mechanical properties of Ti -15333 . TABLE.2 Ti -15333 (%) chemical composition Ti V Al Cr Sn 76 15 3 3 3 TABLE.3 Ti-15333Material properties Parameter Unit value Solution treated and aged Density, ρ (kg/m3 ) 4900 Young’s modulus, E (GPa) 87 Tensile strength, UTS (Mpa) 1200 Thermal conductivity, k (W/Km) 8.08 Hardness (Rockwell B) 95 65 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 75. TABLE.4 Cutting conditionsdetails Experimental conditions Workpiece Ti-15333 , length 5mm Cutting tool TiAlN Cutting speed,(m/min) 10, 20, 30, 40, 50,60 Feed rate, 0.1 mm/rev Depth of cut, 0.3 mm Vibration parameters 20 kHz, 8µm Dry cutting C. Cutting Conditions Finite element E simulations have been carried out for overall 12 cases of experiments for both CT and UAT, consisting of 6 different cutting speeds while keeping depth of cut and feed rate constant. Table.4 presents simulation matrix for tool wear analysis. III. FINITE ELEMENT ANALYSIS In this section the results of finite element simulations are conferredand deliberated.Resultant values of stresses and temperatures are studied which can eventually affected the tool wear and tool life. Fig. 3a, 3b and 3c shows effective stresses on UAT and CT processes respectively.We can see theaverage effective stresses on tool are comparatively lower in UAT than CTfor one complete vibrational cycle. But on the other side at penetration stage, high stresses are noted. Fig. 3c shows the stresses on the tool during retraction time.This concludes that at penetration stage in UAT, the tool wear rate is high. A. Stresses on tool In case of CT, the tool is continuouslyin connectionwith the workpiece whereas in UAT the cutting process can be divided into four main stages in each single cycle due to the ultrasonic vibration setup. They are listed as approach, chip contact, penetration and unloading. In present discussion we are assuming the tool as at penetration stage and at retraction stage in each vibration cycle. The number of vibrations cycles depends on the total simulation time and also on frequency and amplitude as well. We are considering effective stresses against the time domain for our analysis of tool wear because of the ultrasonic vibration cycles and stress fluctuations in UAT. Initially, in UAT,when the cutting tool approaches the chip, it generates smaller effective stresses of 190 MPawhich later on increases with an increase oftool and chip contact. The higher Fig. 3a Stresses-effective in CT Fig. 3b Stresses-effective in UAT at penetration stage Fig. 3c Stresses-effective in UAT retraction stage stressesof 3140 MPa are noted at the penetration stage of the tool, which is much more than 3090 Mpa that in CT, see Fig. 3a and 3b. On the other side, at retraction stage, the effective stresses again reduce to the lower value of 2540 MPa. It can be concluded that the tool is facing higher stresses at the 66 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 76. contact point, resultinghigher tool wear and reduction in tool life as comparable to that in case of CT. The stresses distribution on cutting toolalong the rake face and the flank face directions in UAT and CT are shown in Fig. 4 B. Tool Temperature During turning processes an elevated temperature normally generated in the cutting regiondue tothe plastic deformation and high friction of tool and workpiece contact. It is also a fact that high temperature is considered the most important factor for tool wear and tool life in machining processes. Therefore evaluation of temperature is an important characteristic for tool wear and tool life during CT and UAT. Comparative analysis of temperature distribution on cutting tool for both UAT and CT are shownin Fig. 5. The simulation result enhances the previous work carried out on temperature effects on cutting tool in CT and UAT. Higher temperatureshave been noticed at the penetration stage in UAT.This raising in cutting temperature directly affects the tool coating regardingthermal softening like reduction in yield stress,thermal conductivity, Young’s modulus, coefficient of thermal expansion and specific heat. As the cooling time is very short and vibration to the tool also increase its temperature, which further suffice the wear of the tool. It is also noticed thatif thecutting speedhas increasedfrom 10 m/min to high speed value, the resultant temperature also increased in UAT. This temperature increment is due to the decrease in retraction time. The same pattern has been noticed for other cutting speeds i.e. 30 m/min., 40 m/min, 50 m/min and 60 m/min. From the above discussions it can be urge that tool wear is increases in UAT with the temperature increment at different speed points. IV. CONCLUSIONS A 3D finite element model has been developed in DEFORM 3D V 6.1 for the simulation of Ti-15333 in CT and UAT. The results shows noticeable increase in cutting stresses in UAT as compare to that in CT. At the penetration stage there is much higher difference in the effective stresses on the cutting tool, but at retraction stage considerable lower stresses are noticed. State variable distributions also show the notable difference in the effective stresses on cutting edge of the tools in cases of CT and UAT processes. Further, as increase in speed from lower to higher value the contact between tool and workpiece reduces in UAT, resulting high temperatures in cutting zone, which affected the tool wear and tool life by increasing the wear in UAT as compared to CT Fig. 4 Effective stresses in cutting and radial direction in both CT and UAT Fig. 5 Temperature details in CT and UAT V. REFERENCES 1. Muhammad, R., et al. Finite-element analysis of forces in drilling of Ti-alloys at elevated temperature. in Solid State Phenomena. 2012. Trans Tech Publ. 2. Muhammad, R., et al., 3D modeling of drilling process of AISI 1010 steel. Journal of Machining and Forming Technologies ISSN, 2010. 1947: p. 4369. 3. Muhammad, R., et al., Thermally enhanced ultrasonically assisted machining of Ti alloy. CIRP Journal of Manufacturing Science and Technology, 2014. 7(2): p. 159-167. 0 500 1000 1500 2000 2500 0 0.5 1 1.5 Stresses-Effective(MPa) Distance (mm) Stresses along the rake face in CT Stresses along the flank face in UAT Stressesalong the flank face in CT Stresses along the rake face in UAT 80 150 220 290 360 430 500 570 0 10 20 30 40 50 60 Temperature(ºC) Speed (m/min) Tool Temperature (°C) in UAT Tool Temperature (°C) in CT 67 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 77. 4. Maurotto, A.,et al., Ti Alloy with Enhanced Machinability in UAT Turning. Metallurgical and Materials Transactions A, 2014. 45(6): p. 2768-2775. 5. Arrazola, P.-J., et al., Machinability of titanium alloys (Ti6Al4V and Ti555. 3). Journal of Materials Processing Technology, 2009. 209(5): p. 2223-2230. 6. Veiga, C., J. Davim, and A. Loureiro, Review on machinability of titanium alloys: the process perspective. Reviews on Advanced Materials Science, 2013. 34(2): p. 148-164. 7. Ezugwu, E., High speed machining of aero-engine alloys. Journal of the Brazilian society of mechanical sciences and engineering, 2004. 26(1): p. 1-11. 8. Dandekar, C.R., Y.C. Shin, and J. Barnes, Machinability improvement of titanium alloy (Ti– 6Al–4V) via LAM and hybrid machining. International Journal of Machine Tools and Manufacture, 2010. 50(2): p. 174-182. 9. Zlatin, N. and M. Field, Procedures and precautions in machining titanium alloys, in Titanium Science and Technology. 1973, Springer. p. 489-504. 10. Muhammad, R., et al. Effect of cutting conditions on temperature generated in drilling process: A FEA approach. in Advanced Materials Research. 2011. Trans Tech Publ. 11. Maurotto, A., et al. Recent developments in ultrasonically assisted machining of advanced alloys. in Proceedings of the 4th CIRP International Conference on High Performance Cutting (HPC2010). 2010. 12. Kumabe, J., et al., Ultrasonic superposition vibration cutting of ceramics. Precision Engineering, 1989. 11(2): p. 71-77. 13. Muhammad, R., et al., Analysis of a free machining α+ β titanium alloy using conventional and ultrasonically assisted turning. Journal of Materials Processing Technology, 2014. 214(4): p. 906-915. 14. Muhammad, R., A. Roy, and V.V. Silberschmidt, Finite element modelling of conventional and hybrid oblique turning processes of titanium alloy. Procedia CIRP, 2013. 8: p. 510-515. 15. Kim, G.D. and B.G. Loh, Characteristics of elliptical vibration cutting in micro-V grooving with variations in the elliptical cutting locus and excitation frequency. Journal of micromechanics and microengineering, 2008. 18(2): p. 025002. 16. Bouchelaghem, H., et al., Wear behaviour of CBN tool when turning hardened AISI D3 steel. Mechanika, 2007. 65(3). 17. Fnides, B., M. Yallese, and H. Aouici, Hard turning of hot work steel AISI H11: Evaluation of cutting pressures, resulting force and temperature. Mechanika, 2008. 4(72): p. 59-73. 18. Skiedraitė, I., et al., Ultrasonic application in turning process of different types of metals. 19. Maurotto, A., et al., Comparing machinability of Ti- 15-3-3-3 and Ni-625 alloys in UAT. Procedia CIRP, 2012. 1: p. 330-335. 68 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 78. Metamaterials in AerospaceIndustry: Recent Advances and Prospects Rabia Zafar, Saima Shabbir Institute of Space Technology, Islamabad, Pakistan drsaimashabbir@gmail.com rabiazafar38@gmail.com Azeem Zafar National University of Science and Technology, Islamabad, Pakistan azeemgeologist@gmail.com Abstract— This article offers a review of the existing metamaterials and their concepts, recent advances, and research directions. It also describes applications of metamaterials design in aerospace and defense sector as an emerging technology. Acoustic metamaterials and metamaterial based antenna present advantages over conventional noise absorbers and antenna technology employed in aerospace components. Gain enhancement of antennas and efficiency of structures are described in this article with reference to improvements and advancements made so far in academia. A rational approach to metamaterials science and engineering can revolutionize aerospace design and render better prospects for military applications. Keywords—Acoustics, Aerospace, Metamaterials I. INTRODUCTION For the past few decades research in metamaterials is gaining an increasing interest [1, 2] due to counterintuitive electromagnetic properties incorporated into material structure. Principally, Metamaterials are left handed materials which exhibit negative refractive index, unusual diamagnetic properties, with fractional effective permeability and a very large in-plane effective permittivity [3, 4]. The permeability can be attributed to the Eddy current phenomenon induced in conductive inclusions and permittivity is attributed to an increased value of dielectric constant. For a material to be optically transparent, the incident wavelength will be refracted off centre the superficial surface. The angle of refraction is smaller than angle of incident as demonstrated in Fig. 1. The velocity of incident wavelength is greater in space than the velocity of transmitted wavelength [5] and object will appear transparent. Metamaterials find immense applications in the field of science and engineering including wireless communication, bio sensing, seismology, electromagnetic devices and perfect lenses [6]. Moreover, metamaterial absorbers and emitters, super lenses and cloaking devices for military camouflage are under high consideration. A. Development history Developmental history of metamaterials began by 20th century in 1904, when experiments on negative group velocities were carried out by Pocklington [7] and progressed through the mid of 20th century contributed by distinguished scientists [8, 9]. John Pendry in 1990’s made significant contributions in the development of metamaterials theory [6]. Metamaterials do not exist naturally with the exception of metallic element Bismuth. Plasmas and noble metals show this behavior in infra red or visible spectrum [10]. B. Existing metamaterials Metamaterials cover a narrow range of wavelength owing to the fact that they can respond below red region of visible wavelength. Based upon the medium properties, metamaterials can be classified as; • Negative index metamaterials (NIM) which include double negative (DNG- ε and µ are negative) and single negative metamaterials (SNG-either ε or µ is negative), • Double positive (DPS- ε and µ are positive), • Epsilon negative (ENG- ε is negative and µ is positive) and µ negative (MNG ε is positive and µ is negative). MNGs show magneto optic effect in which electromagnetic wave propagation takes place in a quazistatic altered magnetic field [11]. Based on the response to a particular wavelength metamaterials are further classified as photonic metamaterials, acoustic metamaterials, seismic metamaterials, electromagnetic metamaterials, terahertz metamaterials, tunable metamaterials, elastic metamaterials, chiral metamaterials and quantum metamaterials. 69 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 79. Fig. 1. Raydiagram of a metamaterial and a conventional material II. MICRO FABRICATION AND APPLICATIONS Metamaterials industry is emerging as a leading technology in future for defense and aerospace sectors by 2020. Currently, there are only few methods available for the fabrication and characterization which involve complex principles of microelectronics and lithographic deposition. The characterization of metamaterials is rather expensive, which poses high need for capital investment in instrumentation. Due to these limitations there are only a few metamaterials which show negative effective permeability and large permittivity and successfully characterized as metamaterials. A. Polymer based metamaterials Metamaterials are composed of polymer matrix filled with, copper, gold or silver flakes. They are formed by stacking alternating layers of ENGs/MNGs type of metamaterials separated by space. Metamaterials for interference shielding applications were fabricated by pulsed current electrolysis routes and by carefully controlling the concentration of copper metal flakes as inclusions in polymer matrix composites [12]. B. Metamaterial coatings Metamaterials multipole coatings are being applied by carefully tailoring the surface properties ranging from the lowest ordered multipole coatings for microscopic bodies to higher ordered multipoles for macroscopic bodies. Concealing a macroscopic body becomes more difficult as domain size increases and concealing the particles in a closed proximity is difficult to achieve practically. C. Metamaterial emitters In a single band metamaterial a noble metal such as gold or silver is deposited by electron beam lithography on a silicon substrate followed by curing the spun coated benzocyclobutene on metallic top. Dual band absorber is fabricated by lithographic deposition with only difference that the dielectric space is occupied by Al2O3. Such emitters find applications in energy harvesting devices [13]. D. Metamaterials and seismic cloak The seismic wave guided metamaterials can be made by stacking sequences of macromolecular polymer based split ring resonators overlaid with other meta atoms in a consecutive fashion. Such a metamaterial can be used to absorb seismic waves generated by earth quake and can efficiently refract back or absorb seismic waves protecting building infrastructure. E. Stealth applications Implementation of metamaterials in stealth technology finds its use to disguise moving bodies from radar detection; either by absorbing or by scattering all the radiations from the radar detection system. F. Acoustic cloaking Acoustic metamaterials are ordered artificial structures which can absorb and manipulate sonic, infrasonic and ultrasonic waves at will [14]. Bulk modulus and mass density counteract with permittivity and permeability. This analogous property is under high consideration and research is being done to direct and control sound waves at will by tailoring these two parameters. 2D acoustic cloaking has been optimized by placing five arrays of cylindrical lattices along a common central axis to achieve omnidirectional cloaking devices. G. Metamaterials antenna gain Metamaterials are being extensively studied to find applications in wireless communication and improved antenna performance i.e. gain enhancement, miniaturization of size, directivity and bandwidth broadening. Antenna is dispersive in nature and acts as an electrically small resonator. In a conventional antenna surface waves propagate to increase cross polarization effects, low gain, narrow bandwidth and frequency range. The quest continues as how to improve the output of conventional antenna by incorporating metamaterial and which particular type of metamaterial can improve one or more parameters? An important consideration to achieve antenna gain is to incorporate double negative DNG metamaterials and single negative metamaterials into antenna structure to harvest efficient yield and enhanced bandwidth. SNGs are dispersive in nature [15]. Using this principle behavior SNGs can be alternated with either MNG layer or ENG layer to get a suitable combination of optical properties, enhanced dispersion and miniature antenna size. Finally, to reduce surface wave propagation it is necessary to minimize the size of antenna and for this purpose photonic band gap metamaterials can be employed. III. Discussion and outlook The demonstration of recent advancements and analysis of metamaterials with all possible applications and future prospects in both academia and industry is interesting. Various antenna parameters have been studied and improved by Metamaterial Conventional materials Incident ray 70 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 80. different researchers suchas gain enhancement, dispersion, and bandwidth broadening and miniaturizing the size. Studies revealed that unit cell metamaterials of omega structure can be combined to form a slab and used to cover antenna surface and increases directive gain to 11.5 dBi [16]. Metamaterial based broadband antenna were demonstrated to show -1 dBi antenna gain at frequency of 2.5 GHz. with -1 dBi [17]. Electrically small antenna were demonstrated by combining layers of ENG metamaterials alternated with ordered multipole resonant structures to increase radiation efficiency and high impedance [18]. Metamaterials show exceptional physical properties which are overwhelming from both theoretical and academic perspective and definitely find relevance to aerospace and defense applications. A. Metamaterial airborne design In any purposeful review the electromagnetic properties of metamaterials, which strongly impinge airborne radome design cannot be neglected. Overall antenna performance is strongly dependent on the mechanical and electromagnetic properties of dielectrics. Unfortunately, there are limited number of materials which compromise over low dielectric constant and high mechanical strength. Significant research in the radome design is necessary in both spheres of metamaterials design and electromagnetism to circumvent radome, which may pose restrictions in metamaterials airborne design in near future. An outlook of current projects shows that the developed nations are involved in understanding the metamaterial concepts and their practical applications. For example, BAE systems has tested metamaterial based invisible vehicles which can disguise in infra red and microwave frequency range. Other international research groups are aiming at developing nano sized plasmonic devices [19]. U.S. Army Research office in collaboration with university students is developing liquid crystals tested in the range of 1- 110GHz. Other tunable devices made by liquid crystals are in the process of development such as tunable antenna [20]. IV. FUTURE CHALLENGES To address the future challenges metamaterials research must transcend the theoretical investigations and transform into practice. Practical applications make transitions through terahertz (THz), infra red and visible region of electromagnetic spectrum. THz metamaterials can be tailored by changing meta atoms at nanoscale and open new horizons of understanding and research in photonics and electronic band gap materials. Challenges related to efficiently generating and detecting broadband THz waves has primarily limited their use. Currently existing metamaterials respond to a narrow band of EM spectrum below red region of visible spectrum. This requires the development of metamaterials to extend their response in visible, micro and radio wave region of EM spectrum in both theory and practice. Thus metamaterials design can miniaturize antenna design [22] and dispersion coupled with improved satellite communication and can revolutionize the defense system. V. SUMMARY In this review we have taken a brief outlook of existing metamaterials, their fabrication routes, current projects and ongoing research to enhance performance and applicability of metamaterial devices around the world. Survey of metamaterials clearly demonstrated that metamaterials can be used for improving the performance of conventional antennas and can revolutionize the communication system. Additionally, a brief review of the current work and development in the field of metamaterials is provided. We also examined the on-going projects on metamaterial to enhance the performance of antennas and various novel metamaterials. Various future challenges are also considered. ACKNOWLEDGMENT Authors would like to acknowledge Department of Materials Science and Engineering, Institute of Space Technology, Islamabad for encouragement and moral support to write this review paper. REFERENCES [1] A. J. Hoffman, L. Alekseyev, S. S. Howard, K. J. Franz, D. Wasserman, V. A. Podolskiy, E. E. Narimanov, D. L. Sivco, and C. Gmach, “Negative refraction in semiconductor metamaterials,” Nat. Mater, vol. 6, pp. 946-950, October 2007. [2] N. Fang, D. Xi, J. Xu, M. Ambati, W. Srituravanich, C. Sun, and X. Zhang, “Ultrasonic metamaterials with negative modulus,” Nat. Mater, vol. 5, pp. 452- 456, April 2006. [3] R. A. Shelby, D. R. Smith, S. Shultz, and S. C. Nemat-Nasser, "Microwave transmission through a two-dimensional, isotropic, left-handed metamaterial," Appl. Phys. Lett. vol. 78, pp. 489-491, January 2001. [4] D. R. Smith, W. J. Padilla, D. Vier, S. C. Nemat- Nasser, and S. Schultz, "Composite Medium with Simultaneously Negative Permeability and Permittivity". Phys. Rev. Lett, vol. 84, pp. 4184- 4187, May. 2000. [5] K. Sanderson, “Materials science: Unexpected tricks of the light,” Nature, vol. 446, pp. 364-365, March 2007. [6] J. B. Pendry, “Negative refraction makes perfect lens,” Phys. Rev. Lett., vol. 85, pp. 3966-3969, October 2000. [7] H. C. Pocklington, “Growth of a Wave-group when the Group-velocity is Negative”, Nature, vol. 71, pp. 607-608, April 1905. [8] L. I. Mandel’shtam, “Group velocity in a crystal lattice,” Zh. Eksp. Teor. Fiz, vol. 15, pp. 475-478, April 1945. [9] V. G. Veselago, “The electrodynamics of substances with simultaneously negative values of ε and µ,” Sov. Phys., Usp, vol.10, pp. 509-514, February 1968. [10]V. A. Podolskiy, L. Alekseev, and E. E. Narimanov, “Strongly anisotropic media: The THz perspectives 71 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 81. of left-handed materials,”J. Mod. Opt, vol. 52, pp. 2343-2349, January 2005. [11]Z. Qiu, and S. D. Bader, “Surface Magneto-Optic Kerr Effect. Characterization of Materials,” 2nd ed., John Wiley & Sons, 2012, pp.1-10. [12]P. Los, A. Lukomska, S. Kowalska, R. Jeziorska, P. Zaprzalski, and J. Krupka, “Metamaterials based on polymer dispersions of nanoparticles and particles of copper obtained by cathodic current pulse electrolysis.” Materials Science-Poland, vol. 29, pp. 35-40, March 2011. [13]X. Liu, T. Tyler, T. Starr, A. F. Starr, N. M. Jokerst, and W. J. Padilla, “Taming the Blackbody with Infrared Metamaterials as Selective Thermal Emitters,” vol. 107, pp. 045901-4, July 2011. [14]C. Jo, J. Jeong, B. J. Kwon, K. C. Park, and I. K. Oh, “Omnidirectional two-dimensional acoustic cloak by axisymmetric cylindrical lattices,” Wave Motion, vol. 54, , pp. 157-169, April 2015. [15]G. V. Eleftheriades, and K. G. Balmain, “Negative- refraction metamaterials: fundamental principles and applications.” 1st ed., Wiley-IEEE Press. 2005. [16]A. S. Ahmad, O. Adib, J. J. Mohd, H. B. Noor, A. A. Robiatun, and F. F. Mohd, “Small patch antenna on omega structure met material,” Eur. J. Sci. Res. vol. 43, pp. 527-537, June 2010. [17]M. Palandoken, A. Grede, and H. Henke, “Broadband Microstrip Antenna with Left-Handed Metamaterials,” IEEE Trans. Antennas Propag, vol. 57, February 2009. [18]A. Erentok, and R. W. Ziolkowski, “Metamaterial inspired Efficient Electrically Small Antennas,” IEEE Trans. Antennas Propag, vol. 56, pp. 691-707, March 2008. [19]J. B. Pendry, A. J. Holden, D. J. Robbins, and W. J. Stewart, “Low frequency plasmons in thin wire structures,” J. Phys. Condens. Matter, vol. 10, pp. 4785, June 1998. [20]D. R. Smith, “Research group- Novel, electromagnetic materials program,” January 2005. 72 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 82. Finite Element Simulationof Composite Body Armor Numan Khan1 , Abdul Shakoor1 1 Dept. of Mechanical Engineering University of Engineering and Technology, Peshawar, Pakistan Write2numan@gmail.com Riaz Muhammad2 , Naseer Ahmed2 2 Dept. of Mechanical Engineering CECOS University of IT & Emerging Sciences, Peshawar, Pakistan. Vadim V. Silberschmidt3 3 School of Mechanical and Manufacturing Engineering, Loughborough University, Loughborough, UK Abstract—Finite element simulation of composite body armor for impact loading conditions is presented in this paper. The 7.62 mm armor piercing projectile with the velocity of 800 m/s is used for impacting on the target material. The numerical model is developed in ANSYS AUTODYN using Johnson– Cook material model for projectile and backing steel material. The Johnson–Holmquist material model is used for ceramic material. Different simulations are performed for different configurations of composite materials and the results of projectile velocity, kinetic energy, internal energy of target plate and perforation time is obtained. Finite element simulation of 6 mm ceramic backed by 2 mm steel plate has shown full penetration of bullet while the other backed by 4, 6 and 8 mm respectively has prevented complete penetration. The numerical results show that ceramic material is responsible for breaking projectile tip and absorbing most of the kinetic energy. The drop of about 66% in projectile kinetic energy is observed with increase of backing steel by 2 mm. Finally optimum ceramic/steel material thickness ratio of 1.5 with light weight is recommended for single body armor. Keywords- Numerical simulation; Ballistic impact; Projectile; Armor plate; Johnson–Cook material model; Johnson–Holmquist model I. INTRODUCTION Ever since the discovery of metals and alloys, different type of protection is made against the enemies attack for arresting high-velocity projectiles impact by mankind [1, 2]. The development of light weight and effective body armor is the need of the day due to the current law and order situation in the world and particularly in Pakistan. For the mobility of solders light weight efficient and cost effective body armor is necessary to be developed. Most of the ballistic materials shows nonlinear deformation and dynamic behavior under impact loading [3]. Material properties like strain, strain rate, materials hardening and thermal softening play a vital role in the design process [2]. With the passage of time different researchers have developed various ballistic resistive materials. Normally single high strength metallic plates are used for body armors. However, multi layered plate configuration are often used to meet the design specifications which cannot be fulfilled by metallic single plate. Zukas et al. [4] studied that a thick uniform steel plate was more resistant to hemispherical projectile than composite multi layered target of identical thickness due to free slip between plates. It was also observed that by increasing the number of plates in multi layered target plate reduces its resistance, which is much closed to the data obtained from the experiments of Almohandes et al. [5]. A detailed analysis of ballistic performance of composite and hybrid armor plates were performed using ANSYS AUTODYN software by Yohannes Regassa et al. [6]. Ceramic faced armor plate was used for observation of critical penetration velocity, which shows good agreement with experimental results. Teng et al. [7] has presented the performance of composite and single layered steel armor under ballistic conditions and concluded that nose shape, velocity and mass of projectile are the key parameters effecting body armor performance. Similarly, the effect of nose shape and hardness on failure mechanism of the armor is presented in [8]. Studies of H. Kurtaran et al [9]. has shown full penetration of semispherical nose shape projectile through 2mm thick steel plate at velocity of 500m/s. However the backing of front plate with a 20 mm thick plate of same material and two plates at different orientation angle has prevented full penetration. it was further observed that the results of Johnson-Cook Material model are more accurate than the plastic kinematic model for evaluation of the fracture damage and thermal softening [9]. Various analytical models have been developed by S. Feli [10]. To analyze the performance of Alumina plate backed by composite material made of Twaron fibers. During impact the fragmentation of ceramic conoid from ceramic tile and decrease in the semi-angle of ceramic decreases with increase of initial velocity was absorbed. The research for design of light weight and efficient body armor required more efforts for further improvement to increase its safety, performance and cost effectiveness. The current study demonstrates the finite element simulation of projectile piercing ceramic steel target armor. For fragmentation, the ceramic plate is modeled with Johnson-Holmquist model where brittle failure, strain rate effect and enormous deformation are considered. While backing steel is modeled using Johnson-cook model in which strain rate hardening effects, thermal softening and materials fracture are reflected [9]. 73 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 83. II. FINITE ELEMENTMODELING The performance of composite body armor under impact loading is investigated by explicit finite element model using ANSYS AUTODYN [11, 12]. To reduce the computational cost a small circular element of diameter 60 mm is considered for analysis. Initially the thickness of the target plate is kept constant to 6 mm whereas actual bullet dimensions are used for bullet modeling in the current analysis. For addressing the increase of damage effect of bullet on target plate with decrees in distance, these simulations are carried out at minimum possible distance. Similarly for simulation of the actual phenomenon of AK- 47 bullet with velocity range of 720 to 750 m/s, the projectile is fired with a velocity of 800 m/s and its rotational speed is neglected in the current study. Finite element model for different configurations of ceramic and steel are modeled as shown in Table 2. The thickness of the front ceramic material is kept constant to 6 mm while the 4340 steel of different thickness 2, 4, 6 and 8 mm are used as a backing material. To complete the perforation of the plate the total time of simulation is considered as 50 µs. the 3D model of target armor and 7.62 mm diameter projectile are shown in the Fig. 1. In current study the simulations are carried out on core i3 PC with 2.2 GHz processor. The time taken by one simulation is about between 8 to 12 hours of CPU time. A. Finite element mesh A 3-D finite element mesh for both target plate and projectile is generated by using automatic mesh generation feature in ANSYS workbench. As can be seen form Fig. 2 the mesh of inner region of plate is kept finer than outer region. In the mesh the maximum aspect ratio is kept below five. Both the bullet and plate are meshed with hexagonal elements of different size between 0.5 to 1.5 mm. The total number of elements for the projectile and armor are 148,000 including, 52,000 elements for projectile and 96,000 for armor plate. B. Material model Based on the material characteristics different material models are used for different materials in bullet impact simulation [1] In the current study, Johnson-Holmquist model represented by equation 1 [13] is used for the ceramic materials. This model is representing relationship between the intact and fractured strength, pressure and volume. The various parameter of the model are presented in Table.1. [ ( ) ][ ] (1) Where A, B, n, C are material constants and is representing the normalized strength of material. is the normalized pressure. The damage D can be represented by equation 2. ∑ ( ) (2) Here the equivalent plastic strain increment and is the increment in the volumetric compaction strain. Fig. 1. 3D model of projectile and target armor plate Fig. 2. The FE mesh details of projectile and target armor plate Similarly, Johnson-cook material model is used for the steel part of the armor. The main feature of the Johnson Cook model is representation[14-16] of the yield strength of materials on high strain, strain rate, temperature and material hardening, and is represented by the equation 3 [17]: ( )( )( - ) (3) Where A, B, C, n and m are constants and is representing value of effective stress, is effective plastic strain, is normalized effective plastic strain The term can be defined as: (4) In which and are used for melting and room temperatures, respectively. Similarly, the damage of element in Johnson-cook failure model is given by ∑ (5) Where shows the change of equivalent plastic strain and is the strain at fracture. The term can be represented by the equation 6. [ ( )][ ][ ] ( ) 74 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 84. Table. 1 Strengthand failure data for SiC using Johnson-Holmquist model [18] Parameters Units Values Density (g/cm3 ) 3.215 Equation of State Data (Linear) Bulk Modulus-A1 GPa 220 Strength Data Modulus of Rigidity GPa 193 Elastic Limit (Hugoniot) GPa 11.7 Intact Strength Constant-S1 GPa 7.1 Intact Strength Constant-P1 GPa 2.5 Intact Strength Constant-S2 GPa 12.2 Intact Strength Constant-P2 GPa 10 Constant of Strain Rate-C None 0.009 Fracture Strength (Max)-Sfmax GPa 1.3 Constant of Failed Strength-σ None 0.4 Failure Data Tensile Limit-T GPa -0.75 Constant of Damage- Efmax None 1.2 Constant of Damage- P3 GPa 99.75 Constant of Buckling-β None 1 The fracture constants are expressed by D1, D2, D3, D4 and D5. For describing the pressure volume relationships Mie–Gruneisen equation of state is used with combination of Johnson-cook model [9]. Constant for Johnson-cook model and Mie–Gruneisen equation of state model are given in Table 3. III. RESULTS AND DISCUSSION Results of finite element simulation for projectile velocity of 800 m/s and different configurations as shown in Table.2 are discussed in this section. The variation in bullet velocity during penetrations for different configurations is shown in Fig. 3. The config 1, 2 and 3 has maximum residual velocities of about 580 m/s, 230 m/s and 80 m/s respectively after impact while in config 4 projectile has been stopped and velocity is reached to 0 m/s at 50 µs. this shows that the percent reduction in projectile velocity is about 60% with the increase of backing material thickness by 2 mm. this reduction is mainly due to the impact of projectile on ceramic front material. Table. 2 Composition and areal density for configuration of ceramic and 4340 steel plate Configuration Front SiC plate Backing steel plate Areal density (kg/m2 ) Config 1 6 mm 2 mm 34.99 Config 2 6 mm 4 mm 50.69 Config 3 6 mm 6 mm 66.39 Config 4 6 mm 8 mm 82.09 Fig. 3. Variation in projectile velocity In the same way the Fig .4 gives the dissipation of kinetic energy for config 1 to 4 after impact of projectile. It is observed that maximum amount of kinetic energy is dissipated in ceramic fragmentation and the rest of residual kinetic energy is observed by the backing steel plate. The config 4 dissipates maximum amount of energy as compared to others as it goes through ceramic material. The percent reduction in kinetic energy is about 66 % with increase of thickness by 2 mm of backing material. This discloses that the backing steel plate is responsible for absorption of residual kinetic energy of the projectile and stop penetration. Fig. 4. Projectile kinetic energy The variation in internal energy of target armor plate can be observed form Fig .5 which shows the increase for different configurations. The Fig. 5a to Fig. 5d reveals that in all cases the maximum amount of increase in internal energy of ceramic plate is occurred. It is also observed that the increase in internal energy of ceramic front plate is about 20 % with increase of backing plate thickness by 2 mm. Similarly the maximum increase in average internal energy is observed in config 4 as shown in the Fig. 5. Thus by increase of backing 75 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 85. plate thickness maximumamount of kinetic can be absorbed in target armor plate. . Table.3 Johnson- Cook model constants for 4340 steel[19] Density, ρ (g/cm3 ) Elasticity modulus, E (MPa) Strength / Damage constants A (MPa) B (MPa) N c m 7.83 210000 910 586 0.26 0.014 1.03 Failure Constants D1 D2 D3 D4 D5 -0.8 2.1 -0.5 0.002 0.61 Mie–Gruneisen equation of state constants Γ0 C1 (m/s) 1.6 3574 S1 1.91 Ref. Temp (K) 293 Fig. 5. Change Internal Energy for (a) Config 1, (b) Config 2, (c) Config 3, (d) Config 4 76 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 86. Similarly Fig. 6.shows the bullet penetration for configuration 1 to 4. The config 1 shows full penetration of bullet at velocity of 800 m/s whereas in config 2 the penetration is stopped moderately. However, the bullet is fully stopped by increasing the backing material thickness as shown in config 3 and 4 in Fig. 6. Throughout the simulations the damage mechanisms are observed in the armor materials. Firstly, Silicon carbide material which is very brittle tends to shatter upon impact. It is also observed in the simulations that the cause of failure of silicon carbide is tensile stress wave that passes through the ceramic. Secondly, the tip of the projectile become blunt upon impact with ceramic material and adopted the shape like mushroom as shown in Fig. 7. The distortion in projectile shape is increasing with in increase in thickness as clear form Fig. 7. Another important point is the deceleration of projectile upon advancement in ceramic material. This reveals the fact that ceramic material is contributing mainly in the breaking of projectile tip and kinetic energy absorption. Finally the areal density for different configuration is listed. Areal density can found by adding the multiplication of the density and thickness values of ceramic and backing steel material [1]. Upon comparison with steel body armor of thickness 25.4 mm, about 100 kg per square meter can be reduced. Thus reducing the total weight can greatly improve the mobility of soldiers. Config1Config2Config3Config4 t=0 µs t=10 µs t=20 µs t=30 µs t=40 µs t=50 µs Fig. 6. Bullet penetration for different configurations at different time 77 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 87. Fig. 7. Bulletshape at 50 µs after impact on target armor plate IV. CONCLUSIONS In this work light weight ceramic composite body armor is analyzed for impact loading of AK 47. A composite material composed of Silicon Carbide and 4340 steel is suggested and is simulated for various thicknesses. Based on the simulation results, maximum reduction in kinetic energy and increase in internal energy with increase of backing plate thickness, the final optimum thickness ratio of 1.5 for ceramic/4340 steel configuration is suggested. It was also concluded that by reducing the thickness of the backing materials can significantly reduce the weight of the body armor, improve strength and will make it easy for the security forces to carry it. As clear form table. 2 the weight of about 15 kg per square meter can be saved if 2 mm less thickness is used for backing material. V. ACKNOWLEDGMENT The authors would like to gratefully acknowledge Dr. Rizwan Alim Mufti (Head of the Department) Pakistan Institute of Engineering and Applied Sciences (PIEAS) for his help in simulations. VI. REFERENCES [1]. Grujicic, M., et al., A computational analysis of the ballistic performance of light-weight hybrid composite armors. Applied Surface Science, 2006. 253(2): p. 730- 745. [2]. Lee, M. and Y. Yoo, Analysis of ceramic/metal armour systems. International Journal of Impact Engineering, 2001. 25(9): p. 819-829. [3]. Børvik, T., S. Dey, and A. Clausen, Perforation resistance of five different high-strength steel plates subjected to small-arms projectiles. International Journal of Impact Engineering, 2009. 36(7): p. 948-964. [4]. Zukas, J.A. and D.R. Scheffler, Impact effects in multilayered plates. International Journal of Solids and Structures, 2001. 38(19): p. 3321-3328. [5]. Almohandes, A.A., M.S. Abdel-Kader, and A.M. Eleiche, Experimental investigation of the ballistic resistance of steel-fiberglass reinforced polyester laminated plates. Composites Part B: Engineering, 1996. 27(5): p. 447-458. [6]. Regassa, Y., G. Likeleh, and R. Uppala, Modeling and Simulation of Bullet Resistant Composite Body Armor. [7]. Teng, X., T. Wierzbicki, and M. Huang, Ballistic resistance of double-layered armor plates. International journal of impact engineering, 2008. 35(8): p. 870-884. [8]. Corran, R., P. Shadbolt, and C. Ruiz, Impact loading of plates—an experimental investigation. International Journal of Impact Engineering, 1983. 1(1): p. 3-22. [9]. Kurtaran, H., M. Buyuk, and A. Eskandarian, Ballistic impact simulation of GT model vehicle door using finite element method. Theoretical and Applied Fracture Mechanics, 2003. 40(2): p. 113-121. [10]. Feli, S. and M. Asgari, Finite element simulation of ceramic/composite armor under ballistic impact. Composites Part B: Engineering, 2011. 42(4): p. 771- 780. [11]. Fawaz, Z., W. Zheng, and K. Behdinan, Numerical simulation of normal and oblique ballistic impact on ceramic composite armours. Composite Structures, 2004. 63(3): p. 387-395. [12]. Sadanandan, S. and J. Hetherington, Characterisation of ceramic/steel and ceramic/aluminium armours subjected to oblique impact. International journal of impact engineering, 1997. 19(9): p. 811-819. [13]. Johnson, G.R., Numerical algorithms and material models for high-velocity impact computations. International Journal of Impact Engineering, 2011. 38(6): p. 456-472. [14]. Muhammad, R., et al. Effect of cutting conditions on temperature generated in drilling process: A FEA approach. in Advanced Materials Research. 2011. Trans Tech Publ. [15]. Muhammad, R., et al. Finite-element analysis of forces in drilling of Ti-alloys at elevated temperature. in Solid State Phenomena. 2012. Trans Tech Publ. [16]. Muhammad, R., et al., 3D modeling of drilling process of AISI 1010 steel. Journal of Machining and Forming Technologies ISSN, 2010. 1947: p. 4369. [17]. Zukas, J.A., High velocity impact dynamics. 1990: Wiley-Interscience. [18]. Holmquist, T.J. and G.R. Johnson, Response of silicon carbide to high velocity impact. Journal of applied physics, 2002. 91(9): p. 5858-5866. [19]. Holmquist, T.J., D.W. Templeton, and K.D. Bishnoi, Constitutive modeling of aluminum nitride for large strain, high-strain rate, and high-pressure applications. International Journal of Impact Engineering, 2001. 25(3): p. 211-231. 78 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 88. Incorporate GNSS withAndroid & Improve the Search and Rescue Operations Atiq ur Rehman Aeronautics and Astronautics Institute of Space Technology (IST) Islamabad, Pakistan attiqpk2@gmail.com Najam Abbas Naqvi Aeronautics and Astronautics Institute of Space Technology (IST) Islamabad, Pakistan najm_naqvi@yahoo.com Abstract — Real time positioning or LBS (Location Based Services) is very important factor that improved the quality of services that’s offered by the institutions in various area, currently those services are provided using the sophisticated, expensive and dedicated devices that are used only for specific applications i.e. LBS or Real-Time Position determinations. In this research using ICT (integrated Communication Technologies) devices i.e. Android based Smart Phones and Tablets/ PDA’s (Personal Digital Assistants) that contain sensors i.e. GNSS receiver, Compass, Gyroscope and Accelerometers etc. Using open source tools a cost effective solutions is developed for market. Rescue Operation, Disaster Management needs real-time positioning services. The GNSS is freely available resource; currently the GNSS market is growing quickly and penetrates it in all areas. Now days the GNSS sensors are installed in wrest watches, phones, tablets and other communications devices. Android is an open source platform and a famous operating system of smart device. Freely available resources are used to integrate both of them android & GNSS to create the application for population to improve, timely and efficient emergency or disaster support management. Index Terms — GNSS, Cellular Communication, Smart Devices and Android I. ANDROID Android is an operating system (OS) based on the Linux and currently developed by Google corporation. With a user interface based on direct manipulation, Android is designed primarily for touch screen devices such as smart phones tablet PDA’s etc, with specialized user interfaces for televisions (Android TV), cars (Android Auto), and wrist watches (Android Wear). The OS uses touch inputs that loosely correspond to real-world actions, like swiping, tapping, pinching, and reverse pinching to manipulate on-screen objects, and a virtual keyboard. Despite being primarily designed for touch-screen input, as regular PCs and other electronics. Android's source code is released by Google under open source licenses, although most Android devices ultimately ship with a combination of open source and proprietary software. Initially developed by Android, Inc., which Google backed financially and later bought in 2005, Android was unveiled in 2007 along with the founding of the Open Handset Alliance, a consortium of hardware, software, and telecommunication companies devoted to advancing open standards for mobile communication devices. Android is popular with technology companies which require a ready-made, low-cost and customizable operating system for high-tech devices. Android's open nature has encouraged a large community of developers and enthusiasts to use the open-source code as a foundation for community- driven projects, which add new features for advanced users or bring Android to devices which were officially, released running other operating systems. The operating system's success has made it a target for patent litigation as part of the so-called "smart phone wars" between technology companies. The android operating system symbol is shown in the fig-1. Fig-1 II. GNSS The term “Global Navigation Satellite System” (GNSS) is refers to a constellation of satellites that provide signals from space transmitting positioning and timing data information. A GNSS provides global coverage. The GNSS receivers determine location by using the timing and positioning data encoded in the signals transmitted from satellites. In the world the 4 types of Navigation systems that are designed and developed by different countries, the name and their 79 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 89. developing countries asThe United States of America GNSS is known as GPS (Global Positioning System) or NAVSTAR. Russian GNSS is known as GLONASS in Russian Language known as “Globalnaya Navigatsionnaya Sputnikovaya Sistema” is provide the global coverage around the globe as examples of GNSS. Europe is in the process of launching its own independent GNSS, system name as Galileo. The Galileo Navigation system expected to provide the full worldwide coverage by 2019. The People Republic of China also developing its own GNSS system known as BeiDou Satellite Navigation System (BDS) and also known as COMPASS or BeiDou-2, will be a global satellite navigation system. BeiDou administration planned to begin serving global customers upon its completion in 2020. In this study we discusses the US based navigation system popularly known as GPS. because the GPS is provide its coverage around the globe. III. GNSS ARCHITECTURE The Global Navigation satellite systems (GNSS) consist of three major parts known as segments. One is The Space Segment, second is The Control Segment and last one is The User Segment. These are illustrated in Fig-2. Fig-2 Space Segment: The GNSS Space Segment is formed by a satellite constellation with sufficient satellites to ensure that the users will have, at least, 4 simultaneous satellites in view from any point at the Earth surface at any time. The nominal GNSS satellite constellation consists of 24 space vehicles (Satellites) (GPS constellation consists of six (6) orbital planes inclined at 55o associated with equator plane, the orbital altitude is 20,500KMs. GLONASS constellation is consists in Three (3) orbital planes inclined at 64.8o associated with equator plane, the orbital altitude is 19,100KMs. GALILEIO constellation is consists in Three (3) orbital planes inclined at 56o associated with equator plane, the orbital altitude is 23,220KMs. BeiDou constellation is consists in Three (3) orbital planes inclined at 55.5o associated with equator plane, the orbital altitude is 27,878KMs) The main functions of the Space Segment are to transmit radio-navigation signals (navigation Data) stored and retransmit the navigation message provided/uploaded by the ground Control Segment. Control Segments: The Control Segment or Operational Control System, OCS, is responsible for the proper operations of the GNSS system. The main tasks performed by the CS are the followings: • Monitor and control of satellites orbital parameters. • Monitor health and the status of the satellite subsystems i.e. solar arrays, battery power, stabilization etc • Activation of spare satellites. • Regular Update of clock corrections, ephemeris, almanac and navigation data. • Resolve satellite anomalies. • Controlling Selective Availability (SA) and Anti-Spoofing (AS) services • Passive tracking of the satellites. The Control segments comprise of major parts these areas. • Master Control Station. • Upload Antennas Stations • Monitoring Stations. User Segment The User Segment consists of GNSS user terminal that contain the L-band radio receiver/ processors and antennas which received GNSS navigation signals, Ephemeris and Almanac data to determine the accurate pseudoranges. After receiving the previous mentioned data then solve the navigation equations in order to obtain their coordinates. A GNSS receiver is a device capable of determining the user position, velocity and precise time (PVT) by processing the signal broadcasted by GNSS satellites. IV. APPLICATION DEVELOPERS Android is popular with technology companies which require a ready-made, low-cost and customizable operating system for high-tech devices. Android's open nature has encouraged a large community of developers and enthusiasts to use the open-source code as a foundation for community- driven projects, which add new features for advanced users or bring Android to devices which were officially, released running other operating systems. The operating system's success has made it a target for patent litigation as part of the so-called "smart devices wars" between technology companies. Android's default user interface is based on direct manipulation, using touch inputs, that loosely correspond to real-world actions, like swiping, tapping, pinching, and reverse pinching to manipulate on-screen objects, and a virtual keyboard. The response to user input is designed to be immediate and provides a fluid touch interface, often using the 80 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 90. vibration capabilities ofthe device to provide haptic feedback to the user. Internal hardware such as accelerometers, gyroscopes and proximity sensors are used by some applications to respond to additional user actions, Android devices boot to the home screen, the primary navigation and information point on the device, which is similar to the desktop found on PCs. Android home screens are typically made up of app icons and widgets; app icons launch the associated applications, whereas widgets display live, auto-updating content such as the weather forecast, the user's email inbox, or a news ticker directly on the home screen. Third-party apps available on Google Play and other app stores can extensively re-theme the home screen, and even mimic the look of other operating systems, such as Windows or IOS.[59] Most manufacturers, and some wireless carriers, customized the look and feel of their Android devices to differentiate themselves from their competitors. Various types of Application developer suits are used in market but Software Development Kit (SDK) is the famous platform to use for application development in android based operating devices. App Work Flow To develop apps for Android, you use a set of tools that are included in Android Studio. In addition to using the tools from Android Studio, can also access most of the SDK tools from the command line. Developing with Android Studio is the preferred method because it can directly invoke the tools that need while developing applications. However we choose to develop with another IDE or a simple text editor and invoke the tools on the command line or with scripts. This is a less streamlined way to develop because you will sometimes have to call command line tools manually, but you will have access to the same number of features that you would have in Android Studio. The basic steps for developing applications (with or without Android Studio) are shown in Fig-3. The development steps encompass four development phases, which include:. A. Environment Setup During this phase installed and set up development environment. And also create Android Virtual Devices (AVDs) and connect hardware devices on which install the development applications. B. Project Setup and Development During this phase we set up and develop Android Studio project and application modules, which contain all of the source code and resource files for designed application. C. Debugging and Testing During this phase build an application into a debug able .apk packages that can install and run on the emulator or an Android-powered device. Android Studio uses a build system based on Gradle that provides flexibility, customized build variants, dependency resolution, and much more. If using another IDE, build an application using Gradle and install it on a device using adb. Next, with Android Studio debug application using the Android Debug Monitor and device log messages (logact) along with the IntelliJ IDEA intelligent coding features. We can also use a JDWP-compliant debugger along with the debugging and logging tools that are provided with the Android SDK. And an Last, test your application using various Android SDK testing tools. For more information, see Test your application with the Testing and Instrumentation framework Fig-3 D. Publishing Publishing is the general process that makes your Android applications available to the users. When you publish an Android application you perform two main tasks: Prepare the application for release. During the preparation step you build a release version of your application, which users can download and install on their Android-powered devices. Usually, release application through an application marketplace, such as Google Play. However, we can also release applications by sending them directly to users or by letting users download them from your own website. V. PROPOSED APPLICATION Few years earlier the firms have no choice to manage & update changes their area maps by timely surveying. The rescue response on various emergency conditions response teams are managed by the area an number of rescue stations are managed regain, City and country because no one knwn the batter location of the area maps and totally dependent by the information that provided by the caller, the caller informed the condition of emergency and inform his location too, the call receiver/operator note down information and pointed the position on given map and then lookout the nearest rescue operation and mobilized them towards effected areas or parties. The mention procedure required much time to evacuate the 81 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 91. effected partied fromeffected place. Since 1992 the first GNSS is lunched worldwide by Unites States department of defense knows as Global Positioning System (GPS), this system is designed for military requirement but the system provides the navigation information round the clock globally with low accuracy for civilian users. Initially the GPS receivers huge in size and quit expensive, Now days the receivers are available in reduces in sizes and the decreases the price too. The reduced size receiver are used/ installed in the smart devices i.e. cell phones, tablets, wrest watches etc. as on proposed work the main aim is build up an android based application, that’s incorporate the GNSS, android & Cellular Communication to reduced the time for rescue/ emergency relief operation by fixing the exact position of the effected place or parties to save lives and also save them resources. A. Application Flow The android application functional block diagram is shown in the Fig-4, the functional diagram is operational when the app is installed on the android based smart device either it’s a cell phone or the tablet/PDA, whatever its brand but must have a GNSS receiver, now days devices have GPS/ GLONASS supported receivers. Start Direct Launched Launched by Dialed Numbers Enable GNSS Receiver Determine Position with time Position Fixed Determine Position with time No YES Position Stored Sending Source Generation .nav Extention InternetSMS Sent to Destination Server Via Internet Sent to Destination Server Via SMS Displayed Message “Position Sent” End Process Standby mode Repeat loop 3 times Disable GNSS Receiver Fig-4 After the installation is done on a device either it’s a cell phone or tablet. The application is standby mode because the power is important factor in the smart devices. The application is lunched by the App icon at app manager screen or by dialed fixed or predefined number patters i.e. 1122, 115 & 114 etc. The application starts and performed its initial operation that is the enable the GNSS receiver, in this step the physical GNSS receiver is power on and a popup on the screen the GPS enables. In the next block is performed operation i.e. tracked the presented satellites (4 signals) over the sky, and compute its position, this process take some time up 30 seconds initial for position fixation. It performed frequently operations until the position is fixed. In next step, but it’s performed the parallel with previous one (position fixing), its desired the information from the user (depend on the user of application i.e. courier services, vaccinations, the community health programs or emergency Search and rescue relief operation) Information competed with coordinates, now the information is saved at local storage with automatically generated label (either its date with time or with number series). The stored information is may be required to be converted on the required format (i.e. .nav, .txt or .ext etc extension files or nay other format according to the requirement) and also added the cell phone number for easy identification of the position, after suitable conversion and addition in the file, the information is ready to send their central information center using the internet or the SMS, tasks is repeated at three times in four minutes for better position for rescue operations. Once the task is complete the user can disconnect the call the application is automatically disabled the gnss receiver and goes to standby mode. CONCLUSION Currently, GNSS resources are available freely without any cost and GNSS market is growing quickly and has been expanded in vast areas of applications. Nowadays, GNSS receivers are installed in wrest watches, phones, tablets, satellites, aero planes and other communications devices. Android has an open source code and resource and can easily be integrated with GNSS. Applications for the local market can be produced i.e. health, search and rescue will be improved and enhanced in Pakistan. ACKNOWLEDGMENT I personally thank Dr Najam Abbass department of aeronautics and astronautics institute of space technology Islamabad, Pakistan for his kind effort for completion of the research project and express my warm thanks to Mr. Hassan Ali for his aspiring guidance, and sincerely grateful to them for sharing their enlightening views on a number of matters related to the project. REFERENCES 1] HOANG, V. D., HA, D. T., & PHUONG, X. Q. (2013). A Vehicle Monitoring and Navigation System Design based on Android Smartphone (Space, Aeronautical and Navigational Electronics). 2] Marti, E., Garcia, J., & Molina, J. M. (2014, July). Navigation capabilities of mid-cost GNSS/INS vs. smartphone: Analysis and comparison in urban navigation scenarios. In Information Fusion (FUSION), 2014 17th International Conference on (pp. 1-7). IEEE. 82 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 92. 3] Djuric, U.,Abolmasov, B., Dragana, P., Marjanovic, M., & Kuzmic, P. PORTABLE GEOTECHNICS–USING ANDROID SMARTPHONES AND TABLETS FOR GEOTECHNICAL FIELD INVESTIGATIONS. 4] developer.android.com 5] http://en.wikipedia.org/wiki/ Android_(operating_system) 83 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 93. Geostatistical Analysis onSeismic Data over North-Western Regions of Pakistan, Afghanistan and Eastern Regions of Iran & Tajikistan Malik Abid Hussain Khokhar1 , Asad Ali1 , Sadaf Javed1 & Razia Rani2 1 Department of Space Science, Institute of Space Technology, Islamabad 44000, Pakistan 1 abidmalikgeo@gmail.com, 1 asad06@gmail.com, 1 nomi4all786@gmail.com, 2 Department of Geography, Government College University, Lahore 2 raziaranigeo@gmail.com Abstract – Earthquake, it is a process which usually take place anywhere in the world. But most of the time this phenomena takes place at the place which is weak part of the earth - faults. Earthquakes occur where faults are but magnitude and depth remain less at particular places. Factually, it is considered that earthquake activity is normally related to particular activity of tectonic processes under the earth. The Seismic geomorphic structures contain particular basic source mechanisms and with regard to seismic activity, these bases are considered as a solution of tectonic regional planes. In this way, different seismic activities solutions bind their strong dependence with a recognized fault linage system. Furthermore, these connected seismic activities give us benefits of attaining an objective of absolute possible “depth and magnitude” of earthquake. The seism-o-tectonic model is an essential parameter that covers the assessment of earthquake activity. The tectonic framework of the said study region is extended from the fault system on the west to northwest Himalaya and boundary of Hindu Kush-Pamir terrain to the seismic belt of the geographical borders of Pakistan, Afghanistan, Iran and Tajikistan. Sequel to this, a geostatistical analysis are conducted by utilizing variogram estimations, anisotropy and kriging methods for seismic depth conversion and magnitude on the said sub-surfaces and interval velocity fields in a dependable way. Index Terms— Geostatistics, Earthquake, Seismicity, Magnitude, Variogram, Kriging. I. INTRODUCTION Earthquake take place in the faults areas where their intensity increases if there is weak surface prevailed then it bring about the source of a number of earthquakes as a natural calamity. Various features of geology in the region are direct manifestation of subsurface of the earth dynamics that’s why it is important to understand variety of problem related to physical endogenic processes aspects of happening deep down the earth surface. Earthquakes create faulting and folding due to above mentioned processes and bring about severe natural disasters in the affected areas. These happened due to primary and secondary and long waves that determined depth and intensity of the earthquakes[1]. The area is extended to Himalayan arc which encompass about 2,500 km from northwest to southeast and includes from east to west the width of the belt varies from 250-350 km. The colossal Himalayas and the Karakoram, representing the biggest concentration of lithospheric masses, grew south of the Pamir that goes to Iran from touching southwestern parts of Pakistan and the western regions of Afghanistan. The study area contains a captivating geological record of Precambrian to present and terminate both east and west with spectacular bends. The region covers the foot hills, lesser Himalayas, Greater Himalayas and Higher Tajikistan Mountainous Zones. Changes in the Earth's orography and consequent geomorphic changes are directly related to the ongoing seismic event. This collisional events have long been argued to be liable for geological, geochemical and seismic magnitudes of global extent[2]. The exogenic / eperogenic processes are still going on with the approximate rate of a few centimeters per year imperceptibly with continued erosional and denudation factors of internal movements. The eroded material from its rugged topography is regularly shed into different depositional settings within the Himalayas. In this region, the stress and strain caused due to plate movements which cause frequent earthquakes enormous loss of human life. Landslides and irregular thrusting have been also cause of frequent natural disasters. Deforestation and road construction have also been witnessed as a reason of landslides. Earthquakes occurred within the Baluchistan Range (Pakistan) and Iran ranges with the Mw6.4 which generated a complex and poorly located seismic sequence. These earthquakes mostly affected Pishin and Ziarat districts along with the borders areas of Iran, an area where several active faults have previously been identified. To create the relationship between these earthquakes variograms models have been established for geostatistical analysis”[3]. In the paper, geostatistical analysis (variogram estimations, anisotropy, kriging and cross validation) on Depth and Magnitude of the earthquake data on the subject regions are applied that display spatial structure ranges and directional ranges. II. DATA AND STUDY AREA Borders of Pakistan, Afghanistan, Iran and Tajikistan are the study area is chosen where earthquakes are frequent due to weak subsurface. Area is bounded in between 230 .37’N to 410 N and 440 E to 790 E. The points in the map are showing the number of earthquakes are frequent due to weak subsurface. Area is bounded in between 230 .37’N to 410 N and 440 E to 790 E. The points in the map are showing the number of earthquakes that occurred in between 2014 to March 2015. 84 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 94. Sample size ofeach country is equally taken and total samples are 150. Major concentration of the earthquakes is the North & the Western regions of Pakistan and offshore region in the South, Eastern and the Southern regions of Iran, boarder regions of Afghanistan, Tajikistan and Pakistan. Magnitude of these earthquakes is in between 4 to 5.5. Only, Iran is the country where magnitude is detected 6.6 that occur in the southern regions of the country. Fig.1. Location of the Study Area III. METHODOLOGY A. UTM Conversion of Coordinates and De-trending Data Initially, coordinates in the data were in Longitudes and Latitudes which were converted Universal Transverse Mercator Projection (UTM) WGS 1984 coordinates by giving appropriate zones as per the limits of the earmarked countries. This conversion was necessary to apply other geo-statistical analysis which are being explicated in the successive pages of this paper. After UTM conversion, trend was removed from the data before applying variogram models. Trend is removed from following equation 1. 2 2 ( ) ( ) ( y) (1)x y I x I y I x B. Variogram Selection/ Estimation The variogram or the spatial correlation function describes the nature/shape and construction of the arbitrary component of the concerned variable that is the difference between the true depth of the earthquake and its trend. So, variogram relies greatly on the trend. The variogram is specified by the standard error, the correlation range and finally some functional forms such as Linear, Exponential or Gaussian models. Variogram models are here to show the correlation of values of each other i.e. depth and magnitude. TABLE I. CALCULATION OF STANDARD ERROR FOR THE SELECTION OF BEST MODEL FOR PAKISTAN AND TAJIKISTAN TABLE II. CALCULATION OF STANDARD ERROR FOR THE SELECTION OF BEST MODEL FOR IRAN AND AFGHANISTAN For the selection of variograms, exponential model is most suitable to choose for the results of depths on the regions of Iran and Afghanistan and this model is selected because of lowest standard error 47.03481 S.E (Standard Error) among all the calculated models. Glancing on the table 2, it is evident that linear model is suitable for the variograms of Magnitude of the earthquake data as it has lowest standard error 0.48379274. Figure 2, 3, 4 and 5 are illustrating the variogram fit on the Depth and Magnitude. Observation on the trend of successive variograms are depicting cyclical trend where green lines are fitting lines of the model concerned and red lines in the variograms are showing estimated models. Variograms Model. 21 ( ) ( ) ( ) ( ) (2) 2 h m h z x z x hi i Exponential Model. ( ) 1 exp (3) h h c a Gaussian Model 2 ( ) 1 exp (4) h h c a Linear Model (5)( ) hh c a Equation 2 which denotes as: - ( )h is variogram, m(h) is number of point which making pairs, xi + h, xi : variables x at location i, & i + h: lag vector. Nugget (C0) is the start of variogram which starts from zero up till its constant figure which is known as sill (C). Partial Sill is denoted as C1 whereas variogram sills are combination of partial sill and Types of Models Depth Magnitude Linear 1.064598 1.064598 Exponential 1. 064594 1.064600 Spherical 1.064596 1.064596 Gaussian 1.064596 1.064586 Types of Models Depth Magnitude Linear 47.90590 0.48379274 Exponential 47.03481 0.9713586 Spherical 47.95570 0.5032916 Gaussian 47.78765 0.5004617 Circular 48.00123 0.5035408 85 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 95. nugget such as0 1C C C . Nugget effect is minimized to zero C(0). Owing to this condition, variogram is not enough for spatial variation analysis & suitable model must be fitted on it[4]. In the equation number 2, 3, 4 and 5 are known as the models as mentioned against each. Here, ‘c’ is a Sill and ‘h’ is lag specified from the distance/ weight derived from the coordinates and given values at the coordinates. C. Kriging Interpolation provide Kriging methods besides Kernel Density etc. In geostatistical methods, Kriging is also “Best Linear Unbiased Predictor" (BLUP) which satisfies a certain criterion and provide minimum variance[5]. The kriging is based on the weighted moving average[6]. 0 0 (6) var Z x Z xii Z It is an estimated iable at the spatial location x location Z x observed parameter of Depth and Magnitudei every predicted form nearest neighbore of spatial points weights applied to these neighboring po ts a i t in Unity of sum is the weights are constrained so Equation 6 is an unbiased estimator[7]. IV. RESULTS Results in the successive pages are over the seismic data of the study areas. After the application of various models of variograms as per the appropriate requirement of model of the variogram, Ordinary Kriging analysis are applied. It is because of its real implementation and OK (Ordinary Kriging) is appropriate applicable where spatial sampled data is large. This method comprehensively evolves on sampled observations instead of complete population where trend assessment of spatial identity always re-assessed so that a smooth and accommodative model may arise. For depth and magnitude analysis, Ordinary Kriging (OK) remains always very suitable, but it does not work properly if we have very small sample data for using in moving neighborhood observations and so it bring about the reason of further good and useful results[9]. The results which showing the variograms are in cyclical with the models of Exponentials, Linear and Gaussian. These were derived on the basis of lowest standard that bring about the enormous appropriate cause of validity of findings for any data basing upon geostatistical methods. Results are restricted to bifurcated due to large area and large sample size. Initially, sample size of the data was accumulatively 600 which split to two parts of 300 each. Data was full of trend which has been shown in the variograms models and anisotropy plots. These plots are derives from the R language and ArcGIS was also incorporated for visual interpretations which has been useful to predict the nearby areas by known points. By viewing, figures 12, 13, and 14 are predicting depth and magnitudes of earthquakes by the application of Kriging and Kriging is found the most suitable predictors because it is showing very accurate results. For comprehensive geostatistical analysis, omnidirectional variograms are not taken to compute the directional trend in the data which require to analyze trend of anisotropy in the sampled data. So sequel to these indicators directional variograms are calculated with the angle tolerance 150 from 00 to 1650 . Figure 6 and 7 are interpreting anisotropic trends as well as directional trend in the data for depth and magnitude of seismic data on the subject regions, where point- paired values are evidently displayed with respect to their directions Finding enormous trend of direction, it is evident that direction is not uniform to a specific side rather all sides which depicts that compactness of the lithosphere which is not too hard to sustain the shocks of frequent earthquakes that’s the reason for tremendous isotropic trend. Beside these, another way of determining said trend, there is a method of Rose- Diagram that is used for computing anisotropy within the map having 3600 angular direction, under mentioned figures are related to said directional trend which are showing the direction of earthquake magnitude and its depth in the subject regions. With respect to different angels, an unequivocal variation in the above shown rose diagrams of Pakistan & Tajikistan Regions and Iran & Afghanistan regions. When the data of Magnitude values and depth values of occurred earthquakes of the subject regions are slanted in a side then both plots show only that certain direction either 00 degree or 900 degree whereas the trends elucidate the level of depth in different directions specified rocks of weaken lithosphere. For comprehending the intensity of the depth and magnitude of the earthquake, figure 6 and 7 exhibit variogram factor parameters with all possible angles. Numerical figures of the said variograms are showing directional variogram for relating parameters of same dataset within anisotropic influence. Magnitude and depth have different numbers of values in view of the sill, partial sill and nuggets and range is comparatively less differ in the whole dataset because these are based on the Standard Error’s values which are being selected on the basis of lowest vales and models are also applied on the appropriations where Depths have 700 at 1500 and sill is 12000 that is the maximum figure in case of Pakistan and Tajikistan regions and on the same angles Iran and Afghanistan have the lowest observation at this angle but this regions has more than 6000 sill vales at 1180 and nugget values is 1500. Deviations proceed to the resultant anisotropic effect in the data values. Similarly, Magnitude data is of the both regions have different sill effects, nugget and rang in terms of their direction significances. Merely, directional variograms are not sufficient for geostatistical analysis for computations for interpolation methods like Kriging and Inverse Distance Weighted analysis but other analysis of variograms are also needed to interpolated directional ranges and anisotropic corrected variograms of current data. Above discussed directional ranges are estimated directional coverage of predicted surface areas along with appropriate model fittings. Figure 14 is showing Kriging plot of depth values of earthquake data in all the areas as mentioned in topic and figure 13 is established using Geostatistical Analyst Tool of ArcGIS. Figures 12 is serving a clear depiction of variables Depth and Magnitudes of whole data set with the 86 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 96. complete area ofconcentration within UTM coordinates i.e. x- limits 7524 to 1506500 and y-limits 2501400, 4359300; where filled dots are the known points and the hallow points are estimated points as estimated with the help of Ordinary Kriging method of prediction and the same derived from the R language.. These figures are showing all four areas but with cut-off extents. If we concentrate on the results, we will be satisfied by the results of predicted Kriging because of its plots and cross validation results. TABLE III. CROSS VALIDATION OF DEPTH AND MAGNITUDE ON PAKISTAN AND TAJIKISTAN REGION Pakistan and Tajikistan Regions ( Total 300 Sample Points were Collected ) Cross Validation on Depth Attributes Min Max Mean Median 1st Qu 3rd Qu MPE MPE/ MEAN RMSE RMSE/STD RMSE/IQR Observed 1.80 250 64.47 41.55 10 111.5 - - - - - Predicted 9.13 212.8 65.87 47.33 20.17 109.3 -1.411 -0.2188 26.6608 0.4699597 0.2624747 Residual -97.4 121.77 -1.4110 0.022 -11.09 9.930 - - - - - Cross Validation on Magnitude Observed 4 5.80 4.456 4.40 4.20 4.60 - - - - - Predicted 4.32 4.70 4.456 4.369 4.374 4.553 -1.18 -2.6589 0.349549 1 0.9820722 0.8738728 Residual -0.63 1.369 -0.00118 -0.062 -0.27 0.166 - - - - - 87 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 97. TABLE IV. CROSSVALIDATION OF DEPTH AND MAGNITUDE ON IRAN AND AFGHANISTAN REGIONS Fig. 2. Exponential Variogram Model on Depth – Pakistan and Tajikistan Regions Iran And Afghanistan Regions (Total 300 Sample Points were Collected) Cross Validation on Depth Attributes Min Max Mean Median 1st Qu 3rd Qu MPE MPE/ MEAN RMSE RMSE/STD RMSE/IQR Observed 10 265.6 70.32 33 33 98.45 - - - - - Predicted 9.33 243.9 71.570 38.46 33.15 99.82 -1.252 -0.0178 34.711 0.541 0.5303518 Residual -155 146.3 -1.253 -1.825 -10.73 3.993 - - - - - Cross Validation on Magnitude Observed 4 6.6 4.537 4.5 4.2 4.7 - - - - - Predicted 4.23 4.978 4.538 4.523 4.466 4.583 -0.058 -0.0012 0.4161 0.9743606 0.8323063 Residual -0.94 -0.28 -0.00058 -0.050 -0.286 0.1830 - - - - - 88 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 98. Fig. 3. ExponentialVariogram Model on Depth – Iran and Afghanistan Regions Fig.4. Gaussian Variogram Model on Magnitude – Pakistan and Tajikistan Regions 89 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 99. Fig.5. Linear VariogramModel on Magnitude – Iran and Afghanistan Regions Fig.6. Directional Variograms on Depth Values – Pakistan and Tajikistan Regions 90 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 100. Fig.7 Directional Variogramson Depth Values – Iran and Afghanistan Regions Fig.8 Interpolated Directional Ranges on Depth Values –Pakistan and Tajikistan Regions 91 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 101. Fig.9. Interpolated DirectionalRanges on Magnitude Values –Pakistan & Tajikistan Regions Fig.10. Anisotropy-Corrected Variograms for Depth - Iran & Afghanistan Regions 92 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 102. Fig.11. Anisotropy-Corrected Variogramsfor Magnitude - Iran & Afghanistan Regions Fig.12. Area of Plot for Prediction of Depth and Magnitude 93 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 103. Fig. 13. Krigingon Depth Values with Depth Unit Fig.14. Kriging on Magnitude Values with Unit 94 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 104. Fig.15. Cross Validationof Depth and Magnitude Predicted Values Fig.16. Cross Validation of Standard Error 95 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 105. Fig.17. Cross validationof Error Values of Depth Fig.18. Normal QQ Plot 96 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 106. V. ROSS VALIDATION Cross-validation,it is also known as the appreciation with the name of rotation estimation, that is a technique of model validation to evaluate geostatistical analysis for generalization of an independent dataset [10]. It is mainly used in settings where the goal is prediction, and one wants to estimate how correctly/ perfectly a predictive model to perform in practice into implementations of cross validation. It is the tool of computing the validation of correctness and authentication of the methodology of geostatistical analysis like Kriging and Kernel Density etc. It is important to mention over here that the main reasons for using cross-validation instead of using the conventional validation is that the error (e.g. Root Mean Square Error) on the dataset. Because the conventional validation is not a useful estimator of model performance and thus the error on the observational dataset does not properly represent the assessment of model performance[10]. This way to tool is not only the part of geostatistics methods but also it is a part of Geostatistical Analyst tool in ArcGIS. Sequel to this, following are few plots which are derived from Geostatistical Analyst Tool that includes predicted values, error, and standard error and QQ plot. QQ plot is showing the distribution of dataset. Figure 15 is the depiction of distribution of predicted observations and the sample observations with their cross- validation. VI. CONCLUSION On the subject topic of four countries (i.e. Pakistan, Iran, Afghanistan and Tajikistan), six hundred sample points/ observations were collected and only three parameter are used which are coordinates, magnitudes and depth of the earthquakes. Geostatistical analysis are applied on these parameters. With the help of R language, major geostatistical analysis (variogram model analysis, anisotropic analysis and Kriging etc.) are conducted; however ArcGIS is also incorporated for visual interpretation with the utilization of geostatistical analyst tools. Taking all the analysis methods on board, it is concluded that Ordinary Kriging is the valid/ best estimation method because it presents accurate smooth canvass of surface and minimum standard error estimation as revealed in cross validation plots and table number 3 and 4. For correctness of the application of Kriging cross the validation on depth and magnitude of earthquake data is carried out as well. Seismic spatial point data are gigantic source of information to see the location and concentration of earthquake where we can estimate disaster damage which help us to layout decision making efforts for benefiting mankind. Semi-variogram and other variogram models analysis of seismic data are very useful for estimating destructions and directions of waves of earthquakes as showing in the figures 2,3,4, 5 and it is also significant for direction of earthquakes. The paper has shown that semi-variogram variables of range, nugget and sill which can be easily used to pinpoint earthquake spatial deviation and consequently the location of severely damaged areas as well. Hence, it is concluded that Northeastern regions of Afghanistan, Eastern regions of Tajikistan are major concentrated areas of earthquake and more deepened than the other regions of Pakistan and Iran. However, as for as magnitudes of the earthquakes are concerned, eastern and east- southern parts of Iran are very vulnerable due to weaken parts of Lithospheric compactness which can give room for more severe earthquakes with higher intensity of 6.6. Southern parts of Pakistan with the borders of Iran and western Balochistan are also vulnerable places for future earthquakes. ACKNOWLEDGEMENT The authors are indebted to Dr. Asad Ali who guided them for writing this paper and he reviewed the paper as a whole. His precious time from his busy schedule and commendable cooperation is highly appreciable in conducting this research work. REFERENCES [1] N. L. and T. M. Gary Mavko, “Seismic Methods For Imaging Physical Properties of the Earth,” pp. 0–3, 1995. [2] S. Development and I. M. Region, The Tajik Pamirs Challenges of Sustainable Development in an Isolated Mountain Region. . [3] D. L. P. Parasad, Journal of Nepal Geological Society, vol. 45. . [4] C. E. G. James, R., “Use Of Geostatistics For Accurate Mapping Of Earthquake Ground Motion,” no. 1988, pp. 31– 40, 1989. [5] V. De Rubeis, P. Tosi, C. Gasparini, and A. Solipaca, “Application of Kriging Technique to Seismic Intensity Data,” vol. 95, no. 2, pp. 540–548, 2005. [6] M. Franklin, “Solution to Ordinary and Universal Kriging Equations,” pp. 1–5, 2014. [7] R. Giraldo, P. Delicado, and J. Mateu, “Geostatistics for Functional Data : An Ordinary Kriging Approach.” [8] K. Chang, Introduction To Geographic Information Systems, Sixth Edit. Printed in Singapore: McGraw-Hill, 2012, p. 418. [9] P. Abrahamsen, “GEOSTATISTICS FOR SEISMIC DEPTH CONVERSION 3 . Kriging in the geophysical literature 4 . Seismic depth conversion 5 . Velocity or depth prediction 6 . Variogram estimation,” pp. 1–9, 1996. [10] S. Geisser, Predictive Inference. Chapman and Hall, 1993, p. 3471. 97 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 107. GIS for estimatingoptimized water demand using sustainable water resource management for a planned city Rabia Tabassum,1,2 , Mudassar Hassan Arsalan2,3 , Arif Inam Osmani3 , Anam Khalid2 1 Science and Humanities Department National University of Computer and Emerging Sciences (FAST), Karachi, Pakistan 2 Geoinformaticsand Sustainable Development Research Lab Institute of Space and Planetary Astrophysics, University of Karachi, Pakistan 3 Osmani & Company (Pvt.) Ltd., Karachi, Pakistan . Abstract Water, a vital natural resource, is the topic of modern researcher and getting momentum throughout the world due to its scarcity. Many regions of the world have reached a point at which present water resources are already being over-used, as indicated by the depletion of groundwater aquifer and rivers which no longer to reach the sea. The alternate water supply approach becomes more important to achieve sustainable urban development and to reduce the energy as well. The idea of using modern techniques like remote sensing and geographical information systems (GIS) in water resource management studies is quite mature. In the last two decades, RS and GIS have been broadly used for the preparation of different sorts of thematic layers and integrating them for multiple purposes, especially for watershed analysis, ground water recharge and water demand optimizations. The DHA City Karachi (DCK), the study area, has been envisioned as the first sustainable city of Pakistan which will probably serve as a model for the future independent planned cities in the country and abroad. The domestic water requirement of 54 gpcd has been considered as a standard of Karachi Water and Sewerage Board for the metropolitan city and average water demand per unit area is calculated as 0.09 g/ft2 for DCK. This study consists of two parts, in the first part water demand estimation is done for the DCK land use; Residential, commercial, Mixed use, Amenities, Recreation, Utilities, Transportation, Agriculture farming and Reserved spaces. Water in gallon per capita per day (gpcd) and gallon per unit area per day (gal/ft2 ) are considered as water coefficients for the water demand calculation. DCK water demand on time scale is estimated as per its development and inhabitation strategy, which is going to develop in three phases; Short Term (2012-2015), Mid Term (2015-2020) and Long Term (2020-2030). Secondly, to estimate optimized water demand for the land use, the most feasible and easy to deploy strategies and methods like efficiency practices (behavior practice and Technology practices) and gray water recycle and reuse is used. By implementing these water reduction methods on DCK estimated water demand, it is evident that recycling and reuse of waste water have a positive impact on ecological sustainability. Index Terms: Water demand efficiency, Recycle and reuse of gray water, Water demand, Alternate water resources, GIS. I. INTRODUCTION Water is one of the most vital natural resources that somewhat different from the other, as it is viewed as a key to prosperity and wealth. Water has played a crucial role in the growth, function and location of communities[2]. Egypt, South Africa and Pakistan are the most water-limited nation among the twenty five most populous countries. According to the World Bank data, Pakistan’s river system has a storage capacity of 30 days of water usage, India stores 120 days, while US stores 900 days of river water. In international and national agendas, sustainable water policies search is at its peak. Approximately 40% of the world’s population will probably experience shortages in water by 2050. When water could not find in the desire quality and amount at the right place and time, water problem may exist [3]. World’s arid and semiarid regions have motivated the development of the innovative management measures due to water shortage. At large, the major part of Pakistan is arid and semi-arid, therefore water management system improvement is necessary. The systems to manipulate and present spatial data can be considered as Geographical Information systems (GIS), which are capable of performing a range of function such as data capture, data storage, data cleaning, data finding and retrieval, statistical and spatial analysis, and data display [4]. GIS applications as efficacious technology in water resources management for storing, managing, displaying spatial data are constantly raising. These applications, including supply and sewer system modeling, surface hydrology and ground water modeling, storm water and nonpoint source pollution modeling for urban and agricultural areas. Because of the spatial nature of the data for design, planning, operation and maintenance of water and sewer system in urban areas, GIS technology used for such systems to provide benefit significantly. The unique characteristics like family size, lot size, property value, and soil property, which vary from on different geographical locations, are associated with consumers that affect demand[5]. In hydrological modeling, GIS could be useful to provide the information needed for a watershed runoff analysis. The integrated technique using 98 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 108. GIS and physicalbased distributed hydrological models considered as an effective technique in the field of watershed –based water resource management[6]. II. LITERATURE REVIEW In the last few decades, GIS use started to proliferate in the field water resources management and planning. Mark R. Leipnik and Hugo A. Loaiciga [7] discussed eachaspects of GIS and focused on the other facets of GIS pertinent to the planning and management of water resources. Geographical Information Systems (GIS) have become an integral and beneficial tool in many analyses such as statistical and spatial analysis in water resources management [4, 8, 9]. Panagopoulos, et al. [10] estimated the future water requirement based on population growth using multi-criteria decision analysis in which Analytical Hierarchy Process (AHP) and GIS techniques were involved. AHP has proved as a useful tool for planners that assess future water demands and it can be suitable for planning of a national water management system. Udovyk [11] ‘s research revealed that GIS couple with modeling have potential to make a significant contribution to integrated water resources management (IWRM) and also notified the issue involved in using GIS as part of IWRM. GIS considered as a powerful tool to interpolate and model the spatial data, enable to overlay and compare the environment and population data for quantifying the population at risk, and also communicate these to decision makers and other stakeholders[11]. A particular aspect of water resource management problems, needs a specific approach to the development of GIS data structure. By constructing a conceptual GIS data model to integrate the physical and logical components of the modeling problem into an operational framework, Daene C. Mc Kinney [12] expended the function of GIS that implements a tight linkage between Water resources management and GIS model Ahmad [13] introduced a new approach for Simulation of Water Resources Systems called spatial system dynamics (SSD which presented for modeling feedback dynamics processes in time and space. GIS and System dynamics are couple to develop this model. In the study presented by Allen, et al. [14], GIS is used for verifying and calibration a continuous simulation model with an event –driven model; they realized that GIS is useful in enhancing the modeling efforts because of increased accuracy, and minimization of human error with time. One of the GIS application is future forecasting of water demand, Shamsi [15] presented this application and established the criteria for future development with the help of potential development map, a digital elevation model, a zoning map, and water ,sewer and road network maps. Same works could be made for sewerage flows for the rapidly growing population areas. For calculating and projecting the average daily sewerage flows from the different land use such as residential, commercial, and industrial, Kirshen [16] used data handling properties of GIS. For the development of a master plan of waste water system, Giguere [17] used GIS and discovered its efficiency to bring the relevant data from present databases into geographic analysis environment, view information and, create report. The master plan predicts future wastewater flows, assesses existing treatment and collection system potential, constructs a sewer rehabilitation program, assesses and identifies alternatives to expand wastewater facilities, and makes a phased plan to finance the required improvement. Moutal [18] presented same work for different country. The automated mapping system was used to record, update and database development for planning, design and operations. Domestic water demand considered as a complex element of different factors, such as socioeconomic and physical characteristics, urban planning strategies, infrastructure, and public water policy. In GIS environment, different thematic layers are used such as road network distance, distance from the city center, distance from the coastline, topographic slope, land use/land cover, Urban Plan, population density and existing water supply and sewerage system that correspond for the planning factors. This study is basically for newly planned city and GIS tools not only used for the master plan of the city, but also for estimating water demand on spatial and temporal bases. The main emphasis of the study is the estimation of optimized water demand using sustainable water management approach. III. STUDY AREA DHA City Karachi (DCK) has been envisioned as “The First Sustainable City of Pakistan”, which will serve as a model for the future independent cities in the country and abroad. It is with the aim to provide its citizens a safe, comfortable, efficient, and beautiful lifestyle. DCK is strategically located on the Karachi-Hyderabad Super Highway in District Malir at the eastern border of Karachi as an independent urban center in the vicinity of Karachi Metropolis. The city is planned on an area spanning 11,668 Acres, comprising residential, commercial and mixed-use elements. In geographical characteristics topography and distances are the main factors affecting the potential water sources. DCK is naturally drained by Abdar and SukkanNullah (small non- perennial water tributaries of Malir River). Their productive aquifer is recharged from the same water streams and small water reservoirs developed on them. It extends from latitudes 240 57 N to 250 2 N and longitudes 670 24’ E to 670 32’ E, and is a sub-tropical location in character. The study area has a moderate climate and experiences bulk rainfall in the duration of monsoon season, but the average rainfall is recorded as only 7inches per annum. In summer, May and June are the hottest months and temperature often reaches the 43°C. January is considered as the coldest month with temperature 5°C. As a part of Karachi district, DCK may get its share of water from the bulk water supply by Karachi 99 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 109. Water and SewageBoard. But the current water budget of KW &SB’s has been already in stress and demand is more than supply. A. DCK land Use Planning The DHA City Karachi is the study area of this research and Figure 1 shows its land use. For DHA city Karachi the entire planning approach focuses on Sustainability Design Principles, including the concepts of connectivity, efficiency, renewability, minimization of externalities and carry capacity[19]. These goals of sustainable design have been adopted in the overall districts planning approach by maintaining the ecological integrity of site nullahs, reserved spaces and natural contours[20]. Master plan of DCK cumulatively covers approximately 11,668 Acres of land and comprises of eight major land use types as shown in Figure 2. IV. METHODOLOGY The basic aim of this study is the estimation of optimized water demand by using alternate water resources for DCK in GIS environment. For this purpose, the master plan of DCK was converted into GIS format. ArcGIS software was used to translate files from AutoCAD drawing files to the ArcGIS shape file. The master plan of DCK was consisting of detailed parcels/plots, transportation corridors, reserved spaces for natural drainage, open and green spaces etc. Land use classification was applied to parcels / plots, according to the detailed master plan categories on different levels, such as the major level comprised Residential, commercial, mixed use, amenities, recreation, utilities, transportation, agricultural and reserved spaces. The minor / detailed land use maps were also developed for each individual major land use class according to DCK planning. The flow chart of land use layer formation is shown in Figure 3. The water demand for DCK was calculated in the spatial domain, on the basis of water in gallon per capita per day (gpcd) and gallon per unit area per day (gal/ft2 ), which was measured according to the land use types. On a time scale DCK water demand is estimated as per its development and inhabitation strategy, which is going to develop in three phases; Short-term, Midterm and Long-term. Sewage volume is calculated, which gives the amount of water that can be reused for the non-potable use. Also by implementing the water reduction method such as efficiency practices and grey water recycling, optimized water demand is estimated. A. Criteria of Water demand estimation According to United Nation, in 2012 approximately 50% of the world's population lived in urban areas and this percentage is expected to swell to 60 % by 2030[21]. Therefore, water demand in commercial, domestic, and industrial areas are consequently growing in cities and result the water scarcity[22]. For newly planned city, water supply authorities Figure 1: Study Area “DHA City Karachi with its land use Figure 2: Land use types and % of Area of DCK Figure 3: Flow chart of land use layers formation 100 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 110. are needed tosupply water to their consumers with a certain degree of reliability. Non uniform distribution of water will cause water scarcity in some area and excess in other areas which result the obstacle the natural growth of the city. Therefore, proper criteria setting is necessary to know the water demand in spatial and temporal domain instead of computing demand as a whole which is the main objective of this study. Municipal water use commonly consists of residential, industrial, commercial, and public uses, as well as some minor use for other purposes like system losses, line cleaning and firefighting. In developing countries public water withdrawal is only 13-26 gpcd in contrast, many developed countries’ water withdrawal is 79-158gpcd. Municipal water demand is exactly related to the amount of water withdrawn by population in town, cities, housing estates, public and, domestic service enterprises. The public demand consumes high quality water from the city water supply system and as its supply also includes water for industry that provides directly for the requirements of urban populations. The portion of municipal water demand that is used in the home is the residential or domestic water demand and worldwide typical domestic use of water in comparison with residential plot of 500 square yards is given in Table 1. This water use is highly dependent on socioeconomic and environmental conditions and including toilet flush, bathing, cooking, drinking, watering lawn, and other uses. The average water demand varies from 25 to 120 gpcd whereas the most commonly used range is 50-80gpcd for the world urban areas. In many developed countries residential water use can constitute over half of total municipal water use and consumption. Residential water use is also directly linked to the general public health. Improving the health of the poor is one of the main goals of water and sanitation projects. Therefore, household water use is always the top priority in the municipal water supply[23]. Estimation of water demand based on per-capita consumption is one of the common approaches. Karachi Water and Sewerage Board is the main source of water supply for Karachi city. DCK’s current estimated water demand stands at 45MGD. Initially the average domestic water requirement has assumed as 54 gallons per capita per day for the DCK which is a reasonable value in comparison with per capita daily water demand for the different world cities, e.g. Tokyo –Japan (77gpcd) New York-USA (161 gpcd) Colombo-Sri Lanka (60gpcd) and Mumbai- India (43gpcd) [24]. Most of the people who have purchased plots of land for residential or commercial purposes in the DCK are likely to belong to the affluent society of Pakistan. The houses or dwellings constructed by them are considered to be of superior quality depicting a higher standard of living, which requires a large amount of per capita water consumption. It is anticipated that this may necessitate a higher per capita water supply to be provided for the water supply system in this area. In comparison to the general average provision normally adopted by KWSB for their water supply project as for Karachi city, which is only 36gpcd[25]. Commercial, Industrial, and Institutional (CII) sectors are considered as the significant contributors to Public water demand. Historically utilities have relied on water use coefficient, which is the number of employees for estimating CII water use, but it is difficult to obtain this information at affine resolution. To overcome this situation through spatial, physical and economic property-based information, Miguel A. Morales [26] developed a methodology to estimate CII water use. For calculating commercial water demand, a survey has been conducted in different commercial areas of Karachi. With the help of a questionnaire, commercial water demand has been estimated on the basis of water uses for drinking, cleaning, washroom and other[27]. The result shows that water demand per employee has been recorded as 2.3 gal per employee per day, whereas the water demand per unit area was 0.031gal/ ft2 . Generally, the commercial water demand is about 10-20% of total water demand. The average water demand as per planning of the DCK commercial plot is given in Figure 4. Table 1: World Wide Typical domestic use of water in comparison with residential plot. Types of Water Use World Average (%) Residential 500 sq.yd plot (%) Toilet Flush, including its leakage 33 30 Shower and bath 28 25 Wash basin 11 10 Kitchen 9 10 Drinking, cooking (2 - 6) Dishwashing (3 - 5) Garbage Disposal (0 - 6) Laundry and washing machine 16 15 Lawn sprinkling and miscellaneous 3 10 *Source: Interim Report-1 “feasibility report of DCK alternate water supply” [1] Figure 4 : Average Water Demand as per planning of DCK commercial plot. 101 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 111. Water use inpublic building such as community hall, Marriage hall, Mess club, culture and art, school etc. as well as in public services including fire station, street washing and landscape irrigation are considered as a Public water use. Usually 5-10% of Municipal water demand use in public area[24]. As per planning of DCK the average water demand for different public building are given in Table 2. The public water demand may become more precise when the actual detailed building structure is planned. V. RESULTS A. Water Demand for DCK A key input in municipal water services planning and design is the estimation of present water demand, and the prediction of future water demand. In distribution of domestic water equally to all regions with proper pressure, a uniform spatial water distribution system will be very helpful[28]. As the total average water use when DCK will be fully developed expected to reach up to 45MGD and therefore water demand per unit area is around 0.09g/ft2 . Land use planning is used for estimation of water demand for DCK. Table 3 shows the distribution of water requirement for the land use of DCK. In the whole water requirement, the biggest share is of residential use (41% of total water demand) which occupies the major share of land in DCK (37% of land use). However the water demand per square feet for residential area is quite low (0.09g/ft2 ). The further distribution of water requirement on various sizes of plot in the DCK residential area is shown in Figure 5. It shows that approximately 46% water will be required for R-5 (i.e. plots of 500 square yards). In the DCK almost 2% area is designated for commercial plots of various sizes in community centers and in the central business district. However, some of the largest plots with higher FAR are planned as the Mall Zone in a commercial area and has the highest water demand as compared with the other commercial areas as shown in Figure 6. Its vertical height (FAR) and horizontal size (plot size) with density of water users (population) are the major water demand controlling factors. In the overall DCK the second highest water requirement will be for commercial plots, which is almost 16.8% of the total water demand. Although most of the mixed use plots in community centers (C3, C4 and C5) are comparatively small in size, however large designated plots in the CBD and South Zone are the major role players due to their sized and respective FAR. Further distribution of water demand for mixed use plots is given in map of Figure 8. It has Table 2 Average water demand for public water use in DCK Land Use Average Water Requirement Community Hall, Marrige Hall, 0.15 gallon per sqft per day Mess , Club, Culture & Art 0.15 gallon per sqft per day Mosque 0.5 gallon per sqft per day Hospital 250 gallon per bed DCK Clinics 0.15 gallon per sqft per day of covered area Collage/ School(day scholar, without boarding) 17 gallon per student per day University 34 gallon per student per day Glof club 0.02 gallon pers qft per day Lake View Park 0.02 gallon per sqft per day Park and Playground 0.01 gallon per sqft per day Theme Park 0.01 gallon per sqft per day Nursery 0.02 gallon per sqft per day Road Landscaping 0.006 gallon per sqft per day Table 3: Average Water demand of fully developed DCK S- No Land use Types Water Demand(MGD) Water Demand % 1 Residential 18.6 41.2 2 Commercial 7.6 16.8 3 Mixed use(Res. Cum Com) 7.4 16.4 4 Amenities 3.7 8.3 5 Recreation 0.7 1.7 6 Utilities 5.4 12.0 7 Transportation 1.0 2.3 8 Agricultural and Reserved Spaces 0.6 1.2 Total 45.0 100.0 Figure 5: Distribution of Water demand (MGD) in Residential Land use Figure 6: Distribution of water requirement (MGD) in Commercial Land use. 102 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 112. been observed thatthe water need for mixed use plots is 16.4% of the total water demand of DCK and almost same as for commercial plots. A long list of required amenities and special used is planned in DCK that covers around 7.3% of DCK land. It consists of Educational institutions, public building, religious center, health facilities and some reserved amenities for future requirements. The average water requirement per unit area for amenities is 0.16 g/ft2 with an overall requirement of 8.3% of total water demand. Educational institutions are the main water demanding units due to the major strength of water users (students). Further detail of average water requirement is given in map of Figure 7. The main requirement of water in recreational use is related to the landscaping and green space development. The estimated water demand coefficient is 0.1 g/ft2 for recreational areas. Approximately 1.7% of DCK land is designate for recreational purpose. This share will not only serve to the local recreational demand, but also cater for regional requirements of high quality amusement and sport facilities such as theme Parks, golf course and specialized sport centers as shown in Figure 10. There are some reserved areas within DCK which are required to be kept green that includes right of way (ROW) of oil and gas pipelines passing through the DCK reserved areas for agriculture and some open and green spaces along the storm water drainage corridors. These areas will require around 1.2% of the total demand with low water requirement per unit area (approximately 0.01g/ft2 ). The water distribution map of these areas is given in Figure 9. As a first sustainable city of Pakistan, DCK is planned in a great comprehensive manner that is going to minimize its dependencies up to the highest extent on the resources outside. In this connection its entire energy requirement is planned to produce at the site with the blend of renewable and conventional energy sources. Similarly, the water requirement is going to be fulfilled by optimizing the conjunctive use of alternate water supply with regular water supply from the bulk water source. Almost 3.3% of the land have been designated for exclusive use of utilities to manage in a sustainable manner. The overall average daily water requirement for the complete city is around 12%, with the lion's share for various processes and in conventional power generation (i.e. 79% of the total utility requirement). Its per unit area requirement is quite higher (0.34 g/ft2). Further details are given in map of Figure 11. In the whole DCK urban Green is in focus. With good architectural design, landscaping is planned with multi- Figure 8: Distribution of Water requirement in Mixed use Figure 7: Distribution of Water requirement in Planned Amenities and Special Use Figure 10: Distribution of Water requirement in Recreation Figure 9: Distribution of Water requirement in Agriculture and Reserved Spaces Figure 11: Distribution of Water requirement for Utilities 103 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 113. parameter. Fruits, flowers,flower colors, fragrance, birds’ attraction and water requirement are the main selection parameter for soft landscaping. In this connection other than recreational facilities, transpiration corridors are also planned with soft landscaping. Transportation corridors cover around 34% of DCK land with different row and road categories. The water requirement for transportation corridors landscaping and other minor uses will be around 2.3% of total water demand with highest share at streets. The details are given in map of Figure 13. 1) Spatial Analysis of Water Demand According to zonal distribution, DCK comprises of four main zones; DCK sectors, South Zone, Down Town and Gateway, as well as miscellaneous area as shown in Figure 14. In zonal distribution more than half (60%) of water share is required for residential sectors with a comparatively low water requirement per unit (0.07g/ft2 ). Second largest requirement of water is in south zone that is almost 15.2 % of the total demand with a comparatively higher value per unit area (0.15g/ft2 ). The highest water requirement per unit area in City Gateway and Downtown (i.e. 0.27g/ft2 ), that covers around 14% share. Further sectors wise water requirement are given in Figure 12 and Figure 15 that show the sector wise split of land use water demand. 2) Temporal Analysis of Water Demand Many of engineering applications and operational management of water distribution system are based on the future water consumption. Therefore, it is necessary to forecast water demand not only on spatial scale but also with development time considerations. It is very hard to get a reliable estimation of future water demand without a clear understanding of historical and current trends of water use patterns [29]. Commonly, water demand prediction is divided into three categories: long - term, medium –term and short- term periods. Water forecasting in future years used for distribution system design refers to long-term period. Seasonal water consumption refers as median-term period, whereas short –term period concern with daily and hourly water demand. Domestic water use by households is a dominant municipal water requirement in urban areas. Water utilities to make system plan, design and asset management are important for municipal water demand, short- and long- term forecasting. Long term water demand is useful for management and operation, whereas short term is essential for municipal water demand short and long term forecasting. Water demand prediction accuracy achievement is challenging, however , because the predicting model must simultaneously consider multiple factors associated with economic development, population growth and migration, climate changes, and even consumer behavioral patterns[30]. The population growth rate of the city is required for forecasting the domestic water requirements. Future population prediction is based on the knowledge about the city, its development of surrounding areas, environments, potential trade expansion and, infrastructure and communication network development. As per DCK development strategy, it is going to develop in three phases; Short-Term (2012-2015), Mid- Term (2015-2020) and Long- Term (2020-2030). On the basis of population growth of Figure 13: Distribution of Water requirement in Transportation use Figure 14: Zonal Wise water demand Figure 15: Sectors wise water demand of DCK Figure 12: Sector wise split of land use Water Requirement. Table 4: Water requirement with respect to DCK development phases Development plants Water requirement after completion (MG) Yearly Water Resource Development requirement(MG) Short-Term (2012-2015) 11.1 3.7 Mid-Term (2015-2020) 20.7 4.14 Long-Term (2020-2030) 11.5 1.15 104 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 114. DCK, water demandfor these three phases has been estimated as shown in Table 4 and Figure 16 is represented its spatial distribution. By considering development plans it is estimated that for short- term development approximately 11.1 MG water will be required that will serve the complete demand of the short- term planned area. However, its water source development will be tricky in the beginning plan; just for the short term development plan average annual water resource development requirement is around 3.7MG per year. However the actual use of water will be started with the occupancy and use of land for the respective purpose. It is estimated that in initial stage only water is required for construction purpose and only 10 to 15 % of water will be required in DCK Camp site till the end of short term development plan. It is better to develop around 1.5 to 3 MG water resource by the end of short term development plan. B. Reduced water demand A demand reduction measure serves to reduce water demand, water losses, peak water demand, and non-essential water uses. Reducing water consumption in the home, market and at public places is a simple and easy way to decrease water and energy bills and reduce household’s impact on the environment. Conserving scarce water resources helps reduce the need to dam rivers, reduce waste water produced and treated at sewage plants, lower energy requirements for treating and transporting water and waste water, and reduce greenhouse emission. 1) Strategy to Reduction water demand For the efficient use of water at usage point is more effective to save water and reduce overall municipal water demand. There are many methods and strategy for water reduction, but most feasible and easy to deploy are; Use of Water efficient plumbing devices or fixtures and Recycling or re-use of grey water a) Use of Water efficient plumbing devices or fixtures In the last two decades the use of water efficient plumbing fixture has been introduced throughout the world considering water scarcity. It has been observed that water efficient appliances in the home can reduce indoor water consumption by 35-50%[31]. Typically used such devices and reported water efficiencies are given in the Table 5. Error! Reference source not found. gives figure for water demand expected saving in water demand at DCK pertaining to use of water efficient devices. It is clear from the above table that the use of the aforementioned devices can save 44 % water and the average water requirement will be reduced to 30 gpcd. The use of such devices can either be regulated in building codes for their increased use. It is also suggested that new home building should be guided to use these devices through media campaigns. The land use like amenities and commercial building give more saving by using water conservation devices. b) Recycling and re-use of Grey water For grey water recycling, the overriding principle is safe usage of recycling water. Therefore, we have not taken grey water recycling where its safe usage can be questioned or its consumption is not required. For plot sizes smaller than 500sqyrds, in any land use, have not been considered because they do not have any green area and grey water usage for landscaping is suggested only for large size plots. Health facilities have not been incorporated for calculation of grey water because of the hazardous nature of medical waste. Therefore, grey water volume has been computed for residential plots R5, R10 and R20, commercial plots (C5, C10, and C20), mixed use (M5, M10, and M20), as well as planned amenities. For this purpose, grey water shall be treated sites and reused for landscaping. Amenities especially educational institutes, produce a lot of grey water. This phenomenon Figure 16: Water Demand for DCK development phases Table 5: Water demand using water efficiency devices Name of efficient device Frequency of use(per person) Daily Water use without water conservation device (gal/person) Daily Water use with water saving device (gal/person) Daily Water saving with water saving device (gal/person) Annual Water saving (gal/person) Low-flush Toilet (1.6gpf) 5.1 flushes /day 204 8.2 12.2 4453 Low- volume showerhead (2.5gpm) 5.3 minutes/day 15.9 13.3 2.6 949 Low- volume Faucet (1.5gpm) 4 minutes/day 12 6 6 2190 Front-loading washing machine (27gpl) 0.37 loads/day 18.9 10 8.9 3249 Water efficient dishwasher (7gpl) 0.1 loads/day 1.1 0.7 0.4 146 Total 68.3 38.2 30.1 10987 *Assumes conventional toilet at 4gpf, showerheads at 3gpm,faucet at 3gpm, washing machine at 51gpl, and dishwasher at 11gpl[1] 105 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 115. coupled with thefact that these amenities are usually surrounded by many green spaced requires for the efficient storage and usage of this grey water. Hence, it is suggested that firstly smart taps and such fixtures be attached to lower the water consumption and secondly to utilize the volume of grey water produced to irrigate neighboring green areas. c) Water conservation at power plant: Power house usually consumes a very large amount of high quality water. The expected water requirement without water conservation is about 79% of total utility water demand. This high water usage is normally not provided through municipal network. Therefore, according to the standard procedure, the energy generation plant will manage its own water resource based on deep well ground water, reclaimed water and limited volume of municipal resources. It is expected that power plant will not have designated quota from the municipal water supply. For water conservation; it will be recycled inside the plan and it is further suggested that most economical water use practice are deployed. The initial assessment is based on wet cooling, which amount to 4.2 mgpd, whereas dry cooling process will take only 0.58 mgpd. The other possibility is to recycle with adopting wet cooling, which will give saving of 1.1 mgpd. C. Optimized Water Demand The estimated water demand for DCK is 45MGD which can be reduced by using recycled Grey water and use of efficient devices. For this purpose water demand is calculated on the bases of three scenarios. The first scenario is based on the grey water use for landscaping and gardening within and outside plots by the user of the building. For making it more practical it is assumed to construct the grey water tanks separately for storage and making it mandatory for all the residential, commercial, mixed use and the amenities plot holder having plot area 500sq yards or more. This is only possible when an owner of houses installs separate drains for grey and non grey water. This scenario will impact the reduction up to the 7.5% of water required for the landscaping etc. In adding to the first scenario of using grey water, if DCK enforces the use of water efficient fixtures having rate of water, saving up to 44% in the kitchen sink, faucets, showers, flush, taps and laundry, with monitoring and auditing this enforcement may produce a large impact on water demand reduction (about 43.3%). In the third scenario only water efficient devices are used that provide a moderate reduction in water demand approximately 39.3%, which is slightly less than second scenario. The results of these three scenarios for different land use of DCK are shown in Table 6. VI. DISCUSSION AND CONCLUSION Water scarcity has become a major problem in arid and semi- arid regions therefore it is necessary to get rid this problem. As implementation of the new water resource has least probability due to its effect on the environment, therefore alternate supply and demand reduction is the only solutions to overcome water scarcity. Sustainable water resources management is considered as an important application of GIS and remote sensing, especially for the newly planned city. Using this technology, this research not only help to estimate the water demand for the new planned city with respect to its land use, but also provided an optimized water demand by using alternate water resources such as reuse or grey water and using efficient devices. The study results show that residential Land use has the maximum water demand of 41% of total municipal water demand, whereas the recreation and agriculture areas are required comparatively the lowest water demand with only 1%. As DCK is considered as a first sustainable city of Pakistan and for maintaining its high standard of life style Table 6: Water Reduction by using three scenarios S- No Land use Types Estimated Water demand (%) Water Demand Reduction (%) Scenario-1 Scenario-2 Scenario-3 (Use of Grey Water) (Use of both grey water and efficient devices) (Use of efficient water saving devices) 1 Residential 41.3 10.2 49.5 44.6 2 Commercial 16.9 5.3 38.2 34.2 3 Mixed use(Res. Cum Com) 16.4 5.4 44.6 41.9 4 Amenities 8.2 18.9 45.9 32.4 5 Recreation 1.6 0.0 85.7 85.7 6 Utilities 12.0 0.0 5.6 5.6 7 Transportation 2.2 00.0 00.0 0.0 8 Agricultural and Reserved Spaces 1.3 00.0 00.0 0.0 Total 100.0 7.6 43.3 39.3 106 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 116. comparatively high waterdemand is essential. Karachi water and sewerage supply is the only water resource for DCK which already supply insufficient water for Karachi city. Reuse or recycling of waste water and use of efficient devices serve as efficient and valuable ways to cope with the scarcity of water and can reduce the water demand. The estimated waste water for the DCK is 66% of the total water demand, taken from the land use; residential, commercial, mixed use and amenities, can reuse after the treated for landscaping and agriculture outside the houses. For implementing these alternate water resources, well-structured planning is necessary. As for reuse of wastewater, black water drain system should be separated with grey water and forcefully implementation of water efficient fixture is essential for the householder. There will be 45MGD water required for the whole planning of DCK and water demands by using these alternate resources are estimated. The results of the first scenario based on reuse of grey water able of only 7.9 % reduction of water demand. Whereas in the third scenario, efficient device use show that efficient fixtures can reduce water demand up to 39%. The second scenario, based on reuse of grey water with the use of efficient devices, is considered as an optimized water demand because of the maximum reduction in water demand about 43%. After successful deployment of the suggestion without failure to operate and maintain the system could result in significant reduction in water demand. Considering the scarcity of water in the area, we expect that these critical management practices will be adopted and implemented. For further study this research can be extended to estimate the portable and non-portable water use from the reused waste water. This research provides a sustainable water resource management system which is the requirement of the better life style free from water scarcity which is expected to increase in future. VII. ACKNOWLEDGEMENT The authors are grateful to DHA City Karachi, Pakistan and Osmani & Company Pvt. Ltd. for sharing planning information and other relevant literature related to the DCK. Authors are also thankful to the GIS and planning team of Osmani & Company Pvt. Ltd. for their great support. VIII. REFERENCES [1] OCL, "Interim Report -1 , " Feasibility study for alternate water supply stretegy including rechagre, reuse and small dams " Osmani Company (PVT) limited and Pakistan Defences Officers Housing Authority Karachi 2013. [2] F. Arbu s, . a. . Garc a- ali as, and . art ne - Espiñeira, "Estimation of residential water demand: a state- of-the-art review," The Journal of Socio-Economics, vol. 32, pp. 81-102, 2003. [3] F. A. Ward, "Decision support for water policy: a review of economic concepts and tools," Water Policy, vol. 9, p. 1, 2007. [4] O. Udovyk, "GIS for intergrated water resourses management," Integrated Urban Water Resources ManagementNATO Security through Science Series., pp. 35-42, 2006. [5] R. H. a. H. R. Vassilios A. Tsihrintzis, Fuentes., "Use of Geographic Information Systems (GIS) in Water Resources: A Review," Water Resources Management, vol. 10, pp. 251-277, 1996. [6] Z. X. Xu, K. Ito, G. A. Schultz, and a. J. Y. Li, "Integrated hydrologica modeling and GIS in water resources management.," Journal of computing in civil engineering, vol. 15, pp. 217-223, 2001. [7] K. K. K. Mark R. Leipnik, and and Hugo A. Loaiciga, "Implementation of GIS for Water resources planning and management," Journal of Water Resources Planning and Management, vol. 119, pp. 184-205, 1993. [8] A. Gemitzi, V. A. Tsihrintzis, O. Christou, and C. Petalas, "Use of GIS in siting stabilization pond facilities for domestic wastewater treatment," J Environ Manage, vol. 82, pp. 155-66, Jan 2007. [9] K. V. Kalivas DP, Karantounias G, " A GIS for the assessment of the spatio-temporal changes of the Kotychi lagoon, western Peloponnese, Greece.," Water Resourece Management, vol. 17, pp. 19-36, 2003. [10] G. P. Panagopoulos, G. D. Bathrellos, H. D. Skilodimou, and F. A. Martsouka, "Mapping Urban Water Demands Using Multi-Criteria Analysis and GIS," Water Resources Management, vol. 26, pp. 1347-1363, 2012. [11] O. Udovyk, "GIS for intergrated water resourses management," Integrated Urban Water Resources Management, pp. 35-42, 2006. [12] X. C. Daene C. Mc Kinney, "Linking GIS and water resources management models:an object-oriented method," Environmental Modelling & Software, vol. 17, pp. 413- 425, 2002. [13] S. a. S. Ahmad, S. , "Spatial System Dynamics: New Approach for Simulation of Water Resources Systems," Journal of Computing in Civil Engineering, vol. 18, pp. 331–340, 2004. [14] L. Allen, J. Christian-Smith, and M. Palaniappan, "Overview of Greywater Reuse: The Potential ofGreywater Systems to Aid Sustainable WaterManagement," Pacific Institute 654 13th Street, Preservation Park Oakland, California.2010. [15] U. M. Shamsi, "GIS, Remote Sensing, and Master Water Plan: A Case Study," In Computing in Civil Engineering and Geographic Information Systems Symposium, pp. 695- 702, 1992. [16] P. Kirshen, Nvule, D., and Corliss, J, "Estimating urban and suburban sewerage flows withassistance of GIS technology," in Proc. ASCE Water Resow Plan. Manage. Conf, Baltimore, MD, 2-6 August,, 1992, pp. 208-212. [17] P. R. a. B. S. Giguere, "Application of GIS to waste water system mmmaster planning in Albuquerque," in ASCE 107 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 117. National Conference CivilEng Appl. Remote Sensing GZS, Washington DC,, 1991, pp. 115-119. [18] H. P. a. B. outal, D. , "Updating New York city’s sewer and water main distributionsystems: practical applications of GIS," in ASCE National Conference Civil Eng Appl. Remote Sensing GZS, Washington DC,, 1991, pp. 155-164. [19] OCL, "DHA city Karachi planning report," Osmani Company (PVT) Limited and Pakistan Defences Officers Housing Authority Karachi2012. [20] OCL, "DHA city Karachi ecology Report," Osmani Company (PVT) Limited and Pakistan Defences Officers Housing Authority Karachi2012. [21] U. Nation, "The Millennium Development Goals Report," UNITED NATIONS2012. [22] P. R. Philippe Loubet , Eleonore Loiseau , and V. Bellon- Maurel, "Life cycle assessments of urban water systems A comparative analysis of selected peer-reviewed literature," Water Reseach, vol. 64, pp. 187-202, 2014. [23] T. Lu, "Research of domestic water consumption: a field study in Harbin, China," Master of Science of Loughborough University, Water, Engineering and Development Centre Department of Civil and BuildingEngineering, 2007. [24] E. M. M. Syed R. Qasim, Guang Zhu, Water Works Engineering: Planning, Design, and Operation: Prentice- Hall Of India Pvt. Limited, 2000. [25] M. Ahmad, "Pakistan Water and Sanitation Operators Directory," Water and Sanitation program. 2012. [26] J. P. H. Miguel A. Morales, Kenneth R. Friedman, and Jacoueline M. Martin., "Estimating commercial, industrial,and institutional water use on the basis of heated building area," Journal American Water Works Association vol. 103, 2011. [27] R. Tabassum, "Title," unpublished|. [28] K. H. V. Durga Rao, "Multi-criteria spatial decision analysis for forecasting urban water requirements: a case study of Dehradun city, India," Landscape and Urban Planning, vol. 71, pp. 163-174, 2005. [29] M. A. Al-Noaimi, "Decelopment of water resources in Bahrain: A combined approch of supply demand analysis," PDh in School of Earth, Ocean, and Environmental Sciences, University of Plymouth. [30] C. Qi and N. B. Chang, "System dynamics modeling for municipal water demand estimation in an urban region under uncertain economic impacts," J Environ Manage, vol. 92, pp. 1628-41, Jun 2011. [31] D. Inman and P. Jeffrey, "A review of residential water conservation tool performance and influences on implementation effectiveness," Urban Water Journal, vol. 3, pp. 127-143, 2006. 108 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 118. Optimized Threshold Calculationbased on received signal characteristics for Blanking Non-linearity at OFDM Receivers Ferheen Ayaz SUPARCO Institute of Technical Training, HR Dte Gen, Space and Upper Atmosphere Research Commission, Karachi, Pakistan ferheen@msn.com Zehra Ali Space Electronics Wing, Space and Upper Atmosphere Research Commission, Karachi, Pakistan Zehra.aly82@gmail.com Abstract—In satellite communications, OFDM (Orthogonal Frequency Division Multiplexing) is evolving quickly as a well- established scheme. OFDM based communication system performance, despite of its robustness and high efficiency, is greatly affected by the addition of Impulsive Noise (IN). Additive impulsive noise deteriorates the transmitted signal and overall performance results poor when signal to noise ratio (SNR) and symbol error rate (SER) are considered. Blanking nonlinearity scheme is among the simplest methods to reduce the adverse effect of impulsive noise at OFDM receivers. The performance of blanking non-linearity is based upon threshold selection. In blanking, the received samples whose magnitude exceeds a certain fixed threshold are considered to be IN-affected and are therefore set to zero. The value of threshold plays a very important role in detecting the samples affected by IN. Setting a fixed threshold does not always determine the IN-affected samples truly as the OFDM distribution characteristics vary throughout the signal. Several theoretical methods to optimize the threshold level to improve system performance have been reported. This paper reviews some of the threshold optimization methods and proposes a new method based on some characteristic distribution of the received signal including median and peak. This proposed optimized threshold can be used practically and results in higher or comparable Signal to Impulsive Noise ratio than the other methods which are incapable to be adopted in practical OFDM receivers. Keywords—OFDM; impulsive noise; blanking non-linearity; threshold; median; peak I. INTRODUCTION Orthogonal Frequency Division Multiplexing (OFDM) is a rapidly growing modulation scheme and is widely adopted in many wireless applications [1]. Performance of OFDM modulation scheme for GNSS (Global Navigation Satellite Systems) is one of the latest research topics [2]. Other applications of OFDM include 4G mobile satellite systems [3] and European Digital Terrestrial Video Broadcasting (DVB-T) [4]. Despite of many advantages of OFDM communication systems, the performance degradation in them still persists due to high-power additive Impulsive Noise (IN) pulses and their random occurrence in a channel [5]. DVB-T systems are affected significantly by Impulsive Noise [6]. Performance degradation in OFDM systems due to impulsive noise result in high Symbol Error Rate (SER) and low Signal to Impulsive Noise Ratio (SINR) [7]. A number of schemes to eliminate or reduce impulsive noise from the received signal have been reported. These schemes aim to increase the Signal to Impulsive Noise Ratio (SINR) of the overall system. The simple approach to mitigate impulsive noise from a received signal is to introduce a memoryless non-linearity at OFDM receiver before demodulation. Blanking, clipping and hybrid clipping- blanking schemes have been devised using this approach. Among all the non-linearity schemes, blanking is the easiest to implement as it only converts the high amplitude samples corrupted by IN to zero [7-8]. The blanking non-linearity scheme revolves around the decision of whether the received sample is affected by impulsive noise or not and then changing it to zero. The decision is taken by a fixed parameter known as threshold level or simply threshold. The threshold level indicates if IN has corrupted the received signal sample or the original sample value is retained at the receiver exactly as transmitted. If the received sample’s amplitude exceeds the threshold level, it is considered to be corrupted by IN and therefore it is multiplied by zero. If the threshold level is higher than the sample’s amplitude, it is assumed to be the original sample value unaffected by IN and hence multiplied by one to provide no change to the sample [9]. Fixing a value of threshold is very important as it defines the overall SINR of the system. A low value of threshold may blank the original signal samples which were initially not corrupted by IN and a high threshold value may pass the IN affected samples without blanking [10]. Fig. 1 illustrates the two values of threshold to be used with blanking nonlinearity on a signal affected by impulsive noise. Since the performance of blanking non-linearity is greatly affected by the selection of threshold, the choice of an optimized threshold is a crucial decision. Optimized threshold has been calculated in literature by different methods [11-12]. However, theoretical calculations of optimized threshold 109 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 119. cannot be usedat practical OFDM receivers where probability of IN occurrence is unknown. In this article, we review the two threshold optimization techniques for blanking nonlinearity; first is adaptive threshold optimization method which maximizes the Signal to Impulsive Noise Ratio (SINR) [11] and second is dynamic peak based threshold estimation method [12]. We also propose a new method for calculating threshold based upon median and peak of the received signal. The proposed technique of threshold calculation has the advantage over the other two schemes as it can be used at practical OFDM receivers without the information of original signal transmitted, probability of impulsive noise occurrence and maximum SINR achievable. 0 2 4 6 8 10 12 x10 4 0 2 4 6 8 10 12 OFDM symbols corrupted by IN Amplitude OFDM symbols Impulsive Noise Threshold 1 (Too low) Threshold 2 (Too high) Fig. 1. Two threshold levels considered for blanking IN-affected samples. The organization of the remaining paper is as follows: OFDM transmission system with blanking non-linearity scheme is described in Section II. Section III overviews the two threshold optimization techniques and proposes Median Based Threshold (MBT). Section IV includes simulation results in which MBT is compared with the other threshold optimizing techniques with respect to Signal to Impulsive Noise Ratio (SINR) and Symbol Error Rate (SER). Conclusion is presented in Section V. II. THE OFDM TRANSMISSION SYSTEM OFDM transmission is represented in the form of block diagram as shown in Fig. 2. Fig. 2. OFDM transmission system with blanking scheme at receiver. In OFDM transmission, the time-domain form of transmitted signal s(t) is found by applying inverse Fourier transform to the frequency-domain signal by using 11 2 . . ( ) exp , 0 t T S0 N k t s t S j k TN k S (1) Where S k denotes frequency-domain signal, N is number of subcarriers, 1j , and sT is active symbol interval. Impulsive Noise (IN) is represented as ki and is included to the transmitted signal using Bernoulli-Gaussian random process as follows: , 0 k N-1k k ki b g (2) Where ‘ 1kb ’ or ‘ 0kb ’ according to Bernoulli process of sequence. The probability for ‘ 1kb ’ is p and ‘ 0kb ’ 1 p . Whereas, kg is additive white Gaussian noise (AWGN) with variance 22 (1/ 2) [ ].i kE g Hence, the signal-to-IN ratio (SIR) can be given by; 2 1010log (1/ )iSIR . It is worthwhile mentioning here that in this paper we represent theoretical signal to impulsive noise ratio as SIR and practical signal to impulsive noise ratio as SINR. The time-domain signal ( kr ) received after addition of IN and AWGN can be represented as if 0 , if 1 k k k k k k s b r s i b (3) The received signal is passed to the blanking non-linearity block. The output of blanking non-linearity ky is then fed to the OFDM demodulator for the further processing and is represented as , 0 k N-1k k ky r d (4) Where kd represents the blanking non-linearity scheme and is given as 1 0 k k k r T d r T (5) Where T is the blanking threshold level and 0, 1, ..., 1k N . Optimizing the value of T for increasing effectiveness of blanking non-linearity is described in the next section. III. THRESHOLD OPTIMIZATION A. Overview of Threshold Optimization Techniques An optimized threshold is very important for efficient blanking of Impulsive Noise. Two threshold optimizing methods are discussed in this section. 110 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 120. An Adaptive Threshold(AT) method is reported in [11]. AT is the threshold which gives maximum signal to impulsive noise ratio (SINR) according to (6). max( ( ))AT SINR T (6) AT for an OFDM signal is found by applying Blanking nonlinearity to IN affected received symbols at all thresholds varying from 0 to 15 with an increment of 0.1 and calculating SINR each time. The threshold which gives maximum SINR is denoted by AT as shown in Fig. 3. 0 5 10 15 0 5 10 15 20 Threshold SINR AT at max(SINR) Fig. 3. SINR after applying blanking nonlinearity with respect to threshold and AT, SIR=-15 dB, p=0.01. Another optimized threshold is known as Dynamic Peak Based Threshold (DPBT) [12]. In DPBT estimation method the peak of the OFDM signal is calculated before it is transmitted through a noisy channel. At OFDM receiver, blanking nonlinearity scheme is used with threshold equal to the peak of the original signal as shown in equation (7). max( ( )), 0 k N-1kDPBT s dB (7) Where DPBT denotes the threshold in Dynamic Peak based Threshold Estimation Method. Both of the optimized thresholds; AT and DPBT cannot be used practically as it is impossible to find the exact peak of transmitted signal and maximum SINR at OFDM receiver [12]. B. The Proposed Median Based Threshold (MBT) Calculation Method Due to practical shortcomings of the techniques discussed above, a new technique is therefore devised to calculate an optimized threshold, known as MBT, by using the distribution characteristics of the received signal which are its peak and median. Fig. 4 illustrates the method to calculate MBT. Figure 4. Method to calculate Median Based Threshold. MBT is given as: max( ) ( ) ,0 k N-1k kr median r MBT (8) Where kr is received signal, max(rk) denotes the peak of the received signal and is a threshold slope to be fixed at the receiver. is set to achieve high SINR. Its optimum value lies in the range for 4 to 6. In our simulations, it is kept 4 which was found to be optimized by using hit and trial method. Setting an optimal value of is a better approach as compared to selecting a fixed optimal threshold because a threshold value may lie from 1 to 15 but the range of β is comparatively lesser and does not vary with different probabilities of IN occurrence, unlike a fixed threshold. Performance comparison of the three methods is shown in the next section. IV. SIMULATION RESULTS The OFDM system is simulated using MATLAB with 16- QAM modulation, SIR=-15dB, number of subcarriers, N = 10^5, and 22 (1/ 2) [ ]i kE g . Probability (p) of presence of IN is taken as p=0.01, which suggests that 1% of the received samples are corrupted by Impulsive Noise. The output SINR is obtained by 2 2 [ ] [ ] k k k E s SINR E y s (9) Fig. 5 shows the SINR achieved by three optimized 111 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 121. thresholds; AT, DPBTand MBT. It is noted that MBT mostly results in a higher SINR than DPBT but in a lower SINR than AT. However, the difference in SINR between MBT and AT is not less than -0.64dB, which is quite negligible. This difference in terms of SINR (Gain/Loss) is further illustrated in Fig. 6. MBT results in a maximum of SINR gain of 1.2 dB when compared with DPBT and 0.02 dB gain when compared with AT . The relative SINR gain/loss is evaluated as ( / ) ( / )DPBT MBT DPBTSINR Gain Loss SINR SINR (10) ( / ) ( / )AT MBT ATSINR Gain Loss SINR SINR (11) Fig. 5. SINR from 100 simulations after applying blanking nonlinearity using DPBT, AT and MBT. From Fig. 5 and Fig. 6, it is clear that the performance of AT is better than MBT and DPBT. Therefore, MBT is desired to be closer to AT in order to achieve optimum performance. Fig. 7 shows that, as desired, the threshold values obtained by MBT are near to AT which maximize SINR as compared to the values of DPBT or peak of transmitted OFDM signal. System performance of blanking nonlinearity with the three optimized thresholds is also compared in terms of SER by using 2 1 3 1 1 2(1 ) ( ) 1 SER Q mm (12) Where m is modulation index, Q denotes Q function and γ denotes SINR. Fig. 8 shows that MBT outperforms DPBT in resulting SER of the system but the SER of MBT is slightly higher than or equal to that resulted by AT . 0 10 20 30 40 50 60 70 80 90 100 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 Number of simulations SINRGain/Loss(dB) MBT with AT MBT with DPBT Fig. 6. SINR Gain/Loss of MBT relative to DPBT and AT after applying blanking nonlinearity. 16 16.5 17 17.5 18 18.5 6.5 7 7.5 8 8.5 9 9.5 SINR(dB) Threshold AT DPBT MBT Fig. 7. Thresholds (AT, DPBT, and MBT) relative to SINR after applying blanking nonlinearity. To check the authenticity of the proposed technique for threshold optimization, SER was also obtained using AT, DPBT and MBT with blanking non-linearity and making the OFDM signal corrupted by impulsive noise with probability of 0.03. p=0.03 implies that 3% of the OFDM signal samples would be affected by Impulsive Noise. The plot of SER with respect to SINR in Fig. 9 shows that the performance of MBT is close to AT. However, the performance of DPBT is degraded, which means that DPBT is not feasible of all probabilities of Impulsive Noise occurrence. 112 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 122. 5.5 6 6.57 7.5 8 8.5 9 10 -0.4 10 -0.3 SINR(dB) SER AT DPBT MBT Fig. 8. SER with respect to SINR after applying blanking nonlinearity using AT, DPBT and MBT, p=0.01. 16 16.5 17 17.5 18 18.5 10 -4 10 -3 10 -2 SINR (dB) SER AT DPBT MBT Fig. 9. SER with respect to SINR after applying blanking nonlinearity using AT, DPBT and MBT, p=0.03. V. CONCLUSION Two methods of optimizing thresholds of blanking nonlinearity at OFDM receivers are reviewed. Both of the methods are purely theoretical and infeasible to use in practical OFDM receivers. One of the methods suggested peak of transmitted OFDM signal as the optimized threshold but it only worked for a single probability of impulsive noise occurrence and did not give high SINR at other probabilities. A new method of calculating optimized threshold based on peak and median of the received signal samples is proposed which can be used practically. This method results in higher SINR and lower SER than the method which calculates threshold by determining the original signal peak for different probabilities of Impulsive Noise occurrence. The proposed method is sometimes incapable of achieving the maximum SINR as that achieved by adaptive threshold optimization method. However, the SINR achieved by this method is always close to the maximum and acceptable in practical scenarios. Acknowledgment We gratefully acknowledge Training Division, HR Dte Gen, Pakistan Space and Upper Atmosphere Research Commission for their kind coordination in publishing the article. References [1] S. H. Han, and J. H. Lee, “An overview of peak-to-average power ratio reduction techniques for multicarrier transmission,” IEEE Personal Communications, vol. 12, no. 2, April 2005, pp. 56–65. [2] X. Liu, M. Liang, Y. Morton, P. Closas, T. Zhang, and Z. Hong, “Performance evaluation of MSK and OFDM modulations for future GNSS signals,” GPS Solutions, vol. 18, no. 2, January 2014, pp. 163- 175. [3] A. Vaneli-Corali, G. E. Corazza, G. K. Karagiannidist, P. T. Mathiopoulosl, D. S. Michalopoulost, C. Mosquera, S. Papaharalabost, and S. Sca1ise, “Satellite Communications: Research Trends and Open Issues,” International Workshop on Satellite and Sapce Communication (IWSSC’07), IEEE, 2007, pp. 71-75. [4] S.V. Zhidkov, “Impulsive noise suppression in OFDM-based communication systems,” Consumer Electronics, IEEE Transactions, 2003, vol. 49, no. 4, pp. 944-948. [5] M. Katayama, T. Yamazato, & H. Okada, “A mathematical model of noise in narrowband power line communication systems,” IEEE Journal on Selected Areas in Communications, vol. 24, no. 7, 2006, pp. 1267- 1276. [6] Y. Wu, “Performance comparison of ATSC 8-VSB and DVB-T COFDM transmission systems for digital television terrestrial broadcasting,” IEEE Transactions on Consumer Electronics, ,vol. 45, no. 3, 1999, pp. 916-924. [7] S.V. Zhidkov, “Analysis and comparison of several simple impulsive noise mitigation schemes for OFDM receivers,” IEEE Transactions on Communications, vol. 56, no. 1, 2008, pp. 5-9. [8] U. Epple, K. Shibli, and M. Schnell, “Investigation of blanking nonlinearity in OFDM systems,” Iternational Conference on Communications (ICC), June 2011, pp. 1-5. [9] S.V. Zhidkov, “Performance analysis and optimization of OFDM receiver with blanking nonlinearity in impulsive noise environment,” IEEE Transactions on Vehicular Technology, ,vol. 55, no. 1, 2006, pp. 234-242. [10] S.V. Zhidkov, “On the analysis of OFDM receiver with blanking nonlinearity in impulsive noise channels,” Proceedings of International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS 2004), November 2004, pp. 492-496. [11] U. Epple & M. Schnell, “Adaptive threshold optimization for a blanking nonlinearity in OFDM receivers,” Global Communications Conference (GLOBECOM 2012), IEEE, December 2012, pp. 3661-3666. [12] E. Alsusa & K. M. Rabie, “Dynamic peak-based threshold estimation method for mitigating impulsive noise in power-line communication systems,” IEEE Trans. on Power line, vol. 28, no. 4, October 2013, pp. 2201-2208. This article is sponsored by Pakistan Space and Upper Atmosphere Research Commission 113 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 123. A Review onMobile Wireless Communication Networks (0G to upcoming generations) F.J.Sheikh 1 Dept. of CS & IT University of Lahore, Lahore farzanjavedsheikh@gmail.com K.S.Mughal 1 Dept. of CS & IT University of Lahore, Gujarat khadija.shakeel10@gmail.com I.T.Cheema1 Dept. of CS & IT University of Lahore, Gujarat cheema.cheemas@gmail.com Mujahid Afzal2 Dept. of CS & IT University of Lahore, Gujarat mujahid.afzal@uog.edu.pk Abstract—Advancements in mobile communication have modernized the way people communicate, interact, and share data and messages to each other. Latest technologies provide these facilities to user in a short time. The technology of mobile communication took its start from zero generation (0G) and now fifth generation (5G) is near to overcome the world. Over as far back as decade, remote innovation organization needs undergone gigantic Growth. Zero generation is known as mobile radio telephone, the predecessor of first generation. The 1st generation (1G) has covered the essential portable voice. The 2nd generation (2G) has announced the size and enabled the network to transfer data, messages and MMS and followed by the third generation (3G), which fulfill its mission to send data at high speed and enough flexible to support 5 major radio technologies, this further acknowledged by the fourth generation (4G) which broadened the high velocity advancements and high versatility through LTE (Long Term Evolution) and the fifth generation (5G) mobile network correspond the framework with high data transfer capacity and wide scope region. This paper will present the evaluation of mobile wireless communication from 0G to upcoming technologies. Evaluation include the wireless access techniques used in all generations of mobile wireless, and IPWireless technology used in all technology, purpose of LTE in 4G. What are the reasons and limitations that the new generations or technologies are now replacing the old ones? This paper also explores the need of upcoming generations that will present the image of future wireless interaction in mobiles. Keywords —1G; 2G; 3G; 4G; 5G; MMS; radio technologies; LTE. I. INTRODUCTION Data transmission without the use of improved electrical conductor or wires over a distance is known as Wireless communication. The area elaborate might be near (a limited meter) or far away (thousands or millions of km for broadcasting infrastructures). Some terms are often reduced to “Wireless” due to indistinct framework that includes many types of static, portable, movable, cellular phones, PDA and wireless network. From the past few periods, technology of mobile wireless (knowledge of many genera’s) named as 0G to 4G. The operations of 5G technology or new upcoming technologies can make improvement in usage of internet and browsing data on WWW. Respectively every genera has more or less principles, capabilities, procedures and innovative functions over others that separate them in others. The figure of cell phone customers are growing with passage of time due to new structures. At Bell’s test center, which was a cellular system (1960-70), the development of wireless communication was considered as a product for the improvement of extremely consistent, tiny and solid state RF hardware. The development of the mobile system, offer few benefits: Function in least bandwidth, it also offer customer and consumer fulfillment by the use of variety adeptness feature. In comparison of 'wired', wireless interaction are, in many circumstances, low-cost to configure. Cost feature is also reliable. Because of the geographical area covered, number of customers cannot build blocking probability. So capacity structure must be high for mobile telephone. II. ZERO GENERATION TECHNOLOGY (0G) In earlier days, there were only one or two stations available for mobile operators to set up a call. First modern cellular mobile telephony technology was “Mobile transistor telephone system”. Since these structures are staying the prototype of the first generation of cellular telephone and indicates as zero generation system (0G). Technologies used in 0G methods involved Push to Talk (PTT), Mobile Telephone System (MTS), Improved Mobile Telephone Service (IMTS), Advanced Mobile Telephone System (AMTS), Norwegian for Offending Land Mobile Telephone (OLT), Public Land Mobile Telephony) and Swedish abbreviation for Mobile telephony system(MTD). Push to Talk system are Half-duplex communication that can be two way handshaking or present connection that is called Push to talk and it is working on personal computer to portable through internet. MTS framework was work bolstered in just as instructions, if one calls by using public switch telephone network (PSTN). Versatile administrator who might request one's portable number for call and the other numbers being called place the specific call. When receiver was picked up from frame, IMTS units form a specific dial bell and with this approach it appeared as wired phone and less a mobile phone. Almost 40-60 miles area was protected by IMTS in diameter. The frequency channel in bigger cities are large while rural stations have one or two stations. 114 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 124. AMTS overawedall the problems that occur from IMTS and works on 900 MHz In Norwegh, OLT which worked on 160 MHz VHF band was the first land mobile telephone network. In regularity inflection it worked on 160 -162 MHz for the mobile portion and in base station it is 168- 170MHz. MTD include high speed internet structure with fixed wireless facility that deals constantly on internet access. III. FIRST GENERATION (1G) The 1st genera of mobile interaction was presented in the 1980 and analog transmission was used as communication. The 1st mobile system (1979) developed by Nippon Telephone and Telegraph (NTT) in Tokyo, Japan. Mobile Telephones (NMT) and Total Access Communication Systems (TACS) were two most popular system of analogue system [17]. Frequency division multiple access (FDMA) with channel capacity of 30 KHz and frequency band 824-894 MHz use incidence variation system for radio communication based is derived from a technology called Advance Mobile Phone Service (AMPS) [14, 15]. Fig 1. 1G Mobile Phone [6] The initial step of mobile communication is NTM. It is in the result of heavy and increased usage of physical cellular communication. The variations of NMT are NMT-450 and NMT-900. The figures specifies usage of band rate correspondingly. NMT-900 contain other Channels as compared to NMT-450. NMT cells cast-off full duplex program. Permitting in real-time acceptance and vocal sound communication [14]. AMPS was a reporter cellular telephone framework typical creates in 1983 by Bell Labs, America. It was an initial genera mobile phone program which use isolated frequencies in place of individually discussion. It was changed commencing more seasoned arrangements by "back-end" setup of calls usefulness that permitted bigger telephones quantities to be bolstered concluded geological region. In 1983, an aggregate of 40MHz of range in the 800 MHz band was assigned by FCC for AMPS. In 1989, additional 10 MHz has been charged to AMPS because of ascent of cell framework limit. Uplink frequency for AMPS remains 824- 849MHz and downlink frequency stays at 869-894MHz. Independently AMPS frequency is 30 KHz wide [16]. Fig 2. Architecture of Advance Mobile Phone Services (AMPS) [6] IV. SECOND GENERATION (2G) After first genera similarity mobile statement system 2G mobile system was announced around 1991.concept of 2G based on many base stations where each station circulated consistently over the world to communicating with the users. To converse more and more users various access procedures are used i.e. FDMA, TDMA, and CDMA [10]. 2G technology make procedures of compression decompression procedure (codec) and family members of this generation are 2G (GSM, GPRS, EDGE). Digital signal usage among recipients and tower build the framework ability in two forms 1) Digital sound data be able to smashed and multiplexed proficiently in this way consenting more data is delivered into the equal quantity of radio data transmission. 2) Less radio power is released while using digital systems. Therefore creation of small size of and altering better mobiles in similar quantity of space. Two standard was used for 2nd genera technology – CDMA and TDMA on center of multiplexing. In drill, TDMA and CDMA systems share with FDMA. TDMA expression is utilized to assign frameworks, which partition the channel into frequency openings, later on isolate every rate space into various time openings. Also, CDMA is truly a cross breed of CDMA and FDMA where the station is initially partitioned into recurrence openings. 115 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 125. Fig 3. 2GMobile Phone [6] GPRS is a cell phone wireless technology established amongst prototype, 2G, and predecessor, 3G) [12]. It’s a radio innovation for GSM systems that upgrades bundle changing over conventions, set up period for ISP systems, choice to pay according to the sent data, instead of association period. In packet transferring method, material to be delivered is damaged into packets, size of the packets is in kb, after that these are moved through network among various ends, each data packet contain address of destination. Data transmission and network connection is the provision of GPRS. EDGE is an ordinal cell phone technique that actions as a bolt-on enrichment to General Packet Radio Service (GPRS) networks. It is ideal as compared to GSM because it carries packet switch data and circuit switch data [11]. Black berry N97 and N95 mobile phones are famous because of EDGE technology. By using EDGE technology, no additional hardware and software resources are required for making EDGE technology work. Any of the extra expense are not required for developing such technology [13]. Fig 4. GPRS Architecture [6] V. THIRD GENERATION (3G) 3G is the innovative generations for cellular interaction facilities, its services area centered on procedural values of IMT-2000 containing the consistency, accuracy and also speed (rate in which data is transferred) that is minimum 200kb per second [11]. Outside of cell phone, more data transfer ratio permitted 3G links in personal computers, digital gaming equipment, tablets, and smart phones including all devices that are able to take benefit by using high and fast quality of internet usage. Working capacity of devices and functionality of devices can be increased many times with the usage of fast and reliable internet connection. Better safety and less threat is offered while connecting and transferring data to other wireless devices by using 3G [2, 3]. Three basic technologies are covered by 3G that is CDMA2000, TD-SCDMA - Time- division Synchronous Code-division Multiple Access and W- CDMA (UMTS) Wideband Code Division Multiple Access. The vital role of 3G knowhow is, it offers high transfer of data transmission and improved speed. Packet-switching technique is used in 3G that is considered more realistic and also much faster as compared to the previous used technologies and techniques, issue with it is it requires a totally different groundwork as compared to the 2G systems. Fig 5. CDMA Architecture [6] Mobile communication market is increasing on a very fast pace, usage of mobile phone industry is increasing in leaps and bounds. Almost one billion customers globally are benefiting from cellular phone industry, in which two-thirds are GSM customers. One of the firmest growing wireless technology is known as CDMA, it is developing its subscribers base signify cantle around the world. Nowadays CDMA subscribers are crossing the line of 200 million users. 3G networks enable network workers to deal customers an extensive variety of latest progressive facilities and also realizing better and superior connection ability through healthier spectral efficiency. Fig 6. 3G Mobile Phone [6] 116 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 126. VI. FOURTH GENERATION(4G) 4G technology uses 3G as a inheritor, developed for the broadband capabilities with the services of 3G to enable the high data rate streaming i.e.20 Mb/s per customer, Quality of services, Fast Speed, Faster Response Time, Routing efficiency, Higher Bandwidth and Lower Power consumption. Speed for low mobility is 1 gigabit per second (Gbit/s) and 100 megabits per second (Mbit/s) for peak speed. NTT Do Co Mo was first successor to achieve. Downloading of packet transmission at a moving speed of almost 20km/h is 1Gbps in Tokyo, Japan on June 23rd, 2005 [2]. 4G technology is combination of IP-Based voice, Seamless access, Personalization, Quality of services. Its main purpose is to access network anytime, anywhere and anyhow. 4G wireless communication technology is furthermore discussed as “MAGIC”. MAGIC word stands for mobile multimedia, any- where, global mobility solutions over, integrated wireless, customized services. Need for this platform is to integrate pervious mobile technologies entirely which are present that is GSM, GPRS, IMT-2000, Wi-Fi, Bluetooth, WCDMA, HSDPA, CDMA2000 and EVDO, provide correspondence to user expectations for various services. Mobile WiMAX regularities and Long Term Evolution (LTE) rules were deployed for 4G but failed to satisfy the IMT- Advanced requirement, may well though measured as 4G. Later on LTE Advanced (latest release of LTE) standardized by the 3GPP (Third Generation Partnership Project) and 802.16m uniformed by IEEE (that is WiMAX). Demand of IMT-Advanced for the generations of interactive mobile services by providing a global platform is increased to gain unified messaging, quicker information admittance, improved roaming competencies and broadband multimedia [16]. A. LTE Advanced LTE is not such a new technology but some essential enhancements in LTE to improve existing LTE network. It was officially candidate of ITU-T (International Telecommunication Union- Telecommunication), meeting the requirements of the IMT-Advances standard and also standardized by 3GPP. LTE Advanced has 30bps/Hz peak spectrum downlink value and 15 Bps/Hz peak spectrum uplink value, can be enhanced if multiple-input multiple- output (MIMO) (that is antenna arrays) remain used. Both coordinated multipoint (COMP) and MIMO remained recognized such as the crucial procedures for LTE [2]. Fig 7. LTE Advance Architecture [17] VII. FIFTH GENERATION (5G) 5G (Mobile and wireless Network) can make the perfect wireless real world without any limitation is known as World Wide Wireless Web (WWWW). Beyond the 4G/IMT-Advanced standards 5G symbolizes the next utmost chief level of cellular phone technology genera. 5G is still not considered and used as an official term on behalf of any specific requirement and design and also not publically used by telecommunication companies or standardization such ITU-R (International Telecommunication Union-Recommendation), 3GPP, WiMaxForum (Worldwide Interoperability for Microwave Access) [2]. Every new release of technology enhances the system performance with some advance capabilities. 5G technology must be something else than previous technologies, whose purpose is not only to increase the throughput but also consider the other requirements. Improved coverage of geographical zone and high rate for data transfer at the edge of the mobile phone Truncated power and battery consumption Accessibility of numerous data transmission tracks Nearby 1 Gbps data rate is definitely thinkable Additional Security Energy effectiveness and spectral efficiency are good [17] 117 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 127. Fig 8. 5GMobile Phone [17] 5G mobile technology will be powerful and demanding in future due to including most powerful features. Also it will change the concept of using mobile phones with more bandwidth. Customers certainly not have experience of such great and fast tracking technology. The architecture of the fifth generation consists of cellular networks interoperability (contains of a customer station and number of autonomous radio access technologies (RAT)) and wireless IP based model [5]. User terminal has a significant role in this 5G technology system. Proposed 5G architecture is shown in fig 8. 5G technology deal with camera, large phone memory, MP3 recording, dialing speed, video player, audio player and many creative, innovative and imaginative technologies [13]. Fig 9. Functional Architecture for 5G- Network [18] VIII. SIXTH GENERATION (6G) The 6G technology will combine the 5G satellite network and mobile wireless system. Satellite network include telecommunication satellite, navigational satellite, imaging satellite, environmental information system. These system used for data, voice, video broadcasting and internet; weather; environmental information Collection; and global positional system (GPS) respectively [2]. At the same time Fifth generation will focus on user terminals but it is necessary for terminals to have access of different wireless technologies. Roaming and handoff will be the big issue of 6G because of the different network stations. Solving the problem of handoff and roaming is still a question. But the charges of mobile call will be relatively high [8]. IX. UPCOMING GENERATIONS Upcoming technologies will provide the real image of wireless world with less boundaries. Artificial intelligence and ubiquitous network will provide universal computing from anywhere, anytime, and any device. User can instantly linked to several technologies and perfectly move between them. X. COMPARISON We compare all the technologies of Mobile Technology in term of data bandwidth, standards, technology, multiplexing, switching see table 1. [6] XI. CONCLUSION Wireless communication technologies has captured the market since its first start. This study provides the better understanding of the previous and present technologies along with their performances, portals, advantage and disadvantages of one generation to other and also predict incoming future technologies. . REFERENCES [1] F. R. Y. Chengchao Liang, "Wireless Virtulization for Next Generation Mobile Cellular Networks," in IEEE Wireless Communications, February 2015. [2] P. Sharma, "Evolution of Mobile Wireless Communication Networks- 1G to 5G as well as Future Prospective of Next Generation Communication Network," in International Journal of Computer Science and Mobile Computing, August 2013. [3] E. M. I. A. E. U. M. A. Engr. Muhammad Farooq, "Future Generations of Mobile Communication Networks," in Academy of Contemporary Research Journal, January 2013. [4] A. G. Roopali Sood, "Digital Society from 1G to 5G: A Comparative Study," in International Journal of Application or Innovation in Engineering & Management (IJAIEM), February 2014. [5] R. Rajan, R. Baweja and R. Kumar, "Review on Different Multiple Access Technique Used in Wireless Communication," in International Journal of Research (IJR), November 2014. [6] P. S.Venkata Krishna Kumar, "A Study of Wireless Mobile Technology," in International Journal of Advanced Research in Computer Science and Software Engineering, January 2014. [7] V. G. APANA BHAT, "EVOLUTION OF 4G: A STUDY," International Journal of Innovative Research in Computer Science & Engineering (IJIRCSE), February 2015. [8] A. S. J. R. Sarmistha Mondal, "A Survey on Evolution of Wireless Generations 0G to 7G," International Journal of Advance Research in Science and Engineering- IJARSE,. [9] A. A. J. I. Nabeel ur Rehman, "3G Mobile Communication Networks," EXPLORE SUMMER 2006, 2006. [10] N. Schmitz, "The Path To 4G Will Take Many Turns," 1 March 2005. [Online]. Available: http://electronicdesign.com/4g/path-4g-will-take- many-turns. [11] A. G. R. S. O. Z. Xichun Li, "The Future of Mobile Wireless Communication Networks," 2009 International Conference on Communication Software and Networks, 2009. [12] A. R. Mishra, Fundamentals of Cellular Network Planning and Optimisation: 2G/2.5G/3G... Evolution to 4G, John Wiley & Sons ©2004, 2004. [13] A. F. Molisch, Wireless Communications, Wiley, 2010. [14] F. ADACHI, "Wireless Past and Future--Evolving Mobile Communications Systems," IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences, 2001. 118 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 128. [15] D. Y.L. D. J. S. D. Amit Kumar, "Evolution of Mobile Wireless Communication Networks: 1G to 4G," International Journal of Electronics & Communication Technology, December 2010. [16] A. Pashtan, "wireless terrestrial communication: cellular telephony," TELECOMMUNICATION SYSTEM AND TECHNOLOGIES, 2006. [17] G. M. K. A. Arun Agarwal, "The 5th Generation Mobile Wireless Networks- Key Concepts, Network Architecture and Challenges," American Journal of Electrical and Electronic Engineering,, vol. 3, no. 2, pp. 22-28, February 2015. [18] [Online].Available:http://www.prismatelecomtesting.com/products/prod ucts-overview/radio-interface-simulation/uesim/. TABLE 1. COMPARISON OF MOBILE TECHNOLOGIES Generation Features 1G 2G 3G 4G 5G Years 1980s 1990s 2000s 2010s 2020s Data Bandwidth 2kbps 64kbps 2Mbps 200Mbps 1Gbps Standards AMPS TDMA, CDMA,GSM, GPRS WCDMA Single unified standards Single unified standards Technology Analog cellular Digital Cellular Broadband with CDMA, IP Technology Unified IP & Seamless combination of broadband, LAN,WAN & WLAN Unified IP & Seamless combination of broadband, LAN,WAN & WLAN,WWWW Service Mobile technology (voice) Digital Voice, SMS, Higher Capacity Packetized Integrated high quality audio, video & data Dynamic information access, wearable Devices Dynamic information access, wearable Devices with AI capabilities Multiplexing FDMA TDMA,CDMA CDMA CDMA CDMA Switching Circuit Circuit & Packet Packet All Packet All Packet 119 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 129. Intelligent Detection ofDistributed Flooding Attack in Wireless Mesh Network M.Altaf Khan Institute of Information Technology Kohat University of Science & Technology Kohat, Pakistan altaf_apc@yahoo.com Shafi Ullah Khan Institute of Information Technology Kohat University of Science & Technology Kohat, Pakistan skkust@hotmail.co.uk Amjad Mehmood Institute of Information Technology Kohat University of Science & Technology Kohat, Pakistan amjadiitkust@gmail.com Zeeshan Iqbal Institute of Information Technology Kohat University of Science & Technology Kohat, Pakistan Zeeshan.zn@gmail.com Abstract—Multi-hop wireless mesh network (WMN) is a growing and promising technology that fulfills the requirements of next generation wireless networking by providing an extensive assortment of applications that were not assumed to be supported by a wireless network. In WMN, a Gateway is the only way of integrating it with other wireless and wired networks. Due to this critical role, gateways in WMNs are targeted by diverse security threats. A distributed denial of service (DDoS) attack is one of the prime security threats to WMN. DDOS attack is an explicit attempt to prevent the legitimate users from accessing their service. In a distributed flooding attack, gateway can be easily exploited by the attackers with the aim to reduce the available network bandwidth for the valid users. Hence network performance is reduced and also destruct the services provided to the end users. Numbers of detection mechanisms have been provided by researchers in this regard. However, due to advancements in the attacking tools and techniques, more intelligent efforts are required in terms of precision and competence, to protect the WMNs from such attacks. In this paper, an intelligent mechanism called Distributed Flooding Attack Detector (DFAD) is proposed with the aim to detect distributed flooding attacks in WMNs. Network simulator (NS2) is being used to evaluate the performance of DFAD. Keywords—Wireless Mesh Network, Distributed Denial of Service Attack, Mobile Ad-hoc Network, Artificial Neural Network, Back-Propagation I. INTRODUCTION A multi-hop and multi-radio wireless mesh network (WMN) is a decentralized, dynamically self-organized and self-configured network. Connections among the nodes in WMN establish without any central controlling body using partial or complete mesh topology [1]. Two types of nodes are existed in a typical WMN i.e. mesh node (MN) and mesh router (MR). MN may either be static or mobile having one wireless interface. MRs forms a multi-hop backbone of the WMN by providing self-configured and self-healed links. Sometimes, in WMN, a MR may have both wired and wireless interfaces which enables it to provide functionality of gateway. Unlike conventional routers, MR exhibits some additional routing functions to bear mesh networking. Moreover, due to multi-hop support, mesh routers provide more coverage with low transmission power than conventional routers. WMNs can be classified into three main architectures i.e. infrastructure- based, infrastructure-less and hybrid WMN [2]. Infrastructure- based WMN is composed of both MNs and MRs. Infrastructure-less WMN seems to be a pure mobile ad-hoc network (MANET), because it is consisted of only MNs that may also carry out routing and self-configuration to form a peer to peer wireless network. Hybrid WMN is composed of MNs and MR, where MRs are used to form backbone of WMN and MNs offer routing abilities in order to achieve an enhanced coverage and connectivity in the interior WMN. In both infrastructure-based and hybrid WMNs, through mesh gateway (MG), a node in WMN can communicates with other existing networks. Hence, MG plays a vital rule for instigation or termination of traffic in WMN. So it is mostly targeted by intruders during different attacks. Once MG is being compromised, all the provided services of WMN are affected and hence degrades its performance. WMNs have made life of its users easier. However, some serious security threats may be faced by the WMN that may compromise the confidentiality, integrity or availability of a WMN [3]. Confidentiality is being compromised through several attacks like eavesdropping, traffic analysis, corrupt access point, war driving attack, and brute force attack. Black- hole, worm-hole, grey-hole etc. are the attacks through which integrity of WMNs are being compromised. Availability is compromised by many DOS attacks including TCP SYN- flooding attack, UDP-flooding attack and smarf attack, where intruders try to prevent normal services or to make the network resources unavailable for the legitimate users. When a DOS attack is implemented by thousands of attackers simultaneously, with the aim to increase the severity of attack, it is identified as DDOS attack. An intruder launches such attack by sending a massive bulk of ineffectual packets concurrently towards victim, through huge number of compromised nodes also known as zombies [4]. These zombies may either be internal or external to the victim’s network. Distributed flooding attack (DFA) is a one of the classical approach to launch DDOS attack in WMNs using 120 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 130. tools like Trinoo,Tribe Flood Network (TFN), Shaft, Knight and Trinity [5]. Anomaly based DDoS attack is being detected in [6]. Features of malicious packets are analyzed using radial base function neural network (RBFNN). A vector of seven features is being used for activation of RBFNN and classification of incoming packets flows into classes of attack traffic or normal flows. Proposed method is evaluated through datasets of UCLAD. Proposed method classified the incoming traffic into class of normal or attack traffic, but it cannot classify and identify the type of attack. A statistical approach, based on values of numerous features is implemented in [7] to detect attack traffic. Proposed method used a module for features extraction and then extracted features are stored in a database for identification and classification of normal and malicious packets flows. A novel approach is presented in [8] to detect flooding and IP spoofing attack. Prevention of flooding attack is accomplished through considering clock values of individual node of network. A combined data mining approach is used in [9] against DDoS attack for detection of protocol anomaly. In the proposed method for extraction of features network traffic is utilized and then a classification algorithm based on data mining is used to cluster incoming traffic into normal and attack flows. For detection of such attack numerous techniques are proposed so far, but the techniques that are based on intelligent schemes are proved to be more proficient alternatives in this regard [10,11,12] . Still there is a need to propose more mechanisms to handle this attack. Our proposed mechanism is more intelligent and reliable then the mechanism proposed earlier for detection of DFA in WMNs. II. PROPOSED MECHANISM Due to self-configuration and dynamic changes in topology of WMN, ANN is considered to be the most suitable technique amongst all other intelligent techniques to detect distributed flooding attack [5]. Our proposed mechanism i.e. distributed flooding attack detector (DFAD) is based on ANN technique and is adaptive in nature. ANN used the concepts of biological neural networks found in human brain. With the assistance of this network a brain involves millions of interconnected neurons to fully control the human body. Fig.1 depicts the basic structure of a biological neuron. It illustrates some important components of a neuron. Actual cell body of the neuron is constructed by Soma. It is composed of a nucleus and plasma. Information regarding to innate traits is provided by the nucleolus and plasma generates material required by neurons. Dendrites and Axons are used during communication with other neurons. Through synaptic terminal an Axon spreads out received signal to other neurons [13]. Following the same concepts, ANN is a network of interconnected nodes (neurons) that provide required solution after performing parallel processing. Structure of earliest ANN called perceptron, is given in the Fig.2. Using equations (1,2,3), If Sum of the product of binary inputs (Z1, Z2…. Zn) and their associated weights (W1, W2…....Wn) is found greater or equivalent to “0” then “f” i.e. an activation function will turn out “1” as output if not, “0” will be the output of the perceptron [14]. (1) ( ) (∑ ( )) (2) ( ) { (3) Two types of structures are existed in ANNs i.e. feed- forward and feed-back neural networks. Main difference between both structures is the flow of signals. Feed-forward neural network allows information flow in just forward direction while a feed-back neural network allows information to be flowed in both forward and backward directions[15]. These structures are mainly distributed in three main layers: input-layer (IL), hidden-layer (HL) and output-layers (OL). IL receives input for the network and the OL provides the results of the network. HL offers the neural network with its aptitude to generalize. Moreover, HL may be composed of more than one layer. There may be different number of neurons in each layer. As quantity of layers and nodes in HL increases, complexity of ANN will be increased. On the other hand, probability of accurate learning is increased. A feed-forward ANN is illustrated in Fig.3. Soma Dendrites Axon Nucleus Synaptic Terminals Output (0/1) X1 X2 X3 Xn Fig.1. Structure of a biological neuron Fig.2. Structure of a perceptron 121 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 131. Learning of ANNcan be performed through either supervised or un-supervised learning algorithms. In supervised learning, an output is provided with each input during training while in unsupervised learning algorithms, target outputs are not provided with input data during training. Hence ANNs perform data compression or clustering to categorize patterns of the same attributes into same output clusters [16]. In feed-forward ANNs, most commonly used supervised learning algorithms may include delta rule, perceptron and back-propagation (BP). For learning a feed-forward ANN, BP algorithm is most commonly used amongst entire algorithms. Delta rule and perceptron algorithms are useful to resolve merely linear problems, while BP algorithm is able to provide learning in complicated non-linear problems. During training, in BP based ANN, signals are transferred in forward direction i.e. from input to OL. Error between actual and target output in the OL is calculated in each pass. In BP, learning process includes forward-pass and backward-pass. Input data is moved from IL to OL in forward-pass. Most commonly sigmoid activation function is used in BP algorithms to generate results from each HL and OL node by using following equation (4). Output of this sigmoid function lies between “0” and “1” ( ) (4) Error (E) between target output (T) and actual output (O) is calculated at the end of each forward-pass. Gradient descent technique is used to reduce this error ‘E’ between Oi with respect to Ti. The value of E, gradient descent and required changes in the weights of each connection values are computed through following equations respectively. ∑ ( ) (5) (6) ( ) (7) (8) In equation (6), ‘GrD’ represents value of gradient descent, its value concludes that either to raise or reduce value of weights and biases. ‘ ’ denotes momentum and ‘ ’ represents the learning rate. In equation (7) and (8), and represents the desired change in weights associated with hidden-to-output layers or input-to-hidden layer connections and biases respectively in the backward-pass. In our DFAD, we used multiple layer perceptron (MLP) having three inputs nodes in the IL, because we aimed to check the MG after each five seconds. Each input value represents received number of packets. After trying different number of nodes and different number of layers in the HL, finally we got our results with higher detection rates and lower false rates, using one HL having four nodes in it. As we are interested to distribute the incoming flows at MG into three different categories therefore we used three nodes in the OL to generate one of the desire output i.e. [1,0,0], [0,1,0], [0,0,1] to represent normal, low distributed flooding (LDFA) attack and sever distributed flooding attack (HDFA) flows respectively. During training of DFAD, a vector of eight (08) input values is provided. Arrangement of the input vector is [4,3,7,0,0,1], where [4,3,7] are the amount of packets received in each second and the remaining values of the input vector i.e. [0,0,1] represents the target output for the given input values. We used three different target outputs with each input vectors i.e. [0, 0, 1], [0, 1, 0] and [1, 0, 0], Where [0, 0, 1] is imagined as normal flows, [0, 1, 0] is considered to be an intermediate distributed flooding attack and [1, 0, 0] is expected to be a high distributed flooding attack flows. A minute detail of training dataset is given in the Table.1. Input Vectors 4, 3, 7, 0, 0, 1, 31, 34, 36, 0, 1, 0, 42, 51, 65, 1, 0, 0, 4, 6, 12, 0, 0, 1, 33, 26, 22, 0, 1, 0, 46, 54, 57,1, 0, 0, 16, 6, 17 0, 0, 1, 37, 35, 46, 0, 1, 0, 54, 58, 56, 1, 0, 0,……..….. Weights and Biases 0.11, 0.21, 0.13, 0.14, 0.25, 0.36, 0.17, 0.8, 0.19,-3.0, -4.0, -2.0, 1.6, 1.8, 1.3, 1.6, 1.2, 1.1, 1.7, 2.1,2.3,-2.3, -7.0, -5.0 Assortment of values for initial weights and biases is an actually very tough job. Numbers of forward and backward passes to generate the desired outputs from the network depend on these values. In training, maximum 100 epochs are used with each input vector to minimize the error between target output and actual output. Dataset used to train the network, is consisted of 5000 of input vectors, having all three types of flows, mentioned above. After completion of training, updated values of weights and biases are used by DFAD to distribute the input vectors in desired category in testing phase. Algorithm of proposed mechanism is given in Fig.4. and Fig.5. Main difference between training and testing phase is that, both passes i.e. forward and backward passes of BP algorithms are used in training phase while in testing phase only forward-pass is used. Moreover, in testing phase input vector is entered in the system without target output values. Inputs IL HL OL Table.1. Training dataset Fig.3. Feed-forward artificial neural network 122 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 132. Dataset used totrain the network, is consisted of 5000 of input vectors, having all three types of flows, mentioned above. In training, maximum 100 epochs are used with each input vector to minimize the error between target output and actual output. After completion of training, updated values of weights and biases are used by DFAD to distribute the input vectors in desired category in testing phase. III. SIMULATION RESULTS NS2 [17] is considered to be a best choice for simulations of wired and wireless networks therefore we carried out our simulations in NS2.34. Effectiveness and accuracy of DFAD is tested by performing simulations on NS2. Following parameters are set for our simulations. ======================= Training: ================================== 1. Initialize all weights and biases with small random numbers 2. For ∀ Input vectors in the training set 3. Input Current pattern and target output to the network 4. // Propagated the input forward through the network: 5. For ∀ node in the Hidden layer i. Calculate the sum of product of weights and inputs to the node using Eq.2 ii. Add the biase of each node to the calculated sum iii. Calculate the output using Eq.4 for each node 6. Next 7. For ∀ node in the Output layer i. Calculate the sum of product of weights and inputs to the node using Eq.2 ii. Add the biase of each node to the calculated sum iii. Calculate the output using Eq.4 foreach node 8. Next 9. Calculate sum of error between target and actual output using Eq.5 10. IF ((maximum number of iterations (epochs) < 100) && (Error>=0.25)) 11. // Propagate the errors backward through the network 12. For ∀ node in the output layer i. Calculate Gradient value for node in the output layer using Eq.6 ii. Update each node's weight and biase values in the output layer using Eq.7 and Eq.8 13. Next 14. For∀ node in the hidden layer i. Calculate the Gradient of node's in the hidden layer using Eq.6 ii. Update each node's weight and biase value in the hidden layer using Eq.7 and Eq.8 15. Next 16. With updated weights and biases repeat from Step-5 17. endif 18. Next // select next input vector for training ===================================== Testing: ===================================== 1. Assign all Updated weights and biases obtained from training phase 2. Input Current pattern to the network 3. // Propagated the input forward through the network: 4. For ∀ node in the Hidden layer i. Calculate the sum of product of weights and inputs to the node using Eq.2 ii. Add the biase of each node to the calculated sum iii. Calculate the output using Eq.4 for each node 5. Next 6. For ∀ node in the Output layer i. Calculate the sum of product of weights and inputs to the node using Eq.2 ii. Add the biase of each node to the calculated sum iii. Calculate the output using Eq.4 for each node 7. Next 8. // Output Layer: Node1’s output: out1, Node2’s output: out2,Node3’s output: out1; 9. IF(out1>out2&&out1> out3) 10. printf (“Normal flow”); 11. elseif (out2> out1&&out2> out3) 12. printf(“LDFA-Traffic”); 13. else 14. printf(“HDFA-Traffic”); 15. endif Fig.5. Testing phase Fig.4. Training phase 123 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 133. Simulation time 60sec Number of nodes 100 MG 01 MR (Fixed nodes) 09 Mobile nodes 90 Speed of mobile nodes 5 m/sec Simulation area 500 X 500 Nodes transmission range 250 m To perform training of the DFAD, 40 nodes are used to generate flows with different data rates towards gateway. To generate normal packets flows, we set transmission rates of these nodes to 8Kbps, low and high distributed flooding attack flows is generated with the transmission rates of 15 Kbps to 35 Kbps respectively. Network animator (NAM) i.e. an animation tool usually used by NS2 to generate a view of the simulation is being shown in Fig.5. It shows that packets flows garneted by both mesh nodes and attacking nodes will reach to the gateway through mesh routers. The nodes which are not in the range of mesh routers will use nearby mesh nodes to transfer the packets towards the gateway. Extensive simulations with different data rates were carried out for both UDP and TCP flows. Simulation time is 60 seconds for each experiment. Initial experiment is performed with 40 nodes where each node transmitted data with rate of 08Kbps.In second experiment, there were 30 normal nodes and 10 malicious nodes (attacker). Each normal node transmitted with the rate of 08 Kbps and malicious nodes transmitted with 20 Kbps. In our last experiment, there were 20 normal nodes and 20 malicious nodes. In this experiment the transmission rate of malicious nodes was increased to 35 Kbps. In each experiment the target of the flows was gateway node. After each experiment, DFAD distributed the received flows of mesh gateway accordingly as Normal, LDFA or HDFA attack flows. Distribution of both TCP and UDP flows in each experiment is given in Fig. (7, 8). Furthermore, it is evident from Fig. (9, 10) that with increasing the flows transmission rates in each experiment, packet dropping rate is also increased. Time (Sec) Normal- Flows LDFA-Flows HDFA-Flows Packets Normal Flows LDFAFlows HDFAFlows Time(Sec) Packets Time (Sec) Packets HDFA-Pkt.Drop.Rate LDFA-Pkt.Drop.Rate Normal-Pkt.Drop.Rate Fig.6. Output of NAM Fig.7. Throughput of UDP Flows Fig.8. Throughput of TCP Flows Fig.9. Packet Drops of UDP flows Table.2. Simulation parameters 124 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 134. IV. CONCLUSION ANDFUTURE WORK In WMNs, mesh gateways are mostly susceptible to DDOS attacks. Numerous techniques are proposed to detect distributed flooding attack in WMNs, but still there is a need of an intelligent security mechanism to avoid the influence of intruders on WMNs. An intelligent mechanism based on ANN, is being proposed in this paper to classify incoming flows at gateway into three different categories. Simulations results verified that DFAD is efficient and precise in classifying the incoming flows. In future, DFAD will differentiate flash crowd form attack flows. Moreover, we will also identify and trace back source of the ongoing distributed flooding attack. ACKNOWLEDGEMENT The authors extend their gratitude to the Research Centre, Institute of IT, Kohat University of Science & Technology and the Higher Education Commission of Pakistan, for funding this research work. REFERENCES [1] S. Khan, K. K. Loo, and Z. Din, "Framework for intrusion detection in IEEE 802.11 wireless mesh," International Arab Journal of Information Technology, vol. 7, no. 4, pp. 435-439, 2010. [2] D. Benyamina, A. Hafid, and M Gendreau, "Wireless mesh networks design — a survey," IEEE Communications Surveys & Tutorials, vol. 14, no. 2, pp. 299-310, 2011. [3] F. Xing and W. Wang, "Understanding dynamic denial of service attack in mobile ad hoc networks," in IEEE Military communication conference (MILCOM), 2006. [4] H. Beitollahi and G. Deconinck, "Analyzing well-known countermeasures against distributed denial of service attacks," Computer Communications, vol. 35, no. 11, pp. 1312–1332, 2012. [5] B B Gupta, Joshi, and Manoj , "Distributed denial of service prevention echniques," International Journal of Computer and Electrical Engineering, vol. 2, no. 2, pp. 268-276, 2010. [6] K Reyhaneh and F Ahmad, "An anomaly-based method for ddos attacks detection using rbf neural networks," International Conference on Network and Electronics Engineering IPCST, vol. 11, pp. 44-48, 2011. [7] Thwe Thwe Oo and Thandar Phyu, "A statistical approach to classify and identify DDoS Attacks using UCLA dataset," International Journal of Advanced Research in Computer Engineering & Technology (IJARCET), vol. 2, no. 5, pp. 1766- 1770, 2013. [8] D. J. Jingle and E. B. Rajsingh, "Defending IP spoofing attack and TCP SYN flooding attack in next generation multi-hop wireless networks," International Journal of Information & Network Security (IJINS), vol. 2, no. 2, pp. 160-166, 2013. [9] Thwe Thwe Oo and Thandar Phyu, "DDoS detection system based on a combined data mining approach," 4th International Conference on Science and Engineering, pp. 315-319, 2013. [10] G. Wang, J. Hao, J. Ma, and L. Huang, "A new approach to intrusion detection using artificial neural," Expert Systems with Applications, vol. 37, no. 9, pp. 6225-6232, 2010. [11] S. Khan and K. K Loo, "Real-time cross-layer design for large- scale flood detection and attack traceback," Network Security, vol. 23, no. 5, pp. 9-16, 2009. [12] D. Novikov, R. V. Yampolskiy, and L. Reznik, "Artificial intelligence approaches for intrusion," in IEEE Long Island Systems, Applications and Technology Conference (LISAT 2006), Long Island, NY, 2006. [13] E. Gelenbe, "Learning in the recurrent random neural network," Neural Computation, vol. 5, no. 1, pp. 154–164, 1993. [14] S. M. A. Burney, M. S. A. Khan, and Dr. T. A. Jilani, "Feature deduction and ensemble design of parallel neural networks for intrusion detection system," International Journal of Computer Science and Network Security (IJCSNS), vol. 10, no. 10, pp. 259-267, 2010. [15] G. OKE and G. Loukas, "A denial of service detector based on maximum likelihood detection and the random neural network," The Computer Journal, vol. 50, no. 6, pp. 717-727, 2007. [16] B. B. Gupta and R. C. Misra,M. Joshi, "ANN based scheme to predict number of zombies in a DDOS attack," International Journal of Network Security, vol. 14, no. 2, pp. 61-70, 2012. [17] The network simulator - ns-2 web page,"http://nsnam.isi.edu/ nsnam/index.php/Main_Page". Time(Sec) Packets HDFA-Pkt.Drop.Rate LDFA-Pkt.Drop.Rate Normal-Pkt.Drop.Rate Fig.10. Packet Drops of UDP flows 125 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 135. Touch Panel BasedModern Restaurants Automation using Zigbee Technology 1,* Aamir Nawaz,2 Faiz Jalil Institute of Engineering & Technology Gomal University DIKhan, Pakistan 1 engr.aamir09@gu.edu.pk 2 fj_kulachvi@yahoo.com Abstract In present scenario, automation is in high demand. Management of present restaurants and hotels are much dependent on manpower. This project deals with automation of restaurants with wireless touch panel based menu systems. This project reduces manpower required for taking order from customers. Full menu of all restaurant’s eatable items are displayed on touch panel for selection. Customers have to put their orders through touch panels which will be received in kitchen without any interaction of waiters. Zigbee has been used for wireless link between touch panels of kitchen and restaurant tables. PIC microcontroller has been used for coding of menu display on touch panel and transmitting and receiving on zigbee modules. Project has been implemented and tested in software such as Proteus 8. Furthermore, hardware implementation has been done on PCB layout. By using embedded system with wireless Technology whole procedure will built. The order from selected table number go to server then serve items to selected table customer .We can speed up the serving and attract customer with non – volatile memory capable of retaining data for over ten years .By dumping the code in microcontroller ISP is used Index Terms—Zigbee, Restaurant automation, Micrcontroller I. INTRODUCTION Restaurant business is one of the most profitable businesses. Therefore, the importance of food serving is of great significance. Over the years, food and the relative job of serving has grown so much that need for facilitation and automation has been increased. industries. Nowadays, when profit is the prime concern and its measurement has increased from million to billions, every bit is done to increase profit. We have made an effort to bring technology into the dining menu of customers. This research aims to not only improve the business of restaurants but also to incorporate the essence of science in dining menu. Arun in [1], implemented a train anti-collision and level crossing system based on zigbee technology. He has created four modules which are train system module, control center system module, signaling post system module and level crossing at gate module. He has simulated proposed system in Proteus electronic simulation package. Ramasamy [2] has proposed a fire avoiding system on running train. In his proposed system, train driver will know in case of fire in any section of train and stop it instantly to take necessary action. In [3], Blaho proposed a zigbee wireless network to control speed and other parameters of train model. He has created m- files and Simulink environment which can be used for teaching purposes. In [4], Vinodhini proposed a restaurant system for two way ordering system. He has provided table of customer with touch screen and kitchen with liquid color display (LCD). He has used ARM cortex M3 for simulation. In [5], Wang has introduced zigbee technology for establishing wireless sensor networks (WSN) and its applications in industry and home appliances. Bhargave in [6], has introduced a system for restaurant using android technology. He proposed integration of Order System Kitchen Order Ticket (KOT), Billing System and Customer Relationship Management System (CRM). Implementation of such integration has increased quality and speed of restaurant service. Wahab in [7], has presented a new development in restaurant ordering system. He built up network system integration for connecting customer table with management panel. He used microcontroller with VB and RS 232 communication port. Proposed system performed well in network based system. The proposed system will have a complete Touch Panel device in restaurants and hotels that will help to reduce the human interaction and save time. This project implements a system that is on customer table by having touch panel menu ordering system. We have one touch technology by selecting the items from menu list display in LCD on customer table The basic concept of the project is to reduce human effort for the customers and also to deliver them with best facilities. Wireless communication link between customer’s table and kitchen has been established for placing order on customer’s side and verifying it on kitchen side. Furthermore, bill details 126 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 136. are also displayedon customer’s screen. Management can change menu by editing it and adding more items or changing cost of each item as per market rates. This paper is organized as explained further. Section 2 contains problem statement. Section 3 contains working of Zigbee technology. Section 4 contains overview of proposed system for restaurants. Section 5 contains comparison of proposed system with previous systems. Section 6 explains simulation and results. Section 7 contains conclusion of this paper. II. PROBLEM STATEMENT The operational system is very tedious .Manual order placements with the help of waiters have been carried out in restaurants. Menu card are normally presented on to the customer out of whom he can select available eatable items. All this process is cumbersome for customers from order placement till bill payment. Moreover, changing of menu cards or adding new items to menu, causes problems for responsible persons to update menu cards frequently. Therefore, existing system is time taking and requires a lot of human effort. The existing system is non- computerized. Such manual systems lead to problems because the waiter might not understand what the customer had ordered therefore serving him/her a different menu. This could be quite embarrassing as it will lead to customer dissatisfaction and ruin of restaurant reputation. III. ZIGBEE TECHNOLOGY ZigBee Module is a low-cost, low-power, wireless mesh networking standard. With minimum amount the technology to be widely deployed in wireless control and monitoring applications, the low power-usage allows longer life with smaller batteries, and the mesh networking provides high reliability and larger range This network device having qualities of electric power- saving, reliability, low cost,large capacity and security, and it has very wide application in various fields of automatic control. The target application domains are aimed at industry, home automation, telemetry and remote control Technical Data o Use Industry Standard for environment o Maximum Transmission Distance: o Internodes Barrier Free: 200 meters( 658feet) Wireless Frequency: 2.4G ISM License-free Frequency Band Channel Mode: 16 Channels Can Be Specified or the Best Channel Can Be Automatically Selected Antenna Configuration: Built-in 2.4G Ceramics Antenna Node Type: Center Node, Routing Node, Terminal Node, or Software Set Serial Rate: 1200-115200 Send Mode: Broadcast Send or Destination Address Send Working Voltage DC-3.3V Peak Current 40mA Best features is Low power consumption. At the low power consumption standby mode, two No.5 dry-charged batteries can support one node to work 6 ~ 24 months, or even longer. Low cost. Because of dramatically simplifying the protocol. It is protocol patent fee, free and low rate. Zigbee can work at the low rate of 20 ~ 250 kbps. Short range transmission range is generally between 10 ~ 200 m ( 32.8feet~ 656feet) .Short time delay. The response speed of Zigbee is very fast, in general, it merely need 10ms from the into work state and it merely need 20 ms from nodes connect into the network state. For high capacity Zigbee can adopt star topology, tree topology and mesh network structure, composing of up to 65, 000 network node. For high security. Zigbee provides a three- tier security model, including the use of Worry-Free Security settings, the access control list ( ACL) to prevent illegally accessing the data and Advanced Encryption Standard ( AES 128) symmetry password. For License-free frequency band. Zigbee adopts direct sequence spread spectrum of the industrial scientific medical (ISM) band. 2. 4 GHz (global). Sensor Network (WSN) Automatic meter reading system of water, gas, heat, electricity meters Intellectual traffic control, signal lights control and street lights control Fire safety alarm, building monitoring Catering order, canteen’ s sale of food system Access control, time and attendance system PTZ monitoring, engine room equipment monitoring Store & Logistics, laser guns, bar code reader RFID remote data transmission Meeting ballot system IV. OVERVIEW OF PROPOSED SYSTEM The proposed system has been developed for making restaurant ordering system easy. This system is also capable of keeping record of customer’s orders. The proposed systems support restaurant operation by performing following tasks: To put, view or change submitted order on the customer’s table module. To display menu on customer’s module interface. To allow kitchen’s module to give feedback about placed order of customer (whether ordered items are available). The proposed system will handle efficiently all problems faced by previous manual systems. Furthermore, system will store and analyze information, which will be provided by customer and kitchen staff and perform task automatically. The proposed system has also following added features that can increase productivity of the system: Data handling with ease and accuracy Reduction in paper work done by restaurant waiters or managers. 127 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 137. A. Zigbee transceiverat customer end (on table) . The customer module is play key role our project since the customers are the vital aspect for any business firm. The customer module consists of sensitive touch hardware. Our project was implemented using FRIENDLYARM. The best features of the hardware includes impressive menu, comprising of self-generated item number, item name, item description, an interesting feature is the item picture, tax, price and quantity selection checkbox allowing a maximum level of 10. After selecting the item the customer selects the add order button upon which the page refreshes and when he clicks the view order button he can see the item he has ordered along with the cost and if the customer wishes he can edit the order and start ordering afresh otherwise he can confirm the order by pressing the place order button, and a message will come and after taking food the customer can also send his valuable. Figure 1: Block diagram of customer end module 1) Implementation Microcontroller pic 16F877A is place in middle and all components are connected with this microcontroller firstly connected with DC supply interfacing via 10k resister five pins are connected with lcd and eight pins are interfacing with touch pad and three pins are connected with zigbee module and eight pins are first connected with 10 k resisters then coneected with led and these interfacing are with red lines to short these components Figure 2: Customer end module operating B. Zigbee transceiver at restaurant management end (in kitchen) As soon as the customer confirms his order, the order is displayed in the Central Kitchen . The order comprises of the table number, order number and dish name. The status of incoming order is made pending and as soon as the chef prepares the food he can set the status as finished. The importance of this module is that it plays a vital part in time reducibility. Figure 3: Block diagram of kitchen end module 1) Implementation After deep studying and analysis of these components interfacing one by one to gain the desire results first available all components and microcontroller is the main in which all other components interfaced As microcontroller is placed and given the power supply with two pins and zigbee module is connected with three pins and eight pins are connected with LCD and some pins are connected with resistors of 10 k and then interfacing with led different is interfacing three relays on this side with resistors that if some other components are placed with this u can easily attach like ATM dipper etc. 128 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 138. Figure 4: Kitchenend module operational V. COMPARISON OF PROPOSED SYSTEM WITH PREVIOUS SYSTEMS Table 1 shows comparison results on the basis of different features considered for comparison. For comparison, Pixel point [9], LRS[10], Eric software[8] and Smart order system[7] are considered. As it can be seen that proposed system outperforms all other systems in comparison. TABLE 1. COMPARISON OF PROPOSED SYSTEM WITH PREVIOUS SYSTEMS Features Pixel point LRS Eric soft Smart order system Proposed System Graphics NO NO NO NO YES ATM Sweeper NO NO NO NO YES Prices NO NO NO NO YES Group Orders NO NO NO YES YES Status of Ordering NO NO NO NO YES Wireless Network YES YES YES NO YES Touch Screen YES NO YES NO YES VI. SIMULATION AND RESULT EXPLORATION In restaurant, customer’s table module has a touch screen display in which menu is shown. When this menu is pressed, full menu of restaurant eatables are shown on screen to be selected. Once order is selected and forwarded to kitchen, it uses zigbee transmitter and receiver to deliver that information to kitchen. Kitchen module will receive order which will be displayed on LCD. Kitchen person will confirm order by pressing confirm button or reject any unavailable eatable items from menu. Then, customer will get information regarding unavailability of any items and can re-order. Furthermore, waiter can be called by pressing “Call waiter” button. The proposed system can be used with line following robots or conveyer belts to deliver order till customers. This will reduce more human effort as there will be no waiter required to deliver order to tables. Other benefits are: Provision of facility for handling text electronically using powerful processors to produce elegant and error free documents. In addition to storing the organization’s operational data on disk drive for back up. With the installed software, product ordering and delivery was made easier. The systematic approaches used during each phase of the software development provides a clear road map that would be of immense help to anyone carrying out research work in this area. VII. CONCLUSION This research contributes to recent working environment of restaurants and can change it from previous manual system to an automated restaurant with a lot of facilities to customers as well as restaurant management. Furthermore, it will be helpful in attracting customers to see and avail such facilities. Zigbee , as a wireless link between customer’s table and kitchen can be replaced by any other wireless technologies for larger restaurants which supersedes zigbee’s range of communication. Line following robots or conveyer belts can be used to deliver eatable items to customers instead of waiters. REFERENCES [1] Arun P., Sabarinath G., “Implementation of ZigBee based Train Anti -Collision And Level Crossing Protection System for Indian Railways." Int. J. of Lat. Trends in Eng. and Tech., Vol.2, pp: 12-18, January, 2013. 129 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 139. [2] R. P.Ramasamay, M. P. Kumar, S. S. Kumar, R. R. Raman, “Avoidance of fire accident on running Train using ZigBee Wireless Sensor Network”, Int. J. of Inform. and Comp. Tech., Vol.3, pp: 583-592, 2013. [3] M. Blaho, J. Ozimy, M. Ernek, M. Foltin, “Zigbee implementation in Railroad model”. [4] B. Vinodhini, K. Abinaya, R. Roja, M. Rajeshwari, “Wireless Two-way Restaurant Ordering System via Touch Screen”, Vol.3, pp: 01-05, 2014. [5] W. Wang, G. He, J. Wan, “Research on Zigbee wireless communication technology”, Int. Conf. on Elect. and Cont. Eng. (ICECE), 2011. [6] Bhargave, A., Jadhav, N., Joshi, A., Oke, P., & Lahane, M. S. (2013). “Digital Ordering System for Restaurant Using Android”, International Journal of Scientific and Research Publication, Vol.3, 2013. [7] Wahab, M. H. A., Kadir, H. A., Ahmad, N., Mutalib, A. A., & Mohsin, M. F. M. (2008, August).” Implementation of network- based smart order system”, International Symposium on Information Technology, 2008. ITSim 2008. (Vol. 1, pp. 1-7). IEEE. [8] .[8] M.H.A. Wahab, H.A. Kadir, N. Ahmad, A.A. Mutalib and M.F.M. Mohsin, “Implementation of network-based smart order system,” International Symposium on Information Technology 2008 [9] PAR PixelPoint “PixelPoint POS Brochure [Accessed: 11 May 2011] [10] Advanced Analytical, Inc (October 2004) “LRS Restaurant Server Pager”, [Accessed: 11 May 2011] 130 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 140. Introducing space plantbiology to students through hands-on activities using clinostat Ferheen Ayaz SUPARCO Institute of Technical Training, HR Dte Gen, Space and Upper Atmosphere Research Commission, Karachi, Pakistan ferheen@msn.com Qurat-ul-Ain SUPARCO Institute of Technical Training, HR Dte Gen, Space and Upper Atmosphere Research Commission, Karachi, Pakistan ain_haider@hotmail.com Abstract—Space biology is an important topic for research and education in order to explore future life in space. Plant growth and development largely depends upon gravitational acceleration on earth. The growth process of a plant in space is significantly different from that on earth due to the absence of gravity, which is referred as zero gravity or microgravity. One- axis clinostat simulates microgravity environment on earth. On a horizontal clinostat, an object is rotated perpendicular to the force of gravity. Although the gravity is present on an object rotated by clinostat, its continuous reorientation nullifies the net effect of gravitational acceleration. One-axis/horizontal clinostat is a useful tool for space education. The simulated microgravity created by the clinostat, despite its limitations, causes significant changes in plant growth which may lead to introduce school and college level students about plant life in space and creating their interest in space biology. This paper describes some of the experiments conducted on one axis clinostat, by a group of teachers and students of SUPARCO Institute of Technical Training, which proved to be a fruitful activity resulting in learning of gravitropism basics of plants and seeds. The article is aimed towards helping the educators in creating hands-on learning activities for students to have an idea about life in space by making use of facilities available on earth. The orientation of roots grown from radish seeds (Raphanus sativus), area and length of roots and shoots grown from pea seedlings (Pisum sativum), length and bending angle of shoots of Marigold plant (Tagetes patula) and growth rate and direction of wheatgrass (Thinopyrum intermedium) were observed on the clinostat. The roots of radish seeds did not grow downwards in the direction of gravity but oriented themselves at random angles on the clinostat. Pea seedlings also showed different growth rates on clinostat as compared to their growth on earth. The experiment on Marigold plant resulted in faster growth in length and larger bending angle of clinorotated plant shoots as compared with stationary plant shoots. The growth rate of wheatgrass on earth and on clinostat was nearly same but their bending angles were different in both conditions. All the experiments provoked the minds of students to make inferences on seeds and plants growth in space. The experiments described in the paper create inquisitiveness among teachers and students to observe many other parameters in plant growth under microgravity. Keywords—clinostat; gravitropism; microgravity I. INTRODUCTION Considering the long term goals of space agencies and seeking habitation of extraterrestrial surfaces; one of the important concerns is to analyze the development of plants in space. The reason behind this concern is the dependence of plants and animals upon each other for their survival and flourishing of ecosystem [1]. Study of plant growth in space is different from conventional plant growth on earth mainly because of the fact that it is greatly influenced by gravitational acceleration. Recent experiments in space have resulted in modified growth rates of roots and shoots of different plants and the major reason of this modification is the absence of gravity in space [2]. In order to promote space exploration related activities and establish plant life in space, one of the key initiatives is to promote space education among young students. Creating interest of students in space sciences may lead to increase their curiosity towards exploring life support systems in space and planetary surfaces. Due to limited facilities of space research in underdeveloped countries, students are least interested in space biology. The students are less inclined to study space sciences because of unavailability of resources and opportunities to do practical work in this field. The practical work for science students plays a very important role in their learning [3]. As it is difficult to create practical setups of complete space environment on earth, establishing a simulated microgravity environment can provide excellent practical opportunity for researchers and students to study life sciences and plant and animal life in space [4]. Fig. 1. One-axis clinostat to simulate microgravity on earth. Microgravity for a very short period of time can be created on earth by elevator, drop tower, rocket and aircraft [5] but changes in plant growth cannot be observed by these methods as their growth process is comparatively slower. The efficient instruments for conducting biology experiments in simulated 131 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 141. microgravity environment are2D and 3D clinostats [6]. Keeping in view the practical needs of space education, SUPARCO (Space and Upper Atmosphere Research Commission) Institute of Technical Training acquired 2D clinostat, as shown in Fig. 1, under ZGIP (Zero Gravity Instrument Project) initiated by United Nations Human Space Technology Initiative (HSTI) in 2013. A 2D or one axis clinostat has one rotational axis which rotates perpendicular to the direction of gravity vector in order to create the effect of microgravity and is suitable for studying the behavior of germinating seeds and analyzing plant growth [7]. Although horizontal clinostat does not completely create zero gravity environment on the object under study [8], it can still be used for educational purposes to demonstrate changes in plant growth by alteration of gravity. This article describes some experiments conducted on seeds and plants using clinostat at SUPARCO Institute of Technical Training by a group of 2 instructors and 10 students aged between 17 to 19 years. The objective of describing the experiments is to promote science teachers and students to study space biology practically and to provoke the minds of interested learners to bring up similar but new ideas to experiment and discover effects of gravity on plants. The organization of the rest of the paper is as follows: Section II describes the experiments performed using clinostat, Section III discusses results of the experiments and conclusion is mentioned in Section IV. II. EXPERIMENTS PERFORMED USING CLINOSTAT A. The radish seeds The first experiment to study gravitropism on the roots of radish seeds (Raphanus sativus) was conducted as described in [7]. Nine samples of radish seeds were placed on each of the three Petri dishes. The Petri dishes were half filled with agar- agar solution to aggravate the process of germination. Humidity, temperature and light conditions were kept same for each Petri dish. After 48 hours, when the roots started to grow downward, first Petri dish was placed vertically with roots along the direction of gravity, second was turned 90o so that the roots were held perpendicular to the gravity vector and the third was attached to clinostat using double sided tape. The rotation disc of clinostat was held horizontally. The placement of Perti dishes is illustrated in Fig. 2. The clinostat was set to rotate with a speed of 10 r.p.m. for 4 hours and photographs of each Petri dish were captured using a digital camera. After the experiment, the photographs were observed using Image J software. The orientation of roots was analyzed in terms of angle created as the roots bent during the experiment. Fig. 2 (a). First Petri dish with roots parallel to gravity vector. Fig. 2 (b). Second Petri dish with roots perpendicular to gravity vector. Fig. 2 (c). Third Petri dish attached to clinostat. B. The pea seedlings This experiment was designed by the working group to observe not only the roots, but also the shoots grown from pea seeds (Pisum sativum). Four test tubes were half filled with the agar-agar solution made by the procedure defined in [7]. A pea seed was placed inside each test tube such that it lies exactly in the middle of the tube as shown in Fig. 3. After 7 days, the shoots and roots of the pea seeds were grown. Two test tubes were placed horizontally perpendicular to the gravity vector and the other two test tubes were attached to clinostat. The 132 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 142. clinostat was setto rotate on horizontal axis at 15 r.p.m. for 72 hours in a dark room. The area and length of both root and shoot of each pea seedling was observed at the end of experiment. Fig. 3. Placement of pea seed in test tube. C. The Marigold plant The growth of shoots of Marigold plant (Tagetes patula) in simulated microgravity produced by clinostat was compared with the growth in normal gravity conditions. In this experiment, six small Marigold plants were observed. Initially, the length of the shoot of each plant was measured. The stripes with the interval of 5mm were made on each complete shoot using permanent marker. Six test tubes were filled with water and a Marigold plant was inserted in each test tube such that roots were dipped inside water and shoots appeared from just outside the tube as shown in Fig. 4. Three test tubes were placed horizontally perpendicular to the gravity vector and other three test tubes were attached to clinostat. The clinostat was set to rotate on horizontal axis at 5 r.p.m. for 4 hours in a dark room. The length and bending of each shoot was observed at the end of experiment. Fig. 4. Placement of Marigold plant in test tube. D. The wheatgrass The fact that wheat is one of the major food crops of Pakistan motivated the working group to observe its growth behavior on clinostat. Wheat seeds (Thinopyrum intermedium) were soaked in water overnight and then grown in two Petri dishes filled with soil for 10 days under direct sunlight and normal gravity conditions of earth. When the grass grew out of the Petri dishes, one of the dishes was placed on earth with grass growing in the direction opposite to the gravity and the other was attached with the clinostat, as shown in Fig. 5. The clinostat was set to rotate on horizontal axis at 10 r.p.m. for 4 days in a room with a 100 Watt yellow light bulb turned on continuously. The photographs of both Petri dishes were captured each day. The soil in Petri dishes was watered each day using a spray bottle. At the end of experiment, the length and bending angle of some samples of wheatgrass were observed in the photographs using Image J software. Fig. 5. Experimental setup for observing growth of wheatgrass under normal gravity on earth and simulated microgravity produced by clinostat. III. RESULTS A. The radish seeds The results of the experiment were as follows: The roots in first Petri dish grew in the direction of gravity and the effect is called positive gravitropism, therefore their bending angle was approximately zero degree. Fig. 6 shows that the bending angle of second Petri dish (90-degree turned) was gradually decreasing which shows their growth towards the direction of gravity with the passage of time. However, the decline in the angle was not at a constant rate. Fig. 6 does not show a regular pattern in bending angle of third Petri dish (clinorotated) and hence we can conclude that in microgravity, the roots may grow in any random direction and not necessarily downwards. Summary of the result is presented in Table I. 133 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 143. Fig. 6. Bendingangle of roots of radish seeds on 90o turned and clinorotated Petri dish with respect to time. TABLE I. BENDING ANGLE OF ROOTS OF RADISH SEEDS Time (minutes) Average angle of nine seeds (degrees) 90o turned Clinorotated 0 35.3897 9.241 30 28.2403 14.584 60 22.9933 15.775 90 26.0908 15.798 120 24.8322 14.128 150 25.962 6.935 180 24.0473 12.757 210 21.8095 15.607 B. The pea seedlings The results of the pea seedlings grown for 3 days on horizontal clinostat at 15 rpm compared with stationary seedlings were as follows: Roots from clinorotated plants had a smaller cross sectional area compared to the stationary control with a difference of -3% whereas shoots from clinorotated plants had a larger diameter relative to the stationary control with a difference of 23%. Clinorotated pea seedlings showed an increase in root length by 3% and increase in shoot length by 27% as compared to stationary control (1-g on earth). Summary of the result is presented in Table II. TABLE II. RESULT OF EXPERIMENT ON PEA SEEDLINGS Stationary Clinorotated Individual test tubes Average Individual test tubes Average Roots Area (mm square) 0.046 0.0435 0.048 0.042 0.041 0.036 Length (mm) 7.56 7.1755 7.91 7.3615 6.791 6.813 Shoots Area (mm square) 0.046 0.0345 0.061 0.0425 0.023 0.024 Length (mm) 7.507 5.65 9.919 7.1995 3.793 4.48 C. The Marigold plant On an average, the shoots of Marigold plant grew more in 4 hours on clinostat as compared to earth. Table III and the graph in Fig. 7 summarize the result. The following important points were resulted: The average growth of shoots on earth was 0.034 cm and on clinostat was 0.305 cm. Bending of shoots was observed both on clinostat and on earth, but the shoots on clinostat bent at larger angles than the shoots which were placed horizontally on earth. TABLE III. GROWTH OF SHOOTS OF MARIGOLD PLANT Gravity condition Test tube Initial length of shoot (cm) L1 Final length of shoot (after 4 hours) (cm) L2 Difference (cm) L2-L1 Angle of bending after four hours (degrees) 1-g 1 5.3 5.454 0.154 -0.256 2 5.6 5.602 0.002 11.911 3 6.3 6.385 0.085 1.193 Simulated micro- gravity by clinostat 4 5.4 5.655 0.255 28.944 5 4.9 5.428 0.528 30.42 6 5.6 5.733 0.133 30.379 Fig. 7(a). Growth in length of shoots of Marigold plant. Fig. 7(b). Bending angles of shoots of Marigold plant. D. The wheatgrass The results of the experiment are as follows: As shown in Fig. 8(a), the length of wheatgrass showed similar growth rate on earth and on clinostat and it was decreasing per day. Calculated growth rate on earth 134 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 144. was 0.118 cm/hourfrom Day1 to Day2, 0.026 cm/hour from Day2 to Day3 and 0.024 cm/hour from Day 3 to Day4. It was almost the same on clinostat. The calculations came out to be 0.112 cm/hour from Day1 to Day2, 0.027 cm/hour from Day2 to Day3 and 0.026 cm/hour from Day 3 to Day4. As shown in Fig. 8(b); on earth, the wheatgrass grew almost straight and angles with respect to the direction of gravity were in the range of 0 to ±3o , showing almost no bending. On clinostat, the angles were varying in the range to 10o to 14o which is the evident proof of bending. Summary of the result is presented in Table IV. TABLE IV. LENGTH AND ANGLE OF WHEATGRASS DRUING 5 DAYS OF EXPERIMENT Measurement of: Average Length (cm) Average Angle (degrees) Day On earth On clinostat On earth On clinostat 1 3.69 2.30 0.931 13.73 2 6.54 5.00 -0.69 8.77 3 7.19 5.67 2.81 10.80 4 7.77 6.31 -2.44 8.39 Fig. 8(a). Length of wheatgrass per day on earth and on clinostat. Fig. 8(b). Angle of wheatgrass per day on earth and on clinostat. IV. CONCLUSION Four experiments on different seeds and plants were conducted on clinostat. The results show that all plants do not exhibit similar behavior in microgravity. Absence of gravity speeded up the growth rate in pea seedlings and Marigold plant but did not influence wheatgrass. The diversity of the experiments described in this article encourages the students to try out experiments on other plants and seeds available to them. Various other parameters like temperature, humidity, type of soil, amount of water and nutrients provided can also be modified to discover new effects of gravity and microgravity on plants. Acknowledgment The study is carried out as a part of Zero Gravity Instrument Project initiated by United Nations Office for Outer Space Affairs (UNOOSA). We gratefully acknowledge UNOOSA for donating clinostat, Training Division, HR Dte Gen, Pakistan Space and Upper Atmosphere Research Commission for their coordination in publishing the article and Horticulture Cell, Pakistan Space and Upper Atmosphere Research Commission for assisting in plant growth. References [1] R. Ferl, R. Wheeler, H. G. Levine, and A. L. Paul, “Plants in space.” Current opinion in plant biology, vol. 5, no. 3, 2002, pp. 258-263. [2] T. Hoson, “Plant growth and morphogenesis under different gravity conditions: Relevance to plant life in space,” Life, vol. 4, no. 2, 2014, pp. 205-216. [3] V. N. Lunetta, A. Hofstein, and, M. P. Clough, “Learning and teaching in the school science laboratory: An analysis of research, theory, and practice,” Handbook of research on science education, 2007, pp. 393– 441. [4] A. Niu, M. Ochiai, and H. J. Haubold, “The United Nations Human Space Technology Initiative (HSTI): Science Activities,” IAC-12- A2.5.11, 2012. [5] M. J. Rogers, G. L. Vogt, and M. J. Wargo, “Microgravity: A Teacher's Guide with Activities in Science, Mathematics, and Technology,” NASA, EG-1997-08-110-HQ, 1997. [6] M. Cogli, “The fast rotating clinostat: a history of its use in gravitational biology and a comparison of ground-based and flight experiment results,” Gravitational and Space Research, vol. 5, no. 2, 2007. [7] “Programme on Space Applications: Teachers Guide to Plant Experiments in Microgravity,” United Nations Office for Outer Space Affairs, 2013. [8] A.H. Brown, O.A. Dahl, and D.K. Chapman, “"Limitation on the use of the horizontal clinostat as a gravity compensator," Plant Physiology, vol 58, no. 2, pp. 127-130, 1976. Sponsored by Pakistan Space and Upper Atmosphere Research Commission. 135 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 145. The Political andEconomic Feasibility of Current Space Resource Management Policies Faizan Muhammad Lahore, Pakistan faizan.muhammad7@outlook.com Muhammad bin Munim Lahore,Pakistan muhammadbinmunim@gmail.com Suleman Saleem Lahore, Pakistan msulemansaleem727@gmail.com Abstract—At the rapid pace with which the interest in utilization of extra-terrestrial resources is developing, it would be no surprise if such projects are practically begun by the next decade. This, however, poses a unique dilemma to the scientists and politicians alike. There is no global consensus as yet regarding the specific rules and regulations that would come into play, nor is there any international framework to establish a legal foothold in the matter. To further complicate matters, not only are national governmental space agencies interested in such projects, but private companies too wish to take part in the process. Countries, such as the USA for example, have a relatively elaborate procedure for private companies but many others do not. The 1967 “Treaty on Principles Governing the Activities of States in the Exploration and Use of Outer Space, including the Moon and Other Celestial Bodies” by the United Nations was successful but the 1979 “Agreement Governing the Activities of States on the Moon and Other Celestial Bodies” turned out to be a disappointment. The flaws that caused it to be treated that way were identified and addressed. The feasibility and possible consequences of the aforementioned treaties and legislature were analyzed with respect to future plans of government and private agencies as well as relevant historic parallels. Certain limitations were exposed in the current system which may lead to escalated controversies in the future. Suggestions have also been made to improve or redesign certain aspects of existing systems to work together efficiently and smoothly all over the globe. Index Terms—Space Econopolitics, Space Ownership, Space Law INTRODUCTION Due to the fact that the Moon Treaty has not been ratified by any nation that has achieved spaceflight (and will thus play an active role in space expeditions in the near future), the Outer Space Treaty of 1967 holds greater importance than the Moon Treaty and hence currently forms the legal framework of international space law . The treaty prohibits any military activity in space, confining the use of the moon, its orbits, and other celestial bodies to peaceful purposes only. Furthermore, the treaty explicitly forbids any claim of sovereignty by any nation by any means whatsoever. The clauses of the Moon Treaty are very similar to those of the Outer Space Treaty, the major difference being that the Moon Treaty calls for solidification the management of Space under an international regime. It provides equal rights to all the states in terms of conducting research in space. Furthermore it bans commercial activity in space, giving the power of resource extraction and allocation to “international regime” approved parties only. Another distinction of the Moon Treaty is that unlike the Outer Space treaty, it also prohibits ownership of any part of the celestial bodies by private parties not necessarily tied to any nation. In the following sections, the incentives and proposed plans of various countries and organizations regarding lunar and asteroid mining will be discussed as case studies in order to put emphasis on how current treaties governing space exploration and resource management are not feasible. To impress upon the urgency of the matter, the reasons for controversies will be discussed, leading to the importance of treaties in the international landscape and through the analysis of treaties and treaty organizations a parallel will be drawn between the historic treaties and conflict and the current space situation to show that the current treaties governing space exploration and resources management solutions are not viable. Finally some suggestions will be made to improve the state of current space resource management policies to help avoid any potential future conflicts. LUNAR RESOURCES A. Helium-3 Extraction TABLE 1.PROPERTIES AND MAJOR APPLICATIONS OF HELIUM-3 Property Application Does not solidify, even at 0K, at atmospheric pressure Used for Liquid cooling of Low Temperature Superconductors Second lightest element, after Hydrogen Used as a substitute for Hydrogen in filling balloons 136 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 146. Helium is theelement with the smallest molecular size Used for Leak detection Has extremely low solubility in water Used as a diving gas Helium-3 is not radioactive Used as a heat transfer medium in fusion reactors The importance of the noble gas being realized, it is important to mention that there is a shortage of Helium on Earth. Currently, Helium supplies are obtained from the extraction of natural gas. The United States produces about 75% of the world’s Helium and roughly 30% of the world’s Helium supply comes from the US federal Helium reserve. The problem is that Helium reserves are being depleted at a rapid rate because of wide use of the gas. Walter Nelson, director of helium sourcing for Air Products and Chemicals Inc. predicts the usable life of the US reserves to last till 2018-2020. This will have a global impact, especially on healthcare and small- scale scientific research. MRI machines need Helium to operate and they will become difficult to maintain with a shortage of Helium, ultimately making it very difficult to obtain MRI’s. Furthermore, Harrison Schmitt, a geologist and astronaut, is of the opinion that Helium-3 can prove to be a very useful alternative for deuterium and tritium in nuclear fusion reactors for power generation, as the reaction involving Helium-3 does not result in the emission of neutrons, making it clean and non- polluting. Since there are not enough supplies on the Earth to provide the Helium-3 needed to carry out commercial power generation through fusion, in his book “Return to the Moon” Harrison Schmitt argues that the moon should be mined for Helium-3. Since the moon is continuously bombarded with solar wind, it has a large amount of Helium-3, though in small concentrations of about 10-20 parts per billion. Schmitt outlines a plan to for mining helium-3 on the Moon to feed a commercial industry for generating power, though it will be very expensive. Even if many consider Schmitt’s proposal to be unrealistic, it is a fact that he is not the only one who thinks mining the moon for Helium-3 is a good idea; The leader of Beijing's space program, has said generating power via nuclear fusion using He-3 could "solve energy demand for 10,000 years at least". So China is seriously considering it. B. Rare Earth Elements Rare Earth Elements, or REE for short, are 17 elements in the periodic table which are widely used in electronics. Their name tends to be misleading; they are relatively plentiful in the earth’s crust. [11] The mining of REE was once dominated by the United States from 1965 to 1985, until The People’s Republic of China started dominating the market. In 2010, a “Rare Earths Trade Dispute” occurred, with China on one side and many other countries, especially the US, on the other [10]. China had increased the restrictions on its exports of REE by reducing its export quota by 40% [2][12]. This resulted in a major increase in the prices of REE outside the country. Although the Chinese government argued that the reduction in the quota was for the benefit of the environment (REE mining is highly polluting), critics were unconvinced. Due to the importance of the REE in the military, the question arose in the US of supply vulnerability of these elements [2][13]. There were also accusations that China had banned the exports of REE to Japan after the 2010 Senkaku boat collision incident. The matter of the Chinese restrictions was brought to the Dispute Settlement Body of the WTO in 2012, by the United States. China had joined the WTO in 11 December 2001. When it joined the WTO, China signed an accession treaty, which did not allow such restrictions on exports unless the goods in question were stated, which was not the case here. China defended itself by arguing that article XX (b) of the GATT allowed exceptions if the restrictions were put in place for the protection of human, animal or plant life or health. The majority of the panel was of the view that the exceptions in article XX (b) did not justify China’s reduction of export quotas and that these measures were not necessary for the protection of human, animal or plant life or health in this case. So the WTO ruled against China in 2014 and the country removed the restrictions in 2015 [10]. There are rocks rich in Potassium, REE, and Phosphorous along with other elements, referred to as KREEP, on the moon’s surface. According to the deputy director of the Rock Mechanics and Explosives Research Centre at the Missouri University of Science and Technology in Rolla, Leslie Gertsch, although REE themselves aren’t detectable on the moon’s surface, KREEP is. Detection of components of KREEP, such as Thorium, on the lunar surface helps locate associated REE due to similar geochemical properties that caused them to crystallize under the same conditions, Gertsch says. Dale Boucher, director of innovation at the Canada-based Northern Centre for Advanced Technology Inc., in Sudbury, Ontario, stated that the feasibility of mining REE on the moon can only really be judged after exploration intended for this exact purpose. It is difficult to assess it otherwise, as there are many factors which determine its feasibility, including cost of separating REE obtained from ores from each other, cost of transportation of said REE, cost of refining them etc. (North America seems to have abundant REE deposits but it is just not profitable to mine them). But it is still in the realm of possibilities. If a country does indeed intend to mine REE from the moon’s surface in the near future, there’s a risk that a scenario similar to the Rare Earths Trade Dispute might occur C. Colonization The United States, Japan, and Russia have shown interest in the establishment of colonies on the moon [16]. The plans seemed to have gained more ground on November 13, 2009, 137 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 147. when NASA announcedthe discovery of ice caps at the south pole of the moon [15]. The reasons given by NASA for the exploration of the moon are: 1. Human Civilization: Extend human presence to the Moon to enable eventual settlement 2. Scientific Knowledge: Pursue scientific activities that address fundamental questions about the history of Earth, the solar system and the universe; and therefore, about our place in them 3. Exploration Preparation: Test technologies, systems, flight operations and exploration techniques to reduce the risks and increase the productivity of future missions to Mars and beyond 4. Global Partnerships: Provide a challenging, shared and peaceful activity that unites nations in pursuit of common objectives 5. Economic Expansion: Expand Earth's economic sphere, and conduct lunar activities with benefits to life on the home planet 6. Public Engagement: Use a lively space exploration program to engage the public, encourage students and help develop the high-technology workforce that will be required to address the challenges of tomorrow [3][4][5] The 1st and 3rd reasons show their intention to colonize the moon in the future. In January 2012 Republican candidate for President of the United States of America, Newt Gingrich, proposed the establishment of a US moon colony by 2020, though the plan faced much criticism.[16] Russia’s intentions seem to be different regarding the 3rd reason. “The Moon is not an intermediate point in the race; this process has the beginning, but has no end. We are going to the moon forever.” stated Russia’s Deputy PM Dmitry Rogozin who is in charge of space and defense industries. In 2007, the Soviet Union announced its plan to establish a permanent moon base by 2025 [14]. The main focus seems to be lunar tourism [16]. According to a government draft that Izvestia, a Russian newspaper, claims to have obtained, Russia has a 3- step plan for manning the moon [14][17]: 1. Sending a robotic craft to the moon, possibly as early as 2016, 2. Sending manned missions to orbit the moon by 2028, 3. A base will be set up by 2030, using resources from the moon. Japan announced its plans of setting up a moon base in 2006. They estimated that the base in question will be established by 2030. Their main motivation, they claim, is the development of robotics. [9] D. Future Missions of Major Parties: [20] 1. United States (NASA) a. Exploration Mission-1: An unmanned flight of the Orion (MPCV), using NASA’s Space Launch System (SLS). The aim will be to qualify the Heavy Lift Launch Vehicle (HLV) and Beyond Earth Orbit (BEO) Orion for transportation of humans into deep space. The estimated date of launch is December 17, 2017. The mission is expected to last 7-10 days [18]. b. Exploration Mission-2: The first manned mission of NASA’s BEO. The crew will consist of 4 people. The aim will be to test Orion’s life support systems and to validate its crew operations. The mission is expected to be carried out by the year 2021 [19]. 2. Russia (ROSCOSMOS) a. Luna-25: To land on the South Pole of the moon, and test the lander technology and communications systems b. Luna-26: To orbit the moon for detailed mapping of the moon’s surface and search for a new landing site for future missions and to measure the lunar atmosphere. c. Luna-27: To go to the south pole of the moon for testing drilling systems and analyzing contents. d. Luna-28: To return soil samples to the earth e. Luna-29: To use a rover to take samples from different locations on the moon’s surface [21][22][23][24][25] 3. China a. Chang’e 4: A backup probe for Chang’e 3 [6], which put which put a lander and the Jade Rabbit rover on the moon in 2013. The expected year of launch is 2020. [26][28] b. Chang’e 5: To return lunar samples [27][29] c. Chang’e 6: A backup for Chang’e 5. [27] 4. Japan (JAXA) An attempt at Japan’s first lunar landing, an unmanned flight, is planned for 2019, though the budget for the plan is not available yet. [30][31][32] 5. India (ISRO) Chandrayaan-2: Terrain mapping by the employment of new technology for better imaging of the lunar surface. 138 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 148. [33][34] 6. European Union(ESA) ESA plans to send a lander near the moon’s South Pole in 2018. The purpose of the mission is to probe the moon’s surface and to test new technologies for future landings.[7] 7. Private Google Lunar X Prize: Astrobotic Technology, a privately held American company, plans on launching a private launcher in 2015, with the aim of winning the Google Lunar X Prize, which was created in 2017 to encourage space entrepreneurs to create a new era of affordable access to the Moon and beyond. [8] E. Conclusion As it can be vividly seen, there are several detailed and ambitious projects ready to explore the moon and benefit from its resources. Keeping that in mind it becomes more important than ever to firmly establish policies right now that are acceptable to all parties to avoid any conflicts in the future. ASTEROID MINING Besides the moon, asteroids also have potential resources which scientists are looking to explore. Most asteroids in the solar system are found between Mars and Jupiter. This region is termed the asteroid belt or main belt. Asteroids with orbits that bring them within 1.3 AU (121 million miles/195 million kilometres) of the Sun are known as Earth-approaching or near-Earth asteroids (NEAs). These are the asteroids which people are looking to mine. Most or all asteroids found in the main belt appear to be NEAs. Asteroids are usually categorized according to their albedo (how reflective they are), composition derived from spectral features in their reflected sunlight, and inferred similarities to known meteorite types. There are 3 main categories: 1. C-type (carbonaceous): More than 75% of known asteroids belong to this category. Albedo: 0.03-0.09 Composition: Thought to be similar to the sun’s composition, except for the absence of hydrogen, helium and other volatile elements and compounds. Because of their low temperatures, they are estimated to contain up to 22% water. Found in: the main belt’s outer region. 2. S-type (silicaceous): About 17% of known asteroids belong to this category. Albedo: 0.10-0.22 Composition: metallic iron mixed with iron- and magnesium-silicates. Found in: the main belt’s inner region. 3. M-type (metallic): Albedo: 0.10-0.18 Composition: mainly metallic iron. Found in: the main belt’s middle region. [35] A. Motivations for Asteroid Mining Based on consumption rates in both developed and developing countries, it is estimated that terrestrial reserves of elements needed for modern industry will be exhausted within 50–60 years. Consequently, 44 elements are expected to face supply limitations in the coming years, including phosphorus, antimony, zinc, tin, silver, lead, indium, gold, and copper. All these elements are crucial for industry [36][37][38]. Some of the applications of a few of these elements are noted in the following table: TABLE 2.ELEMENTS FROM ASTEROIDS AND MAJOR APPLICATIONS Element Applications Copper . Used in electrical equipment such as wiring and motors. . Copper Sulfate is used as an agricultural poison and as an algaecide in water purification. . Fehling’s solution, a solution containing copper, and other copper containing solutions are used in chemistry for qualitative analysis. . Copper alloys are used in making coins. Zinc . Used to galvanize objects to prevent rusting. . Used to produce die-castings. . Used in alloys such as brass, nickel silver and aluminum solder. . Zinc oxide is used in the manufacture of paints, rubber, cosmetics, pharmaceuticals, plastics, inks, soaps, batteries, textiles and electrical equipment. . Zinc sulfide is used in making luminous paints, fluorescent lights 139 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 149. and x-ray screens. Lead. Used for car batteries, pigments, ammunition, cable sheathing, weights for lifting, weight belts for diving, lead crystal glass, radiation protection and in some solders. . Used to store corrosive liquids. . Used in architecture, for roofing and in stained glass windows. Tin . Used to coat other metals to prevent corrosion. . A niobium-tin alloy is used for superconducting magnets. . Tin salts sprayed onto glass are used to produce electrically conductive coatings. [39] B. Mining Considerations There are three options when considering mining asteroids: 1. Bringing the raw material to earth. 2. Processing the raw material on the asteroid itself and then bringing the processed material to earth. 3. Transport the asteroid to a safe orbit around the Moon, Earth or to the ISS. This method is estimated to be the one which allows the most materials to be used. [40][41] C. Proposed Mining Projects There are four main parties currently proposing projects for mining asteroids: 1. NASA (US) An upcoming mission, termed OSIRIS-Rex, scheduled for launch in 2016, will involve travelling to a near-Earth asteroid called Bennu and bringing a small (about 60g) sample back to Earth for study. The destination will be reached in 2018 and the sample will be transported to the earth in 2023.[49][50] NASA Institute for Advanced Concepts (NIAC) announced the Robotic Asteroid Prospector (RAP) project on in September 2012. The purpose of RAP is to evaluate the feasibility of asteroid mining in terms of means, methods, and systems. Their stated strategy is: “Initial prospecting missions will be robotic, although early industrial mining missions will require both humans and robots on-site. There will be too many judgment calls and qualitative decisions on-site for robots alone. Humans will need the practice of mining in space with robots to learn how to innovate and to work efficiently, reliably, and safely.” [42] 2. Planetary Resources (Private, US) The company announced on April 24, 2012 its plans to mine NEA. They plan to launch prospective spacecraft for collecting data about asteroids within 7 years. The company plans to build new technology which will reduce the price of space travel so that it can be successful in its mission. [46][47][48] 4. Deep Space Industries (Private, US) On January 22, 2013, this private company announced its plans to launch spacecraft in 2015 for prospecting suitable asteroids and returning samples to the earth, and to begin mining NEA by 2023. [43][44][45] 5. Kepler Energy and Space Engineering (KESE) They aim to build and send a 4-module Automated Mining System (AMS) to a small asteroid with a simple digging tool to collect approximately 40 tons of samples by the end of the decade, using existing guidance, navigation and anchoring technologies.[51] D. Conclusion It becomes clear after looking at the motivations and future plans of various parties (countries as well as private parties) that a future marathon for space resources is highly probable. Unfortunately, the Outer Space and Moon Treaty hinder the steps made towards the mentioned endeavours as they face many legal barriers. For now, only the hurdles provided by the outer space treaty will be discussed, as the moon treaty has not been signed by major influential parties. Specific parts of the two treaties will be discussed in more detail later on in the paper. Based in part on the Antarctic Treaty, the Outer Space Treaty only appears to permit research activities on the moon and other celestial bodies, though this is debatable. “The idea that you can't claim sovereignty is not necessarily incompatible with the right to go conduct mining operations. The high seas are not subject to any sovereignty, but people can go and fish there.” said international lawyer and space-law expert Timothy Nelson in an interview with space.com, referring to an article in the 1967 treaty stating: “Outer space, including the moon and other celestial bodies, is not subject to national appropriation by claim of sovereignty”. Even then, the treaty hinders research, mainly owing to a lack of incentive. Due to the statement that no particular party can claim any ownership rights to the resources, there will be a lack of commercial concern, ultimately leading to a lack of funding for research, hence lowering the amount of potential research which can be carried out. 140 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 150. HISTORICAL ANALYSIS The ever-evolvingeconomic landscape presents a unique challenge in the world of politics. The one of keeping peace and making sure that the resources of a state are not illegitimately taken away from it. The solution is an old one; treaties and organizations to enforce the treaties. In this section conflicts and resources as a main reason for conflicts will be discussed, along with how they can be avoided. This will lead to the importance of treaties in the international landscape and through the analysis of treaties and treaty organizations a parallel will be drawn between the historic treaties and conflict and the current space situation to show that the current treaties governing space exploration and resources management are not viable. A. Wars History has seen many wars fought over resources. Hitler’s campaign in 20th century, before the Second World War is a perfect example of a war fought over resources. When Hitler came to power in 1933 his whole agenda centered on making Germany a powerful sovereign country again. There were two important aspects to this. One was regaining areas with great economic potential, such as Alsace and Loraine on the western border; they were highly industrialized zones and had Iron ore and Coal deposits. Another was to get more land. This was a part of Lebensraum (literally ‘living space’), a German ideology proposing expansion to give more space to the German people. This generally advocated expansion on the Eastern front. [52] Another event is the period in history known as the “Scramble for Africa” between 1881 and 1914. It is the period of invasion, occupation, colonization and annexation of African territories by European empires. Africa had a lot of natural resources, ranging from coal to diamonds, and a huge untapped potential. Furthermore the African population was also used as a source of free labor through slavery. As expected, with such resources available, it was not long before war broke out, not only the war against the natives to take control of the area, but also between different European powers to get the biggest piece of the pie. [53] The most recent, and one that is still relevant today, is of the Oil War. Oil war is the term given to all the conflicts over petroleum resources. Oil has become very important in the modern world, as primary source of energy its significance is unprecedented. Petroleum resources alone can run a country’s economy such as the Middle East states, while on the other hand import of oil can cripple a country. The result of all this is that oil became an extremely important resource in the international landscape, a resource that countries wage wars over. There have been a number of conflicts over oil [54]. The Iran-Iraq war (1980-1988), the Gulf War (1990-1991) and the Afghan War (1978-present) are major examples of this. Some conflicts that have control of petroleum resources as an ulterior motive are still present today. The most recent major crisis was Russia’s annexation of Ukraine, this not only demonstrates Russia’s interest in controlling the oil pipeline routes, but also that of United States [55]. And perhaps the most important is the creation of ISIS. ISIS sits on an oil rich land, and even though no state can legally buy oil from them, they do have a steady income through oil sales. This location of ISIS and the possible effect of their oil sales made it important to other foreign powers even before they began global terrorism. [56][57]. As it can be seen, resources are an impetus for war. And as the previous section elaborated, space has a lot of resources available and a space race between various parties is highly probable. Whereas history shows, there is an equally high probability of conflict as a result of this. And in order to avoid this we need a comprehensive treaty system. B. A Failed Treaty As with anything else, it is important to learn from the mistakes already made by others. The same goes with treaties. It is crucial to see where the previous treaties went wrong and to rectify them. In hope of avoiding future conflicts from turning violent and it is of immense importance when considered that now the domain of humans is not just restricted to the Earth, but also includes the space. The Treaty of Versailles of 1919 is one of the most important treaties ever signed. It effectively brought an end of the First World War, though unknown at the time it was signed, it was a significant factor in the start of the Second World War. [58] The treaty’s ultimate failure was when Germany invaded Poland in 1939, dragging the whole of Europe into the Second World War. There were a number of reasons for the failure of the treaty. One was that the treaty had been made by the winners of the war and it was very profitable for them i.e. it was a biased treaty where everyone was not on the table. The treaty was generally against one country (Germany) and the actors at the time felt that as long as one rogue state was controlled, the peace would not be disturbed, but in reality limitations were needed for all the countries. Another reason was that there was no way to reinforce the terms of the treaty, such as the clauses that called for all-around disarmament. None of the countries wanted to reduce armaments, so no one was there to enforce the decision either. And whereas countries like Britain and France had agreed to keep Germany in check, they were not ready to go to war for it or even impose sections that would hurt their own economy. Hence it can be seen that the ultimate weakness of the treaty was that there was no one to stop any country from breaking it.[52] Most of these reasons are generic, and very similar reasons lie behind the failure of almost all treaties. And so when planning for the future, especially in the case of forming 141 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 151. treaties that haveuniversal importance, like as those related to the space, such weaknesses in treaties of the past should be avoided. A few of the common shortcomings that must be avoided are the treaty should not be directed against just a certain country or to control just the rogue states, but be equally applied to everyone; furthermore the treaty needs to absolutely binding, and the powerful states that promise to direct it, must be legally compelled to do. And most importantly there should be measures that make it impossible for states to violate or ignore the treaty. C. A Failed Organization An organization created along with Treaty of Versailles was the League of Nations. It was the predecessor to the United Nations. It had two main aims, one to maintain international peace through “collective security” and the second to promote international cooperation to solve social and economic problems. [59] The former was the more important function of the body. The world had just come out of the First World War, and sought peace and stability. The main idea was to discuss the problems as they rose, and solve them through an understanding of all the countries through a unanimous effort. Though the League looked very good on paper, it was not so successful in reality, with the Second World War an epitome of its failure. There are a number of reasons why the organization failed. Firstly it did not include all the countries, major exclusions were the US, Germany, and USSR. And there were serious weaknesses in the covenant, like unanimous decisions were needed and the League had no force of its own. The result was that it could advise on how to solve problems, but it could not force a state. Furthermore it was a very British and French affair; these were the two most powerful countries at the time and the league was always under their control. As a result the League was prone to letting allies of the major states get away with anything; major examples of this are the Japanese invasion of Manchuria (1931) and the Italian invasion of Abyssinia (1935). [52] Once again it is important to take care of these few parameters in order to form a comprehensive organization that would be much more equipped to deal with problems in a manner that would avoid conflict. The “Regime” is an organization was proposed by the Moon Treaty to for the use of the Moon and other celestial bodies, and it is yet to be formed. Keeping the weaknesses of past organizations in mind, especially those that led to the failure of the League of Nations, the framework of the Regime should be such that it covers all states equally and does not hinder it in its function, yet allows it to be decisive. All countries should get a say in it, but decisions should not be unanimous as it would render it impractical, and similarly it should not give powerful states the veto power, as it too could hinder the organization. D. A Successful Treaty Antarctica is the closest example of space on Earth. Antarctica, like space, has no native inhabitants. This makes the Antarctica Treaty System immensely important. Signed in 1959, the treaty established a few ground rules for every state regarding Antarctica. First off the geographical area of Antarctica was determined. And it was agreed that the area will be used for peaceful purposes only. Freedom of scientific investigation and cooperation was guaranteed. And importantly it prohibited new territorial sovereignty claims, but did not dispute nor recognize the old claims. [60] Further additions were made to the initial treaty, among them, the most important was the Convention on the Regulation of Antarctic Mineral Resource Activities (1988), which made mining illegal in Antarctica even though there coal deposits and hydrocarbons along with other minerals [61]. And the Protocol on Environmental Protection to the Antarctic Treaty (1991) was to insure the protection of the Antarctic environment and ecosystem.[62] The Antarctic treaty has been a success so far. It has managed to peacefully promote Antarctica as a research area and kept it from being militarized. And it has also successfully prevented any conflicts over territory by not recognizing or disputing preexisting claims, and forbidding new claims. Furthermore the environment of Antarctica has been protected and it remains the only continent that has not been affected by the human population. The Antarctic Treaty has been successful, however that does not make it a perfect model to work with in terms of a space treaty. Unlike Antarctica, outer space is not a small potential contributor to world’s demands for mineral resources. Moreover, outer space also does not pose any environment threats to Earth like industrialization of Antarctica could. Furthermore, research in space might very well soon depend on extraction and utilization of space resources to fuel it. Yet on the other hand, the Antarctic Treaty has prevented conflict in the region in an exemplary way, and the features of the treaty that guaranteed this should be replicated which will be discussed latter. The current treaties are inspired by the Antarctic Treaty but success cannot be replicated this way since the conditions are far too different. As a result, the UN must recognize the difference to move forward to full utilization of outer space’s potential, yet in such a manner that global peace is not jeopardized. FLAWS IN THE TREATIES A. The Res Communis Controversy One of the chief controversies is the Res Communis ideology employed especially in the Moon Treaty 1979 [65]. The phrase “common heritage of mankind” contributes to the overall vagueness of the treaty and hence fuels the insecurity of interested nations. Since moon and celestial objects are a “province of mankind” [66], any form of active military involvement is forbidden. However, considering the rise of terrorist organizations like ISIS and rogue nations like North 142 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 152. Korea, both increasinglyusing sophisticated technology; it seems dubious that space would remain untainted from terrorist activities in the future. In the case of USA, however, the reasons for disagreement were political. The treaty put developed countries in a place where they or their private companies could be attacked for using outer space resources by developing countries. This concerned scientists, entrepreneurs and politicians alike as it not only meant a decreased interest in outer space by the government and private groups but also a resulting decline of space science research because if the moon and celestial bodies are to remain unusable, then there is little incentive for them to work. It is true, however, that it could be done under the watch of an “international regime” [67] but that too raises the fear of blackmail since non-space-faring countries are in a majority and most likely to lobby themselves to obtain economic advantages. It is expected that the “international regime” would be similar to “The Enterprise” proposed in Agreement of the Law of the Sea Convention 1994[68]. The Enterprise, by its very nature, was to receive a portion of wealth obtained by the parties exploiting marine resources and distribute it amongst developing countries. Such a system could prove discouraging for space resources as there is a much higher capital required and much greater risk of failure, combining this with low profits means that space resources may remain untouched for a long time even after sufficient technology has been developed. Outer Space Treaty enforced the idea that no state could claim any region in space [69], however it was not very specific about private companies and groups. But the Moon Treaty attempted to make sure that even private groups could not lay claim to any region in space. As expected, it raised a lot of opposition from the business community and it became one of the most controversial points of the treaty. B. Property and Ownership It has been agreed that ownership of space and celestial objects is not possible for countries but there is an apparently silent approval for private groups. This raises further questions and controversies. Firstly, how can the ownership be declared if there is no national government to maintain and uphold the rights of the property owner. Such a situation would force the owners to establish their own system of defense which may lead to an arms race or recurring physical conflict in space. As it is later suggested, the establishment of an organization like “The Enterprise” is critical. Moreover, the UN should be prepared to accept the eventual necessity of a peace-keeping force in space to assist the organization in enforcement of its mandate. C. Lack of Space Involvement On the other end of the spectrum, another hurdle in development of sustainable space policies is a lack of awareness and interest by some countries stemming from a lack of space involvement. These countries often have not developed a significant space involvement and hence deem the development of such policies to be pointless in the present, as the delegate of Argentina replied to COPUOS question regarding definition of Outer Space: “The present magnitude of space and aviation activities and technologies has not given rise to a need for the adoption of specific provisions delimiting outer space [70]”. Though Argentina has CONAE as its space agency, its future plans, however, do not include any form of space exploration, simply satellite management [71], unlike countries like China, Japan, USA, Russia and India as elaborated in Section II. The case of Argentina is not singular but it represents the hesitation of many countries that have too weak space involvement to be able to make decisions concerning their role in it. Hence to play it safe they decide not to indulge in the process of any policy making which they may find even slightly restrictive in the future. Consequently, a large number of developing and some developed countries are not a part of policy making because their national experts and politicians do not have the knowledge and understanding of the importance and significance of space policy making, especially those pertaining to resources because since Moon and celestial bodies have been declared a common heritage of mankind it is important that all countries understand their resulting rights and limitations D. Confidentiality and Privacy The Moon Treaty states “…all space vehicles, equipment, facilities, stations and installations on the moon shall be open to other States Parties…” [72]This could give rise to a lot of possible misuses and controversies such as technology theft, infringement of privacy and right to confidentiality. If it is possible for any state to inspect the equipment of any other state then it is possible that this may be used to replicate the technology of the first state. It poses many questions about the right of having intellectual properties in space. Even though countries and private groups may accept space as a common heritage but it will be unfair for them to lose the rights to all their individual research and development in space. Since such research is extremely expensive, it is very likely that such a jurisdiction would very severely dent the interest in it. Parties operating in space and celestial bodies will also have to disclose the specifics of their activities even though they may be well within their rights. It too opens up debate on human rights, among other things. If, for example, a researcher is studying the health of astronauts on moon, he may have to disclose data regarding this research, which may contain personal information of astronauts. This can be taken as a violation of Universal Declaration of Human Rights [73]. E. Lack of Violation Protocols Both the treaties fail to outline a mechanism that is to be followed if the treaties are breached by a country. United Nations would most likely be unable to undertake any legal or physical action since the Security Council would need 143 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 153. unanimous votes todo so. As a result the treaties appear to restrict less influential and less powerful countries only, since the stronger ones can easily get off without any rebuttal as they can diplomatically procure a veto. The example of Antarctic Treaty needs to be considered here [74]. Interestingly enough, Antarctic treaty was the inspiration for Outer Space treaty and yet such an integral component was left out. According to Antarctic Treaty, in case there is a clash of purposes between two parties, the matter can be taken into International Court of Justice from where the matter could be resolved. Not only does it safeguard the integrity of the treaty but it also makes it more trustable, one of the many reasons of its success. The two treaties were missing a core element of a successful international agreement, the violation protocols and as it was witnessed with the failure of Moon Treaty, other articles are meaningless without them. F. Stifling Space Research and Funding It is true that Outer Space Treaty intends to maximize research and interest in outer space; however, its actual effects have been different. As it has been previously established, the restrictions collectively work to discourage investors. This in turn leads to a direct decrease in the funding of researchers and scientists who are researching on space related areas. Apart from nationally run agencies, few research groups are able to sustain themselves. Consequently, the research that has been done is nowhere near the research that could have been done without the policies. If some of the most restrictive clauses are removed, there would be an immediate rise in research, jobs, interest and awareness regarding space and celestial bodies. Space is an area that captures the imagination of many young scientists but they are often unable to find employment or even advanced education in their area of interest. SOLUTIONS A. Establishment of Extraction-Friendly Framework The current policies need to be modified to encourage both the private and government space agencies to attempt extraction of resources from regions beyond Earth. To that end, several changes need to be made. Private parties should be allowed to have at least temporary property rights so their reservations about lack of control can be met. It makes sense to allow property rights on, for example, asteroids, because after the extraction is over there will be little left of it. Parties should also have a greater degree of privacy and confidentiality. Specific rules and regulations should be framed to ensure that the rights to intellectual property and personal privacy of astronauts are conserved. Moreover the specifics of a regime like “The Enterprise” for space should be discussed and decided. Special attention should be given in making sure that space agencies are not restricted unfairly by the share they have to provide to such a regime. The shares from the extraction should be based on a model such that small scale extraction is bound to a lesser percentage and large scale extraction to a greater percentage. B. Increasing Space Awareness Established agencies like the NASA should work harder to develop a sense of understanding and importance of space resources and hence the importance of policy making governing it. Only after the governments and experts of individual nations realize the significance of such policies and how they are critical in avoiding future clashes can the UN as a whole move forward. It may also be important to get international public support for the policies to pressurize the governments into making decisions in the present than letting them hang on for the future. This requires the use of education and awareness mechanisms through television, newspapers and public talks. C. International Space Property Rights Due to the sheer enormity of space, abundance of celestial bodies and rapid shifting, it may never be possible to divide regions of space in a manner that is fair to all. However, the common heritage ideology may prove extremely impractical in the future. As an analysis of history shows, such a concept is extremely naïve when considered alongside the thousands of wars that people have waged for land and resources. Space agencies, both private and government, need to have a tangible incentive for them to risk a great amount of capital and manpower in attempting to reach for the moon and beyond. Loss of the right to ownership can discourage space exploration. They should be entitled to own, use and sell the resources that they have extracted, to at least some degree. However, concerns of developing nations should also be addressed pertaining to property rights. Their current reduced state of development does not mean that they lose any chance of being a part of the future. Taking both sides into consideration, one method that seems feasible is to employ a procedure whereby regions of space are provided as property only during the period they are being actively used by the party to discourage gluttony. . D. Developing Violation Protocols To uphold the good intentions, enforce articles and develop trust, there needs to be a strong and independent system of justice which ensures that the rights of less powerful and less influential countries remain preserved. To that end, a section of the International Court of Justice could be formed to deal with space law. However, for its smooth operation the policies need to be much more detailed. In fact, it would be wise to draw up an internationally acceptable constitution governing activities in space and celestial objects. Not only would it act as a code of conduct for 144 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 154. parties interested inspace resource extraction but would also help quickly resolve any minor conflicts that may arise. E. Using International Diplomacy Positively Moon Treaty was treated very coldly by the international community. However, an active interest of a single party such as the USA could have totally reshaped its outcome as the decisions of such influential nations are followed by their close allies as well. In fact, an agreement on an international issue between USA, China and Russia almost always gains nearly unanimous global support. As a result bloc leaders like USA and China should take a responsive and positive role in the development of space policies. By considering the interests of their allies, many of those who do not have an independent space program, they can help shape a globally acceptable and beneficial treaty. REFERENCES [1] [Online] Available: www.islandone.org/Treaties/BH766.html [2] Leonard David. (2010, October 4). Is Mining Rare Minerals on the Moon Vital to National Security? [Online]. Available: http://sservi.nasa.gov/articles/is-mining-rare-minerals-on-the- moon-vital-to-national-security/ [3] Beyond Earth [Online]. Available: http://www.nasa.gov/exploration/home/why_moon.html [4] (December 2006). Lunar Exploration Objectives (1sted.)[Online]. Available: http://www.nasa.gov/pdf/163560main_LunarExplorationObjecti ves.pdf [5] [Online]. Available: http://www.permanent.com/lunar-mining- the-moon.html [6] (2013, December 16). China plans to launch Chang'e 5 in 2017 [Online]. Available: http://news.xinhuanet.com/english/china/2013- 12/16/c_132971252.htm [7] (2010, September 16). NEXT STEP FOR ESA’S FIRST MOON LANDER [Online]. Available: http://www.esa.int/Our_Activities/Human_Spaceflight/HSF_Dir ectors_Corner/Next_step_for_ESA_s_first_Moon_lander [8] [Online]. Available: http://lunar.xprize.org/ [9] (2003, August 3). Japan Plans Moon Base By 2030 [Online]. Available: http://www.moondaily.com/reports/Japan_Plans_Moon_Base_B y_2030_999.html [10] China — Measures Related to the Exportation of Rare Earths, Tungsten and Molybdenum [Online]. Available: https://www.wto.org/english/tratop_e/dispu_e/cases_e/ds431_e. htm [11] Kieran Mulvaney. (2012, March 14). The Fight Over Rare Earths: A Primer [Online]. Available: http://news.discovery.com/earth/the-fight-over-rare-earths-a- primer-120314.htm [12] Lesley Stahl. (2015, March 22). Modern life’s devices under china’s grip?[Online]. Available: http://www.cbsnews.com/news/rare-earth-elements-china- monopoly-60-minutes-lesley-stahl/ [13] Leonard David. (2010, October 4). Is Mining Rare Minerals on the Moon Vital to National Security? [Online]. Available: http://www.space.com/9250-mining-rare-minerals-moon-vital- national-security.html [14] (2014, May 9). Russia will begin Moon colonization in 2030 - draft space program [Online]. Available: http://rt.com/news/157800-russia-moon-colonization-plan/ [15] Pat Cahill. (2014, January 20). Mount Holyoke professor Darby Dyar says a colony on the moon is no fantasy [Online]. Available: http://www.masslive.com/living/index.ssf/2014/01/mount_holyo ke_professor_part_of_massive_study_of_outer_space.html [16] DeAnnaKerley. 15 Ambitious Plans to Colonize the Moon [Online]. Available: http://mentalfloss.com/article/53588/15- ambitious-plans-colonize-moon [17] Anna Dolgov. (2014, May 8). Russia Plans to Colonize Moon by 2030, Newspaper Reports [Online]. Available: http://www.themoscowtimes.com/article/499806.html [18] Chris Bergin. (2012, February 29). Exploration Mission 1: SLS and Orion mission to the Moon outlined [Online]. Available: http://www.nasaspaceflight.com/2012/02/exploration-mission-1- sls-orion-debut-mission-moon-outlined/ [19] Chris Bergin. (2012, October 30). EM-2: Orion crew to spend up to four days in Lunar Orbit [Online]. Available: http://www.nasaspaceflight.com/2012/10/em-2-orion-crew- spend-four-days-lunar-orbit/ [20] Future planetary missions [Online]. Available: http://spaceexp.tumblr.com/future-missions [21] Russian Moon missions face three-year delay [Online]. http://www.russianspaceweb.com/spacecraft_planetary_2014.ht ml [22] (2014, October 10). SRI: Russia will begin launching automatic stations to the Moon in 2018 [Online]. Available: http://ria.ru/space/20141003/1026822990.html#ixzz3eXhKtcXA http://ria.ru/space/20141003/1026822990.html [23] Russian Lunar Polar Probes, 2015-2020 [Online]. Available: http://www.permanent.com/russian-lunar-polar-probes.html [24] (2014, August 3) Manned Moon Mission to Cost Russia $2.8 Bln – Space Research Institute [Online]. Available: http://sputniknews.com/russia/20140803/191677434.html#ixzz3 eXiSLeokhttp://sputniknews.com/russia/20140803/191677434.h tml [25] (2014, October 10) Russian Luna-25 Mission to Cost Billions: Space Agency [Online]. Available: http://sputniknews.com/russia/20141010/193919188.html#ixzz3 eXj6Nr00http://sputniknews.com/russia/20141010/193919188.h tml [26] (2015, May 22). China Plans First Ever Landing On The Lunar Far Side [Online]. Available: http://www.spacedaily.com/reports/China_Plans_First_Ever_La nding_on_the_Dark_Side_of_the_Moon_999.html [27] Leonard David. (2015, March 17). China Outlines New Rockets, Space Station and Moon Plans [Online]. Available: http://www.space.com/28809-china-rocket-family-moon- plans.html [28] Alexandra Witze. (2014, March 19). China's Moon rover awake but immobile [Online]. Available: http://www.nature.com/news/china-s-moon-rover-awake-but- immobile-1.14906 [29] Bradley Perret. (2013, December 16). China Plans 2017 Lunar Sample-Return Mission [Online]. Available: http://aviationweek.com/space/china-plans-2017-lunar-sample- return-mission [30] (2015, April 19). JAXA plans Japan’s first moon shot in 2018 [Online]. Available: http://www.japantimes.co.jp/news/2015/04/19/national/jaxa- plans-japans-first-moon-shot-in-2018/#.VY5YSruqqkp [31] Dianne Depra. (2015, April 4). Japan Aiming For Moon Landing By 2018 According To JAXA [Online]. Available: http://www.techtimes.com/articles/48266/20150424/japan- aiming-for-moon-landing-by-2018-according-to-jaxa.htm [32] (2015, April 22). JAPAN TO ATTEMPT UNMANNED LUNAR LANDING IN 2018 [Online]. Available: http://www.spaceflightinsider.com/organizations/jaxa/japans- jaxa-to-attempt-lunar-landing-in-2018/ 145 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 155. [33] (2010, Sep5). ‘We’re launching Chandrayaan-2 for a total coverage of the moon’ [Online]. Available: http://timesofindia.indiatimes.com/home/opinion/interviews/We re-launching-Chandrayaan-2-for-a-total-coverage-of-the- moon/articleshow/6501413.cms?referral=PM# [34] India to launch Chandrayaan-2 by 2017 [Online]. Available: http://www.thehindu.com/sci-tech/science/india-to-launch- chandrayaanii-by-2017/article5562361.ece?ref=sliderNews [35] Asteroids [Online]. Available: http://nssdc.gsfc.nasa.gov/planetary/text/asteroids.txt [36] Stephen Shaw. (2012, August 21). Posts Tagged ‘C-type asteroids’ Asteroid Mining [Online]. Available: http://www.astronomysource.com/tag/c-type-asteroids/ [37] Endangered Elements [Online]. Available: http://www.acs.org/content/acs/en/greenchemistry/research- innovation/research-topics/endangered-elements.html [38] David Cohen. (2007, May 23). Earth's natural wealth: an audit [Online]. Available: http://www.sciencearchive.org.au/nova/newscientist/027ns_005. htm [39] Periodic Table [Online]. Available: http://www.rsc.org/periodic- table/ [40] Stephen Harris. (2013, April 8). Your questions answered: asteroid mining [Online]. Available: http://www.theengineer.co.uk/aerospace/in-depth/your- questions-answered-asteroid-mining/1015966.article [41] (2012, April 2). Asteroid Retrieval Feasibility Study [Online]. Available: http://kiss.caltech.edu/study/asteroid/asteroid_final_report.pdf [42] Marc Cohen. Robotic Asteroid Prospector (RAP) Staged from L-1: Start of the Deep Space Economy [Online]. Available: http://www.nasa.gov/offices/oct/early_stage_innovation/niac/20 12_phase_I_fellows_cohen.html [43] Mike wall. (2013, January 22). Asteroid-Mining Project Aims for Deep-Space Colonies [Online]. Available: http://www.space.com/19368-asteroid-mining-deep-space- industries.html [44] (2013, January 22). Commercial Asteroid Hunters announce plans for new Robotic Exploration Fleet [Online]. Available: http://www.prweb.com/releases/2013/1/prweb10346181.htm [45] Taylor Soper. (2013, January 22). Deep Space Industries entering asteroid-mining world, creates competition for Planetary Resources [Online]. Available: http://www.geekwire.com/2013/deep-space-industries-entering- asteroidmining-world-creates-competition-planetary-resources/ [46] http://www.planetaryresources.com/technology/#space-based- observation [47] (2012, April 24). Company aims to mine resource-rich asteroids [Online]. Available: http://lightyears.blogs.cnn.com/2012/04/24/companies-plan-to- mine-precious-metals-in-space/?hpt=hp_t3 [48] (2012, April 24). Plans for asteroid mining emerge [Online]. Available: http://www.bbc.com/news/science-environment- 17827347 [49] (2013, August 8). New NASA Mission to Help Us Learn How to Mine Asteroids [Online]. Available: https://www.nasa.gov/content/goddard/new-nasa-mission-to- help-us-learn-how-to-mine-asteroids [50] (2015, June 22). OSIRIS-REx Team Prepares for Next Step in NASA’s Asteroid Sample Return Mission [Online]. Available: https://www.nasa.gov/feature/goddard/osiris-rex-team-prepares- for-next-step-in-nasa-s-asteroid-sample-return-mission [51] Michael Buet and Robert Frantz. ASTEROID MINING: WHY AND HOW? [Online]. Available: http://www.friendsofbest.org/wp- content/uploads/2015/05/ASTEROID-MINING_WHY-AND- HOW-Article_03.pdf [52] Norman Lowe, Mastering Modern World History, 4th edition, New York: Palgave Macmillan, 2007 [53] Andrina Stiles, "Bismark's Germany" in The Unification f Germany, 1st edition, London, England: Hodder& Stoughton, 1998, pp. 85-99 [54] John Lowe and Robert Pearce, "Colonial Rivalries, 1870-1914" in Rivalry and Accord - International Relations 1870-1914, 1st edition, London, England: Hodder& Stoughton, 2007, pp. 39-69 [55] Michael Klare, The Race for What's Left: The Global Scramble for the World's Last Resources, 1st edition, New York: Picador, 2013. [56] Dr. Nafeez Ahmed. (2014, March 6), Ukraine crisis is about Great Power oil, gas pipeline rivalry [Online]. Available: http://www.theguardian.com/environment/earth- insight/2014/mar/06/ukraine-crisis-great-power-oil-gas-rivals- pipelines [57] Keith Johnson. (2014, September 24), U.S. Strikes ISIS Oil Installations [Online]. Available: http://foreignpolicy.com/2014/09/24/u-s-strikes-isis-oil- installations/ [58] Treaty of Versailles, Paris, 28 June 1919, League of Nations Treaty Series, available from https://en.wikisource.org/wiki/Treaty_of_Versailles [59] Covanent of League of Nations, available at https://en.wikisource.org/wiki/Covenant_of_the_League_of_Nat ions [60] Antarctic Treaty, Washington, 1 December 1959, available from http://disarmament.un.org/treaties/t/antarctic/text [61] Environmental Treaties and Resource Indicators (ENTRI), Available From:http://sedac.ciesin.columbia.edu/entri/texts/acrc/cramra.txt .html [62] Protocol on Environmental Protection to the Antarctic Treaty, 1991, available from http://www.antarctica.ac.uk/about_antarctica/geopolitical/treaty/ update_1991.php [63] Outer Space Treaty, Washington DC, 10 October 1967, available from http://www.unoosa.org/oosa/en/ourwork/spacelaw/principles/leg al-principles.html [64] Moon Treaty, New York, 18 December 1979, available from http://www.unoosa.org/oosa/en/ourwork/spacelaw/treaties/moon -agreement.html [65] Para 1, Article 11, Agreement Governing the Activities of States on the Moon and Other Celestial Bodies, United Nations, 1979 [66] Article 1, Treaty On Principles Governing The Activities Of States In The Exploration And Use Of Outer Space, Including The Moon And Other Celestial Bodies, United Nations, 1967 [67] Para 5 -7, Article 11, Agreement Governing the Activities of States on the Moon and Other Celestial Bodies, United Nations, 1979 [68] Part XI, Agreement of the Law of the Sea Convention, United Nations, 1994 [69] Article 2, Treaty On Principles Governing The Activities Of States In The Exploration And Use Of Outer Space, Including The Moon And Other Celestial Bodies, United Nations, 1967 [70] A/AC.105/889/Add.14, Committee on the Peaceful Uses of Outer Space, United Nations General Assembly [71] CONAE website : http://www.conae.gov.ar [72] Para 1, Article 15, Agreement Governing the Activities of States on the Moon and Other Celestial Bodies, United Nations, 1967 [73] Article 12, Universal Declaration of Human Rights,1948 [74] Para 2 Article 11, The Antarctic Treaty, 1959 146 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 156. Feasibility Study toInstall Firefighting Equipment on a Cargo Helicopter S. Shahid1 , B. Saeed2 , S. Javed3 Department of Aerospace Engineering National University of Sciences and Technology Pakistan 1 msohailshahid135@gmail.com 3 saad.javaid@gmail.com Abstract— Aerial firefighting has seen recent interest with events such as forest and domestic fires seen on the rise across the world. Resources such as fixed wing aircraft and helicopters have been employed for use in firefighting setups. Aerial firefighting began around 1920 with first attempts at dropping water aircraft onto a fire. Most of these attempts were unsuccessful during this era, but they provided a platform for future initiatives. In 1935, the Aerial Fire Control experimental Project was created. At this point, aircraft became important for fire detection. Inspired by agricultural spraying techniques, “helitanker” soon became the next progression in aerial firefighting. With firefighting helicopters came “heli-rapellers” – firefighters who rappelled down to the fire as a ground force. In the 1960s, surplus military aircraft became another effective way to combat wildfires. Planes like the B- 25, Douglas B-26 and Lockheed PSV could carry 1000 gallons of water. Similarly some helicopters use fire watch cameras for the surveillance of fire. Its infrared thermal imager can detect heat of a wild fire even through thick smoke. The aim of this project is to carry out feasibility study of installation of firefighting equipment on MI-171 helicopter and determine its practicability. A study will be carried out that whether MI-171 can be used for firefighting purposes. MI-171 has the main advantage of its reliability. By virtue of its qualities MI-171 is successfully operated in different conditions – whether it is heat or cold, rain or snow, mountings or desert, sea or land. It has wide range of mission accomplishments like transportation of up to 4000 kg cargo inside cargo compartment, transportation of up to 4000 kg cargo on external sling, SAR missions, fire-fighting, patrolling, construction works, offshore oil rig flights, ambulance mission etc. Keeping all the above traits in view MI-171 can be made capable of fighting the wild fires. As it has the ability to carry a lot of weight inside its cargo cabin and mounted on external sling, these areas will be analyzed especially. For this first of all volume and weight capacity of cargo section of MI 171 helicopters will be carried out. After that all the forces applied on the helicopter structure will be estimated. Based on these forces the aircraft modified CAD model will be analyzed numerically in ANSYS software. Main focus will be on the structural and aerodynamic analysis of the helicopter with water tank installed inside the cargo section. For structural analysis, Workbench software will be used. The results will be validated by solving the problems analytically. Index Terms— Aerospace, Firefighting, Helitanker Stress analysis INTRODUCTION Wildfires are nature’s most dangerous, terrifying and recurring phenomenon. Similarly periodic wildfires are a natural part of echo system dynamics. A large fir has an average cost of between $2.1 million and $4.5 million (which includes suppression costs, social costs and the magnitude of insured losses). On the other hand a successful initial attack on the same large fire saves $3.3 million. Aerial firefighting has seen recent interest with deadly events such as forest and domestic fires went on the rise across the globe. Resources such as fixed wing aircraft and helicopters have been employed for use in firefighting setups. Aerial firefighting is the use of aerial resources to suppress fires. Many aircrafts like fixed-wing, helicopters, smoke jumpers and rappellers are used for this purpose. A wide variety of terminology is used in the media for the aircrafts used in aerial fighting. The term “air tanker” is used for the fixed wing aircraft generally. One of the major contributor towards aerial firefighting is helicopter. Helicopters can be fitted with buckets or tanks, called helitanker for firefighting purpose. Some of them are also fitted with foam cannons which are mounted on their cannons. Aerial firefighting began around 1920 with first attempts at dropping water tanks onto a fire. Most of these attempts were unsuccessful during that duration, but they provided a platform for initiatives in the future. In 1935, the Aerial Fire Control experimental Project was created. At this point, aircraft became important for fire detection. Inspired by agricultural spraying techniques, “helitanker” soon became the next progression in aerial firefighting. With firefighting helicopters came “heli-rapellers” –firefighters who rappelled down to the fire as a ground force. In the 1960s, surplus military aircraft became another effective way to suppress wildfires. Aircrafts like the B-25, Douglas B-26 and Lockheed PSV were able to carry 1000 gallons of water. Similarly some helicopters use fire-watch cameras for fire surveillance. Its infrared thermal 147 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 157. imager can detectheat of a wild fire even through thick smoke. The aim of this project is to carry out feasibility study of installation of firefighting equipment on MI-171 helicopter and determine its practicability. A study will be carried out that whether MI-171 can be used for firefighting purposes. By virtue of its qualities MI-171 is successfully operated in different conditions – whether it is heat or cold, rain or snow, mountings or desert, sea or land. It has wide range of mission accomplishments like transportation of up to 4000 kg cargo inside cargo compartment, transportation of up to 4000 kg cargo on external sling, SAR missions, fire-fighting, patrolling, construction works, offshore oil rig flights, ambulance mission etc. a design of the firefighting equipment, that is to be installed will be given first, and then the whole design will be analyzed. METHODOLOGY Firefighting helicopters are classified in two categories based on firefighting mechanism and apparatus Helitanker Bambi bucket Both have their own advantages and disadvantages which are as follows Firefighting with helitanker aircraft speed limits are not that much affect while Bambi bucket has the limitation of maximum speed i.e. o 80 – 120 kph (with empty bambi bucket) o 60 – 180 kph (with filled bambi bucket) o 60 – 80 kph (while dumping water) If we exceed these limits the bucket starts swinging forth and backward, and even it can touch the tail rotor of the helicopter During operation aircraft CG is not disturbed in case of helitanker while the CG shifts in case of bambi bucket Extra care is required during take-off and landing In Bambi bucket, the fire retardant can only be concentrated at the point exactly below the opening of the bucket, no other direction is possible, as it falls with action of gravity Based on the above discussion we selected to operate from a water tank (helitanker) installed the cargo section of our Helicopter. Now for this first of all the volume and weight capacity of cargo compartment were evaluated from manuals of the aircraft. Volume capacity of helicopter 12.5 𝑚3 and weight capacity is 4000 kg. First of all a water tank was designed in CATIA software. Its volume capacity was 3200𝑚3 . To avoid column buckling and resist the circumferential stresses, bulk heads and stringers were integrated with internal skin of the water tank. These bulk heads are having hole so that water can easily pass in between the section of the tank. These bulk head structures also avoid sloshing effect. Similarly to resist axial and bending loads formers were integrated on internal skin of the water tank. Figure 1: Helitanker Figure 1: Helitanker Now to install this equipment on the floor of cargo compartment, there are mounts available on it for the auxiliary fuel tanks. Auxiliary fuel tanks are extra fuel tanks which are carried by the helicopters to increase the range and endurance. Once the fuel from the main fuel tanks is consumed completely, then these are connected to the engine. Generally the volume capacity of a single tank is almost 800𝑚3 . A cargo helicopter can carry 4 auxiliary fuel tanks. But the distances among them are according to the dimension of the stands of auxiliary fuel tanks. A model simulating the floor of the aircraft was generated in CATIA software. The hard points are prominent in the following diagram. Figure 2: Floor of Cargo Compartment Figure 2: Floor of Cargo Compartment 148 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 158. So a newstand was designed according to the dimensions of helitanker. The stand was installed on a layer of CFRP sheet which has the mounts in accordance with dimensions of cargo section. Figure 3: Stand Figure 3: Stand Water will be released from emergency hatch of cargo aircraft which will go down with action of gravity. For this purpose an opening is provided in the CFRP sheet. This sheet will fix on the floor with the help of screws just as the stands for auxiliary fuel tanks are fixed. The dimensions and the distance between the threaded holes is in accordance with hard points available on the floor. Similarly the stand to hold the water tank is fixed on the same sheet. This whole package when combined together completes the structural design of the firefighting equipment. This whole structure is fixed on the hard points already available for auxiliary fuel tanks. Figure 4: Complete Design. As our main focus is on the weight constraint during the design phase so a high strength to weight ratio composite “CFRP” is selected for manufacturing and further analysis. Material properties will be computed later on. Figure 4: Complete Design For pumping of water, an Electric In-Line axial flow pump will be used in whish an axial flow pump is mounted on the free end of a snorkel which is fixed on the onboard storage tank on the other end. Figure 5: Snorkel Pump. We could also use the hydraulic centrifugal pump but then we would have faced some weight penalty. Similarly the low power requirement of electrical motor makes it suitable for small helicopters also. In addition to its advantage of low weight, it is axially balanced with snorkel, and thus easily supported and in- line with axis of the snorkel. This electric pump can deliver 1000 gpm through a 6” diameter snorkel. The snorkel length varies from 12 to 15 feet. It operates on 7.5 horsepower. It will draw this power from onboard generator. In this invention, an electrical power cable extends from the electrical source of helicopter along the outside of the snorkel and goes into the protective shrouding on the pump unit. Once the pump is lowered into the water reservoirs such as lakes, swimming pools etc. the shrouding protects the pump unit from any hazardous material such as rocks or wall of pool during filling. Figure 5: Snorkel Pump Now to spray the fire retardant on the burning fire, doors are extended from the helitanker through emergency hatch of the floor. The maximum flow rate from this opening will be 1.19 m3 /s. Once the doors are opened, the fire retardant will fall directly from the helitanker with action of gravity. These doors are controlled by microprocessor which actually adjusts for airspeed and allow the doors to open for the flow of water. The flow rate matches the particular coverage selected by the pilot himself. Figure 6: Helitanker Figure 6: Helitanker 149 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 159. MATERIAL PROPERTIES For analysis,material properties of CFRP were calculated. To have symmetric properties in all the three direction, Quasiisotropic (0, 90, +45, and 45)Figure 7: Stacking Sequence sequence was selected as stacking sequence for our material Figure 7: Stacking Sequence To find out the properties of CFRP, “ANALYSIS AND DPERFORMAANCE OF FIBRE CCOMPOSISTEES”, written by BHAGWWAN D. AGAARWWAL was concerned, in CFRP carbon is fiber while epoxy was as matrix, following calculations were carried out The expression for elastic modulus is as 𝐸𝑐 = 𝐸𝑓 𝑉𝑓 + 𝐸 𝑚 𝑉𝑚 We have 𝐸𝑓 = 220 𝐺𝑃𝑎 𝐸 𝑚 = 3.54𝐺𝑃𝑎 𝑉𝑓 = 0.5 𝑉𝑚 = 0.5 By putting these values 𝐸𝑐 = 220 ∗ 0.5 + 3.54 ∗ 0.5 𝐸𝑐 = 112 𝐺𝑃𝑎 Now to find out elastic modulus in other two direction 𝐸 𝑇 𝐸 𝑚 = 1 + Ƹƞ𝑉𝑓 1 − Ƹƞ𝑉𝑓 ƞ = 𝐸𝑓 𝐸 𝑚 ⁄ − 1 𝐸𝑓 𝐸 𝑚 ⁄ + Ƹ Putting values in above eqn we get ƞ = .95 𝐸 𝑇 = 17.61 𝐺𝑃𝑎 Similarly 𝐺 𝐿𝑇 becomes 𝐺 𝐿𝑇 = 3.60 𝐺𝑃𝑎 To find out the elastic modulus we will find out the A matrix for above sequence For this we need Q matrix for each ply 𝑄 = [ 𝑄11 𝑄12 𝑄16 𝑄21 𝑄22 𝑄26 𝑄61 𝑄62 𝑄66 ] 𝑄11 = 𝐸𝐿 1 − 𝑣 𝐿𝑇 𝑣 𝑇𝐿 𝑄11 = 134.5 𝐺𝑃𝑎 𝑄22 = 𝐸 𝑇 1 − 𝑣 𝐿𝑇 𝑣 𝑇𝐿 𝑄22 = 17.7 𝐺𝑃𝑎 𝑄12 = 4.619 𝐺𝑃𝑎 𝑄66 = 𝐺 𝐿𝑇 = 3.69 𝐺𝑃𝑎 The Q matrix becomes for 0 deg ply becomes 𝑄 = [ 134.5 4.6 0 4.6 17.75 0 0 0 3.6 ] For lamina at any angle the Q matrix becomes 𝑄̅ = [ 𝑄̅11 𝑄̅12 𝑄̅16 𝑄̅21 𝑄̅22 𝑄̅26 𝑄̅61 𝑄̅62 𝑄̅66 ] Above values can be calculated from eqns as follows 𝑄̅11 = 𝑄11 𝑐𝑜𝑠4 𝜃 +𝑄22 𝑠𝑖𝑛4 𝜃 + 2(𝑄12 + 2𝑄66) 𝑠𝑖𝑛2 𝜃 𝑐𝑜𝑠2 𝜃 𝑄̅22 = 𝑄11 𝑠𝑖𝑛4 𝜃 +𝑄22 𝑐𝑜𝑠4 𝜃 + 2(𝑄12 + 2𝑄66) 𝑠𝑖𝑛2 𝜃 𝑐𝑜𝑠2 𝜃 𝑄̅12 = 𝑄12(𝑐𝑜𝑠4 𝜃 + 𝑠𝑖𝑛4 𝜃) + (𝑄11 + 𝑄12 − 4𝑄66) 𝑠𝑖𝑛2 𝜃 𝑐𝑜𝑠2 𝜃 𝑄̅66 = 𝑄66(𝑐𝑜𝑠4 𝜃 + 𝑠𝑖𝑛4 𝜃) + (𝑄11 + 𝑄22 − 2𝑄12 − 2𝑄66) 𝑠𝑖𝑛2 𝜃 𝑐𝑜𝑠2 𝜃 𝑄̅66 = (𝑄11 − 𝑄12 − 2𝑄66 )𝑐𝑜𝑠3 𝜃 sin 𝜃 + (𝑄22 − 𝑄12 − 2𝑄66) 𝑠𝑖𝑛3 𝜃 cos 𝜃 𝑄̅66 = (𝑄11 − 𝑄12 − 2𝑄66 )𝑠𝑖𝑛3 𝜃 cos 𝜃 + (𝑄22 − 𝑄12 − 2𝑄66) 𝑐𝑜𝑠3 𝜃 sin 𝜃 Q matrix for 45 deg lamina becomes 𝑄̅ = [ 44.14 36.76 29.09 36.76 43.96 29.09 29.09 29.09 3.6 ] 150 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 160. Q matrix for-45 deg lamina becomes 𝑄̅ = ⌈ 44.14 36.76 −29.09 36.76 43.96 −29.09 −29.09 −29.09 35.76 ⌉ Q matrix for 90 deg lamina becomes 𝑄̅ = [ 17.75 4.6 0 4.6 134.5 0 0 0 3.6 ] Now material properties can be found as 𝐸 𝑋 = 𝐴11 𝐴22 − 𝐴12 2 𝐴22 𝑡 𝐸 𝑦 = 𝐴11 𝐴22 − 𝐴12 2 𝐴11 𝑡 𝑣𝑥𝑦 = 𝐴12 𝐴22 𝑣 𝑦𝑥 = 𝐴12 𝐴11 𝐺𝑥𝑦 = 𝐴66 𝑡 So from above expressions we found out the matrix as follows Total thickness is “10 mm” & number of lamiae are 20, so the total thickness of each angled ply becomes 2.5 mm [𝐴] = 2.5 ∗ [[𝑄̅]0 + [𝑄̅]90 + [𝑄̅]45 + [𝑄̅]−45] Putting all the related values in the above equation we get the A matrix as ⌈𝐴⌉ = [ 601.25 206.8 0 206.8 600.4 0 0 0 196.8 ] 𝐺𝑃𝑎 Putting the values from above in the expression of Ex, Ey, Vxy, Vxy & Gxy, we get 𝐸𝑥 = 53 𝐺𝑃𝑎 𝐸 𝑦 = 53 𝐺𝑃𝑎 𝐺𝑥𝑦 = 19.68 𝐺𝑃𝑎 𝑣𝑥𝑦 = 0.3449 𝑣𝑥𝑦 = 0.3439 The above calculated values were verified by using NASTRAN software. Figure 8: Nastran Pattern Results These values will be embedded in ANSYS Workbench as engineering data in orthotropic properties. Further computation will be carried out in the next steps. Figure 8: Nastran Pattern Results COMPUTATIONAL ANALYSIS To check out for the structural integrity of the design, this CATIA file was converted in “IGES” file. This igs file was imported in ANSYS workbench. The static structural analysis was carried out using different conditions. First of all the water tank, structural analysis of each part is carried out separately. The design of water alone was analyzed. The water of pressure 151 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 161. was applied oninternal skin of the water tank. The pressure applied was 2565 pa. Now to carry out these simulations fine meshing was created. After that, the structure was put on simulations. The following result was obtained.Figure 9: Simple Analysis of Helitanker Figure 9: Simple Analysis of Helitanker The values shown along the colored graph are “Von- Mises” or equivalent stresses. According to these results the maximum stress lies on the internal corner of the opening of the tank. This maximum stress is in quite safe limits. Similarly the frame was analyzed separately and the force was applied which would be applied on it when the water tank is filled. Figure 10: Static Structural Analysis of Stand Figure 10: Static Structural Analysis of Stand From the above results, still it is evident that maximum point of maximum equivalent stress, the material will not fail. After these simulations, the whole design, completely assembled, was analyzed in the static position (Helicopter is at rest with tank filled with water). The forces applied were changed into pressure form because these forces were distributed across the surfaces. Figure 11: Static Analysis of Complete Structure Figure 11: Static Analysis of Complete Structure The results obtained still shows that the point of maximum stress is located at the corner of the door. The maximum Von- Misses stress still under the safe considerations. Now to carry out the analysis while the aircraft is in the move climb or any bank, values of gs was embedded in the initial conditions. For this, first acceleration was converted into inertial loads and this inertial load was applied as the nodal force on the bodies with pink color in the following diagramFigure 12: initial Conditions for Dynamic Analysis Figure 12: initial Conditions for Dynamic Analysis 152 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 162. After running thesimulations, following results were given by ANSYS workbench. Values in the green boxes shows Von Misses stress at that location.Figure 13: Results from 4g Loads Figure 13: Results from 4g Loads The maximum equivalent stress shows an abrupt rise due to the acceleration. Similarly the total maximum deformation was too high. Similar analysis was carried out on inertial load with respect to 5g acceleration. The following results were obtained. When simulations for 5g inertial loads were applied on the whole structure, the maximum equivalent stress increased by a big amount. This value is still under the elastic limit range, so the design will not fail. Figure 14: Results of 5g Loads Figure 14: Results of 5g Loads Figure 15: Deformation at 5g Loads Frrom the results of total deformation the maximum deformation, the maximum deformation is 0.11176 m which is too high and strucure will eventually break. So with this design, we have to keep ouselve within 4g limit. We must try to stay maximum at 3g. this holds good for all manuvers, translations and rolls. CONCLUSION: With this we can conclude that any cargo helicopter with weight lifting capacity of 4000 kg or more can be modified so that it can be used in Aerial firefighting. For that a helitanker is installed inside the cargo section of the aircraft, which will by a snorkel from the water reservoirs easily. If we use light composites like CFRP, it will have enough strength to resist flight variations up to 3g. The helitanker will have the capacity to carry 3200 liters of water. Compared with firefighting helicopters, it will be an efficient helicopter. REFERENCES [1] F. Qian, V. Strusevich, I. Gribkovskaia, and Ø. Halskau, "Minimization of passenger takeoff and landing risk in offshore helicopter transportation: Models, approaches and analysis," Omega, vol. 51, pp. 93-106, 2015. [2] D. G. Nichols Sr, "Electric in-line snorkel pump for helicopter tanker and method of operation," ed: Google Patents, 2001. [3] E. G. Keating, A. R. Morral, C. C. Price, D. Woods, D. M. Norton, C. Panis, et al., Air Attack Against Wildfires: Rand Corporation, 2012. [4] B. D. Agarwal, L. J. Broutman, and K. Chandrashekhara, Analysis and performance of fiber composites: John Wiley & Sons, 2006. 153 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 163. Design and Optimizationof S-Band Wilkinson Power Divider for Transceiver Applications Muhammad Saad Sohail1 , Talha Khan2 1, 2, Space and Upper Atmosphere Research Commission (SUPARCO), Karachi, Pakistan 1 saad435@hotmail.com,2 aquarius_132@hotmail.com Abstract—This paper presents a high performance and compact size single band Wilkinson Power divider for wireless applications. Basic three-port equations are used for designing the Wilkinson at the operating frequency range, further tuning is done by converting the lengths and impedances of each section of conventional Wilkinson power divider into single band T-shaped sections. ADS 2011 is used for the schematic design and EM simulations are performed using Momentum. The reflection coefficients are analyzed at the frequency range of 2- 2.25GHz.The computed and the measured S-parameters are in a very good agreement. However, substrate cost is the leading aspect for micro strip passive circuits, there is a requirement to make use of cheaper materials such as FR4 without compromising the performance. In this paper, FR4 has been used as a substrate material for the development of Wilkinson power dividers operating in portions of the S-band (2.5-2 GHz). Keywords—Micro strip, T section, Wilkinson Power Divider, Optimize Transceiver. I. INTRODUCTION Wilkinson power divider is a commonly used device in microwave applications. Quarter wave transformers are used in Wilkinson power dividers which are easy to develop on printed circuit board using inexpensive method whereas still providing high level of performance. As PCB transmission line is employed, lumped and micro-strip line can be used for accuracy in results. The designed power dividers are simple in approach yet provide excellent performance in terms of insertion loss, port matching and isolation. WPD divides input power into two output ports. In ideal conditions, this split is lossless. The divider requires all ports to be matched. Since, three port are never reciprocal and matched at the same time, a resistor is introduced which absorbs energy due to any mismatch. This resistor also isolates output ports which makes this device to also be used as a power combiner. II. BRIEF INTRODUCTION TO WILKINSON POWER DIVIDER Loss less power dividers can be a formed a WPD if the output ports are matched. Reflected power is dissipated. Input power can be split into two or more in-phase signals with the same amplitude. Micro-strip based equal split WPD is feasible to make without any lumped components. The design considered in this paper is micro strip based, as shown below in Fig. 1. The equivalent transmission line circuit is shown in Fig. 1. Design for center frequency of 2-2.25 GHz and Z0 = 50 Ω requires the isolation resistor to be 2Z0= 100 Ω and the impedance of the quarter-lambda transmission line split section to be √2 Z0 = 70.70 Ω. Fig.1.Impedance description of Wilkinson power divider S-parameters are measured by applying signal to each port. The reflection power waves are measured at remaining ports. Calculation of S-parameters is the ratio between incident and reflected power waves. S-parameter for an ideal Wilkinson Power Divider has the following properties: • All ports are matched (S11, S22 and S33 equals 0). • Ports 2 and 3 are isolated (S23 and S32 equals 0) • The power present at port 1 is equally divided between ports 2 and 3 (S12, S13, S21and S31 equals -3 dB with a 90° phase shift) • The circuit is reciprocal (Sij equals Sji) III. DESIGN Initial design problem was to analyze a non-optimum Wilkinson Power Divider and then to optimize it using Agilent ADS and produce the usual values for line impedance (Z) and isolation resistance (R). For this task, the single stage Wilkinson Power Divider was implemented as shown in figure (3) using the ideal transmission line model. To determine the initial values of line impedance (Z) and isolation resistance (R), two variables were defined as: r1 = 50 154 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 164. Ω opt {10Ω to 500 Ω}, z1 = 50 Ω opt {20 Ω to 150Ω} The nominal values of r1 and z1 were specified to be 50 Ω with r1 ranging between 10 Ω to 500 Ω and z1 ranging between 20 Ω to 150 Ω. The nominal value was defined to be taken as the seed value whereas the range was defined as an optimization window. The simulation settings as mentioned below were set to obtain the S – parameters, Start = 1.0 GHz; Step = 250 MHz; Stop = 4.0 GHz The Optimization goals were set as: dB (S11) < – 35; dB(S13) = – 3.01 and dB(S12) < –50. LineCalc tool of ADS 2011 was used to convert the ideal micro strip lines to the practical values as shown in Figure 2. Fig.2. LineCalc tool of ADS for converting ideal transmission lines to micro strip line Quarter wave transmission line (λg/4) is designed the required frequency. The length λg/4 could be folded using the curved structure to reduce the size of design as shown in Fig. 3. Fig.3. Schematic design of Wilkinson Power divider Layout was generated as shown in Figure 4 using Fr4 substrate and EM simulations were performed in Momentum. The results were slightly deviated due to EM effects. Therefore, co-simulation was performed to optimize the results. Fig.4. Layout design of final Wilkinson Power divider IV. SIMULATION AND COMPARISON RESULTS The design was the simulated and optimized on FR4 and AD450 substrates for comparison reasons having same thickness and relative permittivity but different loss tangents. The various properties of the substrates used are shown in following tables. Substrate parameters for Fr4 Er 4.5 H 0.508mm T 17um Cond 5.88e7 Rough 0mm TanD 0.016 Substrate parameters for AD450 Er 4.5 H 0.508mm T 17um Cond 5.88e7 Rough 0mm TanD 0.0045 The final measurement results of S-parameters as obtained from Vector Network Analyzer (VNA) for both the substrates are compared in Figures 5, 6, 7 and 8: 155 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 165. Fig.5. Reflection parametersusing AD450 substrate Fig 6.Reflection parameters using AD450 substrate Fig.7.Reflection parameters using FR4 substrate Fig.8.Reflection parameters using FR4 substrate As can be seen from the results of above plots, that performance of Wilkinson power divider is almost same for both the substrates i.e. low cost FR4 substrate and high cost and low loss AD450 substrate. Therefore, it can be said that the design fits well for any substrate and is very useful for the cheaper substrates without compromising the performance. V. CONCLUSION RF transceiver commonly have a Wilkinson power divider (WPD) or combiner circuit, techniques for designing WPD are diverse. The paper shows comparison of WPD design on two substrates having different loss tangent. FR4 being cheap and easily available is defined as equally comparable to AD450 in simulations. Next step is the fabrication and testing of these WPD. The final circuit occupies less space and has good matching results. With the thickness of 0.5mm, FR4 has comparable behavior to specialized substrates and can be considered as a robust and cost effective alternative for applications in S-band. REFERENCES [1] Cripps, S.C., RF Power Amplifiers for Wireless Communications 2nded, Norwood,MA: Artech House, 2006 [2] David M. Pozar, Microwave Engineering, John Wiley & Sons Inc., 2005 [3]Guillermo Gonzalez,”Microwave Transistor Amplifer Analysis and design, Prentice Hall [4] Maas, Stephen A.Nonlinear microwave and RF circuits, “Artech House microwave library”, 1997. 2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.452.00 2.50 -36 -34 -32 -30 -28 -26 -24 -38 -22 freq, GHz dB(S(1,1)) dB(S(2,2)) m1 freq= dB(S(2,1))=-3.272 2.250GHz m2 freq= dB(S(3,3))=-36.784 2.250GHz 2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.452.00 2.50 -35 -30 -25 -20 -15 -10 -5 -40 0 freq, GHz dB(S(2,1)) m1 dB(S(3,3)) m2 m1 freq= dB(S(2,1))=-3.272 2.250GHz m2 freq= dB(S(3,3))=-36.784 2.250GHz 156 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 166. OPTIMIZE MANUFACTURING OF UNIDIRECTIONALCARBON PREPREGS FOR SPACE APPLICATIONS Mateen Tariq Ahmad Bashir, Dr. Sohaib Akbar, Dr. Sajid Mirza Pakistan Space and Upper Atmosphere Research Commission Karachi, Pakistan Mateen_90@hotmail.com Abstract— High performance composite prepregs are used globally in aerospace applications for faster and easier manufacturing. Manufacturing composite parts using unidirectional carbon prepregs offers flexibility to optimize the design and layers, resulting in superior properties. Using prepregs as manufacturing medium, parts can be manufactured with low void content. Unidirectional prepregs give control over thickness hence high specific modulus and strength can be achieved. Fatigue and acoustic performance of composite is strictly related to type of reinforcement. Unidirectional reinforcement plies offer good fatigue and acoustic performance making them suitable candidates for space manufacturing. Superior properties make them viable for manufacturing of various parts of Launch vehicles such as payload fairing, motor casings, fuel tanks and shims for flexible nozzle assembly. This paper deals with the in-house manufacturing of unidirectional carbon prepregs. The prepregs were manufactured using carbon fiber tows and epoxy based prepreg resins. The fiber from tow was mechanically placed parallel to each other resulting in unidirectional carbon fiber prepregs. A solvent dip process using resin bath and roller was used to manufacture prepregs. To stock the batch of prepreg, cold storage system was utilized. The process was developed using customized machinery and equipment. Tests on prepregs showed satisfactory results in terms of volume and weight fractions. Finally the prepregs were cured and formed into unidirectional sheets for further tests. Prepregs were cured in autoclave utilizing vacuum bagging operation. The cured sheets were subsequently tested which showed competent results hence asserting the material as suitable candidate for use in space applications. Index Terms— Unidirectional Carbon, Prepregs, Polymer matrix composites. I. INTRODUCTION Space applications demand materials with high specific stiffness and strength so that weight of structure may be minimized. This need of space industry has lead towards development of many advance fiber reinforced composite materials. With the development of light weight composite structures, many structural challenges can be addressed. Among various advance composite materials, unidirectional fiber reinforced composites offer better control on properties [1]. With the use of unidirectional laminates, desired strength can be achieved by placement of laminates as be modeled design. Moreover degree of anisotropic behavior may also be altered using unidirectional laminates. Prepreg technology allows fabrication of intricate part with high quality. Prepregs may be used manually as in hand layup process of with automated placement machines. For manufacturing of parts for space applications consistent and high quality is required which can be achieved by using prepregs. The desired volume fraction of fiber for prepreg for space application is about 55-60% [2]. Choice of matrix and reinforcement depends on the application. Carbon fiber impregnated with epoxy matrix is used in many applications as this combination offers superior mechanical properties. Carbon composites with epoxy matrix have been used extensively in producing aerospace parts such as landing gear doors, flaps, aileron and other structural parts. Unidirectional carbon fiber- epoxy composites offer competitive fatigue properties in low earth orbits [3]. With superior properties, unidirectional carbon fiber composites may be used as face sheets for satellite payload fairing where high strength and high fatigue life is required. As the space environment offers vacuum thermal cycling, superior thermos-mechanical properties of unidirectional carbon fiber composites make them viable candidate for manufacturing various satellite parts This paper is distributed in two parts, manufacturing and testing. The former explains the in-house manufacturing of carbon fiber prepregs and the subsequent sheet forming, whereas the later describes the testing results of manufactured unidirectional carbon fiber sheets. II. EXPERIMENTAL A. Materials Material used for this study were oriented towards high specific strength. Carbon fiber tow (10K) was used as reinforcement for prepreg forming. Carbon fiber was chosen due to its high strength and low coefficient of thermal expansion. Choice of resin system depends upon the application hence for high strength application, epoxy based resin is desired. For this study, Swancor 2552 prepreg resin was chosen due to its low viscosity and good impregnation properties. 157 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 167. B. Manufacturing ofprepregs Manufacturing of prepregs was performed using in-house developed fiber impregnation machine. The scheme of manufacturing involves impregnation of dry fiber and then placement of fibers parallel to each other to form a continuous sheet. The manufacturing process of prepregs is somewhat related to the process of filament winding. Dry fiber was firstly impregnated with resin using a drum-roller mechanism same as for filament winding process. The schematics of impregnation of fiber is given in fig 1. Fig. 1: Schematic for Impregnation of Carbon Fiber Tow After impregnation of fiber with prepreg resin, the fibers is passed through pressing rollers which align the filaments in longitudinal direction and avoid any overlap of filaments within fiber layer. This process is desired because it flattens the fiber and aligns it so that perfectly straight and low thickness sheet of prepreg may be achieved. The last stage of prepreg setup consists of a large rotating mandrel, 2400 mm in diameter containing a release agent sprayed non-stick paper. The wet fiber was wound over the mandrel in such a manner that neither the two parallel fibers overlap each other nor there a gap in between them. This perfect alignment of fibers side by side is achieved by controlling the operating parameters with respect to band width of fiber. Subsequently after the desired width of wound has been achieved, the fibers were cut in circumferential direction and unwound so that flat sheet of prepreg is formed. Fig 2 shows a prepreg sheet. Fig. 2: Prepreg sheet The open surface of prepreg was covered with non-stick paper and prepreg was wrapped in roll form. Afterwards, the roll was sent to cold storage to store until use. The specification of prepregs and process parameters is given in table1. TABLE I. PROCESS PARAMETERS AND SPECIFICATIONS Fiber Carbon fiber 12K Matrix Epoxy-Swancor 2552 Width 355 mm Length 6000 mm No. of Spools 01 Roller Speed 36.9 mm/min Dry Fiber Weight 32 g Releasing agent Mold-Viz Spray C. Curing of prepreg sheets Prepregs were drawn from cold storage for manufacturing of composite structure. Prepregs were unwrapped and protective covering was removed from one side. Placement of prepregs on mold is a intricate job. As the Fibers are not linked with each other mechanically, it becomes difficult to place a laminate of unidirectional fiber on surface. Back non-stick paper layer acts as a supporting medium to bind the fibers together. Fiber were placed on mold manually with hand-layup and afterwards, paper layer was removed. The placement direction of fibers depend upon the design of structure. The sample for mechanical testing was prepared with unidirectional layer giving its maximum properties in one direction. Curing of unidirectional sheet was carried out in autoclave using air pressure and temperature up to 150 o C. The whole process was carried out in 2-3 hours resulting in rigid unidirectional carbon fiber reinforced epoxy sheet. Fig 3 shows a cured sample with unidirectional laminates. Fig. 3: Cured unidirectional carbon fiber sheet 158 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 168. D. Testing ofComposite samples Two type of tests were carried out in this study. Initially, prepregs were tested for calculating resin content. Four samples of prepregs before curing were tested to calculate average value. Mechanical testing was performed on cured sheet to obtain strength parameters. Tensile test according to standard ASTM- D 3039 [4] was performed on a universal testing machine. III. RESULTS AND DISCUSSION A. Resin Contents Resin contents were calculated in terms of total weight. Tests results indicates average value of 40% resin content. Also the presence of solvent was calculated as approximately 3% within the matrix system. B. Mechanical testing Tensile test was performed on cured unidirectional carbon samples. The average results are summarized in table 2. TABLE II. RESULTS OF MECHANICAL TESTING Breaking load 23 KN Ultimate Tensile Strength 1440 MPa Modulus of elasticity 97 GPa Elongation 1.4 % IV. CONCLUSION Unidirectional carbon fiber prepregs were manufactured using in-house designed prepreg machine. The prepreg machine was designed on the concept of filament winding. Dry fiber from spool was impregnated in a resin bath and then wrapped around a large mandrel. Semi cured prepreg sheets were removed from mandrel and stored in low temperature environment for preservation. The sheets were then withdrawn and cured in unidirectional laminates to form rigid sheet for subsequent testing. Test results of prepreg sheet and cured composite part shows satisfactory results. Prepregs contain 40% resin which is a nominal value. Moreover, cured sheets offered excellent mechanical properties. Hence composite structure with light weight and high strength was formed using unidirectional carbon fiber prepregs. REFERENCES [1] Purslow D., “The shear properties of unidirectional carbon fiber reinforced plastics and their experimental determination, Aeronautical research council current papers”, 1977 Retrieved from: naca.central.cranfield.ac.uk/reports/arc/cp/1381.pdf [2] Larberg Ylva, “Deformability of Unidirectional Prepreg Materials, Licentiate Thesis Stockholm”, Sweden. 2009 Retrieved from: www.diva-portal.org/smash/get/diva2:224958/ FULLTEXT01.pdf [3] A. Anvari, “Fatigue Life Prediction Of Unidirectional Carbon Fiber/Epoxy Composite In Earth Orbit”, Int. J. of Appl. Math and Mech. 10 (5): 58-85, 2014 Retrieved from: ijamm.bc.cityu.edu.hk/ijamm/outbox/Y2014V10N5P58C25132 388.pdf [4] ASTM D 3039, “Standard Test Method for Tensile Properties of Polymer Matrix Composite Materials”, 159 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 169. Electrical Power ConditioningUnit Design for Space Qualified C-Band Receiver in GEO Satellite Applications Taimoor Zahid Institute of Space Technology, (I.S.T) Islamabad, Pakistan Taimoor187@yahoo.com Muhammad Azeem Aslam Beihang University, (BUAA) Beijing, China Ms.azeem@gmail.com Abstract— The competition in commercial satellite market requires the Electrical Power Subsystem (E.P.S) components to be more and more adaptive to the payload specific characteristics. The Electrical Power Conditioning Unit (EPC) is an important part of GEO Satellites, especially in payload receivers. Because of input/output voltage, reliability, anti- radiation, in-rush current protection, TM/TC, EMC, bus voltage ripples and many more requirements which makes its design complex and lengthy for an RF point of view. Apart from these design driven factors the design should also be cost effective and mechanically stable. The C-band microwave frequencies have excellent performance in most physical aspects like rain fade, atmospheric attenuation at low elevation angles and cheaper bandwidth. This aim of this paper is to describe the functions and design of Electronic Power Conditioning (EPC) module for a C- Band Receiver for a given set of space qualified requirements. To fulfil the above said requirements a new generation of EPC design is presented by implementing half bridge topology and more efficient rectification and regulation techniques. The output voltage ripples have been minimized to 25mV and input filter is designed by using Differential Mode (DM) and Common Mode (CM) filtering techniques to overcome EMC/EMI interference. Under voltage and over current protections are also implemented to protect satellite in case of short circuits or lower main bus voltage. De-rating analysis is performed by considering the ECSS‐Q‐ST‐30‐11C space standard. It also describes the compliance with performance requirements for RF, electrical and mechanical interface during the EPC module design for C- Band Receiver. The design presented in this paper complies with performance requirements for RF, electrical and mechanical interface. Considering all the above mentioned factors the design, simulation results and analyses are presented in paper. Index Terms— EPC, GEO, EPS, and EMC/EMI. I. INTRODUCTION The Electronic/Electrical Power Conditioner (EPC) is actually a multiple output DC-DC converter which provides power to a TWTA or an SSPA from the satellite power bus. The power provided to both amplifiers is normally from more than one source and is well regulated. The power requirement depends on the efficiency of the amplifier and transmitter power requirement. This aim of this paper is to describe the functions and design of EPC module for C- Band Receiver. It also describes the compliance with performance requirements for RF, electrical and mechanical interface during the EPC module design for C-Band Receiver. The main functions of EPC unit are: Conversion of DC bus Voltage to a different high stability output Voltage. Controlling the sequence between positive voltage and negative voltage. Low output voltage-ripples. Perform TC ON/OFF, Telemetry etc. Perform over-current protection, under-voltage protection. Fulfil EMC requirements. Fulfil high reliability requirement. For the power bus voltage there is an operational and a non- operational requirement. The operational voltage ranges are given below: Power bus voltage range of the normal work mode is 90～110 (Nominal value is 100V). The safety power bus voltage range：85～90V and 110～115V In the normal voltage range of the power bus, the equipment meets all performance requirements. In the safety voltage range of power bus (85～90V, 110～115V), basic equipment functions keep working. The non-operational power bus DC voltage range is 0V-115V. In the non-operational DC voltage range, the equipment should not be damaged. Variation slope is less than 10V/msec. The output current and voltage requirements are given in the table below: V1 V2 V3 Remarks Voltage or Voltage Range (V) -5 +12（after CW7812） +6 Voltage Tolerance △V (V) 0.2 0.2 0.2 Ripple（P-P）(mV) 50 50 50 Current or current range（mA） SS in vacuum 5-15 580-810 165- 230 In this range, first value is nominal, SS in ambient 5-15 750-1050 165- 230 160 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 170. Warm up (10 min) 5-15 1000-1400 165- 230 second valueis by adding 40% margin. Requiring voltage regulator block CW79L05 CW7812 Whether the negative voltage is first out when switch ON and negative voltage last when switch OFF Yes Table.1 Output Current and Voltage Requirements The equipment will be commanded ON/OFF through pulse signal. Command signal is a positive voltage pulse. It is necessary to take anti-interference measures for the input to perform remote control command. The TC circuits will not operate when supplied with a pulse of 100µs duration, up to 5V amplitude and a repetition rate of one per 13 msec. The TC circuit cannot operate correctly when subjected to the nominal pulse with one break period of 1µsec which occur at any time during the TC pulse period. The signal waveform and timing are described in the figure below. Figure.1 Telecomand signal waveform profile The values of tr and tf be equal to or lower than 50µs. The telecommand requirements of the C-band receiver are given in the table below: Parameters Requirements Pulse Voltage-True State 10V ~ 12V Pulse Voltage-False State 0 ~ 0.2V Pulse Duration 100±2ms Maximum current < 1.2mA Input Impedance ≥10KΩ Conducted Susceptibility 12V/100 µs pulse width/10Hz Conducted Emission Appropriate suppress switching transient due to inductive load which may cause the driving source capabilities to be exceeded Table. 2 Telecommand Requirements The Digital Telemetry (BL) features are: 1) Output Voltage High state corresponding to logic "1" level: 10.0V~12.0V Low state corresponds to logic "0" level：0V~0.2V Exceptions conditions do not exceed 0~+15V。 2) Output impedance：≤5kΩ 3) Output short circuit protection The short circuit of output and ground do not cause any permanent damage and also do not affect the performance of other parts of equipment. With the interconnecting harness(es) disconnected, the equipment should have a minimum DC resistance of 1MΩ between the power input leads and the equipment case. It should also have secondary power supply circuits connected to the equipment case. These circuits should be isolated from the primary power circuits by a minimum DC resistance of 1MΩ. The equipment should have the command return isolated from the power return and also from secondary ground. The equipment must continue to operate normally without degradation with either a short or open circuit on one or all of telemetry output lines and must not be damaged by either short or open circuit on the command or power lines. To avoid failure, a fuse must be used to protect the main bus. The over current protection circuit (fuse) must start/be used when bus current goes up to 2A. Current consumption of the equipment should be limited to twice the maximum steady state current within 5ms. Receiver must incorporate under-voltage protection circuitry that will automatically turn off the receiver when the primary power input voltage falls below 85V. The response time does not exceed 20ms.When bus voltage reaches the nominal value, the equipment can be turned on by telecommand only. The design of the equipment is done while taking into account the consideration of the ground storage for two years. Failure Rate < 400 Fits @500 C for the lifetime being 15 years. The temperature should be as follows: Operating temperature range：-50 C～500 C Acceptance temperature range：-100 C～550 C Start-up Temperature: -150 C～+650 C. Equipment boot inrush current requirements for design and test (without evaluation) are shown in the figure below. Idc is a maximum constant current under normal conditions. 161 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 171. Figure.2 In-rush currentLimits. II. MODULE DESIGN DESCRIPTION EPC unit is subdivided into four major parts which are as under: Main DC-DC converter unit Auxiliary DC-DC converter TC Circuit TM Circuit The block diagram of the EPC module is shown in figure below: Figure.3 Block Diagram of EPC. A. Main DC-DC Unit The main function of this unit is to produce the output voltages (-5V, 12V +6V) for C-band RF and the local oscillator chain at different output power levels and 12V for under voltage / over current protection (voltage used in different modules on EPC PCB board). It is implemented using half bridge topology using rectifier, output filter and voltage regulator modules. The output current and voltage requirements mentioned above have been fulfilled by this unit. The result for different output voltage levels and output ripple (50mV) has been simulated and can be seen from figures below. These simulation results are before linear regulator (LR), because it is difficult to model LR in P-Spice. LR is used on 12V and -5V secondary lines. After LR the secondary voltages will be at nominal values (12V,-5V and 6V). The output ripples can be seen in figures below. Figure 4 Secondary 6V Ripples Level (20mV max.) Figure 5 Secondary 15V Ripples (35mV max.) B. Auxiliary DC-DC Converter The main function of the auxiliary DC-DC converter is to ensure the output voltage to be -5V. It has two outputs one is 15V and the other is -15V, -15V is for the output line of -5V to make sure that -5V appears first and 15V is for the main DC-DC unit to start it. The operating time is not more than 100ms because the TC ON pulse time is 100ms.The output results are shown as under. 162 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 172. Figure 6. AuxiliaryConverter Output (15V and -15V) C. TC ON/OFF Circuit This unit is used for switching on and off the EPC and all output voltages only when Telecommand Unit (TCU) sends command to EPC. This circuit generates 15V signal to auxiliary converter PWM IC to start EPC and generates 5V pulse for main DC-DC converter PWM to shut down EPC. According to the requirement it can only be operated when telecommand of 100ms is at the input of this unit. The TC circuit does not operate when supplied with a pulse of 100µs duration, up to 5V amplitude and a repetition rate of one per 13 msec. The TC circuit operates correctly when subjected to the nominal pulse with one break period of 1µsec which occurs at any time during the TC pulse period. The simulation results can be seen from figures below. Figure.7 TC ON Circuit Output Figure.8 TC OFF Circuit Output TC OFF signal in figure above is at the input of main PWM IC to turn it off. But when there is no TC OFF signal the voltage across main PWM IC increases slowly. In real-time circuit, the voltage across PWM cannot rise till TC ON signal comes from the ground station. Hence The EPC shall be at off state. When 12V TC ON signal appears at the input of Auxiliary DC-DC converter it generates signal to main PWM to start. D. TM Circuit This circuit provides a constant telemetry from 10 to 12V when receiver is working and 0V when receiver is off. The TM signal will be received for EPC when EPC is working. When EPC is not working there will be no TM signal which can be seen from the figure below. Figure 9 Output Telemetry Signal III. CONCLUSION In this paper a design of the Electrical Power Conditioning Unit (EPC) which is an important part of GEO Satellites, especially in communication satellite payload receivers using half bridge topology and more efficient rectification and regulation techniques is presented. The aim of this paper is to 163 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 173. describe the designand analyze results of EPC module for new generation communication satellite payload receivers for a given set of specific space qualified requirements. The output voltage ripples have been minimized to 25mV is presented. The input/output voltages, in-rush current protection, TM and TC characteristics, bus voltage ripples and other requirements which make its design complex and significant for RF point of view. Both the output current and voltage requirements are presented and met successfully. The design presented is also compliant with the TC requirements. The under voltage and over current protections are also implemented to protect satellite in case of short circuits or lower main bus voltage. All the four major modules of EPC (main DC-DC converter unit, auxiliary DC-DC converter, TC circuit and TM circuit) are discussed and analyzed along with their simulated results. REFERENCES [1] Ioana-Monica, Pop-Calimanu, Prutianu Florin, and Popescu Viorel. "Design and simulation of DC/DC boost converter used for distributed sensing system based on a multidrop sensor network with RS485 interface", 2012 10th International Symposium on Electronics and Telecommunications, 2012. [2] Pequet, E.; Delporte, P.; Fayt, P.; Gak, M.; Canon, T., "ESA qualified EPC for telecommunication satellites TWTA," Vacuum Electronics Conference, 2000. Abstracts. International , vol., no., pp.2 pp.,, 2-4 May 2000. [3] R. E. Sorace, V. S. Reinhardt, and S. A. Vaughn, “High-speed digital-to- RF converter,” U.S. Patent 5 668 842, Sept. 16, 1997. [4] Castiaux, J.P.; Bury, P.; Liegeois, B., "Power conditioning units for high power geostationary satellites," Power Electronics Specialists Conference, 1997. PESC '97 Record., 28th Annual IEEE, vol.1, no., pp.722,733 vol.1, 22-27 Jun 1997. [5] http://www.space-airbusds.com [6] https://www.thalesgroup.com [7] http://www.sstl.co.uk [8] Garrigos, A.; Carrasco, J.A.; Blanes, J.M.; Sanchis-Kilders, E., "A power conditioning unit for high power GEO satellites based on the sequential switching shunt series regulator," Electrotechnical Conference, 2006. MELECON 2006. IEEE Mediterranean , vol., no., pp.1186,1189, 16-19 May 2006. 164 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 174. DSP Based Electro-HydraulicActuator Control with Irretraceable Feedback Error Compensation Najam-Ud-Din Ahmed, Syed Aleem Azhar, Khadim Hussain Abstract—This paper describes the interfacing detail of Electro-Hydraulic actuator with DSP and addresses the issues which commonly occur during development. The purpose of designing digital controller on DSP is it’s easy configurability via software and requirement of less analog part. Scenarios where a dc-dc converter used as voltage source may induce erroneous feedback in the feedback control system due to its dual output mismatch. And despite a good PID controller real position of manipulator may have errors which are irretraceable even by the intelligent controller schemes. A judicious and intelligent use of available DSP peripherals like ADCs, DACs etc and minor adjustment of circuit can effectively increase the real time accuracy of whole system against irretraceable errors. Keywords—DSP based PID, PID Controller, Irretraceable error, Erroneous feedback compensation I. INTRODUCTION It is common practice to use DSP or any digital controller to control analog manipulators and actuators in mechatronic applications [3]. However, noise and supply voltage accuracy dependability of analog systems is always associated with them and affect the accuracy of the overall system. There are many advanced as well as conventional control schemes to increase the performance of the actuator [2] but these schemes might be less affective if the feedback is erroneous or we have high tolerances in the supply voltages. This paper discusses such a scenario where a dc-dc converter used as voltage source may induce errors in the feedback system due to its dual output mismatch. And despite a good controller real position of manipulator may have errors which are irretraceable by the intelligent controller schemes. In addition, use of digital controller specially a DSP helps in generation of real time rate for interfacing. Although various techniques are available for getting precise band-gap reference voltage [1] but DSP based controller also adds noise immunity which is always associated with analog systems. II. PROBLEM STATEMENT Two hydraulic pistons are used in a robotic arm for high speed application prototype. Feedback is based on linear potentiometer for which dual supply is required. It is known that 1% error in supply voltage will produce same error in the feedback which will be not traceable in potentiometer based feedbacks and still be added in the real position of actuator. Output Voltage (Vout) at potentiometer viper is, (1) (2) (3) (4) (5) ( ) (6) This error may be induced due to two factors 1. Due to tolerance in supply 2. Due to mismatch in dual supply There is always a pinch of tolerance in supply and it is always expensive to use high grade supply unit or 165 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 175. reference generation ICs.Thermal drifts associated with the electronics circuits make this tolerance/error always changing. Moreover, it is common observation that plus side load is always greater than negative side load of the supply or any DC-DC converters which are dependent upon percentage load for its good regulation. Hence, there is always supply mismatch between its dual outputs. This condition becomes more severe when there is a gain in the feedback loop. A current loop is added in the system to compensate the noise affect associate with the analog electro-hydraulic valve. An op-amp based current loop will serve the purpose which was connected with DSP DAC output. Nonlinearity may also occur in feedback if the feedback is just beyond the measurable range of the ADC. For example, an ADC of ±2.5 volt is being used and feedback voltage of 2.6 volt will appear 2.5volt to the controller algorithm and position corresponding to 0.1Volt increase will never be compensated. This situation may occur due to saturation of ADC. III. HARDWARE DESCRIPTION AND DSP ALGORITHM DSP D-Module 21065L D-sign T is used for this application along with S-module. In order to control two actuators four ADC input and two DAC output are used. Two ADCs are required for taking feedback position of both actuators, two DAC output are required to drive each actuator to the desired calculated position. The ADC and DAC we used have the 16-bit resolution and have different full-scale voltage ranges of +/-2.5V, +/-5V and +/-10V. In order to utilize full resolution of ADC the reference voltage range provided to the actuator’s potentiometer position sensor and the ADC full scale voltage range must be matched. We select the ADC range and correspondingly reference voltage range to +/-10V because the external noise produces little impact on this large value than using +/-5V or +/-2.5V range of ADC. Two more ADCs are required for measuring reference voltage supplied to the actuator’s potentiometer for position sensing. A. Error compensations and implementation The ADC available on S-module has 16-bit resolution and has a bi-polar input of a range of +/- 10V. Actuator gives position feedback in the form of varying resistance. To utilize the full resolution of ADC, we have to supply +/-10V to the potentiometer. Non-compensated feedback position = ADC input (1 or 2) Position Error (to be used in PID eq.) = Setpoint - Non-compensated feedback position This relation is valid where there is no mis-match in ±10volt references. And tolerance is zero or very close to zero. This is not possible in real scenario. Due to possible supply tolerances outside the range; ADC will not be capable to measure the voltage beyond +/-10V due to its saturation. Therefore, +/-9.5V has been provided for reference voltage instead of +/-10V to the potentiometer. This will decrease the resolution of ADC by some amount but provides the protection for the possible overflow of reference voltage outside the ADC range. Moreover, it will indirectly provide fault tolerance against open/short connection of ADCs and feedback potentiometers. Now there are two errors which need compensation. Full scale reference input error due to ±9.5 volt instead of ±10 volt (instead of 20 it is 19 volt) Reference mismatch error due to tolerances in +9.5 and -9.5 voltage references (creating offset) To compensate these errors we need to exactly find the current reference voltages and then apply the compensation as follows, ( ) ( ) ( 7) ( ) ( ( ) ( ) ) Now position error used in PID loop is Err calculated as, And desired output as, Desired_Actuator_position= ( ) ( ) ( ) ( 8) Where, kp = proportional-gain, ki = integral-gain, kd = differential- gain. The tuning of digitally implemented PID has made it much easier as compare to using of analog PID. 166 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 176. B. Saturation limit Asaturation limit is a safety measure must always be applied in the algorithm to avoid overflow conditions and generation of undesired & unknown erroneous DAC output. For 16 bit DAC we have saturation applied as, If (Desired_Actuator_Position > 32767) Desired_Actuator_Position = 32767 ElseIf (Desired_Actuator_Position < -32767) Desired_Actuator_Position = -32767 Else Desired_Actuator_Position= Desired_Actuator_Position Finally, the above calculated value is sent to the corresponding DAC-Channel. The scanning frequency of the above calculation is done at the rate of 10 KHz. It means that ADC sampling, the calculation for the desired output in PID LOOP and updating on DAC output are all done at 10 KHz frequency. To ensure the constant sampling rate, we do all the calculation at the 10 KHz Timer- Interrupt. C. Real time clock rate and RS422 Communication In the 10 KHz loop, we also generate different patterns of set-point on time-base. These set-points contain sine-wave, square-wave, triangular-wave and stair-case waveform. To evaluate the performance and post-analysis of the above control we sent real-time values of set-point and feedback along with the absolute time in seconds in the 2msec interval to the Rs-422 port of PC at the baud-rate of 115200.These values are arranged in the packet form and have the total length of 16-bytes. The above packet is transmitted by the DSP and received by the application program built on the Labview. As in the table shown above, the starting two bytes used to synchronize the incoming packet and the last byte contain the checksum of the data bytes only excluding the synchronization byte. The incoming packet considered to be valid if calculated checksum and the received checksum contained in the last byte of the packet matches. These values are displayed in real-time in the application and also logging the values in the text file for post-analysis. IV. DSP PROGRAM FLOWCHART Fig.1. Main program Loop START Initialize Dsp and Peripherals If (Time_Delay > 2msec) False True Prepare packet to transfer serially If (Transmit-Buffer = Empty) False Send Data-Packet serially If (60-sec Profile is not running & Character is received at serial-port ) True Receive Command True If (Command is for Sine-wave, Square- wave or for profile & Command is new) Save absolute-time into offset-time and new- command into previous-command Reset Timer_Count to Zero. END True False True False Main Program Loop 167 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 177. Fig.2. 10kHz InterruptLoop Fig.3. PID Loop PID Function START Calculate offset generated by voltage-level converter card by using the formula given below. (+Vref)+(-Vref) / 2 Remove offset-error generated from both actuators feedback positions. Calculate Error for both actuators, Error = Set-point – Corrected Feedback Apply Gain-Control on both actuators, Desired-Position = (kp * Err) + (ki * ) + (kd * ) Apply Saturation limit to Actuator 1 desired-position to prevent overflow from 16-bit integer value. (Because the DAC has 16-bit output) Apply Saturation limit to Actuator 2 desired-position to prevent overflow from 16-bit integer value. (Because the DAC has 16-bit output) Output desired calculated position to respective DAC channels. END 10khz Interrupt Loop START Calculate absolute-time and relative-time Read Feedback position of Actuators and Reference supply voltages using A/D converter. If (Command ==1) Generate Set-Point(Profile) of both actuators contain various waveform of duration of 60-seconds. True If (Command ==2) Generate Set-Point of both actuators to value of 25mm. True If (Command ==3) Generate Set-Point of both actuators to value of -25mm. True If (Command ==4) Generate Set-Point of both actuators to value of square-wave of 2Hz. True If (Command ==5) Generate Set-Point of both actuators to value of sine-wave of 2Hz. True If (Command ==6) Generate Set-Point of both actuators to value of sine-wave of 8Hz. True If (Command ==7) Generate Set-Point of both actuators to zero millimiter(“0” mm). True Call PID( ) function that perform on both actuators position and also output desired actuators position on DAC. END 168 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 178. V. TEST PROFILE Acomplete test profile is a good practice to test the response of actuator. This profile depicts and incorporates all or nearly all type of responses which an actuator/ manipulator may experience during its course of operation. A test profile for such purpose is generated and preprogrammed in the DSP as shown in fig 4. Fig.4. Test Profile This profile contains 1. 2Hz 25mm sine & square set point 2. 25mm Triangular set point 3. Staircase Set point 4. High frequency low amplitude set point VI. RESULTS Fig.5. Low Frequency high Amplitude Sine Wave Fig.6. Square Wave Response Fig.7. Step Input Response Fig.8. Higher Frequency Low Amplitude Sine Results clearly shows good dynamic response of electro hydraulic system VII. CONCLUSION In this paper, we have utilized available ADCs and high speed performance of DSP for compensation of feedback errors. Use of extra available ADCs will decrease the additional cost of high grade circuitry for generation of highly accurate voltage references. A simple resistive based feedback sensor will further decrease the cost and interfacing effort in analog/digital position feedback setups. Use of already available ADCs will not only simplify the 169 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 179. circuit but alsothe errors associated with use of additional components in any circuit. REFERENCES: [1] LinHaiCui. Design of a High Precision Band-gap Voltage Reference. Proceedings of the 2011 International Conference on Electronic & Mechanical Engineering and Information Technology. IEEE, 2011 : 2187~2190. [2] Sergey Edward Lyshevski and Trevor C Smith. Tracking Control of Direct-Drive Servos. Proceeding of the 2011 International Conference on Decision and Control and European Control Conference(CDC- ECC).Orlando, FL, USA, IEEE, 2011 : 1602~1607. [3]Hai-tao Wang, Ze-Zhang and Xiang-yu Liu. Design of Control System for Brushless DC Motor based on TMS320F28335.Proceeding of the 2011 Third International Conference on Measuring Technology and Mechatronics Automation. IEEE, 2011 : 954~958. [4]D-module D-sign T manual. [5]VisualDSP++ 4.0 - C/C++ Compiler and Library Manual for SHARC Processors. [6]ADSP-21065L SHARC DSP - User's Manual and Technical Reference 170 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 180. Design for TestApproach using FPGA for BPSK Modem Syed Jahanzeb Hussain Pirzada1 , Muhammad Faisal Arain1 , Rahman Mehboob1 1 Space and Upper Atmosphere Research Commission (SUPARCO), Pakistan. Abstract— Implementation of modem has always been the key interest for researchers working in the field of communication. Nowadays, at low frequencies (in kilohertz range) the modulation and demodulation is performed digitally. As, modem is utilized for telemetry and telecommand communication between satellite and ground station. This work provides the design of test bed for modem. The testing is performed using ECSS based Standards. This paper explains the design of test bed for testability of Binary Phase Shift Keying modem implemented on Field Programmable Gate Arrays. The testing of FPGA based modulator was performed using signal analyzer. On the other hand the testing of FPGA based demodulator was complicated and needs to be design on seperate hardware; as it contains carrier recovery, bit syncronization and bit recovery. Hence, digital signal processor and FPGA based co-processing test bed was design and developed for tesing of demodulator. The experimental results show the architecture and testing results of the proposed approach. The proposed methodology describes the problems encountered during testing and proposed solutions to it. Such as in demodulation, the bit recovered from demodulation can be analyzed on oscilloscope but it is cumbersome. Therefore, an asynchronous bit recovery algorithm is employed with serial communication for representation and verification. Keywords- DSP, BPSK, modulator , demodulator. I. INTRODUCTION Digital modulation is utilized to convert digital symbols into waveforms to make it compatible with the characteristics of transmission channel [3]. In digital modulation, a baseband signal is changed into band pass signal compatible for transmission. Similarly, demodulation is utilized in receiving end to recognize the detected data. Binary Phase Shift Keying (BPSK) is employed in transmission and reception of telemetry and telecommand data in satellites. It is the simplest form of phase shift keying (PSK). Specially, at low frequencies (in kilohertz range) the modulation and demodulation is performed digitally. Generally it consists of two phases one is inphase with the transmitted carrier and other is out of phase with the transmitted carrier as represented in Figure 1. The two conditions are generally represented as logic 0 and logic 1, the formula associated with it represented in eq-1 and eq-2 [4]. Figure 1: Phase change of BPSK signal. A Test bed is essential for rigorous testing of the system. Implementation of digital modulation schemes have become more and more attractive and simple with the used of digital signal processors (DSP) and Field Programmable Gate Arrays (FPGA). Recently, researcher are utilizing the processing power of FPGA and DSP together to implement systems. Testing of BPSK modem is very important as all the data transmitted through the transmission channel depends on the correctness of implmentation of baseband modualtion. The remaining paper is organized as follow. Section II provides the previous work. Section III explains the proposed methodology. Section IV explains experimental results. Section V provides conclution of this paper. II. PREVIOUS WORK Many researchers have designed and developed BPSK modulator and demodulator in the Past [2] [5] [7] [8]. The testing with genralized equipments are usually quoted in many implementation. Including some results are acquired in the form of software simulations. In this paper a testbed based testing mechanism is proposed. Although it is formed using traditional testing equipment but incombination with custom solution. III. PROPOSED METHODOLOGY Testing of BPSK modem is proposed in this paper. The test setup is composed of two sets of modems. One is the designed BPSK modem to be utilized in telemetry and telecommand module and other is for testing of designed BPSK modem. For testing of BPSK modulator and demodulator designed on FPGA to be utilized for telemetry and telecommand. A testing 171 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 181. scheme is devlopedwhich consists of a BPSK modulator with variable frequency and data rate. This modulator is designed on Digital signal processor. Similarly an FPGA is utilized for debuging of the BPSK demodulator under test. On the other hand, for testing of FPGA based Modulator Network analyzer is utilized. The block diagram of the proposed setup is shown in Figure 2. FPGA Based BPSK Modulator FPGA Based BPSK Demodulator Network Analyzer Based BPSK Demodulator DSP and FPGA Based BPSK Modulator and Debugging Designed FPGA TestBed Figure 2: Block Diagram for BPSK Test Bed. IV. EXPERIMENTAL RESULTS Testing of BPSK modulator is significantly important for reliable operation of satellite. Testing of telemetry modulator data is acquired through on-board computer on FPGA according to ECSS standard, the data is then converted to non- return to zero (NRZ) format and modulated. The modulated signal is then tested using signal analyzer. A signal analyzer has built in BPSK demodulator. The demodulator demodulates the data and show constellation diagram. Let us first illustrate the ideal result for BPSK constellation diagram in Figure 3. It represents the possible symbols that may be selected by a given modulation scheme as points in the complex plane. Measured constellation diagrams can be used to recognize the type of interference and distortion in a signal. For analyzing the spectrum and constellation diagram of BPSK modulated signal, Rhodes and Schwartz™ signal analyzer was used. Signal analyzer contains a BPSK demodulator, which can demodulate the signal accurately on wide range of frequencies and construct constellation diagram. Figure 4 shows the constellation diagram obtained from testing our BPSK modulator with signal analyzer. That shows constellation points position is very similar with the ideal results. Figure 3: Constellation Diagram of BPSK modulator Figure 4: Signal analyser results for constellation diagram Table 1 shows that the magnitude error and phase error are in allowable number for satellite communication. Owing to which the signal to noise ratio (SNR) is in the acceptable range for space communication. BPSK modulated signal spectrum as analyzed by signal analyzer is shown in figure 5. Spectrum shows that the difference of frequency of two peaks close to the center is 2 kHz which authenticates the data rate of 2kbps .As the symbol rate is equal to the bit rate. As shown in the equation eq.3 and eq.4. R(s) = R (b) / n (3) For BPSK, (n = 1) R(s) = R (b) (4) where R(s) is the symbol rate, R (b) is the bit rate and n represent the number of phase changes. In case of BPSK modulator the number of phase changes is 1. So symbol rate is equal to bit rate. 172 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 182. TABLE I PERFORMANCE CHARACTERISTICSOF BPSK MODULATOR S.No Modulation Accuracy Category Results Peak Unit 1 EVM 0.662 1.665 % 2 Magnitude Error 0.561 1.661 % 3 Phase Error 0.20 0.82 % 4 Amplitude Drop 0.01 - dB 5 Origin Offset -73.86 - dB 6 Gain Imbalance 0.00 - dB 7 Quadrature Error 0.00 - deg 8 SNR (MER) 43.58 - dB 9 RHO 0.999 - - Figure 5: Spectrum representation of BPSK modulator on signal analyzer. Testing of demodulator is performed with DSP based BPSK modulator and FPGA based debugging equipment. Result of simulation on Code Composer Studios (Software used for programming TI DSP Kits) are shown in Figure 6, Upper half of the figure shows input converted to NRZ signal and in the lower half corresponds to modulated BPSK signal [1]. The BPSK modulator implemented on Texas Instruments TMS320C6713 DSP kit has variable frequency and data rate configurations. Althought in this testing application data rate and frequency is fixed. The data from BPSK modulator is provided through test computer serially and it modulates the data. The modulated data is then input to FPGA based BPSK demodulator. After the demodulation the data is transfered to FPGA based Debugger for viewing the retreived output. V. PROBLEMS AND SOLUTIONS DURING TESTING Several problems are encountered during the testing of BPSK modem implemented on FPGA. Firstly, The bit recovered after demodulation are continous, in long data stream it will be difficult to monitor it in osciloscope. Therefore, a uart based debug ging is implemented on FPGA for validation of demodulator performance. Secondly, Analog to digital converter has been tested seperately but on integration with FPGA and DSP based modualator the rescaling was required. Lastly, there was a transmission rate missmatch between ADC and uart transmission. So the data from ADC is stored in memory and then transmitted through uart transmission. Figure 6: CCS Graph results of NRZ converted data and BPSK modulated signal. VI. CONCLUSION Our experiment showed testbed setup for BPSK modem. Test setup for testing of BPSK modem using signal analyzer and DSP/ FPGA test setup was proposed. The problems encountered are discussed. ACKNOWLEDGMENT This work was supported by Space and Upper Atmosphere Research Commission (SUPARCO). Authors also acknowledge the efforts of Mr. Muhammad Faisal and Mr. Razi Iqbal for continuous support and efforts for completion of this research work. REFERENCES [1] G.Susinderrajan, S.Saran Kumar, T.Shankar, K.Sreeram, P.V.Ramakrishna, "Design of a Low Cost All-Digital PSK/PM Satellite Telemetry Transceiver", IEEE Electron Device Letters, vol. 20, 569– 571, Nov. 1999. [2] S. Ruque, I. Ruiz, E. Carrión, "Simulation and implementation of the BPSK modulation on a FPGA Xilinx Spartan 3 xcs200-4ftp256, using Simulink and the System Generator blockset for DSP/FPGA", IEEE Electron Device Letters, vol. 20,569–571, Nov. 1999. [3] B.Sklar, “DigitalCommunications Fundamentals and Applications”, 2nd ed. Englewood Cliffs, NJ: Prentice-Hall PTR, 2001. [4] S. Johanna, Ruque, D.Ruiz, Carlos, “Simulation and implementation of BPSK system on a FPGA board using system generator blockset for DSP/FPGA”, School of Electronics and Communication, Technical university Loja 2005. [5] C.J. Harsha, D.H. Sandeep, A.S. Mali,"Hardware Implementation of BPSK system on Virtex2-Pro FPGA using Xilinx System Generatror".International Refereed Journal of Engineering and Science (IRJES), vol. 2, pp 18-24, January 2013. 173 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 183. [6] J.G. Proakis,Digital Communication, 5th edition, McGraw Hill New York, 2001. [7] H. Malik, D.R. Rotake, M. Mahajan,"Design and Implementation of BPSK Modulator and Demodulator using Vhdl". IOSR Journal of Electronics and communication Engineering (IOSR-JECE), Vol. 9, pp 98-105, May-Jun 2014. [8] N. P. Shrirao, A.P. Thakare,"Design of Digital Modulators:BASK, BPSK and BFSK using VHDL". International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE), Vol.3 january 2013. 174 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 184. Design of aFuzzy Logic Water Level Controller Taimoor Zahid Institute of Space Technology, Islamabad, Pakistan taimoor187@yahoo.com Nosheen Zafar COMSATS Institute of Information Technology, Islamabad, Pakistan nosh_zaf@hotmail.com Abstract— This paper analyzes the effectiveness of water level control using the concepts of Fuzzy Logic. We have designed a Fuzzy Logic Controller for Water Tank Level Control by using the linear control valve, the level transmitter and the fuzzy control rules built in the Fuzzy Logic Tool-Box and Simulink in MATLAB and have then compared the control effect with that of a Proportional–integral–derivative (PID). This design can be very valuable for the industries where water level control is extensively used and a slight deviation can lead to major accidents and huge losses in revenue. Therefore, it has become necessary to develop an accurate and cost effective control system. The purpose of such controller is to sustain a set point at a particular value and be able to dynamically accept the new values. Sometimes steady-state error and overshoot can no longer be taken into account by the PID Controller due to increase in the complexity, hysteresis, nonlinearity and coupling of control object. While on the other hand, with the development of the Intelligent Control Theory, the application of Fuzzy Controller has become very popular in practical industrial automation applications and these problems can be successfully dealt with using Fuzzy Control. Keywords— Water Level Control, Fuzzy Logic, PID, Tank, Mamdani. I. INTRODUCTION During the past several years, Fuzzy Control has emerged as one of the most active and fruitful areas of research in the application of Fuzzy Set Theory [1]. Due to its robust, precise and orderly approach for operating the controller, Fuzzy Logic can be used in several applications. The field of Control System Engineering has benefited the most from the success of Fuzzy Set Theory. The reason behind this unexpected success is high performance rate and low cost. Although PID-Controllers are easy to install but their performance can deteriorate quite fast when used in non-linear systems [2]. Moreover, the selection of the parameters of PID is also a difficult task. If one does not choose the fine parameters, it will badly impact the working of controller. Fuzzy Control gives more attention to various parameters, such as the time of response, the error of steadying and overshoots [3]. This paper shows the development and implementation of Fuzzy Logic Controller for Flow Control Application because Fuzzy Controller can achieve greater control results where limit of overshoot is strict [4]. We have taken the level of the water tank as an object and designed a Fuzzy Controller in MATLAB. The Simulink models are generated to see the working of tank. The controllers are then added in those models to check and understand the working of the designed controllers. The control effect is examined and compared with the effect of PID Controller. After designing the PID and Fuzzy Controllers, the results are compared. This approach of simulation by MATLAB allows us to solve more complicated problems and develop the measuring states to improve modeling and analysis of system and to allow more control in the designs [5]. The overall goal of this paper is to show which controller is better for flow and control applications. II. THE TANK PROCESS Input, Process, Feedback and Output are the four steps of tank process. 1) Input: A square-shaped waveform of amplitude 0.5 and frequency 0.1 Hz is generated from the signal generator as given by (1). y(t) = Amp * waveform(freq, t) (1) 2) Process: The Tank-Process includes valve, tank and controller. The water is pumped into the tank via valve with controllable degree of opening. Water flows out of the tank through a hole in the bottom with area 0.05 m2 . The height of the tank is 2m, and the cross-sectional area of the tank is 1m2 . 175 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 185. Technically, the waterlevel in the tank can be described mathematically using (2). = √ – (2) Where h is the water level [m], is the influent flow rate [m3 /h] and and are parameters depending on the tank and outflow areas. The influent flow rate depends on the degree of valve opening. In simulations, the water level in the tank is controlled towards some pre-specified value (reference signal) by letting the controller manipulate the degree of valve opening. Eq. (2) will be used in designing the Simulink model for controlling the water level for the designed tank. Figure.1 Water Tank System 3) Feedback: The signal is generated from the signal generator and is further processed through controller. This output from the process becomes a feedback. This feedback goes to a summer named error. It is named so because it calculates the difference between actual and desired water level. The actual value is the one which is coming through feedback and the desired is the one which is coming straight through the signal generator. 4) Output: The output can be generated via: (a) To workspace: It writes the input to the specified array or structure in MATLAB’s main workspace. The “To Workspace” block takes the signal as input and writes the data of the signal into the workspace of MATLAB. Data is not available until the simulation is paused or stopped. This block writes the data to an internal buffer during the simulation and that data is written to the workspace when a simulation is completed or paused. (b) Comparison: This block generates a plot and that plot gives the comparison between the reference signal and system output. (c) S-function: This block provides a direct access to the additional parameters to be passed to S-function. The functions parameters can be stated as variables separated by commas or as MATLAB expression. This function displays the number of I/O ports plus the names of specific S-functions. In our case, the built-in file named “tankdemo” is used for generating the animation. The S-function used in this application refers to the “animtank” which is built-in in the MATLAB directory. ANIMTANK stands for “Animation of water tank system”. III. PID-CONTROL OF TANK WATER LEVEL A. PID Controller PID controller refers to a Proportional–integral– derivative controller. In order to overcome the shortcoming of self-tuning PID controllers, fuzzy logic and neural networks are utilized [2]. The conventional PID is commonly utilized in controlling the level, but the parameter of those controllers must be turned by tuning method either in time response or frequency response to meet their required performances [6, 7]. The block diagram of PID Controller with constants is shown in Figure. 2. P element: e (t) is the proportional to the error at the instant t, which is the present error. I element: ∫ dt is proportional to the integral of the error up to the instant t, which can be interpreted as the accumulation as the past error. D element: is proportional to the derivative of the error at the instant t, which can be interpreted as the prediction of the future error. [8] Figure.2 Block diagram of PID Controller 176 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 186. B. Simulink Modelof Water Level Controller Considering the tank process described above, the Simulink Model is designed the as shown in the Figure.4 and is saved with the name of “tankpid.mdl.” Figure.4 illustrates that the process includes PID Controller, valve and water tank. 1) Controller: The PID controller is the built-in block in the Simulink. Its block diagram representation is shown in the Figure.3. As described earlier, it has three parts namely proportional, integral and derivative. The input is an error signal (difference between the reference signal and system output) which when comes into the controller is divided into three respective parts and output from each of these parts is then summed up to give a stable output. The output of this block is the weighted sum of the input signal, the integral of the input signal and the derivative of the input signal [9]. In this case, this controller is acting as a PD-controller as the parameter which set for the integrator is zero. Figure.3 Block Diagram for PID-Controller 2) Valve: The water is pumped into the tank via valve with controllable degree of opening. 0 means the valve is closed and 1 means that the valve is fully open. The more the valve is open (the closer the value is to one), the larger the inflow to the tank. As illustrated in Figure. 5, the product of the condition provided in the limited integrator and inflow rate of the water gives the rate at which water flows out. One of the important parts of the valve is limited integrator. The reason behind using the limited integrator is to stop the output from surpassing specific levels. The function of this block is that, as soon as the output surpasses the limits, it turns off the integral action for preventing the integral wind up. The limits can be changed for the period of simulation. Figure.5 Simulink Model for Valve Figure.4 Simulink Model for PID- Controller 3) Tank: The water level in the tank can be described mathematically using (2) and tank thus designed is shown Figure 6. 177 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 187. Figure.6 Simulink Modelof Water Tank C. Execution After the completion of the Tank-Design Procedure, the next step is Execution. Executing the command “tankpid” in MATLAB opens a Simulink program where PID control of the tank process is simulated. The desired water level (reference signal) and the actual water level are saved into the workspace of MATLAB in the variable comp. By executing the MATLAB command “plot (comp.time,comp.signals.values)”, the reference signal and the actual water level are plotted against simulated time. D. Results The results obtained from the Simulink model of PID Controller are shown in the Figure.7. Figure.7 Output of PID- Controller These results show that the level of water reaches to the wanted output after passing some overshoots and oscillations. The goal behind developing this water level control to illustrate that the Fuzzy Logic Controller can be used to control the water level and to compare its control performance to the conventional PID Controller. For this purpose, we have replaced the PID controller with Fuzzy Controller in the Simulink model used in the next sections of this paper. IV. FUZZY LOGIC CONTROL OF THE WATER LEVEL TANK The Fuzzy Logic Controller is developed through following two approaches: (a) Adaptive Neuro-Fuzzy Inference System (ANFIS) Approach: For a particular system, a FLC can be designed in different manners. In the ANFIS approach, the FLC is trained to produce the response similar to that of the PID controller and once the training reaches the required error tolerance, the PID controller will be totally replaced by the FLC. (b) Design of a Fuzzy Inference System (FIS): FIS System is developed by using Mamdani approach using Fuzzy Logic Toolbox in MATLAB. This involves designing and tuning of the membership functions, I/O rules, and the de-fuzzification technique. A) Fuzzy Controller Design The Fuzzy Controller for the tank process can be designed by using the graphical tool in MATLAB. First of all, a Fuzzy Controller (a FIS) is constructed which uses the ‘level’ and ‘rate’ as input. The ‘output’ of the controller is the change in valve opening. The general working of tank lies in its overflow and underflow conditions. Taking care of these circumstances, a brief description of the number and shapes of input and output membership functions and rule base is given in the following steps: 1) The Fuzzy Mechanism: As shown in the Figure.8, the system has two inputs and one output. The linguistic term used for the input is ‘level’ and ‘rate’ and for the output is ‘valve’. Figure.8 Block diagram of Water Tank (a) Input 1: For this FLC, the level (control error) is defined as input of this Fuzzy Inference System (FIS).This input basically defines the control error for the water level. 178 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 188. The input rangeis [-1, 1]. It consists of three membership functions namely high, good and low. Trimf is used for ‘good’, and Trapmf for ‘high’ and ‘low’ as a nominal function. These membership functions are shown in Figure.9. Figure.9 Membership function for input variable ‘level’ (b) Input 2: The rate (flow rate) is defined as the second input. Its range is [-1, 1]. The membership functions are named as ‘rising’ and ‘falling’. Trapmf is used as a nominal function. Figure.10 Membership function input variable ‘rate’ (c) Output: The designed system has only one output given by the linguistic variable ‘valve’. It basically represents the position of the valve by accounting for the change in valve opening. The output space range is [-1 1]. The membership function for the output is shown in Figure.11. Figure.11 Membership function for output variable ‘valve’ 2) Rule Base: The input-output relationships are used to develop the IF-THEN rules for the FIS [10]. From the two input variables and one output variable total of four following rules are developed. 1. If (level is low) then (valve is open_fast) 2. If (level is high) then (valve is close_fast) Figure.12 Simulink Model of Fuzzy Controller of Water Tank 3. If (level is good) and (rate is rising) then (valve is close_slow) 4. If (level is good) and (rate is falling) then (valve is open_slow) B. Design of the Simulink Model 179 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 189. After setting theparameters of the FIS membership function and making the fuzzy controller; the FIS file named “tank.fis” is exported to the workspace. The reason behind the fact that the derivative of the process variable is taken instead of the derivative of error is that the derivative mode amplifies the sudden changes in controller input signal and cause large variation in the controller output. [11]. Now the tank with Fuzzy Controller can be simulated by using model shown in Figure.12 and the output comes out to be as shown in Figure.13. Figure.13 Output of Fuzzy Controller V. CONCLUSION In plant process control, the measurement and control of flow are two very important factors. But these two factors are susceptible to disturbances because of the different variables affecting the flow rate. Therefore for optimizing the performance of a plant, the flow controllers are required to be robust. Fuzzy Logic Controller is proposed to replace the conventional PID controller due to its superior applicability and robustness [12]. The concept of Fuzzy Logic is a representation of the human thinking and decision making process. The experimental results show that Fuzzy Control is better than PID Control not only because of simplicity of its implementation, better control performance, robustness and overall stability but also because it gives a reasonable consideration to variety of parameters like overshoot and time of response etc and for this reason, it tunes and develops the FIS more intuitively as compared to the PID Controller. REFERENCES [1] L.A. Zadeh, “Fuzzy Sets”, Informal Control, Vol. 8, Pp. 338- 353, 1965. [2] Y. F. Chan, W. Chan and H. T. Mok "Adaptive Neurofuzzy Network Based PI Controllers with Multi-objective Functions",Next-Generation Applied Intelligence, Vol. 5579, Pp. 604-613, 2009. [3] Qi Li, Yanjun Fang, Jizhong Song, Ji Wang, "The Application of Fuzzy Control in Liquid Level System", 2010 International Conference on Measuring Technology and Mechatronics Automation, 03/2010. [4] Wangnipparnto S. and Tunyasrirut S., "Level Control in Horizontal Tank by Fuzzy Logic Controller", 2006 SICE-ICASE International Joint Conference, 10/2006. [5] Duane Hanselman and Bruce Littlefield, "Book: Mastering Matlab® 5", ISBN# 0-13-858366-8, Prentice-Hall, 1998. [6] W. K. Ho, C.C. Hang, and J. H .Zhou, "Performance And Gain And Phase Margins Of Well-Known Pituning Formulas," Accepted For Publication In IEEE Trans. Contr. Syst. Techno, 1995. [7] J. G .Ziegler and N. B. Nichols, "Optimum Settings for Automatic Controller," By ASME Trans. Vol. 64, Pp.759-768, 1942. [8] Aravind, Sekhar R, and B R Vinod, "A generalized method for improving the performance of controllers by combining PID and DELTA rule", 2012 Annual IEEE India Conference (INDICON), 2012. [9] “MATLAB documentation,” Mathworks (www.mathworks.com ) [10] Pathmanathan E, and R Ibrahim., "Development and implementation of Fuzzy Logic Controller for Flow Control Application", 2010 International Conference on Intelligent and Advanced Systems (ICIAS), 2010. [11] Silviu Ionita and Emil Sofron, “The Fuzzy Model for Aircraft Landing Control”, Prentice Hall, 2001 [12] Ronald R. Yager and Dimitar P. Filev, “Essentials Of Fuzzy Modeling And Control”, Johan Wiley & Sons, 2002. [13] D. Wu, F. Karray, I. Song, ‘Water level control by Fuzzy Logic and Neural Networks’, IEEE Conference on Control Applications, pp.3134-39, 2005. [14] Namrata Dey, Ria Mandal and M. Monica Subashini, “Design and Implementation of a Water Level Controller using Fuzzy Logic”, International Journal of Engineering and Technology (IJET), Jun-Jul 2013. [15] J. Yen, "Fuzzy logic-a modern perspective", IEEE Transactions on Knowledge and Data Engineering, 1999 180 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 190. The Use ofRadioactive Isotopes and Nuclear Reactors in Space Application: Propulsion and Power concept Gohar Ali Hira High school and Science College Shahdherai Swat, Khyberpakhtunkhwa, Pakistan. Email:105goharswati@gmail.com Abstract As we know that men have insatiable thirst in exploration and discovery especially to unfold the mysteries of universe. There are lots of frontiers that could be covered in space exploration but unfortunately, there are some reasons which affect it. Firstly, the distances to be covered are extremely large, this large distance in outer space makes it very difficult to prospect for an astronaut involved in such a mission. Even a simple mission to far planets like Uranus, Neptune or even to nearest planet Venus would take a long time. Hence in case of Venus, it would need to take more than one year for such a mission. However, to explore outer space of solar system like to go to our nearest star “PROXEMA CENTAURI” it would take more than millions of years to reach there. Moreover, the second problem is greater amount of energy required, which need for an astronaut onboard for any outpost installation. Luckily, the availability of nuclear technology allows us to overcome these two problems. In this paper, it is demonstrated that by using nuclear reactors or Radioactive isotope decay like Americium-241 with half life of 432.7 years or Uranium -235 using Concept of propulsion and power can provide enough energy to travel vast distances in space and the energy need for an astronaut at ISS would be easily available. Keywords: nuclear propulsion, nuclear rocket, space power application, nuclear power in space station. I. Introduction The huge distance in space becomes very difficult for a space scientist and for astronaut to cover that. The reason is the fuel need for such a mission not finds easily because to goes out of solar system would takes million of year. Even a simple mission in our solar system to the terrestrial planet like to Venus or to Mercury would also takes hundreds of years so, the one problem which effects our mission the most is the propulsion need for a space shuttle to covers such a distance, And secondly we also know that energy requires for an astronauts onboard for any outpost Installation also becomes difficult. Yeah it would be tempting to believe that all power in space could be supplied by solar means since the sun is available and free. However, in many cases the mission some time takes place in the dark place and solar panel is not always suitable for a mission (figure 1). 181 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 191. Nuclear reactors haveprovided electrical power for some of the U.S. space program’s greatest successes, including the Apollo lunar landings and the Viking Landers that searched for life on Mars. RTGs made possible NASA’s celebrated Voyager explorations of Jupiter, Saturn, Uranus and Neptune, RTG Power sources are enabling the Galileo mission to Jupiter, the international Ulysses mission studying the Sun’s Polar Regions, and the Cassini mission to Saturn. Pioneer 10 was launched from Cape Kennedy on 2 March 1972. It was the first interplanetary probe, successfully navigating the asteroid belt before Making rendezvous with Jupiter and Saturn The probe was equipped with an array of instruments for measuring such phenomena as the solar wind and the magnetic and radiation fields surrounding Jupiter. Nuclear reactors being a reliable and competitive source of energy they have some risk too and as well it also becomes difficult to do in some condition. The overcome of these problems are also demonstrated in this paper. II. Space power concept: To remove the waste, to lift a load and as well as for communication in International space station or on ground you will need sufficient power to do so for a long time. Moreover it would be very difficult for to generate electricity by thermal or chemical means in microgravity environment [5]. However with the availability of radioactive isotopes and as well as with the help of nuclear reactors the Electricity and all the other requirements in ISS would be easily available without any difficulty for several years [5] For the production of electricity heat energy can be converted by many devices but the precise one is to use Thermo electric generator which typically works on Seebak Concept i.e dV = SAB . (Kh - Kc) Where dV is Voltage difference, SAB is two dissimilar metals, & Kh - Kc is temperature gradient. In this way Nuclear reactors can provide infinite power for almost any time. However, they are not practicable for applications below 15 kilo watt (kW). Figure 1 : regimes of possible space power applicability 182 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 192. Radioisotopes are consideredto be best used for continuous supply of low levels (up to 5 kW) of power or in combinations up to many times this value. For this reason, especially for long interplanetary missions, the use of radioisotopes for communications and the powering of experiments are preferred. The nuclear process shown in Fig. 2 can either be a critical reactor or radioisotope fuel source such as plutonium oxide. In either case the heat can be converted to electricity either statically through thermoelectric or a thermionic converter, or dynamically using a turbine generator in one of several heat cycles (Rankin, Stirling, Brayton).. The nuclear workhorses for current space missions are the RTGs and the TEGs powered by radioisotopes in the Russian Federation that provide electricity through static (and therefore reliable) conversion at power levels of up to half a kilowatt, or more by combining modules. And that energy can be used for to run machinery. [2]. Space Propulsion concept Nuclear reactor can also be used in rocket propulsion system. Propulsion system in space is to change the position Or velocity (v), of space craft under the strong influence of gravity. [6] When a space craft moves it generally term is momentum (mv) so, the basic concept of propulsion is to change the momentum (mv) of a space craft, and change in momentum is known as impulse (Imp). So, the basic aim of propulsion in space is to create impulse, to measure the impulse is often term is specific impulse (Isp).[6] Specific impulse also known the exhaust velocity (ve) (OR) the impulse per unit weight ( ) but while discussing the engine in space there is no weight so it can become impulse per unit mass The difference arises here only by the acceleration due to gravity (g).Where its value is 2 on earth. However the value of g and reaction mass is not important while the vehicle discussing in space. [4]. The two parameter specific impulse and the mass of the rocket during launch and then in orbit is measure the efficiency of propulsion energy. Where the equation for the measure of specific impulse as seen as (1) (1) Where represent mass of propellant and represent mass flow. Here it is seen that mass of propellant is inversely proportional to specific impulse, So, In chemical propulsion there is used hydrogen and oxygen which provides a specific impulse about 4400 m/s with mass ratio of earth escape of 15.[1] Figure 2.The conversion of heat energy into kinetic or to mechanical energy 183 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 193. However americium ORhydrogen heated with fission reactor can achieves twice the impulse with mass ratio of 3.2. And with different core it can be generate much more impulse i.e. seven times having mass ratio of only 1.2 [1]. III. Nuclear rocket Nuclear rocket is the rocket which use nuclear reactors usually fission concept instead of chemical propulsion to generate thrust it include nuclear thermal rocket (NTR), Nuclear electric rocket (NER) or Hybrid (NTR/NER). [2] These rocket are propelled by the force of nuclear explosion, usually liquid hydrogen is heated via nuclear fission which subsequently can be expanded in nozzle and thus accelerated at a high ejection velocity (6,000 to 10,000 m/s) to generate thrust. There are also many design are find for nuclear thermal propulsion i.e. liquid, solid or gas core rocket. [6] They use solid, liquid or gas core reactors respectively. Electro thermal propulsion rocket have been use in many orbital mission in this type of rocket the propellant is heated electrically (heat energy of nuclear reactor are converted into electricity) and then it accelerate the ionized gas via electrostatic force at supersonic velocity having range from 1,000 to 5,000 m/s and thrust range 0.01 to 0.5 N. But it has low Specific impulse therefore it can be use only for interplanetary missions, however, it can be use for long time instead of chemical rocket [2] The nuclear rocket has many advantages which overcome on chemical rocket. Like it can be use in much complicated mission even it can be travels to the edge of our solar system or even to our nearest cluster, nuclear space craft can be also used for to carry heavy payload or to carry flyby, rover etc to the far planet in our solar system, because it provides a large specific impulse as a result faster travel and the probability for complicated missions can be met, secondly nuclear reactor has low molecular weight which can increase the propulsive force per unit of propellant flow and allowing us for increase the proportion of total weight of payload [3]. IV. Microgravity effect on core reactor In microgravity environment every mass tends to be in state of motion (try to revolve from heavenly bodies near them due to centripetal force) So, The big challenge of core reactors in microgravity environment is to control the flow of fission. It seems like very difficult but under certain design it is possible to keep the reaction self sustain. So the level of energy would be maintained. In core reactors usually the flow is turbulent due to the randomness motion of molecule. And hence it becomes difficult to keep flow within specified boundaries. For this reason Reynolds number flow are preferred for microgravity environment. If it is not under the specified boundaries then the chain reaction times would be increases due to uncontrolled condition in the core reactor. And some parts of the core reactor are exposed and hot spot would be appear on that as well some parts of the core reactors would be damage due to high temperature because the distribution of heat is not possible in such circumstance. 184 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 194. Figure 3 Precisecore reactors for microgravity environment 1, Reflector moderator 2, Gaseous fissile zone 3, Working medium flow zone, 4, Fissile material diminution replenishment 5, working medium Inlet. To overcome on this problem a special core is design (shown in figure 3) to sustain fission reactor under control especially with this core the possibility of environmental hazard would be reduce. And various high temperature fuels are possible with this design. in The fissile material for example Uranium or americium would be located in the center of cavity enclosed by the neutron moderator reflector. The working gas flow is close to the cavity walls and it is heated by the high temperature plasma radiation which causes the fission reaction to be self sustained under microgravity conditions. [8] V. Radiation hazards of Nuclear rocket Expose of a person to radiation does not mean that the person would get cancer. People are exposing to radiation on daily basis like on lesser extent from human activities i.e.-rays. And radiation comes out from natural environment i.e. in the earth cosmic rays and radon. The radiation expose is measures in unit of Dose called millirem. Over the course of year the average person would expose to a total of 360 millirem to natural radiation .In nuclear thermal rocket some of the radioactive element would be release into the atmosphere The particles which are hazardous to people would remain high in the atmosphere for long course of time gradually these particles being spread thinly across the world. And eventually making their way on the surface especially on ocean since these materials are Insoluble once it reaches to the surface most of it would become trapped by the soil or by the ocean and not pose a health hazard. Thus most of the release materials won’t be breath by people. But the small amount of released material would be breath by the people and as well this small amount will be also distributed across the world. So, In this way the amount to be breath will be much less the person may receive it less than one millirem in 50 years. This small radiation is negligible compare to the 15,000 millirem a person will get it from natural resources over the period of 50 years. [7] VI. Result and conclusion In this paper we have emphasize on the importance of nuclear reactors in space based application, where the nuclear reactors and radioactive isotopes would provides infinite heat energy for long durations during journey of the space craft and that energy can be converted into other form of energy and the need for power and propulsion would be met. But in microgravity environment it becomes difficult to do but however with the help of gas core reactors and other special core design would allow us to do so in microgravity environment too. And the space exploration would be become more precious and peaceful. 185 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 195. VII. Acknowledgment I wouldlike to thank national space agency SUPARCO Space awareness society including Mr. Qamar Abbas and Mr. Tahir hussain and My School principal Mr. Sami ullah for to guide me as well for their support. I also like to express my gratitude with SUPARCO propulsion team for did my conceptual analysis. I am also grateful to Mr. Rahman u din and Mr. Zulfiqar Ali for Insatiable support and guide. References [1]. Rocket propulsion elements: introduction to the engineering of rockets by George P. Sutton, Oscar biblarz___7th edition. [2].The role of nuclear power and nuclear propulsion in peaceful exploration of space by International atomic energy agency Austria September 2005. [3]. Basic of space flight by National Aeronautic and space administration sec 2, 2002. [4]. Advance space propulsion for the 21st Century by Robert H .Frisbi a report from 9 August 2003 to 15 September 2003[5].Civilian usage of nuclear reactors in space volume 1, science and global security, 1989, [6] Thruster precisely guide by Hiss, M martin and K, Rachule October 2002. [7] Space craft power for cassini (http://www.jpl.nasa.gov/cassini/) [8] Fission fragments heating for space propulsion by C. Rubbia 186 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 196. Impact of ThermalAging on Microstructure and Mechanical Properties of high Sn Content, Sn-Pb Solders Muhammad Aamir1 , Riaz Muhammad1 , Naseer Ahmed1 1 Dept. of Mechanical Engineering CECOS University of IT & Emerging Sciences, Peshawar, Pakistan. aamir@cecos.edu.pk Khurshid Alam2 2 Department of Mechanical and Industrial Engineering, Sultan Qaboos University, Muscat, Oman. Abstract — The microstructure and mechanical properties of solder joints in electronic components are evolving when exposed to thermal aging. The current work is based on the investigation of high Sn content Sn-Pb solders i.e. 96Sn-04Pb. Scanning Electron microscopy was carried out to observe the microstructure examination followed by chemical composition with elemental mapping using Electron dispersive X-ray. Mechanical properties were examined using tensile tests. The effect of thermal aging on microstructure and mechanical properties has also been studied. The results show that mechanical properties are highly influenced by its microstructure after thermal aging. Due to aging effect, the microstructure becomes coarsen and ultimately the mechanical properties including yield strength and ultimate tensile strength degrades but increase in ductility is obtained. Keywords—96Sn-04Pb, Microstructure examination, Mechanical properties. I. INTRODUCTION The use of the solder alloys as metal joint can be dated back to thousands of years and advancement in the electronics industry are continuously in progress. However, the use of surface mount technology (SMT) in printed circuit boards (PCB) means that solders performs a leading role to make electronic connections [1]. Surface mounting of the components means that during impact loading the joint transfer whole momentum of a component to the board. Therefore, mechanical properties of the solder joint are of great importance. Another important factor is the accelerated aging due to relatively high temperature under severe service condition. This also makes the joint in high fraction of their melting point which results in microstructural changes. Thermal aging is also responsible for reduction in strength through grain growth after microstructural coarsening process [2-6]. The mechanism for the decrease in strength in solder alloys after aging is associated with coarsening of microstructure. Coarser microstructure leads to fewer grain boundaries to block the dislocation movements which ultimately cause reduction in strength [7]. Solder joints are of great important in the electronic industry, they are not only used to physically hold assemblies together but also used for transmission of electrical signals [8]. When electronic device is in use, the solder connections faces mechanical stresses. The on off switching of the system also leads cyclic mechanical load and ultimately results in stresses in solder joint [9, 10]. The decrease in the strength and microstructure coarsening are also the result of high homologous temperature and ultimately solder alloys will undergo creep even at room temperature [11]. Pb based solders, particularly 63Sn-37Pb or near eutectic Sn-Pb, has been used in electronic industry for a long time. This is due to the fact that Pb bearing solders have good metallurgical, mechanical, solder ability, reliability, low cast, physical, mechanical and metallurgical properties [12]. The use of Pb in Sn-Pb solder is to reduce the surface tension which ultimately results in improvement of wettability of the solder joint [13, 14]. Sn is used as it is ductile in nature and having good resistance to corrosion. High Sn concentration also lead to high tensile and shear strength [15]. However, the concern about Pb toxicity and legislation to ban on Pb bearing electronic products results in moving the electronic industry towards environmental friendly green products [16-18]. Therefore, electronic industries and academic institutes accepted a challenge for characterizing new solders in terms their mechanical properties, material prosperities and manufacturing process compatibility with components and printed circuit boards (PCB) [19]. The current work is based on developing and characterizing a new solder with high Sn contents. The Pb content has been limited upto 04% instead of 37% conventional Pb solders. The reason for limiting Pb upto 04% is to get the properties of Pb and make the environment as green as possible. The microstructure including intermetallic particles and mechanical properties have been examined at room temperature and after thermal aging at 100°C for 50 hours. 187 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 197. II. EXPERIMETAL WORK Preparationof good sample is necessary for experimental study for better characterization and mechanical properties. Samples of 250g ingots of pure metals with composition (wt. %): 96Sn-04Pb is prepared by putting them in the crucible and heated in an oven for 30 minutes at room temperature to get the molten metal. A die composed of two parts of aluminum and a central steel plate have been used to get dog-bone specimen after the alloy is then cooled in air. The final dog bone shape specimen is shown in Fig. 1 [20]. In order to analyze the microstructure, the specimen is cut into pieces and mounted in bakelite powder using mounting press for proper handling and avoiding any distortion. The temperature for using bakelite was is at 160◦ C for 09 minutes with 130 Kg/cm2 . The mounted samples are initially grinded with different grit sizes of silicon carbide water proof sand papers. During grinding process, heat is generated so tap paper is used for lubrication purpose to avoid any damage to the surface. After sand paper grinding, the specimens are polished with polycrystalline diamond suspension with abrasive particle size of 6 μm, 1 μm and 0.25 μm to give extra shine to the surface and then cleaned with distilled water to remove any residue from the specimen left during polishing. The specimens are then etched for 45 seconds using 5% hydrochloric acid and 95% ethanol. Proper care is necessary for good observation of grain particles and grain boundaries for examination of morphology under Scan electron microcopy. Before Scan electron microscopy (SEM), samples are coated with silver paste and placed in sputter coater and vacuum evaporator for gold coating to make the material conductive. The SEM images have been taken at different location and magnification followed by Electron dispersive X-ray (EDX) to confirm the chemical composition of the alloy. For thermal aging, samples for microstructure examination and dog bone specimen are exposed to 100°C for 50 hours in oven. The aging process is necessary to get the better understanding of performance during extreme operating conditions. For tensile testing, universal testing machine is used to determine the mechanical properties. The data has been taken at room temperature. The universal testing machine with dog bone specimen is shown in Fig. 2. Fig. 1 Dog bone specimen [20] Fig. 2 Universal testing machine with dog-bone shape specimen III. RESULTS AND DISCUSSION A. Effect of Thermal Aging on Microstructure The microstructure of 96Pb-04Pb alloy is described by low and high magnification of images at different resolutions, the images are taken for the as casted and after thermal aging to examine the effect of changes in the morphology. The “as casted” SEM images are shown in Fig. 3 and “thermally aged” in Fig. 4. The chemical composition is confirmed by Electron dispersive x-ray shown in Fig. 5. The microstructure of 96Sn-04Pb is composed of soft Sn matrix and hard IMCs of Pb. The black zone plate shape is the Sn matrix and white color rods shaped are the IMCs of Pb. The elemental mapping of Sn matrix and Pb particles are shown in Fig. 6. These IMCs are usually brittle in nature in comparison with soft Sn matrix and responsible for changes in mechanical properties. The growth of these IMCs results in coarsening of the alloy. The size of these IMCs depends on many parameters including composition of alloy, environmental conditions under sever temperature during service and cooling rate. A highly temperature aged alloy can produce coarse microstructure. By closely observing the SEM images, it is observed that the Pb white particles become wider and distributed in Sn matrix after thermal aging. Since the properties of the solder joints are highly microstructure dependent which changes reasonably during its life time so, it is reasonably expected that after thermal aging the mechanical properties would be worsen. Increase in the growth of the grain size after thermal aging may also leads in reducing the strength because larger grains create fewer obstacles per unit area to the movement of dislocation. 188 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 198. Fig. 3 SEMimages at various magnification of “as casted alloy” Fig. 4 SEM images at various magnification of thermally aged alloy Fig. 5 EDX Analysis of alloy Fig. 6 elemental mapping of Sn matrix and Pb particles B. Effect of Isothermal Aging on Mechanical Properties As thermal aging affects the microstructure of the solder alloy and ultimately influences mechanical properties including yield strength and tensile strength, therefore the impact of thermal aging over these properties is discussed in this study. Yield strength, ultimate tensile strength and elongation to failure are presented in Fig. 7 for the “as casted” and “thermally aged” samples and summarized in Table. I. Since the microstructure becomes coarse with thermal aging, therefore, decrease in yield strength and ultimate tensile strength is noted. It is investigated that 17.4 % reduction in yield strength and 25.5% reduction in ultimate tensile strength have been occurred but 32.1% increase in elongation is found. Elongation to failure is the measure of ductility of material so a ductile material offers a high elongation. It means that after thermal aging the ductility of the alloy is increases. For larger grains, small boundary area per unit volume exists therefore, a larger grain size results in increasing the ductility. 189 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 199. Fig. 7 Mechanicalproperties of “as casted and “thermally aged” alloy TABLE I. TENSILE TEST DATA Specimen Tensile Test Description Condition Alloy Without Thermal Aging Alloy With Isothermal aging at 100C up to 50 hours Elongation @ Peak (mm) 1.0090 1.4870 Load @ Peak (N) 435.00 324.00 Strain @ Yield (%) 1.2333 1.3946 Strain @ Peak (%) 1.6817 2.0231 Stress @ Yield (N/mm2 ) 16.856 13.919 Stress @ Peak (N/mm2 ) 19.595 14.595 IV. CONCLUSIONS It is concluded that from the results and observations that the properties of the solder joint are highly microstructure dependent and a strong relationship exists between microstructural and mechanical properties. It has been observed that thermal aging leads to coarsen the microstructure of the alloys and ultimately affects the mechanical properties. Mechanical properties including yield strength and ultimate tensile strength decreases but considerable increase in ductility have been examined. This also concluded that thermal aging is closely related to mechanical properties. Therefore, great care must be taken regarding microstructure evolution and mechanical properties in designing and developing solders but for electronics green and environment friendly Pb free solders must be introduced. REFERENCES [1]. Prymark, J., et al., Fundamentals of microsystems packaging. ed. RR Tummala, Mcgraw-Hill Book Company, New York, USA, 2001: p. 420. [2]. Jung, K. and H. Conrad, Microstructure coarsening during static annealing of 60Sn40Pb solder joints: I stereology. Journal of Electronic Materials, 2001. 30(10): p. 1294-1302. [3]. Kang, J. and H. Conrad, Microstructure coarsening during static annealing of 60Sn40Pb solder joints: II Eutectic coarsening kinetics. Journal of Electronic Materials, 2001. 30(10): p. 1303. [4]. Jung, K. and H. Conrad, Microstructure coarsening during static annealing of 60Sn40Pb solder joints: III intermetallic compound growth kientics. Journal of electronic materials, 2001. 30(10): p. 1308-1312. [5]. Basaran, C. and Y. Wen, Coarsening in Bga solder balls: modeling and experimental evaluation. Journal of Electronic Packaging, 2003. 125(3): p. 426-430. [6]. Tang, H. and C. Basaran, Influence of microstructure coarsening on thermomechanical fatigue behavior of Pb/Sn eutectic solder joints. International Journal of Damage Mechanics, 2001. 10(3): p. 235-255. [7]. Ma, H., et al. The influence of elevated temperature aging on reliability of lead free solder joints. in Electronic Components and Technology Conference, 2007. ECTC'07. Proceedings. 57th. 2007. IEEE. [8]. Efzan, E. and A. Marini, A review of solder evolution in electronic application. International Journal of Engineering, 2012. 1(1): p. 2305-8269. [9]. Frear, D., D. Grivas, and J. Morris, A microstructural study of the thermal fatigue failures of 60Sn-40Pb solder joints. Journal of Electronic Materials, 1988. 17(2): p. 171-180. [10]. Hacke, P., A. Sprecher, and H. Conrad, Microstructure coarsening during thermo- mechanical fatigue of Pb-Sn solder joints. Journal of Electronic Materials, 1997. 26(7): p. 774-782. [11]. Wassink, R.K., Soldering in Electronics, 1989. Electrochemical Publication. 177. [12]. Sharif, A. and Y. Chan, Dissolution kinetics of BGA Sn–Pb and Sn–Ag solders with Cu substrates during reflow. Materials Science and Engineering: B, 2004. 106(2): p. 126-131. [13]. Abtew, M. and G. Selvaduray, Lead-free solders in microelectronics. Materials Science and Engineering: R: Reports, 2000. 27(5): p. 95-141. [14]. Zeng, K. and K.-N. Tu, Six cases of reliability study of Pb-free solder joints in electronic packaging technology. Materials science and engineering: R: reports, 2002. 38(2): p. 55-105. [15]. Needleman, H.L., et al., The long-term effects of exposure to low doses of lead in childhood: an 11- year follow-up report. New England journal of medicine, 1990. 322(2): p. 83-88. [16]. Noor, E.E.M., et al., Wettability and strength of In– Bi–Sn lead-free solder alloy on copper substrate. Journal of Alloys and Compounds, 2010. 507(1): p. 290-296. [17]. Abadin, H., et al., Toxicological profile for lead. Atlanta (GA): Agency for Toxic Substances and Disease Registry, 2007. 190 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 200. [18]. Harrison, M.,J. Vincent, and H. Steen, Lead-free reflow soldering for electronics assembly. Soldering & Surface Mount Technology, 2001. 13(3): p. 21-38. [19]. Smith III, E.B., Environmental Impacts and Toxicity of Lead Free Solders. op. cit, 1999: p. 2. [20]. Sadiq, M., R. Pesci, and M. Cherkaoui, Impact of thermal aging on the microstructure evolution and mechanical properties of lanthanum-doped Tin- Silver-Copper lead-free solders. Journal of electronic materials, 2013. 42(3): p. 492-501. 191 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 201. Effect of aryldiazonium salt fucntionalization on the electrical properties of MWCNTs and MWCNTs/CF reinforced polymer composite Madni Shifa, Fawad Tariq, Kiran Ali, Rasheed Ahmed Baloch Quality Management Directorate General Pakistan Space and Upper Atmosphere Research Commission (SUPARCO) 75270 Karachi, Pakistan Email: madnishifa@yahoo.com Abstract— Multiwall carbon nanotubes (MWCNTs) are kind of inert and hydrophobic material. These MWCNTs do not disperse properly and tend to agglomerate in polymer matrices due to high surface area and strong van der Waal forces between themselves. Numerous studies have been conducted in the last few years to resolve the issue of CNTs dispersion and solubilization in polymer matrix through functionalization process. However, most of the functionalization treatments introduce defects and results in deterioration of electrical properties. In this study, MWCNTs were covalently functionalized by direct grafting of aryl diazonium salt (BF4N2-C6H4-NO2) and effect of functionlization on electrical behavior was assessed under applied electric field in an electrolytic cell. The MWCNTs were dispersed ultrasonically in 2-propanol which act as an electrolyte in electrolytic cell. Voltage was applied on copper electrodes and movement of MWCNTs inside the electrolyte was carefully observed in-situ under optical microscope. In case of functionalized MWCNTs (f-MWCNTs), no effect was observed under applied voltage up to 100 V whereas un-functionalized MWCNTs (p- MWCNTs) tend to align themselves in regular fashion showing conducting behavior. SEM analysis was performed to see the consequence of functionalization treatment on morphology of MWCNTs and optical microscopy was carried out to see the effect of functionalization on the dispersion of MWCNTs in cured and uncured state of polymer matrix. MWCNT/CF reinforced polymer composites were fabricated using hand layup and compression molding technique. DC electrical conductivity & electromagnetic interference shielding effectiveness (EMI SE) testing were performed to examine the effect of MWCNTs functionalization on fabricated samples. Test results showed that the functionalization degraded the electrical properties of MWCNTs. Keywords— Aryl diazonium salt, Functionalization, Morphology of MWCNTs, SEM analysis, EMI SE I. INTRODUCTION Carbon nanotubes (CNTs) are carbon based advance nano- materials with unique set of structure and properties. Commonly two types of CNTs are being used i.e. single wall carbon nanotubes (SWCNTs) and multi wall carbon nanotubes (MWCNTs). Outer diameter of SWCNTs is around 1-2 nm, while the outer shell diameter of MWCNTs ranges from 10- 100 nm. A typical length range of CNTs varies from 10-1000 μm and have aspect ratio of about 1000:1 [1]. CNTs are one of the most promising materials known to men; their mechanical, thermal and electrical properties are superb in all aspects [2]. Considering electrical properties of CNTs, electrical resistivity is 0.4-0.7 Ωm, maximum current density 107-109 Acm-2 and the band gap of CNTs is 0-0.7 eV depending upon their nature. Owing to exceptional electrical properties in conjunction with high aspect ratio and low density (one sixth time lower than steel), CNTs have gained great interest for developing lightweight electrically conductive polymer nanocomposite [3, 4]. CNT/epoxy polymer composites have been widely used in advance military aircrafts, fairings, missiles, personal communications and electronic packaging in which CNTs act as nano-fillers in epoxy matrix [5].High electrical conductivity and nanometer size of CNTs imply very low level of percolation threshold in CNT/epoxy nanocomposite [6]. Exceptional electrical properties of CNTs directly related to high aspect ratio and defect free structure of CNTs. Advance applications of epoxy polymer composite demands high electrical conductivity, electromagnetic interference (EMI) shielding and electrostatic dissipation (ESD) characteristics. Electromagnetic waves produced from some electronic component decrease the performance of nearby electronic system by introducing noise, disturbance, data loss, etc. Hence, there is an intense need to shield the sensitive electronic circuits form undesired signals through EMI shielding material. One basic criteria of EMI shielding material is that it should be electrically conductive. CNT/epoxy composite is found to be promising candidate in this regard [4, 7-12]. However, one key issue of CNT based polymer nanocomposite is the dispersion of CNTs in polymer matrix. Dispersion includes separation of CNTs bundles and then stabilization of CNTs in polymer matrix. As CNTs agglomerates into bundles due to high surface area and these bundles exhibits inferior properties as compare to individual CNT. Another key challenge is the good interfacial bonding between CNTs and polymer matrix. As CNTs are inert and atomically smooth material so there is a lack of bonding between polymer matrix and CNTs [13-15]. Best route to overcome stated problems is to attach functional groups through surface modification treatments of the CNTs. There are numerous approaches to functionalize CNTs depending upon their end use and the choice of functional group. Among 192 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 202. them, modification ofCNTs through aryl dizonium salt is more versatile and easy with variety of functional groups. However, fucntionalization process induces defects and decreases the aspect ratio of CNTs thereby degrading the electrical properties like conductivity, EMI shielding and ESD of CNT reinforced polymer composite [13, 16-18]. Most of the literature discusses the effect of acidic functionalization on the electrical properties of CNT based polymer composite. However, the effect of functionalization treatment through aryl diazonium salt on electrical properties of MWCNTs and MWCNT/CF reinforced composite has been seldom reported. Therefore a detailed study has been conducted in which MWCNTs were functionalized through aryl diazonium salt to attach covalent functional group. Effect of applied electric field on the electrical behavior of f- MWCNT and p-MWCNTs in electrolyte has been in-situ observed under light microscope and discussed here in detail. Moreover, electrical conductivity and EMI shielding effectiveness of f-MWCNTs/CF reinforced polymer composite and p-MWCNTs/CF reinforced polymer composite was also evaluated and compared. II. MATERIALS AND METHOD A. Materials MWCNTs used in this study were purchased from Cheap Tube Inc, US and were synthesized through CCVD (Catalytic Chemical Vapor Deposition). Specification of MWCNTs given by manufacturer: outer diameter 10-20 nm, length 10-30 µm, purity level > 95%, ash content <1.5%, specific surface area 180-230 m2 /g and bulk density 0.22 g/cm3 . Specifications details of carbon woven fabric which was used as reinforcement in epoxy matrix are: areal density 325±5, thickness 0.5±0.005 mm, breaking strength > 270Kg/5cm in wrap direction, weave style satin 3, 3K tow and equal no of threads in weft and warp direction. Thermoset epoxy resin (Bisphenol-A) containing a reactive diluent was chosen as matrix because of its long pot life, low viscosity and multifunctional properties. Cycloaliphatic amine was selected as hardener and used in the ratio of 10:3.5 by weight with epoxy resin. B. Functionalization of MWCNTs Functionalization of MWCNTs was carried out through a facile method in which 5-10 mg MWCNTs were dispersed in acetronitrile through bath sonication. After 30 minutes of sonication, 0.15-5 equivalent of diazonium salt (BF4N2-C6H4- NO2) was added and stirring of mixture was carried out for 5 days at room temperature. Then whole mixture was filtered through 0.45 micron size PTFE filter paper and washed several time with acetronitrile to remove unreacted dizonum salt. Functionalized MWCNTs were dried at room temperature under vacuum for 24 h before use [19]. Functionalization of MWCNTs was visually confirmed by dispersing 5 mg of pristine and functionalized MWCNTs in water in different tubes and kept for 24 h. After 24 h pristine MWCNTs were settle down at the bottom of tube whereas the functionalized Fig. 1. Schematic illustration of MWCNTs fucntionalization with 4- Nitrobenzenediazonium tetrafluoroborate. Fig. 2. Suspension of F (f-MWCNTs) and P (p-MWCNTs) in water after 24 h. MWCNTs were remained suspended which confirmed the functionlization of MWCNTs as shown in Fig. 2 whereas Fig. 1 demonstrates fucntionalization mechanism of MWCNTs. C. Fabrication of MWCNTs/CF polymer composite Functionalized MWCNT (0.2 wt% of matrix) was added to ethanol and sonicated in sonication bath for 30 min to open up MWCNTs bundles. Then epoxy resin was added to the f- MWCNTs contained ethanol and further sonicated for 30 min at 80⁰C to evaporate ethanol and disperse MWCNTs in epoxy resin. Upon cooling of MWCNT filled epoxy resin, hardener was added and solution was magnetically stirred for 10 min to thoroughly mix the hardener in the solution. Finally, 193 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 203. Fig. 3. Schematicof MWCNTs dispersion in matrix followed by hand layup on carbon fabric & curing of sample. Fig. 4. MWCNTs/CF reinforced polymer composite fabrication through compression molding technique. composite was fabricated through hand layup and compression molding technique as shown in Fig. 3. Aluminum die was in-house prepared for fabrication of composite laminates through compression molding method (Fig. 3a). Three layers of carbon fabric were cut according to the dimension (75 mm width & 150 mm length) and placed in the cavity of lower part (female part) of aluminum die one by one and impregnated with MWCNT filled epoxy resin through applicator brush. Upper part (male part) of the die was placed on the lower part and aligned through guided pins. Sufficient pressure was applied by hand tightening of guided pins so that excess resin escaped out through drain holes. Total 3 layers of carbon fabric along with MWCNT filled epoxy were used to get final 2 mm thick composite. Composite sample was cured by placing the aluminum die in an oven at 80°C and 120°C for 30 min and at 160°C for 2 h. After curing process, sample was removed from the die and trimmed to get smooth edges (Fig. 4b & 4c). Similar process was also carried out to fabricate p- MWCNTs/CF reinforced polymer composite samples. III. CHARATERIZATION TECHNIQUES A. Field emission scanning electron microscopy (FE-SEM) FE-SEM (MIRA 3 TESCAN) was carried out in order to observe the effect of functionalization on the morphology of MWCNTs. Functionalized MWCNTs were dispersed in ethanol through bath sonication , and one drop was taken on glass slide and gold coated to make the sample electrically conductive. Glass slide was observed under FE-SEM at high magnification. Similar process was also employed to examine the p-MWCNTs morphology under FE-SEM. B. In-situ observation of MWCNTs under applied electric field Functionalized MWCNTs were dispersed in 2-propanol through sonication process which act as an electrolyte in electrolytic cell. Electrolytic cell was designed by connecting two copper wires with plastic jar through holes (Fig. 5a). One wire was connected to the positive terminal of battery and second wire was connected to the negative terminal of the battery which serves as anode and cathode respectively. DC voltage of 20V, 40V, 60V, 80V and 100V was applied and the movement of MWCNTs in electrolyte was observed under light microscope (Optika, Italy) at 100x magnification as shown Fig. 5b. Similar process was employed to assess the effect of applied electric field on the behavior of p-MWCNTs in electrolyte. C. Optical microscopy Optical microscopy was also carried out in order to see the dispersion state of MWCNTs in epoxy matrix. A drop of f- MWCNTs filled epoxy matrix was taken on slide, gently pressed by placing another slide on it and observed under optical microscope. Slides were then placed in oven and allowed to cure at prescribed curing cycle. Slides were again 194 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 204. Fig. 5. Pictureof (a) Electrolytic cell (b) In-situ observation of MWCNTs under electric field through light microscope. examined after curing under the microscope to witness the effect of curing cycle on dispersion state. Similar process was carried out with p-MWCNTs containing epoxy resin. D. DC electrical conductivity measurements DC electrical conductivity of fabricated samples (both f- MWCNTs/CF and p-MWCNTs/CF) was measured at room temperature in accordance with ASTM standard D4496-98 and compared with the results of neat CF reinforced polymer composite. Standard four point probe method was used to determine the surface and volume conductivity via four Kelvin pins and LCR meter. E. EMI shielding effectiveness (EMI SE) test EMI SE test was carried out in Anechoic chamber using calibrated Vector Network Analyzer (Rohde & Schwarz) over the frequency range of 7-14 GHz. Attenuation was measured by recording S21 parameter (return loss) between source and receiving antenna inside the anechoic chamber without any sample. Then test sample was placed between the antennas (source & receiving) and attenuation was again measured by noting down the value of S21 parameter. The difference between S21 values of sample and without sample gives the value of attenuation (i.e. EMI SE) caused by the sample. Attenuation is described in terms of shielding effectiveness (SE), measured in decibel (db) and shows the reflective nature of the sample. Testing was performed on equally dimensional samples (150 x 75 x 2 mm) of f-MWCNTS/CF reinforce polymer composite and p-MWCNTs/CF reinforced polymer composite. IV. RESULTS AND DISCUSSION A. FE-SEM Figure 6a and 6b shows the high resolution SEM images of f-MWCNTs and p-MWCNTs respectively. Opening of end caps and shortening of length was evident in case of f- MWCNTs as shown in Fig. 6(a) while Fig. 6(b) demonstrates that p-MWCNTs have long length and damage-free structure. High aspect ratio and defect free structure of nanotubes is prime requirement for exhibiting good electrical conductivity. Morphological study through FE-SEM confirms that the aryl diazonium salt functionlization has destroyed the physical structure and decreased the aspect ratio of nanotubes, which in turn decreased the electrical conductivity of MWCNTs. B. In-situ observation of MWCNTs under applied electric field When DC electric field was applied up to 80 V no significant effect was observed on MWCNTs under microscope. However at 100 V fraction of p-MEWCNTs moved from cathode to anode and network of nanotubes is formed between electrodes. Under applied electric field MWCNTs experience force due to polarization effect and move from cathode to anode get discharged and absorbed on the anode. Then adsorbed MWCNTs become the source of adsorption of further nanotubes because of high electric field at the tip which is connected to anode. In this way, conductive pathway of p-MWCNTs is formed from anode to cathode showing the high conducting nature of p-MWCNTs as shown in Fig. 7 (a). In case of f-MWCNTs most of the nanotubes under applied electric field attached to the anode without development of conductive pathway because of imbalance polarization effect due to structural damage during fucntionalization process. Similar effect was also observed with acid functionlized MWCNTs under applied electric field by other researchers [20]. Transfer of electric current is not possible between the electrodes without formation of conductive path suggesting the poor electrical conductivity of f-MWCNTs. Arrows head show the absence of conductive pathway in Fig. 7 (b). C. Optical microscopy Figure 8 (a, b) shows the optical images of 0.2 wt % f- MWCNTs and p-MWCNTs filled epoxy matrix in uncured and cured state. Functionalized MWCNTs were more uniformly distributed in the matrix and less amount of sub- micron size aggregates were present as compared to p- MWCNTs-epoxy matrix in uncured state. f-MWCNTs were dispersed homogenously in matrix during sonication process due to the bonding between epoxy and functional group grafted on MWCNTs. However, in case of p-MWCNTs homogenous distribution was difficult in matrix because of 195 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 205. Fig. 6. FE-SEMimages of (a) f-MWCNTs and (b) p-MWCNTs. high electrostatic forces present between MWCNTs. Similarly, in cured state amount of aggregates in f-MWCNTs was lowered than that of p-MWCNTs in epoxy matrix. Since curing is a lengthy polymerization process, p-MWCNTs find time to get re-entangled due to vander walls forces before their Fig. 7. Optical images showing the comparison of conductive pathway formation of (a) p-MWCNTs (b) f-MWCNTS motion get seize during polymerization. However, f- MWCNTs remained suspended in the matrix and very few aggregates formed during curing stage because bonding between epoxy and functional group resist re-agglomeration and coagulation of f-MWCNTs. D. Eelectrical conductivity Conventional polymers are usually electrical insulators and do not fulfill the requirements of EMI and ESD characteristics. Polymers are made conductive generally by addition of conductive fillers to protect against EM waves and electrostatic charge (ESC). Fig. 9 shows the bar graph of volume and surface electrical conductivity of neat CF reinforced polymer composite, p-MWCNT/CF reinforced polymer composite and f-MWCNT/CF reinforced polymer composite. Conductivity of neat CF reinforced polymer composite is due to the conductive nature of carbon fabric. Figure 9 illustrates that the volume as well as surface conductivity both significantly increased by the addition of 0.2 wt% of p-MWCNTs. This increment in conductivity was attributed to high aspect ratio and formation of 3d conductive network of MWCNTs inside the composite. However, addition of 0.2 wt% f-MWCNTs has not increased the conductivity of polymer composite up to appreciable extent. It was concluded from the conductivity measurements that the functionalization process has in fact adversely affected the conductive nature of MWCNTs by shorting the length and creating defects in the structure of MWCNTs which in turn ruined the conductivity of composite sample. E. EMI-SE Electromagnetic radiations can be attenuated by three basic mechanisms: reflection, absorption and multiple reflections. Figure 10 exhibits the SE of f-MWCNTs/CF and p- MWCNTs/CF reinforced polymer composites against EM waves at different frequencies. EMI SE is sensitive to 196 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 206. Fig. 8. Opticalimages showing the dispersion of (a) f-MWCNTs and (b) p- MWCNTs in uncured and cured state of polymer matrix. Fig. 9. Volume and surface dc electrical conductivity of composite samples. frequency and dependent on the type of MWCNTs addition. Absorption and multiple reflections play an important role in shielding against EM waves but dominating mechanism at lower frequency is reflection which is directly related to electrical conductivity of the sample [21]. Layered structure of composite is responsible for the lost of EM waves at the interface between carbon fabric and the matrix through multiple reflection phenomena. And some part of EM waves was absorbed by MWCNTs but mainly EM waves were attenuated by reflection mechanism. As shown in Fig. 10 that the p-MWCNTs composite demonstrated overall better EMI SE which is governed by high electrical conductivity of p- MWCNTs. For reflection of EM waves, 3D conductive pathway is required which is better provided by p-MWCNTs than f-MWCNTs. Maximum SE was observed by p-MWCNTs sample at frequency range of 7-9 GHz with a peak about 70 db at a 7.75 GHz frequency. However, in the frequency range of 10-11.5 GHz f-MWCNTs composite performed similar way as p-MWCNTs sample and maximum SE about 46 db was observed at 11.4 GHz in case of f-MWCNTs sample. Minimum attenuation of about 14.8 db was observed at 13.7 GHz by p-MWCNTs sample and in case of f-MWCNTs minimum attenuation of 9.4 db was attained at the frequency of 13.5 GHz. From the test results, it can be inferred that p- MWCNTs composite performed better than f-MWCNTs in shielding EM waves in the whole tested frequency range with the maxima at around 70 db. V. CONCLUSION Effect of functionalization by aryl diazonium salt on electrical properties of MWCNTs was assessed through applied electric field test in electrolytic cell, electrical conductivity measurement and EMI SE test. Following conclusions were drawn on the basis of test results: SEM microscopy showed that the functionalization process has damaged the physical structure of nanotubes. It was found from the optical microscopy that the dispersion of f-MWCNTs in the host matrix was improved. Amount of aggregates of MWCNTs increased upon curing in both cases (p-MWCNTs & f-MWCNTs) but f- MWCNTs showed more resistance to agglomerate formation during curing stage. It was deduced that high aspect ratio and defect-free structure of MWCNTs is essential for formation of conductive pathway under applied electric field. Pristine MWCNTs aligned themselves in a regular fashion under applied electric field whereas the f-MWCNTs do not respond appreciably to the applied field. DC electrical conductivity measurements showed that the conductivity of f-MWCNTs was decreased as a result of functionalization. The reason for deterioration in electrical conductivity was attributed to the introduction of defects in the MWCNT sidewalls and shortening of length. Pristine MWCNTs based composite sample outclassed the f-MWCNTs composite in EMI SE testing over the tested frequency range of 7-14 GHz. 3D conductive pathway formation of MWCNTs inside the matrix is the prime requirement for exhibiting good electrical properties and this in turn drastically affected by the choice of functionalization procedure. 197 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 207. Fig. 10. Comparisonof shielding effectiveness of f-MWCNTs/CF and p-MWCNTs/CF reinforced polymer composite. ACKNOWLEDGMENT Authors would like to thank Mr. Ahmed Bilal (Chairman, SUPARCO) for approval and provision of facilities. The authors also gratefully acknowledge Mr. Wasim Nawaz for performing EMI testing and Ms Noureen Owais for providing assistance in electrical conductivity measurements of the samples. REFERENCES [1] Z.L. Wang, W. H. Manu and Dechang Li, “Structure and growth of CNT films by Pyrolysis,” Cheml Phys Lett, V. 316(5,6) , pp-349, 2000. [2] Ajayan PM, Zhou OZ, “Applications of carbon nanotubes. Carbon Nanotubes,”. pp.391-425, 200. [3] R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, C. Xu, Y. Hee Lee, S. G. Ki, A. G. Rinzler, D. T. Colbert, G. E. Scuseria, D. Tomanek, and R, E. Smalley, Thesis, “Crystalline Ropes of Mettaic Carbon Nanotubes,” Science Vol. 273. No. 5274, pp. 483-487, 1996. [4] Prabhakar R, Bandaru “Electrical Properties and Applications of Carbon Nanotube Structures,” Reviews journal of Nanoscience and Nanotechnology Vol.7, pp.1–29, 2007. [5] Christopher Kingston, Richard Zepp, Anthony Andrady, et al. “Review- Release characteristics of selected carbon nanotube polymer composites,”CARBON Vol. 68 pp.33–57, 2014. [6] Sandler JKW, Kirk JE, Kinloch IA, Shaffer MSP, Windle AH, “Ultra- lowelectrical percolation threshold in carbon-nanotube-epoxy composites.” Polymer, 2003, Vol. 44, pp.5893-9, 2003. [7] Baughman R. H, Zakhidov A. A, de Heer, W. A, “Carbon nanotubes-the route toward applications,” Science, Vol. 297, pp. 787-792, 2002. ISSN 0036-8075 [8] Bauhofer, W., & Kovacs, J. Z, “A review and analysis of electrical percolation in carbon nanotube polymer composites,” Compos. Sci. Technol., Vol. 69, pp. 1486-1498,ISSN 0266- 3538 [9] Singh BP, Choudhary V, Saini P, Mathur RB, “Designing of epoxy composites reinforced with carbon nanotubes grown carbon fiber fabric for improved electromagnetic interference shielding,” AIP Adv 2012;2:6 [10] Mathur RB, Singh B.P, Tiwari P.K, Gupta T.K, Choudhary V “ Enhancement in the thermomechanical properties of carbon fibre-carbon nanotubes-epoxy hybrid composites,” International Journal of Nanotechnology, Vol. 9, pp.1040-1049, 2012. [11] Huang Y, Li N, Ma Y, Feng D, Li F, He X, et al, “The influence of single-walled carbon nanotube structure on the electromagnetic interference shielding efficiency of its epoxy composites,” Carbon,Vol.45, pp.1614-21, 2007. [12] Li N, Huang Y, Du F, He XB, Lin X, Gao HJ, et al. “Electromagnetic interference (EMI) shielding of single- walled carbon nanotube epoxy composites,” Nano Letters, Vol.6, pp.1141-5, 2006. [13] Kannan Balasubramanian and Marko Burghard, “Chemically Functionalized Carbon Nanotubes reviews,” Small, Vol. 1, No. 2, pp. 180–92, 2005. [14] A.KANAPITSAS, C.TSONOS, “Study of electrical / dielectric and thermomechanical properties of polymer–carbon nanotubes nanocomposites,” In Proceedings of the 1st WSEAS International Conference on NANOTECHNOLOGY. [15] R.B. Mathu r, Shaila ja Pande, B.P. Sing h, T.L. Dhami, “Electrical and Mechanical Properties of Multi-Walled Carbon Nanotubes Reinforced PMMA and PS Composites,” POLYMER COMPOSITES-2008 [16] C.VELASCO-SANTOS, A.L.MARTINEZ-HERNANDEZ and V. M. CASTANO, “Review Carbon nanotube-polymernanocomposites: The role of interfaces Composite Interfaces,” Vol. 11, No. 8-9, pp. 567–586, 2005. [17] Najiba Abdullah Al-Hamdani, “Preparation and Electrical Properties of Epoxy Resin Reinforced with Functionalized Carbon Nanotubes,” IOSR Journal of Applied Physics (IOSR-JAP) Volume 6, Issue 4 Ver. I, pp. 54-56, Jul-Aug. 2014 [18] Jeffrey L. Bahr, Jiping Yang, Dmitry V. Kosynkin, Michael J. Bronikowski, Richard E. Smalley, and James M. Tour, “ Functionalization of Carbon Nanotubes by Electrochemical Reduction of 198 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 208. Aryl Diazonium Salts:A Bucky Pap,” J. Am. Chem. Soc,Vol. 123, 6536-6542, 2001. [19] Anastasia A. Golosova, Christine M. Papadakis and Rainer Jordan “Chemical functionalization of carbon nanotubes with aryl diazonium salts” Mater. Res. Soc. Symp. Proc. Vol. 1362 © 2011 Materials Research Society DOI: 10.1557/opl.2011.1141 [20] Abu Bakar Sulong, Nurhamidi Muhamad, Jaafar Sahari, Rizauddin Ramli, Baba Md Deros, Joohyuk Park “Electrical Conductivity Behaviour of Chemical Functionalized MWCNTs Epoxy Nanocomposites” European Journal of Scientific Research ISSN 1450- 216X Vol.29 No.1, pp.13-21, 2009. [21] Simon Rea,David Linton, Eddie Orr, Jonathan McConnell“Electromagnetic Shielding Properties of Carbon Fibre Composites in Avionic Systems,”Microwave Review, June, 2005 199 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 209. Design of Cband Slotted Waveguide Antenna with High Impedance Bandwidth and Improved Reflection Coefficient for SAR applications Sania Nazir1 , Masab Iqbal2 , A.K.M. Shafaat Ali3 Satellite Research and Development Center, Space and Upper Atmosphere Research Commission, Karachi, Pakistan sania_tl@yahoo.com, masabiqbal@gmail.com,shafaat_32@yahoo.co.uk Abstract—With the advancement of technology, synthetic aperture radar (SAR) has acquired great importance in the field of remote sensing. For optimum performance of SAR systems, selection and design of antenna choice is very critical. In certain applications, specially for airborne purpose, SAR antenna should be efficient, conformal, medium to high gain, capable of providing wide beam in range while narrow beam in azimuth direction and either mechanically or electronically steerable. These performance parameters can be met with slotted waveguide antennas and are an ideal choice for radar and Airborne SAR applications. Slotted waveguide arrays can be either standing wave (resonant) or travelling wave (non- resonant). In this paper, a slotted waveguide antenna which resonates at C band is presented. This paper discusses the design of a bottom fed linear resonant slotted wave guide antenna array with longitudinal shunt slot elements. This design is optimized using advanced EM simulation software in order to achieve improved impedance bandwidth and good reflection coefficient. This improvement is achieved by introducing non-uniform inter element center to center spacing and slot center to short end spacing as odd multiple of guide wavelength (λg/4). Reflection coefficient value at the desired frequency comes out to be < -35dB with an impedance bandwidth of > 100 MHz. Measurement results are in conformance with the simulated results with a gain of approximately 20 dB. Keywords: Slotted Waveguide, reflection coefficient, impedance bandwidth. I. INTRODUCTION In this advanced era of remote sensing and imaging, synthetic aperture radar (SAR) has attained great significance due to its all weather, day/night imaging capability. It is basically a technique of generating the effect of long antenna with physically small antenna. It consists of long array of successive and coherent radar signals initially transmitted and then received by physically small antenna every time when it moves along a flight path. Thus antenna plays an important role in the overall performance of SAR system. For SAR applications, the antenna should have following characteristics: Efficiency: It is essential for SAR antenna that the received echo signal must be strong enough to provide required signal to noise ratio(SNR). For receiving echo signals with adequate SNR the antenna efficiency must be higher. High gain: In SAR systems, specially pulsed SAR systems whether space borne or air borne, it is required to transmit very high power short duration pulses to achieve high effective isotropic radiated power(EIRP). These high power pulses are generated by either TWTA or SSPA. If transmitter is not capable of generating very high peak power then required EIRP can be achieved by using high gain antenna. Wide beam in Range and narrow beam in Azimuth: SAR is a two dimension imaging system, along track direction (azimuth) and cross track direction (range). The swath size which is covered by the system is specified in terms of minimum and maximum look angles in range direction. It can be stated that range foot print defines the swath size to be covered on ground. Hence the beam must be wide in the range direction to meet the requirement of higher swath width. Mechanical or electronic steering: SAR can be operated in several imaging modes such as strip map, spot light and scan SAR. Among these modes, strip map mode has fixed antenna beam in azimuth as well as range direction. While for spot light and scan SAR which provide high resolution and larger swath width respectively antenna beam must be either mechanical or electronic steerable to achieve high resolution and swath width especially in space borne SAR systems.In this paper, a resonant C band slotted waveguide antenna is presented used for Strip Map Air borne SAR applications. II. ANTENNA DESIGN The proposed antenna is a slotted waveguide antenna working at C band. Operating frequency range is in between is 5.2 to 5.4 GHZ. This slotted waveguide antenna is resonant, linear antenna array with longitudinal shunt slot elements cut along the length of waveguide. Waveguide end is short 200 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 210. circuited which inturn converts waveguide into a resonant standing wave structure. Thus the antenna beam is fixed and antenna array is non- scanning. Resonant arrays can be either end fed or center fed. Proposed antenna is end fed where one end of waveguide is fed while other end is terminated in short. The antenna configuration is shown in figure 1. In this configuration the some of the most important parameters which affect the performance of antenna are; (a) number of slots which determine gain (b) slot length which determines the resonant frequency (c) inter slot spacing and slot location in the waveguide. For resonant arrays it is recommended that the inter element center to center spacing must be uniform, which is equal to half of guide wavelength (λg/2) and the distance from short wall of waveguide to the center of last slot should be equal to quarter of guide wavelength (λg/4). Where λg is the guide wavelength and is given by Eq1. λg = λ / λc 2 -λ2 (1) Where, λc=2a= Cut off wavelength a= Longest dimension of rectangular waveguide λ= Operating wavelength This design is optimized for achieving higher impedance bandwidth and reflection coefficient with non-uniform spacing and last slot center to short end spacing as odd multiple of quarter guide wavelength (λg /4). Slots along the length of waveguide are modeled in such a way that the center to center slot spacing between consecutive slots is (λg), while the center to center inter element spacing between alternate slots is lesser and higher than half of guide wavelength. Last slot center to short end spacing is either 3λg /4 or 5λg /4 or higher instead of λg /4 in order to achieve higher reflection coefficient and impedance bandwidth. In the proposed design, non-uniform spacing between the last slot center to short end spacing is 5λg /4 and the length of waveguide is 826 mm with 22 longitudinal shunt slot elements for achieving gain of approximately 20 dB. Figure 1 Basic structure of resonant slotted waveguide antenna array with shunt slot element III. SIMULATION RESULTS The proposed antenna is modeled, simulated and analyzed in Advance EM Simulation software. Isometric view of the modeled structure is shown in figure 2. After optimization, this antenna provides a gain of 20 dB. The simulated gain plots are shown in figure 3 and 4. Figure 2 HFSS model of proposed antenna Figure 3 Simulated gain plot Figure 4 Simulated 3D radiation plot 201 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 211. Figure 5 SimulatedReflection coefficient (S11) plot The simulated reflection coefficient is less than -40 dB at desired frequency while the impedance bandwidth is greater than 100 MHz. Figure 5 shows the reflection coefficient plot. IV. MEASUREMENT RESULTS The proposed antenna is fabricated on wire cut machine in order to have the slots fabricated with precision along with maintaining very strict inter element spacing. The fabricated antenna is shown in figure 6. The measured reflection coefficient plot is shown in figure 7. Measured reflection coefficient at desired frequency is -43 dB while impedance bandwidth is 145 MHz. Antenna radiation pattern measurement was carried out in Anechoic Chamber Test facility. Measured azimuth and elevation cuts are shown in figure 8 and 9 the measured gain at our desired frequency is about 20 dB. Figure 6 Fabricated slotted waveguide antenna array Figure 7 Measured Reflection coefficient Figure 8 Measured Azimuth cut Figure 9 Measured Elevation Cut V. CONCLUSION A resonant C band slotted waveguide antenna is designed, fabricated and tested as per design requirement. Gain of the antenna is approximately 20 dB with -43 dB reflection coefficient value. Impedance bandwidth achieved is 145 MHz for VSWR < 2. VI. ACKNOWLEDGMENT First I would like to thank Almighty for all His blessings and providing me the strength for the successful completion of this task. I would also like to thank SUPARCO for providing me the opportunity and necessary support and guidance from the team members for accomplishing this task. VII. REFERENCES [1] Antenna engineering Handbook Richard Clayton Johnson. [2] Waveguide slot antenna array by Ronald A Gilbert BAE systems. [3] W1 GHZ Microwave Antenna Book Paul Wade W1GHZ. [4] Resonant length calculation and radiation pattern synthesis of longitudinal slot antenna in rectangular waveguide By M. Mondal, A. Chakrabarty Progress In Electromagnetics Research Letters, Vol. 3, 187–195, 2008. [5] A Design of Slotted Waveguide Antenna Array Operated at X band Kuo-Lun Hung, His Tseng Chou. Wireless Information Technology and Systems (ICWITS), 2010 IEEE International Conference. 202 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 212. Risk Area Mappingand Identification of Hotspots on the Road-Network of Lahore Spatial Pattern and comparision of Road Traffic Accident with population and commercial area of district Lahore. Uzair ul Hassan Shah GIS Developer The Urban Unit Lahore, Pakistan Uzairshah.gis@gmail.com Muhammad Nauman Hussain GIS Research Associate The Urban Unit Lahore, Pakistan Nauman.gis11@gmail.com Abstract— if we talk about big cities first thing that comes to our mind is crowded roads and accidents. We daily read it in newspaper that many people died or injured were being rescued by the Rescue teams but these type of news are becoming part of our life, in other words it’s not strange for us anymore. Ratio of accidents is increasing day by day because of many reasons like over speeding, inexperience, intake of drugs, mobile phone use, and negligence of drivers and to avoid the use of seat belts. The current study is an effort of road accidents investigation in Lahore district. For this purpose, road accident data of March, 2014, Lahore district is used that was stored by Rescue 1122. The density function available in the spatial analysis tool of the Arc Map 10.2 software was applied to identify the accident hotspots. Both simple and kernel density tools were applied for the assessment of risk areas on major roads of Lahore. Spatial analyses for finding out risk areas of these road segments. Furthermore a relation is drawn between the highly populated area and the accident density. Also a comparison is made among the road accidents and the commercials areas present in the district. Based on the result, spatial distribution of Risks along the Major road network of Lahore is marked on the map and necessary measures are proposed to reduce the future accidents. This geo-spatial data is vastly helpful for better planning of combating strategies against road accidents. Keywords— Road Traffic Accidents, Road Risk Assessment, GIS I. INTRODUCTION Transport system plays major role in the growth and development of a nation to make safe transportation of persons and goods from one location to another. Fatality/injury ratio in road accidents in major cities of the world is alarmingly high. Road accidents occasionally cause big tragedies by killing a lot of people. There are many causes behind these accidents. Persons travelling on the busy road are highly exposed to risk of accident. Risk mapping has been carried out based on the collected geo- tagged points and analyzed quality data of road traffic accidents for managing further activities aimed at improvement of traffic safety. A significant precaution to save lives from the event of road accident is provision of emergency/traffic control services at appropriate distances. Limitations that explain the poor outcome for people involved in road traffic accident in Pakistan have been identified as inappropriate dedicated transportation and traffic density is most in urbanized cities. Road traffic accident should be examined also from spatial dimension by using GIS techniques. GIS (Geographical Information System) can be put in effective use for accident analysis. GIS technology is used for managing locational and relational information and it is able to manipulate and visually display trends & data required for easy comparison and decision making regarding future planning and development of traffic system. Development of a GIS system to analyze the traffic accident has been pursed towards improving the effectiveness and efficiency of traffic accident. II. STUDY AREA Lahore is the capital of Punjab Province and is the most urbanized and thickly populated district in Punjab province. According to the 2013 census (Punjab Development Statistic 2013 Bureau of Statistic, Lahore) has 12,218,345 population with 81.17% urban portion. The total area is 1,772 square kilometer. Nearby Districts are Kasur, Nankana Sahib and Sheikhpura. Major Roads in the District Boundary are National Highway (N5), Ferozpur road, Grand Trank Road, Ring Road, Canal Road, Mall Road, Jail Road, Multan Road, Allama Iqbal Road, Raiwind Road, other primary and secondary roads with different lengths,widths and Traffic Densities. 203 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 213. III. METHODS A. DataCollection In order to identify the accident locations in Lahore district, following data were collected and used. RTA (Road Traffic Accident) location data were collected from the control room of Rescue 1122 (Emergency service) in Lahore. Accident Report for the month of march 2014 The analysis result is based on the available data collected from Rescue 1122 databases of district Lahore. The Population data (in the form of landscan image) is gathered from The Urban Unit. That raster is further clipped according to the study area. The commercial areas are identified by the Urban Unit. The quality of data determines largely the quality of research result. B. Digitizing Digitizing is the process of encoding the geographical features in digital form as X, Y coordinates on the geo- referenced satellite image of district Lahore. Road network is digitized as line feature and accident location as point feature in Arc Map 10.1. The exact location of accident is identified by using online Google maps, Bing maps and distance from the relative location (Shops, Petrol pumps and traffic signals etc.) by using distance measuring tool in Arc Map 10.1. The maps showing the accident location in the figure II. C. Assigning Attributes All vector data (i.e. line, point feature) had attributes. Here roads were labeled with road name and total accidents on the road segment that is from one intersection to another by spatial join tool in Arc Map 10.1 point feature contain following attributes: Date of Accident occurrence. Exact Time of occurrence. Day of occurrence. Place of accidents. D. Statisctis In the present study a general statistical analysis was carried out in terms of accident occurrence on the primary and secondary roads of Lahore at different times, locations and causes of accidents. There are total 4528 accident cases in rescue 1122database from March 01, 2014 to March 31, 2014.Among 4528 accidents, 718 pedestrian, 1315 passengers, 1998 Drivers and 325 under age drivers became victims measuring 15.9%, 29.0%, 44.1% & 7.20% respectively of total reported accidents. TABLE 1: VICTIM DETAIL Victim Detail Accidents % Pedestrian 718 15.9 Passenger 1315 29.0 Driver 1998 44.1 Under Age Driver 325 7.20 In 4528 accidents: 2231 bikes, 365 cars, 14 busses, 39 trucks, 177 vans, 479 rickshaws and 277 other vehicles were involved in the events, measuring 49.20%, 8.0%, 0.3%, 0.8%, 3.0%, 10.5% & 6.10% respectively of total reported accidents. TABLE 2: VEHICLE DETAIL Vehicle Detail Accidents % Bikes 2231 49.2 Cars 365 8.0 Busses 14 0.3 Trucks 39 0.8 Vans 177 3.0 Rickshaws 479 10.5 Others 277 6.1 Distribution of injuries as reported in the rescue database: 23 spinal injury cases, 335 head injury, 202 leg fractures, and 130 multiple fractures and 3341 minor injuries. FIGURE I- THE URBAN AREA OF DISTRICT LAHORE WITH MAJOR ROADS FIGURE II- MAP SHOWING THE ACCIDENT LOCATIONS IN DISTRICT LAHORE DURING MARCH 2014 204 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 214. FIGURE III-:RTA INMARCH 2014 TABLE 3: DISTRIBUTION OF RTA INJURY Injury Victim % Spinal Injury 23 0.5 Head Injury 335 7.3 Leg Fracture 202 4.4 Multiple Fracture 130 2.8 Minor Injury 3341 73.7 Causes of Road Traffic accidents reported in the database: 1756 over speed cases, 734 carelessness, 485 wrong turn, 328 U turn, 1 one wheeling, and 4 tyre brust, 274 others accidents causes. TABLE 4: DISTRIBUTION OF REASON CAUSE ACCIDENTS Causes of RTA Victim % Over speed 1756 38.7 Carelessness 734 16.21 Wrong Turn 485 10.7 U Turn 328 7.2 One Wheeling 1 0.2 Tyre brust 4 0.8 Others 1220 26.9 Patient condition on the spot: 16 patient died on spot 2761 victim were alive and unstable, 1756 alive and stable. TABLE 5: DISTRIBUTION OF VICTIM CONDITION ON SPOT Condition Victim % Dead 16 0.35 Alive& unstable 2761 60.97 Alive & Stable 1751 38.67 Age of victims effected by the road traffic accident: 153 cases of age 1 ~ 10Y, 854 were 11 ~ 20Y, 1389 were 21 ~ 30Y, 692 were 31 ~ 40Y, 485 were 41 ~ 50Y, 284 were 51 ~ 60Y, 174 were above 60Years. TABLE 6: DISTRIBUTION OF VICTIM AGE Age Victim % 1Y-10Y 153 3.79 11Y-20Y 854 21.18 21Y-30Y 1389 34.45 31Y-40Y 692 17.16 41Y-50Y 485 12.03 51Y-60Y 284 7.04 Above 60Y 174 4.31 IV. RESULT AND ANALYSIS Compiling the data and performing statistical analysis shows the high level of traffic accidents in this major city of Pakistan. These statistical results are represented in the form of charts are given below. Above charts show the RTA take place in March 2014. It illustrates the randomness of the accidents in Lahore city. But the charts did not explain the spatial feature of accidents i.e. where was that accident happened. To show the location of RTA, maps are produced which are displayed as follow. 205 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 215. FIGURE VII- RELATIONBETWEEN RTA AND POPULATION Kernel density is applied to get the strength of traffic accident per unit area. In kernel density all those points are considered in calculating density which are fall with in the neighborhood of a particular cell. If no point fall within the neighborhood of a particular cell, that cell is assigned no data. The map shows spatial distribution of RTA w.r.t days of a week. This shows the venerable areas of accident during different days of a week which intends helpful for the traffic authorities mitigate the venerability of accidents. The above figure contain 4 maps demonstrate different time zones of a day. A day is divided in to four time zones i.e. from 12AM to 6AM, 6AM to 12PM, 12PM to 6 PM, 6PM to 12AM. Each map shows the density of accidents in the specific time zone. The above map is the comparison of accidents and population density. The population is represented in a tone of red bright color which turns darker when we move towards the denser population. The accidents density is also shown in atone changing from green to red (red as most venerable area). The comparison shows that most of the accidents occurs in highly populated areas. Similarly accident strength is high on major/main roads (in blue color lines) of the city. FIGURE IVI- KERNEL DENSITY MAP FIGURE IV- RTA SPATIAL DISTRIBUTION W.R.T DAYS FIGURE IVI- RTA DENSITY DURING DIFFERENT TIME ZONES OF A DAY FIGURE VIII- RTA COMPARISON W.R.T COMMERTIAL AREA FIGURE IV-RTA DENSITY MAP 206 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 216. For Cities likeLahore having major trading points may have risk of more RTA. To understand this problem above map illustrate the RTA comparison with commercial area. The map clearly shows that the majority of accident happened near commercial areas. Another picture of same issue reflect the comparison of RTA density with commercial points (Figure XI). This map demonstrates that red zones for accidents are nearby commercial areas. V. CONCLUSION AND RECOMMENDATIONS This work gives an insight of the present scenario of the traffic condition in Lahore and shows out the most accident prone roads. The result shows that total road traffic accidents, dense population areas, commercial areas are the important variables to take into the consideration in examining traffic accidents. Spatial data can be shared with transportation, emergency services and other government organizations. This spatial database can be helpful for expert system to make new traffic plans and rules. It also provides accurate recommendations to vehicle drivers, the police and to local monitoring authorities. In order to curtail road traffic accident on Lahore roads the following recommendations are necessary: Drivers and bikers should be trained as the mean of effective dealing with road accidents. Traffic rules must be deployed by the law enforcement agencies in the district and country. Road safety education and seminars should be held by the transport departments. Drivers should drive within speed limit and with a speed unswerving with road condition. There must have proper parking plazas in all major commercial zones and markets. VI. ACKNOWLEDGMENT This study is sponsored by The Urban Unit (Urban Sector Planning and Management Service Unit USPMSU). The authors would like to thanks Mr Niazm-ud-din (GIS Manager at Urban Unit) for providing data and Mr. Ehsan Saqib (Sr. GIS Development Specialist), Dr. Ather Ashraf (Sr. GIS Specialist) for helping in different analysis. VII. REFERENCES [1] Bureau of Statistic Bos, “Punjab Development Statistics 2013”,www.bos.gop.pk. [2] Deepthi Jayan.K, B. Ganeshkumar,”Identification of Accident Hotspot: A GIS Based Implementation for Kannur District, Kerala” , vol. 1. ISSN 0976-4380. [3] Dr, A.J.Aderamo, “Spatial Pattern of Road Traffic Accident Casualties in Nigeria,”, vol. III, ISSN 2039-2117. [4] Harnam Singh,A.D Aggarwal, “Fatal Road Traffic Accident among Young Children,” J Indian Acad Forensic Med, 32(4), ISSN 0971-0973. [5] The Urban Unit, USPMSU Lahore, www.urbanunit.gov.pk FIGURE IX- RTA DENSITY COMPARISON WRT COMMERCIAL AREA 207 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 217. ANALYSIS OF RECENTDROUGHT BASED ON NDVI AND METEOROLOGICAL DATA A case study of Tharparkar district, Sindh Muhammad Arslan1 , Rao Zahid2 , Badar Ghauri3 Department of Remote Sensing and GISc Institute of Space Technology Karachi, Pakistan marslanhafeez100@gmail.com Abstract— Thar region of Pakistan has often undergone short and long period of drought. In this study, Satellite Remote Sensing and meteorological data were used to detect the drought. Agricultural drought was identified based on the pre and post monsoon Normalized Difference Vegetation Index (NDVI) with 250meter spatial resolution from MODIS terra satellite from 2008 to 2014. Departure from the mean values of pre and post monsoon NDVI was categorized to determine the agricultural drought. Whereas meteorological drought was identified based on the annual precipitation data for the same period provided by the Pakistan Meteorological Department (PMD). Pre and post monsoon mean NDVI for the year 2014 was then compared with the previous years. Pre monsoon mean NDVI values of 2008 to 2013 was 0.115956, while post monsoon mean NDVI values of 2008 to 2013 were around 0.221801. On the other hand, pre and post mean NDVI for the year 2014 were observed as 0.127617 and 0.156314 respectively. Mean post-monsoon NDVI value drastically dropped in the year 2014. Satellite based mean NDVI, PMD annual rainfall data and field observation showed that a short term drought occurred in year 2014. This was also confirmed by the local population during the field visit. Keywords— Drought, MODIS, NDVI, precipitation. I. INTRODUCTION Drought is classified into four classes. Hydrological drought, meteorological drought, agricultural drought and socio-economic drought. No universal definition of drought is still known. Drought is defined as a period of abnormally dry weather which eventually affects in a change of vegetation of an area [1]. Drought is also defined as it usually occurs when the rainfall or runoff is found below a mean truncation level, derived from the long term rainfall series [2]. For human society and ecosystem drought is a significant adverse climatic events. Since 1850, the mean global surface air temperature has increased by 0.76o C [3], by the end of the twenty first century, it is expected that the temperature will increase by 1.5 o C – 6.4 o C [4]. Drought may increase under the warm climate [5]. The frequency and the intensity of drought have increased in past decades in many regions of the world [6, 7]. This recurring of drought has negative impact in these areas in term of annual loss of vegetation and food. Drought monitoring has gained importance in many countries especially developed countries due to the serious social, environmental and economic ramifications [8]. Conventionally, drought monitoring was based on meteorological station observations that obviously does not have a continuous spatial coverage for the monitoring of comprehensive drought. Most of the studies have used satellite data to monitor hazards including, drought, flood, earth quake and so on [9, 10]. Satellite data have frequently been used to monitor and detect the drought phenomenon [1]. Satellite data provides earth observation for monitoring the impacts of drought in a precise and effective way. There exists a time lag between precipitation events and its response on the vegetation. The time gap between the occurrence of precipitation and the time precipitation water reaches the plant roots. It varies from 01 to 12 weeks depending on vegetation type [11]. There are various studies that have focused on NDVI and precipitation relationship. This relationship varies spatially, when the NDVI values changes with respect to precipitation [12]. Spatial distribution and natural condition of vegetation are the function of environmental variables such as precipitation. Strong relationship is occurred between NDVI and rainfall [11]. Precise and consistent vegetation mapping is essential and necessary for the management of water resources and mitigation of drought events. A. Objectives to identify the NDVI and meteorological based drought. to compare 2014 year recent drought with the previous years to evaluate the effects of drought in term of socio- economic condition of the local people II. MATERIALS AND METHODS A. Study Area Tharparkar district of Sindh province consists of deserted land and its area is 19,638 square kilometers. Administratively this district is divided into four talukas. These are Diplo, Nagarparkar, Mithi and Chachro shown in Fig1. It has 208 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 218. population of 9,14000 and density of 46.17 per square kilometer. Figure 1. Map of the study area B. Data Sources Satellite data included Moderate Resolution Imaging Spectro-radiometer (MODIS) 16 days composite of Normalized Difference Vegetation Index (NDVI) from 2008 to 2014. Meteorological annual rainfall data of the duration of Mithi observatory, Tharparkar was obtained from Pakistan Meteorological Department (PMD). Side by side a field survey was also undertaken in order to observe the drought condition in this area. TABLE 1. DATA USED AND ITS SOURCES Data Description Sources MODIS Product NDVI MODIS Satellite Rainfall Data Mithi Observatory PMD Study area Condition Agriculture and Livestock Field Observation C. Methodology Following were the main steps taken to accomplish the research study and as illustrated in Fig.2. After having acquired all the necessary data (described above), pre-processing of the data was done. Mean NDVI of the pre and post monsoon seasons were used for the study. Information on fodder and trend of agriculture activities collected from the field of Tharparkar district on February 2015. MODIS mean NDVI, field observations and PMD annual rainfall data were integrated in order to analyze the drought severity condition specifically for the year 2014. Figure 2. Methodological framework chart III. RESULTS AND DISCUSSIONS Pre monsoon mean NDVI value of 2008 to 2013 was 0.115956, while post monsoon mean NDVI of 2008 to 2013 was 0.221801. On the other hand, pre and post-monsoon mean NDVI for the year 2014 was observed as 0.127617 and 0.156314 respectively. Post-monsoon mean NDVI value dropped drastically in 2014 compared to post-monsoon mean NDVI value for the period 2008-2013 shown in fig.3 and 4. Annual rainfall (PMD data) also dropped drastically in 2012, 2013 and 2014 as 227mm, 189.1mm and 193 mm respectively compared to 2011 which was 1310 mm (shown in fig.5). In 2014, annual rainfall increased a little by 4 mm, but still this indicated drought in Tharparkar district. Lack of rainfall caused loss of natural vegetation. Quantitatively, livestock directly depends on the natural vegetation for fodder (as shown in fig.6). In Tharparkar, agriculture activity mainly depends on the mercy of rainfall. During the field survey carried out, no agriculture activity was observed because drought was the main reason affecting local agriculture (shown in fig.7). Locals were waiting for the rainfall and they left their ploughed land in expectation of rain. The satellite NDVI, PMD annual rainfall and field based observations showed that a short term drought might have occurred due to the lack of rainfall compared to 2011 and a loss of natural vegetation in post monsoon season of 2014. The 2014 drought was also confirmed by the local population during field visits shown in Fig 8(d). Most of the vegetation cover reduced and thus developing a seasonal drought 2014 in Tharparkar. Based on satellite, field and weather data, 2014 drought had occurred in Tharparkar District, Sindh. Figure 3. Post monsoon mean NDVI (2012-2014) Figure 4. Pre and post monsoon mean NDVI values (2008-2014) 2008 20102009 2011 20132012 2014 209 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 219. Figure 5. Annualrainfall provided by PMD (2009-2014) Figure 6. Means of food for the livestock Figure 7. Reduction in agriculture Figure 8. (a and b) Livestock of the area, (c) How local people collect water (d) Interview from the local people related to drought (e and f) Showing local people in the picture IV. CONCLUSION Drought is a recurring phenomenon and it can be examined by satellites. Long term satellite data could help the severity and onset of drought due to its spatial and temporal variability. Effective and efficient drought mitigation could be achieved through a well-defined drought monitoring system. ACKNOWLEDGMENT This research is part of a project named “Development of Decision Support System for Drought Monitoring in Sindh” funded by the Pakistan Higher Education Commission (HEC). The authors acknowledged the financial grant of HEC. Thanks are also due to Pakistan Meteorological Department (PMD) for provision of data. We also thank NASA site for giving data. REFERENCES [1] Kogan, Felix N. "Global drought watch from space." Bulletin of the American Meteorological Society 78.4 (1997): 621-636. [2] Dracup, John A., Kil Seong Lee, and Edwin G. Paulson. "On the statistical characteristics of drought events." Water resources research 16.2 (1980): 289-296. [3] Jones, P. D., Trenberth, K., Ambenje, P., Bojariu, R., Easterling, D., Klein, T., ... & Zhai, P. (2007). Observations: surface and atmospheric climate change. In Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 235-336. [4] Solomon, Susan, ed. Climate change 2007-the physical science basis: Working group I contribution to the fourth assessment report of the IPCC. Vol. 4. Cambridge University Press, 2007. [5] Pachauri, Rajendra K., and A. Reisinger. "Synthesis Report." IPCC. Ginebra (2007). [6] Hulme, Mike, and Mick Kelly. "Exploring the links between desertification and climate change." Environment: Science and Policy for Sustainable Development 35.6 (1993): 4-45. [7] McCarthy, James J. Climate change 2001: impacts, adaptation, and vulnerability: contribution of Working Group II to the third assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press, 2001. [8] M Amin, Owrangi, et al. "Drought monitoring methodology based on AVHRR images and SPOT vegetation maps." Journal of water resource and protection 2011 (2011). [9] Anderson, James Richard. A land use and land cover classification system for use with remote sensor data. Vol. 964. US Government Printing Office, 1976. [10] Peters, Albert J., et al. "Drought monitoring with NDVI-based standardized vegetation index." Photogrammetric engineering and remote sensing 68.1 (2002): 71-75. [11] Chopra, Parual. "Drought risk assessment using remote sensing and GIS: a case study of Gujarat." ME thesis, Indian Institute of Remote Sensing, Dehradun & International Institute for Geo-information & Earth Observation, the Netherlands (2006). [12] Nicholson, S. E., and T. J. Farrar. "The influence of soil type on the relationships between NDVI, rainfall, and soil moisture in semiarid Botswana. I. NDVI response to rainfall." Remote Sensing of Environment 50.2 (1994): 107-120. 210 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 220. Assessment of UrbanGrowth of Karachi: From A Tiny Town to A Meta City of the World Ali Jan Hassan1 , Mudassar H. Arsalan2 , Hira Fatima Institute of Space and Planetary Astrophysics University of Karachi Karachi, Pakistan 1 Alijan.shaaan@gmail.com 2 mharsalan@uok.edu.pk Abstract— Karachi is the only meta city of Pakistan and capital of Sindh. It plays a key role in the country’s economy by contributing 65% of total revenue. However the sustainability of this city of lights is at risk due to mismanagement and development without proper long term planning. The increased urbanization has brought several severe issues such as increased demographic density, management challenges, resource deficit and ecological pressures, etc. The sprawl of the city is uncontrolled spreading of urban developments. Remote sensing data and GIS are now widely being used for change detection analysis and applications. Urban sprawl is one of them, which can be analyzed efficiently and cost effectively by utilizing satellite data available online in the form of present as well as historical images in different spatial resolutions. In this study change detection technique was used to assess the urban sprawl through the integration of historical maps with Landsat-5 TM sensor image of 1989 and Landsat-8 OLI images of 2014. With the aid of these images we monitored, measured and analyzed the urban sprawl of Karachi. Despite of many unsustainable conditions in the city, its population growth is tremendous i.e. approximately 4 % per annum. This growth is the result of the high natural increase as well as substantial migrations from other parts of the country. It is assessed that the living population in the mid 2014 was around 21 million, which was far greater than of 1998 i.e. 9.2 million. In the last 25 years, Karachi has extended around 15 percent at the rate of 2 square kilometers per year to accommodate its increasing population with considerable high density. Index Terms— Urban sprawl, Change detection, Sustainability, Satellite imagery, GIS, RS. I. INTRODUCTION Sustainability is not a new concept, it was discussed, analyzed and even practiced at all levels, from country to individual. The term ‘Sustainable Urban Developed’ (SUD) emerged from the Brundtland Report [1], according to this report, SUD is defined as “Development that meets the needs of today without compromising the needs of the future.” Many organizations such as UN-Habitat, WHO, EEA, etc., started to work on SUD indicators. Indicators for sustainable development provide valuable information, indicating rather a development is on the way of sustainability or degradation [2]. Planners, developers and stakeholders use the SUD indicators to better understand the current development situation as well as to predict the future development scenario in context of sustainability[3] Urban Sprawl is one of the main indicators of SUD, that can determine the increment area of Urban land, as well as the densification of the city [4] Urbanization and Urban Sprawl are two different terminologies. Urbanization is a process of increasing number of people that migrate from rural to urban areas, results in the physical growth of urban areas regardless of developments occur horizontally or vertically [5], whereas Urban Sprawl is defined as the expansion of urban area in the form of new development on isolated tracts, separated from other areas by vacant land [6]. The socioeconomic conditions of any place are the driving forces for urbanization hence it occurs more tremendously in areas that are economically well established [7]. Urbanization is an inevitable process due to economic development and rapid population growth [8] In the last part of the 20th century and the beginning of the 21st century human population has increased tremendously. On average, at every second 4 or 5 children are born, while 1 or 2 people dies, this rate of birth and death means a net gain of 2.5 persons (on average) on the Earth [9]. Unfortunately, this increase has been occurring in the developing part of the world, where lack of facilities, poverty, pollution and crime rate are at its peak [9]. Pakistan is in the list of developing countries and is badly affected by terrorism, energy crises and law and order situation in the country [11]. It is an atomic power, bearing world’s toughest army, having a vast amount of resources of coal, oil and other minerals [12]. But due to miss-management, irregularities and irresponsibleness of bureaucracy, stakeholders and political leaders, the complete potential of the resources is not being properly utilized [13]. Pakistan is the world’s sixth most populous country today [14] With an estimated population of 188 million, and 2.05 percent annual 211 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 221. growth rate, Pakistanis expected to become the fifth largely populated nation by 2050 [14]. Table 1. World’s Most Populous Cities [14] 2014 2015 Country population (Millions) Country population (Millions) China 1368 India 1628 India 1226 China 1437 USA 320 USA 420 Indonesia 255 Nigeria 299 Brazil 203 Pakistan 295 Pakistan 188 Indonesia 285 A. Study Area Today Karachi is the largest and most populous city of Pakistan, lies in the south bounded by the Arabian Sea. It covers a total area of 3600 square kilometers that includes Built-up land, Vegetation Coverage, Industrial built-up, commercial areas and a lot of recreational areas. The Karachi port serves as a gateway for all the import/export activity by sea. The industrial and economic activity in Karachi support Pakistan a 65% of revenue [15]. Karachi appeared as a mega- city with the beginning of this century with an exponential growth in its population. The residents of Karachi are blessed with the sea breeze that blows throughout the year. Because it is close to the Sea the temperature remains normal, the mean annual temperature is in between, the wind normally blows towards North-East. There are many commercial centers exist in the city, which enabled the it to develop as an important industrial and economic center that attract the people not only from Pakistan but all over the world. B. Objectives The Objective of this study was to determine the urban footprint for the year 1989 and 2014 using remote sensing data and GIS, to figure out the extension of Built-up land within 25 years (1989-2010). II. MATERIALS AND METHODS The simple methodology adopted for this study comprises of three steps: Data collection, data processing and urban sprawl analysis. The brief account of these steps is given below. A. Data Collection This study is an application of remote sensing and GIS in Urban sprawl analysis. As recent as well as historical remote sensing data is freely available on the internet. Further, using GIS on this Remote sensing data different analysis can be done efficiently and cost effectively. For this study Landsat 5 TM sensor clod free image of Karachi (Path-Row 152-42) captured in November 1989 and Landsat 8 OLI sensor image captured in January 2014 were acquired from the Landsat archives available at the USGS website. B. Data Processing The pre-processing was carried out for the development of AOI. Downloaded images of the study area were geometrically corrected and projected to Universal Transverse Mercator (UTM) projection system, zone 32. The spheroid and datum were referenced to WGS1984. C. Methodological Framework for Urban Sprawl The change detection technique was adopted to assess the change in land use and land cover (LULC) of the study area. Collected and processed Landsat 5 image of 1989 and Landsat 8 image of 2014 were used while the main focus was on built- Fig.1. Population growth in developing and developed nations, 1750 – 2100 [10] Fig. 2. Study Area- Karachi, Pakistan 212 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 222. up area change.To extract the Urban built up the more accurately supervised maximum likelihood classification was done. The Maximum Likelihood Classification is one of the supervised classification approaches for pattern recognition in which the probability of a pixel belonging to each of a predefined set of classes is calculated, then pixel is assigned to the class for which the probability is the highest [16]. ML is based on the Bayesian probability formula: P(x, w) = P(w|x) P(x) = P(x|w) P(w) Where x and w are called events, P(x, w) is the probability of coexistence of events x and w, while P(x) and P(w) are the prior probabilities of events x and w and P(w|x) is the conditional probability of event x given event w [16]. By carefully observing the patterns in the images, it was concluded that the classification should be done for 4 classes i.e. Urban Land, (including dense, green as well as sparse Urban areas), Vegetation (including dense and sparse vegetation as the study area had a vast mangrove forest towards its south), Water bodies and Open/barren land. Maximum Likelihood Classification was done on natural color composite image (Band Combination 3, 2, 1) by training 100 samples of four classes distributed randomly in the image. Classes were identified by visual interpretation. From this a thematic map of four classes was generated. The same procedure was applied to the other image. 1) Accuracy Assessment To check the level of confidence in the classification results an error matrix is usually generated. This is some time called the confusion matrix, in which a matrix presents the relationship between the classes in the reference maps and classified images. Estimating classification accuracy enables us to determine classification performance [16], but there is a lack of information, because it leads to overall accuracy of the classes not of individual class. To assess the accuracy of individual class the concept of Producer’s and User’s accuracy was used in this paper. A multivariate index, Kappa Coefficient was also used to assess the difference between actual agreement and change agreement [16]. III. RESULTS Due to the highly favorable geographical location, Karachi has always been a trade route for the merchants of India, but it gained its importance in the British rule. The development of perennial irrigation schemes by the British in Punjab and Sindh mainly accounted for the growth of Karachi along with that the development of railways, making the quick transportation of the agricultural material that in turn increased in agricultural production which was exported through Karachi which made it the largest exporter of wheat and cotton to India [17]. In 1935, when Sindh separated from Bombay, Karachi became the capital of Sindh, Government offices and trade organizations were shifted from Bombay to Karachi. In 1947, Pakistan got independence and Karachi became first capital Pakistan. Table 2 Karachi Population Growth [18] Year Population Relative change No. of year in between two consecutive census Population change (%) 1901 136,297 136,297 10 1911 186,771 323,068 10 37 1921 244,162 430,933 10 30.2 1931 300,779 544,941 10 23.2 1941 435,887 135,108 10 44.9 1951 1,137,667 701,780 10 161 1961 2,044,044 906,377 10 79.67 1972 3,606,746 1,562,702 11 76.5 1981 5,437,984 1,831,238 9 50.8 1998 9,802,134 4,540,422 17 86.29 2001 13,500,000 3,697,866 3 37.7 2005 15,119,000 1,619,000 4 11.9 2014* 20,228,112 5,109,112 9 25.5 From the below graph we can see that the population of Karachi was 9.2 million in 1998, there is a tremendous increase of around 12 million in 2014[18]. Fig. 3. Methodological Framework of Urban Sprawl Fig. 4. The Population Trend of Karachi 1901-2014 213 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 223. There was asignificant increase of 161% from 1947-1951, because of the migration of 600,000 refugees from India. The settlement of those refugees was a challenge for newly built local government. The refugees started to live on the periphery and within the city itself, as a result the environment of Karachi totally changed not only demographically but also ethnically [17]. The irresponsibilities of higher authorities and lack of planning and strategy the refugees unwillingly started to live in squatter settlements and katchi abadis. After 1958, five master plans have been developed for the development of the city, but the implementation of those plans has always been an issue because of political instability within the country. The Karachi City, administratively is spread over an area of approximately 3,600 sq. km. More than half of this area (approximately 2455.5 acres) consists of vacant land. The land- cover features such as water and vegetation, changes throughout the year because it depends on the Climatic conditions at a particular time. The Urban Built-up area has extended from 343 square km in 1989 to 396 square km in 2014 that is a valuable change of 53 km in 25 years. This reason behind this rapid increase in population within the city is the migration from Interior Sindh, Punjab and Khyber Pakhtoonkhuwa to find better opportunities in the mega city. There is a 100 square km decrease in open land from which 50 sq. km land is converted into built-up land The actual built-up land is 396 square km but as per document figures by the City District Government of Karachi (CDGK) [19], the built-up land was 1300 square km, it was a combined built-up and developed open spaces. Developed open spaces are those housing societies which were initiated in 1970s, 80s, and 90s, however, even after their infrastructure developed they are vacant and hardly 5% occupied as shown in table 3. Table 3 Occupancy Status of New Housing Scheme, [19] Serial. No. Name of Scheme Year of Notification Current Occupancy Status (%) 1 Scheme no. 25- A 1980 5% 2 Scheme no. 33 1971 20% 3 Scheme no. 42 1983 5% 4 Scheme – 43 1986 0% 5 Scheme – 45 1986 5% From the table below, we can see that there is a 15 % increase in built-up land and 3% decrease in open land. This growth is mainly in built-up area and decrease in open land. Table 4 Area Cover by each class for the years 1989 and 2014 in Sq. km Serial No. Class Area (1989) Sq. km Area (2014) Sq. km Change Sq. km Change (%) 1 Built-up 343 396 53 15 2 Open Land 2556 2455 -100 -3 3 Vegetation 748 849 -100 13 4 Water 1014 972 -41 -4 Table 5 Error Matrix of Land Cover (Karachi - 1989) Class Built- up Open Land Veget ation Wat er Row Sum User's Accura cy Prod ucer's Accu racy Built- up 14 0 1 0 15 93.33 93.33 Open Land 1 45 9 0 55 81.82 100 Veget ation 0 0 10 0 10 100 45.45 Water 0 0 2 18 20 90 100 Colu mn Sum 15 45 22 18 100 Overall Accuracy 87% Kappa 0.328 Fig. 8. Land Cover Change of Karachi (1989-2014) Fig. 9. Land Cover change % of Karachi (1989 - 2014) 214 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 224. Table 6 ErrorMatrix of Land Cover (Karachi - 1989) Class Built- up Open Land Veget ation Wate r Row Sum User's Accurac y Produ cer's Accur acy Built- up 8 4 0 0 12 66.67 88.89 Open Land 1 44 5 2 52 84.62 88 Vegeta tion 0 0 9 0 9 100 64.29 Water 0 2 0 25 27 92.59 92.59 Colum n Sum 9 50 14 27 100 Overall Accuracy 86% Kappa 0.3563 IV. DISCUSSION AND CONCLUSION This study reveals that according to CDGK (2007) the population of the city reaches to around 21 million and if we consider the whole of Karachi area that is 3,600 km, the population density is 6000 persons per square kilometer about 25% of its population has increased in last 10 years. However, the Urban built-up area has extended from 343 square km in 1989 to 396 square km in 2014 that is a change of 53 square km in 25 years. It shows the average annual growth rate is slightly above 2 square kilometers per year. If we consider average population density is 50,000 per square kilometer, the city can accommodate only 0.1 million people every year. However, it is not so, as the average population increase is 5 times higher than its accommodating capacity. It indicates the pressure of population is more towards the existing neighborhood in the form of densification and slum area's growth. That results in depletion of resources, degradation in environmental conditions and lack of recreational areas within the city center. All these factors produce bad impacts on social, environmental and economic conditions of the city and makes difficult to achieve sustainable development. This study was conducted to assess the conditions of Urban Sprawl in Karachi that is an effective indicator of SUD, however many other indicators can also be identified and assessed, by implementing the techniques of RS and GIS. Fig. 10. Growth of Karachi 1938-1992 [20] 215 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 225. Fig. 10. LandCover Map of Karachi 1989 Fig. 11. Land Cover Map of 2014 216 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 226. REFERENCES [1] World Commission,"Report of the World Commission on Environment and Development: Our Common Future Acronyms and Note on Terminology Chairman’s Foreword," ed: Oxford University Press, 1987. [2] M. Javadian, H. Shamskooshki, and M. Momeni, "Application of sustainable urban development in environmental suitability analysis of educational land use by using AHP and GIS in Tehran," Procedia Engineering, vol. 21, pp. 72-80, 2011. [3] N. Chrysoulakis, C. Feigenwinter, D. Triantakonstantis, I. Penyevskiy, A. Tal, E. Parlow, et al., "A Conceptual List of Indicators for Urban Planning and Management Based on Earth Observation," ISPRS International Journal of Geo-Information, vol. 3, pp. 980-1002, 2014. [4] Sustainable Cities International, Indicators for sustainability- How cities are monitoring Sustainability, 2012. [5] UN, "United Nations Conference on Environment & Development, Rio de Janerio, Brazil, 3 to 14 June 1992," 1992. [6] J. I. IGBOKWE, I. C. EZEOMEDO, and E. J, "Investigation of Urban Sprawl Using Remote Sensing and GIS: A case of Onitsha and its Environments in Southeast, Nigeria," International Journal of Remote Sensing & Geoscience (IJRSG), vol. 2, 2013. [7] Boundless. (2015, July). The Process of Urbanization. Available: https://www.boundless.com/sociology/textbooks/boundless- sociology-textbook/population-and-urbanization- 17/urbanization-and-the-development-of-cities-123/the-process- of-urbanization-695-3433/ [8] S. Deep and A. Saklani, "Urban sprawl modeling using cellular automata," The Egyptian Journal of Remote Sensing and Space Science, vol. 17, pp. 179-187, 2014. [9] W. P. Cunningham and M. A. Cunningham, Principles of environmental science: inquiry & applications: McGraw-Hill, 2011. [10] D. B. Botkin, E. A. Keller, and D. B. Rosenthal, Environmental science: Wiley, 2012. [11] I. Rana, "Foreigners driven away by terrorism, energy crisis," in The Express Tribune, ed: @etribune, 2011. [12] H. Naz, Environment of Pakistan, 2006. [13] I. A. Sodhar. (2011, July). Pakistan Rich in Natural Resources But Poor in their Management. Available: http://jworldtimes.com/Article/92011_Pakistan_Rich_in_Natura l_Resources_But_Poor_in_their_Management [14] U. S. C. Bureau. (2011, July). International Programs, Country Rank. Available: https://www.census.gov/population/international/data/countryra nk/rank.php [15] A. Hassan. (2015, July). Ishaq Dar’s budget defies economic vision: Karachi Chamber of Commerce and Industry. Available: http://asianetpakistan.com/business-news/general-business- news/203382/ishaq-dars-budget-defies-economic-vision- karachi-chamber-of-commerce-and-industry-3/ [16] P. Mather and B. Tso, Classification methods for remotely sensed data: CRC press, 2009. [17] A. Hasan and M. Mohib, "The case of Karachi, Pakistan," 2003. [18] Government of Pakistan, "Population Census Reports." [19] CDGK, "Karachi Strategic Development Plan 2020," 2007 [20] M. S. Faruqui, Karachi, Physical Situation of Human Settlements: Master Plan and Environmental Control Department, Karachi Development Authority, 1982. 217 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 227. JUPITER, THE GASGIANT ABDUR RAQEEB GAZIANI, POSTER COMPETITION Jupiter is the first of gas planets in our solar system. It travel around sun once every 11.86years. Jupiter is so large that over 1300 earths can be packed in its volume. Here we will discuss some bold points of Jupiter: THE SURFACE: Is mostly of gaseous hydrogen and the layer is about 20,000 km thick. Jupiter is one of the brightest plant in the sky with the great red spot on surface. Convection currents and the Jupiter’s rapid rotation create the bright zones and dark belts. MOONS OF JUPITER There are at least 63 natural satellites or moons orbit Jupiter. The four major moons of Jupiter: o Io: Io is about the same size as Earth’s moon. It orbits Jupiter once every 1.77 days. Io has nine giant erupting volcanoes on its surface and up to 200 smaller volcanoes. There are also mountains up to 10 km high but these are not volcanic.Io is thought to be volcanically active because of huge high tidal forces created by Jupiter. o Europa: Europa is the fourth largest moon of Jupiter. It orbits Jupiter once 3.55 days Europa has a thin outer layer of ice and a layered internal structure, probably with a metallic core. Its surface is relatively smooth, with no mountains and very few craters. Some astronomers think that a layer of liquid water may be present below the ice-covered surface. o Ganymede: Ganymede is the largest moon in solar system. Its diameter is largest than the planet mercury. It orbits Jupiter in synchronous rotation once every 7.16 days at a distance of about one million km.The crust is thought to be 75 km thick and contain an outer layer of ice. The Hubble space telescope has recently found evidence oxygen on Ganymede. o Callisto: It is second largest moon orbiting Jupiter. It is just slightly smaller than planet Mercury, but has only one-third its mass.Callisto orbits Jupiter once every 16.69 days. Its surface is a dark, ancients and icy crust, covered with many old impact craters. It probably cooled very rapidly. Voyager’s instrument measured a temperature range from -180ºC during day time to - 315ºC at night. It has little internal structure, composed of about 40% ice and 60% rocky iron. MAGNETIC FIELD: Jupiter’s magnetic field is about 20 times stronger than Earth’s. Its magnetosphere extends only a few million kilometres towards sun, but it extend more 650 million kilometres away from sun. Jupiter’s magnetosphere is pushed by the solar winds. The magnetic poles are offset from the rotational axis by nearly 10º. DENSITY AND COMPOSITION: Jupiter is more than 318 times more massive than the Earth and has twice as much mass as all the other planets combined. Jupiter’s average density is lower than that of Earth –only 1.33g/cm³ compared to Earth’s 5.52g/cm³. 218 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 228. Identification of PostDisaster Scenario Using Double Threshold Energy Detection Shakeel Ahmed Waqas* , Fahad Shamshad Electrical Engineering Department, Military College of Signals, National University of Science and Technology, Islamabad, Pakistan * Shakeel_aw@yahoo.com Nisar ali Department of Telecommunication Engineering, University of Engineering and Technology, Peshawar, Pakistan Abstract—The disaster can be naturally or manmade. Telecommunication cellular system can be completely destroyed by this disaster. The main priority is to provide connection to disaster areas before the rescue team are being arrived and save as much lives as possible. In this paper, we presented a new algorithm which enable mobile station to automatic detect the disaster scenario and switch to emergency mode. We used double threshold energy detection to reduce the false alarm. The performance of proposed algorithm is evaluated using Monte Carlo Simulation in term of probability of detecting post disaster scenario, probability of miss detection and probability of false alarm. Keywords— Post Disaster Communication, Emergency Mode, Normal Mode, Double Threshold, Energy Detection I. INTRODUCTION The telecommunication existing network can be damaged or destroyed by manmade or natural disaster. In result, that area will have no connection with national communication infrastructure. 50% death occurs in two hours after disaster [1]. The lives can be saved if there are some mean of communication before the arrival of rescue team. The survivors need urgent help after the disaster. Therefore, post disaster communication is required to connect the disaster area with whole country [1]. The post disaster communication can be achieved by using Ad-hoc network. Ad-hoc network is used by many researchers for emergency network. Wireless Emergency Network can be obtained by multi hop Ad-hoc network. In Wireless Emergency Network, there is ability of self-healing [2]. But in disaster scenario, the resources are limited. Ad-hoc network cannot reuse the available resources and also unstable. In [1], the authors proposed that using cluster based Ad-hoc network, the network become stable and resources can be reuse. Thus, it increase the capacity. In cluster based network, cluster head will forward the data of cluster members. In [2], the authors proposed route repairing for Wireless Emergency Network making it more stable, energy efficient and reliable. In [3], authors presented cross layer framework for the post disaster scenario. The authors presented an adaptive protocol selector which select appropriate protocol for each layer. Furthermore, the authors in [4] proposed three algorithms using single threshold based on cognitive radio, cellular network and artificial neural network to detect the post disaster scenario. But there is possibility of false alarm if the mobile is in deep fading area. The authors in [5], proposed double threshold energy detection algorithm for cognitive network and proved that false alarm can be reduced by using double threshold. In the paper, we presented a new algorithm using double threshold for detecting post disaster scenario. We used energy detection technique for the decision making. In this paper, we presented the performance of our proposed algorithm using BPSK modulation scheme over AWGN channel. The performance is analyzed in term of probability of detection post disaster scenario, probability of misdetection of post disaster scenario and probability of false alarm. All the results are evaluated using Monte Carlo simulation in MATLAB. II. SYSTEM MODEL The system investigated in this paper is shown in figure 1. Our system model consists of one mobile station and two base stations. The mobile station is denoted by MS and the two base stations are denoted by BS1 and BS2. MS is connected with BS1 and BS2 is adjacent base station. The distance between the MS and BS1 is denoted by d1. Similarly, the distance between the MS and BS2 is denoted by d2. In our system model, d1 is 2 km and d2 is 3 km. AWGN channel is considered between MS, BS1 and BS2. The path loss exponent is considered as n = 2. All the nodes (MS, BS1 and BS2) are mounted with single antenna. It is assumed that MS has already installed software based identification module. There are two modes on which 219 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 229. MS can operate.One is Emergency Mode and second is Normal Mode. In Normal Mode, the MS will operate normally Fig1. System Model and will be connected with its mobile service provider infrastructure recognizing itself as a part of cellular network. In Emergency Mode, the MS will stop broadcasting messages to BS and switch to Ad-hoc mode. Switching the MS from one mode to another, MS regularly sense the signal coming from connected base station BS1 and adjacent base station BS2. Here we used energy detection technique and double threshold to take decision and switch the MS to appropriate mode. The detected energy from BS1 is denoted by connectedV and the detected energy from BS2 is denoted by adjacentV . III. DETECTION ALGORITHM Our proposed algorithm uses two thresholds i.e. 0 and 1 . The decision making is totally depends on detected energies connectedV & adjacentV and values of threshold. Algorithm: 1. If the connectedV and adjacentV both are greater than 1 , then switch to Normal Mode. 2. If the connectedV and adjacentV both are less than 0 , then switch to Emergency Mode. 3. If the connectedV and adjacentV both are between 0 and 1 , then don’t take any decision. Scan again 4. If connectedV is greater than 1 and adjacentV is less than 0 , in this case, no decision would be taken. Scan again to take better decision 5. If connectedV is less than 0 and adjacentV is greater than 1 , again it will be uncertain Fig2. Double threshold Energy Detection algorithm for post disaster identification The advantage of our algorithm is that it reduce the false alarm. If the MS is in deep fading areas where the signals are weakly received, the MS will not switch to Emergency mode. According to our proposed algorithm, the MS cannot change its mode easily. It will change its mode when the energy of connected BS and adjacent BS both are higher than threshold 1 or lower than threshold 0 . IV. RESULTS AND DISCUSSION The system model shown in previous section is simulated using Monte Carlo simulation. Results are discussed in this section. The time bandwidth factor is taken as 2. Average SNR (signal to noise ratio) is 14 dB. The path loss is considered. The path loss exponent is 2. Modulation used for transmission is BPSK. The distance between MS and BS1 is 2 km and distance between MS and BS2 is 3 km. AWGN noise is considered in channel. Fig3. Probability of detection post disaster of scenario The figure3 show the probability of detecting post disaster scenario for different threshold 0 . As clear from simulation result that for 0 dB 0 , the probability of detecting post disaster scenario is zero. Means we MS will not switch to Emergency mode. As we increase the value of 0 , the Energy Emergency Mode Normal ModeUncertain 0 1 1BS 2BS 1d 2dAWGN AWGN MS 220 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 230. probability increases. Thepost disaster scenario can be easily detected if 0 is high. The probability of miss detection of post disaster scenario is shown in figure4. From the result, it is obvious that we will surely miss the detection of post disaster scenario if the threshold 1 is kept very low. Keeping the 1 low, then greater probability of switching to Normal Mode and low probability of switching to Emergency Mode. As the threshold 1 increases, then we can easily detect the post disaster scenario. Therefore, the probability of miss detection decreases as the threshold 1 increases. The probability of false alarm of our proposed algorithm is shown in figure5. To calculate false alarm we assume that SNR Fig4. Probability of miss detection of post disaster scenario Fig5. Probability of false alarm is 0 dB. According to our simulation result, it is stated that if the thresholds are low then the probability of false alarm increases. We get 100 % probability of false alarm at 0 dB threshold. As the threshold increases, the probability of false alarm decrease. We get 0% false alarm for threshold equal to or greater than 10 dB. V. CONCLUSION In this paper, we presented double threshold energy detection algorithm for detecting post disaster scenario. This algorithm enable mobile station to detect the post disaster efficiently and take right decision in switching its mode from normal to emergency or emergency to normal. Decision is made by considering the energy of signals received from connected Base station and from adjacent base station. The performance is presented in term of probability of detection of post disaster scenario, probability of miss detection of post disaster scenario and probability of false alarm. In future, we will work future on identification of post disaster scenario using double threshold considering the adjacent channels. REFERENCES [1] Sonia Majid and Kazi Ahmed, “Cluster-based Communications System for Immediate Post-disaster Scenario” Journal of Communication, Vol. 4, No. 5, June, 2009. [2] Xin Ma, Quan-yi Huang and Yan Kang, “A New Algorithm for Stable Communication in Wireless Emergency Network” International Conference on Computer and Management, pp. 1-4, May 2011, Wuhan, China. [3] Sonia Majid and Kazi Ahmed, “Cross-layer Framework for Post-disaster Communications” IEEE International Conference on Telecommunications and Malaysia International Conference on Communications, 14-17 May 2007, Penang, Malaysia [4] Sonia Majid and Kazi Ahmed, “CIP- Cognitive identification of post-disaster communication” IEEE conference on Telecommunications, pp. 43-47, June 2012, Thailand [5] Jinbo Wu, Tao Luo, Jianfeng Li and Guangxin Yue “A Cooperative Double-threshold Energy Detection Algorithm in Cognitive Radio Systems” IEEE conference on Wireless Communications, Networking and Mobile Computing, pp.1-4, Beijing, China. [6] Haitao Wang, Lihua Song, Jianzhou Li, Jiaxin Deng, “Wireless Self-organizing Emergency Communication Network based on Network Clustering and Information Ferry” , 2nd Internation Conference on consumer Electronics, Communications and Networks, pp. 3704- 3707, Yiching, China. [7] Sonia Majid and Kazi Ahmed, “Post-disaster Communications: A Cognitive Agent Approach” 7th International Conference on Networking, pp- 23-30 Cancun Mexico, April 2008. [8] Marco Di Felice, Luca Bedogni, Angelo Trotta, Luciano Bononi, Fabio Panzieri, Giuseppe Ruggeri, Gianluca Aloi, Valeria Loscr, Pasquale Pace, “Smartphones Like Stem Cells: Cooperation and Evolution for Emergency Communication in Post-Disaster Scenarios” First IEEE International Black Sea Conference on Communications and Networking 2013 (IEEE BlackSeaCom 2013), Batumi, Georgia. [9] Jinbo Wu, Tao Luo and Guangxin Yue, “An Energy Detection Algorithm Based on Double-threshold in Cognitive Radio Systems” The 1st International Conference on Information Science and Engineering (ICISE2009), pp. 493-496, Nanjing China. 221 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 231. Design and Analysisof Magnetic MEMS Accelerometer for Inertial Navigation M. Ameer Umar Malik, Umer Izhar, Umar Shahbaz, Javaid Iqbal Department of Mechatronics Engineering College of Electrical and Mechanical Engineering, NUST Islamabad, Pakistan ameerumar@yahoo.com, umerizhar@yahoo.com Abstract- Accelerometer is an important sensor of inertial navigation system. Focus of new developments is a low cost, small size sensor with high performance that meets the requirements of navigation. In this paper a multi wafer magnetic MEMS accelerometer design is proposed for inertial navigation. Capacitive pick off is used to sense deflection of sensor mass and permanent magnet & coil is used for rebalancing of deflected mass of the sensor. Total two sides capacitive area of sensor is 18 mm2 with gap of 8µm. Permanent magnet with thickness of 330 µm and diameter of 700 µm and soft magnet material coating is used. Out of plane coil with total 64 turns, 8 layers and cross- section area 20µm x20µm is used. Fabrication process outlines for both, the magnetic system and coils are proposed separately. Stress analysis for flexures of sensing pendulum has been carried out. Close loop frequency response analysis is carried out with a proposed PID controller. Key Words- Magnetic MEMS, Accelerometer I. INTRODUCTION Accelerometers are the important sensors of inertial navigation systems. Traditional high accuracy electromechanical accelerometers with quartz pendulum are highly accurate with resolution of <1µg and one year repeatability of ≤ 50 µg with input range of ± 50 g but also are very expensive [1]. Whereas the best available MEMS devices with input range of ±10g have resolution of up to 250 µg and max repeatability of 25mg [2] but cost of MEMS sensors is very low as compared to traditional electromechanical accelerometers. Although accuracy of MEMS accelerometers is yet not comparable to quartz flexure accelerometers but due to their mass production process these provide very economical solution particularly for low accuracy navigation applications. MEMS accelerometers are widely being used in automobiles but their accuracy is not enough for high accuracy inertial navigation use. Variety of MEMS accelerometers are being developed and reported for navigation applications. Most of these are capacitive devices with electrostatic rebalancing mechanisms. Recently reported ultra precision MEMS accelerometer developed by Colibrys is also a capacitive device working in close loop based on electro static rebalancing [3-4]. In this paper a magnetic rebalancing mechanism for a MEMS accelerometer is designed to get navigation grade performance. Permanent magnets are being used in MEMS for their repeatable performance. As the techniques of permanent magnets fabrication in MEMS are being developed their use is also increasing in high performance devices [5]. II. NAVIGATION GRADE ACCELEROMETERS High performance accelerometers are required for high accuracy navigation applications. Different performance parameters can be specified for inertial navigation grade accelerometers. In broad terms accelerometer performance can be specified by two parameters i.e. bias and scale factor [1-6]. Accelerometers for navigation require high level of bias stability, repeatability, temperature sensitivity and fine resolution. Requirements of sensor parameters and comparison for different requirements are discussed in [7]. Some of critical parameters required for high accuracy navigation application are given in table-1. A variety of transduction mechanisms have been used in MEMS accelerometers. Initially piezo-resistors were used in MEMS accelerometers [7]. Then other principles like capacitive, tunneling current, thermal, optical, electromagnetic and piezoelectric etc are also utilized with their some advantages and unavoidable disadvantages [8-9]. Capacitive based sensing gives greater design flexibility than piezoelectric quartz technology [6]. The capacitive devices are very popular for their high precision. TABLE:1 INERTIAL NAVIGATION REQUIREMENTS Parameters Values Range ± 25 g Bias Stability ≤50 µg Scale Factor Stability ≤ 50 ppm Resolution ≤ 5µg Frequency Range DC-200 Hz Max. Shock in 1msec > 20 g Temperature Range -40°C to 80 °C III. PERMANENT MAGNETS IN MEMS Use of electrostatic force is very common in microelectronic systems due to its better force scaling at micro scale level and also technology is highly developed as compared to electromagnetic systems at micro level.[10] Scaling differences of electrostatic and electromagnetic forces 222 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 232. is discussed indetails by Edward P Furlani [5]. Magnetic field from electromagnet unfavorably scales with diminishing sizing, whereas the fields from permanent magnet are scale invariant. Source of magnetic field in electromagnet is current but the magnetic field in permanent magnet is residual magnetization which arises at atomic level [5]. For actuation electromagnetic is more stable for high force and for large gaps [11]. Use of permanent magnets was limited in MEMS due to difficulties of their fabrication at micro scale level but now it is increasing as different fabrication technologies are being reported. Larger actuation distance can be achieved by using magnets as compared to electrostatic actuation[5,12,13]. Permanent magnet is used in traditional navigation grade electromechanical accelerometers. In MEMS accelerometer to get µg resolution and low noise of accelerometer a heavier proof mass is recommended and to rebalance this proof mass an electromagnetic system with a permanent magnet can be used. IV. PROPOSED DESIGN Accelerometer can be constructed and operated in open loop or closed loop configuration. The basic construction of accelerometer is such that a proof mass is suspended in case and confined to a zero position by means of a spring, when the acceleration is applied to the case of sensor, proof mass is deflected with respect to its null position within the case, this deflection of proof mass is proportional to specific applied force along input axis. More accurate output can be obtained by applying a rebalancing mechanism for deflected proof mass [14]. This rebalancing mechanism can be either electromagnetic or electrostatic. In traditional electromechanical accelerometers the rebalance mechanism is electromagnetic while in MEMS accelerometers it is electrostatic. A permanent magnet based MEMS accelerometer design is proposed and analyzed. It is a three wafer design in which the centre wafer with proof mass is sandwiched between two symmetrical permanent magnet wafers. The bench mark of the design is to use permanent magnet for rebalancing of proof mass of MEMS sensor instead of electrostatic rebalancing. Permanent magnet with volume of 0.7mm3 has been selected to get bulk fabrication with powder size of 5~ 50 µm and this can be accommodated in 500 µm thick wafer. Proof mass includes the rebalance coils and capacitance area for pickoffs. The designed proof mass has total area of 16 mm2 with thickness of 500 µm and coils mass is additional. This heavy proof mass results in a low Brownian noise [15] and greater stability and repeatability. The proof mass is suspended by two flexure beams attached to the fixed frame of centre wafer. In response to the applied input acceleration the proof mass deflects and change in capacitance is detected. One side capacitance area for pickoff is 9 mm2 which is large as compared to in plane MEMS accelerometers, this result in fine pickoff resolution. Rebalance force is applied by the passing current through the coils present on the proof mass. This current is the output of the accelerometer giving information about the acceleration applied to the system. This can be modeled as a second order mass damping system. Proof mass inertia is ‘J’, beams have effective stiffness ‘K’ and damping is ‘D’. Transfer function is given as H(s) = Where ‘a’ is applied acceleration and ‘Θ’ is angular deflection of proof mass. Fig.1(a) Centre wafer with proof mass. (b) Three wafer device For an applied input acceleration the total force depends on the total proof mass of pendulum part. This input force is to be balanced by coils. Lorentz force of the coils with current flowing through is the rebalance force. The Lorentz force on a conductor of length ‘L’ and with the current ‘i’ in an externally generated magnetic flux density ‘B’ perpendicular to the wire is given by F = Li × B. Rebalance force is directly proportional to the magnetic field available in the working gap of coil. To get maximum force magnetic field and coil placement need to be optimized. Few analytical methods to get axial force between a cylindrical magnet and a thick coil are discussed in ref [16]. To get maximum magnetic field in working gap of coil a magnetic loop is designed. Also to get maximum force, magnetic field direction should be perpendicular to the current flowing through the coil. Different configurations of magnetic system with permanent magnet and soft magnet materials are proposed, performance of each is evaluated to select the best possible configuration. Magnetic flux density plot of selected configuration is shown in in figure-2. Field variation in two dimensions is simulated to get optimized location of coil. Variation of flux density along z direction is plotted in figure-3. The resolution of 10 µg is selected as a target. Stiffness hinges are designed so that the smallest input of 10 µg produces a deflection of 4.33e-7 rad. This minimum deflection is used to estimate the overall gain required for pickoff system. Overall pickoff gain of 2.32e6 is used and gain values of different portions of pickoff system are calculated. Details are discussed in closed loop analysis of system. All the parameter values are calculated for this proposed design. Stiffness - K= 2.32 g. mm / rad. Damping - D = 1.16 g-s-mm. Inertia - J = 0.0000307 g-mm-s2 223 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 233. Fig.2 Surface Plotof Magnetic Flux Density V. FABRICATION PROCESSES Design of this sensor system consists of three wafers as shown in figure-4. Three wafers will be manufactured separately and then bonded together to form a sensor. Two wafers with permanent magnet are symmetrical. Coil is fabricated on both sides of the third wafer. This third wafer is to be fixed between two wafers with permanent magnet. Fabrication processes for both magnet and proof mass coil are proposed. Permanent magnet is the key part in this design and fabrication of permanent magnets has always been challenging in bulk batch fabrication of MEMS. Permanent magnets are produced by metallurgy process of mechanically pressing and heating. But this process is yet not compatible with MEMS production processes. Fig.3 Magnetic Flux density along z direction In MEMS conventionally permanent magnets are fabricated by vapor deposition methods like sputtering, evaporation, pulse laser deposition or by electrochemical deposition. By these methods a layer of only few micrometers can be deposited. Unconventional strategies like using magnetic powders of 5~50 micrometer sizes are being adopted to fabricate magnets of required size at wafer levels. Permanent magnets fabrication using magnetic powder and fabrication at wafer level is a choice for this design. Fig.4 Three Wafers of Magnetic MEMS accelerometer Details and performance of this method is discussed in details by Ololade D Oniku in ref [12]. Conventional deposition of micro magnets with results is described in [13, 17]. Nd-Fe-B powder of size 15~ 50 µm can be used to get the size required in this design and this can provide remanence Br of 0.5 ~ 0.7 Tesla. Doctor’s blade technique as discussed in ref [12] can be easily adopted at wafer fabrication for this design. For fabrication of required size permanent magnet powder fabrication is the best suited technique available. Fabrication steps for complete magnetic system are proposed as Cavity etching as per dimensions of magnet and working clearance using DRIE process. Coating of soft magnet material with thickness of 30 ~ 40 µm. Fill sacrificial material and form a cavity for permanent magnet powder. Permanent magnet fabrication in cavity using powder. Coating of soft magnet material with thickness of 30 ~ 40 µm above magnet. Removal of sacrificial material to form working gap for coil. Fig.5 Permanent Magnet Fabrication (a) Cavity etching for soft magnet coating. (b) Coating of soft magnet material. (c) Cavity forming for permanent magnet. (d) Permanent magnet fabrication using powder. (e) Coating of soft magnet material above magnet. (f) Removal of sacrificial material layer to form working gap. Fabrication of rebalance coil in multi layers is required on the centre wafer. Fabrications of spiral coils have been reported and one such multi layer coil with 36 turns in two layers has 224 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 234. been reported byChong H Ahn, and Mark G Allen [18]. The rebalance torquer coil required in this design is of 64 turns with 8 layers. The spiral pattern is Cu electroplated in 8 layers with each layer separated by an insulating layer. Fabrication steps for coil are proposed as Pattern planner spiral coil on wafer and electroplate copper coil layer. Sputter 1st insulating layer then pattern for 02 copper links for 2nd layer and electroplate copper links. Pattern for planner spiral 2nd layer on 1st insulating layer and electroplate 2nd copper coil layer. Sputter 2nd insulating layer then pattern for 02 copper links (different location as compared to step ‘b’) for 3nd layer and electroplate copper links. Pattern for planer spiral 3rd layer on 2nd insulating layer and electroplate 3rd copper coil layer. Repeat this process to complete 8 copper coil layers. Etching of insulating layer. This coil is to be fabricated on both sides of wafer. Two coils are connected in series therefore via connection across wafer thickness is also made. Capacitive area coating and connection pads are connected through two flexure hinges. Fig.6 Coil Fabrication processes (a) Pattern planer spiral coil on wafer and electroplate copper coil layer. (b) Sputter 1st insulating layer then pattern for 02 copper links for 2nd layer and electroplate copper links. (c) Pattern for planer spiral 2nd layer on 1st insulating layer and electroplate 2nd copper coil layer. (d) Sputter 2nd insulating layer then pattern for 02 copper links (different location as compare to step ‘b’) for 3nd layer and electroplate copper links. (e) Pattern for planer spiral 3rd layer on 2nd insulating layer and electroplate 3rd copper coil layer. (f) Repeat this process to complete 8 copper coil layers. (g) Etching of insulation layer. VI. ANALYSIS A. Structure Stress Analysis Sensor structure can be subjected to different forces during manufacturing, storage and operational life. Analysis of designed structure is carried out to evaluate its survival/ performances in different conditions. Flexures are the critical parts that are under stress, in free condition when centre part is not fixed between two wafers the centre part can bend freely and subjected to maximum stress as shown in figure-7. Input moment is defined as ‘M’ where ‘P’ is the mass of proof mass part of centre wafer. σ = as here are 2 flexure with section modulus ‘S’ σB = = 5.4012 Kg/ mm2 σA = = 6.3832 Kg/ mm2 Also deflection and angle of point ‘B’ in free condition are calculated. δB = ( + ) = 0.0111 mm Θ = ( + ) = 0.0433 rad Maximum deflection of point ‘C’ in free condition is calculated as δC = δB + Θ *2L = 0.2494 mm But in assembled form this maximum deflection is restricted by the fixed wafers. Maximum deflection possible in assembled form is the gap provided for capacitance pickoff. Stress analysis in the condition when centre wafer is fixed between two fixed wafers is also carried out. Fig.7 Static condition when centre wafer is free This condition is analyzed by using static equilibrium conditions and general solutions for deflection of statically determinate beams [19]. RA + RB = nP --- 1 2L RB + M = nPL --- 2 – + ( – ) 2L = 0 --- 3 Moment is calculated by solving three equations and stress due this moment ‘M’ is calculated σB = M/2S = 0.2381 225 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 235. Kg/mm2 . This stressis produced by application of 1g input and is much less than the stress calculated in free condition. From this analysis it is concluded that in assembled form the flexures can bear much higher inputs before failure. Fig.8 Static equilibrium condition when centre wafer is fixed between two wafers. Stress analysis and deflection analysis is performed in FEM software and the results are verified. Results of FEM analysis for the free condition are shown in figure-9. Fig.9 FEM stress analysis for free condition B. Mechanical Noise Intrinsic noise of sensor can be due to mechanical and electronic components. Mechanical–thermal noise due to Brownian motion is usually analyzed for MEMS sensor and is expressed total noise equivalent acceleration (TNEA) as √ Where “kB” is Boltzmann’s constant, “T” is absolute temperature, “w” is natural frequency and “Q” is quality factor of flexure. High mass and quality factors lead to low noise. Noise level of this design is estimates as 4.3µg/√Hz. C. Closed Loop System Analysis Resolution, measurement range, bandwidth, non linearity and noise are the parameters to be analyzed for performance of sensor. Simulation analysis of the designed sensor is performed to find out the behavior of sensor for critical performance parameters. Measurement range is the minimum and maximum acceleration that sensor can response along its input axis with certain accuracy. As spring stiffness is designed lowest to get minimum acceleration detection, in open loop operation the small distance between the moving capacitance surface and the fixed surface limits the maximum measurement range. Performance of open loop accelerometer is very limited concerning to measurement range, bandwidth and linearity [20]. In close loop operation maximum measurement range is limited by the feedback force that can be applied to rebalance the deflected mass. This design can only be operated in closed loop with high gain pick off system with negative feedback. To achieve desired system dynamics PID controller coefficients should be carefully designed [21]. Stiffness of flexure is very small to detect smallest input. The deflection produced by smallest input is to be detected by the pickoff system. Gains of pickoff blocks are calculated to detect smallest input. Closed loop analysis of the design is performed in Matlab. Architecture of close loop with a PID controller HCONT is shown in figure-10. Pick off gain and other parameter used for control and analysis are HCA = 10 A/V. HPA = 6880 pF/Arc-min, KPSA = 0,2V/pF with τPSA = 6.37e-6 sec. KTM =201 g-mm/A with τTM = 3e-5 sec. Steady state sensitivity of sensor depends on mass of pendulum ‘m’, length of pendulum ‘l’ and KTM of rebalance torque system. As these parameters remain constant for a sensor for this system steady state sensitivity if calculated as Kss = ml / KTM = 0.1005 /201= 0.0005A. Fig.10 Close loop architecture with output configuration Fig.11 frequency response plot of closed loop architecture -350 -300 -250 -200 -150 -100 -50 Magnitude(dB) 10 2 10 3 10 4 10 5 10 6 10 7 -450 -360 -270 -180 -90 0 90 Phase(deg) Bode Diagram Gm = 75.5 dB (at 1.66e+004 rad/sec) , Pm = Inf Frequency (rad/sec) i a pFα HPA (pF/arc- min) Mi n ΔM V - + HTM – Coil & Magnet MTM ml HPSA (V/pF) HCA (A/V) i Gp HCONT 226 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 236. Frequency response analysisshows a bandwidth of 1.18e+4 rad/sec. Step response shows a rise time of 1.5154e-004 sec with 27.5 % overshoot for this output configuration as is shown in figure.12. Fig.12 Step response of close loop architecture Output response for a square input signal of 100 Hz is shown in figure. Fig.13 Response of close loop architecture Maximum deflection of proof mass is limited by the capacitance gap. In close loop operation the deflection of proof mass must be minimum and less than the capacitance gap to avoid electrical contact. To control the angular deflection the close loop analysis is performed by transfer function as per configuration is shown in figure 13. Fig.13 Close loop architecture for angular deflection Deflection for a step input signal of different g is shown in figure-14. Maximum deflection at input signal of 10 g is less than 1.5e-3 rad. Minimum deflection of proof mass ensures reduced non linearity of output. Fig.14 Deflection of sensor against step input PID control parameters can be further optimized to achieve best close loop performance of accelerometer. VII. CONCLUSION In this paper a new magnetic MEMS accelerometer is designed with capacitive pickoff using permanent magnet and coil for rebalancing of proof mass instead of traditional electrostatic rebalancing of MEMS accelerometers. Resolution of 10 µg can be achieved to fulfill navigation requirements. Heavier proof mass results in reduced noise and also can be rebalanced by electromagnetic force. Silicon structure provided good thermal stability and magnets also provide low thermal sensitivity as in traditional electromechanical accelerometers. Permanent magnet can be fabricated in the bulk fabrication of MEMS devices. Fabrication processes of permanent magnet and coil are also proposed. In future study the device can be fabricated and different performance parameters can be studied. REFERENCES [1] Honeywell, Q-Flex Navigation Accelerometers, http://www.inertialsensors.com [2] Colibrys Single axis analog accelerometer http://www.colibrys.com [3] Y.Dong, P.Zwahlen, A.M Nguyen,R. Frosio, F. Rudolf,“ Ultra high precision MEMS accelerometer” Transducers’11, Beijing, China,June5-9,2011. [4] P. Zwahlen, Y Dong, A-M. Nguyen, F. Rudolf, J-M Stauffer, P. Ullah, V. Ragot, “Breakthrough in High Performance Inertial Navigation Grade Sigma-Delta MEMS Accelerometer” [5] E.P Furlani, Permanent Magnet and Electromechanical Devices, Material analysis and Applications Academic press 2001. [6] Ralph Hopkins, Joseph Miola, Roy Setterlund, Bruce Dow, William Sawyer, “The Silicon Oscillating Accelerometer: A High- Performance MEMS Accelerometer for Precision Navigation and Strategic Guidance Application” Institute of Navigation 61st Annual Meeting Cambridge, MA, June 27-29, 2005. [7] Navid Yazdi, Farrokh Ayazi, Khalil Najafi, “Micro-machined Inertial Sensors” Proceedings of IEEE Vol. 86, No. 8 Aug 1998. [8] Cenk Acar and Andrei M Shkel, “Experimental evaluation and comparative analysis of commercial variable-capacitance MEMS accelerometers” Step Response Time (sec) Amplitude 0 1 2 3 4 5 6 x 10 -3 0 1 2 3 4 5 6 7 8 x 10 -4 System: Gcsys Peak amplitude: 0.000638 Overshoot (%): 27.5 At time (sec): 0.000434 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 -1 0 1 2 3 4 5 6 x 10 -4 Linear Simulation Results Time (sec) Amplitude 0 0.005 0.01 0.015 0.02 0.025 0 0.5 1 1.5 x 10 -3 Step Response- Displacement of pendulum Time (sec) Amplitude(rad) 10 g 8 g 6 g 4 g 2 g 1 g Gp α HTM Coil & Magnet i HPA (pF/arc- min) HPSA (V/pF) HCA (A/V) pFVi HCONT ml - + ΔM Min MTM a 227 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 237. [9] Teodor LucianGrigorie “The Matlab/Simulink Modeling and Numerical Simulation of an Analog Capacitive Micro- Accelerometer. Part1: Open loop ” MEMSTECH’ May 2008 Polyana UKRAINE [10] Qingkun Zhou, Azhar Iqbal, Pinhas Ben-Tzavi, Dapeng Fan, “Design, Analysis, and optimization of a magnetic micro actuators” Proceedings of the ASME 2009, November13-19 Florida ,USA. [11] Tsung-Shune Chin, “Permanent magnet films for application in micro-electromechanical systems” Journal of Magnetism and Magnetic Materials 209 (2000) 75-79. [12] Ololade D Oniku, Benjamin J Bowers, Sheetal b Shetye, Naigang Wang and David P Arnold “Permanent magnet microstructures using dry-pressed magnetic powders” Journal of Micromechanics and Microengineering published 12-june 2013. [13] David P. Arnold., Naigang Wang., “Permanent Magnet for MEMS”. Journal of Electromechanical Systems. Vol.18. No. 6, December 2009. [14] D.H.Titterton, J.L. Weston “Strapdown inertial navigation technology” 1997 Peter Peregrinus Ltd. [15] Ki-Ho Han , Young-Ho Cho, “ Self-Balanced Navigation- Grade Capacitive Micro-accelerometer Using Branched Finger Electrodes and Their Performance for Varying Sense Voltage and Pressure ” Journal of Microelectromechanical systems, Vol. 12, No.1, February 2003. [16] Will Robertson, Ben Cazzolato, Anthony Zander, “Axial Force Between a Thick Coil and a Cylindrical Permanent Magnet: Optimizing the Geometry of an Electromagnetic Actuator” IEEE Transactions on Magnetics Vol. 48 No.9, September 2012. [17] Paul McGuiness David Jezersek, Spomenka Kobe, “100- µ- thick Nd-Fe-B magnets for MEMS applications produced via a low-temperature sintering route.” [18] Chong H. Ahn, Mark G. Allen. “Micro- machined Planar Inductors on Silicon Wafers for MEMS Applications”. IEEE Transactions on Industrial Electronics Vol. 45, No.6 December 1998. [19] Andrew Pytel, Ferdinand L. Singer, “Strength of Materials” [20] Teodor Lucian Grigorie “The Matlab/Simulink Modeling and Numerical Simulation of an Analog Capacitive Micro- Accelerometer. Part2: Closed loop” MEMSTECH’ May 2008 Polyana UKRAINE [21] Yin Liang, Liu Xiaowei, Chen Weiping, Zhou Zhiping, “High resolution interface circuit for closed-loop accelerometer” Journal of semiconductor Vol.32, No.4 April 2011. 228 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 238. Feasibility Assessment ofRunning JP-8 Fuel in Diesel Engine M. Shahan1 , Ali Sarosh2 ,S. Javaid3 Department of Aerospace Engineering National University of Sciences and Technology Pakistan 1 heavy.cavalry.archer@gmail.com 2 alisarosh@cae.nust.edu.pk 3 saad.javaid@gmail.com Abstract— Single fuel concept (SFC) demonstrates the need for a safe and cost effective fuel that can be utilized by light-duty as well as specialized heavy-duty vehicles. Logistics and pipeline operations are greatly simplified when SFC is implemented for vehicles. The study considers the technical feasibility of using aviation fuel JP-8 in diesel engines. It compares the effects on fuel injection, combustion, engine performance and emissions when using JP-8 fuel instead of diesel inside a compression ignition engine. The performance of both fuels at various operating conditions is calculated. Besides this, a detailed review of chemical kinetics of diesel and JP-8 and the effect of thermodynamic properties on engine performance and emission is also presented. In addition, combustion analysis of both the fuels is done in which endothermic enthalpies due to air and water at various fuel-air ratios are calculated. The results reflect that use of JP-8 in diesel engine leads to a penalty of torque and fuel consumption primarily due to lower density and viscosity. The reduced cetane number causes increase in ignition delay and premixed combustion of JP- 8. Modifications to engine are proposed so as to match the performance of both the fuels. Analysis show that implementation of SFC is best suited for supplication in heavy-duty diesel engine vehicles Index Terms— JP-8, Enthalpy, Cetane number, Ignition delay, Pre-mixed combustion INTRODUCTION Single fuel policy states that Jet Propellant 8 (JP-8) fuel must be used in all air and ground battle field vehicles. According to a research, 38.6% of army logistics supply consists of fuel. JP-8 is considered as a safe fuel for storage as well as during operations. The US army and NATO will be using JP-8 fuel in their aircrafts and ground vehicles till 2025[1].Using JP-8 in ground vehicles offered my advantages. It reduced nozzle fouling, exhaust emissions, water entrainment and microbial growth inside the fuel tanks and increased fuel filter replacement intervals, storage stability and enhanced oil change and filter replacement intervals[2]. JP-8 is a jet fuel and belongs to the class of kerosene. The flash point of JP-8 is 100 degrees Fahrenheit. It is specified by MIL-DTL-83133. It was first used by the NATO countries having a NATO code F-34. JP-8 is almost similar to Jet-A1 fuel except that it has three military additives. Tests were conducted on standard diesel cycle to give an overall idea of replacing JP-8 fuel with diesel [3].Using JP-8 in ground battle field equipment leads to many advantages. Logistics are simplified, oil lubricity is increased and exhaust emissions are reduced. Single fuel concept also carried some draw backs as well. Wear in fuel injection pumps exhaust valves increases with the use of JP-8[4].Briefing charts regarding the properties of JP-8 and other military fuels, different additives that are used in fuels and their effect on fuel properties has been presented by Schmitigal, Joel Tebbe and Jill in [5]. Similarly, world jet fuel specifications have been given in [6]. A brief review of the previous work done is this regard is given below: Laboratory evaluation of JP-8 in diesel engine was conducted as given in [7] in which JP-8 fuel was found to be satisfactory for use in diesel engine but the maximum engine power was reduced and leak in fuel injection lines was observed. Compensation was done by increasing thermal efficiency. Engine testing with various blends of JP-8 was conducted by April Covington with the perception that in future an EPU will be designed to optimize engine performance with JP-8[8] Direct imaging and two color thermometry technique was applied to verify the emission trends for JP-8 and fossil diesel fuel in an optically accessible single cylinder diesel engine. Image analysis focusing was done to determine the characteristics of combustion. Results verified that JP-8 emitted less smoke and increased HC emissions [9]. Tests with petroleum, hydro processed and Fischer Tropsch diesel and jet fuels were carried out in a direct injection single cylinder diesel engine. Decrease in ignition delay was quantified by increasing the derived cetane number (DCN) of fuel. DCN decreased the ignition delay[10] Systematic study of ignition delay for different fuels in 2.44 L heavy duty single cylinder diesel engine was conducted over a wide range of temperatures and densities. Results indicated that fuel properties relating to spray break have a nominal effect on the ignition delay period [11]. 229 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 239. Oxides ofnitrogen are harmful to the environment. During combustion, oxides of nitrogen are formed depending on the temperature and pressure. Hence there is a need to reduce nitrogen oxide emissions from diesel engine. Control on NOx emissions was obtained by introducing varying quantities of cetane improver additive in diesel fuel by enhancing the ignition quality of fuel[12] Mie-scattering technique is used in [13] to determine the spray characteristics of fuels which includes spray angle and spray tip penetration. Spray characteristics are in turn correlated to combustion and emission trends of fuels Tests conducted on S60 engine revealed that JP-8 reduced nitrogen oxide and particulate matter signatures in the exhaust emission and hence it is environmentally friendly. Torque and fuel economy penalty were compensated by increasing the pulse width of fuel [1]. Tests results carried out on 60 kW direct injection diesel engine running alternatively on diesel, JP-8 and JP-8 treated with varying quantities 2-ethylhexyl nitrate showed that ignition delay decreases with the addition of cetane improver additives [14]. Sound analysis regarding maximum heat release rate, engine efficiency, cylinder pressure and exhaust were also presented. Analysis of the usage of blends of biodiesel and JP-8 in diesel engine as given in [15] revealed that specific fuel consumption decreases and NOx emissions increase by increasing the quantity of biodiesel in test fuels. It was concluded that these fuel blends can be effectively used inside diesel engine. METHODOLOGY The basic aim was the verification of the results whether JP- 8 could be used as a replacement of diesel fuel in diesel engine. Main objective was to understand the effects on combustion and performance of engine when replacing diesel with JP-8 in diesel engine. The approach which was followed is as described: Analysis of actual diesel and JP-8 fuel properties was done so that effect of each property on engine performance can easily be quantified Heat of combustion analysis of both fuels was performed to obtain an indirect insight about the thermodynamic parameters of fuels and power output of engine when replacing JP-8 with diesel fuel Dynamometer testing of engine is done to determine the performance parameters such as torque and power output Assessment of fuel economy and strategies for matching the power output with JP-8 with different adjustments FUEL PROPERTIES Density effects the composition of fuel. Greater density indicates lower volatility. Density of diesel is greater than JP- 8.It relates to the fuel pressure and dynamic start of fuel injection in the engine[16] Cetane number provides a measure of ignition delay of fuel. This property indicates the quantity of straight chain hydrocarbons in fuel. Lower cetane number implies lower proportions of straight chain hydrocarbons and in turn longer ignition delay[17]. Heating value of a fuel determines its energy content. Higher heating value results in higher power output of engine [11]. JP- 8 has higher heating value than diesel but its lower density leads to lower heating value on volume basis. Viscosity provides a measure of resistance to flow. Working of fuel injection pump is directly related to viscosity of fuel. Pump wear and leakage in injection pumps increases due to lower viscosity. Generation of spray pattern is also governed by the viscosity of fuel[18]. Volatility of fuel is governed by distillation curve. Heavier fuel results in incomplete vaporization and combustion and it causes soot and smoke quantity to be increased[18]. Hydrocarbons having multiple benzene rings are called Poly Aromatic Hydrocarbons (PAH’s). Diesel fuel contains higher proportions of aromatics. Aromatics effect density, cetane number and soot formation characteristics of fuel [17, 19, 20]. Environmental Protection Agency (EPA) requires that sulfur content in the fuel should be as low as possible. JP-8 has lower sulfur content as compared to diesel fuel so it is environmentally friendly as compared to diesel. CHEMICAL TESTING OF FUELS Samples of JP-8 and diesel were taken and complete chemical testing was done so that engine performance trends can be co-related to chemical properties of fuels. Following results were obtained from the chemical testing of JP-8. Chemical testing of JP-8 Table 1 Test tests of JP-8 Test Method Test Limits Test Result Appearance Color, Saybolt ASTM D156 Report +29 Composition Total Acidity, mg KOH/gm ASTM D3242 0.015 max 0.007 Sulfur, Total % mass ASTM D4294 0.3 max 0.006 Sulfur, Mercaptan, % mass ASTM D3227 0.003 max 0.0003 Volatility Initial boiling point, °C ASTM D86 Report 146.9 10% vol. Recovery, °C ASTM D86 205 max 166.6 20% vol. Recovery, °C ASTM D86 Report 174.1 230 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 240. 50% vol. Recovery,°C ASTM D86 Report 188.6 90% vol. Recovery, °C ASTM D86 Report 214.9 End point, °C ASTM D86 300 max 228.9 Residue, % Vol ASTM D86 1.5 max 0.5 Loss, % vol ASTM D86 1.5 max 0.5 Flash point, °C IP 170 38 min 40 Density @ 15°C, kg/𝑚3 ASTM 1298 775 to 840 802.2 Specific Gravity @15.6/15.6 °C ASTM D1296 Report 0.8025 Fluidity Freezing point, °C ASTM D2386 -47 max -50 Viscosity @ -20°C, cst ASTM D445 8 max 3.574 Combustion Hydrogen Contents, % mass ASTM D3343 Report 13.62 Smoke point, mm ASTM D1322 19 min 23 Specific Energy net, MJ/kg ASTM D3338 42.8min 43.13 Naphthalene, % vol ASTM D1840 3 max 1.66 Contamination Existent gum, mg/100ml ASTM D381 7 max 1 Particulate Contamination, mg/L ASTM D6452 1 max 0.8 Chemical testing of diesel Table 2 Test results of diesel Test Method Test Limits Results Appearance Color ASTM D1500 3 max 2.5 Composition Total Acidity, mg KOH/gm D974 0.5 max 0.01 Sulfur, Total % wt. D129 1.0 max 0.08 Volatility Flash point, °C D93 66 max 60 Specific Gravity @ 15.6/15.6 °C ASTM D1298 Report 0.833 Fluidity Pour point, °C ASTM D97 -7 max -7 Viscosity @`20°C, cst ASTM D445 6.5 max 3.86 Cloud point, °C ASTM D2500 9 max 4 Contamination Ash content, %wt ASTM D482 0.01 max 0.001 Sediment, %wt. ASTM D473 0.01 max 0.0008 COMBUSTION ANALYSIS OF FUELS Combustion analysis of fuels was done on Chemical Kinetic Analyzer to obtain an indirect insight about the thermodynamic parameters of fuels and the power output of engine when replacing JP-8 with diesel fuel Figure 1 Chemical Kinetics Analyzer The chemical kinetic analyzer works on the principle of heat exchanging mechanism. Fuel is allowed to burn inside the combustion chamber. Air is supplied to aid in the combustion of fuel. Water flows through the outer walls of the combustion chamber. When fuel burns, heat of combustion is absorbed by air and water. In this way, we can calculate the heat released by the fuel and the efficiency of combustion. We can also control the amount of fuel and air that flows inside the combustion chamber so that we can analyze the combustion efficiency at different air to fuel ratios. 231 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 241. Figure 2 SchematicDiagram of Chemical Kinetics Analyzer Test Results Figure 3 Heat of Combustion of Diesel and JP-8 Figure 4 Combustion Efficiency of Diesel and JP-8 Comparison of Heat of Combustion and Combustion Efficiency of Diesel and JP-8 is given in Figure 3 and Figure 4. It clearly indicates that when JP-8 is used in Diesel engine, there will be a compromise in power output. Heat release profile for diesel is greater than JP-8 by about 5.6% and hence diesel engine will give less power output when replaced with JP-8. Combustion efficiency of both the fuels increases by increasing the equivalence ratio. However, efficiency increase for JP-8 is higher than diesel DYNAMOMETER TESTING The engine used in experiment is ISUZU Truck 6BD1 102 kW engine. Complete specifications of engine are given in table Table 3 Engine Specifications Cylinder 6 Displacement 5.785L Bore (mm) 102 Stroke (mm) 118 Compression ratio 16.5:1 Advertised Power/Speed (kW/rpm) 102/3000 Max torque/Speed (Nm/rpm) 402/1600-1800 Idle speed (rpm) ≤750 Ignition Sequence 1-5-3-6-2-4 Net Weight (kg) 555 Intake Valve Open at 18° (B.T.D.C.) Intake Valve Close at 46° (A.B.D.C.) Exhaust Valve Open at 48° (B.B.D.C.) Exhaust Valve Close at 16° (A.T.D.C) The dynamometer system used is capable to measuring engine torque, RPM, power, fuel flow rate and oil pressure inside the engine Figure 5 Schematic Diagram of Dynamometer Testing (1) Engine, (2) Dynamometer, (3) Shaft, (4) Flywheel, (5) Exhaust Pipe, (6) Control Unit, (7) Fuel Measurement System Density of JP-8 is less as compared to diesel and engine is fueled on volume basis instead of mass. The compensation for density can be done by increasing the duration of injection of JP- 8 so that same mass of both the fuels is injected inside the combustion chamber and hence its power output can be 232 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 242. increased. Viscosity andvolatility also affects the performance of fuel injector. At the same time some adjustments can be made to cater for the lower cetane number of JP-8 by changing the injection timing. In this way, full understanding of the effect of fuel properties on engine performance can be understood. Correlation of performance trends with test results is done in order to explain them and devise strategies for matching engine output of JP-8 to that of diesel fuel. Speed-Load combinations used during the testing of engine are as follows: 1250 rpm, 40% load (low rpm - low load) 1600 rpm, 50% load (mid rpm - mid load) Testing at 1250 rpm, 40% load (low rpm - low load) Figure 6 Fueling Rate @1250 RPM and 40% Load Figure 7 Brake Power @1250 RPM and 40% Load Figure 8 Brake Specific Fuel Consumption @1250 RPM and 40% Load Lower density of JP-8 and leakage loses in the fuel injection pump are two significant factors relating to the less fueling of JP-8. Initially the fuel flow rate of JP-8 is 12% less than baseline JP-8 and then it is increased by increasing the pulse width. After the modification the difference is reduced to 5.4%. Similarly the reduction in torque is due to lower fueling rate of JP-8 in spite of the fact that JP-8 has a higher heating value than diesel on mass basis. The increase in brake specific fuel consumption (BSFC) is mainly due to increased frictional and mechanical loses while running the diesel engine on JP-8. The reduction in fueling rate can be compensated by increasing the duration of injection of JP-8 so that same mass of both the fuels is injected. The results of increased pulse width are then compared with baseline diesel fuel. Similarly the quality of combustion can be improved by advancement in injection timing of JP-8 to cater for its lower cetane number. Combustion starts earlier in the cycle. Consequent result is the reduction in mechanical loses and increased brake power. The effect of increasing pulse width and injection timing is the increased torque and consequently higher power output. BSFC is reduced when injection pulse width is increased. In this way, performance of JP-8 can be match with baseline diesel fuel. Table 4 Comparison of Brake power and BSFC Diesel JP-8 Increased pulse width Increased pulse width + crank angle % Difference with baseline Diesel Brake power 16 11 14.5 15.2 5% BSFC 242.30 7 309.9 5 252.99 241.34 0.38% 233 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 243. Testing at 1600rpm, 50% load (mid rpm - mid load) Figure 9 Fueling Rate @1600 RPM and 50% Load Figure 10 Brake Power @1600 RPM and 50% Load Figure 11 Brake Specific Fuel Consumption (B.S.F.C.) @1600 RPM and 50% Load Table 5 Brake power and BSFC at 1600 RPM Diesel JP-8 % Difference Fueling rate(kg/s) 0.002545 0.002775 -9% Brake power (kW) 41 41.8 -1.91% BSFC (g/kWh) 223.503 237.931 -6.45% At 1600 RPM, adjustment of pulse width is not required because fueling rate is already higher due to increased load on the fuel injector and higher pressure buildup. Hence the power output of JP-8 is very much the same as that of diesel fuel. Brake Specific Fuel Consumption of JP-8 is higher as compared to diesel attributed to the lower cetane number of JP-8 and increased mechanical loses as evident from figure above. But at 1600 RPM the temperature of engine is higher when compared with diesel because of higher heating value of JP-8 and frictional loses inside the engine. RESULTS AND CONCLUSION The basic aim is to the study of feasibility whether JP-8 fuel can be used in diesel engine and the effects on engine performance and fuel combustion. The methodology adopted explains the rationale behind the differences in performance achieved by two different fuels. The results of this experimental study can be summarized as follows. Properties of fuels differ from location to location due to atmosphere and climate condition and the test limits are set according to the environment where testing is to be performed. Properties of fuels have a major effect on the performance parameters of engine like cetane number governs the ignition delay period and lower viscosity of JP-8 causes lower fueling rate, lower pressure build up and leakage loses. Although JP-8 has higher energy content as compared to diesel but it results in reduction of power output because of its lower density as the fuel is injected on volume basis inside engine. This fact is verified by the combustion analysis performed on Chemical Kinetics Analyzer where diesel showed 5% greater heat output When used inside the diesel engine, there is reduction in power and torque output due to lower fueling rate of JP-8 and lower viscosity. For example, in the above case at 1200 rpm, fueling rate of diesel is 12% higher than JP-8. At low rpm, performance of JP-8 can be matched by increasing the mass of fuel injected and advancing the crank angle. The difference in power remained only about 5% when the pulse width and crank angle adjustment was applied. At high rpm, there is no need for any engine modification. JP-8 gives the same output as that of 234 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 244. Diesel because ofhigher pressure build up and load on the fuel pump. But the engine heats up earlier because of greater frictional loss. Brake Specific Fuel Consumption for JP-8 is higher than diesel fuel when used in diesel engine due to higher frictional and mechanical loses and lower cetane number of JP-8. Analysis show that implementation of SFC is best suited for supplication in heavy-duty diesel engine vehicles REFERENCES [1] G. Fernandes, J. Fuschetto, Z. Filipi, D. Assanis, and H. McKee, "Impact of military JP-8 fuel on heavy-duty diesel engine performance and emissions," Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, vol. 221, pp. 957-970, 2007. [2] D. G. Weir, "Strategic Implications for a Single-Fuel Concept," DTIC Document1996. [3] A. Command, "JP-8: the single fuel forward: an information compendium," Warren, MI, 2001. [4] A. F. Montemayor, L. L. Stavinoha, S. J. Lestz, and M. E. LePera, "Potential Benefits from the Use of JP-8 Fuel in Military Ground Equipment," DTIC Document1989. [5] J. Schmitigal and J. Tebbe, "JP-8 and other military fuels," DTIC Document2011. [6] E. Aviation, "World jet fuel specifications with avgas supplement," Machelen: ExxonMobil Aviation, 2008. [7] W. E. Likos, E. C. Owens, and S. J. Lestz, "Laboratory evaluation of Mil-T-83133 JP-8 fuel in army diesel engines," DTIC Document1988. [8] A. Covington, "The investigation of combustion and emissions of JP8 fuel in an auxiliary power unit," 2011. [9] J. Lee, H. Oh, and C. Bae, "Combustion process of JP- 8 and fossil Diesel fuel in a heavy duty diesel engine using two- color thermometry," Fuel, vol. 102, pp. 264-273, 2012. [10] S. Gowdagiri, X. M. Cesari, M. Huang, and M. A. Oehlschlaeger, "A diesel engine study of conventional and alternative diesel and jet fuels: Ignition and emissions characteristics," Fuel, vol. 136, pp. 253-260, 2014. [11] D. A. Rothamer and L. Murphy, "Systematic study of ignition delay for jet fuels and diesel fuel in a heavy-duty diesel engine," Proceedings of the Combustion Institute, vol. 34, pp. 3021-3029, 2013. [12] K. Velmurugan and S. Gowthamn, "Effect of CETANE Improver Additives on Emissions." [13] J. Lee and C. Bae, "Application of JP-8 in a heavy duty diesel engine," Fuel, vol. 90, pp. 1762-1770, 2011. [14] G. Labeckas, S. Slavinskas, and V. Vilutiene, "Effect of the Cetane Number Improving Additive on Combustion, Performance, and Emissions of a DI Diesel Engine Operating on JP-8 Fuel," Journal of Energy Engineering, 2014. [15] A. Uyumaz, H. Solmaz, E. Yılmaz, H. Yamık, and S. Polat, "Experimental examination of the effects of military aviation fuel JP-8 and biodiesel fuel blends on the engine performance, exhaust emissions and combustion in a direct injection engine," Fuel Processing Technology, vol. 128, pp. 158-165, 2014. [16] D. Kouremenos, D. Hountalas, and A. Kouremenos, "Experimental investigation of the effect of fuel composition on the formation of pollutants in direct injection diesel engines," SAE Technical Paper 0148-7191, 1999. [17] Y. Kidoguchi, C. Yang, and K. Miwa, "Effects of fuel properties on combustion and emission characteristics of a direct-injection diesel engine," SAE Technical Paper 0148- 7191, 2000. [18] T. J. Callahan, T. W. Ryan, L. G. Dodge, and J. A. Schwalb, "Effects of fuel properties on diesel spray characteristics," SAE Technical Paper1987. [19] M. Lapuerta, J. Hernandez, R. Ballesteros, and A. Durán, "Composition and size of diesel particulate emissions from a commercial European engine tested with present and future fuels," Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, vol. 217, pp. 907-919, 2003. [20] T. L. Ullman, K. B. Spreen, and R. L. Mason, "Effects of cetane number, cetane improver, aromatics, and oxygenates on 1994 heavy-duty diesel engine emissions," SAE Technical Paper 0148-7191, 1994. . 235 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 245. 1 Artificial Intelligence Robot MuhammadUsman Saleem Muhammad Shaheer Sajid Muhammad Mohid Collage of earth and atmospheric science LCAS LCAS University of the Punjab Lahore Pakistan Lahore Pakistan Lahore Pakistan shaheersajid15@gmail.com mmoohhiidd111@gmail.com Osman.geomatics@gmail.com Abstract: A.I.R or Artificial intelligence robot is a machine which can be operated in two modes: manual and automatic. This machine is specially designed to suit the needs for both domestic and industrial purposes. A.I.R can be control and operated within 10m of radius. A.I.R can detect obstacles within distance of 1.5m. Then it will search the path without obstacles. For automatics mode, it has been programed in such a way, that we can define a line for its path then it will be follow this path throughout its journey. Through the concentration of smoke with in its domains, A.I.R can detect the location where the smoke occurred. Its can transmit videos to the receiver station through the camera installed on it. This video can be investigated on a laptop. Manual control can be done with the help of receiver. Keywords: Control system, artificial intelligence, security robots, and robotics. I.INTRODUCTION Now-a-days, Robotics has become one of the essential tools aiming to make human lives much more compatible and easier. Over the past years, many different mobile robots have been created which directly affect the scientific community; such as crawling and legged robot. This paper deals with a wheel robot [1]. Similarly, A.I.R has specifications which too aim for the same outcome. The basic functions of A.I.R include; obstacle detection, path following, line following and smoke/fire detectors. Multiple mobile robots have advances on single mobile robot [2]. Hence we make it a multiple mobile robot. Alongside, our robot has the ability to be controlled directly by laptop and also by a remote controller. It may also be viewed by a light weight wireless camera. At this moment, it is worth mentioning that A.I.R, though small in size, has been very efficiently to build, also making it a low cost autonomous robot. This actually helps large companies to buy these robots and get access to their numerous advantages and applications, cheaply. These functions hold immense importance for what we hoped to achieve. The use of obstacle detectors and line following functions actually put up a vast frame of use of the A.I.R. Examples include, factories, Metro Politian [1] and even for military purposes. The fire detectors and path detector functions allow the robot to be used as a fire escaping gadget, public safety, health care management system respectively [3]. A.I.R can be operated automatically and manually. Arduino is used to control A.I.R through laptop. This robot is 60% recycled. II. OBECTIVES To build an artificial intelligence robot for the public safety, road navigation, health care management system, fire detection, industrial and military purpose. Primary objectives were to connect this robot to laptop and get on the spot information regarding to defense purposes. III. METHODS AND METHODOLOGY A. Remote Controller A.I.R can be operated through remote control .It has a receiver with its paired transmitters, which work on the principle of radio waves. A transmitter sends radio waves to the receiver which are then decoded and further processed (see fig 1). There are four channels of the receiver which are attached to the motor deriver circuit that derives the motor (see fig.2 & 6). The radio circuit is a four data transceiver circuit, 2 controls for each motor: forward and backward. Hence increasing control over the robot. B. Laptop Connection In order to connect this robot to the laptop we have used Arduino. The transmitter has a simple 4 channel encoder circuit (see fig.2) that is connected to a development board called Arduino (see fig. 3). The Arduino is programmed in such a way to control it via a computer (see fig.4). Hence the robot is a computer controlled one. 236 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 246. 2 Figure 1: RemoteControl transmitter Figure 2: Remote controller transmitter Figure 3: Arduino connections used. Figure 4: Connectivity of the Arduino with radio transmitter. The programming is done using LabVIEW, visual programming software (see fig. 5). The program firsts establishes serial communication with Arduino. There are 8 pins of the Arduino board being used for the movement of the robot: 2 for each motion type i. e. forward, backward, right and left (see fig.5).These pins are attached to the 4 channels of the transmitter. So when the program runs, the user can control the motion of the robot. The receiver is a decoder with a radio receiver module tuned to the frequency of the transmitter. It has decoder which decodes the signal and gives out 4 channels. The four channels are then fed into a motor driver circuit, 2 channels for each motor i.e. forward and backward control for each motor (see fig.5). C. Line following/Path following Figure 7 shows the line following circuit on board of A.I.R. The path following is a simple mechanism of reflection over 237 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 247. 3 bright and dullsurfaces. It is based on the working of a light dependent resistor (see fig.8) over such surfaces. Figure 5: Programming in LabVIEW to conduct the action on robot with keyword commands. Figure 6: Radio receiver circuit on the A.I.R. Table 1 shows the table sheet used for line following sensor. The robot has a pair of light dependent resistors which control the motors independently (i.e. one connected to one motor). The circuit which controls it is an operational amplifier used as a comparator (see fig.8).The comparator compares voltage of the light dependent resistor with the reference voltage across it. When light falls on the light dependent resistor the output of the amplifier turns high. There are two such circuits each for one motor. The outputs are fed to the forward driving input of the motor driver. The path detector has made circular in shape and it is making of white color so that when A.I.R moves over the surface it can detect its specific path (see fig 9). When both sensors are on white surface they turn the outputs of both the operational amplifiers high. These outputs are attached to the forward motion control pins of the motor driver. Hence both the motors move forward (see fig. 10). Figure 7: Line following circuit on board of A.I.R. Table 1: Specifications of the Line Following Circuit Specifications Explanation LM741*2 Operational amplifier used as comparator Light dependent resistor*2 to sense light reflected from floor 10 kilo ohms*4 Resistor to set reference voltage 10 kilo Ohms*2 Variable resistor to adjust threshold level Two blue LED’s To transmit light on the floor 470 ohms Resistor limit current through LED’s Figure 8: Line following circuit with L.E.D.s and Motor drivers used in A.I.R. 238 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 248. 4 Figure 9: Linedetector on the A.I.R. Figure 10: Mechanism of line following in forward motion When one of the light dependent resistor fall on black surface its resistance increases, this sets the output of the corresponding amplifier low. The operational amplifier attached to the forward pin of the corresponding motor shuts down the motor. Figure 11: Mechanism of the line following during turn. Hence that motor is off while the other is on, this caused the A.I.R to turn, in the direction to the off motor so that the both the sensors are back on the white surface (see fig.11). D. Obstacle detection Obstacle detection is the procedure of detection of obstacles in the path of robot (see fig.12).Obstacle detection is based on transmission and reflection of infrared rays. The infrared rays fall on an object and reflect back (see fig. 13, 14 & 15). The sensor receives the reflected rays and turns its output high or low. This is used to drive the transistor which energizes the relay. The relay acts as a switch and breaks the positive connection of the line following circuit (see fig.13). The circuit turns off and the wheels stop. Hence the AIR stops when an obstacle comes in front of it. A.I.R detects the obstacle within 1 meter of range and automatically stops 20cm before the obstacle. The transmitter circuit is a dual oscillator to oscillate first on infrared LED at 38 kHz and then oscillate at 38 kHz frequency. This is so because the receiver used is a 38 kHz receiver and requires a dual oscillator to work. The receiver circuit comprises of a 3 pin infrared receiver. When it receives infrared rays it turns its output pin high and after amplified by a transistor a relay is energized. Table 2: Specifications of transmitter for Obstacle detection Specifications Explanation 555 timers *2 Integrated circuit configured as A stable multi-vibrator 1 kilo ohms Resistor to set frequency 1.5 kilo ohms Resistor to set frequency 330 ohms Resistor to limit current through led 56 kilo ohms Resistor to set frequency 220 kilo ohms Resistor to set frequency 470 ohms Resistor to drive the transistor 10 nF Capacitor to set frequency of oscillator 4.7 nF Capacitor to set frequency of oscillator Infrared L.E.D Transmitter BC547 Transistor to drive the first oscillator 239 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 249. 5 Figure 12: Obstacledetection sensor on board of A.I.R Table 3: Specifications of the receiver for Obstacle detection Figure 13: Receiver circuit of obstacle detection sensor. Figure 14: Infrared Transmitter for obstacle detection used in A.I.R Figure 15: Principle of working infrared sensor to detect obstacle in its path. E. Motor driver The motor driver circuit is a circuit based on the motor driver integrated circuit L293D. It is a quad-H bridge motor driver integrated circuit that can control two D.C motors in both directions independently (see fig.16). Figure 16: Electric motors used in the A.I.R Specifications Explanation BC547 Transistor to drive relay 330 ohms Resistor limit current through infrared sensor 470 ohms Resistor drive the transistor Tsop1738 Infrared sensor module 240 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 250. 6 F. Smoke sensor A.I.Rhas mounted with a gas sensor (smoke sensor) to detect any leakage of gas. These gases are particular to methane, propane and L.P.G (see table 4). The sensor has radioactive isotope of the element Americium which acts as the heating element (see fig 17) for the gases .The change also appears to be the output voltage. This voltage can be measured to determine the concentration of gas or even drive an alarm circuit (see fig.18). Table 4: Smoke sensor specifications. Specifications Explanation MQ-2 LPG, methane, propane, carbon dioxide sensor 10 Kilo ohms Resistor to adjust the sensitivity of the gas sensor. Figure 17: Gas sensor used in the A.I.R Figure 18: Testing of gas sensor G. Wireless camera The camera is basically operated by Bluetooth. An app is used to transfer live video from one device to another. The app used is an android based one called Camera Remote. The app is installed on two devices. One is configured as a camera while other is configured as the display. IV. APPLICATIONS OF A.I.R Wireless live video streaming Can be used in hospitals for transporting patients and other equipment through a line following system Obstacle sensor is used to detect obstacles by vehicles For spying purposes by several agencies. By the military for making robot based weapons Can be used in industries to transport materials. For security purposes RESULTS AND CONCLUSIONS This paper mainly describes the mobile robot which has achieved the mentioned objectives for this study. Line tracking makes this robot as path specific machines. It detects the obstacle with in 1.5m of its radius, the magnitude of the obstacle detection is going to increase within 1.5m of its radius, A.I.R will stop before the obstacle and will be able to change path through laptop commands. Sensing the smoke in its buffer zone is only possible when these gas molecules come in contact with the gas sensor. Due to available sized antenna, A.I.R can be controlled automatically within 20-27 m of domain. But in future, this range can be increased so that its area of application increases. Figure 19: A.I.R used for the research. 241 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 251. 7 ACKNOWLEDGEMENT We would liketo acknowledge the contribution of Residence Officer, Prof. Dr. Sajid Rashid Ahmad, Principal of Collage of earth and atmospheric science, University of the Punjab, Lahore Pakistan for kind contribution for the plagiarism evaluation for this study. Also we would like to express our deepest gratitude to Dr.Najam Abbas, Sectary ICASE 2015, Institute of Space technology, Islamabad Pakistan for his very kind contribution in publication of this effort to the public. REFERENCES [1] R. Chandra Kumar, M. Saddam Khan, D. Kumar, R. Birua, S. Mondal and M. Kr.Parai, 'OBSTACLE AVOIDING ROBOT – A PROMISING ONE', International Journal of Advanced Research in Electrical, Electronics and Instrumentation Engineering, vol. 2, no. 4, p. 1430, 2013. [2] S. Shan, 'Mobile Robots Based Intelligent Fire Detection and Escaping System', JOURNAL OF COMPUTERS, vol. 8, no. 5, p. 1298, 2013. [3] D. Punetha, N. Kumar and V. Mehta, 'Development and Applications of Line Following Robot Based Health Care Management System', International Journal of Advanced Research in Computer Engineering & Technology, vol. 2, no. 8, p. 2446, 2013 242 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 252. An Intelligent Approachfor Edge Detection in Noisy Images using Fuzzy Logic Izhar1,2 , Fayyaz3 1 Mechanical Engineering Department, CECOS University of IT and Emerging Sciences, Peshawar, Pakistan 2 Institute of Mechatronics Engineering, University of Engineering and Technology, Peshawar, Pakistan 1 izhar@cecos.edu.pk R. Muhammad1 and N. Ahmed1 3 Department of Mechanical Engineering, Pakistan Institute of Engineering and Applied Sciences, Islamabad, Pakistan Abstract—Edge detection has widespread applications in the field of feature extraction, machine learning, computer vision, pattern recognition, remote sensing and satellite image analysis. In satellite images it detects various regions of interest for different applications. Though it has useful applications, but it is a complicated task and it becomes more arduous when it comes to noisy images. This paper reports an intelligent approach for edge detection in noisy images. The proposed edge detection approach is based on fuzzy logic. The proposed algorithm employs a 3 x 3 window mask and a fuzzy inference system. The values acquired from the window mask are subjected to the fuzzy rules for edge detection. The proposed technique is successfully tested on noise free as well as on noisy images. The experimental results are also compared with the reported established edge detection techniques such as Laplacian of Gaussian (LOG), Roberts, Prewitt, Sobel and previously proposed fuzzy based algorithm. The proposed technique when tested on a grayscale image having 24 dB ‘salt and pepper’ noise, detected 106 false edge pixels. However, in comparison the reported edge detection techniques like Sobel, Prewitt, LOG and previously developed a fuzzy based technique have detected 3678, 4435, 835 and 2852 false edge pixels, respectively. From the experimental results it is evident that the proposed edge detection technique provides better solution to the edge detection problems in noisy images. Index Terms— Edge Detection, Noisy Images, Fuzzy Logic I. INTRODUCTION Edges in an image are contours produced as a result of abrupt or sudden alteration in any of the (multiple) characteristics at pixel level. These changes could be observed due to change in texture, color, shade or light absorption. These characteristics could further lead in estimating the orientation, size, depth and surface features in an image [1]. Edge detection has numerous applications in the field of robotics, medical image analysis, geographical science, pattern recognition, and military technology [2]-[7] etc. It is often the case that images embody high frequency noise or irrelevant data which inhibits the detection of continuous edge points [8] since edge itself is a composition of high frequency data. The noise produces false flags, which often mislead the techniques for an edge. Several algorithms are employed for the detection of edges in images [9]-[14]. The motivation behind each technique is to overcome the limitations in previous methodologies. The conventional techniques like Canny, Sobel, Robert, Prewitt and LOG [9]-[14] incorporate the use of spatial differential filters utilising local gradient. These filters recognise an edge as an abrupt change of grey scale pixel intensities. These techniques are well established. These techniques process the data in relatively short time and are computationally efficient. However, the conventional techniques are susceptible to affect generated by noise during edge detection. Jiange and Bunke [15] proposed an approximation of scan lines method for edge detection. The technique in comparison to other segmentation techniques produced considerably good results. Genming and Bouzong [16] proposed a 5x5 window mask for the detection of edges based on a fixed threshold level. However, their limitation was its inadaptability to regions with varying greyscale due to fixed threshold point. Latest techniques incorporate the use of artificial immune system, artificial neural networks, genetic algorithms with particle swarm optimization and ant colony optimization [17]- [19]. Fuzzy set theory is another technique that is employed for edge detection [20]-[21]. The method performs logical and mathematical reasoning based on approximations rather than crisp values. The technique therefore significantly reduces the complexity of problems where fixed values cannot be predicted. Kaur et al. [22] proposed a fuzzy logic based edge detection technique employing a 2×2 window mask. Sixteen fuzzy rules were defined for edge detection in the algorithm .The results for edge detection was appreciable in noise free images. However when subjected to noisy image the technique detected the false edges produced due to noise as well. To address the short comings of the reported edge detection techniques, a fuzzy logic based technique is proposed in this work. The technique employs a 3×3 window mask guided by fuzzy rule set for edge detection in noisy images. In our previous work [23], we have used trapezoidal and gaussian 243 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 253. membership functions (MFs),however, in this work we have preferred Pi and generalized bell shape MF to develop an effective filter that is convenient to apply to both noise free and noisy images. Moreover, algorithm developed in this work is tested on noisy images having different noise levels. II. PROPOSED METHODOLOGY In Fig. 1 the conceptual framework of the proposed fuzzy logic based edge detection is shown. The developed edge detection technique for noisy images is based on fuzzy logic. To extract the greyscale values of neighborhood pixels from the input image, a 3x3 window mask is designed. The grayscale values of the neighborhood pixels acquired from the mask are pre-processed before applying to the fuzzy inference system. A fuzzy inference system is designed that takes these processed values as input. These values are subsequently converted into the fuzzy plane. A fuzzy rule base is defined that determine and show the edge pixels in the output image. The output of the system is computed based on centroid method and defuzzification is performed based on Mamdani implication. Fig. 1. Conceptual framework of the proposed edge detection approach. A. Window Mask for Scanning the Image A 3x3 window mask is designed for scanning the image, in the proposed approach as shown in Fig. 2. The mask takes the grayscale values, of eight neighbourhood pixels with the central pixel, as the out pixel. Before applying to the fuzzy inference system, the grayscale values obtained from the mask are pre-processed. Fig. 3 shows the processed mask, where – for x = 1, 2, 3,…,8. B. Fuzzy Membership Functions In fuzzy inference system, MFs play a key role. In the fuzzy set fuzziness is measured using MFs as they are the key constituents of the fuzzy set theory. Based on the nature of problem the type and shape of the MF should carefully be selected as they have effects on the fuzzy inference system. In the developed edge detection technique, for the input data, Pi MFs are used, because they show reasonably improved results with minimum error [24]. Moreover for the output data, 244 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 254. generalized bell MFsare used because they are smooth and non-zero at all points. Fig. 2. window mask for scanning the image. Equation (1) represents the standard Pi MF. { ( ) ( ) ( ) ( ) 1 where , , and are the various parameters of Pi MF, and its details are depicted in Fig. 3(a). Equation (2) represents the generalized bell MF. | | 2 where ‘k’, ‘l’, and ‘m’ are the different parameters of the generalized bell MF and its details are shown in Fig. 3(b). Fig. 3. MF plots (a) Pi, (b) Generalized Bell. C. Fuzzy Sets Each input, to fuzzy inference system is divided into two fuzzy sets, Less and More. The output (pixel), from the fuzzy inference system is also divided into two fuzzy sets, Non-Edge and Edge. In Figs. 4 and 5, the associated MFs with the input and output fuzzy sets are shown respectively. Fig. 4. MFs of the input variable ∆Pj. Fig. 5. MFs of the output pixel 245 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 255. Table I liststhe various terminologies and parameters of both the input and output fuzzy sets. TABLE I. PARAMETERS AND TERMINOLOGIES OF INPUT AND OUTPUT FUZZY SETS Linguistic Variable Parameter Range MF Type Fuzzy Input ∆P1 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P2 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P3 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P4 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P5 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P6 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P7 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Input ∆P8 Less [0 0 23 47] [0 255] Pi MF More [23 47 255 255] Fuzzy Output G0 Non-Edge [3 2 10] [0 255] Generalized Bell MFEdge [3 2 248] D. Fuzzy Knowledge Base Fuzzy knowledge base or rule base in fuzzy inference system is a set of linguistic descriptions [25]. Fuzzy rule base plays a key role in fuzzy inference system as it makes conclusions related to classifying an input or stabilizing and adjusting the output. Fuzzy rule base for the proposed edge detection algorithm consists of the following linguistic descriptions as listed in Table II. III. RESULTS AND DISCUSSION The developed edge detection technique is tested on a number of grayscale images including noise free and noisy images. In noise free greyscale images the developed technique has successfully detected all type of edges as shown in Fig. 6. In Fig. 6 (a) grayscale image of size 185 x 232 pixels having five different regions covered by four boundary lines is shown. As shown in Fig. 6 (c), the proposed technique for edge detection has detected these four boundary lines (edges) successfully. Similarly, as shown in Fig. 6 (b) the proposed method has successfully detected edges in the greyscale (flower) image. TABLE II. FUZZY KNOWLEDGE BASE FOR THE DEVELOPED EDGE DETECTION TECHNIQUE Rules Statement R1 If ∆G1==More && ∆G2==More && ∆G8==Less then G0 == Edge R2 If ∆G1==More && ∆G4==More && ∆G8==Less then G0 == Edge R3 If ∆G2==More && ∆G3==More && ∆G8==Less then G0 == Edge R4 If ∆G4==More && ∆G6==More && ∆G8==Less then G0 == Edge R5 If ∆G1==More && ∆G2==More && ∆G7==Less then G0 == Edge R6 If ∆G1==More && ∆G4==More && ∆G7==Less then G0 == Edge R7 If ∆G2==More && ∆G3==More && ∆G7==Less then G0 == Edge R8 If ∆G4==More && ∆G6==More && ∆G7==Less then G0 == Edge R9 If ∆G1==More && ∆G2==More && ∆G9==Less then G0 == Edge R10 If ∆G1==More && ∆G4==More && ∆G9==Less then G0 == Edge R11 If ∆G2==More && ∆G3==More && ∆G9==Less then G0 == Edge R12 If ∆G4==More && ∆G6==More && ∆G9==Less then G0 == Edge Fig. 6. Tested images: (a) Image having four different regions, (b) Flower, (c) Edge detection in image having four different regions, and (d) Edge detection in flower image. 246 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 256. The developed edgedetection technique has the advantage of detecting edges in the noisy images as previously discussed (in the introduction). This is verified by detecting edges in an image having 24 dB ‘salt and pepper’ noise. To compute the noise level in an image through peak signal to noise ratio (PSNR), first based on equation (3) the mean square error (MSE) is computed [26]-[27]. ∑ ∑[ ] 3 Where and represents the input noise free and noisy images respectively. While and indicates the total number of rows and columns of the input images respectively. Finally the expression for the computation of noise level becomes as in (4). [ ] 4 Where denotes the maximum possible intensity value of the pixel in the input image. The value of for eight bit unsigned integer data type image is 255. The developed edge detection technique is applied to a size 500 x 232 pixels image having ‘salt and pepper’ noise at a level of 24 dB. The simulation results are compared with other conventional and reported edge detection algorithms as shown in Fig. 7. Fig. 7. Comparison of experimental results in noisy image, (a) Original image, (b) Noisy image, (c) Sobel edge detection (d) Prewitt edge detection, (e) LoG edge detection, (f) Previously developed fuzzy based edge detection technique [22], (g) The proposed method. From the experimental results it is clear that the proposed fuzzy based edge detection algorithm has detected a very few false edge pixels in comparison to the other reported edge detection techniques. The number of false edge pixels detected by different reported edge detection techniques, when subjected to images having various noise levels is shown in Fig. 8. From Fig. 8 it is evident that the developed edge detection technique when subject to a noisy image of 500 x 232 size and 24 dB noise level has detected 106 false edge pixels, while other edge detection techniques for instance, Sobel, Prewitt, LOG, and previously developed fuzzy logic technique, 247 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 257. after fine tuningtheir parameters have detected 3678, 4435, 835 and 2852 false edge pixels respectively. Fig. 8. False edge detected pixels in a standard image of 500 x 232 pixels: A comparision. IV. CONCLUSION AND FUTURE WORK This paper proposes and demonstrates a fuzzy logic based edge detection algorithm for noisy images. The developed technique employs a 3×3 mask guided by fuzzy rule set for edge detection in noisy images. The developed technique has successfully detected all the edge pixels in noise free and noisy images. The developed algorithm is also compared with other conventional and previously developed fuzzy logic based edge detection techniques. The developed edge detection algorithm when subjected to a 500 x 232 size grayscale image having 24 dB ‘salt and pepper’ noise has detected a very few false edge pixels (106), while the reported edge detection techniques like Sobel, Prewitt, LOG and previously developed fuzzy logic have detected 3678, 4435, 835 and 2852 respectively. From the experimental results it is obvious that in case of noisy images the proposed technique provides better results. In future an investigation on how to incorporate artificial immune system with fuzzy logic to develop a hybrid technique for edge detection is under consideration. REFERENCES [1] V. Torre and T. A. Poggio, “On edge detection.,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 8, no. 2, pp. 147–63, Feb. 1986. [2] G. O. B. Siliciano, L. Sciavicco, L. Villani, “Modelling, Planning and Control,” in Robotics, Springer, 2010, pp. 415–418. [3] Sirikan Chucherd, “Edge detection of medical image processing using vector field analysis,” in 11th International Joint Conference on Computer Science and Software Engineering (JCSSE), 2014, pp. 58–63. [4] H. Lin, P. Du, C. Zhao, and N. Shu, “Edge detection method of remote sensing images based on mathematical morphology of multi-structure elements,” Chinese Geogr. Sci., vol. 14, no. 3, pp. 263–268, Sep. 2004. [5] R. Rulaningtyas and K. Ain, “Edge detection for brain tumor pattern recognition,” in International Conference on Instrumentation, Communication, Information Technology, and Biomedical Engineering 2009, 2009, pp. 1–3. [6] A. A. Goshtasby, 2-D and 3-D Image Registration: for Medical, Remote Sensing, and Industrial Applications. Wiley, 2005, pp. 34–39. [7] A. Loft, K. E. Jensen, J. Löfgren, S. Daugaard, and M. M. Petersen, “PET/MRI for Preoperative Planning in Patients with Soft Tissue Sarcoma: A Technical Report of Two Patients.,” Case Rep. Med., vol. 2013, p. 791078, Jan. 2013. [8] M. Setayesh, M. Zhang, and M. Johnston, “Effects of static and dynamic topologies in Particle Swarm Optimisation for edge detection in noisy images,” in 2012 IEEE Congress on Evolutionary Computation, 2012, pp. 1–8. [9] J. Canny, “A Computational Approach to Edge Detection,” IEEE Trans. Pattern Anal. Mach. Intell., vol. PAMI-8, no. 6, pp. 679–698, Nov. 1986. [10] C. Y. H. X. Zhang Jin-Yu, “Edge detection of images based on improved Sobel operator and genetic algorithms,” in 2009 International Conference on Image Analysis and Signal Processing, 2009, pp. 31–35. [11] A. Rosenfeld, “The Max Roberts Operator is a Hueckel-Type Edge Detector,” IEEE Trans. Pattern Anal. Mach. Intell., vol. PAMI-3, no. 1, pp. 101–103, Jan. 1981. [12] R. A. Kirsch, “Computer determination of the constituent structure of biological images.,” Comput. Biomed. Res., vol. 4, no. 3, pp. 315–28, Jun. 1971. [13] L. Yang, X. Wu, D. Zhao, H. Li, and J. Zhai, “An improved Prewitt algorithm for edge detection based on noised image,” in 2011 4th International Congress on Image and Signal Processing, 2011, pp. 1197– 1200. [14] F. Ulupinar and G. Medioni, “Refining edges detected by a LoG operator,” in Proceedings CVPR ’88: The Computer Society Conference on Computer Vision and Pattern Recognition, 1988, pp. 202–207. [15] X. Jiang and H. Bunke, “Edge Detection in Range Images Based on Scan Line Approximation,” Comput. Vis. Image Underst., vol. 73, no. 2, pp. 183–199, Feb. 1999. [16] C. Genming and Y. Baozong, “A new edge detector with thinning and noise resisting abilities,” J. Electron., vol. 6, no. 4, pp. 314–319, Oct. 1989. [17] A. Naumenko, V. Lukin, and K. Egiazarian, “SAR-image edge detection using artificial neural network,” in 2012 International Conference on Mathematical Methods in Electromagnetic Theory, 2012, pp. 508–512. [18] O. P. Verma and R. Sharma, “An optimal edge detection using universal law of gravity and ant colony algorithm,” in 2011 World Congress on Information and Communication Technologies, 2011, pp. 507–511. [19] M. Setayesh, M. Zhang, and M. Johnston, “Effects of static and dynamic topologies in Particle Swarm Optimisation for edge detection in noisy images,” in 2012 IEEE Congress on Evolutionary Computation, 2012. [20] C.-C. L. Yau-Hwang , Chang-Shing Lee, “A new fuzzy edge detection method for image enhancement,” in Proceedings of 6th International Fuzzy Systems Conference, 1997, vol. 2, pp. 1069–1074. [21] S. E. El-Khamy, M. Lotfy, and N. El-Yamany, “A modified fuzzy Sobel edge detector,” in Proceedings of the Seventeenth National Radio Science Conference. 17th NRSC’2000 (IEEE Cat. No.00EX396), 2000. [22] K. Kaur, V. Mutenja, and I. S. Gill, “Fuzzy Logic Based Image Edge Detection Algorithm in MATLAB,” Int. J. Comput. Appl., vol. 1, no. 22, pp. 57–60, Feb. 2010. [23] S. Mandal, J. Choudhury and S. Chaudhuri, “In Search of Suitable Fuzzy Membership Function in Prediction of Time Series Data,” IJCSI International Journal of Computer Science Issues, vol. 9, no.3, 2012. [24] Izhar, I. U. Haq, K. Shah, T. Khan, K. Azam, S. Anwar, “Fuzzy Logic Based Edge Detection for Noisy Images” Technical Journal, UET Taxila, vol. 20 (SII), no. II (S), 2015. [25] O. Karray, and C. De Silva, Soft computingand intelligent systems design: theory, tools and applications, Pearson Addison Wesely, Essex CM20 2JE, England, 2004. [26] F. Luisier, T. Blu and M. Unser “Image Denoising in Mixed Poisson- Gaussian Noise,” IEEE Transactions on Image Processing, vol. 20, no. 3, pp. 696-708, 2011. [27] J. Liu, Z. Huan and H. Huang, “Image restoration under mixed noise using globally convex segmentation,” Journal of Visual Communication and Image Representation, vol. 22, no.3, pp. 263-270, 2011. 248 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 258. A Survey ofActive ITU-R P-Series Propagation Models Sundus Najib1 , Ibrar Ali Khan2 , Muhammad Waleed Aftab1 , Naveed Mufti1 1 Department of Telecommunication Engineering 2 Department of Electrical Engineering University of Engineering and Technology, Peshawar Peshawar, Pakistan. sundus.najib@gmail.com Rafia Mehreen Department of Telecommunication Engineering University of Engineering and Technology, Taxila Taxila, Pakistan Abstract— ITU-R (International Telecommunication Union) P-series (Propagation-series) recommendations encapsulate the propagation models and data vital for modeling of any radio communication path, including space and terrestrial paths. For a new researcher, an understanding of these recommendations prior to their application in various scenarios is desirable. This paper presents an introduction to propagation, radio channel impairments and previous review studies, followed by review of thirty-one recommendations applicable to numerous indoor/outdoor propagation scenarios. It is observed that some outdoor propagation models can be used for partial indoor scenarios while the only indoor model provided by ITU-R series can also be used for various outdoor setups. Certain recommendations are interconnected with other recommendations, developing a need for their comprehensive study. Keywords— ITU-R; Radiowave Propagation; Indoor/Outdoor propagation models; P-Series. I. INTRODUCTION The International Telecommunication Union (ITU)’s Radiocommunication Sector primarily works for global harmonization of the use of radio frequency spectrum. In addition to supervising the international radio spectrum regulation and satellite orbits resources, ITU-R also develops Radiocommunication standards and recommendations for the efficient use of spectrum in Radiocommunication systems [1]. Radio spectrum has a vast range of applications in considerable operational environments. Three main mechanisms namely refraction, diffraction and scattering make it possible for the radio waves to propagate beyond the horizon in terrestrial communication [2]. Certain factors can have different effects along the path of transmission, resulting in the loss of information signal. For communication to succeed, modeling of the radio spectrum/system to overcome and minimize these channel impairments is vital [3]. Many modeling techniques for radio channel modeling have been put forth by researchers from time to time for effective radio communication. ITU-R encourages research and investigations into different aspects of radiocommunication by proposing questions (ITU-R Questions), constituting Study Groups, seeking input from Member States and Industry and ultimately converting different inputs into ITU-R Recommendations. It takes years of research and hard work for developing and modifying each recommendation, thus making it valid and reliable. The Recommendations are further classified on the basis of ‘services’ [1]. There are currently seventy-eight active recommendations dealing with Radiowave Propagation [4]. Each recommendation is valid for specific conditions and frequency bands. At the time of writing, there is no known comparison/review of the ITU-R Recommendations pertaining to propagation. It is felt that, for a new researcher, it becomes difficult to understand, sort and pin point the most relevant ITU-R propagation model for his/her modeling/ scenario/investigation. This review paper is aimed at providing a comparative summary of the propagation models contained within the ITU-R P-series Recommendations. It is hoped that this would help the researchers to get an overview of the propagation models within the P-series and enable them to select the most relevant recommendation for modeling their specific scenario. The order of remaining paper is as follows: Section-II explains basic propagation mechanisms, followed by radio channel impairments. Section III discusses previous review studies while in Section-IV, ITU-R indoor and outdoor radio propagation models are discussed. Finally Section-V and Section-VI puts forward the guidelines and conclusions along with future work of this review paper. II. PROPAGATION MECHANISMS AND RADIO CHANNEL IMPAIRMENTS A. Propagation Mechanisms 1) Refraction: In radio propagation, the atmospheric bending of the radio signal away from the straight path is termed as atmospheric refraction [2]. A drop in temperature and pressure with increasing altitudes results in decreasing value of refractive index with height, contributing to bending of the wave towards the ground. Atmospheric refraction is calculated in units of ‘N’ [2]: Where, is the refractive index. 249 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 259. Fig. 1. RefractionTypes [2] In order to follow the earth’s curvature, the change in refractivity with height, termed as vertical refractivity gradient, denoted by dN/dh, should be equal to -157 N units/kilometer [2]. When the gradient exceeds -157 N- units/kilometer, trapping of signals inside ducts can occur. A less negative refractivity gradient causes signals to be super- refracted while gradients more positive than -40 N- units/kilometer can cause sub-refraction. Fig-1 shows different refractive types and associated values of gradients. 2) Diffraction: Diffraction occurs due to the obstructions present in the path of transmitter and receiver antennas [3]. Objects with sharp irregularities and impenetrable bodies such as buildings, vegetation, terrain etc. are the main reason for the diffraction of radio signal. Three common procedures for diffraction modeling are Bullington method, Epstein-Peterson method and Deygout (also known as the ‘principle edge method’) [2]. Bullington method is widely used in ITU-R P-Series Propagation Models for diffraction calculations which is based on constructing a comparable knife edge at the junction of transmitter/receiver horizons. 3) Scattering: When radio signal encounters objects whose size are of the order or less than that of propagating wave, it gets scattered [3]. For long range Very High Frequency (VHF)/Ultra High Frequency (UHF) propagation, tropospheric scattering plays an important role, where signals are scattered to the receiver located beyond the horizon. A limit to the propagation range is impinged by the joint bulk made by the overlap of transmitter/receiver antenna patterns which needs to be in the troposphere [2]. B. Radio Channel Impairments 1) Attenuation: Attenuation can be defined as the gradual degradation of signal power during its transmission over a radio path [5]. It is measured in units of decibels (dB). Attenuated power can be calculated by: Where, Ps= Signal power at transmitter (Watt) Pd=Signal Power at Receiver (Watt) 2) Free Space Loss (FSL): FSL is assumed for simplification purposes. A signal can be visualized as spreading out from a transmitter. As it travels away from the source, it spreads out in the form of a sphere. The surface area of the sphere increases as distance from transmitter increases. Following the law of the conservation of energy, the signal intensity at a point must decrease as the surface area of the sphere increases [2]. 3) Multipath Effects: Due to the irregularities and non-uniform behavior of the radio channel along with the straight path in which the signal is transmitted, signal may reach the receiver by following other directions. The resultant signal at the receiver is then the combined effect of these multipath signals which either interfere constructively or destructively, depending on the path differences and antenna heights. 4) Fading: Fading is a random process that may vary with position, frequency and time. Fading encountered due to multipath effects can be categorized as multipath fading while the shadowing of radio signal from different objects cause shadowing fading. 5) Signal Penetration into Buildings: Radio signals are attenuated when they come in contact with a building/hill etc. Shape of the building, frequency of operation and materials of the building greatly affect the penetration phenomena. Normally signals are more attenuated by metals than plywood/concrete. Aluminum acts as a reflector medium while gypsum and asbestos have higher signal absorbing capabilities than other materials [6]. III. PREVIOUS REVIEW STUDIES Sumit Joshi et al [7] and Govind Sati et al [8] present reviews of some commonly used empirical, stochastic and deterministic propagation models, namely FS path loss model, Okumura model, COST 231 Hata model, Stanford University Interim (SUI) Model, ECC-33 model (suitable for urban environments) and COST-231 Walfish-Bertoni model for small height buildings. Furthermore, Sumit Joshi et al [7] provided a MATLAB simulated comparison of the models they tested and generated an error factor to be added with Bertoni’s model to achieve higher fidelity for scenarios under their investigation. Pooja Rani et al also reviewed above mentioned models along with two additional propagation models i.e. Longley-Rice model and Two-Ray Ground model [9]. A performance comparison of Longley-Rice Model, ITU- R P.1546 terrestrial Model and HATA-Davidson Model using Height Above Average Terrain (HAAT) for Digital Video Broadcasting Television shows that ITU-R P.1546 250 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 260. recommendation is onlyreliable for distances shorter than 50 kilometers while the Longley-Rice model incorporated with terrain information can lead to better radio modeling performance but at the cost of computational efficiency [10]. Empirical and deterministic propagation models for terrestrial communication are reviewed by Magdy F. Iskander et al [11] and ray-tracing methods for enhanced deterministic models are discussed thoroughly. Unlike other studies, Tapan K.Sarkar et al [12] provide a more comprehensive review of several statistical and site-specific models that shed light on various scenarios of small- scale/ large-scale fading. The study proposed a new numerical approach for prediction of propagation mechanisms. For indoor environment, an advance ray tracing method is put forward for locating the rays’ paths while Finite Difference Time Domain (FDTD) method is used for analyzing the propagation of radio wave through the walls of building. All the above referenced comparative studies focus on similar indoor and outdoor propagation models. No review has been conducted for discussing only ITU-R propagation models. This presents an urge to address these models for convenience of other researchers. IV. DISCUSSION OF ITU-R INDOOR/OUTDOOR PROPAGATION MODELS At the time of writing, ITU-R Propagation Section has in force seventy eight recommendations, related to indoor propagation models, outdoor propagation models, and global maps along with data, attenuation models, diffraction models, noise model and list of other factors that should be considered in modeling of radio paths [4]. Our review study focuses on the Indoor and outdoor models. In this regard, a total of thirty-one recommendations have been studied and reviewed in this paper. The recommendations’ numbers along with area of concern, frequency of operation and path length applicability are listed in Table-1. A. Indoor Propagation Model Unlike many outdoor models, ITU-R has only published a single Indoor Propagation model, as part of its P.1238 Recommendation [13]. The model, can however, be applied to partial outdoor scenarios [14]. This model covers the mechanisms and factors related to the radio propagation of indoor signals operating in 900 MHz to 100 GHz frequency range. The Path loss models are divided in two categories, namely site-general (little site information needed) models and site specific (detailed information of building) models. The models assume that the portable user equipment and the base station are located inside the building premises. Statistical and deterministic delay spread models are also discussed. The impact of polarization and antenna radiation patterns is discussed for three different cases i.e. LOS case, obstructed LOS case and cases where portable radio terminals are randomly oriented resulting in the scatter of transmitted polarization energy into orthogonal polarizations. Dependence of signal strength and quality on location of the transmitter and receiver, effect of building’s structure and its materials’ permittivity, and the effects of motion of objects in room are also thoroughly discussed in this recommendation. B. Outdoor Propagation Models: 1) Aeronautical and Space Models P.528 is an aeronautical and satellite communication based model, calculating basic transmission loss [15] of aeronautical and satellite services operating in frequency range of 125- 15500 MHz. It uses an interpolation method to determine basic transmission loss data, from set of curves that are valid for ground-satellite, ground-air, air-air, satellite-satellite and air-satellite links. This transmission loss is predicted on the basis of distance between antennas, the heights of the antennas above mean sea level, the frequency, and the time percentage. P.682 provides guidance on Earth -Space aeronautical mobile telecommunication systems planning by incorporating models to be used with P.618 [18]. The models therein take account of effects generated by signal multipath and scattering from the surface of the earth [16]. The propagation mechanisms discussed in P.1409 [17] should be taken into consideration when planning a radio communication system comprising of stations at high altitudes and other higher level platforms in stratosphere, and in the studies of sharing and compatibility. Recommendations P.531 [19], P.618 [18], P.619 [20], P.679 [21], P.680 [22], P.681 [23] and P.840 [24] are developed for planning space, satellite and earth-satellite systems. P.531 provides data and methods that help in satellite systems planning by evaluating the ionosphere’s effects on Earth-space paths at frequency range from 0.1 GHz to 12 GHz. P.618 also contributes to the planning of earth-space communication planning. P.619 puts forward model to be used for calculating interference between earth and space stations. Recommendation P.679 uses methods from P.618 and additional data to model broadcast satellite systems. P.680 also uses methods from P.618 and incorporates other data for designing Earth-space nautical mobile satellite systems. P.681 deals with modeling of land mobile telecommunication systems between Earth and space. P.840 incorporates the models, graphs and global maps needed for the calculation of attenuation caused fog and clouds in the atmosphere. 2) Terrestrial Models Total of twenty terrestrial propagation models are discussed in this section. P.452 [25] addresses the microwave interference and propagation loss encountered between earth stations operating at frequency range of 0.1 GHz to 50 GHz. Recommendation P.525 concerns the free space planning and provides guidance on the attenuation effects that should be considered in designing a radio link that assumes free space environment [26]. For evaluating the diffraction effects on the received signal strength, P.526 recommends a number of models that are applicable to various obstacle types and variety of path geometries such as spherical earth surface and irregular terrain [27]. P.530 [28] comprehensively describes the propagation effects that should be tackled when planning a fixed digital line of sight (LOS) link. ELF, VLF and LF terrestrial radio waves are confined within the space between the Earth and the ionosphere in order for them to travel to greater distances. For this purpose, recommendation P.684 proposes methods for developing services in the frequency range below 150 KHz [29]. P.842 [30] focuses on the issue of 251 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 261. reliability and compatibilityin planning and design of High Frequency (HF) radio communication systems. Considering the fact that scattering from ionization caused by meteor trails can provide a convenient way of communication at HF and VHF, Recommendation P.843 [31] tackles with this issue of meteor-burst propagation. P.844 provides information that should be considered when designing radio communication systems that rely on sharing of frequencies within the 30 MHz-3 GHz Range [32]. Planning of broadcasting services falling in the LF and MF bands i.e. 150,000 Hz to 1700,000 Hz, having path lengths from 50 km to 12000 km, are discussed in P.1147 [33]. P.1321 [34] shed light on the characteristics of LF/MF ground-wave and sky-wave propagation that might cause an effect on the digital modulation techniques used in those bands. In designing and planning of terrestrial land mobile and broadcasting services, propagation characteristics whose impact cannot be ignored in the proper modeling of these services are mentioned in P.1406 [35]. Recommendation P.1410 [36] addresses radio communication systems operating in frequency ranges of 3GHz to 60 GHz (broadband services) and provides guidelines for line-of-sight and alternative non-LOS propagation mechanisms. P.1411 [14] is a model for short range propagation in an outdoor scenario over the frequency spectrum of 300 KHz to 100 GHz but it can be used as good modeling technique for cases where some indoor factors are also involved. A method constituting of interpolation and extrapolation techniques from field strength curves that are based on empirical data is presented in recommendation P.1546 which can be applied to all the polarization types [37]. These curves are plotted as a function of path length, effective height of the antenna, frequency of operation and percentage time of the year. This method predicts the radio propagation behavior in point-to-area scenarios for land based services operational in the frequency band of 30,000 KHz to 300 MHz. Recommendation P.1791 [38] provides procedures valid for frequency ranges from 1 GHz to 10 GHz, to calculate path loss for Ultra-wideband (UWB) indoor and outdoor environments for both direct and obstructed paths, and estimating the power received by a conventional narrow-band receiver from an ultra-wideband transmitter. The method presented in P.1812 [39] gives a thorough study of terrain for point to area services which are land based and operate in the frequencies between 30,000 KHz to 300 MHz and paths starting from 0.25 kilometer and going up to 3000 km distance, with both terminals having maximum height of 3 km above ground. Radio communication can also be achieved by using optical and infrared spectra. For taking advantage of this resource, Recommendation P.1814 provides propagation estimation procedures for terrestrial free-space optical systems planning. Attenuation caused by the presence of fog in the atmosphere, rain, snowfall and attenuation of signal during normal sunny days is also addressed in this recommendation. It also covers scintillation and impairments by sunlight [40]. P.1816 [41] is an urban/sub-urban propagation model that guides on the forecasting of time based profiles and profiles of spatial elements for broadband land mobile services operating in the frequencies between 700 MHz to 9000 MHz and distances 0.05 km-3 km. In 2013, ITU-R published a wide- range terrestrial propagation model, P.2001 [42], that predicts the Path Loss due to both enhancement of the radio signal and occurrence of fading mechanism over the range from zero percent to hundred percent of an averaged year and is applicable to the frequency range from 30 MHz to 50 GHz. Finally, P.2040 deals with the issue of impact of building’s infrastructure on propagation of the radiowaves i.e. materials’ types used in its construction and the shape of building [43]. V. GUIDELINES FOR RESEARCHERS We recommend following procedure to be adopted for selecting a proper recommendation prior to modeling a scenario. 1. ITU presents two categories of recommendations i.e. terrestrial and space models, each dealing with separate set of environmental conditions. Thus foremost, it is important that scenario should be distinguished whether it is a terrestrial or space dominant environment. 2. The communication type must also be specified whether it is point-to-point, point to area or indoor communication scenario. 3. On the basis of working environment, select a suitable recommendation that covers the range of operational frequency and distance. It is of crucial importance to consider frequency limitation because certain parameters in recommendations are frequency dependent and if not taken care of, can produce inaccurate results. 4. Shortlisted recommendations should then be assessed in terms of scope of that recommendation, whether it tackles with the desired metrics and purpose of study. For example, if interference measurement is the required analysis then one should consider interference modeling recommendations. 5. The final selected model should then be read in detail and desired input parameters be collected. 6. Some recommendations rely on other recommendations for calculating certain parameters. A thorough study of referenced models is also needed. Although recommendations provide step by step procedures for calculating each parameter, they require intense mathematical calculations and a good understanding of complex mathematical problems. VI. CONCLUSIONS AND FUTURE WORK The propagation models and relevant procedures contained within the ITU-R Propagation Series (P-Series) Recommendations comprehensively address all the possible radio propagation environments from indoor to outdoor space, urban/suburban and aeronautical cases. A detailed understanding of these recommendations prior to their use in one’s research is needed. Recommendations rely on each other for the calculation of certain atmospheric and modeling parameters. Certain short range outdoor models and the sole indoor model available, can be used for partial indoor/outdoor scenarios. Also each recommendation covers some specific 252 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 262. environment and frequencyof operation, limiting its area of use. Researchers are encouraged to benefit from the comparisons and information contained in this review in making decisions about most suitable/applicable models for the scenarios under investigation. The latest ITU-R Propagation Model, P.2001, presents a wide-range propagation model for general purpose terrestrial communication. It covers frequency range from 30 MHz to 50 GHz and distances from 3 km to at least 1000 km. Unlike most of the previous models, the model predicts path loss effectively in the range from 0% to 100% of an average year. The model can be compared with other relevant propagation models for a study on a specific outdoor scenarios. TABLE I. ITU INDOOR/OUTDOOR RECOMMENDATIONS’ SPECIFICATIONS [1,44] S.No. Recommendation Number Model type Scope Frequency of operation Distance 1 P.452 Terrestrial Interference prediction and path loss estimation 0.1 GHz-50 GHz Up to radio horizon and beyond 2 P.525 Terrestrial FS attenuation calculation ------- ------- 3 P.526 Terrestrial Diffraction calculation ------ ------- 4 P.528 Aeronautical/ satellite services Transmission loss calculation/ path loss 125 MHz- 15500 MHz 0 km- 1800 km 5 P.530 Terrestrial Path loss and diversity enhancement in Digital fixed LOS links planning 150 MHz- 100 GHz 200 km 6 P.531 Satellite System Ionospheric propagation 0.1 GHz- 12 GHz ------- 7 P.618 Earth-space model Estimation of path loss, amount of diversity gain and cross polarization discrimination (XPD) in Telecommunication services 1 GHz- 55 GHz All possible orbital heights 8 P.619 Earth-space model Interference calculation --------- ------- 9 P.679 Satellite Effects of surrounding environment and path loss prediction in Broadcasting satellite systems 0.5 GHz- 5.1 GHz All possible orbital heights 10 P.680 Satellite Fade/fade duration and interference calculation for Maritime mobile telecommunication 0.8 GHz- 8 GHz All possible orbital heights 11 P.681 Satellite Fade and path loss estimation in Land-mobile telecommunication 0.8 GHz- 20 GHz All possible orbital heights 12 P.682 Aeronautical Fade and multipath calculation in Aeronautical mobile telecommunication systems 1 GHz- 2 GHz (sea), 1 GHz- 3 GHz (ground) All possible orbital heights 13 P.684 Terrestrial ELF, VLF and LF terrestrial radio services 30 kHz- 150 kHz 0 km- 16000 km 14 P.840 Earth-space model Attenuation due to clouds and fog >10 GHz ------- 15 P.842 Terrestrial Reliability/compatibility prediction ---------- ------- 16 P.843 Terrestrial Transmission loss, meteor flux variations, antenna and frequency considerations in Meteor-burst communication 30 MHz-100 MHz 100 km- 1000 km 17 P.844 Terrestrial Ionospheric communication 30 MHz-3 GHz ------- 18 P.1147 Terrestrial Estimating sky-wave field strength in Broadcast services 150 kHz-1700 kHz 50 km- 12000 km 19 P.1238 Terrestrial Amount of path loss and delay spread in case of Indoor propagation 900 MHz - 100 GHz Buildings, interiors 20 P.1321 Terrestrial Digital modulation ----------- ------- 21 P.1406 Terrestrial Land mobile and broadcasting services ------------ ------- 22 P.1409 Aeronautical High altitude radio communication systems ---------- ------- 23 P.1410 Terrestrial Coverage and reduction in coverage due to rain in Broadband radio access 3 GHz- 60 GHz 0 km- 5 km 24 P.1411 Terrestrial Path loss and delay spread in Out-door short range propagation 300MHz- 100 GHz Less than 1 km 25 P.1546 Terrestrial Field strength measurement in Point-to-area propagation 30 MHz- 3000 MHz 1 km- 1000 km 26 P.1791 Terrestrial UWB path loss for indoor/outdoor cases 1 GHz- 10 GHz ------- 27 P.1812 Terrestrial Field strength calculation in Point-to-area propagation 30 MHz- 3000 MHz Up to radio horizon and beyond 28 P.1814 Terrestrial Calculating beam spreading, absorption and scattering loss and scintillation in FSO radio communication systems 20 THz- 375 THz No limit 29 P.1816 Terrestrial Urban/suburban planning 0.7 GHz-9 GHz ------- 30 P.2001 Terrestrial Path loss due to signal enhancement and fading 30 MHz- 50 GHz 3 km- 1000 km 31 P.2040 Terrestrial Building materials’ properties and structures ---------- ------- 253 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 263. REFERENCES [1] “International TelecommunicationUnion”, (ITU), www.itu.int, 2/2/2015, <http://www.itu.int/en/Pages/default.aspx> [2] “Propagation Tutorial”, (Mike Willis), 4/2/2015, <http://www.mike- willis.com/Tutorial/propagation .html > [3] Andrea Goldsmith, “Path Loss and Shadowing,” in Wireless Communications, Cambridge, United Kingdom: CUP, 2005. [4] “Radiowave propagation”, (ITU 2008), Accessed 2/2/2015,< http://www.itu.int/rec/R-REC-P/en> [5] L. K. Bandyopadhyay, S. K. Chaulya, and P. K. Mishra, “Mines Communication Technique,” in Wireless Communication in Underground Mines: RFID-based Sensor Networking, Springer Science & Business Media, 2009. [6] Masri, T., Chew, S.P., Wong, C.P., and Lias, K., “A study of signal penetration into building materials,” IEEE 1st International Conference on Computers, Communications, & Signal Processing, 2005. CCSP 2005, Pages: 21-24, Kuala Lumpur, Malaysia. [7] Sumit Joshi, and Vishal Gupta, “A review on empirical data collection and analysis of Bertoni’s model at 1.8 GHz,” International Journal of Computer Applications (IJCA), 2012, Volume 56 - Number 6, pages: 17-23, October 2012. [8] Govind Sati, and Sonika Singh, “A review on outdoor propagation models in radio communication,” International Journal of Computer Engineering and Science (IJCES), Volume: 4, Issue: 2, pages: 64-68, October 2014. [9] Pooja Rani, Vinit Chauhan, Sudhir Kumar, and Dinesh Sharma, “A review on wireless propagation models,” International Journal of Engineering and Innovative Technology (IJEIT), Volume: 3, Issue: 11, May 2014. [10] Stylianos Kasampalis, Pavlos I. Lazaridis, Zaharias D. Zaharis, Aristotelis Bizopoulos, Spyridon Zettas, and John Cosmas, “Comparison of Longley-Rice, ITU-R P.1546 and Hata-Davidson propagation models for DVB-T coverage prediction,” 2014 IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB), pages: 1-4, Beijing, China, June 2014. [11] Magdy F. Iskander, and Zhengqing Yun, “Propagation Prediction Models for Wireless Communication Systems,” IEEE Transactions on Microwave Theory and Techniques, VOL. 50, NO. 3, pages: 662 – 673, March 2002. [12] Tapan K.Sarkar, Zhong Ji, Kyungjung Kim, Abdellatif Medouri, and Magdalena Salazar-Palma, “A survey of various propagation models for mobile communication,” IEEE Antennas and Propagation Magazine, Volume: 45, Issue: 03, pages: 239 – 307, June 2003. [13] ITU-R P.1238-5 (2007), “Propagation data and prediction methods for the planning of indoor radio communication system and radio local area networks in the frequency range 900 MHz to 100 GHz”, ITU-R Recommendation [14] ITU-R P.1411-7 (2013) ,”Propagation data and prediction methods for the planning of short-range outdoor Radiocommunication systems and radio local area networks in the frequency range 300 MHz to 100 GHz”, ITU-R Recommendation [15] ITU-R P.528-3 (2012),”Propagation curves for aeronautical mobile and radio navigation services using the VHF, UHF and SHF bands”, ITU-R Recommendation [16] ITU-R P.682-3 (2012),” Propagation data required for the design of Earth-space aeronautical mobile telecommunication systems ”, ITU- R Recommendation [17] ITU-R P.1409-1(2012),”Propagation data and prediction methods for systems using high altitude platform stations and other elevated stations in the stratosphere at frequencies greater than about 1 GHz”, ITU-R Recommendation [18] ITU-R P.618-11(2013),”Propagation data and prediction methods required for the design of Earth-space telecommunication systems”, ITU-R Recommendation [19] ITU-R P.531-12(2013),”Ionospheric propagation data and prediction methods required for the design of satellite services and systems”, ITU-R Recommendation [20] ITU-R P.619-1(1992),”Propagation data required for the evaluation of interference between stations in space and those on the surface of the Earth”, ITU-R Recommendation [21] ITU-R P.679-3(2001),”Propagation data required for the design of broadcasting-satellite systems”, ITU-R Recommendation [22] ITU-R P.680-3(1999),”Propagation data required for the design of Earth-space maritime mobile telecommunication systems”, ITU-R Recommendation [23] ITU-R P.681-3(1997),” Propagation data required for the design of Earth-space maritime mobile telecommunication systems”, ITU-R Recommendation [24] ITU-R P.840-6(2013),”Attenuation due to clouds and fog”, ITU-R Recommendation [25] ITU-R P.452.15(2013),”Prediction procedure for the evaluation of interference between stations on the surface of the Earth at the frequencies above about 0.1 GHz”, ITU-R Recommendation [26] ITU-R P.525-2 (1994),” Calculation of free-space attenuation”, ITU- R Recommendation [27] ITU-R P.526-13(2013),”Propagation by diffraction”, ITU-R Recommendation [28] ITU-R P.530-15(2013),” Propagation data and prediction methods required for the design of terrestrial line-of-sight systems”, ITU-R Recommendation [29] ITU-R P.684-6(2012),”Prediction of field at the frequencies below about 150 kHz”, ITU-R Recommendation [30] ITU-R P.842-5(2013),”Computation of reliability and compatibility of HF radio systems”, ITU-R Recommendation [31] ITU-R P.843-1(1997),”Communication by meteor-burst propagation”, ITU-R Recommendation [32] ITU-R P.844-1(94),”Ionospheric factors affecting frequency sharing in the VHF and UHF bands (30 MHz-3 GHz)”, ITU-R Recommendation [33] ITU-R P.1147-2(2003),”Prediction of sky-wave field strength at frequencies between about 150 and 1700 kHz”, ITU-R Recommendation [34] ITU-R P.1321(2013),”Propagation factors affecting systems using digital modulation techniques at LF and MF”, ITU-R Recommendation [35] ITU-R P.1406-1(2007),” Propagation effects relating to terrestrial land mobile and broadcasting services in the VHF and UHF bands”, ITU-R Recommendation [36] ITU-R P.1410-5 (2012),”Propagation data and prediction methods required for the design of terrestrial broadband radio access systems operating in a frequency range from 3 to 60 GHz”, ITU-R Recommendation [37] ITU-R P.1546-5 (2013) ,”Method for point-to-area predictions for terrestrial services in the frequency range 30 MHz to 3 000 MHz”, ITU-R Recommendation [38] ITU-R P.1791-0 (2007),” Propagation prediction methods for assessment of the impact of ultra-wideband devices “, ITU-R Recommendation [39] ITU-R P.1812-3(2013),” A path-specific propagation prediction method for point-to-area terrestrial services in the VHF and UHF bands”, ITU-R Recommendation. [40] ITU-R P.1814-0 (2007) ,” Prediction methods required for the design of terrestrial free-space optical links”, ITU-R Recommendation [41] ITU-R P.1816-2 (2013),” The prediction of the time and the spatial profile for broadband land mobile services using UHF and SHF bands”, ITU-R Recommendation [42] ITU-R P. 2001-1 (2013),” A general purpose wide-range terrestrial propagation model in the frequency range 30 MHz to 50 GHz”, ITU- R Recommendation [43] ITU-R P.2040 (2013),” Effects of building materials and structures on radiowave propagation above about 100 MHz”, ITU-R Recommendation [44] ITU-R P.1144-6 (2012), “Guide to the application of the propagation methods of Radiocommunication Study Group 3”, Recommendation. 254 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 264. Issues of AlphaMechanism in Subsonic Wind Tunnel Testing for a Hybrid Buoyant Aircraft Experimental Fluid Dynamics A.U.Haque* , W. Asrar, E. Sulaeman, J.S. Mohamed Ali and A.B. Embong IIUM-LSWT, International Islamic University Malaysia (IIUM) Kuala Lumpur, Malaysia * anwar.haque@live.iium.edu.my Abstract— In subsonic wind tunnels with large test sections, scaled down models are attached to the balance with the help of model support system. Such a system mainly includes a main strut and an auxiliary strut, also known as pitch rod. Flow measurements of such a system are required to be subtracted from the raw data of the wind tunnel. For this purpose, it is important to find the contribution of model support system separately. Potential issues and limitations related to such testing are discussed in the light of technological constraints. Special focus is given on the streamlined profile of main strut and wind tunnel testing of a hybrid buoyant aircraft. In this paper we experimentally simulate a situation in which height of both these struts are kept constant. Wind tunnels tests are conducted to find the aerodynamic and stability characteristics at constant angle of attack as well as yaw angle. An asymmetric pattern is found in the pitch and yaw stability for a defined array of beta sweep. Such a support system not contributes towards drag but also in lift as well as in side force. Index Terms— Wind Tunnel Testing, Pitch Rod, Aerodynamic and stability measurements, Data acquisition and reduction system, Strut Arrangement I. INTRODUCTION It is well known that the experimental data obtained from a wind tunnel is subject to different corrections. Correction due to model support group is among one of important correction for which wind tunnel is run for model off and strut on conditions. Struts are usually geometrically tapered with less diameter and chord at the tip than that at the root which is attached with the balance. Moreover, they are of cylindrical or streamlined shapes. Such struts have more frontal area and have complex interaction of the flow when scaled down model is attached at the tip of it. One of the prospective solutions is to have tapered struts of profile similar to a diamond shaped airfoil. Diamond shaped struts are usually used in supersonic [1-3] and hypersonic wind tunnels [4-5] in which there is requirement to hold the model from the outer profile and no interaction of the strut with the base of the model. However, to the authors’ best knowledge, its utilization in subsonic speed has hardly been seen in the literature. Most of the earlier work done is to compare the effects of different arrangement of streamlines, round or cylindrical struts with pitch rod at fixed position of angle of attack [6-7]. But such a method is not quick measuring method as the individual has to change the position of pitch rod at every required angle of attack and hence forming different “V” which require additional fillets to be installed for all angular setting of the struts tested individually [6]. Also, in comparison with the main strut the dimensions especially the diameter and the thickness of the pitch rod is usually quite less. One end of the pitch rod is usually connected with the gear system of alpha mechanism which allows the rod to swing in a plane perpendicular to the turntable axis. If the tip of the pitch rod exposed to the free stream velocity is not physically attached with the main strut than there might be some effect of flow induced vibration on the overall results. The drag of a model mounted only on main strut in a wind tunnel is always “more than the individual sum of the isolated strut and isolated body”[8]. Isolated strut are used when model is either fixed at certain position or the operator manually change the position with the help of instruments like inclinometer. Also it is usually used when the geometric dimensions of the scaled down model quite small to have auxuilary strut to be position at the aft part of the model. Examples of few models tested earlier on single strut are shown in Fig. 1. It is important to take separate measurement of the strut arrangement to get a more general picture of the estimated parameters of main strut with axillary strut for defined range of side-slip angle and angle of attack. Such a strut arrangement can cause complex interference flow fields including the increase in drag and pitching moment. This work aims to quantify the magnitude of such effects and to give a first and knowledge of general arrangement of the experimental setup for fixed height of the pitch rod. Due to more interference and drag due to strut arranged in tandem [9] form and further discussion on it or other two or three strut arrangement is beyond the scope of present work. In early times, the manufacturing and assembly techniques for wind tunnel models are not so sophisticated as that of the present time. Moreover, the size of the test section is also small. Such technological constrains limits the overall length of the model and pitch rods has to positioned at some inclined angle to attached its tip with the base of the model. In order to 255 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 265. define any arrayof input angle of attack; the pitch rod has to move in a plane perpendicular to the tunnel’s floor. (a) Delta wing model (b) Airship Model (c) Hybrid lifting hull model Fig. 1. Few models tested in IIUM-LSWT Now days, through precise machining, scaled down models in large dimensions can be manufactured by using combination of metal and composite. For such models, the height of auxiliary strut increases for negative angle of attack and decreases for positive angle of attack. In this way there will be multiple numbers of cases to estimate the contribution of model support only. One of the optimum solutions to keep minimum number of blow downs is to keep the height of the pitch rod equal to that of the main strut. Complete details of this prospective solution are discussed below with the help of an example of on-going wind tunnel testing program of a hybrid buoyant aircraft. Such aircraft are powered by fuel and has the ability to generate certain amount of aerodynamic lift for the weight which is not balanced by the aerostatic lift. II. PROSPECTIVE SOLUTION In the case of aircraft the operational angle of attack is usually limited to 16-18 degree on the positive side and -6 degree on the negative side. Hence the cases for experimental investigation for aerodynamic and stability characteristics through wind tunnel testing will be more for the positive one. In order to define a positive angle of attack, the alpha mechanism pushes the pitch rod in the downward direction. By doing so, the tail of the aircraft moves towards the tunnel floor and it is vice versa for the situation where negative angle of attack is required. In comparison with conventional aircraft, hybrid buoyant aircraft may need to fly at altitude less than that of pressure height [10]. Although a lifting fuselage can provide additional lift and hence high lift to drag ratio for the cruise and take-off segment of the flight but there may exist a scenario in which a pilot may need to give negative angle of attack [11]. In order to simulate such conditions in wind tunnel, there is always a requirement to do more testing at negative angle of attack for which the pitch rod has to move up to attain a height equal to that of the main strut. For data reduction; all the aerodynamic and stability parameters has to non- dimensionalized by reference dimensions of the wing. All such dimensions, obtained from the CAD model (actual as well as scaled down) are tabulated below in Tab. 1 for quick reference. A schematic view of the model attached with the diamond shaped strut and pitch rod attached at its base is shown in Fig. 2. It is important to note that the reference area value is obtained through an iterative analysis work for which lift from the fuselage is added into the main wing by using an equivalent wing with an airfoil as that of the wing [12]. TABLE. 1 REFERENCE DIMENSIONS OF HB AIRCRAFT Wing Area Span Chord m2 m m Origional 27.5 20 1.5 Scaled down 0.086 1 0.075 It is pertinent to highlight that the scope of present research work of model support system is not only limited to the aerospace application like that of hybrid buoyant aircraft, discussed above. But can also be applied to any body in which there is requirement to evaluate the effect of defining the body at some angle to the free stream velocity. Estimation of 256 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 266. parachute’s drag [13],bird’s aerodynamics [8,14], car’s aerodynamics [15-16], study of lift generated wakes [17] are few examples of it. Fig. 2. Isometric view of HB aircraft installed in test section (a) Side view (b) Front View Fig. 3. Positive incidence of the model with the help of pitch rod For such cases, it is simply replacing the terms of reference area for aerodynamic forces and reference dimension for the moments. For example the non-dimensional quantity F (where F is an aerodynamic force) can be expressed in the non- dimensionalized form, Eq. (1). In this relationship, instead of taking the actual exposed area of the body, the platform area of the wing is usually used as S. SV F CF 2 2 1 (1) Where, CF F,V, S and is coefficient of aerodynamic force, aerodynamic force, free stream velocity, reference area and density of air respectively. Similarly, the aerodynamic moments also can be expressed in the form of non-dimensional coefficients. Since a moment is the product of a force and a length it follows that a non-dimensional form for a moment is SlV Q CQ 2 2 1 (2) Where, Q is any aerodynamic moment about center of gravity and l is a reference length. III. EXPERIMENTAL SETUP In order to find the aerodynamic and static stability behavior of model support system discussed in the previous section, tests are conducted in IIUM-LSWT at 40 m/s. IIUM- LSWT is a closed-loop wind tunnel with a test section of dimensions 1.5m × 2.3m × 6m and maximum speed of 50 m/s [18]. A pictorial view of the test section is shown in Fig. 4(a) [19]. In this tunnel, the balance section is placed on the floor to have some height for the external balance. Both the struts (main as well as auxiliary i.e. pitch rod for alpha mechanism) are attached with the balance. The same can be observed from Fig. 4(b) in yellow colour. This wind tunnel is currently one of the largest wind tunnels in Malaysia and has excellent flow. The limitations of this experimental setup and its repurcation of the data obtained from DARCS system of the tunnel are as follows: a. It is mandatory to have the pitch rod for defining angle of attack range. Consequently, in order to quantify their aerodynamic and stability effects; no comparisons are made between results from diamond shaped strut with and without the pitch rod strut. b. No tests are performed with the model attached to the defined model support system. Therefore the investigation reported in this paper is limited to the estimation of struts along contribution and interference arising from combination of model-strut arrangement is beyond the scope of presented work. c. No tests for angle of attack are possible as the model support system is rigidity fixed with the balance. IV. EXPERIMENTAL SETUP In order to evaluate the effectiveness of the proposed engineering method, Fig. (5); wind tunnel blow downs are carried out at different velocities. All the results are plotted in the form of graph to get first-hand knowledge about the general trends of aerodynamic and stability characteristics of model support system. First the wind tunnel is run for a series of blow 257 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 267. downs (α=0, β=0)by increasing V from 30 to 50 m/s and with step velocity of 5 m/s. (a) IIUM-LSWT (b) Balance Compartment Fig. 4 Details of IIUM-LSWT Fig. 5 Proposed arrangement of model support struts It can be observed from Fig. 6 that for the said range of velocity, all the aerodynamic and stability coefficients remain almost constant. However, an increase in CL at V=45 m/s and increase in CD at V=40 m/s are observed. -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 30 35 40 45 50 Coefficients V (m/s) CL CD Cm Cy Cn Fig. 6 Variation in aerodynamic and stability coefficients with increase in the free stream velocity -40 -30 -20 -10 0 10 20 30 40 50 -20 -15 -10 -5 0 5 10 15 20 Forces(N) (degree) Lift Side Drag (a) Overall aerodynamic forces -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 -20 -15 -10 -5 0 5 10 15 20 Coefficients (degree) CL Cn CD (b) Overall aerodynamic coefficients Fig. 7 Aerodynamic properties of model support system All the moments and its coefficients are estimated about the moment reference centre (MRC) of the balance. Values of 258 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 268. coefficient of yawand side force are negative, whereas the sign of pitching moment is positive. For β sweep test; the model strut assembly remained rested on tunnel turntable and rotated at -20o to +20o range with 50 increments. As more attention is given for defining the problem description than the flow conditions. Hence, this technical note does not attempt to predict the aerodynamic and stability coefficients for beta sweep at different free stream velocities. A drastic increase in side force is observed at α = 150 , Fig. 7(b). This figure also shows a continuous increase in CL for defined range of and a slight droppage in value of CD for β = ±100 . However, after β = ±100 , CD increases linearly till β = ±200 . An asymmetric pattern is observed in overall trends of pitch and yaw moment, Fig. 8(a). Same thing is reflected in corresponding coefficient plots as well i.e. mC vs and nC vs plots, respectively. However, general behavior is more sinusoidal for the plot of nC vs . Fig. 8(c). Value of mC first increases till for β = 50 but a drastic decrease in its value is observed till β = 150 , Fig. 8(b). But its magnitude is not consistent with that of its value at β = 150 . Values of nC for . ±200 are almost constant but a sudden increase in its value is found at β = 50 . Since the overall behavior of these two stability coefficients is non-linear. Therefore, it is quite difficult to comment on the overall slopes of mC and nC . -5 15 35 55 75 95 115 -20 -15 -10 -5 0 5 10 15 20 Moment(N.m) (degree) Pitch Moment Yaw Moment (a) Plots of pitch and yaw moment due to sweep 0 0.2 0.4 0.6 0.8 1 1.2 -20 -15 -10 -5 0 5 10 15 20 Cm β (degree) (b) mC vs -0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02 0.025 -20 -15 -10 -5 0 5 10 15 20 Cn β (degree) (c) nC vs Fig. 8 Stability characteristics of model support system Sometimes, the aerodynamic force of the body alone is defined as the difference between the aerodynamic force of the body attached with the strut and results of strut alone testing. However, this is not the true representation as there is always some effect of the interference due to model support system as well. For example, the complete analytical relationship of the drag force is expressed as follows [8]: ISBm DDDD (1) Where, Dm, Ds and DI are drag of the body, strut and due to interference effects respectively. Drag due to tare is usually automatically subtracted in online calculations by the DARCS and hence it is not account far in the above equation. DI would take out any interference effects due to supports and can be estimated in future by subtracting the model-on wind-on condition from the model off-wind on condition. This is a shortcut (not perfect) to avoid lengthy procedure involving measurements with inverted model [20]. V. CONCLUSION For measurements of aerodynamic and stability coefficients of only model support system placed normal to the flow, a method is used where both the struts are at same height and attached through a thin wire. For defined range of velocity, all the aerodynamic and stability coefficients remain almost constant for α = 00 . Diamond shaped profile of the mains strut acts as a symmetric airfoil, which result in significant contribution towards the side force. For side force rest of the forces and moments, the overall trend is not symmetric. Specially, an oscillating behaviour is observed in the trend of nC vs . Such effects of the measured aerodynamic and stability characteristics caused by main strut and the pitch rod are of considerable magnitude and should be excluded from the test results of complete configuration. Although the presented results of proposed experimental setup furnish some interesting finding but its interference with the model deserved much more attention. 259 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 269. ACKNOWLEDGMENT The support ofthe Ministry of Science, Technology and Innovation (MOSTI), Malaysia, under the grant 06-01-08- SF0189 is gratefully acknowledged. Authors are thankful to Dr. Fadhil Hasim and his team for providing the diamond shaped strut for the experimental work. REFERENCES [1] D., Biermann, & W. H. Herrnstein, The interference between struts in various combinations. NASA Report No. NASA-L- 468, 1933 [2] E. P., Jones, S. M., Townsend, Y., Guy, & T. E. McLaughlin, Reduction of the wave drag of a blunt body by means of a standoff spike. AIAA Aerosp. Sei. Meet., 38th, Reno, NV, 2000 [3] S. Feldman, Some shock tube experiments on the chemical kinetics of air at high temperatures. Journal of Fluid Mechanics, 3(03), 225-242, 1957. [4] K. L., Mikkelsen, & D. F. Long, Refurbishment, Calibration, and Initial Testing in the ASE Hypersonic Wind Tunnel. In 43rd AIAA Aerospace Sciences Meeting and Exhibit, 2005. [5] M. H. Bertram, (1950). Investigation of the pressure-ratio requirements of the Langley 11-inch hypersonic tunnel with a variable-geometry diffuser, 1950. [6] D., Biermann, & W. H. Herrnstein, The interference between struts in various combinations. NASA Report No. NASA-L- 468, 1933 [7] V., Nguyen, Y., Drolet, & Watt, G. (1995, January). Interference of various support strut configurations in wind tunnel tests on a model submarine. In 33rd Aerospace Sciences Meeting and Exhibit (p. 443). [8] V.A, Tucker, "Measuring aerodynamic interference drag between a bird body and the mounting strut of a drag balance." The Journal of Experimental Biology 154.1 (1990): 439-461. [9] R. C. Pankhurst, (1958). Wind-Tunnel Interference Due to Model Supports. Aircraft Engineering and Aerospace Technology, 30(2), 55-55. [10] A.U., Haque, W. Asrar, A.A, Omar, E., Sulaeman and J.S, Ali, "Assessment of Engine’s Power Budget for Hydrogen Powered Hybrid Buoyant Aircraft", Journal of Propulsion and Power Research (in press), 2015 [11] A.U., Haque, W. Asrar, A.A, Omar, E., Sulaeman and J.S, Ali, "Aerostatic and Aerodynamic Modules of a Hybrid Buoyant Aircraft: an Analytical Approach", IIUM Engineering Journal, IIUM Engineering Journal, Vol 16, No. 1 (2015). [12] A.U., Haque, W. Asrar, A.A, Omar, E., Sulaeman and J.S, Ali, “Power-off static stability analysis of a clean configuration of a hybrid buoyant aircraft”, 7th Ankara Internatıonal aerospace conference, 11-13 September 2015 - METU, Ankara Turkey. [13] J. A., Weiberg, & K. W. Mort, Wind-tunnel tests of a series of parachutes designed for controllable gliding flight. National Aeronautics and Space Administration, 1967 [14] G. R.,Spedding, M.Rosén, & A. Hedenström, A family of vortex wakes generated by a thrush nightingale in free flight in a wind tunnel over its entire natural range of flight speeds. Journal of Experimental Biology, 206(14), 2313-2344, 2013 [15] A. Cogotti, Ground effect simulation for full-scale cars in the pininfarina wind tunnel (No. 950996). SAE Technical Paper.1995 [16] B. Hetherington, and D. B. Sims-Williams, Support Strut Interference Effects on Passenger and Racing Car Wind Tunnel Models. No. 2006-01-0565. SAE Technical Paper, 2006. [17] J., Rossow, J. N., Sacco, P. A., Askins, L. S., Bisbee and S. M Smith, Wind-tunnel measurements of hazard posed by lift- generated wakes. Journal of aircraft, 32(2), 278-284, 1995. [18] F. Hasim, R. Rusyadi, W. I. Surya, W., Asrar, A. A.Omar, , J., Syed Mohamed Ali, & R. Kafafy, The IIUM low speed wind tunnel. Proceedings of EnCon2008 2nd Engineering Conference on Sustainable Engineering Infrastructures Development & Management, December 18-19, 2008, Kuching, Sarawak, Malaysia [19] User’s Manual 6-Component External Balance for International Islamic University Malaysia Wind Tunnel, S & V Teknik Sdn. Bhd and Aero Engineering, 2015. [20] J. Barlow, W. Rae, A. Pope, Low Speed Wind Tunnel Testing, Jon Wiley & Sons, 1999. 260 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 270. Fairing Separation DynamicAnalysis Using Analytical Approach Muhammad Tanveer Iqbal Institute of Space Technology Islamabad, Pakistan mrtanveeriqbal@gmail.com Dr. Abdul Majid Institute of Space Technology Islamabad, Pakistan majid363@yahoo.com Abstract— Fairing is the upper portion of launch vehicle, usually made in two halves. It is used to protect the satellite from loadings (thermal, aerodynamic and acoustic) of harsh environment during ascent. Separation of fairing at particular altitude is most critical part of launch vehicle flight which dictates the success or failure of the whole mission. This paper presents the designing of spring for separation system of fairing using 1dof dynamic equations. Material properties of base ring and spring are considered and integrated in dynamic system of equations. These equations are modeled and evaluated. Effects on separation time due to variations in spring stiffness and compressed length are analyzed. After analysis , best spring parameters are selected on the basis of fairing separation requirements. Developed analytical method is also validated through professional fairing separation software. Keywords—satellite launch vehicle; fairing jettison; heat shield; dynamic analysis; analytical approach; separation springs I. INTRODUCTION Fairing is designed to shield the satellite from acoustic, aerodynamic and especially thermal loads. To minimize these loadings inside the fairing, honeycomb and carbon fiber sandwich structure is frequently employed. Separation of fairing is typically done at altitude where free molecular heat flux is less than 1135W/m2 to avoid damages to satellite [1-3]. Bending load and dynamic pressure are not considered for fairing separation as their utmost values are achieved prior in the flight [4]. Successful mission is heavily depends on fairing separation. Many missions were failed just due to no or late separation of fairing. In the March 4, 2011, Glory climate satellite launched by Taurus launch vehicle was failed to achieve specific orbital altitude due to failure of fairing separation and collapsed back to earth [5]. Two halves of fairing are separated by using pyro-bolts and gas/spring thrusters. In the given case, spring thrusters are selected for designing jettison mechanism. Two main considerations for designing separation mechanism are that the magnitude of force produced by spring system must be enough to separate the fairing's halves in given time at very high acceleration and after separation, there should be no collision of opened halves with the launch vehicle [6]. Both are directly related to spring stiffness. If spring stiffness is low, sufficient force would not be produced to open the fairing halves in given time. The high spring stiffness will produce the high angular velocity in fairing halves and cause its collision with launch vehicle. Therefore, optimized value of stiffness must be evaluated. In this paper, along with stiffness, other design parameters of spring used in fairing jettison mechanism are also evaluated. For this purpose, analytical equations for rotational dynamics are derived using laws of motion. Then using Range-Kutta method, these equations are solved, simulated and analyzed. Acquired results are validated through professional software used for fairing separation calculations.. II. FAIRING SEPARATION MECHANISM Jettison is started by lateral separation mechanism that unlocks the fairing bottom from launch vehicle. Then pyro- bolts of longitudinal separation mechanism are exploded that separate both halves of fairing, followed force is released by compressed springs which rotate the halves about hinges. After reaching certain angle, fairing starts free fall rotation about its center of gravity. Fairing and hinge structures under consideration are illustrated in Fig. 1(a) and 1(b). Fig. 1(a): Fairing structure 261 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 271. Fig. 1(b): Fairingmodel along with spring mechanism and hinge structure III. FAIRING SEPARATION DYNAMICS Dynamic equations are derived for 1-DOF system. For analytical derivation, fairing is considered as rigid body. During separation, hinge point's linear movement is neglected. Frame moves with constant linear acceleration fixed on launch vehicle, is taken as reference [7]. Pitch angle of launch vehicle is taken as zero. Free body diagram of left half of fairing before fairing separation is illustrated in Fig. 2. Free body diagram of left half of fairing during separation is shown in Fig. 3. Fig. 3: Free body diagram of left half of fairing before separation Fig. 4: Free body diagram of left half of fairing during separation For left half of fairing, summation of all the moments about the hinge point is given by: (1) After re-arranging (1); (2) Where, 'Izz ' is the moment of inertia of left half about hinge. 'Fs ' is the magnitude of force applied by all springs. 'ds' is the moment arm of spring force 'Fs '. 'Mfair' is the mass of fairing half. 'a' is the linear acceleration applied on the fairing due to instantaneous thrust of launch vehicle, which is considered as constant during separation. ' ' is the angle of rotation of fairing half. 'Fs' is calculated using equation given below. (3) Where, 'K' is stiffness of one spring. 'Xtot' is the total compressed length of a spring. 'Ns' is the number of springs. In given case, 2 springs are considered in separation mechanism. 'Xdef' is the instantaneous uncompressed length of spring during separation and it is found using cosine law of triangle given by, (4) Equations (1) to (4) are coupled through ' '. These inter- linked dynamic equations are solved by using Range-Kutta method. Same is used for dynamic solution of right half of fairing with trivial modifications. Theoretically, force exerted by the spring system 'Fs ' on the base ring and the reaction force 'Fr ' generated by the ring on the spring system is equal. But in practical, "action is not equal to reaction" due to inelastic contact of two bodies [8-9]. Spring impingement on the base ring causes the absorption of energy in the ring. Amount of energy absorbed depend on the materials of spring and ring. Thus reaction force produced by ring will always be less than the spring force. Harder material absorbs less energy and vice versa. Force loss factor 'e' is considered to make analytical solution closer to real results. Therefore 'Fr ' is used instead of 'Fs ' in equation (2). (5) Equations (1) and (2) are valid prior to fairing separation from hinge. After fairing separation, analogous equations are required to derive for motion of fairing about its center of gravity. IV. RESULTS AND DISCUSSION Equations are implemented and solved in MATLAB. Spring parameters including spring stiffness, spring material and compressed length, are evaluated to satisfy given separation requirements given in table 1. 262 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 272. TABLE 1: FAIRINGSEPARATION REQUIREMENTS Fairing Jettison Angle 70 deg Fairing Jettison Time < 0.8 sec Typical values of system parameters used in developed methodology are given in table 2. TABLE 2: SYSTEM PARAMETERS Mass of Fairing Half 137 kg Moment of Inertia of one half about hinge 700 Nm2 Location of Xcg 1.75 m Location of Ycg 0.39 m Linear Acceleration of Fairing due to Thrust 46.44 m/sec2 Gravitational force at Separation Altitude 9.464 m/sec2 Moment Arm of Spring 0.61 m Earlier discussed spring loss factor is evaluated from professional mechanics software. Values for different materials of spring and base ring in-contact are given in table 3. TABLE 3: SPRING FORCE LOSS FACTOR Material of Spring Material of Base Ring Force Loss Factor Aluminum Aluminum 0.09 Aluminum Steel 0.012 In our case, 0.012 is used because material considered for spring is aluminum and for base ring is steel. Different cases are simulated and analyzed for various spring constants and spring compressed lengths given in table 4 and 5. TABLE 4: EFFECTS ON SEPARATION TIME DUE TO VARIATION IN SPRING COMPRESSED LENGTH S. No Spring Stiffness (N/m) Spring Compressed Length (m) Max. Reaction Force 'Fr' on each half of Fairing (N) Separation Time (sec) 1 100,000 0.04 7904 No Separated 2 0.05 9880 No Separated 3 0.06 11856 1.1 4 0.07 13832 0.89 5 0.08 15808 0.78 TABLE 5: EFFECTS ON SEPARATION TIME DUE TO VARIATION IN SPRING STIFFNESS S. No Spring Stiffness (N/m) Spring Compressed Length (m) Max. Reaction Force 'Fr' on each half of Fairing (N) Separation Time (sec) 1 70,000 0.08 11066 0.96 2 80,000 0.08 12646 0.88 3 90,000 0.08 14228 0.83 4 100,000 0.08 15808 0.78 On the basis of results, spring having stiffness of 100,000 N/m and compressed length of 0.08m is selected since it fulfills fairing separation requirements with sufficient margin. Fig. 5 shows the effect of stiffness and compressed length of spring on separation time of fairing. Fig. 2: Effects of Spring Parameters on Separation Time For validation, results computed by analytical method are compared with professional fairing separation software. Fig. 6 and 7 authenticates analytical method. Fig. 3: Angular Velocity about Z-axis of Left Half of Fairing Fairing's half separated from hinge 263 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 273. Fig. 4: ReactionForce of Left Half of Fairing Fairing's opening angle is presented in Fig. 8. At 0.78 sec, fairing achieved rotation of 70o so it is released from hinge, satisfying the requirements given in table 1. Fig. 5: Opening Angle Theta of Left Half of Fairing Fig. 9 presents the movement of fairing half in body frame. Red curve shows the cg variation of fairing's half during separation. Fig. 6: Fairing's half Movement during Separation V. CONCLUSION Results show that designed analytical approach for fairing separation dynamic analysis is fairly close to professional software approach. Inclusion of material properties of spring and base ring made results more comparable. Among various springs of different stiffness and compressed lengths, best one (having stiffness of 100,000 N/m and compressed length of 0.08m) is chosen on the basis of given fairing separation parameters. 3dof and 6dof systems will be considered; modeled and simulated in future that will provide complete dynamics of fairing separation along all axes. ACKNOWLEDGMENT We would also like to show our gratitude to Dr. Imran Afzal for providing insight and expertise that greatly assisted the research. REFERENCES [1] Steven J. Isakowitz, Joshua B. Hopkins, Joseph P. Hopkins Jr., “International Reference Guide to Space Launch Systems,” 4th ed., AIAA, pp. 341, 2004. [2] “Soyuz User's Manual,” Ariane Space, vol 2 rev. 0 ,March 2012. [3] “Atlas Launch System Mission Planner's Guide, Atlas V Addendum(AVMPG),” International Launch Services, pp. 2-7, Dec. 1999. [4] Antonio Pagano, “Global Launcher Trajectory Optimization for Lunar Base Settlement,” Delft University of Technology, May 2010. [5] spacenews.com,‘Orbitals-glory-launch-failure-review-nears-conclusion’, 2011. [Online]. Available: http://spacenews.com/orbitals-glory-launch- failure-review-nears-conclusion/. [Accessed: Aug. 9, 2015]. [6] Choong-Seok Oh, Byung-Chan Sun, Yong-Kyu Park and Woong-Rae Roh, “Payload Fairing Separation Analysis Using Constraint Force Equation,” International Conference on Control, Automation and Systems, Gyeonggi-do, Korea, Oct. 27-30, 2010. [7] Raymond H. Schuett, Bruce A. Appleby, Jack D. Martin, “Dynamic Load Analysis of Space Vehicle System ,Launch and Exit Phase,” General Dynamics Convair Division, Report No. GDC-DDE66-012, June 1966. Fairing's half separated from hinge Path of CG [9] en.wikipedia.org, 'Inelastic Collision', 2015. [Online]. Available: https://en.wikipedia.org/wiki/Inelastic_Collision. [Accessed: Aug. 8, 2015]. [8] W. Bauer, G.D. Westfall, “Physics for Scientists and Engineers,” 3rd ed., McGraw Hill, 2008. 264 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 274. Abstract— In thispaper, a robust attitude control scheme for satellite launch vehicle (SLV) is presented that is based on inverse dynamics control augmented by sliding mode neural network compensator. In this proposed scheme, the inverse dynamics control acts as the primary control agent that computes the control moment required to not only stabilize the launch vehicle but also to implement the steering commands obtained from guidance scheme. This control action is derived solely by the inversion of launch vehicle dynamics; therefore, it is vulnerable to errors resulting from the uncertainties, simplifying assumptions, nonlinearities and various local and environmental disturbances all of which cannot be accurately modeled. This problem is addressed by the secondary control agent of the proposed control scheme i.e. a sliding mode neural network compensator. This nonlinear compensator is independent of the system dynamics and therefore ensures stability against the un-modeled dynamics and the disturbances acting on the launch vehicle. Along with this hybrid control architecture, a guidance scheme is also developed that reshapes the reference attitude profile on the basis of any deviation from the desired trajectory of the launch vehicle so that the desired orbit is reached. Lyapunov stability criterion is incorporated in the control loop to guarantee system stability and the efficacy of guidance and control scheme is verified on a six degree of freedom simulation model developed exclusively in Matlab/Simulink® using reference data for a four stage launch vehicle. In these simulations, the launch vehicle is subjected to various severe combinations of disturbances and parametric variations and the control system successfully compensates all the disturbances and efficiently tracks the desired attitude profile and attains the targeted orbital parameters. Keywords—Launch Vehicle Control, Dynamic Inversion, Inverse Dynamic Control, Guidance and Control, Sliding Mode Neural Network Compensator NOMENCLATURE m = Mass of vehicle α= Angle-of-Attack β= Side slip angle δ= Deflection angle J= Moment of inertia ω = Angular rate g = Gravitational acceleration P = Thrust force xr = Dist. from nose to nozzle center of gravity xd = Dist. from nose to vehicle center of pressure zr = Dist. from central line of vehicle body to parallel central line of thrust pipe xz = Dist. from nose to vehicle center of gravity V = Air stream velocity of vehicle C = Side force L= Lift force D = Drag force φ= Pitch angle ψ = Yaw angle γ = Roll angle θ = Flight Path Angle 𝜗 = Local flight path angle σ = Heading angle I. INTRODUCTION The access to space in a reliable and cost effective way is a critical requirement these days. This ability depends largely on the performance of the guidance and control system of the launch vehicle. The classical controllers for the ascent phase of SLV depend heavily on the accuracy of its dynamic modeling in order to perform as anticipated. Because, it may not satisfy the stability and performance requirements if the SLV is outside of its desired trajectory envelop. But, due to the various uncertainties, nonlinearities and disturbances acting on the system, the dynamics of SLV cannot be precisely modeled. Therefore, the controller needs to have the ability to perform even with inaccurate dynamic model. The controller of a launch vehicle has to satisfy three often contradictory requirements that are to stabilize the vehicle, reduce trajectory deviations by efficiently implementing the steering commands from guidance scheme and to minimize the angle of attack in the high dynamic pressure region to avoid structural overload. During the last few years, a lot of work has been reported in the field of adaptive and nonlinear control of launch vehicle [1] [2] [3] [4] [5]. Such techniques have the adaptability to adjust their output in accordance with the varying dynamics of the system while retaining the desired performance and stability. Similarly, various techniques have been devised to enhance the robustness of the inverse dynamic controller [6] [7] against modeling uncertainties. Launch Vehicle Control based on Dynamic Inversion with Sliding Mode Neural Network augmentation Saqib Alam SUPARCO, Pakistan e-mail: alam.saqib@gmail.com Syed Minhaj un Nabi Jafri SUPARCO, Pakistan e-mail: jafri_minhaj@yahoo.com 265 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 275. This research isaimed at demonstration of the simplicity and effectiveness of inverse dynamic controller when it is aided with a sliding mode neural network compensator in developing the autopilot of SLV. The advantage of using this architecture is that the primary control produces just enough control moment that is required for vehicle stabilization and trajectory following resulting in a simplified and efficient control law. Whereas all the additional factors are taken care of by the secondary control, hence ensuring robustness and adaptability. The weights of the sliding mode neural network are tuned adaptively in synchronization with the changing environment. In order to inject a satellite in a desired orbit, an external guidance loop is also required alongside the inner control loop to provide the steering commands to the controller on the basis of trajectory deviations that may occur due to environmental and local disturbances. A simplified guidance scheme is proposed here for successful accomplishment of orbital parameters. The paper is organized as follows. Section II illustrates the mathematical modeling of SLV. Section III describes the design of inverse dynamic control with sliding mode neural network compensator for attitude control of SLV along with the guidance loop. The simulation results are presented in Section IV and Section V provides the concluding remarks. II. DYNAMIC MODELING OF SLV To analyze ascent flight trajectory, 6DOF simulation model is developed in Simulink MATLAB in which the data of four stage SLV is incorporated. The symbols definition of SLV is depicted in Fig.1. Fig. 1 Symbol’s definitions Table I depicts the four stage data used to simulate the trajectory of SLV.[1] TABLE I. LAUNCH VEHICLE DATA Parameters 1st stage 2nd stage 3rd stage 4th stage Liftoff Mass (ton) 31.34 9.31 2.87 0.91 Fuel Mass (ton) 18.76 5.55 1.72 0.54 Thrust (kN) 593 262 94 30 Mass Flow rate (kg/s) 265 103 37 13 Burn Time (s) 70.7 53.8 46.9 46.9 Stage Dia. (m) 1.3 1.0 0.7 0.5 The launch vehicle experiences various aerodynamic forces and moments during its flight that act as disturbances and cause trajectory deviations. In order to simulate these effects, trajectory of 500 km orbit is considered with injection velocity of 7612 m/s2 , DATCOM is used to compute aerodynamic coefficients, mass flow rate and thrust are considered as constant. The equations of motion are formulated considering the rigid body dynamics, and are given as follows [2]: sin cos 4 cos cosmV mg P D (1) cos 4 sin cos 2 cos mV mg P P L (2) sin sin 4 sin cos2 C mV mg P P (3) 2 ( )y z y zdx m kr x x x x J Jm qS lPz J VJ J (4) 2 ( ) ( ) 2 ( ) dy m kz x d z y x z y y y R z y m qS lJ J C x x J J VJ P x x J (5) 2 ( ) ( ) 2 ( ) x y d z dz m k z y x z z z R z z J J L x x m qS l J J VJ P x x J (6) Where, III. DESIGN OF IDSNN CONTROL For Launch vehicle, the proposed control architecture 266 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 276. consists of feed-forwardterm based on Inverse Dynamics of SLV and feedback term in which sliding mode based neural network compensator is employed. A. Problem formulation In order to apply dynamic inversion to on the attitude dynamics of launch vehicle, it is assumed that the inertial rates , ,p q r are equivalent to body rates , , neglecting the effect of earth rotation. Therefore, the dynamic equations 4-6 can be written as: (7) Where, , , , , , , represent aerodynamic moments, damping moment and control moment respectively. B. Inverse Dynamic Control The dynamic inversion methodology to formulate control input diu requires two steps. The first step is the state and input transformation to convert nonlinearities of the system dynamics into its linear time invariant form: z az bv (8) The second step is the designing of the control input v using linear control methodology. To achieve stable tracking control, three variables of interest i.e. ( , , )diy are chosen as output variables. Transform system to get the form as shown in Equ.8, Output vector is differentiated twice to get input vector [8]: di di di diy M N u (9) Where, T diy , 0 0 ( ) 0 0 0 0 0 ( ) 0 0 0 0 0 0 diM 0 0 0 0 0 0 diN and diu The desired input to the system can be written as: 1 ( )di di di diu N M v (10) Where, ( )dI dIy v v v v (11) To design the linear controller to control the output states ( , , )diy , the linear controller is written as: ( ) ( ) ( ) ( ) ( ) ( ) d d d p d DI d d d p d d d d p d v k k v v k k v k k (12) C. Feedback term based on Sliding mode Neural Network In practice, it seems to be impossible to model the exact dynamics of the system. Since the dynamic inversion control explicitly utilizes the modeled dynamics, it is assumed that 0 ,aero 0damp and 0con represent aerodynamic, damping and control moment of the nominal model respectively. The difference in actual and nominal model is represented by term for which the moments can be written as: 0 0 0 , ,aero aero aero damp damp damp con con con (13) By substituting these values, Equ. 7 can be written as: 0 0 0 0 0 0 0 0 ( , ) ( , , ) ( , , ) damp con damp con aero damp con aero damp con aero damp con aero damp con f f f (14) In order to estimate the function ( , )damp con f , for roll and ( , , )aero damp con f for pitch and yaw, adaptive Sliding mode Neural Network compensator (SNN) is proposed [9]. By employing this, the control deflection computed by the proposed controller is written as: ˆ(.)diu f (15) Radial Basis Neural Network compensator is designed on the basis of sliding surface and the surface depends on error and its rate. RBF networks are adaptively used to approximate the uncertain f. ( )T i if w h s 267 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 277. Where h(s) isthe Gaussian function of neural network based on sliding surface. 2 2 exp , 1,2,...5. 2 i i i s c h s i (16) Fig. 2 SNN structure The sliding surface is defined as:[8] *S C e e (17) Let the nonlinear system is defined as: ( , )x f x x gu dt We know that s ce e ds e ce x x ce (18) ˆ ( )dx fx gu dt ce ksign s The control input can be written as: 1 ˆ ( )du x fx dt ce ksign s g ˆ( ) ( ) ( )s f x f x dt ksign s ( ) ( )s f x dt ksign s (19) Where ˆ ˆ( ) ( ) ( )T T T f f f w h s w h s w h s Lyapunov function is defined as: 21 1 2 2 T L s w w (20) The derivation is written as: T L ss w w ˆ( ( ) ) ( ( ))T L w sh s w s dt ksign s (21) Where ε is approximation error of neural network and γ is a positive coefficient. The stability Criteria is defined as: 0LL The adaptive law is selected as 1 ˆ ( )w sh s Now we can write: ( ( )) ( )L s dt ksign s s dt k s (22) Since approximation error is sufficiently small and k dt We get 0L D. External Guidance loop: In order to achieve the desired orbital parameters, an external guidance loop is incorporated in the autopilot of SLV. The guidance scheme performs the online reshaping of the reference attitude profiles in the exo-atmospheric phase. It computes deviations (Δx, Δy, Δz) from the intended trajectory and on the basis of these deviations, it formulates the new reference attitude angles accordingly. The controller in turn tracks the updated reference profiles to attain the desired orbit. dx x x (23) dy y y (24) dz z z (25) 1 tan y x (26) 1 tan z x (27) d ref (28) d ref (29) The complete block diagram of the proposed control algorithm with external guidance loop is depicted below: 268 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 278. Fig. 3 Guidance& Control architecture IV. SIMULATION RESULTS The performance evaluation of the proposed control architecture was carried out on a six degree of freedom simulation software developed in MATLAB Simulink. The simulation result of pitch channel for the first 30 seconds of flight including the pitch over phase under nominal flight conditions is shown in Fig. 4. Fig. 4 Pitch over Attitude profile with Compensator It can be seen from Fig. 4 that the proposed controller is efficiently tracking the reference pitch angle profile. The robustness analysis against parametric variation of the control scheme is carried out by the incorporation of 30% random variation in mass properties of the SLV. Fig. 5 Attitude profile without Compensator with Uncertainty Fig. 6 Attitude profile with Compensator with Uncertainty 0 5 10 15 20 25 30 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Time [Sec] Deflection[Deg] Without compensator With compensator Fig. 7 Comparison of Commands with Uncertainty Fig. 5-7 show the comparison of controller performance with and without the sliding mode neural network compensator and their resultant outputs under parametric variations. It can be seen that the hybrid control structure is immune to modeling 269 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 279. uncertainties. The effectiveness ofthe proposed guidance scheme with online trajectory reshaping is depicted in Fig. 8 that shows the error in orbital attitude with and without guidance under parametric variations. The guided SLV accurately attains the desired orbit. Fig. 8 Injection Altitude with uncertainties Fig. 9 shows the pitch profile for the whole mission from lift-off to orbit injection including online attitude reshaping from guidance scheme to attain the desired orbital parameters. 0 100 200 300 400 500 600 -50 0 50 100 Time [sec] PitchAngle[deg/sec] Actual Reshaping Reference Fig. 9 Pitch profile with uncertainties V. CONCLUSION An adaptive and robust control strategy for SLV autopilot is presented in this paper that is based on inverse dynamic control aided with a sliding mode neural network compensator. Lyapunov stability criterion is utilized to ensure the stability of the control algorithm. The simulation results prove that the proposed controller not only guarantees stability but also efficiently tacks the reference attitude profile and the steering commands from the guidance loop even under parametric variations. The guidance scheme performs the online reshaping in the exo-atmospheric phase of flight in case of deviations from the intended trajectory to meet the desired orbital parameters accurately. REFERENCES [1] Uzair Ansari. "Hybrid Genetic Algorithm fuzzy rule based guidance and control for launch vehicle", 2011 11th International Conference on Intelligent Systems Design and Applications, 11/2011 [2] Ansari, Uzair, Saqib Alam, and Syed Minhaj un Nabi Jafri. "Trajectory optimization and adaptive fuzzy based Launch Vehicle attitude control", 2012 20th Mediterranean Conference on Control & Automation (MED), 2012. [3] Filho, W. C. L., "Application of Adaptive Control Sounding Rocket Attitude", Proceedings of International Conference on Intelligent Autonomous Control in Aerospace, Beijing, China, (1995). [4] Zhihua Qu, “Robust Control of Nonlinear Uncertain Systems” John Wiley & Sons, Inc. 1998. [5] Yuri Shtessel, James McDuffie, Mark Jackson, Charles Hall, Don Krupp, Michael Gallaher, and N. Douglas Hendrix, "Sliding Mode Control of the X33Vehicle in Launch and Re-entry Modes," AIAA98-4414, Proceedings of AIAA Guidance, Navigation, and Control Conference, Boston, MA, August 10-12, 1998 [6] Ito D, Ward DT and Valasek J. Robust dynamic inversion controller design and analysis for the X-38. In: AIAA conference on guidance, navigation and control, Canada, 6–9 August 2001, paper no. AIAA-2001- 4380. [7] David B. Doman and Anhtuan D. Ngo “Dynamic Inversion Based Adaptive/Reconfigurable Control of the X-33 on Ascent” Journal of Guidance Control and Dynamics, 2002. [8] Abhijit Das, Frank L. Lewis and Kamesh Subbarao (2011). Sliding Mode Approach to Control Quadrotor Using Dynamic Inversion, Challenges and Paradigms in Applied Robust Control, ISBN: 978-953-307-338-5. [9] Feng G (1995), “A compensating scheme for robot tracking based on neural networks”. Robot Auton Syst 15(3):199–206. 270 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 280. Design and Developmentof Low cost Motor Drive for Hub Wheel based Electric Vehicles Syed Wasif Ali Shah SUPARCO-Karachi-Pakistan wasif_iiee@yahoo.com Muhammad Asim SUPARCO-Karachi-Pakistan m_asim2000@hotmail.com Abstract—Robotic mobility platforms are serving a wide range of civil, commercial, domestic and military applications. This paper describes the design and development of low cost motor drive, applicable for wide range of permanent magnet DC (PMDC) motors for electric vehicles and especially for the hub wheel motors as they exhibit a distinctive frequency behavior. This controller comprises of 3 units: First, a logic unit providing control signals; Second, a sensing unit for over current and batteries protection: third, a high power switching unit for driving motors. Though applicable to wide range of unmanned ground vehicles (UGVs) but here it is implemented in a differential drive autonomous UGV built on hub motor wheels with a speed of 1.88 m/sec, endurance of 2 hours and payload capacity of more than 100 kg. Features, performance and development cost of the drive in comparison with the similar commercially available drives is also discussed. Keywords— Mobility platforms; UGV; Hub Wheel; Motor Drive; I. INTRODUCTION Unmanned ground vehicles (UGVs) are the mobility platforms that have enormous civil, commercial, military, domestic and urban applications [1][2] ranging from rescue, surveillance, reconnaissance, mine detection, fire fighting, material handling, agriculture, handicapped assistance, border security, special weapons and tactics (SWAT) robot, manipulation, exploration, automation, transportation, handling in dangerous environment and many others. UGVs may be based on different propulsion systems like engine based, electric or hybrid drive mechanism [3]. Similarly different applications and features require different traction and locomotion systems [4] depending upon the required speed, tracks, obstacle traversing, energy efficiency, load bearing capacity, mechanical and control complexity and many others. To meet different applications and fulfill different tasks through UGVs, varieties of motorized actuation systems are available and applied; also different mechanical structures and their level of complexity have different actuation systems [5].DC motors amongst them are most widely used actuator in electric vehicles due to their low cost, simple and economical control and weight to performance ratio [6].Varieties of DC motor drives are available having different options, ratings and control interfaces [7][8][9]. The motor drive described here has the ability to drive wide range of permanent magnet brushed DC motors with power rating of up to 1.8 KW. Particularly we have developed a UGV shown in fig: 1 with the dimensions of 72 cm in length, 91 cm in width and 20 cm in height, having weight of 90 kg and payload capacity of more than 100 kg build on the Hub motorized wheels. This UGV is primarily designed for autonomous transportation with GPS navigation with maximum speed of 1.88 m/sec. The UGV has the differential drive mechanism with zero turning radius. The motor drive has shown the ability to smoothly control the hub motor’s torque, rpm and direction. The advantage of hub wheel motor in building the UGV is its mechanical simplicity and installment, compactness giving more room to other components, less complex structure of motion transmission eliminating axels and drivelines, reduced weight giving more efficiency and endurance and ability of high torque at low rpm [10]. The implemented hub motor has the following major parameters given in tab-1. Table 1: Hub Motor Parameters Parameter Value Power 250W Voltage 24 Vdc RPM 120 Radius 6 in Figure 1: The differential drive UGV build with Hub motorized wheels 271 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 281. The objective ofthe design is to develop a high current low cost PMDC motor driver primarily for heavy robot building and applications since available commercial drives are either very expensive or inadequate for large applications. The presented motor drive can be used for a variety of PMDC motors. Though primarily developed for UGVs but suitable for wide range of industrial and commercial applications like blowers, pumps, valves etc and other actuators. II. DRIVE DESIGN The design of the motor drive is consisted of three main units as shown in block diagram fig-2. Despite these units a voltage regulation circuitry is added to provide stable power to logic unit and optional 12V fan. It’s a high efficient switching regulator which also supplies off the board power through connector to external logic controller eliminating the need of another power source required for microcontrollers, Radio Control (RC) receivers and other signal sources. A. Logic Unit The logic layer bridges the input signal to the power unit to drive motor and simultaneously gets input from sensing unit to monitor current flow through the drive and battery voltage level. This unit accepts the TTL level pulse width modulated (PWM) and logic signals and translates to the control signals for power unit for forward, reverse direction, speed, brake and enable/disable commands. The PWM and intelligence signals are provided externally. The drive can be interfaced with any microcontroller, timing circuitry or simply with a computer easily. A variety of other signal sources can also be interfaced by adding the converter circuitry to PWM like it can be interfaced with RC receivers to operate wirelessly with RC radio, serial, I2C and other interfaces as shown in fig-3. The logic unit is electrically isolated from the power unit to avoid any damage to the control circuitry. Figure 2: The block diagram of the motor drive consisting of three main units. Figure 3: The logic unit accepting virtually all compatible signal sources. B. Sensing Unit As shown in the block diagram fig: 2, sensing unit monitors the current being drawn from the power source to the motor through power unit as well the voltage level of the attached batteries. Whenever the vehicle stucks in the rough terrain or passing through the uneven surfaces, it leads to draw the extra amount of current or even stall conditions which may damage the motor and the power unit of the drive. So it ensures the safety by limiting the current to the motor as adjusted by the user. The hall sensor is employed to sense the current. Upper maximum limit can be set in this unit above which it gives the signal to the logic layer which forces all the output levels to zero regardless of the input signals and turns off the operation. It also checks the batteries voltage level and perform the same action to turn off the drive if it falls below the certain level to avoid damage to the system and batteries life. A graph of trip voltages against the adjusted peak current is shown in fig: 4. C. Power Unit High power fast switching unit is the last stage of the drive. It receives direction, speed and brake commands and 13-50 volt power and can pulsate it at max up to 16 KHz rate. In result it drives the motor with max of 160 A of continuous current. The block diagram of the power unit is shown if fig-5. Figure 4: Trip voltage against the adjusted peak current. 272 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 282. The FET driveraccepts signals from the logic unit and drives the FET bridge circuitry. It can be seen that block diagram is in full bridge configuration. The bridge is made of the N- Channel power MOSFETs as shown in one of the four legs in fig: 6. FET driver has built in voltage booster. It outs four signals to drive the full bridge circuitry and accepts the modified TTL signals allowing a variety of signal sources to drive. The FET bridge driver provides voltage above the positive voltage source in to the gate of high side FETs through H1 and H2 to turn on the MOSFETs. FETs usually are sensitive to voltage levels. A higher or lower level of voltage may damage it. Therefore Zener diodes are added in the circuit as shown in fig: 6 to protect the gates of MOSFETs. Two Zener diodes are used to clip the upper and lower level of voltages. Also these diodes clip the fast charging and discharging gate capacitance due to transistors switching. Figure 5: The high switching MOSFET power unit Single to several MOSFETs can be used in one leg of the bridge depending upon the motor current requirement. A single MOSFET used here can bear up to 40 A of continous current hence four can stand the load of 160 A continous. This drive can be used in parallel configuration to drive the even more higher power and current motors and other loads upto 300 A without much worrying about dissipated heat management. A small resistor is added between the FET and driver to limit and control the possible large gate current flow due to multiple MOSFETs used in single leg as shown in fig: 6. Schottkey diodes are placed in parallel with the gate resistor, starts to conduct when the FETs are turning off by enhancing its time and eliminates the possiblilty of shoot- through situation. Motors and other inductive loads creates problem for the drives due to voltage spikes and noise of brushes and commutators. Transient voltage suppressores (TVS) are sued as the zener diodes to safely handle the high current voltage spikes coming from motors. TVS are connected parallel to battery voltages to clip the spikes across the board. Additional filteration of the power source is done Figure 6: Power unit bridge circuitry showing of the leg. with the electrolytic capacitors and snubber circuits. Optional fan output of 12 V is provided to connect the cooling fan to manage the dissipated heat through the MOSFETs due to high current. III. HUB MOTOR FREQUENCY RESPONSE The most commercially available motor drives don’t provide the option to change the frequency of the switching power to the motor. They change polarity to alter the direction and pulse width to vary the speed at fixed frequency as it usually not required, whereas the hub motor has a distinct frequency behavior as compared to other PMDC motors. It consumes different power at different switching frequencies. Under no load conditions, the motor draws higher current at low switching frequencies which even increases till 1.5 kHz but drastically decreases up to 4 kHz and then gradually further decreases and almost becomes steady at higher frequencies as shown in fig: 7. The drive features the variable frequency (to be provided by the external controller) up to 16 kHz depending upon the number of MOSFETs used in the power unit. Higher the numbers of MOSFETs limit the maximum switching frequency. IV. DEVELOPMENT COST Since the drive is designed to achieve a low cost dc motor controller while differential drive UGVs require two parallel motors to be excited. Hence two drives are employeed in our plateform. The drive is build from the commercially and easily available components in the local market hence making it serviceable and replaceable. The total development cost of the motor drive is listed in the tab: 2. Figure 7: Frequency response of the hub motor showing current variation at different switching frequencies 273 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 283. Table 2: Costingof the motor drive. Component Appromimate Cost Drive PCB $10 Drive Components $48 Development $7 Total $65 Table 3: Specifications of the motor drive Specification value Supply voltage 13V to 50V Output Current (continuous) 160A Output Current (surge) >400 Switching Frequency Upto 16 kHz (with 4 MOSFESTs) Weight 100 g (without fan) MOSFETs 16 (4 per leg) On Resistance .0026 ohm max at 25C Cooling Through optional 40 CFM 12 volt fan Dimension (mm) 115(L)X80(W)X35(H) Parameters and features of the drive are given in tab: 3. The drive has competitive features and performance if compared with similar commercially available PMDC motor drives whose prices range from $200-$400 [7][8][9]. The drive is thoroughly tested with different motors and loading conditions for different applications. It conforms to the specifications especially under harsh and open environment of robotic mobility platforms for long durations. Different features can also be added in this design with minimal marginal cost which otherwise incurs heavy cost if procured. V. CONCLUSION The paper discussed the design and development of low cost PMDC brushed motor drive. Particularly implemented in differential drive UGV with heavy payload capacity, build with hub motorized wheel showing distinctive frequency response. The developed controller consists of three units: logic unit, sensing unit and high switching power unit. The drive can be driven virtually with any microcontroller and other signal sources like PC and wirelessly with RC signals. It senses the current being drawn and the battery voltage levels and further can be enhanced to monitor other motor parameters like temperature. Features and development cost of the drive are also discussed and compared with similar drives available commercially. The drive is very economical with easy service and replacement. VI. REFERENCES [1] V. Sezer, C. Dikilita, Z. Ercan, H. Heceoglu, A.Buner, A.Apak, M.Gokasan and A. Mugan “Conversion of a conventional electric automobile into an unmanned ground vehicle (UGV)” IEEE international conference on Mechatronics, pp. 564-569, (2012) [2] A. Bouhraoua, N.Merah, M. Al-Dajani and M. Elshafi “Design and development of an unmanned ground vehicle for security applications” 7th International symposium on mechatronics and its applications, pp. 1-6, (2010) [3] LinoGuzzella, Antonio Sciarretta “Vehicle propulsion systems” Introduction to Modeling and Optimization, 2nded. [4] L Bruzzone and G. Quaglia “ Review Article: Locomotion systems for ground mobile robots in unstructured environment” Mech:sci:, 3,49-62, 2012, www.mech-sci.net/3/49/2012/ [5] William Brown “Improving the electromechanical performance of Robot Arms on Unmanned Ground Vehicles” GRRC Technical report 2010-01, 12/20/2009 [6] Nasser Hashemnia and Behzad Asaei“Comparative study of using different electric motors in the electric vehicles” proceedings of the 2008 international conference on electrical machines, paper ID 1257. [7] Roboteq, Inc., “Brushed DC motor controller”, Online: http://roboteq.com, 2015. [8] Robot Power. “Robot power products”, Online: http://robotpower.com, 2015. [9] VEX Robotics. “Victor Motor Controller”. Online: http://vexrobotics.com, 2015. [10] Machine Design. “A look at the advantages of hub motors as well as disadvantages”, Online: http://machinedesgin.com, 2015. 274 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 284. Preparation, Structure andDielectric/Piezoelectric Properties of BiScO3-PbTiO3-Pb(Mn1/3Nb2/3)O3 High Temperature Piezoelectric Ceramics Shahid Karim1 , Jinting Tan1 , Adil Murtaza2 , Shahid Sultan2 , Khalida Kareem2 , Muhammad Nouman Shaikh3 1 Electronic Material research laboratory & International Center of dielectric research, School of Electronics and Information Engineering, Xi’an Jiaotong University, Xi’an 710049, China 2 School of Sciences, Frontier Institute of Science and Technology, MOE Key Laboratory for Nonequilibrium Synthesis and Modulation of Condensed Matter, State Key Laboratory for Mechanical Behaviour of Materials, Xi’an Jiaotong University, Xi’an 710049, China 3 Xi’an Jiaotong University, Xi’an 710049, China Abstract—With the rapid growth of science and technology, more piezoelectric devices should be used in high-temperature environment. Commercially using PZT piezoelectric ceramics have a low curie temperature (<150°C). So the present study is focused on synthesis and characterization of high Curie temperature piezoelectric ceramics. As a kind of excellent high Curie temperature piezoelectric ceramics BiScO3-PbTiO3 (BS- PT) system at MPB reveals high Curie temperature piezoelectric properties. Pb(Mn1/3Nb2/3)O3-PbTiO3 (PMnN-PT) system ceramics have excellent piezoelectric properties at MPB with low Curie temperature (150°C). 0.1PMnN-BSPT ceramics show the optimal piezoelectric properties with d33=200pC/N, kp=0.40, Tc=338°C, Ec=22.75kV/cm and Pr=21.85µC/cm2 . The 0.05PMnN- BSPT ceramics indicate the optimal piezoelectric properties with d33=250pC/N, kp=0.54, Tc=373°C, Ec=14.63kV/cm and Pr=11.51µC/cm2 . Also 0.06PMnN-BSPT system with 64mol.% PT indicates the optimal piezoelectric properties with d33=210pC/N, kp=0.49, Tc=328°C, Ec=20.21kV/cm and Pr=8.35µC/cm2 . So the introduction of PMnN reduces the curie temperature; 10mol.% of PMnN indicates the “hard” behavior as a function of reducing the piezoelectric properties; 6mol.% of PMnN indicates the “soft” behavior with increasing the piezoelectric properties. Index Terms—Morphotropic phase boundary (MPB), Curie temperature, Piezoelectric Properties, High-temperature piezoelectric ceramics, Binary system ceramics I. INTRODUCTION Piezoelectric ceramic is a kind of important functional ceramic material, which can realize the conversion between mechanical energy and electric energy which are widely used in manufacturing, micro displacement transducer and ultrasonic generator etc. With the development of science and technology, a lot of electrical and electronic equipment which are used in high temperature piezoelectric devices increased the demand. Study on the high temperature, with excellent piezoelectric properties of piezoelectric ceramic materials have become a hot research topic of functional ceramics industry. The commercial applications of lead zirconate titanate (PZT) in piezoelectric ceramics range of Curie temperature system is 250-380°C, and the temperature is limited to 150°C or less, which is far less than high temperature piezoelectric devices. For a long time, a lot of piezoelectric crystal materials due to the use of high temperature and excellent piezoelectric properties have become an important part of high temperature piezoelectric applications [4]. Therefore high electromechanical power applications requires such ceramics with better electrical properties such as the piezoelectric strain constant (d33>200 pC/N), low dielectric loss (tanδ) mechanical quality factor (Qm>1000), planar electromechanical coupling factor (kp> 0.5) [1-3]. Lead niobate (PbNb2O6) tungsten bronze structure was first discovered, it has tetragonal tungsten bronze structure [5]. This system has large anisotropy piezoelectric coefficient, quality factor, low and high Curie temperature (570°C) is not easy to depolarization, therefore particularly suitable for the manufacture of high temperature transducer. But pure lead niobate piezoelectric activity difference exists, the electromechanical coupling coefficient is low, and is not easy to sinter, and its high temperature ferroelectric phase at room temperature is not stable. There is irregular grain growth and phase transformation during sintering and cooling process caused by the large volume change, so it is difficult to prepare the excellent performance of the pure lead niobate piezoelectric ceramic materials [6]. Through the substitution of bismuth layer structure piezoelectric ceramic modification can improve its piezoelectric activity. Korzunova [5] in the Bi3TiNbO9 system, using Ti4+ , Ta6+ or W6+ instead of B in the Nb5+ , using K+ to replace Bi3+ , the piezoelectric constant 5 pC/N is increased to 24 pC/N. II. EXPERIMENTAL PROCEDURE Mixture grade (>99.9% purity) powders of Bi2O3, PbO, Sc2O3, TiO2, MnCO3, and Nb2O5 were used to make xPbTiO3- (0.9-x)BiScO3-(0.05 to 0.15)Pb(Mn1/3Nb2/3)O3, (x=0.64). The mixture of MnCO3 and Nb2O5 was ball-milled in a polyethylene jar for 4h by high purity tetragonal zirconia polycrystalline (3Y-TZP) ball media (Tosoh Co., Tokyo, Japan). The mixed slurry was dehydrated and condensed for calcination and phase formation at 800°C, 850°C and 900°C for 5h. The Mn1/3Nb2/3O2 phase was confirmed by X-ray diffractometer (XRD) after calcination, to obtain fine particles the calcined piece was ball milled again for 4 h. After preparation of Mn1/3Nb2/3O2 powder, PbO, Bi2O3, Sc2O3, and 275 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 285. TiO2 were mixedwith it and ball milled to make final compositions. Pellets (12.7x3mm) were compressed under a pressure of 100 MPa by using a hydraulic press. Pressureless sintering was performed in a high purity alumina crucible at 1100°C, 1150°C and 1200°C for 2h. To prevent from reaction with alumina crucible, the specimens were put on a Pt-foil inside crucible. After sintering, apparent densities were measured by the Archimedes method. Electrode formation was done by sputtering silver on the lapped surfaces of pallets. By applying dc field of 3.5 kV/mm for 10 min, electroded specimens were poled at 120°C in silicone oil. Dielectric properties were acquired by measuring the capacitance and loss using LCR meter (HP model 4284). To determine Curie temperature, capacitance measurements were completed as a function of temperature in an automated temperature controlled furnace linked with a computer for data acquirement. The piezoelectric coefficient (d33) was noted from one-day aged samples by using Berlincourt-d33 meter. Electromechanical coupling factor (kp) and mechanical quality factor (Qm) were obtained by using an impedance analyzer (Model HP4194) built on the IEEE standards [12]. Each experimental step is given in Fig. 1. Make the composition of Mn1/3Nb2/3O2 by the molecular weight of MnCO3, and Nb2O5 Calcinate it at 1000'C for 5 hours Mix the powders of Mn1/3Nb2/ 3O2, PbO, Bi2O3, Sc2O3 and TiO2 and take six compositions of 5%, 6%, 8%, 10%, 12% and 15% Calcinate each sample at different temperature 800'C, 850'C and 900'C respectively Blend the powders by mixing with PVA+ Water and take the specific grain size of powders to make pallets Sintered each sample at different temperatures 1150'C, and 1200'C Make the pallets shape smooth and then measure the apparent densities of each sample silver was pasted for the electrode formation The electroded specimens were poled in silicone oil at 120 o C by applying a d.c. field of 3.5 kV/mm for 10 min piezoelectric coefficient d33 was recorded using a Berlincourt d33 meter, dielectric properties were obtained using an LCR meter Ball milled,4h Ball milled,4h XRD XRD Fig. 1. Flow chart of experiment III. RESULTS AND DISCUSSIONS A. Sintering behavior and microstructures Fig. 2, shows Powder XRD for the samples 5, 6, 8, 10, 12 and 15% of PMnN calcined at 800°C, 850°C and 900°C. It is clearly shown at 8% of PMnN the XRD pattern is smooth. Also it is studied that all the samples show tetragonal and rhombohedral phases. Fig. 3, shows sintered XRD for the samples 5, 6, 8, 10, 12 and 15% PMnN which are sintered at 1150°C. There is clearly shown at all doping percentage of PMnN the XRD pattern is smooth. Also it is seen from the angle 44 to 46 at the peak of 200 of all the samples show tetragonal and rhombohedral phases. Fig. 4, shows sintered XRD for the samples 5, 6, 8, 10, 12 and 15% PMnN which are sintered at 1200°C. There is clearly shown at all doping percentage of PMnN the XRD pattern is smooth. Also it can be seen from the angle 44 to 46 at 200 peak, all samples show tetragonal and rhombohedral phases. Fig. 2. Powder X-ray Diffraction Analysis of Samples Fig. 3. XRD of the samples of x=5, 6, 8, 10, 12, 15% respectively sintered at 1150°C Fig. 4. XRD of the samples of x=5, 6, 8, 10, 12, 15% respectively sintered at 1200°C The microstructures of the composites were studied by Scanning electron microscope (SEM). The SEM micrographs of the samples (a) 5%; (b) 6%; (c) 8%; (d) 10%; (e) 12% and (f) 15% of PMnN sintered at 1150°C for 2h are shown in Fig. 5. They showed different microstructures distinctly; grain size of each sample is (a)3.6; (b)2.8; (c)3; (d)3.2; (e)3.7; and (f)2; in micrometers respectively and grain boundaries of sample (c)8% are more clear and significant. The SEM micrographs of the samples 6% of PMnN sintered at (a) 1100°C; (b) 1150°C and (c) 1200°C respectively for 2h are shown in Fig. 6. They showed different microstructures distinctly; grain size of each 276 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 286. sample is (a)1.6;(b)2.8; (c)4; in micrometers respectively. The SEM micrographs of the samples 8% of PMnN sintered at (a) 1100°C; (b) 1150°C and (c) 1200°C respectively for 2h are shown in Fig. 7. They showed different microstructures distinctly; grain size of each sample is (a)1.6; (b)2.6; (c)4.2; in micrometers respectively. The SEM micrographs of the samples 12% of PMnN sintered at (a) 1100°C; (b) 1150°C and (c) 1200°C respectively for 2h are shown in Fig. 8. They showed different microstructures distinctly; grain size of each sample is (a)1.6; (b)3.8; (c)5; in micrometers respectively. It is clearly seen that in all samples of Fig. 6, 7 and 8 there is no significant change in the grain size of samples sintered at 1100°C but the grain boundary is clear by increasing the content of PMnN. Each SEM microstructure of the sintered body did not show any second phase and with increasing temperature, grain size was increased and grain boundaries are more clear and significant. Nevertheless the microstructures of the every composition are uniform. Fig. 5. Sample (a) 5%; (b) 6%; (c) 8%; (d) 10%; (e) 12% and (f) 15% of PMnN sintered at 1150°C respectively Fig. 6. SEM microstructures of Samples 6% of PMnN sintered at (a) 1100°C, (b) 1150°C and (c) 1200°C respectively Fig. 7. SEM microstructures of Samples 8% of PMnN sintered at (a) 1100°C, (b) 1150°C and (c) 1200°C respectively Fig. 8. SEM microstructures of Samples 12% of PMnN sintered at (a) 1100°C, (b) 1150°C and (c) 1200°C respectively B. Bulk Density The density of undoped BSPT and from 5 to 15 mol.% of PMnN of all the samples are given as shown in Fig. 9. It is clearly seen that the density of sample at 5% is more than others. C. Dielectric Properties Fig. 10, illustrate dielectric constant as a function of temperature for BSPT-(5-15%)PMnN ceramics at different frequencies (1kHz~100kHz). It is difficult to see that the 0.64PT content of ceramic samples with diffuse phase transition and frequency dispersion phenomenon, the width of dielectric peaks in the vicinity of the characteristic temperature change with the increase of frequency, the dielectric constant, dielectric peaks shifted to higher temperature. With the increase of PMnN content, the dielectric peaks become sharp and frequency dispersion phenomenon gradually disappeared. Fig. 11, is the medium temperature of BSPT-PMnN ceramics samples in the 1 kHz spectrum. It can be seen that at different content of PMnN, sometimes the Curie peak becomes sharp, dielectric constant, characteristic temperature are sometime increased and sometime decreased. At 5% of PMnN there is highest temperature and highest dielectric constant in all compositions as shown in Fig. 10. D. Piezoelectric Properties The electromechanical coupling factor kp was obtained by using an impedance analyzer. Electromechanical properties are measured from the admittance spectra at low field ac drive are shown in Fig. 12, as a function of PMnN compositions. Near the MPB boundary, 64% PT the sample of 5 and 15% of PMnN there is a significant increase in kp factor, at 12% there is decrease in kp factor as clearly shown in Fig. 12. However the consistency of electromechanical properties lowers considerably while driving at high-temperature. According to high Curie temperature and the coercive field of BSPT, this composition is predicted as high power piezoelectric materials, if we can increase the mechanical quality factor Qm. Near the MPB boundary, 64% PT the sample of 5 and 10% of PMnN there is clearly increment in quality 277 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 287. factor (Qm) asshown in Fig. 13. So, it has been seen that replacement of Pb(Mn1/3Nb2/3)O3 is meaningful to improve mechanical quality factor Qm of BS-PT. All the properties show highest magnitude at compositions near the MPB boundary at 64% PT. There is significant increase in the magnitude of d33 for the samples of BSPT and other compositions of 5, 6, 8, 10, 12 and 15% of PMnN respectively. At these compositions the magnitude of d33 are 320, 250, 210, 220, 200, 74 and 160pC/N respectively as shown in Fig. 14. Dielectric and piezoelectric properties of all samples are given in table I. Fig. 9. Apparent densities at different temperatures 1150ºC and 1200ºC Fig. 10. BSPT-PMnN system ceramics dielectric temperature under different frequency spectrum Fig. 11. Dielectric constant as a function of temperature for BSPT-PMnN ceramics at frequency of 1 kHz Fig. 12. Electromechanical Coupling Factor at different doping percentages 5, 6, 8, 10, 12 and 15% of PMnN Fig. 13. Mechanical Quality Factor at different doping percentages 5, 6, 8, 10, 12 and 15% of PMnN Fig. 14. Piezoelectric constant for all samples at doping %age of 5, 6, 8, 10, 12 and 15 of PMnN 278 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 288. TABLE I. Dielectricand Piezoelectric Properties with Curie temperature Systems d33 (pC/N) Tc (°C) kp Qm εr εmax 5% 250 373 0.54 30 1018 12820 6% 210 328 0.49 15.6 761 4874 8% 220 302 0.42 27.06 1149 5104 10% 200 338 0.40 59.60 931 9809 12% 74 286 0.35 14.5 764 3169 15% 160 326 0.50 13.38 927 10248 E. Ferroelectric Properties Polarization is saturated PS at high electric fields and there is no more motion of the domain walls. Area of the hysteresis curve shows heat density generated in the material at one cycle. It is clearly seen that the 10% of PMnN has high polarization value as well as electric field as shown in Fig. 15. If the residual magnetization is high when magnetic intensity will zero, it means that material have strong value of magnetization. If the value of residual magnetization is less, it means it will not retain the magnetization for long time after removing the magnetic intensity. The results of remanent polarization and coercive field were given in tab. II. When the electric field began to decrease because of the crystal stress domain deviate from the direction of polarization, the electric field is reduced to zero, still remain in the polarization direction of ferroelectric domains, thus the macro reflects the remnant polarization, so that the material is polarized. When the electric field is reversed, the remnant polarization disappears, so the coercive field can be regarded as the minimum field strength of the domain. Fig. 15. Polarization of all compositions with 5, 6, 8, 10, 12 and 15% of PMnN TABLE II. Remanent Polarization and Coercive Field System Remanent Polarization (Pr/µC/cm2 ) Coercive Field (kV/cm) 5% 11.50 14.63 6% 8.36 20.21 8% 10.99 11.03 10% 21.85 22.75 12% 9.42 13.81 15% 9.82 19.78 IV. EFFECTS OF ANNEALING ON DIELECTRIC AND PIEZOELECTRIC PROPERTIES OF BSPT-PMNN CERAMICS Microstructure of ceramics sturdily depends on annealing process and it defines the ceramic’s physical performance. Density, homogeneity and grains size in the annealing steps are significant to attain the preferred electrical properties [7]. It will influence the crystallographic phase, microstructure and electrical properties of the resultant thin films [8-10]. The temperatures of pre-annealing and annealing were determined by the results of the thermal analysis and XRD, respectively. A. SEM Analysis After Annealing The microstructure of the compositions after annealed was investigated by SEM. The SEM graphs of the samples (a) 5%; (b) 6%; (c) 8%; (d) 10%; (e) 12% and (f) 15% of PMnN respectively annealed at 800°C for 2h are shown in Fig. 16. SEM microstructures of the annealed body are not showing any second phase. They have clearly displayed different microstructures; grain size of each sample is (a)3.2; (b)2.8; (c)7.2; (d)4; (e)2.2; and (f)2; in micrometers respectively and grain boundaries of sample (c)8% are more clear and significant. They showed various microstructures clearly with annealed temperature of sample 8% of PMnN, grain size was increased and grain boundaries of sample (a) and (d) are also clear and significant. Nevertheless the microstructures of the every composition are uniform. B. Piezoelectric Properties After Annealing The piezoelectric coefficient d33 was noted from one-day aged samples by a Berlincourt d33 meter (IAAS ZJ-2, Beijing, China). All the properties displayed highest magnitude of compositions near MPB boundary that is 0.64PT. As compared to BS-PT ceramic, there is noticeable decrease in the magnitude of d33 [11]. Also there is noticeable increase in the magnitude of d33 after annealing for the samples of BSPT and other compositions of 5, 6 and 12% of PMnN respectively. At these compositions the magnitude of d33 are 490, 320, 270, 230, 210, 120 and 165pC/N respectively as shown in Fig. 17. The electromechanical coupling factor kp was measured using an impedance analyzer (Model HP 4194) based on the IEEE standards [12]. Electromechanical resonance properties determined after annealed at 800°C for 2h from the admittance spectra at low field ac drive are shown in Fig. 18, as a function of PMnN composition. Near the MPB boundary, 64% PT the sample of 10 and 12% of PMnN there is a bit increase in kp factor, at 5 and 15% there is decrease in kp factor as clearly shown in Fig. 18. The factor Qm was calculated by using an impedance analyzer. As a consequence of the higher amount of coercive field and Curie temperature of BS-PT, this composition is predictable as high power piezoelectric material, if we can increase the value of quality factor Qm. Near the MPB boundary, 64% PT the sample of 5, 10, 12 and 15% of PMnN there is clearly increment in quality factor (Qm) as shown in Fig. 19. So, the replacement of Pb(Mn1/3Nb2/3)O3 is effective to improve the mechanical quality factor Qm of BS-PT system. 279 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 289. C. Ferroelectric PropertiesAfter Annealing There is a dependency of the piezoceramic properties on the achieved polarization level that affects the uniformity of the results due to edge effects, achieved from length poled resonators, long bar and the standard in-plane poled shear plate and the thickness poled resonators; like the thin disk or plates. It is studied that all the results for polarization have increased significantly after annealing temperature. There clearly shown in Fig. 20, that every sample has a large difference in electric field and polarization. And the saturation point of sample at 8% of PMnN is very small and the curve is steep. The saturation of 6% of PMnN has larger value than others. It has been studied that the value of remnant polarization and coercive field were increased conspicuously after annealing temperature. And for the sample 5% of PMnN there is no difference in the value of coercive field before and after annealing. All the extracted data for polarization is mentioned in tab. III. Fig. 16. SEM micro graphs are shown of Samples (a) 5%; (b) 6%; (c) 8%; (d) 10%; (e) 12% and (f) 15% of PMnN respectively after annealing Fig. 17. Piezoelectric Constant after annealing temperature of 800°C Fig. 18. Electromechanical coupling factor of composition of 0, 5, 10, 12 and 15% of PMnN Fig. 19. Mechanical Quality Factor of compositions of 0, 5, 10, 12 and 15% of PMnN Fig. 20. Polarization Characteristics of 5, 6, 10, 12 and 15% of PMnN 280 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 290. TABLE III. RemnantPolarization and Coercive Field after annealing System Remnant Polarization (Pr/µC/cm2 ) Coercive Field (kV/cm) 5% 23.30 -14.66 6% 24.93 -17.02 10% 25.63 -20.55 12% 17.35 -17.01 15% 21.29 -21.17 BSPT 33.98 -17.33 V. CONCLUSION BSPT-PMnN dense ceramic samples were prepared by the columbite pre-production process. XRD results show that, at the appropriate sintering temperatures, two systems were obtained with pure perovskite structure. In systems of 0.05PMnN-BSPT, 0.06PMnN-BSPT, 0.08PMnN-BSPT, 0.1PMnN-BSPT, 0.12PMnN-BSPT and 0.15PMnN-BSPT when the content of PT is 64% is the tetragonal-phase existence of morphotropic phase boundary region. SEM images show that the ceramic sample’s grain sizes are in uniform distribution at the appropriate sintering temperatures, the grain size is about 1µm. When the content of PMnN is low, the PMnN-BSPT system shows the characteristics of typical relaxor ferroelectrics, with the increase of PMnN content, gradually transition from ferroelectric relaxor to normal ferroelectrics. Due to low content of PMnN, PMnN- BSPT system does not show obvious relaxation characteristics. With the increase of PMnN content, the remnant polarization Pr, first increased and then decreased. The coercive field system gradually increases with the increase of PMnN content. The piezoelectric properties of the optimal compositions of 0.1PMnN-BS-PT and 0.05PMnN- BSPT system in the morphotropic phase boundary region are: d33=250pC/N, kp =0.54 and the Curie temperature of Tc=373°C. The piezoelectric properties of the optimal composition of 0.05PMnN-0.95BSPT system in the morphotropic phase boundary region are better than other optimal compositions. After annealing at 800°C for 2h, the resident strain can be quited and the order degree of lattice may be higher than that without annealing process. Annealing process may affect the domain size and domain switch, domain size will be larger after annealing process. The value of Qm increase evidently after annealing process. ACKNOWLEDGMENT I would like to thank my father and teachers for their kind and humble guidance from the very start to the end of the paper writing. I would not have accomplished this without their constant support and feedback; they made it a thorough learning and pleasant experience for me. I admire the guidance and assistance provided by research associates and fellows. REFERENCES [1] Ueha, S., “Ultrasonic motors: theory and applications,” Vol. 29. 1993: Oxford University Press, USA. [2] Chen, H., “The influence of Pb(Mg1/3Nb2/3)O3 content on the piezoelectric and dielectric properties of PMNN-PZT system near the MPB. in Applications of Ferroelectrics,” 2002. ISAF 2002. Proceedings of the 13th IEEE International Symposium on. 2002. IEEE. [3] DU, H. l., “Effect of sintering temperature and composition on microstructure and properties of PMS-PZT ceramics,” Transactions of Nonferrous Metals Society of China, Vol. 16: pp.165-169, 2006. [4] Weis, R. and T. Gaylord, “Lithium niobate: summary of physical properties and crystal structure,” Applied Physics A, Vol. 37(4), pp. 191-203, 1985. [5] Korzunova, L. V., and L. A. Shebanov, “New perovskite-like high temperature ferroelectrics,” Ferroelectrics, vol. 93.1 pp. 111-115, 1989. [6] B. Frit and J. P. Mercurio, “The crystal chemistry and dielectric properties of the Aurivillius family of complex bismuth oxides with perovskite-like layered structures,” Journal of Alloys and Compounds, vol. 188, pp. 27-35, 1992. [7] M. Ghasemifard, “Dielectric, piezoelectric and electrical study of 0.65PMN-0.20PZT-0.15PT relaxor ceramic,” The European Physical Journal Applied Physics, Vol. 54, No. 2 , pp. 20701- 20707, 2011. [8] Laughlin, Brian, Jon Ihlefeld, and Jon‐Paul Maria, “Preparation of sputtered (Bax, Sr1−x)TiO3 thin films directly on copper,” Journal of the American Ceramic Society, vol. 88.9.pp 2652- 2654 2005. [9] Malic, Barbara, “Processing and dielectric characterization of Ba0.3 Sr0.7 TiO3 thin films on alumina substrates,” Journal of the European Ceramic Society 27, pp. 2945-2948, 2007. [10] Tahan, Danielle M., Ahmad Safari, and Lisa C. Klein., “Preparation and characterization of BaxSr1‐x, TiO3 thin films by a Sol‐Gel technique,” Journal of the American Ceramic Society, vol. 79.6, pp.1593-1598, 1996. [11] Chen, Jun, “Temperature dependence of piezoelectric properties of high-TC Bi(Mg1/2Ti1/2)O3-PbTiO3,” Journal of Applied Physics, vol. 106.3 pp.034109-034109, 2009. [12] IEEE “Standard on piezoelectricity,” IEEE Standard 176-1978 (IEEE, New York, 1978). 281 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 291. A Comprehensive Studyon QoS for Mobile Ad Hoc Network Abdur Rasheed Irfan Nazeer Fayyaz Khalid Department of Computer Science Department of Computer Science Department of Computer Science University of Gujrat University of Gujrat University of Gujrat Gujrat, Pakistan Gujrat, Pakistan Gujrat, Pakistan abdur.rasheed@live.com irfan6200@yahoo.com fayyazrao@yahoo.com ABSTACT----A Mobile Ad Hoc Network (MANET) is a collection of nodes which has no centralized administration and without the assistant of base station. All the nodes have equal rights and have their own resources in MANET. QoS is a challenging issue now-a- days due to bandwidth constraints and dynamic topology in MANET. A lot of work done on supporting on QoS in internet and other networks but most of them is not suitable for MANET. Most of work is about the QoS routing, QoS Mac and resource reservation, there is a little work discuss on QoS models and its routing protocols. In this article, we review the QoS models and share some knowledge about QoS routing protocols and other related work on QoS in MANET. The purpose of this article is to give the appropriate knowledge about QoS for MANET at the same platform. Keywords: QoS, Mobile Ad Hoc Networks I. Introduction Wireless network has become a popular network now-a-days. Wireless communication has been making our life easy and comfortable. Wireless communication has been categorized into two primary models; one is traditional wireless network in which a fix infrastructure is used to communicate between one hop to another. In conventional wireless network, much of the nodes are mobiles and connected with the help of fixed backbone nodes using wireless medium(radio waves). In this type wireless network a base station is used for communication. But in some situation (i.e. storm disaster, battle field) there is no fix infrastructure is used. Then we used a specific type of network which is called “Mobile Ad Hoc Network”. A mobile ad hoc network (MANET) [1][2], is a dynamic wireless system in which all nodes are mobile and nodes can move in any direction, independent of each other. According to [3], Mobile Ad hoc Networks are class of networks whose contains characteristics is, Physical characteristics –As MANET is a wireless network in which nodes can create link in any time, bandwidth and Delay due to the nodes are mobile. Organization – As mobile nodes are distributed so resource estimation is not predefined in MANET Dynamic Topology – As the nodes have dynamic topology so it has no fixed architecture. MANET is the combination of different mobiles nodes which does not have any server, router or access points for communication. It has low cast wireless network and easy for deployment. MANET is used in different situations in our life i.e., Military operations, Rescue search operations, communication between mobiles vehicles in smart transportation, mobiles phones, laptops and smart watches etc. As shown in Fig.1 MANET provides a flexible and seamless access to internet. Due to in increasing popularity in MANET, 282 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 292. QoS has becomethe main task of research in MANET. QoS is a challenging issue now-a-days due to dynamic topology and bandwidth constraints. A lot of work is done on supporting on QoS for internet and other networks but most of them are not suitable for MANET. Due to autonomous wireless network, MANET has its own protocols. In this literature, we discuss QoS for MANET includes QoS models and also discuss some important protocols in MANET. There are many researches on QoS on the internet. Due to this, it is difficult to get appropriate knowledge on QoS in a specific location. This paper provides a comprehensive knowledge on QoS models for MANET and some QoS routing protocols. In first part, we shall discuss some QoS models which are specified an architecture in which we discuss some services that can be used in MANET. In second part, we discuss some QoS protocols for MANET. These protocol are divided in the based on route discovery, metrics and network architecture. Then we gives a comparison of these protocols on the bases on route discovery, metrics and network architecture. II. QoS models and MANETs QoS model defines the architecture which provides a platform to different services; which is used in networks. QoS models use different situations like dynamic topology etc. QoS model consider the exiting internet architecture for MANET. In this section, we explain the different QoS models especially DiffServ and one of the newly proposed QoS Model FQMM. A. IntServ and MANETs The idea of IntServ[4] is taken from the wired network architecture and B-ISDN i.e. have the virtual circuit mechanism. Virtual connection is maintains through signaling protocol RSVP (Resource Reservation Protocol)[4].RSVP is very important protocol in all QoS support routers. IntServ provides two types of services; one is Guaranteed [5] Service and the Other is Controlled Load Service [6]. Here some pros and cons of IntServ mentioned below: a) Scalability IntServ provides per-flow granularity so when state information is increases the number of lows is also increased. This requires storing a lot storage capacity and processing is increases, it results a problem that's called scalability problem of IntServ. The scalability problem is very less in current MANETs because there is less number of users, less bandwidth and small size of network. On the other Hand, in future the use of MANETs is increasing rapidly and the size and speed of the MANETs increases so the Scalability problem is also increased with time so IntServ is not suitable in MANETs. b) Signaling Signaling protocols like Resource Reservation protocol have three steps; connection maintenance, connection tear-down and connection establishment. Corson[7] assumes and then suggest that a larger number of links is given to the network to control overhead .For MANETs with link capacities, the overheads of connection maintenance and dynamic topology usually a large for connection establishing so that RSVP signaling protocol is not suitable in MANETs. c) Router mechanisms IntServ is required a large number of routers so that all router have packet scheduler, classifier, admission control routine and RSVP that's why it is undesired able in MANETs. B. DiffServ and MANETs Like ATM and IntServ, the tenet of DiffServ [8] is used a relative-priority scheme to soften the hard requirements of hard QoS models. DiffServ is specially designed for the deficiencies of IntServ and RSVP model like scalability problems in internet backbone [9]. DiffServ is defined limited classes for overcome this problem of IntServ. Much kind of 283 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 293. services are supportedin DiffServ model which are following [10]: a) Premium Service Premium Service is considered to give the end-to-end bandwidth services, low jitter, low loss and low delay [11]. b) Assured Service Assured service is used to provide better reliability for application. Other, it is more qualitative service than quantitative and easy to implement. c) Olympic Service Three types of services are provided in Olympic services which are Gold, Silver and Bronze with decreasing quality [9]. Premium service is not suitable in MANETs because it implementation in dynamic environment is not possible. DiffServ is considered a defined solution to MANETS QoS model because it provides Assured Service and it is lightweight in inner router which is suitable to MANETs. Because of this, DiffServ is designed for wired networks and have some problems in MANETs which are follows DiffServ is not sure in to the boundary nodes for this; it requires a lot of storage capacity. DiffServ provides a Service Level Agreement (SLA) which implementation is not possible in MANETs. SLA is an agreement between customer and the internet service Provider (ISP) to receive Differentiated Services. SLA obeys the partial or whole traffic rules [12]; but in MANETs, it is not possible to negotiate the traffic rules. C. Flexible Quality of Service Model (FQMM) and MANETs FQMM is the first QoS model which is preferred in MANETs. FQMM is proposed in [13] and it tries to take advantage from IntServ and DiffServ models both. FQMM is designed for average size MANETs which comprises less than 50 nodes using at non- hierarchical topology. However, FQMM is a hybrid approach because it has combined features from IntServ and DiffServ. In FQMM infrastructure, it has types of Nodes like DiffServ which follows: Ingress Node Interior Node Egress Node Ingress node is a source node which is a mobile node. Interior node is the intermediate node that sends the data of other mobile nodes. Egress node is the node which is received data and it is the destination node. Note that the role of Ingress node is changing due to change in its position and network traffic. Resource determination and allocation is done to the various mobile nodes by using provisioning in FQMM. Provisioning is at top level and based on utilization of the network in present internet. Provisioning is used to realize ideal per-flow by RSVP in IntServ and is used in DiffServ as per-class by human agents rather than RSVP or other. But on the other hand, FQMM introduced a hybrid scheme as per-flow as well as per class. Traffic with highest priority is given per-flow provisioning while other priority is based on per- classes. Although it is the first efficient model in MANETs but there is further need to study that how sessions is entertained as per-flow due to bandwidth constraints in MANETs. III. QoS Aware-Routing Protocols In this section, we discuss some QoS aware routing protocols. After this, we discuss their functionality and then see their comparisons with each other with different aspects. The QoS Aware-Routing Protocols are follows: A. Optimized link stat routing Protocol OLSR [14, 15] is a proactive routing protocols and the idea of OLSR is given by IETF MANET[3] working group. OLSR is the further study of Link stat routing protocols. The idea of OLSR is based on MPRs (Multi Point Relays).Every MPR have two hops distance which is the part of communicating nodes. If a node has no MPR then it cannot transfer data further. Consider that MPR have larger distance than a node which is not a MPR then we may miss some good quality links and that's reason, it is difficult to achieve a good quality of service in OLSR. 284 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 294. B. Predictive LocationBased QoS Routing Protocol PLBQR [16] is a proactive routing protocol which is used in MANET.PLBQR is totally based on location prediction of a node a particular instant in ad hoc wireless networks. Location prediction is used to find the geographical location of a node at particular instant 'tf' when the receiver node picks a packet in future. The value of 'tf' is calculated by calculating delay in communication. No resources are reserved during the propagation from source to destination but QOS is controlled. This QoS routing protocol is helped the nodes to update location prediction and delay during communication. C. Ticket Based QoS Routing Protocol Ticket Based QoS Routing protocol [17] is a reactive protocol which is used in MANETs. The basic idea of Ticket Based QoS Routing protocols is to use tickets to find the best path which may delay constrained or bandwidth constrained. This protocols is used two types of tickets, one is yellow ticket and the other is green. The purpose of yellow tickets is to find the best path with delay/bandwidth constraints. The other green ticket purpose is to find the low cost route. The source node has a lot the number of tickets to establish a QOS aware route. The tickets may be less or more it depends on the route constraints to find the best path. The Drawback of this routing protocol is that the neighbor nodes incorporated to each other with resource estimation that’s why it requires more memory. When a route is established to the destination node acknowledged even resource reservation is established. D. Ad Hoc QoS on Demand Routing The idea of AQOR [18] protocol is to find the best route which is fulfills the constraints of end to end delay with specific bandwidth. In route discovery, the route request is sends to nearby node and attaches the constraints with the packet and if both nodes satisfy then this request is rebroadcasted to the next hop with some expiry time. If the node is not received the reply in given time then entry will be deleted. If the nodes do not want to delete path the only intermediate nodes will reply. Once the route is found then resource reservation is takes place on the found route. If the route is deleted the new route request is sent. E. Adaptive QoS Routing Algorithm ADQR [19] is a reactive protocol. Adaptive QoS Routing algorithm protocol is used to establish a route between two nodes which have longer life time. In these protocols, the route breakage and route creation is bases on signal power. Bandwidth is considered to the lower layer. Route maintenance contains two ways. One is pre-routing and the other is rerouting. Pre-routing is used when the value of the threshold is greater than signal power and rerouting is used when the signal power is very low. For example of this protocol is; if a node want to communicate with other node then it sends a RREQ packet to nearby node and then intermediate node attached his address to this packet and forward to the next. If the packet is reached to the exact destination then destination node checked that the signal power. Signal power ensures that the route is enough or it disjoints with others .QoS_Reserve packet is sent to the current route back and QoS Acknowledgment is sent to the source. [3] F. Trigger-based Distributed-QoS Routing Protocol TDR [20] is a reactive protocol which is a location based protocol. In this protocol every host node contains two types of databases. One is for local neighbor and other is activity based. Hosts are passed out the mobility and location information and the neighbor nodes is stored the power level in their local database. Activity-based data base is stored the every session routing information and updated in every session. Route discovery is done through the neighboring power level and compared this with threshold time. Route maintenance is done like ADQR protocol [3]. G. Bandwidth Estimation QoS Routing Protocol BEQR [21] is a reactive protocol which is inspired by AODV routing protocol [23].The main objective of this protocol is to give soft QoS for service with the bandwidth constraint. In this two schemes is used one is Feedback scheme and the other is Admission 285 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 295. scheme. In Feedbackscheme is used for updating constraints when a node have no bandwidth its own. Admission Scheme is used for route discovery with satisfied bandwidth constraints. In this protocol bandwidth is calculated with ratio of free and busy times. Resource reservation is not considered. It is design for soft QoS not hard. H. Core Extraction Distributed Ad hoc Routing Protocol CEDAR [23, 3] was proposed by R. Siva Kumar [23].CEDAR is a QoS aware routing protocol and it is a hybrid routing protocol. In this protocol nodes are selected as Core node and they are responsible for route maintenance, computation and QoS providing by using distributed algorithm. Link state information depends on network state whether it is dynamic or stable. Route computation is on-demand based and Cedar contain following three features which is very important: 1. Destination location discovery and establishing the core path to destination 2. Core nodes provides the stable and short path between source to destination and also a core path guideline 3. If the path is break then it re-established as well as typology changes. CEDAR considers that the resource is reserved. [3]The route maintenance in CEDAR is done as source initiated route maintenance or dynamic route maintenance. I. INSIGNIA INSIGIA [24] is a hybrid protocol which is proposed for adaptive services in ad-hoc networks. Adaptive services entertains that types of services which requires minimum QoS guarantee (like minimum bandwidth) and it varies when resources is sufficient. INSIGNIA in-band signaling is the very important component which is entertained fast reservation to deliver the adaptive services[25].Routing module is responsible for finding routes between nodes and sends the packets to the next node(intermediate).In any time of transmission when topology is changed it is established a new path. Every data packet has a optimal QoS field and admission control module is responsible for resource reservation; it is refreshed periodically. If an application is used to want minimum bandwidth then it bandwidth indicator is "MIN", if the application is used to want maximum bandwidth then indicator indicates "MAX". J. Forward Algorithm FA [26] belongs to the on-demand routing protocols which is used on demand algorithms with bandwidth as QoS constraints parameter. FA is the modified version of AODV [22] is used for route discovery and some additional information is attached before sending to the next hope. During route discovery, the value of local maxima is calculated and forward [27]. FA is used in many other reactive protocols like TORA [28] and DSR [29] for route discovery. The Route maintenance is managed by using old routes in it. 286 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 296. IV. Comparison ofQoS Aware-Routing Protocols A comparison of QoS aware-routing protocols is given below: Routing Protocols QoS Support Redundant Route Route Maintenance Route Breakage Prediction Network Architecture Route Discovery OLSR X X X Hierarchical Proactive PLBQR X X Location Prediction Proactive Ticket Based X Flat Reactive AQOR X X Flat Reactive ADQR Flat Reactive TDR X Location Based Reactive BEQR X X X Flat Reactive CEDAR X X Hierarchical Hybrid INGIGNIA X X Flat Reactive FA X X Flat Reactive Conclusion In first part of this paper, we have described three QoS Models (IntServ, DiffServ and FQMM).All three models are suitable for different situations. All three have some pros and cons; but FQMM is a flexible QoS model for MANETs. To our knowledge, FQMM was the first such QoS model proposed for MANETs. It provides a hybrid provisioning scheme and a relative and adaptive traffic profile. In second part of this paper, we described some important QoS aware-routing protocols in ad hoc wireless networks which is an active research now-a-days. In this paper, we discussed some characteristics of the QoS routing protocols and compare the characteristics of these protocols in the form of table. The aim of this review is to give an appropriate knowledge on QoS for MANETs at the single paper. References [1] Loay Abusalah, Ashfaq Khokhar, and Mohsen Guizani “A Survey of Secure Mobile Ad Hoc Routing Protocols” IEEE communications surveys & tutorials, VOL. 10, NO. 4, FOURTH QUARTER, 2008, pp 78-93. [2] Sunil Taneja and Ashwani Kush” A Survey of Routing Protocols in Mobile Ad Hoc Networks”, International Journal of Innovation, Management and Technology, Vol. 1, No. 3, August 2010. [3] Lei Chen and Wendi B. Heinzelman, “A Survey of Routing Protocols that Support QoS in Mobile Ad Hoc Networks”, IEEE Network November/December 2007. [4] R. Braden, D. Clark, and S. Shenker. Integrated Services in the Internet architecture - an Overview. IETF RFC1633, June 1994. [5] S. Shenker, C. Partridge and R. Guerin “Specification of Guaranteed Quality of service” RFC 2212 September 1997 [6] J. Wroclawski, “Specification of the Controlled-Load Network Element Sevice ” RFC 2211, Sept.1997 [7] M.S. Corson. Issues in supporting quality of service in mobile ad hoc networks. In IFIP 5th Int.Workshop on Quality of Service (IWQOS'97),May 1997. [8] S. Blake. An architecture for Di_erentiated Services. IETF RFC2475, December 1998. [9] Xipeng Xiao Nad Lionel M. Ni “Internet QoS: a big Picture,” IEEE Network magazine , March 1999. [10] Kui Wu and Janelle Harms “Quality of Services in Mobile Ad Hoc Networks” [11] K. Nichols, V. Jacobson, and L. Zhang. A two-bit Di_erentiated Services architecture for the Internet. IETF RFC2638, July 1999 [12] S. Blake, “An Architecture for Differentiated Services,” IETF RFC 2475, December 1998. [13] H. Xiao , W.K.G. Seah, A. Lo and K. C. Chua, “A flexible Quality of Service Model for Mobile Ad hoc Networks” IEEE VTC2000-spring . Tokyo, Japan May 2000. [14] Y. Ge, T. Kunz, and L. Lamont, “Quality of Service Routing in Ad-Hoc Networks Using OLSR,” Proc. 36th Hawaii Int’l. Conf. Sys. Sci., Jan. 2003. [15] P. Jacquet, P. Muhlethaler, A. Qayyum, A. Laouiti, L. Viennot, T. Clauseen, "Optimized Link State Routing Protocol draft-ietf-manet-olsr-05.txt", INTERNET-DRAFT, IETF MANET Working Group. [16] S.H.Shah and K.Nahrstedt, “Predictive Location- Based QoS Routing in Mobile Ad Hoc Networks,” Proceedings of IEEE ICC 2002,vol.2, pp.1022-1027, May 2002 287 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 297. [17] S. Chenand K. Nahrstedt, “Distributed Quality-of-Service in Ad Hoc Networks,” IEEE JSAC, vol. 17, no. 8, 1999. [18] Q. Xue and A. Ganz, “Ad hoc QoS On-demand Routing (AQOR) in Mobile Ad hoc Networks,” J. Parallel and Distrib. Comp., vol. 62,no. 2, Feb.2003, pp. 154–65. [19] Y. Hwang and P. Varshney, “An Adaptive QoS Routing Protocol with Dispersity for Ad-hoc Networks,” Proc. 36th Hawaii Int’l.Conf. Sys. Sci., Jan. 2003. [20] S. De et al., “Trigger-Based Distributed QoS Routing in Mobile Ad Hoc Networks,” ACM SIGMOBILE Mobile Comp. and Commun.Rev., vol. 6, no. 3, July 2002, pp. 22–35. [21] Lei Chen and Wendi B. Heinzelman, “QoS-Aware Routing Based on Bandwidth Estimation for Mobile Ad Hoc Networks,. IEEE Journal on Selected Areas in Communications, 23(3):561_572, March 2005. [22] C. E. Perkins and E. M. Royer. “Ad-hoc On-Demand Distance Vector Routing”, In Proceedings of the 2nd IEEE Workshop on Mobile Computing Systems and Applications, pages 90_100, New Orleans, LA, Feb. 1999 [23] R. Sivakumar, P. Sinha, and V. Bharghavan, “CEDAR: a core extraction distributed ad hoc routing algorithm” IEEE Journal on Selected Areas in Communication, Vol. 17, no. 8, pp 1454-65 August1999. [24] X. Zhang, S.B Le, A. Gahng-Seop and A.T Campbell, “INSIGNIA: an IP based quality of service framework for mobile ad hoc networks.”Journal of Parallel and Distributed Computing, Vol. 60, no. 4,pp. 374-406, April 2000. [25] C.Siva Ram Murthy,B.S.Manoj, “Ad Hoc Wireless Networks Architectures and Protocols” Prentice Hall Communications Engineering and Emerging Technologies Series. [26] Chenxi Zhu and M. Scott Corson. “QoS routing for mobile ad hoc networks”. Technical report, Flarion Technologies, Bedminster, New Jersey 07921, 2002. [27] Philipp Becker, “QoS Routing Protocols for Mobile Ad-hoc Networks – A Survey”, TU KL Technical Report 368/08, A publication by University of Kaiserslautern [28] V.D. Park and M.S. Corson, “A Highly Adaptive Distributed Routing Algorithm for Mobile Wireless Networks”, Proceedings of IEEE INFOCOM 1997, pp.1405-1413, April 1997. [29] D.B.Johnson and D.A. Maltz, “Dynamic Source Routing in Ad Hoc Wireless Networks”, Mobile Computing, Kluwer Academic Publishers,vol.353,spp.153-181,1996. 288 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 298. Comparison of MaximumLikelihood Classification Before and After Applying Weierstrass Transform Muhammad Shoaib, Zaka Ur Rehman, Muhammad Qasim Abstract— The aim of this paper is to use Maximum Likelihood (ML) Classification on multispectral data by means of qualitative and quantitative approaches. Maximum Likelihood is a supervised classification algorithm which is based on the Classical Bayes theorem. It makes use of a discriminant function to assign pixel to the class with the highest likelihood. Class means vector and covariance matrix are the key inputs to the function and can be estimated from training pixels of a particular class. As Maximum Likelihood need some assumptions before it has to be applied on the data. In this paper we will compare the results of Maximum Likelihood Classification (ML) before apply the Weierstrass Transform and apply Weierstrass Transform and will see the difference between the accuracy on training pixels of high resolution Quickbird satellite image. Principle Component analysis (PCA) is also used for dimension reduction and also used to check the variation in bands. The results shows that the separation between mean of the classes in the decision space is to be the main factor that leads to the high classification accuracy of Maximum Likelihood (ML) after using Weierstrass Transform than without using it. Index Terms— Maximum Likelihood Classificaion (ML), Weierstrass Transform, Bayes Theorem, Supervised Classification, Principle Component Analysis (PCA). I. INTRODUCTION Satellite images are widely used for various purposes and with the advancement in Satellite technology every country wants to solve their problem by using satellite images. Satellite images can be used in various disciplines such as Metrology, Environmental sciences, GIS, Geology, Geophysics, Engineering and Construction, Defense and Intelligence and the list goes on. Satellite images contains maximum information about our planet and main problem is that how can we retrieve the useful information satellite images. Researchers can have several problems such how to classify the land cover types, depth of Aerosol particles in atmosphere, Forest Climates, Flood predictions and many more. Researcher solve these problems by analyzing the satellite images. In this article our main aim is to classify the land cover types by using Maximum Likelihood Classification algorithm. First we classify land cover types without using Weierstrass Transform (WT) and then we compare the results by applying Weierstrass Transform (WT). Maximum Likelihood (ML) is a supervised classification method derived from the Bayes theorem, which states that the a posteriori distribution P(i|θ), i.e., the probability that a pixel with feature vector θ belongs to class i, is given by: where P(θ|i), is the likelihood function, P(i) is the a priori information, i.e., the probability that class i occurs in the study area and P(θ) is the probability that is observed, which can be written as: where M is the number of classes. P(θ) is often treated as a normalization constant to ensure ∑P(i|θ) sums to 1. Pixel x is assigned to class i by the rule: Each pixel is assigned to the class with the highest likelihood or labeled as unclassified if the probability values are all below a threshold set by the user [1]. Maximum Likelihood (ML) assumes that the data within a given class i obeys a multivariate Gaussian distribution. Frist we implement Maximum Likelihood (ML) classification algorithm without checking the assumption of normality and then we apply Maximum Likelihood (ML) classification after using Weierstrass Transform on the image data and we have seen that after applying Weierstrass Transform our data is normally distributed. The accuracy after applying Weierstrass Transform (WT) is increased because the distribution of the data is normal because for ML classification data has to be normal. II. STUDY AREA We use Quickbird satellite image for our analysis. Quickbird is high-resolution commercial satellite which is owned by DigitalGlobe and launched in 2001. Quickbird uses Aerospace's Global Imaging System 2000 (BGIS 2000).The Quickbird satellite collected panchromatic (black and white) at 61cm resolutin and multispectral imagery at 2.44-(at 450 km) to 1.63-meter (at 300km) resolution, as orbit altitude is lowered during the end of mission life [2]. The Quickbird satellite image which we use for our analysis is sample image of Rome Italy with coordinates with four bands which are blue (450-520nm), green (520-600nm), red (630-690 nm) and NIR (760-900 nm). 289 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 299. III. METHODOLOGY Methodology isthe systematic, theoretical analysis of the methods applied to a field of study. It comprises the theoretical analysis of the body of methods and principles associated with a branch of knowledge (wiki). The detail methodology is given in the figure 1. Fig. 1. Data Processing flowchart IV. PREPROCESSING Preprocessing is a common name for operations with images at lowest level of abstraction both input and output are intensity images. The main aim of preprocessing is an improvement of the image data that suppresses unwanted distortions or enhances some image features important for further processing. In this paper we have to do classification so we do not need to do any radiometric calibration or anything else. But we have to check the distribution of each band whether it is normal or not which is the main assumption for ML classification. A. Distribution Check We use image histogram to check the distribution of each band. An "image histogram" is a type of histogram that acts as a graphical representation of the tonal distribution in a digital image. It plots the number of pixels for each tonal value. By looking at the histogram for a specific image a viewer will be able to judge the entire tonal distribution at a glance. We make a separate histogram for each band to check the distribution of each band in image. Blue Band 0 2000 4000 6000 8000 10000 12000 14000 Blue Band Histogram 0 50 100 150 200 250 Green Band 0 2000 4000 6000 8000 10000 12000 14000 Green Band Histogram 0 100 200 Red Band 0 2000 4000 6000 8000 10000 12000 14000 Red Band Histogram 0 100 200 NIR Band 0 2000 4000 6000 8000 10000 12000 NIR Band Histogram 0 100 200 Fig. 2. Histograms with their respective bands Preprocessing Distribution Check Weierstrass Transform PCA Training Samples Results ML without WT ML with WT Conclusions 290 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 300. In the figure2 we can see that the distribution of each band is not Gaussian and we know that for ML classification algorithm the distribution of each has to be Gaussian. In this case distribution of first three bands are positively skewed while the histogram of band 4 is looks like Gaussian but not exactly Gaussian. So as our aim is to check the accuracy of classification before making data Gaussian and after making it Gaussian. We use Weierstrass Transform to make the distribution Gaussian. B. Weirstrass Transform As we have seen that distribution of our data is not follow Gaussian distribution so make that Gaussian we have to use transformation to make pixel values Gaussian. Researchers uses various types of transformation to make the data Gaussian. In this paper we use different type of transformation called Weirstrass Transform. The Weirstrass transformation, also known as the Gauss transform, the Gauss Weirstrass transform and the Hille transform, is intimately related to the solution of the heat equation for one- dimensional flow. It is also a special case of convolution-transform, yet it is considered a singular case of convolution-transform theory developed by Hirschmann and Widder [3]. The Weirstrass transformation F(x) of a function f(y) defined as: the convolution of f with the Gaussian function . Instead of F(x) we also write W[f](x). Note that F(x) need not exist for every real number x, because the defining integral may fail to converge [3]. When Gaussian filter modifies the input signal by using Gaussian function; this is known as Weierstrass transformation. We can write Gaussian function mathematically in 2D as follow: where ρ is the correlation between X and Y and where σ_x>0 and σ_y>0. In this case, In the bivariate case, the first equivalent condition for multivariate normality can be made less restrictive: it is sufficient to verify that countably many distinct linear combinations of X and Y are normal in order to conclude that the vector [X Y]′ is bivariate normal. Gaussian filtering is used to blur images and remove noise and detail. Gaussian filter uses the above Gaussian function which is given in equation 1. After applying Gauss transform or Wierstrass transform we can compare original false color composite (FCC) image with transformed false color composite (FCC) image. (a) original false color composite (FCC) image (b) Transformed false color composite (FCC) image Fig. 3. In figure 3 we can see the clear difference between original FCC image with transformed FCC image. After applying transformation we can see that the transformed image is more blur than original image. 291 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 301. Blue Band 0 0.5 1 1.5 2 x 10 4 BlueBand Histogram 0 0.5 1 (a) Green Band 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 x 10 4 Green Band Histogram 0 0.5 1 (b) Red Band 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 x 10 4 Red Band Histogram 0 0.5 1 (c) NIR Band 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 NIR Band Histogram 0 0.5 1 (d) Fig. 4. (a) Histogram of band 1 of transformed image, (b) Histogram of band 2 of transformed image, (c) Histogram of band 3 of transformed image, (d) Histogram of band 4 of transformed image In Figure 4 if we look at the histograms of different bands of transformed false color composite (FCC) image we can examine that the distribution of each band follow Gaussian distribution. We can see now pixel values of image follow standard Gaussian distribution. C. Dimension Reduction Dimension reduction in satellite image processing reducing the number of bands which have same information in nature. Dimension Reduction is used when we have large number of bands. We have Quickbird image which have 4 bands and we apply PCA to check which band we have to ignore. We apply PCA on transformed FCC image. Fig. 4. Principle Components of four bands of transformed image. 292 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 302. In the abovefigure we can see that band 3(Red) and band 4(NIR) have same information so we have to neglect band 3 or band 4 from our analysis because they have same information in the image for this purpose we will use classification on three bands that are Blue, Green and NIR. D. Training Samples Training areas were established by choosing one or more polygons for each class. Pixels fall within the training area were taken to be the training pixels for a particular class. In order to select a good training area for a class, the important properties taken into consideration are its uniformity and how well they represent the same class throughout the whole image [4]. We use Stratified random sampling for selecting the training pixels. A stratified random sample is a population sample that requires the population to be divided into smaller groups, called 'strata'. Random samples can be taken from each stratum, or group. V. RESULTS Maximum Likelihood considered to be an most accurate algorithm when we compare it to classical algorithm such as parallelepiped classification algorithm. But as we know that ML classification algorithm assumes that distribution of each band has to be Gaussian. we had seen that our band's distribution was not Gaussian so we make distribution Gaussian by applying Gauss transform or Weirstrass transform. Now we will check what is the effect on accuracy of classification classes when the distribution is Gaussian and when not Gaussian. A. Classification Before Weirstrass Transform The outcome of ML classification after assigning the classes with suitable colors, is shown in Fig.5 (a): Grass (green), Urban Area (blue), Roads(yellow), Soil (cyan), Trees (red) and Unclassified (Black). The areas in terms of percentage and square meters were also computed; the classes with the largest area is Urban Area. In this case the distribution of our data is not Gaussian and we can clearly see in the Fig.5. (a) that some buildings are wrongly classified as Roads because we are applying ML classification on non normal data. Class Color Area in % Area in Meters²) Color Unclassified: [Black] 6.44% (26,783.6400 Meters²) Trees [Red] 14.20% (59,095.0800 Meters²) Grass [Green] 13.63% (56,719.0800 Meters²) Urban Area [Blue] 31.45% (130,868.2800 Meters²) Roads [Yellow] 30.31% (126,133.9200 Meters²) Soil [Cyan] 3.97% (16,532.6400 Meters²) (a) Class Color Area in % Area in Meters²) Color Unclassified: [Black] 10.30% (42,851.1600 Meters²) Trees [Red] 13.75% (57,232.0800 Meters²) Grass [Green] 10.34% (43,005.2400 Meters²) Urban Area [Blue] 44.79% (186,373.0800 Meters²) Roads [Yellow] 19.39% (80,677.4400 Meters²) Soil [Cyan] 1.44% (5,993.6400 Meters²) (b) Fig. 5. (a) Maximum Likelihood classification an original image, (b) Maximum Likelihood classification on transformed image 293 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 303. TABLE I. CONFUSIONMATRIX OF ORIGINAL CLASSIFIED IMAGE Class Color Trees Grass Urban Area Roads Soil Color Unclassified [Black] 0.4 0.05 7.84 2.33 0 Trees [Red] 94.38 8.06 0.94 5.28 0 Grass [Green] 1.64 88.71 2.66 0.03 3.41 Urban Area [Blue] 0 1.06 70.71 1.19 2 Roads [Yellow 3.58 0 4.61 91.17 0 Soil [Cyan] 0 2.11 13.23 0 94.59 Total 100 100 100 100 100 Ground Truth (Percent) Overall Accuracy = (18318/21394) = 85.62% Kappa Coefficient = 0.8135 TABLE II. CONFUSION MATRIX OF TRANSFORMED CLASSIFIED IMAGE Class Color Trees Grass Urban Area Roads Soil Color Unclassified [Black] 0.85 0.36 4.36 0.52 0 Trees [Red] 98.51 4.72 0.09 0.08 0 Grass [Green] 0.65 94.71 1.92 0 0.3 Urban Area [Blue] 0 0.2 86.5 1.33 0.52 Roads [Yellow] 0 0 3.19 98.07 0 Soil [Cyan] 0 0 3.94 0 99.19 Total 100 100 100 100 100 Ground Truth (Percent) Overall Accuracy = (20156/21394) = 94.2133% Kappa Coefficient = 0.9236 Accuracy assessment of the ML classification was determined by means of a confusion matrix (sometimes called error matrix), which compares, on a class-by class basis, the relationship between reference data (ground truth) and the corresponding results of a classification [1]. The results are given in Table 1. The diagonal elements in Table 1 represent the percentage of correctly assigned and are also known as the producer accuracy. Producer accuracy is a measure of the accuracy of a particular classification scheme and shows the percentage of a particular ground class that is correctly classified. The Kappa coefficient, κ is a second measure of classification accuracy which incorporates the off-diagonal elements as well as the diagonal terms to give a more robust assessment of accuracy than overall accuracy. The ML classification yielded an overall accuracy of 85.62% and kappa coefficient 0.8135. B. Classification After Weirstrass Transform When we apply Gauss transform on data we have already seen that our band's distribution changes from skew symmetric to approximately Gaussian. When we apply ML classification algorithm on transformed image we have seen that the overall accuracy increases from 85.62% to 94.21% because pixel values of image data follows Gaussian distribution and also the value of Kappa coefficient, κ increases from 0.8135 to 0.9236. In Table 2 we can see the results of classification after applying the Weirstrass Transform or Gauss transform. So when our data follows Gaussian distribution we can say that ML classification produce better. VI. CONCLUSIONS In this study, detail analyses of ML classification for tropical land covers in Rome have been carried out, in which lead to a number of conclusions. ML classifies the classes that exist in the study area with a good agreement when distribution follows Gaussian and when distributions is not Gaussian we can see that overall accuracy decreases almost 9%. ML classified the study area into 5 classes, with accuracy 88% (κ= 0.8425). ML classifies pixels based on known properties of each cover type, but the generated classes may not be statistically separable. 294 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 304. ACKNOWLEDGMENT The authors wouldlike to thanks the Dr Waqas A. Qazi for his valuable input and ideas, Muhammad Qasim for helping with the implementation of Weirstrass transform. REFERENCES [1] T.M. Lillesand, R.W. Kiefer and J.W. Chipman, Remote Sensing and Image Interpretation, John Wiley & Sons, Hoboken, NJ, USA, 2004. [2] DigitalGlobe.com. DigitalGlobe. 12 February 2014. Retrieved 19 June 2014. [3] Ahmed I. Zayed, Handbook of Function and Generalized Function Transformations, Chapter 18. p.322-324. [4] J.B. Campbell, Introduction to remote sensing, Taylor & Francis, London, 2002. Muhammad Shoaib is Master student of RS and GIS in the department of Space Science in Institute of Space Technology Islamabad, Pakistan. Zaka Ur Rehman is Master student of RS and GIS in the department of Space Science in Institute of Space Technology Islamabad, Pakistan. Muhammad Qasim received the Masters degree in Mathematics from Comsats Institute of information technology Abbottabad, Pakistan. 295 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 305. Spatio -Temporal Analysisof Shoreline Changes along Makran Coast Using Remote Sensing and Geographical Information System Hira Fatima1 , Mudassar H. Arsalan, Anam Khalid, Kashif Marjan, Mukesh Kumar Geo-informatics and Sustainable Development Research Lab Institute of Space and Planetary Astrophysics University of Karachi Karachi, Pakistan 1 hirafatima1407@gmail.com Abstract—Satellite Images of Makran coast captured by Landsat 5 TM in 1987, 2000 and 2010 were integrated in a GIS (Geographical Information System) to perform spatio-temporal analysis of shoreline changes, coastal accretion and erosion, and coastline length. It was found that Remote sensing data and GIS are time-saving and cost effective for such analyses. Makran shoreline has changed noticeably from 1987 to 2010. During 1987-2010, erosion was dominant along western Makran, it was eroded at the rate of 0.2 sq. km per year. However accretion was dominant along Eastern Makran coast, it was advanced at the rate of 0.5 sq. km per year. During the long time span of 23 years (1987-2010) Gwadar port was eroded 4 sq. km which was quite higher than that of Jiwani i.e. 2.4 sq. km. However Ormara, Pasni and Astola Island were relatively stable places. Index Terms— Shoreline change detection, Erosion - accretion analysis, Image interpretation, Band ratio index, Satellite images, Geospatial technologies, GIS, RS, Geo-informatics. I. INTRODUCTION A shoreline being the interface between land and water is a dynamic feature on the earth's surface. It is always changing due to sediments continually being eroded from one place and deposited to another. Many factors play a part behind these changes, they may be natural like waves, winds, tides and storms as well as human influenced such as dredging, mineral exploration, vegetation removal, infrastructure development etc. These natural processes and human activities can change shoreline on small time scales i.e. days and years or long time scales such as decades or centuries. Monitoring shoreline changes, is very crucial in coastal management and planning. The rate of these changes i.e. erosion and accretion, can be used to forecast future shoreline trends. Decision makers need this information to assess the feasibility of a project and Government organizations for disaster risk management [1-5]. Remote sensing and Geographical Information System are widely used for shoreline change detection analysis. The spectral, spatial and temporal resolutions, repetitive coverage, synoptic view and cost effectiveness of remote sensing data has made it more favorable than data collected by conventional techniques. The shoreline change detection analysis is now widely being done using diversified data such as historical topographical maps, aerial photographs, Digital Elevation Models (DEM), and satellite images. Satellite images from different satellites (such as Landsat 5, Landsat 7, Landsat 8, IKONOS, Quick bird, IRS etc.) have been used for such analysis [5-11]. Further, the data management, visualization and data analysis capabilities of GIS are applied on the data derived from satellite images to get desirable results [12-14]. The initial and crucial step in the shoreline change detection is the selection of appropriate feature that can be treated as shoreline proxy or indicator that can represent the current position and can act as a reference to monitor the change in different periods [2, 15]. These proxy shorelines are typically of two types: a visible and distinguishable line in coastal imagery known as High Water Line (HWL) or intersection of a tidal datum with coastal profile (say MHW, MLW, etc.) [2]. Among all the vertical levels that are used in the shoreline change detection, HWL is very common as it is easily interpretable in photos [2]. After identification, shoreline can be extracted by manual digitization or some automated techniques in GIS [13, 16]. Further, the shoreline movement, erosion and accretion rates can be calculated and its future trends can be predicted [17, 18]. Literature review shows shoreline change detection is now considered as an important part of sustainable coastal resource management and environment conservation, hence it is monitored throughout the World [19]. Table 1 gives a summary of several shore detection studies done throughout the world for different parts of coasts such as beaches, river 296 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 306. mouths, deltas, lakes,estuaries and even islands, and geospatial data and techniques used in these case studies. Over the 100 years, about 70% of world’s sandy coasts (which constitute about 20% of the world’s coastline) have been retreating while 20-30% have been stable and less the 10% have been advancing [20]. Coastal erosion is notable along the sandy beaches of Pakistan [21]. Literature review shows that very few shoreline change detection studies for coastal areas of Pakistan have been done [22, 23] and there is a big research gap in this dimension. This paper aims to bridge this research gap. The Pakistan coast consists of Makran coast (the major portion) and the Sindh coast. Sindh coast is further divided into Indus deltaic coastline (a network of 17 minor creeks) and the Karachi coastline that extends from West of Indus delta to the outfall of the Hub River from where the Makran coast starts and extends up to the Iranian border [21, 24-26]. The coastal length of Pakistan found in literature is ambiguous. It varies from 825 to 1200 km in different sources (see Table 2). Literature review shows that reason behind this ambiguity is coastline paradox. This phenomena was first discovered by Lewis Fry Richardson [27]. Coastline does not have a well-defined length because of its fractal like characteristics. The coastline has features at very different scales, hence the length of coastline increases every time when it is measured at a smaller scale [28]. In short, the length of coastline depends on the method used to measure it, the scale and the resolution of data used [28]. The length of coastline also varies as the shoreline evolves. –with some places say beaches getting sediments to deposit while others getting eroded [29]. However, with the advent of geospatial technologies and high resolution remote sensing data world’s coastlines can be measured more precisely and this paradox can be handled to some extent. Not only the length of coastline, but the area of sea can also be monitored efficiently and cost effectively using remote sensing data and GIS [30]. In this study, the length of Makran coast, the study area, in different years (i.e. 1987, 2000 and 2010) were calculated from Landsat TM satellite images. However, more precise coastline length measurements can be made using high resolution data, such as SPOT [31]. A. Study Area The coastal area of Pakistan selected for this study is Makran coast (see Figure 2 and 6). Most of the Makran coast is undeveloped, with deserted beaches and some fishing villages. Further, it is characterized by unique landforms such as mud flats, rocky cliffs, bays, lagoons, deltas, as well as of the narrow strip of mountains, which is known as Makran ranges. Makran ranges cover area of about 400 km long and 250 km wide. These mountains have elevation up to 1500 m above mean sea level [26, 32]. The Makran coast is blessed with scenic beauty of several beaches: Somiani, Hingol River, Ormara, Pasni, Gwadar etc. The main towns and fishing ports of the study area are Ormara, Pasni and Jiwani. Gwadar deep sea port is the economic region of Balochistan coast. It is a major destination in the Pakistan-China economic corridor [32, 33]. Makran coast differs Sindh coast in several ways. Indus is the only major river of Pakistan and its delta lie at the Sindh coast, while at Makran coast only small rivers i.e. Hingol, Hab, Basul and Dasht do exist. The depths along the Sindh coast change moderately, while the Makran coast is steep. The average rainfall on the Makran coast is about 10 mm while it is twice at the Sindh coast [34]. Makran coast is less vulnerable to sea level rise than Sindh coastal area as the Makran coast is uplifting about 1-2 mm/year due to the subduction of Indian oceanic plate [22]. Coastal erosion is one of the typical effect of sea level rise [35]. A very threatening subduction zone is located about 100 km away from Makran coast that can cause tsunami [36, 37]. In 1945 tsunami, Las Bela State of Balochistan (which is now Western Pakistan) was affected badly, particularly the Pasni and Ormara towns [38]. Pakistan has mixed semi-diurnal tides, however the tidal range varies throughout the coastline [22]. The tidal amplitudes are greater in the Indus deltaic region as the sea water flows into creeks with high velocity during flood and ebb tides, however the tidal amplitudes gradually decreases towards west (Makran coast) [22, 39]. During the Southwest monsoon season the direction of the prevailing ocean current is clockwise and counter clockwise during the Northeast monsoon season in the Arabian Sea [34]. Generally, the salinity value is approximately 36 ppt [36]. In this study, we examined different bands of Landsat 5 TM, their combination and ratios for shoreline delineation. We calculated shoreline length for different years and studied coastline paradox phenomenon. We analyzed the long and short term shoreline changes and associated erosion and accretion along the Makran coast using Landsat TM data acquired in 1987, 2000 and 2010. II. MATERIALS AND METHODS In this study, we used Landsat 5 TM data and two commercial software (i.e. ArcGIS 10.3 and ERDAS Imagine 9.2) for data management, analysis and visualization. The objectives of this study were achieved by simple steps shown in the figure 3.Fig. 1. Coastline Paradox Illustration 297 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 307. Table 1 ShorelineChange Detection Studies throughout the World Serial # Author Year Study Area Technique used Data Used 1 Shetty [40] 2015 Mangalore Coast, Netravathi-Gurupur and Mulky-pavanje Spits Visual interpretation and on screen digitization Landsat 5, IRS and Landsat 8 2 Natesan [41] 2015 Tamil Nadu, India Layer Stacking, digitization Landsat MSS, Landsat TM, Landsat ETM+ and Landsat OLI 3 Hegde [42] 2015 Karnataka Coast, India Histogram slicing of infrared band, unsupervised classification and digitization Landsat TM, Landsat ETM+, Landsat OLI-TIRS 4 Aedla [16] 2015 Netravati-Gurpur River mouth Modified Self–Adaptive Plateau Histogram Equalization with Mean Threshold IRS data 5 Mahapat [10] 2013 South Gujarat Coast False color composite (bands 2, 3, 4), visual interpretation and digitization Landsat MSS, Landsat TM, Landsat ETM+, IRS 6 Choudharey [43] 2013 from Karwar to Gokarna - South West coast of India Geo referencing, digitization Toposheet and IRS data 7 Murray [8] 2013 Wotje Atoll, Marshall Islands Mosaicking, Digitization Aerial photographs, IKONOS, QuickBird, GeoEye-1 8 Rio [4] 2013 SW Spain Georeferencing, polynomial correction and digitization Aerial Photographs 9 Das [44] 2013 Shankarpur - Mandarmoni Coast line, West Bengal Band ratio mode between mid infrared and green, digitization Topomap, Landsat TM, Landsat ETM+ and Google Earth imagery 10 Azaz [45] 2012 Wedam_Alsahel area, Batinah, Oman Geometric correction, Image degradation, digitization IRS, IKONOS 11 Faiboon [46] 2012 Chalatat Beach in Songkhla Province, Thailand Geometric correction, resampling, Overlay Landsat ETM+, Landsat TM, 12 Eludoyin [1] 2012 Bonny Island, Nigeria Layer stacking, visual interpretation and digitization Landsat TM, Nigersat data 13 Kumaravel [47] 2012 Cuddalore district, East coast of Tamil Nadu, India Georeferencing, layer stacking, digitization Toposheets, IRS data 14 Kuleli [6] 2011 Coastal Ramsar Wetlands of Turkey Top of atmosphere segmentation algorithm, NDWI, Automatic thresholding and shoreline extraction Landsat MSS, Landsat TM, Landsat ETM+ 15 Adegoke [3] 2010 Niger Delta, Nigeria Radiometric and geometric correction, georeferencing, mosaicing, Histogram equalization, Band combination 6,4,2 and digitization Landsat TM, Landsat ETM+ data 16 Li [48] 2010 Pearl River Estuary, China Georeferencing, Mosaicking, Layer stacking, slicing, classification and digitization Topographic map, nautical map, Landsat TM, Landsat MSS, Landsat TM+, Spot XS 17 Maiti [5] 2009 Bay of Bengal, Eastern India Georeferencing, gray level thresholding and segmentation by the edge enhancement technique of NIR bands, Digitization Toposheets, Landsat MSS, Landsat TM, Landsat ETM+ and ASTER 18 Kumar [49] 2009 Mangalore Coast, India Georeferencing, layer stacking, digitization Toposheets, IRS data 19 Rajamanickkam 2006 Kallar and Vaipar Coast, Tamilnadu, India, Edge enhancement, level slicing, NDVI, digitization Topo sheets and IRS data 20 Klien [50] 2006 Position adjacent to Hadera power station, Israel Rectification and georeferencing, Digitization Aerial Photographs 21 Makota [13] 2004 Kunduchi Area, Tanzania Visual interpretation and on screen digitization Aerial Photographs 22 Frazier [51] 2000 Channel and flood plain of Murrumbidgee River Single band density slicing, multi- spectral maximum likelihood algorithm, digitization Landsat 5 TM data, Aerial Photographs 298 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 308. Table 2 Lengthof Pakistan Coastline from different sources Serial # Author/ Organization Year Length of Pakistan Coastline km Length of Makran Coastline km Length of Sindh Coastline km 1 UNEP [34] 1986 825 ___ ___ 2 Quraishee [39] Tariq [52] Rashid [23] 1986 2000 2012 990 ___ ___ 3 Pakistan Tourism Development Corporation [53] 2006 1046 ___ ___ 4 Jalal [54] 2008 1050 ___ ___ 5 Wildlife of Pakistan [32] 2006 1050 800 250 6 Pike [36] Hafeez [55] 2014 2007 1050 700 350 7 Einfopedia [56] 2010 1100 771 329 8 Encyclopedia Britannica [25] 2001 ___ ___ 350 9 ESCAP [21] 1996 ___ ___ 370 10 WWF- Pakistan [24] 2008 ___ 1000 ___ 11 Himalayan Holidays [26] 2002 ___ 754 ___ 12 Jalal [33] 2008 1200 ___ ___ A. Data Collection For this study, we utilized Landsat 5 TM images of the study area (Path/Row: 156-42, 156, 43, 155-42, 155-43, 154- 42, 154-43, 153- 42 and 153- 43) acquired in 1987, 2000 and 2010. Landsat 5 was a joint effort of NASA and USGS, launched on 1st March 1984. Landsat 5 was put in a circular, sun synchronous, near polar orbit at the altitude of 705.3 km. Landsat 5 carried TM (Thematic Mapper) and MSS (Multi Spectral Scanner System) sensors. It took approximately 16 days to scan the whole earth's surface. It used to cross equator at 9:45 a.m. (+/- 15 min). [57] Landsat 5 was decommissioned on 5th June 2013, due to several mechanical failures. Landsat 5, in its 29 year, 3 months, 4 day journey orbited 150,000 times, transmitted over 2.5 million images. Several significant events were captured by Landsat 5, such as 1986’s Chernobyl incident, 2004’s devastating tsunami in Southeast Asia, etc. [58] Landsat 5 set a Guinness world record for “Longest operating Earth observation satellite”. Landsat 5 Thematic mapper sensor data is available on USGS websites for free. [59] To assess short term (in a span of 10 and 13 years) and long term (in a span of 23 years) shoreline change detection, cloud free and low tide data of Landsat 5 TM were downloaded from the USGS EarthExplorer website. Further characteristics of satellite data (Landsat 5 TM) used in this study is summarized in table 3. Boundaries of water bodies can be delineated more accurately using fine resolution spatial data for example IKONOS and SPOT, with spatial resolution of less than 1 m. Although IKONOS and SPOT are commercial satellites and their data is not available free of cost. However, all the Landsat data in USGS archive is available online for free. The summary of collected data is given in table 4. B. Data Processing After data collection, data was processed to make it ready for our analysis. As the study area was very large (i.e. about 990 km long) the study area was divided into two parts i.e. Western Makran Coast (Path/Row: 155-42, 155- 43, 156-42 and 156-43) and Eastern Makran Coast (Path/Row: 154-42, 154-43, 153-42 and 153-43) (See figure 6). All the images of Western and Eastern study area were geometrically corrected, stacked and mosaicked using ERDAS Imagine 9.2. All the collected and processed data, and result files of shoreline change detection and erosion - accretion analysis were stored in ArcGIS geodatabase. All images and shape files used in this study were projected to Universal Transverse Mercator (UTM) projection system, zone 41 and the spheroid and datum were referenced to WGS1984. C. Shoreline Identification Water behaves like a semi-transparent medium for the electromagnetic radiation, hence when EM radiation strikes water surface, it gets reflected, transmitted or absorbed in water. The spectral responses depend on the wavelength of the radiation as well as the physical and chemical properties of the water. When water is in the liquid form, it shows greater reflectance in the visible region between 0.4μm and 0.6μm however wavelengths beyond 0.7μm are completely absorbed. Water absorbs most of the energy in NIR and MIR wavelengths and appears dark in these bands, in contrast to this land and vegetation show higher reflectance in these Fig. 2 Some Glimpses of Study Area- Makran Coast, Pakistan 299 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 309. wavelengths and appearbright in these bands. This property of Water is used to differentiate it from land and vegetation in multi-spectral imagery [7, 9, 62-64]. Figure 5 shows the spectral profiles of land, water and vegetation drawn in ERDAS Imagine 9.2 from a Landsat 5 TM image. Figure 7 shows images of Miani Hor -a part of Makran Coast- in different bands of the Landsat 5 TM. In first three bands the contrast between water and other surface features are not very remarkable. On the contrary, the IR bands (bands 4 and 5) show a sharp contrast between them as water reflects very less in the IR region of the EMR spectrum. Among all the bands, band 5 gives the most outstanding contrast between water and other surface features. However, different band combinations have their own tendency to differentiate land and water. Figure 7 shows that Mid IR and Green band ratio is very suitable for shoreline delineation. Band ratio is a digital image processing technique in which DN value of one band is divided by that of any other band. It is used to enhance the contrast between different features. It also eliminates the shadowing effects [63]. Band ratio has its application in several fields such as minerals, soil, vegetation, water bodies etc. [64]. Three different band ratio images can be combined to make color composites. This will highlight certain features in distinctive colors [65]. The initial and crucial step in the shoreline change detection is the selection of shoreline proxy and then its identification. Water boundary delineation is one of the applications of Remote Sensing. In this study, we considered High Water Line (WHL) as a shoreline proxy. The WHL is defined as the intersection of land with water surface at an elevation of high water [66]. High water line has been used in several studies for shoreline delineation [4, 10, 13, 41, 43, 44]. To identify the high water line (WHL), the mid infrared (band 5) and green (band 2) band ratio index methodology was adopted. D. Shoreline Extraction Band ratio was calculated for every image, then the high water line was extracted by manual digitization on 1:50,000 scales in ArcGIS 10.3. This scale was kept consistent to in all digitization work throughout the study to maintain the consistency. The temporal movement (1987-2010) of shoreline is shown in figure 8 and 9. E. Shoreline Length Measurement The length of shorelines (extracted from band ratio images of 1987, 2000 and 2010) were measured using calculate geometry function in ArcGIS 10.3. F. Shoreline Change/Movement Shoreline change was analyzed by overlay technique. Extracted shorelines of 1987, 2000 and 2010 of eastern and western Makran coast were overlaid on each other to visualize the shoreline movement. G. Erosion and Accretion Measurement Erosion is the transportation process of soil or decomposed rock material by natural forces (such as storms, flooding, waves and tides) while accretion is a slight and unnoticeable accumulation of soil or decomposed rock by natural phenomenon. Coastal accretion occurs when the alluvion deposit upon the shore. [68] Fig. 4. Data Processing Fig. 3 Materials and Methods Fig. 5. Spectral Profiles of Land, Water and Vegetation drawn in Erdas Imagine 9.2 from Landsat 5 TM image 300 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 310. Erosion and accretionanalysis for short time spans (i.e. 1987-2000 and 2000-2010) and long time span (i.e. 1987- 2010) for Western and Eastern Makran Coast were done using ArcGIS 10.3. The Features to polygon tool was applied to 1987 and 2000 shorelines to identify the stable and unstable areas. This tool gave unstable areas as polygons. From these unstable areas (polygons) the advanced (accretion) and retreat (erosion) areas were identified and their areas were calculated using ArcGIS 10.3. The same steps were followed to do the accretion and erosion measurements for another short time span (i.e. 2000-2010) and a 23 year long time span (i.e. 1987-2010). To better understand the results were normalized (See table 6-8). Figures 10-17 show the developed accretion and erosion maps. H. Visualization GIS has provided us a new way of looking and communicating our results. Visualizing results for better understanding has become very common these days. All the results of this study were visualized using ArcGIS 10.3. The division of study area in eastern and western part, movement of shoreline and accretion - erosion in different time spans are shown in figures 6, 8 and 9, and 10-17respectivelyy. III. RESULTS AND DISCUSSIONS Coastline paradox is a globally accepted phenomenon. The length of coastline changes when measured from remote sensing data of different resolution or when measured on different scales. It also changes when measured on the same scale from remote sensing data of the same resolution at different time interval. The reason behind this temporal variability is the change in shape of coastline associated with erosion and accretion. Table 5 shows the length of Makran coastline measured in different years from Landsat TM images. However the length of coastline was measured on the same scale from remote sensing data of same spatial resolution images, the length of coastline was found different in different years due to change in coastline shape with associated erosion and accretion. Spatio-temporal shoreline movement and accretion - erosion are shown in figures 8-17. A. Accretion - erosion along Western Makran Coast The western Makran coast accretion - erosion analysis results are presented in table 6. The accretion - erosion analysis result is geographically presented in figure 10. 1) Accretion and Erosion between 1987-2000 The accretion - erosion analysis shows during 1987 to 2000 (i.e. 13 year time span) about 9.4 sq. Km land was eroded while 12.6 sq. Km land area was advanced. Hence the total change was 3.2 sq. Km land advancement. During 1987-2000 the accretion rate per year was about 1 sq. Km, while the erosion rate was 0.7 sq. Km which was less than accretion rate. Hence, in this short time span (1987-2000) accretion was more dominant than erosion along the western coast and the land advanced at a rate of 0.3 sq. Km per year. 2) Accretion and Erosion between 2000-2010 The accretion - erosion analysis shows during the short time span of 10 years (i.e. 2000- 2010) about 11 sq. Km land was eroded while land area was advanced by 4.8 sq. km along western Makran coast. Hence the total change was 6.2 sq. Km erosion. During 2000-2010 the accretion rate per year was about 0.4 sq. Km, while the erosion rate was 1.1 sq. Km, which was quite higher than accretion rate. Hence the erosion was far dominant than accretion in this time span and the Western Makran coast receded at the rate of 0.7 sq. km per year. 3) Accretion and Erosion between 1987-2010 The accretion - erosion analysis shows during the long time span of 23 years (i.e. 1987- 2010) about 13.6 sq. Km land was eroded while land was advanced by 9.2 sq. km. Hence the total change was 4.4 sq. Km erosion. During 1987-2010 the accretion rate per year was about 0.4 sq. Km, while the erosion rate was 0.6 sq. Km which was greater than accretion rate. Hence the erosion was more dominant than accretion in this 23 year long time span (1987-2010) and the coastal area of Western Makran was eroded at the rate of 0.2 sq. Km per year. Fig. 6. Study Area- Makran Coast, Pakistan 301 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 311. 4) Accretion -erosion Analysis Results on Local Scale a) Gwadar Gwadar (25°07′35″N 62°19′21″E) is a city located on the southwestern coastline of Pakistan. Gwadar was nominated as the winter capital of Balochistan in 2011. Gwadar port located at the mouth of the Persian Gulf. It is a third deep sea port of Pakistan, inaugurated to increase the port capacity of Pakistan [69]. The erosion - accretion analysis shows that Gwadar coast has faced a noticeable erosion problem. During 1987-2000 about 4.5 sq. Km land was eroded at Gwadar port and no noticeable accretion took place while 1.5 sq. Km land was eroded and 0.5 sq. Km land was advanced at Western Gwadar bay. During 2000-2010 about 1.6 sq. Km land was advanced at Gwadar port while no noticeable erosion took place meanwhile, no noticeable erosion and accretion took place at Eastern or Western Gwadar bay. In long time span of 23 years (1987-2010) about 4 sq. Km land was eroded at Gwadar port while 1.5 sq. Km land was advanced meanwhile 1.2 sq. Km land at western and 0.4 sq. Km land at eastern Gwadar bay was eroded while no noticeable accretion took place (See figure 11). b) Pasni Pasni (25°16′N 63°28′E) is a moderate size town and fishing port located at the Makran coast [70]. During the time span of 13 years (1987-2000) about 0.2 sq. Km land was eroded while 4.5 sq. Km land was advanced at Pasni and its vicinity. During the short time span of 10 years (2000-2010) about 0.3 sq. Km land was advanced while 5.8 sq. Km land was eroded at Pasni and its vicinity which is greater than the accretion took place during 1987-2000. During the long time Fig. 7. Landsat 5 TM Images of Miani Hor-a part of Makran Coast- in different Spectral bands, band combinations and band ratio using ERDAS Imagine 9.2 302 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 312. Table 3 Landsat5 TM and MSS Sensor Characteristics Table 4 Summary of Collected Data Satellite (Sensor) Path/Row Year Tide Phase Landsat 5 (TM Sensor) 153-42, 153-43 154-42, 154-43 155-42, 155-43 156-42, 156-43 1987 Low2000 2010 Table 5 Makran Coastline Length measured from Landsat 5 TM images Year Scale Spatial Resolu- tion m Western Makran coast length Km Eastern Makran coast length Km Makran coast total length km Shoreline proxy 2010 1:50,000 30 335 634 969 WHL at low tide 2000 337 661 998 1987 344 667 1011 span of 23 years (1987-2010) about 0.5 sq. Km land was eroded as well as 0.5 sq. Km land was advanced which showed a balance in erosion and accretion but at a long time span (See figure 12). a) Jiwani Jiwani (25°03′N 61°44′E) is a town and commercial port. It is located near the Pak-Iranian border [71]. The accretion - erosion analysis shows that the coastal area of Jiwani and its vicinity is under influence of natural forces as well human induced factors which have been causing coastal erosion. During 1987-2000 about 5.2 sq. Km land was advanced while no considerable erosion took place. During 2000-2010 about 4.8 sq. Km land was eroded (near about accretion took place during 1987-2000) while only 0.8 land was advanced at Jiwani and vicinity. In 23 year long time span (1987-2010) about 5.2 sq. Km land was advanced while 2.4 sq. Km land was eroded (See figure 13). B. Accretion - erosion along Eastern Makran Coast The western Makran coast accretion - erosion analysis results are given in table 7. The accretion - erosion analysis visualization is shown in figure 14. 1) Accretion and Erosion between 1987-2000 The accretion - erosion analysis shows during 1987 to 2000 (i.e. 13 year time span) about 8.5 sq. Km land was eroded while 12 sq. Km land area was advanced at the east Makran coast. Hence the total change was 3.5 sq. Km accretion. During 1987-2000 the accretion rate per year was about 0.9 sq. Km, while the erosion rate was 0.7 sq. Km which was not exactly same, but close to the accretion rate. Hence, in this short time span there was a balance in the accretion and erosion to some extent. In this (1987-2000) thirteen year period the eastern Makran coast was advanced at a rate of 0.3 sq. Km per year. 2) Accretion and Erosion between 2000-2010 The accretion - erosion analysis shows during the short time span of 10 years (i.e. 2000- 2010) about 5.5 sq. Km land was eroded while land was advanced by 16 sq. km at eastern Makran coast. Hence the total change was 10.5 sq. Km accretion. During 2000-2010 the erosion rate per year was about 0.5 sq. Km, while the accretion rate was 1.6 sq. Km, which was quite higher than erosion rate. Hence the accretion was far dominant than erosion in this time span and land advanced at the rate of 1.1 sq. km per year. 3) Accretion and Erosion between 1987-2010 The accretion - erosion analysis shows during the long time span of 23 years (i.e. 1987- 2010) about 6.8 sq. Km land was eroded while land was advanced by 17.5 sq. km at eastern Makran coast. Hence the total change was 10.7 sq. km erosion. During 1987-2010 the accretion rate per year was about 0.8 sq. Km, while the erosion rate was 0.3 sq. Km which was lower than accretion rate. Hence the accretion was quite dominant than erosion in this 23 year long time span (1987- 2010) and the coastal area of Eastern Makran was advanced at the rate of 0.5 sq. Km per year. Satellite Sensors Swath (Km) Scene size (Km) Resolution Band Spectral Coverage (μm) Temporal Spectral Spatial (Meters) Landsat 5 MSS 180 180 x 170 16 days 4 82 Band 4 – Green 0.5 – 0.6 82 Band 5 – Red 0.6 – 0.7 82 Band 6 – Near IR 0.7 – 0.8 82 Band 7 – Near IR 0.8 – 1.1 TM 185 170 x 183 16 days 7 30 Band 1 – Blue 0.45 – 0.52 30 Band 2 – Green 0.52 – 0.6 30 Band 3 – Red 0.63 – 0.69 30 Band 4 – Near IR 0.76 – 0.90 30 Band 5 – Mid IR 1.55 – 1.75 120 Band 6 –Thermal 10.40 – 12.50 30 Band 7 – Mid IR 2.08 – 2.35 303 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 313. 4) Accretion -erosion Analysis Result on Local Scale a) Hingol River Hingol River (25° 23′N 65° 28′E) is located in the Gwadar district of Balochistan. It is the longest river in Balochistan (about 560 km long). Unlike the majority of streams in Balochistan, which flow only in rare rain, Hingol River flows all year long [72]. The erosion - accretion analysis of mouth of the Hingol River shows that during 1987-2000 about 1.3 sq. Km land was eroded there while 1.2 Sq. km land was advanced in its vicinity. During 2000-2010 about 0.4 sq. Km land was advanced while only 0.2 Sq. km land was eroded at the mouth of Hingol River and about 0.6 Sq. km in the vicinity. In long time span of 23 years (1987-2010) about 0.2 sq. Km land was eroded while 0.3 sq. Km land was advanced at the mouth of the Hingol River while in its vicinity about 0.6 Sq. km land advancement took place and 1.2 sq. km land was eroded. (See figure 15). b) Ormara Ormara (25°12′53″N 64°38′2″E) is an old coastal town located in the Gwadar district of Balochistan. Ormara has a fish harbor and a port [73]. The erosion - accretion analysis of Ormara shows that during 1987-2000 about 0.6 sq. Km land Fig. 8. Western Makran Coast Temporal Shoreline Movement 1987-2010 Fig. 9. Eastern Makran Coast Temporal Shoreine Movement 1987-2010 304 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 314. was eroded while4 Sq. km land was advanced at Ormara. During 2000-2010 about 1 sq. Km land was advanced at Ormara port while only 1.5 Sq. km land was eroded. In long time span of 23 years (1987-2010) about 0.5 sq. Km land was eroded at Ormara port while 3.7 sq. Km land was advanced. (See figure 16). c) Islands in Study Area A piece of sub-continental land that is surrounded by water is known as the Island. Islands are classified into two main types i.e. continental islands and oceanic islands. They may be artificial as well [74]. Pakistan has several natural islands off the Sindh and Balochistan (Makran) coasts. The islands off the Makran coast are Malan island (a volcanic mud island that Fig. 10. Western Makran Coast Accretion Erosion 1987-2010 305 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 315. appears and disappears),Zalzala island (appeared offshore Gwadar followed by an earthquake of 7.7 magnitude in September 2013) and the Astola island [75]. ASTOLA ISLAND Astola Island (25°7′21″N 63°50′51″E) is Pakistan’s largest offshore island. It is also known as Haft Talar Island (Island of seven Hills) with the highest point 246 feet above the sea level [76]. It is about 6.7 km long and with a maximum width of 2.3 km and 6.7 Sq. km approximated area [76]. However the area of Astola Island estimated in this study was 2.2 sq. km with shoreline length 10.2 km. This study reveals that during short (i.e. 1987-2000 and 2000- 2010) and long time (1987-2010) spans, very nominal accretion and erosion took place along the shoreline of Astola Island. During the short time span of 13 years (i.e. 1987-2000) only 0.13 sq. km land was advanced while 0.02 sq. km land was eroded. During ten year time span (2000-2010) only 0.04 sq. km land was advanced while 0.01 sq. km land was eroded. During the long time span of 23 years (i.e. 1987-2010) 0.1 sq. km land advanced while only 0.04 sq. km land was eroded. d) Lagoons in Study Area The lagoon is a shallow water body separated from the larger water body by reefs and small islands [77]. Lagoons are classified into coastal and oceanic lagoons. Coastal lagoons are further classified into three main types i.e. choked, leaky and restricted [78]. Lagoons are found in a great range of sizes, the size of a smallest lagoon is about a hectare while the largest (i.e. Lagoa Patos in Brazil) is about 10,000 km2 . Usually the length of the lagoon is greater than its width [79]. Lagoons are typical coastal features found parallel to the coastlines all around the world say Glenrock lagoon in Australia, Lagoa dos Patos lagoon in Brazil, Apoyo lagoon natural reserve in Nicaragua, Blue lagoon in Turkey, Venetian lagoon in Veneto, Miani Hor and Khor Kalmat along Makran coast, Pakistan [77, 80, 81]. Fig. 11. Gwadar Accretion Erosion 1987-2010 306 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 316. KHOR KALMAT Khor kalmatlagoon (25° 24′N 64° 06′E) is located along the eastern Makran coast. Basol River flows southward in the Gwadar District. It drains a deserted Makran region, with its river mouth at Khor Kalmat lagoon [81]. MIANI HOR Miani Hor (25° 32′N 66° 12′E) is a swampy lagoon and a Ramsar wetland, located in Lasbela district [80]. Both Miani Hor and Khor Kalmat were not included in the accretion - erosion analysis of this study as lagoons are extremely dynamic ecosystem [78]. Further, they have their own dynamics different from that of other coastal features [79]. A study in which 7 year beach profile data was analyzed, showed that the lagoon shorelines were far dynamic than the ocean shorelines [82]. The flow of water in lagoons is slow and sluggish and the flushing time depends on the type of lagoon [78]. C. Accretion - erosion along Makran Coast The accretion - erosion analysis results of western and eastern Makran coast were combined to understand the accretion - erosion scenario of entire Makran coast. D. Accretion and Erosion between 1987-2000 The accretion - erosion analysis shows during 1987 to 2000 (i.e. 13 year time span) about 17.9 sq. Km land was eroded while land was advanced by 24.6 sq. Km at Makran coast. Hence the total change was 6. 7 km accretion. During 1987- 2000 the accretion rate per year was about 1.9 sq. Km, while the erosion rate was 1.6 sq. Km which was not exactly same, but close to the accretion rate. Hence, in this short time span there were balance in the accretion and erosion to some extent. However, in this (1987-2000) thirteen year period the Makran coastal area advanced at a rate of 0.5 sq. Km per year. Fig. 12. Pasni and Vicinity Accretion Erosion 1987-2010 307 Fourth International Conference on Aerospace Science & Engineering (ICASE 2015) Proceedings
- 317. 1) Accretion andErosion between 2000-2010 The accretion - erosion analysis shows during the short time span of 10 years (i.e. 2000- 2010) about 16.5 sq. Km land was eroded while land was advanced by 20.8 sq. km at Makran coast. Hence the total change was 4.3 sq. Km accretion. During 2000-2010 the erosion rate per year was about 1.7 sq. Km, while the accretion r