1. OPTIMIZATION OF MACHINING PROCESS
USING ANN MODEL
Abhishek Srivastava(141703)
Adarsh Sharma (141704)
Aditya Singh Gaur (141708)
Rahul kumar (141803)
Project guide-
Dr. Y.K. Modi
2. CONTENT
Introduction to Design of Experiment (DOE)
Introduction to Artificial neural network (ANN)
Literature survey
References
4. 14-1: Introduction
4
• An experiment is a test or series of tests.
• The design of an experiment plays a major role in
the eventual solution of the problem.
• In a factorial experimental design, experimental
trials (or runs) are performed at all combinations of
the factor levels.
• The analysis of variance (ANOVA) will be used as
one of the primary tools for statistical data analysis.
6. Role of DOE in Process Improvement
DOE is a formal mathematical method for systematically
planning and conducting scientific studies that change
experimental variables together in order to determine their effect
of a given response.
DOE makes controlled changes to input variables in order to
gain maximum amounts of information on cause and effect
relationships with a minimum sample size.
DOE is readily supported by numerous statistical software
packages available on the market.
6
7. DOE: Objectives
Determine influential variables (factors)
Determine where to set influential factors to optimize response
Determine where to set influential factors to minimize response
variability
Determine where to set influential factors to minimize the effect of
the uncontrollable factors
O
E
7
8. DOE: Applications in Process
Development
Improve process yield
Reduce variability
Reduce development time
Reduce overall costs
O
E
8
9. DOE: Applications in Design
Evaluate and compare alternatives
Evaluate material alternatives
Product robustness
Determine key design parameters
O
E
9
10. Four Basic Steps of DOE
10
Cycle of Experimentation
COLLECT
Observe
Code
PLAN
Questions
Design
Scope
PRESENT
Answer Questions
Graphically
Mathematically
State Uncertainty
Recommend
ANALYZE
Plot
Plot
Plot
Model
Conclude
11. 3-PRINCIPLES OF DOE
Replication – repetition of a basic experiment without changing
any factor settings, allows the experimenter to estimate the
experimental error (noise) in the system used to determine
whether observed differences in the data are “real” or “just
noise”, allows the experimenter to obtain more statistical power
(ability to identify small effects)
Randomization – a statistical tool used to minimize potential
uncontrollable biases in the experiment by randomly assigning
material, people, order that experimental trials are conducted, or
any other factor not under the control of the experimenter.
Results in “averaging out” the effects of the extraneous factors
that may be present in order to minimize the risk of these factors
affecting the experimental results.
11
12. Blocking – technique used to increase the precision of an experiment
by breaking the experiment into homogeneous segments (blocks) in
order to control any potential block to block variability (multiple lots
of raw material, several shifts, several machines, several inspectors).
Any effects on the experimental results as a result of the blocking
factor will be identified and minimized.
12
13. Metrology considerations for industrial
designed experiments
Accuracy:- It refers to the degree of closeness between the
measured value and the true value or reference value.
Precision:- It is a measure of the scatter of results of several
observations and is not related to the true value
Stability:- A measurement system is said to be stable if the
measurements do not change over time.
Capability:- A measurement system is capable if the measurements
are free from bias (accurate) and sensitive.
14. A practical methodology for DOE
The methodology of DOE is fundamentally divided into
four phases. These are:
1. planning phase
2. designing phase
3. conducting phase and
4. analyzing phase.
15. Planning phase
(a) Problem recognition and formulation.
(b) Selection of response or quality characteristic.
(c) Selection of process variables or design parameters.
(d) Classification of process variables
(e) Determining the levels of process variables
(f) List all the interactions of interest
16. Designing phase
In this phase, one may select the most appropriate
design for the experiment. Experiments can be
statistically designed using classical approach
advocated by Sir Ronald Fisher, orthogonal array
approach advocated by Dr Genichi Taguchi or
variables search approach promoted by Dr Dorian
Shainin
17. Conducting phase
Selection of suitable location for carrying out the
experiment.
Availability of materials/parts, operators, machines,
etc. required for carrying out the experiment.
Assessment of the viability of an action in monetary
terms by utilising cost benefit analysis
18. Analyzing phase
Determine the design parameters or process variables
that affect the mean process performance.
Determine the design parameters or process variables
that influence performance variability.
Determine the design parameter levels that yield the
optimum performance.
Determine whether further improvement is possible.
20. ANN
ANN possesses a large number of processing elements
called nodes/neurons which operate in parallel.
Neurons are connected with others by connection link.
Each link is associated with weights which contain
information about the input signal.
Each neuron has an internal state of its own which is a
function of the inputs that neuron receives- Activation
level
21. Fundamental concept
NN are constructed and implemented to model the
human brain.
Performs various tasks such as pattern-matching,
classification, optimization function, approximation,
vector quantization and data clustering.
These tasks are difficult for traditional computers
22. Brain ANN
Speed Few ms. Few Nano sec. massive ||el
processing
Size and complexity 1011 neurons & 1015
interconnections
Depends on designer
Storage capacity Stores information in its
interconnection or in
synapse.
No Loss of memory
Contiguous memory
locations
loss of memory may
happen sometimes.
Tolerance Has fault tolerance less fault tolerance Info.
gets disrupted when
interconnections are
disconnected
Control mechanism Complicated involves
chemicals in biological
neuron
Simpler in ANN
Comparison between brain verses computer
26. APPLICATION
Functional approximation, including time series prediction and
modelling.
Call control- answer an incoming call (speaker-ON) with a
swipe of the hand while driving.
pattern and sequence recognition, pattern detection and
sequential decision making.
Skip tracks or control volume on your media player using
simple hand motions.
Data processing, including filtering, clustering, blind signal
separation and compression.
27. ADVANTAGES OF
NEURAL NETWORKS
Adaptive learning: A neural networks have the ability to learn how to
do things.
Self-Organisation: A neural network or ANN can create its own
representation of the information it receives during learning.
Real Time Operation: In neural network or ANN computations can be
carried out in parallel.
Pattern recognition is a powerful technique for the data security. Neural
networks learn to recognize the patterns which exist in the data set.
The system is developed by learning rather than programming. Neural
networks teach themselves the patterns in the data freeing the analyst for
more interesting work.
28. LIMITATIONS OF NEURAL
NETWORK
ANN or Neural Networks is not a daily life problem
solver.
There is no structured methodology available.
There is no single standardized paradigm for Neural
Networks development.
The Output Quality of an ANN can be unpredictable.
Many ANN Systems does not describe how they solve
the problems.
Nature of ANN is like a Black box.
30. RESEARCH PAPERS
TOPIC AUTHOR
USE OF ANN AND RSM
TO MODEL, PREDICT
AND
OPTIMIZE THE
PERFORMANCE
PARAMETERS FOR
TURNING WASPALOY
Kishore Jawale,
Dr. P Subhash Chandra
Bose, Prof. C S P Rao
International Journal of
Advanced Engineering
Research and Studies E-
ISSN2249–8974
Prediction and control of
surface roughness in CNC
lathe using artificial neural
network
Durmus Karayel journal of materials
processing technology 209
( 2009) 3125–3137
Investigation of cutting
parameters of surface
roughness for a non-ferrous
material using artificial
neural network in CNC
turning
C. Natarajan, S. Muthu
and P. Karuppuswamy
Journal of Mechanical
Engineering Research Vol.
3(1), pp. 1-14, January 2011
31. USE OF ANN AND RSM TO MODEL, PREDICT AND
OPTIMIZE THE PERFORMANCE PARAMETERS FOR
TURNING WASPALOY
Artificial neural networks (ANNs) and Response Surface
Methodology (RSM) are often used in manufacturing field for
modelling complex relationships which are difficult to describe with
physical models.
This paper aims to apply Taguchi method for the optimization of
ANN model. A case study of modelling resultant Cutting Force,
Surface Finish and Cutting Temperature in turning process is used to
demonstrate implementation of this approach.
Genetic Algorithms are also used to optimize the performance
parameter obtained from ANN model.
32. USE OF ANN AND RSM TO MODEL, PREDICT AND
OPTIMIZE THE PERFORMANCE PARAMETERS FOR
TURNING WASPALOY
MATERIAL &MACHINE TOOLS-
Waspaloy (Nickel 58%, chromium 19%, cobalt 13%, molybdenum
4%, titanium 3%, aluminium 1.4%) is hard turned on retrofitted
VDF CNC lathe with Multi-layered PVD coated cemented tungsten
carbide Inserts.
A double clamp-type tool holder is used to provide rigidity.
The measurements of average surface roughness (Ra) are made with
HANDYSURF E 35 B instrument.
Piezo-electric dynamometer is used to measure the cutting forces.
. Infrared thermometer KIRAY 300 is used for temperature measure
at the cutting zone during machining.
33. USE OF ANN AND RSM TO MODEL, PREDICT AND
OPTIMIZE THE PERFORMANCE PARAMETERS FOR
TURNING WASPALOY
Parameters used in this paper are:
Speed
Feed
Depth of cut
34. USE OF ANN AND RSM TO MODEL, PREDICT AND
OPTIMIZE THE PERFORMANCE PARAMETERS FOR
TURNING WASPALOY
CONCLUSION-
This higher predictive accuracy of ANN can be attributed to its
universal ability to approximate nonlinearity of the system whereas
RSM is only restricted to second-order polynomial.
The main advantage of RSM over ANN is its ability to exhibit the
factor contributions from the coefficients in the regression model.
This ability is powerful in identifying the insignificant main factors and
interaction factors or insignificant quadratic terms in the model and
thereby can reduce the complexity of the problem.
35. Investigation of cutting parameters of surface
roughness for a non-ferrous material using artificial
neural network in CNC turning
Surface roughness, an indicator of surface quality is one of the most
specified customer requirements in a machining process.
For efficient use of machine tools, optimum cutting parameters
(speed, feed and depth of cut) are required. So it is necessary to find
a suitable optimization method which can find optimum values of
cutting parameters for minimizing surface roughness.
36. Investigation of cutting parameters of surface
roughness for a non-ferrous material using artificial
neural network in CNC turning
Material -
Brass C26000 material
Parameters-
Spindle speed(rpm)
Feed rate(mm/rev)
Depth of cut(mm)
37. Investigation of cutting parameters of surface
roughness for a non-ferrous material using artificial
neural network in CNC turning
CONCLUSION-
The surface roughness could be effectively predicted by using spindle
speed, feed rate, and depth of cut as the input variables.
Considering the individual parameters, feed rate had been found to be the
most influencing parameter, followed by spindle speed and depth of cut.
As the spindle speed increases for lower feed rates, the surface roughness
decreases
Model (including interaction terms), considering the interaction
between the individual parameters, could achieve an accuracy of
75.6%.
38. Prediction and control of surface roughness in CNC
lathe using artificial neural network
In this study, a neural network approach is presented for the prediction
and control of surface roughness in a computer numerically controlled
(CNC) lathe.
A feed forward multilayered neural network was developed and the
network model was trained.
The adaptive learning rate was used. Therefore, the learning rate was
not selected before training and it was adjusted during training to
minimize training time.
39. Prediction and control of surface roughness in CNC
lathe using artificial neural network
Material-
St 50.2 steel
Size-Ø25mm×100mm
PARAMETERS-
Depth of cut (mm)
Cutting speed (m/min)
Feed rate (mm/rev)
40. Prediction and control of surface roughness in CNC
lathe using artificial neural network
CONCLUSION-
• The feed rate is a dominant parameter and the surface roughness
increases rapidly with the increase in feed rate.
• The cutting speed has a critical value for which the best surface
quality can be achieved. Below this critical value, the surface
roughness decreases with increasing cutting speed and after this
value, the surface roughness increases with increasing cutting speed.
• The effect of depth of cut on surface roughness is not regular and
has a variable character.
• ANN can produce an accurate relationship between cutting
parameters and surface roughness.
42. REFERENCES-
Design of Experiments for Engineers and Scientists by Jiju Antony
Research Papers-
Investigation of cutting parameters of surface roughness for a non-
ferrous material using artificial neural network in CNC turning
Prediction and control of surface roughness in CNC lathe using artificial
neural network
USE OF ANN AND RSM TO MODEL, PREDICT AND OPTIMIZE THE
PERFORMANCE PARAMETERS FOR TURNING WASPALOY