This document summarizes research on using active and passive flow control techniques to improve the performance of Darrieus wind turbines. Wind tunnel experiments were conducted on airfoil sections to characterize the effects of passive "bugs tape" and steady/pulsed boundary layer suction via actuators. A double multiple streamtube model was developed and used to predict turbine performance with and without flow control effects implemented. Results showed that passive flow control improved performance at low wind speeds but degraded it at high winds, while optimized active flow control via pulsed suction increased turbine power output and efficiency across a range of operating conditions.

1. 1
Flow Control Techniques for Improved
Performance of a
Darrieus Wind Turbine
Nir Morgulis
Adviser: Prof. Avi Seifert
Meadow Aerodynamics Laboratory
School of Mechanical Engineering,
Faculty of Engineering
Tel Aviv University, ISRAEL
Funding: Gordon family fund.
Scholarship: Ministry of national infrastructure, energy and water
resources Technical staff: Shlomo Paster, Tomer Bachar, Avraham Blas,
Eli Kronish and Mark Vasserman
Lab Colleagues: Vitali Palei,Victor Troshin, Ori Friedland, Gadi Lubinsky,
Dima Sarkorov, Danny Dolgopyat, Artur Minasyan, Liad Marom, Gideon
Luther Lee, Chen Rossert, Deborah Toubiann.
11-Nov, 2014, Tel-Aviv University
(Neumayer station, Antarctica) (M. Islam et al, (2008))

2. 2
Talk Outline
• Background and motivation
• Wind tunnel measurements of GOE-222 airfoil section
 Experimental Setup
 Passive Flow control
 AFC- Steady State Boundary Layer Suction
 AFC- Pulsed Suction
• Turbine Performance Prediction
 Double Multiple Stream Tube Model
 Passive Flow Control Effect on Performance
 Implementation of Tailored AFC
• Summary and Future Work (windpowerzeyu.com)

3. 3
Background
• Today, most of our energy comes from fossil fuels
(contaminating, limited amount, politics)
• Renewable energy sources are free (once installed),
naturally refurbished and nonpolluting
• Wind turbines types:
Horizontal axis wind turbines
Vertical axis wind turbine
• Drag machines (“Savonius”)
• Lift machines (“Darrieus”)
(windenergy.com) (greenenergyreporter.com)
(cleangreenenergyzone.com)

4. Background
4
Passive Flow Control Active Flow Control
• Doesn’t require energy investment
• Cannot react to changes in flow
• Require energy investment
• Can react to changes in flow
(Prandtl,1904)
(Glauert GlasII,1945)
(Yehoshua and Seifert, 2003, 2006)
(Van Dyke)
Trip Wire
(Gloster Javelin)

5. 5
Motivation
• Improve Darrieus VAWT turbine
performance.
• Characterize airfoil at wind
conditions relevant to the
operation of a VAWT.
• Test the effects of flow control on
airfoil performance.
• Implementation of flow control
effects in performance prediction
models
• Optimize the AFC for more
efficient turbine performance.
Maor Avnaim (MSc TAU, 2011)

6. Experimental Setup - Airfoil
6
• GOE-222 airfoil section
• Chord length of 165mm
• Max thickness 18.5% at 29.2%
chord.
• Actuator ports of 1mm diameter
with 10mm spacing, located at
10% chord length
Actuator Ports
10mm

7. Wind Tunnel Experimental Setup
7
(Seifert et al 1993)
Knapp-Meadow wind tunnel at TAU
Unsteady Air Flow Valve

8. Baseline Airfoil Characteristics
8
-0.2
0.3
0.8
1.3
1.8
-10 0 10 20
Cl
AoA[deg]
Lift Vs. AoARe=75k
Re=100k
Re=150k
Re=200k
Re=300k
Re=400k
-0.2
0.3
0.8
1.3
1.8
0.01 0.03 0.05 0.07 0.09
Cl
Cd
Lift Vs. Drag
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 0.2 0.4 0.6 0.8 1
Cp
X/c
Pressure Distribution
Re=100k 7deg
8deg
9deg
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 0.2 0.4 0.6 0.8 1
Cp
X/c
Pressure Distribution
Re=200k AoA=6deg
AoA=10deg
AoA=14deg

9. Passive Flow Control: “Bugs”
9
-0.2
0.3
0.8
1.3
1.8
-10 0 10 20
Cl
AoA[deg]
Lift Vs. AoABase line
Passive porosity
Bugs10
Bugs20
Bugs 30
-0.2
0.3
0.8
1.3
1.8
0.01 0.03 0.05 0.07 0.09
Cl
Cd
Lift Vs. Drag
Base line
Passive porosity
Bugs10
Bugs20
Bugs 30
At lower wind velocities:
• Passive flow control promoted
laminar-turbulent transition
• Prevented laminar flow separation
bubble that degrades airfoil
performance
“Bugs” tape
Re=100k

10. Passive Flow Control: “Bugs”
At higher wind velocities:
• No laminar flow problems are encountered
• Passive flow control degraded the airfoil performance
• Conclusion: PFC must be tuned and used at a single Re otherwise
it degrades performance 10
-0.2
0.3
0.8
1.3
1.8
-10 0 10 20
Cl
AoA[deg]
Lift Vs. AoABase line
Passive porosity
Bugs10
Bugs20
Bugs 30
-0.2
0.3
0.8
1.3
1.8
0.01 0.03 0.05 0.07 0.09
Cl
Cd
Lift Vs. Drag
Base line
Passive porosity
Bugs10
Bugs20
Bugs 30
Re=200k

11. AFC - Steady Boundary Layer Suction
• 300-2500 Pa sub-atmospheric airfoil cavity
pressures were tested
• Positive effect at high and low Re numbers
• The positive effect of the suction actuators
reaches saturation at low values
11
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
-10 -6 -2 2 6 10 14 18 22
Cl
AoA[deg]
Lift Vs. AoA
Base Line
Cmiu avg-0.0023
Cmiu avg-0.0037
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0.01 0.02 0.03 0.04 0.05
Cl
Cd
Lift Vs. Drag
Base Line
Cmiu avg-0.0023
Cmiu avg-0.0037
2







V
V
A
A
C
jets
foil
jets

foil
jetsjets
wakeDcorrD
AV
AV
CC

  2.
Re=200k

12. AFC - Steady Boundary Layer Suction
12
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 0.2 0.4 0.6 0.8 1
Cp
x/c
Cµ=0.002, Cl=1.783, Cd=0.066
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
0 0.2 0.4 0.6 0.8 1
Cp
x/c
Cµ=0, Cl=1.471, Cd=0.100
x/cSep~0.5 x/cSep~0.6
Lower Min(Cp)
Higher “Base” Cp
Suction
Re=200k, AOA=14deg

13. AFC – Pulsed Suction
• A series of measurements at
AoA=10deg, 100k<Re<400k
• Actuation frequency can enhance the
effect of steady suction by ~40%
• Optimal frequency was found at F
+
=0.6
13
-800
-600
-400
-200
0
200
0 0.02 0.04 0.06
InletPressure[Pa]
t[sec]
Pulsed Suction



V
cf
F
*

14. Double Multiple Streamtube Model
• The Blade Element Momentum (BEM) model presented in 1981 by
Paraschivoiu
• Each streamtube is defined as
• Every streamtube is s divided to upstream and downstream parts, each
calculated as an actuator disk.
14

(M. Islam et al, 2008)
(M. Paraschivoiu et al, 2009)

15. Double Multiple Streamtube Model
15

 2
2
2
cossin 






V
R
VW auu













cos)//()/(
sin
tan 1
VVVR au
u
 VVV aue 2
Upstream half-cycle
Downstream half-cycle

 2
2
2
cossin 






e
add
V
R
VW







 



cos)//()/(
sin
tan 1
eade
d
VVVR
(M. Islam et al,2008)
We begin with an initial value of induced velocity. Actual Va will be found in
an iterative process

16. Double Multiple Streamtube Model
16
)sin()cos(  dln CCC 
)cos()sin(  dlt CCC 
Local normal and tangential force coefficients
 
 2
2
2
2
sincossec
8

 

dCC
V
W
R
Nc
f tn
au
u
up
Momentum balance equation
up
au
fV
V


 

Momentum balance function
Process is repeated until the induced velocity is converged for each
Streamtube half cycle
2
0
1
( )
2at tF F d

 

  atQ NF R QP 0
After we know the tangential forces, turbine power can be calculated:

17. Double Multiple Streamtube Model
• The DMST model offers relativally good accuracy with a short calculation
time (when compared to CFD or other methods.
• Model’s equations can be manipulated to include measured experimented
results.
17
(Paraschivoiu, 2009)

18. Double Multiple Streamtube Model
The current research is focused on turbines that can be effective in
urban settings with 1m to 2m radius
18


V
R
TSR

3
00
5.0 

VA
P
P
P
C
sweptwind
p

(greenenergyreporter.com)

19. Passive FC Effect On Turbine Performance
19
• Passive flow control can be effective for turbines with small radius in low
wind speed conditions
• Example: At V=5m/s max Cp is increased by 10% for a 1m radius turbine
• At higher wind speeds, passive flow control degrades performance
Number of blades-3, Airfoil type- GOE222, Chord length- 0.165m, Radius=1m
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
1.5 2 2.5 3 3.5
Cp
TSR
V=5m/s
Base
Bugs20
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
1.5 2 2.5 3 3.5
Cp
TSR
V=8m/s
Base
Bugs20

20. AFC effect on turbine performance
The AFC is activated constantly during a blade cycle
20
2
0
1
( )
2at tF F d

 

 
atQ NF R
QP 0-15.00
-10.00
-5.00
0.00
5.00
10.00
15.00
20.00
25.00
-90 10 110 210
Ft[N]
Azimuthal Angle[deg]
Tangential Forces On a Single Blade
R=1.4m, V=8m/s, TSR=2.7
Base-Line, Mechanical Power=166[w]
Cmiu=0.001, Mechanical Power=203.3[w]
Cmiu=0.002, Mechanical Power=236.1[w]

21. AFC effect on turbine performance
21
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
1.5 2 2.5 3 3.5 4
Cpt
TSR
N=3, GOE222, c=0.165m,
R=1.4m, Vo =8m/s
Base
Cmiu=0.001
Cmiu=0.002
Energy investment in actuation reduces total system efficiency
Optimization is required

22. Implementation Of Conditioned AFC
22
0
10
20
30
40
50
60
70
80
90
100
300
310
320
330
340
350
360
0 0.0005 0.001 0.0015 0.002
InvestedPower[Watt]
ProducedPower[Watt]
Cmiu
P-blade
P-net
P-jet
-14.25 deg
)(),(),(),(  CCCCCCCWWC ttnn 
   RCCCWACP tfoilblade )()(
2
1
),( 2
)(
2
1
),( 3
  CuACP jjetsjetsjets 
)(]),([ jetsbladeoptopt PPMaxCP 

23. Implementation of Conditioned AFC
• An overall improvement at all tip speed ratios was achieved at both fast
and slow wind conditions
• Improvement of 10.5% for the maximum efficiency
23
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
1.5 2 2.5 3 3.5
Cpt
TSR
V=5m/s
Base-line
Optimized
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
1.5 2 2.5 3 3.5
Cpt
TSR
V=8m/s
Base-line
Optimized
Number of blades-3, Airfoil type- GOE222, Chord length- 0.165m, Radius=1m

24. Implementation of Conditioned AFC
Efficiency improvement on turbines with larger radiuses as well (in
contrast to the passive flow control), at the lower TSRs.
24
-0.05
0.05
0.15
0.25
0.35
0.45
1.5 2 2.5 3 3.5 4
Cpt
TSR
N=3, GOE222, c=0.165m,
R=1.4m, Vo =5m/s
Base-line
Optimized

25. Summary and Future Work
• Wind turbine performance enhancement using flow control demonstrated
• Flow control effects on turbine performance were modeled
• Passive flow control is a single design tool
• Enhanced airfoil performance translated to better turbine performance
• Model included energy based AFC scheduling
• AFC proved to be a versatile tool for performance improvement, when
properly used
• Conditioned boundary layer suction can increase max Cp by approx. 10%.
• At sub optimal TSR it significantly enhance efficiency
Future Work
• Modeling of pulsed suction and actuation frequency in DMST model
• Inclusion of dynamic stall and blades interaction for higher fidelity model
• Experimental validation on a real 1-2m radius turbine
25

26. Additional slide 1
26