2nd european forum on flow control - univ … · 2nd european forum on flow control effc poitiers,...
TRANSCRIPT
GDR CDD Toulouse 21/11/06
2nd EUROPEAN FORUM on FLOW CONTROLEFFC
Poitiers, France May-July 2006
Organized byLaboratoire d’Etudes AérodynamiquesCNRS - Université de Poitiers – ENSMA
Chairmen: JP Bonnet and F Alvi (FSU)
Sponsored by :AIRBUS within the AIRNet programme
Berlin/Madrid/Manchester/Poitiers –
Control of Aerodynamic Flows for the Environmentally Driven Aircraft
CAFEDA
Centre National de la Recherche Scientifique (GDR « contrôle des décollements »)
GDR CDD Toulouse 21/11/06
European Forum on Flow Control,Network for training and research
• The European Forum on Flow Control (EFFC) is devoted to common research and exchange of ideas during relatively long periods (3 months) involving resarchers at a post doctoral level, engaged into research in flow control for several years. All the conditions for real synergysm will be given both for theoretical, numerical or experimental approaches.
• Positions will be available for durations up to 3 months for researchers in flow control. Participants to the programme can be any active european researcher in the domain. A selection being is based on research programmes proposed by the candidates. The list of the selected topic is given, but researchers not directly implied in the topics can be eligible for wider exchanges during the sessions.
GDR CDD Toulouse 21/11/06
The EFFC-2nd session, May-July 2006
JP Bonnet, F Alvi, J Delville and P Jordan
• A. Experimental study of unsteadyseparation on a NACA 0015 airfoil
• B. Low order description of turbulent flowin view of flow control (EFFC-1 cont’d)Towards Quiet turbulence
GDR CDD Toulouse 21/11/06
Collaborative research programme EFFC 2006
1st group: experiments
– Experimental control of flow separation on a NACA 0015 wing (CNRS GDR) in the 2.4 x 2.4 Poitiers windtunnel.
• 7 Participants: C Atkinson & S Trevor (Monash, Australia ), F Alvi& V Kumar (Florida State Univ. USA), A Seifert and O Stanlov(TAU, Israel ), L Gomes (Univ Manchester, G-B)
• WL Siaw, S Bourgois, J Tensi, JP Bonnet, LEA
GDR CDD Toulouse 21/11/06
Collaborative research programme EFFC 2006
2nd group: theory and computation :(J Delville and P Jordan + 10 participants)Post processing and closed-loop methods for
turbulent flow controlTowards Quiet turbulence (J Freund)
Jonhatan FREUND Urbana
WEI Stanford
Caroline Braud IMFL Lille
Laurent CORDIER Nancy/Poitiers
Oksana STALNOV & A SEIFERT TAU Tel Aviv
Maja WÄNSTRÖM & W GEORGE Chalmers
Dandy ESCHRICHT TU Berlin
Michael SCHLEGEL TU Berlin
Bernd NOACK TU Berlin
GDR CDD Toulouse 21/11/06
Main results of the collaborative Studies on Flow Separation Control over a NACA0015
W.L. Siauw , J.P. Bonnet, J. Tensi, J.M. Breux, W.H. Khoo,Poitiers University (LEA-ENSMA-CNRS)
A. Seifert, O. Stalnov, Tel Aviv University (TAU)B. V. Kumar, F.S. Alvi, Florida State University (FMRL)
C.H. Atkinson, Monash University (LTRAC)
L.D. Gomes Manchester University
GDR CDD Toulouse 21/11/06
Presentation Outline
• Introduction• Model, Fluidic Actuators & Test Conditions• Results
– Angled & Normal Steady Jets– Zero-Net-Mass-Flux (ZNMF) Jets– Development of Multi Orifice Single Chamber (ZNMF)
Jets• Conclusion• Future Work
GDR CDD Toulouse 21/11/06
Introduction
• Objectives – Study the effects of different fluidic actuators on the
same NACA0015 airfoil– Obtain an approximation of the timescales of
attachement & separation in view of re-active flowcontrol
– Develop a multi orifice single chamber syntheticZNMF actuator
GDR CDD Toulouse 21/11/06
Test Facility
• Closed loop tunnel• Test section 2.4m by
2.6m• Turbulence intensity =
0.4% at 40m/s• Instrumentation
– Force measurement– Pressure measurement– Wake suvey
Test section 2.4m by 2.6m
GDR CDD Toulouse 21/11/06
Model & Test Conditions
• Model specification– NACA0015 - 0.35m (chord) &
2.4m (span)– Model turbulated by 8 micron
carborandum at x/c of 2% to 4%.
• Flow conditions– Free-stream velocity = 40m/s– Re = 0.96 million
wind
Plan view of wing
2.4m
0.35
m
Externalbalance
GDR CDD Toulouse 21/11/06
Fluidic Actuators Specifications
• Orifices distributedapproximately 1 third of the airfoilspan
2.4m
0.35
m
Single row of orifices machined onto a removeable cover
560.3piezo-electric
Amplitude/plused modulated
“Normal” ZNMF Jet (1mm diameter)
640.3continuousSteady “Normal” Jets (0.5mm diameter)
510.3
normal to surface
continuousSteady “Normal” Jets (1mm diameter)
440.330deg (pitch), 60deg (yaw)
pressurized cavity
continuousSteady “Angled” Jets (1mm diameter)
Number of Orifices
Position (x/c)
Jet Orientation
Means of Deployment
Mode of Deployment
GDR CDD Toulouse 21/11/06
Angled & Normal SteadyJets
Poitiers University (LEA-ENSMA)Monash University (LTRC)
Florida State University (FMRL)
Participants : W.L. Siauw , J.P. Bonnet, J. Tensi, J.M. Breux, W.H. Khoo, F.S . Alvi, V. Kumar, C.H. Atkinson, T. Stephens, L.D. Gomes
GDR CDD Toulouse 21/11/06
Results : Baseline surface flow visualization with orifices cover ed
• Quasi 2D separation up to 11 o.• Transition to 3D separation
near 12o.• Formation of 3 stall cells
starts at 13° onwards.• Cl & Cd comparison were
made with orifices exposed.
-0.1
0.1
0.3
0.5
0.7
0.9
1.1
0 2 4 6 8 10 12 14 16alpha (deg)
CL
baseline with orifices covered
baseline with orifices (directed jets) exposed
00.040.080.120.16
0 2 4 6 8 10 12 14 16alpha (deg)
Cd
baseline with orifice (directed jet) exposed
baseline with orifice covered
• With orifice exposed : higher Cl after 8 o
(possible that directedorifices behave as a mechanical vortex generator)
• Starts to stall at 9 o
Incidence 11°
Incidence 15°
Flow direction
Flow direction
GDR CDD Toulouse 21/11/06
Model Overview
Cavity integrated on the
under side of the cover
Interior of NACA0015
Single row of jet orifices orientated at 30 o (pitch) and 60o (yaw)
30o
60o
GDR CDD Toulouse 21/11/06
Results (Angled Steady Jet, orifice diameter = 1mm at x/c=0.3) : Surface flow visualization at 12 o & 15o
α =12°baseline
α =12°, C µ~0.36%, VR~3.5
• (12o to 16o) Full attachment wasobserved.
• At 15°, stall cells atthe central portion was eliminated.
• In all cases, adjacent stall cells(if exist) becomingstronger.
α =15°baseline α =15°, C µ~0.36%, VR~3.5
Flow direction
Flow direction
Flow direction
Flow direction
Orifice, x/c=0.3Orifice, x/c=0.3
Orifice, x/c=0.3Orifice, x/c=0.3
GDR CDD Toulouse 21/11/06
• At Cµ~0.36%, Cl improvebetween 3% and 14% (16 o).
• Stall characteristic is more gradual when jets weredeployed
Results (Angled Steady Jet, orifice diameter = 1mm at x/c=0.3) : Lift coefficient comparison during actuator deployment
0.85
0.9
0.95
1
1.05
1.1
1.15
9 10 11 12 13 14 15 16 17
alpha (deg)
CL
baseline with orifices exposedC_mu~0.27%, VR=2.5C_mu~0.36%, VR=3.5
30o60o
* Lift improvement could be 3 times more since jets were deployed over 1/3 of airfoil span
GDR CDD Toulouse 21/11/06
Results (Angled Steady Jet, orifice diameter = 1mm at x/c=0.3) : Drag coefficient comparison dring actuator deployment
• Cd reduced significantly for incidences above 8 o(up to 55%)
• Cd(Cµ~0.36%) > d(Cµ~0.27%)
30o
60o
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
9 10 11 12 13 14 15 16
alpha (deg)
Cd
baseline C_mu~0.27%, VR=2.5 C_mu~0.36%, VR=3.5
GDR CDD Toulouse 21/11/06
Results (Directed Steady Jet, orifice diameter = 1mm at x/c=0.3 ) : Time scale of attachment & separation at 11 o incidence
• Use error function for curve fitting : f(x) = 1/2*(d+b*erf(1.77245/a*(x+c)))
• Typical time for attachment= 0.1127sec (T+~42 based on sep length or 18 actuator/TE)
• Typical time for separation = 0.207sec (T+~84 or 34)
• (16 and 26 for Darabi & Wignanski)
11mm
10mm
Hotwire position
separation
α =11°baseline α =11°, C µ=0.27%, VR=2.5
Orifice, x/c=0.3
sepration, x/c=0.85
Orifice, x/c=0.3
Flow direction Flow direction
GDR CDD Toulouse 21/11/06
General model:f(x) = 0.5*(d+b*erf(1.77245/a*(x+c)))
Coefficients (with 95% confidence bounds):a = -0.1127 (-0.114, -0.1114)b = -13.89 (-13.91, -13.87)c = -3.714 (-3.714, -3.713)d = 28.61 (28.59, 28.64)
Goodness of fit:SSE: 1.449e+005R-square: 0.9587Adjusted R-square: 0.9587RMSE: 1.39
40m/s
GDR CDD Toulouse 21/11/06
General model:f(x) = 0.5*(d+b*erf(1.77245/a*(x+c)))
Coefficients (with 95% confidence bounds):a = 0.207 (0.2054, 0.2086)b = -13.69 (-13.7, -13.68)c = -2.325 (-2.325, -2.324)d = 28.29 (28.28, 28.3)
Goodness of fit:SSE: 3.843e+005R-square: 0.9654Adjusted R-square: 0.9654RMSE: 1.269
40m/s
GDR CDD Toulouse 21/11/06
Results (Normal Steady Jet, orifice diameter 1mm at x/c=0.3) : Comparison of Cl with/without Actuation
• No change in Cl atincidences < 8 o
• Improvement in Cl - up to 5%
• Cl increases with C µ for incidences 8 o to 12o
* Lift improvement could be 3 times more since jets were deployed over 1/3 of airfoil span
90o, 1mm orifice
0.8
0.85
0.9
0.95
1
8 9 10 11 12 13 14 15 16
Incidence (deg)
CL
Baseline
,C_mu~0.03%VR~0.9
,C_mu~0.05%VR~1.1
,C_mu~0.8%VR~1.3
,C_mu~0.14%VR~1.8
GDR CDD Toulouse 21/11/06
Results (Normal Steady Jet, orifice diameter 1mm at x/c=0.3) : Comparison of Cl with/without Actuation
• For incidence less than12o, Cd seems to bemarginally reduced(uncertainty in Cd~0.0026).• For incidences above12o, Cd has the tendencyof increasing with C µ
except at 15 o)
Comparison of Cd at different C_mu
0.03
0.05
0.07
0.09
0.11
0.13
0.15
8 9 10 11 12 13 14 15 16incidence (deg)
Cd
baseline
C_mu~0.03%
C_mu~0.05%
C_mu~0.08%
C_mu~0.14%
90o, 1mm orifice
GDR CDD Toulouse 21/11/06
Results (Normal Steady Jet, orifice diameter 0.5mm at x/c=0.3) : Surface flow visualization at 12 o
• Flow attached over 85% ofairfoil after actuator deployed• Stall cells, on adjancent sidesof separation delay zone, start to form and/or intensify
α =12°baseline
α =12°, C µ~0.4%, VR~8
Flow direction
Flow direction90o, 0.5mm
orifice
GDR CDD Toulouse 21/11/06
Results (Normal Steady Jet, orifice diameter 0.5mm at x/c=0.3) : Comparison of Cl with/without Actuation
• Jet deployed at soniccondition,
– Cµ ~ 0.4%, VR ~ 8• Cl_max shifted to higherincidences• 3% to 7% improvement in Cl
beyond AOA = 8 o
• Cd reduced by 15% -22% • For fixed incidence Cl increases ~ linearly with C µ.
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
8 9 10 11 12 13 14 15 16
Incidence (degree)
Cd
0.8
0.85
0.9
0.95
1
1.05
1.1
CL
Drag Baseline Drag C_mu~0.4%Lift Baseline Lift C_mu~0.4%
* Lift improvement could be 3 times more since jets were deployed over 1/3 of airfoil span
90o, 0.5mm orifice
Cl
Cd
GDR CDD Toulouse 21/11/06
• Angled Steady Jet with diameter of 1mm ( 0.27%<Cµ<0.36%, 2.5<VR<3.5)– Cl improves by 5% - 16% depending upon incidence– Cd reduced 30% to 50% – Typical time for attachment/separation at 11deg ~ 0.1sec (T+=42).
• Normal Steady Jet with diameter of 1mm (0.03%<Cµ<0.14%, 0.9<VR<1.8)– Cl improves by 3% - 5% depending upon incidence– Increases Cd for incidences above 12 o, except at 15 o
• Normal Steady Jet with diameter of 0.5mm (Cµ~0.4%, VR~ 8)– Cl improves by 3% - 8% depending upon incidence– Cd reduced 15% - 22%– More effective after 8 o incidence (no separation before this incidence)
Results Summary : Steady Jets
GDR CDD Toulouse 21/11/06
Zero-Net-Mass-Flux JetsTel-Aviv University (TAU)
participants: A. Seifert and O. Stalnov
GDR CDD Toulouse 21/11/06
• Integral part of the model cover• 14 localized PZT actuators• Four, 1mm diameter holes for
each localized actuator• Perpendicular to surface• Located at x/c=0.3
ZNMF actuator - wing Integration
GDR CDD Toulouse 21/11/06
• Conditions at lower speed:– Reynolds number = 0.25x10 6
free stream velocity = 10m/s– Incidence = 13 o
– Cµ=0.32%, VR=3– Provides stall control
• Amplitude modulation (AM)– Actuator operated at 1.95kHz
(F+=30 )– Modulated with a sine wave
at 41 Hz (F+=0.6)• Effects on Cl & Cd
– Better drag reductioncharacteristics at lowincidences
– Cl improvement up to 4%(could be 3 times better)
Effects of ZNMF Deployment
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
8 9 10 11 12 13 14 15 16Incidence(deg)
Cl Cl baseline
Cl, AM, F+=0.6
Cl, F+=30
00.10.20.30.40.50.60.70.80.9
1
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Cd
ClCl baseline
Cl, AM, F+=0.6
Cl, F+=30
GDR CDD Toulouse 21/11/06
• Conditions– Re = 0.25mil, free-
stream velocity = 10m/s
– Incidence = 13 o
– Pulse modulated at0.5Hz
• Good correlationbetween movement of tuffs and actuatordeployment
Effects of Synthetic Jet Deployment
GDR CDD Toulouse 21/11/06
• Conditions– Re = 0.25 mil, free stream velocity = 10m/s– Model incidence = 13 o
• Operation principal – Actuator Peak velocity = 30m/s– Operate the PZT at 1.95kHz– Modulated by pulse at 0.5Hz– Measure static pressure near TE
Measurement of Step Response
PZT located at x/c=0.3Kulite pressure
transducer located atx/c=0.83
GDR CDD Toulouse 21/11/06
• PZT operation– PZT actuated at 1.95kHz corresponding to F +=30– Pulse modulated at low frequency (0.5Hz)
• Estimated response time during ON state = 0.2sec (T +=11) • Estimated response time during OFF state = 0.6sec (T +=33)• Comparable with Darabi and Wygnanski JFM 2004• Independency against actuator configuration not obvious
Estimation of Response Time
Ensembled Average of 10 Events
GDR CDD Toulouse 21/11/06
• Cµ=0.02%, VR=3
• Max of 1% improvement in Cl
• No change in Cd
Effects of ZNMF Deployment at Re=1 million
Re=1e6, d=1mm, S=12mm, x/c(holes)=0.3, F+~8
0.86
0.88
0.9
0.92
0.94
0.96
0.98
1
1.02
0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Cd
Cl
baseline
Amplitudemodulated
* Lift improvement could be 3 times more since jets were deployed over 1/3 of airfoil span
GDR CDD Toulouse 21/11/06
• AM provides significant drag reduction at low incidences
• The experimental setup enablesaccurate and repeatabledetermination of the flow stepresponse
• Clear visual correlation with the lowfrequency component of the AM signal
• Orifice configuration not optimized
Results Summary : TAU ZNMF Actuator
GDR CDD Toulouse 21/11/06
Development of Multi Orifice Single Chamber
Zero-Net-Mass-Flux JetsManchester University
participants: L.D Gomes
GDR CDD Toulouse 21/11/06
– Measurement made in the absence of cross flow fro; the original actuator from Manchester
– Response time:
• Reponse time during ON state is of the order of a few millisec.
• Reponse time during OFF state is 3 to 4 times larger (compared to ON state).
Time Response of Cylindrical Chamber Single Orifice Actuator
– Characteristics are well within the timescales of separation and attachment (order of 0.1sec to 0.5sec)
GDR CDD Toulouse 21/11/06
• Study and enhance the performance of a multi-orifice single chamber actuator
Design of Multi-Orifice Single Chamber Actuator
Design of the first prototype
GDR CDD Toulouse 21/11/06
– Test conditions:• Excitation frequency = 1,270Hz
• Excitation voltage = 140Vpp
• Height of hotwire probe from surface ≈2Do
• Orifice diameter, D0 = 1.2mm• Hotwire diameter = 1.0mm
– Averaged peak jet velocity reached = 50m/s
– Further work to exceed 100m/s• Optimization of chamber
volume, increasing input voltage range (up to +/-250V) to further augment the output vel.
• Assessment of actuator power consumption and efficiency
Characteristics of Multi-Orifice Single Chamber Actuator
7 orifices chamber
Hotwire
Orifices
GDR CDD Toulouse 21/11/06
– Conclusion:• Explored the ability of steady & ZNMF jets in the control of separated
flow over a NACA0015.• Developed a preliminary version of a multi orifice single chamber ZNMF
actuator.• More importantly, the test/research program allows a multi-national
(France, Singapore, Australia, USA, UK and Israel) cooperative effort to study the control of flow separation and develop actuator/devices that controls it .
– Future Work• Implementation of optimized ZNMF onto the NACA0015 and to gain
better insight on the response time during attachment/separation.
• Continual effort to make detail study on the effects of angled steady jets using PIV (ENSMA).
• Other participants are welcome for future EFFC.
Conclusion & Future Work