2nd european forum on flow control - univ … · 2nd european forum on flow control effc poitiers,...

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 »)


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.


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


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


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


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


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


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


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


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


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


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


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


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


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


• 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


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


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


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


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


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


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


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


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


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%


90o, 0.5mm orifice

Cl

Cd


• 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


Zero-Net-Mass-Flux JetsTel-Aviv University (TAU)

participants: A. Seifert and O. Stalnov


• 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


• 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


• 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


• 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


• 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


• 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



• 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


Development of Multi Orifice Single Chamber

Zero-Net-Mass-Flux JetsManchester University

participants: L.D Gomes


– 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)


• Study and enhance the performance of a multi-orifice single chamber actuator

Design of Multi-Orifice Single Chamber Actuator

Design of the first prototype


– 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


– 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

2nd european forum on flow control - univ … · 2nd european forum on flow control effc poitiers,...

Documents