yrt

ntaintyra

and-rotare-rotde

rticyerhee vor cthery lise-bowi

1

da

Ogaerodynamic characteristics are affected 1,2. In order to ascer-tain the underlying performance characteristics of race car frontw

fa

flim

s

v

lw

c

e

ta

gdt

ber of different fields, such as vortex generators VGs, VG jets,spanwise cylinder, and wall heat transfer techniques 1518.

JsK

J

Downloaings, a number of experimental investigations have been per-ormed for an isolated wing in ground effect; a single-element35, a double-element wing 68, a gurney flap fitted wing9,10, a wing in a wake flow 11,12, interactions between a wingnd wheel 13, and a dynamically motioned wing 14.

The experiments of Zerihan and Zhang 4,5 revealed importantow characteristics of a single-element wing in ground effect us-

ng a number of experimental methodologies, including forceeasurement, surface flow visualization, surface pressure mea-

urement, laser Doppler anemometry LDA, and particle imageelocimetry PIV. A moving ground was used to properly simu-ate the conditions on a race track, where a moving belt rig in theind tunnel runs at the same speed as the freestream in order to

ontrol the boundary layer on the ground. Zerihan and Zhang 4xamined the ride height sensitivity of the downforce characteris-ics; the downforce behavior with respect to the ride height shows

downforce enhancement as a wing is moved closer to theround, and then a downforce reduction is shown by further re-uction in the ride height. Surface flow visualization revealed thathe downforce reduction phenomenon is induced by a breakdown

Lin 19,20 suggested that effective devices for separation con-trol are those that generate streamwise vortices, such as thoseproduced by VGs. Large-scale vortex generators LVGs, whosedevice height is of the same order as the boundary layer thickness,have been used to control flow separation by transferring the mo-mentum from the outer flow to boundary layer flow. Sub-boundary layer vortex generators SVGs, however, appear to bemore advantageous in terms of their effectiveness with a lowerdevice height, which is just fraction of the boundary layer thick-ness 1921. Lin 19 exhibited that optimally placed vane typeSVGs with a device height of 20% of the boundary layer thick-ness performs as well as vane type LVGs with device height of80% of the boundary layer, but with less device drag.

Pauley and Eaton 22 studied the mean streamwise develop-ment of pairs of longitudinal vortices, generated by delta vanetype VGs with a device height of 150% of the boundary layerthickness, in a turbulent boundary layer. Several vortex configu-rations were investigated, i.e., corotating vortices and counter-rotating vortices, inducing common-downwash and common-upwash. The streamwise decay of the peak vorticity showed thatthe corotating vortices decay quicker than the common-downwashvortices. The arrayed counter-rotating vortices show a slightlyslower decay rate in the vortex circulation than the other configu-rations. Godard and Stanislas 23 attempted an optimization fordelta vane type SVGs. The optimized counter-rotating sub-boundary layer vortex generators CtSVGs shows higher effec-tiveness than the corotating sub-boundary layer vortex generatorsCoSVGs.

1Present address: Williams F1.2Present address: TotalSim LLC.Contributed by the Fluids Engineering Division of ASME for publication in the

OURNAL OF FLUIDS ENGINEERING. Manuscript received March 20, 2009; final manu-cript received October 1, 2009; published online November 19, 2009. Editor: Josephatz.

ournal of Fluids Engineering DECEMBER 2009, Vol. 131 / 121103-1Copyright 2009 by ASMEYuichi Kuyae-mail: yuichi@soton.ac.uk

Kenji TakedaSenior Lecturer

e-mail: ktakeda@soton.ac.uk

Xin ZhangProfessor

e-mail: x.zhang1@soton.ac.uk

Scott Beeton1

Ted Pandaleon2

School of Engineering Sciences,University of Southampton,

Southampton, Hampshire SO17 1BJ, UK

Flow PhWith VoEffectThis paper experimeon an inverted wingflow visualization. Atunnel over a widesub-boundary layercomprising counterheight and spacingerators and counterat the center of eachinduce horseshoe voing sub-boundary lacontrol. Increasing tsignificant horseshoscale vortex generatwing equipped withrotating sub-boundaance in the spanwcounter-rotating subcontrol for race car

IntroductionRace car performance can be significantly enhanced by the ad-

ition of inverted wings to create downforce, improving tractionnd cornering ability. For open-wheel race series such as Formulane and Indy Racing, inverted wings and diffusers operate inround effect, that is, in close proximity to the road, and theirded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMsics of a Race Car Wingex Generators in Ground

lly investigates the use of vortex generators for separation controlground effect using off-surface flow measurements and surface

pical racing car wing geometry is tested in a rolling road windnge of incidences and ride heights. Rectangular vane type oflarge-scale vortex generators are attached to the suction surface,

ating and corotating configurations. The effects of both deviceexamined. The counter-rotating sub-boundary layer vortex gen-

ating large-scale vortex generators suppress the flow separationvice pair, while the counter-rotating large-scale vortex generatorses between each device where the flow is separated. The corotat-vortex generators tested here show little evidence of separation

spacing of the counter-rotating sublayer vortex generator inducesortices, comparable to those seen in the counter-rotating large-ase. Wake surveys show significant spanwise variance behind thecounter-rotating large-scale vortex generators, while the counter-ayer vortex generator configuration shows a relatively small vari-direction. The flow characteristics revealed here suggest thatundary layer vortex generators can provide effective separationngs in ground effect. DOI: 10.1115/1.4000423

of edge vortices at the end plates and flow separation on the suc-tion surface of the wing, which is induced by the large adversepressure gradient.

If a flow is in a large adverse pressure gradient region, thestreamwise momentum of the flow is reduced, and the flow mayseparate from the wall 15,16. A number of separation controlmethods exist and have been successfully implemented in a num-E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

ibpa2wswRtrtm

2

fvnsaccbra0evic

suT

profile, type LS1-0413, and is manufactured by carbon fibercomposite. The model has a span of 1100 mm, constant chord c of0.4

Freestream U

inVG

1

DownloaAs described in Ref. 24, the use of VGs for an inverted wingn ground effect for separation control can be effective in terms ofoth downforce and efficiency. This paper investigates the flowhysics that induce such force improvements, examining time-veraged and unsteady on- and off-surface flow features. Sectiondescribes the experimental testing procedure in a wind tunnel,here a moving ground condition is properly simulated. The re-

ults of surface flow visualization and PIV measurement of theake are described in Sec. 3, followed by discussion in Sec. 4.ectangular vane type of SVGs and LVGs are tested on the suc-

ion surface of the wing, comprising counter-rotating and co-otating configurations. Different device spacings are also studiedo see how this affects the flow characteristics. Concluding re-

arks are included in Sec. 5.

Experimental Setup2.1 Test Facility. The experiments described here are per-

ormed in the 2.1m1.5m closed section wind tunnel at the Uni-ersity of Southampton. This wind tunnel has been used for aumber of ground effect aerodynamics studies by different re-earchers, including Zerihan and Zhang 4,5,7,8,10 and Zhangnd coworkers 2527. The tunnel is of conventional return cir-uit design, and is equipped with a moving belt rig and a three-omponent overhead balance system. The 3.2m1.5m movingelt is controlled by slots and suction system for boundary layeremoval, which gives 99.8% of the freestream velocity at 2 mmbove the belt. The turbulence intensity of the freestream is about.3%. Further descriptions of the wind tunnel are given by Burgint al. 28. For the experiments presented here, the freestreamelocity U and moving belt speed are set at 30 m/s, correspond-ng to the Reynolds number Re of 450,000 based on the winghord c.

2.2 Experimental Models. Figure 1 shows a schematic of theingle-element wing geometry and installation used. The modelsed is an 80% scale model of the main element of the 1998yrrell 026 F1 car front wing, which is based on a NASA GAW

0.2x/c

0.40 0.6 0.8 1.0 1.2

0.2

0

-0.2

y/c

h VG U

Moving ground

End plate

2.6

Fig. 1 Schematic of single-element wing, end plate, and VG

za(a) zb

4hVG dVG

4hVG

15

x/c=0.537

Fig. 2 Configurations of VGs on wmeasurement: a counter-rotatingfrom bottom to top.

21103-2 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASM223.4 mm, and finite trailing edge of 1.65 mm, and is the samemodel used by Zerihan and Zhang 4,5,10. The origin of thecoordinate system is set at the leading edge of the wing. For thesurface flow visualization, generic end plates of 250mm100mm4mm are attached on both ends of the wing, mean-while a size of 400mm170mm4mm glass end plates are usedfor the PIV measurement. The span of the wing is long enoughsuch that edge vortices induced around the end plates do not affecta flow around the center portion of the wing span where the PIVmeasurement is performed. Therefore, the size difference of theend plates between the PIV measurement and flow visualization isdeemed negligible for wake flow characteristics at the center por-tion. The incidence is measured relative to a line from thetrailing edge to the most swelled point on the pressure surfacewhich corresponds to 2.6 deg relative to the chord line. The trueincidence is therefore equal to the measured incidence plus 2.6deg. When the incidence is 1 deg, corresponding to the true inci-dence of 3.6 deg, the upper and lower edges of the end plates areparallel to the ground. The incidence is varied by rotating aboutthe quarter chord position. The ride height h is defined by thedistance from the lowest point on the suction surface of the wingto the moving ground as the incidence is fixed at 1 deg. The rideheight and incidence are fixed at h /c=0.090 and 1 deg in all themeasurements conducted here, respectively.

Rectangular vane type VGs are employed here, comprisingthree configurations, which are the CtSVGs, counter-rotatinglarge-scale vortex generators CtLVGs, and CoSVGs. The SVGand LVG have a device height of 2 mm hVG /c=0.009 and 6mm hVG /c=0.027, respectively. The VGs are made of alumi-num plate with 0.6 mm thickness and built in pairs separated by4hVG at the trailing edge of the VGs. The vanes are oriented at15 deg relative to the streamwise direction, comprising thecounter-rotating or co-rotating VG configuration. Pairs of VGs areput side by side along the span of the wing, as shown in Fig. 2.The VGs are attached on the suction surface of the wing such thatthe trailing edge of the VGs is fixed at x /c=0.537. The height tolength ratio of the vanes is fixed at 1:4. For the CtLVG andCoSVG configurations, the device spacing dVG between each de-vice pair of the VGs is fixed at 4hVG, while close- and wide-spacings of 2hVG and 8hVG are also examined with the CtSVGconfiguration in addition to the reference-spacing of 4hVG. Unlessthere is a particular notation, CtSVG represents a CtSVG configu-ration with the reference-spacing of 4hVG, which is the same de-vice spacing as the other VG configurations.

2.3 Experimental Methods. The flow visualization is under-taken using a mixture of liquid paraffin and fine powder InvisibleBlue T70 painted on the suction surface of the wing.

PIV measurement is performed in the xy-plane along the centerline of the wing using a Dantec Dynamics FlowMap PIV2100Processor Denmark 29. The laser is a New Wave ResearchGemini dual Nd-YAG laser Fremont, CA, USA mounted down-stream of the wing to illuminate the wake section. The laser sheet

(b)

4hVG

zc

4hVG

15

x/c=0.537

dVG

g and laser sheet positions in PIVs and b co-rotating VGs. Flow is

Transactions of the ASMEE license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

ptdatisHalo1icsmfiT

B

J

Downloaosition is altered for different VG geometries, as shown in Fig. 2,o compare the difference of the wake structure in the spanwiseirection. For the counter-rotating VGs, two laser sheet positionsre employed, which are at the center of the device pair za and athe center between the device pair zb. A single laser sheet positions used at the center of the device pair zc for the co-rotating VGs,ince the vanes are mounted periodically. A Dantec Dynamicsi-Sense CCD camera with a resolution of 12801024 pixels

nd a 105 mm lens is used. The time interval between the twoaser pulses is 15 s, and 500 pairs of snapshots are continuouslybtained. The flow is seeded by smoke particles with a size of1.5 m. FlowManeger 4.10 software is used for postprocess-

ng to calculate the velocity field with the adaptive cross-orrelation algorithm, which uses iteration steps for offsetting theecond window for cross-correlation analysis, using four refine-ent steps with a local median validation function to obtain anal interrogation area size of 3232 pixels without filtering.he overlap between interrogation areas is set at 50%.

(a)

(c)

C

D

Fig. 3 Surface flow visualization on suction=1 deg and h /c=0.090: a clean, b CtSVtom to top.

ournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASM2.4 Uncertainty. The uncertainty of the PIV measurement isestimated by assessing the normalized maximum velocity deficitof the clean wing at x /c=1.5. The 500 PIV snapshots for eachmeasurement are divided into five subsets containing 100 snap-shots, and the uncertainty is assessed between the five subsetsusing procedures described in Refs. 30,31. According to the pro-cedures, the standard error of the measurement is given as 0.008,and hence, the uncertainty with the 95% confidence is given as0.017, where the coverage factor of 2 is used.

3 Results3.1 Surface Flow Visualization. Surface flow visualization

results are presented here to ascertain key flow physics that drivethe force and pressure characteristics presented by Kuya et al.24. Figure 3 illustrates results of the surface flow visualizationon the suction surface around the centre portion of the wing span

(b)

(d)

(b)

A

rface around centre portion of wing span atc CtLVG, and d CoSVG. Flow is from bot-

DECEMBER 2009, Vol. 131 / 121103-3suG,E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

afl

lbTsecms

cclooctvvcsiiwaisebdsstVirieei

surd

1

Downloat =1 deg and h /c=0.090 for the four configurations. Theow direction is from bottom to top in the figures.All of the configurations show that the flow transitions from

aminar to turbulent at about 30% chord via a short reattachmentubble in agreement with the results of Zerihan and Zhang 4.he flow pattern near the end plates not shown here features thepanwise component of flow toward the wing center due to thedge vortices induced around the end plates. The edge vorticesreate a downwash on the suction surface, which pumps the mo-entum into the boundary layer flow, and thus, there is no flow

eparation around the spanwise end of the wing.For the clean wing in Fig. 3a, characteristic horseshoe vorti-

es and flow separation can be seen between 65% and 80%hords. The vortices induced by the counter-rotating VGs areikely to flow straight, since the spanwise component of the sec-ndary flow induced by the VG-generated vortices cancels eachther near the surface. Meanwhile, the flow pattern of the CoSVGonfiguration features the spanwise component of the flow due tohe lateral component of the secondary flow of the VG-generatedortices. In general, since a secondary flow induced by co-rotatingortices flows in the same direction near a surface, the spanwiseomponent is enhanced. Downwash regions can be seen down-tream of each pair of the CtSVGs as represented by the region An Fig. 3b. At the center of each device pair, the flow separations suppressed by the vortices, where each vane induces a down-ash on the suction surface. The vortices allow the flow to remain

ttached up to 95% chord, eventually breaking down at the trail-ng edge; the close proximity to the trailing edge inferring that thetrength of the VG-generated vortices is nearly optimal. Betweenach device pair, an upwash region can be observed as representedy the region B in Fig. 3b induced by the VG-generated vorticesetaching from the suction surface; the upwash enhances the floweparation. For the CtLVG configuration in Fig. 3c, the floweparation is entirely suppressed at the center of each device pairo the trailing edge due to the strong downwash induced by theG-generated vortices. In this case the horseshoe vortices are

nduced between each device pair at 70% chord, as denoted by theegion C. The rotational direction of the counter-rotating vorticess likely to induce the flow outwards, deviating from the center ofach device pair, as represented by the region D. The flow fromach device pair meets at the center of the device pairs, as shownn the region C, where the upwash induced by the VG-generated

(a)

F

E

Fig. 4 Surface flow visualization on suctioneffect of VG device spacing at =1 deg anspacing. Flow is from bottom to top.

21103-4 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMvortices is enhanced by the two vortices, resulting in the separatedflow with the horseshoe vortices. The CoSVG configuration ap-parently shows flow separation downstream of the VGs over thewing span at about 80% chord, as shown in Fig. 3d. The sepa-rated flow pattern is more two-dimensional rather than three-dimensional as featured by the horseshoe vortices on the cleanwing. As will be discussed later, the reduced effectiveness of theCoSVGs regarding the separation control is likely to be due toquicker decay of co-rotating vortices compared with counter-rotating vortices.

The close- and wide-spacings dVG=2hVG and 8hVG are exam-ined in addition to the reference-spacing dVG=4hVG in theCtSVG configuration. Figure 4 shows results of the surface flowvisualization on the suction surface of the CtSVG configurationswith close- and wide-spacings at =1 deg and h /c=0.090. Forthe close-spacing CtSVG configuration, the flow pattern is similarto that of the reference-spacing CtSVG configuration; the regionsE and F in Fig. 4a correspond to the regions A and B in thereference-spacing CtSVG configuration shown in Fig. 3b. Forthe wide-spacing CtSVG configuration, the horseshoe vortices canbe observed between each device pair, as represented by the re-gion G in Fig. 4b, resulting in a flow pattern similar to that of theCtLVG configuration. When the VGs are spaced closer togetherrelative to the vortex size, the interaction with neighboring vortexpairs is significantly enhanced. On the contrary, the vortices gen-erated by the wide-spacing CtSVG configuration appear to moveapart, inducing separated flow comprising horseshoe vortices be-tween each device pair.

3.2 Wake Characteristics. Results of the PIV measurementof the wake are presented here. Figure 5 shows mean streamwisevelocity contours of the four configurations at =1 deg andh /c=0.090, time-averaged over 500 PIV snapshots. The velocityis normalized by the freestream velocity. The results in the vicin-ity of the trailing edge of the wing and moving ground are ex-cluded due to reflections from the solid surfaces. The figures forthe counter-rotating VG configurations Figs. 5b and 5c showprofiles at two spanwise positions z=za and zb. At z=za, thevortices from the counter-rotating VGs induce downwash on thesuction surface, while the upwash, which detaches from the suc-tion surface, is induced at z=zb.

(b)

G

face around centre portion of wing span forh /c=0.090: a close-spacing and b wide-

Transactions of the ASMEE license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

tflrgFwstlCtdtpebcCt

1.2

zc

Counter-rotating Co-rotating

zaU U

xz=

h

J

DownloaFor the baseline clean wing case in Fig. 5a, the wake becomeshicker as it flows downstream, reducing the velocity deficit. Theow between the wake and the moving ground is accelerated as aesult of ground effect, inducing boundary layer growth on theround see Fig. 6. For the counter-rotating VG configurations inig. 5b, the wake profiles at z=za and zb show different features,ith the CtLVG configuration exhibiting more variance in the

panwise direction compared with the CtSVG configuration. Bothhe profiles of the CtSVG configuration at z=za and zb show simi-ar wake thicknesses to the clean wing. The velocity deficit of thetSVG configuration tends to persist farther downstream, while

he wake of the clean wing shows relatively faster decay of theeficit. For the wake profile at z=za, the downwash near the suc-ion surface induced by the VG-generated vortices appears to sup-resses the flow separation accelerating the flow near the trailingdge, compared with the clean wing. For the accelerated flowetween the wing and the ground, both profiles of the CtSVGonfiguration show similar distributions to the clean wing. ThetLVG configuration exhibits a significant effect of the VGs on

he wake structure, the wake profiles at z=za and zb being very

(a)

1.2 1.1

1.1

1.0

1.0

1.0

0.9

0.9

0.8

0.80.70.60.50.40.3

0.1

0

-0.1

y/c

1.11.0 1.2 1.3 1.4 1.5 1.6

1.11.2

0.1

0

-0.1

1.11.0

0.1

0

-0.1

1.11.0 1.2 1.3 1.4 1.5 1.6

1.1

1.00.9

1.2

1.00.9

1.0

0.1

0

-0.1

1.11.0

1.1

0.6

1.2

zb

(b)

y/c

y/c

y/c

x/c

(c)

1.0x/cz=za

Fig. 5 Mean streamwise velocity contours at =1 deg andz=za and zb, and d CoSVG at z=zc

u/U0.5

-0.1

y/c

0.6 0.7 0.8 0.9 1.0 1.

0.1

0

y/c

Fig. 6 Mean streamwise velocity profile

ournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMdifferent, as shown in Fig. 5c. The profile at z=za shows a nar-row wake distribution downstream of the trailing edge as the re-sult of the suppression of the flow separation. Meanwhile, a sig-nificantly thicker wake is shown at z=zb. For the CoSVGconfiguration in Fig. 5d, the wake profile shows a similar distri-bution to that of the clean wing, while the CoSVG configurationgenerates a slightly thicker wake than the clean wing.

Figure 6 shows the mean streamwise velocity profiles at x /c=1.5 of the four configurations at =1 deg and h /c=0.090. Forthe counter-rotating VG configurations, profiles at the two span-wise positions z=za and zb are given. Characteristics of the ve-locity profiles are summarized in Table 1. The thickness of thewake is defined as the distance between the upper and lowerpoints of the wake where the velocity is 99% of the freestream.Adding the CtSVGs to the wing reduces the thickness and in-creases the maximum velocity deficit of the wake compared withthe clean wing. The maximum velocity deficit increases fromumin /U=0.70 y /c=0.048 for the clean wing to umin /U=0.60 y /c=0.055 at z=za and umin /U=0.61 y /c=0.053 at

/c

1.0

1.00.9

0.90.8

0.8

0.70.6

1.3 1.4 1.5 1.6

0.7

1.0

1.1

1.0

1.1

0.90.80.7

0.60.50.40.3

0.1

0

-0.1

1.11.0 1.2 1.3 1.4 1.5 1.6

0.90.8

1.0

1.0

0.1

0

-0.1

1.11.0 1.2 1.3 1.4 1.5 1.6

1.1

1.0

1.00.9

0.90.8

0.8

0.70.6

0.7

1.2

0.50.4

1.0

1.3 1.4 1.5 1.6

1.0

1.00.9

0.90.8

0.8

0.7

0.6

1.0

y/c

y/c

zax/cz=zb

(d)x/cz=zc

/czb

/c=0.090: a clean, b CtSVG at z=za and zb, c CtLVG at

Counter-rotating

za

zb

Uzc

Co-rotating

U

CleanCtSVG(z=za)CtSVG(z=zb)CtLVG(z=za)CtLVG(z=zb)CoSVG(z=zc)

t x /c=1.5 at =1 deg and h /c=0.090

DECEMBER 2009, Vol. 131 / 121103-51

s ax1.2

0.7

z=E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

zfCpdCttbsdzw=

=

t

larger wake compared with the clean wing. The maximum veloc-ity deficit increases from umin /U=0.70 y /c=0.048 for the

Table 1 Wake characteristics at x /c=1.5 at =1 deg andh /c=0.090

u

y

x1

1xz

z

=1

1

Downloa=zb for the CtSVG configuration. The wake thickness is reducedrom wake /c=0.19 for the clean wing to wake /c=0.17 for thetSVG configuration. Of great interest here is that both wakerofiles of the CtSVG configuration show a remarkably similaristribution, thus, less variance in the spanwise direction. For thetLVG configuration, the profiles obviously show different fea-

ures from each other. The variance of the wake structure indicateshat the wake flows are highly affected by the vortices generatedy the CtLVGs. The velocity deficit of the CtLVG at z=za ismaller than the others, and another small velocity deficit can beetected around y /c=0.10. The wake thickness of the CtLVG at=za including the small deficit is similar to that of the cleaning. Meanwhile, a significantly thicker wake is observed at zzb. The maximum velocity deficit changes from umin /U=0.70y /c=0.048 for the clean wing to umin /U=0.86 y /c0.043 at z=za, and umin /U=0.63 y /c=0.037 at z=zb for

he CtLVG configuration. The CoSVG configuration shows a

Clean CtSVG CtLVG CoSVG

z=za z=zb z=za z=zb

wake /c 0.19 0.17 0.17 0.19 0.22 0.20min /U 0.70 0.60 0.61 0.86 0.63 0.62/c at umin 0.048 0.055 0.053 0.043 0.037 0.032

(c)

0.1

0

-0.1

1.0x/c

0.1

0

-0.1

1.0

(b)

y/c

y/c

(a)x/c

0.1

0

-0.1

1.1 1.2 1.31.0

y/c

0.1

0

-0.1

1.1 1.2 1.31.0

y/c

zc

Counter-rotating Co-rotating

za

zb

U U

z=za

-40z

Fig. 7 Instantaneous spanwise vorticity distributions at c CtLVG at z=za and zb, and d CoSVG at z=zc

21103-6 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMclean wing to umin /U=0.62 y /c=0.032 for the CoSVG con-figuration. For the accelerated flow between the wake and ground,the counter-rotating configurations show slightly higher velocity,whereas the CoSVG configuration presents lower velocity, com-pared with the clean wing. The boundary layer growth detected onthe moving ground shows a similar rate in all four profiles.

Figure 7 shows instantaneous spanwise vorticity contours of thefour configurations at =1 deg and h /c=0.090 to illustratesome unsteady flow features behind the wing. While only onesnapshot is shown, the characteristics described here are evidentin the whole set of captured PIV data. The vorticity is normalizedby the freestream velocity and wing chord. The figures for thecounter-rotating VG configurations Figs. 7b and 7c show dis-tributions at two spanwise positions z=za and zb. The vortexstructure behind the clean wing is characterized by the presence ofvortex shedding and a shear layer between regions of positive andnegative vorticity, as shown in Fig. 7a. For both distributions ofthe CtSVG configuration in Fig. 7b, the vortices appear tospread relatively wider and the size of each shed vortex is typi-cally smaller compared with the clean wing case. The differencebetween the profiles at z=za and zb is not obvious. The CtLVGconfiguration shows a significant impact of the CtLVGs on thevorticity structure in Fig. 7c. At z=za, the vortices distributeonly downstream of the trailing edge, and the shear layer is ob-served clearly. The size of each vortex is relatively small. Mean-while, a wide range of the vortex shedding is observed at z=zb.The structure of shed vortices is much more irregular compared

0.1

0

-0.1

x/c1.1 1.2 1.31.0

/c.1 1.2 1.3

0.1

0

-0.1

x/c1.1 1.2 1.31.0

(d)

.1 1.2 1.3/c

y/c

y/c

=za z=zb

z=zc=zb

40

deg and h /c=0.090: a clean, b CtSVG at z=za and zb,

Transactions of the ASME0E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

wFdc

itrtcadvctoo

Bogdptttzzvmsvsvsdt

nd

J

Downloaith the others, and the shear layer cannot be observed clearly.or the co-rotating VG configurations in Fig. 7d, the structure ofiscrete vortices of the CoSVG configuration is similar to thelean wing.

In addition to the instantaneous unsteady features of the vortic-ty, the mean spanwise vorticity contours of the four configura-ions, time-averaged over 500 PIV snapshots, at =1 deg andh /c=0.090, are shown in Fig. 8. The figures for the counter-otating VG configurations Figs. 8b and 8c show distribu-ions at two spanwise positions z=za and zb. The mean vorticityontours clearly explain where the positive and negative vorticesnd the shear layers are generated, and how they diffuse as theyevelop downstream. For the clean wing in Fig. 8a, the positiveorticity is thicker than the negative vorticity at a region ofx /c1.3 due to the flow separation on the suction surface. Bothontours for the CtSVG configuration show a characteristic vor-icity distribution in Fig. 8b. Two positive vorticity regions arebserved: one is along the suction surface of the wing, and thether is detected underneath the suction surface at y /c0.02.etween these positive vorticity regions, the negative vorticity isbserved underneath the suction surface in addition to the oneenerated around the pressure surface. The difference between theistributions at z=za and zb is not obvious; however, at z=za, theositive vorticity region near the trailing edge flows slightly fur-her downstream than that at z=zb. Additionally, the positive vor-icity underneath the suction surface at z=za shows lower degreehan the value of the clean wing and the CtSVG configuration at=zb. These discrepancies of the distributions between z=za andb are due to the suppression of flow separation at z=za as re-ealed by the surface flow visualization. The effect of the VGs isore obvious in the CtLVG configuration in Fig. 8c, as can be

een in other wake surveys. At z=za, thin distributions of theorticity are observed downstream of the trailing edge due to theuppression of the separation. Meanwhile at z=zb, the negativeorticity generated from the pressure surface and underneath theuction surface results in a thicker distribution downstream. Bothistributions of the CtLVG configuration show two positive vor-icity regions, as observed in the CtSVG configurations. One is

Fig. 8 Mean spanwise vorticity distributions at =1 deg aat z=za and zb, and d CoSVG at z=zc

ournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMshown on the suction surface of the wing, and the other under-neath the suction surface at y /c0 z=za and at y /c0.05 z=zb. The value of the positive vorticity at y /c0.05 z=zb is obviously higher than that at y /c0 z=za dueto the flow separation at z=zb. The vorticity distribution of theCoSVG configuration in Fig. 8d is remarkably similar to that ofthe clean wing. The similarity between the clean wing andCoSVG configuration can also be seen in the mean velocity dis-tributions.

4 DiscussionKuya et al. 24 showed that both the counter-rotating configu-

rations can increase downforce compared with the clean wingwhen the wing is operated in the ground effect regime, and inparticular, the use of the CtSVGs has advantages both in thedownforce and efficiency under some conditions, while theCtLVG configuration indicates less efficiency. The CoSVG con-figuration deteriorates the wing performance in all cases. The ex-perimentally investigated characteristics, including surface flowvisualization and wake surveys, explain the physical mechanismof the separation control and the advantage of the CtSVGs com-pared with the others.

The vortices generated by the VGs induce downwash and up-wash to the suction surface, which mixes the outer flow and theflow in the boundary layer. The downwash generated at the centerof each device pair transfers the high momentum of the outer flowinto the boundary layer flow, leading to the suppression of flowseparation, as can be seen in the results of the surface flow visu-alization of the counter-rotating VG configurations. However, theinteraction between the neighboring co-rotating vortices tend tocancel each others downwash and upwash effects, resulting in amore rapid decay of the vortex, and enhances the lateral compo-nent of the flow. Therefore, the co-rotating vortices do not persistfurther downstream, resulting in little effect of separation control.The more rapid decay of the co-rotating vortices compared withthe counter-rotating vortices captured here is in good agreementwith the investigation of Pauley and Eaton 22. Thus, it is con-

h /c=0.090: a clean, b CtSVG at z=za and zb, c CtLVG

DECEMBER 2009, Vol. 131 / 121103-7E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

ceivable that co-rotating VGs with wider device spacings, whichhave less interaction between each vortex, could have a morefavorable effect for separation control.

gvbcttfidtvcClwridw

ssrCpaacscC

ttcwed

ciaVc

5

ifc

tices induced by the close- or reference-spacing CtSVGs isrestricted by the interaction of the vortices existing in theirneighbors.

1

DownloaFor separation control, important factors regarding the VG-enerated vortices are their strength and size. The surface flowisualization reveals that the vortices induced by the CtSVGsreak down at 95% chord, indicating that the vortices are suffi-iently strong, but not excessively so, to make the flow overcomehe adverse pressure gradient region, leading to efficient separa-ion control at the condition tested. Meanwhile, the CtLVG con-guration not only suppresses the flow separation due to theownwash on the suction surface, but also induces horseshoe vor-ices where the flow is separated by the upwash. The wake sur-eys performed by the PIV measurement show that the CtSVGonfiguration possesses a small spanwise variance, while thetLVG configuration exhibits significant spanwise variances. The

arge variance of the CtLVG configuration indicates that vorticesith excessive strength and size have not only favorable effects

egarding separation control but also significant penalties. Accord-ngly, the device height of the VGs is important not only for theevice drag but also for the strength and size of the vortices,hich the VGs produce.Regarding the effect of the device spacing, Kuya et al. 24

howed that the close-spacing CtSVG configuration exhibits aimilar effect on the downforce and efficiency with respect to theeference-spacing CtSVG configuration, while the wide-spacingtSVG configuration shows less downforce and efficiency, com-ared with the other spacing CtSVG configurations. Those char-cteristics are consistent with the results of the surface flow visu-lization investigated here; the close-spacing CtSVGonfiguration shows a similar flow pattern to that of the reference-pacing CtSVG configuration, while the wide-spacing CtSVGonfiguration leads to the horseshoe vortices, as shown in thetLVG configuration.Accordingly, as shown in the investigation of Kuya et al. 24,

he CtSVG configuration exhibits the best performance thanks tohe nearly optimal strength and size of the counter-rotating vorti-es to overcome the adverse pressure gradient region induced hereith a small drag penalty. The CoSVG configuration, however,

xhibits a negligible separation control capability due to the rapidecay of the co-rotating vortices.

Although the device height and spacing of the VGs arehanged, the orientation angle is fixed at 15 deg in the currentnvestigation. The strength of the VG-generated vortices may beltered by adjusting the orientation angle of the vane; optimizedG parameters for the rectangular vane type could be found via

hanging the device height, spacing, and orientation angle.

Concluding RemarksAn experimental investigation of the flow characteristics of an

nverted wing with VGs in ground effect is performed, using sur-ace flow visualization and wake surveys, and the following con-lusions are drawn:

Surface flow visualization of the clean wing captures thecharacteristic horseshoe vortices and flow separation down-stream of 6580% chord. Both the counter-rotating configu-rations suppress the flow separation at the center of eachdevice pair, while the CtLVGs induce the horseshoe vorticesbetween each device where the flow is separated. TheCoSVG configuration shows the flow separation down-stream of the VGs over the wing span at 80% chord.

The close- and wide-spacings dVG=2hVG and 8hVG areexamined in addition to the reference-spacing dVG=4hVGin the CtSVG configuration. The counter-rotating vorticesinduced by the wide-spacing CtSVGs are likely to spreadoutward and induce the horseshoe vortices as the CtLVGconfiguration features; meanwhile, the spreading of the vor-

21103-8 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASM Wake flow surveys obtained by the PIV measurement revealsignificant spanwise variances in the wake behind the wingequipped with the CtLVGs, while the CtSVG configurationshows a relatively small variance in the spanwise direction.The CoSVG configuration shows very similar distributionsto those of the clean wing.

The flow physics investigated here suggests advantages of ause of the CtSVG configuration for the separation control.

AcknowledgmentY. Kuya gratefully acknowledges the financial support of the

Ministry of Education, Culture, Sports, Science, and Technologyof Japan and the School of Engineering Sciences, University ofSouthampton. The authors would like to thank Mr. Mike Tudor-Pole for his assistance with the experiments.

NomenclatureRoman Symbols

c wing chorddVG device spacing of the vortex generatorhVG device height of the vortex generator

h wing ride heightRe Reynolds number =Uc /U freestream velocity

u , v Cartesian components of velocity streamwiseand lateral directions

umin maximum velocity deficitx ,y ,z Cartesian tensor system streamwise, lateral,

and spanwise directionsza ,zb ,zc PIV laser sheet positions

Greek Symbols wing incidence

wake wake thickness dynamic viscosity

densityz nondimensional spanwise vorticity =v /x

u /yc /U

References1 Zhang, X., Toet, W., and Zerihan, J., 2006, Ground Effect Aerodynamics of

Race Cars, Appl. Mech. Rev., 59, pp. 3349.2 Katz, J., 2006, Aerodynamics of Race Cars, Annu. Rev. Fluid Mech., 38,

pp. 2763.3 Knowles, K., Donoghue, D. T., and Finnis, M. V., 1994, A Study of Wings in

Ground Effect, Loughborough University Conference on Vehicle Aerodynam-ics, Vol. 22, pp. 113.

4 Zerihan, J., and Zhang, X., 2000, Aerodynamics of a Single Element Wing inGround Effect, J. Aircr., 376, pp. 10581064.

5 Zhang, X., and Zerihan, J., 2003, Off-Surface Aerodynamic Measurements ofa Wing in Ground Effect, J. Aircr., 404, pp. 716725.

6 Ranzenbach, R., Barlow, J. B., and Diaz, R. H., 1997, Multi-Element Airfoilin Ground EffectAn Experimental and Computational Study, AIAA PaperNo. 1997-2238.

7 Zhang, X., and Zerihan, J., 2003, Aerodynamics of a Double-Element Wingin Ground Effect, AIAA J., 416, pp. 10071016.

8 Zhang, X., and Zerihan, J., 2004, Edge Vortices of a Double-Element Wing inGround Effect, J. Aircr., 415, pp. 11271137.

9 Katz, J., and Largman, R., 1989, Effect of 90 Degree Flap on the Aerody-namics of a Two-Element Airfoil, ASME J. Fluids Eng., 111, pp. 9394.

10 Zerihan, J., and Zhang, X., 2001, Aerodynamics of Gurney Flaps on a Wingin Ground Effect, AIAA J., 395, pp. 772780.

11 Soso, M. D., and Wilson, P. A., 2006, Aerodynamics of a Wing in GroundEffect in Generic Racing Car Wake Flows, Proc. Inst. Mech. Eng., D J.Automob. Eng., 2201, pp. 113.

12 Soso, M. D., and Wilson, P. A., 2008, The Influence of an Upstream Diffuseron a Downstream Wing in Ground Effect, Proc. Inst. Mech. Eng., Part D J.Automob. Eng., 2224, pp. 551563.

13 Diasinos, S., and Gatto, A., 2008, Experimental Investigation Into Wing Spanand Angle-of-Attack Effects on Sub-Scale Race Car Wing/Wheel Interaction

Transactions of the ASMEE license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Aerodynamics, Exp. Fluids, 453, pp. 537546.14 Coe, D., Chipperfield, A., and Williams, C., 2006, Transient Wing in Ground

Effect Aerodynamics: Comparisons of Static and Dynamic Testing, SixthMIRA International Vehicle Aerodynamics Conference, pp. 404410.

15 Gad-el-Hak, M., 1990, Control of Low-Speed Airfoil Aerodynamics, AIAAJ., 289, pp. 15371552.

16 Gad-el-Hak, M., and Bushnell, D. M., 1991, Separation Control: Review,ASME J. Fluids Eng., 113, pp. 530.

17 Lin, J. C., Howard, F. G., and Bushnell, D. M., 1990, Investigation of SeveralPassive and Active Methods for Turbulent Flow Separation Control, AIAAPaper No. 1990-1598.

18 Lin, J. C., Selby, G. V., and Howard, F. G., 1991, Exploratory Study ofVortex-Generating Devices for Turbulent Flow Separation Control, AIAAPaper No. 1991-0042.

19 Lin, J. C., 1999, Control of Turbulent Boundary-Layer Separation UsingMicro-Vortex Generators, AIAA Paper No. 1999-3404.

20 Lin, J. C., 2002, Review of Research on Low-Profile Vortex Generators toControl Boundary-Layer Separation, Prog. Aerosp. Sci., 3845, pp. 389420.

21 Lin, J. C., Howard, F. G., and Selby, G. V., 1990, Small Submerged VortexGenerators for Turbulent Flow Separation Control, J. Spacecr. Rockets,275, pp. 503507.

22 Pauley, W. R., and Eaton, J. K., 1988, Experimental Study of the Develop-ment of Longitudinal Vortex Pairs Embedded in a Turbulent Boundary Layer,

AIAA J., 267, pp. 816823.23 Godard, G., and Stanislas, M., 2006, Control of a Decelerating Boundary

Layer. Part 1: Optimization of Passive Vortex Generators, Aerosp. Sci. Tech-nol., 103, pp. 181191.

24 Kuya, Y., Takeda, K., Zhang, X., Beeton S., and Pandaleon, T., 2009, FlowSeparation Control on a Race Car Wing With Vortex Generators in GroundEffect, ASME J. Fluids Eng., 131, p. 121102.

25 Senior, A. E., and Zhang, X., 2001, The Force and Pressure of a Diffuser-Equipped Bluff Body in Ground Effect, ASME J. Fluids Eng., 123, pp.105111.

26 Ruhrmann, A., and Zhang, X., 2003, Influence of Diffuser Angle on a BluffBody in Ground Effect, ASME J. Fluids Eng., 125, pp. 332338.

27 Zhang, X., Senior, A., and Ruhrmann, A., 2004, Vortices Behind a BluffBody With an Upswept Aft Section in Ground Effect, Int. J. Heat Fluid Flow,25, pp. 19.

28 Burgin, K., Adey, P. C., and Beatham, J. P., 1986, Wind Tunnel Tests on RoadVehicle Models Using a Moving Belt Simulation of Ground Effect, J. WindEng. Ind. Aerodyn., 22, pp. 227236.

29 2000, FlowMap Particle Image Velocimetry Instrumentation, 5th ed., DantecDynamics, Skovlunde, Denmark.

30 Abernethy, R. B., Benedict, R. P., and Dowdell, R. B., 1985, ASME Mea-surement Uncertainty, ASME J. Fluids Eng., 107, pp. 161164.

31 Bell, S., 1999, A Beginners Guide to Uncertainty of Measurment, Measure-ment Good Practice Guide, 112, pp. 133.

J

Downloaournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMDECEMBER 2009, Vol. 131 / 121103-9E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm