protrotigs olarof

vor3%leaph26

owtexratota

vevely

1

idancogciverted wing, modeling a race car front wing, in ground effect viavortex generator VG attachment. VGs are known to be capableof controlling flow separation in adverse pressure gradient flows

waaItlgtddlrteFw

ratio. In general, flow separation is generated downstream ofmaximum suction point, and even though the effect of the sepa-ration on lift or downforce of a wing is not significant, a thicker

JsK

J

Downloa3,4.A number of experimentally investigated studies for the front

ing have been previously carried out 511. Typical features ofsingle-element wing in ground effect are described by Zerihan

nd Zhang 5 and Zhang and Zerihan 6 using a moving ground.n wind tunnel testing, use of a moving ground rig, which runs athe same speed as the freestream in order to control the boundaryayer on the ground, is necessary for race car aerodynamics inround effect to properly simulate the real conditions on a racerack. Zerihan and Zhang 5 showed not only an increase ofownforce as the ride height is reduced but also captured theownforce reduction phenomenon, when the wing is mounted be-ow the height where the maximum downforce is generated. Theesults of surface flow visualization revealed the downforce reduc-ion phenomenon by showing a breakdown of edge vortices at thend plates and flow separation on the suction surface of the wing.low separation is induced by the large adverse pressure gradient,here the streamwise momentum of the flow is reduced, and the

turbulent boundary layer, yielded by the separation, induces a sig-nificant increase in drag, thus, adversely affects the performanceof the wing 12.

A number of separation control methods exists. Lin 12,13suggested that effective devices for separation control are thosethat generate streamwise vortices such as those produced by VGs.The VGs commonly used for aeronautical applications are at-tached upstream of the separation line and generate streamwisevortices, which accelerate mixing of streamwise momentum of thefreestream and the flow in a boundary layer in order to overcomethe large adverse pressure gradient. Sub-boundary layer vortexgenerators SVGs have a device height between 10% and 50% ofthe boundary layer thickness according to the classification of Lin12, and are more advantageous in terms of effectiveness and lessdrag penalty compared with large-scale vortex generators LVGs,whose device height is of the same order as the boundary layerthickness 1214. Several experimental investigations of VGshave been performed to improve wing performance in freestream13,1517. Garcia and Katz 18 and Katz and Morey 19 inves-tigated LVGs attached to a flat plate in ground effect, simulatingan application of VGs on lower surface of race cars. There is,however, a lack of investigation of an inverted wing with VGs forseparation control in ground effect. There have been computa-tional studies regarding VG applications; a flat plate 20,21,

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 / 121102-1Copyright 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 SO17 1BJ, UK

Flow SeRace CaGeneratFlow separation conexperimentally invespressure distributionco-rotating rectanguthe wing. The effectsub-boundary layeron the wing deliver 2compared with the cdownforce reductionerators exhibit up tonamic efficiency at lcounter-rotating vorrotating vortex genevortex generators, ntype, can be effectidownforce for relati

IntroductionIn open-wheel racing series, such as Formula 1 and Indy Rac-

ng, both a front wing and rear wing are inverted to produceownforce, that is, negative lift, leading to an enhancement ofcceleration and cornering performance of the cars. The aerody-amic performance plays a significant role in open-wheel racears and is one of the most important factors to be developed1,2. The performance of aerodynamic devices is altered whenperating in close proximity to a solid boundary, known as theround effect regime, and different flow features are exhibitedompared with the freestream condition. This study attempts tomprove aerodynamic performance of a generic single-element in-ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMaration Control on aWing With Vortexrs in Ground Effectl using vortex generators on an inverted wing in ground effect isated, and its performance is characterized in terms of forces andver a range of incidence and ride height. Counter-rotating and-vane type vortex generators are tested on the suction surface ofdevice height and spacing is investigated. The counter-rotating

tex generators and counter-rotating large-scale vortex generatorsand 10% improvements in the maximum downforce, respectively,

n wing, at an incidence of one degree, and delay the onset of theenomenon. The counter-rotating sub-boundary layer vortex gen-% improvement in downforce and 10% improvement in aerody-

ride heights. Chordwise pressure measurement confirms that bothgenerator configurations suppress flow separation, while the co-

ors exhibit negligible effectiveness. This work shows that a use ofbly of the counter-rotating sub-boundary layer vortex generatorat controlling flow separation, with a resultant improvement in

low drag penalty. DOI: 10.1115/1.4000420

flow separates from the wall 3,4. At a moderate ride height, flowseparation was observed near the trailing edge of the wing, and asthe wing is brought close to the ground through the ride height ofthe maximum downforce, the region of the flow separation in-creases, resulting in the loss of the downforce. Zhang and Zerihan6 studied characteristics of the edge vortices with respect to theride height. It was shown that the edge vortices break down whenthe wing reaches maximum downforce height. The investigationssuggest that separation control on the suction surface of a wingand a development of end plates could improve the aerodynamicperformance of the wing, controlling the flow and edge vortices,leading to higher downforce or more efficient downforce-to-dragE license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

wVs

pttbacicu

2

fvnaimTsfswpa4

suTpc

223.4 mm, and finite trailing edge of 1.65 mm and is the samemodel used by Zerihan and Zhang 5,9 and Zhang and Zerihan6. The origin of the coordinate system is set at the leading edge

0.4Freestream U

g: to

1

Downloaings 22, and source term models 2224. An application ofGs to an inverted wing in ground effect was computationally

tudied by Kuya et al. 25.This paper investigates contribution of VGs on aerodynamic

erformance of an inverted wing in ground effect, describingime-averaged force and pressure characteristics. The experimen-al testing is performed in a wind tunnel equipped with a movingelt rig. Rectangular vane type SVGs and LVGs are used andttached on the suction surface of the wing in counter-rotating ando-rotating configurations. Different device spacings are also stud-ed with the SVGs. Kuya et al. 26 described the detailed flowharacteristics of an inverted wing with VGs in ground effect,sing surface flow visualization and particle image velocimetry.

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, including Zerihannd Zhang 5,9, Zhang and Zerihan 68, Senior and Zhang27, Ruhrmann and Zhang 28, and Zhang et al. 29. The tunnels of conventional return circuit design and is equipped with a

oving belt rig and a three-component overhead balance system.he 3.2m1.5m moving belt is controlled by slots and suctionystem for boundary layer removal, which gives 99.8% of thereestream velocity at 2 mm above the belt. The turbulence inten-ity of the freestream is about 0.3%. Further descriptions of theind tunnel are given by Burgin et al. 30. For the experimentsresented here, the freestream velocity U and moving belt speedre set at 30 m/s, corresponding to Reynolds number Re of50,000 based on the wing chord 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 GAWrofile, type LS1-0413, and is manufactured by carbon fiberomposite. The model has a span of 1100 mm, constant chord c of

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 Configuration of VGs on winrotating VGs. Flow is from bottom to

21102-2 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMof the wing. Generic end plates of 250mm100mm4mm areattached on both ends of the wing throughout the experiment. Theincidence is measured relative to a line from the trailing edge tothe most swelled point on the pressure surface which correspondsto 2.6 deg relative to the chord line. The true incidence is there-fore equal to the measured incidence plus 2.6 deg. When the in-cidence is 1 deg, corresponding to the true incidence of 3.6 deg,the upper and lower edges of the end plates are parallel to theground. The incidence is varied by rotating about the quarterchord position. The ride height h is defined by the distance fromthe lowest point on the suction surface of the wing to the movingground as the incidence is fixed at 1 deg.

Rectangular vane type VGs are employed here comprising threeconfigurations, which are counter-rotating sub-boundary layervortex generators CtSVGs, counter-rotating large-scale vortexgenerators CtLVGs, and co-rotating sub-boundary layer vortexgenerators CoSVGs. The SVG and LVG have a device height of2 mm hVG /c=0.009 and 6 mm hVG /c=0.027, respectively.The VGs are made of aluminum plate with 0.6 mm thickness andbuilt in pairs separated by 4hVG at the trailing edge of the VGs.Although carbon fiber may be used in race car applications, ef-fects due to the difference of VG material are deemed negligible.The vanes are oriented at 15 deg relative to the streamwisedirection, comprising the counter-rotating or co-rotating VG con-figuration. Pairs of VGs are put side by side along the span of thewing, as shown in Fig. 2. The VGs are attached on the suctionsurface of the wing such that the trailing edge of the VGs is fixedat x /c=0.537. The height to length ratio of the vanes is fixed at1:4. For the CtLVG and CoSVG configurations, the device spac-ing dVG between each device pair of the VGs is fixed at 4hVG,while close- and wide-spacings of 2hVG and 8hVG are also exam-ined with the CtSVG configuration in addition to the reference-spacing of 4hVG. Unless there is a particular notation, CtSVGrepresents a CtSVG configuration with the reference-spacing of4hVG, which is the same device spacing as the other VG configu-rations.

2.3 Experimental Methods. The time-averaged downforceand drag are measured by a three-component overhead balancesystem. The coefficients of downforce and drag are calculated byusing the corrected force values, taking into account variances ofair density because of changes in the freestream temperature andpressure during the tests. The CL quoted in the investigation rep-resents the downforce coefficient. The ride height of the wing isaltered between 0.045c and 0.448c. The incidence is swept from1 deg to 17 deg.

A pressure tapped wing model is employed for the surface pres-sure measurement on the wing in the chordwise direction. Thetapped wing has the same geometry as the wing used in the forcemeasurement. The pressure tapped array has an orientational angleof 22.5 deg regarding the streamwise direction. Twenty-five and23 pressure taps are, respectively, mounted on the suction surfaceand pressure surface around the center portion of the wing. An

(b)

4hVG

zc

4hVG

15

x/c=0.537

dVG

a counter-rotating VGs and b co-p.

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

atnTs

umilattscFimtdt0pac

3

ah

arihcrhfic

scdfibitib

CleanCtSVG

0.14

(

CleanCtSVG

e h

J

Downloaverage in the spanwise direction is taken for the measurement ofhe VG configurations since the flow of the VG configurations isot two-dimensional and is described in detail in the Appendix.he ride height and incidence are fixed at 0.090c and 1 deg, re-pectively.

2.4 Repeatability and Uncertainty. The repeatability andncertainties are estimated for the force and pressure measure-ents using procedures described in Refs. 31,32. The repeatabil-

ty of the measurements is categorized into short, medium, andong-term repeatability. In the short term repeatability test, datare continuously taken during a single wind tunnel run withouturning the wind tunnel off. The medium-term repeatability isested by starting from scratch settings of experimental parametersuch as the incidence and ride height of the wing. The test isonducted with several wind tunnel runs over a couple of hours.or the long-term repeatability test, a set of data is taken in an

nterval of more than 1 month. The standard error for the short,edium, and long terms is given as 0.0016, 0.0018, and 0.043 for

he downforce coefficient, 0.00019, 0.00028, and 0.0064 for therag coefficient, respectively, while for the pressure coefficient,he standard error for each term is, respectively, given as 0.035,.034, and 0.067. The uncertainty of the downforce, drag, andressure coefficients for the medium-term are, respectively, givens 0.0035, 0.0006, and 0.068 with the 95% confidence, where theoverage factor of 2 is used.

Results

3.1 Force Characteristics. Figure 3 shows the downforcend drag coefficients of the four different configurations examinedere at various ride heights at =1 deg.All the configurations show a downforce enhancement regime

nd reach the maximum downforce, followed by a downforceeduction regime as the wing is moved closer to the ground. Ofnterest here is that both counter-rotating VG configurations ex-ibit higher downforce below h /c=0.134 compared with thelean wing, meanwhile the CoSVG configuration exhibits a dete-ioration of performance at almost all ride heights. For rideeights above h /c=0.134, both the counter-rotating VG con-gurations show very similar downforce to that yielded by thelean wing. At h /c=0.134, Zerihan and Zhang 5 discussedmall regions of flow separation on the suction surface of thelean wing at 95% chord that appears to grow as the ride height isecreased. The higher downforce of the counter-rotating VG con-gurations could be due to the suppression of the flow separationelow h /c=0.134 26. Further evidence for this are presentedn the pressure distributions discussed as follows. In addition, al-hough it may be inferred that the downforce increase is due to thencreased suction by the VG-generated vortices, the evidence thatoth the counter-rotating VG configurations produce very similar

0.1 0.2 0.3 0.4 0.50

CL

1.5

2.0

1.0

CtLVGCoSVG

(a) h/c0

Fig. 3 Force characteristics at various riddrag

ournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMdownforce to the clean wing for ride heights above h /c=0.134indicate that the effect of the increased suction due to the VG-generated vortices is not significant; the downforce enhancementdue to the VGs attachment could be mainly led by the suppressionof flow separation rather than the increased suction due to theVG-generated vortices. The use of CtSVGs and CtLVGs on thewing delivers 23% and 10% improvements in the maximumdownforce, respectively. Additionally, the counter-rotating VGconfigurations show a delay of the downforce reduction phenom-enon; while, for the clean wing, the downforce reaches a maxi-mum at h /c=0.090 and drops below this ride height, both thecounter-rotating VG configurations reach the maximum value ath /c=0.067 and the downforce reduction phenomenon are ob-served below this height. The downforce of the counter-rotatingVG configurations show almost the same values above h /c=0.134, and the CtSVG configuration generates higher downforcethan the CtLVG configuration below h /c=0.134. The downforceof the CoSVG configuration reaches the maximum values ath /c=0.090, the same as the clean wing but with lowermagnitude.

The drag indicates an increase of value as the wing is movedcloser to the ground. The gradient of the curves becomes steeperas the wing is mounted below h /c=0.134. For the clean wing,the increase of drag is due to an enhancement of the pressure drag;the streamwise component of the force generated by the pressureon the wing. As the wing moves closer to the ground the forceenhancement phenomenon and flow separation on the suction sur-face lead to enhancement of the pressure drag. All of the VGconfigurations generate higher drag than the clean wing at all theride heights. For the CoSVG configuration, since the CoSVGsexhibits a deterioration in performance of the wing as shown inthe downforce, the increase in the pressure drag should be rela-tively small. Therefore, the contribution of the drag increase in theCoSVG configuration is led by the device drag, which is com-posed of the pressure drag and skin friction drag of the VGs. TheCtLVG configuration generates higher total drag than the CtSVGconfiguration. By the comparison between the CtSVG and CtLVGconfigurations, it can be deduced that the contribution of the de-vice drag of the counter-rotating VGs to the total drag enhance-ment is higher than that of the pressure drag. The CtSVG configu-ration generates a higher pressure difference between the pressureand suction surfaces than the CtLVG configuration, and hence thepressure drag of the CtSVG configuration should be higher thanthat of the CtLVG configuration. The total drag is, however,higher for the CtLVG configuration compared with the CtSVGconfiguration at all the ride heights examined here.

Figure 4 shows contour maps of downforce with respect to theincidence and ride height of the four configurations. The down-force contours show that both the counter-rotating VG configura-

0.1 0.2 0.3 0.4 0.50

0.12

0.10

0.08

0.06

0.04

CD

b)

CtLVGCoSVG

h/c

eights at =1 deg: a downforce and b

DECEMBER 2009, Vol. 131 / 121102-3E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

ttdpc

=

CbtltTiVm

drBid2

CL1.1 2.4

pe

Tw

1

Downloaions produce higher downforce, showing steep gradient of con-our lines at low incidences, and reaching the maximumownforce region distributing relatively lower incidences com-ared with the clean wing. The CtSVG configuration also indi-ates another high downforce region around h /c=0.090 and 3 deg addition to the region around the center of the plot. TheoSVG configuration shows the similar degree of the downforceut a smaller high downforce region than clean wing, indicatinghe CoSVG configuration deteriorates the performance. Table 1ists increases in the maximum downforce via the attachment ofhe VGs against the clean wing at h /c=0.090, 0.179, and 0.269.he counter-rotating VG configurations apparently exhibit better

mprovements in the maximum downforce than the co-rotatingG configurations. The CoSVG configuration reduces the maxi-um downforce at all the three ride heights.For the counter-rotating configurations which exhibit higher

ownforce than the clean wing, downforce and downforce-to-dragatio improvements relative to the clean wing are shown in Fig. 5.oth the counter-rotating configurations show an improvement

n the downforce. The CtSVG configuration shows a higherownforce region at 112 deg at all the ride heights, and6% maximum improvement is obtained at =3 deg and h /c

(a)

2.0

1.8

1.9

1.7

1.6

1.5

1.7

1.6

1.5

1.4

1.3

0.20

0.25

0.15

0.10

(c)

0.20

0.25

0.15

0.10

2.0

1.81.9

1.7

1.6

1.5

1.4

1.3

1.9

1.8

1.7

1.6

2.1

2.2

2.3

1.6

1.5

2.0

h/c

h/c

105 15(deg)

105 15(deg)

Fig. 4 Downforce characteristics with resb CtSVG, c CtLVG, and d CoSVG

able 1 Increase in maximum downforce relative to cleaning %

h /c CtSVG CtLVG CoSVG

0.090 15.0 8.7 1.60.179 9.6 15.0 0.10.269 11.5 11.5 1.9

21102-4 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASM=0.090. Meanwhile, the CtLVG configuration indicates a widerregion of improvement at 114 deg than the CtSVG con-figuration, and 16% maximum improvement is achieved at =3 deg and h /c=0.090, and at =7 deg and h /c=0.179. Ofnote here is that the CtSVG configuration exhibits regions ofhigher efficiency with respect to the clean wing with improvementof the downforce, obtaining 10% maximum improvement at =7 deg and h /c=0.269, while the CtLVG configuration showsentirely lower efficiency. The effective region of the CtSVG con-figuration is observed at 26 deg when the ride height islow and at 58 deg when the ride height is high. This is asthe strength of adverse pressure gradient on the suction surface issimilar between the conditions at low ride height with low inci-dence and at high ride height with relatively high incidence. Ingeneral it is known that adverse pressure gradient, which leads toflow separation, is induced when a wing is operated close to aground or at high incidence. This suggests that a use of the CtS-VGs has advantages both in the downforce and efficiency againstcertain level of the adverse pressure gradient.

The close- and wide-spacings dVG=2hVG and 8hVG are exam-ined in addition to the reference-spacing dVG=4hVG in theCtSVG configuration. Figure 6 shows the effect of the devicespacing on improvements of downforce and downforce-to-dragrelative to the clean wing. The close-spacing CtSVG configurationshows a similar distribution of the improved region of the down-force to the reference-spacing CtSVG configuration; however,higher improvement is observed in most of region. A 24% maxi-mum improvement is obtained at =3 deg and h /c=0.090,which is slightly less than that of the reference-spacing CtSVGconfiguration. The distribution of the efficiency is rather differentbetween the close- and reference-spacing CtSVG configuration;

(d)

0.20

0.25

0.15

0.10

2.0

1.7

1.6

1.5

1.4

1.3

1.9

1.81.7

1.6

(b)

0.20

0.25

0.15

0.10

2.0

1.81.9

1.7

1.6

1.51.4

1.3

1.9

1.8

1.7

1.6

2.1

2.2

2.0

1.5

2.22.3

h/c

h/c

105 15(deg)

105 15(deg)

ct to incidence and ride height: a clean,

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

hiwdpr

Rwaa

mtpiamtt

tViaattsfi

CL(%)-26 260

L/D(%)-36 360

J

Downloaowever, the effective region is similar with a 10% maximummprovement at =7 deg and h /c=0.269. Meanwhile, theide-spacing CtSVG configuration shows less improvement inownforce and similar or less efficiency compared with the im-rovements of the close- and reference-spacing CtSVG configu-ations.

3.2 Averaged-Chordwise Surface Pressure Distributions.esults of the pressure measurement with the pressure tappeding are presented here. Figure 7 shows distributions of the

veraged-chordwise surface pressure on the wing at =1 deg,nd h /c=0.090 corresponding to the clean wing reaches theaximum downforce at =1 deg Fig. 3a. The vertical lines in

he figure represent the leading and trailing edges of the VGs. Theressure distributions of the VG configurations takes the averagen the spanwise direction, since the flows are not two-dimensional,s revealed by surface flow visualization 26. While the currentethod does not represent the pure chordwise pressure distribu-

ions, it provides a representative measurement to determine therend characteristics.

All the configurations produce apparently similar pressure dis-ributions on the pressure surface, while both the counter-rotatingG configurations show slightly higher pressure around the trail-

ng edge. Evidence of a short reattachment bubble can be seen atbout 30% chord on the suction surface in all of the configurationsnd is captured by the surface flow visualization in Ref. 26. Forhe counter-rotating VG configurations, the VGs enhance the suc-ion compared with the clean wing. The CtSVG configurationhows higher suction on the suction surface than the CtLVG con-guration, leading to the higher value of downforce, as shown in

(a)

(b)

CL

0.20

0.25

0.15

0.10

0.20

0.25

0.15

0.10

CL

-2

2

10

10

8 64 0

-4

-6

12

14

12

864

20

-8

-12

-1414

-2

2

10

10

8 6

4

0

-4

-6

12

1412

86

42

0

1618202224

h/c

h/c

105 15(deg)

105 15(deg)

Fig. 5 Effect of counter-rotating VGs onprovements relative to clean wing: a CtS

ournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMthe force characteristics. A similar effect can be seen when a Gur-ney flap is fitted at the trailing edge of a wing, which can enhancethe suction over the entire suction surface. The Gurney flap, how-ever, also increases pressure on the pressure surface because thepresence of the flap at the trailing edge decelerates the velocity onthe pressure surface 9. The CoSVG configuration shows verysimilar distribution to that of the clean wing and produces lesssuction at some points than the clean wing. The gradient of thepressure on the suction surface of the clean wing suddenlychanges at about 75% chord, showing a constant value region nearthe trailing edge. Both the counter-rotating VG configurations,however, show the pressure recovery at an almost constant gradi-ent toward the trailing edge and eliminate the constant value re-gion, indicating that flow separation is suppressed by the VGs.The pressure recovery of the CoSVG configuration indicates asignificantly small or negligible effect of the CoSVGs regardingthe separation control.

4 DiscussionThe experimental result of the force characteristics at

=1 deg suggests that both the counter-rotating configurations canincrease downforce compared with the clean wing when the wingis operated in the ground effect regime. Meanwhile, the VGs alsoincrease both the pressure and device drag. The results of down-force and downforce-to-drag ratio improvements relative to theclean wing suggest that a use of the CtSVGs has advantages bothin downforce and efficiency under some conditions, but theCtLVG configuration indicates less efficiency than the clean wingand CtSVG configuration. For race car applications, tracks require

L/D

0.20

0.25

0.15

0.10

0.20

0.25

0.15

0.10

L/D

8

4

0

-4

-8

-12 -4

0

4

-4-8

-16

-20

-32-28-24

-8

-12

-36

-4

-4

-8

-28

-12

-16

-20

-24

-32

-8

h/c

h/c

105 15(deg)

105 15(deg)

wnforce and downforce-to-drag ratio im-and b CtLVG

DECEMBER 2009, Vol. 131 / 121102-5doVGE license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

duwccpw

iC

0CL(%)

-26 260L/D(%)

-42 42

Fw

1

Downloaifferent levels of downforce and downforce-to-drag ratio. These of the VGs may have more favorable effects in a scenariohere high downforce is demanded. In this research, the CtSVG

onfiguration shows best performance in terms of the separationontrol due to the high downforce increase and relatively low dragenalty, meanwhile the CoSVG configuration deteriorates theing performance.The close- and wide-spacings dVG=2hVG and 8hVG are exam-

ned in addition to the reference-spacing dVG=4hVG in thetSVG configuration. The close-spacing CtSVG configuration

(a)

(b)

CL

CL

0.20

0.25

0.15

0.10

0.20

0.25

0.15

0.10

-2

2

10

10

8

6

40

-4

12

14

12

8

64

2

1618

2022

14

0

-2

2

10

8

6

4

-4

12

14

2

0

0

0

h/c

h/c

105 15(deg)

105 15(deg)

Fig. 6 Effect of VG device spacing on dowments relative to clean wing: a close-spa

0.2 0.4x/c

0 0.6 0.8 1

CP

1

0

-2

-3

-4

-5

-6

-1

CleanCtSVGCtLVGCoSVG

ig. 7 Averaged-chordwise surface pressure distributions oning at =1 deg and h /c=0.090

21102-6 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMshows similar level of improvement to the reference-spacingCtSVG configuration in downforce and efficiency relative to theclean wing. Meanwhile, the wide-spacing CtSVG configurationshows less improvement in downforce and similar or less effi-ciency compared with that of the close- and reference-spacingCtSVG configurations.

The suppression of flow separation due to the counter-rotatingVGs can be seen in the chordwise pressure distributions. Whenflow separation has occurred, the pressure distribution plateaus, asthe clean wing shows in this investigation. Meanwhile, both thecounter-rotating VG configurations show moderate pressure re-coveries, indicating a reduced or an absence of flow separation.The CoSVG configuration exhibits a lesser effect of the vorticeson separation control, exhibiting a similar pressure recovery trendto that of the clean wing.

The detailed flow characteristics studied by Kuya et al. 26explains the above features. The vortices generated by the VGsinduce the downwash and upwash to the suction surface, whichmixes the outer flow and boundary layer flow. The downwashgenerated at the center of each device pair pumps the high mo-mentum of the outer flow into the boundary layer flow leading tothe suppression of flow separation. The co-rotating vortices, how-ever, cancel the downwash and upwash having opposite flow di-rections to each other at the center between each vane, leading toa more rapid decay of the vortices and thus the negligibly smalleffect on the separation control.

A further possible advantage of the application of VGs to a racecar front wing is that VGs could eliminate hysteresis of the down-force curve with varying ride height. Storms and Ross 17 ob-served an elimination of hysteresis of lift via wishbone type VGs

L/D

L/D

0.20

0.25

0.15

0.10

0.20

0.25

0.15

0.10-4

-12

0

-4-8

-16

-20

-32

-28

-24

0

8

4

-4

-8

-12

0-4

-8

-16

-32-28

-24

0

h/c

h/c

105 15(deg)

105 15(deg)

rce and downforce-to-drag ratio improve-g CtSVG and b wide-spacing CtSVG

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

performed in a freestream condition. On a race track, a race car isin an unsteady flow condition and the distance between the wingand ground continuously varies. VGs may also be beneficial incTtTsaoltotscpr

5

vis

Symmetric linePressure tapped arrayMirrored pressure tapped array

J

Downloaases where active incidence changes are initiated on a wing.hese dynamic changes make the hysteresis in force and some-

imes make the wing impossible to recover flow condition 11.he VGs, therefore, could help the wing to produce much moretable characteristics. An application of VGs with a Gurney flap islso of great interest. Zerihan and Zhang 9 showed earlier stallf a Gurney flap fitted wing compared with a clean wing due to itsarger adverse pressure gradient on the suction surface. In addi-ion, although an inverted single-element wing is employed andperated in ground effect, in the current investigation, an applica-ion of VGs to a multi-element wing is also of great interest andhould have some advantages, as indicated in freestream tests13,15,17. For the global aerodynamic characteristics of a racear, a front wing with VGs could potentially help to improve theerformance of components downstream such as a diffuser andadiator.

Concluding RemarksContribution of VGs on aerodynamic performance of an in-

erted wing in ground effect is experimentally examined regard-ng force and pressure characteristics, and the following conclu-ions are drawn.

VGs have an effect on flow separation control in an adversepressure gradient, particularly at low incidence 5 deg and low ride height h /c0.12, and medium in-cidence 510 deg and medium ride height 0.15h /c0.20 examined here, leading to the improvement ofthe downforce.

The attachment of CtSVGs and CtLVGs on the wing respec-tively delivers 23% and 10% improvements in the maxi-mum downforce relative to the clean wing at =1 deg. TheCtSVG configuration shows best performance in terms ofthe separation control due to the its high downforce increaseand relatively low drag penalty, meanwhile the CoSVG con-figuration deteriorates the wing aerodynamic performance.

The results of downforce and downforce-to-drag ratio im-provements relative to the clean wing suggest that a use ofthe CtSVGs has advantages both in the downforce and effi-ciency under several conditions, indicating a 26% and a10% maximum improvement of the downforce and effi-ciency. Meanwhile, the CtLVG configuration indicates lessefficiency than the clean wing and CtSVG configuration.

The close- and wide-spacings dVG=2hVG and 8hVG areexamined in addition to the reference-spacing dVG=4hVGin the CtSVG configuration. The close-spacing CtSVG con-figuration shows a similar level of improvement of down-force and efficiency relative to the clean wing to thereference-spacing CtSVG configuration. Meanwhile, thewide-spacing CtSVG configuration shows less improvementin downforce, and similar or less efficiency compared withthat of the close- and reference-spacing CtSVG configura-tions.

The chordwise pressure distribution features flow separationon the suction surface of the clean wing where the pressurecurve shows a plateau. The counter-rotating VG configura-tions show a moderate slope of the pressure recovery towardthe trailing edge and eliminate the constant value region,indicating no or less separation. Meanwhile, the CoSVGconfiguration exhibits a similar pressure recovery trend tothat of the clean wing.

VGs may help to reduce hysteresis effect due to flow sepa-ration. Combinational application of VGs and a Gurney flap,or a multi-element wing may also improve performance.

ournal of Fluids Engineeringded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMAcknowledgmentY.Kuya gratefully acknowledge 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 through the experiments.

NomenclatureRoman Symbols

c wing chordCD drag coefficient =2D /U

2 SCL downforce coefficient =2L /U

2 SCP pressure coefficient =2p p /U

2 D drag

dVG device spacing of vortex generatorhVG device height of vortex generator

h wing ride heightL downforcep pressure

p freestream static pressureRe Reynolds number =Uc /

S wing planform areaU freestream velocity

x ,y ,z Cartesian tensor system; streamwise, lateral,and spanwise directions

Greek Symbols wing incidence dynamic viscosity density

Appendix: Surface Pressure MeasurementAn average in the spanwise direction is taken for the measure-

ment of the VG configurations since the flow of the VG configu-rations is not two-dimensional. Additionally, the pressure tappedarray orients at 22.5 deg with respect to the streamwise directionfor ease of manufacture and to work more pressure taps on thesurface. Figures 8 and 9 respectively, illustrate the feature of thepressure measurement undertaken in the investigation for counter-rotating and co-rotating VG configurations.

For the counter-rotating VG configurations, the pressure tappedarray passes through center of each device pair or between eachdevice pair, as shown in Figs. 8a and 8b, respectively. Figure 8also illustrates the symmetric lines between each vane and a mir-rored pressure tapped array, which is a mirror image of the pres-sure tapped array with respect to the symmetric lines. When aconsideration regarding the mirrored pressure tapped array istaken and the average is taken between the two arrangements ofthe pressure tapped array, the process can provide an averagelyinterpolated pressure distribution along the center of a single vane,as shown in Fig. 8c.

(a) (b) (c)

Fig. 8 Schematic of pressure tapped array and mirrored pres-sure tapped array of counter-rotating VG configuration

DECEMBER 2009, Vol. 131 / 121102-7E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Similarly, for the co-rotating VG configuration, the pressuremeasurement is taken with two co-rotating VG configurations,which compose mirroring configurations of each other, as illus-trated in Figs. 9a and 9b, and the periodic lines are consideredinstead of the symmetric lines. An image of the mirrored pressuretapped array in the co-rotating VG configuration given in Fig. 9cshows the absolutely same as that of the counter-rotating VG con-figuration given in Fig. 8c. Therefore, the procedure also canprovide the averagely interpolated pressure distribution along thecenter of a single vane.

R

Automob. Eng., 2201, pp. 113.11 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.

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

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

14 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.

15 Lin, J. C., Robinson, S. K., McGhee, R. J., and Valarezo, W. O., 1994, Sepa-ration Control on High-Lift Airfoils Via Micro-Vortex Generators, J. Aircr.,316, pp. 13171323.

16 Storms, B. L., and Jang, C. S., 1994, Lift Enhancement of an Airfoil Using aGurney Flap and Vortex Generators, J. Aircr., 313, pp. 542547.

17 Storms, B. L., and Ross, J. C., 1995, Experimental Study of Lift-EnhancingTabs on a Two-Element Airfoil, J. Aircr., 325, pp. 10721078.

18 Garcia, D. L., and Katz, J., 2003, Trapped Vortex in Ground Effect, AIAA J.,414, pp. 674678.

19 Katz, J. and Morey, F., 2008, Aerodynamics of Large-Scale Vortex Generatorin Ground Effect, ASME J. Fluids Eng., 1307, p. 071101.

20 Allan, B. G., Yao, C.-S., and Lin, J. C., 2002, Numerical Simulations ofVortex Gner-ator Vanes and Jets on a Flat Plate, AIAA Paper No. 20023160.

21 Allan, B. G., Yao, C.-S., and Lin, J. C., 2002, Simulation of EmbeddedStreamwise Vortices on a Flat Plate, NASA/CR-2002-211654, ICASE ReportNo. 2002-14.

22 Brunet, V., Francois, C., Garnier, E., and Pruvost, M., 2006, Experimentaland Numerical Investigations of Vortex Generators Effects, AIAA Paper No.20063027.

23 Bender, E. E., Anderson, B. H., and Yagle, P. J., 1999, Vortex GeneratorModeling for Navier-Stokes Codes, Paper No. FEDSM99-6919.

24 May, N. E., 2001, A New Vortex Generator Model for Use in Complex

Periodic linePressure tapped arrayMirrored pressure tapped array

(a) (b) (c)

Fig. 9 Schematic of pressure tapped array and mirrored pres-sure tapped array of co-rotating VG configuration

1

Downloaeferences1 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 Gad-el-Hak, M., 1990, Control of Low-Speed Airfoil Aerodynamics, AIAA

J., 289, pp. 15371552.4 Gad-el-Hak, M., and Bushnell, D. M., 1991, Separation Control: Review,

ASME J. Fluids Eng., 113, pp. 530.5 Zerihan, J., and Zhang, X., 2000, Aerodynamics of a Single Element Wing in

Ground Effect, J. Aircr., 376, pp. 10581064.6 Zhang, X., and Zerihan, J., 2003, Off-Surface Aerodynamic Measurements of

a Wing in Ground Effect, J. Aircr., 404, pp. 716725.7 Zhang, X., and Zerihan, J., 2003, Aerodynamics of a Double-Element Wing

in Ground Effect, AIAA J., 416, pp. 10071016.8 Zhang, X., and Zerihan, J., 2004, Edge Vortices of a Double-Element Wing in

Ground Effect, J. Aircr., 415, pp. 11271137.9 Zerihan, J., and Zhang, X., 2001, Aerodynamics of Gurney Flaps on a Wing

in Ground Effect, AIAA J., 395, pp. 772780.10 Soso, M. D., and Wilson, P. A., 2006, Aerodynamics of a Wing in Ground

Effect in Generic Racing Car Wake Flows, Proc. Inst. Mech. Eng., Part D J.

21102-8 / Vol. 131, DECEMBER 2009ded 11 Mar 2010 to 152.78.214.194. Redistribution subject to ASMConfiguration CFD Solvers, AIAA Paper No. 20012434.25 Kuya, Y., Takeda, K., and Zhang, X., Computational Investigation of a Race

Car Wing with Vortex Generators in Ground Effect, ASME J. Fluids Eng.,submitted.

26 Kuya, Y., Takeda, K., Zhang, X., Beeton S., and Pandaleon, T., 2009, FlowPhysics of a Race Car Wing with Vortex Generators in Ground Effect, ASMEJ. Fluids Eng., 131, p. 121103.

27 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.

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

29 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.

30 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.

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

32 Bell, S., 1999, A Beginners Guide to Uncertainty of Measurement, Mea-surement Good Practice Guide, 112, pp. 133.

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