REPORT DOCUMENTATION PAGE 0MB No. 07o4-18A

Pub] rpr-I v r-oe fv -1 ~ekliv f xa m - , ,at?o k, j- rj-4 , ,. e esp.r~e nuwoiriq tt~e ume tor reviewing nstr cons e~rhinq emr ng ca6 Qurcp,.gaiittrg al' ma'sd-' 1.h ddla nteea. ', ,wj i rt -. e" j mivrmdvn >tfd %orrmerns rN~arairq this burden esumate or in tier 4i-pect oft his

collektlurA A 11 I m,.ivt ,&T-0r4 vi ieou.fl f kr,,> Dua n tv.1,~jr~v riedaquirle- er , es, irecturte for Antflt 0 n QflU eravon$ dria ttepvr1 2 45~ Ij efe(50flOasg ,s H1- ... e ' e 4. -w q n. A 22G -4302, and w tO f ,e I, dr.. 6dent y , Pae. per RRdetion Proiet (0104-0 188), ',asongton. "C 20503

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED

06 Jun 88 Conference Presentation4. TITLE AND SUBTITLE 5. FUNDING NUMBERSUnsteady Aerodynamic Forces at TA 2307-FI-38Low Airfoil Pitching Rates00

6. AUTHOR(S)J. Albertson, T. Troutt, C. Kedzie

7. PERFORMINGQRGANIZATION NAME(S), AND ADDRESS(ES) 8. PERFORMING ORGANIZATIONREPORT NUMBER

S F.J. Seiler Research Laboratory FJSRL-PR-90-0015USAF Academy CO 80840-6528

9. SPONSORING, MONITORING AGFNCY NAME(S) JD AD L31i 10. SPONSORING, MONITORING

TAGENCY REPORT NUMBER

11. SUPPLEMENTARY NOTES OE

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Distribution Unlimited

13. ABSTRACT (Maximum 200 words)\L-t-Experiments were conducted on a NACA-0015 airfoil undergoing low

constant pitch rates to study the effects of dynamic stall formationon the airfoil upper surface pressure field. The airfoil was pitchedabout pivot locations of 0.25c, 05c, and 0.75c at nondimensionalpitch rates below 0.2. Lift and drag coefficients were evaluated forall cases, and smoke flow visualization at low pitch rates was

CD studied for the quarter chord pivot location. Results indicate thatthe greatest increases in lift due to the pitching motion occur priorto the nondimensional pitch rate of 0.1 for all three pivot

L1j locations. The effects of pitch rate on the maximum lift and dragvalues appear similar for the three pivot locations studied. Lift todrag ratios show significant enhancement even at very lownondimensional pitch rates. Flow visualization indicateG that theleading edge dynamic stall vortex is present even at very lownondimensional pitch rates. (e6luoO : ACwO Oc, )iNf/4ra AITC ilt

14. SUBJECT TERMS 15. NUMBER OF PAGESflow visualization 9

f9%o!rces' a .... 16. PRICE CODE

17. SECURITY CLASSIFICATION '8, SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT

OF REPORT OF THIS PAGE OF ABSTRACTUNCLASSIFIED UNCLASSIFIED UNCLASSIFIED NONE

NSN 7540.01-280-5500 Standard Form 298 (Rev 2-89)- P(ee(tibd b ANSt Std M04~

• * . A

88-2579-CP

UNSTEADY AERODYNAMIC FORCES AT LOW AIRFOIL PITCHING RATES

Julie A. Albertson*and

rimothy R. Troutt "

Department of Mechanical and Materials EngineeringPullman;WA99164-2920

Christopher R. Kedvie' °*

Frank J. Seiler Research LaboratoryUSAF Academy, Colorado Springs, CO 80840-6528

Abstract Introduction

Experiments were conducted on a NACA-O015 airfoil The flow phenomena associated with unsteady airfoil

undergoing low constant pitch rates to study the effects of motion have been a major research focus for the past two

dynamic stall formation on the airfoil upper surface pres- decades.' Specific attention has been given to dynamic stall

sure field. The airfoil was pitched about pivot locations of and its associated effects on the airfoil surface pressure

0.25c, 0.50c, and 0.75c at nondimensional pitch rates, c+, field.2' 3 The enhanced aerodynamic forces resulting from

below 0.2. Lift and drag coeflicients were evaluated for all pitching motions may lead to improved maneuverability of

cases, and smoke flow visualization at low pitch rates was high-performance aircraft. 4

studied for the quarter chord pivot location. Results indi- The phenomena of dynamic stall occurs when airfoils

cate that the greatest increases in lift due to thr pitching are pitch up rapidly beyond their static stall angiv of attack.

motion occur prior to the nondimensional pitch rate of 0.1 As the angle of attack increases, the flow separates and a

for all three pivot locations. The effects of pitch rate on the region of vorticity develops. The dynanic stall vortex forms

maximum lift and drag values appear similar for the three near the airfoil leading edge, grows a- the airfoil continues

pivot locations studied. Lift to drag ratios show significant its upward motion, and eventually moves along the airfoil

enhancement even at very low nondiimensional pitch rates, surface and detaches. Numerous studies have been done

Flow visualization indicates that the leading edge dynamic on sinusoidally oscillating airfoils to analyze the dynamicstall vortex" is present even at very low nondiniensional stall vortex characteristics and the corresponding pressure

pitcl rates, field.5.6,7.8Recent experimental studies have concentrated on the

simpler case of airfoils undergoing constant pitching mo-lonmenclature tions. These investigations have focused on the developing

unsteady aerodynamic forces and the dynamic stall phenomn-ena occurring on the surface. Dramatic increases in lift and

c airfoil chord drag from corresponding maximum static levels occur over a

Cd airfoil pressure drag coefficient wide range of nondimensional pitch rates, ae. 9'10 ' FlowS lvisualization studies have also shown direct correlations be-C1, airfoil lift coeflicienmt

tween the dynamic stall vortex anti the tipper surface pres-L/D airfoil lift to drag ratio = Gr,/C sure field.12 "13 '14

Re chord Reynolds number = 9 Experimental results are presented here to clarify the

UM freestrean velocity effects of pivot locations on airfoil performance at relativelya pitch rate low nondimensional pitch rates. The concentration on lowa + nondimnensional pitch rate= nondimensional pitch rates pursued in this work stems fromV kinematic viscosity previous experimental observations that the dependence of

unsteady aerodynamnic forces on pitch rate is especially strong

in this region. 15 This is one of the first attempts to corre-late the effect of pivot location on the unsteady aerodyna||icforces at these low nondimensional pitch rates. In addition,the influence of pitching motion and pivot location on liftto drag ratios is explored to determine relationships between

Research Assistant, Student Member AIAAAssociate Professor, Member AIAAResearch Associate

454 I

90 10 30

the upper surface pressure field on the airfoil and the dy- The relationship between the maximum lift coefficientsnamic stall phenomena. The associated behavior of the dy- and the nondimensional pitch rate is shown in figure 2 fornamic stall vortex at low nondimensional pitch rates is also three different pivot locations. The behavior of the max-examined using flow visualization results. imium lift coefficient with respect to cr is similar for the

three pivot locations. Maximum lift values increase rapidlybetween a+ = 0 and 0.05, and then level off for the remain-

Experimental Methods der of the pitch rate range studied. This finding is similar toresult- reported by Francis and Keesee t 7 concerning airfoils

The experiments discussed here were conducted at the pitching in a similar a+ range with pivot locations of 0.32cU.S. Air Force Academy in the Frank J. Seiler Research Lab- and 0.38c. Maximum lift coefficients are equal within the

oratory's 0.91 rn x 0.91 in low speed wind tunnel. The model indicated errors for all three pivot locations.used was a NAdA 0015 airfoil with a 15 cm chord and a5 8 The iel~iiiip'btivih the draig c6efflci nt and-tne

cm span. The airfoil was pitched at a constant rate front 0° attack angle corresponding to ot+ values between 0 and 0.15

to 60' around pivot locations of 25%, 50% and 75% of the is given in figure 3 for the 0.25c pivot location. Two peaks

chord. The airfoil motion was accomplished using a stepper are apparent in the a+ = 0.03 curve, wikh the latter peak

motor assembly controlled by a MASSCOMP 5500 micro- being slightly higher in magnitude than the first. Two peaks

computer system. Instantaneous surface pressure measure-, are again evident in the curve for cr+ = 0.1, although in this

ments were made using eighteen Endevco 8507-2 miniature case it is the first peak which coincides with the maximum

pressure transducers mounted in close-coupled connection to drag. Only one peak occurs in the curve at a+ = 0.15. The

surface pressure ports. magnitude of the maximum drag coefficient and the attack

Flow visualization was obtained using a 0.013 cm di- angle at which it occurs increases with increasing nondimen-

ameter tungsten wire located 30 cm upstream of the airfoil sional pitch rate for the four pitch rates shown.

leading edge and perpendicular to the airfoil span. A smoke The relationship between the maximum drag coefficientproducing oil was coated on the wire and then heated electri- and the nondimensional pitch rate can also be seen in figure.

cally to create uniform streaklines. ligh speed movies were 4. All three pivot locations follow the same trend althoughtaken with a LOCAM 1116 mm high speed movie camera op-. values for the 0.50c pivot location deviate slightly from pat-crating at 20.0 frames per second. Illumination was provided terns followed by-the other two cases between a+ = 0.075by three Strobrite stroboscopic lights synchronized with the, and 0.15. The maxinum drag coefficients increase rapidly tocamera. Pressure meastirements were taken with a tunnel a local maximum at a+ = 0.075 for all three pivot locations.speed of 7.62 m/s, corresponding to a chord Reynolds num-. Values gradually increase again between a+ = 0.1 and 0.2.

'ber of 60000, while flow visualization was done with a tunnel The attack angles coinciding to the maxinuinn lift andspeed of 6.10 m/s, corresponding to a chord Reynolds num- drag coefficients as a function of a+ are shown in figure 5 forber of 47000, allowing higher resolution. A range of nondi- pitching about the quarter chord. The attack angle corre-mensional pitch rates a+, from 0 to 0.2 was evaluated in sponding to maximum lift curve exhibits behavior similar todetail. It has been shown previously that a+ can be treated the maximum lift curve shown in figure 2. The maximum liftas a similarity parameter in flows such as the one studied attack angle increases between a+ = 0 and 0.075, and thenhere. 16 begins to level off. The attack angle corresponding to peak

drag follows the same trend as the peak drag curve shown infigure 4. The maximum drag attack angle increases rapidly

Experinental ilesuLtLs between a+ = 0 and 0.075, decreases between c4 values of0.075 and 0.1, and then begins to increase in a manner sim-ilar to the maximum lift attack angles. The attack anglescorresponding to maxinun drag coefficient are significantly

Pressure Measurements higher than those coinciding to maxinmum lift coefficient be-tween a+ values of 0.03 and 0.1.

The change in lift coefficient with respect to angle of The effects ofpitching motion and pivot location on the t10

attack is presented for the quarter chord pivot location in airfoil lift to drag ratios were also evalnated. Lift to drag ra-figure 1 at nondimensional pitch rates, a+ between 0 and tios as a function of attack angle for the quarter chord pivot INPECOM0.15. A significant increase in maxinmun lift coefficient from location are shown in figure 6 at the nondinmensional pitchthe static case is evident even at an a+ of 0.03. Maximum lift rates of 0, 0.02, and 0.1. Values for time static cas -are greatercoefficients continue to increase for the subsequent a+ values than those for the two dynamic cases between attackanglesof 0.1 and 0.15, although the differential between maxinmum of 0' and 140, where 14' corresponds approximately to staticlift coefficients diminishes as the nondimensional pitch rate stall. Lift to drag ratios for the static case drop off rapidly 'orincreases. after stall, but lift to drag ratios at corresponding attack an-

The angle of attack at which maximum lift occurs also gles for the dynamic cases are higher than time static ratios.increases as the nondimnensional pitch rate grows. Two peaks The range over which lift to drag ratios are greater than the "are apparent in the lift coefficient curve at a+ = 0.03, with static case increases with increasing pitch rate. Although ex- 1 3the greatest lift value for that pitch rate occurring at the act values cannot be deternmined due to experimental errors, LO i|Ifirst peak. Only one significant peak is evident at a+ values an increase in lift to drag ratios for the dynamic cases overof 0.1 and 0.15. There are slight peaks in the ce+ = 0.03 values for time static case is clearly evident. Both the curveand 0.1 curves at attack angles higher than the maxinun at a+ = 0.02 and at a+ = 0.1 asymptote towards the staticlift angles, but it be noted that these peaks are not greater values later in the pitching cyc!e. )l/than the experimental error. Figure 7 shows lift to drag ratios as a functioni of nondi- Lty Codo

- and/or

455 V1.3zecial

mensional pitch rate for attack angles between 50 and 200. namic stall vortex can again be seen at a = 200 by close ob-The highest cverall lift to drag ratios occur at ca = 50, with servation, but it's subsequent growth is noticeably delayed.values at this attack angle being approximately constant be- At ca = 250, given in figure Ile, the dynamic stall vortex istween ct+ = 0.005 and 0.1, and then decreasing. For an clear but occupies a smaller surface region than it did forattack angle of of = 150 lift to drag ratios are approximately the corresponding attack angle for or+ = 0.03. The vortex isconstant between cr+ = 0.005 and 0.075 and then begin to 'still involved with the airfoil surface at at = 300, but has de-decrease. For the attack angle cc = 15° lift to drag ratios tached by a = 35°. A loose trailing edge vortex is evident atincrease rapidly between a+ = 0 and 0.005 and then remain this attack angle, shown in figure 1 g, and a second dynamicconstant for the higher pitch rates. Values for a = 200 show stall vortex appears to be forming.a slight increasing trend between a+ - 0 and 0.02 and then The presence of more than one dynamic stall vorlex atbecome constant. It should be noted for attack angles of 50 a4 = 0.03 and 0.05 seems to have a direct effect on the shapeand 100 that initial lift to drag ratios decrease from the static of the lift anddrag coefficient curves, since the peaks in thesecase, ct+ 0, to the dynamic case of a + = 0.005, whereas curves coincide with the attack angles at which the dynamicthey increase in this range for an attack angle of 15*. This stall vortex is apparent on the airfoil upper surface. It hasis caused by the occurrence of static stall near a = 140 and been shown previously that movement of the dynamic stallthe subsequent drop in static lift to drag ratios. vorti-es on the airfoil is closely connecled to the aerodynamic

An evaluation was also made to determine the effect of forces generated, and that peaks in the drag coefficient withpivot location on lift to drag ratios. Graphs for cy+ = 0.02 respect to attack angle curves correspond to detachment ofand 0.1 are shown in Figures 8a and 8b respectively. Prior to, the dynamic stall vortex from the airfoil surface.' 5

ca = 140, lift to drag ratios are slightly lower for the 0.25c and0.75c pivot points than the 0.50c case at an a+ of 0.02, shownin figure 8a. Lift to drag values for the 0.75c pivot location Conclusionsare the lowest of the three pivot cases in this region. Thistrend continues for the case of a+ = 0.1, given in figure 8b.Lift to drag ratios for the 0.75c pivot location are consistently The results of this study indicate that the greatest en-

lower than the other two cases, and values for the 0.50c pivot hancement of lift due to the constant pitching motion of the

locations are higher than those for 0.25c and 0.75c positions NACA-0015 airfoil occur at low nondinensional pitch rates,

prior to c = 200. Lift to drag ratios are similar for all three namely a+ below 0.1, for Ihe three pivot points evaluated.

pivot locations for attack angles greater titan a = 200. The range of nondimensional pitch rates between cc+ = 0.03and 0.1 coincides with the region where the occurrence ofmaximum lift significantly precedes maximum drag in the

-10 W -VQ.ySualLj~Qf pitching cycle. Lift to drag ratios were found to depend onpivot position, with values being consistently lower for thethree-quarter chord pivot location than for the other two

Smoke flow visualization for the quarter chord pivot cases. It was found that lift to drag ratios are greater for thepoint is sliowi in figure 9 for cx+ = 0.01. The flow ap. dyiamic cases than the static case at high angles of attack.pears symmetric at a = 5'. At & = 10, a separation region Flow visualization showed that the dynamic stall vortex is

has formed near the trailing edge and is moving upstream, present eve at a nondimensional pitch rate of 0.01, and that

The flow is separated over a large percentage of the airfoil presene o usqen vorticesconrthe upper a l sur-surfce t a= 1'. losel deine dyami stll ortx Lthe presence of subsequent vortices on the upper airfoil sur-surfce t a= 15. Alooelydefied lynmic tal votex face during the pitching cycle creates local maxima in the

covering virtually the entire upper surface is clearly evident

at c = 200. Unfortunately, the flow visualization methods lift and drag coefficients.

used in this experiment did not allow flow observations athigher attack angles for this nondimuensional pitch rate.

Flow visualization for a+ = 0.03 is presented in figure Acknowledgenents.10 for the quarter chord pivot point. The flow follows a si-ilar pattern as a+ = 0.01, although respective changes occur This research was sponsored by tIhe Air Force Officeslightly later in the pitching cycle, A separation region is of Scientific Research/AFSC, United States Air Force, un-beginning to grow at c = 15', and covers a greater portion der Contract F49620-85-C- 0013. The authors gratefully ac-of the airfoil surface at a = 200. Close observation at this knowled,e the support of Major J.M. Walker and Dr. M.C.attack angle also reveals evidence of a dynamic stall vortex Robinson. They also thank Captain E. Stephen for his help -

forming at the leading edge. This vortex is even more evi- with experimental preparations, and the supporLstaff6f'hedent at a = 250, and covers approximately half of the airfoil Frank J. Seiler Research Laboratory for general assistancechord. Between angles of attack of 250 and 30' the dynamic with experimental facilities.stall vortex is disrupted by a counter-clockwise vortex whichforms at the airfoil trailing edge. At the attack angle of 300shown in figure 10f, the dynamic stall vortex has detached References

from the airfoil surface and caii be seen in the tipper lefthiam! corner of the figure. A second dynamic stall vortex isbeginning to form at a'= 30, and covers the entire airfoil 1. McCroskey, W.J., "Unsteady Airfoils," Anr Rev Fluidsurface at a = 350. This vortex is not as clearly defined and McIcl-i 14, pp. 285-311, 1982.ha5 a shorter surface residence time than the first vortex. 2. Reynolds, W.C. and Carr, L.W., "Review of Unsteady,

Flow visualir-,tion at a+ = 0.05, shown in figure 11, Driven, Separated Flows," AIAA Paper No. 85-0527,exhibits ilo; behavior similar to that at a = 0.03. The dy- March 1985.

456 I

3. Gad-el-llak, M., "Unsteady Separation on Lifting Sur-faces," Appl. Mech. Rev., Vol. 40, No. 4, pp. 441-452,April 1987.

4. llerbst, W.B., "Supermaneuverability," Proceedings fromWorkshop on Unsteady Separated Flows, IJSAFA, De-partment of Aeronautical Engineering,oUniversity of Col-orado, Boulder, CO, pp. 1-9, August 1983.

5. McAlister, K.W. and Carr, L.W., "Water Tunnel Visual-izations of Dynamic Stall," J. Fluids Engr, 101, pp. 376-380, September 1979.

6. Adler, J.N., Robinson, M.C., Luttges, M.W., andKennedy, D.A., "Visualizing Unsteady Separated Flows,"Flow Visualization lII, Hemispheie Publishing,New York, pp. 342-347, 1985.

7. Robinson, M.C. and Luttges, M.W., "Unsteady Flow Sep-aration and Attachment Induced by Pitching Airfoils,"AIAA Paper No. 83-0131, June 1983.

8. Gad-el-llak, M. and IHo, C.M., "Three-Dimensional Ef-fects on a Pitching Lifting Surface," AIAA Paper No. 85-0041, January 1985.

9. Walker, J.M., Itelin, II.E. and Strickland, J.11., "An Ex-,perimental Investigation on an Airfoil Undergoing Large'Amplitude Pitching Motions," AIAA Paper No. 85-0039,January-1985.

10. Jumper, E;J., Shreck, S.J., and Dinmick, R.L., "Lift-Curve Characteristics for an Airfoil Pitching at a Con-stant Rate," AIAA Paper No. 86-0117, January 1986.

11. Walker, J.M., and Chou, D.C., "Forced Unsteady VortexFlows Driven by Pitching Airfoils," AIAA Paper No. 87-

331, January 1987.12. helin, I[.E. and Walter, J.M., "Interrelated Effects of

Pitch Rate and Pivot Point on Airfoil Dynamic Stall,"AIAA Paper No. 85-0130, January 1985.

13. Strickland, J.HI. and Graham, G.M., "Force Coefficientsfor a NACA-0015 Airfoil Undergoing Constant Pitch RateMotions," AIAA J., Vol. 25, No. 4, pp. 622-624, April1987.

14. Alb.ertson, J.A., Troutt, T.R.,.Siuru, W.D., and Walker,J.M., "Dynamic Stall Vortex Development and the Sur-face Pressure Field of a Pitching Airfoil," AIAA PaperNo. 87-1333, June 1987.

15. Walker, J.M., Helin, 1i.E., und Chou, D.C., "UnsteadySurface Pressure Measurements on a Pitching Airfoil,"AIAA Paper No. 85-0532, March 1985.

16. Cook, R.J., "Similarity Condition§ for Flows About Pitch-ing Airfoils," FJSRL-TM-87-0003, June 1987.

17. Francis, M.S. and Keesze, J.X., "Airfoil Dynamic StallPerformance with, Large Amplitude Motions," A.IAAJ..,Vol. 23, No. 11, pp. 1653-1659, November 1985.

47457 '

3.0

IUncert. PIVOTSEs'. - 4- Uncert. 00.25c

2.5- .. 0.15 Est. 0 0.50c0 0.75c

2.0- ..

/V -0.03 ~CL

1.5-/ '

10 ,-~

0 0 005 0.10 015 020

0 +

jtigurii 4. Maxinni drag coefficient as a func-

Figure 1. Lift coefficient as a Function of iittack -tinonoinnsnaptcrte

angle. Pivot 0 .25c.

4'vr 6CUncert 0 0.25C fnetEst. 0 0.50C ',Est.0 0.75c

a 30

20

0 005 0.10 0.15 0.20a+

Figure 2. Maximum lift coefficient as a function 0 005 0.10 0.15 0.20

nonlineJsional pitch rate. Ot+

2.5 Et inlul lift and dIrag coefficient as a0s4-igre5 Anle of tc orriesoningtcah

2.0-0.15rate. P'ivot location 0 .25c.

CD 1.5

1.0-

0.5-

0 -0 10 20 30 40 50 60

Figure 3. Drag coefficient as a function of attackanle. Pivot 0 .25c.

4S8

50 50-

40 Urcert. 4 PIVTx Uncert.

30 /D30 j 0.75c

L/D L/

2 0 -~ 20 II to

10

08 L .-A- ... 0 10 20 30 '10 50 s00 10 20 30 40 s0 60

Figure 6. Lift to drag ratios as a function ofr 00attack angle. Pivot location =0.25c. 5

Uncert.

40.t

30-

20LID Q50C

LID 20- 0.25S

R0 03 t- %C 5c

0P 110 20 30 40 50 60

00.05 0.10 0.15 0.10

a + b) c, = 0.1

Figure 7. Lift to drag ratios as a function ofnondi(uflnsional pitch rate. Pivot Figure 8. Lift to drag ratios as a function of

attack anile.

459

0C a

b) a =10' d =.20'

Figure 9. Smoke flow visualization. Pivot loca-tion =0.25c. a+ =0.01.

460 A

a) a=50e) or= 250

b) a 100 f) 300

c) a, 150 g) ar 35

Figure 10. Suuioke flow vistiaizattion. Plivot loca-

tioni 0.25c. a+ - 0.03.

d)a=200

461 ( /

a) =o5 e) a =250

b)a100 :f) a =300

c) a= 15 0

g) a = 350

Figure 11. Smoke flow visualization. Pivot loca-tion = 0.25c. a+ = 0.05.

d) a = 200

462