aiaa 2003-0922 recommended experimental procedures …mln/ltrs-pdfs/nasa-aiaa-2003... · 2003. 1....

For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics1801 Alexander Bell Drive, Suite 500, Reston, VA 20191

RECOMMENDED EXPERIMENTALPROCEDURES FOR EVALUATION OFABRUPT WING STALL CHARACTERISTICS

41st AIAA Aerospace SciencesMeeting and Exhibit

AIAA 2003-0922

F. J. Capone, R. M. Hall, D. B. Owens,J. E. Lamar and S. N. McMillinNASA Langley Research CenterHampton, Virginia

6-9 January 2003Reno, Nevada

AIAA-2003-0922

Recommended Experimental Procedures For EvaluationOf Abrupt Wing Stall Characteristics

F. J. Capone†, R. M. Hall*, D. B. Owens§,J. E. Lamar‡, and S. N. McMillin**

NASA Langley Research Center, Hampton, Virginia

ABSTRACT

This paper presents a review of the experimental program under the Abrupt Wing Stall (AWS)Program. Candidate figures of merit from conventional static tunnel tests are summarized andcorrelated with data obtained in unique free-to-roll tests. Where possible, free-to-roll results arealso correlated with flight data. Based on extensive studies of static experimental figures of meritin the Abrupt Wing Stall Program for four different aircraft configurations, no one specific figure ofmerit consistently flagged a warning of potential lateral activity when actual activity was seen tooccur in the free-to-roll experiments. However, these studies pointed out the importance ofmeasuring and recording the root mean square signals of the force balance.

SYMBOLS AND ABBREVIATIONS

AFOM alternate figure of meritAWS abrupt wing stallCA axial force coefficientCL lift coefficientCLa lift-curve slope, per degreeCi rolling-moment coefficient

Ci,rms root mean square of CiCN normal-force coefficientCN,rms root mean square of CNCm pitching-moment coefficientCn yawing-moment coefficientCLWB left-wing bending-moment coefficientC LWB,rms root mean square of CLWBCRWB right-wing bending-moment coefficientC RWB,rms root mean square of CRWB rCY side-force coefficient

FOM figure of meritFTR free-to-rollLERX leading-edge root extensionM Mach numberRMS root mean squareTFOM traditional figure of meritt timea angle of attack, degda aileron angle, degdle leading-edge flap angle, degdte trailing-edge flap angle, degf model roll angle, degq model pitch angle, degSFOM,3 summation figure of merit (ref. 9)SFOM,22 summation figure of merit (ref. 9)pFOM product figure of merit (ref. 9)16-ft TT 16-Foot Transonic Tunnel

INTRODUCTION

A joint NASA/Navy/Air Force Abrupt WingStall Program (AWS) was established after severalpre-production F/A-18E/F aircraft experiencedsevere wing-drop motions during thedevelopment stage. A Blue Ribbon Paneldetermined that a poor understanding of thephenomena causing the problem existed, andmade the recommendation to: “Initiate a nationalresearch effort to thoroughly and systematically

_________________________________†Senior Research Engineer,*Senior Research Engineer, Associate Fellow AIAA§ Aerospace Engineer, Member AIAA‡ Aerospace Engineer, Member AIAA**Senior Research Engineer, Associate Fellow AIAA______________________________________This material is declared a work of the U. S. Governmentand is not subject to copyright protection in the UnitedStates.

AIAA 2003-0922

American Institute of Aerodynamics and Astronautics

2

study the wing drop phenomena.” The problemarea addressed by the AWS Program1-17 is theunexpected occurrence of highly undesirablelateral-directional motions at high-subsonic andtransonic maneuvering conditions. One of therecommendations made by Chambers2 was “todefine and assess candidate figures of merit forthe prediction of wing drop and wing rock fromexperimental methods.”

The interpretation of experimental dataobtained in wind tunnels for conditions involvinghighly separated wing flows has always been verysubjective and open to many opinions as to whatis happening. One aircraft that went through anextensive experimental program because ofuncommanded wing rock/drop was the Harrier18.The potential existence of wing rock/drop wasinterpreted from wind-tunnel rolling-moment datarecorded with on-line continuous pen recordersas shown in figure 1. For example, a continuousoscillation of the pen was interpreted to be wingrock, whereas a steady departure was consideredto be wing drop. Note the relatively smooth liftcurves with no abrupt breaks.

This paper will review the scope of theexperimental program conducted under theAbrupt Wing Stall (AWS) Program. Four differentaircraft configurations that either do or do notexhibit uncommanded lateral motions weretested. Along with the pre-production F/A-18E,the AV-8B was chosen as a configuration thatexhibits wing drop. The two configurations that donot have wing-drop were the F/A-18C and the F-16C. Candidate figures of merit from conventionalstatic tunnel tests are summarized and correlatedwith data obtained in unique free-to-roll tests.Where possible, free-to-roll results are alsocorrelated with flight data. Recommendations asto how to conduct an experimental program onfuture vehicles are made.

In order to obtain approval for releasingthis paper to the public, quantitative informationhas been removed from most vertical scales as perguidelines from the Department of Defense.

ABRUPT WING STALL EXPERIMENTALPROGRAM

An extensive experimental wind-tunneltest program has been conducted as part of theAWS program. Early in the AWS program1, aseries of tests were conducted on an 8-percentpre-production F/A-18E model with new set ofwings that were heavily instrumented with wingbending-moment gauges, static-pressure taps,and unsteady pressure gauges. Both static anddynamic force balance measurements were made,

as well a tunnel entry devoted to using thepressure sensitive paint technique3,5. These testswere conducted prior to the operational readinessof the free-to-roll rig10.

One of the objectives of the AWSprogram was to test four different aircraft that areknown to either exhibit uncommanded lateralmotions or not. Along with the pre-productionF/A-18E, the AV-8B was chosen because itexhibited wing drop behavior. The twoconfigurations that did not have wing-drop werethe F/A-18C and the F-16C. The overallgeometric dimensions of these models are alsoshown in figure 2 as well as a summary of thescope of experimental measurements made in theprogram.

All four models were tested in the 16-FootTransonic Tunnel over an eleven-week period,during which both static and free-to-roll data weretaken. For this portion of the overall investigation,the F/A-18E model was tested with the baselinewings that did not contain any of theinstrumentation previously noted for the newwings. Details of the experimental proceduresused for both the static and free-to-rollinvestigations can be found in reference 10.

WHAT STATIC TESTING IS NECESSARY?

Data GranularityOne important aspect of experimental

wind tunnel testing has always been “how muchdata is necessary”. One major conclusion of thisprogram has been that it is necessary to take datain increments of angle of attack at least as small as0.5° in the region where wing stall occurs. Thisrequirement is illustrated in figure 3, where it maybe noted that if the data had been taken either in2° or 1° increments, the sharp break in the lift curvewould not have been measured, and the potentialfor wing drop may have been missed.

Static Forces and MomentsOne question that arose during the AWS

Program was whether one could rely solely onfigures of merit derived from static data takenduring a transonic model test to provide thecertainty needed that a new aircraft would or wouldnot be susceptible to uncommanded lateralmotions during its flight operations. Static figuresof merit, or FOMs, are very important because theyare intended to predict wing drop/rock behaviorand, consequently, will enter into the decision ofdoing FTR testing or not. Comparing static testresults with the FTR response data has provided a

AIAA 2003-0922


3

rational basis for assessing the merits of usingstandard test techniques for the prediction ofAWS events. A review of selected results fromthe investigations of four configurations will bemade in order to define which static testmeasurement techniques should be employed tomeet these requirements.

The sets of data for the four models werechosen as being representative of when bothtraditional (TFOMs) and alternate (AFOMs) figuresof merit indicated that AWS activity would bepresented and where these figures of meritindicated some activity, but none was present.Three of the TFOMs chosen for this analysis werelift coefficient CL, rolling-moment coefficient Ci androot mean square of rolling moment Ci,rms. Inaddition, three of the AFOM from reference 9 arealso included.

One is now faced with the problem ofdefining which criteria are indicative of an AWSevent. Figure 4 presents criteria that have beenused traditionally to predict the onset of an AWSevent. First, sharp breaks in lift-curve slope havetraditionally been used as an indicator of an abruptstall, which because of its nature may beasymmetric. Sudden or rapid variations in eitherrolling-moment coefficient or root mean square ofrolling moment can signify rapid flow topologychanges as angle of attack changes. The angle ofattack at which these curves have a spike or peakis used as the criteria to indicate AWS activity forthese two TFOMs and also for the AFOMs.However, as will be shown, where AWS activity isinitiated is a matter of conjecture, and it could beargued that this activity starts at some angle ofattack prior to that indicated by the peak. Forthose cases in which wing bending moments weremeasured, a change in the slope of these curvesis used as the criteria.

Angles of attack at which AWS activity areindicated are then summarized in a “stoplight”chart. The stoplight chart also presents color-coded results measured during the pitch-pausephase of free-to-roll testing. During FTR pitch-pause tests, the model was released from a wings-level condition, and the resulting motions wererecorded. A discussion of the color-coded ratingsystem is given in the Free-To-Roll Figure of Meritsection of this paper.

AV-8B . Four configurations of the AV-8Bhave been selected for discussion, and includethose with the trailing-edge flap at 10°, 100%LERX at M = 0.5 and 0.75, and with the trailingedge flap at 25°, 100% LERX and 65% LERX at M= 0.50. The TFOMs and AFOMs are presented infigures 5 to 8 and compared to FTR results in

figures 9 to 12. Note that symbols have not beenused on any data figures for clarity. However, datawas recorded in increments of angle of attack atleast as small as 0.5°. Data shown on the stoplightcharts are presented as a function of model pitchangle q rather than angle of attack, since duringFTR testing q can be held constant whereas angleof attack a varies during the model rolling motions.When a or q = 0°, these angles differ only bytunnel upflow angularity that averaged about+0.10°.

For the trailing edge flap at 10°, there is amarked difference in the magnitude of the lift-curve slope decrease between a Mach number of0.50 and 0.75 (fig. 5). At M = 0.50, there is a slightdrop in the lift curve slope at a = 7.5°, whereasthere is a much sharper decrease in slope at a =8.5° at M = 0.75 that could be indicative of AWSactivity. However, as can be seen in figure 10,free-to-roll results showed no FTR lateral activity inthis angle of attack range. The changes in wingbending moments noted at a = 8° on figure 10 areconsistent with the decrease in lift curve slope.Also shown in figures 9 and 10 are the CFDpredictions8 of the break in the lift curve at M = 0.5and 0.75. As can be seen, the predicted break inlift curve slope at M = 0.50 occurred about 3°higher than that measured. However, goodagreement is noted between the CFD predictionand wind tunnel at M = 0.75.

Figure 5 also shows a difference in thecharacter of rolling-moment coefficient between M= 0.5 and 0.75. At M = 0.5, no FTR lateral activitywas noted, despite the sharp increase in rollingmoment indicated in figure 5 at a = 16°. The verylarge asymmetries in rolling moment that werepresent at M = 0.75, are more typical of thebehavior one might expect where uncommandedlateral motions might be present and, in fact, wereexperienced during the FTR tests (fig. 10).

At M = 0.5, there is a gradual increase inrolling-moment RMS, starting of about a = 12°.The increase in RMS is probably more anindication of buffet onset than any AWS activity.Buffet onset also occurs where there is a break inthe axial force coefficient CA curve. As seen infigure 5, the break in CA also occurs at angle ofattack of 12°. At M = 0.75, buffet onset occurs at a= 8°. Using the criteria previously described, AWSactivity would be possible at a = 12° where therolling moment RMS curve peaks for M = 0.75.Note that rolling-moment RMS returns to a lowerlevel at an angle of attack greater than 14° afterthe wing has totally stalled. Based on theforegoing results, it should be obvious that tryingto predict AWS activity depends on looking atmore that a single balance component.

AIAA 2003-0922


4

A summary o f the free-to-rollcharacteristics for the AV-8B model for M = 0.5with 25° flaps and the 100% LERX is presented infigure 11. For this configuration, some of theTFOMS and AFOMs predict AWS activity, butthese predictions seem to occur randomly atvarious angles of attack.

However, for the AV-8B with the 65%LERX (fig. 12), several of the figures of meritpredict AWS activity at q = 14 to 14.5°, and severeAWS activity was experienced in the FTR tests.Note that there were also false predictions ofactivity near a = 10°. Wing rock occurred with flapson schedule at q = 14° to 14.5°.

For the AV-8B, both rolling moment andRMS rolling moment appear to be the significantfigures of merit, and the results suggest that anyfuture program include these measurements.The results of the static portion of the tests on theAV-8B also suggest a need to conduct FTR tests.

F/A-18C . The traditional and alternatefigures of merit for the F/A-18C with dle = 6°, dte = 8°and da = 0° (6°/8°/0°) at M = 0.80 to 0.90 arepresented in figures 13 and 14 respectively. Thefirst angle corresponds to leading-edge flapdeflection, the second angle corresponds to thetrailing-edge flap deflection, and the third anglecorresponds to the aileron bias. These data aresummarized in figure 15. At all Mach numbers,there are very significant breaks in the lift curvesthat are in marked contrast to those shownpreviously for the AV-8B. Figure 15 indicates thatsevere FTR lateral activity occurred at the anglesof attack at which the lift curves break, eventhough there are lift-curve breaks at lower a’s atwhich there is no activity. Note that FTR lateralactivity always occurred at angles of attack at whichthe flaps are not on schedule. An examination offigure 15 suggests the difficulty of predicting AWSactivity from the TFOMs and AFOMs. At all threeMach numbers, the figures of merit sometimesreliably predicted activity, and sometimes theyfalsely predicted activity. Also shown in figure 15,are the CFD predictions8 of the break in the liftcurve at M = 0.85 and 0.9. As can be seen, thepredicted break in lift curve slope at M = 0.90occurred about 1° lower than that measured.

For the F/A-18C configuration, lift-curveslope, rolling moment and RMS rolling momentappear to be significant FOMs, such that anyfuture program should include thesemeasurements.

Pre-product ion F /A-18E . Thetraditional and alternate figures of merit for the F/A-18E with flap settings of 6°/8°/4°, 10°/10°/5°, and

15°/10°/5° at M = 0.90 are presented in figures 16and 17. Figure 18 presents some typical wingbending-moment data that were measured duringan earlier entry in the 16FTT. These data aresummarized in figures 19 to 21. The lift curves forthe 6°/8°/4° and 10°/10°/5° flap settings have verysignificant breaks, similar to the F-18C. However,FTR lateral activity was already present, havingstarted some 1° to 1.5° sooner than what wouldhave been predicted by the breaks in lift curves(figs 19 and 20). For the 10°/10°/5° flap setting, allthe figures of merit including the CFD FOM7 lineup and essentially predicted lateral activity. It isinteresting to note that even though the timeaveraged Cl data indicates significant variation witha, the unsteady Cl,rms data is smoother. Note thatwhile FTR lateral activity is present with flaps onschedule, these results are for a model of the pre-production F/A-18E aircraft1 and not arerepresentative of the production aircraft1. For thisconfiguration, lift-curve slope and rolling momenttended to predict AWS activity at angles of attackhigher than where FTR lateral activity was present.However, in the absence of any FTR data, theresults of the static portion of this investigation onthe pre-production F/A-18E clearly show a needfor FTR testing of this configuration.

F-16C . The traditional and alternatefigures of merit for the F-16C with dle = 10° and dte =0° at M = 0.80 and 0.90 are presented in figures22 and 23 respectively. These data aresummarized in figure 24. Except for somerelatively small peaks noted for the Cl,rms, curves,this was a relatively benign configuration.

For the F-16C configuration, the TFOMspredicted that some AWS activity would bepresent, but no FTR lateral activity was present.

Wing PressuresA series of tests were conducted early in

the AWS program (ref. 1), on an 8-percent F/A-18E model with new set of wings that were heavilyinstrumented with static-pressure taps, andunsteady-pressure gauges. These wings werenot used during the free-to-roll tests.

Steady Pressures . An example of thesteady pressure distributions acquired on the F/A-18E model is given in figure 25. Distributions arepresented at angles of attack both below wing stalland above wing stall for a flap setting of 10°/10°/5°,where the first angle corresponds to leading-edgeflap deflection, the second angle corresponds tothe trailing-edge flap deflection, and the thirdangle corresponds to the aileron bias. Thepressure data are presented for various spanwise

AIAA 2003-0922


5

wing stations, or butt lines, seen in the sketch infigure 25. The first pressure distribution, shownby the circles, was taken at an angle of attack of8.0°, before the stall occurred in the mid-wingregion. Abrupt stall occurred immediatelythereafter as angle of attack increased by a smallincrement. The squares are for an angle of attackof 9.0°, which is after the stall for the wing panel atthe mid-wing span locations. As seen during thestall process, significant regions of low pressures--that is, coefficients toward the top of the verticalaxis--collapse as the separation process movesforward up and over the mid-wing region.Evidence of the stall process is apparent in RowsA, E, G, and I, but is clearest in Row G and isconsistent with legacy oil flow images. Furtherdetails of this investigation are discussed byMcMillin3. This type of information is invaluable indetermining the flow physics associated with AWSevents, and the critical spanwise stations wherethe separation is most severe.

Unsteady Pressures An analysis ofthe unsteady pressure measurements has beenmade by Schuster5. Figure 26 shows pressurecoefficient time history data acquired at a singlepressure transducer on the E-row near the centerof the wing box at Mach 0.9. The location of thistransducer is circled on the image of the planformat the bottom of the figure. Time histories areplotted in one-degree angle of attack incrementsfrom 6.5° to 9.5°. The pressure coefficient plottedin this figure is the complete pressure coefficient,as opposed to just the fluctuating component ofthe pressure.

This figure clearly shows the progressionof the shock wave forward on the wing as theangle of attack is increased into the AWS region.At 6.5°, the pressures measured by thetransducer are very stable and constant across thetime slice. At 7.5°, the first hint of a shock movingonto this chordwise location is seen in the discretespikes in the pressure time history. By 8.5°, thespikes are much more prevalent, and finally at 9.5°the time history is saturated with pressure spikesas the shock moves back and forth across thepressure transducer. Hwang and Pi19 observed asimilar unsteady pressure character in their buffetand wing rock analysis of a model of the F-5Aaircraft during transonic wind-tunnel tests.

Further research5 is required into howunsteady pressures might be readily used toscreen for AWS. Generally, a normal test programcannot afford to include unsteady pressuretransducers nor the time to record and analyze thedata. However, a definite recommendation is thatif unsteady pressure transducers are required,they should be included on both wings of the

aircraft as opposed to just the single wing in thisstudy. Lateral phenomena could be readilyextracted and separated from longitudinalphenomena using time synchronized pressuredata from both wings. This would likely provide anentirely new insight into the AWS phenomenon.In addition, the overall coverage of unsteadytransducers should be increased over that used inthe present study. This would probably require alarger scale model and it would surely require amore complex and capable dynamic dataacquisition system than used in this analysis.

Pressure Sensitive PaintIn addition to obtaining steady and

unsteady pressure measurements, pressuresensitive paint (PSP) imaging was used to gatherglobal information on the pressures influencingthe wing drop. This technique, which offers theadvantage of continuous pressure informationacross the wing (in contrast to the discretepressure taps seen in figure 25), was highlysuccessful in this transonic application. As seenin figure 27, the pressure pattern correlates wellwith comparable Veridian 8-ft Transonic Tunnel oilflows, despite the fact that the Veridian test used adifferent model wing than that tested at Langley.The correlation of PSP images with force andmoment data is reported in more detail byMcMillin5. As discussed in reference 3, trade offsare required in operational test time using thetechnique, since considerable setup time isrequired between runs for model cool-down,image calibrations, etc.

CORRELATION OFSTATIC AND FREE-TO-ROLL BALANCE

MEASUREMENTS

One of the FTR major requirements10 inthe design of the free-to-roll rig was that the forcebalance was to be retained and used duringtesting. Even though there were risks involvedwith this requirement, having the force balanceproved to be invaluable, in particular from a safetystandpoint in being able to monitor model loadsduring FTR testing. Another concern was that theforces and moments measured by the balanceduring free-to-roll testing would be erroneous.However, this turned out not to be the case. Acomparison of the measured forces and momentsbetween FTR and static testing is presented infigures 28 to 31. Data for the static tests werepitch-pause, whereas the FTR tests wereconducted using the pitch sweep mode. Theseresults are for the largest model tested, the AV-8B, with the 100% and 65% LERX and with 10°

AIAA 2003-0922


6

and 25° trailing edge flap deflection. Excellentagreement is shown for the two test modes for thelongitudinal data for those cases when no free-to-roll activity is present. However, when free-to-rollactivity is present, large variations in someaerodynamic coefficients do occur due to largechanges in bank angle.

FREE-TO-ROLL TO FLIGHTCORRELATIONS

Free-To-Roll Figure of MeritIn order to discern the level of lateral

activity from free-to-roll tests, figures of merit werealso developed. This figure of merit needs toresolve “signif icant” activity from theinconsequential. Taking into account amplitudealone could lead to the wrong conclusions sincethe motion may have a large amplitude change butthe motion is so slow that it would be easilycontrolled. Taking into account just themagnitude of rates or accelerations alone couldalso lead to the wrong conclusions since a largeacceleration with a small amplitude oscillationmight not be controllable but would not lead to alarge deviation in the aircraft trajectory. Or, theacceleration might be favorable if it is returning theaircraft to a wings level condition.

Initially, amplitude measured during thepitch-pause phase of free-to-roll testing, was usedas a figure of merit. During FTR pitch-pause tests,the model was released from a wings-levelcondition, and the resulting motions were ratedbased on an arbitrary color-coded rating system.This system was based on the ensuing amplitudeof bank angle, where green represented anamplitude of less than 10°, yellow an amplitudebetween 10° and 20° and red any amplitudegreater than 20°. Initial ratings for a selectedconfiguration of each of the four models testedare presented in figure 32. Note the large numberof “yellow” and “red’ events that occur for the F/A-18C and F/A-18E at low angles of attack. Theseevents can generally be characterized as beingslow-period large-amplitude motions.

Therefore, a new figure of merit11 wasdeveloped that accounted for both amplitude andrate. The FOM (pP-V)max is defined as the absolutevalue of the amplitude change from a peak to itsnearest valley divided by the time it takes to rollthrough this amplitude. This method captureswing drops that have no overshoots and wing rockthat has sinusoidal motion. A color-coded systemwas also devised for this FOM where greenrepresented values of (pP-V)max less than 50, yellowhad values between 50 and 100 and red any value

greater than 100. While still somewhat arbitrary,the various levels were established after a reviewof the results from the four models showed datafalling predominately within these three bands.

The current ratings for the same fourconfigurations shown previously in figure 32 arepresented in figure 33. Although the same rangeis used for all the models, there is no expectationthat the level of lateral activity means the same forall airplanes given their different sizes and inertias.As can be seen, the previous the low a “red”severe events for the F/A-18E have now become“green” not significant events. Also note that the“green” event occurring at a = 15.5° (fig. 32) forthe AV-8B has become a “red” event.

Roll Rate CorrelationA correlation of roll rate to roll acceleration

has been performed for each of the fourconfigurations tested during the FTR tests andthese results are presented in figure 34. All datacollected during the pitch-pause phase of the FTRtests are given for each configuration. Note that allconfigurations have been plotted to the samescales. The highest roll rates measured were forthe F-18C at angles of attack in which the flapswere off schedule. There appears to be no effectof Mach number, except for the AV-8B at M = 0.3.(Note the data with the circle symbols.)

Roll rates for the F/A-18E determined inthe wind tunnel and flight12 are shown in figure 35.A direct comparison of these data cannot be madebecause the wind tunnel model was notdynamically scaled to the aircraft, and the aircraftstability augmentation system was notrepresented in the FTR test technique. However,the trends of roll rate variation with acceleration arevery similar.

Angle of Attack CorrelationA comparison of FTR activity to flight data

for the pre-production F/A-18E is shown in figure36 for Mach numbers of 0.8 and 0.9. The datashown is where both the airplane and model hadapproximately the same flap settings and Machnumber at the time of a lateral activity event. Ascan be seen, there is good agreement betweenthe wind tunnel and flight. However, thiscorrelation only shows unacceptable lateralactivity, not the type of lateral activity. Note thatwhile FTR lateral activity is present with flaps onschedule, these results are for a model of the pre-production F/A-18E aircraft1 and are notrepresentative of the production aircraft1.

AIAA 2003-0922


7

Figures 37 and 38 shows a comparison ofangles of attack for wing rock/drop between flightand wind tunnel for the AV-8B with the 100%LERX and the 65% LERX. Flight data for the100% LERX is only available for Mach numbersgreater than 0.5, whereas flight data exists for the65% LERX at Mach numbers from about 0.22 to0.90. The ranges of wind-tunnel data presentedare for “yellow or red” conditions. As can be seen,there is excellent agreement between the flightand wind tunnel data. The most noticeabledifference between 100% and 65% LERX occursat M = 0.5 where, unfortunately, there is no flightor wind tunnel data for 100% LERX. It should benoted, that any wing drop/rock activity noted is notconsidered an AV-8B aircraft operational problem.

Roll Damping CorrelationA correlation of roll damping between

flight20 and those measured in the wind tunnel arepresented in figure 39. Roll damping is extractedfrom the roll angle time histories of the resultingmotions when the model is released from someinitial roll angle. Good agreement is noted in figure39.

RECOMMENDATIONS

Based on extensive studies of staticexperimental figures of merit in the AWS Programfor four different aircraft configurations, no onespecific static FOM consistently flagged a warningof potential lateral activity only when actual activitywas seen in the FTR experiments and,conversely, never flagged when FTR activity didnot occur. However, these studies pointed outthe importance of measuring and recording theRMS signals of the force balance. As reported byLamar23, however, many variables can serve as“relatively” dependable static FOMS when takenas a group. Included in this group should bemeasurements of wing bending for correlations tothe corresponding CFD FOMs involving half-planerolling moment15.

If the experimental static FOMs indicate apotential lateral activity, it is recommended toproceed to FTR testing. It is essential that the FTRtest program include at least three testapproaches; continuous sweep, pitch pause, androll offsets. Some further refinement of free-to-rolltest procedures are expected as further analysisof the results reaches completion.

REFERENCES

1. Hall, R. M.; and Woodson, S: Introduction tothe Abrupt Wing Stall (AWS) Program. AIAA-2003-0589, January 2003.

2 . Chambers, J. R.; and Hall, R. M.: HistoricalReview of Uncommanded Lateral-DirectionalMotions at Transonic Speeds. AIAA-2003-0590. January 2003.

3. McMillin, S. N; Hall, R. M.; and Lamar, J. E.:Understanding Abrupt Wing Stall withExperimental Methods. AIAA-2003-0591,January 2003.

4. Woodson, S.; Green, B; Chung, J.; Grove, D.;Parikh, P.; and Forsythe, J.: UnderstandingAbrupt Wing Stall (AWS) with CFD. AIAA-2003-0592, January 2003.

5 . Schuster, D.; and Byrd, J.: TransonicUnsteady Aerodynamics of the F/A-18E atConditions Promoting Abrupt Wing Stall.AIAA-2003-0593, January 2003.

6 . Forsythe, J.; and Woodson, S.: UnsteadyCFD Calculations of Abrupt Wing Stall UsingDetached- Eddy Simulation. AIAA-2003-0594, January 2003.

7 . Parikh, P; and Chung, J.: A ComputationalStudy of the AWS Characteristics for VariousFighter Jets: Part I, F/A-18E & F-16C. AIAA-2003-0746, January 2003.

8. Chung, J.; and Parikh, P.: A ComputationalStudy of the Abrupt Wing Stall (AWS)Characteristics for Various Fighter Jets: Part II,AV-8B and F/A-18C. AIAA-2003-0747,January 2003.

9. Lamar, J. E.; Capone, F. J.; and Hall, R. M.:AWS Figure of Merit (FOM) DevelopedParameters from Static, Transonic ModelTests. AIAA-2003-0745, January 2003.

10. Capone, F. J; Owens, D. B.; and Hall, R. M:Development of a Free- To- Roll TransonicTest Capability. AIAA-2003-0749. January2003.

11. Owens, B; Capone, F. J.; Hall, R. M.; Brandon,J; and Cunningham, K: Free-To-Roll Analysisof Abrupt Wing Stall on Military Aircraft atT ranson ic Speeds . AIAA-2003-0750,January 2003.

1 2 . Roesch, M; and Randall, B.: Flight TestAssessment Of Lateral Activity. AIAA-2003-0748, January 2003.

13 . Green, B; and Ott, J.: F/A-18C to E WingMorphing Study for the Abrupt Wing StallProgram. AIAA-2003-0925, January, 2003.

14 . Kokolios; A., and Cook, S.: Use of PilotedSimulation for Evaluation of Abrupt Wing StallCharacteristics. AIAA-2003-0924, January2003.

AIAA 2003-0922


8

15. Woodson, S.; Green, B.; Chung, J; Grove, D.;P a r i k h , P ; a n d Forsythe, J.:Recommendations for CFD Procedures forPredicting Abrupt Wing Stall (AWS). AIAA-2003-0923, January, 2003.

1 6 . Cook, S.; Chambers, J.; Kokolios, A.;Niewoehner, R.; Owens, B.; and Roesch, M.:An Integrated Approach to Assessment ofAbrupt Wing Stall for Advanced Aircraft. AIAA-2003-0926, January 2003.

17. Hall, R. M.; Woodson, S.; and Chambers, J.R.: Accomplishments of the AWS Programand Future Requirements. AIAA-2003-0927,January, 2003.

18. Bore, Cliff L.: Post-Stall Aerodynamics of theHarrier GR1. AGARD-CP-102 Fluid Dynamicsof Aircraft Stalling. 1972.

19 . Hwang, C.; and Pi, W.S.: “Investigation ofSteady and Fluctuating Pressures AssociatedWith the Transonic Buffeting and Wing Rockof a One-Seventh Scale Model of the F-5AAircraft.” NASA Contractor Report 3061,November 1978. NOTE: Subsequentlysummarized, reduced, and published as:“Some Observations on the Mechanism ofAircraft Wing Rock.” AIAA Paper 78-1456,AIAA Aircraft Systems and TechnologyConference. Los Angeles, CA. August 21-23,1978.

20. Stevenson, Scott W.; Holl, David; and Roman,Alan: Parameter identification of AV-8BWingborne Aerodynamics for Flight SimulatorModel Updates. AIAA-92-4506-CP, August1992.

Static Forces & Moments

Force & Moment, rms

Unsteady Forces & Moments

Wing Bending Moments

Wing Bending Moments, rms

Free-To-Roll

Wing Pressures

Unsteady Wing Pressures

Pressure Sensitive Paint

Length, in.

Wing Area, ft2

Span, ft

Weight, lbs

Inertia, s-ft2

Scale, %

81.27

5.18

4.55

490

4

15

Yes

Yes

Yes

Yes

Yes

No

No

No

No

37.25

1.33

2.07

56

0.2

6.67

Yes

Yes

Yes

Yes

Yes

No

No

No

No

39.18

1.44

2.25

55

0.2

6

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

54.99

3.2

3.34

185

1.2

8

Yes

Yes

Yes

No

No

No

No

No

No

AV- 8B F/A-18E F-16CF/A-18C

CL

Cl

Wing dropin flightRoll unsteadiness

wing rock in flight

Figure 1. Roll unsteadiness of the Harrier fromexperimental wind tunnel data (ref. 18).

Figure 2. Scope of Abrupt Wing Stall experimental investigations.

9

10 11 12 13 14 15 16 17 18

Increment2.0°1.0°0.5°

CL

α, deg

Figure 4. Definition of traditional figures of merit.

Figure 3. Effect of increment in angle of attack on test data.

CL

Cl,rms

Cl

Break

Spike

Peak

α, deg

10

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

CA

M

0.499 0.748

δte, deg

10.0 10.0

LERX, %

100 100

CLα

CL

Figure 5. Traditional figures of merit for the AV-8B.

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

Cι,rms

Cι

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,22

M

0.499 0.748

δte, deg

10.0 10.0

LERX, %

100 100

ΣFOM,22 = SQRT((Cι2 + Cι,rms

2)/ 2)

Figure 6. Alternate figures of merit for the AV-8B. 11

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,3

ΣFOM,3 = SQRT((CN,rms2 + Cι

2 + Cι,rms2)/ 3)

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΠFOM AWS ↑

ΠFOM = |CN,rms x Cι,rms x ∆Cι|

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

CA

M

0.499 0.499

δte, deg

25.0 25.0

LERX, %

100 65

CLα

CL

Figure 7. Traditional figures of merit for the AV-8B.

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

Cι,rms

Cι

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,22

M

0.499 0.499

δte, deg

25.0 25.0

LERX, %

100 65


2)/ 2)

Figure 8. Alternate figures of merit for the AV-8B. 12

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,3


2 + Cι,rms2)/ 3)

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΠFOM AWS ↑


6 8 10 12 14 16 18θ, deg

Flaps on schedule

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break

CFD CL break

Flight

θ indicated for AWS activity

θ for slope sign change

Not significant

Moderate

Severe

FTR Rate FOMNo data/effect

Flight

6 8 10 12 14 16 18θ, deg

Flaps on schedule

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break

CFD CL break



Not significant

Moderate

Severe


Figure 9. Stoplight chart for AV-8B with 100% LERX, δf = 10°, M = 0.50.


13

6 8 10 12 14 16 18θ, deg

Flaps on schedule

Flight data



Not significant

Moderate

Severe


Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break

Flight


6 8 10 12 14 16 18θ, deg

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break



Not significant

Moderate

Severe


Flaps on schedule

Figure 11. Stoplight chart for AV-8B with100% LERX, δf = 25°, M = 0.50.

14

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

CA

M

0.798 0.848 0.898

δle, deg

6 6 6

δte, deg

8.0 8.0 8.0

δa, deg

0.0 0.0 0.0

CLα

CL

Figure 13. Traditional figures of merit for the F/A-18C.

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

Cι,rms

Cι

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,22

M

0.798 0.848 0.898

δle, deg

6 6 6

δte, deg

8.0 8.0 8.0

δa, deg

0.0 0.0 0.0


2)/ 2)

Figure 14. Alternate figures of merit for the F/A-18C. 15

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,3


2 + Cι,rms2)/ 3)

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΠFOM AWS ↑


6 8 10 12 14 16 18θ, deg

M = 0.90

Flaps on schedule

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

Free-to-roll

Cl,rms peak

CL break

CFD CL break

Figure 15. Stoplight chart for F/A-18C, δle = 6°, δte = 8°, δa= 0°.

6 8 10 12 14 16 18θ, deg

M = 0.80

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

Free-to-roll

Cl,rms peak

CL break

θ indicated for AWS activityNot significant

Moderate

Severe


6 8 10 12 14 16 18θ, deg

Flaps on schedule

M = 0.85

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

Free-to-roll

Cl,rms peak

CL break

CFD CL break

16

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

CA

M

0.898 0.897 0.897

δle, deg

6.0 10.0 15.0

δte, deg

8.0 10.0 10.0

δa, deg

4 5 5

CLα

CL

Figure 16. Traditional figures of merit for the pre-production F/A-18E.

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

Cι,rms

Cι

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,22

M

0.898 0.897 0.897

δle, deg

6.0 10.0 15.0

δte, deg

8.0 10.0 10.0

δa, deg

4 5 5


2)/ 2)

Figure 17. Alternate figures of merit for the pre-production F/A-18E. 17

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,3


2 + Cι,rms2)/ 3)

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΠFOM AWS ↑


0 2 4 6 8 10 12 14 16

CLWB,rms

CLWB

α, deg

δle, deg δte, deg δa, deg 6 8 410 10 5

0 2 4 6 8 10 12 14 16

CRWB,rms

CRWB

α, deg

Figure 18. Wing bending moment characteristics for the pre-production F/A-18E, M = 0.90.

6 8 10 12 14 16 18θ, deg

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break


θ for slope change

Not significant

Moderate

Severe


Figure 19. Stoplight chart for pre-productiion F/A-18E, δle = 6°, δte = 8°, δa = 4°, M = 0.90.

18

6 8 10 12 14 16 18θ, deg

Flaps on schedule

θ indicated for AWS activityNot significant

Moderate

Severe


Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

Free-to-roll

Cl,rms peak

CL break

M = 0.90

Figure 21. Stoplight chart for pre-production F/A-18E, δle = 15°, δte = 10°, δa = 5°, M = 0.90.

6 8 10 12 14 16 18θ, deg

Flaps on schedule

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break

CFD CL break



Not significant

Moderate

Severe


M = 0.80

Figure 20. Stoplight chart for pre-production F/A-18E, δle = 10°, δte = 10°, δa= 5°, M = 0.90.

19

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

CA

M

0.798 0.898

δle, deg

10.0 10.0

δte, deg

0.0 0.0

CLα

CL

Figure 22. Traditional figures of merit for the F-16C.

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

Cι,rms

Cι

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,22

M

0.798 0.898

δle, deg

10.0 10.0

δte, deg

0.0 0.0


2)/ 2)

Figure 23. Alternate figures of merit for the F-16C. 20

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΣFOM,3


2 + Cι,rms2)/ 3)

0 2 4 6 8 10 12 14 16 18 20 22 α, deg

ΠFOM AWS ↑


6 8 10 12 14 16 18θ, deg

Flaps on schedule

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break

M = 0.90

6 8 10 12 14 16 18θ, deg Flaps on schedule

Cl spike

ΣFOM,3

ΣFOM,22

ΠFOM

CLWBCRWB

Free-to-roll

Cl,rms peak

CL break

M = 0.80


θ for slope change

Not significant

Moderate

Severe


Figure 24. Stoplight chart for F-16C, δle = 10°, δte = 0°.

Figure 25. Steady pressures for F/A-18E, δle = 10°, δte = 10°, δa= 5°, M = 0.80.

Circles, α = 8°, squares, α = 9°

21

Figure 26. Pressure coefficient time history at a single point on the wingfor a series of angles of attack, M = 0.90.

Figure 27. Comparison of PSP image to oil flow photograph.

22

6 8 10 12 14 16 18 20 22 24 θ, deg

Cm

Figure 28. Comparison of static and free-to-roll aerodynamic characteristics for the AV-8B. 100% LERX, δte = 10.0°, M = 0.50.

CA

CN

Type FTR

Static

6 8 10 12 14 16 18 20 22 24 θ, deg

CY

Cn

Cι

6 8 10 12 14 16 18 20 22 24 θ, deg

Cm


23

CA

CN

Type FTR

Static

6 8 10 12 14 16 18 20 22 24 θ, deg

CY

Cn

Cι

6 8 10 12 14 16 18 20 22 24 θ, deg

Cm


CA

CN

Type FTR

Static

6 8 10 12 14 16 18 20 22 24 θ, deg

CY

Cn

Cι

6 8 10 12 14 16 18 20 22 24 θ, deg

Cm


24

CA

CN

Type FTR

Static

6 8 10 12 14 16 18 20 22 24 θ, deg

CY

Cn

Cι

Roll Acceleration, deg/sec2

Rol

l Rat

e, d

eg/s

ec AV-8B


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec F-16C


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec F/A-18C

Figure 34. Roll rate characterisitcs for the four configurations tested. 25


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec Pre-production F/A-18E


Rol

l Rat

e, d

eg/s

ec


Rol

l Rat

e, d

eg/s

ec

0 2 4 6 8 10 12 14 16 18 20 22 θ, deg

φ max

, deg

Model AV-8B

F/A-18C F/A-18E F-16C

Flaps 25°/65% LERX

6°/8°/0° 20°/10°/0°

5°/0°

M 0.50 0.85 0.90 0.80

Figure 32. Initial ratings for a selected configuration of the models tested.

0 2 4 6 8 10 12 14 16 18 20 22 θ, deg

φ max

, deg

0 2 4 6 8 10 12 14 16 18 20 22 θ, deg

φ max

, deg

0 2 4 6 8 10 12 14 16 18 20 22 θ, deg

ppv

Model AV-8B

F/A-18C F/A-18E F-16C

Flaps 25°/65% LERX

6°/8°/0° 20°/10°/0°

5°/0°

M 0.50 0.85 0.90 0.80

Figure 33. Current ratings for a selected configuration of the models tested.

0 2 4 6 8 10 12 14 16 18 20 22 θ, deg

ppv

26

Figure 35. Comparison of free-to-roll and flight roll rate for the pre-production F/A-18E.

Max

RM

S r

oll r

ate

Max RMS roll acceleration

Free-to-Roll Flight

10

12

14

16

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00

Flight DataFTR Activity (on flap schedule)

Figure 36. Comparison of free-to-roll and flight data for the pre-production F/A-18E.

Ang

le o

f atta

ck, d

eg

Mach number

27

Figure 37. Comparison of free-to-roll and flight data for the AV-8B, 100% LERX.

10

15

20

25

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

25° Flap

10° Flap

15° Flap

100% LERX - AV-8B at extreme of flight envelope

δte, deg FTR Rate FOM 10 Moderate 10 Severe 15 Moderate 15 Severe 25 Moderate 25 Severe

Flight

Mach number

Flap schedule

Figure 38. Comparison of free-to-roll and flight data for the AV-8B, 65% LERX.

10

15

20

25

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

δte, deg FTR Rate FOM 10 Moderate 10 Severe 25 Severe

25° Flap

10° Flap

15° Flap

Mach number

Flight

Flap schedule

65% LERX - AV-8B at extreme of flight envelope

28

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0 5 10 15 20 25

FlightWind Tunnel

Unstable

Stable

o

Figure 38. Comparison of free-to-roll and flight roll damping for the AV-8B.

Angle of attack, deg

Rol

l dam

ping

aiaa 2003-0922 recommended experimental procedures …mln/ltrs-pdfs/nasa-aiaa-2003... · 2003. 1....

Documents