An Evaluation Technique for an F/A-18 Aircraft Loads Model Using F/A-18 Systems Research Aircraft Flight Data

Candida D. Olney and Heather HillebrandtNASA Dryden Flight Research CenterEdwards, California

Eric Y. ReichenbachThe Boeing Company, Phantom WorksSt. Louis, Missouri

July 2000

The NASA STI Program Office…in Profile

Since its founding, NASA has been dedicatedto the advancement of aeronautics and space science. The NASA Scientific and Technical Information (STI) Program Office plays a keypart in helping NASA maintain thisimportant role.

The NASA STI Program Office is operated byLangley Research Center, the lead center forNASA’s scientific and technical information.The NASA STI Program Office provides access to the NASA STI Database, the largest collectionof aeronautical and space science STI in theworld. The Program Office is also NASA’s institutional mechanism for disseminating theresults of its research and development activities. These results are published by NASA in theNASA STI Report Series, which includes the following report types:

• TECHNICAL PUBLICATION. Reports of completed research or a major significantphase of research that present the results of NASA programs and include extensive dataor theoretical analysis. Includes compilations of significant scientific and technical data and information deemed to be of continuing reference value. NASA’s counterpart of peer-reviewed formal professional papers but has less stringent limitations on manuscriptlength and extent of graphic presentations.

• TECHNICAL MEMORANDUM. Scientificand technical findings that are preliminary orof specialized interest, e.g., quick releasereports, working papers, and bibliographiesthat contain minimal annotation. Does notcontain extensive analysis.

• CONTRACTOR REPORT. Scientific and technical findings by NASA-sponsored contractors and grantees.

• CONFERENCE PUBLICATION. Collected papers from scientific andtechnical conferences, symposia, seminars,or other meetings sponsored or cosponsoredby NASA.

• SPECIAL PUBLICATION. Scientific,technical, or historical information fromNASA programs, projects, and mission,often concerned with subjects havingsubstantial public interest.

• TECHNICAL TRANSLATION. English- language translations of foreign scientific and technical material pertinent toNASA’s mission.

Specialized services that complement the STIProgram Office’s diverse offerings include creating custom thesauri, building customizeddatabases, organizing and publishing researchresults…even providing videos.

For more information about the NASA STIProgram Office, see the following:

• Access the NASA STI Program Home Pageat http://www.sti.nasa.gov

• E-mail your question via the Internet to help@sti.nasa.gov

• Fax your question to the NASA Access HelpDesk at (301) 621-0134

• Telephone the NASA Access Help Desk at(301) 621-0390

• Write to:NASA Access Help DeskNASA Center for AeroSpace Information7121 Standard DriveHanover, MD 21076-1320

NASA/TM-2000-209028

An Evaluation Technique for an F/A-18 Aircraft Loads Model Using F/A-18 Systems Research Aircraft Flight Data

Candida D. Olney and Heather HillebrandtNASA Dryden Flight Research CenterEdwards, California

Eric Y. ReichenbachThe Boeing Company, Phantom WorksSt. Louis, Missouri

July 2000

National Aeronautics andSpace Administration

Dryden Flight Research CenterEdwards, California 93523-0273

NOTICEUse of trade names or names of manufacturers in this document does not constitute an official endorsementof such products or manufacturers, either expressed or implied, by the National Aeronautics andSpace Administration.

Available from the following:

NASA Center for AeroSpace Information (CASI) National Technical Information Service (NTIS)7121 Standard Drive 5285 Port Royal RoadHanover, MD 21076-1320 Springfield, VA 22161-2171(301) 621-0390 (703) 487-4650

AN EVALUATION TECHNIQUE FOR AN F/A-18 AIRCRAFTLOADS MODEL USING F/A-18 SYSTEMS RESEARCH

AIRCRAFT FLIGHT DATA

Candida D. Olney and Heather HillebrandtNASA Dryden Flight Research Center

Edwards, California

Eric Y. ReichenbachThe Boeing Company, Phantom Works

St. Louis, Missouri

Abstract

A limited evaluation of the F/A-18 baselineloads model was performed on the SystemsResearch Aircraft at NASA Dryden FlightResearch Center (Edwards, California).Boeing developed the F/A-18 loads modelusing a linear aeroelastic analysis inconjunction with a flight simulator todetermine loads at discrete locations on theaircraft. This experiment was designed sothat analysis of doublets could be used toestablish aircraft aerodynamic and loadsresponse at 20 flight conditions.Instrumentation on the right outboard leadingedge flap, left aileron, and left stabilatormeasured the hinge moment so thatcomparisons could be made between in-flight-measured hinge moments and loadsmodel-predicted values at these locations.Comparisons showed that the differencebetween the loads model-predicted andin-flight-measured hinge moments was up to130 percent of the flight limit load. Astepwise regression technique was used todetermine new loads derivatives. Thesederivatives were placed in the loads model,which reduced the error to within 10 percentof the flight limit load. This paper discusses

the flight test methodology, a process fordetermining loads coefficients, and the directcomparisons of predicted and measuredhinge moments and loads coefficients.

Nomenclature

AAW active aeroelastic wing

dad differential ailerondeflection, radians

dail symmetric ailerondeflection, radians

dfilef differential inboard leadingedge flap deflection,radians

dfolef differential outboard leadingedge flap, radians

dftail differential stabilatordeflection, radians

dilef symmetric inboard leadingedge flap deflection,radians

diftef differential trailing edge flapdeflection, radians

2

dolef symmetric outboard leadingedge flap deflection,radians

drud differential rudderdeflection, radians

dstab symmetric stabilatordeflection, radians

dtef symmetric trailing edge flapdeflection, radians

FLLa flight limit load, absolutevalue

Hmx hinge moment derivative asa result ofnondimensionalizedparameter x, percent offlight limit load, absolutevalue/unit of x

Hp pressure altitude, ft

i incidence

LEF leading edge flap

Ny lateral acceleration, g

Nz vertical acceleration, g

OBES onboard excitation system

P nondimensionalized roll rate

Pdot nondimensionalized rollacceleration

psf pounds per square foot

Q nondimensionalized pitchrate

qact actual dynamic pressure,psf

Qdot nondimensionalized pitchacceleration

qtarget target dynamic pressure,psf

R nondimensionalized yawrate

Rdot nondimensionalized yawacceleration

ROLEFHMact actual right outboard leadingedge flap hinge moment,percent of flight limitload, absolute value

rudtoe symmetric rudderdeflection, radians

SRA Systems Research Aircraft

TEF trailing edge flap

α angle of attack, radians

β angle of sideslip, radians

Introduction

NASA Dryden Flight Research Center(Edwards, California) performed a limitedF/A-18 baseline loads model evaluation aspart of the active aeroelastic wing (AAW)risk reduction experiment on the SystemsResearch Aircraft (SRA).1 The ongoingAAW2 program uses an F/A-18 with wingsmodified for reduced torsional stiffness sothat tools may be developed thatconcurrently integrate control and structuraldesign to save structural weight, reduce drag,improve cruise and roll performance, andincrease fatigue life. The AAW risk reductionexperiment grew out of the programdevelopment to test techniques to be usedduring the program for identifying individualcontrol surface effectiveness and loadsderivatives and for performing flight fluttertesting. This experiment used the F/A-18SRA aircraft because this highlyinstrumented research vehicle has an easilymodified flight control system similar to theAAW aircraft. The two fundamental

3

differences between the SRA and AAWaircraft are as follows:

• The SRA wings have the same stiffnessas a production F/A-18 has, while theAAW aircraft wings have reducedtorsional stiffness.

• The SRA inboard and outboard leadingedge flaps (LEF) operate together, whilethe AAW LEFs operate separately.

Boeing developed the F/A-18 loads modelfor use with the flight simulator to determineloads at discrete locations on the aircraft.The model consists of FORTRAN code thatuses a loads database consisting of loadscoefficients with aircraft flight parameters todetermine loads at 36 discrete locations onthe aircraft. The standard F/A-18 loadsdatabase, which for this paper will be calledthe baseline database, was then modified toreflect the increased flexibility of the AAWaircraft so that it could be used in developingthe AAW control laws. The loads models forboth the AAW aircraft and the baselineF/A-18 predicted higher-than-expected loads,especially for the outboard LEF, which hasbecome the limiting factor in the AAWcontrol law development. This conservatismresults from the fact that both loads modeldatabases were obtained using a finiteelement model and a doublet-latticeaerodynamic model. Although an attemptwas made to modify the baseline databasewith loads obtained during the original flighttest of the F/A-18, only limited changescould be made because of the nature of theflight test maneuvers performed.3 Analysishas shown that high-fidelity loads andaerodynamics models are required to makethe design tools that are being developedduring the AAW program work. Thepurposes of this experiment were todetermine the level and sources of

conservatism inherent in the baselinedatabase and to develop a method formodifying the loads model database based onin-flight measured loads.

The F/A-18 loads model allows 36 discreteloads to be calculated at 20 flight conditions.During this experiment, three controlsurfaces were instrumented for hingemoment: the right outboard LEF, the leftaileron, and the left stabilator. Each of thecontrol surfaces performed doublets so thataerodynamic and load responses to thesedeflections could be determined. This paperfocuses on evaluating the right outboard LEFhinge moment from flight test in comparisonwith the loads model-predicted hingemoment. The flight test methodology isdiscussed; a process for determining loadscoefficients is discussed; and directcomparisons of predicted and measuredhinge moments and loads coefficients arepresented.

Note that use of trade names or names ofmanufacturers in this document does notconstitute an official endorsement of suchproducts or manufacturers, either expressedor implied, by the National Aeronautics andSpace Administration.

Flight Test Methodology

The design of the AAW risk reductionexperiment facilitates the determination ofindividual control surface effects on theoverall aerodynamics of the F/A-18 aircraft.A series of doublet maneuvers wereperformed at 20 different flight conditions,during which the onboard excitation system(OBES)4 commanded a series of independentcontrol surface deflections. In addition to theaerodynamic parameters, three controlsurfaces—the right outboard LEF, the left

4

stabilator and the left aileron—wereinstrumented and calibrated to measure hingemoments. The hinge moments weremonitored in real time in the control roomduring each flight. The following will discusseach component of the flight testmethodology that was used in thisexperiment including flight conditions, OBESmaneuvers, instrumentation, and controlroom monitoring.

This experiment consisted of performingdoublet maneuvers at 20 different flightconditions or test points. The flightconditions included a Mach range of 0.85 to1.3 and an altitude range of 5,000 to25,000 ft as seen in figure 1. After obtaininga specified flight condition and trimming theaircraft for level flight, the pilot entered andengaged the correct OBES maneuver to beflown. Each OBES maneuver consisted of aseries of longitudinal or lateral-directionaldoublets. The OBES commanded the typeand size of doublet that was to beperformed.

Figure 2 shows the order and direction ofthese doublets. Figure 2(a) shows a typicallongitudinal maneuver consisting of a seriesof single-surface doublets for the left control

surfaces; for simplicity, the right controlsurface deflections, which were equal inmagnitude and direction, are not shown. Eachlongitudinal maneuver consists of asymmetric LEF doublet followed by asymmetric trailing edge flap (TEF) doublet, asymmetric aileron doublet, and a symmetricstabilator doublet. Between each doublet,was a 5-second (sec) pause. Figure 2(b)shows a typical lateral-directional maneuver;again, the right control surface deflections,which were equal in magnitude and oppositein direction, are not shown. Each lateral-directional maneuver consists of an anti-symmetric rudder doublet followed by ananti-symmetric LEF doublet, an anti-symmetric TEF doublet, an anti-symmetricaileron doublet, and an anti-symmetricstabilator doublet. Either small, medium, orlarge longitudinal and lateral-directionaldoublets were flown at each flight condition.Doublet size was based on expected aircraftaerodynamic and load response. Figure 1shows how the doublet sizes variedthroughout the flight envelope. Thefollowing table lists the type of doublet withits corresponding control surface deflectionsin degrees for each of the six maneuvers.

Doublet control surface deflections.

Maneuver Rudder LEF TEF Aileron Stabilator

Large lateral-directional ±4 ±3 ±5 ±6 ±3

Large longitudinal N/A +3 +5 ±6 ±1

Medium lateral-directional ±4 ±2 ±4 ±5 ±3

Medium longitudinal N/A +2 +4 ±5 ±1

Small lateral-directional ±4 ±1 ±3 ±4 ±3

Small longitudinal N/A +1 +3 ±4 ±1

5

Test points were performed in order ofincreasing dynamic pressure so that themaneuvers could be evaluated for loadsbefore proceeding to higher loadingconditions. As previously mentioned, eachdoublet input was separated by a 5-secpause to verify that each control surfacereturned to its trim deflection. This approachenabled safe conduct of the flight test whileobtaining hinge moment data at everyrequired test condition in both thelongitudinal and lateral directions by varyingone parameter at a time.

As stated previously, three control surfaceson the SRA aircraft were instrumented andcalibrated for hinge moment: the rightoutboard LEF, the left aileron, and the leftstabilator. Figure 3 illustrates the threelocations on the aircraft that wereinstrumented for this experiment. Both theleft aileron and left stabilator rod-ends wereinstrumented with sixteen strain gagesconfigured into four, four-active-arm bendingbridges. Figure 4 shows the placement of thebridges on the left stabilator rod-end.Figure 4(a) shows the locations of theprimary and spare compression bridges,while figure 4(b) shows the locations of theprimary and spare tension bridges. The leftaileron rod-end strain gage placement wassimilar to that of the left stabilator rod-end.The strain gage outputs were calibrated withload by placing the stabilator and aileron rod-ends in a load test machine and applyingknown tensile and compressive loads.

The right outboard LEF was instrumentedwith twelve strain gages configured intothree, four-active-arm bending bridges on thelugs of the transmission. Figure 5 shows thegage placement on the right outboard LEFtransmission. After the transmission and flapwere reinstalled on the aircraft, these straingages were calibrated by applying known

compressive loads to the right outboard LEFusing a hydraulic jack, load cell, and loadpad. Before each flight, a ground check wasperformed in which load was applied toverify that all of the instrumentation wasfunctioning. Phasing maneuvers were alsoperformed in-flight to ensure that theinstrumentation was functioning correctly.

During flight testing, data from all threeinstrumented control surfaces weretelemetered to the control room for real-timemonitoring. Among the parameters observedduring flight were the right and left outboardLEF positions, the left aileron and the leftstabilator hinge moments, the right outboardLEF hinge moment, and outputs from allthree strain gage bridges on the rightoutboard LEF. All listed parameters wereoutput to and viewed on an eight-channelstrip chart. In addition, all control surfacedeflections, vertical acceleration (Nz), angleof attack (α), angle of sideslip (β), Machnumber, altitude (Hp), pitch rate (Q), rollrate (P), and yaw rate (R) were monitored inreal time in the control room. Monitoring thecontrol surface deflection allowed the testengineer to call directly to the pilot toterminate the maneuver if a control surfaceactuator were to stall during high dynamicpressure test points. If a control surfaceactuator had stalled, the maneuver wouldhave been unusable for analysis. In addition,a stalled control surface actuator could havecaused a structural overload conditionbecause of the nature of the control surfacedoublets. Monitoring also included observingreal-time measured hinge moments andpredicted hinge moments as percentages ofthe absolute value of the flight limit load(designated percentage of FLLa). Thepredicted hinge moments were calculated inreal time by the loads model using basicaircraft parameters and flight conditions

6

telemetered from the aircraft. In this way,measured and predicted hinge moments forall three instrumented control surfaces werecompared during flight.

The control room display incorporatedwarnings to indicate the percentage of theabsolute value of the flight limit load, thatwas reached during flight. Had the measuredhinge moment reached 80 percent of FLLa,the value displayed would have turnedyellow, and if it had reached 100 percent ofFLLa, the value displayed would have turnedred. In addition to these safety-of-flightwarnings, there was a research requirementfor a warning to flash if a measured hingemoment exceeded its predicted hinge momentby 10 percent of FLLa so that the testengineer could call to terminate the testpoint.

Data Processing Methodology

Data processing for this experimentconsisted of several stages: The first stagewas the preliminary analysis in which directcomparisons of the measured and predictedhinge moments were made. After this stagewas complete, time histories of eachmaneuver were processed through a stepwiseregression technique5 that had beenimplemented in MATLAB by Dr. E. A.Morelli at NASA Langley Research Center(Hampton, Virginia). This regressiontechnique related each of the aircraftparameters to the measured hinge moment.The most highly correlated parameter wasretained in the model, and the estimated loadresulting from that parameter was removedfrom the overall hinge moment. The nextmost highly correlated parameter was thenselected, and the estimated load resultingfrom that parameter was removed from themeasured hinge moment. This process was

repeated for each selected aircraft parameter.However, after each parameter was selectedfor retention in the regression model, thepreviously selected parameters werereevaluated to ensure that they were stillsignificant in predicting the overall load. Thisanalysis stage established loads coefficients,which will be called regressed coefficients.The regressed coefficients then werecompared with the coefficients contained inthe baseline database, which will be calledbaseline coefficients.

The final stage of the analysis consisted ofreplacing the baseline coefficients with theregressed coefficients. The flight data werethen rerun through the loads model, and thepost-regression hinge moment wascalculated. The results of using thistechnique on the right outboard LEF hingemoment is discussed in the following section.The following paragraphs describe the firsttwo stages of this procedure in greater detail.

First Stage

As previously mentioned, the first stage ofthe analysis consisted of direct comparisonsof the measured and predicted hingemoments. This task was performed bycreating time history plots of the hingemoment data. The histories allowed theoverall conservatism to be evaluated at eachtest point and insight to be gained as to thesources of the conservatism in the predictedhinge moment. After the overallconservatism was established, the sources ofthe differences in the predicted and measuredhinge moments were established, which wasaccomplished using the stepwise regressiontechnique.

7

Second Stage

The second stage of the analysis was to usethe stepwise regression technique todetermine the loads derivatives to beincluded in the baseline database. To explainthe process used to determine the loadscoefficients based on the flight data, it isimportant to first understand how the loadsmodel predicts the load for a particular flightcondition. The baseline database consists ofloads derivatives that are used in the loadsmodel to calculate load at a particularlocation on the aircraft. The loads model cancalculate 36 discrete loads on the F/A-18 asseen in figure 6. These loads include controlsurface hinge moments; wing root and wingfold shear, bending, and torque; and thehorizontal and vertical tail shear, bending,and torque for each side of the aircraft.

The loads model determines the range ofMach number and altitude in which theaircraft is flying. This range establisheswhich set of loads derivatives to use at thattest point. These derivatives are then used tocalculate the load at the target Mach numberand altitude combination. The load iscalculated by multiplying the derivatives bytheir respective flight parameter and addingthem together to determine the overall load atthat location.

During flight test, it is difficult to establishand maintain an exact flight condition.Therefore, it was necessary to calculate theactual load based on the load at the targetMach number and altitude. To calculate theactual load, the aerodynamic contributors tothe overall load are multiplied by the ratio ofthe actual dynamic pressure (qact) to thetarget dynamic pressure (qtarget). Thecomplete computation is shown inequation (1) for the determination of theright outboard LEF hinge moment

(ROLEFHMact).

ROLEFHMact = (qact/qtarget)[Hm i(i)+ Hmα(α) + HmQ(Q)

+ Hmdilef(dilef)+ Hmdolef(dolef)+ Hmdtef(dtef)+ Hmdail(dail)+ Hmrudtoe(rudtoe)+ Hmdstab(dstab)– {Hmβ(β) + HmP(P)+ HmR(R) (1)

+ Hmdfilef(dfilef)+ Hmdfolef(dfolef)+ Hmdiftef(diftef)+ Hmdad(dad)+ Hmdrud(drud)+ Hmdftail(dftail)}]+ [HmNz(Nz)+ HmQdot(Qdot)– {HmNy(Ny)+ HmPdot(Pdot)+ HmRdot(Rdot)}]

Because of the nature of the database, it isrequired to subtract the lateral-directionalterms from the longitudinal components forthe right wing as shown in equation (1).

Prior to calculating the loads derivatives, itwas necessary to modify the flight data sothat the aircraft parameters match the loadsmodel-expected input parameters. Forexample, the control surface symmetric anddifferential deflections between the right andleft wings had to be calculated from themeasured control surface deflections. Inaddition, all of the aircraft parameters werenondimensionalized, and the aerodynamicloads contributors were multiplied by theratio of the actual dynamic pressure to thetarget dynamic pressure. These modifiedparameters were then loaded into the

8

stepwise regression program along with themeasured hinge moment for each maneuver ateach flight condition. Each longitudinal andlateral-directional maneuver was individuallyanalyzed.

The first step in this regression techniquecalculated an intercorrelation matrix whichestablished any parameter interdependencies.Highly correlated parameters were evaluated,and the dependent parameter was removedfrom the regression model. Because thedoublet maneuvers consist of independentcontrol surface deflections, theintercorrelations are relatively small.However, rudder deflections were eliminatedfrom the longitudinal regression becausethere were no symmetric rudder doublets,thereby causing a gap in the database.Because the contribution of the symmetricrudder deflection to the overall outboard LEFhinge moment was small, however, the errorintroduced in this case was considerednegligible.

After the initial selection of parameters wascomplete, the second step was to determinethe correlation of the flight parameters withthe measured hinge moment. The mosthighly correlated parameter is selected and,using least squares, the loads coefficient forthat parameter is determined. After removingthe load caused by that parameter from themodel, the loads coefficient for the next mosthighly correlated parameter is determined,updating and reevaluating the significance ofalready calculated derivatives as required toobtain a best-fit solution.

The final selection of parameters requiredmultiple runs of the regression technique tobe performed, in which different parameterswere retained in the model. The criteria forretaining parameters were as follows:(1) Primary parameters had to contribute to

the overall load and (2) secondaryparameters had to contribute to the loadwithout adversely affecting the primaryparameter derivatives. The regressiontechnique was applied to each of thelongitudinal and lateral-directionalmaneuvers, and new loads derivatives wereestablished.

Flight Test Results

This section compares the measured rightoutboard LEF hinge moment with the loadsmodel-predicted hinge moment andrespective loads derivatives. All of the loadsresults in this section are given as apercentage of the absolute value of the flightlimit load. The first stage of this analysisconsisted of directly comparing the in-flight-measured and loads model-predicted hingemoments, which are called measured hingemoments and predicted hinge moments,respectively. Comparison of the measuredand predicted hinge moments showed thatthe loads model was calculating, as expected,a conservative estimate of the actual hingemoments for both longitudinal and lateral-directional maneuvers in both the subsonicand supersonic segments of the flight regime.

For subsonic flight conditions, on average,the loads model predicted the right outboardLEF hinge moment 70 percent of FLLa

higher than that measured during trimmedflight. As the doublets were performed,however, the difference between thepredicted and measured loads increased. Asseen in figure 7(a), which depicts a series oflongitudinal doublets at Mach 0.95 and analtitude of 10,000 ft, the load differenceincreased to as much as 130 percent of FLLa

during the LEF deflection. The influence ofthe various control surface deflections on theoverall load can be seen easily in figure 7(b),

9

which shows the same maneuver with thetrimmed loads for both the measured andpredicted hinge moments removed. As theLEFs deflect, the difference between thechanges in the load for the measured andpredicted hinge moments was approximately65 percent of FLLa.

The next control surface deflections were theTEFs. This doublet produced a difference ofonly 10 percent of FLLa, therefore showingthat the loads derivatives associated with theTEF doublet are not as conservative as theLEF doublet. The aileron deflectionsproduced a difference of 60 percent of FLLa,and the stabilator doublets caused adifference of almost 70 percent of FLLa.Figures 8(a) and 8(b) were produced usingthe same procedure to look at the lateral-directional doublets for the same flightcondition. Figure 8(a) shows a similardifference in load when looking purely at theload resulting from trimmed flight.

Figure 8(b) then looks at the changes in theload resulting from lateral-directional controlsurface deflections when the trimmed load isremoved. The change in load resulting fromthe rudder deflections differs by only about5 percent of FLLa. The LEF deflections,however, cause a difference of as much as65 percent of FLLa. The TEF deflectionshows a difference once again of less than5 percent of FLLa, and the aileron caused adifference of approximately 10 percent ofFLLa. The stabilator deflection, however,caused a hinge moment difference ofapproximately the same magnitude, but inthe opposite direction from the measuredhinge moment.

With these observations, it was necessary toperform the second stage of the analysis.The second stage used the stepwiseregression technique to determine the actual

contributions of each input parameter so thatboth the flight data regressed and baselineloads derivatives could be compared. Asmentioned in the previous section, theparameters to be maintained within theregression model had to be determined. First,to determine which parameters to retainwithin the regression model, theintercorrelation of the flight parameters wasevaluated. Because of the independent natureof the control surface deflections, theintercorrelations were relatively low. Second,the significant parameters had to bedetermined. For the longitudinal maneuvers,only symmetric parameters were retained atthis stage, excluding the symmetric rudderdeflection because no symmetric rudderdoublets were performed. For the lateral-directional maneuvers, only the anti-symmetric terms were retained in theregression model.

Final selection required that the stepwiseregression technique be used to evaluate theaddition of each of these terms in theirrespective regression models. Figure 9displays the results of this evaluation for theMach 0.95 and an altitude of 10,000 ft forlongitudinal doublets. This evaluation wasmade in six steps. The first step was toinclude only the LEF symmetric deflection.Second, all of the control surfaces wereincluded and the resulting regressed hingemoment was compared with the regressedhinge moment from the case in which onlythe LEF deflection was included and themeasured hinge moment (fig. 9(a)). The loadwas more accurately estimated when allcontrol surfaces, except for the rudder, wereincluded than when only the LEF was usedin the regression model. The third step wasto include angle of attack and the fourth toinclude pitch rate in the regression model.The inclusion of each of these parameters

10

contributed to the regressed hinge momentwithout adversely affecting the derivativesbeing calculated (fig. 9(b)). Figure 9(c) showsthe final two steps which include the verticalacceleration and then the pitch acceleration.As each parameter was included, theresulting regressed hinge moment was able tomore accurately estimate the actual measuredhinge moment without adversely affectingthe primary control surface parameters.Therefore, all of these parameters wereretained within the regressor model. For thelongitudinal case, the final equation was asfollows:

ROLEFHMact = (qact/qtarget)[Hm i(i)

+ Hmα(α) + HmQ(Q)+ Hmdilef(dilef)+ Hmdolef(dolef)+ Hmdtef(dtef) (2)+ Hmdail(dail)+ Hmdstab(dstab)]+ [HmNz(Nz)+ HmQdot(Qdot)]

The same procedure was applied for thelateral-directional maneuvers, and thefollowing equation was established foranalysis of the lateral-directional doublets:

ROLEFHMact = (qact/qtarget)[Hm i(i)

– {Hmβ(β) + HmP(P)+ HmR(R)+ Hmdfilef(dfilef)+ Hmdfolef(dfolef)+ Hmdiftef(diftef) (3)+ Hmdad(dad)+ Hmdrud(drud)+ Hmdftail(dftail)}]– [HmNy(Ny)+ HmPdot(Pdot)+ HmRdot(Rdot)]

After the required parameters wereestablished, the loads derivatives were

obtained for this flight condition.Figure 10(a) shows the baseline database andthe regressed coefficients for the trimmedflight condition, which corresponds withHm i, at Mach 0.95. The exact value of thecoefficient to be used was determined bylooking at the trends and consistency of thedata. Figure 10(b) shows the inboard LEFregressed and baseline database coefficientsat Mach 0.95. Previously, direct comparisonhad estimated that the hinge moment on theright outboard LEF was overpredicted by asmuch as 130 percent of FLLa during the LEFdeflections. Figure 10(b) demonstrates thatone source of this conservatism is the loadassociated with the deflection of the inboardLEF by showing the difference in theregressed and the baseline databasederivatives resulting from inboard LEFdeflection.

After derivatives were obtained from boththe longitudinal and lateral-directionaldoublets, the outboard LEF derivatives werereplaced within the loads model and theflight maneuvers were rerun through themodel. Figure 11(a) shows the regressedhinge moment compared with the measuredand predicted hinge moments for thelongitudinal maneuver, and figure 11(b)contains the regressed hinge moment alongwith the in-flight measured and predictedhinge moments for the lateral-directionalmaneuver. The stepwise regression techniquecould calculate a hinge moment that waswithin 10 percent of FLLa of the measuredhinge moment, in contrast to the 70 to130 percent of FLLa difference from thebaseline database derivatives.

The flight regime also included numeroussupersonic test points. This section usesdata collected at Mach 1.1 and an altitude of10,000 ft to demonstrate the use of these

11

techniques when evaluating flight data. Thiscondition was chosen because the loadsmodel predicted the highest hinge momentsat that test condition. Figure 12(a) shows theoverall predicted and measured hingemoments for a longitudinal maneuver at thiscondition. For this condition, the differencein the measured and predicted hingemoments due to trimmed flight isapproximately 30 percent of FLLa.

Figure 12(b) demonstrates the differences inthe load resulting from the control surfacedeflections after removing the load caused bytrimmed flight. When the LEF is deflected,the difference in the measured and predictedloads is approximately 55 percent of FLLa,while the TEF and aileron deflections cause adifference within 10 percent of FLLa, and thestabilator doublet cause a difference ofapproximately 50 percent of FLLa.

The hinge moments resulting from thelateral-directional doublets at this conditionare in figure 13(a), with their respectivedeltas shown in figure 13(b). The trimmedload, as estimated in figure 13(a), agrees withthat seen during the longitudinal maneuver.The delta loads, as shown in figure 13(b),establish that the differences between thepredicted and measured hinge moments areapproximately 10 percent of FLLa during therudder deflection, 70 percent of FLLa duringthe LEF deflections, 5 percent of FLLa

during the TEF deflections, and 10 percentof FLLa during the aileron deflections.However, the difference during the stabilatordeflections is essentially zero.

The stepwise regression was then used toestablish regressed loads derivatives.Figure 14(a) compares the right outboardLEF hinge moment calculated using theregressed derivatives with the measured andoriginal predicted hinge moments for the

longitudinal maneuver. The regressedderivatives cause the measured and regressedhinge moments to agree to within 2 percentof FLLa in contrast to the 30 to 80 percentof FLLa shown previously with the originalpredicted hinge moment. Figure 14(b)displays a reduction in the differencebetween the predicted and measured hingemoment, from up to 70 percent to less than2 percent of FLLa when the regressedderivatives are used to calculate hingemoment, and the measured, regressed, andpredicted hinge moment time histories areplotted for the lateral-directional maneuver.

Conclusions

The active aeroelastic wing risk reductionexperiment on the F/A-18 SRA aircraftsuccessfully established the conservatisminherent in the baseline F/A-18 loads modelusing the outboard leading edge flap (LEF)hinge moment as an example. The loadsmodel was shown to be overall moreconservative within the subsonic regime thanit was in the supersonic regime. Thecomparisons in this experiment also showedthat the main sources of the conservativeprediction were included in the trimmedflight load and during the LEF deflections forthe right outboard LEF. In the subsonicregime, the longitudinal maneuversoverpredicted the hinge moment by 70 to130 percent of flight limit load (FLLa), withthe lateral-directional maneuvers being lessconservative. In the supersonic regime theloads model overpredicted the right outboardLEF hinge moment by 30 to 100 percent ofFLLa. The lateral-directional maneuvers weremore conservative than the longitudinalexcept during the LEF deflections. Thestepwise regression technique was able toproduce new loads derivatives that couldpredict the hinge moment within 10 percent

12

of FLLa subsonically and 2 percent of FLLa

supersonically.

References

1. Sitz, Joel R., “F-18 Systems ResearchAircraft Facility,” SAE Technical PaperSeries 922063, 1992.

2. Pendleton, Ed, Denis Bessette, PeteField, Gerry Miller, and Kenneth Griffin,“The Active Aeroelastic Wing FlightResearch Program,” AIAA 98-1972,1998.

3. Heinemann, D. A., “F/A-18Demonstration Data Report: Structural,”

vol. 4 rev. B, May 1, 1984. (Prepared byMcDonnell-Douglas Corporation for theU. S. Department of Energy.)

4. Regenie, Victoria A., Michael Earls,Jeanette Le, and Michael Thomson,Experience with ADA on the F-18 HighAlpha Research Vehicle Flight TestProgram, NASA TM-104259,Oct. 1992.

5. Draper, N. R. and H. Smith, AppliedRegression Analysis, John Wiley & Sons,Inc., 1966.

Figure 1. Flight envelope with doublet sizes.

13

(a) Large longitudinal doublets.

(b) Large lateral-directional doublets.

Figure 2. Control surface deflections over time.

14

Figure 3. Measured control surface locations (positive directions shown).

(a) Compression bridges

Figure 4. Left stabilator rod-end with strain gage bridges installed.

15

(b) Tension bridges

Figure 4. Concluded.

Figure 5. Right outboard leading edge flap transmission with strain gage bridges installed.

16

Figure 6. Loads model-calculated loads locations (positive direction shown).

(a) Overall hinge moment.

Figure 7. Hinge moments at Mach 0.95 and at an altitude of 10,000 ft.

17

(b) With trimmed load removed.

Figure 7. Concluded.

(a) Overall hinge moment.

Figure 8. Hinge moments resulting from large lateral-directionaldoublets at Mach 0.95 and an altitude of 10,000 ft.

18

(b) With trimmed load removed.

Figure 8. Concluded.

(a) Hinge moments regressed with outboard leading edge flap and all control surfaces.

Figure 9. Measured and regressed hinge moments as a function of timeat Mach 0.95 and at an altitude of 10,000 ft.

19

(b) Regressed hinge moments with angle of attack and pitch rate added.

(c) Regressed hinge moments with vertical and pitch accelerations added.

Figure 9. Concluded.

20

(a) Trimmed flight loads derivatives.

(b) Loads derivatives caused by symmetric inboard leading edge flap deflection.

Figure 10. Loads derivatives as a function of dynamic pressure at Mach 0.95.

21

(a) As a result of large longitudinal doublets.

(b) As a result of large lateral-directional doublets.

Figure 11. Hinge moment as a function of time at Mach 0.95 and at an altitude of 10,000 ft.

22

(a) Overall hinge moment.

(b) With trimmed load removed.

Figure 12. Hinge moment as a result of small longitudinal doublets at Mach 1.1 and at an altitude of 10,000 ft.

23

(a) Overall hinge moment.

(b) With trimmed load removed.

Figure 13. Hinge moment as a result of small lateral-directional doublets at Mach 1.1 and at an altitude of 10,000 ft.

24

(a) As a result of small longitudinal doublets.

(b) As a result of small lateral-directional doublets.

Figure 14. Hinge moment as a function of time at Mach 1.1 and at an altitude of 10,000 ft.

REPORT DOCUMENTATION PAGE Form ApprovedOMB No. 0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington,VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.

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

4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

6. AUTHOR(S)

8. PERFORMING ORGANIZATION REPORT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING AGENCY REPORT NUMBER

11. SUPPLEMENTARY NOTES

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

13. ABSTRACT (Maximum 200 words)

14. SUBJECT TERMS 15. NUMBER OF PAGES

16. PRICE CODE

17. SECURITY CLASSIFICATION OF REPORT

18. SECURITY CLASSIFICATION OF THIS PAGE

19. SECURITY CLASSIFICATION OF ABSTRACT

20. LIMITATION OF ABSTRACT

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)Prescribed by ANSI Std. Z39-18298-102

An Evaluation Technique for an F/A-18 Aircraft Loads Model UsingF/A-18 Systems Research Aircraft Flight Data

WU 529-61-14-E8-14-00-AAW

Candida D. Olney, Heather Hillebrandt, and Eric Y. Reichenbach

NASA Dryden Flight Research CenterP.O. Box 273Edwards, California 93523-0273

H-2423

National Aeronautics and Space AdministrationWashington, DC 20546-0001 NASA/TM-2000-209028

A limited evaluation of the F/A-18 baseline loads model was performed on the Systems Research Aircraft atNASA Dryden Flight Research Center (Edwards, California). Boeing developed the F/A-18 loads model usinga linear aeroelastic analysis in conjunction with a flight simulator to determine loads at discrete locations onthe aircraft. This experiment was designed so that analysis of doublets could be used to establish aircraftaerodynamic and loads response at 20 flight conditions. Instrumentation on the right outboard leading edgeflap, left aileron, and left stabilator measured the hinge moment so that comparisons could be made betweenin-flight-measured hinge moments and loads model-predicted values at these locations. Comparisons showedthat the difference between the loads model-predicted and in-flight-measured hinge moments was up to 130percent of the flight limit load. A stepwise regression technique was used to determine new loads derivatives.These derivatives were placed in the loads model, which reduced the error to within 10 percent of the flightlimit load. This paper discusses the flight test methodology, a process for determining loads coefficients, andthe direct comparisons of predicted and measured hinge moments and loads coefficients.

Active aeroelastic wing, F/A-18, Flight test, Loads, Simulation

A03

30

Unclassified Unclassified Unclassified Unlimited

July 2000 Technical Memorandum

Prepared for the 31st SFTE Annual Symposium 2000, Torino, Italy, September 18–22, 2000. Candida D. Olney and Heather Hillebrandt, NASA Dryden Flight Research Center, Edwards, California, and Eric Y. Reichenbach, The Boeing Company, Phantom Works, St. Louis, Missouri.

Unclassified—UnlimitedSubject Category 05

This report is available at http://www.dfrc.nasa.gov/DTRS/