NASA NACA-TN-3554-1956 A preliminary investigation of the effects of frequency and amplitude on the rolling derivatives of an unswept-wing model oscillating in roll《频率和振幅对非扫掠机翼模型滚转.pdf

1、.JNATIONALADVISORYCOMMITIEEFOR AERONAUTICSA PREIiUMtNARYTECHNICAL NOTE 3554LINVESTIGATION OF THE EFFECTS OF FREQUENCYAND AMPLITUDE ON THE ROLIXNG DERIVATIVES OF ANUNSWEPT-WING MODEL 0SC!ILIJM71NG IN ROLLBy Lewis R. Fisher, Jacob H. Liechtenstein,and Katherine D. WilliamsLangley Aeronautical Laborato

2、ry-.Langley va9WashingtonJanuary 1956I., .-d1L -J-.= -”- . . . . .- . . . . . . Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-1P.- -.-_TECHLIBRARYKAFB,NMNMIONALADVISORYCOMMITTEE,FORAERONAUTICS lug!lll!lHl#BullHluDObbb7S.WHNICAL NOTE3554A PRELIMINAR

3、YINVESTIGATIONOF THEEFFECTSOF FREQUENCYANDAMPL13!UDEON THEROLLINGDERIVATIVESOFANUNSWETT-WINGMODELOSCILLATINGINROLLBy LewisR. Fisher,JacobandKatherineD.,SUMMARYH. Liechtenstein,WilliamsA modelwithseparablewingandtailassemblywas oscillatedinroll through a rangeof frequenciesandamplitudesof oscillation

4、foran angleof attackof0 andat onefYequencyandamplitudefortwohigheranglesof attackin orderto determinetheeffectsof theunsteadymotionon therollingstabilityderivativesof themdel anditscomponents.A variationof frequencyor amplitudeof oscillationin therangecoveredat an angleof attackof 0 hadno importsnte

5、ffecton either.theyawingmomentduetorollingor thedampinginrollforthisunswept-wingairplaneconfiguration”.The onlyappreciablevalueofyawingmomentduetorollingwas shownby thefuselage-tailcombination.Thisconfigurationexperienceda reductioninmagnitudeof thederivativeaseitherthefrequencyor theamplitudeof the

6、oscillationincreased. .Thevaluesof therolMng derivativesobtainedby oscillationwereconsistentwiththevaluesmeasuredby meansof conventionalrolling-flowtestsat an sngleof attackof 0. Forthemodelwiththewingata highangleof attack,theoscillatoryyawingmoiuentdueto rollingwas differentfromthatobtainedunderst

7、eady-stateconditions.INTRODUCTIONAs partof a continuinginvestigationof theeffectsofunsteadymotionon thelateralstabilityderivativesof airplanemodels,testsweremadein theLangleystabilitytunnelat low speedsto determinetheeffectsof freqyencyandamplitudeon theyawingmomentduetorollinnandthedsmpinginrollfor

8、an unswept-wingairplanemodel. Thesetests,whichwerepreliminaryin nature,involvedtheforcedoscillationinrollof themodelaboutitslongitudinalwindsxisthrougha rangeofI- - - -. - -. .-. - - - - - . . Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.-. . . .

9、 . - . 2 NACATN 3554frequenciesandamplitudesofmotion. Theresul% obtainedwerepri-marilyforan singleof attackof OO;however,someresultsforhigheranglesof attacksrepresented.Steady-statederivativesweremeasuredby meansof testsmadewiththemcdelstationaryinrollingflowandwiththemodelrollingsteadilyat severalr

10、otaryvelocitiesin straightflow. Thesesteadyresultsareregardedas zero-fkequencyoscillationdataandformthebasisfora comparisonof theunsteady-stateandthesteady-staterollingderivatives.Theoreticalvaluesforthe steady-staterollingderivativeswerealsousedfor comparisonwiththeexperimentaldata.Themodelusedin t

11、hesetestshada wingof aspectratio6 sadatailassenibly,eitherofwhichcouldbe separatedfromthefuselage.Testswereconductedforthefuselagealone,iPorthewing-offconfig-uration,forthetail-offconfiguration,andforthe completemcdel.The contributionof theverticaltailto thedampinginyaw of a modelof similarconfigura

12、tionmeasuredduringfreeoscillationinyaw isreportedinreference1 andduringforcedoscillationinyaw inreference2.The datasentedinthe.4SYMMLS-.arereferredto the stibiliw systemof axesandme pre-formof coefficientsof theforcesandmomentsaboutapointwhichcorrespondsto thenormallocationof thequarter-chordpointof

13、 thewingmeanaerodynamicchordof thetideltested. (Seefig.1.) The coefficientsand symbolsusedhereinaredefinedas follows:2A aspectratio, .5b Wing Spanj ft”% dm.g coefficient,*% liftcoefficient,*.Cz rolling-momentcoefficient, thisphotographshowsa typicalmotionof themodelduringoscillation.Theamplitudeof t

14、herollingmotionwas variedby adjustingthe eccentricityof the crankat theflywheel.Thefrequencyof oscillationwas variedby a speedcontrolthatregulatedthevoltageto thedrivingmotor.For thosetestsinwhichthemodelwas forcedto rollsteadily,thecrankmechanismwasreplaced.bya V-beltandpulleysystemwithonepulleyat

15、thecenterof thefwheel andtheotherat themodelsupport. (Seefig.6.) Ming thesetests,itwasnecessarytobalancethe systemaboutitsaxisofrotationby meansof the counterweightshownin figure6.In orderto eliminatetheuse of slipringsin the steady-roingtestsof themodel,the strain-gagewires(fig.6)wereextendeddown-s

16、treamabout feetwheretheyweretiedto a lengthof shockchordbeforebeingledoutsidethetestsection.Theremainingendof theshockchordwas tiedto a downstreamtunnelsupport.As the stingrevolvedandthe strain-gagewirestwisted,the shockchordstretchedto keepthetensioninthewiressmall. Thewiresweredisconnectedtheother

17、steady-statetestsinvolvedrollingtheairflowpastthe stationerymodel. In thelattermethod,momentsweremeasuredbothby the strain-gagebalancewhilethemodelwas supportedas in theoscillationtestsandby a mechanicalsix-componentwind-tunnelbalancewhilethemodelwasmountedon a single-strutsupport.The secondof these

18、is the stand-ardprocedureemployedin theLangleystabilitytunnelformeasuringtherolling-stabilityderivativesofmodelsandis describedinreference3.TestConditionsAll testswereconductedat a dynamicpressureof 24.9poundspersqparefootwhichcorrespondsto a free-streamvelocityof 145 feetpersecond(understandardcond

19、itions),a Reynoldsnuniberof442,000basedon thewingmeanaerodynamicchord,anda Machnuxriberof 0.13. Themodelwas testedat anglesof attackof Oo,4, and8by boththeoscillationandthe steady-statetestprocedures;however,becausethereductionoftheoscillographdataprovedto be an extremelylaboriousprocess,itwas deeme

20、dadvisableto restrictthe scopeof theinvestigationto thefollowingrepresentativecases:Therangeofcommonlya, deg f, Cps $.,dego o.5to 4.0 ?51 *5 to tioO, 4)0and 8 1 $5frequenciesof theoscillationtestswerechosenso thatthethereduced-frequencyparameter encompassedtherange2Vencounteredin thelateraloscillati

21、onsof airpbnes. This- . _ - - - -Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.-. _ ._8 NACATN3554.-parametervariedfrom = 0.030 to = 0.243. Theactualfiequen.tiesof oscillationwere0.5,1.0,1.5,2.0,3.0,and4.0 cyclespersecond. The largeamplitudesof os

22、cillationwere chosenpurposel.ytoinsurethattheyawingmomentwouldbe ofmeasurablemagnitude.TheseamplitudeswereHo, tlOO,t15,andii?(l”.Forthetestsinwhichthemciielwasrolledat constentvelocity,the circularvelocitieswere0, tO.50,M.75, andfl.00revolutionsper= of 0, -K).030, to.046, anati.061.secondwhichcorres

23、pondto valuesof 2VForthosetestsinwhichthemodelwas stationarywhilethe.air-streauwas srted a rollipgveloci, thevaluesof pb were0.057,0.029, 0.008, 0, -o.025, -043, and -0.065 in the caseof the sting-mountedmodeland0, W.023, fo.046,andtO.063in the caseof the strut-mountedmodel.The followingmodelconfigu

24、rationswereteited: fuselage,wing,andtail(designatedFWT),fuselageandwing (designatedEW),fuselageandtail(designatedl?l),andfuselage(designatedF). .REDUCTIONOF DATA .The equationsof equilibriumofrollingandyawingmomentsforamodelmountedon a strain-gagebalanceandhavinga motionaboutitsrollaxissreandwhere L

25、 and N arethemomentsmeasuredby the strain-gagebalance.Sinceforthishermonicmotion - .-. . . - -.- . .Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-2PNACATN 3554 9then,at thetimewhen = O,j = .Qb2 ofSimilsrly,forthedWP* inroll,2 (L)oc3p=-_tipVSb2ofd t

26、hewingitselfcontributesonlya smaLl%negativevalueto . Ihthepresenceof thewing,however,theloadingon theverticaltail.dueto rollingis effectivelyovercomebyan oppositeloadingdueto sidewashfromtherollingwingas discussedinreference4.In orderofmagnitude,valuesof Cnp obtainedl%omoscillationtestsagreeverywell

27、,withoneexception,withthe steady-statevaluesobtainedinrollingflowat an angleof attackof OO. The exceptionoccurredforthecompletemodelforwhicha smallpositive Cnp WaSmeasuredinrollingflowwhereastheoscillatoryvalueswerealsosmallbutnegative(fig.9(a). These-rolling-flowresultswereforthemodelsupportedon th

28、eoscillationstingin the samemanneras itwas fortheoscillationtests.Increasingtheangleof attackof themodelfromOO to 8 at 1 cycleper secondreducedthelargevalueof c% forthefuselageandtailby about70percent(fig.9(c). No particul= effectof angleof attackwas shown,however,on therathersmallvaluesof Cnp exhib

29、itedbytheotherconfigurations.The incrementsin p contributedby thetailbothin thepres-enceandin theabsenceof thewingare shownin figureU. togetherwithsteady-statevaluescalculatedforthetail. The tail-alonevalueofCnp was calculatedfromthedataof reference5. Thisvalueis expectedto be scmewhatlargefora tail

30、surfacebecausetheloadingsgiveninreference5 areforonesemispanof a completerollingwingand containsomeloadcarriedoverfromtheothersemispan.Thepresenceof thewingwas takenintocomiderationby usingthemethodpresentedinref-erence4 to estimatethe sidewashat thetaildueto the load. Theestimatedreductionenceof th

31、ewingisAISO shownClear,Cnp resultingfromoscild.ation.inininanthetailcontributionto C% dueto-thepres-goodagreementtiththeexperimentalresults.figure-n.,is the overallreductionto thetailincreasein eitherfrequencyor amplitudeof. . .- -. - . . . . . . .-Provided by IHSNot for ResaleNo reproduction or net

32、working permitted without license from IHS-,-,-12 NACATN 3554Dampinginroll.- Theresultsshownin figure10 indicatethatatan angleof attackof 0 thereexeno importanteffectsofunsteadymotiononthedaqing h rold-fortheunswept-wingmodeltested. Thevaluesof Clp obtainedby oscillationaxe consistentwiththoseobtain

33、edby rolllmgfluwwiththemodelsupportedon theoscillationsting. Thewingitselduced thelargestincrementof Clp. Thepresenceofthewingreducedthe contributionof.thetailto Ctp ,asit didfor%P, butthiseffectis insificant becauseof thenessof thetailcontributionto Czp.An estimateof the steady-statecontributionoft

34、o C2P msybe madeby methodssimilarto thoseforbution. Sucha calculationpredictsa valueof LCzpcompsxativesmsU-theverticsltail the C% contri-= -0.044 forthetdl in theabsenceof thewinganda valueof CClp= -0.027 forthetailin thepresenceof thez. Thesevaluesare inagreamentwiththe incrementswhichcambe obtaine

35、dfromfigure10.An increasein angleof attackfromOO to 8reducedthedampinginrollofthewingby a considerableamount(fig.1O(C). Thisreduc-tionin Czp isroughlyproportionalto the changein thewinglift-curvesiopebetweentheseangles. (Seefig.8.)On the qualityof the oscillationdata.- In orderto recordthesmaU yawin

36、gmomentsdueto rolling,itwas necessarythat the sensi-tivi of the strain-gageyawing-momentbeambe ccmparativel.yhigh.Thishighsensitivityresultedin a naturaltrequencyinyaw of theorderof 20 cyclesper secondforthe straingagewithmodelattached.Becausethemodelwas excitedby wind-tunnelturbulenceorby unsteadyv

37、ortexflowoffthemodel,theoscillographrecordsshowa tracewitha frequencyof 20 cyclesper seconddueto thesedisturbancessuperim-posedon thelowerfrequencytracedueto theforcedoscillation.(Seefig.7.) At an angleof attackof0,the superimposednoisehadanamplitudecomparableinmagnitudeto thesmptude of theoscillato

38、rymomentanditwas veryregularovera longperiodof time. Itwas notdifficult,therefore,becauseof thisregulari,to eliminatethehigh-frequas a result,the superimposednoisewas ofmuchhigherfreqpencyandofmuchsmalleramplitudestn thatforthe corre-spondingyawing-momenttiaceas sh_ti_f igure7-.Although-the-nofseaga

39、inassumedgreateramplitudesandlessuniformim at a = 8,ingeneral,theoscillatoryvaluesof Czp are consideredtobe reliableforti anglesof attack.Comparisonofresultsobtainedby differenttechniques.-A compari-sonof theoscillatoryderivatives(for f = 1 cps andthe steady-statederivativesmeasuredby twomethods(rro

40、llingmodel)is shownin figure13. In general,thesetwo steady-statetechniquesgiveapproximatelythe samevaluesof % and Clpat a= 0,withtheexceptionof a differencein C% forthemodelswiththewing. Thisdifferencemy be duein“partto differencesinsuppmt-strutinterference.Althoughthepreviousdiscussionhas shownthat

41、,for a = Oo,frequencyeffectsweresmell.,thedifferencein at a= 8 betweenthe steady-stateandtheoscillationresultsmsyindicatelargeeffectsof frequencyat thehighanglesof attack. Apreviousinvestigation(ref.6)has shownthat,forwingsforwhichpartialseparationhas occurredduringunsteadymotion,an aerodynamiclagw

42、existwhichcontributesto themomentsactingon thewing. Suchan aerodynamiclag causetherolling-stabili derivativeseitherofthewingitselfor of thetailor fuselagein thepresenceto,be markedlydifferentfromthe steady-statederivativesot attack.CONCLUSIONSAn unswept-wingmodel,whichwas tested-asa fuselageof thewi

43、ngat highanglesalone, afuselage-tail-conibifortheotherconfigurationstested,yawingmomentdueto rollingwas smalland showedno importanteffectof fre-quencyor amplitude.Thelargeyawingmomentdueto rollingof thefuselage-tailconibinationwasreducedto a smallvalueinthepresenceof thewing. Thiswing-interferenceef

44、fecton thetailcontributiontothederivativecanbe accuratelyestimatedby meansof existingsteady-statetheory.2. FYequencyor amplitudehadno noticeableeffecton themagnitudeof thedampinginrolJforthemodelor anyof its componentsat anangleof attackof0.3. The rollingderivativesofthemodelanditscomponentsmeasured

45、by theoscilllationtestsweregeneraldyconsistentat lowanglesofattackwiththederivativesmeasuredby steady-statetests. At a highangleof attackthe oscillatoryyawingmomentdueto rollingforthemodelwiththewingwas differentfromthatobtaineduudersteady-stateconditions.LangleyAeronauticalLaborat,NationalAdvisoryC

46、ommitteeforAeronautics,_ey Field,Vs.,October19,1955. _“.- .Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACATN 3554 15.-.,a71REFERENCES-1. Bird,JohnD.,Fisher,LewisR., andHubbard,SadieM.: SomeEffectsofFYequencyon theContributionof a VerticalTailto

47、theFreeAerodynamicDsmpingof a ModelOscillatingin1953. (SersedesNACATN 2657.)2. Fisher,LewisR., andWolhart,WalterD.: Son #o = 5.(C) a = 8; o = t15.Figure7.- Sampleoscillogrtiphrecordsforthreeamglesof attack.W+ F+ Tjf=lcps.-. - .-. - - - .Provided by IHSNot for ResaleNo reproduction or networking perm

48、itted without license from IHS-,-,-24cma71a16 f = 1 Cps.Figure9.- The effectsof frequency,amplitude,andangleof attackon “theyawingmomentdueto ro12ing. RolMng-flowresultsareformodelmountedon stingsupport. - - - . . .-.- .- . . . - - -Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. . - -_26 NACATN 3554zo?2-.4-.6 0 Q4 08 L2 L6 2.0 2.4 28 32 .36 40.20.2-.4-