NASATechnicalMemorandum106511_ResearchMemorandum-Laboratory_ARL_TR-381/H--37InfluenceofGearDesignParametersonGearboxRadiatedNoise-2-:_4O,aOo,ZC_00FredB.OswaldLewisResearchCleveland,OhioandDennisCenterRTownsend_-L....¢13........................td3tZ.OMarkJ.Valco.........U.S.ArmyCleveland,andRobertH.Spencer,RaymondBoeingHelicoptersPhiladelphia,PennsylvaniaLDrago,andJoseph_=.......W.Lenski,_:_Jr.Research.LaboratoryOhioUJU.ZwtbiZu'l,_aA0IillI.,I..-Illii.IJ_.#1.i..Ii'o-._C_Z_u.IUI--i,u,Ib--IZv1Zu'l0Preparedforthe1994InternationalGearingConferencesponsoredbytheUnitedKingdomNationalUniversityofNewcastleUponTyne,England,GearMetrologyLaborato_September7-9,1994_=U.S.ARMYNationalAeronauticsandRESEARCHLABORATORYSpaceAdministrationInfluenceofGearDesignParametersonGearboxRadiatedNoiseFredB.OswaldandDennisP.TownsendOhio44135NASALewisResearchCenter,Cleveland,MarkI.ValcoU.S.ArmyResearchLaboratory,RobertLewisResearchCenter,Cleveland,Ohio44135H.Spencer,RaymondJ.DragoandJosephW.tenski,Jr.BoeingHelicopters,Philadelphia,Pennsylvania19142SYNOPSISSpurandhelicalgearsweretestedintheNASAgear-noiserigtocomparethenoiseproducedbydifferentgeardesigns.Soundpowermeasurementswereperformedundercontrolledconditionsforamatrixofoperatingconditions.Soundpowerwascomputedfromnear-fieldacousticintensityscanstakenjustabovethetopsurfaceofthegearbox.Testgearsincludedfourspurandfivehelicalgeardesigns.Thegearsweredesignedtobeasnearlyidenticalaspossibleexceptfordeliberatedifferencesintoothgeometryandcontadratio.Testresultsarepresentedasnarrow-bandsoundpowerspectraandaschartscomparingthevariousdesigns.1INTRODUCTIONAmajorsourceofhelicoptercabinnoise(whichhasbeenmeasuredatover100decibelssoundpressurelevel)isthegearbox.ReductionofthisnoiseisaNASAandUSArmygoal.ArequirementfortheArmy/NASAAdvancedRotorcraftTransmissionprojectwasa10dBnoisereductioncomparedtocurrentdesigns.Themainexcitingforceswhichproducegearnoisearethemeshingmission.Whileforcesofthegearteethmanyfactorsinfluenceinthetrans-transmissiondealtwithinthedesignoftheoverallsystem,especiallythesupportbearings,gearblankdesign,andhousingstructure.Doublehelicalgears,whichcancelthethrustloadsfromeachhelixwithinthegearblank,providerelieffromnetthrustproblems.However,thenoisepro-pertiesofdoublehelicalgearshavenotbeenreported.Noninvolutetoothformshavebeeninvestigatedforpossibleuseinhelicoptertransmissionsinrecentyears.Testingofhighprofilecontactratio,noninvolutetoothformgears,(HCR-NIF),hasshownthattheloadcapacitycanbesubstantiallyhigherthanthatofconventionalinvolutegearsandthebendingloadcapacity(athighloads)wasatleastequaltothatoftheinvolutegears[2].Theseinvestigations,however,havecenteredalmostuniversallyontheloadcapacityandnotonnoisegeneration.ThisprogramwasconductedaspartoftheAdvancedRotorcraftTransmissionproject[3].Itsobjectivewastodefine,bycontrolledtesting,theeffectonnoiselevelsduetochangesintheprofileandfacecontactratiosandthegeartoothform.Thesefactorswerevariedbothseparatelyandincombination.Thetestgearconfigurationswereselectedtoberepre-sentativeofthoseusedinhelicoptertransmissions.Thetestgeardesignsincludefourdifferenttypesofspurgears(low-andhigh-contact-ratioinbothinvoluteandnon-involuteprofiles)aswellasfivedifferenthelical(singleanddouble)geardesignswithvariousprofileandfacecontactratios.Thegearsweredesignedtobeasnearlyidenticalaspossibleexceptfordeliberatedifferencesintooth-geometryandcontactratio.Testingwasconductedundercontrolledconditions(torque,speed,oilflow,temperatures,etc.).Acousticintensitymeasurementsweretakenwiththeaidofarobottoinsurerepeatabilityofmeasurementsbetweengearsetsandtominimizetheinfluenceofoperatortechnique.Resultspresentedhereincludetrendsofthenoise,thesimplefactremainsthatifthebasicexcitingforcesarereducedandnoamplifyingfactorsarepresent,theoverallnoiselevelofthesystemwillbereduced.Amongtheseveralwaysinwhichthegeartoothmeshingforcesmaybereduced,twoofthemostdirectlyapplicabletohelicoptertransmissionsaretheformoftheteethandtheoverallcontactratio.Bothapproachesareattractiveforanaerospaceapplicationsince,unlikesoundabsorbingtreatments,theseapproacheshavethepotentialforreducingnoisewithoutreducingperform-anceorincreasingoverallsystemweight.Bothapproachesalsoofferthepossibilityofimprovinggearperformanceintermsoflongerlife,higherloadcapacity,greaterreliability,andreducedweightwhilesimultaneouslyreducingnoiselevels.Helicalgears,ascomparedtospurgears,typicallyproducelowernoiselevels.Winter[1]providesaconcisesummaryonthevariationofexcitationlevelswithfacecontactratio.Thereislittleotherdefinitivedata,foraccurate,groundgears,whichdefinesthenoiseadvantageofhelicalgears.Similarly,anecdotalinfor-mationindicatesthathighercontactratios,bothfaceandprofile,alsotendtoreducenoiselevelsbut,again,harddatawasnotreadilyavailable.Whilehelicalgearsprovidesomenoisereduction,theirusealsogeneratesathrustloadwhichmustbesoundpoweratmeshfrequencyandnarrow-bandDuringthemanufactureofthetestgears,thedoublespectraofsoundpower.PreliminaryresultsfromthishelicalgeardrawingswentoutwithadraftingerrorprogramwereearlierpresentedbyDrago[4].suchthatbothhelicesweremanufacturedwiththesamehand.Theresultantgearset(knownofficiallyas\"spread2TESTGEARSsinglehelicalgears\"andunofficiallyas\"OOPS\"gears),areshownintheupperrightcornerofFig.1.AlthoughthesegearsprobablywouldnotbeusedinaproductionEightsetsoftestgearsweredesigned.Fouroftheseareenvironment,wedecidedtotestonepairofthemspurgears.Twosetshaveaninvolutetoothformandtwoutilizeanoninvolute,constantradiusofcurvatureanyway.toothform.Thefourhelicalgearsetsincludevariousprofileandfacecontactratios.Allgearsweredesigned3APPARATUSANDPROCEDUREinaccordancewithstandardaerospacepracticesothat,exceptforsize,theyarerepresentativeoftypicalheli-3.1TestFacilitycoptergears.TheeightgeardesignsaresummarizedinTable1andareshowninFig.1.Additionaltestpara-TheNASALewisgearnoiserig(Fig.2)wasusedformetersareshowninTable2.thesetests.Thisrigfeaturesasingle-meshgearboxFigure1alsoshowsagearsetwhichisnotlistedinpoweredbya150kW(200hp)variablespeedelectricTable1.Thiswasnotoneoftheplannedtestvariants.motor.Aneddy-currentdynamometerloadstheoutputTable1TestGearConfigurationsTransverseContactRatioConfigurationToothFormTypePressureAngleProfileFaceTotalt.SpurBaselineInvoluteSpur251.30.01.32.I-ICRSpurInvoluteSpur202.10.02.13.HelicalBaselineInvolute20°Helical251.31.152..DoubleHelicalInvolute350Helical251.32.33.65.HelicalInvolute27°Helical251.31.62.96.HCRHelicalInvolute34°Helical202.12.14.27.Nil=SpurBaselineNonInvoluteSpur251.30.01.38.NIF-I-ICRSpurNonlnvoluteSpur202.10.02.1Fig.1-Testgears2Table2TestgearparametersNo.Teeth25and31Transversemodule,mm3.175(8)(diametrialpitch,in1)Facewidth,ram(in)31.8(1.25)100%inputspeed,rpm5000100%inputtorque,N-m,(in-lb)256(2269)100%power,kW(hp)134(180)shaft.Thegearboxcanbeoperatedatspeedsupto6000rpm.Therigwasbuilttocarryoutfundamentalstudiesofgearnoiseandthedynamicbehaviorofgearsystems.ItisFig.3-TestgearboxandRAIMSrobotdesignedtoallowtestingofvariousconfigurationsofgears,Theacousticintensityprobeconsistsofapairofphase-bearings,dampersandsupports.Toreduceunwantedmatched6mmmicrophonesmountedface-to-facewitha6reflectionofnoise,acousticalbafflescoveredtestcellwalls,mmspacer.Theprobehasafrequencyrange(_1dB)offloor,andothernonmovingsurfaces.Thematerialatten-300-10000Hz.Measurementsweremadeatadistanceofuatesreflectedsoundby20dBormoreforfrequenciesof60mmbetweentheacousticcenterofthemicrophonesand500Hzandabove.thegearboxtop.Ateachoperatingcondition,theintensityspectracollectedfromthetwentynodesofthegridwereaveraged,thenmultipliedbytheareatocomputean801-linesoundpowerspectrum.Theareawasassumedtobetheareaofthegridplusone-halfadditionalrowandcolumnofelementsor0.0910m2:Theactualareaofthetopis0.1034m2.Wedidnotextendthemeasurementgridcompletelytotheedgesofthegearboxtopbecause(1)theedgeofthetopwasboltedtoastiffmountingflangewhichwouldnotallowmuchmovement,and(2)measurementstakenclosetotheedgeofthetopwouldbeaffectedbynoiseradiatedfromthesidesofthebox.Fig.2-GearnoiserigNoisemeasurementsfromthegearboxsideswerenotattemptedforthefollowingreasons:(I)thetopisnotas3.2InstrumentationandTestProcedurestiffasthesides;thus,noiseradiationfromthetopExperimentalmodaltestresultsfromaprevioustestingdominatesatmostfrequencies;(2)thenumberofmeasure-program[5]providedthefirstfivenaturalfrequenciesandmentlocationswerereduced;and(3)shaftingandothermodesofvibrationofthegearboxtop.Thenaturalfreq-projectionsmadesuchmeasurementsdifficult.uencieswerecheckedtoassurethatgearmeshfrequenciesSoundpowermeasurementsweremadeoveramatrixofdidnotcoincidewithimportantmodesofthegearbox.ninetestconditions:3speeds(60,80,and100percentofAlso,frompreviousanalyticalwork,weknowthattorsion-5000rpm)andat3torquelevels(60,80and100percentalmodesofthegearsystemarewellabovethe6000rpmofthereferencetorque256N-m(2269in-lb)).Duringspeedlimitoftherig.eachintensityscan,thespeedwasheldtowithin+5rpmAcousticintensitymeasurementswereperformed,underandtorqueto+2N-re.Atleastfivecompletesetsofscansstable,steady-stateoperatingconditions,withtheaidofawereperformedoneachgearset.computer-controlledrobotdesignatedR.AIMS(RoboticAcousticintensitydatawererecordedovertheAcousticIntensityMeasurementSystem).TheRAIMSsoft-bandwidth6-7296Hz.Onthe801lineanalyzer,thisware(1)commandedtherobottomoveanintensityprobeproducedalinespacingof8Hz.Wechosethisfrequencyoveraprescribedmeasurementgrid;(2)recordedacousticrangebecauseitincludesthefirstthreeharmonicsofgearintensityspectraintheanalyzerforeachnodeofthegrid;meshingfrequencyforthespeedrange(3000-5000rpm).and(3)transmittedthespectratothecomputerforstorageondisk.Thegearbox,robotandintensityprobeareillust-3.3ProcessingSoundPowerDataratedinFig.3.RAIMSismorecompletelydescribedinThesoundpowerdatacapturedbythemethodoutlinedreferences[6]and[7].aboveconsistsofmanydatafilesofsoundpowerspectra.SamplespectraforthefourspurgearconfigurationsareejgshowninFig.4andspectraforthefivehelicalgearconfigurationsareshowninFig.5.Eachspectrumincludes59thefirstthreeharmonicsofgearmeshfrequency.The3e.harmonicfrequenciesaremarkedwitha\"o\"onthetopborder.Eachharmonicissurroundedbyseveralsidebands.Ik,+Themostprominentsidebandswererelatedtothepinionshaftfrequency.Gearshaftsidebandswerenotprominent.Tocharacterizethemeasurements,wedecidedtoreduceeach80l-linesoundpowerspectrumtoafewnumbersthatwouldrepresentthegearmeshnoise.Wecallthesenum-berstheharmonicsoundpowerlevels.Weconsideredfiveteee_f,m__,_,3+._o4ore,,t_55Be,_,t+_methodsfordeterminingtheharmonicsoundpowerlevel:Fe'e_.uenc,d,Hz(1)RecordonlythevalueatthemeshfrequencyFig.4-Spectraforspurgears(frombottom,configur-harmonic.Thismeanstoignoresidebandseventhoughations1,2,7,8)at100%speed,100%torque.theywereoftensignificant.(2)Checktheharmonicfrequencyandseveralsidebandsandrecordthehighestvalue.79(3)Addtogetherthevalueswithinafixed-widthfrequencybandcenteredonthemeshfrequency.Thismeansmoresidebandswouldbeincludedatlowerspeedswherethesidebandspacingisless.(4)Similarto(3)exceptthesizeofthefrequencybandwouldvarywithspeed.Thismeansthenumberofvaluest_addedtogetherwouldnotbeconstant.(5)Addthevaluesatthemeshfrequencyandatafixednumberofsidebandsoneachsideofthemeshfrequency.Alternative5waschosenforcalculatingharmonicsoundFr\"eque.,nct..j,Hzpowerlevels.Weusedthreepairsofsidebandsatpinionshaftspacing(i.e.,7peaks).SoundpowervalueswereFig.5-Spectraforhelicalgears(frombottom,eonfigur-convertedtoWattspriortocalculationofsums.afio_3,5,6,4and\"OOl_\")at11111%speed,11111%torque.TOreduceeffectsofspeeddri_andSignalleakagewetookthevalueatthepeakplustwofrequencylinesoneachimside.Inotherwords,weaddedtogether5valuesateachpeak.Sincesevenpeakswereused,35values(5x7)wereaddedtogethertoproduceeachharmonicsoundpower_Nlevel.Figure6showsthedata(markedwithsymbols\"*\"/and\"+\")usedtocomputeoneharmonicsoundpowerlevel.ThisisfromthetoptraceinFig.4nearthefirstharmonicat2083Hz.(Wedeliberatelychoseanunusualexamplewhereonesidebandishigherthanthemeshfreq-uency.)Thesidebandspacing(at5000rpm)is83Hz.,,18ii+IS192+2122+2\"124thusthereareabout10analyzerlinespersideband.AtFr'e:qumr_c_,kHzlowerspeeds,therearefeweranalyzerlinespersideband.Fig.6-EnlargementofportionoftopspectrumtnFig.4.3.4DataSamplingwhereTobeassuredthatdatafromeachgearsetcanbereliablyCt=confidencelimit,dBcomparedwithdatafromothergears,weneededtohavet=probabilitydistribution(\"Studentt\"distribution)sufficientrecordstoestablisha95%confidencelevelof=standarddeviationofdata,dB+1dB.Thisiswellbeyondthepracticaldifference(i.e.,n=numberofsamples(typically5)achangeofabout3dB)whichnormalhearingcandetect.WeperformedatleastfivecompletesetsofseamonValuesforthe\"t\"distributioncanbefoundinanyeachgearpairtested.Fromthesesetsofmeasurements,standardstatisticstext.Wechosea95percentconfidencewecomputedmeanvaluesandconfidencelimitsofthehar-levelwhichcorrespondstoaprobabilitylevelof0.05.Themonicsoundpowerlevel.(Forthecalculationofmeanandnumberofdegreesoffreedominthetdistributionistheconfidencelimit,dBvalueswereused.Wedidnotconvertnumberofsamplesminus1.backtoWatts.)Theconfidencelimitwascalculatedfrom:Themeanvaluesofthethreeharmonicsoundpowerlevelswereusedtocomputeasingle\"composite\"noisect=t(++/_)levelforeachtestconditionbyaddingthesoundpower(inWatts)andthenconvertingtodB.Itisthesecompositeg5valuesthatwecompareforthevariousgearconfigurations.Toestimatetheeffectduetosample-to-samplevariation,twosetsofgearsforeachdesignwerefabricatedandtested.Eachgearwasinspectedinaccordancewithtypicalproductionhelicopterstandards.Theoverallaccuracyofthegearswasconsistentwithproductionhelicoptergearsofsimilarsizeandconfiguration.Thevariationbetweenthesetsofgearsisreasonablytypicalofnormalproductionforgearsinthesamemanufacturinglot.Lottolotvariations(nottestedhere)maybehigherbuttheoveralltrendoftheeffectshouldbeaboutthesame.Alargedifferenceinnoiselevelissometimesobservedonproductiongearboxessimplyasaresultofrebuildingthemafterdisassemblyforinspection,eventhoughnopartswerechanged.Consideringthiseffect,inadditiontothemanufacturingvariabilitychecks,wealsocheckedforvariabilityduetodisassemblyandreassembly.Wecheckedforvariabilitybytestingthree\"builds\"ofthefirstgearset.Eachbuildusedexactlythesamepartsandeachwasaccomplishedbythesametechnicianusingthesametools,4RESULTSAverylargeamountofdatawascollectedduringthistestprogram.AnoverviewofallthedataispresentedinthecompositenoiselevelbarchartsofFigs.7-8.andparts.|*o].65lO0/100_d1_!ITo_m1_)Fig.7-Spurgearcomposite100noiselevels100/11010011004.1SpurGearstoothratio.reactingFig.8-Helicalgearcompositenoiselevelsthethrust)thusthisresultisespeciallyinterestingWetestedgearswithbothinvoluteandnoninvoluteform,andwithbothstandardandhighprofilecontactThoughthenoiselevels(Fig.7)generallyincreasedwithspeed,ingeneral,thehighcontactratiospurgears(configs.2,8)were2dBquieterthanthestandardcontactratiogears(configs.1,7)regardlessofthetoothform.Similarly,theinvolutetoothformgears(configs.1,2)werequieter(by3-4dB)thantheirnoninvolutecounterparts(configs.7,8).4.2HelicalGearsThesinglehelicalgearsincludethreedifferenthelixanglesandbothstandardandhighprofilecontactratios.Asinthespurgears,anincreaseinthecontactratiocorrelateswithadecreaseinthenoiselevel.Increasingthefacecontactratiofromabout1.15(config.3),to1.6(config.5),decreasesthenoiselevelsubstantiallyineverycase,thoughtheresultsathigherspeedsaremoredramaticthanatlowerspeeds.Also,ateveryoperatingcondition,thecompositenoiselevelofahelicalgear(Fig.8)islessthanthelevelforaspurgearwithsimilarprofilecontactratio.Thecombinationofahighprofileandhighfacecontactratiofurtherdecreasesgearnoise.Indeed,thehighprofileandhighfacecontactratiodesign(config.6)withprofileandfacecontactratiosof2.1,and2.1respectivelywasthelowestnoisegeneratoratalmosteveryoperatingcondition.Helicalgearsusedinhelicopterstendtohaverelativelylowfacecontactratios(helixanglesarekeptlowtominimizethrustloadingandtheextraweightassociatedwithsinceitsuggeststhatitmaybepossibletotradeoffhelixangleagainstincreasingprofilecontactratiotoimprovethenoiselevelwithouttheweightpenaltyassociatedwithaccomplishingthesamereductionwithhelixanglealone.Asurprisingresult,thedoublehelicalgearsetwasnoisier(by4dBonaverage)thanitssinglehelical(OOPSgear)counterpart.TheOOPSgearsetisessentiallyasinglehelicalsetwithagapinthemiddleofthetoothface.Itseffectivefacecontactratioissimilartothatofthehighcontactratiohelicalgears(config.6).Thedoublehelicalphenomenaappearstoberelatedtoaxialshuttlingwhichoccursasthedoublehelicalpinionmovestobalanceoutthenetthrustloading.Theshuttlingisduetothepresenceofsmallmismatchesintherelativepositionsoftheteethoneachhelix.Nomatterhowaccuratethegearis,somemismatchwillalwaysbepresent,thusthisisanunavoidablephenomena.Whilethethrustbalancingcharacteristicofadoublehelicalgearisavaluabledesignfeaturesinceitgreatlysimplifiesthebearingsystem,apriceispaidintermsofnoiseandvibrationasthegearsetshuttlesbackandforth.Sincetheperhelixfacecontactratio,facewidth,profilecontactratio,etc.areidenticalfortheOOPSandthedoublehelicalgearsets,theonlyoperationaldifferenceisthelackofaxialshuttling.Thedoublehelicalsetwillbeinaconstantequilibriumseekingstatebecauseofthetheoretic-1.Themostsignificantfactorfornoisereduction,withinaallyzeronetthrustloadwhiletheOOPSgearsetwillrungeardesigner'scontrol,wasfoundtobethetotalcontactinafixedaxialpositionduetothenetthrustload.ratio.GearnoisemaybereducedbyincreasingeithertheThistestprovidessomeinsightintothemagnitudeoftheprofileorfacecontactratio.noisepenaltywhichispaidwhendoubleratherthanequi-valentsinglehelicalgearsareused.Sincethesetestgears2.Thenon-involutetoothformspurgearswerefoundtoareallveryaccurate(typicalforhelicoptergears),itshouldhavea3-4dBnoisepenaltycomparedtotheirconventionalbeobviousthatalargerpenaltywouldbepaidifgearsofinvolutecounterparts.lesserqualityweretobeusedbecausethelowerthegearqualityis,themoreshuttlingwouldbelikelytooccur.3.Thehigh-contact-ratiospurgears(witha58percentincreaseinprofilecontactratio)showedanaveragenoise4.3Sample,Build,andSpecimenVariationsreductionofabout2dBoverstandardgears.Wetookatleastfivesetsofnoisescansateach4.Thenoiselevelofdoublehelicalgearsaveragedabout4operatingcondition.OurgoalwastoobtainconfidencedBhigherthanotherwisesimilarsinglehelicalgears.limitswithin1dBforeachvalueofharmonicsoundpowerlevel.Thisgoalwasmetonabout60percentofthetest5.Innoisereductiontests,variationduetounintendedsets.effectssuchastestingdifferentpartspecimensorevenDuringothertesting,theauthorshavenotedsignificantreassemblywiththesamepartsmaybeofthesameordervariationsinthemeasured(andperceived)noiseleveloftheofmagnitudeastheeffectofdeliberatedesignchanges.samegearsystembeforeandafterdisassembly.Insomeeases,thisvariationwasofconsiderablemagnitude.ToREFERENCESinvestigatethisphenomena,thefirstsetofbaselinespurgears(config.1),wasassembled,tested,disassembled,re-(i)Winter,H.,Gerber.H.,Muller,R.:\"Investigationsontheassembledandthentestedagain.ThisprocesswasrepeatedExcitationofVibrationsandNoiseatSpurandHelicaluntilthegearshadbeentestedthreetimes.Gears\Proc.ASME19Int.PowerTrans.GearingConf.,ThelargestminimumtomaximumbuildvariationwasChicago,IL,Apr25-28,19,pp765-7727.8dB(atthehighspeed,lowtorquecondition)Whiletheminimumbuildvariationwas0.7dB(atthemediumspeed(2)Townsend,DennisP.,Baber,BedB.,andNagy,Andrew:condition).Theaveragebuildvariationwasabout3dB.\"EvaluationofHigh-ContactRatioSpurGearsWithProfileWhilenorealpatternisapparent,itappearsthattheModifications,\"NASATP-1458,Sep.1979.variationdecreasedslightlywithincreasingload.Sincewetestedtwosamplesofeachoftheeightgear(3)Bill,R.C.:\"SummaryHighlightsoftheAdvancedRotorcraftTransmission(ART)Program,\"AIAAPAPER92-3362,Julydesigns,wecancomparethe\"build\"variationtothevaria-1992tionbetween\"identical_parts.Fortheeightgeardesigns,theaveragepart-to-partvariationinthecompositenoise(4)Drago,R.J.,Lenski,J.W.,Spencer,R.H.,Valco,M.,andlevelswas2.8riB.OnewouldexpectthevariationbetweenOswald,F.B.:\"TheRelativeNoiseLevelsofParallelAxissamplesofthesameparttoequalorexceedthevariationGearSetswithVariousContactRatiosandGearToothfromrebuildingthesameparts.The\"build\"testwasForms,\"AGMAreport93-FTM-11,Oct.1993(also,NASAperformedatthebeginningofthetestprogram.IncreasedTM-1031).experiencemayhavereducedthevariationforlatertests.Thefactorsconsideredabovepointoutthedifficultyin(5)Oswald,F.B.,Zakrajsek,J.J.,Townsend,D.P.,Atherton,defininganoisereductioneffortinthatthevariationsdueW.,andLin,H.H.:\"EffectofOperatingConditionsontounintendedeffectsareoftenofthesameorderofmagnit-GearboxNoise,\"ASMEpublicationDE-Vol.43-2,pp.669-674,Amer.Soc.Mech.Engr.Sep.1992(also,NASATM-udeofthechangeswhichmaybeattributedtogearconfig-105331).urationortreatment.Suchdifferencesshouldexceedthevariationsduetosampleandbuildeffectsandthoseobserv-(6)Flanagan,P.M.,andAtherton,W.J.,1985,\"InvestigationedamongdifferentspecimensofthesamepartbeforetheyonExperimentalTechniquestoDetect,LocateandQuantifycanbeconsideredsignificantofthemselves.GearNoiseinHelicopterTransmissions',NASACR-3847.(7)Atherton,WilliamI.,Pintz,Adam,andLewicki,DavidG.,CONCLUSIONS1987,\"AutomatedAcousticIntensityMeasurementsandtheEffectofGearToothProfdeonNoise',MechanicalNinedifferentspurandhelicalgeardesignsweretestedinSignatureAnalysis-MachineryVibration,.Flow-inducedVibration,andAcousticNoiseAnalysis,S.Braun,etal.,theNASAgear-noiserigtocomparethenoiseradiatededs.,ASME,pp.109-113(also,NASATM-100155).fromthegearboxtopforthevariousgeardesigns.SoundpowermeasurementsweremadeundercontroUedconditionsforamatrixofoperatingconditions.Thefollowingconclusionsweremade:,,1FormApprovedREPORTDOCUMENTATIONPAGEOMBNo.0704-0188Publicreportingburdenforthiscollectionofinformationisestimatedtoaverage1hourperresponse,includingthetimeforreviewinginstructions,searchingexistingdatasources,gatheringandmaintainingthedataneeded,andcompletingandreviewingtheco.llectionofinformation;Sendcommentsr.egardingthisburdenestimateoranyetherasp_t,_,ofthiscollectionofinformation,includingsuggestionsforreducingthisburden,toWashingtonHeadquarters:_ervlces,u=rectorateformtormat=onuperatJonsanoHepons,1_,Je_ersonDavisHighway,Suite1204,Adlngton,VA22202-4302.andtotheOfficeofManagementandBudget.PaperworkReductionProject(0704-0188),Washington,DC20503.1.AGENCYUSEONLY(Leaveblank)2.REPORTDATE3.REPORTTYPEANDDATESCOVEREDMarch1994TechnicalMemorandum4.TITLEANDSUBTITLE5.FUNDINGNUMBERSInfluenceofGearDesignParametersonGearboxRadiatedNoiseWU-505--62-366.AUTHOR(S)1L162211A47FredB.Oswald,DennisP.Townsend,MarkJ.Valco,RobertH.Spencer,RaymondJ.Drago,andJosephW.Lenski,Jr.7.PERFORMINGORGANIZATIONNAME(S)ANDADDRESS(ES)8.PERFORMINGORGANIZATIONNASALewisResearchCenterREPORTNUMBERCleveland,Ohio44135-3191andE-8622VehiclePropulsionDirectorateU.S.ArmyResearchLaboratoryCleveland,Ohio44135-31919.SPONSORING/MONITORINGAGENCYNAME(S)ANDADDRESS(ES)10.SPONSORINCdMONITORINGAGENCYREPORTNUMBERNationalAeronauticsandSpaceAdministrationWashington,D.C.206--0001andNASATM-106511U.S.ArmyResearchLaboratoryARL-TR-381Adelphi,Maryland20783-114511.SUPPLEMENTARYNO3\"ESPreparedforthe1994InternationalGearingConferencesponsoredbyticUnitedKingdomNationalGearMetrologyLaboratory,UniversityofNewcastleUponTyne,England,September7-9,1994.FredB.OswaldandDennisP.Townsend,NASALewisResearchCenter;MarkJ.Valco,VehiclePropulsionDirectorate,U.S.ArmyResearchLaboratory,Cleveland,Ohio;andRobertH.Spencer,RaymondJ.Drago,andJosephW.Lenski,Jr.,BoeingHelicopters,Philadelphia,Pennsylvania.Responsibleperson,FredB.Oswald,organizationcode2730,(216)433-3957.12a.DISTRIBUTION/AVAILABILITYSTATEMENT12b.DISTRIBUTIONCODEUnclassified-UnlimitedSubjectCategory3713.ABSTRACT(Maximum200words)SpurandhelicalgearsweretestedintheNASAgear-noiserigtocomparethenoiseproducedbydifferentgeardesigns.Soundpowermeasurementswereperformedundercontrolledconditionsforamatrixofoperatingconditions.Soundpowerwascomputedfromnear-fieldacousticintensityscanstakenjustabovethetopsurfaceofthegearbox.Testgearsincludedfourspurandfivehelicalgeardesigns.Thegearsweredesignedtobeasnearlyidenticalaspossibleexceptfordeliberatedifferencesintoothgeometryandcontactratio.Testresultsarepresentedasafunctionofthegeardesignandoperatingconditionsintheformofsoundpowerchartsandasnarrow-bandspectra.14.SUBJECTTERMS15.NUMBEROFPAGES8Gearnoise;Spurgears;Helicalgears;Acousticintensity;Soundpower16.PRICECODEA0217.SECURITYCLASSIFICATION18.SECURITYCLASSIFICATION19.SECURITYCLASSIRCATION20.LIMITATIONOFABSTRACTOFREPORTOFTHISPAGEOFABSTRACTUnclassifiedUnclassifiedUnclassifiedNSN70-01-280-5500StandardForm298(Rev.2-)PrescribedbyANSIStd.Z39-18298-102
