外文翻译--直齿圆柱齿轮啮合过程中的声发射源 英文版.pdf
DGZfP-ProceedingsBB90-CDLecture46EWGAE2004469ThesourceofAcousticEmissionduringmeshingofspurgearsCheeKeongTan,DavidMbaSchoolofEngineering,CranfieldUniversity,Bedfordshire.MK430AL,UKTel+44(0)1234-754681,Fax:+44(0)1234-751566.E-mail:c.k.tancranfield.ac.ukandd.mbacranfield.ac.ukAbstractConditionmonitoringofgearswithvibrationanalysisandSpectrometricOilAnalysisPrograms(SOAP)arewidelyacceptedandusedinthemilitaryaviationindustry.WhilsttheliteraturetodatehasshowntheapplicationoftheAcousticEmission(AE)techniquetogearboxhealthmonitoringstillbeinitsinfancy,thereexistopportunitiestodevelopthetechniqueintoacomplementarydiagnostictool.Indevelopingsuchatool,itisimperativethatanunderstandingofthesourceofAEactivityisestablished.ThispaperpresentsexperimentalresultsthatexploretheAEsourceduringagearmeshandhasledtoquiteuniqueandsignificantobservations.Keyword:AcousticEmission,asperitycontact,conditionmonitoring,geardefectdiagnosis,machinehealthmonitoring,rollingfriction,slidingfriction.LiteraturereviewsApplicationofvibrationanalysistogearfaultdiagnosisandmonitoringhasbeenwidelyinvestigatedanditsusageinindustryiswellestablished.Thisisparticularlyreflectedintheaviationindustrywherehelicopterengines,transmissionsystems,drivetrainsandrotorsystemshaveadoptedvibrationanalysisforhealthmonitoring.However,researchintheapplicationofAEtogearfaultdetectionandmonitoringislimited.Themostcommonfailuresencounteredinoperationalgearboxesincludemicro-pitting,pitting,scuffingandabrasivewear.Althoughtoothfractureandbendingfatiguearerare,thecriticalityofsuchafailurehasdrawnhugeattentionfromresearchers.InexaminationoftheAEtechniqueonthesefailuremodes,mostresearchershaveoptedforsimulatedpitdefects.Singh1,2,Tandon3andSiores4performedtheirexperimentsusingsimulatedpits,whilstToutountzakis5,Sentoku6andMiyachika7allowednaturaldefectssuchaspittingtooccurduringthetests.Theconclusiondrawnfromalltheseexperimentswereencouraging;AEtechniquewasabletodetectbothseededandnaturaldefects.Amongtheseresearchers,onlyToutountzakisandSentokuemployedaslipringtotransmittheAEdatafromtherotatingAEsensortotheacquisitionsystems,therebyofferingadirecttransmissionpath.OthersmountedtheirAEsensorsonthebearingorgearboxcasingand,claimedsuccessfulidentificationofdefectivegears.ThepapersreviewedhaveillustratedthepotentialandviabilityoftheAEtechniqueinbecomingausefuldiagnostictoolinconditionmonitoringofgears.However,nonehaveDGZfP-ProceedingsBB90-CDLecture46EWGAE2004470investigatedthesourceofAEactivityduringthegearmesh.AnunderstandingofthefundamentalAEsourcemechanisminmeshinggearsisofvitalimportanceindevelopingthistechnique.ObservationofAEburstsduringgearmeshAEdatawasrecordedduringsimulateddefectgeartests.AninterestingobservationwastheAEtransientsassociatedwiththegearmesh.Withasamplingrateof10MHzandrotationalspeedofapproximately742rpm,only16meshingteethsignatureswererecordedperacquisitionwindow.Figure1showsthetimedomainofanAEsignaturerecordedduringthetestsundertwodifferentloadconditionsclearlyshowingtheAEtransientresponseassociatedwithgearmeshingof16teeth.ThisAEtransientresponseledtheauthorstoinvestigatethepossiblesourceofAEactivityduetothemeshinggear.TheresultspresentedwereacquiredfromtheAEsensormountedonthepinionwheelandAEdatawascapturedwithacommercialdataacquisitioncard(PAC)viaaslipring.00.0050.010.0150.020.025-0.1-0.0500.050.1Amplitude(v)Time(seconds)00.0050.010.0150.020.025-0.1-0.0500.050.1Time(seconds)Amplitude(v)Figure1TimedomainofanAEsignatureshowingclearlytheAEtransientresponseassociatedwithgearmeshingof16teethfortherotationalspeedof745rpmundervariousloads(pre-amplification20dB)AESourceMechanismsFromtheobservationspresentedthusfar,theauthorsbelievetherearethreepossiblesourcesofAEactivityduringthegearmesh:toothresonance,secondarypressurepeakinlubricatedgearsand/orasperitycontacts.DGZfP-ProceedingsBB90-CDLecture46EWGAE2004471ToothresonanceEstimationofthetoothresonancefrequencywasaccomplishedbymodellingthegeartoothascantileverbeamandspringmasssystem.Basedonthegeartoothgeometry,loadingconditionsandontheassumptionthattheloadistransmittedalongthepitch-line,thetoothresonancefrequencywascalculatedat75kHz.Thisisbelowthefrequencyrange(100to1200kHz)oftheAEsensoremployed.Inactualgearsystems,alowermomentofinertiawouldbeexpectedresultinginalowernaturalfrequencythanthatcalculated.ThiseliminatedgeartoothresonanceasasourceofAEactivityduringgearmesh.SecondarypressurepeakinthelubricantTheotherpossiblesourceofAEmechanisminthemeshinggearscouldoccurasaresultofthepressuredistributionbetweenthegearteethsurfacesandthelubricatingoilfilm.Thispressuredistributionisstronglyinfluencedbythegearoperatingconditionssuchasload,speedandmaterialproperties,asdepictedinfigure29.Fromthisfigure,theoccurrenceofalocalpressurepeakfarinexcessoftheHertzainmaximumisobserved.Thishastheeffectofreducingthefilmthicknessatthepositionofthesecondarypressurepeak.Thedecreaseinfilmthicknessatthisparticularpositioncanbeabrupt,dependingonthesurfacevelocity.ThissuddenincreaseinpressureordecreaseinfilmthicknesscouldbeasourceofAEactivityalthoughtheauthorscannotprovethisatpresent.Itmaybeworthnotingthatloadhasnoinfluenceonthefilmthicknessorlevelofsecondarypeak.Iftheraisetimeforthesecondarypressurepeakisintheorderof0.8to10µs,thesecondarypressurepeakcouldgenerateAEtransientburstasobservedfromthegearmeshAEbursts.Thispostulationassumesisothermalconditionsunderpurerolling,howeverthisisnotthecasewithinthegearmeshwhererollingandslidingarebothexperienced.FrictionandAsperityContactsDuringthegearmesh;sliding,rollingoracombinationofbothwilloccur.Asthegearteethsurfacesarelimitedtomanufacturingcapabilities(approximately2µm)asperitycontactswilloccur10duringmeshingonalmostallgears,particularlyasthecalculatedoilfilmthickness(approximately1µm)inthisinstanceislessthanthecompositeroughness.DGZfP-ProceedingsBB90-CDLecture46EWGAE2004472Figure2Pressuredistributionsforacompressiblelubricantwithincreasingspeedsfromplots(0)to(5)9.Duringexperiments,Smith11notedtransientshockpulsesduringthegearmesh.Itwasconcludedthattheseshockswereattributedtoasperitycontact.Whilstasingleasperitymodelwaspresentedastheprobablecauseoftheshocks,thelikelihoodofsuchscenarioinpracticeislimited,multiplecontactswillbepresent.However,itwasshownthatbasedonasperitywidthof5µmandslidingandrollingvelocitiesoftheorderof500mm/s,theraisetimeforsuchatransienteventwas10µs.ItmustbenotedthatthesensorsemployedbySmithhadanaturalfrequencyof50kHzandisoutsidetherangeofAE.Forthisparticularinvestigation,anasperitywidthofapproximately2µm,withapitch-linerollingvelocityoflessthan2000mm/s,canprovidearaisetimelessthan10µswhichwillbedetectedbytheAEsensoremployed.AsthegearmeshinvolvesrollingandslidingofmatinggearsitisworthnotingthatBonesetal8suggestedthatAEactivitywasmostlikelyduetoasperitycontactduringsliding.InrelatingAEtoslidingfrictionDornfeld12etalhaveshownthehighsensitivityofAEtoslidingspeedandappliedload.ItwasnotedthatthebasicmechanismforAEgenerationwastheelasticdeformationofthematerialatasperitycontacts.Therangeofsurfacefinishforthematerialsinvestigatedwasfrom1to4µm;comparablewiththegearstestedinthisstudy.ItwasnotedthatthestrengthofAEamplitudewasdependentonsurfaceroughness,inadditiontoslidingspeedandappliedload.TheobservationsofSmith11,Bones8andDornfeld12pointtoasperitycontactasthemainsourceofAEactivityforlubricatedanddrycontacts.AssuchconstanttemperaturetestswereperformedbytheauthorstoaidAEsourceidentificationduringgearmeshing.Sincetheoiltemperatureswerekeptconstant,theoilviscosityandfilmthicknesswillremainconstantduringtheexperiment.ItwasthoughtthatthismayconfirmtheauthorsviewthattheAEsourceisduetofriction.Thegearboxwasrunat745rpmwithaloadof220Nmfor5hoursatwhichtimetheoiltemperaturestabilised.Thegearboxwasbroughttoastopandadjustedtono-loadLocalpressurepeak