Modeling-indexing-phase-of-borg-warner-gearbox-syn

MODELLINGINDEXINGPHASEOFBORG-WARNERGEARBOXSYNCHROMESHMECHANISMИССЛЕДОВАНИЕПРОЦЕССОВОТЦЕПЛЕНИЯСИНХРОННЫХПЕРЕДАЧТИП«BORG-WARNER»PhDst.LovasL.1,Prof.Dr.Sci.Dir.PlayD.2,Prof.Dr.MárialigetiJ.1,Prof.Dr.RigalJ-F.3BudapestUniversityofTechnologyandEconomics1,Budapest,Hungary-FederalMogulOperationsFrance2,Veurey-Voroize,France-INSA-Lyon3,Lyon,FranceE-mail:[e-mail],[e-mail],[e-mail],[e-mail]:Thispaperdealswithmodellingofindexingphaseofmanualgearboxsynchromeshmechanisms.Afterthepresentationofthestateoftheart,anewapproachofmodellingisdescribed,basedonthermalandmechanicaldeformationsofthesynchroring.Dynamicalbehaviourofthegearchangingmechanismisincludedinthemodel.Datafromnumericalsimulationsfitwellthemeasuredones,confirmingthistypeofapproach.KEYWORDS:SYNCHROMESH,DOUBLEBUMP,NUMERICALSIMULATION,GEARBOX1.INTRODUCTIONGearchanginginmanualgearboxesisacomplicatedprocess.Synchromeshmechanismsinsurethesefunctionsandfromaglobalpointofview,theysatisfiedautomotiveengineersinthelast30years,solittleattentionwaspaidtoitsstudy.Today,increasingpowerofengines,moresophisticatedusercomfortrequirements,androbotizationofmanualgearchanging,needmoredetailedstudiesofgearchanging.Highchangingforcesareneededbothduringsynchronisationandduringindexingphase.Theindexingphasehappensjustafterthesynchronisationphase,whenthesleevepassesthroughthesplinesofthesynchroring,andheadstotheclutchingteethofthegeartobeengaged(Fig.1).Fig.1SketchofpartsduringindexingphaseFig.2Measureddata,gearchanging1-2Duringthisoperation,thephenomenoncalled“2ndbump”[7]or“doublebump”[2,7]takesplace.Itisvisualisedbyaforcepeakinthediagramwheremeasuredaxialforceappliedonthesleeveispresentedversustime(Fig.2).Handlingitbecomesaseriousproblem,becauseinourknowledge,neitherthepredictionofitsoccurrencenorthecalculationofitsamplitude,arepossible.Sometimes,peaksmaybehigherthanthesynchronisationforcelevel.Designerscannotreachtheneededcomfortlevelrequirement,andtheyhavedifficultieswithsizingofrobotizedchangingmechanism.Inthispaper,doublebumpissupposedtobealwayspresent,andonetriestodefineitsamplitudebasedonmeasureddataandnumericalsimulationresults.2.STATEOFTHEARTManyauthorsmadevarioustheoriesaboutoriginsofdoublebump.Gotoetal.[4]simulatedthetorsionalbehaviourofageartrain.Theirinvestigationsshowedthedependencybetweenbumpandtangentialclearancesinthegeartrainchainsituatedafterthesynchromeshmechanism.Theysupposedthatclearancesprovokeangularvelocitydifferences.Therefore,todecreasethebumpforcepeak,thetangentialclearanceshavetobeminimised.ShinbataandNakamura[11]performedasimilarstudybysupposingthatthedoublebumpcomesmainlyfromgearcontactsidechangingduetotangentialclearances.Theydefinedtimeconditionsallowingavoidingtheformationofdoublebumps.Ifsidechanginghappensaftertheengagementofthesynchroringandbeforetheengagementofthegear,thenthebumpwillbehigh.Thebumpmaybedecreasedbychoosinganappropriatevalueofaxialforce.Moreover,itdisappearsforagivenrangeofaxialforceifthegearsplinechamferangleishigherthan68°.Kimetal.[5]havedifferentideasabouttheoriginsofthedoublebump.Theyusedadetailedmodelfromthegearlevertothegear,inordertosimulategearchanging.Theysupposedthatlinkingmechanismconnectingthegearlevertotheshiftforkbecomescompressedduringsynchronisation.Thenitisreleased,andtherelatingimpulsiveforceproducesashockbetweenthesleeveandthegearsplinechamfer,whichisfeltasthedoublebump.Itisrecognisedthattheamplitudeofthebumpdependsontherelativepositionofsleeveandgearsplines,thusitisconsideredasarandomevent.D’Orazioetal.[2]madeamulti-bodymodelofgearchangingsimilartothatofKimetal.Theymodelledthelinkingmechanismwithclearancesandtookintoaccountforkelasticity,butneglectedgeartraintangentialclearances.BearingandoilchurninglossesTangentialvelocityofthegearSleevesplinesSynchroringsplinesGearsplinesNNAxialvelocityofthesleeveTangentialdisplacementofthegearFtgFtgsynchronisationindexingphasewerealsoincluded.Frommeasurementsandnumericalsimulationstheydeducedthatthedoublebumpforcedecreaseswithincreasingsynchronisatingaxialforceasitisshowninref.[11].Theysupposedfirstlythatlargedistributionofbumpforcevaluescomesfromtherandomrelativechamferposition,andsecondlythatthestartingphaseofgearchanginghasaneffectonthedoublebump.3.THEDEVELOPEDMODELInthispaperanewapproachispresented,comparedtotheprecedingmodels.Generalconditionsareasfollows:•Inertiaofthesynchronisedpartsismuchsmallerthanthatofthesynchronisingparts,•Therearenoclearancesinthegeartrain,•Thereisnoangularvelocitydifferencebetweenthesynchroringandthegearaftersynchronisation.Thebasicassumptionisthatthesynchroringisstuckatthesynchrogearconeattheendofthesynchronisationphase.Thishasbeenverifiedbyvisualobservationsduringtestswithvideocamera.Thereforeaninitialpressureexistsatthebeginningoftheindexingphase.Toovercomethispressureandtoseparatetheringfromthegearcone,thefollowingtorqueisneeded:M=µ·p·A·r1Whereµ–frictioncoefficientbetweenconicalsurfacesoftheringandthegear,p–pressure,A–conicalsurface,r1–meanradiusofthecone.InthisequationAandr1arecons