中英文翻译Word文件下载.docx

中英文翻译Word文件下载.docx

《中英文翻译Word文件下载.docx》由会员分享，可在线阅读，更多相关《中英文翻译Word文件下载.docx（12页珍藏版）》请在冰豆网上搜索。

ChrisK.Mechefske&

QiaoSun

AbstractAstheautomotiveindustrybecomesincreasinglycompetitive,partsmanufacturersareunderextremepressuretoimprovethequalityoftheirparts,whileatthesametimereducingcosts.Themethodcurrentlyusedtodetectfailuresinautomotivelightassembliesduringvibrationendurancetestinginvolvesmanualinspectiononlyafterthetestiscompleted.Inmostcases,thismethoddoesnotallowthefirstonsetoffailuretobedetermined.Subsequentdamagecausedbytheongoingtestaftertheinitialfailureoftenobliteratesthecluestowhatfailedfirst,thusdiminishingtheusefulnessoftheinformationcollectedasinputtodesignimprovementiterations.Anadaptable,reliable,andlow-costreal-timemonitoringanddiagnosticsystemthatwouldinterruptthetestingoperationatthefirstonsetofafailureisdesired.Thispaperdescribesanaccelerometer-basedandamicrophone-basedmonitoringsystemforautomotivelightassemblyfailuredetectionduringendurancetesting.Thevibrationandacousticsignalanalysismethodsarealsodescribed.Preliminaryresultsfromthesetwosystemsshowasignificantdifferencebetweenhealthyandfaultyfoglightassemblies,whichcanbedetected.Thissuggeststhatthereispotentialtodeveloptheseandpotentiallyothermeasuresforuseinanadaptable,reliable,andlow-costreal-timemonitoringanddiagnosticsystemthatwouldinterruptthetestingoperationatthefirstonsetofafailure.Furthertestingisplannedtodeterminethesensitivityofthemethodsusedinthisstudytodetecttheearliestonsetoffailuredamageduringendurancetestingofautomotivelightassemblies.

KeywordsFailuredetection.Vibrations.Acoustics.Endurancetesting

1Introduction

Oneofthequalitycontroltestsforautomotivelightassembliesistosubjectthemtovibrationtestingonashakertable.Thesetestscanlastforaslongas24h.Tobeaccepted,theassembliesmustcompletethetestwithoutanybreakage.Currently,whenassembliesdobreak,itisnotpossibletoknowwhenthefirstbreakagehasoccurredandwhattheinitialfailureis（alltestsrunthefulltimespecifiedregardlessofwhenthebreakageoccurs）.Amonitoringsystemisrequiredthatwouldinterrupttheshakertableoperationatthefirstonsetofafailure.Inthisway,informationastotheinitialfailure（causeandlocation）canbeusedasusefulfeedbacktotheiterativedesignprocess.

Theobjectiveofthisresearchprojectistoinvestigatewhethervibrationandacousticmonitoringtechniquescanbeusedtodetectmechanicalfailuresinautomotiveheadlightandtaillightassembliesastheyundergoshakertabletesting.Afurtheraspectofthisstudyistoinvestigatedifferentmeasuredsignalanalysismethodsthatcouldbeusedtoautomatethefailuredetectionandinterruptthetestinprogress.AshakertablehasbeenconstructedatQueen’sUniversityinKingston,Ontario,Canada,whichmeetstheSocietyofAutomotiveEngineersguidelines[1,2],andhasenabledthecollectionoforiginaldata.Selectionofsuitablesensorshasbeencompleted[3,4].Analysisofthedatacollectedsofarhasshownthattheaccelerometer-basedandthedual-microphonemethodsweresuccessfulatfailuredetection.Aremainingchallengetobesolvedistheneedtoimprovethequalityoftheresultsbycontrollingthebackgroundnoise.

2Monitoringanddiagnosticsystems

Theaccelerometer-basedandmicrophone-baseddiagnostictestingsystemsweretestedinalaboratoryenvironment.Eachtestingsystemconsistsofthreemaincomponents:

（1）shakertable（mechanicalvibrator）,

（2）monitoringhardware（sensor,computer,etc.）,（3）diagnosticsoftware.

Theshakertableormechanicalvibrator（seeFig.1）isusedtogeneratevibrationswithafixedimpactstrengthandexcitationfrequency.ThetestconditionsandthedesignoftheshakertablefollowedspecificationssetbytheSocietyofAutomotiveEngineers[1,2].Theshakertableconsistsofa1.5HPelectricalmotor,aframe,acam,andafollowerframe/platform.Thetestedlightassembliesareattachedtothefollowerplatform.Duringthetesting,themotordrivesacamatarotationspeedof750±

10rpm,togiveavibrationexcitationfrequencyofapproximately12.5Hz.Theamplitudeoftheoscillationsis3.0±

0.25mm.Eachrotationofthecamforcestheplatform（andattachedlightassembly）upwardsby3.0mm（±

0.25mm）andthenallowsittofalldownwardsandimpactthebaseplate.ThedropisactuallyforcedbythefourspringsattachedtotheframeasshowninFig.1.Theshakertableisaduplicateofthatusedintheindustry.

Themonitoringsystemconsistsofthreemaincomponents:

sensors/transducers,datacollectionhardware,anddataanalysissoftware.Forthiscase,thesensorsusedwerethreeaccelerometersandtwomicrophones.ThedatacollectiondeviceusedwasanNIPXI-1031equippedwithanNI4472DAQcard.ThedataanalysissoftwarewasprogrammedinNILabview™usingtheNISoundandVibrationToolKit™.

Theautomotivelightassembliesarevibratedwiththeshakertable.Themeasureddynamicresponseoftheautomotivelightassembliesshouldchangewhenacrackorotherfailuredevelops,acomponentbecomeslooseordetached,orwhenrattlingisintroducedduetolooseorcompletelydetachedcomponents.Theselectionofthethreedifferentkindsoftransducersandthesetupofthemicrophone-basedandaccelerometer-baseddynamicdiagnosticsystemsaredescribedinthenextsection.

2.1Transducerselection

Sixcommonitemsneedtobeaddressedwhenselectingtransducers:

sensitivity,linearity,frequencyrange,maximumshock,weight,andmountingmethods.Inparticular,forthiscase,considerationwasfocusedonthefrequencyrange,maximumshock,weight,andmountingmethodsbecauseoftheextrememaximumimpactshockthattheaccelerometerswillbeexposedto.

Theworkingfrequencyrangeoftransducerswaschosentoincludethevibrationfrequenciesoftheshakertableandlightassembly,aswellasthefrequencyofvibrationsfromfailures.Themotordrivesacamthrough“V”beltsatarotationspeedof750±

10rpm,resultinginaprimaryvibratingfrequencyofapproximately12.5Hz.Thevibrationfrequenciesofthelightassemblieswithfailuresdevelopingareunknownbutareexpectedtobeabove12.5Hz.Thevibrationresponsesarenotdependentontheexcitationfrequencyoftheshakertablebutwillbeinarangeabove12.5Hzandupto10kHz.Accordingtoearlierexperiments[3,4],themaximummeasuredvibrationlevelwasfoundtobeapproximately151×

g（wheregistheaccelerationduetogravity）.Therefore,thechosentransducer’smaximumshockshouldbehigherthan151×

g,ifthetransducerisacontactsensor.

Anotherconsiderationinchoosingtransducersistheweight.Sensorsmountedonthelightassemblyaddextramasstothemeasuredsystem.Themassofasingletransducerandthenumberofthetransducersmayhavepotentialinfluenceonthecharacteristicsoftheautomotivelightassembliesandshouldbeconsidered.Consequently,thetotalmassoftransducersshouldbeaslightaspossibletoreduceconsequentinfluence.However,smallertransducersusuallyhaveloweroverallsensitivity（signaltonoiseratio）andhigherprices.The“ruleofthumb”relatingthemassratiobetweentransducerandthemeasuredsubjectsuggeststhattheratioofmassesshouldbeatleast1:

100.Thisvaluewasconfirmedthroughdiscussionswithsensormanufacturersandvendors.AlltheautomotivelampassembliesthatwereusedfortestingatQueen’sUniversityweighfrom0.87kg（foglightassemblies）to11.89kg（headlightassemblies）.Sothetransducerweightshouldbelessthan8.7g.

Mountingisalsoanimportantfactorthathastobeconsideredforthisproject.Accelerometermountinglimitationsincludeirregulargeometryofthelightassemblies,theplasticmaterialofthelightassemblies,thecurvedface,andlimitedaccessibilityforscrewmountings.Theonlysuitablemountingforthetransducersisadhesivemounting.Therefore,mountingprovisionforanypotentialtransducersmustbeadhesivemountingornoncontact.Duringtesting,oscillationamplitudeoftheshakertableisrelativelylarge（3.0±

0.25mm）,whichrequiresastrongadhesiveforgluingtransducersinplace.However,thetransducersneedtobeputonandtakenofffrequentlytomakedifferentmeasurementsfordifferentlightassemblies.Consideringallthese,anindirectadhesivemountingsystemisabetterchoiceforthecontactsensorsusedinthisproject.

Accordingtotheaforementionedselectionconditions,twokindsofminiaturetransducershavebeenselected.Figure2showsselectedtransducers.Theyarebothminiaturesensors.Table1liststhekeyfactors,suchasweight,maximumshock,mountingmethods,measurementfrequencyrange,andcostofthetwoselectedtransducers.

2.2Microphone-basedsystem

Astheprevioussectionpointedout,themicrophone-baseddiagnosticsystemconsistsoftransducermicrophones,anNI4472DAQcardandNIPXI1031datacollectorandLabview™andNISoundandVibrationToolKit™analysissoftware（SeeFig.3）.

Themicrophoneusedisaself-containedunitforacousticmeasurements.Itcomprisesamicrophone,preamplifier,preamplifierpowersupply,andpolarizationvoltagesupply.Themicrophonerespondstothesoundpressurecausedbyvibrationoftheautomotivelightassemblieswhenitisvibratedontheshakertable.ThegeneratedelectricalsignalfromthemicrophoneisthencollectedandamplifiedbyNI4472DAQcardandpreliminarilyanalyzedusingaLabviewprogram.Thesamplingrateis40,000samplespersecondwithdatacollectedover2-speriods.Theshakertableusuallyr