文献翻译原文通过实验设计优化微注射成型工艺Word文档格式.docx
《文献翻译原文通过实验设计优化微注射成型工艺Word文档格式.docx》由会员分享,可在线阅读,更多相关《文献翻译原文通过实验设计优化微注射成型工艺Word文档格式.docx(12页珍藏版)》请在冰豆网上搜索。
专业:
机械设计制造及其自动化
学生姓名:
学号:
指导教师单位:
姓名:
职称:
2014年5月26日
1.Introduction
MicroInjectionMoulding(MIM)isarelativelynewtechnologywhichispopularintheindustryformicromanufacturebecauseofitsmassproductioncapabilityandlowcomponentcost.InordertoachievethehighestqualitycomponentswithminimalcostsusingMIMitisimportanttounderstandtheprocessandidentifytheeffectsofdifferentindependentparameters.OneofthemethodsthatcanbeemployedtoinvestigatetheoveralloperationofMIMisDesignofExperiments(DoE).Ingeneral,DoEcanbeusedtocollectdatafromanyprocessandgainanunderstandingoftheprocessthroughdataanalysis.Thisprocedurecanhelptooptimisetheprocessandeventuallyleadtoqualityimprovements.
Thispaperisorganizedasfollows.TheMIMprocessisdescribedinSection2.InSection3theDoEisintroduced.Thecollectionofexperimentaldataisexplainedinsection4followedbyresultsanddata
analysisinsection5.Thediscussionofresultsispresentedinsection6.Finallythepaperendswithconclusionsgiveninsection7.
2.MicroInjectionMoulding(MIM)
Micro-injectionmoulding[1]isarelativelynewtechnologyinthemanufacturingworld,andassuch,itneedstobethoroughlyinvestigated.AccordingtoMicro-powderinjectionmoulding,conductedbyLiuet.
al.[2],micro-systemtechnologywerewidelyusedinthenew21stcenturybecauseofitssuccessfulapplicationsinmanydifferentfields,e.g.influidic,medical,opticalandtelecommunications.Presentedwithmass
productioncapabilityandlowcomponentcost,maketheMIMtechnologytobeoneofthekeyproductionprocessesformicromanufacturing.TheComponentsofMIMfallintooneofthefollowingtwocategories:
TypeA:
Overallsizelessthan1mm
TypeB:
Microfeaturelessthan200um.
InitialworkonDoEanddataanalysisonMIM,conductedbyShaet.al.[3],primarilyfocusedontheanalysisof5differentfactors(themeltandmould
temperature,injectionspeed,pressureandflowstatus)affectingtheachievableaspectratiosinthreedifferentpolymermaterials.Theaspectratioistheratioofalongerdimensiontoitsshorterdimensionofaspeciallydesignedmicrofeatureforthisexperiment.TheirstudyconcludedthatMeltTemperature(Tb)andInjectionSpeed(Vi)werethekeyfactorsaffectingtheaspectratiosachievableinreplicatingmicrofeaturesinallthreepolymersmaterials.
TheeffectoftoolsurfacequalityinMIM,conductedbyGriffithset.al.[4],primarilyfocusedonthefactorsaffectingtheflowbehaviorandalsotheinteraction
betweenthemeltflowandthetoolsurface.
Thefindingsoftheseearlierinvestigationsaretakenintoconsiderationinthisstudy.
Fig1showsapictureofaMIMmachine.TheplanningofDoEandthedataanalysiswascarriedoutusingthestatisticalsoftwarepackage“Minitab16”.
3.DesignofExperiments(DoE)
ThetechniqueofdefiningandinvestigatingallpossibleconditionsinanexperimentinvolvingmultiplefactorsisknownastheDesignofExperiments.
ThetwotypesofDoEthatarewidelyusedaretheFactorialdesignandTaguchiMethod.AccordingtoMinitabdesignofexperiment[6],Factorialdesignisa
typeofdesignedexperimentthatallowsforthesimultaneousstudyoftheeffectsthatseveralfactorsmayhaveonaresponse.Whenperforminganexperiment,varyingthelevelsofallfactorssimultaneouslyratherthanoneatatime,allowsforthestudyofinteractionsbetweenthefactors.
Inafullfactorialexperiment,responsesaremeasuredatallcombinationsoftheexperimentalfactorlevels.Thecombinationsoffactorlevelsrepresenttheconditionsatwhichresponseswillbemeasured.Eachexperimentalconditioniscalled“run”andtheresponsemeasurementanobservation.Theentiresetofrunsisthe“design”.
Tominimizetimeandcost,itispossibletoexcludesomeofthefactorlevelcombinations.Factorialdesignsinwhichoneormorelevelcombinationsareexcludedarecalledfractionalfactorialdesigns.
Fractionalfactorialdesignsareusefulinfactorscreeningbecausetheyreducethenumberofrunstoamanageablesize.Therunsthatareperformedareaselectedsubsetorfractionofthefullfactorialdesign.ButRoy[7]mentionsthatusingfullfactorialandfractionalfactorialDoEmaycontributetothefollowingissues:
●Theexperimentsbecomeunwieldyincostand
timewhenthenumberofvariableislarge;
●Twodesignsforthesameexperimentmayyield
differentresults;
●Thedesignsnormallydonotpermit
determinationofthecontributionofeachfactor;
●Theinterpretationofexperimentwithalarge
numberoffactorsmaybequitedifficult.
Hence,Taguchimethodwasdevelopedinordertoovercomesomeoftheseissues.Taguchimethodisthetechniqueofdefiningandinvestigatingallpossibleconditionsinanexperimentinvolvingmultiplefactors.
TaguchimethodwasfirstintroducedbyDr.GenichiTaguchiaftertheSecondWorldWar[8,9].Hecameupwiththreebasicconcepts[7]:
1.Qualityshouldbedesignedintotheproductandnotinspectedintoit.
2.Qualityisbestachievedbyminimisingthedeviationfromatarget.Theproductshouldbesodesignedthatitisimmunetouncontrollableenvironmentalfactors.
3.Thecostofqualityshouldbemeasuredasafunctionofdeviationfromthestandardandthelossesshouldbemeasuresystem-wide.
Dr.TaguchisetupathreestageprocesstoachievetheenhancementofproductqualitybyDoEbasedupontheconceptsabove,namely,Systemdesign,Parameter
design,andTolerancedesign.
Forthefirststage,systemdesignistodeterminethesuitableworkinglevelsofdesignfactors.Itincludesdesignandtestofasystembasedonselectedmaterials,
partsandnominalproduct/processparameters.
Parameterdesignisforfindingthefactorlevelthatcanachievethebestperformanceoftheproduct/process.
Thelaststagewhichisthetolerancedesignistodecreasethetoleranceoffactorswhichissignificantlyaffectingtheproduct/process.
AspecialsetofarrayscalledOrthogonalArrays(OAs)wereconstructedtolayouttheexperiment.TheOAsimplifytheexperimentdesignprocess.Itisdoneby
selectingthemostsuitableOA,assigningthefactorstotheappropriatecolumns,anddescribingthecombinationsoftheindividualexperimentscalledthetrialconditions.
InthisstudyafractionalfactorialDoEwasconductedincombinationwithTaguchi’designconceptsforqualityenhancement.
4.CollectionofExperimentalData
Theexperimentwasdesignedandset-upasdefinedbySha,et.al.[10].Thisaimofthisexperimentistoanalysetheeffectsofsixfactorsontheachievableaspectratiosandfindthemostsignificantfactorsinordertoreachtheoptimalsettingswhichwouldgivethehighestaspectratios.Fig.2showsthetestpartwithmicrofeaturesintheformoflegswithtwolevelofwidth(W),200or500um,anddepth(D),700()or100um()wherethefeatureshavingthesamedepth,D1orD2,weregroupedononesideofthepart.
Threedifferentmaterials,namely,semi-crystallinepolymerssuchaspolypropylene(PP),polyoxymethylene(POM)andanamorphouspolymersuchasacrylonitrilebutadiene-styrene(ABS)wereinthisstudy.Theparametersinvestigatedwerebarreltemperature(Tb),mouldtemperature(Tm),injectionspeed(Vi),holding
pressure(Ph),theexistenceofairevacuation(Va)andthewidth(W)ofmicro-legs.
Theaspectratios,i.e.theratiosbetweenthelengthofthemicrofeatureandtheirdepths,D1orD2,aremeasuredduringtheexperiment.Theaveragevaluesof24measuredresponseswiththesameWandD(twoperpart)whileapplyingtheprocesssettinggiveninTable1areusedinthisstudy.
5.ResultsandDataAnalysis
A2-levelsixfactorsfractionalfactorialdesign(26-2)wasappliedinthisexperiment.TheDoEwasusedtoidentifythefactorsthatwereactiveandsignificanttostudythefillingofmicrochannels.ThepurposeofthisexerciseistolookattheresultsoftheDoEresponsesinordertounderstandtheprocessandselectthesignificantfactorswiththeirappropriatesettingswhicharenecessaryforoptimalperformance.
5.1.Results
ThemeasuredexperimentalresponsesfortheDoEfortheratiosbetweenthelengthofthemeltfillsandthedepthofthechannels,D1orD2arerecordedinTable2.ThevalueofD1andD2shownonthetablearetheaveragevaluesof24measurements.
5.2.DataAnalysis
Thestatisticalsoftwarepackage“Minitab16”wasusedtoanalysetheresultsobtainedfromtheexperiment.TheresultoftheanalysisforPPforboththecasesofD1andD2isgiveninTable3.
InTable3the“Effect”columnshowsthepositiveornegativeeffectofthefactoronthemeasuredresponse.Hencethehighertheeffectthemoresignificantthefactorinconsiderationwillbe.The“effect”columndeterminesthefactors’relativestrength,the“p-values”determinewhichofthefactorsarestatisticallysignificant.InthisstudythevaluesinthePcolumnoftheEstimatedEffectsandCoefficientstableareusedtodeterminewhichoftheeffectsaresignificant.Tomakeadecisionconcerningwhichfactorsaresignificant,furtheranalysisisnecessaryandthiswillbediscussedinthenextsection.Atypicalvalueforthesignificancewaschosentobe0.05throughoutthisstudy.
6.DiscussionofResults
TheaboveresultswereutilisedtoproducemoreevidencetosupporttheclaimsforstrongfactorswhichmatterthemostfortheMIMprocess.
Using=0.05,forPPD1,thep-valuesfoundforTbis0.038andViis0.009indicatethatthemaineffectsfromthesetwosinglefactorsTbandViaresignificant,i.e.theirp