外文翻译--三轴和五轴表面仿形铣削加工 英文版.pdf
ThreeandfiveaxesmillingofsculpturedsurfacesR.Baptistaa,*,1,J.F.AntuneSimoesb,2aInstitutoSuperiorTe´cnico,TechnicalUniversityofLisboa,Libson,PortugalbEscolaSuperiordeTecnologia/Inst.Polite´cnicodeSetu´bal,SchoolofTechnology/PolytechnicInstituteofSetu´bal,Setu´bal,PortugalAccepted17December1999AbstractTheaimofthisworkistheanalysisoftheinfluenceofthemillingparametersonthesurfacefinish.Thisworkisofinteresttobothscienceandindustry.Ononehand,ittriestoreducethemanualpolishingtime,whichrepresentsahighpercentageoftheproductiontimeofmoulds.Ontheotherhand,whiletheliteraturemainlyreferstosteelmoulds,thisworkwascarriedoutonaluminiumbecauseofitsadvantagesformachininganditssuitabilityformoulds.Insmallbatchproductionthemanufacturersareinterestedintheproductionofaluminiummoulds.A5axisCNCmillingmachinewasusedforthesurfacefinishingofseveralparts,previously3axismachined,withinequalcuttingconditions.Differentmachiningconditionson3and5axesfinishingoperationsweretested.Thecomparisonoftheresultsallowstoconcludethatabettersurfacefinishisachievedwith5axismillingusinganendmillinclinedinthefeedinsteadofthetraditional3axismillingwithaballnosecutter.TheapplicationoftheDesignofExperimentstechniquetogetherwithmultiplelinearregressionenablestheestablishmentofamathematicalmodeloftheprocessthatgivestheprocessparametervaluesthatleadtotheminimummachiningtimeinordertoachieveacertainroughness.Keywords:Threeandfiveaxesmilling;Tooltypeandorientation;Surfacefinishandmachiningtime1.IntroductionThedevelopmentoftheaeronauticandautomobileindus-triesbroughtnewtechnologicalchallenges,relatedtothegrowingcomplexityoftheproductsandthenewgeometriesmodelledinCADsystems.Thesemorecomplexgeometriesimposenewchallengingmanufacturingsituationsforthemillingofmouldsandstampingtools.Thisisanimportantreasonforthedevelopmentofnewmillingtechnology,namely5axismilling.Thestudyofthesenewmillingtechniqueisveryimpor-tantinordertoidentifyinwhatconditionseachoneofthesetechniquescangivethebestresult.Manualpolishingofsculpturedsurfacesisalmostneces-sarytoobtainthedesiredsurfaceroughness.Thiskindofoperationisaconsequenceoftheinabilityofthemillingprocesstoachieveeconomicallytherequiredsurfacequalityandinourdaysismoreandmoreaffectedbythedecreasingnumberofqualifiedworkers.The5axismillingis,now,oneofthetechniquesthatcangiveenormousadvantagesinmachiningoperations,potentialforsignificantreductionoftherequiredmanualfinish.Inthispaper,theroughness(Ra)ofthemachiningsurfaceisrelatedtotheoperatingparameters,appealingtoamath-ematical-statisticmodel.Thepartgeometrywasselectedinordertorepresent,inablockofsmalldimensions,allthedifferentgeometriesthatcanbefoundinmouldsandstampingtools.Theequationsrelatingtheroughnessleveltothemachin-ingconditions(feedpertoothandstepover),onparticularpartarea,weredetermined.Asexpected,thedevelopedmathematicalmodelexpressesadequatelytheexperimentalresultsandenablestheutilisationofanoptimisationcriter-ion,machiningtimeinthiscase.Thisfactemphasisestheimportanceofthemathematicalmodelstoestablishthemachiningparametersaimingtheoptimaluseoftheresources.TheexperimentalworkplanningandtheanalysisoftheresultsweredoneaccordingtotheDesignofExperimentsmethod.2.ExperimentalworkThisworktargetsaspecificindustrialproblemthefinishingofsculpturedsurfacebymillingoperation.TheJournalofMaterialsProcessingTechnology103(2000)398403*Correspondingauthor.1M.S.O.,AssistantProfessoroftheIST.2C.P.,AssistantoftheEST.workcarriedoutconsistedonthecomparisonoftheper-formanceof5axismillingcomparativetothe3axismillingofsculpturedsurfaces.Alloftheprocessvariableshavebeenidentifiedandthemostrelevantoneswereselectedforstudy.Thefinalobjectiveistoquantifytheirvaluesinordertoobtaintheminimummachiningtimeforafixedroughness.2.1.IdentificationandselectionofprocessvariablesTheprocessparametersselectedfortheexperimentalworkwerethestepover,thefeedpertooth,thefeeddirectionandthe3andthe5axesmillingtechnology.Asthemillingmachinehadalowrotationlimitof4.200rpm,andthetoolcouldworkinhigherspeed,itwasdecidedtokeepthecuttingspeedconstant.Also,thetooldiameterwaskeptconstantandasmalldiameterof10and12mm,suchasoccurswithindustrialpractice,werechoseninordertoavoidgeometricalinterferencebetweenthetoolandthepart.Thecuttingdepthwasalsomaintainedconstantnotonlytoreducethenumberofvariablesbutalsobecauseinfinishingoperationsthedepthisalwayssmall.2.2.EquipmentThemillingoperationswereperformedona5axisCNCmillingmachine(HermleUWF1202H)equippedwithanHeidenhainTNC425CNCcontroller.MasterCamversion5.54wasusedtocreatetheCNCpartprograms.Multisurfmodulefor3axisandFlowlinemodulefor5axis,wereused.3.Experimentalset-upTheexperimentalworkwasdividedintotwophases.Thefirstphasewasthe3axismillinganalysis,withaballnosemill.Thesecondphasewasthe5axismilling,withtwokindsoftools,aballnoseandanendmill.Thecontrolparametersconsideredwerethesurfaceroughness,dimensionalaccuracyandshapeofthepartgeometry,andmachiningtime.Thepartsgeometryhasfourdifferentshapes:aconvexshape,aconcaveshape,anhorizontalplaneshapeandaninclinedplaneshape(Fig.1).Althoughthisstudyisabout5axismilling,theparthadacurvatureonlyinonedirection(thedoublecurvaturewillbringonlymorecomplexity,andthismodelcanalsobeappliedinthatcase).So,itwasdecidedtoinvolveonlyonerotationaxis(AaxisrotationoverthelinearXaxis).Thepartsmaterialwasanaluminiumalloyofthe7.000series,usedinthemanufacturingofplasticsinjectionmouldsforsmallandmediumseriesproduction.3.1.ThreeaxismillingInitially,theimportanceofthestepoverandthefeedpertoothparametersonfinishingoperationswereinvestigatedbythreeaxismilling.Afterthisinitialstudy,thefeeddirectionimportancewasanalysed.Thisparameterwasstudiedconsideringthreedifferentdirections:(a)thedirectionof08;(b)thedirectionof458;(c)thedirectionof908(contouring)(Fig.2).Asthenextpictureprintsout,oneofthecharacteristicsin3axismillingoperationsisthatthelead/lagangleisnotconstant.Thisangleisdefinedbetweenthetoolaxisandthesurfacenormal.On3axismillingoperationsitispossibletoidentifythreedifferentcuttingtypes:thedownwardcutting,theplanecutting,andtheupwardcutting,Fig.31.Thecuttingspeedisnotconstantalongthemachiningpath.Fig.3alsoshowsthatthecuttingspeeddistributioninthetooledgeisdifferentforeachoneofthethreecuttingtypes.3.2.FiveaxismillingOnthefinishingoperationsby5axismillingthelead(bf>0)/lag(bf<0)angleimportance(withdifferentpositivevalues)tothequalityofthemachinedsurfacewasanalysed.Theresultsofaballnosemillinoppositiontoanendmillwerealsocompared.Fig.1.Surfacepart.Fig.2.Feeddirection.R.Baptista,J.F.AntuneSimoes/JournalofMaterialsProcessingTechnology103(2000)398403399TheCAD/CAMsystemusedcreatesthe5axisCNCpartprograms,withthetoolaxisperpendiculartothesurface(Fig.4).Usually,onmillingoperationswithballnosemill,themaximumroughnessvalueoccursalongadirectionperpen-diculartothefeeddirection.So,replacingtheballnosemillbyanendmillinclinedinthefeeddirection,itwillbepossibletoreducethescallopdimension,andthatisdirectlyrelatedtothesurfaceroughness(Fig.5).Inthe5axismillingexperiments,thelead/lagangle(bf)valueswerechosentakingintoaccountthevaluesrecom-mendedintheliterature(theballnosemill,with158;theendmill,with48)(Fig.6)1.4.ExperimentalworkandanalysisofresultsFirstsomeexperimentswerecarriedoutinordertodefinethebestrangeofvaluesfortheprocessparameterstobeinvestigated.4.1.Preliminaryexperiments3axismillingInitially,thestepoverandthefeedpertoothvalueswereselectedfromthetoolcatalogue.However,theroughnessresultswerenotsatisfactorybecausetheywerenotindust-riallyacceptableforfinishingoperation.4.1.1.StepoverandfeedpertoothAnothergroupofexperiments(theresultsareshowninTables1and2fortheinclinedplaneshape,Fig.1)withdifferentvaluesforstepoverandfeedpertoothweredefinedlater.Theresults(feedpertoothandstepover,respectively)forthemachiningtimeandtheroughnessrepresentedinFigs.7and8showmoreclearlythatuntil1mmroughnessthereisnoimportantaggravationofthemachiningtime;Fig.3.Lead/lagangle(bf).Fig.4.5Axismillingoperation.Fig.5.(a)Ballnosemill(bf0);(b)Endmill(bf>0)(Thetoolshasthesamediameterandstepover.)Fig.6.(a)Inclinationoftheballnosemill(bf158);(b)inclinationoftheendmill(bf48).400R.Baptista,J.F.AntuneSimoes/JournalofMaterialsProcessingTechnology103(2000)398403therearedifferentvaluesofstepoverandfeedpertoothleadingtothesameroughnessvalue,butwithdifferentmachiningtime;forthestepoverandfeedpertoothparameters,thereisalowerlimitwhereadramaticaggravationonthemachin-ingtimebegins.Processmodel.Withthelastresultsfortheinclinedplaneshapeaprocessmodelwasdeveloped.RoughnessRa0:7810:4676logStepover1:2436logFeedpertooth(1)Eq.(1)relatestheroughnesswiththestepoverandthefeedpertooth.Thisequationwasobtainedbyamultiplelinearregression(thelogarithmicapproachdisplayedthebestresults).Machiningtime921:94=Stepover74:98Feedpertooth310:8(2)Eq.(2)relatesthemachiningtimetothestepoverandfeedpertooth.Itisanapplicationoftheuniformmovementequation.Afterthedefinitionoftherequiredroughnessvalue,inthiscasea1mmroughnesswaschosen,theprocessvariablesfortheminimummachiningtimewereoptimised(Eq.(3),Fig.9).Table1Averageroughness,Ra(mm)Stepover(mm)Feedpertooth(mm/tooth)0.0300.0400.0500.0600.0800.1320.1390.1700.0630.290.550.1000.461.150.1550.471.360.2000.630.700.751.630.2500.850.850.901.420.2750.981.430.3001.081.110.3251.480.3501.270.3751.640.4401.551.631.710.5001.940.6002.270.7003.10Table2Machiningtime(s)Stepover(mm)Feedpertooth(mm/tooth)0.0300.0400.0500.0600.0800.1320.1390.1700.0631.5606250.1007402390.1556751470.200482305250930.250390254208960.275223880.3003251720.325750.3502800.375650.440225134860.5001040.600870.70074Fig.7.Stepover(mm).Fig.8.Feedpertooth(mm/tooth).Fig.9.Minimummachiningtime.R.Baptista,J.F.AntuneSimoes/JournalofMaterialsProcessingTechnology103(2000)398403401