外文翻译--一个激光束加工（LBM方法）数据库的切割瓷砖 英文版.pdf

JournalofMaterialsProcessingTechnology84(1998)4755Alaserbeammachining(LBM)databaseforthecuttingofceramictileI.Black*,S.A.J.Livingstone,K.L.ChuaDepartmentofMechanicalandChemicalEngineering,Heriot-WattUni6ersity,Riccarton,EdinburghEH144AS,UKReceived13December1997AbstractThispapercoversthecuttingofcommercially-availableceramictilesusingaCO2lasercuttingmachine,withtheobjectofproducingalaserbeammachining(LBM)databasethatcontainstheessentialparameterinformationfortheirsuccessfulprocessing.Variouslasercuttingparameterswereinvestigatedthatwouldgenerateacutinceramictilewhichrequiredminimalpost-treatment.Theeffectsofvariousshieldgases,ofmulti-passcuttingandofunderwatercuttingwerealsoexamined.©1998ElsevierScienceS.A.Allrightsreserved.Keywords:CO2；Lasercutting；Ceramicmaterials；Advancedmanufacturingprocesses1.IntroductionandbackgroundManualmethodsofcuttingceramictilesareverysimilartothatforglass,i.e.scribingthematerialswithtungsten-carbidetippedcutter,followedbytheapplica-tionofabendingmomentalongthescribedlinetoinitiatecontrolledfracture.However,manualtech-niquesarelimitedtostraight-linecuttingandrelativelylarge-radiuscuts.Internalandundercutprofilesarenearlyimpossibletoproducewithscoringalone(withthepossibleexceptionofinternalcircles)；moresophis-ticatedmethodshavingtobeappliedtoachievetheseprofiles.Traditionally,diamond-saw,hydrodynamic(waterjet)orultrasonicmachiningareusedtocreatecomplexgeometriesinceramictiles,buttheseprocessesareverytimeconsumingandexpensive.Forexample,typicaldiamond-sawcuttingspeedsareintheorderof20mmmin11,whileultrasonicdrillingofAl2O3takesover30sperhole2.ThemostcriticalfactorarisingfromuseofaCO2lasertocutceramictilesiscrackdamage,whichisessentiallycausedbyahightemperaturegradientwithintheceramicsubstrateduringthecuttingprocess.Thesecracksreducethestrengthandaresourcesforcriticalcrackgrowth,whichmayresultinpartialorcompletefailureofthetilesubstrate3.Thusareduc-tionofprocess-inducedcrackformationisparamountfortherealisticcommercialuseoflaserstocutceramictiles.2.LasercuttingparametersLasermachiningofanymaterialisacomplexprocessinvolvingmanydifferentparametersthatwhichallneedtoworkinconsorttoproduceaqualitymachiningoperation4,parameterssuchas:(i)laserpowerinput；(ii)focalsetting；(iii)assistgastypeandpressure；(iv)nozzleconfiguration；(v)workpiecethickness；and(vi)optophysicalproperties.Previousresearchwithintheauthorsdepartment1,5,6hasalsodemonstratedthecriticalityoftheaboveparametersinefficientlasercutting.2.1.LaserpowerLaserpowerdependsonthetypeoflaserused.Fortheworkreportedinthispaper,aFerrantiMF400CNClasercutterwasemployed,ratedatapoweroutputof400W.However,duetoupgrading,themaximumbeampowerachievablewasbetween520and*Correspondingauthor.Fax:441314513129；e-mail:i.blackhw.ac.uk0924-0136:98:$-seefrontmatter©1998ElsevierScienceS.A.Allrightsreserved.PIIS0924-0136(98)00078-8I.Blacketal.:JournalofMaterialsProcessingTechnology84(1998)475548530Wincontinuouswave(CW)cuttingmode.Thelaseralsohadtheabilitytoworkinpulsemode(PM)andsuper-pulsemode(SPM；Fig.1).Todeterminetheequivalentpoweroutputduringpulsingoperation,apowerversespulsingchartwasusedinconjunctionwiththefollowingbasicequation9:PrPl:Psf1:(PlPr)Althoughthelasercuttercouldoperatebetweenfre-quenciesof50and5000Hz,avalueof500Hzwasrecommendedinpreviouswork1,5.Sincethissettingprovedtobesuccessful,onlylimitedinvestigationintootherfrequencieswascarriedout(at250Hz,750and100Hz).2.2.CuttingspeedTheCNCtableusedwiththeFerrantiMF400lasercutterhadamaximumfeedrateof10000mmmin1.Previouswork6indicatedthatfeedratesabove6000mmmin1provedtobeunstableforanystandardisedtesting.Theoptimumcuttingspeedvariedwiththepowersettingand,moreimportantly,withthethicknessoftheworkpiece.2.3.ShieldgastypeandpressureCompressedair,argon,nitrogenandoxygenwereusedasshieldgasesduringcutting,withpmax:4bar.Differentshieldgaseswereusedtoexaminedtheireffectoncutqualityafterprocessing,sincetheshieldgasnotonlycoolsandcutedgesandremovesmoltenmaterial,butalsogeneratesachemicalreactionwiththesub-stratematerial7.Theresultsofthischemicalreactiondifferforeachtypeofshieldgasused.Fortestpurposespwasvariedinstepsof0.5barfrom1to2.5bar,theninstepsof0.2barfrom2.6bartothemaximumattainablegaspressure.2.4.NozzleconfigurationThenozzlediametercontributesdirectlytothemaxi-mumachievablegaspressureandhencetothemassflowrateofthegaswasimportantfortheeconomicsofcutting,especiallywhenusingcylindersofargonandnitrogen.Onlycircularprofilesforthenozzleexitswereavailable(0.6mm5Ns520mm),butthisuniformnozzleexitgeometryallowedcuttinginanydirection.2.5.NozzleheightandfocalpositioningTheheightatwhichthenozzlewassetwasgovernedbythepositionofthefocalpoint.TheFerrantiMF400lasercutteronlypossessedalongfocallengthof110Fig.1.Cuttingmodes.mm(originallyashortfocallengthof46mmwasavailablebeforeupgrading)andthislengthcouldbealteredby95mm.Ifthenozzleheightwasincorrectlysetthebeamwouldclipthenozzleandreducetheequivalentpoweroutputtotheworkpiece6.Forthebulkofthetestingthefocalheightwassetsothefocalpointwasonthejob,i.e.onthetopsurfaceoftheworkpiece.Thisconditionobviouslygovernedtheposi-tionofthenozzleabovetheworkpiece.3.ExperimentalprocedureSixtypesofSi:Al2O3-basedceramictileswereexam-ined(Table1),originatingfromdifferentcountries.Notethatthecompositionofthetilesvaried,asdidthethickness,butallpossessedasurfaceglazeandinthecaseofthe7.5,8.6and9.2mmSpanishtilestheglazewasdoublelayered.3.1.Set-upprocedureSincetherewasaneedforstandardtestingcondi-tions,thefollowingprocedurewasimplementedbeforethestartoftesting:(i)thebeampowerwasvalidatedtospecification,i.e.520530Wdevelopedatfullpower(CW),althoughthisdroppedtoaround50WafterTable1Typesofceramictileusedts(mm)TiletypeBodycolour3.7BrazilianWhite4.7WhitePeruvianLightredItalian5.2SpanishRed5.74Spanish7.5RedRedSpanish8.69.2RedSpanishI.Blacketal.:JournalofMaterialsProcessingTechnology84(1998)475549about1hoftesting；(ii)thenozzleandthefocallenswerecheckedtoensurethattheywereingoodcondi-tion,i.e.cleanandundamaged；(iii)theshieldgaspressureregulatorandshieldgastankswereturnedontopreventdamagetothefocallens；(iv)thelaserbeamwascentredwithinthenozzleusingasquaretest,alowerenergyinputinPMbeingusedtocutasquareonamildsteel,thesparkingdensitythatresultedfromcuttingbeingcheckedtoseeifitwasequallydistributedaboutthecutline；and(v)thefocalpointwassetforitsdesiredpositioning,i.e.onthejob.3.2.TestingAstraight-linetest(SLT)wasusedtoevaluatethevariablelaserparametersforfullthrough-cutting(FTC).Angularcuttingwasconfiguredtoinvestigatehowthematerialreactedduringcuttingoftightgeome-try.Circulartestingandsquaretestingweredevisedtodeterminetheeffectsresultingfromcuttingvariousgeometries.TheSLTallowedforthecombinedtestingoftwoseparateparametersononetestpiece,uponcompletiontheresultsbeingpresentautomaticallyinacuttingmatrixintheformoftheresultingcuts.PandVarethemostimportantlaserparameters,astheydictatetheamountofenergyinputperunitlengthofcut,thereforetheywerepairedfortheSLT,aswerepandNSwhichgovernthemassflowrateoftheshieldgas.FortheP:Vtestruns,thepowerwasheldconstantwhilethecuttingspeedwasincreasedalongthecut(Fig.2(a).ThelengthofcutatconstantcuttingspeedhadtobeofsufficientmagnitudetoaccommodatetheaccelerationordecelerationoftheCNCtablebetweenfeedchanges:previouswork6indicatedthat50mmwasadequate.Interpretingtheresultswasmadeeasierduetotheirtabularformat,withthecuttingmatrixshowingclearlyanytrendsorpatternsoccurringduetothechangesinparametersettings.TheSLTalsoal-lowedalargenumberofcutstobecarriedoutoverashorttime-frame.Thisprovedadvantageous,asthelasertendedtodriftfromitsinitialsettingswithtime.Precautionshadtotakentoavoidlocalisedheatinginthetilefromcontinuouscloseproximitycutting,asachangeintilebodytemperaturewouldinvalidateanyresultingdata.Initially,a20mmseparationbetweencutswasusedandthisprovedsufficient.Inordertostudyhowclosethecutscouldbemadetoeachother,theseparationbetweencutswasreducedbyincrementsof2mmfromaninitial20mmspacing.DuringtheSLTtheotherlaserparametershadtobeheldconstant6.ForPversusV,fwasheldat500HzwithNS1.2mmandp3bar.Thebeamfocalpointremainedonthejob.TheresultsfromtheP:VcuttingmatrixdeterminedthefixedvaluesforthecuttingspeedandpulsesettingsforthesucceedingSLT.FortheNS:pFig.2.Testingconfiguration:(a)straight-linetesting；(b)angulartesting；(c)circulartesting；(d)squaretesting.cuttingmatrix,thenozzlesizeremainedconstantalongthex-axis(refertoFig.2(a)whilepwasincreasedinstepsof0.2barfrom2barinthey-axis(thecutseparationremainedconstantat20mm).Anewmatrixwascreatedsubsequentlyforeachnozzlesize.Angulartesting(Fig.2(b)wasusedtoinvestigatehowthecutmaterialreactedtosustainedexposurefromthelaserbeamduringthemachiningoftightgeometries(i.e.whereseveralcutsaremadeincloseproximitytoeachother).TheproximitytestmentionedforSLTdetermineshowcloseparallellinescanbecuttoeachother,whereasangulartestingisusedtodeter-minehowthecuttingofacuteangleseffectsthecutquality.Theanglescutfromaworkpiecewerereducedfrom45°to10°andthecorrespondingsurfacefinishquality(SFQ)wasnoted.I.Blacketal.:JournalofMaterialsProcessingTechnology84(1998)475550Table2Multi-passcuttingparametersPlCuttingmodePsNo.ofpassesLastcutCW60FTC9000100SPMFTC100Table3GradingofSFQGrading1Nocrackinginsurfaceglaze,solidsharpcutedgeMinimalglazecracking(WcB2mm)withslight2lossofsharpnessincutedgeMediumcracking(2mmBWcB4mm)andslight3damagetounglazedtilesubstrateSignificantdamagetoglazecoating(Wc6mm),4heavydamagetounglazedsubstratecausingflakingintheglazedsurface5Sameas4butwiththeformationofcracksinthetilesmainbodyleadingtostructuralfailureinapartofthetile(usuallyattheendofacutorwithin8mmofthetileedge).Therearetworeasonsforconductingsquareandcirculartesting(Fig.2(c)and(d):first,todeterminetheoptimummethodoflaser-beamintroductiontointernalcutprofiles；andsecondly,todetermineiftherewasanylimitationinthedimensionofthesizeofsquareorholecut.Ifnotcorrectlyintroduced,thelaserbeamwouldcauseaninternally-cutprofiletofailatthepointofintroduction,duetothebriefbutexcessivethermalgradientinducedfromcutting(i.e.thermalshock).Therefore,utilisingmethodsofbeamintroduc-tion,suchastrepanning,ontoaprofileenabledcom-plexgeometriestobeinvestigated.Whatalsobecameapparentduringtestingwastheimportanceofthepositionofbeamextractionfromthecutprofileandthepositionofthebeamstartingpointrelativetothegeometry,i.e.whetheritwasatacorneroronastraightedge.3.3.Multi-passandunderwatercuttingMulti-passcuttingwasbegunwithalowpower(P100W)laserbeam.Thefirstpassproducedawelldefinedblindkerfinthesubstrate,followedbyasecondpasstocutdeeperandsoon.Theprocesswasrepeateduntilthekerfwasabout20mmdeepandthenthelaserpowerwasswitchedto500WanddothefinalFTC.Theobjectiveofmulti-passcuttingwastoreducether-maloverloadbyuseoflessinputenergyperunitlength.TheparametersusedinthistestaregiveninTable2.Underwatercuttingwasconductedwiththeobjectiveofreducingtheinfluenceofheataroundthecutareaandalsotoexaminedtheeffectoncutqualitythroughacceleratedheatdissipationusingwater8.Thece-ramictilewasplacedunderwaterandthenozzlewasalsodippedinwater,theshieldgaspressurepreventinganywaterfromenteringthenozzlejetchamber.4.CutqualityMaterialproperties,laserparametersandworkpiecegeometryhaveasignificanteffectonthefinalresultofthelasercuttingprocess.Cutqualityisessentiallychar-acterisedbysurfaceroughnessanddrossheight,whereascracklengthdictatesthestrengthreductioninthesubstrate(Fig.3).TheoverallSFQattheglazesurfacewasclassifiedaccordingtothegradingscalegiveninTable3.Therefore,thequalityofthecutsurfaceandedgeweremeasuredwithrespectedto:(i)surfaceroughness；(ii)surfacefinishand；(iii)drossadherence.Fig.3.Qualitycriteriaforthelasercuttingofceramictiles.