ofg,isprocess.ultrasonictheultrasonicKeywords:Ultrasonicmachining；Turning；Finiteelementmodelling；Microstructuretheexistent,conventionalturning(CT)technology.frequencyultrasonicvibration,superimposedontheconventionalmovementofthecuttingtool(Fig.1),hasprovedtobeeectiveinmachiningintractablemetalalloysaswellasbrittlematerials,suchasceramicsandcuttingforces7,8.trolsystem.Thissystemstabilisestheturningprocesswithultrasonicvibrationandmakesthisprocesshighlycontrollable.Thedetaileddescriptionofthisnovelcontrolsystemisgivenin9,10.TheexperimentalpartofthispaperstudiesUATwithautoresonantcontrolincomparisontoconventionalturning.AnotherimportantissueconcerningUATisthe*Correspondingauthor.Tel.:+44-1509-227504；fax:+44-1509-Conventionalmachiningofmodernnickel-andtita-nium-basedsuperalloys,usedinaerospaceapplications,causeshightooltemperaturesandsubsequentfastwearofcuttingedgesevenatrelativelylowcuttingspeeds.Agrowingdemandformachiningtheseintractablematerialsrequiresnewadvancedturningtechnologies.Suchatechnologywasintroducedin1960s:high-Nevertheless,uptothepresentdayUAThasnotbeenwidelyintroducedintoindustrialenvironment.ThemainreasonforitissensitivityoftheUATprocesstotheloadappliedtothecuttingtip,resultinginthelossofcuttingeciencywhentheloadchangesoradierenttipisused.However,thislimitationhasrecentlybeeneliminatedwiththeinventionoftheautoresonantcon-1.IntroductionTurningisamachiningprocess,whereathinsurfacelayerofthetreatedmaterialisremovedfromawork-piecebyasharpwedge-shapedcuttingtoolformingacylindricalsurface.Thistechnologyhasbeenusedforcenturiesmainlyforcuttingvarioustypesofmetallicmaterials.However,intherecentyears,arangeofnewalloysandcompositematerialshasbeendevelopedforvariousengineeringapplications.Manyofthesenewmaterialsbecomemuchmorediculttocutwithglass.Thistechnology,calledultrasonicallyassistedturning(UAT),demonstratesarangeofbenetsinmachininghardmetalalloys:adecreaseincuttingforcesofuptoseveraltimes14,improvementinsurfacenishbyupto50%comparedtoCT5andnoisereduction6.Asformachiningofbrittlematerials,ceramicsandglasspresentlyrequireprolongedandexpensivepost-processingtoobtainthesurfacequalityrequiredforopticalcomponents；UATallowsobtainingmirrorsurfacenishinmachiningthesematerialsaswellasconsiderablereductionintoolwearandaverageUltrasonicallyassistedturningandexperimentalV.I.Babitsky,A.V.Mitrofanov,WolfsonSchoolofMechanicalandManufacturingEngineerinAbstractUltrasonicallyassistedturningofmodernaviationmaterialsamplitudeaC2515lm)superimposedonthecuttingtoolmovement.nonlinearresonantmodeofvibrationthroughoutthecuttingworkpiecesmachinedconventionallyandwiththesuperimposedprocessandnanoindentationanalysesofthemicrostructureofmodelprovidesnumericalcomparisonbetweenconventionalandcuttingforcesandcontactconditionsattheworkpiece/toolinterface.C2112004ElsevierB.V.Allrightsreserved.Ultrasonics42(2004)227502.E-mailaddress:v.silberschmidtlboro.ac.uk(V.V.Silberschmidt).0041-624X/$-seefrontmatterC2112004ElsevierB.V.Allrightsreserved.doi:10.1016/j.ultras.2004.02.001aviationmaterials:simulationsstudyV.V.Silberschmidt*LoughboroughUniversity,LeicestershireLE113TU,UKconductedwithultrasonicvibration(frequencyfC2520kHz,AnautoresonantcontrolsystemisusedtomaintainthestableExperimentalcomparisonofroughnessandroundnessforvibration,resultsofhigh-speedlmingoftheturningmachinedmaterialarepresented.Thesuggestednite-elementturningofInconel718intermsofstress/strainstate,8186/locate/ultrasmechanicsofthisprocess.Thereareonlyafewsourcesoftheworkpiece(paralleltotheX-axis,Fig.1b),orinthefeeddirection,i.e.alongtheaxisoftheworkpiece(Z-axis,Fig.1b).Aself-sustainedresonantmodeofvibrationofthiscuttingsystemisimplementedviatheautoresonantcontrolsystem,whichisdescribedindetailin9,10.ArangeofturningtestshasbeenconductedtocomparetheusageofUATandCTformachiningavi-ationmaterials.Thedetaileddescriptionofthesetestscanbefoundin5.AmongthematerialsusedforthetestsisInconel718ahigh-gradeheat-resistantNi-basedsuperalloywidelyusedintheaerospaceindustry.Thismaterialisveryabrasiveandcausesthetoolbluntingandhighcuttingtemperatureswhenmachinedconven-tionally.Thesurfacequalityobtainedbyturningisoneofthecrucialfactorsinmetalcuttingandisextremelysensitivetoanychangesinthemachiningprocess.ThesurfacenishofspecimensiscomparedintermsofaverageroughnessmeasurementRaandmeasurementofroundness(thepeak-to-valleymeasure),usingtheTay-s42(2004)8186Fig.1.Experimentalsetupforultrasonicallyassistedturning(a),andaschemeofrelativemotionoftheworkpieceandcuttingtool82V.I.Babitskyetal./Ultrasonicattemptingtodescribetheprocessesintheworkpiece/cuttingtoolinteractionzoneandtheirinuenceonthestructureofthemachinedmaterial3,6,11.Theseworksstudymostlythedynamicsoftheultrasonicmachineunitandnottheresponseofthetreatedmaterialtothistechnology,whileaclearunderstandingofmechanicalprocessesinthematerialduringUATwouldcertainlyallowafurtherdevelopmentoftheUATtechnology.ThemainaimofthispaperistostudyexperimentallyandnumericallythematerialmechanicsoftheUATprocess.2.ExperimentalstudiesTheexperimentalsetupusedtostudyUATisshowninFig.1.Theworkpieceisclampedinthechuckoftheuniversallatheandrotateswithaconstantspeed.Highfrequencyelectricimpulses,fedtotheinputoftheultrasonictransducer,excitevibrationinpiezoceramicringsduetothepiezoelectriceect.Thevibrationamplitudeisintensiedintheconcentratorandtrans-mittedtothetoolholderatthethinendofthecon-centrator.Resultantvibrationofthecuttingtipxedinthetoolholderreaches15lm(i.e.30lmpeak-to-peak)atafrequencyofabout20kHz.ThevibrationcanbeappliedeitherinthedirectiontangentialtothesurfaceinorthogonalUATwithtangentialvibration(b).lorHobsonTalysurf4surfacemeasurementinstrument.Thefollowingcuttingparametersareusedtomachinetestedspecimens:depthofcutd0:8mm,feedrates0:05mm/rev,andcuttingspeedv17m/min.ThesameparametersareusedforbothUATandCT,withsuperimposedultrasonicvibrationinthefeeddirectionappliedforUAT.Fig.2ashowsrepresentativeaxialprolesofthemachinedsurfaceoftheInconel718.ItisobviousthatmagnitudesofRaarereducedbynearly50%forspeci-mensmachinedwithUAT.Furthermore,theregularityofthesurfaceproleisgreatlyimproved,asthesurfaceFig.2.SurfacequalityofInconel718specimensmachinedwithUATandCT:axialsurfaceproles(a),roundnessproles(b).Cuttingparameters:d0:8mm,s0:05mm/rev,v17m/min.direction.Apparently,thereasonfortheseimprovementsisthes42(2004)818683changeofthenatureofthecuttingprocess,whichistransformedintotheonewithmultiple-impacthigh-frequencyinteractionbetweenthecuttingtoolandchipduetoappliedultrasonicvibration.Thisleadstochangesinmaterialdeformationprocessesandfric-tionforces,andincreaseinthedynamicstinessofthelathe-tool-workpiecesystem6,11duetothevibrationfrequencylevelsconsiderablyexceedingitsnaturalfre-quency.Inadditiontomeasurementsofthesurfacequality,themicrostructureofthemachinedsurfacehasbeeninvestigated.Inconel718workpiecesaremachinedun-derthesamecuttingconditions(v3:6m/min,d0:1mm,s0:03mm/rev)withapplicationofultrasonicvibrationintangentialdirectionandwithoutit.Then,nanoindentationanalysesofthesurfacelayersareper-formedwiththeNanoTestPlatformmadebyMicroMaterialsLtd.Accordingtotheresultsofthesetests,thewidthofthehardenedsurfacelayer,whichresultsfromtheextensivedeformationandhightemperaturepro-cessesduringtheturningprocedures,fortheultrasoni-callymachinedspecimenishalfthesizethatoftheconventionallymachinedone(40and80lm,respec-tively).Furthermore,theaveragehardnessofthislayerforUAT(about15GPa)isahalfofthatforCTandconsiderablyclosertothehardnessoftheuntreatedmaterial(about7GPa).Thehardnessofthematerialnonlinearlyincreaseswithariseintheleveloftheresidualplasticstrains.Hence,nanoindentationtestsindicatelowerresidualstrainsinthesurfacelayerforworkpiecesmachinedwithUATandaconclusioncanbedrawnthattheUATprocedureisconsiderablymoredelicatetotheworkpiecematerial.3.NumericalanalysisofUATFiniteelement(FE)simulationsareamajortoolformodellingofmachiningprocesses.Ithasbeenusedformodellingofturningforsome30years.Theoverviewofthestateoftheartinmetalcuttingsimulationscanbefoundin13,14.However,uptotheauthorsknowledge,nomodelsforUAThavebeendevelopeduntilnow.Thetwo-dimensionalFEmodelofbothCTandUATde-becomessmootherintheaxialdirection.Aconsiderableimprovementisalsoobtainedforroundnessofma-chinedworkpieces(Fig.2b):apeak-to-valleyvalueofroundnessmeasures4.20lmforCT,whereasitattainsonly1.89lmforUAT.Hence,theroundnessisim-provedby40%whenultrasonicvibrationissuperim-poseduponthemovementofthecuttingtool.Itisworthnoticingthatsimilarresultshavebeenobtainedbyotherresearchers7,12utilisingvibrationinthetangentialV.I.Babitskyetal./UltrasonicscribedinthispaperisbasedonthecommercialFEcodeMSC.Marc15.Anorthogonalturningprocess,i.e.thecuttingprocesswherethetooledgeisnormaltobothcuttingandfeeddirections,withtangentialvibrationisconsidered.Fig.1bshowsaschemeofthemodelledrelativemotionoftheworkpieceandcuttingtool；therotationaxisofthecylindricalworkpieceisorthogonaltotheplaneofthegure.Theworkpiecemoveswithaconstantvelocity,whereasthetoolvibratesharmonicallyarounditsequilibriumpositionwithfrequencyf20kHzandamplitudea15lm,correspondingtothevaluesusedinexperimentalstudies.Otherparametersofsimulationsare:uncutchipthicknesst10:1mm(whichcorrespondstothedepthofcut),rakeangleofthetoolc10C176,cuttingspeedV9m/min.Suchparametersofvibrationandofthecuttingprocessprovideseparationofthecutterfromthechipwithineachcycleofultrasonicvibration.ThematerialconstantsforagedInconel718aretakenfrom16.Kinematicalboundaryconditionsfortheworkpieceareappliedtoitsleft,rightandbottomsides(Fig.1b),whereasitstopsurfaceisfree:VxjAHV；VxjFGV；VxjHGV；VyjHG0:Thermalboundaryconditionsincludeconvectiveheattransferfromtheworkpiece,chipandtoolfreesurfacestotheenvironment:C0koT=onhTC0T1,wherekistheconductivity,hisaconvectiveheattransfercoe-cient,T1istheambienttemperature.ThethermaluxpassingfromthechiptothecutteralongthecontactlengthLc(Fig.1b)isdescribedasfollows:qHTchipC0Ttool,whereHisacontactheattransfercoecient,TchipandTtoolarechipandtoolsurfacetemperatures,respectively.Themodeltakesintoconsiderationthefollowingfactors,importantformetalturningsimulationsandaectingstressandstraingeneration:(1)contactinter-actionandfrictionatthetool-chipinterface；(2)non-linearmaterialbehaviour,includingstrain-rateeects,namelythedependenceofthematerialsyieldstressonstrainrates；(3)thermomechanicalcoupling,er-connectionbetweenmechanicalandthermalpartsoftheproblem.AsfollowsfromFEsimulations,theUATprocessduringonecycleofvibrationcouldbedividedintofourmainstages.Duringtherststage(Fig.3a),thecutterapproachesthechip；inthesecondstage,thecuttingtoolcontactsthechipandstartspenetratingintothework-piececausingthechipseparation.Theattainmentofthemaximumpenetrationdepthischaracterizedbythehighestlevelofgeneratedstressesintheprocesszoneandmarkstheendofthesecondstage(Fig.3b).Thefollowingstageisunloading:thevelocitydirectionofthetoolchangesanditmovesbackwards,butremainsincontactwiththechipevenafterthemomentwhenthespeedofthetoolexceedsthecuttingspeed(duetotheelasticspring-backofthechip).Duringthisphase,framespersecondwiththeareaoftheimagecomprisingabout4mm2.Fig.4ademonstratesaframeoftheUATlmingshowinganinteractionbetweenthecuttingtipandworkpiece.ThedierencesbetweenCTandUATinchipseparationmanifestinsuchspecicfeaturesofthetheelasticstrainsintheprocesszonedecrease.Thelaststage,startingwiththefullseparationofthecuttingedgefromthechip,isthewithdrawalofthecutterfromthechip(Fig.3c).Theintermittentcharacterofthechip-cuttingtoolcontactdeterminesthemaindierencesinthestressdistributionforCTandUAT.ThestressstateduringCTisnearlyquasistatic,asshowninnumericalsimulations(Fig.3d),withthehighestequivalentstressesconcen-tratedinprimaryandsecondaryshearzones,i.e.zonesaroundlineBE(Fig.1b)andnexttotherakefaceEK.Conversely,thestressstateinUATisinherentlytran-sient:stressesreachmaximumlevels,similartothoseofFig.3.DistributionofequivalentstressesduringUATatdierentmomentsofasinglecycleofvibration:cutterapproachingthechip(a),cutterinfullcontactwiththechip(b),andcuttermovingawayfromthechip(c)andCT(d).84V.I.Babitskyetal./UltrasonicCT,duringthepenetrationpartofthecycleofultrasonicvibration(Fig.3b),whereasthestressmagnitudeissig-nicantlylowerduringtheotherstagesofthecycle(Fig.3aandc),whenthecuttermovesawayfromthechipornotincontactwithit.Hence,averagestressesgeneratedinthematerialand,consequently,theintegrallevelofinteractionforcesbetweenthecuttingtoolandwork-pieceareconsiderablysmallerforUAT.Thisexplainsareductioninaveragecuttingforces(byseveraltimes)reportedinmanyexperimentalstudies1,3,4.4.StudyofchipformationAchipformationprocessisoneofthemostimpor-tantcharacteristicsinmetalcutting