Management Systems in Production Engineering

2012,No3(7),pp3437

IMPACTTESTSOFMICROMILLINGTOOLMOUNTEDINMICROMILLINGMACHINESPINDLE
MarcinMATUSZAK WestPomeranianUniversityofTechnology

Abstract: Methodofperformingimpacttestoftoolmountedinmicromillingmachinespindleispresented.Duetoverysmalltool dimensionsperformingimpacttestinclassicalwayisimpossible.Accelerometercannotbeusedforimpulseresponse measurement.Formeasurementoftooldisplacementlaservibrometerisused.Frequencyresponsefunc onwasmea suredintwodirec onsinsevenpointsofmicromillingtool.Addi onallyfrequencyresponsefunc oninthreepointsof machinespindleismeasured.Resonantfrequenciesandtheiramplitudeforpointsontoolandonmachinespindleare compared.Resultsofperformedimpacttestsareshown.Conclusionsarisingfromperformedimpacttestsarepresen ted.

Keywords:micromilling,impacttest,frequencyresponsefunc on,micromillingtool

INTRODUCTION Frequencyresponsefunc onoftoolmountedinmilling machine spindle measured on the tool p gives the possi bility of process dynamic analysis. It can be used eg. for processstabilitylobesevalua on[1]. Common method used in macro scale of es ma ng frequency response func on of tool mounted in machine spindleisimpacttestperformedontool p.Inmostcases tool displacement is measured with accelerometer a ached to tool p. Due to very small micromilling tool dimensionsperformingimpacttestontool pcanbedi cult. Excita on of tool p with micro impact hammer is impossible. Usage of accelerometer for tool displacement measurementisalsounreasonable. Therearesomeknownmethodsofes ma ngfrequen cyresponsefunc ononmillingandmicromillingtool p[2, 3,4,5,6,7,8].Usuallythesemethodsusescomplexmathe ma calmodelsandarebasedonsynthesisofexperimental and analy cal data [2, 3 ,4, 5, 6, 7, 8]. The main disad vantageofthesemethodsisdicultexperimentalverica onofcalculatedtransferfunc on. Inthepaperisproposedtooldisplacementmeasure ment with laser vibrometer. Laser vibrometry do not re quires a achment of accelerometer which ads extra mass to the measured object. Laser vibrometry also provides velocity signal that requires one integra on in case to get displacement signal. Accelera on signal requires double integra ontoobtaindisplacementsignal.

Proposed method can be used for verica on of fre quency response func on obtained from analy cal meth ods. From described method can be obtained frequency response func on from point on tool and tool p. Due to largeimpulseforcehammerdimensionsandverysmalltool pdimensionsfrequencyresponsefunc onfromtool pto tool pcannotbeobtaineddirectly. EXPERIMENTSETUP Experiment was performed on prototype threeaxial micromillingmachine SNTMCMZUT1(Fig.1a) whichwas buildinMechatronicsCentreofWestPomeranianUniversi tyofTechnology. LSM SCADAS III data acquisi on device was used for signalregistra on.ImpulseforcehammerPCB086E80was used for tool excita on. Tool displacement was measured with Polytec PSV400 Scanning Vibrometer. LMS Test Lab so ware wasused for signal processing and frequency re sponsefunc onevalua on. Impact tests were performed in two direc ons. (X and Y). For measurement of tool displacement in Y direc on there was used addi onal mirror from Renishaw interfer ometer.Toolresponsewasmeasuredinpoints17(Fig.1b), spindle response was measured in points AC. Excita on from impulse force hammer was applied in point 5 of mi cromillingtool.ForperformedtestsmicromillingtoolKyoc era2FESM00501004wasused.Tool pdiameterwas0.5 mm.

ManagementSystemsinProduc onEngineering3(7)/201235
M.MATUSZAKImpacttestsofmicromillingtoolmountedinmicromillingmachinespindle

Fig.1.a)schema cviewofmicromillingmachine,b)17pointsoftooldisplacementmeasurement,AC,pointsofspindledisplacement measuremen

Fig.2.Impacttestoftool:a)impacthammerandtool,b)micromillingmachineandlaservibrometer

EXPERIMENTRESULTS Comparison of frequency response func on for points1,5and7inX(horizontal)direc onisshownong 3. There can be seen three main resonant frequencies at 2100 Hz, 2700 Hz and 11 kHz. Resonant frequencies for measurement points 1, 5 and 7 on tool in X direc on are the same. Only their amplitude diers. On the tool p (measurement point 7) highest amplitude has third reso nant frequency (11 kHz) and rst resonant frequency (2100 Hz). First two resonant frequencies (2100 Hz and 2700 Hz) have the similar amplitude for points 1 and 5. Thirdresonantfrequency(11kHz)hashigheramplitudein point5thaninpoint1.

Frequencyresponsefunc onsforpoints1,5,and7inY (ver cal)direc onisshownongure4.Inthisdirec oncan beobservedtwomainresonantfrequenciesat2650Hzand 11kHz.Highestamplitudesforbothrstandsecondreso nantfrequenciesareatpoint7. Dierence between frequency response func ons ob tainedforXandYdirec onsisonlyinnumberofresonant frequencies. In X direc on there are three resonant fre quencies and in Y direc on are two resonant frequencies. Both in X and Y direc ons amplitude in point 7 for last (11kHz)resonantfrequencyismuchhigherthanforpoints 1 and 5. The reason is probably that highest resonant fre quencyisgeneratesbythintool p.

36ManagementSystemsinProduc onEngineering3(7)/2012 M.MATUSZAKImpacttestsofmicromillingtoolmountedinmicromillingmachinespindle

Fig.3.Transferfunc oninXdirec onforpoints1,5and7

Fig.4.Transferfunc oninYdirec onforpoints1,5and7

Fig5.Transferfunc onforpointsCand1inXdirec on

ManagementSystemsinProduc onEngineering3(7)/201237
M.MATUSZAKImpacttestsofmicromillingtoolmountedinmicromillingmachinespindle

Comparisonoffrequencyresponsefunc onforpoint1 oftool,whichisclosesttothemachinespindleandpointC ofmachinespindleisshownongure5.Resonantfrequen cies for these points are dierent. In point 1 can be seen addi onal third resonant frequency at 2700 Hz. This fre quencydonotoccursinpointC.Alsolastresonantfrequen cyisdierentforpoint1andpointC.Thesedierencesin resonant frequencies are probably result of spindletool connec onwhichisnotperfectlys andhassomedamp ing. SUMMARY Performed impact tests show that laser vibrometer can be used for micromilling tool displacement measure mentwhichgivespossibilityoffrequencyresponsefunc on calcula on.Thereisapossibilityofdisplacementmeasure mentofmicromillingtool pof0.5mmdiameter.Transfer func ons from point (5) located on tool to points 17 was evaluated.Addi onallytransferfunc onsfromtooltoma chinespindlewerees mated. ResonantfrequenciesbothforXandYdirec onfor dierent (1, 5, 7) points of tool are the same. Only ampli tudeforthesefrequenciesisdierent.Toolissolidpartof material with very low damping coecient which should result in same resonant frequencies for all measurement points. Measurementpoint C whichin on micromilling ma chine spindle has dierent resonant frequencies than fre quenciesonmicromillingtool.Moreovertherecanbeseen (Fig. 5) addi onal resonant frequency at 2700 Hz that oc curs only on micromilling tool and cannot be observed on machinespindle.Probablysourceofthisfrequencyisspring masssystemthatiscreatedbytoolchucks nessandtool mass. This phenomenon cannot be neglected in tool fre

quency response func on es ma on especially when is usedsynthesisofanaly calandexperimentalresults. REFERENCES [1]Chae J., Park S.S., Freiheit T.: Inves ga on of micro cu ng opera ons. Interna onal Journal of Machine Tools&Manufacture.No.46,2006,pp.313332. [2]Malekian M., Park S. S.. Mar n B.G. Jun: Modeling of dynamicmicromillingcu ngforces.Interna onalJour nalofMachineTools&Manufacture.No.49,2009,pp. 586598. [3]MatuszakM.,PowakaB.:Receptancecouplingmethod formicromillingmachinedynamicssimula on.Proceed ings of 13th interna onal symposium of students and youngmechanicalengineersAdvancesinchemicaland mechanicalengineering.Gdask2010,pp.262267. [4]MatuszakM.,PowakaB.:Moliwocibadaniadynamiki procesu mikrofrezowania. Ekonomika i Organizacja Przedsibiorstwa.Nr4(723),s.7071+CD. [5]MatuszakM.,PowakaB.:Wybraneproblemybadawcze waciwoci dynamicznych obrabiarki do mikroskra wania. Modelowanie inynierskie. T. 8, nr 39, 2010, s. 151158. [6]ChiHungCheng,SchmitzT.L.,NagarajArakere,Duncan G. S.: An approach for micro end mill frequency re sponsepredic ons.Proceedingsof2005ASMEInterna onalMechanicalEngineeringCongressandExposi on. ASME,Orlando2005,pp.11391145 [7]Mar n B. G. Jun, Xinyu Liu, Richard E. DeVor, Shiv G. Kapoor:Inves ga onoftheDynamicsofMicroendMill ingPartI:ModelDevelopment.JournalofManufactur ing Science and Engineering. November 2006, vol. 128, pp.893900. [8]Filiz S, O. Burak Ozdoganlar: Microendmill Dynamics Including the Actual Fluted Geometry and Setup Er

mgrin.MarcinMatuszak WestPomeranianUniversityofTechnology FacultyofmechanicalEngineeringandMechatronics ul.Piastw19,70310Szczecin,POLAND email:marcin.matuszak@zut.edu.pl