Investigation of laser consolidation process for metal powder by two-color pyrometer and high-speed video camera

(1)

Investigation of laser consolidation process for metal powder by two‑color pyrometer and high‑speed video camera

著者 Furumoto Tatsuaki, Ueda Takashi, Alkahari Mohd Rizal, Hosokawa Akira

journal or

publication title

CIRP Annals ‑ Manufacturing Technology

volume 62

number 1

page range 223‑226

year 2013‑01‑01

URL http://hdl.handle.net/2297/34658

doi: 10.1016/j.cirp.2013.03.032

(2)

Investigation of laser consolidation process for metal powder by two-color pyrometer and high-speed video camera

Tatsuaki Furumoto

^a

, Takashi Ueda (1)

^a,

*, Mohd Rizal Alkahari

^b,c

, Akira Hosokawa

^a

aFacultyofMechanicalEngineering,InstituteofScienceandEngineering,KanazawaUniversity,Kakuma-Machi,Kanazawa,Ishikawa920-1192,Japan

bDivisionofInnovativeTechnologyandScience,GraduateSchoolofNaturalScienceandTechnology,KanazawaUniversity,Japan

cFacultyofMechanicalEngineering,UniversitiTeknikalMalaysiaMelaka(UTeM),Malaysia

1. Introduction

Layeredmanufacturingtechnologywaspresentedinitiallyby Kodamaasanewrapidprototypingtechnologytofabricatethree- dimensional plastic models [1]. This technique became the commontechnologyforadditivemanufacturing(AM)toproduce prototypes,toolsand functional endproducts witha varietyof components,suchaspolymer,ceramicandmetalpowder[2].The developmentofthree-dimensionalCADsystemalsocontributedto the commercial uses of thelayered manufacturing techniques.

Indeed, the use of metal powder in layered manufacturing is especially remarkable because the structures obtained are sufﬁcientlystrongforpracticalapplication[3].Laserconsolidation of metal powder is classiﬁed into several types of mechanism accordingtotheenergydensityonareairradiatedbylaserbeam [4].Selectivelasersintering(SLS)andselectivelasermelting(SLM) enable the fabrication of three-dimensional models from the powdered materials by selectively heating and fusing powder particlesusingalaserbeamirradiation.

ProblemsencounteredinSLSandSLMaretheaccuracyofthe structuresobtainedbythelayeredmanufacturingprocesses.The dimensionalaccuracyandthesurfacequalityofthestructuresare inferior to conventional technologies such as a milling and grinding,and theseareimportant limitations tolayeredmanu- facturingtechnology. The improvementof these problems was proposedsymptomaticallysuchasanemploymentofsecondary laserirradiationwithoutdepositinganewmetalpowderlayer[5]

and anapplication of theatmospheric plasma spraying system (APS)on theconsolidatedstructure[6].Additionally, Abeet al.

developedamultitaskingmachineinwhichalaserconsolidationof metal powder and an end milling of the edge of consolidated structurewereperformedalternately[7].

The principal factor which causes the deformation of the consolidatedstructureinSLSandSLMistheresidualstress,which isinducedbythethermalgradientbroughtonbyrapidheatingand cooling during layer consolidation. Shiomiet al. measuredthe distributionofresidualstresswithintheconsolidatedstructurein theSLMprocess[8].Theyreportedthatthetensileresidualstress remaininginthe surfacelayer oftheconsolidatedstructurewas improvedbyre-scanningofthelaserbeamwithoutdepositionof newpowder.Theinﬂuenceofthethicknessofsubstrateandthe heightofconsolidatedstructureontheresidualstressinthesurface layerwasalsoinvestigatedinourpreviousstudy[9].Itwasshown thatthedeformationoftheconsolidatedstructurewasrelatedtothe valueofresidualstressandthatresidualstressincreasedwiththe increaseinthenumberoflayers.Itisquitedifﬁculttoeliminatethe residualstressinducedinsidetheconsolidatedstructurealthough pre-heatingofthepowderbediseffectiveinreducingit.

In order to improve the dimensional accuracy causally in layeredmanufacturingtechnology,itisessentialtomonitorand visualizethelaserirradiationareaduringpowderconsolidation.

Thisresearchfocusesonthemeasurementofsurfacetemperature offerrousbasedmetalpowderbyirradiationwithYb:ﬁberlaser beam.Surfacetemperaturewasmeasuredwithtwo-colorpyrometer employing an optical ﬁber with a different acceptable wavelength of InAs and InSb detectors, which were developed bytheauthors[10].Additionally,thepowdersurfaceduringlaser irradiationwasmonitoredwithhighspeedvideocamerainorder toclarifytheconsolidationmechanism.

2. Experimentalmethod 2.1. Two-colorpyrometer

Aschematicillustrationofthetwo-colorpyrometerisshownin Fig. 1 [10]. The pyrometer is composed of an optical ﬁber, a condenser,anopticalﬁlterandtwotypesof infrareddetectors;

*Correspondingauthor.

ARTICLE INFO

Keywords:

Additivemanufacturing Temperature

Monitoring

ABSTRACT

This paper deals withthe measurementof surface temperatureon metal powder during the laser consolidationprocesswithtwo-colorpyrometer.Additionally,theaspectofselectivelasersintering(SLS) andselectivelasermelting(SLM)ofmetalpowderisvisualizedwithhighspeedvideocamera.Asaresult,the surface temperature during the laser irradiation was ranged 1520–18108C and the consolidation phenomenawasclassiﬁedaccordingtothemeltingpointofmetalpowder.Themetalpowderattheheating processcoheredintermittentlytothemeltpoolalthoughthelaserbeamwascontinuouslyirradiatedtothe powdersurface.

ß2013CIRP.

0007-8506/$–seefrontmatterß2013CIRP.

http://dx.doi.org/10.1016/j.cirp.2013.03.032

(3)

namely,anInAsdetectorandanInSbdetector.Thesedetectorsare mountedinasandwichconfiguration,witheachdetectorhavinga different range of acceptable wavelength. The infrared energy radiatedfromthetargetisledtothetwo-colordetectorthrougha chalcogenide glass fiber and converted to an electric signal, amplified and stored in a digital memory. The frequency characteristicsofthepyrometerareaflatresponseforsinewave from10Hzto100kHz,sothatthepyrometerhassufficientspeed tomeasurethelaserirradiationarea[11].Bytakingtheratioof outputsignals, theinfluence of theemissivityoccurring in the surfacecharacteristics atthe laser irradiationarea is negligible [12].To protectthetwo-colordetectorfromlaser irradiation,a germaniumopticalfilterwithathicknessof1mmwasapplied.

Thisﬁltercutoffnearlyallwavelengthsbelow1600nm.

Thecalibrationwascarriedoutusingtwodifferentmethods.In therange upto10008C, theknown uniform temperature on a radiationsurface wasmeasured with a thermocouple. Alumina wasusedasatargetmaterial.Intherangeexceeding10008C,the minimumoutput,atwhich thesurface ofthespecimenmelted when the laser beam was irradiated, was measured. As target materials,SiC,withameltingpointof22008C,andasteelsheet withameltingpointof15108Cwereused.

Fig. 2showsthecalibrationcurveobtainedbytakingtheratio oftheoutputvoltages ofInAsand InSbdetectors.Bytakingthe ratioofoutputsignals,theinﬂuenceoftheemissivityoccurringin the surface characteristics at the laser irradiation area is eliminated.Thesolid line indicates thecurvederived from the sensitivityofthepartswhichcomposethetwo-colorpyrometer.

Theexperimentalpointscoincidedwellwiththecalculatedcurve.

Theinterpretationofratiototemperaturewascalculatedwitha solidline.Therangeoftemperaturemeasurementwiththetwo- colorpyrometerwas400–22008C.

2.2. Experimentalsetup

Theexperimental arrangementisschematicallyillustratedin Fig.3,andtheexperimentalconditionsaresummarizedinTable1.

Thisequipment wasa self-designed consolidated system com- posedof laser equipmentand consolidation facility of a metal

powder. The laser beam used was a continuous Yb:ﬁber laser (SUNXLtd.:LP-F10)withamaximumlaserpower of40W.The intensityofrelativedensityofthelaserbeamradiatedfromthe galvanometermirrorformedaGaussianshape.Thebeamdiameter at the point where the intensity falls to 1/e² times from the maximumvaluewas45

m

m.Thesubstrateusedtoconsolidatethe metalpowderwascoldrolledsteel,anditssurfacewasprocessed by sandblasting withan average grain size of #46 in order to improvethewettingpropertyofthemeltedpowder.Thevesselfor the consolidation was ﬁlled with nitrogen to prevent the oxidizationofthemetalpowderduringlaserirradiation.

Themetalpowderusedwasamixtureof70%steelpowder,20%

copperphosphorous alloypowderand 10%nickelpowder.Each powderwaspreparedbythegasatomizationmethod.Themetal powderwasdepositedonthesubstrateusingalevelingblade,and its thickness couldbe controlled by the precision stage at the requiredthickness.

The surface temperature during line consolidation on the deposited metal powder was measured. The laser beam was irradiatedtothepowderedsurfacethroughagalvanometermirror atthefocusspot,and scannedusing programmedNCdata.The chalcogenideglassfiberofthetwo-colorpyrometerwassetata distanceof4.2mmfromthelaserirradiationareaandatanangle of458tothepowderedsurface.Thecenterofthelaserirradiation areaalwayspassedthecenteraxisoftheglassfiber.Hence,the fiberreceivedtheinfraredraysradiatedfromthelaserirradiation areawhenthelaserbeampassedthroughthetargetareaofthe fiber.Theacceptance angleof thefiberwas248. Therefore,the passing length through the target area was 4.1mm [13]. The influenceoftheirradiationparameteronsurfacetemperaturewas investigated.

Fig.1.Fundamentalstructureoftwo-colorpyrometer.

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8

0 400 800 1200 1600 2000 2400

Fiber type: NSG

Core material: Chalcogenide glass Core diameter: 380µm

Filter: Ge

Temperature o C

Ratio of output voltage (InAs/InSb) Al2O

3

Steel sheet SiC

Fig.2.Calibrationcurveobtained.

Table1

Experimentalconditions.

Lasersource

Wavelength [nm] 1070

Power [W] 1–40

Beamdiameteratfocalspot [mm] 45

Scanspeed [mm/s] 1–88

Substrate

Material Coldrolledsteel

Thickness [mm] 3.2

SurfaceroughnessRa [mm] 3.5

Metalpowder

Material Fe,Cu,Ni

Shape Irregular

Particlemeandiameter [mm] 25

Bulkdensityonsubstrate [kg/m³] 4200 Absorptionratiooflaserbeam [%] 25

Thermalconductivity [W/mK] 0.14

Layerthickness [mm] 0.05

Highspeedvideocamera

Recordingspeed [fps] 10,000

Resolution 768648

Lightsource Metalhalidelamp

Fig.3.Experimentalsetupfortemperaturemeasurement.

T.Furumotoetal./CIRPAnnals-ManufacturingTechnologyxxx(2013)xxx–xxx 2

(4)

3.1. Consolidationmechanismofmetalpowder

Fig.4showstheconsolidationaspectintheSLMprocessofmetal powderduringlaserbeamirradiation.Theschematicillustrationsof powderconsolidationineachimagearealsorepresented.Herethe laserpowerwassetat40Wwhereasthescanspeedwas45mm/s.

Theseimageswererecordedbythehigh-speedvideocameraata timeintervalof4ms.Whenthelaserbeamwasirradiatedtothe powdersurface,thepowderwasheatedandthemoltenareawas formedasshownin(a). Thecircumferentialpowderofthelaser beamspotwasalsoaffectedbytheconductedheatinducedbylaser beamirradiation,andthesepowderscoheredtothecenterofthe laserbeamduetothesurfacetensionofmoltenpowder,asindicated in(b).Thecohesionofcircumferentialmetalpowdertothemolten area occurred intermittently although the laser beam was con- tinuouslyirradiatedtothepowdersurface.Inadditiontothesurface tensioninducedbythemeltingofthemetalpowder,theshapeofthe molten powder was strongly influenced by the adhesion force inducedbythewettingproperties.Thelaserbeamwasscannedon thepowdersurfaceandthemoltenareawasformedinthelaserscan direction,sothatthe moltenspherical powder was growing up gradually,asdepictedin(c).Whenthelaserbeampassedthrough the moltenarea whichwas sufficientlylarge duetothe surface tension,the molten powder resultedin the solidification repre- sentedin(d).Theincreaseofheatcapacityinducedbythegrowthof themoltenpowderalsopromotedthesolidificationofthemolten powder.Inadditiontothis,newlymoltenpowderwasalsoformed near the solidified powder. These processes were periodically repeatedaccordingtotheprogressoflaserbeamirradiation.

3.2. Surfacetemperatureduringlaserbeamirradiation

Fig. 5showsthevariationoftemperatureduringlaserbeam irradiationwiththeenergydensity.Herethelayerthicknessonthe substratewas0.05mm.Theenergydensitywascalculatedusing laser power, scan speed and beam spot diameter. The typical outputsignalofInSbdetectorandtemperaturehistoryobtained fromthetwo-colorpyrometerarealsorepresentedinFig.5.When the laser beam reached the target area of the pyrometer, the infraredenergy wasimmediately detected bythe detectorand convertedintoanelectricsignal.Theoutputsignalincreasedasthe

laserbeamapproachedthecenterofthetargetarea,andreached itsmaximumvaluewhenthelaserbeamreachedthecenterofthe targetarea.Theoutputsignalthendecreasedgraduallyafterthe laser beam passed the center. The variation in temperature obtainedtakingtheratioofoutputsignalsshowedsimilartrends withthatoftheoutputsignal.Thelaserbeamwasnotdetectedin thedetectorduetotheworkingofagermaniumﬁlter.

Thetemperatureonthemetalpowderwasgreatlyinﬂuencedby energydensityasshowninFig. 5.Thesurfacetemperatureatan energydensityof2.5J/mm²was15208C,and18108Cat10.1J/mm². Fig. 6 compares the image of the consolidated structure to investigatetheinﬂuenceofenergydensityonmetalconsolidation.

Schematicillustrationsofpowderconsolidationarealsoshown.Fig.

6(a)showstheenergydensityat2.5J/mm²andFig. 6(b)showsthe energydensityat10.1J/mm².Whentheenergydensitywas2.5J/

mm²,thesurfacetemperaturewas15208Candwaslowerthanthe melting pointof steelpowder. Therefore, the steelpowder was partiallymeltedandtheshapeofthesolidiﬁedstructurewasrough.

Ontheotherhand,whentheenergydensitywasincreasedto10.1J/

mm²,thesurfacetemperaturewas18108C.Thesteelpowderwas fully melted and the shape of the consolidated structure was sphericalduetothesurfacetension.Themeltingpointofthesteel powderwhichwasthemaincomponentofthepowdermixturewas

Fig.5.Variationoftemperaturewithenergydensity.

Fig.4.Consolidationaspectofmetalpowder. Fig.6.Comparisonoftheinﬂuenceofenergydensityonconsolidation.

(5)

15408C,and the meltingpoint ofcopperand nickelwhich was includedinthepowdermixturewaslowerthanthatofsteelpowder.

Theconsolidationphenomena ofpowder mixture were strongly relatedtosurfacetemperature.Especially,themeltingpointofsteel powderwasimportantfactortoclassifytheconsolidationphenom- ena of powder mixture. Additionally, it was also shown that consolidationbehaviorcouldbecontrolledbymeasuringthesurface temperatureonthepowdermixturenon-destructively.

3.3. Visualizationofthesourceofresidualstressonmetalpowder Fig. 7 shows an image and schematic representation of consolidationwhenthelaserpowerwas40Wandthescanspeed ofthelaserbeamwas1mm/s.Duringcoolingprocessofthemetal powder,theconsolidatedstructureshrunktowardthecenterofthe structure. The shrinkageof consolidated structure started after 250msoflaserirradiation.Itwasestimatedthattheshrinkageof theconsolidatedstructureoccurredwhenthesurfacetemperature wassufﬁcientlylow. Theshrinkage aftersolidiﬁcation ofmetal powdercaused residualstresson theconsolidated surface.The sourceofresidualstressonthemetalpowderwasvisualizedbythe observationofpowderconsolidationduringlaserbeamirradiation.

Fig. 8showsthevariationofthewidthofmoltenpowderwith energydensity.Additionally,thedifferenceofwidthinmoltenarea andconsolidatedstructureineachenergydensityisalsoshownin it.Herethewidthofmoltenareaandconsolidatedstructurewere bothdefinedasthewidthintowhichallconsolidatedstructures wereincludedinside thedouble parallellines witha reference lengthof2.6mm.Thewidthofmoltenareawassignificantlylarger thanthatofthespotdiameterofthelaserbeam,anditsvaluewas increasedwithenergydensityduetotheconductedheatinduced bylaser beamirradiation. Onthe otherhand,thedifference of widthinmoltenareaandconsolidatedstructureintheprocessof solidification was almost constant under each condition. The reduced distance was the most important parameter in the generationoftheresidualstress.Shiomiet al.reportedthatthe residualstressontheconsolidatedstructurewasnotaffectedby thelaserconditions[8].Itwasalsoshowninourpreviouspaper thattheparameterthatinfluencedresidualstresswasthenumber oflayersintheconsolidatedstructure[9].Sincethegenerationof

residual stress could not be avoided in the process of laser consolidation, it wasnecessary toperformpostprocess on the consolidatedstructuretoimproveresidualstress.

4. Conclusions

Inthispaper,thesurfacetemperatureofthepowdermixturein metallic additivemanufacturing duringlaserbeam irradiation wasmeasuredbytwo-colorpyrometeremployingopticalﬁber, andtheaspectof thepowderconsolidationwasmonitoredby high-speed video camera in orderto clarify the consolidation mechanism. Additionally, the inﬂuence of energy density on consolidation characteristics was investigated experimentally.

Theresultsobtainedwereasfollows:

1.Thetwo-colorpyrometerwithopticalﬁbermadeitpossibleto measurethesurfacetemperatureattheirradiatedareaoflaser beamwhilethemetalpowderwasconsolidated.Themeasured temperatureduringthelaserirradiationrangedfrom15208Cto 18108C.

2.Theconsolidationphenomenaofpowdermixturewerestrongly relatedtosurfacetemperature,andthepossibilityofcontrolling consolidationbehaviorbymeasuringthesurfacetemperature wasdemonstrated.

3.Inthesolidiﬁcationprocessofmoltenpowder,thecohesionof circumferential metal powder to the molten area occurred intermittently although the laser beam was continuously irradiatedtothepowdersurface.

4.The consolidatedstructure shrunk during cooling process of metal powder,and its shrinkage occurred when the surface temperaturewassufﬁcientlylow.Thesourceofresidualstress on the metal powder was visualized by the observation of powderconsolidationduringlaserbeamirradiation.

Acknowledgments

Theauthors wouldlike toexpress their sinceregratitudeto Panasonic Corporation for providing layered manufacturing equipment,theirsupportandpreciousadvice.

References

[1]KodamaH (1981)AutomaticMethodforFabricatingaThree-dimensional PlasticModelwithPhoto-hardeningPolymer.ReviewofScientiﬁcInstruments 52(10):1770–1773.

[2]KruthJP(1991)MaterialIncreaseManufacturingbyRapidPrototypingTech- niques.AnnalsoftheCIRP40(2):603–614.

[3]LevyGN,SchindelR,KruthJP(2003)RapidManufacturingandRapidTooling withLayerManufacturing(LM)Technologies,StateoftheArtandFuture Perspectives.AnnalsoftheCIRP52(2):589–609.

[4]KruthJP,LevyG,KlockeF,ChildsTHC(2007)ConsolidationPhenomenainLaser andPowder-BedBasedLayeredManufacturing.AnnalsoftheCIRP56(2):730–

759.

[5]YasaE,KruthJP,DeckersJ(2011)ManufacturingbyCombiningSelectiveLaser Meltingand SelectiveLaserErosion/Laser Re-melting.Annals of the CIRP 60(1):263–266.

[6]ZhangB,ZhuL,LiaoH,CoddetC(2012)ImprovementofSurfacePropertiesof SLMPartsbyAtmosphericPlasmaSprayingCoating.AppliedSurfaceScience 263:777–782.

[7]AbeS,HigashiY,FuwaI,YoshidaN,YoneyamaT(2006)Milling-Combined LaserMetalSinteringSystemandProductionofInjectionMoldswithSophis- ticated Functions.ICPE2006,11thInternational Conferenceof Precisionand Engineering,16–18August,Tokyo,Japan,288–293.

[8]ShiomiM,OsakadaK,NakamuraK,YamashitaT,AbeF(2004)ResidualStress withinMetallicModelMadebySelectiveLaserMeltingProcess.Annalsofthe CIRP53(1):195–198.

[9]Furumoto T,UedaT, AzizMSA, HosokawaA, TanakaR (2010)Study on ReductionofResidualStressInducedduringRapidToolingProcess,Inﬂuence ofHeatingConditions onResidualStress. KeyEngineeringMaterials447–

448:785–789.

[10]UedaT,SentokuE,YamadaK,HosokawaA(2005)TemperatureMeasurement inLaserFormingofSheetMetal.AnnalsoftheCIRP46(1):117–122.

[11]UedaT,HosokawaA,OdaK,YamadaK(2001)TemperatureofFlankFaceof CuttingToolinHighSpeedMilling.AnnalsoftheCIRP50(1):37–40.

[12]DaviesMA,UedaT,M’SaoubiR,MullanyB,CookeAL(2007)OntheMeasure- ment ofTemperature inMaterial RemovalProcesses.Annals of the CIRP 56(2):581–604.

[13]Ueda T,YamadaK,Nakayama K (1997)Temperatureof Work Materials IrradiatedwithCO2Laser.AnnalsoftheCIRP46(1):117–122.

Fig.7.Consolidationaspectofmetalpowder.

Fig.8.Variationofmoltenwidthandshrunkdistancewithenergydensity.

T.Furumotoetal./CIRPAnnals-ManufacturingTechnologyxxx(2013)xxx–xxx 4