1-s2.0-S2214860414000256-main

Available online at www.sciencedirect.com

ScienceDirect

Additive Manufacturing 5 (2015) 3139

Approximation of absolute surface temperatbe no

yar A. W

W.M. Keck Center for 3D Innovation, The University of Texas at El Paso, 500 W. University Ave., Engineering Building Room 108, El Paso, TX 79968, USAAccepted 11 December 2014

Available online 7 January 2015

Abstract

Additive material usathe process infrared (IRabsolute surmelting (EBand raking) to determineparameters f(determinedfactors assocenvironmena 3.77% diftemperaturedetermine pfor considerthe fabricati 2014 Else

Keywords: E

1. Introdu

Electron(AM) procmetal powdthe past de

CorresponE-mail ad

http://dx.doi.o2214-8604/manufacturing (AM) has several possible advantages over traditional manufacturing including increased design freedom, reducedge, and shorter lead-times. A noteworthy capability of AM is the ability to monitor the process during material deposition and interruptduring fabrication if necessary. Recently, such monitoring, feedback, and control have been made possible by implementing in situ) thermography in powder bed fusion AM technologies. The purpose of the current research was to investigate the acquisition offace temperatures using in situ IR imaging of the melted or solid surfaces layer-by-layer during fabrication within an electron beamM) system. The thermal camera was synchronized with the systems signal voltages of three synchronized events (pre-heating, melting,to automatically capture images. To acquire absolute temperature values from the IR images, a calibration procedure was established

the solid materials emissivity and reflected temperature or mean radiant temperature of the build chamber, which are necessary inputor the IR camera. A blackbody radiator was fabricated via EBM and was used as a tool to determine the emissivity of Ti6Al4V

to be 0.26 in the temperature range of the current study). Furthermore, a mathematical model was developed to determine the viewiated with the systems interior (e.g. heat shielding) that were used in calculating the mean radiant temperature of the manufacturing

t (342 C). Experimental validation of the model was performed using a thermocouple embedded during fabrication that showedference in temperature. A temperature difference of 366 C (1038 C vs. 672 C) was observed when comparing uncorrected IR

data with corrected temperature data. Upon validation of the IR parameters for a melted area, experimentation was conducted to alsoowder emissivity (found to be 0.50). The thermal model presented here can be modified and implemented in other AM technologiesation of radiation energy to acquire absolute temperatures of layered surfaces, leading to improved thermal monitoring and control ofon process.vier B.V. All rights reserved.

lectron beam melting; Infrared; Monitoring; Emissivity; Surface temperature

ction

beam melting (EBM) is an additive manufacturingess for direct-metal freeform fabrication that useser as the precursor to build solid metal parts. Within

cade, the development of processing parameters for

ding author. Tel.: +1 915 747 6837.dress: [email protected] (J. Mireles).

several alloys (copper, niobium, iron, TiAl, TiNb, and nickel-based superalloys) has made them suitable for fabrication byEBM technology [1,9,13,1720]. Although EBM has been usedin the commercial fabrication of parts for the medical andaerospace industries [8,24]; opportunities still exist to make AMsystems more repeatable and reproducible for the production ofhigh-quality products that may be qualified during fabricationthrough the use of monitoring, feedback and control. Christensenet al. [5] discussed qualification of EBM technology for ortho-pedic applications and a need for systematic process monitoring

rg/10.1016/j.addma.2014.12.001 2014 Elsevier B.V. All rights reserved.d fusion additive manufacturing techthermograph

Emmanuel Rodriguez, Jorge Mireles , CesMireya A. Perez, Ryan Bure measurements of powderlogy using in situ infrared

. Terrazas, David Espalin,icker

32 E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139

to understamicrostrucand discusswithin the pbe possible(through reparametersdemonstratby developtem wheremicroscopyprocessinggraded micshowing a Rodriguez itoring usinprocessingcess in EBMdue to a laif excess pr

In addittems have bsystems aroable to idenature measshortly, thirately deterdue to promorphologing a radiaIR studies imaging syporosity wiwas designvent metallof metal ducontinuouscamera an

using a sining Kaptonthe EBM pother reseanology [2,which integthe objectivlead to defeto acquire afaces of papreviously mal modela parts sumeasureme

benefits toother powdprovides dimaging toperatures.

thods

lectr

EBM F75builder (1abriclly isingo 5l4VA) anpowdvely

(up tvely ild ps is r

herm

IR cs, In

and n tha fiepatia

atog wansmpe o

nismw tolting

that winde cam

co

n in IR cn au

Nationd uurinuent

cam

imags; howect flecondias to ergynd process variables. Murr et al. [16] showed thetural anisotropy that is inherent in EBM productsed these differences arise from temperature gradientsrocess. Murr et al. [16] further described that it may

to selectively grade the alloys mechanical propertiessulting microstructures) with control of processing. Extending the work of Murr et al. [16], Mireles [14]ed that microstructural control was possible in EBMing a closed-loop automatic feedback control sys-

microstructure control was verified through optical. In this work, successful monitoring and control of

temperature was employed to intentionally create arostructure of Ti6Al4V EBM-fabricated parts [15]difference in alpha width from 0.65 m to 1.31 m.[21] further demonstrated the benefits of in situ mon-g IR thermography to enable development of EBM

parameters for copper, a complicated material to pro- because of its thermal expansion and heat capacity

rge thermal expansion property that causes warpingocessing temperature is used.

ion to those briefly described above, IR imaging sys-een installed in a number of powder bed fusion AMund the world. However, the authors have not beentify any published work suggesting absolute temper-urements have been acquired. As will be describeds is most likely a result of the complexity with accu-mining emissivity (a difficult parameter to measure

cess temperature variations and changes in surfaceies) of the surfaces under study as well as develop-tion model of the build chamber. Examples of priorinclude Schwerdtfeger et al. [22] in which an IR-stem was developed to monitor the EBM process forthin a parts melt surface. The aforementioned systemed with an integrated shuttered mechanism to pre-ization (metal coating that occurs due to evaporationring melting) of the viewing glass which prevented

monitoring. Dinwiddie et al. [7] also installed an IRd developed two shutterless imaging systems (onegle mirror periscope and another using a revolv-

film) that helped provide continuous monitoring ofrocess. Additionally, IR imaging has been used byrchers to monitor selective laser melting AM tech-11]. It is important to note that these researchers,rated IR in AM systems, used thermal imaging withe of searching for surface abnormalitieswhich mayctive products. The purpose of the current work wasbsolute temperature measurements of the build sur-rts under fabrication using in situ IR imaging withdetermined material emissivity and a developed ther-

that considers the effects of radiant temperature onrface temperature. Obtaining absolute temperaturents during the EBM process can provide numerous

the quality of parts being produced using this ander-based AM processes. As a result, the followingetails of the method developed for EBM with IR

measure accurate layer-by-layer part surface tem-

2. Me

2.1. E

TheASTMrial to chambin its f(typicaform ulayer tTi6A(8.8 mmetal selectipowereffectithe buproces

2.2. T

TheSystempixels betweeing in and a sitionedviewinhigh trused tymechawindothe meensure

on theand th25). Ais showby thestep. Aware (1 secotured dsubseq

2.3. IR

IR imagethe eff(e.g. resuch ccamer

ant enon beam melting process

process uses a metal powder (typically Ti6Al4V, CoCr, or Grade 2 Titanium), as the precursor mate-

solid parts in a layerwise fashion inside a vacuum04 Torr). The Arcam A2 system follows four stepsation process: (1) depositing a layer of metal powdern the range 0.050.20 mm thick) across the build plat-

the machines raking mechanism, (2) preheating the0% of the metals melting temperature (760 C for) using the electron beam gun at a low beam currentd high scan speed (14,600 mm/s) to lightly sinter theer (also helps reduce the parts residual stresses), (3)

melting the preheated powder by increasing the beamo 17 mA) and reducing the scan speed (500 mm/s) toreach the materials melting point, and (4) loweringlatform a height equal to one layer thickness. Thisepeated until the part is complete.

al imaging system installation

amera utilized in this study was a FLIR SC645 (FLIRc., Wilsonville, OR) with a resolution of 640 480

a temperature range of up to 2000 C. The distancee camera and the powder bed was 330 mm, result-ld of view of 274 mm by 206 mm (with a 25 lens),l resolution of 175 m/pixel. The camera was pos-

p the build chamber and used a zinc-selenide (ZnSe)indow. This window type was selected due to itsittance of the IR wavelengths; it is the most widelyf window in IR imaging [10]. A mechanical shutter

was installed inside the chamber covering the ZnSe protect it from metallization that may occur during

of metal powder particles. The shutter also served to images were not obscured from unwanted particlesows surface. The angle between the surface normalera was 18 (with an incline allowance of up to

mputer-aided design (CAD) rendering of this setup Fig. 1. Images of the build surface were capturedamera following the completion of the melt processtomated system was developed using LabVIEW soft-nal Instruments, Austin, TX) to open the shutter for

pon completion of the melt step; images were cap-g the 1 s interval. This process was repeated for each

layer.

era calibration

ers are capable of providing quantitative thermalever, their measurement accuracy is dependent upon

of environmental conditions and surface propertiestivity and emissivity). Measuring and accounting fortions is critical to the calibration procedure of IRobtain absolute temperature measurements. The radi-

received from an object will be a function of its

E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139 33

cham

temperaturtion energytransmissiothat were mthe target, faces), andviewing glaare the amemissivity.

2.3.1. ExteExterna

the IR camtion, yielditransmissiobration sou

acedated

prevIR cckboing astalled u

the Fig. 1. CAD rendering of the Arcam A2 vacuum

e, spectral emissivity (a surfaces ability to emit radia-), reflections from its surroundings, and atmosphericn. Fig. 2 is a schematic representation of the sourceseasured by the IR camera. These sources include

radiation from surroundings (e.g. heat shield sur- emission and absorption from the environment (e.g.ss). The main sources of thermographic disturbance

bient sources, or reflecting sources, and the targets

was pland he(0.99),to the the blaassignthen inmodifithat ofrnal optics transmissionl optics transmission was a parameter assigned toera that supported its internal measurement func-ng a temperature value. To determine the percentn of the ZnSe window, a blackbody hot plate cali-rce (Omega model: BB-2A, Stamford, Connecticut)

2.3.2. MeaSurface

sures mean

view factoration energstriking a

Fig. 2. Schematic of the various sources of radiation thatber with IR Camera.

inside the build chamber atop the build platform to 300 C. The emissivity value for the blackbodyiously provided by the manufacturer, was assignedamera. An average temperature for the surface ofdy was recorded without the ZnSe window present,n external transmission of 100%. The window wased and the external optics transmission parameter wasntil the average temperature of the surface equaledtemperature measured without the viewing window.n radiant temperature calculationradiation can be measured by determining the enclo-

radiant temperature, or reflected temperature using calculations, or a geometric consideration for radi-y. It was important to first identify the reflections

surface. The Arcam A2 system uses a heat shield

an IR camera measures.

34 E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139

Fig. 3. CA coupl

placed atopinsulator tosource of robserved bof the six-sfaces (fronsurface, wemay be defiradiation heity being mequation:

Tmr = 4

where FSeach wall (jsurface s ththe j-th walthe view fa

F12 = 1A

where A istwo walls, and the vec

After idtion properwere speciassumed ttransparentshields areters and difassumption[4]. Also, isothermal were assum

[4]. The vculated indalgebra (u

en vA)

ed dmoc26, Ot, OmFig.

passof wriateof ted foratursum

ar shd to

pai

Emi rea

to eries consD rendering of the Arcam A2 heat shield used (bottom view) showing thermo

the build surface (Fig. 3) that serves as a thermal the build platform. The heat shield was the primaryeflection onto the build surface, or the surface beingy the IR camera. Thus, only the surface radiationsurface enclosure, comprised of five independent sur-t, left, rear, right, and top shield) as well as the buildre considered. The mean radiant temperature (Tmr)ned as the temperature of an isothermal cavity whoseat transfer to the sample is the same as the actual cav-easured [6], and was determined from the following

jT 4j FSj (1)

j is the view factor between the target surface (S) and), representing the fraction of the radiation leaving theat strikes the j-th wall. Tj represents the temperature ofl [6]. In Eq. (1), the basic formulation for determiningctor between two walls (F12) is:

1

A1

A2

cos 1 cos 2r2

dA1dA2 (2)

the area of each wall, r is the distance between the

and thick, MrecordK ther116U-cemen

walls (by thelation appropatures averagtempewas as

and reexposeof each

2.3.3. Any

abilityand vais not and is the angle between each walls normal vectortor pointing toward each wall.entifying the enclosure, the thermal state and radia-ties of each surface (shield walls) of the enclosurefied [23]. In the current analysis, the surfaces wereo be opaque, diffuse, and gray; therefore, non-

(which is a reasonable assumption since the heat solid steel sheets). The surfaces were diffuse emit-fuse reflectors independent of wavelength (a common

made in radiation problems to simplify analysis)each surface of the enclosure was assumed to haveproperties and the incoming and outgoing radiationsed to be uniformly distributed over each surface

iew factor for each wall of the enclosure was cal-ependently by using published formulae and plots,tilizing the basic formulation shown by Eq. (2)),

morphologis a blackbwhich radiaple reflectiothe materiaan internal samples in sivity valueconstant inremoved frcavity. Thethe hose ofopening fotion of pressintered pasintered poe locations (a) and isometric view of heat shield assembly (b).

erified using MATLAB software (MathWorks, Nat-[21]. Shield wall temperature measurements wereuring three different builds of Ti6Al4V using typeouples (Range: 270 to 1372 C, Model: HKQIN-

mega Stamford, CT) cemented (high-temperatureegabond 400, Omega, Stamford, CT) to the shield

3). The use of only three thermocouples was limited-through that was installed which allowed the instal-iring inside the system while still maintaining an

vacuum environment for fabrication. The temper-he shield walls were measured during each build,r each wall, and used to calculate the mean radiant

e using the view factor model. For this experiment, ited that the left and right shields as well as the frontields had the same temperatures (since they were

the same thermal environment and the surface areared shield was the same).

ssivity calculationl surface has a distinct emissivity surface property (themit radiation energy and denoted ), that is unitlessbetween 0 and 1 [4,12]. The emissivity of a surfacetant; instead, it varies with temperature and surface

y. A common tool used in determining emissivityody cavity, or radiator. A blackbody cavity is one intion enters through an opening and undergoes multi-ns within the cavity before exiting [4]. To determine

ls (Ti6Al4V) emissivity, a rectangular prism withcavity was fabricated via EBM. Construction of theseEBM was purposefully done to acquire the true emis-, since the unique EBM produced part surface is fairly

every Ti6Al4V build. Once fabricated, the part wasom the machine for thorough cleaning of the internal

powder inside the cavity was removed by directing Arcams powder recovery system toward the cavitysr 15 min. This is a system which uses a combina-surized air and metal powder (Ti6Al4V) to removerticles from a part. After the cavity was cleaned of allwder, it was used as a blackbody radiator.

E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139 35

Fig. 4. (a) CA ity, (btools and cem

To calcuheated to ma temperatu420, Cornichamber onon the hot ptop surfacesured by ceside walls omocouple tto yield a teblackbody ture measuthermocouVIEW as th

Emissiv

= T4r T

T 4S Twhere Tr iis the absoltemperaturcam ReseaWilsonvillecing a rectasurface andbody radiatby placing orifice (Figradiator dessivity (leve[3]. In thesorifice and to the spotThe mean iment descThermacam(Fig. 4b) w

orificemiscour to

etersatur

a CAge oalid

discrmin

xper

verifon hs andn propareded reme

o aD rendering (wire frame view) of blackbody radiator model and its internal cavented-on thermocouple.

late the emissivity of the material, the blackbody waseasure the thermal radiation coming from its orifice atre of 200300 C. A hot plate (Corning Model PC-

ng, Tewksbury, MA) was placed inside the vacuum top of the build platform. The blackbody was placedlate and a type K thermocouple was cemented on its

(Fig. 4b). Heat shield wall temperatures were mea-menting a second type K thermocouple to one of thef the heat shield enclosure, and the third type K ther-o the top shield of the enclosure. The hot plate was setmperature of 180 C and 200 C. IR images of thespecimen were captured and thermocouple tempera-rements were recorded using a National Instrumentsple input module (NI 9213, Austin, TX) and Lab-e data acquisition software.

ity was computed by the following equation [6]:4mr

4mr

(3)

s the radiant temperature of the targets surface, TSute surface temperature, and Tmr is the mean radiante (in temperature units of Kelvin). Using Therma-

of the as the thermoradiatoparamtempershowsIR imaUpon vwill beto dete

2.4. E

To radiatimodelficatioto comembedmeasu

cess. T

rcher Professional software (FLIR Systems, Inc.,, OR), the radiant temperature was measured by pla-ngular region of interest on the blackbody specimens

assigning it an emissivity of 0.99 to acquire black-ion. The absolute surface temperature was measureda spot meter tool directly on the blackbody radiator. 4a). According to Castrejon et al. [3] the blackbodyign (Fig. 4a) selected in this research yields an emis-l of blackness) at the cavitys orifice of 0.994 0.2e tests this value of emissivity was assigned to thean absolute temperature was obtained corresponding

meter tool placed within the orifice area (Fig. 4b).radiant temperature was determined by the exper-ribed in the previous section using Eq. (1). Using

software, the emissivity of the region of interestas adjusted until the temperature was equal to that

EBM on asurements in the expe270 to 13CT) througor build plawhich protfrom the hing of this sthe solid pding it wit(shown in were recordware and aIt is importpartially ex) IR image of the blackbody radiator with software measurement

e. The resulting emissivity value was the value usedsivity of melted (solid) Ti6Al4V. Additionally, a

ple was cemented on the surface of the blackbodybe used as a validation tool. The validity of input

was verified by comparing the IR cameras outpute measurement to that of the thermocouple. Fig. 4aD rendering of the blackbody radiator and Fig. 4b, anf the blackbody radiator at an elevated temperature.ation of the solid emissivity, a similar procedure thatussed later (using a thermocouple) was implementede the emissivity of the Ti6Al4V powder.

imental IR parameter verication

y the aforementioned theoretical models, that is, theeat transfer theory used to derive the view factor

the blackbody radiator model, an experimental veri-cedure was carried out. The experiment was designed

the temperature measurements from a thermocoupleduring fabrication and the IR cameras temperaturents taken consecutively during the embedding pro-

ccomplish this, a solid cube was fabricated using

thermocouple to make real-time temperature mea-while simultaneously acquiring IR images. Step oneriment was to insert a type K thermocouple (Range:72 C, Model: HKQIN-116U-26, Omega Stamford,h a drilled hole in the center of the EBM start plate,tform. The thermocouple contains an Inconel sheathects the junction from damage that may be incurredigh-power electron beam. Fig. 5a is a CAD render-etup. When the manufacturing process commenced,

art was manufactured on the thermocouple, embed-hin the solid part as the layered process continuedFig. 5b). Thermocouple temperature measurementsed using National Instruments data acquisition soft-

thermocouple input module (NI 9213, Austin, TX).ant to note that the tip of the thermocouple was onlyposed (2 mm) into the Z-axis build plane to prevent

36 E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139

Fig. 5. (a) CA platsolid, and (c)

Fig. 6. IR immeasure powd

bending orture measu(about 0.5 its counterpoint was EBM buildtured at layinstalled th1.40 mm, ameasureme

important tbeam scanat layer 19and 21. Fu(Table 4) arlayer to layusing a holpowder (exment showused as anand the emthe IR tempD rendering of thermocouple placement through drilled hole within the buildembedded thermocouple within sintered powder of an EBM-fabricated part.age of EBM part with embedded thermocouple to verify solid emissivity measuremer emissivity (b).

breaking of the powder re-coater blades. Tempera-rements recorded at the thermocouple junction pointmm from the thermocouple tip) were compared topart IR measurements. The thermocouple junction0.15 mm in diameter which corresponds to three

layers of IR measurements. Therefore IR images cap-ers 19, 20, and 21 (representing layers closest to theermocouple) or corresponding Z-heights 1.33 mm,nd 1.47 mm were used to compare IR temperaturents to thermocouple temperature measurements. It iso note that, for the experiment in Fig. 6a, the electronned over the thermocouple, which was still exposed

and fully covered by solid Ti6Al4V at layers 20rthermore, temperature differences between layerse most likely due to differences in scan direction fromer. The same validation procedures were followed

low cube to determine the emissivity of the sinteredperimental setup shown in Fig. 5c). For the experi-

n in Fig. 5c, the temperature of the thermocouple was accurate representation of the powder temperatureissivity of the surrounding powder was adjusted untilerature agreed with that of the thermocouple.

Upon dricated usiparts in EBa melted cyuntil build ically logsa single thform. Bothcorrespondtransmittanusing KimGA) and is

3. Results

3.1. Extern

Table 1 and the retransmissiodow showeemission. Tform, (b) EBM-fabricated part with embedded thermocouple forents (a) and embedded thermocouple within sintered powder to

etermining the IR parameters, cylinders were fab-ng the standard processing parameters to fabricateM using Ti6Al4V. The average temperature fromlinder was taken at every other layer and recorded

completion. Additionally, the EBM system automat- temperature measurements every 10 layers usingermocouple that is located below the build plat-

temperature measurements were plotted against theing layer and compared. To prevent reduction ince, the ZnSe glass was cleaned every three builds

wipes (Kimberly-Clark Global Sales, Inc., Roswell,opropyl-alcohol.

and discussion

al optics transmission

summarizes the parameters assigned to the cameracorded temperatures in testing the external opticsn. The temperature measurements without a win-d the ideal temperature, or the temperature withouthe temperature with the ZnSe window installed was

E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139 37

Table 1External optics testing of ZnSe window IR image parameters assigned.

Experiment Emissivity Ext. optics transmission Avg. temp. (C) Std. dev. (C)No window 0.99 1.00 309 5ZnSe window 0.99 1.00 298 2ZnSe window (adjusted) 0.99 0.94 309 2

reduced (11 C) due to the windows emission, and thus thepercent transmission camera parameter was modified until thetemperature with and without the glass installed was equal.The external optics transmission was found to be 94%. In thecameras internal measurement function, anything that was nottransmission was considered emission. Therefore, 6% of theradiation reaching the detector was from the window. Differentwindows have different emissions, thus new calculations willbe required if a different window is used. The experiment wasrepeated afthe shutter transmissiothat the shtion and fumaintained

3.2. Mean

The meaEq. (1). A stemperaturcalculationDue to theleft and rigresulting ma standard fTi6Al4Vcalculationthe processbeyond thobe of benefithe vacuumusing the IRfor each co

Table 2Mean radiant

View factors

Fsurface-top shieFsurface-left shieFsurface-rear shieFsurface-right shiFsurface-front sh

radiant temperature is calculated for every build. This can alsobe extended to produce automatic layer-by-layer mean radianttemperature measurements for further accuracy.

3.3. Emissivity studies

The determined emissivity for the solid (after the melt step)Ti6Al4Vage emissiv

M fre [ing

cons

d thocesthus

fabrermione

or o

g, or

valu

usingn equ

em

ed p ands de

shoe em

eters embremeig.

ity da

(solid(solidal. ter 5000 h of use to determine the effectiveness ofmechanism and it was found that the external opticsn was maintained at 94%. Thus, it can be concludedutter mechanism protects the glass from metalliza-rther allows an adequate level of transmission to be.

radiant temperature

n radiant temperature was calculated by employingummary of all view factors found, the average shieldes recorded during Ti6Al4V processing, and the

of the mean radiant temperature is given in Table 2. symmetrical configuration of the heat shield, theht shields were assigned the same view factor. Theean radiant temperature (Tmr = 342 C) can be used asor IR in situ measurements when processing standard

builds with EBM. It is important to note that a re- of mean radiant temperature would be required ifing conditions for Ti6Al4V using EBM changese recommended by Arcam. Furthermore, it wouldt to permanently install a set of thermocouples within

chamber, such as on the heat shield walls, when camera as this would ensure that shield temperaturesrresponding build are utilized and the correct mean

temperature calculations.

Average shield wall temperatures T4Fs jld = 0.05 Top shield avg. temp. (K) = 584 6.1E+09ld = 0.26 Left shield avg. temp. (K) = 641 4.3E+10ld = 0.18 Rear shield avg. temp. (K) = 638 3.0E+10eld = 0.26 Right shield avg. temp. (K) = 641 4.3E+10ield = 0.12 Front shield avg. temp. (K) = 638 2.0E+10

the EBliteratuAccordnearlydize anthat prment, duringity detundergcursor

meltin

3.4. E

By missioand andescribcamera

setup aFig. 6awith thparamcoupleMeasuest in F

Table 3Emissiv

Sample

EBM 1 EBM 2 Yang et

j

T 4Fsj = 1.4E+11

Tmr(K) = 4

jT 4Fsj = 615

Tmr(C) = 342

Yang et al. Yang et al. Yang et al. Yang et al. Yang et al. Yang et al. was measured to be 0.26. Table 3 lists the aver-ity measured from two different experiments usingabricated blackbody, including data obtained from25], and shows a close similarity in measurements.to Yang et al. [25] emissivity of Ti6Al4V remainstant up to 760 C before the material begins to oxi-e emissivity increases rapidly. It is important to notesing Ti6Al4V in EBM occurs in a vacuum environ-thermal oxidation is either limited or does not occurication. It is also important to note that the emissiv-ned applies only to the surface of the part after it hasmelting and will not be the same for the powder pre-ther surface morphologies such as the surface during

liquidus phase.

ation of experimental IR parameter verication

the calculated input parameters (atmospheric trans-al to 0.94, mean radiant temperature equal to 342 C,issivity of 0.26) determined using the previouslyrocedures, temperature measurements, using the IR

the embedded thermocouple with the experimentalscribed in Section 2.2, were acquired and compared.ws an IR image capture of the EBM part fabrication

bedded thermocouple used to verify the radiation. Fig. 6b shows the IR image captured of the thermo-edded within the sintered powder of a hollow cube.nts were obtained for the described region of inter-

6a, and Table 4 provides the results for 3 fabrication

ta of Ti6Al4V (solid).Temperature (C) Emissivity ()

) 177 0.26) 198 0.26

165 0.25

250 0.27350 0.23450 0.23550 0.24700 0.25750 0.27

38 E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139

Table 4Thermocouple and IR simultaneous temperature measurements for solid part.

Layer number Thermocouple (C) IR camera (C) % Difference19 730 703 3.7220 717 689 3.8421 712 685 3.76

Average 3.77

layers. As can be seen in the table, the IR camera and ther-mocouple measurements agree within an average difference of3.77%.

Emissivity of sintered powder was measured using the meth-ods described in Section 2.2 and IR temperature data was takenfrom the region of interest described in Fig. 6b. The mean radi-ant temperature determined in Section 3.2 was used for all IRtemperature calculations. Table 5 shows the results obtainedwhere the original IR temperature was corrected using differ-ent emissivity values until it was in close agreement with thethermocouple measurement that was taken simultaneously. Theresulting emissivity value for Ti6Al4V sintered powder wasdetermined to be 0.50. Powder emissivity will tend to vary fromthe particle size, shape, and packing used, thus the emissiv-ity determined is specific to Ti6Al4V powder used in thisstudy (purchased from Arcam AB). With results provided inTable 5, it ider throughupon depothe recentlyexcess coo

microstrucusing this iprocess by gradients d

3.5. Absolute temperature measurements

The temperature measurements from the IR camera and thoseobtained from the EBM system for a 35 mm tall build areplotted in Fig. 7. From the graph, it can be observed that thethermocouple temperature and IR temperatures were in closeagreement with each other at the beginning of fabrication (whenthe thermocouple is closest to the build surface). As fabrica-tion continued, the IR camera showed that the temperature atthe top surface increased while the temperature below the buildplatform decreased as more layers were fabricated. The resultsfrom Fig. 7 correlate closely to the expected behavior. That is, asfabrication progresses, the thermocouple below the build plat-form moves away from the heat source, or the pre-heating andmelting cycles. Subsequently, the build environments tempera-ture increases as the build progresses since more heat is added byeach layers pre-heating and melting cycles. Furthermore, a plotof uncorrected IR temperature data is shown using default cam-era parameters where the radiant temperature is set to 26.9 C.A difference of 366 C was evident when comparing the cor-rected temperature data using the developed thermal model tothe uncorrected temperature data using default camera settings.It is also important to note that the uncorrected temperature dataand the thermocouple data from the fabrication process do notagree.

resu

mpery siat thcou

e use

ricame ias a

Table 5Thermocoupl

Layer numbe Emis

19 0.50 20 0.50 21 0.50

Fig. 7. EBM from the bottos now possible to determine temperature of the pow-out processing. It was found that powder temperature

sition was 430 C. Such thermal gradient between melted part and newly deposited powder can cause

ling that leads to thermal stresses and undesirabletures (e.g. martensitic phase in Ti6Al4V). Bynformation, it may become possible to optimize theadding a powder heating system that reduces thermalue to powder deposition.

Theface tecan va

note ththermothat aring fabon a tiatures

e and IR simultaneous temperature measurements for sintered powder.

r IR camera, C (uncorrected) Thermocouple (C) 994 680 996 678 993 676 processing temperature measurements for temperature measurements using default IRm of the build platform.lts obtained here show the need to monitor the sur-rature during fabrication since the thermal behaviorgnificantly during processing. It is also important toe EBM system in its commercial state relies on the

ple measurements to provide absolute temperaturesd for parameter (e.g. beam power) adjustments dur-tion. The EBM software performs algorithms basedndependent heat equation to predict surface temper-

function of the thermocouple measurement under

sivity IR camera, C (corrected) % Difference678 0.25680 0.29677 0.15

Average 0.23

camera parameters, corrected parameters, and thermocouple data

E. Rodriguez et al. / Additive Manufacturing 5 (2015) 3139 39

the build platform, the parts geometry, and the Z-height. IRmeasurements yield real-time absolute surface temperature datathat is otherwise unachievable through the predictive mathemat-ical algorithms. The IR measurements obtained here showedthat the thermocouple temperature is different from the IRtemperature, specifically as a build progresses, thus the useof a surface temperature for parameter adjustments should beconsidered. Furthermore, it is important to ensure absolute tem-perature measurements are being obtained, specifically whenusing temptrol, and/omaterials.

4. Conclu

To use Itor must protemperatursupport anabsolute tewas develoments for Tpost-meltinview factorThe view fthe reflecteture measuradiant temEBM yieldtured IR imidentify abSuch theorbe used in graphy. Fubetween prsurements,EBM-fabri

Reference

[1] Amato KCompari(Alloy 62012:34

[2] Bayle F,high spemicro- a

[3] Castrejoopment, Instrum

[4] Cengel YMcGraw

[5] Christensen A, Kircher R, Lippincott A. Qualification of electron beammelted Ti6Al4V for orthopaedic applications. In: Materials and pro-cesses for medical devices conference. 2008. p. 4853.

[6] Coppa P, Consorti A. Normal emissivity of samples surrounded by surfacesat diverse temperatures. Measurement 2005:12431.

[7] Dinwiddie R, Dehoff R, Lloyd P, Lowe L, Ulrich J. Thermographic in situprocess monitoring of the electron beam melting technology used in addi-tive manufacturing. In: Thermosense: thermal infrared applications XXXV.Baltimore, MD: SPIE; 2013. p. 19.

[8] Gibson I, Rosen DW, Stucker B. Additive manufacturing technologies.New York: Springer; 2010.

rnandcturericateSS. 1: IRISauss H

melt101ldaguting. Nrtinezcturetallogreles ng infblishinreles electroricatirr LE

al. Minufac6Alrr LE, et a

ed by0:188rr LEcrostrm merr LEcturee iron85.

mirez et alricatictron drigueelectrblishinwerding e

arrow Corphlers

ng J, Sment Ti6Alerature measurements for layerwise monitoring, con-r parameter development for processing of novel

sions

R imaging with a certain level of accuracy, the opera-vide the camera with the targets emissivity, reflected

e, and optics transmission as they are necessary to infrared cameras measurement functions to obtainmperature measurements. A step-by-step procedureped in this research to obtain emissivity measure-i6Al4V material pre-melting (powder form) andg. A thermal model was established to determine thes associated with the Arcam A2s thermal enclosure.actors were found to be necessary when computingd temperature and thus, with shield wall tempera-rements, it became possible to determine the meanperature. The integrated thermal-imaging system ins a higher level of feedback that can allow for cap-ages to be analyzed and processed as a means to

solute thermal non-uniformity on the parts surface.y and calculations expressed in this research canother AM technologies that use in situ IR thermo-ture work will consist of establishing a correlationocessing parameters, IR surface temperature mea-

and mechanical and microstructural properties forcated parts.

s

N, Hernandez J, Murr LE, Martinez E, Gaytan SM, Sindo PW.son of microstructures and properties for a Ni-base superalloy25) fabricated by electron and laser beam melting. J Mater Sci Res1.

Doubenskaia M. Selective laser melting process monitoring withed infrared camera and pyrometer. Fundamentals of laser assistednd nanotechnologies. In: SPIE. 2008. p. 18.n-Garcia R, Castrejon-Pita JR, Castrejon-Pita AA. Design, devel-and evaluation of a simple backbody radiative source. Rev Sci

2010;81(5). p. 055106.A. Heat and mass transfer: a practical approach. Boston, MA:

Hill; 2007.

[9] Hestrufab

[10] IRIFL

[11] Krtive999

[12] Mates

[13] MastruMe

[14] MiusiPu

[15] Miin fab

[16] Muet ma

Ti[17] Mu

DHcat201

[18] MuMibea

[19] Mustrupur376

[20] RaBI,fabele

[21] Roin Pu

[22] Schdur

[23] Sping

[24] Wo[25] Ya

eleof ez J, Li SJ, Martinez E, Murr LE, Pan XM, Amato KN, et al. Micro-s and hardness properties for beta phase Ti24Nb4Zr7.9Sn alloyd by electron beam melting. J Mater Sci Technol 2013:10117.0 things you need to know about infrared windows. Bradenton,S, Inc.; 2013., Eschey C, Zaeh MF. Thermography for monitoring the selec-

ing process. In: Solid freeform fabrication symposium. 2012. p.4.e X. Theory and practice of infrared technology for nondestructiveew York, NY: John Wiley & Sons; 2001.

E, Murr LE, Hernandez J, Pan X, Amato K, Frigola P, et al. Micro-s of niobium components fabricated by electron beam melting.r Microstruct Anal 2013;2(3):1839.

J. Process study and control of electron beam melting technologyrared thermography. Ann Arbor, MI: Proquest, UMI Dissertationsg; 2013.

J, Terrazas C, Medina F, Wicker RB. Automatic feedback controln beam melting using infrared thermography. In: Solid freeform

on symposium. 2013. p. 70817., Esquivel EV, Quinones SA, Gaytan SM, Lopez MI, Martinez EY,crostructures and mechanical properties of electron beam-rapidtured Ti6Al4V biomedical prototypes compared to wrought4V. Mater Charact 2009:96105., Gaytan SM, Ceylan A, Martinez E, Martinez JL, Hernandez

l. Characterization of titanium aluminide alloy components fabri- additive manufacturing using electron beam melting. Acta Mater794.

, Martinez E, Pan XM, Gaytan SM, Castro JA, Terrazas CA, et al.uctures of Rene 142 nickel-based superalloy fabricated by electronlting. Acta Mater 2013:428996., Martinez E, Pan X, Meng C, Yang J, Li S, et al. Micro-

s and properties of solid and reticulated mesh components of fabricated by electron beam melting. J Mater Res Technol 2013:

DA, Murr LE, Martinez E, Hernandez DH, Martinez JL, Machado. Novel precipitate-microstructural architecture developed in theon of solid copper components by additive manufacturing usingbeam melting. Acta Mater 2011:408899.z E. Development of a thermal imaging feedback control system

on beam melting. Ann Arbor, MI: ProQuest, UMI Dissertationsg; 2013.

tfeger J, Singer RF, Korner C. In situ flaw detection by IR-imaginglectron beam melting. Rapid Prototyp J 2012:25963.

EM. Radiation heat transfer. New York, NY: Hemisphere Publish-oration; 1978.

T. Wohlers report. Fort Collins, CO: Wohlers Associates; 2013.un S, Brandt M, Yan W. Experimental investigation and 3D finite

prediction of the heat affected zone during laser assisted machining4V alloy. J Mater Process Technol 2010:221522.

Approximation of absolute surface temperature measurements of powder bed fusion additive manufacturing technology using in...1 Introduction2 Methods2.1 Electron beam melting process2.2 Thermal imaging system installation2.3 IR camera calibration2.3.1 External optics transmission2.3.2 Mean radiant temperature calculation2.3.3 Emissivity calculation

2.4 Experimental IR parameter verification

3 Results and discussion3.1 External optics transmission3.2 Mean radiant temperature3.3 Emissivity studies3.4 Evaluation of experimental IR parameter verification3.5 Absolute temperature measurements

4 ConclusionsReferences