Influence of Powder Surface Contamination in the Ni

Article Inﬂuence of Powder Surface Contamination in the Ni-Based Superalloy Alloy718 Fabricated by Selective Laser Melting and Hot Isostatic Pressing

Yen-Ling Kuo * and Koji Kakehi

Department of Mechanical Engineering, Tokyo Metropolitan University, 1-1 Minami-Osawa, Hachioji-Shi, Tokyo 192-0397, Japan; [email protected] * Correspondence: [email protected]; Tel./Fax: +81-42-677-2709

Received: 9 May 2017; Accepted: 7 September 2017; Published: 13 September 2017

Abstract: The aim of this study was to gain a deep understanding of the microstructure-mechanical relationship between solid-state sintering and full-melting processes. The IN718 superalloy was fabricated by hot isostatic pressing (HIP) and selective laser melting (SLM). Continuous precipitates were clearly localized along the prior particle boundary (PPB) in the HIP materials, while SLM materials showed a microstructure free of PPB. The mechanical properties of specimens that underwent SLM + solution treatment and aging were comparable to those of conventional wrought specimens both at room temperature and 650 ◦C. However, a drop was observed in the ductility of HIP material at 650 ◦C. The brittle particles along the PPB were found to affect the HIP materials’ creep life and ductility during solid-state sintering.

Keywords: superalloy; surface contamination; prior particle boundary; hot isostatic pressing; selective laser melting

1. Introduction IN718 superalloy is widely used in gas turbine and related aerospace applications due to its excellent hot corrosion resistance, good weldability, and high mechanical properties—e.g., creep and fatigue strengths [1–3]. Macro segregation occurs during the conventional ingot metallurgy route due to melt-related problems such as segregation and porosity, which lead to degradation of the mechanical properties of the alloys [4,5]. One solution to this problem was to minimize segregation through rapid solidiﬁcation of the metal or through the solid-state sintering process. A powder metallurgy technology—i.e., hot isostatic pressing (HIP)—was developed to eliminate segregation and macro-porosity, and to manufacture net shape components with homogeneous microstructure and improved mechanical properties [6]. Moreover, the temperature involved in the HIP process allows segregants to diffuse, thus enhancing the mechanical properties. After this type of solid-state sintering process, special care—i.e., rolling, extrusion and forging—is required to mitigate the detrimental effects of prior particle boundary (PPB) precipitation [7]. The so-called PPB problem results from surface contamination with pre-alloyed powder, and particles such as oxides, carbides, oxy-carbides and oxycarbonitrides are formed along the PPB [8–10]. These precipitates along the PPB are brittle and thus provide an easy fracture path, and so the undesirable effects of PPBs have been focused on because they limit the applications of such consolidated products in HIPed materials [11]. Although many efforts have been made to decrease the effects of PPBs to extend the life of such products in applications in the aerospace industry, no effective and reliable method has yet been determined. Hence, an additive manufacturing technology, selective laser melting (SLM), was used in the current research. The laser beam irradiates metal powders until they fully melt in the SLM process, which has the beneﬁt of causing PPB breakdown. SLM is applicable to components of fully dense bulk material with complex

Metals 2017, 7, 367; doi:10.3390/met7090367 www.mdpi.com/journal/metals Metals 2017, 7, 367 2 of 13 geometry [12]. In spite of the fact that the process is capable of making full density (~98–99%) parts, post-processes such as HIP were commonly used to eliminate the small amount of residual porosity for critical applications where high strength and fatigue resistance were necessary [13,14]. The obvious characteristics of the mechanical properties of SLM parts include anisotropy; the reasons for the special characteristics of IN718 built up by SLM—e.g., the high dislocation density, subgrain boundary and a rowMetals of interdendritic2017, 7, 367 δ-phase precipitates—have been explained in our previous2 studies of 13 from several perspectivesdense bulk material [15,16 with]. However,complex geometry research [12]. relatedIn spite of to the the fact surface that the contaminationprocess is capable onof powder in SLM materialsmaking full isdensity limited. (~98–99%) Therefore, parts, post-process the currentes such study as HIP focused were co onmmonly surface used contaminationto eliminate and on the consolidationthe small amount behaviour of residual between porosity SLMfor critical and conventionalapplications where HIP. high In strength order to and gain fatigue an in-depth understandingresistance of were the necessary surface contamination[13,14]. The obvious by char theacteristics powder of boththe mechanical in the SLM properties and of HIP SLM processes, parts include anisotropy; the reasons for the special characteristics of IN718 built up by SLM—e.g., the microstructure-mechanicalthe high dislocation density, relationship subgrain boundary of IN718 and a nickel-based row of interdendritic alloy fabricatedδ-phase precipitates— by HIP and SLM processeshave were been compared explained in in theour currentprevious study.studies from several perspectives [15,16]. However, research related to the surface contamination on powder in SLM materials is limited. Therefore, the current 2. Materialsstudy and focused Experimental on surface contamination Procedure and on the consolidation behaviour between SLM and conventional HIP. In order to gain an in-depth understanding of the surface contamination by the Thepowder prealloyed both IN718in the SLM superalloy and HIP processes, powder wasthe microstructure-mechanical prepared by vacuum meltingrelationship and of atomizedIN718 into powdersnickel-based using the gas-atomizingalloy fabricated by method. HIP and SLM The pr gas-atomizedocesses were compared IN718 powdersin the current used study. for both the HIP process and the SLM process were typically spherical particles that were ﬁner and had a broad 2. Materials and Experimental Procedure particle size distribution below 60 µm (Figure1), which may provide high packing density and reduce the percentageThe of prealloyed voids [ 17IN718,18]. superalloy The chemical powder compositionwas prepared by of vacuum IN718 melting powder and usedatomized in theinto HIP and powders using the gas-atomizing method. The gas-atomized IN718 powders used for both the HIP SLM processes is shown in Table1. The oxygen contents in the powders used for HIP and SLM were process and the SLM process were typically spherical particles that were finer and had a broad comparable.particle For size the distribution HIP process, below the 60 powders μm (Figure were 1), entirely which may canned provide into high SUS304 packing stainless density steel and capsules −3 which wasreduce heated the andpercentage evacuated of voids to [17,18]. 1.0 × 10 The chemicPa byal an composition oil-diffusion of IN718 pump. powder An used optimal in the HIP HIP condition which consistingand SLM ofprocesses a soaking is shown temperature in Table 1. ofThe 1180 oxyg◦enC contents and a pressure in the powders of 175 used MPa for forHIP 4 and h was SLM used [19]. On the otherwere hand,comparable. a set For of the SLM HIP process process, parametersthe powders were (laser entirely power: canned 400 into W; SUS304 scanning stainless speed: steel 7.0 m/s; capsules which was heated and evacuated to 1.0 × 10−3 Pa by an oil-diffusion pump. An optimal HIP µ µ µ layer pitch:condition 20 m; which layer consisting thickness: of a 40 soakingm;beam temperature diameter: of 1180 100 °C andm; atmosphere:a pressure of 175 pure MPa 99.9999% for 4 h argon) 3 was utilizedwas toused fabricate [19]. On anthe INother 718 hand, cube a set with of SLM the dimensionsprocess parameters35 × (laser35 × power:35 mm 400. W; The scanning SLM cube was built usingspeed: the 7.0 alternating m/s; layer pitch: scan 20 vector μm; layer with thickness: a 66.7◦ 40rotation μm; beam of diameter: the scanning 100 μm; direction.atmosphere: The pure scanning strategy of99.9999% SLM process argon) was that utilized were to used fabricate are illustratedan IN 718 cube in with Figure the2 dimensionsa. After the 35 HIP× 35 × and 35 mm SLM3. The processes, the specimensSLM cube were was subjected built using tothe solution alternating treatment scan vector and with aginga 66.7° (STA):rotation aof solution the scanning treatment direction. at 980 ◦C The scanning strategy of SLM process that were used are illustrated in Figure 2a. After the HIP and ◦ for 1 h/airSLM cooling processes, (AC) the to specimens room temperature were subjected and to a solution two-step treatment aging treatmentand aging (STA): consisting a solution of 718 C for ◦ ◦ 8 h/furnacetreatment cooling at 980 to621 °C forC 1 and h/air holding cooling (AC) at 621 to roomC for temperature 10 h before and AC a totwo-step room aging temperature. treatment The heat histories ofconsisting the specimens of 718 °C arefor shown8 h/furnace in Figurecooling3 toa,b. 621 Tensile °C and testholding and at creep 621 °C test for specimens 10 h before wereAC to cut from the HIPedroom ingots temperature. and cubed The using heat histories a spark of cutter, the spec andimens the gageare shown length in Figure of each 3a,b. was Tensile 19.6 ×test2.8 and× 3.0 mm3 creep test specimens were cut from the HIPed ingots and cubed using a spark cutter, and the gage (Figure2b). The tensile tests were conducted at room temperature (25 ◦C) and at 650 ◦C under a strain length of each was 19.6 × 2.8 × 3.0 mm3 (Figure 2b). The tensile tests were conducted at room −4 −1 ◦ rate of 4.25temperature× 10 s(25 °C). The and creep at 650 °C test under was a conductedstrain rate of at4.25 650 × 10−C4 s− and1. The 550 creep MPa. test was The conducted microstructures were observedat 650 by°C aand scanning 550 MPa. electron The microstructures microscope were (HITACHI observed S-3700N, by a scanning Tokyo, electron Japan), microscope and transmission electron microscope(HITACHI S-3700N, (TEM, Tokyo, JEOL JEM-3200FS,Japan), and transmission Tokyo, Japan). electron Inverse microscope pole (TEM, ﬁgures JEOL (IPF) JEM-3200FS, were calculated from theTokyo, orientation Japan). measurementsInverse pole figures by (IPF) electron were backscattercalculated from diffraction the orientation (EBSD, measurements Oxford Instruments, by electron backscatter diffraction (EBSD, Oxford Instruments, Oxfordshire, UK). X-ray diffraction Oxfordshire,(XRD, UK). Rigaku X-ray Ultima diffraction IV, Tokyo, (XRD, Japan) Rigaku was used Ultima to determine IV, Tokyo, the relative Japan) amounts was used of phases to determine in the relative amountsthe specimens. of phases in the specimens.

Figure 1. Cont. Metals 2017, 7, 367 3 of 13 Metals 2017, 7, 367 3 of 13 Metals 2017, 7, 367 3 of 13 Metals 2017, 7, 367 3 of 13

Figure 1.FigureSecondary 1. Secondary electron electron (SE) (SE) images images of ofgas-atomized gas-atomized IN718 IN718 powders powders used for used (a) hot for isostatic (a) hot isostatic pressingFigure 1.pressing (HIP) SecondarySecondary and (HIP) (b electronelectron)and selective (b) selective (SE)(SE) laser imagesimages laser melting melting ofof (SLM);gas-atomized (SLM); SE images images IN718 of ( ofc) particles (powdersc) particles along used the along prior for the (particlea)) prior hothot isostaticisostatic particle boundary (PPB) in an as-HIPed specimen and (d) an as-built SLM specimen. boundarypressing (HIP) (PPB) and in an (b) as-HIPed) selectiveselective specimen laserlaser meltingmelting and (SLM); ((SLM);d) an as-built SESE imagesimages SLM ofof specimen.((c)) particlesparticles alongalong thethe priorprior particleparticle boundary (PPB) in an as-HIPed specimen and (d) an as-built SLM specimen. boundary (PPB) inTable an as-HIPed 1. Chemical specimen composition and of (IN718d) an powdas-builter used SLM in specimen.HIP and SLM (mass %). Table 1. Chemical composition of IN718 powder used in HIP and SLM (mass %). IN718 Cr Nb Mo Ti Al Co Cu C Si, Mn P, S B O Fe Ni Table 1. Chemical composition of IN718 powder used in HIP and SLM (mass %). HIP Table 18.4 1. 4.98Chemical 3.3 composition0.84 0.57 0.1of IN7180.01 powd0.026 er used 0.08 in 0.02 HIP and0.001 SLM0.014 (mass 17.2 %). Bal. SLM 19.6 5.05 2.85 1.10 0.46 0.03 0.05 0.04 0.04 0.0 0.002 0.019 Bal. 52.59 IN718IN718 Cr Cr Nb Nb Mo Mo Ti Ti Al Al Co Co Cu Cu C C Si, Si, Mn Mn P, S P, S B B O O Fe Fe Ni Ni IN718 Cr Nb Mo Ti Al Co Cu C Si, Mn P, S B O Fe Ni HIP 18.4 4.98 3.3 0.84 0.57 0.1 0.01 0.026 0.08 0.02 0.001 0.014 17.2 Bal. HIPHIP 18.4 18.4 4.98 4.98 3.3 3.3 0.84 0.84 0.57 0.57 0.1 0.1 0.01 0.01 0.0260.026 0.080.08 0.02 0.02 0.0010.001 0.014 0.014 17.217.2 Bal. Bal. SLM 19.6 5.05 2.85 1.10 0.46 0.03 0.05 0.04 0.04 0.0 0.002 0.019 Bal. 52.59 SLMSLM 19.6 19.6 5.05 5.05 2.85 2.85 1.10 1.10 0.46 0.46 0.03 0.03 0.05 0.05 0.04 0.04 0.04 0.04 0.0 0.0 0.0020.002 0.019 0.019 Bal. Bal. 52.59 52.59

Figure 2. Illustrations of (a) scanning strategy of the SLM process and (b) dimensions of the mechanical test specimen (mm).

Figure 2. Illustrations of (a) scanning strategy of the SLM process and (b) dimensions of the Figure 2.2.Illustrations Illustrations of of (a )( scanninga) scanning strategy strategy of the of SLM the process SLM process and (b) dimensionsand (b) dimensions of the mechanical of the mechanical test specimen (mm). testmechanical specimen test (mm). specimen (mm).

Figure 3. Heat history of the (a) HIP process and (b) SLM process.

Figure 3. Heat history of the (a) HIP process and (b) SLM process. Figure 3. HeatHeat historyhistory ofof thethe ((a)) HIPHIP processprocess andand ((b)) SLMSLM process.process.

Metals 2017, 7, 367 4 of 13

3. Results

3.1. Microstructures of as-HIPed Specimen and as-Built Specimen Densities of HIP and SLM, C&W alloys were 8.19, 8.18, and 8.20 g/cm3 respectively, as determined by the Archimedes principle. Investigation by X-ray diffraction (XRD) revealed the major peaks of γ matrix, while the oxide peaks such as Al2O3 or TiO2 could not be identified due to the limited volume fraction (Figure4). Earlier studies showed comparable results which did not show the oxides peaks in the as-HIPed and as-built SLM specimens [19–21]. However, the clearly continuous precipitates were localized along the PPB in the as-HIPed specimen from the SEM observations (Figure1c). The precipitates along the PPB in the as-HIPed specimen were mainly oxides (Figure5a,c). Al, which easily reacts with oxygen, was found to be the major metallic element in an oxide layer of Al2O3 (Figure5c). By contrast, the as-built specimen built up by the SLM process showed a free-PPB microstructure (Figure1d). Oxides and carbides induced by oxygen and carbon contamination of pre-alloyed powder were rarely found in the as-built SLM specimen (Figure5b). EBSD measurements were used to establish the crystallographic orientation of the γ matrix for all of the specimens. The orientations were colored in terms of a [001] inverse pole figure coloring scheme as shown in the legend (Figure6). IPF maps of the as-HIPed and as-built specimens are shown in Figure6a,b by means of EBSD, respectively. The equiaxed grains with an average size of 8.8 µm were observed in the Metals 2017, 7, 367 4 of 13 as-HIPed specimen (Figure6a), whereas the columnar grains were observed in the as-built specimen built up by the3. SLM Results process (Figure6b). These columnar grains were attributed to the epitaxial growth and heat flux during3.1. Microstructures the solidification of as-HIPed Specimen in the and SLM as-Built process Specimen [22]. Densities of HIP and SLM, C&W alloys were 8.19, 8.18, and 8.20 g/cm3 respectively, as 3.2. Influence ofdetermined the Heat-Treatment by the Archimedes Process principle. Investigation by X-ray diffraction (XRD) revealed the major peaks of γ matrix, while the oxide peaks such as Al2O3 or TiO2 could not be identified due to the limited IPF mapsvolume of the fraction HIP (Figure + STA 4). Earlier specimen studies showed and comparable SLM + STAresults specimenwhich did not show are shownthe oxides in Figure6c,d, respectively. Thepeaks SLM in the +as-HIPed STA specimen and as-built hadSLM specim IPF imagesens [19–21]. similar However, to the those clearly of continuous the as-built specimen precipitates were localized along the PPB in the as-HIPed specimen from the SEM observations (Figure6b,d). For(Figure the 1c). HIP The precipitates materials, along the the equiaxed PPB in the as-HIPed grains specimen structure were could mainly oxides be observed (Figure both before and after the heat-treatment5a,c). Al, which easily process reacts with (Figure oxygen, 6wasa,c). found The to be fine the major equiaxial metallic grainselement in were an oxide observed in the layer of Al2O3 (Figure 5c). By contrast, the as-built specimen built up by the SLM process showed a HIP materials becausefree-PPB microstructure of the presence (Figure of 1d). the Oxides PPB, and which carbides would induced impede by oxygen grain and growth carbon [23]. Moreover, carbides alongcontamination the PPB appeared of pre-alloyed to powder increase were rarely after found heat in the treatment as-built SLM when specimen compared (Figure 5b). with as-HIPed EBSD measurements were used to establish the crystallographic orientation of the γ matrix for all of material (Figurethe7 specimens.). The previous The orientations oxides were colored along in the terms PPB of a [001] promoted inverse pole the figure subsequent coloring scheme precipitation of carbides, sinceas specific shown in the oxides legend act(Figure as 6). nuclei IPF maps for of thethe as-HIPed formation and as-built of MC-type specimens are carbides shown in or M23C6-type carbides alongFigure the PPB 6a and (Figure Figure 6b7 ).by Duringmeans of EBSD, the heat-treatmentrespectively. The equiaxed process, grains titaniumwith an average and size carbon of within the 8.8 μm were observed in the as-HIPed specimen (Figure 6a), whereas the columnar grains were powder wouldobserved migrate in the to as-built the specific specimen oxidesbuilt up by on the the SLM particle process (Figure surfaces, 6b). These resulting columnar grains in the formation of stable titaniumwere oxy-carbides attributed to alongthe epitaxial the PPBgrowth [24 and]. heat flux during the solidification in the SLM process [22].

Figure 4. X-ray diffraction (XRD) patterns of as-HIPed and as-built SLM samples. Figure 4. X-ray diffraction (XRD) patterns of as-HIPed and as-built SLM samples.

Metals 2017, 7, 367 5 of 13

Figure 5. Transmission electron microscopy (TEM) bright field micrograph of (a) as-HIPed specimen and (b) as-builtFigure 5.SLM Transmission specimen; electron Digital microscopy maps (TEM) of brightenergy field dispersive micrograph ofspectroscopy (a) as-HIPed specimen (EDS) analysis Figure 5. Transmissionand (b) as-built electronSLM specimen; microscopy Digital maps (TEM) of brightenergy ﬁelddispersive micrograph spectroscopy of ((EDS)a) as-HIPed analysis specimen showingand (b) as-built distributionshowing SLM distribution specimen; of Al, ofO, Al, DigitalTi O, and Ti mapsandNb Nb elements of elements energy within dispersive (c )( cas-HIPed) as-HIPed spectroscopy specimen specimen (EDS)and (d ) analysis andas-built (d ) showingas-built SLMdistribution specimen.SLM of specimen. Al, O, Ti and Nb elements within (c) as-HIPed specimen and (d) as-built SLM specimen.

Figure 6. Inverse pole figures (IPFs) of (a) as-HIPed; (b) as-built SLM; (c) HIP + solution treatment and aging (HIP + STA) and (d) SLM + STA specimens were analyzed using the orientation measurements by electron backscatter diffraction (EBSD). FigureFigure3.2. 6. InverseInfluenceInverse poleofpole the figuresHeat-Treatmentfigures (IPFs)(IPFs) ofProcess of ( a()a as-HIPed;) as-HIPed; (b ()b as-built) as-built SLM; SLM; (c) ( HIPc) HIP + solution + solution treatment treatment and andaging aging (HIPIPF +(HIP STA)maps +and ofSTA) the (d )HIPand SLM + ( +STAd) STA SLM specimen specimens + STA and specimenswereSLM analyzed+ STA werespecimen using analyzed theare orientationshown using in Figurethe measurements orientation 6c,d, measurementsby electronrespectively. backscatter by The electron SLM diffraction + ba STAckscatter specimen (EBSD). diffraction had IPF images (EBSD). similar to those of the as-built specimen (Figure 6b,d). For the HIP materials, the equiaxed grains structure could be observed both before and after the heat-treatment process (Figure 6a,c). The fine equiaxial grains were observed in the HIP materials 3.2.3.3. Influence Tensile Properties of the Heat-Treatment Process ◦ IPFEngineering maps of stress-strainthe HIP + STA curves specimen and the and tensile SLM properties + STA specimen at room temperatureare shown in and Figure 650 C6c,d, are respectively.shown in Figures The SLM8 and +9 STA, Tables specimen2 and3 ,had respectively. IPF images At similar room temperature,to those of the the as-built as-built specimen specimen (Figure of 6b,d).SLM materialFor the HIP showed materials, lower tensilethe equiaxed properties grains than structure those of thecould as-HIPed be observed material both (Figure before8). and After after the theSTA heat-treatment heat-treatment process process, (Figure theproof 6a,c). andThe fine tensile equiaxial strengths grains of thewere SLM observed materials in the were HIP improvedmaterials significantly compared with the as-built SLM materials (Table2 and Figure8). The tensile properties in the heat-treated materials were comparable to those of the conventionally wrought specimen at room temperature, as shown in Table2. The principal strengthening phase in IN718 was the γ00 precipitate that promoted the strengths of heat-treated materials (Figure 10). The ductility of SLM + STA materials Metals 2017, 7, 367 6 of 13 at room temperature was related to the fine dimples present on the fracture surface (Figure 11b). MetalsIt was 2017 also, 7, noted367 that a mixed fracture surface which consisted of fine dimples and a few particle6 of 13 boundary facets was observed in the HIP + STA material (Figure 11a). The strengths and ductility of becausethe SLM of + the STA presencespecimen of werethe PPB, comparable which would to those impede of the grain conventionally growth [23]. wrought Moreover, specimens, carbides both along at theroom PPB temperature appeared to and increase 650 ◦C, after as shownheat treatment in Tables when2 and compared3. However, with the as-HIPed ductility material of the HIP(Figure + STA 7). Thespecimen previous was oxides much along lower the than PPB those promoted of the conventionallythe subsequent wroughtprecipitation specimen of carbides, and the sinceSLM specific + STA ◦ oxidesspecimen act atas 650 nucleiC (Tablefor the3). formation The ductility of MC-type of the HIP carbides + STA or specimen M23C6-type was carbides only one-tenth along the of thatPPB (Figureof the cast 7). During and wrought the heat-treatment specimen at process, 650 ◦C tita (Tablenium3). and A fracture carbon within surface the was powder observed would along migrate the tograin the boundaries specific oxides covered on withthe particle PPB in thesurfaces, HIP + STAresulting specimen in the (Figure formation 11c), whereas,of stable fracture titanium surface oxy- carbidesspecimens along were the observed PPB [24]. along the dendrite structure in the SLM + STA (Figure 11d).

Figure 7. TEMTEM bright bright field ﬁeld micrographs micrographs of of(a) (particlesa) particles along along the thePPB; PPB; (b) growth (b) growth behavior behavior of oxy- of carbides;oxy-carbides; (c) EDS (c) EDSdigital digital maps maps showing showing the distribution the distribution of elements of elements within within the HIPed the HIPed material material after theafter heat-treatment the heat-treatment process. process. Metals 2017, 7, 367 7 of 13 3.3. Tensile Properties Engineering stress-strain curves and the tensile properties at room temperature and 650 °C are shown in Figures 8 and 9, Tables 2 and 3, respectively. At room temperature, the as-built specimen of SLM material showed lower tensile properties than those of the as-HIPed material (Figure 8). After the STA heat-treatment process, the proof and tensile strengths of the SLM materials were improved significantly compared with the as-built SLM materials (Table 2 and Figure 8). The tensile properties in the heat-treated materials were comparable to those of the conventionally wrought specimen at room temperature, as shown in Table 2. The principal strengthening phase in IN718 was the γ″ precipitate that promoted the strengths of heat-treated materials (Figure 10). The ductility of SLM + STA materials at room temperature was related to the fine dimples present on the fracture surface (Figure 11b). It was also noted that a mixed fracture surface which consisted of fine dimples and a few particle boundary facets was observed in the HIP + STA material (Figure 11a). The strengths and ductility of the SLM + STA specimen were comparable to those of the conventionally wrought specimens, both at room temperature and 650 °C, as shown in Tables 2 and 3. However, the ductility of the HIP + STA specimen was much lower than those of the conventionally wrought specimen and Figure 8. Engineering stress-strain curves of HIP specimen and SLM specimens at room temperature. the SLMFigure + STA 8. Engineering specimen stress-strain at 650 °C (Table curves 3).of HIPThe spec ductilityimen and of the SLM HIP specimens + STA specimenat room temperature. was only one- tenth of that of the cast and wrought specimen at 650 °C (Table 3). A fracture surface was observed along the grain boundaries covered with PPB in the HIP + STA specimen (Figure 11c), whereas, fracture surface specimens were observed along the dendrite structure in the SLM + STA (Figure 11d).

Figure 9. Engineering stress-strain curves of HIP specimen and SLM specimens at 650 °C.

Table 2. Tensile properties at room temperature.

σ0.2 (MPa) σT (MPa) εf (%) As-HIPed 833 1225 32.1 As-built SLM 677 1023 28.1 HIP + STA 1125 1473 21.2 SLM + STA 1271 1425 18.6 Wrought + STA 1216 1435 24.0

Table 3. Tensile properties at 650 °C.

σ0.2 (MPa) σT (MPa) εf (%) As-HIPed 844 978.0 6.8 As-built SLM 594 862.0 25.1 HIP + STA 1031 1093 0.9 SLM + STA 1042 1142 10.1 Wrought + STA 1028 1139 10.5

Metals 2017, 7, 367 7 of 13

MetalsFigure2017 8., 7Engineering, 367 stress-strain curves of HIP specimen and SLM specimens at room temperature.7 of 13

FigureFigure 9. Engineering 9. Engineering stress-strain stress-strain curves curves ofof HIPHIP specimen specimen and and SLM SLM specimens specimens at 650 at ◦650C. °C.

TableTable 2. 2. TensileTensile properties at at room room temperature. temperature.

σ0.2 σ(MPa)0.2 (MPa) σσTT (MPa)(MPa) εf (%)εf (%) As-HIPedAs-HIPed 833833 12251225 32.1 32.1 As-builtAs-built SLM SLM 677 677 1023 1023 28.1 28.1 HIP +HIP STA + STA 1125 1125 1473 1473 21.2 21.2 SLMSLM + STA + STA 1271 1271 1425 1425 18.6 18.6 Wrought + STA 1216 1435 24.0 Wrought + STA 1216 1435 24.0

Table 3. Tensile properties at 650 ◦C. Table 3. Tensile properties at 650 °C.

σ0.2 σ(MPa)0.2 (MPa) σσTT (MPa)(MPa) εf (%)εf (%) As-HIPedAs-HIPed 844844 978.0978.0 6.8 6.8 As-builtAs-built SLM SLM 594 594 862.0 862.0 25.1 25.1 HIP + STA 1031 1093 0.9 HIP + STA 1031 1093 0.9 SLM + STA 1042 1142 10.1 WroughtSLM + STA + STA 1028 1042 1139 1142 10.1 10.5 Metals 2017, 7, 367 Wrought + STA 1028 1139 10.5 8 of 13

Figure 10. TEM bright field micrograph of strengthening phase precipitate within (a) HIP + STA and Figure 10. TEM bright ﬁeld micrograph of strengthening phase precipitate within (a) HIP + STA and (b) SLM + STA specimens and (c) high magnification of area I in (b). (b) SLM + STA specimens and (c) high magniﬁcation of area I in (b).

Figure 11. Fracture surfaces of (a) HIP + STA specimen and (b) SLM + STA specimen tested at room temperature; fracture surfaces of (c) HIP + STA specimen and (d) SLM + STA specimen tested at 650 °C.

3.4. Creep Properties The creep curves of heat-treated specimens at 650 °C and 550 MPa are shown in Figure 12. As for the SLM + STA specimen, the creep rupture life was approximately 140 h, while the HIP + STA specimen exhibited a rupture life of only 20 h. The fracture surface was observed along the grain boundaries covered with PPB in the HIP + STA specimen as shown in Figure 13a.

Metals 2017, 7, 367 8 of 13

Figure 10. TEM bright field micrograph of strengthening phase precipitate within (a) HIP + STA and Metals 2017, 7, 367 8 of 13 (b) SLM + STA specimens and (c) high magnification of area I in (b).

FigureFigure 11. Fracture 11. Fracture surfaces surfaces of of (a ()a HIP) HIP + + STASTA specimenspecimen and and (b ()b SLM) SLM + STA + STA specimen specimen tested tested at room at room ◦ temperature;temperature; fracture fracture surfaces surfaces of ( cof) HIP(c) HIP + STA + STA specimen specimen and and (d) ( SLMd) SLM + STA + STA specimen specimen tested tested at 650at C. 650 °C. 3.4. Creep Properties 3.4. Creep Properties The creep curves of heat-treated specimens at 650 ◦C and 550 MPa are shown in Figure 12. As for The creep curves of heat-treated specimens at 650 °C and 550 MPa are shown in Figure 12. As the SLM + STA specimen, the creep rupture life was approximately 140 h, while the HIP + STA for the SLM + STA specimen, the creep rupture life was approximately 140 h, while the HIP + STA specimen exhibited a rupture life of only 20 h. The fracture surface was observed along the grain specimen exhibited a rupture life of only 20 h. The fracture surface was observed along the grain Metals 2017, 7, 367 9 of 13 boundariesboundaries covered covered with with PPB PPB in in the the HIP HIP + + STASTA specimen as as shown shown in inFigure Figure 13a. 13 a.

Metals 2017, 7, 367 9 of 13

◦ FigureFigure 12.Figure Creep 12. 12.Creep Creep curves curves curves of ofHIP of HIP HIP + + STA+ STASTA specimen specimen and and SLM SLMSLM + STA + + STA STA at 650 at at 650 °C 650 andC °C and550 and MPa. 550 550 MPa. MPa.

FigureFigure 13. Fracture13. Fracture surfaces surfaces of of (a ()a HIP) HIP + + STASTA specimen and and (d ()d SLM) SLM + STA + STA specimen specimen tested tested at 650 at °C 650 ◦C and 550and MPa.550 MPa. Figure4. Discussion 13. Fracture surfaces of (a) HIP + STA specimen and (d) SLM + STA specimen tested at 650 °C and 550 MPa. 4.1. Grain Morphologies of SLM Materials 4. DiscussionEquiaxed grains with an average size of 8.8 μm were observed in the HIP specimens (Figure 6a), whereas columnar grains were observed in the SLM specimens (Figure 6b). The columnar grains were attributed to the epitaxial growth which resulted from the partial melting of the back of the 4.1. Grainprevious Morphologies layer to of promote SLM Materials growth from the prior grains of the previous layer in the [001] direction Equiaxed[25]. In grainsaddition, with fine anand average equiaxed size grains of with 8.8 μa mgrain were size observed of less than in 1 μthem wereHIP alsospecimens observed (Figurein 6a), the SLM specimens (Figure 6b). Fine grains can be formed by the constitutional supercooling whereas columnar grains were observed in the SLM specimens (Figure 6b). The columnar grains mechanism [26]. were attributed to the epitaxial growth which resulted from the partial melting of the back of the previous 4.2.layer The toFormation promote of PPB growth in HIP Materialsfrom the prior grains of the previous layer in the [001] direction [25]. In addition,The presence fine and of equiaxedelements with grains high withoxygen a affinigrainties size in superalloys of less than lead 1s μtom the were production also observed of in the SLMhighly specimens stable oxides (Figure at the 6b).powder Fine surface, grains or so-c canalled be PPB, formed in the consolidated by the constitutional products. The effects supercooling mechanismof PPBs [26]. are known to have detrimental effects on the mechanical properties of HIP materials. If the amount of oxygen is limited, the formation of Al2O3 and TiO2 is expected to occur first, before the other elements are oxidized, due to the order of Gibbs energies for the oxides [27,28] (see Figure 14). These 4.2. The Formation of PPB in HIP Materials oxides would promote the subsequent precipitation of MC-type carbides or M23C6-type carbides The alongpresence the PPB of (Figureelements 7). These with oxides high actoxygen as hetero affinigeneousties innuclei superalloys sites, which lead haves ato strong the productioneffect of on the nucleation and growth behavior of the carbide. During the heat-treatment process, carbide- highly stableforming oxides elements at the within powder the powdersurface, would or so-c migratealled PPB,to the in specific the consolidated oxides of particle products. surfaces, The effects of PPBs areresulting known in stable to have oxy-carbides detrimental on the effectsPPB [24]. on the mechanical properties of HIP materials. If the amount of oxygen is limited, the formation of Al2O3 and TiO2 is expected to occur first, before the other elements are oxidized, due to the order of Gibbs energies for the oxides [27,28] (see Figure 14). These oxides would promote the subsequent precipitation of MC-type carbides or M23C6-type carbides along the PPB (Figure 7). These oxides act as heterogeneous nuclei sites, which have a strong effect on the nucleation and growth behavior of the carbide. During the heat-treatment process, carbide- forming elements within the powder would migrate to the specific oxides of particle surfaces, resulting in stable oxy-carbides on the PPB [24].

Metals 2017, 7, 367 9 of 13

4. Discussion

4.1. Grain Morphologies of SLM Materials Equiaxed grains with an average size of 8.8 µm were observed in the HIP specimens (Figure6a), whereas columnar grains were observed in the SLM specimens (Figure6b). The columnar grains were attributed to the epitaxial growth which resulted from the partial melting of the back of the previous layer to promote growth from the prior grains of the previous layer in the [001] direction [25]. In addition, ﬁne and equiaxed grains with a grain size of less than 1 µm were also observed in the SLM specimens (Figure6b). Fine grains can be formed by the constitutional supercooling mechanism [26].

4.2. The Formation of PPB in HIP Materials The presence of elements with high oxygen affinities in superalloys leads to the production of highly stable oxides at the powder surface, or so-called PPB, in the consolidated products. The effects of PPBs are known to have detrimental effects on the mechanical properties of HIP materials. If the amount of oxygen is limited, the formation of Al2O3 and TiO2 is expected to occur first, before the other elements are oxidized, due to the order of Gibbs energies for the oxides [27,28] (see Figure 14). These oxides would promote the subsequent precipitation of MC-type carbides or M23C6-type carbides along the PPB (Figure7). These oxides act as heterogeneous nuclei sites, which have a strong effect on the nucleation and growth behavior of the carbide. During the heat-treatment process, carbide-forming elements within the powder would migrate to the specific oxides of particle surfaces, resulting in Metals 2017, 7, 367 10 of 13 stable oxy-carbides on the PPB [24].

Figure 14. Illustration of equilibrium oxygen partial pressurepressure for somesome metalmetal oxides/metaloxides/metal systems, systems, Reproducedreproduced withwith permissionpermission fromfrom [[28].28]. Elsevier,Elsevier, 2012.2012.

4.3. The The Effects Effects of of PPB PPB on the Mechanical Properties of HIP Specimens at 650 °C◦C The strengths and ductility of the SLM + STA STA specimen were comparable to to those of the conventionally wrought specimen,specimen, bothboth atat roomroom temperature temperature and and 650 650◦ C,°C, as as shown shown in in Tables Table2 and2 and3. TableHowever, 3. However, a drop was a drop observed was observed in the ductility in the ductility of HIP + of STA HIP material + STA material at 650 ◦C at (Table 650 °C3). (Table The fracture 3). The fracturesurfaces surfaces appeared appeared completely completely along the along PPB (Figure the PPB 11 (Figurec). The presence11c). The of presence a PPB decorated of a PPBwith decorated brittle withoxide brittle and carbide oxide and particles carbide resulted particles in premature resulted in failure premature [11,29]. failure The degree [11,29]. of ductilityThe degree of aof material ductility is ofrelated a material to the is relaxation related to process the relaxation that occurs process at the that crack-tip. occurs at When the crack-tip. brittle non-metallic When brittle particles non-metallic which particlestend to lead which to tend interface to lead decohesion to interface are decohesion dispersed are along dispersed the boundaries along the of boundaries the matrix of content the matrix [30], contentsuch as [30], carbides such as or carbides oxides along or oxides the PPBalong in the HIP PPB materials, in HIP materials, the process the pr ofocess crack of nucleation crack nucleation at the atgrain the boundarygrain boundary takes place takes more place easily more at easily 650 ◦ C.at Due650 °C. tothe Due interface to the interface between incoherentbetween incoherent particles, particles,the matrix the becomes matrix a preferredbecomes sitea preferred for crack initiationsite for crack [23]. Theinitiation crack would[23]. The begin crack to propagate would begin rapidly to propagate rapidly when the stress reaches a critical value. The presence and specific location of the precipitates along the PPB weaken the metallic bonding between two adjacent powders, which leads to worse mechanical properties [31], contributing to the predominance of fractures along the PPB in HIP materials at 650 °C (Figure 11). Interestingly, the HIP + STA material strengthened by precipitation showed only one-seventh of the ductility shown by the as-HIPed specimen at 650 °C (Table 3). The drop in the ductility of the HIP + STA material was observed under both the tensile test and creep test at 650 °C. As the temperature increases above 750 °C, the coherent relationship between γ and γ″ could be lost and be replaced by the incoherent δ phase with the same Ni3Nb composition [16]. The γ″ phase will be metastable, and the conversion reaction of γ/γ″ → δ accelerates at temperatures above 750 °C [32]. These needle- shaped and brittle δ phases are generally undesirable due to their adverse effects on mechanical properties [33], which lead to “δ-phase embrittlement” [16,34] (Figure 12). Furthermore, the previous oxides along the PPB would promote the subsequent precipitation of carbides along the PPB (Figure 7). During the heat-treatment process, carbide-forming elements within the powder would migrate to the specific oxides of particle surfaces, resulting in stable oxy-carbides on the PPB [24]. The carbides would grow during the heat-treatment process, which includes enough diffusion time for the required carbide-forming elements to be obtained (Figures 7a,b). These script-like carbides or needle- shaped δ phases generally would have a negative effect on alloy performance. Since the oxides, carbides and needle-shaped δ phases are closely spaced along the PPB (Figure 7a), they provide a continuous path for crack propagation. At temperatures above 980 °C, Al2O3 forms as a protective oxide scale [35]. After one hour of isothermal oxidation in CM-247LC, continuous Al2O3 formed at 1150 °C, whereas, NiO formed at 1000 °C [36]. This means that the HIP temperature of 1180 °C would be suitable for forming Al2O3 in Ni-based superalloys. The main reason for this is that PPB precipitate

Metals 2017, 7, 367 10 of 13 when the stress reaches a critical value. The presence and speciﬁc location of the precipitates along the PPB weaken the metallic bonding between two adjacent powders, which leads to worse mechanical properties [31], contributing to the predominance of fractures along the PPB in HIP materials at 650 ◦C (Figure 11). Interestingly, the HIP + STA material strengthened by precipitation showed only one-seventh of the ductility shown by the as-HIPed specimen at 650 ◦C (Table3). The drop in the ductility of the HIP + STA material was observed under both the tensile test and creep test at 650 ◦C. As the temperature increases above 750 ◦C, the coherent relationship between γ and γ00 could be lost and 00 be replaced by the incoherent δ phase with the same Ni3Nb composition [16]. The γ phase will be metastable, and the conversion reaction of γ/γ00 → δ accelerates at temperatures above 750 ◦C[32]. These needle-shaped and brittle δ phases are generally undesirable due to their adverse effects on mechanical properties [33], which lead to “δ-phase embrittlement” [16,34] (Figure 12). Furthermore, the previous oxides along the PPB would promote the subsequent precipitation of carbides along the PPB (Figure7). During the heat-treatment process, carbide-forming elements within the powder would migrate to the speciﬁc oxides of particle surfaces, resulting in stable oxy-carbides on the PPB [24]. The carbides would grow during the heat-treatment process, which includes enough diffusion time for the required carbide-forming elements to be obtained (Figure7a,b). These script-like carbides or needle-shaped δ phases generally would have a negative effect on alloy performance. Since the oxides, carbides and needle-shaped δ phases are closely spaced along the PPB (Figure7a), they provide a ◦ continuous path for crack propagation. At temperatures above 980 C, Al2O3 forms as a protective oxide scale [35]. After one hour of isothermal oxidation in CM-247LC, continuous Al2O3 formed at 1150 ◦C, whereas, NiO formed at 1000 ◦C[36]. This means that the HIP temperature of 1180 ◦C would be suitable for forming Al2O3 in Ni-based superalloys. The main reason for this is that PPB precipitate networks formed during HIPing would be maintained at the HIP temperature for a long time (Figure3a). As a result, the HIP + STA material would exhibit premature fracturing with poor ductility at 650 ◦C.

4.4. Free-PPB Microstructure in SLM Materials Precipitates were localized along the PPB in both the as-HIPed and HIP + STA specimens (Figure1), whereas SLM materials showed a free-PPB microstructure despite having a comparable oxygen content (Table1). The difference in microstructure can be explained as follows. The HIP process is a solid-state diffusion process occurring below the material’s melting temperature. The presence in superalloys of elements with high oxygen affinities brings about highly stable oxides at the powder surface. These oxides along the PPB would become diffusion barriers, resulting in a slower diffusion process [37]. As for the SLM process, the laser beam irradiates metallic powder in order to form a layer-by-layer material in a chamber with an inert gas atmosphere. The main reason that limited oxides are observed would be that the precipitates along the PPB precipitates network are churned and spread uniformly. Another layer of metallic powder is then fused and deposited, and the previous layer absorbs the extra power, resulting in the remelting of a part of the previous layer [26]. This remelting process is analogous to dilution, which provides a continuous solid–liquid interface free of contaminants or oxides and a sufficient bonding between layers [26,38,39]. Furthermore, the evaporation of Al and the ejection of pure-Al-rich particles with only 0.4 wt % oxygen were observed during the SLM process [40]. These results show that Al is easy to vaporize, and that Al2O3 does not easily form during the SLM process because the time exposed to high temperatures above 1150 ◦C is limited (Figure3b). From the view point of thermodynamics, Figure 14 shows the equilibrium oxygen partial pressure for some metal oxides/metal systems [28]. Figure 14 clearly indicates that every oxide could be avoided when the optimal temperature and oxygen partial pressure are reached. Under the conditions below each curve in the figure (low oxygen partial pressure and high temperature), the oxide is decomposed. On the other hand, above each curve (i.e., under high oxygen partial pressure and low temperature conditions) the metal is easily oxidized [28]. In the current study, since the temperature was maintained Metals 2017, 7, 367 11 of 13 in the HIP process (1180 ◦C for 4 h), a certain degree of oxidation cannot be avoided under normal HIP conditions. HIP was performed in a chamber with a vacuum which contained contaminants in the form of water and oxygen. As a result, metals could be easily oxidized during the HIP process (Figures3a and 14). On the other hand, the minimum temperature in a high-energy laser-beam-focused zone ranges from 1260 to 1336 ◦C (melting point of Alloy718) during the SLM process [35]. As a result of the decomposition of contaminant compounds, such compounds might appear at low density in SLM materials (Figures1 and5).

5. Conclusions The following conclusions can be drawn from this work:

(1) Continuous precipitates were clearly localized along the PPB in HIP materials, while SLM materials showed a microstructure free of PPBs and an extremely low number of contaminants. (2) The columnar grains in the SLM specimens were attributed to epitaxial growth. (3) The strengths and ductilities of the SLM + STA specimen were comparable to those of the conventionally wrought specimen both at room temperature and 650 ◦C. However, a drop was observed in the ductility of the HIP material at 650 ◦C. (4) The brittle precipitates formed in the HIP materials, such as oxy-carbides and δ phases, tended to lead to interface decohesion and resulted in premature fracture with poor ductility at 650 ◦C. (5) The main reason that PPB precipitate networks formed during HIPing is that the HIP temperature was held at a temperature suitable for forming Al2O3 for a long time. (6) In the SLM process, Al2O3, which became a nucleation site of precipitates, is not easily formed, and the PPB precipitates network is churned and spread uniformly during the process.

Acknowledgments: This research was supported by the ALCA Program of the Japan Science and Technology Agency, JST, and a Grant-in-Aid for Scientific Research (16K06799) from the Japan Society for the Promotion of Science. This support is gratefully acknowledged. Author Contributions: Yen-Ling Kuo performed the experiments, analyzed the data and wrote the paper; Koji Kakehi designed the experiments and wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Yeh, A.C.; Lu, K.W.; Kuo, C.M.; Bor, H.Y.; Wei, C.N. Effect of serrated grain boundaries on the creep property of Inconel 718 superalloy. Mater. Sci. Eng. A 2011, 530, 525–529. [CrossRef] 2. Chamanfar, A.; Sarrat, L.; Jahazi, M.; Asadi, M.; Weck, A.; Koul, A.K. Microstructural characteristics of forged and heat treated Inconel-718 disks. Mater. Des. 2013, 52, 791–800. [CrossRef] 3. Ghosh, S.; Yadav, S.; Das, G. Study of standard heat treatment on mechanical properties of Inconel 718 using ball indentation technique. Mater. Lett. 2008, 62, 2619–2622. [CrossRef] 4. Bai, Q.; Lin, J.; Tian, G.; Dean, T.A.J. Review and analysis of powder prior boundary (PPB) formation in powder metallurgy processes for nickel-based super alloys. J. Powder Metall. Min. 2015, 4.[CrossRef] 5. Lee, S.C.; Chang, S.H.; Tang, T.P.; Ho, H.H.; Chen, J.K. Improvement in the microstructure and tensile properties of Inconel 718 superalloy by HIP treatment. Mater. Trans. 2006, 47, 2877–2881. [CrossRef] 6. Rao, G.A.; Kumar, M. High performance stainless steel via powder metallurgy hot isostatic pressing. Mater. Sci. Technol. 1997, 13, 1027–1031. [CrossRef] 7. Gessinger, G.H. Powder Metallurgy of Superalloys; Butteriworth & Co.: London, UK, 1984. 8. Maurer, G.E.; Castledine, W.; Schweizer, F.A.; Mancuso, S. Development of HIP consolidated P/M superalloys for conventional forging to gas turbine engine components. In Proceedings of the 8th International Symposium on Superalloys, Pittsburgh, PA, USA, 28 September–1 October 2014; pp. 645–652. 9. Furrer, D.; Fecht, H. Ni-based superalloys for turbine discs. JOM 1999, 51, 14–17. [CrossRef] 10. Rao, G.A.; Srinivas, M.; Sarma, D.S. Effect of solution treatment temperature on microstructure and mechanical properties of hot isostatically pressed superalloy Inconel* 718. Mater. Sci. Technol. 2004, 20, 1161–1170. [CrossRef] Metals 2017, 7, 367 12 of 13

11. Bai, Q.; Lin, J.; Jiang, J.; Dean, T.A.; Zou, J.; Tian, G. A study of direct forging process for powder superalloys. Mater. Sci. Eng. A 2015, 621, 68–75. [CrossRef] 12. Haase, C.; Bültmann, J.; Hof, J.; Ziegler, S.; Bremen, S.; Hinke, C.; Schwedt, A.; Prahl, U.; Bleck, W. Exploiting process-related advantages of selective laser melting for the production of high-manganese steel. Materials 2017, 10, 56. [CrossRef][PubMed] 13. Ellyson, B.; Brochu, M.; Brochu, M. Characterization of bending vibration fatigue of SLM fabricated Ti-6Al-4V. Int. J. Fatigue 2011, 19, 389–395. [CrossRef] 14. AlMangour, B.; Grzesiakc, D.; Yang, J.M. Selective laser melting of TiB2/H13 steel nanocomposites: Influence of hot isostatic pressing post-treatment. J. Mater. Process. Technol. 2017, 244, 344–353. [CrossRef] 15. Kuo, Y.L.; Horikawa, S.; Kakehi, K. The effect of interdendritic δ phase on the mechanical properties of Alloy 718 built up by additive manufacturing. Mater. Des. 2017, 116, 411–418. [CrossRef] 16. Kuo, Y.L.; Horikawa, S.; Kakehi, K. Effects of build direction and heat treatment on creep properties of Ni-base superalloy built up by additive manufacturing. Scr. Mater. 2017, 129, 74–78. [CrossRef] 17. Donachie, M.J.; Donachie, S.J. Superalloys: A Technical Guide, 2nd ed.; ASM International: Geauga County, OH, USA, 2002. 18. Zhang, J.; Singer, R.F. Hot tearing of nickel-based superalloys during directional solidification. Acta Mater. 2002, 50, 1869–1879. [CrossRef] 19. Chang, S.; Lee, S.; Tang, T.; Ho, H. Influences of soaking time in hot isostatic pressing on strength of inconel 718 superalloy. Mater. Trans. 2006, 47, 426–432. [CrossRef] 20. Jia, Q.; Gu, D. Selective laser melting additive manufacturing of Inconel 718 superalloy parts: Densification, microstructure and properties. J. Alloys Compd. 2014, 585, 713–721. [CrossRef] 21. Zhang, B.; Lee, X.; Bai, J.; Guo, J.; Wang, P.; Sun, C.; Nai, M.; Qi, G.; Wei, J. Study of selective laser melting (SLM) Inconel 718 part surface improvement by electrochemical polishing. Mater. Des. 2017, 116, 531–537. [CrossRef] 22. Kanagarajah, P.; Brenne, F.; Niendorf, T.; Maier, H.J. Inconel939 processed by selective laser melting: Effect of microstructure and temperature on the mechanical properties under static and cyclic loading. Mater. Sci. Eng. A 2013, 588, 188–195. [CrossRef] 23. Rao, G.A.; Kumar, M.; Srinivas, M.; Sarma, D.S. Effect of standard heat treatment on the microstructure and mechanical properties of hot isostatically pressed superalloy inconel 718. Mater. Sci. Eng. A 2003, A355, 114–125. [CrossRef] 24. Crompton, J.S.; Hertzberg, R.W. Analysis of second phase particles in a powder metallurgy HIP nickel-base superalloy. J. Mater. Sci. 1986, 21, 3445–3454. [CrossRef] 25. Olakanmi, E.O.; Cochrane, R.F.; Dalgarno, K.W. A review on selective laser sintering/melting (SLS/SLM) of aluminium alloy powders: Processing, microstructure, and properties. Prog. Mater. Sci. 2015, 74, 401–477. [CrossRef] 26. DuPont, J.N.; Lippold, J.C.; Kiser, S.D. Welding Metallurgy and Weldability of Nickel Base Alloys; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2009. 27. He, L.M.; Su, Y.-F.; Allard, L.F.; Lance, M.J.; Lee, W.Y. Effects of preoxidation on the nucleation and growth behavior of chemically vapor-deposited α-Al2O3 on a single-crystal Ni-based superalloy. Metall. Trans. A 2004, 35, 1113–1124. [CrossRef] 28. Hryha, E.; Dudrova, E.; Nyborg, L. On-line control of processing atmospheres for proper sintering of oxidation-sensitive PM steels. J. Mater. Process. Technol. 2012, 212, 977–987. [CrossRef] 29. Wallace, W.; Wiebe, W.; Whelan, W.P.; Dainty, R.V.; Terada, T. The effect of grain-boundary structure on the tensile fracture behavior of hot-isostatically pressed 713 LC Alloy Compacts. Powder Metall. 1972, 16, 416–435. [CrossRef] 30. Tetelman, A.S.; Mcevily, A.J., Jr. Fracture of Structural Materials; John Wiley & Sons Inc.: New York, NY, USA, 1967; pp. 193–233. 31. Broek, D. Elementary Engineering Fracture Mechanics; Martinus Nijhoff: Leiden, The Netherlands, 1984. 32. Idell, Y.; Levine, L.E.; Allen, A.J.; Zhang, F.; Campbell, C.E.; Olson, G.B.; Gong, J.; Snyder, D.R.; Deutchman, H.Z. Unexpected δ-phase formation in additive-manufactured Ni-based superalloy. JOM 2016, 68, 950–959. [CrossRef] 33. Campo, E.; Turco, C.; Catena, V. The correlation between heat treatment, structure and mechanical characteristics in Inconel 718. Metall. Sci. Technol. 1985, 3, 16–21. Metals 2017, 7, 367 13 of 13

34. Radavich, J.F.; Korth, G.E.; Antolovich, S.D.; Stusrud, R.W.; MacKay, R.A.; Khan, T.; Kissinger, R.D.; Klarstrom, D.L. High temperature degradation of Alloy 718 after longtime exposures. In Proceedings of the 7th International Symposium on Superalloys, Seven Springs, PA, USA, 20–24 September 1992; pp. 497–506. 35. Davis, J.R. ASM Specialty Handbook: Heat-Resistant Materials; ASM International: Geauga County, OH, USA, 1997. 36. Chiou, M.S.; Jian, S.R.; Yeh, A.C.; Kuo, C.M. Effects of aluminum addition on the high temperature oxidation behavior of CM-247LC Ni-based superalloy. Int. J. Electrochem. Sci. 2015, 10, 5981–5993. 37. German, R. Powder Metallurgy and Particulate Materials Processing: The Processes, Materials, Products, Properties and Applications; Metal Powder Industries Federation: Princeton, NJ, USA, 2005. 38. Abe, F.; Osakada, K.; Shiomi, M.; Matsumoto, M.; Shiomi, M. The manufacturing of hard tools from metallic powders by selective laser melting. J. Mater. Process. Technol. 2001, 111, 210–213. [CrossRef] 39. Das, S.; Beaman, J.J. Direct Selective Laser Sintering of Metals. U.S. Patent 6,676,892 B2, 13 January 2004. 40. Wang, X.; Read, N.; Carter, L.N.; Ward, R.M.; Attallah, M.M. Defect formation and mitigation in selective laser melting of high γ0 Ni-base superalloys. In Proceedings of the 13th International Symposium on Superalloys, Seven Springs, PA, USA, 11–15 September 2016; pp. 351–358.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).