J. Laser Appl. 31, 031201 (2019); https://doi.org/10.2351/1.5085206 31, 031201

© 2019 Laser Institute of America.

Microstructure evolution during selectivelaser melting of metallic materials: A reviewCite as: J. Laser Appl. 31, 031201 (2019); https://doi.org/10.2351/1.5085206Submitted: 10 December 2018 . Accepted: 13 May 2019 . Published Online: 31 May 2019

Xing Zhang , Christopher J. Yocom, Bo Mao , and Yiliang Liao

Microstructure evolution during selective lasermelting of metallic materials: A review

Cite as: J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206

View Online Export Citation CrossMarkSubmitted: 10 December 2018 · Accepted: 13 May 2019 ·Published Online: 31 May 2019

Xing Zhang, Christopher J. Yocom, Bo Mao, and Yiliang Liaoa)

AFFILIATIONS

Department of Mechanical Engineering, University of Nevada, Reno, Nevada 89557

a)Author to whom correspondence should be addressed; electronic mail: yliao@unr.edu

ABSTRACT

Selective laser melting (SLM) is an additive manufacturing technology that uses a laser beam to melt powder materials together layerby layer for solid part fabrication. Due to its superior rapid prototyping capability of three-dimensional structures, SLM has beenused for widespread industrial applications including aerospace, automotive, electronics, and biomedical devices. As a state-of-the-arttechnology, ongoing investigations are being conducted to improve the efficiency and effectiveness of SLM. In particular, understand-ing of microstructure evolution during SLM is essential to achieve improved process control and ensure the performance of laser-fab-ricated components. This paper is to review the recent research and development progress in SLM of metallic materials with a focuson the process–microstructure relationship. The grain growth and porosity evolution as affected by laser processing parameters in theSLM process are discussed. Phase transformation in SLM of steel and titanium alloys is studied. The formation of precipitates inSLM of titanium, nickel, and aluminum/magnesium alloys is reviewed. The balling phenomenon and cracking behaviors during SLMare discussed. In addition, the recent development of computational modeling of microstructure evolution during SLM isinvestigated.

Key words: selective laser melting, microstructure evolution, grain growth, porosity, computational modeling

© 2019 Laser Institute of America. https://doi.org/10.2351/1.5085206

I. INTRODUCTION

Selective laser melting (SLM) is an additive manufacturingtechnology that uses a laser beam to melt powder materialstogether layer by layer for three-dimensional (3D) solid partfabrication. At the beginning of the process, a 3D computer-aided design model describing both internal and externalgeometry features of the part is developed and sliced into astack of 2D layers, which are loaded into the SLM equipment.A thin layer of powder material is then spread on a substrateplate via a powder delivery system and later melted togetherwith a high-energy laser beam scanning over the selected area.By applying another thin layer of powder on top of the previ-ously processed layer and repeating the scanning process, thesuccessive layers are formed and metallurgically bonded withthe previous layer. The designed part is, thus, fabricated layerby layer as shown in Fig. 1.1 Compared to conventional manu-facturing techniques such as casting and forging, SLM holdsadvantages including (1) rapid prototyping, (2) the capability

of producing components with customized or complexgeometry, (3) the ability to produce novel structural designs forincreased functionality, and (4) the joining of dissimilar metalsthrough metallurgical means.

Due to these advantages, SLM has been used for widespreadpractical industrial applications, such as aerospace, automotive,electronics, and biomedical devices. Yakout et al.2 produced anairfoil blade of stainless steel 316L using SLM. Song et al.3 manu-factured the metal die of an automobile deck part at a low cost anda short process cycle. Wits et al.4 utilized SLM to fabricate a low-weight and high-pressure micropump for small satellite applica-tions. Hou et al.,5 Wong et al.,6 and Zenou et al.7 employed thistechnology to manufacture electronics such as printed circuitboards, heat sinks, and transparent conductors. SLM has also beenapplied in the medical field8–11 to manufacture customized bioim-plants. The most recent application of SLM is to fabricate novelmaterials such as functionally graded materials12–14 and metalmatrix nanocomposites.15–17

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-1

© 2019 Laser Institute of America

However, SLM is a complicated process that suffers from theintense laser–material interaction, molten pool instability, largedegree of shrinkage, and complex residual stress.18 The propertiesand performance of a component fabricated via SLM are closelyrelated to microstructure characteristics and determined by variousconsiderations. From the perspective of metallurgical bonding,specific considerations should be given to the wetting behavior,19

Marangoni effect,20 and capillary forces.21 In addition, rapidthermal cycles and directional solidification can generate micro-structure that differs from conventionally processed materials.

More importantly, microstructure evolution is highly affected bylaser processing conditions such as laser parameters and scanningstrategy. Therefore, precision control of the microstructure evolu-tion during SLM remains a major challenge, which significantlyaffects the performance of as-built parts.

With the increased demand for SLM methods inmanufacturing,19,22–26 it is of primary importance to gain a betterunderstanding of the process–microstructure relationship in order toachieve improved process control and ensure the performance oflaser-fabricated components (Fig. 2). This paper is to review therecent research and development progress in SLM of metallic materi-als with a focus on the process–microstructure relationship. The evo-lution of grain and pore structures are studied by reviewingpreviously conducted experiments, along with the effects of variousprocess parameters. Phase transformation in SLM of steel and tita-nium alloys is studied. The formation of precipitates in SLM of tita-nium, nickel, and aluminum/magnesium (Al/Mg) alloys is reviewed.The balling phenomenon and cracking behaviors during SLM arediscussed. In addition, computational modeling methodologies forpredicting microstructure evolution during SLM are reviewed with afocus on the growth of the grain structure. The knowledgegained from this review paper could provide insights and guidancefor optimization and further development of the SLM process.

II. GRAIN GROWTH IN THE SLM PROCESS

A. Mechanism of grain growth

During SLM of metallic materials, the evolution of the grainstructure in terms of grain size, morphology, and orientation isconsidered as the key microstructure feature determining themechanical performance of 3D-printed parts. Similar to traditionalwelding processes, the grain growth during SLM is a localized sol-idification process that is initiated by epitaxial growth proceeded bycompetitive growth toward the center of the melt pool. As the

FIG. 1. Schematic of the SLM setup. Adapted with permission from K.-H. Leitz,P. Singer, A. Plankensteiner, B. Tabernig, H. Kestler, and L. Sigl, Met. PowderRep. 72, 331–338 (2017). Copyright 2017, Elsevier.

FIG. 2. Correlation among the SLMprocess parameters, microstructure, andas-built part properties.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-2

© 2019 Laser Institute of America

traveling laser beam interacts with the material, a melt pool forms.After the laser beam leaves the area, solidification takes place viaepitaxial growth of the partially melted grains from the substrate orprevious layer. The shape of the liquid–solid interface in terms ofplanar, cellular, columnar dendritic, or equiaxed dendritic fronts isgoverned by the temperature gradient G, solidification rate R,undercooling ΔT, and solute diffusion coefficient DL (Fig. 3).27–30

According to the solidification theory, the grain morphology andsubstructure (fineness/size) are strongly affected by the ratios G/Rand G⋅R (cooling rate), respectively.19 At the bottom of the meltpool where the solidification rate R≈ 0, the ratio of G/R is infinite,leading to a planar-dominated grain structure. As the ratio, G/Rdecreases until G/R < ΔT/DL, the cellular solidification takes placenear the planar region. Further decreasing of the G/R ratio leads tothe columnar dendritic formation. The equiaxed solidification typi-cally happens near the surface of the melt pool where the tempera-ture gradient is relatively low.

In addition to the grain morphology, the grain growthdirections are aligned in the thermal gradient direction as the solid-ification proceeds. As shown in Fig. 4, at the early stage of solidifi-cation, the epitaxial grains have multiple orientations. However, thecrystallographic effect affects the grain structure by favoring growthalong particular crystallographic orientations.31,32 For instance, themetallic materials with the face-centered cubic structure have a pre-ferred growth along the h100i crystallographic orientations. Thus,the grains with the same or close orientation as the temperaturegradient grow faster, while the grains with other orientations gradu-ally stop growing due to the competitive growth of differentgrains.33 In addition, the horizontal grain growth is suppressedwhen the adjacent grains impinge on each other, and the high rate

of elongated grain growth results in little chance of nucleationoccurring in the liquid ahead of the columnar zone.34 As a result,this competitive growth phenomenon leads to the unidirectionalcolumnar grain structure during solidification, which is a typicalmicrostructure observed in SLM-built products. Nevertheless, byadjusting laser parameters and/or applying methods to promotenucleation (such as adding impurities like nanomaterials and utiliz-ing dynamic approaches like arc oscillation), the equiaxed dendriticgrowth can take place due to the massive increase of multiorienta-tion nucleation sites, resulting in finer grain structures.35–43

B. Effects of laser parameters and scanning strategieson grain growth

In order to investigate the effect of laser parameters on micro-structure evolution during SLM, researchers usually consider acombined parameter—energy density, which can be expressed as

E ¼ P=(v � h� t), (1)

where P is the laser power, v is the scanning speed, h is the hatchspacing, and t is the layer thickness. These laser parameters playimportant roles in determining the temperature gradient, coolingrate, and local heat flow direction, which in turn considerably affectthe grain morphology, size, and orientation. Monroy et al.44

studied the impacts of laser power and scanning speed on the grainsize during SLM of cobalt alloys, and Sun et al.45 conducted similarinvestigation on stainless steel 316L. It was found that the grain sizeincreased with increasing laser power due to a higherenergy-induced coarsening effect, and the grain size decreased withincreasing scanning speed due to shorter exposure time and fastersolidification. Yadroitsev et al.46 characterized the grain growth ofstainless steel 316L during SLM in terms of primary cell spacingand misorientation angle, which are defined as the distancebetween adjacent grains and the angle between the laser travelingdirection and the grain growth direction, respectively. It wasreported that the cell spacing decreased by 25%–40% as the scan-ning speed increased from 0.08 to 0.28 m/s. In addition, tworegions with different misorientation angles (58°–69° and 10°) were

FIG. 3. Schematic of the effects of temperature gradient G and solidificationrate R on grain morphology. Adapted with permission from J. Dupont,Fundamentals of Weld Solidification (ASM International, Materials Park, OH,2011). Copyright 2011, ASM International.

FIG. 4. Schematic of grain growth in SLM including initial epitaxial growth andsubsequent competitive growth. Adapted with permission from S. Kou, WeldingMetallurgy (Wiley, New York, 2003). Copyright 2003, John Wiley and Sons.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-3

© 2019 Laser Institute of America

found while processing with a low scanning speed. On the otherhand, increasing the scanning speed resulted in only one misorien-tation angle of 58°–69°. Dezfoli et al.34 obtained similar conclusionsdescribing the influence of scanning speed during SLM ongrain size and grain orientation of Ti64 (Ti-6Al-4V). In addition,the effect of hatching spacing on grain morphology ofSLM-manufactured CoCrMo alloys was analyzed by Olsen et al.47

In the SLM process, small equiaxed grains are often found in thecenter of the melt tracks due to the constitutional undercooling.As the hatch spacing decreased, the new melt track would remeltthe center area, leading to large epitaxial grain growth acrossthe entire area.

Besides laser parameters, scanning strategies play another criti-cal role in determining the heat flow direction and, thus, the graingrowth direction. Generally, as illustrated in Fig. 5, four differentscanning strategies48 can be applied in the SLM process: unidirec-tional, zigzag, crosshatching, and “island” scanning strategies. Thijset al.49 investigated the grain growth during SLM of Ti64 asaffected by scanning strategies including unidirectional, zigzag, andcrosshatching. It was found that all methods produced identicalgrain growth directions. In specific, for the unidirectional scanningstrategy [as observed in the side view of Fig. 6(b)], the elongatedgrains tilted 19° away from the building direction. For the zigzagscanning, a grain structure with a herringbone pattern wasobserved [Fig. 6(c)] due to the alternating scanning direction, andelongated grains [Fig. 6(d)] were found to be generally tilted alongthe build direction. For the cross-hatching strategy, the grainsbecame more equiaxed, and a grain structure of the grid planpattern was observed [Fig. 6(e)] while two grain growth directions,parallel to the building direction and tilted at 25°, were identified[Fig. 6(f )]. Arisoy et al.50 conducted SLM experiments of nickelalloy IN625 using cross-hatching strategy with various rotationangles between consecutive layers. It was reported that compared tothe crosshatching with a rotation angle of 90°, the scanning strategywith a rotation angle of 67° resulted in a reduction of the averagegrain size, and the grain growth directions were less affectedby laser power and scanning speed. Carter et al.51 studied the

FIG. 5. Overview of different scanning strategies: (a) unidirectional, (b) zigzag,(c) crosshatching, and (d) island scanning pattern. Adapted with permissionfrom L. Thijs, K. Kempen, J.-P. Kruth, and J. Van Humbeeck, Acta Mater.61, 1809–1819 (2013). Copyright 2013, Elsevier.

FIG. 6. Effects of the scanning strate-gies on grain growth direction duringSLM of Ti64. Sample scanned withunidirectional strategy: (a) top view and(b) side view. Sample scanned withzigzag strategy: (c) top view and (d)side view. Sample scanned with cross-hatching strategy: (e) top view and(f ) side view. Adapted with permissionfrom L. Thijs, F. Verhaeghe, T. Craeghs,J. Van Humbeeck, and J.-P. Kruth, ActaMater. 58, 3303–3312 (2010). Copyright2010, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-4

© 2019 Laser Institute of America

influence of island scanning strategy on the grain structure ofSLM-fabricated nickel superalloy CM247LC. A bimodal grainstructure with large elongated columnar grains in the buildingdirection and fine grains in the edges of islands was observed. Itwas claimed that each scanning island was heated by the bandheating effect instead of point heating, resulting in the appearanceof fine grains to the higher cooling rate at the edges of islands dueto surrounding cold material.

C. Effects of additives on grain growth

The growth of the grain structure during SLM can be signifi-cantly affected by adding additives into the powder bed, such assecond phase nanoparticles. By providing a large amount of hetero-geneous nucleation sites in front of the liquid–solid interface, suit-able nanoparticles can considerably decrease the criticalundercooling needed for producing ideal equiaxed structures.52

Such effect of nanoparticles on microstructure evolution was firstlyobserved in other laser-based techniques such as laser cladding andlaser surface alloying. Li et al.35 compared the grain structure ofcobalt-based alloy coatings manufactured by laser cladding withand without nano-Y2O3. It was found that for the as-built partwithout nanoadditives, the directional columnar dendrites growingalong the temperature gradient direction can be easily observed atthe liquid–solid interface [Fig. 7(a)]. Contrarily, by mixing cobaltpowders with Y2O3 nanoparticles (0.5 wt. %), multiorientationequiaxed dendrites formed rather than columnar dendrites, asshown in Fig. 7(b). It was reported that such “columnar-to-equiaxed transition” became more evident by increasing nanoparti-cle quantities. For instance, the equiaxed grains can be observedacross the entire section by adding 1.0 wt. % of nano-Y2O3. Zhanget al.36 conducted similar experiments to study the effect of CeO2

microparticles/nanoparticles on the grain growth of Ni-based alloycoatings fabricated using laser cladding. It was found that a coatinglayer with a finer grain structure can be achieved by adding nano-additives, as compared to use microadditives. Later on, nanoparti-cles have been introduced to the SLM process to manufacturenanocomposites with excellent mechanical properties. Li et al.37

studied SLM of TiB2 nanoparticle-decorated AlSi10Mg alloys. Itwas concluded that by adding TiB2 nanoparticles, the obtainedaverage grain size (2 μm) was much smaller than that of theas-built sample without nanoparticles (∼10 μm). Recently, a high-strength aluminum alloy (AA) with a fine equiaxed grain structure

was 3D-printed using SLM, particularly through mixing AA micro-powders with hydrogen-stabilized zirconium nanoparticles.38

Ma et al.39 employed SLM to manufacture Ni/Al2O3 nanocompo-site, and it was proved that nanoparticles could also prevent thegrain coarsening within the heat-affected zone (substrate or previ-ous layer) by altering thermal conditions. However, it is also worthmentioning that some studies have shown that nanoparticles donot necessarily lead to grain refinement. For example, Streubelet al.53 found that the addition of nano-Y2O3 did not affect thegrain morphology of as-built steel part, which was attributed to thepoor wettability and lattice mismatching between the nanoparticleand metal matrix. Clearly, the role of nanoparticles in microstruc-ture evolution during SLM still needs to be further investigated.

D. Controlled grain growth by other strategies

Coupling SLM with dynamic approaches (such as electromag-netic vibration, ultrasonic vibration, transverse arc oscillation, etc.)might serve as an alternative methodology to tune the grainstructure of SLM-fabricated components. For instance, Huanget al.40 obtained Ti64 with finer grains by integrating SLM withelectromagnetic vibrations. It was claimed the electromagneticvibration-induced electromagnetic force effect, Peltier effect, andJoule heating effect may all contribute to the grain refinement. Todate, little effort has been put into the investigation of coupling SLMprocess with dynamic approaches. However, it is worth mentioningthat researchers have successfully conducted experiments on similarprocesses like laser surface melting and welding, which present apromising future for introducing various dynamic approaches toSLM. Zhou et al.41 examined the effect of coupling laser surfacemelting with an electromagnetic field (EMF) on the microstructureevolution of AZ91D magnesium alloys. As shown in Fig. 8, it wasobserved that without EMF, the melted layer mainly consisted ofcolumnar dendritic and cellular grains; while with EMF, the meltedlayer had a uniform microstructure with finer equiaxed grains. Xuet al.42 obtained fine grain structure through ultrasonic vibration-assisted laser surface melting. It was found that with the increase ofvibration amplitude, more dendrites were broken up and functionedas nuclei due to ultrasonic cavitation and acoustic streaming stirring.As the vibration amplitude approached 80%, the columnar crystalalmost disappeared. Yuan et al.43 conducted welding experimentswith transverse arc oscillation to refine the grain structure of AZ31Bmagnesium alloys. In this work, the effects of oscillation amplitude,

FIG. 7. Grain structure of cobalt-basedalloy coatings manufactured by SLM:(a) without nanoadditives and (b) with0.5 wt. % nano-Y2O3. Adapted withpermission from L. Mingxi, H. Yizhu,and Y. Xiaomin, Appl. Surf. Sci. 252,2882–2887 (2006). Copyright 2006,Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-5

© 2019 Laser Institute of America

frequency, and torch travel speed on grain refining were investigated.It was claimed that the grain refinement was attributed to the den-drite fragmentation induced by the remelting of dendrite arms ratherthan the mechanical breakup.

Due to the rapid solicitation process, directional columnardendrite is the typical grain morphology in SLM-built parts,leading to anisotropic mechanical properties.54,55 Recent researchindicates that altering the powder composition could serve as analternative method to obtain ultrafine grains and address the aniso-tropic effect in SLM. For example, Montero-Sistiaga et al.56 investi-gated the effects of Si on the grain size distribution during SLM ofAl7075, as shown in Fig. 9. It was found that the grain size wasreduced from 50–100 to 5 μm by adding 4 wt. % Si. Further investi-gations are needed to elucidate the mechanism responsible for thepowder composition effect on the grain growth in SLM.

III. EVOLUTION OF PORE STRUCTURES IN THE SLMPROCESS

In addition to the evolution of grains, the formation of porestructures during SLM needs to be investigated in order for achiev-ing 3D-printed parts with desired properties. Considerable researchand development effort has been put on SLM process design andoptimization to eliminate or minimize the porosity, since a highmaterial density ensures the mechanical integrity, which is criticalfor widespread industrial applications, particularly for those usedas load-bearing structural components. On the other hand, SLM ofporous materials attracts attention in recent years, since the porestructures with designed pore size and density are promising forapplications in thermal management and mass/energy transport.Therefore, it is of specific importance to gain an understanding ofthe mechanism responsible for the formation of pore structuresduring SLM of metallic materials.

A. Porosity-formation mechanism during SLM

Similar to conventional metal casting processes, the formationof pore structures in SLM is mainly attributed to the liquid–solidphase transformation during solidification, which is stronglyaffected by various factors including the thermal history (coolingrate), material intrinsic properties, part geometry, etc. In a studyconducted by Thijs et al.,49 it was found that the porosity evolutionin SLM of Ti64 was strongly affected by the accumulation inpowder denudation around the melt pool within a layer and the

accumulation of surface roughness across the layers. While for SLMof AlSi10Mg, deep keyholes were observed at the start/endingpoints of the scanning track due to the heat accumulation. Thesekeyholes with a low thermal stability collapse during laser process-ing, resulting in the formation of large pores.48 Vilaro et al.57 con-tributed the small spherical pores in SLM-fabricated Ti64 to theexistence of gas that may dissolve in the melt pool and remainthere due to the rapid solidification. Qiu et al.58 observed openpores on the melting surface of SLM-fabricated AlSi10Mg and

FIG. 8. Grain structure of AZ91D mag-nesium alloys after laser surfacemelting (a) without and (b) with anelectromagnetic field. Adapted withpermission from J. Zhou, J. Xu,S. Huang, Z. Hu, X. Meng, andX. Feng, Surf. Coat. Technol. 309,212–219 (2017). Copyright 2017,Elsevier.

FIG. 9. Electron backscattered diffraction results of AA7075 fabricated by SLM,with the addition of (a) 0, (b) 1, (c) 2, (d) 3, and (e) 4 wt. % Si, respectively.Adapted with permission from M. L. M. Sistiaga, R. Mertens, B. Vrancken,X. Wang, B. Van Hooreweder, J.-P. Kruth, and J. Van Humbeeck, J. Mater.Process. Technol. 238, 437–445 (2016). Copyright 2016, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-6

© 2019 Laser Institute of America

claimed that the formation of open pores was attributed to insuffi-cient feeding of molten material to solidification fronts and/or theincomplete remelting of the localized surface. They also argued thatthe violent interaction between laser and material would make themelt pool become turbulent, causing splashing, and resulting inhigh porosity.59 Panwisawas et al.60 concluded that for SLM ofTi64, the spherical or ellipsoidal pore formation depended on theinteracting domain area and flow velocity direction, and the flat orelongated internal pore formation was due to the incompletebonding between layers. In addition, Gong et al.61 claimed that theexcessive temperature would induce serious evaporation and formgas bubbles. The bubbles far beneath the surface may not havesufficient time to rise and escape during laser processing, leading tothe formation of pores.

B. Effects of laser parameters on porosity evolution

It is well accepted that the porosity is greatly influenced by theenergy density related parameters, such as laser power, scanningspeed, hatch spacing, and layer thickness, since these parametershave significant impacts on the temperature evolution and liquid/gas flow dynamics within the processed area. As shown in Fig. 10,there are optimal windows for laser power and scanning speed toobtain minimized porosity during SLM of Ti64. It was observedthat given a low laser power or a high scanning speed (low-energydensity), irregular pores or crevices were formed due to insufficientmelting and the balling phenomenon [Fig. 11(a)]; while given ahigh laser power or a low scanning speed (high-energy density),spherical pores were generated due to evaporated material and dis-solved inert gases becoming trapped in the melted material duringthe solidification process [Fig. 11(c)].62–64 In addition, the porositytypically increases with the widening of hatch spacing and thicken-ing of the powder layer.65–67 Aboulkhair et al.66 investigated theeffect of changing hatch spacing from 50 to 250 μm on the porosityduring SLM of AlSi10Mg. The gaps between adjacent tracks wereclearly observed once the hatch spacing exceeded 150 μm due tothe reduced intralayer bonding. Stoffregen et al.65 found that theporosity of SLM-built stainless steel part increased from 2.0% to11.4% as the layer thickness increased from 20 to 80 μm due to theinsufficient penetration depth of laser energy. Ma et al.68 suggestedthat the powder layer thickness should be less than 100 μm such

that the residual gas at the bottom of the melt pool can escape intime during the rapid solidification.

C. Porosity evolution as affected by scanning strategies

Scanning strategy is another key factor that affects the porosityin SLM-built parts. As aforementioned, pores can be formed withinthe unstable melt pool during solidification via various mechanisms.For the unidirectional and zigzag scanning strategies, a longer scan-ning distance increases the instability of the melt pool and leads tothe accumulation of defects like porosity. To address this issue, adouble-scanning strategy66,69 was developed to reduce the porosity byremelting the solidified layer or melting the remaining powder parti-cles. Alboulkhair et al.66 investigated the effect of double-scanningstrategy on porosity during SLM of AlSi10Mg. It was found that thesecond scanning reduced porosity by 4% at most. However, thedouble-scanning could introduce excessive thermal energy, leading tothe generation of additional pores. An alternative solution, called the“refill” or “knitting” strategy,70 is proposed to reduce porosity withinthe overlapping zone. During the SLM process, the surface tensiontends to drive the geometry of the melt material toward a sphericalshape, leading to the porosity between scanning tracks.71 The knittingstrategy is utilized to refill the pores/gaps in the overlapping area byscanning the next layer at an offset position. The cross-hatching strat-egy employs a similar technique as the method of repairing defectsbetween adjacent scanning tracks in the previous layer, resulting in astronger interlayer bonding and a lower porosity.72 The island-scanning strategy, which was developed to effectively decrease theresidual stress at first,58,73 may reduce the porosity during the SLMprocess as well because of short scanning distances. It was found thatthe effectiveness of reducing porosity using island-scanning strategywas strongly affected by the individual island size.74,75

D. Porosity evolution as affected by powdercharacteristics

In addition to laser parameters and scanning strategy, powdercharacteristics in terms of size distribution, morphology, and impuri-ties have considerable influences on the formation of pore structuresduring SLM. For instance, surfaces of SLM-built parts fabricatedusing different sizes of M2 high speed steel powders are shown inFig. 12.76 Micropores were clearly observed in SLM-built parts if the

FIG. 10. Effect of (a) laser power and(b) scanning speed on total porosityduring SLM of Ti64. Adapted withpermission from G. Kasperovich, J.Haubrich, J. Gussone, and G. Requena,Mater. Des. 105, 160–170 (2016).Copyright 2016, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-7

© 2019 Laser Institute of America

powder size was too large (>150 μm), since the laser energy inputwas not sufficient to fully melt powder particles to form a densestructure [Fig. 12(b)]. Moreover, coarse powders can lead to largevoids among particles.77 On the other hand, micropores were alsoeasily found in the parts if using the powders with a size too small(<38 μm) [Fig. 12(c)]. This is attributed to the enhanced laser–mate-rial interaction induced by higher surface energy or aggravating gasentrapment, and dewetting behavior induced by a higher level ofoxygen.19,76 Therefore, it was concluded that a proper powder sizeneeds to be selected in order for minimizing the porosity duringSLM. Li et al.78 investigated the densification behavior ofgas-atomized 316L stainless steel powders during SLM. It was foundthat the samples fabricated using gas-atomized powders exhibitedless porosity as compared to those produced from water-atomizedpowders. This was accredited to the lower oxygen content andhigher packing density of gas-atomized powder.

IV. PHASE TRANSFORMATION AND PRECIPITATION

A. Phase transformation during SLM of steels andtitanium alloys

For SLM of steels and titanium alloys, the phase transforma-tion plays another important role in determining the performance

of laser 3D-printed parts. The formation and distribution ofvarious phases in as-built parts are governed by the rapid heatingand cooling cycles during laser processing.

1. SLM of steels

During SLM of steels, the phase transformation is highlydetermined by the heating-induced austenitization andcooling-induced self-quenching effect. Normally, SLM of austeniticsteel powders, such as 304L and 316L, leads to a fully austeniticmicrostructure due to the rapid melting and cooling cycles.79,80

However, Yadollahi et al.81 observed a small volume fraction of δferrite (5–10 vol. %) distributed homogeneously within the austen-ite grains for direct laser deposited 316L stainless steel. It wasclaimed that the generation of ferrite phase was attributed to thecombing effect of rapid solidification and chemical composition.For martensitic steels, it was reported that different amounts ofmetastable austenite were observed after SLM.82–85 Faccgini et al.82

identified 72 vol. % metastable austenite and 28 vol. % twinnedmartensite in the parts fabricated by SLM using precipitation-hardening stainless steel powders (17-4PH). Given different laserprocessing parameters, LeBrun et al.83 found 36 vol. % austenite inthe SLM-fabricated 17-4PH samples. It was claimed that the

FIG. 11. Effect of laser energy densityon pore structure during SLM ofTi64. Adapted with permission from G.Kasperovich, J. Haubrich, J. Gussone,and G. Requena, Mater. Des. 105,160–170 (2016). Copyright 2016,Elsevier.

FIG. 12. Surfaces of SLM-built parts fabricated using different sizes of M2 high speed steel powders: (a) 53–150, (b) >150, and (c) <38 μm. Adapted with permission fromH. Niu and I. Chang, J. Mater. Sci. 35, 31–38 (2000). Copyright 2000, Springer Nature.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-8

© 2019 Laser Institute of America

residual thermal stress in the bulk material was the reason for theformation of the mechanically stabilized metastable austenitephase. SLM of AISI 420 stainless steel was conducted byKrakhmalev et al.,85 and it was found that the austenite volumefraction varied with processing regions due to different temperaturefields during laser processing. The upper layers exhibited 21 vol. %retained austenite, while the lower layers had a higher amount ofaustenite (57 vol. %) due to austenite reversion or growth ofretained austenite, which was induced by in situ carbon partition-ing and the diffusion process. A similar phenomenon in terms ofretained austenite was also observed in SLM of other martensiticsteels such as 18-Ni300 (Ref. 84) and AISI420.85

In addition, the effect of powder composition on phase trans-formation during SLM of steels was studied by Jerrard et al.80 Byvarying the ratio of 316L and 17-4PH stainless steel powder mix-tures, it was found that with the increase of 17-4PH content, theprecipitation of α-phase ferrite and cementite (Fe3C) appeared dueto the existence of interface boundary between austenitic and mar-tensitic grains. Once the weight percentage of 17-4PH powdersincreased to a critical value (75 wt. %), the final microstructure con-sisted of both austenite and martensite. When 17-4PH powderreached 100 wt. %, a tempered martensitic lath structure wasobserved.

2. SLM of titanium and its alloys

For SLM of titanium (Ti) and its alloys, various phases includ-ing α (hcp), β (bcc), α0 (hcp martensite), and α00 (fc-orthorhombic)can be manufactured.86 Chen et al.87 conducted SLM of Ti64 usingthe commercial Ti64 powders, consisting of hcp α-Ti phase andbcc β-Ti phase. It was found that the β phase was transformed intomartensitic α0 phase after laser processing. The similar results werealso reported in Ref. 88. Gu et al.18 studied the effect of laser scan-ning speed on the phase transformation during SLM of pure Ti. Itwas reported that a high laser scanning speed (400 mm/s) resultedin the formation of martensitic α0 phase. Xu et al.89 investigatedhow the phase transformation of SLM of Ti64 was affected by thelayer thickness, focal offset distance, and energy density. It wasfound that the reduction of layer thickness resulted in finer β grainsdue to the faster cooling condition. By decreasing the focal offsetdistance or increasing the energy density, an ultrafine lamellar (α +β) structure, rather than α0 martensite, was gradually prevailed. ForSLM of Ti-Nb alloys, Wang et al.90 studied the effect of Nb contenton the phase transformation. Given a Ti-Nb powder mixture of 15wt. % Nb, both α0 and β phases were observed in the SLM-builtparts. As the Nb content increased, the stability of the β phase wasenhanced and martensitic transformation (β→ α0) was restricted.Once the Nb content increased to 25 at. %, only the β phase wasobserved.

B. Precipitation during SLM

Precipitation can take place simultaneously with phase trans-formation during SLM of titanium alloys. It is also commonlyobserved in SLM-built nickel, aluminum, and magnesium alloys.The type and characteristics of precipitates which affect themechanical properties of as-built parts can be controlled by

adjusting laser processing parameters and/or the powder chemicalcomposition.

For SLM of titanium alloys, Chlebus et al.91 observed uni-formly distributed precipitates of the β-AlNbTi2 phase inSLM-fabricated Ti-6Al-7Nb alloys. It was claimed that suchuniform distribution of precipitates might be attributed to themicrosegregation of solute in β grains, leading to the enhancedtensile and compressive strengths of the as-built parts due to theprecipitation-hardening effect. Li et al.92 claimed the precipitationkinetics in SLM of Ti-45Al-2Cr-5Nb alloys as β→ α2 (Ti3Al) + γ(TiAl) and disorder-order transformation (β→ B2). It was foundthat by increasing the laser scanning speed, the contents of B2 andγ precipitates were increased with a decrease of the content of theα2 matrix phase.

In the study of SLM of Ni-based superalloys, Idell et al.93 andJia and Gu94 both identified the nucleation of γ0 phase precipitates[Ni3(Al,Ti)], while the unexpected δ-phase precipitates (Ni3Nb)were also found in the matrix93 due to the significant microsegrega-tion of niobium caused by rapid heating and cooling cycles. It wasclaimed that the nucleation of δ-phase precipitates would berestricted when the local content of niobium was less than 6.8%molar fraction in the SLM process. It was also concluded that byincreasing the laser energy, the coarsening effect led to a larger sizeof γ0 precipitates, and as the sample building up, the previous layerexperienced an aging treatment, which further promoted the for-mation of the γ0 phase.94

For SLM of Al/Mg alloys, the final products usually consist ofprimary α-Al/α-Mg and intermetallic phases depending on thechemical composition of alloy powder.95–97 Sun et al.96 optimizedthe laser energy in SLM of Al-8.5Fe-1.3V-1.7Si alloys to obtain highcooling rates (>105 °C/s), leading to the formation of Al12(Fe, V)3Siphase rather than a brittle θ-Al13Fe4 phase, which forms at lowcooling rates (<103 °C/s). The Al-20Si-5Fe-3Cu-1Mg alloy was man-ufactured using SLM by Ma et al.97 It was found that onlyδ-Al4FeSi2 phase was observed in the as-built part, while the as-castalloy typically consisted of both β-Al5FeSi and δ-Al4FeSi2. Shuaiet al.98 attributed the precipitation of Mg7Zn3 in the SLM-fabricatedMg alloy ZK60 to the rapid solidification, which hindered thedecomposition of this high temperature phase (Mg7Zn3→ α-Mg +MgZn). It was claimed that the uniformly distributed Mg7Zn3 phasein ZK60 could contribute to the improvement of strength.

In addition to the aforementioned different precipitates, therapid cooling rate of the SLM process can also lead to the forma-tion of a supersaturated solid solution and suppress the precipita-tion, especially for Al-Si and Al-Sc alloys. Maeshima and Oh-ishi99

conducted SLM experiments of AlSi10Mg material and found thatthe Si content in α-Al was 2.2 at. %, which significantly exceededthe maximum solubility of 1.4 at. % in the equilibrium state ofAlSi10Mg. Zhang et al.100 investigated the microstructure evolutionduring SLM of Sc modified Al–Mg alloys. It was found that Al3(Sc,Zr) precipitates were obtained at the bottom of molten pool at ascanning speed of 1800 mm/s, while increasing the scanning speedto 3000 mm/s resulted in the formation of supersaturated solid sol-ution and no precipitates were observed. Moreover, the as-builtsupersaturated alloys can be used for subsequent heat treatmentto improve mechanical properties,101 which deserves furtherinvestigation.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-9

© 2019 Laser Institute of America

V. BALLING AND CRACKING

Apart from the evolution of grain structure, pore structureand metallic phases during SLM, various microstructure defectsneed to be studied, as those defects could result in poor quality,low dimension accuracy, and limited mechanical properties ofas-built parts.

A. Balling phenomenon

Balling effect,102–111 as shown in Fig. 13, is a common phe-nomenon during SLM, which is detrimental to part formation. Itcan lead to discontinuous tracks and incomplete interline bonding,as well as the porosity, and even delamination. Generally, ballingphenomenon takes place with a decrease of surface energy, but theformation mechanisms are rather complicated.

Gu and Shen110 conducted an experimental study on SLM of316L stainless steel, in which the balling phenomenon was consid-ered as two categories based on the ball size. The coarsened ballswere caused by the limited liquid formation due to a low laserenergy input, which resulted in a high viscosity of the liquid–solidmixture and, thus, hindered the liquid flow and particle arrange-ment, eventually leading to the aggregation of particles and the for-mation of balls in the scale of laser spot size. On the other hand,the micrometer-scaled balls (∼10 μm) were formed by small drop-lets splashing from the surface of unstable molten tracks when laserprocessing was conducted with a high scanning speed. Niu andChnag106 concluded that a high laser scanning speed during SLMwould promote the breaking up of molten tracks into balls. Zhouet al.108 studied the balling phenomenon in SLM of tungsten. Itwas claimed that the slow wetting and spreading speed of tungsten,caused by its high surface tension and viscosity, were the main

reasons for the large balling tendency. In addition, the oxidation oftungsten melt pool also affected the surface tension gradient andthe thermocapillary convection, which favored the balling initia-tion. Das109 investigated how the balling phenomenon was affectedby the wettability. It was concluded that the oxidation on the sub-strate can significantly reduce the wettability of liquid metal. As aresult, the molten material tended to form spheres to decreasesurface energy. SLM of 316L stainless was also investigated experi-mentally by Shen et al.,112 and it was observed that the ballsformed during the initial scanning track were larger than the onesformed during the subsequent scanning tracks. Such a phenome-non was explained by the sharp temperature gradient between theheated area and unprocessed area for the initial scan, and lowertemperature gradient generated for the subsequent scanning due tothe overlap between scanning tracks. The effects of laser power,scanning speed, layer thickness, and oxygen content on ballingphenomenon were discussed thoroughly by Gu and Shen110 and Liet al.111 In general, by adjusting laser processing parameters prop-erly, such as increasing the laser power or reducing the scanningspeed, layer thickness, hatch spacing, and/or oxidation, the ballingeffect during SLM process can be alleviated.

B. Cracking

Cracking is another major concern associated with the rapidheating and cooling cycles, large solidification shrinkage, and highcoefficient of thermal expansion.19,63,113,114 Given its complicatedforming mechanisms, it is usually difficult to eliminate cracks onlyby the optimization of SLM processing parameters.115,116

Carter et al.116 characterized the cracks in SLM-fabricatedNi-based superalloy CM247LC into two categories: solidificationcracks growing both longitudinally and transversely to the builddirection [Fig. 14(a)] and grain boundary cracks lying along grainboundaries [Fig. 14(b)]. Three mechanisms, including liquationcracking, strain-age cracking, and ductility-dip cracking, were pro-posed to explain the formation of grain boundary cracks. Theeffects of laser power intensity, scanning speed, and hatch spacingon the cracking density and morphology were studied. It wasreported that a high laser intensity led to the solidification crackingwhile a low laser intensity led to the grain boundary cracking.Additionally, the solidification cracking was increased by decreasingthe hatch spacing. Wang et al.117 studied the crack growth asaffected by different scanning strategies during SLM of molybde-num powders. By rotating the laser scanning direction betweenadjacent layers, the columnar grain structure transferred into theinterlocked structure, which greatly shortened the growth path forcracks and caused the crack deviation. Chemical compositions wereidentified as another important factor affecting cracking duringSLM of Hastelloy-X alloys.118,119 It was found that the minor ele-ments, such as Mn, Si, and S, could lead to a decrease of solidifica-tion temperature and a microsegregation at grain boundaries,resulting in the crack initiation. Cloots et al.120 investigated howcracking was affected by a laser beam profile. It was reported thatfor SLM-built Ni-based superalloy IN738LC, the crack density wasreduced from 0.27 to 0.11 mm/mm2 by using a laser beam with aDoughnut profile rather than a Gaussian profile. This was

FIG. 13. Surface of SLM-built part from stainless steel powder with typicalballing phenomenon. Adapted with permission from D. Gu and Y. Shen, Mater.Des. 30, 2903–2910 (2009). Copyright 2009, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-10

© 2019 Laser Institute of America

attributed to the difference in melt shape and grain structure whenapplying different beam profiles.

C. Approaches to eliminate or minimize balling andcracking

The balling effect and cracking behavior are significant issuesrestricting the applications of SLM. Some researchers claim that itis difficult to completely eliminate balling and cracking by simplytuning the laser parameters and scanning strategy. Alternativeapproaches to tackle the balling and cracking issues are of criticalneeds. To alleviate balling effects, improving the wettability of thepowder material has been proved to be an effective method. Forinstance, Agarwala et al.121 developed a multiphase powderapproach by prealloying phosphorus with a copper alloy toimprove the wetting characteristics, resulting in the as-built samplewith smooth surface finish, as shown in Fig. 15. To eliminate crack-ing during SLM, preheating substrate is commonly recommendedfor stainless steels and Ti alloys.122,123 However, it is ineffective for

metals like aluminum alloys, due to the severe level of crackinginduced by the large solidification temperature range, high thermalexpansion coefficient, and large solidification shrinkage. Chemicaladditives serve as another possible approach for preventing crack-ing. As shown in Fig. 16, it was found by Zhang et al.124 that theaddition of 2 wt. % Zr to Al-Cu-Mg alloy powder significantlyreduced and eventually eliminated the crack formation duringSLM, which was attributed to the formation of Al3Zr precipitatesand grain refinement. Further systematic investigations are neces-sary to identify and optimize approaches for eliminating or mini-mizing balling and/or cracking in SLM of metallic materials.

VI. COMPUTATIONAL MODELING OFMICROSTRUCTURE EVOLUTION IN THE SLM PROCESS

Due to the rapid heating and cooling cycles during the SLMprocess, in situ observation of microstructure evolution duringlaser processing is of specific difficulties. While the parametricstudy by experiments is time- and cost-consuming, the computa-tional modeling offers the exceptional capability of bridging theprocess–microstructure relationship and, therefore, provides guid-ance and insights for process design and optimization. For SLMmodeling, the primary focus has been placed on the grain structureevolution. In the literature, the grain growth during SLM can besimulated using three computation approaches including MonteCarlo (MC), cellular automaton (CA), and phase field (PF)simulation.

A. Monte Carlo simulation

As one of the earliest methods for microstructural modeling,MC simulation is widely used in casting and welding processes. Itutilizes the change in the system energy as a criterion to generatethe probability of grain growth, i.e., P = exp(−ΔE/kBT) if ΔE > 0 orP = 1 if ΔE≤ 1, where ΔE is the change in total energy, kB isBoltzmann’s constant, and T is the local temperature.125 Rodgerset al.126 developed a modified kinetic MC 3D model incorporatingthe molten zone movement to predict the grain evolution in multi-layer SLM. The shape of molten zone and heat-affected zone, thetemperature gradient, and the scanning direction were taken intoaccount in this model. The temperature-dependent grain boundarymobility was incorporated into the probability function. It wasreported that the simulated results were in good agreement withexperimental findings, as illustrated in Fig. 17. MC method

FIG. 14. Crack details under differentlaser energy densities: (a) solidificationcracks in a “high-energy” build and (b)grain boundary cracks in a “low-energy” build. Adapted with permissionfrom L. N. Carter, M. M. Attallah, andR. C. Reed, Superalloys 2012, 577–586 (2012). Copyright 2012, JohnWiley and Sons.

FIG. 15. The surfaces of SLM-built Cu alloy samples (a) without phosphorusshowing balling and rough surface and (b) with phosphorus showing smoothsurface. Adapted with permission from M. Agarwala, D. Bourell, J. Beaman, H.Marcus, and J. Barlow, Rapid Prototyping J. 1, 26–36 (1995). Copyright 1995,Emerald Group Publishing.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-11

© 2019 Laser Institute of America

provides the exceptional capability of large scale simulation with alow computational cost. However, such a probability-based methodlacks physical principles and hence is unable to predict detailedmicrostructure characteristics such as the dendrite front.

B. Cellular automaton simulation

For CA simulation, each computational cell has a solid frac-tion fs representing the local state (liquid if fs = 0, solid if fs = 1, andinterface if 0 < fs < 1), as illustrated in Fig. 18.127 The interface con-sists of one layer of the CA cell that separates solid and liquid.Generally, a decenter square algorithm is used for the CAmodel,128 and the microstructure evolution is controlled by theinterface growth rate obtained from coupled mass or heat equa-tions. Many researchers34,128–131 have adopted a CA-finite elementcoupled model to predict the microstructure evolution in a largeregion. The temperature profile induced by laser energy is usuallyacquired first on the macroscale. Then, on the microscale CAmodel, the density of nucleus and the interface growth rate (tip

velocity) are calculated based on the real-time undercooling, andthe nucleation sites are updated to the grids at each time step. Yinand Felicelli129 used the CA model to predict the effects of process-ing parameters (including the laser scanning speed, layer thickness,and substrate size) on the dendrite morphology during SLM of asingle layer Fe-0.13 wt. % alloy. Nie et al.130 simulated the micro-structure evolution during SLM of Nb-bearing nickel-based super-alloy IN718. The coefficient relating the growth angle and thepreferential crystallographic orientation was introduced to capturethe growth of dendrites. Lopez-Botello et al.131 further applied thismethod to simulate a multilayer SLM of AA2024. As one of themost advanced models, CA is coupled with lattice Boltzmann (LB)method for SLM simulation, where the temperature history is pre-dicted using the LB method with consideration on the capillarity,wetting behavior, and melt pool dynamics,132,133 as shown inFig. 19. The CA method is exceptional due to its highly computa-tionally efficient and capability of simulating anisotropic effectsduring solidification. However, it fails to provide a detailed descrip-tion of dendritic morphology, because the tip curvature and

FIG. 16. The cross sections ofSLM-built Al-Cu-Mg parts fabricatedusing different scanning speeds with/without the addition of Zr. Adapted withpermission from H. Zhang, H. Zhu,X. Nie, J. Yin, Z. Hu, and X. Zeng, Scr.Mater. 134, 6–10 (2017). Copyright2017, Elsevier.

FIG. 17. Comparision of experimentaland MC simulation results of SLM ofIN718 with different scanning strate-gies: (a) unidirectional scan and (b)bidirectional scan. Adapted with per-mission from T. M. Rodgers, J. D.Madison, and V. Tikare, Comput.Mater. Sci. 135, 78–89 (2017).Copyright 2017, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-12

© 2019 Laser Institute of America

secondary dendrite arm are greatly affected by grid meshing sincethe interface only consists of one layer of cells.134 It is also not suit-able to simulate complex processes such as solidification of themulticomponent alloy due to the constraints of the growth kineticsformulation.135

C. Phase field simulation

In PF simulation, an order parameter (w) is used to representthe state of grids. For the solid phase, w = 1, and for the liquidphase, w = 0. For the interface, 0 < w < 1, which usually consists of afew cells and, thus, provides a smooth transition of material prop-erties across the interface. The microstructure evolution is obtainedby the solution of a set of functions including the phase field equa-tion, solute, and/or temperature conserved equations. In the workof Cao and Choi,136 laser processing of H13 tool steel was simu-lated by coupling PF microscale model with the macroscale tem-perature field obtained from a volume of the fluid based algorithm.The microstructure evolution was predicted by solving the phasefield equation, composition equation, and energy equation withdimensionless temperature and composition. The solidification ofpure metal and binary alloy was both analyzed, and the effects ofmicrocell location and undercooling on the microstructure evolu-tion were studied. Keller et al.137 assumed a local steady-state con-dition during SLM by using an imposed temperature field, T(x,t) =T0 +G(x –Vst) as an input for the phase field equation, where tem-perature gradient G and velocity Vs were considered to be constant.The microstructure evolution of Ni-Nb binary alloys was simulatedby solving the phase field equation and the composition field equa-tion. Fallah et al.138 used a similar method to predict the micro-structure evolution by solving a phase field equation and asupersaturation field equation. The dendrite growth morphology,especially the dendrite arm spacing, during directional solidificationof Ti-Nb alloys was studied. Kundin et al.139 successfully predictedthe grain structure during SLM of IN718 alloys. The model wasformulated with a general phase field equation and multiplediffusion equations for each component. The temperature-dependent parameters used in the diffusion equations including theequilibrium concentration in liquid and solid phases were

recalculated in each time step. The modeling results were validatedby the steady-state growth problem taking solute trapping intoaccount. The effects of undercooling on dendritic tip radius andvelocity were predicted for multicomponent alloy solidification.Recently, Acharya et al.140 used a computational fluid dynamicsmodel to describe the characteristics of molten pool and to intro-duce the laser heat source as affected by the scanning speed, Vx, inthe phase field equation, composition equation, and energy equa-tion. This model was used to predict the evolution of microstruc-ture features during SLM of IN718 alloys, including the dendriticfragmentation, dendritic size, and dendritic orientation. As shownin Fig. 20, the dendritic growth was simulated in both longitudinaland transverse cross sections. The effect of scanning speed was ana-lyzed via simulation results and further discussed in terms of multi-layer build. The PF method does not require the tracking ofinterface position, and hence the number of governing equations isminimized. It is suitable to simulate complex microstructure andmulticomponent alloy. However, it is restricted due to the smallscale simulation with a relatively high computational cost.

VII. CURRENT CHALLENGES AND FUTURE RESEARCHDIRECTIONS

As a promising 3D printing technology, SLM has been exten-sively applied to different fields and experimented on various mate-rials. However, the rapid processing makes the microstructureevolution complicated with issues such as anisotropic mechanicalbehaviors, balling, and/or cracking.

The anisotropic mechanical properties of SLM-built parts aremainly attributed to anisotropic grain orientations aligned with thethermal gradient directions as the solidification proceeds. The heatflow direction and temperature gradient as affected by variousprocess parameters play major roles in determining the graingrowth during the SLM process. Although the effects of laser powerand scanning speed on the grain growth have been extensivelystudied for different materials, a few efforts have been put onunderstanding effects of other laser parameters (hatching spacingand layer thickness)141,142 and powder characteristics (size distribu-tion and chemical composition). Additives and/or dynamic

FIG. 18. Configuration of CA cellsincluding solid ( fs = 1), liquid ( fs = 0),and interface (0 < fs < 1). Adapted withpermission from A. Choudhury, K.Reuther, E. Wesner, A. August,B. Nestler, and M. Rettenmayr,Comput. Mater. Sci. 55, 263–268(2012). Copyright 2012, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-13

© 2019 Laser Institute of America

FIG. 19. Predicted grain structure inSLM of IN718 using CA simulation,where the temperature input isobtained from the LB model. Adaptedwith permission from A. Rai, M. Markl,and C. Körner, Comput. Mater. Sci.124, 37–48 (2016) Copyright 2016,Elsevier.

FIG. 20. Prediction of microstructureevolution during SLM of IN718 alloys:(a) schematic of scanning track andrelated cross-section views, and PF-simulated dendritic growth in (b) longi-tudinal cross section and (c) transversecross section. Adapted with permissionfrom R. Acharya, J. A. Sharon, andA. Staroselsky, Acta Mater. 124, 360–371 (2017). Copyright 2017, Elsevier.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-14

© 2019 Laser Institute of America

approaches integrated into the SLM process provide potential solu-tions to address the issue of anisotropic mechanical properties bygenerating equiaxed grains all over the melting area. Nevertheless,the uniform distribution of nanoadditives in the powder bedremains a challenge, while the integration of dynamic approacheswith SLM brings difficulties in manufacturing process design.Altering the powder composition could serve as an alternativemethod to obtain ultrafine grains and address the anisotropic effectin SLM; however, further investigations are needed to elucidate thefundamental mechanism responsible for the powder compositioneffect. On the other hand, in order to tackle the balling and/orcracking challenges, in addition to design and optimization of laserparameter and scanning strategy, alternative approaches are of criti-cal needs, such as improving the wettability of the powder materi-als, employing of preheating substrate, and adding chemicaladditives. However, the lack of systematic investigations makesthese approaches less effective.

Porous structures offer a variety of advantages, such as lowthermal conductivity, low density, short diffusion path, and largesurface area of reaction. The SLM-fabricated porous structure hasshown its promising future due to the combination of advantagesof laser-based rapid prototyping and unique properties of porousmaterial toward applications in mass, momentum, and energytransport. Recently, a few investigations have been conducted witha focus on porous structure fabrication using SLM by partialmelting143–145 or adding pore-forming agent;146 however, the prop-erties of as-built porous parts have been barely studied. It is envi-sioned that future studies aiming at advancing the understandingof process–microstructure–property relationship of SLM-fabricatedporous structures will open new avenues for applications of SLMtechnology.

For SLM of metallic materials with phase transformation and/or precipitation phenomena, it is feasible to control the phasetransformation or precipitation during SLM by adjusting processingconditions, leading to tunable bulk properties such as hardness,yield strength,147 and magnetic performance.148 However, due tothe rapid melting and solidification, the phase transformation andprecipitation in the SLM process can be very different as comparedto those in conventional manufacturing processes like casting andheat treatment. Thermal dynamics and physical metallurgy studieson phase transformation and precipitation during SLM will be thefirst investigation step toward a fundamental understanding ofprocess–structure–property relationship during SLM of metallicmaterials, for which the phase volume fraction and/or precipitatecharacteristics play critical roles on determining their propertiesand multiphysics performance.

Computational methods exhibit their advantages in processoptimization and microstructure prediction. The physics-basedmodeling efforts nowadays, such as CA and PF, have mainlyfocused on the microstructure evolution during the single scan.However, the SLM process involves remelting of previous scannedtracks and solidified layers, leading to the grain reformation.Therefore, it is necessary to further develop models for multiscan/multilayer to effectively predict the entire microstructure in 3Dcomponents.149–151 Moreover, the microstructure simulation typi-cally needs temperature history as the input. Currently, the temper-ature profile is often acquired from a simple heat conduction

model by considering the powder bed as a bulk material.Nevertheless, the material properties such as thermal conductivityand absorptivity can change as the powder particles melt/consoli-date together. In addition, when the laser beam is directed onto thepowder bed, the loose powder particles would result in multipleabsorptions and scatterings of the beam energy,152 rather than asimple absorption as a bulk material. It would be more accurate forpredicting the microstructure evolution by improving the tempera-ture model with consideration on these important facts.Furthermore, researchers have been dedicating to simulating thegrain growth during SLM, while there is very little research thatinvolves the simulation of porosity formation,153 phase transforma-tion,151 or precipitation during SLM.

Considering the aforementioned current challenges in SLMprocess design and optimization, future investigations are neededto address the following knowledge and technology gaps:

1. The SLM process design and optimization toward 3D printingof porous structure is of great importance. 3D-printed compo-nents with controlled porosity find many applications includingwater filtration, thermal storage, vacuum insulation, and bioac-tive implant.154 However, for the SLM process, most researchershave focused on manufacturing “full dense” or “pores-free” partvia adjusting process parameters or post-treatment conditions.To date, there are few works on manufacturing porous structurewith SLM process,143,145,155 let alone the properties of theas-built porous parts. The capability and feasibility of fabricatingporous components using SLM is promising and needs to befurther investigated.

2. Adding chemical additives into commercial alloy systems hasbeen proved to be an effective method for tuning the grain size,generating the strengthening second phase, and alleviatingballing or cracking. In addition, the formation of new phasescould affect the melt pool by altering the temperature distribu-tion and thermal behavior, such as dynamic viscosity, capillaryinstability, and surface tension, leading to beneficial effects onreducing porosity and residual stress. Furthermore, the additionsmake SLM possible for processing some unprintable metallicmaterials, which is of enormous significance to expand thematerial library for SLM and hence enlarge the applicationareas. Future research with a focus on adding chemical additivesinto alloy systems will further advance understanding, control,and optimization of the SLM process.

3. Computational methods are powerful tools for process designsince it is difficult to directly observe the microstructure evolu-tion during the SLM process. The future effort is necessary tofurther develop models for multiscan/multilayer simulation toeffectively predict the entire microstructure in 3D-printed com-ponents. Moreover, the prediction of temperature profile duringSLM can be further improved by considering the powder sizeeffect on the absorption and scattering of laser beam energy.

VIII. CONCLUSION

In this paper, the microstructure evolution during SLM ofmetallic materials is reviewed. The effects of processing conditions(laser parameters, scanning strategies, additives, and dynamic

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-15

© 2019 Laser Institute of America

approaches) on the grain growth during SLM are reviewed in termsof grain morphology, size, and orientation. The formation of porestructures during SLM as affected by laser parameters, scanningstrategies, and powder characteristics is discussed. In addition, thephase transformation during SLM of steels and titanium alloys isreviewed. The precipitation during SLM of titanium, nickel, andAl/Mg alloys is studied. Moreover, since the balling and crackingare complicated and difficult to avoid during SLM, the formationmechanisms and classification are discussed in detail. Finally,various computational methods including MC, CA, and PFapproaches to predict the grain structure evolution during the SLMprocess are introduced. This review summarizes the currentadvances in understanding of the process–microstructure relation-ship during SLM of metallic materials. It provides guidance forprocess design and optimization in SLM of metallic materials.

ACKNOWLEDGMENT

The authors appreciate financial support by start-up fundingfrom the Department of Mechanical Engineering at the Universityof Nevada, Reno.

REFERENCES1K.-H. Leitz, P. Singer, A. Plankensteiner, B. Tabernig, H. Kestler, and L. Sigl,“Multi-physical simulation of selective laser melting,” Met. Powder Rep. 72,331–338 (2017).2M. Yakout, M. Elbestawi, and S. C. Veldhuis, “On the characterization of stain-less steel 316L parts produced by selective laser melting,” Int. J. Adv. Manuf.Technol. 95, 1953–1974 (2017).3Y. Song, Y. Yan, R. Zhang, D. Xu, and F. Wang, “Manufacture of the die of anautomobile deck part based on rapid prototyping and rapid tooling technology,”J. Mater. Process. Technol. 120, 237–242 (2002).4W. W. Wits, S. J. Weitkamp, and J. van Es, “Metal additive manufacturing of ahigh-pressure micro-pump,” Procedia CIRP 7, 252–257 (2013).5S. Hou, S. Qi, D. A. Hutt, J. R. Tyrer, M. Mu, and Z. Zhou, “Three dimensionalprinted electronic devices realised by selective laser melting of copper/high-density-polyethylene powder mixtures,” J. Mater. Process. Technol. 254,310–324 (2018).6M. Wong, I. Owen, and C. J. Sutcliffe, “Pressure loss and heat transfer throughheat sinks produced by selective laser melting,” Heat Transf. Eng. 30, 1068–1076(2009).7M. Zenou, O. Ermak, A. Saar, and Z. Kotler, “Laser sintering of copper nano-particles,” J. Phys. D Appl. Phys. 47, 025501 (2013).8T. Hayashi, K. Maekawa, M. Tamura, and K. Hanyu, “Selective laser sinteringmethod using titanium powder sheet toward fabrication of porous bone substi-tutes,” JSME Int. J. Ser. A Solid Mech. Mater. Eng. 48, 369–375 (2005).9M. A. Stoodley, J. R. Abbott, and D. A. Simpson, “Titanium cranioplasty using3-D computer modelling of skull defects,” J. Clin. Neurosci. 3, 149–155 (1996).10D. A. Hollander, M. Von Walter, T. Wirtz, R. Sellei, B. Schmidt-Rohlfing,O. Paar, and H.-J. Erli, “Structural, mechanical and in vitro characterization ofindividually structured Ti–6Al–4V produced by direct laser forming,”Biomaterials 27, 955–963 (2006).11M. Kanazawa, M. Iwaki, S. Minakuchi, and N. Nomura, “Fabrication of tita-nium alloy frameworks for complete dentures by selective laser melting,”J. Prosthet. Dent. 112, 1441–1447 (2014).12V. Beal, P. Erasenthiran, C. H. Ahrens, and P. Dickens, “Evaluating the use offunctionally graded materials inserts produced by selective laser melting on theinjection moulding of plastics parts,” Proc. Inst. Mech. Eng. B J. Eng. Manuf.221, 945–954 (2007).

13M. Erdal, S. Dag, Y. Jande, and C. Tekin, “Manufacturing of functionallygraded porous products by selective laser sintering,” Mater. Sci. Forum 631–632,253–258 (2010).14Z. Liu, D. Zhang, S. Sing, C. Chua, and L. Loh, “Interfacial characterization ofSLM parts in multi-material processing: Metallurgical diffusion between 316Lstainless steel and C18400 copper alloy,” Mater. Charact. 94, 116–125 (2014).15D. Lin, C. R. Liu, and G. J. Cheng, “Single-layer graphene oxide reinforcedmetal matrix composites by laser sintering: Microstructure and mechanical prop-erty enhancement,” Acta Mater. 80, 183–193 (2014).16D. Lin, C. Ye, Y. Liao, S. Suslov, R. Liu, and G. J. Cheng, “Mechanism offatigue performance enhancement in a laser sintered superhard nanoparticlesreinforced nanocomposite followed by laser shock peening,” J. Appl. Phys. 113,133509 (2013).17D. Lin, C. Richard Liu, and G. J. Cheng, “Laser sintering of separated and uni-formly distributed multiwall carbon nanotubes integrated iron nanocomposites,”J. Appl. Phys. 115, 113513 (2014).18D. Gu, Y.-C. Hagedorn, W. Meiners, G. Meng, R. J. S. Batista, K. Wissenbach,and R. Poprawe, “Densification behavior, microstructure evolution, and wearperformance of selective laser melting processed commercially pure titanium,”Acta Mater. 60, 3849–3860 (2012).19E. O. t. Olakanmi, R. Cochrane, and K. Dalgarno, “A review on selective lasersintering/melting (SLS/SLM) of aluminium alloy powders: Processing, micro-structure, and properties,” Prog. Mater. Sci. 74, 401–477 (2015).20P. Yuan and D. Gu, “Molten pool behaviour and its physical mechanismduring selective laser melting of TiC/AlSi10Mg nanocomposites: Simulation andexperiments,” J. Phys. D Appl. Phys. 48, 035303 (2015).21P. Clementz and J.-N. Pernin, “Homogenization modeling of capillary forcesin selective laser sintering,” Int. J. Eng. Sci. 41, 2305–2333 (2003).22H. Shipley, D. McDonnell, M. Culleton, R. Lupoi, G. O’Donnell, andD. Trimble, “Optimisation of process parameters to address fundamental chal-lenges during selective laser melting of Ti-6Al-4V: A review,” Int. J. Mach. ToolsManuf. 128, 1–20 (2018).23Z. X. Khoo, Y. Liu, J. An, C. K. Chua, Y. F. Shen, and C. N. Kuo, “A review ofselective laser melted NiTi shape memory alloy,” Materials 11, pii: E519 (2018).24X. Wang, X. Gong, and K. Chou, “Review on powder-bed laser additive manu-facturing of Inconel 718 parts,” Proc. Inst. Mech. Eng. Part B J. Eng. Manuf.231, 1890–1903 (2017).25L. C. Zhang and H. Attar, “Selective laser melting of titanium alloys and tita-nium matrix composites for biomedical applications: A review,” Adv. Eng.Mater. 18, 463–475 (2016).26L.-C. Zhang, H. Attar, M. Calin, and J. Eckert, “Review on manufacture byselective laser melting and properties of titanium based materials for biomedicalapplications,” Mater. Technol. 31, 66–76 (2016).27T. W. Eagar and A. D. Mazzeo, Welding Process Fundamentals (ASMInternational, Materials Park, OH, 2011).28F. Yan, W. Xiong, and E. J. Faierson, “Grain structure control of additivelymanufactured metallic materials,” Materials 10, 1260 (2017).29J. Ion, Laser Processing of Engineering Materials: Principles, Procedure andIndustrial Application (Elsevier, New York, 2005).30J. Dupont, Fundamentals of Weld Solidification (ASM International, MaterialsPark, OH, 2011).31S. David, S. Babu, and J. Vitek, “Welding: Solidification and microstructure,”J. Miner. Met. Mater. Soc. 55, 14–20 (2003).32S. Kou, Welding Metallurgy (Wiley, New York, 2003).33A. R. A. Dezfoli, W.-S. Hwang, J. Augusto, A. K. Shukur, and S. Tzeng,“Modeling of poly-crystalline silicon ingot crystallization during casting and the-oretical suggestion for ingot quality improvement,” Mater. Sci. Semicond.Process. 53, 36–46 (2016).34A. R. A. Dezfoli, W.-S. Hwang, W.-C. Huang, and T.-W. Tsai, “Determinationand controlling of grain structure of metals after laser incidence: Theoreticalapproach,” Sci. Rep. 7, 41527 (2017).35M. Li, Y. He, and X. Yuan, “Effect of nano-Y2O3 on microstructure of lasercladding cobalt-based alloy coatings,” Appl. Surf. Sci. 252, 2882–2887 (2006).

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-16

© 2019 Laser Institute of America

36S. H. Zhang, M. X. Li, T. Y. Cho, J. H. Yoon, C. G. Lee, and Y. Z. He, “Laserclad Ni-base alloy added nano-and micron-size CeO2 composites,” Opt. LaserTechnol. 40, 716–722 (2008).37X. Li, G. Ji, Z. Chen, A. Addad, Y. Wu, H. Wang, J. Vleugels, J. VanHumbeeck, and J.-P. Kruth, “Selective laser melting of nano-TiB2 decoratedAlSi10Mg alloy with high fracture strength and ductility,” Acta Mater. 129,183–193 (2017).38J. H. Martin, B. D. Yahata, J. M. Hundley, J. A. Mayer, T. A. Schaedler, andT. M. Pollock, “3D printing of high-strength aluminium alloys,” Nature 549,365–369 (2017).39C. Ma, L. Chen, C. Cao, and X. Li, “Nanoparticle-induced unusual meltingand solidification behaviours of metals,” Nat. Commun. 8, 14178 (2017).40W.-C. Huang, C.-S. Chuang, C.-C. Lin, C.-H. Wu, D.-Y. Lin, S.-H. Liu,W.-P. Tseng, and J.-B. Horng, “Microstructure-controllable laser additive manu-facturing process for metal products,” Phys. Procedia 56, 58–63 (2014).41J. Zhou, J. Xu, S. Huang, Z. Hu, X. Meng, and X. Feng, “Effect of laser surfacemelting with alternating magnetic field on wear and corrosion resistance of mag-nesium alloy,” Surf. Coat. Technol. 309, 212–219 (2017).42J. Xu, J. Zhou, W. Tan, S. Huang, S. Wang, and W. He, “Study on laser surfacemelting of AZ31B magnesium alloy with different ultrasonic vibration ampli-tude,” Corros. Eng. Sci. Technol. 53, 73–79 (2017).43T. Yuan, Z. Luo, and S. Kou, “Grain refining of magnesium welds by arc oscil-lation,” Acta Mater. 116, 166–176 (2016).44K. P. Monroy, J. Delgado, L. Sereno, J. Ciurana, and N. J. Hendrichs, “Effectsof the selective laser melting manufacturing process on the properties ofCoCrMo single tracks,” Met. Mater. Int. 20, 873–884 (2014).45D. Sun, X. Li, and W. Tan, presented in the 28th Annual International SolidFreeform Fabrication Symposium, Austin, TX, 7–9 August 2017.46I. Yadroitsev, P. Krakhmalev, I. Yadroitsava, S. Johansson, and I. Smurov,“Energy input effect on morphology and microstructure of selective laser meltingsingle track from metallic powder,” J. Mater. Process. Technol. 213, 606–613(2013).47J. Olsen, X. Zhou, Y. Zhong, L. Liu, D. Wang, C. Yu, Y. Wang, K. Li, L. Xing,and J. Ma, presented in IOP Conference Series: Materials Science andEngineering, Risø, Denmark, 4–8 September 2017.48L. Thijs, K. Kempen, J.-P. Kruth, and J. Van Humbeeck, “Fine-structured alu-minium products with controllable texture by selective laser melting of pre-alloyed AlSi10Mg powder,” Acta Mater. 61, 1809–1819 (2013).49L. Thijs, F. Verhaeghe, T. Craeghs, J. Van Humbeeck, and J.-P. Kruth, “Astudy of the microstructural evolution during selective laser melting of Ti–6Al–4V,” Acta Mater. 58, 3303–3312 (2010).50Y. M. Arısoy, L. E. Criales, T. Özel, B. Lane, S. Moylan, and A. Donmez,“Influence of scan strategy and process parameters on microstructure and itsoptimization in additively manufactured nickel alloy 625 via laser powder bedfusion,” Int. J. Adv. Manuf. Technol. 90, 1393–1417 (2017).51L. N. Carter, C. Martin, P. J. Withers, and M. M. Attallah, “The influence ofthe laser scan strategy on grain structure and cracking behaviour in SLMpowder-bed fabricated nickel superalloy,” J. Alloys Compd. 615, 338–347 (2014).52M. Easton and D. Stjohn, “Grain refinement of aluminum alloys: Part I. Thenucleant and solute paradigms—A review of the literature,” Metall. Mater. Trans.A 30, 1613–1623 (1999).53R. Streubel, M. B. Wilms, C. Doñate-Buendía, A. Weisheit, S. Barcikowski,J. H. Schleifenbaum, and B. Gökce, “Depositing laser-generated nanoparticles onpowders for additive manufacturing of oxide dispersed strengthened alloy partsvia laser metal deposition,” Jpn. J. Appl. Phys. 57, 040310 (2018).54B. E. Carroll, T. A. Palmer, and A. M. Beese, “Anisotropic tensile behavior ofTi–6Al–4V components fabricated with directed energy deposition additivemanufacturing,” Acta Mater. 87, 309–320 (2015).55N. Yoshigai, K. Kudo, F. Tsumori, T. Osada, and H. Miura, “Anisotropicmechanical properties of Ni-base superalloy compacts by direct laser formingtechnology,” J. Jpn. Soc. Powder Powder Metallur. 63, 427–433 (2016).56M. L. M. Sistiaga, R. Mertens, B. Vrancken, X. Wang, B. Van Hooreweder,J.-P. Kruth, and J. Van Humbeeck, “Changing the alloy composition of Al7075

for better processability by selective laser melting,” J. Mater. Process. Technol.238, 437–445 (2016).57T. Vilaro, C. Colin, and J.-D. Bartout, “As-fabricated and heat-treated micro-structures of the Ti-6Al-4V alloy processed by selective laser melting,” Metall.Mater. Trans. A 42, 3190–3199 (2011).58C. Qiu, N. J. Adkins, and M. M. Attallah, “Microstructure and tensile proper-ties of selectively laser-melted and of HIPed laser-melted Ti–6Al–4V,” Mater.Sci. Eng. A 578, 230–239 (2013).59C. Qiu, S. Yue, N. J. Adkins, M. Ward, H. Hassanin, P. D. Lee, P. J. Withers,and M. M. Attallah, “Influence of processing conditions on strut structure andcompressive properties of cellular lattice structures fabricated by selective lasermelting,” Mater. Sci. Eng. A 628, 188–197 (2015).60C. Panwisawas, C. Qiu, Y. Sovani, J. Brooks, M. Attallah, and H. Basoalto, “Onthe role of thermal fluid dynamics into the evolution of porosity during selectivelaser melting,” Scr. Mater. 105, 14–17 (2015).61H. Gong, K. Rafi, H. Gu, T. Starr, and B. Stucker, “Analysis of defect genera-tion in Ti–6Al–4V parts made using powder bed fusion additive manufacturingprocesses,” Addit. Manuf. 1, 87–98 (2014).62H. Gu, H. Gong, D. Pal, K. Rafi, T. Starr, and B. Stucker, presented in2013 Solid Freeform Fabrication Symposium, Austin, TX, 12–14 August2013.63B. Zhang, Y. Li, and Q. Bai, “Defect formation mechanisms in selective lasermelting: A review,” Chin. J. Mech. Eng. 30, 515–527 (2017).64G. Kasperovich, J. Haubrich, J. Gussone, and G. Requena, “Correlationbetween porosity and processing parameters in TiAl6V4 produced by selectivelaser melting,” Mater. Des. 105, 160–170 (2016).65H. Stoffregen, J. Fischer, C. Siedelhofer, and E. Abele, presented in SolidFreeform Fabrication Proceedings, Austin, TX, 8–10 August 2011.66N. T. Aboulkhair, N. M. Everitt, I. Ashcroft, and C. Tuck, “Reducing porosityin AlSi10Mg parts processed by selective laser melting,” Addit. Manuf. 1, 77–86(2014).67A. Ahmed, M. Wahab, A. Raus, K. Kamarudin, Q. Bakhsh, and D. Ali,“Effects of selective laser melting parameters on relative density of AlSi10Mg,”Int. J. Eng. Technol. 8, 2552–2557 (2017).68M. Ma, Z. Wang, M. Gao, and X. Zeng, “Layer thickness dependence of per-formance in high-power selective laser melting of 1Cr18Ni9Ti stainless steel,”J. Mater. Process. Technol. 215, 142–150 (2015).69B. AlMangour, D. Grzesiak, and J. Yang, presented in the 26th AnnualInternational Solid Freeform Fabrication Symposium, Austin, TX, 8–10 August2016.70R. Morgan, A. Papworth, C. Sutcliffe, P. Fox, and W. O’neill, “High densitynet shape components by direct laser re-melting of single-phase powders,”J. Mater. Sci. 37, 3093–3100 (2002).71W. Meiners, K. Wissenbach, and R. Propawe, presented in Proceedings of theLANE, Erlangen, 23–26 September 1997.72D. Hagedorn-Hansen, M. Bezuidenhout, D. Dimitrov, and T. Oosthuizen,“The effects of selective laser melting scan strategies on deviation of hybridparts,” S. Afr. J. Ind. Eng. 28, 200–212 (2017).73G. Casalino, S. Campanelli, N. Contuzzi, and A. Ludovico, “Experimentalinvestigation and statistical optimisation of the selective laser melting process ofa maraging steel,” Opt. Laser Technol. 65, 151–158 (2015).74Y. Lu, S. Wu, Y. Gan, T. Huang, C. Yang, L. Junjie, and J. Lin, “Study on themicrostructure, mechanical property and residual stress of SLM Inconel-718alloy manufactured by differing island scanning strategy,” Opt. Laser Technol.75, 197–206 (2015).75N. Read, W. Wang, K. Essa, and M. M. Attallah, “Selective laser melting ofAlSi10Mg alloy: Process optimisation and mechanical properties development,”Mater. Des. 65, 417–424 (2015).76H. Niu and I. Chang, “Selective laser sintering of gas atomized M2 high speedsteel powder,” J. Mater. Sci. 35, 31–38 (2000).77S. Wang, Y. Liu, W. Shi, B. Qi, J. Yang, F. Zhang, D. Han, and Y. Ma,“Research on high layer thickness fabricated of 316L by selective laser melting,”Materials 10, pii: E1055 (2017).

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-17

© 2019 Laser Institute of America

78R. Li, Y. Shi, Z. Wang, L. Wang, J. Liu, and W. Jiang, “Densification behaviorof gas and water atomized 316L stainless steel powder during selective lasermelting,” Appl. Surf. Sci. 256, 4350–4356 (2010).79K. Abd-Elghany and D. Bourell, “Property evaluation of 304L stainless steelfabricated by selective laser melting,” Rapid Prototyping J. 18, 420–428 (2012).80P. Jerrard, L. Hao, and K. Evans, “Experimental investigation into selectivelaser melting of austenitic and martensitic stainless steel powder mixtures,” Proc.Inst. Mech. Eng. B J. Eng. Manuf. 223, 1409–1416 (2009).81A. Yadollahi, N. Shamsaei, S. M. Thompson, and D. W. Seely, “Effects ofprocess time interval and heat treatment on the mechanical and microstructuralproperties of direct laser deposited 316L stainless steel,” Mater. Sci. Eng. A 644,171–183 (2015).82L. Facchini, N. Vicente, I. Lonardelli, E. Magalini, P. Robotti, and A. Molinari,“Metastable austenite in 17–4 precipitation-hardening stainless steel produced byselective laser melting,” Adv. Eng. Mater. 12, 184–188 (2010).83T. LeBrun, T. Nakamoto, K. Horikawa, and H. Kobayashi, “Effect of retainedaustenite on subsequent thermal processing and resultant mechanical propertiesof selective laser melted 17–4 PH stainless steel,” Mater. Des. 81, 44–53 (2015).84K. Kempen, E. Yasa, L. Thijs, J.-P. Kruth, and J. Van Humbeeck,“Microstructure and mechanical properties of selective laser melted 18Ni-300steel,” Phys. Procedia 12, 255–263 (2011).85P. Krakhmalev, I. Yadroitsava, G. Fredriksson, and I. Yadroitsev, “In situ heattreatment in selective laser melted martensitic AISI 420 stainless steels,” Mater.Des. 87, 380–385 (2015).86L. Murr, S. Quinones, S. Gaytan, M. Lopez, A. Rodela, E. Martinez,D. Hernandez, E. Martinez, F. Medina, and R. Wicker, “Microstructure andmechanical behavior of Ti–6Al–4V produced by rapid-layer manufacturing, forbiomedical applications,” J. Mech. Behav. Biomed. Mater. 2, 20–32 (2009).87S. Chen, J. Huang, C. Pan, C. Lin, T. Yang, Y. Huang, C. Ou, L. Chen, D. Lin,and H. Lin, “Microstructure and mechanical properties of open-cell porousTi-6Al-4V fabricated by selective laser melting,” J. Alloys Compd. 713, 248–254(2017).88R. Mulay, J. Moore, J. Florando, N. Barton, and M. Kumar, “Microstructureand mechanical properties of Ti–6Al–4V: Mill-annealed versus direct metal lasermelted alloys,” Mater. Sci. Eng. A 666, 43–47 (2016).89W. Xu, M. Brandt, S. Sun, J. Elambasseril, Q. Liu, K. Latham, K. Xia, andM. Qian, “Additive manufacturing of strong and ductile Ti–6Al–4V by selectivelaser melting via in situ martensite decomposition,” Acta Mater. 85, 74–84(2015).90Q. Wang, C. Han, T. Choma, Q. Wei, C. Yan, B. Song, and Y. Shi, “Effect ofNb content on microstructure, property and in vitro apatite-forming capabilityof Ti-Nb alloys fabricated via selective laser melting,” Mater. Des. 126, 268–277(2017).91E. Chlebus, B. Kuźnicka, T. Kurzynowski, and B. Dybała, “Microstructure andmechanical behaviour of Ti–6Al–7Nb alloy produced by selective laser melting,”Mater. Charact. 62, 488–495 (2011).92W. Li, J. Liu, Y. Zhou, S. Li, S. Wen, Q. Wei, C. Yan, and Y. Shi, “Effect oflaser scanning speed on a Ti-45Al-2Cr-5Nb alloy processed by selective lasermelting: Microstructure, phase and mechanical properties,” J. Alloys Compd.688, 626–636 (2016).93Y. Idell, L. E. Levine, A. J. Allen, F. Zhang, C. E. Campbell, G. Olson, J. Gong,D. Snyder, and H. Deutchman, “Unexpected δ-phase formation in additive-manufactured Ni-based superalloy,” J. Miner. Met. Mater. Soc. 68, 950–959(2016).94Q. Jia and D. Gu, “Selective laser melting additive manufacturing of Inconel718 superalloy parts: Densification, microstructure and properties,” J. AlloysCompd. 585, 713–721 (2014).95M. Tang, P. C. Pistorius, S. Narra, and J. L. Beuth, “Rapid solidification:Selective laser melting of AlSi10Mg,” J. Miner. Met. Mater. Soc. 68, 960–966(2016).96S. Sun, L. Zheng, Y. Liu, J. Liu, and H. Zhang, “Selective laser melting ofAl-Fe-V-Si heat-resistant aluminum alloy powder: Modeling and experiments,”Int. J. Adv. Manuf. Technol. 80, 1787–1797 (2015).

97P. Ma, Y. Jia, K. G. Prashanth, S. Scudino, Z. Yu, and J. Eckert,“Microstructure and phase formation in Al–20Si–5Fe–3Cu–1Mg synthesized byselective laser melting,” J. Alloys Compd. 657, 430–435 (2016).98C. Shuai, Y. Yang, P. Wu, X. Lin, Y. Liu, Y. Zhou, P. Feng, X. Liu, and S. Peng,“Laser rapid solidification improves corrosion behavior of Mg-Zn-Zr alloy,”J. Alloys Compd. 691, 961–969 (2017).99T. Maeshima and K. Oh-ishi, “Solute clustering and supersaturated solid solu-tion of AlSi10Mg alloy fabricated by selective laser melting,” Heliyon 5, e01186(2019).100H. Zhang, D. Gu, J. Yang, D. Dai, T. Zhao, C. Hong, A. Gasser, andR. Poprawe, “Selective laser melting of rare earth element Sc modified aluminumalloy: Thermodynamics of precipitation behavior and its influence on mechani-cal properties,” Addit. Manuf. 23, 1–12 (2018).101R. Li, H. Chen, H. Zhu, M. Wang, C. Chen, and T. Yuan, “Effect of agingtreatment on the microstructure and mechanical properties of Al-3.02 Mg-0.2Sc-0.1 Zr alloy printed by selective laser melting,” Mater. Des. 168, 107668(2019).102V. Gunenthiram, P. Peyre, M. Schneider, M. Dal, F. Coste, and R. Fabbro,“Analysis of laser–melt pool–powder bed interaction during the selective lasermelting of a stainless steel,” J. Laser Appl. 29, 022303 (2017).103D. Gu and Y. Shen, “Balling phenomena during direct laser sintering of multi-component Cu-based metal powder,” J. Alloys Compd. 432, 163–166 (2007).104D. Gu, Y. Shen, J. Yang, and Y. Wang, “Effects of processing parameters ondirect laser sintering of multicomponent Cu based metal powder,” Mater. Sci.Technol. 22, 1449–1455 (2006).105H. Gong, H. Gu, K. Zeng, J. Dilip, D. Pal, B. Stucker, D. Christiansen,J. Beuth, and J. J. Lewandowski, presented in Solid Freeform FabricationSymposium, Austin, TX, 4–6 August 2014.106H. Niu and I. Chang, “Instability of scan tracks of selective laser sintering ofhigh speed steel powder,” Scr. Mater. 41, 1229–1234 (1999).107N. K. Tolochko, S. E. Mozzharov, I. A. Yadroitsev, T. Laoui, L. Froyen,V. I. Titov, and M. B. Ignatiev, “Balling processes during selective laser treatmentof powders,” Rapid Prototyping J. 10, 78–87 (2004).108X. Zhou, X. Liu, D. Zhang, Z. Shen, and W. Liu, “Balling phenomena inselective laser melted tungsten,” J. Mater. Process. Technol. 222, 33–42 (2015).109S. Das, “Physical aspects of process control in selective laser sintering ofmetals,” Adv. Eng. Mater. 5, 701–711 (2003).110D. Gu and Y. Shen, “Balling phenomena in direct laser sintering of stainlesssteel powder: Metallurgical mechanisms and control methods,” Mater. Des. 30,2903–2910 (2009).111R. Li, J. Liu, Y. Shi, L. Wang, and W. Jiang, “Balling behavior of stainless steeland nickel powder during selective laser melting process,” Int. J. Adv. Manuf.Technol. 59, 1025–1035 (2012).112Y. F. Shen, D. Gu, and Y. Pan, “Balling process in selective laser sintering 316stainless steel powder,” Key. Eng. Mater. 315–316, 357–360 (2006).113X. Cao, W. Wallace, C. Poon, and J.-P. Immarigeon, “Research and progressin laser welding of wrought aluminum alloys. I. Laser welding processes,” Mater.Manuf. Process. 18, 1–22 (2003).114J. Chen, X. Lin, T. Wang, H. Yang, and W. Huang, “The hot cracking mecha-nism of 316 L stainless steel cladding in rapid laser forming process,” Rare MetalMater. Eng. 32, 183–186 (2003).115L. N. Carter, K. Essa, and M. M. Attallah, “Optimisation of selective lasermelting for a high temperature Ni-superalloy,” Rapid Prototyping J. 21, 423–432(2015).116L. N. Carter, M. M. Attallah, and R. C. Reed, “Laser powder bed fabricationof nickel-base superalloys: Influence of parameters; characterisation, quantifica-tion and mitigation of cracking,” Superalloys 2012, 577–586 (2012).117D. Wang, C. Yu, J. Ma, W. Liu, and Z. Shen, “Densification and crack sup-pression in selective laser melting of pure molybdenum,” Mater. Des. 129, 44–52(2017).118D. Tomus, T. Jarvis, X. Wu, J. Mei, P. Rometsch, E. Herny, J.-F. Rideau, andS. Vaillant, “Controlling the microstructure of Hastelloy-X components manu-factured by selective laser melting,” Phys. Procedia 41, 823–827 (2013).

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-18

© 2019 Laser Institute of America

119N. J. Harrison, I. Todd, and K. Mumtaz, “Reduction of micro-cracking innickel superalloys processed by selective laser melting: A fundamental alloydesign approach,” Acta Mater. 94, 59–68 (2015).120M. Cloots, P. J. Uggowitzer, and K. Wegener, “Investigations on the micro-structure and crack formation of IN738LC samples processed by selective lasermelting using Gaussian and doughnut profiles,” Mater. Des. 89, 770–784 (2016).121M. Agarwala, D. Bourell, J. Beaman, H. Marcus, and J. Barlow, “Direct selec-tive laser sintering of metals,” Rapid Prototyping J. 1, 26–36 (1995).122K. Kempen, L. Thijs, B. Vrancken, S. Buls, J. Van Humbeeck, and J. Kruth, pre-sented in Proceedings of the 24th International Solid Freeform FabricationSymposium. Laboratory for Freeform Fabrication, Austin, TX, 12–14 August 2013.123R. Mertens, S. Dadbakhsh, J. Van Humbeeck, and J.-P. Kruth, “Applicationof base plate preheating during selective laser melting,” Procedia CIRP 74, 5–11(2018).124H. Zhang, H. Zhu, X. Nie, J. Yin, Z. Hu, and X. Zeng, “Effect of zirconiumaddition on crack, microstructure and mechanical behavior of selective lasermelted Al-Cu-Mg alloy,” Scr. Mater. 134, 6–10 (2017).125T. M. Rodgers, J. D. Madison, V. Tikare, and M. C. Maguire, “Predictingmesoscale microstructural evolution in electron beam welding,” J. Miner. Met.Mater. Soc. 68, 1419–1426 (2016).126T. M. Rodgers, J. D. Madison, and V. Tikare, “Simulation of metal additivemanufacturing microstructures using kinetic Monte Carlo,” Comput. Mater. Sci.135, 78–89 (2017).127A. Choudhury, K. Reuther, E. Wesner, A. August, B. Nestler, andM. Rettenmayr, “Comparison of phase-field and cellular automaton models fordendritic solidification in Al–Cu alloy,” Comput. Mater. Sci. 55, 263–268 (2012).128A. Zinoviev, O. Zinovieva, V. Ploshikhin, V. Romanova, and R. Balokhonov,“Evolution of grain structure during laser additive manufacturing. Simulation bya cellular automata method,” Mater. Des. 106, 321–329 (2016).129H. Yin and S. Felicelli, “Dendrite growth simulation during solidification inthe LENS process,” Acta Mater. 58, 1455–1465 (2010).130P. Nie, O. Ojo, and Z. Li, “Numerical modeling of microstructure evolutionduring laser additive manufacturing of a nickel-based superalloy,” Acta Mater.77, 85–95 (2014).131O. Lopez-Botello, U. Martinez-Hernandez, J. Ramírez, C. Pinna, andK. Mumtaz, “Two-dimensional simulation of grain structure growth withinselective laser melted AA-2024,” Mater. Des. 113, 369–376 (2017).132A. Rai, M. Markl, and C. Körner, “A coupled cellular automaton—LatticeBoltzmann model for grain structure simulation during additive manufacturing,”Comput. Mater. Sci. 124, 37–48 (2016).133A. Rai, H. Helmer, and C. Körner, “Simulation of grain structure evolutionduring powder bed based additive manufacturing,” Addit. Manuf. 13, 124–134(2017).134W. Chen, “A computational multiscale integrated model for microstructuresimulation in selective laser melting,” Thesis, Hong Kong University of Scienceand Technology, 2016.135W. Tan, N. S. Bailey, and Y. C. Shin, “A novel integrated model combiningcellular automata and phase field methods for microstructure evolution duringsolidification of multi-component and multi-phase alloys,” Comput. Mater. Sci.50, 2573–2585 (2011).136Y. Cao and J. Choi, “Solidification microstructure evolution model for lasercladding process,” J. Heat Transf. 129, 852–863 (2007).137T. Keller, G. Lindwall, S. Ghosh, L. Ma, B. M. Lane, F. Zhang, U. R. Kattner,E. A. Lass, J. C. Heigel, and Y. Idell, “Application of finite element, phase-field,and calphad-based methods to additive manufacturing of Ni-based superalloys,”Acta Mater. 139, 244–253 (2017).

138V. Fallah, M. Amoorezaei, N. Provatas, S. Corbin, and A. Khajepour,“Phase-field simulation of solidification morphology in laser powder depositionof Ti–Nb alloys,” Acta Mater. 60, 1633–1646 (2012).139J. Kundin, L. Mushongera, and H. Emmerich, “Phase-field modeling ofmicrostructure formation during rapid solidification in Inconel 718 superalloy,”Acta Mater. 95, 343–356 (2015).140R. Acharya, J. A. Sharon, and A. Staroselsky, “Prediction of microstructure inlaser powder bed fusion process,” Acta Mater. 124, 360–371 (2017).141A. Popovich, V. S. Sufiiarov, E. Borisov, I. Polozov, and D. Masaylo, “Designand manufacturing of tailored microstructure with selective laser melting,”Mater. Phys. Mech. 38, 1–10 (2018).142J. Yin, G. Peng, C. Chen, J. Yang, H. Zhu, L. Ke, Z. Wang, D. Wang, M. Ma,and G. Wang, “Thermal behavior and grain growth orientation duringselective laser melting of Ti-6Al-4 V alloy,” J. Mater. Process. Technol. 260,57–65 (2018).143S. Chang, L. Li, L. Lu, and J. Y. H. Fuh, “Selective laser sintering of poroussilica enabled by carbon additive,” Materials 10, 1313 (2017).144I. A. van Hengel, M. Riool, L. E. Fratila-Apachitei, J. Witte-Bouma, E. Farrell,A. A. Zadpoor, S. A. Zaat, and I. Apachitei, “Selective laser melting porousmetallic implants with immobilized silver nanoparticles kill and prevent biofilmformation by methicillin-resistant Staphylococcus aureus,” Biomaterials 140,1–15 (2017).145D. Pattanayak, T. Matsushita, H. Takadama, A. Fukuda, M. Takemoto,S. Fujibayashi, K. Sasaki, N. Nishida, T. Nakamura, and T. Kokubo, “Fabricationof bioactive porous Ti metal with structure similar to human cancellous bone byselective laser melting,” Bioceramics Dev. Appl. 1, 1–3 (2011).146M. Yan, X. Tian, G. Peng, Y. Cao, and D. Li, “Hierarchically porous materialsprepared by selective laser sintering,” Mater. Des. 135, 62–68 (2017).147T. L. Starr, K. Rafi, B. Stucker, and C. M. Scherzer, “Controlling phase com-position in selective laser melted stainless steels,” Proc. of SFF Symposium439–446 (2012).148K. P. Davidson and S. Singamneni, “Magnetic characterization of selectivelaser-melted Saf 2507 duplex stainless steel,” J. Miner. Met. Mater. Soc. 69,569–574 (2017).149J. Akram, P. Chalavadi, D. Pal, and B. Stucker, “Understanding grain evolutionin additive manufacturing through modeling,” Addit. Manuf. 21, 255–268 (2018).150J. A. Koepf, M. R. Gotterbarm, M. Markl, and C. Körner, “3D multi-layergrain structure simulation of powder bed fusion additive manufacturing,” ActaMater. 152, 119–126 (2018).151S. Chen, Y. Xu, and Y. Jiao, “A hybrid finite-element and cellular-automatonframework for modeling 3D microstructure of Ti–6Al–4V alloy during solid–solid phase transformation in additive manufacturing,” Model. Simul. Mater. Sci.Eng. 26, 045011 (2018).152H.-C. Tran, Y.-L. Lo, and M.-H. Huang, “Analysis of scattering and absorp-tion characteristics of metal powder layer for selective laser sintering,” IEEEASME Trans. Mech. 22, 1807–1817 (2017).153M. Xia, D. Gu, G. Yu, D. Dai, H. Chen, and Q. Shi, “Porosity evolution andits thermodynamic mechanism of randomly packed powder-bed during selectivelaser melting of Inconel 718 alloy,” Int. J. Mach. Tools Manuf. 116, 96–106(2017).154S. Rashidi, J. A. Esfahani, and N. Karimi, “Porous materials in buildingenergy technologies—A review of the applications, modelling and experiments,”Renew. Sust. Energy Rev. 91, 229–247 (2018).155F. Calignano, G. Cattano, L. Iuliano, and D. Manfredi, presented inInternational Conference on Additive Manufacturing in Products andApplications, ETH Zürich, Switzerland, 13–15 September 2017.

Journal ofLaser Applications REVIEW scitation.org/journal/jla

J. Laser Appl. 31, 031201 (2019); doi: 10.2351/1.5085206 31, 031201-19

© 2019 Laser Institute of America