See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327608577

A Review of Selective Laser Melting of Aluminum Alloys: Processing,

Microstructure, Property and Developing Trends

Article in Journal of Materials Science and Technology -Shenyang- · February 2019

DOI: 10.1016/j.jmst.2018.09.004

CITATIONS READS

373 8,243

5 authors, including:

Jinliang Zhang Bo Song

Huazhong University of Science and Technology Huazhong University of Science and Technology
5 PUBLICATIONS 543 CITATIONS 102 PUBLICATIONS 3,854 CITATIONS

SEE PROFILE SEE PROFILE

Qingsong Wei Yusheng Shi

Huazhong University of Science and Technology Huazhong University of Science and Technology
193 PUBLICATIONS 5,051 CITATIONS 405 PUBLICATIONS 9,424 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Structured light 3D measurement View project

rapid development of drip emitters View project

All content following this page was uploaded by Bo Song on 09 January 2019.

The user has requested enhancement of the downloaded file.

 Journal of Materials Science & Technology 35 (2019) 270–284

Contents lists available at ScienceDirect

Journal of Materials Science & Technology

journal homepage: www.jmst.org

Invited Review

A review of selective laser melting of aluminum alloys: Processing,

microstructure, property and developing trends
Jinliang Zhang a , Bo Song a,∗ , Qingsong Wei a , Dave Bourell b , Yusheng Shi a
a
State Key Laboratory of Materials Processing and Die & Mould Technology, School of Materials Science and Engineering, Huazhong University of Science
and Technology, Wuhan, 430074, China
b
Dave Bourell Laboratory for Freeform Fabrication, Mechanical Engineering Department, The University of Texas at Austin, Austin, TX, 78712, USA

a r t i c l e i n f o a b s t r a c t

Article history: Selective laser melting (SLM) is an attractive rapid prototyping technology for the fabrication of metallic
Received 29 March 2018 components with complex structure and high performance. Aluminum alloy, one of the most perva-
Received in revised form 26 April 2018 sive structural materials, is well known for high specific strength and good corrosion resistance. But the
Accepted 22 May 2018
poor laser formability of aluminum alloy restricts its application. There are problems such as limited
Available online 12 September 2018
processable materials, immature process conditions and metallurgical defects on SLM processing alu-
minum alloys. Some efforts have been made to solve the above problems. This paper discusses the current
Keywords:
research status both related to the scientific understanding and technology applications. The paper begins
Selective laser melting
Aluminum alloy with a brief introduction of basic concepts of aluminum alloys and technology characterization of laser
Metallurgical defects selective melting. In addition, solidification theory of SLM process and formation mechanism of metal-
Mechanical properties lurgical defects are discussed. Then, the current research status of microstructure, properties and heat
Heat treatment treatment of SLM processing aluminum alloys is systematically reviewed respectively. Lastly, a future
Developing trend outlook is given at the end of this review paper.
© 2018 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science &
Technology.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
2. Solidification theory and metallurgical defects of laser selective melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
2.1. Solidification theory of laser selective melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
2.2. Formation mechanism and control methods of metallurgical defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
2.2.1. Balling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
2.2.2. Porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
2.2.3. Residual stress and cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
2.2.4. Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
2.2.5. Loss of alloying elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
3. Solidification theory and metallurgical defects of laser selective melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
3.1. Al-Si series alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
3.1.1. Microstructure characteristics of SLMed Al-Si alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
3.1.2. Mechanical properties of SLMed Al-Si alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
3.1.3. Effect of heat treatment on the microstructure and properties of SLMed Al-Si alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
3.2. Other aluminum alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
4. Prospective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

∗ Corresponding author.
E-mail address: bosong@hust.edu.cn (B. Song).

https://doi.org/10.1016/j.jmst.2018.09.004
1005-0302/© 2018 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science & Technology.
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 271

1. Introduction

Aluminum alloys, the most widely utilized metal structural

materials, after iron and steel, have the potential for application
and development in the aviation, aerospace, automobile, naval,
weapons and power electronics fields due to their low density, high
specific strength, easy processing, and good corrosion resistance as
well as their excellent electric and thermal conductivity [1–3].
Based on their composition, microstructure and process charac-
teristics, aluminum alloys can be categorized into cast aluminum
alloys and wrought aluminum alloys. In general, the alloying ele-
ment content of cast aluminum alloys is 10%–12%, while that of
wrought aluminum alloys is 1%–2% (although, in some cases, it can Fig. 2. Schematic diagram of selective laser melting [14].
be as high as 6%–8%) [4]. According to whether the alloy responds
to heat treatment by precipitation hardening, aluminum alloys can gral forming of complex structural components not only reduces
be further divided into heat treatable and non-heat treatable alu- the time and tooling for fabrication and assembly of small and
minum alloys, as shown in Fig. 1. The properties of commercial medium size components but also decreases the weight and stress
aluminum alloys are shown in Table 1. 2XXX (Al-Cu or Al-Cu-Mg), concentrations normally associated with welding and other joining
6XXX (Al-Mg-Si) and 7XXX (Al-Zn-Mg) series alloys belong to heat approaches [11–13]. The production of aluminum alloy parts with
treatable alloys, whose strength can be enhanced by various heat diverse structures, high dimensional accuracy and near net shape
treatment processes [4–6]. The 1XXX, 3XXX, 4XXX and 5XXX series will be a major research and development objective in the future.
alloys are non-heat treatable aluminum alloys that can be strength- Selective laser melting (SLM) is considered one of the most
ened by solid solution and mechanical deformation processes [4,5]. promising additive manufacturing (AM) technologies, whose pro-
At present, aluminum alloy structural parts are mainly manufac- cessing schematic diagram is shown in Fig. 2. SLM utilizes a
tured by traditional methods such as casting, forging, extrusion and high-energy laser beam to completely melt metal powder in a pro-
powder metallurgy [7]. Although aluminum alloy products made by tective atmosphere along the laser path, and this molten metal
the above traditional processes have been extensively used, there rapidly solidifies [14]. By repeating this step and overlapping layer
are still many problems in the process of production and their by layer, a three-dimensional component is eventually formed.
applications. First, the low cooling rate in the casting processes Such a layerwise approach has a unique advantage in the integrated
makes the microstructure coarse, and many defects exist, such formation of complex structured and thin-walled components
as offset defects, shrinkage porosity, slag inclusion and element [15,16]. The process of welding and riveting is eliminated, and no
segregation in cast aluminum alloys, that lower the mechanical mold is needed, thus shortening the design and production times
properties of the parts [8–10]. Second, the preparation and form- [17]. Recycling the metallic powder increases the feedstock utiliza-
ing process of high performance aluminum alloy components are tion rate and reduces the production cost. In addition, the SLMed
separated, which causes a long process chain with limited flexi- parts have excellent quality and performance. The laser interacts
bility. In addition, with the modern industrial developments, the with the metallic powder to form a small molten pool on a scale
structure and performance requirements of aluminum alloy parts of approximately 100 ␮m [18]. The cooling rate of the molten
are continuously increasing. For example, to meet the engineering pool reaches 103 -108 K/s due to the rapid movement of the laser
requirements of high thermal conductivity, light weight and high at 100–1000 mm/s [19]. Such a rapid cooling rate inhibits grain
load carrying capacity, the thermal protection system of aerospace growth and segregation of the alloying elements. Together with
vehicle engines commonly uses a lattice or cellular structure. Inte- the stirring action of Marangoni flow [20] in the molten pool, a fine

Fig. 1. Classification of cast aluminum alloys and wrought aluminum alloys: (a) cast aluminium alloys. (b) wrought aluminium alloys. Non-heat treatable aluminium alloys
are in blue and heat treatable aluminum alloys are in red.
272 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

Table 1
Properties of aluminium series alloys.

Series Types elements Performance

1××× non-heat treatable Pure Al (≥99.00%) Low strength, good corrosion resistance and conductivity, easy processing
2××× heat treatable Al-Cu/Al-Cu-Mg Hard-aluminium alloy. High strength, good heat resistance, poor corrosion resistance
3××× non-heat treatable Al-Mn Antirust aluminium alloy. Low strength, cold-working-hardening, good plasticity and weldability
4××× non-heat treatable Al-Si High silicon, low melting point, good weldability, good heat and wear resistance
5××× non-heat treatable Al-Mg High magnesium, good corrosion resistance and weldability
6××× heat treatable Al-Mg-Si Medium strength, good formability, weldability and machinability
7××× heat treatable Al-Zn-Mg Very high strength, cannot be welded, poor corrosion resistance
8××× / Other elements /
9××× / Spare alloys /

Fig. 3. Diagram of interaction between laser and powder bed [28].

uniform microstructure is formed that significantly improves the

strength and toughness. The non-equilibrium solidification process
increases the solid solution limit of the alloy elements in the matrix,
and new metastable or even amorphous phases may be generated Fig. 4. Epitaxial solidification and competitive growth of welded molten zone [37].
[21–23]. However, the surface quality and dimensional accuracy of
the SLMed parts are typically not sufficient to meet the demands
of industrial production, so post-treatment processes, such as sur- material and growth toward the welding centerline [31,32]. Sim-
face treatment and machining, are required, which increase the ilarly, under the condition of melt superheating, homogeneous
production time [24]. On the other hand, the large temperature nucleation hardly occurs in the SLM process, and nucleation is
gradients and complex heat transfer due to the cyclic processing always enhanced at the solid-liquid interface between the grain
of the laser beam result in the directional growth of grains, so the surface of the parent metal and the liquid metal [33,34]. Once crys-
microstructure and properties of the alloy tend to be anisotropic tallization is initiated at the boundary of the molten pool, grains
[25]. The quality of the SLMed parts depends on the selection of the continue to grow toward the interior of the melt in the form of
technological parameters, such as the laser power, scanning speed, columnar crystals [34]. The crystallization orientation most con-
scanning spacing and layer thickness. These process parameters ducive to the grain growth of aluminum is <001> [35]. Grains
are influenced by the material characteristics, powder fluidity, par- growing along the <001> orientation and perpendicular to the
ticle size/shape/distribution and the type and spot size of the laser isotherms of the molten pool boundary, where the heat dissi-
[26,27]. The use of improper process parameters may cause balling, pates fastest, can keep growing to the interior of the molten pool.
pores, cracks and low density. However, grains growing along the <001> orientation but in the
direction of the boundary isotherm are inhibited and grow only a
2. Solidification theory and metallurgical defects of laser short distance [36,37]. This phenomenon is known as preferential
selective melting growth.
The cooling and solidification mode of the molten pool mainly
2.1. Solidification theory of laser selective melting depends on the laser energy density and the interaction period of
the laser and the material [38]. The temperature gradient is given as
When the laser contacts the metal powder, a series of complex G=dT/dx, the solidification rate is given as R=dx/dt, and the cooling
physical and chemical phenomena occur during the rapid melting rate is given as dT/dt = G·R. A higher cooling rate G·R can improve
and solidification processes, such as the absorption and scattering the undercooling of the melt and refine the grains [39,40]. When
of laser energy, heat transfer, phase transition and melt flow in the the ratio of the temperature gradient and the solidification rate
molten pool [28,29], as described in Fig. 3. The thermodynamic and G/R increases, the crystal morphology changes from dendritic to
kinetic behavior of the molten pool can be changed by adjusting cellular and finally to planar crystal [19,41]. The scale and the type
the processing parameters. The size and shape of the grains and of microstructure are controlled by G·R and G/R, respectively.
the content and composition of the phases are controlled to obtain The microstructure undergoes a morphological evolution from
the desired microstructure and mechanical properties [30]. Like- planar to cellular, cellular dendrite and an equiaxed crystal with
wise, to explain the microstructure and property evolution laws of the successive decreases of G·R [19,41]. Therefore, the solidification
SLMed parts, the thermal behavior and solidification theory under structure often has a variety of grain types in the SLM solidification
the processing conditions must be clearly defined. process. As shown in Fig. 5, when the temperature gradient G1 at
As shown in Fig. 4, the growth of grains in fusion welding is the front of the solid-liquid interface is high, the actual temper-
considered to be initiated by the epitaxial solidification of the base ature in the liquid phase is higher than the liquidus temperature
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 273

Fig. 5. Conditions of constitutional supercooling.

TL , and a constitutional supercooling zone will not appear. The

nuclei growing forward are remelted by the superheated liquid
metal, and the solidification interface is smooth with no solute
segregation in the grains. Thus, a planar microstructure is formed.
On the condition that the temperature gradient G2 of the liquid
phase becomes lower, a constitutional supercooling zone may be
formed. Many parallel small grains form at the solidification inter-
face and protrude into the supercooled liquid due to the unstable Fig. 6. Grain map measured by electron backscattered diffraction technology [44].

state of the planar crystal interface. The solute is expelled toward

the lateral subgrain boundaries, and the liquidus temperature of time and the temperatures of the melt and base material, which
the subgrain boundary decreases. As a result, a bundle of parallel induces low temperature gradient and cooling rate of the solidifica-
prismatic substructures with hexagonal cross sections are formed tion interface to cause a coarse microstructure [43,46]. In contrast,
in the grain. In this case, a cellular grain structure is formed. If the a low laser power and a high scanning velocity inhibit the growth of
temperature gradient G3 further decreases, the cellular crystalline grains. The cooling rate also has an important influence on the den-
microstructure can penetrate deep into the interior of the liquid drite arm spacing size [45]. The relationship between the dendrite
for a longer distance, and constitutional supercooling also occurs arm size and the cooling rate is revealed as:
in the transverse direction. Secondary dendrites may appear on the
=c·T −n (1)
primary dendrites, which is a characteristic of cellular dendrites.
If the temperature gradient G4 continues to decrease, the consti- where c is the constant of the alloy and n is the rate exponent. This
tutional supercooling zone further expands, and dendritic grains equation indicates that the dendrite size decreases with increas-
are formed. The contact surface of the secondary dendrites is the ing cooling rate. Additionally, the sub-boundary formation can be
boundary of the grains. The larger the solidification rate is, the affected by a fast cooling rate. As a result, the surface always has a
smaller the dendrite spacing is [42]. At the same time, nucleation higher hardness and a greater wear resistance value due to the finer
could also occur in the liquid to produce new grains, which finally precipitate phase particles being evenly distributed, finer dendrites
grow into equiaxed grains. Because of the Gaussian distribution and the occurrence of sub-boundaries compared with the core area.
of the laser energy, the temperature gradient and crystallization
rate, as well as the distribution of the undercooling, are different 2.2. Formation mechanism and control methods of metallurgical
in the different zones of the molten pool [43]. Therefore, various defects
grain morphologies might be expected to appear in the solidified
microstructure. At the molten pool boundary, it is hard to induce 2.2.1. Balling
constitutional supercooling resulting from the large temperature According to the minimum principle of surface energy, a liquid
gradient and slow solidification velocity, so planar grain forma- metal will shrink into a spherical shape driven by surface ten-
tion is most common. With the grains growing toward the melt sion when it is in poor contact with a substrate. This is termed
pool center, the temperature gradients decrease, and the grain the balling phenomenon [47]. The balling phenomenon results in a
growth rate, mass fraction of solute and constitutional supercooling rough surface on the solidified layer accompanied by a large num-
zone increase gradually. Correspondingly, the substructures in the ber of pores, which reduces the density and quality of the parts
columnar grains are cellular grains, cellular dendritic grains and and may even damage the powder roller or scraper, which fails the
dendritic grains, successively. When the grains grow toward the process [48,49].
center of the molten pool, equiaxed grains may form eventually Generally, the balling phenomenon can be attributed to the poor
[23]. The above theory has been confirmed by the literature [44], wettability and droplet splashing during processing. As shown in
as shown in Fig. 6. Fig. 7, the molten pool can be divided into an upper part consisting
The extent of the grain refinement depends on the cooling rate of the molten powder and a lower part consisting of the molten sub-
G·R. A higher cooling rate in the core of the molten pool results strate material [50]. The gas-liquid interface of the upper molten
in finer grains than those at the edge of the molten pool [45], pool tends to promote the balling phenomenon, while the lower
which can also be observed in Fig. 6. Therefore, the microstructure melt pool hinders the balling trend of the upper part. If there is
of SLMed alloys can be controlled by adjusting the technological enough melt in the lower molten pool, the balling tendency of the
parameters. A low laser scanning velocity and a high laser power upper melt pool will be completely suppressed [50,51]. Therefore,
can increase the laser energy density and the interaction time a high energy density can form sufficient liquid metal and helps
between the laser and material, thus increasing the solidification to mitigate the balling phenomenon. Moreover, a high tempera-
274 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

Fig. 7. Schematic diagram of balling phenomenon. (a) Less molten substrate material. (b) More molten substrate material.

always tend to exist in the minimum surface energy state, the gas
pores are generally spherical and fine [63,68,69]. The main source
of this type of porosity is the inert gas in the protective atmosphere
involved in the molten pool that is incorporated into the part via
surface turbulence, the gas in the powder particle gaps, or the vapor
of the low melting compounds in the alloy that fails to escape from
the molten pool [63,70].
Gu et al. [71] studied the effect of the shielding gas on the molten
pool dynamics, the evaporated material velocity field and the resul-
tant surface morphology by numerical simulations, as shown in
Fig. 9. It was found that when a He protective atmosphere was
used, the velocity field vector of the evaporated metal is in a hori-
zontal or downward direction, and the nonuniform recoil pressure
Fig. 8. Schematic diagram of droplet splash during SLM [14]. generated on the surface of the molten pool results in the occur-
rence of keyholing. When an Ar protective atmosphere is used, the
ture can lead to low viscosity and good fluidity of the liquid metal, upward velocity field vectors make it difficult to trap gases, so a uni-
thus enhancing the wettability of the melt and the solidified layers. form recoil pressure and a stable condition of the molten pool are
However, if the laser energy is too high, excessive liquid metal will formed, which leads to a relatively flat surface morphology. When a
lead not only to balling but also to deformation of the parts due to N2 protective atmosphere is applied, the metal vapor tends to flow
residual stress formation [52,53]. toward the laser-powder interaction region, resulting in material
Excessive temperature will also cause evaporation of the mate- stacking and poor surface quality. In the actual SLM processing of
rial. As shown in Fig. 8, the rapid formation of a gaseous phase AlSi12, the defects of the samples are fewer in atmospheres of Ar
produces a large recoil pressure in the molten pool, causing some and N2 . In an atmosphere of He, however, there are some pore clus-
melt to escape in the form of a metal jet [54]. The metal jet is bro- ters in the microstructure in local areas. Because these areas are
ken up into droplets by metallic vapor and the laser beam, finally relatively small, the overall density values of the SLMed samples
forming metal balls. In addition, due to the influence of the metal are not significantly affected, but lower mechanical properties are
vapor, the unmelted metal powder around the molten pool may be produced for the SLMed samples, especially the ductility [72]. The
dispersed and splashed sideways [14]. porosity of the fracture morphology under a He atmosphere was
Laser remelting may provide a means for the metal balls to significantly higher than those under Ar and N2 . Cracks will first
rewet the substrate, reducing the balling phenomenon, but it simul- form and expand rapidly in the pore concentration areas, resulting
taneously increases the manufacturing time [48,55,56]. Substrate in the fracture of the material.
preheating can improve the wettability between the melt and sub- In the previous section, it has been mentioned that if the balling
strate and thus restrain the shrinkage effect during solidification, phenomenon is severe on a certain layer, it will inevitably lead
resulting in good metallurgical bonding [57]. However, when the to a ‘ripple effect’ such that large amounts of balling and porosity
preheating temperature is too high, some powder particles may occur on the next layer. This results in the formation of metallur-
form agglomerates attached to the molten pool, causing an even gical defects and low density in the final part. Therefore, strategies
more severe balling phenomenon [58]. The preheating temperature to diminish balling by adjusting the process parameters, substrate
of aluminum alloys is commonly within the range of 50–200 ◦ C. preheating and remelting are equally applicable for reducing the
porosity. Sufficient liquid metal supply and long molten pool life-
time enhance complete filling of the pores during SLM processing.
2.2.2. Porosity
For aluminum alloys, the water absorption of aluminum powder is
The pores in the SLMed parts are divided into three types: fusion
strong, and the solubility values of hydrogen in solid and liquid alu-
errors, gas pores and shrinkage pores [59–61]. Fusion errors are
minum are different [73,74]. The powder material should therefore
attributed to insufficient laser energy density, which causes incom-
be sufficiently dried before processing to prevent the formation of
plete remelting of the last solidified layer and a poor metallurgical
hydrogen porosity during rapid solidification.
bond [62]. These pores are irregularly shaped and are common
at the interfaces of the layers. Their size and number are greatly
affected by the processing parameters. The gas entrapped in the 2.2.3. Residual stress and cracking
gaps of powder particles in the powder mass contributes to the Rapid melting and solidification during SLM manufacturing pro-
formation of fusion errors [63]. When laser scanning is progress- cessing lead to a higher cooling rate and temperature gradient in
ing, the gas escapes and thereby results in unstable scanning paths. the molten pool [19]. Owing to the continuous thermal cycling and
With the formation of these cavities, the fluid force in the molten complex physical/chemical reactions, an inhomogeneous heat dis-
pool tends to balance with the vapor pressure in the cavity, caus- tribution leads to the thermal expansion and contraction of the
ing the liquid metal to collapse and form periodic pores [64–66]. solidified structure [29]. Therefore, the SLM process is inevitably
Shrinkage porosity is mainly attributed to an insufficient supply of accompanied by high residual stress levels. High residual stress
liquid metal during the solidification process [62,67]. Since bubbles zones formed around the molten pool may lead to cracking, delami-
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 275

Fig. 9. Diagram of velocity field of evaporated metal. (a) He. (b) Ar. (c) N2 [71].

nation, fatigue failure and thermal deformation [75,76]. As a result, [95]. Crack initiation is more likely to occur. The addition of alloy-
the dimensional accuracy and mechanical properties of the parts ing elements, which are beneficial to narrowing the solidification
will be significantly affected [24]. Crack formation in aluminum temperature range, can modify the composition of the molten pool
alloys is even aggravated due to its high thermal conductivity, wide and prevent cracks [96]. For example, 5XXX series aluminum alloys
solidification temperature range, high thermal expansion coeffi- have a lower cracking tendency than the 2XXX, 6XXX and 7XXX
cient and severe solidification shrinkage [77]. series aluminum alloys in laser processing.
The stresses formed in the SLM process are mainly classified into Shiomil et al. [79] found that laser remelting and substrate pre-
thermal stresses and structural stresses [78]. The thermal stress is heating can effectively reduce the residual stress by 55% and 40%,
caused by the uneven heating of the laser and the resultant dif- respectively, during processing due to the reduction of the cooling
ferent amounts of thermal expansion and shrinkage deformation rate. The formation of cracks also depends on the choice of process
between the regions near to and away from the molten pool [79,80]. parameters. Only when the energy density is exactly the optimum
When phase transformation occurs in the SLM process, the material value or within the optimum range can a crack-free, dense SLMed
will undergo volume expansions and contractions arising from the part be manufactured. When the energy density is low, a disordered
different specific volume of phases, thereby resulting in structural solidification front of liquid metal and severe balling will lead to
stresses [81]. Thermal stress is the primary cause of crack initi- crack formation [67]. To diminish residual stress and crack forma-
ation in SLM [82]. When the stresses inside the part exceed its tion, sufficient liquid phase should be ensured to backfill the cracks
yield strength, either plastic deformation occurs, possibly result- and take up the solidification strain [97]. At extremely high energy
ing in part distortion, or cracks may form to relieve the stress [83]. density values, low liquid viscosity and long liquid lifetime will
Considering the complexity of the SLM process and the difficulty result in an increase of thermal stress [98]. Excessively rapid cooling
of experimental measurement, finite element simulation methods rates should also be avoided when setting the process parameters.
are often used to predict the distribution and evolution of residual Too high a cooling rate will accelerate the development of ther-
stress. mal strain and will increase the stress gradient, thus enhancing the
Similar to the welding of aluminum alloys, cracking in SLM can crack initiation rate [78]. Furthermore, excessively rapid cooling
be divided into liquation cracking and solidification cracking [84]. also reduces the time available for the liquid to refill and heal the
Liquation cracks often occur in alloys with high amounts of alloy- cracks.
ing elements [85]. These alloys will precipitate many low melting
point eutectics in the heat affected zone and overlapping regions 2.2.4. Oxidation
between layers that will be remelted under the peak temperature of Although parts are processed under a protective atmosphere,
the thermal cycle [86]. Thus, cracking occurs under tensile stress. At there is still approximately 0.1% oxygen in the real production pro-
present, there have been few studies of the formation mechanism cess due to air filling in the gaps between the powder particles.
and influencing factors of liquation cracks produced by SLM in var- Similar to the casting process, oxide inclusions in SLM mainly come
ious kinds of metallic materials. In the final stage of solidification, from two sources. One is due to the partial oxidation of the pow-
hot cracking caused by the residual liquid film between the crystals der raw material. The other arises from oxygen entrapped from
in the mushy zones is termed solidification cracking [87]. Solidifi- the atmosphere by the surface turbulent flow of the molten pool
cation cracking is closely related to the range of the solidification [63,70]. Al2 O3 formed during SLM processing of aluminum alloys
temperature and the content of liquid during solidification [88,89]. hinders the fusion of the powder particles, lowers the metallurgi-
Ductility-dip cracking (DDC) is one of the key mechanisms of crack cal bonding effect between the solidified layers and scanning paths,
formation during the SLM process and often occurs at a moderate aggravates the balling phenomenon and reduces the densification
temperature where the ductility and tensile properties are rela- of the alloy [99,100]. Oxide inclusions often induce crack initiation,
tively low [90]. The DDC mechanism usually induces the formation lowering the mechanical properties of the parts [99,101].
of intergranular cracks, and crack initiation is more likely to occur in Louvis et al. [102] used a NaOH solution to deeply corrode 6061
the presence of high angle grain boundaries [31,91,92]. Because the aluminum alloy, revealing the oxidation mechanism and oxide film
solidification strain is directly related to the solidification temper- morphology of the SLMed aluminum alloy. As shown in Fig. 10(a)
ature range where the ductility and tensile properties are lower, and (b), oxide films can form on all sides of the molten pool in the
alloys with a wide temperature range of solidification are more process of laser processing. The oxide films on the upper surface
likely to exhibit crack initiation [93,94]. Regarding the SLM process, evaporate and form fumes escaping from the molten pool under the
owing to the lack of diffusion in the non-equilibrium rapid solid- laser beam. The oxide films on the lower side of the molten pool are
ification process, the solidus and liquidus temperatures decrease, broken up by the stirring action of the Marangoni flow. Thus, the
and the temperature range of solidification becomes wider [84]. “walls” of oxide film around the weld pool are retained, and oxide
The residual liquid in the mushy zone along the grain boundaries films with a network distribution are observed after deep corrosion,
becomes film-like in shape, which results in strain concentration as shown in Fig. 10(c). The existence of these oxide films reduces the
276 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

Fig. 10. Formation mechanism and morphology of oxide film [102]. (a) Oxide disruption of the molten pool. (b) The walls of oxide films. (c) SEM image of oxide films after
deep corrosion.

wettability of the melt and substrate, and the overlapping portions spontaneously nucleated particles in the molten pool, which is an
of the adjacent tracks form closed or semi-closed pores, in which important cause for the development of columnar grains [36].
some melt or unmelted powder particles are trapped. The energy density is identified as a key parameter affecting
To reduce oxidation in SLM, besides maintaining sufficiently low vaporization. High loss of alloying elements usually occurs at high
oxygen partial pressure, the metal powder must be clean and dry. energy density due to overheating and evaporation of volatile
Under laser action, moisture on the powder surface can be decom- elements [53,105]. However, the chemical composition variation
posed into H2 to produce hydrogen porosity and oxygen combined within an SLMed part is a function of the evaporation rate of the
with aluminum melt to form alumina [103]. elements and the volume of the molten pool. Although the evapo-
ration rate of the elements increases with the energy density, the
small volume of liquid metal causes a high specific surface area
2.2.5. Loss of alloying elements of the molten pool, so the evaporation rate of alloying elements
In additive manufacturing using laser or electron beam melting is also enhanced at too low an energy density level [108]. There-
technologies, the high superheat of the molten pool and evapora- fore, the combination of moderate laser power with a high scanning
tion losses of alloying elements will inevitably lead to the deviation speed is the key to avoiding the evaporation of low-melting-point,
of the original alloy composition and reduction of performance low-vapor-pressure elements.
[104–106].
Serious losses and uneven distribution of Al in as-SLMed TiAl
3. Solidification theory and metallurgical defects of laser
samples have been observed, and the corresponding Al loss of sin-
selective melting
gle tracks varies from 5.73 to 0.32 at% [53]. Significant Al losses and
segregation have also been reported in the literature in the selective
Most research on SLM-fabricated metals focuses on ferrous-,
electron beam melting (SEBM) manufacturing process [104,105].
nickel- and titanium-based alloys [49,109,110]. With the develop-
The loss of tin is observed on the fracture surfaces of SLMed Cu-
ment of research and technology, the SLM process of aluminum,
4Sn specimens [107] . These samples simultaneously display step
copper- and magnesium-based materials has begun to mature
surfaces and dimples, thus implying that the production of brit-
[107,111–113]. Compared with iron-, nickel- and titanium-based
tle cleavage fracture and ductile fracture can be ascribed to the
alloys, the difficulties of processing aluminum alloys by SLM are as
segregation of tin in the sample. The dissipation of elemental Mg
follows [84,102,114].
or Zn from an aluminum alloy not only reduces the precipitation
strengthening effect and mechanical properties but also decreases
the stability of the scanning tracks [108]. Therefore, the loss of (1) The poor fluidity of aluminum alloy powder induces the for-
alloying elements is also an important reference criterion for the mation of agglomeration when spreading the powder, which
optimization of the process parameters. In addition, the loss of results in an uneven thickness of the powder layers and affects
alloying elements may cause a reduction of the amount of non- layer quality.
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 277

above Td , these particles begin to melt. At temperatures above Tb ,

the molten Al and Si are uniformly distributed in the liquid phase. Td
(AlSi10Mg: 1020 ± 30 ◦ C, AlSi12: 1080 ± 30 ◦ C) and Tb (AlSi10Mg:
1170 + 30 ◦ C, AlSi12: 1090 ± 30 ◦ C) are both significantly higher
than the eutectic temperature of the Al-Si alloy (557 ◦ C) [119,120].
In the molten pool, the temperature of most areas is between Td
and Tb , so most of the melt is in an overheated state, which leads
to an uneven distribution of Al and Si. In addition, the short inter-
action time between the laser and material, vibration of the liquid
phase and capillary action also contribute to inhomogeneity in the
microstructure [120]. The Al-Si phase diagram shows that the solid-
ification of AlSi10Mg undergoes a phase transition reaction L→L+˛
and a eutectic reaction L→˛+Si successively, and the solid solubility
of Si in the Al matrix reaches 8.89 at.% during SLM processing [117].
Fig. 11. Al-Si phase diagram.
During solidification, Si particles whose melting point is higher are
first heterogeneously nucleated in the melt pool. With decreas-
(2) The laser reflectivity of aluminum reaches 91%. The thermal ing temperature, ␣-Al nucleates and grows in the depleted region
conductivity of aluminum is 237 W/(mK), which is eleven times around the silicon particles. The continuous solidification of ␣-Al
the value for Ti and five times the value for Fe. A higher cool- leads to a gradual increase in the concentration of Si in the residual
ing rate leads to crack initiation. Less heat accumulation during liquid phase, which makes the liquid composition gradually move
processing is more likely to cause the formation of metallurgical toward the eutectic range, forming the Al-Si eutectic [121–123].
defects such as pores and cracks. Therefore, a higher laser power A large temperature gradient in the molten pool leads to a great
is required to form dense parts, so higher power demands on undercooling of the melt, resulting in the fiber morphology of the
the equipment are required. Si crystal. Therefore, the microstructure has a grid-like pattern of
(3) Aluminum oxidizes easily. The existence of oxide films will fibrous Si embedded in a supersaturated Al matrix in the SLMed
induce surface passivation of the molten pool and promote AlSi10Mg alloy. Fig. 12(a) and (b) present a network of ultrafine
the formation of metallurgical defects, which results in strict cellular structures with an average diameter of ∼500 nm, which is
requirements for the vacuum or oxygen partial pressure of the often observed in studies of SLMed Al-Si alloys [124,125]. From the
processing environment. inverse polar figure obtained by Electron Backscattered Diffraction
(4) The moisture absorption of aluminum alloy powder is strong, (Fig. 12(c)), the average grain size is revealed as ∼10 ␮m, which
and the solubility values of hydrogen in solid and liquid alu- is a magnitude larger than the cell size [125]. The Si phase of cast
minum are very different. Thus, hydrogen porosity is easily Al-Si alloy is rod-like or acicular, and the microstructure is coarser,
produced when the melt solidifies rapidly. as shown in Fig. 12(d). Compared to the Al-Si alloys made by con-
(5) The high thermal expansion coefficient and wide solidification ventional methods, the ultrafine eutectic microstructure formed by
temperature range of aluminum may generate a large amount SLM has better mechanical properties.
of residual stress during rapid solidification, which causes the
parts to crack and deform. 3.1.2. Mechanical properties of SLMed Al-Si alloys
(6) Aluminum alloys often contain Mg and other low melting point The microstructure of the Al-Si alloy, such as the size and
compounds. The fluctuation of alloy composition and crack for- morphology of the Si particles and the intermetallic compounds,
mation occur easily during laser cycling. dictates the mechanical properties of the parts [126]. At present,
studies on the mechanical properties of SLMed aluminum alloys
At present, SLM-fabricated aluminum alloys are mostly cast- mostly involve hardness and tensile properties, which are sum-
ing grade Al-Si alloys, especially AlSi10Mg and Al-Si12, whose marized in Table 2. The Vickers hardness of SLMed AlSi10Mg is
castability and weldability are relative good. The demand for approximately 130–150 HV, which is double that of the EN 1706
more aluminum alloys and the development of SLM technology die casting alloy [127]. Residual stresses are not always detrimen-
have resulted in studies of Al-Mg, Al-Cu-Mg, Al-Si-Mg, zirconium- tal to the manufacturing process of SLM. They can also improve the
modified aluminum alloys and other wrought aluminum alloys. hardness of a component if at a reasonable level [80,128]. This is
In this chapter, the research progress on the microstructure and another reason for the increase in hardness of SLM specimens. The
properties of various SLMed aluminum alloys is summarized. increase of hardness also improves the wear resistance of SLM parts.
The hypereutectic AlSi50 aluminum alloy has a hardness value of
3.1. Al-Si series alloys 188 HV. As a reinforcement material in the sliding process, the pri-
mary Si improves the friction and wear properties of the alloy [129].
Al-Si casting alloys are widely used in the automotive and power However, with such a high content of Si, macro-segregation of the
transmission industries [115]. Fig. 11 shows that the compositions primary Si is observed due to Marangoni convection and the tem-
of AlSi10Mg and AlSi12 are near the eutectic points, where the melt- perature difference between the interior and exterior of the molten
ing point is lower and solidification temperature range is narrower. pool [130,131].
Therefore, near-eutectic Al-Si cast alloys have good casting proper- Owing to the fine grain strengthening and solid solution
ties and less shrinkage porosity, so they are more suitable for SLM strengthening associated with SLM processing, the yield strength
processing [116]. (YS) and ultimate tensile strength (UTS) of SLMed specimens
are superior to those of Al-Si alloys fabricated by conven-
3.1.1. Microstructure characteristics of SLMed Al-Si alloys tional powder metallurgy [125] and casting methods [132,136],
There are two characteristic temperatures in the melting and but the plasticity is reduced. Similar ultimate tensile and yield
solidification process of Al-Si alloys: the dissolution temperature strengths were obtained from different research groups in both
(Td ) and the branching temperature (Tb ) [117,118]. When the tem- cases (UTS = 380–450 MPa; YS = 250–350 MPa). The differences in
perature is below Td , the liquid phase contains Al and Si-rich mechanical properties are mainly attributed to the diverse SLM pro-
particles separated from the powder. When the temperature is cess conditions, which lead to some deviations in the densities and
278 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

Fig. 12. (a) SLMed Al-Si alloy microstructure at low magnification. (b) SLMed Al-Si alloy microstructure at high magnification. (c) Inverse polar figure obtained by EBSD [125].
(d) Casting Al-Si alloy microstructure [124].

Table 2
Mechanical properties of SLMed Al-Si alloys from different literatures.

Material Condition Hardness(HV) Yield strength(MPa) Ultimate Tensile Strength(MPa) Elongation(%) Reference

AlSi10Mg As-SLMed — ∼270 ∼375 ∼4 [132]

As-SLMed ∼136 – ∼396 ∼3.5 [133]
As-SLMed 139-146 ∼360 ∼6 [134]
As-SLMed ∼133 ∼322 ∼434 ∼5.3 [117]
SLM + solution at 450 ◦ C ∼90 ∼196 ∼282 ∼13.4 [117]
SLM++solution at 550 ◦ C ∼60 ∼90 ∼168 ∼23.7 [117]
SLM + T6 ∼78 – ∼187 ∼19.5 [117]
As-SLMed 125 ∼268 ∼333 ∼1.4 [135]
SLM + T6 ∼103 ∼239 ∼292 ∼3.9 [135]
As-SLMed ∼311 ∼391 ∼7.2 [136]
As-SLMed ∼300 ∼455 ∼5.4 [136]
As-SLMed ∼255 ∼377 ∼1.2 [137]
SLM + annealing ∼158 ∼256 ∼9.9 [137]
SLM + T6 ∼210 ∼284 ∼4.9 [137]
AlSi12 SLM + solution ∼110 ∼190 ∼25 [118]
As-SLMed ∼260 ∼380 ∼3 [138]
SLM + annealing ∼95 ∼140 ∼15 [138]
AlSi9Mg As-SLMed 328 379 ∼8.1 [138]
AlSi7Mg0.3 As-SLMed ∼200 ∼400 12-17 [139]
AlSiMg0.75 As-SLMed ∼150 354.9 427.7 2.54 [140]
SLM + annealing ∼110 275.4 360.2 4.57 [140]

microstructures of the samples, such as imperfections, texture and the best combination of parameters, a fatigue limit of ∼200 MPa
cell size [141]. (with no substrate heating, 0◦ orientation and T6 heat treatment)
The fatigue property is a critical attribute for reliable appli- could be attained. The combination of 300 ◦ C platform heating
cations in functional industrial components. However, studies on and peak hardening is a valuable approach to increase the fatigue
the fatigue properties of SLMed aluminum alloys are much fewer resistance (and the static tensile strength) and neutralize the dif-
than those on their static properties. Brandl et al. [142] inves- ferences in fatigue life for the 0◦ , 45◦ , and 90◦ directions. The
tigated the effect of the platform temperature, build direction breakthrough cracks always start at the surface or subsurface
and heat treatment on the fatigue strength of AlSi01Mg. Peak (pores, non-melted spots), which has also been reported for SLMed
hardening has the most considerable and building direction the AlSi12 alloy [143], Ti6Al4V [144,145] and 316 L stainless steel
least significant impact on the fatigue resistance. By choosing [146].
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 279

Table 3
Mechanical properties of AlSi12 via SLM and casting in literature [124].
√ √
Test condition YS (MPa) UTS(MPa) ef (%) Kq (MPa m) Kth (MPa m) m FS(MPa) FS/UTS

AS 270.1±10 325±20 4.4±0.7 46.7 1.1 3.1 60 0.22

AS⊥ 274.8±8 296.1±20 2.2±0.3 37.9 1.4 3.7 — —
HS 153.4±5 228±13 5.3±0.7 21.7 2.0 3.1 110 0.41
HS⊥ 150.3±17 210.1±20 4.2±0.3 19.3 3.1 3.7 — —
CS 262.4±17 330.7±15 3.9±0.6 47.0 1.3 3.4 70 0.22
CS⊥ 276.6±15 302.7±15 2.3±0.3 34.5 1.3 3.9 — —
CC 104.2±11 192.3±15 9±0.5 11.1 3.4 5.4 94.5 0.49

Fig. 13. Schematic diagram of scanning strategy [124]: (a) unidirectional scanning strategy. (b) checkerboard scanning strategy.

The directional solidification of SLM leads to preferential growth the driving force on the crack tip, thereby increasing the tough-
in the <100> direction, and the resultant intense texture in the ness of the alloy [150]. Therefore, the SLM process can increase the
alloy leads to anisotropy [35]. By reducing the layer thickness and strength and toughness of aluminum alloy simultaneously.
the scan spacing to increase the amount of remelting of the prior The fatigue crack propagation of a specimen follows the Paris
solidified structure, grains are more likely to grow along the build law:
or Z direction, which can reduce the anisotropy of the alloy [35].
da/dN=c(K)m (2)
The mechanical property characterization and anisotropy of SLMed
AlSi12 alloy compared with casting has been comprehensively where c and m are material constants, N is the cycle number of alter-
explored and examined in the literature [118]. Selected mechanical nating stresses, a is the crack length, and (da/dN) is called the fatigue
properties of SLMed and cast AlSi12 alloys are shown in Table 3. crack growth rate. The above parameters are listed in Table 3. Due
AS represents the SLMed alloy with the unidirectional scanning to the low strength of CC, the circular plastic zone ahead of the crack
strategy as seen in Fig. 13(a). CS represents the SLMed alloy using tip is larger, which accelerates crack closure and reduces the driving
the checkerboard scanning strategy shown as Fig. 13(b).  and ⊥ force of crack growth. Therefore, Kth is larger. The higher m value
represent parallel and vertical processing directions, respectively. of CC compared with the SLMed alloys is due to the presence of
Table 3 shows that SLMed alloys are almost isotropic in yield coarse Si dendrites in the microstructure. Fracture and debonding
strength (YS) and ultimate tensile strength (UTS), which agrees in the circular plastic zone ahead of the crack tip enhance the crack
with the results found in literature [132,136]. However, ef in growth rate of each fatigue cycle. Higher Kth and m values of the
the  direction is apparently higher than that in the ⊥ direction. cast specimens indicate higher fatigue crack initiation resistance,
The toughness of the SLMed alloy is anisotropic, which is mainly but a crack will grow rapidly once it is initiated [151]. For unnotched
related to the plastic deformation occurring in the microstructure. fatigue properties, the FS/UTS ratio for CC is relative high. The lower
First, plasticity will cause strain localization, debonding and cav- FS of SLMed alloys may be due to high tensile residual stresses and
ity nucleation at the interface of ␣-Al/Si [124]. Upon further strain, unmelted powder particles or pores in the microstructure. Crack
coalescence of these micropores results in crack formation and frac- initiation always starts at defects (the unmelted powder and pores)
ture. Because of the existence of the coarse Si phase in the heat due to stress conditions and the local plastic deformation caused
affected zone, cracks are commonly formed along the laser track. by surface discontinuities [152–154]. Pore formation reduces the
Therefore, the relative orientation between the scanning direction effective bearing area perpendicular to the layer (Z direction) and
and the loading direction has a great influence on the toughness of causes stress concentrations (notch effect), which lead to a decrease
the specimen. In addition, as shown in Fig. 14, the laser tracks in the in the static and dynamic strength in the Z direction [142].
 direction are more tortuous than those in the ⊥ direction [147],
which results in higher ef , Kq and UTS values. Unmelted powder 3.1.3. Effect of heat treatment on the microstructure and
particles due to oxidation and micropores are the main cause of the properties of SLMed Al-Si alloy
cracking of specimens [148,149]. Although the ductility is low, the Li et al. [118] has investigated the effect of the heat treat-
presence of molten pool boundaries enhances the tortuosity of the ment time on the microstructure and mechanical properties of
crack paths and leads to significant crack deflection, which reduces Al-Si alloys. After heat treatment, the alloy microstructure is trans-
280 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

Fig. 14. Possible crack propagation tracks of SLM specimens [147]. (a)  direction. (b) ⊥direction.

Fig. 15. Microstructure of Al-Si12 after heat treatment [118]. (a) 15 min, (b) 30 min, (c) 2 h, (d) 4 h.

formed from ␣-Al with a Si network into coarse Si particles conventional casting method, the spherical Si particles can reduce
uniformly distributed in the Al matrix. As the heat treatment time the local shear and restrain crack initiation and propagation when
increases, fine Si particles are precipitated from the supersaturated deformed. An increase of the Si particle size and a decrease of the
Al matrix and are deposited on the surface of larger particles. Due number of Si particles reduce the local stress and strain during ten-
to the Ostwald ripening and coalescence of adjacent small Si parti- sile loading. The residual stress level is also greatly reduced after
cles, continued growth of the Si particles occurs. The amount and heat treatment, so the tensile plasticity of the samples is enhanced.
particle size of eutectic Si increase, but the number of eutectic Si As shown in Table 3, the elongation rate reported in the literature
particles decreases, and the difference between a fine grain zone [118] reaches 25%.
and a coarse grain zone is less obvious, as shown in Fig. 15. The The mechanical properties of AlSi12 after heat treatment are
heat treatment temperature has a similar effect in that a higher shown in Table 3. HS represents the alloy processed by SLM with
temperature induces a coarser microstructure and a larger size but unidirectional scanning and subsequent heat treatment (300 ◦ C,
smaller number of Si particles [138]. 6 h). From Table 3, it is seen that after heat treatment, YS and UTS
The morphology of the eutectic Si phase plays a significant role decrease remarkably but remain at approximately twice the value
in the mechanical properties of AlSi12 alloys. In addition to the for CC. The ductility, especially in the vertical direction, increases
interface energy ␥(Al/Si) between Al and Si, other kinetic or thermo- substantially, which is consistent with the above analysis. After the
dynamic factors, such as the wettability (contact angle ) and the Si is precipitated from the Al matrix, the influence of the molten
local concentration of Al and Si atoms, are also of importance to the pool boundary on crack propagation is reduced, which decreases
solidification eutectic microstructure [155–157]. The Si-rich and the toughness of the sample after heat treatment. The reduction
Al-rich regions in the melt are mostly retained during rapid cool- of Kq does not lead to a decrease of Kth ; conversely, it increased
ing, and their growth during the cooling process is greatly limited. by 100% and 200%, respectively, in the  and ⊥ directions. This is
There is an orientational relationship between Al and Si of (111) Si mainly due to the decrease of the yield strength after annealing,
 (200) Al [118]. When heat treatment is carried out at a high tem- which results in more obvious crack blunting and promotes crack
perature, the Si phase undergoes thermally activated growth; thus, closure and a rise of Kth . For steady-state fatigue crack growth,
the Si phase can grow along the stable close-packed plane {111}, the size scale of the cyclic crack tip field is much larger than that
and the precipitated Si phase is spherical. Compared with the Si of the microstructure, so the crack propagation rate (m) is insen-
phase in flake- or rod-like or acicular shape manufactured by the sitive to the microstructural parameters. As the annealing process
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 281

reduces the tensile residual stress produced during processing, FS Zr nucleants were chosen to coat pre-alloyed gas-atomized 7075
rises, which is different from the common case that FS has the same and 6061 spherical powders via an electrostatic assembly tech-
trend as YS. nique. In situ synthesized fine Al3 Zr particles provided a large
number of ideal low-energy heterogeneous nucleation sites and
3.2. Other aluminum alloys induced equiaxed microstructures, which easily accommodate the
thermal contraction strains associated with solidification, result-
Wrought aluminum alloys have a large solidification tempera- ing in an alloy system that is highly tear resistant. Finally, the
ture range, high crack sensitivity and poor formability. As a result, mechanical properties of additively manufactured 7075 are within
analogous to welding processes, columnar grains and hot tearing the expected bounds for wrought counterparts. This metallurgical
cracks are formed under direct laser irradiation. approach offers a good reference for other crack-susceptible alloys.
Al7075 and Al6061 alloys are the most widely used com- By optimizing the processing of 2XXX series Al-Cu-Mg alloy,
mercial aluminum alloys due to either high strength or medium Zhang et al. [163] found that a dense sample (99.8%) could
strength but good machinability. Many attempts have been made to be obtained using the laser energy density threshold value of
manufacture high-performance crack-free alloys, but cracking and 340 J/mm3 . Cu-rich intermetallic compounds are homogeneously
pore formation are extensive during SLM processing [74,102,158]. distributed in the Al matrix and become coarser at the bound-
The strength of 7075 obtained via SLM is only 200 MPa, and the ary of the molten pool. The ultimate tensile strength and yield
strength is not changed after heat treatment [159]. Bourell [160] strength are 402 MPa and 276 MPa, respectively. Heat treatment
has explored the use of an elemental powder feedstock to mitigate has a remarkable influence on the mechanical properties of SLMed
crack formation in Al6061. Here, the methodology involves using Al-Cu-Mg alloy [164]. When the heat treatment temperature rises
approximately 98% pure Al powder with either small amounts from 480◦ C to 540◦ C, the microstructure becomes coarser but more
of Mg and Si or Mg2 Si particles to eliminate the negative effects homogeneous, as shown in Fig. 16(c). With the increase of the
of elemental Mg in the microstructure. When SLM processed, heat treatment solutionizing temperature, UTS, YS and elongation
the solidification is controlled by the pure Al, which results in a increase by approximately 15%, 22% and 47%, respectively, which is
microstructure of Al with isolated particles of Mg and Si (or Mg2 Si). attributed to precipitation strengthening of the S phase (Al2 CuMg).
The 6061 microstructure is restored during a post-build homoge- When the temperature is further increased to 560◦ C, the tensile
nization heat treatment, similar to the solutionizing heat treatment properties decrease. Because of the low degree of saturation after
of conventionally processed Al6061. Si additions can help improve slow quenching, the size and distribution of the precipitates are no
melt flowability, reduce the thermal expansion coefficient and longer helpful for strengthening [165]. Thus, compared with water
decrease the eutectic temperature and temperature range of solid- quenching, UTS, YS and elongation are reduced by 70%, 65% and 56%,
ification. The addition of Si to the original powder can remarkably respectively, when using air cooling. Under the optimum heat treat-
increase the density of the samples, but tensile properties are ment regime (540 ◦ C, 1 h, water quenched), UTS, YS and elongation
not reported [160,161]. HRL Laboratories [162] demonstrated that are 532 MPa, 338 MPa and 13%, respectively. To reduce microcrack-
crack formation can be resolved by introducing nanoparticles of ing and improve the mechanical properties of the alloy, 2 wt% Zr
nucleants that control solidification during laser additive manufac- powder was added to the Al-Cu-Mg powder [166]. The addition of
turing. They indicated that solidification shrinkage of interdendritic Zr leads to the formation of more low melting point phases, which
fluid trapped between the cellular or dendritic grains produced cav- can backfill cracks in the final stage of solidification and reduce
ities and hot tearing cracks. Equiaxed grains could reduce the effect crack susceptibility [167]. Al3 Zr particles make the grains change
of trapped liquid, as the grains behaved as a low-resistance granular from columnar crystals to ultrafine equiaxed grains (Fig. 16(a), (b)
solid. Based on the crystallographic theories, hydrogen-stabilized and (d)). The fine grains increase the total grain boundary area per

Fig. 16. SEM images of Al–Cu–Mg samples. (a) and (b) SLMed Al–Cu–Mg samples [163]. (c) Al-Cu-Mg alloy after heat treatment [164]. (d) SLMed Zr/Al-Cu-Mg alloy [166].
282 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

unit volume, which strengthens the material and suppresses inter- erties. Aluminum alloys also have excellent thermal and electrical
granular crack formation [168]. Moreover, with the addition of Zr, properties. Research on these comprehensive properties and on
the boundary misorientation distribution shifts from a low angle the anisotropy of SLMed aluminum alloys will be important in the
to a high angle, and the distribution nicely follows the ideal distri- future. On the other hand, the manufacturing flexibility of the SLM
bution for random orientation, which shows that the Zr/Al-Cu-Mg process can be realized by incorporating lattice and porous struc-
sample fabricated by SLM has low anisotropy. Finally, no obvi- tures and other complex part geometries. Since SLM technology is
ous microcracking was observed in the specimens, and the yield interdisciplinary in nature, it is meaningful to combine the advan-
strength increased to 446 ± 4.3 MPa due to the combined effect tages of SLM with mechanical and structural design to realize a
of fine grain strengthening and precipitation strengthening. The complete process from the design to the manufacturing of complete
ultimate strength increased to 451 ± 3.6 MPa. structures or functional components.
The type of SLM equipment used for processing can also affect
defect formation. Studies at the Universities of Nottingham and
Acknowledgements
Erlangen have shown suppression of crack formation in wrought
Al alloys when a Renishaw fabricator is used. This is attributed to a
This work was sponsored by National Key Research and
high energy density and a small melt pool. In this regard, the “AM-
Development Program “Additive Manufacturing and Laser Manu-
processability” of the alloys is similar to their weldability, for which
facturing” (No. 2016YFB1100101), Natural and Science Foundation
certain welding processes are successful and others are not.
of China (Grant Nos. 51775208, 51505166), Hubei Science Fund
for Distinguished Young Scholars (No. 0216110085), Wuhan Morn-
ing Light Plan of Youth Science and Technology (No. 0216110066),
4. Prospective
Graduates’ Innovation Fund, Huazhong University of Science and
Technology (No. 5003110027), Fundamental Research Funds for
To summarize, the present development obstacles for laser
the Central University (No. 2017JYCXJJ004), and the Academic fron-
selective melting manufacturing of aluminum alloys mainly
tier youth team at Huazhong University of Science and Technology
include four aspects: the aluminum alloy types suitable for SLM
(HUST).
are few, the technological conditions are immature, metallurgical
defects are difficult to control and performance studies have not
been systematic. References
At present, the research studies on aluminum alloy SLM manu-
[1] J.C. Williams, E.A.S. Jr, Acta Mater. 51 (2003) 5775–5799.
facturing have mostly been developed around the influence of the [2] A. Heinz, A. Haszler, C. Keidel, S. Moldenhauer, R. Benedictus, W.S. Miller,
technological parameters and the heat treatment process on the Mater. Sci. Eng. A 280 (2000) 102–107.
microstructure and mechanical properties of Al-Si alloys. Except [3] W. Xu, Y. Luo, Wei. Zhang, M. Fu, J. Mater. Sci. Technol. 34 (2018) 173–184.
[4] C. Kammer, Aluminium Handbook, Aluminium-Verlag Marketing &
for the studies on the 2XXX series Al-Cu-Mg alloy [163,164,166], Kommunikation GmbH, Germany, 1999, pp. 61–73.
Al7075 and Al6061 [162], investigations of other casting and [5] E.A.S. Jr, J.T. Staley, Prog. Aerosp. Sci. 32 (1996) 131–172.
wrought aluminum alloys, especially the high strength and heat- [6] S. Fu, Y. Zhang, H. Liu, D. Yi, B. Wang, Y. Jiang, Z. Chen, N. Qi, J. Mater. Sci.
Technol. 34 (2018) 335–343.
resistant aluminum alloys that are in great demand in the aerospace
[7] C. Cui, Y. Shen, F. Meng, J. Mater. Sci. Technol. 16 (2000) 619–626.
field, have not been reported in detail. In addition, most aluminum [8] S. Ranganath, J. Mater. Sci. Technol. 32 (1997) 1–16.
alloys manufactured by SLM are at present based on the alloying [9] S.C. Tjong, Z.Y. Ma, Mater. Sci. Eng. R 29 (2000) 49–113.
[10] L. He, J. Kang, T. Huang, K. Rong, Heat Treat. Met. 31 (2004) 69–72.
composition of commercial grades. Non-equilibrium solidification
[11] W. Li, K. Yang, S. Yin, X. Yang, Y. Xu, R. Lupoi, J. Mater. Sci. Technol. 34 (2018)
occurring during the SLM process, however, leads to a significantly 440–457.
enhanced solid solution limit of the alloy elements in the matrix [12] R.S. Mishra, Z.Y. Ma, Mater. Sci. Eng. R 50 (2005) 1–78.
metal, which makes it difficult to achieve the desired properties [13] F. Nie, H. Dong, S. Chen, P. Li, L. Wang, Z. Zhao, X. Li, H. Zhang, J. Mater. Sci.
Technol. 34 (2018) 551–560.
according to the composition of the traditional grades. Therefore, [14] Y. Liu, Y. Yang, S. Mai, D. Wang, C. Song, Mater. Des. 87 (2015) 797–806.
there is a need to broaden the scope of the applicable aluminum [15] J.P. Kruth, L. Froyen, J.V. Vaerenbergh, P. Mercelis, M. Rombouts, B. Lauwers,
alloy types for SLM and design new materials whose composition J. Mater. Process. Tech. 149 (2004) 616–622.
[16] J.P. Kruth, G. Levy, F. Klocke, T.H.C. Childs, CIRP Ann. Manuf. Tech. 56 (2007)
is suitable for the process characteristics of laser additive manufac- 730–759.
turing. [17] L.E. Murr, S.M. Gaytan, D.A. Ramirez, E. Martinez, J. Hernandez, K.N. Amato,
A high level of residual stress and metallurgical defects, such as P.W. Shindo, F.R. Medina, R.B. Wicker, J. Mater. Sci. Technol. 28 (2012) 1–14.
[18] N. Contuzzi, S.L. Campanelli, A.D. Ludovico, Int. J. Simul. Model 10 (2011)
balling, pores and cracks, commonly occur in SLM-fabricated alu- 113–121.
minum alloys, which not only reduces the surface roughness and [19] S.L. Sing, J. An, W.Y. Yeong, F.E. Wiria, J. Orthop. Res. 34 (2016) 369–385.
dimensional accuracy but also results in poor mechanical proper- [20] D. Gu, P. Yuan, J. Appl. Phys. 118 (2016) 401–477.
[21] S. Pauly, L. Löber, R. Petters, M. Stoica, S. Scudino, U. Kühn, J. Eckert, Mater.
ties of the components. However, understanding of the formation
Today 16 (2013) 37–41.
mechanism of the above defects is still incomplete. The methods [22] Y. Zhang, J. Zhang, Q. Yan, L. Zhang, M. Wang, B. Song, Y. Shi, Scripta Mater.
of improving SLM technology mainly focus on process parame- 148 (2018) 20–23.
[23] K.G. Prashanth, J. Eckert, J. Alloy Compd. 707 (2016) 27–34.
ter adjustment, substrate preheating, laser remelting and scanning
[24] Y. Yang, J. Liu, Z. Luo, D. Wang, Rapid Prototyp. J. 18 (2012) 482–489.
strategy variation. Few systematic studies have been conducted [25] B. Song, S. Dong, P. Coddet, H. Liao, C. Coddet, Mater. Des. 53 (2014) 1–7.
to improve part quality via dual-laser online annealing and multi- [26] P. Chalker, A. Clare, S. Davies, C.J. Sutcliffe, S. Tsopanos, Int. J. Mech. Mater.
laser scanning. Moreover, the above improvements are mostly from Des. 4 (2008) 181–187.
[27] G. Levy, N. Herres, A.B. Spierings, D. Bourell, Rapid Prototyp. J. 17 (2010)
the viewpoint of technology. The scientific characterization of these 195–202.
methods has not been explained and analyzed in detail. Theoretical [28] Q. Shi, D. Gu, M. Xia, S. Cao, T. Rong, Opt. Laser Technol. 84 (2016) 9–22.
computer simulations and experiments are needed to advance our [29] D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Int. Mater. Rev. 57 (2012)
133–164.
understanding and to improve part quality. [30] M. Peel, A. Steuwer, M. Preuss, P.J. Withers, Acta Mater. 51 (2003)
Because of the directionality of laser scanning, components tend 4791–4801.
to show some degree of anisotropy. Thus far, most research on [31] C.V. Robino, J.R. Michael, M.C. Maguire, Weld. J. 77 (1998) 446S–457S.
[32] T.W. Nelson, J.C. Lippold, M.J. Mills, Weld. J. 78 (1999) 329S–337S.
the performance of SLMed aluminum alloys has focused on the [33] C. Wang, J. Zhang, L. Liu, H. Fu, J. Mater. Sci. Technol. 27 (2011) 668–672.
mechanical properties, especially the tensile properties. Relatively [34] M. Gäumann, S. Henry, F. Cléton, J.D. Wagnière, W. Kurz, Mater. Sci. Eng. A
few studies have been conducted on the fracture and fatigue prop- 271 (1999) 232–241.
 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284 283

[35] L. Thijs, K. Kempen, J.P. Kruth, J.V. Humbeeck, Acta Mater. 61 (2013) [90] L.N. Carter, C. Martin, P.J. Withers, M.M. Attallah, J. Alloy Compd. 615 (2014)
1809–1819. 338–347.
[36] S. Wu, Y. Liu, Principle of Materials Forming, China Machine Press, Beijing, [91] M.G. Collins, J.C. Lippold, Weld. J. 83 (2003), 39/S-49/S.
2008, pp. 123–154 (in Chinese). [92] M.G. Collins, A.J. Ramirez, J.C. Lippold, Weld. J. 83 (2003), 39/S-49/S.
[37] G. Çam, S. Mistikoglu, J. Mater. Eng. Perform. 23 (2014) 1936–1953. [93] N. Coniglio, C.E. Cross, Int. Mater. Rev. 58 (2013) 375–397.
[38] W. Huang, C. Chuang, C. Lin, C. Wu, D. Lin, S. Liu, W. Tseng, J. Horng, Phys. [94] P.V. Witzendorff, S. Kaierle, O. Suttmann, L. Overmeyer, Metall. Mater. Trans.
Procedia 56 (2014) 58–63. A 46 (2015) 1678–1688.
[39] H. Fu, J. Mater. Sci. Technol. 17 (2001) 299–302. [95] L. Thijs, F. Verhaeghe, T. Craeghs, J.V. Humbeeck, J.P. Kruth, Acta Mater. 58
[40] H. Yamagata, W. Kasprzak, M. Aniolek, H. Kurita, J.H. Sokolowski, J. Mater. (2010) 3303–3312.
Process. Tech. 203 (2008) 333–341. [96] H. Zhao, D.R. White, T. Debroy, Int. Mater. Rev. 44 (1999) 238–266.
[41] S. Guo, Solidification Behavior of Spray Deposited Al-Zn-Mg-Cu Alloys and [97] D. Choo, H. Moon, T. Kang, S. Lee, Mater. Trans. A 32 (2001) 2249–2258.
Their Microstructural Evolution During Thermal Processing, Ph.D. Thesis, [98] Y. Liu, Y. Yang, D. Wang, Int. J. Adv. Manuf. Tech. 87 (2016) 647–656.
Harbin Institute of Technology, 2011 (in Chinese). [99] A. Jalar, N.A.C. Lah, N.K. Othman, R. Shamsudin, A.R. Daud, S.R.S. Bakar, Adv.
[42] X. Xu, Y. Zhao, H. Hou, F. Liu, J. Alloy Compd. 744 (2018) 740–749. Mater. Sci. Eng. 2014 (2015) 1–10.
[43] Q. Zhang, H. Xue, Q. Tang, S. Pan, M. Rettenmayr, M. Zhu, Comp. Mater. Sci. [100] M. Simonelli, C. Tuck, N.T. Aboulkhair, I. Maskery, I. Ashcroft, R.D. Wildman,
146 (2018) 204–212. R. Hague, Metall. Mater. Trans. A 46 (2015) 3842–3851.
[44] S. Liu, H. Zhu, G. Peng, J. Yin, X. Zeng, Mater. Des. 142 (2018) 319–328. [101] Q. Jia, D. Gu, Opt. Laser Technol. 62 (2014) 161–171.
[45] Y. Liu, Z. Liu, Y. Jiang, G. Wang, Y. Yang, L. Zhang, J. Alloy Compd. 735 (2018) [102] E. Louvis, P. Fox, C.J. Sutcliffe, J. Mater, Progress. Tech. 211 (2011) 275–284.
1414–1421. [103] C. Weingarten, D. Buchbinder, N. Pirch, W. Meiners, K. Wissenbach, R.
[46] P. Sharifi, J. Jamali, K. Sadayappan, J.T. Wood, J. Mater. Process. Tech. 34 Poprawe, J. Mater, Process. Tech. 221 (2015) 112–120.
(2018) 324–334. [104] A. Klassen, V.E. Forster, C. Körner,Modell. Simul. Mater. Sci. Eng. 25 (2016),
[47] D. Bourell, H.L. Marcus, J.W. Barlow, J.J. Beaman, Int. J. Powder Met. 28 025033.
(1992) 369–381. [105] A. Klassen, V.E. Forster, V. Juechter, C. Körner, J. Mater. Process. Tech. 247
[48] R. Li, Research on the Key Basic Issues in Selective Laser Melting of Matallic (2017) 280–288.
Powder, Ph.D. Thesis, Huazhong University of Science and Technology, 2010 [106] C. Panwisawas, C. Qiu, M.J. Anderson, Y. Sovani, R.P. Turner, M.M. Attallah,
(in Chinese). J.W. Brooks, H.C. Basoalto, Comp. Mater. Sci. 126 (2017) 479–490.
[49] C.Y. Yap, C.K. Chua, Z. Dong, Z. Liu, D. Zhang, L.E. Loh, S.L. Sing, Appl. Phys. [107] Z. Mao, D. Zhang, P. Wei, K. Zhang, Materials 10 (2017) 333.
Rev. 2 (2015), 518-187. [108] M. Pastor, H. Zhao, R.P. Martukanitz, T. Debroy, Weld. J. 78 (1999) 207s–216s.
[50] W. Wu, Y. Yang, D. Wang, J. South China Univ. Techno. Nat. Sci. Ed. 38 (2010) [109] M. Rombouts, J.P. Kruth, L. Froyen, P. Mercelis, CIRP Ann. Manuf. Technol. 55
110–115 (in Chinese). (2006) 187–192.
[51] X. Zhang, X. Liu, D. Zhang, Z. Shen, W. Liu, J. Mater. Process. Tech. 222 (2015) [110] H. Shipley, D. Mcdonnell, M. Culleton, R. Lupoi, G. O’Donnell, D. Trimble, Int.
33–42. J. Mach. Tool Manu. 128 (2018) 1–20.
[52] M. Shiomi, K. Osakada, K. Nakamura, T. Yamashita, F. Abe, CIRP Ann. Manuf. [111] I. Yadroitsev, A. Gusarov, I. Yadroitsava, I. Smurov, J. Mater. Process. Tech.
Techn. 53 (2004) 195–198. 210 (2010) 1624–1631.
[53] X. Shi, S. Ma, C. Liu, Q. Wu, Opt. Laser Technol. 90 (2017) 71–79. [112] E.O. Olakanmi, J. Mater. Process. Tech. 213 (2013) 1387–1405.
[54] J. Yang, J. Han, H. Yu, J. Yin, M. Gao, Z. Wang, X. Zeng, Mater. Des. 110 (2016) [113] C.C. Ng, M.M. Savalani, M.L. Lau, H.C. Man, Appl. Surf. Sci. 257 (2011)
558–570. 7447–7454.
[55] E. Yasa, J.P. Kruth, J. Deckers, CIRP Ann. Manuf. Techn. 60 (2011) 263–266. [114] X. Wang, Process Parameters and Properties of Selective Laser Melting Al-Si
[56] E. Yasa, J.P. Kruth, Procedia Eng. 19 (2011) 389–395. Alloys, Ph.D. Thesis, China University of Geosciences (Beijing), Beijing, 2014
[57] B. Song, S. Dong, H. Liao, C. Coddet, Mater. Res. Innov. 16 (2012) 321–325. (in Chinese).
[58] I. Yadroitsev, P. Krakhmalev, I. Yadroitsava, S. Johansson, I. Smurov, J. Mater. [115] H. Ye, J. Mater. Eng. Perform. 12 (2003) 288–297.
Process. Tech. 213 (2013) 606–613. [116] Q. Wang, C.J. Davidson, J. Mater. Sci. 26 (2001) 739–750.
[59] D. Gu, Y.C. Hagedorn, W. Meiners, G. Meng, R.J.S. Batista, K. Wissenbach, R. [117] W. Li, S. Li, J. Liu, A. Zhang, Y. Zhou, Q. Wei, C. Yan, Y. Shi, Mater. Sci. Eng. A
Poprawe, Acta Mater. 60 (2012) 3849–3860. 663 (2016) 116–125.
[60] Q. Jia, D. Gu, J. Alloy Compd. 585 (2014) 713–721. [118] X. Li, X. Wang, M. Saunders, A. Suvorova, L. Zhang, Y. Liu, M. Fang, Z. Huang,
[61] D. Dai, D. Gu, Mater. Des. 55 (2014) 482–491. T.B. Sercombe, Acta Mater. 95 (2015) 74–82.
[62] R. Xiao, X. Zhang, J. Manuf. Process. 16 (2014) 166–175. [119] M. Calvo-Dahlborg, P.S. Popel, M.J. Kramer, M. Besser, J.R. Morris, U.
[63] M. Xia, D. Gu, G. Yu, D. Dai, H. Chen, Q. Shi, Int. J. Mach. Tool Manu. 116 Dahlborg, J. Alloy Compd. 550 (2013) 9–22.
(2017) 96–106. [120] W. Liu, L. Ye, K. Liu, J. Appl. Phys. 109 (2011), 043109-043109-5.
[64] K. Monroy, J. Delgado, J. Ciurana, Procedia Eng. 63 (2013) 361–369. [121] S.D. Mcdonald, K. Nogita, A.K. Dahle, Acta Mater. 52 (2004) 4273–4280.
[65] P.A. Kobryn, E.H. Moore, S.L. Semiatin, Scr. Mater. 43 (2000) 299–305. [122] M. Gremaud, D.R. Allen, M. Rappaz, J.H. Perepezko, Acta Mater. 44 (1996)
[66] M. Ma, Z. Wang, M. Gao, X. Zeng, J. Mater, Process. Tech. 215 (2015) 142–150. 2669–2681.
[67] E.O. Olakanmi, R.F. Cochrane, K.W. Dalgarno, J. Mater. Process. Tech. 211 [123] Y.T. Pei, J.T.M.D. Hosson, Acta Mater. 48 (2000) 2617–2624.
(2011) 113–121. [124] J. Suryawanshi, K.G. Prashanth, S. Scudino, J. Eckert, O. Prakash, U.
[68] N.T. Aboulkhair, N.M. Everitt, I. Ashcroft, C. Tuck, Addit. Manuf. s1-4 (2014) Ramamurty, Acta Mater. 115 (2016) 285–294.
77–86. [125] B. Chen, S.K. Moon, X. Yao, G. Bi, J. Shen, J. Umeda, K. Kondoh, Scripta Mater.
[69] Y. Fu, N. Kang, H. Liao, Y. Gao, C. Coddet, Intermetallics 86 (2017) 51–58. 141 (2017) 45–49.
[70] W.E. King, H.D. Barth, V.M. Castillo, G.F. Gallegos, J.W. Gibbs, D.E. Hahn, C. [126] D.A. Lados, D. Apelian, J.K. Donald, Acta Mater. 54 (2006) 1475–1486.
Kamath, A.M. Rubenchik, J. Mater. Process. Tech. 214 (2014) 2915–2925. [127] D. Buchbinder, H. Schleifenbaum, S. Heidrich, W. Meiners, J. Bültmann, Phys.
[71] D. Dai, D. Gu, Appl. Surf. Sci. 355 (2015) 310–319. Procedia 12 (2011) 271–278.
[72] X. Wang, L. Zhang, M. Fang, T.B. Sercombe, Mater. Sci. Eng. A 597 (2014) [128] D. Gu, W. Meiners, Mater. Sci. Eng. A 527 (2010) 7585–7592.
370–375. [129] N. Kang, P. Coddet, C. Chen, Y. Wang, H. Liao, C. Coddet, Mater. Des. 99
[73] X. Cao, W. Wallace, C. Poon, J.P. Immarigeon, Mater. Manuf. Prog. 18 (2003) (2016) 120–126.
23–49. [130] N. Kang, P. Coddet, H. Liao, C. Coddet, J. Alloy Compd. 662 (2016) 259–262.
[74] N. Kaufmann, M. Imran, T.M. Wischeropp, C. Emmelmann, S. Siddique, F. [131] T. Kimura, T. Nakamoto, M. Mizuno, H. Araki, Mater. Sci. Eng. A 682 (2017)
Walther, Phys. Procedia 83 (2016) 918–926. 593–602.
[75] A. Hussein, L. Hao, C. Yan, R. Everson, Mater. Des. 52 (2013) 638–647. [132] N. Read, W. Wang, K. Essa, M.M. Attallah, Mater. Des. 65 (2015) 417–424.
[76] C. Li, C. Fu, Y. Guo, F. Fang, J. Mater. Process. Tech. 229 (2016) 703–712. [133] K. Kempen, L. Thijs, J.V. Humbeeck, J.P. Kruth, Mater. Sci. Tech. Lond. 31
[77] G. Çam, Int. Mater. Rev. 56 (2014) 1–48. (2015) 917–923.
[78] C.M. Chen, R. Kovacevic, Int. J. Mach. Tool Manu. 43 (2003) 1319–1326. [134] P. Wei, Z. Wei, Z. Chen, J. Du, Y. He, J. Li, Y. Zhou, Appl. Surf. Sci. 408 (2017)
[79] D. Deng, H. Murakawa, Mater. Sci. 37 (2006) 269–277. 38–50.
[80] P. Mercelis, J.P. Kruth, Rapid Prototyp. J. 12 (2006) 254–265. [135] N.T. Aboulkhair, I. Maskery, C. Tuck, I. Ashcroft, N.M. Everitt, Mater. Sci. Eng.
[81] D. Kim, J. Hwang, E. Kim, Y. Heo, W. Woo, S.H. Choi, J. Alloy Compd. 714 A 667 (2016) 139–146.
(2017) 687–697. [136] L. Wang, X. Jiang, M. Guo, X. Zhu, B. Yan, Mater. Sci. Tech. Lond 33 (2017) 1–9.
[82] L.E. Lindgren, Comput. Method Appl. M. 195 (2006) 6710–6736. [137] M. Fousová, D. Dvorský, A. Michalcová, D. Vojtěch, Mater. Charact. 137
[83] D. Ouyang, N. Li, W. Xing, J. Zhang, L. Liu, Intermetallics 90 (2017) 128–134. (2018) 119–126.
[84] E.O. Olakanmi, R.F. Cochrane, K.W. Dalgarno, Prog. Mater. Sci. 74 (2015) [138] K.G. Prashanth, S. Scudino, H.J. Klauss, K.B. Surreddi, L. Löber, Z. Wang, A.K.
401–477. Chaubey, U. Kühn, J. Eckert, Mater. Sci. Eng. A 590 (2014) 153–160.
[85] C. Huang, S. Kou, Weld. J. 84 (2004) 63s–71s. [139] T. Kimura, T. Nakamoto, Mater. Des. 89 (2016) 1294–1301.
[86] Y. Chen, K. Zhang, J. Huang, S.R.E. Hosseini, Z. Li, Mater. Des. 90 (2016) [140] Y. Bai, Y. Yang, Z. Xiao, M. Zhang, D. Wang, Mater. Des. 140 (2018) 257–266.
586–594. [141] F. Trevisan, F. Calignano, M. Lorusso, J. Pakkanen, A. Aversa, E.P. Ambrosio,
[87] D.G. Eskin, L. Suyitno, Katgerman, Prog. Mater. Sci. 49 (2004) 629–711. M. Lombardi, P. Fino, D. Manfredi, Materials 10 (2017) 1–23.
[88] J.H. Dudas, Weld. J. 45 (1966). [142] E. Brandl, U. Heckenberger, V. Holzinger, D. Buchbinder, Mater. Des. 34
[89] R. Liu, Z. Dong, Y. Pan, Chin. J. Nonferrous Met. 16 (2006) 110–116 (in (2012) 159–169.
Chinese). [143] S. Siddique, M. Imran, F. Walther, Int. J. Fatigue 94 (2016) 246–254.
284 J. Zhang et al. / Journal of Materials Science & Technology 35 (2019) 270–284

[144] S. Leuders, M. Vollmer, F. Brenne, T. Tröster, T. Niendorf, Metall. Mater. [158] P. Wang, H. Li, K.G. Prashanth, J. Eckert, S. Scudino, J. Alloy Compd. 707
Trans. A 46 (2015) 3816–3823. (2016) 287–290.
[145] A. Riemer, S. Leaders, H.A. Richard, T. Tröster, Materialprufung 55 (2013) [159] W. Reschetnik, J.P. Brüggemann, M.E. Aydinöz, O. Grydin, K.P. Hoyer, G.
537–543. Kullmer, H.A. Richard, Procedia Struct. Integr. 2 (2016) 3040–3048.
[146] A. Riemer, S. Leuders, M. Thöne, H.A. Richard, T. Tröster, T. Niendorf, Eng. [160] C.E. Roberts, D. Bourell, T. Watt, J. Cohen, Phys. Procedia 83 (2016)
Fract. Mech. 120 (2014) 15–25. 909–917.
[147] S. Wen, S. Li, Q. Wei, C. Yan, S. Zhang, Y. Shi, J. Mater. Process. Tech. 214 [161] M.L.M. Sistiaga, R. Mertens, B. Vrancken, X. Wang, B.V. Hooreweder, J.P.
(2014) 2660–2667. Kruth, J.V. Humbeeck, J. Mater. Process. Tech. 238 (2016) 437–445.
[148] Z. Wang, K. Guan, M. Gao, X. Li, X. Chen, X. Zeng, J. Alloy Compd. 513 (2012) [162] J.H. Martin, B.D. Yahata, J.M. Hundley, J.A. Mayer, T.A. Schaedler, T.M.
518–523. Pollock, Nature 549 (2017) 365.
[149] H. Gong, K. Rafi, H. Gu, G.D.J. Ram, T. Starr, B. Stucker, Mater. Des. 86 (2015) [163] H. Zhang, H. Zhu, T. Qi, Z. Hu, X. Zeng, Mater. Sci. Eng. A 656 (2016) 47–54.
545–554. [164] H. Zhang, H. Zhu, X. Nie, T. Qi, Z. Hu, X. Zeng, in: B. Gu, H. Helvajian, A. Pique
[150] R.I. Stephens, A. Fatemi, R.R. Stephens, H.O. Fuchs, J. Mater. Sci. Technol. 103 (Eds.), Fabrication and Heat Treatment of High Strength Al-Cu-Mg Alloy
(1981) 415–432. Processed Using Selective Laser Melting, E-Publishing, Inc., San Francisco,
[151] S. Suresh, Fatigue Fract. Eng. M. 28 (1998) 1153–1160. 2016, pp. 97380X1-97380X7.
[152] H.R. Ammar, A.M. Samuel, F.H. Samuel, Int. J. Fatigue 30 (2008) 1024–1035. [165] E.M. Elgallad, P. Shen, Z. Zhang, X. Chen, Mater. Des. 61 (2014) 133–140.
[153] G. Qian, Y. Hong, C. Zhou, Eng. Fail. Anal. 17 (2010) 1517–1525. [166] H. Zhang, H. Zhu, X. Nie, J. Yin, Z. Hu, X. Zeng, Scripta Mater. 134 (2017)
[154] A. Leon, E. Aghion, Mater. Charact. 131 (2017) 188–194. 6–10.
[155] D. Qiu, M. Zhang, J. Alloy Compd. 586 (2014) 39–44. [167] F.M. Ghaini, M. Sheikhi, M.J. Torkamany, J. Sabbaghzadeh, Mater. Sci. Eng. A
[156] S. Hegde, K.N. Prabhu, J. Mater. Sci. 43 (2008) 3009–3027. 519 (2009) 167–171.
[157] M. Asta, C. Beckermann, A. Karma, W. Kurz, R. Napolitano, M. Plapp, G. [168] E.A. Jägle, Z. Sheng, L. Wu, L. Lu, J. Risse, A. Weishelt, D. Raabe, JOM-US 68
Purdy, M. Rappaz, R. Trivedi, Acta Mater. 57 (2009) 941–971. (2016) 943–949.

View publication stats