Materials Science & Engineering R 146 (2021) 100648

Contents lists available at ScienceDirect

Materials Science & Engineering R

journal homepage: www.elsevier.com/locate/mser

Additive manufacturing of metallic lattice structures: Unconstrained

design, accurate fabrication, fascinated performances, and challenges
Liang-Yu Chen a, Shun-Xing Liang c, Yujing Liu d, Lai-Chang Zhang b, *
a
School of Materials Science and Engineering, Jiangsu University of Science and Technology, Zhenjiang, Jiangsu, 212003, China
b
School of Engineering, Edith Cowan University, 270 Joondalup Drive, Joondalup, Perth, WA, 6027, Australia
c
Technical Chemistry I and Center for Nanointegration Duisburg-Essen (CENIDE), University of Duisburg-Essen, Universitaetsstrasse 7, Essen, 45141, Germany
d
School of Materials Science and Engineering, Changsha University of Science and Technology, Changsha, Hunan, 410114, China

A R T I C L E I N F O A B S T R A C T

Keywords: Lattice structures, which are also known as architected cellular structures, have been applied in various industrial
Additive manufacturing sectors, owing to their fascinated performances, such as low elastic modulus, high stiffness-to-weight ratio, low
Lattice structures thermal expansion coefficient, and large specific surface area. The lattice structures fabricated by conventional
Metallic materials
manufacturing technologies always involve complicated process control, additional assembly steps, or other
Topography optimization
Bioimplant
uncontrollable factors. Furthermore, limited types of unit cells can be used to construct lattice structures when
Energy absorber using conventional processes. Fortunately, additive manufacturing technology, based on a layer-by-layer process
from computer-aided design models, demonstrates the unique capability and flexibility and provides an ideal
platform in manufacturing complex components like lattice structures, resulting in an effective reduction in the
processing time to actual application and minimum of material waste. Therefore, additive manufacturing relieves
the constraint of structure design and provides accurate fabrication for lattice structures with good quality. This
work systematically presents an overview of conventional manufacturing methods and novel additive
manufacturing technologies for metallic lattice structures. Afterward, the design, optimization, a variety of
properties, and applications of metallic lattice structures produced by additive manufacturing are elaborated. By
summarizing state-of-the-art progress of the additively manufactured metallic lattice structures, limitations and
future perspectives are also discussed.

structure is an porous network in which the pores are interconnected

[2]. By contrast, the pores in closed-cell cellular structure are isolated.
1. Introduction Both of them have significantly lower weight than the solid materials in
the identical volume [3]. Natural cellular structures have been found
In human history, bulk materials, such as metals, wood, ceramics, with combined advantages of low density, high specific strength, and
and glasses, have been widely used for thousands of years. Among these multifunctionality [4,5]. As an example, the bones of birds are light­
materials, metals as easily approachable resources usually possess weight cellular structures but with a high specific strength, which fa­
attractive properties, such as high strength and large ductility, high cilitates their flying [6]. Although human-made bulk metals and alloys
workability, high melting point, desired chemical and dimensional sta­ are superior to the natural biological-based materials in the mechanical
bility, and excellent thermal/electrical conductivity [1]. Metals have properties, the synthetic cellular structures with complex geometries are
been playing a prominent role in many industrial sectors, including less sophisticated. As such, cellular structures made of metals and alloys
military, traffic, architecture, energy industry, biomedical industry, etc. (and other materials as well) have received considerable attention and
However, bulk metallic materials are frequently used in the past. Rare have been well developed in recent years [7–9].
metallic cellular structures had been used in the past [1]. Cellular Metallic cellular structures have the advantages of lower fuel con­
structures are usually highly porous, which have been defined as the sumption, larger surface area, better mechanical and other perfor­
solid materials containing a high volume fraction of pores. According to mances in a wide range of industries, such as automobile, aerospace,
the pore type, the cellular structures can be classified as open-cell sports, and biomedical sectors [10–13]. Traditionally, cellular structures
cellular structure and close-cell cellular structures. Open-cell cellular

* Corresponding author.
E-mail addresses: [email protected], [email protected] (L.-C. Zhang).

https://doi.org/10.1016/j.mser.2021.100648
Received 11 August 2021; Received in revised form 24 September 2021; Accepted 13 October 2021
Available online 28 October 2021
0927-796X/© 2021 Elsevier B.V. All rights reserved.
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Nomenclature FDMS fused deposition modeling and sintering

FFF fused filament fabrication
3D three-dimensional HIP hot-isostatic pressing
β-TCP β-tricalcium phosphate L-PBF laser powder bed fusion
AM additive manufacturing LENS laser engineered net shaping
AFCC all face-centered cubic LOM laminated object manufacturing
BJ binder jetting LSM level set method
BCC body-centered cubic Micro-CT micro computed tomography
BCCZ BCC with Z strut MMCs metal matrix composites
BESO bi-directional evolutionary structural optimization PBF powder bed fusion
CAD computer-aided design PEEK printing poly-ether-ether-ketone
CFRP carbon fiber reinforced plastics PFCC face-centered cubic with Z-strut
CLAD 3D laser cladding PLA polylactic acid
CLS conformal lattice structure PLGA polylactic-co-glycolic acid
CSAM cold spray additive manufacturing PM powder metallurgy
DED directed energy deposition RP rapid prototyping
DIW direct ink writing r-SBF revised simulated body fluid
DMD direct metal deposition SEM scanning electron microscopy
EBM electron beam melting SIMP solid isotropic material with penalization
ECAM electrochemical additive manufacturing SLM selective laser melting
ESO evolutionary structural optimization SMA shape memory alloy
F2BCC two face-centered cubic combined in a BCC way SLS selective laser sintering
FBCCZ FCC with BCCZ TEM transmission electron microscopy
FBCCZXYZ FBCCZ with X- and Y- strut TPMS triply periodic minimal surface
FCC face-centered cubic UAM ultrasonic additive manufacturing
FCCZ face-centered cubic with Z-strut WAAM wire and arc additive manufacturing
FDM fused deposition modeling

are manufactured by investment casting [14], wire-woven methods [28,29], and composites [30,31], can be used as the feedstock. There­
[15], foaming [16], powder metallurgy [17], and other methods [13]. fore, significantly broad applications, including aerospace [32], auto­
However, the conventional manufacturing methods have apparent re­ mobile [33], biomedicine [34], construction [35], catalyst [36], and
straints in fabricating complex designed structures and/or the possibility other industrial sectors [37], have been achieved. Combining these ad­
of architectures. In brief, the produced cellular structures by conven­ vantages of AM, multi-scale cellular structures with unprecedented ge­
tional manufacturing technologies have two main disadvantages: (i) ometries are readily produced and exhibit better performance. For
stochastic structure, and (ii) limited architecture for use. On the one example, the AM manufactured AlSi10Mg foams with a cellular struc­
hand, due to the imprecisely uncontrolled pores and wall thicknesses, ture and a porosity of 86.5 % achieve a 60 % higher heat transfer co­
stochastic cellular structures have unstable performances, which is un­ efficient compared with the conventionally manufactured metal foams
desirable in actual applications. On the other hand, non-stochastic with stochastic structures [38]; the transportation of reactants is
structures produced by conventional manufacturing methods usually improved in three-dimensional (3D) catalysts with cellular structures
require further assembly or bonding steps, which indeed complicates the because a cellular structure offers a larger surface area and open pores
manufacturing process and even renders difficulty in manufacturing for more active sites, resulting in enhanced catalytic performances
cellular structures, thereby limiting the availability of types of archi­ [39–41].
tectures. Given the fact that these disadvantages largely limit the de­ The exact control of the performances of cellular structures with
velopments and applications of metallic cellular structures, the cellular stochastic geometries is still difficult. To overcome this disadvantage of
structures with non-stochastic geometries are expected to be manufac­ cellular structures, lattice structure is therefore developed. A lattice
tured by a more easy-operation, time-saving, and structure-flexibility structure, which is a type of cellular structures with non-stochastic ge­
method. ometry, has an architecture formed by regularly geometric arrange­
Fortunately, the development of additive manufacturing (AM) ments of periodic unit cells over a space [42]. The lattice structure is
technologies enables the achievement of cellular structures with regular previously defined as a framework with non-stochastic structural
geometries. AM technologies, which build components in a layer-by- crossed materials and typically developed from truss structures [43].
layer way according to their corresponding computer-aided design However, the configurations of truss structures are limited. Developing
(CAD) models [18–20], were firstly invented in the 1980s. These AM more types of unit cells may upgrade the lattice structures with better
technologies get rid of the constraints from conventional manufacturing and desired properties [44]. Unit cells that are used to construct lattice
methods and could fabricate the components with more complex ge­ structures are arranged according to specific space groups. Therefore,
ometries. The entire AM process only takes a few steps from the many types of lattice structures have unit cells with the basic structures
modeling in a computer to direct production by AM equipment with of body-centered cubic (BCC), face-centered cubic (FCC), and their de­
minimal machinery [21]. Therefore, the costs of labor, materials, and rivatives (more types of lattice structures would be introduced in Section
time are greatly saved. With the assistance of emerging AM technolo­ 4). Of course, the unit cells may not be merely cubic but they fill a
gies, more architectures from designed models with unprecedented ge­ desired volume periodically. As such, the repeatable unit cells are
ometries are realized. Meanwhile, thanks to the highly improved important for lattice structures. Now, the lattice structure has been
accuracy by AM, the designed model can be achieved to the greatest known to be a non-stochastic structure in three-dimensional space and is
extent [22,23]. With the development of AM technologies, different named architected cellular structure. The differences in the stochastic
types of materials, such as metals [24,25], ceramics [26,27], polymers and non-stochastic structures are exampled in Fig. 1. Fig. 1a is a Mg

2
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 1. Illustrations of stochastic and non-stochastic structures: (a) a Mg foam produced by direct foaming (reproduced with permission from Ref. [45], Copyright
(2016), Elsevier), and (b) a schematic of a lattice structure model with 4 × 4 × 4 tetrakaidecahedron cells (reproduced with permission from Ref. [46], Copyright
(2015), Elsevier).

foam with a stochastic feature produced by direct foaming, while Fig. 1b research on topology optimization and AM technologies in recent years
is a schematic of a lattice structure model with 4 × 4 × 4 tetrakaide­ are reviewed; multi-scale or hierarchical structural optimization and
cahedron cells [45,46]. Non-stochastic feature endows the lattice topology optimization regarding AM constraints are systematically
structures with their high controllability in architectural characteristics summarized. Guo et al. [18] provided a comprehensive understanding of
and superior properties, including low elastic modulus [47,48], negative AM technologies and materials used; the starting materials can be
Poisson ratio [49,50], high stiffness-to-weight ratio [51,52], low ther­ categorized into liquid, filament/paste, powder, and solid sheet, and the
mal expansion coefficient [53,54], large surface area [55,56], tremen­ types of materials can be classified into polymers, metals, ceramics, and
dous internal pores [57–59], etc. Such superior properties make lattice composites. Mahmoud et al. [74] gave an overview of lattice structures
structures be applied in many fields, such as biomedical engineering and functionally graded materials applications in AM of orthopedic
[21,60–63], aerospace and automotive lightweight structures [51,52, implants. This review discusses how both technologies can significantly
64], energy absorber [65–67], heat exchanger [68,69], acoustic insu­ enhance the performances of implants from a mechanical and biological
lator [57–59,69], and catalytic applications [55,56,70]. It is worth point of view. Similarly, the reviews regarding the applications of lattice
noting that lattice structures made from metals or other artificial syn­ structures produced by AM technologies in specific fields have also been
thetic materials exhibit better performances than cellular structures to reported [75–77]. Nevertheless, an overview of metallic lattice struc­
some extent. Therefore, lattice structures produced by AM technologies tures, which has comprehensive coverage of the birth, design, perfor­
have received increasing attention during recent years. mances, and applications as well as the manufacturing by conventional
In recent years, several review articles have addressed the research manufacturing technologies and additive manufacturing technologies, is
progress and outcomes concerning the lattice structures produced by still absent. As such, a timely review addressing the recent advances and
AM technologies. For example, Helou et al. [71] systematically reviewed proposing insightful perspectives in the field of additive manufacturing
the design, analysis, and manufacturing of lattice structures. This review focusing on metallic lattice structures is highly desirable. In this article,
concluded the significant limitations in the lattice structure definition a brief introduction to the concept of lattice structures and their features
and proposed a re-definition of lattice structures to incorporate new and applications are first presented. By comparing the characteristics of
literature recently produced. Zhu et al. [72,73] dedicated to analyzing conventional cellular structures, the definition of lattice structures is
the topology optimization of lattice structure as well as other parts stressed. Afterward, because the manufacturability of AM technologies
produced AM technologies. The main content and applications of leads to the tremendous development of metallic lattice structures, the

3
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

detailed introduction and comparison are made between conventional foams. The blowing agents decompose in the melt and release gas to
manufacturing technologies and additive manufacturing technologies. produce bubbles [16]. Taking the fabrication of Al foam as an example,
The corresponding design methods and optimization methods of lattice Ca is added into an Al melt at 680 ◦ C and the viscosity of melt can be
structure are also reviewed. Finally, a variety of the applications of increased by a factor of up to five because of the generation of CaO and
metallic lattice structures in different fields are elaborated. Based on the CaAl2O4 [8,96]. After the viscosity of melt reaches the required value,
reviewed findings, the limitations/challenges and future opportunities titanium hydride (TiH2), which is used as the blowing agent, is added to
of AM technologies and metallic lattice structures are also discussed. the melt to produce H2 in the hot viscous liquid [97]. The other steps of
This review would provide an insight to the additive manufacturing of the process are similar to the foaming by gas injection. The homogeneity
metallic lattice structures and deliver huge information from the birth, of produced foam can be improved by adjusting the processing param­
design, manufacturing, and performance of lattice structures to their eters. Similar to TiH2, zirconium hydride (ZrH2) and tungsten powder
various applications. have also been employed as blowing agents [98,99]. The foam produced
by this method has a trading name “Alporas”, and is more homogeneous
2. Conventional technologies for metallic cellular and lattice than the counterparts produced by gas injection method. However, it is
structures still difficult to control the gas bubbles in the melt due to their random
distribution. Meanwhile, the sizes of gas bubbles are significantly
2.1. Direct foaming different. Although “Alporas” has a better homogeneous pore structure
in macroscale, the precise regulations of the size and distribution of
The direct foaming method, which is a liquid state process, produces pores are still unavailable.
metal foams by creating gas bubbles in the metallic melts and conserves The induction of pores is used to tailor the mechanical properties of
the profile of the corresponding liquid metallic foams [78,79]. By metals. For metallic foams (as well as other metallic porous materials),
injecting the compressed gas, the created gas bubbles are prone to rise to the mechanical properties, including compressive strength, fatigue
the surface of metallic melts due to the buoyancy forces. However, the behavior, elastic modulus, and flexural properties, are the important
high viscosity of the molten metal may hamper the rise of gas bubbles. considerations for their performance [100–102]. Unlike bulk materials,
Therefore, fine ceramic powder or alloying elements, which are favor­ pores and voids are present in the cellular structures. The definition of
able for the formation of stable particles, are commonly used to enhance porosity is the volume fraction of pore/void in a solid material [21]. In
the viscosity of the metallic melt. The earliest trace of the direct foaming such a situation, an important concept, which is named as relative
method was in the 1940s [80]. Afterward, different metals and alloys, density, should be understood. The relative density (ρs ) of a cellular
such as Al [78,81,82], Cu [83–85], Ni [86,87], and Mg [88], have been structure is calculated as the mass of the cellular structure divided by its
used to produce metal foams by this method. Generally, the direct volume. In some reports, relative density is also called the density of a
foaming method can be classified into foaming by gas injection [89,90] porous structure [21,103]. The relationship between porosity and rela­
and foaming with blowing agents [91,92]. In both methods, gas bubbles tive density can be expressed as:
are created by gas or blowing agents. ρs
Foaming by gas injection requires the preparation of metallic melt porosity = (1 − ) (1)
ρ
containing SiC, or oxide particles. It is similar to the preparation of or­
dinary metal matrix composites (MMCs). The metallic melt should have where ρ is the density of the bulk material. Gibson and Ashby [104,105]
a uniform distribution of the reinforcements and the reinforcements are devoted to investigating the mechanical properties of cellular structures
fully wetted by melts. Fig. 2 presents a schematic illustration of foaming and developed a series of equations to describe the relationship between
by gas injection [80]. The first step requires the preparation of metallic the relative density of cellular structures and their mechanical proper­
melt in a container. In the second step, the foam is manufactured by ties, which are listed in Table 1. Meanwhile, the distribution and sizes of
injecting gases (air, nitrogen, argon) into melt using specially designed pores would lead to anisotropy in the mechanical properties and stress
rotating impellers or vibrating nozzles [93–95]. The impellers or nozzles concentration. Edwin et al. [106] prepared closed-cell Al foams and they
are used to produce fine bubbles in the melt and to homogeneously found that the foam with relatively large pores has a more apparent
distribute these fine bubbles. The mixture of bubbles and melt floats up anisotropy in the mechanical properties. Such a finding is ascribed to the
to the surface of the liquid and turns into a relatively dry liquid foam. In random distribution of pore size and position in the foam. The direct
such a situation, the semi-solid foam can be pulled out from the liquid foaming technology has still encountered difficulty in the control of
melt. When the semi-solid foam cools down and solidifies, this process is pores. Wang et al. [107] investigated the influence of pores on the crack
finished. initiation in tensile and cyclic loadings of a A319 alloy foam. Compared
Using blowing agents is an alternative way to produce metallic to small pores, large pores provide lots of strain concentration zones for

Fig. 2. A schematic illustration of foaming by gas injection. (reproduced with permission from Ref. [80], Copyright (2001), Elsevier).

4
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 1 strength. Meanwhile, the meshes with the honeycomb unit cells show
Equations for mechanical properties of metal foams proposed by Gibson and the highest bending resistance, while the meshes with the square unit
Ashby. ρ is the density of bulk material; ρs is the relative density of cellular cells demonstrate the lowest bending resistance. Although the products
structure; E and Es are the elastic moduli of bulk and cellular structure respe­ in Ref [112]. have the rudiment of the lattice structures, the wax pat­
tively; σ c and σc,s are the compressive yield strength of bulk and cellular struc­ terns are still too coarse (in millimeter) to fabricate products with fine
ture respectively; σ t is the tensile strength; G is the shear modulus; εD is the
and complex structures.
densification strain (reproduced with permission from Ref. [100], Copyright
However, fabricating customized or complex components needs
(2012), Elsevier).
huge tooling costs for producing wax patterns in conventional invest­
Property Open-cell cellular structure Closed-cell cellular structure ment casting. Until recent years, most wax patterns are still handmade
Elastic modulus E/ES = (0.1 − 4)∙(ρ/ρs )2 E/ES = (0.1 − 1.0)∙[0.5∙ by experienced people [113]. A skilled toolmaker should estimate a
(ρ/ρs )2 + 0.3∙(ρ/ρs )] variety of mold designs because design errors or repeating processes are
Compressive σc /σc,s = (0.1 − 1.0)∙ σc /σc,s = (0.1 − 1.0)∙[0.5∙ usually costly and time-consuming. Especially for the parts with
yield strength
(ρ/ρs )3/2 (ρ/ρs )2/3 + 0.3∙(ρ/ρs )] extremely complex geometry (e.g., cellular structure or lattice struc­
Tensile strength σt = (1.1 − 1.4)∙σc σt = (1.1 − 1.4)∙σc ture), it is rather difficult to produce patterns for conventional invest­
Shear modulus G = 3/8∙E G = 3/8∙E ment casting. In this case, rapid prototyping (RP) techniques become
Densification εD = (0.9 − 1.0)∙[1 − 1.4∙ εD = useful tools to assist the design and manufacturing in conventional in­
strain (ρ/ρs ) (0.9 − 1.0)∙[1 − 1.4∙(ρ/ρs )3 ] vestment casting [113]. RP techniques also belong to AM, which can
+ 0.4∙(ρ/ρs )3 ] directly convert a 3D CAD model into a solid physical model (the
detailed information with respect to additive manufacturing would be
introduced in Section 3.) [114]. Using RP techniques to produce patterns
crack initiation in both tensile and cyclic loadings. Because lattice is greatly favorable for the design and manufacturing in conventional
structures have a periodic distribution of unit cells, most unit cells in a investment casting. With the assistance of RP techniques, producing
lattice structure are theoretically identical. However, the direct foaming lattice structures from investment casting also becomes possible. Such a
method creates pores in the metallic melt via a stochastic manner, method is also named additive manufacturing assisted investment
depiste blowing gas or blowing agents being used. Meanwhile, the casting [115]. Xue et al. [116] designed auxetic unit cells with different
formed pores in metallic melt are still mobilizable. Hence, after the so­ parameters (Fig. 3a and b) and fabricated a lattice structure pattern by
lidification of metallic melt, the pores have a random distribution. 3D printing using photosensitive resin (Fig. 3c). The fabricated photo­
Moreover, the sizes of pores are non-uniform. Although the use of im­ sensitive resin lattice structure pattern is placed in a stainless-steel
pellers or nozzles can reduce the difference in the sizes of pores, a wide container and plaster slurry is poured to fill the space between the lat­
distribution of the sizes of pores is still found. As a result, such features in tice pattern and container. After the consolidation of the plaster slurry,
direct foaming lead to the significant difficulty in the production of the pattern is burned up. As such, a plaster shell mold of lattice structure
lattice structures. is completed. Afterward, molten Al is poured into the mold. The
stainless-steel container is sealed and compressed air is fed to make the
2.2. Investment casting molten Al fill the space of mold. Finally, a lattice structure is produced.
The compressive tests indicate that both the elastic modulus and
It is well accepted that investment casting can produce parts with a compressive strength increase with decreasing the strut length (Fig. 3d).
good surface finish, dimensional accuracy, and complex shapes [108]. Such a result can be easily understood by relating the strut length to the
Therefore, investment casting has been used to fabricate weapons, relative density of lattice structures. Fig. 3e is the stress-strain curves of
jewelry, and art casting in the past. Generally, a wax model of the part is the lattice structure with different diameters. With increasing the strut
made at the initial stage of investment casting. A few wax models are diameter, both the relative density of lattice structure and the
assembled and rammed in a sand mold. The sand mold is subsequently compressive strength increase. The investment casting with the assis­
heated so that the wax can be drained out from the sand mold to produce tance of additive manufacturing can successfully produce lattice struc­
a cavity inside. The production of the cavity further allows the molten tures with tailorable properties. However, the unit cells in the produced
metals to be poured inside. After cooling, the solidified metals are de­ Al lattice structure have a relatively large size, which is in the millimeter
tached from the mold for subsequent post-processing. Owing to their level. To ensure the filling of molten metals in the mold, the pattern for
sufficient green strength/fired strength, high thermal shock resistance, lattice structure cannot be very small in the size of unit cell. In short,
chemical stability, and limited creep, ceramic shells could withstand the investment casting is a promising technology for producing many
weight of cast metal and prevent cracking during metal pouring. metallic parts. The principle of investment casting is mold making and
Therefore, ceramic shell investment casting is subsequently developed melt pouring. So far, the molds for lattice structures may be fabricated
[109,110]. The high chemical and shape stability of ceramic shells also by RP technologies. However, elaborated lattice structures with signif­
provides high precision for the parts with complex geometries (e.g., icantly thin walls or thin struts are still not yet reported. The difficulties
turbine blades, turbocharger wheels, and hip replacement implants) in producing such elaborated structures are the fabrication of mold and
[109,110]. Cellular structures are also produced by investment casting. pouring the melt into the mold. Therefore, the advances of technologies
Brothers et al. [111] used polyurethane foams as the precursors to in these two aspects may facilitate the production of various lattice
produce open-cell metallic foams with controlled and continuous den­ structures by investment casting.
sity gradients by investment casting. The produced cylindrical sample
has different densities of two ends. The sizes of pores in the sample have 2.3. Wire-woven methods
a graded distribution from one side to the other side. However, all pores
are in irregular shapes and the control in the sizes of pores is unfavor­ Wire-woven methods, which are also known as metal wire ap­
able. Montenegro et al. [112] produced three NiTi shape memory alloy proaches, use wires and rods as the raw materials to weave metallic
(SMA) meshes with honeycomb, square, and circle unit cells by invest­ structures with a repetitive symmetric geometry, producing multi-
ment casting using wax mesh patterns. The produced meshes have layered structures with fine cells [117–119]. Using metallic wires
similar sizes to the wax patterns. The tensile behavior of the produced and/or rods as the raw materials has several advantages. First, the ma­
meshes is different due to the difference in unit cells. The meshes with terials with few defects are easily obtained from wires and/or rods, and
the circle unit cells have the highest tensile strength and stiffness, while almost all metals can be processed into wires and rods [120]. Second,
the meshes with the honeycomb unit cells exhibit the lowest tensile wires and/or rods are prone to be handled in the manufacture of cells.

5
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 3. Auxetic lattice structures produced by additive manufacturing assisted investment casting: (a) designed parameters, (b) illustration of unit cells, (c) a 3D
auxetic photosensitive resin pattern, (d) a 3D auxetic Al lattice structure, and (e) the axial section of the produced structures, and the stress-strain curves of the lattice
structure with (f) different strut length L and (g) different diameter D (reproduced with permission from Ref. [116], Copyright (2018), Elsevier).

Lots of work has reported the metallic meshes fabricated by wires discontinuous, machining on its external surface is frequently required.
and/or rods using conventional wire-woven methods [119,121]. Post-treatment is often optional. For example, brazing often induces
Wire-woven metals have periodic structures and therefore exhibit better annealing of the wires, resulting in a degradation in the strength of
capability in heat dissipation [122], absorption of vibration [123], and wire-woven metallic structures (e.g., carbon steel and Al alloy) [127].
lightweight load [124] compared to bulk metals. Regardless of raw Therefore, quenching, and aging processes are carried out to enhance
materials used, the topologies of wire-woven metals determine their the strength of wire-woven metallic structures. In addition, painting
mechanical properties [117]. To fabricate wire-woven metals, a and/or plating may be employed to endow the wire-woven metallic
textile-based weaving process is employed, and a specific pattern is structures with better corrosion resistance [130].
created by wire/rod segments [121,125]. In this specific pattern, plenty Thanks to their low density, high specific strength, high elastic
of plain-woven meshes are stacked node to node and connected to modulus, and other useful properties, wire-woven metallic structures
produce a core of lattice structure [15,121,126,127]. have been employed in many industrial applications, such as sandwich
The fabrication processes of the wire-woven metals assembled by core [131], heat transfer media [132], brake disk core [133], catalyst
such patterns are similar, which mainly include three steps, i.e., support [134], reinforcements of matrices [125], and seismic dampers
assembling, fixing, and machining [126]. Of course, some treatments, [135]. However, the limited patterns in wire-woven metallic structures
such as pre-forming, post-forming, and post-treatment, may be cannot meet the requirement of lattice structure with complex geome­
employed in the above three steps. In the pre-forming step, wires are try. 3D metallic lattice structures were reported to be produced by
formed to a required shape according to the a specifically designed joining bending wires/rods [119,136]. By this method, many unit cells,
pattern [126]. In the assembling step, the pre-formed wires are assem­ including cuboid, pyramidal, hexagon, truss-like, prismatic,
bled to be meshes. The assembling process is always conducted by an body-centered cubic, and face-centered cubic cells, are constructed. The
assembling machine and a specific pattern generally has a correspond­ produced metallic lattice structures have a significantly low relative
ing assembling machine [117]. Post-forming, which may be employed density (< 10 %) and comparable strength depending on the design,
after the assembling step, is used to further deform the meshes into the bending accuracy, optimum dipping mixture [119]. Khoda et al. [119]
required structure [13]. After the assembling, the structure is fixed at the reported the AISI1006 low carbon steel lattice structures produced by
intersections between the wires to acquire its best performance in a way wire-woven methods. Fig. 4a shows the fabricated physical products.
(e.g., strength). The commonly used methods for fixing are soldering, The relative densities are 8.6 % and 6.8 % for 8 mm cell, 6.4 % and 5.3 %
brazing, and transient liquid phase bonding. [117,128,129]. After the for 10 mm cell, and 5.0 % and 3.8 % for 12 mm cell with 2 × 2 × 1 and 2
fixing, the intersections of the structure are bonded. Because the surface × 2 × 2 cells, respectively. Fig. 4b and c are the stress versus strain
of a wire-woven metallic structure may be inhomogeneous and curves of 2 × 2 × 2 cell and 2 × 2 × 1 cell lattice structures for Z-axis

6
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 4. AISI1006 low carbon steel lattice structures produced by the wire-woven method: (a) the fabricated products, and stress versus strain curves of (b) 2 × 2 × 2
cell and (c) 2 × 2 × 1 cell lattice structures for Z-axis loading. σ is the nominal strength, ε is the plastic strain and E is the elastic modulus (reproduced with permission
from Ref. [119], Copyright (2021), Springer).

loading. It is found that the strength and elastic moduli of the lattice metallic parts by pressing and sintering with few chemical reactivity
structures decrease with increasing the cell size or reducing the relative constraints at low temperatures [137–139]. During the PM process, a
density. In addition, the strength, plastic strain, and elastic modulus for homogeneous mixture is prepared by mixing the metallic powder with
all lattice structures with 2 × 2 × 1 cells are higher than the corre­ lubricants or additives [140]. Afterward, the mixture is compacted in a
sponding ones for all lattice structures with 2 × 2 × 2 cells, which is die at pressures between 140 and 950 MPa, depending on the densities
attributed to the increased relative density of the lattice structures with and compressibility of the materials. In the sintering step, the compact is
2 × 2 × 1 cells. Note that, the wire-woven methods involve too many placed in a controlled-atmosphere furnace. The compact is heated to the
steps to produce lattice structure. Because the lattice structures are made sintering temperature. After holding at the sintering temperature for
of one or several metallic wires or rods, the sizes (especially the diam­ several hours, the compact is cooled down in furnace. As such, a bulk
eter) of wires or rods would play an important role in determining the part is produced. Such a technology is significantly useful for producing
final dimensions of the produced lattice structure. Hence, the lattice bulk metallic materials, including Ti [141–145], Al [146–148], Mg
structures with complex geometries and fine pores are still inaccessible. [149,150], steel [151,152], high/medium entropy alloy [153,154], etc.
On the other side, if the wires or rods are significantly fine (down to However, although PM has relatively precise control of process
several microns) in diameter, the fabrication processes would become variables, it is still significantly difficult in the production of lattice
extremely difficult. Moreover, because the metallic wires or rods cannot structures. Therefore, PM techniques are commonly employed to pro­
be self-bonded, post-treatment is always employed, which decreases the duce porous metallic materials with stochastic structures. It should be
production efficiency. Most important of all, the configurations of lattice noted that many investigations have shown the production of porous
structure are significantly limited when wire-woven methods is used. metallic materials by loose sintering of powder [155], space holder
method [156–158], spark plasma sintering [159,160], metal injection
2.4. Powder metallurgy with space holder molding [161]. Among these technologies, the space holder method has
relatively higher controllability for producing cellular structures. To
Powder metallurgy (PM), which is a solid-state process, can produce produce cellular structure by the space holder method, five main steps

7
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

are involved: powder selection, mixing, compaction, sintering, and arrangement of space holders is still random although they seem to be
removal of the space holders. Spacer holders, which cannot react with homogenous in macroscale. Therefore, merely using the space holder
the base material, should be selected first. Space holders (particles) are method to produce a lattice structure becomes less possible. On this
employed as pore formers to generate pores in metallic cellular struc­ basis, Ryan et al. [170] have improved the manufacturing possibility by
tures [162–164]. Hence, the selection of space holders leads to the building a porous wax template based on rapid prototyping and
controllability of porosity parameters, such as the morphology, rough­ combining powder metallurgy to produce a Ti lattice structure, as shown
ness, and volume fraction of pores. Generally, polymethyl methacrylate, in Fig. 6. The size and distribution of pores produced by combining
ammonium hydrogen carbonate, TiH2, NaCl, (NH2)2CO, and C12H22O11 powder metallurgy and rapid prototyping are relatively uniform
are frequently used as space holders (pore-forming agents) and polyester compared to the conventional powder metallurgy with space holders.
resin is commonly used as the binder [101]. Afterward, metallic powder, Such a method provides a new idea to produce the parts with relatively
space holders, and blinders should be evenly mixed, which determines regular geometries. However, a porous wax template should be pro­
the pore distribution and mechanical properties of the cellular structure. duced first. Therefore, extra equipment (or instruments) is required. In
The mixture is then compacted in a die to generate a green body (the the meanwhile, as mentioned in Section 2.1, the wax template is not
compacted powder before sintering). Some compaction methods have enough elaborate to fabricate lattice structures with fine pores and
been employed, such as uniaxial die pressing [165], isostatic pressing complex geometries. Totally, like the AM-assisted investment casting,
[166], and injection molding [167]. The pressure should be controlled using powder metallurgy technology meets the difficulties in pattern
to achieve a balance between the shape of compacted space holders and making. The patterns with elaborated structures may be produced by
the strength of the cellular structure. Subsequently, the space holders are other technologies, such as RP technology. However, the powder spread
removed by heat treatment [168]. This step determines the purity and in the elaborated structures is extremely difficult. On the one hand, the
the strength of the cellular structure produced. Finally, the cellular powder is always spherical or nearly spherical and void is inevitable in
structure is sintered and its mechanical properties can be tailored by the powder stacking. This is a generic problem in powder technology.
controlling the sintering temperature [169]. On the other hand, the decrease in the particle size of powder can reduce
Fig. 5 shows the typical Ti parts produced by powder metallurgy with the void in the powder stacking as well as the produced parts. However,
the space holders. It can be found that the rigorous uniform size and the cost of the powder with significantly low size may increase the total
distribution of pores are still not be achieved. The pores are produced by cost of the final products. As it is well known, among the mainstream
the space holders. However, space holders are not identical but have a metallic feedstocks avaible in powder market, the cost of powder is
distribution in the size range. Meanwhile, during the mixing process, the significantly higher than that in other forms (such as wires, rods, and

Fig. 5. Osteosinter samples produced by space holder method and the porous characterization via micro-CT (reproduced with permission from Ref. [168], Copyright
(2019), Elsevier).

8
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 6. A Ti lattice structure produced by combining rapid prototyping and powder metallurgy: (a) top and side profiles of a Ti lattice structure with 59.1 % porosity
and (b) scanning electron micrographs of Ti lattice structures with the pore sizes of 200, 300, and 400 mm. (reproduced with permission from Ref. [170], Copyright
(2008), Elsevier).

sheets). Therefore, powder metallurgy technology is incapable of pro­ 3. Additive manufacturing technologies for metallic lattice
ducing lattice structures with complex geometries. structures

3.1. An overview for additive manufacturing technologies

2.5. Other technologies
Additive manufacturing (AM), which is commonly referred to as 3D
In addition to the aforementioned methods, other technologies have
printing, has been rapidly developing over the last three decades. In
been employed to manufacture lattice structures, such as high-
comparison to “additive manufacturing”, conventional manufacturing
temperature forming and diffusion bonding [171], electroless plating
(also most called subtractive manufacturing) requires the processing
[172], sacrificial mold coating [173], interlocking grid assembly [11,
schedules to produce a component, which more or less wastes the ma­
174,175], eutectic reaction [176,177], folding of a perforated metal
terials used. The AM process has almost no or limited materials waste,
sheet [178,179] and lattice block by stacking [180]. Fig. 7 shows the
attributed to its layer-wise additive nature of the manufacturing
lattice structures manufactured by the above methods. However, such
manner. The AM process produces 3D parts from their CAD models in a
methods always have disadvantages on the precision controllability,
layer-by-layer mode. Before AM processing, the CAD model to be used
fabrication of complex designed structure, or the requirement of further
for AM is first sliced by a defined layer thickness. A serial of slices are
assembly. Furthermore, nodes are commonly disconnected in the
produced in sequence and constructed into a 3D part that is presented by
manufacturing processes [11,174,176,181]. Such shortcomings hamper
the CAD model used. In this way, a complex 3D part could be con­
the manufacturing and development of lattice structures.
structed by a series of two-dimensional (2D) slices. Therefore, additive
Table 2 summarizes some conventional methods for producing
manufacturing can realize those CAD designs with complex dimensions
metallic cellular structures or lattice structures as well as the main
that could not be produced by conventional manufacturing technolo­
features of these methods. In brief, all conventional methods show
gies. As a result, additive manufacturing provides a better solution to the
inherent limitations, such as the presence of impurity phases, low
parts having complex geometry and the pre-designed CAD models
porosity level, and unfavorable control over the size, shape, and distri­
extend the architectures of the end-use parts, leading to superior flexi­
bution of the pores in the lattice structures. Hence, new technologies
bility than conventional processing for producing parts with geometric
that can overcome these shortages are urgently required. The develop­
and material complexities, especially for producing lattice structures.
ment of additive manufacturing technologies has made breakthroughs
A variety of AM technologies have been developed during the recent
for manufacturing lattice structures in recent years.

9
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

decades; some of them are suitable for manufacturing metals and alloys
and these metal additive manufacturing technolgoes include powder
bed fusion (PBF) [19,185–193], direct energy deposition (DED)
[194–196], fused deposition modeling (FDM) [28,30,197], and lami­
nated object manufacturing (LOM) [198–200], direct ink writing (DIW)
[201–204], binder jetting [205–208], joule print [75,209], liquid metal
additive manufacturing [210,211], electrochemical additive
manufacturing (ECAM) [212–216], and cold spray additive
manufacturing (CSAM) [217–220]. Each AM technology has its advan­
tages and limitations, and each may or may not meet the requirements
for the fabrication of lattice structures. Meanwhile, the feedstock used
for these AM technologies can be broadly categorized as (1) liquid, (2)
filament/paste, (3) powder, and (4) solid sheet. Table 3 summarizes a
variety of AM technologies that are suitable for manufacturing metallic
parts, the feedstock, their processing mechanisms, and typical materials
during the AM process.

3.2. Mainstream additive manufacturing techniques for metallic lattice

structures

3.2.1. Powder bed fusion

PBF technologies mainly include selective laser sintering (SLS), se­
lective laser melting (SLM), and electron beam melting (EBM) [19,185,
186,225–227]. Both SLS and SLM belong to laser powder bed fusion
(L-PBF) [228]. In Refs. [62,196,229,230], such L-PBF technologies are
also named as direct metal printing (DMP). The L-PBF technologies
employ a laser beam as the heat source and scan the powder bed at a
controlled speed and selected locations. In comparison to the partial
melting of powder in the SLS process, the SLM process can make the full
melting of powder [18]. Besides, EBM technology is operated at a similar
Fig. 7. Lattice structures manufactured by various technologies: (a) lattice
procedure as SLS and SLM but EBM uses an electron beam as a heat
structure built by assembling, (reproduced with permission from Ref. [175],
Copyright (2012), Massachusetts Institute of Technology), (b) free-standing source [19]. Many metals and alloys, such as Ti and Ti alloys [231–235],
liquid metal, (reproduced with permission from Ref. [182], Copyright (2013), Al alloys [185,236–238], Ni alloys [239], Mg alloys [240], silicon brass
Wiley), and (c) folding of a perforated metal sheet (reproduced with permission [241], steels [242], NiTi alloys [243,244], Co–Cr-based alloys [245],
from Ref. [178], Copyright (2006), Royal Society). metallic glass [246] and composites [247–249], can be produced in the
PBF process.
Fig. 8a schematically illustrates the powder bed fusion processes. All

Table 2
Some conventional methods for producing metallic cellular structures or lattice structures as well as the main principles of these methods.
Method Material Feedstock Principle Disadvantage Refs.

Direct foaming Metal Molten metals It is a liquid state process. Gas bubbles are created in the The distribution, size, and number of [78,79]
metallic melts and conserved their profile to produce the gas bubbles are random. Therefore, a
corresponding liquid metallic foam. The produced porous regular structure cannot be obtained.
materials have open cells and close cells.
Eutectic reaction Metal Wire or rod Lattice can be produced by stacking wires or rods The patterns of lattice structures are [176,177]
according to the designed patterns. Then the wires or rods significantly limited.
are bonded together by a eutectic reaction. The produced
lattice structures only have open cells.
Electroless plating Metal Solution A template should be prepared first. Then the electroless The pattern of lattice structure is [172]
plating process is conducted. Afterward, the template is determined by the template.
removed.
Folding of a Metal Sheet Lattice can be produced by folding a perforated or Types of patterns are limited and post- [178,179]
perforated metal expanded metal sheet and the lattice structure can be assembly is always required. Fine
sheet created by bonding the lattice together. The produced structures are unavailable
porous materials have open cells.
Interlocking grid Metal, Fiber Fibers are assembled by interlocking. The produced Fibers should be custom-made and fine [11,174,
assembly composite, porous materials only have open cells. structures are unavailable 175]
polymer
Investment casting Metal Any type of Wax models are fabricated in a sand mold and heated up Low accuracy for producing the parts [108,109,
materials to to be melted and drained out. The molten metal is then with significantly fine structures; long 110,183]
produce molten poured into the sand mold. After cooling, the solidified production period.
metals materials are separated and finished. The produced
porous materials have open cells.
Powder metallurgy Metal Powder Space holders are employed as pore formers to generate The distribution, size, and number of [162,163,
with space pores in metallic products. The produced porous space holders are random. Therefore, a 164]
holder materials have open cells and close cells. regular structure cannot be obtained.
Wire-woven Metal Wire or rod Metallic wires and rods are woven to fabricate the parts. Types of patterns are limited and post- [117,118,
method The produced porous materials have open cells. assembly is always required. 119,176,
177,184]

10
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 3 Note that, it is of importance to ensure the homogeneous thickness of

Various additive manufacturing technologies for producing metallic parts, the the powder layer, which has a significant influence on the built parts.
feedstock, and their processing mechanisms. Generally, sufficient powder with controlled volume can be dropped by
Technologies Feedstock Processing Typical Refs. a hopper or achieved by lifting the feed cartridge to a defined height
mechanisms materials [189]. Subsequently, a counter-rotating roller, wiper, or doctor blade
Binder jetting Powder Material is Metals, [205, uniformly spread out the powder in the build platform. The stage for the
(BJ) bonded by ceramics, 206, spread of powder is pertinent during the process. The flow of the powder
jetted binders polymers, and 207, and the avoidance of particle cloud formation are critical to the quality
other powder 208]
of the produced parts. Processing parameters, such as layer thickness (t),
Cold spray Powder The powder is Metals, ceramics [217,
additive sprayed from a 218, laser/electron beam power (P), scanning speed (v), and hatching space
manufacturing nozzle and 219, (h), along with scanning strategy, spot size (d) and platform pre-heating
(CSAM) deposited on a 220] temperature, are the key factors for the final quality of the produced
substrate at a parts, which were widely reported in literature, e.g., Ref. [31,250,256,
very high speed.
Direct energy Powder or Partial melting Metals, ceramics [194,
257]. Fig. 9 reveals the detailed processing parameters involved in the
deposition wire or full melting 195, PBF technologies. To produce a part with good performance, a optimal
(DED) and 196, combination of these parameters should be considered. Furthermore,
solidification by 221] the processing strategies could affect the final quality of the manufac­
cooling
tured parts [258,259].
Direct ink writing Paste The material is Ceramic/metal [201,
(DIW) extruded slurries, 202, According to the melting points, the thermal conductivity, and the
directly without polymers, and 203, laser adsorption in different materials, the processing windows may be
melting or sol-gel inks 204] varied. Of all, the energy density of laser or electron beam is the
solidification threshold for the densification of the materials. Hence, several equations
Electrochemical Electrolyte Material is Metals, ceramics [212,
for calculating the energy density in the PBF process have been put
additive deposited by 213,
manufacturing ultra-micro 214, forward:
(ECAM) electrodes 215,
P
216] Ev = (2)
Fused deposition Filament, Filament melted Metals, [28, vht
modeling wire, or in nozzle and thermoplastic, 30,
(FDM) paste solidification by wax 197, 4P
El = (3)
cooling 222, π vd2
223]
Joule print Filament, Filament melted Metals, ceramics [75, P
wire, or in nozzle and 209] Es = (4)
vd
paste solidification by
cooling
where Ev is the volumetric energy density [260], El is the linear input
Liquid metal Liquid Material is Metals [210,
additive melted before 211]
energy density [261], and Es is the surface energy density [262]. From
manufacturing extrusion and these equations, one can find that the parameter of energy density is an
solidified by integrative processing parameter. Theoretically, almost all metal­
cooling s/alloys can be processed by SLM/EBM because both laser and electron
Laminated object Solid sheet Sheets are Papers, plastics, [198,
beam have high energy input. In the SLM/EBM process, a sufficient
manufacturing stacked and metals 199,
(LOM) bonded by the 200, energy density is required for melting metallic powder and therefore a
diffusion 224] high density for the manufactured parts. Take the laser scan speed as an
Powder bed Powder Partial melting Metals, [19, example, Liu et al. [263] produced Ti–24Nb–4Zr–8Sn lattice structures
fusion (PBF) or full melting ceramics, 185, by SLM using different laser scan speeds. Although an identical CAD
and Thermoplastic, 186,
solidification by wax 187,
model is used, the fabricated lattice structures are not entirely identical.
cooling 188, Fig. 10a and b respectively show the top and side view of single unit cells
189, after SLM fabrication. After careful measurements, it is found that both
190, X-Y plane and Z plane strut thicknesses decrease with lowering the scan
191,
speed (Fig. 10c). The increased thickness of Z plane struts leads to the
192]
enhanced compressive strength of lattice structures (Fig. 10d). However,
Liu et al. [263] also pointed out that higher energy density results in
processes are conducted in a chamber with a protective atmosphere more evaporation of Sn during the process, and hence the lattice
(usually argon atmosphere or nitrogen) for SLS and SLM [18,250,251] structures produced at the scan speed of 500 mm/s have the highest
or vacuum for EBM [19,252] to prevent the oxidation during density of pores. The pores in the lattice structures would reduce their
manufacturing. The powder layer is deposited and leveled in the fatigue strength [47]. Therefore, according to the applications of lattice
chamber first. A laser or an electron beam is then focused on and scans structures, the processing parameters may be varied even if the identical
the powder and the CAD structure is located on the slice/layer to induce CAD model is employed. Laser scan speed also would influence the
a fast fusion, after which metallic melt cools down to become the solid surface roughnesses of the produced products. Gockel et al. [264]
materials according to the pre-designed CAD model. The scan path of the investigated the surface roughness of Alloy 718 cylinders produced
laser or electron beam for each layer is determined by the corresponding using SLM with different laser powers and scan speeds. They found that
slice of the CAD model and the selected scanning strategy. Afterward, the surface roughness of the samples increases with increasing the laser
the powder bed is lowered by a defined layer thickness and a new scan speed when the laser power is fixed, or the surface roughness of the
powder layer is deposited and flattened. Such a process is repeated layer samples decreases with increasing the laser power when the laser scan
by layer until the completion of manufacturing the entire part. Owing to speed is constant, which is attributed to the decline in energy density.
their high accuracy, PBF technologies have great advantages in fabri­ Correspondingly, the samples with lower surface roughness have better
cating lattice structures with significantly complex geometries [21,253, fatigue performance.
254]. In addition, because of high temperature gradients, residual stresses

11
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 8. Schematic illustrations of various metal additive manufacturing technologies: (a) powder bed fusion, (b) direct metal deposition, (c) fuse deposition
modeling, (d) laminated object manufacturing, (e) direct ink writing (reproduced with permission from Ref. [207], Copyright (2019), Elsevier) and (f) binder jetting
(reproduced with permission from Ref. [255], Copyright (2017), Wiley).

Fig. 9. Detailed processing parameters involved in the powder bed fusion technologies.

are inevitably produced in the parts produced by SLS/SLM. In compar­ incomplete melting, and cracks) [47,263,269]. Such defects lower the
ison, residual stresses is relatively relieved in the parts produced by EBM density of the produced parts and act as stress raisers, thereby degrading
because the building plate in the EBM system always maintains a rela­ the properties and performances of the produced parts. Therefore,
tively high temperature (therefore a relatively low temperature optimized processing parameters can be used to enhance the perfor­
gradient) during the EBM process. Due to the layer-by-layer method, a mances of the produced lattice structures to a large extent [263].
large amount of thermal expansion and shrinkage during the cycle of However, the PBF technologies have a considerable number of pro­
heating and cooling can also produce residual stresses [19,265]. Re­ cessing parameters, which are relatively difficult to optimize.
sidual stresses may cause unexpected results of the produced parts, such
as distortion [266], cracking [267], and delamination [268]. Lattice 3.2.2. Directed energy deposition
structures are primarily formed by periodic unit cells with edges and DED technology was firstly developed by Sandia National Labora­
faces, which introduce porosity into a solid. Consequently, lattice tories in the USA and was commercialized by Optomec Corporation in
structures generally have a lower strength compared with the solid 1998 [270]. DED technologies include laser engineered net shaping
materials occupied the same space. As such, high residual stresses may (LENS) [271–273], wire and arc additive manufacturing (WAAM) [274],
be more destructive to lattice structures. Meanwhile, lattice structures 3D laser cladding (CLAD) [275–278]. Differing from other AM pro­
produced by PBF technologies may contains defects (porosity, cesses, DED technologies supply the blown powder or fed wire into the

12
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 10. (a) Top view and (b) side view of single-cell unit produced at different laser scan speeds, (c) the thickness of struts in the X–Y plane and Z plane at different
laser scan speeds, and (d) the corresponding compressive performance of Ti–24Nb–4Zr–8Sn lattice structures as a function of Z plane strut thickness. (reproduced
with permission from Ref. [263], Copyright (2015), Elsevier).

path of a scanning laser or an electron beam [279,280]. The DED process an optimized processing window and hence produce the parts with
creates an active melt pool by melting the deposited materials better robustness, less brittleness, and less cracking possibility [288,
(exceeding 3000 ◦ C). Almost any metal in powder or wire form can be 289].
used as feedstocks. Multiple nozzles are employed during the DED However, the poor surface finish and low build accuracy are the
process. Inert gas is used to reduce the oxidation of materials [279]. disadvantages that need to be taken into account in the LENS technol­
Such a process is similar to laser cladding. Thanks to a 3-7-axis machine ogy. During the LENS process, partially melted particles are also
system, the materials can be deposited from any angle [279]. Fig. 8b frequently found on the surface of the produced parts, which increases
shows the schematic illustration of the DED process. According to the their surface roughness and may lower the fatigue life of the produced
pre-designed CAD model, the near-net-shape components can be pro­ metallic parts [287,290]. The successive deposition is significantly
duced. Therefore, DED technology can produce complex porous struc­ determined by the precedent layer geometry. If the precedent layer
tures. Many types of powder of metals and alloys, including Ti and Ti deviates from the designed one, the dimension of the melt pool may
alloys [281,282], Co–Cr-based alloys [283], steels [284], Inconel [221], change because of the incorrect laser focal point [291,292]. Although
high-entropy alloys [285], NiTi alloys [286], can be employed in the the build accuracy can be improved by optimizing the processing pa­
DED process. rameters for a given metallic material, the LENS-produced components
Among all DED technologies, LENS, also named as direct metal still have lower accuracy than the PBF-produced parts. Therefore,
deposition (DMD), is the most frequently used DED technology [271]. post-processing treatments, such as heat treatment, hot isostatic press­
Compared with the widely used PBF technologies, LENS technology can ing, machining and so on, are always required for the LENS-produced
build a large part with a volume of 900 × 1500 × 900 mm3 [271]. parts (as well as the parts produced by other DED technologies)
Interestingly, the LENS process is not limited to manufaturing parts on a [293–296]. Another disadvantage is the non-uniformity in the micro­
levelled platform but can build on the other parts. Hence, LENS is structure and macrostructure of produced parts caused by uneven hea­
promising to repair worn or damaged components [287]. LENS can ting/cooling rate during the LENS process. Although uneven
produce the parts with multiple materials to obtain graded functions, heating/cooling rate caused by the large temperature gradients between
owing to the multiple nozzles [283]. The heat source of laser has a high the ambient environment and melt pool is found in all metallic AM
energy input and hence LENS can also produce parts using materials technologies, such a situation is more deteriorative during the LENS
with high melting points. It was reported that the LENS process can process, especially in the layers near the substrate. Therefore, the re­
accomplish better interfacial bonding between the fabricated layers in sidual stress is significantly higher in the LENS-produced parts than in

13
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

the PBF-produced parts, which would influence the part precision or LENS and investigated the processing parameters on their porosities and
even result in the part collapse during the LENS process. mechanical properties. They found that metallic lattice structures with
The main controllable parameters during the LENS process include interconnected porosities down to 70 vol% can be successfully manu­
laser power, scan speed, powder feed rate, and hatch distance [271]. factured. The porosities of lattice structures can be tailored by varying
Such parameters and their interaction determine the quality of the laser scan speed, powder feed rate, and laser power. The produced lat­
produced parts. Laser power may be the most important parameter tice structures with the porosity of 35–42 vol% demonstrate close
during the LENS process, which determines the formation and the moduli and yield strength to the human cortical bone [300]. Baranowski
dimension of the melt pool and therefore the mechanical properties of et al. [301] prepared two honeycomb lattice structures using Ti–6Al–4 V
the produced parts. An appropriate increase in the laser power can avoid by LENS and investigated the deformation behavior of the samples
partial melting zones; therefore, a dense structure is obtained [281]. under quasi-static loading. The honeycomb lattice structures have an
Regarding the scan speed, it refers to the relative translational speed of identical wall thickness of 0.7 mm but different diameters of 3 mm and 5
the laser focal point. Like PBF, increasing the scan speed would decrease mm. Meanwhile, they also simulated the deformation process of the
the input energy density for constant laser power and frequency. Such a samples and correlated their mechanical properties using LS-Dyna
parameter also influences the dimension of the melt pool therefore the software. Both experimental and simulation results showed that the
mechanical properties of the produced part [284]. Theoretically, the honeycomb lattice structure with a diameter of 3 mm exhibit about 3
increase in scan speed can save build time. However, the decreased input times the strength and energy absorption than the counterpart with a
energy density may result in partially melted regions, which hampers diameter of 5 mm. As shown in Fig. 11a, Antolak-Dudka et al. [302] also
the metallurgical bonding between layers [297]. Therefore, a threshold manufactured Ti–6Al–4 V honeycomb lattice structures with a wall
of scan speed is required for a specific material to ensure the quality of thickness of 0.7 mm and diameters of 3 mm, 4 mm, 5 mm, and 6 mm and
produced parts. Powder feed rate is related to the speed and volume of investigated their static and dynamic loading behavior. The Ti–6Al–4 V
powder blown into the melt pool [288]. Hatch distance can be consid­ honeycomb lattice structure with a diameter of 3 mm has the highest
ered as a cross-sectional distance between two scan lines, which de­ deformation energy in both static and dynamic loading conditions
termines the attainable smallest feature size during the LENS process (Fig. 11b). Such results are ascribed to the highest relative density of the
[298]. For the fabrication of lattice structures, hatch distance also has a lattice structures with a diameter of 3 mm.
great influence on the shape and distribution of pores [299]. Although both PBF and DED are well known AM technologies for
There are some examples of lattice structures produced by LENS. producing metallic parts, there exisit distinct differences between PBF
Krishna et al. [300] produced cubic lattice structures using CP–Ti by ad DED. On the one hand, the PBF process should be conducted in a

Fig. 11. Comparison of deformation energy between dynamic and quasi-static conditions for Ti-6Al-4 V honeycomb lattice structures with different cell sizes: (a)
four different cells and (b) deformation energy curves (reproduced with permission from Ref. [302], Copyright (2019), MDPI).

14
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

highly controlled atmosphere and the entire build process run in a lattice structures only have a slight dimensional change from the
chamber [250]. Therefore, the dimensions of the produced parts are FDM-produced pattern and only a small number of gas entrapment de­
limited by space offered by this chamber. In comparison, the DED pro­ fects are found in the produced lattice structures. Very recently, an
cess can be considered as an advanced laser cladding process without a innovative FDM process, which is known as fused deposition modeling
chamber and inert gas is employed to protect the materials from and sintering (FDMS), was developed in 2019 [242,316]. The FDMS
oxidation [279]. Hence, the DED process can produce larger parts than process uses the filaments of metal/polymer composites with a high
the PBF process. On the other hand, DED has a low manufacturing ac­ fraction of metal powder as feedstock and the printer is further opti­
curacy because the successive layer strongly depends on its precedent mized based on the conventional FDM printer. Consequently, the pro­
layer during the DED process [291,292]. Any deviation of accuracy duced parts are debinded to remove the polymer binder and sintered to
would lead to incorrect laser focal point. By contrast, the powder bed has fuse the metallic particles to form a dense part [316]. Based on this
an identical height during the PBF process and stable energy input can concept, stainless steel 316 L [316], AZ91 Mg alloys [317], Cu [318],
be realized during the PBF process [303]. Hence, higher accuracy of bronze [318], Al [318], and Co-based diamond composite [319] have
produced parts can be obtained by PBF than by DED. Such differences been successfully fabricated. As an alternative, the FDMS process can
make DED and PBF have various targeted applications. Owing to its produce metallic parts at low energy consumption and low cost because
ability for producing larger parts, DED can be employed to produce the temperature during the process is slightly higher than the melting
complex components in engineering applications, such as satellites, point of polymers (usually below 300 ◦ C). Although the sintering process
helicopters, and jet engines [304]. Such large parts may not be fabri­ and debinding process are required, they can be conducted in batches.
cated in the chamber during the PBF process. PBF can produce the parts Mohammadizadeh et al. [318] reported that the FDMS-produced Cu has
with the requirement of high accuracies, such as the metallic lattice comparable or even slightly higher strength than the SLM- and
structures for implant applications [303]. EBM-produced Cu. However, to date, there are still rare reports on the
metallic lattice structures produced by FDM. Fortunately, the develop­
3.2.3. Fused deposition modeling ment of the FDMS technology becomes a candle of hope for producing
FDM technique was developed by Stratasys and Scott Crump in 1989 metallic lattice structures using the FDM process.
[305]. Afterward, FDM was commercially applied in the early 1990s.
Unlike the PBF technologies, FDM technologies use filaments and/or 3.2.4. Laminated object manufacturing
pastes as feedstocks, such as plastics, metallic powder, ceramics, and Laminated object manufacturing (LOM) is one of the sheet lamina­
composites [30,306]. FDM technologies are primarily used for tion AM processes, which was developed by a California-based compa­
modeling, prototyping, and manufacturing. As shown in Fig. 8c, the ny—Helisys Inc. [35]. LOM technologies use the sheets of materials
filament/paste materials are fed into a nozzle and heated to their (such as metal, paper, ceramics, and polymer) as feedstock and bond
melting temperatures by liquefier. The nozzle moves at a pre-designed them together to a 3D object [320,321]. LOM methods adopt the
path across a base substrate and extrudes the fibers in a layer-by-layer concept of “cut-then-paste” and employ the “paste-then-cut” approach,
way to build a three-dimensional structure. After the deposition of leading to an inherent support mechanism. During the LOM process,
materials, the materials solidify into solids and form the final products sheets are precisely cut and stacked according to the pre-designed ge­
with a pre-designed geometry. According to the ASTM F42 terminology, ometry [322]. A localized energy source, such as ultrasound or laser, is
the FDM process can also be called fused filament fabrication (FFF) or used to bond the stacks. Ultrasonic additive manufacturing (UAM) is the
material extrusion additive manufacturing [30]. most commonly used LOM technology for producing metallic parts
The mechanism of the FDM process depends on the solidification [323]. By applying ultrasound and pressure on sheet stacks at room
time of deposited fibers [30,307]. At the very beginning of the process, temperature, the stacked sheets would be bonded by the diffusion be­
the extruded fiber interacts with the base according to the designed tween interfaces during the UAM process. This process is also known as
path. The subsequent fibers interact with the formed deposited ones and ultrasonic welding. In this way, a 3D object is formed without using a
generate a bond. Voids may be produced among the deposited fibers if heat source, which is remarkably different from other AM technologies.
the cooling is too fast, which significantly declines the mechanical Currently, the UAM technique is primarily used for Al, Cu, stainless
properties of the produced parts [308]. The main processing parameters steel, and Ti [324–326].
in the FDM process include nozzle diameter, layer thickness, raster The detailed working process of UAM is shown in Fig. 8d [322].
width, raster orientation, build orientation, infill pattern, print velocity, Sheets are set up and stacked on a base plate. A sonotrode in
air gap, infill density, and the number of contours [306]. These pa­ high-frequency vibration is employed to roll and provide a mechanical
rameters can significantly influence the mechanical properties, surface load on the sheet. This process also produces asperities on the surface of
roughness, thermal expansion, and dimensional accuracy of the pro­ the sheet [327]. When the next sheet is applied, the asperities become a
duced parts [309–312]. bonded interface under the effect of the vibrating sonotrode. Conse­
Initially, FDM is a low-cost technology developed for the fabrication quently, a new metallic sheet is bonded to the previously built parts. For
of thermoplastic polymers. The FDM process has a relatively low metallic parts, the temperature locally increases during the UAM process
extrusion temperature compared with other AM technologies [306]. For and the heating process is greatly influenced by vibration amplitude
example, printing the first layer of polypropylene should maintain the because a higher amplitude causes a higher plastic shear strain at the
nozzle at 165 ◦ C and printing poly-ether-ether-ketone (PEEK) should asperities [322,325]. The cooling between the fabrication of each layer
retain the nozzle over 350 ◦ C [313,314]. However, the working tem­ is in a short period in order to reduce thermal residual stress. In the back
peratures in FDM are still quite low for melting the metallic materials of the fabrication of all layers, the part is cut from the base plate and
having high melting points. As such, it is currently difficult to directly polished for a better finish.
obtain metallic lattice structures produced by FDM. As mentioned in Note that, the interface bonding strength should be carefully
Section 2.2, FDM can produce wax or polymer patterns with complex considered because the interface delamination determines the failure
shapes (e.g., lattice structures) for investment casting [116]. The mold, and fracture of the UAM-produced parts in the building direction [328].
which is used for manufacturing metallic lattice structures, can be The strength of the UAM-produced parts in the building direction is
fabricated according to the produced patterns. Therefore, the patterns significantly weaker than that in the other two directions [328]. Many
produced by FDM can be considered as an intermediate product for investigations have reported the anisotropic mechanical properties of
manufacturing metallic lattice structures. Similar work was reported by UAM-produced metallic parts [328–330]. Such a problem may be
Carneiro et al. [315], who produced polylactic acid (PLA) patterns by minimized by optimizing the processing parameters, such as roller
FDM and fabricated the A356 Al alloy lattice structures. The produced Al temperature, roller speed, the surrounding temperature, and the base

15
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

plate temperature [224]. Roller temperature and speed influence the better printing process. When the printing ink has been extruded from
transfer of heat from the roller to the sheet (as well as the previously the nozzle, it should then become a highly viscous state because the
fabricated parts), while the surrounding temperature and base plate shape of materials has to be maintained on the build platform [332].
temperature affect the rate of heat loss of the built part. These param­ Therefore, the printing inks used in the DIW process are pseudoplastic
eters jointly influence the thermal history of the produced parts there­ fluids, including slurries, polymers, and sol-gel inks [333].
fore their mechanical properties. Substantially, DIW was initially developed for printing ceramic
Compared with other AM technologies, LOM technologies cost less materials [334]. With the development of DIW technology, a wider
because metallic sheets are inexpensive than powder, wires, and fila­ range of materials are available. Other materials, such as polymers [335,
ments. The sheet-construction method allows different metals to be used 336], carbon materials [337–339], and zeolites [40,340,341], are also
during the LOM process, which is still a major challenge for other AM printed by DIW. Until recent years, metals and alloys can be also printed
technologies. The produced parts are finally precisely cut in both the by the DIW process [70,76,342]. Because the DIW process has no
rolling direction and vibrating direction. However, the accuracy is still melting and solidification process, metals and alloys cannot be directly
difficult to control in the build direction because the layer thickness is printed. Compared with the FDM process using the paste/filament as
changed under the mechanical load. Based on such characteristics, the feedstock, the DIW process employs the solution as feedstock. Metallic
LOM process is frequently employed to handle the relatively large parts, particles with significantly small sizes are dissolved in a solvent to
and thus becomes less possible to directly produce a lattice structure. prepare printing ink. For example, to obtain Fe and Ni lattice structures,
Luong et al. [331] fabricated a three-dimensional laser-induced gra­ Taylor et al. [342] used Fe and Ni metallic particles with average sizes of
phene foam using LOM. Zhang et al. [198] used organic mesh sheets to 5.9 mm and 4.4 mm, respectively, to prepare printing inks using
fabricate the ceramics with regular micron-scale pore structures by the polylactic-co-glycolic acid copolymer as the solvent. Although the DIW
LOM process. However, to the best of the authors’ knowledge, there are process is convenient to produce the parts with designed shapes, it also
still rare reports concerning the metallic lattice structure produced by has distinct disadvantages. Because the printing inks are prepared from a
LOM. solvent, post-treatments, such as drying and sintering, are always
required [343]. Therefore, the shrinkage of samples is ineludible. As
3.2.5. Direct ink writing shown in Fig. 12a-d, Elsayed et al. [344] prepared Ti–6Al–4 V lattice
Direct ink writing (DIW), also known as robocasting, is another structures by DIW. Different processing parameters, including the
extrusion-based AM process. Fig. 8e shows the schematic illustration of diameter of filaments (df) and distance between filaments (φf), were
the DIW process. Similar to the FDM process, the materials are extruded employed. After sintering, lots of pores are uniformly distributed in the
from a nozzle during the DIW process but without melting and solidi­ lattice structures (Fig. 12e). These pores are produced by the evapora­
fication [332]. Therefore, the DIW process is easily operated and has a tion of the solvent. By the measurement of micro-CT, a 3D representa­
relatively low cost owing to its low energy consumption. Printing inks, tion of the distribution of closed-cell pores can be obtained (Fig. 12f).
which are not directly used, should be prepared first before the DIW The pores have a Gaussian distribution and d90 is 134.3 μm (Fig. 12g).
process. The printing ink should have a low viscosity and hence a good Because of the evaporation of the solvent, the lattice structures are
fluidity can be obtained. However, a higher fluidity does not suggest a shrunk after sintering. The shrinkage of lattice structures decreases with

Fig. 12. Ti–6Al–4 V lattice structures produced by direct ink writing: (a) schematic illustration of filament diameter (df) and distance between filaments (φf), (b)-(d)
optical images of lattice structures with df = 250 μm with an increasing df/φf ratio, and (e) volumetric shrinkage (ΔV) and (f) yield strength (σ y) of the lattice
structures after sintering versus the df/φf ratio (reproduced with permission from Ref. [344], Copyright (2019), Elsevier).

16
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

increasing the df/φf ratio (Fig. 12h). Fig. 12i is the results of quasi-static solid design. Hence, the mechanical properties of this component ob­
compression tests of lattice structures and shows a large deviation be­ tained from the compressive test and three-point bending test are lower
tween results in a group of samples. This result may account for the poor than those obtained from its numerical model. In addition, because of
geometrical repeatability of the produced lattice structures fabricated by the instinct features of this technology, the metallic parts produced by
DIW. Therefore, the fabrication accuracy of DIW needs to be improved BJ have a lower precision in dimensions than those produced by other
in future research. During the sintering, carbide is produced in the lat­ technologies (e.g., PBF and DED).
tice structure [344], where carbon is mainly from the solvent. In addi­
tion, the fully metallic parts with lattice structures are still not 3.2.7. Other additive manufacturing technologies for metallic production
achievable by the DIW method. Other additive manufacturing technologies for metallic part pro­
duction have also been reported. Joule printing, which is similar to DED,
3.2.6. Binder jetting was developed in recent years [75,209]. However, Joule printing melts
Binder jetting (BJ) is a similar process to the DIW process. The DIW the wires or filaments by current rather than laser or electron beam,
process is extrusion-based, while the binder jetting process is droplet- which results in a higher manufacturing rate. In addition, liquid metal
based [205–208]. During the BJ process, a liquid binding agent is additive manufacturing is a technology that melts the metals first and
dispersed on the powder to print a two-dimensional pattern on a powder drops the liquid metal from a jet [210,211]. This technology is devel­
bed [207]. This technology was invented in 1993 at the Massachusetts oped based on the generation of the magnetohydrodynamic droplet. The
Institute of Technology [345]. As an AM process, BJ also possesses a molten metals in the jet may be higher than 1000 ◦ C. The parts are then
considerable number of advantages. The BJ process can use any produced layer by layer. Another AM technique called electrochemical
powder-based feedstock with a binder, such as metals [346–348], ce­ additive manufacturing (ECAM) is a novel non-thermal process for
ramics [349–351], biomaterials [352,353], and polymers [354–356]. producing macro-, micro-, and nanoscale metallic parts [212–216]. Such
Meanwhile, the print rate of BJ is relatively high compared with other a technology combines electrochemical deposition by ultra-micron
AM processes [357]. As shown in Fig. 8f for a typical BJ system, the parts electrodes and additive manufacturing technology to produce metallic
are produced by jetting the liquid binding agent to the powder bed in a parts directly from pre-designed models. Furthermore, cold spray ad­
layer-by-layer way. After the production of a layer, the powder bed is ditive manufacturing (CSAM) is a solid-state supersonic deposition
lowered to provide space for the next layer. Such a process is similar to method that can achieve mass production, fabrication, and restoration
the SLM and EBM processes and is repeated until the production of the of metallic components [217–220]. Cold spray technology, which is
entire part is finished. Due to the existence of a binder, the produced used in CSAM, is different from thermal spray technology and processes
parts always need to be post-processed [358]. The typical the materials at a relatively low temperature [379–384]. CSAM can
post-processing includes powder removal [359], debinding [360], sin­ generate the profile of the component close to the final geometry with a
tering [361], infiltration [362]. small tolerance. The surfaces of final products have to be processed to
Metallic components (steels [361,363–368], Inconel [369–372], Cu obtain the required functional surfaces [217]. In a summary, recent
[357,373], Ti and Ti alloys [374,375], WC-12Co [376], etc.) can be advances have suggested that the development of different AM tech­
produced by BJ, which occupies a large percentage of BJ products. In nologies has been booming in recent years, and the AM technologies
many industries, although the produced parts with a high relative hold the promise to design and manufacture more complicated metallic
density are required, the parts produced by BJ may not meet the re­ parts. However, not all AM technologies are suitable for manufacturing
quirements for specific applications because BJ-produced parts have metallic lattice structures. Some AM technologies (such as stereo­
relative high porosity. Therefore, hot isostatic pressing (HIP) is often lithography [385], multi-jet fusion [386], rapid freeze prototyping
employed to improve the relative density of as-fabricated BJ parts [377, [387]) cannot use metallic powder or wire as feedstock and some AM
378]. In addition, the BJ process has been employed to manufacture technologies (such as LOM [198] and CSAM [381,382]) cannot fabricate
lattice structures, where Vangapally et al. [365] designed and fabricated the metallic structures with refined unit cells. So far, metallic structures
a series of 316 L steel lattice structures with cubic unit cells, and they are mostly produced by PBF, DIW, and BJ.
found that the mechanical properties of the final products are affected
by both sintering time and temperature. As shown in Fig. 13, Tang et al. 3.2.8. Post-treatment
[368] also printed 316 L steel lattice structures with cubic unit cells by The AM-produced metallic parts often have print defects, poor sur­
the BJ method and they found that the metal powder is not fully merged face finish, and undesired microstructure. Meanwhile, a functional
after sintering. The fabricated parts do not have full density despite the surface may also be expected. Therefore, post-treatments, including HIP,
heat treatment, and surface modification, are sometimes employed.
(1) HIP: hot isostatic pressing is usually employed to densify castings,
sintered components, consolidating powder, and interfacial bonding
[388]. Hence, this technology can further densify the AM-produced
parts because print defects are inevitable in the as-manufactured con­
dition [263]. Qian et al. [389] reported that HIP can close or heal the
pores and lack-of-fusion defects of SLM-produced Ti–6Al–4 V alloy,
thereby further enhancing its tensile ductility by about 10 %. However,
HIP only can be conducted on bulk materials. If metallic lattice struc­
tures is subject to HIP, they may be deformed to some extent (e.g.,
distortion of open cells). Therefore, so far, rare reports concerning the
HIP on lattice structures have been found.
(2) Heat treatment: due to the fast cooling rate during the AM process
(such as SLM and DED), significant internal thermal stresses are pro­
duced in the structure. Heat treatment is often conducted to reduce the
internal thermal stresses of the produced parts and manupluate their
microstructures. For example, SLM-produced Ti–6Al–4 V always ex­
Fig. 13. SEM image for a 316 L steel lattice structure produced by binder hibits dominated α′ martensite in the microstructure [232]. The Ti alloys
jetting and sintering. The inset is the sintered sample at a macro-scale view with such a microstructure always have relatively high yield stress but
(reproduced with permission from [368] copyright (2016), Taylor & Francis). low ductility. After being heated within the α + β region, the

17
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

morphology and size of α-phase can be controlled and the size of the 4. Design of metallic lattice structures
prior β phase is almost not influenced [390]. Vrancken et al. [390]
produced Ti–6Al–4 V alloys by SLM and heated the samples ranging In the past, even though many complex architectures with potential
from 500 ◦ C ~ 1050 ◦ C. The SLM-produced Ti–6Al–4 V alloy has a excellent performances can be designed, their fabrications are still
tensile strength of 1267 ± 5 MPa and strain of 7.28 ± 1.12 %. After restrained by conventional manufacturing methods. Fortunately, owing
heat-treated below 995 ◦ C (the β-transus temperature), the dominated α′ to the advanced manufacturing capability, the development of AM
microstructure is converted to a lamellar mixture consisting of α and β technologies in return facilitates the design of more complex architec­
phases. When heat-treated above the β-transus temperature, the growth tures. AM technologies offer better control and flexibility in the
of grains takes place and large β grains are formed. They pointed out that manufacturing process, thereby pushing the investigations toward to­
the ductility of SLM-produced Ti–6Al–4 V alloy increases from 7.36 ± pology optimization with the aim to optimized geometry for a specific
1.32%–12.84 ± 1.36 % after the heat treatment of 850 ◦ C for 2 h. Li et al. set of requirements. Therefore, thanks to the manufacturing capability
[391] prepared AlSi10Mg alloys by SLM and heated the samples ranging offered by AM, the design approach of cellular structures has gradually
from 450 ◦ C ~ 550 ◦ C. With the increased temperature of heat treat­ shifted from the conventional design for manufacturability to the
ment, the strength of the samples decreases, while their ductility in­ advanced design for functionality. This shift leads to an intensive focus
creases. On the one hand, the internal thermal stresses of samples on architectured cellular structures, namely, lattice structures, which
decrease after heat treatment. On the other hand, the growth of grains have better performances than stochastic cellular structures. In the view
leads to a decrease in the strength of samples. Although the above in­ of structure design, a lattice structure is a collection of unit cells
vestigations use bulk metallic materials as the targets, the principle is repeated in a certain spatial pattern. Therefore, both unit cell design and
the same to the metallic lattice structures. Liu et al. [47,48] fabricated pattern design should be considered in the design of lattice structures.
Ti–24Nb–4Zr–8Sn lattice structures by EBM. The produced lattices
structures have a mixed microstructure of α + β phases. After 4.1. Unit cell design
heat-treated at β-phase region, only β phase is found in the microstruc­
ture of sample. Correspondingly, the annealed sample has an decrease in The unit cell is the smallest unit composed of the whole lattice
the elastic modulus but a higher ductility compared with the as-built structure [398]. As such, the design of unit cell has an important effect
one. Therefore, the heat treatment is mainly used to reduce the inter­ on the characteristics of a lattice structure, such as mechanical response
nal residual stresses of produced components and tailor their [399], specific surface area [55], stiffness [400,401], pore size [401],
microstructures. and other properties [58,59,65,402]. Moreover, the formability of lat­
(3) Surface modification: AM-produced metallic parts may have a tice structures by AM should be considered when designing unit cells.
poor surface finish. For instance, the SLM-produced lattice structures According to the topologies, lattice structures can be further categorized
have attached a large number of partially melted particles [392]. Such as strut-based lattice structures, triply periodic minimal surface (TPMS)
attached particles would increase the surface roughness of metallic lattice structures, and shell lattice structures [403].
lattice structures and have an adverse effect on their mechanical prop­
erties, especially fatigue strength. Therefore, the improvement of the 4.1.1. Strut-based unit cell
surface quality is required. Some methods, such as blasting [392,393], Strut-based lattice structure has a periodic arrangement of unit cells
chemical etching [394], and electrochemical polishing [394] have been constituted of struts and nodes [404]. The strut-based lattice structure is
proposed to remove the attached particles. As a result, the surface also named as beam-based lattice structure. The simplicity of design
roughness of lattice structure is reduced. However, Pyka et al. [394] makes these strut-based topologies more widely adopted [405]. Sehar­
pointed out that electrochemical polishing would also decrease the ing et al. [406] have concluded the most commonly used metallic unit
diameter of the strut and therefore reduce the strength of lattice struc­ cells to build strut-based lattice structures. The main features of these
tures. In such a situation, the surface roughness and the strength of common metallic unit cells are listed in Table 4. Apparently, different
lattice structures should be balanced. On the other aspect, surface lattice structures have various features therefore potential applications.
functionality may be required for metallic lattice structures to improve The strut-based unit cells for lattice structures could be expressed by
their performances. For example, although Ti and Ti alloys are widely the numbers of struts (s) and nodes (n). These two numbers determine
used as biomedical implants, they have low abilities for protein the Maxwell number (M) of the unit cell. The relationship of s, n, and M
adsorption and interrogation of neutrophils and macrophages due to is shown as follows [427]:
their bioinertness [60]. Due to their convoluted surfaces, many con­
M = s − 3n + 6 (5)
ventional surface modification methods for bulk materials cannot be
employed for lattice structures [395]. To conduct the surface modifi­ If M < 0, the number of struts is not enough to equilibrate the
cation for lattice structures, some surface modification methods have external forces induced by the nodes, the bending-dominated behavior
been reported. Chai et al. [396] used perfusion electrodeposition to of the structure, resulted from the bending stresses in the struts, would
deposit calcium phosphate on Ti–6Al–4 V lattice structures and the take place. In contrast, if M ≥ 0, no bending takes place at the nodes and
biological performance of the modified Ti–6Al–4 V lattice structures axial tension and compression in struts can equilibrate external forces,
significantly increases. Das et al. [397] fabricated Ti lattice structures by and hence the structure is stretch-dominated [239]. By comparison, if
LENS and modified their surface by anodization. Hence, a uniform the structures have the same mass, the stretch-dominated structures are
nanoporous TiO2 layer with the thickness of 300 nm and pore diameters stiffer and the bending-dominated structures are more compliant (more
of 50 nm was prepared on the surfaces of Ti lattice structures, which detailed information on deformation mechanism is described in Section
enhanced the precipitation of apatite in simulated body fluids. Similar 5.1) [105]. Designing the unit cells for strut-based lattice structure is
work was reported to modify the surfaces of lattice structures for the based on the parametric algorithm of the software used [36,441]. The
applications of catalysts [255,342]. Indeed, the convoluted surface is the MATLAB software has a parametric algorithm code that designs the
primary difficulty for surface modification of lattice structures. There­ strut-based lattice structure by the following parameters: cell type, cell
fore, as mentioned above, the methods for surface modification of lattice size, strut diameter, and elongation angle. Visual programming lan­
structures frequently employ the solutions to ensure the contact between guage Grasshopper provides a new parametric algorithm to design
lattice structure and effective media. strut-based lattice structures, which can run within the Rhinoceros 3D
CAD software. Such a parametric algorithm can create the lattices
structures with any shape and has a great convenience in altering the
geometrical parameters, such as cells size, struts diameter, gradient for

18
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 4
Most commonly used strut-based unit cells to construct metallic lattice structures and their main features (partially reproduced with permission from Ref. [406],
Copyright (2014), SAGE).
Unit cell Topology Fabrication method Features Preferred applications Refs.

Auxetic AM-assisted Negative Poisson coefficient; Energy adsorber, wine bottle cork, [116,
investment casting, significantly high energy absorption. sensor, filter, sandwich panel core, 407,408,
EBM, SLS, SLM. scaffolds, light-weight structure, 409,410]
impact protecting apparatus.

All face-centered cubic SLM, EBM. Symmetrical in X, Y, Z axes; high Energy adsorber, lightweight structure. [231]
(AFCC) stiffness; suitable for energy absorption.

Body-centered cubic (BCC) SLM, EBM, SLS. Symmetrical in XYZ axes; isotropic in X, Lightweight structure, energy [239,
Y, Z, XY, YZ, XZ, XYZ directions; eight adsorber, bone implant. 411,412,
struts connected at the center of the 413]
cube;

BCC with Z strut (BCCZ) SLM, EBM, SLS. BCC with four Z-strut reinforcements; Lightweight structure, energy [412,
isotropic in X, Y, YZ, XZ directions; adsorber. 414]
anisotropic in other directions.

Cubic SLM, EBM, SLS, BJ, A cubic frame is formed by twelve Bone implant, catalytic structure, [255,
DIW struts; stress concentration may take energy adsorber. 344,368,
place in this structure. 415,416]

Diamond SLM, EBM, SLS, Isotropic in XY, YZ, XZ directions; the Bone implant, lightweight structure. [417,
manufacturing process of overhanging 418,419,
struts requires support. 420]

Edge-centered cubic SLM, EBM Symmetrical in X, Y, Z axes; struts are Bone implant, lightweight structure, [405]
connected at every edge of the cubes. energy adsorber.

Face-centered cubic (FCC) SLM, EBM Symmetrical in X, Y, Z axes; isotropic in Bone implant, lightweight structure, [405,
X, Y, Z, XY, YZ, XZ directions. energy adsorber. 421]

Face-centered cubic with Z- SLM, EBM FCC with Z-strut reinforcements; Bone implant, energy adsorber, [421,
strut (FCCZ/PFCC) isotropic in X, Y, YZ, XZ directions; lightweight structure. 422,423]
anisotropic in other directions.

FBCCZ with X- and Y- strut SLM, EBM FBCCZ with X-strut and Y-strut Bone implant, energy adsorber, [231,
(FBCCXYZ) reinforcements; isotropic in X, Y, Z, XY, lightweight structure. 423]
YZ, XZ, XYZ directions.

FCC with BCCZ (FBCCZ) SLM, EBM Combination of FCC and BCCZ; Bone implant, energy adsorber, [403,
isotropic in X, Y, YZ, XZ directions; lightweight structure. 421,423]
anisotropic in other directions.

G7 SLM, EBM Relative low strength and low elastic Bone implant. [264]
modulus.

Hexagonal Honeycomb DED, SLM, EBM, Anisotropic structure; suitable for Energy adsorber, lightweight structure. [10,424,
energy absorption. 425,426]

Octahedron SLM, EBM Rarely used Bone implant. [403]

Octet-truss SLM, EBM AFCC is combined with the octahedron; Heat exchanger, bone implant, [427,
high stiffness. lightweight structure. 428,429]

Optimized SLM, EBM High stiffness; High strength Bone implant, lightweight structure. [430]
(continued on next page)

19
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 4 (continued )
Unit cell Topology Fabrication method Features Preferred applications Refs.

Rhombic dodecahedron SLM, EBM High stiffness; high energy absorption. Bone implant, energy adsorber. [2,431,
432]

Tetrahedron SLM, EBM Cubic with diagonal-strut Bone implant, energy adsorber, [403,
reinforcements. lightweight structure. 433]

Tetrakaidecahedron Rarely reported Closed-cell lattice structure; suitable for Energy adsorber, heat exchanger. [434,
space-filling materials. 435,436]

Truncated cuboctahedron SLM, EBM, High fatigue resistance. Energy adsorber, bone implant. [437,
438]

Two face-centered cubic SLM, EBM Two FCC unit cells are combined with Bone implant, energy adsorber, [422,
combined in a BCC way one BCC unit cell; high energy lightweight structure. 439,440]
(F2BCC) absorption.

struts diameter, sample size [442]. Fig. 14 shows the design route for design [447,448]. TPMS defines a periodic structure along with three
gradient lattice structure of Ti–6Al–4 V, designed by Grasshopper [443]. coordinates with zero average curvature of the surface [448]. The
Such a gradient lattice structure has increased strut diameters from top definition of TPMS is based on implicit equations which define a set of
to bottom. In the PBF process, different struts in the unit cells have zeros of a function (F(x, y, z) = 0) and determine an array of points
various orientation angles to the powder bed. These orientation angles located on the surface of the unit cell. The first example of TPMS was
would influence the build quality of struts [444]. Generally, the hori­ found by Schwarz in 1865 [449]. Subsequently, Neovius also discovered
zontal struts have the lowest quality, while the vertical struts are easy to a new type of TPMS in 1883 [449]. Schwarz and Neovius found five
build [61]. types of TPMS, namely, Schwarz primitive, Schwarz diamond, Schwarz
Substantially, the unit cell has a significant influence on the me­ hexagonal, Schwarz crossed layers of parallels and Neovius surface.
chanical properties of the produced lattice structures, even if the lattice Afterward, the Gyroid surface, which is the most famous type of TPMS,
structures have similar relative densities and are produced by the same was described by Schoen in 1970 [450]. In many ways, the Gyroid
technology. Li et al. [445] designed three types of Ti–6Al–4 V lattice surface, Schwarz diamond, and Neovius surface are the most commonly
structures with cubic, G7 and rhombic dodecahedron unit cells and used TPMS [447], which are displayed in Fig. 16.
produced a series of samples with different densities by EBM (Fig. 15a). All types of TPMS can be expressed using implicit functions [452].
Fig. 15b shows the mechanical properties of these lattice structures. One For Gyroid surface:
can find that the elastic moduli of lattice structures increase linearly in
F(x, y, z) = cos(x)∙sin(y) + cos(y)∙sin(z) + cos(z)∙sin(x) + α (6)
the range of 0.5− 15 GPa with increasing the relative density. Such
values are comparable with those of trabecular (0.05− 3 GPa) and For Schwarz diamond:

F(x, y, z) = sin(x)sin(y)∙sin(z) + sin(x)∙cos(y)∙cos(z) + cos(x)∙sin(y)∙cos(z) + cos(x)∙cos(y)∙sin(z) + α (7)

cortical bone tissue (10− 25 GPa) [63]. For the same density, the lattice
structure with cubic unit cell exhibits the highest elastic modulus, while For Neovius surface:
the counterpart with G7 unit cell has the lowest elastic modulus. Like the
F(x, y, z) = 3∙(cos(x) + cos(y) + cos(z) ) + 4∙(cos(x)∙cos(y)∙cos(z) ) + α
elastic moduli, the compressive strength of lattice structures also has a
linear relation to the density. The lattice structures with cubic and G7 (8)
unit cells and the same density exhibit the highest and lowest strength,
where x, y, and z are the variants in Cartesian coordinates [452], α is
respectively. Liu et al. [446] also reported the different energy absorp­
used to determine the solid fraction of designed part [103]. α is the
tions of Ti–24Nb–4Zr–8Sn lattice structures with different unit cells and
offset, which can be both positive and negative, giving the desired solid
similar densities produced by SLM. The topology optimized sample has
fraction of the unit cell. Such parameters determine the configurations of
the highest compressive strength and highest energy absorption. Such
TPMS lattice structures and therefore their properties.
unit cells always have different deformation behavior, which would be
Meanwhile, periodicity is also closely related to the properties of
discussed in Section 5.1 in detail.
TPMS lattice structures. Periodicity is defined by the k values (kx , ky , kz ),
which can be calculated by the following equation:
4.1.2. Triply periodic minimal surface
The unit cells based on triply periodic minimal surface (TPMS) have ki = 2π
ni
(9)
also been widely investigated. TPMS is an effective approach in unit cell Li

20
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 14. Design route for strut-based gradient lattice structure. (reproduced with permission from Ref. [443], Copyright (2018), Elsevier).

where ni is the number of cell repetitions and Li is the absolute size of the lattice structures because the TPMS unit cells can be adjusted by
lattice structure in x, y, or z directions. changing the parameters in the implicit functions [448]. Therefore, the
Unlike strut-based lattice structures, the orientation of TPMS should TPMS lattice structure has received increasing attention in recent years
also be considered. When a TPMS lattice structure is used in an aqueous [26,455,456,459–461].
environment (such as heat exchangers [453] and bone implants [21]),
the open pores may influence the flow direction of fluids. Therefore, 4.1.3. Shell unit cell
some specific properties, such as the fluid flowing in heat exchangers The unit cells composed of plates rather than struts or continuous
[453] and the body fluid flowing in bone implants, would be enhanced surfaces are called shell unit cells or shellular, which was firstly reported
or controlled to some extent [454]. To design TPMS unit cells, eligible in 2015 [469,470]. Such a type of unit cell is also known as TPMS-like
software is required. However, the conventional CAD system used for lattice structures although their surfaces may not always keep zero in
SLM, EBM, etc. is difficult to design implicit surfaces. In this case, mean curvature. Fig. 17 shows the typical shell lattice structures [471].
MathMod (K3DSurf), which is free software, is frequently used for It can be found that the unit cells of shell lattice structures can be also
creating a TPMS unit cell in academic research. Another software Sim­ analogous to crystalline structures, which is similar to the strut-based
pleware ScanIP is commercially used and can employ a CAD module to lattice structure. Each unit cell of shell lattice structure has an original
create lattice structures based on implicit functions. configuration of strut-based lattice structure. However, the plates in the
Like the strut-based lattice structures, the metallic TPMS lattice unit cells of shell lattice structure result in closed pores, which is largely
structure produced by AM has been widely reported [448,455,456]. As different from the strut-based lattice structures with open pores. On this
listed in Table 5, various metallic TPMS lattice structures have been basis, Tancogne-Dejean et al. [69] reported that the shell lattice struc­
fabricated and investigated. TPMS lattice structures generally have tures with closed pores have a better elastic performance than the
better mechanical properties (e.g., fatigue strength, stiffness, and strut-based lattice structures with open pores, even though they have
roughness) than strut-based lattice structures [457,458]. This is an similar relative densities. Tancogne-Dejean et al. [69] also printed
important advantage of TPMS lattice structures. Meanwhile, TPMS unit polymer shell and strut-based lattice structures with 30 % relative
cells are better and easier defined compared to the ones for strut-based density by direct laser writing and found that the shell lattice structures

Fig. 15. Three types of Ti–6Al–4 V lattice structures with cubic, G7 and rhombic dodecahedron unit cells and different densities produced by EBM: (a) the cor­
responding prototype blocks, and (b) dynamic Young’s modulus (by open symbols) and static Young’s modulus (by solid symbols), (c) compressive strength and (d)
the relationship between dynamic modulus and compressive strength (reproduced with permission from Ref. [445], Copyright (2014), Elsevier).

21
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 16. Typical triply periodic minimal surface unit cells: (a) Gyroid surface, (b) Schwarz diamond, and (c) Neovius surface. (reproduced with permission from
Ref. [451], Copyright (2019), Elsevier).

have about 2–3 times elastic moduli of the strut-based lattice structures 4.2.1. Direct patterning
with the same porosity level. Such two lattice structures demonstrate Direct patterning is relatively simple and convenient in which the
different failure modes; the shells and struts first fail as a result of unit cells are designed as cubic elements and simply repeated in the
buckling in the shell and strut-based lattice structures along the loading lattice structures in most cases [476]. Fig. 18 schematically shows the
direction, respectively. For the shell lattice structures, the diagonal direct patterning method and some actual examples (i.e., cubic, topol­
plates experience deformation as well; while for the strut-based lattice ogy optimized, and rhombic dodecahedron lattice structures) designed
structures, the diagonal struts remain straight, even at significantly large by this method [446]. Apparently, regardless of the lattice structure to
strains. It is well known that shells maintain their loading carrying ca­ be constructed, the unit cells are only simply repeated in a 3D space.
pacity even after buckling [472], which could explain why the shell Such a method is suitable for the design of lattice structures produced by
lattice structures have higher moduli than strut-based lattice structures. AM technologies [47,62,468,477]. Meanwhile, as shown in Fig. 14, the
However, up to date, rare reports on the metallic shell lattice structures direct patterning of gradient strut-based lattice structure can be ach­
are found. The investigations mainly focus on the simulations and ieved by altering the geometric parameters.
polymers [469,471]. Berger et al. [469] reported the shell lattice The above examples are based on the strut-based lattice structures.
structures may reach the theoretical limit of isotropic elastic stiffness. For TPMS, the creation of a lattice structure is relatively easier. To
Nevertheless, shell lattice structures with closed pores cannot allow fluid enlarge the TPMS, it can be achieved by extending the ranges of the
flow and therefore this type of lattice structure is not suitable for heat variables of implicit functions. Generally, the range from 0 to 1 is used to
exchanger. define the unit cell. When the range is extended to 0–3, the created
The design and manufacture of shell lattice structures by AM tech­ lattice structure would have 3 × 3 × 3 unit cells [449]. The gradient
nologies have similar routes as the strut-based lattice structures [473]. TPMS lattice structures can be achieved by introducing a linear term to
However, given the reported advantages of shell lattice structures, some the implicit functions. The period of TPMS lattice structures can be
technical difficulties have encountered in their manufacturing process extended in x, y and z directions. If the term of x is changed to kx, the
because the powder removal in closed pores is usually difficult when TPMS lattice structure reveals a porosity gradient in x direction [303].
powder-based AM technologies are adopted [471]. As such, the lattice The value of k determines the porosity gradient in the corresponding
structures with open pores receive more interest in a variety of in­ direction. However, using direct patterning to produce lattice structures
dustries [470,473,474]. may not meet the requirement of the geometries of the final products
because the unit cells have edges and faces. Alternatively, Boolean
4.2. Pattern design subtraction is used to subtract the lattice structure [42]. The overall
lattice structure is simply subtracted according to the profile of the
The unit cells repeat in a 3D space and make up the lattice structures. produced components. In this way, the outer unit cells are patchy. As
The way to arrange unit cells refers to pattern design [475], which de­ such, conformal patterning has received considerable attention.
termines the final configuration of lattice structures. Generally, three
design methods, namely, direct patterning, conformal patterning, and 4.2.2. Conformal patterning
topology optimization, are commonly applied. To create a lattice structure with a specific shape, all unit cells must
be constrained in a design space. As mentioned in Section 4.2.1, the
lattice structure built by direct patterning always produces corner

Table 5
Various metallic triply periodic minimal surface (TPMS) lattice structures produced by different additive manufacturing technologies and their mechanical properties
(obtained from compressive tests). E: Young’s modulus; σy: yield strength.
Unit cell Fabrication method Material Porosity (%) E (GPa) σy (MPa) Refs.

Gyroid SLM CP-Ti 72.4 1.47− 2.30 49.0− 53.3 [462]

Gyroid SLM Ti-6Al-4V 80− 95 0.13− 1.25 6.5− 81.3 [454]
Gyroid EBM Ti-6Al-4V 82 0.64 ± 0.05 13.2 ± 1.6 [393]
Diamond SLM Ti-6Al-4V 80− 95 0.12− 1.25 4.7− 69.2 [454]
Schwartz SLM Ti-6Al-4V 51 22.3 ± 2.3 160 ± 12.6 [463]
F-RD SLM Ti-6Al-4V 45 2.81 173.9 [464]
I-WP SLM Ti-6Al-4V 45 2.52 125.0 [464]
Schwartz SLM Ti-25Ta 25− 64 14.3− 36.1 86.0− 319.0 [465]
Gyroid SLM NiTi 78.5 – 29.0 [466]
Gyroid SLM 316 L 85 0.24− 0.31 12.4− 17.0 [467]
Schwarz SLM Fe-35 Mn 42 34.0 ± 2.0 89.0 ± 2.0 [468]

22
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 17. Typical shell lattice structures: (a) octet-truss shell lattice structure, (b) rhombic dodecahedral shell lattice structure, (c) cubic shell lattice structures with
octahedral cells, (d) simple cubic shell lattice structure, (e) octet-truss lattice structure, (f) rhombic dodecahedral lattice structure, and (g) simple cubic lattice
structure (reproduced with permission from Ref. [471], Copyright (2019), Elsevier).

angles, which are against the design of curved surface. Conformal several steps. Considering the algorithm of Algorithm Construct
patterning can converge the unit cells in a designed space and retain Conformal Mesh as an example, the part boundary of the model is
their integrity, thereby generating desirable curved surfaces of lattice divided into relatively flat regions because it is prone to control the mesh
structures. For a comparison, Fig. 19 shows the difference between the generation [479]. For each region, the offset of the object boundary
parts created by direct patterning and conformal patterning, where the should be firstly computed, which is a collection of points at a specified
latter one constructs a more homogenous and curved surface. distance away from the starting surface. Positive offset or negative offset
To create a conformal lattice structure, several main steps are leads to a larger or smaller object, respectively. A tri-parameter volume
required [478]. After the creation of a 3D solid part, a conformal mesh is between the original surface and its offset is subsequently constructed.
computed by a meshing algorithm, followed by the filling of cells in this Both surface and volume are parameterized. A surface parameterization
conformal mesh. At this stage, a model of conformal lattice structure is would be imposed on some regions which need to be reinforced [480].
generated. Note that the computation of conformal mesh needs to follow As such, a lofted volume is constructed between the original and the

Fig. 18. Schematic illustrations of (a) direct patterning method and (b)-(d) some actual AM-produced examples designed by direct patterning. (reproduced with
permission from Ref. [446], Copyright (2018), Elsevier).

23
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 19. 2D parts created by (a) direct patterning and (b) conformal patterning.

offset surfaces. The parameterized volume is then divided into individ­ configurations which are hardly obtained by conventional ideas [72].
ual elements by conformal mesh. For a 3D conformal lattice structure, Owing to its independent configuration, topology optimization has a
the hexahedron is widely used as the individual element [481,482]. broader design space. Therefore, topology optimization has been
When the selection of the number of individual elements and their frequently employed in the automotive, aerospace, and biomedical in­
typical sizes is finished, the algorithm computes the increment. Finally, dustries [73,483–486]. One of the advantages of the optimized structure
the match of region boundaries may be done by adjusting node positions is the efficient use of material to obtain the maximum performance ac­
and/or by adding individual elements. The as-constructed individual cording to a set of constraints. During the topology optimization process,
elements may not match well because the regions are separately iteration is used to determine whether the material in the designed mesh
parameterized. Thus, hexahedral, or tetrahedral elements can be added is retained or not based on the finite element method.
in the gaps between meshes in neighboring regions. The construction of During the past three decades, lots of topology optimization
the model of conformal lattice structure is then finished. methods, such as the density-based method [487,488], the solid
isotropic material with penalization method (SIMP) [489,490], the
4.2.3. Topology optimization evolutionary structural optimization (ESO) [491,492], bi-directional
Topology optimization is to find optimized geometry and is evolutionary structural optimization (BESO) [493–495], and the level
frequently used in both unit cell design and pattern design [37]. Origi­ set method (LSM) [496,497], surface optimization [419], have been
nally, topology optimization creates innovative high-performance reported. For the density-based method, the discrete optimization

Fig. 20. Examples of optimized lattice struc­

tures: (a) a bridge-type structure, which has
retrieved local optimal material topologies and
is designed by bi-directional evolutionary
structural optimization method (reproduced
with permission from Ref. [504], Copyright
(2015), Elsevier), (b) a lattice structure with
diamond unit cells optimized by surface opti­
mization, (c) the optimized diamond unit cells
are controlled by rod diameter (D) and
optimized-radius (R), and (d) the lattice struc­
tures with optimized diamond unit cells have
higher strength than the original lattice struc­
tures with non-optimization at the same rela­
tive density (reproduced with permission from
Ref. [419], Copyright (2018), MDPI).

24
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

problem is changed into the continuous optimization problem to relax Table 6

the binary design form. However, the complication of the mathematical Software for the design of lattice structures.
problem results in difficulty in optimization [498]. Hence, the SIMP Software Company Features Cooperativity
method was developed by Bendsøe and Kikuchi [499]. The elastic
AutoCAD Autodesk Versatile and robust; Available for
modulus of the element is introduced into the SIMP method and expo­ the incapability of Simpleware ScanIP,
nentially penalized according to the density variables compared to the triply periodic minimal Selective Space
density-based method. The ESO method is based on the heuristic stra­ surface design; Structures, Altair
tegies and sets the inefficient material to void until satisfying the inefficient generation of OptiStruct, Autodesk
a large-scale lattice within, and other 3D
requirement of needed material volume [500]. However, effective so­ structure. printing software.
lutions sometimes are difficult to be obtained. Therefore, the BESO MathMod (or Free Allowing to plot 3D Unavailable.
method was developed to extend the ESO method, which allows to add K3DSurf) software mathematical surfaces
or remove material to change the structure [501,502]. LSM method described by implicit or
parametric equations;
employs high-dimensional level set functions to optimize the structure
offering a very large
via iterating the Hamilton-Jacobi equation [503]. database of the model.
For lattice structure, all topology optimization methods control the Simpleware Synopsys Providing a Available for
material distribution to obtain a better performance. As shown in ScanIP comprehensive AutoCAD and other
Fig. 20a, a bridge-type structure is optimized by the BESO method, segmentation software 3D printing software.
environment for
which has retrieved local optimal material topologies [504]. This processing 3D image
structure is achieved on multiscale structural topology optimization. data; Generating lattice
Local materials are optimized based on the current loading status. structures within a
Apparently, topologically optimized structures have significantly com­ given part based on
implicit functions by
plex geometry, which is rather difficult to be manufactured by con­
cooperating with
ventional methods (e.g., investment casting, powder metallurgy, AutoCAD.
wire-woven method in Section 2.1). In this situation, AM technologies Selective Space Autodesk Possessing a standard Available for
enhance the possibility to fabricate topology optimization structures and Structures Netfabb unit cell library and AutoCAD.
realize the potential of topology optimization. Liu et al. [419] used the (3S) allows the user to
define unit cell type;
surface optimization method to optimize a lattice structure with dia­
Solving the difficult
mond unit cells for orthopedic applications. The nodes in the lattice task of creating
structures are modified to have a gradient rod diameter and smooth complex structures
surface compared with the straight beam-like nodes, aiming to fewer within parts; Building
repetitive structure
stress concentrations in the deformation (Fig. 20b). The optimized dia­
cells following the
mond unit cells are controlled by rod diameter (D) and optimized radius model of repetitive
(R) (Fig. 20c). They printed the samples by SLM using Ti–6Al–4 V and spatial lattice structures
investigated their mechanical properties by compressive test. The lattice in nature.
structures with optimized diamond unit cells demonstrate higher Altair Altair Being a proven, modern Available for
OptiStruct Engineering structural solver with AutoCAD and other
strength than the original lattice structures with non-optimization at the
comprehensive, 3D printing software.
same relative density (Fig. 20d). The enhancement in strength is accurate, and scalable
attributed to the release of stress concentration from the optimized solutions for linear and
surface, which changes the failure mechanism of Ti–6Al–4 V lattice nonlinear analyses
across statics and
structures from the pattern of bottom-up collapse by layer to that of the
dynamics, vibrations,
diagonal shear band [419]. acoustics, fatigue, heat
transfer, and multi-
4.3. Software for pattern design physics disciplines;
Being used to optimize
the lattice structures.
Proper use of software gives the efficient pattern design of lattice Autodesk Autodesk Being a set of Available for
structures, and some common software is listed in Table 6. Each soft­ within generative design AutoCAD.
ware has its own advantage for designing and optimizing patterns. software solutions
aimed at engineering
users who wish to build
5. Mechanical properties of metallic lattice structures fabricated
and 3D print
by additive manufacturing lightweight objects for
the automotive,
Metallic lattice structures are primarily developed for tailoring the medical implant,
mechanical properties of products. Therefore, this section only describes aerospace, and
industrial equipment
the mechanical properties of metallic lattice structures, and other markets; Being used to
properties are introduced in Section 6. For various purposes, different optimize the lattice
mechanical properties of metallic lattice structures may receive exten­ structures.
sive attention. For example, the applications of lightweight structures Conformal Paramount Being an innovative Available for
Lattice method for reinforcing AutoCAD.
require high specific stiffness and strength, while the applications of
Structures components with
energy absorbers need large deformation at a relatively lower stress (CLS(TM)) designed performance
level. In addition, fatigue resistance is also important because the to achieve substantial
metallic lattice structures may bear a cyclic load in some applications (e. weight to strength
g., load-bearing biomedical devices). In such a situation, most service ratios.
3-Matic STL Materialize Cleaning up rough data Available for
failures of metallic lattice structures can be ascribed to fatigue, which for simulation or AutoCAD and other
takes place at a significantly lower stress level relative to their tensile/ convert the mesh back 3D printing software.
compressive strength [505]. Generally, the mechanical properties of (continued on next page)
metallic lattice structures have been investigated systematically and

25
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 6 (continued ) continuous collapse of struts until the densification takes place. Because
Software Company Features Cooperativity such two lattice structures have different deformation behavior, their
applications also have different targets. The stretch-dominated struc­
to CAD; Enhancing the
design by creating 3D
tures are suitable for the applications of lightweight structures which
textures, lattice require high specific stiffness and strength. By comparison, for the ap­
structures, and plications of energy absorbers, the bending-dominated structures are the
conformal structures. first design priority because they have large deformation at relatively
lower stress levels.
Indeed, the deformation of strut-based lattice structures can also be
they are primarily influenced by their unit cells, porosities, materials
considered as the buckling and bending deformation of struts (respec­
used, and manufacturing methods adopted.
tively for stretch-dominated structures and bending-dominated struc­
tures). However, both buckling deformation and bending deformation
5.1. Effect of unit cells are always coupled during the deformation of strut-based lattice struc­
tures owing to their complex geometries. Li et al. [445] investigated the
Due to the existence of different unit cells, metallic lattice structures deformation behavior of EBM-produced Ti–6Al–4 V lattice structures
have demonstrated different deformation behavior compared with bulk with cubic, G7, and rhombic dodecahedron unit cells. Fig. 22 shows the
metals. As shown in Fig. 15, although the various unit cells have iden­
tical relative densities, they still exhibit distinct strength. Take the strut-
based lattice structures as examples, stretch and bending may take place
when the lattice structures are deformed [105]. Therefore, the
strut-based lattice structures can be roughly divided into the
bending-dominated structure and the stretch-dominated structure.
Fig. 21 illustrates the deformation behavior of the bending-dominated
structure and stretch-dominated structure [105]. The
bending-dominated structure is elastically deformed up to its elastic
limit. From the point of plasticity, the structure may be plastically
deformed, buckling or fracture, resulting in a broad plateau region
presented in the stress-strain curves (Fig. 21a). The bending-dominated
structures often have a relatively lower yield strength but higher energy
absorption. Because the internal pores are almost closed and the struts
merge, the stress dramatically increases. This phenomenon is called
densification. For the stretch-dominated structures, their modulus and
yield strength are relatively higher compared with the
bending-dominated structures with the same relative density (Fig. 21b).
A softening post-yielding response is found after yielding due to the

Fig. 22. (a) A schematic illustration depicts the buckling/bending vectors, (b)
Fig. 21. Illustration of deformation behavior of the stretch-dominant structure unit cells designed to increase the bending component of the load applied on
and bending-dominant structure: (a) bending-dominant structure and (b) the struts, and (c) compressive results of the designed lattice structures
stretch-dominant structure. (reproduced with permission from Ref. [445], Copyright (2014), Elsevier).

26
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

schematic illustration to depict the buckling/bending vectors and unit The void/pores are unable to resist the external force. Therefore, lower
cells designed to increase the bending component of the load applied on porosity (therefore higher relative density) would generally result in the
the struts. According to the direction of applied force, the buckling/­ higher mechanical properties of lattice structures made of identical
bending vectors can be analyzed (Fig. 22a). For cubic unit cells, only a material [119]. In other words, the increased relative density of lattice
buckling vector is found. For G7 and rhombic dodecahedron unit cells, structures is associated with increased mechanical properties. As such,
both buckling/bending vectors are found and both vectors have an angle the mechanical properties of lattice structures can be tailored by varying
with the applied load. Fig. 22b shows the designed unit cell for their relative density besides the types of unit cells. However, the
increasing the bending component. The angles between the applied load increased relative density of lattice structure would also result in the
and the beam (for the bending component) are designed to be 36 ◦ , 30 ◦ , increased consumption of material and the increased weight of the final
and 23 ◦ . According to the compressive test, the stress plateau of the products, which is opposite to the concept of lightweight. Meanwhile,
structures becomes smooth with decreasing the angle, while their the material used also has an important influence on the mechanical
strength and moduli correspondingly decrease (Fig. 22c). Energy ab­ properties of AM-produced metallic components (including lattice
sorption is a critical consideration for the fatigue properties of lattice structures). The materials have their natures on mechanical properties.
structures, which should be considered from the unit cell design. Liu For example, pure Ti has a higher strength than pure Mg; almost all Ti
et al. [446] used SLM to produce Ti–24Nb–4Zr–8Sn lattice structures alloys have higher strength than pure Ti. Therefore, the selection of
with cubic, topology optimized, and rhombic dodecahedron unit cells, materials used would congenitally determine the level of mechanical
respectively. The compressive tests showed that the lattice structures properties of metallic lattice structures. Liu et al. [48,225,430,446]
with topology optimized unit cells exhibit a low Young’s modulus (~2.3 conducted a series of investigations with respect to the fatigue properties
GPa) and high compressive strength (~58 MPa) together with high of Ti alloy lattice structures. They used EBM to prepare
elastic energy absorption. Therefore, the mechanical properties of lattice Ti–24Nb–4Zr–8Sn lattice structures with rhombic dodecahedron unit
structures can be tailored by the unit cells used without changing the cells and investigated the effect of porosity on the compressive proper­
geometry of unit cells and relative density. Therefore, designing proper ties and fatigue properties of the produced lattice structures (Fig. 23a).
unit cells is an efficient way to enhance the mechanical properties of Ti–6Al–4 V lattice structure with the same unit cells and 75 % porosity
metallic lattice structures. was used as reference. As shown in Fig. 23b, both compressive strength
There are still rare systematical investigations on the deformation and fatigue strength of lattice structures decrease with increasing the
mechanism of TPMS and shell lattice structures. Nevertheless, the porosity [48]. They stated that the decreased porosity would increase
deformation mechanisms of TPMS and shell lattice structures can be the materials used (relative density) in a unit cell and therefore the
deduced from strut-based lattice structures. For shell lattice structures, strength of the lattice structure. For the identical porosity (75 %),
they are evolved from the strut-based lattice structures [469,470]. The Ti–24Nb–4Zr–8Sn lattice structures exhibit higher compressive and fa­
deformation behavior of strut-based lattice structures depends on the tigue strength compared with the Ti–6Al–4 V counterpart with the same
angle between the applied force and struts. This basic principle is also porosity built from the same unit shape (Fig. 23c). CP–Ti lattice struc­
suitable for shell lattice structures [69]. The angle between the applied ture with topology optimized unit cells shows excellent fatigue proper­
force and struts/plates would determine the bending behavior or stretch ties with an ultra-high normalized fatigue life of ~0.65 at 106 cycles and
behavior in shell lattice structures. As mentioned in Section 4.3.1, the a low density of 1.3 g/cm3 despite the Ti–6Al–4 V counterparts with
difference in the deformation mechanism between the strut-based lattice same porosity and same unit cell having higher compressive strength.
structures and the shell lattice structures can be ascertained based on The reason for this phenomenon is the reduction in fatigue crack
their different failure modes. The loading is difficult to be transferred propagation rate, owing to the high plasticity of CP–Ti and the occur­
among different struts in the strut-based lattice structures, while rence of fatigue crack deflection (Fig. 23d-f). Similar results were also
different struts or plates can stand up to the load synergistically [69]. found by other researchers [507,508]. Ahmadi et al. [509] reported that
Therefore, the shell lattice structures have higher stiffness than the the order for the fatigue resistance is as follows: Co–Cr > CP–Ti > pure
strut-based lattice structures at the same relative density. TPMS lattice Ta > Ti–6Al–4 V for the metallic lattice structures with identical patterns
structures are not limited by loading direction and show triply isotropic but with different materials. Liu et al. [225] found that non-layer-wise
[506]. Meanwhile, multi-directional loading can be carried by TPMS fracture can be achieved in the SLM-produced β-type
lattice structures and improved controllability of platform strength is Ti–25Nb–3Zr–3Mo–2Sn (TLM) lattice structure with 50 % porosity cubic
therefore obtained. Wang et al. [506] built the models of Schwarz, unit cells. Such a finding is attributed to the stress-induced martensitic
Schoen, and Neovius TPMS lattice structures and numerically investi­ transformation (β→α′′ ) accompanied by twinning, leading to the uni­
gated their collapse modes. Their results showed that the TMPS lattice form deformation without cracks even under a strain of ~50 %
structures can obtain more reliable mechanical properties when the cell (Fig. 24a). Thanks to the non-layer-wise fracture behavior,
number is no less than 8 × 8. The most central unit cell is squeezed by six SLM-produced TLM lattice structure with 50 % porosity cubic unit cells
surrounding unit cells. Therefore, the entire collapse firstly takes place reveals higher failure strain than the lattice structures with 50 %
in the central region and subsequently relates to the adjacent unit cells. porosity cubic unit cells produced other materials (e.g., Ti-6Al-4 V,
In the meantime, the strength of TMPS lattice structures is also related to AlSi12Mg, and Co-Cr alloys) (Fig. 24b) [225]. However, unlike bulk
the angle between the applied force and the surfaces. Although all sur­ metals, the influence of the material used on their fatigue properties is
faces are curved, a lower angle between the applied force and the sur­ significantly complicated. Because different materials have different
faces always results in lower stiffness and strength of TMPS lattice natures, such as melting point, flowability, and thermal conductivity, it
structures. Such a feature is commonly found for all the strut-based is rather difficult to fabricate identical lattice structures with the merely
lattice structures, the shell lattice structures and the TMPS lattice difference in the materials; there are always some differences in
structures. microstructural features, e.g., porosities and surface roughness, there­
fore mechanical properties (and other properties as well), even though
5.2. Effect of porosity and materials same manufacturing technologies and same processing parameters are
adopted. Note that, further understanding regarding the influence of
Porosities of metallic lattice structures also have a significant influ­ unit cells and materials on the mechanical properties of lattice structures
ence on their mechanical properties because the porosity determines the is needed.
relative density of lattice structures. As mentioned earlier in Section 2.1,
porosity is the volume ratio of pores/voids in the bulk material [21].
When a lattice structure is loaded, the loading only acts on the materials.

27
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 23. (a) Finite element models for the stress distribution of the rhombic dodecahedron lattice structures with 67.9 % and 91.2 % porosity at the 3% compressive
strain, (b) typical compressive stress-strain curves, and (c) normalized S-N curves for Ti–24Nb–4Zr–8Sn and Ti–6Al–4 V samples with different porosities (reproduced
with permission from Ref. [48], Copyright (2017), Elsevier), (d) compressive stress-strain curves of CP–Ti and Ti–6Al–4 V lattice structures with topology-optimized
and rhombic dodecahedron unit cells and (e) their corresponding normalized S-N curves, and (f) morphology of fatigue deflection on the surface of
topology-optimized CP–Ti lattice structure (reproduced with permission from Ref. [430], Copyright (2020), Elsevier).

Fig. 24. (a) 2D XRD patterns and the corresponding azimuthally (0-360 ◦ ) unrolled diffraction images of the SLM-produced Ti–25Nb–3Zr–3Mo–2Sn (TLM) lattice
structures with 50 % porosity cubic unit cells under different loadings and (b) the relationship between the porosity and failure strain of different cubic lattice
structures from different materials (reproduced with permission from Ref. [225], Copyright (2021), Elsevier).

5.3. Effect of manufacturing energy density of the SLM process. Therefore, more Sn evaporates dur­
ing the SLM process, leading to the production of pores in the lattice
Manufacturing (or processing) is a complicated influence factor on structures. As a result, Liu et al. [47] found that the cyclic ratcheting and
the mechanical properties of metallic lattice structures because lots of the surface properties are the main factors determining the fatigue
processing parameters are involved during the AM process. For a behavior of the EBM- or SLM-produced lattice structures at the lower
designed lattice structure model, the physical product may be signifi­ stress levels, and both EBM or SLM lattice structures exhibit similar fa­
cantly different resulted from the different AM technologies adopted or tigue properties. In comparison, the SLM lattice structures have a lower
even the different parameters used. Fig. 25 shows micro-CT recon­ and more variable fatigue life at higher stress levels because the crack
structed images of EBM- and SLM-produced Ti–24Nb–4Zr–8Sn lattice initiation and propagation take place from pores.
structures [47]. The SLM-produced Ti–24Nb–4Zr–8Sn lattice structure Because of the differences in the materials and structures, the pro­
contains a higher number of pores. Such a difference is mainly caused by duced lattice structures would also have various surface roughnesses
the different spot sizes during the EBM and SLM processes (200 μm for despite the same AM technologies with the same parameters used.
electron beam and 40 μm for laser), which results in a higher local Ahmadi et al. [509] reported that the metallic strut-based lattice

28
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 25. Micro-CT reconstructed images of the EBM- and SLM-produced Ti–24Nb–4Zr–8Sn lattice structures: (a), (b) the strut outside surfaces, (c), (d) the defects in
the solid struts, and (e) the size and count distribution of the defects inside the lattice structures as a function of equivalent diameter (reproduced with permission
from Ref. [47], Copyright (2016), Elsevier).

structures with a smoother surface have a higher fatigue resistance than different AM technologies, which gives a comparison among the unit
their counterparts with a relatively rougher surface. Such a phenomenon cells, porosity, and fabrication methods [511]. Low fatigue life is still a
was also found for the TPMS lattice structures. The TPMS lattice struc­ question for the AM-produced metallic lattice structures [446,512,513].
tures have curved shapes, which may mitigate stress concentration to a Unlike bulk metals, the lattice structures are difficult to be physically
certain degree [103,510]. Hence, the fatigue resistance is enhanced. polished to reduce their surface roughnesses. Therefore, stress concen­
Yánez et al. [511] found the crack initiation on the TPMS lattice struc­ tration is prone to be generated, which facilitates crack initiation and
ture and confirmed that the stress concentrators are located at surface accelerates crack propagation.
defects by finite element analyses. Therefore, cracks may be also pro­ Different manufacturing technologies may result in the different
duced if the TPMS lattice structure has a high surface roughness. Table 7 microstructures of the produced lattice structures. Heinl et al. [518] and
compares the fatigue strength of different lattice structures produced by Hrabe et al. [515] respectively fabricate Ti–6Al–4 V lattice structures

29
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 7
Fatigue properties of different lattice structures produced by additive manufacturing. The data were obtained from compression− compression fatigue tests. The
superscripts of S, H, and A indicate the post− treatments of sand− blast, hot isostatic pressure, and annealing, respectively (reproduced with permission from Ref. [511],
Copyright (2020), Elsevier).
Material Unit cell Fabrication method Porosity (%) Cycles Fatigue strength/yield strength Refs.
6
Ti–6Al–4V Cellular gyroid EBM 90 10 0.21–0.23 [511]
Ti–6Al–4V Sheet-based gyroid SLM 85− 87 106 0.3–0.5 [514]
Ti–6Al–4V Sheet-based gyroid SLM 52− 66 7 × 105 0.6 [457]
Ti–6Al–4V Diamond EBM 60–83 106 0.15–0.25 [515]
Ti–6Al–4V Diamond SLM 67–89 2 × 105 0.32 [508]
Ti–6Al–4V Diamond SLM 66–89 106 0.23 [508,509]
Ti–6Al–4V Diamond SLM 67–72 106 0.19 [516]
Ti–6Al–4V Diamond SLMS 67–72 106 0.2 [516]
Ti–6Al–4V Diamond SLMH 67–72 106 0.21 [516]
Ti–6Al–4V Diamond SLMA 67–72 106 0.23 [516]
Ti–6Al–4V Rhombic dodecahedron SLM 66–84 2 × 105 0.2 [508,517]
Ti–6Al–4V Rhombic dodecahedron SLM 66–84 106 0.15 [508,509]
Ti–6Al–4V Truncated cuboctahedron SLM 67–80 2 × 105 0.36 [508]
CP–Ti Rhombic dodecahedron SLM 67–82 106 0.42 [509]
316 L Cellular gyroid SLM 85 2 × 106 0.48 [459]
316 L Cellular gyroid SLMS 85 2 × 106 0.56 [459]
NiTi Cellular gyroid SLMA 75 106 0.34 [466]
NiTi Sheet-based gyroid SLMA 73 106 0.52 [466]
NiTi Octahedron SLMA 74 106 0.13 [466]
Co–Cr Diamond SLM 60–79 106 0.5 [509]

Co–Cr Truncated cuboctahedron SLM 61–76 106 0.7 [509]

Co–Cr Rhombic dodecahedron SLM 59–70 106 0.6 [509]

with diamond unit cells by SLM and EBM. The SLM-produced Ti–6Al–4 alloys, Co–Cr-based alloys, Mg alloys, and Zn alloys, have been used as
V sample with 60 % porosity has the yield strength of 112.7 ± 0.6 MPa, biomedical implants to replace the dysfunctional hard tissues, owing to
while the EBM-produced Ti–6Al–4 V counterpart has the yield strength their high strength, excellent corrosion resistance and excellent
of 99.7 ± 2.9 MPa. The higher yield strength of the SLM-produced biocompatibility [63,523–530]. Although these metallic materials meet
Ti–6Al–4 V sample is attributed to the dominated α′ phase in the the requirement of strength, they possess significantly higher elastic
microstructure. The α′ phase is frequently observed in the hexagonal moduli than the human bones [282,513,531–533]. Such a difference in
metals after fast cooling from β region to α region and enhance the elastic moduli would cause the so-called “stress-shielding effect” phe­
strength of the alloys [519–522]. Therefore, owing to the fast-cooling nomenon in the human body and hamper the transfer of required stress
rate in the SLM process, the SLM-produced Ti–6Al–4 V sample has the from implant material to the adjacent bone [534–536]. Consequently,
higher strength than the EBM-produced counterpart. The SLM and EBM the bone cells would be adsorbed around the implants, leading to bone
processes have high precision in the production of the lattice structure. death and the failure of implantation [537,538]. As such, endeavors
In comparison, other AM technologies available for metallic lattice have been made to reduce the elastic moduli of implanted materials to
structures, such as BJ and DIW, have relatively low precision. Li et al. improve their mechacompatibility (i.e., mechanical compatibility) and
[62] produced a Fe lattice structure with diamond unit cells and 80 % sustainability in the human body. For example, designing and devel­
porosity. The produced Fe lattice structures show a significantly low oping novel β Ti alloys with lower elastic moduli is a receptible way for
strength of 40 MPa and low elastic moduli of 1.6–1.8 GPa. The Fe lattice the biomedical industry [226,539–544]. However, to the best knowl­
structures produced by BJ and DIW have relatively low porosities. edge of the authors, all new β-type Ti alloys still have higher elastic
Hence, the Fe lattice structures produced by BJ and DIW have a rela­ moduli than the human bones [526,539].
tively high strength of 190 and 141 MPa. Meanwhile, as mentioned in In 1972, Weber and White put forward the concept of porous ma­
Section 3.2.6, the parts produced by BJ would have a loose structure and terials to osseointegration [545]. Afterward, porous materials, including
a significantly high surface roughness due to the evaporation of the ceramics, polymers, and metallic materials, were widely investigated
binder during sintering (Fig. 11). Such a morphology of structure cannot since the 1970s. Porous ceramics and polymers cannot meet the re­
obtain satisfying mechanical properties. quirements under load-bearing conditions, because porous ceramics
have intrinsic brittleness and may fracture under loading while porous
6. Applications and performances of metallic lattice structures polymers cannot withstand the mechanical force in joints. Therefore,
produced by additive manufacturing porous metallic materials were focused on by researchers because of
their sufficient mechanical strength and good biocompatibility required
Metallic lattice structures are appealing for their fascinated sets of for implants. Porous metallic biomaterials with reduced elastic moduli
mechanical and physical performances. The low weight associated with have been rapidly developed during recent years [168,177,546].
a relatively high stiffness or a high energy absorption is commonly Generally, the modulus and the strength of a material are inversely
designed. The applications of metallic lattice structures can primarily be proportional to the porosity level in the material. Therefore, many
classified into two categories: structural applications or functional ap­ porous metallic biomaterials are produced by powder metallurgy [140],
plications. Structural applications can include biomedical implants, foaming [547], investment casting [548], etc. because these methods
load-bearing parts (e.g., lightweight structures), or energy absorbers, can produce porous materials. However, as mentioned earlier in Section
whereas functional applications can refer to heat exchanges, catalysts, 2, conventional manufacturing technologies have low controllability in
or acoustic insulators. the porosity and distribution and size of pores in the cellular structures.
Meanwhile, as the hard tissues in the human body, bones have a
6.1. Biomedical implants hierarchical structure from macroscale to sub-nanoscale and each
structural level reveals various mechanical, biological, and chemical
Since the 1920s, bulk metallic materials, such as stainless steels, Ti functions [60]. Bones can be classified into two categories: cortical bone

30
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

and cancellous bone. Cortical bone has 3~5% space for blood vessels, fabricated by AM technology [557]. Fig. 26a shows the designed model
osteocytes, canaliculi, etc., while cancellous bone has a large space that created by CAD software and optimized by finite element analysis and
ranges from 50 % to 95 % [549]. The space of cancellous bone is filled density-based topology optimization. Fig. 26b represents the fabricated
with marrow and the pores in the cancellous bone range from 200 μm to implant. From design to fabrication, AM technology can almost restore
1000 μm [60,550]. Therefore, porosity, pore size, and pore inter­ the model to an ideal implant and have a precision fabrication.
connectivity are essential considerations for an implant, which signifi­ Currently, extensive endeavors have been dedicated to investigating
cantly influences the mechanical properties and biocompatibility. The metallic lattice structures for biomedical applications. Pure Ti
mechanical properties of lattice structures (or cellular structures) have [558–560], Ti alloys [561–565], Co–Cr-based alloys [566,567], stainless
been discussed in the previous sections. Here, the effect of pores on the steel [568–570], NiTi alloys [571–575], and Ta [34,576–578] are the
biocompatibility of lattice structures is stressed in this section. It is well most commonly used bio-inert metals for fabricating lattice structures
known that porous biomedical implants have notable biocompatibility, by AM technologies. Fig. 27 shows some lattice structures produced by
which can enhance bone regeneration [551]. A suitable porous AM technologies using bio-inert metals, where some of them have been
biomedical implant needs to be open-cellular and interconnected, successfully employed in clinic applications [34,579]. Meanwhile, a
facilitating cell migration and blood vessel formation [21]. Ataee et al. considerable number of investigations have shown the understanding of
[393] suggested that the porosity of porous biomedical implants has a the structure-property relation in metallic lattice structures using
minimum requirement of 60 %. Torres-Sanchez et al. [552] investigated various AM fabrication methods and designed lattice patterns [47,466,
the biocompatibility of porous Ti implants (seeded with osteosarcoma 580–582]. For example, Liu et al. [47] compared the microstructures,
osteoblasts 143B and incubated for 12 days) and the pore sizes of these defects, and mechanical behavior of β-type Ti–24Nb–4Zr–8Sn rhombic
implants have four different pore size ranges, i.e., 45–106 μm, 106–212 dodecahedron structures (porosity: 75 %) fabricated by SLM and EBM,
μm, 212–300 μm, and 300–500 μm. It is found that small pores are and found that EBM- and SLM-produced Ti–24Nb–4Zr–8Sn rhombic
beneficial for cell attachment and large pores can promote cell prolif­ dodecahedron structures have α + β phases and a single β phase,
eration [552]. Compared with other implants, the Ti implant with the respectively, in the microstructure. The faster cooling rate of
pore size of 45–106 μm has a higher cell growth rate in the first three Ti–24Nb–4Zr–8Sn alloy during the SLM process triggers the generation
days and a lower cell growth rate after the first three days. The Ti of fine β dendrites, leading to a higher compressive strength (50 ± 0.9
implant with a pore size of 300–500 μm has a tremendously higher cell MPa) than the EBM-produced counterpart with α + β phases
growth rate after the first day. Moreover, Bose et al. [553] pointed out (compressive strength: 45 ± 1.1 MPa). In addition, the elastic moduli of
that the porous implants with combined macro-pores (over 50 μm) and lattice structures can be altered by adjusting the porosity. Wu et al.
micro-pores (below 20 μm) would show better biocompatibility than [580] used EBM to produce gradient Ti–6Al–4 V lattice structures with
those only having macro-pores. The micro-pores result in the increasd 70 %~90 % porosity and the produced lattice structures show the elastic
surface area of the implant for osteogenic protein adsorption and oste­ moduli close to natural bones. Regarding the effects of the types of unit
oblast cell attachment, enhancing bone regeneration. cells on the properties of lattice structures, Speirs et al. [466] found that
Metallic lattice structures manufactured by AM technologies which the NiTi lattice structures with cellular gyroid and sheet gyroid unit cells
have more accurate fabrication ability well overcome the drawbacks of have better static mechanical properties than those with octahedron unit
cellular structures produced by conventional manufacturing technolo­ cells at identical volume fractions. Hence, such NiTi lattice structures
gies [303,554–556]. For example, as shown in Fig. 26, a with cellular gyroid and sheet gyroid unit cells may be superior to the
topology-optimized lattice structure with gradient unit cells was ones with octahedron unit cells in load-bearing applications.

Fig. 26. A topology optimized lattice structure with gradient unit cells fabricated by additive manufacturing technology: (a) designed model and (b) fabricated
implant (reproduced with permission from Ref. [557], Copyright (2016), Elsevier).

31
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

used SLM to manufacture a 316 L stainless steel lattice structure with 87

% porosity for the potential applications of trabecular bone and the
results indicated that the produced sample showed an elastic modulus of
0.15 GPa and compressive yield strength of ~3 MPa. Table 8 summa­
rizes the lattice structures of primary bioinert metals produced by AM
technologies as well as their mechanical properties. Apparently, the
strength of the lattice structures is influenced by the fabrication method,
materials used, porosity, and unit cell. For the identical material and
unit cell, the SLM-produced lattice structures have a higher static me­
chanical strength than the EBM-produced ones when the lattice struc­
tures have similar porosity level, owing to the better surface roughness
of the produced parts [47]. However, Liu et al. [47] found that the
vaporization of Sn takes place during the fabrication of
Ti–24Nb–4Zr–8Sn by both SLM and EBM, which results in the produc­
tion of defects in the lattice structure. Because of a smaller laser spot size
and a faster cooling rate, the Sn vapor is prone to be trapped during the
SLM process. The SLM-produced Ti–24Nb–4Zr–8Sn lattice structures
have 10 times of defects compared with the counterparts produced by
EBM. Therefore, the SLM-produced Ti–24Nb–4Zr–8Sn lattice structures
show a lower fatigue strength than the EBM-produced one. The brittle
layer-wise fracture is a major obstacle restricting the development of
metallic lattice structures, which is commonly found and may induce
catastrophic failure under occasional overloading.
There are rare reports with respect to the clinic applications of
biodegradable metal lattice structures although they have been suc­
cessfully fabricated by AM technologies [62,196,468,477,555,598].
Biodegradable metals mainly consist of Mg [599,600], Fe [62,477,598,
601], and Zn [555]. Fig. 28 shows examples of lattice structures made of
biodegradable metals [62,468,555,601–604]. Fig. 28a and b show two
examples of WE43 Mg lattice structures produced by SLM. Fig. 28c is the
magnified SEM image of the structure in Fig. 28b. Fig. 28d-g display the
pure Zn lattice structures produced by SLM potentially used for the
implant and intravascular stent. Fig. 28h is the pure Fe lattice structure
produced by SLM. Fig. 28i andj represent the pure Fe lattice structure
produced by DIW; the produced structure is not elaborated compared to
pure Fe lattice structures produced by SLM. Fig. 28k shows the Fe–35 Mn
lattice structure produced by SLM used for bones for load-bearing ap­
plications. Unlike bioinert metals, biodegradable metals such as Mg and
Zn have low melting points of 650 ◦ C and 420 ◦ C, respectively [604].
Severe evaporation may take place during the AM process [605,606].
Pre-heating of the substrate and the chamber with overpressure has been
reported to overcome the severe evaporation of Mg alloys during the AM
process [240,607]. Until 2015, the first SLM-produced AZ91 Mg lattice
structure was prepared by Jauer et al. [608]. Later in 2018, Li et al.
[602] prepared a WE43 Mg lattice structure with diamond unit cells and
firstly reported its in vitro performance. After 4 weeks of biodegradation
in simulated body fluid with 5% fetal bovine serum, the WE43 Mg lattice
structure had a loss of 20 % volume fraction (Fig. 29a). Meanwhile, the
WE43 Mg lattice structure revealed a less than 25 % cytotoxicity in vitro
after 72 h cultivation of MG-63cells (Fig. 29b). Compared with Mg, Zn
has a lower melting point and hence the evaporation of Zn-based alloys
should be carefully considered during the AM process. Demir et al. [609]
obtained SLM-produced Zn cubic parts with 98 % density in the open air
Fig. 27. Various metallic lattice structures produced additive manufacturing:
with an auxiliary purge gas stream. Wen et al. [604] used SLM to
(a) NiTi lattice structures with the octahedron, cellular gyroid and sheet gyroid
unit cells (reproduced with permission from Ref. [466], Copyright (2017),
fabricate pure Zn lattice structure for cardiovascular stents with
Elsevier), (b) and (c) 316 stainless steel lattice structure produced by selective different diameters (i.e., 5, 4, 3, and 2 mm) and the formation quality of
laser melting and corresponding SEM image (reproduced with permission from the produced lattice structure is acceptable. Wen et al. [603] also
Ref. [582], Copyright (2016), Elsevier). fabricated three types of Zn lattice structures for potential applications
on implant osseointegration and demonstrated the availability of SLM
Lattice structures produced by other metallic materials are also re­ for all specific lattice structures. Due to the low melting point, it is still
ported. Cutolo et al. [581] investigated the influence of layer thickness difficult to fabricate a Zn-based lattice structure with high quality
and post-process treatments on the fatigue properties of SLM-produced because Zn-based alloys are prone to evaporate during the AM process.
CoCr lattice structure with diamond unit cell and found that Fe is another biodegradable metal, which has similar properties to
increasing the layer thickness led to a significant improvement in the stainless steel. Fe has a relatively lower corrosion rate than Mg and Zn in
productivity for CoCr parts (about 40%–50%). Čapek et al. [582] also the human body. Therefore, bulk Fe cannot be used as a biodegradable
implant. Nevertheless, the Fe lattice structure has a larger surface area

32
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 8
Summary of unit cells, porosity, elastic modulus (E), compressive strength (σ max), and yield strength (σ y) of additive manufacturing produced lattice structures from
several bioinert metals.
Material Fabrication Method Unit cell Porosity (%) E (GPa) σy (MPa) σmax (MPa) Refs.

CP-Ti (Grade 1) SLM Dodecahedron 66− 81 0.58− 2.61 8.6− 36.5 – [583]
CP-Ti (Grade 1) SLM Gyroid 72.4 1.47− 2.30 49.0− 53.3 – [462]
CP-Ti (Grade 2) SLM Cellular 37 13 ± 3 113 ± 0.3 – [584]
Ti-6Al-4V SLM Cubic 72.8 12.6 ± 3.3 16.6 ± 0.5 – [585]
Ti-6Al-4V EBM Cubic 80 1.6 ± 0.3 22.0 ± 1.1 29.3 ± 0.8 [12]
Ti-6Al-4V EBM Diamond 60− 87 0.4− 6.5 11.4− 99.7 16.3− 118.8 [518]
Ti-6Al-4V SLM Diamond 65 2.481 76.2 – [586]
Ti-6Al-4V SLM Truncated cube 80− 91 0.99− 3.19 10− 40 – [587]
Ti-6Al-4V SLM Truncated cuboctahedron 64− 81 2.37− 4.62 25− 100 – [587]
Ti-6Al-4V SLM Rhombic dodecahedron 68− 89 0.22− 2.97 7− 88 – [587]
Ti-6Al-4V EBM Rhombic dodecahedron 62.7− 83.8 – – 20− 90 [588]
Ti-6Al-4V EBM Rhombic dodecahedron 58− 88 0.5− 6.5 – 10− 100 [445]
Ti-6Al-4V SLM Rhombic cuboctahedron 68− 89 2.23− 4.40 39− 93 – [587]
Ti-6Al-4V SLM Octet truss 50− 75 1.2− 4.6 34− 172 – [589]
Ti-6Al-4V SLM Tetrahedron 50− 75 1.9− 4.3 57− 156 – [589]
Ti-6Al-4V EBM Honeycomb 66 2.5 ± 0.5 – 116 ± 10 [590]
Ti-6Al-4V SLM Cuboctahedron 58 2.77 119.4 – [586]
Ti-6Al-4V SLM Twist struts 55− 61 3.4− 26.3 103− 402 – [591]
Ti-6Al-4V EBM Hatched structure 60 12.9 ± 0.9 107.5 ± 3.6 148.4 ± 3.5 [12]
Ti-6Al-4V EBM G7 58− 88 0.5− 5 – 8− 80 [592]
Ti-6Al-4V SLM F-RD 45 2.81 173.9 – [593]
Ti-6Al-4V SLM I-WP 45 2.52 125 – [593]
Ti-6Al-4V SLM Gyroid 80− 95 0.13− 1.25 6.5− 81.3 – [454]
Ti-6Al-4V EBM Gyroid 82 0.64 ± 0.05 13.16 ± 1.6 24.4 ± 2.4 [393]
Ti-6Al-4V SLM Diamond TPMS 80− 95 0.12− 1.25 4.7− 69.2 – [454]
Ti-6Al-4V EBM Solid cellular foam 55− 82 0.75− 9.97 – – [594]
Ti-6Al-4V EBM Hollow cellular foam 55− 89 0.48− 9.77 – – [594]
Ti-24Nb-4Zr-8Sn SLM Rhombic dodecahedron 75 0.95 ± 0.05 – 50 ± 0.9 [47]
Ti-24Nb-4Zr-8Sn EBM Rhombic dodecahedron 75 1.34 ± 0.04 – 45 ± 1.1 [47]
Ti-24Nb-4Zr-8Sn EBM G7 70 0.7 ± 0.1 – 35 ± 2 [595]
Ti-35Zr-28Nb SLM Face centered cubic 83 1.1 ± 0.1 – – [423]
Ti-35Zr-28Nb SLM Face and body centered cubic 50 1.3 ± 01 – – [423]
NiTi SLM Octahedron 79.3 – 21 – [466]
NiTi SLM Gyroid 78.5 – 29 – [466]
Ti-25Ta SLM Schwartz 25− 64 14.3− 36.1 86− 319 – [465]
Co-28Cr-6Mo SLM Cubic 81 7.18 ± 0.47 – 271.5 ± 20.9 [596]
Co-28Cr-6Mo SLM Face centered cubic 76 12.88 ± 0.64 – 501.7 ± 12.6 [596]
Co-28Cr-6Mo SLM Body centered cubic 77 11.81 ± 0.59 – 453.9 ± 28.4 [596]
Co-28Cr-6Mo SLM Spherical hollow cubic 77 8.79 ± 0.28 – 315.5 ± 17.6 [596]
316 L stainless steel SLM Cubic 87 0.15 3 – [578]
316 L stainless steel SLM Gyroid 85 0.24− 0.31 12.4− 17.0 – [467]
Ta SLM Dodecahedron 80 1.22 ± 0.07 12.7 ± 0.6 – [578]
Ta SLM Diamond 70 3.10 ± 0.03 – – [597]

and therefore exhibits a higher corrosion rate compared to their bulk produce Fe lattice structures with diamond unit cells and investigated
ones [610]. As such, the Fe lattice structure is expected to promote their corrosion fatigue properties. The designed relative density is 20 %
biodegradable ability and its biological application. Yang et al. [601] and the relative density of the real product is 22.3 %. Apparently, the
printed pure Fe lattice structure by a precise 3D plotted system. They fatigue resistance of Fe lattice structures decreases in the revised simu­
modified the surface of lattice structure by hydroxyapatite, resulting in lated body fluid (r-SBF) compared to that in the air (Fig. 31a). After
better bioactivity in load-bearing bone tissue applications. Li et al. [62] fatigue tests, no evident differences between samples are noted. At 2%
printed Fe lattice structures with diamond unit cells by SLM and sys­ strain before failure, both samples tested in air and r-SBF show any
tematically investigated their degradation and corrosion behavior. As apparent visual changes (Fig. 31b). At 5% strain before failure, both
shown in Fig. 30, the as-printed Fe lattice structure has a gradual samples slightly slip sideways. Fig. 30c-e shows the samples after fatigue
decrease in both elastic modulus and yield strength when immersed in test until 5% strain in the air at 0.9 yield strength; the cracks are blunting
revised simulated body fluid [62]. After 28-day immersion, 7% and 5% instead of sharply propagating through the whole strut. Fig. 31f-h dis­
reduction in the elastic modulus and yield strength, respectively, are plays the samples after fatigue test until 5% strain in r-SBF at 0.9 yield
observed. Such results demonstrate that SLM-printed Fe lattice struc­ strength and Fig. 31i-k is the samples after 3 × 106 cycles of fatigue test
tures are promising for biomedical applications. Chou et al. [611] firstly in r-SBF at 0.65 yield strength. Due to the localized degradation, cracks
fabricated Fe–Mn lattice structures by BJ in 2013. The produced Fe–Mn are produced with pits. In turn, the cyclic loading significantly accel­
lattice structure shows similar tensile properties to those of natural erates the degradation rate. Similar results were also reported in
bones, reducing the risk of stress-shielding effect and a significantly high Ref. [196], which investigates the corrosion fatigue behavior of the Zn
corrosion rate than pure Fe. Tables 9 and 10 summarize the lattice lattice structures. For metallic lattice structures used as implants,
structures of primary biodegraded metals produced by AM and their corrosion fatigue behavior is vital for their service time. Nevertheless,
mechanical properties and biodegradation behavior. the investigations with respect to corrosion fatigue behavior of metallic
Bone implants are expected to undergo a huge number of cycles of lattice structures are still in the initial stage.
loading and may have fatigue fracture even the loading on the materials Both in vitro and in vivo tests for metallic lattice structures have been
is much lower than the yield strength of the materials used. Such a conducted. Fig. 32 shows some in vitro tests for metallic lattice struc­
phenomenon is called corrosion fatigue. Li et al. [477] used SLM to tures. Fig. 32a shows the indirect cytotoxicity tests of SLM-produced

33
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 28. Biodegradable metallic lattice structures produced by additive manufacturing technologies: (a)-(b) WE43 Mg lattice structure produced by selective laser
melting and (c) SEM image of (b) (reproduced with permission from Refs. [555,602], Copyright (2018, 2019), Elsevier), (d)-(g) pure Zn lattice structures produced by
selective laser melting (reproduced with permission from Ref. [603,604], Copyright (2018, 2019), Elsevier), (h) pure Fe lattice structures produced by selective laser
melting (reproduced with permission from Ref. [62], Copyright (2018), Elsevier), and (i)-(j) pure Fe lattice structures produced by direct ink writing (reproduced
with permission from Ref. [601], Copyright (2018), ACS publications) and (k) Fe–35 Mn lattice structures produced by selective laser melting (reproduced with
permission from Ref. [468], Copyright (2020), Elsevier).

Ti–6Al–4 V, WE43 Mg alloy, pure Fe, and pure Zn lattice structures using vitro test (Fig. 33b), SLM-produced Fe–35 Mn lattice structure shows
MG63 cells after incubation for 72 h [229]. The SLM-produced WE43 Mg good osteointegration after 4-week implantation in a rat cranial bone,
alloy and Zn lattice structures respectively show < 25 % and < 15 % where bone apposition is observed on the lattice structure (Fig. 33c)
non-viable cells, while the SLM-produced pure Fe lattice structure shows [468]. Ray et al. [620] also reported that the Fe lattice structures with a
40 % non-viable cells. In comparison, the SLM-produced Ti–6Al–4 V FeSr coating demonstrate a significant increase in bone formation
exhibits better biocompatibility. Fig. 32b represents the live/dead compared to a FeBiP coating and bare Fe lattice structures after 6-week
staining of rabbit bone marrow mesenchymal stem cells after 1 and 3 implantation in a rat femur (Fig. 33d). Such results suggest the need of
days of culture on Fe lattice structures produced by direct ink writing. applications of bioactive coatings on Fe-related biomedical materials.
Apparently, most of the rabbit bone marrow mesenchymal stem cells on
the Fe lattice structure are dead after 1 and 3 days of culture. Similar
6.2. Lightweight structures
results were reported in Ref. [62]. By contrast, the Fe lattice structure
coated with hydroxyapatite has a significant proliferation of rabbit bone
The “lightweight concept” is to reduce the weight of a structure to
marrow mesenchymal stem cells. Based on these results, it can be noted
reach a balance between its performance and reliability. Light weight is
that the biocompatibility of Fe is relatively low. Fig. 33 represents some
an important role in nature, which has a goal to achieve the lowest
results of in vivo tests of metallic lattice structures. Mg powder,
consumption with the highest efficiency [621]. Investigating light­
polylactic-co-glycolic acid (PLGA), β-tricalcium phosphate (β-TCP) are
weight structures involves multidisciplinary subjects, including mate­
employed to fabricate a novel PLGA/TCP/Mg lattice structure by
rials mechanics, materials science, mathematics, computational
low-temperature rapid prototyping (Fig. 33a) [618]. With the addition
technology, and preparation technique [622,623]. The lightweight
of Mg, the PLGA/TCP/Mg lattice structure shows 56.3 % higher mean
concept was first used in aircraft manufacturing and the
bone volume than the PLGA/TCP counterpart. The micro-CT 2D (the red
aerospace-related industry [624,625]. The lightweight strategies not
arrows indicate new bone cells) and 3D reconstruction models (Fig. 33b)
only provide sufficient strength and safety but also save the consump­
illustrate the bone formation on Mg lattice structures after the implan­
tion of materials and energy. Therefore, such targeted manners lead to a
tation for 16 weeks [619]. The Mg lattice structure with larger pores is
significant increase in high demands in the new technologies and
found to promote vascularization and results in more mature bone for­
knowledge. Subsequently, the “lightweight concept” is rapidly
mation in a rabbit model. Therefore, it is significant to consider the ef­
employed for the high-performance locomotives [626], ships [9,627],
fect of the geometry of lattice structure on bone formation. Although it is
aerospace industries [622,628], biomedical industries [629], wind
mentioned that most of the rabbit bone marrow mesenchymal stem cells
power plants [630], automobile body manufacturing [631–633], and
on the Fe lattice structure are dead after 1 and 3 days of culture in an in
machine tool manufacturing [634]. The component designed and

34
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

produced by the “lightweight concept” may have higher speeds, a higher

precision, and a longer service time. Meanwhile, the concept of light­
weight also has a great economic value, which decreases the cost of
production and saves energy consumption [635–637].
Conventionally, the “lightweight concept” involves three critical
criteria: lightweight materials, lightweight structure, and lightweight
manufacturing. The most frequently used lightweight materials include
Al alloys [148,638], Mg alloys [639,640], Ti alloys [623,641], and
high-strength steels [642,643]. Such materials have a lighter weight but
a higher specific strength than others, which are particularly desirable
for the application of lightweight materials. Under the preconditions of
the functional purposes and safety factors, the lightweight design is to
optimize each structural parameter and reduce their mass [72,493–495,
623]. Lots of optimization methods have been presented in Section
4.2.3. The optimization of lightweight structures includes size [73],
shape [644], topology [497] and bionics [645,646]. Such optimization
aspects mainly depend on the mathematical models, such as conven­
tional algorithm, finite element algorithm, or intelligent algorithm, and
could be theoretically achieved. New developed lightweight structures
with lattice infill are promising in the applications. Similar to the con­
ventional lightweight structure, the ones with lattice infill are also
needed to be optimized for lightweight applications. However, the
optimized structures may not be available for fabrication by conven­
tional methods because the optimization process may be significantly
complex or precise. Conventional methods for producing lightweight
structures include welding, cutting, and other technologies [623,647]. It
would be substantially hard when using conventional methods to pro­
duce a lightweight structure with high refinement requirements. In
contrast, AM technologies provide advanced alternative solutions for
fabricating optimized lightweight structures. Therefore, AM technolo­
gies can be categorized as one of the lightweight manufacturing
Fig. 29. Biodegradation and cytotoxicity of WE43 Mg lattice structure in
simulated body fluid with 5% fetal bovine serum: (a) volume change after 28-
methods. Fig. 34 shows primary energy consumption and energy in­
day immersion and (b) relative cellular activity after 72 h cultivation of MG- tensity for the AM-produced aircraft components and the conventionally
63cells (reproduced with permission from Ref. [602], Copyright manufactured aircraft components, which also highlights the
(2018), Elsevier). energy-saving potential [648]. For each component, the AM-produced
aircraft components have significantly lower primary energy used
compared to the conventionally manufactured aircraft components. The

Fig. 30. Mechanical behavior of Fe lattice structures immersed in revised simulated body fluid for 28 days: (a) compressive properties of as-built and as-polished
lattice structure (inset is the specimen at the start and a strain of 60 %), (b) compressive curves of lattice structures after 28-day immersion, (c) changes in elastic
modulus and (d) yield strength with immersion time. (reproduced with permission from Ref. [62], Copyright (2018), Elsevier).

35
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 9
Summary of unit cells, porosity, elastic modulus (E), ultimate strength (σ max), and yield strength (σy), and hardness (H) of additive manufacturing produced lattice
structures from primary biodegraded metals.
Material Fabrication method Porosity (%) Unit cell E (GPa) σy (MPa) σmax (MPa) H (HV) Ref.

Mg SLM Cubic – – – 48.3 [612]

Mg-Ca SLM 6~38 Cubic – – 112 C 60–68 [613]
WE43 Mg SLM 76 Lattice – – 15 C – [614]
WE43 Mg SLM 67 Diamond 0.8 23 C 27 C – [602]
Fe SLM 67 Cubic – 70 ± 4 C 135 ± 5 C – [615]
Fe SLM 85 Diamond 1 11 C – – [598]
Fe SLM 70 Diamond 1 29 C – – [598]
Fe SLM 70 Diamond 2 31 C – – [598]
Fe SLM 59 Diamond 3 54 C – – [598]
Fe SLM 73 Diamond 1.6 – 1.8 24 C 40 C – [62]

T T
Fe-Mn BJ 39 – – 228 ± 11 190 ± 26 39 ± 1 [616]
Fe+HA DIW – Cubic 1.25 – 141 C – [601]
Zn SLM 73 Diamond 0.4 4C – – [229]
Zn SLM 69 ± 2 Diamond 0.5 6C – – [229]
Zn SLM 62 ± 3 Diamond 0.8 11 C – – [229]

Superscript “C” means under compression and superscript “T” indicates in tension condition.

Table 10
Summary of the biodegradation behavior after 28-day immersion test for some additive manufacturing produced lattice structures from primary biodegraded metals.
Material Fabrication method Unit cell Porosity Testing medium Weight Corrosion rate Corrosion density Ref.
(%) loss (%) (mm/y) (μA/cm2)

WE43 SLM Diamond 67 Revised simulated 21 0.2 – [602]

Mg body fluid
Fe SLM Cubic 67 Simulated body 5 0.09 ± 0.02 – [615]
fluid
Fe 3D printing mold + Truncated octahedron 81 Simulated body 5 – 61 ± 2 [617]
pressureless microwave fluid
sintering
Fe 3D printing mold + Truncated octahedron 60 Simulated body 4 – 88 ± 2 [617]
pressureless microwave fluid
sintering
Fe 3D printing mold + Truncated octahedron 46 Simulated body 6 – 141 ± 2 [617]
pressureless microwave fluid
sintering
Fe 3D printing mold + Cubic 51 Simulated body 6 – 94 ± 3 [617]
pressureless microwave fluid
sintering
Fe SLM Diamond (0.2 mm strut 85 Revised simulated 17 0.2 – [598]
thickness) body fluid
Fe SLM Diamond (strut thickness 70 Revised simulated 10 0.14 – [598]
graded from 0.2 mm to 0.4 body fluid
mm)
Fe SLM Diamond (strut thickness 70 Revised simulated 9 0.17 – [598]
graded from 0.4 mm to 0.2 body fluid
mm)
Fe SLM Diamond (0.4 mm strut 59 Revised simulated 5 0.11 – [598]
thickness) body fluid
Zn SLM Diamond 73 Revised simulated 12 0.17 – [598]
body fluid
Zn SLM Diamond 69 Revised simulated 8 0.14 – [598]
body fluid
Zn SLM Diamond 62 Revised simulated 7 0.13 – [598]
body fluid
Zn SLM Diamond 73 Revised simulated 5 0.07 – [598]
body fluid
Zn SLM Diamond 69 Revised simulated 3 0.06 – [598]
body fluid
Zn SLM Diamond 62 Revised simulated 4 0.07 – [598]
body fluid
Zn SLM Diamond 62 Revised simulated – – 45 ± 2 [229]
body fluid

bracket has roughly 2 GJ savings, whereas the seat buckle has about 70 ratio of lightweight and mechanical efficiency. Lots of metallic light­
MJ savings. The energy savings from the AM-produced aircraft com­ weight structures with lattice infill are produced by various AM tech­
ponents are mainly attributed to the reductions in the used resource nologies during recent years [72,406,623,649–652]. The convenience
production energy in the AM process and the reduced weight associated and flexibility of AM technologies lead to the possibility of the pro­
with optimized geometries. duction of metallic lightweight structures with novel design concepts.
Lattice structures have received considerable attention for their high Fig. 35 shows some examples of metallic lightweight structures with

36
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 31. Corrosion fatigue properties of Fe lattice structures produced direct energy deposition: (a) S-N curve, (b) macro-scale appearance after fatigue tests, (c)-(e)
the samples after fatigue test until 5% strain in the air at 0.9 yield strength, (f)-(h) the samples after fatigue test until 5% strain in r-SBF at 0.9 yield strength, and (i)-
(k) the samples after 3 × 106 cycles of fatigue test in r-SBF at 0.65 yield strength (reproduced with permission from Ref. [477], Copyright (2019), Elsevier).

Fig. 32. The in vitro biocompatibility of addi­

tively manufactured metallic lattice structures:
(a) the indirect cytotoxicity tests of selective
laser melting produced Ti–6Al–4 V, WE43 Mg
alloy, pure Fe, and pure Zn lattice structures
using MG63 cells after incubation for 72 h
(reproduced with permission from Ref. [229],
Copyright (2020), Elsevier), and (b) the live/­
dead staining of rabbit bone marrow mesen­
chymal stem cells after 1 and 3 days of culture
on Fe lattice structures produced by direct ink
writing (reproduced with permission from
Ref. [601], Copyright (2018), ACS
publications).

lattice infill [72,649,650]. These examples carry out the lightweight illustrates an example of the tetrahedron lattice structure for a 3D
concept; the optimized lightweight structures are produced by light­ Michell beam, a close look at this lattice structure, and a printed struc­
weight manufacturing using lightweight materials. Fig. 35a-35c ture using 316 L stainless steel [649]. This lattice structure model is

37
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 33. The in vitro biocompatibility of metallic lattice structures: (a) a novel PLGA/TCP/Mg lattice structure by low-temperature rapid prototyping with 56.3 %
higher mean bone volume than the PLGA/TCP counterpart (reproduced with permission from Ref. [618], Copyright (2019), Elsevier), (b) the micro-CT 2D (the red
arrows indicate new bone cells) and 3D reconstruction models, illustrating the bone formation on Mg lattice structures after the implantation for 16 weeks
(reproduced with permission from Ref. [619], Copyright (2016), Springer), (c) SLM-produced Fe–35 Mn lattice structure shows good osteointegration after 4-week
implantation in a rat cranial bone (reproduced with permission from Ref. [468], Copyright (2020), Elsevier), and (d) the Fe lattice structure with a FeSr coating
shows a significant increase in the bone formation compared to a FeBiP coating and bare Fe lattice structure after 6-week implantation in a rat femur (reproduced
with permission from Ref. [620], Copyright (2018), Elsevier). (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article).

composed of 905 hexahedra and approximately 308,000 facets. Fig. 35d 6.3. Energy absorber
is an automotive control arm produced by binder jetting using 420
stainless steel [650] and Fig. 35e is a pillow bracket printed by SLM In the fields of vehicle, ship, construction, and aerospace industries,
using a Ti alloy [650]. Fig. 35f-h shows a satellite bracket filled with energy-absorbing structures continuously pursue fuel-saving, less gas
lattice fill produced by SLM. Such a satellite bracket has reductions in emission, improved structural integrity, and passenger safety in the past
dynamic response by 25 % and weight by 17 % compared to the coun­ because a considerable number of impact events, such as traffic acci­
terpart with the original design [72]. Although the optimization is the dents, natural collisions, and earthquakes, take place [29,67,655–658].
theoretical basis for obtaining high performance of a lightweight Energy absorbers are the systems that totally or partially convert kinetic
structure with lattice infill, the manufacturing is still difficult for con­ energy into other forms [659]. The energy conversions of energy ab­
ventional methods from the models to physical objects. As a class of sorbers are either reversible or irreversible. For example, the pressure
advanced preparation technologies, AM technologies can fabricate the energies incompressible fluids and elastic strain energy in solids are
objects based on the constructed models. Table 11 lists the specific reversible, while the plastic deformation energy is irreversible [659].
strength of some lattice structures produced by different AM technolo­ Energy absorbers may be produced by a variety of materials, such as
gies using various unit cells and materials. One can note that the unit metals and alloys [659,660], polymers [29,661], and concretes
cells and materials play an important role in the specific strength of [662–664]. In this section, only metallic energy absorbers and
lattice structures. The deep integration of optimization methods and AM energy-absorbing structures are considered. The applications of energy
technologies gives an effective solution to the advanced design and absorbers and energy-absorbing structures are diverse, such as crash­
manufacturing of novel lightweight structures with high performance worthiness of vehicles [665–667], crash barrier design [668–670], the
and multifunction. safety of nuclear reactors [671], collision damage to road bridges [672,

38
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 34. Primary energy consumption and energy intensity for the AM-produced aircraft components and the conventionally manufactured aircraft components
(reproduced with permission from Ref. [648], Copyright (2016), Elsevier). AM indicates additive manufacturing and CM is conventional manufacturing.

Fig. 35. Metallic lightweight structures with

lattice infill produced by various additive
manufacturing technologies: (a)-(c) an example
of the tetrahedron lattice structure for a 3D
Michell beam, a close look at up view of this
lattice structure, and a printed structure using
316 L stainless steel (reproduced with permis­
sion from Ref. [649], Copyright (2016), Elsev­
ier), (d) an automotive control arm produced by
binder jetting using 420 stainless steel (repro­
duced with permission from Ref. [650], Copy­
right (2016), Elsevier), (e) a pillow bracket
printed by SLM using Ti alloy (reproduced with
permission from Ref. [650], Copyright (2016),
Elsevier), and (f)-(h) a satellite bracket filled
with lattice fill produced of SLM (reproduced
with permission from Ref. [72], Copyright
(2020), Elsevier).

39
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Table 11
The specific strength of some lattice structures produced by different AM technologies using various unit cells and materials.
Unit cell Fabrication method Material Relative density (kg m− 3) Specific strength (kN m kg− 1) Ref.
3
Cubic EBM Ti-6Al-4V 0.88 × 10 32.39− 37.63 [653]
Cubic SLM Co-28Cr-6Mo 1.54 × 103 176.32 ± 13.57 [625]
Cubic SLM Fe 2.59 × 103 52.12 ± 1.93 [642]
Diamond SLM WE43 Mg 0.59 × 103 45.76 [630]
Diamond EBM Ti-6Al-4V 0.57− 1.76 × 103 28.60− 67.50 [557]
Rhombic dodecahedron EBM Ti-6Al-4V 0.71− 1.64 × 103 28.17− 54.88 [621]
Rhombic dodecahedron EBM Ti-6Al-4V 0.53− 1.85 × 103 18.87− 54.05 [469]
Rhombic dodecahedron EBM Ti-24Nb-4Zr-8Sn 1.34 × 103 33.58 ± 0.82 [48]
Rhombic dodecahedron SLM Ti-24Nb-4Zr-8Sn 1.34 × 103 37.31 ± 0.67 [48]
G7 EBM Ti-6Al-4V 0.53− 1.85 × 103 15.09− 43.24 [623]
G7 EBM Ti-24Nb-4Zr-8Sn 1.61 × 103 21.74 ± 1.24 [624]
Honeycomb EBM Ti-6Al-4V 1.50 × 103 77.33 ± 6.67 [654]
Hatched structure EBM Ti-6Al-4V 1.76 × 103 84.32 ± 1.99 [653]
Gyroid EBM Ti-6Al-4V 0.79 × 103 30.89 ± 3.04 [421]
Face-centered cubic SLM Co-28Cr-6Mo 1.94 × 103 258.60 ± 6.49 [625]
Body-centered cubic SLM Co-28Cr-6Mo 1.86 × 103 244.03 ± 15.26 [625]
Spherical hollow cubic SLM Co-28Cr-6Mo 1.86 × 103 169.62 ± 9.44 [625]
Lattice SLM WE43 Mg 0.43 × 103 34.88 [641]

Fig. 36. AlSi10Mg lattice structures produced by selective laser melting for energy absorption applications: (a) the configurations of unit cells, (b) produced L4-1, L4-
2, L4-3, and L4-4 lattice structures, (c) compression failure modes of L4-1, L4-2, L4-3, and L4-4 lattice structures, (d)-(g) the relationship between relative density (ρ*)
of (d) max energy absorption per unit, (e) specific strength, (f) specific energy absorption and (g) energy absorption efficiency. (reproduced with permission from
Ref. [692], Copyright (2021), Elsevier).

673], and offshore structures [674,675]. conventional ones, have been reported [431,455,687,688]. These
The structures (or shapes) of energy absorbers have a significant metallic lattice structures for energy absorbers applications are made of
influence on their performance. The most commonly employed struc­ stainless steel [455,688,689], Ti alloys [431,660,690], and Al alloys
tures (or shapes) of conventional energy absorbers are tubes [676,677], [439,687,690–693]. As shown in Fig. 36a, Zhang et al. [692] produced
frusta [678,679], multi-corner columns [680,681], sandwich plates AlSi10Mg alloy by SLM and designed a series of lattice structures with
[657,682], honeycomb cells [683,684] and struts [685,686]. Tubes, rotational symmetry of triple, quadruple, and hex, which are denoted as
frusta, and multi-corner columns are frequently used as structural ele­ L3, L4, and L6, respectively. Each type of lattice structure (named as 1, 2,
ments in the past. With the development of preparation technologies, a 3, and 4) has a distinguished relative density. Using L4-1, L4-2, L4-3, and
more selection of structures can be built, including sandwich plates, L4-4 as examples, the energy absorption characteristics of those samples
honeycomb cells, and struts. In recent years, energy absorbers with (Fig. 36b) showed that each type of lattice structure displays different
metallic lattice structures, which have better performances than the failure modes. Each sample reveals 45 ◦ diagonal shear bands because

40
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

the maximum shear stress is generated at the angle of 45 ◦ to the strength for all lattice structures. The energy per unit volume is not
compression direction. The defect in the sample would accelerate the enhanced by increasing the number of structural layers and therefore
formation of shear bands. L4-1 has a global failure. Owing to the exis­ their specific energy absorption decreases. Harris et al. [689] employed
tence of Z-struts, other lattice structures have a layer-by-layer failure SLM to manufacture metallic hybrid lattice structures with lattice-wall
mode. Their energy absorption characteristics are also enhanced by square honeycomb using 316 L stainless steel and the samples exhibit
increasing the relative densities of lattice structures (Fig. 36c and d). Lei high strength and great specific energy (energy absorption per unit
et al. [693] fabricated two different types of multi-layer AlSi10Mg alloy volume).
sandwiches with lattice structures by SLM. The lattice structures have It was reported that the lattice structures with TPMS would have
BCC and BCCZ unit cells. The results showed that all lattice structures higher mechanical properties compared to the lattice structures with
display a primary failure mode of layer-by-layer crushing and shear struts [455,660]. Zhao et al. [660] compared the mechanical properties
deformation. The crash load efficiencies of lattice structures increase as of BCC lattice structures with and without TPMS produced by SLM using
increasing the layer number. However, during the energy absorption Ti–6Al–4 V. As shown in Fig. 37a, both samples have different unit cells
process, the plateau stresses are lower than their initial crushing and are successfully produced by SLM. Apparently, the BCC lattice
structures with TPMS have higher energy absorption and cumulative
energy absorption than the corresponding BCC lattice structures without
TPMS (Fig. 37b and 37c). Especially, the BCC TPMS lattice structures
with 30 % volume fraction (i.e., 70 % porosity) have about 3.5 times
cumulative energy absorption than the counterparts without TPMS.
Such an enhancement is attributed to the struts of BCC lattice structures
with TPMS are deformed by axial force and have a brittle fracture,
whereas struts of BCC lattice structures without TPMS are influenced by
bending moment and have cracks on nodes. The energy absorptions of
strut-based lattice structures are also different. Liu et al. [446] investi­
gated the energy absorptions of Ti–24Nb–4Zr–8Sn lattice structures with
cubic, optimized, and rhombic dodecahedron unit cells. The
Ti–24Nb–4Zr–8Sn lattice structures with rhombic dodecahedron unit
cells have a lower strength and energy absorption compared to the other
two lattice structures (Fig. 38a, b, and c). According to the analysis of
finite element modeling, the local stress is higher at the node of the
lattice structure with a rhombic dodecahedron unit cell (Fig. 38d).
Therefore, the different energy absorption values for these three lattice
structures are attributed to different stress distribution and local stress
concentration in the lattice structures.
In the past, lots of models have been designed but many of them
could not be successfully fabricated due to the limitation of capability
offered by conventional preparation methods. Up to date, the manu­
facturability of these designs is realized by AM technologies. Among
existing lattice structures, the auxetic structure is one of the strut-based
lattice structures, which has unique mechanical properties, resulted
from a negative Poisson’s ratio [116,407–410]. A negative Poisson’s
ratio would result in higher indentation resistance, higher shear resis­
tance, and higher fracture toughness [694–696]. Therefore, the auxetic
structure has a significantly high energy absorption ability under
roading and therefore such a type of structure is frequently used as en­
ergy absorbers [428]. The general topography of the auxetic structure
has been mentioned in Section 4.1.1. Previously, metallic auxetic
structures are produced by volumetric compression of conventional
metal foams [49]. Unfortunately, such a method leads to very limited
control over the resultant mechanical properties, including Poisson’s
ratio. In recent years, well-defined auxetic structures of arbitrary ge­
ometry can be fabricated by AM technologies in a highly controlled way.
Several examples have been reported. Schwerdtfeger et al. [409] used
EBM to fabricate Ti–6Al–4 V auxetic structures and investigated their
auxetic properties by compressive test. The basic element of the fabri­
cated auxetic structures can be considered as an inverted tetrapod. Such
elements are used to fill a plane and then the planes are stacked by
turning each layer by 60 ◦ C. The resultant auxetic structure is aniso­
tropic. The produced auxetic structures have the Poisson’s ratio ranging
from -0.4 to -0.2. Zhang et al. [697] printed Al-12Si auxetic structures
Fig. 37. BCC lattice structures of Ti–6Al–4 V with and without TPMS and their with re-entrant honeycomb patterns by SLM and investigated their dy­
energy absorption characteristics: (a) unit cells, constructed models and pro­
namic tensile behavior. During the dynamic loading, the tensile defor­
duced samples, (b) energy absorption per unit, and (c) cumulative energy ab­
mation takes place from the loading end to the fixed end in the produced
sorption at different volume fractions. BT indicates the body-centered cubic
lattice structures with the triply periodic minimal surface while B means the auxetic structures. Yang et al. [698] also built Ti–6Al–4 V auxetic
body-centered cubic lattice structure without triply periodic minimal surface. structures with re-entrant honeycomb pattern by EBM and the samples
10, 20, 30 illustrate the volume fractions of lattice structures. (reproduced with have four configurations, namely, the length of vertical struts and the
permission from Ref. [660], Copyright (2021), MDPI). re-entrant struts, the strut cross section, and the re-entrant angle. The

41
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 38. The stress-strain curves for multiple cyclic load-unloading curves and finite element modeling results for three lattice structures with cubic, optimized, and
rhombic dodecahedron unit cells: (a) the stress-strain curves at a range of 1–6% strain levels, (b) the stress-strain curves at 1% strain level, (c) the stress-strain curves
at 2% strain level and (d) the finite element modeling stress distribution at 1% strain level. (reproduced with permission from Ref. [446], Copyright (2018), Elsevier).

Fig. 39. Gravimetric energy absorption obtained from (a) compression tests and (b) shear tests for the lattice structures with different densities and materials.
(reproduced with permission from Ref. [440], Copyright (2016), Elsevier).

mechanical tests indicated that the elastic moduli of Ti–6Al–4 V auxetic 6.4. Catalysts
structures decrease with increasing the length of re-entrant struts. The
Poisson’s ratios are different in various directions, which range from AM technologies also extend the applications of catalysts because
− 0.57 to − 0.09. Ullah et al. [440] compared the energy absorption of considerable attention has been paid to the relationships between
lattice structures with different densities and materials, as shown in catalyst structure and catalytic performance [55,705,706]. Current
Fig. 39. The data were extracted from Refs. [699–704]. It was found that active heterogeneous catalysts mostly focus on particles in nano- and
SLM-produced Kagome structures have relatively lower density and micro-scale due to their size reduction with enlarged specific surface
higher energy absorption compared with the carbon fiber reinforced area, thus providing more catalytic active sites [707–709]. However, the
plastics (CFRP) pyramid structures, foams, and other metal-based aggregation of nanoparticles usually results in catalytic deactivation
structures. Some hybrid CFRP and steel honeycombs have higher en­ with unstable performance, and the supporting materials are usually
ergy absorption while they are nearly two times heavier. To balance the needed for particles in the specific application. The application of thin
energy absorption and weight is also a significant consideration for rods/wires and ribbons may provide alternative solutions [710,711],
design a lightweight structure. but the lower surface-to-volume ratio in these catalysts is unavoidable.
Thus, further post-treatment such as dealloying to produce porous

42
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

structure is needed to improve the catalytic performance [712,713]. In surface area. In such a situation, directly printing metal/metal oxide
contrast, the self-supported bulk catalysts with 3D lattice structures structures could be a good choice. As shown in Fig. 40, Shannon et al.
directly manufactured by AM, hold the promise to provide much more [342] employed DIW to fabricate 3D structures in a cylindrical shape
reliable catalysts in industrialization (wastewater treatment, water with 10 mm in diameter and 25 mm in height using inks of Fe2O3, Fe,
splitting, fuel cells, supercapacitors, etc.). NiO, and Ni, respectively. The metal inks are prepared by dissolving
The catalysts with lattice structures produced by AM technologies metal particles in a PLGA copolymer. After printing, the cylindrical
show lots of advantages. For example, three-dimensional catalysts with metal structures are heated at 300 ◦ C for 1 h to remove PLGA and sin­
lattice structures have a larger surface area and open pores, which tered at 900 ◦ C for 2 h for Fe structure and 1300 ◦ C for 4 h for Ni
provide the possibility to obtain more active sites on the catalyst surface. structure. Compared to metallic structure, a reduction stage is added to
The transport of reactants is improved and the catalytic performances of oxide structure, which is used to reduce the oxide to metal. After sin­
catalysts with lattice structures are significantly enhanced [39–41]. tering, both metallic and oxide structures have underwent changes in
Meanwhile, fine lattice structures of catalyst frames prepared by AM mass and dimension (Fig. 40a). By comparison, metallic structures have
technologies are cost-effective with the high surface area to mass ratio fewer changes in both mass and dimension because fewer oxygen ions
because the manufacturing prices for both equipment and feedstock are reduced and lost.
have been largely reduced in recent years. Owing to the flexibility of AM Due to the uniquely disordered atomic packing structure with
technologies, the structures can be customized and precisely controlled metastable nature leading to a high concentration of coordinately un­
to accommodate the application requirements. AM technologies can also saturated sites, metallic glasses (also known as amorphous alloys) have
provide the solution of one-time synthesis of the metallic lattice struc­ received extensive attention in catalytic applications [716–719]. As a
tures in a relatively simple way, which avoids the complex processes in result, in recent years, increasing researchers have focused on the
the conventional methods [76]. combination of catalytically active metallic glasses and AM technologies
Three types of AM technologies, including PBF (such as SLM, EBM, to achieve novel bulk catalyst design [718,720,721]. In 2018, Yang et al.
and SLS), FDM, and DIW, have been commonly employed for preparing [41,76] applied SLM to fabricate a lattice structure of Zr55Cu30Ni5Al10
metallic lattice structures as catalysts [287,714]. The metallic lattice metallic glass (Fig. 41), which was further dealloyed in a solution con­
structures may not have the catalytic capability; however, the catalysts taining HF and H2SO4 to prepare the nanoporous structure of Cu. In this
can be synthesized by the deposition of the catalytic active component way, the synthesized catalysts showed a 660-fold larger specific surface
on the produced structures. Avril et al. [715] used SLM to print area and a 14 times higher reaction rate constant in pollutant degra­
helical-shaped 3D stainless steel electrodes. The catalytic active dation than traditional ribbon and powder catalysts. In 2020, Liang et al.
component of IrO2 is prepared by immersing the 3D stainless steel [55] also used SLM to print a lattice structure (with rhombic dodeca­
electrode in a IrCl4 solution added oxalic acid and hydrogen peroxide. hedron microstructure) using Fe70Cr5Ni3Mo3W9Si5B5 metallic glass
Afterward, potential scans between -0.6 V and +0.6 V (vs Ag/AgCl) at a composite powders, which has in-situ α-Fe precipitates dispersed in the
scan rate of 0.05 V/s are conducted. As such, a functional IrO2 film is amorphous matrix (Fig. 42). In comparison with the application in the
synthesized. Such an electrode has good catalytic properties for the Fenton-like process, the SLM-produced porous Fe-based metallic glass
oxygen evolution reaction. Avril et al. [39] also printed a 3D static mixer matrix composite shows ultrahigh reusability (at least 45 times) in sul­
and deposited catalyst using electroplating or cold spraying, aiming to fate radical-based reaction without apparent decay of catalytic activity
generate a catalytic reactor system. As reported, the manufacturing cost during the degradation of cibacron brilliant red 3B-A dye (Fig. 42b-e).
of this system is significantly lower than that of other processes. More importantly, such AM-produced lattice-structural catalysts have a
Although such a process of printing and deposition is useful for pro­ higher reaction rate constant and lower activation energy than con­
ducing metallic lattice structures with active components, it is not a ventional catalysts, leading to a promising overall catalytic ability for
facile method to synthesize the catalysts using structures with low industrialization. As a bottom-up method, AM technology opens a new

Fig. 40. Metallic and oxide lattice structures produced by direct ink writing: (a) changes in mass and dimensions after post-treatment, (b) and (f) Fe2O3 lattice
structures, (c) and (g) Fe lattice structures, (d) and (h) NiO lattice structures, (e) and (i) Ni lattice structures. All scale bars are 2 mm (reproduced with permission
from Ref. [342], Copyright (2016), Elsevier).

43
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 41. A lattice structure of Zr55Cu30Ni5Al10 metallic glass produced by selective laser melting: (a) photographs of the lattice structure before and after dealloying
in a HF and H2SO4 mixture solution for 60 h, SEM images of (b) the as-fabricated lattice structure, (c) the nanoporous surface on the lattice structure after dealloying,
and (d) cross-sectional morphology of the nanoporous surface. (reproduced with permission from Ref. [342], Copyright (2016), Elsevier).

door to prepare high-efficient lattice-structural catalysts in a facile and voids in the lattice structures also enable the flow of fluid through the
economical way. material and therefore the heat exchangers with lattice structures may
have enhanced effective thermal conductivity [733–735]. The overall
heat transfer is determined by the porosity and surface area density of
6.5. Acoustic insulator and heat exchanger the lattice structure, while the orientation of the lattice structure has
little influence on heat transfer. In order to improve the performance of
Lattice structures are also able to be applied as acoustic insulators heat exchangers, AM technologies provide the possibility for lattice
[722–724]. Topological acoustics indicate that the materials with lattice structure manufacturing to promote the porosity and surface area den­
structures exhibit better acoustic insulation properties [57,722]. How­ sity [77,731,736–738]. Ho et al. [739] employed SLM to manufacture a
ever, due to the limitation of conventional preparation methods, the novel heat exchanger with a rhombi-octet lattice structure (Fig. 44a).
acoustic properties of lattice structures are mainly theoretically inves­ The heat exchanger has a 45 % higher thermal conductance compared
tigated [723,725,726]. In comparison, AM technologies provide a better with a conventional fin-tube heat exchanger. Wong et al. [740] used
solution to investigate acoustic insulators with lattice structures. As SLM to fabricate five different types of heat exchangers with fins of
shown in Fig. 43a, Sun et al. [727] fabricated lattice structures with circular, rectangular, rectangular round, elliptical shapes, and lattice
Dode-medium unit cells by SLM using 316 L stainless steel. Two samples structures. Their performances were investigated at fixed mass flow rate
with different dimensional parameters of unit cells were prepared, conditions. When the mass flow rate is 4.3 × 103 kg/s, the heat
where one of them has a larger lattice length of 2 times than the other. At exchanger with a lattice structure only can dissipate 54 % heat of that is
the same time, a bulk sample with the identical size and material was dissipated by the heat exchanger with rectangular fins, although the
prepared under the same parameters, which was used as a reference. The heat exchanger with lattice structure has the largest surface area to
acoustic properties, including sound insulation and sound absorption of volume ratio. The large temperature gradient in thin struts reduces fin
these three samples, were investigated. Both samples with lattice efficiency and the orientation of lattice channels minimizes the flow
structures have better sound absorption than the bulk one in the fre­ interactions with fibers (Fig. 44b). Wong and Leong [429] investigated
quency range of 1–6.3 kHz (Fig. 43b). The lattice structure with a the heat exchangers with octet-truss lattice structures, which have
small-sized unit cell shows better sound absorption than the counterpart different unit cell lengths and the same porosity. The heat exchangers
with a large-sized unit cell. This work demonstrates that lattice struc­ produced by SLM using AlSi10Mg alloy are shown in Fig. 44c. The re­
tures are promising in the applications of acoustic insulators. Indeed, sults showed that the surface area to volume ratios are not influenced by
acoustic insulators are commonly made by polymers because polymers the unit cell length of lattice structures. Even so, the AM-produced heat
have better acoustic adsorption than metals [728,729]. Therefore, there exchangers with lattice structures still have some limitations, including
are still limited investigations with respect to the acoustic properties of surface roughness, resolution, high equipment costs, post-processing
metallic lattice structures [727,730]. difficulties in complex shapes [741]. Moreover, the heat exchanger
Similar to acoustic insulators, heat exchangers, which are functional with multi-scale fluid flow and heat transfer is difficult in the designing
structures, have also been fabricated by AM technologies using lattice and fabrication.
structures, owing to their high surface-volume ratios, high heat transfer
capability, and high specific strength [77,731,732]. Interconnected

44
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 42. Lattice structures of Fe70Cr5Ni3Mo3W9Si5B5 metallic

glass composites with rhombic dodecahedron unit cell pro­
duced by selective laser melting: (a) lattice structures produced
by different laser scan rates, the reusability of produced lattice
structure (at a laser scan rate of 2000 mm/s) (b) in Fenton-like
process (H2O2) and (c) in sulfate radical-based reaction (per­
sulfate) for degrading BR3B-A dye, (d) and (e) are corre­
sponding dye removal rates in H2O2 and sulfate radical-based
reaction, respectively, at 5 and 10 min. (reproduced with
permission from Ref. [55], Copyright (2020), Elsevier).

7. Limitations and challenges Ti–24Nb–4Zr–8Sn lattice structures with rhombic dodecahe­

dron unit cells. Fig. 45a and c are the SEM images of the defects
Although metallic lattice structures have demonstrated excellent on the surfaces of lattice structures. By the analysis of electron
functional flexibilities, metallic lattice structures are difficult to be microprobe, it is found that Sn is prone to be concentrated in the
manufactured by conventional manufacturing due to their complex defects (Fig. 45b and d). Sn is evaporated during the AM process.
geometries in nature. Undoubtedly, AM technologies open up a new era As a result, defects are produced and the surface roughness
in the design and fabrication of complex metallic lattice structures. increases.
However, AM technologies are not omnipotent and also have some (2) For some AM technologies, such as DED and FDM, the accuracy of
limitations and challenges for producing metallic lattice structures for a produced metallic lattice structures should be improved [743].
variety of industries. Based on these limitations and challenges, some The geometries of produced metallic lattice structures strongly
considerations in AM technologies and lattice structures should be depend on each built layer. If the geometry and position of one
noted. layer deviate from the designed model, the subsequent layers also
have deviations [744]. In such a situation, the produced lattice
(1) The surface finish of parts produced by AM technologies is not structures may not have good symmetry.
satisfactory. Post-treatments after the manufacturing are often (3) Some AM technologies, such as SLM and EBM, should work in a
required. Especially for powder technologies, the unmelted or controlled atmosphere. Therefore, the sizes of produced metallic
semi-melted powder may be attached to the produced metallic lattice structures are significantly limited. Large components,
lattice structures. Hence, the surface roughness of the produced which are used in heavy industries, cannot be produced by SLM
metallic lattice structures increases [742]. To some extent, the and EBM.
attached powder may become the nucleation sites for cracking (4) Some AM technologies, such as DIW and BJ, produce the parts
when the produced parts are deformed. On the other hand, the using bonding matters (such as PLGA or other adhesives) mixed
evaporation of elements with low melting points (e.g., Sn) also with metallic particles. Therefore, sintering or debinding is
would generate defects on the surfaces of lattice structures [47]. commonly employed after manufacturing. However, when the
Fig. 45 shows the surface states of EBM- and SLM-produced binder is removed, the obtained metallic lattice structures would

45
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

compared with the ones produced by other AM technologies

[207].
(5) Feedstocks are still limited for AM technologies. Therefore, it is
difficult to produce metallic lattice structures using new mate­
rials. For example, the metallic powder should be atomized from
an ingot with identical composition, which increases the cost of
the products. Similarly, lattice structures produced by composites
may have better mechanical properties. However, the feedstocks
produced by composites are rarely reported. The AM-produced
parts by mixed powder may have a low homogeneity in the
microstructure therefore properties [745].
(6) The design and manufacturing of metallic lattice structures
involve a lot of subjects, such as materials science, mechanics,
topology optimization, finite element analysis, heat transfer
theory, and so on. Therefore, producing a metallic lattice struc­
ture with good performance may require cooperations. A
comprehensive and reliable database containing detailed infor­
mation on the needed properties is still yet to be set up. Such
information can be used as the “target” of design and topology
optimization.
(7) Topology optimization is a complicated process. Metallic lattice
structures may fail to be manufactured by AM technologies if the
designed model exceeds the processability of the equipment.
Furthermore, the equipment generally has limited resolution.
Therefore, a significantly fine metallic lattice structure may be
designed but not successfully fabricated. Meanwhile, the critical
Fig. 43. 316 L stainless steel acoustic insulators produced by selective laser
angle of overhanging structure and the difficulty in removing the
melting: (a) two sandwich samples with different lattice sizes and a bulk one
and (b) sound transmission loss of three samples. (reproduced with permission
supporting structure are still the constraints and limitations of
from Ref. [727], Copyright (2020), MDPI). current AM technologies. Hence, new topology optimization al­
gorithms should be developed to meet the demand of current AM
technologies.
have a relatively lower relative density compared with those
(8) Although AM technologies can produce a variety of metallic lat­
manufactured by other AM technologies. As such, the mechanical
tice structures, some types are still not manufacturable. For
properties of metallic lattice structures produced by such AM
example, pentamode metamaterials are a new class of 3D me­
technologies may not be satisfied. Meanwhile, the obtained
chanical metamaterials, which has an unusual elastic property
metallic lattice structures may have higher surface roughnesses

Fig. 44. Heat exchangers with lattice structures: (a1)-(a4) a novel heat exchanger with rhombi-octet lattice structures produced by selective laser melting from
design, unit cell to real products (reproduced with permission from Ref. [739], Copyright (2020), Elsevier), (b1)-(b6) five different types of heat exchangers with fins
of circular, rectangular, elliptical shapes and lattice structures produced by selective laser melting (reproduced with permission from Ref. [740], Copyright (2009),
Elsevier) and (c1)-(c4) the octet-truss lattice-structural heat exchangers with different unit cell lengths and same porosity (reproduced with permission from
Ref. [429], Copyright (2018), Elsevier).

46
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 45. The surface states of Ti–24Nb–4Zr–8Sn lattice structures with rhombic dodecahedron unit cells produced by electron beam melting and selective laser
melting: (a) and (c) SEM images of the defects, (b) and (d) the analysis of electron microprobe for elemental maps of Sn near a defect (reproduced with permission
from Ref. [47], Copyright (2016), Elsevier).

with vanishing shear modulus [746,747]. A pentamode meta­ (10) There is a lack of software for shell lattice structures and
material is prone to be deformed in five independent modes advanced functionally graded lattice structures. These two types
because it only has a non-zero eigenvalue from its sixth-order of lattice structures are demonstrated to be useful in many in­
elasticity matrix. As shown in Fig. 46a and b, the lattice of pen­ dustries. Therefore, new software is urgently needed for these
tamode metamaterials consists of double-cone bars and the two lattice structures.
connection of lattice is point-to-point. The production of pen­ (11) Surface modification is widely used for metallic components to
tamode metamaterials has a significantly high requirement, the further improve their performance. However, it is challenging in
physical products always have deviations from the ideal model the surface modification of the metallic lattice structures.
(Fig. 46c and d). Such a result stems from AM technologies. The
AM technologies have limited precision during production due to 8. Summary and outlook
their spot sizes and/or rough surface finish. The connection of
lattice in pentamode metamaterials cannot become a real point in During the past 30 years, the rapid development of AM technologies
the real products. In addition, the high residual stresses in the brings a significant revolution in material processing. The layer-wise
processing resulting from the high temperature gradient and fast production method with a computer-aided model enables the additive
cooling together with the high surface roughnesses would have a manufacturing technologies to fabricate the complex-shaped parts and
significant influence on the ideal appearance of the produced components which are not easily produced by the conventional pro­
pentamode metamaterials. All these factors mostly lead to the cessing methods. Meanwhile, additive manufacturing technologies
breaking of nodes in pentamode metamaterials therefore unsuc­ greatly save the costs of labor, materials, and time. Therefore, such
cessful manufacturing of pentamode metamaterials. technologies have been employed in many important fields, and boost
(9) Although all types of lattice structures have a similar principle in the advances of architected cellular structure, namely, lattice structure.
the deformation behavior, different configurations of various This article first introduces the definition of lattice structure and
lattice structures also lead to distinct deformation process. So far, gives a comparison between lattice structure and conventional cellular
only a few of investigations on the deformation mechanism of structure. The lattice structure has regular geometrical arrangements of
strut-based lattice structures have been reported. There are still in periodic unit cells over an area and/or space. Such a feature endows the
a short of the systematical investigations on the deformation lattice structures with their high controllability in architectural char­
mechanisms of TPMS and shell lattice structures. acteristics and superior properties. The conventional processing

47
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Fig. 46. (a) a double-cone bar, (b) the lattice consisted of double-cone bars, (c) the designed model of a pentamode metamaterial, and (d) the additively-
manufactured pentamode metamaterial. The inset is the SEM image of a magnified view of produced pentamode metamaterial (reproduced with permission from
Ref. [746], Copyright (2021), Elsevier).

methods cannot build non-stochastic structures or need the further as­ lattice structures is discussed from the features of unit cells. As such, the
sembly to produce lattice structures with limited architectures. There­ structures can be divided into bending- and stretch-dominated struc­
fore, the conventional methods fail to manufacture the lattice structures tures. Bending-dominated structures have high energy absorption and
with fine and complex geometry. Subsequently, the commonly used stretch-dominated structures have high stiffness. Metallic lattice struc­
additive manufacturing technologies, such as powder bed fusion, direct tures have been successfully manufactured and applied in many indus­
energy deposition, fused deposition modeling, laminated object trial sectors. In this review, the applications of metallic lattice structures
manufacturing, direct ink writing, and binder jetting, for metallic lattice in biomedical implants, lightweight structures, heat exchangers, energy
structures are introduced from the principle and feedstock to the pro­ absorbers, and catalysts are addressed in detail. The performance of a
cessing. The flexibility and accuracy of additive manufacturing tech­ metallic lattice structure highly depends on the material, porosity, and
nologies provide significant convenience in manufacturing metallic architecture. As reviewed, metallic lattice structures have better per­
lattice structures. Therefore, a considerable number of works with formances than the counterparts produced by conventional methods
respect to the metallic lattice structure have been reported. with stochastic structures. Therefore, with the development of additive
In the meantime, the unit cell design and optimization are the manufacturing technologies, metallic lattice structures may replace
important preconditions for producing a metallic lattice structure with conventional cellular structures and become a mainstream direction for
good performance. Generally, three types of unit cells, namely, strut- industrial applications.
based unit cells, triply periodic minimal surfaces, and shell unit cells, In the current decade, more and more attention has been taken to the
are commonly used to construct lattice structures. The arrangements of combination of metallic lattice structures and additive manufacturing
unit cells in the area or space constitute the patterns of lattice structures. technologies. However, there are still some limitations and challenges in
The pattern of lattice structure can be generated by direct patterning or the design and fabrication of metallic lattice structures by additive
conformal patterning. The unit cells are considered as cubic elements manufacturing technologies. Additive manufacturing technologies are
and simply repeated in lattice structures in the former method and the not omnipotent. Further understandings are required to ensure the
latter method conform the unit cells in a designed space and retain their reliability of the products to be employed in various industrial sectors.
integrity. Although these two methods could simply design the patterns More investigations should focus on the modeling, the optimization, and
of lattice structures, topology optimization is still required in actual the relationships among the materials, processing parameters, archi­
applications. tectures, and the performances of metallic lattice structures.
As a bottom-up method, additive manufacturing technology provides
new solutions for preparing metallic lattice structures from modeling to
process economically and conveniently. The deformation behavior of

48
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

Declaration of Competing Interest [39] A. Avril, C.H. Hornung, A. Urban, D. Fraser, M. Horne, J.P. Veder, J. Tsanaktsidis,
T. Rodopoulos, C. Henry, D.R. Gunasegaram, React. Chem. Eng. 2 (2017)
180–188.
The authors declare that they have no known competing financial [40] J. Lefevere, M. Gysen, S. Mullens, V. Meynen, J. Van Noyen, Catal. Today 216
interests or personal relationships that could have appeared to influence (2013) 18–23.
the work reported in this paper. [41] C. Yang, C. Zhang, L. Liu, J. Mater. Chem. A 6 (2018) 20992–21002.
[42] W. Tao, M.C. Leu, Design of lattice structure for additive manufacturing, 2016.
International Symposium on Flexible Automation (ISFA), 2016, pp. 325–332.
Acknowledgements Cleveland, OH.
[43] Y. Tang, G. Dong, Q. Zhou, Y.F. Zhao, IEEE Trans. Autom. Sci. Eng. 15 (2018)
1546–1562.
The authors would like to acknowledge the financial support pro­ [44] R.M. Gorguluarslan, U.N. Gandhi, R. Mandapati, S.-K. Choi, Comput. Des. Appl.
vided by Jiangsu Province six talent peaks project (Grant No. XCL-117), 13 (2016) 50–62.
National Natural Science Foundation of China (No.52001030), Natural [45] P. Li, N.V. Nguyen, H. Hao, Mater. Des. 89 (2016) 636–641.
[46] A. Szczurek, A. Ortona, L. Ferrari, E. Rezaei, G. Medjahdi, V. Fierro, D. Bychanok,
Science Foundation of Jiangsu (Grant No. BK20201456), the financial P. Kuzhir, A. Celzard, Carbon 88 (2015) 70–85.
support from Alexander von Humboldt Foundation Research Fellowship [47] Y.J. Liu, S.J. Li, H.L. Wang, W.T. Hou, Y.L. Hao, R. Yang, T.B. Sercombe, L.
for postdoctoral researchers, Open Foundation of Guangxi Key Labora­ C. Zhang, Acta Mater. 113 (2016) 56–67.
[48] Y.J. Liu, H.L. Wang, S.J. Li, S.G. Wang, W.J. Wang, W.T. Hou, Y.L. Hao, R. Yang,
tory of Processing for Non-ferrous Metals and Featured Materials, L.C. Zhang, Acta Mater. 126 (2017) 58–66.
Guangxi University (Grant No. 2020GXYSOF01, 2021GXYSOF03), and [49] R. Lakes, Science 235 (1987) 1038.
the Australian Research Council through Discovery Project [50] Z. Rueger, C.S. Ha, R.S. Lakes, Physica Status Solidi B: Basic Res. 256 (2019),
1800512.
(DP110101653, DP130103592). The authors are grateful to Shujun Li,
[51] A. Großmann, J. Gosmann, C. Mittelstedt, Mater. Sci. Eng. A 766 (2019), 138356.
Yulin Hao, Tim Sercombe, Yu-Wei Cui, Bo-Yuan Qin, Yu-Hang Chu, Nan [52] L. Boccarusso, M. Durante, A. Langella, Composites Part B.: Eng. 146 (2018)
Li, and Hao Liu. 165–175.
[53] J. Chen, Q. Gao, A. Sanson, X. Jiang, Q. Huang, A. Carnera, C.G. Rodriguez,
L. Olivi, L. Wang, L. Hu, K. Lin, Y. Ren, Z. Lin, C. Wang, L. Gu, J. Deng, J.
References P. Attfield, X. Xing, Nat. Commun. 8 (2017) 14441.
[54] E. Boatti, N. Vasios, K. Bertoldi, Adv. Mater. 29 (2017), 1700360.
[1] K. Lu, Science 328 (2010) 319. [55] S.-X. Liang, X. Wang, W. Zhang, Y.-J. Liu, W. Wang, L.-C. Zhang, Appl. Mater.
[2] S. Babaee, B.H. Jahromi, A. Ajdari, H. Nayeb-Hashemi, A. Vaziri, Acta Mater. 60 Today 19 (2020), 100543.
(2012) 2873–2885. [56] C. Parra-Cabrera, C. Achille, S. Kuhn, R. Ameloot, Chem. Soc. Rev. 47 (2018)
[3] A.P. Roberts, E.J. Garboczi, Acta Mater. 49 (2001) 189–197. 209–230.
[4] J. Crocker, Mater. Technol. 23 (2008) 174–178. [57] M. Weiner, X. Ni, M. Li, A. Alù, A.B. Khanikaev, Sci. Adv. 6 (2020), eaay4166.
[5] F. Baino, M. Ferraris, Int. J. Appl. Ceram. Technol. 14 (2017) 507–520. [58] H. Xue, Y. Ge, H.-X. Sun, Q. Wang, D. Jia, Y.-J. Guan, S.-Q. Yuan, Y. Chong,
[6] Z. Jian-zhong, W. Jiu-gen, M. Jia-ju, J. Zhejiang Univ. Sci. A 6 (2005) 1095–1099. B. Zhang, Nat. Commun. 11 (2020) 2442.
[7] G. Dercz, I. Matuła, M. Zubko, A. Kazek-Kęsik, J. Maszybrocka, W. Simka, [59] X. Ni, M. Li, M. Weiner, A. Alù, A.B. Khanikaev, Nat. Commun. 11 (2020) 2108.
J. Dercz, P. Świec, I. Jendrzejewska, Mater. Charact. 142 (2018) 124–136. [60] X. Wang, S. Xu, S. Zhou, W. Xu, M. Leary, P. Choong, M. Qian, M. Brandt, Y.
[8] J. Banhart, JOM 52 (2000) 22–27. M. Xie, Biomaterials 83 (2016) 127–141.
[9] V. Crupi, G. Epasto, E. Guglielmino, Mar. Struct. 30 (2013) 74–96. [61] A.A. Zadpoor, J. Mater. Chem. B 7 (2019) 4088–4117.
[10] A. Ajdari, H. Nayeb-Hashemi, A. Vaziri, Int. J. Solids Struct. 48 (2011) 506–516. [62] Y. Li, H. Jahr, K. Lietaert, P. Pavanram, A. Yilmaz, L.I. Fockaert, M.A. Leeflang,
[11] K.C. Cheung, N. Gershenfeld, Science 341 (2013) 1219. B. Pouran, Y. Gonzalez-Garcia, H. Weinans, J.M.C. Mol, J. Zhou, A.A. Zadpoor,
[12] P. Heinl, L. Müller, C. Körner, R.F. Singer, F.A. Müller, Acta Biomater. 4 (2008) Acta Biomater. 77 (2018) 380–393.
1536–1544. [63] L.C. Zhang, L.Y. Chen, Adv. Eng. Mater. 21 (2019), 1801215.
[13] H.N.G. Wadley, N.A. Fleck, A.G. Evans, Compos. Sci. Technol. 63 (2003) [64] K. Wei, Y. Peng, K. Wang, S. Duan, X. Yang, W. Wen, Compos. Struct. 188 (2018)
2331–2343. 287–296.
[14] C.M. Cheah, C.K. Chua, C.W. Lee, C. Feng, K. Totong, Int. J. Adv. Manuf. Technol. [65] X. Li, Y.H. Tan, H.J. Willy, P. Wang, W. Lu, M. Cagirici, C.Y.A. Ong, T.S. Herng,
25 (2005) 308–320. J. Wei, J. Ding, Mater. Des. 178 (2019), 107881.
[15] K. Sharp, D. Mungalov, J. Brown, Procedia Mater. Sci. 4 (2014) 15–20. [66] C. Yang, Q.M. Li, Mech. Mater. 148 (2020), 103536.
[16] Z.M. Ariff, L.O. Afolabi, L.O. Salmazo, M.A. Rodriguez-Perez, J. Mater. Res. [67] F.N. Habib, P. Iovenitti, S.H. Masood, M. Nikzad, Mater. Des. 155 (2018) 86–98.
Technol. 9 (2020) 9929–9940. [68] M. Pelanconi, M. Barbato, S. Zavattoni, G.L. Vignoles, A. Ortona, Mater. Des. 163
[17] B. Arifvianto, J. Zhou, Materials 7 (2014) 3588–3622. (2019), 107539.
[18] N. Guo, M.C. Leu, Front. Mech. Eng. 8 (2013) 215–243. [69] T. Tancogne-Dejean, M. Diamantopoulou, M.B. Gorji, C. Bonatti, D. Mohr, Adv.
[19] L.C. Zhang, Y. Liu, S. Li, Y. Hao, Adv. Eng. Mater. 20 (2018), 1700842. Mater. 30 (2018), 1803334.
[20] L.C. Zhang, D. Klemm, J. Eckert, Y.L. Hao, T.B. Sercombe, Scr. Mater. 65 (2011) [70] C.R. Tubío, J. Azuaje, L. Escalante, A. Coelho, F. Guitián, E. Sotelo, A. Gil,
21–24. J. Catal. 334 (2016) 110–115.
[21] L. Yuan, S. Ding, C. Wen, Bioact. Mater. 4 (2019) 56–70. [71] M. Helou, S. Kara, Int. J. Comput. Integr. Manuf. 31 (2018) 243–261.
[22] A. Kumar, L. Collini, A. Daurel, J.-Y. Jeng, Addit. Manuf. 33 (2020), 101168. [72] J. Zhu, H. Zhou, C. Wang, L. Zhou, S. Yuan, W. Zhang, Chinese J. Aeronaut. 34
[23] S. Xu, J. Shen, S. Zhou, X. Huang, Y.M. Xie, Mater. Des. 93 (2016) 443–447. (2020) 91–110.
[24] A.J.J.N. Clarke, Nature 576 (2019) 41–42. [73] J.-H. Zhu, W.-H. Zhang, L. Xia, Arch. Comput. Methods Eng. 23 (2016) 595–622.
[25] D. Herzog, V. Seyda, E. Wycisk, C. Emmelmann, Acta Mater. 117 (2016) 371–392. [74] D. Mahmoud, M.A. Elbestawi, J. Manuf. Mater. Process. 1 (2017) 13.
[26] S. Vijayavenkataraman, L.Y. Kuan, W.F. Lu, Mater. Des. 191 (2020), 108602. [75] M.A. Gibson, N.M. Mykulowycz, J. Shim, R. Fontana, P. Schmitt, A. Roberts,
[27] K. Zhang, K. Wei, J. Chen, B. Liang, D. Fang, R. He, J. Eur. Ceram. Soc. 41 (2021) J. Ketkaew, L. Shao, W. Chen, P. Bordeenithikasem, J.S. Myerberg, R. Fulop, M.
2796–2806. D. Verminski, E.M. Sachs, Y.-M. Chiang, C.A. Schuh, A. John Hart, J. Schroers,
[28] B. Hussey, P. Nikaeen, M.D. Dixon, M. Akobi, A. Khattab, L. Cheng, Z. Wang, J. Li, Mater. Today 21 (2018) 697–702.
T. He, P. Zhang, Compos. Part B Eng. 183 (2020), 107700. [76] C. Yang, C. Zhang, W. Xing, L. Liu, Intermetallics 94 (2018) 22–28.
[29] A.K. Mishra, H. Chavan, A. Kumar, Mater. Today Proc. 6 (2021) 1–21. [77] S. Chekurov, J. Kajaste, K. Saari, H. Kauranne, M. Pietola, J. Partanen, Prog.
[30] P.K. Penumakala, J. Santo, A. Thomas, Compos. Part B Eng. 201 (2020), 108336. Addit. Manuf. 4 (2019) 55–61.
[31] H. Attar, M. Bönisch, M. Calin, L.C. Zhang, S. Scudino, J. Eckert, Acta Mater. 76 [78] R. Huang, S. Ma, M. Zhang, J. Xu, Z. Wang, Mater. Sci. Eng. A 756 (2019)
(2014) 13–22. 302–311.
[32] J.C. Najmon, S. Raeisi, A. Tovar, Review of additive manufacturing technologies [79] T. Luukkonen, J. Yliniemi, H. Sreenivasan, K. Ohenoja, M. Finnilä, G. Franchin,
and applications in the aerospace industry. Additive Manufacturing for the P. Colombo, Sci. Rep. 10 (2020) 7233.
Aerospace Industry, Elsevier, 2019, pp. 7–31. [80] J. Banhart, Prog. Mater. Sci. 46 (2001) 559–632.
[33] D. Böckin, A.-M. Tillman, J. Clean. Prod. 226 (2019) 977–987. [81] G. Kırmızı, H. Arık, H. Çinici, Compos. Part B Eng. 164 (2019) 345–357.
[34] H. Qian, T. Lei, P. Lei, Y. Hu, Tissue Eng. Part B Rev. 27 (2021) 166–180. [82] R. Kumar, H. Jain, S. Sriram, A. Chaudhary, A. Khare, V.A.N. Ch, D.P. Mondal,
[35] D. Delgado Camacho, P. Clayton, W.J. O’Brien, C. Seepersad, M. Juenger, Mater. Chem. Phys. 240 (2020), 122274.
R. Ferron, S. Salamone, Autom. Constr. 89 (2018) 110–119. [83] H. Wang, Y. Min, P. Li, J. Yang, J. Li, Electrochim. Acta 283 (2018) 54–62.
[36] O. Santoliquido, G. Bianchi, P. Dimopoulos Eggenschwiler, A. Ortona, Int. J. [84] Y. Liu, X. Cao, L. Cui, Y. Zhong, R. Zheng, D. Wei, C. Barrow, J.M. Razal, W. Yang,
Appl. Ceram. Technol. 14 (2017) 1164–1173. J. Liu, J. Power Sources 437 (2019), 226897.
[37] J. Liu, A.T. Gaynor, S. Chen, Z. Kang, K. Suresh, A. Takezawa, L. Li, J. Kato, [85] C. Li, X. Shi, S. Liang, X. Ma, M. Han, X. Wu, J. Zhou, Chem. Eng. J. 379 (2020),
J. Tang, C.C.L. Wang, L. Cheng, X. Liang, A.C. To, Struct. Multidiscip. Optim. 57 122248.
(2018) 2457–2483. [86] Q. Chen, J. Li, C. Liao, G. Hu, Y. Fu, O.K. Asare, S. Shi, Z. Liu, L. Zhou, L. Mai,
[38] J. Broughton, Y.K. Joshi, J. Heat Transfer 142 (2020), 082201. J. Mater. Chem. A 6 (2018) 19488–19494.
[87] P. Tan, B. Chen, H. Xu, W. Cai, W. He, M. Ni, Appl. Catal. B 241 (2019) 104–112.

49
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[88] G. Anbuchezhiyan, B. Mohan, D. Sathianarayanan, T. Muthuramalingam, [143] W. Xu, X. Lu, J. Tian, C. Huang, M. Chen, Y. Yan, L. Wang, X. Qu, C. Wen,
J. Alloys. Compd. 719 (2017) 125–132. J. Mater. Sci. Technol. 41 (2020) 191–198.
[89] K. Heim, F. García-Moreno, J. Banhart, Scr. Mater. 153 (2018) 54–58. [144] L.C. Zhang, J. Xu, E. Ma, Mater. Sci. Eng. A 434 (2006) 280–288.
[90] N. Wang, E. Maire, X. Chen, J. Adrien, Y. Li, Y. Amani, L. Hu, Y. Cheng, Mater. [145] P. Yu, L.C. Zhang, W.Y. Zhang, J. Das, K.B. Kim, J. Eckert, Mater. Sci. Eng. A 444
Charact. 147 (2019) 11–20. (2007) 206–213.
[91] H. Wu, G. Zhao, G. Wang, W. Zhang, Y. Li, Mater. Des. 144 (2018) 331–342. [146] L. Wang, B. Dong, F. Qiu, R. Geng, Q. Zou, H. Yang, Q. Li, Z. Xu, Q. Zhao,
[92] Á. Kmetty, K. Litauszki, D. Réti, Appl. Sci. 8 (2018) 1960. Q. Jiang, J. Mater. Res. Technol. 9 (2020) 641–653.
[93] G.Q. Tang, A.R. Kovscek, Transp. Porous Media 65 (2006) 287–307. [147] H.-Y. Yang, Z. Wang, L.-Y. Chen, S.-L. Shu, F. Qiu, L.-C. Zhang, Compos. Part B
[94] Y. Zeng, R. Farajzadeh, A.A. Eftekhari, S. Vincent-Bonnieu, A. Muthuswamy, W. Eng. 209 (2021), 108605.
R. Rossen, G.J. Hirasaki, S.L. Biswal, Langmuir 32 (2016) 6239–6245. [148] H.-Y. Yang, Y.-F. Yan, T.-S. Liu, B.-X. Dong, L.-Y. Chen, S.-L. Shu, F. Qiu, Q.-
[95] Z. Derikvand, M. Riazi, J. Mol. Liq. 224 (2016) 1311–1318. C. Jiang, L.-C. Zhang, Compos. Part B Eng. 225 (2021), 109265.
[96] T. Miyoshi, M. Itoh, S. Akiyama, A. Kitahara, Adv. Eng. Mater. 2 (2000) 179–183. [149] V. Kavimani, K.S. Prakash, T. Thankachan, Compos. Part B Eng. 162 (2019)
[97] A. Ibrahim, C. Körner, R.F. Singer, Adv. Eng. Mater. 10 (2008) 845–848. 508–521.
[98] M.D. Demetriou, G. Duan, C. Veazey, K. De Blauwe, W.L. Johnson, Scr. Mater. 57 [150] L. Zhao, G. Ma, P. Jin, Z. Yu, J. Alloys. Compd. 810 (2019), 151843.
(2007) 9–12. [151] X. Han, Z. Zhang, J. Hou, S.J. Thrush, G.C. Barber, Q. Zou, H. Yang, F. Qiu,
[99] G.J. Davies, S. Zhen, J. Mater. Sci. 18 (1983) 1899–1911. J. Mater. Res. Technol. 9 (2020) 12293–12307.
[100] B.H. Smith, S. Szyniszewski, J.F. Hajjar, B.W. Schafer, S.R. Arwade, J. Constr. [152] X. Han, Z. Zhang, Y. Rong, S.J. Thrush, G.C. Barber, H. Yang, F. Qiu, J. Mater. Res.
Steel Res. 71 (2012) 1–10. Technol. 9 (2020) 1357–1364.
[101] B. Zhao, A.K. Gain, W. Ding, L. Zhang, X. Li, Y. Fu, Int. J. Adv. Manuf. Technol. 95 [153] I. Moravcik, J. Cizek, Z. Kovacova, J. Nejezchlebova, M. Kitzmantel, E. Neubauer,
(2018) 2641–2659. I. Kubena, V. Hornik, I. Dlouhy, Mater. Sci. Eng. A 701 (2017) 370–380.
[102] S. Guddati, A.S.K. Kiran, M. Leavy, S. Ramakrishna, Int. J. Adv. Manuf. Technol. [154] X. Liang, Q. Wu, H. Li, R. Wang, L. Kang, B. Liu, L. Wang, J. Alloys. Compd. 862
105 (2019) 193–215. (2021), 158602.
[103] A. Yánez, A. Cuadrado, O. Martel, H. Afonso, D. Monopoli, Mater. Des. 140 [155] A. Civantos, M. Giner, P. Trueba, S. Lascano, M.-J. Montoya-García, C. Arévalo,
(2018) 21–29. M.Á. Vázquez, J.P. Allain, Y. Torres, Metals 10 (2020) 696.
[104] L.J. Gibson, M.F. Ashby, G.N. Karam, U. Wegst, H.R. Shercliff, Proc. R. Soc. Lond. [156] X. Yang, Q. Hu, J. Du, H. Song, T. Zou, J. Sha, C. He, N. Zhao, Int. J. Fatigue 121
A Math. Phys. Sci. 450 (1995) 141–162. (2019) 272–280.
[105] M.F. Ashby, Philosoph. Trans. Roy. Soc. A: Math. Phys. Eng. Sci. 364 (2006) [157] G. Jia, Y. Hou, C. Chen, J. Niu, H. Zhang, H. Huang, M. Xiong, G. Yuan, Mater.
15–30. Des. 140 (2018) 106–113.
[106] R.E. Raj, B.S.S. Daniel, J. Alloys. Compd. 467 (2009) 550–556. [158] Y. Shu, A. Suzuki, N. Takata, M. Kobashi, J. Mater. Process. Technol. 264 (2019)
[107] L. Wang, N. Limodin, A. El Bartali, J.-F. Witz, R. Seghir, J.-Y. Buffiere, 182–189.
E. Charkaluk, Mater. Sci. Eng. A 673 (2016) 362–372. [159] S. Mohammad Bagheri, M. Vajdi, F. Sadegh Moghanlou, M. Sakkaki,
[108] S. Singh, R. Singh, Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 230 (2016) M. Mohammadi, M. Shokouhimehr, M. Shahedi Asl, Ceram. Int. 46 (2020)
2143–2164. 7615–7624.
[109] S.O. Barnett, Foundry Trade J. Int. 11 (1988) 33. [160] M. Shahedi Asl, Z. Ahmadi, A. Sabahi Namini, A. Babapoor, A. Motallebzadeh,
[110] S. Jones, C. Yuan, J. Mater. Process. Technol. 135 (2003) 258–265. Ceram. Int. 45 (2019) 19808–19821.
[111] A.H. Brothers, D.C. Dunand, Adv. Eng. Mater. 8 (2006) 805–809. [161] Nd.F. Daudt, M. Bram, A.P.C. Barbosa, A.M. Laptev, C. Alves, J. Mater. Process.
[112] E.O.S. Montenegro, E.N.D. Grassi, J.B. Simões, P.C.S. da Silva, C.J. De Araújo, Technol. 239 (2017) 202–209.
Smart Mater. Struct. 29 (2020), 125008. [162] W. Niu, C. Bai, G. Qiu, Q. Wang, Mater. Sci. Eng. A 506 (2009) 148–151.
[113] S. Pattnaik, P.K. Jha, D.B. Karunakar, Proc. Inst. Mech. Eng. Part L J. Mater. Des. [163] S. Lascano, C. Arévalo, I. Montealegre-Melendez, S. Muñoz, J.A. Rodriguez-Ortiz,
Appl. 228 (2013) 249–277. P. Trueba, Y. Torres, Appl. Sci. 9 (2019) 982.
[114] A. Bandyopadhyay, B. Heer, Mater. Sci. Eng. R Rep. 129 (2018) 1–16. [164] W. Xu, J. Tian, Z. Liu, X. Lu, M.D. Hayat, Y. Yan, Z. Li, X. Qu, C. Wen, Mater. Sci.
[115] V.H. Carneiro, S.D. Rawson, H. Puga, J. Meireles, P.J. Withers, Addit. Manuf. 33 Eng. C 105 (2019), 110015.
(2020), 101085. [165] D.L. Zhang, C.C. Koch, R.O. Scattergood, Mater. Sci. Eng. A 516 (2009) 270–275.
[116] Y. Xue, X. Wang, W. Wang, X. Zhong, F. Han, Mater. Sci. Eng. A 722 (2018) [166] J. Wu, R. Guo, L. Xu, Z. Lu, Y. Cui, R. Yang, J. Mater. Sci. Technol. 33 (2017)
255–262. 172–178.
[117] K.-J. Kang, Prog. Mater. Sci. 69 (2015) 213–307. [167] A.C. Gonçalves, J. Mater. Process. Technol. 118 (2001) 193–198.
[118] C.R. Simovski, P.A. Belov, A.V. Atrashchenko, Y.S. Kivshar, Adv. Mater. 24 (2012) [168] A. Rodriguez-Contreras, M. Punset, J.A. Calero, F.J. Gil, E. Ruperez, J.M. Manero,
4229–4248. J. Mater. Sci. Technol. 76 (2021) 129–149.
[119] B. Khoda, A.M.M.N. Ahsan, A.N. Shovon, A.I. Alam, Sci. Rep. 11 (2021) 434. [169] C.E. Wen, M. Mabuchi, Y. Yamada, K. Shimojima, Y. Chino, T. Asahina, Scr.
[120] J. Fu, Y. Su, Y.-X. Qin, Y. Zheng, Y. Wang, D. Zhu, Biomaterials 230 (2020), Mater. 45 (2001) 1147–1153.
119641. [170] G.E. Ryan, A.S. Pandit, D.P. Apatsidis, Biomaterials 29 (2008) 3625–3635.
[121] D.J. Sypeck, H.N.G. Wadley, J. Mater. Res. 16 (2001) 890–897. [171] D.T. Queheillalt, H.N.G. Wadley, Mater. Des. 30 (2009) 1966–1975.
[122] T.J. Lu, L. Valdevit, A.G. Evans, Prog. Mater. Sci. 50 (2005) 789–815. [172] T.A. Schaedler, A.J. Jacobsen, A. Torrents, A.E. Sorensen, J. Lian, J.R. Greer,
[123] S. Li, J.-S. Yang, L.-Z. Wu, G.-C. Yu, L.-J. Feng, Mar. Struct. 63 (2019) 84–98. L. Valdevit, W.B. Carter, Science 334 (2011) 962.
[124] Y.-E. Lim, J.-H. Park, K. Park, Int. J. Precis. Eng. Manuf. Technol. 5 (2018) [173] X. Zheng, H. Lee, T.H. Weisgraber, M. Shusteff, J. DeOtte, E.B. Duoss, J.D. Kuntz,
499–506. M.M. Biener, Q. Ge, J.A. Jackson, S.O. Kucheyev, N.X. Fang, C.M. Spadaccini,
[125] C. Sennewald, S. Kaina, D. Weck, A. Gruhl, M. Thieme, G. Hoffmann, G. Stephani, Science 344 (2014) 1373.
R. Böhm, C. Cherif, O. Andersen, B. Kieback, W.A. Hufenbach, Adv. Eng. Mater. [174] S. Jiang, F. Sun, X. Zhang, H. Fan, Compos. Struct. 176 (2017) 55–71.
16 (2014) 1234–1242. [175] C.K. Chun-Wai, Digital Cellular Solids: Reconfigurable Composite Materials,
[126] Y.-H. Lee, B.-K. Lee, I. Jeon, K.-J. Kang, Acta Mater. 55 (2007) 6084–6094. Department of Architecture, vol Doctor, Massachusetts Institute of Technology,
[127] B.K. Lee, K.J. Kang, Adv. Eng. Mater. 10 (2008) 835–839. Massachusetts Institute of Technology, 2012, p. 147.
[128] W.D. MacDonald, T.W. Eagar, Annual Rev. Mater. Sci. 22 (1992) 23–46. [176] L. Wang, L. Xie, L.-C. Zhang, L. Chen, Z. Ding, Y. Lv, W. Zhang, W. Lu, D. Zhang,
[129] G.O. Cook, C.D. Sorensen, J. Mater. Sci. 46 (2011) 5305–5323. Acta Mater. 143 (2018) 214–226.
[130] M.G. Lee, J.W. Yoon, S.M. Han, Y.S. Suh, K.J. Kang, Procedia Mater. Sci. 4 (2014) [177] S. Liu, J. Liu, L. Wang, R.L.W. Ma, Y. Zhong, W. Lu, L.C. Zhang, Scr. Mater. 181
203–208. (2020) 121–126.
[131] H.J. Rathbun, F.W. Zok, S.A. Waltner, C. Mercer, A.G. Evans, D.T. Queheillalt, H. [178] H.N.G. Wadley, Philosoph. Trans. Roy. Soc. A Math. Phys. Eng. Sci. 364 (2006)
N.G. Wadley, Acta Mater. 54 (2006) 5509–5518. 31–68.
[132] J. Tian, T.J. Lu, H.P. Hodson, D.T. Queheillalt, H.N.G. Wadley, Int. J. Heat Mass [179] M.A. Ablat, A. Qattawi, Int. J. Adv. Manuf. Technol. 102 (2019) 615–633.
Transf. 50 (2007) 2521–2536. [180] H.N.G. Wadley, K.P. Dharmasena, M.R. O’Masta, J.J. Wetzel, Int. J. Impact Eng.
[133] H.B. Yan, T. Mew, M.G. Lee, K.J. Kang, T.J. Lu, F.W. Kienhöfer, T. Kim, J. Heat 62 (2013) 114–128.
Transfer 137 (2015), 022601. [181] W. Chen, Z. Liu, H.M. Robinson, J. Schroers, Acta Mater. 73 (2014) 259–274.
[134] B.-C. Choi, J.-W. Jeong, J.-H. Joo, K.-J. Kang, Adv. Eng. Mater. 11 (2009) [182] C. Ladd, J.-H. So, J. Muth, M.D. Dickey, Adv. Mater. 25 (2013) 5081–5085.
536–540. [183] S. Pattnaik, D.B. Karunakar, P.K. Jha, J. Mater. Process. Technol. 212 (2012)
[135] H.K. Miyamoto, R.D. Hanson, Structure 1 (2004) 16–18. 2332–2348.
[136] B. Khoda, A.M.M.N. Ahsan, 3D Print. Addit. Manuf. 8 (2021) 111–125. [184] L. Wang, C. Wang, L.-C. Zhang, L. Chen, W. Lu, D. Zhang, Sci. Rep. 6 (2016)
[137] S. Parvizi, S.M. Hashemi, F. Asgarinia, M. Nematollahi, M. Elahinia, Prog. Mater. 23905.
Sci. 117 (2021), 100739. [185] E.O. Olakanmi, R.F. Cochrane, K.W. Dalgarno, Prog. Mater. Sci. 74 (2015)
[138] C. Yang, M.D. Zhu, X. Luo, L.H. Liu, W.W. Zhang, Y. Long, Z.Y. Xiao, Z.Q. Fu, L. 401–477.
C. Zhang, E.J. Lavernia, Scr. Mater. 139 (2017) 96–99. [186] W.S.W. Harun, N.S. Manam, M.S.I.N. Kamariah, S. Sharif, A.H. Zulkifly, I. Ahmad,
[139] C. Yang, L.M. Kang, X.X. Li, W.W. Zhang, D.T. Zhang, Z.Q. Fu, Y.Y. Li, L.C. Zhang, H. Miura, Powder Technol. 331 (2018) 74–97.
E.J. Lavernia, Acta Mater. 132 (2017) 491–502. [187] M.J. Matthews, G. Guss, S.A. Khairallah, A.M. Rubenchik, P.J. Depond, W.E. King,
[140] Y. Li, C. Yang, H. Zhao, S. Qu, X. Li, Y. Li, Materials 7 (2014) 1709–1800. Acta Mater. 114 (2016) 33–42.
[141] L. Bolzoni, E.M. Ruiz-Navas, E. Gordo, Mater. Sci. Eng. A 687 (2017) 47–53. [188] S. Vock, B. Klöden, A. Kirchner, T. Weißgärber, B. Kieback, Prog. Addit. Manuf. 4
[142] Y. Alshammari, F. Yang, L. Bolzoni, J. Mech. Behav. Biomed. Mater. 91 (2019) (2019) 383–397.
391–397. [189] L. Dowling, J. Kennedy, S. O’Shaughnessy, D. Trimble, Mater. Des. 186 (2020),
108346.

50
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[190] C. Han, Q. Fang, Y. Shi, S.B. Tor, C.K. Chua, K. Zhou, Adv. Mater. 32 (2020), [239] M. Leary, M. Mazur, H. Williams, E. Yang, A. Alghamdi, B. Lozanovski, X. Zhang,
1903855. D. Shidid, L. Farahbod-Sternahl, G. Witt, I. Kelbassa, P. Choong, M. Qian,
[191] Q. Tan, J. Zhang, Q. Sun, Z. Fan, G. Li, Y. Yin, Y. Liu, M.-X. Zhang, Acta Mater. M. Brandt, Mater. Des. 157 (2018) 179–199.
196 (2020) 1–16. [240] M.M. Savalani, J.M. Pizarro, Rapid Prototyp. J. 22 (2016) 115–122.
[192] A. Plotkowski, K. Sisco, S. Bahl, A. Shyam, Y. Yang, L. Allard, P. Nandwana, A. [241] C. Yang, Y.J. Zhao, L.M. Kang, D.D. Li, W.W. Zhang, L.C. Zhang, Mater. Lett. 210
M. Rossy, R.R. Dehoff, Acta Mater. 196 (2020) 595–608. (2018) 169–172.
[193] Y. Zhao, C. Wu, S. Zhou, J. Yang, W. Li, L.-C. Zhang, J. Alloys. Compd. 863 [242] H. Gong, D. Snelling, K. Kardel, A. Carrano, JOM 71 (2019) 880–885.
(2021), 158721. [243] Y. Yang, J.B. Zhan, Z.Z. Sun, H.L. Wang, J.X. Lin, Y.J. Liu, L.C. Zhang, J. Alloys.
[194] N.E. Putra, M.A. Leeflang, M. Minneboo, P. Taheri, L.E. Fratila-Apachitei, J.M. Compd. 804 (2019) 220–229.
C. Mol, J. Zhou, A.A. Zadpoor, Acta Biomater. 121 (2021) 741–756. [244] H.Z. Lu, L.H. Liu, C. Yang, X. Luo, C.H. Song, Z. Wang, J. Wang, Y.D. Su, Y.
[195] D.K. Mishra, P.M. Pandey, Mater. Sci. Eng. A 804 (2021), 140759. F. Ding, L.C. Zhang, Y.Y. Li, J. Mater. Sci. Technol. 101 (2022) 205–216.
[196] Y. Li, W. Li, F.S.L. Bobbert, K. Lietaert, J.H. Dong, M.A. Leeflang, J. Zhou, A. [245] Y. Lu, S. Wu, Y. Gan, S. Zhang, S. Guo, J. Lin, J. Lin, J. Mech. Behav. Biomed.
A. Zadpoor, Acta Biomater. 106 (2020) 439–449. Mater. 55 (2016) 179–190.
[197] W.J. Choi, K.S. Hwang, H.J. Kwon, C. Lee, C.H. Kim, T.H. Kim, S.W. Heo, J.- [246] X.P. Li, C.W. Kang, H. Huang, L.C. Zhang, T.B. Sercombe, Mater. Sci. Eng. A 606
H. Kim, J.-Y. Lee, Mater. Sci. Eng. C 110 (2020), 110693. (2014) 370–379.
[198] G. Zhang, J. Guo, H. Chen, Y. Cao, J. Eur. Ceram. Soc. 41 (2021) 2790–2795. [247] H. Attar, K.G. Prashanth, L.-C. Zhang, M. Calin, I.V. Okulov, S. Scudino, C. Yang,
[199] L. Shen, H.B. Wu, F. Liu, J.L. Brosmer, G. Shen, X. Wang, J.I. Zink, Q. Xiao, M. Cai, J. Eckert, J. Mater. Sci. Technol. 31 (2015) 1001–1005.
G. Wang, Y. Lu, B. Dunn, Adv. Mater. 30 (2018), 1707476. [248] H. Attar, M. Bönisch, M. Calin, L.C. Zhang, K. Zhuravleva, A. Funk, S. Scudino,
[200] G. Zhang, H. Chen, S. Yang, Y. Guo, N. Li, H. Zhou, Y. Cao, J. Eur. Ceram. Soc. 38 C. Yang, J. Eckert, J. Mater. Res. 29 (2014) 1941–1950.
(2018) 4014–4019. [249] S. Zhou, M. Xie, C. Wu, Y. Yi, D. Chen, L.-C. Zhang, J. Mater. Sci. Technol. 104
[201] T. Wu, P. Jiang, X. Zhang, Y. Guo, Z. Ji, X. Jia, X. Wang, F. Zhou, W. Liu, Mater. (2022) 81–87.
Des. 180 (2019), 107947. [250] L.-C. Zhang, H. Attar, Adv. Eng. Mater. 18 (2016) 463–475.
[202] X. Peng, X. Kuang, D.J. Roach, Y. Wang, C.M. Hamel, C. Lu, H.J. Qi, Addit. Manuf. [251] X.J. Wang, L.C. Zhang, M.H. Fang, T.B. Sercombe, Mater. Sci. Eng. A 597 (2014)
40 (2021), 101911. 370–375.
[203] E. Peng, D. Zhang, J. Ding, Adv. Mater. 30 (2018), 1802404. [252] Y. Bai, X. Gai, S. Li, L.-C. Zhang, Y. Liu, Y. Hao, X. Zhang, R. Yang, Y. Gao, Corros.
[204] L. Tabard, V. Garnier, E. Prud’Homme, E.J. Courtial, S. Meille, J. Adrien, Sci. 123 (2017) 289–296.
Y. Jorand, L. Gremillard, Addit. Manuf. 38 (2021), 101776. [253] V.J. Challis, X. Xu, L.C. Zhang, A.P. Roberts, J.F. Grotowski, T.B. Sercombe,
[205] K.K.B. Hon, L. Li, I.M. Hutchings, CIRP Annals 57 (2008) 601–620. Mater. Des. 63 (2014) 783–788.
[206] J.A. Lewis, Adv. Funct. Mater. 16 (2006) 2193–2204. [254] L.-C. Zhang, H. Attar, M. Calin, J. Eckert, Mater. Technol. 31 (2016) 66–76.
[207] M. Ziaee, N.B. Crane, Addit. Manuf. 28 (2019) 781–801. [255] X. Zhou, C.-j. Liu, Adv. Funct. Mater. 27 (2017), 1701134.
[208] M. Li, W. Du, A. Elwany, Z. Pei, C. Ma, J. Manuf. Sci. Eng. 142 (2020), 090801. [256] J.P. Kruth, P. Mercelis, J. Van Vaerenbergh, L. Froyen, M. Rombouts, Rapid
[209] N. Chawake, L.D. Pinto, A.K. Srivastav, K. Akkiraju, B.S. Murty, R.S. Kottada, Scr. Prototyp. J. 11 (2005) 26–36.
Mater. 93 (2014) 52–55. [257] L.C. Zhang, T.B. Sercombe, Key Eng. Mater. 520 (2012) 226–233.
[210] U. Daalkhaijav, O.D. Yirmibesoglu, S. Walker, Y. Mengüç, Adv. Mater. Technol. 3 [258] Y. Yang, Z.G. Wu, B.Y. Shen, M.Z. Wu, Z.S. Yuan, C.Y. Wang, L.C. Zhang, J. Mater.
(2018), 1700351. Process. Technol. 296 (2021), 117177.
[211] M. Simonelli, N. Aboulkhair, M. Rasa, M. East, C. Tuck, R. Wildman, O. Salomons, [259] W. Fan, H. Tan, F. Zhang, Z. Feng, Y. Wang, L.-C. Zhang, X. Lin, W. Huang,
R. Hague, Addit. Manuf. 30 (2019), 100930. J. Mater. Sci. Technol. 94 (2021) 32–46.
[212] A. Kamaraj, S. Lewis, M. Sundaram, Procedia CIRP 42 (2016) 788–792. [260] S. Siddique, M. Imran, E. Wycisk, C. Emmelmann, F. Walther, J. Mater. Process.
[213] A.M. Brant, M.M. Sundaram, A.B. Kamaraj, J. Manuf. Sci. Eng. 137 (2015), Technol. 221 (2015) 205–213.
011018. [261] Y. Liu, Y. Yang, S. Mai, D. Wang, C. Song, Mater. Des. 87 (2015) 797–806.
[214] M.M. Sundaram, A.B. Kamaraj, V.S. Kumar, J. Manuf. Sci. Eng. 137 (2015), [262] M. Das, V.K. Balla, D. Basu, S. Bose, A. Bandyopadhyay, Scr. Mater. 63 (2010)
021006. 438–441.
[215] A. Brant, M. Sundaram, Procedia Manuf. 5 (2016) 928–943. [263] Y.J. Liu, X.P. Li, L.C. Zhang, T.B. Sercombe, Mater. Sci. Eng. A 642 (2015)
[216] A.B. Kamaraj, M. Sundaram, J. Appl. Electrochem. 48 (2018) 463–469. 268–278.
[217] G. Prashar, H. Vasudev, J. Clean. Prod. 310 (2021), 127606. [264] J. Gockel, L. Sheridan, B. Koerper, B. Whip, Int. J. Fatigue 124 (2019) 380–388.
[218] N. Hutasoit, R.A.R. Rashid, S. Palanisamy, A. Duguid, Int. J. Adv. Manuf. Technol. [265] D.D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Int. Mater. Rev. 57 (2012)
110 (2020) 2341–2357. 133–164.
[219] B. Marzbanrad, E. Toyserkani, H. Jahed, J. Mater. Process. Technol. 289 (2021), [266] J.-P. Kruth, J. Deckers, E. Yasa, R. Wauthlé, Proc. Inst. Mech. Eng. Part B J. Eng.
116928. Manuf. 226 (2012) 980–991.
[220] D. Vanerio, J. Kondas, M. Guagliano, S. Bagherifard, J. Manuf. Process. 67 (2021) [267] R. Engeli, T. Etter, S. Hövel, K. Wegener, J. Mater. Process. Technol. 229 (2016)
521–534. 484–491.
[221] H. Rao, R.P. Oleksak, K. Favara, A. Harooni, B. Dutta, D. Maurice, Addit. Manuf. [268] I. Yadroitsev, I. Yadroitsava, Virtual Phys. Prototyp. 10 (2015) 67–76.
31 (2020), 100932. [269] C. Qiu, N.J.E. Adkins, M.M. Attallah, Mater. Sci. Eng. A 578 (2013) 230–239.
[222] H. Klippstein, A. Diaz De Cerio Sanchez, H. Hassanin, Y. Zweiri, L. Seneviratne, [270] T. Wohlers, T. Gornet, Wohlers Rep. 24 (2014) 118.
Adv. Eng. Mater. 20 (2018), 1700552. [271] M. Izadi, A. Farzaneh, M. Mohammed, I. Gibson, B. Rolfe, Rapid Prototyp. J. 26
[223] K.S. Boparai, R. Singh, H. Singh, Rapid Prototyp. J. 22 (2016) 281–299. (2020) 1059–1078.
[224] R. Gupta, M. Dalakoti, A. Narasimhulu, Adv. Mater. Eng. Manuf. Proc. 1 (2020) [272] F. Zhang, P. Gao, H. Tan, Y. Li, Y. Chen, M. Mei, A.T. Clare, L.-C. Zhang, J. Mater.
31–39. Sci. Technol. 88 (2021) 132–142.
[225] Y.J. Liu, J.S. Zhang, X.C. Liu, X. Wu, J.C. Wang, Y.S. Zhang, L.Q. Wang, L. [273] L. Xie, H. Guo, Y. Song, C. Liu, Z. Wang, L. Hua, L. Wang, L.-C. Zhang, Mater.
C. Zhang, Mater. Sci. Eng. A 822 (2021), 141696. Charact. 161 (2020), 110137.
[226] P. Qin, L.Y. Chen, C.H. Zhao, Y.J. Liu, C.D. Cao, H. Sun, L.C. Zhang, Corros. Sci. [274] T. Abe, H. Sasahara, Addit. Manuf. 28 (2019) 639–648.
189 (2021), 109609. [275] F. Zhang, Y. Qiu, T. Hu, A.T. Clare, Y. Li, L.-C. Zhang, Mater. Sci. Eng. A 796
[227] M. Xie, S. Zhou, S. Zhao, J. Jin, D. Chen, L.-C. Zhang, J. Alloys. Compd. 838 (2020), 140063.
(2020), 155592. [276] S. Zhao, S. Zhou, M. Xie, X. Dai, D. Chen, L.-C. Zhang, J. Mater. Res. Technol. 8
[228] X. Luo, C. Yang, Z.Q. Fu, L.H. Liu, H.Z. Lu, H.W. Ma, Z. Wang, D.D. Li, L.C. Zhang, (2019) 2001–2010.
Y.Y. Li, Mater. Sci. Eng. A 823 (2021), 141731. [277] J. Lei, C. Shi, S. Zhou, Z. Gu, L.C. Zhang, Surf. Coat. Technol. 334 (2018)
[229] Y. Li, P. Pavanram, J. Zhou, K. Lietaert, F.S.L. Bobbert, Y. Kubo, M.A. Leeflang, 274–285.
H. Jahr, A.A. Zadpoor, Biomater. Sci. 8 (2020) 2404–2419. [278] C. Shi, J. Lei, S. Zhou, X. Dai, L.C. Zhang, J. Alloys. Compd. 744 (2018) 146–155.
[230] D.P. Laverty, M.B.M. Thomas, P. Clark, L.D. Addy, Dent. Update 43 (2016) [279] D.R. Feenstra, R. Banerjee, H.L. Fraser, A. Huang, A. Molotnikov, N. Birbilis, Curr.
826–835. Opin. Solid State Mater. Sci. 25 (2021), 100924.
[231] L. Bai, J. Zhang, X. Chen, C. Yi, R. Chen, Z. Zhang, Materials 11 (2018) 1856. [280] A. Singh, S. Kapil, M. Das, Addit. Manuf. 35 (2020), 101388.
[232] L.-Y. Chen, H.-Y. Zhang, C. Zheng, H.-Y. Yang, P. Qin, C. Zhao, S. Lu, S.-X. Liang, [281] V.K. Balla, P.D. DeVasConCellos, W. Xue, S. Bose, A. Bandyopadhyay, Acta
L. Chai, L.-C. Zhang, Mater. Des. 208 (2021), 109907. Biomater. 5 (2009) 1831–1837.
[233] P. Qin, L.Y. Chen, Y.J. Liu, Z. Jia, S.X. Liang, C.H. Zhao, H. Sun, L.C. Zhang, [282] J. Wang, Y. Liu, C.D. Rabadia, S.-X. Liang, T.B. Sercombe, L.-C. Zhang, J. Mater.
Corros. Sci. 191 (2021), 109728. Sci. Technol. 61 (2021) 221–233.
[234] Y.J. Liu, Y.S. Zhang, L.C. Zhang, Materialia 6 (2019), 100299. [283] F.A. España, V.K. Balla, S. Bose, A. Bandyopadhyay, Mater. Sci. Eng. C 30 (2010)
[235] J.C. Wang, Y.J. Liu, S.X. Liang, Y.S. Zhang, L.Q. Wang, T.B. Sercombe, L.C. Zhang, 50–57.
J. Mater. Sci. Technol. 105 (2022) 1–16. [284] J.B. Ferguson, B.F. Schultz, A.D. Moghadam, P.K. Rohatgi, J. Manuf. Process. 19
[236] Y. Yang, Y. Chen, J. Zhang, X. Gu, P. Qin, N. Dai, X. Li, J.-P. Kruth, L.-C. Zhang, (2015) 163–170.
Mater. Des. 146 (2018) 239–248. [285] I. Kunce, M. Polanski, K. Karczewski, T. Plocinski, K.J. Kurzydlowski, J. Alloys.
[237] Y.J. Liu, Z. Liu, Y. Jiang, G.W. Wang, Y. Yang, L.C. Zhang, J. Alloys. Compd. 735 Compd. 648 (2015) 751–758.
(2018) 1414–1421. [286] B.V. Krishna, S. Bose, A. Bandyopadhyay, J. Biomed. Mater. Res. Part B Appl.
[238] X.P. Li, X.J. Wang, M. Saunders, A. Suvorova, L.C. Zhang, Y.J. Liu, M.H. Fang, Z. Biomater. 89B (2009) 481–490.
H. Huang, T.B. Sercombe, Acta Mater. 95 (2015) 74–82. [287] W.E. Frazier, J. Mater. Eng. Perform. 23 (2014) 1917–1928.
[288] A. Bandyopadhyay, F. Espana, V.K. Balla, S. Bose, Y. Ohgami, N.M. Davies, Acta
Biomater. 6 (2010) 1640–1648.

51
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[289] V.K. Balla, A. Bandyopadhyay, Surf. Coat. Technol. 205 (2010) 2661–2667. [338] K.-S. Ha, G. Kwak, K.-W. Jun, J. Hwang, J. Lee, Chem. Commun. 49 (2013)
[290] G.P. Dinda, L. Song, J. Mazumder, Metall. Mater. Trans. A 39 (2008) 2914–2922. 5141–5143.
[291] N. Shamsaei, A. Yadollahi, L. Bian, S.M. Thompson, Addit. Manuf. 8 (2015) [339] Z. Wang, C.-J. Liu, Nano Energy 11 (2015) 277–293.
12–35. [340] H. Thakkar, S. Eastman, A. Hajari, A.A. Rownaghi, J.C. Knox, F. Rezaei, ACS Appl.
[292] Z.L. Lu, D.C. Li, B.H. Lu, A.F. Zhang, G.X. Zhu, G. Pi, Opt. Lasers Eng. 48 (2010) Mater. Interfaces 8 (2016) 27753–27761.
519–525. [341] S. Couck, J. Lefevere, S. Mullens, L. Protasova, V. Meynen, G. Desmet, G.V. Baron,
[293] Y. Zhai, H. Galarraga, D.A. Lados, Eng. Fail. Anal. 69 (2016) 3–14. J.F.M. Denayer, Chem. Eng. J. 308 (2017) 719–726.
[294] A.J. Sterling, B. Torries, N. Shamsaei, S.M. Thompson, D.W. Seely, Mater. Sci. [342] S.L. Taylor, A.E. Jakus, R.N. Shah, D.C. Dunand, Adv. Eng. Mater. 19 (2017),
Eng. A 655 (2016) 100–112. 1600365.
[295] J. Joseph, P. Hodgson, T. Jarvis, X. Wu, N. Stanford, D.M. Fabijanic, Mater. Sci. [343] C. Zhu, T.Y.-J. Han, E.B. Duoss, A.M. Golobic, J.D. Kuntz, C.M. Spadaccini, M.
Eng. A 733 (2018) 59–70. A. Worsley, Nat. Commun. 6 (2015) 6962.
[296] N.D. Stepanov, D.G. Shaysultanov, N.Y. Yurchenko, S.V. Zherebtsov, A. [344] H. Elsayed, N. Novak, M. Vesenjak, F. Zanini, S. Carmignato, L. Biasetto, Mater.
N. Ladygin, G.A. Salishchev, M.A. Tikhonovsky, Mater. Sci. Eng. A 636 (2015) Sci. Eng. A 787 (2020), 139484.
188–195. [345] E. Sachs, M. Cima, P. Williams, D. Brancazio, J. Cornie, J. Eng. Ind. 114 (1992)
[297] F.A. España, V.K. Balla, A. Bandyopadhyay, Surf. Coat. Technol. 204 (2010) 481–488.
2510–2517. [346] Y. Bai, C.B. Williams, Mater. Des. 147 (2018) 146–156.
[298] V.K. Balla, S. Bodhak, S. Bose, A. Bandyopadhyay, Acta Biomater. 6 (2010) [347] A. Mostafaei, E.L. Stevens, J.J. Ference, D.E. Schmidt, M. Chmielus, Addit. Manuf.
3349–3359. 21 (2018) 63–68.
[299] B.V. Krishna, S. Bose, A. Bandyopadhyay, Metall. Mater. Trans. A 38 (2007) [348] D. Karlsson, G. Lindwall, A. Lundbäck, M. Amnebrink, M. Boström, L. Riekehr,
1096–1103. M. Schuisky, M. Sahlberg, U. Jansson, Addit. Manuf. 27 (2019) 72–79.
[300] B.V. Krishna, S. Bose, A. Bandyopadhyay, Acta Biomater. 3 (2007) 997–1006. [349] X. Lv, F. Ye, L. Cheng, S. Fan, Y. Liu, Ceram. Int. 45 (2019) 12609–12624.
[301] P. Baranowski, P. Płatek, A. Antolak-Dudka, M. Sarzyński, M. Kucewicz, [350] W. Du, X. Ren, C. Ma, Z. Pei, Mater. Lett. 234 (2019) 327–330.
T. Durejko, J. Małachowski, J. Janiszewski, T. Czujko, Addit. Manuf. 25 (2019) [351] S. Huang, C. Ye, H. Zhao, Z. Fan, Addit. Manuf. 29 (2019), 100802.
307–316. [352] E. Sheydaeian, M. Vlasea, A. Woo, R. Pilliar, E. Hu, E. Toyserkani, J. Biomed.
[302] A. Antolak-Dudka, P. Płatek, T. Durejko, P. Baranowski, J. Małachowski, Mater. Res. Part B Appl. Biomater. 105 (2017) 828–835.
M. Sarzyński, T. Czujko, Materials 12 (2019) 1225. [353] Z. Zhou, F. Buchanan, C. Mitchell, N. Dunne, Mater. Sci. Eng. C 38 (2014) 1–10.
[303] B. Liao, R.F. Xia, W. Li, D. Lu, Z.M. Jin, J. Mater. Eng. Perform. 30 (2021) [354] S. Holland, T. Foster, W. MacNaughtan, C. Tuck, J. Food Eng. 220 (2018) 12–19.
4993–5004. [355] S. Holland, C. Tuck, T. Foster, Carbohydr. Polym. 200 (2018) 229–238.
[304] S. Mohd Yusuf, S. Cutler, N. Gao, Metals 9 (2019) 1286. [356] C. Polzin, S. Spath, H. Seitz, Rapid Prototyp. J. 19 (2013) 37–43.
[305] D.K. Yadav, R. Srivastava, S. Dev, Mater. Today Proc. 26 (2020) 2089–2093. [357] Y. Bai, C.B. Williams, Rapid Prototyp. J. 21 (2015) 177–185.
[306] V. Dhinakaran, K.P. Manoj Kumar, P.M. Bupathi Ram, M. Ravichandran, [358] S.M. Gaytan, M.A. Cadena, H. Karim, D. Delfin, Y. Lin, D. Espalin, E. MacDonald,
M. Vinayagamoorthy, Mater. Today Proc. 27 (2020) 752–756. R.B. Wicker, Ceram. Int. 41 (2015) 6610–6619.
[307] F. Daver, K.P.M. Lee, M. Brandt, R. Shanks, Compos. Sci. Technol. 168 (2018) [359] P. Lima, A. Zocca, W. Acchar, J. Günster, J. Eur. Ceram. Soc. 38 (2018)
230–237. 3395–3400.
[308] S. Vyavahare, S. Teraiya, D. Panghal, S. Kumar, Rapid Prototyp. J. 26 (2020) [360] I.M. Somasundram, A. Cendrowicz, D.I. Wilson, M.L. Johns, Chem. Eng. Sci. 63
176–201. (2008) 3802–3809.
[309] J. Torres, J. Cotelo, J. Karl, A.P. Gordon, JOM 67 (2015) 1183–1193. [361] T. Do, P. Kwon, C.S. Shin, Int. J. Mach. Tools Manuf. 121 (2017) 50–60.
[310] X. Liu, M. Zhang, S. Li, L. Si, J. Peng, Y. Hu, Int. J. Adv. Manuf. Technol. 89 [362] B.D. Kernan, E.M. Sachs, S.M. Allen, A. Lorenz, C. Sachs, L. Raffenbeul,
(2017) 2387–2397. A. Pettavino, Metall. Mater. Trans. A 36 (2005) 2815–2827.
[311] R.J. Zaldivar, D.B. Witkin, T. McLouth, D.N. Patel, K. Schmitt, J.P. Nokes, Addit. [363] N.B. Crane, J. Wilkes, E. Sachs, S.M. Allen, Rapid Prototyp. J. 12 (2006) 266–274.
Manuf. 13 (2017) 71–80. [364] A. Lores, N. Azurmendi, I. Agote, E. Zuza, Powder Metall. 62 (2019) 267–296.
[312] G. Dong, G. Wijaya, Y. Tang, Y.F. Zhao, Addit. Manuf. 19 (2018) 62–72. [365] S. Vangapally, K. Agarwal, A. Sheldon, S. Cai, Procedia Manuf. 10 (2017)
[313] M. Spoerk, J. Sapkota, G. Weingrill, T. Fischinger, F. Arbeiter, C. Holzer, 750–759.
Macromol. Mater. Eng. 302 (2017), 1700143. [366] T. Do, C.S. Shin, D. Stetsko, G. VanConant, A. Vartanian, S. Pei, P. Kwon, Procedia
[314] M. Rinaldi, T. Ghidini, F. Cecchini, A. Brandao, F. Nanni, Compos. Part B Eng. 145 Manuf. 1 (2015) 263–272.
(2018) 162–172. [367] M. Ziaee, E.M. Tridas, N.B. Crane, JOM 69 (2017) 592–596.
[315] V.H. Carneiro, S.D. Rawson, H. Puga, P.J. Withers, Sci. Rep. 11 (2021) 4974. [368] Y. Tang, Y. Zhou, T. Hoff, M. Garon, Y.F. Zhao, Mater. Sci. Technol. 32 (2016)
[316] B. Liu, Y. Wang, Z. Lin, T. Zhang, Mater. Lett. 263 (2020), 127252. 648–656.
[317] A.S. Vishwanath, K. Rane, J. Schaper, M. Strano, R. Casati, Powder Metall. [369] P. Nandwana, A.M. Elliott, D. Siddel, A. Merriman, W.H. Peter, S.S. Babu, Curr.
(2021), https://doi.org/10.1080/00325899.2021.1911457. Opin. Solid State Mater. Sci. 21 (2017) 207–218.
[318] M. Mohammadizadeh, H. Lu, I. Fidan, K. Tantawi, A. Gupta, S. Hasanov, Z. Zhang, [370] A. Mostafaei, E.L. Stevens, E.T. Hughes, S.D. Biery, C. Hilla, M. Chmielus, Mater.
F. Alifui-Segbaya, A. Rennie, Inventions 5 (2020) 44. Des. 108 (2016) 126–135.
[319] Z. Su, S. Zhang, L.-l. Liu, J. Wu, J. Alloys. Compd. 871 (2021), 159569. [371] M. Turker, D. Godlinski, F. Petzoldt, Mater. Charact. 59 (2008) 1728–1735.
[320] L. Weisensel, N. Travitzky, H. Sieber, P. Greil, Adv. Eng. Mater. 6 (2004) [372] A. Mostafaei, S.H.V.R. Neelapu, C. Kisailus, L.M. Nath, T.D.B. Jacobs,
899–903. M. Chmielus, Addit. Manuf. 24 (2018) 200–209.
[321] B. Dermeik, N. Travitzky, Adv. Eng. Mater. 22 (2020), 2000256. [373] Y. Bai, G. Wagner, C.B. Williams, J. Manuf. Sci. Eng. 139 (8) (2017), 081019.
[322] Y. Zhang, L. Wu, X. Guo, S. Kane, Y. Deng, Y.-G. Jung, J.-H. Lee, J. Zhang, [374] E. Sheydaeian, K. Sarikhani, P. Chen, E. Toyserkani, Mater. Des. 135 (2017)
J. Mater. Eng. Perform. 27 (2018) 1–13. 142–150.
[323] T. Obikawa, M. Yoshino, J. Shinozuka, J. Mater. Process. Technol. 89-90 (1999) [375] Y. Xiong, C. Qian, J. Sun, Dent. Mater. J. 31 (2012) 815–820.
171–176. [376] R.K. Enneti, K.C. Prough, T.A. Wolfe, A. Klein, N. Studley, J.L. Trasorras, Int. J.
[324] S. Shimizu, H.T. Fujii, Y.S. Sato, H. Kokawa, M.R. Sriraman, S.S. Babu, Acta Refract. Metals Hard Mater. 71 (2018) 28–35.
Mater. 74 (2014) 234–243. [377] A. Kumar, Y. Bai, A. Eklund, C.B. Williams, Procedia Manuf. 10 (2017) 935–944.
[325] R.R. Dehoff, S.S. Babu, Acta Mater. 58 (2010) 4305–4315. [378] A. Yegyan Kumar, Y. Bai, A. Eklund, C.B. Williams, Addit. Manuf. 24 (2018)
[326] A. Hehr, M. Norfolk, Rapid Prototyp. J. 26 (2020) 445–458. 115–124.
[327] H.T. Fujii, M.R. Sriraman, S.S. Babu, Metall. Mater. Trans. A 42 (2011) [379] P. Sang, L.-Y. Chen, C. Zhao, Z.-X. Wang, H. Wang, S. Lu, D. Song, J.-H. Xu, L.-
4045–4055. C. Zhang, Metals 9 (2019) 1342.
[328] M.N. Gussev, N. Sridharan, M. Norfolk, K.A. Terrani, S.S. Babu, Mater. Sci. Eng. A [380] L.-Y. Chen, T. Xu, H. Wang, P. Sang, S. Lu, Z.-X. Wang, S. Chen, L.-C. Zhang, Surf.
684 (2017) 606–616. Coat. Technol. 358 (2019) 467–480.
[329] F. Kümmel, T. Hausöl, H.W. Höppel, M. Göken, Acta Mater. 120 (2016) 150–158. [381] M.R. Rokni, S.R. Nutt, C.A. Widener, V.K. Champagne, R.H. Hrabe, J. Therm.
[330] N. Sridharan, M. Gussev, R. Seibert, C. Parish, M. Norfolk, K. Terrani, S.S. Babu, Spray Technol. 26 (2017) 1308–1355.
Acta Mater. 117 (2016) 228–237. [382] A. Moridi, S.M. Hassani-Gangaraj, M. Guagliano, M. Dao, Surf. Eng. 30 (2014)
[331] D.X. Luong, A.K. Subramanian, G.A.L. Silva, J. Yoon, S. Cofer, K. Yang, P. 369–395.
S. Owuor, T. Wang, Z. Wang, J. Lou, P.M. Ajayan, J.M. Tour, Adv. Mater. 30 [383] J. Sha, L.-Y. Chen, Y.-T. Liu, Z.-J. Yao, S. Lu, Z.-X. Wang, Q.-H. Zang, S.-H. Mao,
(2018), 1707416. L.-C. Zhang, Metals 10 (2020) 1688.
[332] L. Hao, D. Tang, T. Sun, W. Xiong, Z. Feng, K.E. Evans, Y. Li, Int. J. Precis. Eng. [384] L.-Y. Chen, H. Wang, C. Zhao, S. Lu, Z.-X. Wang, J. Sha, S. Chen, L.-C. Zhang, Surf.
Manuf. Technol. 8 (2021) 665–685. Coat. Technol. 369 (2019) 31–43.
[333] I. Salaoru, Z. Zhou, P. Morris, G.J. Gibbons, J. Appl. Polym. Sci. 133 (2016) [385] F.P.W. Melchels, J. Feijen, D.W. Grijpma, Biomaterials 31 (2010) 6121–6130.
43572. [386] F. Sillani, R.G. Kleijnen, M. Vetterli, M. Schmid, K. Wegener, Addit. Manuf. 27
[334] K. Cai, J. Sun, Q. Li, R. Wang, B. Li, J. Zhou, Appl. Phys. A 102 (2011) 501–507. (2019) 32–41.
[335] J. Seo, D.I. Kushner, M.A. Hickner, ACS Appl. Mater. Interfaces 8 (2016) [387] F.D. Bryant, G. Sui, M.C. Leu, Rapid Prototyp. J. 9 (2003) 19–23.
16656–16663. [388] H.V. Atkinson, S. Davies, Metall. Mater. Trans. A 31 (2000) 2981–3000.
[336] E. Fantino, A. Chiappone, I. Roppolo, D. Manfredi, R. Bongiovanni, C.F. Pirri, [389] M. Qian, W. Xu, M. Brandt, H.P. Tang, MRS Bull. 41 (2016) 775–784.
F. Calignano, Adv. Mater. 28 (2016) 3712–3717. [390] B. Vrancken, L. Thijs, J.-P. Kruth, J. Van Humbeeck, J. Alloys. Compd. 541 (2012)
[337] C. Liang, H. Xie, V. Schwartz, J. Howe, S. Dai, S.H. Overbury, J. Am. Chem. Soc. 177–185.
131 (2009) 7735–7741. [391] W. Li, S. Li, J. Liu, A. Zhang, Y. Zhou, Q. Wei, C. Yan, Y. Shi, Mater. Sci. Eng. A
663 (2016) 116–125.

52
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[392] C. Yan, L. Hao, A. Hussein, Q. Wei, Y. Shi, Mater. Sci. Eng. C 75 (2017) [446] Y.J. Liu, S.J. Li, L.C. Zhang, Y.L. Hao, T.B. Sercombe, Scr. Mater. 153 (2018)
1515–1524. 99–103.
[393] A. Ataee, Y. Li, D. Fraser, G. Song, C. Wen, Mater. Des. 137 (2018) 345–354. [447] M.M. Sychov, L.A. Lebedev, S.V. Dyachenko, L.A. Nefedova, Acta Astronaut. 150
[394] G. Pyka, A. Burakowski, G. Kerckhofs, M. Moesen, S. Van Bael, J. Schrooten, (2018) 81–84.
M. Wevers, Adv. Eng. Mater. 14 (2012) 363–370. [448] D.J. Yoo, Biomaterials 32 (2011) 7741–7754.
[395] C. Xu, L.-Y. Chen, C.-B. Zheng, H.-Y. Zhang, C.-H. Zhao, Z.-X. Wang, S. Lu, J.- [449] L. Han, S. Che, Adv. Mater. 30 (2018), 1705708.
W. Zhang, L.-C. Zhang, Adv. Eng. Mater. 23 (2021), 2001433. [450] H. Karcher, Manuscr. Math. 64 (1989) 291–357.
[396] Y.C. Chai, G. Kerckhofs, S.J. Roberts, S. Van Bael, E. Schepers, J. Vleugels, F. [451] T. Maconachie, M. Leary, B. Lozanovski, X. Zhang, M. Qian, O. Faruque,
P. Luyten, J. Schrooten, Biomaterials 33 (2012) 4044–4058. M. Brandt, Mater. Des. 183 (2019), 108137.
[397] K. Das, V.K. Balla, A. Bandyopadhyay, S. Bose, Scr. Mater. 59 (2008) 822–825. [452] C. Fee, Curr. Opin. Chem. Eng. 18 (2017) 10–15.
[398] A. du Plessis, I. Yadroitsava, I. Yadroitsev, S.G. le Roux, D.C. Blaine, Virtual Phys. [453] D.A. Clarke, F. Dolamore, C.J. Fee, P. Galvosas, D.J. Holland, Chem. Eng. Sci. 231
Prototyp. 13 (2018) 266–281. (2021), 116264.
[399] A. Zargarian, M. Esfahanian, J. Kadkhodapour, S. Ziaei-Rad, D. Zamani, Theor. [454] C. Yan, L. Hao, A. Hussein, P. Young, J. Mech. Behav. Biomed. Mater. 51 (2015)
Appl. Fract. Mech. 100 (2019) 225–232. 61–73.
[400] R. Wu, P.C.E. Roberts, S. Lyu, F. Zheng, C. Soutis, C. Diver, D. Zhou, L. Li, Z. Deng, [455] L. Zhang, S. Feih, S. Daynes, S. Chang, M.Y. Wang, J. Wei, W.F. Lu, Addit. Manuf.
Adv. Funct. Mater. 31 (2021), 2008252. 23 (2018) 505–515.
[401] L. Zhang, B. Song, L. Yang, Y. Shi, Acta Biomater. 112 (2020) 298–315. [456] O. Al-Ketan, R.K. Abu Al-Rub, Adv. Eng. Mater. 21 (2019), 1900524.
[402] H. Yan, Q. Zhang, W. Chen, G. Xie, J. Dang, T.J. Lu, Appl. Therm. Eng. 166 [457] F.S.L. Bobbert, K. Lietaert, A.A. Eftekhari, B. Pouran, S.M. Ahmadi, H. Weinans, A.
(2020), 114687. A. Zadpoor, Acta Biomater. 53 (2017) 572–584.
[403] X.Z. Zhang, M. Leary, H.P. Tang, T. Song, M. Qian, Curr. Opin. Solid State Mater. [458] H. Montazerian, E. Davoodi, M. Asadi-Eydivand, J. Kadkhodapour, M. Solati-
Sci. 22 (2018) 75–99. Hashjin, Mater. Des. 126 (2017) 98–114.
[404] W.P. Syam, W. Jianwei, B. Zhao, I. Maskery, W. Elmadih, R. Leach, Precis. Eng. 52 [459] L. Yang, C. Yan, W. Cao, Z. Liu, B. Song, S. Wen, C. Zhang, Y. Shi, S. Yang, Acta
(2018) 494–506. Mater. 181 (2019) 49–66.
[405] Z. Xiao, Y. Yang, R. Xiao, Y. Bai, C. Song, D. Wang, Mater. Des. 143 (2018) 27–37. [460] S.Z. Khan, S.H. Masood, E. Ibrahim, Z. Ahmad, Virtual Phys. Prototyp. 14 (2019)
[406] A. Seharing, A.H. Azman, S. Abdullah, Adv. Mech. Eng. 12 (2014) 1–21. 360–370.
[407] F. Warmuth, F. Osmanlic, L. Adler, M.A. Lodes, C. Körner, Smart Mater. Struct. 26 [461] Z. Cheng, R. Xu, P.-X. Jiang, Int. J. Heat Mass Transf. 170 (2021), 120902.
(2016), 025013. [462] A. Ataee, Y. Li, M. Brandt, C. Wen, Acta Mater. 158 (2018) 354–368.
[408] S. Yuan, F. Shen, J. Bai, C.K. Chua, J. Wei, K. Zhou, Mater. Des. 120 (2017) [463] N. Soro, H. Attar, X. Wu, M.S. Dargusch, Mater. Sci. Eng. A 745 (2019) 195–202.
317–327. [464] J. Kadkhodapour, H. Montazerian, A.C. Darabi, A.P. Anaraki, S.M. Ahmadi, A.
[409] J. Schwerdtfeger, P. Heinl, R.F. Singer, C. Körner, Phys. Status Solidi B 247 (2010) A. Zadpoor, S. Schmauder, J. Mech. Behav. Biomed. Mater. 50 (2015) 180–191.
269–272. [465] N. Soro, H. Attar, E. Brodie, M. Veidt, A. Molotnikov, M.S. Dargusch, J. Mech.
[410] I. Eldesouky, O. Harrysson, H. West, H. Elhofy, Addit. Manuf. 17 (2017) 169–175. Behav. Biomed. Mater. 97 (2019) 149–158.
[411] Q. Feng, Q. Tang, Z. Liu, Y. Liu, R. Setchi, Proc. Inst. Mech. Eng. Part B J. Eng. [466] M. Speirs, B. Van Hooreweder, J. Van Humbeeck, J.P. Kruth, J. Mech. Behav.
Manuf. 232 (2016) 1719–1730. Biomed. Mater. 70 (2017) 53–59.
[412] I. Maskery, A. Hussey, A. Panesar, A. Aremu, C. Tuck, I. Ashcroft, R. Hague, [467] C. Yan, L. Hao, A. Hussein, D. Raymont, Int. J. Mach. Tools Manuf. 62 (2012)
J. Cell. Plast. 53 (2016) 151–165. 32–38.
[413] E. Onal, J.E. Frith, M. Jurg, X. Wu, A. Molotnikov, Metals 8 (2018) 200. [468] D. Carluccio, C. Xu, J. Venezuela, Y. Cao, D. Kent, M. Bermingham, A.G. Demir,
[414] M. Smith, Z. Guan, W.J. Cantwell, Int. J. Mech. Sci. 67 (2013) 28–41. B. Previtali, Q. Ye, M. Dargusch, Acta Biomater. 103 (2020) 346–360.
[415] N. Tanlak, D.F. De Lange, W. Van Paepegem, Mater. Des. 133 (2017) 549–558. [469] J.B. Berger, H.N.G. Wadley, R.M. McMeeking, Nature 543 (2017) 533–537.
[416] E. Sallica-Leva, A.L. Jardini, J.B. Fogagnolo, J. Mech. Behav. Biomed. Mater. 26 [470] S.C. Han, J.W. Lee, K. Kang, Adv. Mater. 27 (2015) 5506–5511.
(2013) 98–108. [471] C. Bonatti, D. Mohr, J. Mech. Phys. Solids 122 (2019) 1–26.
[417] B. Van Hooreweder, J.-P. Kruth, CIRP Annals 66 (2017) 221–224. [472] M. Friedrich, M. Kružík, Arch. Ration. Mech. Anal. 238 (2020) 489–540.
[418] C. Neff, N. Hopkinson, N.B. Crane, Addit. Manuf. 22 (2018) 807–816. [473] H. Yu, J. Huang, B. Zou, W. Shao, J. Liu, Virtual Phys. Prototyp. 15 (2020) 35–48.
[419] F. Liu, D.Z. Zhang, P. Zhang, M. Zhao, S. Jafar, Materials 11 (2018) 374. [474] X. Chen, Q. Ji, J. Wei, H. Tan, J. Yu, P. Zhang, V. Laude, M. Kadic, Int. J. Mech.
[420] B. Jetté, V. Brailovski, M. Dumas, C. Simoneau, P. Terriault, J. Mech. Behav. Sci. 169 (2020), 105288.
Biomed. Mater. 77 (2018) 58–72. [475] Y. Wang, L. Zhang, S. Daynes, H. Zhang, S. Feih, M.Y. Wang, Mater. Des. 142
[421] M. Leary, M. Mazur, J. Elambasseril, M. McMillan, T. Chirent, Y. Sun, M. Qian, (2018) 114–123.
M. Easton, M. Brandt, Mater. Des. 98 (2016) 344–357. [476] D.B. Shao, S.C. Chen, Nano Lett. 6 (2006) 2279–2283.
[422] A. Aremu, I. Maskery, C. Tuck, I. Ashcroft, R. Wildman, R. Hague, A comparative [477] Y. Li, K. Lietaert, W. Li, X.Y. Zhang, M.A. Leeflang, J. Zhou, A.A. Zadpoor, Corros.
finite element study of cubic unit cells for selective laser melting, in: 25th Annual Sci. 156 (2019) 106–116.
International Solid Freeform Fabrication Symposium - An Additive Manufacturing [478] B. Wintiba, B. Sonon, K. Ehab Moustafa Kamel, T.J. Massart, Compos. Struct. 180
Conference, University of Texas at Austin, 2014, pp. 1238–1249. (2017) 955–971.
[423] Y. Li, Y. Ding, K. Munir, J. Lin, M. Brandt, A. Atrens, Y. Xiao, J.R. Kanwar, C. Wen, [479] J. Nguyen, S.I. Park, D.W. Rosen, L. Folgar, J. Williams, Conformal lattice
Acta Biomater. 87 (2019) 273–284. structure design and fabrication. Solid Freeform Fabrication Symposium, 2012,
[424] J. Ju, J.D. Summers, Mater. Des. 32 (2011) 512–524. pp. 138–161. Austin, TX.
[425] D. Mousanezhad, R. Ghosh, A. Ajdari, A.M.S. Hamouda, H. Nayeb-Hashemi, [480] Y. Liang, F. Zhao, D.-J. Yoo, B. Zheng, Rapid Prototyp. J. 26 (2020) 1005–1017.
A. Vaziri, Int. J. Mech. Sci. 89 (2014) 413–422. [481] Y. Chen, Comput. Des. 39 (2007) 975–986.
[426] S.Y. Choy, C.-N. Sun, K.F. Leong, J. Wei, Addit. Manuf. 16 (2017) 213–224. [482] D. Marschall, H. Rippl, F. Ehrhart, M. Schagerl, Mater. Des. 190 (2020), 108539.
[427] V.S. Deshpande, N.A. Fleck, M.F. Ashby, J. Mech. Phys. Solids 49 (2001) [483] B. Zhu, X. Zhang, H. Zhang, J. Liang, H. Zang, H. Li, R. Wang, Mech. Mach.
1747–1769. Theory 143 (2020), 103622.
[428] T. Tancogne-Dejean, A.B. Spierings, D. Mohr, Acta Mater. 116 (2016) 14–28. [484] T. Dbouk, Appl. Therm. Eng. 112 (2017) 841–854.
[429] K.K. Wong, K.C. Leong, Int. J. Heat Mass Transf. 121 (2018) 46–63. [485] T. Pham, P. Kwon, S. Foster, Energies 14 (2021) 283.
[430] Y.J. Liu, D.C. Ren, S.J. Li, H. Wang, L.C. Zhang, T.B. Sercombe, Addit. Manuf. 32 [486] V.J. Challis, A.P. Roberts, J.F. Grotowski, L.C. Zhang, T.B. Sercombe, Adv. Eng.
(2020), 101060. Mater. 12 (2010) 1106–1110.
[431] L. Xiao, W. Song, C. Wang, H. Liu, H. Tang, J. Wang, Mater. Sci. Eng. A 640 [487] A. Sohouli, A. Kefal, A. Abdelhamid, M. Yildiz, A. Suleman, Struct. Multidiscip.
(2015) 375–384. Optim. 62 (2020) 2375–2389.
[432] X. Cao, S. Duan, J. Liang, W. Wen, D. Fang, Int. J. Mech. Sci. 145 (2018) 53–63. [488] S.B. Dilgen, C.B. Dilgen, D.R. Fuhrman, O. Sigmund, B.S. Lazarov, Struct.
[433] Y. Wang, S. Arabnejad, M. Tanzer, D. Pasini, J. Mech. Des. 140 (2018), 111406. Multidiscip. Optim. 57 (2018) 1905–1918.
[434] H.X. Zhu, J.F. Knott, N.J. Mills, J. Mech. Phys. Solids 45 (1997) 319–343. [489] J. Hu, Y. Luo, S. Liu, Compos. Struct. (2021), 114224.
[435] S.K. Nammi, P. Myler, G. Edwards, Mater. Des. 31 (2010) 712–722. [490] M. Tarek, T. Ray, Comput. Methods Appl. Mech. Eng. 363 (2020), 112880.
[436] J.G. Fourie, J.P. Du Plessis, Chem. Eng. Sci. 57 (2002) 2781–2789. [491] X. Liu, W.-J. Yi, Q.S. Li, P.-S. Shen, J. Constr. Steel Res. 64 (2008) 305–311.
[437] S.L. Sing, W.Y. Yeong, F.E. Wiria, B.Y. Tay, Exp. Mech. 56 (2016) 735–748. [492] P. Tanskanen, Comput. Methods Appl. Mech. Eng. 191 (2002) 5485–5498.
[438] A. Zargarian, M. Esfahanian, J. Kadkhodapour, S. Ziaei-Rad, Mater. Sci. Eng. C 60 [493] L. Xia, Q. Xia, X. Huang, Y.M. Xie, Arch. Comput. Methods Eng. 25 (2018)
(2016) 339–347. 437–478.
[439] D.S.J. Al-Saedi, S.H. Masood, M. Faizan-Ur-Rab, A. Alomarah, P. Ponnusamy, [494] L. Xia, L. Zhang, Q. Xia, T. Shi, Comput. Methods Appl. Mech. Eng. 333 (2018)
Mater. Des. 144 (2018) 32–44. 356–370.
[440] I. Ullah, M. Brandt, S. Feih, Mater. Des. 92 (2016) 937–948. [495] D.J. Munk, T. Kipouros, G.A. Vio, G.T. Parks, G.P. Steven, Struct. Multidiscip.
[441] A. Ortona, C. D’Angelo, S. Gianella, D. Gaia, Mater. Lett. 80 (2012) 95–98. Optim. 57 (2018) 665–688.
[442] S. Saremi, S. Mirjalili, A. Lewis, Adv. Eng. Softw. 105 (2017) 30–47. [496] H. Liu, H. Zong, T. Shi, Q. Xia, Comput. Methods Appl. Mech. Eng. 367 (2020),
[443] M. Zhang, Y. Yang, D. Wang, Z. Xiao, C. Song, C. Weng, Mater. Sci. Eng. A 736 113154.
(2018) 288–297. [497] P. Wei, Z. Li, X. Li, M.Y. Wang, Struct. Multidiscip. Optim. 58 (2018) 831–849.
[444] R. Wauthle, B. Vrancken, B. Beynaerts, K. Jorissen, J. Schrooten, J.-P. Kruth, [498] M.P. Bendsøe, N. Kikuchi, Comput. Methods Appl. Mech. Eng. 71 (1988)
J. Van Humbeeck, Addit. Manuf. 5 (2015) 77–84. 197–224.
[445] S.J. Li, Q.S. Xu, Z. Wang, W.T. Hou, Y.L. Hao, R. Yang, L.E. Murr, Acta Biomater. [499] M.P. Bendsøe, O. Sigmund, Arch. Appl. Mech. 69 (1999) 635–654.
10 (2014) 4537–4547. [500] Y.M. Xie, G.P. Steven, Comput. Struct. 49 (1993) 885–896.
[501] X. Huang, Y.M. Xie, Finite Elem. Anal. Des. 43 (2007) 1039–1049.

53
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[502] J.H. Zhu, W.H. Zhang, K.P. Qiu, Comput. Mech. 40 (2006) 97. [550] J.-Y. Rho, L. Kuhn-Spearing, P. Zioupos, Med. Eng. Phys. 20 (1998) 92–102.
[503] M.Y. Wang, X. Wang, D. Guo, Comput. Methods Appl. Mech. Eng. 192 (2003) [551] J. Wieding, T. Lindner, P. Bergschmidt, R. Bader, Biomaterials 46 (2015) 35–47.
227–246. [552] C. Torres-Sanchez, F.R.A. Al Mushref, M. Norrito, K. Yendall, Y. Liu, P.P. Conway,
[504] L. Xia, P. Breitkopf, Comput. Methods Appl. Mech. Eng. 286 (2015) 147–167. Mater. Sci. Eng. C 77 (2017) 219–228.
[505] Q. Pan, H. Zhou, Q. Lu, H. Gao, L. Lu, Nature 551 (2017) 214–217. [553] S. Bose, M. Roy, A. Bandyopadhyay, Trends Biotechnol. 30 (2012) 546–554.
[506] Z. Wang, X. Wang, T. Gao, C. Shi, Mech. Adv. Mater. Struct. 28 (2021) [554] A. Velasco-Hogan, J. Xu, M.A. Meyers, Adv. Mater. 30 (2018), 1800940.
2057–2069. [555] Y. Qin, P. Wen, H. Guo, D. Xia, Y. Zheng, L. Jauer, R. Poprawe, M. Voshage, J.
[507] S. Zhao, S.J. Li, W.T. Hou, Y.L. Hao, R. Yang, R.D.K. Misra, J. Mech. Behav. H. Schleifenbaum, Acta Biomater. 98 (2019) 3–22.
Biomed. Mater. 59 (2016) 251–264. [556] Y. Li, H. Jahr, J. Zhou, A.A. Zadpoor, Acta Biomater. 115 (2020) 29–50.
[508] S. Amin Yavari, S.M. Ahmadi, R. Wauthle, B. Pouran, J. Schrooten, H. Weinans, A. [557] E. Garner, H.M.A. Kolken, C.C.L. Wang, A.A. Zadpoor, J. Wu, Addit. Manuf. 26
A. Zadpoor, J. Mech. Behav. Biomed. Mater. 43 (2015) 91–100. (2019) 65–75.
[509] S.M. Ahmadi, R. Hedayati, Y. Li, K. Lietaert, N. Tümer, A. Fatemi, C.D. Rans, [558] H. Attar, K.G. Prashanth, A.K. Chaubey, M. Calin, L.C. Zhang, S. Scudino,
B. Pouran, H. Weinans, A.A. Zadpoor, Acta Biomater. 65 (2018) 292–304. J. Eckert, Mater. Lett. 142 (2015) 38–41.
[510] M. Afshar, A.P. Anaraki, H. Montazerian, J. Kadkhodapour, J. Mech. Behav. [559] S. Roy, N. Khutia, D. Das, M. Das, V.K. Balla, A. Bandyopadhyay, A.
Biomed. Mater. 62 (2016) 481–494. R. Chowdhury, Mater. Sci. Eng. C 64 (2016) 436–443.
[511] A. Yánez, M.P. Fiorucci, A. Cuadrado, O. Martel, D. Monopoli, Int. J. Fatigue 138 [560] H. Attar, M. Calin, L.C. Zhang, S. Scudino, J. Eckert, Mater. Sci. Eng. A 593 (2014)
(2020), 105702. 170–177.
[512] S. Zhao, S.J. Li, S.G. Wang, W.T. Hou, Y. Li, L.C. Zhang, Y.L. Hao, R. Yang, R.D.K. [561] Y. Liu, S. Li, W. Hou, S. Wang, Y. Hao, R. Yang, T.B. Sercombe, L.-C. Zhang,
K. Misra, L.E. Murr, Acta Mater. 150 (2018) 1–15. J. Mater. Sci. Technol. 32 (2016) 505–508.
[513] C.D. Rabadia, Y.J. Liu, C.H. Zhao, J.C. Wang, S.F. Jawed, L.Q. Wang, L.Y. Chen, [562] N. Hafeez, J. Liu, L. Wang, D. Wei, Y. Tang, W. Lu, L.-C. Zhang, Addit. Manuf. 34
H. Sun, L.C. Zhang, Mater. Sci. Eng. A 766 (2019), 138340. (2020), 101264.
[514] C.N. Kelly, J. Francovich, S. Julmi, D. Safranski, R.E. Guldberg, H.J. Maier, [563] W. Yuan, W. Hou, S. Li, Y. Hao, R. Yang, L.-C. Zhang, Y. Zhu, J. Mater. Sci.
K. Gall, Acta Biomater. 94 (2019) 610–626. Technol. 34 (2018) 1127–1131.
[515] N.W. Hrabe, P. Heinl, B. Flinn, C. Körner, R.K. Bordia, J. Biomed. Mater. Res. Part [564] N. Dai, L.-C. Zhang, J. Zhang, X. Zhang, Q. Ni, Y. Chen, M. Wu, C. Yang, Corros.
B Appl. Biomater. 99B (2011) 313–320. Sci. 111 (2016) 703–710.
[516] S.M. Ahmadi, R. Kumar, E.V. Borisov, R. Petrov, S. Leeflang, Y. Li, N. Tümer, [565] N. Dai, L.-C. Zhang, J. Zhang, Q. Chen, M. Wu, Corros. Sci. 102 (2016) 484–489.
R. Huizenga, C. Ayas, A.A. Zadpoor, V.A. Popovich, Acta Biomater. 83 (2019) [566] B.V. Hooreweder, K. Lietaert, B. Neirinck, N. Lippiatt, M. Wevers, J. Mech. Behav.
153–166. Biomed. Mater. 70 (2017) 60–67.
[517] S. Amin Yavari, R. Wauthle, J. van der Stok, A.C. Riemslag, M. Janssen, [567] P. Caravaggi, E. Liverani, A. Leardini, A. Fortunato, C. Belvedere, F. Baruffaldi,
M. Mulier, J.P. Kruth, J. Schrooten, H. Weinans, A.A. Zadpoor, Mater. Sci. Eng. C M. Fini, A. Parrilli, M. Mattioli-Belmonte, L. Tomesani, S. Pagani, J. Biomed.
33 (2013) 4849–4858. Mater. Res. Part B Appl. Biomater. 107 (2019) 2343–2353.
[518] P. Heinl, C. Körner, R.F. Singer, Adv. Eng. Mater. 10 (2008) 882–888. [568] T. Zhong, K. He, H. Li, L. Yang, Mater. Des. 181 (2019), 108076.
[519] L. Chen, J. Li, Y. Zhang, W. Lu, L.C. Zhang, L. Wang, D. Zhang, J. Nucl. Sci. [569] C. Yan, L. Hao, A. Hussein, P. Young, D. Raymont, Mater. Des. 55 (2014)
Technol. 53 (2016) 496–507. 533–541.
[520] L. Chen, Q. Zeng, J. Li, J. Lu, Y. Zhang, L.-C. Zhang, X. Qin, W. Lu, L. Zhang, [570] R. Li, J. Liu, Y. Shi, M. Du, Z. Xie, J. Mater. Eng. Perform. 19 (2010) 666–671.
L. Wang, D. Zhang, Mater. Des. 92 (2016) 888–896. [571] S. Bernard, V. Krishna Balla, S. Bose, A. Bandyopadhyay, J. Mech. Behav. Biomed.
[521] L.Y. Chen, P. Sang, L. Zhang, D. Song, Y.Q. Chu, L. Chai, L.C. Zhang, Metals 8 Mater. 13 (2012) 62–68.
(2018) 759. [572] M.H. Elahinia, M. Hashemi, M. Tabesh, S.B. Bhaduri, Prog. Mater. Sci. 57 (2012)
[522] Y. Chen, J. Zhang, N. Dai, P. Qin, H. Attar, L.-C. Zhang, Electrochim. Acta 232 911–946.
(2017) 89–97. [573] Z. Gorgin Karaji, M. Speirs, S. Dadbakhsh, J.P. Kruth, H. Weinans, A.A. Zadpoor,
[523] M. Geetha, A.K. Singh, R. Asokamani, A.K. Gogia, Prog. Mater. Sci. 54 (2009) S. Amin Yavari, ACS Appl. Mater. Interfaces 9 (2017) 1293–1304.
397–425. [574] Q. Yang, K. Sun, C. Yang, M. Sun, H. Peng, X. Shen, S. Huang, J. Chen, J. Alloys.
[524] E. Zhang, X. Zhao, J. Hu, R. Wang, S. Fu, G. Qin, Bioact. Mater. 6 (2021) Compd. 858 (2021), 157674.
2569–2612. [575] H. Ibrahim, A. Jahadakbar, A. Dehghan, N.S. Moghaddam, A. Amerinatanzi,
[525] L.E. Murr, J. Mater. Res. Technol. 9 (2020) 1087–1103. M. Elahinia, Metals 8 (2018) 164.
[526] P. Bocchetta, L.-Y. Chen, J.D. Tardelli, A.C. Reis, F. Almeraya-Calderón, P. Leo, [576] Y. Guo, K. Xie, W. Jiang, L. Wang, G. Li, S. Zhao, W. Wu, Y. Hao, ACS Biomater.
Coatings 11 (2021) 487. Sci. Eng. 5 (2019) 1123–1133.
[527] L. Chai, Y. Zhu, X. Hu, K.L. Murty, N. Guo, L.-Y. Chen, Y. Ma, L.-C. Zhang, Scr. [577] A. Bandyopadhyay, I. Mitra, A. Shivaram, N. Dasgupta, S. Bose, Addit. Manuf. 28
Mater. 196 (2021), 113761. (2019) 259–266.
[528] L. Guan, X. Yang, Y. Li, C. Shi, D. Yan, J. Huang, L.-C. Zhang, J. Appl. [578] R. Wauthle, J. van der Stok, S. Amin Yavari, J. Van Humbeeck, J.-P. Kruth, A.
Electrochem. (2021). A. Zadpoor, H. Weinans, M. Mulier, J. Schrooten, Acta Biomater. 14 (2015)
[529] L. Wang, L. Xie, Y. Lv, L.-C. Zhang, L. Chen, Q. Meng, J. Qu, D. Zhang, W. Lu, Acta 217–225.
Mater. 131 (2017) 499–510. [579] J. Yang, H. Cai, J. Lv, K. Zhang, H. Leng, C. Sun, Z. Wang, Z. Liu, Spine 39 (2014)
[530] N. Dai, J. Zhang, Y. Chen, L.-C. Zhang, J. Electrochem. Soc. 164 (2017) E486–E492.
C428–C434. [580] Y.C. Wu, C.N. Kuo, M.Y. Shie, Y.L. Su, L.J. Wei, S.Y. Chen, J.C. Huang, Mater. Des.
[531] D. Kuroda, M. Niinomi, M. Morinaga, Y. Kato, T. Yashiro, Mater. Sci. Eng. A 243 158 (2018) 256–265.
(1998) 244–249. [581] A. Cutolo, B. Neirinck, K. Lietaert, C. de Formanoir, B. Van Hooreweder, Addit.
[532] J. Jin, S. Zhou, Y. Zhao, Q. Zhang, X. Wang, W. Li, D. Chen, L.-C. Zhang, Opt. Manuf. 23 (2018) 498–504.
Laser Technol. 134 (2021), 106644. [582] J. Čapek, M. Machová, M. Fousová, J. Kubásek, D. Vojtěch, J. Fojt, E. Jablonská,
[533] N. Hafeez, S. Liu, E. Lu, L. Wang, R. Liu, W. lu, L.C. Zhang, J. Alloys. Compd. 790 J. Lipov, T. Ruml, Mater. Sci. Eng. C 69 (2016) 631–639.
(2019) 117–126. [583] R. Wauthle, S.M. Ahmadi, S. Amin Yavari, M. Mulier, A.A. Zadpoor, H. Weinans,
[534] M. Niinomi, Mater. Sci. Eng. A 243 (1998) 231–236. J. Van Humbeeck, J.-P. Kruth, J. Schrooten, Mater. Sci. Eng. C 54 (2015) 94–100.
[535] T. Majima, K. Yasuda, T. Tsuchida, K. Tanaka, K. Miyakawa, A. Minami, [584] H. Attar, L. Löber, A. Funk, M. Calin, L.C. Zhang, K.G. Prashanth, S. Scudino, Y.
K. Hayashi, J. Orthop. Sci. 8 (2003) 836–841. S. Zhang, J. Eckert, Mater. Sci. Eng. A 625 (2015) 350–356.
[536] C.D. Rabadia, Y.J. Liu, L.Y. Chen, S.F. Jawed, L.Q. Wang, H. Sun, L.C. Zhang, [585] Y. Liu, J. Zhang, Q. Tan, Y. Yin, M. Li, M.-X. Zhang, Opt. Laser Technol. 134
Mater. Des. 179 (2019), 107891. (2021), 106649.
[537] D. Sumner, T. Turner, R. Igloria, R. Urban, J. Galante, J. Biomech. 31 (1998) [586] J.K. Chen, M.W. Wu, T.L. Cheng, P.H. Chiang, Mater. Sci. Eng. C 100 (2019)
909–917. 781–788.
[538] H.Z. Lu, C. Yang, X. Luo, H.W. Ma, B. Song, Y.Y. Li, L.C. Zhang, Mater. Sci. Eng. A [587] S.M. Ahmadi, S.A. Yavari, R. Wauthle, B. Pouran, J. Schrooten, H. Weinans, A.
763 (2019), 138166. A. Zadpoor, Materials 8 (2015) 1871–1896.
[539] L.-Y. Chen, Y.-W. Cui, L.-C. Zhang, Metals 10 (2020) 1139. [588] X.Y. Cheng, S.J. Li, L.E. Murr, Z.B. Zhang, Y.L. Hao, R. Yang, F. Medina, R.
[540] S. Liu, S. Han, L. Zhang, L.-Y. Chen, L. Wang, L. Zhang, Y. Tang, J. Liu, H. Tang, B. Wicker, J. Mech. Behav. Biomed. Mater. 16 (2012) 153–162.
L.-C. Zhang, Compos. Part B Eng. 200 (2020), 108358. [589] S. Arabnejad, R. Burnett Johnston, J.A. Pura, B. Singh, M. Tanzer, D. Pasini, Acta
[541] Y.-W. Cui, L.-Y. Chen, X.-X. Liu, Curr. Nanosci. 17 (2021) 241–256. Biomater. 30 (2016) 345–356.
[542] L.-C. Zhang, L.-Y. Chen, L. Wang, Adv. Eng. Mater. 22 (2020), 1901258. [590] X. Li, C. Wang, W. Zhang, Y. Li, Mater. Lett. 63 (2009) 403–405.
[543] X. Luo, L.H. Liu, C. Yang, H.Z. Lu, H.W. Ma, Z. Wang, D.D. Li, L.C. Zhang, Y.Y. Li, [591] V. Weißmann, R. Bader, H. Hansmann, N. Laufer, Mater. Des. 95 (2016) 188–197.
J. Mater. Sci. Technol. 68 (2021) 112–123. [592] S. Li, Q. Xu, Z. Wang, W. Hou, Y. Hao, R. Yang, L.J.Ab. Murr, Acta Biomater. 10
[544] P. Qin, Y. Chen, Y.-J. Liu, J. Zhang, L.-Y. Chen, Y. Li, X. Zhang, C. Cao, H. Sun, L.- (2014) 4537–4547.
C. Zhang, ACS Biomater. Sci. Eng. 5 (2019) 1141–1149. [593] J. Kadkhodapour, H. Montazerian, A.C. Darabi, A. Zargarian, S. Schmauder,
[545] R.A. White, J.N. Weber, E.W. White, Science 176 (1972) 922. J. Mech. Behav. Biomed. Mater. 70 (2017) 28–42.
[546] P. Qin, Y. Liu, T.B. Sercombe, Y. Li, C. Zhang, C. Cao, H. Sun, L.-C. Zhang, ACS [594] L.E. Murr, S.M. Gaytan, F. Medina, E. Martinez, J.L. Martinez, D.H. Hernandez, B.
Biomater. Sci. Eng. 4 (2018) 2633–2642. I. Machado, D.A. Ramirez, R.B. Wicker, Mater. Sci. Eng. A 527 (2010) 1861–1868.
[547] A. Ibrahim, F. Zhang, E. Otterstein, E. Burkel, Mater. Des. 32 (2011) 146–153. [595] Y. Liu, S. Li, W. Hou, S. Wang, Y. Hao, R. Yang, T.B. Sercombe, L.-C. Zhang,
[548] B.T. Prayoga, R. Dharmastiti, F. Akbar, Suyitno, J. Mech. Sci. Technol. 32 (2018) J. Mater. Sci. Technol. 32 (2016) 505–508.
149–156. [596] C. Han, C. Yan, S. Wen, T. Xu, S. Li, J. Liu, Q. Wei, Y. Shi, Rapid Prototyp. J. 23
[549] V. Karageorgiou, D. Kaplan, Biomaterials 26 (2005) 5474–5491. (2017) 16–27.

54
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[597] S.L. Sing, S. Wang, S. Agarwala, F.E. Wiria, T.M.H. Ha, W.Y. Yeong, Int. J. [645] H. Yu, Z. Han, J. Zhang, S. Zhang, Wear 482-483 (2021), 203955.
Bioprinting 3 (2017) 007. [646] L. Liu, J. Zhang, G. Zhao, Z. Wang, W. Chen, Struct. Multidiscip. Optim. 62 (2020)
[598] Y. Li, H. Jahr, P. Pavanram, F.S.L. Bobbert, U. Puggi, X.Y. Zhang, B. Pouran, M. 175–192.
A. Leeflang, H. Weinans, J. Zhou, A.A. Zadpoor, Acta Biomater. 96 (2019) [647] Z. Li, S.L. Gobbi, J. Mater. Process. Technol. 70 (1997) 137–144.
646–661. [648] R. Huang, M. Riddle, D. Graziano, J. Warren, S. Das, S. Nimbalkar, J. Cresko,
[599] V. Manakari, G. Parande, M. Gupta, Metals 7 (2017) 2. E. Masanet, J. Clean. Prod. 135 (2016) 1559–1570.
[600] W.-n. Zhang, L.-z. Wang, Z.-x. Feng, Y.-m. Chen, Optik 207 (2020), 163842. [649] J. Robbins, S.J. Owen, B.W. Clark, T.E. Voth, Addit. Manuf. 12 (2016) 296–304.
[601] C. Yang, Z. Huan, X. Wang, C. Wu, J. Chang, ACS Biomater. Sci. Eng. 4 (2018) [650] P. Zhang, J. Liu, A.C. To, Scr. Mater. 135 (2017) 148–152.
608–616. [651] S.K. Moon, Y.E. Tan, J. Hwang, Y.-J. Yoon, Int. J. Precis. Eng. Manuf. Technol. 1
[602] Y. Li, J. Zhou, P. Pavanram, M.A. Leeflang, L.I. Fockaert, B. Pouran, N. Tümer, K. (2014) 223–228.
U. Schröder, J.M.C. Mol, H. Weinans, H. Jahr, A.A. Zadpoor, Acta Biomater. 67 [652] L. Chougrani, J.-P. Pernot, P. Véron, S. Abed, Comput. Des. 90 (2017) 95–104.
(2018) 378–392. [653] P. Heinl, L. Müller, C. Körner, R.F. Singer, F.A. Müller, Acta Biomater. 4 (2008)
[603] P. Wen, Y. Qin, Y. Chen, M. Voshage, L. Jauer, R. Poprawe, J.H. Schleifenbaum, 1536–1544.
J. Mater. Sci. Technol. 35 (2019) 368–376. [654] T.-T. Duan, F. Qiu, L. Zhu, T.-T. Li, H.-Y. Yang, Z.-H. Bai, M. Zha, L.-Y. Shan, Q.-
[604] P. Wen, M. Voshage, L. Jauer, Y. Chen, Y. Qin, R. Poprawe, J.H. Schleifenbaum, C. Jiang, Materialia 12 (2020), 100769.
Mater. Des. 155 (2018) 36–45. [655] F. Xu, X. Zhang, H. Zhang, Eng. Struct. 171 (2018) 309–325.
[605] C.C. Ng, M.M. Savalani, H.C. Man, I. Gibson, Virtual Phys. Prototyp. 5 (2010) [656] M. Bodaghi, A. Serjouei, A. Zolfagharian, M. Fotouhi, H. Rahman, D. Durand, Int.
13–19. J. Mech. Sci. 173 (2020), 105451.
[606] J. Kim, S. Oh, H. Ki, J. Mater. Process. Technol. 232 (2016) 131–141. [657] H. Yazdani Sarvestani, A.H. Akbarzadeh, A. Mirbolghasemi, K. Hermenean,
[607] M. Gieseke, C. Noelke, S. Kaierle, V. Wesling, H. Haferkamp, in: N. Hort, S. Mater. Des. 160 (2018) 179–193.
N. Mathaudhu, N.R. Neelameggham, M. Alderman (Eds.), Magnesium Technology [658] L. Meng, J. Shi, C. Yang, T. Gao, Y. Hou, L. Song, D. Gu, J. Zhu, P. Breitkopf,
2013, Springer International Publishing, Cham, 2013, pp. 65–68. W. Zhang, Extreme Mech. Lett. 40 (2020), 100869.
[608] L. Jauer, W. Meiners, S. Vervoort, C. Gayer, N.A. Zumdick, D. Zander, Selective [659] A.A.A. Alghamdi, Thin-walled Struct. 39 (2001) 189–213.
laser melting of magnesium alloys, European congress and exhibition on powder [660] M. Zhao, F. Liu, G. Fu, D.Z. Zhang, T. Zhang, H. Zhou, Materials 11 (2018) 2411.
metallurgy, in: European PM Conference Proceedings, The European Powder [661] D. Li, W. Liao, N. Dai, Y.M. Xie, Materials 12 (2019) 2183.
Metallurgy Association, Shrewsbury, 2016, pp. 1–6. [662] N. Toader, W. Sobek, K.G. Nickel, J. Bionic Eng. 14 (2017) 369–378.
[609] A.G. Demir, L. Monguzzi, B. Previtali, Addit. Manuf. 15 (2017) 20–28. [663] R. Saxena, S. Siddique, T. Gupta, R.K. Sharma, S. Chaudhary, Constr. Build.
[610] M. Moravej, A. Purnama, M. Fiset, J. Couet, D. Mantovani, Acta Biomater. 6 Mater. 176 (2018) 415–421.
(2010) 1843–1851. [664] D.-Y. Yoo, M.-J. Kim, Constr. Build. Mater. 209 (2019) 354–363.
[611] D.-T. Chou, D. Wells, D. Hong, B. Lee, H. Kuhn, P.N. Kumta, Acta Biomater. 9 [665] Q.Q. Li, E. Li, T. Chen, L. Wu, G.Q. Wang, Z.C. He, Thin-walled Struct. 162 (2021),
(2013) 8593–8603. 107588.
[612] Y. Yang, P. Wu, X. Lin, Y. Liu, H. Bian, Y. Zhou, C. Gao, C. Shuai, Virtual Phys. [666] L. Yu, X. Gu, L. Qian, P. Jiang, W. Wang, M. Yu, Thin-walled Struct. 161 (2021),
Prototyp. 11 (2016) 173–181. 107410.
[613] C. Liu, M. Zhang, C. Chen, Mater. Sci. Eng. A 703 (2017) 359–371. [667] G. Gao, S. Wang, Int. J. Crashworthiness 24 (2019) 664–676.
[614] A. Kopp, T. Derra, M. Müther, L. Jauer, J.H. Schleifenbaum, M. Voshage, O. Jung, [668] A. Mehrara Molan, M. Moomen, K. Ksaibati, J. Safety Res. 75 (2020) 155–165.
R. Smeets, N. Kröger, Acta Biomater. 98 (2019) 23–35. [669] S. Gopinath, A. Prakash, F.A. A. K, Int. J. Crashworthiness (2020) 1–7.
[615] C. Shuai, W. Yang, Y. Yang, H. Pan, C. He, F. Qi, D. Xie, H. Liang, Mater. Res. [670] D. Bruski, S. Burzyński, J. Chróścielewski, K. Jamroz, Ł. Pachocki, W. Witkowski,
Express 7 (2020), 015404. K. Wilde, Eng. Fail. Anal. 104 (2019) 227–246.
[616] D. Hong, D.-T. Chou, O.I. Velikokhatnyi, A. Roy, B. Lee, I. Swink, I. Issaev, H. [671] A. Al-Othman, N.N. Darwish, M. Qasim, M. Tawalbeh, N.A. Darwish, N. Hilal,
A. Kuhn, P.N. Kumta, Acta Biomater. 45 (2016) 375–386. Desalination 457 (2019) 39–61.
[617] P. Sharma, K.G. Jain, P.M. Pandey, S. Mohanty, Mater. Sci. Eng. C 106 (2020), [672] J. Pan, H. Fang, M.C. Xu, Y.F. Wu, Thin-walled Struct. 130 (2018) 85–100.
110247. [673] J. Pan, H. Fang, M.C. Xu, X.Z. Xue, Thin-walled Struct. 157 (2020), 107010.
[618] Y. Lai, Y. Li, H. Cao, J. Long, X. Wang, L. Li, C. Li, Q. Jia, B. Teng, T. Tang, J. Peng, [674] Z. Yu, J. Amdahl, D. Kristiansen, P.T. Bore, Ocean. Eng. 194 (2019), 106653.
D. Eglin, M. Alini, D.W. Grijpma, G. Richards, L. Qin, Biomaterials 197 (2019) [675] T. Moan, J. Amdahl, G. Ersdal, Mar. Struct. 63 (2019) 480–494.
207–219. [676] A.B.M. Supian, S.M. Sapuan, M.Y.M. Zuhri, E.S. Zainudin, H.H. Ya, Def. Technol.
[619] M.-q. Cheng, T. Wahafu, G.-f. Jiang, W. Liu, Y.-q. Qiao, X.-c. Peng, T. Cheng, X.- 14 (2018) 291–305.
l. Zhang, G. He, X.-y. Liu, Sci. Rep. 6 (2016) 24134. [677] A. Sadighi, A. Eyvazian, M. Asgari, A.M. Hamouda, Thin-walled Struct. 145
[620] S. Ray, U. Thormann, M. Eichelroth, M. Budak, C. Biehl, M. Rupp, U. Sommer, (2019), 106418.
T. El Khassawna, F.I. Alagboso, M. Kampschulte, M. Rohnke, A. Henß, K. Peppler, [678] H. Jiang, Y. Ren, J. Xiang, Thin-walled Struct. 124 (2018) 278–290.
V. Linke, P. Quadbeck, A. Voigt, F. Stenger, D. Karl, R. Schnettler, C. Heiss, K. [679] N. Gan, S. Yao, H. Dong, Y. Xiong, D. Liu, D. Pu, Thin-walled Struct. 122 (2018)
S. Lips, V. Alt, Biomaterials 157 (2018) 1–16. 147–157.
[621] J. Wang, Y. Li, G. Hu, M. Yang, Appl. Sci. 9 (2019) 5322. [680] A. Malekshahi, K.H. Shirazi, M. Shishehsaz, J. Mech. 35 (2019) 315–326.
[622] D.F.O. Braga, S.M.O. Tavares, L.F.M. da Silva, P.M.G.P. Moreira, P.M.S.T. de [681] A. Jusuf, L. Gunawan, T. Dirgantara, F. Mubarhak, in: A. Praveen Kumar,
Castro, Prog. Aerosp. Sci. 69 (2014) 29–39. T. Dirgantara, P.V. Krishna (Eds.), Advances in Lightweight Materials and
[623] J. Plocher, A. Panesar, Mater. Des. 183 (2019), 108164. Structures, Springer, Singapore, Singapore, 2020, pp. 31–41.
[624] J. Sobieszczanski-Sobieski, R.T. Haftka, Struct. Optim. 14 (1997) 1–23. [682] J. Cui, R. Ye, N. Zhao, J. Wu, M. Wang, Thin-walled Struct. 130 (2018) 520–534.
[625] L. Zhu, T.-S. Liu, T.-T. Duan, T.-T. Li, F. Qiu, H.-Y. Yang, Z.-H. Bai, Y.-Y. Liu, Q.- [683] N.S. Ha, G. Lu, X. Xiang, J. Mater. Sci. 54 (2019) 6286–6300.
C. Jiang, Mater. Des. 181 (2019), 107945. [684] D. Hu, Y. Wang, B. Song, L. Dang, Z. Zhang, Compos. Part B Eng. 162 (2019)
[626] J. Vitins, Transp. Res. Rec. 2289 (2012) 42–46. 21–32.
[627] S. Benson, J. Downes, R.S. Dow, Mar. Struct. 31 (2013) 44–62. [685] S. Zhu, L. Ma, B. Wang, J. Hu, Z. Zhou, Mater. Today Commun. 14 (2018)
[628] S. Shu, Y. Jin, B. Zhang, L. Chen, H. Yang, E. Dong, Int. J. Mod. Phys. B 33 (2019), 273–281.
1940048. [686] J. Mueller, K. Shea, Extreme Mech. Lett. 25 (2018) 7–15.
[629] J. Wu, N. Aage, R. Westermann, O. Sigmund, IEEE Trans. Vis. Comput. Graph. 24 [687] A. Suzuki, K. Sekizawa, M. Liu, N. Takata, M. Kobashi, Adv. Eng. Mater. 21
(2018) 1127–1140. (2019), 1900571.
[630] A. Korompili, Q. Wu, H. Zhao, Renewable Sustainable Energy Rev. 59 (2016) [688] Y. Shen, S. McKown, S. Tsopanos, C.J. Sutcliffe, R.A.W. Mines, W.J. Cantwell,
1405–1414. J. Sandw. Struct. Mater. 12 (2009) 159–180.
[631] F. Xiong, D. Wang, S. Zhang, K. Cai, S. Wang, F. Lu, Struct. Multidiscip. Optim. 57 [689] J.A. Harris, R.E. Winter, G.J. McShane, Int. J. Impact Eng. 104 (2017) 177–191.
(2018) 441–461. [690] C.G. Ferro, S. Varetti, G. De Pasquale, P. Maggiore, Mater. Today Commun. 15
[632] Y. Li, Z. Lin, A. Jiang, G. Chen, Mater. Des. 24 (2003) 177–182. (2018) 185–189.
[633] H. Yang, E. Dong, B. Zhang, Y. Yuan, S. Shu, Scanning 2017 (2017), 9314740. [691] R. Vrána, O. Červinek, P. Maňas, D. Koutný, D. Paloušek, Materials 11 (2018)
[634] F. Birk, F. Ali, M. Weigold, E. Abele, K. Schützer, Int. J. Adv. Manuf. Technol. 108 2129.
(2020) 3915–3924. [692] H. Zhang, H. Zhou, Z. Zhou, H. Zeng, X. Zhang, J. Yang, H. Lei, F. Han, Int. J.
[635] M. Kurpinska, B. Grzyl, A. Kristowski, Materials 12 (2019) 3629. Solids Struct. 226-227 (2021), 111082.
[636] A. Salim, S. Ghosh, S. Lim, Sensors 18 (2018) 3049. [693] H. Lei, C. Li, J. Meng, H. Zhou, Y. Liu, X. Zhang, P. Wang, D. Fang, Mater. Des.
[637] H. Yang, P. Ji, C. Zhan, Y. Jin, S. Shu, Int. J. Mod. Phys. B 33 (2019), 1940051. 169 (2019), 107685.
[638] X. Dong, P. Li, S. Amirkhanlou, S. Ji, P.S. Popel, U. Dahlborg, M. Calvo-Dahlborg, [694] R.S. Lakes, K. Elms, J. Compos. Mater. 27 (1993) 1193–1202.
Sci. Rep. 10 (2020) 12979. [695] A. Alderson, K.L. Alderson, Proc. Inst. Mech. Eng. G J. Aerosp. Eng. 221 (2007)
[639] D. Kumar, R.K. Phanden, L. Thakur, Mater. Today Proc. 38 (2021) 359–364. 565–575.
[640] W.A. Counts, M. Friák, D. Raabe, J. Neugebauer, Acta Mater. 57 (2009) 69–76. [696] J.B. Choi, R.S. Lakes, Int. J. Fract. 80 (1996) 73–83.
[641] J.S. Weaver, S.R. Kalidindi, U.G.K. Wegst, Acta Mater. 132 (2017) 182–192. [697] J. Zhang, G. Lu, Int. J. Impact Eng. 139 (2020), 103497.
[642] F. Qiu, T.-S. Liu, X. Zhang, F. Chang, S.-L. Shu, H.-Y. Yang, Q.-L. Zhao, Q.- [698] L. Yang, O. Harrysson, H. West, D. Cormier, Int. J. Solids Struct. 69-70 (2015)
C. Jiang, China Foundry 17 (2020) 111–126. 475–490.
[643] B. Wang, G.C. Barber, F. Qiu, Q. Zou, H. Yang, J. Mater. Res. Technol. 9 (2020) [699] H.N.G. Wadley, Adv. Eng. Mater. 4 (2002) 726–733.
1054–1069. [700] S. Chiras, D.R. Mumm, A.G. Evans, N. Wicks, J.W. Hutchinson, K. Dharmasena, H.
[644] W. Zhang, L. Zhao, S. Cai, Comput. Methods Appl. Mech. Eng. 294 (2015) N.G. Wadley, S. Fichter, Int. J. Solids Struct. 39 (2002) 4093–4115.
359–383.

55
L.-Y. Chen et al. Materials Science & Engineering R 146 (2021) 100648

[701] T. George, V.S. Deshpande, K. Sharp, H.N.G. Wadley, Compos. Struct. 108 (2014) [742] C. Qiu, S. Yue, N.J.E. Adkins, M. Ward, H. Hassanin, P.D. Lee, P.J. Withers, M.
696–710. M. Attallah, Mater. Sci. Eng. A 628 (2015) 188–197.
[702] R. Gümrük, R.A.W. Mines, S. Karadeniz, Mater. Sci. Eng. A 586 (2013) 392–406. [743] D.-G. Ahn, Int. J. Precis. Eng. Manuf. Technol. 8 (2021) 703–742.
[703] J. Zheng, L. Zhao, H. Fan, Compos. Part B Eng. 43 (2012) 1516–1522. [744] D.-S. Shim, G.-Y. Baek, J.-S. Seo, G.-Y. Shin, K.-P. Kim, K.-Y. Lee, Opt. Laser
[704] H. Fan, Z. Qu, Z. Xia, F. Sun, Mater. Des. 58 (2014) 363–367. Technol. 86 (2016) 69–78.
[705] Z. Niu, N. Becknell, Y. Yu, D. Kim, C. Chen, N. Kornienko, G.A. Somorjai, P. Yang, [745] J.C. Wang, Y.J. Liu, P. Qin, S.X. Liang, T.B. Sercombe, L.C. Zhang, Mater. Sci. Eng.
Nat. Mater. 15 (2016) 1188–1194. A 760 (2019) 214–224.
[706] X. Li, T. Fan, H. Zhou, S.-K. Chow, W. Zhang, D. Zhang, Q. Guo, H. Ogawa, Adv. [746] S. Wu, Z. Luo, Z. Li, S. Liu, L.-C. Zhang, Comput. Methods Appl. Mech. Eng. 377
Funct. Mater. 19 (2009) 45–56. (2021), 113708.
[707] S.-X. Liang, Q. Zhang, Z. Jia, W. Zhang, W. Wang, L.-C. Zhang, J. Colloid Interface [747] Z. Li, Z. Luo, L.-C. Zhang, C.-H. Wang, Mater. Des. 202 (2021), 109523.
Sci. 581 (2021) 860–873.
[708] S.-X. Liang, S. Salamon, S. Zerebecki, L.-C. Zhang, Z. Jia, H. Wende,
S. Reichenberger, S. Barcikowski, Scr. Mater. 203 (2021), 114094.
Liang-Yu Chen is an Associate Professor in the School of Ma­
[709] S.-X. Liang, L.-C. Zhang, S. Reichenberger, S. Barcikowski, Phys. Chem. Chem.
Phys. 23 (2021) 11121–11154. terial Science and Engineering at Jiangsu University of Science
[710] S.X. Liang, Z. Jia, Y.J. Liu, W. Zhang, W. Wang, J. Lu, L.C. Zhang, Adv. Mater. 30 and Technology. He received his PhD degree from Shanghai
(2018), 1802764. Jiaotong University in 2016. He has been a visit professor in
[711] S.-X. Liang, W. Zhang, L. Zhang, W. Wang, L.-C. Zhang, Sustain. Mater. Technol. school of in Edith Cowan University during 2018. He is an
associated editor in Frontiers in Materials. His current research
22 (2019), e00126.
[712] Z. Jia, K. Nomoto, Q. Wang, C. Kong, L. Sun, L.-C. Zhang, S.-X. Liang, J. Lu, J. interest focuses on the metal additive manufacturing, plasma
J. Kruzic, Adv. Funct. Mater. 31 (2021), 2101586. sprayed coatings, and corrosion science of metallic materials.
[713] Q. Zhang, S.-X. Liang, Z. Jia, W. Zhang, W. Wang, L.-C. Zhang, J. Mater. Sci. Up to date, he has published over 50 refereed international
Technol. 61 (2021) 159–168. journal papers (high-cited paper = 2) and h-index of 25.
[714] X. Liu, R. Jervis, R.C. Maher, I.J. Villar-Garcia, M. Naylor-Marlow, P.R. Shearing,
M. Ouyang, L. Cohen, N.P. Brandon, B. Wu, Adv. Mater. Technol. 1 (2016),
1600167.
[715] A. Ambrosi, J.G.S. Moo, M. Pumera, Adv. Funct. Mater. 26 (2016) 698–703.
[716] Z. Jia, Q. Wang, L. Sun, Q. Wang, L.-C. Zhang, G. Wu, J.-H. Luan, Z.-B. Jiao,
A. Wang, S.-X. Liang, M. Gu, J. Lu, Adv. Funct. Mater. 29 (2019), 1807857. Shunxing Liang received his PhD degree in Materials &
[717] S.X. Liang, Z. Jia, W.C. Zhang, X.F. Li, W.M. Wang, H.C. Lin, L.C. Zhang, Appl. Chemistry at Edith Cowan University (ECU), Australia, in the
Catal. B. Environ. 221 (2018) 108–118. group of Prof. Lai-Chang Zhang in 2020. Currently, he is
[718] Z. Jia, X. Duan, P. Qin, W. Zhang, W. Wang, C. Yang, H. Sun, S. Wang, L.C. Zhang, working as Alexander von Humboldt Research Fellow at the
Adv. Funct. Mater. 27 (2017), 1702258. University of Duisburg-Essen (UDE). Since he started his
[719] L.C. Zhang, Z. Jia, F. Lyu, S.X. Liang, J. Lu, Prog. Mater. Sci. 105 (2019), 100576. research in 2016, he has published 23 refereed international
[720] S.-X. Liang, W. Zhang, W. Wang, G. Jia, W. Yang, L.-C. Zhang, J. Phys. Chem. journal papers. His research focuses on advanced catalytic
Solids 132 (2019) 89–98. function of metallic glasses and nanocrystallization, as well as
[721] Z. Jia, J.C. Wang, S.X. Liang, W.C. Zhang, W.M. Wang, L.C. Zhang, J. Alloys. pulsed laser synthesis and processing of functional amorphous
Compd. 728 (2017) 525–533. metal nanoparticles.
[722] Z. Yang, F. Gao, X. Shi, X. Lin, Z. Gao, Y. Chong, B. Zhang, Phys. Rev. Lett. 114
(2015), 114301.
[723] H. Xue, Y. Yang, G. Liu, F. Gao, Y. Chong, B. Zhang, Phys. Rev. Lett. 122 (2019),
244301.
[724] Y. Chen, X. Liu, G. Hu, Sci. Rep. 5 (2015) 15745.
[725] G. Liao, C. Luan, Z. Wang, J. Liu, X. Yao, J. Fu, Adv. Mater. Technol. 6 (2021),
2000787. Yujing Liu a Professor of Institute of Metals, College of Mate­
[726] Y. Chen, B. Zhao, X. Liu, G. Hu, Extreme Mech. Lett. 40 (2020), 100916. rial Science and Engineering, Changsha University of Science &
[727] X. Sun, F. Jiang, J. Wang, Metals 10 (2020) 111. Technology. His current research interest focuses on the addi­
[728] T.G. Zieliński, K.C. Opiela, P. Pawłowski, N. Dauchez, T. Boutin, J. Kennedy, tive manufacturing via selective laser melting and electron
D. Trimble, H. Rice, B. Van Damme, G. Hannema, R. Wróbel, S. Kim, S. Ghaffari beam melting technology, aiming to understand the structure-
Mosanenzadeh, N.X. Fang, J. Yang, B. Briere de La Hosseraye, M.C.J. Hornikx, process-property relationship of 3D printable metallic mate­
E. Salze, M.-A. Galland, R. Boonen, A. Carvalho de Sousa, E. Deckers, M. Gaborit, rials and explore their practical applications in aerospace and
J.-P. Groby, Addit. Manuf. 36 (2020), 101564. biomedical components, etc.
[729] S. Kuschmitz, T.P. Ring, H. Watschke, S.C. Langer, T. Vietor, Materials 14 (2021)
1747.
[730] Y. Ibrahim, Z. Li, C.M. Davies, C. Maharaj, J.P. Dear, P.A. Hooper, Addit. Manuf.
24 (2018) 566–576.
[731] N. Lebaal, A. SettaR, S. Roth, S. Gomes, Mech. Adv. Mater. Struct. 28 (2020) 1–9.
[732] K.N. Son, J.A. Weibel, V. Kumaresan, S.V. Garimella, Int. J. Heat Mass Transf. 115
(2017) 619–629.
[733] J. Gullbrand, K. Balantrapu, R. Wirtz, Thermal Characterization of Open Lattice
Structures Used as Heat Exchanger Surfaces, in: 43rd AIAA Aerospace Sciences Prof. Lai-Chang ZHANG received his PhD in Materials Science
Meeting and Exhibit, American Institute of Aeronautics and Astronautics, 2005. & Engineering at Institute of Metal Research, Chinese Academy
[734] K.J. Maloney, K.D. Fink, T.A. Schaedler, J.A. Kolodziejska, A.J. Jacobsen, C. of Sciences, China, in 2005. After that, he was an Alexander
S. Roper, Int. J. Heat Mass Transf. 55 (2012) 2486–2493. von Humboldt Research Fellow at both Darmstadt University
[735] X. Bai, Z. Zheng, A. Nakayama, Int. J. Heat Mass Transf. 143 (2019), 118525. of Technology and Leibniz-Institute for Solid State and Mate­
[736] M. Alsulami, M. Mortazavi, S.A. Niknam, D. Li, Int. J. Adv. Manuf. Technol. 110 rials Research Dresden (IFW Dresden), Germany until 2007.
(2020) 865–873. From 2007–2009, he was a Research Fellow at University of
[737] R.P.P. da Silva, M.V.V. Mortean, K.V. de Paiva, L.E. Beckedorff, J.L.G. Oliveira, F. Wollongong (Australia). Since September 2009, he had been a
G. Brandão, A.S. Monteiro, C.S. Carvalho, H.R. Oliveira, D.G. Borges, V. Research Assistant Professor at The University of Western
L. Chastinet, Appl. Therm. Eng. 193 (2021), 116973. Australia (Australia) until June 2012. Afterward, he was an
[738] S.A. Niknam, M. Mortazavi, D. Li, Int. J. Adv. Manuf. Technol. 112 (2021) Associate Professor (2012–2018) and Professor (since 2018) &
601–618. Program Leader Mechanical Engineering at Edith Cowan Uni­
[739] J.Y. Ho, K.C. Leong, T.N. Wong, Int. J. Heat Mass Transf. 150 (2020), 119262. versity (Australia). His research interests include metal additive manufacturing, alloy
[740] M. Wong, I. Owen, C.J. Sutcliffe, A. Puri, Int. J. Heat Mass Transf. 52 (2009) development, light-weight alloys, nanocrystalline materials and metallic glasses, and
281–288. nanomaterials for water treatment. He has published about 300 international journal
[741] J.R. McDonough, Therm. Sci. Eng. Prog. 19 (2020), 100594. papers with a total of 27 ESI Highly Cited or Hot Papers and his H-index is 60 in Scopus.

56