1 s2.0 S1359028617301870 Main

Current Opinion in Solid State & Materials Science 22 (2018) 75–99
Contents lists available at ScienceDirect
Current Opinion in Solid State & Materials Science

journal homepage: www.elsevier.com/locate/cossms
Selective electron beam manufactured Ti-6Al-4V lattice structures for T

orthopedic implant applications: Current status and outstanding challenges
⁎ ⁎
X.Z. Zhanga,b,c, M. Learyc, H.P. Tangb, , T. Songc, M. Qianb,c,
a
School of Materials Science and Engineering, Northeastern University, Shenyang 110819, China
b
State Key Laboratory of Porous Metal Materials, Northwest Institute for Nonferrous Metal Research, Xi’an 710016, China
c
Centre for Additive Manufacturing, School of Engineering, RMIT University, Melbourne, VIC 3000, Australia
A R T I C LE I N FO A B S T R A C T
Keywords: Additively manufactured Ti-6Al-4V lattices display unique mechanical and biological properties by virtue of
Additive manufacturing their engineered structure. These attributes enable the innovative design of patient-specific medical implants
Ti-6Al-4V that (i) are conformal to the intended surgical geometry, (ii) mimic the mechanical properties of natural bone,
SEBM and (iii) provide superior biological interaction to traditional implants. Selective electron beam melting (SEBM)
Electron beam
is an established metal additive manufacturing (AM) process that has enabled the design and fabrication of a
Lattice structure
variety of novel intricate lattices for implant applications over the last 15 years. This article reviews the technical
Implants
Titanium and clinical characteristics of SEBM Ti-6Al-4V lattices, including (i) the SEBM process and its capabilities, (ii) the
structures of human bones with an exhaustive list of corresponding mechanical properties from literature, (iii)
the mechanical properties of SEBM Ti-6Al-4V lattices of various designs and their shortcomings when compared
to human bones, (iv) microstructural control of SEBM Ti-6Al-4V lattices for improved performance, (v) the
lattice manufacturability and associated geometric errors, and (vi) clinical cases. Existing literature on the
mechanical response of SEBM Ti-6Al-4V lattice structures is exhaustively evaluated for documentation quality
using established theoretical models. This extensive data-set allows novel insights into the effect of lattice design
on mechanical response that is not possible with the individual data; and provides a comprehensive database for
those who are actively involved in patient-specific SEBM implant design. On this basis, outstanding challenges
and research opportunities for SEBM Ti-6Al-4V lattices in the biomedical domain are identified and discussed.
1. Introduction made up of connected struts or plates [3]. The unit cell size can vary
from nanometers to centimeters. Lattices should contain no isolated
Metallic implants help improve the quality of life and longevity of pores while having high porosity (e.g., > 50% or even 70%) [4].
human beings. Among all the applications of metallic implants, the However, additive manufacturing (AM) can produce lower porosity
requirements for spinal, hip and knee replacements are significant [1]. lattices without containing isolated voids. Lattice design provides an
The actual clinical application determines the property requirements effective way of relegating the elastic modulus and strength of a solid
for implants, e.g., femoral implants call for superior compressive and material, as well as an innovative biological form of fixation by pro-
tensile strengths while dental implants require high fatigue life and moting bone tissue ingrowth into its open-celled structures [2]. In ad-
wear resistance. In addition, implants are expected to serve a lifetime dition, lattice-like open cellular structures are more conductive to in-
without failure or revision surgery. One challenge is that the elastic serting antibiotics to eliminate or control infection [5,6].
modulus of most current metallic implants is much higher than that of Table 1 summarizes the 23 most common metallic lattices assessed
the bone to be replaced, e.g., the elastic modulus of human bone varies to date with structural details. The strut of a lattice can be circular,
from 0.02–40 GPa compared to 110 GPa for commercially pure tita- triangular, and hexagonal. Lattices present fundamentally different
nium (CP-Ti), 190–210 GPa for Co-Cr alloys, and 210–253 GPa for 316L mechanical and biological behavior from their dense parent materials
stainless steel [2]. This discrepancy can entail serious stress shielding [51] and have found a variety of applications for light weighting and
[1]. energy absorption [3,52]. In particular, they have been widely used as
Lattices can be defined as three-dimensional (3D) open-celled bone implants [53–57] because they can enable replication of both the
structures formed by the arrangement of a repeating typical unit cell biological and mechanical properties of those naturally occurring
⁎
Corresponding authors.
E-mail addresses: thpfys@126.com (H.P. Tang), ma.qian@rmit.edu.au (M. Qian).
https://doi.org/10.1016/j.cossms.2018.05.002
Received 5 October 2017; Received in revised form 30 March 2018; Accepted 13 May 2018
Available online 22 May 2018
1359-0286/ © 2018 Elsevier Ltd. All rights reserved.
X.Z. Zhang et al. Current Opinion in Solid State & Materials Science 22 (2018) 75–99
Nomenclature TLIF transforaminal lumbar interbody fusion

d strut diameter
AM additive manufacturing UTS ultimate tensile strength
SEBM selective electron beam melting TYS tensile yield strength
BCC body centered cubic UCS ultimate compressive strength
SLM selective laser melting CYS compressive yield strength
BCCZ body centered cubic with vertical struts C, C1, C2 constant
FCC face centered cubic E elastic modulus of the cellular solid
FCCZ face centered cubic with vertical struts Es elastic modulus of the pore-free solid
FBCCZ face and body centered cubic with vertical struts ρ density of the cellular solid
FBCCXYZ face and body centered cubic with horizontal and vertical ρs density of the pore-free solid
struts σ collapse strength of the cellular solid
ALIF anterior lumbar interbody fusion σy,s yield strength of the pore-free solid
PLIT posterior lumbar interbody fusion
cellular materials. Fig. 1 shows some recent United States (US) Food ∼15000 mm/s; beam current: ∼40 mA; acceleration voltage:
and Drug Administration (FDA)-approved Ti lattices, of which the 60 kV) is then used to preheat the entire powder bed (Preheat I) and
Emerging Implant Technologies (EIT) Ti lattice interbody cages have subsequently the selected area (Preheat II) (Fig. 2a).
been used in over 10,000 cases in 15 countries [58]. Consistent with the (3) Selective melting: a much slower scanning speed (∼4500 mm/s)
23 different unit cells illustrated in Table 1, the lattice struts in the and a smaller beam current (∼20 mA) are used for selective
implants shown in Fig. 1 were designed and built in a variety of dia- melting, where a contour strategy (a Multibeam™ technology) is
meters and inclination angles. As will be discussed in Section 6.1 Lattice used to fuse the perimeter of each layer section before a hatching
manufacturability by SEBM, the minimum strut diameter or inclination strategy fills in the section area (Fig. 2b) [71]. The Multibeam™
angle that can be built with high quality reflects the manufacturability technology allows multi- melt pools to be maintained concurrently
of the SEBM process. in order to improve surface finish [70,72], e.g., 50 melt pools for
As shown in Table 1, research on lattices has focused on Ti alloys outer contour and 10 for inner contours [71]. Hatching is used to
and more specifically on Ti-6Al-4V. This is because Ti alloys are the melt the lattice area using a “snaking” and/or “point” strategy
material of choice for most medical and dental applications today [64]. chosen automatically according to the strut diameter (d) (e.g.,
AM allows the cost-effective fabrication of patient-specific lattices. In “snaking” for d > 1 mm). To avoid local overheating of the hot
particular, selective electron beam melting (SEBM) is most suited to the region (see Fig. 2b from ② to ③), the hatching track turns on au-
AM of reactive metals such as Ti because of its high vacuum printing tomatically by increasing the beam speed to keep a constant energy.
chamber [65]. In addition, SEBM offers higher production rates Besides, the hatching speed will increase during melting of the
(80 cm3·h−1) than selective laser melting (SLM) (20–40 cm3·h−1) and overhanging part to compensate for the insulating effect of the
more flexibility in powder size selection (up to 180 µm but usually powder bed [73]. In addition, in the hatching process of lattices
45–105 µm vs. 20–45 µm for SLM). This difference in size range cor- (Fig. 2b), the line offset is often set to be greater than that used for a
responds to a noticeable difference in affordability for Ti powder. An- bulk part (e.g., 0.2 mm vs. 0.1 mm at the layer thickness of 50 µm).
other important advantage is that owing to the high powder bed tem- For SEBM of a bulk part, in order to achieve overlap between the
perature, SEBM Ti-6Al-4V lattices or parts do not normally need post hatching lines, the line offset should be smaller than the beam spot
stress-relief-annealing [66,67]. In fact, the world’s first AM metal (Ti- size. However, since lattices contain a lot of voids and d is small, the
6Al-4V) acetabula cup, named the Delta-TT Cup, for use in a hip re- use of a large line offset is more efficient. The beam spot size in the
placement was produced using the SEBM process in 2007 by Lima- “point” strategy is larger than that used in the “snaking” strategy; as
Corporate and Arcam and has shown great success after 10 years of a result the resulting surface finish of the lattice struts is rougher.
surgery [68]. The surface of the Ti-6Al-4V Delta-TT Cup had a lattice- (4) Lowering the platform for spreading the next layer of powder: After
like feature. SEBM of the current layer, the platform is lowered by 50–150 µm
This paper provides a comprehensive review of all major aspects of for spreading the next layer of Ti-6Al-4V powder.
SEBM Ti-6Al-4V lattices for medical applications and identifies research
and clinical opportunities that can expand their applications. In addi- These four steps are repeated until the entire part is built. A finished
tion, available data on the mechanical performance of both SEBM Ti- part is often cooled in the powder bed to room temperature where air is
6Al-4V lattices and human bones are summarized in detail, with a view flooded only at ≤100 °C. It is typically enveloped by a shell of lightly
to providing a database for those who are actively involved in patient- sintered powder particles, most of which can be removed by com-
specific implant design. pressed air for recycling after sieving.
2. The SEBM process of Ti-6Al-4V lattices

3. Properties of human bones
The SEBM process was patented by Arcam AB in 1997 [69]. Today
In order to design a functional lattice implant, it is imperative to
Arcam SEBM systems include Models of S12, A1, A2, A2x, and A2xx
comprehend the properties of human bones to be integrated with.
(tungsten cathode) and Models of Q10, Q10 plus, Q20, and Q20 plus
Human bones have a varied arrangement of structures at different
(LaB6 cathode) [69]. A typical SEBM process of Ti-6Al-4V lattices
length scales which work in concert to perform diverse mechanical,
consists of four cyclic steps:
biological and chemical functions [75]. In general, human bones are
hierarchically structured (Fig. 3 [76]) and they derive their function
(1) Spreading a layer of powder: The first layer of Ti-6Al-4V powder is
through such a hierarchical organization. The hierarchical levels are
spread on a stainless steel or titanium platform that is preheated to
composed of (i) the macrostructure (e.g., cortical and cancellous bone);
∼730 °C [70].
(ii) the microstructure ranging from 10 µm to 500 µm (e.g., haversian
(2) Preheating of powder: A defocused electron beam (scanning speed:
systems, osteons, single trabecular); (iii) the fine-microstructure in the
76
Table 1
Additively manufactured metal lattice structures from literature. (See below-mentioned references for further information.)
Lattice structure unit cell Lattice structure Material AM methods Ref.
Auxetic Ti-6Al-4V SEBM [7-12]
BCC Ti-6Al-4V SLM [ 13]
Al-12Si-Mg SLM [14]
BCCZ Al-12Si-Mg SLM [14]
Cross 1 Auxetic Auxetic Ti-6Al-4V SEBM [ 15]
Cubic Ti-6Al-4V SEBM [16-19]
Ti-6Al-4V SLM [13, 20, 21]
316L stainless steel SLM [22, 23]
BCC BCCZ
BCC BCCZ Pure Titanium SLM [24]
Al-12Si-Mg SLM [14]
Diamond Ti-6Al-4V SEBM [25-28]
Ti-6Al-4V SLM [20, 29-31]
Cross 1 Cubic
Pure Ta SLM [32]
Cross 1 Cubic
Dode thin Ti-6Al-4V SEBM [15]
Co-29Cr-6Mo SEBM [33]
Ni-21Cr-9Mo SEBM [33]
Diamond Dode thin
Dodecahedron Diamond Dode thin Ti-6Al-4V SEBM [34-36]
Ti2448 SEBM [37]
(Ti-23.9Nb-3.9Zr-
8.2Sn-0.19O)
Dodecahedron Ti-6Al-4V SLM [38-40]
FCC Al-12Si-Mg SLM [14]
FCCZ Ti-6Al-4V SLM [ 13]
Dodecahedron G7 Al-12Si-Mg SLM [14]
FCC FCCZ
FBCCZ Ti-6Al-4V SLM [13]
Al-12Si-Mg SLM [14]
FBCCXYZ FCC FCCZ Ti-6Al-4V SLM [13]
G7 Ti-6Al-4V SEBM [15, 16, 41]
FBCCZ FBCCXYZ
Gyroid Ti-6Al-4V SLM [ 31]
316L stainless steel SLM [42]
Hatched FBCCZ FBCCXYZ Ti-6Al-4V SEBM [27]
Hexagonal G7 Gyroid Ti-6Al-4V SEBM [43, 44]
Ti-6Al-4V SLM [45, 46]
Gyroid Hatched Pure Titanium SLM [45]

Honeycomb Ti-6Al-4V SEBM [47]
Hexagonal Honeycomb
Octahedron Ti-6Al-4V SEBM [ 48]
Octet-truss Ti-6Al-4V SEBM [ 49]
T e t r ah e d r o n Ti-6Al-4V SLM [ 50]
Hexagonal Honeycomb
Triangular Ti-6Al-4V SLM [40]
Truncated Octahedron Octet-truss Ti-6Al-4V SLM [20]
cuboctahedron
Rectangular Ti-6Al-4V SLM [40]
Octet-truss Tetrahedron
Tetrahedron
Truncated cuboctahedron
Truncated cuboctahedron
77
Fig. 1. Examples of FDA-approved Ti-6Al-4V lattice implants. (a) By EIT, Germany [58]. (b) By Renovis, USA [59]. (c) By 4WEB Medical, USA [60]. (d) By Joimax,
Germany [61]. (e) By K2M, USA [62]. (f) By Solidscape, USA [63].
Fig. 2. (a) Preheating strategy [70,74]. (b) Melting strategy [71].
range of 1–10 µm (e.g., lamellae), (iv) the sub-microstructure (e.g., fi- (1) The ultimate strength and elastic modulus of cortical bone are both
brillar collagen and embedded mineral), and (v) the nanostructure that much higher than the respective properties of cancellous bone. In
is below a few hundred nanometers (e.g., molecular structure of con- addition, different parts of the cortical bone do not normally show
stituent elements, such as mineral, collagen, and non-collagenous or- the same mechanical properties, e.g., both the tensile and com-
ganic proteins). pressive properties of tibial are clearly higher than those of femur
Tables 2–4 exhaustively summarize relevant literature data on the (Tables 2 and 3). In that regard, Ti-6Al-4V lattices with graded
age- and gender-related mechanical properties of human bones, in- properties are necessary. Most property data listed in Tables 2 and 3
cluding a statistical summary of observed values for both cortical and focus on tensile properties as clinical experiences and experimental
cancellous bones. In analyzing the strength data, it is useful to also evidence indicate that most fracture failures of cortical bone arise
consider the density or porosity of each category of bone for corre- from tensile stresses and strains caused by bending or torsion [78].
sponding lattice design. The following general observations are most The cortical bone is subjected to both tensile and compressive
notable: stresses and strains during bending or torsion. Failure or fracture is
78
Fig. 3. Hierarchical structures of human cortical and cancellous bone [76,77].
initiated in the area of the bone under tension because the tensile but there is no distinct difference.
strength of the bone is considerably lower than its compressive (3) Most human bones are moist during mechanical testing except for a
strength [79]. few that are specified in Table 4. By comparing the mechanical
(2) The tensile properties of cortical fibular and femoral bone degrade properties of vertebra obtained in dry and wet conditions, the dry
with increasing age, but the tibia shows no regular change in tensile bone was found to be stiffer, stronger and considerably more brittle.
properties with increasing age except for the ultimate strain The compressive yield strength of both cortical and cancellous
(Table 2). The ultimate tensile and compressive strains of human bones is not easy to obtain by experiment and hence no data have
bones decrease with increasing age. Some differences do exist in the been listed in Tables 3 and 4.
mechanical properties of human bones between female and male, (4) The ultimate compressive strength and strain of cancellous tibia
Table 2
Tensile properties and test conditions of human cortical bone [2].
Bone Age TYS (MPa) UTS (MPa) Elastic modulus Ultimate strain Density Source of bone (moist) Cross head speed (mm/ Ref.
(GPa) (%) (g·cm−3) min)
Fibula 33 male –a 100 19.20 2.10 1.91 Embalmed body 1.10 [78]
59 female – 80 15.20 1.19 1.73 Embalmed body 1.10 [78]
Humerus 15–89 male – 149 15.60 1.20 1.77 – 0.50 [80,81]

15–89 – 151 16.10 1.90 1.72 – 0.50 [80,81]
female
Tibia 41.5 male – 106 18.90 1.76 1.96 Embalmed body 1.10 [78]
71 male – 84 16.20 1.56 1.83 Embalmed body 1.10 [78]
20–29 126 161 18.90 4.00 1.83b Frozen – [79]
30–39 129 154 27.00 3.90 1.83b Frozen – [79]
40–49 140 170 28.80 2.90 1.83b Frozen – [79]
50–59 133 164 23.10 3.10 1.83b Frozen – [79]
60–69 124 147 19.90 2.70 1.83b Frozen – [79]
70–79 120 145 19.90 2.70 1.83b Frozen – [79]
80–89 131 156 29.20 2.30 1.83b Frozen – [79]
Femur 41.5 male – 102 14.90 1.32 1.91 Embalmed body 1.10 [78]
71 male – 68 13.60 1.07 1.85 Embalmed body 1.10 [78]
15–89 male – 141 15.20 2.00 1.90 – 0.50 [80,81]
15–89 – 134 15.00 1.80 1.80 – 0.50 [80,81]
female
20–29 120 140 17.00 3.40 1.85c Frozen – [79]
30–39 120 136 17.60 3.20 1.85c Frozen – [79]
40–49 121 139 17.70 3.00 1.85c Frozen – [79]
50–59 111 131 16.60 2.80 1.85c Frozen – [79]
60–69 112 129 17.10 2.50 1.85c Frozen – [79]
70–79 111 129 16.30 2.50 1.85c Frozen – [79]
80–89 104 120 15.60 2.40 1.85c Frozen – [79]
Note:
a
–: not available.
b
Taken as 1.83 g·cm−3 based on Ref. [78].
c
Taken as 1.85 g·cm−3 based on Ref. [78].
79
Table 3 (UTS) > 170 MPa, tensile yield strength (TYS) > 140 MPa and
Compressive strength and elastic modulus of human cortical bone [79]. tensile strain > 2.9% (as shown bold in Table 2) and ultimate
Bone Age UCS Elastic modulus Density Source of bone compressive strength (UCS) > 213 MPa (as shown bold in Table 3)
(MPa) (GPa) (g·cm−3) (moist) at the same time. Besides, their elastic modulus should be lower
than or similar to 28.8 GPa to match that of the tibia bone [79].
Tibia 20–29 –a – 1.83b Frozen
(2) Cortical femurs exhibit not only good tensile strengths (e.g.
30–39 213 35.30 1.83b Frozen
40–49 204 30.60 1.83b Frozen
UTS > 140 MPa, TYS > 120 MPa) but also impressive tensile
50–59 192 24.50 1.83b Frozen strain (e.g. > 3%, Table 2). It can be challenging for SEBM Ti-6Al-
60–69 183 25.10 1.83b Frozen 4V lattice implants to replace them when elastic modulus
70–79 183 26.70 1.83b Frozen (15–30 GPa) match is also satisfied, which often requires high
80–89 197 25.90 1.83b Frozen
porosity of the lattice.
Femur 20–29 209 18.10 1.85c Frozen (3) At first glance, SEBM Ti-6Al-4V lattices can easily offer the me-
30–39 209 18.60 1.85c Frozen
chanical properties required by the highly porous human cancellous
40–49 200 18.70 1.85c Frozen
50–59 192 18.20 1.85c Frozen bone (density: 0.19–0.3 g·cm−3). For example, the highest UCS of
60–69 179 15.90 1.85c Frozen cancellous bone is 10.6 MPa with compressive strain of 2.48% and
70–79 190 18.00 1.85c Frozen elastic modulus of 0.65 GPa at the density of 0.25 g·cm−3 (bold in
80–89 180 15.40 1.85c Frozen Table 4). However, it should be pointed out that if density match is
also met, then it implies that the relative density of the SEBM Ti-
Note:
a
–: not available.
6Al-4V lattice should be only about 5.5%, namely 94.5% of por-
b
Taken as 1.83 g·cm−3 based on Ref. [78]. osity. Under such a substantially high porosity level, it can be
c
Taken as 1.85 g·cm−3 based on Ref. [78]. equally challenging for SEBM Ti-6Al-4V lattices to satisfy the me-
chanical properties of cancellous bone. Section 4 will further dis-
decrease with increasing age and there is almost no difference in cuss these issues.
elastic modulus between young and old people, which can simplify
the design of the elastic modulus issue in patient-specific lattices. 4. Mechanical properties of SEBM Ti-6Al-4V lattices
With an increase in bone density, the ultimate strength, strain and
elastic modulus of the cancellous bone show an increasing trend in A bone implant is expected to exhibit mechanical strengths higher
both dried and moist conditions. than, and elastic modulus similar to, that of the bone concerned. Its
(5) The anisotropic issue also needs to be fully understood. For ex- fatigue strength can determine the long-term success of the implant
ample, even though the lumbar spines show similar densities in under cyclic loading conditions. In that regard, a lattice structure with a
vertical and horizontal directions, the ultimate compressive balanced combination of strength (tensile, compressive or fatigue) and
strength and elastic modulus in the vertical direction are found to elastic modulus (close to that of the bone) should be considered in order
be much higher than that in the horizontal direction, while the to avoid loosening of the implants for long-term service. They are af-
ultimate compressive strain is the other way around. The aniso- fected by the parent material, lattice topology, relative density, local
tropic properties of human cancellous bone indicate that if a Ti-6Al- geometry, microstructure and defects [88].
4V lattice structure designed for implant applications is anisotropic,
e.g., those with a diamond or dode thin unit cell (Table 1), its
mechanical properties should also be made similarly anisotropic for 4.1. Compressive performance of SEBM Ti-6Al-4V lattices
mechanical matching.
The mechanical responses of SEBM Ti-6Al-4V lattices have been as-
These consolidated properties constitute a basis for the design of sessed under different conditions [11,12,17,28,34,35,41,44,55,89–91].
patient-specific implant lattices. Specifically, the following properties Fig. 4 shows the stress-strain response for various SEBM Ti-6Al-4V lattices.
are expected for SEBM Ti-6Al-4V lattices: The stress-strain curves of diamond, rhombic dodecahedral, cubic, auxetic
and G7 lattices (densities: 0.13–1.6 g cm−3) are similar to those of elastic-
(1) For cortical tibia, the SEBM Ti-6Al-4V lattice implants at the density brittle foam, while for G7 lattices with higher densities (1.8 g/cm3), their
of 1.83 g·cm−3 should exhibit ultimate tensile strength stress-strain curves are similar to those of elastic-plastic foam [4]. The
compressive stress-strain curve shown for the octet-truss lattice in Fig. 4f is
Table 4
Compressive properties and test conditions of human cancellous bone [2].
Bone Age CYS (MPa) UCS Elastic modulus Ultimate strain Density Source of Test condition of Cross head speed Ref.
(MPa) (GPa) (%) (g·cm−3) bone bone (mm/min)
Lumbar 14–89 male –a 4.60 0.06 6.70 0.20 – Dried 0.05 [82]
vertebra 14–89 female – 2.70 0.04 6.10 0.20 – Dried 0.05 [82]
Tibial head 14–89 male – 3.90 0.03 8.30 0.22 – Dried 0.05 [82]
14–89 female – 2.20 0.02 6.90 0.22 – Dried 0.05 [82]
Tibia 16–39 – 10.60 0.65 2.48 – Frozen Moist 0.02 [83]
40–59 – 9.86 0.83 2.12 – Frozen Moist 0.02 [83]
60–83 – 7.27 0.61 2.05 – Frozen Moist 0.02 [83]
Proximal tibia 59–82 – 5.33 0.45 – 0.29 Frozen Moist – [84]
Femur 58–83 – 7.36 0.39 – 0.50 Frozen Moist – [85]
Lumbar spine 15–87 (Vertical) – 2.45 0.07 7.40 0.25 Frozen Moist – [86]
15–87 – 0.88 0.02 8.50 0.24 Frozen Moist – [86]
(Horizontal)
71–84 1.37 1.55 0.02 7.40 0.19 Frozen Moist 5.00 [87]
Note:
a
–: not available.
80
Fig. 4. Nominal compressive stress-strain curves of SEBM Ti-6Al-4V lattices with different unit cells. (a) Diamond [28]. (b) Rhombic dodecahedron [41]. (c) Cubic
[41]. (d) G7 [41]. (e) Auxetic [10]. (f) Octet-truss [49]. The inset in (a) indicates Stages I, II and III.
obtained from successive loading/unloading up to nearly 5% of strain. curve of Stage I while yield strength is defined as the stress at which
The stress-strain curves shown in Fig. 4 display three common lattice is no longer elastically recoverable. In case the yield point cannot
stages, i.e., Stage I: linear elasticity, corresponding to unit cell strut be defined from the stress-strain curve, the yield strength of the lattice
bending or stretching; Stage II: stress plateau, corresponding to pro- structure can be determined from the stress-strain curve using the σ0.2
gressive unit cell collapse by plastic yielding, elastic buckling or brittle approach as has been used for a solid material (σ0.2, Fig. 4a). As can be
crushing, depending on unit cell structure and nature of the solid from seen from Fig. 4, the ultimate compressive strain of a Ti-6Al-4V lattice
which the lattice is made; and Stage III: densification [3,92,93]. During is often shown to be exceptional, e.g., up to 93% [35]. This can be
Stage III, the stress rises sharply and the opposite sides of the lattice misleading. In fact, the compressive strain at the first fracture plays a
structure are forced into contact and behave increasingly like a solid more important role in determining the mechanical performance of the
structure. The elastic modulus is the initial slope of the stress-strain lattice. In this regard, the compressive strain used in the following
81
analysis refers to that corresponding to the first compressive fracture in discussed in Section 4.3.
the lattice. The accumulated strain (ε) vs. fatigue cycle (N) and the associated
According to the compressive stress-strain curves shown in Fig. 4, stress plots are shown in Fig. 6(a–c) for three distinct lattice topologies
lattice response can be divided into bending (a, b, c, d and e) and with similar densities and strut diameters. In general, three stages are
stretch (f) dominated behavior. Stretch domination means that during observed [16]. Stage I is featured by an initial increase in strain within
compression, it is the struts carrying tension that fail first [3], while for just a few cycles (N < 10). This is followed by the plateau region of
bending-dominated lattice, the deformation of its struts during com- Stage II (also referred to as the “incubation” region) in which strain is
pression can occur through either bending, (for rhombic dodecahedron almost constant over a wide cycle range. In Stage III, a rapid strain in-
and G7 lattices) or buckling (for cubic lattices) [41]. The final failure is crease is observed associated with strut collapse and catastrophic
due to crush bands which are perpendicular or 45° to the compressive structural failure. However, with an increase in applied stress, the en-
direction [17,26,41,94]. A stretch-dominated lattice shows much durance strain increased at the cost of decreasing the number of cycles
higher UCS and elastic modulus than a bending-dominated lattice of the for the same lattice structures (Fig. 6a–c). This will be further discussed
same relative density [3,95]. In either case, fracture under compressive in Section 4.3.
loading mostly occurs from the nodes in lattices [16,41,47,96]. The fatigue properties are mainly determined by the interaction of
Fig. 5a–d shows the stress distribution in honeycomb and octahedron the cyclic ratcheting of the struts (the stage II incubation region) and the
lattices by simulation. Despite different unit cell sizes and shapes, the fatigue crack initiation and growth through the struts [17]. Cyclic
stress is always concentrated at the node, which undergoes plastic de- ratcheting, ∂ɛ/∂N, refers to the progressive accumulation of strain in
formation during compression [47]. lattices resulting from cyclic strut bending [16]. It depends on cell to-
pology and local geometry, e.g., the deformation of a cubic lattice
4.2. Fatigue performance of SEBM Ti-6Al-4V lattices structure is due primarily to buckling (Fig. 6d). SEBM Ti-6Al-4V cubic
lattices displayed the highest fatigue strength in Fig. 6e [17].
Resistance to fatigue crack initiation and growth is a critical design The elastic modulus of a lattice is related primarily to whether its
constraint for the applications of patient-specific metallic implants, lattice struts are in the state of bending or stretching (Section 4.1)
even though experimental data on the fatigue strength of different [100–103], where bending-dominated lattices, which show two types of
human bones are scare. To date, the fatigue performance of SEBM Ti- basic responses, i.e., bending or buckling (Section 4.4), tend to have
6Al-4V lattices has been investigated under different conditions lower elastic modulus than stretch-dominated lattices [41]. The mod-
[16,17,26,48,88,97–99]. Similar to the fatigue data reported on solid ulus of a lattice can be readily designed to match that of the bone (see
materials, many uncertainties exist and inconsistent fatigue results have Fig. 9c), many such low modulus Ti-6Al-4V lattices show poor fatigue
been reported for SEBM Ti-6Al-4V lattices as shown in Table 5. A strengths due to high porosity [20,38]. This shortcoming can be re-
comparison of their fatigue strengths with those of human bones will be garded as an important constraint on the design of lattices that intend
Fig. 5. Simulation results of stress distribution (a, b, c and d) and fatigue failure patterns (e and f) of SEBM Ti-6Al-4V lattices under compression with a unit cell of (a)
honeycomb-2.5 mm, (b) honeycomb-2.0 mm, (c) octahedron-2.0 mm (d) is a high magnification that shows stress concentrated at the node [47], (e) diamond and (f)
truncated cuboctahedron [88].
82
Table 5
Summary of compressive mechanical property test results for SEBM Ti-6Al-4V lattices in the as-built state without surface treatment.
Density Porosity (%) Lattice structure CYS (MPa) UCS (MPa) Elastic modulus Fatigue Compressive Strain Number of Dimensions Dimensions Composition Composition Reference
(g·cm−3) topology (GPa) strength strain (%) rate specimens (nominal) (measured) (nominal) (actual)
X.Z. Zhang et al.
(MPa) (s−1)
0.52 88.26 Cross 1 –a – 0.58b – – – – yc – y – [15]

1.31 70.32 Cubic – 83.13 ± 10.25 – – 4 – y y y y y [19]
1.31 70.43 Cubic 74.6 ± 3.5 83.8 ± 4.2 1.13 ± 0.04 – 15 10−3 y y – y – [106]
1.34 70.32 Cubic 83.13 ± 10.25 – 0.20 – – – y y y y – [18]
1.63 63.21 Cubic 155.9 ± 10.6 196 ± 7.1 14.90 ± 0.10b 105 (106) – 10−3 y y y y – [16]
1.75 60.41 Cubic – 117.05 ± 5.54 – – 3.5 – y y y y y [19]
1.75 60.41 Cubic 117.05 ± 5.54 – 0.26 – – – y y y y – [18]
2.18 50.75 Cubic – 163.02 ± 11.98 – – 4 – y y y y y [19]
2.18 50.75 Cubic 163.02 ± 11.98 – 0.23 – – – y y y y – [18]
2.23 49.75 Cubic – 7.28 ± 0.93 – – 6.8 – y y y y y [19]
0.17 96.16 Dodecahedron – 0.8 – – – – y y – y – [36]
0.22 95.03 Dodecahedron – 1.2 ± 0.4 – – – – y y – y – [36]
0.32 92.78 Dodecahedron – – 0.05 ± 0.005 – – – y y – y – [36]
0.35 92.10 Dodecahedron – 3.0 ± 0.2 – – – – y y – y – [36]
0.51 88.49 Dodecahedron – 26.59 ± 1.32 1.710 ± 0.06 – – – y y y y y [34]
0.53 88.04 Dodecahedron – – 0.26 – – – y y – y – [36]
(XY)d
(XY)
(Z)e
f
0.53 88.04 Dodecahedron – 31.75 ± 1.24 2.41 ± 0.05 – – – y y y y y [34]
0.62 86.00 Dodecahedron – 12.4 0.54b – 11 10−3 y y – y – [35]
83
0.82 81.49 Dodecahedron – 75.16 ± 4.11 5.59 ± 0.33f – – – y y y y y [34]
1.18 73.36 Dodecahedron – 59.4 2.99b – 5.5 10−3 y y – y – [35]
1.68 62.08 Dodecahedron 77.2 ± 6.5 112 ± 2.8 6.30 ± 0.10b 18 (106) – 10−3 y y y y – [16]
1.59 64.11 Dode thin – – 6.15b) – – – – y – y – [15]
0.58 86.91 Diamond 11.4 ± 0.09 16.3 ± 0.04 0.40 ± 0.01 – 4.5 – y y – y y [28]
0.71 83.97 Diamond 19.1 ± 0.23 26 ± 0.14 0.80 ± 0.02 – 4.5 – y y – y y [28]
0.75 83.07 Diamond 19.1 – – 3 (106) – – y y y y – [26]
0.8 81.94 Diamond 22.7 ± 0.76 30.7 ± 0.6 1.00 ± 0.02 – 4.5 – y y – y y [28]
0.8 81.94 Diamond 23.2 ± 0.78 31.6 ± 1.25 0.90 ± 0.03 – 4.5 – y y – y y [28]
0.83 81.26 Diamond (Z) 16.1 ± 0.4 21 ± 0.7 0.90 ± 0.10 – – – y y – y – [27]
0.84 81.04 Diamond (XY) 22 ± 1.1 29.3 ± 0.8 1.60 ± 0.30 – – – y y – y – [27]
0.93 79.01 Diamond 37 ± 1 – 1.13 ± 0.03 – – – y y y y – [25]
1.02 76.98 Diamond 37.8 ± 0.48 49.3 ± 0.53 1.80 ± 0.08 – 4.5 – y y – y y [28]
1.06 76.07 Diamond 41.73 – – 10 (106) – – y y y y – [26]
1.11 74.94 Diamond 46.2 ± 0.56 58.6 ± 0.79 2.30 ± 0.03 – 4.5 – y y – y y [28]
1.24 72.01 Diamond 50.6 ± 0.01 62.9 ± 0.42 3.00 ± 0.06 – 4.5 – y y – y y [28]
1.33 69.98 Diamond 62.87 – – 19 (106) – – y y y y – [26]
1.37 69.07 Diamond 89 ± 4 – 2.48 ± 0.40 – – – y y y y – [25]
(continued on next page)

Table 5 (continued)
Density Porosity (%) Lattice structure CYS (MPa) UCS (MPa) Elastic modulus Fatigue Compressive Strain Number of Dimensions Dimensions Composition Composition Reference
(g·cm−3) topology (GPa) strength strain (%) rate specimens (nominal) (measured) (nominal) (actual)
X.Z. Zhang et al.
(MPa) (s−1)
1.46 67.04 Diamond 46 ± 5 – 2.28 ± 0.84 – – – y y y y – [25]

1.59 64.11 Diamond 82.4 ± 1.39 98.4 ± 3.77 5.10 ± 0.02 – 4.5 – y y – y y [28]
1.73 60.95 Diamond 101.39 – – 40 (106) – – y y y y – [26]
1.77 60.05 Diamond 99.7 ± 2.91 118.8 ± 3.74 6.50 ± 0.09 – 4.5 – y y – y y [28]
1.77 60.05 Diamond 112.73 – – 45 (106) – – y y y y – [26]
1.9 57.11 Diamond 160 ± 2 – 5.38 ± 0.37 – – – y y y y – [25]
1.11 74.94 G6 – – 3.67b – – – – y – y – [15]
1.57 64.56 G7 59.6 ± 2.4 61.0 ± 1.4 2.40 ± 0.10b 14 (106) – 10−3 y y y y – [16]
1.83 58.69 G7 – – 6.74b – – – – y – y – [15]
1.83 58.69 G7rg – – 3.03b – – – – y – y – [15]
1.79 59.59 Hatched (XY) 107.5 ± 3.6 148.4 ± 3.5 12.90 ± 0.90 – – – y y – y – [27]
1.79 59.59 Hatched (Z) 49.6 ± 20.6 127.1 ± 29.2 3.90 ± 2.10 – – – y y – y – [27]
0.23 94.81 Hexagonal (Z) – 2.2 0.03 – – – y y – y – [43]
0.33 92.55 Hexagonal (Z) – 2.5 0.03 – – – y y – y – [43]
0.24 94.58 Hexagonal (XY) – 2.3 0.05 – – – y y – y – [43]
0.49 88.94 Hexagonal (Z) – 6.7 0.10 – – – y y – y – [44]
0.33 92.55 Hexagonal (XY) – 4.1 0.11 – – – y y – y – [43]
0.49 88.94 Hexagonal (Z) – 5.3 0.07 – – – y y – y – [44]
0.49 88.94 Hexagonal (XY) – 7.3 0.20 – – – y y – y – [44]
0.49 88.94 Hexagonal (XY) – 9.3 0.21 – – – y y – y – [44]
0.49 88.94 Hexagonal (XY) – 9.7 0.22 – – – y y – y – [44]
0.5 88.71 Hexagonal (Z) – 6 0.08 – – – y y – y – [43]
0.5 88.71 Hexagonal (XY) – 8.8 0.21 – – – y y – y – [43]
84
Note:
a
–: data not stated.
b
Modulus tested by the dynamic method.
c
y: data provided.
d
XY: The direction of compression is parallel to build direction.
e
Z: The direction of compression is perpendicular to build direction.
f
Modulus tested by four-point flexure testing method.
g
G7r: a 90° rotation about the joined pyramid axis in a G7 lattice [15].
Values in italic: excluded from Figs. 7 and 8.
Fig. 6. Accumulated strains versus cycle numbers (a–c) and stresses (d–e) for SEBM Ti-6Al-4V lattices. (a) Cubic. (b) G7 and (c) rhombic dodecahedron unit cells. (d)
Cyclic ratcheting rate. (e) S-N curves of the three lattices [16]. Inset indicates unit cell topology. The extra inset in (b) indicates Stages I, II and III.
to match the modulus of bone. whole, thereby identifying and reducing the influence of outliers that
The initiation and propagation of fatigue cracks within a strut is are clearly incompatible with established theoretical models.
another critical factor affecting the fatigue strength. In all the cases To date, 11 types of SEBM Ti-6Al-4V lattices have been studied with
studied, fatigue striations are evident on the rough fracture surfaces of a focus on cubic, dodecahedron, diamond and hexagonal lattices. The
the struts, and fatigue cracks initiate near partially fused or attached Ti- reported densities are in the range of 1.3–2.3 g·cm−3 for cubic lattices;
6Al-4V powder particles on the strut surface [16,99]. Crack initiation is 0.1–2.2 g·cm−3 for dodecahedron lattices; 0.5–1.9 g·cm−3 for diamond
also observed around internal pores in the struts [16]. The influence of lattices; and 0.2–0.5 g·cm−3 (the lowest) for hexagonal lattices. It was
defects on fatigue properties of SEBM Ti-6Al-4V lattices will be dis- noticed that with similar densities (1.63–1.68 g·cm−3), the yield
cussed in Section 6.4. Fig. 5e and f shows the fatigue failure patterns of strength, ultimate strength and fatigue strength of cubic lattices are
two SEBM Ti-6Al-4V lattices predicted by simulation. Similar to that consistently higher than respective properties of dodecahedron lattices,
observed under uniaxial compressive testing conditions [17,26,94], the especially the fatigue strength being about sevenfold higher. This can
crush band of SEBM Ti-6Al-4V lattices after fatigue testing is also 45° to be attributed to the distinct structural difference in lattice unit cell
the loading direction [16,88]. assuming the surface roughness and internal defects are similar under
the same SEBM conditions. Owing to the difficulties of producing
consistent and reliable experimental data for lattices, some important
4.3. Mechanical comparison with human bones
information is still missing, such as compressive strain, strain rate, and
even yield strength in some cases. In particular, there are scant data on
The available experimental data on the mechanical performance of
fatigue strength of lattices due further to the associated experimental
SEBM Ti-6Al-4V lattices are exhaustively summarized in Table 5 in
time and cost. This lack of experimental data places a significant lim-
relation to density, topology, strength, modulus, fatigue strength and
itation on the clinical applications of SEBM Ti-6Al-4V lattices.
compressive strain. This and the following two sections provide an
To provide insight into the practical design of lattice implants, the
extensive analysis of this data-set. The intention is twofold: (i) to pro-
mechanical properties of SEBM Ti-6Al-4V lattices and human bones are
vide a previously unavailable repository for the design of commercial
plotted together in Fig. 7a–c vs. density or porosity using detailed data
implant structures; and (ii) to generate by unifying such a wealth of
in Tables 2–5. The porosity of human bone is calculated by using the
data new insights into the achievable performance of SEMB Ti-6Al-4V
bone density of 1.99 g·cm−3 as the theoretical density [75]. The fol-
lattices that are not possible when the associated data are observed in
lowing observations can be drawn:
isolation.
To provide an archival quality resource, and to ensure valid insights
(1) Even though the experimental data were obtained by different re-
are generated, it is imperative that the data quality be ensured [104].
searchers, the UCS, CYS and modulus values of SEBM Ti-6Al-4V
This is achieved by two methods in this research. Firstly, data are
lattices all increase with increasing density, showing a correct
considered only from peer reviewed data sources and reference is made
trend. In fact, as will be shown below, they all follow the Gibson-
to the ISO standard for testing and reporting mechanical property data
Ashby model, indicative of the quality of the experimental data
for porous and cellular metals (ISO 13314) [105]. Secondly, unifying
plotted in Fig. 7.
test data within a single repository allows the data to be assessed as a
85
50 years but about 10 MPa lower than that of the bones of younger
than 40 years (Table 3). Cubic lattices with porosity of 50.75%
achieved a UCS of 163.02 ± 11.98 MPa and hatched XY with
porosity of 59.59% exhibited a UCS of 148.40 ± 3.50 MPa
(Table 5).
(3) CYS data are scarce for all human bones (only one result of human
cancellous bone, Fig. 7b) but cubic and diamond SEBM Ti-6Al-4V
lattices can achieve high CYSs (Table 5), and the three highest
values are about 20 MPa higher than the reported TYS of human
cortical tibia and femur. The CYSs of these two SEBM lattices are
expected to be similar to those of human cortical tibia and femur.
(4) Human bone experiences tensile and compressive stresses and
strains alternately every day, especially the tibia and femur.
However, no tensile property data have been reported for SEBM Ti-
6Al-4V lattices. Significant further research is still needed in that
regard.
(5) As can be seen from Fig. 7c, the modulus of SEBM Ti-6Al-4V lattices
is comparable to that of human cancellous bone (density is much
higher) but much lower than that of human cortical bone (density is
similar). The modulus of SEBM Ti-6Al-4V can be relegated through
adjusting porosity and/or the unit cell structure. In addition, testing
methods can noticeably affect the modulus data obtained. The static
modulus and dynamic modulus are similar for SEBM Ti-6Al-4V
lattices with similar density but the flexure modulus can be much
higher.
(6) The fatigue performance can decide the lifespan of a lattice implant.
Limited fatigue data (Table 5) indicates that SEBM Ti-6Al-4V lat-
tices can meet the requirements for fatigue strength as dental im-
plants by ISO 14801, i.e., a minimum of 5 million cycles of com-
pressive loading at 2–15 Hz [107]. However, only SEBM Ti-6Al-4V
cubic lattices (porosity: 63.21%) can offer fatigue strengths
(105 MPa measured at 106 cycles) higher than that of human cor-
tical femur (60 MPa measured at 106 cycles) [108]. On the other
hand, the fatigue strength of human cortical tibia was reported to
be 160 MPa at 106 cycles [109]. No SEBM Ti-6Al-4V lattices re-
ported so far have achieved such high fatigue strength. Reducing
porosity is expected to improve fatigue strength at the cost of in-
creased density.
(7) The ratio of fatigue strength to UCS for SEBM Ti-6Al-4V lattices falls
in the range of 0.15–0.55, which is lower than 0.65 for dense Ti-
6Al-4V [16,110,111]. This can be attributed to the combined effect
of the (i) open-celled high porosity, (ii) rough strut surfaces (Section
6.3), (iii) internal defects (Section 6.4), (iv) notch sensitivity of Ti-
6Al-4V [24,95], and (v) inhomogeneous microstructure (Section 5)
[112].
(8) No data has been reported on the compressive strain of human
cortical bone, whose tensile strain is in the range of 1–4% (Table 3).
Again, only SEBM Ti-6Al-4V cubic lattices can endure accumulated
strain up to 2% at 106 cycles (Fig. 6a–c), which is comparable to the
tensile strain of human cortical fibula and humerus bone.
These comparisons show the deficiencies of current SEBM Ti-6Al-4V

lattices for implant applications, as well as their high potential. By
taking into account the compressive strengths (both UCS and CYS),
modulus, strain, and fatigue strength, SEBM Ti-6Al-4V cubic lattices
(porosity: 63%) have proved to be the best and can replace human
cancellous bone, cortical fibula and humerus. However, in order to be
Fig. 7. Literature data on (a) ultimate compressive strength (UCS) and ultimate able to replace human cortical tibia and femur, both the compressive
tensile strength (UTS). (b) Compressive yield strength and tensile yield strength and fatigue strength of SEBM Ti-6Al-4V cubic lattices still need
strength, and (c) modulus vs. density for SEBM Ti-6Al-4V lattices and human
to be largely improved.
bones.
Up till now, the mechanical properties of 11 different types of SEBM
Ti-6Al-4V lattices have been reported. It is possible that some of the
(2) The UCSs of SEBM Ti-6Al-4V lattices are comparable to, or higher remaining 12 different types of lattice structures listed in Table 1 may
than, that of human cancellous bone (Fig. 7a). The highest UCS display better mechanical properties than the cubic lattice. This needs
reported is 196 ± 7.1 MPa (Table 5, cubic; porosity: 63%), which to be verified and can be costly. Besides, no comprehensive mechanical
is similar to that of human cortical tibia and femur of older than property data have been reported for stretch-dominated SEBM Ti-6Al-
86
4V lattices, which offer higher modulus and UCS than bending-domi- (1) All the property data listed in Table 5 were obtained from bending-
nated ones at the same density (Section 4.4). In addition, it is necessary dominated SEBM Ti-6Al-4V lattices (no experimental data on
to produce systematic tensile property data on SEBM Ti-6Al-4V lattices stretch-dominated lattices). The exponent obtained (2.14 in
for a detailed comparison with those of human bones for better me- Fig. 8a″) is close to the theoretical value of 2.0 for buckling (Eq. (3))
chanical match. In brief, cubic lattices with a high relative density have but clearly higher than 1.5 for bending (Eq. (2)), indicating that the
the potential to satisfy most requirements for strength. However, in- dominated mechanical mechanism is buckling. The slightly higher
creasing density implies reducing the void size, which could lead to exponent (2.14 vs. 2.0) can be attributed to the α′-martensite mi-
more powder particles entrapped in the lattice causing negative im- crostructure in the lattice (to be discussed in Section 5), which leads
pacts. Some of the challenges will be discussed in Section 8. to higher strength but lower ductility. In addition, geometric
variability in lattice strut geometry (Section 6.2) can make it dif-
4.4. Comparison with the Gibson-Ashby model ficult to measure strut cross-section thereby affecting the experi-
mental results.
A number of predictive models have been proposed to relate void (2) The similar exponents for SEBM Ti-6Al-4V lattices (2.14 in Fig. 8a″)
topology and material property parameters to mechanical properties. and human bones (1.94 in Fig. 8a′) suggest that the lattices can
The most notable one is the Gibson-Ashby model [92], described by respond similarly to external loading as human bones.
Eqs. (1)–(6) in Table 6 for different cellular solids. They have been (3) The exponent for the E−ρ relationship for the lattices (2.95 in
successfully applied to metallic foams made from Al and its alloys, Mg Fig. 8b″) is close to that of human bone (2.87 in Fig. 8b′). In con-
alloys and Zn alloys [4]. Their applicability to SEBM Ti-6Al-4V lattices junction with their similar exponents for the σ−ρ relationship (2.14
remains unknown due to the lack of sufficient experimental data for a vs. 1.94), it can be concluded that the SEBM Ti-6Al-4V lattices
detailed assessment. shown in Table 5 can behave similarly as human bones. Both the
Fig. 8 plots the ultimate strength vs. density and static modulus vs. exponents for the bone (2.87) and that for the lattices (2.95) deviate
density on the logarithmic scale for both SEBM Ti-6Al-4V lattices and substantially from the theoretical value of 2 (Eq. (1)). This implies
human bones using the data listed in Tables 2–5. It should be pointed that the model may need to be improved. Technically, the elastic
out that the densities of human cortical and cancellous bones fall in the response of a lattice includes both linear and non-linear elastic
ranges of 1.72–1.91 and 0.19–0.50 g·cm−3, respectively. As a result, the zones. Moreover, it is often difficult to clearly define the linear
mechanical property data shown in Fig. 8a′ and b′ for these two types of elastic zone from the stress strain curve for a lattice. Both issues can
bones are limited to the ranges of 0.54–0.65 (i.e., ln1.72–ln1.91) and entail uncertainties in the experimental measurements.
−1.67 to −0.69 (i.e., ln0.19–ln0.50) in ln (Density) axis, respectively.
The following observations can be made with the Gibson-Ashby model: In summary, the Gibson-Ashby model is applicable to SEBM lattices
for certain cases but significant deviations exist for other cases.
(1) The high values of R2 (0.92–0.99) indicate that the Gibson-Ashby
model can be used to describe SEBM Ti-6Al-4V lattices and human 4.5. Ti-6Al-4V auxetic lattice structure manufactured by SEBM
bones. This observation also indicates that experimental data re-
ported by different researchers are consistent and of high quality. Unlike all other lattices discussed above, the transverse dimension
(2) Based on exponents shown in Fig. 8, Table 7 lists the strength and of an auxetic lattice structure diminishes under compression. This un-
modulus data together with the calculated Gibson-Ashby constants ique property, which implies a negative Poisson’s ratio, ν, can enable
(C1 and C2) for a further assessment of the experimental data. About innovative designs for applications including high sandwich panels,
72% of C1 fall in the range of 820–1450 and the corresponding UCS radome frames, seat cushion, bulk heads of pressure vessels, aerospace
data can be regarded as more effective data. Replotting this dataset fillers and implants [8–12].
in Fig. 8a″ increased the value of R2 from 0.92 to 0.99, which is A notable application of the auxetic structures in the biomedical
remarkable and leads to the Gibson-Ashby model for the strength of area is that of the artificial auxetic blood vessel, which can increase in
SEBM Ti-6Al-4V lattices in the form of Eq. (7), where the pore-free wall thickness when a pulse of blood flows through it, thereby avoiding
density of Ti-6Al-4V is taken to be 4.43 g·cm−3. Similarly, the more eventual rupture due to wall thinning [113]. Another application is the
effective modulus data are replotted in Fig. 8b″ with an improved auxetic dental floss, which offers the ability to expand so as to timely fit
value of R2 = 0.99, which leads to Eq. (8) for the compressive the widely differing gaps between teeth [114]. In addition, it can have
modulus of SEBM Ti-6Al-4V lattices. in-built drug release features [114]. SEBM enables the manufacture of
intricate auxetic lattice structures (anisotropic, see Table 1) with a high
ρ 2.14
σ = C1 ∗ ⎛ ⎞ degree of control over the resulting mechanical properties.
⎝ 4.43 ⎠ (7)
Fig. 9 shows the compression mechanical properties obtained from
ρ 2.95 SEBM Ti-6Al-4V auxetic lattices. Compressive loading was applied to
E = C2 ∗ ⎛ ⎞ each face individually to determine the anisotropic mechanical prop-
⎝ 4.43 ⎠ (8)
erties (Static modulus and Poisson’s ratio) of the auxetic lattice. As
Table 6
The expression of Gibson-Ashby model.
Mechanical properties of cellular solid Bending-dominated cellular solid Stretch-dominated cellular solids
Modulus (E) ρ
2 ρ
E = C∗ES ⎛ ⎞ or E ∝ ρ2 (1) E = C∗ES ⎛ ⎞ or E ∝ ρ (4)
⎝ ρS ⎠ ⎝ ρS ⎠
A cellular solid that made of ductile materials, bending-dominated behavior
Collapse strength (σ) 1.5 ρ
ρ
σ = C∗σy,s ⎛ ⎞ or σ ∝ ρ1.5 (2) σ = C∗σy,s ⎛ ⎞ or σ ∝ ρ (5)
⎝ ρS ⎠ ⎝ ρS ⎠
A cellular solid with slender struts collapses by elastic buckling, buckling-dominated behavior
Buckling strength (σ) 2 2
ρ ρ
σ = C∗Es ⎛ ⎞ or σ ∝ ρ2 (3) σ = C∗Es ⎛ ⎞ or σ ∝ ρ2 (6)
⎝ ρS ⎠ ⎝ ρS ⎠
87
Lattice structure Human bone
Fig. 8. Log-log plots of literature data on UCS and UTS and modulus vs. density for SEBM Ti-6Al-4V lattice (a, b, a″, b″) and human bone (a′, b′). The data shown in
(a″) and (b″) have been validated. Insets indicate unit cell topology analyzed in this figure.
shown in Fig. 9a, the modulus values obtained from both experiment unique SEBM Ti-6Al-4V auxetic lattices. However, owing to the rough
and simulation increase progressively with increasing relative density, surface (Section 6.3) and defects (Section 6.4), the predicted values can
indicating that the Gibson-Ashby model can also be applied to the deviate substantially from experimental data. The exponents for the
88
Fig. 9. Static modulus and Poisson’s ratio of SEBM Ti-6Al-4V auxetic lattices [11]. (a) Comparison of static modulus data obtained from simulation and experiment
vs. relative density for four different orientations. (b) Legends. (c) Dependence of Poisson’s ratio (ν) on relative density, where εi and εj are the resulting strain values
perpendicular and parallel to the force direction, respectively. (d) Dependence of Poisson’s ratio on the internal angle of the structure at a relative density of 10%.
E−ρ relationship (Eq. (1)) for cell orientations of 1, 2 and 3 are in the 5.1. Microstructure of SEBM Ti-6Al-4V lattices
range of 2.2–2.5, close to 2.0, suggesting bending-dominated de-
formation [11,12]. However, for the cell orientation at 45°, which SEBM of Ti-6Al-4V is featured by a high powder bed temperature
corresponds to rotating the x3-orientation around the x1-axis of the (build platform is preheated to 730 °C), which is kept above 500 °C
structure by 45°, the exponents obtained by simulation (2.5) and during the entire SEBM process. The microstructure in each layer (apart
measurement (2.6) are greater than 2.0 but close to that of human from the last few layers) undergoes a complex thermal process during
bones (2.87). This is consistent with the results discussed in Section 4.4. SEBM. In addition, it is affected by the geometry-related local cooling
Poisson’s ratio shows clear anisotropy at each relative density from 30% rate. Fig. 10a–d compares the microstructures of 1.0 mm diameter
to 50% (Fig. 9c) and each internal angle from 40° to 100° (Fig. 9d). SEBM Ti-6Al-4V struts with inclination angles of 0 and 90° [89]. The
However, the degree of anisotropy diminishes with increasing relative location-specific microstructure includes that (i) columnar prior β
density (Fig. 9c). It also decreases with increasing internal angle be- grains in the strut are parallel to the build direction where inside the
tween struts (Fig. 9d). These observations fortify the unique deforma- columnar β grains are α and α′ platelets (Fig. 10b), (ii) α + β basket-
tion features of Ti-6Al-4V auxetic lattices in relation to other types of waves are observed in the strut upper section vertical to the build di-
lattices. rection (Fig. 10c), and (iii) α′-martensite close to the stainless steel
Other design considerations for biomedical auxetic lattices include substrate due to the faster cooling conditions therefore (Fig. 10d). These
[8,9] (i) the re-entrant angle θ has a more profound influence on represent the general microstructural features in SEBM Ti-6Al-4V lat-
compressive strength than does the H/L ratio (H: length of the vertical tices. It is location specific and strut-diameter-related.
strut; L: length of the internal strut), and (ii) the relative density affects
not only the compressive strength and modulus but also the sign of
Poisson’s ratio. 5.2. Effect of heat treatment on SEBM Ti-6Al-4V lattice microstructure
α′-martensite offers high strength but low ductility. Post heat

5. Microstructure of SEBM Ti-6Al-4V lattices and effects of post treatment in either the α + β or β region (β-transus = 994 °C for Ti-Al-
heat treatment 4V) can eliminate the α′ phase and produce a variety of microstructures
containing α and β phases depending on the cooling rate (see Table 8).
During SEBM of Ti-6Al-4V, the melt pool solidifies as the single β The ductility improves but the strength decreases. The columnar prior β
phase. Depending on the subsequent cooling rate, the β phase can grains still exist after heat treatment (e.g., 960 °C/2h/furnace cooling)
transform into α′-martensite, massive αm and α. Table 8 summarizes in the α + β region (Fig. 10e) [89]. However, annealing in the β-phase
the cooling rates reported for the formation of different phases in Ti- region (e.g., 1200 °C/2h/furnace cooling) can eliminate the columnar
6Al-4V when cooled from the β phase field. Owing to their small strut grains and the lattice microstructure obtained is α + β basketwaves
diameters, α′-martensite is commonly observed in the struts of SEBM [89]. The α-lath thickness can increase substantially (e.g., from
Ti-6Al-4V lattices. Post-SEBM heat treatment can change the marten- 0.9 ± 0.5 µm to 6.3 ± 1.5 µm after 960 °C/2h/furnace cooling [89]),
sitic structure for more balanced mechanical properties. which reduces the yield strength by a modified Hall-Petch relationship
[65]. A slight increase in oxygen during post heat treatment may occur
but Ti-6Al-4V manufactured from powder, its oxygen content can
89
Table 7
Validation of the UCS and modulus data using the Gibson-Ashby model.
Lattice topology Density, ρ (g·cm−3) UCS, σucs (MPa) ρ
2.19 Static modulus, E (MPa) ρ
2.86
C1 = σUCS / ⎛ ⎞ C2 = E/ ⎛ ⎞
⎝ ρS ⎠ ⎝ ρS ⎠
Cubic 1.31 83.13 1198 –a –

Cubic 1.31 83.8 1208 1130 36,847
Cubic 1.34 – – 200 6152
Cubic 1.63 196 1751 – –
Cubic 1.75 117.05 895 – –
Cubic 1.75 – – 260 3627
Cubic 2.18 163.02 770 – –
Dodecahedron 0.17 0.8 1009 – –
Dodecahedron 0.32 – – 50 91,273
Dodecahedron 0.51 26.59 3025 – –
Dodecahedron (XY) 0.53 – – 260 113,220
Dodecahedron (XY) 0.53 – – 440 190,002
Dodecahedron (Z) 0.53 – – 130 57,695
Dodecahedron 0.53 31.75 3321 – –
Dodecahedron 0.73 20.9 1084 – –
Dodecahedron 0.82 75.16 3023 – –
Dodecahedron 0.9 73.26 2403 – –
Dodecahedron 1.17 131.6 2430 – –
Dodecahedron 1.18 59.4 1076 – –
Dodecahedron 1.18 62.79 1138 – –
Dodecahedron 1.23 84.28 1395 – –
Dodecahedron 1.29 120.08 1790 – –
Dodecahedron 1.68 112.8 943 – –
Dodecahedron 1.68 112 936 – –
Dodecahedron 1.75 – – 12,750 181,594
Dodecahedron 1.76 – – 11,990 168,023
Dodecahedron 1.78 – – 11,290 153,195
Diamond 0.58 16.3 1399 400 134,082
Diamond 0.71 26 1433 800 150,385
Diamond 0.8 30.7 1303 1000 133,622
Diamond 0.8 31.6 1341 900 120,260
Diamond (Z) 0.83 21 822 900 108,242
Diamond (XY) 0.84 29.3 1117 1600 185,950
Diamond 0.93 – – 1130 98,159
Diamond 1.02 49.3 1229 1800 120,057
Diamond 1.06 – – – –
Diamond 1.11 58.6 1214 2300 120,453
Diamond 1.24 62.9 1023 3000 114,459
Diamond 1.37 – – 2480 71,145
Diamond 1.46 – – 2280 54,526
Diamond 1.59 98.4 928 5100 95,563
Diamond 1.77 118.8 886 6500 89,624
Diamond 1.9 – – 5380 60,571
G7 1.57 61.0 592 – –
Hatched (XY) 1.79 148.4 1080 12,900 172,244
Hatched (Z) 1.79 127.1 925 3900 52,074
Hexagonal (Z) 0.23 2.2 1432 26 121,369
Hexagonal (Z) 0.33 2.5 738 30 50,789
Hexagonal (XY) 0.24 2.3 1364 50 222,863
Hexagonal (Z) 0.49 6.7 832 100 53,903
Hexagonal (XY) 0.33 4.1 1210 110 178,771
Hexagonal (Z) 0.49 5.3 658 70 38,044
Hexagonal (XY) 0.49 7.3 907 200 109,892
Hexagonal (XY) 0.49 9.3 1155 210 114,632
Hexagonal (XY) 0.49 9.7 1205 220 119,599
Hexagonal (Z) 0.5 6 713 80 40,536
Hexagonal (XY) 0.5 8.8 1046 210 108,282
Note:
a
–: data not available.
Values in italic: excluded from Fig. 8a″ and b″.
tolerate up to 0.32% prior to a clear reduction in tensile ductility Arcam A2 and 140 µm for the Arcam Q20. It is thus challenging to
[119,120]. manufacture struts thinner than respective beam size. The unstable
melt flow and layer-by-layer AM process (see Fig. 11) could add extra
geometric deviations from the digital model and easily lead to rough
6. Geometric errors associated with SEBM Ti-6Al-4V lattices
surfaces and defects [70,121,122]. This section discusses a few im-
portant issues in geometric errors.
The electron beam used for SEBM is about 250 µm in diameter for
90
Table 8
Experimental critical cooling rates for phase transformations in Ti-6Al-4V [115].
Composition of Ti-6Al-4V (wt.%) Critical cooling rate dT/dt (°C/s) Transformation Microstructural features Ref.
Ti-6.04Al-4.03V 420 < dT/dt < 525 β → α′ Fully martensitic (α′) [116]
Impurities: 0.12Fe-0.09O-0.03C-0.009N-0.023H 20 < dT/dt < 420 β → αm Massive transformation (patch-like αm phases)
Quenched from 1027 °C dT/dt < 20 β→α Coarse lamellar α/β and GB α
Ti-6.1Al-4.3V dT/dt ≥ 18.0 β → α′ Fully martensitic (α′) [117]

Impurities: 0.16Fe-0.01C 3.5 < dT/dt < 18.0 β→α Lamellar α/β and GB α
Quenched from 1027 °C dT/dt < 2.0 β→α Lamellar α/β and GB α
Ti-6.5Al-4.4V dT/dt ≥23.1 β → α′ Fully martensitic (α′) [118]

Impurities: 0.13O-0.15Fe-0.023C-0.017N-0.0024H 7.3 < dT/dt < 23.1 β → α Lamellar α/β and GB α
Quenched from 1020 °C 0.065 < dT/dt < 7.3 β → α Lamellar α/β and GB α
0.015 < dT/dt < 0.065 β → α Coarse lamellar α/β and GB α
dT/dt ≤ 0.012 β → α Equiaxed α with β
6.1. Lattice manufacturability by SEBM strut vertically (θ = 90°) is always preferred which ensures L = d . Re-
arranging Eq. (9) yields
AM produces a metal lattice from an idealized digital model. L h
However, many process factors including the powder size, layer = 1− cosθ 0° < θ < 90°
d/sinθ d (10)
thickness, beam diameter, thermal distortion and residual stress hinder
the full realization of such digital models [34]. The manufacturability A study of the literature data on d, h and θ [31,40] revealed that
of SEBM thus remains to be a concern for the production of precise L/(d/sinθ) should be greater than about 0.90 for AM of quality struts by
lattices for biomedical applications [77]. Table 9 summarizes detailed SEBM. According to Eq. (10) this requires that
literature data for the manufacture of Ti-6Al-4V lattices, including h
process parameters, powder characteristics, and lattice structure geo- cosθ ⩽ 0.10 0° < θ < 90°
d (11)
metry. These process variables affect the manufacturability including,
dimensional accuracy (Section 6.2), roughness (Section 6.3), and the Eq. (11) can be regarded as a generic requirement for AM of quality
occurrence of lack-of-fusion defects (Section 6.4). It should be noted struts and for understanding various experimental observations re-
that there are other factors which can also affect the manufacturability ported in the literature, e.g., the smallest diameter of a quality strut
of Ti-6Al-4V lattices, e.g., the time interval between two layers and the discussed in Refs. [34,43,44].
AM atmosphere. This layer-by-layer AM process or stair-case affects not only the
Fig. 11a and b schematically illustrates the correlation of the strut manufacturability but also the surface quality of a strut (will be dis-
inclination angle (θ), diameter (d) of the strut, layer thickness (h) and cussed Section 6.3), as shown in Fig. 11c and d.
the strut-strut overlap (L) between each two layers, where the re-
lationship between L and d, h, and θ is defined by Eq. (9). 6.2. Dimensional accuracy
d h The dimensional accuracy of a SEBM Ti-6Al-4V lattice can be

L= − 0° < θ < 90°
sinθ tanθ (9) compromised by thermal stresses, achievable process parameters and
the layer-by-layer nature of the SEBM process. It can be assessed by
The continuity of a strut requires that the value of L be as close to comparing the actual strut diameter dactual with the design diameter
d/sinθ as possible (L ⩽ d/sinθ , Fig. 11) and in that regard building a dtarget, as shown in Table 10. In general, fine struts with dtarget ranging
Fig. 10. Optical microstructures of SEBM Ti-6Al-4V cubic lattice before (a–d) and after heat treatment (e–f). (a) Three locations of interest before heat treatment were
selected. (b) The mid-height of a vertical strut. (c) The top section of a horizontal strut. (d) A region close to the stainless steel platform. (e) 960 °C/2h/furnace cooling
(primary α plus α + β) and (f) 1200 °C/2h/furnace cooling (lamellar α + β). (The axis indicates the build direction) [89].
91
Fig. 11. Effect of inclination angle on strut continuity. (a) High inclination angle and (b) low inclination angle [36]. The surface roughness of different SEBM Ti-6Al-
4V lattices: (c) cubic [41]. (d) Octet-truss and (e) octet-truss after chemical etching [49]. The red circulars indicate the influence of stair-case on strut roughness.
from 0.2 mm to 0.8 mm (except the 0.722 mm) were found to be 6.3. Surface roughness
oversized while coarse struts with dtarget ≥ 0.9 mm were consistently
undersized. In particular, for small values of dtarget, e.g., 0.2 mm, dactual Surface roughness (Ra) affects the static strength, fatigue perfor-
can reach 2dtarget. However, the influence of inclination angle on dtarget mance and functional properties, such as fluid flowability, optical
remains unclear. Since the dimensional accuracy of lattice structure is properties, frictional behavior and heat transfer efficiency
of paramount importance for implant applications, a better under- [57,125–129]. The surface of a SEBM lattice is inherently rough
standing of the uncertainty in dimensional accuracy is necessary for the (Fig. 11c and d) due to a number of factors. These include (i) coarser
design and fabrication of Ti-6Al-4V lattices. powder (45–105 µm) (20–45 µm for SLM), (ii) large electron beam size
Table 9
Parameters summary of SEBM Ti-6Al-4V lattices.
Machine Stainless steel Preheating Melting Layer Mean Distribution Size of Strut Research Ref.
mode platform thickness powder (µm) strut angle (°) intent
temperature (°C) Beam Scan Beam Scan (µm) size (µm) (mm)
current speed current speed
(mA) (mm/s) (mA) (mm/s)
Q10 –a – – 3.0 450 50 – 45–105 0.3 0, 90 A/L/Mb [7]

A2 – 0–38 14,600 1.7–3.0 200 100 – 45–100 0.8, 0.9 20, 45, A/L/M [8–10,43]
90
A2 – – – 1.9–6.0 160 100 70 ∼105 – 60, 90 L [12]
A2 650 – – – – – 30 – 45 M [17]
A1 730 – – – – – 50 – 0.5 30, 36, L/M [16,41,96]
90
A1 720 – – – – 50 – – 2, 5, 10 20, 30, L [123]
40, 50,
60
A1 700 – – – – 50 77 – 1 0, 45, 90 L [124]
A1 730 0.4–15 – 0.4–15 – 50 – 45–100 0.1, 0.2, 60 M [94]
0.3, 0.4
S12 750 – – – – – – – 0.7 60 M [44]
S12 – – – – – 100 – 45–100 – 90 A/L/M [91]
S12 – – – – – – 70 – – 109.5 M [26]
S12 – – – 2 240, 70 70 – 0.42, 109.5 M [28]
160, 133 0.54
S12 – – – – – 100 70 – – 109.5 M [27]
S12 640 – – – 200 70 – 45–100 0.2, 0.5, 109.5 M [25]
0.8
Note:
a
–: not available.
b
Aerospace, Lattice, Medical.
92
Table 10
Comparison between the design and actual diameters of SEBM Ti-6Al-4V struts. Some struts are from a lattice and other are made as single strut.
Lattice structure Single strut, inclination angle dtarget (mm) dactual (mm) Deviation (mm) (%) Ref.
Diamond 0.200 0.398 ± 0.037 0.198 ± 0.037 99.00 [25]

Dodecahedron 0.454 0.559 ± 0.025 0.105 ± 0.025 23.13 [34]
Diamond 0.500 0.600 ± 0.060 0.100 ± 0.060 20.00 [25]
Dodecahedron 0.575 0.590 ± 0.036 0.015 ± 0.036 2.61 [34]
Dodecahedron 0.722 0.573 ± 0.030 −0.149 ± 0.030 −20.64 [34]
Diamond 0.800 0.769 ± 0.042 0.031 ± 0.042 3.88 [25]
Cubic 0.800 0.81 ± 0.252 0.010 ± 0.252 1.25 [89]
Dodecahedron 0.905 0.777 ± 0.035 −0.128 ± 0.035 14.14 [34]
Cubic 1.000 0.970 ± 0.023 −0.030 ± 0.023 3.09 [89]
Cubic 0° 1.000 0.518 −0.482 −48.20 [124]
Cubic 45° 1.000 0.414 −0.586 −58.60 [124]
Cubic 90° 1.000 0.458 −0.542 −54.20 [124]
Dodecahedron 1.153 1.003 ± 0.015 −0.15 ± 0.015 −13.01 [34]
Dodecahedron 1.361 1.172 ± 0.026 −0.189 ± 0.026 −13.89 [34]
Dodecahedron 1.408 1.229 ± 0.025 −0.179 ± 0.025 −12.71 [34]
Cubic 1.500 1.480 ± 0.019 −0.020 ± 0.019 −1.33 [89]
Dodecahedron 1.732 1.595 ± 0.038 −0.137 ± 0.038 −7.91 [34]
Cubic 1.800 1.780 ± 0.028 −0.020 ± 0.028 −1.11 [89]
Dodecahedron 2.096 1.959 ± 0.023 −0.137 ± 0.023 −6.54 [34]
(140–250 µm) (70–115 µm for SLM), (iii) large layer thickness (50/ 30 µm at θ = 0° and 35 µm at θ = 90°.
70 µm vs. 30/50 µm for SLM), (iv) unavoidable adherence of partially A rough surface can be beneficial to enhance osseointegration
fused powder [130] (see Fig. 11c and d), (v) intrinsic stair-case effect [131,132]. Human bone marrow can detect changes in roughness of the
(Fig. 11a and b), and (vi) the dynamic, unstable melt flow [126,129]. In order of magnitude 0.6 µm [133]. However, an increased implant sur-
addition, scanning strategy, parameter settings and strut geometry also face could lead to an increased release of Al and V ions [134], which
play a role [129]. As a result, the Ra value is typically ∼40 µm for SEBM interfere with mineralization of the adjacent bone [135]. Besides, a
Ti-6Al-4V vs. ∼25 µm for SLM Ti-6Al-4V. Lhuissier et al. [53] char- highly rough surface can increase the development of bacteria in areas
acterized the surface roughness of 1.6 mm diameter single struts in a with low blood flow [136]. Mechanically, high surface roughness can
SEBM Ti-6Al-4V octet-truss lattice, and found that the Ra value was noticeably decrease tensile strength and ductility [126,137,138].
Fig. 12. Defect distribution in SEBM Ti-6Al-4V lattices and the struts. (a) Cubic [16]. (b) Vertical strut of 1.0 mm diameter [126]. (The red spot in a and b are the
pores). (c) Vertical strut of 1.66 mm diameter [49]. (Arrows indicate the lack-of-fusion defects).
93
Surface polishing works [139] but it is challenging to deal with those

inaccessible internal structures. In that regard, chemical etching
[49,53] remains to be an important solution in that regard (Fig. 11d
and e).
6.4. Defects in SEBM Ti-6Al-4V lattices
Similar to dense SEBM Ti-6Al-4V, the major defects in SEBM Ti-6Al-

4V lattices are voids (Fig. 12a and b) and lack-of-fusion features
(Fig. 12c) which can be larger than 100 µm in size and are typical ir-
regular in shape [73,140]. Closed pores ranging from 30 to 80 µm were
found in SEBM Ti-6Al-4V cubic, G7 and rhombic dodecahedron lattices
[16]. They are larger than the approximate 15 µm critical size for dense
Ti alloys below which there is negligible influence on tensile and HCF
performance [141]. These large voids and irregularly shaped lack-of-
fusion defects can be a killer to the fatigue performance of SEBM lat-
tices [142,143].
The primary reasons for the formation of these defects can be as-
cribed to (i) unstable melt flow and evaporation of chemical elements
[65], (ii) internal pores from feedstock powder [123,124,144,145], (iii)
insufficient melting of local layer-layer contacts, and (iv) the 8% liquid-
to-solid shrinkage for Ti-6Al-4V [146]. Hot isostatic pressing (HIP) can
shrink or heal them to below the resolution limit of X-ray micro-com-
puter tomography (1.5 µm). However, surface connected pores could be
expanded after HIP [73]. This can be an issue for lattices due to the
small strut diameters.
7. Biomedical applications of SEBM Ti-6Al-4V lattices
There have been significant clinical applications of SEBM Ti-6Al-4V

lattices since 2002. Discussed below are just three selected examples.
Fig. 14. SEBM Ti-6Al-4V hemi-pelvic [148].
Case study 1: SEBM Ti-6Al-4V ankle implants
treatment, was implanted into a patient’s foot on 30 January 2015 at
Based on patient-specific CT scan data a custom implant was de- Guangzhou Hospital of Traditional Chinese Medicine in Guangzhou
signed to replace resected ankle bone, as shown in Fig. 13 [147]. The City, China [147]. The implant has functioned as required after more
lattice part in the implant had 90% porosity, which is conformal to the than two years of implantation and the patient walks normally today.
medullary cavity with non-uniform triangular cells ranging from about
2.5 mm to 15 mm in size, and a 99% dense section, which is patient
specific to the cortical bone. The implant was manufactured by SEBM Case study 2: SEBM Ti-6Al-4V hemi-pelvic
using extra-low interstitial (ELI) Ti-6Al-4V powder and has a mass of
about 350 g and an overall length of 150 mm. A Ti-6Al-4V hemi-pelvic with a lattice structure, as shown in Fig. 14
The as-built Ti-6Al-4V ankle part, without subsequent surface [148], was designed and fabricated using SEBM for a patient. The
Fig. 13. SEBM Ti-6Al-4V ankle lattice implant. (a) Computer-aided design (CAD) model. (b) As-built implant. (c) Post implantation CT scan [147].
94
design considerations can be found from Ref. [148]. After SEBM, the as- implants can well satisfy the requirements. As pointed out in Section 3,
built hemi-pelvic was further processed using computer numerical different human cortical bones can show different properties (Tables 2
control (CNC) machining for several needs including adding threads for and 3). Novel lattice designs with a topology gradient could provide an
locking screw fixation [148]. The final implant was reported to have the innovative solution to satisfying all such property requirements. Re-
dimensions of 80 mm × 143 mm × 150 mm. The surgery occurred search on graded Ti-6Al-4V lattices has recently received attention
successfully in June 2016 [148]. In particular, the locking screws and [25,96].
hip cup machined areas functioned as required. The off-the-shelf
ceramic insert fitted the hip cup perfectly allowing the hip component
to articulate correctly to the insert. The surgeon reported that the pa- 8.2. Improving the ductility of SEBM Ti-6Al-4V lattices
tient recovered well post-operatively.
No tensile property data have been reported on SEBM Ti-6Al-4V
Case study 3: SEBM Ti-6Al-4V cervical vertebral fusion cages lattices due to their rough strut surfaces, internal defects and in-
completely decomposed α′-martensitic microstructure, which together
The cervical vertebral fusion cage shown in Fig. 15 was designed eliminate tensile ductility. In fact, even the ductility under compression
with internal lattice structures to promote bone ingrowth while redu- (at the first fracture) is also limited (mostly between 3.5% and 5.0%,
cing mass, cost and build time [147]. The lattice designed has 80% Table 5). Post-SEBM heat treatment may improve ductility. The lack of
porosity and the cell size was 0.45 mm. The cage weighs 90 g with di- ductility remains to be a major concern for the wider applications of
mensions of 40 mm × 15 mm. It was implanted into a patient’s neck SEBM Ti-6Al-4V lattices and needs to be addressed properly.
(C2–C4) on 27 November 2015 at the 301 Hospital in Beijing [147].
The patient recovered well after surgery and is in good condition today
after more than two years of implantation. 8.3. Manufacturing issues associated with Ti-6Al-4V lattices by SEBM
Compared with AM of dense components, AM of Ti-6Al-4V lattices

8. Outstanding challenges and future directions by SEBM is still far from mature due to both their specific geometrical
complexity and fine scales (strut diameter < 1.0 mm). For example, Eq.
Although SEBM provides an unprecedented opportunity for in- (11) imposes a strict condition for SEBM of quality struts. Chemical
novative design of Ti-6Al-4V lattices, challenges exist that restrict their etching is effective to reduce the roughness to only ∼10 µm [126] but
wider application in a clinical setting including mechanical properties, high fatigue performance requires a much smoother surface. The same
manufacturing capability, limited manufacturable cavity size and concern applies to the nodes in lattices, which are the most stress-
bioactive surface treatment. These challenges and the associated re- concentrated areas (Section 4.1), despite the use of a different SEBM
search opportunities are described below. process (known as the net process) developed specifically by Arcam for
SEBM of Ti-6Al-4V lattices. For example, the nodes manufactured are
8.1. Improving the strength of SEBM Ti-6Al-4V lattices far from smooth and contains many unmelted powder particles (Fig. 11)
[16,41,49]. In that regard, lattice-specific SEBM parameters that can
As discussed in the previous sections, SEBM Ti-6Al-4V lattices can realize the optimal build geometry with desired quality should be de-
be made to show similar mechanical response to human bone and have veloped. The current relative density that can be achieved for SEBM of
been used clinically with success for more than 10 years [68]. But their Ti-6Al-4V is in the range of 99.4–99.8% [146]. This means that pores
strengths remain below those of the strongest human cortical bone. For and lack-of-fusion defects are unavoidable in the struts. The use of pore-
example, even for the currently manufactured best-performance SEBM free feedstock powder is expected to reduce internal porosity. Through
Ti-6Al-4V cubic lattices (porosity: 63.21%; density: 1.63 g·cm−3), their careful control of the SEBM process, it is also possible to minimize the
compressive strength and fatigue strength are still inferior to the re- presence of lack-of-fusion defects. This depends on a more compre-
spective strengths of human cortical tibia and femur bones. Increasing hensive understanding of the SEBM process, which can offer greater
their relative density in conjunction with the use of stretch-dominated insight into the design and manufacture of high-quality lattices for
lattice design [50] or using ultrahigh-strength titanium alloys to make preferred mechanical properties. In addition, there exists a significant
SEBM lattice structures should be able to address this issue according to lack of design tools specifically for the design of patient-specific lattices.
the Gibson-Ashby equation (Eq. (7)). On the other hand, it should be This limitation necessitates manual intervention and sub-optimal de-
noted that there exist 58 different types of human bones and only a few sign techniques that unnecessarily add complexity, risk, and cost to the
of their implants call for superior strength, as shown in Tables 2 and 3. fabrication of AM implants.
For those which do not require high strength, SEBM Ti-6Al-4V lattice
Fig. 15. SEBM Ti-6Al-4V cervical vertebral fusion cage. (a) CAD model including internal lattice structures. (b) As-built implant. (c) Post implantation CT scan [147].
95
8.4. Manufacturable cavity size by SEBM and biomedical needs Appendix A. Supplementary material
The minimum cavity size that can be manufactured in a SEBM Ti-6Al- Supplementary data associated with this article can be found, in the
4V lattice is about 450 µm [27], greater than most cavities in human bone online version, at http://dx.doi.org/10.1016/j.cossms.2018.05.002.
(Fig. 3). This discrepancy limits the manufacture of more biomimetic
lattices. Researchers have concluded that pore size in the range of References
150–250 µm is optimal for (i) the expression and production of collagen
and aggrecan, (ii) supplying oxygen from blood vessels to neighboring [1] M. Geetha, A.K. Singh, R. Asokamani, A.K. Gogia, Ti based biomaterials, the ul-
cells, and (iii) enhancing bone formation and cartilage mechanical prop- timate choice for orthopaedic implants – a review, Prog. Mater Sci. 54 (3) (2009)
397–425.
erties [149–152]. The larger size range (300–400 µm) of pores is more [2] Y. Li, C. Yang, H. Zhao, S. Qu, X. Li, Y. Li, New developments of Ti-based alloys for
suitable for both angiogenesis and bone ingrowth [153,154]. In either biomedical applications, Materials 7 (3) (2014) 1709–1800.
case, it is demanding for the current SEBM practice. One latent risk of [3] M.F. Ashby, The properties of foams and lattices, Philos. Trans. Ser. A, Math. Phys.
Eng. Sci. 364 (1838) (2006) 15–30.
manufacturing small cavities by SEBM is how to remove the unmelted [4] L.J. Gibson, M.F. Ashby, Cellular Solids-Structure and Properties, Cambridge
powder particles from them in a lattice. These particles may have to stay in University Press, 1999.
the lattice and then risk entering the patient’s body. [5] L.E. Murr, Open-cellular metal implant design and fabrication for biomechanical
compatibility with bone using electron beam melting, J. Mech. Behav. Biomed.
As shown in Fig. 3, bone is a hierarchically structured porous ma-
Mater. (2017).
terial ranging from the macroscale to nanoscale [77]. Emerging evi- [6] M.B. Bezuidenhout, D.M. Dimitrov, A.D. van Staden, G.A. Oosthuizen, L.M. Dicks,
dence has confirmed the crucial role that micro- and nanoscale porous Titanium-based hip stems with drug delivery functionality through additive
manufacturing, BioMed Res. Int. 2015 (2015) 1–11.
structures play in regulating cell morphology, adhesion, differentiation,
[7] F. Warmuth, F. Osmanlic, L. Adler, M.A. Lodes, C. Körner, Fabrication and char-
migration, and proliferation [77]. One unique opportunity to produce acterisation of a fully auxetic 3D lattice structure via selective electron beam
lattices with a hierarchical porous structure is to combine SEBM with melting, Smart Mater. Struct. 26 (2) (2017) 025013.
dealloying [156], which is a diffusion-controlled selective dissolution [8] L. Yang, O. Harrysson, H. West, D. Cormier, Mechanical properties of 3D re-en-
trant honeycomb auxetic structures realized via additive manufacturing, Int. J.
and self-organization process [157–160]. Recent innovations using Solids Struct. 69–70 (2015) 475–490.
molten metals as the dissolution medium provide a generic method of [9] L. Yang, O. Harrysson, H. West, D. Cormier, Modeling of uniaxial compression in a
manufacturing nanoporous materials from nearly any metal, for ex- 3D periodic re-entrant lattice structure, J. Mater. Sci. 48 (4) (2012) 1413–1422.
[10] L. Yang, O. Harrysson, H. West, D. Cormier, Compressive properties of Ti–6Al–4V
ample, Ti [161], Ta [162], Nb [163] and Cr [164]. The integration of auxetic mesh structures made by electron beam melting, Acta Mater. 60 (8) (2012)
SEBM and dealloying has the potential to produce hierarchical porous 3370–3379.
structures that replicate the properties of human bones at macro-to- [11] J. Schwerdtfeger, F. Schury, M. Stingl, F. Wein, R.F. Singer, C. Körner, Mechanical
characterisation of a periodic auxetic structure produced by SEBM, Phys. Status
nano scales. The effect of dealloying on implant mechanical properties Solidi (b) 249 (7) (2012) 1347–1352.
is yet to be quantified [2]. [12] J. Schwerdtfeger, P. Heinl, R.F. Singer, C. Körner, Auxetic cellular structures
through selective electron-beam melting, Phys. Status Solidi (b) 247 (2) (2010)
269–272.
8.5. SEBM Ti-6Al-4V lattice structures with bioactive surface [13] M. Mazur, M. Leary, S. Sun, M. Vcelka, D. Shidid, M. Brandt, Deformation and
failure behaviour of Ti-6Al-4V lattice structures manufactured by selective laser
The surface of a SEBM Ti-6Al-4V lattice is bio-inert and the survival melting (SLM), Int. J. Adv. Manuf. Technol. 84 (5–8) (2016) 1391–1411.
[14] M. Leary, M. Mazur, J. Elambasseril, M. McMillan, T. Chirent, Y. Sun, M. Qian,
of cells on it can be an issue, in addition to bacterial infection. It has
M. Easton, M. Brandt, Selective laser melting (SLM) of AlSi12Mg lattice structures,
been shown that a simple surface treatment with HCl and NaOH can Mater. Des. 98 (2016) 344–357.
improve the biological activity of SEBM Ti-6Al-4V lattice implants, [15] L.E. Murr, S.M. Gaytan, F. Medina, H. Lopez, E. Martinez, B.I. Machado,
leading to easier fixation around the concerned bone and improving D.H. Hernandez, L. Martinez, M.I. Lopez, R.B. Wicker, J. Bracke, Next-generation
biomedical implants using additive manufacturing of complex, cellular and
long-term stability [27,91,165]. Coating is another option to enhance functional mesh arrays, Philos. Trans. Ser. A, Math. Phys. Eng. Sci. 368 (1917)
the vascularization and osseointegration [166]. The antibacterial (2010) 1999–2032.
properties of SEBM Ti-6Al-4V lattice implants could be improved by [16] S. Zhao, S.J. Li, W.T. Hou, Y.L. Hao, R. Yang, R.D. Misra, The influence of cell
morphology on the compressive fatigue behavior of Ti-6Al-4V meshes fabricated
introducing silver or copper on the surface of the lattice implant by electron beam melting, J. Mech. Behav. Biomed. Mater. 59 (2016) 251–264.
[167–169]. Silver (Ag) is known for against a broad spectrum of bac- [17] S.J. Li, L.E. Murr, X.Y. Cheng, Z.B. Zhang, Y.L. Hao, R. Yang, F. Medina,
teria while copper is expected to protect the implant from bacterial R.B. Wicker, Compression fatigue behavior of Ti–6Al–4V mesh arrays fabricated
by electron beam melting, Acta Mater. 60 (3) (2012) 793–802.
infection without degradation the properties of the implant. [18] J. Parthasarathy, B. Starly, S. Raman, A design for the additive manufacture of
functionally graded porous structures with tailored mechanical properties for
8.6. SEBM of β-titanium alloy lattice implants biomedical applications, J. Manuf. Processes 13 (2) (2011) 160–170.
[19] J. Parthasarathy, B. Starly, S. Raman, A. Christensen, Mechanical evaluation of
porous titanium (Ti6Al4V) structures with electron beam melting (EBM), J. Mech.
A wide variety of Al-, V- and Ni-free new β-Ti alloys have been Behav. Biomed. Mater. 3 (3) (2010) 249–259.
developed over the last two decades for biomedical applications [2]. [20] S. Amin Yavari, S.M. Ahmadi, R. Wauthle, B. Pouran, J. Schrooten, H. Weinans,
A.A. Zadpoor, Relationship between unit cell type and porosity and the fatigue
They are more biocompatible than Ti-6Al-4V and possess lower Young’s
behavior of selective laser melted meta-biomaterials, J. Mech. Behav. Biomed.
modulus. In addition, some of them further offer good shape memory or Mater. 43 (2015) 91–100.
super elasticity properties. A recent study has reported promising me- [21] E. Sallica-Leva, A.L. Jardini, J.B. Fogagnolo, Microstructure and mechanical be-
chanical properties of SEBM β-Ti alloy lattices [37]. They represent a havior of porous Ti-6Al-4V parts obtained by selective laser melting, J. Mech.
Behav. Biomed. Mater. 26 (2013) 98–108.
new class of titanium lattice materials that have the potential to en- [22] T. Niendorf, F. Brenne, M. Schaper, Lattice structures manufactured by SLM: on
hance both the mechanical and biological performance of lattice im- the effect of geometrical dimensions on microstructure evolution during proces-
plants. Challenges lie in how to translate their unique properties into sing, Metall. Mater. Trans. B 45 (4) (2014) 1181–1185.
[23] S. Tsopanos, R.A.W. Mines, S. McKown, Y. Shen, W.J. Cantwell, W. Brooks,
lattice forms. C.J. Sutcliffe, The influence of processing parameters on the mechanical properties
of selectively laser melted stainless steel microlattice structures, J. Manuf. Sci.
Acknowledgements Eng. 132 (4) (2010) 041011.
[24] R. Stamp, P. Fox, W. O'Neill, E. Jones, C. Sutcliffe, The development of a scanning
strategy for the manufacture of porous biomaterials by selective laser melting, J.
This project was supported by the National Natural Science Mater. Sci. Mater. Med. 20 (9) (2009) 1839–1848.
Foundation of China through No. 51528401 and the Australian [25] W. van Grunsven, E. Hernandez-Nava, G. Reilly, R. Goodall, Fabrication and
mechanical characterisation of titanium lattices with graded porosity, Metals 4 (3)
Research Council (ARC) for funding through ARC LP 140100607. X.Z. (2014) 401–409.
Zhang acknowledges the scholarship received from the China [26] N.W. Hrabe, P. Heinl, B. Flinn, C. Korner, R.K. Bordia, Compression-compression
Scholarship Council (CSC). fatigue of selective electron beam melted cellular titanium (Ti-6Al-4V), J. Biomed.
96
Mater. Res. Part B, Appl. Biomater. 99 (2) (2011) 313–320. [53] P. Lhuissier, C. de Formanoir, G. Martin, R. Dendievel, S. Godet, Geometrical
[27] P. Heinl, L. Muller, C. Korner, R.F. Singer, F.A. Muller, Cellular Ti-6Al-4V struc- control of lattice structures produced by EBM through chemical etching:
tures with interconnected macro porosity for bone implants fabricated by selective Investigations at the scale of individual struts, Mater. Des. 110 (2016) 485–493.
electron beam melting, Acta Biomater. 4 (5) (2008) 1536–1544. [54] H. Hasib, O.L.A. Harrysson, H.A. West, Powder removal from Ti-6Al-4V cellular
[28] P. Heinl, C. Körner, R.F. Singer, Selective electron beam melting of cellular tita- structures fabricated via electron beam melting, Jom 67 (3) (2015) 639–646.
nium: mechanical properties, Adv. Eng. Mater. 10 (9) (2008) 882–888. [55] M. Suard, P. Lhuissier, R. Dendievel, J.J. Blandin, F. Vignat, F. Villeneuve,
[29] N. Taniguchi, S. Fujibayashi, M. Takemoto, K. Sasaki, B. Otsuki, T. Nakamura, Towards stiffness prediction of cellular structures made by electron beam melting
T. Matsushita, T. Kokubo, S. Matsuda, Effect of pore size on bone ingrowth into (EBM), Powder Metall. 57 (3) (2014) 190–195.
porous titanium implants fabricated by additive manufacturing: an in vivo ex- [56] S.-I. Park, D.W. Rosen, S.-K. Choi, C.E. Duty, Effective mechanical properties of
periment, Mater. Sci. Eng. C, Mater. Biol. Appl. 59 (2016) 690–701. lattice material fabricated by material extrusion additive manufacturing, Addit.
[30] R. Wauthle, B. Vrancken, B. Beynaerts, K. Jorissen, J. Schrooten, J.-P. Kruth, Manuf. 1–4 (2014) 12–23.
J. Van Humbeeck, Effects of build orientation and heat treatment on the micro- [57] G. Strano, L. Hao, R.M. Everson, K.E. Evans, Surface roughness analysis, modelling
structure and mechanical properties of selective laser melted Ti6Al4V lattice and prediction in selective laser melting, J. Mater. Process. Technol. 213 (4)
structures, Addit. Manuf. 5 (2015) 77–84. (2013) 589–597.
[31] V.J. Challis, X. Xu, L.C. Zhang, A.P. Roberts, J.F. Grotowski, T.B. Sercombe, High [58] EIT, < http://www.prweb.com/releases/prweb14517869.htm > (accessed: July,
specific strength and stiffness structures produced using selective laser melting, 2017).
Mater. Des. 63 (2014) 783–788. [59] Renovis, < http://www.renovis-surgical.com/2016/03/renovis-surgical-receives-
[32] http://www.metal-am.com/news/003768.html#sthash.d6uXDQND.dpuf. fda-clearance-for-first-ever-3d-printed-titanium-stand-alone-cervical-cage/ > (ac-
[33] L. Murr, S. Li, Y. Tian, K. Amato, E. Martinez, F. Medina, Open-cellular Co-base cessed: March, 2016).
and Ni-base superalloys fabricated by electron beam melting, Materials 4 (12) [60] 4WEB, < https://4webmedical.com/about/ > (accessed: July, 2016).
(2011) 782–790. [61] Joimax, < http://www.3ders.org/articles/20150714-spine-device-company-
[34] T.J. Horn, O.L.A. Harrysson, D.J. Marcellin-Little, H.A. West, B.D.X. Lascelles, joimax-gets-fda-approval-for-3d-printed-spinal-implant.html > (accessed: July,
R. Aman, Flexural properties of Ti6Al4V rhombic dodecahedron open cellular 2015).
structures fabricated with electron beam melting, Addit. Manuf. 1–4 (2014) 2–11. [62] K2M, < http://investors.k2m.com/releasedetail.cfm?ReleaseID=973648 > (ac-
[35] X.Y. Cheng, S.J. Li, L.E. Murr, Z.B. Zhang, Y.L. Hao, R. Yang, F. Medina, cessed: June, 2016).
R.B. Wicker, Compression deformation behavior of Ti-6Al-4V alloy with cellular [63] Solidscape, < https://www.solid-scape.com/experiences/california-medical-
structures fabricated by electron beam melting, J. Mech. Behav. Biomed. Mater. 16 company-goes-high-tech-with-cadcam/ > (accessed: March, 2015).
(2012) 153–162. [64] F.H. Froes, M. Qian, Titanium in Medical and Dental Applications, Woodhead
[36] O.L.A. Harrysson, O. Cansizoglu, D.J. Marcellin-Little, D.R. Cormier, H.A. West, Publishing, Elsevier, 2017.
Direct metal fabrication of titanium implants with tailored materials and me- [65] M. Qian, W. Xu, M. Brandt, H. Tang, Additive manufacturing and post-processing
chanical properties using electron beam melting technology, Mater. Sci. Eng., C 28 of Ti-6Al-4V for superior mechanical properties, MRS Bull. (2016) 775–784.
(3) (2008) 366–373. [66] H.P. Tang, J. Wang, C.N. Song, N. Liu, L. Jia, J. Elambasseril, M. Qian,
[37] Y.J. Liu, H.L. Wang, S.J. Li, S.G. Wang, W.J. Wang, W.T. Hou, Y.L. Hao, R. Yang, Microstructure, mechanical properties, and flatness of SEBM Ti-6Al-4V Sheet in
L.C. Zhang, Compressive and fatigue behavior of beta-type titanium porous As-built and hot isostatically pressed conditions, JOM 69 (3) (2017).
structures fabricated by electron beam melting, Acta Mater. 126 (2017) 58–66. [67] S.L. Lu, H.P. Tang, Y.P. Ning, N. Liu, D.H. StJohn, M. Qian, Microstructure and
[38] S.A. Yavari, R. Wauthle, J. van der Stok, A.C. Riemslag, M. Janssen, M. Mulier, mechanical properties of long Ti-6Al-4V rods additively manufactured by selective
J.P. Kruth, J. Schrooten, H. Weinans, A.A. Zadpoor, Fatigue behavior of porous electron beam melting out of a deep powder bed and the effect of subsequent hot
biomaterials manufactured using selective laser melting, Mater. Sci. Eng. C, Mater. isostatic pressing, Metall. Mater. Trans. A 46 (9) (2015) 3824–3834.
Biol. Appl. 33 (8) (2013) 4849–4858. [68] Metal AM, < http://www.metal-am.com/celebrating-ten-years-of-metal-
[39] J. Van der Stok, O.P. Van der Jagt, S. Amin Yavari, M.F. De Haas, J.H. Waarsing, additively-manufactured-hip-cups/ > (accessed: March 2018).
H. Jahr, E.M. Van Lieshout, P. Patka, J.A. Verhaar, A.A. Zadpoor, H. Weinans, [69] http://www.arcam.com/company/about-arcam/history/.
Selective laser melting-produced porous titanium scaffolds regenerate bone in [70] C. Körner, Additive manufacturing of metallic components by selective electron
critical size cortical bone defects, J. Orthopaed. Res.: Off. Publ. Orthopaed. Res. beam melting—a review, Int. Mater. Rev. (2016) 1–17.
Soc. 31 (5) (2013) 792–799. [71] S. Tammas-Williams, H. Zhao, F. Léonard, F. Derguti, I. Todd, P.B. Prangnell, XCT
[40] G. Campoli, M.S. Borleffs, S. Amin Yavari, R. Wauthle, H. Weinans, A.A. Zadpoor, analysis of the influence of melt strategies on defect population in Ti–6Al–4V
Mechanical properties of open-cell metallic biomaterials manufactured using ad- components manufactured by Selective Electron Beam Melting, Mater. Charact.
ditive manufacturing, Mater. Des. 49 (2013) 957–965. 102 (2015) 47–61.
[41] S.J. Li, Q.S. Xu, Z. Wang, W.T. Hou, Y.L. Hao, R. Yang, L.E. Murr, Influence of cell [72] http://www.arcam.com/technology/electron-beam-melting/.
shape on mechanical properties of Ti-6Al-4V meshes fabricated by electron beam [73] S. Tammas-Williams, P.J. Withers, I. Todd, P.B. Prangnell, The effectiveness of hot
melting method, Acta Biomater. 10 (10) (2014) 4537–4547. isostatic pressing for closing porosity in titanium parts manufactured by selective
[42] C. Yan, L. Hao, A. Hussein, P. Young, D. Raymont, Advanced lightweight 316L electron beam melting, Metall. Mater. Trans. A 47 (5) (2016) 1939–1946.
stainless steel cellular lattice structures fabricated via selective laser melting, [74] Arcam training session, 2012.
Mater. Des. 55 (2014) 533–541. [75] J. Black, G. Hastings, Handbook of Biomaterial Properties, Thomson Science, UK,
[43] O. Cansizoglu, O. Harrysson, D. Cormier, H. West, T. Mahale, Properties of 1998.
Ti–6Al–4V non-stochastic lattice structures fabricated via electron beam melting, [76] J.-Y. Rho, L. Kuhn-Speaing, P. Zioupos, Mechanical properties and the hierarchical
Mater. Sci. Eng., A 492 (1–2) (2008) 468–474. structure of bone, Med. Eng. Phys. 20 (1998) 92–102.
[44] D.C.O. Cansizoglu, O. Harrysson, An evaluation of non-stochastic lattice structures [77] X. Wang, S. Xu, S. Zhou, W. Xu, M. Leary, P. Choong, M. Qian, M. Brandt, Y.M. Xie,
fabricated via electron beam melting, Solid Freeform Fabric. (2006) 209–219. Topological design and additive manufacturing of porous metals for bone scaffolds
[45] L.A. Dobrzański, A.D. Dobrzańska-Danikiewicz, T.G. Gaweł, A. Achtelik-Franczak, and orthopaedic implants: a review, Biomaterials 83 (2016) 127–141.
Selective laser sintering and melting of pristine titanium and titanium Ti6Al4V [78] F.G. Evans, Mechanical properties and histology of cortical bone from younger and
alloy powders and selection of chemical environment for etching of such mate- older men, Anatom. Rec. 185 (1976) 1–11.
rials, Arch. Metall. Mater. 60 (3) (2015). [79] A.B.D.R.M. Martens, Aging of bone tissue: mechanical properties, J. Bone Joint
[46] S. Van Bael, Y.C. Chai, S. Truscello, M. Moesen, G. Kerckhofs, H. Van Oosterwyck, Surg. 58 (1) (1976) 82–86.
J.P. Kruth, J. Schrooten, The effect of pore geometry on the in vitro biological [80] O. Lindahl, A.G.H. Lindgren, Cortical Bone in Man 1.Variation of the Amount and
behavior of human periosteum-derived cells seeded on selective laser-melted Density with Age and Sex, Acta Orthop. Scand. 38 (1–4) (1967) 133–140.
Ti6Al4V bone scaffolds, Acta Biomater. 8 (7) (2012) 2824–2834. [81] O. Lindahl, A.G.H. Lindgren, Cortical Bone in Man II.Variation in Tensile Strength
[47] M. Jamshidinia, L. Wang, W. Tong, R. Kovacevic, The bio-compatible dental im- with Age and Sex, Acta Orthop. Scand. 38 (1–4) (1967) 141–147.
plant designed by using non-stochastic porosity produced by Electron Beam [82] O. Lindahl, Mechanical Properties of Dried Defatted Spongy Bone, Acta Orthop.
Melting® (EBM), J. Mater. Process. Technol. 214 (8) (2014) 1728–1739. Scand. 47 (1) (1976) 11–19.
[48] H.B. Hasib, Mechanical Behavior of Non-Stochastic Ti-6Al-4V Cellular Structures [83] M. Ding, M. Dalstra, C.C. Danielsen, Age variations in the properties of human
Produced via Electron Beam Melting (EBM), North Carolina State University, tibial trabecular bone, J. Bone Joint Surg. 79 (1997) 995–1002.
2011. [84] L. Frank, I. Hvid, B. Pongsoipetch, Energy absorptive properties of human trabe-
[49] C. de Formanoir, M. Suard, R. Dendievel, G. Martin, S. Godet, Improving the cular bone l, J. Orthopaed. Res. 7 (3) (1989) 432–439.
mechanical efficiency of electron beam melted titanium lattice structures by [85] R. Antonius, H. Zilch, G. Bergmann, R. Kolbel, Material properties of femoral
chemical etching, Addit. Manuf. 11 (2016) 71–76. cancellous bone in axial loading Part I: Time independent properties, Arch.
[50] S. Arabnejad, R. Burnett Johnston, J.A. Pura, B. Singh, M. Tanzer, D. Pasini, High- Orthopaed. Traum. Surg. 97 (2) (1980) 95–102.
strength porous biomaterials for bone replacement: a strategy to assess the in- [86] L. Mosekilde, L. Mosekilde, C.C. Danielsen, Biomechanical competence of ver-
terplay between cell morphology, mechanical properties, bone ingrowth and tebral trabecular bone in relation to ash density and age in normal individuals,
manufacturing constraints, Acta Biomater. 30 (2016) 345–356. Bone 8 (2) (1987) 79–85.
[51] X. Zheng, W. Smith, J. Jackson, B. Moran, H. Cui, D. Chen, J. Ye, N. Fang, [87] T.H. Hansson, T.S. Keller, M.M. Panjabi, A study of the compressive properties of
N. Rodriguez, T. Weisgraber, C.M. Spadaccini, Multiscale metallic metamaterials, lumbar vertebral trabeculae: effects of tissue characteristics, Spine 12 (1) (1987)
Nat. Mater. 15 (10) (2016) 1100–1106. 56–62.
[52] S.K. Maitin, L.J. Gibson, M.F. Ashby, Deformation and energy absorption diagrams [88] A. Zargarian, M. Esfahanian, J. Kadkhodapour, S. Ziaei-Rad, Numerical simulation
for cellular solids, Acta Metall. 32 (11) (1984) 1963–1975. of the fatigue behavior of additive manufactured titanium porous lattice
97
structures, Mater. Sci. Eng. C, Mater. Biol. Appl. 60 (2016) 339–347. additive manufacturing of metallic implants: a review on processes, materials and
[89] E. Hernández-Nava, C.J. Smith, F. Derguti, S. Tammas-Williams, F. Leonard, designs, J. Orthopaed. Res.: Off. Publ. Orthopaed. Res. Soc. 34 (3) (2016)
P.J. Withers, I. Todd, R. Goodall, The effect of defects on the mechanical response 369–385.
of Ti-6Al-4V cubic lattice structures fabricated by electron beam melting, Acta [122] X. Gong, T. Anderson, K. Chou, Review on powder-based electron beam additive
Mater. 108 (2016) 279–292. manufacturing technology, Manuf. Rev. 1 (2014) 2.
[90] L.E. Murr, K.N. Amato, S.J. Li, Y.X. Tian, X.Y. Cheng, S.M. Gaytan, E. Martinez, [123] C.J. Smith, F. Derguti, E. Hernandez Nava, M. Thomas, S. Tammas-Williams,
P.W. Shindo, F. Medina, R.B. Wicker, Microstructure and mechanical properties of S. Gulizia, D. Fraser, I. Todd, Dimensional accuracy of Electron Beam Melting
open-cellular biomaterials prototypes for total knee replacement implants fabri- (EBM) additive manufacture with regard to weight optimized truss structures, J.
cated by electron beam melting, J. Mech. Behav. Biomed. Mater. 4 (7) (2011) Mater. Process. Technol. 229 (2016) 128–138.
1396–1411. [124] M. Suard, G. Martin, P. Lhuissier, R. Dendievel, F. Vignat, J.J. Blandin,
[91] P. Heinl, A. Rottmair, C. Körner, R.F. Singer, Cellular titanium by selective elec- F. Villeneuve, Mechanical equivalent diameter of single struts for the stiffness
tron beam melting, Adv. Eng. Mater. 9 (5) (2007) 360–364. prediction of lattice structures produced by Electron Beam Melting, Addit. Manuf.
[92] A.G.E.M.F. Ashby, N.A. Fleck, L.J. Gibson, Metal Foams: A Design Guide, Elsevier, 8 (2015) 124–131.
2000. [125] K. Mustafa, A. Wennerberg, J. Wroblewski, Determining optimal surface rough-
[93] M.F. Ashby, R.E.M. Medalist, The mechanical properties of cellular solids, Metall. ness of TiO2 blasted titanium implant material for attachment, proliferation and
Trans. A 14 (9) (1983) 1755–1769. differentiation of cells derived from human mandibular alveolar bone, Clin. Oral.
[94] Y. Quan, P. Drescher, F. Zhang, Cellular Ti6Al4V with carbon nanotube-like Impl. Res. 12 (2001) 515–525.
structures fabricated by selective electron beam melting, Rapid Prototyp. J. 20 (6) [126] Y.Y. Sun, S. Gulizia, C.H. Oh, D. Fraser, M. Leary, Y.F. Yang, M. Qian, The
(2014) 541–550. Influence of As-built surface conditions on mechanical properties of Ti-6Al-4V
[95] H.D. Carlton, J. Lind, M.C. Messner, N.A. Volkoff-Shoemaker, H.S. Barnard, additively manufactured by selective electron beam melting, Jom 68 (3) (2016)
N.R. Barton, M. Kumar, Mapping local deformation behavior in single cell metal 791–798.
lattice structures, Acta Mater. 129 (2017) 239–250. [127] G. Pyka, G. Kerckhofs, I. Papantoniou, M. Speirs, J. Schrooten, M. Wevers, Surface
[96] S. Li, S. Zhao, W. Hou, C. Teng, Y. Hao, Y. Li, R. Yang, R.D.K. Misra, Functionally roughness and morphology customization of additive manufactured open porous
graded Ti-6Al-4V meshes with high strength and energy absorption, Adv. Eng. Ti6Al4V structures, Materials 6 (10) (2013) 4737–4757.
Mater. 18 (1) (2016) 34–38. [128] G.A. Longhitano, M.A. Larosa, A.L.J. Munhoz, C.A.D.C. Zavaglia, M.C.F. Ierardi,
[97] M. Jamshidinia, L. Wang, W. Tong, R. Ajlouni, R. Kovacevic, Fatigue properties of Surface finishes for Ti-6Al-4V alloy produced by direct metal laser sintering,
a dental implant produced by electron beam melting® (EBM), J. Mater. Process. Mater. Res. 18 (4) (2015) 838–842.
Technol. 226 (2015) 255–263. [129] A. Safdar, H.Z. He, L.Y. Wei, A.S. Luis, E.C.D. Paz, Effect of process parameters
[98] R.K. Mahdi Jamshidinia, The numerical modeling of fatigue properties of a bio- settings and thickness on surface roughness of EBM produced Ti-6Al-4V, Papid
compatible dental implant produced by Electron Beam Melting®(EBM), in: Prototyp. J. 18 (5) (2012) 401–408.
Proceeding of the Twenty Forth Annual International Solid Freeform Fabrication [130] J. Karlsson, A. Snis, H. Engqvist, J. Lausmaa, Characterization and comparison of
Symposium, August 12–14, Austin, TX, 2013, p. 15. materials produced by Electron Beam Melting (EBM) of two different Ti–6Al–4V
[99] K.S. Chan, M. Koike, R.L. Mason, T. Okabe, Fatigue life of titanium alloys fabri- powder fractions, J. Mater. Process. Technol. 213 (12) (2013) 2109–2118.
cated by additive layer manufacturing techniques for dental implants, Metall. [131] J. Springer, In vitro strudies of titanium alloy discs, fabricated with electron beam
Mater. Trans. A 44 (2) (2012) 1010–1022. melting for improving transdermal osseointegrated implant surfaces, Dissertation
[100] G. Menges, F. Knipschild, Estimation of mechanical properties for rigid poly- North Carolina State University, 2010.
urethane foams, Polym. Eng. Sci. 15 (1975) 623–627. [132] S.J. Ferguson, N. Broggini, M. Wieland, Biomechanical evaluation of the inter-
[101] W.L. Ko, Deformation of foamed elastomers, J. Cell. Plast. 1 (1965) 45–50. facial strength of a chemically modified sandblasted and acid-etched titanium
[102] L.J. Gibson, M.F. Ashby, G.S. Schajer, The mechanics of two-dimensional cellular surface, J. Biomed. Mater. Res. 78 (2) (2006) 291–297.
materials, Proc. R. Soc. Lond. Ser. A, Math. Phys. Sci. 382 (1782) (1982) 25–42. [133] D.D. Deligianni, N.D. Katsala, P.G. Koutsoukos, Y.F. Missirlis, Effect of surface
[103] L.J. Gibson M.F. Ashby The mechanics of three-dimensional cellular materials roughness of the titanium alloy Ti-6Al-4V on human bone marrow cell response
Proc. R. Soc. Lond. Ser. A, Math. Phys. Sci. 382 1782 1982 43-59. and on protein adsorption, Biomaterials 22 (11) (2001) 1241–1251.
[104] M. Leary, C. Burvill, Applicability of published data for fatigue-limited design, [134] A. Wennerberg, A. Ektessabi, T. Albrektsson, A 1-year fellow-up of implants of
Qual. Reliab. Eng. Int. 25 (8) (2009) 921–932. differing surface roughness placed in rabbit bone, Int. J. Oral & Maxllofac.
[105] ISO 13314, Mechanical testing of metals – ductility testing – compression test for Implants 12 (4) (1997) 486–494.
porous and cellular metals, 2011. [135] N.C. Blumenthal, V. Cosma, Inhibition of apatite formation by titanium and va-
[106] X. Li, Y. Luo, C. Wang, Preparation and characterization of porous Ti6Al4V/algi- nadium ions, J. Biomed. Mater. Res. Part A 23 (S13) (1989) 13–22.
nate hybrid implant by combination of electron beam melting and freeze-drying, [136] I. Drstvensek, N.I. Hren, T. Strojnik, Application of rapid prototyping in cranio-
Mater. Lett. 81 (2012) 23–26. maxilofacial surgery procedures, Int. J. Biol. Biomed. Eng. 1 (2) (2008) 29–38.
[107] ISO 14801, Dentistry – implants – dynamic loading test for endosseous dental [137] S. Palanivel, A.K. Dutt, E.J. Faierson, R.S. Mishra, Spatially dependent properties
implants, 2016. in a laser additive manufactured Ti–6Al–4V component, Mater. Sci. Eng., A 654
[108] P. Zioupos, M. Gresle, K. Winwood, Fatigue strength of human cortical bone: age, (2016) 39–52.
physical, and material heterogeneity effects, J. Biomed. Mater. Res. Part A 86 (3) [138] G. Kasperovich, J. Hausmann, Improvement of fatigue resistance and ductility of
(2008) 627–636. TiAl6V4 processed by selective laser melting, J. Mater. Process. Technol. 220
[109] K.K. Choi, S.A. Goldstein, A comparison of the fatigue behavior of human trabe- (2015) 202–214.
cular and cortical bone tissue, J. Biomech. 25 (12) (1992) 1371–1381. [139] G. Pyka, A. Burakowski, G. Kerckhofs, M. Moesen, S. Van Bael, J. Schrooten,
[110] K. Rüdinger, D. Fischer, Relationship between primary alpha content, tensile M. Wevers, Surface modification of Ti6Al4V open porous structures produced by
properties and high cycle fatigue behaviour of Ti-6Al-4V, Titanium: Sci. Technol. 4 additive manufacturing, Adv. Eng. Mater. 14 (6) (2012) 363–370.
(1985) 2123–2130. [140] R. Cunningham, S.P. Narra, T. Ozturk, J. Beuth, A.D. Rollett, Evaluating the effect
[111] G.K. Haritos, N. Theodore, D.B. Lanning, Notch size effects in HCF behavior of of processing parameters on porosity in electron beam melted Ti-6Al-4V via syn-
Ti–6Al–4V, Int. J. Fatigue 21 (7) (1999) 643–652. chrotron X-ray microtomography, Jom 68 (3) (2016) 765–771.
[112] G.Q. Wu, C.L. Shi, W. Sha, A.X. Sha, H.R. Jiang, Effect of microstructure on the [141] T. Saito, T. Furuta, Sintered powdered titanium alloy and method of producing the
fatigue properties of Ti–6Al–4V titanium alloys, Mater. Des. 46 (2013) 668–674. same United State, 1995.
[113] K.E. Evans, A. Alderson, Auxetic materials: functional materials and structures [142] H. Galarraga, D.A. Lados, R.R. Dehoff, M.M. Kirka, P. Nandwana, Effects of the
from lateral thinking, Adv. Mater. 12 (9) (2000) 617–628. microstructure and porosity on properties of Ti-6Al-4V ELI alloy fabricated by
[114] Y. Li, H. Hong, A review on auxetic structures and polymeric materials, Sci. Res. electron beam melting (EBM), Addit. Manuf. 10 (2016) 47–57.
Essays 5 (10) (2010) 1052–1063. [143] J.A. Slotwinski, E.J. Garboczi, K.M. Hebenstreit, Porosity measurements and
[115] I. Polmear, D.H. St John, J.F. Nie, M. Qian, Light Alloys: Metallurgy of Light analysis for metal additive manufacturing process control, J. Res. National Inst.
Metals, Elsevier, Butterworth-Heinemann, 2017 (Chapter 7). Stand. Technol. 119 (2014) 494–528.
[116] T. Ahmed, H.J. Rack, Phase transformations during cooling in α + β titanium [144] H. Gong, K. Rafi, H. Gu, T. Starr, B. Stucker, Analysis of defect generation in
alloys, Mater. Sci. Eng., A 243 (1998) 206–211. Ti–6Al–4V parts made using powder bed fusion additive manufacturing processes,
[117] J. Sieniawski, W. Ziaja, K. Kubiak, M. Motyk, Titanium Alloys-Advances in Addit. Manuf. 1–4 (2014) 87–98.
Properties Control, In Tech, 2013. [145] M. Svensson, U. Ackelid, Titanium alloys manufactured with electron beam
[118] R. Dąbrowski, The kinetics of phase transformations during continuous cooling of melting: mechanical and chemical properties, Fabric. Process. Med. Dev. Conf.
the Ti6Al4V alloy from the single-phase β range, Arch. Metall. Mater. 56 (3) (2009) 189–194.
(2011). [146] Y.Y. Sun, S. Gulizia, D. Fraser, C.H. Oh, S.L. Lu, M. Qian, Layer additive production
[119] M. Yan, W. Xu, M.S. Dargusch, H.P. Tang, M. Brandt, M. Qian, Review of effect of or manufacturing of thick sections of Ti-6Al-4V by selective electron beam melting
oxygen on room temperature ductility of titanium and titanium alloys, Powder (SEBM), Jom (2017) 1–8.
Metall. 57 (4) (2014) 251–257. [147] H.P. Tang, P. Zhao, C.S. Xiang, N. Liu, L. Jia, Ti-6Al-4V orthopedic implants made
[120] M. Yan, M.S. Dargusch, T. Ebel, M. Qian, A transmission electron microscopy and by selective electron beam melting, in: F.H. Froes, M. Qian (Eds.), In Titanium for
three-dimensional atom probe study of the oxygen-induced fine microstructural Medical and Dental Applications, Woodhead Publishing, Elsevier, 2017, pp. 239-
features in as-sintered Ti–6Al–4V and their impacts on ductility, Acta Mater. 68 249.
(2014) 196–206. [148] R.G. Thompson, Anatomics 3D printed titanium implants from head to heel, in: F.
[121] S.L. Sing, J. An, W.Y. Yeong, F.E. Wiria, Laser and electron-beam powder-bed H. Froes, M. Qian (Eds.), In Titanium for Medical and Dental Applications,
98
Woodhead Publishing, Elsevier, 2017, pp. 225-237. [161] T. Wada, A.D. Setyawan, K. Yubuta, H. Kato, Nano- to submicro-porous β-Ti alloy
[149] Q. Zhang, H. Lu, N. Kawazoe, G. Chen, Pore size effect of collagen scaffolds on prepared from dealloying in a metallic melt, Scr. Mater. 65 (6) (2011) 532–535.
cartilage regeneration, Acta Biomater. 10 (5) (2014) 2005–2013. [162] I. McCue, S. Ryan, K. Hemker, X. Xu, N. Li, M. Chen, J. Erlebacher, Size effects in
[150] J. Rouwkema, N.C. Rivron, C.A. van Blitterswijk, Vascularization in tissue en- the mechanical properties of bulk bicontinuous Ta/Cu nanocomposites made by
gineering, Trends Biotechnol. 26 (8) (2008) 434–441. liquid metal dealloying, Adv. Eng. Mater. 18 (1) (2016) 46–50.
[151] J.J. Folkman, M. Hochberg, Self-regulation of growth in three dimensions, J. Exp. [163] J.W. Kim, T. Wada, S.G. Kim, H. Kato, Sub-micron porous niobium solid electro-
Med. 138 (4) (1973) 745–753. lytic capacitor prepared by dealloying in a metallic melt, Mater. Lett. 116 (2014)
[152] C.K. Colton, Implantable biohybrid artificial organs, Cell Transplant. 4 (4) (1995) 223–226.
415–436. [164] T. Wada, H. Kato, Three-dimensional open-cell macroporous iron, chromium and
[153] Y. Kuboki, Q.Q. Jin, H. Takita, Geometry of carriers controlling phenotypic ex- ferritic stainless steel, Scr. Mater. 68 (9) (2013) 723–726.
pression in BMP-induced osteogenesis and chondrogenesis, J. Bone Joint Surg. 83- [165] X. Li, C. Wang, W. Zhang, Y. Li, Fabrication and characterization of porous
A (1 suppl 2) (2001) S105–S115. Ti6Al4V parts for biomedical applications using electron beam melting process,
[154] E. Tsuruga, H. Takita, H. Itoh, Y. Wakisaka, Y. Kuboki, Pore size of porous hy- Mater. Lett. 63 (3–4) (2009) 403–405.
droxyapatite as the cell-substratum controls BMP-induced osteogenesis, J. [166] J. Lv, P. Xiu, J. Tan, Z.J. Jia, H. Cai, Z.J. Liu, Enhanced angiogenesis and osteo-
Biochem. 121 (2) (1997) 317–324. genesis in critical bone defects by the controlled release of BMP-2 and VEGF:
[156] L.Y. Chen, X.Y. Lang, M.W. Chen, Dealloyed nanoporous metals, in: Q. Xu (Ed.), implantation of electron beam melting-fabricated porous Ti6Al4V scaffolds in-
Nanoporous Materials: Synthesis and Applications, Taylor and Francis Group, corporating growth factor-doped fibrin glue, Biomed. Mater. 10 (3) (2015)
2012, pp. 125–181. 035013.
[157] J. Erlebacher, M.J. Aziz, A. Karma, N. Dimitrov, K. Sieradzki, Evolution of nano- [167] S. Mei, H. Wang, W. Wang, L. Tong, H. Pan, C. Ruan, Q. Ma, M. Liu, H. Yang,
porosity in dealloying, Nature 410 (6827) (2001) 450–453. L. Zhang, Y. Cheng, Y. Zhang, L. Zhao, P.K. Chu, Antibacterial effects and bio-
[158] T. Song, M. Yan, Z. Shi, A. Atrens, M. Qian, Creation of bimodal porous copper compatibility of titanium surfaces with graded silver incorporation in titania na-
materials by an annealing-electrochemical dealloying approach, Electrochim. Acta notubes, Biomaterials 35 (14) (2014) 4255–4265.
164 (2015) 288–296. [168] L. Zhao, H. Wang, K. Huo, L. Cui, W. Zhang, H. Ni, Y. Zhang, Z. Wu, P.K. Chu,
[159] N.S. Dhillon, J. Buongiorno, K.K. Varanasi, Critical heat flux maxima during Antibacterial nano-structured titania coating incorporated with silver nano-
boiling crisis on textured surfaces, Nat. Commun. 6 (2015) 8247. particles, Biomaterials 32 (24) (2011) 5706–5716.
[160] T.T. Song, Y.L. Gao, Z.H. Zhang, Q.J. Zhai, Microstructure and phase evolution [169] M. Ip, S.L. Lui, V.K. Poon, I. Lung, A. Burd, Antimicrobial activities of silver
during the dealloying of bi-phase Al–Ag alloy, Corros. Sci. 68 (2013) 256–262. dressings: an in vitro comparison, J. Med. Microbiol. 55 (Pt 1) (2006) 59–63.
99

1 s2.0 S1359028617301870 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S1359028617301870 Main

Uploaded by

Copyright:

Available Formats

Current Opinion in Solid State & Materials Science 22 (2018) 75–99

Contents lists available at ScienceDirect

Current Opinion in Solid State & Materials Science

Selective electron beam manufactured Ti-6Al-4V lattice structures for T

Nomenclature TLIF transforaminal lumbar interbody fusion

2. The SEBM process of Ti-6Al-4V lattices

Gyroid Hatched Pure Titanium SLM [45]

Fig. 2. (a) Preheating strategy [70,74]. (b) Melting strategy [71].

Fig. 3. Hierarchical structures of human cortical and cancellous bone [76,77].

Humerus 15–89 male – 149 15.60 1.20 1.77 – 0.50 [80,81]

0.52 88.26 Cross 1 –a – 0.58b – – – – yc – y – [15]

(continued on next page)

1.46 67.04 Diamond 46 ± 5 – 2.28 ± 0.84 – – – y y y y – [25]

These comparisons show the deﬁciencies of current SEBM Ti-6Al-4V

Lattice structure Human bone

α′-martensite oﬀers high strength but low ductility. Post heat

Cubic 1.31 83.13 1198 –a –

Ti-6.1Al-4.3V dT/dt ≥ 18.0 β → α′ Fully martensitic (α′) [117]

Ti-6.5Al-4.4V dT/dt ≥23.1 β → α′ Fully martensitic (α′) [118]

d h The dimensional accuracy of a SEBM Ti-6Al-4V lattice can be

Q10 –a – – 3.0 450 50 – 45–105 0.3 0, 90 A/L/Mb [7]

Diamond 0.200 0.398 ± 0.037 0.198 ± 0.037 99.00 [25]

Surface polishing works [139] but it is challenging to deal with those

6.4. Defects in SEBM Ti-6Al-4V lattices

Similar to dense SEBM Ti-6Al-4V, the major defects in SEBM Ti-6Al-

7. Biomedical applications of SEBM Ti-6Al-4V lattices

There have been signiﬁcant clinical applications of SEBM Ti-6Al-4V

Compared with AM of dense components, AM of Ti-6Al-4V lattices

You might also like