materials

Article
Manufacturing Feasibility and Forming Properties of
Cu-4Sn in Selective Laser Melting
Zhongfa Mao 1 , David Z. Zhang 1,2, *, Peitang Wei 1 and Kaifei Zhang 1
1 State Key Laboratory on Mechanical Transmission, Chongqing University, Chongqing 400044, China;
zhongfamao@cqu.edu.cn (Z.M.); peitangwei@cqu.edu.cn (P.W.); kaifeizhang@cqu.edu.cn (K.Z.)
2 College of Engineering, Mathematics and Physical Sciences, University of Exeter, North Park Road,
Exeter EX4 4QF, UK
* Correspondence: zhangzw@cqu.edu.cn; Tel.: +86-23-6510-2537

Academic Editor: Guillermo Requena

Received: 27 February 2017; Accepted: 15 March 2017; Published: 24 March 2017

Abstract: Copper alloys, combined with selective laser melting (SLM) technology, have attracted
increasing attention in aerospace engineering, automobile, and medical fields. However, there are
some difficulties in SLM forming owing to low laser absorption and excellent thermal conductivity.
It is, therefore, necessary to explore a copper alloy in SLM. In this research, manufacturing feasibility
and forming properties of Cu-4Sn in SLM were investigated through a systematic experimental
approach. Single-track experiments were used to narrow down processing parameter windows.
A Greco-Latin square design with orthogonal parameter arrays was employed to control forming
qualities of specimens. Analysis of variance was applied to establish statistical relationships, which
described the effects of different processing parameters (i.e., laser power, scanning speed, and
hatch space) on relative density (RD) and Vickers hardness of specimens. It was found that Cu-4Sn
specimens were successfully manufactured by SLM for the first time and both its RD and Vickers
hardness were mainly determined by the laser power. The maximum value of RD exceeded 93%
theoretical density and the maximum value of Vickers hardness reached 118 HV 0.3/5. The best
tensile strength of 316–320 MPa is inferior to that of pressure-processed Cu-4Sn and can be improved
further by reducing defects.

Keywords: selective laser melting; additive manufacturing; copper alloy; relative density;
Vickers hardness

1. Introduction
Selective laser melting (SLM) is a unique additive manufacturing (AM) technology that enables
us to produce dense metal parts by means of layer-by-layer construction using metal powders
based on CAD models [1–3]. In comparison with traditional manufacturing methods, SLM offers
a wide range of advantages. It can directly manufacture highly complex parts that are difficult
or unable to be fabricated by conventional methods, with mechanical properties comparable to
those of wrought materials. It does not require expensive tooling and machining in manufacturing
and, therefore, saves manufacturing costs and lead-time [4,5]. It is also a favorable method for the
fabrication of functionally-graded multi-material parts [6]. Furthermore, SLM technology offers high
degrees of freedom for designers throughout the product development, from conceptual design to
low-volume trial production [7]. These advantages make SLM a very useful method for the fabrication
of biocompatible implants, conformal cooling channels of heat exchangers, light-weight structures
in biomedicine, molds, the aerospace industry, and other fields [2,8]. However, current studies on
SLM are limited and mainly focus on steel and iron-based alloys, titanium and its alloys, and Inconel

Materials 2017, 10, 333; doi:10.3390/ma10040333 www.mdpi.com/journal/materials

Materials 2017, 10, 333 2 of 12

and nickel-based alloys [2]. Copper alloys have recently appeared as a new type of potential material
for SLM.
Copper alloys, compared with other metal alloys, exhibit moderate mechanical properties, high
corrosion resistance, excellent electrical and thermal conductivity, as well as outstanding machinability
and formability [9]. As such, these alloys are widely used in electronics, machinery, aerospace, defense,
and other industrial fields, such as heat exchangers for various types of equipment, high-precision
springs and bearings, electronic connectors, plastic deformation tools, and propulsion devices in
marine applications [10,11]. However, fabrication methods of copper alloys are limited in conventional
powder metallurgy technology, cold compaction, sintering, and infiltration. These methods are not
appropriate for manufacturing parts with intricate shapes. Therefore, a large number of valuable
studies were reported about copper alloys in direct metal laser sintering (DLMS), including balling
phenomenon, densification mechanisms, and microstructural characteristics during the sintering
process [12,13]. Meanwhile, the combination of design flexibility, excellent process capabilities, and
the fully dense mechanism offered by SLM makes SLM a very attractive potential method for creating
copper alloy components. It was reported that thin-wall W-Cu alloy components could be directly
manufactured by SLM and influences of different energy input on the thickness of single-track walls
were previously analyzed [14]. The research on Cu-10Sn bronze for SLM showed that its yield and
ultimate strengths were much higher than those attained by casting and were accompanied by a
significant improvement in ductility [11]. Moreover, multi-materials including copper alloy, such as
316L stainless steel and C18400, AlSi10Mg and C18400, were explored by SLM and the interfacial
character was fully analyzed [15,16]. Cu-Cr-Zr-Ti alloy specimens were also fabricated by SLM and their
microstructure and mechanical properties were investigated and compared to hot-rolled samples [17].
There are quite a few difficulties in manufacturing copper alloy parts in SLM. The high thermal
conductivity and reflectivity of copper alloys at the laser result in significant heat loss and inadequate
melting of powder [16]. A systematic experimental process to effectively identify appropriate SLM
processing parameters for copper alloys is minimally discussed in the literature. Therefore, in order
to explore the manufacturing process of copper alloys in SLM, in this work the influences of SLM
processing parameters on forming properties of Cu-4Sn are investigated using experimental methods.
The Cu-4Sn bronze was selected for further investigation of microstructure and mechanical properties
in SLM because it has the minimum casting shrinkage coefficient among non-ferrous metals and has
relatively slight dendrite segregation among tin-copper alloys, as the segregation depends on the level
of tin.

2. Materials and Methods

2.1. Experimental Materials

The feedstock material used for this investigation was gas-atomized Cu-4Sn powder produced
by SNDVARY (Wuxi SNDVARY New Powder Materials Technology Co., Ltd., Wuxi, China) and its
nominal chemical composition (wt %) was: Cu = balance, Sn = 3.95%, impurities = 0.2%, maximum.
The morphology of the powder was observed using scanning electron microscopy VEGA2 (TESCAN,
Brno, Czech Republic). The raw powder is almost spherical with minor satellites, as shown in
Figure 1a. Particle size distribution was measured using a laser particle size analyzer BT-9300Z
(Bettersize Instruments Ltd., Dandong, China). The result in Figure 1b shows an essential log-normal
distribution with particle sizes of 11.66 µm (D10), 31.81 µm (D50), and 63.36 µm (D90).
Materials 2017, 10, 333 3 of 12
Materials 2017, 10, 333 3 of 12

Figure 1. (a) SEM morphology and (b) particle size distribution of the Cu-4Sn powder.
Figure 1. (a) SEM morphology and (b) particle size distribution of the Cu-4Sn powder.

2.2. Experimental Equipment

2.2. Experimental Equipment
Experiments were performed on the commercial SLM machine EOSINT M280 (EOS GmbH,
Experiments were performed on the commercial SLM machine EOSINT M280 (EOS GmbH,
Krailling, Germany). The source of radiation is a single-mode continuous wave ytterbium fiber laser
Krailling, Germany). The source of radiation is a single-mode continuous wave ytterbium fiber laser
YLR-200 (IPG Photonics, Oxford, MA, USA) operating at a wavelength of 1.07 μm and producing a
YLR-200 (IPG Photonics, Oxford, MS, USA) operating at a wavelength of 1.07 µm and producing
laser beam with an energy intensity distribution with a Gaussian profile. The main characteristics of
a laser beam with an energy intensity distribution with a Gaussian profile. The main characteristics of
this machine are as follows: the maximum laser power is 195 W; the maximum laser scanning speed
this machine are as follows: the maximum laser power is 195 W; the maximum laser scanning speed is
is 7 m/s and the laser spot size on the surface of the powder bed is 100 μm. The process chamber
7 m/s and the laser spot size on the surface of the powder bed is 100 µm. The process chamber provides
provides a closed environment filled by inert gas, such as nitrogen or argon, and the substrate
a closed environment filled by inert gas, such as nitrogen or argon, and the substrate temperature can
temperature can be chosen according to different materials.
be chosen according to different materials.
2.3. Experimental Methods
2.3. Experimental Methods
First, single-track experiments with different combinations of processing parameters were
First, single-track experiments with different combinations of processing parameters were
conducted to preliminarily narrow down selection windows during the exploration stage. Some key
conducted to preliminarily narrow down selection windows during the exploration stage. Some key
factors were considered as variables and the others were fixed in the experiment. The linear energy
factors were considered as variables and the others were fixed in the experiment. The linear energy
density (LED), defined by the ratio of laser power (LP) and scanning speed (SS), is a key factor in
density (LED), defined by the ratio of laser power (LP) and scanning speed (SS), is a key factor in
SLM [18,19]. The processing parameters used for the single-track experiments are summarized in
SLM [18,19]. The processing parameters used for the single-track experiments are summarized in Table 1.
Table 1. In order to minimize the effect of oxygen content on single tracks, it was not until the oxygen
In order to minimize the effect of oxygen content on single tracks, it was not until the oxygen content
content was less than 0.1% that the argon supply to the chamber was stopped. Geometrical
was less than 0.1% that the argon supply to the chamber was stopped. Geometrical morphologies of
morphologies of single-tracks were surveyed by a Digital Microscope VHX-1000 series (Keyence,
single-tracks were surveyed by a Digital Microscope VHX-1000 series (Keyence, Osaka, Japan).
Osaka, Japan).
Table 1. SLM processing parameters used for single-track experiment.
Table 1. SLM processing parameters used for single-track experiment.
Parameters
Parameters Value
Value Increment
Increment
LED
LED 100–1000 J/mJ/m
100–1000 100 J/m
100 J/m
LP
LP 50–195 W
50–195 W 2525WW
SS 0.05–1.95 m/s -
SS
Layer thickness
0.05–1.95
20 µm
m/s -
—
Layer temperature
Preheating thickness 80 20
◦ C μm -—
Preheating
Substrate temperature
material Die Steel80 °C
(MS1) -—
Atmosphere
Substrate material Argon (Oxygen level(MS1)
Die Steel < 0.1%) -—
Atmosphere Argon (Oxygen level < 0.1%) —
Based on single-track experiments, a Greco-Latin square design experiment was performed for the
Based of
fabrication onthe
single-track experiments,
Cu-4Sn bulk a Greco-Latin
specimens with dimensionssquare design
of 5 mm experiment
× 5 mm × 5 mm.was performed
In this for
experiment,
the fabrication of the Cu-4Sn bulk specimens with dimensions of 5 mm × 5 mm × 5 mm. In this
experiment, the major processing parameters would be further redefined. The LP varied from 100 to
Materials 2017, 10, 333 4 of 12

the major processing parameters would be further redefined. The LP varied from 100 to 195 W with a
step of 25 W;
Materials the10,SS
2017, 333varied from 100 to 300 mm/s with a step of 50 mm/s; and the hatch space 4 of 12 (HS)

varied from 100 to 180 µm with a step of 20 µm. The scanning stripe width was fixed at 3 mm and other
195 W with a step of 25 W; the SS varied from 100 to 300 mm/s with a step of 50 mm/s; and the hatch
parameters were
Materials 2017, 10,kept
333 the same as in the single-track experiment. The cross-hatching scanning 4 ofstrategy
12
space (HS) varied from 100 to 180 μm with a step of 20 μm. The scanning stripe width ◦was fixed at
was employed
3 mm
whereby
and other
parallel alternative scan vectors were overlaid at an angle of 67 to the previous
195 W with a stepparameters
of 25 W; the were kept the
SS varied same100astoin300
themm/s
single-track experiment. The and
cross-hatching
deposited layer,
scanning as shown
strategy was in Figure
employed 2. Thisfrom
whereby scanning
parallel strategy
alternative
with a step
has been
scan vectors
of 50 mm/s;
proven
weretooverlaid
be an the hatch
effective
at fixed
an anglemethod
space (HS) varied from 100 to 180 μm with a step of 20 μm. The scanning stripe width was at
of improving the surface roughness and reducing defects in SLM [20]. Densities of as-built specimens
3of
mm67°andto the previous
other parametersdeposited
were layer, as same
kept the shown asininFigure 2. This scanning
the single-track strategy
experiment. Thehas been proven
cross-hatching
werescanning
measured
to be an strategyby an
effective was
electronic
method balance
of improving
employed
using
whereby the
the
surface
parallel
Archimedes
roughness
alternative scanand
method.
reducing
vectors
Relative
defects in
were overlaid
density
atSLM
(RD) is
[20].
an angle
Densities
presented as a of as-built
percentage specimens
of measured were measured
density and by an electronic
theoretical densitybalance
(8.92 using
g/cm 3the
). Archimedes
Vickers hardness
of 67° to the previous deposited layer, as shown in Figure 2. This scanning strategy has been proven
was tomethod.
measured Relative
be an effective density
on polished
method (RD)
of is presented
cross-sections
improving of
theeachas aspecimen
surface percentage
roughness ofand
using measured
a hardness
reducing density
testerand
defects theoretical
NH-5L
in SLM (Shanghai
[20].
density
EVERONE (8.92
DensitiesPrecision g/cm
of as-built
3). Vickers hardness was measured on polished cross-sections of each specimen
Instruments
specimens were Co., measured
Ltd., Shanghai, China) with
by an electronic a 300using
balance g load
the for 5 s. For each
Archimedes
using a Relative
method. hardnessdensity
tester NH-5L
(RD) is (Shanghai
presented EVERONE
as 10
a percentagePrecisionmeasured
Instruments Co., and
Ltd.,theoretical
Shanghai,
condition, all specimens were measured at least times andof results averageddensity
after eliminating false
China) with a 300 g load for 5 s. For each condition, all specimens were measured at least 10 times
data,density
so as to (8.92 g/cm
avoid the). measuring
3 Vickers hardness was measured on polished cross-sections of each specimen
deviation.
and results
using averaged
a hardness after
tester eliminating
NH-5L false EVERONE
(Shanghai data, so as to avoid the
Precision measuring deviation.
Instruments Co., Ltd., Shanghai,
China) with a 300 g load for 5 s. For each condition, all specimens were measured at least 10 times
and results averaged after eliminating false data, so as to avoid the measuring deviation.

Figure 2. Scanning strategy used in SLM experiments.

Figure 2. Scanning strategy used in SLM experiments.
According to the optimal processing parametric set in the Greco-Latin square design experiment,
Figure 2. Scanning strategy used in SLM experiments.
tensile specimens
According were fabricated
to the optimal processing andparametric
the configuration
set in thefollowed the Chinese
Greco-Latin square GB/T228.1-2010
design experiment,
standard, as shown
Accordingwere in Figure
to thefabricated 3. Tensile
optimal processing strengths of these
parametric set followed specimens
in the Greco-Latin were
square evaluated
design using an
experiment,
tensile specimens and the configuration the Chinese GB/T228.1-2010 standard,
electronic
tensile universal
specimens tester
were CMT5105
fabricated (MTS,
and Eden
the Prairie,
configurationMN, USA) and
followed fracture
the faces
Chinese were examined
GB/T228.1-2010
as shown in Figure 3. Tensile strengths of these specimens were evaluated using an electronic universal
by scanning electroninmicroscopy VEGA2strengths
(TESCAN, of Brno, Czech Republic). Simultaneously, the
testerstandard,
CMT5105
chemical
as (MTS,
shown Eden
compositions
Figure
were
3. MN,
Prairie, Tensile
USA)
characterized by and
the
thesefaces
fracture
energy
specimens
were
dispersive
were evaluated
examined
spectroscopy
using an
by scanning
(EDS) electron
technique.
electronic universal tester CMT5105 (MTS, Eden Prairie, MN, USA) and fracture faces were examined
microscopy VEGA2 (TESCAN, Brno, Czech Republic). Simultaneously, the chemical compositions
by scanning electron microscopy VEGA2 (TESCAN, Brno, Czech Republic). Simultaneously, the
were characterized by the energy dispersive spectroscopy (EDS) technique.
chemical compositions were characterized by the energy dispersive spectroscopy (EDS) technique.

Figure 3. Configuration of tensile specimens.

3. Results and Discussion

Figure 3. Configuration of tensile specimens.
Figure 3. Configuration of tensile specimens.
3.1. Single-Track Experiments
3. Results and Discussion
3. ResultsFigure 4 shows morphologies of single-tracks produced by SLM on the MS1 substrate during
and Discussion
3.1. Single-Track
single-track Experiments
experiments. The horizontal coordinate represents LED with the values (decreasing from
3.1. Single-Track
1400, 1100, Experiments
800, to 500 J/m)
Figure 4 shows morphologies and the vertical coordinate
of single-tracks represents
produced LP with
by SLM on the
the values (increasing
MS1 substrate from
during
100, 150,
single-track to 195 W).
experiments. It can be seen
The horizontal from Figure 4 that
coordinate represents solidified single-tracks present different
Figure 4 shows morphologies of single-tracks producedLED by with
SLMthe onvalues (decreasing
the MS1 substratefromduring
characteristics
1400, 1100, 800, under
to 500 different
J/m) and processing
the vertical conditions.represents
coordinate For the same
LP LP,the
with thevalues
widths(increasing
of single tracks
from
single-track
decrease experiments.
as the The horizontal coordinate represents LED with the values (decreasing are from
100, 150, to 195 LED
W). decreases
It can be (i.e.,
seenafrom
decrease as the
Figure SS increases).
4 that solidified Irregularities
single-tracks and distortions
present different
1400,observed
1100, 800,when to 500 J/m)
the LED and the
decreases vertical
to 500 J/m, coordinate
as illustrated represents
in same
the inset LP with the values (increasing
characteristics under different processing conditions. For the LP,(LED is 500 J/m,
the widths LP is 100
of single W)
tracks
from of100, 150,4.to
Figure 195phenomenon
This W). It can be can seen
be from Figure
explained by the4fact
that solidified
that the less single-tracks
laser energy present
irradiated ondifferent
the
decrease as the LED decreases (i.e., a decrease as the SS increases). Irregularities and distortions are
characteristics
powders,when
observed under
the different
narrower
the LED processing
the heat-affected
decreases conditions.
zone,
to 500 J/m, Forineven
which can
as illustrated the
the same LP, to
give (LED
inset rise the widths
J/m, LPofmelting
isincomplete
500 issingle tracks
100 W)of
powdery
decrease material.
as the
of Figure A similar
LEDphenomenon
4. This phenomenon
decreases (i.e.,
can be was
a decrease previously
explained as
by the reported
SS increases).
the fact [21]. For the
that the lessIrregularitiessame LED in Figure
and distortions
laser energy irradiated 4,
on the are
powders, the narrower the heat-affected zone, which can even give rise to incomplete melting of
powdery material. A similar phenomenon was previously reported [21]. For the same LED in Figure 4,
Materials 2017, 10, 333 5 of 12

observed when the LED decreases to 500 J/m, as illustrated in the inset (LED is 500 J/m, LP is 100 W)
of Figure 4. This phenomenon can be explained by the fact that the less laser energy irradiated on
the powders, the narrower the heat-affected zone, which can even give rise to incomplete melting of
powdery material. A similar phenomenon was previously reported [21]. For the same LED in Figure 4,
Materials 2017, 10, 333
a smoother and more uniform surface morphology can be seen with the increase of the LP, 5asof shown 12
in
the inseta (LED
smootheris 1400
and moreJ/m,uniform
LP is 195 W).morphology
surface It is emphasized that
can be seen thethe
with surface
increaseroughness
of the LP, ashas worsened
shown
in both low
in the inset (LED is 1400 J/m, LP is 195 W). It is emphasized that the surface roughness has worsened can be
LP and high LED, such as the single track (LED is 1400 J/m, LP is 100 W), which
attributed to the
in both lowunstable
LP and highmolten poolaswhere
LED, such a strong
the single Marangoni
track (LED flow
is 1400 J/m, LP was
is 100formed in can
W), which virtue
be of the
attributed
decreasing to theon
SS. Based unstable molten poolexperiments,
the single-track where a stronga Marangoni flow wasparameter
rough processing formed in virtue
window,of thehigh LP,
decreasing SS. Based on the single-track experiments, a rough processing parameter window, high
and large LED can be acquired.
LP, and large LED can be acquired.

Figure 4. The morphologies of single tracks with different processing parameters.

Figure 4. The morphologies of single tracks with different processing parameters.
3.2. The Experiment of Greco-Latin Square Design
3.2. The Experiment of Greco-Latin
Figure 5 shows Square
the surface Designof solidified Cu-4Sn single-layers and the corresponding
appearances
bulk specimens, respectively. Thanks to the Greco-Latin square design used in this study, the number
Figure 5 shows the surface appearances of solidified Cu-4Sn single-layers and the corresponding
of experiments is effectively reduced without influencing the analysis of experimental results.
bulk specimens, respectively.
The experimental orthogonalThanksarray isto the inGreco-Latin
given Table 2. The square design
fabricated layers used in this study,
and specimens are the
numberdistributed
of experiments is effectively
in the regular sequencereduced
whereby without influencing
LP decreases along thethey-axis,
analysis of experimental
SS increases along the results.
The experimental
x-axis and HSorthogonal array to
arranges according is the
given in Table
orthogonal 2. The fabricated layers and specimens are
method.
distributed Figures 5a revealssequence
in the regular that qualities of singleLP
whereby layers are significantly
decreases along affected by different
the y-axis, processing
SS increases along the
parameters. The LP, the energy intensity of the direct heat source, greatly determines the degree of
x-axis and HS arranges according to the orthogonal method.
melting of the present powder layer and re-melting depth of the previous layer. It shows a trend in
Figure 5a reveals that qualities of single layers are significantly affected by different processing
Figure 5a that incomplete melting becomes more evident with the decrease of LP. The SS,
parameters. The LP, the
characterizing the interactive
energy intensity of thethe
time between direct
laserheat
beamsource, greatlymaterial,
and powdery determines has athe degree of
slight
meltinginfluence
of the present powder
on the qualities of layer and re-melting
the single-layer. depth of the
The HS represents theoverlapping
previous layer.
level ofItthe
shows a trend in
previous
Figure 5atrack
thatand the presentmelting
incomplete track, which influences
becomes morethe joint strength
evident with the between twoofadjacent
decrease LP. Thesingle-tracks.
SS, characterizing
Some black particles (i.e., incompletely melted powders) can be seen when
the interactive time between the laser beam and powdery material, has a slight influence on the HS is wider than
thethe
qualities
Materials 2017, 10, 333 6 of 12

of the single-layer. The HS represents the overlapping level of the previous track and the present
track, which influences the joint strength between two adjacent single-tracks. Some black particles
Materials 2017, 10, 333 6 of 12
(i.e., incompletely melted powders) can be seen when the HS is wider than the width of a single-track,
as shown in the
width of acase in Figureas5ashown
single-track, whereby HS
in the = 180
case µm, LP
in Figure 5a =whereby
150 W andHS =SS180= μm,
200 mm/s.
LP = 150However,
W and final
properties of specimens produced by SLM depend strongly on the interaction
SS = 200 mm/s. However, final properties of specimens produced by SLM depend strongly on the of various parameters
interaction
rather than of various parameters
the influence of a singlerather than
factor, the influence
which can beofvalidated
a single factor, which canthrough
hereinafter be validated
RD values
of thesehereinafter
specimens through RD values
in Figure of these
5b. Here, specimens
it is in Figure
worth noting 5b. the
that Here, it is worthinnoting
difference thatcolor
surface the of the
difference in surface color of the specimens between Figure 5a,b is due to different photographing
specimens between Figure 5a,b is due to different photographing environmental conditions.
environmental conditions.

Figure 5. The morphologies of (a) single-layers and (b) specimens with various processing parameters.
Figure 5. The morphologies of (a) single-layers and (b) specimens with various processing parameters.
The measured results of density and Vickers hardness are elaborated in Table 2. From Table 2,
the maximum RD of 93.68% theoretical density is obtained in the standard order 22, with parametric
The measured results of density and Vickers hardness are elaborated in Table 2. From Table 2,
combinations of LP = 195 W, SS = 150 mm/s and HS = 100 μm. Meanwhile, the maximum Vickers
the maximum RD of 93.68% theoretical density is obtained in the standard order 22, with parametric
hardness of 118 HV 0.3/5 is acquired corresponding to the standard order 23, which is significantly
combinations of LP
higher than the=traditional
195 W, SS = 150 mm/s and
pressure-processed HS = 100 µm. Meanwhile, the maximum Vickers
Cu-4Sn.
hardness of 118 HV 0.3/5 is acquired corresponding to the standard order 23, which is significantly
higher than theTable 2. Experiment
traditional design matrix and results
pressure-processed of the Greco-Latin square design experiment.
Cu-4Sn.
Factors Responses
Standard Run
Table 2. Experiment
Order Order design
LP (W)
matrix and results
SS (mm/s)
of the Greco-Latin
HS (μm)
Density square
RD (%)
design experiment.
Vickers Hardness
(g/cm3) (HV 0.3/5)
1 6 100 100 100 7.577 86.71 79
Factors Responses
2
Standard Run5 100 150 120 8.282 86.61 89
3
Order 19
Order 100 200 140 Density
7.657 87.10 Vickers
82 Hardness
LP (W) SS (mm/s) HS (µm) RD (%)
4 3 100 250 160 (g/cm3 )
8.211 85.84 (HV
83 0.3/5)
15 61 100100 300 100 180 100 7.717
7.577 85.04
86.71 81 79
26 20
5 125100 100 150 120 120 7.728
8.282 89.57
86.61 95 89
37 13
19 125100 150 200 140 140 8.296
7.657 88.51
87.10 86 82
48 15
3 125100 200 250 160 160 7.991
8.211 88.85
85.84 87 83
59 18
1 125100 250 300 180 180 8.284
7.717 87.29
85.04 92 81
610 208 125125 300 100 100 120 8.010
7.728 89.59
89.57 92 95
711 13
24 150125 100 150 140 140 8.296
8.000 88.51
90.09 95 86
812 15
22 150125 150 200 160 160 7.991
8.135 88.85
89.92 98 87
913 18
11 150125 200 250 180 180 8.284
7.820 87.29
89.69 95 92
1014 8
23 150125 250 300 100 100 8.010
8.213 89.59
91.12 97 92
1115 24
12 150150 300 100 120 140 8.000
7.925 90.09
91.20 98 95
1216 22
25 175150 100 150 160 160 8.135
8.257 89.92
90.68 95 98
1317 11
10 175150 150 200 180 180 7.820
8.244 89.69
91.03 10195
1418 23
17 175150 200 250 100 100 8.213
7.786 91.12
92.42 10297
1519 12
16 175150 250 300 120 120 7.925
7.769 91.20
92.37 10698
1620 254 175175 300 100 140 160 8.257
7.943 90.68
92.22 10395
1721
10
14 195
175 100
150 180
180 8.244
8.347
91.03
92.41
101
100
18 17 175 200 100 7.786 92.42 102
22 21 195 150 100 7.966 93.68 104
19 16 175 250 120 7.769 92.37 106
23 2 195 200 120 8.128 93.04 118
20 4 175 300 140 7.943 92.22 103
24 9 195 250 140 8.017 92.87 104
21 14 195 100 180 8.347 92.41 100
25 7 195 300 160 8.049 93.10 103
22 21 195 150 100 7.966 93.68 104
23 2 195 200 120 8.128 93.04 118
24 9 195 250 140 8.017 92.87 104
25 7 195 300 160 8.049 93.10 103
Materials 2017, 10, 333 7 of 12

Materials 2017, 10, 333 7 of 12

Materials
The main2017, 10, 333
effects plots of RD and Vickers hardness are shown in Figures 6 and 7, respectively. 7 of 12
It is
clear from
The Figure 6 thatplots
main effects the RDof RDmainly
and depends on the LP
Vickers hardness areand increases
shown significantly
in Figures 6 and 7, as LP increases
respectively.
The main effects plots of RD and Vickers hardness are shown in Figures 6 and 7, respectively.
It is100
from clear
to from
195 W.Figure 6 that the
In contrast, theRD SSmainly
has few depends
impactsononthe LPRD
the and increases
and its inset significantly
illustrates as LP
a minor
It is clear from Figure 6 that the RD mainly depends on the LP and increases significantly as LP
increases from
fluctuation 100 to 90%
at around 195 W. withIn contrast,
the increasethe SS ofhas
SS few
fromimpacts
100 to on
300the RD and
mm/s. Theits inset
RD has illustrates a
a negative
increases from 100 to 195 W. In contrast, the SS has few impacts on the RD and its inset illustrates a
minor fluctuation
correlation with theat HS,around
which 90% withathe
shows increasegentle
relatively of SS from 100 to 300
downward mm/s. The
tendency with RD thehas a negative
increase of HS
minor fluctuation at around 90% with the increase of SS from 100 to 300 mm/s. The RD has a negative
correlation
from 100 to 180with theBy
µm. HS, which shows
comparing a relatively
Figure gentle
6 to Figure 7, downward
the effect oftendency with the
LP on Vickers increase presents
hardness of HS
correlation with the HS, which shows a relatively gentle downward tendency with the increase of HS
from
basically 100 to 180
thetosame μm. By
trends comparing Figure 6 to Figure 7, the effect of LP on Vickers hardness presents
from 100 180 μm. By as that on RD.
comparing The6differences
Figure to Figure 7,inthe Figure
effect7of
are
LPthat Vickers hardness
on Vickers hardnesspresents
plots have
basically
peak the
pointsthe same
versus trends as that on RD. The differences in Figure 7 are that Vickers hardness plots
basically sameSS and HS,
trends respectively,
as that on RD. Thewhich meansinrelatively
differences Figure 7 areoptimal processing
that Vickers parameters
hardness plots
have
of have peak
the Vickers points
hardness versus SS
(i.e.,SS and
SS and HS,
= 200HS, mm/srespectively, which
and HS =which120 µm).means relatively
The relatively
phenomenon optimal can beprocessing
explained
peak points versus respectively, means optimal processing
parameters of the Vickers hardness (i.e., SS = 200 mm/s and HS = 120 μm). The phenomenon can be
byparameters
the solidification time being
of the Vickers hardnessdecreased
(i.e., SS as themm/s
= 200 SS increases,
and HS =resulting
120 μm). in Thethe formation can
phenomenon of abefine
explained by the solidification time being decreased as the SS increases, resulting in the formation of
crystal microstructure
explained on the surface
by the solidification of specimens,
time being decreasedwhileas theaSSfurther increase
increases, of the
resulting inSStheinhibits
formationinverse
of
a fine crystal microstructure on the surface of specimens, while a further increase of the SS inhibits
segregation of the
a fine crystal tin element and
microstructure on thethesurface
generation of the intermetallic
of specimens, compounds
while a further increase fromof theperitectic
SS inhibitsand
inverse segregation of the tin element and the generation of the intermetallic compounds from
eutectic reactions. The effect of HS on Vickers hardness can be attributed
inverse segregation of the tin element and the generation of the intermetallic compounds from to thermal cycling, which
peritectic and eutectic reactions. The effect of HS on Vickers hardness can be attributed to thermal
peritectic
hascycling, and eutectic reactions.
a significant The effect of HS ontransformation
Vickers hardness can be attributed to thermal
which influence on crystal
has a significant microstructure
influence on crystal microstructure andtransformation
element redistribution
and element in the
cycling, which
overlapping zone. has a significant influence on crystal microstructure transformation and element
redistribution in the overlapping zone.
redistribution in the overlapping zone.
93
93
92
(%)

92
(%)

91
RD

91
RD

90
ofof

90
Mean

89
Mean

89
88
88
87
87
86
86 100 125 150 175 195 100 150 200 250 300 100 120 140 160 180
100 125LP150
(W)175 SS (mm/s)
195 100 150 HS 140
200 250 300 100 120 (μm)
160 180
LP (W) SS (mm/s) HS (μm)
Figure6.6.Main
Figure Maineffects
effectsplot
plot of
of processing
processing parameters
parametersfor
forthe
themean
meanvalue
valueofof
relative density.
relative density.
Figure 6. Main effects plot of processing parameters for the mean value
μ of relative density.
μ
0.3/5)
0.3/5)

105
105
(HV
(HV

100
100
hardness
hardness

95
95
Vickers

90
Vickers

90

85
ofof

85
Mean
Mean

80
80 100 125 150 175 195 100 150 200 250 300 100 120 140 160 180
100 125LP150
(W)175 195 100 SS
150 (mm/s) HS140
200 250 300 100 120 (μm)
160 180
LP (W) SS (mm/s) HS (μm)
Figure 7. Main effects plot of processing parameters for the mean value of Vickers hardness.
Figure
Figure 7.7.Main
Maineffects
effectsplot
plotof
ofprocessing
processing parameters
parameters for
forthe
themean
meanvalue
valueofofVickers
Vickershardness.
hardness.
3.3. Analysis of Variance
3.3. Analysis of Variance
3.3. Analysis of Variance
In order to more accurately analyze the statistical influences of various factors, analysis of
In order to more accurately analyze the statistical influences of various factors, analysis of
variance
In order(ANOVA) was carried
to more accurately out using
analyze a generalinfluences
the statistical linear model based on
of various theseanalysis
factors, RD andofVickers
variance
variance (ANOVA) was carried out using a general linear model based on these RD and Vickers
hardnesswas
(ANOVA) values. There
carried are three
out using impact
a general factors
linear containing
model based onLP, SS,RD
these HS,and
and two corresponding
Vickers hardness values.
hardness values. There are three impact factors containing LP, SS, HS, and two corresponding
responses
There containing
are three RD and
impact factors Vickers hardness
containing LP, SS, HS,in and
ANOVA. Factorial information
two corresponding responses and results ofRD
containing
responses containing RD and Vickers hardness in ANOVA. Factorial information and results of
andANOVA are
Vickersare exhibited
hardness in Table
in ANOVA. 3. The p-value and F-value were used to judge the significance
exhibited inlevels
ANOVA exhibited in Table 3.Factorial information
The p-value andwere
and F-value results of ANOVA
used aresignificance
to judge the Table 3.
levels
Materials 2017, 10, 333 8 of 12

The p-value and F-value were used to judge the significance levels of the factors. A p-value less than
0.05 indicates that the influence of this factor is statistically significant. Furthermore, the p-value less
than 0.01 is termed highly statistically significant [22]. The F-value is the ratio of mean square errors
from various factors and experimental errors and when it is below 1 it suggests that the effect of a factor
is smaller than that of experimental error; therefore, the factor can be neglected in this experiment [23].
In Table 3, both the F-value (170.59) and p-value (0) of the LP in RD ANOVA prove that the LP is the
most significant impact factor, followed by the F-value (10.74) and p-value (0.001) of the HS. It is noted
that the SS has no influence on RD results because the p-value (0.586) is greater than 0.25 and the
F-value (0.73) is below 1. The Vickers hardness ANOVA maintains a pronounced similarity with RD
ANOVA. In the Vickers hardness ANOVA, the LP and the HS convey the conspicuous impacts on the
Vickers hardness, and the SS also has no effect on the Vickers hardness.

Table 3. General linear model factors and results of ANOVA.

Factors Information
Factor Type Levels Values
LP (W) Fixed 5 100, 125, 150, 175, 195
SS (mm/s) Fixed 5 100, 150, 200, 250, 300
HS (µm) Fixed 5 100, 120, 140, 160, 180
RD ANOVA
Source DF Adj SS Adj MS F-Value p-Value
LP (W) 4 139.236 34.8090 170.59 0.000
SS (mm/s) 4 0.599 0.1499 0.73 0.586
HS (µm) 4 8.767 2.1918 10.74 0.001
Error 12 2.449 0.2041 - -
Total 24 151.051 - - -
Vickers Hardness ANOVA
Source DF Adj SS Adj MS F-Value p-Value
LP (W) 4 1646.8 411.7 41.31 0.000
SS (mm/s) 4 48.8 12.2 1.22 0.351
HS (µm) 4 216.8 54.2 5.44 0.010
Error 12 119.6 9.967 - -
Total 24 2032 - - -

Residual plots for RD and Vickers hardness are presented in Figures 8 and 9, which contain the
normal probability plot of residuals, residuals versus fitted values, the histogram of the residuals, and
residuals versus the order of data, respectively. The points in Figure 8a form a straight line, meaning
that the residuals of measured RD values with predicted values are normally distributed. Figure 8c
intuitively shows a normal distribution pattern. A random distribution of residuals in Figure 8b
suggests that this experimental model tends to yield the random error instead of an outlier. In addition,
Figure 8d is a plot of all residuals in the order that the experimental results were collected and can be
used to identify the non-random error and, here, it displays a beneficial experimental process control
based on the SPC control chart standard. However, besides the similarity of normal distribution in
Figure 9, the differences in these plots are concerned. The residual range (−5 to 5) becomes wider than
that (−0.6 to 0.6) in Figure 8 and individual outliers may be produced, as shown in Figure 9b,c, which
can be ascribed to the resultant porosity in these specimens, leading to the non-random error.
Materials 2017, 10, 333 9 of 12
Materials 2017, 10, 333 9 of 12
Materials 2017, 10, 333 9 of 12
(a) (b)
99 (a) 0.6 (b)
99
90 0.6
0.3

Residual
Percent
90 0.3

Residual
0.0
Percent

50
50 0.0
-0.3
10 -0.3
10 -0.6
1 -0.6
-0.8 -0.4 0.0 0.4 0.8 86 88 90 92 94
1
-0.8 -0.4 Residual
0.0 0.4 0.8 86 88
Fitted Value 92
90 94
Residual Fitted Value
(c) (d)
8 (c) 0.6 (d)
Frequency

Residual
68 0.6
0.3
Frequency

Residual
46 0.0
0.3
0.0
-0.3
24
-0.3
-0.6
02
-0.6 2 4 6 8 0 2 4 6 8 0 2 4
0 -0.6 -0.4 -0.2 0.0 0.2 0. 4 0 .6 1 1 1 1 1 2 2 2
.6 .4 .2 0.0 0.2 0. 4 0 .6 2 4 6 8 10 12 1 4 16 18 20 2 2 24
Observation Order
-0 -0 -0 Residual
Residual Observation Order
Figure 8. The residual plots for RD: (a) normal probability plot of residuals; (b) residuals versus fitted
Figure 8. The residual plots for RD: (a) normal probability plot of residuals; (b) residuals versus fitted
Figure (c)
values; 8. The residualofplots
histogram for RD: (a)and
the residuals; normal probability
(d) residuals plotorder
versus of residuals;
of data.(b) residuals versus fitted
values; (c) histogram of the residuals; and (d) residuals versus order of data.
values; (c) histogram of the residuals; and (d) residuals versus order of data.
(a) (b)
99 (a) 5.0 (b)
99
90 5.0
Residual

2.5
Percent

90
Residual

2.5
Percent

50 0.0
50 0.0
10 -2.5
10 -2.5
-5.0
1
-5.0 -2.5 0.0 2.5 5.0 -5.0 80 90 100 110
1
-5.0 -2.5 Residual
0.0 2.5 5.0 80 90
Fitted 100
Value 110
Residual Fitted Value
(c) (d)
6.0 (c) 5.0 (d)
Frequency

5.0
Residual

6.0
4.5 2.5
Frequency

Residual

4.5
3.0 2.5
0.0
3.0
1.5 0.0
-2.5
1.5
0.0 -2.5
-5.0
-4 -2 0 2 4 -5.0 2 4 6 8 10 1 2 14 16 18 20 22 24
0.0
-4 -2 Residual
0 2 4 2 4 Observation 6 18 20 22 24
6 8 10 1 2 14 1Order
Residual Observation Order
Figure 9. The residual plots for Vickers hardness: (a) normal probability plot of residuals; (b) residuals
Figurefitted
versus 9. Thevalues;
residual
(c)plots for Vickers
histogram of thehardness:
residuals;(a)and
normal probability
(d) residuals plotorder
versus of residuals;
of data.(b) residuals
Figure 9. The residual plots for Vickers hardness: (a) normal probability plot of residuals; (b) residuals
versus fitted values; (c) histogram of the residuals; and (d) residuals versus order of data.
versus fitted values; (c) histogram of the residuals; and (d) residuals versus order of data.
3.4. Forming Properties
3.4. Forming Properties
Figure 10
3.4. Forming shows the surface morphology of a Cu-4Sn specimen with 93.68% RD value in the
Properties
Figure 10
Greco-Latin square shows the surface
design morphology
experiment. of a Cu-4Sn
It is obvious that highspecimen with 93.68%
surface roughness andRD value in the
discontinuous
Figure
Greco-Latin 10 shows
square the surface
design experiment.morphology of
It isAobvious a Cu-4Sn specimen with 93.68% RD value in the
single tracks can be observed in Figure 10a. numberthat high surface
of defects, such asroughness
pores andand pitsdiscontinuous
(overlapping
Greco-Latin
single tracks square design
can be observed experiment.
in Figure It is obvious that high surface roughness and discontinuous
defects) are produced, contributing to 10a. A numberlow
the relatively of defects,
RD. The such as pores
reasons for and
thesepits (overlapping
defects can be
single tracks can be observed in Figure 10a. A numberlow of defects, such as pores and pitsdefects
(overlapping
ascribed to the inadequate laser energy, therefore, resulting in insufficient liquid phasecan
defects) are produced, contributing to the relatively RD. The reasons for these andbe
defects) are produced,
ascribed to thecausing contributing
inadequate to
laser energy, the relatively
therefore, low RD.
resulting The reasons
in insufficient for these defects
liquid tension
phase andcan
and
simultaneously a significant shrinkage of the molten pool due to the high surface
be ascribed to the inadequate lasershrinkage
energy, therefore, resulting in to
insufficient liquid tension
phase and
large viscosity of the liquid phase along with the resultant Marangoni flow. Additionally, itand
simultaneously causing a significant of the molten pool due the high surface is
simultaneously
large viscosity causing a significant shrinkage of
thethe molten pool due to flow.
the high surface tension
responsible for aof the liquid
non-uniform phase along
deposition with
in the previous resultant
layer onMarangoni Additionally,
account of characteristically it is
rapidly
and large viscosity
responsible of the liquid phase along
for aanon-uniform with the resultant Marangoni flow. Additionally, it is
solidifying with limitation ofdeposition
spreading in thethe previous
molten poollayer onoverlapping
in the account of characteristically
track-track position rapidly
(as
solidifying with a limitation of spreading the molten pool in the overlapping
shown in Figure 10b). As a consequence, according to the origins of these defects and above ANOVA, track-track position (as
shown in Figure 10b). As a consequence, according to the origins of these defects and above ANOVA,
Materials 2017, 10, 333 10 of 12

responsible for a non-uniform deposition in the previous layer on account of characteristically rapidly
solidifying with a limitation of spreading the molten pool in the overlapping track-track position
Materials 2017, 10, 333 10 of 12
(as shown in Figure 10b). As a consequence, according to the origins of these defects and above ANOVA,
the densities of Cu-4Sn parts manufactured by SLM can be improved further through tailoring a higher
the densities of Cu-4Sn parts manufactured by SLM can be improved further through tailoring a
energy intensity with a high LP while decreasing SS or narrowing HS are unfavorable methods
higher energy intensity with a high LP while decreasing SS or narrowing HS are unfavorable methods
since they tend to form the unstable molten pool (as shown in Figure 4) and seriously influence the
since they tend to form the unstable molten pool (as shown in Figure 4) and seriously influence the
production efficiency in the industrial application. The scaly image on the single tracks can contribute
production efficiency in the industrial application. The scaly image on the single tracks can contribute
to identifying the scanning direction, as shown by the arrow in Figure 10b.
to identifying the scanning direction, as shown by the arrow in Figure 10b.

Figure 10. SEM surface morphologies of the Cu-4Sn with the RD 93.68%: (a) At a low magnification;
Figure 10. SEM surface morphologies of the Cu-4Sn with the RD 93.68%: (a) At a low magnification;
and (b) At a high magnification.
and (b) At a high magnification.

The tensile specimens of Cu-4Sn were fabricated using relatively optimal parameters set in the
above experiments. An An attained
attained ultimate
ultimate tensile
tensile strength
strength (UTS)
(UTS) 316–320
316–320 MPa
MPa is is inferior
inferior to that of
traditional pressure-processed Cu-4Sn (not less than 410 MPa). Figure 11 shows the fracture surfaces
of these tensile samples
samples inin order
order toto further
further understand
understand the the fracture
fracture mechanism.
mechanism. These samples
simultaneously display two features
features in the Figure
Figure 11: step surfaces (Part 1) and dimples (Part 2), thus
implying two fracture mechanisms (i.e., brittle cleavage fracture and ductile fracture). The production
of two different fracture mechanisms can be ascribed to the segregation of tin in the sample sample (Figure
(Figure 11)
11)
where point 1 of of the
the step
step surfaces
surfacesholds
holdsthethesame
sametin tinlevel
levelasasraw
rawmaterials,
materials,while
while copper
copper content
content is
is extremely
extremely high
high on on point
point 2 of2 the
of the dimples,
dimples, as presented
as presented in Table
in Table 4. Some
4. Some porespores
and and grooves
grooves are
are also
also observed
observed in these
in these specimens,
specimens, which
which cancan explain
explain thethelowlow UTS.Generally,
UTS. Generally,these
thesedefects
defects are
are crack
initiators and influence the overall structural performance during tensile testing. Therefore, defects
should be reduced for the further improvement
improvement of of UTS
UTS to to meet
meet thethe needs
needs of
of applications.
applications.

4. The chemical compositions of different point on the

Table 4.
Table the fracture
fracture surface.
surface.

Weight
Weight Percentage
Percentage (%)
(wt %)
Elements
Elements Point 1 Point 2
Point 1 Point 2
Cu 95.33 100
Cu 95.33 100
Sn Sn 4.674.67 00
Total
Total 100100 100
100
Materials 2017, 10, 333 11 of 12
Materials 2017, 10, 333 11 of 12

Figure 11. SEM micrograph of fracture surfaces during tensile testing.

Figure 11. SEM micrograph of fracture surfaces during tensile testing.

4. Conclusions
4. Conclusions
Relatively optimal SLM processing parameters of Cu-4Sn based on the EOSINT M280 equipment
Relatively optimal SLM processing parameters of Cu-4Sn based on the EOSINT M280 equipment
were successfully identified and bulk specimens successfully manufactured for the first time using a
were successfully identified and bulk specimens successfully manufactured for the first time using
systematic experimental approach. This proposed experimental approach consists of a single-track
a systematic experimental approach. This proposed experimental approach consists of a single-track
experiment and Greco-Latin square design experiment. The obtained conclusions are as follows:
experiment and Greco-Latin square design experiment. The obtained conclusions are as follows:
1. The best RD value (93.68%) was obtained using an optimal SLM processing parametric set of
1. LP The= best RDSS
195 W, value
= 50(93.68%)
mm/s and wasHSobtained
= 60 μm.using an optimal
The best VickersSLM processing
hardness (118 HV parametric
0.3/5) wasset of
also
LP = 195 W, SS = 50 mm/s and HS = 60 µm. The best Vickers hardness (118 HV 0.3/5)
acquired using an SLM processing parametric set (i.e., LP = 195 W, SS = 200 mm/s, and HS = 120 μm). was also
acquiredthe
2. During using an SLM processing
experimental parametric
process, set (i.e.,
single-track LP = 195 W, SS
experiments = 200
with mm/s, and
different HS = 120 µm).
combinations of
2. processing
During theparameters
experimental process, single-track experiments with different combinations
can effectively narrow down selection windows for various parameters. of
processing parameters
Greco-Latin square design canexperiments
effectively narrow down selection
with orthogonal windows
parameter arraysfor various
can parameters.
effectively reduce
Greco-Latin square design experiments with orthogonal
the number of experiments to acquire optimal processing parameters. parameter arrays can effectively reduce
3. Thethe number
ANOVAofpermits
experimentsone totoidentify
acquirestatistical
optimal processing
influences parameters.
of processing parameters on RD and
3. Vickers
The ANOVA permits one to identify statistical influences
hardness. It is noted that RD and Vickers hardness ofof
processing parameters
Cu-4Sn specimens on RD and
produced by
Vickers
SLM hardness.
depend stronglyIt ison
noted that RD and Vickers hardness of Cu-4Sn specimens produced by
the LP.
4. TheSLMhighest
dependtensile
strongly on theachieved
strength LP. for Cu-4Sn in SLM is only 316–320 MPa, which is much
4. less
Thethan
highest
thattensile strengthpressure-processed
of traditional achieved for Cu-4Sn in SLM
Cu-4Sn. is main
The only 316–320
reason for MPa,
this which
is due is
to much
some
less thandefects
internal that ofoftraditional
specimens, pressure-processed Cu-4Sn.
such as grooves, pores, andThe mainThese
balling. reasondefects
for thiscan
is due to some
be inhibited
internal defects
effectively of specimens,
by further improving such
LP.as grooves, pores, and balling. These defects can be inhibited
effectively by further improving LP.
Acknowledgments: The authors acknowledge the National High-tech Research and Development Program of
China (863 Program:The
Acknowledgments: 2015AA042501) for financial
authors acknowledge thesupport of High-tech
National this work. Research and Development Program of
China (863 Program: 2015AA042501) for financial support of this work.
Author Contributions: Zhongfa Mao, David Z. Zhang conceived and designed the experiments; Zhongfa Mao
Author Contributions:
and Kaifei Zhongfa
Zhang performed the Mao, David Z.Zhongfa
experiments; Zhang conceived and designed
Mao and Peitang the experiments;
Wei analyzed the data; Zhongfa
Zhongfa Mao
Mao
and Kaifei Zhang performed the experiments; Zhongfa Mao and Peitang Wei analyzed the data; Zhongfa Mao
wrote the paper.
wrote the paper.
Conflicts of Interest:
Conflicts of Interest: The
The authors
authors declare
declare no
no conflict
conflict of
of interest.
interest.

References
References
1.
1. Kruth,
Kruth, J.P.;
J.P.; Levy,
Levy, G.; Klocke, F.;
G.; Klocke, F.; Childs,
Childs, T.H.C.
T.H.C. Consolidation
Consolidation phenomena
phenomena in laser and
in laser and powder-bed
powder-bed based
based
layered manufacturing. CIRP Ann. Manuf. Technol. 2007, 56, 730–759.
layered manufacturing. CIRP Ann. Manuf. Technol. 2007, 56, 730–759. [CrossRef]
2. Yap, C.Y.; Chua, C.K.; Dong, Z.L.; Liu, Z.H.; Zhang, D.Q.; Loh, L.E.; Sing, S.L. Review of selective laser
melting: Materials and applications. Appl.
Appl. Phys.
Phys. Rev.
Rev. 2015,
2015,2,
2,041101.
041101. [CrossRef]
3. Wang, Z.M.; Guan, K.; Gao, M.; Li, X.Y.; Chen, X.F.; Zeng, X.Y. The microstructure and mechanical
properties of deposited-in718 by selective laser melting. J. Alloys Compd. 2012, 513, 518–523.
Materials 2017, 10, 333 12 of 12

3. Wang, Z.M.; Guan, K.; Gao, M.; Li, X.Y.; Chen, X.F.; Zeng, X.Y. The microstructure and mechanical properties
of deposited-in718 by selective laser melting. J. Alloys Compd. 2012, 513, 518–523. [CrossRef]
4. Edwards, P.; Ramulu, M. Fatigue performance evaluation of selective laser melted Ti–6Al–4V. Mater. Sci.
Eng. A 2014, 598, 327–337. [CrossRef]
5. Thijs, L.; Verhaeghe, F.; Craeghs, T.; Van Humbeeck, J.; Kruth, J.P. A study of the micro structural evolution
during selective laser melting of ti-6al-4v. Acta Mater. 2010, 58, 3303–3312. [CrossRef]
6. Mumtaz, K.A.; Hopkinson, N. Laser melting functionally graded composition of waspaloy® and zirconia
powders. J. Mater. Sci. 2007, 42, 7647–7656. [CrossRef]
7. Wang, D.; Yang, Y.; Liu, R.; Xiao, D.; Sun, J. Study on the designing rules and processability of porous
structure based on selective laser melting (SLM). J. Mater. Process. Technol. 2013, 213, 1734–1742. [CrossRef]
8. Wang, D.; Wang, Y.M.; Wang, J.H.; Song, C.H.; Yang, Y.Q.; Zhang, Z.M.; Lin, H.; Zhen, Y.Q.; Liao, S.X. Design
and fabrication of a precision template for spine surgery using selective laser melting (SLM). Materials 2016,
9, 608. [CrossRef]
9. Caron, R.N. Copper: Alloying. In Reference Module in Materials Science and Materials Engineering; Elsevier:
Amsterdam, The Netherlands, 2016.
10. Barik, R.C.; Wharton, J.A.; Wood, R.J.K.; Tan, K.S.; Stokes, K.R. Erosion and erosion–corrosion performance
of cast and thermally sprayed nickel–aluminium bronze. Wear 2005, 259, 230–242. [CrossRef]
11. Scudino, S.; Unterdörfer, C.; Prashanth, K.G.; Attar, H.; Ellendt, N.; Uhlenwinkel, V.; Eckert, J. Additive
manufacturing of cu–10sn bronze. Mater. Lett. 2015, 156, 202–204. [CrossRef]
12. Gu, D.; Shen, Y. Balling phenomena during direct laser sintering of multi-component cu-based metal powder.
J. Alloys Compd. 2007, 432, 163–166. [CrossRef]
13. Gu, D.; Shen, Y.; Lu, Z. Microstructural characteristics and formation mechanism of direct laser-sintered
cu-based alloys reinforced with ni particles. Mater. Des. 2009, 30, 2099–2107. [CrossRef]
14. Song, C.H.; Yang, Y.Q.; Liu, Y.; Luo, Z.Y.; Yu, J.K. Study on manufacturing of W-Cu alloy thin wall parts by
selective laser melting. Int. J. Adv. Manuf. Technol. 2015, 78, 885–893. [CrossRef]
15. Sing, S.L.; Lam, L.P.; Zhang, D.Q.; Liu, Z.H.; Chua, C.K. Interfacial characterization of slm parts in
multi-material processing: Intermetallic phase formation between alsi10mg and c18400 copper alloy.
Mater. Charact. 2015, 107, 220–227. [CrossRef]
16. Liu, Z.H.; Zhang, D.Q.; Sing, S.L.; Chua, C.K.; Loh, L.E. Interfacial characterization of slm parts in
multi-material processing: Metallurgical diffusion between 316l stainless steel and c18400 copper alloy.
Mater. Charact. 2014, 94, 116–125. [CrossRef]
17. Popovich, A.; Sufiiarov, V.; Polozov, I.; Borisov, E.; Masaylo, D.; Orlov, A. Microstructure and mechanical
properties of additive manufactured copper alloy. Mater. Lett. 2016, 179, 38–41. [CrossRef]
18. Wang, X.J.; Zhang, L.C.; Fang, M.H.; Sercombe, T.B. The effect of atmosphere on the structure and properties
of a selective laser melted Al-12Si alloy. Mater. Sci. Eng. Struct. Mater. Prop. Microstruct. Process. 2014, 597,
370–375. [CrossRef]
19. Gu, D.; Shen, Y. Effects of processing parameters on consolidation and microstructure of W–Cu components
by dmls. J. Alloys Compd. 2009, 473, 107–115. [CrossRef]
20. Yadroitsev, I.; Smurov, I. Selective laser melting technology: From the single laser melted track stability to 3d
parts of complex shape. Phys. Procedia 2010, 5, 551–560. [CrossRef]
21. Yadroitsev, I.; Gusarov, A.; Yadroitsava, I.; Smurov, I. Single track formation in selective laser melting of
metal powders. J. Mater. Process. Technol. 2010, 210, 1624–1631. [CrossRef]
22. Oehlert, G.W. A First Course in Design and Analysis of Experiments; Kluwer Academic Publishers: New York,
NY, USA, 2000; pp. 45–50.
23. Yadroitsev, I.; Yadroitsava, I.; Bertrand, P.; Smurov, I. Factor analysis of selective laser melting process
parameters and geometrical characteristics of synthesized single tracks. Rapid Prototyp. J. 2012, 18, 201–208.
[CrossRef]

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