materials

Article
Wire Arc Additive Manufacturing of AZ31
Magnesium Alloy: Grain Refinement by
Adjusting Pulse Frequency
Jing Guo 1 , Yong Zhou 1 , Changmeng Liu 1, *, Qianru Wu 1 , Xianping Chen 2 and Jiping Lu 1
1 School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China;
guojingcn@hotmail.com (J.G.); bitzhouy@bit.edu.cn (Y.Z.); qrwu@foxmail.com (Q.W.);
jipinglu@bit.edu.cn (J.L.)
2 Beijing Institute of Astronautical Systems Engineering, Beijing 100076, China; chenxianping2000@gmail.com
* Correspondence: liuchangmeng@bit.edu.cn; Tel.: +86-10-6891-5097

Academic Editor: Guillermo Requena

Received: 26 August 2016; Accepted: 5 October 2016; Published: 9 October 2016

Abstract: Wire arc additive manufacturing (WAAM) offers a potential approach to fabricate
large-scale magnesium alloy components with low cost and high efficiency, although this topic
is yet to be reported in literature. In this study, WAAM is preliminarily applied to fabricate AZ31
magnesium. Fully dense AZ31 magnesium alloy components are successfully obtained. Meanwhile,
to refine grains and obtain good mechanical properties, the effects of pulse frequency (1, 2, 5, 10,
100, and 500 Hz) on the macrostructure, microstructure and tensile properties are investigated.
The results indicate that pulse frequency can result in the change of weld pool oscillations and
cooling rate. This further leads to the change of the grain size, grain shape, as well as the tensile
properties. Meanwhile, due to the resonance of the weld pool at 5 Hz and 10 Hz, the samples have
poor geometry accuracy but contain finer equiaxed grains (21 µm) and exhibit higher ultimate tensile
strength (260 MPa) and yield strength (102 MPa), which are similar to those of the forged AZ31 alloy.
Moreover, the elongation of all samples is above 23%.

Keywords: wire arc additive manufacturing; AZ31 magnesium alloy; grain; mechanical properties

1. Introduction
As the lightest structural alloys, magnesium alloys have attracted great attention in recent
years. They possess many excellent properties, such as high specific strength, good mechanic
damping properties, electromagnetic shielding properties, castability, machining property, weldability,
recyclability, biodegradability etc. [1–4]. Therefore, they have been widely used in aerospace,
aircraft, automotive, electronics, and other industries [5,6]. For example, in the automotive industry,
magnesium alloys play an important role in reducing weight and realizing energy saving and emission
reduction by replacing steel and aluminum parts [7–9]. AZ31 magnesium alloy is one of the Mg–Al–Zn
(AZ) alloys, and typically contains 2.5~3.5 wt % Al, 0.6~1.4 wt % Zn and 0.2~1.0 wt % Mn. At present,
AZ31 magnesium alloy is the most widely used among all wrought magnesium alloys.
Currently, to realize further lightweight, developing monolithic components is an important
trend for the application of magnesium alloy. However, the monolithic components generally exhibit
large scale and complex shape. It is difficult to fabricate them using conventional forging and casting
techniques [10–12]. Thus, new manufacturing methods are highly desired to produce monolithic
magnesium alloy components.
Additive manufacturing (AM) is a technique that fabricates components layer upon layer by
adding metallic materials vice conventional subtractive manufacturing [13–15]. It is a promising

Materials 2016, 9, 823; doi:10.3390/ma9100823 www.mdpi.com/journal/materials

Materials 2016, 9, 823 2 of 13

method for manufacturing complex and large scale monolithic components. During AM, a heat source
(arc, electronic beam or laser) moves along a specific path and melts metal powder or wire to form the
components [13]. In recent years, metal AM technology has been developed rapidly due to its ability
of near-net shape fabrication of monolithic components and the possibility of reducing lead-time,
energy consumption and cost [14,16–19]. Up to now, only a few researches on manufacturing pure
magnesium [20,21], magnesium alloys [10,11] or the mixture of magnesium [22] using selective laser
melting (SLM) have been reported. However, during SLM, most of laser energy is reflected owing to
the low absorption of magnesium [23], which results in poor efficiency and increases the risk of the
breakdown of the laser device. Besides, the fine powders of magnesium alloys for SLM are flammable
and explosive.
Wire arc additive manufacturing (WAAM) is another choice of AM techniques. To fabricate parts
or components, metal wire is fed at a constant speed and melted by arc onto a substrate or the previous
deposition layers [24]. Compared with the power-based AM technique, WAAM has a much higher
deposition rate, higher material usage efficiency and lower cost, which reveals the potential to fabricate
large scale components [24–28]. Hence, WAAM has attracted great attention and has been reported
in many papers to fabricate various materials, such as steel [29], nickel aluminum bronze alloy [30],
titanium alloys [17,18,31], aluminum alloys [32,33] etc. However, to the best of our knowledge,
studies focused on the WAAM of magnesium or its alloys have not been reported. Investigation of the
WAAM of AZ31 magnesium alloy may provide an alternative method, allowing a broader application
of magnesium and its alloys.
Hence, in this study, the WAAM of AZ31 magnesium alloy is investigated. Meanwhile, one of the
main challenges for the widespread use of AM is to obtain fine grains and good mechanical properties.
This study focuses on achieving this goal by adjusting pulse frequency. Thus, the influence of pulse
frequency on macrostructure, microstructure and tensile properties of wire arc additive manufactured
AZ31 magnesium alloy is disclosed in this paper.

2. Experimental Procedures

2.1. Experimental Setup and Manufacturing Process

Figure 1 schematically depicted the experimental setup developed for wire arc additive
manufacturing. It was composed of gas tungsten arc welding (GTAW) equipment, a wire feeder,
a computer, argon gas, a three-dimensional workbench and working chamber.
AZ31 wire with a diameter of 1.2 mm was used in this study. The rolled AZ31 plate with the size
of 150 × 150 × 5 mm3 (length × width × height) was used as a substrate, which was mechanically
cleaned and then fixed on the workbench before the deposition process.
To investigate the influence of pulse frequency, six wire arc additive manufactured walls of about
100 mm in length and 10~20 mm in height were manufactured by GTAW equipment (Make: Miller,
Appleton, WI, USA; Model: Dynasty 350) with a WF-007A cool wire feeder (Make: WEILD, Guangzhou,
China) in an argon atmosphere. Each wall was made up of eight layers. No preheating was conducted
before the deposition process and the first layer was manufactured without a pulse current in order
to obtain good shape. Other layers were manufactured with different pulse frequencies. There was
about a one-minute period to cool down between depositing each layer. According to careful analysis
of previous work about the GTAW of AZ31 magnesium alloy [7,34–36], and multiple trail experiments,
optimized fabrication parameters were determined. Six different pulse frequencies were selected,
and other fabrication parameters were kept constant. The main deposition parameters used in this
study are shown in Table 1.
Materials 2016, 9, 823 3 of 13
Materials 2016, 9, 823 3 of 12

Figure
Figure1.1.Schematic
Schematicof
ofthe
theexperimental
experimentalsetup
setupdeveloped
developedfor
forwire
wirearc
arcadditive
additive manufacturing.
manufacturing.

Table
Table1.
1.Deposition
Depositionparameters
parametersused
usedin
inthis
thisstudy.
study.

Deposition Parameters Values

Deposition Parameters Values
Type of welding current alternating current (AC)
Type of welding current alternating current (AC)
Alternating current frequency 400 Hz
Alternating current frequency 400 Hz
Electrode positive
Electrode (EP)(EP)
positive 136A A
136
Electrode negative
Electrode (EN)(EN)
negative 9191AA
Electrode
Electrode negative
negative ratioratio 80%
80%
Pulse Pulse frequency
frequency 1,
1, 2,2,5,5,10,
10,100
100and and500500
HzHz
Peak time ratio 50%
Peak time ratio
Base-to-peak current ratio 50%
30%
Base-to-peak current
Wire ratio
feed rate 30%
2 m/min
WireDeposition
feed rate speed 200 2mm/min
m/min
DepositionArgon speed 99.999% purity
200 mm/min
Shield gas flow rate 20 L/min
Argon 99.999% purity
Arc length 3 mm
Shield gas flow rate
Tungsten electrode diameter 20 mm
2.4 L/min
Arc length 3 mm
Tungsten electrode diameter 2.4 mm
The waveform used in this study was a combination of pulse and AC waveform as shown in
Figure 2. The average current can be calculated by the following relation:
The waveform used in this study was a combination of pulse and AC waveform as shown in
Figure 2. The average current ( IEP × tcan be calculated by the following relation:
EP + IEN × t EN ) × t p + ( IEP × t EP + IEN × t EN ) × k bp × tb
Iaverage = (1)
× + × × t+ p + tb × + × × ×
= (1)
+
where Iaverage is the average current, IEP is the electrode positive, IEN is the electrode negative, tEP is the
where Iaverage
electrode is the average
positive duration,current,
tEN is the IEP electrode
is the electrode positive,
negative duration, IEN tis the electrode negative, tEP is the
p is the peak current duration, tb is
electrode positive duration, t EN is the electrode negative duration,
the base current duration and kbp is the base-to-peak current ratio. Furthermore, tp is the peak current duration,
the average tb is
current in
the
thisbase
studycurrent
was 65duration
A. and kbp is the base-to-peak current ratio. Furthermore, the average current
in this study was 65 A.
Materials 2016, 9, 823 4 of 13
Materials 2016, 9, 823 4 of 12

Figure 2. Schematic of the waveform used in this study.

Figure 2. Schematic of the waveform used in this study.
2.2. Research Methodology
2.2. Research Methodology
The influence of pulse frequency is analyzed from three aspects: macrostructure, microstructure
The influence
and tensile of pulse
properties. The frequency
height of theis analyzed
wire arcfrom threemanufactured
additive aspects: macrostructure,
walls was microstructure
measured and
and tensile
was dividedproperties.
by the total The height of
number of the wire
layers toarc additive
obtain manufactured
the average walls wasIn
layer thickness. measured and was
consideration of
divided
the low by the totalaccuracy
geometry number of of layers
walls, to obtain thewidth
maximum average layer thickness.
is chosen as another In reference
consideration of the low
for geometric
geometryin
accuracy accuracy
this study.of walls, maximum width is chosen as another reference for geometric accuracy in
this study.
Figure 3 shows the sample preparation procedures. Cross sections were cut for microstructure
Figure 3as
observations shows
shown theinsample
Figure preparation procedures.specimens
3b. The metallographic Cross sections were cut for
were mounted, microstructure
polished with SiC
observations as shown in Figure 3b. The metallographic specimens were
papers (180, 400, 600, 800, 1000, 1500, 2000 grit), and then electrolytic polished in ethanol mounted, polished
90%withand
SiC papersacid
perchloric (180,
10% 400, 600, 800,
for about one 1000,
minute1500,
at −302000 grit),
°C and 15 V.and thenthat,
After electrolytic
they werepolished
etched inina mixture
ethanol
90% −30 ◦
of 5 gand perchloric
picric acid, 5 mL acid 10%acid,
acetic for 10
about one minute
mL diluted wateratand 100 CmLand 15 V. The
ethanol. After that, they were
microstructures of
etched inwere
samples a mixture of 5 g picric
characterized acid, 5microscopy
by optical mL acetic acid,
(Make: 10 mL
Leica,diluted water
Wetzlar, and 100Model:
Germany; mL ethanol.
Leica
The microstructures
DM4000M). The first of samples
(bottom) and were
last characterized by optical
(top) several layers microscopy
have different (Make: Leica,
microstructures Wetzlar,
due to the
Germany; Model: Leica DM4000M). The first (bottom) and last (top) several
incomplete thermal cycles [37]. To avoid this influence, the middle of the walls was selected to reveal layers have different
microstructures
the microstructures dueandto the incomplete
analyze thermal
the tensile cycles [37]. To avoid this influence, the middle of the
properties.
wallsInwas selected to reveal the microstructures
order to quantitatively investigate the effect and analyze
of pulsethe tensile properties.
frequency on microstructures, the line
In order to quantitatively investigate the effect of pulse
intercept method is adopted to measure the grain size and the grain aspect frequency on microstructures, the line
ratio. The details are
intercept method
described in Figure is 4.
adopted
Firstly,to measure
three the grain
test lines were size
evenlyandplaced
the grain
in theaspect ratio. The details
microstructures along theare
described in Figure 4. Firstly, three test lines were evenly placed in the
horizontal and vertical directions. Secondly, the interceptions were counted. An intersection pointmicrostructures along the
horizontal and
represents that vertical
one testdirections.
line cuts Secondly, the interceptions
a grain boundary. were counted.
The counting numberAn is intersection
set to 1 for point
each
represents that
intersection, one each
1 for test line cuts a grain
tangential boundary.0.5
intersection, The counting
when one endnumber is set
of the testtoline
1 forends
eachexactly
intersection,
on a
1 for each tangential intersection, 0.5 when one end of the test line ends exactly
grain boundary, and 1.5 when the intersection occurs at a triple point. Thirdly, the grain size is on a grain boundary,
and 1.5 when
obtained by, the intersection occurs at a triple point. Thirdly, the grain size is obtained by,

l
d == (2)
(2)
N
where d is grain size, l is the length of the test line and N is the number of the intersections. Finally,
where d is grain size, l is the length of the test line and N is the number of the intersections. Finally,
at least three different microstructures of every sample were measured. The grain size was calculated
at least three different microstructures of every sample were measured. The grain size was calculated by
by averaging the measured values. In this way, grain size along the horizontal and vertical direction
averaging the measured values. In this way, grain size along the horizontal and vertical direction was
was obtained respectively. Different grain sizes along different directions were observed in this
obtained respectively. Different grain sizes along different directions were observed in this experiment.
experiment. Therefore, the grain aspect ratio, defined as the grain size along the vertical direction
Therefore, the grain aspect ratio, defined as the grain size along the vertical direction divided by the
divided by the grain size along the horizontal direction, is analyzed, which to some extent reflects
grain size along the horizontal direction, is analyzed, which to some extent reflects the anisotropic
the anisotropic mechanical properties.
mechanical properties.
Materials 2016, 9, 823 5 of 13
Materials
Materials 2016,
2016, 9,
9, 823
823 5
5 of
of 12
12

Figure
Figure 3. Sample preparation:
3. Sample
Sample preparation: (a)
preparation: (a) manufacturing
(a) manufacturing procedure
manufacturing procedure of of tensile
tensile specimens;
specimens; (b)
(b) cross
cross sections
sections for
for
microstructure observations
observations and
and (c)
(c) dimensions
dimensions of
of tensile
tensile specimens.
specimens.
microstructure observations and (c) dimensions of tensile specimens.

Figure
Figure 4.
4. Sketch
Figure 4. Sketch of
Sketch of the
of the line
the line intercept
line intercept method.
method.
intercept method.

Figure
Figure 3a 3a presents
presents the
the manufacturing
manufacturing procedure
procedure ofof the
the tensile
tensile specimens.
specimens. These
These specimens
specimens were
were
Figure 3a presents the manufacturing procedure of the tensile specimens. These specimens
cut
cut from the middle parts of the walls along the longitudinal direction and then machined to the
from the middle parts of the walls along the longitudinal direction and then machined to the
were cut from the middle parts of the walls along the longitudinal direction and then machined to
required
required dimensions
dimensions by by electric
electric discharge
discharge wire
wire cutting.
cutting. The
The tensile
tensile properties
properties were
were evaluated
evaluated byby an
an
the required dimensions by electric discharge wire cutting. The tensile properties were evaluated
electronic universal
electronic universal material
material testing
testing machine
machine (Make:
(Make: Instron,
Instron, Darmstadt,
Darmstadt, Germany;
Germany; Model:
Model: Instron
Instron
by an electronic universal material testing machine (Make: Instron, Darmstadt, Germany; Model:
5966)
5966) at
at a
a strain
strain rate
rate of
of 9.8
9.8 ×× 10 −4 s
10−4 −1 at
s−1 at room
room temperature.
temperature. The
The tensile
tensile properties
properties were
were calculated
calculated by by
Instron 5966) at a strain rate of 9.8 × 10−4 s−1 at room temperature. The tensile properties were calculated
averaging the
averaging the measured
measured data
data ofof three
three tensile
tensile samples.
samples. The
The dimensions
dimensions ofof tensile
tensile specimens
specimens are
are shown
shown
by averaging the measured data of three tensile samples. The dimensions of tensile specimens are shown
in Figure
in Figure 3c.
3c. Besides,
Besides, the
the fracture
fracture surfaces
surfaces of of the
the tensile
tensile specimens
specimens were
were characterized
characterized by
by aa scanning
scanning
in Figure 3c. Besides, the fracture surfaces of the tensile specimens were characterized by a scanning
electron microscope (SEM) (Make: JEOL, Tokyo, Japan; Model:
electron microscope (SEM) (Make: JEOL, Tokyo, Japan; Model: JSM-6610LV). JSM-6610LV).
electron microscope (SEM) (Make: JEOL, Tokyo, Japan; Model: JSM-6610LV).
3.
3. Results
Results and Discussion
3. Resultsand
andDiscussion
Discussion

3.1. Macrostructure
3.1. Macrostructure
Figure
Figure 55 presents
presents the
the surface
surface morphologies
morphologies of of the additively
additively manufactured
the additively manufactured samples
samples fabricated
fabricated
at different pulse frequencies. Coarse periodic
at different pulse frequencies. Coarse periodic weld ripples are observed at the surface
periodic weld ripples are observed at the surface of the of the walls
walls
deposited by
deposited by low
low pulse
pulse frequency
frequency as
frequency as shown
as shown in
shown in Figure
in Figure 5e,f.
Figure 5e,f. With
5e,f. With the
With the increase
the increase of
increase of the
of the pulse
the pulse frequency,
pulse frequency,
frequency,
periodic
periodic weld
periodic weld ripples
weldripples become
ripplesbecome
become finer.
finer.
finer. Moreover,
Moreover,
Moreover, at high
at high
at high pulsepulse
pulse frequency,
frequency,
frequency, there
there there are
are so are
manyso many
so weld weld
manyripples
weld
ripples
ripples that they seem to become continuous. The wall surface is smoother as shown in Figure 5a,b.
that they seem to become continuous. The wall surface is smoother as shown in Figure 5a,b.
It can
It can be
be seen
seen that
that pulse
pulse frequency
frequency hashas aa significant
significant effect
effect on
on the
the surface
surface morphologies.
morphologies. WeldWeld ripples
ripples
Materials 2016, 9, 823 6 of 13

that they seem to become continuous. The wall surface is smoother as shown in Figure 5a,b. It6 of
Materials 2016, 9, 823
can12
be seen that pulse
Materials 2016, 9, 823 frequency has a significant effect on the surface morphologies. Weld ripples 6 ofare
12
indications
are indicationsof solidified droplets.
of solidified During
droplets. the welding
During process,
the welding the wire
process, theiswire
melted at the peak
is melted at thecurrent
peak
are indications
period and the of solidified
molten dropletdroplets.
is During
adhered to the
the welding
wire due toprocess,
the the
surface wire is
tension
current period and the molten droplet is adhered to the wire due to the surface tension at the base melted
at the at
base the peak
current
current
current period
period. When the
period. and
When thethe
base molten
current droplet is adhered
turns into
base current turns the
intopeakto peak
the the wire
current, due previous
to the
previous
current, surface
molten tension
droplet
molten at drops
drops
droplet the
onbase
the
on
current
substrate period.
or the
the substrate orWhen
previousthedeposition
basedeposition
the previous current turns
layers. intoHence,
Hence,
layers. the
at peak
at current,
a low apulse previous
frequency,
low pulse molten
there
frequency, droplet
isthere
more drops
istime
more on
fortime
the
the substrate
droplet
for the droplet ortothe
to become previous
bigger,
become deposition
resulting
bigger, in weld
resultinglayers.
ripplesHence,
in weld being at a being
low and
coarser
ripples pulse frequency,
less
coarser than less there
andthose is more
at high
than those at time
frequency.
high
for the
Besides,
frequency.droplet to
theBesides, become
number of theweldbigger,
ripples
number resulting
of indicates in weld ripples
the number
weld ripples indicates being
of the
weld coarser and less
pool oscillations.
number of weld pool than those
As pulse at high
frequency
oscillations. As
frequency.
increases, Besides,
the molten the number
pool is of
stirred weld
more ripples indicates
intensely.
pulse frequency increases, the molten pool is stirred more intensely. the number of weld pool oscillations. As
pulse frequency increases, the molten pool is stirred more intensely.

Figure 5.
Figure 5. Surface
Surface morphologies
morphologies of the samples
of the samples deposited
deposited by
by different pulse frequencies:
different pulse frequencies: (a)
(a) 500
500 Hz;
Hz;
Figure
(b) 100 5.
Hz;Surface
(c) 10 morphologies
Hz; (d) 5 Hz; of2the
(e) Hz samples
and (f) 1deposited
Hz. by different pulse frequencies: (a) 500 Hz;
(b) 100 Hz; (c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.
(b) 100 Hz; (c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.
Figure 6 displays the cross sections of the samples deposited by different pulse frequencies. At
Figure 66 displays
Figure displaysthe thecross
cross sectionsofofthe the samples deposited by by different pulse frequencies.
a lower pulse frequency (F < 5), asections
multilayer sample samples deposited
is deposited with different
few changes pulse infrequencies.
its width from At
aAt a lower
lower pulsepulse frequency
frequency (F < (F
5),<a5), a multilayer
multilayer samplesampleis is deposited
deposited with with
few
the substrate to the top as shown in Figure 6e,f. At the frequencies of 5 Hz and 10 Hz, a multilayer
few changes
changes in its in its width
width from
fromsubstrate
the the substrate
to thetotop
the as
top as shown in Figure6e,f.6e,f.
At Atthethe frequencies ofof5 5HzHzandand10 10Hz,
Hz,aa multilayer
multilayer
sample is deposited with shown
its widthin Figure
becoming larger frequencies
from the substrate to the top as shown in
sample isisdeposited
sample deposited with its width
with its widthbecoming
becominglargerlarger
from thefromsubstrate to the top
the sample
substrate toasthe
showntopasin Figure
asits
shown 6c,d.
Figure 6c,d. At a higher pulse frequency (F > 10), a multilayer is deposited width in is
At a higher
Figure 6c,d. pulse
At a frequency
higher (F >frequency
pulse 10), a multilayer
(F > 10),sample
a is deposited
multilayer sample as its
is width
deposited is smaller
as its than
width that
is
smaller than that of samples deposited by 5 Hz and 10 Hz and changes slightly from the substrate to
of samples
smaller than deposited
that of by 5 Hzdeposited
and 10 Hz and changes
and 10slightly andfrom the substrate to the thetop as shown
the top as shown in samples
Figure 6a,b. Figure by 5 Hz
7 quantitatively Hz
describeschanges slightly
the effect of thefrom
pulse substrate
frequency to
on
in Figure
the top as 6a,b.
shown Figure
in 7 quantitatively
Figure 6a,b. Figure describes
7 the effect
quantitatively of the pulse
describes the frequency
effect of the on thefrequency
pulse geometryon of
the geometry of deposited layers. It can be seen that as the pulse frequency increases, the layer
deposited
the geometrylayers. It can
of deposited be seen that as the pulse frequency increases, the layer thickness decreases
thickness decreases firstly layers.
and then It can be seenand
increases thatremains
as the pulse
stable,frequency
while theincreases,
maximum the width
layer
firstly anddecreases
thickness then increasesfirstlyand andremains
then stable, while
increases and the maximum
remains width
stable, whileincreases
the firstly andwidth
maximum then
increases firstly and then decreases to a certain value. The sample fabricated at the frequency of
decreases firstly
increases to a certain
and thenvalue. The samplea fabricated
decreases at the frequency of 10 Hz has the smallest layer
10 Hz has the smallest layer thicknessto(1.7 certain
mm) and value. The
the largest sample
maximum fabricated
width at (7.3the frequency
mm). of
thickness (1.7 mm) and the largest maximum width (7.3 mm).
10 Hz has the smallest layer thickness (1.7 mm) and the largest maximum width (7.3 mm).

Figure 6. Cross sections of the samples deposited by different pulse frequencies: (a) 500 Hz; (b) 100
Figure 6.
Hz; (c) 6. Cross
10Cross sections
Hz; (d) 5 Hz; (e)of 2the
Hz samples 1deposited
and (f)deposited
Hz. by different pulse frequencies: (a) 500 Hz; (b) 100
Figure sections of the samples by different pulse frequencies: (a) 500 Hz; (b) 100 Hz;
Hz; (c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.
(c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.
Materials 2016, 9, 823 7 of 13
Materials 2016, 9, 823 7 of 12

Materials 2016, 9, 823 7 of 12

Figure 7. Effect of pulse frequency on the geometry of deposited

deposited layers.
layers.

Within the bounds of statistical variation, the geometry of the 1, 2, 100, and 500 Hz samples are
Within the bounds of 7.statistical
Figure variation,
Effect of pulse frequencythe geometry
on the geometryof the 1, 2, layers.
of deposited 100, and 500 Hz samples
all virtually identical. The only ones that are statistically different are the 5 Hz and 10 Hz samples.
are all virtually identical. The only ones that are statistically different are the 5 Hz and 10 Hz
The geometryWithin of
thedeposited
bounds layers isvariation,
of statistical determined by the of
the geometry shape1,of the andweld500pool. Pulse are
current
samples. The geometry of deposited layers is determined by thethe shape 2, 100,
of the weld Hz samples
pool. Pulse current
promotes the fluid
all virtually flow
identical.and agitation
Theagitation
only onesof of the
that weld pool [24], and with the increase of pulse frequency,
promotes the fluid flow and theare statistically
weld pool [24],different
and with are the
the5increase
Hz and 10 of Hz samples.
pulse frequency,
the molten pool is stirred
The geometry more intensely
of deposited as mentioned
layers is determined above.
by the shapeIt isoflikely that pool.
the weld a frequency of 5–10 Hz
Pulse current
the molten pool is stirred more intensely as mentioned above. It is likely that a frequency of 5–10 Hz
causespromotes
the weld thepool
fluid to
flow and agitation
oscillate of the weld
at its natural pool [24],This
frequency. and withcausesthe the
increase
oddof pulse frequency,
shaped deposits with
causes the weld pool to oscillate at its natural frequency. This causes the odd shaped deposits with
the molten
these frequencies. pool is stirred more intensely as mentioned above. It is likely that a frequency of 5–10 Hz
thesecauses
frequencies.
the weld pool to oscillate at its natural frequency. This causes the odd shaped deposits with
these frequencies.
3.2. Microstructure
Microstructure
3.2.
Figure 88 exhibits
exhibits the
3.2. Microstructure
Figure the microstructures
microstructures of of the
the samples
samples deposited
deposited by by different
different pulse
pulse frequencies.
frequencies.
No pore defect
No pore defect
Figure is is observed
observed
8 exhibits theinin these microstructure
these microstructure
microstructures images
images
of the samples and fully
and fully
deposited dense
by dense AZ31
AZ31
different pulsemagnesium
magnesium alloy
frequencies.alloy
components
No pore are
components are obtained.
defect is observed
obtained. It
It is noted
isinnoted that
these that there are
microstructure
there are huge
huge differences
imagesdifferences between
and fully dense
between these
AZ31these microstructures
magnesium alloy
microstructures
in terms of
components grain.
are Table
obtained. 2 summarizes
It is noted thatthe
theregrain
are size
huge of samples
differences
in terms of grain. Table 2 summarizes the grain size of samples fabricated by different pulse fabricated
between these by different
microstructures pulse
frequencies.
in terms With
of the
grain. enhancement
Table 2 of pulse
summarizes frequency,
the grain grain
size of size
samplesdecreases firstly
fabricated
frequencies. With the enhancement of pulse frequency, grain size decreases firstly and then increases. by and then
different increases.
pulse
frequencies. Withof
The microstructures
microstructures thesamples
enhancement of pulse
fabricated byfrequency, grain
Hzsize
aredecreases firstly
andand then increases.
The of samples fabricated by 55 Hz
Hz and
and 1010 Hz are fairly fine
fairly fine and uniform,
uniform, whose grain
whose grain
size The
is 21 microstructures
μm as shown of
in samples
Figure fabricated
8c,d. Coarseby 5 Hz and
grains 10
can Hz
be are fairly
observed fine
in and uniform,
Figure whose
8a,b,f. The grain
grain size
size is 21 µm as shown in Figure 8c,d. Coarse grains can be observed in Figure 8a,b,f. The grain size of
size is 21 μm as shown in Figure 8c,d. Coarse grains can be observed in Figure 8a,b,f. The grain size
of the sample fabricated by 100 Hz
the sample fabricated by 100 Hz is even up to 39 µm. is even up to 39 μm.
of the sample fabricated by 100 Hz is even up to 39 μm.

Figure 8. Microstructures of the samples deposited by different pulse frequencies: (a) 500 Hz; (b) 100
Figure 8. Microstructures of the samples deposited by different pulse frequencies: (a) 500 Hz; (b) 100 Hz;
FigureHz;8.(c) 10 Hz; (d) 5 Hz; (e)the
Microstructures 2 Hzsamples
and (f) 1 Hz.
(c) 10 Hz; (d) 5 Hz; (e) 2 Hzofand (f) 1 Hz. deposited by different pulse frequencies: (a) 500 Hz; (b) 100
Hz; (c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.
Materials 2016, 9, 823 8 of 13

Table 2. Grain size and grain aspect ratio of the samples fabricated by different pulse frequencies.

Pulse Frequency (Hz) Grain Size (µm) Grain Aspect Ratio

1 31 ± 4 1.09 ± 0.04
2 23 ± 1 1.15 ± 0.06
5 21 ± 1 1.02 ± 0.12
10 21 ± 1 0.98 ± 0.11
100 39 ± 4 1.22 ± 0.21
500 37 ± 3 1.14 ± 0.10

Besides, Table 2 summarizes the grain aspect ratio of samples fabricated by different pulse
frequencies. The grain aspect ratio reflects whether there are differences in grain size along the
horizontal and vertical direction, which indicate anisotropic mechanical properties. The grain aspect
ratio of the samples fabricated at 5 Hz and 10 Hz is close to 1, which means the same grain size along the
horizontal and vertical direction as shown in Figure 8c,d and indicates isotropic mechanical properties.
The grain aspect ratio of the 1, 2, 100, and 500 Hz samples is slightly larger than 1, which indicates
slightly anisotropic mechanical properties.
Apparently, pulse frequency has an appreciable influence on microstructure, especially grain
size and grain aspect ratio. There are mainly two factors of pulse frequency contributing to the grain
growing behavior.
Firstly, pulse current can cause the stirring of the molten pool, leading to refining grains.
Increased pulse current results in an increased plasma momentum and increases the value of the
electromagnetic force. Plasma momentum creates arc pressure and shearing force on the surface
of the weld pool. Electromagnetic force inside the weld pool pushes the fluid radially inwards
and then down the axis [34]. This can produce enough weld pool oscillations and cause dendrite
fragmentation to provide more heterogeneous nucleation sites so as to refine the grain size [13] as
shown in Figure 9b. As the pulse frequency grows, weld pool oscillations increase, which is consistent
with the phenomenon of the weld ripples, and the grain refinement effect is increased.
Secondly, the pulse current can cause the high cooling rate of the molten pool, resulting in refining
grains. The peak current is mainly used to input heat to melt the metal wire while the base current
is much smaller than the peak current and is used to maintain the arc. During the base current,
the heat input was suddenly reduced and therefore the cooling rate of the molten pool is high causing
heterogeneous nucleation to refine grains [38,39] as shown in Figure 9c. When pulse frequency changes,
total heat input remains constant. With the increase of pulse frequency, arc pressure is becoming
higher, resulting in a more constricted and longer arc, so that arc is large with a conical shape at first,
and then becomes smaller and finally even appears to be steady in a near cylindrical shape. Meanwhile,
as the pulse frequency increases, the duration of the base current becomes shorter and the time for
heat dissipation is reduced. In a word, higher pulse frequency results in a larger effective heat input,
higher temperature and lower cooling rate [34].
In the case of the two reasons above, there exists an optimum pulse frequency of maximum grain
refinement due to both more weld pool oscillations and a higher cooling rate of the molten pool,
leading to more heterogeneous nucleation sites as shown in Figure 9d. In this study, the optimum
pulse frequency is found to be 5 Hz and 10 Hz. The samples fabricated at 5 Hz and 10 Hz contain
finer equiaxed grains (21 µm) whose grain aspect ratio is close to 1 at the same time. Meanwhile,
the resonance of the weld pool at these frequencies, as mentioned previously, greatly promotes
grain refinement.
Materials 2016, 9, 823 9 of 13
Materials 2016, 9, 823 9 of 12

Figure
Figure 9.9.Schematic
Schematicdiagram
diagram of effect
of the the effect of frequency
of pulse pulse frequency on nucleation:
on nucleation: (a) conventional
(a) conventional nucleation
nucleation without pulse current; (b) nucleation under weld pool oscillations; (c) nucleation
without pulse current; (b) nucleation under weld pool oscillations; (c) nucleation under high cooling under
high cooling
rate and rate and (d)
(d) nucleation nucleation
rate curves. rate curves.

3.3.
3.3. Tensile Properties
Tensile Properties
Tensile testing was
Tensile testing was conducted
conducted to to discover
discover thethe effect
effect of
of pulse
pulse frequency
frequency on onmechanical
mechanicalproperties.
properties.
Table 3 summarizes the tensile properties of the samples fabricated by different pulse
Table 3 summarizes the tensile properties of the samples fabricated by different pulse frequencies,
frequencies,
including ultimatetensile
including ultimate tensile strength
strength (UTS),
(UTS), yieldyield strength
strength (YS) and(YS) and elongation
elongation (EL). The(EL).
tensileThe tensile
properties
properties change significantly with the variation of pulse frequency (UTS
change significantly with the variation of pulse frequency (UTS is from 221 to 263 MPa and YS is from 221 to 263 MPa is
and YS is from 79 to 104 MPa). It should be noted that the samples fabricated
from 79 to 104 MPa). It should be noted that the samples fabricated at 5 Hz and 10 Hz exhibit higherat 5 Hz and 10 Hz
exhibit
UTS (260higher
MPa)UTSand(260 MPa)
YS (102 and YS
MPa), (102are
which MPa), which
similar are similar
to those of the to thoseAZ31
forged of thealloy
forged AZ31UTS
whose alloy
is
whose UTS is 234 MPa and YS is 131 MPa according to ASTM standard B91-12
234 MPa and YS is 131 MPa according to ASTM standard B91-12 [40]. Moreover, the elongation of all [40]. Moreover, the
elongation of all samples
samples is above is above
23%, which 23%,good
indicates which indicates
plastic good plastic property.
property.

Table
Table 3.
3. Tensile properties of
Tensile properties of the
the samples
samples fabricated
fabricated by
by different
different pulse
pulse frequencies.
frequencies.

Pulse Frequency (Hz) UTS (MPa) YS (MPa) EL (%)

Pulse Frequency (Hz) UTS (MPa) YS (MPa) EL (%)
1 229 ± 3 90 ± 1 26.3 ± 0.3
12 232±±31
229 89 ±
90 ± 11 26.3 ±
26.7 0.3
± 1.2
2 232 ± 1 89 ± 1 26.7 ± 1.2
5
5 258 ± 5
258 ± 5
100 ± 3
100 ± 3
25.6 ± 3.1
25.6 ± 3.1
10
10 263 ±±55
263 104
104 ±± 55 23.0
23.0 ± 3.7
± 3.7
100
100 221±±11
221 11 82 ±
82 ±6 23.4 ±
23.4 ± 5.8
500
500 231±±55
231 79 ±
79 ± 22 27.3 ±
27.3 0.3
± 0.3

Figure
Figure 1010showsshows thethe
SEMSEM fractographs
fractographs of tensile
of tensile samplessamples
depositeddeposited
by different bypulse
different pulse
frequencies.
frequencies. It is clear that all fracture surfaces present a dimple rupture. Compared
It is clear that all fracture surfaces present a dimple rupture. Compared with others, the fracture with others, the
fracture surfaces of the tensile specimens manufactured at 5 Hz and 10 Hz contain
surfaces of the tensile specimens manufactured at 5 Hz and 10 Hz contain a lot of much finer dimples a lot of much finer
dimples
which whichrelatively
indicate indicate relatively higher
higher tensile tensile The
strength. strength. The fractographs
fractographs are consistent
are consistent with tensilewith tensile
properties
properties as well as the
as well as the microstructure. microstructure.
The
The samples
samples fabricated
fabricated atat 55 Hz
Hz and
and 1010 Hz
Hz contain
contain finer
finer equiaxed
equiaxed grains
grains with
with nono pore
pore defect
defect and
and
achieve
achieve thethe maximum
maximumtensile tensilestrength
strengthinincomparison
comparisontotoothers.
others.Moreover,
Moreover, thethe
larger
larger thethe
grain size
grain is,
size
the worse the tensile strength becomes. Many researchers have studied the
is, the worse the tensile strength becomes. Many researchers have studied the relationship between relationship between
mechanical
mechanical properties
properties and and microstructure.
microstructure. The The mechanical
mechanical property
property of
of the
the alloy
alloy is directly affected
is directly affected
by the average grain size according to the Hall–Petch equation [41,42]. Therefore,
by the average grain size according to the Hall–Petch equation [41,42]. Therefore, the formation of the formation of
finer grains of the samples contributes to higher tensile strength and coarser grains
finer grains of the samples contributes to higher tensile strength and coarser grains easily lead to easily lead to poor
tensile
poor properties.
tensile properties. Grain size size
Grain depends
depends on on
thetheprocess
processparameters
parametersand
andpulse
pulsefrequency
frequencycan can refine
refine
grains and enhance tensile properties by controlling weld pool oscillations and the cooling rate of the
molten pool.
Materials 2016, 9, 823 10 of 13

grains and enhance tensile properties by controlling weld pool oscillations and the cooling rate of the
molten
Materials pool.
2016, 9, 823 10 of 12

Figure 10.
Figure 10. SEM
SEM fractographs
fractographs of
of tensile
tensile samples
samples deposited
deposited by
by different pulse frequencies:
different pulse frequencies: (a)
(a) 500
500 Hz;
Hz;
(b) 100 Hz; (c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.
(b) 100 Hz; (c) 10 Hz; (d) 5 Hz; (e) 2 Hz and (f) 1 Hz.

It can be concluded that pulse frequency has a remarkable influence on microstructure and
It can be concluded that pulse frequency has a remarkable influence on microstructure and
mechanical properties of wire arc additive manufactured AZ31 magnesium alloy. Fine microstructure
mechanical properties of wire arc additive manufactured AZ31 magnesium alloy. Fine microstructure
and good mechanical properties can be obtained by adjusting pulse frequency. The results indicate
and good mechanical properties can be obtained by adjusting pulse frequency. The results indicate
that optimum pulse frequency is found to be 5 Hz and 10 Hz, which leads to finer equiaxed grains
that optimum pulse frequency is found to be 5 Hz and 10 Hz, which leads to finer equiaxed grains and
and better mechanical properties compared with other pulse frequencies. Although the shape of
better mechanical properties compared with other pulse frequencies. Although the shape of samples
samples fabricated at 5 Hz and 10 Hz is poor, good geometry accuracy can be reached by adjusting
fabricated at 5 Hz and 10 Hz is poor, good geometry accuracy can be reached by adjusting the relative
the relative manufacturing process. The arc current can control the geometry of deposited layers [32]
manufacturing process. The arc current can control the geometry of deposited layers [32] and reducing
and reducing the arc current is often used to obtain good geometry accuracy during deposition
the arc current is often used to obtain good geometry accuracy during deposition [17,18].
[17,18].
4. Conclusions
4. Conclusions
Wire arc additive manufacturing of AZ31 magnesium alloy has been carried out in this
Wire arc additive manufacturing of AZ31 magnesium alloy has been carried out in this
investigation. The effects of pulse frequency on macrostructure, microstructure and tensile properties
investigation. The effects of pulse frequency on macrostructure, microstructure and tensile properties
of the arc additive manufactured samples are analyzed. From this investigation, the important
of the arc additive manufactured samples are analyzed. From this investigation, the important
conclusions are as follows:
conclusions are as follows:
(1) Fully dense AZ31 magnesium alloy components are successfully fabricated using WAAM.
(1) Fully dense AZ31 magnesium alloy components are successfully fabricated using WAAM.
WAMM has the potential to manufacture magnesium alloys components.
WAMM has the potential to manufacture magnesium alloys components.
(2) Pulse frequency
(2) Pulse frequency has has aa significant
significant influence
influence on on macrostructure,
macrostructure, microstructure
microstructure and and tensile
tensile
properties of the wire arc additive manufactured samples. The shape
properties of the wire arc additive manufactured samples. The shape of samples fabricated of samples fabricated at
5atHz5 Hz
andand10 10
HzHz is odd
is odd due
due toto theweld
the weldpool
pooloscillating
oscillatingatatits
itsnatural
natural frequency.
frequency. AsAs the
the pulse
pulse
frequencygrows,
frequency grows,weldweldpoolpooloscillations
oscillationsincrease
increaseandand the
the cooling
cooling rate
rate decreases
decreases leading
leading to to the
the
fact that grain size decreases firstly and then increases, meanwhile, the grain aspect ratio is close
fact that grain size decreases firstly and then increases, meanwhile, the grain aspect ratio is close
toto11firstly
firstlyand
andthen
thenaway
awayfrom
from1. 1. Grain
Grainrefinement
refinement reaches
reaches the
the maximum
maximum at at 5–10
5–10 Hz,
Hz, owing
owing to to
resonanceofofthe
resonance the weld
weld pool
pool atat these
these frequencies
frequencies as as mentioned
mentioned previously. Correspondingly,
previously. Correspondingly,
tensilestrength
tensile strengthincreases
increasesfirstly
firstlyand
andthen
thendecreases.
decreases.
(3) Finegrains
(3) Fine grainsand
andgood
goodtensile
tensileproperties
propertiescancanbe beobtained
obtained at
ataacertain
certain pulse
pulse frequency.
frequency. The
The samples
samples
fabricated at 5 Hz and 10 Hz contain finer equiaxed grains (21 µm) and exhibit
fabricated at 5 Hz and 10 Hz contain finer equiaxed grains (21 μm) and exhibit higher ultimate higher ultimate
tensilestrength
tensile strength(260
(260MPa)
MPa)and andyield
yieldstrength
strength(102(102 MPa),
MPa), which
which areare similar
similar to
to those
those of
of the
the forged
forged
AZ31 alloy. Moreover, the elongation of all samples
AZ31 alloy. Moreover, the elongation of all samples is above 23%. is above 23%.

Acknowledgments: The work was financially supported by the National Natural Science Foundation of China
(51505033), Beijing Natural Science Foundation (3162027), and Excellent Young Scholars Research Fund of
Beijing Institute of Technology (2015YG0302).
Materials 2016, 9, 823 11 of 13

Acknowledgments: The work was financially supported by the National Natural Science Foundation of China
(51505033), Beijing Natural Science Foundation (3162027), and Excellent Young Scholars Research Fund of Beijing
Institute of Technology (2015YG0302).
Author Contributions: Jing Guo performed all experiments and wrote this manuscript. Changmeng Liu designed
the research, helped analyze the experimental data and gave some constructive suggestions. Yong Zhou,
Qianru Wu, Xianping Chen and Jiping Lu participated in the discussion on the results and guided the writing of
the article.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Haferkamp, H.; Boehm, R.; Holzkamp, U.; Jaschik, C.; Kaese, V.; Niemeyer, M. Alloy development,
processing and applications in magnesium lithium alloys. Mater. Trans. 2001, 42, 1160–1166. [CrossRef]
2. Mordike, B.L.; Ebert, T. Magnesium: Properties–applications–potential. Mater. Sci. Eng. A 2001, 302, 37–45.
[CrossRef]
3. Chai, F.; Zhang, D.; Li, Y. Effect of thermal history on microstructures and mechanical properties of az31
magnesium alloy prepared by friction stir processing. Materials 2014, 7, 1573–1589. [CrossRef]
4. Zhang, J.; Xu, C.; Jing, Y.; Lv, S.; Liu, S.; Fang, D.; Zhuang, J.; Zhang, M.; Wu, R. New horizon for high
performance mg-based biomaterial with uniform degradation behavior: Formation of stacking faults. Sci. Rep.
2015, 5, 13933. [CrossRef] [PubMed]
5. Cao, X.; Xiao, M.; Jahazi, M.; Immarigeon, J.P. Continuous wave nd:Yag laser welding of sand-cast ze41a-t5
magnesium alloys. Mater. Manuf. Process. 2005, 20, 987–1004. [CrossRef]
6. Feliu, J.S.; Samaniego, A.; Bermudez, E.; El-Hadad, A.; Llorente, I.; Galván, J. Effect of native oxide film on
commercial magnesium alloys substrates and carbonate conversion coating growth and corrosion resistance.
Materials 2014, 7, 2534–2560. [CrossRef]
7. Padmanaban, G.; Balasubramanian, V. Optimization of pulsed current gas tungsten arc welding process
parameters to attain maximum tensile strength in az31b magnesium alloy. Trans. Nonferr. Met. Soc. China
2011, 21, 467–476. [CrossRef]
8. Easton, M.; Beer, A.; Barnett, M.; Davies, C.; Dunlop, G.; Durandet, Y.; Blacket, S.; Hilditch, T.; Beggs, P.
Magnesium alloy applications in automotive structures. JOM 2008, 60, 57–62. [CrossRef]
9. Zhu, X.; Liu, Y.; Wang, Q.; Liu, J. Influence of sulfate-reducing bacteria on the corrosion residual strength of
an az91d magnesium alloy. Materials 2014, 7, 7118–7129. [CrossRef]
10. Wei, K.; Gao, M.; Wang, Z.; Zeng, X. Effect of energy input on formability, microstructure and mechanical
properties of selective laser melted az91d magnesium alloy. Mater. Sci. Eng. A 2014, 611, 212–222. [CrossRef]
11. Wei, K.; Wang, Z.; Zeng, X. Influence of element vaporization on formability, composition, microstructure,
and mechanical performance of the selective laser melted mg–zn–zr components. Mater. Lett. 2015, 156,
187–190. [CrossRef]
12. Lancaster, R.; Davies, G.; Illsley, H.; Jeffs, S.; Baxter, G. Structural integrity of an electron beam melted
titanium alloy. Materials 2016, 9, 470. [CrossRef]
13. Wang, F.; Williams, S.; Rush, M. Morphology investigation on direct current pulsed gas tungsten arc welded
additive layer manufactured ti6al4v alloy. Int. J. Adv. Manuf. Technol. 2011, 57, 597–603. [CrossRef]
14. Frazier, W.E. Metal additive manufacturing: A review. J. Mater. Eng. Perform. 2014, 23, 1917–1928. [CrossRef]
15. Chen, C.-H.; Liu, J.; Chua, C.-K.; Chou, S.-M.; Shyu, V.; Chen, J.-P. Cartilage tissue engineering with silk
fibroin scaffolds fabricated by indirect additive manufacturing technology. Materials 2014, 7, 2104–2119.
[CrossRef]
16. Antonysamy, A.A.; Meyer, J.; Prangnell, P.B. Effect of build geometry on the β-grain structure and texture
in additive manufacture of Ti 6Al 4V by selective electron beam melting. Mater. Charact. 2013, 84, 153–168.
[CrossRef]
17. Zhang, J.; Wang, X.; Paddea, S.; Zhang, X. Fatigue crack propagation behaviour in wire + arc additive
manufactured Ti-6Al-4V: Effects of microstructure and residual stress. Mater. Des. 2016, 90, 551–561.
[CrossRef]
18. Lin, J.J.; Lv, Y.H.; Liu, Y.X.; Xu, B.S.; Sun, Z.; Li, Z.G.; Wu, Y.X. Microstructural evolution and mechanical
properties of Ti-6Al-4V wall deposited by pulsed plasma arc additive manufacturing. Mater. Des. 2016, 102,
30–40. [CrossRef]
Materials 2016, 9, 823 12 of 13

19. Liu, C.; Lu, Y.; Tian, X.; Liu, D. Influence of continuous grain boundary α on ductility of laser melting
deposited titanium alloys. Mater. Sci. Eng. A 2016, 661, 145–151. [CrossRef]
20. Chi, C.N.; Savalani, M.; Man, H.C. Fabrication of magnesium using selective laser melting technique.
Rapid Prototyp. J. 2011, 17, 479–490.
21. Hu, D.; Wang, Y.; Zhang, D.; Hao, L.; Jiang, J.; Li, Z.; Chen, Y. Experimental investigation on selective laser
melting of bulk net-shape pure magnesium. Mater. Manuf. Process. 2015, 30, 1298–1304. [CrossRef]
22. Zhang, B.; Liao, H.; Coddet, C. Effects of processing parameters on properties of selective laser melting
Mg–9%Al powder mixture. Mater. Des. 2012, 34, 753–758. [CrossRef]
23. Al-Kazzaz, H.; Medraj, M.; Cao, X.; Jahazi, M. Nd:Yag laser welding of aerospace grade ZE41A magnesium
alloy: Modeling and experimental investigations. Mater. Chem. Phys. 2008, 109, 61–76. [CrossRef]
24. Wang, F.; Williams, S.; Colegrove, P.; Antonysamy, A.A. Microstructure and mechanical properties of wire
and arc additive manufactured Ti-6Al-4V. Metall. Mater. Trans. A 2012, 44, 968–977. [CrossRef]
25. Martina, F.; Mehnen, J.; Williams, S.W.; Colegrove, P.; Wang, F. Investigation of the benefits of plasma
deposition for the additive layer manufacture of Ti-6Al-4V. J. Mater. Process. Technol. 2012, 212, 1377–1386.
[CrossRef]
26. Ding, D.; Pan, Z.; Cuiuri, D.; Li, H. Wire-feed additive manufacturing of metal components: Technologies,
developments and future interests. Int. J. Adv. Manuf. Technol. 2015, 81, 465–481. [CrossRef]
27. Brandl, E.; Baufeld, B.; Leyens, C.; Gault, R. Additive manufactured Ti-6Al-4V using welding wire:
Comparison of laser and arc beam deposition and evaluation with respect to aerospace material specifications.
Phys. Proced. 2010, 5, 595–606. [CrossRef]
28. Shen, C.; Pan, Z.; Ma, Y.; Cuiuri, D.; Li, H. Fabrication of iron-rich Fe–Al intermetallics using the wire-arc
additive manufacturing process. Addit. Manuf. 2015, 7, 20–26. [CrossRef]
29. Xiong, J.; Zhang, G.; Zhang, W. Forming appearance analysis in multi-layer single-pass gmaw-based additive
manufacturing. Int. J. Adv. Manuf. Technol. 2015, 80, 1767–1776. [CrossRef]
30. Ding, D.; Pan, Z.; van Duin, S.; Li, H.; Shen, C. Fabricating superior nial bronze components through wire
arc additive manufacturing. Materials 2016, 9, 652. [CrossRef]
31. Bermingham, M.J.; Kent, D.; Zhan, H.; StJohn, D.H.; Dargusch, M.S. Controlling the microstructure and
properties of wire arc additive manufactured Ti-6Al-4V with trace boron additions. Acta Mater. 2015, 91,
289–303. [CrossRef]
32. Wang, H.; Jiang, W.; Ouyang, J.; Kovacevic, R. Rapid prototyping of 4043 Al-alloy parts by VP-GTAW.
J. Mater. Process. Technol. 2004, 148, 93–102. [CrossRef]
33. Ouyang, J.H.; Wang, H.; Kovacevic, R. Rapid prototyping of 5356-aluminum alloy based on variable polarity
gas tungsten arc welding: Process control and microstructure. Mater. Manuf. Process. 2002, 17, 103–124.
[CrossRef]
34. Subravel, V.; Padmanaban, G.; Balasubramanian, V. Effect of pulse frequency on tensile properties and
microstructural characteristics of gas tungsten arc welded AZ31B magnesium alloy. Trans. Indian Inst. Met.
2014, 68, 353–362. [CrossRef]
35. Padmanaban, G.; Balasubramanian, V. Influences of pulsed current parameters on mechanical and
metallurgical properties of gas tungsten arc welded AZ31B magnesium alloys. Met. Mater. Int. 2011,
17, 831–839. [CrossRef]
36. Subravel, V.; Padmanaban, G.; Balasubramanian, V. Effect of welding speed on microstructural characteristics
and tensile properties of gta welded AZ31B magnesium alloy. Trans. Nonferr. Met. Soc. China 2014, 24,
2776–2784. [CrossRef]
37. Liu, C.M.; Wang, H.M.; Tian, X.J.; Tang, H.B.; Liu, D. Microstructure and tensile properties of laser melting
deposited Ti–5Al–5Mo–5V–1Cr–1Fe near β titanium alloy. Mater. Sci. Eng. A 2013, 586, 323–329. [CrossRef]
38. Sharir, Y.; Pelleg, J.; Grill, A. Effect of arc vibration and current pulses on microstructure and mechanical
properties of tig tantalum welds. Met. Technol. 2013, 5, 190–196. [CrossRef]
39. Kou, S. Welding Metallurgy, 2nd ed.; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2003.
40. ASTM; B91-12; ASTM International: West Conshohocken, PA, USA, 2012.
Materials 2016, 9, 823 13 of 13

41. Jia, W.; Ma, L.; Tang, Y.; Le, Q.; Fu, L. Relationship between microstructure and properties during multi-pass,
variable routes and different initial temperatures hot flat rolling of AZ31B magnesium alloy. Mater. Des.
2016, 103, 171–182. [CrossRef]
42. Zhu, Y.Z.; Wang, S.Z.; Li, B.L.; Yin, Z.M.; Wan, Q.; Liu, P. Grain growth and microstructure evolution based
mechanical property predicted by a modified hall–petch equation in hot worked Ni76Cr19AlTiCo alloy.
Mater. Des. 2014, 55, 456–462. [CrossRef]

© 2016 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/).