Journal of Alloys and Compounds 753 (2018) 247e255

Contents lists available at ScienceDirect

Journal of Alloys and Compounds

journal homepage: http://www.elsevier.com/locate/jalcom

Optimising the mechanical properties of Ti-6Al-4V components

produced by wire þ arc additive manufacturing with post-process
heat treatments
M.J. Bermingham a, *, L. Nicastro a, D. Kent b, Y. Chen c, d, M.S. Dargusch a
a
Centre for Advanced Materials Processing and Manufacturing, School of Mechanical and Mining Engineering, The University of Queensland, St Lucia,
Queensland, 4072, Australia
b
School of Science and Engineering, University of the Sunshine Coast, Maroochydore DC, Queensland, 4558, Australia
c
School of Materials, University of Manchester, Manchester, M13 9PL, UK
d
Research Complex at Harwell, Rutherford Appleton Laboratory, Harwell, OX11 0FA, Oxfordshire, UK

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

Article history: Wire þ Arc Additive Manufacturing (WAAM) is a promising manufacturing process for producing large
Received 11 November 2017 aerospace components. Based on welding technology, the process is highly affordable, has a very high
Received in revised form deposition rate and is not limited by chamber size. Ti-6Al-4V is a promising candidate material for this
11 April 2018
technology given that it is extensively used in aerospace applications and some large, high buy-fly ratio
Accepted 12 April 2018
Available online 21 April 2018
components can be more efficiently produced by WAAM than via the conventional machining from billet
approach. There is currently limited knowledge about whether additional post processes including heat
treatments and hot isostatic pressing are necessary to unlock the optimal mechanical properties of Ti-
Keywords:
Metals and alloys
6Al-4V components produced by WAAM. This work explores a range of different post process treat-
Liquid-solid reactions ments and the effects on the microstructure and tensile properties of Ti-6Al-4V components produced by
Mechanical properties WAAM. The relatively slow cooling rate (10-20Ks1) during the b-a transformation produced Wid-
Microstructure manst€atten-a and offered an optimal balance between strength and ductility. Hot Isostatic Pressing
(HIPing) removed gas porosity but was not effective in improving strength or ductility. Residual tensile
stresses in as-built components severely impair ductility and should be removed through stress relief
treatments.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction
Additive Manufacturing (AM) has emerged as a promising
manufacturing process for producing bespoke components,
particularly from high cost materials. Powder bed AM technologies
such as Selective Laser Melting (SLM) are relatively mature and
have received significant attention in the production of Ti-6Al-4V
components. AM technologies incorporating wire feedstock are
beginning to emerge and offer advantages including high deposi-
tion rates and the potential for large part sizes. The wire based AM
processes share a number of similarities in terms of the size, shape
and quality of potential parts but differ in terms of the heat source
(electron beam, laser beam, plasma arc, electric arc etc.) and the
chamber requirements (vacuum, inert gas or atmosphere exposed)
[1,2]. Unlike powder bed AM technologies such as SLM, which are
characterised by very small melt pools and very rapid heating and
cooling rates approaching 103-108 Ks1 [3], wire based AM tech-
nologies involve melting a very large volume of metal1 under
slower scan speeds and results in considerably higher deposition
rates, slower cooling rates and generally coarser microstructures
[2]. Of these technologies, Wire þ Arc Additive Manufacturing
(WAAM) is amongst the most affordable because it based on
inexpensive Gas Tungsten Arc Welding technology [4]. Further-
more, despite being an out of chamber process exposed to the at-
mosphere, it is possible to fabricate high quality titanium parts with
minimal oxygen pick-up using low cost localised inert gas shielding
[5]. While unlikely to be a net shape AM process (post machining is
required to achieve necessary tolerances), wire based AM

1
* Corresponding author. The melt pool diameter can exceed 10 mm during WAAM, compared to SLM
E-mail address: m.bermingham@uq.edu.au (M.J. Bermingham). which can approach 0.1 mm in diameter.

https://doi.org/10.1016/j.jallcom.2018.04.158
0925-8388/© 2018 Elsevier B.V. All rights reserved.
248 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255
technologies such as WAAM have considerable potential in fabri-
cating large-scale ribbed components for aerospace applications as
this substantially reduces the machining required if the component
was to be otherwise produced from a wrought billet.
A notable difference between wire based AM processes and
established powder-bed AM processes is the size of the molten
pool, slower torch travel speed and higher material deposition rates
which subsequently determine important solidification parameters
such as the temperature gradient G, solid growth rate R and the
cooling rate (product of G and R). Bontha et al. [6] modelled laser
deposition of Ti-6Al-4V (LENS process) and found that commonly
used small scale laser deposition parameters (representative of
powder bed technologies) will produce temperature gradients well
above 10,000 Kcm1 and growth rates of the order of 0.1cms1
resulting in fully columnar microstructures. However, the temper-
ature gradients substantially reduce by more than an order of
magnitude (<1000 Kcm1) when larger scale AM processes were
modelled (that tend to be more representative of conditions
experienced during WAAM) and it was predicted that equiaxed
coarse grained microstructures could form. The scale of the WAAM
process means that cooling rates and therefore microstructures and
properties will naturally vary from SLM.
It is now widely accepted that the rapid cooling during SLM of
Ti-6Al-4V under most processing conditions results in an as-built
microstructure containing very fine acicular a or a0 with corre-
sponding high strength and moderate to low ductility on account of
the small slip length [7]. It is also widely understood that the AM of
Ti-6Al-4V produces highly textured columnar grains [8] and this, as
well as the incomplete fusion of powder particles during manu-
facture [9,10], can cause substantial anisotropy [11]. Many re-
searchers have shown that post-build heat treatments and Hot
Isostatic Pressing (HIPing) are effective in improving ductility at the
expense of some strength of Ti-6Al-4V produced by powder bed
AM processes [9,12e14]. HIPing in particular is very effective in
improving the fatigue life and ductility especially during certain
SLM processes where unmelted powder and porosity remain. Other
researchers have shown that the selection of processing parameters
for SLM (such as scan rate) has substantial influence on the
densification of Ti-6Al-4V and when optimised can result in as-
built components with negligible porosity and improved ductility
[15,16].
While a detailed understanding of the effect of processing pa-
rameters and post process treatments of Ti-6Al-4V produced by
popular powder bed AM technologies is beginning to emerge, there
is still uncertainty surrounding how these same factors influence
the microstructure and mechanical properties of Ti-6Al-4V pro-
duced by WAAM. Initial research has shown that WAAM can pro-
duce high quality Ti-6Al-4V components with mechanical
properties equalling or exceeding other AM processes in the as-
built state. For example, the slower cooling rates during WAAM
prevents the formation of a0 martensite [17] and this greatly im-
proves the ductility of as-built products which is reported to be in
the range of 7e10% in the horizontal direction (normal to build
direction) and 12e16% in the vertical direction [4,18e20]. In
contrast, when using optimised scan parameters that minimise the
presence of porosity during SLM or LENS the typical ductility of Ti-
6Al-4V is in the range of 11e12% [15,16]. However, when processing
parameters are not necessary optimised during SLM the ductility is
generally less than 10%. In powder bed AM processes the ductility is
known to be extremely sensitive to the presence of lack of fusion
defects and porosity but these can be greatly eliminated by post-
process HIPing. For example, Kobryn and Semiatin [9] reported
ductility as low as 0.8% in stress-relieved Ti-6Al-4V produced by
LENS and this poor result was attributed to lack of fusion defects.
However, after being subjected to HIP the porosity was closed and
ductility greatly improved to almost 12%.
In light of the coarser microstructure produced and differences
in the process it remains unclear whether Ti-6Al-4V produced by
WAAM necessitates additional post process heat treatment or
expensive HIPing processes. Very limited work exists on the effects
of heat treatment on the mechanical properties of Ti-6Al-4V pro-
duced by WAAM. Brandl and co-workers [20] performed heat
treatment at 873 K and reported limited improvement on the me-
chanical properties compared with the as-built state but at 1116 K
the ductility improved considerably at a marginal expense of
strength. Bermingham et al. [21] also performed heat treatment at
1323 K on Ti-6Al-4V produced by WAAM and found that it had no
influence on the compressive ductility of Ti-6Al-4V but did
considerably reduce the compressive strength. To the authors' best
knowledge there is no literature available which investigates the
effect of HIPing on the microstructure and tensile properties of Ti-
6Al-4V produced by WAAM and compares this to the as-built or
other heat treated conditions. The fact that WAAM is not as sus-
ceptible to the same incomplete fusion defects as SLM or LENS
(blown powder) may indicate that expensive HIPing processes are
not necessary; although on the other hand the potential for
entrapment of gas porosity remains. Furthermore, the cooling rates
during WAAM are much lower than typical powder bed processes
so the tendency to form a0 martensite is reduced. The purpose of
this research is to compare the microstructure and mechanical
properties of Ti-6Al-4V produced by the WAAM process in the as-
built, stress relieved, heat treated and HIPed conditions and iden-
tify the post build processing conditions that yield the most optimal
tensile performance.
2. Experimental method
The aim of this work is to compare the microstructures and
mechanical properties of Ti-6Al-4V components produced by Wire
þ Arc Additive Manufacturing that are subjected to different post-
build treatments. Details about the equipment and base experi-
mental set up for the WAAM process used here are available else-
where [21]. Ti-6Al-4V wire was used as the feedstock (initial
composition: 5.95 wt% aluminium, 4.02 wt% vanadium, 0.07 wt%
oxygen) and a wrought Ti-6Al-4V base plate was used as a substrate
for the deposits. A trailing shield supplying high purity argon was
used to protect the components as they cool which resulted in
contamination free surfaces (the components had a shiny, silver
surface appearance). Recent research has demonstrated negligible
oxygen pick-up from the atmosphere when using the same trailing
shield and deposition conditions for Ti-6Al-4V [5]. Each deposit
was created by moving the welding torch in a linear direction and
feeding wire into the molten pool, which subsequently solidifies to
make a layer. A subsequent layer was then deposited over the first
by increasing the height of the torch. The temperature at the end of
deposition2 was measured using a non-contact IR pyrometer, cali-
brated for emissivity against ultra-high purity titanium (99.995%
purity). Calibration was performed by melting a stationary pool of
the high purity titanium and experimentally determining the
emissivity against the measured thermal arrests occurring at
known phase transformation temperatures as it cooled (L/S at
1668  C and b/a at 882  C). The average emissivity from three
calibration tests was 0.2907 at the solidification temperature and
0.2909 at the b/a transformation temperature for the high purity
2
The temperature was measured at the end of the layer after the arc was
terminated. It was not possible to measure during deposition as radiation from the
arc interfered with the pyrometer, and the trailing shield prevented direct mea-
surement because it enclosed the cooling metal.
 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255 249

titanium. It was assumed that the emissivity would be comparable Table 2

between Ti-6Al-4V and pure titanium because both were well Chemical analysis of the bulk components (measured in the centre of component
away from the edge). O and N determined by LECO combustion and Al, V, Fe
shielded by inert gas and their surfaces were oxide free. After this, determined by ICP-AES.
deposition continued until the approximate dimensions of the final
build was approximately 180mm  12mm  12 mm (not measuring Sample O N Al V Fe

the base plate), which corresponded to four layers. The deposition wt% wt% wt% wt% wt%
parameters used for this research are identical to those reported Initial Wire feedstock 0.07 0.003 5.95 4.02 0.09
elsewhere [22]. As Built 0.07 0.009 5.90 3.68 0.09
Components were compared in the as-built condition, stress As Built þ Stress relief 0.08 0.007 5.73 3.63 0.11
Solution Treated þ Aged 0.11 0.017 5.90 3.75 0.09
relieved condition, heat-treated condition and HIPed condition. The
Vacuum Heat Treated 0.11 0.007 5.82 3.67 0.09
HIPing process is a heat treatment which simultaneously coarsens HIP 0.11 0.010 5.45 3.62 0.08
the microstructure through annealing while closing porosity. In
order to separate the effects of microstructure and porosity on the
resulting mechanical properties it was necessary to perform an
3. Results and discussion
additional test that replicated this heat treatment without closing
the porosity. This was achieved with a vacuum heat treatment
As expected the post process heat treatments had a large in-
under identical heating and cooling cycles as that used during the
fluence over the microstructure produced and corresponding ten-
HIP process. The slow cooling rate and long exposure time at high
sile properties. The deposition parameters used in this study
temperature during the HIP and vacuum heat treatments is ex-
produced wide components (approximately 12 mm) that contained
pected to coarsen the microstructure which may reduce strength
coarse prior-b grains often exceeding several millimetres in size
while increasing ductility. As such it was desired to explore a
(see Fig. 1C). Heat treatments did not change the morphology or
different heat treatment that could preference strength over
size of the prior-b grains but there was a significant change in the
ductility. Solution treatment and ageing (STA) is one such heat
size of the a-grain structure which are shown in Fig. 2. It is well
treatment used for Ti-6Al-4V components and involves firstly so-
known that a0 -martensite forms in Ti-6Al-4V when cooling rates
lution treating to develop a larger b-phase fraction followed by
exceed 410Ks1 during the b/a transformation [24]. In the present
quenching and subsequent ageing to decompose the unstable b and
study, the cooling rate during the solid state transformation is
promote higher strength. The STA heat treatment was selected as
much lower (between 10 and 20Ks1 as shown in Fig. 3) and as such
the final condition to evaluate the microstructure and mechanical €tten-a. In contrast, it
the as-built condition contains fine Widmansta
properties of Ti-6Al-4V produced by WAAM. A summary of the heat
has been frequently reported that the cooling rates are fast enough
treated conditions is shown in Table 1. The chemistry of the
during SLM to produce a0 -martensite (indicating cooling rates
component is presented in Table 2.
>410Ks1). This demonstrates one potential advantage of the
Tensile bars were wire cut from the components using modified
slower to cool wire arc AM process in that it is not necessary to
geometries of ASTM E8M (flat dogbone, 1.5 mm  5 mm gauge
perform a post-build heat treatment in order to remove the un-
cross-section x 16 mm gauge length) and tested in an Instron at a
desired martensite. However, the consequence of starting with a
crosshead speed of 0.5 mm/min. The orientation of the tensile bars
relatively coarse Widmansta €ten-a in the as-built state is that
was in the ‘longitudinal/horizontal build direction’, meaning that
additional annealing treatments such as HIPing will coarsen the a/b
the tensile axis was normal to the build direction (refer to Fig. 1).
grain size which for Ti-6Al-4V is largely controlled by the diffusion
This is frequently reported as the least ductile test orientation of Ti-
of vanadium through the b-phase [25]. A solution heat treatment
6Al-4V components produced by WAAM having between 20 and
on the other hand will result in a higher b-fraction which is then
60% lower ductility compared to the transverse/vertical testing
quenched to either retain a supersaturated b-phase or produce a0
orientation [4,18e20,23] and as such it was selected here to eval-
martensite. These metastable phases then decompose during sub-
uate a worst-case scenario ductility. A minimum of four tensile
sequent ageing and fine acicular a forms.
specimens were tested for each condition. Samples were also
The tensile properties for each test condition are presented in
removed from the component and prepared using conventional
Fig. 4 and averaged in Table 3. The tensile properties are in broad
techniques for optical and electron microscopy.
agreement with expectations given that the a-plate thickness and
The effectiveness of Hot Isostatic Pressing on removing porosity
colony size determines the effective slip length in lamellar micro-
was investigated by examining HIPed and non-HIPed components
structures, and therefore, higher strength is achieved under con-
with an Inveon multimodality PET-CT imaging scanner (Siemens)
ditions that promote faster cooling rates. The exception to this is the
(Micro-CT). The system can deliver high resolution CT images
as-built condition (no stress relief) which had a significantly lower
(<10 mm for a field-of-view <20 mm) with a maximum field-of-
tensile strength (~6e7%) and ductility (~33%) compared with the
view of 80 mm  50 mm (with a resolution of approximately
stress-relieved equivalent, despite both conditions having the same
50 mm). The voltage and current used in this study were 300 kV and
apparent a-phase size and morphology. Clearly, the presence of
80 mA, respectively. The cross sections of the sample were recorded
residual tensile stresses have a major deleterious effect in this
in three directions with step sizes of 0.001 mm.
instance. Although the magnitude of the residual stresses was not

Table 1
Summary of the different post process conditions investigated.

Condition Details

As-Built Residual stresses are maintained

As-Built þ Stress relief Stress relieved at 753 K for 2 h
Hot Isostatic Pressing HIPed at 1200 K for 2 h s dwell @ 1500Bar argon pressure with 5 Kmin1 heating and cooling rate
Vacuum Annealing 1200 K for 2 h s dwell with 5 Kmin1 heating and cooling rate (same as HIPed without pressure)
Solution Treatment þ Age Solution treated at 1240 K for 1 h, water quenched then aged at 868 K for 2 h and air cooled.
250 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255

Fig. 1. Geometry of the tensile bars machined from the components, (B) photograph showing the orientation of tensile bars with respect to build direction, (C), cross section view of
the macrostructure which reveals large prior-b grains in as-built þ stress relieved condition and (D), higher magnification SEM image of Widmanst€atten-a microstructure within
large prior-b grain.

measured, many others have modelled and measured the devel-
opment of residual stress during additive manufacturing of similar
single pass multi-layer components and report that substantial
tensile stresses develop in the longitudinal direction (i.e. the
orientation that the tensile bars were obtained from in this work,
refer to Fig. 1B). For instance, Wang et al. [26] modelled the residual
stress state in single pass 6 layer high aluminium components
produced by Wire Arc Additive Manufacturing (approximate di-
mensions 100 mm long, 10 mm wide and 10 mm high) and pre-
dicted significant tensile stresses that can exceed 75% of the
material's yield strength. A very similar stress profile was reported
by Cao, Charghouri and Nash [27] who modelled and experimen-
tally validated the residual stresses developed in single pass 6 layer
high Ti-6Al-4V components produced by electron beam AM and
measured tensile stresses exceeding 300 MPa in the longitudinal
direction. Martina et al. [28] also measured significant tensile
residual stresses (up to 500 MPa) during Wire Arc Additive
Manufacturing of Ti-6Al-4V. Such tensile stresses will have adverse
effects on the ductility of the material because it can assist in the
opening of cracks. Zhang et al. [29] performed fatigue crack prop-
agation tests on Ti-6Al-4V produced by Wire Arc Additive
Manufacturing and found that crack growth rates are strongly
influenced by residual stresses. Vrancken et al. [30] also reported
that the tensile residual stresses produced during SLM have sig-
nificant detrimental influences on the fatigue crack growth rates in
Ti-6Al-4V.
3.1. Effect of hot isostatic pressing
The benefits of HIPing titanium components produced during
powder bed additive manufacturing processes such as SLM has
been previously discussed by others. Namely, the process is greatly
M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255 251

beneficial in removing the defects associated with unmelted

powder particles and/or unconsolidated layers that can limit
ductility to under εf ¼ 1% [9]. However, the tensile testing here
indicated negligible difference between the HIP and Vacuum
annealing in terms of average strength or ductility in the Ti-6Al-4V
components produced by wire arc additive manufacturing. This
result is somewhat surprising given that HIPed components are
generally expected to have better ductility than non-HIPed
components.
This result could indicate that either porosity does not form
during WAAM (and therefore HIPing makes no difference), or
alternatively, all components contain porosity and the HIPing
process was not effective in removing this porosity. To investigate
further, high resolution X-ray Micro-CT was performed on small
sections3 of HIPed and non-HIPed components and a typical scan is
shown in Fig. 5. It is clear that the non-HIPed (Vacuum annealed)
component contains spherical porosity, with the largest pore
detected throughout the specimen approximately 350 mm in
diameter. However, it is important to note that large porosity such
as this is rare and most of the analysed sample either contains no
porosity or very small pores beyond resolution of the X-Ray
equipment. The spherical nature of the porosity indicates that it is
most likely entrapped gas porosity. It is unclear where this porosity
originates (i.e. argon gas or other) but the highly localised nature of
it could suggest that sources may include sections of unclean wire
feedstock or substrate. For example, Wang et al. [4] demonstrated
that even touching Ti-6Al-4V wire with a human hand prior to
WAAM drastically increased spherical gas porosity with a similar
size to the pores observed here. Whatever the source of this
localised porosity, it is clear that the HIPing process was very
effective in removing it as no porosity was detected post HIP
(Fig. 5b).
The localised nature of the large pores in non-HIPed (Vacuum
annealed) samples may also account for the observed variations in
tensile ductility. Although the vacuum annealed and HIPed speci-
mens had similar mean tensile ductility (εf ¼ 11.5% and εf ¼ 11.0%
respectively), the vacuum annealed specimens had more variation
than the HIPed specimens which would be expected if tensile bars
were machined from sections randomly containing larger or more
frequent porosity than other sections. It is also possible, given the
localised nature of the porosity detected by X-ray Micro-CT, that
some tensile specimens are machined from regions that are
porosity free, or at least, only contain small pores that do not in-
fluence tensile ductility. The size of porosity is very important and it
has been reported that small scattered pores in GTAW Ti-6Al-4V
welds do not effect tensile properties [31]. Although in the pre-
sent study the removal of large porosity through HIPing did not
measurably improve the mean tensile ductility, it did improve
consistency, which is very important from a certification
perspective.
Perhaps a more important consideration for removing porosity
through HIPing is improved fatigue properties. It is understood that
internal defects such as porosity considerably reduce fatigue life
given that it is the crack initiation process that consumes most of
the fatigue life. Therefore, the size and distribution of porosity will

Fig. 2. Typical microstructures produced for each condition: (A) as-built; (B) as-
built þ stress relieved; (C) Solution Treated þ aged; (D) vacuum anneal; (E) HIP. The
microstructure after all heat treatments is predominately Widmanst€atten/acicular-a,
however, the thickness of the a-laths (and b-regions) varies significantly.

3
The sections examined in Micro-CT were approximately 3.5 mm thick x 12 mm
wide x 14 mm tall.
252 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255

most influential on fatigue properties.

The tensile strength of the components subjected to vacuum
annealing and HIPing was less than the minimum strength speci-
fied in ASTM F1108 (860 MPa), however, the stress relieved as-built
and the Solution Treated and Aged components exceeded this
threshold. The microstructural factors that determine tensile
strength of Ti-6Al-4V are well known (a-colony size and a-lath size
in Widmansta €tten structures which determines ease of slip [7]) and
these are in turn influenced by the cooling rate during the b/a
transformation. The relatively slow cooling rate experienced during
WAAM in this work of 10e20 Ks1 during the transformation
(approximately 90e100Ks1 during solidification) produces
coarser grained Widmansta €tten-a compared to powder bed AM
processes which are known to have much higher cooling rates. In
fact the cooling rate experienced here is more than an order of
Fig. 3. Cooling curve of layer 3 during wire arc additive manufacturing. The cooling magnitude slower than some powder bed AM processes given that
rate prior to the b to a transformation is in the range of 10e20Ks1 which is well below a0 martensite is frequently reported in those cases [13,33] which
the critical cooling rate required to form a0 martensite. Consequently, the a-grains
necessitates cooling rates in excess of 410Ks1 during the b/a
adopt the diffusion controlled Widmanst€atten morphology.
transformation [24]. Additionally, the already coarse-grained a-
phase in the as-built WAAM condition continues to coarsen
through grain growth during additional vacuum annealing/HIPing
and this further reduces tensile strength. A similar trend was
observed in laser deposited Ti-6Al-4V components subjected to
heat treatment at 1473 K [34].
Another key microstructural parameter is the prior-b grain size
because this in turn limits the size of colony-a structures that can
form. When Ti-6Al-4V is cooled through the b-transus, clusters of
lamellar a belonging to the same crystallographic variant can form
colonies that grow in size until they are impeded by other a variants
growing within the parent b-grain or until impeded by the b grain
boundary itself. Thus large b-grains have the potential to foster
large a-colonies and since the colony size determines slip length, it
is desirable to avoid the formation of large b grains during solidi-
fication. Fig. 1 shows the cross-section macrostructure of a typical
component produced in this study (representative of all since all
were built under the same conditions). It is clear that the prior-b
grain size is quite large (exceeding several millimetres in some
Fig. 4. Results of tensile testing for each test condition. For reference, the results are cases) which results from the large thermal input of the WAAM
presented along with the minimum ductility and tensile strength requirements for Ti-
6Al-4V defined in ASTM F1108.
process, the slow cooling rate during solidification (90-100Ks1)
and potential grain growth during subsequent layer deposition.
Despite the large b-grain size it is found that large colonies are not
Table 3 prevalent and instead the majority of a-phase adopts a Wid-
Average mechanical properties for each test condition. Error bars indicate ±1 mansta €tten morphology (refer to Fig. 2 which shows a-clusters are
standard deviation.
only a few laths wide and of the order of a few micrometers). Using
0.2% sy (MPa) UTS (MPa) εf % techniques developed by Tilley et al. [35], the volume fraction of
As-Built þ Stress Relief 766 ± 11.9 872 ± 11.7 10.9 ± 2.46 colony-a in the WAAM built components was measured to be
As-Built (no Stress Relief) 710 ± 4.00 820 ± 6.23 7.18 ± 0.932 approximately 20% (the remainder considered Widmanst€ atten-a),
Solution Treated þ Age 858 ± 39.2 918 ± 30.4 5.85 ± 1.10 where for the purposes of measurement a colony was defined as a
Vacuum Heat Treatment 721 ± 16.4 810 ± 9.04 11.5 ± 4.15 cluster of 5 or more similarly oriented a-laths. In Widmansta €tten
Hot Isostatic Press 712 ± 11.5 800 ± 11.6 11.0 ± 1.68
structures that contain a mixture of misorientated a-laths, the slip
length can be approximated by the width of the a-lath (a-grain
size) between a-grains of different orientations as is the case for
be a critically important factor influencing the fatigue life of Ti-6Al- acicular martensitic a’ [36]. This high volume fraction of Wid-
4V WAAM components. Oh et al. [32] demonstrated this point by mansta €tten-a in the WAAM microstructure accounts for the
showing that large spherical porosity greater than 50 mm in diam- observed mechanical properties that comply with ASTM F1108 (as-
eter had a measurable negative effect on high-cycle fatigue life of built þ stress relieved) despite very large b-grain sizes.
Ti-6Al-4V welds but concluded that welds containing a distribution It is important to note that the processing parameters in the
of porosity smaller than 5 mm had no effect on fatigue properties. present study resulted in very large melt pools (greater than 10 mm
Similarly, Baufeld et al. [23] investigated the high-cycle fatigue diameter) as a consequence of high heat input (approximately
properties of Ti-6Al-4V WAAM components and found that 1.48 kJ/mm determined from experimental parameters) and this
porosity did not influence fatigue life except for in one instance may not be representative of the cooling rates experienced in other
when a large spherical pore approximately 170 mm in diameter WAAM processes using different parameters. For example, Wu et al.
reduced fatigue life. In Ti-6Al-4V components containing small [37] recently used WAAM to produce Ti-6Al-4V components with
sized pores, both Oh et al. and Baufeld et al. suggested that it was processing parameters that produced much lower heat input
microstructural factors, particularly the a-lamella size, that were (~0.15 kJ/mm) and this created smaller melt pools that cooled faster
 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255
Fig. 5. Typical MicroCT slice showing porosity in non-HIPed (vacuum annealed) specimen (A) and lack of porosity post-HIP (B). For each specimen, three images show repre-
sentative ‘slices’ in the top view, front view and side view orientations. No porosity was detected through any slice in the HIPed specimen, whereas random spherical porosity of
varying size was found throughout the entirety of the non-HIPed specimen.
and produced finer a-grain size. Nevertheless, in a general sense,
WAAM is prone to forming much coarser a-grain size and therefore
lower tensile strength than higher cooling rate AM processes such
as SLM. Fig. 6 shows the Hall-Petch relationship between a-grain
size and tensile strength for this work and prominent works from
the literature showing this trend. The a-grain size represents the
width of a-laths and this follows the Hall-Petch trend in Wid-
manst€ atten structures as this dictates the approximate slip length
and is proportional to cooling rate. Unlike high cooling rate AM
technologies such as SLM that sometimes require annealing to
remove a0 martensite and/or coarsen the alpha-grain size in order
to lower strength and improve ductility [13], in the present case for
WAAM it appears that such heat treatments may not be required in
order to achieve this balance. Furthermore, for WAAM with high
heat input such as the present study, it may even be necessary for
solution treatment plus ageing heat treatments (designed to refine
Fig. 6. Yield strength as a function of a-grain size for this work and a selection of
related works. Some of the work from the literature did not quantify a-grain size and
approximate values were measured from images provided. Heat treatments are
labelled and all other data points are “as-built”. Note that the laser build conditions in
the SLM study performed by Xu et al. [15] was controlled so that a0 martensite was not
formed and that the microstructures consisted of very fine lamellar aþb which allows
comparison to be made with WAAM that produces coarse lamellar aþb.
253
the a-grain size) to better optimise properties.
3.2. Fracture surfaces
Images of the fracture surfaces are shown in Fig. 7 for the two
individual test coupons with the extremes in ductility (i.e. the
vacuum annealed with 16.5% ductility and the Solution Treated and
Aged with 4.8% ductility). Both of these, as well as all other tests in-
between, show evidence of ductile fracture with dimples evident
that result from micro-void coalescence. The fracture surface of the
most ductile test confirms the presence of large pores (with the
largest pore diameter exceeding 200 mm). Given that the removal of
these pores by HIPing did not measurably improve ductility, it can
be concluded that other factors (i.e. microstructure) have had more
influence over tensile ductility than porosity.
Analogous with tensile strength, the ductility of the components
was most sensitive to the scale of the a-grain size. The size of the
dimples on the fracture surfaces were found to correlate with the
grain size (z10 mm in vacuum heat treated vs z 3 mm in solution
treated and aged, Fig. 7). This is consistent with the findings of
others and indicates that the smaller grain size provided an
increased density of nucleation sites to initiate microvoid coales-
cence [38,39].
4. Conclusions
This work investigated the effect of post build heat treatments
and Hot Isostatic Pressing on the strength and ductility of Ti-6Al-4V
produced by Wire Arc Additive Manufacturing (WAAM). The
different thermal processes were found to directly influence the
mechanical properties through a Hall-Petch relationship with the
a-grain size. The conclusions from the study are listed below:
 WAAM is prone to the entrapment of spherical gas porosity
which ranges in size but can exceed 350 mm in diameter (di-
mensions measured by X-ray MicroCT). Hot Isostatic Pressing is
very effective in removing this porosity, however, this gave no
measureable improvement in strength or ductility.
254 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255
Fig. 7. SEM of fracture surfaces for A) most ductile sample of all tests (Vacuum heat treated εf ¼ 16.5%) and least ductile of all tests (Solution Treated þ Aged, εf ¼ 4.8%). All fracture
surfaces show ductile dimple fracture and some contain evidence of porosity as present in A (indicated by arrows).
 The processing parameters during WAAM (current, deposition
speed etc.) are particularly influential on the grain size and
mechanical properties. Unlike other higher cooling rate AM
processes that generally require post-build annealing to remove
martensite and achieve acceptable ductility, the relatively slow
cooling rate of 10e20Ks1 during the b/a transformation in
WAAM as-built components produced a grain size that offers a
good balance between strength and ductility that exceeds the
requirements outlined in ASTM F1108 (after stress relief). High
temperature annealing (1200 K for 2 h) during vacuum heat
treatment or HIPing coarsened the microstructure which
reduced strength by around 6e7% compared to the as-built
component while offering little to no improvement in
ductility. Solution Treatment at 1240 K followed by quenching
and ageing at 686 K for 2 h refined the as-built grain size which
increased strength by around 12% but reduced ductility by more
than 30%.
 Residual stresses substantially impair the tensile properties of
as-built components. A short post-build stress relief at 653 K for
2 h improved the ductility by over 30% while not coarsening the
grain size. This condition fulfilled the strength and ductility
requirements specified in ASTM F1108.
Acknowledgements
The authors would like to acknowledge to support of the School
of Mechanical and Mining Engineering and the Queensland Centre
for Advanced Materials Processing and Manufacturing (AMPAM).
MJB acknowledges the support of the Australian Research Council
Discovery Program and is in receipt of Discover Early Career
Researcher Award (DE160100260). MJB, DK and MSD also
acknowledge the support of the Australian Research Council
Research Hub for Advanced Manufacturing of Medical Devices
(IH150100024) and the Australian Research Council Research Hub
to Transform Transforming Australia's Manufacturing Industry
through High Value Additive Manufacturing (IH130100008). All
authors acknowledge the support of Dr Tom Jarvis and the Monash
Centre for Additive Manufacturing for assistance with Hot Isostatic
Pressing.
References
[1] D. Ding, Z. Pan, D. Cuiuri, H. Li, Wire-feed additive manufacturing of metal
components: technologies, developments and future interests, Int. J. Adv.
Manuf. Technol. 81 (2015) 465e481.
[2] W.E. Frazier, Metal additive manufacturing: a review, J. Mater. Eng. Perform.
23 (2014) 1917e1928.
[3] D.D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Laser additive manufacturing
of metallic components: Materials, processes and mechanisms, Int. Mater.
Rev. 57 (2012) 133e164.
[4] F. Wang, S. Williams, P. Colegrove, A.A. Antonysamy, Microstructure and
mechanical properties of wire and arc additive manufactured Ti-6Al-4V,
Metall. Mater. Trans. 44 (2013) 968e977.
[5] M.J. Bermingham, J. Thomson-Larkins, D.H. St John, M.S. Dargusch, Sensitivity
of Ti-6Al-4V components to oxidation during out of chamber wire þ arc ad-
ditive manufacturing, J. Mater. Process. Technol. 258 (2018) 29e37.
[6] S. Bontha, N.W. Klingbeil, P.A. Kobryn, H.L. Fraser, Effects of process variables
and size-scale on solidification microstructure in beam-based fabrication of
bulky 3D structures, Mater. Sci. Eng. A 513e514 (2009) 311e318.
[7] G. Lütjering, J.C. Williams, Titanium, Springer, New York, 2003.
[8] A.A. Antonysamy, J. Meyer, P.B. Prangnell, Effect of build geometry on the b-
grain structure and texture in additive manufacture of Ti6Al4V by selective
electron beam melting, Mater. Char. 84 (2013) 153e168.
[9] P. Kobryn, S. Semiatin, Mechanical properties of laser-deposited Ti-6Al-4V, in:
Solid Freeform Fabrication Proceedings, Austin, 2001, pp. 179e186.
[10] T. Vilaro, C. Colin, J.D. Bartout, As-Fabricated and heat-treated microstructures
of the Ti-6Al-4V alloy processed by selective laser melting, Metall. Mater.
Trans. 42 (2011) 3190e3199.
[11] B.E. Carroll, T.A. Palmer, A.M. Beese, Anisotropic tensile behavior of Ti-6Al-4V
components fabricated with directed energy deposition additive
manufacturing, Acta Mater. 87 (2015) 309e320.
[12] B. Vrancken, L. Thijs, J.P. Kruth, J. Van Humbeeck, Heat treatment of Ti6Al4V
produced by selective laser melting: microstructure and mechanical proper-
ties, J. Alloys Compd. 541 (2012) 177e185.
[13] L. Facchini, E. Magalini, P. Robotti, A. Molinari, S. Ho € ges, K. Wissenbach,
Ductility of a Ti-6Al-4V alloy produced by selective laser melting of prealloyed
powders, Rapid Prototyp. J. 16 (2010) 450e459.
[14] S. Leuders, M. Tho € ne, A. Riemer, T. Niendorf, T. Tro € ster, H.A. Richard,
H.J. Maier, On the mechanical behaviour of titanium alloy TiAl6V4
 M.J. Bermingham et al. / Journal of Alloys and Compounds 753 (2018) 247e255
manufactured by selective laser melting: fatigue resistance and crack growth
performance, Int. J. Fatig. 48 (2013) 300e307.
[15] W. Xu, E.W. Lui, A. Pateras, M. Qian, M. Brandt, In situ tailoring microstructure
in additively manufactured Ti-6Al-4V for superior mechanical performance,
Acta Mater. 125 (2017) 390e400.
[16] G. Kasperovich, J. Hausmann, Improvement of fatigue resistance and ductility
of TiAl6V4 processed by selective laser melting, J. Mater. Process. Technol. 220
(2015) 202e214.
[17] A.K. Swarnakar, O. Van Der Biest, B. Baufeld, Young's modulus and damping in
dependence on temperature of Ti-6Al-4V components fabricated by shaped
metal deposition, J. Mater. Sci. 46 (2011) 3802e3811.
[18] B. Baufeld, O. Van der Biest, Mechanical properties of Ti-6Al-4V specimens
produced by shaped metal deposition, Sci. Technol. Adv. Mater. 10 (2009)
015008.
[19] B. Baufeld, O.V.D. Biest, R. Gault, Additive manufacturing of Ti-6Al-4V com-
ponents by shaped metal deposition: microstructure and mechanical prop-
erties, Mater. Des. 31 (2010) S106eS111.
[20] E. Brandl, B. Baufeld, C. Leyens, R. Gault, Additive manufactured Ti-6Al-4V
using welding wire: comparison of laser and arc beam deposition and eval-
uation with respect to aerospace material specifications, Phys. Procedia
(2010) 595e606.
[21] M.J. Bermingham, D. Kent, H. Zhan, D.H. St John, M.S. Dargusch, Controlling
the microstructure and properties of wire arc additive manufactured Ti-6Al-
4V with trace boron additions, Acta Mater. 91 (2015) 289e303.
[22] M.J. Bermingham, S.D. McDonald, M.S. Dargusch, Effect of trace lanthanum
hexaboride and boron additions on microstructure, tensile properties and
anisotropy of Ti-6Al-4V produced by additive manufacturing, Mater. Sci. Eng.
A 719 (2018) 1e11.
[23] B. Baufeld, E. Brandl, O. van der Biest, Wire based additive layer
manufacturing: comparison of microstructure and mechanical properties of
Tie6Ale4V components fabricated by laser-beam deposition and shaped
metal deposition, J. Mater. Process. Technol. 211 (2011) 1146e1158.
[24] T. Ahmed, H.J. Rack, Phase transformations during cooling in aþb titanium
alloys, Mater. Sci. Eng. A 243 (1998) 206e211.
[25] S.L. Semiatin, B.C. Kirby, G.A. Salishchev, Coarsening behavior of an alpha-beta
titanium alloy, Metall. Mater. Trans. 35 (2004) 2809e2819.
[26] X. Wang, A. Wang, Three-dimensional finite element analysis with clamping
in wire and arc additive manufacturing, in: Modelling Symposium (EMS),
2016, European, IEEE, 2016, pp. 104e108.
255
[27] J. Cao, M.A. Gharghouri, P. Nash, Finite-element analysis and experimental
validation of thermal residual stress and distortion in electron beam additive
manufactured Ti-6Al-4V build plates, J. Mater. Process. Technol. 237 (2016)
409e419.
[28] F. Martina, M.J. Roy, B.A. Szost, S. Terzi, P.A. Colegrove, S.W. Williams,
P.J. Withers, J. Meyer, M. Hofmann, Residual stress of as-deposited and rolled
wireþarc additive manufacturing Tie6Ale4V components, Mater. Sci. Tech-
nol. 32 (2016) 1439e1448.
[29] J. Zhang, X. Wang, S. Paddea, X. Zhang, Fatigue crack propagation behaviour in
wireþarc additive manufactured Ti-6Al-4V: effects of microstructure and
residual stress, Mater. Des. 90 (2016) 551e561.
[30] T. Lyman, E. Boyer, J. Carnes, P. Unterweiser (Eds.), Atlas of Microstructures of
Industrial Alloys, Metals Handbook American Society for Metals, Metals Park,
OH, 1972.
[31] A.B. Short, Gas tungsten arc welding of a þ b titanium alloys: a review, Mater.
Sci. Technol. 25 (2009) 309e324.
[32] J. Oh, N.J. Kim, S. Lee, E.W. Lee, Correlation of fatigue properties and micro-
structure in investment cast Ti-6Al-4V welds, Mater. Sci. Eng. A 340 (2003)
232e242.
[33] S. Al-Bermani, M. Blackmore, W. Zhang, I. Todd, The origin of microstructural
diversity, texture, and mechanical properties in electron beam melted Ti-6Al-
4V, Metall. Mater. Trans. 41 (2010) 3422e3434.
[34] E. Brandl, F. Palm, V. Michailov, B. Viehweger, C. Leyens, Mechanical proper-
ties of additive manufactured titanium (Tie6Ale4V) blocks deposited by a
solid-state laser and wire, Mater. Des. 32 (2011) 4665e4675.
[35] J. Tiley, T. Searles, E. Lee, S. Kar, R. Banerjee, J.C. Russ, H.L. Fraser, Quantifi-
cation of microstructural features in a/b titanium alloys, Mater. Sci. Eng. A 372
(2004) 191e198.
[36] G. Luetjering, Influence of processing on microstructure and mechanical
properties of (aþb) titanium alloys, Mater. Sci. Eng. 243 (1998) 32e45.
[37] Q. Wu, Z. Ma, G. Chen, C. Liu, D. Ma, S. Ma, Obtaining fine microstructure and
unsupported overhangs by low heat input pulse arc additive manufacturing,
J. Manuf. Process. 27 (2017) 198e206.
[38] Y.G. Ko, D.H. Shin, K.-T. Park, C.S. Lee, An analysis of the strain hardening
behavior of ultra-fine grain pure titanium, Scr. Mater. 54 (2006) 1785e1789.
[39] M.-H. Cai, C.-Y. Lee, Y.-K. Lee, Effect of grain size on tensile properties of fine-
grained metastable b titanium alloys fabricated by stress-induced martensite
and its reverse transformations, Scr. Mater. 66 (2012) 606e609.