Optics and Laser Technology xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Optics and Laser Technology

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

Full length article

Effect of laser welding process on the microstructure, mechanical

properties and residual stresses in Ti-5Al-2.5Sn alloy
Massab Junaid a,⇑, Fahd Nawaz Khan b, Khalid Rahman a, Mirza Nadeem Baig c
a
Faculty of Mechanical Engineering, Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi 23640, Pakistan
b
Faculty of Materials and Chemical Engineering, Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi 23640, Pakistan
c
Pakistan Welding Institute, PO Box 1781, Islamabad, Pakistan

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

Article history: Pulsed Nd-YAG laser was employed in bead on plate configuration for welding of 1.6 mm thick Ti-5Al-
Received 28 April 2017 2.5Sn alloy sheet. The effect of laser processing parameters on the weld pool shape, pulse overlap, oxide
Received in revised form 18 June 2017 formation, and microstructure were studied using scanning electron and optical microscope. It was found
Accepted 7 July 2017
out that laser peak power had a significant influence on the FZ oxygen contents and grain size whereas,
Available online xxxx
both peak power and heat input per unit length were important in defining the weld pool shape.
Processing parameters for full penetration welds with acceptable joint properties and low oxygen con-
Keywords:
tents were selected for further study in terms of residual stresses and mechanical properties.
Titanium alloys
Laser welding
Formation of acicular a and a0 martensite in fusion and heat affected zone, led to an increase in micro-
Residual stresses hardness by about 55 HV0.2 as compared to base metal. Induced residual stresses were found to be sig-
Microstructure nificantly less than the yield strength resulting in plate deformation less than 1 mm. Transverse residual
Fracture behavior stresses present at different depths below the surface tend to counter effect each other resulting in tensile
Tensile properties strength of welded specimen becoming nearly equal to that of the base metal.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction equipment such as liquid hydrogen tanks and high pressure ves-
sels, operating at such low temperatures [8–10].
Titanium alloys are mostly used in complex engineering struc- Nd-YAG laser beam welding (LBW) is efficient and controllable
tures, especially in aerospace industry owing to their better as the desirable welding mode can be obtained by an optimum
strength to weight ratios, remarkable resistance to corrosion, wear parametric combination of pulse energy, pulse waveform, pulse
and high temperature creep phenomenon [1–3]. The current focus duration, pulse frequency and welding speed. In addition, the pul-
of aerospace industry is to replace traditional riveting process with sating effect leads to intermittent cooling of welds resulting in
welding which would not only save extra weight of thousands of grain refinement [11,12]. Squillace et al. [13] showed that for
rivets but also additional overlapping of sheets could be avoided. LBW joints of Ti-6Al-4V, the underfill depth has a strong depen-
This will significantly reduce the overall weight of airframe struc- dence on weld heat input. The weld morphology and microstruc-
ture [4]. Therefore, for manufacturing of airframe components, the tural characteristics of laser weld beads are strongly dependent
selected titanium alloy should possess good weldability. Ti-5Al- on weld speed and have a major influence on resultant mechanical
2.5Sn is an a titanium alloy composed of low cost alloying ele- properties [14,15]. Cao and Jahazi [14] found the tensile strength of
ments as compared to Ti-6Al-4V. Ti-5Al-2.5Sn alloy has excellent LBW Ti-6Al-4V joints to be similar to that of the base metal, with a
weldability and is mostly used in airframe applications, especially loss in ductility. Gao et al. [16] showed that for Nd-YAG LBW of Ti-
on the aerospace structural component near engines and leading 6Al-4V, large stresses were accumulated at the a/b interface during
edge of the wings where it can sustain temperatures up to 480 °C tensile testing and voids initiated at the a/b boundaries. Further-
[5,6]. This alloy also possesses good ductility and fracture tough- more, it was reported that a heterogeneity in strain distribution
ness at extremely low temperatures (252 °C) where the well- at the HAZ/ base metal interface reduced strength of the welded
known Ti-6Al-4V alloy loses its ductility [7]. In the past, this alloy joint [17].
has been extensively used in annealed conditions for cryogenic The mechanical properties of titanium welds are more compli-
cated than other alloys and are not only governed by the
⇑ Corresponding author. microstructure, but are also strongly dependent on the microele-
E-mail address: [email protected] (M. Junaid). ment content and strengthening mechanisms [18]. The strengthen-

http://dx.doi.org/10.1016/j.optlastec.2017.07.010
0030-3992/Ó 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
2 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

Nomenclature

Nd-YAG neodymium-doped yttrium aluminum garnet SEM scanning electron microscopy

FZ fusion zone PO pulse overlap
HAZ heat affected zone XRD X-ray diffraction
Cp Ti commercially pure titanium UTS ultimate tensile strength
OM optical microscopy YS yield strength

ing mechanism observed in welding of a and a + b titanium alloys Laser welding induces microstructural changes and the post
is different from that of b alloys [19]. For a and a + b titanium weld air quenching generates residual stresses; both of which are
alloys, the fusion zone (FZ) in welded joints undergoes fast cooling important from the aspect of mechanical properties of the welded
leading to formation of a0 martensite phase which acts as a structure. It has been observed that very limited literature is avail-
strengthening precipitate since it has a higher hardness value able on the study of welding Ti-5Al-2.5Sn alloy, which is an alpha
[13,20]. However in b titanium alloys, the formation of strengthen- titanium alloy. The available studies are mostly focused on the
ing precipitates like a0 is suppressed because of high concentration welding of the well-known alpha-beta alloy, Ti-6Al-4V. The work
of b stabilizers [21], resulting in reduced hardness of the FZ [19]. of Akman et al. [25], Cao and Jahazi [14] and Squillace et al. [13]
Furthermore, the microhardness profile over the width of FZ, was focused on investigating the effect of pulsed laser welding
HAZ and base metal is important because it gives an estimated parameters on the weld pool shape and properties of Ti-6Al-4V
strength of the weldment [20]. Chamnafar et al. [22] reported the alloy. However, neither of them studied the residual stress distri-
formation of a0 martensite due to rapid solidification of molten bution and associated deformations due to the welding phe-
pool in LBW of Ti-6242 alloy. This led to a significant increase in nomenon. The study of residual stresses and deformations is
hardness of the FZ with joint strength approximately equal to par- important especially for thin sheets as they have low stiffness
ent metal. Baruah and Bag [23] showed that for thin sheet of Ti- and distort easily under the influence of residual stresses. Hence,
6Al-4V alloy, optimum combination of microplasma arc weld it becomes difficult to achieve the desired shape of the welded
parameters could lead to production of high quality joints, with component. Moreover, Akman et al. [25] performed the parametric
strength superior to that of base metal. In another study by the investigation at a fixed welding speed of 5 mm/s whereas, Squil-
same author, it was reported that an increase of heat input in micro lace et al. [13] only investigated the effect of heat input per unit
plasma arc welding of Ti-6Al-4V led to growth of lamellar a in the length by varying the welding speed and laser power at a fixed
FZ and resulted in a decrease of tensile strength [24]. Furthermore, pulsed width (viz. constant peak power). Similarly in the work of
the penetration depth and geometry of the pulsed LBW weldments Cao and Jahazi [14], only the influence of welding speed at a fixed
can be controlled in a precise manner by controlling the pulse laser power was studied. Furthermore, in all these studies, the
duration and laser beam energy [25]. Akman et al. [25] observed effect of welding parameters on the pulse overlap and oxygen con-
in LBW of Ti-6Al-4V alloy weldments that the penetration depth tents was also not studied. The aim of the present work is to inves-
is strongly dependent on the peak power which is governed by tigate the effect of both peak power and heat input per unit length
the laser pulse-on duration. Moreover, they also concluded that by varying the welding speed, average power and pulse width of
the microhardness profile was strongly dependent on the extent Nd-YAG laser welding on the pulse overlap weld pool shape,
of martensite formation, which was directly related to the cooling microstructure and oxygen contents in the weld zone of Ti-5Al-
rate. An increase in peak power, by reducing the pulse-on time at 2.5Sn thin sheet. The weld showing acceptable joint performance
the same average power, increased the cooling rate in the FZ which in terms of weld pool shape, FZ oxidation, pulse overlap and
led to an increased microhardnes. Baruah and Bag [26] also studied microstructure was selected for further analysis in terms of micro-
variation of weld pool dimensions with variation in pulse energy hardness, mechanical properties and residual stresses.
for laser pulsed microwelding of 0.5 mm thin Ti-6Al-4V sheet. They
reported a change of weld pool shape from trapezoidal to hourglass 2. Material and methods
as the heat input increased from 53.3 J/mm to 63.3 J/mm.
During welding of titanium alloys, the use of shielding gas pro- As-annealed sheets of Ti-5Al-2.5Sn alloy (ASTM B265 grade 6)
tects the metal surface from reacting with atmospheric oxygen as with a sample size of 100 mm  80 mm  1.6 mm provided by
the temperatures are high [27]. Improper shielding of titanium Continental Steel and Tube Company USA, were used. The nominal
alloys during welding may lead to porosity and embrittlement of composition of the alloy based on wt.% is shown in Table 1. Sam-
the welded joint [26]. For efficient protection, the flow of shielding ples were welded in Bead on Plate configuration using Nd-YAG
gas is maintained even after the completion of welding operation LBW without clamping. Argon was used as a shielding gas to avoid
which results in fast cooling of the weld metal until an ambient the contact of atmospheric oxygen with the alloy surface.
temperature is reached [28,29]. Hence, the welded structure expe- The review of literature suggests that average power, speed and
riences expansion and contraction during the heating and cooling pulse width are the critical parameters having major influence on
thermal cycles. This eventually leads to non-uniform plastic defor- the peak power and input heat energy per unit length of the mol-
mations and induced residual stress [30]. During welding of thin ten weld pool [25,26,36]. Therefore welding was performed at a
walled components, the generation of residual stresses makes the range of average power, welding speed and pulse duration. In
control of final part dimensions difficult and unpredictable [31]. pulsed form of welding, the weld bead is comprised of a series of
These residual stresses may also cause a reduction in buckling overlapping spots. Pulse overlap was calculated by measuring the
strength, fatigue damage, and fracture strength associated with dimensions of overlapping spots under optical microscope and
stress corrosion cracking [32]. The accurate measurement of resid- using Eq. (1) [37]
ual stress distribution is of prime importance as these induced
residual stresses affect the tensile properties of the welded struc- ½S  S0 
% Overlap ¼  100% ð1Þ
ture [31–35]. S

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 3

Table 1
Chemical composition of Ti-5Al-2.5Sn (wt%).

Ti Al Sn V Mo Fe C N Si O
Bal. 5.34 2.56 0.02 <0.01 0.27 0.01 0.01 0.03 0.17

where S is the major diameter of the weld spot and S0 is the length analyzed using Energy dispersive X-ray (EDX) setup attached to the
of weld spot not overlapped by successive welding spot. Heat input SEM. For each of the weld bead, three different points were ana-
is dependent on average power (Pav ) and welding speed (U) by Eq. lyzed using EDX and the average values were used for further anal-
(2). ysis. Microhardness distribution across the weld zone was
measured on the metallographic samples by Tukon Model 300
P av
H¼ ð2Þ Vickers hardness machine with a weight of 200 gf and dwell time
U of 10 s. FZ, HAZ and base metal were sectioned from the metallo-
Moreover, the average laser power (P av ), pulse frequency (f), graphic samples for XRD analysis. Philips PW 3710 mpd control
peak power (P peak ), and pulse duration (t) are related by Eq. (3) X-ray diffractometer was used for the identification of different
phases developed in the welded zone during the joining process.
P av Cu Ka radiations (k = 1.54 Å) at 40 kV and 25 mA and 2h angle
Ppeak ¼ ð3Þ
ft was varied from 20° to 80° with a step size of 0.05°.
For all the experiments, pulse frequency was fixed at 8 Hz and As the plate was not clamped during welding, a deformation
the standoff distance was maintained at 4 mm and 6 welded sam- and distortion of plate geometry was observed which was recorded
ples were chosen corresponding to different parametric combina- using a dial gauge, with a least count of 0.01 mm. Hole-drill strain
tions as given in Table 2. The influence of weld parameters on measurement method (setup shown in Fig. 1a) was employed to
the physical appearance, weld zone size, pulse overlap and oxide measure through thickness residual stresses at various distances
formation in the FZ was studied using various characterizing tech- from the weld centerline. Integral method was used to perform
niques. The weldment showing satisfactory physical appearance strain to stress transformation, in which the strain measured at a
(absence of excessive material spattering, underfill, and burnout certain depth is integral of the strain components measured at
of the weld bead) with complete penetration and acceptable level all the depth increments below it. This relationship is described
of oxidation was selected for further study of residual stresses and as follows
tensile testing. Z
Specimens for metallography and tensile tests were segmented 1þm h

ðhÞ ¼ AðH; hÞrðHÞdH; 06H6h ð4Þ
from the welded plate in transverse direction using wire electric E 0
discharge machining (EDM). ASTM standard E8M-04 was followed
to prepare the specimens for tensile testing with a gauge length of where AðH; hÞ is the normalized strain relaxation [38]. According to
25 mm. The tensile testing was performed using 30 kN Instron Rossini et al. [39], Hole-drill strain measurement is considered to be
machine with a strain rate of 2.5 mm/min. In order to examine a reliable and accurate technique for measuring residual stresses.
the microstructure of welded section, metallographic samples The drilling process in hole-drill strain measurement may lead to
were ground using water abrasive paper up to 4000 grit, mechan- development of some plastic strains near the edge, resulting in an
ically polished using 1 lm diamond suspension paste. Further- additional residual stress. This additional stress was taken as
more, these samples were etched first with Kroll solution (6% 20 MPa and was calculated by applying hole-drill strain measure-
HNO3 and 2% HF by volume in distilled water) followed by further ment in the plate before welding. Table 3 shows the detailed
etching with 2% HF. The samples were studied under Olympus description of strain gauge rosettes used which were manufactured
BH2-UMA optical microscope up to 500, using polarized light by Vishay Measurement Group, Raleigh, NC, USA. As shown in
and sensitive tint filter. Mira 3 Tescan PC controlled scanning elec- Fig. 1b, ‘‘Type B” rosette was used near the weld centerline whereas
tron microscope (SEM), at a vacum of 106 Pa was used for higher ‘‘Type A” rosette was used at locations away from the weld center-
magnification images. SEM analyses were performed on the metal- line. The standard procedure for calculations of ASTM E837 was
lographic samples in back scattered electron (BSE) and in-beam adopted. Residual stresses which were relieved during the drilling
mode whereas secondary electron (SE) along with in-beam mode process, were recorded using Strain smart software and Model
were used for fractured surface analysis of tensile specimens. In SCXI-1520 multi-channel data acquisition system made by National
order to measure the elemental distribution especially for oxygen Instruments. H-Drill software was used for data reduction purposes
(weight percent), each of the weldment and the parent metal were which is based on ASTM Standard E-837.

Table 2
Welding parameters for LBW.

Average power (W) Welding speed (mm/min) Pulse duration (ms) Peak power (kW) Heat input (J/mm)
Weldment 1 297 160 8 4.6 111.4
Weldment 2 228.4 160 10 2.9 85.7
Weldment 3 297 160 3 12.4 111.4
Weldment 4 320 180 6 6.7 106.7
Weldment 5 297 200 6 6.2 89.1
Weldment 6 320 200 8 5.0 96.0
Weldment 7 240 160 8 3.8 90.0
Weldment 8 171.3 180 6 3.6 57.1
Weldment 9 114.2 180 6 2.4 38.1
Weldment 10 91.4 160 8 1.4 34.3
Weldment 11 297 140 4 9.3 127.3

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
4 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

Fig. 1. (a) Residual stress measuring setup and (b) locations of strain gauges on the welded plate.

Table 3
Strain gauge rosette specifications.

Strain guage rosette pattern Dimensions (mm) Resistance (O) Rosette type
Typical Hole Matrix
diameter
Min Max Length Width
1.5 2 9.6 12.2 120 ± 0.4 B

CEA-XX-062UM-120

10.7 10.7 A

EA-XX-062RE-120

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 5

3. Results and discussion value of pulse frequency, increasing the welding speed leads to a
decrease in pulse overlap [26,36]. In the current work, pulse over-
The microstructure of as-received Ti-5Al-2.5Sn alloy is shown in lap was found to be approx. 91%, 85%, 79% and 70% for welds per-
Fig. 2. It can be observed that equiaxed a grains are embedded in formed at a speed of 140 mm/min, 160 mm/min 180 mm/min and
the prior beta matrix. The equiaxed grains have an average size 200 mm/min respectively.
of about 9.3 lm as measured using ImageJ software which is based A comparison of weldment 1 and weldment 3 in Fig. 3 shows
on the method of linear interception. Similar microstructure has the effect of varying the pulse width under constant average power
also been reported by Tan et al. [8] for Ti-5Al-2.5Sn sheet with and welding speed. A reduction of pulse width from 8 ms to 3 ms
thickness of 1.5 mm. in weldment 3, increased the peak power by approximately 2.6
Table 2 shows the welding conditions corresponding to 11 times. This increased peak power led to overheating of the weld
welded samples which were chosen for analysis. The appearance bead and excessive melting as evident by the bead appearance in
of weld surfaces for different welding conditions is shown in Fig. 3c. Furthermore, this excessive melting of the weld pool led
Fig. 3. Weldment 1, 2 and 3 in both these figures were performed to material expulsion and spattering at the top surface and an
at a welding speed of 160 mm/min, with different value of average undercut at the bottom as shown in Fig. 4a. Squillace et al. [13]
power and pulse width. Weldment 4, 5, and 6 were performed at a reported that increase of power density of LBW above a certain
relatively high welding speed which led to a decrease in pulse level increases the vapor pressure above the keyhole and results
overlap as shown in Fig. 3. It has been reported that for a fixed in flow of molten metal outside the pool or can open a channel
at the weld bottom.
It is well established that the weld pool width and penetration
is not only dependent on the heat input but also on the pulse
energy and peak power [25,26]. Fig. 4 presents weld macrographs
showing a variation in weld pool shape and penetration due to
change in heat input per unit length and peak power. At a fixed
value of welding speed and pulse width, the weld penetration
increases with increase of heat input and reaches to full depth at
a heat input per unit length of approximately 90–95 J/mm as
shown in Fig. 5a. Moreover, the top width of the weld pool also
increases consistently with an increase of heat input per unit
length until the weld pool reaches full depth (Fig. 5a). Fig. 5b
shows the effect of increase in peak power on the top and bottom
width of full penetration weldments. It can be seen that at a fixed
value of pulse energy (37.1 J), an increase of peak power from
4.6 kW to 9.3 kW resulted in an increase of bottom width of the
weld pool by approximately 166% however, the top width of the
weld pool remained nearly constant. This is attributed to the
increase of heat transfer in thickness direction once the molten
pool reaches full depth. It was also interesting to note that further
Fig. 2. Microstructure of the as received Ti-5Al-2.5Sn alloy. increase in peak power from 9.3 to 12.4 kW resulted in a decrease

Fig. 3. Weld bead appearance at different welding conditions (a) weldment 1, (b) weldment 2, (c) weldment 3, (d) weldment 4, (e) weldment 5 and (f) weldment 6.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
6 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

Fig. 4. Weld pool shape at different conditions of heat input per unit length and peak power (a) 111.4 J/mm, 12.4 kW, (b) 127 J/mm, 9.3 kW and (c) 57.1 J/mm, 3.6 kW (Weld
pool boundary is shown by the yellow arrows). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5. Weld pool dimensions at different conditions of heat input per unit length and peak power.

of bottom width of weld pool by approximately 31% (Fig. 5b). This minimal contamination and adequate shielding as opposed to blue,
is due to the fact that weldment 3 (peak power of 12.4 kW) had a purple and grey appearance. In another work, blue colored bead
lower heat input per unit length of 111 J/mm as compared to 127 J/ appearance in LBW Cp Ti was reported to contain highest oxygen
mm for weldment 11 (peak power of 9.3 kW) as shown in Fig. 4. contents while silver color showed lowest concentration of oxides
During the solidification process, growth of prior b grains occurs [42]. Although the weld bead color indicates the level of contami-
in the direction of heat flow [40]. For a higher welding speed of 200 nation due to poor shielding of the weld pool [27], however, EDX
mm/min, corresponding to minimum bead overlap (compare analysis will provide a much better estimate of FZ oxidation. In
weldment 6 and weldment 3 in Fig. 3), the prior b grains appeared the current study, although the flow of shielding gas was
to be growing radially from a central point as shown in Fig. 6a. employed, however significant oxidation was observed in some
However, as shown in Fig. 6b, the welding speed was relatively less of the welded samples as compared to parent metal. Fig. 7a shows
(160 mm/min) and the direction of grain growth was as shown by that the oxygen concentration was about 3.7% in the parent metal.
the arrows. This trend was observed owing to the combined effect Among the welded samples, oxygen concentration was lowest
of heat flow towards the base metal region and also towards the (5.3%) in FZ of weldment 10 with a minimum heat input per unit
trailing edge of the heat source. The influence of welding speed length and peak power of 34.3 J/mm and 1.4 kW respectively. As
on grain growth was also studied by Liu et al. [40] for laser welding shown in Fig. 7c, an increase of heat input to 111.4 J/mm at a fixed
of Cp Ti. He reported a transition from heterogeneous to an epitax- welding speed and pulse width resulted in a little variation of oxy-
ial grain growth perpendicular to the weld centerline by increasing gen contents. This was due the fact that there was no significant
the weld speed from 2 m/min to 8 m/min leading to a reduction of variation in peak power of selected weldments and it was in the
misorientation angles. However they did not study the effect on range 1.4–4.6 kW. However, an increase of peak power to
grain orientation when the welding speed was further increased 12.4 kW resulted in a significant increase of oxygen contents to
beyond 8 m/min. This behavior was also observed to some extent approximately 25% as shown in Fig. 7d, signifying the dominant
in the present study, where increasing the welding speed from influence of peak power on the level of oxide formation during
160 mm/min to 200 mm/min resulted in an increase of misorien- the welding process. This is due to the fact that oxygen concentra-
tation angle. tion in FZ is dependent on the peak temperatures attained during
Fig. 7 shows the result of EDX analysis to measure the concen- the welding process. An increase in welding temperatures by
tration of oxygen in the FZ of welded samples and the parent increasing the peak power levels will lead to increase in oxide for-
metal. The oxide formation in FZ during welding of titanium alloys mation in the FZ [26].
is very important, since it influences the mechanical properties of Fig. 8 presents the influence of welding conditions on the evo-
the weldments [24]. Lathabai et al. [41] reported that for TIG weld- lution of prior b grains and the average grain sizes in FZ of LBW
ing of Cp Ti, bright silvery appearance of the weld bead showed weldments. Akman et al. [25] reported that a decrease in prior b

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 7

Fig. 6. Evolution of prior b grains at different welding conditions (a) weldment 6 and (b) weldment 3.

Fig. 7. EDX analysis for different welding conditions (a) as-received sample, (b) heat input per unit length of 111.4 J/mm and peak power of 4.6 kW, (c) oxygen percentage at
different condition of heat input per unit length and (d) oxygen percentage at different condition of peak power.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
8 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

Fig. 8. FZ microstructure under different condition of heat input per unit length and peak power (a) 111.4 J/mm and 4.64 kW, (b) 85.7 J/mm and 2.9 kW, (c) variation of grain
size with heat input per unit length and (d) variation of grain size with peak power.

grain size increased the hardness and tensile strength of Ti-6Al-4V becomes steeper. Akman et al. [25] also reported that at a constant
weldments. Furthermore, they also concluded that high cooling average power, any significant increase in peak power by reducing
rates can be achieved by performing the welding operation at rel- the pulse width will result in higher target temperatures and
atively low average power and heat input. It is well established increased cooling rates. It can be further observed in Fig. 8d that
that the grain size is strongly related to the cooling rate achieved the grain size increased by approximately 43% when the peak
during solidification of the weldments and an increase in cooling power increases from 6.2 to 9.3 kW. This trend is attribute to the
rate reduces the prior b grain size [43]. In the current study, weld- substantial increase in heat input per unit length by 38 J/mm when
ment 10 exhibited the minimum grain size of 33 lm since it was the peak power was increased due to which relatively lower cool-
performed at the lowest heat input per unit length and peak power ing rates were attained.
of 34.3 J/mm and 1.4 kW, respectively. The maximum grain size of For Further analysis, the welding conditions of weldment 1
92 lm was observed for weldment 1 which was performed at heat were selected as the optimum combination of welding parameters
input per unit length and peak power of 111.4 J/mm and 4.6 kW. As since they resulted in minimum oxygen concentration among all
shown in Fig. 8c, an increase of heat input per unit length by 77 J/ the full penetration welds and an acceptable level of pulse overlap.
mm at a fixed welding speed and pulse width, resulted in an The weldment had a V-shaped bead having a top and bottom width
increase of prior b grain size by approximately 178%. This is due of approximately 2.01 mm and 0.52 mm, respectively. For the cho-
to the fact that increase of heat input per unit length decreases sen sample, detailed microstructural and phase analysis was per-
the cooling rate [25]. Although the increase of heat input in formed using SEM and XRD. Residual stress analysis, plate
Fig. 8a also led to an increase of peak power by 3.2 kW, however deformation measurement, microhardness measurement, tensile
a much better estimate of the effect of increase in peak power on testing and fractography were performed systematically.
the prior b grain size is shown in Fig. 8d. This figure shows the rela- Fig. 9 shows in detail the evolved microstructures in FZ, HAZ
tionship between peak power and grain size for a fixed value of and base metal of the selected weld sample. Fig. 9a and b shows
pulse energy. The heat input per unit length for each data point the FZ/HAZ/base metal interfaces using optical microscope and
is also shown which varies from 89.1 J/mm to 127.2 J/mm in the SEM (BSE with in-beam mode) while Fig. 9c, d and e shows the
peak power range of 4.6–12.4 kW. It can be seen that for a fixed higher magnification images of FZ. Each of these zones (FZ, HAZ
value of heat input per unit length of 111.4 kW, increasing the peak and base metal) experienced thermal cycles of different magni-
power by 7.8 kW resulted in a decrease of grain size by 36%. For a tudes and therefore different microstructures in different areas.
fixed heat input per unit length, increase in peak power increases In Fig. 9a, the HAZ microstructure changes from equiaxed a grains
the value of peak temperature achieved and thermal gradient near the base metal to a much refined acicular a and a0 martensite

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 9

Fig. 9. Variation of microstructures in different zones of LBW joint (a) Optical Micrograph of base metal/HAZ/FZ interface, (b) SEM images of base metal/HAZ/FZ interface, (c)
OM image of FZ, (d) SEM image of FZ and (e) SEM image at 5000 of FZ martensite.

near the HAZ/FZ interface. The equiaxed grains in base metal nature of the heating process, intermittent cooling occurs which
appeared to become elongated and relatively refined in the HAZ. further increases the cooling rate and refine the grains [46,47]. This
The as-received base metal was in annealed form and had equiaxed a0 martensite in FZ of pulsed LBW was also reported for Cp-Ti [48],
grains whereas the post welding cooling process was unidirec- Ti-6Al-4V [14,20,49] and Ti-6Al-2Sn-4Zr-2Mo [22].
tional which resulted in columnar grains in HAZ. Such a significant Although in Fig. 3a, the weld bead appeared to be free from
variation of microstructure in HAZ was also observed in previous pores, however, in the current study, as shown in Fig. 9e some
studies [13,14,20,25] and was attributed to the high cooling rate micro pores were observed at a much higher magnification of
near the FZ/HAZ interface and relatively low cooling rate near 5000. This suggested that the defects were at a very small scale
the base metal/HAZ boundary. Fig. 9a further shows the existence owing to the adequate shielding of weld area. Moreover, it is
of a very narrow HAZ, also reported by Yunlian et al. [44] and Gao expected that the tensile properties will not be affected signifi-
et al. [20] for LBW of Cp Ti and Ti-6Al-4V alloy, respectively. This is cantly due to these small scale defects as defects like micro poros-
due to very focused and dense laser heat source resulting in ity are commonly observed in laser welding of titanium alloys
reduced overall heat input required to achieve complete [14,50,51].
penetration. The aforementioned microstructure evolution was supported
In the FZ (Fig. 9c and d), however, nearly complete formation of by the XRD performed on both base metal and FZ for the determi-
a0 martensite was observed. During welding, FZ experienced tem- nation of different phases. In Fig. 10a and b the XRD spectrum for
perature far above the melting point of the alloy whereas in HAZ, base metal and FZ are shown respectively in which all the domi-
the temperature remained below the melting point, but above nant peaks are associated with hexagonal closed pack (hcp) struc-
the btransus. During the cooling cycle in FZ and HAZ, b phase is ture of a phase. The analysis of base metal microstructure in Fig. 2
retained for a short period [45] and then it is converted in a confirms that b phase is present at the boundary of equiaxed a
diffusion-less manner to needle like a0 martensite if the cooling phase. However b phase in a titanium alloys is very difficult to
rate is greater than 410 °C/s [43]. Such a high cooling rate is observe as it has a very weak peak in the XRD spectrum [52].
expected in pulsed LBW because of two reasons; firstly because The hcp pattern in Fig. 10a and b can be attributed to both a and
of the dense laser heat source [20] and secondly due to the pulsed a0 martensite as they have similar lattice parameters [53]. The only

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
10 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

(a)

1.2
(b)
1.0

Angle of deformation
0.8

(degree)
0.6

0.4

0.2

0.0
Fig. 10. XRD results (a) fusion zone and (b) base metal. 0 20 40 60 80 100
Longitudnal coordinate (X/mm)

observed difference between the spectra of FZ and base metal was Fig. 12. (a) Plate deformation and (b) angular deformation.
the increase in peak intensities at 38.5° and a decrease in peak
intensity at 70.85° of FZ spectrum. In both Fig. 10a and b the peak martensite led to a decrease in intensity of peak at 70.85° in
at 70.85° is attributed to a phase and the peak at 38.5° is associated Fig. 10a and an increase in a0 martensite which was marked by
to both a and a0 martensite. The conversion of a phase to a0 an increase in peak intensity at 38.5°.

Fig. 11. Longitudinal and transverse residual stresses at (a) surface, (b) depth of 0.4 mm and (c) depth of 0.8 mm from surface and at various locations from the weld center
line.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 11

assumed in the residual stress profile. It was reasonable to assume

that data points at the same distance but at either sides of the cen-
terline would have the same value of residual stress. These types of
symmetric approaches have been adopted in a number of research
studies on weldments [20,54,55]. At the weld center line, presence
of weld bead made the top surface of the plate uneven; therefore
strain gauge rosette could not be installed. The closest measure-
ment point was at 3 mm from the weld center line which was
nearly at the HAZ/base metal interface. The relieved strains in
hole-drill strain measurement at each of the rosette location were
recorded at specific depth intervals of 0.4 mm and 0.8 mm. This
gave an estimate of through thickness residual stress profile with
reasonable accuracy [56]. During welding process, the moving
laser instantly melts the area ahead of it while the flow of shielding
gas suddenly cools the weld areas behind the heat source. This
simultaneous heating and cooling leads to non-uniform plastic
deformation of the weld zone, however as it is structurally con-
nected with the parent metal, the constraint to maintain original
Fig. 13. Micro hardness distribution in FZ, HAZ and base metal. shape induces residual stress. Fig. 11 shows that the induced lon-
gitudinal and transverse residual stresses were of magnitude less
than 100 MPa and were significantly less than the yield strength
Fig. 11 shows the longitudinal residual stress and transverse (950 MPa) of alloy under study. This is primarily attributed to over-
residual stress distribution at different locations across the weld all low heat input of the LBW process as concluded by Zhang et al.
center line. As shown in Fig. 1, two rosettes were installed at the [54]. Moreover the unclamped setup during welding caused some
right side of weld centerline at a distance of 3 mm and 13 mm from of the stresses to be relieved when the plate shrunk and deformed
the centerline. Similarly three rosettes were installed on the left during the cooling process [57].
side of the weld centerline at a distance of 8 mm, 18 mm and 23 Residual stresses are generated primarily due to differential
mm from the centerline. Since the boundary conditions at both plastic flow and thermal strains, however phase change also plays
end of the weld centerline are similar, a symmetric approach was a significant role in residual stress profile [58]. In Fig. 11, the max-

Fig. 14. Results of tensile testing (a) stress-strain curve, (b) fractured samples and (c) overview of fractured surfaces.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
12 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

imum values of longitudinal and transverse residual stress are not Fig. 12 shows the deformation of the plate after laser welding
close the weld center line as expected. This behavior was explained process. The plates were not clamped during the welding opera-
by Rossini et al. [39] as a combination of shrinkage, quenching and tion, therefore the deformation took place under the influence
phase changes. The residual stress will be maximum near the weld of residual stresses which were generated as the molten weld
center line if pure thermal strains are involved and no phase pool solidifies [60]. In the unclamped setup, some of the residual
change take place. However, inclusion of phase change complicates stresses were relieved as the plate deforms and the overall mag-
the residual stress distribution leading to maximum value of resid- nitude was less than a fully clamped setup [57]. The extent of
ual stress away from the weld centerline. For titanium alloys, the deformation in weldments is associated with the heat input. Since
welding process led to formation of a0 martensite and acicular a in pulsed LBW process, the overall heat input was relatively low,
from b phase [45]. This caused the bcc (b) to hcp (a0 , a) transforma- the measured deformations were less than 1 mm as shown in
tion which further led to a local increase in volume of the lattice Fig. 12a. Moreover, the welding process involved deformations
[59]. Hence at the weld center line, this phase change either in both longitudinal and transverse directions which are evident
induced compressive longitudinal and transverse residual stress from the angle of deformation at different positions as shown
or may lead to reduction of tensile residual stress. It can also be in Fig. 12b.
observed that in the surface and bottom layers, both longitudinal The formation of finely dispersed a0 martensite and acicular a in
and transverse residual stresses were of compressive nature. How- FZ and HAZ significantly affects the microhardness distribution in
ever longitudinal and transverse residual stresses present at 0.4 both the a and a + b titanium alloys in a similar manner [19].
mm below the surface were mainly tensile in nature. According to Zeng and Bieler, 2005 [61], the hardness of these

Fig. 15. SEM fractography of the tensile specimens (a) LBW sample, (b) higher magnification image of LBW sample and (c) base metal.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 13

increases in the sequence: a0 martensite > acicular a > b. Wang and by reducing the pulse width to 3 ms, at constant average power
Wu [62] also reported that hardness of b phase is less than a and a0 resulted in excessive pool melting, expulsion and spattering at
martensite for a + b titanium alloys. The microhardness distribu- the top side of the weld bead and an undercut at the bottom.
tion across the welded zone is shown in Fig. 13, in which the 2. The increase of peak power at a fixed value of heat input per
dashed lines represent the average Vickers hardness in FZ, HAZ unit length resulted in an increase of oxygen concentration in
and base metal. The strengthening mechanism leading to the the FZ. The maximum oxygen contents were found to be
increase in hardness for the Ti-5Al-2.5Sn can be observed in approx. 25.2% at a peak power and heat input of 12.37 kW
Fig. 13, where the average hardness of FZ is about 56 HV more than and 111.4 J/mm respectively. Moreover, the prior b grain size
the hardness of base metal. This increase is related to the forma- reduced by 36% by increasing the peak power by 7.8 kW at a
tion of high strength a0 martensite in FZ of LBW joint. Furthermore constant heat input per unit length of 111.4 J/mm. Furthermore,
a very steep gradient of 72.6HV/mm was observed in the HAZ due the grain growth shifted from a uni-directional at lower weld-
to its small width as compared to width of FZ. ing speed to multi directional at higher welding speed.
The results for tensile testing are presented in Fig. 14a. As 3. For the selected welded sample with lowest oxygen content,
shown in Fig. 14b, LBW samples were fractured at the base metal fast cooling rate associated with full penetration pulsed LBW
which suggested that the weld zone was stronger than the base of Ti-5Al-2.5Sn alloy led to much refined, a0 martensitic struc-
metal. The strength of FZ and HAZ depends to a larger extent on ture in the FZ, and in the HAZ/FZ interface.
the microhardness value. Gao et al. [20] showed that for continu- 4. The tensile tests of the welded specimen resulted in a ductile
ous CO2 LBW of Ti-6Al-4V alloy sheet, the Vickers hardness of FZ fracture in the un-welded section, due to trans-granular micro
was more than HAZ and base metal and fracture occurred at the void coalescence. This suggested increased strength of the
base metal. They further showed that TIG weldment fractured at welded zone which was attributed to increase in Vickers hard-
the FZ during tensile tests because the hardness of FZ was less than ness of the FZ by 55HV compared to the base metal due to for-
base metal. Pasang et al. [19] showed that for Ti-5Al-5V-5Mo-3Cr, mation of a0 martensite. Moreover, LBW joints showed
a b titanium alloy, the joints fabricated by EBW, LBW and TIG weld- approximately the same elongation as that of the base metal
ing had less tensile strength than the parent metal and all fractured samples.
at the welded zone as the hardness at the weld zone was less than 5. The low heat input in pulsed LBW induced mostly compressive
the base metal. The strengthening mechanism in a + b and a tita- longitudinal and transverse residual stresses of magnitude less
nium alloys is the same, however the formation of strengthening than 100 MPa which was much less than the alloy’s yield
precipitates of a0 is suppressed in b titanium alloys [19,52]. strength. Furthermore the plate deformations associated with
It was expected that the compressive nature of transverse resid- the welding operation was less than a millimeter indicating
ual stress would increase the UTS or YS of the LBW sample. How- low thermal shrinkage.
ever transverse residual stress was of compressive nature only at 6. The microstructural evolution complicated the residual stress
the surface and at 0.8 mm depth, whereas at 0.4 mm depth, trans- distribution across the weld center line. Residual stresses were
verse residual stress was of tensile nature with a maximum value found to be non-uniform along the thickness and due to their
of about 70 MPa at 13 mm from the weld centerline. These oppos- low magnitude, tensile properties were not significantly
ing magnitudes counter balanced the effect of compressive layers affected.
resulting in no significant change of UTS and YS.
Fig. 15 shows the SEM images of fractured surfaces for both
LBW and base metal samples. Both samples have similar and Acknowledgments
uneven topographic features at the fractured surface (Fig. 14c).
Features of LBW sample appears in Fig. 15a and b in which the The authors would like to thank Ghulam Ishaq Khan Institute of
marked area in Fig. 15a is shown at a higher magnification in Engineering Sciences and Technology for providing funding for this
Fig. 15b. The fractured surfaces for both welded and base metal project. Furthermore the author wish to thank Pakistan Welding
samples had a dull and matte appearance, with dimpled ruptures Institute and Pakistan Institute of Engineering and Applied
indicating a predominantly ductile fracture mode, with a minor Sciences for use of research facilities.
influence of brittle fracture as evident from the fewer cracks. How-
ever, the presence of small number of cracks suggested the pres-
ence of fine dimples which confirmed that micro void References
coalescence was the dominant fracture mechanism. At different
[1] S.D. Luo, M. Qian, Microwave processing of titanium and titanium alloys for
locations the micro void coalescence led to trans-granular fracture, structural, biomedical and shape memory applications: current status and
therefore cleavage facets were formed as shown in Fig. 15a–c. challenges, Mater. Manuf. Process. (2016) 1–15, http://dx.doi.org/10.1080/
Since the fracture took place in the base metal where the dominant 10426914.2016.1257800.
[2] A.I. Khorev, Development of titanium sheet alloys for welded structures
a phase has coarse equiaxed grains, the origination of fine dimples working at high temperatures, Weld. Int. 30 (2016) 389–394, http://dx.doi.org/
can be attributed to strain hardening during tensile tests. The 10.1080/09507116.2015.1090169.
micro voids were most likely generated at the a/b grain boundary [3] A. Revathi, S. Magesh, V.K. Balla, M. Das, G. Manivasagam, Current advances in
enhancement of wear and corrosion resistance of titanium alloys – a review,
where there was a mismatch of strain hardening between the two Mater. Technol. 31 (2016) 696–704, http://dx.doi.org/10.1080/
different phases [22]. Moreover tear ridges, pores and cracks were 10667857.2016.1212780.
also visible at few locations. [4] R. Humberto Mota de Siqueira, A. Capella de Oliveira, R. Riva, A. Jorge Abdalla,
M. Sérgio Fernandes de Lima, Comparing mechanical behaviour of aluminium
welds produced by laser beam welding (LBW), friction stir welding (FSW), and
4. Conclusions riveting for aeronautical structures, Weld. Int. 30 (2016) 497–503, http://dx.
doi.org/10.1080/09507116.2015.1096501.
[5] R.R. Boyer, An overview on the use of titanium in the aerospace industry,
1. LBW performed at different welding conditions showed that the Mater. Sci. Eng., A 213 (1996) 103–114, http://dx.doi.org/10.1016/0921-5093
weld pool width is strongly dependent on both the peak power (96)10233-1.
and heat input per unit length. A maximum pulse overlap of [6] http://www.suppliersonline.com/Research/Property/result.asp?FamilyID=16&
MetalID=2522&Chemical=1&Physical=1&Mechanical=1, (n.d.). http://www.
approx. 91% was achieved by reducing the welding speed to suppliersonline.com/Research/Property/result.asp?FamilyID=16&MetalID=
140 mm/min. Moreover, increase in peak power to 12.37 kW 2522&Chemical=1&Physical=1&Mechanical=1.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
14 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx

[7] E.E. Billinghurst Jr., Tensile properties of cast titanium alloys: Titanium-6Al-4V [33] S. Song, P. Dong, Residual stresses at weld repairs and effects of repair
ELI and Titanium-5Al-2.5Sn ELI, NASA Technical Paper, 1992. <http:// geometry, Sci. Technol. Weld. Joining 22 (2017) 265–277, http://dx.doi.org/
ntrs.nasa.gov/search.jsp?R=19930000884> (accessed October 4, 2016). 10.1080/13621718.2016.1224544.
[8] M.J. Tan, G.W. Chen, S. Thiruvarudchelvan, High temperature deformation in [34] C. Liu, Q. Ge, D. Chen, F. Gao, J. Zou, Residual stress variation in a thick welded
Ti–5Al–2.5Sn alloy, J. Mater. Process. Technol. 192 (2007) 434–438, http://dx. joint after ultrasonic impact treatment, Sci. Technol. Weld. Joining 21 (2016)
doi.org/10.1016/j.jmatprotec.2007.04.027. 624–631, http://dx.doi.org/10.1080/13621718.2016.1149932.
[9] Q.Y. Sun, H.C. Gu, Tensile and low-cycle fatigue behavior of commercially pure [35] S. Bate, M. Smith, Determination of residual stresses in welded components by
titanium and Ti–5Al–2.5Sn alloy at 293 and 77 K, Mater. Sci. Eng., A 316 (2001) finite element analysis, Mater. Sci. Technol. 32 (2016) 1505–1516, http://dx.
80–86, http://dx.doi.org/10.1016/S0921-5093(01)01249-7. doi.org/10.1080/02670836.2016.1209882.
[10] Q.Y. Sun, S.J. Song, R.H. Zhu, H.C. Gu, Toughening of titanium alloys by [36] H. Chmelíčková, H. Šebestová, Pulsed Laser Welding, Nd YAG Laser, 2012, pp.
twinning and martensite transformation, J. Mater. Sci. 37 (2002) 2543–2547, 41–58. <http://www.intechopen.com/books/nd-yag-laser/pulsed-laser-
http://dx.doi.org/10.1023/A:1015456026919. welding>.
[11] A. Costa, R. Miranda, L. Quintino, D. Yapp, Analysis of beam material [37] Y.F. Tzeng, Parametric analysis of the pulsed Nd:YAG laser seam-welding
interaction in welding of titanium with fiber lasers, Mater. Manuf. Process process, J. Mater. Process. Technol. 102 (2000) 40–47, http://dx.doi.org/
22 (2007) 798–803, http://dx.doi.org/10.1080/10426910701446671. 10.1016/S0924-0136(00)00447-7.
[12] F.M. Ghaini, M.J. Hamedi, M.J. Torkamany, J. Sabbaghzadeh, Weld metal [38] G.S. Schajer, Measurement of non-uniform residual stresses using the hole-
microstructural characteristics in pulsed Nd: YAG laser welding, Scripta Mater. drilling method. Part II. Practical application of the integral method, J. Eng
56 (2007) 955–958, http://dx.doi.org/10.1016/j.scriptamat.2007.02.019. Mater Technol, Transact ASME 110 (1988) 344–349, http://dx.doi.org/10.1115/
[13] A. Squillace, U. Prisco, S. Ciliberto, A. Astarita, Effect of welding parameters on 1.3226059.
morphology and mechanical properties of Ti-6Al-4V laser beam welded butt [39] N.S. Rossini, M. Dassisti, K.Y. Benyounis, A.G. Olabi, Methods of measuring
joints, J. Mater. Process. Technol. 212 (2012) 427–436, http://dx.doi.org/ residual stresses in components, Mater. Des. 35 (2012) 572–588, http://dx.doi.
10.1016/j.jmatprotec.2011.10.005. org/10.1016/j.matdes.2011.08.022.
[14] X.Ã. Cao, M. Jahazi, Effect of welding speed on butt joint quality of Ti – 6Al – 4V [40] H. Liu, K. Nakata, J.X. Zhang, N. Yamamoto, J. Liao, Microstructural evolution of
alloy welded using a high-power Nd: YAG laser, Opt. Lasers Eng. 47 (2009) fusion zone in laser beam welds of pure titanium, Mater. Charact. 65 (2011) 1–
1231–1241, http://dx.doi.org/10.1016/j.optlaseng.2009.05.010. 7, http://dx.doi.org/10.1016/j.matchar.2011.12.010.
[15] M.Y. Krasnoperov, R.R.G.M. Pieters, I.M. Richardson, Weld pool geometry [41] S. Lathabai, B.L. Jarvis, K.J. Barton, Comparison of keyhole and conventional gas
during keyhole laser welding of thin steel sheets, Sci. Technol. Weld. Joining 9 tungsten arc welds in commercially pure titanium, Mater. Sci. Eng., A 299
(2004) 501–506, http://dx.doi.org/10.1179/136217104225021733. (2001) 81–93, http://dx.doi.org/10.1016/S0921-5093(00)01408-8.
[16] X. Gao, L. Zhang, J. Liu, Effects of weld cross-section profiles and microstructure [42] X. Li, Effects of oxygen contamination in the argon shielding gas in laser
on properties of pulsed Nd: YAG laser welding of Ti6Al4V sheet (2014) 895– welding of commercially pure titanium thin sheet, 2005, pp. 3437–3443.
903. http://dx.doi.org/10.1007/s00170-014-5722-x. [43] T. Ahmed, H.J. Rack, Phase transformations during cooling in a + b titanium
[17] G. Krolczyk, A. Sedmak, U. Kumar, S. Chattopadhyaya, A.K. Das, A. Pramanik, alloys, Mater. Sci. Eng., A 243 (1998) 206–211, http://dx.doi.org/10.1016/
Study of heat-affected zone and mechanical properties of Nd-YAG laser S0921-5093(97)00802-2.
welding process of thin titanium alloy sheet, Natur. Resour. Eng. 1 (2016) 51– [44] Q. Yunlian, D. Ju, H. Quan, Z. Liying, Electron beam welding, laser beam
58, http://dx.doi.org/10.1080/23802693.2016.1246278. welding and gas tungsten arc welding of titanium sheet, Mater. Sci. Eng. A 280
[18] H. Liu, K. Nakata, N. Yamamoto, J. Liao, Mechanical properties and (2000) 177–181, http://dx.doi.org/10.1016/S0921-5093(99)00662-0.
strengthening mechanisms in laser beam welds of pure titanium, Sci. [45] J.W. Elmer, T.A. Palmer, S.S. Babu, W. Zhang, T. DebRoy, Phase transformation
Technol. Weld. Joining 16 (2011) 581–585, http://dx.doi.org/10.1179/ dynamics during welding of Ti-6Al-4V, J. Appl. Phys. (2004), http://dx.doi.org/
1362171811Y.0000000054. 10.1063/1.1737476.
[19] T. Pasang, J.M. S??nchez Amaya, Y. Tao, M.R. Amaya-Vazquez, F.J. Botana, J.C. [46] B. Mehdi, R. Badji, V. Ji, B. Allili, D. Bradai, F. Deschaux-Beaume, F. Soulié,
Sabol, W.Z. Misiolek, O. Kamiya, Comparison of Ti-5Al-5V-5Mo-3Cr welds Microstructure and residual stresses in Ti-6Al-4V alloy pulsed and unpulsed
performed by laser beam, electron beam and gas tungsten arc welding, TIG welds, J. Mater. Process. Technol. 231 (2016) 441–448, http://dx.doi.org/
Procedia Eng. 63 (2013) 397–404, http://dx.doi.org/10.1016/j. 10.1016/j.jmatprotec.2016.01.018.
proeng.2013.08.202. [47] N.K. Babu, S.G. Sundara, R. Mythili, S. Saroja, Correlation of microstructure
[20] X.-L. Gao, L.-J. Zhang, J. Liu, J.-X. Zhang, A comparative study of pulsed Nd:YAG with mechanical properties of TIG weldments of Ti – 6Al – 4V made with and
laser welding and TIG welding of thin Ti6Al4V titanium alloy plate, Mater. Sci. without current pulsing, vol. 58, 2007, pp. 581–587. 10.1016/
Eng., A 559 (2013) 14–21, http://dx.doi.org/10.1016/j.msea.2012.06.016. j.matchar.2006.07.001.
[21] D. Eylon, R. Boyer, D.A. Koss, M. and M.S.T.C. Minerals, M. and M.S.A.M. (122nd [48] Z. Sun, I. Annergren, D. Pan, T. Mai, Effect of laser surface remelting on the
: 1993 : D. Minerals, Beta titanium alloys in the 1990’s : proceedings of a corrosion behavior of commercially pure titanium sheet, Mater. Sci. Eng., A
Symposium on Beta Titanium Alloys, Minerals, Metals Mater. Soc., 1993. 345 (2003) 293–300, http://dx.doi.org/10.1016/S0921-5093(02)00477-X.
[22] A. Chamanfar, T. Pasang, A. Ventura, W.Z. Misiolek, Mechanical properties and [49] T.S. Balasubramanian, M. Balakrishnan, V. Balasubramanian, M.A.M.
microstructure of laser welded Ti-6Al-2Sn-4Zr-2Mo (Ti6242) titanium alloy, Manickam, Influence of welding processes on microstructure, tensile and
Mater. Sci. Eng., A 663 (2016) 213–224, http://dx.doi.org/10.1016/j. impact properties of Ti-6Al-4V alloy joints, Transact. Nonferrous Metals Soc.
msea.2016.02.068. China (English Ed.) 21 (2011) 1253–1262, http://dx.doi.org/10.1016/S1003-
[23] M. Baruah, S. Bag, Influence of heat input in microwelding of titanium alloy by 6326(11)60850-9.
micro plasma arc, J. Mater. Process. Technol. 231 (2016) 100–112. [50] R. Xiao, X. Zhang, Problems and issues in laser beam welding of aluminum–
[24] M. Baruah, S. Bag, Microstructural influence on mechanical properties in lithium alloys, J. Manufact. Process. 16 (2014) 166–175, http://dx.doi.org/
plasma microwelding of Ti6Al4V alloy, J. Mater. Eng. Perform. 25 (2016) 4718– 10.1016/j.jmapro.2013.10.005.
4728, http://dx.doi.org/10.1007/s11665-016-2333-8. [51] S. Meshram, T. Mohandas, Welding studies on a near-alpha titanium alloy,
[25] E. Akman, A. Demir, T. Canel, T. Sinmazçelik, Laser welding of Ti6Al4V titanium Defence Sci. J. 61 (2011) 590–596, http://dx.doi.org/10.14429/dsj.61.648.
alloys, J. Mater. Process. Technol. 209 (2009) 3705–3713, http://dx.doi.org/ [52] M.R. Amaya-Vazquez, J.M. Sánchez-Amaya, Z. Boukha, F.J. Botana,
10.1016/j.jmatprotec.2008.08.026. Microstructure, microhardness and corrosion resistance of remelted TiG2
[26] M. Baruah, S. Bag, Influence of pulsation in thermo-mechanical analysis on and Ti6Al4V by a high power diode laser, Corros. Sci. 56 (2012) 36–48, http://
laser micro-welding of Ti6Al4V alloy, Opt. Laser Technol. 90 (2017) 40–51, dx.doi.org/10.1016/j.corsci.2011.11.006.
http://dx.doi.org/10.1016/j.optlastec.2016.11.006. [53] S. Malinov, W. Sha, Z. Guo, C. Tang, A. Long, Synchrotron X-ray diffraction
[27] A. Karpagaraj, N. Siva shanmugam, K. Sankaranarayanasamy, Some studies on study of the phase transformations in titanium alloys, Mater. Charact. 48
mechanical properties and microstructural characterization of automated TIG (2002) 279–295, http://dx.doi.org/10.1016/S1044-5803(02)00286-3.
welding of thin commercially pure titanium sheets, Mater. Sci. Eng., A 640 [54] J.X. Zhang, Y. Xue, S.L. Gong, Residual welding stresses in laser beam and
(2015) 180–189, http://dx.doi.org/10.1016/j.msea.2015.05.056. tungsten inert gas weldments of titanium alloy, Sci. Technol. Weld. Joining 10
[28] S. Bannour, K. Abderrazak, H. Mhiri, G. Le Palec, Effects of temperature- (2005) 643–646, http://dx.doi.org/10.1179/174329305X48374.
dependent material properties and shielding gas on molten pool formation [55] A.R. Kohandehghan, S. Serajzadeh, A.H. Kokabi, A study on residual stresses in
during continuous laser welding of AZ91 magnesium alloy, Opt. Laser Technol. gas tungsten arc welding of AA5251, Mater. Manufact. Process. 25 (2010)
44 (2012) 2459–2468, http://dx.doi.org/10.1016/j.optlastec.2012.03.034. 1242–1250, http://dx.doi.org/10.1080/10426914.2010.481004.
[29] H. Sanati, F. Reshadi, G. Faraji, N. Soltani, E. Zalnezhad, Proceedings of the [56] G.S. Schajer, Application of Finite Element Calculations to Residual Stress
institution of mechanical engineers Part B, J. Eng. Manuf. (2014), http://dx.doi. Measurements, vol. 103, 2014, pp. 157–163.
org/10.1177/0954405414534432. [57] N. Ma, H. Huang, H. Murakawa, Effect of jig constraint position and pitch on
[30] J.A. Goldak, M. Akhlaghi, Computational welding mechanics, Springer Sci. Bus. welding deformation, J. Mater. Process. Technol. 221 (2015) 154–162, http://
Media (2005), http://dx.doi.org/10.1007/b101137. dx.doi.org/10.1016/j.jmatprotec.2015.02.022.
[31] S. Sarafan, J.-B. Lévesque, P. Wanjara, J. Gholipour, H. Champliaud, Distortion [58] W. Guo, J.A. Francis, L. Li, A.N. Vasileiou, D. Crowther, A. Thompson, Residual
and residual stresses in electron beam-welded hydroelectric turbine materials, stress distributions in laser and gas-metal-arc welded high-strength steel
Sci. Technol. Weld. Joining 21 (2016) 473–478, http://dx.doi.org/10.1080/ plates, Mater. Sci. Technol. 32 (2016) 1449–1461, http://dx.doi.org/10.1080/
13621718.2015.1132127. 02670836.2016.1175687.
[32] E.S. Costa, P.D.C. Assunção, E.B.F. Dos Santos, L.G. Feio, M.S.Q. Bittencourt, E.M. [59] B. Appolaire, A. Settefrati, E. Aeby-Gautier, Stress and strain fields associated
Braga, Residual stresses in cold-wire gas metal arc welding, Sci. Technol. Weld. with the formation of a00 in near-b titanium alloys, Mater. Today: Proceed. 2
Joining (2017) 1–8, http://dx.doi.org/10.1080/13621718.2017.1306014. (2015) S589–S592, http://dx.doi.org/10.1016/j.matpr.2015.07.353.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010
 M. Junaid et al. / Optics and Laser Technology xxx (2017) xxx–xxx 15

[60] K. Fahlström, O. Andersson, A. Melander, L. Karlsson, L.-E. Svensson, Khalid Rahman did his PhD in Electrohydrodynamic jet
Correlation between laser welding sequence and distortions for thin sheet printing of electronics devices from JNU, South Korea.
structures, Sci. Technol. Weld. Joining 22 (2017) 150–156, http://dx.doi.org/ His research interest is in additive manufacturing pro-
10.1080/13621718.2016.1207046. cesses.
[61] L. Zeng, T.R. Bieler, Effects of working, heat treatment, and aging on
microstructural evolution and crystallographic texture of a, a0 , a00 and b
phases in Ti–6Al–4V wire, Mater. Sci. Eng., A 392 (2005) 403–414, http://dx.
doi.org/10.1016/j.msea.2004.09.072.
[62] S. Wang, X. Wu, Investigation on the microstructure and mechanical
properties of Ti – 6Al – 4V alloy joints with electron beam welding, Mater.
Des. 36 (2012) 663–670, http://dx.doi.org/10.1016/j.matdes.2011.11.068.

Massab Junaid is pursuing his PhD at GIK Institute on

the welding of titanium alloys using advanced joining
techniques. His research interest includes mechanical Mirza Nadeem Baig did his PhD from GIK Institute of
properties of welded joints, and FE simulations of Engineering Sciences and Technology, Pakistan in 2014.
welding process. Presently, he is working as Head Academics at Pakistan
Welding Institute, Islamabad and has more than
24 years of experience in the field of welding and join-
ing. His current research interests are mainly in the area
of joining of Ti, Ni, Al and their alloys using laser, TIG
and plasma welding processes.

Fahd Nawaz Khan did his PhD from Northumbria

University, Newcastle Upon Tyne, UK. In (2011). His
research interests are primarily in the area of machining
of high performance alloys such as Titanium and Nickel
alloys. In his PhD, he worked on the bandsawing of
these advanced alloys, using un-coated and coated
carbide bandsaws. This project was a combination of
mechanical and materials engineering.

Please cite this article in press as: M. Junaid et al., Effect of laser welding process on the microstructure, mechanical properties and residual stresses in Ti-
5Al-2.5Sn alloy, Opt. Laser Technol. (2017), http://dx.doi.org/10.1016/j.optlastec.2017.07.010