Materials Sciences and Applicatio ns, 2011, 2, 251-257
doi:10.4236/msa.2011.24032 Published Online April 2011 (
Copyright © 2011 SciRes. MSA
Process Parameters Optimization of an Aluminium
Alloy with Pulsed Gas Tungsten Arc Welding
(GTAW) Using Gas Mixtures
Pawan Kumar1, Kishor Purushottamrao Kolhe2, Sashikant Janardan Morey3,
Chanchal Kumar Datta1
1Department of Mechanical Engineering, Delhi College of Engineering, Delhi, India; 2 Department of Mechanical Engineering, Agricul-
tural University, Ratnagiri, India; 3Department of Production Engineering, College of Engineering and Technology, Akola, India.
Email: {kishor_kolhe, ckd5929},
Received December 7th, 2010; revised February 1st, 2011; accepted March 4th, 2011.
This paper demonstrates the enhancement of mechanical properties and effective optimization of pulsed GTAW process
parameters on aluminium alloy 60 61 using sinusoidal AC wave with argon p lus helium gas mixtures. Modified Ta guchi
Method (MTM) was employed to formulate experimental layout and to study effect of process parameter optimization
on mechanical properties of the weld joints. Microstructural characterization of weld joint was carried out to under-
stand the str uctural pro perty correlat i o n wi th process p ara me t e rs.
Keywords: Pulsed GTAW, Aluminium Alloy 6061, Gas Mixtures, Mechanical Properties , Modified Taguchi Method,
1. Introduction
Aluminium alloy 6061 (Al-Mg-Si alloy) is widely used
in the fabrication of food processing equipment, chemi-
cal containers, passenger cars, road tankers and railway
transport systems due to its high strength, excellent
weldability and resistance to corrosion. As these Al al-
loys (6xxx series) are precipitation-hardened, they suffer
from a strength reduction in heat affected zone (HAZ)
and like most face-centered cubic metals, they do not
exhibit clear endurance limit [1]. GTAW is an arc weld-
ing process, where arc is produced between non-consum-
able tungsten electrode and base metal. Pulsed GTA
welding process is frequently used for welding of alu-
minium alloys as heat input during welding can be pre-
cisely controlled. This process is strongly characterized
by bead geometry, which plays an important role in de-
termining mechanical properties of the weld [2]. Pulse
process variables are controlling factor for heat input,
which in turn leads to grain refinement in fusion zone,
width reduction of HAZ, segregation of alloying ele-
ments, reducing hot cracking sensitivity and residual
stresses [3,4].
Improved mechanical properties of weld are achieved
by using current pulsing due to the grain refinement oc-
curring in the fusion zone. The main aim of pulsing is to
achieve maximum penetration without excessive heat
built-up. The use of high current pulses is to penetrate
deep and cater for longer arc period at lower current.
Deep penetration in pulsed current welding is produced
by arc pressure at peak for longer durations [4]. In addi-
tion to this argon-helium gas mixture offers certain ad-
vantages by increasing heat input of the arc during weld-
ing. Argon is known for stable arc with better arc ignition
whereas helium provides higher thermal conductivity.
There exists a linear relationship between heat input of a
weld and maximum temperature at a given distance from
the weld centre line. It shows that pulsed arc welds are
cooler and exhibit less thermal distortion than conven-
tional GTA welds of the same penetration [5]. Heat input
is typically calculated as follows:
H = [60 E I]/1000 S
where H = Heat Input (kJ/mm), E = Arc Voltage (Volts),
I = Current (Amps) and S = Travel Speed (mm/min).
The evolution of microstructure in weld fusion zone is
influenced by current pulsing and cyclic variations of
energy input into the weld pool causing thermal fluctua-
tions. Consequently this leads to periodic interruption in
solidification process. As pulse current decays, solid liq-
Process Parameters Optimization of an Aluminium Alloy with Pulsed Gas Tungsten Arc Welding
(GTAW) Using Gas Mixtures
uid interface advances towards the arc and becomes sus-
ceptible for disturbances in the arc formation. As current
increases again in the subsequent pulse, the growth of
dendrites can also occur. Current pulsing also results in
periodic variations of the arc forces resulting in addi-
tional fluid flow, which lowers the temperature in front
of the solidifying interface. Furthermore, the temperature
fluctuations inherent in pulsed welding leads to a contin-
ual change in the weld pool size and shape favoring
growth of new grains. It is to be understood that effective
heat input for unit volume of weld pool should be con-
siderably less in pulse current welds and thus expecting
the average weld pool temperatures to be low [6].
A weldment basically consists of five microstructur-
ally distinct regions normally identified as fusion zone,
unmixed region, partially melted region, HAZ, and unaf-
fected base metal. The HAZ is the portion of weld joint
which experiences peak temperatures high enough to
produce solid-state microstructural changes but they are
too low to cause any melting. Every point of the weld-
ment in the HAZ relative to the fusion line is subjected to
unique thermal experience during welding, in terms of
both maximum temperature and cooling rate. Thus, each
point has its own microstructural features and corrosion
susceptibility. Partially melted region extends usually
one or two grains into the HAZ relative to the fusion line.
It is characterized by grain boundary liquation, which
may result in liquation cracking. An increase in the pulse
frequency refines grain structure of weld metal using
pulses of short duration. At a given frequency, long pulse
duration produces coarser grain structure than shorter
pulse duration. Further increase in the peak current leads
to coarseness of grain structure [7].
Taguchi method is a powerful tool for design of
high-quality systems, widely used for improving quality
without increasing cost and with minimum experimenta-
tion. It provides a simple, efficient and systematic ap-
proach to optimize designs for performance, quality and
cost. This method is valuable when process parameters
are qualitative and discrete. The parameter design based
on the Taguchi method can optimize the quality charac-
teristics through the settings of process parameters and
reduce the sensitivity of the system performance to the
sources of variation [8]. In fact, Taguchi method had
been designed to optimize a single quality characteristic.
However, Modified Taguchi method (MTM) is used for
several quality characteristics taken together into consid-
eration for selection of process parameters [9,10].
2. Experimental Procedure
The material under investigation is 5.00 mm thick Al
alloy 6061. A non-consumable tungsten electrode of 2.4
mm Φ shielded by argon gas is used to strike the arc with
base metal. Filler rods (31.5 mm Φ) of Aluminium alloy
5183 are recommend for welding of this alloy for getting
maximum strength and elongation. The chemical compo-
sition of base metal and filler rod are tabulated in Table
1. Sample plates of size 300 × 150 × 5 mm were prepared
by milling and EDM wire-cut machines. Welding of the
samples was carried out on Automatic Pulse GTAW Tri-
ton 220 V AC/DC. In order to remove oil, moisture and
oxide layer from base metal, they were thoroughly wire
brushed, cleaned with acetone and preheated at 150˚C in
the oven. The quality of weld is based on the process
parameters, such as pulse current in the range of 150 -
210 A, base current in the range of 75 - 135 A, pulse fre-
quency in the range of 50 - 150 Hz, pulse-on-time in the
range of 30% - 90% and percentage of He in Ar + He
mixtures in a range of 10% - 50%.
Tensile specimens of required dimension as per ASTM
E8M were separated out from welded coupon plates and
tests were carried out on 400 kN computer controlled
Universal Testing Machine. The specimens were loaded
at the rate of 1.5 kN/minute as per ASTM specifications,
so that the tensile specimen undergoes deformation. All
specimens finally fail after necking and the influence of
load on displacement profile was recorded. Higher the
tensile properties have better quality characteristics.
At the same time for microhardness measurement,
samples of transverse cross-section of joint were taken
from the weld coupons. The specimens formicrostruc-
tural characterization were mechanically polished using
220, 320, 400, 600 and 1000 grit waterproof SiC emery
papers and alumina grade-II paste. Microhardness tests
were carried out on a Leco Digital Microhardness Tester
with 50 gf load and 15 second dwell time incorporated
with diamond indenter. Final polishing was carried out
using 3 and 1 μm diamond paste. To reveal the macro-
structures of the welded sample, deep etching technique
using 10% HF acid solution in water was used. However,
for further revealing of microstructures, Keller’s reagent
was used as etchants. Metallurgical microscope coupled
with Image Analysis system was used for the micro-
structural characterization and to measure the porosity
level in the weldment.
Table 1. Chemical composition of Base metal and filler rod.
Weight Percentage
Alloys Mg Si Cu Zn Mn Al
6061 1.1 0.7 0.25 0.25 - Rest
5183 4.55 - - 0.1 0.65 Rest
opyright © 2011 SciRes. MSA
Process Parameters Optimization of an Aluminium Alloy with Pulsed Gas Tungsten Arc Welding 253
(GTAW) Using Gas Mixtures
3. Selection of Process Parameters and Their
The bead profile and quality of the GTA weld is greatly
dependent on the joint design, welding position and
proper selection of process parameters, such as current
polarity, pulse current, secondary current (base current),
pulse frequency, pulse duty cycle, welding voltage,
welding speed, arc length and shielding gas flow rate.
The thermal behaviour of weld is governed by arc char-
acteristics. The metal transfer during welding signifi-
cantly affects the weld bead geometry, chemistry and
microstructure [11]. Welding parameters are selected on
the basis of literature survey.
A large number of trial runs were carried on
5mm-thick aluminium alloy 6061 samples to find out the
feasible working limits of pulsed current GTAW pa-
rameters. After examining bead morphology of samples,
the upper and lower limit of the peak current, base cur-
rent, pulse duty cycle and pulse frequency were decided.
The process parameters and their levels are shown in
Table 2. Some constant process parameters used during
welding are shown in Table 3. The bead contour, bead
appearance and weld quality of the weld joint were in-
spected in order to identify the welding parameters. From
the above analysis, following observations are made:
1) If the peak current (P) was less than 150 A, incom-
plete penetration and lack of fusion were observed. At the
same time, if P exceeds 210 A, the defects like under cut,
spatters and overheating were observed.
2) If base current (B) was lower than 75 A, heat input
was very less because of very short arc length. On the
other hand, for B greater than 135 A, arc became unstable
and lengthy.
3) If the pulse frequency (F) was lower than 50 Hz, the
bead contour and bead appearance was not of good quality.
However, if the F was larger than 150 Hz, the harsh sound
emanated from welding machine.
4) If the pulse duty cycle (T) was less than 30%, the
generated heat input was not sufficient to melt the base
metal. On the contrary, if T was larger than 90%, over
melting of base & filler metals and overheating of tung-
sten electrode were observed.
5) If the helium (X) in mixture of argon and helium was
less than 10%, the arc stability, bead penetration and bead
appearance were poor. On the other hand if X was greater
than 50%, gas consumption gas mixture per kg weld
deposition was very high.
4. Optimum Selection of Process Parameters
by MTM
Modified Taguchi Method (MTM) is used to find out the
optimal process parameters to enhance the mechanical
Table 2. Process Parameters and their limiting Values.
Symbol Process
Parameter Unit 1 2 3 4 5
P Peak Current A 150 165 180195210
B Base Current A 75 90 105120135
F Pulse Frequency Hz 50 75 100125150
T Pulse Duty
Cycle % 30 45 60 7590
X Percentage of
He in Ar % 10 20 30 4050
Table 3. Constant Process Parameters.
Process Parameters Constant Values
Shielding gas (Argon) flow rate, l/min 8.0
Filler rod diameter, mm 3.15
Electrode material 98% W + 2% Zr
Electrode diameter, mm 2.4
AC Balance, %age 0.00 (Sinusoidal)
properties by conducting minimum experimental runs.
4.1. Orthogonal Array
In this study, four levels were chosen for four factors.
Limiting values of these factors are tabulated in Table 2.
There are 20 degrees of freedom and hence L25 or-
thogonal array was used. Experimental layout for con-
ducting test is shown in Table 4.
Table 4. Experimental Layout using L25 Orthogonal Array.
Levels of Process Parameters
Expt No. P B F T X
1 1 1 1 1 1
2 1 2 2 2 2
3 1 3 3 3 3
4 1 4 4 4 4
5 1 5 5 5 5
6 2 1 2 3 4
7 2 2 3 4 5
8 2 3 4 5 1
9 2 4 5 1 2
10 2 5 1 2 3
11 3 1 3 5 2
12 3 2 4 1 3
13 3 3 5 2 4
14 3 4 1 3 5
15 3 5 2 4 1
16 4 1 4 2 5
17 4 2 5 3 1
18 4 3 1 4 2
19 4 4 2 5 3
20 4 5 3 1 4
21 5 1 5 4 3
22 5 2 1 5 4
23 5 3 2 1 5
24 5
4 3 2 1
25 5 5 4 3 2
opyright © 2011 SciRes. MSA
Process Parameters Optimization of an Aluminium Alloy with Pulsed Gas Tungsten Arc Welding
(GTAW) Using Gas Mixtures
4.2. Experimental Results
The experiments were conducted as per orthogonal array
and results of the mechanical properties as ultimate tensile
strength (UTS), microhardness of weld zone and per-
centage elongation are tabulated in Table 5. All tensile
specimens fractured at weld joint, therefore their me-
chanical properties correspond to that of weld joint. Also
microhardness profile of three samples (03, 08 and 11)
were measured at a distance of 0.5 mm across the weld as
shown in Figure 1 and plotted in Figure 2.
Microstructures of above three samples were charac-
terized by image analysis system for the measurement of
porosity level and grain size of the weld zone and HAZ. It
was observed that the grain size of the sample 11 was less
then 08 and 03 samples in both the HAZ and weld area.
Also the porosity level in the weld joint of D11 was
comparatively lower than that of the 08 and 03 samples.
Microstructures and characterization data of 11, 08 and 03
samples (weld zone and HAZ) are shown in Figure 3 and
Table 6 respectively.
For optimizing the process parameters, four quality
characteristics were considered for a single characteristic.
Hence the Modified Taguchi Method (MTM) is used.
Weighted response of mechanical properties was obtained
by adding weights to the responses as a single quality
Table 5. Experi me ntal results for the mechanical properties.
No. UTS (MPa) Microhardness
(VPN) Elongation (%age)
01. 256.40 114.1 08.71
02. 267.47 112.6 12.94
03. 283.36 098.7 10.00
04. 288.08 112.5 10.90
05. 298.75 112.9 09.70
06. 293.76 105.8 07.84
07. 290.74 097.8 09.00
08. 352.85 110.6 09.08
09. 272.12 120.8 09.92
10. 386.07 099.2 10.70
11. 290.74 109.0 10.00
12. 272.63 112.4 10.36
13. 282.62 109.3 12.06
14. 280.75 100.5 10.86
15. 300.95 103.9 08.12
16. 300.18 105.7 09.20
17. 288.05 107.0 11.30
18. 282.42 102.2 09.32
19. 290.74 115.4 10.58
20. 278.27 101.4 09.90
21. 295.69 108.1 10.60
22. 293.02 095.6 11.04
23. 294.00 102.5 09.62
24. 301.50 097.0 10.48
25. 310.03 099.5 10.58
Figure 1. Macrograph of weld joint showing schematic of
microhardness me asurements .
Table 6. Microstructure characterization results of welded
Grain Size in
μm (HAZ)
Grain Size in μm
(Weld Zone)
Porosity in
03 49 ± 5 17 ± 3 3
08 37 ± 5 15 ± 3 2.5
11 29 ± 5 12 ± 3 0.8
characteristic. Also weights for UTS, microhardness in-
weld zone and elongation at break were selected as 0.640,
0.254 and 0.106 respectively with help of AHP (Ana-
lytical Hierarchy Process). Weighted response is tabulated
in Table 7.
4.3. Analysis of Variance (ANOVA) Calculations
Analysis of variance is most important tool for calculat-
ing responsible factors, which significantly affects me-
chanical properties. For determining these affect on
process parameters, F-test was performed. Results of
ANOVA and percentage contributions by each process
parameters are tabulated in Table 8. Response graphs
Table 7. Weighted response for mechanical properties.
Expt No.Weighted ResponseExpt No. Weighted Response
1. 185.78 14. 196.99
2. 192.57 15. 209.69
3. 197.97 16. 209.84
4. 206.64 17. 203.19
5. 210.99 18. 198.26
6. 205.84 19. 206.98
7. 202.02 20. 195.63
8. 242.75 21. 207.98
9. 197.18 22. 203.06
10. 259.82 23. 205.33
11. 205.17 24. 208.45
12. 195.27 25. 214.26
13. 200.65
opyright © 2011 SciRes. MSA
Process Parameters Optimization of an Aluminium Alloy with Pulsed Gas Tungsten Arc Welding
(GTAW) Using Gas Mixtures
Copyright © 2011 SciRes. MSA
Figure 2. Microhardness Profile of GTA Welded Aluminum Alloy 6061.
Figure 3. (a) 03, (b) 08, (c) 11, microstructures of HAZ and weld z one at 100X and (d) 03, (e) 08, (f) 11, microstructures of weld
zone at 200X.
Table 8. Results of analysis of variance.
Welding parameter Deg of freedom Sum of square Mean squareF Contributed %age
Pulse Current (A) 4 1621.34 405.34 2.902 29.31
Base Current (A) 4 992.89 248.22 1.777 17.95
Pulse Frequency (Hz) 4 611.07 152.77 1.094 11.08
Pulse Duty Cycle (%) 4 1205.99 301.49 2.158 21.80
Percentage of He in Ar 4 541.21 135.30 0.969 09.79
Error 4 558.79 139.69 10.10
Total 24 5531.29
(Figure 4) are drawn from response table indicated Ta-
ble 9, to identify the significant levels of each factor.
4.4. Confirmation Test
Optimal level of process parameters was predicted using
response graph and ANOVA. Process parameters and
their levels which affect mechanical properties are pulse
current at level 2, background current at level 5, pulse
frequency at level 4, pulse duty cycle at level 2 and per-
centage of Helium in Argon at level 3. The obtained re-
sults were verified for the improvement in multiple qual-
ity characteristics by conducting a confirmation test based
on results obtained in Tabl e 10.
5. Discussion
ANOVA calculations shown in Table 8, shows that pulse
Process Parameters Optimization of an Aluminium Alloy with Pulsed Gas Tungsten Arc Welding
(GTAW) Using Gas Mixtures
Figure 4. Response graph of mechanical properties.
current, base current, pulse duty cycle, frequency and
percentage of Helium in Argon plays significant role on
microstructure and mechanical properties of welded Al
alloy 6061, but pulse current plays the maximum role i.e.
29.31%. In this investigation, pulse current of 165 A,
background current of 135 A, pulse frequency of 125 Hz,
pulse duty cycle of 45% and 30% of Helium with Argon
affects the maximum to the mechanical properties.
As shown in Figure 2, all three samples show almost
same trend of microhardness across weld zone, HAZ and
base metal. Lower microhardness was observed in the
weld zone in all samples because of using low hardness of
filler rod, coarse dendritic solidified microstructure and
interdendritic segregated phase. Microhardness of HAZ is
also lower than base metal but at the fusion zone a higher
value can be seen. This increase in microhardness is due to
the fast cooling rate experienced near the weld area, which
produces a super saturated solid solution. During subse-
quent cooling and natural ageing phenomena very fine
precipitate of Mg2Si may have formed near weld area
Table 9. Response table for mechanical properties.
parameter Level 1 Level 2Level 3 Level 4Level 5
Current (A) 198.79 221.52 201.54 202.78207.82
Current (A) 202.92 199.22 208.99 203.25218.08
Frequency (Hz) 208.78 204.08201.85 214.87203.99
Pulse Duty
Cycle (%) 195.84 214.27203.65 204.92213.79
of He in Ar 209.97 201.49213.60 202.36205.03
Table 10. Confirmation test results.
Optimal level of
process parameters
P2 B5 F4 T2 X3
P2 B5 F4 T2 X3 389.31 121.4 12.98
of HAZ.
From microstructures as shown in Figure 3, it is ap-
parent that sample 11 results in fine dendritic micro-
structure and lower porosity level as compared to other
sample (03 and 08). Therefore from the microstructural
characterization it is observed that the pulse parameters
play an important role in development of fine micro-
6. Conclusions
An experimental study has been conducted to understand
the effect of process parameters of pulsed current GTA
welding on aluminium alloy weldments. Five important
process parameters of pulsed current GTA welding were
used with gas mixtures to optimize the four quality char-
acteristics of mechanical properties. Modified Taguchi
Method was used with success to identify the optimum
parameters. Experimental results coupled with ANOVA
results proved that pulse current is having pronounced
effect on the multiple quality characteristics of the me-
chanical properties and microstructural characterization.
Generation of fine dendritic microstructure resulted in
improvement of mechanical properties. The confirmation
experiments conducted with predicted levels of factors
proved to be worthy.
[1] T. Senthilkumar, V. Balasubramanian and M. Y. S. Babu,
“Effect of Pulsed Current GTAW Parameters on Fusion
Zone Microstructure of AA6061,” International Journal
of Metals & Materials, Vol. 13, No. 4, 2007, pp. 345-351.
[2] S. C. Juang and Y. S. Tarng, “Process Parameter Selec-
tion for Optimizing the Weld Pool Geometry in the TIG
Welding of Stainless Steel,” Journal of Material Process
Technology, Vol. 122, No. 1, 2002, pp. 33-37.
[3] A. A. Mohamed, “Optimisation of Weld Bead Dimen-
sions in GTAW of Al-Mg Alloy,” Materials and Manu-
facturing Processes, Vol. 16, No. 5, 2001, pp. 725-736.
[4] M. Balasubramanian, V. Jayabalan and V. Balasubrama-
nian, “Optimizing the Pulsed Current GTAW Parameters
to Attain Maximum Impact Toughness,” Materials and
Manufacturing Processes, Vol. 23, No. 1-2, 2008, pp.
69-73. doi:10.1080/10426910701524584
[5] R. E. Leitner, H. Mcelhinney and E. I. Pruitt, “Aninvesti-
gation of Pulsed Welding Variables,” Welding Journal,
Vol. 52, No. 9, 1973, pp. 405-410.
[6] G. M. Reddy, A. A. Gokhale and K. P. Rao, “Optimization
of Pulse Frequency in Pulsed Current Gas Tungsten Arc
Welding of Aluminium-Lithium Alloy Sheets,” Journal of
Material Science and Technology, Vol. 14, 1998, pp.
[7] R. Manti, D. K. Dwivedi and A. Agarwal, “Microstruc-
opyright © 2011 SciRes. MSA
Process Parameters Optimization of an Aluminium Alloy with Pulsed Gas Tungsten Arc Welding
(GTAW) Using Gas Mixtures
Copyright © 2011 SciRes. MSA
ture and Hardness of Al-Mg-Si Weldments Produced by
Pulse GTA Welding,” International Journal of Advance
Manufacturing Technology, Vol. 36, No. 3-4, 2008, pp.
263-269. doi:10.1007/s00170-006-0849-z
[8] Y. S. Tarng and W. H. Yang, “Optimization of the Weld
Bead Geometry in GTA Welding by the Taguchi
Method,” International journal of Advance Manufactur-
ing Technology, Vol. 14, 1998, pp. 549-554.
[9] A. Kumar and S. Sundarrajan, “Optimization of Pulsed
TIG Welding Process Parameters on Mechanical Proper-
ties of AA 5456 Al Alloy Weldments,” Materials and
Design, Vol. 30, No. 4, 2009, pp. 1288-1297.
[10] A. Kumar and S. Sundarrajan, “Selection of Welding
Process Parameters for the Optimum but Joint Strength of
an Aluminum Alloy,” Material and Manufacturing Proc-
ess, Vol. 21, No. 8, 2006, pp. 789-793.
[11] M. Balasubramanian and V. Balasubramanain, “Process
Parameter Optimisation of the Pulsed GTAW of Titanium
Alloy,” Journal of Material Science and Technology, Vol.
24, No. 3, 2008, pp. 423-426.