Académique Documents
Professionnel Documents
Culture Documents
& Design
Materials and Design 28 (2007) 20802092
www.elsevier.com/locate/matdes
b,*
,
M.Y. Sanavullah
Department of Automobile Technology, School of Engineering and Technology. Bharathidasan University, Tiruchirappalli 620 024, Tamil Nadu, India
b
Department of Manufacturing Engineering, Annamalai University, Annamalainagar 608 002, Tamil Nadu, India
c
V.M.K.V. Engineering College, Salem 636 308, Tamil Nadu, India
Received 12 September 2005; accepted 24 May 2006
Available online 4 August 2006
Abstract
Medium strength aluminium alloy (AlMgSi alloy) has gathered wide acceptance in the fabrication of light weight structures requiring a high strength-to-weight ratio, such as transportable bridge girders, military vehicles, road tankers and railway transport systems. In
any structural application of this alloy consideration its weldability is of utmost importance as welding is largely used for joining of structural components. The preferred welding process of aluminium alloy is frequently tungsten inert gas (TIG) welding due to its comparatively easier applicability and better economy. In the case of single pass TIG welding of thinner section of this alloy, the pulsed current
has been found benecial due to its advantages over the conventional continuous current process. The use of pulsed current parameters
has been found to improve the mechanical properties of the welds compared to those of continuous current welds of this alloy due to
grain renement occurring in the fusion zone. Many considerations come into the picture and one need to carefully balance various pulse
current parameters to arrive at an optimum combination. Hence, in this investigation an attempt has been made to study the inuence of
pulsed current TIG welding parameters on tensile properties of AA 6061 aluminium alloy weldments.
2006 Elsevier Ltd. All rights reserved.
Keywords: Pulsed current; Tungsten inert gas welding; Medium strength aluminium alloy; Tensile strength; Design of experiments; Analysis of variance
1. Introduction
Weld fusion zones typically exhibit coarse columnar
grains because of the prevailing thermal conditions during
weld metal solidication. This often results in inferior weld
mechanical properties and poor resistance to hot cracking.
It is thus highly desirable to control solidication structure
in welds and such control is often very dicult because of
higher temperatures and higher thermal gradients in welds
in relation to castings and the epitaxial nature of the
growth process. Nevertheless, several methods for rening
weld fusion zones have been tried with some success in the
past: inoculation with heterogeneous nucleants [1], microc-
ooler additions, surface nucleation induced by gas impingement and introduction of physical disturbance techniques
such as torch vibration [2].
The use of inoculants for rening the weld fusion zones
was, as a matter of fact, not as successful as in castings
because of the extremely high temperatures involved in
welding and also due to the undesirable eects of inoculating elements on weld mechanical properties at the levels
required for producing grain renement. Other techniques
like surface nucleation and microcooler additions were also
turned down because of the complicated welding set-ups
and procedures associated with their use. In this process,
two relatively new techniques namely, magnetic arc oscillation and current pulsing, have gained wide popularity
because of their striking promise and the relative ease with
which these techniques can be applied to actual industrial
situations with only minor modications of the existing
welding equipment [3].
(v)
(vi)
(vii)
(viii)
2081
2082
Table 1
Important factors and their levels
S. no.
Factor
Unit
Notation
Levels
Low (1)
High (+1)
1.
2.
3.
4.
Peak current
Base current
Pulse frequency
Pulse on time
A
A
Hz
%
P
B
F
T
160
80
2
40
180
90
6
60
Table 2
Experimental design matrix and the results
Expt. no.
YS
(MPa)
TS
(MPa)
EL
(%)
NTS
(MPa)
NSR
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
1
+1
1
+1
1
+1
1
+1
1
+1
1
+1
1
+1
1
+1
1
1
+1
+1
1
1
+1
+1
1
1
+1
+1
1
1
+1
+1
1
1
1
1
+1
+1
+1
+1
1
1
1
1
+1
+1
+1
+1
1
1
1
1
1
1
1
1
+1
+1
+1
+1
+1
+1
+1
+1
225
261
171
198
243
288
207
225
198
234
153
180
207
252
189
207
250
290
190
220
270
320
230
250
220
260
170
200
230
280
210
230
10
11.6
7.6
8.8
10.8
12.8
9.2
10
8.8
10.4
6.8
8
9.2
11.2
8.4
9.2
295
355
200
250
325
400
265
290
245
310
175
215
265
340
230
260
1.18
1.22
1.06
1.13
1.20
1.25
1.15
1.17
1.12
1.19
1.04
1.08
1.14
1.21
1.10
1.14
where Xi is the required coded value of a factor of any value X from Xmin to Xmax, Xmin is the lower level of the factor and Xmax is the upper level of the factor.
2.4. Conducting the experiments and recording the responses
The base metal used in this investigation is a medium
strength aluminium alloy of AA 6061 grade. The chemical
composition of the base metal was obtained using a vacuum spectrometer (ARL-Model: 3460). Sparks were
ignited at various locations of the base metal sample and
their spectrum was analysed for the estimation of alloying
Type of
material
Mg
Mn
Fe
Si
Cu
Al
Base metal
(AA 6061)
All weld metal
(AA 4043)
0.689
0.331
0.230
0.531
0.305
Bal.
0.050
0.220
0.050
5.0
0.120
Bal.
Table 3b
Mechanical properties of base metal and all weld metal
Joint type
Yield
strength
(MPa)
Ultimate
tensile
strength (MPa)
Elongation
(%)
Vickers
hardness
(0.05 kg)
Base metal
(AA 6061)
All weld metal
(AA 4043)
270
310
10
240
140
210
260
Table 6
Yates algorithm to calculate sum of squares (SS) for elongation (EL)
[+]
[]
[+]
[]
[1]
[2]
[3]
[4]
SS
225
261
171
198
243
288
207
225
198
234
153
180
207
252
189
207
486
369
531
432
432
333
459
396
36
27
45
18
36
27
45
18
855
963
765
855
63
63
63
63
117
99
99
63
9
27
9
27
1818
1620
126
126
216
162
36
36
108
90
0
0
18
36
18
18
3438
252
378
72
198
0
54
36
198
0
54
0
18
0
18
0
738,740
3969
8930.3
324
2450.3
0
182.25
81
2450.3
0
182.25
0
20.25
0
20.25
0
1
P
B
PB
F
PF
BF
PBF
T
PT
BT
PBT
FT
PFT
BFT
PBFT
Table 5
Yates algorithm to calculate sum of squares (SS) for tensile strength (TS)
[+]
[]
[1]
[2]
[3]
[4]
SS
250
290
190
220
270
320
230
250
220
260
170
200
230
280
210
230
540
410
590
480
480
370
510
440
40
30
50
20
40
30
50
20
950
1070
850
950
70
70
70
70
130
110
110
70
10
30
10
30
2020
1800
140
140
240
180
40
40
120
100
0
0
20
40
20
20
3820
280
420
80
220
0
60
40
220
0
60
0
20
0
20
0
912,025
4900
11,025
400
3025
0
225
100
3025
0
225
0
25
0
25
0
1
P
B
PB
F
PF
BF
PBF
T
PT
BT
PBT
FT
PFT
BFT
PBFT
[1]
[2]
[3]
[4]
SS
10
11.6
7.6
8.8
10.8
12.8
9.2
10
8.8
10.4
6.8
8
9.2
11.2
8.4
9.2
21.6
16.4
23.6
19.2
19.2
14.8
20.4
17.6
1.6
1.2
2
0.8
1.6
1.2
2
0.8
38
42.8
34
38
2.8
2.8
2.8
2.8
5.2
4.4
4.4
2.8
0.4
1.2
0.4
1.2
80.8
72
5.6
5.6
9.6
7.2
1.6
1.6
4.8
4
0
0
0.8
1.6
0.8
0.8
152.8
11.2
16.8
3.2
8.8
0
2.4
1.6
8.8
0
2.4
0
0.8
0
0.8
0
1459.2
7.84
17.64
0.64
4.84
0
0.36
0.16
4.84
0
0.36
0.004
0.04
0
0.04
0.005
1
P
B
PB
F
PF
BF
PBF
T
PT
BT
PBT
FT
PFT
BFT
PBFT
Table 7
Yates algorithm to calculate sum of squares (SS) for notch tensile strength
(NTS)
Table 4
Yates algorithm to calculate sum of squares (SS) for yield strength (YS)
Y
2083
[+]
[]
[1]
[2]
[3]
[4]
SS
295
355
200
250
325
400
265
290
245
310
175
215
265
340
230
260
650
450
725
555
555
390
605
490
60
50
75
25
65
40
75
30
1100
1280
945
1095
110
100
105
105
200
170
165
115
10
50
25
45
2380
2040
210
210
370
280
60
70
180
150
10
0
30
50
40
20
4420
420
650
130
330
10
80
60
340
0
90
10
30
10
20
20
1E + 06
11,025
26,406
1056.3
6806.3
6.25
400
225
7225
0
506.25
6.25
56.25
6.25
25
25
1
P
B
PB
F
PF
BF
PBF
T
PT
BT
PBT
FT
PFT
BFT
PBFT
Table 8
Yates algorithm to calculate sum of squares (SS) for notch strength ratio
(NSR)
[+]
[]
[1]
[2]
[3]
[4]
SS
1.18
1.22
1.06
1.13
1.2
1.25
1.15
1.17
1.12
1.19
1.04
1.08
1.14
1.21
1.1
1.14
2.4
2.19
2.45
2.32
2.31
2.12
2.35
2.24
0.04
0.07
0.05
0.02
0.07
0.04
0.07
0.04
4.59
4.77
4.43
4.59
0.11
0.07
0.11
0.11
0.21
0.13
0.19
0.11
0.03
0.03
0.03
0.03
9.36
9.02
0.18
0.22
0.34
0.3
0
0.06
0.18
0.16
0.04
0
0.08
0.08
0.06
0
18.38
0.4
0.64
0.06
0.34
0.04
0.16
0.06
0.34
0.04
0.04
0.06
0.02
0.04
9E 16
0.06
21.114
0.01
0.0256
0.0002
0.0072
1E 04
0.0016
0.0002
0.0072
1E 04
0.0001
0.0002
2E 05
1E 04
5E 32
0.0002
1
P
B
PB
F
PF
BF
PBF
T
PT
BT
PBT
FT
PFT
BFT
PBFT
2084
Sum of
squares (SS)
Degrees
of freedom
(d.o.f)
Similarly mathematical models have been developed to predict the tensile strength and elongation of the PCTIG
welded joints by incorporating pulsed current parameters
and they are given below:
Tensile strength; TS f239 17:5P 26:3B
13:75F 13:75T 5PB
3:75BF 3:75BTg MPa 6
Table 11
ANOVA (analysis of variance) test results for elongation
Mean
squares
[SS/d.o.f]
Fratio
[MS/error]
Factors
Sum of
squares (SS)
Degrees
of freedom
(d.o.f)
Mean squares
[SS/d.o.f]
Fratio
[MS/error]
Main factors
P
B
F
T
3969
8930.25
2450.25
2450.25
1
1
1
1
3969
8930.25
2450.25
2450.25
196
441
121
121
Main factors
P
B
F
T
7.84
17.64
4.84
4.84
1
1
1
1
7.84
17.64
4.84
4.84
196
441
121
121
Two factors
PB
a
PF
a
PT
BF
BT
a
FT
324
0
0
182.25
182.25
20.25
1
1
1
1
1
1
324
0
0
182.25
182.25
20.25
16
0
0
9
9
1
Two factors
PB
a
PF
a
PT
BF
BT
a
FT
0.64
0
0
0.36
0.36
0.04
1
1
1
1
1
1
0.64
0
0
0.36
0.36
0.04
16
0
0
9
9
1
101.25
18609.8
5
15
20.25
0.2
36.76
5
15
0.04
Error
Total
Error
Total
a
F(1,5,0.95) = 6.41. Therefore, PF, PT and FT are not signicant at 95%
condence level.
a
F(1,5,0.95) = 6.41. Therefore, PF, PT and FT are not signicant at 95%
condence level.
Table 10
ANOVA (analysis of variance) test results for tensile strength
Table 12
ANOVA (analysis of variance) test results for notch tensile strength
Factors
Main factors
P
B
F
T
Two factors
PB
a
PF
a
PT
BF
BT
a
FT
Error
Total
Sum of
squares (SS)
Degrees
of freedom
(d.o.f)
Mean
squares
[SS/d.o.f]
Fratio
[MS/error]
Factors
4900
11,025
3025
3025
1
1
1
1
4900
11,025
3025
3025
196
441
121
121
Main factors
P
B
F
T
400
0
0
225
225
25
1
1
1
1
1
1
400
0
0
225
225
25
16
0
0
9
9
1
Two factors
PB
a
PF
a
PT
BF
BT
a
FT
125
22,975
5
15
125
a
F(1,5,0.95) = 6.41. Therefore, PF, PT and FT are not signicant at 95%
condence level.
Error
Total
Sum of
squares (SS)
Degrees of
freedom
(d.o.f)
Mean squares
[SS/d.o.f]
Fratio
[MS/error]
191.7391
459.2391
118.3696
125.6522
11,025
26406.3
6806.25
7225
1
1
1
1
11,025
26406.3
6806.25
7225
1056.25
6.25
0
400
506.25
56.25
1
1
1
1
1
1
1056.25
6.25
0
400
506.25
56.25
5
15
57.5
287.5
53,775
18.36957
0.108696
0
6.956522
8.804348
0.978261
a
F(1,5,0.95) = 6.41. Therefore, PF, PT and FT are not signicant at 95%
condence level.
2085
Table 13
ANOVA (analysis of variance) test results for notch strength ratio
Table 14
Coecient of correlation for the developed mathematical models
Factors
S. no.
Correlation
coecient
1.
2.
3.
4.
5.
0.92
0.94
0.90
0.96
0.88
Sum of
Degrees of
Mean squares Fratio
[MS/error]
squares (SS) freedom (d.o.f) [SS/d.o.f]
Main factors
P
B
F
T
0.01
0.0256
0.00722
0.00722
1
1
1
1
0.01
0.0256
0.00722
0.00722
Two factors
a
PB
a
PF
a
PT
BF
a
BT
a
FT
0.00023
1E 04
1E 04
0.0016
0.0001
2.5E 05
1
1
1
1
1
1
0.00023
1E 04
1E 04
0.0016
0.0001
2.5E 05
Error
Total
0.00078
0.05298
64.51613
165.1613
46.6129
46.6129
1.451613
0.645161
0.645161
10.32258
0.645161
0.16129
5
15
F(1,5,0.95) = 6.41. Therefore, PB, PF, PT, BT and FT are not signicant
at 95% condence level.
5:63BTg MPa
280.00
Frequency = 2 Hz
Pulse on Time = 40 %
Frequency = 6 Hz
Pulse on Time = 40 %
280.00
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
240.00
240.00
200.00
200.00
160.00
150.00
160.00
170.00
180.00
150.00
190.00
160.00
170.00
180.00
190.00
Frequency = 4 Hz
Pulse on Time = 50 %
Frequency = 6 Hz
Pulse on Time = 60 %
260.00
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
240.00
200.00
240.00
220.00
200.00
160.00
180.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
170.00
180.00
190.00
2086
indicates high correlation exist between experimental values and predicted values. Similarly coecient of correlation has been calculated for all the developed models and
their values are presented in Table 14.
4. Discussion
YSp YS
Frequency = 2 Hz
Pulse on Time = 40 %
Frequency = 6 Hz
Pulse on Time = 40 %
320.00
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
320.00
280.00
240.00
280.00
240.00
200.00
150.00
160.00
170.00
180.00
190.00
150.00
180.00
190.00
Frequency = 6 Hz
Pulse on Time = 60 %
B = 80 Amps
B = 80 Amps
280.00
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
170.00
Frequency = 4 Hz
Pulse on Time = 50 %
280.00
160.00
240.00
200.00
240.00
200.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
170.00
180.00
190.00
2087
14.00
Frequency = 2 Hz
Pulse on Time = 40 %
12.00
Frequency = 6 Hz
Pulse on Time = 40 %
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
Elongation %
Elongation %
12.00
10.00
10.00
8.00
8.00
6.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
14.00
180.00
190.00
180.00
190.00
14.00
Frequency = 4 Hz
Pulse on Time = 50 %
Frequency = 6 Hz
Pulse on Time = 60 %
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
12.00
Elongation %
12.00
Elongation %
170.00
10.00
8.00
10.00
8.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
170.00
During tensile tests all the specimens were found to fracture within the weld region. Thus it may be assumed that
the ultimate tensile strength is primarily the ultimate tensile
strength of the weld. The use of pulsed current welding
improves the strength of the weld over that observed for
the case of continuous current welding. The renement of
microstructure due to the pulsed current welding results
in an uniform distribution of the ne precipitates more
eectively governed by its zinc pick up enhancing the
amount of precipitates in the matrix. Similar observation
has been made by other investigators also [14,15].
In general, hardness in the fusion zone is the lowest
due to the as-cast nature of the microstructure, which is
characterised by coarse dendritic grains, interdendritic
segregate phases, and the lack of strengthening phases.
Hardness is slightly higher in pulsed current welds as
2088
Frequency = 6 Hz
Pulse on Time = 40 %
360.00
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
320.00
280.00
B = 90 Amps
360.00
320.00
280.00
240.00
200.00
240.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
170.00
180.00
190.00
180.00
190.00
360.00
Frequency = 6 Hz
Pulse on Time = 60 %
Frequency = 4 Hz
Pulse on Time = 50 %
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
320.00
B = 90 Amps
320.00
280.00
B = 90 Amps
280.00
240.00
240.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
170.00
2089
1.25
Frequency = 2 Hz
Pulse on Time = 40 %
Frequency = 6 Hz
Pulse on Time = 40 %
B = 80 Amps
B = 80 Amps
B = 85 Amps
1.20
B = 85 Amps
B = 90 Amps
B = 90 Amps
1.20
1.10
1.15
1.10
1.00
1.05
150.00
160.00
170.00
180.00
190.00
150.00
170.00
180.00
190.00
170.00
180.00
190.00
1.20
Frequency = 4 Hz
Pulse on Time = 50 %
Frequency = 6 Hz
Pulse on Time = 60 %
B = 80 Amps
B = 80 Amps
B = 85 Amps
B = 85 Amps
B = 90 Amps
B = 90 Amps
1.15
1.15
160.00
1.10
1.05
1.10
1.05
1.00
1.00
150.00
160.00
170.00
180.00
190.00
150.00
160.00
with which these techniques can be applied to actual industrial situations with only minor modications to the existing welding equipment [20].
In general, the formation of equiaxed grain structure
in CCTIG weld is known to be dicult because of the
remelting of heterogeneous nuclei or growth centers
ahead of the solidliquid interface. This is due to the high
temperatures in the liquid, thus making survival nuclei
dicult. The evolution of microstructure in weld fusion
zone is also inuenced in many ways by current pulsing,
principally, the cyclic variations of energy input into the
weld pool cause thermal uctuations, one consequence
of which is the periodic interruption in the solidication
process. As the pulse peak current decays the solidliquid
interface advances towards the arc and increasingly
becomes vulnerable to any disturbances in the arc form.
As current increases again in the subsequent pulse,
2090
2091
Fig. 6 (continued )
2092
[10] Box GEP, Hunter WH, Hunter JS. Statistics for experiments. USA:
Wiley; 1978.
[11] Montgomery DC. Design and analysis of experiments. New York:
Wiley; 1991.
[12] Ravindra J, Parmar RS. Mathematical model to predict weld bead
geometry for ux cored arc welding. Metal Construct 1987;19:4552.
[13] Miller I, Freund JE, Johnson R. Probability and statistics for
engineers. New Delhi: Prentice of Hall of India Pvt. Ltd.; 1999.
[14] Ghosh PK, Vijay Sharma M. Chemical composition and microstructure in pulsed MIG welded AlZnMg alloy. J Mater Trans, JIM
1991;32:14550.
[15] Potluri NB, Ghosh PK, Gupta PC, Reddy YS. Studies on weld metal
characteristics and their inuences on tensile and fatigue properties of
pulsed current GMA welded AlZnMg alloy. Weld Res Suppl
1996:62s70s.
[16] Janaki Ram GD, Mitra TK, Shankar V. Microstructural renement
through inoculation of type 7020 AlZnMg alloy welds and its eect
on hot cracking and tensile property. J Mater Process Technol
2003;142:17481.
[17] Kou S, Le Y. Nucleation mechanism and grain rening of weld metal.
Weld J 1986:6570.
[18] Norman AF, Hyde K, Costello F, Thompson S, Birley S, Pragnell
PB. Examination of the eect of Sc on 2000 and 7000 series
aluminium castings: for improvements in fusion welding. Mater Sci
Eng 2003;1:1881998.
[19] Shinoda T, Ueno Y, Masumoto I. Eect of pulsed welding current on
solidication cracking in austenitic stainless steel welds. Trans Jpn
Weld Soc 1990;21:1823.
[20] Madhusudhan Reddy G. Welding of aluminium and alloys. In:
Proceedings of ISTE summer school on recent developments in
materials joining, Annamalai University; 2001.
[21] Lin DC, Wang TS, Srivatsan TS. A Mechanism for the formation of
equiaxed grains of aluminiumlithium alloy 2090. Mater Sci Eng A
2003;335:3049.