Vous êtes sur la page 1sur 13

Materials

& Design
Materials and Design 28 (2007) 20802092
www.elsevier.com/locate/matdes

Inuences of pulsed current tungsten inert gas welding parameters


on the tensile properties of AA 6061 aluminium alloy
T. Senthil Kumar a, V. Balasubramanian
a

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-

Corresponding author. Tel.: +91 4144 241147(R), +91 4144 239734


(O); fax: +91 4144 238275/238080.
E-mail address: visvabalu@yahoo.com (V. Balasubramanian).
0261-3069/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.matdes.2006.05.027

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].

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

Pulsed current tungsten inert gas (PCTIG) welding,


developed in 1950s, is a variation of tungsten inert gas
(TIG) welding which involves cycling of the welding current from a high level to a low level at a selected regular frequency. The high level of the peak current is generally
selected to give adequate penetration and bead contour,
while the low level of the background current is set at a
level sucient to maintain a stable arc. This permits arc
energy to be used eciently to fuse a spot of controlled
dimensions in a short time producing the weld as a series
of overlapping nuggets and limits the wastage of heat by
conduction into the adjacent parent material as in normal
constant current welding. In contrast to constant current
welding, the fact that heat energy required to melt the base
material is supplied only during peak current pulses for
brief intervals of time allows the heat to dissipate into the
base material leading to a narrower heat aected zone
(HAZ). The technique has secured a niche for itself in specic applications such as in welding of root passes of tubes,
and in welding thin sheets, where precise control over penetration and heat input are required to avoid burn through.
Extensive research has been performed in this process
and reported advantages include improved bead contour,
greater tolerance to heat sink variations, lower heat input
requirements, reduced residual stresses and distortion [4].
Metallurgical advantages of pulsed current welding frequently reported in literature include renement of fusion
zone grain size and substructure, reduced width of HAZ,
control of segregation, etc. [5]. All these factors will help
in improving mechanical properties. Current pulsing has
been used by several investigators to obtain grain renement in weld fusion zones and improvement in weld
mechanical properties [6,7]. However, reported research
work related to the eect of pulsed current parameters on
mechanical and metallurgical properties are very scanty.
Moreover, no systematic study has been reported so far
to analyse the inuence of pulsed current parameters on
mechanical and metallurgical properties.
Hence, in this investigation an attempt has been made to
develop mathematical models to predict the eects of
PCTIG welding parameters on tensile properties of medium strength AA 6061 aluminium alloy using statistical
tools such as design of experiments, analysis of variance
and regression analysis.
2. Scheme of investigation
In order to achieve the desired aim, the present investigation has been planned in the following sequence:

(v)
(vi)
(vii)
(viii)

2081

Recording the responses.


Developing mathematical models.
Identifying the signicant factors.
Checking the adequacy of the developed models.

2.1. Identifying the important parameters


From the literatures [58] and the previous work [9]
done in our laboratory, the predominant factors which
are having greater inuence on fusion zone grain renement of PCTIG welding process have been identied. They
are: (i) peak current; (ii) background current; (iii) pulse frequency; (iv) pulse on time.
2.2. Finding the working limits of the parameters
A large number of trial runs have been carried out using
5 mm thick rolled plates of AA 6061 aluminium alloy to
nd out the feasible working limits of PCTIG welding
parameters. AA 4043 (Al5%Si) aluminium alloy of
3 mm diameter has been used as the ller metal. Dierent
combinations of pulsed current parameters have been used
to carryout the trial runs. The bead contour, bead appearance and weld quality have been inspected to identify the
working limits of the welding parameters. From the above
analysis following observations have been made:
(i) If peak current is less than 160 A, then incomplete
penetration and lack of fusion have been observed.
At the same time, if peak current is greater than
180 A, then undercut and spatter have been observed
on the weld bead surface.
(ii) If background current is lower than 80 A, then the
arc length is found to be very short and addition of
ller metal becomes inconvenient. On the other hand,
if the background current is greater than 90 A, then
arc becomes unstable and arc wandering is observed
due to increased arc length.
(iii) If pulse frequency is less than 2 Hz, then the bead
appearance and bead contours are appear to be similar to that of constant current weld beads. Further, if
pulse frequency is greater than 6 Hz, then more arc
glare and arc spatter have been experienced.
(iv) If pulse on time is lower than 40%, then weld nugget
formation is not so smooth due to incomplete melting
of ller metal. On the contrary, if the pulse on time is
greater than 60%, then overmelting of ller metal and
overheating of tungsten electrode have been noticed.
2.3. Developing the experimental design matrix

(i) Identifying the important PCTIG welding parameters


which are having inuence on fusion zone grain
renement and tensile properties.
(ii) Finding the upper and lower limits of the identied
parameters.
(iii) Developing the experimental design matrix.
(iv) Conducting the experiments as per the design matrix.

By considering all the above conditions, the feasible limits


of the parameters have been chosen in such a way that the
AA 6061 aluminium alloy should be welded without any
weld defects. Due to narrow ranges of factors, it has been
decided to use two level, full factorial design matrix to optimise the experimental conditions. Table 1 presents the

2082

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092


Table 3a
Chemical composition (wt%) of base metal and all weld metal

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

ranges of factors considered and Table 2 shows the 16 set of


coded conditions used to form the design matrix 24 (two levels and four factors) factorial design. The 16 experimental
conditions (rows) have been formed for main eects by using
the formula 2nc  1 for the low (1) and high (+1) values;
where nc refers to the column number. For example, in
Table 2, the rst four rows are coded as 1 and next four
rows are coded as +1, alternatively, in the third column
[because nc = 3 and therefore 23  1 = 4]. The method of
designing such matrix is dealt elsewhere [10,11].
For the convenience of recording and processing the
experimental data, upper and lower levels of the factors
have been coded as +1 and 1, respectively and the coded
values of any intermediate levels can be calculated using
the following expression [12]:
X i X  X max X min =2=X max  X min =2

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

elements. The chemical composition of the base metal in


weight percent is given in Tables 3a and 3b. Tensile specimens were prepared as per the ASTM E8M-90a (ASTM,
1991a) guidelines. Tensile tests were carried out in
100 kN, electro-mechanical controlled Universal Testing
Machine (Make: FIE-BLUE STAR). The specimen was
loaded at the rate of 1.5 kN/min as per ASTM specications, so that tensile specimen undergoes deformation.
The specimen nally fails after necking and the load versus
displacement was recorded. The 0.2% oset yield strength
was derived from the diagram. Notched tensile specimens
were used to evaluate notch tensile strength and notch
strength ratio. Notch strength ratio is the ratio between
tensile strength of notched specimen and tensile strength
of unnotched specimen. At each experimental condition,
three specimens were tested and average values are presented in Table 2.
3. Developing mathematical models
Representing the tensile properties, say yield strength of
the joint by YS, the response function can be expressed as
[1012]
YS f Peak current; Base current; Pulse
frequency; Pulse on time
YS f P; B; F; T

The model selected includes the eects of main factors and


rst order interaction of all factors. It is a portion of power
series polynomial expressed as follows:
YS b0 b1 P b2 B b3 F b4 T b5 PB
b6 PF b7 PT b8 BF b9 BT
b10 FT b11 PBF b12 PBT b13 PFT
b14 BFT b15 PBFT

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

where b0 is the average of responses (yield strength); b1, b2,


b3, . . ., b15 are the coecients that depend on respective
main and interaction factors which are calculated by using
following expression [10]:
X
bi
X i  Y i =n
4
where i varies from 1 to n, in which Xi is the corresponding coded value of a factor and Yi is the corresponding response output value (yield strength) obtained from the
experiment and n is the total number of combinations
considered (in this case n = 16).
Analysis of variance (ANOVA) method has been
applied to nd out the signicance of main factors and
interaction factors. The higher order interactions (three
factor interactions and four factor interactions) are practically insignicant and hence not considered [13]. Yates
algorithm has been used to calculate sum of squares.
Tables 48 represent the Yates algorithm and in the
column marked (1), the upper half is obtained by adding

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

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

successive pairs of treatments and the lower half is


obtained by subtracting successive pairs. Columns (2), (3)
and (4) are obtained in the same manner from the entries
in columns (1), (2) and (3), respectively. Each sum of
square is obtained by squaring the corresponding eect
total and dividing the result by r 2nf, where r is number
of replicates (trials) and nf is the number of chosen factors. Further details regarding ANOVA method and Yates
algorithm are dealt with elsewhere [10,11].
3.1. Final mathematical models
ANOVA test results are presented in Tables 913. From
the ANOVA test results, it is evident that all the main factors (P, B, F, T) and few interaction factors (PB, BF and
Table 9
ANOVA (analysis of variance) test results for yield strength
Factors

Sum of
squares (SS)

Degrees
of freedom
(d.o.f)

BT) are considered to be signicant. Hence the nal model


is developed including only these signicant factors and it
is given below:
Yield strength; YS f215 15:75P  23:6B
12:38F  12:38T  4:5PB
3:38BF 3:38BTg MPa

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.

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

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.

Name of the mathematical model

Correlation
coecient

1.
2.
3.
4.
5.

Yield strength (YS)


Tensile strength (TS)
Elongation (EL)
Notch tensile strength (NTS)
Notch strength ratio (NSR)

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

Elongation; EL f9:6 0:7P  1:05B 0:55F


 0:55T  0:2PB 0:15BF
0:15BTg%

Notch tensile strength; NTS


f276 26:25P  40:6B 20:63F

5
15

 21:25T  8:13PB 5BF

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

Yield Strength, MPa

Yield Strength, MPa

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

Peak Current, Amps

Peak Current, Amps

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

Yield Strength, MPa

Yield Strength, MPa

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

Peak Current, Amps

Fig. 1. Eect of pulsed current parameters on yield strength.

170.00

Peak Current, Amps

180.00

190.00

2086

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

Notch strength ratio; NSR


f1:1 0:025P  0:04B 0:021F
 0:021T 0:01BFg

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.

3.2. Checking adequacy of the developed model

4. Discussion

Coecient of correlation r is used to nd how close the


predicted and experimental values lie and it is calculated
using the following expression:
r2 Explained variation=Total variation
.X
X
2
2
YSe  YS
10

YSp  YS

Mathematical models developed in the above section


have been written in C program and the developed C programs have been used to estimate the yield strength, tensile
strength, elongation, notch tensile strength and notch
strength ratio of the PCTIG welded AA 6061 aluminium
alloy welds for dierent combinations of pulsed current
parameters. Predicted values are plotted as graphs and they
are displayed in Figs. 15. The plotted graphs can be eectively used to understand the eect of pulsed current
parameter such as peak current, base current, pulse frequency and pulse on time on tensile properties of TIG
welded AA 6061 aluminium alloy joints.

where YSp is predicted (using the above model) yield


strength value for the given factors; YSe is the experimental
value for the corresponding factors; YS is the average of
experimental yield strength life values. The value of r
for the above-developed model is found to be 0.92, which

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

Tensile Strength, MPa

Tensile Strength, MPa

320.00

280.00

240.00

280.00

240.00

200.00

150.00

160.00

170.00

180.00

190.00

150.00

Peak Current, Amps

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

Tensile Strength, MPa

B = 90 Amps

Tensile Strength, MPa

170.00

Peak Current, Amps

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

Peak Current, Amps

180.00

190.00

150.00

160.00

170.00

Peak Current, Amps

Fig. 2. Eect of pulsed current parameters on tensile strength.

180.00

190.00

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

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

Peak Current, Amps

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

Peak Current, Amps

10.00

8.00

10.00

8.00

150.00

160.00

170.00

180.00

190.00

150.00

160.00

Peak Current, Amps

170.00

Peak Current, Amps

Fig. 3. Eect of pulsed current parameters on elongation.

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

compared to continuous current welds and this could


be due to the rened microstructure and low segregation
of strengthening phases. The moderately higher hardness
of pulsed current welds close to the fusion boundary is
possibly due to a large fraction of alloying elements in
solid solution at the end of the weld thermal cycle,
thereby giving conditions for extensive age hardening.
This can be explained as follows: at the fusion boundary,
precipitate dissolution occurs as the particles are exposed
to temperatures higher than 400 C during heating and
cooling as a result of welding. Dissolution process
enriches the solid solution of the aluminium matrix with
Mg, and Si. This resulted in increasing hardness. The
necessity of pulsed current welding and the mechanism
of grain renement due to pulsed current welding are discussed below.

2088

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092


400.00
Frequency = 2 Hz
Pulse on Time = 40 %

Frequency = 6 Hz
Pulse on Time = 40 %

360.00
B = 80 Amps

B = 80 Amps

B = 85 Amps

B = 85 Amps

Notch Tensile Strength, MPa

Notch Tensile Strength, MPa

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

Peak Current, Amps

160.00

170.00

180.00

190.00

180.00

190.00

Peak Current, Amps

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

Notch Tensile Strength, MPa

Notch Tensile Strength, MPa

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

Peak Current, Amps

150.00

160.00

170.00

Peak Current, Amps

Fig. 4. Eect of pulsed current parameters on notch tensile strength.

AlMgSi (AA 6061) alloys are known to be susceptible


to hot cracking. Control of solidication cracking is an
important consideration in the welding of these alloys and
this aspect often dictates the choice of ller material. They
are usually welded with ller material of type Al5%Mg
and less commonly, with Al5%Si, neither of which
responds to a post weld aging treatment. Filler materials
which are age hardenable and at the same time also resist
hot cracking have also been suggested and occasionally
used. Also, special techniques such as synchronous rolling
and the use of trailing heat sink have been suggested for
reducing tensile welding stresses and hence hot cracking of
high strength aluminium alloys during fusion welding [16].
Solidication cracking occurs when the thermal stresses
that build up during freezing exceed the strength of the
solidifying weld metal. Methods that are commonly used
to reduce the tendency for solidication cracking include:

altering weld metal composition, through the addition of


a ller wire, close process control, and controlling the grain
structure within the fusion zone. It is widely accepted that
by changing the welds grain structure, from coarse columnar to ne equiaxed, better cohesion strength can be
obtained, and the remaining eutectic liquid present during
the nal stages of solidication can be fed more easily and
heal any cracks that may form may be healed [17,18].
Another way of reducing the susceptibility to solidication cracking is through fusion zone grain renement,
which confers the further benet that the weld metal
mechanical properties are improved. Various grain renement techniques have been discussed in the literature for
aluminium alloy welds, e.g., electromagnetic stirring, current pulsing, torch vibration and inoculation. Of these,
pulsed current welding technique has gained wide popularity because of their striking promise and the relative ease

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

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

Notch Strength Ratio

Notch Strength Ratio

1.20

1.10

1.15

1.10

1.00

1.05
150.00

160.00

170.00

180.00

190.00

150.00

Peak Current, Amps


1.20

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

Notch Strength Ratio

1.15

Notch Strength Ratio

160.00

1.10

1.05

1.10

1.05

1.00

1.00
150.00

160.00

170.00

180.00

190.00

Peak Current, Amps

150.00

160.00

Peak Current, Amps

Fig. 5. Eect of pulsed current parameters on notch strength ratio.

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,

growth is arrested and remelting of the growing dendrites


can also occur. Current pulsing also results in periodic
variations in the arc forces and hence an additional uid
ows that lowers temperatures in front of the solidifying
interface. Furthermore, the temperature uctuations
inherent in pulsed welding lead to a continual change in
the weld pool size and shape favouring the growth of
new grains. It is also to be noted that eective heat input
for unit volume of the weld pool would be considerably
less in pulse current welds for which reason the average
weld pool temperatures are expected to be low [19,20].
Fig. 6 shows the optical micrographs of weld region of
some of the joints fabricated by pulsed current welding
technique.
It is important to note that while dendrite fragmentation has frequently been cited as a possible mechanism,
evidence for the same has not been hitherto established/

2090

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

demonstrated. It has been sometimes suggested that the


mechanism of dendrite break-up may not be eective in
welding because of the small size of the fusion welds and
the ne interdendrite spacing in the weld microstructure.
Thus grain renement observed in the PCTIG welds is
therefore believed to be due to other eects of pulsing on
the weld pool shape, uid ow and temperatures. The continual change in the weld pool shape is particularly important. As the direction of maximum thermal gradient at the
solidliquid interface changes continuously, newer grains
successively become favourably oriented. Thus, while each
grain grows only a small distance, more grains grow resulting in a ne grained structure [17].

The weld pool solidication during fusion welding


begins with the epitaxial growth of grains from partially
melted zone grains along the fusion boundary, at the
interface between the base metal and fusion zone. The
partially melted grains provide excellent sites for growth
with the growth rate exceeding the nucleation rate in this
region. Epitaxial growth across the weld pool results in
long and oriented columnar grains. The epitaxial grains
are the end result of continuing growth of the partially
melted grains from the fusion boundary. Epitaxial growth
requires that a minimal degree of undercooling prevail. In
contrast, the nucleation of new grains both at and near
the fusion boundary necessitates a free energy barrier to

Fig. 6. Micrographs of fusion zone region (100).

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

2091

Fig. 6 (continued )

be overcome. Consequently, no undercooling is necessary


for nucleation. To initiate nucleation in the weld deposit
and concurrently promote epitaxial grain renement, it
is essential to either increase the driving force, i.e., degree
of undercooling or reduce the free energy barrier by introducing trace amounts of zirconium or titanium to the aluminium weld pool [21].
5. Conclusions
The inuences of pulsed current parameters such as
peak current, base current, pulse frequency and pulse on
time on tensile properties of pulsed current TIG welded

AA 6061 aluminium alloy have been analysed in detail


and following conclusions have been obtained:
(i) Factorial experimentation technique is more convenient to predict the eect of pulsed current welding
parameters on tensile properties of TIG welded aluminium alloy joints. Analysis of variance method is
more appropriate to nd out the signicant main
and interaction factors of pulsed current TIG welding
process.
(ii) In general, peak current and pulse frequency are
having directly proportional relationship with the
tensile properties of the welded joints, i.e., if the

2092

T. Senthil Kumar et al. / Materials and Design 28 (2007) 20802092

peak current is increased, then the tensile strength is


increasing and similar eect is observed when frequency is increased.
(iii) However, base current and pulse on time is having
inversely proportional relationship with the tensile
strength, i.e., if the base current is raised, then the
tensile strength is decreasing and similar inuence is
noticed when pulse on time is increased.
References
[1] Simpson RP. Weld J 1977;56(3):67s72s.
[2] Garland JG. Metal Construct 1974;6(4):1218.
[3] Prasad Rao K. In: Proceedings of national conference on recent
advances in materials processing, Annamalai Nagar, India; 2001. p.
17696.
[4] Ravi Vishnu P. Weld World 1995;35(4):21420.
[5] Gokhale AA, Tzavaras DJ, Brody HD, Ecer GM. In: Proceedings of
conference on grain renement in casting and welds. St. Louis (MO):
TMS-AIME; 1982. p. 22347.
[6] Madhusudhan Reddy G, Gokhale AA, Prasad Rao K. J Mater Sci
1997;32(1993):411726.
[7] Yamamoto H. Weld Int 1993;7(6):45662.
[8] Madhusudhan Reddy G, Gokhale AA, Prasad Rao K. Optimization of pulse frequency in pulsed current gas tungsten arc welding
of aluminium lithium alloy sheets. J Mater Sci Technol 1998;14:
616.
[9] Ravisankar V, Balasubramanian V. Inuences of pulsed current
welding parameters on tensile and impact behaviour of AlMgSi
alloy weldments. In: Proceedings of international conference on
IMPLAST, New Delhi, India; 2003. p. 22432.

[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.

Vous aimerez peut-être aussi