You are on page 1of 8

See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/248251088

The influence of double pulse on porosity


formation in aluminum GMAW
ARTICLE in JOURNAL OF MATERIALS PROCESSING TECHNOLOGY FEBRUARY 2006
Impact Factor: 2.04 DOI: 10.1016/j.jmatprotec.2005.07.008

CITATIONS

DOWNLOADS

VIEWS

16

85

2 AUTHORS, INCLUDING:
Americo Scotti
Hgskolan Vst
79 PUBLICATIONS 189 CITATIONS
SEE PROFILE

Available from: Americo Scotti


Retrieved on: 09 August 2015

Journal of Materials Processing Technology 171 (2006) 366372

The influence of double pulse on porosity formation in aluminum GMAW


Celina Leal Mendes da Silva a , Americo Scotti b,
a

Department of Metallurgical Engineering, Federal Center for Technological Education of Para, Av. Almirante Barroso, 1155, 66093-020 Belem, Brazil
b Mechanical Engineering Faculty, Federal University of Uberlandia, Campus Santa Monica, 38400-902 Uberlandia, MG, Brazil
Received 18 February 2005; accepted 7 July 2005

Abstract
Pulsed GMAW (P-GMAW) has been recognized as an efficient alternative for minimizing porosity in aluminum welding. The double-pulsed
GMAW (DP-GMAW) technique is a variation of the pulsed GMAW technique, in which the pulsing current aimed to metal transfer control is
overlapped by a thermal pulsation, which in turn means pool control (similar to pulsed GTAW). Despite some advantages claimed by this reasonably
new technique, one could expect more porosity when compared with P-GMAW. Thus, in this work a comparative study on the performance of these
techniques at different parameter combinations and conditions favorable to generate porosity is presented. Porosity quantification was assessed by
gravimetric and radiographic procedures, while the pores were characterized by macrographic analysis. The results indicate that the DP-GMAW
technique maintains the capability of porosity minimization credited to the conventional P-GMAW technique and its advantages can be better
exploited in industrial applications.
2005 Elsevier B.V. All rights reserved.
Keywords: Welding; Pulsed GMAW; MIG/MAG; Aluminum; Porosity

1. Introduction
The choice of a specific welding process must take into
account some requirements, such as joint quality, levels of productivity and the costs involved in equipment operation and
acquisition. Unfortunately, the search for improvements on one
of these necessities may lead to the detriment of others. For
instance, trials on increasing productivity can lead to drop in
quality, and vice-versa. Thus, to better qualify the term, quality
would be the result of the capability of the process in adequating
to operational barriers, which are inherent to each situation. In
the case of aluminum welding, one remarkable barrier would be
the tendency for porosity generation.
It is widely recognized that hydrogen is the dominant cause
of porosity in aluminum alloy welds. Gases such as oxygen
and nitrogen present lower solubility in molten aluminum in
comparison to hydrogen [1] and, as a consequence, they usually play no important role in porosity generation. However,
in a micro gravity environment with vacuum, these gases can
assume importance. In a study on aluminum welding in this

Corresponding author. Tel.: +55 34 3222 8789; fax: +55 34 3239 4192.
E-mail addresses: clm.silva@cefetpa.br (C.L.M. da Silva),
ascotti@mecanica.ufu.br (A. Scotti).
0924-0136/$ see front matter 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmatprotec.2005.07.008

environment, Fujii et al. [2] carried out experiments with electron beam and GTAW processes and some unexpected results
emerged. These workers report that in all runs, including the
ones carried out in bead-on-plate condition, several pores were
observed. This porosity was linked to the formation of Al2 O
gas bubbles at high vacuum (the oxygen would come from the
superficial oxide layer).
One of the most comprehensive reviews on porosity in aluminum alloy welding was published two decades ago, by Devletian and Wood [3], and it can still be considered up-to-date.
These reviewers cite that the ratio between the maximum solubility of hydrogen in superheated molten and in the metal at the
solidification temperature is far higher for aluminum than for
any other structural materials, despite the disagreement among
researchers on solubility data. These authors believe that the
varying solubility data is the influence of alloy additions. They
also show several examples of the alloying elements influences
on the porosity concentration after welding. That means that
porosity in aluminum welds is very much related to the bead
chemical composition.
The most important source of hydrogen entering the arc column of welds includes: hydrogen contained with the filler metal
and parent metal; hydrogen bearing contamination or hydrated
oxide films on the surface of the filler and base metals; hydrogen or water vapor in the shielding gas [3]. Thus, cleanliness of

C.L.M. da Silva, A. Scotti / Journal of Materials Processing Technology 171 (2006) 366372

wire and workpiece surfaces is extremely important for reaching


aluminum sound welds.
However, the mechanisms of pores formation in aluminum
welding is based not only on entrapped hydrogen. Hooijmans
and Den Ouden [4] remark that the hydrogen concentration
in the weld pool at any moment is the result of two mutually
independent processes: inflow and outflow of hydrogen. That
is, porosity is a consequence of a dynamic equilibrium between
the rate of hydrogen absorption and rejection. And the control
of this equilibrium is not an easy or predictable task. That is
the reason why Devletian and Wood [3] state that size, shape,
distribution and amount of hydrogen pores generated in the weld
are generically dependent upon the solidification mode, cooling
rate, welding parameters, bead shape, shielding gas and external
pressure. And no conclusive results have been presented upon
the effect of, for instance, the welding parameters.
Porosity contents in welds can be measured by different
approaches, such as radiographic, gravimetric and microscopic
procedures. Morais [5] used the gravimetric method to evaluate the influence of different GMAW metal transfer modes on
porosity generation in an AA5052 alloy. He concluded that controlled metal transfer (pulsed GMAW) leads to lesser porosity
than welds carried out with the standard transfer modes. The
worst results were found with short-circuiting transfer.
Recently, a new welding technique was developed as a derivation of the pulsed GMAW (P-GMAW), in which a low frequency
current pulsation, typical of pulsed GTAW, is superimposed a
pulsed current for metal transfer control, characteristic of the
P-GMAW. This new technique has been named double pulsed
GMAW (DP-GMAW) or P-GMAW with thermal pulsation.
Commercially it has been referred to by different names, such
as Inpulse, Alu-Plus, etc. The DP-GMAW can be considered a
means of increasing productivity without quality drops, a fact
not always conceived, as shown in the first paragraph.
The principle of the most popular version of the DP-GMAW
is explained through the schematic oscillogram presented in
Fig. 1. Double pulsation is represented by the alternation of
low frequency periods of thermal base ( b ) and thermal pulse
( p ). During each of the thermal periods, there is a coexistence
of a current pulsation at higher frequency, in which the pulse
current and time (duration) (Ip and tp ) are kept the same and the

Fig. 1. Schematic oscillogram of current in the DP-GMAW technique.

367

base current and time (Ib and tb ) are adjusted as a function of the
desired average currents at each thermal period. The variation
of the mean current between b and p (Im b and Im p , respectively) is synchronized with a wire feed speed variation (WFS b
and WFS p , respectively), so that the arc length is maintained
constant. There is a resultant weighted mean current (Imt ) and a
weighted mean wire feed speed (WFSt ).
The role of a low frequency current pulsation (thermal pulsation) is to control the weld pool. While the heat input is higher
(thermal pulse), there is an effective fusion of the parent metal.
On the other hand, when the heat input is lowered (thermal base),
surface tension and viscosity of the molten metal get higher,
making it more difficult for the pool to melt through. There
must be equilibrium between the b and p duration so that the
volume and shape of the weld bead do not get patchy [6].
However, the consequent cyclic variation of the arc dimensions at low frequency can, in theory, mimic arc instabilities,
breaking the harmony of the original technique (P-GMAW)
and making DP-GMAW more prone to porosity. Thus, it is
worthwhile to exploit better the DP-GMAW technique, especially concerning its ability for generating pores, before claiming
the advantages of this new technique as a productivity and
quality booster. This work aimed to carry out a comparative
study between the techniques P-GMAW (considered appropriate
concerning porosity generation) and DP-GMAW (claimed for
higher productivity), in relation to porosity on AA5052 welding.
2. Materials and methods
As the objective of this work is to study comparatively the ability of the
two techniques for generating pores, conditions more prone to this defect were
looked for methodologically, yet not in excess; either too little or too much
porosity could disguise the effect of the techniques. However, at the same time,
the welds should be close to real conditions, in order to transport the results
of this study to workshops. According to Devletian and Wood [3], welding in
overhead position is more prone to porosity. Thus, a special gadget for holding
the test-plates (item 3 in Fig. 2) was fixed to an automated XY coordinate table
(item 2, with high precision, for guarantying repeatability). This coordinate
table is commanded by a computational program, which synchronizes the torch
movements with the trigger of a multi-process electronic power supply (item 1).
The delimited positions (at the striking and extinction of the arc) and the travel
speed (TS), both set in the program, defined the length of each weld bead. This
was necessary for making sure that all beads were exactly the same length, since
aluminum has low robustness in relation to thermal stability.
In order to measure the values of the wire feed speed, which must change
quickly with time in the DP-GMAW technique, an encoder connected to the drive
shaft of the feeder rolls was used. The output electrical signal, corresponding to
the imposed wire feed speed, was conditioned in a dedicate electronic interface.
A data acquisition system, working at 10 kHz per channel and at 12 bits, was
used for acquiring the instantaneous values of the wire feed speed, as well as
the arc voltage and welding current.
A push-pull gun (to prevent wire feeding problems) was used, inclined 100
forward in a pushing direction. A 1.0-mm-diameter AWS ER4043 wire was
employed, shielded by argon at 15 l/min. The weldments were carried out with
DCEP, with the electronic power supply set either for pulsed GMAW (P-GMAW)
or for double pulsed GMAW (DP-GMAW). For the pulse parameter settings, a
pulse current (Ip ) of 200 A and a pulse duration (tp ) of 2.2 ms were chosen so that
the condition of one drop per pulse (ODPP) would be achieved for the given wire
and shielding gas [7]. The certification of ODPP condition was made through
laser shadowgraphy and high-speed filming [8]. Possible short-circuits were
minimized by programming the power supply with a short-circuiting voltage
and current of, respectively, 7 V and 400 A.

368

C.L.M. da Silva, A. Scotti / Journal of Materials Processing Technology 171 (2006) 366372

Fig. 2. Representation of the experimental rig.

Fig. 3. Test-plate used in the experiments: (1) dimensions (mm); (2) sectioning; (3) geometric parameters.
The experimental runs were planned to be accomplished in two steps, in
which the second one would be a validation of the first step results. All test-plates
were made of an aluminum-magnesium alloy of the class AA5052-F, measuring
250 mm 50 mm 6.35 mm, as illustrated in Fig. 3, part 1. The plates were
previously cleaned with water and soap (no organic solvent). In order to intensify
even further porosity generation, a 60 -included-angle groove, with 1-mm-wide
root, was longitudinally machined in the plates.
In the first step, the test-plates were welded using two runs (root and capping),
for which the groove was prepared with a depth (Pc) of 5 mm and contact tipto-groove bottom distances (CTWD) of 15 and 17 mm were used, respectively,
for the root and capping runs. In the second step, some welding conditions were
altered so as to enlarge the achievements of the result validation. Thus, only
one run was deposited to fill the groove, in which Pc was reduced to 3 mm and
CTWD was kept at 15 mm.
In each experimental step, three levels of average current were used (higher
reliability). However, to filter the effect of different heat inputs on porosity
generation, the same volume of deposited material per run was used. To be more
precise, heat input should be referred to as HI = UI/TS, where U, arc voltage; I,
welding current; TS, travel speed; , thermal efficiency. Nevertheless, as thermal
efficiency is challenging to determine, the authors of this work believe that a
constant volume of deposited material per run would be an adequate parameter
for comparison in different welding conditions. To reach such a condition, a ratio
between wire feed speed (WFS) and travel speed (TS) was maintained equal to a

constant (K). The values of K for the two experimental steps were, respectively,
equal to 18.5 and 20 (K was calculated so that the groove would be filled with
the right amount of material).
Three replications were used for each combination of welding conditions
(WCond) and welding techniques (WTech), totaling 18 runs for each experimental step. The setting parameters for both P-GMAW and DP-GMAW are
presented in Tables 1 and 2, respectively, for each experimental step.
After the weldments, the test-plates were transversely sectioned into four
segments, as illustrated in Fig. 3, part 2. The extremes (P1 e P4) were discharged.
Measurements of bead penetration (P), total area (At , i.e., the melted area), metal
deposited area (Ad ) and penetrated area (Ap ) were carried out over the transverse
sections a, b and c (which were metalographicaly prepared, etched by
Tuckers). From these measurements, the calculation of dilution (Dil), amongst
others, was possible. All these parameters are illustrated in Fig. 3, part 3. It
must be pointed out that for step 1 (two runs) the penetrated area (Ap ) represents
the penetrated area of both runs and the dilution refers to the two runs as it
were only one. Using a caliper, bead reinforcement (R) and width (W) were
measured over the beads on segment 3. From these measurements, the calculation
of the geometric indexes R/W (which gives a notion of the bead convexity) and
W/PR (which indicates the bead slenderness), where PR = P + R, was achievable.
The remaining geometric parameters were used in the precision confirmation
of the methodology and in one of the methods to quantify porosity, as seen
later.

Table 1
Setting parameters used during Step 1
WTech

P-GMAW

Parameters

WCond1 (Imt /TS)

DP-GMAW ( p = 0.6 s e b = 0.4 s)


WCond2 (Imt /TS)

WCond3 (Imt /TS)

Periods
Imt
Im p e Im b (A)
Ib (A)
tb (ms)
WFSt (m/min)
WFSp and WFSb (m/min)
TS (mm/min)

86
30
4.6
5.4

94
34
4
6

WCond1 (Imt /TS)

WCond2 (Imt /TS)

Wcond3 (Imt /TS)

110
44
3.1

70
22
6.1

110
44
3.1

90
32
4.2

6.5

4.5

6.7

5.7

102
39
3.5
6.4

86
50
20
11.2
5.4
3.6

297

327

349

94

297

102

5.9
320

p
110
44
3.1
6.3
6.7
344

C.L.M. da Silva, A. Scotti / Journal of Materials Processing Technology 171 (2006) 366372

369

Table 2
Setting parameters used during Step 2
WTech

P-GMAW

Parameters

Wcond1 (Imt /TS)

DP-GMAW ( p = b = 0.5 s)
WCond2 (Imt /TS)

WCond3 (Imt /TS)

Periods
Imt
Im p e Im b (A)
Ib (A)
tb (ms)
WFSt (m/min)
WFS p and WFS b (m/min)
TS (m/min)

80

90

27
5.1
5.3

32
4.3
5.7

WCond1 (Imt /TS)

WCond2 (Imt /TS)

Wcond3 (Imt /TS)

110
44
3.1

70
6.1
4.5

110
44
3.1

90
32
4.2

6.8

4.6

6.9

5.7

100

80
50
20
11.2

38
3.6
6.4

90

5.2
3.4

264

284

320

Segment P2 had the bead reinforcement machined off and its surface, after
metalographic preparation, together with the transverse section a, were analyzed under a microscope to characterize the pores. The pore quantification was
carried out by two methods (radiography and gravimetry) upon segment 3 (P3).
The radiographies were visualized through a negatoscope, aided by a magnifying glass, allowing the counting of the number of pores (NP). Then, the same
segment was machined for preparing the sample for the gravimetric analysis.
The gravimetric method was developed for estimating the volume of voids (Vv )
in solid materials (see, for instance, Saperstein et al. [9]), based on differential
weighting of the samples, first in air, then in water, to calculate the specific
density. More details of the application of both methods in this work can be
found in an other publication [10], in which systematic studies for verifying the
influence of the radiographic plane (from either the top or the side of the bead)
and the fabrication sequence of the sample for gravimetry (taking out either
the bead or the parent metal) on the results were accomplished. The robustness
of each method was also assessed. The results suggest that the position from
where the radiography is taken does not influence the amount of observed pores.
They also suggest that gravimetry is more robust than radiography (confirming
Saperstein), which does not discredit the latter.
All obtained results (porosity and geometric related) were submitted to variance analysis (ANOVA).

3. Results and discussion


3.1. Step 1
Table 3 presents the monitored mean values of current and
voltage and wire feed speed. It is also presented the values of
the parameter Imt /TS calculated from the monitored current and
the adjusted travel speed (considering the precision of the gun
manipulator, the TS as set can be considered correct). The resulting values of Imt /TS suggest that the heat input was practically
the same for all welds. The arc voltage around 24 V for the first

p
100

5,8

259

110
44
3.1
6.3

289

6.8
313

run and around 25 V for the capping run depict constancy in the
arc length, despite the different sets for Imt and WFS.
In Table 4 are shown the measured geometric parameters. As
the values of the deposited area (Ad ) are close to each other, one
can say that the approach of making WFSt /TS (=18.5) constant
has worked out. On the other hand, the values of the penetrated
area (Ap ) became different, despite the fact that Ad was the same.
This means that the experimental methodology was good enough
for distinguishing the influence of welding parameters on the
geometry and, if some difference between the two techniques
has to happen, there is no reason why it should not.
The influence of the welding technique (WTech) on porosity
generation is illustrated in Fig. 4, where a tendency for higher
porosity from the DP-GMAW technique can be noticed when the
analysis is made by the radiographic results (average number of
poresNP ), while no tendency is withdrawn when the analysis
is made by the gravimetric data (average volume of voidsVv ).
All welds contained pores, but according to the ASME SEC VIII
standard, all welds would be approved. Then, the objective of
having porosity, but not in excess, was achieved.
It is important to point out that the standard deviations were
higher for the radiographic results, i.e., the precision of the gravimetric procedure is higher. The statistic ANOVA was applied
on the data and do not point to any significant influence of
the welding techniques on porosity, either from radiographic
or gravimetric data.
However, as said before, one could expect more porosity in
DP-GMAW than in P-GMAW (H2 inflow). As the technique
showed not to influence porosity, it was considered important
to assess the behavior of the geometric parameters, in order to

Table 3
Monitored and calculated parameters during the experimental Step 1
WTech/WCond

First Run (filling)

Second run (capping)

Monitored

DP-GMAW/WCond1
DP-GMAW/WCond2
DP-GMAW/WCond3
P-GMAW/WCond1
P-GMAW/WCond2
P-GMAW/WCond3

Calculated

Monitored

Calculated

Imt (A)

U (V)

WFSt (m/min)

Imt /TS (A min/m)

Imt (A)

U (V)

WFSt (m/min)

Imt /TS (A min/m)

88
97
104
88
96
104

24.4
24.5
24.4
24.0
23.8
24.5

5.6
6.0
6.4
5.5
6.2
6.6

296
303
302
296
293
298

87
97
103
87
95
102

25.4
25.1
24.6
24.5
25.4
25.5

5.6
6.0
6.4
5.5
6.2
6.6

293
303
299
297
291
292

C.L.M. da Silva, A. Scotti / Journal of Materials Processing Technology 171 (2006) 366372

370

Table 4
Average values and standard deviations (%) of the geometric parameters (Step 1)
WTech/Wcond

Adm (mm2 )

% Ad

Apm (mm2 )

% Ap

R/W

% R/W

W/PR

% W/PR

DP-GMAW/WCond1
DP-GMAW/WCond2
DP-GMAW/WCond3
P-GMAW/WCond1
P-GMAW/WCond2
P-GMAW/WCond3

32
31
30
31
29
30

2.7
3.1
5.3
6.9
1.5
4.0

3.8
7.1
4.0
2.2
3.2
4.8

38.1
18.8
47.0
74.8
73.3
30.1

0.18
0.17
0.20
0.25
0.24
0.21

17.5
15.1
10.6
20.0
8.8
12.6

1.55
1.66
1.44
1.30
1.27
1.46

6
6
2
7
9
8

turbulent molten metal flow, fact that would help gas escape too.
Another point that would make the gas entrapment lower in DPGMA welds is the breakage of the columnar growths (which
happens between the transition between the thermal pulse and
base), reducing the grain size and splitting the pores between
them.
Fig. 5 illustrates the pore (rounded) disposition in the welds
made by the two techniques. One can observe larger pores at
the central line of the bead when P-GMAW was used (Fig. 5left). In Fig. 5-right, the pores, in smaller sizes, are disposed at
the solidification lines created by the mentioned breakage of the
columnar growths.
3.2. Step 2
Fig. 4. The relationship between the welding conditions and porosity for PGMAW and DP-GMAW during STEP 1, using data from both radiography and
gravimetry (TS is dependant of the technique for a given Imt see Table 1).

understand better the outcomes so far (H2 outflow). The higher


R/W, which represents the bead convexity, the more difficult it
is for the gas to escape (more porosity). In general, P-GMAW
shows higher mean values of this geometric index. However,
the ANOVA indicated that only at WCond2 the technique has
significance on R/W.
Concerning the index W/PR, which indicates the bead slenderness, one could expect that the more slender the bead (lower
ratio W/PR), the more difficult for the gas to escape. The W/PR
values (Table 4) suggest that the DP-GMAW technique should
present less difficulty for pore trapping. The ANOVA shows that
this is not significant only at WCond3.
Thus, if on one side the DP-GMAW technique should absorb
more gas due to the cyclic variation of the arc shape, on the
other side its resulting bead geometry seems to favor the gas
discharge. In addition, the double pulsation might generate a

Some welding conditions were changed to the experimental


step 2. Firstly, the thermal period set was altered (from p / b
equal to 0.6/0.4 s to 0.5/0.5 s), what lead to a lower average current for any parameter combination. According to Woods [11],
even though referring to GTAW, lower current reduces porosity
in aluminum alloy welding, since, as much the pool volume, as
H2 absorption due to external sources, lessen. Just one run was
used to prevent the effect of pore migration from the first run
to the second. Finally, the ration WFSt /TS was increased to 20,
since the groove was shallower than for the STEP 1 conditions.
Table 5 presents the monitored mean values of current and
voltage and wire feed speed, as much as the calculated parameter Imt /TS. Similarly to Step 1, the Imt /TS values suggest that
the heat input was approximately the same for all runs. Again,
the regularity of the arc voltages (around 24 V) indicates an
arc length constancy. Table 6 presents the measured geometric
parameters.
The influence of the technique on porosity generation is better
visualized in Fig. 6. One can notice that the technique P-GMAW
showed to be slightly less susceptible to porosity when quantified

Fig. 5. Pores disposition in welds: (left) P-GMAW e (right) DP-GMAW.

C.L.M. da Silva, A. Scotti / Journal of Materials Processing Technology 171 (2006) 366372
Table 5
Monitored and calculated parameters during the experimental Step 2
WTech/WCond

DP-GMAW/WCond1
DP-GMAW/WCond2
DP-GMAW/WCond3
P-GMAW/WCond1
P-GMAW/WCond2
P-GMAW/WCond3

Monitored

Calculated

Imt (A)

U (V)

WFSt
(m/min)

Imt /TS
(A min/m)

81
92
101
82
91
102

24.1
24.3
24.6
24.2
24.7
24.5

5.4
5.7
6.2
5.3
5.8
6.5

296
306
322
311
320
319

Fig. 6. The relationship between the welding conditions and porosity for PGMAW and DP-GMAW during STEP 2, using data from both radiography and
gravimetry (TS is dependant of the technique for a given Imt see Table 2).

by radiography. However, the radiographic methods presented


higher standard deviations, ratifying the finding in the previous step. Similarly to STEP 1, all welds contain pores, yet
approved by the ASME SEC VIII standard. This re-confirms
having reached the experimental purpose of having pores in all
welds at an adequate rate. However, by comparing Vv % presented in Figs. 4 and 6, one can notice that the volume of voids
increased in STEP 2 in comparison to STEP 1.
Once more, the outcome analysis starts with the verification
of whether or not the methodologies premises had been fulfilled,
that is, whether or not the deposited areas would be the same for
all conditions, so that the comparison between the techniques
and the welding conditions could be realized. According to the
Ad values in Table 6, one can say the premises were satisfied.

371

The variance analysis was again applied on the porosity


and geometric indexes data. Likewise STEP 1, only in one
condition (WCond2), and based on the radiographic analysis,
the technique present significant influence on porosity formation, suggesting that the DP-GMAW technique is more prone
to pores. However, assuming the higher data dispersion of the
radiographic method, and based on the same results obtained
with the gravimetric method during STEP 1 and STEP 2,
it can be inferred that there is no significant influence of
the welding technique on porosity generation (or, at most, a
slight tendency favoring more porosity with DP-GMAW). These
results disagree with those of Matsui and Shionoya [12]. It is
important to mention that the pulsation frequencies employed
by these workers were much higher than those used in this
work.
Thus, the suspicion that the wire feed speed cyclic variations that take place in the DP-GMAW technique would lead
to arc instability (oscillation of the arc geometry), generating
more porosity, seems again not to be confirmed. It is important
to point out that this possible double action (gas is more easily absorbed and released at the same time) was already used
by Zakharov and Yushchenko [13] to justify the same level of
porosity with welding current increases. For a second time, it
was considered important to assess the influence of both techniques on the geometric parameters, namely R/W and W/PR
indexes.
Observing Table 6, one can presume that DP-GMAW would
have a slight tendency for presenting a lower R/W ratio and
a higher W/PR ratio than P-GMAW, expecting less gas trapping with the former technique. That may be one strong reason for the DP-GMAW technique to have produced no higher
pores levels. In addition, only spherical pores were found in
the weld bead transverse sections, regardless of the technique
(Fig. 7). However, the resulting pores from the technique PGMAW were found mostly in the central region (Fig. 7a), while
the pores from the DP-GMAW technique were more scattered
(Fig. 7b).
Independently of the technique, some pores were also
observed on the interface parent metal/deposited metal, as illustrated in Fig. 8; aluminum alloys have high thermal conductivities, consequently their solidification occurs rapidly from the
parent metal towards the weld pool centerline, trapping pores.
Naturally, one can suppose that in some cases, a very unstable
arc, for instance, the action of gas absorption can overcome the
gas release action, and the DP-GMAW technique would present
more porosity than the P-GMAW technique.

Table 6
Average values and standard deviations (%) of the geometric parameters (Step 2)
WTech/WCond

Adm (mm2 )

% Ad

Apm (mm2 )

% Ap

R/W

% R/W

W/PR

% W/PR

DP-GMAW/WCond1
DP-GMAW/WCond2
DP-GMAW/WCond3
P-GMAW/WCond1
P-GMAW/WCond2
P-GMAW/WCond3

15.1
15.8
15.6
15.8
15.5
15.7

1.3
3.7
6.2
3.6
7.4
1.7

3
4.3
5.0
1.6
4.4
5.5

29.1
5.9
18.7
25.8
25.8
21.6

0.12
0.15
0.14
0.20
0.17
0.17

5.5
6.2
9.5
14.9
2.2
4.7

2.3
2.0
1.9
1.7
1.7
1.7

2.5
7.2
4.8
9.8
7.8
2.7

372

C.L.M. da Silva, A. Scotti / Journal of Materials Processing Technology 171 (2006) 366372

Fig. 7. Transverse section of two STEP 2 samples, welded with Imt 80 A: (a) P-GMAW: (b) DP-GMAW.

and White Martins Gases Industriais, for providing the parent


metal and shielding gas, respectively.
References

Fig. 8. Interface parent metal/deposited metal in welds carried out with the PGMAW at Imt 80 A.

4. Conclusions
In the conditions established in this work, it can be deduced
that:
The proposed methodology for comparing two welding techniques in relation to porosity generation in aluminum, based
on the correspondence of the thermal conditions and on the
promotion of pores at a regular amount, is adequate;
The double pulsed GMAW (DP-GMAW) technique, in spite
of having theoretically higher potential for porosity generation, does not increase the porosity susceptibility in aluminum welding, when compared with the pulsed GMAW
technique;
However, welding equipment that does not provide fast
response during the wire feed speed transitions between thermal pulse and base and/or a parameter setting that does not
guaranty the same arc length at the thermal pulses and bases,
may not reproduce the same findings.
Acknowledgements
The authors would like to acknowledge the importance of the
PADCTIII research program, though the CEMAT 620094/97-4
project, due to its financial support. They are grateful to ALCAN

[1] R.J. Shore, R.B. McCauley, Effects of porosity on high strength aluminum 7039, Welding J. 49 (July (10)) (1970) 311s321s.
[2] H. Fujii, Y. Aoki, K. Nogi, Electron Beam, Gas tungsten arc welding
under microgravity, Trans. JWRI 30 (1) (2001) 105109.
[3] J.H. Devletian, W.E. Wood, Factors affecting porosity in aluminum
weldsa review, Welding Research Council Bulletin 290, USA, December 1983, p. 20.
[4] J.W. Hooijmans, G. Den Ouden, A model of hydrogen absorption during
GTA welding, Welding J. 76 (7) (1997) 264s268s.
[5] F.C. Morais, Influence of operational factors on porosity formation in
aluminum MIG welding, MSc. Thesis, Federal University of Uberlandia,
Brazil, June 2001 (in Portuguese).
[6] C.L. Mendes da Silva, A. Scotti, Assesment of the thermal pulsation periods on the weld bead formation in aluminum welding by the
DP-GMAW technique, in: I Brazilian Congress of Manufacturing Engineering COBEF 2001, ABCM, Brazil, April, 2001 (in Portuguese).
[7] L.O. Vilarinho, A. Scotti, An alternative algorithm for synergic pulsed
GMAW of aluminum, Austr. Welding J. 45 (2) (2000) 3644.
[8] P.S.S. Balsamo, L.O. Vilarinho, M. Vilela, A. Scotti, Development of
an experimental technique for studying metal transfer in welding: synchronised shadowgraphy, Int. J. Joining Mater. 12 (2) (2000) 4859.
[9] Z.P. Saperstein, G. Prescott, G.E.W. Monrea, Porosity in aluminum
welds, Welding J. 43 (October (10)) (1964) 443s453s.
[10] C.L. Mendes da Silva, Evaluation of the thermal pulsation technique
in aluminum welding, P.hD. Thesis, Federal University of Uberlandia,
Brazil, January 2003 (in Portuguese).
[11] R.A. Woods, Porosity and hydrogen absorption in aluminum welds,
Welding J. 53 (March (3)) (1974) 97s108s.
[12] H. Matsui, S. Shionoy, Pulsed arc welding apparatus having a consumable electrode wire, USA Patent no.: 5.473.139, Toyota Jidosha
Kabushiki Kaisha, Toyota, Japan, 1993.
[13] W.S. Zakharov, K.A. Yushchenko, Effects of the baking temperature
of electrodes on the porosity in welds in high-alloys steels, Automatic
Welding (April) (1984) 5356.
Dr. Celina Leal Mendes da Silva is a lecturer at Federal Center for Technological Education of Para, Brazil. She is graduated as mechanical engineer
and received her Dr. Eng. degree in 2003 at Federal University of Uberlandia,
Brazil.
Dr. Americo Scotti is a professor in the Mechanical Engineering Faculty
at Federal University of Uberlandia, Brazil. He is Mech. Eng., M.Sc. and
received his P.hD. degree at Cranfield University (UK) on welding technology.
He has been lecturing, researching and consulting on welding for 23 years.
He also has 2 years of industrial experience with WMGI, a Brazilian branch
of Praxair Inc. His current research interests are related to the application of
novel arc welding technologies into industrial fields.

You might also like