Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:26 PM Page 120
WELDING RESEARCH
Cold Metal Transfer (CMT) GMAW
of Zinc­Coated Steel
Welding conditions for quality welding were obtained based on understanding the porosity
formation mechanism and mechanical strength evaluation
BY M D. R. U. AHSAN, Y. R. KIM, R. ASHIRI, Y. J. CHO, C. JEONG, AND Y. D. PARK
ABSTRACT
This paper investigates the feasibility of cold metal transfer (CMT) gas metal arc
(GMA) welding of zinc­coated steel in a lap­joint configuration with no root opening over
a wide range of heat inputs from 150 to 550 J/mm in terms of porosity formation and
mechanical performance. The quantitative analysis of porosity by x­ray radiographic test­
ing shows the heat input conditions below 250 J/mm and above 350 J/mm cause the
least amount of porosity. The temperature­dependent properties of the weld pool vis­
cosity and vapor pressure of zinc determine the porosity formation, growth, and escape
mechanisms at different heat inputs. These factors were characterized by using high­
speed imaging analysis and infrared (IR) thermography of the weld pool. Secondly, to en­
sure joint integrity, samples welded at different heat input conditions were mechanically
tested to identify the minimum required penetration depth leading to base metal failure.
It was found that a minimum heat input of 200 J/mm is required for ensuring that mini­
mum penetration. Finally, the heat input criteria achieved from the porosity analysis and
mechanical testing were combined and two sets of optimized welding conditions — one
set is in the low heat inputs, ranging from 200 to 250 J/mm, and another set is in the
high heat inputs, ranging from 350 to 550 J/mm — were derived. The guideline proposed
by this work can be used for selecting suitable welding conditions based on the require­
ments, which can ensure low porosity in the weldments and adequate mechanical
strength of the joints.
KEYWORDS
• Zinc­Coated Steels • Cold Metal Transfer (CMT) • Gas Metal Arc Welding (GMAW)
• Porosity • Mechanical Properties
Introduction
In order to improve fuel efficiency
and reduce CO2 emissions by reducing
vehicle weight, without compromising
safety, leading automakers are extensively using thinner sheets of highstrength steels (Ref. 1). These steels
are often coated with zinc-based coatings as they protect the steel compo-
nents from aqueous corrosion by offering barrier protection and galvanic
protection (Ref. 2).
Gas metal arc welding (GMAW) is a
commonly used method for joining
plastically deformed and closed structural components for the automotive
industry (Refs. 3, 4). In order to
achieve high productivity and avoid
melt-through in thin sheet materials,
welding conditions with fast welding
speeds ensuring low heat input are required. Cold metal transfer (CMT) is a
recently developed automated GMAW
process associating short circuit transfer as the droplet transfer method
(Ref. 5). The CMT process engages
with the mechanically controlled retraction of the wire to assist the
droplet transfer during short circuiting without aid from the electromagnetic force, which allows reducing the
current without affecting the material
transfer and thus creates a lower heat
input for the same amount of deposition compared to other GMAW
processes (Ref. 6). The controlled
method of material deposition and
higher melting coefficient compared
to conventional welding processes
make CMT a suitable choice for joining
steel sheet materials with a higher
productivity (Ref. 7).
However, the melting temperature
of steels is about 1500°C, which is significantly higher than the boiling temperature of zinc (906°C) (Ref. 8). This
situation causes formation of highpressure zinc vapor at the interface of
zinc-coated sheets during welding
(Ref. 9). If this high-pressure zinc vapor is not vented out properly, it will
be trapped in the weld pool and form
weld discontinuities in the form of
porosities (Refs. 10–13), which is severe, especially in the lap-joint weld
geometry with no root opening (Ref.
MD. R. U. AHSAN and Y. D. PARK (corresponding author, ypark@deu.ac.kr) are, respectively, student and professor of the Dept. of Advanced Materials Engi­
neering, Dong­Eui University, Busan, Korea. Y. R. KIM is with the Hyundai Motor Co., Gyeonggi­do, Korea. R. ASHIRI is a visiting researcher with the Department
of Advanced Materials Engineering, Dong­Eui University, Busan, Korea, invited from the Dept. of Materials Engineering, Isfahan University of Technology, Is­
fahan, Iran. Y. J. CHO is with the Ship Infrared Signature Research Center, Dept. of Naval Architecture and Ocean Engineering, Dong­Eui University, Busan,
Korea. C. JEONG is with the Dept. of Mechanical Engineering, Stevens Institute of Technology, Castle Point on Hudson, Hoboken, N.J.
120-s WELDING JOURNAL / APRIL 2016, VOL. 95
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 121
WELDING RESEARCH
Fig. 1 — A — Stress vs. time curves at different heat input condi­
tions; B — image from x­ray RT test of the respective weldments
with the exact location of tensile test sample on weld bead marked.
1). The presence of porosities in the
weld bead severely affects the joint integrity in terms of both strength and
fatigue performance (Ref. 14). Existing works show entrapment of zinc vapor in the weld pool can be avoided if
the solidification is delayed enough to
allow time for the zinc vapor to escape
from the weld pool (Ref. 15). This requires high heat input. However, high
heat input in fusion welding processes
gives rise to the chance of meltthrough, hence it is undesirable (Refs.
12, 13, 16).
Existing works focusing on mitigating porosity discontinuities in the
weld bead include removal of the zinc
coating as the simplest approach as
suggested by AWS
(Ref. 17). Other researchers also looked
for an alternate to
zinc as a coating material, such as nickel, which has a higher boiling point and hence does not
evaporate during welding (Refs. 18,
19). The porosity issue has also been
dealt with by using metal cored wires
(Refs. 13, 16), flux cored wires (Ref.
20), or changing the composition of
solid wires (Refs. 21, 22). Significant
efforts are also made by modifying the
process parameters in order to attain a
better control over the process and increase the heat input to reduce porosities in the weld bead (Refs. 20, 21).
In the works of Matsui et al. (Refs.
23, 24), the effect of the root opening
between specimens, welding position,
tilt angle of the torch, contact tip-toworkpiece distance (CTWD), and base
plate angle on porosity formation were
extensively studied. The effects of
Table 1 — Welding Conditions Experienced for Understanding the Porosity Formation Behavior in CMT­GMAW
19.7
18.5
18.4
18.3
18.2
17.6
16.9
16.2
15.5
14.9
14.6
13.9
13.1
12.7
12.4
40
50
60
Constant Speed
263
252
241
231
220
211
200
190
180
170
161
150
141
130
120
Voltage
(V)
Welding Speed (cm/min)
70
80
90
100
110
120
Melt-Through
Current
(A)
Constant Current and Voltage
Incomplete Fusion
APRIL 2016 / WELDING JOURNAL 121-s
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 122
WELDING RESEARCH
A
B
Fig. 2 — Porosity analysis data as function of the following: A — Welding current (with corresponding voltage) at a constant welding
speed [60 cm/min (15.75 in./min)]; B — welding speed at a constant welding current and voltage (200 A and 16.9 V).
welding speed and welding voltage on
porosity formation were also studied
in their work. However, these specific
studies have been conducted in the
bead-on-plate condition, which is far
from a real weld used in the automotive industry.
Although extensive research has
been conducted on fusion welding of
zinc-coated steels, all these prior studies focused on mitigating porosity discontinuities without any in-depth
study to disclose the mechanisms and
phenomena involved in the porosity
formation in relation to a wide range
of welding parameters. Moreover,
none of the existing works involve the
CMT process. In this work, a systematic study is performed on porosity formation, growth, and escape phenomena in CMT-GMA welding of zinc-coated steels over a wide range of welding
parameters (welding current, voltage,
and welding speed) consisting of
low, medium, and high-heat-input
conditions.
The weld pool behavior during
welding and the quality of welds depend largely on weld pool temperature. Infrared (IR) thermography using
an infrared camera can detect the IR
energy emitted from the surface and
convert it into a two-dimensional
image of temperature distribution
showing real-time weld pool temperature during the process (Ref. 25). Insitu study of the weld pool using highspeed camera and infrared camera
were performed in order to support
the experimental data and to explain
the mechanisms involved in porosity
formation, growth, and escape phenomena in CMT-GMA welding of zinccoated steel. With the understanding
of the mechanism, a set of welding
conditions were derived that ensure
the least possible amount of porosity.
In order to obtain welds with reliable mechanical responses (static and
dynamic) in the arc welding process,
adequate penetration and a discontinuity-free weld are needed. Therefore,
CMT-GMA weld joints were also put
through a tensile test to determine a
critical penetration depth to ensure
adequate mechanical properties and
sort out a range of welding conditions
that ensure adequate strength of the
weld joint. Finally, the outcome of the
study on porosity and the mechanical
properties were combined to yield optimized welding conditions with adequate penetration and less porosity,
ensuring the integrity of the welds.
Experimental Procedure
Experiments on porosity were conducted using hot-rolled steel sheets of
440 MPa ultimate tensile strength and
2.3 mm thickness with a galvannealed
coating thickness of 10 μm. The samples were welded using ER 70S-3 solid
welding wires 1.2 mm in diameter. As
single lap-joint geometry is the most
susceptible one to porosity formation,
this geometry with a 25-mm overlap
was selected for the current study.
Throughout the study, a constant welding length of 100 mm was used. As was
found in the study by Matsui et al.
Table 2 — Tensile Test Results Showing the Strength and Failure Modes of the Joints with Different Penetrations
Heat Input (J/mm)
Penetration (mm)
Sample 1
152
167
185
203
223
0.01
0.10
0.20
0.28
0.38
122-s WELDING JOURNAL / APRIL 2016, VOL. 95
331
382
337
447
451
UTS (MPa)
Sample 2
Sample 3
311
292
359
443
455
313
376
384
441
445
Sample 1
Mode1
Mode 1
Mode 1
Mode 2
Mode 2
Failure Mode
Sample 2
Mode 1
Mode 1
Mode 1
Mode 2
Mode 2
Sample 3
Mode 1
Mode 1
Mode 1
Mode 2
Mode 2
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 123
WELDING RESEARCH
Fig. 3 — Porosity analysis data as a function of heat input for
CMT­GMAW (Ref. 28).
Fig. 4 — Porosity analysis data as a function of heat input for the stan­
dard short circuit GMAW process.
B
A
C
Fig. 5 — Analysis at high­heat­input condition: A — Frames
from high­speed camera movie showing porosity escape phe­
nomenon; B — weld pool temperature distribution obtained
by infrared camera; C — schematic illustrating the steps of
zinc vapor bubble formation, growth, and escape; D — x­ray
RT film and bead cross section of the sample welded at a high­
heat­input condition.
(Refs. 23, 24), a root opening of 0.2 mm
between the sheets in lap-joint geometry can provide a passage for high-pressure zinc vapor to escape and thus can
significantly affect the amount of
porosity in the weld bead. Therefore,
specimens were spot welded prior to the
experiments in order to ensure zero
root opening and avoid ambiguity of
data. The drag technique with a drag angle of 80 deg and torch angle of 60 deg
was used throughout the study.
In this current work, wide ranges of
welding parameters, including welding
current, welding voltage, and welding
speed, were considered to study porosity formation mechanism and behav-
ior. The CMT process works based on a
synergic program, allowing the user to
only modify wire feed speed. The machine itself selects the suitable welding
current and voltage based on the synergic program. In this work, wire feed
speed was used ranging from 3 to 11
m/min, which varied the welding current and welding voltage in the ranges
of 120 to 263 A and 12 to 20 V,
respectively. The welding speeds used
ranged from 40 to 120 centimeters per
minute (cm/min) (15.75–47.25
in./min) with an interval of 10
cm/min (4 in./min).
To quantify the amount of porosity
in the weld bead, welded samples were
D
put through x-ray radiographic testing
(RT) with a Rigaku 250EGS industrial
x-ray system. X-ray RT results were obtained in films, which were digitalized
for the analysis of porosity. Commercially available image analysis software, Image-Pro Plus®, was used for
quantification of porosity.
For the quantification of porosities
in weldments, previous researchers
have followed two approaches: The
first approach involves measuring the
summation of the area of all the
porosities in the x-ray RT image and
quantifying it as the percentage of the
total area of the weld bead in the x-ray
RT image (Ref. 14). The second apAPRIL 2016 / WELDING JOURNAL 123-s
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 124
WELDING RESEARCH
B
C
A
Fig. 6 — Analysis at medium­heat­input condition: A — Frames from
high­speed camera movie showing porosity escape phenomenon; B
— weld pool temperature distribution obtained by infrared camera;
C — schematic illustrating the steps of zinc vapor bubble formation,
growth, and escape; D — x­ray RT film and bead cross section of the
sample welded at a medium­heat­input condition.
proach considers counting the number
of porosities from the x-ray RT results
(Refs. 21, 23, 24). As the latter approach does not consider the size of
porosities, in this study the first approach was followed and porosities
were quantified as area percentage.
However, x-ray RT results were obtained in two-dimensional plane and
can reveal the size of porosities without disclosing their positions. The
welded samples were cross sectioned,
polished, and etched with 2% picral
solution. The subsequent macroscopic
analyses were performed with an
Olympus SZ61 macroscope at 10
magnification to find out the location
for porosity formation.
Moreover, a high-speed imaging
technique was engaged to study the
weld pool surface to visualize the zinc
vapor escape phenomenon and its frequency. A collimated monochromatic
LED light source was used to illuminate the weld pool for high-speed imaging at 4000 frames per second with
a Photron MC2-type high-speed video
camera. The high-speed camera was
set up parallel to the welding gun focusing on the weld pool right behind
the arc. To study the weld pool temperature at different heat input conditions, a FLIR A655sc infrared camera
was used. Data from infrared imaging
were used for
comparing the
differences in
weld pool temperature at different welding
conditions.
For the study
of mechanical
properties, hotrolled steel sheets of 440 MPa ultimate
tensile strength and 2.3 mm thickness
were used. As the presence of porosity
can lead to irregular and fluctuating
results in the tensile test and prevent
the selection of critical condition, uncoated samples were used for the tensile test with the
KSU-10M tensile testing machine.
Other welding conditions were kept
constant.
Results and Discussion
Porosity in GMAW of Zinc­
Coated Steel
Effect of Porosity on the Mechanical
Properties
In a sound welding condition, the
mechanical properties of weld joints of
zinc-coated steels usually match those
124-s WELDING JOURNAL / APRIL 2016, VOL. 95
D
of uncoated steels (Ref. 27). Data from
tensile tests performed on weld joints
at different welding conditions are
presented in Fig. 1. As the ultimate
tensile strength (UTS) of base metal
and welding wire are 440 and 540
MPa, respectively, failure is more likely
to occur at base metal with an UTS of
about 440 MPa. However, the stress
vs. strain curves produced from the
tensile test data presented in Fig. 1A
show a scattered behavior and failures
were found in weld zones as well as
base metal for different welding conditions. X-ray RT analysis of samples
presented in Fig. 1B shows that the
sample, which failed through the weld
zone at a lower stress, has a large
amount of porosity in the weld bead.
The presence of weld discontinuities
like porosity degrades the weld quality
as the cracks initiate along the edge of
the pores (Ref. 1).
Moreover, the stress vs. strain
curves presented in Fig. 1A show a sig-
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 125
WELDING RESEARCH
B
C
A
Fig. 7 — Analysis at low­heat­input condition: A — Frames
from high­speed camera movie showing porosity escape phe­
nomenon; B — weld pool temperature distribution obtained by
infrared camera; C — schematic diagram illustrating the steps
of zinc vapor bubble formation, growth, and escape; D — x­ray
RT film and bead cross section of the sample welded at a low­
heat­input condition.
nificant difference in the failure strain
of samples with and without porosities in the weld bead. The sample with
a large amount of porosity had a very
low failure strain (2.82%) compared to
the weld joints with little or no porosity (17.72 and 19.92%, respectively).
Hence, the presence of porosity in the
weld bead greatly reduced the energy
absorption capacity of the joint, which
is a significant requirement for ensuring vehicle safety. On a similar study
conducted by Joaquin et al. (Ref. 14)
on the mechanical properties of
GMAW lap-joint geometry of zinccoated steel, fluctuating results in the
tensile strength were observed due to
the presence of porosity in weldments.
Porosity Formation Behavior in
GMAW
To understand the porosity formation mechanism and behavior in CMTGMAW of zinc-coated steels, a wide
range of welding conditions were studied as presented in Table 1. The CMT
process functions based on a synergic
program, which allows the user to only
change wire feed speed, and the CMT
machine chooses the suitable current
and voltage. Therefore, in this work, at
the fixed welding speed, wire feed
speed was changed in a way that the
D
current changed by 10 A (from 120 to
263 A), letting the machine choose the
respective voltage. In addition, at the
fixed welding current and voltage,
welding speed was varied at an interval of 10 cm/min (4 in./min) in the
range of 40 to 120 cm/min (15.75 to
47.25 in./min).
Samples prepared according to the
experiment plan presented in Table 1
were put through an x-ray RT test for
the purpose of porosity evaluation.
The amount of porosity in area percent as function of welding current,
voltage, and welding speed are presented in the graphs in Fig. 2. All the
graphs in Fig. 2 have similar trends for
the porosity amount. At the lower
ends, the amount of porosity is low,
which then shows a rapid increment in
the medium ranges of the parameters
and again decreases at the higher
ends. In order to compare both data
sets as a function of one parameter,
the calculated heat inputs were used.
The amount of heat input was calculated using Rosenthal’s simplified approach. When the amount of porosity
is plotted as a function of heat input,
both data sets yield an identical trend:
having a lesser amount of porosity in
low and high heat inputs and a relatively higher amount of porosity in the
medium-heat-input range, as shown in
Fig. 3 (Ref. 28).
According to the ISO standard for
porosity, the critical amount of porosity is equal to 2% of the total bead area
in weld joints (Ref. 29). The porosity
analysis data in Fig. 3 indicates the
amount of porosity in the weldment
was lower than the critical value in low
heat inputs ranging from 150 to 250
J/mm. As the heat input rose beyond
250 J/mm, the amount of porosity
rapidly increased and reached its maximum at around 300 J/mm heat input.
Further increment in the heat input
showed a decreasing trend of porosity,
which fell below the critical once again
beyond the heat input of 350 J/mm.
Based on the described trend of porosity accumulation in the weld bead, the
whole heat input range was divided
into three distinguished sections as
low, medium, and high heat input for
the present study. Based on this, welding conditions resulting in heat inputs
less than 250 J/mm were considered
as low-heat-input conditions. Heat inputs between 250 and 350 J/mm,
which are susceptible to porosity accumulation, were marked as mediumheat-input conditions. Heat inputs beyond 350 J/mm were considered highheat-input conditions. The lesser
amount of porosity at high heat inputs
attributed to the slow solidification
APRIL 2016 / WELDING JOURNAL 125-s
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 126
WELDING RESEARCH
A
B
Fig. 8 — A — Amount of porosity as function of heat input (verification of the proposed mechanism); B — contour plotting for choosing
suitable CMT­GMAW conditions with less porosity.
A
B
Fig. 9 — A — Three­dimensional schematic illustrating the prepa­
ration of the tensile test specimens; B — bead cross sections at
different heat inputs with different penetration depths.
rate, which provided a longer time for
the zinc vapor to escape from the weld
pool compared to other heat inputs
(Refs. 12, 15).
As explained by Ribic et al. (Ref.
15), the thermophysical properties of
the material, power density of the
heat source, and the heat input per
unit length govern the solidification
rate of the weld pool. As the material
and heat source were fixed throughout this study, heat input per unit
length was the factor determining the
solidification rate. At high heat inputs, the solidification rate is slow
enough to allow the zinc vapor to escape the weld pool, causing less
porosity. This phenomenon is supported by other published research
regarding the porosity issue in fusion
welding of zinc-coated steels (Refs.
13, 16, 30). As the heat input decreases up to the medium range, the
amount of porosity starts to rise,
which is also supported by the previous works
(Refs. 12, 15, 20).
However, all the
previous works
have used the
terms “high heat
input” and “low
heat input” without indicating the
exact range in actual values (Refs. 12,
13, 20). Beside, the weldable heat input range changes depending on the
materials’ thermophysical properties
and thickness for the same welding
process. Hence, in order to compare
CMT-GMAW process with other
processes and to explain less porosity
in low heat inputs, a better understanding on the weldable heat input
126-s WELDING JOURNAL / APRIL 2016, VOL. 95
range of conventional short-circuit
process and porosity formation behavior over that range are required.
To achieve that understanding, a
set of samples were welded using the
same shielding gas over a wide range
of heat inputs, which showed that at
heat inputs less than approximately
250 J/mm, conventional GMAW cannot produce adequate deposition to
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 127
WELDING RESEARCH
A
B
Porosity
Formation Mech­
anisms at Differ­
ent Heat Input
Conditions
As zinc has a
boiling temperature that is much
lower than the
melting temperature of iron, the
zinc coating present on the surface
of specimens
evaporates when
exposed to the
C
arc. In the lap
joint configuration, a high-pressure zinc vapor
Fig. 10 — A — Typical failure mode at penetration < 10% of base
metal thickness; B — typical failure mode at penetration >10% of
starts to form at
base metal thickness; C — tensile strength as a function of pene­
the faying intertration depth.
face of the specimens (Ref. 21).
The amount of
create a sound weld joint. Weldability
zinc vapor formation depends on the
of the CMT process at heat inputs less
extent of the region over the boiling
than 250 J/mm is the result of the
point of zinc, which is proportional to
controlled method of metal transfer
the amount of heat input (Ref. 1). In
and its higher melting efficiency comthe absence of a root opening in the
pared to conventional arc welding
lap-joint configuration, zinc vapor has
processes, as explained by Pickin et
only one way to be vented out and that
al. (Ref. 7). The obtained conventionis through the weld pool (Refs. 1, 21).
al GMA welded samples were put
The attempts made by Masaharu Sato
through an x-ray RT test and the reto identify critical locations of zinc
sults of porosity evaluation were plotcoating for porosity formation by placted as the function of heat input in
ing a thin zinc leaf on different locathe graph shown in Fig. 4. Within the
tions of the lap-joint geometry to simweldable range starting from approxiulate the zinc coating (Ref. 31) conmately 250 J/mm, the amount of
cluded that the zinc present at the fayporosity is similarly high as in CMTing interface of the specimens is reGMA welded samples up to 350
sponsible for porosity formation in
J/mm, which decrease sharply at the
weldments of zinc-coated steels. Once
high heat inputs beyond 350 J/mm.
the zinc vapor enters the weld pool, it
takes the form of a bubble. This bubble contains a high-pressure zinc vapor, which lets it grow in size and then
rise up due to buoyancy. Like all
gaseous matter, the vapor pressure of
zinc also has strong temperature dependency and shows higher vapor
pressure at higher temperatures,
which is governed by the ClausiusClapeyron equation (Refs. 8, 9).
Weld pool temperature depends on
the amount of heat input causing a
higher temperature weld pool at higher heat inputs (Ref. 26). The maximum
temperature of CMT-GMAW was
measured within 1650° to 2700°C with
different heat input conditions, which
is close to the measurement made by
Kozakov et al. (Ref. 32). From this
measured temperature range, the approximate range of vapor pressure can
be calculated within 63 to 125 MPa.
While vapor pressure affects the
growth and rising up of the zinc vapor
bubble by assisting in the process, viscosity of molten metal affects the
process by acting as the resisting force
(Ref. 21). The viscosity of molten metal is also temperature dependent.
The Arrhenius equation shows this
dependency as follows:
Eμ
μ = μo
e RT
(1)
where  is the viscosity at the studied
temperature, ois the viscosity at the
reference temperature, E is the temperature coefficient of viscosity, R is
the universal gas constant, and T is
the study temperature in Kelvin (Ref.
33). From the stated equation, it can
be said that the viscosity of molten
APRIL 2016 / WELDING JOURNAL 127-s
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 128
WELDING RESEARCH
A
B
Fig. 11 — A — Penetration depth as function of heat input; B — contour plot for choosing a suitable CMT­GMA welding condition ensuring
adequate penetration.
metal decreases with the increase in
temperature. The weld pool temperature depends on the welding parameters or heat input (Ref. 33). Within the
measured range of weld pool temperature, the calculated viscosity using the
provided equation varies from 4.00 
10–3 to 1.62  10–3 kg/ms. Therefore,
at higher heat inputs, weld pool temperature is higher, resulting in a less
viscous weld pool and vice versa. The
zinc vapor bubble’s growth and rise
due to vapor pressure and buoyancy
depends on the viscosity and the time
required to reach solidification temperature of the weld pool (Ref. 15).
The size and location of pores are the
results of the competition of viscosity
against buoyancy and vapor pressure
within the time required to reach solidification temperature.
Based on this concept, schematic
diagrams of zinc vapor behavior with
respect to different heat input conditions are presented in Figs. 5–7 along
with corresponding x-ray RT results,
bead cross sections, and in-situ study
evidence from high-speed camera (to
study the zinc vapor bubble escape behavior) and infrared camera (to measure the weld pool temperature) in order to explain the porosity formation
and growth behaviors in CMT-GMAW
of the studied zinc-coated steel at different heat input conditions.
The in-situ high-speed imaging
study of the weld pool gives out the
zinc vapor escape phenomenon from
the weld pool. Frames from the study
conducted at a high-heat-input condi-
tion are presented in Fig. 5A, which
shows the zinc vapor escape phenomena from the large-sized weld pool. In
the high-heat-input conditions, the
frequency of zinc vapor escaping from
the weld pool was counted to be about
25 times per second. The temperature
measurement of the weld pool surface
using the infrared camera (Fig. 5B)
shows the peak temperature of the
weld pool was about 2700°C and the
arc and the region just behind the arc
were at maximum temperature. A
schematic representation of zinc vapor
formation, growth, and escape mechanism is presented in Fig. 5C. The weld
pool temperature was highest in the
high inputs within the experimental
range. This induced a less viscous weld
pool with a viscosity of about 1.62 
10–3 kg/ms, less than the half of the
viscosity in the low-heat-input cases
(4.00  10–3 kg/ms). Moreover, as calculated, the vapor pressure of zinc in
high heat inputs was about 125 MPa,
which is also about twice the vapor
pressure at low-heat-input conditions.
Because of high vapor pressure inside
the zinc vapor bubble in a less viscous
weld pool, the zinc vapor bubbles rapidly grew in size and rose up faster. Beside, due to the high temperature of
the weld pool, it needed a longer time
to reach its solidification temperature.
Due to these facts, even though zinc
vapor bubbles cross a longer distance
to reach the surface, they can escape
the weld pool before solidification.
This is claimed because the high-speed
camera recorded the highest zinc va-
128-s WELDING JOURNAL / APRIL 2016, VOL. 95
por bubble escape frequency at this
condition. As a result, a porosity-free
weld bead can be obtained at high inputs as evident from the x-ray RT image and bead cross section shown in
Fig. 5D.
Frames from a similar high-speed
camera study of the weld pool with
medium-heat-input conditions are
presented in Fig. 6A. From the highspeed camera frames, zinc vapor escape behavior is also evident in medium heat inputs, although the zinc vapor escape location is far from the arcing location compared to the high heat
inputs. The zinc vapor escape initiation (at 3.5 ms) was also delayed compared to the high-heat-input case (at
3.0 ms) indicating a longer time was
required for the zinc vapor bubble to
rise to the surface. Moreover, the frequency of zinc vapor escaping was
much lower than that of the highheat-input condition (5 times per second).
The weld pool surface temperature
measurement data obtained by the IR
camera in Fig. 6B shows the highest
temperature was about 2100°C, which
is significantly lower than the highheat-input case. Hence, the weld pool
is more viscous with an approximate
viscosity of 2.6  10–3 kg/ms. At this
weld pool temperature, the vapor pressure of zinc was calculated to be about
91 MPa, significantly lower than the
high-heat-input case. A zinc vapor
bubble with lower vapor pressure in a
more viscous weld pool takes longer to
rise to the weld pool surface and es-
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 129
WELDING RESEARCH
Fig. 12 — Weldability envelope diagram presenting different zones in the whole experi­
mental range.
cape. This explains the delayed escape
of the zinc vapor bubble from the weld
pool at a location far behind the arc
compared to that of the high-heatinput case.
Based on this evidence, the mechanisms for zinc vapor formation and
growth in medium-heat-input conditions are explained in the schematic
representation of Fig. 6C. In medium
heat inputs, the zinc vapor bubble formation starts in a similar manner as in
high heat inputs. However, due to the
significantly lower weld pool temperature compared to the high-heat-input
condition, the viscosity of the weld
pool was higher and the vapor pressure was lower than the high-heatinput case as calculated. Therefore, the
zinc vapor bubble faced more resistance and needed a longer time to reach
the weld pool surface, even though it
had to cross a shorter distance due to a
lower amount of deposition compared
to high-heat-input conditions. Hence,
the zinc vapor bubble could rise up
and grow in size; however, the weld
pool temperature was not high enough
to delay the solidification initiation
until the zinc vapor bubble reached
the surface and completely vented out
from the weld pool and the molten
pool again filled the void the escaped
zinc vapor left behind. When solidification starts before the zinc vapor
bubble reaches the surface, it produces
discontinuities in the form of internal
porosities (Ref. 21), which are not
seen from the surface.
As a result, large-sized porosities
(about 750 m in diameter) are evident in the x-ray RT images in Fig. 6D.
However, some irregular escaping of
zinc vapor bubbles was observed in
high-speed camera analysis. In such
cases, if the solidification starts before
the weld pool closes down, there will
be pits (porosities open to the surface)
(Ref. 21). The bead cross sections in
Fig. 6D show the presence of large
porosity right below the surface in the
first one and the presence of pit in the
second one, as explained.
The heat input range specified as
low-heat-input conditions in the current work was unexplored in earlier
studies (Refs. 9–14) as the standard
short-circuit arc welding process is unable to produce adequate deposition
required for making a weld joint possible at this heat input range. The porosity evaluation results show (Fig. 3)
that the amount of porosity is less
than the critical value in this range.
However, the high-speed camera study
of the weld pool presented in Fig. 7A
showed no porosity escape phenomenon in this condition. Infrared temperature measurement of the weld
pool surface reveals its maximum
temperature is about 1650°C, which is
relatively close to the solidification
temperature.
The infrared image in Fig. 7B also
shows a small region over the solidification temperature of molten steel,
and images from the high-speed camera in Fig. 7A also show a smaller-sized
weld pool at this heat input condition.
Yet the lower amount of porosity is
the subject of the small size of porosity as evident in x-ray RT image of Fig.
7D.
The bead cross section in Fig. 7D
reveals the porosity formation location is near the root of the weld bead,
far away from the weld bead surface.
The mechanisms involved in porosity
formation in low heat inputs are explained with a schematic in Fig. 7C. As
illustrated in the schematic, in the low
heat inputs, the zinc vapor bubble formation initiates in a similar manner as
in the high- and medium-heat-input
conditions. However, due to low heat
input, the amount of zinc evaporation
is lower. The weld pool temperature is
lower than in the medium- and highheat-input conditions. The low temperature of the weld pool causes less
vapor pressure inside the zinc vapor
bubble, which is about 62 MPa, about
half of the vapor pressure exerted at
the high-heat-input conditions. Beside, at the low-temperature weld
pool, the viscosity of the molten metal
(approximately 4.00  10–3 kg/ms) is
more than twice the viscosity at high
heat inputs. The lower vapor pressure
of zinc in a high-viscosity weld pool results in slower growth of the zinc vapor bubble. The small-sized bubbles
need a longer time to rise to the surface due to weaker buoyancy than the
larger-size zinc vapor bubbles formed
in medium- and high-heat-input conditions and higher viscosity of the
weld pool (Ref. 34).
Moreover, the low temperature of
the weld pool also leads to its early solidification. As a result of these, smallsized porosities with an average diameter of 250 m are formed near the
root of the weld bead as evident in the
bead cross section shown in Fig. 7D.
APRIL 2016 / WELDING JOURNAL 129-s
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 130
WELDING RESEARCH
Beside, as the porosities cannot rise to
the surface, no porosity escape phenomena are observed at these heat inputs. In the quantitative characterization of porosity in laser welding of
stainless steel performed by Madison
et al. (Ref. 35), a similar behavior in
porosity formation location and size
was observed. In that work, smallsized pores were evident at a location
far away from the weld bead surface.
As the heat input gradually increases,
larger-sized pores start to appear at a
gradually closer position to the weld
bead surface.
Optimized Welding Conditions for
Porosity
As discussed in the previous section, the heat inputs ranging from 250
to 350 J/mm, which are marked as
medium heat inputs, are susceptible to
porosity formation. The low-heatinput (below 250 J/mm) and highheat-input (over 350 J/mm) conditions have significantly less porosity.
To perform a validation test on this
range, all welding conditions yielding
heat inputs within 250 to 350 J/mm
along with one condition in the highheat-input range and one condition in
the low-heat-input range at each welding speed within the range were evaluated for porosity. The porosity evaluation results were accumulated in the
graph in Fig. 8A. The graph shows the
porosity evaluation data agree well
with the proposed mechanism and behavior discussed in the previous section. All the porosity evaluation data
of welds performed at heat inputs
within 250 to 350 J/mm show a
porosity amount of more than 2% of
the total bead area, whereas other conditions (low- and high-heat-input conditions) always show an amount of
porosity less than the critical amount
(2% of bead area).
Based on this validation study, it
can be concluded the proposed mechanisms are sound. However, the synergic program in the CMT-GMAW
process increases the amount of deposition along with the welding current
and voltage. As a result, the welding
conditions with high heat inputs can
lead to overdesigning the weld joint,
exceeding the actual requirements. An
independent control over current and
voltage apart from the wire feed rate
as found in double-electrode GMAW
(Refs. 36-42) would have allowed
changing the heat input without increasing the amount of deposition,
hence avoiding overdesigning the weld
joint.
With the porosity evaluation data, a
contour mapping was prepared and is
presented in Fig. 8B. By following the
color coding in the contour plot, welding conditions susceptible to porosity
formation can be identified. Moreover,
using this diagram as a guideline, suitable welding conditions causing less
porosity can be selected based on the
requirement. In the contour map, the
welding conditions falling from the
yellow to red color regions must be
avoided due to the high amount of
porosity, whereas the ash-colored regions do not have weldability due either to incomplete fusion or meltthough. Welding conditions in the ashcolored region of the low heat inputs
do not produce adequate deposition to
produce a weld joint. On the other
hand, the welding conditions in the
ash-colored region at the high-heatinput side produce too high a heat input and cause melt-through. The welding conditions leading to any other
color in the map will produce a weld
joint with porosity less than 2% of total bead area and are hence acceptable.
Mechanical Properties
In the previous section, a range of
optimized welding conditions was derived based on the porosity analysis results. However, a weld joint must ensure adequate mechanical strength as
well. To achieve a guideline for acceptable welding conditions for adequate
mechanical strength, tensile testing was
performed in the current work. As
shown in Fig. 1, tensile testing with
zinc-coated samples can lead to ambiguous results due to the presence of porosity, hence prohibiting researchers from
deriving any practical useful guideline.
To overcome this difficulty, tensile tests
were performed using uncoated materials with the same mechanical properties
and composition. For the tensile
strength test, tensile samples were prepared with the dimensions presented in
Fig. 9A, according to the ASTM E8M-03
standard (Ref. 43).
A set of samples with different heat
inputs and hence different leg lengths
130-s WELDING JOURNAL / APRIL 2016, VOL. 95
and penetration depths as shown in
Fig. 9B were put through a tensileshear test at a constant testing speed
of 5 mm/min. The samples at all the
welding conditions had a leg length
more than 90% of base metal thickness, which is the minimum leg length
requirement as specified by the AWS
D8.8 standard (Ref. 44). For each
welding condition, three tensile samples were machined from the same
weldment.
The results from the tensile-shear
tests are presented in Table 2, which
shows two types of failure mode. The
first type of failure mode is referred to
as “mode 1,” which is shown in Fig.
10A. The penetration depth was less
and under tensile loading, the weld
bead separated from the lower sheet.
As the leg length fulfilled the minimum requirement, this kind of failure
was attributed to inadequate joint
penetration (Refs. 7, 45). Beside, as
mentioned by Lawrence and Cox (Ref.
46), a significant drop in the tensile
strength and fracture strain is seen in
the samples with inadequate penetration. Moreover, in CMT-GMAW, the
leg length and penetration increase simultaneously as the amount of deposition increases with the amount of
heat input. Hence, in this work, penetration depth was considered as the
critical parameter for a weld to ensure
adequate mechanical strength. From
the data in Table 2, it is evident that
“mode 1” is the usual failure mode at
penetration depths less than 0.20 mm.
In the case of the samples with
higher penetration depths, the failure
location moves toward the base metal
as a shown in Fig. 10B. In this work,
failure at the base metal was classified
as “mode 2.” The tensile test data from
Table 2 are plotted as a function of
penetration depth in order to relate
the tensile strength with penetration
depth in the graph of Fig. 10C. The
graph shows gradual increments in the
tensile strength with increasing penetration depth until the penetration
depth is less than 0.23 mm. However,
as the penetration depth increases beyond 0.23 mm, which is 10% of the
base metal thickness, the ultimate tensile strength stops increasing as the
failure location shifts toward the base
metal. Hence, the penetration depth
equal to 10% of the base metal thickness was considered the minimum re-
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 131
WELDING RESEARCH
quirement to ensure adequate mechanical strength in a CMT-GMA weld
joint. A similar guideline (10% of base
metal thickness) as minimum requirement for penetration depth was also
used by O. Schwedler et al. (Ref. 47) to
set acceptance criterion for sound
welding in CMT-GMAW.
Based on the criteria set for minimum penetration depth ensuring adequate mechanical strength, a range of
bead cross sections were checked for
penetration depth. In the graph presented in Fig. 11A, penetration depth
was plotted as a function of heat input, which shows an increasing trend
in penetration depth with heat input
and a heat input of 200 J/mm was
considered to be the critical value to
yield a penetration greater than 10%
of base metal thickness. To verify this
condition, all the welding conditions
right above 200 J/mm within the total
welding range shown in Table 1 were
checked for penetration depth. The results of the verification study is accumulated in the contour plot in Fig.
11B. The contour plot shows the
whole weldable range that has been divided into three distinguished zones
with black lines: welding conditions
that had inadequate heat input to produce sufficient penetration, welding
conditions with adequate penetration,
and extremely high heat inputs that
will lead to melt-through. Using this
contour plot, suitable welding conditions can be chosen to ensure minimum adequate penetration and,
hence, joint strength.
Optimized Welding Conditions
From the study conducted on
porosity formation, growth, and escape mechanisms in CMT-GMAW of
zinc-coated steels, heat inputs below
250 J/mm (low heat inputs) or above
350 J/mm (high heat inputs) had the
amount of porosity less than the critical amount of 2% of total bead area.
However, not all the welding conditions in these ranges might be suitable
from the mechanical performance
standpoint due to inadequate joint
strength. To ensure joint efficiency
and optimized welding conditions,
tensile tests on samples with different
heat input conditions, and hence, different penetration depths, were carried out. The tensile testing results
show a minimum penetration equivalent to 10% of base metal thickness is
enough to ensure mechanical efficiency (see Fig. 10C). The relationship between heat input and penetration
depth shows that heat inputs equal to
or above 200 J/mm always yields a
penetration equal or higher than the
set critical condition (10% of base
metal thickness).
The conditions derived from the individual studies performed on porosity
and mechanical porosity evaluation
are plotted in contours in Figs. 8B and
11B. Both contours show optimum
weldable regions considering either
porosity or mechanical joint efficiency.
Both of the mentioned contours are
plotted within the same range of welding parameters, hence, their optimum
weldable ranges are super positioned
to derive the common weldable region,
which is presented as the weldability
envelope diagram shown in Fig. 12.
The envelope diagram shows two regions, one ranging from 200 to 250
J/mm heat input, falling into the lowheat-input conditions (lower enclosed
section by green line). The other region ranges from 350 to 550 J/mm,
falling into the high-heat-input conditions (upper enclosed section by green
line). Both regions ensure adequate
penetration depth to ensure joint efficiency and a less-than-critical amount
of porosity, which are recommended
for next experiences.
Conclusions
This paper explains a study on the
porosity formation, growth, and escape behavior, and the mechanical
properties of weld joints in CMTGMAW over a wide range of heat inputs (150–550 J/mm). At the three
distinctive heat input regions, [low
(below 250 J/mm), medium (250–350
J/mm), and high (over 350 J/mm)],
the maximum weld pool temperatures
were found to be approximately 1650°,
2100°, and 2700°C.
In this work, porosity formation,
growth, and escape mechanisms were
explained using the viscosity of the
weld pool and the vapor pressure of
zinc. With the low heat inputs, due to
the lower temperature of the weld pool
within the range, low vapor pressure
of zinc, and high viscosity of the weld
pool, the zinc vapor bubble could not
grow and rise. As a result, small-sized
pores were formed near the weld root
and the total amount of porosity was
less. With the high heat inputs, the
high temperature of the weld pool assisted the zinc vapor bubble to grow in
size, rise up, and escape by creating
the least viscous weld pool, maintaining high vapor pressure of zinc, and
delaying the solidification. With the
medium heat inputs, the weld pool
temperature was proved not to be high
enough to delay solidification until the
zinc, vapor could escape. However, it
was high enough to maintain a low
viscosity and high vapor pressure of
zinc to let the zinc vapor bubble grow
and rise. The combination of these factors led to a high amount of porosity
accumulation in the weldment in this
range.
The mechanical properties evaluation with a tensile test showed penetration depth equivalent to 10% base
metal thickness was enough to provide
adequate joint strength. The bead
cross-section analysis at different heat
inputs showed a minimum heat input
of 200 J/mm is required to achieve
penetration above 10% of base metal
thickness.
By combining experimental data
from porosity and mechanical properties evaluation, two distinctive set of
optimized welding conditions were derived: one at the low heat input ranging
from 200 to 250 J/mm and the other at
the high heat inputs, starting from 350
J/mm and rising up to 550 J/mm.
References
1. Yang, S., Wang, J., Carlson, B. E., and
Zhang, J. 2013. Vacuum-assisted laser
welding of zinc-coated steels in a gap-free
lap joint configuration. Welding Journal
92(7): 197-s to 204-s.
2. Marder, A. R. 2000. The metallurgy
of zinc-coated steel. Progress in Material
Science 45: 191–271.
3. Ramazani, A., Mukherjee, K., Abdurakhmanov, A., Prahl, U., Schleser, M.,
Reisgen, U., and Bleck, W. 2014. Micromacro-characterization and modeling of
mechanical properties of gas metal arc
welded (GMAW) DP600 steel. Material Science & Engineering A 589: 1–14.
4. Ramazani, A., Li, Y., Mukherjee, K.,
Prahl, U., Bleck, W., Abdurakhmanov, A.,
Schleser, M., and Reisgen, U. 2013. Microstructure evolution simulation in hot
rolled DP600 steel during gas metal arc
APRIL 2016 / WELDING JOURNAL 131-s
Ahsan Paper 201590 April 2016_Layout 1 3/10/16 4:27 PM Page 132
WELDING RESEARCH
welding. Computational Materials Science
68: 107–116.
5. Pickin, C. G., and Young, K. 2006.
Evaluation of cold metal transfer (CMT)
process for welding aluminum alloy. Science
and Technology of Welding and Joining 11(5):
583–565.
6. Zhang, H. T., Feng, J. C., He, P.,
Zhang, B. B., Chen, J. M., and Wang, L.
2009. The arc characteristics and metal
transfer behavior of cold metal transfer
and its use in joining aluminum to zinccoated steel. Material Science & Engineering
A 499: 111–113.
7. Pickin, C. G., Williams, S. W., and
Lunt, M. 2011. Characterization of the
cold metal transfer (CMT) process and its
application for low dilution cladding. Journal of Material Processing Technology 211:
496–502.
8. Yang, S. L., and Kovacevic, R. 2009.
Welding of galvanized dual-phase 980 steel
in a gap-free lap joint configuration. Welding Journal 88(8): 168-s to 178-s.
9. Tzeng, Y.-F. 1999. Pulsed Nd:YAG
laser seam welding of zinc-coated steel.
Welding Journal 78(7): 238-s to 244-s.
10. Pieters, R. R. G. M., and Richardson, I. M. 2005. Laser welding of zinc-coated steel in overlap configuration with zero
gap. Science and Technology of Welding and
Joining 10(4): 142–144.
11. Jeric, A., Grabec, I., and Govekar, E.
2009. Laser droplet welding of zinc-coated
steel sheets. Science and Technology of Welding and Joining 14(4): 662–668.
12. Fiore, S., Barhorst, S., and King, T.
H. 2014. A new filler metal for producing
high-speed welds in thin-gauge coated
steels with minimal porosity. Proc. Conf.
Sheet Metal Welding Conference XVI, Livonia, Mich., American Welding Society (Detroit Section), Paper 4B-1.
13. Narayanan, B. K., Henry, J., Liao, Y.
C., and Galiher, D. 2014. Solutions for
welding zinc-coated steels. Proc. Conf. Sheet
Metal Welding Conference XVI, Livonia,
Mich., American Welding Society (Detroit
Section), Paper 4B-2.
14. Joaquin, A., Elliott, A. N. A., Jiang,
C., Rajan, V., Hartman, D., and Karas, C.
2007. Gas metal arc welding of coated advanced high strength steel (AHSS) — Developments for improved weld quality. SAE Int.
15. Ribic, B., Palmer, T. A., and DebRoy
T. 2009. Problems and issues in laser-arc
hybrid welding. International Materials Review 54: 223–244.
16. Keegan, J. M., Nagarajan, S., Daemen, R. A., and Bundy, J. 1999. Metal-core
weld wire for welding galvanized steels.
U.S. Patent No. 5857141 A.
17. AWS WZC/D19.0-72, Welding ZincCoated Steels. 1972. Miami, Fla.: American
Welding Society.
18. Pennington, E. J. 1987. Laser welding of galvanized steel. U.S. Patent No.
4642446.
19. Williams, S. W., Salter, P. L., Scott,
G., and Harris, S. J. 1993. New welding
process for galvanized steel, ftwc. 26th Int.
Symp. Automotive Technology and Automation. Aachen, Germany, pp. 49–56.
20. Jud, R. W. 1984. Joining of galvanized and galvannealed steels. SAE International.
21. Izutani, S., Yamazaki, K., and Suzuki, R. 2013. New welding process, J-Solution™ Zn, suitable for galvanized steel in
the automotive industry. KOBELCO Technol. Review, p. 32.
22. Kodama, S., Ishida, Y., Furusako, S.,
Saito, M., Miyazaki, Y., and Nose, T. 2013.
Arc Welding Technology for Automotive Steel
Sheets. Nippon steel technical report no.
103.
23. Matsui, H., Suzuki, H., and Yamada,
M. 1997. Reduction of blowholes in high
speed arc welding of hot dipping galvanized steel sheets. Journal of Japan Welding
Society, pp. 476–483.
24. Matsui, H., and Yamada, M. 1997.
Fundamental examination in arc welding
procedure of hot-dip zinc-coated carbon
steel sheets. Journal of Japan Welding Society, pp. 484–493.
25. Frappier, R., Benoit, A., Paillard, P.,
Baudin, T., Le Gall, R., and Dupuy, T. 2014.
Quantitative infrared analysis of welding
processes: Temperature measurement during RSW and CMT-MIG welding. Science
and Technology of Welding and Joining.
19(1): 38–43.
26. Alfaro, S. C. A., Vargas, J. A. R., Carvalho, G. C. D., Gonc, G., and Souza, A. D.
2015. Characterization of humping in the
GTA welding process using infrared images. Journal of Materials Processing Technology 223: 216–224.
27. Gregory, E.N. 1971. The mechanical
properties of welds in zinc coated steel.
1971. Welding Journal 50(10): 445-s to
450-s.
28. Ahsan, Md. R. U., Kim, Y. R., Kim,
C. H., Kim, J. W., and Park, Y. D. Porosity
formation mechanisms in cold metal transfer (CMT) gas metal arc welding (GMAW)
of zinc-coated steels. Science and Technology
of Welding and Joining.
29. International Standard: ISO 5817.
30. Baardsen, E. L. 1976. Method of
welding galvanized steel. U.S. Patent No.
3969604 A.
31. Sato, M. 2015. Serial Lecture: Easy
description of weld arc phenomena for
welders, Part 4: Control of phenomena associated with consumable electrodes and
weld pool (2). Journal of Welding for AsiaPacific 19(2): 9–11.
32. Kozakov, R., Schopp, H., Gott, G.,
Sperl, A., Wilhelmand, G., and Uhrlandt, D.
2013. Weld pool temperatures of steel
S235 while applying a controlled short-circuit gas metal arc welding process and vari-
132-s WELDING JOURNAL / APRIL 2016, VOL. 95
ous shielding gases. Journal of Physics D:
Applied Physics 46(47): 1–12.
33. Lancaster, J. F. 1986. The Physics of
Welding, 2nd edition, Chapter 2, International Institute of Welding, Pergamon
Press, pp. 40, 41.
34. Zhao, H., and DebRoy, T. 2001. Pore
formation during laser beam welding of
die-cast magnesium Alloy AM60B —
mechanism and remedy. Welding Journal
80(8): 204-s to 210-s.
35. Madison, J. D., and Aagesen, L. K.
2012. Quantitative characterization of
porosity in laser welds of stainless steel.
Scripta Materalia 67: 783–786.
36. Lu, Y., Chen, J. S., Zhang, Y. M., and
Kvidahl, L. 2014. Predictive control based
double-electrode submerged arc welding
for fillet joints. Journal of Manufacturing
Processes 16: 415–426.
37. Lu, Y., Chen, J. S., Shi, Y., Li, X. R.,
Chen, J. S., Kvidahl, L., and Zhang, Y. M.
2014. Double-electrode arc welding
process: Principle, variants, control and developments. Journal of Manufacturing
Processes 16(1): 93–108.
38. Lu, Y., Zhang, Y. M., and Kvidahl, L.
2013. Heat input reduction in fillet welding using bypass and gap. Welding Journal
92(12): 390-s to 400-s.
39. Shi, Y., Liu, X., Zhang, Y. M., and
Johnson, M. 2008. Analysis of metal transfer and correlated influences in dual-bypass GMAW of aluminum. Welding Journal
87(9): 229-s to 236-s.
40. Li, K., and Zhang, Y. M. 2008. Consumable double-electrode GMAW Part II:
Monitoring, modeling, and control. Welding Journal 87(2): 44-s to 50-s.
41. Li, K., and Zhang, Y. M. 2008. Consumable double-electrode GMAW Part I: The
process. Welding Journal 87(1): 11-s to 17-s.
42. Li, K., Chen, J., and Zhang, Y. M.
2007. Double electrode GMAW process
and control. Welding Journal 86(8): 231-s
to 237-s.
43. ASTM Standard Test Methods for
Tension Testing of Metallic Materials [Metric], Designation: E 8M-03.
44. AWS D8.8M: Specification for Automotive Weld Quality Arc-Welding of Steel.
Miami, Fla.: American Welding Society.
45. Pickin, C. 2007. Premium automotive research and development project:
Cold metal transfer (CMT) brazing of galvanised steel sheets. Warwick Manufacturing Group (WMG), University of Warwick.
46. Lawrence, F. V., and Cox, E. P. 1976.
Influence of inadequate joint penetration
on tensile behavior of A514 steel welds.
Welding Journal 55(5): 113-s to 120-s.
47. Schwedler, O., Schlosser, B., and
Jüttner, S. 2015. Welding of press hardened steel sheet using controlled GMA
welding processes with regard to manufacturing influences. IIW-Doc. No. SC-Auto81-15, pp. 1 to 8.