Int J Adv Manuf Technol (2010) 48:143–154

DOI 10.1007/s00170-009-2270-x

ORIGINAL ARTICLE

Seam gap bridging of laser based processes for the welding

of aluminium sheets for industrial applications
Bernard Johan Aalderink · B. Pathiraj ·
R. G. K. M. Aarts

Received: 16 November 2008 / Accepted: 13 August 2009 / Published online: 29 August 2009
© Springer-Verlag London Limited 2009

Abstract Laser welding has a large potential for the process can bridge even gaps up to 1.0 mm. Most welds
production of tailor welded blanks in the automotive had a high tensile strength. However, during Erichsen
industry, due to the low heat input and deep penetra- Cupping tests, the deformation of the welds is signifi-
tion. However, due to the small laser spot and melt cantly lower as compared to the base material.
pool, laser-based welding processes in general have a
low tolerance for seam gaps. In this paper, five laser- Keywords Laser welding · Hybrid welding ·
based welding techniques are compared for their gap Twin spot welding · Aluminium · Gap bridging
bridging capabilities: single-spot laser welding, twin-
spot laser welding, single-spot laser welding with cold
wire feeding, twin-spot laser welding with cold wire 1 Introduction
feeding and laser/GMA hybrid welding. Welding ex-
periments were performed on 1.1- and 2.1-mm-thick The use of aluminium tailor welded blanks (TWBs) is
AA5182 aluminium sheets. The resulting welds were an important way to reduce the weight of a car body
evaluated using visual inspection, cross sectional analy- [1, 2]. Laser welding is potentially well-suited for the
sis with optical microscopy, tensile tests and Erichsen production of TWBs due to the relatively high weld-
Cupping tests. The results show that the use of a filler ing speed compared to traditional welding processes,
wire is indispensable to increase the gap tolerance. A the non-contact nature of the welding process and the
proper alignment of this wire with the laser spot(s) is suitability for process automation [3]. In the production
crucial. With the single spot laser welding with cold of TWBs, sheets are typically welded in a butt weld
wire feeding, a gap up to 0.6 mm could be bridged as configuration. In this weld configuration, the maximum
opposed to a maximum allowable gap width of 0.2 mm allowable misalignment between laser spot and weld
for single-spot laser welding without filler wire. For seam is about 0.2 to 0.3 mm [4, 5]. This imposes high
2.1-mm-thick sheets, the laser/GMA hybrid welding demands on work-piece clamping and manipulator ac-
curacy [6]. It is also important that the sheets fit-up well
and that the air gap between the sheets is minimised.
This requires a good edge preparation [4, 5]. An in-
creased tolerance for these seam gaps could potentially
B. J. Aalderink (B) improve the production yields and laser weld quality in
Materials Innovation Institute, P.O. Box 217, 7500 AE,
an industrial environment.
Enschede, The Netherlands
e-mail: [email protected] Although many publications provide parameter
studies of different laser-based welding processes, a
B. Pathiraj · R. G. K. M. Aarts comparative overview of the various processes with
Department of Mechanical Automation, Faculty of
Engineering Technology Applied Laser Technology,
practical guidelines for industry is still lacking. The
University of Twente, P.O. Box 217, 7500 AE, Enschede, research described in this paper focusses on the applica-
The Netherlands bility of five Nd:YAG laser-based welding processes for
144 Int J Adv Manuf Technol (2010) 48:143–154

aluminium TWB production in the presence of a seam Figure 1 gives a schematic representation of the five
gap. Welding experiments on 1.1- and 2.1-mm-thick different laser-based welding processes. For the SSCW
AA5182-O/H111 sheets in a butt weld configuration welding, the SS welding set-up additionally equipped
were carried out using single-spot laser welding (SS) with a Dinse DIX WDE 315 push–push wire feeder
[3], twin-spot laser welding (TS) [7–9], single-spot laser was used. During the experiments, a standard AA5356
welding with cold wire feeding (SSCW) [10, 11], twin- filler wire (1.2 mm diameter, Table 1) was fed from the
spot laser welding with cold wire feeding (TSCW) [4, 5] front. The angle between the wire feeding nozzle and
and laser/GMA hybrid welding (LGH) processes [12, the work-piece was 35◦ . The wire touches the work-
13]. Details of the welding experiments are provided in piece at the point of incidence of the laser beam on the
Section 2. sheet surface.
Section 3 describes a series of practical weld quality The TS experiments (both TS and TSCW) were
tests that have been selected in this research that in- performed using the same laser source, wire feeder
clude visual inspection of the weld surface, microscopic and manipulator as the SS and SSCW experiments.
observation of weld cross sections, tensile strength tests However, in these experiments, a BEO70 welding head
and Erichsen Cupping tests. Finally, a summary of the (with a 200-mm collimator lens and a 150-mm focus
most important conclusions is presented. lens) was equipped with a beam splitting prism. This
prism was placed in the collimated laser beam inside the
welding head such that the laser power was equally dis-
2 Welding experiments tributed over the two laser spots. The distance between
the spots at the focal position was 1.5 mm. The laser
The different welding processes and the experimen- spots were projected side-by-side onto the work piece
tal set-ups used during welding are presented in and were symmetrical to the weld seam. For the TSCW
Section 2.1, followed by a description of the process experiments, the wire feeding set-up was identical to
parameters used during the welding experiments. the set-up used in the SSCW experiments. Also, a 1.2-
mm AA5356 filler wire was used, which was aligned
symmetrically with respect to the weld seam and the
2.1 Experimental set-up
laser spots.
The LGH welding experiments were carried out
Table 1 lists the chemical composition of the AA5182
using a standard Fronius LaserHybrid welding head.
base material. The SS welding experiments were car-
This welding head combines a Fronius GMA welding
ried out using a 4-kW Trumpf THL4006D industrial
torch connected to a Fronius TPS5000 digital GMA
laser. This laser was connected to an optical fibre with a
welding source, with a Trumpf BEO70 laser welding
core diameter of 0.6 mm. The laser light was projected
head. The laser welding head was equipped with a
onto the work piece using a standard Trumpf BEO70
200-mm collimator lens and a 220-mm focussing lens,
welding head with a 200-mm collimator lens and a
resulting in a focal diameter of 0.66 mm. In this set-
150-mm focus lens, which resulted in a focal diameter
up, the GMA torch was mounted in a trailing position.
of 0.45 mm. Unless mentioned otherwise, the welding
The angle between the GMA torch and the work-piece
head was moved over the sample using a gantry-type
was 55◦ , and the angle between the optical axis of
manipulator. In this set-up, the optical axis of the weld-
the laser welding head and the work-piece normal was
ing head was normal to the sample surface.
5◦ . The distance between the centre of the MIG arc
and the laser beam at the sample surface was 3.0 mm.
The wire was identical to the wire used for the cold
Table 1 The chemical composition of the AA5182 base material
(courtesy of Corus) and the AA5356 filler wire (specifications by wire feeding experiments. The hybrid welding head was
Thyssen welding) manipulated using a Stäubli RX170HP robot arm.
Element Content (wt.%)
AA5182 (sheet) AA5356 (wire)
2.2 Welding process parameters
Mg 4.90 4.5–5.5
Mn 0.26 0.1–0.5
Cu 0.11 – To investigate the gap bridging capabilities of the dif-
Si 0.08 – ferent welding processes, a butt weld configuration was
Ti 0.01 <0.15 used. In this configuration, similar sheets of 1.1 or
Cr – <0.3 2.1 mm thickness were joined. The seams were always
Al Bal. Bal. welded parallel to the rolling direction of the sheets.
Int J Adv Manuf Technol (2010) 48:143–154 145

Fig. 1 A schematic overview of the five different welding processes

In both cases, a constant pre-set gap between the and arc power (Pg ) were used, which have been deter-
sheets was maintained. For each configuration, opti- mined experimentally. In Table 2, an overview of the
mised values for the laser power (P), welding speed (v) optimal welding parameters for the different processes

Table 2 The welding parameters used during the gap bridging experiments for the different welding processes on AA5182 aluminium
sheets
SS TS SSCW TSCW LGH
Sheet thickness b (mm) 1.1 2.1 1.1 2.1 1.1 2.1 1.1 2.1 1.1 2.1
Laser power P (kW) 4.0 4.0 4.0 4.0 4.0 4.0 3.5 4.0 2.5 3.5
Focal diameter ds (mm) 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.66 0.66
Welding speed v (mm/s) 160 100 140 60 160 80 100 60 120 100
Wire speeda vw (mm/s) – – – – 100–165 100–215 110–190 45–170 40 115
Arc power Pg (kW) – – – – – – – – 0.5 2.0
Heat input qi (J/mm) 25 40 29 67 25 50 35 67 25 55
a For cold wire feeding, the optimal wire speed depends on the gap width
146 Int J Adv Manuf Technol (2010) 48:143–154

is given. Parameter qi is the combined heat input during is mainly due to the large interaction zone of the arc
welding (energy per unit weld length supplied by the with the work piece.
welding equipment) which is calculated as follows: With the LGH welding process, a black deposit is vis-
ible next to the weld bead. The other welding processes
P + Pg
qi = . (1) do not show this, but they do show a grey deposit
v
covering the bottom surface of the weld. These deposits
To obtain a satisfactory weld quality, the wire speed vw mostly contain aluminium oxide and magnesium oxide,
needs to be adjusted for each value of the pre-set gap. with some traces of metallic aluminium and magnesium
During the SS, TS, SSCW and TSCW experiments, [16], and are formed by evaporation and condensa-
argon shielding gas was used with a top flow of 1,360 l/h tion of aluminium and magnesium elements during the
and a back flow of 340 l/h. During the LGH experi- welding process. Since the temperatures involved in the
ments, the weld was shielded using an argon top flow arc of the LGH welding process are higher than those in
of 900 l/h and an argon back flow of 480 l/h. Prelimi- the plume and keyhole of the laser welding process, the
nary experiments have indicated that the optimal weld evaporation will be stronger. Near the interaction zone
quality is achieved with the focal plane on or slightly of the laser and arc with the material, the shielding gas
under the top sheet surface (0 to 0.2 mm), which is in prevents oxidation of these metallic elements. Further
correspondence with the results of others [14]. In this away, the aluminium and magnesium will oxidise and
study, the focal plane was chosen to coincide with the deposit on the colder base material. Normally, these
top surface of the sheets. oxides are white. However, in this case, the particle
size is so small that the deposit entraps light, which
gives it a dark appearance. The grey deposit on the
3 Weld quality evaluation bottom surface of the laser welds indicates that the
gas shielding on the bottom was inadequate. These
For industrial applications, a good weld quality assess- deposits have no direct effect on the weld strength and
ment is important. Since, in this study, different welding performance [16, 17].
processes are compared, a practical means of weld qual- In Table 3, the maximum allowable gaps for each
ity assessment had to be chosen. A visual inspection of of the different welding processes are given. These
the weld surfaces can reveal macro defects like cracks, values have been determined from visual inspections
blow and through holes, and hence was performed as of welds with different pre-set gap sizes, ranging from
the first step in the quality assessment (Section 3.1). 0 to 1.2 mm. In this evaluation, the weld is considered
Typical welding defects in aluminium sheets and acceptable if, at maximum, two small (≤2.0 mm) or one
their quality assessment are defined in international large (>2.0 mm) fully penetrated hole(s) is present in
standard ISO 13919-2 [15]. For any specific application, the weld seam. The samples were welded over a length
not all geometric imperfections defined in this standard of 220 mm, without using a tag weld. Since start and
are equally relevant and are practical to assess. Cracks, stop effects of the welding processes are outside the
undercut, excess weld material, excess penetration and scope of this research, the first and last 20 mm of the
shrinkage grooves were selected to be relevant for welds are not included in the evaluation.
TWB applications (Section 3.2). The SS welding experiments show a dramatic de-
The weld quality will depend further on material- crease in weld quality when the gap is over 0.2 mm.
related properties like strength, microstructure and An important cause for welding defects in this case is
hardness. Assessment of tensile strength and formabil- the lack of gap filling by molten metal. Since the SS
ity based on Erichsen Cupping tests were chosen for melt pool is small (about 1.5 mm wide) compared to
this application (Sections 3.3 and 3.4). other welding processes, there is only a limited amount
of liquid metal available to fill the gap. This results in
3.1 Visual inspection of weld surfaces deep undercuts and holes.
It is difficult for the SS process to initiate a stable
Figure 2 shows typical appearances of weld beads for welding process in the presence of a gap. Since the laser
each of the five welding processes. The weld beads spot is symmetrically aligned with the seam, at the start
resulting from the SS and SSCW welding processes are of the weld, a large part of the laser radiation escapes
narrower than those from the TSCW welding process. through the gap between the sheets. Only after some
For the twin spot processes, the parallel placement time, enough liquid metal is formed to fill the gap in
of the laser spots results in a wider melt pool. The front of the laser beam, preventing such escape of laser
widest welds are produced by the LGH process, which radiation.
Int J Adv Manuf Technol (2010) 48:143–154 147

Fig. 2 Typical weld beads for the different laser-based welding 2.1 mm, s = 0.2 mm; SSCW b = 1.1 mm, s = 0.0 mm; TSCW
processes. The sheet thickness b and the gap width s before b = 1.1 mm, s = 1.0 mm and LGH b = 2.1 mm, s = 1.0 mm
welding are as follows. SS b = 1.1 mm, s = 0.2 mm; TS b =

With the TS welding experiments, the 1.5-mm dis- the melt pool of the TS process is wider, this process
tance between the spots ensures that all laser light is shows an unacceptable weld quality due to undercut
projected onto the base material and that the start-up and holes for even the smallest gap. The occurrence of
of the welding process is less problematic. Although undercut can be understood as the gap has to be filled
with molten material. A gap will decrease the amount
of molten metal between the two keyholes, which will
join into a single large keyhole. Such a large keyhole
Table 3 The maximum allowable gap width the different will start to elongate due to the surface tension pressure
processes estimated from a visual inspection of the weld [5, 18]. This elongated keyhole shape is not stable and
SS TS SSCW TSCW LGH will result in holes in the weld bead. This demonstrates
1.1/1.1 mm 0.2 mm 0 mm 0.4 mm 0.6 mm – that the TS process with a parallel configuration is not
2.1/2.1 mm 0.2 mm – 0.6 mm 0.6 mm 1.0 mm suited for laser welding of AA5182 aluminium sheets
148 Int J Adv Manuf Technol (2010) 48:143–154

and that the use of filler wire is essential to bridge gaps. Dark band
For the TSCW process, the wire assures that the two
keyholes do not join, which results in a stable welding
process.
By applying wire feeding, the gap tolerance of the
welding process increases significantly. SSCW welding
is especially suited if a high processing speed is required
while the maximum gap size is not larger than 0.4 mm
in the case of 1.1-mm sheets and 0.6 mm in the case
of 2.1-mm sheets. For larger gaps, not enough base SS
material is molten to form a good joint, especially at
the weld root. For these cases, the joining becomes
more a brazing process. For the TSCW welding process,
the gap bridging capability is 0.6 mm for both 1.1- and
2.1-mm sheets. However, the welding speed for the
TSCW process is lower than for the SSCW process.
Experiments show that, for both the SSCW and the
TSCW processes, it is occasionally possible to bridge
gaps up to 1.0 mm, but these results cannot be reliably
reproduced.
Although the SSCW and TSCW are significantly TS
more tolerant for gaps, both processes are quite sen-
sitive for wire misalignments. The sensitivity of these
processes for alignment deviations of the wire with re-
spect to the laser spot increases with an increasing gap
width. For gaps over 0.6 mm, the required positioning
accuracy of the wire is very high (<0.1 mm). In this case,
even the smallest variation in wire position will cause
a welding defect. This problem could be overcome by
actively actuating the wire laterally [19].
Welding experiments with the LGH welding process
for 1.1-mm sheets show that it is possible to achieve SSCW
a good weld quality. However, the process parameter
range is very small. The resulting weld quality for 1.1-
mm-thick sheets, even without a gap, is lower than that
of the other processes. For 2.1-mm sheets, the LGH
welding process gives a much better weld quality. Gaps
up to 1.0 mm could be bridged without problems. This
relatively poor performance for the thinner sheets is
mainly due to the wide melt pool in combination with
the small sheet thickness, causing an unstable melt pool,
which is very susceptible to welding defects [5, 18].
The weld bead geometry, porosity content and the TSCW 1 mm
micro-structure in and around the weld can be evalu-
ated by inspection of cross sections of a welded sample. Fig. 3 Cross sections of the welds of the different processes for
In Fig. 4, the heat-affected zone (HAZ) and melt zone 1.1 mm thick AA5182 sheets
(MZ) are indicated.
In Figs. 3 and 4, a dark band at the centre of the
sheets indicates a difference in micro-structure com- this. The effect is more pronounced with the 1.1-mm
pared to the area closer to the sheet surfaces. These sheets.
differences are mainly due to an inhomogeneous plastic For all welds, the base material and MZ can be
deformation of the material during the cold rolling clearly distinguished. During the welding of the 2.1-mm
process [20]. Segregation in castings also contributes to sheets, the heat input qi is much higher than that for the
Int J Adv Manuf Technol (2010) 48:143–154 149

Base material MZ Dark band The MZ of the SS process has the same width at the
top and bottom surfaces of the weld. For the TS, SSCW
and TSCW processes, the MZ is a little bit wider at the
top than at the bottom. For the LGH process, the MZ is
much wider at the top than at the bottom. This is mainly
due to the fact that, in the GMA process, the heat input
is much greater at the surface region and the keyhole
formation is entirely due to the laser process alone.

SS
HAZ 3.2 Cross sectional weld geometry inspection

The amount of undercut and overfill can be compared

using relevant criteria from the ISO 13919-2 norm [15].
A weld can be qualified as quality level B (stringent),
C (intermediate) or D (moderate). The norm classifies
weld imperfections for all sorts of weld configurations,
and only a number of criteria are useful in the context
of butt welds. In this work, the following criteria from
ISO norm 13919-2 were chosen to be relevant: cracks,
undercut, excess weld material, excess penetration and
SSCW
shrinkage grooves (Fig. 5).
Pore

Undercut

TSCW Cracks

Excess
penetration
Excess weld
material

LGH 1 mm

Fig. 4 Cross sections of the welds of the different processes for

2.1 mm thick AA5182 sheets

1.1-mm sheets, resulting in a larger HAZ. For the 2.1-

mm sheets, the HAZ is visible in the cross sections. For
Shrinkage grooves
the welds in 1.1-mm sheets, there is hardly any HAZ Fig. 5 Schematic overview of the selected geometrical criteria
visible. from ISO norm 13919-2
150 Int J Adv Manuf Technol (2010) 48:143–154

Evaluation of the cross sections shows that none of examined cross section of the weld of the LGH process
these welds suffer from cracks. The welds do occa- with b = 2.1 mm and s = 0.4 mm.
sionally show undercut, excess weld material, excess Quality level B welds can be produced with the
penetration and shrinkage grooves. The presence of TSCW process for 2.1-mm sheets if there is no gap
overfill increases the load-bearing surface of the joint present. For gaps between 0.2 and 0.4 mm, the quality
but, especially in combination with shrinkage grooves, level drops to level C. Even larger gaps result in level
can also initiate local stress concentrations near the D welds based on the amount of undercut and excess
edge of the MZ. penetration. For 1.1-mm sheets joined with TSCW, the
In Table 4, the resulting quality levels of the welds restrictions on undercut and excess penetration posed
are given. Note that only cross sections were made of by the ISO norm level B cannot be achieved repro-
welds that show a reasonably good surface quality as ducibly.
assessed by the visual inspection. All other welds are For the LGH process, the requirements on the un-
indicated as not applicable (n.a.). The overall quality dercut and the excess penetration are the bottlenecks.
level is determined by considering the lowest quality For gaps below 0.4 mm, excess penetration limits the
level of the different criteria. For example, if a weld has overall quality level. For higher gap widths, the quality
quality level C based on the shrinkage grooves, but has level is limited by the undercut. For the 2.1-mm sheets,
a quality level D based on the excess weld material, the quality level C is the highest obtainable weld quality.
assigned overall quality level is D. In general, quality level B can only be obtained when
Table 4 shows that the SS process results in a moder- using the TS or SSCW (1.1-mm sheets) and SSCW or
ate weld quality at best. In all cases, the undercut, which TSCW (2.1-mm sheets) processes, all with zero gap. In
is due to evaporation and weld spatter, is too large to all cases, the presence of a gap deteriorates the quality
meet the requirements for the higher quality levels. The level of the weld as defined by the ISO norm.
larger volume of melt pool of the TS welding process
will lead to lesser undercut. For the 1.1-mm sheets, the
TS process can therefore meet the requirements for 3.3 Tensile tests
quality level B. However, as was already pointed out,
this process has no tolerance for seam gaps and is not The ultimate tensile strength (σb ) was determined ac-
suited for welding of the 2.1-mm sheets. cording to the ASTM E 8M-98 norm [21]. All tensile
For gaps under 0.2 mm, the SSCW process can also tests were performed on samples cut transverse to the
deliver weld quality of level B. For larger gaps, the qual- welding direction.
ity level is C or lower, based on the criteria of excess The σb of the base material is determined to be
weld material and shrinkage grooves. Table 4 shows 275 MPa in the case of 1.1 mm AA5182 sheets and
that, for a sheet thickness of 2.1 mm and s = 0.6 mm, 284 MPa in the case of 2.1 mm AA5182 sheets. The
no data are available (indicated with –). Inspection of tensile strength is indicated as a fraction of the tensile
the cross section indicates a quality level B, but visual strength of the base material (σb% ).
inspection of the weld bead shows that the weld quality The tensile test results of the welds from the different
varies strongly over the weld length and that the actual welding processes are shown in Figs. 6 and 7. From
weld quality is lower. This was also the case for the each weld, three tensile tests have been performed.

Table 4 The overall quality Sheet thickness (b ) Gap width (s)

levels (B: high quality, C:
intermediate quality, D: (mm) 0 mm 0.2 mm 0.4 mm 0.6 mm 0.8 mm 1.0 mm
moderate quality, x: not SS 1.1 D x n.a. n.a. n.a. n.a.
passed, –: no data available, TS B n.a. n.a. n.a. n.a. n.a.
n.a.: not applicable) of the SSCW B C C n.a. n.a. n.a.
welds with different gaps s, TSCW D C D D n.a. n.a.
according to the ISO 13919-2 LGH n.a. n.a. n.a. n.a. n.a. n.a.
norm, determined from
SS 2.1 x D n.a. n.a. n.a. n.a.
inspection of the cross
TS n.a. n.a. n.a. n.a. n.a. n.a.
sections
SSCW B C x – n.a. n.a.
TSCW B C D D n.a. n.a.
LGH C C – D D D
Int J Adv Manuf Technol (2010) 48:143–154 151

Fig. 6 The tensile test data 1.2 1.2

from welds of the different
processes for 1.1-mm-thick 1 1
AA5182 sheets. The
variations found in σb% found 0.8 0.8
for the different gap values s

σ b%

σ b%
are indicated with the error 0.6 0.6
bars. The base material σb
value is 275 MPa 0.4 0.4

0.2 0.2
SS TS
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]
1.2 1.2

1 1

0.8 0.8
σ b%

σ b%
0.6 0.6

0.4 0.4

0.2 0.2
SSCW TSCW
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]

Fig. 7 The tensile test data 1.2 1.2

from welds of the different
processes for 2.1-mm-thick 1 1
AA5182 sheets. The
variations found in σb% found 0.8 0.8
for the different gap values s
σ b%

σ b%

are indicated with the error 0.6 0.6

bars. The base material σb
value is 284 MPa 0.4 0.4

0.2 0.2
SS SSCW
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]
1.2 1.2

1 1

0.8 0.8
σ b%

σ b%

0.6 0.6

0.4 0.4

0.2 0.2
TSCW LGH
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]
152 Int J Adv Manuf Technol (2010) 48:143–154

Fig. 8 The Erichsen Cupping 1 1

test values for the welds of 0.9 0.9
the different processes for
0.8 0.8
1.1-mm-thick AA5182 sheets.
0.7 0.7
The variations found in I E%
found for the different gap 0.6 0.6

I E%

I E%
values s are indicated with the 0.5 0.5
error bars. The base material 0.4 0.4
I E value is 8.9 mm
0.3 0.3
0.2 0.2
0.1 SS 0.1 TS
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]
1 1
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
I E%

I E%
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 SSCW 0.1 TSCW
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]

Fig. 9 The Erichsen Cupping 1 1

test values for the welds of 0.9 0.9
the different processes for
0.8 0.8
2.1-mm-thick AA5182 sheets.
0.7 0.7
The variations found in I E%
found for the different gap 0.6 0.6
I E%

I E%

values s are indicated with the 0.5 0.5

error bars. The base material 0.4 0.4
I E value is 10.8 mm
0.3 0.3
0.2 0.2
0.1 SS 0.1 TS
SSCW
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]
1 1
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
I E%

I E%

0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 SSCW
TSCW 0.1 TSCW
LGH
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
s [mm] s [mm]
Int J Adv Manuf Technol (2010) 48:143–154 153

The scatter in tensile strength values is indicated with the figure indicate the scatter in these measurement
error bars. values.
The cross sections indicate that, for the SS, the TS The index I E% did not show any significant varia-
and the LGH processes for both 1.1- and 2.1-mm sheets, tions with the gap width s, for any welding processes
the load-bearing surface in the MZ decreases with employed (Figs. 8 and 9). This implies that the forma-
increasing gap widths. The tensile tests show that, for bility of the weld does not depend on the applied gap
these welds, the load upon failure also decreases as a width.
function of the gap width and that the welds fail close However, the index shows a strong dependency on
to the fusion line (boundary between MZ and HAZ), the type of laser welding process employed. Figure 8
as expected. shows that, for the welding of 1.1-mm sheets, the SS,
The SSCW welds of the 1.1-mm sheets have excess TS and SSCW processes result in welds with the highest
weld material and excess penetration, which increases formability. For 2.1-mm sheets, the SS and the SSCW
the load-bearing cross sectional area. These samples fail welding processes give the best deformable welds. For
just outside the MZ at a position of local stress concen- all except the SSCW processes, the samples failed in
tration (Fig. 3). For the 2.1-mm sheets, the samples fail the MZ, along the weld line. For the SSCW process,
in the MZ, near the fusion line. the samples generally failed near the fusion line. This
Welds of the 1.1-mm sheets performed with the indicates that, for applications with high local deforma-
TSCW process fail in the base material and the tensile tions of the weld seam, welds produced with all tested
strength values are equal to that of the base material. welding methods will have a high risk of failure during
It is suspected that the excess weld material and pen- the deformation process.
etration increase the load-bearing surface of the MZ,
which decreases the nominal stress in the MZ, resulting
in failure outside the MZ. The tensile test specimens
4 Conclusions
for the 2.1-mm TSCW welds show a similar behaviour
as the 2.1 mm SSCW samples and fail in the MZ, near
For the welding of TWBs in the automotive industry, a
the fusion line.
robust joining method is essential. The low heat input
and deep penetration of laser welding make this poten-
tially a very suitable joining method. The disadvantage
3.4 Erichsen Cupping tests of the standard SS process lies in the low tolerance for
seam gaps. In this paper, the gap bridging capabilities of
Erichsen Cupping tests were performed according to five different laser welding processes were investigated.
the ASTM E 643-84 norm [22]. Figures 8 and 9 show the Based on the welding experiments and subsequent weld
relative Erichsen Cupping index I E% for the welds of quality analysis, the results can be listed as follows:
the different processes. For the 1.1-mm base material,
I E is 8.9 mm. For the 2.1-mm base material, I E is – Table 5 gives an overview of the performance of
10.8 mm. For each weld configuration and gap width, the different laser welding processes towards the
three cupping tests were performed. The error bars in different tests performed. The table can be used as

Table 5 A summary of results obtained on the welds produced by the different laser welding processes
Sheet Gap bridging Quality Tensile Erichsen
thickness capabilitya of bead strengthc Cupping
[mm] [mm] shapeb index
SS 1.1 0.2 Moderate 0.78–0.88 0.74–0.82
TS 0 High 0.92 0.58
SSCW 0.4 High 0.84–0.90 0.70–0.79
TSCW 0.6 Moderate 0.97–0.99 0.36–0.41
SS 2.1 0.2 Low 0.78–0.85 0.54
SSCW 0.6 High 0.92–0.96 0.48–0.58
TSCW 0.6 High 0.87–0.94 0.31–0.37
LGH 1.0 Intermediate 0.77–0.88 0.27–0.35
a The maximum seam gap that can be bridged, based on visual inspection of the weld bead
b Determined from cross sections of welds with no pre-set gap and compared to criteria formulated in ISO norm 13919-2
c The tensile strength and the Erichsen Cupping index values are given as fraction of the base material value
154 Int J Adv Manuf Technol (2010) 48:143–154

a guideline for selection of a laser-based welding 2. van Nieuwerburgh D, Smeyers A, van der Veldt T (2004)
process. Market introduction of Al tailor welded blank technology. In:
Proc. of EALA ’04
– The SSCW has the best overall performance. For 3. Duley WW (1999) Laser welding. Wiley, New York
applications with a sheet thickness around 2 mm 4. Aalderink BJ, Aarts RGKM, Meijer J (2007) Increased gap
and an expected gap size over 0.6 mm, the LGH bridging capabilities using twin spot and hybrid laser/GMA
process might also be an option. welding for AA5182. In: Proc. of LIM 07. AT-Verlag,
Munich, pp 79–83
– The TS process in parallel configuration is not
5. Aalderink BJ (2007) Sensing, monitoring and control of laser
suited for welding sheet material due to the lack of based welding of aluminium sheets. PhD thesis, Enschede,
tolerance for seam gaps. The LGH process is not University of Twente
suited for 1.1-mm aluminium sheets due to the high 6. Hakvoort WBJ (2009) Iterative learning control for LTV
systems with applications to an industrial robot. PhD thesis,
heat input and wide melt pool.
Enschede, University of Twente
– The use of filler wire is necessary to increase the gap 7. Hohenberger B, Chang CL, Schinzel C, Dausinger F, Hügel
bridging capabilities of the laser welding process. H (1999) Laser welding with Nd:YAG-Multi-Beam tech-
– A precise alignment of the filler wire during the nique. In: Proc. of ICALEO ’99, pp D167–D176
8. Iwase T, Shibata K, Sakamoto H, Dausinger F, Hohenberger
process is crucial for a good weld quality. For seam
B, Müller M, Matsunawa A, Seto N (2000) Real time x-ray
gaps over 0.6 mm, the demands on the wire align- observations of dual beam welding of aluminium alloys. In:
ment with the laser spot are too strict to be realised Proc. of ICALEO ’00, pp 26–34
in an industrial environment. 9. Xie J (2001) Dual beam laser welding. Weld J 81(9):223-s–
230-s
– When gaps are presenting a butt weld configura-
10. Schneegans J (1992) Untersuchungen zum laser-
tion, no laser welding process can produce joints strahlschweissen mit zusatzdrahtfürung an un- und
that meet quality level B (most stringent quality niedriglegierten Stälen. PhD thesis, Rheinisch-Westfälische
level according to ISO norm 13919-2). Technische Hochschule Aachen
11. Fuest D (1995) Entwicklung von systemkomponenten zur au-
– The use of filler wire eliminates undercut of the
tomatisierung von CO2 -laserstrahlprozessen mit zusatzdraht.
weld bead and increases the load-bearing surface PhD thesis, Rheinisch-Westfälische Technische Hochschule
(which results in a higher load bearing capacity of Aachen
the weld). 12. Bagger C, Olsen FO (2005) Review of laser hybrid welding. J
Laser Appl 17:1–14
– For all tested welding methods, the produced welds
13. Mahrle A, Beyer E (2006) Hybrid laser beam welding—
have a high risk of failure in the weld seam when classification, characteristics and applications. J Laser Appl
large deformations have to be applied. 18:169–180
– In general, welds with a large reinforcement from 14. Cao X, Wallace W, Poon C, Immarigeon JP (2003) Research
and progress in laser welding of wrought aluminum alloys. I.
overfill around the weld tend to have a high max- Laser welding processes. Mater Manuf Process 18(1):1–22
imum tensile test value, but prove to be poorly 15. ISO (2001) ISO 13919-2: welding—electron and laser beam
deformable. For TWB production, it is therefore welded joints—guidance on quality levels for imperfections—
essential to choose a welding method which re- part 2: aluminium and its weldable alloys
16. Armao F (2003) Invesigation of the smut generation during
sults in welds with both limited undercut/shrinkage
welding aluminium alloy with GMAW. Pract Weld Today
grooves and a small overfill/excess penetration. 7:40–41
17. Jyogan S, Kogane K, Sasabe S, Hirano H, Ochiai T, Okita
T (1998) How to recognize, minimize weld smut. In: Proc. of
Acknowledgements This research was carried out under the Inalco 98. Woodhead, Cambridge, pp 17–25
project number MC8.02116 in the framework of the Research 18. Aalderink BJ, de Lange DF, Aarts RGKM, Meijer J (2007)
Program of the Materials innovation institute M2i (www.m2i.nl), Keyhole shapes during laser welding of thin metal sheets. J
the former Netherlands Institute for Metals Research. The au- Phys D Appl Phys 40:5388–5393
thors want to express their gratitude to Corus for their support 19. Huwer A (1993) Sensorsystem zur erfassung variabler
and to the Joining and Mechanical Behaviour group of the Uni- fügespaltweiten beim laserstrahlschweissen im stumpfstoss.
versity of Delft for their aid in this research. PhD thesis, Aachen, Rheinisch-Westfälische Technische
Hochschule Aachen
20. Wen W, Morris JG (2004) The effect of cold rolling and
annealing on the serrated yielding phenomenon of AA5182
aluminum alloy. Math Sci Eng A 373:204–216
References 21. ASTM International (1998) ASTM E 8M-98: standard test
method for tension testing of metallic materials [Metric]
1. Davies RW, Oliver HE, Smith MT, Grant GJ (1999) Charac- 22. ASTM International (1984) ASTM E 643-84: standard
terizing Al tailor-welded blanks for automotive applications. test method for ball punch deformation of metallic sheet
JOM 51:6–50 material