Optics & Laser Technology 150 (2022) 107939

Contents lists available at ScienceDirect

Optics and Laser Technology

journal homepage: www.elsevier.com/locate/optlastec

An investigation on laser/DP-MIG hybrid weldability of aluminum to Al-Si

coated boron steel
Xiaobing Cao a, Ji’an Duan a, *, Zhou Li a, *, Baoqi Chang a, Cong Xu b, Zheng Xie c
a
College of Mechanical and Electrical Engineering, Central South University, Changsha 410083, China
b
Hunan Hostechn Automation Technology Co., Ltd, Changsha 410000, China
c
Hunan Industry Polytechnic, School of Mechanical Engineering, Changsha 410208, China

A R T I C L E I N F O A B S T R A C T

Keywords: Laser/double pulse-MIG (laser/DP-MIG) hybrid welding-brazing of 5052 aluminum alloy to Al-Si coated boron
Laser/DP-MIG hybrid welding steel was successfully conducted in a lap configuration with an Al-5Si filler. The heat input plays a significant role
Dissimilar metals joining in determining the welding performance. The satisfactory appearance and excellent brazed interface with a
Aluminum
maximum thickness of 4.1 μm was achieved at the optimum heat input of 161.8 J/mm, and the maximum joint
Al-Si coating
Boron steel
strength reached to 168.8 MPa in excess of 75% aluminum resistance. Three kinds of intermetallic compounds
Intermetallic compounds (IMC) phases were nicely identified: τ5-Fe2Al8Si, θ-Fe(Al,Si)3 and η-Fe2(Al,Si)5. Microstructure evolution of
brazed joint showed that microstructure of upper-critical HAZ (UCHAZ) consisted of martensite whereas that of
inter-critical HAZ (ICHAZ) consisted of ferrite and martensite. The two failure modes (interfacial failure (IF) and
HAZ failure (HF)) was associated with the interfacial characteristic and porosity percentage. IF mode was a
typical intergranular fracture, and its cracks initiated from notch of weld joint and then redirected toward the
region of pore concentration. HF mode was a mixture of major ductile fracture and minor cleavage fracture, and
its cracks might initiate from notch of weld joint and then propagated along the IMC throughout the whole
interface.

1. Introduction aluminum and liquid steel at high temperature, leading to a weaker joint
strength [9,10]. For these reasons above, more attractions were mainly
The aluminum/steel hybrid parts in automotive body have great focused on the welding-brazing processes, which involved solid/liquid
potential for making a compromise between decreasing cost and guar­ interface by only liquid aluminum with solid steel. For example, Pour­
anteeing security performance by using the combination of advance anvari and Abbasi [11] concluded that there is a critical heat input to
high strength steel (AHSS) and aluminum [1–3]. The limited weldability obtain an excellent joint during the MIG welding-brazing of aluminum
between steel and aluminum has been reported in many studies. Li et al. to steel. Liu et al. [12] explored the laser welding-brazing of aluminum
[4] showed in their works that joining aluminum to steel remained a to galvanized steel using pure Al powder. Hence, the welding-brazing
difficult task due to the large differences of physical properties including process provides us a promising alternative to effectively suppress the
thermal conductivity and melting points as well as thermal expansion. formation and growth of IMC.
Another factor of low bonding strength of interface is the formation and The laser-MIG hybrid welding-brazing coupled arc and laser beam
growth of brittle intermetallic compounds (IMC) [5]. The formation of heat source in a single molten pool. The synergistic enhancement be­
IMC had a negative impact on the joint strength when its thickness tween arc and laser integrated their respective advantages, making it
excessed 10 μm [6,7]. Such suppressing the formation and growth of strongly suitable for joining of dissimilar metals [13–15]. During the
IMC became an important issue via various joining techniques. The laser-MIG hybrid welding, the laser beam induces the formation of sta­
solid-state friction stir welding (FSW) can reach a better dissimilar Al/ ble arc plasma [16]. The laser-MIG hybrid welding-brazing has unique
steel joint [8]. However, the engineering application of FSW was limited advantages of a higher welding stability, bridge ability and better
by the shape of joint. The conventional fusion welding easily induced the welding formation especially increasing the wettability of aluminum to
formation of large IMC due to the directed interaction of liquid steel [15,17,18]. It is known to all that a lower heat input can effectively

* Corresponding authors.
E-mail addresses: [email protected] (J. Duan), [email protected] (Z. Li).

https://doi.org/10.1016/j.optlastec.2022.107939
Received 25 September 2021; Received in revised form 8 December 2021; Accepted 27 January 2022
Available online 12 February 2022
0030-3992/© 2022 Elsevier Ltd. All rights reserved.
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

Table 1
Chemical compositions of steel and 6061 aluminum alloy (w-%).
Materials C Si Mg Mn Zn B Fe Al

Steel 0.21 0.26 – 1.31 0.01 0.002 Bal. 0.02

Aluminum – 0.67 1.08 0.07 0.01 – 0.29 Bal.
Al-5Si wire – 4.5–6.0 0.05 0.05 0.10 – 0.8 Bal.

Fig. 1. The microstructure of (a) boron steel and (b) Al-Si coating.

limit the formation and growth of IMC. Such it is necessary to maintain a However, the comprehensive studies weren’t discussed during the laser/
lower heat input during laser-MIG hybrid welding-brazing. The double DP-MIG welding-brazing especially its application to join aluminum/
pulse MIG (DP-MIG) is a variation of gas metal arc welding (GMAW), steel joint.
which provides a modified pathway to reduce the arc heat input [19,20]. Most recently, aluminized steel replacing galvanized steel in dis­
Wang et al. [21] demonstrated that the DP-MIG could significantly similar Al/steel joining become a popular alternative. The joining
refine microstructure of fusion zone due to the effect of period alter­ mechanisms of Al alloy/aluminized steel differed from that of Al alloy/
nation of thermal peak and base current. It is reasonable to assume that galvanized steel [23,24]. The Al-Si coating become a research focus for
the formation of IMC can be suppressed by the periodical vibration of the aim of limiting IMC thickness [24–27]. For example, Kang and Kim
molten pool. Therefore, a novel attempt was conducted to develop laser/ [24] observed a lower IMC layer thickness in the CMT of aluminum to
DP-MIG hybrid welding-brazing in order to give a suitable method of aluminized steel rather than galvanized steel. Ding et al. [25] revealed
dissimilar aluminum/steel welding. The possibility of the hybrid laser/ the role of Al-Si coating in refill friction stir spot welding of aluminum to
DP-MIG method was preliminarily investigated in our previous study Al-Si coated steel. Huang et al. [26] conducted the laser welding-brazing
[22]. The laser/DP-MIG welding-brazing could limit IMC and improve of aluminum to Al-Si coated press-hardened steel using Zn-Al15 filler.
wettability compared with conventional MIG welding-brazing. For these cases, the Al-Si coating plays a key role in the improvement of

Fig. 2. The laser/DP-MIG hybrid welding system: (a) and (b) the right view and frontal view of schematic diagram, (c) the current waveform of double pulse MIG, (d)
experimental setup.

2
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

Table 2
The detailed welding parameters.
Specimens Laser power (W) Average current (A) Average voltage (V) Current frequency (Hz) Welding speed (mm/s) Line heat input (J/mm)

S1 500 60 16.5 3 11 135.5

S2 500 65 17.2 3 10 161.8
S3 600 70 17.5 3 9 202.8
S4 900 70 17.5 3 9 236.1

the metallurgic bonding and wettability in dissimilar Al/steel welding.

Therefore, the Al-Si coated boron steel rather than galvanized steel in
this study were selected to achieve an excellent dissimilar Al/steel joint.
The laser/DP-MIG welding-brazing of aluminum to Al-Si coated boron
steel were comprehensively explored by varying the heat input. The
welding appearances, interface joining mechanisms, and microstructure
analysis were comprehensively evaluated. The tensile-shear tests were
carried out, and the two failure modes were discussed.

2. Materials and methods

2.1. Materials

The Al-Si coated boron steel and 5052 aluminum having the same
thickness of 2 mm were selected as base metals. The Al-5Si wire
(ER4043) with a diameter of 1.2 mm was employed as the filler material.
Table 1 lists the chemical composition of filler material and base metals.
As shown in Fig. 1(a), the as-received boron steel is original state, which
is composed of ferrite and pearlite. The microstructure of boron steel
evolves from ferrite and pearlite to fully martensite during press-
hardened process [28]. Fig. 1(b) displays a 30 μm thick Al-Si coating
deposited on boron steel substrate, which is comprised of regions A and
B. According to EDS analysis and relevant literatures [25,29], it seems to
be identified as Al-Si eutectics and Fe2Al7Si, respectively.
Fig. 3. The preparative of tensile-shear and microstructural samples.
2.2. Welding configuration

The dissimilar Al/steel lap joint was successfully obtained using a

continuous fiber laser coupled with a Panasonic YD-350GP5 MIG welder
in double pulse mode. The laser/DP-MIG hybrid welding system
configuration was illustrated in Fig. 2. The schematic diagram of laser/
DP-MIG hybrid welding-brazing in this study was given in Fig. 2(a, b).
The aluminum sheet was placed on the top surface of steel sheet by
maintaining an overlap distance of 12 mm. The welding offset was
limited as 0.5 mm. The gap distance between aluminum and steel sheet
Fig. 4. The schematic diagram of measured regions in brazed interface.
was maintained at 0.1 mm. The angel of laser beam to vertical axis was
5◦ while that of DP-MIG torch to horizonal axis was 70◦ . An interval of
about 5 mm in the welding surface was placed between the laser beam 1 1
f = = (2)
and DP-MIG filler. The interaction of laser beam and electrical arc T T1 + T2
synergistically strengthens the welding performance due to the laser
beam inducing the stable DP-MIG plasma [13,30]. The experimental where Ek is the line heat input, P, I, U, and v are laser power, average
setup is given in Fig. 2(d). current, average voltage, and welding speed, respectively, f is current
Before the welding-brazing, acetone was utilized to eliminate greases frequency, T is a thermal period, which is the sum of the times of thermal
and dusts. Argon shielding gas with a flow rate of 18 L/min was applied peak phase (T1) and thermal base phase (T2).
to prevent oxidation. The wires extension was set as 15 mm. The
extensive preliminary tests were conducted to explore the possible op­
2.3. Analysis methods
timum parameters. The heat input is considered as a significant factor in
determining the welding performance. The heat input can be quantita­
After the welding-brazing, the appearances were examined using
tively calculated by the Eq. (1). Table 2 lists the detailed welding pa­
Keyence VHX-5000 optical microscopy (OM). Fig. 3 exhibits the pre­
rameters selected in this study on the basis of preliminary tests. The
parative of tensile-shear and microstructural samples, which were cut by
different heat input was accomplished by varying laser power, average
a wire electrical discharge machining. The standard procedures of me­
current and welding speed. Fig. 2(c) shows the current waveform of
chanical ground and polish were conducted for the further character­
double pulse MIG. The average current is the average of thermal peak
ization in interface structure and microstructure evolution. Fig. 4 shows
current Itp and base current Itb. The current frequency is defined by the
the schematic of three examined regions approximately locating in the
following Eq. (2).
quarter point of brazed interface. The interface microstructure of head,
P+I×U central and foot regions was identified by scanning electron microscope
Ek = (1)
v (SEM) with an energy dispersive X-ray spectroscopy (EDS). The porosity

3
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

Fig. 5. The welding appearances and cross-sectional macrostructures in different heat input of (a) S1, (b) S2, (c) S3, and (d) S4.

features were well extracted by SEM in secondary electron (SE) mode.

Faverage Faverage
The porosity percentage is calculated by the Eq. (3). T = = (4)
A d×h
∑
N
P = Di /Dwb (3) where T is the tensile-shear strength (MPa), Faverage is the average of
i=1 maximum load (N), d and h are the width and height of tensile-shear
sample (mm).
where the Di is the area of pores, N is the number of pores, Dwb is the area
of weld bead. In addition, the cross-section of joint was etched chemi­
cally to conduct microstructure analysis. The etching chemical reagents 3. Results and discussion
were Keller’s etching reagent (95 ml distilled water, 2.5 ml HNO3, 1.5 ml
HCl, 1 ml HF) for aluminum and 4% nital solution for boron steel. The 3.1. Appearance and cross-sectional macrostructure
microstructure evolution was analyzed by SEM and OM.
Finally, the tensile-shear tests were carried out by SANS CMT5105 Fig. 5 shows the welding appearances and cross-sectional macro­
testing machine with a rate of 2 mm/min at the room temperature. The structures of Al/steel lap joints. Obviously, the heat input plays a sig­
tensile stress of joints isn’t uniform due to the variation in failure modes. nificant role in determining the welding formations. The period
Such the tensile-shear strength is evaluated in terms of the ratio of the alternation of thermal peak and base current contributed to the gener­
average peak load to cross-sectional area, which can be expressed by Eq. ation of well ripples [21]. For the case of low heat input, an uneven and
(4). The fracture surfaces were analyzed by SEM and EDS. The Vickers discontinuous appearance with a lot of defects of spatters and burn areas
hardness was measured with an interval of 0.05 mm or 0.1 mm at an was presented in Fig. 5(a-1). The feature of instable ripples and distinct
indenter load of 100 g for 10 s. surface wave appeared due to the deficient heat input. This is also the
reason that there is no obvious HAZ in near Al-Si coating in Fig. 5(a-2).

4
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

Fig. 6. The interface microstructure of head, central, foot regions with different heat inputs: (a-c) S1, (d-f) S2, (g-i) S3, (j-l) S4.

As for the cross-section of the joint S1 in Fig. 5(a-3), the narrow weld was detected. The hydrogen in aluminum oxides may play an important
width of 4 mm and the pore concentration near the HAZ of welded joint role to induce the formation of pores [31,32]. Another reason may be the
mismatch of heat input, leading to an instable pool. The latter is not
conducive to exhaust pores. These characteristics might suggest a bad
tensile-shear strength. With the increase of heat input, the sound and
stable ripples without obvious defects of spatters, burn and pits was
nicely fabricated in Fig. 5(b-1). Fig. 5(b-2) shows an acceptable weld
width of 5 mm and pore distribution except for some small pores. This is
because the appropriate match of heat input induces the strong peri­
odical force to accelerative the escape of pores. However, when the heat
input increased from 161.8 J/mm to 236.1 J/mm, the ripples feature
became more and more inapparent (Fig. 5(c-1)) or even disappear (Fig. 5
(d-1)). The defects of spatters and pits could be observed in the surface of
weld seam. The HAZ of near Al-Si coating firstly occurred in Fig. 5(c-2)
and then extended to 3.8 mm in Fig. 5(c-3) due to the excessive heat
input. Although the wettability was gradually improved as the heat
input increases from 161.8 J/mm to 236.1 J/mm, the generation of
pores in weld seam, large brittle IMC and residual stress in brazed
interface was extraordinarily facilitated. Particularly, when laser power
increased to 900 W, a laser spot appeared in brazed interface in Fig. 5(d-
3), which was the directed evidence of strong intermetallic reaction
between aluminum and steel. The consecutive cracks spontaneously
propagated along the interface. This interesting fact also revealed that
Fig. 7. The effect of heat input on the thickness of IMC layer.

5
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

Table 3 average thickness of IMC layer. The latter increased with the increase of
The atomic percentage of different zones marked in Fig. 6(d-f). the former due to the accelerated diffusion of Al and Fe atoms across at
Points Fe Al Si Possible phases the Al/steel interface. The detailed analysis will be carried out below.
Fig. 6(a-c) shows the interface microstructure of joint S1. The
1 16.51 77.45 6.04 τ5-Fe2Al8Si
2 24.96 68.72 6.32 θ-Fe(Al, Si)3 lamellar IMC was observed in head region and foot region while a
3 16.93 77.62 5.45 τ5-Fe2Al8Si mixture of acicular and lamellar IMC was observed in central region.
4 25.55 69.48 5.12 θ-Fe(Al, Si)3 During laser/DP-MIG welding-brazing, the highest temperature of
5 28.31 66.56 5.13 η-Fe2(Al, Si)5 hybrid heat source occurred near central region of interface where the
6 17.84 77.24 4.92 τ5-Fe2Al8Si
growth of the IMC layer was sharply driven. The IMC thickness of central
region has reached 4.1 μm. Similarly, it can be seen that the IMC shapes
there is a critical threshold for laser power in this study. of joint S2 (Fig. 6(d-f)) is similar to that of joint S1 (Fig. 6(a-c)), as the
For all the cases discussed above, the welding appearance is posi­ heat input increases from 135.5 J/mm to 161.8 J/mm. It can be found
tively correlated to cross-sectioned macrostructure. An excellent that the IMC thicknesses increased slightly, implying that the formation
appearance normally indicates a satisfactory cross section. On the other of IMC could be slowly controlled within the limit of 161.8 J/mm. The
hand, the inadequate heat input can lead to poor wettability and uneven different IMC layers were marked in white dot line in accordance with
appearance whereas excessive heat input will cause indistinct welding the diffracted color in back-scattered electron (BSE). The details of
ripples and defects of large residual stress and pores. These two cases phases identification will be discussed in following paragraph. Fig. 6(g-
aren’t an ideal one when adjusting heat input. Hence, it is supposed to i) show the interface microstructure of joint S3 when the heat input was
make a compromise in heat input so as to achieve an excellent 202.8 J/mm. The mixture of acicular and lamellar IMC with a thickness
appearance. of 4.1 μm begun to appear at the head region due to the increase of heat
input. The similar structure except for a thicker size of 6.5 μm was
observed in central region. To make matter worse, as the heat input
3.2. Interface microstructure
increased to 236.1 J/mm by increasing laser power, the formation of
IMC layers was quickly developed in excess of 10 μm in Fig. 6(j-l). The
In order to understand the bonding mechanism of interface, the
defects of chipped IMC and consecutive cracks were found, which
interface microstructure of head, central, and foot regions was given in
catastrophically deteriorated the tensile-shear strength. The excessive
Fig. 6. The noticeable IMC layer was observed at the brazed interface
heat input was not conducive to the improvement of welded quality.
due to the metallurgical reactions of aluminum and steel. The temper­
The IMC phases marked in Fig. 6(d-f) was discussed to identify its
ature gradient differentia resulted in the various IMC thicknesses and
types. Table 3 lists the EDS results of different zones marked in Fig. 6(d-
structures from head region to foot region [33]. In vertical comparison
f). In heat region interface, the lamellar IMC can be divided into two
with different heat input, the variation of IMC thickness was summa­
zones according to the different diffraction colors. The Si content of the
rized in Fig. 7. The heat input played a significant role in dominating the

Fig. 8. The microstructure evolution of the joint S2: (a) the welded interface, (b) the magnification of dotted region (b) marked in (a), (c) FZ, (d) the of brazed
interface, (e) UCHAZ (i.e., region (e)) marked in (d), (f) ICHAZ (i.e., region (f)) marked in (d).

6
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

Fig. 9. The microhardness profile of the joint S2.

zone of point 1 adjacent to fusion zone of aluminum is enough to form an

Al-Fe-Si ternary IMC. The latter can be basically identified as τ5-Fe2Al8Si
according to the Al-Si-Fe ternary phase diagram [34] and EDS analysis
(Table 3). The similar discussions have been reported by Springer et al.
[35]. In contrast, the Si content in the zone of point 2 is lack for the Fig. 10. The influence of the heat input on the tensile-shear strength and
formation of an Al-Fe-Si ternary IMC, which tends to cause the formation porosity percentage.
of a Si-containing Al-Fe IMC. The maximum solubility of the Si element
in FeAl3 is 7.4% by means of Si atoms replacing Al atoms [36]. There­ steel when a sound brazed interface is formed. Fig. 8(d) shows the
fore, the phase of point 2 can be considered as θ-Fe(Al,Si)3 according to overall inspection of the brazed interface. The HAZ of the brazed
the atom percentage of point 2 (i.e., 24.96% Fe, 68.72% Al and 6.32% interface is marked by white dot line, which is divided into upper-
Si). Sequentially, the atom percentage of point 3 and 4 in the magnifi­ critical HAZ (UCHAZ) and inter-critical HAZ (ICHAZ). Fig. 8(e) shows
cation of central interface is detected, which is similar to that of point 1 the microstructure of UCHAZ, which consists of martensite (M). This
and 2. Thus it is reasonable to affirm that the phases of point 3 and 4 are area adjacent to the fusion zone experiences a complete austenization
τ5-Fe2Al8Si and θ-Fe(Al,Si)3 respectively. The zone of point 5 adjacent to due to its peak temperature ranging from Ac3 and melting point of steel.
the steel side directly encounters the infiltration of Fe element. The atom The austenite grain has a little time to grow, which causes the trans­
percentage of point 5 is adjusted to 28.31% Fe, 66.56% Al and 5.13% Si. formation of martensite. From the SEM image of ICHAZ (Fig. 8(f)), the
The atom ratio of Al plus Si to Fe is around 2.5. The Si element in Fe2Al5 island martensite (M) is randomly dispersed on the ferrite (F) matrix. An
has a maximum solubility of 6% was reported by Windmann et al. [37]. incomplete austenization in the ICHAZ results in the formation of
Hence, The phases of point 5 seems to well match with η-Fe2(Al,Si)5. In martensite due to its peak temperature experiencing in the range of Ac1
foot region of interface, the phase of point 6 is easily identified as τ5- and Ac3. The similar phenomenon has been discussed by Guo et al. [38],
Fe2Al8Si due to the enough Si content. Finally, three kinds of IMC types they researched the microstructures and mechanical of the S960 high
(i.e., τ5-Fe2Al8Si, θ-Fe(Al,Si)3, and η-Fe2(Al,Si)5) are actually confirmed strength steel using laser welding and GMAW.
in this work.

3.4. Hardness profiles

3.3. Microstructure analysis
Fig. 9 shows microhardness profile of the joint S2 with a measured
The microstructure evolution of the joint S2 was given in Fig. 8. The path in vertical and horizontal direction. Generally, there is a one-to-one
microstructure of the welded interface on aluminum side and the brazed relationship between the microstructure and microhardness. The mea­
interface on steel side were given in Fig. 8(a-c) and Fig. 8(d-f), respec­ surement in vertical direction primarily indicates the hardness variation
tively. From Fig. 8(a, b), the welded joint consists of fusion zone (FZ), of the brazed joint. The hardness measurement skips the thin IMC layers
heat affected zone (HAZ) and base metal (BM). The HAZ can be divided due to the mismatch of IMC thickness and Vicker indenter. The hardness
into columnar grain zone (CGZ) and melting zone (MZ). During the re- of BM steel is around 210 HV while that of UCHAZ is about 550 HV. This
solidification process, the crystal behavior is mainly determined by is due to the transformation from ferrite and pearlite into martensite.
the temperature gradient and solidified rate. The grain of CGZ tends to The fusion zone has a hardness of approximately 80 HV on account of
grow in the direction perpendicular to solid/liquid interface due to the the fine equiaxed grains. On the other hand, the maximum hardness of
maximum temperature gradient of weld pool. This causes the formation about 100 HV locates in the fusion zone/aluminum interface where
of columnar grain. However, when the columnar grain epitaxially ex­ coarse columnar grains present in the CGZ and MZ.
tends toward the FZ, the equiaxed grains (see Fig. 8(c)) is formed due to
the excessive supercooling degree. The latter attributes to the decrease 3.5. Mechanical property
of the temperature gradient and the increase of solidified rate. The
particles in FZ have been affirmed as Al-Si eutectics in our previous work Fig. 10 shows the influence of heat input on tensile-shear strength
[22]. and porosity percentage. As shown in Fig. 10(a), the tendency of tensile-
During welding-brazing process, the formation of solid/liquid shear strength increases first and then decreases as a function of heat
interface is due to the atom diffusion between the liquid fusion zone and input. The maximum tensile-shear strength of 168.8 MPa could be
solid steel. Thus, the peak temperature is less than the melting point of achieved when the heat input is 161.8 J/mm, which reached 80% tensile

7
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

strength of 5052 base materials. Two failure modes (see Fig. 10(b)) are
observed among all the cases: interface failure (IF) and HAZ failure (HF).
The fracture surface of failure modes will be examination in below
section. When the heat input is less than 161.8 J/mm, the tensile-shear
strength increases with the increase of heat input. The failure mode is
predominated by IF mode. The appropriate increase of heat input is
beneficial to improve the weld width with a slight increasing of IMC
Fig. 11. The schematic diagram of stress analysis under the applied load. thickness, strengthening the strength of interface. The previous discus­
sions in Sections 3.1 and 3.2 have been stated these facts. As the heat

Fig. 12. The fracture morphology of the HAZ failure (HF) mode: (a, b) different area of fracture surface, (c) a high magnification of dimples, (d) the EDS result of red
dotted square marked in (a).

Fig. 13. The fracture morphology of the interface failure (IF) mode: (a-c) different area of fracture surface, (d) the EDS result of red dotted square marked in (a).

8
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

input exceeds 161.8 J/mm or higher, the tensile-shear strength gradu­ 4. Conclusions
ally decreases and the porosity percentage maintains in an extremely
high level over 6%. Despite the formation of spontaneous cracks (1) The heat input plays a significant role in determining the welding
inducing by the large brittle IMC and rather high residual stress weakens performance during the laser/DP-MIG hybrid welding-brazing of
the Al/steel interface strength, the joint fractures at the HAZ of welded aluminum to steel. The satisfactory appearance with stable rip­
joint rather than Al/steel interface. This interesting phenomenon may be ples and the maximum joint strength of 168.8 MPa were obtained
explained by the reason that higher porosity percentage (beyond 6%) when the heat input with a well match between laser and arc
sharply depressed the strength near HAZ of welded joint. The pore parameters was 161.8 J/mm.
concentration near HAZ is considered as a preferential region to fracture (2) The IMC thickness increases with the increase of heat input, but
during tensile-shear tests. Hence, the joints basically fracture near HAZ the formation of IMC can be slowly controlled within a low heat
of welded joint. input. Three IMC phases were nicely confirmed in detail: τ5-
The failure mode dramatically transforms from IF to HF when the Fe2Al8Si, θ-Fe(Al,Si)3 and η-Fe2(Al,Si)5.
heat input increases. It can be found that the IF mode does not represent (3) Limiting peak temperature below steel melting point is necessary
a low tensile-shear strength compared with HF mode. To furtherly un­ for the formation of sound brazed interface. The microstructure
derstand the determined factors of the two failure modes, the stress of UCHAZ consists of martensite whereas that of ICHAZ consists
analysis should be carried out. Fig. 11 shows the schematic diagram of of ferrite and martensite due to different peak temperatures.
stress analysis under the loading state. The ultimate stress of interface (i. (4) There are two failure modes among all the cases: interfacial
e., F2) is mainly dependent on the adhesion weld width and IMC failure and HAZ failure. The brazed interface and pore concen­
thickness. Another stress is the ultimate stress of welded joint near HAZ tration are the two weak regions to fracture preferentially ac­
(i.e., FN), which is commonly slashed by the higher porosity percentage. cording to the stress analysis. The IF mode is a typical
This is because the present of pores directly decreases the stressed area intergranular fracture, and its cracks initiates from notch of weld
during the tensile-shear tests. Therefore, the failure modes are deter­ joint and then redirects toward the region of pore concentration.
mined by their competitive relationship. For example, the ultimate The HF mode is a mixture of major ductile fracture and minor
stress of interface in joint S1 is less than that of HAZ in joint S1 due to the cleavage fracture, and its cracks may initiate from notch of weld
relatively narrow weld width, resulting in the occurrence of IF mode. joint and then propagate along the IMC throughout the whole
interface.
3.6. Fracture analysis

Different failure modes presented different fracture positions and Declaration of Competing Interest
fracture surface. In the HAZ failure (HF) mode, the original cracks easily
initiated from notch of weld joint due to the concentration stress under The authors declare that they have no known competing financial
the applied load, and then redirected toward the region of pore con­ interests or personal relationships that could have appeared to influence
centration. This assumption was demonstrated by the fact that HF mode the work reported in this paper.
surface consisted of dimples and dense pores at the parameters of S3
sample in Fig. 12(a). As shown in Fig. 12(b), there is a small cleavage Acknowledgement
terrace around some pore characteristic. Fig. 12(c) displays a high
magnification of dimples feature. These dimples indicate different for­ This study was supported by the National Key Research and Devel­
mation mechanism. The deep dimples with the rough and irregular opment Program of China (2017YFB1104801), the Fundamental
surface may belong to intergranular dimple while the shallow dimples Research Funds for the Central Universities of Central South University
with smooth surface may belong to transgranular dimple. The second- (2021zzts0130), Postgraduate Scientific Research Innovation Project of
phase particle in the bottom of dimple is the direct evidence of ductile Hunan Province (CX20210206).
fracture. Hence, the HF mode is a mixture of major ductile fracture and
minor cleavage fracture. Fig. 12(d) shows that the Si content of dimples References
relatively higher than that of aluminum BM. The presence of pores and
higher Si content was the directed evidence that cracks propagated into [1] P. Wang, X. Chen, Q. Pan, B. Madigan, J. Long, Laser welding dissimilar materials
of aluminum to steel: an overview, Int. J. Adv. Manuf. Technol. 87 (9-12) (2016)
pore concentration. The similar analysis has been reported by Yagati 3081–3090.
et al. [39], they found that cracks should propagate into the weld bead [2] T. Matsuda, R. Hatano, T. Ogura, R. Suzuki, H. Shoji, T. Sano, M. Ohata, A. Hirose,
near HAZ. Effect of mismatch in mechanical properties on interfacial strength of aluminum
alloy/steel dissimilar joints, Mater. Sci. Eng., A 786 (2020), 139437.
Fig. 13 shows the fracture morphology of interface failure (IF). The [3] G. Liu, X. Gao, C. Peng, Y. Huang, H. Fang, Y. Zhang, D. You, N. Zhang,
features of cleavage terrace, tear ridge and river patters were observed Optimization of laser welding of DP780 to Al5052 joints for weld width and lap-
in Fig. 13(a, b), exhibiting a brittle fracture. The direction of cleavage shear force using response surface methodology, Opt. Laser Technol. 126 (2020),
106072.
terrace perpendicular to that of the applied load. The fracture might [4] Y. Li, Y. Liu, J. Yang, First principle calculations and mechanical properties of the
occur in the θ-Fe(Al,Si)3 phase according to the EDS result in Fig. 13(d). intermetallic compounds in a laser welded steel/aluminum joint, Opt. Laser
These suggested an intense tearing by splitting the welded interface Technol. 122 (2020), 105875.
[5] Q. Guan, J. Long, P. Yu, S. Jiang, W. Huang, J. Zhou, Effect of steel to aluminum
from root region to toe region of weld joint during tensile-shear tests. An
laser welding parameters on mechanical properties of weld beads, Opt. Laser
ambiguous grain outline was extracted in Fig. 13(c), which indicated an Technol. 111 (2019) 387–394.
intergranular fracture. Therefore, it can be implied that the cracks may [6] B. Mezrag, F. Deschaux-Beaume, L. Sabatier, B. Wattrisse, M. Benachour,
Microstructure and properties of steel-aluminum cold metal transfer joints,
initiate from notch of weld joint, and then propagate along the IMC
J. Mater. Process. Tech. 277 (2020), 116414.
throughout the whole interface. The defects of microcracks and voids in [7] G. Sierra, P. Peyre, F.D. Beaume, D. Stuart, G. Fras, Steel to aluminium braze
the IMC accelerate this destruction. Ma et al. [40] concluded that the welding by laser process with Al-12Si filler wire, Sci. Technol. Weld. Joining 13 (5)
cracks were easy to propagate in the IMC layers since it initiated from (2008) 430–437.
[8] H. Dong, Y. Li, P. Li, X. Hao, Y. Xia, G. Yang, Inhomogeneous microstructure and
the weld root. mechanical properties of rotary friction welded joints between 5052 aluminum
alloy and 304 stainless steel, J. Mater. Process. Technol. 272 (2019) 17–27.
[9] C. Wang, L. Cui, G. Mi, P. Jiang, X. Shao, Y. Rong, The influence of heat input on
microstructure and mechanical properties for dissimilar welding of galvanized steel
to 6061 aluminum alloy in a zero-gap lap joint configuration, J. Alloy. Compd. 726
(2017) 556–566.

9
X. Cao et al. Optics and Laser Technology 150 (2022) 107939

[10] G. Sierra, P. Peyre, F. Deschaux-Beaume, D. Stuart, G. Fras, Steel to aluminium key- [25] Y. Ding, Z. Shen, A.P. Gerlich, Refill friction stir spot welding of dissimilar
hole laser welding, Mater. Sci. Eng., A 447 (1-2) (2007) 197–208. aluminum alloy and AlSi coated steel, J. Manuf. Process 30 (2017) 353–360.
[11] M. Pouranvari, M. Abbasi, Dissimilar gas tungsten arc weld-brazing of Al/steel [26] R. Huang, C. Tan, Y. Sun, X. Gong, L. Wu, B. Chen, H. Zhao, X. Song, Influence of
using Al-Si filler metal: Microstructure and strengthening mechanisms, J. Alloy. processing window on laser welding-brazing of Al to press-hardened 22MnB5 steel,
Compd. 749 (2018) 121–127. Opt. Laser Technol. 133 (2021), 106566.
[12] J. Liu, S. Jiang, Y. Shi, Y. Kuang, G. Huang, H. Zhang, Laser fusion-brazing of [27] M. Windmann, A. Röttger, H. Kügler, W. Theisen, F. Vollertsen, Laser beam
aluminum alloy to galvanized steel with pure Al filler powder, Opt. Laser Technol. welding of aluminum to Al-base coated high-strength steel 22MnB5, J. Mater.
66 (2015) 1–8. Process. Technol. 217 (2015) 88–95.
[13] G. Casalino, A. Angelastro, P. Perulli, C. Casavola, V. Moramarco, Study on the [28] D.C. Saha, E. Biro, A.P. Gerlich, N.Y. Zhou, Fusion zone microstructure evolution of
fiber laser/TIG weldability of aisi 304 and aisi 410 dissimilar weld, J. Manuf. fiber laser welded press-hardened steels, Scr. Mater. 121 (2016) 18–22.
Process. 35 (2018) 216–225. [29] W.-J. Cheng, C.-J. Wang, Microstructural evolution of intermetallic layer in hot-
[14] A.H. Faraji, M. Moradi, M. Goodarzi, P. Colucci, C. Maletta, An investigation on dipped aluminide mild steel with silicon addition, Surf. Coat. Technol. 205 (19)
capability of hybrid Nd:YAG laser-TIG welding technology for AA2198 Al-Li alloy, (2011) 4726–4731.
Opt. Lasers Eng. 96 (2017) 1–6. [30] U. Dilthey, A. Wieschemann, Prospects by combining and coupling laser beam and
[15] H. Huang, P. Zhang, H. Yan, Z. Liu, Z. Yu, D. Wu, H. Shi, Y. Tian, Research on weld arc welding processes, Weld. World 44 (2000) 37–46.
formation mechanism of laser-MIG arc hybrid welding with butt gap, Opt. Laser [31] S. Yang, J. Zhang, J. Lian, Y. Lei, Welding of aluminum alloy to zinc coated steel by
Technol. 133 (2021), 106530. cold metal transfer, Mater. Des. 49 (2013) 602–612.
[16] T. Yan, X. Zhan, Q. Gao, F. Wang, W. Ling, Influence of laser power on molten pool [32] X. Zhan, Y. Zhao, Z. Liu, Q. Gao, H. Bu, Microstructure and porosity characteristics
flow field of laser-MIG hybrid welded Invar alloy, Opt. Laser Technol. 133 (2021), of 5A06 aluminum alloy joints using laser-MIG hybrid welding, J. Manuf. Process.
106539. 35 (2018) 437–445.
[17] M. Gao, C. Chen, S. Mei, L. Wang, X. Zeng, Parameter optimization and mechanism [33] H. Xia, X. Zhao, C. Tan, B.o. Chen, X. Song, L. Li, Effect of Si content on the
of laser-arc hybrid welding of dissimilar Al alloy and stainless steel, Int. J. Adv. interfacial reactions in laser welded-brazed Al/steel dissimilar butted joint,
Manuf. Technol. 74 (1-4) (2014) 199–208. J. Mater. Process. Technol. 258 (2018) 9–21.
[18] J. Xue, Y. Li, H. Chen, Z. Zhu, Effects of heat input on wettability, interface [34] M.C.J. Marker, B. Skolyszewska-Kühberger, H.S. Effenberger, C. Schmetterer, K.
microstructure and properties of Al/steel butt joint in laser-metal inert-gas hybrid W. Richter, Phase equilibria and structural investigations in the system Al-Fe-Si,
welding-brazing, J. Mater. Process. Technol. 255 (2018) 47–54. Intermetallics 19 (12) (2011) 1919–1929.
[19] J. Li, H. Li, H. Wei, Y. Ni, Effect of pulse on pulse frequency on welding process and [35] H. Springer, A. Kostka, E.J. Payton, D. Raabe, A. Kaysser-Pyzalla, G. Eggeler, On
welding quality of pulse on pulse MIG welding-brazing of aluminum alloys to the formation and growth of intermetallic phases during interdiffusion between
stainless steel, Int. J. Adv. Manuf. Technol. 87 (1-4) (2016) 51–63. low-carbon steel and aluminum alloys, Acta Mater. 59 (4) (2011) 1586–1600.
[20] M. Sen, M. Mukherjee, S.K. Singh, T.K. Pal, Effect of double-pulsed gas metal arc [36] S.P. Gupta, Intermetallic compound formation in Fe-Al-Si ternary system: Part I,
welding (DP-GMAW) process variables on microstructural constituents and Mater. Char. 49 (4) (2002) 269–291.
hardness of low carbon steel weld deposits, J. Manuf. Process. 31 (2018) 424–439. [37] M. Windmann, A. Röttger, W. Theisen, Phase formation at the interface between a
[21] L.L. Wang, H.L. Wei, J.X. Xue, T. DebRoy, A pathway to microstructural refinement boron alloyed steel substrate and an Al-rich coating, Surf. Coat. Technol. 226
through double pulsed gas metal arc welding, Scr. Mater. 134 (2017) 61–65. (2013) 130–139.
[22] X. Cao, Z. Yi, C. Xu, Z. Luo, Ji’an Duan, Study on laser/DP-MIG hybrid welding- [38] W. Guo, L. Li, S. Dong, D. Crowther, A. Thompson, Comparison of microstructure
brazing of aluminum to Al-Si coated boron steel, J. Manuf. Process. 64 (2021) and mechanical properties of ultra-narrow gap laser and gas-metal-arc welded
333–340. S960 high strength steel, Opt. Lasers Eng. 91 (2017) 1–15.
[23] K. Zhang, L. Wu, C. Tan, Y. Sun, B.o. Chen, X. Song, Influence of Al-Si coating on [39] K.P. Yagati, R.N. Bathe, K.V. Rajulapati, K.B. Sankara Rao, G. Padmanabham,
resistance spot welding of Mg to 22MnB5 boron steel, J. Mater. Process. Technol. Fluxless arc weld-brazing of aluminium alloy to steel, J. Mater. Process. Tech. 214
271 (2019) 23–35. (12) (2014) 2949–2959.
[24] M. Kang, C. Kim, Joining Al 5052 alloy to aluminized steel sheet using cold metal [40] H. Ma, G. Qin, L. Wang, X. Meng, L. Chen, Effects of preheat treatment on
transfer process, Mater. Des. 81 (2015) 95–103. microstructure evolution and properties of brazed-fusion welded joint of aluminum
alloy to steel, Mater. Des. 90 (2016) 330–339.

10