j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2

Available online at [Link]

journal homepage: [Link]/locate/jmrt

Original Article

Element diffusion and microstructure evolution at

interface of stainless steel/Ti alloy joint by laser
welding with AgCuTi filler metal

DeShui Yu a, Yan Zhang a,*, Seyed Reza Elmi Hosseini c,**, JianPing Zhou a,
DaQian Sun b
a
School of Mechanical Engineering, Xinjiang University, Wulumuqi 830000, China
b
Key Laboratory of Automobile Materials, School of Materials Science and Engineering, Jilin University, Changchun
130022, China
c
School of Metallurgy and Materials Engineering, Iran University of Science and Technology (IUST), Narmak,
Tehran, Iran

article info abstract

Article history: Joining of 304 stainless steel and TC4 titanium alloy was of great interest for applications in
Received 10 March 2023 the marine, and aerospace fields. However, the direct welding process of steel and titanium
Accepted 23 April 2023 will generate a large number of brittle TieFe intermetallic compounds results in crack in
Available online 26 April 2023 joint. In this study, in order to avoid the formation of TieFe intermetallic compounds, the
connection between stainless steel and titanium alloy is successfully realized by adding
Keywords: AgCuTi filler metal. A section of unmelted stainless steel was retained by laser offset
AgCuTi filler welding, and the filler was diffusion bonding with the base metals on both sides by heat
Stainless steel conduction, ensuring that the filler was not melted. Since Fe and Cu elements have high
Titanium alloy solid solubility in liquid state, so the stainless steel and filler interface formed a good
Laser welding connection. There was no TieFe intermetallic compounds found in the joint, and Ti2Cu and
Diffusion coefficient AgTi2 intermetallic compounds with better ductility was formed at the filler-TC4 interface.
Numerical simulation In order to further explore the element diffusion distribution at the interface of the joint.
The temperature distribution model in the fusion zone was established, and obtained the
temperature field of the joint. The diffusion coefficients of Fe, Cu and Ti elements were
calculated by Arrehenius formula. It was verified that the diffusion of Ti elements in the
filler resulted in the aggregation of Ti elements at the filler-TC4 interface. The micro-
structure mechanism for formation of the brazed weld was proposed.
© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license ([Link]

* Corresponding author.
** Corresponding author.
E-mail addresses: yanzhang4967@[Link] (Y. Zhang), Elmihosseini@[Link] (S.R.E. Hosseini).
[Link]
2238-7854/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://
[Link]/licenses/by-nc-nd/4.0/).
6464 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2

of TC4 to AISI 304L with brass interlayer was attempted by

1. Introduction different pressure. It found that increasing the pressure had
an adverse effect on shear strength for the samples processed
Stainless steel (SS) and titanium (Ti) alloy are widely used for at higher temperatures. Norouzi et al. [14] studied the joining
marine and aerospace manufacturing [1,2], the SS/Ti com- between TC4 and AISI 304 through a copper interlayer by
posites structure can combine the wear resistance and low transient liquid phase welding. The results showed that the
cost of SS with the low density and good corrosion resistance width of the joint was increased and then decreased with
of Ti alloy when SS is welded with Ti alloy, and give full play to enhancing the welding time. The welding time was 60 min,
the complementary advantages of the two metals in perfor- which resulted in a joint solidified isothermally, and the
mance and economy. Therefore, the joining between SS/Ti fracture modes are brittle and mixed brittle-ductile.
can not only reduce the cost, but also obtain components with In order to suppress the growth of the brittle phase, diffu-
excellent comprehensive performance, which has important sion welding, friction stir welding have been widely used to
application value. However, there are more challenges faced join SS and Ti alloy. However, these methods are not suitable
during the joining process of SS/Ti dissimilar metals due to for joining tailored blanks and flexible manufacture due to the
the significant differences in the mechanical, metallurgical, limitation of especial joint configuration. As a heat source of
and physical properties of the metals of the dissimilar com- welding dissimilar metals, laser has been widely applied in
bination [3]. Therefore, it is necessary to obtain high quality the fields of dissimilar metals welding. Laser welding is a joint
joints in order to use their maximum potentials in the SS/Ti method that performs the laser heat to melt and join the
joints. workpieces, and the laser welding has a high energy density
Direct fusion welding processes of dissimilar metals which and a small heating range, which can accurately control the
involve melting of base metals were not successful in joining heat flow and obtain a welded joint with good performance
the SS/Ti joints [4], the formation of a number of intermetallic [15]. Laser welding has been widely used in the same alloy
compounds (IMCs) in the weld pool resulted in the embrit- components, and the strength of welded joints can be close to
tlement of joints [5,6]. These brittle TieFe IMCs are very hard or even higher than that of base metal [16,17]. Laser welding
and weaken the mechanical properties of the SS/Ti joints. has flexible requirements for workpieces, and lap mode can be
Therefore, joining of SS/Ti dissimilar metals by direct fusion used for welding. Chen et al. [18] performed laser welding SS/
welding has a metallurgical challenge due to unavoidable Ti by putting the CP Ti sheet on the top of the SS sheet with an
formation of brittle IMCs [7,8]. So it is necessary to control overlap of 10 mm. Throughout the work, the laser beam was
effectively formation and growth of TieFe IMCs. Gotawala focused on the top surface of SS sheets and positioned at the
et al. [9] welded SS/Ti dissimilar metals by friction stir weld- edge of the Ti sheet. Cracks are generated in the joint when
ing. It was found that the decrease of interface temperature normal rectangular pulse is adopted. When ramp-down pulse
led to the decrease of FeTi layer thickness in the SS/Ti joint is used for welding, the heat input decreases and the cracks
and the increase of tensile strength of the joint under low disappear. Some studies indicated that the formation of brittle
speed welding mode. However, the fracture strain for both the TieFe IMCs can be suppressed by using transition interlayer.
welds was very low due to the presence of brittle FeTi IMC at For this purpose, Nb, V and Cu interlayers were used [19,20].
the SS-Ti interface. Liu and Fujii [10] also used the low rotation Fang et al. [21] investigated the laser welding of TC4 and AISI
speed and high pressure to friction weld TC4 alloy and SUS 316L SS by inserting Cu/Nb bilayer, and there are no TieFe
316L SS, and obtained the SS/Ti joint. The low rotation speed IMCs formed in the joint. At the state of high temperature,
provided a low temperature rising rate and a low peripheral Ti alloy and Nb melt and mix with each other because of high
velocity corresponding to a low shear strain rate between solid solubility. Nevertheless, low solid solubility results in
faying surfaces at the weld interface periphery. It found a thin occurrence of segregation and thermal stress during the pro-
and strong IMCs layer formed in the joint. Both researchers cess of cooling. The high heat input and stress induced-
used low speed friction welding to avoid excessive heat input shrinkage of molten metals cause the rupture of solidified
to suppress the formation of large amounts of TieFe IMCs. grains along molten grain boundary [22]. The laser brazing
Kumar [11] studied the friction weld joining TC4 and 304L SS method can suppress effectively the growth of brittle IMCs.
with copper as the intermediate metal. The process output Hence, tensile strength of the joint by laser brazing was
was compared. The formation of TieFe IMCs was avoided by enhanced significantly. In addition, this method provides
friction welding with 20 mm thick copper, and increased good adaptability and a good relation between weight reduc-
strength of joints. Gao [12] performed the direct diffusion tion and costs in serial production.
welding of TA2 and S316, it found that Cr atoms are enriched In this work, the AgCuTi filler was used for laser welding of
at the interface between SS and Ti alloy, and then react with Ti SS and TC4. The excessive heat input at the interface during
atoms to form CreTi IMCs, which hinders the diffusion path of laser welding can cause element segregation and large ther-
Ti atoms. Hence the main atomic diffusion behavior is the mal stress, so the laser beam is welded to the SS side. The
diffusion of Fe and Cr atoms to the Ti alloy. The distribution joining between the filler metal and the base metals on both
phase of the interface in joint is Fe2Ti, Cr2Ti, FeTi/b-Ti. The sides is realized through the conduction of heat, which avoids
aggregation interface of a large number of brittle IMCs be- the generation of cracks caused by the high-density heat of
comes the most likely fracture zone in the joint. In order to the laser focused on the interface. The Ag element in the filler
avoid the formation of brittle IMCs during diffusion welding can improve the wetting and spreading of the interface, which
joint, researchers also filled the interlayer in the diffusion is conducive to the formation of SS/Ti joints. Ti elements of
welding process. Vigraman [13] performed diffusion welding
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2 6465

with the purity of 99.99% was used as a shielding gas, and its
Table 1 e Compositions of 304 SS, TC4 alloy and AgCuTi
flow rate is 10 L/min. Tensile strength was evaluated by using
filler metal used in this study.
stretching machine (Instron 5967B). The microstructure
Alloys Elements (wt.%)
morphology and chemical compositions of joints were ob-
Ti Al Fe V Ag Cu Cr Ni Mn tained by optical microscope (OM), scanning electron micro-
304 SS e e Bal. e e e 19.5 9.2 2.3 scope (SEM) and energy dispersive spectroscopy (EDS). The
TC4 Bal. 5.5 0.3 3.8 e e e e e whole cross section of joint obtained by VHX-5000 super deep
Filler 1.9 e e e 2.46 Bal. e e e scene 3D microscope.

the filler can improve metallurgical bonding with TC4 alloy. 2.2. Simulation and calculation
Therefore, it is a better choice to use AgCuTi filler to join SS/Ti
dissimilar metals. The use of AgCuTi can effectively avoid the According to the actual welding situation, the mathematical
formation of a large number of TieFe IMCs in the joint, and the model of the welded workpiece is established by SYSWELD.
good ductility of AgeCu based filler metal also reduces the The heat source model adopted the Gaussian body heat
welding stress in the joint. In order to further understand the source. Heat exchange coefficient was 80 W/m2. The nonlinear
interface forming mechanism of the welding process. The change of thermophysical parameters with temperature and
temperature field was simulated by SYSWELD, and the diffu- the influence of latent heat of phase change on temperature
sion coefficient of interface elements was calculated by were considered. According to the temperature of the inter-
Arrehenius formula. The diffusion of Ti element in the inter- face in the simulation result, the numerical calculation of the
face is further expounded, which promotes the theoretical of interface element diffusion is carried out. The diffusion
metallurgical bonding of the SS/Ti joint. driving force of interfacial diffusion elements in this study
depends on the diffusion coefficient D, crystal structure,
concentration gradient, joining state and other interfacial
2. Materials and methods conditions of the base metals and filler elements. The element
diffusion coefficient D depends on the diffusion activation
2.1. Materials energy Q and the heating temperature T. The diffusion acti-
vation energy Q is the minimum energy during the atomic
TC4 alloy and 304 SS plates with thickness of 1 mm were transition in the diffusion process, which is generated to
selected as the laser joining materials. In this study, AgCuTi overcome the effect of the surrounding atoms on it. Diffusion
with the thickness of 0.1 mm was used as filler metals. The coefficient D is the average number of particles passing
compositions of materials and material properties listed in through a certain plane in unit time. Diffusion coefficient is a
Table 1 and Table 2. It can be seen from Table 2 that the dynamic value changing with diffusion process [23]. It is very
thermal conductivity and linear expansion coefficient of the important to determine the diffusion coefficient in this study.
two materials are nearly doubled, which is also the reason Through the diffusion coefficient, the concentration distri-
why the two materials are difficult to weld directly. Fig. 1(a) bution of Fe and Ti elements near the joint interface, the
shows a schematic diagram of the welding process, and formation and width of compounds with the change of hold-
Fig. 1(b) present the details of welding SS/Ti procedures, the ing time and heating temperature can be specifically under-
laser beam focused at 1 mm position on SS side. Fig. 1(c) shows stood. The relationship between diffusion coefficient D,
the dimensions of the weld sheets and tensile test specimen, diffusion activation energy Q and temperature T satisfies the
4
1 mm stretch area used in the present study. Tensile Arrehenius index:
strength of the samples was measured by using stretching Q
D ¼ D0 eRT (1)
machine and the strain rate is 1 mm/min. Two samples of the
2 1
same parameters were tested and the average values were where, D0 is the pre-exponential factor (m $s ), R is the gas
recorded. The surfaces of specimen were ground to 1500 fine constant (8.314 J mol1 K1), Q is activation energy for growth
grit SiC paper to take oil and rust off and then cleaned with (J$mol1), and T is absolute temperature in Kelvin (K).
ethanol to completely remove impurities before welding. The The content of Fe, Ti, Cu atoms in the base metals and filler
fixture ensures that there is no gap when clamping the spec- in this experiment is more than 90%, which is large enough
imen, and prevents the liquid metal from dropping to form relative to the diffusion of atoms in the diffusion process and
holes during the formation of the weld. A 1.2 kW CW fiber has rich Fe, Ti, Cu diffusion atoms. Therefore, the diffusion
laser with wavelength of 1080 nm was employed. The spec- couple calculation model can be considered as infinite mate-
imen was welded at the power of 432 W, a welding speed of rial. Atomic diffusion can be solved by unsteady diffusion
750 mm/min and a defocusing distance of þ2.5 mm. Argon gas equation in one dimensional infinite medium [24,25]. The

Table 2 e Material properties of TC4 and 304 stainless steel.

Material Melting point Thermal conductivity Specific heat capacity Linear expansion coefficient Tensile strength
( C) (W$m1K1) (J$kg1K1) (106 K1) (MPa)
TC4 1650 6.4 536 8.7 1100
304 SS 1454 16.3 500 17.2 835
6466 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2

effect on the base metal was small as same as the brazing

weld process [27], so the grains in the unmelted SS did not
grow significantly. Fig. 2 (c) shows the diffusion model of heat
conduction in the joint. The unmelted SS acts as a bridge to
transfer the heat to the brazed weld, and the austenite den-
dritic grains on both sides of the fusion weld are distributed
vertically to the fusion line (Fig. 2(d)). The offset of laser was
the key to acquire a sound joint. SS/Ti joint remains unmelted
SS and realizes diffusion joining between SS-filler interface
and filler-Ti interface by heat conduction. Through the laser
offset at SS side to remain the unmelted SS avoids the crack
upon solidification occurred due to the formation of brittle
intermetallics and residual thermal stresses during direct
Fig. 1 e Schematic diagram of welding process and welding welding dissimilar metals [28], it helps to relieve the residual
joint. (a) laser offset welding process; (b) the details of stress that exists after high temperature welding at the joint.
welding SS/Ti dissimilar metals procedures; (c) size of Fig. 2(e) shows the macroscopic morphology of the brazed
sheets and tensile specimen. weld, and the brazed weld is similar to the diffusion welded
joint in the literature [13]. The SS-filler interface is uneven,
and there is a reaction layer at the interface, while the filler-Ti
interface on the right side is relatively flat. The AgCuTi filler
initial boundary condition (Eq. (3)) of one-dimensional infinite
has good wetting and spreading properties due to the pres-
unsteady diffusion equation is introduced into one-
ence of Ag and Ti elements, and Cu can ensure that enough
dimensional Fick second law (Eq. (2)), and the error function
heat was transferred to promote the formation of the interface
(Eq. (4)) is introduced to solve the error function (Eq. (5)) of one-
of the joints. As shown in Fig. 2(f), brazed weld of the joint got
dimensional infinite unsteady diffusion equation.
a good metallurgical bonding, and yellow dashed lines in
dC v2 C Fig. 2(f) shows the good spread of filler, the pit on SS side
¼D 2 (2) during grinding was filled during welding by filler. Further-
dt vx
more, AgCuTi filler acted as a barrier to prevent the massive


C0 ðx > 0Þ mutual atomic diffusion of Fe and Ti. The filler helps relieve
Cðx; tÞ ¼ (3)
0 ðx < 0Þ the thermal stress at the interface due to the filler consisting
of copper and silver which has good ductility, and preferential
Zb formation of TieCu IMCs with lower brittleness than TieFe
2 2
erfðbÞ ¼ pffiffiffiffi$ eh dh (4) IMCs in weld.
p
0
3.2. Temperature field simulation
 
C2 þ C1 C2  C1 x
Cðx; tÞ ¼  erf pffiffiffiffiffiffi (5)
2 2 2 Dt The welding process was simulated by SYSWELD simulation
software, and the numerical simulation of the temperature
field of the joint was obtained. The top view of temperature
3. Results and discussion simulation during laser welding shows in Fig. 3(a). The tem-
perature contour on Ti side decreases rapidly due to the
3.1. Morphology of joint thermal conductivity of SS alloy is twice that of Ti. The
Gaussian body heat source was used as the simulated heat
Fig. 2(a) shows the whole morphology of the SS/Ti joint using source model (Fig. 3(b)). In order to ensure a good match be-
AgCuTi filler, and the fusion weld, unmelted SS and brazed tween the actual weld and simulated weld morphology, the
weld in the joint are clearly distinguished. There are no de- heat transfer coefficient and thermal efficiency in the heat
fects such as pores and cracks in the cross section of the SS/Ti source model were corrected. The feasibility of the Gaussian
joint. The microstructure and boundary of the fusion weld can body heat source model was verified, so as to obtained the
be clearly seen by enlargement in Fig. 2(b). There are fine temperature field distribution law of SS/Ti joints, and pro-
columnar dendrites growing along the vertical direction to the vided theoretical basis and data support for calculation of
fusion line in the yellow dotted frame (see Fig. 2(b)). This is diffusion coefficient of elements. Fig. 3(b) shows the cross
because the temperature gradient of the material near the section of fusion weld under simulated state compared with
fusion line is large due to the cooling effect of the base metal the actual fusion weld morphology of joint. It is found that the
[26]. The nucleus formed on the surface of the base metal fusion weld morphology is basically the same, which can
attached to the fusion weld will preferentially grow along the verify the rationality of the Gaussian body heat source
direction of the maximum temperature gradient, and forming selected in the simulation. Therefore, parameters in this
columnar dendritic grains. Compared with the fusion weld, simulation test also can be used for the subsequent study of
the grains in the unmelted SS (heat affected zone) did not the temperature field and stress field of more complex struc-
show obvious grain coarsening due to the welding speed was tural components, such as the shape parameters of the heat
fast and only heated in the local area of the joint. The heat source model, heat transfer coefficient and thermal efficiency.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2 6467

Fig. 2 e (a) Cross-section of the welded joint; (b) OM image of fusion line and microstructure in fusion weld. Schematic
diagram of the (c) heat conduction and (d) formation of weld. (e) OM image of brazed weld; (f) microstructure of the brazed
weld.

Fig. 3 (c) shows the temperature field distribution of the extract the temperature thermal cycle data. Fig. 3(d) shows
joint at two moments. The molten pool only exists in a small that the heating and cooling curve on both sides of the brazed
area, and the isothermal lines are densely distributed in the weld. It can be found that the heating and cooling curves on
area directly affected by the laser heat source. The tempera- both sides of the brazed weld are almost completely over-
ture of purple part at moment 2 has exceeded the melting lapped. The peak temperature difference between SS-filler
point of 304 SS, the SS sheet completely penetrated. Since the and filler-Ti is only 10.5  C, owing to the filler is only 100 mm
laser is concentrated on the SS side, so the temperature is thick and the thermal conductivity of Cu is good, so the
concentrated, which further proves that the atomic motion at temperature difference between the two interfaces is small.
the SS-filler interface is intense and the reaction layer is The laser welding energy density is very large. Furthermore,
formed. It also proves that the welding process can effectively the metal can quickly weld penetration and rapid cooling.
realize the formation of the SS/Ti joint. In order to better Since the temperature gradient of laser welding is large, when
calculate the diffusion coefficient of elements, two points the heat source moves to the welding position, the tempera-
(Fig. 3 (c)) of SS-filler and filler-Ti interface were selected to ture will rise sharply and then decrease rapidly.
6468 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2

Fig. 3 e Temperature field simulation. (a) Top view of simulated temperature contour; (b) Gaussian body heat source and
comparison of simulated with actual weld morphologies; (c) temperature field of joint section at moment 1 and moment 2;
(d) comparison of temperature thermal cycles at moment 1 and moment 2.

3.3. Microstructures analysis Ti elements from the EDS line scanning (Fig. 4(i)), and the
average distribution of Ag element in the brazed weld.
The SEM image of joint is shown in Fig. 4(a). The brazed weld Fig. 5 shows the physical model formation of the brazed
formed defect-free interface. The joint divided into the reac- weld during laser welding. The heat is first transmitted to the
tion layer in the interface of SS-filler (zone I), the brazed weld zone I, and it can be seen from the temperature field (Fig. 3(a))
and the interface of filler-Ti (zone II) (Fig. 4(a)). As shown in that the thermal conductivity of the SS is higher than Ti, so the
Fig. 4(b), it can be seen that the reaction layer about 15 mm in diffusion reaction is intense in zone I. As show in Fig. 5(a), the
zone I. Since the zone I is near the laser heat source, the heat diffusion of Fe and Cu result in the formation of (Cu, Fe) solid
conduction is fast, and the atomic diffusion reaction is solution, g-Fe and a-Cu at the zone I, and it was relatively
intense. Several prominent dendritic grains growing along the disordered at the interface. With the heat conduction to zone
direction of temperature gradient on the left side of the re- II (Fig. 5(b)), the Cu of the filler and Ti of TC4 base metal also
action layer, and vertical growth to the reaction layer. Eutectic diffuse to each other (Fig. 5(c)). The content of Cu and Ti ele-
reaction of Cu and Ti atoms in the zone II and formation of ments are the most in zone II, so Ti2Cu and a-Ti are formed.
TieCu IMCs. EDS analysis of point A, B, C and D was measured According to the EDS line scanning analysis (Fig. 4(i)). It can be
and the results was listed in Table 3. The brazed weld of SS-Ti seen that there is Ag element at the zone II, since EDS point
alloy interface mainly consisted of g-Fe, (Cu,Fe) þ a-Cu, Ti2Cu, scan of B has 1.15% Ag and that the solubility of Ag and Fe is
a-Ti, (Cu, Ag) orderly from the SS to the Ti alloy. There is not only 0.02e0.07
104%, Ag and Cu still exist in a the form of
TieFe IMCs was formed in the joint. Through EDS map scan- solid solution in the weld. Combined with the AgeTi binary
ning of the brazed weld, it can be found that Cu element dif- phase diagram and the simulated temperature field in zone
fuses into zone I. Due to the diffusion of Cu element, the II(Fig. 3(d)), it can be inferred that there will be a small amount
reaction layer severe diffusion with Fe and Cu elements, of AgTi2 at the zone II (Fig. 5(d)).
realizing grain refinement and forming (Cu, Fe) solid solution Fig. 6 shows the microhardness test results of the joint (the
in the reaction layer at zone I. Fe element also has certain loading speed is 45 mm/s and the loading time remained 30 s).
diffusion in the filler (Fig. 4 (e)), but the Fe element did not It was found that the hardness of both sides of the brazed weld
diffuse into Ti alloy. Through the EDS analysis of the bright was lower than that of the base metal. As Fig. 4(i) shows that
layer in zone II, it was found that Ti element agglomerated. the diffusion of Cu element at the interface between SS, Ti
Massive Ti element of the filler migrated to the zone II. It can alloy and filler reduced the hardness of the interface, avoided
be seen that the filler can well prevent the diffusion of Fe and the harm of high hardness brittle IMCs to the joint. The high
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2 6469

Fig. 4 e SEM image of the joint. (a) brazed weld of the joint; (b) microstructure of zone I; (c) microstructure of zoneII. EDS map
scanning of (d) Cu, (e) Fe, (f) Ni, (g) Cr, (h) Ti. (i)EDS line scanning across the joint.

density and concentrated laser power focus on the fusion

weld, the grain size increases, and the grain boundary de-
creases, so the ability to resist dislocation movement of grains
decreases. Finally, the ability to resist local plastic deforma-
tion decreases. Thus, SS melting and resolidifying during
welding, which releases the internal stress and deformation
energy of the base metal during the initial cold working
hardening, resulting in the softening of the fusion weld and
the decrease of the microhardness value. It can be seen that
the hardness of TC4 alloy is higher than SS. The microhard-
ness of the fusion weld is significantly reduced compared with
SS. In addition, compared with the fusion weld, the micro-
hardness of the brazed weld is very low, because the filler it-
self is composed of softer materials. Therefore, the brazed
weld is more easily deformed to reduce the residual stress
inside the SS/Ti joint.

3.4. Diffusion mechanism of interfacial elements

Fig. 5 e Physical model formation of the brazed weld
The diffusion driving force of interfacial elements in brazed
during laser welding.
weld in this study depends on the diffusion coefficient D,
concentration gradient, crystal structure, defect, contact state

Table 3 e Chemical composition of each point.

Point Atomic content (%) Potential phases
Ti Fe Cu Cr Ni Al Ag V
A 72.08 19.64 7.48 g-Fe
B 2.1 95.63 1.02 1.15 (Cu,Fe), (Ag,Cu), a-Cu
C 86.16 8.09 2.35 3.39 Ti2Cu
D 92.12 6.93 a-Ti
6470 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2

diffusion factor D0 and activation energy Q corresponding to

Fe, Cu, Ti elements [29e31].
According to Eq. (2), the self-diffusion coefficients of Fe, Cu
and Ti at different heating temperatures can be calculated. As
shown in Fig. 7(a)e(c), it can be seen that the self-diffusion
coefficients of the three elements increase with the increase
of heating temperature in an exponential form, and the
diffusion of elements increases with the increase of temper-
ature. Comparing the diffusion coefficients of Fe (Fig. 7(a)) and
Cu (Fig. 7(b)), it can be seen that the diffusion coefficient of Cu
at the zone I is twice that of Fe when the temperature in-
creases from 1700 K. Therefore, the diffusion coefficient of Cu
atom is larger than that of Fe, and it is easier to move to the SS
side. Therefore, in the EDS map scanning of Cu element
(Fig. 4(d)), a small amount of Cu atoms were found in the re-
action layer of zone I, and the strength of the interface was
Fig. 6 e Vickers microhardness measurements of the joint. increased due to form the (Cu, Fe) solid solution. By compar-
ison, it is found that at the same heating temperature, the self-
diffusion coefficient of Ti is obviously larger than that of Fe
and other interfacial conditions of Fe, Cu, and Ti elements.
and Cu, and the difference is an order of magnitude. The
The element diffusion coefficient D depends on the diffusion
diffusion rate of Ti atom is much larger than that of Fe and Cu
activation energy Q and the heating temperature T. The
atoms. According to EDS map scanning (Fig. 4(h)) and line
diffusion activation energy Q is the minimum energy during
scanning (Fig. 4(i)), it can be seen that Ti element is not
the atomic transition in the diffusion process, which is
detected at the zone I, and the diffusion coefficient of Ti
generated to overcome the effect of the surrounding atoms on
element in the filler at high temperature is larger, which is
it. Diffusion coefficient D is the average number of particles
concentrated in the zone II.
passing through a certain plane in unit time. Diffusion coef-
ficient is a dynamic value changing with diffusion process [12].
It is very important to determine the diffusion coefficient in
this chapter. The diffusion coefficient can be used to under-
stand the concentration distribution of Fe, Cu and Ti near the
interface of the joint. Through the diffusion coefficient, the
concentration distribution of Fe and Ti elements near the joint
interface, the formation can be specifically understood. The
relationship between diffusion coefficient D, diffusion acti-
vation energy Q and temperature T satisfies the Arrehenius
exponential relationship (Eq. (2)). Table 4 shows the values of

Table 4 e Diffusion factor (D0) and activity energy (Q) of Fe,

Cu, Ti.
Element Diffusion factor D0/m2/s activity energy Q/J/mol
Fe 2.0
104 2.4
105
Cu 4.1
102 2.85
105 Fig. 8 e The curve of Cu element at interface with diffusion
Ti 3.1
105 2.0
105 distance.

Fig. 7 e Relation between self-diffusion coefficient D and temperature T of (a) Fe, (b) Cu and (c) Ti.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2 6471

Fig. 9 e (a) Tensile test curve; (b) SEM image of fracture surface; (c) XRD analysis results of fracture surface.

Due to the agglomeration of Ti element in zone II, TieCu Since the zone I is near the laser heat source, the heat con-
IMCs will be formed by reaction with Cu. Although the plas- duction is fast, and the atomic diffusion reaction is intense.
ticity of TieCu IMCs is higher than that of TieFe, it still be- Zone II contains Ti2Cu and a-Ti and AgTi2, and the concen-
comes the weak zone of the joint. Therefore, the tration of IMCs leads to the interface becoming a weak area of
concentration distribution of Cu and Ti elements in zone II the joint.
was further analyzed. The diffusion behavior of Cu and Ti Through simulation of temperature field and calculation of
elements in zone II was calculated by Eq. (5) in an ideal state, diffusion coefficient, the structure integrality of the joint was
and the variation curve of Cu atomic concentration with studied. It can be concluded that at the same heating tem-
diffusion distance was obtained (Fig. 8). It can be seen from perature, the self-diffusion coefficient of Ti is obviously one
Fig. 8 that the distribution of Cu concentration near the order of magnitude larger than that of Fe and Cu. The diffu-
interface of zone II diffusion couple tends to be level with the sion rate of Ti atom is much larger than that of Fe and Cu
increase of diffusion distance. There is a maximum diffusion atoms, this led to the aggregation of Ti element in zone II.
distance of elements, so the diffusion distance of elements
cannot be unlimitedly increased. The diffusion energy ob-
tained by elements under certain heating temperature is Funding
certain. In the concentration gradient direction, the concen-
tration distribution of elements gradually tends to be stable. This project is supported by the National Natural Science
Foundation of China (Grant No.5210040951).
3.5. Tensile tests and fracture analysis

The tensile curves are presented in Fig. 9(a), and the Declaration of competing interest
maximum tensile strength of the joint is 104 MPa. The fracture
surface at the zone II, Fig. 9(b) shows river pattern on fracture The authors declare that they have no known competing
surface that exhibiting typical brittle characteristics. This is financial interests or personal relationships that could have
due to the Ti-rich phase and TieCu IMCs in the zone II. Based appeared to influence the work reported in this paper.
on the XRD analysis results of the fracture surface (Fig. 9(c)), it
was detected that a-Ti, (Ag, Cu) and Ti2Cu phases in fracture. It
should be noted that there was no TieFe IMCs formed in joint. references
The results show that inserting the AgCuTi filler metal be-
tween SS and Ti alloy would suppress the formation of TieFe
IMCs in the joint, the joining between SS and Ti alloy was [1] Zhang Yan, Zhou JianPing, Sun DaQian, Li HongMei. Three-
realized by the diffusion of Cu element, thereby improving the pass laser welding of Ti alloy-stainless steel using Nb and Ni
quality of the joint. interlayers. J Mater Res Technol 2019;9(2):1467e73.
[2] Hosseini SRE, Feng K, Nie P, Zhang K, Huang J, Li ZG, et al.
Interlayer thickening for development of
laser-welded Ti-SS joint strength. Opt Laser Technol
4. Conclusions 2019;112:379e94.
[3] Prasanthi TN, Sudha C, Raju S, Saroja S. Interdiffusion
The joining of 304 SS to TC4 by AgCuTi filler has been suc- behavior in 304L SS/Ti-5Ta-2Nb system. J Alloys Compd
cessfully achieved. Due to the presence of unmelted filler be- 2019;808:1e13.
tween SS and Ti, the formation of brittle TieFe IMCs could be [4] Song TF, Jiang XS, Mo DF, Shen YZ, Zhu DG. A survey on
dissimilar welding of stainless steel and titanium alloy.
avoided. The unmelted SS was selected as a barrier by heat
Mater Reports 2015;29(11):81e7.
conduction to brazed weld avoids mixing Ti and SS to form [5] Chen S, Zhang M, Huang JH. Microstructure and mechanical
crack, which improved the mechanical properties of the property of laser butt welding of titanium alloy to stainless
joints. A great amount of atomic diffusion occurs at the zone I steel. Mater Des 2014;53:504e11.
and II during welding, 15 mm thick reactive layer formed in [6] Zhang Y, Zhou JP, Sun DQ, Gu XY. Nd:YAG laser welding of
zone I, which contains (Cu, Fe) solid solution, g-Fe and a-Cu. dissimilar metals of titanium alloy to stainless steel without
6472 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 6 4 6 3 e6 4 7 2

filler metal based on a hybrid connection mechanism. J [19] Mannucci A, Tomashchuk I, Bolot R, Mathieu A, Lafaye S.
Mater Res Technol 2020;9(2):1662e72. Microstructure and mechanical properties of dissimilar Ti/
[7] Zhang Y, Chen YK, Zhou JP, Xue RL, Sun DQ, Li HM. Nb/Cu/steel laser joints. Procedia CIRP
Characterization of laser beam offset welding of titanium to 2020;94:644e8.
steel with 38Zn-61Cu alloy filler. Opt Laser Technol [20] Sreh A, Kai FA, Pn A, Ke ZA, Jian H, Zlab C, et al. Interlayer
2020;127(1):106195. thickening for development of laser-welded Ti-SS joint
[8] Rezaee S, Arman A, Jurec ka S, Korpi AG, Mwema F, Luna C, strength. Opt Laser Eng 2019;112:379e94.
et al. Effect of annealing on the micromorphology and [21] Fang YJ, Jiang XS, Song TF, Mo DF, Luo ZP. Pulsed laser
corrosion properties of Ti/SS thin films. Superlattice Microst welding of Ti-6Al-4V titanium alloy to AISI 316L
2020;146:1e8. stainless steel using Cu/Nb bilayer. Mater Lett
[9] Gotawala N, Shrivastava A. Microstructural analysis and 2019;244:163e6.
mechanical behavior of SS 304 and titanium joint from [22] Sufizadeh AR, Akbari Mousavi SAA. Investigation of Nd:YAG
friction stir butt welding. Mater Sci Eng, A 2020:139658. pulsed laser dissimilar welding of AISI 4340 and AISI 316L
[10] Liu H, Fujii H. Ultralow rotation speed produces high-quality stainless steels on weld geometry and mechanical
joint in dissimilar friction welding of Ti-6Al-4V alloy and properties. Mater Des 2014;(57):128e34.
SUS316L stainless steel. Mater Sci Eng, A 2020;800:1e6. [23] Garg SP, Kale GB, Patil RV, Kundu T. Thermodynamic
[11] Kumar R, Balasubramanian M. Analysis and comparison of interdiffusion coefficient in binary systems with
diffusion bonded and friction welded Ti-6Al-4V and stainless intermediate phases. Intermetallies 1999;7(8):901e8.
steel joints with copper as interlayer. Mater Today Proc [24] Xia LF, Zhang ZX. Diffusion in metals. Harbin: Harbin
2019;21:1467e73. institute of technology press; 1989.
[12] Gao WW. Study on microstructure and properties of TA2/ [25] Wiliam F Simth. Foundations of materials science and
S316 diffusion welding joints. Shandong University; 2020. engineering. third ed. New York: McGRAW HILL; 2003.
[13] Vigraman T. Liquid-solid phase reaction products formation [26] Zhang L, Lu JZ, Luo KY, Feng AX, Dai FZ, Zhong JS, et al.
in the diffusion welded joints made between Ti-6Al-4 V and Residual stress, micro-hardness and tensile properties of
AISI 304L with brass interlayer. Mater Today Proc ANSI 304 stainless steel thick sheet by fiber laser welding.
2021;42:607e17. Mater Sci Eng, A 2013;561:136e44.
[14] Norouzi E, Morteza A, Masoud S. Effect of bonding time on [27] Gao Ming, Chen Cong, Wang Lei, Wang Zemin, Zeng Xiaoyan.
the joint properties of transient liquid phase bonding Laser-Arc hybrid welding of dissimilar titanium alloy and
between Ti-6Al-4V and AISI 304. J Alloys Compd stainless steel using copper wire. Metall Mater Trans
2017;701:335e41. 2015;46:2007e20.
[15] Zhou K, Li J, Wang L, Yang H, Wang Z, Wang J. Direct laser [28] Oliveira JP, Panton B, Zeng Z, Andrei CM, Zhou Y,
deposited bulk CoCrFeNiNbx high entropy alloys. Miranda RM, et al. Laser joining of NiTi to Ti6Al4V using a
Intermetallics 2019;114:1e6. Niobium interlayer. Acta Mater 2016;105:9e15.
[16] Cao XJ, Kabir ASH, Wanjara P, Gholipour J, Birur A, Cuddy J. [29] Du Y, Chang YA, Huang BY, Gong WP, Jin ZP, Xu HH, et al.
Global and local mechanical properties of autogenously laser Diffusion coefficients of some solutes in fcc and liquid Al:
welded Ti-6Al-4V. Metall Mater Trans A 2014;45A:1258e72. critical evaluation and correlation. Mater Sci Eng, A
[17] Chen Z, Wang B, Duan B. Mechanical properties and 2003;363:1e2.
microstructure of a high-power laser-welded Ti-6Al-4V [30] Wang QX, Wang JL. Study on the interdiffusion of W-Ti alloy
titanium alloy. J Mater Eng Perform 2020;29:2296e304. and b phase stability. Powder Metall Technol 2018;36:3e8.
[18] Chen HC, Bi G, Bing YL, Cheng CK. Laser welding of CP Ti to [31] Cao F. Diffusion behavior and microstructural evolution on
stainless steel with different temporal pulse shapes. J Mater interfaces in Al/Cu bimetal. ME: Dalian University of
Process Technol 2016. Technology; 2018.