1 s2.0 S2666330920300200 Main

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Journal of Advanced Joining Processes 1 (2020) 100022

Contents lists available at ScienceDirect

Journal of Advanced Joining Processes


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

Use of pure vanadium and niobium/copper inserts for laser welding of


titanium to stainless steel
A. Mannucci a,b, I. Tomashchuk a,∗, A. Mathieu a, R. Bolot a, E. Cicala a, S. Lafaye b, C. Roudeix b
a
Laboratoire Interdisciplinaire Carnot de Bourgogne, UMR CNRS-6303, Université de Bourgogne-Franche Comté, 12 rue de la Fonderie, 71200 Le Creusot, France
b
SME Laser Rhône-Alpes, 49-51 Boulevard Paul Langevin, 38600 Fontaine, France

a r t i c l e i n f o a b s t r a c t

Keywords: Niobium and vanadium have high metallurgical compatibility with titanium and therefore can be used as inserts
Laser welding to avoid the accumulation of brittle intermetallic phases such as Fe2 Ti during the fusion welding of titanium alloys
Titanium with steels. In the present study, the continuous double pass welding of 1 mm thick Ti-6Al-4V alloy and 316 L
Stainless steel
stainless steel plates through several mm wide pure vanadium or niobium insert was studied. In case of a vanadium
Vanadium
insert, a beam offset on the vanadium was found to produce cold crack formation in vanadium/316 L melted
Niobium
Copper zones containing more than 40 wt.% V despite the absence of 𝜎 phase. Whereas a centered beam position and
offset on the steel side produced crack-free melted zones still composed of (Fe, V) solid solutions, and exhibiting
a ductile fracture in the unmelted vanadium insert at UTS of 493 ± 25 MPa. The niobium insert produced brittle
niobium/316 L melted zones with UTS of 160 ± 10 MPa, because of the formation of Fe2 Nb and Fe7 Nb6 layers
at the niobium/316 L interface. The addition of a copper insert between niobium and steel allowed avoiding the
development of Fe-Nb brittle intermetallics thanks to the absence of brittle phases in Cu-Nb system. For optimal
welding conditions, a ductile fracture occurred in the copper interlayer at maximal UTS of 255 ± 10 MPa.

Introduction The use of insert material as a barrier between titanium and steel
represents an interesting approach for both fusion and solid state join-
Titanium and its alloys have a high corrosion resistance, biocompat- ing methods. Very few metals are compatible simultaneously with tita-
ibility and an important specific strength. These properties are useful nium and iron. Only eight commercially available metals form contin-
for chemical, medical and aeronautic industries respectively, but tita- uous solid solutions with titanium: vanadium, niobium, molybdenum,
nium is two to six times more expensive than stainless steel. The cre- tantalum, tungsten, zirconium, hafnium and magnesium (Tomashchuk
ation of sound joints between these materials opens the perspective of and Sallamand, 2018).
an important cost reduction. Laser welding allows assembling many dis- However, among these metals, only vanadium combines a high melt-
similar metal combinations thanks to important thermal gradients and ing point with a good mutual solubility in stainless steel. A brittle
a high geometrical precision (Sun and Ion, 1995). Nevertheless, for tita- 𝜎 phase (FeV) exists in Fe-V system between ~20 and 80 at.% of V
nium/316 L couple it remains challenging due to an important mismatch (Lyakishev, 1997), but it has a low kinetic of formation and therefore
of thermal expansion coefficients (Table 1) and the formation of brittle it can be avoided in laser welding. Ustinovshchikov et al., 2005 deter-
intermetallic compounds (FeTi, Fe2 Ti, Cr2 Ti, Cr7 Fe17 Ti5 ) (Ivanchenko mined that 𝜎 phase is not stable below 650 °C, and below 600 °C a
and Pryadko, 2008). By the application of a sufficiently large laser beam phase separation of (Fe, V) solid solution takes place. Tomashchuk et al.,
offset from the joint line, the precipitation of these harmful phases can 2015 studied Ti-6Al-4V/vanadium/316 L laser joining, and compared
be reduced to the micrometric layers situated at the limit of the melted the effect of welding in one pass, two independent passes and double
zone (Mannucci et al., 2018a). Nevertheless, the tensile behavior of such spot. Two-pass welding was accepted as the best solution resulting in
joints remains brittle, and the resulting average ultimate tensile strength ductile fracture inside the unmelted vanadium insert at 367 MPa. A dou-
(UTS) is quite low (150 MPa). Solid-state methods such as explosion ble spot caused an important annealing of the vanadium and the melted
welding (Gloc et al., 2016), friction (Kimura et al., 2016) or friction stir zones with a composition near that of 𝜎 phase exhibited a partially brit-
welding (Ishida et al., 2015) can produce resistant joints by reduction of tle fracture, although the 𝜎 phase was not detected by X-ray diffraction
intermetallics zones to a submicron level. However, these methods are (XRD). One pass welding with a laser spot centered at the middle of the
not applicable for all joining geometries. insert did not allow avoiding the embrittlement of the melted zone by Fe-


Corresponding author.
E-mail address: [email protected] (I. Tomashchuk).

https://doi.org/10.1016/j.jajp.2020.100022
Received 15 November 2019; Received in revised form 3 April 2020; Accepted 3 April 2020
2666-3309/© 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license.
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Table 1
Physical properties of welded materials.

Properties Units 316 L T40 Ti-6Al-4V Vanadium Niobium Copper

Melting point/solidus liquidus K 1663-1713 1941 1877-1933 2163 2741 1358


Boiling point K 3013 3560 3560 3653 5015 2840
Thermal diffusivity mm2 /s 3.8 9.4 2.9 10 23.4 120
Thermal expansion 10−6 K−1 16.5 8.4 8.6 8.3 7.2 17
Density kg/m3 7980 4510 4430 6100 8750 8960
Ultimate tensile strength MPa 590 345 1000 500 290 250
Yield strength MPa 300 276 975 420 190 60

Ti intermetallics. A similar result was also obtained for one pass electron beam on steel, but it cannot be completely avoided (Mannucci et al.,
beam welding by Adomako et al., 2018, who compared electron beam 2019c). Zhang et al., 2016 proposed another method of laser joining
welding, pulsed laser and continuous laser welding. In the two pass joint of titanium to steel through the niobium interlayer that consists in cre-
obtained by pulsed laser (Adomako et al., 2018), vanadium/steel melted ating a liquid diffusion zone between solid niobium and stainless steel
zone contained cracks, possibly because of a shorter solidification time using the heat flow from nearby titanium/niobium melted zone. Such
and higher strain rates. In continuous two pass laser welding (Adomako interdiffusion interface had the same composition as the intermetallic-
et al., 2018), the vanadium/steel joints with less than 35 at.% V resisted rich layers observed by other authors, but it fractured at a rather high
during the tensile test, and the ductile fracture occurred in the vanadium UTS of 370 MPa. Hot rolling cladding allows further reduction of this
insert at maximal UTS of 375 MPa, when those with more than 35 at.% V intermetallic layer by the application of a sufficiently mild temperature
suffered from embrittlement. The hardness of vanadium/stainless steel cycle, which results in attaining a ductile tensile behavior and a high
melted zones was increasing along with the vanadium content. A sim- strength compared to the weakest base material of bimetallic assembly
ilar result was reported by Nogami et al., 2013 who also studied the (Zhao et al., 2014; Xie et al., 2018).
effect of post weld heat treatment (PWHT) on electron beam welded The use of materials having a good laser weldability with stainless
vanadium/steel joints. The heat treatment of V-rich joints produced an steel, such as copper, nickel, silver etc. as intermediate inserts is limited
important increase in hardness from < 500 HV to > 1000 HV, attributed by the brittle nature of intermetallics formed with titanium. However,
to the formation of 𝜎-(FeV), Ni2 V3 and NiV3 precipitates. The effect of the creation of a multimaterial insert having full compatibility between
laser beam offset from the joint line and PWHT on the microstructures neighboring metals allows attaining good synergetic results compared
of vanadium/steel weld was furtherly studied by Adomako et al., 2019. to standalone materials. The use of 30 μm titanium, 100 μm vanadium
Again, ductile fracture in solid vanadium occurred only in cases with a and 30 μm copper foils allowed obtaining crack free TiAl/stainless steel
low vanadium dilution. It was observed that the welds with an average diffusion bonds with UTS close to that of TiAl alloy (He et al., 2003).
composition <18 at.% V presented a columnar grain structure perpen- The diffusion bonding of Ti-6Al-4V alloy to 316 L stainless steel through
dicular to the limits of the melted zone, when a coarse grain structure 25 μm niobium and 20 μm copper foils allowed attaining remarkably
was observed in the welds containing >28 at.% V. Intermediate V con- high UTS values associated to a ductile fracture, thanks to the absence
centrations resulted in coexistence of both microstructures, with colum- of brittle phases in the Cu-Nb system (Song et al., 2017). Gao et al.,
nar grains situated next to the stainless steel. The EBSD analysis allowed 2019 used a 0.3 mm thick niobium interlayer for the laser welding of
detecting the tiny grains of the 𝜎 phase in the central part of a V-rich Ti-6Al-4V alloy to copper. The beam offset on Ti-6Al-4V side formed a
melted zone, where they acted as the initiators of intergranular cracks. melted zone containing (𝛼+𝛽 Ti,Nb) isolated from the copper side by
Moreover, the melted zones with > 18 at.% V showed a hardening af- the unmelted niobium layer. The joint between the remaining niobium
ter PWHT similar to that reported by Nogami et al., 2013. Mannucci and copper was created by interdiffusion. During the tensile test, the
et al., 2019a studied the effects of the laser spot diameter, the beam ductile fracture occurred in this diffuse interface at UTS close to that of
offset from the joint line and lineic energy on the properties of Ti-6Al- annealed copper. Mannucci et al., 2018b showed the possibility to form
4V/vanadium/316 L welds. It was found that the use of a smaller laser defect-free copper/steel laser joints within a large window of opera-
spot allows reducing the undesirable annealing in the solid vanadium tional parameters. The combination of consecutive titanium/niobium,
HAZ and in this way attaining the UTS of the raw vanadium. The ef- niobium/copper and copper/steel laser welds would allow completely
fect of vanadium content on grain morphology, reported by Adomako avoiding the embrittlement by intermetallic phases. This makes nio-
et al., 2019 was confirmed. Zhang et al., 2018 reported successful Ti- bium/copper combination attractive despite a relatively low UTS of cop-
6Al-4V/vanadium/301 L joining by a pulsed laser: a UTS of 599 MPa per.
was reached by offsetting the laser spot on the vanadium insert for the The present study is an investigation of the microstructure and me-
Ti-6Al-4V/vanadium joint, and on steel for the vanadium/301 L joint. chanical properties of laser welded joints between titanium alloys and
Among titanium-compatible metals, niobium has a thermal expan- 316 L stainless steel using different kinds of inserts. In the first part,
sion coefficient close to that of titanium and a reasonable mismatch the use of a vanadium interlayer is discussed and the detailed inves-
in melting and boiling points (Table 1). Moreover, it is biocompati- tigation of previously observed composition-dependent solidification
ble and less expensive that vanadium. Compared to vanadium, nio- microstructures is performed with EBSD analysis. The explanation of
bium has a much lower solubility in stainless steel and forms inter- changes in microstructure depending on a melted zone composition is
metallic compounds Fe2 Nb and Fe7 Nb6 (Lyakishev, 1997). Hajitabar proposed, based on the evolution of the solidification front profile. As
and Naffakh-Moosavy, 2017 reported a repetitive crack formation in one-interlayer alternative to vanadium, standalone niobium produced
full penetrated niobium/steel joints obtained by electron beam weld- brittle interfaces with statistical microcracks. Multi-insert joining with
ing, despite of testing the different beam offsets from the joint line. An ductile behavior may be a good solution at lower cost. Niobium/copper
intermetallic layer was formed along the interface between the solid inserts have a good metallurgical compatibility with titanium and steel
niobium and the melted zone even under a low niobium dilution. Micro respectively, are less expensive than vanadium, and have not been the
cracks present in this interface acted as fracture initiation sites during object of many studies. Therefore, the use of this combination with two
the tensile test (Hajitabar and Naffakh–Moosavy, 2018). The thickness different thicknesses of copper insert is compared with a standalone nio-
of this intermetallic-rich layer can be reduced by offsetting the laser bium barrier in the second part of the study.
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Table 2
Average composition of welded materials according to the material certifications of providers (wt.%).

Ti-6Al-4V Ti Al V Fe O C N H
remainder 6.21 3.99 0.13 0.13 0.025 0.003 0.003

T40 Ti Fe O C N H Total impurities


remainder 0.042 0.081 0.01 0.008 0.003 < 0.4

316 L Fe Cr Ni Mo Mn C N Si P S
remainder 16.8 10.1 2.03 0,90 0.023 0.040 0.49 0.035 < 0.001

10 J/mm and three different beam offsets from the joint line: 50 μm
offset on 316 L, centered beam position and 50 μm offset on vanadium
(Fig. 2a).

Welding with niobium and niobium/copper inserts


In all experiments, the 1 mm wide insert of 99.9% pure annealed
niobium was joined to T40 commercially pure titanium plate in the
first pass, with 100 μm beam offset on titanium and a lineic energy
of 20 J/mm. During the second pass performed with lineic energy of
20 J/mm (Fig. 2b)
• the edge of the standalone niobium insert was welded to 316 L plate
with 100 μm offset on 316 L;
• the edge of the niobium insert with a fixed 0.2 mm thick and 1 mm
high sheared copper foil was welded to a 316 L plate with the laser
beam centered at the interface between copper foil and steel.

In the case of the 1 mm thick copper insert (Fig. 2c), during the
second pass, this insert was welded to the 316 L plate with zero offset
from the joint line and lineic energy of 24 J/mm. Finally, during the
third pass, two bimetallic assemblies were joined together with a lineic
energy of 30 J/mm and zero offset from the joint line.

Characterization

For tensile tests and fractography studies, the joints were cut in
Fig. 1. Welding configuration and clamping system.
20 mm large strips. Tensile tests (three per welding condition) were
made with a MTS Insight 30 kN tensile test machine, at a cross head
speed of 5 mm/min. Composition analysis and microhardness measure-
Materials and methods ments were performed on the weld cross sections embedded in conduc-
tive resin, grinded and polished down to colloidal silica. Microstruc-
Welding procedure ture and fractography observations were made using a scanning elec-
tronic microscope JSM-6610LA (Jeol), equipped with an energy disper-
Laser welding was performed with a continuous Yb:YAG laser source sive spectroscopy (EDS) analyzer operating at 20 kV. The X-ray diffrac-
TruDisk 6001 (Trumpf) of 6 kW maximal power, 1030 nm wavelength tion (XRD) analysis was made by PANalytical X’Pert PRO, using a Co
and beam quality of 4 mm/mrad. Collimator and focalization lenses had target, with scan steps of 0.0167°, a scanning range of 30–120° and a
focus distances of 200 mm, resulting in a magnification factor of 1. A counting time of 200 s per step. The electron backscatter diffraction
Gaussian laser spot with 100 μm in diameter was focused on the top sur- (EBSD) maps were performed at an acceleration voltage of 15 kV and
face of the plates joined in butt configuration. Constant values of lineic analysis step of 0.5 μm. Vickers microhardness tests were made with a
energies and beam offsets from the joint line were defined, based on the Buehler device, with a 25 g load applied during 10 s.
preliminary studies (Mannucci et al., 2019a). Base metal plates (Table 2)
and inserts were 150 mm long and 1 mm thick. The inserts were laser Results and discussion
cut. Before the welding, all inserts and 50 mm wide plates were polished
on the edges and cleaned with acetone. The plates were pre-assembled Use of vanadium insert
with five laser pulses per dissimilar joint to maintain them during the
continuous welding, complementary to the clamping device (Fig. 1), Global observations
where the plates to weld were clamped under the holding plates. The Two pass welding resulted in Ti-6Al-4V/vanadium (MZTi ) and
top and the bottom faces of the welds were protected by argon flow 316 L/vanadium (MZ316L ) melted zones isolated by an unmelted vana-
with 20 and 10 l/min respectively. The joint line was positioned upon dium layer (Fig. 3). Both melted zones had a shape typical for laser
a notch, were argon circulated for the bottom protection. welding, with an enlargement on the top and the bottom, and did not
contain major macro segregations. Because of the higher thermal dif-
Welding with vanadium insert fusivity and melting point of vanadium compared to other materials
In the first pass, Ti-6Al-4V plates were welded to 2 mm wide inserts (Table 2), it was melted to a lesser degree, which resulted in a sharper
of 99.8% pure laminated vanadium with a lineic energy of 6.8 J/mm and interface at the limit of the MZ. Thanks to the ideal metallurgical com-
zero offset from the joint line. In the second pass, the vanadium inserts patibility between titanium and vanadium, the use of centered beam po-
were welded to 316 L austenitic stainless steel with a lineic energy of sition resulted in a crack-free Ti-6Al-4V/vanadium weld with 41 at.% V
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 2. Scheme of welding experiments with different inserts: two-pass welding with (a) vanadium insert, (b) niobium and niobium +0.2 mm copper insert; (c)
three-pass welding with niobium and 1 mm copper insert. Laser paths are represented by dashed red lines. The pass numbers are indicated at the side of the joints.
When no beam offset was indicated, the laser spot was focused in the joint line. A Δ sign represents the different beam offsets on 316 L.

Fig. 3. Weld cross sections of Ti-6Al-4V/vanadium/316 L joints: (a) MZTi welded with 6.8 J/mm, no offset from the joint line; MZ316L welded with 10 J/mm and
(b) 50 μm offset on 316 L side, (c) no offset from the joint line, (d) 50 μm offset on vanadium side.

with heat affected zones (HAZ) formed on both Ti-6Al-4V and vanadium content results in more long-living melts past laser keyhole. This has a
sides (Fig. 3a). Consequently, the welding conditions for MZTi were kept direct consequence on the microstructure of the welds, as discussed in
constant. According to Nogami et al., 2013, the increase of V amount in the next section.
vanadium/316 L welds produces important embrittlement and therefore
compromises their mechanical properties. Three chosen beam positions Microstructures of the welds
resulted in MZ316L welds with an increasing amount of V comprised be- The melted zone formed between vanadium and Ti-6Al-4V alloy
tween 11 and 47 at.% (Fig. 3b–d). The last V-rich weld suffered from (MZTi ) contained equiaxed grains composed by cubic (Ti, V) solid so-
cracks situated at the nail-head of the melted zone. The convection in lution. The grains were coarser (up to 130 μm) at the center of the weld
the melted zone was insufficient to homogenize totally the local com- (Fig. 6a–b), which allows supposing an excessive recrystallization of the
position of V-rich welds (Fig. 4). No reactive layer was present at the solidified weld. The neighboring unmelted vanadium formed ~100 μm
limits of the melted zones. thick HAZ where the initial laminated microstructure (on the left of
In case of 50 μm offsets of a 100 μm laser spot (Fig. 4a and c), a key- Fig. 6a) was replaced by equiaxed recrystallized grains with diameter
hole forms preferentially in a single material, which vaporization tem- up to 35 μm. The unmelted Ti-6Al-4V alloy formed ~300 μm thick HAZ
perature defines the maximal temperature of the melted zone. The beam containing dominating martensite-like needles of 𝛼-Ti and punctual
offset on vanadium resulted in hotter and larger melted zone than offset nodular inclusions of 𝛽-Ti (Fig. 6c–d).
on stainless steel. On the other hand, in case of zero offset (Fig. 4b), The melted zone MZ316L welded with a 50 μm offset of the laser
a keyhole is shared between two materials. Consequently, the centered spot from the joint line on 316 L contained only 11 at.% V and was
beam position resulted in an intermediate MZ316L width. composed exclusively by a cubic (Fe, V) solid solution. A columnar so-
The top observation of MZ316L welds obtained with different beam lidification perpendicular to the walls of the melted zone, commonly
offsets revealed tear-like solidification front in the steel-rich weld occurring in the tear-like liquid zones, took place in the totality of the
(Fig. 5a), which promotes a columnar solidification perpendicular to weld (Fig. 7a). The microstructure was coarsening towards the center
the joint line. On the other hand, the enrichment of the melted zone in of the weld, with maximal grain size up to 130 μm. Neighboring vana-
vanadium (Fig. 5b–c) produced large and elongated elliptic shapes. Tak- dium formed a ~100 μm thick zone of 60–70 μm equiaxed grains. No
ing into account the identical lineic energy used for these three welds, it important recrystallization was observed in solid 316 L. In the melted
can be concluded that the beam offset on steel induces shorter lifetime zone MZ316L obtained with zero beam offset from the joint line and
of the melt than offset on the vanadium side. The increase of vanadium containing 30 at.% V, two kinds of grains were observed (Fig. 7b), in
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 4. EDS-maps of MZ316L welds obtained with beam


offsets of (a) 50 μm on 316 L side, (b) no offset from the
joint line, (c) 50 μm on vanadium side.

Fig. 5. Top view of the MZ316L welds with vanadium insert (a) 50 μm offset on 316 L, (b) centered beam position, (c) 50 μm offset on vanadium.

agreement with results of Adomako et al., 2019. The columnar grains Finally, the melted zone MZ316L obtained with a 50 μm beam offset
were oriented perpendicularly to the melted zone limit on the 316 side on vanadium and containing 47 at.% V was composed of very coarse (up
and had a maximal diameter of 50 μm. Coarse equiaxed grains with a to 170 μm) equiaxed grains (Fig. 7c). Only a ~10 μm large belt of thin
maximal diameter of 110 μm filled the rest of the weld. However, this grains (~10 μm) remained at the 316 L limit. The whole melted zone
microstructure evolution was not related to the local change of chemi- contained only a cubic (Fe, V) solid solution. This result was confirmed
cal composition: the whole melted zone was composed by an identical by the XRD analysis. The area of coarse grains with a dominating 111
cubic (Fe, V) solid solution. The HAZ in vanadium became twice larger orientation situated at the nail head of the weld contained intergranular
(~200 μm) compared to the previous case (Fig. 7a), but conserved the cracks (Fig. 7d). For the similar V-rich weld, Adomako et al., 2019 ob-
same grain size of 60–70 μm. These recrystallized HAZ grains continued served a small inclusion of 𝜎-(FeV) phase inside of a crack at the center
to grow into the melted zone. of the weld, detectable by the EBSD analysis. However, for the present
Previous observations reported in Mannucci et al., 2019a showed welding condition, no 𝜎 precipitates were found in the cracks. This dif-
that the columnar solidification remains predominant for steel- ference can be attributed to the higher cooling rate that impedes the
vanadium welds containing less than 10 at.% V and is totally replaced by precipitation of the 𝜎 phase. However, the presence of cracks still oc-
coarse equiaxed cells when the average V amount exceeds 40 at.%. Such curred between the coarse grains formed close to the welding axis. Sev-
change in microstructure can be attributed to the evolution of the so- eral neighboring grains present some distortions that can be attributed
lidification front at the rear of the melted zone from the short tear-like to the increased residual stress in the nail-head of the weld, which is in
for steel-rich welds (Fig. 5a) to the elongated elliptic in V-rich welds agreement with the numerical simulations of residual stress, described
(Fig. 5c) and a possible additional recrystallization of V-rich melted in Mannucci et al., 2019b.
zones.
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 6. EBSD analysis of Ti-6Al-4V/vanadium melted zone: band contrast maps (a-c) and inversed pole IPF Y image with phase mapping in HAZTi (b-d). (c-d) are a
detail of HAZ of Ti-6Al-4V, from the dotted rectangle in (a-b).

Fig. 7. EBSD inversed pole IPF Y images of 316 L/vanadium melted zones: (a) 50 μm beam offset on 316 L, (b) zero beam offset, (c–d) 50 μm beam offset on
vanadium (vanadium was not indexed), (d) detail of micro cracks situated at the top of the melted zone.

Mechanical properties of the welds MZ316L was increasing proportionally to the dilution of vanadium, in
The hardness measurements were made approximately at the middle accordance with the results described in Mannucci et al., 2019a.
height of the samples (Fig. 8). The hardness of the base materials was of The V-rich weld obtained with the 50 μm beam offset on vanadium
170 HV for laminated vanadium, 340 HV for Ti-6Al-4V and 200 HV for had a maximal hardness of 550 HV. However, according to Nogami
316 L. The heat affected zone of Ti-6Al-4V had an increased hardness et al., 2013 and Adomako et al., 2019, the precipitation of the 𝜎 phase,
up to 400 HV, due to the martensitic transformation. The remaining for example after a post weld heat treatment, should increase the
vanadium insert showed similar annealed zones near the limits of the hardness of the melted zone up to 1000 HV. Ustinovshchikov et al.,
welds, with a hardness decreasing down to 100 HV. The MZTi had a 2005 observed that under 600 °C, a nanoscale phase separation in a (Fe,
microhardness close to that of the initial Ti-6Al-4V and the hardness of V) solid solution produces the hardening of the alloy. In the conditions
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 8. Microhardness profiles across the Ti-


6Al-4V/vanadium/316 L joints.

Fig. 9. (a) Tensile curves of Ti-6Al-4V/vanadium/316 L joints compared with raw laminated vanadium and (b) the fractography of broken vanadium insert.

of thermal strain, it can be supposed that such hardening alone creates Titanium and stainless steel have low thermal conductivity and develop
local cold cracks in absence of a 𝜎 phase. For the weld with zero beam important convection movements due to Marangoni effect, which results
offset from the joint line, the hardness of columnar grains and equiaxed in the enlargement of the top and bottom of the welds.
grains having the same chemical composition was comprised within a The welding condition for joining titanium with a niobium interlayer
range of 450–480 HV. Finally, the hardness of the MZ316L obtained with was kept constant in all experiments, and the resulting melted zone al-
a 50 μm offset on the steel side of only 310–330 HV. It is noticeable ways remained isolated from the following welds by the unmelted inter-
that in all cases the maximal hardness was measured near the center of layer. A 100 μm beam offset on the titanium side was applied to avoid
the weld, where a lower cooling rate allows more efficient solid state the beam loss in the gap between the joined plates. A more important
transformations. lineic energy was chosen compared to titanium/vanadium weld in or-
The tensile tests were performed for two defect-free conditions: with der to conserve full penetration, since niobium has a higher melting
MZ316Ls welded by a centered beam position and a 50 μm offset on 316 L. point and thermal diffusivity (Table 1). In this condition, the niobium
The samples demonstrated a similar ductile behavior (Fig. 9), having plate was not directly melted by the laser beam, but underwent a strong
average UTS values of 487 ± 7 MPa and 493 ± 25 MPa respectively. brazing with Ti-rich liquid, which resulted in a global Nb amount of
Compared with the average UTS of laminated vanadium, estimated to about 15 at%. Besides several Nb-rich isles situated at the top of the
502 ± 35 MPa, a joint coefficient close to 100 % was reached. The joints titanium/niobium weld, it was globally homogenous (Fig. 10a). No de-
demonstrated a higher maximal displacement compared to raw vana- fects where detected in titanium/niobium melted zones (MZTi/Nb ), ex-
dium, which can be contributed to the presence of large ductile grains cept some minor porosities.
in the HAZ of vanadium as well as in the melted zones. However, in both The preliminary studies (Mannucci et al., 2019c) reported a severe
cases, the fracture occurred in the middle of the unmelted vanadium in- embrittlement in niobium/316 L melted zones that contained more
sert, where a laminated structure was conserved. Therefore, it can be than 10 at.% Nb. For this reason, in the present experiment, to pro-
concluded that the use of 100 μm laser spot does not create weakened duce MZNb/ 316 L , the laser beam was shifted at 100 μm on 316 L side
HAZ in vanadium. (Fig. 10b). The resulting weld was strongly asymmetrical and contained
only 4 at.% Nb.
Use of niobium and niobium/copper inserts In the next experiment, the 0.2 mm wide copper foil added between
niobium and steel was completely melted after a single pass (Fig. 10c),
Global observations in spite of its additional height that formed nail head in the bottom of
Niobium has a high thermal conductivity and a higher melting point the joint. The used foil was too thin to induce a noticeable heat draining
compared to other used metals. Consequently, a low degree of dilution or decrease of the global quantity of the adsorbed laser radiation. The
is attained and the limits of Nb-containing welds are almost vertical. miscibility gap present in the Cu-Fe system (Cao et al., 2011) produced
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 10. General view of the melted zone cross cuts: (a) MZTi/Nb , (b) MZNb/ 316 L , (c) niobium/0.2 mm copper/316 L assembly and (d) niobium/1 mm copper/316 L
assembly. The squares accompanied by letters represent the positions of the corresponding micrographs shown on Fig. 11.

the macroscopic segregations between the melted copper and steel. The Nb(Cr, Ni)2 . The melt matrix (zone 3) with ~4 at.% Nb, contained 𝛼-Fe
melted zone MZCu was dominated by melted copper with some minor cells surrounded by previous eutectics. From this observation, it can be
Fe-rich inclusions. A fragmented steel-rich domain formed next to the concluded that reactive IML is the most critical zone of niobium/316 L
316 L side at the middle-height of the plate. welds, and its thickness should be reduced down to several μm to avoid
The use of 1 mm wide copper insert allowed creating a sound bar- a crack formation. However, a further increase of beam offset on 316 L
rier between the niobium/copper melted zone (MZNb/Cu ) and the cop- would induce a lack of fusion at the bottom of the weld. On the other
per/stainless steel melted zone (MZCu/ 316 L ) that completely avoided hand, the laser offset on niobium side induces a spontaneous fracture,
niobium and stainless steel elements from mixing (Fig. 10d). Conse- in agreement with the results of Hajitabar and Naffakh-Moosavy, 2017.
quently, no Nb–Fe intermetallics formation took place. The high thermal In the case of a 0.2 mm copper insert, the Cu-rich matrix of MZCu
conductivity of copper led to almost vertical interfaces with both melted (Fig. 11c, d, e) contained only 5 at.% Fe and <0.1 at.% Nb. The demixed
zones (Fig. 10d). Because of a high thermal diffusivity and reflectivity of Fe-rich droplets were composed by ~15 at.% Cu, 60 at.% Fe and variable
copper to the Yb:YAG laser radiation, MZCu/ 316 L contained only 17 at.% amount of Nb, up to 2.5 at.%, in accordance with the Cu-Fe-Nb phase
Cu. However, MZNb/Cu contained only 4 at.% Nb in average due to the diagram (Wang et al., 2000). Small Fe-rich droplets contained more Nb
high melting point of niobium compared to copper. than the bigger ones. In bubbles bigger than 15 μm, a secondary phase
separation took place (Fig. 11e), which indicates an undercooling of
Microstructures of the welds and EDS analysis ~200 K (Chen et al., 2007). Besides the demixed copper droplets, the
In the MZTi/Nb , the Nb-rich isles (Fig. 11a) were totally melted and steel-rich zone contained 7.5 at.% Cu.
contained up to 29 at.% Nb, which corresponds to a body centered cubic The IML formed in the niobium/MZCu interface was very heteroge-
BCC (𝛽-Ti,Nb) solid solution. The surrounding melted zone containing neous and did not develop the microstructure previously observed in
~6 at.% Nb consisted of thin mix of hexagonal close packed (HCP) and the Cu-free niobium/316 L IML. The interface width varied from 3 μm
BCC (𝛼+𝛽 Ti,Nb) microstructures. to 50 μm. The largest zones were produced by the collision of Fe-rich
A reactive intermetallic layer (IML) formed at the niobium interface droplets with IML and contained random cracks (Fig. 11d). The thinnest
of MZNb/ 316 L was ~25 μm thick at the top of the weld (Fig. 11b) and homogenous regions of IML (Fig. 11c) were composed by 8 at.%
and ten times thinner at the bottom. Except for the thinnest part, Cr, 35 at.% Fe, 5 at.% Ni, 17.5 at.% Cu and 35 at.% Nb, which cor-
it contained many cold microcracks that did not propagate in solid responds to (Fe, Ni, Cu)7 Nb6 , small amount (Fe, Cr, Ni)2 Nb and (Cr,
niobium or in the melted zone. The thinnest part of the reactive layer Ni)2 Nb phases. In some places, the Fe-rich droplets collided with the IML
had a uniform structure and contained up to ~ 40 at.% Nb, which and formed a new Fe-rich layer with a thin dendritic structure (Fig. 11d),
corresponds to Fe7 Nb6 phase (Raghavan, 2004). In the larger zones, this composed by 15 at.% Cr, 60 at.% Fe, 7.68 at.% Ni, 8 at.% Cu and 8.35
Fe7 Nb6 layer developed a dendritic structure (zone 1 on Fig. 11b). The at.% Nb, which corresponds to the eutectics 𝛼-Fe + (Fe, Cr, Ni)2 Nb.
next segregated layer (zone 2) was composed by 64 at.% Fe and 9 at.% Fig. 11e shows the last stage in the development of the interface, when
Nb, which corresponds to Fe2 Nb/Fe eutectic, with some addition of Fe-rich droplets were almost completely dissolved (except the zone 1).
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 11. Microstructures of Nb-containing


welds: (a) Nb-rich islet in MZTi/Nb ; (b)
niobium/MZNb/ 316 L intermetallic-rich
layer, where 1 - Fe7 Nb6 layer; 2 -
Fe2 Nb + 𝛼-Fe eutectics and 3 - 𝛼-Fe cells
surrounded by previous eutectic; (c-d-e)
different zones of niobium/MZCu inter-
face of niobium/0.2 mm copper/316 L
assembly; (f) niobium/MZNb/Cu interface
and (g) copper/MZCu/ 316 L interface of
niobium/1 mm copper/316 L assembly.

The rest of the interface contained thin inclusions 2 composed of (Fe, In the case of the 1 mm thick copper insert, MZNb/Cu contained
Cr, Ni)2 Nb (8.1 at.% Cr, 4.6 at.% Ni, 34 at.% Fe, 14 at.% Cu), in the ~4 at.% Nb. The very limited solubility of Nb in Cu and the high
bright matrix 3 composed of Fe7 Nb6 , where Fe is partially substituted difference of the melting temperatures between the two metals pro-
by Ni, Cr (Takeyama et al., 2001) and probably Cu (6 at.% Cr, 6.1 at.% duced tiny Nb-rich BCC precipitates and thin dendrites (Fig. 11f). The
Ni, 29 at.% Fe, 14 at.% Cu, 45 at.% Nb). Rare inclusions 4 were formed face centered cubic (FCC) Cu matrix contained around 0.2 at.% Nb.
by the segregated Cu droplets. MZCu/ 316 L contained 17 at% Cu, which segregated in form of micro
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 12. Microhardness profiles across the (a) MZTi/Nb , (b) MZNb/ 316 L , (c) niobium/0.2 mm copper/316 L and (d) niobium/1 mm copper/316 L assemblies.

droplets between 𝛾-Fe columnar microstructures (Fig. 11g). The solid- During all tensile tests, MZTi/Nb was never damaged, and the fracture
ification was oriented perpendicularly to the limits of the melted zone. occurred in niobium/316 L part of the joints, with different mechanisms
Nb-Fe intermetallics were not observed thanks to the unmelted copper for each considered case. The titanium/niobium/316 L dissimilar joint
layer. resisted up to 160 ± 10 MPa (Fig. 13e, blue continuous line), with a
brittle fracture in IML. XRD analysis of the fracture (Fig. 13a) on the
Mechanical properties of the welds melted zone side showed the presence of 𝛼-Fe and high amount of Fe2 Nb
The hardness measurements across the niobium/titanium weld did along with small quantities of 𝛾-Fe and metastable 𝜀-Fe. The XRD of the
not reveal any embrittlement (Fig. 12a). Due to the high Ti content, fracture on the niobium side along the IML showed an important peak
average Vickers hardness of the melted zone (227 ± 20 HV) neared that of Fe7 Nb6 and low intensity peaks of Fe2 Nb and 𝛾-Fe. Some small peaks
of raw titanium (200 ± 40 HV). The niobium plates were initially in were attributed to Nb(Cr, Ni)2 .
annealed state, and the welding did not induce additional softening: an The MZCu resisted up to 162 ± 10 MPa (Fig. 13e, dark dashed line).
average of 95 ± 15 HV was measured near the weld limit and elsewhere The tensile curve is very similar to the MZNb curve, and the fracture
in the solid insert. propagated only in IML developed next to the solid niobium. Fe7 Nb6 ,
Due to the precipitation of Fe2 Nb, MZNb/ 316 L had Vickers hardness Nb(Cr, Ni)2 and a small amount of metastable 𝜀-Fe were found on both
comprised between 375 and 475 HV, which is twice higher compared sides of the fracture (Fig. 13b). Cu was the main phase of the melted
to raw 316 L (Fig. 12b). These measures did not include fissured IML zone side. Its presence on the niobium side is explained by some Cu-rich
due to its thickness and cracks. precipitates in the IML. Cu peaks were enlarged due to the presence of
The hardness of MZCu was very fluctuating (Fig. 12c): soft regions minor quantities of 𝛾-Fe (Fig. 13c). Unlike the Cu-free Nb/316 L weld,
between 118 and 150 HV corresponded to Cu-rich matrix. The steel- 𝛼-Fe from steel-rich matrix was less present. Fe2 Nb is observed in IML,
rich zones had hardness of ~210 HV on 316 L side and around 370 HV but its peak on the melted zone side was very small, which can be ex-
in the center. At niobium/MZCu interface, IML hardness was above plained by very low solubility of niobium in copper that did not allow
800 HV. the formation of Fe + Fe2 Nb eutectic in steel-rich zones. The niobium
Fig. 12d shows that the precipitation of Nb dendrites produced some side also revealed peaks of solid niobium due to the thickness of the
hardening in MZNb/Cu (109 HV) compared to the unmelted copper in- fractured IML.
terlayer (74 ± 4 HV) and solid niobium (95 ± 15 HV). The hardness of The titanium/niobium/1 mm copper/316 L dissimilar joint showed
the copper/316 L stainless steel zone varied between 130 and 207 HV, UTS of 255 ± 10 MPa (Fig. 13e, red dotted line) with a ductile fracture
depending on the local segregation of Cu, which is close to raw stainless in the unmelted copper interlayer (Fig. 13d). The combination of 1 mm
steel (194 HV). The unmelted copper insert underwent visible recrystal- thick copper and niobium inserts prevented the formation of an inter-
lization and showed a hardness of 70 HV, identical to the measurements metallic layer along the niobium edge. In this way, a joint coefficient
in HAZ of standalone copper/316 L welds that correspond to partially of 100% was attained compared to partially annealed copper, and 88%
annealed copper (Mannucci et al., 2018b). compared to annealed niobium.
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

Fig. 13. Fracture surfaces analysis: XRD of (a) niobium/316 L and (b) niobium/0.2 mm copper/316 L fracture surfaces with (c) zoom on Cu and 𝛾-Fe XRD peak at
110 °; (d) the reconstitution of broken niobium/1 mm copper/316 L sample with EDS line analysis; (e) tensile curves of Nb-containing welds.

Conclusions The welds between titanium and vanadium contained cubic (Ti, V)
solid solutions and were free from intermetallic phases. The formation
The use of a vanadium interlayer associated with double pass weld- of martensitic structures in HAZ of Ti-6Al-4V did not induce fractures
ing allowed obtaining resistant joints between titanium and stainless during the tensile test. The welds between vanadium and 316 L
steel with a joint coefficient of 100 % regarding to raw vanadium. A contained only cubic (Fe, V) solid solutions even for low (6 at.% V)
ductile fracture occurred in the residual vanadium insert, when the dilution of insert. The beam offset on 316 L allows crack-free columnar
amount of vanadium in the melted zone did not exceed 30 at.% V. The solidification perpendicular to the limits of the melted zone, when a
presence of HAZ in the unmelted vanadium did not weaken the insert, centered beam position produced columnar structures next to 316 L
and the joint coefficient reached 100% compared to initial laminated side and coarse equiaxed grains in the rest of the melt. The beam offset
vanadium. on vanadium produced a melted zone with 47 at.% V with increased
A. Mannucci, I. Tomashchuk and A. Mathieu et al. Journal of Advanced Joining Processes 1 (2020) 100022

hardness and intergranular cold cracks between coarse recrystallized Hajitabar, A., Naffakh-Moosavy, H., 2017. Electron beam welding of difficult-to-weld
grains. However, EBSD analysis did not reveal the precipitation of the 𝜎 austenitic stainless steel/Nb-based alloy dissimilar joints without interlayer. Vacuum
146, 170–178.
phase. The observed hardening can be attributed to a nanoscale phase Hajitabar, A., Naffakh-Moosavy, H., 2018. The effect of FexNby (x = 2, 7 and y = 1, 6)
separation occurring in non-equilibrium conditions. intermetallics on microstructure and mechanical properties of electron beam welded
The use of niobium insert resulted in dissimilar joints with UTS of Nb-1Zr refractory alloy. Int. J. Refractory Metals Hard Mater. 76, 192–203.
He, P., Feng, J.C., Zhang, B.G., Qian, Y.Y., 2003. A new technology for diffusion bonding
160 ± 10 MPa only, due to the brittle fracture occurring in the inter- intermetallic TiAl to steel with composite barrier layers. Mater.Charact 50, 87–92.
metallic layer at the interface between solid niobium and niobium/316 L Ivanchenko, V., Pryadko, T., 2008. Chromium–iron–titanium. In: Effenberg, G., Ilyenko, S.
melted zone, containing Fe7 Nb6 and Fe2 Nb intermetallics. The welds be- (Eds.), Ternary Alloy Systems. Springer, Berlin, p. 365.
Ishida, K., Gao, Y., Nagatsuka, K., Takahashi, M., Nakata, K., 2015. Microstructures and
tween titanium and niobium systematically resisted during the tensile
mechanical properties of friction stir welded lap joints of commercially pure titanium
test thanks to the absence of brittle phases in the Nb-Ti system. The ad- and 304 stainless steel. J.Alloys Compd. 630, 172–177.
dition of a 0.2 mm thick copper insert between niobium and steel was Kimura, M., Iijima, T., Kusaka, M., Kaizu, K., Fuji, A., 2016. Joining phenomena and tensile
strength of friction welded joint between Ti–6Al–4V titanium alloy and low carbon
not sufficient to prevent the formation of brittle phases and resulted
steel. J.Manufact.Process. 24, 203–211.
in similar UTS values. Finally, the creation of isolated niobium/copper Lyakishev, N.P., 1997. Spravochnik Diagrammy Sostoyania Dvoynyh Metallicheskih sis-
and copper/316 L melted zones with use of a 1 mm thick copper in- tem, Mashinostroenie. Russian Federation, Moscow.
terlayer allowed completely avoid the formation of brittle phases. The Mannucci, A., Tomashchuk, I., Mathieu, A., Cicala, E., Boucheron, T., Bolot, R., Lafaye, S.,
2018a. Direct laser welding of pure titanium to austenitic stainless steel. Procedia
niobium/copper melted zone containing 15 at.% Nb was composed of CIRP 74, 485–490.
Nb dendrites dispersed in a Cu matrix. The copper/316 L melted zone Mannucci, A., Tomashchuk, I., Vignal, V., Sallamand, P., Duband, M., 2018b. Paramet-
containing 17 at.% Cu had a 𝛾-Fe structure with micro segregations of ric study of laser welding of copper to austenitic stainless steel. Procedia CIRP 74,
450–455.
Cu. The ductile fracture took place in the remaining unmelted copper Mannucci, A., Tomashchuk, I., Mathieu, A., Bolot, R., Cicala, E., Lafaye, S., Roudeix, C.,
insert at UTS of 255 ± 10 MPa, which corresponds to 100% joint co- 2019a. Pure vanadium insert for efficient joining of Ti6Al4V to 316L stainless steel
efficient compared to copper and 88 % joint coefficient compared to with continuous Yb:YAG laser. In: Lasers in Manufacturing Conference 2019. Munich,
Germany.
niobium. Mannucci, A., Bolot, R., Mathieu, A., Tomashchuk, I., Cicala, E., Roudeix, C., Lafaye, S.,
2019b. Numerical simulation of residual stresses in laser welding: application to
Declaration of Competing Interest Ti6Al4V/316L steel assembly with vanadium insert. In: Lasers in Manufacturing Con-
ference 2019. Munich, Germany.
Mannucci, A., Tomashchuk, I., Mathieu, A., Bolot, R., Cicala, E., Lafaye, S., Roudeix, C.,
The authors declare that they have no known competing financial 2019c.Use of pure vanadium and niobium inserts for dissimilar welding of titanium
interests or personal relationships that could have appeared to influence to stainless steel.In: Laser Processing for Industry Conference, Colmar, France.42–45
Nogami, S., Miyazaki, J., Hasegawa, A., Nagasaka, T., Muroga, T., 2013. Study on electron
the work reported in this paper.
beam weld joints between pure vanadium and SUS316L stainless steel. J.Nucl.Mater.
442, S562–S566.
Acknowledgments Raghavan, V., 2004. Fe-Nb-Ni (Iron-Niobium-Nickel). J.Phase Equilib.Diffus 25, 552.
Song, T.F., Jiang, X.S., Shao, Z.Y., Fang, Y.J., Mo, D.F., Zhu, D.G., Zhu, M.H., 2017. Mi-
crostructure and mechanical properties of vacuum diffusion bonded joints between
This work was carried out as a part of a joint laboratory project Lab- Ti-6Al-4V titanium alloy and AISI316L stainless steel using Cu/Nb multi-interlayer.
Com FLAMme between Laboratoire Interdisciplinaire Carnot de Bour- Vacuum 145, 68–76.
gogne, University of Bourgogne-Franche-Comté and SME Laser Rhône- Sun, Z., Ion, J.C., 1995. Laser welding of dissimilar metal combinations. J.Mater.Sci. 30,
4205–4214.
Alpes. This project was supported by French National Agency of Re- Takeyama, M., Morita, S., Yamauchi, A., Yamanaka, M., Matsuo, T., 2001. Phase equilibria
search (grant number ANR-14-LAB3-0005). among gamma, Ni3Nb-delta and Fe2Nb-epsilon phases in Ni-Nb-Fe and Ni-Nb-Fe-Cr
Authors would like to thank Dr. Ing. Bianca Frincu and Mr. Philippe systems at elevated temperatures. In: Proceedings of the International Symposium
Superalloys 718, 625, 706 and various derivatives, Pittsburg, USA, pp. 333–344.
Charobert (INDUSTEEL CRMC of Le Creusot, France) for their kind ad- Tomashchuk, I., Grevey, D., Sallamand, P., 2015. Dissimilar laser welding of AISI 316L
vice on sample preparation and carrying out EBSD analysis. stainless steel to Ti6–Al4–6V alloy via pure vanadium interlayer. Mater.Sci.Engineer.A
622, 37–45.
References Tomashchuk, I., Sallamand, P., 2018. Metallurgical strategies for the joining of titanium
alloys with steels. Adv Eng Mater 20, 1700764.
Ustinovshchikov, Y.I., Pushkarev, B.E., Sapegina, I.V., 2005. Mechanism of sigma-phase
Adomako, N.K., Kim, J.O., Lee, S.H., Noh, K.-.O., Kim, J.H., 2018. Dissimilar weld-
formation in the Fe-V system. Inorg.Mater. 41, 822–826.
ing between Ti–6Al–4V and 17-4PH stainless steel using a vanadium interlayer.
Wang, C.P., Liu, X.J., Ohnuma, I., Kainuma, R., Ishida, K., Hao, S.M., 2000. Phase equi-
Mater.Sci.Eng.A 732, 378–397.
libria in the Cu-Fe-Mo and Cu-Fe-Nb systems. J.Phase Equilib. 21, 54.
Adomako, N.K., Kim, J.O., Kim, J.H., 2019. Microstructural evolution and mechanical
Xie, G.M., Yang, D.H., Luo, Z.A., Li, M., Wang, M.K., Misra, R., 2018. The Determining Role
properties of laser beam welded joints between pure V and 17-4PH stainless steel.
of Nb Interlayer on Interfacial Microstructure and Mechanical Properties of Ti/Steel
Mater.Sci.Eng.A 753, 208–217.
Clad Plate by Vacuum Rolling Cladding. Materials (Basel) 11, 1983.
Cao, C.D., Sun, Z., Bai, X.J., Duan, L.B., Zheng, J.B., Wang, F., 2011. Metastable phase
Zhang, Y., Sun, D., Gu, X., Li, H., 2016. A hybrid joint based on two kinds of bonding
diagrams of Cu-based alloy systems with a miscibility gap in undercooled state.
mechanisms for titanium alloy and stainless steel by pulsed laser welding. Mater.Lett.
J.Mater.Sci 46, 6203–6212.
185, 152–155.
Chen, Y.Z., Liu, F., Yang, G.C., Xu, X.Q., Zhou, Y.H., 2007. Rapid solidifica-
Zhang, Y., Sun, D., Gu, X., Li, H., 2018. Microstructure and mechanical property improve-
tion of bulk undercooled hypoperitectic Fe–Cu alloy. J.Alloys Compd. 427,
ment of dissimilar metal joints for TC4 Ti alloy to 301L stainless steel. J Mater Sci 53
L1–L5.
(4), 2942–2955.
Gao, X.-.L., Liu, H., Liu, J., Yu, H., 2019. Laser welding of Ti6Al4V to Cu using a niobium
Zhao, D.S., Yan, J.C., Liu, Y.J., Ji, Z.S., 2014. Interfacial structure and mechanical proper-
interlayer. J.Mater.Process.Technol. 270, 293–305.
ties of hot-roll bonded joints between titanium alloy and stainless steel using niobium
Gloc, M., Wachowski, M., Plocinski, T., Kurzydlowski, K.J., 2016. Microstructural and mi-
interlayer. Trans. Nonferrous Metals Soc. China 24 (9), 2839–2844.
croanalysis investigations of bond titanium grade1/low alloy steel st52-3 N obtained
by explosive welding. J.Alloys Compd. 671, 446–451.

You might also like