Content uploaded by Sunil Bhat
Author content
All content in this area was uploaded by Sunil Bhat on Jun 03, 2023
Content may be subject to copyright.
Technical Report
Metallurgical and mechanical examinations of steel–copper joints arc welded
using bronze and nickel-base superalloy filler materials
M. Velu
⇑
, Sunil Bhat
1
School of Mechanical and Building Sciences, Vellore Institute of Technology University, Vellore 632 014, Tamil Nadu, India
article info
Article history:
Received 17 October 2012
Accepted 30 December 2012
Available online 9 January 2013
abstract
The paper presents metallurgical and mechanical examinations of joints between dissimilar metals viz.
copper (UNSC11000) and alloy steel (En31) obtained by Shielded Metal Arc Welding (SMAW) using
two different filler materials, bronze and nickel-base super alloy. The weld bead of the joint with
bronze-filler displayed porosity, while that with nickel-filler did not. In tension tests, the weldments with
bronze-filler fractured in the centre of the weld, while those with nickel-filler fractured in the heat
affected zone (HAZ) of copper. Since the latter exhibited higher strength than the former, all the major
tests were undertaken over the joints with nickel-filler alone. Scanning Electron Microscopy (SEM) cou-
pled with Energy Dispersive Spectroscopy (EDS) indicated corrugated weld interfaces and favorable ele-
mental diffusions across them. X-ray diffraction (XRD) studies around the weld interfaces did not reveal
any detrimental intermetallic compounds. Transverse bending tests showed that flexural strengths of the
weldments were higher than the tensile strengths. Transverse side bend tests confirmed good ductility of
the joints. Shear strength of the weld-interface (Cu–Ni or Ni–steel) was higher than the yield strength of
weaker metal. Microhardness and Charpy impact values were measured at all the important zones across
the weldment.
Ó2013 Elsevier Ltd. All rights reserved.
1. Introduction
Welding of dissimilar metals is done in order to satisfy different
requirements for performance. Mutual solubility between the dif-
ferent metals to be joined is the primary requirement to obtain
metallurgical bond between them. Alternatively, another metal
(filler-metal) that is soluble with each of the dissimilar metals to
be joined is required to make a joint. If thermal-expansion coeffi-
cients of the two metals to be joined are considerably different,
then the joint may fail due to thermal fatigue. Joining of metals
having different melting temperatures or thermal conductivities
is also difficult because one metal will melt before the other [1].
Recently, successful joining of dissimilar metals with filler metals
was reported [2]. Use of buffer layer between the parent metal
and the weld metal, in welding repair (filling the section removed
by erosion wear) of the former was reported [3].
Copper–steel joints are employed in the field of power genera-
tion and transmission, cryogenics, electrical and electronics to uti-
lize high electrical conductivity of copper and high strength of
steel. Their welding too has not been without problems. Copper
conducts heat energy up to 10 times faster than steels, which in
turn tends to dissipate heat rapidly away from the weld leading
to difficulties in reaching the melting temperature of copper. In
addition, according to Fe–Cu phase diagram, limited solubility of
Cu in Fe exists. Another major problem in their welding is hot
cracking in the heat-affected zone (HAZ) of steel due to copper
penetration into grain boundaries of steel. Although, copper and
steel alloys were bonded successfully by solid-state, non-fusion,
processes namely explosive welding and diffusion bonding with
and without interlayer [4–10], butt joints between these metals
by fusion welding methods without the use of filler materials
[11–15] have produced mixed results, at times necessitating alter-
ations in welding procedures. In one of the cases, 1 mm thick plates
of copper and tool steel were welded by laser beam welding (LBW).
To overcome problems of welding copper to steel, laser beam focus
was offset by 0.2 mm [11] from the joint interface into the steel. A
limited amount of copper was melted by the molten steel. The
amount of copper dissolved in the molten steel was very limited
(<2%) and hence, no microcracks appeared at the interface between
steel and weld metal. Although, the steel plate was completely
melted, a complete metallurgical bond was not obtained at the
interface between the Cu-plate and the weld metal. In another
case, Magnobosco et al. [14] fabricated butt-welded joints between
copper and stainless steel by electron beam welding (EBW). Three
such joints with plate thickness 29, 50 and 71 mm were made. The
results showed copper penetrations into grain boundaries of the
steel base metal in one of the joints. Other joints showed cracks,
0261-3069/$ - see front matter Ó2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.matdes.2012.12.073
⇑
Corresponding author. Mobile: +91 9444551700; fax: +91 0416 2243092.
E-mail addresses: mvelu@vit.ac.in,velu20002001@yahoo.com (M. Velu), sunil.
bhat@vit.ac.in,sbbhat@rediffmail.com (S. Bhat).
1
Mobile: +91 9743299255.
Materials and Design 47 (2013) 793–809
Contents lists available at SciVerse ScienceDirect
Materials and Design
journal homepage: www.elsevier.com/locate/matdes
shrinkage and porosities. Information on the impact energy, ten-
sile, bending and shear strength of these joints is not reported yet.
The problems associated with fusion welding of copper and
steel can be overcome to a large extent by introduction of compat-
ible filler materials and their careful selection. In this direction,
copper and stainless steel plates and copper and low carbon steel
plates have been successfully TIG and arc welded using copper-
base filler-DIN 1753: E S Cu Sn7 [16,17]. However, nickel-base filler
is preferred because nickel has improved mechanical properties
and better solid solubility with both copper and iron without pro-
ducing intermetallic compounds [10,18]. Nickel has been success-
fully tried as the interlayer in solid-state diffusion bonding
between stainless steel and copper [10]. However, use of nickel
as filler metal in butt joining of copper and alloy steel by Shielded
Metal Arc Welding (SMAW) is not reported in the literature. SMAW
is commonly used in industry and can be easily performed [19] in
comparison to other welding techniques. It is also commercially
more viable. Therefore a comprehensive study of the butt joint be-
tween copper (UNSC11000) and alloy steel (En31) with nickel base
superalloy filler metal (SA1752), fabricated by Shielded Metal Arc
Welding (SMAW), was carried out and is reported in the present
work. Examinations of joints made with bronze filler were also
undertaken. Since such joints showed weld porosity and poor
strength in comparison with nickel filler joints, further detailed
investigations were carried out only on weldments with nickel fil-
ler. Microstructure and composition at all the crucial zones, near
and away from the weld-interfaces, were examined. Mechanical
properties of the joint and microhardness at all the zones were
measured. The weld joint made with nickel-filler is found to be
of good quality with strength near to copper.
2. Experimental procedure
About 10-mm thick commercially available plates of alloy steel
(En31) and copper (UNSC11000) were cut into blocks of
150 mm 40 mm 10 mm. Butt joints between these were made
by SMAW process using two types of flux covered, consumable
electrodes, copper-base ‘‘superbronze’’ (bronze-filler) and the nick-
el-base ‘‘superalloy’’ (nickel-filler). The chemical compositions of
steel, copper and filler materials are provided in Table 1. The impor-
tant thermophysical properties [20] of the corresponding metals
are shown in Table 2. These values provide a reference for under-
standing the weldability and diffusion behavior. The edges of the
plates were machined at an angle of 30°to obtain a final 60°-V
butt-weld groove. Weld surfaces were thoroughly cleaned before
welding. Copper plate was preheated to 540 °C to prevent dissipa-
tion of heat into it due to its high thermal conductivity. Electrode
was connected to positive and plates to negative terminal of the
power source. Diameter of the filler rods was 3.15 mm. The welding
voltage (V) and current (I) were recorded during each welding pass.
In addition, the time taken for each pass was noted to determine the
welding speed (S). The heat input was calculated as VI/S. The weld-
ing parameters are given in Table 3. Interpass cleaning was done be-
tween each pass. The welded samples were naturally cooled.
A transverse section was machined out from the welded sam-
ple using wire-cut Electric Discharge Machining (EDM) and pro-
cessed by standard metallographic techniques. A solution of
50 ml concentrated HNO
3
, 50 ml HNO
3
and 10 ml H
2
O was used
as etchant for copper-base metals for studying both the macro
and microstructures. The etchants used for steel were a solution
of 25% HNO
3
in water for macrostructure and 1–20 g sodium
metabisulphite + 100 ml water for microstructure [21].Mixed acid
(equal parts of HCl, HNO
3
and acetic acids) was used as etchant
for nickel-weld for studying both the macro and microstructures
[22]. The macro and microstructures were studied using macro-
scope (Macscope-z) and optical microscope (make: Carl Zeiss)
respectively.
Dilution levels were determined by dividing the fusion zone
into two symmetrical areas. The ratio of geometric cross-sectional
area of the melted copper (A
cu
) to the total melted cross-sectional
area of the filler metal and copper (A
cu
+A
fm
) i.e., one-half of the fu-
sion zone, is the dilution level (D
cu
=A
cu
/(A
cu
+A
fm
)) [23] of the fil-
ler-metal by the copper. Similar calculation is repeated for another
symmetric area to calculate the dilution level of the filler-metal by
steel. Average dilution levels for each weld metal type are pre-
sented in Table 4.
Transverse tensile specimens (length perpendicular to welding
direction) of rectangular cross-section, conforming to ASTM: E8/
E8M–09 standard test methods for tension testing of metallic
materials were machined and tested at room temperature, using
Electronic Tensometer. The gauge length was 25 mm and gauge
cross-section was 6 mm 2 mm. The samples were tested at a
strain rate of 3 mm/min at room temperature.
The weld bead of the joint made with bronze-filler showed
porosity, while that of nickel-filler did not. In tension tests, all
weldments of bronze-filler fractured at the centre of the weld.
Table 1
Chemical composition (wt%) of materials used.
Material Elements (wt%)
CFe NiCrMoMnSiPSNbTiAlCuSn
Alloy steel (EN 31) 0.779 97.236 0.092 1.059 1.059 0.587 0.182 0.035 0.030 – – – –
Copper (UNSC 11000) – – – – – 0.002 – – – – 0.005 99.953 –
Bronze-filler – – – ––––0.24 – – – – 92.73 7.03
Nickel-filler 0.027 2.52 67.6 20.57 – 6.05 0.44 0.01 0.01 2.31 0.1 – 0.16 –
Table 2
Thermophysical properties of materials at room temperature [20].
Metal Crystalline
structure, 20 °C
Atomic radius
(nm)
Most common
valence
Melting point
(°C)
Density, 20 °C
(g/cm
3
)
Thermal conductivity
a
(W/m K)
Coefficient of thermal Expansion
a
(10
6
(°C)
1
)
Fe BCC 0.124 2+ 1538 7.87 51.9 11.7
Cu FCC 0.128 1+ 1085 8.94 388 17
Ni FCC 0.125 2+ 1455 8.90 9.8 12.8
a
For respective alloys.
794 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
Whereas all weldments of nickel-filler fractured in the heat af-
fected zone (HAZ) of weaker parent metals (i.e., copper) with all
weld-interfaces remaining intact. From these observations, it was
concluded that nickel-filler was the best filler material for the joint
between copper and alloy steel. Hence, further tests like SEM–EDS
analysis, XRD studies, longitudinal tensile tests, hot tensile tests,
microhardness measurements, transverse bending tests, side bend
tests, shear tests and Charpy impact tests were done only for the
weldments made with nickel-filler.
Scanning Electron Microscope (SEM) coupled with Energy Dis-
persive Spectroscopy (EDS) is used to record elemental composi-
tion for studying alloying behaviour between metals across the
weld-interfaces. X-ray diffraction (powder XRD, make: Bruker,
model: D8 Advance) was used to identify different phases present
at the interfaces of base metals and the weld metal.
Scheme for removal of test specimens from dissimilar welded
copper–steel weldment is shown in Fig. 1. Longitudinal (in the
direction of welding) tensile specimens of base metals (BMs), heat
affected zones (HAZs) and weld metal (WM); and of rectangular
cross-section, conforming to ASTM-E8 standard were machined
from copper–steel weldments made of nickel-filler metal and
tested at room temperature, using Electronic Tensometer. The
gauge length was 25 mm and gauge cross-section was
6mm5 mm. The samples were tested at a strain rate of 3 mm/
min at room temperature. The plates of copper and alloy steel of
300 mm 40 mm 6 mm were welded using nickel-filler. From
this weldment, two transverse tensile specimens of overall size
of 600 mm 20 mm 6 mm were machined out. The gauge
length was 50 mm and gauge cross-section was 12.5 mm 6 mm.
These specimens were tested at an elevated temperature of 100 °C
and 400 °C. The extra lengths of the specimens were to grip the
specimens far away from the heat source. Microhardness values
were measured across the weld metal. Applied load of 300 g and
dwell time of 10 s were set in hardness tester (make: Matsuzawa,
model: MMT-X7B). Transverse specimens of 80 mm 10 mm
10 mm were machined. A transverse bending test to assess the
flexural strength was carried out using universal testing machine
(UTM). The specimens were simply supported at the ends and
the load was applied on the transverse section of the weld as
shown in Fig. 2. A load was gradually applied at the centre of the
specimen until the fracture of these specimens occurred. A side
bend test to assess the ductility and soundness of the weld joints
was performed on transverse specimens of 80 mm 10 mm
5 mm using UTM. The guidelines of ASTM: E190-92 (reapproved
2008) Standard test method for guided bend test for ductility of
welds and ISO 5177 [24] were followed. The specimens were sup-
ported by £50 mm cylinders kept 24 mm apart as shown in Fig. 3.
A former with £8 mm was used to apply the load gradually on the
transverse section. The load was applied until the maximum bend-
ing-angle (
a
) was reached in these specimens. Transverse speci-
mens of circular cross-section of £6.7 mm and 80 mm length
were machined out and shear tested in UTM as shown in Fig. 4.
The positions of the specimens in the fixture were varied to cause
the shear at the weld-interfaces. The Charpy V-notch impact spec-
imens of 55 mm 10 mm 10 mm, as per ASTM: E23–07a Stan-
dard test methods for notched bar impact testing of metallic
materials were machined out from all important zones of the weld-
ment and tested at room-temperature using impact testing
machine.
3. Results and discussion
3.1. Macrostructures
Fig. 5a–c shows the top-view photographs of the dissimilar
welded samples. The weldment made with bronze-filler has a
non-uniform weld bead with many porosity defects. It also dis-
plays large number of spattered particles on either side of the bead.
Whereas the other with nickel-filler has a relatively uniform weld
bead without common weld defects including hot cracking, under-
cutting, lack of fusion, porosity and slag entrapment. No spattering
occurred in it. Fig. 6a and b displays the macrostructure of the
transverse section of the welds; randomly distributed large poros-
ity defects appear in the weld of the bronze-filler, whereas the
weld of the nickel-filler [2] has no weld defects. It can be observed
from the wavy interfaces that both the filler materials penetrate
well into the base metals. The extent of HAZ of steel, number of
passes, extent of penetration, symmetry of the bead and moderate
convexity of the face reinforcement are the notable features seen in
the weld of nickel-filler. No distinct HAZ of copper is seen.
3.2. Microstructures
Refer Fig. 7. Two weld-interfaces Aand Bare defined in the
specimen, interface Abetween copper and weld metal (nickel-fil-
ler) and interface Bbetween weld metal and alloy steel. Zones 1
to 7 are also identified for discussion of results of various
Table 3
Welding parameters.
Filler materials Pass
number
Current
(A)
Voltage
(V)
Welding
speed
(mm/s)
Heat
input
(kJ/
mm)
Total
heat
input
(kJ/
mm)
Bronze-filler 1 100 30 3.33 0.9 4.28
2 104 23 0.72
3 98 30.4 0.89
4 96 30.8 0.89
5 (Root
pass)
100 29.2 0.88
Nickel-filler 1 120 24 3 0.96 4.77
2 120 22 0.88
3 112 28 1.05
4 120 24 0.96
5 (Root
pass)
120 23 0.92
Welding parameters for
Cu–steel weldments
of Ni-filler used for
hot-tensile tests (V-
groove length was
40 mm and plate
thickness 6 mm)
1 116 25 2 1.45 5.95
2 116 25 1.45
3 112 28 1.57
4 (Root
pass)
114 26 1.48
Table 4
Average dilution levels (%) for the weld metals investigated.
Weld
metal
type
Specimen
No.
Dilution
level by
copper
base
metal (%)
Dilution
level by
alloy steel
base
metal (%)
Average
dilution
level by
copper base
metal (%)
Average
dilution
level by
alloy steel
base metal
(%)
Bronze-
filler
1 21.2 23.07 22.45 19
2 25.42 11.53
3 23.2 19
42022
Nickel-
filler
1 29.82 23.4 22.44 23.23
2 14.9 15.55
33230
41324
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 795
examinations. Zones 1 and 7 represent parent copper and alloy
steel respectively, Zone 4 represents the centre of the weld metal,
Zones 3 and 5 represent weld metal very close to the interface A
and Brespectively, whereas Zones 2 and 6 represent heat affected
zone (HAZ) of copper and steel respectively.
Fig. 8 shows the microstructures at various zones of the weld-
ment made with bronze-filler. Fig. 8a shows a typical microstruc-
ture of copper [10,11] at Zone 1, far away from the weld
interface A. No grain growth is observed in Fig. 8b, the microstruc-
ture of copper at Zone 2, close to weld-interface A. This confirms
that the copper is less affected by the welding heat. Fig. 8c presents
microstructure at weld-interface A. A defect-free interface is
clearly seen that confirms a good joint. Because the bronze-filler
is copper-base, it alloyed well with the copper, without any hot
cracking and porosity. Fig. 8d illustrates microstructure at Zone
3, in bronze–weld metal. It shows coarse-grained structure, with-
out any noticeable secondary phases. Fig. 8e illustrates microstruc-
ture at Zone 4, in the centre of the bronze–weld metal. It shows
that it is unaffected by both the base metals. Fig. 8f shows micro-
structure at Zone 5, closer to the steel. It shows randomly distrib-
uted carbides, formed due to the diffusion of carbon into the weld
metal. Fig. 8g presents microstructure at weld-interface B. This
interface also appears to be sound. Dark and irregular shaped chro-
mium carbide particles are found to be randomly dispersed on
150
82
2
10 10
55
10
10
10 10
5
Copper plate
Steel plate
Discard
Transverse-bend specimen
Charpy V- Notch
Impact specimens
V-Notch
Side-bend specimen
Ni-WM
Cu (BM)
Cu (HAZ)
Ni (WM)
Steel (HAZ)
Steel (BM)
Transverse tensile specimen (Bi-metallic)
Impact specimen (Bi-metallic)
Discard
Transverse-bend
specimen
All dimensions in mm
150
82
10
10 25
100
5
10
Steel plate
Discard
Ni-WM
Copper plate
Discard
Longitudinal tensile
specimen- Cu (BM)
Cu (HAZ)
Ni (WM)
Steel (HAZ)
Steel (BM)
All dimensions in mm
(a)
(b)
Fig. 1. Removal of test specimens from Shielded Metal Arc Welded (SMAW) dissimilar copper and alloy steel: (a) specimens other than longitudinal tensile specimens (b)
longitudinal specimens.
796 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
both the sides of the interface. Microstructure of HAZ in alloy steel,
at Zone 6 is in Fig. 8h. It consists of bainite (side plates), pearlite
(nodules) and ferrite, but less martensite [25, p. 404]. This was be-
cause of the preheating of steel near the V-groove, while the whole
of copper was preheated. Finally, microstructure of parent alloy
steel at Zone 7, far away from the interface B,isatFig. 8i. It consists
of light-etching ferrite and a dark-etching pearlite. The higher
amount of pearlite is due to its higher carbon (0.779 wt%) content.
Fig. 9 shows the microstructures at various zones of the weld-
ment made with nickel-filler. Fig. 9a shows microstructure of cop-
per at Zone 1. It exhibits large, equiaxed twinned grains [10,11,14]
and dispersion of Cu
2
O (dark dots) particles caused by oxygen
penetration during heating. No grain growth is observed in
Fig. 9b, the microstructure of copper at Zone 2, close to weld inter-
face A. This confirms that the copper is less affected by welding
heat. Carbides of nickel–weld metal were distributed in the copper
matrix. These might have suppressed the grain growth in copper,
expected of any HAZ. Fig. 9c presents microstructure at
Fig. 2. Schematic diagram illustrating the method of transverse bending test on arc
welded copper–steel joint made with nickel-filler material.
Fig. 3. Schematic diagram illustrating the method of side bend test on arc welded copper–steel joints made with nickel-filler material.
(a)
Fixed
frame
in UTM
Fixture
(b)
Pulled upwards
Test specimen
Fig. 4. Shear testing of arc welded copper–steel joints made with nickel-filler material: (a) and (b) specimens in the fixture.
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 797
weld-interface A. A defect-free interface [10,26] is clearly seen that
confirms a good joint. Because the nickel is fully soluble with cop-
per in both liquid and solid, it alloyed well with the copper, with-
out any hot cracking and porosity. A corrugated interlocking fusion
line (FL) confirms that the nickel-filler material penetrated well
into the copper matrix. The similar atomic structures of copper
and nickel, face-centred-cubic (fcc) enabled the epitaxial growth
of the dendrites. The fusion zone also shows the distinct grain
boundaries (GB1, GB2) [22, p. 58],[23] formed in the filler side.
Fig. 9d–f displays the weld microstructures of nickel-weld at Zones
3, 4, and 5 respectively. It can be observed that microstructure is
fully austenitic [19,23]. Cluster of dark nodules in Fig. 9d may be
attributed to microsegregation [22, p. 21] of Cu in Ni and carbides
of the Ni-weld. The columnar dendritic [2,23] solidification struc-
ture (Fig. 9e) is clearly formed due to the presence of niobium,
which has a tendency to increase the bulk solidification tempera-
ture range. It is seen that carbides (dark particles in Fig. 9f); NbC,
TiC, MC, M(C, N), M
7
C
3
and M
23
C
6
are formed in interdendritic
[19],[22, p. 17] regions. The presence of coloured regions
(Fig. 9f) can be attributed to the precipitation of
c
0
-Ni
3
(Ti) and
c
00
-Ni
3
Nb phases. These phases are responsible for strengthening
of the nickel-base superalloy weld metal [22, p. 4]. There is no evi-
dence on the formation of topologically closed phases (such as
r
,P,
l
and Laves). Fig. 9g displays the microstructure at the weld-inter-
face B. This interface also appears to be sound [19] without any
porosity or solidification cracks. The absence of porosity, solidifica-
tion cracking and grain growth in the HAZ of steel is due to the
presence of deoxidizers (Mn, Ti), high Mn/S ratio and less carbon
content, and the formation of various grain growth suppressing
carbide formers in the nickel-filler metal respectively. Further,
the use of low input and multipass welding process helped to pro-
duce high tough refined columnar grains in the weld metal and
prevent grain growth in the HAZ of steel. Therefore, high tough
HAZ is produced. The preheating also helped prevention of hydro-
gen cracking [25, p. 394]. The coloured feature in the weld metal
side of the interface can be attributed to the various solid-solution
strengthening and precipitation hardening phases of the nickel-
base ‘superalloy’ filler-metal. The interface features a partially
melted zone (PMZ), of steel that was heated to below the liquidus
temperature but above the solidus temperature [23], so it was only
partially melted. The tendency of the weld metal melting into the
steel is clearly observed. The large dark nodules present adjacent to
the fusion line (FL) are indicative of carbides formed by reaction of
chromium with the carbon. The small dark particles scattered
away from the FL are pearlites, and elongated and needle shaped
cementite (Fe
3
C) particles are indicative of formation of bainite
structure. There is no clear evidence of formation of fully martens-
ite structure. This may be due to the preheating of the steel near
the V-groove, which has reduced the cooling rate and thereby re-
duced the formation of martensite [25, p. 404].Fig. 9h shows the
microstructure of HAZ of steel. It consists of finer pearlite (dark
nodules) and elongated and needle shaped cementite (Fe
3
C) parti-
cles suggestive of bainite structure within ferrite matrix. Thus, the
HAZ of fusion welds in Fe-based alloys has a complex structure
[19]. The distribution of carbides (large dark particles) found to
have diminished, when compared to its presence near the FL.
Fig. 9i shows the typical microstructure of parent steel at Zone 7.
Obviously, the effect of HAZ is absent.
3.3. SEM–EDS analysis
Fig. 10a–e illustrates SEM images coupled with corresponding
EDS results at all the selected zones. Findings from SEM images
Porosity inside the weld
Weld (Bronze-filler)
Alloy steel
Copper Weld (Bronze-filler)
Porosity Spattering
(a)
(b)
Copper
Alloy steel
Ni-Weld
No Spattering
(c)
Fig. 5. Shielded Metal Arc Welded (SMAW) dissimilar copper and alloy steel: (a)
using bronze-filler (b) showing porosity defects appearing after milling into 2 mm
(c) using nickel-filler material respectively.
Porosity
Weld
Copper Alloy steel
(a)
Copper
Weld
Alloy steel
(b)
Fig. 6. Macrostructure of the arc welded dissimilar copper and alloy steel using: (a)
bronze (b) nickel-filler materials respectively.
798 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
are in good agreement with those from optical microscope. Images
confirm perfect bonding of copper with weld metal, and weld me-
tal with alloy steel, without major or noticeable weld defects. The
EDS element analysis (in wt%) at various zones are given in Table 5.
Oxidation during welding process caused the presence of the
element ‘O’ in the weldment. However, the least values of ‘O’ in
nickel–weld metal Zones 3, 4 and 5 when compared to the other
zones are due to the formation of a protective Cr
2
O
3
surface oxide
layer [22, p. 15] to prevent the penetration of oxygen into the weld
metal. This will also result in high corrosion resistance of the weld.
The composition at Zone 1 and 7, are of Cu and alloy steel parent
metals respectively. The Zones 2 and 6 have less Ni composition
(wt%) 4.55 and 0.11 respectively. Because, these represent the par-
tially melted zones (PMZs), inside the HAZs of the respective base
metal, only less amount of nickel can enter it. As was seen in the
microstructures, the PMZ was embedded partially with nickel–
weld metal. Another reason for less amount of Ni in Zones 2 and
6 may be its lower intrinsic diffusion coefficient [10,26] than Cu
Fig. 7. Transverse cross-section of the arc welded dissimilar copper–steel weldment, showing various zones.
Bainite
1: Copper (base metal)
2: HAZ of copper
A: weld interface Cu-WM
3: WM near interface A
4: centre of the WM
5: WM near interface B
B: weld interface WM-steel
6: HAZ of steel
7: Steel (base metal)
(a) (c) (b)
Bronze-WM
(d) (f) (e)
(g) (i) (h)
1 2 A
3 4 5
B 6 7
Cu
Bronze
-WM
Steel
Fig. 8. Microstructures at different zones of arc welded dissimilar Cu–steel weld made using bronze-filler material (magnification 200, bright field and linear look up table).
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 799
and
a
-Fe. Furthermore, diffusion of small quantity of Si, S, Cr and
Mn from weld metal into copper at Zone 2 is evident. The compo-
sition at Zones 3 and 5, confirm that both Cu and Fe have diffused
to Ni-weld in greater quantity. Thus, it is understood that the ten-
dency of atomic diffusion from Cu and steel sides to the Ni side is
greater than that in the reverse direction [10]. At the same time,
the dilution level (28.12 wt%) of the weld metal by Fe from the
steel is found to be within allowable limits. Studies showed that
solidification cracking could occur, if dilution of the weld metal
by Fe is above approximately 40 wt% [22, pp. 373–374]. The com-
position at the centre of the weld metal at Zone 4 shows that it has
not diffused with either of the base metals, because of its faraway
location from both the interfaces Aand B. The composition at weld-
interfaces Aand B, once again confirm the better mixing of the
metals from each other.
3.4. XRD studies
By comparing the compositional analysis of elements through
EDS with equilibrium phase diagrams, the various phases present
at the interfaces can be identified. As per the EDS results, the ma-
jor elements (wt%) present at Cu–Ni interface (interface A) are O
(4.26%), Cr (9.69%), Mn (2.45%), Fe (1.35%), Ni (27.09%), Cu
(51.82%) and Nb (1.91%). The composition levels of elements Cr,
Mn, Fe, Cu and Nb are well within their maximum solid solubility
(wt%) in nickel (Cr-47%, Mn-20%, Fe-100%, Cu-100% and Nb-6%)
[22, p. 17]. Therefore, these elements could have completely dis-
solved in nickel and therefore did not produce any detrimental
brittle compounds. This indicates the presence of a solid solution
of nickel and copper (
a
Ni–Cu
), CuO and no intermetallic com-
pounds [10,26]. These findings are consistent with the binary
phase diagram of Ni–Cu, which shows a complete solid solubility
across the entire composition range. Further, these results were
also confirmed by XRD results in Fig. 11a. The peaks for elements
other than Ni and Cu were not appeared, due to their low inten-
sity. The major elements present at Ni–steel interface (interface
B) are O (18.24%), Cr (7.6%), Mn (2.2%), Fe (47.7%) and Ni
(23.13%). This indicates the presence of
c
Fe–Ni
, FeNi
3
, Ni, Cr,
r
FeC-
rNi
phases [10,19]. These results are consistent with the binary
phase diagram [21, p. 11–287] of Fe–Ni and the ternary phase
diagram of Fe–Ni–Cr system at the temperature range of 650–
900 °C[22, p. 25]. As is seen in Fig. 11b, the presence of these
phases in the fusion line of Ni–steel is confirmed by the XRD
results.
(d)
Carbides/ Microsegregation of
Cu in Ni
3
(c)
Epitaxial
growth
FL
GB 1
GB 2
Ni-WM
Cu
A
(a) (b)
Cu2O
1 2
(e) (f)
SGBS Carbides
Precipitates
5
4
Fig. 9. Microstructures at different zones of arc welded dissimilar Cu–steel weld made with nickel-filler material (magnification 200, bright field and linear look up table).
800 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
3.5. Tensile properties
Three transverse tensile specimens of welded joints made with
bronze-filler, namely E1, E2 and E3 as shown in Fig. 12a, fractured
at the weld during tensile testing at room temperature, as shown
in Fig. 12b. These specimens with porosities at the centre of weld
obviously fractured there, resulting in low stress and elongation,
as shown in the stress–strain plot in Fig. 12c. The stress and strain
values are lower than that of the copper (base metal) as shown in
Fig. 12d. Fracture at tin-bronze interlayer and lower tensile
strength (100–150 MPa) compared to other interlayers during dif-
fusion bonding of stainless steel and copper are reported [7]. The
stress–strain plots of the steel (base metal) are shown in Fig. 12e.
In welded joints made with nickel-filler, steel and weld-inter-
faces remained intact and fracture occurred in the heat affected
zones (HAZs) of copper in all the tested specimens (A1, A2 and
A3) as shown in Fig. 12f. This proved good structural integrity of
the weld joint. The fracture in Cu (HAZ) may be attributed to large
amount of Cu species diffusing into Ni than the reverse, leading to
possible formation of microvoids in Cu, which were observed by
other researchers [10,26]. Strength of the joint was confirmed to
be higher than copper. Stress–strain plots are displayed in
Fig. 12g. Good percent elongation values are obtained. These values
indirectly hint at less intermetallic compounds at weld-interfaces
otherwise the weld-interfaces would have been brittle and there-
fore would have failed before copper. No intermetallic compounds
at copper–nickel interface and minimal intermetallic compounds
at nickel–steel interface, as shown by XRD studies, also support
low brittleness. The average tensile properties of base metals and
weldments are given in Table 6. Weldments made with nickel-filler
showed greater average tensile strength values than with bronze-
filler. The maximum tensile strength of 242.72 MPa observed in
the weldment of Ni-filler is in proportion to that of 233.4 MPa in
laser beam welded (LBW) joint [13]. The strength and ductility of
the weldments made with nickel-filler are comparable to those
of copper base metals. From these results of the nickel-filler, it
can be concluded that it is the preferred choice of filler material
for Cu–steel dissimilar fusion welding.
The stress–stain plots of various zones and transverse tensile
specimen of Cu–steel joint made with nickel-filler were shown in
Fig. 12h. From the plots of copper (base metal) and copper (HAZ),
it is found that the copper was least affected by welding heat.
B
PMZ
Ni-WM SteelFL
Bainite
Carbides
Precipitates
(g)
(i)(h)
Bainite
Pearlite
6 7
Fig. 9. (continued)
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 801
Whereas it was the steel, that showed HAZ affects, by showing
highest stress and strain values. It is important to note higher
strain to fracture value that demonstrates that the HAZ of steel is
not fully brittle. These plots also show that the strength of copper
(e)
(d)
(c)
(b)
(a)
1
2
3
4
5
6
7
A
B
Cu
Ni
Steel
Ni
Fig. 10. SEM–EDS results at different zones of arc welded dissimilar Cu–steel weld made using nickel-filler material.
802 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
is the least of all constituent metals and therefore justify the frac-
ture in it. It is also evident that the nickel-weld is tough and duc-
tile. The important result is that the weldment made with nickel-
filler shows a stress–strain plot close to that of copper (base metal),
but with lesser strain value. This is due to the influence of low-
tough steel on the overall ductility of the weldment.
3.6. Hot tensile properties
Two transverse tensile specimens of welded joints made with
nickel-filler were tested at 100 °C and 400 °C. In both the samples,
the fracture occurred in the HAZ of copper, while steel and weld-
interfaces remained intact as shown in Fig. 13a. This enables the
use of this welded joint for high temperature applications. The
stress–strain plots at respective temperatures are shown in
Fig. 13b. As temperature increases, the maximum stress reduces.
Moreover, strain to fracture increases.
3.7. Microhardness measurements
The microhardness variation was measured on three different
depths and across the fusion lines of the nickel-weld, and is shown
in Fig. 14a. The hardness variation along the depth is very similar
[14], except in the HAZ of steel. The hardness variation is constant
across weld metal zone. These data confirm that the nickel–weld
metal is relatively homogenous, as was confirmed by microstruc-
tural study. The hardness values of copper are less than nickel
and steel. This indicates that its strength is also the least of all
and hence confirms its fracture during tensile testing. The hardness
values are the maximum in HAZ of steel. This can be attributed to
the formation of various carbides (dark nodules) and moderately
hard bainite in it. The hardness values declined gradually until
the distance of 6 mm (might be the width of HAZ) from interface
Band then remained flat throughout the steel (base metal). This
microhardness variation across the weldment is in line with the
previous studies [10,23].
Fig. 14b shows the hardness distribution in all zones, measured
at equidistant points, along the depth, beginning from top to bot-
tom. The values at the Zone 1 and Zone 2 on the copper show
the least variation. This indicates that copper is not affected much
by welding heat. The interface Aand Bshow the hardness values,
which are average of the values at nearby zones, namely Zones 2
and 3, and Zones 5 and 6 respectively. This is due to their mixed
composition. The Zones 3 and 4 in the weld metal show a flat dis-
tribution and confirm the homogeneity of the fusion zone.
Whereas, Zone 5 in the weld metal and very close to the interface
Bshows more and wide varying hardness. This is due to the forma-
tion of
a
-Cr precipitates [23], increase in carbon and iron content
due to diffusion from adjacent steel. The Zone 6, HAZ in the steel
shows a wide variation in the hardness values (300 HV-511.7
HV) with only one value reaching 584.9 HV. However, none of
the values exceeded 625–680 HV, the range of values associated
with the presence of martensite in typical high carbon tool steel
[11]. This hints at the formation of more pearlite and ferrite, with
bainite and no or less martensite in the microstructure of HAZ. Fi-
nally, Zone 7 in the unaffected steel (base metal) shows a mini-
mum variation in the hardness values.
3.8. Transverse bending test outcome
Fig. 15 shows transverse bending tested samples. The fracture
occurred in HAZ of copper, few millimetres from the Cu–WM inter-
face, and confirmed the soundness of the joint. The bending load at
fracture (P
u
), flexural strength (
r
u
) and the deflection are given in
Table 7. The flexural strength (
r
u
) is calculated as 3P
u
L/2bd
2
.
Where, Lis span of the beam, bis width and dis depth of the beam.
The flexural strength is found to be higher than the tensile
strength.
Table 5
EDS element analysis (wt.%).
Zones Element (Wt.%)
OK
a
Si K
a
Al K
a
SK
a
PK
a
Ti K
a
Ca K
a
Cr K
a
Cl K
a
Mn K
a
Fe K
a
Co K
a
Ni K
a
Cu K
a
Nb K
a
1 8.46 0.51 0.30 – – – – – – – – 90.74 –
2 3.36 0.32 – 0.65 – – 0.34 1.50 0.76 0.90 – – 4.55 87.63 –
Weld-interface A4.26 0.44 – 0.76 0.17 0.07 – 9.69 – 2.45 1.35 – 27.09 51.82 1.91
3 2.45 0.52 – – – 0.07 – 13.40 – 2.74 3.72 0.30 39.16 36.27 1.38
4 4.44 0.76 0.14 – – – 20.74 – 5.49 8.82 – 59.62 – –
5 1.24 0.78 – – – – – 17.42 – – 28.12 – 47.67 1.98
Weld-interface B18.24 0.43 – 0.69 – – – 7.61 – 2.20 47.70 – 23.13 – –
6 7.30 0.40 – – – – – 1.30 0.27 – 89.98 – 0.11 – –
7 15.54 0.36 – – – – – 1.13 – 0.52 82.44 – – – –
0
1000
2000
3000
4000
2-Theta
Lin (Counts)
XRD results (Cu-Ni interface)
Ni
Cu
α
Ni-Cu CuO
(a)
0
100
200
300
400
500
600
700
800
900
40 50 60 70 80 90 100 110 120 130 140
40 50 60 70 80 90 100 110 120 130 140
2-Theta
Lin (Counts)
XRD results (Ni-Fe interface)Ni
σ
(Fe, Cr, Ni)
Fe-Ni
FeNi3 Fe
(b)
Fig. 11. XRD results of arc welded dissimilar Cu–steel weld made using nickel-filler
material at (a) Cu–Ni, (b) Ni–steel interfaces respectively.
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 803
Porosity in the bronze-weld
Cu Steel
(a)
Cu Steel
Fracture in bronze-weld
(b)
(c)
0
50
100
150
200
Strain
Stress (MPa)
Specimen: E1
Specimen: E2
Specimen: E3
Cu-Steel weldment
(Bronze-filler)
(d)
0
50
100
150
200
250
300
Strain
Stress (MPa)
Specimen: 1
Specimen: 2
Specimen: 3
Cu (BM)
0
100
200
300
400
500
600
700
800
900
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0 0.1 0.2 0.3 0.4 0.5
Strain
Stress (MPa)
Specimen: 1
Specimen: 2
Specimen: 3
Steel (BM)
(e)
Fig. 12. Results of tensile tests on arc welded dissimilar copper–steel joint: (a) specimens before test (b) fractured specimens (c) stress–strain plots of joints (transverse)
made with bronze-filler; (d) stress–strain plots of copper (e) steel base metals; and (f) fractured specimens (g) stress–strain plots of joints (transverse) and (h) various zones
(longitudinal specimens) of joint made with nickel-filler material combined with stress–strain curve of transverse specimen for comparison.
804 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
(f)
Cu
Fracture in Cu (HAZ)
Ni-WM
Steel
0
50
100
150
200
250
300
Strain
Stress (MPa)
Specimen: A1
Specimen: A2
Specimen: A3
Cu-Steel weldment
(Nickel-filler)
(g)
(h)
0
100
200
300
400
500
600
700
800
900
0 0.1 0.2 0.3 0.4 0.5 0.6
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Strain
Stress (MPa)
Cu (BM)
Cu (HAZ)
Ni (WM)
Steel (HAZ)
Steel (BM)
Cu-Steel weldment
Longitudinal
specimens
Transverse specimen
Fig. 12. (continued)
Table 6
Tensile properties.
Material Sample No. Yield strength (MPa) Ultimate tensile strength (MPa) % Elongation (Gauge length 25) Strain to fracture
Copper (base metal) 1 134 247.8 32 0.68
2 150 259.32 28 0.56
3 150 285.4 40 0.73
Average 144.66 264.17 33.33 0.65
Alloy steel (base metal) 1 481 800 12 0.44
2 448 760 16 0.42
3 427 691 12 0.39
Average 452 750 13.33 0.416
Cu–steel weldment (Bronze-filler)
a
E1 93 122.8 6 0.04
E2 86.6 103.9 4 0.035
E3 115.4 188.4 7 0.09
Average 98.33 138.4 5.67 0.055
Cu–steel weldment (Ni-filler)
b
A1 100 206 12 0.37
A2 107 176 16 0.36
A3 170 242.72 8 0.50
Average 125.66 208.24 12 0.41
a
All specimens E1,E2,E3 fractured at the centre of the weld, due to porosities in the weld.
b
Fracture occurred in the HAZ of copper, in all three specimens.
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 805
Ni-WM
(a)
0
50
100
150
200
250
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Strain
Stress (MPa)
100
400
Hot tensile test on Cu-Steel
weldment (Nickel-filler)
°C
°C
(b)
Fig. 13. Results of transverse tensile tests on arc welded dissimilar copper–steel joints made with nickel-filler at high temperatures (a) fractured specimens, (b) stress–strain
plots.
(b)
(a)
0
100
200
300
400
500
600
700
Distance from centre of nickel-weld (mm)
Microhardness (HV300g)
Top
Centre
Bottom
Copper Ni-Weld
HAZ
Steel
Weld C. L
Interface A
Interface B
0
100
200
300
400
500
600
700
-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18
-16 -14 -12 -10 -8 -6 -4 -2
02468
10 12 14
16
Distance from centre of nickel-weld (mm)
Microhardness (HV300g)
Position: 1 (Top)
2
3
4
5
6
7
8
9
Position: 10 (Bottom)
12
Copper Ni-Weld
A
34
B
6
7
5
Steel
Interface A
Interface B
Weld C. L
1: Cu (BM)
2: Cu (HAZ) of copper
A: weld interface Cu-WM
3: WM near interface A
4: centre of the WM
5: WM near interface B
B: weld interface WM-steel
6: Steel (HAZ)
7: Steel (BM)
Fig. 14. Microhardness variation measured on the transverse section of arc welded dissimilar copper–steel joint made with nickel-filler (a) on three different depths of the
bead (b) in all the important zones, at equidistant points from top to bottom, along the depth of the bead.
806 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
3.9. Side bend test outcome
Fig. 16a and b shows side bend tested samples. The size of the
specimen for this test is in proportion to a standard side bend spec-
imen size 150 mm W10 mm, where ‘W’ is available width. The
initiation of crack (due to maximum load and this is not a weld-de-
fect) at the Cu–WM interface was observed. The maximum bend-
ing load, deflection and bending-angles (
a
) are given in Table 8.
Weld defects are not seen either in the interfaces or in weld. This
confirms that the present welding process is capable of producing
defect-free joints. The specimen size of the test is different from
that of the transverse bending test. The depth to span ratio of 10/
150 of standard bend test specimen is maintained by selecting
the depth of the current specimen as 5 mm and the span as
80 mm. Moreover, the maximum bending-angle (
a
) without allow-
ing the specimen to fracture into two parts is of interest in side
bend test of welded joints. Normally, the maximum bending angle
(
a
) for any successful similar (between same metals) welded joints
is 180°or the specimen will take a ‘U’-shape after the test. How-
ever, the present weldment is a dissimilar joint, therefore, a bend-
ing angle (
a
) lower than 180°is obtained.
3.10. Shear test outcome
The shear-tested samples are displayed in Fig. 17. Two of the
three specimens sheared at and near the Cu–WM interface.
Steel
Ni-WM
Fracture in Cu (HAZ)
Cu
Fig. 15. Transverse bending tested specimens of arc welded dissimilar copper–steel
joints.
Table 7
Transverse bending test results.
Specimen
no.
Specimen
size (mm)
Bending
load (P
u
)
at fracture
(N)
Maximum
deflection
(mm)
Flexural
strength
(
r
u
)
(MPa)
Remarks
18010 10 3500 7 420 Fracture
occurred
in copper
28010 10 3500 15 420
(b)
Crack
Steel
Ni-WM
Cu
(a)
Steel
Ni-WM
Cu
α
α
Fig. 16. Side bend tested specimens of arc welded dissimilar copper–steel joints
showing: (a) bent shape (b) surface under tension during bending.
Table 8
Side bend test results.
Specimen
no.
Specimen size
(mm)
Former position Maximum bending load
(N)
Maximum deflection
(mm)
Bending- angle (
a
)
(°)
Remarks
18010 5 Center of weld metal
(WM)
1340 12 40 Crack initiated near WM–
Cu
28010 5 Near WM–steel
interface
1240 10 35 Interface in all the
specimens.
38010 5 Near WM–copper
interface
1060 17 60
3Shear fracture
1
2
Cu
Steel
Ni-WM
Fig. 17. Shear testing of arc welded copper–steel joints made with nickel-filler
material: sheared specimens.
Table 9
Shear test results.
Specimen
no.
Diameter of
the specimen
(mm)
Area of the
specimen
(mm)
Shear
load
(N)
Shear
stress
(MPa)
Remarks
1 6.7 35.25 7800 221.27 Sheared at
Cu–WM
interface
2 6.7 35.25 13020 369.36 Sheared at
steel–WM
interface
3 6.7 35.25 5860 166.24 Sheared in
copper
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 807
Another sample sheared at the WM–steel interface. The shear
loads and stresses are presented in Table 9. The shear strength of
the interfaces is calculated as shear load/shear area. The shear
strengths of the weld-interfaces are well above the average yield
strength of the copper (base metal). Therefore, the weld-interfaces
will not shear before their yielding occurs. The shear strength of
the weldment (166.24 MPa), when sheared at copper, near to
Cu–Ni interface is more than that of diffusion bonded steel–Cu
couple (145 MPa) [10].
3.11. Impact energy
The Charpy impact tested samples are displayed in Fig. 18. The
welded joint fractured in HAZ of copper, while the steel and weld-
interfaces remained intact as shown in Fig. 18a. This confirms that
the weld joints are stronger than the weaker copper under impact
loading also. The impact specimens machined out from various
zones of the welded joint were subjected to this testing and their
fractured surfaces are displayed in Fig. 18b. The Charpy V-notch
(CVN) impact energy values are listed in Table 10. The large CVN
energy value and fibrous failure surface indicate a ductile mode
of fracture, while the small CVN and shiny texture (or cleavage
character) are indicative of the brittle mode of fracture [23]. The
fracture surfaces of the specimens of copper (base metal), copper
(HAZ) and nickel–weld metal appear fibrous with deep and wide
dimples that support their large CVN values. From this, it can be
concluded that these specimens have failed by a fully ductile frac-
ture. The fine grained and fully austenitic structure of the nickel–
weld metal contributed to its high toughness. The steel (base me-
tal) showed the least CVN energy value and a fully brittle fracture.
Whereas the HAZ of steel showed higher impact energy value than
steel (base metal). This is the indication that the HAZ is more duc-
tile than its base metal. The CVN impact energy of steel (base me-
tal) is the least of all other zones in the welded composite and
therefore if it does not fail by brittle fracture, the welded composite
can resist unpredicted mechanical impacts under service condi-
tions. The use of better grade of alloy steel (EN 24/AISI 4340) with
high-toughness, value can increase the overall toughness of the
welded joint.
Fracture is decided by the magnitude of area under entire
stress–strain curve denoted by Modulus of Toughness. The Modulus
of Toughness values of steel (HAZ) and Cu (HAZ) are estimated
approximately from stress–strain curves (Fig. 12h) as 240 MPa
and 58.5 MPa respectively. Therefore, the fracture (Fig. 18a) oc-
curred in Cu (HAZ) due to its lower Modulus of Toughness as com-
pared to that of steel (HAZ) even though Cu (HAZ) had higher
CVN value than that of steel (HAZ).
4. Conclusions
This paper reports the Shielded Metal Arc Welding (SMAW) of
copper (UNSC 11000) and alloy steel (EN 31) using bronze and
nickel-base superalloy filler materials. The major conclusions
drawn from the present work are listed below:
(1) Use of nickel-base superalloy filler material in arc welding of
alloy steel and copper results in a metallurgically sound
weld joint without defects like hot cracking and porosity
at weld-interfaces and in the fusion zone. Whereas large
porosities were observed in the fusion zone of weld made
with the bronze-filler material. Macro and microstructures,
and SEM figures support the observation convincingly.
Therefore, nickel-base filler material is preferred compared
to bronze-filler material to join copper and steel by arc
welding.
(2) Nickel-weld had fully austenitic microstructure. The colum-
nar dendritic solidification structure; microsegregation of
Cu; carbides of type NbC, TiC, MC, M(C, N), M
7
C
3
and
M
23
C
6
in interdendritic regions; and precipitates of type
a
-
chromium,
c
0
-Ni
3
(Ti) and
c
00
-Ni
3
Nb were other features
present. Epitaxial growth towards the copper was clearly
formed.
Cu Steel
(a)
Fracture in Cu (HAZ) Notch
Ni-WM
(b)
1: Cu (BM)
2: Cu (HAZ)
3: Ni- WM
4: Steel (HAZ)
5: Steel (BM)
5432
1
Fig. 18. Impact tested samples: (a) arc welded copper–steel joint with nickel-filler
and (b) fracture surfaces of samples at important zones of the joint.
Table 10
Charpy V-notch impact energy at room temperature.
Material Sample
no.
Impact
energy
(J)
Average
impact
energy (J)
Type of failure
Copper (base
metal)
1 112 Fully ductile
failure but not
fractured
2 116 126
3 149
Copper (HAZ) 1 63 Fully ductile
failure but not
fractured
27678
396
Weld metal 1 65 Fully ductile
fracture
24762
375
Alloy steel (HAZ) 1 16 Fully brittle
fracture
21518
324
Alloy steel (base
metal)
1 10 Fully brittle
fracture
21014
38
Copper–steel
weldment
(Ni-filler)
1 73 Fully ductile
fracture on copper
side
27395
343
808 M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809
(3) Unmixed, partially melted (PMZ) and heat-affected zones
(HAZs) were not formed in copper, whereas PMZ and HAZ
were formed in steel.
(4) Energy Dispersive Spectroscopy (EDS) results confirmed the
tendency of atomic diffusion from Cu and steel sides to the
Ni side are greater than that in the reverse direction. At
the same time, the dilution level of the weld metal by Fe
from the steel was found to be within allowable limit. The
average dilution level of nickel–weld metal by the base met-
als stood at only 22.44% for copper and 23.23% for steel.
(5) XRD studies showed no brittle intermetallic phases in either
of the weld-interfaces, Cu–Ni and Ni–steel.
(6) During transverse tensile testing, the weldments made with
bronze-filler fractured at the weld due to the presence of the
porosity in the weld.
(7) In all the mechanical tests carried out including transverse
tensile test (both at room and high temperatures), trans-
verse bending, side bend and impact tests, the weldments
made with nickel-filler material showed fracture in HAZ of
copper, very near the Cu–WM interface. This confirms the
integrity of the weld-interfaces. The tensile strengths of
the welded joints were comparable to those of the copper
(base metal).
(8) Microhardness values were the least in copper and highest
in the HAZ of steel. The nickel–weld metal displayed higher
hardness values than copper (base metal) and lower values
than HAZ of steel. The values remained almost constant in
copper and nickel–weld metal. The values rose to a maxi-
mum in the HAZ of steel and dropped gradually until they
tend to remain constant in the steel (base metal). The values
in HAZ of steel hint at the formation of bainitic structure
with less martensite.
(9) The flexural strengths of the welded joints made with
nickel-filler material were greater than their tensile
strengths. The side bend tests showed considerable bent
angles and did not reveal weld defects either at weld-inter-
faces or in the welds.
(10) The shear strengths of the weld-interfaces were found to be
greater than yield strengths.
(11) In the Charpy V-notch impact tests, the steel (base metal)
showed the lowest impact energy. Whereas steel (HAZ)
showed higher value than the base metal, a significant result
in support of the welding process that it did not produce a
brittle HAZ. However, it had reduced impact energy of the
copper near the fusion zone compared to its base metal.
Despite this drop, copper in that zone showed a fully ductile
failure. The nickel–weld metal had a significant toughness
and showed a ductile fracture.
(12) Thus, the arc welded dissimilar copper–steel joints with
nickel-filler material can be employed under service condi-
tions involving mechanical and thermal loads.
References
[1] Wei PS, Chung FK. Unsteady marangoni flow in a molten pool when welding
dissimilar metals. Metal Mater Trans B 2000;31:1387–403.
[2] Devendranath Ramkumar K, Arivazhagan N, Narayanan S. Effect of filler
materials on the performance of gas tungsten arc welded AISI 304 and Monel
400. Mater Des 2012;40:70–9.
[3] Zhang CG, van der Vyver S, Hu XZ, Lu PM. Fatigue crack growth behavior in
weld-repaired high-strength low-alloy steel. Eng Fract Mech
2011;78:1862–75.
[4] Durgutlu A, Gulenc B, Findik F. Examination of copper/stainless steel joints
formed by explosive welding. Mater Des 2005;26:497–507.
[5] Kore SD, Date PP, Kulkarni SV, Kumar S, Rani D, Kulkarni MR, et al. Application
of electromagnetic impact technique for welding copper-to-stainless steel
sheets. Int J Adv Manuf Technol 2011;54:949–55.
[6] Leedy KD, Stubbins JF. Copper alloy–stainless steel bonded laminates for fusion
reactor applications: tensile strength and microstructure. Mater Sci Eng A
2001;297:10–8.
[7] Xiong JT, Xie Q, Li JL, Zhang FS, Huang WD. Diffusion bonding of stainless steel
to copper with tin bronze and gold interlayers. J Mater Eng Perform
2012;21:33–7.
[8] Khalid Imran M, Masooda SH, Brandt M, Bhattacharya S, Mazumder J. Direct
metal deposition (DMD) of H13 tool steel on copper alloy substrate: evaluation
of mechanical properties. Mater Sci Eng A 2011;528:3342–9.
[9] Yilmaz O, Celik H. Electrical and thermal properties of the interface at
diffusion-bonded and soldered 304 stainless steel and copper bimetal. J Mater
Process Technol 2003;141:67–76.
[10] Sabetghadam H, Zarei Hanzaki A, Araee A. Diffusion bonding of 410 stainless
steel to copper using a nickel interlayer. Mater Character 2010;61:626–34.
[11] Mai TA, Spowage AC. Characterization of dissimilar joints in laser welding of
steel–kovar, copper–steel and copper–aluminium. Mater Sci Eng A
2004;374:224–33.
[12] Phanikumar G, Manjini S, Dutta P, Mazumder J, Chattopadhyay K.
Characterization of a continuous CO
2
laser-welded Fe–Cu dissimilar couple.
Metall Mater Trans A 2005;36A:2137–47.
[13] Yao C, Xu B, Zhang X, Huang J, Fu J, Wu Y. Interface microstructure and
mechanical properties of laser welding copper–steel dissimilar joint. Opt Las
Eng 2009;47:807–14.
[14] Magnabosco I, Ferro P, Bonollo F, Arnberg L. An investigation of fusion zone
microstructures in electron beam welding of copper–stainless steel. Mater Sci
Eng A 2006;424:163–73.
[15] Bruzek B, Leidich E. Evaluation of crack growth at the weld interface between
bronze and steel. Int J Fatigue 2007;29:1827–31.
[16] Kahraman N, Durgutlu A. Weldability of 316L stainless steel and copper plates
welded by shielded metal arc and tungsten arc welding processes. Technology
2005;8:43–50.
[17] Durgutlu A, Kahraman N, Gulenc B. Joining of copper and steel plates by
shielded metal arc and tig welding methods and investigation of their
interface properties. J Fac Eng Archi Gazi Univ 2005;20:183–9.
[18] Sun Z, Karppi R. The application of electron beam welding for the joining of
dissimilar metals: an overview. J Mater Process Technol 1996;59:257–67.
[19] Arabi Jeshvaghani R, Harati E, Shamanian M. Effects of surface alloying on
microstructure and wear behavior of ductile iron surface-modified with a
nickel-based alloy using shielded metal arc welding. Mater Des
2011;32:1531–6.
[20] Callister Jr W. Materials science and engineering. New Delhi: Wiley India;
2010.
[21] Gale WF, Totemeier TC. Smithells metals reference book. 8th
ed. Oxford: Elsevier Butterworth-Heinemann; 2004.
[22] Dupont JN, Lippold JC, Kiser SD. Welding metallurgy and weldability of nickel-
base alloys. New Jersey: John Wiley & Sons; 2009.
[23] Naffakh H, Shamanian M, Ashrafizadeh F. Dissimilar welding of AISI 310
austenitic stainless steel to nickel-base alloy Inconel 657. J Mater Process
Technol 2009;209:3628–39.
[24] ISO 5177–1981. Fusion welded butt joints in steel-transverse side bend test,
International Organization for Standardization (ISO); 1981.
[25] Kou Sindo. Welding metallurgy. 2nd ed. New Jersey: John Wiley & Sons; 2003.
[26] Zhang J, Shen Q, Luo G, Li M, Zhang L. Microstructure and bonding strength of
diffusion welding of Mo/Cu joints with Ni interlayer. Mater Des 2012;39:81–6.
M. Velu, S. Bhat / Materials and Design 47 (2013) 793–809 809