R. PALANIVEL et al.: TENSILE-SHEAR-STRENGTH PREDICTION OF FRICTION-MELT-BONDED DISSIMILAR SPOT ...
57–62
TENSILE-SHEAR-STRENGTH PREDICTION OF
FRICTION-MELT-BONDED DISSIMILAR SPOT JOINTS USING
RSM
NAPOVED NATEZNO-STRI@NE TRDNOSTI TORNO VARJENEGA
TO^KOVNEGA SPOJA Z UPORABO RSM
Ramaswamy Palanivel
1,*
, Soundararaja Perumal Vigneshwaran
2
, Yousef Alqurashi
1
,
Mohammad Abdur Rasheed
3
1
Department of Mechanical Engineering, College of Engineering, Shaqra University, Dawadmi, Riyadh, Saudi Arabia
2
Department of Mechanical Engineering, SRM Institute of Science and Technology, Ramapuram Campus, Chennai, Tamil Nadu, India
3
Department of Civil Engineering, College of Engineering, Shaqra University, Dawadmi, Riyadh, Saudi Arabia
Prejem rokopisa – received: 2022-11-13; sprejem za objavo – accepted for publication: 2022-12-13
doi:10.17222/mit.2022.689
The joining of aluminium (Al) and copper (Cu) is an essential spot weld for renewable energy applications due to its excellent
thermal and electrical conductivity. It is challenging to produce a high-quality Al-Cu spot weld using traditional techniques and
the durability of a weldment is also uncertain. A new welding technique that is effective and suitable for spot welding Cu and Al
is friction-melt-bonding (FMB) spot welding. The process parameters employed during welding have a strong impact on the
quality of a weld. The strategy for integrating the FMB process characteristics, such as tool rotational speed and dwell time, into
a mathematical model that may be used for predicting the process parameters of dissimilar spot joints, was the focus of this
study. The findings showed that the tensile shear fracture load (TSFL) was significantly influenced by the tool’s rotational speed
and dwell time. Optimization was carried out using the response surface methodology (RSM) to find the best FMB process pa-
rameters for making a dissimilar spot weld of AA 6061-Cu.
Keywords: friction melt bonding, renewable energy, dissimilar joints, optimization
Spajanje aluminija in bakra s to~kovnim varjenjem je pomembno zaradi njune odli~ne toplotne in elektri~ne prevodnosti.
Izdelava visoko kakovostnih spojev Al–Cu s standardnimi postopki varjenja je zahtevno in tudi trajnost tak{nih spojev je
vpra{ljiva. Nova tehnika varjenja, to je torno varjenje z gnetenjem (FMB; angl.: friction melt bonded) je u~inkovito in primerno
za to~kovno varjenje bakra in aluminija. Procesni parametri uporabljeni med varjenjem imajo mo~an vpliv na kakovost
varjenega spoja. V raziskavi so kombinirali s pomo~jo matemati~nega modela procesne parametre postopka FMB, kot sta hitrost
vrtenje orodja ter ~as zadr`evanja in ju uporabili za napoved optimalnih procesnih parametrov to~kovnega varjenja dveh
razli~nih materialov. Rezultati raziskave so pokazale, da je natezno-stri`na poru{na trdnost (TSFL; angl.: tensile shear-fracture
load) mo~no odvisna od hitrosti vrtenja orodja in njegovega ~asa zadr`evanja v stiku obeh kovin. Optimizacijo so izvr{ili z
uporabo metodologije odgovora povr{ine in s tem ugotovili najbolj{i set procesnih parametrov FMB za to~kovno varjenje dveh
razli~nih kovin, to je zlitine na osnovi aluminija vrste AA 6061in bakra.
Klju~ne besede: torno varjenje z gnetenjem, obnovljiva energija, medsebojno spajanje razli~nih materialov, optimizacija
1 INTRODUCTION
A potential alternative large-scale renewable energy
storage technology is an aluminium battery. Electric mo-
tors, batteries, inverters, wiring and charging stations
heavily rely on copper. It should be emphasised that the
primary uses of spot welding copper (Cu) and aluminium
(Al) are in the renewable-energy sector relating to
bimetals, bus bars, switchgears, heat sinks, etc.
1–3
Tech-
nologies of joining Cu and Al, such as resistance spot
welding (RSW), laser-beam welding (LBW) and ultra-
sonic welding (USW) have been developed. However,
dissimilar spot welding is often challenging due to large
changes in physical parameters such as the thermal ex-
pansion coefficient, melting temperature and the
intermetallic-compound (IMC) formations at the weld
interface.
4
Low resistivity and improved electrical con-
ductivity are characteristics of Cu and Al. As a result, the
time required to make Cu-Al joints by RSW is longer,
which could be prohibitively expensive.
5
Furthermore,
because of its inherent fusion welding feature, a lot of
heat is created, resulting in a lot of deformation.
6
To overcome the aforesaid challenges, friction melt
bonding (FMB), a unique approach that makes use of
significant temperature differences between the materials
to be bonded as provides a solution to the challenges that
occur with the fusion welding process. In FMB, the
plates to be joined are stacked one on top of the other
and fastened with clamps. The top surface of the plate is
pressed up against a flat, revolving tool, which results in
a heat production from friction and deformation. The
temperature of the top plate rises to almost the melting
point of the bottom plate because of the frictional heat
produced.
7
The top and bottom plates consequently melt
and react locally, generating a joint. According to Chen
et al.
8
FMB can generate better dissimilar Ti/Al alloy lap
Materiali in tehnologije / Materials and technology 57 (2023) 1, 57–62 57
UDK 678.029.43:662.2.036 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 57(1)57(2023)
*Corresponding author's e-mail:
rpalanivelme@gmail.com (Ramaswamy Palanivel)

joints and can be utilised for spot welding Ti-6Al-4V/
2A12-T4 dissimilar alloys. This is a potential method for
managing the temperature of the Ti/Al contact and
avoiding the development of a thick, brittle Al
3
Ti IMC.
A thermal model was developed for joining steel and
AA2024–T3 to analyse the thermal cycle and inter-
metallic growth, and the results show less intermetallic
growth in the joints.
9
To better understand the controlled
accumulation of IMCs and sensitivity to hot ripping in
Al-steel welding with additional interlayers, Mena et
al.
10
used FMB. Even so, there is a limited amount of lit-
erature on various FMB joints of two plates. There is
currently no published study on the use of FMB for spot
welding Cu and Al. Mechanical characteristics of the
joints are primarily influenced by FMB process parame-
ters such as the tool rotational speed and dwell time. Me-
chanical qualities of joints must be improved in order to
boost the effectiveness of the FMB process. Determining
the best welding parameters is a challenging task re-
quired to attain a quality weld. Therefore, one such
method is the design of experiments (DOE) used for op-
timising the process parameters for the best weld.
11–14
A powerful statistical technique called the central com-
posite design (CCD) saves significant time and costs
while optimising the process, design and product. RSM
is a technique that can be utilised to make a series of ex-
periments, creating a mathematical model, determining
the ideal set of input parameters, and showing the results
as a graph.
11
The CCD can be used to effectively acquire
the optimal process response and assess the qualified
welding process parameters with the help of the response
surface methodology (RSM) and analysis of variance
(ANOVA). When fewer components and levels are cho-
sen, the DOE is best suitable for the case. This situation
can be made possible by carefully selecting the key pro-
cess parameters and their ranges based on preliminary
tests. Amancio-Filho et al.
15
concluded that the rotational
speed and dwell time are significant process parameters
used for determining the strength of the friction stir spot
welding of AA2024–T3 lap joints. According to Plaine
et al.,
14
the dwell time and rotational speed are the two
greatest imperative process variables in friction stir spot
welding when using an RSM model to determine me-
chanical properties. This study makes an effort to fore-
cast and optimise FMB process parameters to achieve the
highest TSFL for the FMB of dissimilar Al- Cu spot
joints.
2 EXPERIMENTAL PART
Dissimilar metals of AA6061–T6 with a composition
(in w/%) of 96.8 % Al, 0.89 % Zn, 0.91 % Si, 0.65 %
Mn, 0.38 % Fe, with few traces of Cu, Ti, Sn, and com-
mercially pure copper were chosen as the investigated
materials. The procured AA6061 and Cu had dimensions
of (100 × 30 × 4) mm and (100 × 30 × 1.5) mm, respec-
tively. The hardness values of AA6061 and Cu are
102 HV and 100 HV, respectively. The melting points of
AA6061 and Cu are 600 °C and 1084 °C, respectively.
The tool is made of tungsten carbide that has a 20-mm
diameter of the shoulder and no pin. The tool material
has a melting point of 2870 °C, and hardness of
2600 HV. The dwell time and rotational speed were con-
sidered as the input parameters for this study based on
the literature. A plunge depth of 0.5 mm and axial force
of 15 kN were kept constant under all the experimental
conditions. Figure 1 shows the experimental set-up and
sample processed with FMB.
Table 1 displays the design matrix and the input pa-
rameter levels that were ultimately chosen with the use
of trial experiments and results. It consists of 13 experi-
mental runs in a 5-centre point and 8-non-centre point
configuration. The experimentation was done in accor-
dance with the design matrix. A computerised tensile
tester (INSTRON 1195) was used to assess the TSFL af-
ter FMB, and the results are shown in Table 1.
3 RESULTS
3.1 Mathematical model
The input parameters relate to the output using a re-
gression model that takes the impacts of the input param-
eters into account. The impacts of the dwell time and ro-
tational speed on the tensile shear fracture load are
presented in Equation (1).
R. PALANIVEL et al.: TENSILE-SHEAR-STRENGTH PREDICTION OF FRICTION-MELT-BONDED DISSIMILAR SPOT ...
58 Materiali in tehnologije / Materials and technology 57 (2023) 1, 57–62
Figure 1: Experimental set-up with: a) tool rotational direction and
fixture design and b) FMB processed sample

TSFL = F(A, B) (1)
Here,
TSFL – the output (tensile shear fracture load)
A – the rotational speed (min
–1
)
B – the dwell time (s)
The preferred polynomial for the two factors is repre-
sented as Equation (2):
Y = b0 + b1A + b2B + b11A
2
+ b22 B
2
+ b12AB (2)
The constant value for the regression model is b0,
first-order coefficients are b1 and b2, second-order coef-
ficients are b11 and b22. The interaction coefficient is
b12.
Regression analysis and Design Expert 13 software
are used to determine the polynomial’s coefficient val-
ues.
Below are the final developed models as a result of
this FMB (Equation (3)):
TSFL = 11.82 + 0.7809A + 0.6867B – 2.62A
2
–
– 2.60B
2
– 0.1703AB (3)
3.2 ANOVA analysis for the quadratic model
The appropriateness of the established model was
analysed with the help of the analysis of variance
(ANOVA) using Design Expert 13. The investigated
ANOVA results are given in Table 2. The F value of the
developed model is significant as it is 380.38.
Lower values of P (less than 0.05) are the indicators
for significant terms. For this developed model signifi-
cant terms are A, B, A
2
,B
2
. Other terms are insignificant
due to a larger value of P (more than 0.1). The F value
for the lack of fit is 0.0, which suggests it is significant
compared to the absolute error. The R
2
, predicted R
2
and
R. PALANIVEL et al.: TENSILE-SHEAR-STRENGTH PREDICTION OF FRICTION-MELT-BONDED DISSIMILAR SPOT ...
Materiali in tehnologije / Materials and technology 57 (2023) 1, 57–62 59
Table 1: Process parameters and their levels, design matrix with the experimental and predicted response
Process parameter
Levels
–1.41 –1 0 1 1.41
(A) Rotational speed
(min
–1
)
1000 1375 1750 2125 2500
(B) Dwell time (s) 50 68 85 103 120
Design matrix
Trial No.
Rotational speed
(min
–1
)
Dwell time level (s)
Tensile shear fracture load (kN)
Error (%) Experimental
values
Predicted values
1 –1 –1 5.01 4.96 1.01
2 1 –1 7.03 6.86 2.48
3 –1 1 6.91 6.67 3.60
4118 7.89 1.39
5 –1.41 0 5.24 5.47 –4.20
6 1.41 0 8.05 7.68 4.82
7 0 –1.41 5.74 5.64 1.77
8 0 1.41 7.68 7.59 1.19
9 0 0 12.01 11.82 1.61
10 0 0 12.05 11.82 1.95
11 0 0 11.75 11.82 –0.59
12 0 0 11.85 11.82 0.25
13 0 0 12.03 11.82 1.78
Table 2: ANOVA analysis
Source Sum of squares
Degrees of free-
dom
Mean square F-value P-value
Model 92.74 5 18.55 380.38 < 0.0001 Significant
A 4.88 1 4.88 100.05 < 0.0001
B 3.77 1 3.77 77.37 < 0.0001
AB 0.116 1 0.116 2.38 0.1669
A
2
47.83 1 47.83 980.83 < 0.0001
B
2
47.1 1 47.1 965.92 < 0.0001
Residual 0.3413 7 0.0488
Lack of Fit 1.53 × 10
–9
3 5.09 × 10
–10
5.97 × 10
–9
1 Not significant
Pure Error 0.3413 4 0.0853
Cor. Total 93.08 12
R
2
Adjusted R
2
Predicted R
2
Adequate precision
0.9963 0.9937 0.9943 45.7412

adjusted R
2
values are more than 0.9, indicating that the
developed models are adequate. A fair amount of agree-
ment is observed between the predicted and adjusted R
2
.
Adequate precision is calculated by measuring the sig-
nal-to-noise ratio. This ratio must be greater than 4 for a
good mathematical model.
13
An adequate precision value
of 45.74 show an adequate signal for the developed
model. Consequently, these models are utilized to estab-
lish the design. The residuals show a random scatter pat-
tern in Figure 2, which suggests that the developed mod-
els are satisfactory and that no independent or constant
variable assumptions have been broken. Figure 3 com-
pares the expected values of scattered responses to the
experimental values, and the fact that it is near the 45-de-
gree line shows that the predicted values and experimen-
tal values concur.
14
The perturbation plots from Figure 4
show how all the parameters have an impact on the ten-
sile shear strength of welded samples. This graph dis-
plays the evolution of the output or reaction. While all
the input parameters deviate from the reference point, the
other values remain constant.
4 DISCUSSION
4.1 Effect of the FMB process parameters
The factors influencing the TSFL of FMB dissimilar
spot joints prophesied with the help of developed models
are given in Figure 5, indicating a common connection
between the cause and result.
As seen in Figure 5a, the lowest rotational speed
(1000 min
–1
) produces a low TSFL of the FMB. The
TSFL increases proportionally with the rotational speed,
starting at 1000 min
–1
, reaching its maximum at
1750 min
–1
. The TSFL of the joint decreases as a result
of the rotational speed continuing to rise above
1750 min
–1
. The bonded zone is reduced at the lower tool
rotational speed due to a low temperature distribution
and low heat input. This low heat input produces a weak
bonding between the top and bottom plates, so the TSFL
is low. It is obvious that the heat input of the FMB in-
creases with the increasing rotational speed. The bonding
between the top and bottom plates is destroyed by the
additional heat input.
14
The response-surface graphs de-
picted in Figure 5a make it clear that the dwell time af-
fects the TSFL and the TSFL of the FMB of the spot
joint is lower with a lower dwell time (50 s). The TSFL
increases along with the dwell time after the first 50 s,
peaking at 85 s. The TSFL of the spot joint is reduced
when the dwell time rises above 85 s. In general, the
FMB at a higher dwell time in a spot-welding area pro-
duces a short exposure duration, inadequate heat and
poor material flow between the plates, making the joint
weaker. A higher dwell time is associated with low heat
R. PALANIVEL et al.: TENSILE-SHEAR-STRENGTH PREDICTION OF FRICTION-MELT-BONDED DISSIMILAR SPOT ...
60 Materiali in tehnologije / Materials and technology 57 (2023) 1, 57–62
Figure 4: Influence of all the parameters on TSFL of FMB (perturba-
tion plot)
Figure 2: Internally studentized residual values for the predicted val-
ues of TSFL of FMB
Figure 3: Predicted values versus experimental values for TSFL of
FMB

inputs, resulting in faster cooling rates of the welded
joint, hence yielding a lower TSFL.
4.2 Optimizing the FMB parameters
The graphs for the response surface and contour were
developed and presented in Figure 5 to determine the
optimum input condition for the best TSFL. These
graphs are used to predict the range of experimental con-
ditions for better outputs.
11
The maximum TSFL is cal-
culated from the vertex of the response plot. To graphi-
cally depict the area of the ideal parameter settings, a
contour plot is developed as shown in Figure 5b.D e -
veloping these plots is more complex for second-order
responses compared to first-order models. Characterizing
the response surface close to the specified position is
typically required once it has been located. The charac-
terization entails determining if the stationary point is a
saddle point, a maximum or minimum response. The
easiest way to do this is to look at it through a contour
plot. The analysis of a response surface relies heavily on
contour plots. The highest possible TSFL value is found
to be 11.86 kN, using the analysis of the response sur-
faces and contour plots (Figure 5). A tool rotational
speed of 1750 min
–1
and dwell time of 85 s are the corre-
sponding parameters that produce this maximum value.
As long as all of the input parameters have the same de-
grees of freedom, it is possible to rank the influences of
the process parameters on the tensile shear failure load
11
based on their respective F-ratio values shown in Ta-
ble 2. The impact is assumed to be more significant if
the F ratio is higher and vice versa. According to the
F-ratio results, the rotational speed has the most impact
on the tensile shear failure load over the range under
consideration in this study, followed by the dwell time.
5 CONCLUSIONS
The following conclusions are derived from this
work:
• The correlations between the process parameters for
the FMB of dissimilar metals including AA 6061 and
Cu were determined using the RSM.
• Response graphs and contour plots were then used to
demonstrate the influences of the rotational speed
and dwell time on the TSFL of the AA 6061 and Cu
joints.
• Both the rotational speed and dwell time exhibited
significant influences on the dissimilar spot joining
of AA6061 and Cu using FMB.
• The optimum ranges of the rotational speed and
dwell time are determined as 1750 min
–1
and 85 s, re-
spectively, for the FMB of AA6061 and Cu.
Acknowledgment
The authors extend their appreciation to the Deanship
of Scientific Research at the Shaqra University, Saudi
Arabia, for funding this research under project number
SU-ANN-202244.
6 REFERENCES
1
J. Y. Yong, V. K. Ramachandaramurthy, K. M. Tan, N. Mithu-
lananthan, A review on the state-of-the-art technologies of electric
vehicle, its impacts and prospects, Renewable Sustainable Energy
Reviews, 49 (2015), 365–385, doi:10.1016/j.rser.2015.04.130
2
M. Palanisamy, V. P. Parikh, M. H. Parekh, V. G. Pol, Lithium metal
battery pouch cell assembly and prototype demonstration using tai-
lored polypropylene separator, Energy Technology, 8 (2020)6 ,
2000094, doi:10.1002/ente.202000094
3
I. Dinaharan, E. T. Akinlabi, Influence of the tool rotational speed on
the microstructure and joint strength of friction-stir spot-welded pure
copper, Mater. Tehnol., 50 (2016) 5, 791–796, doi:10.17222/mit.
2015.301
4
N. J. Mena, P. J. Jacques, L. Dinga, N. Gauquelin, D. Schryvers, H.
Idrissi, F. Delannay, A. Simar, Enhancement of toughness of
Al-to-steel Friction Melt Bonded welds via metallic interlayers, Ma-
terials Science & Engineering A, 740–741 (2019), 274–284,
doi:10.1016/j.msea.2018.10.101
5
M. J. Brand, P. A. Schmidt, M. F. Zaeh, A. Jossen, Welding tech-
niques for battery cells and resulting electrical contact resistances,
Journal of Energy Storage, 1 (2015), 7–14, doi:10.1016/j.est.2015.
04.001
6
L. C. Campanelli, U. F. H. Suhuddin, A. I .S Antonialli, J. F. dos
Santos, N. G. De Alcantara, C. Bolfarini, Metallurgy and mechanical
performance of AZ31 magnesium alloy friction spot welds, Journal
of Material Processing Technology, 213 (2013) 4, 515–521,
doi:10.1016/j.jmatprotec.2012.11.002
R. PALANIVEL et al.: TENSILE-SHEAR-STRENGTH PREDICTION OF FRICTION-MELT-BONDED DISSIMILAR SPOT ...
Materiali in tehnologije / Materials and technology 57 (2023) 1, 57–62 61
Figure 5: TSFL of FMB: a) response graph and b) contour graph

7
T. Sapanathan, N. J. Mena, I. Sabirov, M. A. Monclús, J. M. M.
Aldareguía, P. Xia, L. Zhao, A. Simar, A new physical simulation
tool to predict the interface of dissimilar aluminum to steel welds
performed by friction melt bonding, Journal of Materials Science &
Technology, 35 (2019) 9, 2048–2057, doi:10.1016/j.jmst.2019.
05.004
8
Y. Chen, H. Deng, H. Liu, T. Zhang, S. Li, S. Wang, C. Chen, A
novel strategy for the reliable joining of Ti6Al4V/2A12-T4 dissimi-
lar alloys via friction melt-bonded spot welding, Materials Letters,
253 (2019), 306–309, doi:10.1016/j.matlet.2019.06.089
9
S. Crucifix, C. V. D. Rest, N. J. Mena, P. J. Jacques, A. Simar,
Modelling thermal cycles and intermetallic growth during friction
melt bonding of ULC steel to aluminium alloy 2024-T3, Science and
Technology of Welding and Joining, 20 (2015) 4, 319–324,
doi:10.1179/1362171815Y.0000000020
10
N. J. Mena, A. Simar, P. J. Jacques, On the interplay between
intermetallic controlled growth and hot tearing susceptibility in
Al-to-steel welding with additional interlayers, Materials and De-
sign, 180 (2019), 107958, doi:10.1016/j.matdes.2019.107958
11
R. Karthikeyan, V. Balasubramanian, Predictions of the optimized
friction stir spot welding process parameters for joining AA2024 alu-
minum alloy using RSM, Int. J. Adv. Manuf. Technol., 51 (2010),
173–183, doi:10.1007/s00170-010-2618-2
12
A. Kadirvel, P. Hariharan, Optimization of the die-sinking mi-
cro-EDM process for multiple performance characteristics using the
Taguchi-based grey relational analysis, Mater. Tehnol., 48 (2014)1,
27–32
13
T. Udayakumar, K. Raja, T. M. Afsal Husain, P. Sathiya, Prediction
and optimization of friction welding parameters for super duplex
stainless steel (UNS S32760) joints, Materials and Design, 53
(2014), 226–235, doi:10.1016/j.matdes.2013.07.002
14
A. H. Plaine, A. R. Gonzalez, U. F. H. Suhuddin, J. F. dos Santos, N.
G. Alcântara, The optimization of friction spot welding process pa-
rameters in AA6181-T4 and Ti6Al4V dissimilar joints, Materials &
Design, 83 (2015), 36–41, doi:10.1016/j.matdes.2015.05.082
15
S. T. Amancio-Filho, A. P. C. Camillo, L. Bergmann, J. F. dos
Santos, S. E. Kuri, N. G. Alcântara, Preliminary investigation of the
microstructure and mechanical behaviour of 2024 aluminium alloy
friction spot welds, Mater. Trans., 52 (2011) 5, 985–991,
doi:10.2320/matertrans.L-MZ201126
R. PALANIVEL et al.: TENSILE-SHEAR-STRENGTH PREDICTION OF FRICTION-MELT-BONDED DISSIMILAR SPOT ...
62 Materiali in tehnologije / Materials and technology 57 (2023) 1, 57–62