See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332850878

Development of an empirical relationship to predict the joint tensile strength

and joint shear strength of diffusion bonded AA6082 aluminium alloy

Article in International Journal of Materials and Product Technology · January 2019

DOI: 10.1504/IJMPT.2019.10021000

CITATIONS READS

0 24

2 authors:

s. Venugopal G. Mahendran
GIET University, GUNUPUR,odhisha RMK Engineering College
11 PUBLICATIONS 32 CITATIONS 16 PUBLICATIONS 355 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by s. Venugopal on 08 September 2022.

The user has requested enhancement of the downloaded file.

Int. J. Materials and Product Technology, Vol. 58, No. 4, 2019 257

Development of an empirical relationship to predict

the joint tensile strength and joint shear strength of
diffusion bonded AA6082 aluminium alloy

S. Venugopal*
Department of Mechanical Engineering,
Sathyabhama University,
Chennai-600119, Tamil Nadu, India
Email: [email protected]
*Corresponding author

G. Mahendran
Department of Mechanical Engineering,
IFET College of Engineering,
Vilupuram-605108, Tamil Nadu, India
Email: [email protected]

Abstract: In this work, empirical relationships were developed to predict the

joint tensile strength and joint shear strength of the rolled plates made of 5 mm
thick aluminium alloy AA6082, joined by diffusion bonding. Three diffusion
bonding parameters such as bonding temperature, bonding pressure and holding
time were used to formulate the relationship. A central composite design was
also put to use to minimise the number of experimental conditions. In addition,
the response surface methodology was employed to develop the relationship.
The relationship thus developed can effectively predict, to the 95% confidence
level, the strength of diffusion bonded AA6082 aluminium alloys. It was
observed that the joint tensile strength and joint shear strength went up with the
increase in either bonding temperature or bonding pressure. The maximum
joint tensile strength and joint shear strength were observed as 36 MPa and
115 MPa at 520°C and 10 MPa pressure. It was evidently due to the diffusion
of atoms in the interface. Metallurgical characterisation was performed using
optical microscopy and SEM.

Keywords: AA6082 aluminium alloy; diffusion bonding; response surface

methodology; optical microscopy; scanning electron microscopy; SEM;
bonding temperature; bonding pressure; holding time; joint tensile strength;
joint shear strength.

Reference to this paper should be made as follows: Venugopal, S. and

Mahendran, G. (2019) ‘Development of an empirical relationship to predict the
joint tensile strength and joint shear strength of diffusion bonded AA6082
aluminium alloy’, Int. J. Materials and Product Technology, Vol. 58, No. 4,
pp.257–274.

Biographical notes: S. Venugopal completed his Diploma in Agricultural

Engineering in Muthiah Polytechnic College, Chidambaram in 1977. He
obtained his Bachelor’s in Mechanical Engineering from the Annamalai
University, Chidambaram, India in 1987. He obtained his Master’s in
Production Engineering from the Annamalai University, Chidambaram, India

Copyright © 2019 Inderscience Enterprises Ltd.

258 S. Venugopal and G. Mahendran

in 1997. He is pursuing his research in Sathyabama University,

Chennai-600119, India. He has published four international journals and four
national journals. He has 27 years industrial experience and 12 years teaching
experience. He is a life member of Institution of Engineers India (IEI), Indian
Welding Society (IWS), Indian Society for Technical Education (ISTE) and
Indian Institute of Metals (IIM).

G. Mahendran completed his PhD in Annamalai University, Chidambaram in

2009, MTech in Advance Manufacturing from the Indian Institute of
Technology, Chennai in 2000, and BE in Mechanical and Production from the
Annamalai University, Chidambaram in 1995. Having 22 years of experience,
he worked from 1996 to 2010 as a Lecturer to Professor in VRS College of
Engineering and Technology and joined in IFET College of Engineering in
December 2011 as the Dean and rose to the position of Principal in October
2013. He has to his credit several publications/presentations in the national and
international journals/symposia. He is a recognised supervisor for guiding PhD
under Anna University and Sathyabama University in the area of joining of
materials.

1 Introduction

Aluminium and its alloys are widely used as structural materials in automotive and
aerospace industry because of their high specific strength and excellent corrosion
resistance quality (Eslami and Taheri, 2011). Due to melting and solidification in the
fusion welding process, the joining of aluminium alloys leads to formation of oxide films
and brittle intermetallic in the weld region (Gronostajski et al., 1994). Solid state welding
process avoids such problems and enhances the joining strength.
Diffusion bonding offers both similar and dissimilar materials less dimensional
tolerance and similar microstructure as the base metal. It is observed from the literature
surveys (Kundu and Chatterjee, 2006; Ghosh et al., 2003) that the joint tensile strength
(JTS) and the joint stiffness increase with increasing bonding temperature and pressure.
The usage of reasonable pressure will cause highly deformed contacting asperities,
leading to the development of planar interfacial voids during diffusion bonding. Due to
severe plastic deformations, the atoms move to the void regions from the adjacent regions
and this atom movement reduces the interfacial void volume. With increase in time
duration, the voids will disappear and interface will become highly dense due to atomic
bonding. Therefore, the microstructure at the bond region looks comparable with the base
metal, with similar properties (Islam and Ridley, 1998).
The microstructure study reveals that increase in the width of the brittle
micro-constituents causes higher mechanical strength. The relocation of grain boundaries
at the interface and the grain growth cause higher JTS (Kundu et al., 2005; Huang et al.,
1999). The variation in the catalysts, micro-constituents and intermetallic phase particles
play a key role in obtaining stronger bonded joints (Shirzadi et al., 2002; Ghosh and
Chatterjee, 2003).
Recent researches in the field of welding (Kenevisi and Mousavi Khoie, 2012; Yeh
and Chuang, 1995) have proved that the diffusion bonding technique is a better process
for joining dissimilar metals. It is further found that the choice of the process parameters
affect the interface grain distribution in compound formation. From the previous
 Development of an empirical relationship 259

researches (Kenevisi and Mousavi Khoie, 2012; Yeh and Chuang, 1995; Feng et al.,
2002), it is observed that the most influencing process parameters are bonding
temperature, pressure and holding time. However, no combined effects of the above
mentioned process parameters are addressed. Also, most of the available literatures (Liu
et al., 2006; Li et al., 2008; Liu et al., 2005; Huang et al., 1988; Juan et al., 2008) focus
on diffusion bonding of aluminium materials with respect to its microstructure analysis,
phase formation studies, hardness surveys at the interface, and their subsequent influence
on the JTS. Hence, the present investigation is carried out to develop an empirical
relationship to predict the tensile bonding strength and bonding shear strength of
diffusion bonded joints of AA6082 aluminium alloys, incorporating the process
parameters like bonding temperature, bonding pressure and holding time.

2 Experimental work

2.1 Fabricating the joints and preparing the specimens

Specimens were prepared from the rolled plates of AA6082 aluminium alloy material
with 5 mm thickness. The chemical composition and mechanical properties of the
specimens were presented in Tables 1 and 2 respectively. An optical micrograph structure
of AA6082 aluminium alloy was shown in Figure 1 which displayed the microstructure
of the base metal (AA6082 aluminium alloy). In addition, the structure showed elongated
grains distributed evenly in the base material with particular orientation. The specimens
of the size 50 mm × 50 mm were cut from the plate by power hacksaw, followed by
milling process. Prior to diffusion bonding (ASTM Standard, 1999; Vander Voort, 2004),
the bonding surfaces of samples were ground flat by 200, 400 and 600 grit SiC papers
and cleaned in acetone. The polished and chemically treated specimens were covered
both at the bottom and at the top with mica sheets (RUBY) in a die made up of H13 tool
steel. All the specimens were placed in a vacuum chamber with a vacuum pressure of
3 Pa. The specimens were then heated using an induction furnace with a heating range of
25ºC/min and the required bonding pressure was applied for the duration of the holding
time. After the completion of the diffusion bonding process, all the samples were cooled
to room temperature before taking out of the chamber. The configuration of the diffusion
bonding setup is shown in Figure 2. The parameters used for the diffusion bonding were
bonding temperature, bonding pressure and holding time.
Table 1 Chemical composition (wt. %) of AA6082 aluminium alloy (as-received)

Si Fe Cu Mn Mg Zn Ti Others Al
0.50 0.70 0.25 0.15 1.1 0.63 0.15 0.04 Balance

Table 2 Physical and mechanical properties of AA6082 aluminium alloy at T6 condition

(as-received)

Yield Shear
Density Melting Ultimate tensile Elongation Poisson’s
strength strength
(g/cm3) point (°C) strength (MPa) (%) ratio
(MPa) (MPa)
2.7 665 305 271 188 12 0.3
260 S. Venugopal and G. Mahendran

Figure 1 Typical microstructure of AA6082 aluminium alloy as-received in as-rolled condition

Figure 2 Diffusion bonding setup (see online version for colours)

2.2 Recording the responses

After the bonding process, the samples were prepared from the Al/Al diffusion bonded
joints using a wire-cut EDM and the tensile test was carried out in 50 kN capacity servo
controlled universal testing machine. Mechanical properties such as JTS and joint
shear strength (JSS) were determined using UTM in accordance with ASTM
Standard D1002-99 (Mahendran et al., 2009). To examine the diffusion layer formation
at the interface, micro structural characterisation was carried out using a light
optical microscope (VERSAMET-3), incorporated with image analysis software
(Clemex-vision). The specimens were polished in a disc polishing machine. The diffusion
 Development of an empirical relationship 261

bonded cross section was etched with a Keller’s solution (3ml HCl, 2 ml HF and 90 ml
distilled water) (Zhao and Zhang, 2008) and the surface was observed at 200X
magnification. The microstructure features of diffusion bonded joints were also
characterised by scanning electron microscopy (SEM). The element distribution at the
joint areas was analysed by an in-built electron probe micro analyser of type
EPMA-1600.

3 Developing the experimental design matrix

3.1 Finding the limits of diffusion bonding parameters

From the literature (Ureña et al., 1992; Dheenadayalan et al., 2015; Kurgan, 2014), the
predominant factors that have the greatest influence on the diffusion bonding of AA6082
aluminium alloy joints have been identified. They are:
1 bonding temperature (T)

2 bonding pressure (P)

3 holding time (t).
A large numbers of trial experiments were conducted to identify the feasible testing
conditions using the diffusion bonding conditions. The following observations were
made:

1 If the bonding temperature is lower than 490°C, no bonding will take place between
the specimens. Diffusion of atoms will not happen under insufficient temperature
[Figure 3(a)].
2 If the bonding temperature is greater than 520°C, the bonding pressure will decrease
linearly after a few minutes. This is due to the melting of surface particles of the
specimens [Figure 3(b)].
3 If the bonding pressure is lower than 5 MPa, no bonding will occur. This is due to
the less number of contacting points (present between the surface asperities) through
which diffusion of atoms should generally occur [Figure 3(c)].

4 If the bonding pressure is greater than 15 MPa, no bonding will occur. At such a
pressure, the plates are deformed plastically, causing a reduction in thickness and
bulging at the outer edges (Figure 3(d)].
5 If the holding time is less than 15 min, no bonding will occur. This is because
sufficient time was not given for the diffusion reaction to take place [Figure 3(e)].
6 If the holding time is higher than 45 min, then excessive grain growth will occur
followed by melting of inter-phase particles as shown in Figure 3(f).
262 S. Venugopal and G. Mahendran

Figure 3 Bonds fabricated using upper and lower limits of process parameters,
(a) temperature < 490°C (b) temperature > 490°C (c) pressure < 5 MPa
(d) pressure > 15 MPa (e) holding time < 15 min (f) holding time > 15 min
(see online version for colours)

(a) (b)

(c) (d)

(e) (f)
 Development of an empirical relationship 263

3.2 Matrix development

Out of the extensive range of factors available, only three factors and a central composite
rotatable design matrix were chosen for the current investigation so as to minimise the
number of experiments. The design matrix consisting of 20 sets of coded setting
(comprising a full replication three factorial of eight points, six corner points and six
centre points) was chosen for this investigation. Table 3 represents the range of factors
considered, and Table 4(a) summarises the 20 sets of coded and actual values used in the
experiments. The results were obtained in triplicate and the average responses were
tabulated. Table 4(b) shows the correlation between the experimental and empirical
results. It is observed that the variance and error between the experimental and empirical
results were very minimal and within the limits of errors. For the ease of observing and
processing the experimental data, the upper and lower levels of the factors were coded
here as +1.682 and –1.682. The coded value of each intermediate value could be
calculated using the following relationship (Fernandus, 2012).

Xi = 1.682 [ 2X − ( X max − X min )] ( X max − X min )

where X is the required coded value of a variable X,

X is any value varies from Xmin to Xmax
Xmin is the lower level of the variable
Xmax is the upper level of the variable.
Table 3 Important factors and their levels

Levels
S. no. Factor Unit Notation
–1.682 –1 0 1 1.682
1 Bonding temperature °C T 490 496 505 514 520
2 Bonding pressure MPa P 5 7 10 13 15
3 Holding time min. t 15 21 30 39 45

Table 4(a) Design matrix and experimental results

Bonding Bonding Bonding Shear

Experiment Holding time
temperature pressure P strength BS strength SS
no. T (min.)
T (°C) (MPa) (MPa) (MPa)
1 496 7 21 24 102
2 514 7 21 33 100
3 496 13 21 26 107
4 514 13 21 30 110
5 496 7 39 23 106
6 514 7 39 31 110
7 496 13 39 29 104
8 514 13 39 34 109
9 490 10 30 24 102
10 520 10 30 36 115
264 S. Venugopal and G. Mahendran

Table 4(a) Design matrix and experimental results (continued)

Bonding Bonding Bonding

Experiment Holding time Shear strength
temperature pressure P strength BS
no. T (min.) SS (MPa)
T (°C) (MPa) (MPa)
11 505 5 30 26 100
12 505 15 30 28 105
13 505 10 15 25 104
14 505 10 45 29 107
15 505 10 30 27 103
16 505 10 30 26 103
17 505 10 30 25 103
18 505 10 30 25 103
19 505 10 30 25 103
20 505 10 30 25 103

Table 4(b) Comparison of empirical model and experimental results

Bonding strength Difference Shear strength Difference

Exp. no.
Experimental Empirical % Experimental Empirical %
1 24 23.45 –2.292 102 101.98 –0.01961
2 33 31.35 –5.000 100 101.78 1.78
3 26 25.03 –3.731 107 107.58 0.5421
4 30 29.57 –1.433 110 110.58 0.5272
5 23 21.93 –4.6522 106 106.3 0.2830
6 31 30.79 –0.6774 110 110.22 0.2
7 29 29.27 0.931 104 103.5 –0.4808
8 34 34.77 2.2647 109 110.62 1.4862
9 24 24.694 2.887 102 108.221 6.098
10 36 35.676 –0.9 115 114.0405 –0.834
11 26 24.0868 –7.3585 100 100.205 0.205
12 28 28.763 2.7242 105 105.2511 0.2391
13 25 25.9807 3.922 104 103.5258 –0.4559
14 29 25.18 0.2607 107 107.1925 0.17998
15 27 25.18 –6.7408 103 103.69 0.6699
16 26 25.18 –3.154 103 103.69 0.6699
17 25 25.18 0.72 103 103.69 0.6699
18 25 25.18 0.72 103 103.69 0.6699
19 25 25.18 0.72 103 103.69 0.6699
20 25 25.18 0.72 103 103.69 0.6699

In the present investigation, to correlate the diffusion bonding parameters with the JTS
and JSS of the joints, a second order quadratic model was developed. The response (JTS
 Development of an empirical relationship 265

and JSS) is a function of bonding temperature (T), bonding pressure (P) and holding time
(t) and expressed as,
Joint Tensile Strength and Joint Shear Strength (JTS and JSS) = f (T, P, t)

The empirical relationship must include the major and interactive effects of all the factors
and the selected polynomial becomes:

Y = bo +  bi xi +  bii xi +  bij xi xj
2 3

For three factors, the selected polynomial is expressed as

JTS/JSS = b0 + b1 (T) + b 2 (P) + b3 (t) + b11 ( T 2 ) + b 22 ( P 2 )
+ b33 ( t 2 ) + b12 (TP) + b13 (Tt) + b 23 (Pt)

where b0 is the average of response (of BS and SS) and b1, b2, b3, …, b11, b12, b13, …, b22,
b23, b33, are the coefficient that depends on the respective major and interactive factors,
which are calculated using the mathematical expression as:

Bi =  (Xi, Yi) n
where ‘i’ ranges from 1 to n, in which Xi is the equivalent coded value of a factor and Yi
is the equivalent response output value (BS and SS) obtained from the experiment and ‘n’
is the total number of combination measured. All the coefficient values are calculated
from central composite rotatable design matrix using DX8 design expert statistical
software package.

4 Results and discussion

4.1 Developing empirical relationship

Significant coefficient was determined keeping 95% confidence level and using the
procedure adopted by Balasubramanian et al. (2013).
The calculated value of JTS and JSS was given below:
Joint tensile strength (JTS) = 25.18 + 3.35(T) + 1.39(P) + 0.92(t) − 0.84(TP)
+0.24(Tt) + 1.44(Pt) + 1.82(T2) + 0.44(P2) + 0.83(t2)

Joint shear strength (JSS) = 103.69 + 1.73(T) + 1.50(P) + 1.09(t) + 0.80(TP)

+1.03(Tt) − 2.10(Pt) + 2.63(T2) − 0.34(P2)
+0.59(t2)

The final empirical relationship was obtained using only the calculated coefficient (@
95% confidence level) for the JTS and JSS of diffusion bonded AA6082 aluminium
alloy.
The analysis of variance (ANOVA) technique was used to find the important main
and interactive factors. The null hypothesis for an ANOVA is that there is no significant
difference within the groups. The alternative hypothesis assumes that there is at least one
significant difference in the groups. After cleaning the data, the researcher must test the
266 S. Venugopal and G. Mahendran

assumptions of ANOVA. They must then calculate the F-ratio and the associated
probability value (p-value). In general, if the p-value associated with the F is smaller than
0.05, then the null hypothesis is rejected and the alternative hypothesis is supported. If
the null hypothesis is rejected, it can be concluded that the means of all the groups are not
equal. Post-hoc tests tell the researcher which groups are different from the others. When
conducting an ANOVA, an attempt must be made to determine whether there is a
statistically significant difference within the groups. If differences are observed, then it
will need further examination to find out where the group differences exist.
Table 5 ANOVA test results

Sum of squares df Mean square F-value p-value prob > F

Source
BS SS BS/SS BS SS BS SS BS SS
Model 267.73 243.70 9 29.75 27.08 27.86 25.03 <0.0001 <0.0001
T 152.89 41.08 1 152.89 41.08 143.18 37.97 <0.0001 0.0001
P 26.47 30.59 1 26.47 30.59 24.79 28.28 0.0006 0.0003
t 11.60 16.26 1 11.60 16.26 10.86 15.03 0.0081 0.0031
TP 5.61 5.12 1 5.61 5.12 5.25 4.73 0.0448 0.0547
Tt 0.45 8.41 1 0.45 8.41 0.42 7.77 0.5303 0.0192
Pt 16.53 35.28 1 16.53 35.28 15.48 32.61 0.0028 0.0002
2
T 47.52 99.58 1 47.52 99.58 44.51 92.04 <0.0001 <0.0001
P2 2.75 1.68 1 2.75 1.68 2.58 1.55 0.1394 0.2414
2
t 9.83 4.93 1 9.83 4.93 9.21 4.56 0.0126 0.0585
Residual 10.68 10.82 10 1.07 1.08
Lack of 7.18 7.49 5 1.44 1.50 2.05 2.25 0.2246 0.1977
fit
Pure 3.50 3.33 5 0.70 0.67
error
Cor 278.41 254.52 19
total

At this point, post-hoc tests could run. These are t-tests examining the mean differences
between the groups. Multiple comparison tests can be conducted to control the type I
error rate. The second order response surface model was developed and the curve fitting
was done using ANOVA. The findings are tabulated in Table 5. The coefficient (r2) was
found to be a good fit for the model. The model F-values of the responses were 27.86 and
25.03 and this implied that the model was significant. The ‘Prob > F’ obtained was less
than 0.05 and it indicated that the model factors were significant. In this case, the JTS
included T, P, t, TP, Pt, T2 and t2 which were the significant model terms; while JSS
included T, P, t, Tt, Pt and T2. ‘Prob > F’ was greater than 0.1, indicating that the model
terms were insignificant. The ‘lack of fit F-value’ of the responses were 2.05 and 2.25
respectively, implying that the lack of fit was not a significant relative to the pure error.
There can be a 20%–23% chance that a ‘lack of fit F-value’ of joint tensile and JSS could
occur due to noise. Each of the observed values was compared with the respective
experimental values. The predicted R-squared of joint tensile and JSS were 0.7819 and
0.7490 which were in reasonable agreement with the adjusted R-squared of 0.9271 and
 Development of an empirical relationship 267

0.9192, i.e., the differences were less than 0.2. Adequacy precision measures the signal-
to-noise ratio. A p-ratio of greater than 4 is desirable. The ratio for the responses is 18
and 9 indicating an adequate signal. The normal probability of the joint tensile and JSS
plotted in Figure 4 revealed that the residuals were falling on a straight line, which
indicated taht the errors were distributed normally. All the observed or calculated results
indicated an excellent suitability of the regression model.

Figure 4 Normal probability plots (a) JST (b) JSS (see online version for colours)

(a) (b)

4.2 Effect of bonding temperature

Figure 5 shows the effect of temperature on JTS and the JSS at a constant bonding
pressure of 10 MPa and a holding time of 30 min respectively. It was observed that with
increase in temperature, the JTS and JSS went up. This was attributed to the grain
growth. It was also observed that increasing the temperature from 490°C to 505°C caused
only a small increase in strength, i.e., 1% increase was obtained. However, when the
bonding temperature was steadily raised from 505°C to 520°C, i.e., 10% increase in
strength was observed. The maximum bonding and shear strength obtained were 36 and
115 MPa respectively at 520°C. Further increase in temperature (over and above 520°C)
decreased the JTS and JSS. Hence, it was determined that the diffusion coefficient of
atoms was associated with temperature (Abbasi et al., 2001). At a temperature greater
than 520°C, the interface of diffusion bonded AA6082 aluminium alloy faced the risk of
nucleation and growth of non-metallic intermetallic compounds (Mg2Si). Since AA6082
was a heat treatable wrought aluminium alloy with eutectic, the strengthening elements of
Mg and Si combined to form Mg2Si eutectic, which was responsible for solutionising and
precipitation in the primary aluminium matrix. The rate of dissolution and precipitation
would depend on the temperature and rate of cooling (ASM, 1999). The presence of
Mg2Si was indicated by X-ray diffraction and EDS. Therefore, the brittleness could
weaken the JTS (He and Liu, 2006; Sunwoo and Lum, 1995). It was observed that at a
temperature higher than 520°C and at holding time of 39 min, the extension of these
intermetallic compounds led to the degradation of joint strength. It was apparent that at
bonding temperature of 520°C, in conjunction with the holding time of 30 min, the
highest JSS value of 115 MPa with a bonding pressure 10 MPa was obtained.
268 S. Venugopal and G. Mahendran

Figure 5 Effect of bonding temperature on JTS and JSS (see online version for colours)

Figure 6 Effect of bonding temperature on microstructure (bonding pressure 10 MPa and holding
time of 30 min), (a) 490°C (b) 520°C (see online version for colours)

50 µm 50 µm

(a) (b)

Figures 6(a) and 6(b) show the microstructure of diffusion bonded AA6082 at 490°C and
520°C respectively. At 490°C, the diffusion bonding line was marginally visible with
larger grains along the diffusion zone; whereas, at 520°C, the bonding line was
indistinguishable within the two adjacent metals. The original bond line was located
vertically in the centre of all the micrographs and the centre bond line was distinct at low
temperatures but partially vanished at elevated temperature, irrespective of the amount of
pressure and duration of the holding time. There was a clear evidence of local migration
and grain growth at the bonding line. Figure 6(a) shows the grain size ranging from 25 to
30 microns at 490°C; and Figure 6(b) shows very large grain sizes varying from 45 to 50
microns at 520°C. This grain growth was in closely comparable to the results of other
research works (Wilkinson, 1988; Hu et al., 1995). The absence or vanishing of bond line
 Development of an empirical relationship 269

suggested that better bonding had occurred. Absence of interfacial voids and close
proximity contact were essential for DB process. This phenomenon had occurred at high
bonding temperatures.

4.3 Effect of bonding pressure

The effect of bonding pressure on the JTS and JSS with a bonding temperature of 505°C
and a holding time 30 min are shown in Figure 7. It was observed that the JTS and JSS
increased with increase in pressure. At a pressure of 15 MPa, higher joint tensile and
shear strength of 28 and 105 MPa were obtained, and, at a low bonding pressure of
5 MPa, the strengths obtained were 24 and 100 MPa respectively.

Figure 7 Effect of bonding pressure on JTS and JSS (see online version for colours)

Figure 8 Effect of bonding pressure on microstructure, (bonding temperature of 505°C and

holding time of 30 min), (a) 5 MPa (b) 15 MPa (see online version for colours)

50 µm 50 µm

(a) (b)
270 S. Venugopal and G. Mahendran

The effect of bonding pressure on strength was less when compared with other
parameters such as temperature and time. The increase in bonding pressure decreased the
holding time required to get good bonds, without considering the bonding temperature.
The influence of pressure with respect to temperature decreased the re-crystallisation
temperature and increased the bonding temperature at the grain boundary and stimulated
the growth of the grains (Hu et al., 1995; Smith, 1990).
Figure 8 shows the effect of bonding pressure on the microstructures of the bonded
joints. It was observed that with the increase in bonding pressure, the change in grain
structure was marginal but the bond line was in close proximity to the two metals.
Figure 8 shows the discontinuous bonding lines at 15 MPa. Equated and homogeneous
grains were seen along the bond line (diffusion line) at 15 MPa pressure. At high bonding
pressures the point of contact expanded with plastic deformation (Askeland, 1994). This
increase in the interface contact increased the rate of flow of atoms through the interface.
Hence, the movement of atoms had created more diffusion paths (He et al., 2002;
Mahendran et al., 2010). It was reported (Pilling and Rediely, 1987) that the actual JSS
requirements of the bonds for aircraft structures were generally in the order of
10–20 MPa.

Figure 9 Effect of holding time on bonding and shear strength (see online version for colours)

4.4 Effect of holding time

Figure 9 shows the effect of holding time on the JTS and JSS of AA6082 aluminium
alloy with bonding temperature of 505°C and a bonding pressure of 10 MPa. The results
showed that when the holding time increased from 30 to 45 min, the JSS increased as did
the JTS. Higher the holding time, faster the atom movement. Hence, diffusion bonding
was found to be a better process. The effect of holding time on the microstructure at a
fixed temperature and pressure is shown in Figure 10. The diffusion zone was hindered
 Development of an empirical relationship 271

by the grain growth. Figure 10(a) and 10 (b) indicate the variation of grain sizes at 15
minutes and 45 minutes respectively. Figure 10(a) shows the grain sizes in the range of
15to 20 microns and Figure 10(b) shows the grain sizes in the range of 30 to 35 microns,
at the holding time of 45 minutes.

Figure 10 Effect of microstructure on holding time (bonding temperature of 505°C and bonding
pressure of 10 MPa), (a) 15 min (b) 45 min (see online version for colours)

50 µm 50 µm

(a) (b)

Figure 11 SEM images of diffusion bonded AA6082 aluminium alloy joints, (a) bonding
temperature of 490°C (b) 520°C with bonding pressure of 10 MPa and holding time of
30 min (see online version for colours)

2 µm
2 µm
Diffusion bonding layer Diffusion bonding layer

(a) (b)

This effect was similar to that of the previous research (Wert et al., 1981).

4.5 Scanning electron microscopy

Figure 11 shows the SEM images of diffusion bonded AA6082 aluminium alloy joints
with bonding temperature of 490°C and 520°C, bonding pressure of 10 MPa and a
holding time of 30 min. The present work has identified a range of conditions under
which sound diffusion bonds obtained and these are shown in Figure 6(b). However, it
was recognised that high temperatures led to a significant reduction in the deformation
potential of the material because of grain growth. It was also observed from the SEM
272 S. Venugopal and G. Mahendran

images that the diffusion bonding layer was uniform in width and the grain sizes were in
excess of 12–14 mm. With the increase in temperature, the width of the diffusion bond
layer decreased. At a temperature of 490°C, the width was measured as 15 to 18 microns.
Similarly, the width of the diffusion bonding layer was observed at 5 to 7 microns with
the temperature exposure to the level of 520°C (Mahendran et al., 2011). It was also
observed that both sides of the parent metal showed an identical microstructure with
constituents of eutectics in the primary aluminium solid solution. The interface zone
showed finer constituents of the eutectics which had dissolved and precipitated. The
original grain orientation of the material completely vanished at the diffusion zone.

5 Conclusions

• Empirical relationships are developed to predict the JTS and JSS of the diffusion
bonded joints of AA6082 aluminium alloys by incorporating important parameters.
The developed relationship can be effectively used to predict the JTS and JSS of the
diffusion bonds with 95% confidence level.
• The JTS and JSS increase with the increase in temperature, pressure and holding
time. The bonding temperature is found to have a greater influence on the JSS and
JTS of the joints followed by bonding pressure and holding time.
• A maximum JTS of 36 MPa and JSS of 115 MPa are obtained under the conditions
of bonding temperature of 520°C, bonding pressure of 10 MPa and holding time of
30 min.

References
Abbasi, M., Taheri, A.K. and Salehi, M.T. (2001) ‘Growth rate of intermetallic compounds in
Al/Cu bimetal produced by cold roll welding process’, J. Alloy Compd., Vol. 319,
pp.233–241.
Askeland, R.D. (1994) The Science and Engineering of Material, 3rd ed., PWS Publishing
Company, Boston.
ASM (1999) Metals Hand Book, Vol. 7.
ASTM Standard (1999) D1002 Standard Test Method for Apparent Shear Strength of
Single – Lap-Joint Adhesively Bonded Metal Specimens by Tension Loading (Metal to Metal),
ASM International.
Balasubramanian, V., Fernandus, M.J. and Senthilkumar, T.T. (2013) ‘Development of processing
windows for diffusion bonding of aluminium/magnesium dissimilar materials’, Weld World,
Vol. 57, pp.523–539, DOI: 10.1007/s40194-013-0048-0.
Dheenadayalan, K, Rajakumar, S. and Balasubramanian, V. (2015) ‘Effect of diffusion bonding.
aluminium dissimilar joints’, Applied Mechanics and Materials, Vol. 787, pp.495–499.
Eslami, P. and Taheri, A.K. (2011) ‘An investigation on diffusion bonding of aluminum to copper
using equal channel angular extrusion process’, Mater Lett., Vol. 65, No. 12, pp.1862–1864.
Feng, J.C., Zhang, B.G., Qian, Y.Y. and He, P. (2002) ‘Micro Structure and strength of diffusion-
bonded joints of Ti Al base alloy to steel’, Material Characterization, Vol. 48, pp.401–406.
Fernandus, J., Senthilkumar, M., Balasubramanian, T. and Rajakumar, V. (2012) ‘Optimising
diffusion bonding parameters to maximize the strength of AA6061 aluminium and AZ31B
magnesium alloy joints’, Materials and Design, Vol. 33, pp.31–41.
 Development of an empirical relationship 273

Ghosh, M. and Chatterjee, S. (2003) ‘Diffusion bonded transition joints of titanium to stainless
steel with improved properties’, Materials Science and Engineering A, Vol. 358, pp.152–158.
Ghosh, M., Bhanumurthy, K., Kale, G.B., Krishnan, J. and Chatterjee, S. (2003) ‘Diffusion
bonding of titanium to 304 stainless steel’, Journal of Nuclear Materials, Vol. 322, Nos. 2–3,
pp.235–241.
Gronostajski, J., Kaczmar, J., Marciniak, H. and Matuszak, A. (1994) ‘Recycling of aluminium
alloys by diffusion bonding process’, Proc. of Int. Conf. the Recycling of Metals, Amsterdam,
pp.287–294.
He, P. and Liu, D. (2006) ‘Mechanism of forming interfacial intermetallic compounds at interface
for solid state diffusion bonding of dissimilar materials’, Mater. Sci. Eng. A, Vol. 437,
pp.430–435.
He, P., Feng, J.C., Zhang, B.G. and Qian, Y.Y. (2002) ‘Microstructure and strength of diffusion
bonded joints of Ti Al base alloy to steel’, Materials Characterization, Vol. 48, pp.401–406,
doi:10.1016/S1044-5803(02) 00319-4.
Hu, W., Ponge, D. and Gottstein, G. (1995) ‘Origin of grain boundary motion duringdiffusion
bonding by hot pressing’, Mater. Sci. Eng. A, Vol. 190, Nos. 1–2, pp.223–229.
Huang, Y., Humphreys, F.J., Ridley, N. and Wang, Z.C. (1988) ‘Diffusion bonding of hot rolled
7075 aluminium alloy’, Mater. Sci. and Tech., Vol. 14, pp.405–410.
Huang, Y., Ridley, N. and Humphreys, F.J. (1999) ‘Diffusion bounding of superplastic 7075
aluminium alloy’, Materials Science and Engineering A, Vol. 266, pp.295–302.
Islam, M.F. and Ridley, N. (1998) ‘Isostatic diffusion bonding of a microduplex stainless steel’,
Scripta Mater., Vol. 38, pp.1187–1193.
Juan, W., Li, Y., Liu, P. and Haoran, G. (2008) ‘Microstructure and XRD analysis in the interface
zone of Mg/Al diffusion bonding’, J. Mater. Process Tech., Vol. 205,
pp.146–150.,doi:10.1016/j.jmatprotec.2007.11.096.
Kenevisi, M.S. and Mousavi Khoie, S.M. (2012) ‘A study on the effect of bonding time on the
properties of Al7075 to Ti–6Al–4 V diffusion bonded joint’, Materials Letters, Vol. 76,
pp.144–146.
Kundu, S. and Chatterjee, S. (2006) ‘Interfacial microstructure and mechanical properties of
diffusion-bonded titanium–stainless steel joints using a nickel interlayer’, Mater. Sci. Eng. A,
Vol. 425, pp.107–113.
Kundu, S., Ghosh, M., Laik, A., Bhanumurthy, K., Kale, G.B. and Chatterjee, S. (2005) ‘Diffusion
bonding of commercially pure titanium to 304 SS using copper interlayer’, Materials Science
and Engineering , Vol. A 407, pp.154–160.
Kurgan, N. (2014) ‘Investigation of the effect of diffusion bonding parameters on microstructure
and mechanical properties of 7075 aluminium alloy’, International Journal of Advanced
Manufacturing Technology, Vol. 71, pp.2115–2124.
Li, Y., Liu, P., Wang, J. and Ma, H. (2008) ‘XRD and SEM analysis near the diffusion bonding
interface of Mg/Al dissimilar materials’, Vac., Vol. 82, pp.15–19,
doi:10.1016/j.vacuum.2007.01.073.
Liu, P., Li, Y., Haoran, G. and Juan, W. (2005) ‘Studies of phase constitution near the interface of
Mg/Al Vacuum diffusion bonding’, Mat. Let., Vol. 59, pp.2001–2005,
doi:10.1016/j.matlet.2005.02.038.
Liu, P., Li, Y., Haoran, G. and Juan, W. (2006) ‘Investigation of interfacial structure of Mg/Al
vacuum diffusion-bonded joint’, Vacuum, Vol. 80, pp.395–399.
Mahendran, G., Balasubramanian, V. and Senthilvelan, T. (2009) ‘Developing diffusion bonding
windows for joining AZ31B magnesium-AA2024 aluminium alloys’, Mater. Des., Vol. 30,
pp.1240–1244.
Mahendran, G., Balasubramanian, V. and Senthilvelan, T. (2010) ‘Influence of diffusion bonding
process parameters on bond characteristics of Mg-Cu dissimilar joints’, Tran. Non Ferr. Soc.
of China, Vol. 20, pp.997–1005.
 274 S. Venugopal and G. Mahendran

Mahendran, G., Balasubramanian, V. and Senthilvelan, T. (2011) ‘Mechanical and metallurgical

properties of diffusion bonded AA2024 aluminium alloy and commercial grade copper’, Elixir
Mech. Engg., Vol. 38, pp.4283–4289.
Pilling, J. and Rediely, N. (1987) ‘Solid state bonding of superplastic AA7475’, Material Science
and Technology, Vol. 3, pp.353–359.
Shirzadi, A.A., Saindrenan, G. and Wallach, E.R. (2002) ‘Flux-free diffusion brazing of
aluminium-based materials using gallium’, Materials Science Forum, pp.396–402,
pp.1579–158.
Smith, W.F. (1990) Principles of Materials Science and Engineering, 2nd ed., pp.290–306,
McGraw-Hill Publishing Company, New York.
Sunwoo, A. and Lum, R. (1995) ‘Super plastic deformation enhanced diffusion bonding of
aluminium alloy 7475’, Scripta Metall. Mater., Vol. 33, No. 4, p.639–644.
Ureña, A., Salazar, J.M., Gómez and Quiñones, J. (1992) ‘Interface reactions and bonding strength
in aluminium alloy (AA-7075)-alumina diffusion bonds’, Journal of Materials Science,
Vol. 27, p.5291–5296.
Vander Voort, G.F. (2004) ‘Metallography and microstructures’, ASM Handbook, ASM
International, Materials Park.
Wert, J.A., Paton, N.E., Hamilton, C.H. and Mahoney, M.W. (1981) ‘Grain refinement in 7075
aluminum by thermomechanical processing’, Metall. Trans., Vol. 12A, pp.1267–1276.
Wilkinson, D.S. (1988) in Hamilton, C.H. and Paton, N.E. (Eds.) ‘Super plasticity and super plastic
forming’, Microstructural Instability during Superplastic Flow, pp.81–91, TMS, Warrendale,
PA.
Yeh, M.S. and Chuang, T.S. (1995) ‘Low pressure diffusion bonding of SAE 316 stainless steel by
inserting a super plastic interlayer’, Scr. Met. Mat., Vol. 33, No. 8, pp.1277–1281,
doi:10.1016/0956-716X(95)00364-2.
Zhao, L.M. and Zhang, Z.D. (2008) ‘Effect of Zn alloy interlayer on interface microstructure and
strength of diffusion-bonded Mg-Al joints’, Scri. Mat., Vol. 58, pp283–286,
doi:10.1016/j.scriptamat.2007.10.006.

View publication stats