M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
861–869
MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS
OF Al/Cu FRICTION-STIR SPOT WELDS
RAZVOJ MIKROSTRUKTURE IN STATISTI^NA ANALIZA
VRTILNO-TORNIH TO^KASTIH ZVAROV Al/Cu
Mukuna Patrick Mubiayi1, Esther Titilayo Akinlabi1,
Mamookho Elizabeth Makhatha2
1University of Johannesburg, Department of Mechanical Engineering Science, Auckland Park Kingsway Campus,
2006 Johannesburg, South Africa
2University of Johannesburg, Department of Metallurgy, School of Mining, Metallurgy and Chemical Engineering, Doornfontein Campus,
2028 Doornfontein, South Africa
patrickmubiayi@gmail.com
Prejem rokopisa – received: 2016-11-18; sprejem za objavo – accepted for publication: 2017-03-16
doi:10.17222/mit.2016.320
In this paper, friction-stir spot welding (FSSW) is performed on 3 mm thick AA1060 and C11000 using different process
parameters and tool geometries. The microstructure- and the microhardness-profile analyses were conducted and the probability
distribution function (PDF) of the obtained microhardness values was determined. Optical images showed a good material
mixing in most of the spot welds produced, whereas the energy-dispersive-spectroscopy (EDS) analysis showed the presence of
intermetallic compounds. Microhardness results revealed that process parameters and tool geometries have significant effects on
the distribution of microhardness values in different locations of the produced spot welds. Furthermore, goodness-of-fit values
showed that most of the R2 values ranged between 0.8842 and 0.9999, which indicated how well the model fits with the
experimental data. On the other hand, the residuals comprised positive and negative runs which also indicated the existence of a
certain correlation with the experimentation.
Keywords: aluminium, copper, friction-stir spot welding, statistical analysis
V prispevku je predstavljeno vrtilno-torno to~kovno varjenje (angl. FSSW) na 3 mm debeli plo~evini AA1060 in C11000 z
uporabo razli~nih procesnih parametrov in razli~no geometrijo orodij. Izvedene so bile analize profila mikrostrukture in
mikrotrdote in dolo~ena je bila porazdelitvena funkcija verjetnosti za dobljene vrednosti mikrotrdote. Posnetki so pokazali
dobro me{anje materiala na ve~ini izdelanih to~kovnih zvarov, medtem ko je EDS-analiza pokazala prisotnost intermetalnih
spojin. Meritve mikrotrdote so pokazale, da imajo procesni parametri in geometrija orodij pomemben vpliv na porazdelitev
mikrotrdote glede na razli~ne lokacije izdelanih to~kovnih zvarov. Nadalje je serija opazovanj oz. ocena pokazala, da je ve~ina
R2 vrednosti rangiranih med 0,8842 in 0,9999. To potrjuje, da se model dobro ujema z eksperimentalnimi podatki. Po drugi
strani pa razlika med pozitivnimi in negativnimi cikli ka`e na obstoj dolo~ene korelacije s preizku{anjem.
Klju~ne besede: aluminij, baker, vrtilno-torno to~kovno varjenje, statisti~na analiza
1 INTRODUCTION
Friction-stir welding (FSW) is a fairly new solid-state
joining technique created and patented by The Welding
Institute (TWI) in 1991 for butt and lap welding of
ferrous and non-ferrous metals and plastics.1 Friction-stir
spot welding (FSSW) is a novel variant of linear
friction-stir welding (FSW) used for spot-welding
applications.2 The FSSW process uses a non-consumable
rotating tool which is plunged into the workpieces to be
welded. Before attaining the selected plunge depth, the
rotating tool is held at the same position for a fixed time,
which is defined as the dwell time. Consequently, the
rotating tool is withdrawn from the welded joint, leaving
a solid-phase friction-stir spot weld behind. Throughout
the FSSW process, the tool penetration depth and the
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 861
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
UDK 620.1:67.017:669.71:621.791 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 51(5)861(2017)
Figure 1: Schematic diagram of the friction-stir spot-welding process3
dwell time fundamentally determine the heat generation,
material plasticization around the tool’s pin, weld
geometry and, hence, the mechanical properties of the
welded joint.2 Figure 1 depicts a schematic illustration
of the FSSW technique.
It should be noted that the FSSW process uses a
non-consumable tool, which is similar to the FSW tool.4
The shoulder produces the frictional heat, whereas the
pin creates the material flow between the work sheets.2,3
In addition to the tool, the other parameters involved in
FSSW include the tool rotation speed, the tool plunge
depth and the tool dwell time. These parameters have an
effect on the strength, the surface texture of the welded
joints2 and the existence of weld defects.3
A typical cross-section of a friction-stir spot weld
displays five different structures including the parent
material (PM), the heat-affected zone (HAZ), the
thermomechanically affected zone (TMAZ), the stir zone
(SZ) and the hook.2
Friction-stir welding (FSW) and friction-stir spot
welding (FSSW) of aluminium and copper have thus far
not been fully investigated due to the huge difference
between their melting temperatures and the high chemi-
cal affinity of both materials which facilitate the forma-
tion of brittle intermetallic Al/Cu phases.6–18 Vickers
hardness measurement is a common technique used to
characterise the hardness of materials and it was reported
that the presence of intermetallics affects the hardness
values of the produced FS welds and FSS welds.11,17 On
the other hand, statistical analyses of Al/Cu friction-stir
welds and spot welds have not been well researched. S.
Akinlabi and E. T. Akinlabi19 conducted statistical ana-
lyses on the data obtained from dissimilar friction-stir
butt welds of aluminium (AA5754) and copper (C11000)
to understand the link between the process parameters
and the properties of the resulting welds. They concluded
that the downward vertical force has a significant effect
on the ultimate tensile strength (UTS) of the produced
welds. A robust relationship between the electrical resis-
tivity and the heat input into the welds was also ob-
served.19
Statistical analyses evaluate the effects of process
parameters on the properties of the produced spot welds
and establish the relationships amongst the process
parameters.
Therefore, in the current study, an effort was made to
further understand the relationships between the process
parameters and the microhardness profiles of FSSW
welds between copper and aluminium using the proba-
bility distribution function (PDF). Furthermore, the
microstructure and the chemical analyses of the pro-
duced spot welds were also studied.
2 MATERIALS AND WELDING PARAMETERS
In this study, AA1060 and C11000 base materials
with dimensions of 3 mm thickness, 600 mm length and
120 mm width were friction-stir spot lap welded. The
chemical compositions of the two parent materials were
determined using a spectrometer. The chemical compo-
sition of the aluminium sheet is as follows: 0.058 %
mass fraction of Si, 0.481 % mass fraction of Fe,
0.011 % mass fraction of Ga, 0.05 % mass fraction of
other elements and the rest is Al. The chemical com-
position of the copper sheet is: 0.137 % mass fraction of
Zn, <0.1 % mass fractions of Pb, 0.02 % mass fraction of
Ni, 0.023 % mass fraction of Al, 0.012 % mass fraction
of Co, 0.077 % mass fraction of B, 0.036 % mass
fraction of Sb, 0.043 % mass fraction of Nb, <0.492 %
mass fraction of other elements and the rest is Cu.
The sheets were friction-stir spot welded in a 30 mm
overlap configuration. The spot welds were produced at
the eNtsa of Nelson Mandela Metropolitan University,
Port Elizabeth, South Africa using an MTS PDS I-Stir.
The tool material used was H13 tool steel hardened to
50–52 HRC with a 4 mm tool pin, 5 mm tool diameter
and 15 mm tool shoulder. The friction-stir spot welds
were produced at rotational speeds of 800 min–1 and
1200 min–1, the tool-shoulder-plunge depths employed
were 0.5 mm and 1 mm at a constant dwell time of 10 s.
The two different tool profiles used in the current study
were flat pin/flat shoulder and conical pin/concave
shoulder tool, designated as FPS and CCS, respectively.
The produced welds were designated as XX_XX_XX
with the first part describing the tool geometry, the
second part indicating the rotational speed and the third
part indicating the shoulder plunge depth.
The weld samples were sectioned using wire elec-
trical discharge machining (WEDM), grinded and
polished, mounted and prepared, using the ASTM
standard metallographic procedure and ASTM Standard
E3-11.20
A solution of FeCl3 (10g) + HCl (6 mL) + ethanol
(C2H5OH) (20 mL) + H2O (80 mL) was used to etch the
copper side of the spot welds, while the aluminium side
was etched with H2O (190 mL) + HNO3 (5 mL) + HCl
(10 mL) + HF (2 mL). The microstructure of the spot
welds was studied using an optical microscope (Olympus
BX51M) equipped with the Stream software. Scanning
electron microscopy combined with energy dispersive
spectroscopy (SEM/EDS) was used to further examine
the microstructure and the chemical analyses, respec-
tively. A TESCAN equipped with Oxford Instruments
X-Max was used for the SEM/EDS analyses. The
Vickers microhardness was determined using a dia-
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
862 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 2: Representation of a spot weld with dashed lines illustrating
the location of the microhardness-profile measurements
mond-pyramid-indenter EMCO Test DuraScan tester.
Two different locations on the spot welds were used,
viz., the top and the bottom, as illustrated in Figure 2.
The microhardness measurements were carried out from
the keyhole for all the different parameters and tool
geometries, to find the probability density function
(PDF) of each one. The Matlab 2014 software program
was used to determine the PDF.
3 RESULTS AND DISCUSSION
The microstructure of the friction-stir spot welds was
studied using an optical microscope and the results are
depicted in Figures 3 and 4.
Figure 3 illustrates the microstructures of the spot
welds produced at: a) 800 min–1 and b) 1200 min–1, a
1 mm shoulder plunge depth using a flat pin and a flat
shoulder tool. On the other hand, Figure 4 shows the
microstructure of the spot weld produced at: a) 800 min–1
and b) 1200 min–1, a 0.5 mm shoulder plunge depth
using a conical pin and a concave shoulder tool.
It can be seen in Figure 3a that there are a copper
ring and a mixture of Al/Cu particles in the stir zone.
There is no palpable welding defect in the weld and
copper is disseminated in this zone with different shapes.
In the upper part of the joint, a large bulk of copper with
irregular shapes can be observed (Figure 3a). The tool
pin was inserted in the aluminium plate and the copper
ring extruded upward from the lower copper plate into
the aluminium plate was observed. This was in agree-
ment with the previous work.15
Additionally, the intermixing of copper and alumi-
nium was not homogenous for different spot welds and
different microstructures were formed in different
regions of the welds. It was reported that the FSW of
dissimilar materials is different from that of similar
materials due to the formation of a complex, intercalated
vortex-like and related flow pattern.21
In Figure 4b, a good interlaced structure can be seen.
This is formed by aluminium and copper, thereby
indicating that the two plates are bonded firmly in this
region, which is composed of a lamellar structure of
copper particles with a streamlined shape of aluminium
strips. In this region, a few disseminated copper particles
were also observed.
The energy-dispersive-spectrometry analyses at the
selected points in the stir zone were recorded using SEM
micrographs (Table 1). Intermetallic compounds were
found in most of the produced welds. Two intermetallic
compounds, viz., Al2Cu and Al3Cu4 were found in the
weld produced at 1200 min–1, 0.5 mm shoulder plunge
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 863
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 4: Optical microscope images showing the macrostructures of
the joints at: a) 800 min–1 and b) 1200 min–1, 0.5 mm shoulder plunge
depth using a conical pin and concave shoulder tool
Figure 3: Optical microscope images showing the macrostructures of
the joints at: a) 800 min–1 and b) 1200 min–1, 1 mm shoulder plunge
depth using a flat pin and flat shoulder tool
dwell time fundamentally determine the heat generation,
material plasticization around the tool’s pin, weld
geometry and, hence, the mechanical properties of the
welded joint.2 Figure 1 depicts a schematic illustration
of the FSSW technique.
It should be noted that the FSSW process uses a
non-consumable tool, which is similar to the FSW tool.4
The shoulder produces the frictional heat, whereas the
pin creates the material flow between the work sheets.2,3
In addition to the tool, the other parameters involved in
FSSW include the tool rotation speed, the tool plunge
depth and the tool dwell time. These parameters have an
effect on the strength, the surface texture of the welded
joints2 and the existence of weld defects.3
A typical cross-section of a friction-stir spot weld
displays five different structures including the parent
material (PM), the heat-affected zone (HAZ), the
thermomechanically affected zone (TMAZ), the stir zone
(SZ) and the hook.2
Friction-stir welding (FSW) and friction-stir spot
welding (FSSW) of aluminium and copper have thus far
not been fully investigated due to the huge difference
between their melting temperatures and the high chemi-
cal affinity of both materials which facilitate the forma-
tion of brittle intermetallic Al/Cu phases.6–18 Vickers
hardness measurement is a common technique used to
characterise the hardness of materials and it was reported
that the presence of intermetallics affects the hardness
values of the produced FS welds and FSS welds.11,17 On
the other hand, statistical analyses of Al/Cu friction-stir
welds and spot welds have not been well researched. S.
Akinlabi and E. T. Akinlabi19 conducted statistical ana-
lyses on the data obtained from dissimilar friction-stir
butt welds of aluminium (AA5754) and copper (C11000)
to understand the link between the process parameters
and the properties of the resulting welds. They concluded
that the downward vertical force has a significant effect
on the ultimate tensile strength (UTS) of the produced
welds. A robust relationship between the electrical resis-
tivity and the heat input into the welds was also ob-
served.19
Statistical analyses evaluate the effects of process
parameters on the properties of the produced spot welds
and establish the relationships amongst the process
parameters.
Therefore, in the current study, an effort was made to
further understand the relationships between the process
parameters and the microhardness profiles of FSSW
welds between copper and aluminium using the proba-
bility distribution function (PDF). Furthermore, the
microstructure and the chemical analyses of the pro-
duced spot welds were also studied.
2 MATERIALS AND WELDING PARAMETERS
In this study, AA1060 and C11000 base materials
with dimensions of 3 mm thickness, 600 mm length and
120 mm width were friction-stir spot lap welded. The
chemical compositions of the two parent materials were
determined using a spectrometer. The chemical compo-
sition of the aluminium sheet is as follows: 0.058 %
mass fraction of Si, 0.481 % mass fraction of Fe,
0.011 % mass fraction of Ga, 0.05 % mass fraction of
other elements and the rest is Al. The chemical com-
position of the copper sheet is: 0.137 % mass fraction of
Zn, <0.1 % mass fractions of Pb, 0.02 % mass fraction of
Ni, 0.023 % mass fraction of Al, 0.012 % mass fraction
of Co, 0.077 % mass fraction of B, 0.036 % mass
fraction of Sb, 0.043 % mass fraction of Nb, <0.492 %
mass fraction of other elements and the rest is Cu.
The sheets were friction-stir spot welded in a 30 mm
overlap configuration. The spot welds were produced at
the eNtsa of Nelson Mandela Metropolitan University,
Port Elizabeth, South Africa using an MTS PDS I-Stir.
The tool material used was H13 tool steel hardened to
50–52 HRC with a 4 mm tool pin, 5 mm tool diameter
and 15 mm tool shoulder. The friction-stir spot welds
were produced at rotational speeds of 800 min–1 and
1200 min–1, the tool-shoulder-plunge depths employed
were 0.5 mm and 1 mm at a constant dwell time of 10 s.
The two different tool profiles used in the current study
were flat pin/flat shoulder and conical pin/concave
shoulder tool, designated as FPS and CCS, respectively.
The produced welds were designated as XX_XX_XX
with the first part describing the tool geometry, the
second part indicating the rotational speed and the third
part indicating the shoulder plunge depth.
The weld samples were sectioned using wire elec-
trical discharge machining (WEDM), grinded and
polished, mounted and prepared, using the ASTM
standard metallographic procedure and ASTM Standard
E3-11.20
A solution of FeCl3 (10g) + HCl (6 mL) + ethanol
(C2H5OH) (20 mL) + H2O (80 mL) was used to etch the
copper side of the spot welds, while the aluminium side
was etched with H2O (190 mL) + HNO3 (5 mL) + HCl
(10 mL) + HF (2 mL). The microstructure of the spot
welds was studied using an optical microscope (Olympus
BX51M) equipped with the Stream software. Scanning
electron microscopy combined with energy dispersive
spectroscopy (SEM/EDS) was used to further examine
the microstructure and the chemical analyses, respec-
tively. A TESCAN equipped with Oxford Instruments
X-Max was used for the SEM/EDS analyses. The
Vickers microhardness was determined using a dia-
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
862 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 2: Representation of a spot weld with dashed lines illustrating
the location of the microhardness-profile measurements
mond-pyramid-indenter EMCO Test DuraScan tester.
Two different locations on the spot welds were used,
viz., the top and the bottom, as illustrated in Figure 2.
The microhardness measurements were carried out from
the keyhole for all the different parameters and tool
geometries, to find the probability density function
(PDF) of each one. The Matlab 2014 software program
was used to determine the PDF.
3 RESULTS AND DISCUSSION
The microstructure of the friction-stir spot welds was
studied using an optical microscope and the results are
depicted in Figures 3 and 4.
Figure 3 illustrates the microstructures of the spot
welds produced at: a) 800 min–1 and b) 1200 min–1, a
1 mm shoulder plunge depth using a flat pin and a flat
shoulder tool. On the other hand, Figure 4 shows the
microstructure of the spot weld produced at: a) 800 min–1
and b) 1200 min–1, a 0.5 mm shoulder plunge depth
using a conical pin and a concave shoulder tool.
It can be seen in Figure 3a that there are a copper
ring and a mixture of Al/Cu particles in the stir zone.
There is no palpable welding defect in the weld and
copper is disseminated in this zone with different shapes.
In the upper part of the joint, a large bulk of copper with
irregular shapes can be observed (Figure 3a). The tool
pin was inserted in the aluminium plate and the copper
ring extruded upward from the lower copper plate into
the aluminium plate was observed. This was in agree-
ment with the previous work.15
Additionally, the intermixing of copper and alumi-
nium was not homogenous for different spot welds and
different microstructures were formed in different
regions of the welds. It was reported that the FSW of
dissimilar materials is different from that of similar
materials due to the formation of a complex, intercalated
vortex-like and related flow pattern.21
In Figure 4b, a good interlaced structure can be seen.
This is formed by aluminium and copper, thereby
indicating that the two plates are bonded firmly in this
region, which is composed of a lamellar structure of
copper particles with a streamlined shape of aluminium
strips. In this region, a few disseminated copper particles
were also observed.
The energy-dispersive-spectrometry analyses at the
selected points in the stir zone were recorded using SEM
micrographs (Table 1). Intermetallic compounds were
found in most of the produced welds. Two intermetallic
compounds, viz., Al2Cu and Al3Cu4 were found in the
weld produced at 1200 min–1, 0.5 mm shoulder plunge
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 863
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 4: Optical microscope images showing the macrostructures of
the joints at: a) 800 min–1 and b) 1200 min–1, 0.5 mm shoulder plunge
depth using a conical pin and concave shoulder tool
Figure 3: Optical microscope images showing the macrostructures of
the joints at: a) 800 min–1 and b) 1200 min–1, 1 mm shoulder plunge
depth using a flat pin and flat shoulder tool
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
864 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Table 1 displays SEM micrographs and the EDS analysis of selected points on the produced friction-spot welds: a) (FFS_1200_0.5), b)
(FFS_1200_1), c) (CCS_1200_0.5) and d) (CCS_1200_1).
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
54.94
48.5
4.34
24.55
41.98
45.06
51.5
95.66
75.45
58.02
-
Al2Cu
-
Al3Cu4
Al2Cu
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
1.46
17.71
36.27
94.53
16.98
98.54
82.29
63.73
5.07
83.02
-
Al4Cu9
AlCu
-
Al4Cu9
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
3.1
7.82
49.49
29.22
13.77
96.9
92.18
50.27
70.78
86.23
-
-
Al2Cu
AlCu
AlCu3
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
2.26
1.33
92.77
37.35
9.15
97.74
98.67
6.97
62.65
90.85
-
-
-
AlCu
-
depth, whereas the weld produced at 1200 min–1 and 1
mm shoulder plunge depth contained Al4Cu9 and AlCu
intermetallic compounds. These intermetallics were
found in the welds produced using a flat pin and flat
shoulder tool. On the other hand, the spot welds pro-
duced using a conical pin and concave shoulder con-
tained Al2Cu, AlCu3 and AlCu intermetallic compounds.
These intermetallic compounds were found in the weld
produced at 1200 min–1 and 0.5 mm shoulder plunge
depth. Only the AlCu intermetallic was found in the
weld produced at 1200 min–1 and 1 mm shoulder plunge
depth, but the concentration of this intermetallic com-
pound was relatively small. It was reported that the
presence of intermetallic compounds could affect the
microhardness profile.17
Figure 5 (a, b, c and d) shows the microhardness
values of the spot welds produced using a flat pin and a
flat shoulder tool, or a conical pin and a concave
shoulder at different process parameters. The micro-
hardness values of the parent materials were in the range
of 86.7–96.3 HV for Cu while for Al, the range was
between 34.6–40.3 HV. In all the samples, high micro-
hardness values were recorded at the top, in the region
close to the keyhole.
It was reported that all the mechanical tests are
subject to large statistical variations, which should be
evaluated.22
The probability distribution function (PDF) of the
Vickers hardness was reported in the literature to
correspond to the Gaussian (or normal) distribution23 and
log-normal distribution.24 A. M. Hassan et al.25 studied
the significance of the process parameters of friction-stir
welding of aluminium-matrix composites to set the
optimum level for each of these parameters and to
further predict which responses are affected when using
analyses of variance (ANOVA).25 The present study used
the Matlab 2014a statistical toolbox to analyse the pro-
bability density function (PDF) of the obtained micro-
hardness results. This was done to understand how
different parameters and tool geometries affect the
probability of obtaining specific microhardness values.
A probability density function (PDF) is a function
that defines the relative possibility for a random variable
to take on a given value. The probability of the random
variable falling within a particular range [a, b] of values
is given with a finite integral of the PDF within that
range [a, b], Equation (1):
p x f x e
b
a x
b
a
( ) ( , , )
( )
= =∫ ∫
−
−μ
 
 

 

1
2
2
2
π
(1)
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 865
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 5: Microhardness distributions along the welds produced using different tools and process parameters: a) flat pin/flat shoulder (FPS), top;
b) bottom; c) conical pin/concave shoulder (CCS), top; d) bottom
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
864 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Table 1 displays SEM micrographs and the EDS analysis of selected points on the produced friction-spot welds: a) (FFS_1200_0.5), b)
(FFS_1200_1), c) (CCS_1200_0.5) and d) (CCS_1200_1).
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
54.94
48.5
4.34
24.55
41.98
45.06
51.5
95.66
75.45
58.02
-
Al2Cu
-
Al3Cu4
Al2Cu
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
1.46
17.71
36.27
94.53
16.98
98.54
82.29
63.73
5.07
83.02
-
Al4Cu9
AlCu
-
Al4Cu9
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
3.1
7.82
49.49
29.22
13.77
96.9
92.18
50.27
70.78
86.23
-
-
Al2Cu
AlCu
AlCu3
Point
Composition Intermetallic
compoundAl Cu
1
2
3
4
5
2.26
1.33
92.77
37.35
9.15
97.74
98.67
6.97
62.65
90.85
-
-
-
AlCu
-
depth, whereas the weld produced at 1200 min–1 and 1
mm shoulder plunge depth contained Al4Cu9 and AlCu
intermetallic compounds. These intermetallics were
found in the welds produced using a flat pin and flat
shoulder tool. On the other hand, the spot welds pro-
duced using a conical pin and concave shoulder con-
tained Al2Cu, AlCu3 and AlCu intermetallic compounds.
These intermetallic compounds were found in the weld
produced at 1200 min–1 and 0.5 mm shoulder plunge
depth. Only the AlCu intermetallic was found in the
weld produced at 1200 min–1 and 1 mm shoulder plunge
depth, but the concentration of this intermetallic com-
pound was relatively small. It was reported that the
presence of intermetallic compounds could affect the
microhardness profile.17
Figure 5 (a, b, c and d) shows the microhardness
values of the spot welds produced using a flat pin and a
flat shoulder tool, or a conical pin and a concave
shoulder at different process parameters. The micro-
hardness values of the parent materials were in the range
of 86.7–96.3 HV for Cu while for Al, the range was
between 34.6–40.3 HV. In all the samples, high micro-
hardness values were recorded at the top, in the region
close to the keyhole.
It was reported that all the mechanical tests are
subject to large statistical variations, which should be
evaluated.22
The probability distribution function (PDF) of the
Vickers hardness was reported in the literature to
correspond to the Gaussian (or normal) distribution23 and
log-normal distribution.24 A. M. Hassan et al.25 studied
the significance of the process parameters of friction-stir
welding of aluminium-matrix composites to set the
optimum level for each of these parameters and to
further predict which responses are affected when using
analyses of variance (ANOVA).25 The present study used
the Matlab 2014a statistical toolbox to analyse the pro-
bability density function (PDF) of the obtained micro-
hardness results. This was done to understand how
different parameters and tool geometries affect the
probability of obtaining specific microhardness values.
A probability density function (PDF) is a function
that defines the relative possibility for a random variable
to take on a given value. The probability of the random
variable falling within a particular range [a, b] of values
is given with a finite integral of the PDF within that
range [a, b], Equation (1):
p x f x e
b
a x
b
a
( ) ( , , )
( )
= =∫ ∫
−
−μ
 
 

 

1
2
2
2
π
(1)
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 865
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 5: Microhardness distributions along the welds produced using different tools and process parameters: a) flat pin/flat shoulder (FPS), top;
b) bottom; c) conical pin/concave shoulder (CCS), top; d) bottom
where  is the standard deviation, 2 is the variance and
μ is the mean.
It is given by the area under the density function,
nonetheless above the horizontal axis and in between the
lowest and highest values of the range. The probability
density function is a non-negative value and its integral
over the entire space is equal to one.
PDF histograms of the microhardness and their fits
for the parent materials, namely, aluminium and copper,
are depicted in Figures 6a and 7b, respectively. The
PDFs of the top and bottom hardness measurements
were investigated, Figures 7a and 7b depict the PDF
histograms of the microhardness for the weld produced
at 800 min–1, 0.5 mm shoulder plunge depth, using a flat
pin and a flat shoulder tool.
It can be seen that the probability of having the
microhardness values between 40 and 45 HV is high in
the histogram, as shown in Figure 7a, which represents
the microhardness measured at the top of the spot weld.
This corresponds mostly to the microhardness of the
aluminium parent material, whereas the possibility of
getting high microhardness values between 50 and 60
HV is low.
On the other hand, the PDF of the bottom measu-
rement (Figure 7b) shows that there is a high possibility
of getting microhardness values between 85 and 95 HV.
This corresponds to the microhardness of the copper-
parent sheets, while the microhardness values between
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
866 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 6: Depicting the microhardness PDF histograms of the parent
materials: a) aluminium, b) copper
Figure 8: PDF histogram of the microhardness of CCS_800_0.5 spot
weld: a) top and b) bottom
Figure 7: PDF histograms of the microhardness of FPS_800_0.5 spot
weld: a) top and b) bottom
100 HV and 105 HV are likely to show a lower possi-
bility of being obtained when using the same process
parameters as those used in this research work. The
possibility of having higher microhardness values
compared to the values of the parent materials in the two
different sheets (copper and aluminium) was observed to
be due to the presence of a mixture of copper and
aluminium in the vicinity of the keyhole. Additionally,
Figure 8 depicts a PDF histogram of the microhardness
measurements ((a) top and (b) bottom) for the weld
produced at 800 min–1, 0.5 mm shoulder plunge depth,
using a conical pin and a concave shoulder.
The results show that there is a higher possibility for
obtaining microhardness values between 30 50 HV and
50 HV, whereas the possibility for high values between
120 and 130 HV is low (Figure 8a). The trend for the
bottom area (Figure 8b) is similar to the one discussed
above for the PDF of the bottom microhardness obtained
using a flat pin and a flat shoulder.
Moreover, when the rotational speed of 1200 min–1 is
increased, the possibility of getting high microhardness
values ranging between 100–110 HV and 90–100 HV
increases for the top area of the spot weld produced
using a conical pin and a concave shoulder, and for the
top and bottom microhardness values, respectively. It can
be seen that the rotational speed and the tool geometry
may influence the possibility of different probability
distributions.
The model shows that, in order to get the probability
in a specific region, the integral of the PDF for the region
of interest need to be computed. The PDF found in the
current research work is a normal distribution (called a
Gaussian distribution as well). In order to get any pro-
bability, we can compute the finite integral of the normal
distribution equation (1).
The goodness of fit and the residuals were also
analysed. The results show that most of the R2 values
range between 0.8842 and 0.9999, which is an indication
of how well the model fits with the experimental data
shown in Table 2.
Table 2: R2 and adjusted R2 of the welds produced using a flat pin/flat
shoulder tool and conical pin/concave shoulder tool for the micro-
hardness measured on the top and at the bottom
Sample ID R2 square R2 square
FPS_800_0.5 0.9896 0.8842
FPS_800_1 0.9996 0.9999
FPS_1200_0.5 0.9999 0.9937
FPS_1200_1 0.9993 0.9298
CCS_800_0.5 0.9924 0.9818
CCS_800_1 0.9976 0.9424
CCS_1200_0.5 0.9893 0.987
CCS_1200_1 0.9997 0.9866
Figures 9a to 9b depict the goodness of fit and the
residuals for the weld produced at 1200 min–1, 1 mm
shoulder plunge depth, using a flat pin and a flat
shoulder (the top-microhardness measurement). In
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 867
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 9: a) goodness of fit and b) the residuals for the spot weld
produced at 1200 min–1, 1 mm shoulder plunge depth, using a flat pin
and a flat shoulder (top-microhardness measurements)
Figure 10: a) goodness of fit and b) the residuals for the spot weld
produced at 1200 min–1, 1 mm shoulder plunge depth, using a conical
pin and a concave shoulder (top-microhardness measurements)
where  is the standard deviation, 2 is the variance and
μ is the mean.
It is given by the area under the density function,
nonetheless above the horizontal axis and in between the
lowest and highest values of the range. The probability
density function is a non-negative value and its integral
over the entire space is equal to one.
PDF histograms of the microhardness and their fits
for the parent materials, namely, aluminium and copper,
are depicted in Figures 6a and 7b, respectively. The
PDFs of the top and bottom hardness measurements
were investigated, Figures 7a and 7b depict the PDF
histograms of the microhardness for the weld produced
at 800 min–1, 0.5 mm shoulder plunge depth, using a flat
pin and a flat shoulder tool.
It can be seen that the probability of having the
microhardness values between 40 and 45 HV is high in
the histogram, as shown in Figure 7a, which represents
the microhardness measured at the top of the spot weld.
This corresponds mostly to the microhardness of the
aluminium parent material, whereas the possibility of
getting high microhardness values between 50 and 60
HV is low.
On the other hand, the PDF of the bottom measu-
rement (Figure 7b) shows that there is a high possibility
of getting microhardness values between 85 and 95 HV.
This corresponds to the microhardness of the copper-
parent sheets, while the microhardness values between
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
866 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 6: Depicting the microhardness PDF histograms of the parent
materials: a) aluminium, b) copper
Figure 8: PDF histogram of the microhardness of CCS_800_0.5 spot
weld: a) top and b) bottom
Figure 7: PDF histograms of the microhardness of FPS_800_0.5 spot
weld: a) top and b) bottom
100 HV and 105 HV are likely to show a lower possi-
bility of being obtained when using the same process
parameters as those used in this research work. The
possibility of having higher microhardness values
compared to the values of the parent materials in the two
different sheets (copper and aluminium) was observed to
be due to the presence of a mixture of copper and
aluminium in the vicinity of the keyhole. Additionally,
Figure 8 depicts a PDF histogram of the microhardness
measurements ((a) top and (b) bottom) for the weld
produced at 800 min–1, 0.5 mm shoulder plunge depth,
using a conical pin and a concave shoulder.
The results show that there is a higher possibility for
obtaining microhardness values between 30 50 HV and
50 HV, whereas the possibility for high values between
120 and 130 HV is low (Figure 8a). The trend for the
bottom area (Figure 8b) is similar to the one discussed
above for the PDF of the bottom microhardness obtained
using a flat pin and a flat shoulder.
Moreover, when the rotational speed of 1200 min–1 is
increased, the possibility of getting high microhardness
values ranging between 100–110 HV and 90–100 HV
increases for the top area of the spot weld produced
using a conical pin and a concave shoulder, and for the
top and bottom microhardness values, respectively. It can
be seen that the rotational speed and the tool geometry
may influence the possibility of different probability
distributions.
The model shows that, in order to get the probability
in a specific region, the integral of the PDF for the region
of interest need to be computed. The PDF found in the
current research work is a normal distribution (called a
Gaussian distribution as well). In order to get any pro-
bability, we can compute the finite integral of the normal
distribution equation (1).
The goodness of fit and the residuals were also
analysed. The results show that most of the R2 values
range between 0.8842 and 0.9999, which is an indication
of how well the model fits with the experimental data
shown in Table 2.
Table 2: R2 and adjusted R2 of the welds produced using a flat pin/flat
shoulder tool and conical pin/concave shoulder tool for the micro-
hardness measured on the top and at the bottom
Sample ID R2 square R2 square
FPS_800_0.5 0.9896 0.8842
FPS_800_1 0.9996 0.9999
FPS_1200_0.5 0.9999 0.9937
FPS_1200_1 0.9993 0.9298
CCS_800_0.5 0.9924 0.9818
CCS_800_1 0.9976 0.9424
CCS_1200_0.5 0.9893 0.987
CCS_1200_1 0.9997 0.9866
Figures 9a to 9b depict the goodness of fit and the
residuals for the weld produced at 1200 min–1, 1 mm
shoulder plunge depth, using a flat pin and a flat
shoulder (the top-microhardness measurement). In
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 867
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 9: a) goodness of fit and b) the residuals for the spot weld
produced at 1200 min–1, 1 mm shoulder plunge depth, using a flat pin
and a flat shoulder (top-microhardness measurements)
Figure 10: a) goodness of fit and b) the residuals for the spot weld
produced at 1200 min–1, 1 mm shoulder plunge depth, using a conical
pin and a concave shoulder (top-microhardness measurements)
addition, Figures 10a to 10b depict the goodness of fit
and the residuals for the weld produced at 1200 min–1, 1
mm shoulder plunge depth, using a conical pin and a
concave shoulder (the top-microhardness measurement).
Besides the R2 values, the residual analysis was also
employed in the study in order to check the adequacy of
the models. Figures 9b and 10b show the residual plots
for the microhardness values obtained at the top, using a
flat pin/flat shoulder and a conical pin/concave shoulder,
using the 1200 min–1 speed and 1 mm shoulder plunge
depth, respectively. It was reported that the tendencies to
have runs of positive and negative residuals indicate the
existence of a certain correlation with the experimen-
tation.26
Tables 3 and 4 present the standard deviation, the
variance and the mean obtained from the statistical
analyses of the measured microhardness values for the
two different positions, namely, the top and bottom of
the spot welds produced with different tools and using
different process parameters.
Table 3: Mean, variance, mu and sigma of the spot samples produced
using a flat pin/flat shoulder tool and a conical pin/concave shoulder
tool for the microhardness taken on the top
Sample ID Mean 2 μ 
FPS_800_0.5 46.543 82.8442 46.5429 9.102
FPS_800_1 75.099 4994.97 4.00167 0.796
FPS_1200_0.5 45.121 441.806 3.71115 0.443
FPS_1200_1 51.731 672.845 3.83391 0.474
CCS_800_0.5 44.575 409.802 3.70341 0.433
CCS_800_1 55.887 813.763 3.90757 0.481
CCS_1200_0.5 45.453 482.889 3.71166 0.458
CCS_1200_1 40.942 248.04 3.64315 0.371
Table 4: Mean, variance, mu and sigma of the spot samples produced
using a flat pin/flat shoulder tool and a conical pin/concave shoulder
tool for the microhardness taken at the bottom
Sample ID Mean 2 μ 
FPS_800_0.5 92.234 32.7196 4.52241 0.062
FPS_800_1 84.7 119.974 84.7 10.95
FPS_1200_0.5 84.7 119.974 84.7 10.95
FPS_1200_1 78.429 58.0637 78.4286 7.62
CCS_800_0.5 89.986 14.2859 89.9857 3.78
CCS_800_1 88.564 102.307 88.5643 10.11
CCS_1200_0.5 86.252 89.6604 4.45128 0.109
CCS_1200_1 82.043 54.8519 4.40318 0.09
It should be noted that the mean is equal to μ if the
distribution is normal. It can be seen that in some cases
in the current work μ and the mean have different values,
which shows that the distributions in some of the
analyses were not normal. The standard deviation 
should be close to zero, but in the current work, the value
of the standard deviation is not close to zero in some of
the cases. This shows that the microhardness values are
not close to the expected values and this could be due to
the microhardness values measured in different locations
of the weld samples, far apart from each other.
This was further suspected due to the presence of
intermetallics, which could have been the cause of the
high microhardness values since intermetallics are
invariably hard and brittle.
Each figure contains residuals versus the distance
(the distance from the keyhole), taking into account the
data and the constant variance of the residuals. In the
plot of residuals versus distance, it is shown that the
models are adequate to predict the responses in an
acceptable manner.
4 CONCLUSIONS
According to the presented results, some conclusions
can be drawn:
• The microstructure of the produced spot welds shows
a good material mixing, the presence of a copper ring
and a mixture of Al/Cu particles present in the stir
zone.
• The EDS analyses of the produced friction-stir spot
welds exhibited the presence of intermetallic com-
pounds, which are known to affect the microhard-
ness.
• The microhardness values obtained at the top were
high for all the samples and this was in the region
close to the spot-weld keyhole. In addition, all the
microhardness values obtained at the bottom of the
samples, in the region close to the keyhole, have
lower values, which were close to the average value
of the copper base material. This occurred for all the
spot welds produced using a conical pin and concave
shoulder.
• The probability-density-function (PDF) histograms
of the microhardness results revealed that the process
parameters and the tool geometries have significant
effects on the distribution of the microhardness
values in different locations of the produced spot
welds.
• Additionally, goodness-of-fit values were also
analysed and these showed that most of the R2 values
ranged between 0.8842 and 0.9999, which is an
indication of how well the model fits with the
produced experimental data.
Acknowledgements
The authors wish to acknowledge the financial
support of the University of Johannesburg and the
assistance from Mr. Riaan Brown (eNtsa at Nelson
Mandela Metropolitan University).
5 REFERENCES
1 W. M. Thomas, E. D. Nicholas, J. C. Needham, M. G. Murch, P.
Temple-Smith, C. J. Dawes, International Patent No. PCT/GB92/
02203, GB patent application No. 9125978.8, 1991
2 H. Badarinarayan, Fundamentals of Friction Stir Spot Welding, PhD
Thesis, 2009, Missouri
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
868 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
3 W. Yuan, R. S. Mishra, B. Carlson, R. Verma, R. K. Mishra, Material
flow and microstructural evolution during friction stir spot welding
of AZ31 magnesium alloy, Materials Science and Engineering, A
543 (2012), 200–209
4 J. Podr`aj, B. Jerman, D. Klob~ar, Welding defects at friction stir
welding, Metalurgija, 54 (2015) 2, 387–389
5 J. M. Timothy, Friction Stir Welding of Commercially Available
Superplastic Aluminium, PhD thesis, 2008, Department of Engi-
neering and Design, Brunel University, Brunel
6 J. Ouyang, E. Yarrapareddy, R. Kovacevic, Microstructural evolution
in the friction stir welded 6061 aluminum alloy (T6-temper condi-
tion) to copper, Journal of Materials Processing Technology, 172
(2006), 110–122
7 P. Liu, Q. Shi, W. Wang, X. Wang, Z. Zhang, Microstructure and
XRD analysis of FSW joints for copper T2/aluminium 5A06
dissimilar materials, Materials Letters, 62 (2008), 4106–4108
8 P. Xue, B. L. Xiao, D. R. Ni, Z. Y. Ma, Enhanced mechanical pro-
perties of friction stir welded dissimilar Al–Cu joint by intermetallic
compounds, Materials Science and Engineering A, 527 (2010),
5723–5727
9 A. Esmaeili, M. K. Besharati Givi, H. R. Zareie Rajani, A metallur-
gical and mechanical study on dissimilar friction stir welding of
aluminum 1050 to brass (CuZn30), Materials Science and Engi-
neering, A 528 (2011), 7093–7102
10 A. Abdollah-Zadeh, T. Saeid, B. Sazgari, Microstructural and mecha-
nical properties of friction stir welded aluminum/copper lap joints,
Journal of Alloys and Compounds, 460 (2008), 535–538
11 E. T. Akinlabi, Characterisation of Dissimilar Friction Stir Welds
between 5754 Aluminium Alloy and C11000 Copper, D-Tech thesis,
Nelson Mandela Metropolitan University, South Africa, 2010
12 M. N. Avettand-Fenoel, R. Taillard, G. Ji, D. Goran, Multiscale
Study of Interfacial Intermetallic Compounds in a Dissimilar Al
6082-T6/Cu Friction-Stir Weld, Metallurgical and Materials
Transactions A, (2012), 4655–4666
13 M. F. X Muthu, V. Jayabalan, Tool Travel Speed Effects on the
Microstructure of Friction Stir Welded Aluminium – Copper Joints,
Journal of Materials Processing Technology, 217 (2015), 105–113
14 M. P. Mubiayi, E. T Akinlabi, Friction Stir Spot Welding between
Copper and Aluminium: Microstructural Evolution, Proceedings of
the International MultiConference of Engineers and Computer
Scientists, Vol II, IMECS 2015, March 18–20, 2015, Hong Kong
15 R. Heideman, C. Johnson, S. Kou, Metallurgical analysis of Al/Cu
friction stir spot welding, Science and Technology of Welding and
Joining, 15 (2010) 7, 597–604
16 U. Özdemir, S. Sayer, Ç. Yeni, Bornova-Izmir, Effect of Pin Pene-
tration Depth on the Mechanical Properties of Friction Stir Spot
Welded Aluminum and Copper, Materials Testing in Joining Techno-
logy, 54 (2012) 4, 233–239
17 M. Shiraly, M. Shamanian, M. R. Toroghinejad, M. Ahmadi Jazani,
Effect of Tool Rotation Rate on Microstructure and Mechanical
Behavior of Friction Stir Spot-Welded Al/Cu Composite, Journal of
Materials Engineering and Performance, 23 (2014) 2, 413–420
18 M. P. Mubiayi, E. T. Akinlabi, Evolving properties of friction stir
spot welds between AA1060 and commercially pure copper C11000,
Transactions of Nonferrous Metals Society of China, 26 (2016),
1852–1862
19 E. T Akinlabi, S. A. Akinlabi, Friction stir welding of dissimilar
materials – statistical analysis of the weld data, Proceedings of the
International MultiConference of Engineers and Computer Scientists,
2012, 1368–1373
20 ASTM Standard E3-11, Standard guide for preparation of
metallographic specimens, ASTM International, West Consho-
hocken, PA, 2011, doi: 10.1520/E0003-11, www.astm.org
21 L. E Murr, A Review of FSW Research on Dissimilar Metal and
Alloy Systems, Journal of Materials Engineering and Performance, 8
(2010)19, 1071–108
22 J. M. Schneider, M. Bigerelle, A. Iost, Statistical analysis of the
Vickers hardness, Materials Science and Engineering, A 262 (1999),
256–263
23 A. L. Yurkov, N. V. Jhuravleva, E. S. Lukin, Kinetic microhardness
measurements of sialon-based ceramics, Journal of Materials
Science, 29 (1994), 6551–6560
24 I. Y. Yanchev, E. P. Trifonova, Analysis of microhardness data in Tlx
ln l-x Se, Journal of Materials Science, 30 (1995), 5576–5580
25 A. M. Hassan, M. Almomani, T. Qasim, A. Ghaithan, Statistical
analysis of some mechanical properties of friction stir welded
aluminium matrix composite, Int. J. Experimental Design and
Process Optimisation, 3 (2012)1, 91–109
26 K. Palanikumar, R. Karthikeyan, Optimal machining conditions for
turning of particulate metal matrix composites using Taguchi and
response surface methodologies, Machining Science and Technology,
10 (2006) 4, 417–433
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
addition, Figures 10a to 10b depict the goodness of fit
and the residuals for the weld produced at 1200 min–1, 1
mm shoulder plunge depth, using a conical pin and a
concave shoulder (the top-microhardness measurement).
Besides the R2 values, the residual analysis was also
employed in the study in order to check the adequacy of
the models. Figures 9b and 10b show the residual plots
for the microhardness values obtained at the top, using a
flat pin/flat shoulder and a conical pin/concave shoulder,
using the 1200 min–1 speed and 1 mm shoulder plunge
depth, respectively. It was reported that the tendencies to
have runs of positive and negative residuals indicate the
existence of a certain correlation with the experimen-
tation.26
Tables 3 and 4 present the standard deviation, the
variance and the mean obtained from the statistical
analyses of the measured microhardness values for the
two different positions, namely, the top and bottom of
the spot welds produced with different tools and using
different process parameters.
Table 3: Mean, variance, mu and sigma of the spot samples produced
using a flat pin/flat shoulder tool and a conical pin/concave shoulder
tool for the microhardness taken on the top
Sample ID Mean 2 μ 
FPS_800_0.5 46.543 82.8442 46.5429 9.102
FPS_800_1 75.099 4994.97 4.00167 0.796
FPS_1200_0.5 45.121 441.806 3.71115 0.443
FPS_1200_1 51.731 672.845 3.83391 0.474
CCS_800_0.5 44.575 409.802 3.70341 0.433
CCS_800_1 55.887 813.763 3.90757 0.481
CCS_1200_0.5 45.453 482.889 3.71166 0.458
CCS_1200_1 40.942 248.04 3.64315 0.371
Table 4: Mean, variance, mu and sigma of the spot samples produced
using a flat pin/flat shoulder tool and a conical pin/concave shoulder
tool for the microhardness taken at the bottom
Sample ID Mean 2 μ 
FPS_800_0.5 92.234 32.7196 4.52241 0.062
FPS_800_1 84.7 119.974 84.7 10.95
FPS_1200_0.5 84.7 119.974 84.7 10.95
FPS_1200_1 78.429 58.0637 78.4286 7.62
CCS_800_0.5 89.986 14.2859 89.9857 3.78
CCS_800_1 88.564 102.307 88.5643 10.11
CCS_1200_0.5 86.252 89.6604 4.45128 0.109
CCS_1200_1 82.043 54.8519 4.40318 0.09
It should be noted that the mean is equal to μ if the
distribution is normal. It can be seen that in some cases
in the current work μ and the mean have different values,
which shows that the distributions in some of the
analyses were not normal. The standard deviation 
should be close to zero, but in the current work, the value
of the standard deviation is not close to zero in some of
the cases. This shows that the microhardness values are
not close to the expected values and this could be due to
the microhardness values measured in different locations
of the weld samples, far apart from each other.
This was further suspected due to the presence of
intermetallics, which could have been the cause of the
high microhardness values since intermetallics are
invariably hard and brittle.
Each figure contains residuals versus the distance
(the distance from the keyhole), taking into account the
data and the constant variance of the residuals. In the
plot of residuals versus distance, it is shown that the
models are adequate to predict the responses in an
acceptable manner.
4 CONCLUSIONS
According to the presented results, some conclusions
can be drawn:
• The microstructure of the produced spot welds shows
a good material mixing, the presence of a copper ring
and a mixture of Al/Cu particles present in the stir
zone.
• The EDS analyses of the produced friction-stir spot
welds exhibited the presence of intermetallic com-
pounds, which are known to affect the microhard-
ness.
• The microhardness values obtained at the top were
high for all the samples and this was in the region
close to the spot-weld keyhole. In addition, all the
microhardness values obtained at the bottom of the
samples, in the region close to the keyhole, have
lower values, which were close to the average value
of the copper base material. This occurred for all the
spot welds produced using a conical pin and concave
shoulder.
• The probability-density-function (PDF) histograms
of the microhardness results revealed that the process
parameters and the tool geometries have significant
effects on the distribution of the microhardness
values in different locations of the produced spot
welds.
• Additionally, goodness-of-fit values were also
analysed and these showed that most of the R2 values
ranged between 0.8842 and 0.9999, which is an
indication of how well the model fits with the
produced experimental data.
Acknowledgements
The authors wish to acknowledge the financial
support of the University of Johannesburg and the
assistance from Mr. Riaan Brown (eNtsa at Nelson
Mandela Metropolitan University).
5 REFERENCES
1 W. M. Thomas, E. D. Nicholas, J. C. Needham, M. G. Murch, P.
Temple-Smith, C. J. Dawes, International Patent No. PCT/GB92/
02203, GB patent application No. 9125978.8, 1991
2 H. Badarinarayan, Fundamentals of Friction Stir Spot Welding, PhD
Thesis, 2009, Missouri
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
868 Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
3 W. Yuan, R. S. Mishra, B. Carlson, R. Verma, R. K. Mishra, Material
flow and microstructural evolution during friction stir spot welding
of AZ31 magnesium alloy, Materials Science and Engineering, A
543 (2012), 200–209
4 J. Podr`aj, B. Jerman, D. Klob~ar, Welding defects at friction stir
welding, Metalurgija, 54 (2015) 2, 387–389
5 J. M. Timothy, Friction Stir Welding of Commercially Available
Superplastic Aluminium, PhD thesis, 2008, Department of Engi-
neering and Design, Brunel University, Brunel
6 J. Ouyang, E. Yarrapareddy, R. Kovacevic, Microstructural evolution
in the friction stir welded 6061 aluminum alloy (T6-temper condi-
tion) to copper, Journal of Materials Processing Technology, 172
(2006), 110–122
7 P. Liu, Q. Shi, W. Wang, X. Wang, Z. Zhang, Microstructure and
XRD analysis of FSW joints for copper T2/aluminium 5A06
dissimilar materials, Materials Letters, 62 (2008), 4106–4108
8 P. Xue, B. L. Xiao, D. R. Ni, Z. Y. Ma, Enhanced mechanical pro-
perties of friction stir welded dissimilar Al–Cu joint by intermetallic
compounds, Materials Science and Engineering A, 527 (2010),
5723–5727
9 A. Esmaeili, M. K. Besharati Givi, H. R. Zareie Rajani, A metallur-
gical and mechanical study on dissimilar friction stir welding of
aluminum 1050 to brass (CuZn30), Materials Science and Engi-
neering, A 528 (2011), 7093–7102
10 A. Abdollah-Zadeh, T. Saeid, B. Sazgari, Microstructural and mecha-
nical properties of friction stir welded aluminum/copper lap joints,
Journal of Alloys and Compounds, 460 (2008), 535–538
11 E. T. Akinlabi, Characterisation of Dissimilar Friction Stir Welds
between 5754 Aluminium Alloy and C11000 Copper, D-Tech thesis,
Nelson Mandela Metropolitan University, South Africa, 2010
12 M. N. Avettand-Fenoel, R. Taillard, G. Ji, D. Goran, Multiscale
Study of Interfacial Intermetallic Compounds in a Dissimilar Al
6082-T6/Cu Friction-Stir Weld, Metallurgical and Materials
Transactions A, (2012), 4655–4666
13 M. F. X Muthu, V. Jayabalan, Tool Travel Speed Effects on the
Microstructure of Friction Stir Welded Aluminium – Copper Joints,
Journal of Materials Processing Technology, 217 (2015), 105–113
14 M. P. Mubiayi, E. T Akinlabi, Friction Stir Spot Welding between
Copper and Aluminium: Microstructural Evolution, Proceedings of
the International MultiConference of Engineers and Computer
Scientists, Vol II, IMECS 2015, March 18–20, 2015, Hong Kong
15 R. Heideman, C. Johnson, S. Kou, Metallurgical analysis of Al/Cu
friction stir spot welding, Science and Technology of Welding and
Joining, 15 (2010) 7, 597–604
16 U. Özdemir, S. Sayer, Ç. Yeni, Bornova-Izmir, Effect of Pin Pene-
tration Depth on the Mechanical Properties of Friction Stir Spot
Welded Aluminum and Copper, Materials Testing in Joining Techno-
logy, 54 (2012) 4, 233–239
17 M. Shiraly, M. Shamanian, M. R. Toroghinejad, M. Ahmadi Jazani,
Effect of Tool Rotation Rate on Microstructure and Mechanical
Behavior of Friction Stir Spot-Welded Al/Cu Composite, Journal of
Materials Engineering and Performance, 23 (2014) 2, 413–420
18 M. P. Mubiayi, E. T. Akinlabi, Evolving properties of friction stir
spot welds between AA1060 and commercially pure copper C11000,
Transactions of Nonferrous Metals Society of China, 26 (2016),
1852–1862
19 E. T Akinlabi, S. A. Akinlabi, Friction stir welding of dissimilar
materials – statistical analysis of the weld data, Proceedings of the
International MultiConference of Engineers and Computer Scientists,
2012, 1368–1373
20 ASTM Standard E3-11, Standard guide for preparation of
metallographic specimens, ASTM International, West Consho-
hocken, PA, 2011, doi: 10.1520/E0003-11, www.astm.org
21 L. E Murr, A Review of FSW Research on Dissimilar Metal and
Alloy Systems, Journal of Materials Engineering and Performance, 8
(2010)19, 1071–108
22 J. M. Schneider, M. Bigerelle, A. Iost, Statistical analysis of the
Vickers hardness, Materials Science and Engineering, A 262 (1999),
256–263
23 A. L. Yurkov, N. V. Jhuravleva, E. S. Lukin, Kinetic microhardness
measurements of sialon-based ceramics, Journal of Materials
Science, 29 (1994), 6551–6560
24 I. Y. Yanchev, E. P. Trifonova, Analysis of microhardness data in Tlx
ln l-x Se, Journal of Materials Science, 30 (1995), 5576–5580
25 A. M. Hassan, M. Almomani, T. Qasim, A. Ghaithan, Statistical
analysis of some mechanical properties of friction stir welded
aluminium matrix composite, Int. J. Experimental Design and
Process Optimisation, 3 (2012)1, 91–109
26 K. Palanikumar, R. Karthikeyan, Optimal machining conditions for
turning of particulate metal matrix composites using Taguchi and
response surface methodologies, Machining Science and Technology,
10 (2006) 4, 417–433
M. P. MUBIAYI et al.: MICROSTRUCTURE EVOLUTION AND STATISTICAL ANALYSIS OF Al/Cu ...
Materiali in tehnologije / Materials and technology 51 (2017) 5, 861–869 869
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS