INVESTIGATION OF PROCESS PARAMETERS FOR T-JOINT

6061-T6 ALUMINUM FRICTION STIR WELDING USING

NANOCOMPOSITES MATERIAL
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

F. F. Mustafa, S. R. Daham
UNIVERSITY OF BAGHDAD

ABSTRACT

In this investigation the effect of four parameters, transverse speed, type of nano-
powders, rotational speed, and groove’s depth in friction stir welding (FSW) T-joints
for aluminum alloy 6061-T6, 3mm plate thickness were taken into consideration.
The process parameters optimized by taguchi technique based on L9 orthogonal
array (OA). The optimum welding process parameters were predicted, and their
percentage of contribution was estimated by applying the analysis of variance
(ANOVA) and signal to noise ratio (S/N) techniques. The optimum process
parameters are determined depending on tensile test in skin and stringers
direction, bending test, and hardness test of the joint. Optical microscope (OM) and
scanning electron microscope (SEM) analysis were used to verify the microstructure
INTERJOIN

and dispersion of nano-powders in welding joint. The optimum conditions of 1550

rpm rotational speed, 15 mm/min transverse speed, Al 2O3 type of powder and 1
mm groove’s depth obtained from the results of ultimate tensile strength of the
skin welded part (UTSskin), which gives the best ultimate tensile stress of 177 MPa.
SEM Micrographic for metal matrix nanocomposite (MMNCs) of all nine experiments
revealed that the nano particles are irregular dispersed in nugget zone due to the
one pass of FSW. The rotational speed of 960 rpm, transverse speed of 15
mm/min, type of powder TiO, and groove’s depth of 1.5 mm gives the greater
hardness value of 80 HV in nugget zone. Analysis of variance shows that the
groove’s depth is the most significant parameter in this investigation.

KEYWORDS: FSW T–joint, AA6061–T6 aluminum alloy, taguchi orthogonal array

(OA), (S/N) ratio, ANOVA.

1. INTRODUCTION

Friction stir welding is a relatively new solid-state joining technique developed

initially by Thomas et al. in 1991 at The Welding Institute (TWI) of the United
Kingdom, for welding aluminum alloys that were difficult to weld by using
traditional welding techniques [1]. Aluminum alloy 6061-T6 is widely utilized in
aircraft, defense, automobiles and marine areas due to their good strength, light
weight and better corrosion properties. But, they exhibits inferior tribological
properties in extensive usage [2,3]. In addition, aluminum based composites
become brittle by the addition of reinforcements such as SiC and Al2O3 ceramic
particles [4]. A proper technique can be employed to refine the microstructure and
homogeneous dispersion of reinforcements on metallic surface as since wear is a
surface deprivation property [5]. Dispersion of reinforcement particles on metal
surface and the control of its dispersal are more difficult to attain by conventional
 surface modification techniques [6]. One of the important features of T-joints
components is that the stiffness and strength of skins can be remarkably reinforced
by stringers without a significant increase in weight [7].

The FSW of T-joints as shown in Fig. 1 is obtained by inserting a specially designed

rotating pin into the clamped blanks and then moving it all along the joint. The pin
is inserted with a tilt angle limiting the contact between the tool shoulder and
blank. As the pin is inserted into the upward sheet (skin), the blank material
undergoes to a local backward and forward extrusion process up to penetrate the
vertical blank (stringer) and the tool shoulder becomes contacted with the upward
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

sheet. Then, the tool rotation determines an increase of the material temperature
due to the friction forces and material deformation. As the tool is moved along the
joint, the two blank materials are softened (no melting is observed) and mixed to
perform the weld joint. [8].

Tavares et al. [9] studied the effects of process parameters such as transverse
speed, vertical downward force, and rotational speed on both T-joint and butt joint,
two types of Al alloys were used in this work. The skin plate is made of Al 6056-T4,
and the stringers plate is made of Al 7000 series. Bending test, the fatigue test,
and microhardness were done on all experiments. Lima et al. [10] studied the
influence of processing parameters on particle dispersal, and hardness were
examined in Al-alloy strengthened with multi-walled carbon nanotubes by FSP. TEM
and SEM proved that nanotubes were entrenched in the lamellae zones of the Al-
alloy stir region and their multi-walled was saved. Shafiei-Zarghani et al. [11] a
new technique has been used to utilize the FSP to induce the nano-sized Al2O3 into
6082 Al alloy to create a composite surface layer. Samples were used in multiple
number of FSP passes from one to four, with and without Al2O3 powder. Azizieh et
at. [12] studied the effects of rotational speed, pin profile and the number of FSP
INTERJOIN

passes on the distribution of nanopowders and matrix microstructure. The grain

refinement of matrix and enhanced distribution of nanopowders were achieved after
each FSP pass. Friction stir process was utilized to produce AZ31/Al2O3
nanocomposites for surface applications. Essam et al. [13] studies the effected of
reinforced composites with mixtures of SiC and Al2O3 particles, 1.25μm in average
size have been fabricated on an A 1050-H24 aluminum plate of 5 mm thickness by
friction stir processing FSP and their wear resistance as a function of relative weight
ratios of the particles. Salehi et al. 2011 [14] studies the used of Design of
experiment (DOE) to deciding the most significant determinant which effect
ultimate tensile strength (UTS) of AA6061/SiC nanocomposites created by FSP. The
influence of four determinant, including transverse speed, rotational speed, pin
profile and tool penetration depth on the ultimate tensile stress was examined by
Taguchi method to specified the best process parameters.

The aim of the present paper is to combine between welding T-joint sections and
create MMNCs in welding region ssimultaneously, and study the effect of transvers
speed, rotational speed, types of nano-powder, and groove’s depth on the
mechanical properties, hardness, and microstructure of the nugget zone.
 III International Congress & 21st Technical Sessions on Welding and Joining Technologies

Fig. 1. The clamping fixture of FSW T-joints process

2. EXPERIMENTAL PROCEDURE

In this study, the T-Joint of AA6061-T6 aluminum alloys with a chemical

composition that listed in Table 1, was performed into the skins and stringers to the
required size of (200×70×3) mm and (200×30×3) mm, respectively, as shown in
Fig. 2. Furthermore, the contact surfaces of blanks were polished to avoid some of
the negative influences of the oxidation film during the welding process. A groove
was fabricated along and on the center of the skin plates with a width of 1.6 mm
and variables depth of 0.5, 1, and 1.5 mm. The tool were machined out of HSS
(AISI M2). They consisted of a shoulder with a diameter of 20 mm, and columnar
INTERJOIN

probe with a diameter of 2.5 mm and length of 4.2 mm. Three types and attributes
of commercially available nanoparticles were used and the details of these powders
are given in Table 2. To prevent the nanoparticles from scattering out the groove
during the FSW, the groove was covered firstly with a modified FSP tool that only
had a same shoulder diameter without any probe. The conventional “Turret Milling
Machine Model: MDM 4VS/4HS/4 S” and its fixture was developed to fix the stringer
under the skin for perform the FSW operation. The tensile test is carried out on a
servo-controlled hydraulic universal testing machine at a load rate of 1.5 KN/min
with 1 mm/min crosshead speed to undergo the deformation until the failure
acquires. Two different kinds of tensile test (Hoop stress test and T-pull test) were
used to evaluate the ultimate tensile strength across skin and stringer of welded
part as shown in Fig. 3.

The microstructure of the surface welded T-joints were examined and observed
carefully normal to the FSW, mechanically polished with diamond metallographic
polishing oil, and etched with Keller reagent (5% HNO3, 3%HCL, 2%HF, and
190%H2O for each 200 ml) for about 45 sec, by (Nikon TECHNO) optical
microscopic (OM), and (TESCAN) Scanning Electron Microscopy (SEM).

The micro-hardness values was measured by a micro-Vickers hardness tester using

a load of 200g for 10 sec. The hardness distribution was measured along two
perpendicular lines through the mid-plane of the polished cross-sections of the
nugget zone (NZ) with a separated distance of 0.25mm.
 III International Congress & 21st Technical Sessions on Welding and Joining Technologies

Fig. 2. (a) Skin plate, (b) Stringer plate, and (c) The groove in the center of the
skin

Types of powder Attributes

Aluminum Oxide Al2O3 Alpha, High Purity 99.9% Size 135 nm
Titanium Oxide TiO2 Rutile, 96%, coated with silicon Size 30 nm
Silicon Carbide SiC Beta, 99%, cubic Size <80 nm

Table (2) Details of nano- powders

INTERJOIN

Fig. 3. (a) Specimen tensile test, (b) Hoop test (c) T-pull test and (d) Clamping

3. DESIGN OF EXPERIMENTAL BASED ON TAGUCHI’S METHODS

3.1 Selection of FSW Parameters and the OA Experiment

Trial experiments were conducted by varying the four parameters (rotation speed,
transverse speed, type of powder, and groove’s depth), with three levels for each
parameter are chosen to study the effect of FSW T-joints on 3mm AA6061-T6
aluminum alloy plates, on the other hand keeping the tilt angle of 2.5 o and the
plunge depth of 0.3mm as a constant parameters. The selected parameters for the
FSW process are listed in Table 3.

Taguchi’s method is very effective to deal with responses influenced by many

parameters. It is a powerful design of experiments tool which reduces drastically
the number of experiments that are required to model and optimize the responses
[15]. The number of process parameters considered under this study is four and
the level of each factor is three. Hence, L9 (3 4) orthogonal array is selected to
perform the most effective experiments (9 experiments from the overall 81
experiments). This array has nine rows and four columns and each row represents
 a trial condition while each column accommodates a specific process parameter
[16, 17]. Table 4. Shows the experimental design based on the OA L9.

3.2 S/N Ratio Technique

In Taguchi method, the experimental values of various responses are further

transformed to signal to noise (S/N) ratio. The S/N ratio was chosen according to
the criterion (larger is better) to maximize the response and to measure deviation
of the response from the mean value. The S/N ratio and the mean squared
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

deviation (MSD) are calculated by using the statistical software (MINITABTM®17),

using Eqs. 1 and 2, [18, 19].

MSD ( ∑ ( )) (1)

(S/N )Ratio ( ∑ ( )) (2)

Where: n = Number of trials, i = Trial number, and j

= Experiment number.

3.3 ANOVA Technique

Analysis of variance (ANOVA) examination is utilized to detect the significance of

INTERJOIN

welding parameters are statically. Where the aim of ANOVA examination is to

inspect the significance of the FSW parameters which affect the mechanical
strength properties. It is gives a clear picture to show how far the process
parameters affects the response and the level of their significance [20].

4. RESULT AND DISCUSSIONS

4.1 Analysis of UTSSkins

The UTS (MPa) from hoop stress test of the standard specimens is considered to
describe the quality of the welded parts. The results have been analyzed by using
the S/N ratio method to give the optimal level for each process parameter that
corresponded to the larger S/N ratio and ensured the maximum value of tensile
strength during the FSW process. For each experiment, the mean of the tensile
strength results (UTS Mean) and S/N ratio are given in Table 4. The main effects of
each parameter at three different levels on the mean and S/N ratio for UTS Skins are
calculated and given in Table 5. The optimum level of parameters that corresponds
to a larger value for mean and S/N ratio is compatible, so that the optimum
conditions will be at rotational speed (level 2) =1550 rpm, transverse speed (level
2) =11 mm/min, types of powder (level 1) =AL2O3, and groove’s depth (level 2)
=1 mm. The variance (Delta) between the greater and lesser value gives the
influence of each factor. In this case, the best important factor that has an
influence on UTSSkins for welded parts is groove’s depth followed by rotational
speed, types of powder, and transverse speed. The predicted values of mean are
computed using Eq. 3, and compered with the obtained values as shown in Table 6.

In order to find the effect of process parameters on various responses, ANOVA is

performed and the calculated F-values for various responses determine the relative
 significances of different process parameters. The value of F test which gives the
significant factors that affected on quality characteristic of the welded parts are
arranged as the groove’s depth that have larger value, and followed by rotational
speed, type of powder, finally transverse speed as shown in ANOVA skin chart Fig.
4.

Exp. A B C D S/N Mean S/N Mean S/N Mean

skin skin stringer stringer Hardness Hardness
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

1 960 6 Al2O3 0.5 44.2021 162.5 39.4129 93.5 35.75 61.36

2 960 11 SiC 1 44.2359 163 39.9561 99.5 35.89 62.71
3 960 15 TiO 1.5 41.9015 124.5 38.1033 80.5 37.69 77.04
4 1550 6 SiC 1.5 42.0318 126.5 37.1240 72.0 35.33 59.63
5 1550 11 TiO 0.5 44.9348 176.5 39.8232 98.0 36.36 66.43
6 1550 15 Al2O3 1 44.9578 177 40.2555 103.0 36.83 69.92
7 2000 6 TiO 1 43.8838 156.5 40.4771 106.0 35.48 60.03
8 2000 11 Al2O3 1.5 42.1684 128.5 35.7560 61.5 35.48 59.52
9 2000 15 SiC 0.5 43.6916 153 37.0226 71.0 36.57 67.81

Table 4. Experimental design based on the OA L9 and experimental results

(Mean and S/N ratio) of hoop stress, T-pull stress and HV.
INTERJOIN

S/N ratio for Skin response S/N ratio for Stringer response S/N ratio for Hardness
Levels A B C D A B C D A B C D
1 43.45 43.37 43.78 44.28 39.16 39.00 38.47 38.75 36.45 35.53 36.03 36.23
2 43.97 43.78 43.32 44.36 39.07 38.51 38.03 40.23 36.18 35.91 35.93 36.07
3 43.25 43.52 43.57 42.03 37.75 38.46 39.47 36.99 35.85 37.04 36.52 36.17
Delta 0.73 0.41 0.46 2.33 1.41 0.54 1.43 3.24 0.60 1.51 0.58 0.16
Rank 2 4 3 1 3 4 2 1 2 1 3 4
Means for Skin response Means for Stringer response Means for Hardness
Levels A B C D A B C D A B C D
1 150.0 148.5 156.0 164.0 91.17 90.50 86.00 87.50 67.04 60.34 63.60 65.20
2 160.0 156.0 147.5 165.5 90.00 86.33 80.83 102.83 65.33 62.89 63.39 64.22
3 146.0 151.5 152.5 126.5 79.50 84.83 94.83 71.33 62.46 71.59 67.84 65.46
Delta 14 7.5 8.5 39.0 11.67 5.67 14.00 31.50 4.59 11.25 4.45 1.18
Rank 2 4 3 1 3 4 2 1 2 1 3 4

Table 5. The main effects of means and S/N ratio for UTSskin, UTSstringer, and HV.

Quality predicted condition predicted value Obtained condition Obtained value

UTSskin A2, B2,C1, D2 181.5 MPa A2, B3,C1, D2 177 MPa
1550 rpm, 11 mm/min, 1550 rpm, 15 mm/min,
Al2O3, and 1 mm. Al2O3, and 1 mm.
UTSstringer A1, B1, C3,D2 117.667 MPa A3, B1, C3,D2 106 MPa
960 rpm, 6 mm/min, 2000 rpm, 6 mm/min,
TiO, and 1 mm. TiO, and 1 mm.
HV A1, B3, C3, D3 77 HV A1, B3, C3, D3 77 HV
960 rpm, 15 mm/min, 960 rpm, 15 mm/min,
TiO, and 1.5 mm TiO, and 1.5 mm
Table 6. Optimum and obtained values for quality characteristics.
 4.2 Analysis of UTSStringers

The UTSStringers (MPa) from T pull test of the standard specimens is considered to
describe the quality of the welded parts. The results have been analyzed using the
same technique previously of S/N ratio method according to the criterion (larger is
better) to maximize the response. Table 4 gives the mean experimental results of
the tensile strength (UTS Mean) and S/N ratio. The main effects of each parameter at
three different levels on the mean and S/N ratio for ultimate tensile strength are
calculated and given in Table 5. The optimum condition will be at a rotational speed
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

(level 1) =960 rpm, transverse speed (level 1) =6 mm/min, types of powder (level
3) =TiO, and groove’s depth (level 2) =1 mm. Table 6 gives the mean of the
predicted and obtained values and their conditions.

The value of F exam is also used to define which process factors have an important
influence on quality characteristic of the welded parts. It appears the significant
factors are the groove’s depth that have larger value, followed by type of powder,
rotational speed and finally transverse speed as shown in ANOVA stringer chart of
Fig. 4.

4.3 Hardness

Figure 5 show the hardness profile along skins and stringers plate. For the stringer
the hardness taken from the top of the skin plate to the end of the stringers. The
shape of hardness profile was appear as “N” shape due to due uneven distribution
of the welding temperature through skin thickness. The maximum value of the
hardness in welding zone was appear far away 1.5 mm from the top surface, where
the agglomeration and distribution of nanopowders concentration in this region. “w”
shape of the hardness profile along the skin plate was appear as shown in Fig. 5, it
INTERJOIN

can be seen that the hardness reach the high value at point 1.75 mm from the
center NZ (experiment No. 3), and as going farther from this point the hardness
decreases till it reaches the minimum value at TMAZ and then increases again up to
unaffected zone. It indicates enhanced hardness inside the composite zone and a
softened area surrounding it. The increase in hardness were not significantly high
because improper dispersion of nanoparticles in the matrix, related to the one pass
of FSW.

The same technique (S/N) ratio method was used to analyze the results and
obtained the optimal level for each process parameter to ensure the maximum
value of microhardness during the FSW process. The microhardness results for the
three trials, mean, and (S/N) ratio are given in Table 4, and the main effects of
each parameter at three different levels on the mean and (S/N) ratio for
microhardness are calculated and given in Table 5. So that the optimum condition
of main at (rotational speed (level 1) =960 rpm, transverse speed (level 3) =15
mm/min, types of powder (level 3) =TiO, and groove’s depth (level 3) =1.5 mm).
Table 6 gives the mean of the predicted and obtained values and their conditions.
From ANOVA hardness chart Fig. 4, the significant factors that affected on quality
characteristic of the welded parts are the transverse speed that have larger value,
and followed by type of powder, rotational speed, finally groove’s depth.
 III International Congress & 21st Technical Sessions on Welding and Joining Technologies
INTERJOIN

Fig. 4. Contribution effect of process parameters for Skin response, Stringer

response, and Micro-Hardness.

4.4 Microstructure

The aim of this examination is to identify the distribution of nanopowders in nugget

zone and the amount of embedded in the matrix, also to understand the effect of
process parameters on the making of the MMNCs. After one pass of FSW with
different relative contents of nanopowders, the SEM micrograph appearance of
nuggets zone for the nine experiments are shown in figure (6). From this figure the
aggregation of nanopowder could be observed in the MMNCs and the nanoparticle
distribution in all the samples fabricated by this process parameters is irregular due
to the one pass of the FSW. A similar effect of more than one pass on the cluster
particles in the MMNCs has been reported by other researchers [21- 24] in which
they found the distribution of nanoparticles have been uniform related to more one
pass of FSW.
According to the result of the UTS skin, the experiment number (6) has adopted as
the best result from the nine experiments. The optical micrographs of the surface
 hybrid composites in three zones and its base metal are shown in Fig. 7. It is
obvious that the grain size distributions within the nugget zone is quite fine grains
than the other regions, were the grains in the stir zone experienced severe plastic
deformation and recrystallization. Six random samples chosen to measure their
affected area, the average area of the base metal and nugget zone are 31116.97
μm², and 17406.97 μm² respectively.

5. CONCLUSIONS
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

In this work, T- joints that fabricated from different welding conditions and the
influence of reinforcement such as Al2O3, SiC and TiO on the mechanical properties
of Al alloy 6061-T6 were investigated. The following conclusions are obtained.

 The addition of nano-powders particles in the region of FSW will

improve the mechanical properties and hardness, but the
agglomeration of the powder in this region reduce the increment of
these properties.
 Tensile properties at optimum condition (i.e. A2B2C1D2 for skin and
A1B1C3D2 for stringer) are less as compared to the base material due
to presence of reinforcement particles which make the matrix brittle.
 Hardness value in the nugget zone is higher than other region, but
remained less than the base metal, due to the cluster of the nano-
powders in welding region and non-uniform distribution.
 The dispersion of nano-powders particles in Al matrix cannot be
uniformly distributed by using one pass of FSW.

REFERENCES
INTERJOIN

[1]. Thomas, W., Nicholas, E., Needham, J., Murch, M., Smith, P., and
Dawes, C., 1995, “Friction Stir Butt Welding,” GB Patent Application
No. 9125978.8, and U.S. Patent No. 5,460,317.
[2]. Bakes H, Benjamin D, Kirkpatrick CW. Metal’s handbook, vol. 2. ASM,
Metals Park: OH; 1979; p. 3–23.
[3]. Ravi N, Sastikumar D, Subramanian N, Nath AK, Masilamani V.
Microhardness and microstructure studies on laser surface alloyed
aluminium alloy with Ni–Cr. Mater Manuf Process 2000;15:395–404.
[4]. Clyne TW, Withers PJ. An introduction to metal matrix composites.
Cambridge: Cambridge University Press; 1993.
[5]. Rabinowicz E. Friction and wear of materials. New York: JohnWiley and
Sons; 1965.
[6]. Budinski KG. Surface engineering for wear resistance. New Jersey:
Prentice-Hall; 1988.
[7]. Zhou, G., Yang, X., Cui, L., Zhang, Z., and Xu, X., 2012, “Study on the
Microstructures and Tensile Behaviors of Friction Stir Welded T-Joints
for AA6061-T4 Alloys,” J. Mater. Eng. Perform. 21(10), pp. 2131–2139.
[8]. Mustafa, Faiz F., Ali H. Kadhym, and Hiba H. Yahya. "Tool Geometries
Optimization for Friction Stir Welding of AA6061-T6 Aluminum Alloy T-
Joint Using Taguchi Method to Improve the Mechanical
Behavior." Journal of Manufacturing Science and Engineering 137.3
(2015): 031018.
[9]. Tavares, S. M. O., et al. "Friction Stir Welding of T-Joints in Dissimilar
Aluminium Alloys." ASME 2008 International Mechanical Engineering
Congress and Exposition. American Society of Mechanical Engineers,
2008.
[10]. Lim, D. K., T. Shibayanagi, and A. P. Gerlich. "Synthesis of multi-
walled CNT reinforced aluminium alloy composite via friction stir
 processing." Materials Science and Engineering: A 507.1 (2009): 194-
199.
[11]. Shafiei-Zarghani, A., S. F. Kashani-Bozorg, and A. Zarei-Hanzaki.
"Microstructures and mechanical properties of Al/Al 2 O 3 surface
nano-composite layer produced by friction stir processing." Materials
Science and Engineering: A 500.1 (2009): 84-91.
[12]. Azizieh, M., A. H. Kokabi, and P. Abachi. "Effect of rotational speed and
probe profile on microstructure and hardness of AZ31/Al 2 O 3
nanocomposites fabricated by friction stir processing." Materials &
Design 32.4 (2011): 2034-2041.
III International Congress & 21st Technical Sessions on Welding and Joining Technologies

[13]. Mahmoud, Essam RI, et al. "Wear characteristics of surface-hybrid-

MMCs layer fabricated on aluminum plate by friction stir
processing." Wear 268.9 (2010): 1111-1121.
[14]. Salehi, M., M. Saadatmand, and J. Aghazadeh Mohandesi.
"Optimization of process parameters for producing AA6061/SiC
nanocomposites by friction stir processing." Transactions of Nonferrous
Metals Society of China 22.5 (2012): 1055-1063.
[15]. A. Devaraju, A. Kumar, A. Kumaraswamy, B. Kotiveerachari. “Influence
of reinforcements (SiC and Al2O3) and rotational speed on wear and
mechanical properties of aluminum alloy 6061-T6 based surface hybrid
composites produced via friction stir processing” Materials and Design
51 (2013) 331–341.
[16]. Vidala, C., Infante, V., Pecas, P., and Vilaca, P., 2013, “Application of
Taguchi Method in the Optimization of Friction Stir Welding Parameters
of an Aeronautic Aluminum Alloy,” Int. J. Adv. Mater. Manuf. Charac.,
3(1), pp. 21–26.
[17]. Lakshminarayanan, A., and Balasubramanian, V., 2008, “Process
Parameters Optimization for Friction Stir Welding of RDE_40 Aluminum
INTERJOIN

Alloy Using Taguchi Technique,” J. Trans. Nonferrous Metals Soc.

China, 18(3), pp. 548–554.
[18]. Roy, R., 1990, A Primer on the Taguchi Method, 1st ed., Van Nostrand
Reinhold, New York.
[19]. Minitab Company, “MINITABTM Statistical Software Release 16,”
Minitab Inc., State College, PA, http://www.minitab.com
[20]. Shojaeefard, Mohammad Hasan, et al. "Application of Taguchi
optimization technique in determining aluminum to brass friction stir
welding parameters."Materials & Design 52 (2013): 587-592.
[21]. Shafiei-Zarghani, A., S. F. Kashani-Bozorg, and A. Zarei-Hanzaki.
"Microstructures and mechanical properties of Al/Al 2O3 surface nano-
composite layer produced by friction stir processing." Materials Science
and Engineering: A 500.1 (2009): 84-91.
[22]. Azizieh, M., A. H. Kokabi, and P. Abachi. "Effect of rotational speed and
probe profile on microstructure and hardness of AZ31/ Al 2O3
nanocomposites fabricated by friction stir processing." Materials &
Design 32.4 (2011): 2034-2041.
[23]. Guo, J. F., et al. "Effects of nano- Al2O3 particle addition on grain
structure evolution and mechanical behaviour of friction-stir-processed
Al." Materials Science and Engineering: A 602 (2014): 143-149.
[24]. Shahi, A., et al. "In situ formation of Al–Al3Ni composites on
commercially pure aluminium by friction stir processing." The
International Journal of Advanced Manufacturing Technology 75.9-12
(2014): 1331-1337.