*Corr. Author’s Address: Delhi Technological University, Department of Mechanical Engineering, Delhi, India, anmolbhatia@ncuindia.edu 311
Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321 Received for review: 2021-03-31
© 2021 Journal of Mechanical Engineering. All rights reserved.  Received revised form: 2021-05-16
DOI:10.5545/sv-jme.2021.7203 Original Scientific Paper Accepted for publication: 2021-05-21
Process Parameters Optimization for Maximizing Tensile Strength 
in Friction Stir-Welded Carbon Steel
Bhatia, A. – Wattal, R.
Anmol Bhatia
1,2,
* – Reeta Wattal
2
1
The NorthCap University, Department of Mechanical Engineering, India 
2
Delhi Technological University, Department of Mechanical Engineering, India
The present study focuses on improving the ultimate tensile strength of friction stir welded carbon steel (AISI 1018). The effect of the process 
parameters (welding speed, tool RPM, and shoulder diameter) on the response parameters (ultimate tensile strength, percentage elongation 
and percentage reduction in area) were studied. Response surface methodology was used to develop the mathematical model for response 
parameters, and the adequacy of the model was checked using analysis of variance (ANOVA). The welding speed and tool RPM were found 
to affect the ultimate tensile strength significantly. The percentage elongation was affected only by welding speed. The percentage reduction 
in the area was affected by welding speed and shoulder diameter. The microstructure and microhardness of the weld have been studied and 
reported in the study. 
Keywords: friction stir welded steel, ANOVA, response surface methodology, ultimate tensile strength, microstructure and 
microhardness
Highlights
• 	 In the present study, experimental analysis was done on friction stir-welded AISI1018 carbon steel joints to determine the effect 
of process parameters on response parameters (i.e., ultimate tensile strength, percentage elongation and percentage reduction 
in area) by developing mathematical models using the response surface methodology. 
• 	 The developed regression models proved to be adequate for determining the output response on friction stir-welded steel joints 
at a 95 % confidence interval. 
• 	 The surface plots are also presented in the study, which illustrates the individual effect of process parameters on the mechanical 
properties of the friction stir-welded steel joint.  
• 	 The microstructure and microhardness of the friction stir-welded steel joint under optimized condition are also presented, and 
the observable pattern is presented in detail. 
0  INTRODUCTION
Steel is the most popular choice of engineers 
because it has a good combination of mechanical 
properties and is the most widely used material for 
industrial applications. Steel is generally welded 
with a fusion-welding process that leads to various 
problems, including porosity, hot cracking, hydrogen 
embrittlement, microstructural changes in the heat-
affected zone, etc. [1]. The reason for the occurrence 
of these defects is mainly due to the inherent melting 
of base metal. 
In order to overcome these defects, the friction 
stir-welding process is considered the best choice as 
it does not involve the melting of workpiece; thus, the 
chance of defects arising due to melting are eliminated. 
The process was patented by The Welding Institute, 
U.K. [2]. The setup consists of a non-consumable 
tool having a shoulder through which a small pin is 
extended. The plates that are to be welded are rigidly 
fixed, and the rotating tool is plunged within the 
joint of plates to the extent the shoulder touches the 
plate surface, resulting in the generation of heat. This 
localized heating results in softening the material 
around the pin. The material moves from the front 
to the back of the pin because of the rotational and 
translational movements of the tool; in this manner, a 
solid-state joint is produced [3]. 
The initial work on friction stir welding was 
done on aluminium and its alloys, as they have a low 
melting point [4] to [6]. Many aluminium alloys like Al 
7075, Al 6063 [7] to [9] were the popular choice of the 
researchers. Then, the microstructural examinations 
were carried out on those alloys [10] and  [11]. Later, 
the process parameter optimization studies were 
reported on friction stir-welded aluminium and its 
alloys [12] and [13]; other studies including corrosion 
analysis and tensile strength optimization were also 
reported for aluminium [14] to [16] and magnesium 
alloys [17] but nowadays the research focus is on 
high-temperature materials. A feasibility study related 
to friction stir welding of mild steel was initially 
presented by Linert et al. [18], in which defect-free 
welds were successfully produced, but excessive tool 
wear was reported; the paper emphasized the need to 
work on the geometry of the tool in order to eliminate 
the issue of tool wear. The study demonstrated that 
the friction stir-welding of mild steel is feasible 

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
312 Bhatia, A. – Wattal, R.
without the loss of tensile properties. Ueji et al. [19] 
butt-welded ultra-fine plain low carbon steel with the 
help of a tungsten carbide tool. The paper reported the 
microstructural characterization and microhardness of 
the joint, but no mechanical testing was performed on 
the weld. 
Thereafter, a study on friction stir-welded 1012 
and 1035 carbon steels was conducted by Fujii et al. 
[20] using a tungsten carbide tool; the paper presented 
the effect of welding speed on the ultimate tensile 
strength of the weld nugget; mechanical properties 
and microstructural examination were also presented. 
An effort was then made to weld high carbon steel 
using FSW by Cui et al. [21]; the study was conducted 
on AISI 1070 material using a tungsten carbide tool. 
Metallurgical examinations were reported in the paper 
with an enhanced joint strength than the base metal. 
The effect on friction stir-welding conditions on the 
properties and microstructure was reported by Miles et 
al. [22]; the paper concluded that desired mechanical 
properties can be achieved by adjusting the process 
parameters. Later, a study was conducted by Ghosh 
et al. [23] in which the effect of process parameters 
was analysed on friction stir-welded medium carbon 
steel; the paper concluded that the strength of the 
stir zone was higher than the parent metal, and HAZ 
was reported to be the weakest section. Also, the 
ductility of the weld was found to be lower than the 
base metal. Lakshminarayanan et al. [24] conducted 
friction stir welding on 5 mm thick AISI 1018 
plates using a tungsten carbide tool and reported an 
increased tensile strength value as compared to the 
base metal. The microstructural examination was 
also reported in the paper in which smaller pearlitic 
structures were observed in the weld region. Ghosh 
et al. [25] performed experimental analysis on lap-
welded friction stir-welded M190 steel sheets. The 
variable parameters were kept as rotational and 
transverse speeds with the range of 600 mm/s to 1200 
mm/s and 0.85 mm/s to 3.39 mm/s, respectively. The 
microstructural examination revealed three regions: 
the nugget and two heat affected zones (HAZ-1 and 
HAZ-2) in which HAZ-1 was reported the weakest 
region constituting ferrite pearlite structures, while 
HAZ-2 contains martensite along with the increased 
value of hardness. The authors also reported that 
the best process parameter window to obtain joint 
efficiency of more than 60 % was 1000 rpm tool 
rotational speed and 1.69 to 3.39 transverse speed. 
Yabuuchi et al. [26] analysed the effect of rotational 
speed on the microstructure and tensile strength of 
friction stir-welded oxide dispersion-strengthened 
steel. The investigations were carried out on the speed 
ranging from 250 rpm to 400 rpm. The study revealed 
that the hardness was reduced when welding was 
carried out at all the speed because of the occurrence 
of recrystallisation. The change in ultimate tensile 
strength was also reported similar to the change in 
hardness, meaning that the joint strength was also 
reduced by grain growth. Karami et al. [27] analysed 
the effect of the tool rotational speed and transverse 
speed on microstructure and tensile behaviour of 
friction stir welded mild steel. A tunnel defect was 
observed in the specimens with lower welding speed 
or higher rotational speeds as the amount of heat input 
was not sufficient to provide enough flowability. 
The microstructural study indicated that the stir zone 
or HAZ had an austenitic structure and, on cooling, 
transforms to ferritic and pearlitic structures. The 
study also revealed that all the specimens exhibit 
high yield strength and a lower value of uniform 
elongation because of the fine microstructure in stir 
zone and heat affected zone of the welded specimen. 
The microstructural and tensile behaviour of friction 
stir-welded trip steel were studied by Mironov et al. 
[28], who reported that the thermal effect of welding 
gave rise to thermal softening of the material in HAZ 
and that martensite formation occurred in the stir 
zone. A premature failure was also reported during the 
tensile test, because of the non-homogeneous strain 
distribution that occurred because of microstructural 
changes. Mahmoudiniya et al. [29] applied a friction 
stir-welding technique to weld DP700 steel at 
different rotational speeds 600 rpm, 800 rpm, and 
1000 rpm. The authors evaluated the microstructural 
and mechanical properties of the weld; they reported 
the occurrence of ferrite bands in the stir zone at 600 
rpm, which lead to reduced strength. For the condition 
of 800 rpm and 1000 rpm, joint failure occurred at 
softened HAZ. The maximum joint strength was 
reported for the rotational speed of 800 rpm. Lee et 
al. [30] studied the friction stir welded medium Mn 
steel using a tungsten carbide tool. The paper reported 
an increase in ultimate tensile strength and total 
elongation because of the increased strain hardening 
rate. 
Although few studies were conducted to analyse 
the effect of process parameters on microstructure 
and mechanical properties for friction stir-
welded steels, there is a gap in understanding the 
mechanical properties and concerned microstructural 
characterization of the friction stir-welded mild 
steel. Therefore, an attempt is made to establish 
empirical relationships to determine the effect of 
tensile strength, percentage elongation and percentage 
reduction in area for a friction stir-welded AISI 1018 

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
313 Process Parameters Optimization for Maximizing Tensile Strength in Friction Stir-Welded Carbon Steel
joint. Analysis of variance was also incorporated to 
check the adequacy of the developed model. Thus, 
the objective of the present study is to investigate 
the effects of process parameters on output response 
parameters such as tensile strength, percentage 
elongation and percentage reduction in area. RSM is 
also incorporated to optimize the process parameters 
for tensile strength.   
1  MATERIAL AND METHODS
A butt-welded joint was produced on friction stir-
welding machine, which is developed by R.V . Machine 
Tools, India. Table 1 shows the specifications of the 
machine. The material chosen for the experimental 
analysis was AISI 1018, with a thickness of 3 mm. The 
steel plate was 200 mm in length, 40 mm in breadth, 
and 3 mm thick. The composition and mechanical 
properties of the material is illustrated in Tables 
2a and b. The friction stir-welding was done with a 
tungsten carbide tool with 7 % cobalt. The sectional 
view of the friction stir-welded plate with the start and 
stop position of the tool is illustrated in Fig. 1. The 
composition of the tool used in the present study was 
made with EDAX analysis and is presented in Fig. 2a.  
Fig. 1.  Photograph of friction stir-welded joint
The dimensions of the tool used are presented in 
Fig. 2b. The profile of the tool pin used is presented 
in Fig. 2c. The length of the pin was kept as 2.6 mm 
so that its proper penetration takes place in the plates. 
A 1.5-degree tool tilt angle was kept throughout the 
experimentation, as shown in Fig. 2d. From the 
literature review, the process parameters chosen for 
the present study were welding speed, tool rpm, and 
shoulder diameter. These parameters are the major 
contributors in generating heat and affecting the 
tensile strength. The selected process parameters and 
their levels are tabulated in Table 3. The range of the 
selected parameters was determined by conducting 
several trials of input parameters in which obtaining 
the defect-free weld was the key factor. This was 
carried out by changing one of factors from minimum 
to maximum, keeping the other factors as constant. 
The observations from the pilot experimentation are 
presented in Table 4. 
Table 1.  Specifications of friction stir welding machine used for 
experimentation
Spindle motor 11 kW, 1440 rpm, 440 V 3-Phase, AC 
Drive, Flange mounting (STARK)
Spindle speed  1440 rpm (max)
Spindle housing tilting Angle (–5° to +5°)
Spindle pulley type Timing Pulley
Z-axis stroke  300 mm
Z-axis thrust 3922.6 N (min)- 39226.6 N (max) 
(Adjustable in steps)
Z-axis feed rate 0 mm/min to 2000 mm/min
X-axis stroke 600 mm
Table 600 mm × 400 mm
‘T’ slot- 18 mm × 3 mm
Tool Holder ISO 40 Taper-Side lock holder
Z-axis feed force feedback Load cells of capacity 5000 kgs, with 
least count 1 kg is provided
Controller PC based control system
Table 2a.  Chemical composition of base metal (results by 
spectrometry)
Material C [wt%] S [wt%] Mn [wt%] P [wt%] Fe [wt%]
AISI 1018 Steel 0.18 0.03 0.65 0.04 Bal
Table 2b.  Mechanical properties of base metal
U.T.S. 
[MPa]
U.T.S. for reduced 
section [MPa]
Elongation 
[%]
Yield strength 
[MPa]
Reduction 
in area [%]
416 435 21.8 280 50.77
Table 3.  Chosen process parameters with their levels
Parameter chosen Notation Units
Levels
1 2 3
Welding speed V mm/min 60 110 210
Tool rpm N mm/min 430 550 750
Shoulder diameter D mm 15 18 20
2  EXPERIMENTATION
The L9 orthogonal array was used as a design matrix 
for the conduction of experiments. Transverse tensile 
test specimens were prepared as per the E-8 standard 
of ASTM [31]. Two tensile tests were conducted on the 
welds: transverse tensile testing and reduced section 
tensile testing. The former is used to locate the region 
of lowest strength in weldment and to determine the 
strength in that region; however, when the strength 
of weld nugget is to be determined then the later one 
that is reduced section tensile test is used in which 
area of the cross-section is deliberately reduced at 

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
314 Bhatia, A. – Wattal, R.
Fig. 2.  a) Results obtained from EDAX analysis of tool, b) dimensions of the tool, c) tool profile, and d) 1.5-degree tool tilt angle used in 
generating the weld samples
Table 4.  Observations made while selecting the range of process parameters
Variable parameter Input parameters Macrostructure Observable remarks
Welding  
speed 
Welding speed < 60 mm/min
Tool RPM = 430 mm/min
Shoulder diameter = 20 mm
Root stuck to the surface because of 
increased contact time of hot metal and 
backing plate.
Welding speed > 210 mm/min
Tool RPM = 430 mm/min
Shoulder diameter = 20 mm
Groove defect was observed on the weld 
surface. This is because of the hindrance 
between the frictional heat and the material 
flow.
Tool RPM
Tool RPM < 430 mm/min
Welding speed = 60 mm/min
Shoulder diameter = 20 mm
Excessive degradation of the tool pin was 
observed. This is due to the breakage of tool 
pin because of excessive heat generation.
Tool RPM >750 mm/min
Welding speed = 60 mm/min
Shoulder diameter = 20 mm
Tunnel defect with excessive flash was 
observed. The excessive flash occurred 
because of high rotational speed. 
Shoulder  
diameter 
Shoulder diameter < 15 mm
Welding speed = 60 mm/min
Tool RPM = 430 mm/min
Surface defect was observed. This is 
because the area between the shoulder pin 
and shoulder diameter was reduced, which 
ultimately reduces the flow of material.
Shoulder diameter > 20 mm
Welding speed = 60 mm/min
Tool RPM = 430 mm/min
Tunnel defect was observed on the welded 
specimen. This is because excessive heat 
was generated and material flow was 
inadequate.

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
315 Process Parameters Optimization for Maximizing Tensile Strength in Friction Stir-Welded Carbon Steel
weld nugget; this ensures that the transverse tensile 
test specimen fractures at the weld nugget region 
although weld nugget may not be the weakest region 
of the weldment. Three tensile test specimens were 
prepared for each welding condition. The samples 
for tensile test, microhardness, and microstructure 
evaluation were extracted from the welded specimen, 
as illustrated in Fig. 3a. 
The dimensions of the transverse tensile test 
specimen and reduced section tensile test specimen 
are shown in Fig. 3b and Fig. 3c, respectively. The 
tensile testing of the specimen was done with a 
Zwick/Roell Z250 computer-controlled machine. 
The speed of testing was restricted between 1.5 mm/
min to 15 mm/min according to the E-8 standard 
of ASTM; the average of the results of ultimate 
percentage elongation and percentage reduction in 
area for three specimens are recorded and tabulated 
in Table 5. The microstructure of the FSW joint for 
the optimized condition was observed with a scanning 
electron microscope. The specimen extracted for 
microstructural evaluation was further polished and 
etched with a 3 % Nital solution. 
Microhardness testing was done as per ASTM 
E-92. Vickers’s hardness at a 100-gram load was 
taken over the wide region of the welded specimen 
under optimized conditions. The fractured tensile 
tested specimen was also observed under a scanning 
electron microscope to observe the fractured surface 
morphology.
3  RESULTS AND DISCUSSIONS
3.1  Developing Mathematical Model 
Tensile strength is a function of process parameters 
such as welding speed (V), tool rpm (N) and shoulder 
diameter (D). As the number of levels of the above 
process parameters was three, so up to second-
order terms have been considered in the regression 
equation. The second-order regression equation used 
to represent the effect is represented by Eq. (1).
          Effect = Constant + b
1
(V) + b
2
(V)
2
 + b
3
(N)
2
 
 + b
4
(N)
2
 +b
5
(D) + b
6
(D)
2
, (1)
where b
1
, b
2
, b
3
, b
4
, b
5
 and b
6
 are the regression 
coefficients. During the regression analysis, some of 
the coefficients may come out to be zero, meaning 
that those parameters do not affect the particular 
mechanical property. Standard t-test and regression 
Fig. 3.  a) Test coupon for taking out samples for mechanical and metallurgical testing, b) dimensions of transverse tensile test specimen, 
and c) dimensions of reduced section tensile test specimen
Table 5.  Run order and corresponding output responses
Run 
order
Coded values Original values
For transverse tensile 
test specimen
For reduced section specimen
V N D V N D UTS [MPa] UTS [MPa] Elongation [%] Reduction in area [%]
1 1 1 1 60 430 15 407.13 538.73 20.49 44.74
2 1 2 2 60 550 18 430.00 460.70 20.62 45.72
3 1 3 3 60 750 20 414.10 526.60 18.84 42.48
4 2 1 2 110 430 18 408.83 533.56 22.00 58.82
5 2 2 3 110 550 20 410.23 525.96 23.26 52.30
6 2 3 1 110 750 15 408.53 549.76 23.82 47.93
7 3 1 3 210 430 20 408.93 556.80 25.53 60.34
8 3 2 1 210 550 15 410.60 525.16 24.29 64.55
9 3 3 2 210 750 18 419.56 564.36 24.09 65.41

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
316 Bhatia, A. – Wattal, R.
analysis was performed using MINITAB software. 
F-test was also performed on the data at a 95 % 
confidence level to prove that the data fit the proposed 
model.
The strength of the weld nugget could only be 
determined by using the reduced section tensile test 
specimen. The final mathematical model developed 
for reduced section specimen (i.e., ultimate tensile 
strength [MPa], elongation [%] and reduction in area 
[%]) is presented below:
 UTS = 624.1 + 20.05(V) – 162.2(N) + 41(N
2
), (2)
 Elongation = 18.32 + 2.326(V), (3)
       Reduction in area = 19.86 + 9.56(V) 
                              + 18.02(D) – 4.59 (D
2
). (4)
Analysis of variance was done to identify the 
significant factors that affect the ultimate tensile 
strength, percentage elongation, and percentage 
reduction in area. The results of ANOV A are presented 
in Table 6.
3.2  Checking the Adequacy of Developed Model
To check the adequacy of the developed model, the 
coefficient of determination (R
E
2
) and the square 
root of the coefficient of determination (R
E
) were 
determined. The closer is the value of R
E
 
to one, the 
better is the model. The sum of the square of residual 
(SS residual), regression model sum of squares (SS 
model), mean square of residual (MS residual), and 
the degree of freedom (D
f 
model and D
f
 residual) 
are used for the calculation of the coefficient of 
determination (R
E
2
). Adjusted R
E
2 
that adjusts the 
number of terms in a model was also calculated. The 
correlation parameters for the developed model for 
the ultimate tensile strength, percentage elongation, 
and percentage reduction in area were calculated 
and are presented in Table 7. The ‘p’ value for all the 
developed models is less than 0.05, which proves that 
the model is adequate.
Table 6a.  ANOVA table for Ultimate tensile strength of reduced 
section specimen
Effect Error t-value p-value Significant
Constant 48.0293 12.5563 0.0000 Yes
Welding speed 8.20525 3.7432 0.0406 Yes
(Welding speed)
2
Pooled
Tool RPM 53.5256 -3.2989 0.0128 Yes
(Tool RPM)
2
13.2450 3.23889 0.0208 Yes
Shoulder diameter Pooled No
(Shoulder diameter)
2
Pooled No
Table 6b.  ANOVA table for percentage elongation of the weld
Effect Error t-value p-value Significant
Constant 0.8655 21.88 0.00000 Yes
Welding speed 0.4021 5.334 0.00092 Yes
(Welding speed)
2
Pooled No
Tool RPM Pooled No
(Tool RPM)
2
Pooled No
Shoulder diameter Pooled No
(Shoulder diameter)
2
Pooled No
Table 6c.  ANOVA table for percentage reduction in area of the weld
Effect Error t-value p-value Significant
Constant 6.993 2.5673 0.028756 Yes
Welding speed 1.028 8.7654 0.000294 Yes
(Welding speed)
2
Pooled No
Tool RPM Pooled No
(Tool RPM)
2
Pooled No
Shoulder diameter 7.567 2.65431 0.041513 Yes
(Shoulder diameter)
2
1.889 -2.65435 0.034563 Yes
Table 7.  Correlation parameters of the developed model
Correlation 
parameter
For UTS
For percentage 
Elongation
For percentage 
reduction in area
Multiple R
E
0.8938 0.97423 0.97231
Multiple R
E
2
0.8038 0.8518 0.93831
Adjusted R
E
2
0.6860 0.8024 0.90132
SS model 5801.13 32.7336 590.971
D
f
  model 1 1 3
MS model 1933.71 31.9065 196.990
SS residual 1530.286 7.1123 38.870
D
f
  residual 5 7 5
MS residual 308.6411 1.0349 7.7741
F 6.83 29.861 25.34
P 0.032 0.00094 0.002
3.3  Optimization of Welding Parameters 
In order to optimize the welding parameters, response 
surface methodology is used. Figs. 4 to 6 represent the 
surface plots and contour plots for UTS, percentage 
elongation, and percentage reduction in area for the 
different combination of process parameters. 
3.3.1  Effect of Process Parameters on UTS of Welds
The temperature generated during welding is inversely 
proportional to welding speed. Thus, recrystallisation 
takes place at a lower temperature when welding is 
carried out at a higher speed, which restricts the grain 
growth after recrystallisation, which leads to a finer 
grain structure. Hence welding speed significantly 

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
317 Process Parameters Optimization for Maximizing Tensile Strength in Friction Stir-Welded Carbon Steel
affects the ultimate tensile strength, and the strength 
of the weld nugget increases with speed, but only at a 
maximum limit of welding speed. Hence, in order to 
achieve the maximum tensile strength, welding speed 
should be kept as high as possible. 
The friction temperature also increases with the 
increase in tool RPM. This high temperature leads to 
softening of the material; therefore, the value of flow 
stress will become lower. Higher strain rates are also 
achieved by increasing tool RPM. When the tool RPM 
is increased initially, this factor is non-dominant as 
the increase in temperature largely governs the size 
of the recrystallized grain. On further increasing the 
tool RPM, the increase in the strain rate takes place, 
and the strength starts to increase, which can be seen 
clearly from Fig. 4.
Fig. 4.  a) Surface plot showing variation of UTS with the process 
parameters, and b) contour plot of UTS showing variations with 
process parameters
Increasing the tool shoulder diameter increases 
the amount of heat generated, but the larger tool 
diameter takes away the heat due to conduction. 
Therefore, due to the larger tool shoulder, the heat is 
spread over the larger area. Thus, this factor has not 
affected the ultimate tensile strength of the weld.
3.3.2  Effect of Process Parameters on Percentage 
Elongation
The percentage elongation depends upon the strain 
hardening behaviour of the material that is, in turn, 
dependent on the grain size. It is evident from Fig. 
5 that the percentage elongation increases with the 
welding speed. This is because of the reduction in 
grain size due to the higher strain rates. The effect of 
Tool RPM in increasing the strain rate and increasing 
the temperature during welding seem to have 
compensated each other within the range; therefore, 
it is independent of variation of tool RPM. Also due 
to the larger tool diameter, heat is spread all over the 
larger area; thus, it does not affect the percentage 
elongation.
Fig. 5.  a) Surface plot showing variation of percentage elongation 
with the process parameters, and b) contour plot of percentage 
elongation showing variations with process parameters
3.3.3  Effect of Process Parameters on Percentage 
Reduction in Area
As seen from Fig. 6, the percentage reduction in area 
increases with the welding speed. The reason for this 
is the reduction in grain size due to the higher strain 
rate. The effects of tool RPM in increasing the strain 
rate and increasing the temperature during welding 
seem to have compensated for each other within the 
range in which process parameters were varied. A 

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
318 Bhatia, A. – Wattal, R.
exhibits ferrites and small grains of pearlites, as 
shown in Fig 7a. The optical micrograph consists of 
regions containing the fine-grained heat-affected zone, 
interface between heat affected zone and stir zone, 
subregion inside the weld nugget as shown by Figs. 
7b to d. The reason behind this ferrite and pearlite 
growth is due to relatively slow cooling rates. Fig. 7d 
represents the formation of refined grain structure in 
the stir zone.
The microstructural analysis reveals that heat 
affected zone consists of ferrite and pearlites. 
The ferrites are represented by the light-etched 
constituents, and pearlite is represented by the dark-
etched constituent. The refinement of grains was 
also observed in the stir zone as a result of which 
refined grain structures are observed in this zone. 
The boundary between the transformational zone 
is successfully identified, which shows the clear 
difference of recrystallisation. 
Fig. 7.  Optical micrograph of various regions of the friction stir 
welded steel; a) base metal (region A), b) heat affected zone 
(region B), c) interface between the heat affected zone and stir 
zone (region C), and d) stir zone (region D)
slight effect of tool shoulder diameter is also observed 
on the percentage reduction in area.
Fig. 6.  a) Surface plot showing variation of percentage reduction 
in area with the process parameters, and b) contour plot of 
percentage reduction in area showing variations with process 
parameters
3.4  Results of the Confirmation Experiment
A predicted maximum tensile strength is 564.37 MPa 
for the values of welding speed 210 mm/min, Tool 
RPM as 750 mm/min and tool shoulder diameter as 20 
mm. The value of UTS as determined experimentally 
under the same conditions was 553.5 MPa, which 
shows the consistency of the model. The results of the 
confirmation experiment are illustrated in Table 8.
3.5  Observations on Microstructure
Fig. 7 illustrates the microstructure of the weld 
obtained under optimized conditions at different 
locations. The location from which the micrographs 
have been taken is shown in Fig. 7. The base metal 
Table 8.  Results of confirmation experiment
Process parameters Ultimate tensile strength
Error 
[%]
Welding speed (V ) 
[mm/min]
Tool RPM (N) 
[mm/min]
Shoulder diameter (D) 
[mm]
Predicted value 
[MPa]
Experimental value 
[MPa]
210 750 20 564.37 553.5 1.86

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
319 Process Parameters Optimization for Maximizing Tensile Strength in Friction Stir-Welded Carbon Steel
3.6  Observation on Microhardness of the Weld
The plan followed for making indentations to conduct 
the microhardness testing is shown in Fig. 8. The 
specimen was polished on transverse section of the 
weld for testing purpose. Indentations were made 
1 mm apart, covering the entire region of weldment 
on the transverse section. A total of 20 mm distance 
was covered during the indentation scheme, which 
was sufficient to cover the weld nugget, HAZ and a 
portion of the unaffected base material.
The results of the microhardness are shown in 
Fig. 9. The hardness taken in the stir zone shows a 
constituently high value. The microhardness values in 
the weld nugget are higher as compared to the base 
metal. Although the hardness values vary across the 
weld nugget, they never fall below the hardness value 
of base metal. The increase in hardness is primarily 
due to the fine grain structure of the weld nugget.
Fig. 8.  Scheme followed for making indentations for conducting 
microhardness test
 
Fig. 9.  Microhardness profile (transverse section) of the specimen 
welded under optimized conditions
3.7  Fractured Surface
Fig. 10 illustrates the microstructure of fractured 
tensile test specimen in which dimpled surfaces are 
predominantly seen over the entire area. The mode of 
failure appears to be ductile fracture, which is further 
confirmed by appearance of sheared surfaces and 
deformed craters. 
Fig. 10.  SEM image of fractured joint region
4  CONCLUSIONS
The major conclusions that can be drawn from the 
present study are:
1. Mathematical models were developed for the 
ultimate tensile strength, percentage elongation 
and percentage reduction in area, using regression 
analysis and ANOVA was used to predict the 
tensile strength of the joint at a 95 % confidence 
interval. 
2. The weld nugget was reported to be the strongest 
region, and the unaffected base metal was 
reported to be the weakest region. 
3. The confirmation experiment for ultimate tensile 
strength depicts that the developed model was 
adequate.
4. The ultimate tensile strength was significantly 
affected by the welding speed and tool RPM. The 
percentage elongation of the weld was affected by 
welding speed only, and the percentage reduction 
in the area was significantly affected by welding 
speed and tool shoulder diameter.
5. The microstructure of the welded joint for the 
optimized conditions was observed. Pearlite and 
ferrites were observed in the heat-affected zone, 
and refined pearlitic structures were observed in 
the stir zone.
6. The microhardness values in the weld nugget 
are higher as compared to the base metal. The 
increase in hardness is primarily due to the fine 
grain structure of the weld nugget.  
7. The fracture morphology of the tensile test 
specimen was also carried out and revealed 
ductile fracture because the dimpled regions were 
observed over the entire area. 

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
320 Bhatia, A. – Wattal, R.
5  ACKNOWLEDGEMENTS
The authors would like to thank DTU, Delhi for 
providing the Experimental setup and Central 
Research Facility, NCU, Gurugram for providing the 
necessary testing facilities required for the present 
study.
6  REFERENCES
[1] AWS. (1996). AWS Handbook, 8
th
 ed., vol.3, American Welding 
Society, Miami.
[2] Thomas, W.M., Nicholas, E.D., Needham, J.C., Murch, M.G., 
Temple-Smith, P., Dawes, C.J. (1995). Friction Welding, Patent 
no. 5460317. United States Patent and Trademark Office 
(USPTO), Washington.
[3] Mishra, R.S., Ma, Z.Y. (2005). Friction stir welding and 
processing. Materials Science and Engineering: R: Reports, 
vol. 50, no. 1-2, p. 1-78, DOI:10.1016/j.mser.2005.07.001.
[4] Hassan, Kh.A.A., Norman, A.F., Price, D.A., Prangnell, P.B. 
(2003). Stability of nugget zone grain structures in high 
strength Al-alloy friction stir welds during solution treatment. 
Acta Materialia, vol. 51, no. 7, p. 1923-1936, DOI:10.1016/
s1359-6454(02)00598-0.
[5] Heinz, B., Skrotzki, B. (2002). Characterization of a friction-
stir-welded aluminum alloy 6013. Metallurgical and Materials 
Transactions B, vol. 33, p. 489-498, DOI:10.1007/s11663-
002-0059-5.
[6] Liu, H.J., Fujii, H., Maeda, M., Nogi, K. (2003). Tensile 
properties and fracture locations of friction-stir-welded joints 
of 2017-T351 aluminum alloy. Journal of Materials Processing 
Technology, vol. 142, no. 3, p. 692-696, DOI:10.1016/s0924-
0136(03)00806-9.
[7] Mahoney, M.W., Rhodes, C.G., Flintoff, J.G., Bingel, W.H., 
Spurling, R.A. (1998). Properties of friction-stir-welded 7075 
T651 aluminum. Metallurgical And Materials Transactions A, 
vol. 29, p. 1955-1964, DOI:10.1007/s11661-998-0021-5.
[8] Sato, Y.S., Kokawa, H. (2001). Distribution of tensile property 
and microstructure in friction stir weld of 6063 aluminum. 
Metallurgical and Materials Transactions A, vol. 32, p. 3023-
3031, DOI:10.1007/s11661-001-0177-8.
[9] Sato, Y.S., Kokawa, H., Enomoto, M., Jogan, S., Hashimoto, 
T. (1999). Precipitation sequence in friction stir weld of 
6063 aluminum during aging. Metallurgical and Materials 
Transactions A, vol. 30, p. 3125-3130, DOI:10.1007/s11661-
999-0223-5.
[10] Sato, Y.S., Urata, M., Kokawa, H., Ikeda, K. (2003). Hall–petch 
relationship in friction stir welds of equal channel angular-
pressed aluminium alloys. Materials Science and Engineering: 
A, vol. 354, no. 1-2, p. 298-305, DOI:10.1016/s0921-
5093(03)00008-x.
[11] Su, J.-Q., Nelson, T.W., Mishra, R., Mahoney, M. (2003). 
Microstructural investigation of friction stir welded 7050-
T651 aluminium. Acta Materialia, vol. 51, no. 3, p. 713-729, 
DOI:10.1016/s1359-6454(02)00449-4.
[12] Lakshminarayanan, A.K., Balasubramanian, V. (2008). 
Process parameters optimization for friction stir welding of 
RDE-40 aluminium alloy using Taguchi technique. Transactions 
of Nonferrous Metals Society of China, vol. 18, no. 3, p. 548-
554, DOI:10.1016/s1003-6326(08)60096-5.
[13] Vijayan, S., Raju, R., Rao, S.R.K. (2010). Multiobjective 
optimization of friction stir welding process parameters on 
aluminum alloy AA 5083 using Taguchi-based grey relation 
analysis. Materials and Manufacturing Processes, vol. 25, no. 
11, p. 1206-1212, DOI:10.1080/10426910903536782.
[14] Elatharasan, G., Senthil Kumar, V.S. (2014). Corrosion analysis 
of friction stir-welded AA 7075 aluminium alloy. Strojniški 
vestnik – Journal of Mechanical Engineering, vol. 60, no. 1, p. 
29-34, DOI:10.5545/sv-jme.2012.711.
[15] Kadaganchi, R., Gankidi, M.R., Gokhale, H. (2015). 
Optimization of process parameters of aluminum alloy AA 
2014-T6 friction stir welds by response surface methodology. 
Defence Technology, vol. 11, no. 3, p. 209-219, DOI:10.1016/j.
dt.2015.03.003.
[16] Babu, N., Karunakaran, N., Balasubramanian, V. (2017). A 
study to estimate the tensile strength of friction stir welded 
AA 5059 aluminium alloy joints. The International Journal 
of Advanced Manufacturing Technology, vol. 93, p. 1-9, 
DOI:10.1007/s00170-015-7391-9.
[17] Satheesh, C., Sevvel, P., Senthil Kumar, R. (2020). 
Experimental identification of optimized process parameters 
for FSW of Az91C mg alloy using quadratic regression models. 
Strojniški vestnik – Journal of Mechanical Engineering, vol. 
66, no. 12, p. 736-751, DOI:10.5545/sv-jme.2020.6929.
[18] Lienert, T.J., Stellwag, W.L., Jr., Grimmett, B.B., Warke, R.W. 
(2003). Friction stir welding studies on mild steel. Supplement 
To The Welding Journal, vol. 82, p. 1s-9s.
[19] Ueji, R., Fujii, H., Cui, L., Nishioka, A., Kunishige, K., Nogi, K. 
(2006). Friction stir welding of ultrafine grained plain low-
carbon steel formed by the martensite process. Materials 
Science and Engineering: A, vol. 423, no. 1-2, p. 324-330, 
DOI:10.1016/j.msea.2006.02.038.
[20] Fujii, H., Cui, L., Tsuji, N., Maeda, M., Nakata, K., Nogi, K. 
(2006). Friction stir welding of carbon steels. Materials 
Science and Engineering: A, vol. 429, no. 1-2, p. 50-57, 
DOI:10.1016/j.msea.2006.04.118.
[21] Cui, L., Fujii, H., Tsuji, N., Nogi, K. (2007). Friction stir welding 
of a high carbon steel. Scripta Materialia, vol. 56, no. 7, p. 
637-640, DOI:10.1016/j.scriptamat.2006.12.004.
[22] Miles, M.P., Nelson, T.W., Steel, R., Olsen, E., Gallagher, M. 
(2009). Effect of friction stir welding conditions on properties 
and microstructures of high strength automotive steel. 
Science and Technology of Welding and Joining, vol. 14, no. 3, 
p. 228-232, DOI:10.1179/136217108x388633.
[23] Ghosh, M., Hussain, M., Kumar Gupta, R. (2012). Effect 
of welding parameters on microstructure and mechanical 
properties of friction stir welded plain carbon steel. ISIJ 
International, vol. 52, no. 3, p. 477-482, DOI:10.2355/
isijinternational.52.477.
[24] Lakshminarayanan, A.K., Balasubramanian, V., Salahuddin, 
M. (2010). Microstructure, tensile and impact toughness 
properties of friction stir welded mild steel. Journal of Iron and 
Steel Research International, vol. 17, p. 68-74, DOI:10.1016/
s1006-706x(10)60186-0.

Strojniški vestnik - Journal of Mechanical Engineering 67(2021)6, 311-321
321 Process Parameters Optimization for Maximizing Tensile Strength in Friction Stir-Welded Carbon Steel
[25] Ghosh, M., Kumar, K., Mishra, R.S. (2012). Process 
optimization for friction-stir-welded martensitic steel. 
Metallurgical and Materials Transactions A, vol. 43, p. 1966-
1975, DOI:10.1007/s11661-012-1084-x.
[26] Yabuuchi, K., Tsuda, N., Kimura, A., Morisada, Y., Fujii, 
H., Serizawa, H., Nogami, S., Hasegawa, A., Nagasaka, T. 
(2014). Effects of tool rotation speed on the mechanical 
properties and microstructure of friction stir welded ODS steel. 
Materials Science and Engineering: A, vol. 595, p. 291-296, 
DOI:10.1016/j.msea.2013.12.022.
[27] Karami, S., Jafarian, H., Eivani, A.R., Kheirandish, S. (2016). 
Engineering tensile properties by controlling welding 
parameters and microstructure in a mild steel processed by 
friction stir welding. Materials Science and Engineering: A, vol. 
670, p. 68-74, DOI:10.1016/j.msea.2016.06.008.
[28] Mironov, S., Sato, Y.S., Yoneyama, S., Kokawa, H., Fujii, 
H.T., Hirano, S. (2018). Microstructure and tensile 
behavior of friction-stir welded trip steel. Materials Science 
and Engineering: A, vol. 717, p. 26-33, DOI:10.1016/j.
msea.2018.01.053.
[29] Mahmoudiniya, M., Kokabi, A.H., Kheirandish, S., Kestens, 
L.A.I. (2018). Microstructure and mechanical properties 
of friction stir welded ferrite-martensite DP700 steel. 
Materials Science and Engineering: A, vol. 737, p. 213-222, 
DOI:10.1016/j.msea.2018.09.013.
[30] Lee, S.-J., Park, T.M., Nam, J.-H., Choi, W.S., Sun, Y., Fujii, H., 
Han, J. (2019). The unexpected stress-strain response of 
medium Mn steel after friction stir welding. Materials Science 
and Engineering: A, vol. 744, p. 340-348, DOI:10.1016/j.
msea.2018.12.041.
[31] Standard test methods for tension testing of metallic 
materials, in E 8M-04,2004, ASTM International: West 
Conshohocken, PA 19428-2959, p. 1-6.