Engineering Research Express

PAPER You may also like

- Influence of thickness of zinc coating on
Experimental investigation and process parameter CMT welding-brazing with AlSi5 alloy wire
Pengli Jin, Zhiping Wang, Sinan Yang et
optimization in cold metal transfer welding for al.

- Macrostructure, microstructure and wear

SS304L using response surface method performance of Al alloy cladding fabricated
by CMT technique
Yinbao Tian, Junqi Shen, Shengsun Hu et
To cite this article: Sirisha Bhadrakali Ainapurapu et al 2023 Eng. Res. Express 5 015076 al.

- Microstructures and mechanical properties

of Ni/Fe dissimilar butt joint welded using
the cold metal transfer
YunLong Tang, Xin Ye, LongCan Ding et
View the article online for updates and enhancements. al.

This content was downloaded from IP address 103.28.246.17 on 16/01/2024 at 16:59

 Eng. Res. Express 5 (2023) 015076 https://doi.org/10.1088/2631-8695/acbd86

PAPER

Experimental investigation and process parameter optimization in

RECEIVED
28 August 2022
cold metal transfer welding for SS304L using response surface
REVISED
12 January 2023
method
ACCEPTED FOR PUBLICATION
20 February 2023
PUBLISHED
Sirisha Bhadrakali Ainapurapu1,2,∗ , Venkata Anantha Ramasastry Devulapalli1 and
27 March 2023 Ram Prabhu Theagarajan3
1
Department of Mechanical Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Andhra Pradesh, 522302, India
2
Department of Mechanical Engineering, Aditya Engineering College(A), Surampalem, Andhra Pradesh, 533437, India
3
Defence Research and Development Organisation (DRDO), CEMILAC, Bangalore, Karnataka, 560093, India
∗
Author to whom any correspondence should be addressed.
E-mail: [email protected], [email protected] and [email protected]

Keywords: CMT welding, Austenitic steel, optimization, response surface methodology

Abstract
Cold Metal Transfer (CMT) welding was used to join SS304L plates using SS308L MIG wire. The
effects of current, the distance between nozzle and tip, welding speed on strength, hardness and
surface roughness of welded zone in the CMT process using response surface methodology were
investigated in the present study. The outcomes show that current is having a significant effect on the
outputs. Also, current interacting with weld speed is showing its mark on the output variables like
tensile strength, hardness and surface roughness. Finally, optimal results are found and the
experimental values are in coincidence with the predicted values.

1. Introduction

Short-circuiting transfer based and a modified form of MIG welding process by name Cold metal transfer
welding process. This was developed in the early 2000s by an Austrian company Fronius. The only difference
between MIG/MAG and CMT is the mechanical droplet cutting method [1]. A method for controlling the
material deposition which leads to less thermal input is used in the CMT welding process [2]. This less heat input
leads to a reduction in solidification defects [3]. An important feature of the CMT process is a short circuit with
low current leading to low heat input.
Low heat input results in inadequate melting time for enough materials, which causes a low volume weld
pool with quick freezing as a result. This affects the dendritic structure and regulates the development of
intermetallic layers [4]. It is proposed that a further decrease in weld heat input can be achieved by combining
welding with a magnetic field for joining dissimilar metals [5].
The short circuit in CMT occurs in a controlled manner with stability. This confirms the stable and
controlled deposition in CMT welding. Desired control of heat input, high welding speed, narrower heat
affected zone etc are the advantages offered by the CMT technique compared to other joining processes [6].
There are four phases in the CMT welding process i.e., two hot phases and two cold phases. The sequence of
those is hot, cold, hot, cold [7].
During the first hot phase, i.e., arcing period, filler metal moves to the weld pool. In the cold phase 1, filler
metal drench in the weld pool, the arc quench leads to the lowering of welding current. the hot phase 2, wire
droplet detachment is assisted by wire backward movement through the short circuit and the current of it is
maintained little. In the last stage, there is a reversal in wire motion.
Researchers studied Cold Metal Transfer welding with the characteristics of arc, transfer of droplet,
weldment properties and microstructure [8].
To improve the quality, flexibility and productivity of the welding performance, process parameter
optimization is necessary. Transient numerical model was developed to analyse the TIG - CMT hybrid welding

© 2023 IOP Publishing Ltd

Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Table 1. Input parameters -levels.

Levels
Process parameter Units
−1 1

Current Amp 95 105

Contact Tip to Wrok distance mm 6 10
Travel Speed mm/sec 4.03 5.817

and further increase the potentiality of weld formation for cladding Inconel [9]. Mathematical models were
developed to show the effects of input parameters on the quality of weldment [10]. Researchers paid attention to
the recognition of the best parametric combination in CMT welding of Al alloys [11]. Effects of Wire feed speed
and Arc length correction on the micro, macrostrucutres and wear performance on Aluminium cladding using
CMT were investigated [12]. Multi-parameter optimization techniques were used to identify the weld quality in
terms of reinforcement of welded joint, width of the bead, penetration depth i.e., DOP, width of the heat affected
zone i.e., HAZ [13]. Also, various mechanical properties were optimized in CMT welding using multi-parameter
optimisation [14]. Various mathematical models were developed in welding using response surface
methodology techniques to predict the weld quality [15]. Full factorial central composite face-cantered design
(CCFCD) was utilized to optimize the tensile properties, microhardness and residual stresses in CMT
welding [16].
CMT welding of Al 6061-T6 and galvanized steel for automobile applications was done with the aid of OA,
ANOVA analysis and optimal parametrs are determined [17]. The desirability functions, design of experiments,
and polynomial models are all used in the RSM approach to obtain the ideal working point with the least trials
[18]. DoE is a group of methods for defining a series of tests whose outcomes depend on the input variables.
Fisher created it in the 1920s, and it was first used to examine how fertilisers affected various crops [19].
Currently, both experimentally and through simulation [20–23]. DOE are implemented to perform
optimizations for both multi and single objectives. Box Behnken design was developed for single, multi
parameter optimisation for machining of Incoloy 800H using Wire EDM [24]. Taguchi optimisation was done
with the aid of ANOVA for various parameters in Magnetic assisted abrasive machining [25].
DoE evaluates the factors’ numbers, combinations, randomization, replication, and blocking to confidently
analyse cause-and-effect correlations. Since the trials in this instance are finite element studies, randomization,
replication, and blocking are not required.
In this work, Response surface methodology is used to identify the process parameters which yield better
quality in weldments in terms of weld strength, hardness and surface roughness along with mathematical model
development.

2. Experimental method and procedure

2.1. Design of experiments

Bead geometry of the deposition of the welded zone is going to be varied based on process parameters: viz
welding current, voltage, contact-to-workpiece distance, and travel speed [26]. Among those process parameters
current, nozzle to tip distance and travel speed are considered for the experimentation purpose. Welding design
experiments and the range of process parameters are identified based on the trial experiments and machine
compatibility. To identify these parametrical effects on Tensile strength, hardness and surface roughness of the
weldment three-factor two-level design was selected. Table 1 shows the input parameters with their levels. A
polynomial equation of second order for every output response was formulated using Design Expert software.
Response Surface Methodology is a group of mathematical, statistical methods which are useful for
identifying the relationship between the input and outputs of a process. Finding an appropriate approximation
for the When the input variables are changed within the chosen ranges, the relationship can be utilised to
anticipate the response. If the response is plotted against any two input variables, the regression model
geometrically reflects the surface. Here, the polynomial expression shown below is used to generate the second
order regression [27, 28].
Y = f (x1, x2, x3,¼ ¼) + e (1)
k k k
Y = b0 + å bj x j + å bjj x j2 åå bij x i x j +e (2)
j=1 j=1 i<j j=2

In equation 1 and 2 Where Y, the Predicted response

2
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 1. SS304L plates and SS308L.

Table 2. Matrix of design.

Current Contact tip to work Travel speed

Run (Amp) distance (mm) (mm/sec)

1 95 10 4.03
2 105 10 4.03
3 105 6 5.817
4 108.409 8 4.9235
5 100 8 3.42082
6 95 10 5.817
7 100 8 4.9235
8 105 6 4.03
9 91.591 8 4.9235
10 100 4.63641 4.9235
11 95 6 5.817
12 100 8 4.9235
13 100 11.3636 4.9235
14 105 10 5.817
15 100 8 4.9235
16 100 8 4.9235
17 100 8 6.42618
18 100 8 4.9235
19 100 8 4.9235
20 95 6 4.03

x1, x2, x3, xj are single parameters

ε is error/noise predicted.
β0 is constant, βj is linear coefficient, βjj is the quadratic coefficient, βij is the ith interaction coefficient, and k
denotes the number of factors.
Using the Central Composite Design (CCD) of RSM, different groups of experiments were designed with the
help of Design Expert software which was tabulated in table 2.

2.2. Materials
Austenitic steels can be welded through all sorts of welding techniques. Among those SS304L has versatile
applications in industry, and domestic fields and attractive features viz high resistance to corrosion, ease of
fabrication, and outstanding formability. SS304L plates with a thickness of 5 mm are taken. They are butt welded
using SS308L wire using CMT welding. These are shown in figure 1.
Chemical composition and Properties of the SS304L and MIG wire are mentioned in table 3.

2.3. Procedure of welding

CMT Welding is performed with the design matrix mentioned in table 2 with varying process parameters. While
doing the welding process as it is unable to set the values of current, CTWD and Speed with decimal points, like

3
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 2. CMT welding setup.

Table 3. Chemical composition and properties of SS304L and

SS308L.

Chemical composition
(Weight %)
Element
SS304L SS308L

Carbon, C 0.019 0.018

Chromium, Cr 18.2 19.7
Silicon, Si 0.5 0.32
Nickel, Ni 8.04 10.1
Iron, Fe Balance Balance
Manganese, Mn 1.53 1.9
Phosphorous, P 0.03 0.017
Sulphur, S 0.03 0.01
PROPERTY VALUE
SS304L SS308L
Shear strength 370–680 MPa 380 MPa
Tensile strength (ultimate) 540–1160 MPa 580 MPa
Tensile strength, (yield) 190–870 MPa 230 MPa
Poisson’s Ratio 0.28 0.28
Youngs Modulus 200 GPa 200 GPa

108.409 Amp, 91.591 Amp, 5.817mm/sec, 4.9235mm/sec etc are set up to the nearest values like 108 amp, 92
Amp, etc. The CMT setup used is shown the figure 2. Shielding gas used is Argon and CO2 mixture with a flow
rate of 18 Litres per minute.Metallurgical evidence for the Weld bead of one joint is shown in figure 3.

2.4. Output performances

Properties like Tensile strength, Hardness and Surface Roughness are measured for the weldments. These tests
were conducted in accordance with ASTM standards.
For measuring tensile strength at the welded zone, the specimen was cut using wire EDM. The specimen was
subjected to a tensile test using the micro tensile testing machine. Hardness test was conducted at the welded zone
using a hardness tester with a load of 150 kgf, and 15 s dwell time. The surface roughness test was done using Talysurf.
Figure 4 shows these setups.
Tensile strength, hardness and Surface roughness test results were tabulated in the following table 4.

4
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 3. Weld beads of the CMT welded joints.

Figure 4. (a) Tensile testing specimen (b) Micro Tensile Testing Machine. (c) Rockwell Hardness tester (d), (e) Hardness Testing areas
(f) Surface Roughness Tester.

3. Results and discussions

3.1. Effect of input parameters on output responses

3.1.1. ANOVA
While doing the analysis the FI model is giving significance for the Lack of fit.Quadratic model is considered for
doing the analysis in which lack of fit is insignificant.

3.1.1.1. Tensile strength

The goal of this work is to gain maximum tensile strength at the welded zone with the input parameters stated
above.
A statistical technique called Analysis of Variance (ANOVA) is used to identify the CMT process parameter
that has the greatest impact on output response characteristics. Here Fisher Ratio test and P- value test with a
significance level of P value < 0.05 are used.
The model is suggested to be significant by the Model F-value of 916.05. An F-value this large might happen
to owe to noise only 0.01% of the time.
Model terms are considered significant when the P-value is less than 0.0500. Current (A), Travel Speed (C)
and their interaction (AC), Interaction of CTWD and Weld speed (BC) are important model parameters in this
scenario. Values larger than 0.1000 denote that the model terms, in this scenario Nozzle to tip distance (NTD),
are insignificant.

5
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Table 4. Output responses for the experimental trials.

Travel speed Tensile Surface rough-

Run Current (Amp) CTWD(mm) (mm/sec) strength (MPa) Hardness (BHN) ness (μ)

1 95 10 4.03 314.71 85.5 16.496

2 105 10 4.03 513.97 69.5 20.027
3 105 6 5.817 240.6 84 33.64
4 108.409 8 4.9235 406.057 80.15 29.5
5 100 8 3.42082 458.563 74 16.56
6 95 10 5.817 394.79 71 18.682
7 100 8 4.9235 372.94 75.78 21.54
8 105 6 4.03 542.64 78.5 28.787
9 91.591 8 4.9235 341.476 71.7 13.4
10 100 4.63641 4.9235 372.455 74.2 24.9
11 95 6 5.817 383.24 71 22.8
12 100 8 4.9235 373.104 75.81 21.556
13 100 11.3636 4.9235 375.079 77.6 18.07
14 105 10 5.817 277.09 82 22.58
15 100 8 4.9235 371.9 75.57 21.48
16 100 8 4.9235 370.776 75.34 21.4224
17 100 8 6.42618 288.971 77.6 26.44
18 100 8 4.9235 370.58 75.302 21.410
19 100 8 4.9235 373.767 75.95 21.595
20 95 6 4.03 327.75 66.3 8.885

Table 5. ANOVA table for tensile strength.

Source Sum of squares df Mean square F-value p-value

Model 97668.84 9 10852.09 504.62 <0.0001

A-current 34568.17 1 34568.17 1607.4 <0.0001
B-CTWD 202.64 1 202.64 9.42 0.0118
C-Travel Speed 19221.27 1 19221.27 893.78 <0.0001
AB 10.83 1 10.83 0.5038 0.4941
AC 56867.1 1 56867.1 2644.71 <0.0001
BC 1006.88 1 1006.88 46.82 <0.0001
A2 6.36 1 6.36 0.2956 0.5986
B2 6.36 1 6.36 0.2957 0.5985
C2 6.36 1 6.36 0.2957 0.5985
Residual 50.802 10 50.802 3.39
Lack of Fit 41.1658 5 8.233 4.815 0.2345
Pure Error 8.56 5 1.71
Cor Total 97719.642 19

Table 6. Fit statistics model for tensile strength.

Standard. Deviation 4.22 R2 0.9976

Mean 373.52 R2 Adjusted 0.9966

C.V.% 1.13 R2 Predicted 0.9827
Adequate Precision 117.0552

According to the R2 statistic, the fitted model accounts for 99.76% of the variation in tensile strength. The
modified R2 value is 99.66%, making it more appropriate for comparing models with various numbers of
independent variables. Adequate precision measures the signal-to- noise ratio and it should be more than 4. In
this case, it is 117.0552 indicating an adequate signal.
Table 5, 6 shows the ANOVA and fit statistics model for Tensile strength.

3.1.1.2. Hardness
To gain maximum hardness at the welded zone for the given inputs following analysis is done.
Following is the table of ANOVA to get the significant values of Hardness at the welded zone.

6
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Table 7. ANOVA for hardness.

Source Sum of squares df Mean square F-value p-value

Model 345.2 9 38.36 4.44 0.0146

A-current 6.45 1 6.45 0.7467 0.4078
B-CTWD 28.86 1 28.86 3.34 0.0976
C-Travel Speed 8.04 1 8.04 0.9302 0.3575
AB 114.01 1 114.01 13.2 0.0046
AC 96.61 1 96.61 11.18 0.0074
BC 18.61 1 18.61 2.15 0.173
A2 0.322 1 0.322 0.0373 0.8508
B2 0.0186 1 0.0186 0.0022 0.9639
C2 0.0984 1 0.0984 0.0114 0.9171
Residual 4.53278 10 8.64
Lack of Fit 1.56234 5 0.312468 4.4321 0.1253
Pure Error 0.3525 5 0.0705
Cor Total 349.73278 19

Table 8. Fit statistics model for hardness.

Std. Dev. 2.58 R2 0.7993

Mean 75.85 Adjusted R2 0.7067

C.V.% 3.40 Predicted R2 −0.5787
Adequate Precision 12.7945

For Hardness, Current(A), Current, Travel speed (AC) interaction, Current, and Contact Tip to work
distance (AB) interaction are significant. Here non-significant factors are CTWD and weld speed.
Here predicted R2 is negative indicating overall mean will be the better predictor for the response. Adequate
precision of 12.7945 indicates that the model is having adequate signal.
ANOVA and fit statistics model for Hardness is shown in tables 7, 8.

3.1.1.3. Surface roughness

Surface Roughness for the welded zone is analysed for minimisation using the following procedures for the
inputs stated above.
Following is the ANOVA table for Surface Roughness:
F-value for the model is 114.34 shows its significance. 0.01% chance is there that a large F- value happens
because of noise.
For Surface Roughness all the input factors along with interactions are significant except interctions of AC and BC.
Predicted R2, 0.8531 is in agreement with the adjusted one and leads to a difference of 0.1197. An adequate
precision value of 45.2096 is greater than 4 indicating an adequate signal.
Table 9, 10 indicates the ANOVA and fit statistics for surface roughness.

3.1.2. Empirical models for output responses

3.1.2.1. Tensile strength

Following equation 3 is the empirical model to express the response i.e. Tensile Strength in terms of input
parameters i,e., Current, Nozzle to tip distance, and weld speed. This empirical models are developed as per
quadratic analysi.

Tensile Strength = - 8687.42055

+ (95.82) A - (42.15384) B + (1780.56035) C
+ (0.116375) AB - (18.87) AC + (6.27798) BC (3)

Where A is Current, B is Nozzle to tip distance, and C is weld speed

Figure 5 shows a scatter plot between the fitted and experimental Tensile strength values of the weldment
and a perturbation plot to show the influence of all single variables at the design space’s central point.
The scattered plots in the first graph are extremely close to the 45° line, indicating that the experimental and
predicted values of tensile strength are comparable. The influence of independent process factors on tensile
strength is displayed on perturbation plots. Tensile strength is shown to decrease as the nozzle to tip distance
increases. With an increase in current and welding speed, there is no appreciable rise in tensile strength.

7
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 5. Predicted versus Actual plot and Perturbation plot.

Table 9. ANOVA for surface roughness.

Sum of Mean
Source squares df square F-value p-value

Model 587.08 9 65.23 52.98 <0.0001

A-Current 272.97 1 272.97 221.71 < 0.0001
B-CTWD 12.43 1 12.43 10.1 0.0099
C-Travel Speed 73.24 1 73.24 59.49 < 0.0001
AB 67.94 1 67.94 55.18 < 0.0001
AC 9.45 1 9.45 7.68 0.0198
BC 24.6 1 24.6 19.98 0.0012
A2 0.0417 1 0.0417 0.0339 0.8577
B2 0.3016 1 0.3016 0.2449 0.6314
C2 0.0063 1 0.0063 0.0051 0.9443
Residual 12.31 10 1.23
Lack of Fit 1.3228 5 0.31567 4.3435 0.1342
Pure Error 0.0283 5 0.0057
Cor Total 588.4028 19

Table 10. Fit statistics model for surface roughness.

Standard. Deviation. 0.9260 R2 0.9814

Mean 21.49 Adjusted R2 0.9728
C.V.% 4.31 Predicted R2 0.8531
Adequate Precision 45.2096

3.1.2.2. Hardness
Equation 4 is the empirical formula derived for the Hardness at the welded zone.
Hardness = 62.98450 - 0.35755(A)
+ 42.46121(B) - 69.788(C ) - 0.377500(AB)
+ 0.777840(AC ) - 0.853386(BC ) (4)

Figure 6 shows the shows a scatter plot between the fitted and experimental Hardness values and a
perturbation plot to show the influence of all single variables at the central point of design space.

8
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 6. Hardness Predicted versus actual plots, Perturbation Plots.

Figure 6 shows the actual versus predicted plot and perturbation plots of Hardness.
Figure 7 shows the actual versus predicted plot and perturbation plots of Surface Roughness
In the left side graph for hardness scattered points are far away from the 45° lines indicating that
experimental values are deviating from predicted values. The perturbation graph shows that output is
responding the same for both B, and C factors. But there is a vast increase in Hardness for Factor A.

3.1.2.3. Surface roughness

Empirical formula derived for the Surface Roughness at the welded zone is mentined in equation 5.

Surface Roughness = - 473.67712 + 4.48465(A)

+ 32.95449(B) + 35.46720(C ) - 0.291413(AB)
- 0.243285(AC ) - 0.981323(BC ) (5)

Surface roughness graph of predicted versus actual values shows that scattered points are near to 45° lines
indicating that experimental values are in accordance with predicted values. The perturbation graph reveals that
output is decreasing for B, while it is increasing with A and C factors.

3.1.3. Contour plots and response graphs

In the study of response surface analysis, contour plots are crucial. All of the operating conditions that produce
the same value of the answer can be discovered using the contour plot and response plot layouts. It served as a
tool for enhancing the CMT welding parameters. The contour plots and response graphs for Tensile strength ,
Hardness and surface roughness can be observed in figures 8–10.

3.1.3.1. Tensile strength

The maximum tensile strength which is the objective of a function is going to be observed in the red coloured
area in different curves of contour plots.

3.1.3.2. Hardness
Maximum Hardness is going to be observed in the red and yellow colour zone in the contour.

3.1.3.3. Surface roughness

Minimum surface roughness is the objective function and it is observed in the blue and green colour zones of
contour and response graphs.

9
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 7. Predicted versus actual plot and perturbation graphs for surface roughness.

3.2. Optimisation
3.2.1. Tensile strength
After observing the individual effects, interaction graphs and response graphs, Current (A), Contact Tip to work
Distance(B), Weld Speed (C) and their interactions are having their effect on the Tensile Strength of the welded
zone. Here Current, CTWD shows a nominal impact whereas Speed is having a considerable effect and is
inversely proportional. In the interactions graph of AC, more curvature is observed revealing that more amount
affects the Tensile strength.

3.2.2. Hardness
For getting the maximum Hardness at the welded zone Current,Contact tip to work distance and Speed shows
direct nominal impact. The interaction of current and weld speed has more impact.

3.2.3. Surface roughness

To achieve minimum surface roughness maximum CTWD is needed. Interactions of AB is having a significant
effect on Surface roughness to achieve a minimum value.

3.3. Optimization rule

Table 11 indicates the optimization rule followed in the experimental work. For getting optimal results goals
were set to get maximum Tensile strength, hardness and minimum surface roughness and limits were fixed for
output results.

3.4. Optimization results

Input process parameters three in number were selected and they were optimised to get optimal output results.
Finally 103 Amperes of current,8 mm of CTWD,4.8 mm s−1 of weld speed were selected in order to get
maximum tensile strength, hardness and minimum surface roughness.These results were tabulated in table 12.

4. Conclusions

Effects of input factors viz Current, contact tip to work distance (CTWD) and Weld speed on the weldment’s
Strength, Hardness and surface roughness after the Cold Metal Transfer welding of SS304L plates using SS308L
using Response Surface Methodology was investigated. Following conclusions are drawn in this experiment and
analysis.

1. Among all the input parameters Current is having a significant effect on the output responses.

10
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 8. Plots of contour and graphs for response to tensile strength.

11
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 9. Contour plots and response graphs for hardness.

12
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Figure 10. Contour plots and response graphs for surface roughness.

13
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

Table 11. Optimization rule.

Parameter Limits

Min Max Goal

Inputs Current 95 105 In range

CTWD 6 10
Weld Speed 4.03 5.817
Outputs Tensile Strength 220 550 Max
Hardness 80 90 Max
Surface Roughness 6 34 Min

Table 12. Optimal results.

Nozzle to tip
Inputs Current distance Weld speed

103Amp 8 mm 4.8 mm s−1

Outputs Tensile Hardness SR
Strength
Actual 392 Mpa 77.3 BHN 23 Microns
Predicted 390 MPa 80 BHN 22Microns

2. Interaction of Current and Weld speed showing the maximum effect on the most important mechanical
properties i.e. Tensile strength, Hardness of the welded zone.
3. If the surface roughness is taken as an important factor nozzle tip distance and its interaction with Current
has to be considered.
4. Finally, a current of 103 Amperes, 8 mm of the nozzle to tip distance and 4.8 mm s−1 are the optimal
values which are yielding a tensile strength of 390 MPa, 80 BHN hardness and 22 microns of surface
roughness

Further this work may be extended to optimise other input parameters to achieve better mechanical properties
and also using various machine learning algorithms.

Data availability statement

All data that support the findings of this study are included within the article (and any supplementary files).

ORCID iDs

Sirisha Bhadrakali Ainapurapu https://orcid.org/0000-0001-9967-2971

Venkata Anantha Ramasastry Devulapalli https://orcid.org/0000-0003-0713-4901

References
[1] Furukawa K 2006 New CMT arc welding process e welding of steel to aluminium dissimilar metals and welding of super-thin
aluminium sheets Weld. Int. 20 440e5
[2] Robert T, Renno V, Andres L and Martinš S 2012 Cold metal transfer (CMT) welding of thin sheet metal products Est. J. Eng. 18 243e50
[3] Selvi S, Vishvaksenan A and Rajasekar E 2018 Cold metal transfer (CMT) technology - an overview Defence Technology 14 28–44
[4] Oliveira A S et al 2021 A detailed forecast of the technologies based on lifecycle analysis of GMAW and CMT welding processes
Sustainability. 13 3766
[5] Sun Q, Li J, Liu Y, Jiang Y, Kang K and Feng J 2018 Arc characteristics and droplet transfer process in CMT welding with a magnetic
field J. Manuf. Processes 32 48–56
[6] Kumar N P, Vendan S A and Shanmugam N S 2016 Investigations on the parametric effects of cold metal transfer process on the
microstructural aspects in AA6061 J. Alloys Compd. 658 255e264
[7] Anik S, Dikicioglu A and Vural M 1994 Koruyucu Gaz Altında Kaynak ve Aluminyum MIG Kaynagi. (Istanbul: Teknisyenler Matbaası)
[8] Huan P C, Wang X N, Zhang J, Hu Z R, Chen W G, Nagaumi H and Di H S 2020 Effect of wire composition on microstructure and
properties of 6063 aluminium alloy hybrid synchronous pulse CMT welded joints Mater. Sci. Eng. A 790 139713

14
Eng. Res. Express 5 (2023) 015076 S B Ainapurapu et al

[9] Wang N, Shen J, Hu S and Liang Y 2020 Numerical analysis of the TIG arc preheating effect in CMT based cladding of Inconel 625
Engineering Research Express. 2 015030
[10] Marić D, Šolić T, Kondić V and Samardžić I 2022 Statistical analysis of MAG-CMT welding parameters and their influence on the Ni-
alloy weld overlay quality on 16Mo3 base material Welding in the World. 66 815–31
[11] Suryaprakash S, Vishal S, Sethu Raman M, Rajendra Kumar S, Umar M and Deepan Bharathi Kannan T 2021 Optimization of welding
parameters in CMT welding of Al 5083 Adv. Mater. Res. (Singapore: Springer) pp 663–72
[12] Tian Y, Shen J, Hu S and Gou J 2020 Macrostructure, microstructure and wear performance of Al alloy cladding fabricated by CMT
technique Engineering Research Express. 2 015026
[13] Aravind A P, Suryaprakash S, Vishal S, Sethuraman M, Kannan T D, Umar M and Rajak S 2020 Optimization of welding parameters in
CMT welding of Al 5083 alloys using VIKOR optimization method InIOP Conference Series: Materials Science and Engineering 912
032035
[14] Bhadrakali A S, Narayana K L and Prabhu T R 2020 Multi response optimization in cold metal transfer welding European Journal of
Molecular & Clinical Medicine. 7 2020
[15] Srivastava S and Garg R K 2017 Process parameter optimization of gas metal arc welding on IS: 2062 mild steel using response surface
methodology J. Manuf. Processes 25 296–305
[16] Koli Y, Yuvaraj N and Aravindan S 2020 Multi-response mathematical modeling for prediction of weld bead geometry of AA6061-T6
using response surface methodology Trans. Indian Inst. Met. 73 645–66
[17] Zope B, Dey T, Jatti V S and Sarnobat S S 2022 Aluminum alloy and galvanized steel CMT weld joints for lightweight automobile
applications Engineering Research Express. 4 035001
[18] Harrington E C 1965 The desirability function Ind Qual Control 4 494–8
[19] Fisher R A and Bennett J H 1990 Statistical methods, experimental design, and scientific inference
[20] Escribano-Garcia R, Lostado-Lorza R, Fernandez-Martínez R, Villanueva-Roldan P and Donald B J M 2014 Improvement in
manufacturing welded products through multiple response surface methodology and data mining techniques Advances in Intelligent
Systems and Computing International Joint Conference SOCO’14-CISIS’14-ICEUTE’14 (Bilbao, Spain: Springer) pp 301–10
[21] Lorza R L, Escribano-Garcia R, Calvo M M and Vidal R M 2016 Improvement in the design of welded joints of EN 235JR low carbon
steel by multiple response surface methodology Metals 6 205
[22] Islam M, Buijk A, Rais-Rohani M and Motoyama K 2015 Process parameter optimization of lap joint fillet weld based on FEM–RSM–
GA integration technique Adv Eng Softw 79 127–36
[23] Lostado R, Escribano García R and Fernandez Martinez R 2016 Optimization of operating conditions for a double-row tapered roller
bearing Int. J. Mech. Mater. Des. 12 353–73
[24] Arya R and Singh H 2022 Optimization of Wire-cut EDM process parameters using TLBO algorithm Engineering Research Express. 4
035051
[25] Gill J S, Singh P and Singh L 2022 Taguchi's design optimization for finishing of plane surface with diamond-based sintered magnetic
abrasives Engineering Research Express. 4 035004
[26] Tahaei A, Vazquez F G, Merlin M, Arizmendi-Morquecho A, Valdes F A R and Garagnani G L 2016 Metallurgical characterization of a
weld bead coating applied by the PTA process on the D2 tool steel Soldag. Inspeção 21 209–19
[27] Qin X P, Gao K, Zhu Z H, Chen X L and Wang Z 2017 Prediction and optimization of phase transformation region after spot continual
induction hardening process using response surface method J. Mater. Eng. Perform. 26 4578–94
[28] Kadaganchi R, Gankidi M R and Gokhale H 2015 Optimization of process parameters of aluminum alloy AA 2014–T6 friction stir
welds by response surface methodology Def. Technol. 11 209–19

15