IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS Related content

- Study on Arc Welding Seam Tracking of
Process Parameter Optimization in TIG Welding of TIG Welding
Jin Wei and Xia Xiang
AISI 4340 Low Alloy Steel Welds by Genetic - Optimization of process parameters in
welding of dissimilar steels using robot TIG
Algorithm welding
G Navaneeswar Reddy and M
VenkataRamana
To cite this article: D. Vijayan and V. Seshagiri Rao 2018 IOP Conf. Ser.: Mater. Sci. Eng. 390
012066 - A Study to Increase Weld Penetration in
P91 Steel During TIG Welding by using
Activating Fluxes
Akhilesh Kumar Singh, Mayank Kumar,
Vidyut Dey et al.

View the article online for updates and enhancements.

This content was downloaded from IP address 196.188.192.117 on 23/12/2020 at 11:16

The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

Process Parameter Optimization in TIG Welding of AISI

4340 Low Alloy Steel Welds by Genetic Algorithm
D.Vijayan1*, V. Seshagiri Rao2
1
Assistant Professor, Department of Mechanical Engineering, Sri Chandrasekharendra
Saraswathi Viswa Mahavidyalaya, Enathur, Kanchipuram, Tamilnadu, India.
2
Professor, St. Joseph’s College of Engineering, Chennai, Tamilnadu, India.

1*
Corresponding author: vijaiand2012@gmail.com

Abstract Weld quality is an essential requirement to the fabrication and construction industries. In this present
investigation, an attempt has been made to optimize the process parameters of Tungsten Inert Gas (TIG)
welding of AISI 4340 low alloy steel. Taguchi L9 orthogonal array is used to design the experiment and
response surface methodology was used to develop mathematical model and the process parameters were
optimized using Genetic Algorithm (GA). The important process parameters namely welding current, voltage,
welding speed, and gas flow rate are considered. The results indicate that the current and voltage have a
maximum influence on improving Ultimate Tensile Strength (UTS) in TIG-welded joint.

1. Introduction
Tungsten Inert Gas (TIG) welding is one of the usual arc welding processes where arc produces
between non-consumable tungsten electrode and workpiece and metal fusion happens. TIG welding is
broadly utilized for joining alloy steels, stainless steels in numerous fabrication industries due to high
caliber and great appearance. Be that as it may, one potential issue related to traditional TIG is the
constraint on workpiece thickness that can be welded in single pass thus offers limited efficiency.
And, TIG welding should have the following characteristics, (1) a straight polarity direct current, (2)
high frequency arc starting, and (3) power supply reduction device (1). Increment in current may not
be shrewd dependably as it will prompt increment in bead width with relatively low pick up in
infiltration. Nickel-Chromium-Molybdenum steels are broadly utilized for joining as a part of various
applications including marine, automotive and machine instruments, pressure vessels, transport,
mining and so on. Impressive measures of spotlight on accomplishing high caliber welded joints of
such materials for similar and dissimilar joining are being investigated at quicker production rates
with higher amount of infiltration. Exertion has been made by a few researchers to modify the TIG
welding process by utilizing fluxes to increase the weld penetration. The depth of penetration amid
TIG welding could be altogether expanded laying an activated flux layer on workpiece surface before
welding, accordingly eliminates the need of edge preparation, reduces number of passes required and
improves efficiency. The procedure is instituted as flux activated TIG welding. Numerous analysts
directed experimental studies to assess the impact of various oxide, halide or mixture of fluxes on
increase in DOP and relating decrease in width or width-to-penetration (WP) on various high nickel-
chromium-molybdenum steels at various instances (2-4). Be that as it may, the welding innovation
has not been exceptionally fruitful to join nickel-chromium-molybdenum steels in commercial
ventures in this way, and related studies are likewise to a great degree constrained. So, it gets to be
important to explore the relevance of welding innovation in joining little scale parts.

2. Materials and Methods

The base materials used for the present investigation were AISI 4340 (ASTM A29), produced by M/s
ALOKINGOTS India Pvt. Ltd, Mumbai, India is a medium carbon, low alloy steel. The chemical
compositions of base metal as received is C – 0.39, Mn – 0.72, Cr – 0.7, Ni – 1.65, Si – 0.15, Mo –
0.2, P – 0.035 and S – 0.04. All AISI 4340 steel plates were sized into 150mm (L) × 75mm (W) ×
5mm (t) were butt welded at different process parameters as per the design matrix. Figure.1
demonstrate the AISI 4340 steel plates to be welded. Before welding, acetone was used to eradicate
the oils that remained on the surface of the specimens. Extensive welding trials were conducted in
autogenous mode with ESAB make A-TIG power source having capacity of 250A with 25% duty
cycle to examine the performance and appearance of the weld.

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

Figure 1. Test specimens of AISI 4340 – medium carbon, low alloy steels

Then, all the steel plates were welded by TIG welding as per the variables designed by orthogonal
array using Taguchi technique. A 3 mm diameter and 2.2 % thoriated tungsten electrode with an air-
cooled torch was used. The welding variables and their corresponding ranges are presented in Table 1.
All the samples were welded under full penetration. The electrode angle was kept as 19 – 22o pure
argon was used as a shielding gas during welding. The flux paste was applied later welding, acetone
evaporated, parting a layer of flux on the surface. The torch was moved along the center line of the
weld specimens.
Table.1 – Welding variables and their range
Welding Gas Flow
Process Current Voltage Speed Rate
Factors (A) (V) (WS) (GFR)
(mm/sec) (lpm)
Levels 70 - 80 12 - 14 55 - 65 8-12

Figure 2. Fabricated AISI 4340 low alloy steel

The fabricated AISI 4340 steel plates after welding are presented in Figure 2 and their corresponding
design matrix is presented in Table 2. After welding, specimens were collected from the center of the
welded plate and subjected to Ultimate Tensile Strength (UTS). Subsequently, for conducting the
hardness test, samples were allowed for grinding and polishing followed by etching in a solution of
35% concentrated HCL at 60 - 80oC for 12–15 minutes to produce a bright surface. Scanning electron
microscope (SEM) (Make: FEI Quanta FEG 200 high resolution – Micro lab, Chennai, Tamilnadu,
India) is used to develop the microstructure of the welded samples. Further, hardness variation in
HAZ, NZ and TMAZ on the effect of TIG welding process parameters the specimens were subjected
to vickers hardness and Izod impact test. Micro-hardness measurements were taken across the
specimen at 300 gf load and 15 s dwell time on Vickers micro-hardness tester (Make: Fuel
Instruments and Engineers (FIE), Mettax laboratory, Chennai, Tamilnadu, India). Indentations were
typically performed at 3 mm spacing.

2
The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

Table 2. Experimental results

Gas
Welding Ultimate
Flow
Current Volatge speed tensile
Std Rate
(A) (V) (WS) strength
(GFR)
(mm/min) (UTS)
(lpm)
1 70 12 55 8 972
2 70 13 60 10 961
3 70 14 65 12 951
4 75 12 60 12 970
5 75 13 65 8 933
6 75 14 55 10 943
7 80 12 65 10 938
8 80 13 55 12 952
9 80 14 60 8 915

3. Optimization of welding parameters

3.1. Response surface methodology

Response Surface Methodology (RSM) is collection of mathematical and statistical techniques for
process development optimization(5). RSM is often accomplished directly for the objectives of
quality improvement including reduction of variability, process enhancement and product
performance (6). The RSM is widely used for setting the parameter and optimization such as
chemical, electronics, manufacturing and metal cutting industries. Lot of process parameters may
affect the output quality responses in the TIG welding process. Therefore, optimization and finding
several independent variables affecting the TIG welding process is quite complex and time-
consuming process. However, using design of experiments and applying regression analysis, process
modelling of the favorite response to numerous independent input variables can be gained. The
response surface can be expressed if all the variables are assumed to be measurable,
y  f ( x1  x2  ....  xk )  
In many engineering analysis, there is a relationship between an output variable on interest ‘y’ and set
of controllable variables ( x1  x2  ....  xk ) , is a noise or experimental error. Usually in most of
the engineering and non-engineering RSM problems the relationship between the input and output
variable are unknown. Thus, finding the suitable approximation for the true functional relationship
between ‘y’ and set of the independent variables is the first step in RSM. Usually the second order
model is utilized in RSM,
k k

y   0    i xi    ii xi2    ij xi x j  
i 1 i 1 i j

The β coefficients can be calculated by means of using least square method in above model. The
second-order model is normally used when the response function is not known or nonlinear (7).

3.2 Genetic algorithm

Genetic algorithm (GA) is used for stochastic problems, which is introduced by Holland (8) based on
mechanics of a biological evolution process. Usually, a decision variable of a coded gene is
represented by a real number, a bit or a string in genetic algorithm. And a corresponding gene of each
parameter creates a chromosome which represents a binary of strings or an array of a problem that can
be varied depending on a treated problem (9). A set of chromosomes is called population. Among
solutions, one population is preferred through a selection operator such as roulette wheel or
tournament selection, where the probability is to be selected relative to individual fitness. The selected
set of individuals is used further to create a new population through genetic operators called mutation
and crossover. Crossover chooses genes from parent chromosomes and builds a new offspring.
Besides, the mutation is preventing all the solutions of the population get trapped into a local

3
The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

optimum. This procedure is repeated under some situations are satisfied, until goodness of best
solution is attained. The GA parameters used in the present investigation is presented in Table 3.

Table 3. GA parameters
Parameters Setting value
Population size 50
Crossover rate 0.8
Mutation rate 0.7
Selection operator Adaptive feasible
Fitness operator Tensile strength

4. Results and Discussion

4.1. Mathematical design of the problem

The maximization of the present fitness function values is subjected to the boundaries of weld
parameters. The weld parameters and their range are shown in Table 4 is selected to show the
limitations of the optimization solution and is given as follows,

Table 4. Weld parameters and their ranges

Parameters Range
Current 70≤A≤80
Voltage 12≤V≤14
Welding speed 55≤WS≤65
Gas flow rate 8≤GFR≤12

In order to apply GA several parameters to be considered while initiating the problem toward the
objective function of the problem. The MATLAB optimization toolbox is used to maximize the TIG
welded AISI4340 steels. The optimal values for the maximization of UTS of TIG welded parameters
obtained from GA is Current (A) = 70A, Voltage (V) = 12V, Welding speed = 55mm/min and Gas
flow rate = 11.807 lpm and it is indicated that the optimal solution is obtained in 52 nd iteration. The
best fitness value obtained from GA for maximizing the UTS of TIG welded AISI 4340 is 988.816
MPa.

Figure 3. GA Convergence plot

Fitness value of the convergence result obtained using GA is graphically presented in Figure 3 that
shows a dotted line signifies that there is no infringement of constrains in each generation. The
maximum tensile strength by GA optimization was observed to be low current (70 A), low voltage (12
V), welding speed (55 mm/s), and the high gas flow rate (11.80 lpm). The optimal values of the
process parameters obtained by the GA is 988.816 MPa. Based on Taguchi L9 orthogonal array nine

4
The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

experimental data was assessed from the experiments. Regression analysis was carried out model for
UTS is,

UTS = +1345.5-2.63×Current -11.83×Voltage -1.50×Welding speed + 4.41× Gas flow rate

The analysis of variance (ANOVA) test was employed to determine the percentage contribution of the
parameters at 95% confidence level and to investigate which process factor significantly affected the
quality characteristic of TIG welding. The result of ANOVA is presented in Table 5. The significance
of each process factor was tested using F – test. In general, if F of experimental runs are higher than
the table 10.128 (from the Standard Fishers table), which means that the varying the process factor
makes a big change on the process performance. In addition, the percentage contribution (P) indicates
the significance of process factors for the response. For a factor with a high values of P, a small
variation will have a great influence on the performance (10). The graphical representation of
percentage contribution is presented in Figure 4, it observed that, current (A) was found to be a major
factor affecting the welding process significantly the mean average of UTS of 38.64%. The voltage
(V), gas flowrate (GFR) and welding speed (F) have an effect on the UTS of 31.21%, 17.39% and
12.54% respectively. And, the predicted and adjusted coefficient of determination (R-Squared) are
99.78% and 99.55% respectively. The “predicted R-Squared” of 99.78% is in reasonable agreement
with the “Adjusted R-Squared” of 99.55%.

Figure 4 Contribution plot

Table 5. ANOVA results

Source df Adj SS Adj MS F value p value
Regression 4 2686.00 671.50 447.67 < 0.0001 Significant
A – Current 1 1040.17 1040.17 693.44 0.003
B – Voltage 1 840.17 840.17 560.11 <0.0001
C – Welding
1 337.50 337.50 225.00 <0.0001
speed
D – Gasflow
1 468.17 468.17 312.11 0.0002
rate
Error 4 6.00 1.50
Total 8 5378.01
S = 1.22 R-Sq = 99.78% R-Squ (adj) = 99.55 R-Sq (Pre) = 98.95%

4.2. Ultimate tensile strength

A SEM image of base materials were taken in order to understand primary phases of the materials and
to ascertain morphological changes in various phases of materials after TIG welding. The developed
SEM images of base materials as presented in Figure 5(a) and Figure 5(b). The temperature of weld
metal with a much lower amount of hot molten mass comparatively to surrounding temperatures were
always low. Hence an air-cooling effect was experienced by the weld metal produces irregular harder

5
The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

(martensite) phases owing to insufficient solidification time. Moreover, a couple of micro holes have
been observed on the weld zone due to rapid solidification of molten metals as shown in Figure 5(a).
This phenomenon is observed in all samples being welded irrespective to the process parameters.
Besides that, from over all micro structures a clear elongated grain of irregular harder martensitic
interfaces was seen in all the fabricated weld joints as shown in Figure 5(b) And in the HAZ region,
exhibited fine acicular ferrite and coarse harder martenstic structure was seen, hence a lower hardness
was observed. The readings were taken in Vickers hardness scale is plotted. Figure 6 shows the
hardness distribution at the cross section of the welded joints of AISI 4340 steel plates. The weld zone
from -5mm to +5mm showed lower hardness value when compared to base materials of AISI 4340
steel plates. Due to more heat capacity of the argon gas flame quickly softening the metals thus leads
the formation of micro holes on the weld zone as shown in Figure 5(a). This tendency in nature,
decreasing the hardness on the weld zone and HAZ. However, the obtained hardness plot is mostly
owing for ductile nature of the joints as evidenced from the fracture locations of welded plates. At
weld nugget zone, the weld metal hardness is achieved was maximum due to the formation of
irregular and harder martensitic structures. The HAZ region exhibited a mix of coarse and fine
microstructures. And due to full penetration welding, cooling rate is slow hence there is no noticeable
changes in the micro structures were observed in the metallurgical properties (4).

Figure 5. (a) Micro holes in TIG weld zone, (b) Harder martenstic interface

Figure 6. Rockwell hardness distribution of fabricated TIG welded samples

The main effects plot for UTS with individual effect of current, welding speed, voltage, and gas flow
rate are presented in Figure 7. It indiactes lower current, welding speed, and voltage maximizing
tensile strength because of lower heat input. The tensile strength increases with increase in gas flow
rate. The reason is that shielding gases are used to prevent atmospheric contamination of the weld
metal. A too low heat input result more weld defects thus tensile strength reduced in the fabricated
weld joints. On the other hand increasing gas flow rate increasing tensile strength of the weld joint.
Since, shielding gas has low ionization potential and is heavier than air, subsequently it giving
exceptional shielding of the molten weld pool which improves the mechanical properties of the weld
joint.

6
 The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

Main Effects Plot for Means

Data Means
Current (A) Volatge (V) Welding speed (mm/min) Gas flow rate (lpm)
965

960

Mean of Means
955

950

945

940
70 75 80 12 13 14 55 60 65 8 10 12

Figure 7. Main effect plot

The interaction effect of welding speed versus current on UTS of TIG welding process parameters is
presented in Figure 8. It can be seen that increasing welding current increases UTS from 927 MPa to
954 MPa. Increasing welding current increases weld penetaration therefore heat input increases
subsequently. Moreover, increasing weld penetration improves fluid flow in the weld pool, which is
driven by the electromagnetic force, surface tension gradient, buoyancy force, and impinging force.
Thus UTS is increases with increasingDesign-Expert®
Design-Expert® Software welding current of the fabricated weld joints.
Software

UTS
UTS 972
972
915
915
X1 = A: Current 974
X1 = C: Welding speed X2 = B: Voltage
X2 = A: Current 969
Actual Factors
Actual Factors 961.25
C: Welding speed = 60.00
B: Voltage = 13.00 958.5 D: Gasflow rate = 10.00
D: Gasflow rate = 10.00
948.5
UTS

948
UTS

935.75
937.5

923
927

55.00 12.00
70
70 57.50 12.50
72.5
72.5
60.00 13.00
75 75
77.5 62.50 C: Welding speed B: Voltage 13.50 77.5
A: Current 80 65.00 14.00 80
A: Current

Figure 8. Welding
Design-Expert®speed,
Software Current vs UTS Figure 9. Welding speed, Gas flow rate vs UTS
UTS
972

915

X1 = D: Gasflow rate
970
X2 = B: Voltage

Actual Factors
A: Current = 75.00 959.25
C: Welding speed = 60.00

948.5
UTS

937.75

927

12.00 12.00
11.00 12.50
10.00 13.00
9.00 13.50
D: Gasflow rate B: Voltage
8.00 14.00

Figure 10. Voltage, Gas flow rate vs UTS

Figure 9. shows the welding current, voltage vs UTS. It is observed that the voltage is directly
responsible to the UTS. For voltage, F value is 694.44 in the ANOVA which is the most significant

7
The 3rd International Conference on Materials and Manufacturing Engineering 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 390 (2018) 012066 doi:10.1088/1757-899X/390/1/012066
1234567890‘’“”

term and consequential term in the TIG welding. In the surface plot, it can be observed that the
incrementing current will decrease the UTS values from 961 MPa to 943 MPa. A change in
microstructure directly affects the mechanical properties of the weld. Consequently, when the heat
input increases highly, the hardness of the welded joint decreases slightly due to microstructure
coarsening. Figure 10. shows surface plot of welding speed, gas flow rate vs. on UTS. When the
voltage varies from 12 to 13.5 V, the UTS extremely decreases from 970 MPa to 950 MPa. In
addition to that, an increasing gas flow rate from 8 to 12 lpm, the UTS increases slightly from 927
MPa to 948.5 MPa. Gases with low ionization potential encourage the ignition of the electric arc, and
those with low thermal conductivity tend to build the arc stability. The corresponding F value is
560.11, which implies that voltage is the most critical term in the conducted A-TIG welding process.

5. Conclusion

In this present investigation, an attempt has been made to optimize the process parameters of A-TIG
welding of AISI 4340 steel. Taguchi L9 orthogonal array has been utilized in the present investigation
to optimize the process parameters of TIG welding process. The mathematical model has been
developed to understand the relationship between the ultimate tensile strength and the individual
factors considered in the experiment namely welding current, voltage, welding speed, and gas flow
rate were used and the ultimate tensile strength of AISI 4340 steel has predicted at 95 % confidence
level. ANOVA has been employed to determine the percentage contributions of the process
parameters. Optimization was carried out with the response surface method based genetic algorithm.
Confirmation test reveals that GA can be utlized effectively to evaluate input process parameters to
get the required ultimate tensile strength.

References

1. Zhu Y-P, Li C-Y, Zhang L-Y. Corrosion resistance and antibacterial activity of different
zones in TA2 weldment by TIG welding. Rare Metals. 2014.
2. Bhattacharya A. Revisiting Arc, Metal Flow Behavior in Flux Activated Tungsten Inert Gas
Welding. Materials and Manufacturing Processes. 2015;31(3):343-51.
3. Singh R, Slathia RS. Effect of Cryogenic Treatment on Sensitization of 304 Stainless Steel in
TIG Welding. Journal of The Institution of Engineers (India): Series C. 2015;97(2):127-30.
4. Joseph J, Muthukumaran S. Optimization of activated TIG welding parameters for improving
weld joint strength of AISI 4135 PM steel by genetic algorithm and simulated annealing. The
International Journal of Advanced Manufacturing Technology. 2015.
5. Vijayan D, Rao V. Friction Stir Welding of Age-Hardenable Aluminum Alloys: A Parametric
Approach Using RSM Based GRA Coupled With PCA. Journal of The Institution of Engineers
(India): Series C. 2014;95(2):127-41.
6. Vijayan D, Seshagiri Rao V. Parametric optimization of friction stir welding process of age
hardenable aluminum alloys−ANFIS modeling. Journal of Central South University.
2016;23(8):1847-57.
7. Palanikumar K. Application of Taguchi and response surface methodologies for surface
roughness in machining glass fiber reinforced plastics by PCD tooling. The International Journal of
Advanced Manufacturing Technology. 2006;36(1-2):19-27.
8. Holland JH. Adaptation in natural and artificial systems: an introductory analysis with
applications to biology, control, and artificial intelligence: U Michigan Press; 1975.
9. D.Vijayan VSR. Tensile Properties Improvement on Friction Stir Welded Age-Hardenable
Aluminum Alloys: An Evolutionary Approach using RSM based GA and SA. Revista Tecnica De La
Facultad De Ingenieria Universidad Del Zulia; 2016. p. 55 - 70.
10. Souissi N, Souissi S, Lecompte J-P, Amar MB, Bradai C, Halouani F. Improvement of
ductility for squeeze cast 2017 A wrought aluminum alloy using the Taguchi method. The
International Journal of Advanced Manufacturing Technology. 2015;78(9-12):2069-77.