International Conference of Advance Research and Innovation (ICARI-2020)

Effects of Different Welding parameters on Weld joint Penetration in TIG

welded Stainless Steel Plate
Mayank Shrivastava, Rajeev Kumar
Department of Mechanical Engineering, IIMT College of Engineering, Greater Noida (UP), India
Abstract: The effects of process parameters of TIG welding on weld depth of penetration of welded joints of stainless steel plate of SS304 grade
has been studied. The optimum combination of process parameters has been suggested to achieve maximum depth of penetration. Design of
experiment (DOE) was used to plan and design the experiment to study the effects of parameter on these stresses. The input parameters considered
in present investigation were welding current, gas flow rate, root face and welding speed. L9 Orthogonal Array (OA) technique was used to
formulate the experimental layout. It was concluded that the maximum depth of penetration (DOP) was found at optimal setting of parameters of
welding current (110 A), gas flow rate (10 LPM), root face (1.5 mm) and welding speed (31.578 mm/min.) and the optimum response value is
5.6961 mm.
Keywords: DOE, OA, S/N ratio, AISI 304 stainless steel, DOP, Taguchi method.

1. Introduction
An austenitic stainless steel has been in use for fabrication work as it inert gas welding (TIG) process to yield favorable bead geometry
has excellent high temperature and corrosion resistance properties. It using Taguchi method and grey relational analysis (GRA). The
has a wide application in chemical, dairy, petrochemicals and process significance of all the factors selected on overall quality
industries. Since 304 stainless steel has superior low temperature characteristics of the welded specimen were evaluated quantitatively
toughness and corrosion resistance, it is used extensively steel out of by analysis of variance (ANOVA). They found that the tensile load,
300 series grade. Among many application of type 304 SS, storing heat affected area (HAZ), area of penetration, bead width and bead
and transportation of liquefied natural gas (LPG) is one of the typical height were improved using GRA in combination with Taguchi
application [1]. TIG welding is a metal joining process in which the method The weld pool geometry of pulsed GTA-welded titanium
arc is established between a non-consumable tungsten electrode and alloy was analyzed by mathematical modeling using four factors, five
the metal being welded as shown in figure 1. Shielding gas (Ar, He, levels, central composite design. The developed model was checked
etc) is used to protect weld pool from the contaminants in the for their adequacy. DOE, ANOVA, and multi-objective optimization
atmosphere. The welding current, welding speeds etc. are the major tool namely, lexicographic method were used to obtain an optimal
process control parameters in TIG welding process. The auxiliary weld pool geometry in titanium alloys, M. Balasubramanian et al.
parameters are the torch angle, welding direction, shielding gas flow (2009) [7]. A back-propagation NN was used to predict the bead
rate etc. Except these variable (primary and auxiliary), wire electrode shape parameters. Also, a genetic algorithm (GA) was used to
diameter and its composition, polarity, and its extension, base optimize the process parameters for desired front height to front width
material composition, type of shielding gas etc. are generally kept ratio and back height to back width ratio, D.S. Nagesh et al. (2010)[8].
constant during welding Kamal pal et al. (2011)[2]. If any one of Intelligent modeling using adaptive neuro fuzzy inference system
these parameters get changes might affect another. (ANFIS) was used for predicting weld bead shape parameters during
activated-flux tungsten inert gas (A-TIG) welding of reduced
activation ferritic martensitic (RAFM) steel. In this work, welding
parameters taken were: current, voltage and torch speed and bead
shape parameters: bead width, depth of penetration, and HAZ width,
S. Vishnuvaradhan et al. (2013) [9]. The properties of welded joints
are affected by a number of parameters. For predicting the quality of
welds, it is required to model the bead shape. A support vector
machine-based fuzzy rules acquisition system (SVM-FRAS) for
modeling of aluminium alloy through the GTAW process was
proposed. It was found that the SVM-FRAS model good
generalization capability as well as high comprehensibility, Xixia
Huang et al. (2009) [10].
2. Materials and methods
Butt weld joints are prepared using GTAW under varied process
Fig.1 Process diagram- TIG Welding [11] parameters of welding as given by L9 orthogonal array of Taguchi
N. Lenin et al. [3] optimized the process parameters of MMAW of method under argon gas shielding. As the metal deposition rate in
dissimilar metals in order to get higher welding strength. Taguchi case of gas tungsten arc welding is mainly governed by welding
methodology was used for analyzing effects of input parameters and current, gas flow rate, welding speed (welding time) and to some
to obtain optimal parameters setting to get high welding strength. L. extent root face. Therefore, these input parameters have been taken
Suresh Kumar [4] studied the mechanical properties of austenitic into consideration for the study and analysis. The three levels of each
stainless steel AISI 304 and AISI 316 and obtained the various of the input parameters have been taken for present study based on
characteristics like strength, hardness, ductility, grain structure, and trial experiment producing sound weld with no defects of porosity
HAZ keeping voltage constant of welded specimen using TIG and and lack of fusion. The selected process parameters and their levels
MIG welding process. S.P. Gadewar [5] investigated the influence of are given in Table1. The experimental set up showing the gas
process parameters of TIG welding like welding current (I), gas flow tungsten arc welding machine and inert gas cylinder has been shown
rate (GFR), specimen thickness (t) on bead geometry of AISI 304 in figure1 and figure 2. The welding machine is Lincoln Electric Italia
stainless steel. The parameters selected for study affected the make with 3 phases, 400V. Stainless steel welded sample joined with
mechanical properties with great extent. Ugur Esme et al. [6] the polarity of direct current electrode negative using different
investigated the multi-objective optimization (MOO) of tungsten process parameters have been depicted in Table 1. The specification
Corresponding Author, of GTAW machine has been revealed in Table 2. The base metal
E-mail address:[email protected]; sheets of dimensions 100mm x 50mm x 6mm were cut on power
Phone No--+91-9911343038 hacksaw machine. As the thickness of the plates is 6mm, a V groove
All rights reserved: http://www.ijari.org butt joint of 450 groove angle is made.

IJARI
99

Electronic copy available at: https://ssrn.com/abstract=3567245

 International Conference of Advance Research and Innovation (ICARI-2020)

Table1: Input parameters and their levels

Manufacturer, Model No. Lincoln Electric Italia,

V260-T
Main Supply Voltage, 400V3phase, 50/60 Hz.,
Frequency & Max. Primary 21 A
Current
Max Observed Power and 14.6 KVA, 80V
No-load voltage
Current with duty cycle 260 A
35%
Current with duty cycle 200 A
60%
Fig. 3: Welded samples
Current with duty cycle 160 A
Table 3: Chemical composition of AISI 304 Stainless Steel
100%
Dimension, Wt., 144×19×25.5 cm, 14/16 (Base Metal: AISI 304 SS)
Production Class kg, IP 21
Element C Cr Mn Ni P S Si Fe

Wt. (%) 0.02 18.90 2.00 10.0 0.043 0.02 0.87 Balance
Table 2: Specification of TIG welding machine
(Filler Metal: ER SS304L)
Process parameters Levels
Element C Cr Mn Ni P S Si Fe
1 2 3
Welding current, A 90 100 110 Wt. (%) 0.03 18-19 2 8-12 0.045 0.03 0.75 Balance
(A)
Gas flow rate, LPM 1.0 1.5 2.0 Table 4: Mechanical Properties of AISI 304 stainless Steel
(B) Gra Densi Mean Thermal Resistiv
Root face, mm (C) 1.0 1.5 2.0 de ty Coefficient of conductivity ity
Welding speed 15.384 21.428 31.578 (g/cc) Thermal (W/m0c) (Ω.cm)a
,mm/min (D) Expansion at 250c t 200c
(µm/m/°C)
AISI 8 16.9 16.2 7.2e-
Weld beads of 50 mm length were deposited along the width using 304 005
2.4 mm filler wires of ER SS304L. Table 3 shows chemical Table 5: Physical Properties of AISI 304 stainless Steel
composition and Table 4 and table 5 shows physical and mechanical Grade Yield Tensile Elongation
properties of AISI 304 stainless Steel used for the study. strength(MPa) Strength(MPa) (%)
AISI 290 621 55
304
2.1 Design of Experiment (DOE)
It is highly essential to design an experiment to determine the effects
of variable and welding parameter on the various welding responses
on a sound basis rather than a commonly employed trial and error
basis in conjunction with a small number of repeat experiments for
confirmation of results. Apart from the trial and error method of
investigation the following techniques and approaches are commonly
employed by researches are Theoretical approach, Qualitative
approach, Qualitative cum dimensional analysis method and General
quantitative approach.
2.2. Plan of investigation
In order to achieve the desired aim, the investigations were planned
to be carried out in the following steps and these are identifying the
welding variables, selection of the useful limits of the welding
Fig.1: Shielding Gas (Argon: 99.9% Pure) parameters, namely, peak current, background current, pulse
frequency and peak current time, developing the design matrix,
Conducting the experiment as per design matrix. And analysis of
results and conclusions.
2.2.1. Identifying the welding variables
The welding variables were identified to develop mathematical
models to predict individual and combined effects of the parameters.
The various parameters selected were Peak Welding Current (A), Gas
flow rate (B), Root Face (C), Welding Time (D).
2.2.2. Selection of process parameters and their limits
The limits of the welding parameters were selected on the basis of
excessive trial runs. The basis of selection of the given range for
various welding parameters was that the resultant weld should have
good bead appearance, configurations and be free from the visual
defects. The three levels selected for each of the four variables are
shown in Table 6.
Fig.2:Welding Power Source

IJARI 100

Electronic copy available at: https://ssrn.com/abstract=3567245

 International Conference of Advance Research and Innovation (ICARI-2020)

Table 6: Levels for various control factors In the present work, the standard S/N ratio used is higher-the-better
Control Levels Unit (HB) as in eqn. (2) as depth of penetration should be as high as
factors 1 2 3 possible
Welding 90 100 110 Ampere R= No. of Repetition
current (A) 3.1.2. Taguchi design experiments in MINITAB
Gas flow 8 10 12 Litre per MINITAB calculates response tables and generates main effects for
rate (B) minute Signal-to-noise ratios (S/N ratios) vs. the control factor and Means
(lpm) (static design) vs. the control factors. In the Taguchi Method the term
Root face 1 1.5 2.0 mm. ‘signal‘ represents the desirable value (Mean) for the output
(C) characteristic and the term ‘noise‘ represents the undesirable value
Welding 15.384 21.428 31.578 mm/min (Standard Deviation) for the output characteristic.
speed (D) S/N ratio used to measure the quality characteristic deviating from
2.2.3 Developing the design matrix the desired value. In S/N ratio, S is de-fined as S= -10 log (M.S.D.)
Factorial design can be represented in the form of design matrix where, M.S.D. is the Mean Square Deviation for the output
where column and row correspond to levels of factors and the characteristic. To obtain optimal welding performance, higher-the
different experimental runs respectively as shown in Table 7. better quality characteristic depth of penetration (DOP) must be
Table 7: Design Matrix taken. As shown in Table 8 the response values for depth of
L9(34) A B C D penetration. The M.S.D. for higher-the -better quality characteristic
1 1 1 1 1 can be expressed as, M.S.D = Σ 1/Pi2, Where, Pi is the value of depth
2 1 2 2 2 of penetration (DOP).
3 1 3 3 3 Table 8: Response values for weld distortion and depth of bead
4 2 1 2 3 penetration (DOP)
5 2 2 3 1 S. Welding Gas Flow Root Welding Depth
6 2 3 1 2 No Current Rate Face Speed of bead
7 3 1 3 2 . (A) (lpm) (mm) (mm/min Penetration
8 3 2 1 3 ) (DOP) (mm)
9 3 3 2 1 1 90 8 1 15.384 241.25
2 90 10 1.5 21.428 36.187
3. Methodology 3 90 12 2 31.578 84.437
3.1 Taguchi Methodology (TM) 4 100 8 1.5 31.578 144.75
In the TM , the optimum level of input process parameters are 5 100 10 2 15.384 241.25
decided on the basis of statistical analysis of experimental results that 6 100 12 1 21.428 241.25
makes the process insensitive to the effect of variations due to 7 110 8 2 21.428 132.687
uncontrollable factors such as environmental temperature, humidity 8 110 10 1 31.578 21.712
and vibration. In this method, the experiments are performed as per 9 110 12 1.5 15.384 579
specially designed experimental matrix known as orthogonal array
(OA) [12, 13]. The selection of OA is based on the total degree of 4. Results and Discussion
freedom (dof) of the process. Mathematically, the dof can be
4.1 Effect of welding parameters on depth of penetration
computed as [13]:
dof = [(number of levels – 1) for each control factor + (number of (DOP) of GTA welded AISI 304 stainless steel plate
levels for A – 1) × (number of levels for B – 1) for each interaction Table 9: Response Table for S/N ratio for DOP (higher-the-better)
+1] (1) Level I GFR RF WS
where, A and B are the interacting control factors. (A) (LPM) (mm) (mm/min)
Taguchi suggests two different routes to carry out the complete
analysis .First, the standard approach, where the results of a single
run or the average of repetitive runs are processed through main effect
1 5.707 4.950 5.403 4.900
and ANOVA analysis (Raw data analysis).The second approach
which Taguchi strongly recommends for multiple runs is to use
signal- to-noise ratio(S/N) for the same steps in the analysis. Taguchi 2 4.557 6.007 5.580 5.547
method uses a statistical measure of performance called signal-to-
noise ratio. The S/N ratio is the ratio of the mean (signal) to the 3 5.777 5.083 5.057 5.593
standard deviation (noise). By maximizing the S/N ratio, the loss
associated can be minimized. The standard S/N ratios generally used
are as follows Higher the better (HB) as in eqn. (2), Lower the better
Delta 1.220 1.057 0.523 0.693
(LB) as in eqn. (3) and Nominal is best (NB) as in eqn. (4).The mean
squared deviation (MSD) is a statistical quantity that reflects the
deviation from the target value. Rank 1 2 4 3

There are nine experiments that were designed as per Taguchi’s

design of experiment. It is expected to get an optimized level for each
factor. The signal-to noise ratio (S/N ratio) should be kept high to get
minimum value of distortion. The S/N ratio is obtained using negative
of logarithmic value of distortion calculated experimentally, which is
continuously decreasing function. So the signal-to-noise ratio is
always kept at maximum value. The calculation of S/N ratio is based
on smaller-the-better model. An additive model of Taguchi method
was taken into consideration to obtain the optimized value for each
response factor. A larger value of S/N ratio corresponds to better
quality characteristics.

IJARI 101

Electronic copy available at: https://ssrn.com/abstract=3567245

 International Conference of Advance Research and Innovation (ICARI-2020)

Fig.3: Percentage Contribution of factors affecting DOP

Table 10: Response Table for mean for DOP
Level I(A) GFR RF(mm) WS
(LPM) (mm/min) Fig.5: Main effects plot of Mean (DOP)
1 15.05 13.88 14.52 13.78 As welding current increases, the heat input and heat transferred to
the weld increases ultimately enhancing depth of bead pe7netration
2 13.17 15.42 14.82 14.76 of GTA welded AISI 304 stainless steel plate.
3 15.18 14.10 14.05 14.85 4.2 Optimal value of depth of penetration (DOP) in the
Delta 2.00 1.54 0.77 1.07
welded specimen
From S/N ratio plot an optimized value for maximum depth of bead
Rank 1 2 4 3 penetration (DOP) was found at I (A) at third level, GFR (lpm) at
second level, root face (mm) at second level and welding speed at
Table 11: ANOVA table for weld DOP
third level. The average of nine experimental results of DOP is (DOP)
Sources DOF SS MS %P avg=5.34666 mm. The optimum value of DOP is calculated as shown:

I(A) 2 2.8158 1.4079 45.9434103 (DOP)optimum=(I)3Avg.+ (GFR)2Avg.+(RF)2Avg.+(WS)3Avg.– 3(DOP)avg.

(5)
GFR (lpm) 2 1.9868 0.9934 32.41743898
(I) 3Avg=4.5566
RF (mm) 2 0.42527 0.212635 6.938878736 (GFR) 2Avg=6.006
WS 2 0.901 0.4505 14.70108343 (RF) 2Avg. =5.580
(mm/min) (WS) 3Avg. =5.5933
Error 0 3(Total DOP) avg. =3x5.34666=16.0398
Total 8 6.13 100.000 Putting all these values in equation (1), we get
Response table for S/N ratio and mean of DOP are shown in Table 9 (DOP) optimal= 5.6961 mm
and Table 10 respectively. From Table 9, it is found that welding 4.3 Confirmation test for depth of penetration (DOP)
current and gas flow rate have maximum effect on DOP of GTA The values from final confirmation test and obtained DOP value is
welded AISI 304 stainless steel plate. The above table gives the mentioned in Table 12. From analysis of result, calculated error is
values of control factors at each level. The mean for depth of bead small of the order of 3.5%. This error is less than 10% which indicates
penetration (DOP) have been shown in Table 4.1. The ANOVA table that the results are in good agreement.
for depth of bead penetration (DOP) has been depicted in Table 11. Table 12: Results of confirmation test for weld distortion
Figure 4 and figure 5 shows the S/N ratio and mean plot Performan Optimal Predicted and Experimental
demonstrating effects of each parameter on DOP of GTA welded ce measure setting of results
AISI 304 stainless steel plate. The percentage contribution of parameters Predict Experimen %Err
different parameters of welding has been revealed in figure 3. ed tal value or
optimal (mm)
value
(mm)
Depth of I3GFR2RF2 5.6961 5.4962 3.5094
penetration WS3
(DOP)

5. Conclusions
This work presents optimization of the process parameters of gas
tungsten arc welding by taking depth of penetration (DOP) as a
response variable. The effects of welding process parameters on
depth of penetration (DOP) using GTAW process have been studied
using Taguchi optimization method to find the optimal process
parameters for depth of penetration (DOP). In this study, selected
parameters were welding current, gas flow rate, root face and welding
Fig. 4: Main effects plot of S/N ratio (DOP) speed. The following conclusion can be drawn for effective welding
It has been observed that the parameters affecting the depth of of stainless steel plate by gas tungsten arc welding process as follows:
penetration relatively high are welding current and gas flow rate 1) Welding current (A) is the most significant parameter for depth of
while the parameters affecting comparatively less are welding speed bead penetration (DOP) during GTA welding. The recommended
and root face. Percentage contribution affecting DOP are welding parametric combination for optimum depth of bead penetration is
current and gas flow rate with the order of 46% and 32% respectively. welding current (110 A), gas flow rate (10 LPM), root face (1.5 mm)
The analysis of variance (ANOVA) for the factors indicated in the and welding speed (31.578 mm/min.) and the optimum response
Table 11 and Table 12 shows that the welding current (I) and gas flow value is 5.6961 mm.
rate (GFR) are the most contributing factor in effecting the DOP of 2) A confirmation experiment was also performed and verified for
GTA welded AISI 304 stainless steel plate and welding speed (WS), the effectiveness of the Taguchi method. The experimental value
root face (RF) have their contribution in decreasing order for DOP. obtained from setting of optimal welding parameters was found

IJARI 102

Electronic copy available at: https://ssrn.com/abstract=3567245

 International Conference of Advance Research and Innovation (ICARI-2020)

5.4962. The % error between predicted optimal and experimental

values of depth of penetration rate was found 3.5094.
References
[1]. Kumar S., Shahi A.S., Effect of heat input on the microstructure
and mechanical properties of gas tungsten arc welded AISI 304
Stainless Steel joints, Materials and Design, 32,2011,3617-
3623.
[2]. Kamal Pal, Surjya K. Pal. Effect of pulse parameters on weld
quality in pulsed gas metal arc welding: A Review, JMEPEG,
20, 2011, 918-931
[3]. Lenin N., Sivakumar M., Vignesh kumar. D, Process parameter
optimization in ARC welding of dissimilar metals, Thammasat
Int. J. Sc. Tech, 15(3), 2010.
[4]. Suresh Kumar L., Verma S.M., Radhakrishna Prasad P., Kiran
kumar P., Siva Shanker T. Experimental Investigation for
welding aspects of AISI 304 and 316 by Taguchi Technique for
the process of TIG and MIG welding, International Journal of
Engineering trends and technology, 2(2), 2011, 28-33.
[5]. Gadewar, S.P., Peravli Swaminadhan, Harkare, M.G., Gawande,
S.H. Experimental investigation of weld characteristics for a
single pass TIG welding with SS304, International Journal of
Engineering Science and Technology, 2(8), 2008, 3676-3686.
[6]. Ugur Esme, Melih Bayramoglu, Yugut Kazancoglu, Sueda
Ozgun, Optimization of weld bead geometry in TIG welding
process using grey relation analysis and Taguchi method,
Materials and technology , 43(3), 2010, 143-149.
[7]. Balasubramanian M., Jayabalan V., Balasubramanian V.
Prediction and optimization of pulsed current gas tungsten arc
welding process parameters to obtain sound weld pool geometry
in titanium alloy using Lexicographic method, JMEPEG, 18,
2009, 871-877.
[8]. Nagesh D.S., Datta G.L. Genetic algorithm for optimization of
welding variables for height to width ratio and application of
ANN for prediction of bead geometry for TIG welding process,
Applied soft computing, 10, 2010, 897-907.
[9]. Vishnuvaradhan S., Chandrasekhar N., Vasuderan M.,
Jayakumar T. Intelligent modeling using Adaptive Neuro Fuzzy
Inference system (ANFIS) for predicting weld bead shape
parameters during A-TIG welding of Reduced Activation
Ferritic-Martensitic (RAFM) steel, 2013.
[10]. Xixia Huang, Fanhuai Shi, Wei Gu, and Shanben Chen. SVM-
based fuzzy rules acquisition system for pulsed GTAW process,
Engineering Applications of Artificial Intelligence, 22, 2009,
1245-1255.
[11]. Manual for Gas tungsten arc welding of Miller Electric Mfg.
Co. USA.
[12]. Ross, P.J. Taguchi technique for quality engineering, New York:
McGraw Hill, 1988.
[13]. Phadke, M.S. Quality engineering using robust design,
Englewood CL: Prentice-Hall, 1989.

IJARI 103

Electronic copy available at: https://ssrn.com/abstract=3567245