DOI : 10.17577/IJERTCONV3IS16031
- Open Access
- Total Downloads : 10
- Authors : P. Surendran, B. Senthilkumar, And T. Kannan
- Paper ID : IJERTCONV3IS16031
- Volume & Issue : TITCON – 2015 (Volume 3 – Issue 16)
- Published (First Online): 30-07-2018
- ISSN (Online) : 2278-0181
- Publisher Name : IJERT
- License:
This work is licensed under a Creative Commons Attribution 4.0 International License
Optimization of bead Geometry in Austenitic Stainless Steel Cladding using Taguchi’s Method and Multi Objective Genetic Algorithm
P. Surendran, B. Senthilkumar, and T. Kannan.
Abstract—-This paper presents the effect of welding parameters on bead geometry in multipass gas metal arc welding (GMAW) process. Three beads are deposited with 20%overlap on the preceding bead. The experimental rune have been conducted by varying welding voltage, wire feed rate, welding speed and NTPD. The objective of this cladding process is to maximize the bead geometry to cover the given surface area in minimum number of welding passes. The bead geometry has been governed by various welding parameters like electrode melting and deposition rate. The use of Taguchi design of experiments enabled to study and rank the process parameters based on their influence on the response. It was found that the wire feed rate contribute more than the other parameters on the weld bead width. The welding speed was decisive in the case of reinforcement height.
Key wordsGas metal arc welding, Taguchi method, Bead geometry, Microstructure studies, Genetic algorithm, overlapping.
1 INTRODUCTION
Welding is the process of joining two metals. It has its application in many fields such as automobile, civil construction, manufacturing sectors etc. It is widely used in the fabrication of boiler, pressure vessel, nuclear reactor etc. precision control of heat input during welding is more important to ensure bead dimensions and corrosion resistance. The weld deposits were normally linked to the main cord by serious of repeated welding process, the overlapping of weld may complicate the situation even more. In welding plates of different sizes are used depending upon its application related to pressure. The gas metal arc welding has gone into various refinements in the past decades. The gas metal arc welding can produce high efficiency with respect to deposition. [1] The shield metal protect the arc from change in metal matrix during its travel from the nozzle to the plate. Stainless steel produce high tensile strength. The use of stainless steel electrode improves the strength of the weld metal. Stainless steel have a good resistance to corrosion. These steel have good impact strength, making it suitable to cryogenic tests.
The welding process engage vast quantity of heat transfer. It would be far difficult to simulate the process theoretically. This process involves standardization of welding process the specimens are welded with different constrains. [13] The temperature that is obtained for the period of multipass welding affects various mechanical and micro structural parameters.
-
20% overlapping model
The single weld be a din joining applications, each layer in the GMAW based deposition process consists of continuous overlapping beads. A simple sketch of the overlapping model is shown in Fig.1. The area of valley and overlapping area in adjacent beads are depicted. The center distance of adjacent bead side fined as d. [3] The view to simplifying the overlapping model, three assumptions are proposed:
-
The cross section profile of single weld be a dissymmetrical.
-
Every weld bead with the same welding parameter has uniform cross section profile.
-
The section profile of a single weld bead remains unchanged during the overlapping of adjacent beads.
-
-
PROPOSAL OF EXPLORATION
The research work was prearranged to be conceded out in the subsequent steps [5]:
-
Evolving the design matrix and steering the experiments as per the design matrix.
-
Recording the responses bead width (W) and height.
-
Emerging the mathematical models to influential the signal to ratio (larger the better).
-
Presenting the effects of ranking the process parameters and the graphical form and scrutinizing the results.
-
Optimization of process parameters to obtain the desired weld bead geometry and dilution.
-
-
EXPERIMENTAL PROCEDURE
The base material used for experimental work is low carbon structural steel IS: 2062:2011(Grade C), the dimensions of the work piece are 20mm thickness and its length and breadth are 12 and 5 inches. The compositions are followed.
Table 1: Chemical composition of the base plate.
|
Material |
C % |
Mn.% |
Elements S % |
P % |
Si % |
|
Max. |
Max |
Max |
Max |
Max |
|
|
IS2062 C |
0.2 |
1.5 |
0.04 |
0.04 |
0.4 |
Cladding material is stainless steel 316L electrode and the diameter is 1.2 mm. The compositions for electrode are followed.
Table 2: Electrode chemical composition.
Electrode C% Cr% Ni% Mo% Mn%Si%
material
316L 0.02 17.8- 11.9 2.2 1.6-20.50-
18.4 0.80
-
GMAW process parameters:
Process factors are plays a very significant vital role in this welding procedure. Parameters to be fixed and change for experimental sage.
Welding speed (WS), welding voltage (WV), Wire feed rate (WFR), Nozzle to plate distance (NTPD)
Fixed variables
These variables are important for the welding process. This will be taken according to the collected works assessment.
-
Percentage overlap 20%
-
Inter-pass temperature – 150°C
-
Shielding gas type and flow rate flow 25 lit/min and composition 100%Ar
Fig. 1: Sketch of overlapping beads.
-
-
TAGUCHI TECHNIQUE
Genichi Taguchi has been identified with the advent of what has come to be termed quality engineering. [3] The goal of quality engineering is to move quality improvement efforts upstream from the production phase to the product/process design stage (off-line).As his loss function establishes; his main concern is deviation of a characteristic from its nominal value. Uncontrollable factors (noise) are often responsible for this aberration and, therefore, Taguchis approach to experimental design has as its goal the design of products/process that are robust to these noise factors.
-
Assumptions of the Taguchi method:
The preservative assumption implies that the individual or main effects of the independent variables on routine parameter are separable. In this statement, the effect of each factor can be linear, quadratic or of higher order, but the typical assumes that there exists no cross product effects (interactions) among the specific factors. That means the effect of independent variable 1 on performance parameter does not depend on the different level settings of any other independent variables and vice versa. If at any time, this assumption is desecrated, then the additive of the main effects does not hold, and the variables act together.
-
Designing an experiment:
The design of an experiment consist the following steps. Selection of independent variables
Range of number of level settings for each independent variable Selection of orthogonal array transmission the independent variables to each column
Conducting the experiments investigating the data Inference.
The Taguchi arrays can be derived or looked up. Small arrays can be pinched out manually; large arrays can be derived from deterministic algorithms. Usually, arrays can be found online. The arrays are desinated by the number of parameters (variables) and the number of levels (states). This is further explained later in this article [11]. Analysis of variance on the collected data from the Taguchi design of experiments can be used to select new parameter values to optimize the performance representative. The data from the arrays can be investigated by plotting the data and performing a graphic analysis, regression, bin yield and Fisher's exact test, or Chi-squared test to test significance [9].
An orthogonal array is a type of experiment where the columns for the
Independent variables are orthogonal to one another. Benefits:
-
Conclusions valid over the entire region spanned by the control factors and their settings
-
Large saving in the experimental effort
-
Analysis is easy
To define an orthogonal array, one must identify:
-
Number of factors to be studied
-
Levels for each factor
-
The specific 2-factor interactions to be estimated
-
The special difficulties that would be encountered in running the experiment
We know that with 2-level full factorial experiments, we. Can estimate variable interactions. When two-level fractional factorial designs are used, we begin to confound our interactions, and often lose the capacity to obtain unconfused estimates of main and interface effects. We have also seen
That if the generators are chosen carefully then knowledge of lower order interactions can be achieved under that assumption that higher order interactions are negligible.
Orthogonal arrays are highly fractionated factorial designs. The information they provide is a function of two things
-
The nature of confounding
-
Assumptions about the corporal system.
-
-
HEAT INPUT:
Heat input = 60 V I KJ/mm
S (1)
V- Welding Voltage (V),
I- Welding Current (Amp), S- Welding Speed (m/min).
In the present study, three 3-level and 4-factor process parameters
i.e. welding voltage, welding speed, wire feed rate and standoff distance are considered. The values of the welding progression parameters are shown. The ranges and levels are fixed based on the screening experiments and AWS reference. The interaction effect between the parameters is not considered.
The total degrees of freedom of all process parameters are 8. The degrees of freedom of the orthogonal array should be greater than or at least equal to the degrees of freedom of all the process
parameters. Hence, L9 (34) Orthogonal array was chosen which
has eight degrees of freedom.
Table 4: Process parameters.
Welding parameters
Units
L1
L2
L3
Voltage
V
20
25
30
Wire feed
m/min
5.0800
7.6200
10.160
Speed
m/min
0.1
0.140
0.180
NTPD m 0.015 0.019 0.023
1
1
1
1
1
2
2
2
1
3
3
3
2
1
2
3
2
2
3
1
2
3
1
2
3
1
3
2
3
2
1
3
3
3
2
1
1
1
1
1
1
2
2
2
1
3
3
3
2
1
2
3
2
2
3
1
2
3
1
2
3
1
3
2
3
2
1
3
3
3
2
1
DESIGN MATRIX
-
INVESTIGATIONAL RESULT
For the experimental design, an orthogonal array is used. It consists of a set of experiments where the settings of several products or process parameters to be studied are changed from one experiment to another. Results obtained from the experimentation are studied with the help of S/N ratio and regression analysis. By using these results, optimal cutting parameters for maximum material removal rate are obtained. The analysis is made using the software MINITAB 16.
Table 5: Mediocre bead width and height.
Ex X1 X2 X3 X4 Average Average (voltage) (feed) (speed) (NTPD height width
(V) (m/min) (m/min) (mm)) (mm) (mm)
1
20
5.0800
0.100
15
7
27.77
2
20
7.6200
0.140
19
6.6
28.42
3
20
10.160
0.180
23
6.64
31.94
4
25
5.0800
0.140
23
5.40
23.9
5
25
7.6200
0.180
15
5.78
26.57
6
25
10.160
0.100
19
6.18
35.78
7
30
5.0800
0.180
19
3.67
22.02
8
30
7.6200
0.100
23
6.82
42.36
9
30
10.160
0.140
15
6.86
43.29
Fig 2. Weld bead specimen.
-
RESULT IN MINITAB:
Taguchi Analysis: W versus V, F, S, NTPD
Table 6: Response Table for Signal to Noise Ratios Larger is better
2
29.40
30.03
29.79
29.00
3
30.71
31.30
28.48
30.06
POC
19.30%
41.04%
27.32%
12.32%
Rank
3
1
2
4
2
29.40
30.03
29.79
29.00
3
30.71
31.30
28.48
30.06
POC
19.30%
41.04%
27.32%
12.32%
Rank
3
1
2
4
level V F S NTPD 1 29.84 27.77 30.83 30.03
Main Effects Plot for SN ratios Data Means
V F
19
ratios
ratios
18
f N
f N
17
20 25 30 200 300 400
n
n
S NTPD
19/p>
18
17
100 140 180 15 19 23
Signal-to-noise: Larger is better
Fig 3: Graph for SN ratio for width.
Response Taguchi Analysis: H versus V, F, S, and NTPD
Table 7 Response Table for Signal to Noise Ratios Larger is better.
Frequency
Frequency
level V F S NTPD
Regression Analysis: W versus V, F, S, NTPD The regression equation is
WD = 10.7 + 0.651 V + 0.0622 F – 0.106 S + 0.024 NTPD
Predictor Coef SE Coef T P
Constant
10.75
13.97 0.77
0.484
V
0.6513
0.3351 1.94
0.124
F
0.06220
0.01676 3.71
0.021
S
-0.10575
0.04189
-2.52
0.065
NTPD
0.0238
0.4189
0.06
0.958
S = 4.10425 R-Sq. = 85.7% R-Sq. (adj) = 71.4%
By using Taguchi L9 orthogonal array the response should be optimized using minitab16 software. Further input parameters the best impact solution tabulated and it should be ranked. The graph shows the contributions to the nine combinations are plotted.
Residual Plots for W
Normal Probability Plot Versus Fits
99
3
Re sid ua l
Re sid ua l
90
Percent
Percent
0
50
-3
10
-6
1 -8 -4 0 4 8 25 30 35 40
Residual Fitted Value
Histogram Versus Order
3 3
1
16.58
14.28
16.47
16.29
2
2
15.24
16.10
15.92
14.50
1
3
14.90
16.33
14.33
15.92
0
POC
21.93%
26.72%
27.93%
23.36%
Rank
4
2
1
3
1
16.58
14.28
16.47
16.29
2
2
15.24
16.10
15.92
14.50
1
3
14.90
16.33
14.33
15.92
0
POC
21.93%
26.72%
27.93%
23.36%
Rank
4
2
1
3
-6 -4 -2 0 2 4
0
Residual
Residual
-3
-6
1 2 3 4 5 6 7 8 9
Residual Observation Order
Fig 5: showing the graph for residual plots.
Residual Plots for H
16.5
16.0
15.5
f SN ratios
f SN ratios
15.0
14.5
Mea n
Mea n
16.5
16.0
15.5
15.0
14.5
Main Effects Plot for SN ratios Data Means
V F
20 25 30 200 300 400
S NTPD
100 140 180 15 19 23
Normal Probability Plot Versus Fits
99
0.5
90
Percent
Percent
0.0
50
-0.5
10
-1.0
1
-1 0 1 5 6 7
Residual Fitted Value
Histogram Versus Order
R
e si d u a l
R
e si d u a l
3 0.5
Frequency
Frequency
2 0.0
1 -0.5
-1.0
0
-1.00 -0.75 -0.50 -0.25 0.00 0.25 0.50 1 2 3 4 5 6 7 8
Signal-to-noise: Larger is better
Residual Observation Order
Fig 4: Graph for SN ratio height. Fig 6: showing the graph for residual plots.
Regression Analysis: H versus V, F, S, NTPD. The regression equation is
H = 6.65 – 0.0614 V + 0.00776 F – 0.0119 S + 0.011 NTPD
Predictor Coef SE Coef T P
Constant 6.654 6.274 1.06 0.367
V -0.06142 0.09848 -0.62 0.577
F 0.007762 0.004924 1.58 0.213
S -0.01193 0.01231 -0.97 0.404
NTPD 0.0111 0.1231 0.09 0.934
S = 0.911727 R-Sq. = 69.0% R-Sq. (adj) = 27.7%
-
-
GENETIC ALGORITHM
Fig 6: Pareto chart.
The Pareto chart shows the result for weld bead width and height. For the equation W and H
WD = 10.7 + 0.651 V + 0.0622 F – 0.106 S + 0.024 NTPD
(2)
H = 6.65 – 0.0614 V + 0.00776 F – 0.0119 S + 0.011 NTPD (3)
Table 7: multi objective GA points W and H
-
CONCLUSION
This is the remaining signal to noise ratio graph plotted and get the secluded input parameters belongs their contribution for bead width.
-
To using this signal to noise ratio larger the better solution welding wire feed rate give more contribution to bead width so it ranked first. It is give 41.04% contribution.
-
Welding Speed is the second raked position and it is give 27.32% contribution to the bead width.
-
Welding voltage is the third ranked position and it is give 19.30% impact to the response.
-
Nozzle to plate distance is the least and final ranked position and it is 12.32% contributing.
The secluded input parameters belongs their contribution for bead height.
-
To using this signal to noise ratio larger the better solution welding speed give more contribution to bead width so it ranked first. It is give 27.93% contribution.
-
Welding wire feed rate is the second raked position and it is give 26.76% contribution to the bead width.
-
Nozzle to plate distance is the third ranked position and it is give 23.36% impact to the response.
-
Welding voltage is the least and final ranked position and it is 21.93% contributing
The multi objective GA Pareto chart is drawn by mat lab 10 for the weld bead width and height.
REFERENCES
-
Mohd. shoeb, Mohd. Pervez, and Pratibha Kamari, Effect of mig welding input process parameters on weld bead geometry on hsla steel, International Journal of
1
-38.3516
-7.9695
20.00649
10.15992
0.100234
15.01442
2
-38.3516
-7.9695
20.00649
10.15992
0.100234
15.01442
3
-45.0257
-6.74719
29.99076
10.15534
0.100287
22.97968
4
-42.9118
-7.28326
27.0302
10.15768
0.10035
15.24317
5
-44.6163
-6.98132
29.62473
10.14834
0.100336
16.78445
6
-40.8857
-7.59228
23.90743
10.15902
0.100269
15.03807
7
-41.0444
-7.52357
24.09859
10.15712
0.100252
16.58083
8
-44.6279
-6.91392
29.52038
10.15542
0.100304
19.23511
9
-44.062
-7.08937
28.79615
10.1526
0.100381
15.92364
10
-45.0257
-6.74719
29.99076
10.15534
0.100287
22.97968
11
-44.9444
-6.84147
29.98479
10.15286
0.100301
20.07115
12
-40.1855
-7.66979
22.81138
10.15623
0.10027
15.88006
13
-40.2756
-7.6379
23.01023
10.15844
0.100732
16.05631
14
-39.8009
-7.72976
22.21768
10.15735
0.100261
15.80643
15
-41.6495
-7.45231
25.05487
10.15711
0.100269
15.9308
16
-38.9373
-7.87866
20.92239
10.15837
0.100305
15.0484
17
-38.4061
-7.94066
20.06409
10.15992
0.100235
15.73023
18
-42.061
-7.39181
25.67238
10.15977
0.100254
15.9898
19
-41.4527
-7.50737
24.77725
10.1597
0.100286
15.06983
20
-43.1548
-7.2162
27.3994
10.14891
0.100328
16.15914
21
-39.3286
-7.77876
21.46107
10.15737
0.100239
16.55212
1
-38.3516
-7.9695
20.00649
10.15992
0.100234
15.01442
2
-38.3516
-7.9695
20.00649
10.15992
0.100234
15.01442
3
-45.0257
-6.74719
29.99076
10.15534
0.100287
22.97968
4
-42.9118
-7.28326
27.0302
10.15768
0.10035
15.24317
5
-44.6163
-6.98132
29.62473
10.14834
0.100336
16.78445
6
-40.8857
-7.59228
23.90743
10.15902
0.100269
15.03807
7
-41.0444
-7.52357
24.09859
10.15712
0.100252
16.58083
8
-44.6279
-6.91392
29.52038
10.15542
0.100304
19.23511
9
-44.062
-7.08937
28.79615
10.1526
0.100381
15.92364
10
-45.0257
-6.74719
29.99076
10.15534
0.100287
22.97968
11
-44.9444
-6.84147
29.98479
10.15286
0.100301
20.07115
12
-40.1855
-7.66979
22.81138
10.15623
0.10027
15.88006
13
-40.2756
-7.6379
23.01023
10.15844
0.100732
16.05631
14
-39.8009
-7.72976
22.21768
10.15735
0.100261
15.80643
15
-41.6495
-7.45231
25.05487
10.15711
0.100269
15.9308
16
-38.9373
-7.87866
20.92239
10.15837
0.100305
15.0484
17
-38.4061
-7.94066
20.06409
10.15992
0.100235
15.73023
18
-42.061
-7.39181
25.67238
10.15977
0.100254
15.9898
19
-41.4527
-7.50737
24.77725
10.1597
0.100286
15.06983
20
-43.1548
-7.2162
27.3994
10.14891
0.100328
16.15914
21
-39.3286
-7.77876
21.46107
10.15737
0.100239
16.55212
Engineering Science and Technology (IJEST), (2013) vol.5, pp.200-212.
-
Dr.Robert M Andrews, Harry Kamping, Henk de Haan,
Otto Jan Huising, and Neil A Millwood, Guidance for mechanized GMAW of onshore pipelines,The Journal of Pipeline Engineering, (2013) vol. 12, pp. 277-219.
-
Grujicic .M, Ramaswami .S, Snipes .J.S, Yen .C.F, Cheese man .B.A, and Montgomery .J.S, Multiphysics modelling and simulations of mil A46100 armor-grade martensitic steel gas metal arc welding process,Journal of Materials Engineering and Performance, (2013) vol. 22.10, pp.1-20.
-
Konkol .P.J, Warren .J.L and Hebert .P.A, Weld ability of HSLA-65 Steel for Ship Structures: The higher strength and improved weld ability of HSLA-65 steel provides advantages over conventional DH-36 steel for ship structures, welding research supplement Sponsored by the
American Welding Society and the Welding Research Council, (1998) vol. 77, pp.361s-371s.
-
Sreeraj .T, Kannan .T, and subhasismaji, Optimization of weld bead geometry for stainless steel cladding deposited by GMAW,American journal of engineering research (AJER), (2013) vol.2, pp.178-187.
-
Janez grum, Zoran bergant, Ivan polajnar, Monitoring the gmaw process by detection of welding current, light intensity and sound pressure,journal of Czech Society for Non-destructive Testing, (2012) pp.291-300.
-
Vipin Kumar, Gajendra Singh, and Mohd.ZaheerKhan
.
.
Yusufzai Effects of Process Parameters of Gas Metal Arc Welding on Dilution in Cladding of Stainless Steel on Mild Steel,InternationalMJournal of Mechanical Engineering, (2013) Vol. 2, pp.127-131
-
Abu-Arqub, Omar, Zaer Abo-Hammour, and Shaher Momani. "Application of continuous genetic algorithm for nonlinear system of second-order boundary value problems." Appl. Math 8, no. 1 (2014): 235-248.
-
Wu, S. Q., Min Ji, Cai-Zhuang Wang, Manh Cuong Nguyen, Xin Zhao, K. Umemoto, R. M. Wentzcovitch, and Kai-Ming Ho. "An adaptive genetic algorithm for crystal structure Prediction." Journal of Physics: Condensed Matter 26, no. 3 (2014): 035402.
-
Cioffi, F., J. I. Hidalgo, R. Fernández, T. Pirling, B. Fernández, D. Gesto, I. Puente Orench, P. Rey,and
G. González-Doncel. "Analysis of the unstressed lattice spacing, d 0, for the determination of the residual stress in a friction stir welded plate of an
Age-hardenable aluminum alloyUse of equilibrium conditions and a genetic algorithm." Acta Materialia 74 (2014): 189-199.
-
Zhang, Yanxi, Xiangdong GAO, and Seiji Katayama.
"Weld appearance prediction with BP neural network improved by genetic algorithm during disk laser welding." Journal of Manufacturing Systems 34 (2015): 53-59.
-
Hariharan, Krishnaswamy, NgocTrung Nguyen, Nirupam Chakraborti, Myoung Gyu Lee, and Frédéric Barlat. "MultiObjective Genetic Algorithm to Optimize Variable Drawbead Geometry for Tailor Welded Blanks Made of Dissimilar Steels." steel research international 85, no. 12 (2014): 1597-1607.
-
Senthilkumar, B., T. Kannan, and R. Madesh. "Optimization of Flux Cored Arc Welding process Parameters by using Genetic algorithm."
In International Conference on Advances in, vol. 2014. 2014.
-
Lenka Kuzmikova, Nathan Larkin, Zengxi Pan, Mark Callaghan, Huijun Li and John Norrish, Investigation into
-
Feasibility of hybrid laser-GMAW process for welding high strength quenched and tempered steel, journal on Faculty of Engineering and Information Sciences, (2012) vol. 2, pp. 1-9.
-
Dislav Kolari, Marie Kolarikova, Karel Kovanda, Marek Pantucek, and Petr Vondrou , Advanced Functions of a Modern Power Source for GMAW Welding of Steel, journal on Acta Polytechnica, (2012) Vol. 52, pp. 83-88.
-
So .W.J, Kang .M.J, and Kim .D.C, Weld ability of pulse GMAW joints of 780 MPa dual-phase steel, journal on archives of material science and engineering. (2010)Vol.41, pp.53-60
-
Gomes .J.H.F, Paiva .A.P, Costa .S.C, Balestrass .P.P, and
Paiva .E.J, Weighted Multivariate Mean Square Error for processes optimization: A case study on flux-cored arc welding for stainless steel claddings,European Journal of Operational Research, (2012) vol. 226, pp. 522-535.
-
Boselli1 .M, Colombo.V, Ghedini.E, Gherardi.M and Sanibondi .P, Two-dimensional time-dependent modelling of fume formation in a pulsed gas metal arc welding process, journal of physics d: applied physics,(2013) vol.46, pp.1-10.
-
Sreeraj .T, Kannan .T, and subhasismaji, Optimization of GMAW Process Parameters Using Particle Swarm Optimization, Research Article (2013) ISRN Metallurgy, pp.1-10.
-
Shukla .B.A and Phafat N.G, Analysis of co 2 welding parameters on the depth of penetration of aisi 1022 steel plates using response surface methodology, International Journal of Mechanical Engineering and Technology (IJMET), (2013) vol.4, pp.31-36.
-
Sreeraj .T, Kannan. T, and Subhasis Maji,, Prediction and Control of Weld Bead Geometry in Gas Metal Arc Welding Process Using Simulated Annealing Algorithm,International Journal Of Computational Engineering Research, (2013) vol.3, pp.213-22.
-