Research Inventy : International Journal of Engineering and Science

Research Inventy: International Journal Of Engineering And Science
Vol.3, Issue 5 (July 2013), PP 30-41
Issn(e): 2278-4721, Issn(p):2319-6483, Www.Researchinventy.Com
30
Prediction and optimization of stainless steel cladding deposited
by GMAW process using response surface methodology, ANN and
PSO
P, Sreeraj1
, T, Kannan 2
, Subhasis Maji3
1
Department of Mechanical Engineering, Valia Koonambaikulathamma College of
Engineering and Technology, Kerala, 692574 India.
2
Principal, SVS College of Engineering,Coimbatore,Tamilnadu,642109 India.
3
Professor, Department of Mechanical Engineering IGNOU, Delhi,110068, India.
ABSTRACT: Now a day’s gas metal arc cladding became an important process because it allows deposition of
thick protective coatings on substrates. This article presents an experimental investigation of the influence of
processing parameters on clad angle in GMAW process. Because of high reliability, easiness in operation, high
penetration good surface finish and high productivity gas metal arc welding (GMAW) became a natural choice
for fabrication industries. This paper presents five level four factor central composite rotatable designs with full
replication technique to predict critical dimensions of clad angle. The clad angle is determined from
mathematical expression relating to clad height and clad width. Using regression analysis a mathematical
model is developed. The developed model has been checked for adequacy and significance. The main and
interaction effects of process variables and clad angle are presented in graphical form. Using artificial neural
network clad angle is predicted. Again using particle Swarm Optimization (PSO) parameters were optimized.
KEY WORDS: GMAW, Weld bead geometry, Multiple Regression, Mathematical model, ANN, PSO.
I. INTRODUCTION
In order to extend the life of many components such as mould and dies, their surface needs continuous
repairing. This will not only increase the life but also reduce the operating cost under desired circumstances. Gas
Metal arc welding is a common tool that can repair the defects of die and mould so has to enhance their service
life [1]. The quality of a weld depends on mechanical properties of the weld metal which in turn depends on
metallurgical characteristics and chemical composition of the weld. The mechanical and metallurgical feature of
weld depends on bead geometry which is directly related to welding process parameters. In other words quality
of weld depends on in process parameters.GMA welding is a multi objective and multifactor metal fabrication
technique. The process parameters have a direct influence on bead geometry.
Fig 1 shows the clad bead geometry. Mechanical strength of clad metal is highly influenced by the composition
of metal but also by clad bead shape. This is an indication of bead geometry and clad angle. It mainly depends
on wire feed rate, welding speed, arc voltage etc. Therefore it is necessary to study the relationship between in
process parameters and bead parameters to study clad bead geometry. This paper highlights the study carried out
to develop mathematical, models to predict clad angle, in stainless steel cladding deposited by GMAW [2].
Figure 1: scheme of a typical bead geometry: clad height H ; clad width W and clad angle, =180-
2arctan (2H/W).

Prediction and optimization of in stainless...
31
II. EXPERIMENTATION
The following machines and consumables were used for the purpose of conducting experiment.
1) A constant current gas metal arc welding machine (Invrtee V 350 – PRO advanced processor with5 – 425
amps output range)
2) Welding manipulator
3) Wire feeder (LF – 74 Model)
4) Filler material Stainless Steel wire of 1.2mm diameter (ER – 308 L).
5) Gas cylinder containing a mixture of 98% argon and 2% of oxygen.
6) Mild steel plate (grade IS – 2062)
Test plates of size 300 x 200 x 20mm were cut from mild steel plate of grade IS – 2062 and one of the surfaces
is cleaned to remove oxide and dirt before cladding. ER-308 L stainless steel wire of 1.2mm diameter was used
for depositing the clad beads through the feeder. Argon gas at a constant flow rate of 16 litres per minute was
used for shielding [3]. The properties of base metal and filler wire are shown in Table 1. The important and
most difficult parameter found from trial run is wire feed rate. The wire feed rate is proportional to current.
Wire feed rate must be greater than critical wire feed rate to achieve pulsed metal transfer. The relationship
found from trial run is shown in equation (1). The formula derived is shown in Fig 2.
Wire feed rate = 0.96742857 *current + 79.1 --------------------- (1)
The selection of the welding electrode wire based on the matching the mechanical properties and physical
characteristics of the base metal, weld size and existing electrode inventory [4]. A candidate material for
cladding which has excellent corrosion resistance and weld ability is stainless steel. These have chloride stress
corrosion cracking resistance and strength significantly greater than other materials. These have good surface
appearance, good radiographic standard quality and minimum electrode wastage. Experimental design used for
this study and importance steps are briefly explained.
Table 1: Chemical Composition of Base Metal and Filler Wire
Elements, Weight %
Materials C SI Mn P S Al Cr Mo Ni
IS 2062 0.150 0.160 0.870 0.015 0.016 0.031 - - -
ER308L 0.03 0.57 1.76 0.021 1.008 - 19.52 0.75 10.02
Figure 2: Relationship between Current and Wire Feed Rate
III. PLAN OF INVESTIGATION
The research work is carried out in the following steps[5] .Identification of factors, finding the limit of
process variables, development of design matrix, conducting experiments as per design matrix, recording
responses, development of mathematical models, checking adequacy of developed models, and predicting the
parameters and optimization process parameters using PSO.

32
3.1 Identification of factors and responses
The basic difference between welding and cladding is the percentage of dilution. The properties of the
cladding is the significantly influenced by dilution obtained. Hence control of dilution is important in cladding
where a low dilution is highly desirable. When dilution is quite low, the final deposit composition will be closer
to that of filler material and hence corrosion resistant properties of cladding will be greatly improved. The
chosen factors have been selected on the basis to get minimal dilution and optimal clad bead geometry [1].
These are wire feed rate (W), welding speed (S), welding gun angle (T), contact tip to work to The following
independently controllable process parameters were found to be affecting output parameters distance (N) and
pinch (Ac), The responses chosen were clad bead width (W) and bead height (H).The responses were chosen
based on the impact of parameters on final composite model.
3.2 Finding the limits of process variables
Working ranges of all selected factors are fixed by conducting trial run. This was carried out by varying
one of factors while keeping the rest of them as constant values. Working range of each process parameters was
decided upon by inspecting the bead for smooth appearance without any visible defects. The upper limit of
given factor was coded as -2. The coded value of intermediate values were calculated using the equation (2)
= ---------------- (2)
Where Xi is the required coded value of parameter X is any value of parameter from Xmin – Xmax. Xmin is the
lower limit of parameters and Xmax is the upper limit parameters [4].
The chosen level of the parameters with their units and notation are given in Table 2.
Table 2: Welding Parameters and their Levels
Parameters Factor Levels
Unit Notation -2 -1 0 1 2
Welding Current A 1 200 225 250 275 300
Welding Speed mm/min S 150 158 166 174 182
Contact tip to work distance mm N 10 14 18 22 26
Welding gun Angle Degree T 70 80 90 100 110
Pinch - Ac -10 -5 0 5 10
3.3 Development of design matrix
Design matrix chosen to conduct the experiments was central composite rotatable design. The design matrix
comprises of full replication of 25
(= 32), Factorial designs. All welding parameters in the intermediate levels (o)
Constitute the central points and combination of each welding parameters at either is highest value (+2) or
lowest (-2) with other parameters of intermediate levels (0) constitute star points. 32 experimental trails were
conducted that make the estimation of linear quadratic and two way interactive effects of process parameters
on clad angle. [5].
Figure 3: GMAW Circuit Diagram

33
Table 3: Design Matrix
Trial Number
Design Matrix
I S N T Ac
1 -1 -1 -1 -1 1
2 1 -1 -1 -1 -1
3 -1 1 -1 -1 -1
4 1 1 -1 -1 1
5 -1 -1 1 -1 -1
6 1 -1 1 -1 1
7 -1 1 1 -1 1
8 1 1 1 -1 -1
9 -1 -1 -1 1 -110 1 -1 -1 1 1
11 -1 1 -1 1 1
12 1 1 -1 1 -1
13 -1 -1 1 1 1
14 1 -1 1 1 -1
15 -1 1 1 1 -1
16 1 1 1 1 1
17 -2 0 0 0 0
18 2 0 0 0 0
19 0 -2 0 0 0
20 0 2 0 0 0
21 0 0 -2 0 0
22 0 0 2 0 0
23 0 0 0 -2 0
24 0 0 0 2 0
25 0 0 0 0 -2
26 0 0 0 0 2
27 0 0 0 0 0
28 0 0 0 0 0
29 0 0 0 0 0
30 0 0 0 0 0
31 0 0 0 0 0
32 0 0 0 0 0
I - Welding current; S - Welding speed; N - Contact tip to work distance; T - Welding gun angle; Ac–
Pinch
3.4 Conducting experiments as per design matrix
In this work Thirty two experimental run were allowed for the estimation of linear quadratic and two-
way interactive effects of correspond each treatment combination of parameters on bead geometry as shown
Table 3 at random. At each run settings for all parameters were disturbed and reset for next deposit. This is very
essential to introduce variability caused by errors in experimental set up. The experiments were conducted at
SVS College of Engineering, Coimbatore, 642109, India.

34
3.5 Recording of Responses
For measuring the clad bead geometry, the transverse section of each weld overlays was cut using band
saw from mid length. Position of the weld and end faces were machined and grinded. The specimen and faces
were polished and etched using a 5% nital solution to display bead dimensions. The clad bead profiles were
traced using a reflective type optical profile projector at a magnification of X10, in M/s Roots Industries Ltd.
Coimbatore. Then the bead dimension such as height of reinforcement and clad bead width were measured [6].
The profiles traced using AUTO CAD software. This is shown in Fig 4.This represents profile of the specimen
(front side).The cladded specimen is shown in Fig. 5. The measured clad bead dimensions and clad angles is
shown in Table 4.
Figure4: Traced Profile of bead geometry
Figure 5: cladded specimen
Table 4: Design Matrix and Observed Values of Clad Bead Geometry
Trial No. Design Matrix Bead Parameters
I S N T Ac W (mm) H (mm) Clad Angle(α)
1 -1 -1 -1 -1 1 6.9743 6.0262 120.056
2 1 -1 -1 -1 -1 7.6549 5.88735 123.303
3 -1 1 -1 -1 -1 6.3456 5.4519 120.189
4 1 1 -1 -1 1 7.7635 6.0684 122.605
5 -1 -1 1 -1 -1 7.2683 5.72055 128.574
6 1 -1 1 -1 1 9.4383 5.9169 118.758
7 -1 1 1 -1 -1 6.0823 5.49205 125.444
8 1 1 1 -1 -1 8.4666 5.9467 120.561
9 -1 -1 -1 1 -1 6.3029 5.9059 120.383
10 1 -1 -1 1 1 7.0136 5.9833 119.800
11 -1 1 -1 1 1 6.2956 5.5105 122.923
12 1 1 -1 1 -1 7.741 5.8752 122.633
13 -1 -1 1 1 1 7.3231 5.72095 127.303
14 1 -1 1 1 -1 9.6171 6.37445 120.788

35
W-Width; H– Clad Height; W - Width; Clad angle
3.6 Development of Mathematical Models
The response function representing any of the clad bead geometry can be expressed as [7, 8, and 9],
Y = f (A, B, C, D, E) ---------------------------- (3)
Where, Y = Response variable
A = Welding current (I) in amps
B = Welding speed (S) in mm/min
C = Contact tip to Work distance (N) in mm
D = Welding gun angle (T) in degrees
E = Pinch (Ac)
The second order surface response model equals can be expressed as below
Y = β0 + β1 A + β2 B + β3 C + β4 D + β5 E + β11 A2
+ β22 B2
+ β33 C2
+ β44 D2
+ β55 E2
+ β12 AB + β13 AC + β14
AD + β15 AE + β23 BC + β24 BD + β25 BE + β34 CD + β35 CE+ β45 DE--------- (4)
Where, β0 is the free term of the regression equation, the coefficient β1,β2,β3,β4 andβ5 is are linear terms, the
coefficients β11,β22, β33,β44 andß55 quadratic terms, and the coefficients β 12,β13,β14,β15 , etc are the interaction
terms. The coefficients were calculated by using MINITAB 15software. After determining the coefficients, the
mathematical model was developed. The developed mathematical model is given as follows.
Clad Angle(α) = 121.049 - 0.1764A + 0.0635B -0.1099C +0.4044D – 0.7652E +0.1324A2
+0.9524B2
+0.02712C2
+0.1822D2
+0.1874E2
+0.22237AB + 0.5690 AC - 0.1737 AD + 0.1877AE-0.02077BC-
0.4371BD+0.2357BE+0.0577CD+0.2429CE+0.4519DE ---------------- (5)
15 -1 1 1 1 -1 6.6335 5.554 133.988
16 1 1 1 1 1 10.514 5.4645 119.638
17 -2 0 0 0 0 6.5557 5.80585 119.275
18 2 0 0 0 0 7.4772 6.65505 120.367
19 0 -2 0 0 0 7.5886 6.4069 123.341
20 0 2 0 0 0 7.5014 5.6782 116.783
21 0 0 -2 0 0 6.1421 6.0976 127.193
22 0 0 2 0 0 8.5647 5.63655 124.320
23 0 0 0 -2 0 7.9575 5.8281 122.363
24 0 0 0 2 0 7.7085 6.07515 124.249
25 0 0 0 0 -2 7.8365 5.74915 124.378
26 0 0 0 0 2 8.2082 5.99005 123.355
27 0 0 0 0 0 7.9371 6.0153 126.991
28 0 0 0 0 0 8.4371 5.69895 129.704
29 0 0 0 0 0 9.323 5.57595 129.991
30 0 0 0 0 0 9.2205 5.61485 129.704
31 0 0 0 0 0 10.059 5.62095 131.760
32 0 0 0 0 0 8.9953 5.7052 117.697

36
3.7 Checking the adequacy of the developed model.
Analysis of variance (ANOVA) technique was used to test the adequacy of the model. As per this
technique, if the F – ratio values of the developed models do not exceed the standard tabulated values for a
desired level of confidence (95%) and the calculated R – ratio values of the developed model exceed the
standard values for a desired level of confidence (95%) then the models are said to be adequate within the
confidence limit [10]. These conditions were satisfied for the developed model. The values are shown in Table
5.
Table 5: Analysis of variance for Testing Adequacy of the Model
Source DF Seq SS Adj SS Adj MS F P
Regression 20 385.28 385.28 19.26 1.45 0.266
Linear 5 82.30 82.30 16.46 1.24 0.354
Square 5 59.04 59.04 11.81 0.89 0.520
Interaction 10 243.94 243.94 24.39 1.84 0.166
Residual Error 11 145.91 145.91 13.26
Lack-of-fit 6 89.77 89.77 14.96 1.33 0.385
Pure Error 5 56.15 56.15 11.23
Total 31 531.19
IV. VALIDATION OF MODEL
To test the accuracy of the models in actual application, conformity test were conducted by assigning
different values for process variables within their working limits but different from design matrix. Specimens
were cut from conformity plates and their bead profiles were measured. The percentage of errors calculated
using the equation (10). This is shown in Table 7..It is found that average error is less than three percent.
Error = ................................................................. (10)
Table 7. Validation tests
Test
no
Process parameter ACTUAL PREDICTED ERROR (%)
I S N T Ac Clad Angle(α) Clad Angle(αp) (αE)
1 -
1.5
-
1.5
-
1.5
-
1.5
1.5 122.580 124.987 -1.5
2 1.5 -
1.5
-
1.5
-
1.5
1.5 120.087 123.567 -2.34
3 -
1.5
1.5 -
1.5
-
1.5
1.5 125.650 122.678 2.38
V. Artificial Neural Networks
Neural network consists of many non-linear computational elements operating in parallel. Basically it
consists of neurons; it represents our biological nervous system. The basic unit of ANN is the neuron. The
neurons are connected to each other by link and are known as synapses which are associated to a weight factor.
An artificial neuron receives signals from other neurons through the connection between them [11]. Each
connection has a synaptic connection strength which is represented by a weight of that connection strength. This
artificial neuron receives a weighted sum of outputs of all neurons to which it is connected. The weighted sum is
then compared with the threshold for an ANN and if it exceeds this threshold ANN fires. When it fires it goes to
higher excitation state and a signal is send down to other connected neurons. The output of a typical neuron is
obtained as a result of non-linear function of weighted sum. It is an adaptable system that can learn relationship
through repeated presentation of data and is capable of generalizing a new previously unseen data. One of the
most popular learning algorithms is the back propagation algorithm. In this study feedback propagation
algorithm was used with a single hidden layer improved with numerical optimization technique called
Levenbery Marguent approximation algorithm (LM) .The topology of architecture of feed forward three Layers
back propagations network is illustrated in Fig 5.

37
Input layer Hidden layer Output layer
Fig 6 Neural network architecture
MAT LAB 7 was used for training the network for the prediction of clad angle [12]. Statistical
mathematical model was used compare results produced by the work. For normalizing the data the goal is to
examine the statistical distribution of values of each net input and outputs are roughly uniform in addition the
value should scaled to match range of input neurons.
This is basically range 0 to 1 in practice it is found to between 01 and 9. In this paper data base are normalized
using the Equation (9)
= 0.1 + )
Xnorm = Normalized value between 0 and 1
X = Value to be normalized
Xmin = Minimum value in the data set range the particular data set rage which is to be normalized.
Xmax = Maximum value in the particular data set range which is to be normalized.
The Levenberg-Marquardt approximation algorithm was found to be the best fit for application because
it can reduce the MSE to a significantly small value and can provide better accuracy of prediction. So neural
network model with feed forward back propagation algorithm and Levenberg-Marqudt approximation algorithm
was trained with data collected for the experiment [13].The data obtained is divided in to two sets, one for
training and other for testing the data in order to avoid over fitting and under fitting of data. First eleven data
from Table 5 used for testing and next seventeen data for training. The lowest MSE obtained for 11 neurons. So
a net work of five input neurons and eleven hidden neurons in a single hidden layer and two output neurons
created for the study. The predicted test data is shown in Table 7. [14].
Table. 7. Actual and predicted parameters (Test)
Trial No. Design Matrix Bead Parameters
I S N T Ac W (mm) H (mm) α αp
1 -1 -1 -1 -1 1 6.9743 6.0262 120.056 123.567
2 1 -1 -1 -1 -1 7.6549 5.88735 123.303 126.780
3 -1 1 -1 -1 -1 6.3456 5.4519 120.189 122.870
4 1 1 -1 -1 1 7.7635 6.0684 122.6056 124.897
5 -1 -1 1 -1 -1 7.2683 5.72055 128.5745 125.657
6 1 -1 1 -1 1 9.4383 5.9169 118.758 120.354
7 -1 1 1 -1 -1 6.0823 5.49205 125.444 127.4531
8 1 1 1 -1 -1 8.4666 5.9467 120.561 121.673
9 -1 -1 -1 1 -1 6.3029 5.9059 120.383 120.243
10 1 -1 -1 1 1 7.0136 5.9833 119.800 125.876
11 -1 1 -1 1 1 6.2956 5.5105 122.923 126.453

38
VI. PARTICLE SWARM OPTIMIZATION
In particle swarm optimization algorithm Pbest is the location of the best solution of a particle achieved
so far. Best is the best location of the best solution that any neighbour achieved so far. Initially random numbers
are generated for each particle and these values are considered as PBest and weights. Velocity is calculated
using the equation (8), and added with the present weights in each link of the neural network. For each particle
the newly calculated weights are compared with PBest weights and the minimum error produced by weights are
stored in PBest[15]. Initial velocity V is assumed to be 1 and GBest is the weights of minimum error produced
particle. New weights are calculated using equation (9).
Velocity[]=wVelocity[]+C1rand1(PBest[]-present[])+C2xrand2(GBest[]-present[]).................(8)
Present [] =present [] +velocity [].............................................................................................. (9)
Where C1 and C2 are two positive constants named learning factors.rand1 and rand2 are two random
functions ranging from [0, 1].w is an inertia weight to control over the impact of previous history of velocities
on current velocities. The operator w plays the role of a balancing the global search and the local search; and
was proposed to decrease linearly with time from a value of 1.4-.5. As such global search starts with a large
weight and then decreases with time to favour local search over global search. When the number of iterations is
equal to the total number of particles, the goal is compared with the error produced by GBest weights. If the
error produced by GBest weights are less than or equal to goal weights in GBest are used for testing and
prediction otherwise weights of minimum error are stored in GBest and iterations are repeated until goal is
reached. Optimization procedure is shown in Fig7.A program with objective function as equation (10) and
constraints taken from Table 2 is written in MATLAB.7 code which is used for optimization in this study
Figure 7 Procedure for proposed PSO to optimize GMAW process parameters
rand1 and rand2 are two random functions in the range [0,1] where C1 and C2 are two positive constants named
learning factors taken as 2 and ‘w’ is the inertial weight taken as 0.5.The parameters used for PSO optimization
are shown in Table 7.
Table 8 Parameters for PSO optimization
Population size 30
Dimension size 5
Inertia weight 0.4–0.9
Velocity factors C1, C2 1.4
Number of iteration allowed 100
6.1. Method for developing PSO model
 Initiate each particle.
 Calculate fitness value of each particle. If the fitness value is better than the best fitness value (Pbest) in
history. Set the current value as new Pbest.
 Calculate Gbest.
 For each particle calculate the particle velocity.

39
6.2. Numerical illustration for developed PSO model.
The numerical illustration for the developed model to find optimal parameters for percentage of
dilution as summarised below [16].
Welding current I = Imin + (Imax - Imin) -rand()
Welding speed S = Smin + (Smax - Smin)- rand()
Contact tip to work
distance
N = Nmin + (Nmax - Nmin) - rand()
Welding gun angle T = Tmin + (Tmax - Tmin) - rand()
Pinch P = Pmin+ (Pmax - Pmin) - rand()
These values are substituted in equation (8) and clad angle is obtained.
X (1) = A = Welding current (I) in Amps
X (2) = B= Welding Speed (S) in mm/min
X (3) = C = Contact to work piece distance (N) in mm
X (4) = D = Welding gun angle (T) in degree
X (5) = E = Pinch (Ac)
Objective function for percentage of dilution which must be minimized was derived from equation 5. The
constants of welding parameters are given Table 2.
Subjected to bounds
200 ≤ X (1) ≤ 300
150 ≤ X (2) ≤ 182
10 ≤ X (3) ≤ 26
70 ≤ X (4) ≤ 110
-10 ≤ X (5) ≤ 10
6.3. Objective Function
f(x)=121.049-0.1764*x(1)+0.0635*x(2)–0.1099*x(3)-0.4044*x(4)-0.7652*x(5)–
0.1324*x(1)^2+0.952*x(2)^2+0.0271*x(3)^2 + 0.1822*x(4)^2 + 0.1874*x(5)^2 + 0.22237*x(1)*x(2) +
0.5690*x(1)*x(3) - 0.1737*x(1)*x(4) + 0.1877*x(1)*x(5) - 0 .02077*x(2)*x(3) - 0.4371*x( 2)*x(4) +
0.2375*x(2)*x(5) + 0.0577*x (3)*x(4) + 0.2429*x (3)*x (5) + 0.451*x(4)*x(5) ------ (10)
(This is the percentage of dilution)
Fig 8. Fitness function graph of clad angle
Table 9 Optimal process parameters
Parameters Range (Coded Value) Actual
Welding current (I) 2 300 A
Welding speed (S) 2 182 mm/min
Contact tip to work distance (N) 2 26 mm
Welding gun angle (T) -1.4 76 degree
Pinch (Ac) 2 10
Clad angle obtained is 105.2532 and optimal process parameters shown in Table 9.

40
Fig 9.Surface plot of clad angle (CA) Vs welding current ad welding speed
Fig 10.Surface plot of clad angle (CA) Vs T and N
Fig. 11. Surface plots of clad angle (CA) Vs I and Ac
Fig. 12. Contour plot of clad angle(CA) Vs Nand T
120
125
-5.0-5.0
-2.5
0.0
2.5
130
135
-2.5
-5.0
2.5
0.0
2.5
CA
I
S
Surface Plot of CA vs I, S
120
125
-5.0-5.0
-2.5
0.0
2.5
130
135
-2.5
-5.0
2.5
0.0
2.5
CA
N
T
Surface Plot of CA vs N, T
120
125
-5.0-5.0
-2.5
0.0
2.5
130
135
-2.5
-5.0
2.5
0.0
2.5
CA
I
Ac
Surface Plot of CA vs I, Ac
T
N
43210-1-2-3-4
4
3
2
1
0
-1
-2
-3
-4
>
–
–
–
< 120
120 124
124 128
128 132
132
CA
Contour Plot of CA vs N, T

41
VII. RESULTS AND DISCUSSIONS
a. A five level five factor full factorial design matrix based on central composite rotatable design technique
was used for the mathematical development of model to predict clad angle of austenitic stainless steel
deposited by GMAW.
b. PSO tool available in MATLAB 7 software was efficiently employed for optimization of clad angle.
c. In cladding by a welding process clad bead geometry is very important for economising the material. This
study effectively used regression and PSO models to predict and optimize clad angle.
d. IN this study two models artificial neural network and PSO system for prediction and optimization of clad
angle in GMAW welding process. In this study it is proved that PSO model is more efficient.
e. Fig 9-11 shows various interaction effects of process parameters on clad angle. Fig 12 shows contour plot of
clad angle, welding gun angle and contact tip to work distance. The optimum clad angle obtained is 105.5
and optimal process parameters shown in Table 9.
VIII. CONCLUSIONS
Based on the above study it can be observed that the developed model can be used to predict clad angle
within the applied limits of process parameters. This method of predicting process parameters can be used to get
optimum clad angle. In this study ANN and PSO was used for achieving optimal clad bead dimensions. In the
case of any cladding process bead geometry plays an important role in determining the properties of the surface
exposed to hostile environments and reducing cost of manufacturing. In this approach the objective function
aimed for predicting clad angle within the constrained limits.
ACKNOWLEDGEMENT
The authors sincerely acknowledge the help and facilities extended to them by the department of
mechanical engineering SVS college of Engineering, Coimbatore, India.
REFERENCES
[1] Kannan,T.;Murugan,N.(2006).Effect of flux cored arc welding process parameters on duplex stainless steel clad quality ,journal
of Material Processing Technology vol.176 pp 230-239.
[2] Kannan,T.; Murugn,N.(2006).Prediction of ferrite number of duplex stainless steel clad metals using RSM, Welding Journal. pp.
91 - 99.
[3] Gunaraj,V.; Murugan, N. (2005). Prediction and control of weld bead geometry and shape relationships in submerged arc welding
of pipes, Journal of Material Processing Technology. Vol. 168, pp. 478 – 487.
[4] Kim, I,S.; Son, K,J.; Yang, Y, S.; Yaragada, P, K, D,V. (2003). Sensitivity analysis for process parameters in GMA welding
process using factorial design method, International Journal of Machine tools and Manufacture. Vol.43, pp. 763 - 769.
[5] Godfrey. C. Onwubolu, Paulo Davim.J,Carlos Oliveira,Cardoso.A Prediction of clad angle in laser cladding by powder using
response surface methodology and scatter search (2007).Optics & Laser Technology. Vol (39) pp.1130-1134.
[6] Serdar Karaoglu.; Abdullah Secgin. (2008). Sensitivity analysis of submerged arc welding process parameters,Journal of
Material Processing Technology. Vol-202, pp 500-507.
[7] Ghosh, P,K.; Gupta, P,C.; Goyal, V,K. (1998) . Stainless steel cladding of structural steel plate using the pulsed current GMAW
process, Welding-314.
[8] Gunaraj, V.; Murugan, N.(1999) .Prediction and comparison of the area of the heat effected zone for the bead on plate and bead
on joint in SAW of pipes,Journal of Material processing Technology. Vol. 95, pp. 246 - 261.
[9] Montgomery,D,C.;(2003).Design and analysis of Experiments,John Wiley & Sons (ASIA) Pvt. Ltd.
[10] Kannan,T.; Yoganath.(2010).Effect of process parameters on clad bead geometry and shape relationships of stainless steel
cladding deposited by GMAW,Int. Journal of Manufacturing Technology.Vol-47, pp 1083-1095.
[11] Edwin Raja Dhas,J.;Kumaanan,S.(2001). Optimization of process parameters of submerged arc welding using non conventional
techniques, Applied soft computing 11, 5198-5204
[12] Giridharan, P.K.; Murugan,N.; (2009). Optimization of pulsed GTA welding process parameters for the welding of AISI 304 L
stainless steel, International Journal of Advanced Manufacturing Technology 40: pp.478 - 489.
[13] Siva, K.; Murugan,N.;Logesh, R. (2009). Optimisation of Weld Bead Geometry in Plasma Transferred Arc Hard faced Austenitic
Stainless Steel Plates using Genetic Algorithm, International Journal of Advanced Manufacturing Technology, 4:pp. 24-30.
[14] Riccardo Poli;James Kennedy; Tim Blackwell.(2007).Particle swarm Optimization an overview , Swarm intelligence 11, pp33-
57
[15] Kim, I,S;Basu,A; Siores, E; (1996) Mathematical Models for control of weld weed penetration in the GMAW Process,
International Journal of Advanced Manufacturing Technology,12:pp. 303-401.
[16] Farhad Kolahan.; Mehdi Heidari; (2010). A New Approach for Predicting and Optimising Weld Bead Geometry in GMAW,
International Journal of Mechanical System Science and Engineering, 2:2: pp.138 -142.

Research Inventy : International Journal of Engineering and Science

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Research Inventy : International Journal of Engineering and Science (20)

More from inventy (20)

Recently uploaded (20)

Research Inventy : International Journal of Engineering and Science