Experimental and Numerical Optimization of Tungsten Inert Gas (TIG) Welding Process Parameters Relative to Mechanical Properties of AISI 1018 Mild Steel Plate

Abstract

In this study, central composite design (CCD) in Design Expert
7.01 software was used to generate statistical design of
experiment (DOE) which is an acceptable design approach in
Response Surface Methodology (RSM). Based on the DOE, an
experimental design matrix having six centre points, six axial
points and eight factorial points resulting in twenty experimental
welding runs were generated as input parameters for
experimental TIG welding process and numerical solution using
RSM prediction and optimization. Requirements for the twenty
experimental input welding runs were maximum UTS, maximum
yield strength, minimum strain and minimum elongation, and
welding run No. 8 met the requirement with the following welding
input parameters: 210 amp current, voltage of 21.00 V, gas flow
rate of 19.00 litters/min and welding speed of 3.75 mm/min.
Significant correlations were observed in the four output results
(UTS, yield strength, strain and elongation) obtained from the
control sample, RSM predicted and experimentally determined
results. The solution was selected by design expert as the optimal
solution with a desirability value of 97.29%. Comparing the
output results obtained from the control samples, welding
experiment and RSM predicted values, significant correlations
were observed in RSM predicted and experimentally determined
welding results than that of the controlled samples. From the
tensile test results obtained, the UTS and yield strength were
observed to decrease as the applied force increased while the
strain and elongation increased in the same trend with the
applied force. Results from RSM and the control samples
followed similar trends.