Application of Response Surface Methodology (RSM) to Predict Penetration Area During TIG Welding at Steady State Condition

Abstract

This study is centered on the application of response surface methodology to predict penetration area during TIG welding at steady state condition. This study developed a mathematical model using the second order quadratic polynomial model to select the best setting for the process parameters to produce the maximum penetration area in TIG welding. The parameters considered in this study are welding current, welding voltage and welding speed. The sampled steel plates were 100 pieces of mild steel coupons measuring 80 x 40 x10mm, the experiment was performed 20 times, using 5 specimens for each run. Argon was used as shielding gas. Tungsten inert gas machine was used in joining the weld materials. Obtained results showed that the second order quadratic polynomial equation is the best fit model which adequately explains the relationship between the processes parameters employed in the study. The results possessed good statistical sufficiency, having a very good strength for predicting the target response. The prediction strength and models reliability factor are in reasonable agreement with each other, the surface plots showed the combined interaction of input parameters on the response, and the ANOVA results revealed that the penetration area is very sensitive to the current parameter. The mathematical models developed have a high goodness of fit, adequate signal, strength which can predict the penetration area. The numerical optimal solution produced welded material having penetration area of 80.43mm² with a welding speed of 2.6mm/s, a current of 190Amp and voltage of 20.73 volt. RSM model selected the quadratic model as the suitable model for the heat input because it has a P value of < 0.05 as analyzed by analysis of variance table.