Development of empirical models for prediction of weld bead geometry in robotic - GMAW

Purpose: The objective of this paper is to determine the input-output relationship of robotic gas metal arc welding process using linear as well as second order linear regression analysis. Design/methodology/approach: Taguchi’s L27, 3 level 4 parameter orthogonal array design of experiments and multiple regression techniques has been utilized for the development of empirical model. Arc current, stick-out, arc voltage and travel speed is taken as input parameters and bead geometry has been taken as output responses. The effects and interaction terms on different responses of these selected welding parameters have been analyzed using ANOVA. Findings: Both techniques results were compared and concluding remarks have been made. The developed empirical model has been found good agreement with the experiment results and predicted error for second order polynomial regression equations lies between 0.58% to 14.86% for bead height, 0.93% to 9.44% for bead width and 0.34% to 2.56% for bead penetration using with actual experimental results. Research limitations/implications: It was noticed that interaction effects have considerable influence on the formation of weld bead geometry, so it cannot be ignored. Originality/value: In this present work, an effort has been made to carry out both first as well as second order linear regression analyses on robotic GMAW by L27, Taguchi’s design of experiments.