Prediction of Angular Distortion in MIG Welding of Stainless Steel 202 Plates

Abstract

Metal inert gas (MIG) welding is a popularly used arc welding process for the joining of materials. The process can weld successfully all the materials for which suitable filler material is available. The process is preferred by the fabrication, automotive, food processing and chemical industries owing to its versatility, all-position welding capability, portability and adaptability to fully automatic and robotic welding. The present study is carried out to understand a very important characteristic of arc welding processes with MIG welding being no exception. During arc welding processes, the weld metal experiences non-uniform and rapid heating and cooling cycles. Moreover, the top surface of the weld receives more heat as compared to the bottom surface of the weld, thereby creating a thermal gradient across the thickness. The high heat content of the top surface results in its greater contraction as compared to its bottom-most layer. The distortion thus occurred in the cross-direction of the weld is known as angular distortion. Angular distortion once occurred may result in poor aesthetics, difficulty in the fitment of welded plates into the parent structure and sometimes rejection altogether. This distortion is very difficult and sometimes economically infeasible to remove during post-weld treatments. It is therefore the endeavour of the weld engineer to keep it to the lowest possible extent. An attempt has been made to relate this distortion with the input welding parameters via a mathematical model so that for any given setting of weld parameters the resulting angular distortion can be predicted and suitable preventive measures can be taken. The experiments have been designed using the statistical approach of central composite rotatable design. The sufficiency of the generated model was verified using analysis of variance (ANOVA) technique. Further, the results were graphically analysed using response surface methodology (RSM).