Soft Computing Techniques in Stainless Steel Welding

Abstract

Advancement in Soft Computing is a live process in which new and newer methods of solving nonlinear systems are continuously emerging. Fuzzy logic systems, artificial neural networks, and evolutionary computation are the main methodologies of soft computing. Soft Computing techniques are being increasingly used in solving the complex problems of welding both at the scientific and engineering level. In welding, soft computing is finding applications in real time control of the welding process, adaptive control of the welding process, weld-pool geometry control, quality monitoring, intelligent sensing, seam tracking control, robotic welding, prediction of microstructure, mechanical properties, residual stresses and distortion in welds, and others. The present article is focused on the author's experiences in using core soft computing techniques in modeling and prediction of microstructures in stainless steel welds, modeling weld-bead geometry, optimization of process parameters during arc welding of stainless steels, and comparison of the performance of different soft computing techniques.