Modeling and optimization of Gas Metal Arc Welding process using statistical methods and Simulated Annealing algorithm

Gas Metal Arc Welding (GMAW) using inert gas is an important welding process for making permanent joints in metal parts. The quality of joint in welding is usually expressed by weld bead geometry which includes bead height, width and penetration. Weld bead geometry and joint quality, in turn, depend on the welding parameter settings. In this paper, using regression modeling, the exact mathematical relationships between welding input parameters and weld bead geometry specifications have been established. Then, based on statistical analyses, the best and most fitted model has been selected. The proposed model in the present research has better fit and is simpler than those in related literature. In the second part of this paper, Simulated Annealing algorithm has been employed in order to determine the optimum set of welding parameters values. The comparison between optimization results and actual values demonstrates that the proposed procedure is quite capable in modeling and optimization of GMAW process.