Among different welding techniques, Submerged Arc Welding (SAW) is one of the most widely used processes employed in metal forming industries. In this paper, a Genetic Algorithm approach is proposed to optimally determine SAW process parameters for any desired weld bead geometry. A five-level factorial technique is employed to relate the important process-control variables (welding voltage, wire feed rate, welding speed and nozzle-to-plate distance) to the bead-quality features (penetration, reinforcement, bead width, total volume of the weld bead and dilution). The adequacy of the proposed approach is verified with ANOVA. Then, the developed models embedded to a GA algorithm to determine the best SAW process parameters for any target values of weld bead geometries. Computational results show that GA method can be used effectively for solving complicated and highly non linear equations in prediction and optimization of welding process parameters.

Keywords