Optimization of process variables of laser sensor assisted robotic GMAW process for mild steel material

In manufacturing industries, industrial robots have been introduced for performing welding process to accommodate intelligent, flexible, and automate welding. It is essential to integrate sensors and welding process parameter modeling for achieving higher weld quality, productivity and reduced cycle time in robotic arc welding process. A new approach formed by combining fuzzy-regression with Enhanced Teaching Learning Based Optimization (ETLBO) algorithm logic has been used in this paper to get optimal robotic welding parameter settings for achieving best weld quality measures. The weld joint quality has been determined by considering measures like weld bead features consisting of depth of penetration, width, height of weld bead, mechanical attributes like ultimate strength, yield strength and microstructural properties microhardness, and Heat Affected Zone (HAZ) width simultaneously. The laser sensor for seam finding has been mounted on welding torch for achieving positional accuracy in every cycle. ANOVA analysis has been performed to detect the crucial welding process variables affecting weld quality for robotic welding significantly. The proposed model has been validated through experimentation with MOTOMAN MA 1440 A welding robot and ArcWorld of C-50 Series arc welding setup and maximized values of weld quality has been obtained.