Regression-driven modeling and optimization of ultrasonic-assisted activated flux tungsten inert gas welding on AISI 316L

This study investigates the ultrasonic-assisted activated tungsten inert gas (UA-TIG) welding process, a promising technique for advanced manufacturing. The objective was to model and optimize UA-TIG parameters to enhance weld geometry, specifically achieving a high depth-to-width (D/W) aspect ratio. Activated by SiO₂ nanoparticles as flux, the process employed central composite design (CCD) to examine the influence of welding current, travel speed, and ultrasonic vibration amplitude across five levels. Bead-on-plate welding tests on AISI 316L stainless steel were supported by simulations to identify optimal ultrasonic zones. Using analysis of variance (ANOVA) and response surface methodology (RSM) for optimization, results revealed robust regression modeling with an error margin below 6%. Compared to tungsten inert gas (TIG) and activated flux TIG (A-TIG) methods, UA-TIG welding achieved a substantial D/W improvement, enhancing the ratio by 320% and 56%, respectively. UA-TIG welding also demonstrated the highest microhardness (210 Vickers) among the tested samples and effectively minimized heat affected zone (HAZ) width, showcasing its superior thermal control and weld quality. This work demonstrates UA-TIG’s effectiveness in achieving superior weld geometry with optimized parameters, indicating its potential for widespread application in precision welding.