Analytical and Neural Network Analysis on Flux-Coated Aluminium Alloy by Activated TIG Welding with Synthesized Nanocomposites

Abstract

This research focused to synthesize the material by the tungsten inert gas (TIG) welding process with support of appropriate flux coating material. Therefore the required amount of flux coating material was utilized to enhance the mechanical properties of the specified localized welded regions. Hence, this study concentrated to select the nano-SiO2 flux particles that were employed for TIG process. This activated TIG welding composes the flux-coated welding on the base metal of AA5083-H111, as this material was highly reactive with SiO2 by the presence of magnesium precipitates and well synthesized after the welding. The post- and preheat treatment process was achieved before and after welding. The selection of activated TIG process parameters composed of strengthened weld specimens along with constant parameters like electrode tip angle and flow rate, respectively. Initially, the process parameters were designed by the statistical analysis of Box Behnken method with support of regression formulation to determine the optimal solution. The maximum tensile strength was attained at the welding process parameters of welding speed (100 mm/min), voltage (13 V), and current (125 amps). The higher hardness was achieved at the process parameters of welding speed (80 mm/min), voltage (12 V), and current (125 amps), respectively. Finally, the neural network approach was utilized to verify the predicted responses of tensile and microhardness properties. The interaction plots, mean plots, and 3D scatter plots were influenced to enhance the process parameters. In this research, mechanical properties were enhanced by the flux-coated SiO2 and the analytical method also advances the optimal parameters.