Optimization of Tensile Strength in AA6063 Friction Stir Welds using ANOVA and ANFIS: A Statistical Analysis of Process Parameters

Abstract

The tensile strength of 12 mm thick joints created by friction stir welding (FSW) on AA6063 was taken into account through statistical optimization of process parameters based on experimental research. The process variables including the spindle speed, the tool pin length, and the tool pin shoulder rise determine how well this friction stir welding type welds. Experiments were conducted as per full factorial and L9 orthogonal array. By using ANOVA (analysis of variance) the effect of parameters and its significance were examined on the tensile strength. The multiple variable regression equation and ANFIS (artificial neuro fuzzy inference system) based mathematical model were used to determine the optimized tensile strength value with the percent error of 23.27% and 6.04% respectively. As evident from the results the tool pin length has the higher impact with percentage contribution ratio (PCR) of 79.65.