Friction stir welding process parameters optimization and weld structure analyses to study mechanical characteristics using hybrid artificial intelligent soft computing techniques

Abstract

Friction stir welding (FSW) is a solid-state fusion welding technique enormously used for welding different aluminum alloy plates. It involves a number of input parameters that significantly affect the different mechanical characteristics of the prepared weld joints. In this study, an attempt is made to optimize the input process parameters of FSW and the diverse mechanical characteristics of the resulting weld. The analysis of ultimate tensile strength (UTS), hardness, and yield strength (YS) is conducted following the test procedures outlined in the American Society for Testing and Materials (ASTM) standards E18-15 and E8-M04. The test specimens are prepared using two different aluminium alloy plates (AA6061-T6 and AA5083-H111). The response surface methodology-based face-centered central composite design (FCCD) strategy is used to develop the experimental design matrix. The input parameters values are varied at three levels. To predict more précised and optimized results, an adaptive neuro-fuzzy inference system (ANFIS) and genetic algorithm (GA), in combination, is exercised by utilizing the Matlab tool. Using field emission scanning electron microscopy (FESEM) images, structural analysis is performed to study behavior at the weld joint area. After conducting the tensile and hardness test, the maximum experimental value of UTS, hardness, and YS are noticed as 325 MPa, 69 HRB, and 265 MPa, respectively. By using the GA-ANFIS hybrid tool, the maximum value of UTS and hardness is noticed as 326.99 MPa and 71.40 HRB, respectively, using triangular TPP with 30.5 mm/minute welding speed and 1200 rpm tool pin profile. The optimized value of YS is observed as 271.56 MPa using cylindrical TPP with 24 mm/minute welding speed and 1600 rpm tool rotational speed. In the present work, FESEM-based images reveal a fine-grain structure in the stir zone for the optimized values of mechanical characteristics. The results obtained using the hybrid soft computing technique GA-ANFIS showed that these techniques are adequate tools for the optimization of FSW input process parameters and mechanical characteristics.