Mechanical and microstructural characteristics of underwater friction stir welded AA 6061-T6 joints using a hybrid GRA-artificial neural network approach


In this paper hybrid grey relations analysis (GRA) and an artificial neural network (ANN) are applied to study the influence of process parameters on the mechanical properties of friction stir welded aluminum alloy 6061-T6. Thirty experiments were performed by varying tool rotation speed, tool traverse speed, and tool tilt angle to study their effects on ultimate tensile strength, yield strength, percentage elongation, and impact strength of FSW joints. GRA was used to convert all responses into the single response variable, i.e., the grey relation grade (GRG). A feed-forward backpropagation ANN with two hidden layers composed of 9 and 7 neurons each was used to simulate the weld joint characteristics in terms of GRG. ANOVA analysis was used to study the influence of process parameters on grey relation grade. It was found that tool rotation speed has a significant impact on weld characteristics, followed by traverse speed and tilt angle. Based on the results it was revealed that tool rotation speed contributes 39.89% to the mechanical properties of underwater friction stir welding of AA 6061-T6, followed by tool traverse speed and tool tilt angle, respectively, by 29.87% and 19.59%. The tensile test demonstrates that the underwater FSW joint is approximately 8% stronger than the conventional air FSW joint due to grain refinement and increased nugget zone hardness because of less heat exposure and absorption.