Adaptive ANN-PSO algorithm on tensile stress prediction for the dissimilar FSW butt joints of AA6061- AZ61 alloy
Abstract
The dissimilar AA6061 and AZ 61 of respective Aluminium and Magnesium grade were joined using friction1stir1welding by changing the weld transverse speed, axial1load and rotational1speed. A numerical model was created by insisting the Response Surface Methodology (RSM) with aid of design expert software for predicting the tensile stress. The weld parameters were revised for attaining the peak value of the tensile stress by implementing the adaptive algorithm of ANN with PSO. The optimization work is focused to find the maximum tensile stress for the specific weld parameters for the butt joint of AA6061 - AZ61. The numerical prediction for tensile strength by adaptive ANN-PSO is in close agreement with experimental values.