Development of mathematical models for predicting mechanical properties of dissimilar DP600 steel and AZ31B magnesium alloy friction stir welds

Abstract

This study presents the development of mathematical models for predicting the mechanical properties of dissimilar DP600 steel-AZ31B magnesium alloy friction stir welded joints. A General Full Factorial Design was employed by considering rotational speed, welding speed, and tool offset as process parameters. Ultimate tensile strength (UTS), yield strength (YS), microhardness (HV), and percentage elongation (%E) were selected as response variables. Second-order polynomial regression models were developed using Python, and their adequacy was evaluated through ANOVA. The developed models exhibited high predictive capability with R² values greater than 0.94 for all responses. The optimum welding condition was obtained at a rotational speed of 765 rpm, welding speed of 31.5 mm/min, and tool offset of 1.5 mm, producing a maximum UTS of 185.708 MPa, YS of 130.506 MPa, microhardness of 207.9 HV, and elongation of 4.508%. The results indicated that rotational speed and tool offset positively influenced the mechanical properties, whereas welding speed showed a comparatively negative effect. The close agreement between experimental and predicted values confirmed the reliability of the developed mathematical models for predicting and optimizing the mechanical performance of dissimilar friction stir welded joints.