RSW parameter optimization and quality prediction of aluminum alloy and low-alloy carbon steel dissimilar joints

Abstract

This study conducts an orthogonal experiment to determine the optimal welding parameters for resistance spot welding (RSW) between aluminum alloy and low-alloy carbon steel sheets. The microstructure and quality indicators (nugget diameter, tensile-shear strength, and indentation rate) of the RSW joints are comprehensively analyzed. Additionally, the quality of the RSW joints, including tensile-shear strength, nugget diameter, and indentation rate, is predicted using the crested porcupine optimizer-support vector regression (CPO-SVR) algorithm. Results from the orthogonal experiment demonstrate that the parameter combination of 16 kA welding current, 250 ms welding time, and 3.5 kN electrode force produces the RSW joints with superior performance, including a tensile-shear strength of 2.285 kN, a nugget diameter of 8.1 mm, and an indentation rate of 42%. Compared with the optimal RSW joint in the control variable experiment, the tensile-shear strength increases by 40 N, the nugget diameter decreases by 0.1 mm, and the indentation rate decreases by 10.6%. The range and variance analysis reveal that welding current has the most significant effect on the tensile-shear strength of the RSW joints, followed by electrode force, whereas welding time has the least effect. For the microstructural analysis, it is found that an elliptical nugget forms at the center of the low-alloy carbon steel side, while a bowl-shaped nugget fully penetrates the aluminum sheet on the aluminum alloy side. A continuous and thin bilayer intermetallic compound (IMC) layer is identified at the interface between aluminum alloy and low-alloy carbon steel, with a thickness of 1 µm. The IMC layer exhibits a needle-like morphology on the aluminum alloy side and a tongue-like morphology on the low-alloy steel side. The CPO-SVR algorithm can accurately predict the quality of RSW joints. For shear strength, the root mean square error (RMSE), mean absolute error (MAE), and the coefficient of determination (R²) values for the test set are measured at 0.089, 0.07, and 0.921, respectively. For nugget diameter, the RMSE, MAE, and R² values for the test set are 0.289, 0.235, and 0.912, respectively. For indentation rate, the RMSE, MAE, and R² values for the test set are 2.339, 1.798, and 0.936, respectively. These findings provide valuable references for the development of high-performance dissimilar RSW joints.