Strength and process parameter optimization of active cooling-assisted laser directed energy deposited CoCrNi medium-entropy alloy based on SOO-SVM and NSGA-II algorithms

Aiming at the heat accumulation problem in the laser directed energy deposition (LDED) of CoCrNi alloy, although active cooling can alleviate thermal effects, an excessively high cooling rate may introduce residual stresses and aggravate the deformation of thin-walled parts. To address this issue, laser power, scanning speed, powder feeding rate, and the temperature of the active cooling device were taken as input variables, while the layer thickness, layer width, and melt pool temperature served as response indicators. Twenty-five sets of orthogonal experiments were carried out based on the Box–Behnken design. A support vector machine prediction model optimized by the stellar oscillation optimization algorithm (SOO-SVM) was established, and the NSGA-II algorithm was employed for multi-objective global optimization. The results show that the optimal combination of process parameters is a laser power of 1492.84 W, a scanning speed of 11.92 mm s −1 , a powder feeding rate of 14.53 g min −1 , and a cooling temperature of 8.83 °C. After optimization, the layer width and thickness increased by 22.1% and 85.4%, respectively, while surface defects were significantly reduced., and the microstructure exhibited uniform equiaxed dendrites, directional dendrites/cellular crystals, and sound columnar-crystal metallurgical bonding. The average microhardness was enhanced by 10.4%, and the longitudinal and transverse tensile strengths of the thin-walled parts were improved by 13.5% and 19.2%, respectively. This study verifies the effectiveness of incorporating cooling temperature into process parameter optimization and provides a theoretical basis for active-cooling-assisted LDED manufacturing of high-performance CoCrNi alloy.