Abstract: The alloying elements and rolling/aging parameters have a critical influence on the microstructure and properties of precipitation-strengthened Cu-Ni-Co-Si alloys. Nevertheless, the complexity of alloy composition, variability in processing parameters, and conflicting relationships among material properties impede the concurrent improvement of overall performance characteristics. To address this, we developed a machine learning design strategy and explored the complex effects of key features through feature engineering and SHapley Additive exPlanations (SHAP) analysis. Subsequently, the NSGA-II algorithm was used to optimize alloy composition, and a multi-step rolling and aging process was developed to further tailor microstructure and performance. The results demonstrate that the alloy exhibits exceptional properties after multi-step rolling and aging. The new alloy achieves ultimate tensile strength, electrical conductivity, and elongation of the values of 911.4 MPa, 44.37 % IACS, and 7.3 %, respectively. The remarkable properties of Cu-Ni-Co-Si-Mg-Ti alloys are primarily ascribed to the retention of specific crystallographic textures, notably Brass and S, alongside the precipitation of alloying elements such as Ni, Co, and Si. Preserving these cold-rolled textures is crucial for maintaining high strength. Additionally, the formation of nanoscale δ-(Ni, Co)₂Si precipitates, enriched with Mg at the interfaces, significantly contributed to precipitation strengthening and dislocation pinning, promoting enhanced conductivity and uniform plastic deformation. These findings establish an interpretable and effective framework for the integrated “composition-process-property” design of precipitation-strengthened copper alloys, providing quantitative insights into the precipitation behavior and texture characteristics of Cu-Ni-Co-Si alloys during rolling and aging.

Key words: Cu-Ni-Co-Si alloys, Machine learning, Microstructure, Mechanical properties, Electrical conductivity