A weight-ensemble dual-target optimization strategy for simultaneously enhancing magnetic and mechanical properties of FeCoNiCuTi high-entropy alloys

While high-entropy alloys are well known for their outstanding mechanical properties, a key research challenge remains in the simultaneous imparting of magnetic, electrical, or other functional characteristics to broaden their range of applications. However, this exploration is hindered by a fundamental microstructural trade‑off: homogenization favors magnetic properties whereas heterogenization enhances mechanical performance. In this study, we develop a novel dual‑property prediction model capable of simultaneously evaluating both soft magnetic and mechanical properties for high-entropy alloys. By employing a weight-ensemble dual-target ensemble strategy, our approach achieves a ~13.8% reduction in the average prediction interval width for enhanced robustness, along with a 3% improvement in prediction accuracy over single models. Guided by the model, we developed a novel FeCoNi-based high-entropy alloy that exhibits a superior combination of high saturation magnetization (113 emu/g) and tensile strength (911 MPa), outperforming current prevalent similar high-entropy alloys. This work establishes a cross-domain integrated model for the simultaneous prediction of mechanical and magnetic properties, thus providing a new way for the synergistic enhancement of dual properties in FeCoNi-based high-entropy soft magnetic alloys.