As the die casting technique has been gradually applied to the preparation of large, complex, and integrated structural parts, there is an urgent need for non-heat-treatable aluminum alloys with superior comprehensive properties. Traditionally, the development cycle of new alloys requires 5 to 10 years using conventional experimental approaches from compositional design to industrial application. With the implementation of the Materials Genome Initiative, computation-driven design methods have substantially improved development efficiency, halving the required timeframe. However, such methods still fall short in enabling rapid product iteration and application. In this study, a multi-objective, layer-by-layer alloy screening strategy under an iterative learning framework was established to accelerate the development of novel non-heat-treatable die casting aluminum alloys with superb mechanical properties. Within 5 months, a novel alloy with ultimate tensile strength (UTS) exceeding 290 MPa, yield strength (YS) over 145 MPa, and elongation (EL) above 12% was successfully designed and industrially validated. The alloy outperforms previously reported counterparts in comprehensive mechanical properties. The performance was validated through flat die casting and flowability casting trials, and further applied in the integrated die casting of a full-scale automotive rear floor component. The successful demonstration clearly shows the high efficiency and practicality of the present design method, and is readily applicable to the design of other advanced structural alloys for parts.
Yang et al. (Tue,) studied this question.