Combined high-throughput computational and experimental screening of printable Al alloys from scraps

Kurzfassung

To achieve goals like climate neutrality by 2050 and a circular economy, material discovery must accelerate, especially as raw material costs and supply chain issues rise. Computational screening, advanced characterization, and machine learning enable this progress. A new Al-based alloy for additive manufacturing (AM) was developed from scrap metal mixtures using high-throughput screening and synchrotron-based validation. Designed for aerospace, it meets requirements such as low hot-cracking sensitivity, short solidification intervals, and suitable strength and ductility. Computationally, ~10,000 alloy compositions were generated by varying scrap ratios. CALPHAD simulations assessed phases, solidification, and mechanical properties, while a random forest model and multi-objective evolutionary algorithm identified optimal alloys. Uncertainty analyses accounted for scrap variability. Experimentally, candidate alloys were cast, remelted, and tested under different solidification rates. Synchrotron tomography provided 3D microstructural insights. A promising composition was processed into powder, and in-situ laser powder bed fusion with microtomography confirmed its AM suitability.