A physics-based machine-learning approach for modeling the temperature-dependent yield strengths of medium- or high-entropy alloys

Machine learning is becoming a powerful tool to accurately predict temperature-dependent yield strengths (YS) of structural materials, particularly for multi-principal-element systems. However, successful machine-learning predictions depend on the use reasonable models. Here, we present comprehensive and up-to-date overview bilinear log model predicting YS medium-entropy or high-entropy alloys (MEAs HEAs). In this model, break temperature, Tbreak, introduced, which can guide design MEAs HEAs with attractive high-temperature properties. Unlike assuming black-box structures, our based underlying physics, incorporated in form priori information. A technique unconstrained global optimization employed enable concurrent parameters over low- regimes, showing that temperature consistent across ultimate strength variety HEA compositions. high-level comparison between MEAs/HEAs those Nickel-based superalloys reveals superior properties selected refractory HEAs. For reliable operations, component, such as turbine blade, made from alloys, may need stay below Tbreak. Once above phase transformations start taking place, alloy begin losing integrity.