Enabling high-strength and oxidation-resistant refractory complex, concentrated alloys via a machine learning for accelerated materials discovery framework

Refractory complex, concentrated alloys (RCCAs) can be defined as refractory-based alloys comprising four or more elements with near equimolar compositions. Some of these alloys have recently been shown to exhibit remarkable high temperature strength, exceeding that of Ni-based alloys and Mo- and Nb-based silicides. Furthermore, these alloys can exhibit superior oxidation resistance compared to traditional refractory-based alloys, but current strategies have not enabled the formation of a protective oxide scale above 1300 ºC. In this work, we will present a new machine learning for accelerated materials discovery (ML-AMD) framework that utilizes multi-fidelity and multi-cost experiments with physics-based modeling. This framework led to the discovery of new ultra-high strength Al-based solid-solution BCC RCCAs, exceeding strengths found in literature. A new oxidation database of RCCAs and preliminary analysis will be presented, and recent efforts in optimizing oxidation resistance in RCCAs will be shown.