AI-driven Plasma-assisted Carbothermal Reduction for the Solid-state Synthesis of Cantor and Refractory High Entropy Alloys

The extraordinary mechanical properties of cantor and refractory high-entropy alloys (RHEAs) are on the brink of revolutionizing high-temperature applications, providing a potential substitute for nickel-based superalloys in gas turbines and hypersonic applications. However, traditional melt-derived processing methods to fabricate these alloys encounter severe challenges, including compositional inhomogeneity resulting from elemental segregation and the high melting points of constituent metals. This research proposes the development of an artificial intelligence (AI)-driven Plasma-Assisted Carbothermal Reduction (PACTR) method to augment the reduction of metal oxides, which might potentially offer two noteworthy advantages: (1) compositional homogeneity of the synthesized alloys, and (2) the ability to rapidly heat the raw material mixtures to the desired temperatures. This research will develop innovative models combining machine learning algorithms with CALPHAD methods, phase-field models, DFT calculations, and MD simulations to find a direct link between the manufacturing process, the microstructure that is made, and the mechanical properties of RHEAs. This research will conduct the fabrication of RHEAs via PACTR based on the predicted compositions, processing parameters, and microstructure properties, refine the computational models, and perform iterative optimization to enhance material properties and reduce energy consumption.