| Abstract Scope |
Predicting material behavior at extreme temperatures is essential for designing advanced structural and functional materials. We present a computational framework combining density functional theory, molecular dynamics, and machine learning to model melting, diffusion, entropy, and high-temperature crystal structures. The SLUSCHI package interfaces with VASP and LAMMPS to simulate complex thermodynamic processes, while the MAPP platform enables property prediction directly from chemical formulas using deep learning. Together, these tools have enabled over 300,000 calculations, including melting point predictions for 5,000+ minerals and design of novel refractory systems like Hf-C-N and KCl-based molten salts. Case studies highlight approaches to entropy calculation, crystal structure prediction, and diffusion modeling, offering insights into disordered and partially melted states. All tools are openly available to the community, supporting a broad range of applications in high-temperature materials discovery and design. |