About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Computational Thermodynamics and Kinetics
|
| Presentation Title |
Foundational Graph Atomic Cluster Expansion for Atomistic Thermodynamic and Kinetic Simulation |
| Author(s) |
Ralf Drautz |
| On-Site Speaker (Planned) |
Ralf Drautz |
| Abstract Scope |
Until recently, atomistic thermodynamic and kinetic simulations were constrained by the computational cost of first-principles methods and the inherent approximations of lattice-based models, limiting studies to small systems and simple chemistries. The development of machine learning interatomic potentials (MLIPs) has been instrumental in overcoming these limitations. This presentation introduces the Graph Atomic Cluster Expansion (GRACE), a framework providing a complete and efficient description of atomic interactions. GRACE unifies many current MLIP approaches and enables the development of foundational parameterizations that span the periodic table.
I will demonstrate that GRACE facilitates robust thermodynamic and kinetic simulations for a wide range of materials, including materials with multi-component compositions. Specific applications will be presented, including computation of phase diagrams and prediction of diffusion kinetics in complex solids. Finally, I will address the incorporation of magnetism and charge and discuss the systematic generation of comprehensive training datasets for creating transferable and accurate foundational MLIPs. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Phase Transformations, Computational Materials Science & Engineering, Modeling and Simulation |