Dresden 2026 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

MM: Fachverband Metall- und Materialphysik

MM 7: Mechanical Properties and Alloy Design II

MM 7.2: Vortrag

Montag, 9. März 2026, 16:00–16:15, SCH/A215

Atomistic modelling of Ni-based superalloys at real chemical complexity — •Aditya Vishwakarma, Sarath Menon, Thomas Hammerschmidt, and Ralf Drautz — Interdisciplinary Centre for Advanced Materials Simulation, Ruhr University Bochum, Germany

Ni-based superalloys are critical high-temperature high-load materials for jet-engine turbine blades. While ab-initio computation has been essential for fundamental understanding (e.g., in binary NiAl), simulating the full chemical complexity of commercial superalloys has remained computationally infeasible. In this work, we demonstrate how Graph Atomic Cluster Expansion (GRACE) foundational models overcome this limitation to investigate the complex phase stability of the γ and γ’ phases in industrially relevant Ni-based superalloys. We compute the binary NiAl phase diagram, successfully reproducing the critical γ-γ’ phase coexistence. We then extend these phase stability calculations to a CMSX-4 type multi-component composition using hybrid Molecular Dynamics/Monte Carlo simulations. Our results reveal the short-range order (SRO) in the γ phase and quantify long-range order (LRO) in the γ’ phase. We specifically find that refractory alloying elements (Ta, W, Ti) exhibit strong LRO alongside Al across the operational temperature range. These findings validate that foundational machine learning interatomic potentials are now capable of calculating complex phase diagrams and capturing subtle ordering and clustering phenomena in multi-component alloys. This capability marks a crucial step toward realizing faster, computationally-driven design of next-generation, single-crystal Ni-based superalloys.

Keywords: Ni-based superalloy, Atomistic, UMLIPs