Parts | Days | Selection | Search | Updates | Downloads | Help

MM: Fachverband Metall- und Materialphysik

MM 7: Mechanical Properties and Alloy Design II

MM 7.3: Talk

Monday, March 9, 2026, 16:15–16:30, SCH/A215

A machine-learning approach to investigate deformation mechanisms in Mo-Si-based alloys — •Julie Hammoud and Karsten Albe — Technische Universität Darmstadt, Darmstadt, Germany

Mo-Si-X alloys have gained significant attention for high-temperature applications such as turbine blades, due to their outstanding creep resistance, thermal stability and corrosion resistance. In general, Mo-Si-X alloys include three phases: two intermetallic phases (Mo₃Si and Mo₅SiX₂) and a Mo solid solution phase Mo_ss. Despite its limited oxidation resistance, the solid solution phase enhances the room temperature fracture toughness of the system. However, the Mo solid solution also incorporates Si-rich subphases with a complex microstructure that leads to an unconventional and still insufficiently understood solid solution strengthening behaviour. Previous studies indicate that small additions of Si (0.1 wt.%) soften Mo, motivating a closer examination of this effect. In this study, we employ a machine-learning interatomic potential (MLIP) based on the Atomic Cluster Expansion (ACE) formalism for the Mo-Si binary system. MLIPs have demonstrated their capability to combine first principles calculations accuracy with the scalability characteristic of empirical potentials. To investigate how Si affects the mechanical response of Mo, we examine two aspects: (i) the concentration dependence of Si on the elastic moduli, analyzed within the framework of the Fleischer model, and (ii) the influence of Si on the behaviour of screw dislocations activity under external loading.

Keywords: Mo-Si-X alloys; Refractory alloys; Machine-learning interatomic potential; Dislocation dynamics; Solid solution softening