Multiscale modeling of metallic glass by a direct coarse-graining from the atomic scale to the continuum level

• Applicants: Dr. Franz Bamer; Professor Dr.-Ing. Jaan-Willem Simon
• Subject Area: Mechanics
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 558389565

Metallic glass is a relatively young material type with extraordinary properties, such as high elastic strength, exceptional hardness, and high corrosion resistance, and although brittle on the macroscale, it can exhibit legitimate plastic flow at the microscale. However, the mechanical origin of these fascinating properties, yet to be fully understood, presents an exciting opportunity for future research. The missing link between the rich, fascinating nanoscale mechanics and the resulting macroscale behavior is a challenge that, once overcome, will provide a complete, physically meaningful description of the material response. Thus, in this project, we present a multiscale modeling framework of metallic glass structures by directly coarse-graining relevant nanoscale features to a thermodynamically consistent continuum-mechanically based macroscale model. The mechanical response behavior on the nanoscale is governed by elementary point defects that are hard to detect in metallic glasses in contrast to crystalline metals. Using an athermal quasistatic, we bypass the limiting time scale factor of molecular simulations and extract essential features, i.e., response predictors, that are crucial for a physically meaningful continuum description. Furthermore, we will extract mechanical nanoscale response information independent of the sample size. We will receive these extracted features on the macroscale and formulate a thermodynamically consistent material model using continuum mechanics. We will perform a failure zone homogenization technique that is not averaged over the sample volume but only the region of interest and is less prone to size effects. We will validate the macroscale model using experimental nanoindentation results from the literature. This approach will allow us to identify the best material models discovered to show which predictors should be incorporated into the macroscale as a state variable.

• DFG Programme: Research Grants