| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Powder Materials Processing and Fundamental Understanding
|
| Presentation Title |
Large Scale DEM Simulations and Computational Performance Evaluation |
| Author(s) |
Vimal Ramanuj, William Carter, Michael Sebok, Andrzej Nycz, Peeyush Nandwana, Jason Mayeur, Yousub Lee |
| On-Site Speaker (Planned) |
Vimal Ramanuj |
| Abstract Scope |
Powder Metallurgy Hot Isostatic Pressing (PM-HIP) via additive manufacturing combines traditional and advanced manufacturing advantages. The PM-HIP process involves filling metal powders into an AM-printed container and increasing part density under high temperature and pressure. This process is sensitive to initial powder packing density and configuration. Measuring and quantifying these in large-scale parts is laborious, time-consuming, and costly. A simulation-based approach using the Discrete Element Method (DEM) offers an efficient alternative. In this research, we focus on DEM simulations of particle filling and the effects of pouring parameters. Open-source DEM simulation codes will be examined, particularly LIGGGHTS and Blaze-DEM on an HPC cluster. A benchmark setup of powder filling in a concentric cylindrical cavity will be used with varying geometries leading up to ~1B particles for rigorous computational performance evaluations. Results will be analyzed for physics capabilities (e.g., contact models) and computational performance (e.g., scalability, GPU performance) of the codes. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Modeling and Simulation, Powder Materials, Computational Materials Science & Engineering |