| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Accelerated Computational Insertion of Structural Materials |
| Author(s) |
Anupam Neogi, Deepankar Pal, Jimmy He, Ali Najafi, Grama Bhashyam |
| On-Site Speaker (Planned) |
Deepankar Pal |
| Abstract Scope |
This research study presents an innovative approach leveraging Multiscale Orientation Homogenization (MOH) and Machine Learning enabled Stiffness (MLeS) generation to reduce computational overhead in Crystal Plasticity simulations.
In context of MOH, a new kernel enabling smoothened 2-dimensional orientation distribution statistical input for 3-dimensional particle collision statistical grain generation simulations will be demonstrated. The resulting distributions lead to larger grain and mesh sizes for efficiency and with least degradation in stress and equivalent plastic strain distributions. Additionally, introductory pointers on upscaling this methodology for multiscale orientation and misorientation histogram distributions will be shared for multiscale MOH recursion.
In context of MLeS, a crystal elastic stiffness matrix will be demonstrated for a given shape, material parameters and orientation distribution of an element. This matrix will be compared against its traditional matrix counterpart for accuracy and efficiency. Additionally, introductory examples on stiffness on-the-fly learning in the nonlinear regime will be shared. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
ICME, Machine Learning, Modeling and Simulation |