About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
Presentation Title |
Accelerating Crystal Plasticity Fatigue Simulations of Additively Manufactured Metals Using the “Materialize” Framework |
Author(s) |
George R. Weber, Joshua Pribe, Brodan Richter, Saikumar Yeratapally, Patrick Leser, Andrew Kitahara, Somnath Ghosh, Edward Glaessgen |
On-Site Speaker (Planned) |
George R. Weber |
Abstract Scope |
Fatigue failure of additively manufactured metals occurs through the cyclic evolution of microscale mechanisms. Methods for micromechanical fatigue prediction through high-fidelity, grain-scale modeling are difficult to develop and often computationally intractable. Two factors inhibiting the development of micromechanical fatigue models are: (1) streamlined integration of physics-based and empirical microstructural data and (2) solving mechanistic constitutive models over long-time scales. In this work, the first factor is addressed through the development of a new Python library, Materialize, built to efficiently link computational materials models across the process-structure-property-performance (PSP) pipeline. The second factor is addressed by developing a multi-time scaling methodology for crystal plasticity models within Materialize. Automatic differentiation is leveraged to relieve meticulous derivations required to solve constitutive models and to develop a material-agnostic, multi-time scaling methodology for accelerating fatigue simulations by orders of magnitude. The outcome is a PSP framework for the exploration of cycle-dependent mechanisms in additive manufacturing applications. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, ICME, Computational Materials Science & Engineering |