About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Verification, Calibration, and Validation Approaches in Modeling the Mechanical Performance of Metallic Materials
|
| Presentation Title |
Multiscale Crystal Plasticity Parameter Calibration: Non-Uniqueness in Bayesian and Multi-Objective Approaches |
| Author(s) |
Harry Smith, William Musinski |
| On-Site Speaker (Planned) |
Harry Smith |
| Abstract Scope |
This study compares crystal plasticity parameter calibration performance across multiple techniques using data from macro and mesoscale data for calibration. Bayesian analysis and multi-objective optimization tools such as sbi, emcee, and pymoo were used to quantify uncertainty and non-uniqueness in calibrated parameters. The parameter calibrations were performed using data from the AFRL Challenge Series #4, which features high energy diffraction microscopy (HEDM) data recorded during an in-situ tensile test of an IN625 specimen. The HEDM data allowed for grain-scale mechanical response validation within the parameter calibration process. Parameter sets were calibrated with macroscale data alone, and with the addition of grain-wise elastic strain data. Notably, different sets of calibrated parameters yielded comparable macro and mesoscale responses but varied at the intragranular scale. Understanding how the calibration process influences the localization of stress on the intragranular scale is critical in understanding when and where materials fail due to plasticity and fatigue. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, Modeling and Simulation, |