| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Verification, Calibration, and Validation Approaches in Modeling the Mechanical Performance of Metallic Materials
|
| Presentation Title |
Uncertainty-Aware Validation in Modeling of Metal Plasticity: Beyond Mean Squared Error |
| Author(s) |
Aaron E. Tallman |
| On-Site Speaker (Planned) |
Aaron E. Tallman |
| Abstract Scope |
Reliable, problem-validated models are essential to informed decision making in safety-critical situations, to design of performance-critical components, and to leveraging limited experimental data. Typically, model validation involves a statistical comparison of model predictions with experimental data to find general agreement, with a small-but-reasonable tolerance for error and bias. The extent of the mismatch is often used to temper expectations of the precision of predictions in practical cases. Unfortunately, an agreement with data alone is not a guarantee of a valid pairing of model and problem. In this work, crystal plasticity finite element simulations of micro-indentation testing are used to generate synthetic testing data, fit by a simple homogenized model. Inverse uncertainty quantification-based approaches are used to provide additional rigorous tests, including analysis of the identifiability of parameters and sensitivity to noise and bias in data. The potential impact of a hybrid experimental-computational uncertainty quantification will be discussed. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Modeling and Simulation, ICME, Mechanical Properties |