About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing Modeling, Simulation and Artificial Intelligence
|
| Presentation Title |
Quantitative Validation Methodology for Grain Growth Models Using Probabilistic Metrics |
| Author(s) |
Arulmurugan Senthilnathan, Pranav Karve, Sankaran Mahadevan |
| On-Site Speaker (Planned) |
Arulmurugan Senthilnathan |
| Abstract Scope |
Macro-scale mechanical properties and performance of a material are predicted using multi-scale micromechanical models that require large-scale microstructure data. High experimental cost in observing a large-scale microstructure necessitates the need for computational models. However, the model form error in describing the grain growth phenomenon and numerical errors propagate to the characteristic features of the predicted microstructure. This work presents a quantitative validation method for grain growth models using probabilistic metrics. Physics-based models (e.g., phase-field) are used to predict the microstructure. First, characteristic features of the microstructure (e.g., grain shape) are quantified using selected descriptors (e.g., sphericity). Probabilistic metrics are then defined to compare the descriptors of predicted and experimental microstructural features. Quantitative model validation methods using probabilistic metrics pave the way for systematic uncertainty aggregation in the property and performance prediction, and support model-assisted qualification and certification of additively manufactured products. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, Characterization, Modeling and Simulation |