About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Frontiers in Solidification X
|
| Presentation Title |
Improving Benchmark Validation with Uncertainty Quantification |
| Author(s) |
Levi Dickson, Caleb Schrad, Miles Johnson, Jack Heeney, Red Deaton, Luke N. Brewer, Matthew J. M. Krane |
| On-Site Speaker (Planned) |
Levi Dickson |
| Abstract Scope |
Benchmark experiments featuring unidirectional solidification of metals in prismatic geometries are commonly used to validate solidification models. Epistemic and aleatoric uncertainty regarding inputs such as material properties and evolving physical features such as shrinkage induced thermal resistance can pose a significant challenge in comparing experimental and simulation results. Overcoming this knowledge gap by obtaining more precise measurements or more sophisticated models is not always practical, however, a range of input conditions based on measured uncertainty can offer insight into a system’s sensitivity to input conditions. This work uses the experimental setup described above on Inconel 718 castings which are compared to results from a finite volume, multicomponent, solidification model. A sensitivity analysis is performed for all input parameters, examining sources of uncertainty resulting in output variability. The results of this sensitivity analysis are discussed in the context of improving experimental design and computational fidelity. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Solidification, Modeling and Simulation, Computational Materials Science & Engineering |