Abstract Scope |
This talk addresses the integration of uncertainty quantification and propagation (UQ/UP) within Integrated Computational Materials Engineering (ICME), focusing on CALPHAD modeling for alloy design. Historically, ICME and CALPHAD have utilized deterministic models, neglecting inherent uncertainties in data and model formulations. We review the current state of the field, introduce basic UQ/UP concepts, and discuss their interpretation through frequentist and Bayesian perspectives. Case studies demonstrate UQ application in model parameterization, and we present methods for uncertainty propagation in predicting thermodynamic properties and phase diagrams. The use of Bayesian frameworks for model selection and fusion is also explored. We conclude by highlighting ongoing challenges and the necessity of managing uncertainty to advance ICME's predictive accuracy and reliability.
*This abstract is being submitted to the Uncertainty quantification in ultra-high temperature materials manufacturing session. |