|About this Abstract
||2016 TMS Annual Meeting & Exhibition
||Computational Methods for Uncertainty Quantification, Model Validation, and Stochastic Predictions
||Exploring the Effects of Micro-texture on Engineering-scale Performance
||John Emery, Richard Field, Jay Carroll, Joseph Bishop
|On-Site Speaker (Planned)
Prediction of structural failure due to strain localization is an essential engineering challenge where ultra-high reliability is required. Heterogeneity at the fine scale can contribute to significant uncertainties in performance, particularly for small-scale components. Practical applications cannot include fine-scale details throughout the problem domain due to exorbitant computational demand. One concurrent multiscale calculation is necessary but not sufficient. There is no way to know if it represents the best or worst case or some average of scenarios.
We present an approach to efficiently propagate uncertainty across length scales employing stochastic reduced-order models. The approach begins at the engineering scale and produces a low-fidelity prediction that is subsequently refined with multiscale simulation. The results presented in the talk will focus on meso-scale simulations. We report on efforts to develop a statistically accurate model of crystallographic texture, including spatial variability, and probe the meso-structural response and its effects on structural performance.
||Planned: A print-only volume