|About this Abstract
||2018 TMS Annual Meeting & Exhibition
||Fracture: 65 Years after the Weibull Distribution and the Williams Singularity
||A Data-driven Approach to Predict Microstructurally Small Crack Evolution
||Kyle Pierson, Jacob Hochhalter, P. Thomas Fletcher, Ashley Spear
|On-Site Speaker (Planned)
Microstructurally small/short cracks (MSCs) have been shown to be inherently three-dimensional, and their local growth rates can exhibit significant variability caused by sensitivity to local microstructure. Results from both experiments and models have shown that stress fields surrounding MSCs can be heterogeneous, varying by an order of magnitude among grains. For these reasons, the applicability of an analytical stress singularity, on which linear elastic fracture mechanics is based, is questionable. Two main challenges associated with predicting failure of structural components include: 1) lack of an agreed-upon, quantitative, crack-driving mechanism or model to predict MSC growth rates, and 2) lack of a computationally tractable framework that allows for the representation of evolving plasticity and fracture with adequate fidelity to accommodate the crack-driving mechanism. This talk will focus on recent efforts to address these challenges through the integration of high-resolution experimental data, high-fidelity numerical modeling, and machine learning approaches.
||Planned: Supplemental Proceedings volume