|About this Abstract
||2019 TMS Annual Meeting & Exhibition
||Computational Approaches for Big Data, Artificial Intelligence and Uncertainty Quantification in Computational Materials Science
||Error Estimation for Stress Distributions and Macroscale Yield Prediction in Polycrystalline Alloys
||Kamalika Chatterjee, Robert Carson, Paul Dawson
|On-Site Speaker (Planned)
The errors in the stress distribution are estimated in virtual polycrystalline samples of alpha-Titanium (hcp). To estimate the error, the discontinuous stress field (element-by-element stresses) over a grain is smoothed by projecting the stress components to the nodes and enforcing inter-element continuity. The differences between the continuous and discontinuous stress fields are utilized to estimate the errors for corresponding elements and grains. Error estimations are performed for different cases of microstructures and sample instantiations. Magnitude of errors correlate with the magnitude of heterogeneity of deformation. Serrated grain boundaries are found to have little effect on error values. Simulations are performed with smooth and serrated grain boundary samples, instantiated with experimental data. From the simulation data, macroscopic yield strength values are extracted from volume elements ranging in size from a fraction of the gage section to its entire length and the property trends are compared across the different instantiation approaches.
||Planned: Supplemental Proceedings volume