|About this Abstract
||2019 TMS Annual Meeting & Exhibition
||Computational Approaches for Big Data, Artificial Intelligence and Uncertainty Quantification in Computational Materials Science
||Uncertainty Quantification in Solidification Modeling of Additive Manufacturing
||Supriyo Ghosh, E. Chin, J. Knap, D. Allaire, R. Arroyave
|On-Site Speaker (Planned)
Microstructures that evolve during an additive manufacturing (AM) solidification process possess various sources of uncertainties, leading to variability in material properties of the final product. Reduction of variability in these microstructures or their descriptors, such as spacing and spatial distribution of species, is therefore essential. Macroscale finite element analysis simulates an AM process in order to determine the solidification conditions in the resulting alloy molten pools. High-throughput mesoscale phase-field simulations are then used to predict the planar, cellular, and eutectic microstructures for the above solidification conditions. Global sensitivity analysis of the input process and model parameters, frequency distribution, correlation analysis, and surrogate modeling are used for uncertainty management. Our effort in linking uncertainty quantification with AM will eventually enable to achieve control over the main parameters that will lead to predictive and proactive microstructure evolution.
||Planned: Supplemental Proceedings volume