|About this Abstract
||2017 TMS Annual Meeting & Exhibition
||Computational Methods and Experimental Approaches for Uncertainty Quantification and Propagation, Model Validation, and Stochastic Predictions
||Quantifying Uncertainty from (Pseudo)potentials for First Principles and Classical Atomistic Simulations
||Mark Tschopp, Efrain Hernandez, Shawn Coleman, Decarlos Taylor, Jennifer Synowczynski-Dunn
|On-Site Speaker (Planned)
How do we quantify uncertainty in first principles or atomistic simulations? In this work, we used different approaches to quantify the influence of the (pseudo)potential (and other parameters) on the structure of boron-based ceramics. One example focuses on coupling different similarity/distance measures with virtual x-ray diffraction profiles to quantify the structural uncertainty due to different (pseudo)potentials for first principles and atomistic simulations. The diffraction profiles highlight subtle structural changes that occur when modeling complex materials using different first-principles techniques, basis-sets, pseudopotentials, and classical interatomic potentials. The ability to quantify the difference between these profiles involved the use of approximately 50 distance and similarity metrics. Other examples focus on uncertainty in the parameterization of classical potentials. The significance of this work is that this methodology can highlight the uncertainty associated with various computational approaches and help to validate structure and quantify the fidelity of various classical potentials in atomistic simulations.