In this work we present a non-parametric Bayesian approach for developing structure-property models for grain boundaries (GBs) with built-in uncertainty quantification (UQ). Using this method we infer a structure-property model for H diffusivity in [100] tilt GBs in Ni at 700 K based on molecular dynamics (MD) data. Once a GB structure-property model is developed, it can be used as an input to mesoscale simulations of the effective properties of polycrystals, microstructure evolution, etc. A significant advantage of the Bayesian approach presented here is that it facilitates propagation of uncertainties from the underlying structure-property model to the output predictions from mesoscale modeling. Leveraging this capability, we perform mesoscale simulations of the effective diffusivity of polycrystals to investigate the interaction between structure-property model uncertainties and GB network structure. We observe a fundamental interaction between crystallographic correlations and spatial correlations in GB networks that causes certain types of microstructures (those with large populations of J2- and J3-type triple junctions) to exhibit intrinsically larger uncertainty in their effective properties. Data and code are provided in supplementary materials.
Read full abstract