A major challenge in multiscale materials simulation is the ab initio prediction of phase stabilities in multi-phase materials. Since it involves complex simulation protocols, the uncertainty of the ab initio input and the error propagation to the desired free energies, transition temperatures and entropy changes is a critical issue. At this level, a combination of model uncertainties, numerical, convergence and statistical errors is present. Already the determination of the equilibrium lattice constant and bulk modulus requires a careful analysis of the fitting of energy-volume curves, going beyond the consideration of standard convergence parameters like cutoff and k-points.
In order to handle this delicate interplay of uncertainties, we introduce the concept of uncertainty phase diagrams. Based on the uncertainty phase diagrams we model the convergence gradients of the contributing errors, to automate the convergence process not only for the error in energy. The modelling of uncertainties in relation to the corresponding ab initio calculation is enabled by our recently developed Python based workbench pyiron. In particular the generic interfaces to simulation codes at different time and length scales and the in-process data management model are used to reduce the technological complexity of our uncertainty propagation model. Our investigations revealed that commonly used rules of the thumb for fitting ground state materials properties become invalid for high precision calculations, as the dominating sources of error change.
Back to Complex High-Dimensional Energy Landscapes