Phase transitions are ubiquitous in materials science, but their simulation in atomistic models involves singificant technical and conceptual challenges. We have developed a framework to analyze atomistic models and draw the least-biased connection to macroscopic thermodynamic quantities, by quantifying fluctuations in the bulk phases and using them to characterize the state of two-phase snapshots of a simulation.
Atomistic Modelling of Solidification
Nucleation from the melt is an activated event, because of the energetic cost associated with the formation of a solid-liquid interface. We use accelerated sampling techniques to estimate this surface energy term for different materials and thermodynamic conditions.