Building models for the plasticity, thermodynamics and kinetics of metals is challenging as subtle aspects of atomic cohesion must be faithfully reproduced, and predictions often require averaging over large, complex configuration ensembles. I will discuss how the energy landscapes of atomic systems can be rapidly explored at scale and “coarse-grained” when the dynamics are thermally activated thus thus scale separated[1,2] and how data-driven techniques, typically used to regress energies for modern cohesive models, can be used to capture a much wider range of properties such as defect entropics or dislocation properties. When the dynamics do not have a clear timescale separation, coarse graining is much more challenging. I will discuss how a data-driven approach can provide a solution, producing efficient surrogate models which can predict the evolution of nanoparticle ensembles and the yielding of complex microstructures, offering new perspectives for multiscale modelling approaches.
 TD Swinburne and D Perez, NPJ Comp. Mat 2020, MSMSE 2022
 TD Swinburne and DJ Wales JCTC 2020, 2022
 C Lapointe et al. PRMat 2020
 TD Swinburne, In Prep.