Recent advances in computational and experimental techniques now permit the evolution of a microstructure to be determined in three dimensions and as a function of time. It is thus possible to employ an experimentally measured microstructure as an initial condition in a simulation and to then compare the predicted structure to that measured experimentally at some later time. Such an approach is a particularly stringent test of simulation and can be used to identify important phenomena that are lost in an averaging process. We shall illustrate this approach using experiments and simulations of solid-state grain growth. Using an experimentally measured grain structure as an initial condition in a phase field model, we compare the shapes of individual grains measured experimentally to those predicted by the simulation at some later time. We find that the phase field simulations reproduce quite accurately the grain morphology and topology in regions of the sample with isotropic grain boundary properties. However, in other regions, we find a clear influence of grain boundary energy anisotropy on the morphological evolution of grains and a disagreement between simulation and experiment.
Copyright. All Rights Reserved.