Representative grain boundaries during anisotropic grain growth

Conventional grain growth models do not account for the variability in grain boundary (GB) mobility, GB energy, and solute segregation energy that depend in detail on the GB structure. Instead, these parameters are typically determined empirically from experimental measurements. In this study, we present a systematic quantitative analysis that accounts for anisotropic GB properties, i.e., GB mobility and solute drag pressure to rationalize the average grain size evolution using representative GB properties under normal grain growth conditions in a non-textured system. We perform two-dimensional phase field simulations to analyze the role of (1) anisotropic mobility, (2) anisotropic segregation, and (3) combined anisotropic mobility and segregation on the average grain size evolution. A corresponding "representative GB"is introduced for each case considering a phenomenological grain growth model that fits the simulated average grain size evolution. A relationship is proposed to determine the representative GB properties for an arbitrary distribution of anisotropic GB properties in the initial microstructure. Simulations with the combined variation of GB mobility and segregation energy suggest that the representative GB mobility and segregation energy determined from independent simulations with varying either GB mobility or segregation energy can be superimposed to determine the average grain size evolution in agreement with the phase field simulations. The limits of validity, i.e., the range of moderate anisotropies, are identified where the phenomenological model with representative GB properties may be applicable, i.e., as long as a mean grain size is an appropriate descriptor of the grain structure.