Motion of grain boundaries incorporating dislocation structure

Luchan Zhang, Yang XiangJournal of the Mechanics and Physics of Solids 2018 · DOI · arXiv

Abstract: This study introduces a continuum model for the motion of low-angle grain boundaries (GBs) that explicitly incorporates underlying dislocation structures. By tracking the evolution of dislocation density potential functions, the model naturally predicts complex behaviors such as grain rotation, tangential sliding, and shape-preserving shrinkage. The framework is validated against discrete dislocation dynamics (DDD) simulations, demonstrating that macroscopic GB kinetics emerge from microscopic defect interactions without the need for phenomenological coupling factors.

Introduction

Most engineering materials are polycrystalline, composed of numerous grains separated by grain boundaries (GBs). The migration of these boundaries is a primary driver of microstructural evolution, influencing mechanical strength, ductility, and thermal stability. Traditionally, GB motion has been modeled as mean curvature flow, where the boundary moves to minimize surface energy. However, for low-angle grain boundaries, which are physically composed of dislocation arrays, this simplification fails to capture essential physics like grain rotation and sliding.

Methodology

The authors develop a continuum dislocation dynamics framework to bridge the gap between discrete microscopic defects and macroscopic boundary motion. Instead of tracking individual dislocations, the model employs dislocation density potential functions. The evolution is governed by two coupled sets of equations: one for the normal velocity of the boundary and another for the transport and reaction of dislocation densities along the interface.

Geometric representation of a grain boundary point P defined by radius R(α) and angle α, showing normal (n) and tangent (T) vectors.

Driving forces are derived from the total system energy, accounting for long-range elastic interactions, local line energy, and external stresses. The model ensures consistency with Frank’s formula, maintaining the geometric relationship between dislocation content and grain misorientation. Validation was performed using material parameters for Aluminum and FCC crystal geometries with specific Burgers vectors.

Key Results and Discussion

A significant finding is that coupling between normal motion and grain rotation emerges naturally from dislocation conservation laws. In the absence of dislocation reactions, the model predicts "shape-preserving shrinkage," where non-circular grains (e.g., ellipses) maintain their aspect ratio while shrinking. Furthermore, for pure coupling, the product of the boundary length ($L$) and misorientation angle ($θ$) remains constant ($Lθ = \text{const}$).

When dislocation reactions are permitted, the model captures the transition from smooth to faceted boundaries. For instance, an initially circular grain may evolve into a hexagonal shape as dislocations segregate into lower-energy planar segments. The continuum results show excellent quantitative agreement with discrete dislocation dynamics (DDD) simulations, providing a computationally efficient alternative for large-scale modeling.

Conclusion and Future Work

While currently limited to 2D tilt boundaries and low-angle regimes, this framework provides a robust physical basis for modeling GB kinetics. The authors demonstrate that long-range elastic interactions act as the "glue" maintaining uniform misorientation during motion. Future extensions to 3D geometries and the inclusion of triple junctions will further enhance its applicability to complex polycrystalline microstructures.

Summary updated

BibTeX
@article{zhang2018motion,
  author = {Luchan Zhang and Yang Xiang},
  title = {{Motion of grain boundaries incorporating dislocation structure}},
  journal = {Journal of the Mechanics and Physics of Solids},
  year = {2018},
  doi = {10.1016/j.jmps.2018.05.001},
}
Figure