Microstructure Evolution Based on Grain Boundary Migration Mechanism

Abstract

Grain growth (i.e., the increase in the mean grain size with time during the annealing of a polycrystal) is the most prevalent kind of microstructure evolution. The classical theory of grain growth focuses on the reduction of the excess energy of a grain boundary (GB) network by GB migration. In this framework, GBs migrate towards their centers of curvature (𝜅); the GB velocity may be expressed as 𝑣 = 𝑀𝛾𝜅, where 𝑀 is the GB mobility and 𝛾 is the (isotropic) GB energy. However, this theory fails to explain many widely observed phenomena, e.g., stress-driven grain growth, grain rotation, GB sliding, shear-coupled GB migration, low correlation between GB velocity and GB mean curvature measured in experiments, etc. Although there have been many proposals for the cause of this failure, it remains a subject of ongoing debate.

The fundamental microscopic mechanism of GB migration is now widely accepted as the nucleation and propagation of disconnections, i.e., line defects constrained on GBs possessing both step and dislocation characters. In this thesis, we first propose a bicrystallography-respecting continuum model for arbitrarily curved two-dimensional GBs, accounting for the role of disconnections. We demonstrate that the driving force for GB migration is not merely capillarity force (or curvature effect), but also includes elastic interactions among disconnection Burgers vectors (i.e. self-elastic interactions) and elastic interactions between disconnections and external stress. We conducted numerical simulations to demonstrate that the self-elastic interactions play an important role in the GB faceting-defaceting behavior and grain rotation in various bicrystal configurations. We further implement the proposed continuum model in a diffuse-interface approach to study the role of disconnections on the large-scale grain growth procedure in polycrystals. Qualitative and quantitative consistency between our numerical simulation results and the existing experimental and atomistic simulation results demonstrates that disconnection flow along GBs and the consequently generated internal stress are the main cause of the failure of the classical, mean curvature flow description of grain growth.

Besides the insufficient driving forces considered in the classical model, we also conduct a series of atomistic simulations to show that the mobility (𝑀) of most GBs is dependent on their migration direction, instead of a constant that depends only on temperature. We demonstrate that the directional GB mobility exists only when the crystal symmetry relationship is broken between the two grains adjacent to the GB, whilst it is irrelevant to whether the GB is flat or curved or the origin of the driving force for migration. We demonstrate that GBs are natural Brownian ratchets, i.e., devices that perform unidirectional motion under non-equilibrium fluctuations. We explore the implications of GB-Brownian ratchet behavior for polycrystalline microstructure evolution through thermo-mechanical processing. Finally, we present a revised equation of motion for GB migration that incorporates all our current findings.

Date of Award	3 Jun 2025
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Jian HAN (Supervisor), David Joseph SROLOVITZ (Supervisor) & Zhaoxuan WU (Supervisor)