Dynamic Recrystallisation

Motivation

The image on the left, adapted from Urai et al. (1986), shows a naturally deformed and recrystallised quartzite with two distinct grain populations: larger grains with irregular boundaries and smaller grains interlocking with them, showing straighter, less irregular boundaries. Distinguishing these populations is vital for quantifying dynamic recrystallisation, which governs microstructural evolution and mechanical behaviour. However, previous methods based on intragranular distortion to separate remnant and recrystallised grains are effective only at low homologous temperatures and fail near the melting point, as in the case of ice.

Our Approach: Grain Boundary Sphericity

In this study, we introduce grain boundary sphericity as a new quantitative parameter to describe grain morphology during dynamic recrystallisation. Grain boundary sphericity measure captures the geometric complexity of grain boundaries; it is defined as the 2-D grain area divided by the product of grain-boundary perimeter and area-equivalent radius, providing a dimensionless indicator of boundary irregularity. This approach provides us with a powerful tool to distinguish recrystallised from remnant grains, assess boundary migration activity, and better link evolving microstructures to the rheological behaviour of ice and other crystalline materials. Click to play the illustrative movie.

Sphericity and Dynamic Recrystallisation

Using microstructural data from up-strain deformation experiments and supplementary datasets, we found that sphericity decreases with increasing grain size up to a threshold, beyond which it either plateaus or increases slightly. The decline reflects small, spherical recrystallised grains becoming more irregular through strain-induced grain-boundary migration (GBM). The plateau at larger sizes indicates remnant grains whose irregularity stabilises as GBM and nucleation reach balance. The threshold grain size therefore represents the maximum size attained by recrystallised grains during deformation.

GBM Rate Irrelevant to Temperature?

By assuming that the smallest nuclei formed at the onset of dynamic recrystallisation and grew to the threshold grain size over the course of each experiment, we estimated the GBM rate. We found that deformed samples exhibit similar values of GBM rate across different temperatures, suggesting that grain growth kinetics are largely temperature-independent under the tested deformation conditions.

GBM Rate = Mobility x Driving Force

Previous studies show that grain-boundary mobility decreases with falling temperature (left) as diffusion and migration slow down. However, the driving force for GBM increases with stress (right), which rises at lower temperatures when the strain rate is constant. These opposing effects likely balance each other, producing similar GBM rates at both high and low temperatures and explaining the limited temperature dependence observed in our deformation experiments.