The Role of Grain Size on Shear Localization Illuminated by Acoustic Emissions

Shear localization within the fault core, as evidenced by grain comminution in fault gouge, plays a crucial role in the initiation of frictional instabilities. To upscale the physics of shear localization and understand the influence of grain size, it is essential to identify the governing physical parameters and micro-mechanisms. In this study, we conducted double-direct shear experiments on quartz fault gouges with varying initial grain sizes (coarse, fine, and bi-disperse mixtures) under constant normal stress and shearing velocity, while continuously monitoring Acoustic Emissions (AE). Microstructural analyses were performed on the deformed samples to complement the mechanical and AE data. Our results reveal that, while the initial grain size and distribution do not substantially alter the steady-state friction coefficient, they significantly influence the early stages of frictional evolution leading to steady state, which is reflected in a different rate and amplitude of AEs. Specifically, coarse grains and bi-disperse mixtures exhibit strain-hardening behavior before reaching steady-state friction, whereas fine grains show strain-weakening behavior. Microstructural observations further indicate that bi-disperse mixtures retard the localization of deformation with increasing fault displacement. The AE data shows a strong dependence on both the average grain size and the evolving state of the gouge layer. Notably, there is a direct correlation between the average AE amplitude and the average grain size within the gouge. These findings suggest that variations in AE attributes are indicative of distinct micro-mechanisms active during different stages of shear localization, which are not fully captured by mechanical data and microstructural analysis alone.

The data uploaded are structured as follows:

ReadMe.pdf explains the different parts of the data repository

'Mechanical_and_acoustic_data' gathers all the mechanical and acoustic catalogs used in the paper

Python files are used to plot the figures of the paper and supplementary information