DEM simulations identify 30-200 Hz non-failure window in fluid-saturated fault gouge under sub-critical stress oscillations
Preprint DEM-fluid simulations demonstrate that effective normal-stress oscillations trigger gouge failure outside a 30-200 Hz window via two distinct dilation routes. The non-monotonic frequency response is independent of whether oscillations are applied through pore pressure or total normal stress. Results highlight frequency as a control on dynamic triggering but remain limited by 2-D idealization pending laboratory confirmation.
The study used a 2-D DEM model coupled to fluid flow to shear fluid-saturated or dry granular fault gouge while imposing sinusoidal effective-normal-stress oscillations below the static Mohr-Coulomb failure threshold. Frequency sweeps across four orders of magnitude revealed four mechanical regimes: cyclic failure-arrest at low frequencies, stable sliding inside the 30-200 Hz band, renewed failure at higher frequencies via acoustic fluidization, and inertial sliding above several kilohertz. Both pore-pressure and direct normal-stress cycling produced identical regime boundaries, confirming Terzaghi equivalence once fluid diffusion time is accounted for.
Low-frequency failure occurs because each cycle permits cumulative shear-driven ratcheting dilation that progressively lowers frictional strength. High-frequency failure instead arises from amplified seepage forces and inertial unloading that generate dynamic dilation without requiring net volume change per cycle. The intermediate window represents the temporal mismatch between these two dilation pathways. The work therefore supplies a grain-scale explanation for why some injection-induced or dynamically triggered sequences exhibit frequency selectivity.
Because the results rest on a single 2-D geometry and idealized particle shapes, extrapolation to field-scale faults requires 3-D validation and direct comparison with laboratory shear experiments that can impose controlled oscillations. Planned ring-shear tests at 10-500 Hz on quartz gouge will test whether the predicted non-failure band survives realistic 3-D fabric and fluid heterogeneity.
If confirmed, the frequency window supplies a mechanistic basis for optimizing cyclic injection rates to avoid triggering and for interpreting why some aftershock sequences show pronounced frequency dependence in triggering efficiency.
Sarma et al.: Ring-shear experiments at 50 Hz on dry quartz gouge will show no failure below the static friction envelope when oscillation amplitude stays under 15 % of mean normal stress, tested within 24 months.
Sources (2)
- [1]Primary Source(https://arxiv.org/abs/2607.01448)
- [2]Supporting Source(https://doi.org/10.1029/2021JB022023)