Mechanisms of failure in a fluid-saturated fault gouge subject to cyclic pore-pressure oscillations

While dynamic fluctuations in effective normal stress affecting fault zones are ubiquitous, arising from both complex natural phenomena like remote dynamic triggering of earthquakes and anthropogenic activities like industrial subsurface fluid injection, the precise influence of these perturbations and specifically their frequencies, on the macroscopic fault strength remains insufficiently characterized.The frequency of these pore-pressure variations is likely a key factor setting the timescale for the drained-to-undrained transition, thereby driving markedly different mechanical responses.

In this work, we present results from a coupled hydromechanical-discrete element model simulating a pre-stressed, fully saturated granular fault gouge subject to cyclic pore-pressure variations across three orders of magnitude in frequency. We observe that fault failure consistently occurs before the system reaches the traditional Mohr-Coulomb failure criterion. This early failure indicates that additional dynamic mechanisms, often neglected in effective stress analyses, play a dominant role in triggering instability. We investigate the driving forces responsible for this pre-Mohr-Coulomb failure and find they evolve distinctly with frequency. We evaluate three primary candidates driving this behavior: 1) seepage forces arising from the pore-pressure gradients, 2) contact weakening induced by granular agitation (vibration), and 3) inertial effects driven by acceleration from cyclic pore-pressurization. Our analysis isolates the contribution of each mechanism across the frequency spectrum, offering a new physical basis for understanding why dynamic pore-pressure perturbations can trigger slip earlier than what static friction laws predict.