Mechanisms for Pore Fluid Stabilization of Fault Propagation and Slip

Sudden motions of fault (i.e., fault propagation and slip) cause earthquakes. Understanding the mechanics of earthquakes requires quantitative knowledge of fault propagation and slip instability, which has long been a focus of experimental rock mechanics. In a classic framework based on the elastic rebound theory, the earthquake cycle includes the interseismic period of strain accumulation and the coseismic period of sudden strain release along a tectonic fault. Geophysical observations reveal diverse behaviorsof fault motions resulted from strain accumulation and release, from aseismic creep to slow slip events (SSEs) to regular earthquakes. Discovery of SSEs during the interseismic period provides a new means to assess the mechanical states of a seismogenic fault between earthquakes. Most seismic studies link SSEs to high pore fluid pressure. Yet, the mechanical link between slow fault slip and high pore fluid pressure is not well understood. We conduct experimental investigation to elucidate the mechanisms responsible for pore fluid stabilization of fault propagation and slip. Our experimental results show that slip events along gouge bearing faults can transform from fast to slow with increasing pore fluid pressures while keeping the effective pressure (i.e., confining pressure minus pore fluid pressure) constant. In these experiments, a layer of fine-grained quartz gouge was placed between the saw-cut surfaces in porous sandstone samples. The saw-cut samples were subject to conventional triaxial loading under a constant effective pressure using various combinations of confining and pore fluid pressures. Different slip events, from dynamic, audible stick-slip to slow, silent slip, with a range of slip rates and stress drops were produced along the gauge-filled saw-cut surface. These results suggest that on the same fault, varying pore fluid pressure alone could result in a range of fault slip behaviors from dynamic to creep. Experimental data further demonstrate that under the same effective pressure, high pore fluid pressure conditions stabilize fault propagation in a wide range of intact rocks including granite, serpentine, and sandstones. In compact rocks (initial porosity <5%) the stabilization effect can be explained by dilatant hardening. When dilatancy occurs faster than fluid diffusion along a propagating fracture, the resultant increase in effective normal stress impedes further fracture growth. In porous sandstones (initial porosity >10%), however, dilatancy hardening alone could not adequately explain the stable post-peak fault growth observed at slow loading rates where drained conditions are achieved. Based on the quantitative microstructural analysis of the deformed samples, we propose that the stable fault growth in highly permeable sandstones manifests stable cracking due to stress corrosion. These results elucidate the important controls of pore fluid on rock strength and fault slip beyond the effective stress law. The results provide a mechanic link between the spatially correlated SSEs and high pore fluid pressure conditions.