Shear‐induced dilatancy of fluid‐saturated faults: Experiment and theory

Abstract

Pore fluid pressure plays an important role in the frictional strength and stability of tectonic faults. We report on laboratory measurements of porosity changes associated with transient increases in shear velocity during frictional sliding within simulated fine‐grained quartz fault gouge (d50 = 127 μm). Experiments were conducted in a novel true triaxial pressure vessel using the double‐direct shear geometry. Shearing velocity step tests were used to measure a dilatancy coefficient (ɛ = Δϕ/Δln(v), where ϕ is porosity and v is shear velocity) under a range of conditions: background shearing rate of 1 μm/s with steps to 3, 10, 30, and 100 μm/s at effective normal stresses from 0.8 to 20 MPa. We find that the dilatancy coefficient ranges from 4.7 × 10−5 to 3.0 × 10−4 and that it does not vary with effective normal stress. We use our measurements to model transient pore fluid depressurization in response to dilation resulting from step changes in shearing velocity. Dilatant hardening requires undrained response with the transition from drained to undrained loading indexed by the ratio of the rate of porosity change to the rate of drained fluid loss. Undrained loading is favored for high slip rates on low‐permeability thick faults with low critical slip distances. Although experimental conditions indicate negligible depressurization due to relatively high system permeability, model results indicate that under feasible, but end‐member conditions, shear‐induced dilation of fault zones could reduce pore pressures or, correspondingly, increase effective normal stresses, by several tens of megapascals. Our results show that transient increases in shearing rate cause fault zone dilation. Such dilation would tend to arrest nucleation of unstable slip. Pore fluid depressurization would exacerbate this effect and could be a significant factor in generation of slow earthquakes, nonvolcanic tremors, and related phenomena.