Laboratory observations of fluid pressure and fault dilatancy controlling earthquake instability
Abstract
Pore pressure changes affect fault stability in natural and induced earthquakes. Proposed effects include thermal pressurization, dilatancy hardening, induced seismicity and slow slip. Yet pore pressure changes are not measured directly in rupture nucleation zones. In laboratory tests with external pore pressure control, restricted hydraulic communication with the fault may lead to undetected differences between fault zone pore pressure and the external control system during accelerating creep or dynamic slip. We report on triaxial deformation of sawcut faults in Westerly granite at normal stresses to 197 MPa on both bare surface and faults containing 1 mm thick quartz gouge. Samples were 76.2 mm-diameter cylinders with a fault inclined 30° to the sample axis. In most tests, faults were hydraulically isolated from the external control system, and pore water in the fault communicated with the external control system through the low permeability granite with a diffusion time constant > 1 hour. Thus, the fault was undrained coseismically, undrained or incompletely drained over the duration of precursors and, in many cases, over recurrence intervals. Internal pore pressure (Pp) was measured with a small pressure sensor embedded in the sample close to the fault.
We report on stick-slip and slow slip deformation events and Pp changes accompanying standard velocity stepping tests. Observations of Pp changes are attributed to velocity-dependent gouge dilatancy and are consistent with dilatancy hardening analyzed by Segall and Rice (1995). Dilatancy-driven changes in effective pressure cause transient shear strength changes that exceed intrinsic rate- and state-dependent strength changes. In some tests involving large jumps in the shear loading rate, subsequent transient weakening and stick-slip coincide with a rapid rise in Pp suggesting a collapse of gouge pore structure. In these tests, the sudden rise in Pp assisted or caused slip instability. Gouge composition and structure, effective normal stress and loading rate can all contribute to the conditions under which compaction or dilation occur, and for isolated faults can control fault stability. These experiments represent a promising first attempt at observing pore fluid/fault interactions with a passive, in situ pressure sensor.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMMR23E0163K
- Keywords:
-
- 3902 Creep and deformation;
- MINERAL PHYSICS;
- 8010 Fractures and faults;
- STRUCTURAL GEOLOGY;
- 8034 Rheology and friction of fault zones;
- STRUCTURAL GEOLOGY;
- 8045 Role of fluids;
- STRUCTURAL GEOLOGY