11:00 AM - 1:00 PM
[SCG44-P01] Experimental investigation of strength recovery under fluid flow
Keywords:strength recovery, frictional experiment
Various seismological and geological observations have focused on the role of fluids in the generation of earthquake, including slow slip. Recent scientific drilling project found the existence of high-pressure aquifer in slow slip regions. The fluid in the fault may not only reduce the fault strength by producing high pore pressure, but may also control the process of fault strength recovery through physico-chemical interaction with the fault material. Here, we conduct laboratory friction experiments to verify the hypothesis that the flow of water suppresses the increase of fault strength.
When a fault slip occurs, the strength of the fault decreases rapidly. It recovers during inter seismic period after the earthquake. The mechanism of the strength recovery the fault during the “hold” period is that the microscopic contact area (asperity) on the fault surface adheres due to fluid-rock interaction such as pressure solution and increase of the real contact area. However, it is not clear whether this strength recovery actually occurs in a fluid flow environment. In particular, it is expected that the pressure of the flowing fluid will change at the same time as the flow path is altered by the healing process, and it will be important to consider how this affects the seismic cycle. We performed slide-hold-slide test under condition of water flow (flow SHS test) using a fluid pressure controlled rotary shear apparatus installed in JAMSTEC Kochi.
Ceramic ball (10 - 300 μm) and crushed Indian sandstone (125-250 μm) were used as simulated fault gouge, and the flow SHS tests were performed under the condition of flowing water at a flow rate of 0.0cc/min and 0.6 cc/min to compare the strength recovery rate. The experiment was conducted under the conditions of a slip velocity of 5 μm/s, a normal stress of 3 MPa, a water pressure of 0.3 - 1.5 MPa, and a hold time of 2 seconds to 48 hours.
In case of without the flow, the frictional strength recovered in proportion to the logarithm of time, as in previous studies (e.g. Dietrich, 1972). On the other hand, the frictional strength did not almost recover in the presence of flow. The water pressure tended to gradually increase with the hold time. When the strength recovery is expressed by the effective friction coefficient, the increase rate of the friction coefficient was almost the same as that when there was no flow. This result was obtained for both ceramic and natural sand, confirming that this inhibition of strength recovery occurs independently of the interaction between the fluid and the gouge. Furthermore, there was an increase in fluid pressure over time, which suggests that the increase in contact area may have blocked the flow path, thereby causing an increase in fluid pressure and consequently no apparent strength recovery. The presentation will include a preliminary discussion of the torque control experiments.
When a fault slip occurs, the strength of the fault decreases rapidly. It recovers during inter seismic period after the earthquake. The mechanism of the strength recovery the fault during the “hold” period is that the microscopic contact area (asperity) on the fault surface adheres due to fluid-rock interaction such as pressure solution and increase of the real contact area. However, it is not clear whether this strength recovery actually occurs in a fluid flow environment. In particular, it is expected that the pressure of the flowing fluid will change at the same time as the flow path is altered by the healing process, and it will be important to consider how this affects the seismic cycle. We performed slide-hold-slide test under condition of water flow (flow SHS test) using a fluid pressure controlled rotary shear apparatus installed in JAMSTEC Kochi.
Ceramic ball (10 - 300 μm) and crushed Indian sandstone (125-250 μm) were used as simulated fault gouge, and the flow SHS tests were performed under the condition of flowing water at a flow rate of 0.0cc/min and 0.6 cc/min to compare the strength recovery rate. The experiment was conducted under the conditions of a slip velocity of 5 μm/s, a normal stress of 3 MPa, a water pressure of 0.3 - 1.5 MPa, and a hold time of 2 seconds to 48 hours.
In case of without the flow, the frictional strength recovered in proportion to the logarithm of time, as in previous studies (e.g. Dietrich, 1972). On the other hand, the frictional strength did not almost recover in the presence of flow. The water pressure tended to gradually increase with the hold time. When the strength recovery is expressed by the effective friction coefficient, the increase rate of the friction coefficient was almost the same as that when there was no flow. This result was obtained for both ceramic and natural sand, confirming that this inhibition of strength recovery occurs independently of the interaction between the fluid and the gouge. Furthermore, there was an increase in fluid pressure over time, which suggests that the increase in contact area may have blocked the flow path, thereby causing an increase in fluid pressure and consequently no apparent strength recovery. The presentation will include a preliminary discussion of the torque control experiments.