Figure 1: Experimental setup. (a) Schematic cross section of the triaxial loading machine. (b) Schematic diagram of the displacement of the load point in Figure 1a as a function of time. The displacement is a sum of a sinusoid (tidal load) and a line (tectonic load).
This work is in collaboration with Nick Beeler and Dave Lockner of the Rock Physics group at the U.S. Geological Survey. Dr. Bartlow built on earlier experiments by Lockner and Beeler, JGR, 1999, which studied the effect of sinusoidal stresses (representing tidal stressing) applied to a laboratory sample of granite using a triaxial loading machine (Fig. 1). A granite sample with a sawcut fault surface is loaded from below by an advancing piston. The rate of piston advancement is chosen to be the sum of a constant and sinusoidally varying velocity (Fig. 1b), to simulate tectonic and tidal loading of fault zones. The sample is pressurized to 50 MPa, and the sawcut fault experiences frictional locking and stick-slip events, similar to small earthquakes (Fig. 2).
The previous experiments found that the amplitude of tidal loading required to affect the timing of stick-slip events as a function of frequency is constant for high frequencies (nucleation dominated regime), and increases with decreasing frequency for periods greater than the nucleation time of stick-slip events (Coulomb threshold failure regime). Dr. Bartlow's new experiments include the addition of pore fluid with varying fluid pressure and effective stress. The results show that pore fluid draining effects become important at high frequencies, and that the transition between the Coulomb threshold and nucleation dominated failure regimes scales linearly with effective stress, as predicted by Beeler and Lockner, JGR, 2003.
Dr. Bartlow's results show that the addition of pressurized pore fluid can cause stick-slip events to to become correlated with tidal loading, without changing the load. This can explain why tectonic tremor, which occurs in areas thought to have highly pressurized fluid, is tidally modulated by regular earthquakes are not.
Figure 2: Coefficient of friction (mu) versus piston (ram) displacement for 3 experiments with varying pore fluid pressure (Pp). The sawtooth pattern represents repeated stick-slip events.