FAULT WEAR EXPERIMENTS, PART I: TRANSIENT AND STEADY-STATE STAGES UNDER CONSTANT SLIP-VELOCITY

Fault wear, which is a natural result of frictional slip, is a complex process that depends on rock properties (e.g., roughness, brittleness, hardness, strength, composition), and loading conditions (e.g., normal stress, slip velocity, slip history, temperature). Fault wear is indicated in the field primarily by the occurrence of gouge zones and slickenside striations, and observations of natural faults show that gouge-thickness is proportional to fault-displacement over 6-8 orders of magnitude (Scholz, 1987). Experimental studies of fault-wear, which were mostly conducted under low-velocity and for short slip distances, indicate strong dependency of the wear-rate on normal stress.

We conducted an experimental study of fault-wear on a rotary shear apparatus with continuous monitoring of wear, friction, and temperature at slip velocities of 0.001-1 m/s and large displacements. Samples of Sierra white granite, approximately 7 cm-diameter, were sheared against each other at normal stress ranging between 0.48 and 6.9 MPa and at slip-velocities up to 0.5 m/s. The wear-rate was calculated from the continuously measured closure across the fault blocks, and presented here by the unit W= [(micron of surface wear) / (meter of slip distance)]. Most experiments display two distinct wear-rate stages: (1) an initial, transient “running-in” stage with high wear-rate of W = 80-250 μm/m; and (2) a steady-state stage with lower wear-rate of W =0.5-20 μm/m. The running-in stage exists for 0.5-1.5 m of slip distance. The experimental results suggest that the steady-state wear-rates are controlled by (1) the normal stress, (2) the velocity-dependent friction, and (3) the loading power (=slip-velocity * shear-stress). We will discuss these controlling relations and their implications to gouge zones in the field. Part II of this study includes wear results of experiments with highly variable slip-velocity that simulate the assumed slip-velocity variations during earthquakes (Chang et al., this meeting).