Thermal Pressurization Weakening in Laboratory Experiments

Abstract

AbstractThermal pressurization (TP) is expected to be a dominant frictional weakening mechanism during earthquakes. However, most of our understanding of TP relies on theoretical studies. Our experimental setup allows us to test the mechanical response of experimental faults under elevated confining and pore pressures in the laboratory using a rotary‐shear apparatus. We observe dynamic weakening by TP of water‐saturated faults in Frederick diabase sliding at subseismic slip rates of 2.5–5 mm/s under an effective normal stress of 25–50 MPa and effective confining pressure of 20–49 MPa. No other weakening mechanism is activated at these conditions in dry samples. The permeability of the samples is varied by heat treatment before each experiment, and the frictional behavior is tested by velocity‐stepping experiments. Dynamic weakening was investigated after both low and high total fault displacements. For low‐displacement samples: (1) the magnitude and rate of weakening increase as the sample permeability decreases and (2) frictional heating is limited during sliding. These observations are consistent with the expected behavior of TP from theory. Weakening of high‐displacement samples is similar in magnitude to low‐displacement faults, but with slower rates of weakening and less dilation. The difference in weakening of the high‐displacement samples may reflect changes in pore compressibility in the fault zone, which would have a significant effect on the frictional response during TP. The activation of TP at modest slip rates in these experiments supports the hypothesis that TP is an active, and likely dominant, dynamic weakening mechanism during earthquakes in the early stages of slip.