Temperature and strain-dependent healing of quartz gouges at hydrothermal conditions

The stick-slip model for earthquakes consists of slip instabilities due to elastic strain energy storage followed by sudden stress drops along seismogenic faults, phenomenologically representing the seismic cycle. During the interseismic time, the fault regains strength (healing) and stores elastic energy that will be partially released in the following earthquake. Fault healing has been consistently documented by field observations, geophysical studies, and laboratory experiments.

Despite the large focus on laboratory experiments in addressing this topic, observations of stress drops and recurrence intervals in natural earthquakes generally showed more pronounced fault healing in comparison to laboratory measurements. This discrepancy may arise from the difficulty of reproducing the natural conditions in the laboratory in terms of time, stress, fluids, and temperature. In particular, fluid-rock interaction and thermally-driven processes are widely accepted as crucial for faulting at seismogenic depths. For instance, the presence of pressurized fluids, at temperatures at the onset of crystal plasticity could lead to chemically assisted healing processes such as compaction and cementation. Although this mechanism finds support in a multitude of field observations, there have been only few systematic laboratory studies reproducing and quantifying the occurrence of incipient cementation in the laboratory seismic cycle. In addition, frictional healing has usually been experimentally measured at relatively low shear strain, often overlooking the “strain history” of laboratory faults.

We present a suite of 15 friction experiments in which we performed Slide-Hold-Slide (SHS) tests to evaluate the healing of quartz gouges under hydrothermal conditions. The temperature range investigated spans from 23 to 400 °C, at different effective normal stresses and fluid pressures. We also documented the role of shear strain in controlling the evolution of frictional healing through systematic repetitions of SHS tests at different amounts of strain of the laboratory fault. Our results indicate that frictional healing is positively dependent on temperature, especially at temperatures corresponding to the onset of crystal plasticity for quartz (> 350 °C). Best fit lines of healing measurements at 400 °C deviate from the classical log-linear time-dependent Dieterich-type healing, following an exponential relationship between ∆μ (frictional healing) and the logarithm of hold time. This suggests that incipient cementation processes play a major role during quasi-stationary, interseismic periods, better reflecting the higher fault healing usually observed in natural environments. In addition, experimental results relative to high strain SHS tests revealed that the “strain history” of laboratory faults exerts a strong control on the evolution of friction during the experiments. These results improve our understanding of a critical healing mechanism, constraining the dependence of frictional healing on temperature and shear strain.