SCEC Award Number 18097 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Experimental Study of Thermal Pressurization and the Role of Fault Roughness
Investigator(s)
Name Organization
Terry Tullis Brown University
Other Participants Nick Beeler, USGS
SCEC Priorities 3f, 2e, 2d SCEC Groups FARM, Seismology, Geology
Report Due Date 03/15/2019 Date Report Submitted 05/16/2019
Project Abstract
 This project investigates the high-speed frictional weakening mechanism of thermal pore-fluid pressurization using laboratory experiments. One goal has been to determine whether faults with roughness similar in amplitude to natural faults could cause enough dilatancy to inhibit the theoretically predicted weakening. Our previous experiments on flat surfaces show weakening in some cases and not in others, suggesting that some uncontrolled variable may explain the lack of reproducibility. In order to attribute lack of weakening with rough surfaces to dilatancy we must be able to show that every time we test a flat surface it undergoes weakening. We have found the likely uncontrolled variable, namely the epoxy that holds the samples into their steel grips does not always act as a fluid pressure seal as is intended, thereby preventing build up of thermally produced fluid pressure. Having identified this problem we have conducted additional experiments. They show the differences between no frictional weakening in dry samples and, for properly-sealed saturated samples, increasing weakening with decreasing permeability. We also find that, following the rapid weakening upon a large step-increase in slip velocity, some unexpected restrengthening occurs. This is apparently due to an unexpected decrease in pore-fluid pressure following its rapid thermally-induced rise. Based on analyzing the stress state in our experiments we believe that tensile stresses develop as the pore fluid pressure increases and that this can not only create a reservoir for the fluid to enter, it also increases the hydraulic diffusivity, both of which could reduce the fluid pressure.
Intellectual Merit The project increases our understanding of the physics of the earthquake process, one of the prime goals of SCEC. This high-speed weakening mechanism has been investigated in many theoretical studies but our research represents the most definitive experimental study of the process and is showing that some important effects are not yet included in the existing theoretical analysis.
Broader Impacts One of the major contributions of SCEC is to perform complex numerical calculations of dynamic rupture during an earthquake and thereby increase our understanding of the origins of strong damaging coseismic ground motions. The stress resisting slip during an earthquake, together with the ambient stress prior to the earthquake, determine the earthquake’s dynamic stress drop and through that the intensity of strong ground motions. Some of the most advanced of these dynamic rupture calculations employ extreme frictional weakening due to the process of thermal pressurization. If this weakening did not occur or were less extreme it would be one factor that would decrease strong ground motions in such models. Our lab measurements are intended to determine whether the theoretical predictions of the amount and rate of weakening due to this mechanism are in fact found experimentally, and are therefore realistic to include in rupture simulations, or whether the predictions need to be altered.
Exemplary Figure Figure 3. Friction and fault normal displacement during a velocity step from base velocity v_0=3.162 μm/s to v=2.5 mm/s, for samples 340 (a) and 346 (b). A transient increase in friction is observed during the step up, followed by dramatic weakening. Afterward the fault regains some of its original strength. Note the big difference in the amount of dilation between these samples, with about 10 times more dilation in sample 346 (a) than sample 340 (b).
Linked Publications

Add missing publication or edit citation shown. Enter the SCEC project ID to link publication.