Although a great deal is known about the location of earthquakes and their danger to society, our knowledge of the underlying physics of how they nucleate and, especially, the physical processes operating as the slipped area on the fault expands during an earthquake is still poorly understood. Greater knowledge of these processes is necessary to better predict seismic shaking danger and, it is hoped, to one day enable prediction of major earthquakes. This EAGER project will use experimental studies and high-resolution electron microscopy to test a new hypothesis of how the slip on continental earthquakes occurs.
Earthquakes are understood to initiate by two distinctly different processes: In the cold, low-pressure, environment of the upper few tens of km within the Earth, earthquakes generally begin by overcoming static friction on pre-existing faults. However, earthquakes also occur continuously to depths approaching 700 km in subducting oceanic lithosphere where the pressure is too high to allow brittle failure. Experiments show that shear failure (faulting) at high pressure requires a mineral reaction that yields a small amount of 'fluid' for their initiation and expansion; the 'fluid' can be either a true fluid (eg. H2O or CO2) or a nanocrystalline solid exhibiting an extremely low viscosity in the solid state. One of the two PIs of this project (Reches) is a leading expert in frictional sliding and the other (Green) is the world leader in high-pressure shear failure. They are joined by two leading experts on the physics of earthquakes of the US Geological Survey (D. Lockner and N. Beeler) at no cost to the project. This EAGER project will test the hypothesis that the process of mineral-reaction-induced shearing instability, the mechanism of faulting at high pressure, can also operate in shallow earthquakes where it is activated by the frictional heating/straining that occurs during initiation of earthquake slip. The team envisions two main ways in which this may occur: (1) Breakdown of clay minerals or carbonates in the fault zone releasing a fluid (water or CO2, respectively) that results in a large drop of the resistance to sliding on the fault; (2) generation of extremely small particles during initiation of sliding that form a nanocrystalline solid that can flow by grain-boundary sliding at seismogenic speeds, as has already been demonstrated for high-pressure faulting. Models of earthquake slip, experiments, and examination of fault zones in the field strongly suggest that shear-heating-induced devolatilization occurs in some earthquakes. The high-pressure experimental observations that such reactions lead to shearing instabilities further suggest that similar processes could enhance shallow earthquakes. Similarly, recent laboratory work in at least two laboratories concludes that powder-lubrication may be a critical part of fault propagation and lubrication. The key question we will test experimentally is whether such shear-heating-induced mineral reactions can lead to rapid drop in friction and/or enhancement of slip under shallow crust conditions.
The team will investigate the role of 'fluid'-producing reactions in fault mechanics. They will activate shear-induced devolatilization in laboratory experiments at the University of Oklahoma by high-speed sliding under a range of normal stresses. They then will characterize by high-resolution Scanning and Transmission electron microscopy at UC Riverside the microstructure of gouge and sliding surface produced in these experiments and compare those microstructures with the 'superplastic' fault-filling materials produced in high-pressure faulting experiments. Preliminary results on carbonate that are very encouraging.
Broader Impacts: This project brings together scientists and graduate students from two university campuses and a federal government lab for an EAGER project with potentially profound consequences for residents of earthquake-prone areas such as California. If this work demonstrates that devolatilization is directly responsible for friction drop and/or that fault gouges of large earthquakes are weak nanocrystalline solids, they will have opened a door that will lead to greater understanding of faulting and potentially will lead to a better understanding of which parts of which faults in a given area such as California are dangerous and which are not. The students of this project (one at OU and the other at UCR), with the likely addition of undergraduate assistants, will receive training on state-of-the-art instrumentation and will participate in research at the frontier of their science.
Project outcomes and findings Intellectual merit The main goal is testing the hypothesis that processes of mineral-reaction may induce earthquake instability at shallow crustal depth by frictional heating/straining of a fault zone. We considered processes like (1) Devolatilization or phase(s) tranformation (clay minerals or carbonates); and (2) generation of gouge with nanometric scale grains during rupturing. It was demonstrated that nanocrystalline material can flow by grain-boundary sliding at seismogenic speeds for high-pressure faulting. To meet this goal, we conducted (a) ultra-microscopic analysis of experimental faults; (b) analysis of the high-velocity shear along carbonate faults; and (c) analysis of universal mechanisms of fault weakening from high- PT to low-PT (PT=pressure-temperature). These analysis led to the following discoveries: The discovery of powder rolls in fault powder. For millennia, human-made machinery incorporated rolls and wheels to reduce frictional resistance, yet similar elements are practically absent in natural systems. We found that tiny, cylindrical rolls spontaneously develop along experimental rock faults leading to drastic dynamic weakening. The experimental faults were composed of six rock types that slipped at velocities approaching 1 m/s and normal stresses up to 14.5 MPa. The fault-slip was localized along highly smooth surfaces that frequently displayed a multitude of cylindrical rolls of diameter ~ 1 micron and length range up to 20 microns. The presence and density of these rolls correlate with the measured friction reduction that reflects a transition to rolling dominated slip. This rolling is an effective mechanism of powder lubrication, and the spontaneous growth of such rolls along crustal faults is likely to control earthquake weakening. Brittle-ductile transition of carbonate faults at high velocity shear. The wear-rate and frictional strength of a fault profoundly affect its geometry and slip stability. We ran an extensive series of shear experiments on three types of carbonate samples at a wide range of slip-velocity and normal stress. The analysis, which focuses on the steady-state stage, reveals that the values of wear-rate and frictional strength depend on both slip-velocity and normal stress. We interpret the observed intensity variations of wear-rate and frictional strength as indicating a brittle to ductile transition associated with frictional heating. Universal mechanisms of fault weakening from high-PT to low-PT. Frictional sliding is strongly dependent on normal stress, hence the increase of pressure with depth prohibits such sliding deeper than ~50km and requires that deeper earthquakes initiate by a di?erent mechanism than crustal ones. However, there is no similar constraint on the sliding mechanisms. Here we developed the first quantitative sliding mechanism that explains the low sliding resistance of high?pressure faulting and that can equally well explain the rapid decline of sliding resistance in high?speed friction experiments. Our analysis revealed that the microstructures of high?pressure fault "gouge" and the "gouge" of high?speed friction experiments that do not melt are topologically indistinguishable – fully dense, randomly?oriented, nanocrystalline solids. These observations of nanometric fault zone at both high-PT and low-PT indicate fast flow by grain?boundary sliding (gbs), as was first shown in ceramic materials. We propose that these high?speed experiments, as well as most earthquakes nucleated at any depth greater than a few kilometers, slide by the same mechanism as high pressure faults – flow of a very thin nanocrystalline "gouge" that forms in the first second or so of sliding and flows by grain boundary sliding. Broader impacts Education: One PhD student (Xiaofenf Chen), and one MS student (Yuval Boneh) were involved in all aspects of the project. Publication: Three journal papers were published or in review. Societal benefits: Better understanding of earthquake processes, as done here, will, on the long run, improve the design and constructions of facilities to withstand seismic risk. Contribution to other disciplines: Our discovery of the "powder-rolling" mechanism along experimental faults (above) is a likely mechanisms of powder lubrication in material sciences and engineering.
