Intellectual Merit:

Understanding and eventually predicting earthquakes requires fundamental, multi-scale, multi-physics models that integrate observations from laboratory experiments, seismology, geodesy, and field studies.  Such models rely on incorporating equations that describe frictional sliding of rocks over a wide range of sliding velocities ? from the slow slip rates on faults characteristic of earthquake nucleation - nanometers to microns per second - to rapid coseismic slip rates of meters per second.  Frictional behavior of rocks at slow slip rates is described by ?rate and state? (RSF) friction laws, which describe how friction varies with sliding velocity (rate) and the time of contact (state) of microscopic protrusions or ?asperities? on a fault surface where contact actually occurs. Though widely incorporated in earthquake models, RSF laws are empirical ? that is, derived from laboratory data, but the underlying physical processes are unknown. This severely limits the reliable application of RSF laws outside the limited range of conditions at which they were obtained in laboratory experiments.  

Rock friction studies show that the state of a frictional interface changes with time (or alternatively, with small amounts of slip) during stationary or nearly stationary contact. An increase in friction with time, or aging, is described by two relations that describe the evolution of state with time and slip, respectively. Aging yields a positive or negative dependence of friction on sliding velocity, the sign of which determines whether unstable frictional sliding (an earthquake) is possible. Despite the importance of aging, its physical mechanisms are unknown. The standard view is that aging results from increases in the real contact area between surfaces due to plastic deformation and creep of asperities.

To fill these gaps, we conducted a wide-ranging, interdisciplinary suite of experiments and computer simulations to explore the mechanisms behind rock friction. Experiments were conducted using atomic force microscopy (AFM) and nanoindentation.  In both methods, a small probe is brought into nanometers- to microns-sized contact with a sample, and the deformation and/or frictional behavior is studied. Our approach was to explore and understand the deformation and friction of a single asperity at a frictional interface, through experiments and simulations, then use the knowledge gained to ?scale up? to macroscopic frictional surfaces using computer models.    

A breakthrough result from our nanoindentation studies is that plastic yielding and creep of quartz is independent of relative humidity (RH) in the range ~0 to 50%.  Previous friction experiments on quartz rocks revealed no increase of friction with time at RH= ~0, but pronounced aging at higher humidity. The lack of aging without water has been attributed to the lack of a weakening effect of water on the plastic deformation of quartz. Frictional aging of quartz thus cannot be due to plastic deformation of asperities, the accepted view, pointing to the importance of chemical bonding.   

In our AFM and simulation studies, we further established chemical bonding as a mechanism of aging.  We: 1) estimated values for energy barriers to the formation of interfacial chemical (siloxane) bonds, and their dependence on the applied stress, using computer simulations; 2) found that unstable stick-slip motion occurs for nanometer-scale single-asperity contacts, as for earthquakes; 3) established a chemical-bonding origin for the slip weakening distance Dc, a critical parameter in RSF laws; 4) developed a new computational model of mechanochemical effects, and predicted a strong effect of temperature on chemical aging; 5) measured aging over very short times (a few milliseconds) to constrain aging mechanisms; and 6) extended a well-established model for the direct effect in RSF to incorporate the effects of aging.  

Broadly, our research established the importance of chemical aging for single-asperity contacts, providing a new view of aging of rocks. We constructed contact models using the knowledge gained to determine how chemical reactions and deformation couple to produce time-dependent friction for rough contacting surfaces. We found that regimes in which a logarithmic dependence of friction on time and a linear dependence of friction on load for chemical ageing of asperities may also exist for rough surfaces in contact ? i.e., for faults in the Earth.

Broader Impacts:

Our results have applicability beyond the Earth sciences, e.g. in micro- and nanoelectromechanical systems, in fabrication processes such as wafer bonding, and in the study of nanotribology. The project educated earth scientists in materials science and mechanics, and immersed materials scientists and solid mechanicians in the earthquake problem. The project supported the education of two PhD students and a post doc at the University of Pennsylvania and one PhD student at Wisconsin. Invited talks were given by the team at major societies, international conferences, and universities. The results were incorporated into multiple classes at UPenn and Wisconsin, conveyed to visitors at Philly Materials Day at Penn, and presented at a Science & Faith event at a local church in Madison, Wisconsin. The project produced 9 refereed publications.

Last Modified: 01/29/2021
Modified by: Izabela Szlufarska