SCEC Award Number 17253 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Experimental Investigation of Multi-scale Flash Weakening
Name Organization
Frederick Chester Texas A&M University Judith Chester Texas A&M University
Other Participants Ms. Monica Barbery
SCEC Priorities 1d, 3c, 2d SCEC Groups CS, SDOT, FARM
Report Due Date 06/15/2018 Date Report Submitted 11/03/2018
Project Abstract
Weakening of faults by flash heating is regarded as one of the more important processes governing fault friction behavior, particularly during the onset of high-velocity slip and the early phases of earthquake rupture development. The purpose of this project was to document the relationship between the characteristics of flash-heated contacts and frictional weakening behavior in order to improve the constitutive descriptions of transient and steady-state flash weakening. We take advantage of our high-speed biaxial rock deformation apparatus equipped with a high-speed thermal imaging camera to document the nature of flash heating of rock surfaces during controlled seismic slip. We machine grooved sliding surfaces on test specimens of Westerly granite in order to dictate the size, life-time and rest-time of mm-scale contacts. We have documented the effect of rest-time on flash temperature of contacts and the relationship to magnitude of flash weakening and steady-state friction. These data remove the uncertainty of the average lifetime and rest-time of contact populations undergoing flash-heating, which permits analysis of constitutive models and determination of material parameters for flash weakening in faults with multi-scale roughness using numerical, thermomechanical models.
Intellectual Merit Although experimental studies have documented weakening at high slip-rates, and that the observed weakening behavior is well-characterized by the steady-state, velocity-weakening relation based on flash heating at micrometer-scale asperity contacts, the evolution of contact geometries is not well-defined making it difficult to develop a velocity-dependent constitutive relation that describes transient friction during the accelerating and decelerating phases of earthquake slip. The steady-state, flash-weakening friction-relations generally assume contact dimensions of micrometers, but there is no reason to assume that frictional contacts that form during seismic slip on natural faults will have the same geometry and normal stress distributions as those of flat, ground surfaces in quasi-static contact. Natural faults surfaces display anisotropic roughness over a range of length scales, which are characteristics difficult to simulate in laboratory experiments. We take advantage of our high-speed biaxial rock deformation apparatus equipped with a high-speed thermal imaging camera to document the nature of flash heating of rock surfaces at seismic slip rates. This sample size is larger than most other high-speed testing instruments, which facilitates study of roughness variation over a larger length scale, in both the slip-parallel and -perpendicular directions. We demonstrate that high-speed friction experiments employing samples with larger surface areas and geometrically machined, rough surfaces provides key, unique data that helps constrain theoretical models of flash-weakening, and is a viable approach to understanding friction on natural fault surfaces that are known to be rough at multiple scales. The outcomes of this work directly address the SCEC science priority of quantifying stress heterogeneity on faults at different spatial scales, correlating the stress concentrations with asperities and geometric complexities, and modeling their influence on rupture initiation, propagation, and arrest.
Broader Impacts This project supports the research of Ms. Monica Barbery (Ph.D. candidate) at Texas A&M, which involves the development of novel experimental techniques to study the fundamental physics of dynamic frictional weakening. The work is contributing to an broader effort to construct and utilize a novel and capable testing instrument that can provide unique types of mechanical and observational data pertaining to physics of earthquake rupture propagation and radiation of seismic energy. The instrument and research advances will be useful to the earthquake physics research community for years to come.
Exemplary Figure Figure 5. Illustration of using flash heating models with experiment results from sliding machined surfaces to determine the local normal stress distribution from thermographs of the frictionally heated surfaces. Given the accurately machined surface geometries, records of displacement, and temperatures derived from imaging as a function of time during the velocity-step experiments, the contact Life-Time, LT, and Rest-Time, RT, history can be determined for every heated spot in the image. In this example, we identify a spot (box, located by white arrow) in the thermograph (a) that has a sliding history of a complete LT, RT, and then LT prior to emergence. In (b), the spot temperature history is plotted as a function of local normal stress assuming spot friction equals macroscopic friction, and local normal stress for the spot is constant during contact LT. These assumptions suggest a local normal stress three times that of the average macroscopic contact stress achieves the observed temperature of ~300˚C; however, less restrictive assumptions and modeling will determine local normal stress distributions more accurately