No abstract

Classical rate‐and‐state friction (RSF) laws are widely applied in modeling earthquake dynamics but generally using empirically determined parameters with little or no knowledge of, or quantitative account for, the controlling physical mechanisms. Here a mechanism‐based microphysical model is developed for describing the frictional behavior of carbonate fault gouge, assuming that the frictional behavior seen in lab experiments is controlled by competing processes of rate‐strengthening intergranular sliding versus contact creep by pressure solution. By solving the controlling equations, derived from kinematic and energy/entropy balance considerations, and employing a microphysical model for rate‐strengthening grain boundary friction plus standard creep equations for pressure solution, we simulate typical lab‐frictional tests, namely, “velocity stepping” and “slide‐hold‐slide” test sequences, for velocity histories and environmental conditions employed in previous experiments. The modeling results capture all of the main features and trends seen in the experimental results, including both steady state and transient aspects of the observed behavior, with reasonable quantitative agreement. To our knowledge, ours is the first mechanism‐based model that can reproduce RSF‐like behavior in terms of microstructurally verifiable processes and state variables. Since it is microphysically based, we believe that our modeling approach can provide an improved framework for extrapolating friction data to natural conditions.

[1] Previous rotary shear experiments, performed on a halite-muscovite fault gouge analogue system have shown that the presence of phyllosilicates, under conditions favoring the operation of cataclasis and pressure solution in the matrix phase, can have major effects on the frictional behavior of gouges. While 100% halite and 100% muscovite samples exhibit rate-independent frictional/brittle behavior, the strength of mixtures containing 10-30% muscovite is both normal stress and sliding velocitydependent. At high sliding velocities (>1 mm s À1 ), such mixtures show unusually marked velocity weakening, along with the development of a structureless, cataclastic microstructure. In the present paper, a micromechanical model is developed in an attempt to explain this behavior. The model assumes a granular flow process involving competition between intergranular dilatation and compaction by pressure solution. The predictions of the model agree favorably with the experimental results. Extension of the model to quartz-mica systems implies that the presence of phyllosilicates plus the operation of pressure solution can strongly promote (unstable) velocity-weakening behavior at rapid slip rates on natural faults, under midcrustal conditions. Static stress drop predictions based on the model agree reasonably well with estimates from seismic observations. Our results may help explain the discrepancy between laboratory-derived rate-and-state friction parameter values, obtained for dry, low-strain and/or single-phase rock systems, and the values for natural fault rocks inferred from seismological data.Citation: Niemeijer, A. R., and C. J. Spiers (2007), A microphysical model for strong velocity weakening in phyllosilicate-bearing fault gouges,

No abstract

It is widely believed that grain size reduction by dynamic recrystallization can lead to major rheological weakening and associated strain localization by bringing about a switch from grain size insensitive dislocation creep to grain size sensitive diffusion creep. Recently, however, we advanced the hypothesis that, rather than a switch, dynamic recrystallization leads to a balance between grain size reduction and grain growth processes set up in the neighborhood of the boundary between the dislocation creep field and the diffusion creep field. In this paper, we compare the predictions implied by our hypothesis with those of other models for dynamic recrystallization. We also evaluate the full range of models against experimental data on a variety of materials. We conclude that a temperature dependence of the relationship between recrystallized grain size and flow stress cannot be neglected a priori. This should be taken into account when estimating natural flow stresses using experimentally calibrated recrystallized grain size piezometers. We also demonstrate experimental support for the field boundary hypothesis. This support implies that significant weakening by grain size reduction in localized shear zones is possible only if caused by a process other than dynamic recrystallization (such as syntectonic reaction or cataclasis) or if grain growth is inhibited.

The rheological properties of rock salt are of fundamental importance in predicting the long-term evolution of salt-based radioactive waste repositories and strategic storage caverns, and in modelling the formation of salt diapirs and associated oil traps. The short-term, high-stress rheology of rock salt is well known from laboratory experiments; however, extrapolation to appropriately low stresses fails to predict the rapid flow seen in certain natural structures. Furthermore, experiments have failed to reproduce the recrystallized microstructure of naturally deformed salt. Here we report experiments indicating that the above discrepancies can be explained by taking into account the influence of trace amounts of brine. Trace brine is always present in natural salt but sometimes escapes during experiments. Our tests on dry dilated salt show more or less conventional dislocation creep behaviour, but brine-bearing samples show marked weakening at low strain rates. This is associated with dynamic recrystallization and a change of deformation mechanism to solution transfer creep. Because natural rock salt always contains some brine, these results cast substantial doubt on the validity of presently accepted dislocation creep laws for predicting the long-term rheological behaviour of salt in nature.

No abstract

No abstract