Grain Fragmentation and Frictional Melting During Initial Experimental Deformation and Implications for Seismic Slip at Shallow Depths

Abstract

During seismic slip, the elastic strain energy released by the wall rocks drives grain fragmentation and flash heating in the slipping zone, resulting in formation of (nano)powders and melt droplets, which lower the fault resistance. With progressive seismic slip, the frictional melt covers the slip surface and behaves as a lubricant reducing the coseismic fault strength. However, the processes associated to the transition from grain fragmentation to bulk frictional melting remain poorly understood. Here we discuss in situ microanalytical investigations performed on experimentally produced solidified frictional melts from the transition regime between grain fragmentation and frictional melting. The experiments were performed on granitic gneiss at seismic slip rates (1.3 and 5 m/s), normal stresses ranging from 3 to 30 MPa. At normal stresses <12 MPa, the apparent friction coefficient μ app (shear stress versus normal stress) evolves in a complex manner with slip: μ app decreases because of flash weakening, increases up to a peak value μ p1~0 .6-1.0, slightly decreases and increases again to a second peak value μ p2~0 .44-0.83, and eventually decreases with displacement to a steady-state value μ ss~0 .3-0.45. In situ synchrotron observations of the solidified frictional melt show abundance of ultrafine quartz grains before μ p2 and enrichment in SiO 2 at μ p2 . Because partial melting occurs on the ultrafine quartz grains and, as a consequence, it suggested that the second re-strengthening (μ p2 ) is induced by the higher viscosity of the melt due to its enrichment in Si from melting of the ultrafine quartz grains derived from grain fragmentation. Key Points:• Granitic rocks sheared at seismic slip velocities and low normal stress show double-strengthening friction evolution before final weakening • Particle size distribution constrained by in situ synchrotron analysis shows the presence of additive ultrafine quartz grains • The re-strengthening was due to viscous frictional melts by quasi-equilibrium melting and the Gibbs-Thomson effect on quartz grains

Numerical Modeling Of Frictional Melting Dynamics Constrained By Surface Micro-roughness

“…This peak stress corresponds to the shear stress strengthening stage due to incipient melt formation, which separates two stages of weakening observed in gabbro experiments (Hirose and Shimamoto, 2005). Similar shear stress evolution is observed at experiments with lower stress and slip rates, yet with one more stage of strengthening preceding the incipient melt formation due to a range of possible mechanisms (Hung et al, 2019). With higher heat production rates in natural fault zones at seismogenic conditions, we expect that there is only one stage of strengthening, which is much less pronounced than observed in experiments.…”

“…While the survivor clasts based on the imaged pseudotachylyte make up around 20% of the total volume, this fractional value is also variable during melting. Multiple stages slip experiments (Hung et al, 2019) show that the solid fraction evolves from 0.20 to 0.11 and returns to 0.20 at the final stage. For our simulations, we assume that the solid fraction is constant at 0.20 throughout the melting, and examine this assumption in the discussion.…”

“…where ⌘ is defined as the average fluidity (reciprocal of viscosity), i.e., ⌘ = (1/w) R w 0 (dx/µ c (x)). This quantity can be evaluated precisely given a temperature profile and a temperaturedependence viscosity model or estimated roughly from published viscosity data (for instance, Hung et al, 2019). it is, however, uncertain if the same happens during faulting nucleated at great depth and hence higher normal stress.…”

Determination of Parameters Characteristic of Dynamic Weakening Mechanisms During Seismic Faulting in Cohesive Rocks

“…Shear stress evolution with slip inferred from experiments performed on silicate‐built bare rocks, such as basalt and gabbro, is relatively well reproduced by the composite model made of flash heating and bulk melting, except of granitoid rocks for which the model yields the highest SNR values at low effective normal stress (≤10 MPa) (Figure S7 in Supporting Information S1). A possible explanation is that our viscosity model does not include the complexity of selective melting typical of frictional melting in granitoid rocks (Di Toro & Pennacchioni, 2004; Hung et al., 2019; Papa et al., 2021). Indeed, dedicated high‐velocity experiments using granitoid rocks showed that at low normal stresses (≤10 MPa) the compositional and structural evolution of the melt layer is extremely complex (Hung et al., 2019).…”

“Bristle-State” Friction: Modeling Slip Initiation and Transient Frictional Evolution From High-Velocity Earthquake Rupture Experiments

“…Motivating data are plotted with a linear time scale with inversions in the Supplement. (Reches and Lockner, 2010), grain fragmentation to weaker phases (Hung et al, 2019;Rattez and Veveakis, 2019), acoustic fluidization (Van der Elst et al, 2012), flash-heating (Goldsby and Tullis, 2011;Kitajima et al, 2011), frictional melt (Niemeijer et al, 2011), and/or other chemical processes (Violay et al, 2013). As the slip deficit is used up, fault slip must decelerate and friction will re-strengthen or heal (Violay et al, 2019), allowing seismologically-inferred self-healing pulsed slip behavior (Heaton, 1990) and larger dynamic than static stress drops (Brune, 1970).…”