Earthquakes generate dynamic stress concentrations near their rupture front and near geometrical irregularities that can exceed the strength of the surrounding material and produce inelastic brittle deformation in the bulk. Co-seismic off-fault failure induces both irreversible deformation (plasticity) and reduction of elastic moduli (damage) with potentially distinct signatures on the radiated wavefield and fault zone structure. Dynamic damage can amplify near-fault motions and produce bimaterial interfaces that affect the subsequent behavior of rupture. These feedback mechanisms between off-fault damage and dynamic rupture are not accounted for in models that represent off-fault dissipation by plastic yielding with unchanged elastic moduli. Moreover, natural faults contain geometrical features (roughness) over a broad range of length scales. The multiscale non-planar geometry of faults can enhance the complexity of dynamic rupture and high frequency wave radiation. Fault roughness also induces stress concentrations that can contribute to the generation of dynamic off-fault damage and affect the overall earthquake energy balance. The overarching goal of this project is to provide, through theoretical and computational modeling, quantitative predictions of the impact of off-fault brittle damage and fault roughness (and their feedback) on observable properties of earthquake rupture, seismic wave radiation and short-term evolution of fault zone structure. The research addresses directly the following fundamental questions about earthquake physics and fault dynamics: What are the intensity and spatial extent of the coseismic reduction of wave velocities around the earthquake source? How do off-fault inelasticity and fault roughness contribute to the apparent scaling of earthquake fracture energy? What are the limits imposed by damage on maximum slip rate and peak ground velocity? Can the across-fault asymmetry of dynamic damage induce significant bimaterial effects on rupture? Can dynamic brittle damage produce stronger wave radiation and generate observable non-double-couple seismic radiation? How does fault roughness affect the complexity of earthquake rupture, the distribution of damage and the properties of high frequency radiation? How is the macroscopic response of a fault related to the coupled mesoscopic properties of damage and roughness?
Current studies of earthquake dynamics by several groups are pushing the frontier beyond the classical model of frictional sliding on a planar fault in elastic media to understand the role of a more realistic and complete set of ingredients. The diversity and complexity of the physical processes involved calls for a step-by-step approach in which a few candidate ingredients are tested at a time. The proposed studies aim to investigate the combined effects on dynamic ruptures and seismic radiation of two fundamental ingredients: non-planar fault geometry (fault roughness on a broad range of scales) and off-fault material damage (reduction of elastic moduli). Other related studies consider at present only large scale, kilometric features of fault geometry, and represent off-fault inelasticity by an ideal plastic rheology. The current project will consider fault geometry over a broad range of scales, from meters to kilometers, and will include co-seismic reduction of elastic moduli in the off-fault yielding process. The results can have transformative impact on studies of properties of dynamic ruptures, limits to the maximum expected ground motion, and inversions of observed seismic data for earthquake source properties. The research may provide new target signals for constraining earthquakes processes and dynamic evolution of fault zone structure.
An improved understanding of properties of dynamic ruptures, high frequency seismic radiation and physical limits to ground motion will contribute to emerging physics-based approaches for earthquake engineering and mitigation of seismic hazard. The incorporation of material damage in the source process can have significant impact on many geophysical studies that are based on derived earthquake source parameters (e.g., fault slip and source mechanisms). The results will help to scale fault properties and processes from the laboratory to natural fault zones. Although our focus is on earthquake dynamics, the studies can also have impact on the solid mechanics and material science communities.
Mature tectonic faults are usually surrounded by a fault zone, i.e. a layer of damaged rocks where seismic waves propagate at lower speed than in intact rocks and where inelastic deformation can take place. Until recently, the off-fault dimension had been largely neglected in earthquake models. This project developed theoretical predictions of the potential effects of fault zones on the rupture process of earthquakes. The approach was essentially based on computer simulations of dynamic rupture on faults embedded in a fault zone. This work required further development of rupture simulation software and involved national and international collabboration. We discovered a new mechanism to explain the short duration of slip during most earthquakes. It results from unloading stresses brought back to the slipping fault by seismic waves reflected inside the fault zone layer. We found that this mechanism induces periodic spatial and temporal patterns that are characteristic features that may be identified in seismological, field and laboratory observations. We investigated how dramatic fault-weakening can contribute to the complexity of earthquake rupture. We found that repeated slip can occur in the hypocentral region if a pulse-like rupture leaves there a stress concentration. We systematically studied the effect of off-fault plastic deformation on dynamic rupture. We developed simple relations between key earthquake source parameters, namely rupture speed and peak slip velocity. his relation contributes to physics-based methods for the assessment of strong ground shaking. We found that plastic strain may distort first-order earthquake source parameters such as the moment tensor. We developed simplified dynamic rupture models of the 2011 Tohoku-Oki earthquake that reproduced some of the key features of its rupture process. The model contains deep brittle asperities embedded in more stable regions to generate overall slow rupture down-dip interpersed with high-frequency radiation bursts. This project provided training and research experience for five graduate students and enabled international collaborations between early career researchers.
