SCEC Award Number 13164 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Rupture growth on faults with low background shear stress
Investigator(s)
Name Organization
Paul Segall Stanford University Eric Dunham Stanford University
Other Participants Stuart Schmitt
SCEC Priorities 3, 3, 3 SCEC Groups FARM, Seismology, SDOT
Report Due Date 03/15/2014 Date Report Submitted N/A
Project Abstract
 Motivated by the discrepancy between observations that earthquakes occur on faults at low shear stress and laboratory experiments that show that faults have high shear strength (at modest slip rates), we investigate the role of heterogeneous stress in the mechanics of slip nucleation and dynamic rupture. In our model, slip nucleates where the ratio of shear stress to effective normal stress $\tau/(\sigma-p_0)$ is high at $\sim$0.7 and then propagates into regions where $\tau/(\sigma-p_0)<0.2$. For such a rupture to be self-sustaining, a strong weakening mechanism must operate during fast slip. We consider two such mechanisms, flash heating of asperity contacts and thermal pressurization of pore fluid. We have performed a suite of numerical simulations in which we enable or disable each of these weakening mechanisms, and compare the output to earthquake source parameters inferred from natural earthquakes. Both dynamic weakening mechanisms are capable of propelling dynamic rupture into regions of low shear stress. Earthquake stress drops may therefore be quite low relative to the static frictional strength of the fault, which is consistent with seismological observations. Our simulated ruptures reveal that flash heating does not permit fracture energy to scale with slip---as is thought to occur for natural earthquakes---unless an additional slip weakening mechanism such as thermal pressurization occurs. This result is consistent with thermal pressurization being active in both small and large earthquakes.
Intellectual Merit This work advances our understanding of the physics of earthquake slip. In particular, we address the SCEC goal of investigating the relative importance of different dynamic weakening mechanisms.
By providing a numerical model for earthquakes at low average shear stress, we begin to reconcile laboratory observations of high fault strength, geomechanical observations of low stress on faults, field observations of negligible heating in fault zones, and seismological observations of low earthquake stress drops.
Successful comparisons of our simulation output to source parameters inferred for natural earthquakes corroborate the findings of this and other theoretical studies.
Our results also provide constraints on stress levels required for rupture growth, which may inform the development of earthquake system simulators that try to reproduce long-term seismicity rates in fault systems. Currently, these simulators cannot include the complexity of dynamic rupture and consequently use \textit{ad hoc} methods to allow quasi-dynamic rupture growth.
Broader Impacts A core goal of SCEC is to develop a physics-based understanding of earthquake phenomena. In particular, physics-based numerical earthquake simulations coupled with accurate stress models and detailed fault maps may someday be capable of providing realistic assessments of seismic hazard within a region. Such information would have significant economic and social value. Our present study considers details in the physics of fault slip that, after years of development in numerical methods and in computational power, may become a useful element in those simulators. It is thus an advance in developing our physical understanding of earthquakes, which is an important first step in addressing a hazard that poses considerable risk to the public.
Exemplary Figure Comparison of small simulated earthquakes with identical initial conditions in which a small region of high shear stress is surrounded by low stress. The top row shows log~$v(x,t)$, where $v$ is slip speed, and the bottom row shows snapshots of slip $\delta$ at 1~ms intervals. In the first column, fault strength obeys a laboratory-derived rate- and state-dependent friction law. The second column adds the dynamic weakening effect of flash heating of asperity contacts, which significantly reduces the friction coefficient. The third column incorporates thermal pressurization with the rate/state effects of the first column, which results in significantly-reduced effective normal stress. The fourth column includes both flash heating and thermal pressurization. At least one dynamic weakening mechanism is required for the rupture to propagate far outside the high-stress region, and both dynamic weakening mechanisms together allow for an even larger rupture.