Presentation by Pr. Greg McLaskey
What slows and stops an earthquake rupture? Insights from large-scale laboratory earthquake experiments
Abstract:
For decades, dynamic earthquake rupture models have employed a fracture energy Gtot ≈ 1 MJ/m2, that is orders of magnitude larger than expectations based on laboratory experiments. Such a high Gtot is required to stop ruptures and to prevent sustained supershear ruptures, which are inconsistent with most earthquake observations. A scale-dependent earthquake fracture energy has been proposed to reconcile these differences, but a physical mechanism for such scaling is lacking, and candidate mechanisms have not been rigorously tested. My research group at Cornell has explored how dynamic rupture interacts with heterogeneous fault properties including a high-strength “bump”, regions of low initial shear stress, and fault sections with velocity strengthening frictional properties. We have found that low initial stress is a much more effective mechanism to stop rupture than high strength and fracture energy. In this talk, I review these meter-scale laboratory earthquake observations and present a model of highly variable initial stress as an alternative to scale dependent fracture energy. The variation in the initial stress must be so large that significant fault sections are under stressed and act as energy sinks. Such a highly variable stress may result from slip on fault surfaces with multi-scale roughness. This model helps reconcile the huge discrepancy between lab measurements and seismic inference without invoking a scale-dependent fracture energy, and motivates a new perspective on seismic observations that have been interpreted as evidence for enormous fracture energy for the past 40 years.
Zoom link:
https://columbiauniversity.zoom.us/j/97930168676?pwd=rKOxlsbIL018W6QnYrCVMqpzgUQ92t.1
