Title:   Slow is fast: Dynamic crack-front distortions mediate the propagation of extended fractures

Abstract:
I am a visiting researcher at ERI from September 2025 through June 2026, hosted by So Ozawa. I will begin by briefly introducing myself and areas of recent work. These include (i) numerical and analytical models for fluid-induced slow slip, in which we show that fault slip can spread in a self-similar manner, like pore fluid pressure; and (ii) analytical understanding of rate-and-state models of the seismic cycle, in which we appropriately determine the linear and non-linear stability of model faults to understand how a given model will behave, a priori, on the basis of the loading conditions and frictional property distribution.

The remainder of this talk will focus on the propagation of an extended crack front through a stiff material. High-speed imaging is used to establish that fluid-driven, penny-shaped, planar mode-I cracks propagate through a series of stick-break instabilities. The forward propagation of an existing fracture front always occurs through an initial rupture, nucleated at some localized position, followed by very rapid transverse expansion of the crack front at velocities that can be controlled experimentally and that can be as high as the Rayleigh-wave speed (~1 km/s). The fluid viscosity controls the length of the break, which correlates with the transverse velocity. This novel observation—that seemingly slow, quasi-static fracture propagation actually occurs through a series of discrete fast, dynamic ruptures—should be applicable for any pre-existing fracture front in other geometries and materials.