Earthquake Nucleation and the Initiation of Frictional Ruptures

Recent experiments have demonstrated that rapid rupture fronts, that are actually shear cracks, mediate the transition to frictional motion. Moreover, once these dynamic rupture fronts ("laboratory earthquakes") are created, their singular form, dynamics and arrest are well-described by fracture mechanics. Ruptures, however, need to be created within initially rough frictional interfaces, before they are able to propagate. This formation process, occurring bellow the critical ('Griffith') length for crack instability, is not described by our current understanding of fracture mechanics.
By conducting controlled nucleation experiments, in which the interface is continuously imaged, we have been able to gain a detailed description of the nucleation process. We find that the expansion of the nucleation patch is qualitatively different than the propagation of the fully formed rupture front. It occurs at extremely slow and constant velocities, and it is 2D in nature. Some of the features of this expansion, such as self-similar evolution and stress-dependent timescales, are general. However, the details of this process are governed by the local conditions at the nucleation region. Due to the slow rates of expansion, local variations in the surface toughness (the 'fracture energy') can influence characteristics such as the exact nucleation point, the shape of the patch, and the stress threshold that is needed for nucleation to occur.
As nucleation is not described by the usual frameworks that are used to explain rupture propagation, understanding the driving mechanism of it is of fundamental importance to questions ranging from earthquake nucleation and prediction to processes governing material failure. We propose a theoretical model for this mechanism, and show that it captures the unique characteristics of the nucleation process as well as the transition to dynamic rupture propagation that is described by standard fracture mechanics.