Slow Slip Processes On Frictional Faults - Simulations in a Laboratory Setting

English Abstract:

We present laboratory observations derived from a direct shear test subjected to a normal stress and a controlled direct shear, where the effective strength heterogeneity is governed by the non-uniform distribution of asperities throughout the interface. We were able to map out the asperities and normal stress upon them using a pressure sensitive film. Prior to rapid sliding, we observed slow slip, which accumulated non-uniformly along the fault. Large, densely distributed asperities retarded slow slip and produce a ‘locked’ section with relatively low shear displacement. Slow slip was measured using slip sensors place at seven locations along the fault and a nucleation zone was seen to grow (at rates ~ 3 to 12 mm/s) from the free edge into a ‘locked’ section of the fault. Slow slip showed intermittent, ‘burst-like’ increases in spectral power between the frequencies of 60 to 150 Hz. These events lasted between ~ 12 seconds and were accompanied by burst-type acoustic emissions lasting ~ 0.5 to 3 μs. A local increase in slip rate was observed after its cessation. This increase in slip rate is consistent with observations of slow slip made before and after the 2013-2014 Boso slow slip event [Fukuda et al. 2014]. Further examination of these events shows a spatiotemporal dependence described by a time delay between adjacent along-strike slip sensors. Foreshocks were recorded using Glaser-type acoustic emission sensors and they occurred at the latter stages of the slow slip phase and indicated a source radius ranging from 0.21 to 1.09 mm – similar to larger asperity size measured with the film. Duration of these foreshocks were 7 orders of magnitude shorter than the ‘burst-like’ signals – similar observations have been made in actual geological settings [Ide, 2007]. These results will aid the development of a mechanistic model of slow slip that will improve the understanding of the interaction of slip and effective fault strength heterogeneity.