Formation of the Near-Fault Damage Zone during Dynamic Rupture in a Crystalline Rock Mass

Abstract

Many models for analyzing the dynamic propagation of seismogenic ruptures are based on solving classical problems of fracture mechanics. It is assumed that the fault is a shear crack with uniformly distributed friction and stress concentration at the crack tip. Known fracture mechanics theories do not describe the formation of damage zones in the lateral direction, i.e. perpendicular to the crack plane. Observational data indicate the presence of a fairly extensive zone of damaged material in the vicinity of the fault. This is the zone of dynamic influence where the material has an increased fracture density, higher permeability and lower elastic wave velocities. A correct assessment of the properties and sizes of zones of dynamic influence is crucial for constructing adequate earthquake preparation models. This paper analyzes regularities of development and quantitative characteristics of the damage zone during dynamic earthquake rupture and quasi-static evolution of the fault. The size and mechanical characteristics of the near-fault damage zone produced by movement along the slip surface can be conveniently estimated by the second invariant of the deviatoric stress tensor (shear intensity). Matching of the calculated value with a certain degree of rock mass damage can be done using measurement data from large-scale explosions, by comparing them with the calculation results. It is shown that coseismic movement along the fault leads to insignificant changes in the properties of the host rock. However, the longitudinal wave velocity near the fault decreases markedly by 30–35%, the permeability increases only by approximately a factor of three, and the increase in the degree of fracturing is almost unnoticeable. This means that the properties of the rock mass change due to the opening of preexisting cracks. Repeated movements do not radically change the characteristic dimensions and properties of the damage zone. It is concluded that the fault-affected zone is formed mainly at the quasi-static stage of the formation of the main fault through the coalescence of individual macrofractures, and future seismogenic movements renew the already existing fractures.