The occurrence and magnitude of surface earthquake faults are generally evaluated empirically based on past observation records and geological survey results. On the other hand, simulation analyses of surface breaks of earthquake faults have been conducted using source models, but most of them assume homogeneous ground, and there are few cases considering the layered velocity structure and damage zones around the fault. In this study, the effects of damage zones and velocity layers on the occurrence of surface earthquake faults and strong ground motions were investigated through dynamic rupture simulation using a three-dimensional finite element method on the Hinagu Fault, where the foreshock of the 2016 Kumamoto Earthquake occurred. The results showed that consideration of damage zones reduces fault slip, lowers seismic moment, and inhibits rupture propagation to the surface.
Past earthquake resistant designs are based strong motions. However, following the 1999 Chi-Chi earthquake in Taiwan and the Kocaeli earthquake in Turkey, which caused damage to many dams and bridges due to faulting induced deformation on the ground surface, the issue of dealing with fault displacement has become an important issue.
The amount of displacement of a surface rupture correlates with the size and length of the fault, and is generally evaluated empirically based on past observations or by static analysis assuming a fault plane and slip volume. On the other hand, dynamic numerical simulations have been used to evaluate surface rupture, which have the advantage that complex fault geometry and ground conditions can be taken into account, and the interconnectedness of adjacent faults and displacement of the ground surface can be evaluated. However, it is not easy to apply this method to seismic design because of the large influence of fault modeling, initial conditions, boundary conditions, and other factors.
Fracture zones (damage zones, DZs) formed by fault movements exist around surface ruptures, and their width and fracture properties vary depending on the earthquake scale. Using a planar 2-D dynamic FEM in which the fault is modelled with joint elements, Fukushima et al. (2010) show that the planar strong ground motion distribution can be well simulated by considering the DZs around the fault in two lateral fault earthquakes.
