AGU Fall Meeting, San Francisco, USA, 14-18 December, 2009
J.P.Vilotte, G. Festa, M. R. Raous, C. Henninger
Seismic rupture of large earthquakes propagates along pre-existing faults that have a complex internal structure (fault zone). The propagation and the radiation of a seismic rupture has long been considered in seismology as a friction dominated process, formulated as a propagating shear crack problem under the assumption of a Barenblatt-type surface energy. Simple friction laws, such as the slip weakening and the rate-and-state laws are commonly used in modeling the rupture process, because they well describe the low frequency content of near-fault strong motion data, whilst they insure finite stress and slip velocity at the rupture tip through the introduction of a characteristic cohesive length scale.
The high frequency signal data provided by the rapidly increasing quality and density of the seismological and geodetic data in the near fault region should provide new informations on smaller scale features of the rupture which require to go beyond the classical frictional shear crack paradigm. At the same time, the rapid development of new sophisticated and efficient numerical methods is providing us with tools for simulation of rupture propagation and high frequency radiation.
In this study, we first provide a short description of recent development of numerical methods based on non smooth contact mechanics and high-order variational approximation of the elastodynamics, together with non classical interface constitutive laws (Raous et al, 1999) that account for interface damage and breakdown processes as well as loading/unloading response generalizing the classical smooth linear slip-weakening friction law.
We investigate here the characteristics of the nucleation, propagation, arrest and high frequency radiation for planar and complex fault geometries and compare them to the features provided by standard friction law, such as slip weakening and rate-and-state laws. We finally investigate the effects related to off-fault dissipation in terms of volumetric dynamical damage, and its potential implication for the scaling of the radiated energy, for different fault geometries.
Finally, we discuss some open issues regarding the multi-scale modeling of earthquake rupture dynamics, and high frequency generation, when retaining a formulation based on surface energy.