FEM and BEM Community Fault Models for S. CA., Coseismic and Post seismic Deformation Benchmarks

  • Kenner, Shelley (PI)

Grants and Contracts Details


Recent significant improvements in the quantity and quality of observations pertaining to shallow crustal earthquakes have warranted models of greater realism and sophistication for the purposes of data interpretation. The need for the realistic description of the co-seismic, post-seismic, and inter-seismic deformation is particularly pressing in tectonically active areas where dense seismic and geodetic networks have been deployed, such as in Southern California. While models based on analytic or semi-analytic solutions for an elastic half-space proved to be very useful for the zeroth-order analysis of relatively sparse datasets, further progress may not be possible without an accurate account for vertical and lateral heterogeneities in the Earth crust, complex rupture geometries, non-linear and time-dependent constitutive properties of the crustal rocks, effects of gravity and topography, etc. These features can be readily modeled using Finite Element (FE), and (partially) Boundary Element (BE) techniques. The significant comlexity and diversity of the existing numerical codes require extensive validation and accuracy checks before the codes can be used for the interpretation of geophysical data. The agreed-upon benchmarks that capture the essential physics of the earthquake-related deformation, except for a few analytical solutions for the elastic half-space, are largerly lacking. We propose to create a set of rigorously tested FEM and BEM benchmarks that can be used by SCEC's modeling community for code testing and validation. We will also explore to what extent the simpler and computationally more efficient BEM solutions may be used to approximate complex time-dependent problems involving, e.g., viscoelastic relaxation below the brittle-ductile transition. Finally, we will work on the development of effective meshing tools for complex fault geometries (co-I Kenner), and establishment of a highresolution BEM model for the exploration of mechanical interactions between faults in Southern California as mapped by the SCEC community's fault model (IGPP postdoctoral associate and co-I Becker). PIs have extensive experience with large-scale finite element and boundary element simulations (Fialko et al., 2002; Fialko and Rubin, 1999; Kenner and Segall, 2000a,b). The finite element calculations will be performed using a commercial 3-D code ABAQUS, which is perhaps the most versatile, and thoroughly documented and tested FEM code suitable for crustal deformation modeling. This work will directly contribute to the ongoing FEM modeling efforts of SCEC's Fault Systems working group. Outcomes of this proposal will include: 1) A set of accurate time-dependent 3-D FEM benchmarks for visco-elastic post-seismic deformation due to simple strike-slip and dip-slip earthquake sources. We will also explore the poro-elastic capabilities of ABAQUS. 2) A boundary element tool for modeling fully relaxed visco-elastic and poro-elastic deformation due to an arbitrary superposition of dislocation sources. 3) A robust finite element mesh generator for reasonably realistic fault geometries, including nearfault mesh refinement, and accurate and computationally efficient handling of the far-field boundary conditions with the use of infinite elements. 4) A thoroughly tested BEM interface to the Community Fault Model (CFM) for Southern California. We intend to address these problems in the framework of the related efforts of the Fault Systems' Deformation Model group, as there is obvious overlap in the FEM and BEM modeling approaches, such as the issue of optimal meshing of fault geometries. The efforts described in this proposal are synergetic with, and substantially levereged by the related projects of the PIs that involve development and use of sophisticated FEM and BEM models.
Effective start/end date1/1/031/31/07


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.