The spectral cell method in nonlinear earthquake modeling

Daniel Giraldo, & Doriam Restrepo

Published August 1, 2017, SCEC Contribution #7283

This study examines the applicability of the spectral cell method (SCM) to compute the nonlinear earthquake response of complex basins. SCM combines fictitious-domain concepts with the spectral-version of the finite element method to solve the wave equations in heterogeneous geophysical domains. Nonlinear behavior is considered by implementing the Mohr-Coulomb and Drucker-Prager yielding criteria. We illustrate the performance of SCM with numerical examples of nonlinear basins exhibiting physically and computationally challenging conditions. The numerical experiments are benchmarked with results from overkill solutions, and using MIDAS GTS NX, a finite element software for geotechnical applications. Our findings show good agreement between the two sets of results. Traditional spectral elements implementations allow points per wavelength as low as
PPW=4.5 for high-order polynomials. Our findings show that in the presence of nonlinearity,
high-order polynomials (p >= 3) require mesh resolutions above of \geq$10 to ensure displacement errors below 10 percent.

Giraldo, D., & Restrepo, D. (2017). The spectral cell method in nonlinear earthquake modeling. Computational Mechanics, 60(6), 883-903. doi: 10.1007/s00466-017-1454-8.