SCEC Award Number 19056 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Verification and Validation of 3D Nonlinear Physics-based Ground Motion Simulations: Phase I
Name Organization
Domniki Asimaki California Institute of Technology Kim Olsen San Diego State University Jamison Steidl University of California, Santa Barbara Michael Barall Invisible Software, Inc. Daniel Roten San Diego State University
Other Participants Elnaz Esmaeilzadeh Seylabi, Caltech
Fabian Bonilla, Universite Paris Est
Doriam Restrepo, Universidad EAFIT
Ricardo Taborda, Universidad EAFIT
Roberto Paolucci, Politecnico di Milano
Fernando Lopez Caballero, Paris-Saclay University
SCEC Priorities 4a, 4b, 4c SCEC Groups GM, CS, EEII
Report Due Date 04/30/2020 Date Report Submitted 05/05/2020
Project Abstract
Significant advances in physics-based earthquake ground motion simulations have allowed us to move towards higher frequencies and finer resolution scales. Some of the most prominent features that higher resolution models are called to account for are the inelastic behavior of geomaterials in the shallow crust, which include hysteresis and permanent or transient ground deformation. These effects span a wide range of scales, affect the amplitude and duration of ground motions over a broad range of frequencies, and are highlighted among the research priorities of SCEC5. This work (in progress) is the first phase of a verification and validation exercise of prominent ground motion codes developed and optimized for large-scale 3D nonlinear ground motion simulations. These include one finite element code (Hercules); one finite differences code (AWP) and two spectral element codes (SEM3D, SPEED). The Garner Valley Downhole Array has been identified as benchmark site for the validation exercises, in 1D and in future phases in 3D. Garner Valley was selected because of the wealth of published studies on site characterization and nonlinear properties, key elements of a successful long term validation study.
Intellectual Merit The intellectual merit lies in the development of a series of verification and validation exercises to evaluate codes and workflow of computing seismic wave propagation in nonlinear media.
Broader Impacts The proposed verification and validation exercise is the first phase of a long-term project that will extend through the end of SCEC5. During Year 1 (this work in progress), the various codes are being verified for idealized linear and nonlinear problems, and will be validated against 1D site response recordings at the Garner Valley downhole array. Dr. Fabian Bonilla was among the lead PIs of PRENOLIN and is an international participant on this effort; PI Asimaki was one of the developers who participated in the benchmark. In addition to the two SCEC High-F codes, two spectral element codes developed by researchers at Paris-Saclay University (SEM3D) and Politecnico di Milano will participate in the exercise, thereby broadening the learning experience and international outreach of our SCEC efforts.
Exemplary Figure Figure 1. Surface velocity time series at KiK-net site KSRH10 obtained from linear and Iwan-type nonlinear simulation using the GPU and CPU versions of AWP. The source time function consists of a Ricker wavelet with a central frequency of 3 Hz and a peak amplitude of 5 cm/s. Nonlinear predictions using the Iwan model were obtained using 10 yield surfaces.