Investigation of Adaptive Time-Stepping Algorithms for Simulating Sequences of Earthquakes and Aseismic Slip (SEAS)

Yanke Song, Valere R. Lambert, & Nadia Lapusta

Published August 12, 2018, SCEC Contribution #8402, 2018 SCEC Annual Meeting Poster #193

Modeling earthquake behavior involves resolving the complex interaction of processes active across timescales ranging from hundreds to thousands of years of tectonic loading, down to milliseconds during the dynamic rupture process. To make such problems computationally feasible, simulations of sequences of earthquakes and aseismic slip (SEAS) employ adaptive time-stepping algorithms in order to optimally sample the temporal evolution of fault behavior. The choice of time steps during the dynamic and quasi-static phases of fault slip can impact the accuracy and long-term evolution of the computed quantities (e.g. fault slip, slip rates, and stresses) as well as the computational cost of such simulations. In order to explore the trade-offs for an optimal time-step selection, we study the simulations of SEAS with various adaptive time-stepping algorithms, comparing long-term fault behavior as well as the computational costs in terms of the total number of simulated time steps. We begin with a quasi-dynamic formulation of a 2D antiplane problem based on the first benchmark simulation from the SEAS code comparison exercise supported by the Southern California Earthquake Center. We plan to expand this study to explore the optimal adaptive time-stepping methods for simulations that include fully dynamic stress interactions and enhanced dynamic weakening mechanisms, such as the thermal pressurization of pore fluids.

Key Words
SEAS, adaptive time-stepping

Citation
Song, Y., Lambert, V. R., & Lapusta, N. (2018, 08). Investigation of Adaptive Time-Stepping Algorithms for Simulating Sequences of Earthquakes and Aseismic Slip (SEAS). Poster Presentation at 2018 SCEC Annual Meeting.

Related Projects & Working Groups
Fault and Rupture Mechanics (FARM)