Collaborative Research: Development of Realistic Seismic Input Motions for Improving the Resilience of Infrastructure to Earthquakes

The ability to reconstruct a seismic wave field in a domain of interest from sparsely-measured seismic ground motion data can help engineers to accurately model potential damage during earthquakes, improve safety, and reduce costs. Realistic seismic ground motions are essential for improving design and assessment of infrastructure by engineers, owners, and regulators. Although a large amount of ground motion data are available from modern sensors (e.g., accelerometers, optical cables, etc.), no established method can reconstruct the full 3 component (3C) incident wave field from the measurements in a three dimensional (3D) near-surface domain. This Disaster Resilience Research Grants (DRRG) project will address this need by developing a new method for reconstructing a full, 3C seismic wave field within a soil/rock volume adjacent to infrastructure from field measurements. The resulting 3C seismic wave field obtained by this approach accounts for local geology and variability, and can be used as a realistic seismic motion input into models of structures and infrastructure to assess their performance during earthquakes.

Current use of one component (1C) motions for horizontal and vertical seismic shaking introduces a number of epistemic, modeling uncertainties into soil-structure interaction analysis. Regional-scale wave models need information about seismic sources, and deep and shallow geology that introduces large epistemic and aleatory, parametric uncertainties in the generated seismic motions. This project will develop a method for resolving these issues and providing accurate, realistic seismic motions that will improve modeling and simulation of earthquake-soil-structure interaction (ESSI) behavior. Consequently, design of infrastructure and lifelines and assessment of their earthquake response will be improved, resulting in increased resilience to seismic loading. The method will be integrated into a public domain program, Real-ESSI simulator (http://real-essi.us). The methodology will be scalable to various types of measurement modes (e.g., full translational 3C, 6C (translational 3C with rotational 3C), vertical-only 1C or the amplitude of full-3C motions measured by accelerometers at discrete locations, surface vibrations measured by vision-based sensors, or 3C motions-along-lines measured by optical cables). An advisory panel will provide feedback on the project to facilitate translation of the research into industrial practice. The PIs will develop online educational material on 'Inverse Modeling for ESSI Systems'. Such educational effort and material will help educate not only students working on this project, but also undergraduate and graduate students worldwide, as well as practicing engineers with interest in modeling of ESSI behavior.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.