If you have any problems related to the accessibility of any content (or if you want to request that a specific publication be accessible), please contact us at firstname.lastname@example.org.
CCEER-22-01: Nonlinear Analysis of Near-Fault Structures Using Physics-Based Simulated Earthquake Ground Motions
CCEER 22-01.pdf (24.61Mb)
AltmetricsView Usage Statistics
This report presents the results of several studies on the use of physics-based earthquake simula- tions in engineering analysis of building structures. Three-dimensional physics-based simulations offer new opportunities to advance our understanding of the impacts of fault rupture characteristics, seismic wave propagation patterns and site conditions on the dynamic response of civil structures to strong ground shaking, especially at short distances from active faults. In pursuit of this goal, we study the characteristics of broadband simulated ground motions in the near-fault region where field recordings are typically sparse, propose procedures to incorporate those simulations into en- gineering analysis, and assess their impacts on structures at the regional scale. The first chapter of this report focuses on the characteristics of near-fault ground motion char- acterized by strong velocity pulses, known as pulse-type records. Over the past two decades, seismic performance assessment methods have incorporated pulse-type records into the analysis of near-fault structures in several ways. We examine the assumptions associated with the empirical classification of ground velocity records as pulse-type or non-pulse, and show that this type of classification may be deficient in representing the potential impacts of ground motion records on building structures. We systematically study the differences between the characteristics of records classified as pulse or non-pulse using over 23,000 simulated records, and identify several inten- sity measures that may be suited for selecting representative records for the analysis of near-fault structures, without distinction between pulse-type and non-pulse ground motions. In the second chapter, we propose a procedure for selecting simulated earthquake records for performance-based seismic design of buildings. We design a deterministic framework to assess the bias associated with employing different intensity measures and selection criteria, and put forth recommended best practices to reduce the bias associated with predicting the seismic demands on structures using a relatively small number of curated records from a large database of simulated ground motions. We conclude that restricting the source distance characteristics associated with the underlying pool of candidate records is the most effective strategy to reduce the bias, assuming the candidate pool is sufficiently large and diverse, and may eliminate reliance on the empirical classification procedures of near-fault records. In the final chapter, we present a case study that involves predicting the regional-scale impacts of a magnitude 7.0 rupture of the Hayward fault on buildings across the San Francisco Bay Area using two fault rupture realizations in a domain representative of the region’s geological character- istics. We examine the effects of increasing the highest resolved frequency in the simulations, and reducing the minimum shear wave speed represented in the domain on the demands on low-, mid- and high-rise buildings. We discuss the important interactions between the site representation and forward directivity effects near the fault, and show that low-rise buildings may be more sensitive to the geological properties than high-rise buildings.
Report No. CCEER-22-01