Skip to main content
eScholarship
Open Access Publications from the University of California

Ground Motion Simulation Validation through Seismic Performance Assessment of Structural Systems

  • Author(s): Munjy, Huda Riadh
  • Advisor(s): Zareian, Farzin
  • et al.
Abstract

Recorded ground motions are traditionally used to represent future earthquakes for the seismic performance assessment and design of structures. However, there is a shortage of recorded ground motions that are able to represent a variety of possible scenarios (specifically large magnitude and short distance events). Structural engineers are hesitant in using simulated ground motions as surrogates for missing natural recordings, believing they are not equivalent to recorded ground motions in accurately estimating seismic demand in structures. This research recommends a validation test for simulated ground motions to increase the engineering community’s confidence in using them in engineering applications.

This dissertation first introduces a novel statistical approach that can be used in current seismic performance assessment methods. Gaussian copulas are used to characterize the dependence structure of demand used for the assessment and calculation of loss in buildings due to earthquakes. Using Gaussian copulas is shown to increase the overall accuracy in seismic performance assessment methodology compared to the procedure that is currently being used, outlined by FEMA P-58 (2015). Following the proposal of a more accurate method to characterize structural demand used in seismic performance assessment, a three step methodology, titled Vector-Based Intensity-measure Method (VBIM), for ground motion simulation validation at the structural response level is proposed. Three case study structures are used for the application of VBIM: 2- and 12-story special steel moment-resisting frame (SMRF) buildings and a two-span, cast-in-place concrete bridge. Results indicate that models of simulated and recorded ground motions that predict structural response based on waveform parameters are similar. The results of this study provide recommendations for ground motion simulators regarding the required accuracy of these key parameters in order for simulated ground motions to accurately predict structural response while also providing several steps of validation that show similarities between recorded and simulated ground motions.

Main Content
Current View