Structural Reliability of Buildings Subjected to Simulated Ground Motions using a Novel Probabilistic Approach

Abstract

The catastrophic power of earthquakes has been reported in numerous documents. For example, the occurrence of certain ground motions as the 1985 Mexico City, 1989 Loma Prieta, 1994 Northridge, and 1995 Kobe, represented economic losses of about 4, 6, 30, and 200 billion dollars, respectively. During the shaking of the above-mentioned and several other earthquakes, an inadequate structural performance has been observed in Steel Moment Resisting Frames (SMRFs). In most of the cases, SMRFs are designed using simplified methods of analysis which are permitted by certain building codes. However, to extract as accurate as possible the structural response of SMRFs subjected to ground motions, the nonlinear behavior of them must be properly extracted, generally using time history analysis. Unfortunately, for certain locations, the main problem is related to the availability of ground motions records. In this sense, an option that is becoming very popular is the use of simulated ground motions. Nevertheless, the use of simulated ground motions in time history analysis is still a research topic under development. The knowledge gap is mainly associated to the lack of approaches to artificially generate ground motions and their proper validation for specific applications as the earthquake-resistant design of SMRFs. Based on the above discussion, an approach that integrates the finite element method, response surface method, and first order reliability analysis is developed to calculate the structural reliability of SMRFs excited by synthetic ground motions generated via the Broadband Platform (BBP). Such a software represents an open-access platform developed by the Southern California Earthquake Center. In this context, the validation of the BBP is performed with the help of a 2-story SMRF excited using real and simulated versions of the 1994 Northridge earthquake. Then, a comparison is presented in terms of response spectrum and structural reliability. Once BBP is properly verified, an adequate set of ground motions is generated for a specific zone in the Los Angeles area. Using such a set of ground motions, 11 of them are properly selected considering a target response spectrum and dynamic properties of a 3-story SMRF. Hence, reliability is calculated for the 3-story SMRF in terms of safety index and probability of failure.