Hide/Show Apps

Probabilistic methods for estimating the seismic deformations of underground structures

Soyman, Kadir Buğra
In seismic design of underground structures, simplified semi-deterministic methods based on the stiffness of the soil and the underground structure are used to estimate the seismic deformations. Generally, peak ground acceleration (PGA) value for different hazard levels are provided by the probabilistic seismic hazard assessment (PSHA) analysis, but a single (and deterministic) PGA value is employed in deformation calculations. The objective of this study is to propose a fully probabilistic framework for estimating the free field deformations (dff) of underground infrastructure, especially the metro stations. For this purpose, stations of two metro lines planned to be constructed in Istanbul are used as case studies. A new seismic source characterization model (SSC) is developed for the segments of North Anatolian Fault Zone that are close to metro stations. This SSC model is combined with current global and regionalized ground motion prediction models in PSHA framework to estimate the design ground motions. Simplified soil profiles of metro stations are developed and incorporated into 1-D equivalent linear (EQL) analysis. Using analysis results, two prediction models are developed for estimation of dff. According to performance-based earthquake engineering framework (PBEE), these models are incorporated into the hazard integral to provide annual rate of exceedance for specified dff levels and 475-year return period dff values are estimated for each station. Analysis results show that the 475-year return period dff values are significantly different than the dff values calculated by current semi-deterministic methods, indicating that fully probabilistic approach should be utilized in seismic design of underground structures.