| Summary: | 碩士 === 國立臺灣大學 === 土木工程學研究所 === 106 === Seismic probabilistic risk assessment (SPRA) has been widely used to compute the frequencies of core damage and release of radiation of nuclear power plant (NNP). In 2011, Huang, Whittaker, and Luco published a new SPRA methodology, different from existing SPRA, including nonlinear response-history analysis of structure to estimate seismic demands for component and the component fragility curves are functions of component response parameters
In the study presented herein, the methodologies of developing ground-motion-based and response-based component fragility curves are presented. A residual heat removal (RHR) piping system which contributes significantly to the seismic risk of a sample NPP has been analyzed experimentally and numerically. System identification of the tested piping system was conducted using Enhanced Frequency Domain Decomposition (EFDD), Multiple Singular Spectrum Analysis (MSSA) and Singular Value Decomposition (SVD). The numerical models validated using the results of system identification analysis and experiments were used to develop response-based fragility curves for the system. Comparing to the results of ground-motion-based component fragility analysis, those of response-based fragility analysis are more conservative and reliable, so the former is recommended for seismic probabilistic risk assessment.
|
|---|
