Loading...

The Effect of Exact and Approximate Distribution of Conditional Spectrum in Probabilistic Estimation of Structural Demands

Shokouhirad, Sahar | 2012

634 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 43723 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Golafshani, Ali Akbar
  7. Abstract:
  8. In this thesis several methodological aspects of both conventional and novel approaches to probabilistic seismic hazard assessment are discussed here along with subjects more strictly pertinent to structural and geotechnical engineering. This work, however, has a unique, ultimate goal: to provide procedures for an effective and practical probabilistic estimation of structural demand of offshore platforms due to seismic ground shaking.In modern multi-level design and rehabilitation procedures, structural performance is in fact quantified by deformation demand levels exceeded with a specified annual frequency of structural life. We refer to this as Probabilistic Seismic Demand Analysis (PSDA). Necessarily, this ambitious task requires one to address the problem from multiple perspectives which range from the characterization of the seismicity in the region around the site, through the evaluation of the level of ground shaking at the surface of the local soil deposit, to the computation of the resulting response of the structure subject to such a ground motion.In this research, the probabilistic seismic demand analysis (PSDA) of offshore platforms in the Persian Gulf region has been investigated in terms of Mean Annual Frequency (MAF) of exceeding a given level of response. This approach combines the site-specific ground motion hazard information with the structural response of interest. Accordingly, in this study has been used from a computationally efficient and theoretically consistent algorithm to select the ground motions that match the site-specific target response spectrum mean and variance. The target spectrum used in this research is the exact and approximate distribution of Conditional Spectrum (CS). On the other hand, the associated conditional spectra are conditioned on the geometric mean of the pseudo-spectral acceleration over a certain range of periods, instead of the target spectral acceleration at a specific period. Hence, the pros and cons of the proposed conditioning Intensity Measure (IM) are compared with the conventional conditioning case and also the exact and approximate distribution of conditional spectrum by estimating the probability of collapse and the drift hazard.All the above analysis to a real offshore platform is located in the South Pars field has been investigated. The results demonstrate that although the structural response of the offshore platform is sensitive to spectral acceleration values at several periods, all the record-selection strategies considered, however, lead to nearly identical probabilistic prediction of structural demands at different IM levels.As a result, if one selectground motions to match the full conditional spectrum (including variability), the sameresults will be obtained no matter which conditioning period and distribution of conditional spectrum to be used for conditionalspectrum calculations
  9. Keywords:
  10. Conditional Spectra ; Probabilistic Seismic Hazard Analysis (PSHA) ; Probabilistic Seismic Demand Analysis (PSDA) ; Jacket-type Offshore Platforms ; Demand Hazard Curve

 Digital Object List

 Bookmark

No TOC