• International journal of steel structures
• /
• v.18 no.5
• /
• pp.1508-1524
• /
• 2018

Hysteretic energy dissipating devices (HEDDs) have been increasingly applied to building construction to improve the seismic performance. The seismic responses of such damped structures are significantly affected by HEDD's structural properties. An accurate investigation on the propagation of HEDD's structural properties is required for reasonable evaluation of the seismic performance of a structure. This study aims to develop simplified methods that can estimate the collective uncertainty-propagation to the seismic response of damped structures employing HEDDs. To achieve this, three- and six-story steel moment-resisting frames were selected and the propagations of the individual HEDD's property-uncertainties were evaluated when they are subjected to various levels of seismic demand. Based on the result of individual uncertainty-propagations, a simplified method is proposed to evaluate the variation of seismic response collectively propagated by HEDD's property-uncertainties and is verified by comparing with the exact collective uncertainty-propagation calculated using the Monte Carlo simulation method. The proposed method, called as a modified SRSS method in this study, is established from a conventional square root of the sum of the squares (SRSS) method with the relative contributions of the individual HEDD's property-uncertainty propagations. This study shows that the modified SRSS method provides a better estimation than the conventional SRSS method and can significantly reduce computational time with reasonable accuracy compared with the Monte Carlo simulation method.

• International Journal of Reliability and Applications
• /
• v.2 no.1
• /
• pp.27-36
• /
• 2001

This Paper proposes a new methodology for assessing the reliability of an accident management, which Is based on the reliability physics and the scheme to generate dynamic event tree. The methodology consists of 3 main steps: screening; uncertainty propagation; and probability estimation. Sensitivity analysis is used for screening the variables of significance. Latin Hypercube sampling technique and MAAP code are used for uncertainty propagation, and the dynamic event tree generation method is used for the estimation of non-success probability of implementing an accident management strategy. This approach is applied in assessing the non-success probability of implementing a cavity flooding strategy, which is to supply water into the reactor cavity using emergency fire systems during the sequence of station blackout at the reference plant.