Files in this item



application/pdfKang_WonHee.pdf (1MB)
(no description provided)PDF


Title:Development and application of new system reliability analysis methods for complex infrastructure systems
Author(s):Kang, Won Hee
Director of Research:Song, Junho
Doctoral Committee Chair(s):Song, Junho
Doctoral Committee Member(s):Wen, Y.K.; Spencer, Billie F., Jr.; Ha, Christopher
Department / Program:Civil & Environmental Eng
Discipline:Civil Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Bayesian method
Bridge network
Complex system
Connectivity analysis
Cut-set system
Finite element reliability analysis
First order reliability method
Fragility Curve
Importance measure
Link-set system
Matrix-based computations
Multivariate normal integral
Parallel system
Progressive failure
Reliability bounds
Second order reliability method
Series system
System reliability
Abstract:The failure event of a structure or lifeline network is often described by a complex logical function of multiple component failure events. Despite significant advances in theories on reliability analysis of individual components and their adoption in practice, the critical knowledge and quantitative methods for reliability assessments of complex system events remain elusive, leading to unknown accuracies in the risk assessment. Such a system reliability analysis is computationally challenging, especially when the definition of the system event is complex, the system has a large number of components, or the component events have significant statistical dependence due to common source effects. To overcome these challenges, this study develops two system reliability analysis methods, termed the Matrix-based System Reliability (MSR) Method and the Sequential Compounding Method (SCM), and applies the methods to risk assessment of complex structural systems and lifeline networks. Unlike existing system reliability analysis methods, the MSR method is applicable to any general system events, and can estimate not only system reliability but also component importance measures and parameter sensitivities of system reliability, which are essential metrics for risk-informed decision-making processes. The MSR method is applied to a bridge transportation network, a highway bridge structural system, and truss structures. The method is further developed to achieve improved efficiency using the first- or second-order reliability method; and to evaluate the sensitivity of the system failure probability with respect to parameters that affect the statistical dependence between the components. These further developments are demonstrated by risk assessment of progressive failures of a generalized Daniels system structure and by finite element system reliability analysis of a bridge pylon system. This study also aims at developing new methods for stochastic damage detection of pipeline networks based on the MSR method. The methods allow for efficient uncertainty quantification of system quantities such as network flow measures and for updating the component damage probabilities based on post-disaster observations on network performance. The accuracy and efficiency of these methods are demonstrated by a water pipeline network with 15 pipes that is subjected to an earthquake event. The sequential compounding method (SCM) is also developed to compute the probability of a general system event described in terms of a multivariate normal distribution. The merit of the SCM is its superior efficiency compared to existing system reliability methods including the MSR method. The accuracy and efficiency of the SCM is tested by a wide range of numerical examples including large systems consisting of 1,000 components. Due to its wide applicability, accuracy and efficiency, the method is expected to enhance the computational capability in various applications of system reliability analysis.
Issue Date:2011-05-25
Rights Information:Copyright 2011 Won Hee Kang
Date Available in IDEALS:2011-05-25
Date Deposited:2011-05

This item appears in the following Collection(s)

Item Statistics