System Reliability Model for Long Span Arch Bridges

Abstract

Every structural system consists of elements with certain dependencies between each other. Current design codes do not account for these interactions and they provide provisions only for particular components. Thus, the design and evaluation procedures do not include additional redundancy and ductility, which can have significant influence on the safety of the whole structure. In this dissertation, the reliability of the 800-ft long span, steel arch bridge is evaluated. The steel arches made of built-up box sections with different properties along the length of the span are the primary components of the structure. Monte Carlo simulation technique is used to determine the reliability indices for all sections of the arch. Additionally, the system reliability is determined for different levels of correlation between segments of the arch. To conduct the reliability analysis, load and resistance models were developed and limit state functions were formulated. Traffic data collected from the Weigh-In-Motion station located near the considered bridge serves the purpose of development for the live load model. Using these records, a uniformly distributed load for critical situation (traffic jam on the bridge) was calculated. In addition to the traffic load case with collected Weigh-In-Motion data, live load cases in accordance with AASHTO Standard Specifications for Highway Bridges and AASHTO LRFD Bridge Design Specifications were implemented in the analysis. Beam-column interaction equations served as a basis for limit state function. In order to obtain values of axial forces and bending moment for segments of the arch, the Finite Element Model of the considered bridge was created. Three types of analyses were used: linear, nonlinear and buckling analysis.