Approximately half of the 575,000 highway bridges in the United States were built before 1940. Forty-two percent of these bridges are classified as deficient. The structural deficiency can be attributed to age, external loading, environmental conditions, material characteristics, or human errors. Limited funds are available for rehabilitation and strengthening. Therefore, there is a need for methods to identify the most sensitive parts of girder bridges to damage. This may help to lower the costs of checking, inspection, repair, and replacement.;In this study, sensitivity functions for damaged bridge girders and systems are developed. The limit state which is considered in the reliability analysis is the ultimate flexural moment. Structural damage is considered on two levels: local and global. Local damage results in a reduction in one or more resistance parameters of an element, such as reinforcement area and yield stress. Global damage is defined as the damage which reduces the ultimate capacity of a structural member, without reference to a specific resistance parameter. The study is carried out on four types of bridge girders: noncomposite steel, composite steel, reinforced concrete, and prestressed concrete. Probabilistic moment-curvature relationships for such girders are generated using Monte Carlo simulation.;System reliability models for girder bridges are developed. They take into account simultaneous occurrence of trucks on the bridge as well as transverse truck position in the lane. The effect of correlation between girder strengths is investigated. The relationship between element and system reliabilities is developed in order to quantify redundancy in girder bridges. This relationship considers the effect of girder spacing, bridge length, and girder size. Bayesian method is employed to update the reliability of existing bridges as new information is obtained from field tests.
展开▼