No abstract

No abstract

No abstract

<p>The main deck of a suspension bridge with a main span in excess of 1000m was analysed probabilistically to determine the implied safety level for the deck with utilization ratios in ULS scaled to 1. The scaling was necessary because other design criteria influence the design as well. The cross-section of the bridge is a closed steel box. Structural reliability analysis was performed in which detailed consideration was given to stochastic modelling of resistance and load variables and of their associated uncertainties. Emphasis was placed on accurate and appropriate modelling of the considered limit states as well as on the traffic loading on the structure. In this regard weigh-in- motion data was employed to develop site-specific live loading models for comparison with design models and load effects. In this way the sensitivity of design assumptions and their influence on the implied safety was assessed. The detailed sensitivities of wind-induced section forces from a full dynamic analysis are utilized, and appropriate stochastic models have been found for the influencing parameters. The paper details all aspects of the reliability analysis and provides information on the evaluated implied safety level of the bridge deck. The results of the sensitivity analysis are instructive in understanding the influence of uncertainties on the results, especially the parameters used in the wind-induced response analysis and the assumptions of the traffic load model.</p>

The paper presents the practical application of the Danish Road Directorate guideline for probabilistic assessment of bridges to an actual bridge structure, which failed to demonstrate sufficient load carrying capacity following deterministic assessment. The bridge, a post-tensioned box girder structure, was built in 1965. In addition to presentation of the safety classification of the structure the results of importance and sensitivity studies are presented which indicate the controlling variables. Following the assessment, bridge specific partial safety factors are calibrated for the critical limit state at the β-point which are compared with the global partial factors prescribed for deterministic assessment. It is important to point out that at no stage does the guideline compromise the safety of the structures presented, rather it provides a mechanism for determining their structure-specific safety for comparison with that deemed legally acceptable.