ABSTRACT
ABSTRACT The Cervena Bridge is a steel truss bridge built in the Czech Republic in 1886. The assessment by the partial factor method in Eurocodes reveals its insufficient reliability. In particular the bridge fails to satisfy the equilibrium limit state when strong wind occurs simultaneously with an unloaded train on the bridge. To avoid structural interventions or expensive traffic restrictions, a model of the bridge is tested in a wind tunnel to obtain force coefficients of wind pressure for the specific shape of the bridge and selected types of light-weight trains. The values of force coefficients based on wind tunnel tests correspond to 70–80% of those provided in EN 1991-1-4. Traffic flow records for the railway line under consideration are utilised to obtain distributions of weights and heights of light-weight trains. Detailed probabilistic analysis reveals that bridge reliability is close to the target level when the wind tunnel force coefficients are considered along with free-field wind and railway traffic records. It is shown that case-specific information may help to reduce numerous “hidden safeties” in reliability assessments, often included in the models for permanent actions and geometry, wind pressure, traffic load, and the combination of the last two.
