Site-specific traffic load modelling for bridge assessment

Assessment of highway bridge structures requires accurate prediction of the load effects that a structure may be expected to resist during its required remaining life. Generally, these load effects are determined from codified models more appropriate to design than assessment. This approach is very conservative for existing structures as it fails to fully recognise the reduced level of uncertainty associated with evaluation of (a) the loads to which an existing structure is subjected and (b) the resistance that an existing structure provides. This paper proposes a more rational approach to load effect prediction for existing structures. In the approach the actual traffic loading to which a structure is subjected is measured on site and is employed in simulation to determine site-specific characteristic load effects. The methodology proposed is dependent upon the availability of unbiased site-specific traffic statistics. This paper presents the results of an experiment performed in Vienna and describes the use of weigh-in-motion (WIM) statistics in the prediction of site-specific characteristic load effects-that is, values with specified probability of exceedance during the required lifetime of the structure, against which the structure may be assessed. The advantage of the approach is demonstrated through reductions in the load effect values. The implication for bridge assessment is reduced rehabilitation costs and the avoidance of unnecessary replacement of serviceable highway structures.