Accuracy of lifetime prediction determined by fatigue monitoring

Since the times of erection both, traffic loads and density on road bridges, have increased significantly. To assess durability and sustainability of sometimes degraded structures, nowadays especially fatigue damage needs to be considered. The residual structural lifetime subjected to fatigue can be estimated by different prediction models. Obviously, the prediction should be as accurate as possible, but varies significantly w.r.t. a general uncertain input. In this contribution, the focus is set on an assessment of these alternatives. In order to minimise input scatter a structural monitoring was conducted on a more than 50-year-old pre-stressed concrete bridge with documented cracks nearby coupling joints. Strains of the pre-stressing strands and the concrete were measured for four weeks under traffic to determine accurate stress amplitudes. Based on the monitoring-results and due to well-documented traffic data available, an accurate load history of the structure is created. It serves for numerical fatigue analyses and includes time-dependent effects, such as creep and shrinkage. Uncertainty potentially introduced by other sources like vibrations and temperature is identified and separated by reference measurements without traffic and a temperature profile. Finally several samples of concrete and pre-stressing strands were taken from the bridge and tested in laboratory. The determined material parameters enable a precise description of the structure in a calibrated finite element model and a reduction of scatter due to uncertain parameters. Finally, two aspects are focused on: The time-dependent structural behaviour which also includes a retrofitting of the structure and the assessment of traffic loads based on measurement data. Both increase structural lifetime significantly and illustrate the helpful effect of monitoring measures.