Bridge model updating using distributed sensor data

One of the challenges of managing bridge infrastructure is developing numerical models that can be used to accurately assess highly redundant bridge systems. One way to refine the model is to use sensor data to perform model updating. However, conventional sensors provide limited data with which to update the model, given the many degrees of freedom associated with indeterminate structures, resulting in a large potential error. Distributed sensing technologies such as digital image correlation and fibre optic strain sensors have the potential to provide more extensive data sets for model updating. This paper presents a case study of a reinforced concrete bridge that was modelled numerically to predict the bridge performance. The bridge was then load tested, and distributed sensor data were acquired. Using the sensor data, the numerical model was updated and refined estimates of the bridge behaviour were obtained. The initial and final models produced estimates of bridge behaviour that differed by an order of magnitude, illustrating the importance of sensor data for some bridge assessments. Additionally, the model indicated that the stiffness of the bridge had increased with time owing to an increase in the elastic modulus of the concrete and the development of compressive stresses.