Cable-stayed bridge model updating through analytical formulation, finite element model and experimental measurements

The evaluation of the dynamic response of cable structures is a topic of relevant interest for both researchers and designers working in the civil engineering field. For the implementation of reliable numerical models able to simulate the actual mechanical behaviour of such structures, the cable tension is a fundamental variable to be assessed accurately. This contribution describes methods and procedures aiming to enhance the reliability of finite element models in reproducing the dynamic response of a cable-stayed pedestrian bridge, through the fusion of static and dynamic experimental data from different technologies. First, the experimental evaluation of the cable static configuration has been used for identifying the cable tension. Indeed, by virtue of a low-order approximation of the catenary suspended cable model (obtained by a perturbation method), a cubic function of the cable static configuration under self-weight has been determined. In this formulation, the quadratic and cubic coefficients have been recognized as tension-sensitive quantities and have been used to analytically evaluate the cable tension. Then, traditional output only dynamic tests, under environmental noise, have been exploited to evaluate the reliability of the computational model of the bridge and to analyze the deck modes mainly involved in the dynamic behaviour. The subsequent manual model updating, targeted at reducing the difference between the experimental and numerical frequencies, showed higher effectiveness if the designed cable tensions are replaced by the identified values in the computational model.