System identification of a R/C bridge based on ambient vibrations and 3D numerical simulations of the entire soil-structure system

The scope of this paper is to identify the parameters affecting the dynamic response of an existing R/C bridge, based on low ambient amplitude vibration measurements and numerical predictions using complex finite element models. For this purpose, the instrumented, 2^nd Kavala Bypass Ravine Bridge constructed along the Egnatia Motorway Greece is studied and a refined three- dimensional (3D) FEM is developed that takes into consideration the coupling and dynamic interaction of the overall superstructure-foundationsoil and deck-abutment-embankment system. The instrumentation schemes and the necessary algorithms applied for computing the modal characteristics of the bridge are discussed, while the modelling assumptions made for the soil-structure system are comparatively assessed and justified for various models of different levels of complexity. Given the large number of the system's degrees of freedom, a manual, modal-based FEM updating method is also presented. The results show good agreement between the measured and computationally predicted dynamic characteristics of the structure. They also show that the accurate estimation of the pier, deck and bearings stiffness is a key parameter for reliable system identification.