Dynamic and static identification of base-isolated bridges using Genetic Algorithms

In the paper, an identification approach based on a Genetic Algorithm (GA) is applied to the case study of a base-isolated, post-tensioned concrete bridge investigated in earlier contributions of literature. It is known that bearing isolators greatly influence the overall response of small- and medium-span bridges under dynamic loads, but in previous works it was seen that the characterisation of their elastic stiffness under small displacements may be inaccurate. In this work, based on in-situ test measurements obtained under static and dynamic loading conditions, inverse techniques based on GAs are successfully applied to the examined structural system, providing an efficient and well-calibrated structural identification of its main properties. Compared to other identification tools and classical correlation techniques, the main advantage deriving from the use of inverse approaches based on GAs typically manifests in the possibility to estimate a greater number of material parameters (e.g. properties of concrete as well as stiffness of the bearing isolators, etc.), and to critically assess the accuracy of the identification. Based on rather good correlation between test measurements and finite element (FE) model updating, it is expected that the same technique could be applied to various structural typologies and systems.