Changing dynamic of bridge aerodynamics

The assurance of safety and reliability of long-span bridges under winds requires the accurate modelling of windinduced effects. In this study, the existing conventional linear and non-linear models of bridge aerodynamics are first systematically reviewed with a focus on the study of the relationships among them. After highlighting the shortcomings of conventional models, three advanced non-linear models, based on artificial neural networks, nonlinear moving average (within the framework of Volterra theory) and Volterra series, are introduced. Example applications of linear and non-linear analysis schemes are presented, underscoring the changing dynamic of bridge aerodynamics under smooth and turbulent wind conditions.