Flow mechanism and force modelling of a central-slotted bridge's self-sustained flutter
Abstract
This study investigates the flutter mechanism of a 5,000m bridge with a wide-slotted deck. Preview full aeroelastic model testing indicated that the flutter of this deck was self-sustained rather than violently destructive. Follow-up sectional model tests confirmed that aerodynamic nonlinearity is the cause of the self-sustained flutter. The study aims to understand the mechanisms behind this nonlinearity and develop computational methods to analyze nonlinear aerodynamics.
To understand the mechanisms, CFD and PIV were used to collect data on the flow fields, pressure distribution, aerodynamic forces, and dynamic responses at 85m/s. The PIV phase-averaged flow fields revealed that flutter is not caused by vortex travel. Instead, it is linked to periodic changes in the separation scale of the windward deck, variations in the inflow conditions of the leeward deck, and alterations in the streamline pattern within the wake. The CFD results showed that self-sustained flutter is influenced by nonlinear aerodynamic forces, particularly the highly nonlinear moments. The third-order frequency plays a crucial role. The windward moment acts as the driving force, while the leeward moment acts as a stabilizing factor that limits the vibration amplitude. The combination of PIV flow fields and CFD pressure distribution revealed that the aerodynamic nonlinearity of the windward deck is caused by the periodic changes in the leading-edge separation scale. The aerodynamic nonlinearity of the leeward deck is attributed to the inflow conditions, which are affected by the periodic changes in the height difference between the two decks, resulting in additional wind loading.
The second part of the study addressed the limitations of existing state-space methods for calculating nonlinear aerodynamics. An updated nonlinear form was proposed, maintaining structural similarity. The polynomial expansion was employed to represent nonlinear hysteresis, and the B-curve fitting method was used to account for the amplitude dependence. The study combined the L-M algorithm and the RK-4 method to develop a novel approach for the nonlinear fitting of aerodynamic forces. By substituting the identified coefficients into the dynamic equations, the study demonstrated that this new method has the potential to accurately predict the self-sustained flutter response under various wind speeds.