A semi-empirical model for vortex-induced vertical forces on a twin-box deck under turbulent wind flow

Abstract

Long-span cable-supported bridges with twin-box decks have been constructed in recent years to enhance the stability of the bridges against flutter, but some of these bridges suffer large-amplitude vortex-induced vibration (VIV). Large-amplitude VIV can cause fatigue damage accumulation in structural components and reduce safety and comfort levels of road vehicles, trains, and pedestrians. Therefore, in the design of long-span bridges with twin-box decks, it is essential to know vortex-induced forces (VIF) and predict vortex-induced responses (VIR). Currently, there are several semi-empirical models for the estimation of VIF but most of them are established based on measured VIR for single-box decks and without considering turbulent wind effects.This paper presents a semi-empirical model for vertical VIF on a twin-box deck with and without turbulent wind flow. First, the two existing semi-empirical single-degree-of-freedom (SDOF) VIF models are discussed, and a refined SDOF model for VIF under smooth wind flow is proposed. The proposed SDOF VIF model includes all the non-conservative motion-induced force terms. An approximate analytical solution for the maximum amplitude of VIR is then deduced based on the proposed VIF model. Second, based on quasi-static assumption, the proposed VIF model is further extended to take account of turbulence wind effects by introducing a turbulence-induced positive damping term into the model. Finally, the validity of the proposed VIF model with and without turbulent wind effects is examined using a newly-developed wind tunnel test technique for an elastically-mounted twin-box section model. The comparative results show that the proposed VIF model can effectively predict the maximum VIR of a twin-box deck under different turbulent fields with different structural damping ratios. It is also found that the maximum VIR decreases nonlinearly with the increase of turbulence intensity.