Modeling vortex-induced vibration of long-span bridges.

摘要

Vortex-induced (locked-in) vibration of long-span bridges is one of the primary aerodynamic problems in bridge-wind engineering. Many efforts have been made in the past to model vortex-induced vibration of circular cylinders (e.g., chimneys, marine cables, silos, and pipelines), but limited modeling attempts have been reported for vortex-induced vibration of non-circular sections representative of long-span bridge decks. Existing methods are not adequate for the prediction of vortex-induced vibration of long-span bridges, so the problem is only partially or approximately addressed in design that sometimes result in costly mitigation measures during or after construction. This type of vibration, while less destructive than flutter, can lead to user discomfort (both physical and psychological), material fatigue, and non-structural failure. Moderate-amplitude vortex-induced vibration, which may occur at relatively low wind speeds, near the natural frequency of a flexible structure is primarily caused by a nonlinear interaction between the oscillation of the body and the fluid wake.; The goal of this research was to develop a suitable model, complementary to those existing for flutter and buffeting, which improves current techniques for the prediction of vortex-induced vibration of non-circular sections representative of bridge decks. Wind tunnel experiments of six spring-mounted rigid section models were performed in order to further advance the understanding of the mechanism of vortex-induced vibration of non-circular sections. A practical semi-empirical model that captured the salient features observed in the above investigations was developed and a procedure for extracting the aeroelastic coefficients of the model was developed. The mathematical model developed for a spring-mounted rigid section was extended for flexible, three-dimensional prototype structures through modal analysis. Additionally, spanwise loss of correlation of the aeroelastic coefficients was incorporated in the model. Long-term, full-scale data measured on the Fred Hartman Bridge (a cable-stayed bridge) were analyzed to detect incidents of vortex-induced response using several criteria that were proposed in this study. The full-scale data were also used to identify the modal damping ratio and frequency of the bridge. Finally, the vortex-induced response of the Fred Hartman Bridge was predicted using the analytical model; the predicted responses demonstrated good agreement with the full-scale responses.