An investigation of the mechanisms causing large-amplitude wind-induced vibrations in stay cables using unsteady surface pressure measurements

摘要

Dry galloping has been observed on in-service bridges and has been reproduced in several wind tunnel experiments of inclined stay cables. In certain cases, the large-amplitude vibrations caused by dry galloping could not be mitigated with damping levels specified by the Post Tensioning Institute (PTI). Wind tunnel investigations were conducted for IHI Corporation on a 1:1 scale sectional model of an inclined bridge stay cable in the 3 m x 6 m Wind Tunnel at the National Research Council Canada in 2015. The purpose of the investigation was to reproduce large amplitude cable galloping, which had been observed at an existing cable-stayed bridge in Japan. The experiment was designed to investigate the influence of different damping levels and the ability of a 5 mm-diameter helical fillet to mitigate dry cable vibrations. The current work identified several important factors that contributed to dry galloping and large-amplitude cable motions. It was shown that the physical mechanisms leading to the onset of large-amplitude motion in the critical and supercritical Reynolds number regimes are distinct. In the critical Reynolds number regime, the drop in drag, increase in lift, and fluctuations in the laminar separation bubbles along the cable destabilize the low pressure lobes, allowing the changes in pressure along the cable to become synchronized with the cable motion in a highly-correlated manner. This type of galloping could not be mitigated with damping greater than that required by the PTI. At high Reynolds numbers, the boundary layer state changed and was associated with the upstream movement of the separation point and an asymmetric pressure distribution, !Carman vortex shedding re-emerged and appeared to combine with low-frequency variations to induce large-amplitude motion. The large-amplitude motion at high Reynolds numbers could be mitigated with additional damping.