Bridge buffeting by skew winds: A quasi-steady case study

Abstract

The buffeting response of a 5000-m-long curved floating bridge is investigated for different mean wind directions, with emphasis on the wind load formulations and use of experimental data. Wind tunnel tests, with a section model of the bridge girder, provided six aerodynamic coefficients for different yaw and inclination angles. A comprehensive 3D buffeting formulation with bivariate polynomial fits of the coefficients is proposed and investigated. For cases where the skew wind data is not available, alternative 2D formulations, including the traditional “cosine rule”, are examined. A finite element model is established together with the three-dimensional wind field, in both frequency- and time-domain, and the structural response is analysed. The response is compared for: linear and non-linear coefficient formulations, different buffeting load formulations, different fitting methods, different coefficients considered and different quasi-steady motion-dependent force formulations. The case study demonstrates limitations of the 2D buffeting formulations, in particular of the “cosine rule”, and further supports the use of 3D buffeting formulations, the gathering of comprehensive data for skew wind loads and the use of constrained polynomials when fitting and extrapolating the data. This study also provides valuable insight on how to perform these analyses and overcome some of the practical challenges.