Full-Bridge Aeroelastic Model Tests in Natural Wind for Buffeting Response Study of Long-Span Bridges

Recently, a new experimental method was proposed for wind engineering research of bridges, i.e. studying the wind-resistant performance of bridges based on larger-scale full bridge aeroelastic models in the natural wind. An Outdoor Experimental Base for Bridge Wind Engineering (OEBBWE) has been built, and several full-bridge aeroelastic models have been constructed to test larger-scale models in more realistic turbulence conditions. This study attempts to address two major questions concerning the use of this method for buffeting investigation of long-span bridges: (1) determining if the wind field characteristics at the Outdoor Experimental Base are suitable for a buffeting investigation, and (2) evaluating the reliability and usefulness of the data (including vibration responses and cable forces) measured on a large-scale aeroelastic model in the natural wind. To that aim, a 280-day wind field monitoring campaign at the OEBBWE and a 10-day buffeting monitoring campaign on a 1:50 scaled aeroelastic cable-stayed bridge model were carried out. Experimental results show that the wind conditions were suitable for studying the buffeting performance of bridges. The wind field at the site showed reasonable resemblance to that of the full-scale conditions while providing a large variety of different turbulence conditions due to the inherent variability in the natural wind conditions. The dynamic responses of the bridge were also studied under natural wind. Primary contributors to the buffeting responses of the bridge model were the first few vertical, torsional, and lateral modes, while contributions from higher-order modes were found insignificant. The root mean square (RMS) acceleration of the bridge model correlates strongly with mean wind speed, while strongly influenced by the turbulence intensity and wind yaw angle. These findings suggest that using aeroelastic scale models in the natural wind is a suitable method for buffeting investigations and that the data measured using the proposed new experimental method were reliable and useful.