Aerodynamic behaviour of bridges.

摘要

For a number of years, under various contracts, the Department of Aeronautics and Fluid Mechanics has been wind-tunnel testing bridge models for static loads. A recent development has been to include dynamic testing of models to determine the stability of the bridge in winds. The interest of the writer was in applying aeroelstic techniques to the prediction of the stability of the bridge models. Tests on section models of a proposed road bridge were carried out in the low speed wind-tunnel of the Aeronautics Department. The unusual feature of the bridge under consideration was its composite arch ribs. In the classic suspension bridges, or cable-stayed bridges, the deck is suspended from cables and the stability resolved using the deck alone in the tests. In the case of the proposed bridge, the deck and the supporting arch rib would interact, and each would contribute to the dynamic behaviour of the bridge as a whole. However, because of the differing modes of motion of the parts it was thought that the aerodynamic stability of the complete structure could be determined from tests of section models of each part. Interaction between the parts would tend to reduce motion and increase stability. The size of the wind-tunnel working section usually prohibits testing of complete models at an acceptable scale. The radius of curvature of the arch rib was such that straight sections could be used for the model with very small errors. The separate section models were tested on the three-component balance to determine the steady wind forces on the bridge, which were also compared with predictions using British Standards data, and then on a dynamic mounting to examine their aerodynamic stability. Both the arch rib and the deck had a low speed resonant vibration caused by the natural frequency of the structure matching that of the shedding of vortex pairs from the top and bottom surfaces. The amplitudes of vibration of both were greatly reduced by cutting holes in the webs of the spanwise girders of the deck, and in the side plates of the arch ribs. These holes bled air from the leading edges, and reduced the strength of the vortices. The deck had a divergent pitch oscillation at high speeds, induced by a vortex phenomenon. The speed at which this occurred was increased by about 30% by adding a trapezoidal fairing to the edge of the roadway parapet. This reduced the strength of the upper surface vortex by smoothing the airflow. The low speed instability can be predicted using the Strouhal number for the structure, and amplitudes of vibration can be estimated for a number of damping levels. The pitch instability could only be determined experimentally, and as it will lead to catestrophic failure of the structure, it is essential that the critical speed is well above that likely to be experienced by the prototype. Detail changes have a very important effect on this motion and extrapolation to the full-size prototype must be done with great care. Much more dynamic experimental data are needed from full-size prototypes to allow more confident predictions to be made from model testing.