Wind Tunnel Investigation on the Vortex Shedding of Cable-Stayed Bridges

• Main Authors: Bo-Wen Yeh, 葉博文
• Other Authors: Chii-Ming Cheng
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/00995710069532762159

碩士 === 淡江大學 === 土木工程學系 === 88 === Construction of flexible long-span bridges (suspended bridges or cable-stayed bridges) has resulted in increase of aeroelastic effects. Vortex shedding is a phenomenon that is caused by flow separation around bluff bodies. These vortices subject an elastic or elastically mounted bluff body to a periodic excitation, which may cause it to vibrate. This oscillation is small except in a range of shedding frequencies bracketing the natural frequency. Large-amplitude oscillations occur in this range that appear to control the shedding process in a fluid-structure interaction phenomenon, and shedding frequencies in this range do not vary, known as “lock-in”. In this way, although such vibrations may not be catastrophic, sustained oscillations at relatively low cross-wind velocity associated with vortex-shedding may cause fatigue to the structure. Scanlan(1981) proposed a semi-empirical nonlinear model for the across-flow, vortex-induced response of a bluff body. Experiments were conducted at T.K.U. Wind Tunnel. Single-degree-of-freedom section models considering lock-in were tested condition in the smooth flow. The first set of experiments were conduced under fixed wind speed, varied mechanical damping, and a second set of tests was performed at a fixed value of the mechanical damping, while the wind speed being varied through the lock-in range. Following Van der Pol nonlinear model at lock-in adapted the section model tests. Ehsan and Scanlan(1990) were not generally applicable to all of the different experimental situations that arise in a wind-tunnel test. The methods were less accurate for situations where observations of the experiment are noisy, which is possible under turbulent flow conditions. Also, the existing methods can not be used in experimental situations where the steady-state amplitude is very small and, therefore, can not be observed in an experiment with sufficient accuracy. Hence, we used identification of vortex-induced-response parameters in time domain to acquire more accurate aerodynamic derivatives. Model parameters were used in the equation of motion of the prototype bridge to obtain its dynamic response. By the way, analytical predictions compare with advanced literature in the paper.