Simplifying the simulation of stochastic wind velocity fields for long cable-stayed bridges

The spectral representation method has been greatly developed as the performance of computer was enhanced. For the time-domain analysis of long cable-stayed bridge, it is required to simulate wind velocity histories at many points arranged on the overall bridge. However, the number of these points is usually so large that the spectral representation method obviously becomes computationally prohibitive. In this paper, a practical method is introduced to simplify stochastic wind velocity field for long cable-stayed bridges. According to the properties of both the structural type and the vibration mode of long cable-stayed bridge, and considering the correlation characteristics of natural wind, an overall three-dimensional (i.e. three components of natural wind) wind velocity field of a long cable-stayed bridge can be simplified into many one-dimensional (i.e. just one component) wind velocity fields, respectively, along pylons and deck. The wind fluctuation along each pylon can be simulated independently by omitting the wind velocity correlation between one pylon and another. Using the explicitly expressed Choleskyʹs decomposition of the cross-spectral density matrix, the wind velocity field along the deck can be simulated relatively fast. The FFT (Fast Fourier Transform) technique is adopted to further improve the computational efficiency. As an example, the stochastic wind field of a three-pylon cable-stayed bridge is simulated. The spectrum and correlation characteristics of the simulated samples are checked, which demonstrates that the simulation results agree with the corresponding targets.