Modeling Multidimensional Multivariate Turbulent Wind Fields Using a Correlated Turbulence Wave Number-Frequency Spectral Representation Method

Accurate and efficient simulation of turbulent stochastic fields lays a solid foundation for dynamic response analysis and reliability evaluation of wind-sensitive structures. In this paper, a correlated-turbulence wave number-frequency spectral representation method (CT-WSRM) is proposed for simulating turbulent wind fields. Turbulent spectra that consider the correlation of turbulence are established using wind data measured during Typhoon Yanhua at the Ma'anshan Yangtze River (MYR) Bridge site in China. Using the established spectra, a customized turbulence wave number-frequency spectra density (WSD) matrix is defined and adopted in the proposed CT-WSRM. The proposed method can be utilized to simulate multidimensional multivariate two dimensional-three variate (2D-3V) spatial-temporal turbulent wind fields. In addition, a dimension-reduction model is introduced to describe turbulent wind fields in the probability density level within three random variables. The fast Fourier transform (FFT) algorithm is also embedded in the CT-WSRM to alleviate the computational burden. The stochastic turbulent wind fields for the MYR Bridge were simulated. Results demonstrated the effectiveness of the proposed method against the measured turbulent spectra. This method can be further utilized in the dynamic reliability analysis, providing structural reliability evaluation from the probabilistic view.

Automated summary

In this paper, a correlated-turbulence wave number-frequency spectral representation method (CT-WSRM) is proposed for simulating turbulent wind fields. Turbulent spectra that consider the correlation of turbulence are established using wind data measured during Typhoon Yanhua at the Ma'anshan Yangtze River (MYR) Bridge site in China. The proposed method can be utilized to simulate multidimensional multivariate two dimensional-three variate (2D-3V) spatial-temporal turbulent wind fields. This method can be further utilized in the dynamic reliability analysis, providing structural reliability evaluation from the probabilistic view.