Wind Speed Field Simulation via Stochastic Harmonic Function Representation Based on Wavenumber-Frequency Spectrum

Simulation of fluctuating wind speed field is of paramount significance in the design of large flexible structures. To circumvent the difficulty due to the decomposition of cross power spectral density (PSD) matrix and the interpolation between discretized spatial points, a wavenumber-frequency joint spectrum-based spectral representation method (SRM) has been developed recently. To further improve the efficiency and accuracy, the stochastic harmonic function (SHF) representation is extended in the present paper for the simulation of stationary and nonstationary fluctuating wind fields in two spatial dimensions. In contrast to the SRM, in addition to the phase angles, the frequencies and wavenumbers are also random variables over partitioned wavenumber-frequency subdomains. Furthermore, a strategy of dependent random frequencies and wavenumbers based on the SHF is proposed so that the number of random variables can be considerably reduced by 3/7. A new acceptance-rejection criterion, which avoids the artificial intervene, is suggested based on the p-power joint spectrum, and the subdomains are correspondingly determined by the Voronoi cell partitioning. For illustrative purposes, two numerical examples for the simulation of stationary and nonstationary fluctuating wind speed fields in two spatial dimensions are addressed, demonstrating the effectiveness of the proposed method in considerably reducing the random variables as well as the computational efforts.