Large Eddy Simulation of the meandering of a wind turbine wake with stochastically generated boundary conditions

Wind turbine wakes are known to be affected by the large atmospheric turbulent scales, which can cause trajectory variations within a wide frequency band. This phenomenon, called meandering, is suspected to be a cause of premature wear on turbines located inside wind farms. This work proposes a method to generate and apply synthetic turbulent velocity series as boundary conditions in a Large Eddy Simulation of an actuator disk in a flow with realistic turbulence characteristics. The stochastic generation method relies on the inverse Short-Time Fourier Transform (STFT) of a random vector field correlated in Fourier space according to the covariance tensor calculated from the homogeneous isotropic spectral tensor. In contrast with a single Fourier transform, the STFT allows the generation of arbitrarily large velocity fields. The generated series are used as boundary values on the inlet as well as on the lateral boundaries of the domain. This allows for sustained turbulent forcing on the whole length of the domain which is especially useful for a small computational domain relative to the size of the dominant turbulent scales.