Unfrozen Skewed Turbulence for Wind Loading on Structures

The paper introduces an algorithm to generate a three-variate four-dimensional wind turbulence field suited for yawed wind dynamic load simulation. At large yaw angles, a relaxation of Taylor's hypothesis of frozen turbulence becomes relevant as well as the flow phase lag in the along-wind direction, which modulates the real and imaginary parts of the coherence. To capture such a general wind action on a structure, a modified spectral representation method is used where the coherence of turbulence is described as a complex-valued function. The one-point and two-point co-spectra are implemented in the simulation setup using a square-root-free Cholesky decomposition of the spectral matrix. The numerical procedure is illustrated based on turbulence characteristics derived from data collected during storm Aina (2017) on the Norwegian coast by three-dimensional sonic anemometers. During this event, a remarkable 3-hour stationary time series with a mean wind speed of 24 m s(-1) at a height of 49 m above ground was recorded. Since no computational grid is needed, the velocity fluctuations with representative spatio-temporal characteristics can be directly simulated on structural elements of slender structures. Such an algorithm may be essential for the design of super-long span bridges in coastal areas.