Large-Eddy Simulations of realistic atmospheric turbulence with the DLR-TAU-code initialized by in situ airborne measurements

In this paper the numerical simulation of a turbulent flow in the atmospheric boundary layer (ABL) with a Large-Eddy Simulation (LES)-model is discussed. The results of this work are intended to be used for the numerical simulation of turbulent flows around an airfoil. To simulate the characteristics of the ABL flow and its influence on the airfoil realistically the flow upstream of the airfoil has to be turbulent with statistical properties that are comparable to those found in atmospheric measurements. To achieve this goal, a method to generate synthetic turbulent wind fields was used to initialize an LES model which is able to simulate the turbulent flow around an airfoil.For the initial turbulent wind field to contain realistic statistics of atmospheric turbulence, data taken with the Helipod system are used. The Helipod is a helicopter-borne measurement probe that is able to take high-resolution measurements of temperature, wind vector and humidity. The statistical properties that are used as input parameters for the turbulence generator are the spectral energy, the correlation matrix and the variances of the three components of the wind vector.The LES model used in this project is the flow solver TAU developed by the German Aerospace Center (DLR). TAU is a compressible computational fluid dynamics (CFDs) tool that is able to compute the flow around obstacles (e.g. parts of aircrafts or even whole aircrafts) on an unstructured grid. Calculations with TAU can be performed in Reynolds-Averaged Navier–Stokes-, LES- or Detached Eddy Simulation-mode using different sub-grid scale models.

► Generation of synthetic turbulence with given statistics.
► Synthetic turbulence based on measurements in the atmospheric boundary layer.
► Initialization of compressible flow model with synthetic atmospheric turbulent wind fields.
► Simulation of the decay of the turbulent flow.
► Results show proper behavior of the statistics for future simulations.