A CFD code comparison of wind turbine wakes

The Science of Making Torque from Wind 2014

Van Der Laan, M.P.^1 ; Storey, R.C.^2 ; Sørensen, N.N.^1 ; Norris, S.E.^3 ; Cater, J.E.^2

Technical University of Denmark, DTU Wind Energy, Risø Campus, Frederiksborgvej 399, Roskilde 4000, Denmark^1

University of Auckland, Department of Engineering Science, 70 Symonds St Grafton, Auckland 1010, New Zealand^2

University of Auckland, Department of Mechanical Engineering, 20 Symonds St Grafton, Auckland 1010, New Zealand^3

关键词: Eddy viscosity turbulence model;    Field measurement;    Grid resolution;    K epsilon models;    Reynolds stress;    Reynolds-averaged navier-stokes simulations;    Velocity deficits;    Wind turbine wakes;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/524/1/012140/pdf
DOI  :  10.1088/1742-6596/524/1/012140

来源: IOP

【 摘 要 】

A comparison is made between the EllipSys3D and SnS CFD codes. Both codes are used to perform Large-Eddy Simulations (LES) of single wind turbine wakes, using the actuator disk method. The comparison shows that both LES models predict similar velocity deficits and stream-wise Reynolds-stresses for four test cases. A grid resolution study, performed in EllipSys3D and SnS, shows that a minimal uniform cell spacing of 1/30 of the rotor diameter is necessary to resolve the wind turbine wake. In addition, the LES-predicted velocity deficits are also compared with Reynolds-Averaged Navier Stokes simulations using EllipSys3D for a test case that is based on field measurements. In these simulations, two eddy viscosity turbulence models are employed: the k-Ε model and the k-Ε-fpmodel. Where the k-Ε model fails to predict the velocity deficit, the results of the k-Ε-fPmodel show good agreement with both LES models and measurements.