A Bouhelal - Mechanical Engineering and Development Laboratory, École Nationale Polytechnique, B.P. ۱۸۲, El-Harrach, Alger, ۱۶۲۰۰, Algeria
A Smaili - Mechanical Engineering and Development Laboratory, École Nationale Polytechnique, B.P. ۱۸۲, El- Harrach, Alger, ۱۶۲۰۰, Algeria
O Guerri - Centre de Développement des Énergies Renouvelables, B.P. ۶۲, Route de l’Observatoire, Bouzaréah, Alger, Algeria
C Masson - Department of Mechanical Engineering, École de Technologie Supérieure, ۱۱۰۰ Notre-Dame Ouest, H۳C۱K۳, Montréal, Québec, Canada

The effects of different Reynolds Averaged Navier Stokes (RANS) turbulence models on two near-wall approaches using high and low Reynolds models on predicting performance of horizontal axis wind turbines (HAWTs) were studied for a range of wind conditions where flow over the rotor varied from fully attached to massively separated flow. This paper s main contribution is in establishing which RANS models can produce quantitatively reliable numerical predictions of turbulent flow around wind turbine rotors. The authors used measurements done by the new MEXICO (Model rotor EXperiments In COntrolled conditions) project in the German Dutch wind tunnels (DNW) in order to validate and test CFD (Computational Fluid Dynamic) codes. Four different RANS turbulence models were considered: Spalart-Allmaras; k − ε (RNG); k − ω SST; and the transition γ-Reθ model. At low wind speeds, it was found that all four models were good predictors of aerodynamic performance, and at high wind speeds, where the swirl effect was modeled using wall function corrections in both equations, the k − ε model was considered to be the best model: it was the most accurate within a reasonable computational time.

Aerodynamic analysis; HAWT; CFD; Turbulence modeling; RANS; low and high Reynolds Models; Near-wall treatment; New MEXICO measurement