Validation and comparison of aerodynamic modelling approaches for wind turbines

Frédéric Blondel (1) , R. Boisard (2) , Malika Milekovic (1) , Gilles Ferrer (1) , C. Lienard (2) , David Teixeira (1)

1 IFPEN - IFP Energies nouvelles
2 ONERA - The French Aerospace Lab [Meudon]

The development of large capacity Floating Offshore Wind Turbines (FOWT) is an interdisciplinary challenge for the design solvers, requiring accurate modelling of both hydrodynamics, elasticity, servodynamics and aerodynamics all together. Floating platforms will induce low-frequency unsteadiness, and for large capacity turbines, the blade induced vibrations will lead to high-frequency unsteadiness. While yawed inflow conditions are still a challenge for commonly used aerodynamic methods such as the Blade Element Momentum method (BEM), the new sources of unsteadiness involved by large turbine scales and floater motions have to be tackled accurately, keeping the computational cost small enough to be compatible with design and certification purposes. In the light of this, this paper will focus on the comparison of three aerodynamic solvers based on BEM and vortex methods, on standard, yawed and unsteady inflow conditions. We will focus here on up-to-date wind tunnel experiments, such as the Unsteady Aerodynamics Experiment (UAE) database and the MexNext international project. 1. Introduction Nowadays, most of the design software rely on the Blade Element Momentum models [10]. Those methods are based on strong hypotheses, partially overcame through analytical corrections, that introduce empiricism in the models. The reliability of the model is then questionable, especially in the framework of large fixed and floating offshore wind turbines, which tend to become larger and larger, reaching 10M W or more [2]. On the other side, more physical models, such as the vortex methods, that require less analytical corrections, are being increasingly used. The development of algorithms and new computational architectures have strongly reduced the computational time of vortex methods, which is crucial in the light of the large databases that have to be tackled for certification purposes. Vortex methods provide different solutions to model the wake, such as vortex points, filaments or panels. The aim of this paper is to compare a BEM solver and two vortex solvers designed to be coupled with the aero-servo-hydro-elastic solver DeepLines Wind TM [8] against up-to-date wind tunnels experiments, identify the cases on which BEM models are not accurate enough, and evaluate the potential of vortex models on those cases. The two lifting-line solvers are based on different wake models, a free-wake vortex filaments model and a free-wake vortex panel wake model. The impact of the wake model on the lifting-line results will be analysed.

Domaines: Mécanique des fluides [physics.class-ph]

Frédéric Blondel, R. Boisard, Malika Milekovic, Gilles Ferrer, C. Lienard, et al.. Validation and comparison of aerodynamic modelling approaches for wind turbines. TORQUE Conference 2016, Oct 2016, Munich, Germany. ⟨10.1088/1742-6596/753/2/022029⟩. ⟨hal-01663647⟩