Optimization of vertical-axis turbine using numerical simulations

Rémi Gosselin
Doctorate student
remi.gosselin.2@ulaval.ca

Nowadays, concerns about the environment push the development of sustainable energy sources, among which wind turbines, and more recently tidal or current turbines. Based on the extraction of power from the wind, tidal or river currents, the potential of these technologies is important not only in Quebec, but also in most of the globe’s coastal regions.

During the development of the first modern wind turbines in the 20s, two types of turbines existed : those with an horizontal axis, the propeller type, and those with a vertical axis, among which the Darrieus turbines that have a particular « egg-beater » shape. Each technology as its advantages and drawbacks, and the simpler, steady, aerodynamic of the propeller turbines, also developped for aircrafts at the time, helped them dominate the wind turbine market and take all the research and development focus from the vertical axis turbines. About fifty years later, during the oil crisis in the 70s, interest for the vertical axis turbines growed as an alternative for the horizontal axis turbines that were starting to show their limit in terms of blade complexity and size. Numerous experimental studies were conducted and still are an important source of data today. These experiments showed the potential of vertical axis turbines in terms of efficiency, but couldn’t solve a major drawback of the technology : vibrations induced by the torque variations on the blade, that shorten dramatically the life of the turbine.

Nowadays, computationnal power is sufficient to compute such unsteady flows in a short time, at least in two dimensions, when using uRANS turbulence modelisation. This power helps develop a better understanding of the aerodynamics of a Darrieus turbine, and make better prediction of a real turbine efficiency, without having to undergo some costly experimental studies. It also allows the development of various solutions to help lower torque ripple and increase the turbine efficiency. Solutions like individual blade pitch control are studied at LMFN to help develop this technology.

The 2D parametric study (turbine solidity, number of blades, etc…) is also completed by a few 3D calculations in order to evaluate the loss in efficiency induced by the finite length of the blades (compared to « infinite » blade in 2D calculations), with or without endplates. However, these calculations remain costly in terms of computational power and do not permit the evaluation of each turbine parameter.