DES simulations of the turbulent wake behind a vertical-axis turbine

Among renewable energies, the tidal energy sector is growing rapidly nowadays. The research group at the Laboratoire de Mécanique des Fluides Numérique (LMFN) from Laval University actively contributes to the development of this technology and has been involved in it for several years,  particularly  regarding the development of new concepts of hydro-kinetic turbines such as one using oscillating hydrofoils named Hydrolienne à Aile Oscillantes (HAO). Other technologies are also considered for tidal applications such as the  well-known horizontal axis turbines (HAH ) but also those with vertical axis (HAV) . It is expected that these turbines be used in clusters, similarly to wind farms, in sites where tidal currents are strong. In order to maximize the energy production of a given site, it is ideal to place the turbines as closely as possible to one another. Thus, the choice of a particular technology does not depend solely on its own isolated efficiency, but the interactions between the turbines themselves must also be taken into account. Indeed, the proximity of the turbines can create a blocking effect and the wake of a turbine can greatly influence the performance of another turbine downstream. Therefore, it is very important to characterize the different types of turbine in regard to the structure of their wake, the turbulence they generate as well as  the more or less rapid dissipation of their speed deficit.

The wake study requires unsteady 3D numerical simulations using advanced techniques of turbulence  modeling. One possible approach is called DES (Detached Eddy Simulation). This  is defined as a hybrid model given that it combines different modeling techniques. A relatively simple turbulence model (RANS approach) is used near the walls of a body and more advanced models (LES approach) are used away from the walls in order to  be able to capture the finer details of the flow in areas of interest such as wake. In addition, the reproduction of the movement of the blades of each type of turbine in the numerical simulations is a challenge. The use of specialized techniques such as non-conformal interfaces in movement makes it possible to maintain a good quality of mesh in strategic locations of the flow.

These numerical approaches are implemented in commercial solvers as well as open-source software, all used at the LMFN. Given the complexity of the calculations needed for this study, these simulations are carried out on the infrastructure of high performance computing of Compute Canada , including supercomputers Colosse and Guillimin from CLUMEQ.