Abstract
Reliably predicting turbine/wake interactions in arrays of tidal stream turbines is paramount to reducing losses of energy yield by optimizing array layout at the design stage. This study focuses on the analysis of turbine performance and rotor wakes based on Navier-Stokes computational fluid dynamics, discussing modeling aspects associated with the method used for incorporating turbine hydrodynamics and turbulent flow effects in simulations. Important factors considered herein are a) the impact of laminar-to-turbulent transition of the blade boundary layers on rotor performance, and b) the sensitivity of the computed wake evolution on the turbine modeling approach. For these analyses, the results of rotor resolved and generalized actuator disk solutions are compared to flume tank measured data of a model turbine performance and wake. The overall agreement of all computed solutions and measured data is good. In the presented Reynoldsaveraged Navier-Stokes analyses, the rotor-resolved analysis predicts the measured rotor wake up to about four rotor diameters behind the turbine well, and seemingly better than the actuator disk model, as expected; thereafter, however, the rotor resolved simulation dissipates the wake more slowly than observed in the experiment, whereas the wake recovery rate predicted by the actuator disk analysis is closer to measured data. The cross-comparison of the geometry resolved analyses and measured data indicate that the blade boundary layers are likely to be fully turbulent, which has a significant detrimental impact on the turbine power. It is also found that the time-averaged solution of the time-dependent analysis and that of the steady flow analysis of the turbine resolved flow field, differ negligibly, as noted in other recent studies.