Comparison of RANS Turbulence Models in Predicting Wake Development in a 2-Dimensional Actuator Disk Model

One of the most popular methodologies used to predict the wake of a tidal stream turbine (TST) is the RANS turbulence models coupled with the actuator disk method. This methodology has been widely adopted in the in the wind industry, since the mid-1990s, to predict wake development of wind turbines. Moreover, the reason for its popularity is its capability to give accurate results at an affordable computational cost, and the application of 2-dimensional actuator disk approach could further reduce the computational cost. In this paper, a number of RANS turbulence models represented by a porous disk were used to simulate the wake development behind a TST, the findings were compared. The models adopted in this work are the Standard k-ε model, the Standard k-ω model, the RNG (Re-Normalised Group) k-ε model, the SST (Shear Stress Transport) k-ω model and the RSM (Reynold Stress Model). The results are also validated against experimental measurements found in literature, with a key focus on comparing the downstream velocity and turbulence intensity. It has shown that the Standard k-ε model is best at predicting downstream wake velocities while the SST k-ω model is better at predicting downstream wake turbulence intensity. Mesh convergence studies were conducted to optimise the computational efficiency for each turbulence model used.