Capturing the Dynamic Stall in H-Type Darrieus Wind Turbines Using Different URANS Turbulence Models

Abstract

The occurrence of dynamic stall phenomenon in an H-type Darrieus wind turbine with low tip speed ratio (TSR) has been numerically investigated on a single-bladed rotor with NACA 0012 airfoil. The Reynolds number (Re) ∼105 at TSR = 2 implicates complex turbulence environment around the blades of the turbine modeling which still remains a challenging problem. Thus, with a motivation to find out a suitable turbulence model to capture the dynamic stall, a comparative study is carried out between three unsteady Reynolds-averaged Navier–Stokes (URANS) models: Spalart–Allmaras (S-A), shear stress transport (SST) k–ω, and transition SST (TSST). It was found that the TSST model predicted the dynamic stall phenomenon the earliest, whereas, the S-A model predicted it the latest. The transitional phenomenon like formation and bursting of the laminar separation bubble (LSB) was best predicted by the TSST model. However, the TSST overpredicts the turbulent boundary layer (BL) roll up from the trailing edge (TE) toward the leading edge (LE). The percentage difference in the power coefficient (Cp) values with respect to the TSST accounted to 16.67% and 60% higher for SST k–ω and S-A models, respectively. The S-A model delays the torque coefficient (Ct) peak prediction by 5 deg and 11 deg azimuthal angle compared with SST k–ω and TSST models, respectively. Overall, it was found that the transitional aspect in TSST model is important in predicting the light stall regime; however, in the deep stall regime SST k–ω model performed well too.