Flow Simulation over the isolated NREL 5-MW rotor with the DLR -THETA and -TAU codes: A comparative assessment

Abstract

The flow simulation over the isolated NREL 5-MW rotor was performed with the DLR THETA code for incompressible flows (v.10.1) and the DLR TAU code for compressible flows (v.2015.1.o). A total of eight main steady simulations were run considering wind speeds between the cut-in and rated rotor speeds. After the analysis of the results from these simulations, the THETA code showed better performance than the TAU code. Its torque predictions were very close to the NREL reference values (-0.65% difference in average against 5.31% for TAU) and they showed convergence 7.5 times faster than the TAU ones. However, important oscillations in the monitoring of torque were detected and analyzed in the THETA simulations. Furthermore, a total of two unsteady simulations were run considering only rated speed. The predictions of torque from these unsteady simulations did not converge after the defined five rotor revolutions. The turbulence models used for the steady and unsteady simulations were the SST k-w for THETA and the Wilcox k-w for TAU. An additional number of six steady simulations were run considering mainly different turbulence models, from which the Spalart-Allmaras model proved to be very competent in both codes.