TY - RPRT
T1 - CFD computations of wind turbine blade loads during standstill operation KNOW-BLADE, Task 3.1 report
AU - Sørensen, Niels N.
AU - Johansen, Jeppe
AU - Conway, S.
PY - 2004
Y1 - 2004
N2 - Two rotors blades are computed during standstill conditions, using two different Navier-Stokes solvers EDGE and EllipSys3D. Both steady and transient linear k-? RANS turbulence models are applied, along with steady non-linear RANS and transient DESsimulations. The STORK 5.0 WPX blade is computed a three different tip pitch angles, 0, 26 and 50 degrees tip pitch angle, while the NREL Phase-VI blade is computed at 90 degrees tip pitch angle. Generally the CFD codes reproduce the measured trends quitewell and the two involved CFD codes give very similar results. The discrepancies observed can be explained by the difference in the applied turbulence models and the fact that the results from one of the solvers are presented as instantaneous valuesinstead of averaged values. The comparison of steady and transient RANS results show that the gain of using time true computations are very limited for this case, with respect to mean quantities. The same can be said for the RANS/DES comparison performedfor the NREL rotor, even though the DES computation shows improved agreement at the tip and root sections. Finally, it is shown that the DES methodology provides a much more physical representation of the heavily stalled part of the flow over blades athigh angles of attack.
AB - Two rotors blades are computed during standstill conditions, using two different Navier-Stokes solvers EDGE and EllipSys3D. Both steady and transient linear k-? RANS turbulence models are applied, along with steady non-linear RANS and transient DESsimulations. The STORK 5.0 WPX blade is computed a three different tip pitch angles, 0, 26 and 50 degrees tip pitch angle, while the NREL Phase-VI blade is computed at 90 degrees tip pitch angle. Generally the CFD codes reproduce the measured trends quitewell and the two involved CFD codes give very similar results. The discrepancies observed can be explained by the difference in the applied turbulence models and the fact that the results from one of the solvers are presented as instantaneous valuesinstead of averaged values. The comparison of steady and transient RANS results show that the gain of using time true computations are very limited for this case, with respect to mean quantities. The same can be said for the RANS/DES comparison performedfor the NREL rotor, even though the DES computation shows improved agreement at the tip and root sections. Finally, it is shown that the DES methodology provides a much more physical representation of the heavily stalled part of the flow over blades athigh angles of attack.
KW - 1-E vind
KW - Risø-R-1465
KW - Risø-R-1465(EN)
M3 - Report
SN - 87-550-3343-1
T3 - Denmark. Forskningscenter Risoe. Risoe-R
BT - CFD computations of wind turbine blade loads during standstill operation KNOW-BLADE, Task 3.1 report
ER -