Cross-correlation-based approach to align turbulent inflow between CFD and lower-fidelity-codes in wind turbine simulations

Abstract

In cases with turbulent inflow, the analysis of the temporal development of forces could provide more insight into differences in calculated fatigue loads of wind turbines. While lifting-line-codes and the Blade-Element-Momentum (BEM) approach do not resolve the inflow, Computational Fluid-Dynamic (CFD) codes resolve the entire flow field. A comparison of time series between the different codes therefore requires a consistent input of the background turbulence. This was enabled by extracting the turbulent velocity field from empty box (without rotor) CFD simulations at the anticipated rotor position. The presence of a rotor in the CFD simulations leads to a delay in the inflow due to the induced velocity. This blockage effect of the rotor was quantified by a cross-correlation. The velocity field extracted from the simulation of the empty box was shifted by the resulting temporal offset before it was used as lifting-line-code or BEM input. Thus, a time-dependent load comparison between the codes could be performed. It was found that the difference in load predictions between CFD and BEM seems to be larger at peak values. For cases with high thrust coefficient or high turbulence intensity, a simpler analytical approach resulted in significantly higher temporal offsets than the cross-correlation.