Turbulence Measurements in Trailing Vortices for B.W.I. Noise Prediction
Devenport, William J.
Glegg, Stewart A. L.
MetadataShow full item record
Blade wake interaction (BWI) noise is the broadband noise generated by the ingestion of turbulent tip vortices into helicopter rotors. Prediction of BWI noise requires knowledge of the turbulence structure of the tip vortex. This report describes a joint investigation to measure that structure and incorporate the results into a noise prediction scheme. Measurements were performed on the tip vortex shed by a single rectangular NACA 0012 half wing placed in a wind tunnel test section. The properties of the vortex were studied for a range of angles of attack (2.5° to 7.5°) and chord Reynolds numbers (130000 to 530000). Initially the vortex was examined for stability and probe interference effects through flow visualization. Then detailed three-component velocity, turbulence and spectral measurements were made using multiple hot wires between 20 and 30 chordlengths downstream of the wing. These measurements show the flow to consist of a small axisymmetric core surrounded by a large non-axisymmetric region in which the wing wake forms a spiral around the core. In contrast to the results of previous workers, most of whom studied vortices generated by split wing configurations, there appeared to be little axisymmetric region of merged turbulent flow. Turbulence levels in the spiral wake decay with downstream distance. They also fall as the core is approached, presumably because of the effects of the increased straining and curvature on the turbulent structures here. Turbulence spectra measured in the wake are very similar in form, regardless of conditions and exact location, and bear a strong resemblance to a von Karman spectrum. At most conditions true turbulence levels in and immediately adjacent to the core are very low. Velocity fluctuations, however, are intense as a consequence of coherent lateral motions of the core and possible wave motions and instabilities travelling along it. Velocity autospectra in the core show the lateral motions to be anisotropic at very low frequencies and isotropic and mid frequencies. A large part of these motions may well be self induced. Circulation profiles in this region show Hoffman and Joubert's semi-logarithmic region, and in one case reveal the core to be fully developed. From the point of view of BWI noise prediction the flow measurements identify three sources of velocity fluctuations; low-frequency anisotropic core motions, mid-frequency isotropic core motions, and turbulence in the surrounding spiral wake. Estimates of the noise produced by these different frequency regimes show that it is the mid-frequency isotropic motions which is the most important mechanism for noise production, but the spectral shape and the predicted directionality for this mechanism are not in agreement with the measurements of BWI noise. The original noise-prediction scheme, based on the concept of a more turbulent vortex, has been shown to be the wrong basis for the correct description of the flow, but in spite of this the results are significantly better than those presented here. This suggests that the flow measured in the wind tunnel does not have the same turbulence upwash spectrum as that encountered by the helicopter rotor. One of the features of a real helicopter rotor which was not considered is the effect of a downstream blade upon the tip vortex. This may cause vortex bursting or alter the flow structure in the rotor disc plane in other ways. Measurements to evaluate this concept are planned for the future.