Aerofoil dipole noise due to flow separation and stall at a low Reynolds number

Abstract

Aerofoil self-noise produced by flow separation and stall is relatively little understood regarding the underlying generation mechanisms. The focus of this work is to provide an improved level of understanding particularly with regard to the dipole noise sources utilising a high-fidelity direct numerical simulation. A NACA0012 aerofoil is considered under three different flow conditions at a Reynolds number Re∞=50,000 and a Mach number M∞=0.4. These include: a pre-stall condition with a laminar separation bubble (α=5∘), near-stall (α=10∘), and fully stalled (α=15∘). The noise radiation in the far-field is significantly increased at low frequencies for the full-stall case for all observer directions which is consistent with previous experimental observations. The dominant source regions for each configuration are identified for low, medium and high frequencies, separately. A number of key findings are made concerning the source characteristics in full-stall which differ considerably from the lower angle of attack cases. It is found that the location of the dominant sources changes more significantly with frequency for the full-stall case. Additionally, for medium to high frequencies the maximum acoustic source amplitude is weaker for the full-stall case, despite comparable levels observed in the far-field. This seemingly contradictory observation highlights the importance of phase variations in the wall pressure fluctuations. For the frequencies considered in this paper it is shown that the full-stall case usually produces a relatively more in-phase source distribution, resulting in a more efficient radiation despite the lower amplitude levels. The important flow structures which are responsible for the dipole sources are also identified through analysis of the pressure field at isolated frequencies. It is found that for low frequencies coherent structures in the shear layer are responsible for the scattering of the wall pressure fluctuations at the TE, which agrees with previous findings in the literature. However, at medium and high frequencies the shear layer structures are found to be relatively weak in the proximity of the TE. This indicates that the noise may be generated through other means, for example scattering of fluctuating pressure induced by vortices shed from the TE.