On the aeroacoustics of a propeller at low Reynolds number subjected to a grid-induced turbulent inflow

Abstract

This work investigates the effect of grid-generated turbulence ingestion on noise generation in a propeller operating at a low Reynolds number using high-fidelity, scale-resolved simulations. The numerical setup reproduces experiments carried out at Delft University of Technology, where inflow turbulence is generated by a grid placed within a duct. It is found that, upstream of the propeller, the longitudinal correlation length of the streamwise velocity component increases with respect to the case without the propeller. The opposite happens for the transversal one. The turbulent inflow impinging on the propeller blades does not alter the mean flow characteristics over the propeller blades, e.g., the mean static pressure coefficient. However, it increases the root mean square of the pressure fluctuations up to the turbulent reattachment point of the laminar separation bubble, while leaving the downstream region mostly unaffected. This causes a broadband increase in the radiated noise in the low-to-mid frequency range, as confirmed by applying Amiet’s noise-prediction model with input data sampled near the propeller blades’ leading edge. The far-field noise spectra are characterized not only by an increase in the broadband noise with respect to the clean inflow case, but also by tonal components at multiples of the blade-passing frequency. It is found that these tones are caused by the footprint of the turbulence grid that introduces flow inhomogeneities at the propeller location for this specific configuration. It is recommended, when performing experiments and simulations, to verify if any footprint of the turbulence grid is present, not only by performing single-point measurements but also by measuring the time-averaged flow field before installing the propeller.