Accuracy of noise prediction of propellers via numerical simulations, analytical methods, and experimental campaigns

Abstract

This work focuses on the assessment of the accuracy of numerical prediction and experimental campaigns on providing the noise emissions of an isolated benchmark propeller. An experimental campaign is carried out with a model low-Reynolds propeller of 0.3 m diameter operating at high RPM, equivalent of a tip-Mach number (M
₁) of 0.37 and an advance ratio (J) of 0.4.Measurements are conducted on an open-test section wind tunnel, surrounded by an anechoic chamber. Simulations are carried out with the commercial software PowerFLOW and aim at reproducing the propeller geometry and conditions. BEMT-based noise estimations are also used to demonstrate the expected results. The discussion is focused on the uncertainties of the experimental campaign, and the current accuracy of numerical and analytical predictions, creating a complete picture of the discrepancies expected when predicting propeller noise levels and potential sources of errors. Results point to an accurate ability of the three methodologies to assess the overall noise emissions. Nevertheless, precise description and measurements of the higher harmonics of the tonal emissions and of the broadband noise levels is still lacking and require improvements in experimental conditions and a detailed assessment of the flow over the propeller.