An Assessment of Current Fan Noise Prediction Capability

In this paper the results of an extensive assessment exercise carried out to establish the current state of the art for predicting fan noise at NASA are presented. Representative codes in the empirical, analytical and computational categories were exercised and assessed against a set of benchmark acoustic data obtained from wind tunnel tests of three model scale fans. The chosen codes were ANOPP representing an empirical capability, RSI representing an analytical capability, and LINFLUX representing a computational aeroacoustics capability. The selected benchmark fans cover a wide range of fan pressure ratios and fan tip speeds, and are representative of modern turbofan engine designs. The assessment results indicate that the ANOPP code can predict fan noise spectrum to within 4 dB of the measurement uncertainty band on a third-octave basis for the low and moderate tip speed fans except at extreme aft emission angles. The RSI code can predict fan broadband noise spectrum to within 1.5 dB of experimental uncertainty band provided the rotor-only contribution is taken into account. The LINFLUX code can predict interaction tone power levels to within experimental uncertainties at low and moderate fan tip speeds, but could deviate by as much as 6.5 dB outside the experimental uncertainty band at the highest tip speeds in some case.