Comparison of Semi-Empirical Noise Models with Flyover Measurements of Operating Aircraft

The contribution of the engine and the airframe to the total noise generated by an aircraft varies with the operating conditions. Semi-empirical models are able to account for such variations but require detailed engine and airframe data as input that is not readily available for most aircraft types and operations. This hinders the validation of these models through comparison between predictions and experimental data. This work investigates the sensitivity of semi-empirical models of engine and airframe noise to slight variations of the input data, representative of uncertainties in geometrical parameters and variability of the aircraft operating conditions during flyovers. In addition, the predictions are compared to measurements of A320, A330, and B777 landings and departures. This, together with the sensitivity analysis, indicates frequency regions where a mismatch between measurements and predictions exists. The deviation between predictions and measurements for landings can be partially explained by the underestimation of the sound pressure level of the higher harmonics of the fan. For takeoff, the models predict lower levels than measured. This is hypothesized to be associated with jet-installation noise, which is not accounted for in the semi-empirical models. The predicted spectra of the Airbus A320 and A330 were adjusted to account for jet installation noise, using levels available in the literature. This resulted in a better agreement between modeled and measured spectra at low frequencies.