Development of an aircraft noise emission model accounting for flight parameters
Existing aircraft noise models are only partly suited for the assessment and optimization of noise abatement flight procedures. Either the models are too simplified, or are based on semi-empirical models which require ambitious flight parameters, or are not publicly available. In this thesis, an aircraft noise emission model for turbofan-powered aircraft was developed with the capability to take the flight configuration (thrust rating, airspeed, and aeroplane configuration) into account. Engine and airframe noise were modeled separately with a small number of model parameters. The approach is universal and thus applicable on many aircraft types.
To establish a data basis for different flight configurations under regular air traffic, extensive acoustical measurements around Zurich airport were realized. The acoustical data were processed with the help of flight data, flight paths, and meteorological data to obtain direction-dependent sound emission levels at the source. By means of this data base and multiple linear regression, three model variants with different level of detail were developed: an advanced model with three-dimensional directivity (3D) and two reduced models without aeroplane configuration. The reduced models, which can be applied if no flight data records are available, are modeled either with a three-dimensional directivity (3Dred) or a two-dimensional directivity (2Dred).
In total, 19 aircraft noise emission models for combinations of engine type and aircraft type were established. The main parameter for the engine noise model is the rotational speed of the engines, which also influences the directivity of the model. For airframe noise, the aircraft Mach number is the main parameter. However, also the landing gear is an important sound source of the airframe. If deployed early during approach and thus at high aircraft Mach numbers, it raises the total sound emission level up to 5 dB. Such local influence of the aeroplane configuration is represented by the 3D model.
A time-step method and a detailed propagation model were applied to simulate 10 524 flight events that were used for the development of the models. A comparison between 3D model and measurement resulted in a slight overestimation of the simulated sound exposure level of 0.1 dB with a standard deviation of 0.9 dB. In comparison, the simulated maximum sound pressure level was slightly underestimated and showed larger variations. The model variant 3Dred showed similar results, because the aeroplane configuration during approach is related to the aircraft Mach number. In contrast, the reduction to a two-dimensional directivity (2Dred) resulted in larger standard deviations.
Based on physical and empirical knowledge about sound generation, appropriate model parameters were chosen and linearized, whereby a physical behavior of the statistical model could be achieved. It was shown that the aircraft noise emission models are capable to reproduce the acoustical measurements with high accuracy. However, extrapolations to unknown flight configurations require validation with independent data in the future. The results of the thesis indicate the model’s great potential for the assessment and optimization of noise abatement flight procedures.