Research On Aircraft Engine Fan Noise Separation Algorithms

During the research and design of a new engine, it is significant to study on the noise contribution of each component in order to lower the integrated engine noise.This paper deals with engine fan noise, which is one of the main noise sources of engine, and particularly study the algorithms to separate the engine fan noise from the static total noise. Through the discussion of the current research situation in the domestic and international, the most popular separation algorithms are studied. With the present forms of engine noise data, the beamforming algorithm and DAMAS,which are developed with the prevailing use of phased array of microphones for noise measurement, are chosen to proceed this paper's research.By the analysis of static total engine noise data provided by GE with the engine type CF34-10 A, the cross-spectrum method of beamforming is used to programming with MATLAB at the beginning, and the preliminary identification of fan noise is achieved. Through the comparison of the result with fan noise characteristics, the separation result is not acceptable for its inaccuracy and poor resolution.As a result, the DAMAS is adopted as a final feasible algorithm based on beamforming. Use the beamforming results as the input of DAMAS, by inverse computing and resolve the matrix equation with Gauss iteration, the new separation result is obtained. DAMAS can basically identify the frequency caracteristics of the broadband fan noise, while the discrete tonal noise separation is not well separated.Comparing the DAMAS result with the beamforming's, the improvement in resolution ratio and accuracy has been obviously seen. At last, use the noise attenuation method to transform the fan noise into the noise signals in the position of the microphones, and compare them with the fan inlet noise data provided by GE. The comparasion result shows that DAMAS can separate the directional information of the low-middle frequency zone of fan inlet noise, and that the identification effect of fan inlet noise frequency caracteristics in small angles is also very good. In the aspect of error, DAMAS have big errors in the zone of large angles and high frequencies, but the influence on the total identification of fan noise is not very important.The research shows that the DAMAS can basically identify the fan noise caracteristics from the total engine noise, so it can provide a reference for the airworthiness certification of the engine and the components.