Final Activity Report Summary - INDUCT (Flow duct acoustics: a practical demonstration to noise source reconstruction)
The simulation and experimental results obtained from the INDUCT project have provided a deeper understanding of the limitation performances of up-to-date inverse methods for the localisation and the reconstruction of ducted acoustic sources for a number of source and duct configurations, which is of particular relevance to design novel duct noise control strategies in the aeronautical industry and in any sectors concerned with acoustic emissions from air-conditioning systems.
Ranges of physical parameters have been found such as the source-sensor separation distance, the frequency of reconstruction, the nature of the measured in-duct data field or the signal-to-noise ratio, under which an accurate and well-resolved experimental reconstruction of ducted sources can be achieved under flow- and no-flow conditions, thus providing guidelines for further identification of duct-borne noise sources. Several characterisation methods such as the Equivalent Source Method (ESM) or high-resolution Focussed Beamforming approaches have been assessed based on an equalisation of the modelled source-sensor transfer paths in order to locate and retrieve a number of ducted source strengths distributions with varying spatial correlation degrees. A clear short-coming of the ESM is the need to assume a prior model of suitably located equivalent sources. A novel method has been proposed based on a spectral decomposition of the unknown source strength as series of the duct modes, and which capture both the location and the strength of the source. It has been assessed to extract from in-duct field measurements the distortion frequencies of a compression driver wall-mounted onto a flow duct acoustic facility. Furthermore, the effect of a lined duct section on the reconstruction has been discussed together with the identification of more real-life aero-acoustic noise sources such as the loading noise due the interaction between a non-uniform circumferential flow entering the inlet and the rotating blades of a subsonic fan.
A key factor that influences the accuracy of the reconstruction is the amount of independent information generated by the sources and which can be measured by the sensors. For the first time, closed-form expressions have been found for the independent source and field modes onto which the radiation operator of baffled planar vibrating structures can be decomposed, thus providing a further insight into the physical parameters that govern the accuracy and resolution of the reconstruction of structure-borne noise sources.
Ranges of physical parameters have been found such as the source-sensor separation distance, the frequency of reconstruction, the nature of the measured in-duct data field or the signal-to-noise ratio, under which an accurate and well-resolved experimental reconstruction of ducted sources can be achieved under flow- and no-flow conditions, thus providing guidelines for further identification of duct-borne noise sources. Several characterisation methods such as the Equivalent Source Method (ESM) or high-resolution Focussed Beamforming approaches have been assessed based on an equalisation of the modelled source-sensor transfer paths in order to locate and retrieve a number of ducted source strengths distributions with varying spatial correlation degrees. A clear short-coming of the ESM is the need to assume a prior model of suitably located equivalent sources. A novel method has been proposed based on a spectral decomposition of the unknown source strength as series of the duct modes, and which capture both the location and the strength of the source. It has been assessed to extract from in-duct field measurements the distortion frequencies of a compression driver wall-mounted onto a flow duct acoustic facility. Furthermore, the effect of a lined duct section on the reconstruction has been discussed together with the identification of more real-life aero-acoustic noise sources such as the loading noise due the interaction between a non-uniform circumferential flow entering the inlet and the rotating blades of a subsonic fan.
A key factor that influences the accuracy of the reconstruction is the amount of independent information generated by the sources and which can be measured by the sensors. For the first time, closed-form expressions have been found for the independent source and field modes onto which the radiation operator of baffled planar vibrating structures can be decomposed, thus providing a further insight into the physical parameters that govern the accuracy and resolution of the reconstruction of structure-borne noise sources.