Investigation of Sound Generation and Transmission Effects Through the ACAT1 Fan Stage using Compressed Sensing-based Mode Analysis

Abstract

In an extensive measurement campaign at the UFFA test rig of AneCom AeroTest in Wildau, Germany, the sound field generated from the ACAT1 fan stage was measured by several dynamic pressure sensor arrays of different geometrical arrangements at the inlet, interstage and bypass section. The sparse ring arrays with irregular and optimised geometries are utilised in combination with the Compressed Sensing-based mode analysis in order to investigate the tonal and broadband sound generation and transmission with high resolution. From the azimuthal mode spectra, the relation between the cut-on mode range at the sources and the dominant modes at the measurement sections is observed and used to determine the contribution of different sound generation mechanisms. With the help of interstage mode analyses strong rotor-coherent contributions due to rotor wake irregularities are identified and the rotor-shielding effect is assessed. The dependencies of the broadband mode spectra with respect to the operation points and working lines are shown and the contribution of stator and rotor bound sources is investigated. Results indicate that mode scattering at the stator vane row has a significant influence on the downstream radiated sound field. Evidence is further found for the occurrence of the acoustic tunnel effect, where cutoff modes excited at the stator leading edge become cut-on during the propagation through the inter-vane channels. This effect triggers mode scattering towards rotor counter-rotating modes and increases the sound radiation from fan stages.

Assessment of a 2D Synthetic Turbulence Method for Predicting the ACAT1 Fan's Broadband Noise

“…For a more detailed description of the experimental setup, please refer to Guérin et al 19 or Behn and Tapken. 33 As the 3D-equivalent simulation delivers plausible results as shown in the previous section, all further analyses are performed using this technique. In order to ensure the best possible comparison with experimental data, a simulation of the full duct containing all 44 stator vanes was computed.…”

Radial mode breakdown of the ACAT1 fan broadband noise generation in the bypass duct using a sparse sensor array

“…For the azimuthal mode decomposition a Compressed Sensing approach is taken by applying the EOMP algorithm 10 , which solves the following minimization problem:…”

“…The study was carried out in the frame of the EU project TurboNoiseBB and contributes to a comprehensive acoustic and aerodynamic data base 10,11 with the objective to enable the detailed characterization of the fan broadband noise generation and propagation mechanisms [12][13][14][15][16] . The ACAT1 fan was operated in two rotor-stator configurations at various shaft speeds on two working lines, among this the acoustic certification points.…”

Analysis of the influence of inflow distortions on turbofan rotor noise

“…where W is a matrix containing the mode transfer functions e imϕ k evaluated for the mth mode order at the kth sensor position. The mode spectrum is obtained by solving the system of linear equations in equation 5 using a Compressed Sensing technique as described in [17] for mode orders in the range from -79 to +79. The fluctuating components are described statistically in terms of mean values and spatial cross-spectra.…”

“…where S pp is the cross-spectral matrix of the pressure signals, S aa is the cross-spectral matrix of the mode amplitudes, and W the mode transfer matrix from equation 5. The mode amplitudes of the fluctuating components are determined by solving equation 6 for the cross-spectral matrix S aa using the Compressed Sensing technique described in [17]. Figure 2 illustrates the global effect of the ICD.…”