Abstract
Unsteady aerodynamic phenomena appearing close to the stall boundary in axial fans and compressors have been the subject of extensive investigations. A particular phenomenon known as “rotating instabilities” typically occurs in the tip region at highly loaded conditions and is often linked to blade vibrations. During rig and engine testing, it is usually identified by the characteristic shape of its pressure spectra. In the 1990s, this shape has been explained as the result of a circumferentially propagating disturbance, which is unsteady in a frame of reference rotating with the disturbance. In this analysis, conclusions regarding its propagation speed and frequency have been made based on the analysis of spectral peaks, and this method is still often used to classify unsteady aerodynamic phenomena. However, in high subsonic and transonic machines, where aeroacoustic and aeroelastic phenomena interact with aerodynamic disturbances, the interpretation of measurements using spectra alone is challenging. The present article aims to demonstrate the difficulties and subtleties associated with the analysis of measurement signals, which need to be overcome to correctly interpret stall precursor signatures. At the example of a recent investigation on a composite fan, the consequences of sensor placement and postprocessing techniques are discussed with a focus on spectral averaging, isolation of non-synchronous phenomena, and multisensor cross correlation methods. It is seen that the interpretation of phenomena based solely on spectral peaks and their spacing can be misleading and that the characteristic features of a rotating instability spectrum do not require an unstable pulsating disturbance.
