Outlet guide vannes fault detection via vibro-acoustic signal processing
Résumé
Turbomachinery is subject to several types of fault. The fault diagnosis literature focuses mainly on the mechanical faults related to shafts and gearboxes and little interest is given to loss of a part of the outlet guide vanes (OGVs). This kind of faults causes an efficiency drop in the engine, as the OGVs role is to recover the mechanical energy of the induced tangential velocity generated by the fan. Thus, the objective of this study is to design simple and efficient indicators able to detect this fault. This aims at avoiding engine dismounting for visual inspection, especially when it comes to stator of compressor or turbine. The impingement of the fan wake on OGVs generates tonal noise caused by unsteady pressure variations on OGVs and the periodic character of the wake. The signature of this noise will change with OGVs removal. Thus, the proposed indicators are mainly based on the variation tracking of the noise features such as the harmonics of the blade passing frequency. Furthermore, more indicators can be deduced from mode decomposition of the rotational modes generated by rotor-stator interaction. A healthy baseline is to be compared with results of several configurations of OGVs removal. Considering that the coupling between the rotor and stator is only acoustical, the emphasis is put on the analysis of acoustic signals. However, from structure vibration point of view, indicators adapted to accelerometric signals are also investigated. This approach is applied on vibro-acoustic signals measured on a turbofan and the first results are rather satisfying. The analysis is still running and the final results will be presented at the congress.
Origine : Fichiers éditeurs autorisés sur une archive ouverte