Fault detection in rotating machines with beamforming: Spatial visualization of diagnosis features - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Mechanical Systems and Signal Processing Année : 2017

Fault detection in rotating machines with beamforming: Spatial visualization of diagnosis features

Résumé

Rotating machines diagnosis is conventionally related to vibration analysis. Sensors are usually placed on the machine to gather information about its components. The recorded signals are then processed through a fault detection algorithm allowing the identification of the failing part. This paper proposes an acoustic-based diagnosis method. A microphone array is used to record the acoustic field radiated by the machine. The main advantage over vibration-based diagnosis is that the contact between the sensors and the machine is no longer required. Moreover, the application of acoustic imaging makes possible the identification of the sources of acoustic radiation on the machine surface. The display of information is then spatially continuous while the accelerometers only give it discrete. Beamforming provides the time-varying signals radiated by the machine as a function of space. Any fault detection tool can be applied to the beamforming output. Spectral kurtosis, which highlights the impulsiveness of a signal as function of frequency, is used in this study. The combination of spectral kurtosis with acoustic imaging makes possible the mapping of the impulsiveness as a function of space and frequency. The efficiency of this approach lays on the source separation in the spatial and frequency domains. These mappings make possible the localization of such impulsive sources. The faulty components of the machine have an impulsive behavior and thus will be highlighted on the mappings. The study presents experimental validations of the method on rotating machines.
Fichier non déposé

Dates et versions

hal-01595797 , version 1 (27-09-2017)

Identifiants

Citer

E. Cardenas Cabada, Quentin Leclere, J. Antoni, Nacer Hamzaoui. Fault detection in rotating machines with beamforming: Spatial visualization of diagnosis features. Mechanical Systems and Signal Processing, 2017, 97, pp.33 - 43. ⟨10.1016/j.ymssp.2017.04.018⟩. ⟨hal-01595797⟩
99 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More