Mechanical Seal End Faces Friction Condition Monitoring Based on Particle Filter

Abstract

The non-contact mechanical seal end faces opening friction condition detection and the measurement of film thickness when the end faces is just-lift-off, which are always key problems for scholars engaged in sealing industry for many years, but there is no effective solution. Acoustic emission (AE) signal generated in the running process of mechanical seal end faces has a plenty of information about the faces contact state. According to this, the thickness measurement of mechanical seal and opening condition monitoring technology by using particle filter are put forward based on the acoustic emission signal. Acoustic emission sensor is installed in the stationary ring seat, used for the indirect outer detection of the dynamic and stationary rings opening condition. The acoustic emission signals are processed by particle filter technology, and then the signal features are extracted in time domain, frequency domain and wavelet packet energy. A BP neural network model is established, the features of signal characteristic used as input of the model. It is finally realized that the mechanical seal end faces opening condition was recognized and the film thickness was measured. Eddy current sensor is installed inside the stationary ring and used for direct measurement of film thickness and verify the results got from the BP neural network. Through the experiment, this method is practical and effective, and which can be used in the monitoring of mechanical seal end faces working condition in industrial field.

The Contact State Monitoring for Seal End Faces Based on Acoustic Emission Detection

“…But because of the unstable spectrum structure of the seal signals and the dispersed distribution of the AE features, both the methods will be ineffective when random contact occurs. Particle filter is a time-domain method for signal denoising, and this method has been applied to measure the film thickness [16]. However, since the particle diversity reduces during the resampling procedure, this method cannot reach the best performance.…”