Identify The Defect Of Bearing Vibration And Noise With Wavelet Packet Energy Spectrum

With the development of automobile industry and appliance, it requires the rolling bearing with little vibration and low noise, so bearing factory pay more attention to how to reduce the bearing vibration and noise. The current bearing diagnosis technique focus on finding causes of bearing invalidation. But for the bearing factory, they are more concerned is which factor cause new bearings having high vibration value and noise. This article wants to find out a availability method to defect the bearing fault rapidly and effectively. With this method the factory can improve the product craft timely in order to enhance the voice and vibration quality.In this article, the vibration model is constructed first, the model based on the tribo-dynamics viewpoint of bearing vibration, it reflects the effect of lubrication. In this paper, the exciting and response characteristics of bearing fault is discussed. The analysis of vibration model build the basic of defect diagnosis. On the other hand, the compatible means of signal analysis is introduced and the eigenvector to token bearing estate is found. The results are as follows:1. Overcoming the linearization hypothesis of classic bearings vibration model, establish the bearing non-linear model. The model considers bearing non-linear dynamics characteristics, establish the vibration equation with lubrication.2. Analyse the impact factor of bearing vibration. According the exciting and response characteristic of the factor to look for the likely expression. With the simulink, the defect trait of the waviness in ballsand rings is discussed.3. The vibration signal of new bearing is not similar with the signal of malfunction bearing, the character of signal of new bearing is analysed , so choose wavelet packet method to dispose the signal.4. With the theory of wavelet packet energy spectrum, raise wavelet packet energy spectrum variety rate. Use this rate as a eigenvector to pick-up the defect, then reconstruct the signal. With the result, bearing defects is found exactitude.