Location:Home > Engineering science > General and Fundamental Mechanics > Rolling Bearing Fault Diagnosis and Vibration Signal Processing

Rolling Bearing Fault Diagnosis and Vibration Signal Processing

Downloads: []
Tutor: XuMinQiang
School: Harbin Institute of Technology
Course: General and Fundamental Mechanics
Keywords: rolling bearing,vibration signal,fault diagnosis,kurtosis,time andfrequency doma
CLC: TH133.33
Type: Master's thesis
Year:  2012
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

As a basic component of machine, rolling bearing plays an important role ofensuring the system work steadily. In order to reduce unnecessary losses and makesure the safety of our workers, it is necessary to position the bearing fault locationand diagnose the fault severity accurately first time. In purpose of this, this articlemainly to analyze the fault vibration signal through experiment and simulation.This paper firstly introduced the vibration signal characteristics of the faultbearing. Combined with the actual bearing of No.205relations derived the signalcharacteristic frequency. The part of simulation established the finite element modelwith Patron and the dynamic model with Adams. We analyzed the effect of thebearing load distribution and natural frequency on fault vibration signal through thefinite element model. We analyzed the bearing fault characteristics of different partand fault signals influencing factors: the fault size, load, speed, gap, with thedynamic model. Proposed the optimization of the sensor measurement position,though the way of using measurement position of45degrees oblique in place ofvertical position to increase the fault sensitivity of the pulse signal measuring.We collected1200set of experimental data with the bearing of No.205. Usingtwo class method of frequency domain analysis and statistical analysis to locate thefault and the severity of the fault diagnosis. The application of statistics needs alarge database as a background, take advantage of the data to divide the failureinterval, to diagnose different failure modes with the method of kurtosis and todistinguish the degrees of fault with the method of P/R. This paper mainly used theimproved multiple envelope of resonance demodulation through analyzing thedomain of time and frequency to verify the diagnostic accuracy of the signal.Accuracy rate of the fault diagnostic is almost90%. In order to extract thecharacteristic signal and filter out the interference signal, this paper using thespectral kurtosis and wavelet packet energy to improve signal to noise ratio.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top