Multiparameter vibration analysis of various defective stages of mechanical components

Transcription

1 SISOM 2009 and Session of the Commission of Acoustics, Bucharest May Multiparameter vibration analysis of various defective stages of mechanical components Author: dr.ing. Doru TURCAN Abstract The present paper presents the multiparameter vibration analysis approach (in two practical cases), highlighting his importance in the daily predictive maintenance activity of vibration diagnosis of assets health. The vibration monitoring activity for assets health diagnosis are commonly used the following parameters: vibration acceleration, vibration velocity, vibration displacement and also some special parameter as enveloped acceleration, spike energy, etc. Those three common vibration parameter are used in the particularly frequency ranges as it is shown in the bellow diagram. The vibration displacement it s commonly used in analyzing low frequency vibration phenomena, the vibration acceleration it s used to analyze high frequency vibration phenomena (such is blade pass frequency, gear mesh frequency, etc) and the vibration velocity it s used for vibration ranging from 10 to 1000 Hz. Vibration acceleration and vibration velocity are generally used to analyze periodic excitation phenomena (misalignment, unbalance, etc) and random phenomena (cavitation, friction, etc). To analyze repetitive impact phenomena (roller bearing early defect, faulty gear mesh, etc) we will use enveloped acceleration. In the following pages I will present a couple of cases where multiparameter vibration analysis was used to evaluate the defect stage. 1. Roller bearing early defect stage of defect (NU320) During his existence the roller bearing experiences vibrations caused by the repetitive impact forces and the periodic excitation forces. In his early stage of defect (for example a defect on an inner race) a roller element will be excited by the repetitive impact forces when he will pass over a region with metallic structure deteriorated or where a small metallic part it s removed from his path. Those small

2 411 Doru TURCAN vibrations will be analyzed using signal demodulation (enveloped acceleration) and FFT, to obtain a spectrum with dominant roller bearing frequencies of defect (fig.1). Fig. 1 Specific roller bearing defect frequencies pattern spectrum The moment when the defect frequencies are showing up in enveloped acceleration spectra it s presented in the fig.2. Fig.2 Roller bearing evolution in time towards first step of defect The vibration velocity measurement toke on the same spot in the same time as the above enveloped acceleration it s having the following spectrum shown in fig.3. Fig. 3 The studied roller bearing vibration velocity spectrum In this last spectrum can be see that the defect marker it s no matching the peak frequencies. So, the vibration velocity will not show an early roller bearing defect in this situation the repetitive forces don t cause the vibration.

3 Multiparameter vibration analysis of various defective stages of mechanical components 412 Fig.4 Vibration velocity spectrum - evolution in time A deeper analysis of fig. 4 waterfall it s reviling the possible cause of the bearing defect. We can see that the fundamental frequency amplitude (1X the rotation frequency of the asset shaft) has increased very much. The unbalance could be the cause in this situation. The evolution in time of the vibration acceleration, measured in the same conditions as the other two parameters, it s showing that this parameter has encountered a change in this spectrum shape (fig.5) and an increasing overall value trend (fig.6). The repetitive impact forces couse are increasing the high frequencies amplitude causing the rise of the overall value of this parameter Fig.5 Spectral evolution in time of vibration acceleration Fig. 6 Overall trend of vibration acceleration In this case the multiparameter vibration analysis reviles the early stage defect of a roller bearing and his possible cause. 2. Gear Mesh Defect Cement mill gearbox first gear defect It s common to have gear mesh frequency (GMF) in acceleration and velocity FFT spectra. In the defect situation in the spectra can be seen multiple harmonics of those frequencies (GMF) and well defined sidebands accompanying them (fig.7). In the this case the specific spectral pattern of the defect it s present in the spectrum and also an increased vibration (velocity) overall value due to the gear mesh defect (fig.8 si fig.9).

4 413 Doru TURCAN Fig.7 Specific spectral pattern of gear mesh defect Fig.8 Time evolution of gear mesh defect frequency Fig.9 Overall trend of vibration velocity The confirmation of the defect stage has come from the analysis of the enveloped acceleration spectra, which has shown the gear mesh defect frequency. This appearance, in the enveloped acceleration spectra, it s caused by the wear gear mesh that is producing repetitive impact forces.

5 Multiparameter vibration analysis of various defective stages of mechanical components 414 Fig.10 Defect gear frequency in the enveloped acceleration spectra Fig.11 Time evolution of gear mesh defect frequency in the enveloped acceleration spectra Fig.12 Overall trend of enveloped acceleration The gear mesh defect it s early present in the velocity and acceleration spectra. In the final stages this defect can be saw also in enveloped acceleration spectrum because of the presence of the repetitive impact forces from the gear mesh process. Conclusion Each vibration parameter has his importance, depending from the mechanical part, in the determination of the defective stage: early, medium or final. Choosing the right parameter, the proper frequency range, the proper measurement point position and correct FFT analyzer set-up it s based on the understanding of the mechanical process and his behavior towards final defect stage.

6 415 Doru TURCAN References 1. Franz Reithuber, Vibration Principles - An Introduction to Spectrum Analysis - 2. Dr. Robert Jones, Vibration Monitoring of Bearings 3. Donald Howieson, Vibration Monitoring: Envelope Signal Processing - 4. B. J. Woodley, Machine Condition Monitoring Sources of Equipment and Services. 5. J. S. Rao and K. Gupta, Introductory Course on Theory and Practice of Mechanical Vibrations, 2nd Edition, New Age International (P) Ltd, Publishers, pp M. P. Srivastava, IRD Mechanalysis, Vibration Monitoring for Predictive Maintenance, Purchase, August B. K. N. Rao, Handbook of Condition Monitoring, 1st edition, Elsevier Advanced technology, UK, pp.76, B. R. Satyan and H. N. Nagarajan, Predictive Maintenance through Vibration Monitoring, Technical article, Noise and Vibration Laboratory, CMTI, 1988.