Possibilities of faults detection of rolling bearings using energetic descriptors of vibrations signals

Abstract

The need for fast and reliable evaluation of technical state of rotating machines forces constant development and research for condition monitoring techniques. The paper presents energetic characteristics of vibration signals as a promising new approach in condition monitoring of rolling bearings. The presented approach is based on application of the differential Teager-Kaiser energy operator. The operator makes possible to the detection of short-time disturbances in the signal which are caused by developing faults. Authors assumed that the energetic characteristics and measures would be good tool for detection of faults and defects in rolling bearings, especially when vibration signals are non-stationary in the amplitude and/or the frequency sense. The paper presents the energetic characteristics of the bearing vibration signal in the form of the time history, the energetic trajectories and measures parameterizing them. The obtained results give ability to determine the basic features of characteristics and measures,. The presentation of qualitative changes in the form of characteristics caused by different kinds of faults of rolling bearings was one of the main aims of the research. From practical point of view the assessment of the sensitivity of the above-mentioned energetic measures on changes in technical condition of bearings was also crucial. The presented results have been obtained by testing the set of tapered roller bearings of the same type and size. New bearings, defective ones and bearings with artificially introduced faults were tested.