Analysis of the Rolling Element Bearing data set of the Center for Intelligent Maintenance Systems of the University of Cincinnati

Due to its importance in the industry, vibration-based diagnosis and prognosis of rolling element bearings (REB) attract more and more attention from the research community. Moreover, it is now possible to test newly developed methods on test rig data that are open-access. The NASA provides a list of data bases concerning prognosis and some REB data-sets, such as the one from the Center for Intelligent Maintenance Systems (1) (IMS), University of Cincinnati. This data-set has so far been used by many authors. However, even if the state of health of the tested bearings is clearly provided and sometimes supported by photographs taken at the end of endurance, the diagnosis from the provided vibration signals is not so obvious. This paper tries to establish what can be clearly concluded from the analysis of the signals. In order to fulfill this objective, several diagnostic techniques are used: time analysis, spectral analysis, blind deconvolution, spectral coherence, envelope spectrum. A diagnosis is therefore returned whenever possible, together with recommendations for prognosis in those cases which have been identified as difficult.