(Dr. Steven Y. Liang, advisor)
"Dynamic Diagnostics and Prognostics of Rolling Element Bearing"
The largest bearing manufacturers, with 26% of the market, sell approximately 3 billion dollars worth of bearings each year. As a consequence of their extensive use and importance, bearing failure is one of the foremost causes of breakdowns of rotating machinery. This type of failure can result in catastrophic consequences in many situations, such as in helicopters, transportation vehicles and so on.
The optimization of bearing maintenance schedules and the prevention
of catastrophic failure hinges on the capability to online reliably assess
bearing defect severity under practical operating conditions and real-time
prognosticate its remaining life. Unfortunately, current available techniques
of bearing diagnostics and prognostics have not been well developed. Therefore,
the overall objective of this research is to address these issues by developing
effective methodologies with fundamental understandings for optimization
of bearing maintenance. Two bearing diagnostic models, which have acceptable
performance to evaluate bearing defect conditions on-line, have been developed.
One bearing prognostic model, which shows the capability to predict bearing
remaining life in real-time, has also been developed. With the implementation
of bearing accelerated life testing, an approach, which is capable of evaluating
failure rates of a bearing under both use and accelerated conditions, has
been developed too. Experimental investigations have been performed to
verify these analytical models.
This research addresses the deficiency of current bearing condition-monitoring techniques in the lack of fundamental understandings. This is a real step toward practical applications of individual bearing life prediction based upon condition-monitoring techniques. The potential applications of this methodology will be real-time safety monitoring of any system with rolling element bearings such as helicopter transmissions and machinery in automatic manufacturing environments. Based on the generalization of the developed methodologies, the techniques have the potential applications to any other mechanical components, such as gears, spindles, cams, etc.