Failure acceleration and condition diagnosis of ball bearings by an excitation mechanism using piezo actuators

Abstract

Damage to the bearings can stop the entire machine, resulting in serious economic losses. In recent years, a machine learning-based diagnostic method that can automatically detect abnormalities has been proposed. However, the development and evaluation of similar methods is computationally expensive and cost intensive. In this study, we propose a novel fault acceleration method using a two-axis piezo actuator that can obtain data from normal to initial abnormal conditions in a short time by applying vibration of arbitrary frequency to a rotating shaft. Such a failure acceleration mechanism has never been proposed before. Using the proposed method, the failure of the bearing was accelerated by applying vibrations, and it was confirmed that the rupture of the oil film resulted in surface-based delamination. The condition of the bearing with internal damage was diagnosed using the acceleration data. The diagnostic results revealed that the failure of the bearing progressed differently depending on the oscillation frequency of the applied vibrations, and that a particular frequency could accelerate failure. The actual damage occurring inside the bearing was confirmed by a cutting-disassembly inspection, and the degree of damage varied depending on the pendulum frequency, similar to the acceleration diagnosis results. Furthermore, based on the results of the cutting-disassembly inspection and load estimation of the bearing, it was confirmed that surface-based delamination damage appeared inside the bearing due to oil film rupture. This is the first paper to present the results of a cutting-disassembly inspection of a ball bearing accelerated to failure by the two-axis piezo actuator during rotation. No other research has investigated in detail the time variation of kurtosis and envelope spectra by acquiring acceleration data of ball bearings up to initial abnormal conditions using the novel failure acceleration mechanism.