A proposal of a technique for correlating defect dimensions to vibration amplitude in bearing monitoring

The capability of early stage detection of a defect is gaining more and more importance because it can help the maintenance process, the cost reduction and the reliability of the systems. The increment of vibration amplitude is a well-known method for evaluating the damage of a component, but it is sometimes difficult to understand the exact level of damage. In other words, the amplitude of vibration cannot be directly connected to the dimension of the defect. In the present paper, based on a non-Hertzian contact algorithm, the spectrum of the pressure distribution in the contact surface between the race and the rolling element is evaluated. Such spectrum is then compared with the acquired spectrum of a vibration response of a defected bearing. The bearing vibration pattern was previously analyzed with monitoring techniques to extract all the damage information. The correlation between the spectrum of the pressure distribution in the defected contact surface and the analyzed spectrum of the damaged bearing highlights a strict relationship. By using that analysis, a precise correlation between defect aspect and dimension and vibration level can be addressed to estimate the level of damaging.