An investigation of a detection method for a subsurface crack in the outer race of a cylindrical roller bearing

As one of major failure modes of roller bearings due to periodic contact forces and external impulse loads, subsurface cracks caused by fatigues may produce catastrophic failures of rotating machines. Investigations of subsurface crack detection methods for roller bearings are very useful for maintenance purposes of these machines. In this study, a new detection method based on the curvature and power spectral density (PSD) of displacements is presented to detect a subsurface crack in the outer race of a cylindrical roller bearing. A dynamic finite element model of the cylindrical roller bearing with a subsurface crack in its outer race is developed using an explicit dynamics finite element software package to obtain the time-domain displacements. Differences of the curvature and PSD of displacements of the bearing without and with the subsurface crack are investigated, which are used to detect the location of the subsurface crack with different sizes in the outer race of the bearing. The results show that differences of the curvature and PSD of displacements from the measurement points on the outer race of the cylindrical roller bearing without and with the subsurface crack can be used to detect the location of the crack.