Investigation on full ceramic bearing fault diagnostics using vibration and AE sensors

Full ceramic bearings are considered the first step towards full ceramic and oil free engines in the future. Few researches on full ceramic bearing fault diagnostics using both vibration and acoustic emission (AE) sensors have been reported. In this paper, a research investigation on full ceramic bearing fault diagnostics using vibration and AE sensors is reported. The research utilizes empirical mode decomposition (EMD) to pre-process both vibration and AE signals and a novel multidimensional vibration and AE fault feature extraction method is developed to generate the condition indicators (CIs). These CIs are used to build a fault classifier using a k-nearest neighbor (KNN) algorithm. Seeded fault tests on full ceramic bearing outer race, inner race, balls, and cage are conducted on a bearing fault diagnostic test rig and both vibration signals and AE burst type signals are collected. The effectiveness of the vibration and AE based diagnostic techniques is validated using real full ceramic bearing seeded fault test data. A comparison of fault diagnostic performance between vibration and AE sensors is provided.