Journal bearing condition monitoring based on the modulation signal bispectrum analysis of vibrations

Journal bearings usually wok under a wide range of operating conditions. However, adverse operating such as transient operations and oil degradation can lead to early defects to the bearings. In this paper, modulation signal bispectrum (MSB) is used to analyse vibration responses from a journal bearing lubricated with three different oils to differentiate abnormal lubrication conditions. MSB magnitude results represent the nonlinear vibration responses, which are due to instable hydrodynamics, asperity excitations and nonlinear transfer paths, with two distinctive bifrequency patterns corresponding to instable lubrication and asperity interactions respectively. Using entropy measures, these instable lubrications are classified to be the low loads cases. Furthermore, average MSB magnitudes are used to differentiate the asperity interactions between asperity collisions and the asperity churns. A higher magnitude in the lower frequency band can indicate the excessive asperity contacts due to lowering viscosities. Meanwhile a higher magnitude in the higher frequency band indicates the extreme fluid frictions.