Abstract Rotating machines are widely used in industry and are susceptible to multiple faults, commonly including rotor unbalance and bearing degradation. Regarding hydrodynamic bearings, oil supply and cooling are critical; however, they cannot be closely monitored in machines without dedicated sensors in the oil lines. Thus, this paper proposes a non-intrusive, model-based method to identify oil inlet flow rates, oil inlet temperature, and rotor unbalance through vibration measurements, without instrumenting the lubrication circuit. The method solves a nonlinear inverse problem by matching selected vibration features with a coupled rotor–thermo-hydrodynamic bearing model to estimate the unknown operating and fault parameters. Unlike most related works that focus on discrete lubrication-state classification, the proposed approach performs continuous, quantitative regression of oil-inlet conditions, which can support the detection of incipient changes that may not be captured by coarse state classification. In addition, the method simultaneously identifies lubrication inputs and rotor unbalance in a coupled rotor–bearing system, addressing a multi-fault setting that is rarely considered in the literature. Numerical results show that the method can track time-varying oil inlet temperature and flow while estimating rotor unbalance from noisy measurements (10 dB SNR), with a maximum error of 5.8% and an average identification time of 25.3 s, supporting online monitoring.
Oliveira et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: