Jae Yoon

Gearbox Tooth Cut Fault Diagnostics Using Acoustic Emission and Vibration Sensors — A Comparative Study

In recent years, acoustic emission (AE) sensors and AE-based techniques have been developed and tested for gearbox fault diagnosis. In general, AE-based techniques require much higher sampling rates than vibration analysis-based techniques for gearbox fault diagnosis. Therefore, it is questionable whether an AE-based technique would give a better or at least the same performance as the vibration analysis-based techniques using the same sampling rate. To answer the question, this paper presents a comparative study for gearbox tooth damage level diagnostics using AE and vibration measurements, the first known attempt to compare the gearbox fault diagnostic performance of AE- and vibration analysis-based approaches using the same sampling rate. Partial tooth cut faults are seeded in a gearbox test rig and experimentally tested in a laboratory. Results have shown that the AE-based approach has the potential to differentiate gear tooth damage levels in comparison with the vibration-based approach. While vibration signals are easily affected by mechanical resonance, the AE signals show more stable performance.

On the Use of a Single Piezoelectric Strain Sensor for Wind Turbine Planetary Gearbox Fault Diagnosis

Planetary gearboxes (PGBs) are widely used in the drivetrain of wind turbines. Any PGB failure could lead to breakdown of the whole drivetrain and major loss of wind turbines. Therefore, PGB fault diagnosis is important in reducing the downtime and maintenance cost and improving the reliability and lifespan of wind turbines. PGB fault diagnosis has been done mostly through vibration analysis over the past years. Vibration signals theoretically have an amplitude modulation (AM) effect caused by time-variant vibration transfer paths due to the rotation of planet carrier and sun gear, and therefore, their spectral structure is complex. Strain sensor signals, on the other hand, are closely correlated to torsional vibration, which is less sensitive to the AM effect caused by rotating vibration transfer path. Thus, it is potentially easy and effective to diagnose PGB faults via stain sensor signal analysis. In this paper, a new method using a single piezoelectric strain sensor for PGB fault diagnosis is presented. The method is validated on a set of seeded localized faults on all gears, namely, sun gear, planetary gear, and ring gear. The validation results have shown a satisfactory PGB fault diagnostic performance using strain sensor signal analysis.

Development of an Efficient Prognostic Estimator

In this paper, the development of a new prognostic estimation technique for on-line gear health management system is described and demonstrated with real spiral bevel gear run-to-failure test data. Unlike conventional particle filter-based prognostic estimation methods, the prognostic technique presented in this paper is a hybrid of the unscented Kalman filter and particle filter. It is designed to improve the processing efficiency whilst the state estimation accuracy is maintained. The unscented Kalman filter is utilized to obtain the “best estimate” of the states of a degrading nonlinear component and the particle filter l-step ahead prediction technique is employed to obtain the remaining useful life of the component. In addition, data mining techniques are applied to efficiently define the system dynamics model, observation model, and predicted measurement information for the prognostic estimator. At last, the feasibility of the presented prognostic estimator is demonstrated with satisfactory results using the actual oil debris mass and health index data obtained from a spiral bevel gear test rig.

Online condition monitoring and remaining useful life prediction of particle contaminated lubrication oil

To increase wind energy production rate, there is a pressing need to improve the wind turbine availability and reduce the operational and maintenance costs. The safety and reliability of a functioning wind turbine depend largely on the protective properties of the lubrication oil for its drive train subassemblies such as gearbox and means for lubrication oil condition monitoring and degradation detection. The purpose of lubrication oil condition monitoring and degradation detection is to determine whether the oil has deteriorated to such a degree that it no longer fulfills its function. In this paper, particle contamination of lubrication oil and the remaining useful life (RUL) of the particle contaminated lubrication oil are investigated. Physical models are developed to quantify the relationship between particle contamination level and the outputs of commercially available online oil dielectric and viscosity sensors. The effectiveness of the developed models is then validated using laboratory experiments. In particular, the remaining useful life prediction of degraded lubrication oil with viscosity and dielectric constant data using particle filtering is presented. A simulation case study is provided to demonstrate the effectiveness of the developed technique.