Research On Vibration Signal Model And Fault Detection Of Planetary Gearbox

Planetary gearboxes are widely used in wind turbines on account of their properties of lower weight, higher transmission ratio and efficiency, bigger bearing capacity in comparison with fixed-shaft gearboxes. In wind turbine, the failure rate of planetary gearboxes is extremely high due to the harsh operation conditions. These gearboxes are often installed far off the ground which makes the maintenance cost very high. Due to these facts, developing the techniques for gearbox condition monitoring and fault diagnosis really makes a lot of sense.When a planetary gearbox operates, the planet gears revolves around not only their own centers, but also the center of the whole planetary stage. Meanwhile, the planet gears mesh simultaneously with both the sun gear and the ring gear. Hence, the vibration of the planetary gearbox is much complicated. What’s more, the time-dependent speed and load owing to the variation of wind add complexity to fault diagnosis.In order to deal with these problems, a series of research has been conducted, and they are listed below.(1) Some common failure types and the vibration mechanism are reviewed at first. Then the frequently-used techniques are summarized, including some methods in time domain, frequency domain, time-frequency domain and order tracking.(2) According to the requirements of the project, the experiment is well designed and finally a planetary gearbox test platform under variable-speed-and-load condition is built.(3) This paper assumes that the whole gearbox revolves around the center of the sun gear at the same speed of the planet carrier. Then the fault diagnosis for planetary gearbox is turned into two simpler parts: the fault diagnosis for fixed-shaft gearbox and the research on the influence of the rotation of the sensor. Based on this assumption, a vibration signal model is developed to detect the failure of the planetary gearbox, and the effectiveness is validated through analyzing a simulated signal and the experimental data.(4) STFT(Short Time Fourier Transform) as a time-frequency method is introduced to analysis the time-dependent frequency data. It is not easy for STFT to detect the failure, thus order analysis is introduced. Order tracking method eliminates the impact of time-varying speed by even-angle resampling. By the analysis of the experimental data, it proves that order spectrum analysis method is useful in planetary gearbox fault diagnosis under time varying operations.(5) A planetary gearbox vibration signal analysis system is developed using MATLAB GUI based on the fault diagnosis theories mentioned in the paper.