Research On Gearbox Vibration Simulation And Fault Diagnosis

Gearbox is one of the widely used components of machinery and equipment. With the rapid development of modern industry, science and technology, the normal operation of the gearbox impact on production practice is growing growing day by day. Therefore, it is imperative to study and development of effective equipment condition monitoring and fault diagnosis technology for equipment management and maintenance.Gearbox vibration signal is one of the main gearbox running state characterization, and the main object of the gearbox running state monitoring and fault diagnosis studying. In this study the gearbox vibration mechanism was researched, and the gearbox vibration response under dynamic excitation force was compared with the test signal. By research the method of gearbox vibration data treatment, the advantages and disadvantages of each method in gearbox fault diagnosis was analyzed and compared. The study further investigated the principle, meaning and application in gearbox fault diagnosis of the wavelet analysis and wavelet packet analysis methods. Due to gearbox failure vibration signals are often non-stationary, the study also proposed the feature extraction method based on Teager energy operator and wavelet packet de-noise. The results of this study include:1) To study the gearbox vibration mechanism, internal excitation force of gear was simulated by Matlab and Adams considering the gear dynamic excitation, such as error excitation, time-varying mesh stiffness excitation and meshing impact excitation’s affect to box vibration. The verification shows that the simulation method was reliable.2) To study NGSD230 type gearbox, the parametric gears model, the gear transmission dynamics simulation model and the gearbox casing model were created using Pro/E.3) The result of gearbox vibration, by Pro/E, Hypermesh and ANSYS, was compared with the measurement of box vibration. The vibration acceleration frequency distribution was substantially the same which proves the correctness of the simulation. Finally, this simulation method was applied to the gearbox fault simulation, and its further validate the correctness of the simulation method can provide a good reference for the vibration of the gearbox.4) The study showed that wavelet packet was useful for the vibration signal de-noise.5) To determine whether the gearboxes are failure or not, if failure then what was the fault source, the signal (the wind turbine gearbox vibration data by test) was analyzed by preprocessing method, wavelet packet de-noise, fast Fourier transform and refinement of the spectral analysis method. It was found that the gearbox did not have an obvious failure.6) In order to effectively extract the characteristic parameters of failure signal, the method of signal characteristic parameter extraction based on wavelet packet de-noise and nonlinear Teager energy operator was proposed. Constructed signal simulation demonstrated the feasibility of this method to extract the signal characteristic parameters, and this method was successfully applied to the eccentric gear gearbox fault diagnosis, which demonstrated the effectiveness of this method to extract characteristic parameter of gear fault gearbox vibration signal.