Helical Gear Fault Feature Extraction And Damage Process Analysis Based On HVD-AR

Gearbox,a basic widely used transmission part,is mainly used to change the speed of revolution and transmission power.However,it prone to failure due to its complex structure and poor working conditions.Therefore,it is necessary to monitor and diagnose the fatigue damage of gearbox.Gearbox fault feature extraction based on vibration signal,its main task is to extract the fault feature from the collected signal.The prerequisite of this process is to separate the fault vibration signals accurately and the fundamental purpose is to extract the fault feature.This paper is based on the helical gear fatigue tes t,its research object is the real time vibration signals in the whole fatigue life of helical gears.It proposes an effective method to de-noise the vibration signals and characteristic indexes of gear damage degree.The changing trend of characteristic index is applied to analysis the fatigue damage process of helical gears.This paper studied the helical gear vibration signal mechanism based on the simplified single tooth meshing model.The helical gear typical faults and its characteristics of vibration signals are analyzed and then reviewed the common fault characteristic indexes.According to the nonlinear and non-stationary characteristics of helical gear vibration signal,this paper proposed the vibration signal de-noising method HVD-AR and focused on the signal reconstruction method for HVD and CK criterion for AR order.The effectiveness of the de-nosing method was verified by the vibration signal Hilbert spectrum for the measured pitting process.According to the mechanism of the impact on the gear vibration signal frequency domain sideband,this paper proposed two characteristic indexes (4 and (4,and was applied to the measured signals.Considering the existing conditions of helical gear fatigue bench test,this paper developed the test scheme and briefly analyzed the test results.At the same time,the vibration signal data of the whole life process were obtained,and it provided the necessary data support for the helical gear fatigue damage process analysis.The validity of the signal de-nosing method was verified by the whole life vibration signal for the helical gear typical damage process,including tooth surface pitting and the tooth broken.The changing trend of characteristic indexes (4 and (4 was comparatively analyzed to the common characteristic indexes,including RMS,kurtosis,ER and FM4,to verify its practicality to reflect the helical gear damage process.The result shows that HVD-AR can effectively de-noise the helical gear vibration signal and the changing trend of characteristic indexes (4 and (4 can effectively reflect the helical gear fatigue damage process,and can detect the early weak fault.