Inverse Calculation Of Gear Dynamic Parameters And Quantitative Fault Diagnosis Method

As one kind of the most important mechanical device,the operation condition of the gear directly affects the production and safety of mechanical.Therefore,how to monitor and diagnose gear failure is the top priority for ensuring the normal operation of machine.The study of fault mechanism is the basis of fault diagnosis,and the purpose is to reveal the evolution law of the gear fault by establishing reasonable gear dynamic model.The fault diagnosis method is a tool to solve the fault problem which is beginning from the practical application of engineering to finds the fault phenomenon and solves the fault problem,and provides a more intuitive and convenient way of troubleshooting for engineers.However,these exist some problems such as the fault mechanism is uncertainty,earliest fault damage is hard to detect,dynamic parameters of gear fault is difficult to obtain,transformation from qualitative fault diagnosis to quantitative problem,mechanism research result and fault diagnosis methods are difficult to combine organically.Thus,this paper fisrt starts with the dymaic modeling of gear system,and the methods of dynamic parameters calculation and quantitative fault diagnosis are studied.The detailed contents are listed as follows:1.Mainly studied the typical incipient damage types in gear failure: the failure mechanism of tooth root cracks and tooth surface pitting and the influence on the gear vibration system.By the construction of dynamic model of gear system which contains time-varying meshing stiffness(TVMS)and friction factor,the evolution discipline of dynamic parameters under the typical failure state are analyzed and the influence of different gear failure types on the dynamic parameters are studied.Due to the different need for fault mechanism analysis of gear system,the time-domain analysis,frequency-domain analysis and statistical characteristic indicators are used to analyze the evolution discipline of dynamic response under the different gear failure types.2.In order to solve the drawbacks of TVMS calculation method,the stress intensity factor(SIF)is introduced into the TVMS calculation of cracked gear.After that,a novel TVMS calculation method of cracked gear based on assist-SIF is proposed,which considers the sensitivity of the stress field intensity to the size of the tooth root crack and combining the advantages of finite element method(FEM)and analytical method(AM).The research shows that the proposed method can effectively obtain the TVMS by comparing the traditional TVMS calculation method.Besides,the dynamic behaviors of cracked gear are studied based on the dynamic model of the gear system including the TVMS results of proposed method.3.To resolve the problem that the dynamic parameters are difficulty calculated in the research of gear failure mechanism and the calculation result of dynamic parameters lacks verification artifice,an inverse calculation approach of dynamic parameters based on the dynamic stress detection of tooth root and optimization theory is proposed.After that,the change discipline of tooth root dynamic stress under the cracked gear is researched by establishing the strain measurement experiment rig of gear failure.Based on the established relationship between the dynamic strain of the tooth root and the dynamic parameters(Friction factor,TVMS),the optimization theory is used to modify the parameter values which are related the experimental measured strain of tooth root and the FEM calculated strain,so as to obtain the dynamic parameters of the experiment rig of gear failure.The research shows that the feasibility of calculating dynamic parameters from experiment data is confirmed by comparing results of proposed method with the traditional method of dynamic parameters.Besides,aiming to cracked gear fault and gear surface pitting fault,the dynamic behaviors of gear failure mechanism is studied by using dynamic parameters of inverse calculation method.4.To resolve the problem that the conclusions obtained from the research of fault mechanism cannot be effectively applied to diagnose the gear fault,a quantitative diagnosis method for the gear cracks,which combines dynamic modeling of gear failure and signal processing method,is proposed.And then the feature extraction method of gear vibration signal based on improved local-oscillatory characteristic decomposition(LOD)is studied.By combining feature extraction method and the support vector regression(SVR)method which is used to construct the quantitative relationship between the failure feature values of dynamic model and the gear crack levels,the quantitative diagnosis of gear crack level can be realized.Aiming at the problem of component selection in adaptive and sparsest time-frequency analysis(ASTFA)method,the principle mode analysis(PMA)method is introduced,and a quantitative diagnosis method of cracked gear based on the PMA-ASTFA is presented.Aiming at the modulation characteristics of gear vibration signal,the minimum entropy deconvolution(MED)is assisted to extract the vibration feature of cracked gear and a quantitative diagnosis method of cracked gear based on the MED-ASTFA method is proposed.After all,the analysis results validate that the approach based on the dynamic modeling of gear failure and signal processing method is effective in quantitative diagnosis for the gear root crack.