Research On The Fault Of Aero-engine Inter-shaft Bearing

Advances in science and technology have greatly shortened the distance between people.The aero-engine is the core part of the aircraft.Dual rotors are the foundation of aero-engines.The inter-shaft bearing is in an important position and is easily exposed to extreme conditions such as high temperature and high speed,and is easily in a fault state.The research on its fault mechanism and diagnosis problems is particularly important.Therefore,this paper uses the dual-rotor system of an aero-engine as a model to study the fault of the inter-shaft bearing.The main research contents are as follows:(1)The dynamics model of the aero-engine dual-rotor system is carried out,and the differential equation of motion of the system is obtained through the energy method and the Lagrange equation.Through the Hertz contact theory and appropriate assumptions,mathematical expressions are used to describe the impact force on the rolling elements at the defects,and the finite element method is used to determine the contact stiffness coefficient between the rolling elements and the raceway.(2)In order to better describe the problem of different sizes of bearing fault defects under actual working conditions,uncertainty is introduced into the dynamic model.The defect size is set as a random variable,and the dynamic characteristics of the system are reflected by the root mean square value of the statistical vibration response.The non-intrusive polynomial chaos expansion method(NIPCE)is introduced to deal with the uncertainty of the problem,and the coefficients of the NIPCE model are obtained by using the stochastic response surface method to complete the establishment of the NIPCE model.(3)Differential equations of motion obtained through Lagrange equations are solved and deterministic analysis is performed.Amplitude-frequency characteristic curves in different directions are obtained.Parameter analysis,including the impact of defect size,rotor eccentricity,damping of the support bearing,the stiffness of the support bearing and the radial clearance of the intermediate bearing,on the dynamic characteristics of the system are carried out,and summarize the law of the amplitude-frequency characteristic curves.(4)Solve the model that introduces the randomness problem,and obtain the probability density function curve of the output random variable through the NIPCE method.This method is compared with Monte Carlo Simulation(MCS)to show the effectiveness of the method.Parameter analysis was carried out to analyze the influence of the mean value and standard deviation of the defect size,the eccentricity of the low-pressure and high-pressure rotor,and the change of the radial clearance of the inter-shaft bearing on the statistical characteristics.(5)Based on Convolutional Neural Network(CNN),combined with the basic methods of deep learning,MSCNN is built to identify and detect fault features.After comparing with CNN,BPNN,SVM and other methods,the accuracy of MSCNN shows its superiority.