HHT Spectrum Analysis And Fault Feature Extraction Of Breathing Cracked Rotor-Bearing System

Rotary machinery is widely used in many fields.The rotor-bearing system is the core component of rotating machinery and is also a key research subject of rotor dynamics.Rotor crack fault is a common failure of rotor-bearing system.It is very necessary to study the dynamic characteristics and extract fault features.In this paper,the breathing cracked rotor-bearing system is modeled as a whole and numerical simulations are performed to analyze the dynamic characteristics of the system.For traditional dynamic nonlinear analysis methods,it is difficult to identify crack faults under low-speed and small cracks,and the HHT analysis method is introduced to extract the fault feature quantities,so as to achieve effective diagnosis of faults at low speed and small cracks.The main works of this article are:(1)Summarized the mathematical model of the current respiratory crack failure and the model is optimized based on the neutral axis migration method.The research shows that compared with the classic crack model,the offset neutral axis model has little difference with the integrated model when the crack angle is small,and the initial closure angle has a larger offset due to the neutral axis than the integrated model.This feature fits the actual situation more.When the crack angle is large,the comprehensive model is similar to the cosine model,but the crack full range of the neutral axis model is obviously increased,which is in line with the actual situation.(2)The numerical solution of the breathing-cracked-rotor-bearing system dynamic equation is made.The main nonlinear dynamic analysis of the system is done.The results show that the subcritical resonance of the cracked rotor at the rotational speed of 1/3 and 1/2 times the critical rotational speed disappears after the bearings affected by the nonlinear oil film force are added to both ends of the cracked rotor.The bifurcation diagram and the waterfall diagram of the system show complex dynamic characteristics at high rotational speeds.The frequency components increase,and the multi-periodic movement alternates with the pseudo-periodic movement.(3)The EMD decomposition of the system model at different crack depths was performed,and the first three high frequency IMF components were analyzed by HHT and marginal spectrum analysis.The calculation results show that with the increase of cracks,the instantaneous frequency domain of the first two high-frequency IMF1 and IMF2 components decomposed by the system response increases,the frequency component increases,the amplitude remains basically unchanged.The frequency of the third highest IMF3 component remained stable over the entire time domain,and the magnitude of the amplitude change was large,and the fluctuation degree increased with time.It can be seen that the impact of the crack on the system is mainly reflected in the low frequency components extracted by the HHT method.(4)The system response extracted by the EMD method to the IMF components and constructs different matrix to calculate singular values as the crack fault feature quantity.Add Gaussian white noise to the original signal to simulate the actual signal.The research shows that the singular value of the matrix constructed by Volterra prediction series method has good sensitivity and stability to crack faults at subcritical speed and small cracks.