Research On The Dynamic Characteristics Of The Rotor-receiver System Of The Combustion Engine Power Turbine Considering The Rolling Bearing Failure

Rolling bearings are the core supporting components of the power turbine rotor system in ship gas turbine engines,and are also the key to ensuring the stability and reliability of the rotor system and the entire machine.Compared with ordinary bearings,engine bearings bear larger loads,higher speeds,and temperatures,and serve under extremely harsh working conditions.Especially with the continuous improvement of modern ship power system performance and the continuous strengthening of engine power density,the bearing working environment has become increasingly harsh,leading to bearing failures that pose a serious threat to the safety of the rotor system and the entire machine.Since the supporting bearings are located inside the engine,their failure problems have the characteristics of strong concealment,difficult diagnosis,and difficult disassembly and maintenance.Therefore,it is very necessary and urgent to explore a convenient and feasible method for diagnosing engine bearing failures.To this end,this paper takes the easily monitored engine box vibration as the starting point,establishes a coupled dynamic analysis model of the fault bearing-rotor-engine box system,and inverses the engine bearing fault through the engine box vibration,providing important theoretical support for early fault prediction,fault diagnosis,and health assessment of engine bearings.The specific research contents are as follows:From the perspective of qualitative diagnosis,this paper establishes a rolling bearing fault feature extraction method combining ICEEMDAN and spectral kurtosis,compares the denoising performance of the traditional EMD method and the ICEEMDAN method on the original signal,and discusses the problem of selecting different bandpass filter ranges using spectral kurtosis.Taking the vibration test signal of a real fault rolling bearing as an example,the above method is used to study the feature extraction results of different fault signals,and the entire process from signal preprocessing to fault feature extraction is analyzed from the perspectives of the original signal,frequency spectrum,and envelope spectrum.By considering the coupled dynamic characteristics between the rolling bearing and the rotor system,a dynamic analysis model of the fault bearing-rotor system of the power turbine end of the engine is established,and the model is verified through experiments.Furthermore,the influence law of the bearing surface fault degree on the vibration response of the bearingrotor system is analyzed through experiments.Finally,based on this model,the dynamic characteristics of the fault bearing-rotor system of a real engine power turbine are further predicted.Taking the engine box vibration of the gas turbine engine as the starting point,and considering the serious fault problem of the rear ball bearing of a certain type of gas turbine,this paper establishes a vibration analysis model of the fault bearing-rotor-engine box of the power turbine end,and verifies the accuracy of the vibration transmission characteristics of the box model through free vibration verification with an actual-scale ANSYS model.The vibration characteristics of the power turbine box under the action of the fault bearing are analyzed.For the problem of rolling bearing faults in gas turbine engines,the health assessment and fault diagnosis of engine bearings are carried out.From the perspective of the internal relationship of vibration signals,a gas turbine power turbine rolling bearing health assessment method based on SVDD is established,revealing the relationship between the engine box vibration signal and the remaining life of the rolling bearing,forming a rolling bearing remaining life prediction curve,and obtaining a gas turbine rolling bearing health assessment and fault diagnosis software,providing theoretical guidance for early diagnosis of rolling bearings.