Key Technologies Of Monitoring And Fault Early Warning For Gas Turbine Machinery

The gas turbine is widely used in industrial, aviation, ships, and itsmaintenance costs are quite high. For the complexity of the structure and poorworking conditions，they are fault sensitive parts of the whole equipmentsystem. Monitoring is necessary to grasp the operating status and functionalcharacteristics of the mechanical system, has an important role in thefoundation to ensure the safe operation of equipment to avoid accidents,reduce downtime maintenance time and costs, and improve their operationalreliability. Through the use of condition monitoring systems, regularmaintenance of equipment maintenance are instead of maintenance based onits running condition. This paper research the key technologies of the earlyfault warning based on aero-engine bearing. The main content andcontributions are as follows:First, Design double-rotor aircraft engine test-bed to study the faultsignaling transmission in complex path. Establishment the experimentaldevice combined with the existing aero-engine rotor system for investigate theweak fault signal transmission in complex path, and fault diagnosis of aircraftengine bearing under different speed. And make the modal analysis of this device.Second, establishment the rotor model using the rotor dynamics softwareDyrobes to making dynamics analysis, analyze its critical speed and rotorstability. The results show that this experimental design can meet the need forbearing fault diagnosis under high speed. Investigate weak fault signaltransmission based on the stress wave. Simulation the test-bed using dynamicanalysis software LS-DYNA to analysis testability of the weak signal. Laid thefoundation for future fault diagnosis based on physical model-based.Third, making experiment on bearing fault diagnosis. Set differentsampling frequency to study the characteristics of bearing fault in differentsampling frequency to realize the earlier fault diagnosis. The results show thatwith the sampling frequency increases, monitoring frequency expand, andrecognition of early fault signal become more accurate and comprehensive.This paper make numerical simulation combined with experiment, aim atachieving a combination fault diagnosis approach based on signalcharacteristics and physical model for early and quantitative bearing faultdiagnosis, and laid the foundation for the further implementation of the lifeprediction as well as health management of the bearing.