Fault Simulation Of Rotor-bearing System Of A Centrifugal Compressor

Centrifugal compressor is an important equipment for industrial development,in particular,it plays an important role in large-scale petrochemical enterprises.However,due to complex structure of the centrifugal compressor that is prone to a variety of failures in its work process.While the core parts of the rotor-bearing system is the main cause of excessive vibration mechanical failure,common fault types of oil whirl and oscillation,supporting loose,cracks,rubbing,etc,in addition to a variety of coupling faults,this will not only affect the life of the compressor and it will also affect the production efficiency,and therefore understanding the failure mechanism and failure characteristics of centrifugal compressors for the conduct fault diagnosis is indispensable.In this paper,based on Shenyang Blower Works,a model of the centrifugal compressor according to the compressor model structures was simplified and the model was based on rayleigh beam finite element modeling methods.A rotor-bearing system model in which the nonlinear oil film force Capone's of sliding bearing model was established.It was first sliding bearing oil film pressure distribution characteristics according to the characteristics of plain bearings,and then simulated using numerical method Newmark-? centrifugal compressor according to a common type of fault.The oil whirl and humor oscillation fault simulation model,local rubbing fault simulation model of rotor system,supporting loose simulation models and loose-rubbing coupling fault simulation model were bulit.The fault characteristics simulation failures and the coupling failure features and single fault characteristics similarities and differences were analysed.Finally,the empirical mode decomposition(EMD)in accordance with Shenyang Blower Works measured friction and centrifugal compressor rotor speed instability data obtained for each component was used and then combined with the fast fourier transform to obtain frequency components of each component fault frequencies.