Test Method And Identification Method Of Stiffness And Damping Of Magnetic Bearings

The parameters of magnetic bearing are the basis of control parameters optimization, criticalspeed prediction, unbalance response calculation and the optimal design of dynamics, the parametersare also the preconditions of the guarantee of magnetic bearing rotor system’s operation stability. Theaim of the paper is to research the test methods of dynamic magnetic bearing parameters. Severaldifferent definitions of dynamic magnetic bearing parameters are introduced and the research focusedon the bearing parameters in linear dynamics system.Accurate magnetic bearing-rotor system model is the basis for bearing parameters identificationand dynamics response prediction. The differential eauqtions of motion of rigid disc, elastic axis andmagnetic bearings are established by mechanical analysis. The material parameters of rotor model areupdated by two different kinds of model updating method-model updating based on modal parametersand model updating based frequency response function. Through the analysis and comparison ofupdated parameters, the more accurate rotor model is selected.With the unbalance response of the system, the transfer matrix model updating method and thefinite element model updating method are studied to identify the dynamic magnetic bearingparameters. Unbalance response of magnetic bearing rotor system with assumed initial supportparameters can be calculated by the transfer matrix method. Based on an optimization algorithm, theminimum error between the calculated and measured unbalance response is achieved. The supportparameters which were used to calculate the unbalance response are the stiffness and damping ofmagnetic bearings which are identified by the transfer matrix model updating. The first procedure offinite element model updating method is the establishment of response surface function. With theresponse surface function, the parameters which make the minimum between response surfacefunction and measured unbalance response are obtained by the GlobalSearch method. The results ofsimulation verify the effectiveness of the updating methods, and the transfer matrix model updatingmethod can get more accurate magnetic bearing parameters than the finite element model updatingmethod.The test platform of magnetic bearing rotor system is built. After the known unbalance is addedto the experiment rig, the corresponding unbalance response signal and speed signal can be measured.The amplitude and phase features can be obtain after the measured data is Kasier band-pass filtered, zero phase filtered and Fourier series fitted. Based on these features, the dynamic parameters ofmagnetic bearings are identified by the model updating method. The finite element model based onthe identified bearing parameters are established to verify the accuracy of the model updating method.The results of simulation verify the accuracy of identified parameters.In addition, a test method based on displacement signal and control current signal is proposed.The test method is based on the linear hypothesis of electromagnetic force and the ideal model of PIDcontrol theory. The dynamic bearing parameters are tested by the method therough displacementamplitude, control current amplitude and phase difference features which are extracted fromexperiment data. With the contrast between identified parameters by model updating method and thetested parameters, the magnetic bearing parameters obtained by two different methods are consistent.