Identification Of Supporting Characteristics Of Flexible Rotor AMBs System

As a typical mechatronic system,active magnetic bearings(AMB)possess several advantages such as no mechanical contact,no mechanical wear,long life,no lubrication,high efficiency and low noise.Therefore,AMBs have been increasingly applied in high-speed rotating machineries.Super high and over bending critical rotating speed are the future directions for rotor AMBs.However,with the increasing rotating speed approaching or even passing bending critical speed,flexible rotor property is more obvious and the system dynamics is more difficult to analyze.Identify the AMBs supporting coefficients is the basis for system dynamics analysis.However,existing identification method is only valid when the rotating speed is far below the first bending critical speed and the identification error will increase with the increasing speed.New identification method based on flexible rotor model is the key to this problem,but the mapping relationship between test responses and supporting coefficients is not clear.To solve the problem,this dissertation focuses on the AMBs supporting characteristics and performs the research in the following four aspects including rotor AMBs mechatronic modeling,influence factors for supporting coefficients,identification method for flexible rotor AMBs system and experimental identification.Firstly,the rotor AMBs mechatronic modeling method is proposed.A model updating method is investigated to solve the shrink effects on rotor,which result the discrepancy between theoretical calculation and modal experiment,and an accurate rotor model is obtained.For the rotor AMBs,the main vibration energy of the rotor is the low-frequency vibration,and the energy of high-frequency vibration attenuates with increasing frequency.Therefore,modal truncation is employed for reduction and a low-order state space equation for mechanical system with AMBs is obtained.The electrical system models are created following frequency test data.Combing the mechanical and electrical models,the rotor AMBs mechatronic model is established.Secondly,the influence factors for supporting coefficients are analyzed.The equivalent stiffness and damping expression is compared for single and multi-degrees AMBs system and the equivalent stiffness and damping expression under PID controller are deduced.The influence of the mechanical parameters including the pole area,the number of turns and the air gap are analyzed.The influence of electronic control parameters including controller,sensor,power amplifier parameters and bias current are analyzed.Thirdly,an identification method for flexible rotor AMBs system is proposed.Aimed at the problem of the theoretical equivalent stiffness damping identification method ignoring the gyroscopic effect,hysteresis,eddy current loss and limited to the static state,the identification method based on the physical stiffness and damping definition is studied.It is found that the identification method based on the rigid rotor model is only valid for rigid rotor below the first bending critical speed.To solve the limitation of the existing rigid rotor model identification method,a new identification approach based on the flexible rotor AMBs model is proposed and the mapping relationship between the parameters to be identified and the experimental responses is confirmed.The proposed method is simulated and the results show that it is not only valid for rigid rotor below the first bending critical speed,but also valid for flexible rotor over the first bending critical speed.It can accurate obtain the bearing direct and cross coefficients considering residual unbalances.Finally,with the proposed flexible rotor AMBs identification method,the experimental identification for flexible rotor AMBs is performed.A digital controller is developed.The difficulties for bending modal vibration control are discussed.With phase compensation,the rotor is successfully run over the first bending critical speed.Using two sets independent experimental unbalance response data,the stiffness and damping coefficients ranging from rigid rotor to flexible rotor speed are identified accurately considering residual unbalances.The influence of noise is analyzed.The results show that the identification method proposed in this paper can identify the stiffness and damping values effectively even with 10% noise.The difference between the theoretical equivalent stiffness damping values and the experimental identified values is compared.It is found that theoretical equivalent stiffness is in consistent with experimental results,but theoretical equivalent damping trend deviates from experimental results.The reason for the damping discrepancy is that the definition of equivalent damping ignores the loss on the rotor.The identified stiffness and damping results are verified and the results show that the proposed algorithm in this paper can accurately identify the AMBs stiffness damping coefficients ranging from rigid rotor to flexible rotor speed.