Self-Identification and Balancing for AMB-Rotor System of Turbomolecular Pumps

The rotor unbalance of the magnetic levitated turbomolecular pump (MLTMP) can lead to undesirable rotor synchronous vibration, and thus seriously affect the instrumentation precision in the ultra-high vacuum condition. The accurate identification and elimination of rotor unbalance is the primary issue for system stable operation at high speeds. This article proposes a method for rotor unbalance self-identification without using external information, and designs a model parameter correction procedure to improve the rotor balancing precision. First, the complex coefficient method is utilized to establish the rotor dynamic model under the regular controller, and the numerical equivalence between the correction mass centrifugal force and magnetic bearings force is used to obtain the rotor unbalance. Then, the rotor synchronous responses based on rotor radial displacement and bearing control currents are extracted through the cross-correlation algorithm. To further improve the rotor unbalance measurement accuracy, the stiffness coefficient values are also corrected by introducing virtual trial weights. Finally, the experimental results conducted on a MLTMP prototype verify the effectiveness of the proposed method in the self-identification of rotor unbalance.