Rotor Balancing for Magnetically Levitated TMPs Integrated with Vibration Self-Sensing of Magnetic Bearings

Research output: Contribution to journalArticlepeer-review

Abstract

Due to the rotor unbalance of the magnetically levitated turbo molecular pump (TMP), this article presents a novel high-precision field balancing method based on regular control mode without trial weight. First, the static/dynamic unbalance vector expressions of the magnetically levitated rotor are analyzed and derived in polar coordinates frame. Second, according to the synchronous currents and displacement responses of the rotor at a fixed speed, the correction weight can be solved by the double-plane correction method based on equivalence principle of the magnetic forces. Then, in order to reduce the deviation between theoretical and actual value of the magnetic bearing parameters, this article proposes a second correction method by correcting the conversion coefficient matrix from online identification procedures. The proposed method, which can eliminate the errors introduced by the oversimplification of dynamic model in regular control, is especially suitable for the magnetically levitated rotor system with strong gyroscopic effects and asymmetric structure. Finally, the experimental results on a magnetically levitated TMP prototype validate the effectiveness of the proposed method. Compared with the original unbalance of rotor, the synchronous current amplitude declined by 99.0%, and the synchronous displacement amplitude also declined by 96.1% with the proposed method. A high-precision field balancing has been achieved.

Original languageEnglish
Pages (from-to)3031-3039
Number of pages9
JournalIEEE/ASME Transactions on Mechatronics
Volume26
Issue number6
DOIs
StatePublished - 1 Dec 2021

Keywords

  • Field balancing method
  • magnetic bearing
  • rotor

Fingerprint

Dive into the research topics of 'Rotor Balancing for Magnetically Levitated TMPs Integrated with Vibration Self-Sensing of Magnetic Bearings'. Together they form a unique fingerprint.

Cite this