Nonlinear Dynamic Analysis of a Rotor System with Aerodynamic Journal Bearings

碩士 === 國立中正大學 === 機械工程所 === 95 === With the development of technology, high rotational speed machines become the mainstream of the mechanical design recently. The aerodynamic bearing is regarded as an ideal bearing which has many superior characteristics than the conventional bearing. The characteri...

Full description

Bibliographic Details
Main Authors: Bo- Wen Deng, 鄧博文
Other Authors: Ting-Nung Shiau
Format: Others
Language:en_US
Published: 2006
Online Access:http://ndltd.ncl.edu.tw/handle/06103251085406505940
Description
Summary:碩士 === 國立中正大學 === 機械工程所 === 95 === With the development of technology, high rotational speed machines become the mainstream of the mechanical design recently. The aerodynamic bearing is regarded as an ideal bearing which has many superior characteristics than the conventional bearing. The characteristics include no waste of contact, no need for lubrication, capability of working on unusual environment, long life of working and so forth. This study is focused on the aerodynamic bearing-rotor system characteristics analysis and bearing parameter identification. In the analysis of system characteristics, Finite Difference Method (FDM) is employed to solve the Reynolds equation, which is used to calculate the nonlinear compressible gas force of the gas bearing. By applying the gas bearing force to system equations of motion, the system response can be determined by the numerical integration method. Using the Recursive Least Square Method and the Kalman Filter Method, bearing parameters can be identified from the system response. The parameters include the damping and stiffness coefficients of the aerodynamic bearing. According to the results of identification, both parameters identified by the Recursive Least Square Method and the Kalman Filter Method are in good agreement. The identified system responses are very close to those from numerical integration. The results show that the aerodynamic bearing force can be precisely identified and the system response can be quickly solved with less computer time.