Numerical analysis of spiral grooved opposed-hemisphere gas bearings: A parametric study

Abstract

Although the spiral grooved opposed-hemisphere gas bearings have been widely applied in supporting the spin axis of inertial grade gyroscopes, the accurate prediction of the performance characteristics of the bearings is still very difficult because of their structural complexity. In this paper, the static characteristics of the bearings considering the coupling effects between the journal and two hemisphere bearings are obtained theoretically. The Reynolds equations modeling the journal and hemisphere bearings are solved simultaneously by the finite element method satisfying the continuous boundary conditions between these three bearings. The fluctuations of the bearing load and reaction torque caused by rotating grooves are shown and can be neglected at large number of grooves. The results show that neglecting the coupling effects between the journal and two hemisphere bearings will result in large discrepancies at high rotational speeds and large eccentricity ratios. And the hydrodynamic effect can be divided into two regions for spiral grooved gas bearings: the groove effect domain region and the converged wedge effect domain region. A parametric study provides efficient guidance for the optimal design of the opposed-hemisphere gas bearing.

Chaotic Behaviour Investigation of a Front Opposed-Hemispherical Spiral-Grooved Air Bearing System

“…These results proved useful for this present study. In 2015, Guangwei et al [19,20] studied the spiralgrooved opposed-hemisphere gas bearing system with a rigid rotor and focused particular attention on its whirl motion. They used finite element method combined with the Finite Difference Method to solve the time-dependent Reynolds equation and analyze the complicated dynamic behaviour of the rotor-bearing system by phase portraits, power spectra, Poincare maps, and bifurcation diagrams obtained from the numerical procedure.…”

“…Next, the Differential Transformation Method is applied to the time domain in (19) to produce discrete time intervals, and the central difference scheme of the Finite Difference Method is then applied to the coordinates of the locations. The following can then be derived:…”