Optimal bearing configuration selection for power generation shaft-trains: A linear and nonlinear dynamics approach

Abstract

This paper proposes a straightforward procedure for defining bearing design configurations in turbine-generator shaft trains. The bearing design inputs specify the bearing type and the pad configuration. The design outputs focus on stability, bearing integrity, and operability of the shaft-train system. The design output is evaluated in two ways: a) linear harmonic analysis utilizing linearized stiffness and damping coefficients for the bearing impedance forces, and b) nonlinear analysis, where the bearing forces are modeled as nonlinear functions of bearing and pedestal kinematics; the response is evaluated by collocation-type method coupled with numerical continuation. Thermohydrodynamic lubrication (THD lubrication) with turbulence correction is considered in the bearing lubrication model.

The results show that all constraints are satisfied, and the optimal bearing configurations include preload and offset, while no specific trend is observed for specific loads. Laminar oil flow is prompted by the optimization through specific bearing diameters. Linear and nonlinear dynamic models do not render identical optimal designs. Linear model tends to be conservative in the design output, while nonlinear dynamic model provides more accurate predictions, accounting for any whirling orbit shape. The results emphasize the necessity of incorporating nonlinear dynamics into standard rotor dynamic calculations for this type of machines.