Surface texturing for hydrodynamic bearings

Abstract

In the present thesis the concept of surface texturing, i.e. the intentional introduction of surface features, is investigated as a means of enhancing the tribological performance of hydrodynamic bearings. An in-depth literature review is conducted, outlining the research effort worldwide, analysing the current understandings on how textures can improve the bearing performance, discussing recommended texture geometries and providing a comparative summary of state of the art modelling techniques to study textured surfaces under hydrodynamic conditions. Based on the findings of this literature review, a fast and robust numerical model is developed that allows to study and mathematically optimize texture patterns for tilting pad thrust bearings under a wide range of conditions. The model is based on a non-uniform finite volume discretization of the Reynolds partial differential equation and includes a mass-conserving cavitation algorithm. Meshes are adaptive and film discontinuities are directly incorporated in the discrete equations to improve accuracy and computational speed. A Gauss-Seidel method with successive relaxation is used to solve the discrete system. A novel bearing equilibrium solver is developed, based on a strategic combination of the Newton-Raphson method, Broyden’s method with Sherman-Morrison formula and a continuation approach with fourth-order RungeKutta technique, where results from the equivalent untextured bearing are used to speed up the computation. Thermal effects are considered through an effective temperature method, while considering the hot-oil-carry-over effect. The model takes advantage of multiple processor cores and allows the execution of arbitrary parametric studies as well as the optimization of texture designs by incorporating an interior-point algorithm. The model is validated by comparison with literature and applied to optimize surface textures for tilting pad thrust bearings under various conditions. Moreover, experiments are carried out on a purposely developed thrust bearing test rig to study the influence of surface texturing on the performance of tilting pad thrust bearings and validate the developed numerical model.

More information

Identifiers

ORCID for Ling Wang: orcid.org/0000-0002-2894-6784

Catalogue record

Export record