Computational Investigations Into Aero-Engine Bearing Chamber Off-Take Flows

Abstract

Within an aero-engine bearing chamber oil is provided to components to lubricate and cool. This oil must be efficiently removed (scavenged) from the chamber to ensure it does not overheat and degrade. Bearing chambers typically contain a sump section with an exit pipe leading to a scavenge pump. In this paper a simplified geometry of a sump section is analysed computationally. Two experimental rigs have been utilised to support this work, although these results are not reported in the present paper. Early data from the experimental work is both qualitative (visualisation), and quantitative (conductance probe measurements of film height, analysis of images); but this work is still ongoing. The results indicate that for the chosen geometry there are three separate regimes of flow dependent on inlet flow conditions. To complement the experimental studies the commercial computational fluid dynamics (CFD) code Fluent 6 has been applied to investigate typical flows found in aeroengine bearing chambers. Using the Volume of fluid (VOF) model of Hirt and Nichols, the CFD investigation has built from fundamental cases to computations for the experimental test rig configurations. This is the focus of the present paper. The first CFD case presented is a cylindrical vessel drained through a centrally placed cylindrical pipe. The geometry was chosen to match that of a published experimental study (Lubin and Springer [1]) and CFD data compares well with the available data. A full 3D model was required and a hexahedral mesh was implemented as it was found to provide a better discretisation. Results presented show that the volume of fluid approach is capable of modelling draining flows. Building on this, a three-dimensional CFD investigation into two phase flow in a double intake scavenge system similar to those found in aero-engines has been conducted. The paper shows that CFD is a promising approach to analysing bearing chamber scavenge flows and further, indicates methods of characterising behaviour.