Oil removal (scavenge) from aeroengine bearing chambers is an ongoing challenge for aeroengine designers. Effective scavenging of oil is necessary to avoid excessive heat generation, degradation of oil properties and deterioration in heat transfer functionality. However the task of oil removal is not trivial. Oil is entrained in a highly rotating environment induced by the rotating shafts. Simply removing a larger volume of fluid from the chamber with the scavenge pump can create higher air flowrates but lead to oil becoming trapped in the chamber and so no reduction in residence volume (the amount of oil present in the chamber during operation).
The University of Nottingham Technology Centre in Gas Turbine Transmissions Systems has been conducting experiments investigating two phase behaviour within a simplified aeroengine bearing chamber operating at ambient pressures with water as the working liquid. The rig is constructed from polycarbonate enabling good visual access. In the chamber offtake region a number of behaviours can be observed relating to the hydraulic uplift and general flow behaviour as gas and liquid exit the chamber. Chandra et al  reports a parametric study using a design of experiments approach into geometrical variants of a shallow offtake region defined by curved approaches and a small offtake volume. Phenomenlogical factors were quantified and used to identify the best performing geometry. Previous work at the UTC [2, 3] has used residence volume as the primary comparative performance parameter.
In this paper residence volume data is obtained for two sumps for which phenomenological data exists. The paper compares performance on the basis of these visual factors with performance on the basis of residence volume and concludes that although both frameworks have value, they do not lead to identical conclusions for all operating conditions. In film dominated cases significant hydraulic uplift usually corresponds to larger residence volume but for droplet dominated cases this is not necessarily so.