Gas Turbine Transmission Systems Research Articles

Scavenge Performance of an Optimized Shallow Sump at Various Flow Conditions

Abstract Oil scavenge flow in aero-engine bearing chamber remains largely a challenging problem for many engine designers. Research campaign on scavenge flow has been conducted by G2TRC—Gas Turbine and Transmissions Research Centre (previously Rolls Royce University Technology Centre in Gas Turbine Transmission Systems) at the University of Nottingham. It was recognized that a deep sump performs better than shallower one due to its ability to “shield” the collected oil in the sump from the shaft windage, thus reducing the amount of oil being picked up by the bulk air rotation. However, such a deep sump design cannot be employed in some engines and especially at certain locations where space is limited. A parametric study combined with phenomenological approach on shallow sump geometry has been conducted and presented in the previous publication, where a certain optimized shallow sump variant was proposed depending on whether the flow in the chamber is wall film dominated or airborne droplets dominated. The parametric phenomenological approach was employed since it can be done relatively quicker than typical data gathering through an experiment. However, the approach relies on qualitative interpretation of the flow features, and its application in bearing chamber flow research has never been validated before. This paper presents the results of quantitative measurements of residence volumes of an optimized shallow sump variant identified in the parametric phenomenological study. Comparison was then made with the residence volumes of some existing engine sumps. It was found that the optimized shallow sump for wall film dominated flow has lower residence volumes compared to some existing engine sumps. In some cases, the residence volume can be reduced by up to 75%. An optimized shallow sump variant for airborne droplets dominated was also identified in the previous parametric phenomenological study, although the residence volume measurement is yet to be conducted. The optimized shallow sump for wall film dominated flow was also identified as a good sump regardless of the flow regime. However, when it was tested in airborne droplets dominated flow, its residence volumes are higher than some of the existing engine sumps. This highlights the importance of considering the flow regime in the bearing chamber in any attempt to optimize a sump geometry.

Study of Gas/Liquid Behavior Within an Aeroengine Bearing Chamber

Aeroengine bearing chambers typically contain bearings, seals, shafts and static parts. Oil is introduced for lubrication and cooling and this creates a two phase flow environment that may contain droplets, mist, film, ligaments, froth or foam and liquid pools. Some regions of the chamber contain a highly rotating air flow such that there are zones where the flow is gravity dominated and zones where it is rotation dominated. The University of Nottingham Technology Centre in Gas Turbine Transmission Systems, is conducting an ongoing experimental program investigating liquid and gas flow behavior in a relevant highly rotating environment. Previously reported work by the UTC has investigated film thickness and residence volume within a simplified chamber consisting of outer cylindrical chamber, inner rotating shaft and cuboid off-take geometry (termed the generic deep sump). Recently, a more aeroengine relevant bearing chamber offtake geometry has been studied. This geometry is similar to one investigated at Purdue University and consists of a “sub-sump” region approached by curved surfaces linked to the bearing chamber. The test chamber consists of an outer, stationary cylinder with an inner rotating shaft. The rig runs at ambient pressure and the working fluid (water) is introduced either via a film generator on the chamber wall or through holes in the shaft. In addition to visual data (high speed and normal video), liquid residence volume within the chamber and film thickness were the two numerical comparators chosen. Data was obtained for a number of liquid supply rates, scavenge ratios and shaft rotation speeds. The data from the current model is compared to that from the earlier studies. The data shows that in contrast to the previously reported generic deep sump study, the residence volume of the curved wall deep sump (CWDS) designs is far less sensitive to shaft speed, liquid supply rate and scavenge ratio. The method of liquid supply only makes a significant difference at the lowest scavenge ratios. Residence volume data for the Nottingham CWDS is comparable, when appropriately scaled, to that for the Purdue design. The film thickness data shows that at the lower shaft speeds investigated the flow is gravity dominated whereas at higher shaft speeds shear dominates.

Effect of Pressure and Oil Mist on Windage Power Loss of a Shrouded Spiral Bevel Gear

In many aeroengines the accessory power offtake is achieved using a spiral bevel gear set running off one of the main shafts. The crown and bevel gears are housed in an internal gearbox and there is significant heat generation within this chamber, some of which is attributed to windage power loss (WPL) generated by the gear. Over the past few years the University of Nottingham Technology Centre (UTC) in Gas Turbine Transmissions has been researching spiral bevel gear windage power loss, both computationally and experimentally, using a purpose-built test rig at the UTC. In this study the test rig has been adapted such that chamber pressures up to 8 bar can be generated. Test data has been obtained that shows the effect on the WPL of chamber pressure, advancing understanding of the relationship between data obtained at ambient and pressurized conditions. Three configurations have been studied: unshrouded gear, a gear with 360 deg shroud and a shrouded crown and pinion meshing pair. Furthermore, the effect of oil mist within the chamber on the WPL has been studied. An oil mist generation system was developed for introducing a fine mist into the chamber and results are presented for varying mist flowrates. A mist measurement system was developed to sample mist mass fraction within the chamber and the data obtained is used to calculate an effective (oil/gas mixture) chamber density. Increasing chamber pressure increases the Reynolds number, moving the system behavior further along the moment coefficient-Reynolds number correlation. The Cm-Re correlation is similar in form to that for a shrouded cone, showing transitional behavior around Re = 2 × 106. Beyond this transition, Cm decreases with increasing Re. Introducing an oil spray has two effects: a reduction in chamber temperature and an increase in the effective density of the chamber fluid. Both effects can be accounted for by calculating Re and Cm based on mixture properties but it seems highly likely that the properties of the fluid under the shroud differ from those of the fluid in the external chamber.

Windage Power Losses From Spiral Bevel Gears With Varying Oil Flows and Shroud Configurations

In many aero-engines, the power to drive accessories is transmitted through high speed bevel gears in a chamber in the center of the engine. The windage power loss (WPL) associated with these gears makes a significant contribution to the overall heat generation within the chamber. Shrouding the gears provides an effective method of reducing this WPL and managing the flow of lubricating oil. Experimental and computational programs at the University of Nottingham Technology Centre in Gas Turbine Transmission Systems are providing an improved understanding of shroud performance and design. This paper presents the results from a pair of shrouded meshing gears run at representative speeds and oil flow in a rig with speed and torque measurement. A previously published study of a single bevel gear operating in air (Johnson et al., 2007, “Experimental Investigation Into Windage Power Loss From a Shrouded Spiral Bevel Gear” ASME Paper No. GT2007-27885) found a reduction in torque of up to 70% from shrouding. In this work, the addition of oil and the pinion gear did not lead to high torque due to the buildup of oil under the shrouds, but the reduction in torque due to fitting the shrouds is significantly less than was found for the same gear in air alone. In order to isolate the various parameters, further testing with a single gear was carried out. A fully (360 deg) shrouded gear shows a big improvement over an unshrouded gear when running in air alone, but much of this benefit disappears as soon as a very small amount of oil is introduced under the shroud. This implies that the oil is recirculating under the shroud. Increasing the oil flow beyond this initial level increases the torque by the amount required to accelerate the oil mass flow up to the peripheral speed of the gear. Providing a full width slot in the shroud downstream of the oil jet allows the oil to escape without any recirculation and restores much of the benefit of the shroud. Further insight into the oil behavior is obtained from torque measurements and observations through a transparent shroud and with various slot configurations. Video observation shows evidence of a vortex flow under the shroud that carries some of the oil toward the inner diameter of the gear. The three main windage contributors, air alone, recirculation of oil under the shroud, and acceleration of the feed oil, are quantified and methods for achieving the optimum design are discussed.