In the drive for greater increases in fuel efficiency and reductions in CO2 emissions from aero engines, an epicyclic reduction gearbox can be used to break the link between the turbine and fan, enabling the engine to run at a higher bypass ratio. However, even small power losses can generate significant amounts of heat, due to the high loads transmitted from the gearbox. A substantial amount of cooling is required to remove this heat and a large part of this is supplied directly to the gear face. Assessing the performance of coolants and minimising the buildup of oil in the system is therefore a critical stage in the design process. Traditionally, finite volume CFD methods have been used to compute flow and heat transfer solutions. More recently, Lagrangian methods such as Smoothed Particle Hydrodynamics (SPH) have also been applied. The Lattice Boltzman Method (LBM) is a mesoscopic particle based method which uses statistical properties of particles based at each point of a lattice to calculate flow properties. This is a fully transient method and allows for a simple and efficient derivation of LES turbulence properties. In this work the Lattice Bolztman Method is used to investigate the impingement of an oil jet on a rotating spur gear. A comparison of LBM simulations is made against published work using other methods such as SPH and CFD — utilising the Volume of Fluid method — as well as a qualitative comparison with published experimental high speed images. These all show an excellent agreement and the simulations take the same order of magnitude of computational power as 3D single phase SPH, but are fully multiphase and have LES turbulence. This method is then used to investigate how changes to the oil feed delivery rate affect the spreading of the oil jet on the gear tooth and the splashing profiles. The potential for applying this method to other scenarios, such as lubricating and cooling meshing gears, is also discussed.

This content is only available via PDF.
You do not currently have access to this content.