Session: 10-08-01: Multiphase Flows and Applications

Paper Number: 150382

150382 - Cfd-Based Analysis of Oil-Mist Lubrication in High-Speed Turbomachinery Bearings 

Optimal utilization of oil for lubrication and energy-loss mitigation in high-speed bearings is critical to the operation of many turbomachines and other devices of engineering relevance.  As a result, it has received substantial attention in the past, mostly in the form of experimental and reduced order modeling studies.  Detailed flow analysis using CFD has been used relatively recently towards this end with encouraging results.  In this presentation we describe such an analysis focused on the introduction of oil, as a mist or spray into a high speed, deep groove ball bearing typical to aerospace applications. The CFD model is developed in Ansys Fluent. The detailed geometry is representative of a class of bearings ubiquitous in the field of turbomachinery and is based on actual flight hardware with more than 1 million hours of operation.  The model also includes all the details of the motion of the geometry which is derived from detailed kinematics analysis.  This includes the definition of the contact angle of the bearing which leads to ball rotational axes that are not collinear with the shaft rotation axis and which lead to a natural pumping action of the bearing. The air flow is assumed to be turbulent, and it is simulated by solving the Unsteady Reynolds-Averaged Navier Stokes (URANS) equations with a k-w turbulence model.  This turbulence model is particularly well suited for rotating flows such as the one envisioned here.  Oil flow is considered under the constraint of small flowrates (<100ml/hr). This oil flow is simulated using Lagrangian approaches.  Specifically, the oil dispersed in the air is assumed to be in the form of droplets captured using the Dispersed Phase Method (DPM).  The liquid covering the walls is simulated by the Lagrangian Film Model (LFM) which is an extension of DPM that uses Lagrangian particles originally used in DPM to discretize and solve the evolution of liquid films on surfaces once the oil hits the surface.  The LFM involves a number of submodels, e.g. for splashing and stripping of droplets.  These will be described in more detail in the full paper.  So will the parameters involved the DPM and k-w models. The approximately 1 million cell grid consists of tetrahedral elements and is properly adapted to the geometry details. The number of Largrangian particles varies strongly with simulation case with up to 1 million particles being simulated. Oil is introduced to the bearing as a spray following a Rosin-Rammler droplet distribution injected at various locations on the air intake side of the bearing.  Analysis is focused on the efficient utilization of the oil, identifying optimal coverage of desirable components, and limiting coverage of non-critical components or loss of oil to the environment.   The sensitivity of these oil flow characteristics to oil droplet size and injection velocity and location is quantified.  Preliminary results indicate that while all parameters play important roles in the outcome, the location of injection is by far the most critical parameter.

Presenting Author: Todd Currier Zulu Pods Inc

Presenting Author Biography: Dr. Currier is Chief Scientific Officer at Zulu Pods Inc, where he oversees the research and development of technologies and products. He is a recognized expert in fluid structure interaction. Past appointments include that of senior researcher at Johns Hopkins University Applied Physics Laboratory, where he developed first principles-based models of hypersonic missiles, and that of lead of Pratt and Whitney’s advanced engine development programs. He received his PhD from the University of Massachusetts, Amherst and is Chateaubriand Fellow for which he conducted research at École Polytechnique in France

Authors:

Marios Soteriou Zulu Pods Inc
Todd Currier Zulu Pods Inc