International Journal of Heat and Mass Transfer, Vol.127, 917-932, 2018
Numerical simulations of two-phase flow in open-cell metal foams with application to aero-engine separators
Oil/air separation is a critical process in aero engines since a two-phase flow is created in bearing chambers with closed-loop oil systems. This work describes an application of a two-phase pore-scale modeling method to open-cell metal foams, which are used in secondary oil separation in aero-engine separators. The dispersed oil-phase flow through open-cell foams was studied using a Lagrangian approach: discrete phase model (DPM) and investigated on foam structures of pore densities 10-50 ppi and porosities 75, 85 and 95%. The simulations were performed over flow velocities ranging from 5 m/s to 50 m/s, and for separate uniform droplets of diameters ranging from 0.5 to 15 mu m, with the commercial CFD analysis package Ansys Fluent. Results indicated that the fraction of non-captured droplets decreased when the flow velocity rose, and the higher the foam porosity, the higher the number of non-captured oil droplets. The flow velocity had a significant impact on the oil capture for intermediate droplet diameters of 1-3 mu m. The fraction of non-captured oil droplets increased as the pore density of the open-cell foam increased at the normalized position of x/d(p) = 5. Foams with a larger pore size and smaller specific surface area showed an increased oil capture when compared with foams with smaller pore sizes and larger specific surface area. For Stokes numbers below 0.1, the oil capture efficiency was small and remained rather constant independent of the Stokes number. However, for 0.1 < Stokes numbers < 1.0, the oil capture efficiency increased substantially with increasing Stokes numbers. Stokes number values above 1.0 were associated with a high capture efficiency, which remained stable with increasing Stokes numbers. (C) 2018 Elsevier Ltd. All rights reserved.
Keywords:Open-cell foams;Two-phase flow;Aero-engine separators;Eulerian-Lagrangian method;DPM simulations