Modeling the Flow and Transport Dynamics in Gasoline Particulate Filters to Improve Filtration Efficiency

We propose a new pore-scale/channel model, or hybrid model, for the fluid flow and particulate transport in gasoline particulate filters (GPFs). GPFs are emission control devices aimed at removing particulate out of the exhaust system of a gasoline direct injection engine. In this study, we consider a wall-flow uncoated GPF, which is made of a bundle of inlet and outlet channels separated by porous walls. The particulate-filled exhaust gas flows into the inlet channels, and passes through the porous walls before exiting out of the outlet channels. We model the flow inside the inlet and outlet channels using the incompressible Navier-Stokes equation coupled with the spatially averaged Navier-Stokes equation for the flow inside the porous walls. For the particulate transport, the coupled advection and spatially averaged advection-reaction equations are used, where the reaction term models the particulate accumulation. Using OpenFOAM, we numerically solve the flow and the transport equations and show that the concentration of deposited particles is nonuniformly distributed along the filter length, with an increase of concentration at the back end of the filter as Reynolds number increases. Images from X-ray computed tomography (XCT)-scanning experiments of the soot-loaded filter show that such a nonuniform distribution is consistent with the prediction obtained from the model. Finally, we show how the proposed model can be employed to optimize the filter design to improve filtration efficiency.