[Weiqiang Liu]

Hi all,

My research topic is diesel particulate filter, namely DPF which is used to remove particles in engine exhaust. I am posting this thread in order to find researchers who have the same project with me so I can ask for help or discuss with.

the filtering material of DPF is porous. therefore after some period of filtering, the porous media might be clogged and the soot inside the DPF need to be burned out. To simulate this burning process with fluent is like simulate combustion in porous media.

Before my simulation, there are some problems confuse me a lot in terms of the computation domain. I've attached pictures of the computation domain in a paper and a real product of DPF to illustrate my case.

1. Besides inlet, outlet and porous media interface, all other boundary conditions are WALL in the paper, which means all exhaust gas from the inlet flows though the filter wall. However, in real situation, only one quarter of exhaust gas flows through one particular filter wall. I am wondering does the author make the inlet velocity to be one quarter of real inlet velocity to approximate the real situation.

2. let me assume the paper plane is x-y plane and the perpendicular direction of paper plane is z axis. A two dimensional case means along the z axis all variables are the same. However, field values along z axis are different since the two opposite filtration walls exist along z axis.

Am I misunderstanding the computation domain or some simplifications of the simulation lead to the two dimensional domain?

Looking forward to your reply.

Best regards.

Weiqiang.