Modeling Filters in Airflow CFD: How to Simplify and Analyze

[ChrisHarvey]

If you wanted to model a filter in an airflow using CFD, how would you go about it? I imagine that an exact model of a filter would be way too complex to model and take vast amounts of time to solve, so there must be some kind of simplification? I was thinking that maybe some known properties of a filter could be applied as an input to the airflow. But the question is then, how do these filter properties become known? Experimentally? Another CFD program? And that's assuming that that actually is the method for solving!

Also, can anybody recommend some decent reading material on filters and CFD and filters?

 

[Astronuc]

CFD, like any FEA, requires properties that have been determined by experiment.

How one models the fluid in a filter or porous media depends the fineness (or conversely, coarseness) of the mesh. Are the FEM mesh cells larger or smaller than characteristic dimension of the porosity of filter cell.

Here is an example - Clean Printing Through CFD
http://www.deskeng.com/Articles/Applications/Clean-Printing-Through-CFD-20050201301.html

I'll look for more examples, but they seem hard to comeby on-line.

 

[Clausius2]

If you wanted to model a filter in an airflow using CFD, how would you go about it? I imagine that an exact model of a filter would be way too complex to model and take vast amounts of time to solve, so there must be some kind of simplification? I was thinking that maybe some known properties of a filter could be applied as an input to the airflow. But the question is then, how do these filter properties become known? Experimentally? Another CFD program? And that's assuming that that actually is the method for solving!

Also, can anybody recommend some decent reading material on filters and CFD and filters?

The point of Astronuc about the flow in a porous media is good. But it depends on how do you want to calculate and what details do you want to capture. The statement of your question is incomplete. One may want to resolve the flow through the filter, or you only want to put a boundary condition equivalent to a filter, for which you need to know what is your filter doing. A filter of what?

Another issue is the turbulence provoked by a filter. Take a look to a turbulence book and read about "Grid Turbulence" and how a grid or porous panel is responsible of a homogeneous and almost isotropic turbulence behind. That's a good point you should take into account for a realistic model of a filter looking downstream.