[deve]

Hi Foamers,

i'm trying to simulate the flow inside a cavity with a two phase solver. The two phases(resin and air) flow through porous media, which works fine.
I now have to implement a semi-permeable membrane (which let's the air pass but not the resin). See picture below:

Membrane.jpg

What i wanted to achieve is that alpha gets 1 little by little over time until it reaches the membrane (green).
I came up with the following ideas and would be happy for EVERY idea you guys might have. (of course i prefer the most simple solution possible, but i guess i chose a rather difficiult task; moreover i searched the forum for possible solutions and realized that i'm not the first person with this problem)

Idea 1)
Define a Face inside my cavity and set alpha uniformly equal 0 so that no more flow can get through the faces. Sadly, that didn't work because i imported a mesh from hypermesh (fluent-type) and while converting it, openfoam ignores the face since inside the mesh, if i understood that correctly.
Is that generally possible?
What do i have to do that Openfoam will accept the face inside?

Idea 2)
Define a new Boundary Condition which keeps sucking air (due to p = 0) but i have to f*cking idea how to do that! It would be find if the membrane is not in the acutal cavity but represented by a BC.
Is that possible without have a PHD in C++ programming?

Idea 3)
Create a new Element type!?!? which is porous but only for fixed alphas


You may see that some(or all) of my ideas are total bullsh*t, feel free to tell me!
Any other ideas are highly welcome!

Greetings
Stefan

[WildeCat]

Hi Deve,

I am currently working on a similar problem, I need a mass transfer boundary condition from a fluid region into a porous medium. I was curious to know if you could solve your problem, and how you solved it, if you dont mind sharing.

Thanks in advance
Max

[WildeCat]

Hello everyone,

I would like to give an update for all people who are attempting to couple to fluid regions.

Two way temperature coupling can be achieved with the thermal boundary conditions. Many different exist. More information can be found on:

http://openfoam.org/release/2-3-0/thermal/

Apart from temperature, in some cases it is necessary to couple also other variables such as rho, U or p. In these cases, as far as I know, no two way coupling exists. If however you know that you only need to transfer information from one region to the other, so not coupling it, you can use mappedField. This boundary condition allows to map the conditions at one patch to another patch on different regions. The source guide may help you to understand this boundary condition better.

http://cpp.openfoam.org/v4/a01446.html

Since the source guide is for OF4.x, but I use OF2.3.x, where it is implemented somewhat differently, I will add here how to implement this boundary condition. Make sure the interface between the regions is a mappedWall.

    <interfacePatchName>
    {
        type                  mappedField;
        sampleMode      nearestPatchFace;
        sampleRegion    <neighbourRegion>;
        samplePatch      <interfacePatchNameOnOtherRegion>;
        fieldName          <vairable>; // scalar or vector
        setAverage        no;
        average             (0 0 0); // 0 for scalar
        value                 uniform (0 0 0); // placeholder, 0 for scalar
    }

I am currently working on writing two way coupling between different regions for U. Depending on my case I may be able to solve it differently, but should i find a suitable solution, I'll try to post it here.

Hope I could help