[spv24]

Hello,

I would like to simulate the inflow situation of a catalyst. The catalyst consists of a typical channel structure. There are a lot of small channels in the catalyst brick.
I think it is a bit difficult and it takes to much time to draw and mesh the catalystbrick. So I am looking for a suitable BC for the front face of the catalyst brick.
What would be a good, realistic BC for simulating the inflow. I need the velocity distribution in the front face plane of the catalyst.

Is it possible to have a BC in a flow domain to force a 1 dimensional flow for simulating the catalyst as separate flow domain? Let the channel direction in the catalyst be x then the U_y and U_z components should be set to zero. Is this possible? Does anybody have better suggestions?

Thanks Sammy.

[uwe]

Hi Sammy,

why don't you model the catalyst as a porous media using the solver rhoPorousFoam.

Uwe

[spv24]

Hi Uwe,

First: I have not enough information about the catalyst brick.

Second: I am not sure how to implement the porous material in OpenFoam. There are the tutorials, but it is not clear to me how to set it up. There is a coordinate system in the porouszone file, but I always thought that OpenFoam is a consequent 3d tool . So the two vectors make no sense for me. I am not sure how to use the Darcy-Forchheimer model. Which parameters I have to fill in? There is not enough explanation for me about the whole porosity model. I am a beginner in OpenFoam!

So I am still looking for an opportunity to restrict the flow in one direction in a specified flow domain. Any suggestions? For Example U_x=U_x U_y=0 U_z=0.

Thanks Sammy

[uwe]

Hi Sammy,
the coordinate system defines the direction of the resistance vector, i.e. you can have a different resistance in x-, y- and z-direction. In your case, set high values for the resistance in y- and z-direction. In x-direction you can fit the value, e.g. to experimental values for the pressure drop. For the explanation of the darcy-forchheimer model please have a look at Mark's answer.
http://www.cfd-online.com/cgi-bin/Op...5313#POST15313

Uwe

[spv24]

Hi,

thanks for the reply. But I still don't understand why they use only two vectors for defining the coordinate System? Is the third one produced automatically as orthonormal system?
Then it would be clear what is going on.

Thanks Sammy

[uwe]

Hi Sammy,
you are right.
Uwe

[olesen]

Samuel,

A third directional vector would not only be annoying to enter, it is also redundant.
Note that in the coordinateRotation, the first directional vector has priority and any rounding errors or non-orthogonality is absorbed into the second directional vector.
The 'first' directional vector is the first that appears in the right-hand rule.
For the combination e1/e2, e1 is primary.
For the combination e2/e3, e2 is primary.
For the combination e3/e1, e3 is primary.
This information is in the coordinateRotation doxygen docs. The other bits can be found in the coordinateSystem doxygen docs, which even has a link to porousZones (and thus indirectly to porousZone).

[spv24]

Thank you very much. I looked for the documentation. But it was not completely clear to me because I am a beginner in OpenFoam.

Thanks a lot!

Sammy