CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Flow in closed box ? (http://www.cfd-online.com/Forums/openfoam-solving/124965-flow-closed-box.html)

boryan1965 October 16, 2013 08:13

Flow in closed box ?
 
2 Attachment(s)
Dear All,

I have tried to run a OF simulation on a flow through/around a pipe which is in the middle of closed box .The case is 3D, incompressible, steady . See the attached picture for details.

When my ''location-in-mesh'' point is inside the pipe , then sHM ignores the box and I get a solution/flow in the pipe only. When the ''location-in-mesh'' point is outside the pipe and inside the box respectively I have a flow/solution in the box ,but no flow in the pipe. See the attached picture - Box_No_Pipe, the pipe had not been meshed at all by SHM...

I would appreciate any ideas! :confused: Is it so hard to simulate a flow in closed circuit?

Thank you in advance!

Kind regards,

Boryan



================================================== ==================

“Success is one percent inspiration, ninety-nine percent perspiration.” -Thomas Edison
================================================== ==================

N.B. Damn right ...:D

startingWithCFD October 16, 2013 08:27

If I understand your problem correctly, maybe it would help to give a bit of thickness to the cylinder, i.e. to define the tube wall as the volume between two concentric cylinders.

boryan1965 October 16, 2013 09:12

Hi, thanks for the prompt reply
 
I fact the pipe wall is 10 mm thick. The whole pipe was removed from the mesh when ''location -in-mesh'' point was in the box, but outside the pipe...

Obviously snappy Hex Mesh renders my pressure INLET/OUTLET as a external boundaries of the flow domain and ignores any geometry outside it, i.e. the pipe...

The general question is how one can possibly simulate such flow :confused:

startingWithCFD October 16, 2013 14:40

Could you attach your case? It would be easier to understand the case this way.

boryan1965 October 17, 2013 04:59

Thanks for the interest!
 
1 Attachment(s)
The case is very simple, it is a part of bigger picture I would like to simulate. See the attached picture.

The pipe is in the middle of a closed box. Pressure INLET / OUTLET BC are applied on both ends of the pipe - 0 Pa & 500PA respectively.

sHM ignores the pipe from the model when ''location-in-mesh'' point is outside the pipe and inside the box.

Question is how one can simulate this in OF...Fluent runs closed circuit flows easily.

Attaching of constant & system OF folders would not be useful as I use a bespoke version of OF.

It seems to me that I need a ''multi-region'' solver...

Regards,

Boryan

startingWithCFD October 17, 2013 11:38

I wanted to see exactly where "inlet" and "outlet" are in your system but you have verified my theory.
As far as I know, OpenFOAM does not allow internal patches, so, as you wrote, it recognizes them as external boundaries.
Also, multi-region solvers do not work with two adjacent fluid zones...

How exactly are you imposing the boundary values over there?

boryan1965 October 17, 2013 11:46

How about a fan BC ?
 
Right now I am thinking about removing the pressure INLET / OUTLET BC's and introducing a fan BC in the middle of the pipe...the fan will drive the flow and hopefully this will solve the issue with the domain boundaries...

Strangely enough there are not many examples of flow in closed circuit ....:confused:

boryan1965 October 22, 2013 10:48

Case solved !
 
1 Attachment(s)
After some heavy pondering over the case is solved now - one has to remove boundary patches as INLET/OUTLET from the stl. geometry and the case folders and introduce an internal face Fan (baffleZone true; ) on the place of the fan...big thanks to all guys who have helped me !


All times are GMT -4. The time now is 05:43.