CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   Can't connect volumes in Gambit (https://www.cfd-online.com/Forums/ansys-meshing/93501-cant-connect-volumes-gambit.html)

villager October 17, 2011 19:14

Can't connect volumes in Gambit
 
Good day, everybody.
I had build the model in Gambit, that consists of 12 spheres.http://www.ega-math.narod.ru/Nquant/IMG/Sphere3a.gif
These spheres intersect with each other (figure 1 doesn't demonstrate intersections, it shows a critical case when spheres only touch each other).
Since intersections of spheres are circles, there are circle apertures beetween spheres.
I had build central sphere with 12 apertures, then, i've needed to copy this elementary cell in twelve direction (volume->copy). When I copy this object, some apertures have to match, forming only one face instead two aperture faces. They do match in Gambit, but I have to connect them (Face->connect). When trying to connect some of this faces, I have warn message: "The call to connect on list of faces resulted in no actual operation" and no change takes place. It appears because of edges of apertures (they have the same coordinates) have different basic vertices (that marked green by default - I use Gambit 2.3.16).
How can I connect faces (I must have only one aperture face on each intersection to let fluid flow between spheres.
I'm just a beginner in CFD :) , I've spent a lot of time trying to solve the shown problem, and would be glad for any advice.
John.

-mAx- October 18, 2011 00:45

Quote:

Originally Posted by villager (Post 328311)
I must have only one aperture face on each intersection to let fluid flow between spheres.

So I assume spheres are solid elements around which you may let fluid flows.
I would suggest you to create all your spheres, move them at their right places (don't be disrupted by intersection).
Once everything is right placed, just unite all your volume.
Be sure all your volumes are real else the boolean operation won't be successfull.
Now you just need to create your outer domain, and substract the volume with your spheres from outer domain.

villager October 18, 2011 05:24

No, I want to let fluid flow through apertures of spheres as shown in this screenshot
http://villager16.narod2.ru/ansamble.png

On this figure flow is allright, because of united volumes. But I need to calculate volumetric flow rate throw the apertures of central sphere. If I unite volumes, aperture faces are deleted and I can't calculate flow rate.
So I don't need to unite volumes, I must connect duplicate (superposed) faces. Without connecting fluid doesn't flow (Gambit specifies these duplicate faces as wall. If I retain them, but manually set them to the interface type, Fluent doesn't solve because of error: " Cannot change interior.x to interior because there is only one adjacent cell thread". So I decided to connect faces, using these advices:
http://www.cfd-online.com/Forums/flu...ll-thread.html
http://www.cfd-online.com/Forums/flu...c-problem.html)
http://villager16.narod2.ru/1.png
I can't connect superposed faces if they have different basic vertices (Gambit automatically creates these vertices at circular edges of apertures when I cut off spherical segments with planes). Is there any solution? Thanks.

-mAx- October 18, 2011 05:39

ok so you want to compute flow within spheres an through apertures.
Unite all your spheres >> you will have just one volume.
Then for each aperture, create the surface by selecting the edge(s) forming your aperture.
Then split the volume with the surface. (check that you get one isolated sphere)
Then repeat to next aperture, etc...

Regarding BC: since splits separate connected volume, your apertures will disappear in fluent BC-lists, since they belong to fluid domain. Nevertheless you can enforce Gambit to affect them with internal or interior BC (don't rember which you have to select) and then you should be able to pick them in Fluent BC Panel

villager October 18, 2011 17:10

That has helped me! Thank you very much for advice, for spending time answering me! :)


All times are GMT -4. The time now is 06:12.