Problem in meshing a pipe within a cylinder
 

Hi,

I am new in Gambit and I did the model of my reactor (UASB),

http://jj.emuch.net/image1/ca/f8/100...500105_157.png.

There is a cylinder with an influent pipe on the bottom and an effluent pipe on the up side, the funnel is used to collect gas produced in the reactor (little gas was produced so it is neglected).

I create a 2-D model in Gambit with two separated faces, while seems that the funnel has little influence on the fluid, should any opration be done to connect the two separated faces? What should be the suitable boundary conditions for the pipes?

Thanks,

Tony

what is this body with trianle basis?
Looks like a spool entering in your cylinder.
If it is a solid body (no flowfield inside the spool), then you have to go back in Gambit and you need to split your cylinder with the spool. Finally the spool should appear as an hollow, which means no flowfield expected in this area.

Thanks for your reply.

It is an inverted funnel (just like a pipe) and the fluid can flow into it. Should I split the cylinder with the funnel and then connect them?

where can the flow enter in this funnel? Are there openings?
Do you have a cross scetion from cad or anything for better understanding?

The funnel is a hollow cone with large opening on the bottom and a narrow tube at the apex, which is used to collect the gas produced, as the opening is lower than the effluent pipe (water level), so the flow can also enter, but probably it will not outflow through the narrow tube at the apex

http://132.68.226.240/english/pdf/Pr...al_Green/2.pdf picture 1 (a) is a sample UASB reactor

http://mbio.asm.org/content/2/4/e00159-11/F6.large.jpg

http://www.uasb.org/discover/uasb-scheme.gif
http://www.engineeringfundamentals.n...Bs/fundam4.gif

These are more visualized exsamples of UASB reactor

ok
post a picture of your gambit geometry (without mesh) and locate your BC, especially on your cone.
The question is: is your cone splitted from the cylinder or is it totally free (not connected).
The cone should split the cylinder, BUT, you need to redefine all the cone edges as wall (except the bottom one). Thus split operation generates internal (or interior) entities for giving the possibility to the fluid for flowing through.

Thanks, seems that it works