Hello Gambit/Fluent users:

I created 3-D helical cooling coils in a tank in AutoCAD and imported it into Gambit. I had to create an imaginary coil around the real coil to "lure" the boundary layer outward into the tank instead of inward into the tube. I am having two problems:

(1) It appears that the mesh in the annulus region (ie. coopered volume from QUAD/PAVE mesh on one endcap face) does not match up with the mesh in the tank (TGRID). I believe that node points between two adjacent objects must match at the same node points. Can you recommend types of meshing that can rectify this?

(2) I defined the imaginary coil as a "porous-jump" in Gambit so that liquid can pass through it with any interference. However, when I import it into Fluent I get the error "cannot create porous-jump, changing to wall. Reason : only one adjacent cell thread." What is an adjacent cell thread?

Any help with one or both of these items would be greatly appreciated! Thank you.

-Eric R.

Re (2): If Fluent notes "only one adjacent cell thread" next to your porous jump it means that there is no mesh connected to the other side of the jump. Therefore it is changed to a wall type face.

Instead of creating a porous jump, why not create a slightly larger coil outside your real coil and split the main volume, creating two real connected volumes with your larger coil? You will be left with face that follows your coil where you can control the mesh density toward the coil.

Basically, if I have understood what you are doing, do the same as you have done except use the larger coil to split the volume.