[Mohsin]

Hello

Gambit gives the following error when I use cooper scheme for volume which has some face splits:

Error: "Splitting of existing mesh on source face is not implemented yet,
try using mesh split functionality on source face"

Has anyone experinced this? and why is it caused and what might be solution to avoid this.

Note: Source faces are correctly specified. Everything seems to be perfect for cooper mesh but this error is irritating...

[-mAx-]

post pictures

[Mohsin]

Thank You Max

Cooper scheme is used and the upper face is plane but the lower face is splitted into segments. The upper face is already meshed and is projected downwards to the faces of splitted volumes (which are not meshed as per the requirement of cooper).

Why is this error coming? when i split the upper face like the lower face then the error disapears but i dont want to split the upper face because then i have to split all the upper geomtery for meshing.

Assistance required. Thanks

[-mAx-]

Originally Posted by Mohsin

The upper face is already meshed and is projected downwards to the faces of splitted volumes (which are not meshed as per the requirement of cooper).

That's the problem: you need to mesh the lower surface first and leave the upper meshless.
The upper surface is "free", and the lower have less freedom since you enforce the mesh to follow some splits.
You cannot enforce cooper to project a "free" surface onto a "resticted" surface.
you understand what I mean?

[Mohsin]

yes i understand but if u make 2 geomteries and split the one of them into 2 then if u apply cooper mesh (in the same manner) then it works. But in this case, it isn't. I mean even if the volumes are split into different volumes, still cooper scheme works in some cases but in this case it is not working. A little confusing...

[-mAx-]

delete the surface mesh on the upper surface (or the volume up to your volume from interest)
And remesh

[Mohsin]

Is it possible to split the upper face's mesh like the lower surface? I tried face and edge mesh splits but was unsuccessful. Also, can virtual edges, faces or volume play a role in this issue? I mean for meshing from upper face to lower face can u think of any possibility?

Thanks

[-mAx-]

You have a problem with volume 93. (2 superposed surfaces)
Go and connect all surfaces.
Then try to mesh.
But I don't understand why you are trying to mesh all your domain with hexa, except impellers with coarse tetra.
I assume you will have less problem, if you merge all your volumes, and you apply Size Functions on your impellers.
Then mesh your volume with tetra-hexcore.

[Mohsin]

I am sorry I didn't get your point

Volume 93 2 superposed surfaces? are there any extra faces? how is it superposed?
Secondly, u mean I should merge all the volumes in the domain and apply size function at the specific impeller regions?

I was meshing it with hex cooper in order to get a good quality mesh.

Could u send me a rough sample of jou. file of my geometry? I'll be highly grateful to u..

[-mAx-]

What is the goal of your simulation?

[Mohsin]

The goal is to get the flow behavior inside the main cylindrical chamber. Droplet evaporation, collection of droplets at the outlet and temperature at the wall etc.

Inside the chamber a highly swirl flow occurs because of the impellers.

[-mAx-]

Then a "fine" mesh on impellers would ba appreciated, and you can even aplly a BL
In you case you have very coarse mesh around impeller
Sans titre.jpg