|
[Sponsors] |
August 14, 2024, 12:30 |
model geometry
|
#1 |
New Member
Join Date: Dec 2023
Posts: 6
Rep Power: 2 |
wall_rep.jpg
Hi all, I wish to create a model representation that two inner tunnels are buried in a main body, as shown in the figure. My questions are 1 - how to merge the branches and the main body as a complete surface? currently, the branches are aligned to the main body, however, there are still curves, indicating these parts are unmerged. 2 - from the prev. question, if such a model is meshed and imported into a cfd software, which side of the wall should be the actual 'wall'? this seems to be a dumb question as currently they are approximated using the shell. However, if the branches carry the thickness, I doubt this would confuse the software. Thanks in advance...! |
|
August 14, 2024, 17:31 |
|
#2 |
Senior Member
Gert-Jan
Join Date: Oct 2012
Location: Europe
Posts: 1,885
Rep Power: 27 |
I don't understand your geometry. When doing CFD, you need to mesh the fluid. To me it is unclear in which part you are interested. The main channel, the two smaller ones, or maybe all three? And in the latter case, are they connected or isolated or somehow connected?
Last edited by Gert-Jan; August 15, 2024 at 03:22. |
|
August 15, 2024, 06:20 |
|
#3 | |
New Member
Join Date: Dec 2023
Posts: 6
Rep Power: 2 |
Quote:
The question is, how to design such surface that is enclosed? I realised that stls have both the inner and outer normals, hence, if i represent this surface using a layer of the shells, the wall of the inner tunnel actually has two layers... should I thicken them? Thank you! |
||
August 15, 2024, 06:41 |
|
#4 |
Senior Member
Gert-Jan
Join Date: Oct 2012
Location: Europe
Posts: 1,885
Rep Power: 27 |
It depends a lot on which software you will be using, but in principle the wall between two volumes can be infinitely thin. Then the CFD package (CFX or Fluent) will detect two volumes that are connected by a surface. You will obtain different surfaces on both sides. So you can apply different boundary conditions on both sides if necessary.
You should: - make sure the 3 volumes are closed volumes. A small hole might be present as long as it is much smaller than the element size your choose. - make 3 body points in each volume. - run the mesher. If you have only 1 volume (meaning the volumes are connected somewhere), then you should create a separate part with the internal wall (name it INT_WALL). Then in "Mesh Setup" you tell ICEM that INT_WALL is a "Īnternal wall" or a "Split Wall". Your challenge will be where the surface touch tangentially. There, the angle between the surfaces will be 0°. This will lead to numerical problems. Especially if you want inflation on all walls. You either fill the narrow space up to a certain extent. Or make a small gap between the walls. You have to violate the geometry. But at least do something. |
|
August 15, 2024, 07:35 |
|
#5 | |
New Member
Join Date: Dec 2023
Posts: 6
Rep Power: 2 |
Quote:
|
||
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
viewFactor radiation model stuck | shach934 | OpenFOAM | 1 | May 25, 2021 11:01 |
Unexpected line in my geometry model | 372316422@qq.com | ANSYS | 0 | March 9, 2016 09:02 |
Wrong flow in ratating domain problem | Sanyo | CFX | 17 | August 15, 2015 06:20 |
Water subcooled boiling | Attesz | CFX | 7 | January 5, 2013 03:32 |
vitual _ real | deneb | FLUENT | 3 | January 22, 2007 04:31 |