snappyHexMesh keeping internal geometry
Hi Everyone !
I am currently trying to run a conjugate heat transfer case based on building a Salome STL geometry, converting it to 2-D-compatible case using OpenFOAM through snappyHexMesh, and iterating this procedure with different STL geometries.
Everything worked well simulating as simpleFoam case, but at this point, I want to do a conjugate heat transfer case, which needs to keep the internal solid cylinder geometry.
Can snappyHexMesh be built so that it may not remove the geometry ? Or do somebody knows another solution to 're-include' this removed geometry easily in my case ? (basic hex mesh with few cells)
This question may sound a bit basic, but I've looked around and couldn't find any answers... any help would be greatly appreciated ! :)
have you checked in the tutorial?
is a case that assembling parts.
I would like to create partition so that the boundary faces and points are merged.
Is it possible?
I did check it, and for some simple geometries (like cylinders), I could get the two separated regions, but more complicated shapes were leading to the creation of "domains" when splitting the mesh to regions, which were not compatible with the fluid/solid composition I wanted...
What are actually those "domains"?
I have been using extrudeMesh functions to extract some specific regions for chtMultiRegion purposes, but it only helps for rather simple geometries... was that what you were meaning?
i have the same problem ! snappy/splitMesh generates not necessary domains*
have you found any solution?
PS: i have submit my case in other post but anybody help me !!!
well, I found some ways to sove this problem, depending how complicated is the geometry.
First one was the use of a software to "extract" the faces out of my complex geometry (such bevel gears: http://www.elektrowerkzeuge-ersatzte...TE704za012.JPG)
something that sHM wasn't doing properly (thus generating a lot of those "domains"). This software is unfortunately not free, but solved this problem (check CastNet, from DHCAE-Tools).
On the other hand, a more simple designed geometry from those gears could be implemented without problems, only by identifying the different domains between solid and fluid area. Fact is that the toal number of domains should never exceed the amount of specified fluid/solid areas...
Does it give an answer to your question?
Thank you so much for your help!
I prove CastNet ......
An other problem is the construction of stl file.
I generate a simple geometric shape (solid) with Salome, explode in faces and then exported the single surface in stl format.
When i collect the stl faces files in one file and run surfaceChek on this file one "error" occurred, the surface is disconnected.
Which is the procedure to create right surface closed stl file from singe stl surfaces (boundary) files ?
PS: is this possible with solid works?
Thank you so much again!
I don't know about SolidWorks, I myself kept on using SalomeMeca for simple geometries, avoiding problems with sHM then.
Could you post your STL file, and also your Salome file as well?
i post my stl files:
1) pipe.zip contain the SALOME files and stl export from salome -> FLUIDO.stl and SOLIDO.stl
2) pipe_composed.zip contain single surface stl files of SOLIDO, FLUIDO and the "composition" of this files to form FLUIDO.stl and SOLIDO.stl
When i run surfaceChek on composed stl files (in pipe_composed.zip) SOLIDO.stl and FLUIDO.stl the surface is not closed.....
Hello to everyone,
I run into a similar issue where the inner geometries were being deleted.
As far as being able to show the entire geometries, I figured out that one need to choose the location of locationInMesh point such that you can see all the geometries from it.
For example, if within your block (i.e., the one you create from blockMeshDict), say you have an open cylinder that contains other geometries, in such a case, one may consider placing the locationInMesh on the open side of the cylinder so that the inner geometries are not hidden from it by the outer cylinder.
I hope this will be helpful to some out there.
Also, It would great that one of the senior members say whether this analysis is correct.
Note I figured this out by going through the propeler tutorial in pimpleDyMFoam.
Best regards to all.
|All times are GMT -4. The time now is 07:05.|