Meshing and zones of a reactor vessel
Hi everyone.
I am working with a reactor vessel in fluent. How to make all the elements of the mesh hexahedral? Also in the reactor vessel, there are three zones with different porosity values and there are nozzles with pipelines that connect to the steam generator and circulation pump. How to create the named selection in this case ? As it is a big geometry I am finding difficulties. I have designed it in SolidWorks for convenience. Do I have to combine the parts of the geometry in SolidWorks and then use it for simulation in fluent? Or do I have to create named selections in fluent and then simulate? |
I suggest you to export the .IGES file from Solidworks to ICEM. You can create the named selections manually by following these steps:
(1) Import the .IGES file to ICEM (2) Right-click on parts, click on "adding to part". Then select each surface item in the GUI and give it a distinct name than the default (ICEM assigns name GEOM, you need to change this for each boundary). Once done this with all your boundaries you can block your fluid region manually. You will need to associate the block edges to the curves of your geometry. (I suggest you to break by connectivity, so you can have all curves defining limits shared by surfaces). Then just mesh manually. This will give you an unstructured mesh with pure hexa elements (depending on your geometry, this might be tedious but will give you a very high-quality hexa mesh) |
How to do the structured mesh in ICEM
Can you tell me how to do the structured mesh in ICEM?
|
Dear stfahim
Well, what I typically do is to export the .IGES file from Solidworks and import it to ICEM just as it is. ICEM recognizes the curves that define the bounding surfaces of your CAD file. Once in ICEM you just have to: (1) Go to the tree (bottom-left side of the screen), right-click on "Parts" and then Create Part->Create Part by selection->Select entities from the screen and assign a name, then click OK. Repeat as necessary and be sure to give distinct names to each named selection. (2) Once with all parts created go to the blocking tab and create a 3D block. ICEM introduces a general bounding box. Watch some youtube videos on blocking, they will teach you how to properly block your geometry. Depending on the complexity of your geometry this may be suitable or not, sometimes is very tedious to define proper blocks, but trust me, this will give you a high quality hexa-mesh, if done with caution. This method is basically like "to chisel" the blocks that define your fluid region. Greetings |
errors from fluent
5 Attachment(s)
Hi there.
I couldn't upload the photos previously. Here I have added those. Previously, I have tried to mesh in Fluent and I got the errors "Patch conforming tetrahedron mesh failed possibly because of bad boundary mesh and A mesh could not be generated using the current meshing options and settings." You can see the errors in the picture along with the design. These errors appeared in the default setting. Then I tried sizing if that works. But that didn't help. Is there any way to solve these errors? And do you think doing in ICEM CFD would be a better choice instead of doing it in Fluent? In that case I'll try to do it in ICEM Thanks a lot for your help. :) |
Well, from the pictures I see that your geometry consists of a central vertical cylinder and 4 pipes coming out from the sides. Depending on the type of analysis you want to perform, I'd test first a mesh with pure tetras, and then I go for hexa.
However, although meshing hexa in your case seems realizable with only extruded O-grids, you might find this process tedious especially if you are not familiar with ICEM, gambit, etc.. As a suggestion never try those automatic-hexa meshing options, most likely they will give you errors. I am not expert with Fluent meshing tool, but if you have access to ICEM, you can mesh tetras by using the global mesh setup. Watch some youtube videos if possible, they will teach you from scratch |
Thanks
Thank you for your help :)
|
All times are GMT -4. The time now is 02:44. |