Thanks for your post and detailed description of your problem! I will have a look at that and also tag my colleagues and well known PowerUsers @Get_Barried & @vgon_alves here who might give you some useful input here.
Aims to imrpove your mesh, I suggest to start implemanting some inflation into fluid-valve contact region. I saw that you tried it but your inflate b.layers in mesh 8 were not loaded successfuly.
You can also add some “vicinity” refinement regions to gradually refine your mesh according to the valve and flow direction adding one more cartesian box in addition to the one added. What I did above is too far away of a good mesh, but I think it can be a good begin. EDIT2: It is running OK at least for 100 [s]. I canceled it with the purpose to preserve core hours . You can find the project [here][Deleted by author]. The mesh is called “90 60 150 mesh”.
Maybe this project helps you to see these steps I mentioned above.
Please, let me know if it was useful in some way.
Just to add on to what was mentioned by Vinícius. If you refer to the figure below you will notice that you basically have a broken mesh based on the very high illegal cell count. This is what is causing your simulation error.
From the same mesh’s meshing log, you will also notice that your layers are not showing up as Vinícius has mentioned.
This post here should help you with regards to layer generation.
The issue with illegal cells seems to be with the geometry. Your pipe has a gap between the inner and outer surface along with the hollow cap that is interfering with the inner and outer pipe surface. The presence of this coupled with the sharp edges along with the high refinement levels all equate to the very high illegal cell count you encountered. Do refer below.
A good way to mitigate this issue would be to either create a simple pipe assembly with just one inner surface and a cap with no thickness or to close the gaps within the geometry (for the cap and the pipe itself) by filling in the gaps.
@vgon_alves and @Get_Barried,
so i make a new geometry, only existing out of surface until the umbrella.
I use your values, refinements etc. Here is the link to the new project.
The picture shows the first test simulation. Did you know how i improve the mesh at the begin of the umbrella ?
Better refinement on that edge will help Kai, also if you add a feature refinement that goes say one level finer than the end, then you have a better chance of inflating the layers around the geometry.