Dear All,
I am doing meshing for high resolution .stl model (I set the finest resolution when importing .stl file from solidworks) under SMS mode and constantly encountering the following errors.
What’s more, I report the bug again that once I click the reset all default options, the tol_snap(0) = 0.0 appears, hope that can be fixed later. Thank you.
The tol_snap bug will be fixed in the 21.3 release coming out today or tomorrow.
Please see attached for an example that runs with uniform mesh:simple_chamber_test1.mfx (17.9 KB)
Please be patient to allow sufficient time for users to reply and don’t post duplicates to get a better chance to receive suggestions to fix an issue.
Thank you Jeff and your model works good but it needs 400 days to simulate 1 second which is horrific. So I want to know whether there exists some methods to simplify the computation of the model with tolerant errors. And what are the principles to do meshing for the model with delicate structure like tubes.
And I want to know whether the other methods will take much less time compared to TFM with acceptable precision.
And I want to know if I have no access to HPC personally, what is the most economic method to rent some computation resources? Thank you.
Thank you.
The initial estimate may not be always accurate. It is based on initial time step and convergence rate. It usually gets better once you pass the initial transient. It ran in less than 2 hours for me, using 160 cores.
You might be able to coarsen the mesh a bit, but the presence of internal tubes will require sufficient mesh resolution to capture the geometry.
Don’t you have access to SCITAS at EPFL?
Unfortunately, our HPC system is reserved for in-house research.
