Hello, all.
I am trying to use the unstructured mesh to get the parameters of diffusion equations and multi-groups flux distribution of a heat pipe reactor. I am basically following the example of this: example.
I use the Cubit and its DAGMC toolbar to generate the mesh file, specifically, I initially to generate the model.xml file which contains the materials, geometry and settings, and then I used the openmc_to_cad to convert the xml file to jou file that Cubit can use to create the surface mesh, and finally generated the dagmc.h5m file (I can’t upload this file).
The running code I write is here: generate_model.py (9.9 KB).
The trouble I encountered is that there is nothing in the output vtk file but the coordinates of points. I think the problem might lie in the step where I generate the mesh file. Can anyone tell me how to generate mesh file correctly?
Thanks in advance!
I have solved this problem by following this Cubit post：solution,
though there are some little problems during converting the exodus file to dagmc file:
mbconvert krusty_mesh.e krusty_mesh.h5m
[0]MOAB ERROR: --------------------- Error Message ------------------------------------
[0]MOAB ERROR: krusty_mesh.e: File is not HDF5!
[0]MOAB ERROR: set_up_read() line 451 in src/io/ReadHDF5.cpp
[0]MOAB ERROR: --------------------- Error Message ------------------------------------
[0]MOAB ERROR: NULL file handle.!
[0]MOAB ERROR: is_error() line 134 in src/io/ReadHDF5.hpp
Read "krusty_mesh.e"
Wrote "krusty_mesh.h5m"
The tally finally is not empty now, so I won’t complain about anything.
Hi,
As a user very familiar with this workflow I suggest saving the file as a .cub file and then converting the .cub to a .exo (dont convert to .e since the utility is not updated) using the mbconvert utility. You can also use mbconvert to create the .h5m from the .cub file as well and this insures you have conformal meshes for transport and tallies. I would also use libmesh as the unstructured mesh library as it actually scales with OpenMP while moab doesn’t do well. Also, Be careful on the number of particles you run because with a large mesh that can need million to billions of particles to get good statistics. Hope this helps
Travis
Thanks for your reply, Travis, it is very helpful!
