DAGMC slice plotter rotating geometry?

Hi all, I’m creating a blanket using paramak, and a plasma source. My model looks like this (iso view):
I then export:
get_ipython().system(‘rm dagmc.h5m’)
my_reactor.export_dagmc_h5m(filename=‘dagmc.h5m’, min_mesh_size=1, max_mesh_size=5)

and plot a slice using the dagmc slice plotter package:

However, I think this plot is wrong. Let me explain why. I created the plasma source and added reflective boundaries around this geometry (X<0, Y>0), and the simulation told me that leakage fraction was 1.0 - so no particles went through my geometry. I plotted the surface spectra tally on the outer sphere and it showed a spike at 14MeV.

I then changed the location of the plasma source and reflective boundaries to (X>0, Y>0), and the leakage fraction drops to 0.27230 and the surface spectra show the energy distribution I’d expect
So I believe that either

  • the DAGMC slice plotter is plotting incorrectly (i.e rotating my geometry unwillingly)

  • Or during the export from DAG geometry to OpenMC, everything is being rotated.

I’m not sure why either would happen. I’m attaching my script here (currently running in neutronics workshop docker container)
surface_spectra (1).py (10.2 KB)

Update: After plotting some other tally results of this geometry with ParaView, it seems the issue really was the slice plotter; indeed ParaView shows the geometry in positive X & Y axes:

Hi @Stefan

Thanks for reporting this! I’m sure @Shimwell will be interested in looking into it. The DAGMC slice plotter tool isn’t something we support directly as part of the OpenMC ecosystem, so I’ll refer you to him to sort this out.

Just to confirm, @Stefan, you’re referring to use of this tool for visualization of the model?

Oh yes sorry i missed this. Thanks for reporting i can fix it on the git repo or if you are keen @Stefan i guess it is this line that needs changing

yes, that’s correct. @Shimwell thanks very much. I’m not sure how to make pull requests to change code on Github so might not be able to fix myself. Thanks for pointing this out.

Did you change the line of code locally and get it to work, if so that would be interesting to know