Convert to OpenMC boundary_type

[zxkjack123]

I converted a model using csg2csg from MCNP to OpenMC. And met the following error:

ERROR: No boundary conditions were applied to any surfaces!
application called MPI_Abort(MPI_COMM_WORLD, -1) - process 0
[unset]: write_line error; fd=-1 buf=:cmd=abort exitcode=-1
:
system msg for write_line failure : Bad file descriptor
Traceback (most recent call last):
  File "input.py", line 157, in <module>
    main(args.geom, args.nps, args.run, args.plot, args.vol)
  File "input.py", line 129, in main
    openmc.run()
  File "/home/zxk/opt/openmc/openmc/executor.py", line 212, in run
    _run(args, output, cwd)
  File "/home/zxk/opt/openmc/openmc/executor.py", line 30, in _run
    ''.join(lines))
subprocess.CalledProcessError: Command 'openmc' died with <Signals.SIGSEGV: 11>.

I think this error is related to the boundary type. We need to define the graveyard using boundary_type='vacuum'. However, for the MCNP input file I used, I didn't find the 'imp:norimp:p`.

[makeclean]

Hmm it should do that if your cells use imp:n as opposed to a data block. Could you share your geometry? Get Outlook for Android<https://aka.ms/ghei36>

…

________________________________ From: Xiaokang Zhang <[email protected]> Sent: Monday, August 5, 2019 5:30:12 PM To: makeclean/csg2csg <[email protected]> Cc: Subscribed <[email protected]> Subject: [makeclean/csg2csg] Convert to OpenMC boundary_type (#28) I converted a model using csg2csg from MCNP to OpenMC. And met the following error: ERROR: No boundary conditions were applied to any surfaces! application called MPI_Abort(MPI_COMM_WORLD, -1) - process 0 [unset]: write_line error; fd=-1 buf=:cmd=abort exitcode=-1 : system msg for write_line failure : Bad file descriptor Traceback (most recent call last): File "input.py", line 157, in <module> main(args.geom, args.nps, args.run, args.plot, args.vol) File "input.py", line 129, in main openmc.run() File "/home/zxk/opt/openmc/openmc/executor.py", line 212, in run _run(args, output, cwd) File "/home/zxk/opt/openmc/openmc/executor.py", line 30, in _run ''.join(lines)) subprocess.CalledProcessError: Command 'openmc' died with <Signals.SIGSEGV: 11>. I think this error is related to the boundary type. In MCNP, we need to define the entire universion, and set the graveyard using imp:n=0. But in OpenMC, it seems that it uses boundary_type='vacuum' to define the boundary of the graveyard. Therefore, when converts the input file from MCNP to OpenMC, we need to detect the boundary surface of the graveyard by imp:n=0 or imp:p=0. And set these surface with type vacuum. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#28?email_source=notifications&email_token=AASTUSRAOJE5SUNA52ZIHUDQDBIRJA5CNFSM4IJM5COKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HDNWVXQ>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AASTUSTYRL5QO7QEB2IXSULQDBIRJANCNFSM4IJM5COA>.

[zxkjack123]

@makeclean Sorry for that, I just checked in the file, there isn't imp:n. It is the FNG model, I will send the file to you via slack.

[makeclean]

@zxkjack123 ah ok, no problem. I haven't implemented block reading for importances or weights yet, feel free to have a go though! I've not got huge amounts of time, and the 'perceived complexity' of the number of different options one has for defining importances has put this down my list, e.g. explicitly 1 2 3 4 5, repeated 1 4R, mulitiplicative 1 3.5M 3.5M, interpolative (sp?) 1 2I 4, im sure there are more :)