Error getting pre-concs values from previous run for restart

[ajhlee]

Error getting pre-concs values from previous run for restart

[katyhuff]

Can you update moose and moltres, and rebuild, then see if you still get this error? New changes in moltres have been made to keep moltres up to date with new versions of moose... It's usually best to use the most updated version.

[ajhlee]

Hi all, just rebuilt moose and moltres to the latest versions. Re-ran the above LOSCA example and the error persisted:
"Error getting elemental values.
[0] ../src/mesh/exodusII_io_helper.C, line 1059, compiled Feb 18 2019 at 20:14:09
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0"

Examining the exodusII_io_helper.C file, line 1059 is:
EX_CHECK_ERR(ex_err, "Error getting elemental values.");

  std::vector<Real> block_elem_var_values(num_elem);
  ex_err = exII::ex_get_elem_var(ex_id,
                                 time_step,
                                 var_index+1,
                                 block_ids[i],
                                 num_elem_this_blk,
                                 block_elem_var_values.data());
  EX_CHECK_ERR(ex_err, "Error getting elemental values.");

[gridley]

Well, this superficially looks deeper than Moltres. The first thing to do is check which branch of MOOSE you pulled from---it could be a bleeding-edge branch failing unit tests. To figure out if it is MOOSE, you could try to run something dead-simple: heat conduction. See this example.

https://github.com/idaholab/moose/blob/next/modules/heat_conduction/test/tests/verify_against_analytical/1D_transient.i

I propose that you take this simulation, and rather than go for 100 seconds, split the simulation into two input files: one for 50 seconds, printing its results to a file, then a second reading the results from the first to run it out to 100. Then you can compare that to the 0-100 run. If the second fails to load or they disagree, you should raise an issue to idaholab/moose. Unfortunately I do not have time to do that at the moment, so good luck! 😄

[lindsayad]

A working path is at libMesh/libmesh#2041

[ajhlee]

@lindsayad Thanks! I recompiled libmesh (and MOLTRES) with the modified exodusII_io_helper.C and there are no more problems with restarting using the previous run's pre-conc values!

[lindsayad]

That's great to hear!

…

On Thu, Feb 21, 2019 at 1:02 AM Alvin Lee ***@***.***> wrote: @lindsayad <https://github.com/lindsayad> Thanks! I recompiled libmesh (and MOLTRES) with the modified exodusII_io_helper.C and there are no more problems with restarting using the previous run's pre-conc values! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#80 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AJxgcDdbJl7BkfRayQGkjyRbA8I1wDblks5vPlKSgaJpZM4bB5_c> .

[gridley]

Does this work now for sure @joonhon? If so the issue could be closed.

[ajhlee]

Hi Gavin, thanks for the reminder. I believe the issue has been resolved. I'll close the issue, Thanks!