OpenMC depletion coupling #921

loganharbour · 2024-06-26T21:11:47Z

Things still to do

Actually verify the test, or setup actual verification?
Also add a test with the openmc-only operator?
Clean up the python imports in depletion.py
Add more comments, maybe a little documentation
Setup a proper clean exit of the CardinalOperator that checks to see if MOOSE is also done, otherwise error
Clean up the state cleanup before running (hdf5 and tallies)

loganharbour · 2024-06-26T21:16:40Z

FYI @aprilnovak and @pshriwise; still a mess but it works!

moosebuild · 2024-06-26T21:39:38Z

Job Documentation on 5e3baba wanted to post the following:

View the site here

This comment will be updated on new commits.

pshriwise

This looks really good!! Thank you for creating this @loganharbour. Mostly line changes here and a couple questions about integrating it into a Cardinal install.

pshriwise · 2024-07-08T15:51:44Z

python/cardinal_operator.py

+        super().__init__(*args, **kwargs)
+
+        self._control = None
+        self.cardinal_cmd = None


Let's move these attribute initializations before the super().__init__ call so we don't get an AttributeError from CardinalOperator.__del__ if the parent constructor fails.

Suggested change

super().__init__(*args, **kwargs)

self._control = None

self.cardinal_cmd = None

self._control = None

self.cardinal_cmd = None

super().__init__(*args, **kwargs)

pshriwise · 2024-07-08T15:53:31Z

python/cardinal_operator.py

+            nuclides = [n for n in t.nuclides]
+            names_path = f'{uo_path}/names'
+            self._control.setControllableVectorString(names_path, nuclides)
+            break


We can remove this now, right?

Suggested change

break

python/cardinal_operator.py

pshriwise · 2024-07-08T15:55:30Z

python/cardinal_operator.py

+        if self._control is None:
+            self.start_cardinal()


Suggesting the tiniest of refactors: This pattern appears often. Do we want to put it into a method like ensure_cardinal?

pshriwise · 2024-07-08T15:59:37Z

test/tests/depletion/depletion.py

+
+# All of these imports are awful and we'll figure
+# out a better way to do it before merging
+try:


It seems fair to me to expect users to have installed the OpenMC Python package used with Cardinal, but it doesn't look like your typical installation as that's going to occur from the submodule. Offhand thoughts:

(optionally) Install the OpenMC Python API as part of the Cardinal build

Include setting of PYTHONPATH to the OpenMC submodule in Cardinal's installation docs.

It'd be awesome if we automatically installed the Python API when installing Cardinal -- it's listed in this issue: #592

pshriwise · 2024-07-08T16:00:17Z

test/tests/depletion/depletion.py

+try:
+    from cardinal_operator import CardinalOperator
+except:
+    cardinal_contrib = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..', '..', 'python'))


see above, this could also be solved by requiring the PYTHONPATH to be set to CARDINAL_HOME/python in the installation notes.

A warning similar to the one about NEKRS_HOME could be provided perhaps.

.gitmodules

moosebuild · 2024-10-29T19:02:47Z

Job Documentation, step Sync to remote on 791f7ac wanted to post the following:

View the site here

This comment will be updated on new commits.

aprilnovak · 2025-01-27T01:13:05Z

@loganharbour would you please rebase this off devel and then update this? Now that #923 was merged, is it time to return to this PR?

…n chain

pshriwise · 2025-02-04T18:12:28Z

Looks like we'll need the endf package in the testing environment for OpenMC. @loganharbour is that easy to add?

loganharbour · 2025-02-06T03:42:09Z

Looks like we'll need the endf package in the testing environment for OpenMC. @loganharbour is that easy to add?

A bit. I've needed to find a way to allow custom installs for apps. I'll get to that tomorrow

aprilnovak · 2025-06-10T20:50:24Z

I'm trying to run this locally so I can learn how to use this feature. Do I need to pip install the new python code?

ajnovak2@ajnovak2-serv-01:~/cardinal/test/tests/depletion$ python depletion.py
Traceback (most recent call last):
  File "/home/ajnovak2/cardinal/test/tests/depletion/depletion.py", line 3, in <module>
    from MooseControl import MooseControl
ModuleNotFoundError: No module named 'MooseControl'

Is it as simple as?

cd cardinal/python
pip install .

loganharbour · 2025-06-10T20:54:24Z

I'm trying to run this locally so I can learn how to use this feature. Do I need to pip install the new python code?

ajnovak2@ajnovak2-serv-01:~/cardinal/test/tests/depletion$ python depletion.py
Traceback (most recent call last):
  File "/home/ajnovak2/cardinal/test/tests/depletion/depletion.py", line 3, in <module>
    from MooseControl import MooseControl
ModuleNotFoundError: No module named 'MooseControl'

Is it as simple as?

cd cardinal/python
pip install .

export PYTHONPATH=/path/to/cardinal/contrib/moose:$PYTHONPATH

aprilnovak · 2025-06-10T21:06:20Z

Actually I think I got a bit farther now. I tried setting export RUNAPP_COMMAND="cardinal-opt -i openmc.i". The simulation starts running but then ends with

MooseControl.app: *** ERROR ***
MooseControl.app: CLI_ARGS:1: (UserObjects/openmc_filter1/filter_id):
MooseControl.app:     Filter 1 does not exist and create_filter is false

I think this is probably due to some source code changes in Cardinal which have happened over the last couple of months from the new Tallies syntax. It seems that the filters have not yet been created by the time the OpenMCDomainFilterEditor is trying to use them.

aprilnovak · 2025-06-10T21:14:48Z

This test case seems to have originally been working after the [Tallies] syntax was added. @nuclearkevin can you identify anything which might have changed since June of last year when this PR was opened? Perhaps we need to add some task dependencies so that the Filters get executed before these editor UOs?

test/tests/depletion/openmc.i

pshriwise · 2025-06-17T03:48:42Z

Actually I think I got a bit farther now. I tried setting export RUNAPP_COMMAND="cardinal-opt -i openmc.i". The simulation starts running but then ends with
MooseControl.app: *** ERROR ***
MooseControl.app: CLI_ARGS:1: (UserObjects/openmc_filter1/filter_id):
MooseControl.app:     Filter 1 does not exist and create_filter is false
I think this is probably due to some source code changes in Cardinal which have happened over the last couple of months from the new Tallies syntax. It seems that the filters have not yet been created by the time the OpenMCDomainFilterEditor is trying to use them.

This is strange to me b/c the create_filter parameter is set to true for all filter editors created by the CardinalOperator class. Is this reflected in the contents of the server.i file that's produced by that object?

nuclearkevin · 2025-06-17T11:48:56Z

This is strange to me b/c the create_filter parameter is set to true for all filter editors created by the CardinalOperator class. Is this reflected in the contents of the server.i file that's produced by that object?

It appears as if server.i sets create_filter = true. This is what I get when I run the depletion example:

cardinal-opt
-i
openmc.i
Controls/web_server/type=WebServerControl
Controls/web_server/execute_on="TIMESTEP_BEGIN TIMESTEP_END"
UserObjects/openmc_mat1/type=OpenMCNuclideDensities
UserObjects/openmc_mat1/material_id=1
UserObjects/openmc_mat1/names="U234 U235 U238 O16 O17 U236"
UserObjects/openmc_mat1/densities="4.989113525358761e-06 0.000558182473562055 0.02240526017807056 0.04592456500869191 1.7412009289815074e-05 2.55674383288874e-06"
UserObjects/openmc_mat2/type=OpenMCNuclideDensities
UserObjects/openmc_mat2/material_id=2
UserObjects/openmc_mat2/names="He3 He4"
UserObjects/openmc_mat2/densities="4.808563327759458e-10 0.00024042768553164014"
UserObjects/openmc_mat3/type=OpenMCNuclideDensities
UserObjects/openmc_mat3/material_id=3
UserObjects/openmc_mat3/names="Cr50 Cr52 Cr53 Cr54 Fe54 Fe56 Fe57 Fe58 Sn112 Sn114 Sn115 Sn116 Sn117 Sn118 Sn119 Sn120 Sn122 Sn124 Zr90 Zr91 Zr92 Zr94 Zr96"
UserObjects/openmc_mat3/densities="3.296179086851863e-06 6.356393983859786e-05 7.207577897639481e-06 1.7941072349310136e-06 6.812040282197712e-06 0.00010693480683272309 2.469584661632909e-06 3.286524014540826e-07 4.512345343746015e-06 3.070250173679606e-06 1.5816575713225518e-06 6.763901433813293e-05 3.572667675644065e-05 0.00011266935877799354 3.995997624250037e-05 0.00015155988781410216 2.153844568759417e-05 2.693460658525746e-05 0.021876516442070146 0.004770741147746929 0.0072921764132952375 0.007389971148193187 0.001190558805034355"
UserObjects/openmc_mat4/type=OpenMCNuclideDensities
UserObjects/openmc_mat4/material_id=4
UserObjects/openmc_mat4/names="B10 B11 H1 H2 O16 O17"
UserObjects/openmc_mat4/densities="8.133537263121287e-06 3.2903482227904384e-05 0.051288285482012676 7.988881769415422e-06 0.024612879876383137 9.331818232259238e-06"
UserObjects/openmc_tally1/type=OpenMCTallyEditor
UserObjects/openmc_tally1/create_tally = true
UserObjects/openmc_tally1/tally_id=1
UserObjects/openmc_tally1/scores=""
UserObjects/openmc_tally1/nuclides=""
UserObjects/openmc_tally1/filter_ids=""
UserObjects/openmc_tally1/multiply_density=false
UserObjects/openmc_filter1/type=OpenMCDomainFilterEditor
UserObjects/openmc_filter1/create_filter = true
UserObjects/openmc_filter1/filter_id="1"
UserObjects/openmc_filter1/filter_type="material"
UserObjects/openmc_filter1/bins="1"

nuclearkevin · 2025-06-17T11:54:51Z

Ah, I know what's going on. create_filter = true should be create_filter=true and create_tally = true should be create_tally=true. Spaces in CLI arguments...

python/cardinal_operator.py

pshriwise · 2025-06-17T12:38:01Z

Ah, I know what's going on. create_filter = true should be create_filter=true and create_tally = true should be create_tally=true. Spaces in CLI arguments...

Ah good catch!

Co-authored-by: Kevin Sawatzky <66632997+nuclearkevin@users.noreply.github.com> Co-authored-by: Patrick Shriwise <pshriwise@gmail.com>

pshriwise · 2025-06-19T14:49:49Z

A couple of Python modules missing in the test env it seems (openmc and endf). Is it possible to add these to the CIVET environment @loganharbour?

loganharbour · 2025-06-20T01:41:38Z

A couple of Python modules missing in the test env it seems (openmc and endf). Is it possible to add these to the CIVET environment @loganharbour?

Yep. Is that it?

pshriwise · 2025-06-20T02:50:48Z

Yep. Is that it?

I think so. Thoughts on using PYTHONPATH for openmc so we capture any updates in future PRs?

The endf package is pretty stable and can be installed from PyPI.

aprilnovak · 2025-06-20T15:34:51Z

Trying to run this still - getting farther, but hitting this:

MooseControl.app: *** ERROR ***
MooseControl.app: Unable to open file "/home/ajnovak2/cardinal/test/tests/depletion/Controls/web_server/type=WebServerControl". Check to make sure that it exists and that you have read permission.
MooseControl.app:
MooseControl.app: Stack frames: 10
MooseControl.app: 0: libMesh::print_trace(std::ostream&)
MooseControl.app: 1: moose::internal::mooseErrorRaw(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >)
MooseControl.app: 2: void mooseError<char const*>(char const*&&)
MooseControl.app: 3: MooseUtils::checkFileReadable(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, bool, bool, bool)
MooseControl.app: 4: Parser::parse()
MooseControl.app: 5: Moose::createMooseApp(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, int, char**)
MooseControl.app: 6: cardinal-opt(+0x5611) [0x56478d22b611]
MooseControl.app: 7: /lib/x86_64-linux-gnu/libc.so.6(+0x29d90) [0x7fed6f251d90]
MooseControl.app: 8: __libc_start_main
MooseControl.app: 9: cardinal-opt(+0x58c5) [0x56478d22b8c5]

@loganharbour are there some files missing from this PR, or something needs to be updated still?

loganharbour · 2025-06-20T16:29:12Z

It looks like arguments aren't being passed right to MooseControl; it looks like it's missing the path to the input file?

aprilnovak · 2025-06-20T19:27:45Z

Can you help with debugging? I'm not sure on how these files are meant to look since I think this is the only example case.

loganharbour force-pushed the radiant_depletion branch from 95b5008 to 5e3baba Compare June 28, 2024 01:51

pshriwise requested changes Jul 8, 2024

View reviewed changes

loganharbour force-pushed the radiant_depletion branch from 08524b4 to 470411b Compare October 29, 2024 18:39

aprilnovak mentioned this pull request Jan 17, 2025

Tally and Domain Filter Editor UOs #923

Merged

4 tasks

pshriwise and others added 5 commits January 29, 2025 16:32

Adding the pincell test model and the test file for a simple depletio…

328b918

…n chain

Update scripts to use MooseControl

85f3c68

Add initial test for depletion; still a work in progress

6594442

Adding updates for use of the tally/filter editors

ba8812e

Updating test input to the new tally system

ea9034b

pshriwise force-pushed the radiant_depletion branch from aa7581f to ea9034b Compare February 3, 2025 22:08

aprilnovak reviewed Jun 10, 2025

View reviewed changes

test/tests/depletion/openmc.i Outdated Show resolved Hide resolved

nuclearkevin reviewed Jun 17, 2025

View reviewed changes

python/cardinal_operator.py Outdated Show resolved Hide resolved

python/cardinal_operator.py Outdated Show resolved Hide resolved

Apply suggestions from code review

791f7ac

Co-authored-by: Kevin Sawatzky <66632997+nuclearkevin@users.noreply.github.com> Co-authored-by: Patrick Shriwise <pshriwise@gmail.com>

OpenMC depletion coupling #921

Are you sure you want to change the base?

OpenMC depletion coupling #921

Uh oh!

Conversation

loganharbour commented Jun 26, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

loganharbour commented Jun 26, 2024

Uh oh!

moosebuild commented Jun 26, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pshriwise left a comment

Choose a reason for hiding this comment

Uh oh!

pshriwise Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

pshriwise Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

Uh oh!

pshriwise Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

pshriwise Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

aprilnovak Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

pshriwise Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

pshriwise Jul 8, 2024

Choose a reason for hiding this comment

Uh oh!

Uh oh!

moosebuild commented Oct 29, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

aprilnovak commented Jan 27, 2025

Uh oh!

pshriwise commented Feb 4, 2025

Uh oh!

loganharbour commented Feb 6, 2025

Uh oh!

aprilnovak commented Jun 10, 2025

Uh oh!

loganharbour commented Jun 10, 2025

Uh oh!

aprilnovak commented Jun 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

aprilnovak commented Jun 10, 2025

Uh oh!

Uh oh!

pshriwise commented Jun 17, 2025

Uh oh!

nuclearkevin commented Jun 17, 2025

Uh oh!

nuclearkevin commented Jun 17, 2025

Uh oh!

Uh oh!

Uh oh!

pshriwise commented Jun 17, 2025

Uh oh!

pshriwise commented Jun 19, 2025

Uh oh!

loganharbour commented Jun 20, 2025

Uh oh!

pshriwise commented Jun 20, 2025

Uh oh!

aprilnovak commented Jun 20, 2025

Uh oh!

loganharbour commented Jun 20, 2025

Uh oh!

aprilnovak commented Jun 20, 2025

Uh oh!

loganharbour commented Jun 26, 2024 •

edited

Loading

moosebuild commented Jun 26, 2024 •

edited

Loading

moosebuild commented Oct 29, 2024 •

edited

Loading

aprilnovak commented Jun 10, 2025 •

edited

Loading