adding SFR test cases #12

magnoxemo · 2024-11-08T16:37:39Z

No description provided.

lewisgross1296

A few things that can be applied throughout. Try and look for repeated code that could be turned into functions.

Also, you should probably remove the .jitcache (not something you need or should even version).

Worth considering doing this PR as a few other ones (for the various assemblies).

lewisgross1296 · 2025-02-21T20:01:08Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+###################################################################
+##########################   Materials     ########################
+###################################################################


Let's consider making a function that returns a dictionary of materials create_SFR_materials() and converting use of the material objects to material_dict["mat_name"]

lewisgross1296 · 2025-02-21T20:01:13Z

SFR/Assembly/inner_assembly/mesh.i

+        pattern ='0  0  0  0  0  0  0  0  0 ;
+                0  0  0  0  0  0  0  0  0  0 ;
+              0  0  0  0  0  0  0  0  0  0  0 ;
+            0  0  0  0  0  0  0  0  0  0  0  0 ;
+           0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+          0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+        0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+       0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+      0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+       0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+         0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+          0  0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+           0  0  0  0  0  0  0  0  0  0  0  0  0 ;
+            0  0  0  0  0  0  0  0  0  0  0  0 ;
+             0  0  0  0  0  0  0  0  0  0  0 ;
+              0  0  0  0  0  0  0  0  0  0 ;
+                0  0  0  0  0  0  0  0  0 '


I personally like to use one space between the entries in the pattern. This looks a little wonky (though I'm sure it generates the correct mesh.

lewisgross1296 · 2025-02-21T20:01:57Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+inner_fuel_material= openmc.Material.mix_materials([u235, u238, pu238, pu239, pu240, pu241, pu242, am241, o16],[0.0019, 0.7509, 0.0046, 0.0612, 0.0383, 0.0106, 0.0134, 0.001, 0.1181],'wo')
+cladding_material = openmc.Material.mix_materials([cu63,Al2O3], [0.997,0.003], 'wo')


Probably applicable other places, but we should maybe Black format your python code

lewisgross1296 · 2025-02-21T20:10:06Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+fuel_or = openmc.ZCylinder(surface_id=1, r=0.943/2) 
+clad_ir = openmc.ZCylinder(surface_id=2, r=0.973/2) 
+clad_or = openmc.ZCylinder(surface_id=3, r=1.073/2) 


Not sure you need these surface_ids here. Also we should probably use variables to define the radii

lewisgross1296 · 2025-02-21T20:14:52Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+# Create the argument parser
+ap = ArgumentParser(description="SFR Pincell Model Generator")
+ap.add_argument('-n', dest='n_axial', type=int, default=100, help='Number of cells in the Z direction')
+ap.add_argument('-e', '--shannon_entropy', action='store_true', help='Add Shannon entropy mesh')
+ap.add_argument('--height', dest='height_of_the_core', type=float, default=30.0, help='Height of the reactor core')
+
+# Parse the arguments
+arguments = ap.parse_args()
+
+arguments=ap.parse_args()
+
+N=arguments.n_axial
+height=arguments.height_of_the_core
+shannon_entropy=arguments.shannon_entropy


let's do this in a main() function and throw a

if __name__ == "__main__": main()

at the bottom.

I think a function like build_SFR_model() that actually does all the model generation should be called form main()

lewisgross1296 · 2025-02-21T20:15:15Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+# settings 
+settings=openmc.Settings()
+settings.batches=200
+settings.inactive=40
+settings.particles=20000
+
+
+if (shannon_entropy):
+  entropy_mesh = openmc.RegularMesh()
+  entropy_mesh.lower_left = lower_left
+  entropy_mesh.upper_right = upper_right
+  entropy_mesh.dimension = (10, 10, 20)
+  settings.entropy_mesh = entropy_mesh
+
+settings.temperature = {'default': 280.0 + 273.15,
+                        'method': 'interpolation',
+                        'range': (294.0, 3000.0),
+                        'tolerance': 1000.0}
+
+settings.export_to_xml()                      


we can make a settings function too

lewisgross1296 · 2025-02-21T20:15:44Z

SFR/Assembly/inner_assembly/mesh.i

+#----------------------------------------------------------------------------------------
+# Assembly geometrical information
+#----------------------------------------------------------------------------------------
+pitch        = 1.25984
+height       = 300.0
+r_fuel       = 0.4715
+t_gap        = 0.0150
+r_clad_inner = 0.4865
+t_clad       = 0.05
+edge_length  =12.1705
+#----------------------------------------------------------------------------------------


would be nice to have a common_input.i and

!include common_input.i

at the top

lewisgross1296 · 2025-02-21T20:27:27Z

SFR/Assembly/outer_assembly/make_SFR_assembly_model.py

+# Define a lattice for outer assemblies
+out_lat = openmc.HexLattice(lattice_id=1, name='outer assembly')
+out_lat.center = (0., 0.)
+out_lat.pitch = (21.08/17,)
+out_lat.orientation = 'y'
+out_lat.outer = sodium_mod_u


consider using a 3D lattice, you can look at the GCMR repo to see an example of how to do this and still have density feedback

aprilnovak · 2025-02-26T01:18:01Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+
+# Create the argument parser
+ap = ArgumentParser(description="SFR Pincell Model Generator")
+ap.add_argument('-n', dest='n_axial', type=int, default=100, help='Number of cells in the Z direction')


How about a default of 1? I think that makes sense as a default because that would be the best performance choice if not needing extra cells for depletion/feedback.

aprilnovak · 2025-02-26T01:19:07Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+
+#Material mixtures
+
+inner_fuel_material= openmc.Material.mix_materials([u235, u238, pu238, pu239, pu240, pu241, pu242, am241, o16],[0.0019, 0.7509, 0.0046, 0.0612, 0.0383, 0.0106, 0.0134, 0.001, 0.1181],'wo')


This would be much more efficient (less code) to create this object using a single openmc.Material (as long as you know the overall fuel density from somewhere?)

aprilnovak · 2025-02-26T01:20:22Z

SFR/Assembly/inner_assembly/make_SFR_assembly_model.py

+
+for i in range (N):
+
+    layer=+z_plane[i]&-z_plane[i+1]


This may have been from an example, but why not use a lattice in the z-direction? I would expect the tracking performance to be a little bit better that way, as opposed to N unique cells.

Agreed. Check out my GCMR code for an example with a 3D lattice. There might be some depletion things you don't need in there but it should serve as a good base. IIRC there's some cardinal tutorials that do it too

This may be based on the approach that @nuclearkevin used for LWR problems?

My bad, I actually didn't know that lattices could be defined in 3D. I'll open a PR at some point over the next few days to update the LWR cases so they use a 3D lattice.

SFR/Assembly/inner_assembly/mesh.i

aprilnovak · 2025-02-26T01:22:00Z

SFR/Assembly/inner_assembly/mesh.i

+        direction = '0.0 0.0 1.0'
+
+        bottom_boundary = '10001'
+        top_boundary = '10000'


I know the reactor module likes to use the sideset 10000 as the default outer boundary (if my memory is correct). Did you check that there's not any other existing sidesets numbered 10000 that might be conflicting with this one?

Thank you April. I removed that default top and bottom name and generated the mesh again which I think looks OK. But I hope someone else could cross check maybe?

aprilnovak · 2025-02-26T01:22:41Z

SFR/Assembly/inner_assembly/openmc.i

+  []
+[]
+
+[Adaptivity]


Because we may want to run a whole series of models through different AMR strategies, can you please update these .i files to use the !include strategy idea from #1?

All the amr studies I have for the SFR models are direct implementation of Kevin's value_jump_rel_error. I can remove the adaptivity block from my SFR models since value_jump_rel_error already does it?

I think April is referring to this

cardinal_amr/README.md

Lines 20 to 25 in 221a7f1

In addition to these models, this repository also contains a series of AMR algorithms - these can be found in `/amr_strategies/*`. To apply one of these AMR strategies to a test model in this repository, follow the steps below:

1. Add an `!include STRATEGY_NAME` to the beginning of the `openmc.i` file in one of the model sub directories, where `STRATEGY_NAME` is the name of a file in `amr_strategies`.

2. Run `python make_openmc_model.py`.

3. Run `cardinal-opt -i mesh.i --mesh-only`.

4. Run `mpiexec -n NUM_PROC cardinal-opt openmc.i --n-threads=NUM_THREADS`, replacing `NUM_PROC` with the number of MPI processors and `NUM_THREADS` with the number of OpenMP threads.

gonuke · 2025-02-26T01:46:58Z

In general, I encourage you to look at the changes that @nuclearkevin made during review to the LWR cases and see how you would make changes to match the design patterns used there.

gonuke · 2025-02-26T01:49:13Z

The *.so files shouldn't be here. You might want to update the .gitignore and get used to adding files explicitly to each commit rather than git add -A (or other shortcuts that can risk adding too many files)

magnoxemo · 2025-02-28T18:51:54Z

I think this is ready for second round of review .

magnoxemo and others added 19 commits October 21, 2024 13:07

adding SFR pincell test case input files

2bc1c95

deleting the mesh_in.e file

1b5da78

adding outer fuel pincell test input files

601f4e4

adding SFR inner pincell test case inputfiles

5d58f0e

adding gas gap in the final pincell

4f56535

adding assembly test case (incomplete)

81d1876

deleting the gas gap in pincell model

c9a190b

mesh works (maybe)

e56e9f3

source distribution needs be declared in the script

9391454

adding outer assembly

a7b3609

fixing some bugs

938047f

now on my machine (or maybe) the SFR inner and outer assembly works

76d1e2e

now works (maybe)

95e5b95

fixed the hexlattice orientation bug

21359cf

changing the marker to or combo to see what happends when I run the code

504b044

setting BooleanCombomarker to or mode

59d71a0

adding z_value_marker input files for SFR pincell

4f73f9e

fixing a typo

1ab5d12

fixing a typo

560871b

lewisgross1296 suggested changes Feb 21, 2025

View reviewed changes

aprilnovak reviewed Feb 26, 2025

View reviewed changes

SFR/Assembly/inner_assembly/mesh.i Outdated Show resolved Hide resolved

aprilnovak reviewed Feb 26, 2025

View reviewed changes

magnoxemo and others added 2 commits February 27, 2025 16:40

Merge branch 'neams-th-coe:main' into walid

a2803ee

adding changes

dfa8f57

applying black format

57f03dc

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

adding SFR test cases #12

adding SFR test cases #12

magnoxemo commented Nov 8, 2024

lewisgross1296 left a comment

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

lewisgross1296 Feb 21, 2025

aprilnovak Feb 26, 2025

aprilnovak Feb 26, 2025 •

edited

Loading

aprilnovak Feb 26, 2025

lewisgross1296 Feb 26, 2025

gonuke Feb 26, 2025

nuclearkevin Feb 27, 2025

aprilnovak Feb 26, 2025

magnoxemo Feb 28, 2025

aprilnovak Feb 26, 2025

magnoxemo Feb 28, 2025

lewisgross1296 Feb 28, 2025

gonuke commented Feb 26, 2025

gonuke commented Feb 26, 2025

magnoxemo commented Feb 28, 2025

		inner_fuel_material= openmc.Material.mix_materials([u235, u238, pu238, pu239, pu240, pu241, pu242, am241, o16],[0.0019, 0.7509, 0.0046, 0.0612, 0.0383, 0.0106, 0.0134, 0.001, 0.1181],'wo')
		cladding_material = openmc.Material.mix_materials([cu63,Al2O3], [0.997,0.003], 'wo')


		#Material mixtures

		inner_fuel_material= openmc.Material.mix_materials([u235, u238, pu238, pu239, pu240, pu241, pu242, am241, o16],[0.0019, 0.7509, 0.0046, 0.0612, 0.0383, 0.0106, 0.0134, 0.001, 0.1181],'wo')

	In addition to these models, this repository also contains a series of AMR algorithms - these can be found in `/amr_strategies/*`. To apply one of these AMR strategies to a test model in this repository, follow the steps below:

	1. Add an `!include STRATEGY_NAME` to the beginning of the `openmc.i` file in one of the model sub directories, where `STRATEGY_NAME` is the name of a file in `amr_strategies`.
	2. Run `python make_openmc_model.py`.
	3. Run `cardinal-opt -i mesh.i --mesh-only`.
	4. Run `mpiexec -n NUM_PROC cardinal-opt openmc.i --n-threads=NUM_THREADS`, replacing `NUM_PROC` with the number of MPI processors and `NUM_THREADS` with the number of OpenMP threads.

adding SFR test cases #12

Are you sure you want to change the base?

adding SFR test cases #12

Conversation

magnoxemo commented Nov 8, 2024

lewisgross1296 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

aprilnovak Feb 26, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gonuke commented Feb 26, 2025

gonuke commented Feb 26, 2025

magnoxemo commented Feb 28, 2025

aprilnovak Feb 26, 2025 •

edited

Loading