Photon Transport Simulation with Two Layers of Material & Unstructured Mesh #27

anu1217 · 2024-10-04T15:02:31Z

Performing OpenMC photon transport using photon source generated by R2S Step2
- Source_Mesh_Reader extracts source density data from R2S-generated meshes
- OpenMC_PhotonTransport contains main simulation
- Photon_Tallytovtk reads data from statepoint file

This model currently uses the relative source intensities between different photon groups as the source strengths (and samples equally between individual mesh elements for any given energy group). However, it should be possible to add spatial variation of source intensity onto the MeshSpatial distribution.

gonuke

Glad to see this is all (mostly) working

OpenMC_PhotonTransport.py

Co-authored-by: Paul Wilson <[email protected]>

- Reading total strengths directly from Source_Mesh_Reader instead of writing a new list in this script

- Brought the MeshBase object inside the loop over the energy bounds

Assign material library path to new function and delete unused lines

These two files can be imported into a neutron or photon transport problem

gonuke

A number of suggestions here. We should meet to discuss more.

I think all of these files should be in the WC_Layers directory, and the neutron transport file should be updated to use the same geometry and materials scripts.

TwoLayers_Geometry.py

gonuke · 2024-10-21T13:46:02Z

TwoLayers_Geometry.py

+@author: Anupama Rajendra
+"""
+
+import openmc


Not clear that we need to import openmc here, if this method is only ever used when importing into another script that also imports openmc

I might be missing something here, but OpenMC_PhotonTransport attempts to access openmc inside TwoLayers_Geometry even when only make_spherical_shells is imported

TwoLayers_Geometry.py

gonuke · 2024-10-21T14:01:24Z

Source_Mesh_Reader.py

+    tstt = file['tstt']
+    elements = tstt['elements']
+    tet_four = elements['Tet4']
+    tags = tet_four['tags']
+    sd = tags['source_density'][:]
+
+    # Write source density to a separate file for each mesh
+    with open(f'source_density_{file_index + 1}.txt', 'w') as source:
+        for tet_element in sd:
+            source.write(' '.join(map(str, tet_element)) + '\n')
+
+    # Calculate summed (and individual mesh) strengths for each photon group
+    summed_strengths = []
+    strengths_list = []
+    for group in range(photon_groups):
+        total_strengths = np.sum(sd[:, group])  # Sum over all mesh elements
+        summed_strengths.append(total_strengths)
+        strengths = sd[:,group]
+        strengths_list.append(strengths)


I recommend putting this in a method that does everything you need for a single file. That kind of modularity enhances readability. You might also want to separate the functions of reading from a file, processing the data, and writing to files, so methods like extract_source_data, arrange_data, write_source_density

Source_Mesh_Reader.py

Photon_TallytoVtk.py

OpenMC_PhotonTransport.py

Co-authored-by: Paul Wilson <[email protected]>

- Created dictionary with material densities - Created OpenMC Material and Materials objects with separate function

gonuke

More suggestions as you chip away at this

TwoLayers_Materials.py

gonuke · 2024-11-10T21:59:19Z

TwoLayers_Geometry.py

+
+# Create geometry
+#Spherical shell:
+def make_spherical_shell(inner_radius_W, outer_radius_W, inner_radius_C, M_1, M_2):    


This makes multiple shells, so maybe change the name.

Could also generalize this to take an inner radius plus layers:

Suggested change

def make_spherical_shell(inner_radius_W, outer_radius_W, inner_radius_C, M_1, M_2):

def make_spherical_shells(inner_radius_W, layers):

'''

Builds a set of nested spherical shells, with an inner_radius and

multiple layers each with a material definition and a thickness

inputs

=====

inner_radius : the radius of the inner surface of the first shell

layers : list of tuples where each tuple is (material, thickness)

'''

I think materials will need to be an input as this function is also assigning materials to the geometry. I'm wondering if this can be done with one set of radii/thickness inputs and the materials from TwoLayers_Materials.

Suggested change

def make_spherical_shell(inner_radius_W, outer_radius_W, inner_radius_C, M_1, M_2):

def make_spherical_shell(radii, materials):

'''

Takes a list of materials and radii (outermost to innermost)

and creates an OpenMC Geometry object.

element_list = list of elements (str) in order from radially outermost to innermost

radii = list of the radii (outermost to innermost) that serve as the boundaries of the

spherical shells; must be in the same order as the specified materials

'''

cells = []

for index, material in enumerate(materials):

outer_sphere = openmc.Sphere(r = radii[index]) # add boundary_type here

inner_sphere = openmc.Sphere(r = radii[index + 1])

element_region = +inner_sphere & -outer_sphere

element_shell = openmc.Cell(fill = materials[index], region = element_region)

cells.append(element_shell)

if material == materials[-1]:

void = openmc.Cell(fill = None, region = -inner_sphere)

cells.append(void)

geometry = openmc.Geometry(cells)

return geometry

I agree that the materials are necessary. My version had them bound in the layers. This kind of data structure can lead to clarity because the data gets bound together in instructive ways. An example of calling it might be:

inner_radius = 10.0 # material_a and material_b are OpenMC Materials layers = [(material_a, 5.0), (material_b, 4.0)] geometry = make_spherical_shells(inner_radius, layers)

The function may be:

def make_spherical_shells(inner_radius, layers): inner_sphere = openmc.Sphere(inner_radius) cells = [openmc.Cell(region=-inner_sphere)] for (material, thickness) in layers: outer_radius = inner_radius + thickness outer_sphere = openmc.Sphere(outer_radius) cells.append(openmc.Cell(fill = material, region = inner_sphere & -outer_sphere)) inner_radius = outer_radius inner_sphere = outer_sphere cells.append(openmc.Cell(region = outer_sphere) geometry = openmc.Geometry(cells) return geometry

(There may be minor syntax errors, but this is the basic idea)

Make dictionary of densities with ALARA element library Co-authored-by: Paul Wilson <[email protected]>

Redefine inner and outer W spheres Co-authored-by: Paul Wilson <[email protected]>

- Redefined make_spherical_shells to take layers and inner_radius as input - Deleted reference to TwoLayers_Materials to maintain generality

gonuke

A few more high level comments here. They will probably result in a lot of code changes, but shouldn't take too long. Conceptually, it's a lot of same changes across many files.

WC_Layers/Source_Mesh_Reader.py

WC_Layers/Photon_TallytoVtk.py

WC_Layers/OpenMC_PhotonTransport.py

…Mesh_Reader_Inputs.yaml

…Input/Conversion_h5m.py

…PyNE_Lib.py

gonuke

I didn't note all the places that comments from other PR's/projects might be applied here...

gonuke · 2024-12-23T19:41:05Z

WC_Layers/TwoLayers_Geometry.py

+        thicknesses: thickness of each OpenMC Material
+        inner_radius: the radius of the innermost spherical shell
+    '''
+    layers = zip(materials, thicknesses)


Might be better to zip these together before calling the function (or as you call the function). This reinforces that these need to be bound to each other in this way.

This function now takes layers as input.

gonuke · 2024-12-23T19:46:46Z

WC_Layers/Source_Mesh_Reader.py

+        inputs = yaml.safe_load(yaml_file)
+
+    source_mesh_list = inputs['source_meshes']
+    num_elements = inputs['num_elements']


Is the number of mesh elements defined by the mesh? Doesn't this introduce the possibility that there is a discrepancy introduced by the user?

We set it up this way so that sd_list could be a numpy array instead of a list. Specifying the wrong value of either the number of mesh elements or photon groups will produce an error

gonuke · 2024-12-23T19:48:45Z

WC_Layers/Source_Mesh_Reader.py

+    sd_list = np.ndarray((len(source_mesh_list), num_elements, photon_groups))
+    for source_index, source_name in enumerate(source_mesh_list):
+         file = h5py.File(source_name, 'r')
+         sd_list[source_index,:] = file['tstt']['elements']['Tet4']['tags']['source_density'][:]


sd_list is initialized as three-dimensional, but you only refer to 2 dimensions here. That may be OK if both colons refer to 2-D data?

Yes, both colons refer to the 2D photon_groups X num_elements data. This is just making a list of the 2D data for each decay step

WC_Layers/Source_Mesh_Reader.py

gonuke · 2024-12-23T19:52:44Z

WC_Layers/Source_Mesh_Reader.py

+
+    source_mesh_list = inputs['source_meshes']
+    num_elements = inputs['num_elements']
+    photon_groups = inputs['photon_groups']


Isn't the number of groups already defined internally to the mesh file?

Same as mentioned above regarding the number of mesh elements

gonuke · 2024-12-23T21:10:27Z

WC_Layers/Photon_TallytoVtk.py

+    summed_axes_dose_filter = inputs['summed_axes_dose_filter']
+    cd = calculate_dose(rs[2], summed_axes_dose_filter)
+
+    summed_axes_mesh_filter = inputs['summed_axes_mesh_filter']


Looking at the YAML, the meaning/interpretation of this is not very obvious/clear.

gonuke · 2024-12-23T21:11:42Z

WC_Layers/OpenMC_PhotonTransport.py

+    total_filter = openmc.MeshFilter(total_mesh)
+
+    photon_tally = openmc.Tally(tally_id=1, name="Photon tally")
+    photon_tally.scores = ['flux', 'elastic', 'absorption']


Are these scores relevant/interesting for your photon tally?

Probably want the bounds from your YAML file

I'm getting rid of the elastic tally as it is 0, but it might be interesting to see how the flux and absorption spectra evolve relative to each other (once bounds is applied as an energy filter)

photon_tally is now set up to record flux and absorption over bounds, but I'm wondering if the absorption score can just be moved to spectrum_tally instead

gonuke · 2024-12-23T21:12:32Z

WC_Layers/OpenMC_PhotonTransport.py

+        OpenMC Tallies object
+    '''
+    particle_filter = openmc.ParticleFilter('photon')
+    total_filter = openmc.MeshFilter(total_mesh)


I'm not sure what total means in these variable names?

I meant 'total' as in the unstructured mesh encompasses the full geometry, but I realize that's not clear from the variable name. Renamed total_mesh as unstructured_mesh.

gonuke · 2024-12-23T21:13:59Z

WC_Layers/OpenMC_PhotonTransport.py

+    dose_filter = openmc.EnergyFunctionFilter(dose_energy, dose)
+
+    # An energy filter is created to assign to the flux tally.
+    energy_filter_flux = openmc.EnergyFilter.from_group_structure("VITAMIN-J-42")


I'm not sure this is relevant for photons??

I thought the 42-group structure was for photons and the 175 one for neutrons...or should this change to match the group structure from ALARA?

gonuke · 2024-12-23T21:14:37Z

WC_Layers/OpenMC_PhotonTransport.py

+
+    spectrum_tally = openmc.Tally(tally_id=2, name="Flux spectrum")
+    # Implementing energy and cell filters for flux spectrum tally
+    spectrum_tally.filters = [cell_filter, total_filter, energy_filter_flux, particle_filter, dose_filter]


maybe you should apply the dose_filter to the photon_tally instead of the spectrum_tally?

If photon_tally is changed to include an energy filter with the ALARA bounds, I'm not sure it would matter which tally the dose filter is associated with

Co-authored-by: Paul Wilson <[email protected]>

anu1217 added 2 commits October 4, 2024 09:50

Add files via upload

23c1469

Deleted unused Energy_List lines

8d1f88b

gonuke requested changes Oct 4, 2024

View reviewed changes

OpenMC_PhotonTransport.py Outdated Show resolved Hide resolved

OpenMC_PhotonTransport.py Outdated Show resolved Hide resolved

OpenMC_PhotonTransport.py Outdated Show resolved Hide resolved

anu1217 and others added 10 commits October 10, 2024 16:19

Update OpenMC_PhotonTransport.py

d7e58a3

Co-authored-by: Paul Wilson <[email protected]>

Delete unused 'Energy_List.txt' line

39e2d4b

Delete print statement

3dfb567

Remove summed strengths text file as output

5bb02ff

Read summed photon strengths from Source_Mesh_Reader

1fdb1b1

- Reading total strengths directly from Source_Mesh_Reader instead of writing a new list in this script

Add source strengths for each individual mesh element

569eb28

- Brought the MeshBase object inside the loop over the energy bounds

Update OpenMC_PhotonTransport.py

1008270

Assign material library path to new function and delete unused lines

Adding materials/geometry files

19e86ef

These two files can be imported into a neutron or photon transport problem

Import materials/geometry from TwoLayers_Materials & TwoLayers_Geometry

6fa1367

Delete materials and geometry functions from current script

c3cf5db

anu1217 requested a review from gonuke October 15, 2024 19:36

gonuke requested changes Oct 21, 2024

View reviewed changes

gonuke reviewed Oct 21, 2024

View reviewed changes

OpenMC_PhotonTransport.py Outdated Show resolved Hide resolved

anu1217 and others added 2 commits October 25, 2024 12:41

Update OpenMC_PhotonTransport.py

e217e00

Co-authored-by: Paul Wilson <[email protected]>

Update TwoLayers_Materials.py

94519f9

- Created dictionary with material densities - Created OpenMC Material and Materials objects with separate function

gonuke requested changes Nov 10, 2024

View reviewed changes

anu1217 and others added 10 commits November 11, 2024 14:04

Update TwoLayers_Materials.py

3854660

Make dictionary of densities with ALARA element library Co-authored-by: Paul Wilson <[email protected]>

Update references to some variables

a3db102

Update TwoLayers_Geometry.py

4199c1b

Redefine inner and outer W spheres Co-authored-by: Paul Wilson <[email protected]>

Import variables from TwoLayers_Materials

fdcb11f

Update directory of OpenMC_PhotonTransport

2e209a6

Update directory of Photon_Tallytovtk

e8a305a

Update directory of TwoLayers_Geometry

3aa85b3

Update directory of TwoLayers_Materials

1cff395

Update directory of Source_Mesh_Reader

64d14d5

Update TwoLayers_Geometry.py

88a9e83

- Redefined make_spherical_shells to take layers and inner_radius as input - Deleted reference to TwoLayers_Materials to maintain generality

gonuke requested changes Nov 14, 2024

View reviewed changes

anu1217 added 17 commits December 16, 2024 11:34

Add files via upload

3b808c1

Merge branch 'cnerg:main' into mesh_photontransport

052b312

Add input to function

771b147

Add docstrings to functions

7f76500

Change variable names and inputs

2a7d52c

Change variable key

6b18246

Update PhotonTransport_Inputs.yaml

a3a9921

Fix typo in run mode

608adba

Update and rename Source_Mesh_Reader_Inputs.yaml to WC_Layers/Source_…

15f64ef

…Mesh_Reader_Inputs.yaml

Fix variable updating to avoid seg fault

8af7f12

Add main() function to compile all functions

8aad344

Calculate dose and add main() function

3c272b8

Update PhotonTransport_Inputs.yaml

95b52b1

Update Source_Mesh_Reader.py

2dde9f3

Create README.md

d07dd9d

Update and rename WC_Layers/Conversion_h5m.py to WC_Layers/R2S_Step1_…

da12a00

…Input/Conversion_h5m.py

Update and rename WC_Layers/PyNE_Lib.py to WC_Layers/R2S_Step1_Input/…

cc2dad7

…PyNE_Lib.py

anu1217 requested a review from gonuke December 20, 2024 17:26

anu1217 marked this pull request as ready for review December 20, 2024 17:28

gonuke requested changes Dec 23, 2024

View reviewed changes

anu1217 marked this pull request as draft January 7, 2025 15:28

anu1217 and others added 8 commits January 8, 2025 14:25

Modularize main() and change total_mesh to unstructured_mesh

589bff0

Modularize main()

d813414

Modularize main()

2ae754a

Change inputs for make_spherical_shells()

9e8bb76

Update WC_Layers/Source_Mesh_Reader.py

d3bc003

Co-authored-by: Paul Wilson <[email protected]>

Change variable names and function inputs

18ca1b5

Change variable names and function inputs

792717e

Add extra parameters

b1e219d

anu1217 marked this pull request as ready for review January 10, 2025 07:04

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Photon Transport Simulation with Two Layers of Material & Unstructured Mesh #27

Photon Transport Simulation with Two Layers of Material & Unstructured Mesh #27

anu1217 commented Oct 4, 2024

gonuke left a comment

gonuke left a comment •

edited

Loading

gonuke Oct 21, 2024

anu1217 Nov 26, 2024

gonuke Oct 21, 2024

gonuke left a comment

gonuke Nov 10, 2024

anu1217 Nov 13, 2024

gonuke Nov 14, 2024 •

edited

Loading

gonuke left a comment

gonuke left a comment

gonuke Dec 23, 2024

anu1217 Jan 8, 2025

gonuke Dec 23, 2024

anu1217 Jan 10, 2025

gonuke Dec 23, 2024

anu1217 Jan 8, 2025

gonuke Dec 23, 2024

anu1217 Jan 10, 2025

gonuke Dec 23, 2024

gonuke Dec 23, 2024

gonuke Dec 23, 2024

anu1217 Jan 9, 2025

anu1217 Jan 10, 2025

gonuke Dec 23, 2024

anu1217 Jan 8, 2025

gonuke Dec 23, 2024

anu1217 Jan 9, 2025

gonuke Dec 23, 2024

anu1217 Jan 9, 2025

-def make_spherical_shell(inner_radius_W, outer_radius_W, inner_radius_C, M_1, M_2):
+def make_spherical_shells(inner_radius_W, layers):
+    '''
+    Builds a set of nested spherical shells, with an inner_radius and
+    multiple layers each with a material definition and a thickness
+    inputs
+    =====
+    inner_radius : the radius of the inner surface of the first shell
+    layers : list of tuples where each tuple is (material, thickness)
+    '''

-def make_spherical_shell(inner_radius_W, outer_radius_W, inner_radius_C, M_1, M_2):
+def make_spherical_shell(radii, materials):
+    '''
+    Takes a list of materials and radii (outermost to innermost)
+    and creates an OpenMC Geometry object.
+    element_list = list of elements (str) in order from radially outermost to innermost
+    radii = list of the radii (outermost to innermost) that serve as the boundaries of the
+    spherical shells; must be in the same order as the specified materials
+    '''
+    cells = []
+    for index, material in enumerate(materials):
+        outer_sphere = openmc.Sphere(r = radii[index]) # add boundary_type here
+        inner_sphere = openmc.Sphere(r = radii[index + 1])
+        element_region = +inner_sphere & -outer_sphere
+        element_shell = openmc.Cell(fill = materials[index], region = element_region)
+        cells.append(element_shell)
+        if material == materials[-1]:
+            void = openmc.Cell(fill = None, region = -inner_sphere)
+            cells.append(void)
+    geometry = openmc.Geometry(cells)
+    return geometry

Photon Transport Simulation with Two Layers of Material & Unstructured Mesh #27

Are you sure you want to change the base?

Photon Transport Simulation with Two Layers of Material & Unstructured Mesh #27

Conversation

anu1217 commented Oct 4, 2024

gonuke left a comment

Choose a reason for hiding this comment

gonuke left a comment • 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

gonuke left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gonuke Nov 14, 2024 • edited Loading

Choose a reason for hiding this comment

gonuke left a comment

Choose a reason for hiding this comment

gonuke 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

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 left a comment •

edited

Loading

gonuke Nov 14, 2024 •

edited

Loading