cad_geometry_example

doc/content/bib/references.bib

@@ -30,3 +30,45 @@ @article{gall2024verification
   year={2024},
   publisher={ANS PBNC Conference}
 }
+
+@article{novak2022_cardinal,
+   author = {A.J. Novak and D. Andrs and P. Shriwise and J. Fang and H. Yuan and D. Shaver and E. Merzari and P.K. Romano and R.C. Martineau},
+    title = {{Coupled {Monte} {Carlo} and Thermal-Fluid Modeling of High Temperature Gas Reactors Using {Cardinal}}},
+  journal = {Annals of Nuclear Energy},
+     year = 2022,
+   volume = 177,
+    pages = {109310},
+      DOI = {10.1016/j.anucene.2022.109310},
+      url = {https://www.sciencedirect.com/science/article/pii/S0306454922003450}
+}
+
+@article{giudicelli2024moose,
+   title = {3.0 - {MOOSE}: Enabling massively parallel multiphysics simulations},
+   author = {Guillaume Giudicelli and Alexander Lindsay and Logan Harbour and Casey Icenhour and
+             Mengnan Li and Joshua E. Hansel and Peter German and Patrick Behne and Oana Marin and
+             Roy H. Stogner and Jason M. Miller and Daniel Schwen and Yaqi Wang and Lynn Munday and
+             Sebastian Schunert and Benjamin W. Spencer and Dewen Yushu and Antonio Recuero and
+             Zachary M. Prince and Max Nezdyur and Tianchen Hu and Yinbin Miao and
+             Yeon Sang Jung and Christopher Matthews and April Novak and Brandon Langley and
+             Timothy Truster and Nuno Nobre and Brian Alger and David Andr{\v{s}} and
+             Fande Kong and Robert Carlsen and Andrew E. Slaughter and John W. Peterson and
+             Derek Gaston and Cody Permann},
+    year = {2024},
+ journal = {{SoftwareX}},
+  volume = {26},
+   pages = {101690},
+    issn = {2352-7110},
+     doi = {https://doi.org/10.1016/j.softx.2024.101690},
+     url = {https://www.sciencedirect.com/science/article/pii/S235271102400061X},
+keywords = {Framework, Finite-element, Finite-volume, Parallel, Multiphysics, Multiscale},
+}
+
+@Article{openmc,
+   author = {P.K. Romano and N.E. Horelik and B.R. Herman and A.G. Nelson and B. Forget and K. Smith},
+    title = {{OpenMC: A State-of-the-Art {Monte} {Carlo} Code for Research and Development}},
+  journal = {Annals of Nuclear Energy},
+     year = 2015,
+   volume = 82,
+    pages = {90--97},
+      DOI = {10.1016/j.anucene.2014.07.048}
+}

doc/content/verification_validation_examples/cad_geometry_model/cad_model.md

@@ -0,0 +1,122 @@
+# CAD-based Geometry Workflow for Multiphysics Fusion Problems Using OpenMC and MOOSE
+
+This demonstration describes a workflow for modeling fusion problems in OpenMC and MOOSE using a computer aided design (CAD)-based geometry workflow.
+It is based on the work published in [!cite](Eltawila2024PBNC). 
+
+In this example, you'll learn how to:
+
+- Couple OpenMC [!cite](openmc) and MOOSE [!cite](giudicelli2024moose) using Cardinal for fixed source Monte Carlo calculations.
+- Use Cardinal [!cite](novak2022_cardinal) to tally values of interest such as tritium production and heating which would be used in MOOSE to solve for the temperature distribution
+
+!media figures/transfers.png
+  id=transfers
+  caption=OpenMC and MOOSE Coupling
+  style=width:60%;margin-left:auto;margin-right:auto
+
+An extremely simplified tokamak was modeled in CAD and was considered for this example. The meshed geometry was prepared using direct accelerated geometry Monte Carlo (DAGMC) for particle transport, and a volumetric mesh was also prepared to be used in MOOSE’s finite element solver and to tally OpenMC results for heat source distribution and tritium production. Cardinal was used to run OpenMC Monte Carlo particle transport within MOOSE framework. The data transfer system transferred heat source and temperature distribution between OpenMC and MOOSE as shown in [transfers], with coupling between neutron transport and heat conduction achieved via Picard iteration.
+
+## Generating the meshes
+
+The CAD model was first developed in FUSION360 and was imported into Cubit to assign blocks, materials, and side sets and generate the mesh ([volumetric_mesh]). A corresponding DAGMC surface mesh ([dagmc]) was exported directly from the meshed geometry in Cubit (by loading the volumetric meshed geometry in Cubit and exporting a DAGMC surface mesh).
+
+In this example, `tmesh_1.e` ([volumetric_mesh]) is the finite element mesh used in MOOSE on which the heat conduction physics is solved. `tmesh_1.h5m` ([dagmc]) is the DAGMC surface mesh used for particle transport in OpenMC (which bounds the surfaces between different materials). Cardinal also allows for mesh tallying for tallying OpenMC results directly on the mesh overlayed on the OpenMC geometry which `tmesh_1.e` ([volumetric_mesh]) could be used for as well as an unstructured volume mesh. This could be used by changing the tally type and adding a mesh template (`tally_type = mesh`, `mesh_template = tmesh_1.e`) in Cardinal input under Problem.
+
+ 
+
+!row! style=display:inline-flex;
+!col! small=12 medium=4 large=3
+
+!media figures/mesh_1.png 
+  style=width:130%;display:block;
+  id=volumetric_mesh 
+  caption=Volumetric mesh
+
+!col-end!
+
+!col! small=12 medium=4 large=3
+
+!media figures/d1.png 
+  style=width:130%;display:block;
+  id=dagmc 
+  caption=DAGMC mesh
+
+!col-end!
+!row-end!
+
+## OpenMC
+
+The OpenMC input files is as follows:
+
+!listing /test/tests/cad_geometry_example/model.py language=python
+
+## Cardinal
+
+The Cardinal input files is as follows:
+
+!listing /test/tests/cad_geometry_example/openmc.i
+
+## FENIX (MOOSE) Heat transfer
+
+The FENIX input files is as follows:
+
+!listing /test/tests/cad_geometry_example/solid.i
+
+## Execution
+
+To generate OpenMC xml files, run:
+
+```
+python model.py
+```
+
+Then to run the coupled calculation:
+
+```
+mpiexec -np 2 fenix-opt -i solid.i --n-threads=2
+```
+
+This will run both MOOSE and OpenMC (with Cardinal) with 2 MPI processes and 2 OpenMP threads per rank. To run the simulation faster, you can increase the parallel processes/threads, or simply decrease the number of particles used in OpenMC. When the simulation has completed, you will have created a number of different output files:
+
+- `solid_out.e`, an Exodus output with the solid mesh and solution from the MOOSE/FENIX input file
+- `solid_out_openmc0.e`, an Exodus output with the OpenMC solution and the data that was ultimately transferred in/out of OpenMC
+
+## Results
+
+ 
+
+The results of this simulation are shown in [temps], [h3production], and [results]. More information - including a mesh sensitivity study and results from a finer mesh - is available in [!cite](Eltawila2024PBNC).
+
+ 
+
+!row! style=display:inline-flex;
+!col! small=12 medium=4 large=3
+
+!media figures/Temps.png 
+  style=width:130%;display:block;
+  id=temps 
+  caption=Temperature distribution
+
+!col-end!
+
+!col! small=12 medium=4 large=3
+
+!media figures/tritium_production.png 
+  style=width:130%;display:block;
+  id=h3production 
+  caption=Tritium production rate density
+
+!col-end!
+!row-end!
+
+ 
+
+!table id=results caption=Results summary
+| Parameter (Units) | Value |
+| :- | :- |
+| Armor Max. Temp. (K)         | 1062.4                   |
+| First Wall Max. Temp. (K)    | 1057.6                   |
+| Breeder Max. Temp. (K)       | 987.4                    |
+| Heat Source (W)              | 2.44 × 10^5^ ± 3 × 10^3^   |
+| Tritium Production (atoms/s) | 4.70 × 10^13^ ± 8 × 10^11^ |
+
+ 

doc/content/verification_validation_examples/index.md

@@ -35,6 +35,5 @@ FENIX is under active development and does not currently have any benchmarking c
 
 # List of example cases
 
-!alert construction title=Under development - no example cases are available yet
-FENIX is under active development and does not currently have any example cases available to users.
+- [CAD-based geometry workflow example for fusion problems using OpenMC in FENIX](cad_geometry_model/cad_model.md) based on [!cite](Eltawila2024PBNC).
 

test/tests/cad_geometry_example/gold/solid_out.csv

@@ -0,0 +1,4 @@
+time,max_T,source_integral
+0,0,0
+1,800,244158.40396983
+2,1058.7439583588,243141.87849132

test/tests/cad_geometry_example/gold/solid_out_openmc0.csv

@@ -0,0 +1,4 @@
+time,heat_source,heat_source_RelativeError,tritium_RelativeError,tritium_production
+0,0,0,0,0
+1,244158.40396983,0.0039604975020543,0.0030447507289791,47082007910425
+2,243141.87849132,0.39003312775355,0.39353117228504,46825507978724

test/tests/cad_geometry_example/model.py

@@ -0,0 +1,100 @@
+import openmc
+import math
+
+mat1 = openmc.Material(name="mat1")
+mat1.set_density('g/cc', 19.30)
+mat1.add_element('W', 1.0)
+
+eurofer = openmc.Material(name="eurofer")
+eurofer.add_element('Fe', 0.011   , 'wo')
+eurofer.add_element('Al', 0.002   , 'wo')
+eurofer.add_element('As', 0.0002  , 'wo')
+eurofer.add_element('B', 0.0012  , 'wo')
+eurofer.add_element('C', 0.0005  , 'wo')
+eurofer.add_element('Co', 0.0005  , 'wo')
+eurofer.add_element('Cr', 0.0005  , 'wo')
+eurofer.add_element('Cu', 0.00005 , 'wo')
+eurofer.add_element('Mn', 0.00005 , 'wo')
+eurofer.add_element('Mo', 0.0001  , 'wo')
+eurofer.add_element('N', 0.0001  , 'wo')
+eurofer.add_element('Nb', 0.00005 , 'wo')
+eurofer.add_element('Ni', 0.0003  , 'wo')
+eurofer.add_element('O', 0.00005 , 'wo')
+eurofer.add_element('P', 0.004   , 'wo')
+eurofer.add_element('S', 0.0001  , 'wo')
+eurofer.add_element('Sb', 0.09    , 'wo')
+eurofer.add_element('Sn', 0.0001  , 'wo')
+eurofer.add_element('Si', 0.0011  , 'wo')
+eurofer.add_element('Ta', 0.00002 , 'wo')
+eurofer.add_element('Ti', 0.0005  , 'wo')
+eurofer.add_element('V', 0       , 'wo')
+eurofer.add_element('W', 0.0001  , 'wo')
+eurofer.add_element('Zr', 0.88698 , 'wo')
+eurofer.set_density("g/cm3", 7.798)
+
+Helium = openmc.Material(name="Helium")
+Helium.add_element('He', 1.0)
+Helium.set_density("kg/m3", 0.166)
+
+mat2 = openmc.Material.mix_materials([eurofer, Helium], [0.65, 0.35], 'ao', name="mat2")
+
+beryllium = openmc.Material(name="beryllium")
+beryllium.add_element('Be', 1.0)
+beryllium.set_density("g/cm3", 1.85)
+
+Li4SiO4 = openmc.Material(name="Li4SiO4")
+Li4SiO4.add_element('Li', 4.0)
+Li4SiO4.add_element('Si', 1.0)
+Li4SiO4.add_element('O', 4.0)
+Li4SiO4.set_density("g/cm3", 2.39)
+
+mat3 = openmc.Material.mix_materials([eurofer, beryllium, Li4SiO4, Helium], [0.1, 0.37, 0.15, 0.38], 'ao',name="mat3")
+
+mats = openmc.Materials([mat1, eurofer, Helium,  mat2, beryllium, Li4SiO4,  mat3])
+mats.export_to_xml()
+
+pz = openmc.Plot()
+pz.basis = 'yz'
+pz.origin = (0.0, 0.0, 0.0)
+pz.width = (200.0, 200.0)
+pz.pixels = (500, 500)
+pz.color_by = 'material'
+
+px = openmc.Plot()
+px.basis = 'xy'
+px.origin = (0.0, 0.0, 0.0)
+px.width = (200, 200)
+px.pixels = (500, 500)
+px.color_by = 'material'
+
+plots = openmc.Plots([pz,px])
+plots.export_to_xml()
+
+settings = openmc.Settings()
+settings.dagmc = True
+settings.batches = 100
+settings.particles = 10000000
+settings.run_mode = "fixed source"
+
+settings.temperature = {'default': 800.0,
+                        'method': 'interpolation',
+                        'range': (294.0, 3000.0),
+                        'tolerance': 1000.0}
+
+source = openmc.Source()
+
+r = openmc.stats.PowerLaw(55, 65, 1.0)
+phi = openmc.stats.Uniform(0.0, 2*math.pi)
+z = openmc.stats.Discrete([0,], [1.0,])
+spatial_dist = openmc.stats.CylindricalIndependent(r, phi, z)
+
+source.angle = openmc.stats.Isotropic()
+source.energy = openmc.stats.Discrete([14.08e6], [1.0])
+source.space=spatial_dist
+settings.source = source
+settings.export_to_xml()
+
+dagmc_univ = openmc.DAGMCUniverse(filename='tmesh_1.h5m')
+
+geometry = openmc.Geometry(root=dagmc_univ)
+geometry.export_to_xml()

test/tests/cad_geometry_example/openmc.i

@@ -0,0 +1,83 @@
+[Mesh]
+  [file]
+    type = FileMeshGenerator
+    file = tmesh_1.e
+  []
+[]
+
+[AuxVariables]
+  [cell_temperature]
+    family = MONOMIAL
+    order = CONSTANT
+  []
+[]
+
+[AuxKernels]
+  [cell_temperature]
+    type = CellTemperatureAux
+    variable = cell_temperature
+  []
+[]
+
+[Problem]
+  type = OpenMCCellAverageProblem
+  tally_type = cell
+  tally_name = 'heat_source H3'
+  lowest_cell_level = 0
+  temperature_blocks = '1 2 3'
+  check_tally_sum = false
+  source_strength = 1e18 # Particles/s.
+  volume_calculation = vol
+  tally_score = 'heating_local H3_production'
+  tally_trigger = 'rel_err none'
+  tally_trigger_threshold = '0.1 0.1'
+  verbose = true
+  max_batches = 10
+  batch_interval = 5
+  particles = 5000
+  output = unrelaxed_tally_std_dev
+  skinner = moab
+[]
+
+[UserObjects]
+  [vol]
+    type = OpenMCVolumeCalculation
+    n_samples = 5000
+  []
+  [moab]
+    type = MoabSkinner
+    temperature_min = 800
+    temperature_max = 1100
+    n_temperature_bins = 10
+    temperature = temp
+    build_graveyard = true
+  []
+[]
+
+[Postprocessors]
+  [heat_source]
+    type = ElementIntegralVariablePostprocessor
+    variable = heat_source
+  []
+  [tritium_production]
+    type = ElementIntegralVariablePostprocessor
+    variable = H3
+  []
+  [tritium_RelativeError]
+    type = TallyRelativeError
+    tally_score = h3_production
+  []
+  [heat_source_RelativeError]
+    type = TallyRelativeError
+    tally_score = heating_local
+  []
+[]
+
+[Executioner]
+  type = Transient
+[]
+
+[Outputs]
+  exodus = true
+  csv = true
+[]