OpenMCyclus

Repository of Cyclus archetypes to couple Cyclus with OpenMC

Installation

Dependencies

You will need to have Cyclus, OpenMC, and their required dependencies. It is recommended to install each of these from source.

Directions to install Cyclus and OpenMC¹ :

conda install -y openssh gxx_linux-64=12.2 gcc_linux-64=12.2 cmake make docker-pycreds git xo python-json-logger python=3.10 glibmm glib libxml2 libxmlpp libblas libcblas liblapack pkg-config coincbc boost-cpp sqlite pcre gettext bzip2 xz setuptools pytest pytables pandas jinja2 cython=0.29 websockets pprintpp hdf5=1.12.2 notebook nb_conda_kernels requests entrypoints pyyaml vtk coverage pytest-cov colorama libpng uncertainties lxml scipy

HDF5_DIR=$CONDA_PREFIX \
pip install --upgrade-strategy only-if-needed --no-binary=h5py h5py

git clone https://github.com/abachma2/cyclus.git

cd cyclus

git checkout python-api

python install.py

cd ../
git clone --recurse-submodules https://github.com/openmc-dev/openmc.git
cd openmc/

mkdir build && cd build

cmake ..

make

sudo make install

cd ../

pip install .

If desired, Cycamore can be installed from here.

Install OpenMCyclus

Clone the repository:

git clone https://github.com/arfc/openmcyclus.git 

To install this archetype library run pip install . from the top level of the directory. To run tests: pytest from the main directory.

Running

This archetype assumes that you have a defined reactor model in OpenMC (.xml) files and the required microscopic cross sections (.csv file) for the model. The cross section data must be saved as a .csv file. These files must all be in the same location. Information about these can be found on the OpenMC docpages.

This archetype is then called during a Cyclus simulation by calling the DepleteReactor archetype from the openmcyclus.DepleteReactor library. The input structure is:

<DepleteReactor>
  <fuel_incommods>
    <val>string</val>
    ...
    <val>string</val>
  </fuel_incommods>
  <fuel_prefs>
    <val>double</val>
    ...
    <val>double</val>
  </fuel_prefs>
  <fuel_outcommods>
    <val>string</val>
    ...
    <val>string</val>
  </fuel_outcommods>
  <fuel_inrecipes>
    <val>string</val> 
    ...
    <val>string</val>
  </fuel_inrecipes>
  <fuel_outrecipes>
    <val>string</val> 
    ...
    <val>string</val>
  </fuel_outrecipes>
  <assem_size>double</assem_size>
  <cycle_time>int</cycle_time>
  <refuel_time>int</refuel_time>
  <n_assem_core>int</n_assem_core>
  <n_assem_batch>int</n_assem_batch>
  <power_cap>double</power_cap>
  <model_path>string</model_path>
  <chain_file>string</chain_file>
</DepleteReactor>

fuel_prefs and fuel_inrecipes must be equal in length to fuel_incommods and fuel_outrecipes must be equal in length to fuel_outcommods.

• The model_path variable is the location of the files for OpenMC (can be relative or absolute path): one-group cross sections, materials, and depletion chain file. If using a relative path, it must be relative to the directory you are running the |Cyclus| input file from, not the location of the file that defines the prototype.
• The archetype assumes that the OpenMC materials are in the file called materials.xml and that the cross section data is in a file called micro_xs.csv.
• The chain_file variable is the depletion chain file, and the user provides the name of this file.
• Each material in the materials.xml file that are fuel materials must be marked as depletable and have the name assembly_#. Define one material for each assembly in the reactor core (matches with n_assem_core),
  the number assigned to each material name is irrelevant, just as long as
  there is one.

Outputs

The results of the simulation will be written to cyclus.sqlite or the file name provided when Cyclus was called.

Footnotes

1. Directions on OpenMC install from source taken from: https://docs.openmc.org/en/stable/quickinstall.html, consult this page for the most up to date instructions. ↩