detailed cad model and simulations of the are (aircraft reactor experiment), operated by oak ridge national laboratory 1954
docs/are.pdf lists reference of the are core design, documented in the ornl reports and located in the repository github.com/openmsr/msr-archive
work-in-progress cad model of the are can be found here on onshape.
note that this work and the cad model is under the GNU General Public License v3.0
CAD_to_openMC is an open-source package to convert CAD geometry (in the form of '.step' files) into an openmc-readable h5m file
These simulations use openmc. automated source installation scripts for linux can be found here
We have set up a set of helper scripts to run a set of standard simulations with the ARE-geometry: k_eff, neutron_flux, ...
All these runs first check if the necessary set of three .h5m-files exist. If they do, they ae used, of not a set will be created on the fly using the helper script are_step_to_h5m.py. Thus, should you wish to regenerate the set (or part of it) you can simply move the files out of the way or run are_step_to_h5m.py manually.
estimates the reactivity or criticality of the reactor geometry.
This run defines an overall tally in the reactor geometry that measures the energy of neutron within it, as well as a set of localized energy-scoring tallies only scoring particles within:
- fuel-salt within the reactor core
- the reactor vessel wall.
- the coolant salt.
Please note that even though the ARE reactor geometry describes a fluid fuel-reactor, this simulation is a static steady-state run, meaning we presently neglect depletion/burn-up and things like delayed neutrons. It is possible to account for such effects, but is considere out-of-scope for this set of sims.
This simulation run is very similar to k_eff, ecept that it adds a set of mesh-tallies on slices through the reactor core. Each plane is set to score the neutron-flux in the regions as a function of spatial coordinate, or more precisely flux binnned in cuboid volume elements in the planes. The tallies predefined in the computations are a set of XY|z\in{-50,-25,0,25,50} slices as well as slices at XZ|y=0, and YZ|z=0. Tallies, and mesh-tallies in particular, create some extra overhead slowing the simulations down somewhat - this is normal.
This type of run is not a simulation run as such, instead it creates a set of geometry slice plots aælong the principal axis. This is useful to make sure that you are in fact simulating the geometry you think you are. Lately, a new tool has been included the openmc-distribution: openmc-plotter, which largely renders doing this type of run manually moot. The plotter does perform a geometry plot-run in the background.
All this is also put together in a jupyter notebook that may be found in the subdirectory jupyter.