NDManager is a python package that helps you manage your nuclear data files in a clean manner.
It provides an API and a CLI for creating and interacting with a database of ENDF6 files, as well as processed data files and depletion chain files in the OpenMC format.
In addition, it encapsulate the sandy nuclear data sampling tool to help you easily generate perturbed files in the OpenMC HDF5 format.
Full documentation is available here.
NDManager is available on PyPI
and can therefore be installed with pip:
$ pip install ndmanagerThere are however a few caveats regarding the dependencies of NDManager. To build processed nuclear data files in the OpenMC HDF5 format, an installation of the NJOY processing code as well as the OpenMC python API is required. These can't be installed with pip so you will need to install them separately.
NDManager also uses sandy to generate perturbed data files, currently
the prefered version is v1.1 and is not available on PyPI, you can
install it manually with to following command:
$ pip install git+https://github.com/luca-fiorito-11/[email protected]NDManager allows you to manage databases for different kinds of files:
- ENDF6 evaluation nuclear data files
- HDF5 processed nuclear data files
- XML depletion chain files
- HDF5 sampled nuclear data files
The directories in which these databases will be stored can be specified
using the following environment variables: NDMANAGER_ENDF6,
NDMANAGER_HDF5, NDMANAGER_CHAINS, NDMANAGER_SAMPLES.
You can also define these paths in a YAML file located at
$HOME/.config/ndmanager/settings.yml:
NDMANAGER_ENDF6: /path/to/endf6
NDMANAGER_HDF5: /path/to/hdf5
NDMANAGER_CHAINS: /path/to/chains
NDMANAGER_SAMPLES: /path/to/samplesIf no path is set, the data will be stored in $HOME/.config/ndmanager
The following sections give a quick overview of the main features of NDManager's modules.
The ndf module acts as a kind of package manager for evaluated nuclear
data files. It uses the IAEA
ENDF-Archive as an
online source of ENDF6 files.
Most of the libraries in this repo have a common directory structure
making them easy to parse. You can list the available libraries using
the ndf list command:
$ ndf list
---------------------------------------------------- Available libraries ----------------------------------------------------
brond22 BROND-2-2 [ ]: BROND-2 USSR evaluated neutron data library, issued in 1992
brond31 BROND-3.1 [✓]: BROND-3.1 Russian evaluated neutron data library, issued in 2016
cendl31 CENDL-3.1 [ ]: CENDL-3.1 Chinese evaluated neutron data library, issued in 2009
cendl32 CENDL-3.2 [ ]: CENDL-3.2 Chinese evaluated neutron data library, issued in 2020
endfb70 ENDF-B-VII.0 [ ]: ENDF/B-VII.0 U.S. Evaluated Nuclear Data Library, issued in 2006
endfb71 ENDF-B-VII.1 [ ]: ENDF/B-VII.1 U.S. Evaluated Nuclear Data Library, issued in 2011
endfb8 ENDF-B-VIII.0 [✓]: ENDF/B-VIII.0 U.S. Evaluated Nuclear Data Library, issued in 2018
endfb81 ENDF-B-VIII.1 [ ]: ENDF/B-VIII.1 U.S. Evaluated Nuclear Data Library, issued in 2024
fendl32b FENDL-3.2b [ ]: FENDL-3.2b Fusion Evaluated Nuclear Data Library, 2022
jeff31 JEFF-3.1 [ ]: JEFF-3.1 Evaluated nuclear data library of the OECD Nuclear Energy Agency
jeff311 JEFF-3.1.1 [✓]: JEFF-3.1 Evaluated nuclear data library of the OECD Nuclear Energy Agency
jeff312 JEFF-3.1.2 [ ]: JEFF-3.1.2 Evaluated nuclear data library of the OECD Nuclear Energy Agency
jeff33 JEFF-3.3 [✓]: JEFF-3.3 Evaluated nuclear data library of the OECD Nuclear Energy Agency, 2017
jendl32 JENDL-3.2 [ ]: JENDL-3.2 Japanese evaluated nuclear data library, 1994
jendl4 JENDL-4.0 [ ]: JENDL-4.0 Japanese evaluated nuclear data library, 2010
jendl5 JENDL-5-Aug2023 [ ]: JENDL-5 Japanese evaluated nuclear data library, 2021
tendl2021 TENDL-2021 [ ]: TENDL-2021 TALYS-based Evaluated Nuclear Data Library, 2021
tendl2023 TENDL-2023 [✓]: TENDL-2023 TALYS-based Evaluated Nuclear Data Library, 2023
-------------------------------------------------------------------------------------------------------------------------------The first time you run this command, the entire IAEA database will be parsed, this can take about a minute.
Subsequent runs of the ndf list command will reuse the cached information so it should go much faster.
By default the list command shows a subset of what is available on the website, and shows you a convenient short name in the first column.
To get the entire list of available libraries you can add the --all tag to the command.
You can then install any library you want with the ndf install command, providing a -j tag followed by an integer allows you parallelize the process:
$ ndf install jeff33 -j 10
JEFF-3.3/n : 100%|████████████████████████████████████████| 562/562 [00:12s]
JEFF-3.3/decay : 100%|████████████████████████████████████████| 3852/3852 [00:23s]
JEFF-3.3/nfpy : 100%|████████████████████████████████████████| 19/19 [00:00s]
JEFF-3.3/sfpy : 100%|████████████████████████████████████████| 3/3 [00:00s]
JEFF-3.3/tsl : 100%|████████████████████████████████████████| 20/20 [00:03s]The ndf remove command will help you uninstall libraries you don't need anymore.
You can use NDManager's python API to get the path to an ENDF6 tape installed using ndf:
import ndmanager
tape = ndmanager.get_endf6("jeff33", "n", "Pu239")The second argument refers to the sublibraries. The ones you will probably care the most about are:
nfor incident neutron data, NSUB=10photofor photoatomic interaction, NSUB=3ardfor atomic relaxation data, NSUB=6tslfor thermal scattering laws, NSUB=12
The Omcer module is your tool to manage processed nuclear data files in the OpenMC HDF5 format.
# jeff33.yml
summary: A library based on the JEFF-3.3 evaluations.
description: |
This defines a library based on the JEFF-3.3 evaluations for
neutron cross-sections and thermal scattering laws, and on
ENDF-B/VIII.0 for gamma photoatomic reactions and atomic
relaxation.
Some evaluations from other libraries are used instead
or in addition to JEFF-3.3 evaluations:
C0: JEFF-3.3 contains C0 and C13 isotopes of carbon. Since
using C0 can cause some issues running the same problem
with multiple libraries, I prefer to remove them and
use ENDF-B/VIII.0 carbon instead. Normally you would also
build TSL libraries using C0, I substitute it for C12.
C12: See C0
C13: See C0.
Ta180: OpenMC wrongly defines natural Ta as containing Ta180
instead of Ta180M because ENDF-B/VIII.0 has no Ta180M
evaluation. Since JEFF-3.3 has no Ta180, materials
containing Ta will crash simulations.
O17: The O17 evaluation of JEFF-3.3 is faulty. I use the ENDF-B/VIII.0
evaluation instead.
name: jeff33 # The name of the future HDF5 library
neutron:
base: jeff33 # The ENDF6 library to use the tapes from by default
temperatures: 250 294 600 900 1200 2500
ommit: C0 # The C0 tape from the base library will be ignored
add: # If added nuclides already exist in the base library, they will be replaced
endfb8: Ta180 C12 C13 O17
photon: # The base library need at least photoatomic tapes, ard tapes are optionnal
base: endfb8
tsl:
base: jeff33
add:
jeff33:
tsl_0031_Graphite.endf6: C12 # Here I specify that the graphite TSL laws must be
# built using the C12 incident neutron file, specified
# in the "neutron" section as coming from endf8Running the ndo build command on this file will build your library, once again the -j tag allows you
to parallelize the process:
$ ndo build jeff33.yml -j 40This will build the HDF5 file for all nuclides and create a cross_sections.xml file that you can provide to OpenMC to use the library.
Sometimes you may want to substitute a single nuclide from a library, for instance to check the impact of a new evaluation of your favorite nuclide. My favorite nuclide is Cl35, and it turns out that the cross-sections of Cl35 in JEFF-3.3 and JENDL-5.0 are very different in the fast domain.
To check the impact of substituting the JENDL-5.0 cross-section in my JEFF-3.3 library, I can write the following input file:
summary: The JEFF-3.3 library with the JENDL-5.0 Cl35 data
description: |
This library reuses the cross-sections available in the jeff33 processed nuclear
data library available in the NDManager database, and substitute the Cl35 file with
a new one created from the JENDL-5.0 ENDF6 tape.
name: jeff33-jendl5-Cl35
neutron:
reuse: jeff33 # This needs to be the name of an HDF5 library installed with Omcer
temperatures: 250 294 600 900 1200 2500
add:
jendl5: Cl35
photon:
reuse: jeff33This will run NJOY to create a new processed file only for Cl35.
The cross_sections.xml file will point to the jeff33 processed library for all other nuclides.
To use your new library with OpenMC you can use NDManager's python API:
from ndmanager.API.openmc import set_nuclear_data
set_nuclear_data("jeff33")
# Proceed with openmc stuffIn addition to creating your own libraries, you can download to ones provided by the official OpenMC website:
$ ndo list
-------------------------------------------------- Installable Libraries ---------------------------------------------------
official/endfb71 ENDF-B/VII.1 [ ]: Official OpenMC library based on ENDF-B/VII.1
official/endfb8 ENDF-B/VIII.0 [ ]: Official OpenMC library based on ENDF-B/VIII.0
official/jeff33 JEFF-3.3 [ ]: Official OpenMC library based on JEFF-3.3
lanl/endfb70 ENDF-B/VII.0 [ ]: ENDF-B/VII.0 based library converted from ACE files distributed with MCNP5/6
lanl/endfb71 ENDF-B/VII.1 [ ]: ENDF-B/VII.1 based library converted from ACE files distributed with MCNP5/6
lanl/endfb8 ENDF-B.VIII.0 [ ]: ENDF-B/VIII.0 based library converted from ACE files distributed by Los Alamos National
lab (LANL)
$ ndo install official/endfb8
Downloading official/endfb8: 100%|████████████████████████████████████████| 3.15G/3.15G [02:03s]
Extracting official/endfb8: 100%|███████████████████████████████████████▉| 12.7G/12.7G [02:17s]And then in a python script
from ndmanager.API.openmc import set_nuclear_data
set_nuclear_data("official/endfb8")
# Proceed with openmc stuffThe Chainer module work in a very similar way to the Omcer module.
Here's what a typical ndc input file looks like:
name: jeff33/fast
description: |
A depletion chain based on the JEFF-3.3 evaluations.
branching_ratios: sfr
n:
base: jeff33
ommit: C0
add:
endfb8: Ta180 C12 C13 O17
decay:
base: jeff33
nfpy:
base: jeff33You can run the ndc build command to generate the XML chain file.
To use the generated chains you can use NDManager's python API:
from ndmanager.API.openmc import set_chain
set_chain("jeff33/fast")
# Proceed with openmc stuffTo set both the cross-section files and the chain file at the same time:
from ndmanager.API.openmc import set_nuclear_data
set_nuclear_data("jeff33", "jeff33/fast")
# Proceed with openmc stuffFinally, you can download the official chain from the OpenMC website:
$ ndc list
---------------------------------------- Installable Chains ----------------------------------------
endfb71/thermal [ ]: A chain based on the ENDF-B/VII.1 evaluation with thermal capture branching
ratios
endfb71/fast [ ]: A chain based on the ENDF-B/VII.1 evaluation with fast capture branching ratios
endfb8/thermal [ ]: A chain based on the ENDF-B/VIII.0 evaluation with thermal capture branching
ratios
endfb8/fast [ ]: A chain based on the ENDF-B/VIII.0 evaluation with thermal capture branching
ratios
casl/thermal [ ]: A simplified chain as described by https://doi.org/10.2172/1256820 with thermal
capture branching ratios
casl/fast [ ]: A simplified chain as described by https://doi.org/10.2172/1256820 with fast
capture branching ratios
$ ndc install endfb8/fast
Downloading endfb8/fast : 100%|████████████████████████████████████████| 26.3M/26.3M [00:01s]