Abstract
This work describes newly developed features of the OpenMC code for nuclear data processing. OpenMC, in addition to being a transport code, includes a rich, extensible Python API that enables programmatic pre- and post-processing. A new openmc.data package in the Python API enables users to parse ENDF and ACE files and convert them to an HDF5 format. With this capability, the OpenMC transport solver now relies on HDF5 nuclear data files rather than ACE files produced from NJOY. Moving to a native format will give much greater flexibility for researchers to explore new methods and algorithms that rely on storing data that is not present in the ACE format. Additionally, the module may serve as an independent implementation of the proposed Generalized Nuclear Data (GND) format in the future.
Highlights
The Monte Carlo method is often applied to particle transport simulations and is considered the most accurate method of solution because of the lack of approximations relative to deterministic methods
NJOY [3] is used to process the ENDF-format data into the ACE format [4], which is used by several Monte Carlo codes including MCNP [5] and Serpent [6]
Much of the design of the class hierarchy has been influenced by the Generalized Nuclear Data (GND) format [9] being proposed by OECD/WPEC Subgroup 38
Summary
The Monte Carlo method is often applied to particle transport simulations and is considered the most accurate method of solution because of the lack of approximations relative to deterministic methods. In order to attain such high accuracy, Monte Carlo methods require a higher level of detail in the nuclear data representation, such as explicit cross sections for all possible reactions, a list of reaction products, and their yields/angle-energy distributions. This data typically originates from nuclear data evaluations as given in the ENDF-6 format [1], for example, the ENDF/B-VII. library [2]. ENDF evaluations cannot be used directly—some processing of the data must be done in order for the Monte Carlo code to be able to sample reaction and product distributions. The present work describes the result of such an effort: the openmc.data package within OpenMC’s Python API
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
