HDF5格式多群截面数据库AXELIB的加工与验证

Processing and validation of the HDF5 format multigroup cross section library AXELIB

针对常用多群截面数据库所使用的文本格式或二进制格式存在的缺点,本文设计了HDF5格式分层存储核数据。基于评价核数据加工处理程序NJOY2016,本文开发了核数据处理模块AXER,对NJOY2016程序加工得到的GENDF格式数据处理来产生HDF5格式的多群截面数据库AXELIB。本文同时开发了PyNjoy2016系统,并使用AXELIB进行了各种数值计算,计算结果与OpenMC程序的进行对比,证明了核数据处理方案的可靠性和加工得到的AXELIB的准确性。研究表明:评价数据库的版本会对数值计算结果产生较大影响,本文方法既保留了二进制格式占用内存空间小读写速度快的优点,又能方便用户进行数据查看。

Given the shortcomings of text or binary format used in the currently commonly used multigroup cross section libraries,the HDF5 format is designed to store nuclear data hierarchically in this study.The HDF5 format not only maintains the advantages of fast reading and storage space saving of the binary format but also facilitates users to read and look up data.Based on the NJOY2016 program used to evaluate and process nuclear data,a new nuclear data processing module AXER is developed to process the GENDF format data produced by the NJOY2016 program and generate the HDF5 format multigroup cross section library An Xs Easy-find Library(AXELIB),which ensures the convenience of cross section data consultation.Moreover,the PyNjoy2016 system is developed to generate the entire set of AXELIB containing hundreds of nuclides automatically.The XPZ code uses the newly generated AXELIB to perform various numerical calculations and compares the calculation results with that of OpenMC to verify the reliability of the data processing scheme and the accuracy of AXELIB.Moreover,the version of the evaluation dataset has a considerable influence on the calculation results.