Development and Validation of Broad-group Shielding Library Based on CENDL-3.2

The neutron and photon shielding is an important part of the reactor design for safety of people and facilities. CENDL-3.2 is the latest Chinese evaluated nuclear data library released by China Nuclear Data Center (CNDC) in 2020. Many nuclides were improved with new evaluation techniques and experi...

Full description

Bibliographic Details
Main Author: SHU Wenyu;CAO Liangzhi
Format: Article
Language:English
Published: Editorial Board of Atomic Energy Science and Technology 2022-05-01
Series:Yuanzineng kexue jishu
Subjects:
Online Access:https://www.aest.org.cn/CN/abstract/abstract21413.shtml
Description
Summary:The neutron and photon shielding is an important part of the reactor design for safety of people and facilities. CENDL-3.2 is the latest Chinese evaluated nuclear data library released by China Nuclear Data Center (CNDC) in 2020. Many nuclides were improved with new evaluation techniques and experimental data. Thanks to the high calculation efficiency, broadgroup shielding libraries are widely used in shielding calculation. To carry on the research on reactor shielding calculation, the codes NECPAtlas and NECPShield were applied to develop the broadgroup shielding library NECLCP29 based on CENDL32. Generating a broadgroup shielding library needs three stages. Firstly, NECLCP199, a finegroup library with 199 neutron groups and 42 photon groups, was produced using NECPAtlas and NECPShield. When producing the finegroup library, the neutron weight function consists of a 1/E spectrum, a fission spectrum and a thermal Maxwellian spectrum. The photon weight function consists of a 1/E spectrum at the middle energies. At the high and low energy, the spectrum goes down. Secondly, a typical 1D PWR model used in producing BUGLE libraries was calculated using NECLCP199 to obtain the finegroup flux distribution in the model. Before calculating the model, the nuclides were selfshielded using several different models, including a PWR pin cell model for the core, an ironwater model for the down comer, a concrete model for the concrete, a carbon steel model for the pressure vessel and a stainlesssteel model for the baffle and barrel. Lastly, the finegroup fluxes in core, down comer, pressured vessel and concrete were chosen to collapse finegroup cross sections of selfshielded nuclides to generate a broadgroup library. Moreover, the finegroup flux in concrete was applied to collapse the finegroup infinite diluted cross sections of all nuclides from NECLCP199. The energy group structure has an important influence on the accuracy and efficiency for shielding calculation. To obtain an improved broadgroup structure for better accuracy and efficiency, particle swarm optimization (PSO) was used. The research on application of PSO in optimizing the energy group structure for shielding calculation was performed, including the combination of PSO process and the generation of the broadgroup library, the mapping of PSO variables and the construction of fitness function. A 29group structure was optimized using PSO. NECLCP29 with the optimized 29group structure was generated. To validate NECLCP29 broadgroup shielding library, some benchmarks in SINBAD, the international shielding benchmark library, were calculated, including Iron88, ASPISNG, and HBR2. The calculated results were compared with the measured results. Moreover, the results of BUGLEB7 and BUGLE96 which are famous international broadgroup shielding libraries were also used for comparison. The numerical results show that the calculated values using NECLCP29 agree well with the measured values. Compared with BUGLEB7 and BUGLE96 whose numbers of neutron groups are all 47, the calculation accuracy of NECL-CP29 is better. Especially in calculating the reaction rate of 197Au(n,γ) in Iron88 benchmark, the results given by NECLCP29 are much closer to the measured values than BUGLEB7. Thanks to the optimized energy group structure with less neutron energy group number, the computational efficiency is obviously improved.
ISSN:1000-6931