Homogenized Fuel Assembly CFD Modelling: A Comparison of Coolant Temperature Fields with a Porous Media Model

IVAN K. UMEZU; DARIO M. GODINO; ANTONELLA L. COSTA; DAMIAN E. RAMAJO; CLAUBIA PEREIRA; CARLOS E. VELASQUEZ

Trieste

Workshop; Joint ICTP-IAEA Advanced School/Workshop on Computational Nuclear Science and Engineering; 2022

International Centre for Theoretical Physics

By using stochastic neutronic simulations codes, such as MCNPX and SERPENT, it is possible to simplify a reactor?s geometrical features by the means of homogenization of its fuel assemblies, which reduces the number of fuel elements by combining them into a single central bigger rod, maintaining the same volume ratios of coolant and fuel material as the original. This approach leads to shorter calculation times when evaluating the system as a whole, given that small local details are omitted. However, when neutronics simulations are carried out decoupled from thermalydraulics, constant averaged temperatures are assumed along the whole reactor core, which might not be completely accurate, especially considering that a power distribution profile is developed throughout the core and differences along the temperature fields are expected.This work focuses on the integration between neutronic and thermal hydraulic calculations, using CFD simulation tools in a hybrid ADS-fission nuclear reactor core. The system evaluated consists of a subcritical lead-cooled fast reactor coupled to a central spallation target as neutron source. The design is a preliminary proposal [3] by the reactor research team in the Department of Nuclear Engineering at Universidade Federal de Minas Gerais (DEN/UFMG) in Brazil. The work employs a CFD methodology suited for thermal-hydraulic calculations in nuclear reactor cores with homogenized fuel assemblies, using the open source code OpenFOAM and a porous-media modeling approach [4]. Each fuel assembly is modeled as a porous region with a central solid fuel rod, equal in size and material thermal properties as used in the neutronic simulation. Furthermore, the core?s axial and radial thermal power distributions were considered as volumetric heat sources and were applied to their respective fuel assemblies. Liquid lead as coolant is modelled with temperature dependent thermal-physical properties, due to the high temperature gradient within the core. Finally, the main goal of this study is to feedback temperature data to the neutronic model and,thus, improve its accuracy. It will be achieved by using the first power distribution approximation (from averaged constant temperatures simulation), as input in the CFD model, which calculates the new temperature field for each homogenized fuel assembly and feeds it back to the neutronic model. As iterations go on, a more accurate temperature field is expected and, thus, a more precise neutronic behavior is also expected.