M. I. Gurevich

The status of MCU-5

The software developed within the framework of the MCU project which implements the Monte Carlo method for solving the transport equations for various kinds of radiation is widely used for calculating nuclear reactors of different types. The final version of MCU-5 is described. The developed libraries of constants are briefly described, and the potentialities of the MCU-5 package modules and the executable codes compiled from them are characterized. Examples of important problems of reactor physics solved with the codes from the MCU family are presented.

Evaluated benchmark experiments at critical assemblies simulating features of an HTHR at the ASTRA facility

The design of the ASTRA facility and critical assemblies that simulate physics features of modular high-temperature reactors (HTHR-Ms) with a graphite moderator and reflectors loaded with fuel particles having multilayer ceramic coatings is described in detail. Geometrical dimensions of the main elements and regions of the critical assemblies, composition of the materials used, and experimental results for various configurations of the critical assemblies are presented. A detailed computational benchmark model allowing for the structural and compositional features of the critical assembly configurations in question is developed on the basis of all the above data. The results are to be used for verification of the neutronics codes used for calculations of high-temperature helium-cooled reactors.

Modeling of critical experiments on the ZR-6 assembly with VVÉR-type uranium-water lattices

ConclusionAn electronic atlas for describing the 2D and 3D mathematical models of critical experiments carried out on the ZR-6 assembly has been developed. A program-generator of the initial data for modeling assemblies using the MCU-RFFI/A program has been produced, by means of which mathematical modeling of 154 configurations can be carried out. The results have been compared with experimental data. For the two-dimensional model the average value of keff over all the configurations amounts to 0.9989±0.0034 (for the KENO-Va program it is 0.9939±0.0048). For the three-dimensional model of 13 configurations an average value of keff of 0.9994±0.0023 has been obtained (for the same assembly in the two-dimensional model the value obtained is 0.99871±0.0030). This enables us to recommend the two-dimensional models investigated for verifying programs of neutron-physical calculations on VVÉR reactors.

Features of the application of the Monte Carlo method to calculations for large RBMK reactors and to model correction on the basis of data from in-core detectors

Statistical errors in sampling neutron fields in physically large systems like an RBMK are analyzed both qualitatively and quantitatively. Recommendations concerning the choice of parameters for calculations are given. A new procedure for Monte Carlo RBMK calculations with model corrections on the basis of data from in-core detectors is proposed. Dedicated software based on the CUDA software and hardware platform is developed for computational research. Results of testing the procedure and software in question via calculations for real RBMK reactors are discussed.

Modeling the heat transfer in geometrically complex media with a volume source

Fuel elements produced from spent fuel are porous media with spatially varying characteristics. A hierarchical discrete structure for the numerical modeling of heat-transfer processes in media with an anisotropic geometry that is characterized by both the microscopic voids and macroscopic changes in the parameters is proposed. The basic unit of the structure at its lower level is a cell that represents the local properties of the medium. The cells have a standard interface that allows one to form three-dimensional networks of such cells. Different types of cells in the network represent macroscopic changes. The potential for parallel processing is analyzed.

Validation of the correctness of the Hald approximation in assessing tolerance

An analysis was performed of the correctness of employing the approximate formula which is widely used when assessing tolerances in the results of regression analysis. The correctness of approximation with the frequently used ratio between the probability and significance level equal to 95/95 is demonstrated. Conditions of application of the formula under stricter requirements, for example, 99/99, are formulated.

Status of PERST-5 package

The methods and algorithms used in the PERST-5 package are described. This package is part of the MCU-5 code and is intended for neutron-physical calculation of the cells and parts of nuclear reactors using a generalized method of first collision probabilities.

Algorithms of NCG geometrical module

The methods and algorithms of the versatile NCG geometrical module used in the MCU code system are described. The NCG geometrical module is based on the Monte Carlo method and intended for solving equations of particle transport. The versatile combinatorial body method, the grid method, and methods of equalized cross sections and grain structures are used for description of the system geometry and calculation of trajectories.

Algorithms for computing the components of the generalized transport tensor taking into account surface currents

Application of the first collision probability (FCP) method to the system with boundary conditions of given currents or white reflection is considered. The equations of the method are generalized for the anisotropic case by introducing several angular modes in each zone.

Algorithms for calculation of the transport tensor components in the generalized first collision probability method

The first collision probability (FCP) method allows generalization by introducing several angular modes in each zone. Algorithms are proposed for calculating the components of the corresponding first collision tensor. These algorithms do not increase the computational complexity beyond that of the matrix calculation algorithm in the traditional FCP method and allow calculating a system described in terms of combinatorial geometry. Fourfold integrals are numerically taken in calculation of a three-dimensional system and twofold integrals are taken for two-dimensional geometry. The geometrical capabilities are ensured by using the standard geometrical module of the Monte Carlo method.