THE APPLICATION OF A MULTI-PHYSICS TOOLKIT TO SPATIAL REACTOR DYNAMICS

Traditionally coupled field nuclear reactor analysis has been carried out using several loosely coupled solvers, each having been developed independently from the others. In the field of multiphysics, the current generation of object-oriented toolkits provides robust close coupling of multiple fields on a single framework. This paper describes the initial results obtained as part of continuing research in the use of the OpenFOAM multi-physics toolkit for reactor dynamics application development. An unstructured, three-dimensional, time-dependent multi-group diffusion code DiffusionFOAM has been developed using the OpenFOAM multi-physics toolkit as a basis. The code is based on the finite-volume methodology and uses a newly developed blockcoupled sparse matrix solver for the coupled solution of the multi-group diffusion equations. A description of this code is given with particular emphasis on the newly developed block-coupled solver, along with a selection of results obtained thus far. The code has performed well, indicating that the OpenFOAM toolkit is suited to reactor dynamics applications. This work has shown that the neutronics and simplified thermal-hydraulics of a reactor may be represented and solved for using a common calculation platform, and opens up the possibility for research into robust closecoupling of neutron diffusion and thermal-fluid calculations. This work has further opened up the possibility for research in a number of other areas, including research into three-dimensional unstructured meshes for reactor dynamics applications. Nuclear reactor analysis deals with the coupled solution of the many physical processes taking place in a nuclear reactor. The solution of these individual physical processes has traditionally been carried out using several loosely-coupled solvers, each having been developed independently from the others. In particular, the calculation of the distribution of neutrons in space, energy and time is traditionally separated completely from the heat transfer calculation. This separation was introduced in the past for a number of reasons, namely; the solution of each class of problem is typically undertaken by specialists in each field, the complexities of the problems differ, and there are numerical differences between the classes of problems being modeled. This separation leads to problems when coupling the solvers. Often differences in data