The core of a Light Water Reactor (LWR) involves many different physics problems, such as neutron transport, fluid dynamics and heat transfer. Inherently, all fields are coupled and to determine the state of the reactor all perspectives need to be considered concurrently.
In many of the currently applied methodologies, the coupled problem is divided in its constituent parts, and consequently solved in a segregated manner. Often the coupling is approximated by static or simplified expressions, whereas in other cases an a posteriori coupling is achieved by combining different tools. Such splitting approaches introduce approximations in terms of the interdependent parameters, and also represent an obstacle for highly resolved coupled calculations.
The reactor core is also a multiscale environment, with important phenomena ranging from the size of the reactor tank itself to scales relevant for the fuel pellet, and further smaller approaching particle scales. Since fully resolved simulations on length scales smaller than the fuel pellet are still considered to be extremely heavy computations, the reactor core problem is commonly solved on larger scales. Unavoidably, such a coarsening introduces homogenization, not only in terms of geometrical details but also in the models used to represent the underlying physics.
In contrast, a fine-mesh tool able to resolve the finer scales would also allow resolved coupled calculations to be performed. Such a coupled tool could be used to assess the approximations in the coarser methods, as well as determining the fine scale, local behavior of the physics in the fuel bundles. Development of fine-mesh tools can give an important contribution to safety since they have the potential for reproducing the physical phenomena of a nuclear system with a higher degree of fidelity.
This project is aimed at developing models and implementing a high-resolution coupled tool for fine scale simulations of the reactor core. This includes formulating a fully consistent model, directly coupling the modeling of neutron transport in the core, of fluid dynamics in the moderator and of conjugate heat transfer between the moderator and the fuel pins.
The developed tool is aimed at better capturing the phenomena, both by resolving the physics and introducing a direct coupling on the fine levels. The primary target are LWRs, i.e. Pressurized Water Reactors (PWRs) and Boiling Water Reactors (BWRs), with BWRs representing the most interesting case due to the large fluid heterogeneities induced by the two-phase flow of concurrently appearing steam and liquid water.
A successful implementation of such a tool relies on the use of high performance computing (HPC), including efficient methods as well as the use of fully parallelized algorithms and solvers. Consequently, the computational aspects are also a major focus in the current project.
The Swedish Centre for Nuclear Technology (Svenskt kärntekniskt centrum - SKC)