Speaker
Description
The Molten Salt Reactor (MSR) is one of the Generation IV reactors, utilizing liquid fuel that flows through the core and other supporting components. Due to its liquid nature, an optimal flow will distribute the reactor power evenly across all pin channels, ensuring efficient heat transfer throughout the system [2]. Therefore, an optimal design is required to ensure flow uniformity throughout the reactor core, from the upper plenum to the lower plenum, supporting both safety and efficiency aspects of the reactor. This study addresses two key research questions: (1) How uniform is the fluid flow distribution, and how can the MSR model design be optimized? and (2) How does fuel variation affect the thermal-hydraulic parameters of the MSR, which remain unknown? These questions are tackled using a Computational Fluid Dynamics (CFD) numerical approach, where licensed Ansys software is used to analyze velocity-based variables, while open-source OpenFOAM software is employed to examine temperature-based variables.
The reactor modeling is conducted using Autodesk Inventor 2024. Each fuel pin has a polygon width of 20 cm, with a central cylinder diameter of 10 cm (small pin) and 16 cm (large pin), and a height of 50 cm. The pins are arranged in seven stacks without gaps to simplify meshing, resulting in a total height of 3.5 meters. For addressing problem one, the solid domain is filled with pure graphite, while the fluid domain contains liquid water. Flow uniformity is evaluated using the Uniformity Index – Area Weighted Average (γₐ) available in Ansys Fluent 2024 R2, where γₐ represents how a field variable varies over a surface, and if γₐ is closer to one, it indicates that the fluid distribution is more uniform throughout the design.
The design optimization was conducted by comparing two inlet and outlet configurations, namely design A (6 inlets and 6 outlets) and design B (12 inlets and 6 outlets. The simulation results that the best uniformity index (γₐ) is achieved in variation-6, at 0.492. The design is then used as the basis for solving problem two using OpenFOAM. The solid domain is configured based on the graphite properties from the IAEA tabulation [4], meanwhile, the fluid domain uses LiF-NaF-KF (mol percentage: 46.5-11.5-52, known as FLiNaK) based on the data tabulation from Sohal et al [5]. The solver used is chtMultiRegionSimpleFoam, a Conjugate Heat Transfer (CHT) steady-state solver that accounts for buoyancy effects, turbulent flow, and compressibility. The simulations result show that models begin converge at iteration 100.000, with the achieved temperature approaching 870 K. This indicates that the turbulence model, whether for one stack, seven stacks, or a full reactor, requires an increasing number of iterations to achieve better convergence. Neverthless, this research requires an in-depth analysis about the impact of different fuel variations, such as 2LiF-BeF₂ (molar percentage 67-33; FLiBe), KCl-MgCl₂ (molar percentage 67-33), and NaNO₃-NaNO₂-KNO₃ (molar percentage 7-40-53).
| Abstract Category | Nuclear Physics |
|---|