HIGH FIDELITY SIMULATION OF BUOYANCY-DRIVEN FLUID FLOW IN FLOURIDE SALT-COOLED HIGH TEMPERATURE REACTORS USING THE SPECTRAL ELEMENT METHOD
Open Access
- Author:
- Nguyen, Tri
- Graduate Program:
- Nuclear Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 08, 2023
- Committee Members:
- Arthur Motta, Professor in Charge/Director of Graduate Studies
Xing Wang, Major Field Member
Fan-Bill Cheung, Outside Unit & Field Member
Elia Merzari, Chair & Dissertation Advisor
Haomin Yuan, Special Member
Saya Lee, Major Field Member - Keywords:
- DNS
LES
Heat Transfer
Heat Transfer correlation
FHR
FLiBe
NekRS
High-Pr
Nek5000 - Abstract:
- The world transition from fossil to clean energy is challenging given that fossil fuels still contribute the most of world energy demand. While renewable energy cannot displace fossil fuels from the electricity mix, nuclear energy is the leading solution for the world energy transition mission. Among advanced nuclear reactor designs, Fluoride salt-cooled high-temperature reactors (FHRs) are considered to have the greatest inherent safety thanks to Molten salts' special thermophysical and neutronic properties. This research investigates buoyancy-driven flow in particular aspects that play a crucial role in the passive safety of advanced FHRs. These aspects include the natural circulation system (NCS), the downcomer, and the reactor core. The GPU-based high-fidelity high-order spectral element code NekRS is used to perform all simulations. First, the simplest possible NCS – thermosiphons have been investigated by a series of Large Eddy Simulation (LES) for a wide range of Rayleigh and Lt/D ratios, where Lt is the total length of the loop, and D is the pipe diameters. The simulation results demonstrate a capability to design thermosiphon loops with the prospect of Kairos Power FHR (KP-FHR) design. Then, Direct Numerical Simulation (DNS) of mixed convection in the downcomer for a wide range of Reynolds, Prandtl, and Richardson numbers was performed following the inputs from Kairos Power. An unprecedented high-fidelity numerical DNS database has been established, and new correlations for mixed convection in the downcomer of FHR with molten salt FLiBe are proposed. The newly proposed correlations could predict the trend of the Nusselt number dataset in the heat transfer deterioration region within +/- 10% of the DNS data confidence interval, which is a considerable improvement compared to all other available correlations in the literature. The generated downcomer DNS database will also help better understand Reynolds-averaged Navier–Stokes (RANS) model performance and physical explanation for turbulence model shortcomings. Finally, we perform a full-core high-fidelity LES with the consideration of the FLiBe variables properties for the test reactor "Hermes" of KP-FHR design. The heat transfer data has been obtained and compared with existing correlations in the literature. A good agreement has been achieved between NekRS results and correlations, providing confidence in the generated data. The Nusselt number calculation results could highlight the potential issues of using the existing correlations and propose the correlation implementation depending on the operational regimes of FHRs. Overall, the high-fidelity data and new correlations have implications for the design and optimization of FHRs, providing valuable insights into flow stability, turbulence characteristics, and heat transfer performance.