Simulation of post-loca heat transfer in nuclear fuel bundle
Hussein, Mohamed Hisham Ibrahim (2024)
Diplomityö
2024
School of Energy Systems, Energiatekniikka
Kaikki oikeudet pidätetään.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2024111492361
https://urn.fi/URN:NBN:fi-fe2024111492361
Tiivistelmä
Thermal hydraulic analysis is essential for ensuring safety in the nuclear industry, requiring reliable codes to predict reactor core behavior under both normal and accident conditions. This thesis examines the sub-channel code Coolant-boiling in Rod Arrays- Two Fluids (CTF), to simulate the top reflood of a nuclear fuel bundle using data from the proprietary Westinghouse G-2 experiment. Three tests from the G-2 experiment, conducted under countercurrent upper head injection conditions, were selected. CTF, originally developed for normal operation of light water reactors (LWRs), had not been validated for top-down flow in Loss of Coolant Accident (LOCA) scenarios, which involve significantly different conditions.
A quarter symmetry model of the G-2 experimental test section was created in CTF, featuring a 30-node axial mesh and a 100-channel radial mesh. The heated rods were modeled with three layers of different materials to replicate the experimental setup, and appropriate power distribution and boundary conditions were applied. To simulate reverse flow, gravity was reversed by overwriting the parameter in the CTF source code.
The comparison of the simulation with the experiment showed that CTF prediction agrees with the measured cladding temperatures near the center of the bundle but underpredicts temperatures at top and bottom ends. The experimental data indicated that rod cooling during upper head injection is a gradual process due to the formation and slow movement of a wet front along the rod length. The code appeared to overestimate the wet front propagation velocity, suggesting possible inaccuracies in the heat transfer coefficient calculations.
The study concludes that CTF, in its current state, cannot reliably simulate low-pressure Upper Head Injection. Further work is required to identify and correct the models causing inaccurate results and to improve numerical stability for top-down flow simulations. Addressing these issues will enhance CTF's capability, making it a more powerful and valuable tool for the nuclear industry.
A quarter symmetry model of the G-2 experimental test section was created in CTF, featuring a 30-node axial mesh and a 100-channel radial mesh. The heated rods were modeled with three layers of different materials to replicate the experimental setup, and appropriate power distribution and boundary conditions were applied. To simulate reverse flow, gravity was reversed by overwriting the parameter in the CTF source code.
The comparison of the simulation with the experiment showed that CTF prediction agrees with the measured cladding temperatures near the center of the bundle but underpredicts temperatures at top and bottom ends. The experimental data indicated that rod cooling during upper head injection is a gradual process due to the formation and slow movement of a wet front along the rod length. The code appeared to overestimate the wet front propagation velocity, suggesting possible inaccuracies in the heat transfer coefficient calculations.
The study concludes that CTF, in its current state, cannot reliably simulate low-pressure Upper Head Injection. Further work is required to identify and correct the models causing inaccurate results and to improve numerical stability for top-down flow simulations. Addressing these issues will enhance CTF's capability, making it a more powerful and valuable tool for the nuclear industry.