Numerical Study of Gas Flow and Diffusion Driven by Buoyancy in a Multi-Openings Compartment
Jinghua Jiang,Peizheng Hu,Lili Tong,Xuewu Cao
DOI: https://doi.org/10.1115/icone31-136244
2024-01-01
Abstract:Abstract During a severe accident in a pressurized water reactor, if the coolant pipe ruptures in a multi-openings compartment, a mixture of hydrogen, vapor, and air migrates and diffuses between compartments, driven by density disparities during the extended period following the cessation of coolant rapid discharge. The same phenomenon occurs during a “wet bypass” in a fusion reactor, a mixture of hydrogen, tritium (HT/HTO), vapor, and air migrates and diffuses between the suppression system, vacuum vessel, and Tokamak gallery. In the scenarios mentioned above, the buoyancy-driven flow resulting from the mixing of light and heavy gases will influence the long-term hydrogen risk and migration of radioactive materials. Previous research has primarily focused on studying the migration behavior of multi-component gas mixtures propelled by jet flow. An experimental device (A4Q-FB) with a multi-openings container was established, to clarify the influence of buoyancy-driven flow on gas migration. The test is carried out with well-controlled boundary conditions, in which, a mixture of helium and vapor with a mass fraction of 2:15 is injected through a vertically downward nozzle into the compartment filled with air. The injection was stopped after 11200 s from the beginning of the experiment, and other boundary conditions were controlled to be unchanged, and the data of the temperature and concentration of the gases inside the container and the flow velocity of the gas mixture in the openings were obtained. In this study, a three-dimensional analysis model for this test was developed, and the predictive accuracy for gas mixture migration and diffusion behavior was assessed. The results indicate a fundamental alignment between the simulation outcomes and the trends observed in the experimental data. This consistency suggests the reliability of the model in the macro-level predictions of phenomena and supports our understanding of the gas mixture migration under buoyancy-driven flow. The simulation results indicate a maximum relative error of 12% in predicting the concentration of vapor at the top and a maximum relative error of 20% in estimating the gas flow velocity within the top openings. The simulation also provides the 3D distribution of velocity and concentration in both container and openings.