Thesis - Open Access
Master of Science (MS)
Department / School
Computaional Fluid Dynamics, Natural Convection, Neutrino
The goal of this research was to simulate the flow, temperature, and impurity concentration within the Deep Underground Neutrino Experiment (DUNE) Single Phase Far Detector using a commercially available computational fluid dynamics (CFD) solver. DUNE is a research collaborative investigating properties of neutrinos in an effort to better understand the origins of matter and behavior of subatomic particles. The Far Detector is a geometrically complex neutrino detector containing: anode plane arrays and cathode plane arrays which induce an electric field within the detection region that causes electrons to drift to the sensing equipment, field cage planes to enclose the neutrino detection region, inlets and outlets for liquid argon flow, ground planes to ground the electric field outside the detection region, a service floor, and other smaller features. High-fidelity models are required to accurately simulate the flow patterns within the detector. This research investigated the effects of: 1) mesh refinement, 2) turbulent Schmidt number, and 3) the boundary condition employed at the liquid-ullage interface, i.e. slip vs. no slip. The effect of mesh refinement was analyzed by comparing the results of six levels of mesh refinement, ranging from 40.8 to 151.6 million cells. The simulation was also completed for turbulent Schmidt numbers of 0.5, 0.9, and 2.0 to determine how this property, which is difficult to quantify, impacted the results. Finally, the results of the simulation were compared for using a slip boundary condition at the liquid-ullage interface to the simulations using a no-slip boundary condition. It was expected that these factors would significantly impact the flow, temperature and impurity concentration within the cryostat and that by comparing the simulation results to experimental data the ideal simulation parameters could be identified and implemented. The computationally generated results of this research are validated using the results of the prototype experimental and simulation data. This thesis research led to three distinct findings. First, appropriate mesh refinement in critical areas, such as near walls or surrounding inlets and outlets led to the outcome that all levels of mesh refinement investigated in this work are able to appropriately capture the movement of liquid argon within the detector. By capturing the complex flow features with local mesh refinements, the impact of global mesh refinement was minimized. Second, the effect of changing the turbulent Schmidt number was negligible, with the impurity concentrations varying less than 0.17% for all turbulent Schmidt numbers in this study. This contradicted the hypothesis that lower turbulent Schmidt numbers would results in greater impurity variance within the detector. This may be due to the extreme purity conditions of the detector but confirms that greater study is necessary to determine to optimum turbulent Schmidt guidelines. Third, the selection of a slip vs. no-slip boundary condition at the ullage-liquid interface yields significantly different flow and consequently different thermal profiles. The no-slip boundary condition leads to the predictions of significantly warmer temperatures in the liquid volume. The slip condition provided results much more consistent with those seen in experiments than the no-slip condition. This is most likely due to the reduced mixing between the warmer gaseous argon and cooler liquid argon caused by forcing the fluid along the interface to remain stationary. The slip condition allows the simulated fluid to move along this interface, which is consistent with conditions in the actual detector.
Library of Congress Subject Headings
Heat -- Convection, Natural.
Computational fluid dynamics.
Number of Pages
South Dakota State University
Christensen, Weston, "Analysis of Natural Convection Flow in a Detector Using Computational Fluid Dynamics" (2019). Electronic Theses and Dissertations. 3667.