Date of Award
Doctor of Philosophy (PhD)
Computational Analysis and Modeling
Hydrogen is one of the best fuels because of its high calorific value and environmental friendliness. However, because of its low density, it has storage problems such as high pressure, large volume requirements, heavy weight and safely risks; this quality prevents its wide usage and commercialization.
Researchers have found that some metal/inter-metallic compounds/alloys, such as Mg, La, LaNi5, ZrV2, Mg2Ni and Ti 2Ni, can react with hydrogen and attain relatively large amounts of hydrogen at a relatively low pressure and near normal temperature. Under certain conditions, hydride can desorb hydrogen quickly. Thus, the metal- hydrogen reaction could be a practical means for hydrogen storage, yet some critical issues, including the amount of hydrogen absorbed/desorbed, thermal stability of the hydride, hydrating/dehydrating kinetics, thermodynamical, thermophysical properties, need to be addressed. This dissertation attempts to simulate the hydrogen absorption/desorption processes in cylindrical metal-hydrogen reactors and, hence, answer some of these issues.
We first considered a two-dimensional (2D) mathematical model that governed the heat and mass transfer in a cylindrical reactor due to the symmetry of the reactor. The model included mass conservation, momentum equations, energy conservation, and absorption/desorption rates in porous hydride. As such, a system of mathematical equations was obtained which described the evolution of temperature, density of hydrogen gas, density of hydride, velocity of hydrogen gas, and pressure of hydrogen gas, as well as the absorption/desorption rates. Because of the nonlinearity of the mathematical equation, the system must be solved numerically.
We then developed an accurate and stable finite difference scheme on a staggered mesh using the Crank-Nicholson finite difference method. The model and numerical method were tested for a cylindrical LaNi5-H 2 reactor. Temperature, density, velocity and pressure distribution and profiles in the absorption/desorption processes were obtained. Results showed that during the absorption process, the chemical reaction released heat, and hence, the temperature increased. A cool temperature surrounding the reactor may be needed to prevent overheating. Conversely, the desorption process showed the reverse phenomenon and absorbed heat. A hot temperature is needed to speed up the desorption process.
It should be pointed out that the reaction rate played an important role in the simulation. The reaction rate is usually obtained empirically, which may have affected the accuracy of the numerical prediction. In this study, we further present a numerical method to determine the critical coefficient in the reaction rate by utilizing the difference between measured temperatures and calculated temperatures.
The minimization procedure consisted of a nonlinear system of equations and was solved iteratively by the modified Levenberg-Marquardt method. The procedure was then tested for a cylindrical LaNi5-H2 reactor. Numerical results showed that, with various initial guesses for the coefficient, the estimated coefficient converged to the realistic coefficient, and the estimated values were close to the realistic coefficient when measurements were subject to random error.
Han, Fei, "" (2014). Dissertation. 242.