INVESTIGATION OF TWO-PHASE FLOW THERMAL-HYDRAULIC BEHAVIOR IN ROD BUNDLE DURING REFLOOD TRANSIENTS BASED ON THE RBHT EXPERIMENTAL DATA
Open Access
- Author:
- Jin, Yue
- Graduate Program:
- Nuclear Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 22, 2019
- Committee Members:
- Fan-Bill B Cheung, Dissertation Advisor/Co-Advisor
Fan-Bill B Cheung, Committee Chair/Co-Chair
Asok Ray, Committee Member
Stephen P Lynch, Committee Member
Cengiz Camci, Outside Member
Kostadin Ivanov, Special Member
Maria Avramova, Special Member
Bao-Wen Yang, Special Member
Stephen Bajorek, Special Member - Keywords:
- Two-Phase Flow
Heat Transfer
Reflood
Liquid Droplets
Mass Quality
DFFB - Abstract:
- Reflood transients within a rod bundle geometry involves complicated thermal-hydraulic phenomena. This scenario could occur during emergency coolant injection for a light water reactor (LWR) under accident conditions. The general features related to the reflood stage, such as the flow and heat transfer regimes, quench front propagation, thermal-hydraulic non-equilibrium, are well known to the mechanical and nuclear engineering communities. However, detailed measurements and/or quantitative predictions of the local two-phase flow structures, interfacial mass and heat transfer processes, spacer grid effects on two-phase flow dynamics, especially for the Dispersed Flow Film Boiling (DFFB) regime, still present a major challenge today. This piece of missing information, once acquired, could contribute significantly to a better understanding of the two-phase flow process as well as the overall rod bundle response to postulated accidents such as the Loss-Of-Coolant-Accident (LOCA). In order to have a better understanding of the reflood transients and to provide reliable experimental data for the two-phase flow mass and heat transfer model development, a series of constant reflood tests were performed at the NRC/PSU RBHT test facility. The tests cover a wide range of conditions that could be encountered during the accident scenario for a LWR, accounting for various parameters that include the system pressure, inlet liquid flooding rate, inlet liquid subcooling, initial peak cladding temperature, and the rod bundle power input. In the current experiment, transient variations of key parameters related to the two-phase flow thermal-hydraulics, such as the mass flow rate, fluid temperature, droplet behavior, as well as the rod bundle response, including the cladding temperature, and quench front propagation along the bundle, are measured using various measuring techniques. Based on these unique experimental results obtained, further data reductions and analyses have been performed to specifically study the two-phase flow characteristics (parametric effects, thermal-hydraulic non-equilibrium, mass quality, liquid droplet dynamics, etc.) and the spacer grid effects (droplet breakup on spacer grid, form loss, etc.) in the DFFB regime. In addition, based on the experimental data, theoretical models (thermal-hydrualic non-equilibrium, two-phase flow mass quality, liquid droplet breakup, and spacer grid pressure drop) were developed describing the two-phase flow interactions as well as the rod bundle heat transfer processes. Another part of this work includes a numerical analysis of the reflood transients using a nuclear reactor thermal-hydraulic sub-channel analysis code, i.e., COBRA-TF. The latter incorporates a two-fluid three-field solution scheme for the two-phase flow by solving a set of nine time-averaged conservation equations. Numerical simulations were performed for the reflood transients covering a wide range of test conditions. The results were compared with the RBHT reflood data set to evaluate the prediction capability of COBRA-TF on reflood transients, especially on the mass and heat transfer process in the DFFB regime. The overall prediction uncertainties of the code for important two-phase flow and heat transfer quantities were determined. Based on the comparison and evaluation to the existing models, further improvements as well as developments of new analysis models could be proposed for the two-phase flow and rod bundle heat transfer analyses. The major outcomes of the current work are: 1). The system parametric effects (including the system pressure, inlet liquid injection rate, inlet liquid subcooling and the rod bundle power input) on the quench front propagation, collapsed liquid level variation, rod bundle overall pressure drop, rod bundle cladding temperature and heat flux, vapor temperature, spacer grid temperature, spacer grid pressure drop as well as on the liquid droplet size and velocity have been investigated in detail. 2). Various theoretical and empirical models have been developed based on the RBHT data, these include the two-phase flow thermal-hydraulic non-equilibrium model, two-phase flow local mass quality model for the DFFB regime, dry spacer grid induced liquid droplet breakup model, and the spacer grid pressure drop model in the DFFB regime. 3). Comprehensive and inclusive COBRA-TF numerical simulations were performed and compared with corresponding RBHT reflood tests to verify the code prediction capability and to validate various two-phase flow mass and heat transfer models incorporated in COBRA-TF. In the experimental part, very comprehensive and detailed two-phase flow measurement has been achieved in the RBHT test facility. It was found that the system pressure, flooding rate and rod bundle power input strongly affect the thermal-hydraulic behavior of the two-phase flow and rod bundle heat transfer. Whereas the inlet liquid subcooling was found to have secondary effect on the mass and heat transport, especially in the DFFB regime. Meanwhile, unique liquid droplet data has been obtained during reflood based on advanced laser imaging system. The entrained liquid droplets were found to have a log-normal distribution in size. As a liquid droplet size becomes larger, its corresponding velocity decreases. While it is relatively difficult to distinguish the most dominating system parametric effects on the liquid droplet size variation other than the quench front location relative to the point of measurement, the droplet velocity was found to be quite different for different flooding rate and system pressure tests. During reflood transients, the two phases were always found to be in significant thermal-hydraulic non-equilibrium but no detailed work were found to quantitatively study this phenonmenon due to the limitation of experimental data. In the current study, the non-equilibrium extent, as characterized by the vapor phase superheat and the two-phase slip ratio, were investigated quantitatively. It was observed that the two-phase slip ratio is not only a function of the distance from the quench front location where most droplets get entrained but also a function of the liquid droplet size. Moreover, if there is thermal non-equilibrium within the two-phase flow, then the hydraulic non-equilibrium is likely to exist. In the theoretical part, the several models developed (two-phase flow mass quality, liquid drople breakup, and spacer grid pressure drop) in the current study were able to predict the two-phase flow mixture thermal-hydraulic behavior with significantly improved accuracy. This is mainly because the development of these models started from the fundamental principles and mechanisms involved in the two-phase flow mass and heat transfer processes and from the adequate consideration of various factors that are important to the current problem. The proposed two-phase flow mass quality correlation was able to predict the data from the current RBHT test and previous tests well within ±10% error. Model developed for the dry spacer grid liquid droplet breakup had an error within ±17%, whereas the spacer grid pressure drop model developed for the DFFB regime had an error span of ±25% (Pa). In the numerical part, the COBRA-TF simulations and their detailed comparisons with the RBHT experiments provide abundant and very informative information for the code performance, based on which a verification and validation study was carried out in order to quantify the prediction uncertainties involved in reflood transients. It has been found that over the reflood conditions explored in the current study, the COBRA-TF was able to predict the rod cladding temperature within ±15% (K) error, the spacer grid and vapor phase temperatures within ±20% (K) error, and the droplet velocity within ±30% (m/s) error span. However, it was also found that COBRA-TF tended to predict early quench of the entire bundle, while under-predicting the DFFB regime void fraction variation. Most specifically, large uncertainties existed for two-phase flow pressure drop calculation. The various results and major findings obtained in this work are significant and informative in the sense that new two-phase flow phenomena and behaviors were either discovered for the first time or investigated extensively due to abundant RBHT reflood data that made this possible. The valuable and unique data obtained from the NRC/PSU RBHT test facility, either filled the void or greatly increased the current experimental data base for reflood transients (especially for the DFFB regime with liquid droplets present), and may serve as benchmark tests that will significantly improve the performance of the numerical simulation tools. Also, the theoretical works performed in the current study provided a detailed characterization of the two-phase flow behavior by relating the basic and local flow and heat transfer quantities (such as the interfacial heat transfer, vapor superheat, slip ratio, droplet size and fluid properties, etc.) to variables that are of importance in the two-phase flow thermal-hydraulic analysis (such as the void fraction, mass quality, ratio of in- and out-coming liquid droplet size at spacer grid location, spacer grid pressure drop, etc.), thus broadening our way of understanding the two-phase flow and providing more choices when tackling the problem at hand. Last, the extensive code verification and validation performed for the COBRA-TF simulation provided the basis and direction for further model improvement and development.