Abstract
The prediction of heat and mass transfer rates is indispensable for two-phase flow industrial processes. One-dimensional system analysis codes commonly use the two-fluid model for system design, performance evaluation, and accident analysis. The interfacial drag force (IDF) term is the dominant contributor to the total interfacial momentum transfer term in the gas and liquid momentum equations. The IDF can be obtained based on two approaches: (1) the Andersen-Chu approach (also known as the IDF based on the drift-flux approach) and (2) the drag coefficient approach. The drag coefficient approach is the preferred method for obtaining the IDF. However, it requires some parameters, which are not well-established, including the interfacial area concentration (IAC) and shape factor (SF).The IAC plays a critical role in mass, momentum, and heat transfers between gas and liquid phases through the interfaces. As the IAC increases, mass and heat transfers between phases increase. Hence, a reliable IAC estimation is important for two-phase flow numerical simulations to conduct engineering designs and safety analyses. Gas bubbles exhibit different flow behaviors depending on their drag characteristics. Based on their drag characteristics, gas bubbles can be classified into two groups that show different characteristics, such as IAC. The drag characteristics of spherical and distorted (group one, G1) bubbles significantly differ from those of cap, slug, and churn-turbulent (group two, G2) bubbles. Two-group modeling provides a general approach for estimating the characteristics of each bubble group, including the IAC, SF, drag coefficient, and relative velocities, across the bubbly-to-beyond-bubbly flow regime.
In Chapter 1, the research background is presented, including the significance of the problem, the research questions, the objective and contributions of this work, and finally, the thesis overview.
In Chapter 2, a two-group drift-flux model was developed to obtain the G1 and G2 velocities in gas-liquid flow in large-diameter pipes. The model is developed using the drift-flux model concept and experimental data. The Group-one and group-two distribution parameters and drift velocities are validated through experimental data. The results show that the proposed two-group drift-flux model supports the concept of drift velocity from bubbly flow to beyond-bubbly flow and is consistent with the one-group and two-group drift-flux models. Moreover, steam-water data are used to validate the applicability of the model in steam-water flow conditions. The proposed two-group drift-flux model advances the predictions of the group-one and group-two void fractions using the superficial gas and liquid velocities in Chapter 3.
In Chapter 3, a two-group area-averaged IAC model for upward vertical dispersed flows through large-diameter pipes is developed. The existing models in the literature are based on flow regime maps which may not be scalable with reliable accuracy. The developed group-one and group-two IAC models are based on the simplified IATE and drift velocity, respectively, which are more reliable to be scaled up. In addition, this study proposes an approach based on the two-group drift-flux model to calculate the two-group void fractions. The existing models in the literature use empirical correlations to obtain group one and group two void fractions. These empirical approaches may not be reliable when the flow conditions vary. In addition, these models are usually complex with several empirical constants. The proposed model for the two-group void fraction is based on the two-group drift-flux model, developed in Chapter 2. The models are evaluated using 156 measured two-group data points in dispersed flow conditions through large-diameter pipes.
In Chapter 4, a two-group (2G) IDF model for dispersed gas-liquid flows in large-diameter pipes is developed using two approaches. The first approach develops a 2G IDF model based on the drag coefficient. The second approach extended Andersen and Chu’s 1G IDF model based on drift-flux parameters to a 2G IDF model. This study developed a G1 drag coefficient model that accounts for the acceleration of G1 bubbles in the wake region of G2 bubbles. Moreover, the shape factor for different bubble shapes was analytically investigated. Results demonstrated that the shape factor significantly decreased with increasing void fraction. This chapter addresses a key missing gap in the development of the drag coefficient approach in the literature, which is the shape factor. This study introduced a 2G shape factor model and highlighted its significance for accurate predictions in dispersed flows.
In Chapter 5, widely used one-dimensional two-fluid model (TFM)-based codes for upward flow in large-diameter pipes are presented. One of their major differences lies in how they treat the IDF in the momentum equations. The IDF can be obtained using Andersen and Chu’s approach, also known as the drift-flux approach (DFA), or the drag coefficient approach (DCA). In the current state of the DCA modeling, the shape factor has been assumed to be unity. This study proposed a shape factor model to make the DCA feasible. In addition, to date, there is no comprehensive comparison between widely used models implemented in the codes. Thus, this study developed a one-dimensional TFM code to evaluate the performance of existing and newly developed models using the same numerical scheme, with 464 data (air-water and steam-water, collected from 5 different sources) up to a void fraction (VF) of 0.696 at low-to-high superficial liquid velocities.
| Date of Award | 13 Jul 2026 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Takashi HIBIKI (Supervisor) |
Cite this
- Standard