Pattern transition of a gas–liquid flow with zero liquid superficial velocity in a vertical tube
Introduction
The gas–liquid two-phase flow has attracted considerable attention over the past decades owing to its wide application in engineering fields such as the petroleum, chemical, nuclear power, process, and related industries (Carvalho, 2006, Morgado et al., 2016). The flow pattern classification is crucial in the modelling of different flow phenomena, heat and mass transfer rates, pressure drop, and momentum loss as the flow characteristics are considerably different among various flow patterns. Consequently, a large number of experimental and theoretical studies have been carried out on flow pattern transitions (Wu et al., 2017).
Generally, for a two-phase flow in a vertical tube, the flow patterns are classified into bubble, slug, churn, and annular flows (Cheng et al., 1998, Taitel et al., 1980). The bubble-to-slug flow transition plays an important role in the vertical gas–liquid two-phase flow. Griffith and Snyder (1964) carried out experiments in vertical tubes having diameters of 19.1 and 25.4 mm and reported that the bubble flow converted to the slug flow in the void fraction range of 0.25–0.3. Taitel et al. (1980) regarded the bubble as a spherical shape. At the transition boundary, the spacing between adjacent bubbles was sufficiently small so that the coalescence rate sharply increased. They assumed that the critical distance between bubbles was half of the bubble radius and proposed a criterion for the bubble-to-slug flow transition:
According to the theory proposed by Taitel et al., 1980, Hasan and Kabir, 1988 derived an expression with measurable parameters:where um, usg, and ɛg are the mixture velocity, gas superficial velocity, and void fraction, respectively, C0 is the velocity distribution coefficient in which the effects of the non-uniform flow and concentration profiles were considered by Zuber and Findlay (1965), and u∞ is the terminal rising velocity of an individual bubble in a stagnant liquid column expressed by Harmathy (1960):
This criterion is widely used to predict the flow pattern owing to its simple form (An et al., 2000). However, the transition criterion based on void fraction is controvertible as the critical void fraction varies with the experimental conditions. Omebereiyari et al. (2007) carried out experiments in a 189-mm vertical tube with naphtha and nitrogen as working fluids and reported that the critical void fraction distinguishing the bubble flow from other flows is 0.68, which is different from 0.25.
The bubble dynamic approach can also be used to understand the flow pattern transition, which is increasingly employed as it considers the fluid viscosity, surface tension, flow rates, and other experimental conditions (Das and Pattanayak, 1994, Majumder, 2016, Wang et al., 2012). A transition criterion was proposed based on a population balance model, which described the bubble behaviours and size distribution to demonstrate the processes of bubble coalescence and breakage (Das and Pattanayak, 1994, Wang et al., 2005). Another similar method of bubble dynamics models the interfacial area transport process by solving interfacial area transport equations (Wang et al., 2012). Although further studies are required to complete this theory, the bubble dynamics provide constructive insights into the bubble-to-slug flow transition (Zhang et al., 2017).
The slug-to-churn flow transition is another critical transition of the gas–liquid two-phase vertical flow (Montoya et al., 2016). Currently, there are four mechanisms proposed to predict the slug-to-churn transition boundary:
- (1)
Flooding mechanism. The slug-to-churn flow transition was connected with the flooding phenomenon (Jayanti and Hewitt, 1992, Mcquillan and Whalley, 1985, Nicklin, 1962). This theory states that large interface waves are formed in the liquid film surrounding a Taylor bubble. When the gas flow rate increases, the transition is triggered when the falling film is destroyed and gas-lifted by the Taylor bubble.
- (2)
Slug aeration mechanism. Kaichiro and Ishii (1984) proposed a transition criterion based on the high aeration in the liquid slug. They reported that the churn flow occurred when the mean void fraction in the slug unit became larger than that in the Taylor bubble region.
- (3)
Slug breakup mechanism. This mechanism is attributed to the slug-to-churn flow transition to bubble coalescence (Brauner and Barnea, 1986). The gas phase exists in the liquid slug in the form of dispersed bubbles owing to the effect of the turbulent force. When the void fraction in the liquid slug reaches 0.25, bubbles approach each other and coalesce, destroying the liquid slug and leading to the churn flow.
- (4)
Entrance effect mechanism. In this theory, the entrance length is defined as the pipe length that is sufficiently long for the formation of a stable slug. If the pipe length is smaller than the entrance length, a stable slug flow cannot form and thus a churn flow eventually occurs (Taitel et al., 1980).
The predictions of the slug-to-churn flow transition based on the above four models are plotted in Fig. 1. Significant differences exist between these models; even the line trend is inverse (comparison of the Kaichiro's model with the other three models), which indicates the difficulty in the choice of a reasonable model to predict the slug-to-churn flow transition. A new mechanism is required to understand the flow pattern transition.
Zero liquid superficial velocity (ZLSV) flow is a special type of gas–liquid two-phase flow characterised by the gas flowing through a liquid column but no liquid flows out of the pipe. It exists in many engineering applications, particularly in petroleum and natural gas fields (An et al., 2000, Guo et al., 2005), such as liquid loading, relief well killing, and pumping oil well. This type of two-phase flow can also occur in bubble columns. Besagni et al. (2017b) and Guédon et al. (2017) achieved considerable contributions to the understanding of the flow pattern transition in bubble columns. Although the lift force has been widely studied to explain the flow pattern transition from a homogeneous flow to a heterogeneous flow in bubble flow (Besagni et al., 2017a, Lucas et al., 2005, Mena et al., 2005, Ruzicka et al., 2001a, 2008, Ruzicka et al., 2001b, Shah et al., 1982, Wilkinson et al., 1992), further investigations on pattern transitions with higher gas flow rates (e.g., bubble-to-slug and slug-to-churn flows) are required. A pattern transition criterion for the ZLSV flow is needed to further understand the hydrodynamic behaviour of the ZLSV flow.
An experiment on the pattern transition of the ZLSV flow is carried out in a vertical transparent tube having a length of L = 12 m and inner diameter of D = 0.1 m. It should be noted that the patterns and hydrodynamic properties of a two-phase flow in a large-diameter tube differ from those in a small-diameter tube (Besagni and Inzoli, 2016b). The dimensionless diameter, D*, is used to distinguish large- and small-diameter tubes (Isao and Mamoru, 1987):
Tubes with dimensionless diameters larger than the critical value of 52 are considered large-diameter tubes (Brooks et al., 2012), where the Taylor bubble is not sustained owing to the Rayleigh–Taylor instabilities. In this experiment, the dimensionless diameter of the vertical tube is 36.79. Therefore, the slug flow is expected to appear.
The rest of this paper is organised as follows. The mechanism models for flow pattern transitions are proposed in Section 2. The experimental arrangement and procedure are described in Section 3. In Section 4, the experimental results are discussed and the comparison with the mechanism model is presented. Conclusions are presented in Section 5.
Section snippets
Transition mechanism
When the superficial gas velocity is larger than the critical value, the bubble flow disappears and the slug flow occurs in the whole test section (Bouré and Mercadier, 1982, Cheng et al., 1998, Song et al., 1995). In the bubble flow, when two or more bubbles coalesce, a Taylor bubble could form. Moreover, in the ZLSV flow, the turbulence force is very small owing to the small Reynolds number of the liquid flow and thus hardly leads to bubble breakdown, which increases the possibility of bubble
Experimental arrangement and procedure
To investigate the characteristics and pattern transitions of the ZLSV flow, a flow loop was constructed including a 12.5 m × 100 mm (internal diameter) vertical transparent tube as shown in Fig. 5. The test section is made of polyvinyl chloride with a maximum pressure resistance of 10 MPa which enables the observation of flow characteristics. The gas is supplied by an air compressor and then is dried by a scrubber. At a low flow rate, it is pumped into the flow loop through an orifice plate
Bubble-to-slug flow transition
- a.
Analysis of the deviation between the published model and experimental data
Fig. 10 shows that the flow pattern transition criterion based on void fraction is in the poor agreement with experimental data. The void fraction at the transition boundary is in the range of 0.06–0.10, considerably smaller than 0.25.
The transition criterion based on void fraction considers that the bubble spacing at the transition boundary is half of the bubble radius, which corresponds to a void fraction of 0.25 (
Conclusions
An experimental study on pattern transitions of the ZLSV flow was carried out. The bubble-to-slug and slug-to-churn flow transition criteria were proposed. The predicted results are in good agreement with the experimental data.
The ZLSV flow pattern transitions occurs more easily. The bubble-to-slug and slug-to-churn flow transition boundaries predicted by current models were larger than the experimental results. The void fraction at the bubble-to-slug flow boundary was in the range of
Acknowledgements
The authors acknowledge the supports of the National Natural Science Foundation–Outstanding Youth Foundation (51622405), the Shandong Natural Science funds for Distinguished Young Scholar (JQ201716), the National Key Basic Research Program of China (973 Program, 2015CB251200), the National Science and Technology Major Project (2016ZX05028-001-003) and Program for Changjiang Scholars and Innovative Research Team in University (IRT_14R58).
References (62)
A unified model for predicting flow-pattern transitions for the whole range of pipe inclinations
Int. J. Multiph. Flow
(1987)- et al.
Bubble size distributions and shapes in annular gap bubble column
Exp. Therm Fluid Sci.
(2016) - et al.
Comprehensive experimental investigation of counter-current bubble column hydrodynamics: holdup, flow regime transition, bubble size distributions and local flow properties
Chem. Eng. Sci.
(2016) - et al.
The dual effect of viscosity on bubble column hydrodynamics
Chem. Eng. Sci.
(2017) - et al.
Computational Fluid-Dynamic modeling of the pseudo-homogeneous flow regime in large-scale bubble columns
Chem. Eng. Sci.
(2017) - et al.
Slug/Churn transition in upward gas-liquid flow
Chem. Eng. Sci.
(1986) - et al.
Two-group drift-flux model for closure of the modified two-fluid model
Int. J. Heat Fluid Flow
(2012) Experimental study of the slug/churn flow transition in a single Taylor bubble
Chem. Eng. Sci.
(2006)- et al.
A study of the bubble-to-slug transition in vertical gas-liquid flow in columns of different diameter
Int. J. Multiph. Flow
(1998) - et al.
Effects of initial bubble size on flow pattern transition in a 28.9 mm diameter column
Int. J. Multiph. Flow
(2002)