Comparison with existing flow regime maps and available transition criteria
Comparison with existing flow regime maps
The flow pattern maps of the nitrogen-water system are compared with three existing flow regime maps, i.e. that of Triplett et al. (1999), of Yang and Shieh (2001) and of Hassan et al. (2005). These three flow pattern maps are selected because they are all based on the systematic experiments of two-phase horizontal flow in microchannels, which are close to the present work. In experiments of Triplett, air and water were used as working fluids, and the test channel was circular with di = 1.097 mm. As shown in Figure 2-10(A), the regime map of Triplett agrees well with ours for the regions of bubbly, slug and annular flow, with the exception for the transition from slug-annular and annular to churn flow. This difference may be caused by different experimental conditions, such as temperature, or by the differences in gas liquid inlet conditions, channel diameter and fabrication. Good agreement was also found in comparison with Hassan’s flow map (Figure 2-10(C)). However, it can be noted from Figure 2-10(B) that the Flow regime map of Yang and Shieh (2001) does not correspond well with the present regime maps.
Comparison with available transition criteria
Many transition models have been proposed by researchers to predict the flow pattern transitions in small channels. Here, we choose two transition models, that of Akbar et al.(2003) and that of Waelchli and Rudolf von Rohr (2006), for reason that the channel size and fluid properties considered are close to our work.
The comparison between Akbar’s model and the present flow regime maps using Weber number as coordinates are displayed in Figures 2-11(A)-(C). It can be noted that the Akbar’s model can well predict the annular flow region under all experimental conditions, but it is not the case for other flow patterns. Note that the dispersed flow in Akbar’s model, which occurs at higher gas and liquid velocities than churn flow, was not observed in the present work. According to the “zone division” of Akbar et al. (2 003), the transition flow zone consists slug-annular flow, a part of unstable slug flow and a part of churn flow. In addition, the surface tension dominated zone includes bubbly flow, slug and plug flow. As a result, the Akbar’s transition model seems to be acceptable for predicting the transitions of flow patterns.
Also note that Weber number is more appropriate to be used as coordinates than superficial velocities, for that the effects of surface tension and channel diameter can be taken into account. But it fails to reveal the effect of liquid viscosity on two phase flow patterns, as shown in Figure 2-11(B).
Figure 2-12 shows a comparison of Waelchli and Rudolf von Rohr’s model with our experimental results. Several features may be observed. Firstly, the transition lines of the model are in poor agreement with the present experimental data. Secondly, the effect of channel diameter on two phase flow patterns can be well revealed by using ReG , L 0.4 as coordinates, as show in Figure 2-12(A). However, they may not be appropriate to predict the influences of liquid physical properties.
Transition criteria based on the present experimental data
Based on the above discussion, it can be concluded that the effects of channel size and liquid physical properties on the gas-liquid flow patterns in microchannels are significant. The existing transition models can not properly correlate our experimental data. As a result, we try to propose here a new transition model.
Gas-liquid flow in horizontal circular microchannels with diameter of 302 µm, 496 µm, and 916 µm was experimentally investigated. Special emphasis has been put on the influences of channel size and liquid physical properties on gas-liquid flow patterns. Flow pattern maps were constructed and compared with existing models in literature in order to discuss the transition trends from one flow pattern to another. A new model was proposed which takes the effects of channel size and liquid physical properties into account. Based on the analysis and discussions above, we list the main observations as follows:
(1). Typical flow patterns, i.e. bubbly, slug, slug-annular, annular and churn flow were observed during all the tests. In addition, transitional flow patterns, such as bubbly-slug and unstable-slug flow as well as wavy flow also appeared. Dispersed bubbly flow and dispersed flow were not observed.
(2). Channel size influences the two-phase flow patterns remarkably. With decreasing channel diameter, the transitions from slug to bubbly flow, to churn and to annular flow occur at higher UGS and ULS. In other words, the surface tension dominated region as defined by Akbar et al. (2003) widened.
(3). Liquid physical properties (viscosity and surface tension) have obvious effects on two-phase flow patterns. On one hand, when liquid viscosity increases, the transitions from slug to churn and to slug-annular flow happen at higher UGS and ULS, while the transition to bubbly flow remains almost unvaried. On the other hand, the smaller the surface tension is, the earlier the transitions from slug to bubbly flow, to churn flow and to slug-annular flow, which can be considered as shrinkage of surface tension dominated region.
(4). The flow pattern maps presented by Triplett et al. (1999) and by Hassan et al. (2005) correspond well to our experimental results. However, the maps of Yang and Shieh (2001) are in poor agreement with the present results.
(5). The Akbar’s transition model can only predict the transition from slug-annular to annular flow under our tested conditions. The transition model proposed by Waelchli and Rudolf von Rohr (2006) can not well predict the present flow regime maps either.
(6). A new model which considers the effects of channel size and liquid physical properties is proposed. Three empirical correlations are proposed for predicting the transitions from slug to bubbly flow, from slug to churn and slug-annular flow, and from churn to slug-annular and annular flow, respectively.
Gas-liquid two phase pressure drop characteristics will be presented in the next chapters, in which the main trend of pressure drop with the UGS and ULS and the dependent of pressure drop on the two-phase flow patterns will be presented. The pressure drop characteristics of pressure drop in the inertia-dominated region and that in surface tension-dominated region will be deeply discussed, respectively.
Table of contents :
CHAPTER 1: LITERATURE REVIEW
1.1 General introduction
1.2 Hydrodynamics of two-phase flow in microchannels
1.2.1 Microchannel definition
1.2.2 Two-phase flow patterns
1.2.3 Two-phase flow regime maps
1.2.4 Effects of channel diameters and liquid properties on two-phase flow patterns
1.2.5 Two-phase pressure drop in microchannels
1.3. Two-phase mass transfer and reaction in microreactors
1.3.1 Two-phase mass transfer in microreactors
1.3.2 Gas-liquid reaction in microreactors
CHAPTER 2: GAS-LIQUID FLOW PATTERNS IN CIRCULAR MICROCHANNELS
2.1 Experimental set-up
2.1.1 Test rig
2.1.2 Test section
2.1.3 Working fluids
2.1.4 Parameters measurement and uncertainty analysis
2.2 Flow patterns and flow regime maps in horizontal circular microchannel with a Y junction
2.2.1 Nitrogen-water flow – Influence of channel size
2.2.2 Nitrogen-CMC solution horizontal flow in circular microchannel – influence of viscocity
2.2.3 Nitrogen-SDS solution and Nitrogen-ethanol horizontal flow in circular microchannel influence of surface tension
2.3 Comparison with existing flow regime maps and available transition criteria
2.3.1 Comparison with existing flow regime maps
2.3.2 Comparison with available transition criteria
2.4 Transition criteria based on the present experimental data
CHAPTER 3: TWO-PHASE PRESSURE DROP MODEL FOR INERTIA-DOMINATED REGION
3.2 Experimental condition and data reduction
3.3 Dependence of pressure drop on the flow regimes
3.4 Two-phase pressure drop in circular horizontal microchannel in inertia-dominated region
3.4.1 The comparison of experimental results with existing homogeneous-flow model correlations
3.4.2 The comparison of experimental results with existing Lockhart-Martinelli model correlations
3.5 New correlation for modified Lockhart-Martinelli model
3.5.1 The influence of channel diameter
3.5.2 The influence of liquid viscosity and surface tension
3.5.3 The proposal of new C correlation
3.5.4 Verification of the modified Lockhart-Martinelli model using presently proposed C correlation
CHAPTER 4: HYDRODYNAMIC CHARACTERISTICS OF TAYLOR FLOW IN CIRCULAR MICROCHANNELS
4.2 Results and discussion
4.2.1 Taylor bubble velocity and void fraction
4.2.2 Formation mechanism of Taylor bubbles
4.2.3 Pressure drop characteristics
CHAPTER 5: OXIDATION OF HYDROGENATE 2-ETHYLTETRAHYDROANTHRAQUINONE IN CIRCULAR MICROCHANNEL
5.2.1 Set up
5.2.2 Working solution and analytical method
5.3 Oxygen-anthraquinone working solution two-phase flow in circular microchannel
5.3.1 Two-phase flow patterns
5.3.2 Pressure drop
5.4 Oxidation of THEAQH2 in circular microchannel
5.4.1 Gas-liquid specific interfacial area
5.4.2 Effects of temperature on oxidation
5.4.3 Effects of operating pressure
5.4.4 Effects of liquid velocities on the oxidation