Journal of Visualization

, Volume 21, Issue 4, pp 585–596 | Cite as

Visualisation of gas–liquid bubbly flows in a large diameter pipe with 90\(^{\circ }\) bend

  • Safari Pour Sirous 
  • Krishna Mohanarangam
  • Sara Vahaji
  • Sherman C. P. Cheung
  • Jiyuan Tu
Regular Paper


Two-phase gas–liquid flows are prevalent in many industries and understanding their behaviour would have significant impact on the efficiency of the systems in which they occur. However, information on two-phase gas–liquid flows in \(90^\circ\) bends is limited in the literature and their flow behaviour is not fully understood. One technique that could assist researchers in exploring flow behaviour is visualisation. Accordingly, in this study a two-phase flow experimental investigation was carried out in a large pipe of diameter 150 mm, using water and air at different superficial velocities in order to visualise the effect of \(90^\circ\) bend on two-phase flow behaviour. As optical methods are not suitable for visualising dense bubbly flows due to overlapping of bubbles, in this study, bubble size distribution and void fraction results were obtained using wire-mesh sensors before and after the bend. The results were then post-processed to visualise the flow field. The instantaneous visualisation of flow shows that gas hold-up migrates from the bottom to top wall of the pipe at the bend when the liquid superficial velocity increases for a fixed superficial gas velocity. An increase in superficial gas velocity shows insignificant influence on the gas hold-up at locations beyond the bend for the investigated conditions. This may be due to the centrifugal force imparted by the bend and hence needs further investigation. Bubble size distribution results before and after the bend indicate that the bend has influence on bubble breakup and coalescence.

Graphical Abstract


Two-phase flow Wire-mesh sensor Void fraction 3D bubble Bubble size distribution Instantaneous flow 

1 Introduction

In recent years, two-phase flow has been studied intensively in vertical (Mishima et al. 1984; Taitel et al. 1980; Qi et al. 2012; Schlegel et al. 2009; Kelessidis et al. 1989) and horizontal pipes (França and Lahey 1992; Taitel and Dukler 1976; Mathure 2006; Bottin et al. 2014) to characterise the flow regimes. However, in most industrial two-phase applications, the flow regime and behaviour caused due to connections such as in a \(90^\circ\) bend has not been fully explored.

In single-phase flow, frictional pressure drop and separation effects are some of the main causes of flow complexity, making it difficult to predict the flow behaviour in the \(90^\circ\) bend. Flow complexity increases to even greater extent through the introduction of more phases into the flow system. Dean (1928) conducted a theoretical study to assess the stability of flow through curved pipes under pressure. The results suggested that the stability of flow through the curves depends upon the flow inertia, viscosity and the ratio of pipe diameter to its radius of curvature.

Detra (1953) studied the secondary flow in a pipe curve using experimental results and developed a theory to explain the secondary flow phenomena within it. This theory also explained the effect of vorticity distribution on streamlines which creates secondary flow phenomena in single-phase flow.

Assessing two-phase flow condition, Yadav et al. (2014a) conducted an experiment to investigate the geometrical effect of the \(90^\circ\) bend on the structure of air and water flow, mostly in the bubbly flow regime. The study focused on the flow parameters and dissipation region in the upward elbow. In their study, a four-element conductivity probe of Kim et al. (2000) was used to measure the void fraction, bubble velocity, and interfacial area concentration in a 50.8-mm-diameter transparent pipe. Their results indicated that the velocity profile of the two phase is close to the single phase in bubbly flow regime for the selected pipe diameter size. Additionally, their void fraction results suggested a symmetry about the vertical axis with the maximum value at a ratio of pipe radius divided by the distance from the centre to the pipe wall. Their study further concluded that the elbow upstream of flow created secondary flow in the bubbly flow regime, and this elbow effect dissipates within a distance of 3–9 times the pipe diameter.

Abdulkadir et al. (2011) also studied the effect of \(90^\circ\) bend. In their study, they used a capacitance wire-mesh sensor to assess the phase distribution of silicon oil and air. The air velocity range was 0.05–4.37 m/s with the liquid superficial velocity range being 0.05–0.38 m/s, in a pipe diameter of 0.067 m and the length of 6 m. The results from the wire-mesh sensor was compared to the void fraction measurement from the electrical capacitance tomography (ECT) and visualised with the aid of a high-speed camera. By increasing the air superficial velocity, the authors identified two-phase distributions through the bend. One was unstable-slug and the other was churn flow regime.

In another study, behaviour of churn-annular flow regime and liquid film fraction in \(180^\circ\) return bend was investigated using electrical conductance tomography. The study indicated the gravitational effect causes liquid film breakup to occur in the case of high gas flow with low liquid flow at \(45^\circ\) of the return bend (Abdulkadir et al. 2012).

Abdulkadir et al. (2013) conducted a comparison study between their experimental data and computational modelling for slug flow in a vertical riser with \(90^\circ\) bend. Their experimental results obtained using electrical capacitance and wire-mesh tomography techniques while volume of fluid (VOF) method and k\(\epsilon\) turbulence model was used in CFD modelling. Their study indicated a good agreement between the experiment results and CFD predication in term of time series and population density function of void fraction.

Almabrok et al. (2016) investigated the influence of \(180^\circ\) return bend on the behaviour of downward gas–liquid flow using wire-mesh sensor and liquid film thickness probe. The results showed that the development of axial downward flow occurs in the distance of 30 pipe diameters after the bend. This development distance is also reported for liquid film thickness in annular flow regime.

In another experimental study, Aliyu et al. (2016) proposed a new empirical correlation to determine the interfacial shear for downward annular flow in the large pipes after U-bend. Aliyu et al. (2016) suggested that 46 times the pipe diameter is a sufficient liquid film development length for downward co-current annual gas–liquid flow, after a U-bend with an internal diameter of 101.6 mm. However, Aliyu et al. (2017) indicated that higher multiple of 47 times the pipe diameter is the appropriate length distance for liquid film development in upward gas–liquid annual flow conditions. Aliyu et al. (2017) also provided an improved correlation to predict the liquid film thickness in the large pipes.

Vieira et al. (2014) used a wire-mesh sensor to measure the void fraction distribution before and after a \(90^\circ\) horizontal bend in a 76.2 mm internal diameter and 18 m long pipe. Their results were used to identify the stratified-slug transition in stratified wavy and annual flow regimes using cross-sectional time averaged void fractions. The presence of the \(90^\circ\) bend showed a slight increase of cross-sectional averaged void fraction after the bend, regardless of liquid viscosity. In addition, the cross-sectional time averaged void fraction distribution changed after the bend. Measurements obtained using high viscosity liquids showed the occurrence of slug flow at very low pressure, low gas and low liquid flow rate. In their research, a PDF (population density function) was adopted to distinguish between non-stable stratified flow and stratified wavy flow.

Yadav et al. (2014b) categorised the dissipation of elbow effect in two-phase flow. They proposed a model to investigate the required pipe length needed after the bend to achieve fully developed horizontal flow. For this proposed, the pipe line was divided into three sections of vertical upward, elbow region and developed horizontal. Based on their results, they formulated an experimental correlation to categorise dissipation due to the effect of an elbow. Their correlation involved a linear function of mixture Reynolds number and the inverse of the second moment of local void fraction.

Qiao et al. (2017) further adjusted the experimental setup of Yadav et al. (2014b) to obtain a modified correlation that enable them to determine the region of uniform distribution where the elbow effect would approach zero. In their correlation, Qiao et al. (2017) were able to show that the elbow effect increases in the transition region (unimodal to bimodal) but decreases in the post-elbow dissipation region (bimodal to single peaked) and stays constant in the asymptotic region (fully developed horizontal flow).

Spedding and Benard (2007) experimentally studied the effect of a pressure drop on a \(90^\circ\) bend when it transitions from a vertical to a horizontal pipe in two-phase flow. The study indicated that the bend creates restriction, which will increase the pressure drop through the vertical pipe. The study also indicated that the pressure drop is more pronounced as the rate of gas flow is increased. It was observed that the slug flow in the vertical pipe before the bend transitions to a stratified flow in the horizontal pipe. In addition, a pressure drop change was also observed during a transition regime from churn to semi-annular flow through the vertical pipe.

Notwithstanding the research outlined above, there remain gaps in the current understanding of two-phase flow through \(90^\circ\) bends in large pipes. This issue can begin to be addressed through the application of flow visualisation. Flow visualisation may, for example, be able to reveal the influence of different liquids and air superficial velocities on the two-phase flow behaviour through \(90^\circ\) bends. However, there are certain flow visualisation techniques, for example, optical imaging, that may not reveal vital information about two-phase flow behaviour, especially when concerned with high gas void fraction regimes. This is due to likelihood of overlapping bubbles in these regimes. Delhaye and Cognet (1984) also highlighted that visualisation techniques require high-resolution imaging to capture small bubbles or droplets moving rapidly in the media. This is because the presence of liquid film on the pipe wall can counteract the required resolution. Therefore, in this study, tomography technique based on the conductivity difference of two fluids is utilised. Even with the presence of overlapping bubbles or liquid film, this high-resolution technique allows for the capture of flow field characteristics, including the velocity, bubble size distribution, and void fraction. This is done with the aid of two individual wire-mesh sensors separated by a distance of 35 mm at two pipe locations i.e., before and after the bend.

Most of the studies so far have been carried out in either vertical or horizontal pipe flows both large and small. Very few studies have been reported on the effect of bends in two-phase flows usually in small diameter pipes. To the best of author’s knowledge, the influence of \(90^\circ\) bend in large industrial sized pipes of 150 mm have not been reported despite its prevalence in many mineral and chemical processing plants where the use of \(90^\circ\) bend is inevitable. It is also worth noting that gas–liquid flow behaviour in small pipes differs to the large pipes (Shen et al. 2014; Farman Ali and Yeung 2015; Sharaf et al. 2016).

2 Experimental setup

The experimental two-phase flow loop was constructed using 150 mm nominal PVC pipe diameter and \(90^\circ\) stainless steel bend with 228.5 mm centre line radius. Filtered water flows from the 5000 L holding tank driven by a 50 HP centrifugal pump. Figure 1 shows the experimental setup for the current study. The water flow is controlled by a variable frequency drive and measured by a Yokogawa® AXF series magnetic flow meter with the accuracy of \(\pm 0.35\%\). Figure 2 shows two-phase flow direction through the bend. Air is injected circumferentially around the inside pipe wall through 50 holes each of 1 mm diameter. Figure 3 shows the gas injection flange. The air mass flow rate is measured using a Micro Motion® Coriolis mass flow meter F-series with accuracy up to \(\pm 0.5\%\).
Fig. 1

Experimental setup

Fig. 2

Wire-mesh sensors location before and after the bend

Fig. 3

Gas injection system

Table 1

Investigated flow conditions

Case no.

Case 1

Case 2

Case 3

Case 4

\(J_{g_s}\) (m/s)





\(J_{l_s}\) (m/s)





The list of flow conditions used in the current study are presented in Table 1. Two liquid superficial velocities of 1.02 and 2 m/s were investigated in this study. The liquid and air superficial velocities studied have been plotted in the air-liquid flow regime map for vertical pipe flow of Mishima et al. (1984) and Taitel et al. (1980) as shown in Fig. 4. The conditions tested in this study have also been plotted against Taitel and Dukler (1976) and Mandhane et al. (1974) flow regime map for horizontal pipe flows as shown in Fig. 5. Measurements were conducted before and after the \(90^\circ\) bend for a duration of 60 s at a 2500 Hz in ambient conditions. Figure 6 shows the basic schematic of the wire-mesh sensor. The post-processed data from the sensor was then used to construct the 3D images of instantaneous bubbly flow inside the pipe as a function of distance. Figure 7 shows the steps taken to create data in MATLAB® (MathWorks 2017) for the 3D visualisation inside the pipe using raw data from the wire-mesh sensor, flow conditions and pipe dimensions.
Fig. 4

Flow map for vertical pipe flow

Fig. 5

Flow map for horizontal pipe flow

Fig. 6

Basic schematic of wire-mesh sensor.

Adapted from Pietruske and Prasser (2007)

Fig. 7

Algorithm for data conversion and 3D visualisation

3 Flow field measurement

Two-phase flow analysis and measurement has been the subject of much research over past decades. Understanding the flow regime and characteristics of two-phase flow in a bubble column and in large industrial pipe lines can be achieved by two general measurement techniques: invasive and non-invasive.

The difficulty of using an optical instrument to study densely dispersed phases or opaque liquid flow along with the operational cost of radiometric tomography (gamma, X-ray) raised interest in the development of the wire-mesh sensor to measure bubble size distribution and phase interaction. This has the advantage of relatively low operating costs compared to non-invasive tomography while also offering higher time resolution (Prasser et al. 2001). The idea for a wire-mesh sensor followed its invention which was intended to measure the percentage of water content in crude oil flow in a conduit. The method is based on measuring the time varying voltage at a number of electrical nodes to determine the conductivity of the flow at each node (Johnson et al. 1984).

The wire-mesh sensor tomography has been further developed incorporating two planes of electrodes separated from each other with the wires in each plane perpendicular to each other. The voltage pulse is transmitted to the first plane (transmitter) and if water (conductive fluid) exists between the planes then a signal will be received by second plane (receiver) (Prasser et al. 1998).

In terms of validity of void fraction results obtained from the wire-mesh sensor, Prasser et al. (2005) compared the void fraction results from the wire-mesh sensor to X-ray tomography and good agreement between the two method was found with low deviation of \(1\%\) for bubbly flow and \(4\%\) for slug flow.

In this study, a dual wire-mesh sensor with an internal diameter of 142.7 mm was used before and after a \(90^\circ\) bend with a 228.5 mm centre line radius. Each sensor consists of 32 individual transmitter and receiver wires of 1 mm in diameter. The distance between each transmitter (or receiver) wire is 4.5 mm. The transmitter and receiver planes are located perpendicular to each other and separated by a distance of 3 mm. The bend is fixed at a \(V/D= 8.4\) (vertical distance-to-pipe diameter ratio) from the air injection location.

4 Results and discussion

4.1 Instantaneous visualisation of void fraction

Visualisation of two-phase flow before and after a bend can reveal vital information on the influence of the bend on the flow structure. The presence of the bend can create unsymmetrical bubble size distribution after the bend or at the entrance of bend in certain flow conditions. For example, the vorticity distribution and secondary flow velocity are strongly influenced by inlet stagnation pressure distribution of a bend (Detra 1953).

Figure 8 shows the instantaneous visualisation of the void fraction before and after the bend for case 1 and case 2, respectively, where the liquid superficial velocity was maintained constant at 1.02 m/s while for case 3 and 4 liquid superficial velocity was maintained at 2 m/s as shown in Fig. 9. The experimental results for void fraction show that the primary cause of flow pattern change after the bend could be due to the increase of liquid superficial velocity. The increase of liquid flow rate causes a shift in the void fraction after the bend while the increase of air flow rate does not substantially change the pattern of void fraction after the bend. It appears that the location of average void before the bend would determine the void location after the bend and liquid superficial velocity strongly influences the location void after the bend.

The results show that the change of superficial gas velocity is insignificant compared to the increase of liquid superficial velocity. The significance of liquid superficial velocity can be seen by considering the Froude number \(Fr=J^2_{\mathrm{sl}}/(gR\sin \theta )\), where \(J_{\mathrm{sl}}\) is the liquid superficial velocity, g is acceleration due to gravity and R is the radius of curvature and \(\theta\) is angle from horizontal plane.
Fig. 8

Instantaneous visualisation of case 1 and case 2 before and after the bend

Fig. 9

Instantaneous visualisation of case 3 and case 4 before and after the bend

Table 2 shows the Froude number based on liquid superficial velocity for the investigated cases where case 1 and case 2 have greater than unity before the bend and less than unity after bend. For case 3 and case 4 the Froude number is greater than unity before and after the bend. Gardner and Neller (1969) experimentally observed that for two-phase flow through a \(90^\circ\) bend with less than unity, air tends to preferentially concentrate towards outside of the bend and for greater than unity air tends to segregate toward the inside of the bend. Our experimental observations align with the findings of Gardner and Neller (1969). In future studies, experimental observation inside the bend would be needed to verify the variation of Froude number as shown in Table 2.
Table 2

Froude number for horizontal to vertical angle (after bend)


\(10^{\circ }\)

\(20^{\circ }\)

\(30^{\circ }\)

\(40^{\circ }\)

\(50^{\circ }\)

\(60^{\circ }\)

\(70^{\circ }\)

\(80^{\circ }\)

\(90^{\circ }\)

\(Fr_{case\, 1,2}\)




















4.2 Cross-sectional view of air void fraction

The average cross-sectional void fraction was measured for a duration of 60 s at a frequency of 2500 Hz before and after the \(90^\circ\) bend. Figure 2 shows the wire-mesh sensor locations before and after the bend and location of axis AA’ and BB’ corresponding to the cross-sectional void fraction. Figures 10 and 11 show the average void fraction results for the liquid superficial velocity of 1.02 m/s at two different superficial gas velocities of 0.14 and 0.22 m/s, respectively. The results indicate that the air occupies the inner radius of the pipe before the bend and appears to be on the top of horizontal pipe after the bend.
Fig. 10

Case 1 cross-sectional average void fraction, before and after \(90^\circ\) bend

Fig. 11

Case 2 cross-sectional average void fraction, before and after \(90^\circ\) bend

Cases 1 and 2 both exhibit similar void fraction pattern before and after the bend where the difference between the cases is due to different air superficial velocities. Figures 12 and 13 show the averaged void fraction for case 3 and case 4, respectively. Both cases 3 and 4, have the liquid superficial velocity of 2 m/s for two gas superficial velocities of 0.14 and 0.22 m/s. The results indicate that the average void fraction is similar for both cases towards the outer radius of the pipe before the bend. After the bend, for both cases (3 and 4), the void fraction continues to be higher towards the inner radius of pipe transitioning to the horizontal section.
Fig. 12

Case 3 cross-sectional average void fraction, before and after \(90^\circ\) bend

Fig. 13

Case 4 cross-sectional average void fraction, before and after \(90^\circ\) bend

4.3 Bubble size distribution

The average bubble size distribution is determined based on integrating the local void (\(\epsilon _{ijk}\)) fraction of identified bubbles of same size class. The local void fraction is also related to volume of bubble and total volume as \(\epsilon _{ijk}=V_{\mathrm{{bubble}}}/V_{\mathrm{{total}}}\). The volume of bubble is calculated by \(V_{\mathrm{{bubble}}}=\Delta X \Delta Y \Delta t \,w_{\mathrm{g}}\sum \epsilon _{ijk}\) where \(\Delta X\), \(\Delta Y\) are the distance between two sensor wires (4.5 mm), \(\Delta t\) is based on the acquisition frequency and \(w_{\mathrm{g}}\) is the average gas phase velocity, thereby providing the volume of the bubble and consequently diameter of the bubbles. The averaged gas velocity is obtained by cross-correlating the local instantaneous void fraction signal from each mesh point of sensors. The averaged cross-correlation value is then used with time shift and distance between the two sensors (35 mm) to obtain the average gas phase velocity. The details algorithm of bubble size distribution calculation can be found in Prasser (2007) and Prasser et al. (1998).

The accuracy of the bubble size distribution results from the wire-mesh sensor was previously studied by Nuryadin et al. (2015). The author concluded that at low superficial liquid velocity (0–0.43 m/s) deceleration of bubbles would contribute to the error in estimation of bubble size. However, as liquid superficial velocity increases to 0.6 m/s the deceleration would reduce and eventually disappear due to increase of bubbles inertial compare to surface tension effect.

Ito et al. (2011) reported an increase of liquid inertia above a certain range will stabilise the bubble shape. They described that this occurs by accelerating the bubbles through sensor wires, against superficial forces. Further, at low liquid inertia the surface tension will decelerate the gas–liquid interface, when in contact with wires, which causes bubble deceleration. Similarly, Wangjiraniran et al. (2003) suggested that for liquid velocity below 0.2 m/s the wire intrusiveness will cause deceleration. Therefore, in this study the liquid superficial velocity are carefully chosen between 1.02 and 2 m/s to limit the intrusiveness effect of wire-mesh sensor on bubble size distributions. To further reduce the effect of wire-mesh sensor intrusiveness, the two individual wire-mesh sensors (separated by 35 mm distance) were first installed before the bend for measurements. Later the setup was moved after the bend and the experiments repeated.

Figures 14 and 15 show the bubble size distribution for cases 1, 2, 3 and 4, respectively. The general trend of the distributions for bubble size indicate that all cases show generally two groups of bubble size classes of 2–20 and 40–80 mm. Case 1 and case 2 size distribution indicates the smaller range bubble diameters after the bend is more pronounced. However, for the case 3, the small range bubble diameter after the bend appeared to be less than before the bend. A notably pattern is also observed for case 4.
Fig. 14

Average bubble size distribution before and after bend

Fig. 15

Average bubble size distribution before and after bend

Figure 15 shows a slightly different bubble size distribution compared to Fig. 14, in which the distribution is more scattered around bubble size group 40–80 mm. The shift from smaller size group (1–10 mm) for cases 1 and 2 to bigger size group (40–80 mm) in cases 3 and 4 indicates that bubble breakup is more pronounced at the lower liquid superficial velocity and or bubble coalescence is more pronounced at higher liquid superficial velocity. Comparing case 3 and case 4, the indication is that the coalescence phenomena is more active for lower gas superficial velocity.

5 Conclusion

In this study, the three-dimensional flow of gas–liquid in a large pipe is visualised before and after a \(90^\circ\) bend. The cross-sectional averaged void fraction and bubble size distribution reveal the effect of the bend for different liquid and air superficial velocities. It can be concluded that the increase in superficial liquid velocity has a strong influence on the location of the air void fraction after the bend. The bubble location is shifted from the top portion of horizontal pipe after the bend to the bottom of the horizontal pipe with increasing liquid superficial velocity. In addition, bubble coalescence is seen to be dominant at lower liquid superficial velocity after the bend while at higher liquid superficial velocity the bubble breakup is more pronounced.



The author acknowledges the support received through the provision of an Australian Government Research Training Program in form of a Ph.D. scholarship and RMIT university support. The author also appreciates CSIROs experimental facilities and technical support received from Mr. Andrew Brent.


  1. Abdulkadir M, Zhao D, Sharaf S, Abdulkareem L, Lowndes IS, Azzopardi BJ (2011) Interrogating the effect of 90deg bends on air-silicone oil flows using advanced instrumentation. Chem Eng Sci 66(11):2453–2467CrossRefGoogle Scholar
  2. Abdulkadir M, Zhao D, Azzi A, Lowndes IS, Azzopardi BJ (2012) Two-phase air–water flow through a large diameter vertical 180\(^{\circ }\) return bend. Chem Eng Sci 79:138–152CrossRefGoogle Scholar
  3. Abdulkadir M, Hernandez-Perez V, Lo S, Lowndes IS, Azzopardi BJ (2013) Comparison of experimental and computational fluid dynamics (CFD) studies of slug flow in a vertical 90 bend. J Comput Multiph Flows 5(4):265–281CrossRefGoogle Scholar
  4. Aliyu AM, Lao L, Almabrok AA, Yeung H (2016) Interfacial shear in adiabatic downward gas/liquid co-current annular flow in pipes. Exp Therm Fluid Sci 72:75–87CrossRefGoogle Scholar
  5. Aliyu AM, Almabrok AA, Baba YD, Lao L, Yeung H, Kim KC (2017) Upward gas/liquid two-phase flow after a U-bend in a large-diameter serpentine pipe. Int J Heat Mass Transf 108:784–800CrossRefGoogle Scholar
  6. Almabrok AA, Aliyu AM, Lao L, Yeung H (2016) Gas/liquid flow behaviours in a downward section of large diameter vertical serpentine pipes. Int J Multiph Flow 78:25–43CrossRefGoogle Scholar
  7. Bottin M, Berlandis JP, Hervieu E, Lance M, Marchand M, Öztürk OC, Serre G (2014) Experimental investigation of a developing two-phase bubbly flow in horizontal pipe. Int J Multiph Flow 60:161–179CrossRefGoogle Scholar
  8. Dean WR (1928) Fluid motion in a curved channel. Proc R Soc A Math Phys Eng Sci 121(787):402–420CrossRefzbMATHGoogle Scholar
  9. Delhaye JM, Cognet G (1984) Measuring techniques in gas–liquid two-phase flowsGoogle Scholar
  10. Detra RW (1953) The secondary flow in curved pipes. The Swiss Federal Institute of Technology, ZurichzbMATHGoogle Scholar
  11. Farman Ali S, Yeung H (2015) Experimental study of two-phase air–water flow in large-diameter vertical pipes. Chem Eng Commun 202(6):823–842CrossRefGoogle Scholar
  12. França F, Lahey RT (1992) The use of drift-flux techniques for the analysis of horizontal two-phase flows. Int J Multiph Flow 18(6):787–801CrossRefzbMATHGoogle Scholar
  13. Gardner GC, Neller PH (1969) Phase distributions in flow of an air-water mixture round bends and past obstructions at the wall of a 76 mm bore tube. Inst Mech Eng (G B) 184:93–101Google Scholar
  14. Ito D, Prasser HM, Kikura H, Aritomi M (2011) Uncertainty and intrusiveness of three-layer wire-mesh sensor. Flow Meas Instrum 22(4):249–256CrossRefGoogle Scholar
  15. Johnson ID (1984) Method and apparatus for measuring water in crude oil. US Patent 4,644,263Google Scholar
  16. Kelessidis VC, Dukler AE, Kelessidist VC, Dukler AE (1989) Modelling flow pattern transitions for upward gas–liquid flow in vertical concentric and eccentric annuli. Int J Multiph Flow 15(2):173–191CrossRefGoogle Scholar
  17. Kim S, Fu X, Wang X, Ishii M (2000) Development of the miniaturized four-sensor conductivity probe and the signal processing scheme. Int J Heat Mass Transf 43(22):4101–4118CrossRefzbMATHGoogle Scholar
  18. Mandhane JM, Gregory GA, Aziz K (1974) A flow pattern map for gas–liquid flow in horizontal pipes. Int J Multiph Flow 1(4):537–553CrossRefGoogle Scholar
  19. Mathure N (2006) Study of flow patterns and void fraction in horizontal two-phase flow. PhD thesis, Oklahoma State UniversityGoogle Scholar
  20. MathWorks (2017) MATLAB® data import and export R2017b.
  21. Mishima K, Ishii M, Kaichiro M, Ishii M, Mishima K, Ishii M (1984) Flow regime transition criteria for upward two-phase flow in vertical tubes. Int J Heat Mass Transf 27(5):723–737CrossRefGoogle Scholar
  22. Nuryadin S, Ignaczak M, Lucas D, Deendarlianto (2015) On the accuracy of wire-mesh sensors in dependence of bubble sizes and liquid flow rates. Exp Therm Fluid Sci 65:73–81CrossRefGoogle Scholar
  23. Pietruske H, Prasser HMM (2007) Wire-mesh sensors for high-resolving two-phase flow studies at high pressures and temperatures. Flow Meas Instrum 18(2):87–94CrossRefGoogle Scholar
  24. Prasser H (2007) Evolution of interfacial area concentration in a vertical air–water flow measured by wire-mesh sensors. Nucl Eng Des 237:1608–1617CrossRefGoogle Scholar
  25. Prasser HM, Zschau J, Bottger A (1998) Development of a wire-mesh sensor for the investigation of two phase flow regimes. In: Annual meeting on nuclear technology ’98, proceedings, pp 119–122Google Scholar
  26. Prasser HM, Scholz D, Zippe C (2001) Bubble size measurement using wire-mesh sensors. Flow Meas Instrum 12(4):299–312CrossRefGoogle Scholar
  27. Prasser HM, Misawa M, Tiseanu I (2005) Comparison between wire-mesh sensor and ultra-fast X-ray tomograph for an air–water flow in a vertical pipe. Flow Meas Instrum 16(2–3):73–83CrossRefGoogle Scholar
  28. Qi FS, Yeoh GH, Cheung SCP, Tu JY, Krepper E, Lucas D (2012) Classification of bubbles in vertical gas–liquid flow: part 1—an analysis of experimental data. Int J Multiph Flow 39:121–134CrossRefGoogle Scholar
  29. Qiao S, Mena D, Kim S (2017) Inlet effects on vertical-downward air–water two-phase flow. Nucl Eng Des 312:375–388CrossRefGoogle Scholar
  30. Schlegel JP, Sawant P, Paranjape S, Ozar B, Hibiki T, Ishii M (2009) Void fraction and flow regime in adiabatic upward two-phase flow in large diameter vertical pipes. Nucl Eng Des 239(2009):2864–2874CrossRefGoogle Scholar
  31. Sharaf S, van der Meulen GP, Agunlejika EO, Azzopardi BJ (2016) Structures in gas–liquid churn flow in a large diameter vertical pipe. Int J Multiph Flow 78:88–103CrossRefGoogle Scholar
  32. Shen X, Schlegel JP, Chen S, Rassame S, Griffiths MJ, Hibiki T, Ishii M (2014) Flow characteristics and void fraction prediction in large diameter pipes. In: Frontiers and progress in multiphase flow I. Springer International Publishing, pp 55–103Google Scholar
  33. Spedding PL, Benard E (2007) Gas liquid two phase flow through a vertical 90 elbow bend. Exp Therm Fluid Sci 31:761–769CrossRefGoogle Scholar
  34. Taitel Y, Dukler AE (1976) A model for predicting flow regime transitions in horizontal and near-horizontal gas-liquid flow. AIChE J 22(1):47–55CrossRefGoogle Scholar
  35. Taitel Y, Bornea D, Dukler AE (1980) Modelling flow pattern transitions for steady upward gas–liquid flow in vertical tubes. AIChE J 26(3):345–354CrossRefGoogle Scholar
  36. Vieira RE, Kesana NR, McLaury BS, Shirazi SA, Torres CF, Schleicher E, Hampel U (2014) Experimental investigation of the effect of 90 standard elbow on horizontal gas–liquid stratified and annular flow characteristics using dual wire-mesh sensors. Exp Therm Fluid Sci 59:72–87CrossRefGoogle Scholar
  37. Wangjiraniran W, MOTEGI Y, Richter S, Kikura H, Aritomi M, Yamamoto K (2003) Intrusive effect of wire mesh tomography on gas–liquid flow measurement. J Nucl Sci Technol 40(11):932–940.
  38. Yadav MS, Kim S, Tien K, Bajorek SM (2014a) Experiments on geometric effects of 90-degree vertical-upward elbow in air water two-phase flow. Int J Multiph Flow 65:98–107CrossRefGoogle Scholar
  39. Yadav MS, Worosz T, Kim S, Tien K, Bajorek SM (2014b) Characterization of the dissipation of elbow effects in bubbly two-phase flows. Int J Multiph Flow 66:101–109CrossRefGoogle Scholar

Copyright information

© The Visualization Society of Japan 2018

Authors and Affiliations

  1. 1.School of EngineeringRMIT University, Bundoora East CampusBundooraAustralia
  2. 2.CSIRO Mineral ResourcesClaytonAustralia

Personalised recommendations