The flow pattern map of a two-phase non-Newtonian liquid–gas flow in the vertical pipe

A map for the determination of flow pattern for two-phase flow of gas and non-Newtonian liquid in the vertical pipe has been presented. Our own experimental data confirm applicability of such a map.     

Review on vertical gas–liquid slug flow

Vertical slug flow is characterized by the rise of long bullet-shaped gas bubbles with a diameter almost matching that of the tube - Taylor bubbles. Liquid slugs separate consecutive Taylor bubbles, which may interact and coalesce if the distance between them is small. Slug flow has numerous industrial applications, being also observed on physiological and geological systems. In spite of the contribution of the development of non-intrusive experimental techniques to a deeper understanding of slug flow features, the complexity of this flow pattern requires the combined use of numerical approaches to overcome some of the optical problems reported in experimental methods, and other limitations related to the flow aperiodic behavior.The need to systematize the large amount of data published on the subject and to understand the limitations of the techniques employed constitutes the motivation for this review. In the present work, literature on vertical gas–liquid slug flow, with Newtonian fluids, from 1943 to 2015, covering theoretical, experimental and numerical approaches, is reviewed. Focus is given to single and trains of Taylor bubbles rising through stagnant and co-current liquids.It should be emphasized, however, that further research still needs to be conducted in some particular areas, namely the hydrodynamics of the liquid film surrounding the Taylor bubbles, the interaction between consecutive bubbles, and a more detailed approach to the flow of Taylor bubbles through co-current liquids.     

Modeling of three-phase heavy oil–water–gas bubbly flow in upward vertical pipes

A bubbly gas–bubbly oil flow pattern may occur when water, heavy oil and gas flow simultaneously in vertical pipes in such a way that water is the continuous phase. In this work, a one-dimensional, thermal, transient two-fluid mathematical model, for such flow, is presented. The model consists of mass, momentum and energy conservation equations for every phase whose numerical solution is based on the finite difference technique in the implicit scheme. The model is able to predict pressure, temperature, volumetric fraction and velocity profiles. For accurate modeling of multiphase flows, the key issue is to specify the adequate closure relationships, thus drag and virtual mass forces for the gas and oil phases were taken into account and special attention was paid on the gas–oil drag force. When this force was included into the model it was found that: (1) such force had the same order of magnitude than the oil drag force and both forces were smaller than the gas drag force, (2) the pressure, gas and oil velocities and gas and oil volume fraction profiles were affected, (3) the numerical stability was increased. The model predictions are in agreement with experimental data reported in literature.     

Two-phase flow patterns across triangular tube bundles for air–water upward flow

This paper presents flow pattern experimental results obtained during two-phase upward flow across a horizontal tube bundle. Experiments were performed for flows across a normal triangular tube bundle with 19 mm OD tubes and transversal pitch of 24 mm. Results were obtained for gas and liquid superficial velocities ranging from 0.13 to 10.00 m/s and 0.02 to 1.50 m/s, respectively. Flow patterns were identified subjectively based on visual observations through side windows, and objectively using the k-means clustering method based on signals of a differential pressure transducer and a capacitive sensor. Bubbles, large bubbles, dispersed bubbles, churn, intermittent and annular flow patterns were identified subjectively. The clustering method satisfactorily identified groups of data corresponding to the distinct flow patterns, which were compared with predictive methods available in the open literature. New predictive methods for transitions between flow patterns are proposed based on the flow patterns identified objectively. The proposed methods predicted accurately the data obtained in the present study as well as experimental results and flow pattern maps available in the open literature for distinct geometries.     

Prediction of the entrained liquid fraction in vertical annular gas–liquid two-phase flow

The study considers the prediction of the entrained liquid fraction in adiabatic gas–liquid annular two-phase flow in vertical pipes. Nine empirical correlations have been tested against an experimental data bank drawn together in this study containing 1504 points for 8 different gas–liquid combinations and 19 different tube diameters from 5.00 mm to 57.1 mm. The correlation of Sawant, Ishii and Mishima and the one of Oliemans, Pots and Trompé were found to best reproduce the available data. A new correlating approach, derived from both physical intuition and dimensional analysis and capable of providing further physical insight into the liquid film atomization process, was proposed and worked better than any of the existing methods. This new correlation is based on the core flow Weber number that is also a controlling dimensionless group in determining the wall shear stress and associated frictional pressure gradient of annular flows.     

Structures in gas–liquid churn flow in a large diameter vertical pipe

Gas–Liquid two phase co-current flow in a vertical riser with an internal diameter of 127 mm was investigated in the churn flow pattern. This paper presents detailed experimental data obtained using a Wire Mesh Sensor. It shows that the most obvious features of the flow are huge waves travelling on the liquid film. Wisps, large tendrils of liquid and the product of incomplete atomisation, which had previously detected in smaller diameter pipes, have also been found in the larger diameter pipe employed here. The output of the Wire Mesh Sensor has been used to determine the overall void fraction. When examined within a drift flux framework, it shows a distribution coefficient of ∼1, in contrast to data for lower gas flow rates. Film thickness time series extracted from the Wire Mesh Sensor output have been examined and the trends of mean film thickness, that of the base film and the wave peaks are presented and discussed. The occurrence of wisps and their frequencies have been quantified.     

Prediction of mono-disperse gas–liquid turbulent flow in a vertical pipe

A two-fluid model in the Eulerian–Eulerian framework has been implemented for the prediction of gas volume fraction, mean phasic velocities, and the liquid phase turbulence properties for gas–liquid upward flow in a vertical pipe. The governing two-fluid transport equations are discretized using the finite volume method and a low Reynolds number $k-\varepsilon$ model is used to predict the turbulence field for the continuous liquid phase. In the present analysis, a fully developed one-dimensional flow is considered where the gas volume fraction profile is predicted using the radial force balance for the bubble phase. The current study investigates: (1) the turbulence modulation terms which represent the effect of bubbles on the liquid phase turbulence in the k–ε transport equations; (2) the role of the bubble induced turbulent viscosity compared to turbulence generated by shear; and (3) the effect of bubble size on the radial forces which results in either a center-peak or a wall-peak in the gas volume fraction profiles. The results obtained from the current simulation are generally in good agreement with the experimental data, and somewhat improved over the predictions of some previous numerical studies.     

Investigation of two-phase flow pattern,liquid holdup and pressure drop in viscous oil–gas flow

An experimental investigation was carried out on viscous oil–gas flow characteristics in a 69 mm internal diameter pipe. Two-phase flow patterns were determined from holdup time-traces and videos of the flow field in a transparent section of the pipe, in which synthetic commercial oils (32 and 100 cP) and sulfur hexafluoride gas (SF₆) were fed at oil superficial velocities from 0.04 to 3 m/s and gas superficial velocities from 0.0075 to 3 m/s.     

Wisp-like structures in vertical gas–liquid pipe flow revealed by wire mesh sensor studies

A conductance wire mesh sensor system has been employed on a vertical 67 mm diameter pipe with the up flow of air and water mixtures. The measuring system provides time and cross-sectionally resolved information about the spatial distribution of the phases. Statistical information can be extracted and used to identify flow patterns. The fully resolved data has revealed a hitherto unreported structure has been seen in churn flow which could be linked to the wisps in wispy-annular flow.     

Performance comparison of artificial neural networks and expert systems applied to flow pattern identification in vertical ascendant gas–liquid flows

Instantaneous readouts of an electrical resistivity probe are taken in an upward vertical air–water mixture. The signals are further processed to render the statistical moments and the probability density functions here used as objective flow pattern indicators. A series of 73 experimental runs have its flow pattern identified by visual inspection assisted by the analyses of the void fraction’s trace and associated probability density function. The flow patterns are classified into six groups and labeled as: bubbly, spherical cap, slug, unstable slug, semi-annular and annular. This work compares and analyzes the performance of artificial neural networks, ANN, and expert systems to flow pattern identification. The employed ANNs are Multiple Layer Perceptrons, Radial Basis Functions and Probabilistic Neural Network, with single and multiple outputs. The performance is gauged by the percentage of right identifications based on experimental observation. The analysis is extended to clustering algorithms to assist the formation of knowledge base employed during the learning stages of the ANNs and expert systems. The performance of the following clustering algorithms: self organized maps, K-means and Fuzzy C-means are also tested against experimental data.     

A theoretical model for gas–liquid slug flow in down inclined ducts

Analytical solutions are obtained for flows in downwardly inclined ducts, partly filled by a liquid and containing finite amplitude moving jumps. A unified theory for both roll waves and periodic slug flows in rounded ducts of arbitrary cross-section is worked out by means of some simplifications. The article is focused on slugs: a set of equations is obtained, which predicts the transition between roll waves and slug regimes and gives access to all flow characteristics without any need of closure laws concerning either the speed of propagation or the slug length. As a result, we gain a new insight on the physical structure of slug flow. The proposed model is valid for sufficient inclination, small pressure gradient along the duct and negligible superficial tension. Owing to assumptions, only main trends and orders of magnitude observed in experiments are to be checked. In this connection the model fits most of the previously published experimental results obtained in ducts of circular cross-section: the domain of occurrence of downwardly propagating slugs is satisfactorily predicted, the limitations in drift velocity and in liquid layer thickness are demonstrated and upwardly propagating slugs are possible.     

Prediction of void fraction for gas–liquid flow in horizontal,upward and downward inclined pipes using artificial neural network

In this work, the ability of artificial neural networks (ANNs) to predict void fraction of gas–liquid two–phase flow in horizontal and inclined pipes was investigated. For this purpose, an ANN model was designed and trained using a total of 301 experimental data points reported in the literature for inclination angles between –20° and +20°. Pipe inclination angle as well as superficial Reynolds number of gas (Re_sg) and liquid (Re_sl) were chosen as input parameters of different structures of multilayer perceptron (MLP) neural networks, while the corresponding void fraction was selected as their output parameter. A hyperbolic tangent sigmoid and a linear function were employed as transfer functions of hidden and output layers, respectively, and Levenberg–Marquardt back propagation algorithm was used to train the networks. By trial–and–error method, a three–layer network with 10 neurons in the hidden layer was achieved as optimal structure of the ANN which made it possible to predict the void fraction with a high accuracy. Mean absolute percent error (MAPE) of 1.81% and coefficient of determination (R²) of 0.9976 for training data and MAPE of 1.52% and R² value of 0.9948 for testing data were obtained. Also for all data, MAPE of 1.95% and R² value of 0.9972 were calculated, and 96% data were within ±5% error band. In addition, the accuracy of the proposed ANN model was compared with the predictions from 17 void fraction correlations available in the literature for different flow patterns and horizontal and inclined flows. For all cases, the proposed ANN model gave better performance than all of the studied correlations. The results confirm the very good capability of the ANNs to predict the void fractions of gas–liquid flow in inclined pipes, regardless of flow pattern. Finally, by performing interpolation using the trained network, the void fraction values for some other conditions were predicted.     

Algebraic turbulence modeling in adiabatic gas–liquid annular two-phase flow

The study considers algebraic turbulence modeling in adiabatic gas–liquid annular two-phase flow. After reviewing the existing literature, two new algebraic turbulence models are proposed for both the liquid film and the droplet laden gas core of annular two-phase flow. Both turbulence models are calibrated with experimental data taken from the open literature and their performance critically assessed. Although the proposed turbulence models reproduce the key parameters of annular flow well (average liquid film thickness and pressure gradient) and the predicted velocity profiles for the core flow compare favorably with available core flow velocity measurements, a more accurate experimental database is required to further improve the models accuracy and range of applicability.     

Effect of gas pulsation on long slugs in horizontal gas–liquid pipe flow

Long liquid slugs reaching a length of several hundreds of pipe diameter may appear when transporting gas and liquid in horizontal or nearly horizontal pipelines. These long slugs may cause system vibration, separator flooding, and operational problems for the downstream processing facilities. Although mainly short hydrodynamic slugs have been observed in offshore gas and oil production fields over the past years, the appearance of the long slugs is becoming more common as many production fields are now more mature and reach end of field life, giving reduced production rates and reduced operational pressure.     

Recognition of gas–liquid two-phase flow patterns based on improved local binary pattern operator

A new method to pattern recognition of gas–liquid two-phase flow regimes based on improved local binary pattern (LBP) operator is proposed in this paper. Five statistic features are computed using the texture pattern matrix obtained from the improved LBP. The support vector machine and back-propagation neural network are trained to flow pattern recognition of five typical gas–liquid flow regimes. Experimental results demonstrate that the proposed method has achieved better recognition accuracy rates than others. It can provide reliable reference for other indirect measurement used to analyze flow patterns by its physical objectivity.     

A coupled model for simulation of the gas–liquid two-phase flow with complex flow patterns

A new model coupling two basic models, the model based on interface tracking method and the two-fluid model, for simulating gas–liquid two-phase flow is presented. The new model can be used to simulate complex multiphase flow in which both large-length-scale interface and small-length-scale gas–liquid interface coexist. By the physical state and the length scale of interface, three phases are divided, including the liquid phase, the large-length-scale-interface phase (LSI phase) and the small-length-scale-interface phase (SSI phase). A unified solution framework shared by the two basic models is built, which makes it convenient to perform the solution process. Based on the unified solution framework, the modified MCBA–SIMPLE algorithm is employed to solve the Navier–Stokes equations for the proposed model. A special treatment called “volume fraction redistribution” is adopted for the special grids containing all three phases. Another treatment is proposed for the advection of large-length-scale interface when some portion of SSI phase coalesces into LSI phase. The movement of the large-length-scale interface is evaluated using VOF/PLIC method. The proposed model is equivalent to the two-fluid model in the zone where only the liquid phase and the SSI phase are present and to the model based on interface tracking method in the zone where only the liquid phase and the LSI phase are present. The characteristics of the proposed model are shown by four problems.     

Two-phase turbulence models for simulating dense gas–particle flows

The two-fluid model is widely adopted in simulations of dense gas–particle flows in engineering facilities. Present two-phase turbulence models for two-fluid modeling are isotropic. However, turbulence in actual gas–particle flows is not isotropic. Moreover, in these models the two-phase velocity correlation is closed using dimensional analysis, leading to discrepancies between the numerical results, theoretical analysis and experiments. To rectify this problem, some two-phase turbulence models were proposed by the authors and are applied to simulate dense gas–particle flows in downers, risers, and horizontal channels; Experimental results validate the simulation results. Among these models the USM-Θ and the two-scale USM models are shown to give a better account of both anisotropic particle turbulence and particle–particle collision using the transport equation model for the two-phase velocity correlation.