The outwardly propagating spherical flames in premixed gas containing water droplets are theoretically studied in this work. The correlations between the flame propagation speed, droplet distribution and flame radius are derived, based on the large activation energy and quasi-planar flame assumptions. With this, flame bifurcation and multiplicity are analysed, focusing on the effects of initial droplet mass loading, evaporative heat loss and Lewis number. Meanwhile, the model can predict different gaseous flame types and liquid droplet distributions, as well as the bifurcations and transitions between them. It is shown that the spherical flame propagation is strongly affected by water droplet properties. When initial loading and/or heat loss coefficient are small, there is only one normal stable flame. Two stable flames arise when they increase, i.e. normal and weak flames. Increased droplet loading mainly affects the weak flame, resulting in decreased propagation speed, increased values of evaporation onset and completion fronts. However, increased heat loss affects both normal and weak flames, and flame bifurcation is observed for large heat loss. Droplet properties also greatly influence the weak flame transition between different regimes. Our results also show that Lewis number has significant influence on droplet-laden spherical flame propagation, in terms of flame bifurcation and regime transition. The Lewis number would affect the flame propagation jointly with the positive stretch rate and/or the evolving temperature gradients near the flame front through the interactions with the dispersed evaporating droplets. Furthermore, the magnitudes of Markstein length of the normal flames decrease when Lewis number approaches unity. However, those of the weak flames are mostly negative, indicating the enchantment over the shown Lewis number range. The larger magnitudes of Markstein length of weak flames show stronger sensitivity to stretch than those of normal flames. Finally, different flame types seen from our theoretical analysis are summarised.

Evaporation rates of low-density water aerosols in a cylindrical acoustic resonator were investigated experimentally, motivated by the potential use of water droplet aerosols in two-phase thermoacoustic devices. Measurements of water aerosol evaporation rates were conducted using a novel direct visualization method combining mass balance quantification based on the light scattering from water droplets and two-dimensional PIV measurements with the droplets acting as the flow visualization particles. Water droplet aerosol evaporation was monitored at a location between the resonator pressure and velocity anti-nodes, while being subjected to a low frequency (110 [Hz]) acoustic standing wave with various acoustic pressure amplitude (APA) values. Application of the acoustic field was found to significantly increase water droplet evaporation rates, exhibiting a linear dependency on the applied APA in the range 600[Pa]

Understanding the near-field region of a spray is integral to optimization and control efforts because this region is where liquid break-up and spray formation occurs, setting the conditions under which the spray dynamics evolve under the gas turbulence and droplet inertia. However, the high optical density of this region complicates measurements; thus, it is not yet well characterized. This paper is intended to compare four of the leading experimental techniques that are being used or developed to study the near-field region of a spray. These techniques are shadowgraphy, tube source X-ray radiography, high-speed synchrotron white-beam X-ray imaging, and synchrotron focused-beam X-ray radiography. Each of these methods is applied to a canonical spray, using the same nozzle, under identical flow conditions. Synchrotron focused-beam radiography shows that a time-averaged Gaussian liquid distribution is a valid approximation very near the nozzle, before the core has broken apart. The Gaussian behavior continues as the spray progresses further downstream, showing self-similarity. A spray angle can be defined from the linear spreading of the Gaussian intensity distribution with downstream distance. The spray angle found from shadowgraphy is validated with focused-beam testing. Additionally, a novel method of estimating the intact length of the spray from different X-ray techniques, that uses broadband illumination, is presented.

Three-dimensional direct numerical simulations of dense suspensions of monodisperse spherical particles in simple shear flow have been performed at particle Reynolds numbers between 0.1 and 0.6. The particles translate and rotate under the influence of the applied shear. The lattice Boltzmann method was used to solve the flow of the interstitial Newtonian liquid, and an immersed boundary method was used to enforce the no-slip boundary condition at the surface of each particle. Short range spring forces were applied between colliding particles over sub-grid scale distances to prevent overlap. We computed the relative apparent viscosity for solids volume fractions up to 38% for several shear rates and particle concentrations and discuss the effects of these variables on particle rotation and cluster formations. The apparent viscosities increase with increasing particle Reynolds number (shear thickening) and solids fraction. As long as the particle Reynolds number is low (0.1), the computed viscosities are in good agreement with experimental measurements, as well as theoretical and empirical equations. For higher Reynolds numbers, we find much higher viscosities, which we relate to slower particle rotation and clustering. Simulations with a sudden change in shear rate also reveal a history (or hysteresis) effect due to the formation of clusters. We quantify the changes in particle rotation and clustering as a function of Reynolds number and volume fraction.

We present a particle method for estimating the curvature of interfaces in volume-of-fluid simulations of multiphase flows. The method is well suited for under-resolved interfaces, and is shown to be more accurate than the parabolic fitting that is employed in such cases. The curvature is computed from the equilibrium positions of particles constrained to circular arcs and attracted to the interface. The proposed particle method is combined with the method of height functions at higher resolutions and it is shown to outperform the current combinations of height functions and parabolic fitting. The algorithm is conceptually simple and straightforward to implement on new and existing software frameworks for multiphase flow simulations thus enhancing their capabilities in challenging flow problems. We evaluate the proposed hybrid method on a number of two- and three-dimension benchmark flow problems and illustrate its capabilities on simulations of flows involving bubble coalescence and turbulent multiphase flows.

The interaction and drag coefficient of three horizontal bubbles with a large bubble (BL) and two side symmetrical small bubbles (BS) in the shear-thinning fluids were simulated numerically using the three-dimensional volume of fluid (3D-VOF) method. The effects of the initial diameter of BL, initial bubble interval and liquid property on the rising velocity and drag coefficient of bubbles were studied. When the BL diameter increased, both the drag coefficients of BL and BS decreased. However, the initial bubble interval increased, the interaction between BL and BS became weakened, which caused the decrease of the drag coefficient of BL but the increase of the drag coefficient of BS. Moreover, the flow index n of liquid had also remarkable influence on bubble drag coefficient, as n decreased, the viscosity of the liquid would reduce, which facilitated the motions of BL and BS, thereby the drag coefficients of both BL and BS decreased. Furthermore, the ratios (φBL and φBS) of drag coefficients of BL and BS to corresponding freely rising single bubble were investigated quantitatively to characterize the interaction between bubbles. The φBL and φBS approached gradually to 1 with the increase of initial bubble interval, indicating that interaction between bubbles became weakened gradually. When the size of BL was increased, the influence of BS on BL would become decreased even negligible, however, the wake effect of BL on BS would be markedly enhanced, thus φBL decreased gradually to approaching 1, and meanwhile φBS would decline from near 1 and then be gradually far away. Besides, the φBL (> 1) decreased to approaching 1 gradually with the increase of flow index n, while the φBS (< 1) increased until close to 1. Moreover, two modified correlations of the bubble drag coefficient for the unequal multi-bubble system in the shear-thinning fluids were proposed, which showed considerable good prediction accuracy.

Spray reactive flow finds application in various technical devices, and due to the complex nature, their optimization is very challenging, requiring proper modeling of turbulence/chemistry interactions as well as of the contribution from spray evaporation. This work presents a study of sub-grid scale combustion models, where relevant assumptions on multiphase coupling and their effects are analyzed in detail. For this purpose, two different flamelet approaches, i.e. progress variable spray flamelet and multi-regime gas flamelet are examined in an implementation coupled with the dynamic thickened flame model, along with which the impact of inlet inhomogeneities condition and droplet evaporation taking into account internal temperature gradient is also investigated. The numerical evaluation is carried out in large eddy simulations of a benchmark ethanol spray flame with partial pre-vaporization, where an Eulerian-Lagrangian numerical framework is adopted. The analysis demonstrated that the flame dynamics under consideration is governed by a close coupling between spray evaporation, turbulent dispersion and unsteady flame propagation at upstream shear layers. Results show that the spray flamelets built from counterflow partially-premixed spray flames achieved a better agreement with experiments, capturing the flame structure in terms of gas-phase temperature, OH mass fraction as well as spray statistics.

The present work aims for the isolated investigation of the influence of system pressure on spray quality of twin-fluid atomizers. An approach of pressure adapted nozzles was applied, allowing for constant mass flows, gas-to-liquid ratio as well as fluid velocities at the nozzle orifice independent of system pressure. Two Newtonian liquids featuring viscosities of 1 and 100mPa · s were used, varying the system pressure from 1 to 16bar for gas velocities of 60, 80 and 100m·s−1. A phase doppler analyzer was applied for measurement of resulting drop size and velocity. Primary breakup morphology was detected by a high-speed camera. Two regions with different dependencies of spray quality on system pressure were identified. Applying pressure adapted nozzles while increasing system pressure, first results in a decrease of droplet size followed by an increase. A maximum of the dynamic pressure of the gas phase was determined at minimum droplet size, which is explained based on the theory of a free jet. The observations are underlined by images of the high-speed camera. Here, a change in breakup morphology from fiber type to a mixture of fiber type and non-axisymmetric Rayleigh type breakup at high system pressure was observed.

Electric field has been proven to be an effective active technique in microfluidic devices for precise manipulating of the microdroplet. In this article, we investigate the Water in Oil droplet formation in a flow-focusing microchannel under AC electric field experimentally and numerically. A three-dimensional numerical model is built combining the Volume of Fraction (VOF) method and the leaky dielectric model, which reveals the droplet formation mechanism under the effects of the electric field. Due to the Maxwell stress induced by the electric field, the sine waveform electric field induces the oscillation at the liquid interfaces, which stimulates the breakup of the disperse phase and thus tunes the droplet size. We analyse the phenomena by the electric capillary number CaE evaluated according to the numerical results. The increase of the electric voltage and the frequency both are able to lift CaE. With the increase of the electric voltage, the droplet generated becomes smaller and the droplet formation turns unstable when CaE>1. The dominating effect of the pressure difference between the disperse phase and the continuous phase shifts from the initially hydrodynamic pressure to the latterly electric field induced one during the evolution of the electric voltage. With a relatively high electric frequency (f ≥5000 Hz, CaE>1), the droplet formation regime transits from dripping to jetting under the constant hydrodynamic conditions. The numerical results show that the surge of the magnitude of the electric body force tends to stretch the disperse phase at liquid interfaces which leads to the transition. This study explored the dynamic mechanism of the droplet formation under AC electric field with different voltages and frequencies which contributes to the in-deep understanding of the coupling effect between the hydrodynamic pressure and the AC electric field induced Maxwell stress and hence, might lead to better control strategy on the promising technology.

Effects of the computational domain size on inertial particle one-point statistics are presented for direct numerical simulations of turbulent open channel flow at a moderate Reynolds number, which are seeded with two-way coupled particles at low volume concentration (less than 1.5×10−3, for such particle load the one-way coupled particles scheme is also valid). Particle one-point statistics across a wide range of Stokes numbers for a small domain (which captures only one or two large-scale motions (LSMs) in the inner layer) and a medium domain (which captures only one or two very large-scale motions (VLSMs) in the outer layer), are compared with those from a reference large domain. Although in single-phase flow the medium domain size simulation reproduces the same fluid one-point statistics as those in a large domain size, in particle-laden flow, comparisons show certain discrepancies in the particle one-point statistics, such as particle accumulation close to the wall (y+<10), maximum values of particle mean-squared streamwise velocity fluctuation, and particle Reynolds shear stress in the inner layer. The difference is larger for moderate Stokes numbers (St+=24.2 and 60.5) compared to low (St+=2.42) and very high (St+=908) Stokes numbers, which is also enhanced by using a small domain size.

Laboratory experiments were conducted to study the dynamics of particle clouds released vertically downward into stagnant water. The importance of nozzle diameter, sand particle mass, and particle size were studied in the form of an aspect ratio and Stokes number. The axial and radial profiles of sand concentration and velocity of particle clouds were measured by an accurate and robust optical probe (PV6). Empirical formulations were developed to explain the variations in sand concentration and velocity profiles. It was found that the zone of jet development was smaller in particle clouds than in single-phase water jets and sand jets. Laboratory measurements also indicated that the centerline sand concentration in particle clouds decreased with a slower rate in comparison to single-phase buoyant and sand jets. Important parameters such as mass, momentum fluxes, and drag and entrainment coefficients were calculated inside particle clouds to better understand the evolution of particle clouds in stagnant water. The radial variation of drag coefficient indicated a particle grouping effect in the core region of particle clouds where the drag coefficients decreased from 0.4 to less than 0.1. The entrainment coefficient decreased non-linearly in the radial direction. A new mathematical correlation was also developed to calculate the distribution of entrainment coefficient inside particle clouds. It was observed that the aspect ratio can significantly alter the radial entrainment coefficient in transverse directions. The inter-particle collision of particle clouds was evaluated by calculating the Bagnold number in both axial and radial directions. It was found that the inter-particle collision occurred for x/do ≤ 10 for St = 0.74 and for St = 0.52, and the Bagnold number values were smaller than 45 for x/do ≥ 20, indicating that a micro viscose regime controls the flow in these particle cloud dynamics.

The boundary element method is adopted to study the ventilating axisymmetric bubble growth from a circumferential slit in a submerged cylindrical wall under vertical flow conditions in this paper. During the bubble growth, mesh subdivision and smoothing techniques are adopted in order to optimize mesh topology. In this study, the effects of inflow velocity, surface tension coefficient, ventilation rate and slit width on the bubble development process are studied. The results show that the bubble shape is mainly affected by the inflow velocity and the surface tension coefficient, the inflow velocity promotes the downward movement of the lower intersection point of the wall and bubble, meanwhile, the surface tension inhibits it. The effect of slit width is mainly reflected in the initial stage of the bubble development. And the bubble size basically increases as the ventilation rate increases. In addition, the effects of the Froude number, Weber number, and ventilation coefficient on bubble growth are analyzed. The maximum dimensionless length and thickness of the bubble increase with the increase of the Froude number and decrease with the increase of the Weber number. As the ventilation coefficient increases, the length decreases first and then stabilizes, the thickness decreases first and then increases.

The study of thermal radiation interacting with particle-laden turbulence is of great importance in a wide range of scientific and engineering applications. The fundamental and applied study of such systems is challenging as a result of the large number of thermo-fluid mechanisms governing the underlying physics. This complexity is significantly reduced by transforming the problem of interest into its scale-free form by means of dimensional analysis techniques. However, the theoretical framework of classical dimensional analysis presents the limitations of not providing a unique set of dimensionless groups, and no support for measuring the relative importance between them. In the interest of addressing these shortfalls for multiphysics turbulent flow applications, we present a semi-empirical dimensional analysis approach to efficiently extract important dimensionless groups from data obtained by means of computational (or laboratory) experiments. The methodology presented is then used to characterize important dimensionless groups in irradiated particle-laden turbulence. The study concludes that two dimensionless groups are responsible for most of the variation in the system’s thermal response, with the absorption of radiation by particles, the radiative energy deposition rate and the turbulent flow mixing the most important thermo-fluid mechanisms. The generality of the results obtained can be leveraged to effectively reduce the dimensionality of irradiated particle-laden turbulent flows in research studies and in the design and optimization of similar systems.

Mass transfer from single carbon-dioxide (CO2) bubbles rising through vertical pipes filled with surfactant−electrolyte mixed aqueous solutions was measured to study the combined effects of surfactant and electrolyte on the mass transfer. Triton X-100 and sodium chloride (NaCl) were used for surfactant and electrolyte, respectively. The surface tension, σeq, at an equilibrium state decreased with increasing the concentration of NaCl at a constant concentration of Triton X-100. The pipe diameters were 12.5 and 18.2mm. A wide range of bubble diameter covered various bubble shapes from ellipsoidal to Taylor bubbles. The combined effects of Triton X-100 and NaCl on the Sherwood number, Sh, showed a different trend from those of alcohol and electrolyte, i.e. bubbles in the Triton X-100−NaCl mixed aqueous solution were not fully-contaminated even with the same σeq as in the condition fully-contaminated with Triton X-100, due to the difference in distributions of surfactant concentrations at the interface. The Sh of bubbles fully-contaminated with Triton X-100 were not affected by the presence of NaCl. The Sh of contaminated Taylor bubbles can be expressed as a combination of functional forms of available Sh correlations for clean and fully-contaminated Taylor bubbles.

Multiphase flow in unsaturated porous media exhibiting hysteretic phenomenon is of great significance to oil and gas recovery, agriculture, and hydrology etc. Water retention curve (WRC) and relative hydraulic conductivity function (RHC) as well as relative air permeability function (RAP) are usually employed to characterize the transport properties. A new fractal capillary model accounting for the hysteresis effect is developed in this paper for the multiphase flow in unsaturated porous media. The analytical expressions for WRC, RHC and RAP under drying and wetting paths are derived accordingly. The proposed fractal capillary model shows good consistent with available experimental results. It has been found that the pore-throat structure in the proposed fractal capillary model takes an important influence on the hysteresis effect and fluid flow, especially for the non-wetting phase. The present work may be helpful for understanding the complex mechanisms of hysteresis effect and transport processes in unsaturated porous media.

Flash-boiling of fuel sprays can occur under injection of superheated fuel into ambient pressure that is lower than the saturation pressure of the fuel and can dramatically alter spray formation due to complex two-phase flow effects and rapid droplet evaporation phenomena. Such phenomena exist in-cylinder at low-load in-city driving conditions where strict engine emission regulations apply, hence the need for faithful flash-boiling fuel spray models by engine designers. To enhance the current modelling capability of superheated fuel sprays, with focus on near-nozzle plume expansion, a flash-boiling breakup modelling approach was developed to introduce the thermal breakup mechanism of droplets caused by nucleation and bubble growth. This model was particularly aimed at sprays where levels of superheat introduced noticeable radial expansion of the plumes upon discharge from the nozzle orifice. The model was able to simulate droplet shattering by introducing Lagrangian child parcels at breakup sites with additional radial velocity components instigated by rapid bubble growth and surface instabilities. Combination of the flash-boiling droplet breakup model with a flash-boiling effective nozzle model that was used as boundary condition for the spray plumes offered a more complete modelling approach, where both in-nozzle phase change effects and near-nozzle flashing through droplet shattering were incorporated into the Eulerian-Lagrangian two-phase computational framework. Sensitivity studies were carried out to investigate important parameters which are inherently difficult to measure experimentally and offered valuable insight into modelling superheated sprays. The model was able to capture important flash-boiling spray characteristics and quantitative validation was achieved through comparison to experimental data in the form of penetration lengths and droplet sizes with a good level of agreement.

Attempts have been made for identification and classification of regimes for upward air-water two-phase flow fusing the response from multiple sensors. A needle type conductivity probe and different combinations of optical probes have been employed. The statistical signatures of the signals have been used for identification of the type of flow and revealing some information regarding the interfacial structure. The signals have further been processed using Principal Component Analysis (PCA) and used to classify the flow regime by k-mean clustering algorithm. Classifications have been made using signals from different individual sensors as well as combining them. Finally, the performance of the predictions made by these sensor combinations have been made using confusion matrix.

The role of surface tension in the mechanism of bubble growth and detachment for a co-flowing air-water two-phase flow in a micro-tube is addressed. A numerical investigation for a horizontal axisymmetric flow with the assumption of zero gravity and an upward flow accounting for gravity contribution is carried out. The continuous liquid phase is flowing in a tube of 500 μm inner diameter and the gas phase is axially injected through a nozzle of 110 μm inner diameter and 210 μm outer diameter. A single-fluid model is used to determine the flow field, solving the continuity and momentum equations associated with the volume of fluid method for interface tracking. An open source software, OpenFOAM, is utilized for solving numerically this problem. The prediction results show that the surface tension plays a double role. First, it keeps the bubble attached to the injection nozzle during bubble growth and neck formation. Then, it destabilizes the interface by pinching off the neck in the immediate vicinity of the nozzle at about a distance of 0.5 the nozzle diameter rather than right at the nozzle exit. In-depth analysis of the mechanism of bubble formation induced by the effect of surface tension is carried out. It is highlighted that this latter acts as an attachment force at the injection nozzle during the bubble growth and it acts over the entire interface of the bubble yielding the formation of a neck. Later, the capillary effects reduce the diameter of the neck until it breaks and yields the detachment of the bubble. Further investigation at the nozzle wall allows depicting the motion of the contact line during the process of bubble growth and its significant effect on the bubble formation.

The collision process between a Newtonian liquid jet and pro/hydrophobic surfaces was investigated in this paper. Based on the theory of impinging-jets, a theoretical model of liquid sheet was established which comprehensively considered the effects of pro/hydrophobicity on surface tension force and the energy loss during the entire impinging process. The sheet parameters such as the size, velocity and thickness distribution were obtained. In order to validate the theoretical model, a corresponding experimental set-up has been developed, using deionized water as experimental working medium and 3 different pro/hydrophobic surfaces as experimental parts. The comparisons showed the theoretical results are in good agreement with the experimental data. In addition, the influence of surface pro/hydrophobicity on the sheet characteristics was theoretically discussed, and the results showed the surface pro/hydrophobicity could significantly affect the size of liquid sheet on solid surface.

In this paper, the cavitating flow around a bluff body is studied both experimentally and numerically. The bluff body has a finite length with semi-circular cross section and is mounted on a surface in the throat of a converging-diverging channel. This set-up creates various 3D flow structures around the body, from cavitation inception to super cavities, at high Reynolds numbers (Re=5.6×104−2.2×105) and low cavitation numbers (σ=0.56−1.69). Earlier studies have shown this flow to be erosive and the erosion pattern varies by changing the flow rate and w/o the cylinder; hence, this study is an attempt to understand different features of the cavitating flow due to the cylinder effect. In the experiments, high-speed imaging is used. Two of the test cases are investigated in more detail through numerical simulations using a homogeneous mixture model. Non-cavitating simulations have also been performed to study the effect of cavitation on the flow field. Based on the observed results, vortex shedding can have different patterns in cavitating flows. While at higher cavitation numbers the vortices are shed in a cyclic pattern, at very low cavitation numbers large fixed cavities are formed in the wake area. For mid-range cavitation numbers a transitional regime is seen in the shedding process. In addition, the vapour structures have a small effect on the flow behaviour for high cavitation numbers, while at lower cavitation numbers they have significant influence on the exerted forces on the bluff body as well as vortical structures and shedding mechanisms. Besides, at very low cavitation numbers, a reverse flow is observed that moves upstream and causes the detachment of the whole cavity from the cylinder. Such a disturbance is not seen in non-cavitating flows.

The initial form of partial cavitation on axisymmetric bodies in some water tunnel tests is a regular system of patches with the approximately constant azimuthal distance between them. The described numerical simulation shows that the origin of such three-dimensional system is in the relatively small non-axisymmetric regular perturbations of the inflow caused by the shapes of non-circular water tunnel cross sections (these sections were squares or hexagons). Computational prediction of cavitation inception number for patch cavitation is carried out by a viscous-inviscid interaction method that combines three-dimensional analysis of inviscid flows with quasi-two-dimensional analysis of viscous flows. Computational results are compared with the experimental data obtained in 1980s in water tunnels of Applied Physics Laboratory and David Taylor Research Center. The satisfactory agreement of computed and measured data is considered as the proof of the described explanation of the origin of patch cavitation.

Cavitation often causes a destructive impact on the performance of hydraulic machinery, such as erosive wear, noise and vibrations of the framework and moving parts of marine propellers, pumps, hydraulic turbines and other equipments, which eventually leads to a degradation of overall system effectiveness. The paper reports on an experimental investigation of a passive method of flow control for different cavitation conditions: starting from the cavitation inception, including quasi-steady partial cavitation with shedding of small-scale vortical structures and finishing by unsteady cloud cavitation. The passive flow control was implemented using miniature vortex generators of a cylindrical type referred to as Cylindrical Cavitating-bubble Generators (CCGs) that were placed on the surface of a benchmark CAV2003 hydrofoil. First, we performed high-speed visualization of cavitation on the suction side of the original hydrofoil (without the control element) to find the cavitation inception point near the leading edge and to analyze the spatial structure and time evolution of partial cavities. In order to improve our understanding of the mechanism of cavitating flow unsteadiness and the effect of CCGs on the cavitation dynamics, we also applied a PIV technique to measure the mean flow velocity profiles and a hydroacoustic pressure transducer to record local pressure pulsations in the hydrofoil wake. As a result, this allowed us to determine the influence of CCGs on turbulent structure of the flow at different cavitation regimes and amplitude-frequency spectra of the pressure pulsations associated with attached cavity length oscillations for unsteady flow conditions. It was revealed that, in the case of unsteady cloud cavitation, CCGs were capable to mitigate large-scale cloud cavities. In addition, a substantial decrease in the amplitude of pressure pulsations was registered for the modified hydrofoil (with the control element). In general, CCGs appeared to be quite effective to hinder the cavitation development and to reduce the strength of side- and middle-entrant jets as the primary mechanisms of unsteady cloud cavitation.

Cyclones are the most widely used separators at present, but it's still difficult to reduce their size to serve as pipe separators. A new type of gas liquid pipe separator for high velocity wet gas is proposed in this paper. The separator is actually a short section of pipe, with a centrifugal device in the center portion and 3 narrow tangential conduits (NTC) in the pipe wall. As gas liquid mixture flows through the centrifugal device downwardly, a strong rotation flow is created. Liquid in the mixture is pushed to the pipe wall by centrifugal force and forms a uniform liquid film with high tangential velocity. Then nearly all the liquid film can directly enter the NTC and be discharged from the pipe relying on its own kinetic energy and inertial, because of the little resistance characteristic of NTC to the liquid film, therefore the whole separation process can be completed within the pipe. The swirl characteristics in the pipe were studied by numerical method, and the simulation results indicated that the swirling liquid film is fairly uniform and the suitable installation position for NTC is about 2.5 pipe diameters downstream of the centrifugal device. A separation model for the pipe separator was established and experiments were carried out to verify the proposed model. The superficial velocity ranges of gas and liquid were 22∼72 m/s and 0.07∼0.54 m/s respectively. The experimental results showed that the separation efficiency always increases with the increase of both gas and liquid superficial velocity, generally it is over 81%, and the maximum attainable is 97%. However, the separation efficiency will begin to drop if liquid superficial velocity exceeds the critical value UCSL, due to the onset of unstable liquid film waves. Experimental results also showed that UCSL increases with the increase of gas velocity.

In this paper, a new formulation based on the continuum surface force model (CSF) is developed for particle methods. In the new formulation, two measures are taken to amend the two drawbacks faced by previous formulations of the CSF model. One measure is that the interface curvature is calculated using the surface divergence of the unit normal instead of the divergence as done in previous formulations. This measure completely removes errors artificially stemming from interior particles that are caused by numerical discretization. Another measure taken is that the size of the particle interaction domain is determined based on the physical characteristic length, rather than taken as a fixed multiple of the particle spacing, as in previous formulations. This treatment is based on the knowledge of the nature of the fractional dimensionality of the surface/curve, for which the measured values of macroscopic characteristics are dependent on the measurement scale. Through the curvature calculation of a unit circle, the relation between the size of the particle interaction domain and the critical size of droplet/bubble breakup is established. The former, which is taken as approximately 0.38 times the latter, is accurate enough to obtain a reasonable interface curvature, even with severe irregularity. The new formulation is tested and verified using three numerical cases, including the oscillation and decay of a square drop under surface tension, the coalescence of two drops after a head-on collision, and the Rayleigh-Taylor instability. It is shown that the new formulation is robust and accurate in calculating the surface tension and is able to adapt to different situations.

Reliable entrainment rate model is of great importance for the prediction of annular flow as it specifies the mass and momentum transfer between the liquid film and the droplet in the gas core. Various entrainment rate models have been proposed, yet few of them have a satisfying performance for a wide range of flow conditions, including gas/liquid phase flow rates, fluid type, channel size, etc. To develop a reliable entrainment rate model that is applicable to various fluids, instrumentation methods and available databases are reviewed in detail. The database for model development is determined by analyzing the instrumentation reliability and eliminating data points with the obvious error. The entrainment rate and the superficial liquid flow rate are nondimensionalized as Reynolds number using Laplace length as length scale, which is a critical wavelength of Taylor instability and related to the droplet size for annular flow. The entrainment rate model is developed based on the new nondimensional numbers, and the performance has been compared with various previous models. The evaluation results show that the prediction accuracy of the new model is better than previous models for a wide range of data.

In this work we introduce a boundary condition for thermal lattice Boltzmann simulations that contain a Dirichlet boundary condition by bouncing back the non-equilibrium distribution of the energy distribution function. To this end the thermal lattice Boltzmann equation is modified by introducing an additional collision term that takes into account the thermal diffusivity and local solid volume fraction of a lattice (partially) covered by the solid phase. Asymptotic analysis of the boundary condition confirms that it is of second order accuracy. The method is validated using (i) an analytical solution for the Nusselt number correlation of a single sphere in an unbounded stationary fluid and (ii) direct numerical simulations of the heat transfer between a fluid and individual particles.

A common feature of wall-bounded turbulent particle-laden flows is enhanced particle concentrations in a thin layer near the wall due to a phenomenon known as turbophoresis. Even at relatively low bulk volume fractions, particle-particle collisions regulate turbophoresis in a critical way, making simulations sensitive to collisional effects. Lagrangian tracking of every particle in the flow can become computationally expensive when the physical number of particles in the system is large. Artificially reducing the number of particles in the simulation can mitigate the computational cost. When particle-particle collisions are an important aspect determining the simulation outcome, as in the case when turbophoresis plays an active role, simply reducing the number of particles in the simulation significantly alters the computed particle statistics. This paper introduces a computational particle treatment for particle-particle collisions which reproduces the results of a full simulation with a reduced number of particles. This is accomplished by artificially enhancing the particle collision radius based on scaling laws for the collision rates. The proposed method retains the use of deterministic collision models and is applicable for both low and high Stokes number regimes.

The air entrapment during droplet impingement is responsible for spontaneous droplet bouncing on an arbitrary solid surface at low Weber numbers. However, for the impact on liquid film surfaces, the outcome would significantly change, making it more favorable for the fabrication of non-wetting lubricant impregnated surfaces (LIS/SLIPS). In this paper, we describe a problem associated with the impact on a liquid surface using a three-phase flow model that captures the details of the gas layer thickness and dynamics of fluid motions. The numerical model was based on the finite volume solution coupled with the volume of fluid method to track the phases. The model was validated with the analytical solution. Consequently, the numerical tool was utilized to investigate the thickness of the entrapped air during the impact process while the behavior of droplet and the immiscible liquid film was quantitatively measured. The morphology of the interfacial gas layer was analyzed for key parameters including impact velocity and film thickness. It was observed that the presence of liquid film can reduce the probability of rupturing the gas layer. The results for the profile of liquid film during droplet impact illustrated that the effect of film thickness can considerably influence the bouncing behavior.

The influence of viscosity and surface tension on bubble dynamics and gas liquid mass transfer has been examined experimentally in a model bubble column using optical measurement methods. Different glycerol-water mixtures, together with the surfactant sodium dodecyl sulphate (SDS), were analysed in this study. Thus the present study covers a viscosity range of 0.86…9.29 mPa•s and a surface tension range of 44.8…72.24 mN/m, accordingly, a Morton number range of 1.71 × 10−11…7.04 × 10−7. Shadowgraphy combined with PTV has been used to determine bubble size, velocity, and bubble shrinkage. The liquid velocity and the dissolved CO2 concentration fields have been analysed by particle-image-velocimetry (PIV) and laser-induced fluorescence (LIF), respectively. The bubble sizes and velocities obtained here show a good agreement with correlations proposed in the literature. With decreasing surface tension and decreasing viscosity, the bubble size is decreased but neither the addition of glycerol nor the addition of surfactant was found to be beneficial to mass transfer from the CO2 bubbles into the liquid. Due to the bubble size and velocity change in different mixtures, the liquid flow field is changed as well. The highest velocities could be found in the high viscosity solutions. The liquid velocity profiles are similar for different viscosities, except in the bottom section of the column where the bubble curtain is more concentrated. These results support the existence of a dual effect of viscosity on bubble column fluid dynamics as already stated in the literature before.

A Diffuse Interface Model (DIM) is employed to model droplet impact on a heated solid surface. The DIM uses an especially constructed solid wall boundary condition which enables simulations with different wetting conditions of the solid surface. The model is also extended to include the effects of surface roughness on the behavior of the contact line dynamics. Multiple simulations are carried out to demonstrate the capabilities of the presented model. The simulation results demonstrate the influence of the wetting properties of the solid, with a higher cooling rate for hydrophilic than for hydrophobic wetting conditions. Surface roughness of the solid surface increases the cooling rate of the solid by enhancing the heat transfer between solid and fluid.

Annular flow with swirl induced by a swirler has been widely used in industry. Swirl decay which closely related to working performance and length of the flow is significantly important for its application in industry. Axial development of annular flow with and without swirl has been investigated in this work by means of experimentation and modelling. A vertical pipe system with 11 meter long and 62 mm inner diameter was carried out to investigate axial development of flow pattern, void fraction and pressure drop in the annular flow with and without swirl. The experimental results show that a swirling annular flow with swirling streak and less disturbed waves is observed downstream of the swirler. Void fraction decreases, total pressure drop increases and the PDF (probability density function) of pressure drop becomes more concentrated in the annular flow with swirl, compared with these in the annular flow without swirl. However, swirling annular flow is gradually transformed to annular flow without swirl along the streamwise direction. To qualitatively predict the decay in an annular flow with swirl, a simplified theoretical model was developed here. The results calculated with the theoretical model were in agreements with experimental results. The obtained results can be used in predicting working performance and length of devices, such as separators and heat exchangers.

Flow behavior of thin liquid films in an annular flow regime is an important element of the thermal hydraulics of a BWR – it controls the heat transfer from the fuel rods to the coolant. Despite its relevance, the flow behavior of dynamic liquid films is not well understood even in adiabatic conditions. To investigate it, unsteady numerical simulations with an interface tracking model were performed in a double subchannel geometry with a P/D ratio of 1.325. Different turbulence models, namely the Large Eddy Simulation (LES) and linear/nonlinear eddy viscosity unsteady-RANS (URANS) models, were tested. A novel approach for generating turbulent inlet conditions in a periodic flow domain was developed to reduce computational efforts. Validation against experimental data revealed shortcomings of the linear eddy-viscosity RANS model in predicting key flow parameters. By capturing the effects of the secondary flow structures in a subchannel geometry, improved predictions were obtained with a non-linear SST k- ω (QCR) turbulence model. Time-averaged liquid film thickness (LFT) and tracer distribution obtained with LES were found to have the best agreement with experimental data. Instantaneous and time-averaged velocity profiles were analyzed to understand the influence of the gas-liquid interface. Secondary flow structures in the subchannel gap region were found to enhance the turbulent mixing of the passive scalar in the liquid film. This finding is relevant towards the prediction of thermal-hydraulic parameters in a multichannel flow assembly by accounting for inter-channel mixing phenomena.

The morphology of liquid-gas flows in corrugated plate heat exchangers is complex due to the intricate channel geometry. To date, only a few experimental visualizations have been performed to study the two-phase flow characteristics in plate heat exchangers. In this paper, we perform pioneering computational fluid dynamics simulations of the adiabatic liquid-gas (water-air) flow in a cross-corrugated channel of a plate heat exchanger. The standard volume-of-fluid technique is used to capture the complex phase-interfaces constructed by the cross-corrugated walls. In order to reduce the computational cost, the computational domain is simplified by a series of assumptions. The bubbly flow, slug flow, churn flow and film flow are modeled by varying the superficial velocities of both phases, and the validity of these patterns is proved by comparison with experimental results. A flow regime map is developed based on the numerical results, and the transitions between the regimes are discussed. The predicted pressure drop shows good agreement with the experimental data as well. The two-phase multiplier for general prediction of the friction factor in the cross-corrugated channel is calibrated. The mean void fraction in the channel is quantified by the numerical simulation. The void fraction model from Zuber and Findlay is found to be applicable to the cross-corrugated channel, which is further modified for general use for these types of channel flows.

The modal and non-modal linear stability analyses of a three-dimensional plane Couette-Poiseuille flow through a porous channel are studied based on the two-domain approach, where fluid and porous layers are treated as distinct layers separated by an interface. The unsteady Darcy-Brinkman equations are used to describe the flow in the porous layer rather than the unsteady Darcy’s equations. In fact, the Brinkman viscous diffusion terms are necessary to capture the momentum boundary layer developed close to the fluid-porous interface. The modal stability analysis is performed under the framework of the Orr-Sommerfeld boundary value problem. On the other hand, the non-modal stability analysis is performed under the framework of the time-dependent initial value problem in terms of normal velocity and normal vorticity components. The Chebyshev spectral collocation method along with the QZ algorithm is implemented to solve the boundary value problem numerically for disturbances of arbitrary wavenumbers. The convergence test of spectrum demonstrates that more Chebyshev polynomials are required to arrest the flow dynamics in the momentum boundary layer once the Couette flow component is turned on. Two different types of modes, so-called the fluid layer mode and the porous layer mode are identified in the modal stability analysis. The most unstable fluid layer mode intensifies while the most unstable porous layer mode attenuates in the presence of the Couette flow component. Further, the mechanism of modal instability is deciphered by using the method of the energy budget. It is found that the energy production term supplies energy from the base flow to the disturbance via the Reynolds stress, and boosts the disturbance kinetic energies for the fluid layer and the porous layer. Moreover, the non-modal stability analysis demonstrates that short time energy growth exists in the parameter space and becomes significant in the presence of the Couette flow component and the permeability of the porous medium.

An overview of gas-liquid slug flow and elongated bubble centring in horizontal circular pipes is given and relevant literature is discussed. A novel hypothesis for the mechanism which induces bubble centring is presented based on the analogy to ideal flow over a solid hemisphere. A model, based on the given hypothesis and various phenomenological simplifications, is outlined and the simplest case is solved and compared to two sources of air-water experimental centring data. The Froude number at which bubble detachment initially occurs is predicted with < 6% error for both cases, using the most reasonable reference point available from past literature. The favourable results found support the fundamental basis of the presented hypothesis and various conclusions are drawn. Discussion is provided on the potential importance of bubble centring in horizontal slug flow involving liquids of relatively high viscosity. Future work is recommended which can further elucidate the theory and importance of bubble centring in horizontal slug flow.

The paper is devoted to a theoretical analysis of a co-current gas-liquid flow between two inclined plates. As a first step, we linearized the Navier-Stokes equations in both phases and carried out a linear stability analysis of the basic steady-state solution over a wide variation of the liquid Reynolds number and the gas superficial velocity. We obtained two modes of the unstable disturbances and computed the wavelength and phase velocity of their neutral and the fastest growing disturbances varying the liquid and gas Reynolds number. The first mode corresponds to the Kapitza's waves at small values of the gas superficial velocity. The second mode of the unstable disturbances corresponds to the transition to a turbulent flow in the gas phase. We found that the co-current gas velocity destabilize the film flow at all values of the inclination angle and distance between the plates considered in the paper. The range of the wavelength of the unstable disturbances essentially increases with the gas velocity increasing. As a second step, we have performed the systematic study of nonlinear wave regimes. We found that the gas flow affects significantly the wave characteristics decreasing the amplitude and increasing the phase velocity. The complex multi-fold and multi-sheet surface, found on the plane of parameters (wavelength and the liquid Reynolds number) for the gravitational film flow, exists at all velocities of the co-current gas flow. We carried out investigation of the “optimal” waves for several folders and presented comparison with the regimes observed in experiments. Using of the strict equations without any additional assumptions is an important feature of this paper.

Partial cavitation over incipient, transitory and periodic regimes is investigated using large eddy simulation (LES) in the (experimental) sharp wedge configuration of Ganesh et al. (2016). The numerical approach is based on a compressible homogeneous mixture model with finite rate mass transfer between the phases. Physical mechanisms of cavity transition observed in the experiments; i.e. re-entrant jet and bubbly shock wave, are both captured in the LES over their respective regimes. Vapor volume fraction data obtained from the LES is quantitatively compared to X-ray densitometry. In the transitory and periodic regimes, void fractions resulting from complex interactions of large regions of vapor in the sheet/cloud show very good comparison with the experiments. In addition, very good agreement with the experiments is obtained for the shedding frequency and the bubbly shock wave propagation speed. In the incipient regime, the qualitative characteristics of the flow (e.g. cavitation inside spanwise vortices in the shear layer) are captured in the simulations. Conditions favoring either the formation of the re-entrant jet or the bubbly shock wave are analyzed by contrasting the LES results between the regimes. In the transitory regime, large pressure recovery from within the cavity to outside, and the resulting high adverse pressure gradient at the cavity closure support the formation of re-entrant jet. In the periodic regime, overall low pressures lead to reduced speed of sound and increased medium compressibility, favoring the propagation of shock waves. In a re-entrant jet cycle, vapor production occurs predominantly in the shear layer, and intermittently within the cavity. In a bubbly shock cycle, vapor production is observed spanning the entire thickness of the cavity. Bubbly shock wave propagation is observed to be initiated by the impingement of the collapse-induced pressure waves from the previously shed cloud. Supersonic Mach numbers are observed in the cavity closure regions, while the regions within the grown cavity are subsonic due to the negligible flow velocities.

In this investigation, the accuracy of the discrete and continuous random walk (DRW, CRW) stochastic models for simulation of fluid (material) point particle, as well as inertial and Brownian particles, was studied. The corresponding dispersion, concentration, and deposition of suspended micro- and nano-particles in turbulent flows were analyzed. First, the DRW model used in the ANSYS-Fluent commercial CFD code for generating instantaneous flow fluctuations in inhomogeneous turbulent flows was evaluated. For this purpose, turbulent flows in a channel were simulated using a Reynolds-averaged Navier-Stokes (RANS) approach in conjunction with the Reynolds Stress Transport turbulence model (RSTM). Then spherical particles with diameters in the range of 30 µm to 10 nm were introduced uniformly in the channel. Under the assumption of one-way coupling, ensembles of particle trajectories for different sizes were generated by solving the particle equation of motion, including the drag and Brownian forces. The DRW stochastic turbulence model of the software was used to include the effects of instantaneous velocity fluctuations on particle motion, and the steady state concentration distribution and deposition velocity of particles of various sizes were evaluated. In addition, the improved CRW model based on the normalized Langevin equation was used in an in-house Matlab code. Comparisons of the predicted results of the DRW model of ANSYS-Fluent with the available experimental data and the DNS simulation results and empirical predictions showed that this model is not able to accurately predict the flow fluctuations seen by the particles in that it leads to unreasonable concentration profiles and time-varying deposition velocities. However, the predictions of the improved CRW model were in good agreement with the experimental data and the DNS results. Possible reasons causing the discrepancies between the DRW predictions and the experimental data were discussed. The improved CRW model was also implemented through user-defined functions into the ANSYS-Fluent code, which resulted in accurate concentration distribution and deposition velocity for different size particles.

The present contribution falls in the scope of high-frequency combustion instabilities occurring in liquid rocket engines. Under subcritical operating conditions, numerical simulations have to render the effects of acoustic waves on the atomisation of liquid jets since these may impact the stability of the engine. Therefore, the present contribution aims at evaluating the ability of a particular numerical strategy adapted to the simulation of two-phase flows to render these interaction mechanisms. The selected strategy is based on the coupling between a diffuse interface method for the simulation of large liquid structures, and a kinetic-based Eulerian model for the description of droplets. The numerical simulation of an air-assisted liquid jet submitted to a transverse acoustic modulation is performed. The flattening of the liquid core under acoustic constraints is retrieved and induces an intensification of its stripping. In addition, thanks to appropriate coupling source terms, the modification of the spray shape as well as periodic oscillations of droplets are retrieved. The numerical strategy is thus proved to be adapted to deal with atomised liquid jets under transverse acoustic modulation and can be used for future numerical studies of high-frequency combustion instabilities.

A fully compressible two-phase flow model consisting of four balance equations including two mass, one momentum, and one internal energy equation, formulated with the mechanical and thermal equilibrium assumptions is developed in this article. This model is closed with a real fluid equation of state (EoS) and has been applied to the simulation of different 1D academic cases, in addition to the 3D Large-Eddy Simulation (LES) of the Engine Combustion Network (ECN) Spray A injector including the needle to target part with and without the phase change (i.e. frozen) assumptions. The obtained numerical results from the model with phase change have proven to be able to accurately predict the liquid, vapor penetrations and rate of injection compared to experimental data. However, the frozen model has presented some uncertainties and deviations in predicting the penetration length as with different measure criterions, even though an excellent agreement can be achieved in the estimation of rate of injection, near-nozzle mass and velocity distribution. Several conclusions are drawn from the simulations: (1) the initial in-nozzle flow has a strong effect on the early jet development; (2) considering phase change is still essential in the high temperature, high pressure (HTHP) injection modelling since it strongly affects the temperature distribution, turbulence intensity and thereby the jet development; (3) significant variations of liquid compressibility factor and density, as well as the cooling effect through the nozzle are highlighted. Overall, the detailed analysis of the numerical results reported in this article may complement the Engine Combustion Network (ECN) experimental database.

Experiments reported till today have shown that narrower channel exhibits stronger evaporation. To clarify how this behavior comes, we proposed a new model for evaporation of wetting liquid film in a narrow channel. With considering heat conduction in liquid film, kinetics and diffusion, we described a local heat and mass transfer in the microscopic liquid-film region which length scale is characterized by minimizing free energy. The innovative point of our model is that we did not assume thermodynamic equilibrium in micro and macroscopic meniscus regions, and introduced an excess pressure term balancing a conventional stress term at gas/liquid interface, which acts as a source term of evaporation. Although a free parameter must be included because local thermal equilibrium was not assumed, our phenomenological model makes clear that non-isothermal driving force causes stronger evaporation in narrower channel.

An investigation of void fraction measurement and prediction in a triangular-array rod bundle geometry was carried out under high pressure conditions (P=5–9 MPa and G=100–350 kg•m−2•s−1). The aims of this work were to expand the existing database and assess and modify the predictive models. This paper is the Part Ⅰ of the study, where the experimental work was presented to measure local chordal and area-average void fractions respectively by a comprehensive measuring system composed by optical probes and gamma-ray densitometry. Non-uniform distribution of void fraction was investigated by local-to-local and local-to-average comparisons of void fractions. The results indicated that non-uniform distribution is not only related to the geometry, but also affected by the transition of flow regimes in rod bundle. The effects of pressure and mass flux on void fractions were also analyzed in detail in consideration of variations of fluid properties and drift-flux parameters. Additionally, both void fraction and interface frequency were compared with the flow regime map in TRACE code. The variation of interface frequency indicated that the droplet entrainment would occur in annular flow at high mass flux. In the Part Ⅱ of this study, an assessment of existing models for void fraction prediction in rod bundles was carried out according to the present data and a newly-developed drift-flux model was proposed.

A thorough understanding of the dynamics of meter-sized airgun-bubbles is very crucial to seabed geophysical exploration. In this study, we use the boundary integral method to investigate the highly non-spherical airgun-bubble dynamics and its corresponding pressure wave emission. Moreover, a model is proposed to also consider the process of air release from the airgun port, which is found to be the most crucial factor to estimate the initial peak of the pressure wave. The numerical simulations show good agreement with experiments, in terms of non-spherical bubble shapes and pressure waves. Thereafter, the effects of the port opening time Topen, airgun firing depth, heat transfer, and gravity are numerically investigated. We find that a smaller Topen leads to a more violent air release that consequently causes stronger high-frequency pressure wave emissions; however, the low-frequency pressure waves are little affected. Additionally, the non-spherical bubble dynamics is highly dependent on the Froude number Fr. Starting from Fr=2, as Fr increases, the jet contains lower kinetic energy, resulting in a stronger energy focusing of the bubble collapse itself and thus a larger pressure peak during the bubble collapse phase. For Fr ≥ 7, the spherical bubble theory becomes an appropriate description of the airgun-bubble. The new findings of this study may provide a reference for practical operations and designing environmentally friendly airguns in the near future.

A homogeneous mixture cavitation model is proposed in this study. The bounded Rayleigh-Plesset equation is modified due to the homogeneous treatment and the compressibility of liquid is considered. The acoustic cavitation in stationary liquid is simulated to validate the improvements in proposed model: firstly, cavitation with up to 3 bubbles is simulated, the predicted bubble dynamics agree well with the predictions by the VOF method; then two experiments of shock-wave-induced cavitation are simulated and good agreements are obtained. More factors need to be considered to extend the application range of the proposed model in future.

Based on the experimental and numerical methods, parametric studies on annulus width and post thickness of a liquid-centered swirl coaxial (LCSC) Injector were performed under atmospheric conditions to evaluate their effects on the internal flow dynamics, self-pulsation characteristics and atomization qualities. Filtered water and dried air were employed as simulant mediums. 2-D unsteady numerical simulations based on the swirl axi-symmetric model were conducted. The instantaneous self-pulsated spray images were captured with a back-lighting photography technique. By means of the Dantec Phase Doppler Anemometry (PDA) system, the atomization qualities including Sauter mean diameter (SMD), velocity of droplets and mass flow rate have been characterized. Results showed that annulus width and post thickness of inner injector together with the recess length have significant influences on self-pulsation. Violent liquid sheet pulsations are caused by the tremendous dynamic pressures imposed from the blocked annular gas stream. Whether self-pulsation occurs or not and self-pulsation intensity essentially depend on the relative magnitude of liquid film angle and recess angle. Increasing the annulus width and post thickness leads to an accompanying decay in self-pulsation. Furthermore, it was found that the mass flow rate distribution of self-pulsated spray is double-peaked while that of stable spray is single-peaked. SMD and velocity distributions of all droplets are more susceptible to the injection conditions rather than the geometrical parameters.

Multiphase flow models are validated by comparison with a relatively good supply of high-quality laboratory data, and a relatively sparse supply of field data, which tends to have poorer quality. One of the principal challenges for multiphase flow models, in terms of uncertainty, is the difference in scale and some of the fluid properties between field and laboratory conditions. Therefore, the models may become unreliable when they are applied to conditions that are very different from those in the laboratory. IFE (Institute for Energy Technology) has recently developed and demonstrated scale-up rules for the most basic multiphase pipe flows. The objective of the work presented in this paper was to select appropriate data from our existing database and design new, scaled laboratory experiments, well-suited to demonstrate (or test) the scaling rules by comparing the results. The data include fluid properties, pipe configurations and flow rates. Besides the observed flow pattern, liquid holdup and pressure gradient are the two main parameters for comparison. IFE's CO2 Flow Loop with a test section inner diameter (ID) of 44 mm, operates for two-phase flows over a large range of pressures and temperatures on the equilibrium line of pure CO2. In order to verify scale-up principles, series of experiments were conducted according to the scaling rules to simulate similar conditions. The experiments were performed with gas-liquid two-phase CO2 for fully-developed, steady-state flow, in a horizontal or near-horizontal pipe. The flow regimes include stratified and annular flows. The experimental results showed that measurements of liquid holdup, and pressure gradient in the CO2 Flow Loop are in excellent agreement with appropriately scaled data from the larger-scale facilities. The results also confirm that the gas-to-liquid density ratio plays an important role. The experiments provide valuable data sets for verifying scaling laws, which are lacking in the literature.

In this work, we present a rigorous derivation of the volume-filtered viscous compressible Navier–Stokes equations for disperse two-phase flows. Compared to incompressible flows, many new unclosed terms appear. These terms are quantified via a posteriori filtering of two-dimensional direct simulations of shock-particle interactions. We demonstrate that the pseudo-turbulent kinetic energy (PTKE) systematically acts to reduce the local gas-phase pressure and consequently increase the local Mach number. Its magnitude varies with volume fraction and filter size, which can be characterized using a Knudsen number based on the filter size and inter-particle spacing. A transport equation for PTKE is derived and closure models are proposed to accurately capture its evolution. The resulting set of volume-filtered equations are implemented within a high-order Eulerian–Lagrangian framework. An interphase coupling strategy consistent with the volume filtered formulation is employed to ensure grid convergence. Finally PTKE obtained from the volume-filtered Eulerian–Lagrangian simulations are compared to a series of two- and three-dimensional direct simulations of shocks passing through stationary particles.

We experimentally investigate the deformation and breakup of droplets interacting with an oblique continuous air stream. A high-speed imaging system is employed to record the trajectories and topological changes of the droplets of different liquids. The droplet size, the orientation of the air nozzle to the horizontal and fluid properties (surface tension and viscosity) are varied to study different breakup modes. We found that droplet possessing initial momentum before entering the continuous air stream exhibits a variation in the required Weber number for the vibrational mode to the bag breakup transition with a change in the angle of the air stream. The critical Weber numbers (Wecr) for the bag-type breakup are obtained as a function of the Eötvös number (Eo), angle of inclination of the air stream (α) and the Ohnesorge number (Oh). It is found that although the droplet follows a rectilinear motion initially that transforms to a curvilinear motion at later times when the droplet undergoes topological changes. The apparent acceleration of the droplet and its size influence the critical Weber number for the bag breakup mode. The departure from the cross-flow arrangement shows a sharp decrease in the critical Weber number for the bag breakup, which asymptotically reaches a value associated with the in-line (opposed) flow configuration for the droplet breakup.

The purpose of this discussion is to pay the attention among International Journal of Multiphase Flow readers. Comments are presented on the study by Li and Hibiki [1] where the researchers presented correlation of frictional pressure drop for two-phase flows in micro and mini single-channels. In this discussion, the two-phase frictional pressure drop equation in homogenous flow model is corrected. In addition, a look on the two-phase mixture density (ρtp) equation is presented. From the current analysis, the newcomers to the field of two-phase flow should avoid this wrong definition of the two-phase density (ρtp).

Turbulent two-phase flows feature different mechanisms for production and dissipation of turbulent kinetic energy compared to the single-phase flows. However, this difference is usually neglected in developing eddy viscosity-based subgrid scale (SGS) models for the two-phase large eddy simulation (LES). In this study, a new approach is presented for the two-phase LES to include the surface tension, which is a production mechanism for the kinetic energy in the small scale motions, into the subgrid eddy viscosity model. We follow the Favre-filtered governing equations of interfacial flows based on the volume of fluid (VOF) approach and derive the transport equation for the turbulent kinetic energy to include the effect of surface tension. The original contribution of this study is to propose a new form for the eddy viscosity based on the mixing length assumption which includes an additional production mechanism of turbulent kinetic energy stemming from the interfacial work i.e. surface tension. The proposed model for eddy viscosity is employed to close all the SGS terms. The model performance is evaluated by means of the a-priori filtering of the fine grid simulation of phase inversion problem. To test the generality of the model at different physical conditions, two different density ratios were considered for the fine grid simulation. The results highlight a significant improvement of the eddy viscosity-based SGS models in prediction of the turbulent kinetic energy for the small unresolved scales particularly for the regions of low shear. Furthermore, the model appears to perform more accurately in the case of low density ratios. This study provides a proper perspective for future SGS models in the context of large eddy simulation of two-phase flows.

Swirling gas-liquid flow has a wide variety of engineering applications in petroleum, chemical, and nuclear industries. A fundamental understanding of swirling flow patterns and their transitions is essential for the proper modelling of pressure drop, mass transfer rate and interfacial area in swirling two-phase flow. Although the gas-liquid flows in straight pipes have been widely studied, little attention has been paid to characteristics at transition boundary of swirling gas-liquid flow. In this paper, based on high-speed camera method using back light source, flow patterns and transition criteria for swirling gas-liquid flow induced by a vane-type swirler were experimentally investigated in a vertical tube of 30 mm I.D., operating at ambient temperature and pressure. According to our observation, five typical flow patterns, i.e. swirling gas column flow, swirling intermittent flow, swirling annular flow, slug flow and churn flow, were identified and the detailed characteristics at the transition boundaries were discussed with flow-regime reconstruction technique. The spatial and temporal evolution of the gas core diameter in swirling flow regimes were also analyzed using image processing technique. Additionally, the flow-regime transition criteria were gained which reveal that the void fraction αg in swirling gas column flow is below 0.30∼0.32 and it attains values of 0.32 to 0.71 in the swirling intermittent flow. When αg is greater than 0.71∼0.74, the swirling intermittent flow transforms into swirling annular flow.