Numerical modeling and experimental investigation of gas–liquid slug formation in a microchannel T-junction

Gas–liquid two-phase flow in a microfluidic T-junction with nearly square microchannels of 113 μm hydraulic diameter was investigated experimentally and numerically. Air and water superficial velocities were 0.018–0.791 m/s and 0.042–0.757 m/s, respectively. Three-dimensional modeling was performed with computational fluid dynamics (CFD) software FLUENT and the volume of fluid (VOF) model. Slug flow (snapping/breaking/jetting) and stratified flow were observed experimentally. Numerically predicted void fraction followed a linear relationship with the homogeneous void fraction, while experimental values depended on the superficial velocity ratio U_G/U_L. Higher experimental velocity slip caused by gas inlet pressure build-up and oscillation caused deviation from numerical predictions. Velocity slip was found to depend on the cross-sectional area coverage of the gas slug, the formation of a liquid film and the presence of liquid at the channel corners. Numerical modeling was found to require improvement to treat the contact angle and contact line slip, and could benefit from the use of a dynamic boundary condition to simulate the compressible gas phase inlet reservoir.     

Droplet size and velocity characteristics of water-air impinging jet atomizer

A water-air impinging jets atomizer is investigated in this study, which consists of flow visualization using high speed photography and mean droplet size and velocity distribution measurements of the spray using Phase Doppler Anemometry (PDA). Topological structures and break up details of the generated spray in the far and near fields are presented with and without air jet and for an impinging angle of 90°. Spray angle increases with the water jet velocity, air flow rate and impinging angle. PDA results indicate that droplet size is smallest in the spray center, with minimum value of Sauter mean diameter (SMD) of 50 µm at the air flow rate of Q_m = 13.50 g/min. SMD of droplets increases towards the spray outer region gradually to about 120 µm. The mean droplet velocity component W along the air-jet axis is highest in the spray center and decreases gradually with increasing distance from the spray center. SMD normalized by the air nozzle diameter is found firstly to decrease with gas-to-liquid mass ratio (GLR) and air-to-liquid momentum ratio (ALMR) and then remain almost constant. Its increasing with aerodynamic Weber number indicates an exponential variation. The study sheds light on the performance of water-air impinging jets atomizers providing useful information for future CFD simulation works.     

Study of material agglomeration during nozzle injection of multiviscosity liquid into a fluidized bed reactor by the electric conductance method

Fluidized bed agglomeration is an important and challenging problem for thermal cracking in fluid cokers. A low coker temperature can be problematic because the bitumen is injected into the fluidized bed with a different viscosity, resulting in formation of agglomerates of varying sizes, which slows the cracking reactions. In the present study, the bed material agglomeration process during nozzle injection of multiviscosity liquid was investigated in a fluidized bed operated at different mass ratios of the atomization gas to the liquid jets (GLR = 1%–3.5%) and gas velocities (3.9U_mf and 5.9U_mf) based on a conductance method using a water–sand system to simulate the hot bitumen–coke system at room temperature. During the tests of liquid-jet dispersion throughout the bed, different agglomeration stages are observed at both gas velocities. The critical amount of tert-butanol in the liquid jets that could lead to severe agglomeration of the bed materials (poor fluidization) at GLR = 1% is about 10 wt% at the low fluidizing gas velocity (3.9U_mf) and 18 wt% at the high gas velocity (5.9U_mf). This study provides a new approach for on-line monitoring of bed agglomeration during liquid injection to guarantee perfect contact between the atomized liquid and the bed particles.     

Effect of Gaussian size distribution width on minimum fluidization velocity in tapered gas–solid fluidized beds

This paper investigated the effect of Gaussian distribution width, average particle diameter, particle loading, and the tapered angle on minimum fluidization velocity (U_mf) by conducting extensive experiments in tapered fluidized beds. Three powders with Gaussian size distribution and different distribution widths were used in the experiments. An increase in U_mf with increasing the average particle diameter, particle loading, and the tapered angle was observed. There was also a nonmonotonic behavior of U_mf as the Gaussian distribution width increased. An empirical correlation including dimensionless groups for predicting U_mf in tapered beds was developed in which the effect of distribution width was considered. The proposed correlation predictions were in good agreement with the experimental data, with a maximum deviation of 16.5% and average and standard deviations of, respectively, 6.4% and 7.4%. The proposed correlation was also compared with three earlier models, and their accuracy was discussed.     

Steady rise of a small spherical gas bubble along the axis of a cylindrical pipe at high Reynolds number

Steady irrotational flow of inviscid liquid of density ρ_l around a spherical gas bubble which lies on the axis of a cylindrical pipe is investigated using the analysis of Smythe (Phys. Fluids 4 (1961) 756). The bubble radius b=qa is assumed small compared to the pipe radius a, and the interfacial tension between gas and liquid is γ. Far from the bubble, in the frame in which the bubble is at rest, the liquid velocity along the pipe is v₀, whereas the liquid velocity at points on the wall closest to the bubble is U_zw=v₀(1+1.776q³+⋯). The decrease in wall pressure as the bubble passes is therefore Δp=1.776ρ_lv₀²q³. When the Weber number W=2bv₀²ρ_l/γ is small, the bubble deforms into an oblate spheroid with aspect ratio χ=1+9W(1+1.59q³)/64. If the fluid viscosity μ is non-zero, and the Reynolds number Re=2v₀ρ_lb/μ is large, a viscous boundary layer develops on the walls of the pipe. This decays algebraically with distance downstream of the bubble, and an exponentially decaying similarity solution is found upstream. The drag D on the bubble is D=12πμv₀b(1−2.21Re^−1/2)(1+1.59q³)+7.66μv₀bRe^1/2q^9/2, larger than that given by Moore (J. Fluid Mech. 16 (1963) 161) for motion in unbounded fluid. At high Reynolds numbers the dissipation within the viscous boundary layers might dominate dissipation in the potential flow away from the pipe walls, but such high Reynolds numbers would not be achieved by a spherical air bubble rising in clean water under terrestrial gravity.     

Hydrodynamic behaviour of Cellets™ (Ph.Eur./USP) in a spouted bed using image processing method

The hydrodynamic behaviour of the spouted bed in the pharmaceutical industries has been found to be less addressed. The present paper has focused on the hydrodynamic characteristics of a spouted bed where the Cellets™ (Ph.Eur./USP) is adopted as the bed material. Experiments are carried out with three different static bed heights (H₀) of shallow depth (2D_i ≤ H₀ < 3D_i) using two different particle sizes. The spouted bed employed with D_i/D₀ of 5 has given the experimental information on external spouting (U_es) by mapping the pressure drop, and fountain height (H_f) against the superficial gas velocity (U_g) is represented with the image contours, which show the intrinsic behaviour. All the 1000 μm and 700 μm particles have been found to exhibit symmetric and asymmetric spouting. With increasing U_g, the fully suspended particles are limited to a certain height in the freeboard region due to the gas-solid cross-flow, which implies the clusters have identified with the image processing method.     

Investigation of nanofluid flow in a vertical channel considering polynomial boundary conditions by Akbari-Ganji's method

In this research, a vertical channel containing a laminar and fully developed nanofluid flow is investigated. The channel surface's boundary conditions for temperature and volume fraction functions are considered qth-order polynomials. The equations related to this problem have been extracted and then solved by the AGM and validated through the Runge-Kutta numerical method and another similar study. In the study, the effect of parameters, including Grashof number, Brownian motion parameter, etc., on the motion, velocity, temperature, and volume fraction of nanofluids have been analyzed. The results demonstrate that increasing the Gr number by 100% will increase the velocity profile function by 78% and decrease the temperature and fraction profiles by 20.87% and 120.75%. Moreover, rising the Brownian motion parameter in five different sizes (0.1, 0.2, 0.3, 0.4, and 0.5) causes lesser velocity, about 24.3% at first and 4.35% at the last level, and a maximum 52.86% increase for temperature and a 24.32% rise for Ψ occurs when N_b rises from 0.1 to 0.2. For all N_t values, at least 55.44%, 18.69%, for F(η), and Ω(η), and 20.23% rise for Ψ(η) function is observed. Furthermore, enlarging the N_r parameter from 0.25 to 0.1 leads F(η) to rise by 199.7%, fluid dimensionless temperature, and dimensional volume fraction to decrease by 18% and 92.3%. In the end, a greater value of q means a more powerful energy source, amplifying all velocity, temperature, and volume fraction functions. The main novelty of this research is the combined convection qth-order polynomials boundary condition applied to the channel walls. Moreover, The AMG semi-analytical method is used as a novel method to solve the governing equations.     

Yield stress and flow behavior of concentrated ferrofluid-based magnetorheological fluids: the influence of composition

In this paper, the magnetorheological (MR) and magnetoviscous properties of ferrofluid-based iron particle suspensions were investigated. The 2.1-µm mean size Fe particles were dispersed in high-concentration transformer oil-based ferrofluid, the iron particle volume fraction in the resulting nano-micro composite magnetorheological fluid samples varying from Φ _Fe = 5 to 40 %. The ferrofluid carrier has φ _p = 23 % solid volume fraction of magnetic nanoparticles stabilized with chemisorbed oleic acid monolayer and without any excess surfactant. In the absence of the field, the ferrofluid has a quasi-Newtonian behavior with a weak shear thinning tendency. The static yield stress shows an increase of about 3 orders of magnitude for an iron particle content of approx. Φ _Fe = 25 % (Φ _tot = 42.25 %), while above this value, a saturation tendency is observed. The dynamic yield stress (Bingham model) also increases with the magnetic induction and the particle volume fraction; however, the saturation of the MR effect is less pronounced. The relative viscosity change has a maximum at Φ _Fe = (10–15) % due to the accelerated increase of the effective viscosity of the composite for higher Fe content. Addition of micrometer-sized iron particles to a concentrated ferrofluid without any supplementary stabilizing agent proved to be a direct and simple way to control the magnetorheological and magnetoviscous behavior, as well as the saturation magnetization of the resulting nano-micro composite fluid to fulfill the requirements of their use in various MR control and rotating seal devices.     

Co-current flow effects on a rising Taylor bubble

The effects of co-current flows on a rising Taylor bubble are systematically investigated by a front tracking method coupled with a finite difference scheme based on a projection approach. Both the upward (the co-current flows the same direction as the buoyancy force) and the downward (the co-current moves in the opposite direction of the buoyancy force) co-currents are examined. It is found that the upward co-current tends to elongate the bubble, while the downward co-current makes the bubble fatter and shorter. For large N_f (the inverse viscosity number), the upward co-current also elongates the skirted tail and makes the tail oscillate, while the downward co-current shortens the tail and even changes a dimpled bottom to a round shape. The upward co-current promotes the separation at the tail, while the downward co-current suppresses the separation. The terminal velocity of the Taylor bubble rising in a moving flow is a linear combination of the mean velocity (U_C) of the co-current and the terminal velocity (U₀) of the bubble rising in the stagnant liquid, and the constant is around 2 which agrees with the literature. The wake length is linearly proportional to the velocity ratio (U_C/U₀). The co-currents affect the distribution of the wall shear stresses near the bubble, but not the maximum.     

Effect of the shape of a body on the intermittency of turbulent flow in an axisymmetrical wake

Experimental data regarding the distribution of the intermittency coefficient of the turbulent flow of an incompressible liquid in the axisymtnetrical wakes behind a sphere or a solid of revolution with an elongation of 8∶1 are presented. The original measurements were carried out at a Reynolds number of Re=U_∞D/v=10⁴ (U_∞) is the velocity of the incident flow, D is the diameter of the middle cross section). The shape of the solid is shown to have a considerable effect on the form of intermittency in the far automodel wake.     

Experimental research on the thermal performance of converging slot holes with different divergence angles

Thermal performances of two kinds of converging slot-hole (console) with different divergence angles have been measured using transient liquid crystal measurement technique which can process the nonuniform initial wall temperature. Four momentum ratios are tested. Consoles with different divergence angles produce different cooling effectiveness distributions in the upstream region. However, the cooling effectiveness distributions of the two consoles are similar in the downstream. The laterally averaged cooling effectiveness results show that the differences between the two consoles are very small and the best momentum ratio for both consoles’ cooling effectiveness distribution is around two. With the momentum ratio increasing, the normalized heat transfer coefficient h/h₀ of both consoles increases, but the h/h₀ value of small divergence case is higher and becomes progressively higher than that of large divergence case. Moreover, the effect of the couple vortices on the heat transfer coefficient distributions is more significant for the large divergence case. Both consoles provide the surface a certain degree of thermal protection, especially in the upstream region. The distributions of heat flux ratio q/q₀ are similar with those of η because the influence of η on q/q₀ is much larger than that of h/h₀ on q/q₀.     

Bubble formation process from a novel nozzle design in liquid cross-flow

An experimental study is reported that investigated the bubble formation from a novel nozzle design in a liquid cross-flow using high speed imaging. Different configurations and orientations of the novel nozzle design were considered over a range of gas-to-liquid flow rate ratios (GLR) from 0.00031 to 0.00204. The results show that for all cases, the novel nozzle generated smaller bubbles at higher detachment frequency compared to the standard nozzle. At low liquid velocities, the novel nozzle generated bubbles that were 30% smaller in size at a detachment frequency 2–3 times higher than that for the standard nozzle. It was also found that the bubble diameter and the detachment frequency are almost independent of the liquid velocity. The underlying physical process of bubble formation and detachment in the novel nozzle under liquid cross-flow was also investigated. It was observed that the process comprised of three phases: expansion, collapse and pinch off. It was also found that the rebound force of the bubble from a side-hole under the influence of liquid drag force and hydrostatic pressure plays a key role in the early bubble detachment. The results demonstrated that the novel nozzle design performs better than the standard nozzle in the liquid cross-flow, especially at high GLRs.     

Wavy structures in compressible mixing layers

Semi-periodic structures namely inclined wavy structures(IWS) are experimentally observed in compressible mixing layers at two convective Mach numbers(Mc=0.11 and 0.47). Flow structures are visualized by the laserinduced planar laser Mie scattering(PLMS) technique. Two methods are developed to investigate the spatial distribution and geometry of IWS:(1) the dominant mode extraction(DME) method, to extract the dominant modes of IWS from the streamwise gray-level fluctuation, and(2) the phase tracking(PT) method, to identify the shape of IWS. The results suggest that pressure perturbations account for the formation of IWS in the initial mixing region and the joint efect of dilatation and coherent vortices enhances IWS in the welldeveloped region. The large transverse(cross-flow) scale of the IWS and their relation to coherent vortices(CV) indicate that the disturbance originated from CV in the mixing center propagates far into the free streams. The DME and the PT method are shown to be the efective tools to study the geometrical features of wavy structures in compressible shear flows.     

Neighborhood Relationships of Widely Distributed and Irregularly Shaped Particles in Partially Dewatered Filter Cakes

A more thorough understanding of the properties of bulk material structures in solid–liquid separation processes is essential to understand better and optimize industrially established processes, such as cake filtration, whose process outcome is mainly dependent on the properties of the bulk material structure. Here, changes of bulk properties like porosity and permeability can originate from local variations in particle size, especially for non-spherical particles. In this study, we mix self-similar fractions of crushed, irregularly shaped Al₂O₃ particles (20 to 90 µm and 55 to 300 µm) to bimodal distributions. These mixtures vary in volume fraction of fines (0, 20, 30, 40, 50, 60 and 100 vol.%). The self-similarity of both systems serves the improved parameter correlation in the case of multimodal distributed particle systems. We use nondestructive 3D X-ray microscopy to capture the filter cake microstructure directly after mechanical dewatering, whereby we give particular attention to packing structure and particle–particle relationships (porosity, coordination number, particle size and corresponding hydraulic isolated liquid areas). Our results reveal widely varying distributions of local porosity and particle contact points. An average coordination number (here 5.84 to 6.04) is no longer a sufficient measure to describe the significant bulk porosity variation (in our case, 40 and 49%). Therefore, the explanation of the correlation is provided on a discrete particle level. While individual particles?<?90 µm had only two or three contacts, others?>?100 µm took up to 25. Due to this higher local coordination number, the liquid load of corresponding particles (liquid volume/particle volume) after mechanical dewatering increases from 0.48 to 1.47.

     

Operating range of magnetic stabilization flow regime for magnetized fluidized bed with Geldart-B magnetizable and nonmagnetizable particles

The magnetic stabilization flow regime could also be created for Geldart-B nonmagnetizable particles provided some magnetizable particles are introduced and the magnetic field is applied. This study aimed to explore the size (d_pM) and density (Ͽ_pM) effects of magnetizable particles on its operating range. The upper limit (U_mbH) could not be determined from the οP_b⿿U_g⿿ curve but could from analyzing the variation of οP_b-fluctuation with increasing U_g. Due to the variation of U_mfH (lower limit) with d_pM and Ͽ_pM, both U_mbH⿿U_mfH and (U_mbH⿿U_mfH)/U_mfH were used to quantify the operating range of magnetic stabilization. U_mbH⿿U_mfH varied hardly with d_pM but increased significantly with decreasing Ͽ_pM. (U_mbH⿿U_mfH)/U_mfH increased as d_pM or Ͽ_pM decreased. It was more difficult for the nonmagnetizable particles to escape from the network formed by the smaller/lighter magnetizable particles. For the same magnitude of change, d_pM had a stronger effect than Ͽ_pM on (U_mbH⿿U_mfH)/U_mfH. Neither U_mbH⿿U_mfH nor (U_mbH⿿U_mfH)/U_mfH varied monotonously with the minimum fluidization velocity of the magnetizable particles, indicating that no straightforward criterion for matching the magnetizable particles to the given nonmagnetizable particles could be established based on their minimum fluidization velocities to maximize the operating range of magnetic stabilization.     

Hydrodynamics of two-phase flow in vertical up and down-flow across a horizontal tube bundle

Flow patterns, void fraction and friction pressure drop measurements were made for an adiabatic, vertical up-and-down, two-phase flow of air–water mixtures across a horizontal in-line, 5×20 tube bundle with a pitch-to-diameter ratio of 1.28. The flow patterns in the cross-flow zones were obtained and flow pattern maps were constructed. The data of average void fraction were less than the values predicted by a homogenous flow model and showed a strong mass velocity effect, but were well-correlated in terms of the Martinelli parameter X_tt and liquid-only Froude number Fr_LO. The two-phase friction multiplier data could be well-correlated with the Martinelli parameter.     

Kinetic modeling of particles in stratified flow – Evaluation of dispersion tensors in inhomogeneous turbulence

The continuum equations for a dilute particle distribution in inhomogeneous turbulence are tested against results from a Langevin particle tracking simulation. Reeks’ version of the kinetic theory is used to generate the mass, momentum and kinetic stress equations for the particle distribution. The particle tracking data are used to directly evaluate the dispersion tensors λ and μ which serve as closure relations for the continuum equations. These exact forms are compared to approximate, local forms. Even for low Stokes numbers (corresponding to low particle inertia defined by τ/τ_p ? 1), the tensor λ is strongly affected by the inhomogeneity and depends on turbulence parameters in the volume corresponding to the particle path dispersion over the particle Lagrangian integral timescale τ. In contrast, the locally homogeneous form of the velocity dispersion tensor μ is a sufficient approximation, since it depends on the dispersion volume over the much smaller particle relaxation time τ_p. It is demonstrated that the body force due to the dispersion vector γ cannot be neglected. In the limit of passive tracers (zero stopping distance), γ is equal to the gradient of λ, if the physical setting is such that we can invoke constant tracer density in this limit.     

Analyses of tensile deformation of nanocrystalline α-Fe2O3+fcc-Al composites using molecular dynamics simulations

The tensile deformation of nanocrystalline α-Fe₂O₃+fcc-Al composites at room temperature is analyzed using molecular dynamics (MD) simulations. The analyses focus on the effects of variations in grain size and phase volume fraction on strength. For comparison purposes, nanostructures of different phase volume fractions at each grain size are given the same grain morphologies and the same grain orientation distribution. Calculations show that the effects of the fraction of grain boundary (GB) atoms and the electrostatic forces between atoms on deformation are strongly correlated with the volume fractions of the Al and Fe₂O₃ phases. In the case of nanocrystalline Al where electrostatic forces are absent, dislocation emission initiates primarily from high-angle GBs. For the composites, dislocations emits from both low-angle and high-angle GBs due to the electrostatic effect of Al-Fe₂O₃ interfaces. The effect of the interfaces is stronger in structures with smaller average grain sizes primarily because of the higher fractions of atoms in interfaces at smaller grain sizes. At all grain sizes, the strength of the composite lies between those of the corresponding nanocrystalline Al and Fe₂O₃ structures. Inverse Hall-Petch (H-P) relations are observed for all structures analyzed due to the fact that GB sliding is the dominant deformation mechanism. The slopes of the inverse H-P relations are strongly influenced by the fraction of GB atoms, atoms associated with defects, and the volume fractions of the Al and Fe₂O₃ phases.     

Atomization of glycerin with a twin-fluid swirl nozzle

Atomization of liquids with high viscosity is always a challenge, especially when small diameter droplets and high liquid flow rates are simultaneously required. In the present research, the performance of a Venturi–vortex twin-fluid swirl nozzle is examined, attending to its capabilities to generate droplets with diameters below 20 µm when atomizing pure glycerin at room temperature. In this nozzle, air is injected tangentially in a central convergent section, and discharges suctioning the liquid fed to a coaxial chamber, here using a gear pump. The resulting spray is visualized and analyzed. Droplet size distributions are measured with a laser diffractometer. As expected, droplet diameter increases with liquid flow rate, and quickly diminishes when air flow rate is increased. Sauter mean diameters (SMD) below 15 µm can be obtained even when atomizing pure glycerin. However, these values are obtained for relatively low glycerin flow rates (∼5 l/h), and with rather wide distributions. For 10 l/h and an air-to-liquid mass flow rate ratio (ALR) of 13.7 more than 26% of the glycerin volume is atomized in droplets smaller than 20 µm. Liquid ligaments are observed near the nozzle exit, but they tend to break up while moving downstream.     

Cross-flow past a pair of nearly in-line cylinders with the upstream cylinder subjected to a transverse harmonic oscillation

An experimental investigation is presented for the cross-flow past a pair of staggered circular cylinders, with the upstream cylinder subject to forced harmonic oscillation transverse to the flow direction. Experiments were conducted in a water tunnel with Reynolds numbers, based on upstream velocity, U, and cylinder diameter, D, in the range 1440⩽Re⩽1680. The longitudinal separation between cylinder centres is L/D=2.0, with a transverse separation (for the mean position of the upstream cylinder) of T/D=0.17; the magnitude of the harmonic oscillation is 0.44D peak-to-peak and the nondimensional frequency range of the excitation is 0.05⩽f_eD/U⩽0.44. Flow visualization of the wake-formation region and hot-film measurements of the wake spectra are used to investigate the wake-formation process. An earlier study showed that stationary cylinders in this nearly in-line configuration straddle two very different flow regimes, the so-called shear-layer reattachment (SLR) and induced separation (IS) regimes. The present study, demonstrates that oscillation of the upstream cylinder causes considerable modification of the flow patterns around the cylinders. In particular, the wake experiences strong periodicities at the frequency of the oscillating cylinder; in addition to the usual fundamental lock-in, both sub- and superharmonic resonances are obtained. It is also observed that, although the flow exhibits regions of SLR and IS for excitation frequencies below the fundamental lock-in, for frequencies above the lock-in range the flow no longer resembles either of these flow regimes and vortices are formed in the gap between the cylinders.