T-型微通道中液滴形成机制的CFD模拟

对微米级窄型T-型微通道中微液滴的形成机制进行了CFD模拟,验证了随毛细准数Ca的增加,液滴的形成会经历"squeezing"和"dripping"机制,且2个机制之间明显的存在着一个"transient"机制。通道壁的润湿性能对液滴的形成过程有显著影响,只有当通道壁更亲连续相时,微液滴才能形成。但与"dripping"机制不同,在"squee-zing"机制下,通道壁的润湿性对形成液滴的体积有明显的影响。     

Dynamics of bubble breakup in a microfluidic T-junction divergence

The aim of this work is to investigate experimentally the bubble breakup in a microfluidic T-junction divergence using a high-speed digital camera and a micro-Particle Image Velocimetry (micro-PIV) system. The breakup and non-breakup of N₂ bubbles in glycerol–water mixtures with several concentrations of sodium dodecyl sulphate (SDS) as surfactant were studied with capillary number ranging from 0.001 to 0.1. The cross section of PMMA square microchannel is 400 μm wide and 400 μm deep. Four various flow patterns were observed at the T-junction by changing gas and liquid flow rates. The dynamics of three various types of symmetric breakup of bubbles were investigated. The symmetric breakup of bubbles type I is mainly controlled by the augmented pressure in liquid phase. The symmetric breakup of bubbles type II is controlled by both the increased pressure and viscous forces. In the symmetric breakup of bubbles type III, a scaling law for the minimum bubble neck and the remaining time during bubble breaking process were found. The transitions between breakup and non-breakup of bubbles were investigated, and a power–law relationship between bubble extension and capillary number was proposed to predict the transitions between adjacent regimes. Our experimental results reveal that the bubble breakup in a microfluidic T-junction divergence is similar to the droplet behaviours in such a device ( [Jullien et al., 2009] , [Leshansky and Pismen, 2009] and [Link et al., 2004] ).     

Film and slug behaviour in intermittent slug–annular microchannel flows

Whilst there are numerous experimental, theoretical and computational studies of Taylor flow in microchannels, the intermittent slug–annular regime has largely been neglected. In this paper time-resolved micro-PIV data are collected and used to study the flow characteristics of a gas–liquid system for flow regimes spanning Taylor to annular flow. The experimental work used a 1.73 mm diameter channel with water and nitrogen as the working fluids, for gas and liquid superficial velocity ranges of 0.35–8.65 m s^?1 (40<Re_G<1000) and 0.071–0.18 m s^?1 (120<Re_L<300), respectively. Time-averaged velocity profiles were obtained in the liquid film surrounding the gas bubbles (or the gas core in the pseudo-annular flow regime) and in the liquid slugs (which changed from regular slugs to annular rings as the gas superficial velocity was increased). These data showed that the velocity in the liquid film relaxed back to an equilibrium value following the passage of each liquid slug or annular ring. In contrast rather flat velocity profiles were observed in the liquid slug. Based on a simple representation of the flow structure, average gas holdups were estimated using independent experimental data obtained by the micro-PIV technique and by direct observation of the flow structure. A phenomenological model of intermittent slug flow, based on the representation of the flow structure as a train of slugs and bubbles moving over a liquid film, is used to interpret the experimental data. The modelling work highlights the different behaviour of the limiting cases of slug and annular flow, in terms of the gas–liquid interfacial shear and its influence on the pressure field.     

An experimental investigation of gas-liquid two-phase flow in single microchannel contactors

This paper describes two-phase flow pattern and pressure drop characteristics during the absorption of CO₂ into water in three horizontal microchannel contactors which consist of Y-type rectangular microchannels having hydraulic diameters of 667, 400 and , respectively. With the help of a high-speed photography system, flow patterns such as bubbly flow, slug flow (including two sub-regimes, Taylor flow and unstable slug flow), slug-annular flow, churn flow and annular flow were observed in these microchannels. The applicability of the currently available correlations for describing flow pattern transitions in microchannels has been examined. Generally, the predicting performance of these correlations deteriorates as the channel diameter further reduces. Toward solving this discrepancy, an empirical correlation based on the superficial Weber numbers was developed to interpret the transition from Taylor flow to unstable slug flow in three microchannels. Taylor bubble formation process in microchannels was found to be in the squeezing regime at lower superficial liquid velocities (Ca ranging from 0.0019 to 0.029) while the transition to the dripping regime was observed at the highest superficial liquid velocity of 1.0 m/s. Lengths of Taylor bubbles formed in the squeezing regime can be well represented by the scaling relation proposed by Garstecki et al. [Formation of droplets and bubbles in a microfluidic T-junction—scaling and mechanism of break-up. Lab on a Chip, 6, 437-446]. For flow patterns including slug-annular flow, annular flow and churn flow, a simple analysis based on the separated flow model has been performed in order to reveal the observed effect of the superficial liquid velocity on two-phase frictional multiplier in the present microchannels. Then, reasonable correlations for the prediction of two-phase frictional pressure drop under these flow patterns were suggested.     

Preparation and characterization of lotus ceramics with different pore sizes and their implication for the generation of microbubbles for CO2 sequestration applications

Four kinds of porous mullite ceramics, named lotus ceramics because of the similarity of their microstructure with lotus roots, were prepared by an extrusion method using rayon fibers of four different diameters (8.1, 9.6, 16.8 and 37.6 μm) as the pore formers. The physicochemical properties of these samples were characterized to test their applicability for the generation of microbubbles. The lotus ceramic samples contained pores of 9.4, 10, 15.6 and 30 μm size and porosities of 45–48%. SEM micrographs confirmed that the cylindrical pores were oriented unidirectionally along the extrusion direction and the degree of alignment was greater with larger fiber diameter. The permeability for gaseous CO₂ increased with increasing pore size from 3×10^?13 to 8×10^?13 m². The four lotus ceramic samples, a commercial air stone (72 μm) and two simple tubes (1000 and 3500 μm) were used to generate microbubbles in water under ambient conditions from a gas mixture of CO₂ and air. It was found that the bubble size could be decreased with bubblers of smaller pore size. In the bubble size measurements for pure CO₂ and air, the air bubbles were larger than the CO₂ bubbles due to partial dissolution of CO₂ into the water during bubbling. In order to generate smaller size bubbles using porous ceramic bubblers, the liquid must penetrate through the pores of the lotus ceramics before the gas is introduced into the system.     

Preparation of monodisperse PEG hydrogel microparticles using a microfluidic flow-focusing device

A microfluidic assisted preparation method of nearly monosized poly(ethylene glycol) (PEG) microparticles has been described. Three types of microfluidic flow-focusing devices with different geometries were fabricated using polydimethylsiloxane (PDMS). Microdroplets of PEG hydrogel were successfully prepared in the microfluidic flow-focusing devices by adjusting the flow rates of the continuous phase, namely, mineral oil, and the dispersed phase, viz., hydrogel solution. Then, the microdroplets of PEG hydrogel were cured by UV irradiation. Various experimental conditions pertaining to the geometry of the microfluidic flow-focusing device, flow rates of the dispersed and continuous phases, and concentration of PEG hydrogel solution were investigated and optimized to fabricate monosized PEG hydrogel microparticles. The prepared PEG microparticles were nearly monosized in the range of 40 μm to 200 μm in diameter according to the above experimental conditions. Then, PEG hydrogel particles laden with microbeads of 6 μm diameter were fabricated using the microfluidic flow-focusing devices with the optimized conditions.     

Gas–liquid two-phase flow division at a micro-T-junction

The phase distribution of a gas–liquid flow through a 1 mm T junction has been studied. Gas superficial velocities of 2.5 and 4.9 m/s and liquid superficial velocities 0.09–0.42 m/s were investigated. Increasing the liquid superficial velocity was shown to decrease the liquid taken off at the side arm. Increasing the gas superficial velocity was found to affect the phase split by increasing the fractional liquid taken off. It was noticed that pressure has no influence in the phase split when it was increased from 0.13 to 0.18 MPa. From examination of data from different pipe sizes, it was seen that the 1 mm T-junction shared similar split characteristics as those observed for larger diameter junctions. Finally, the gas–liquid flow pattern through the junction was observed to be slug for a range of gas and liquid superficial velocities.     

Heat transfer in well-characterised Taylor flow

The flow and heat transfer behaviours of gas–liquid, non-boiling, Taylor flow in the vertical upward direction were studied experimentally using a 2.00 mm diameter channel. Nitrogen and water at atmospheric pressure were employed as the working fluids. Three circular T-junction mixers with different diameters were used to generate gas bubbles and liquid slugs of different lengths (1–220d) with controlled mixture velocities (0.11<U_TP<0.53 m s^?1, 200<Re_TP<1100) and homogeneous void fractions (0.03<β<0.90). High-speed visualization of adiabatic flow and heat transfer rate determination for constant wall heat flux conditions were performed. The heat transfer enhancement brought about by Taylor flow is found to be larger with shorter slugs and higher mixture velocities. An enhancement up to 3.2-fold over the liquid-only flow was observed. Based on the experimental data, a correlation between the apparent slug Nusselt number (Nu_L?) with a Graetz number, where the characteristic length is that of the slug, is proposed.     

Characteristics of microbubbles generated by porous mullite ceramics prepared by an extrusion method using organic fibers as the pore former

Porous mullite ceramics with unidirectionally oriented pores were prepared by an extrusion method using rayon fibers as the pore formers and the characteristics of microbubbles generated by these porous ceramics were investigated. The 1200 mm long ceramics were tubular and of thick or thin types of 20–30 mm inner diameter and 30–50 mm outer diameter, respectively. The thin and thick samples had porosities of 47 and 49% and average pore radii of 7.8 μm. The gas permeabilities of the thick and thin samples were 4.1 × 10^?14 and 5.4 × 10^?14 m², respectively. Microbubbles were generated by introducing N₂ gas through the ceramic tube by immersing it into water. The minimum pressure (bubble point pressure) for generation of microbubbles was 20 kPa, much lower than for other bubble-forming methods. The average microbubble radii ranged from about 70 to 105 μm at flow rates of 0.15–0.25 L/min in the thin sample and 0.3–0.7 L/min in the thick sample. These bubble sizes are much smaller than calculated for a Fritz-type bubble such as generally formed by bubbling from pores and/or orifices. However, the present bubble sizes agree well with the calculated value based on nanobubbles, indicating that bubble formation occurs by mixing the gas with water in small pores. Since microbubbles enhance the dissolution rate of a gas phase in water, they are potentially useful for improving water environments, especially oxygen-deficient water. The effectiveness of gas dissolution in water was confirmed by determining the dissolution behavior of CO₂ gas using these porous ceramics.     

Intensification of bubble column performance by introduction pulsation of liquid

The application of pulsations at determined parameters to a bubble column can significantly modify the process of mass transfer. However, determination of these parameters is so difficult that it has not been done so far in a satisfactory way even for technical purposes. Hydrodynamics of pulsed-bubble columns has been a subject of a few research works published so far. This paper presents results of studies on flow hydrodynamics of gas bubbles in a glass pulsed-bubble column of square cross section with side D = 0.14 m and height H = 2.25 m, filled with water. Experiments were carried out at air superficial velocity UG ranging from 1.6 to 13.9 mm/s. Operating parameters of the pulsed-bubble column were measured in the frequency range f from 0 to 100 Hz and the amplitude of vibration exciter Xp from 0.25 to 2 mm. On the basis of results obtained so far it was shown that the application of pulsations at determined parameters to a bubble column caused a significant increase in the gas holdup. Based on the analysis of results of the measurements, model equations were developed to determine the Sauter diameters of bubbles and gas holdup during the pulsed-bubble column operation.     

The effect of sparger geometry on gas holdup and regime transition points in a bubble column equipped with perforated plate spargers

This paper investigates the effect of sparger geometry on flow regime of a bubble column. The experiments presented in this study were performed under atmospheric pressure with water/air in a cylindrical Plexiglas^® column of 33.0 cm i.d. and 3.0 m height. Three different perforated plate spargers were employed. Hole diameter was varied in the range of 1–3 mm, while the free area was 1.0%.The theory of linear stability is used for the prediction of regime transitions in the bubble column and a comparison has been presented between the predictions and the experimental observations. A good agreement between the predictions and the experimental values of transition gas holdup has been obtained.In addition, the data from the literature has been analyzed. Experimental values of transition gas holdups and predictions by the theory of linear stability have been compared with those of literature.A correlation based on dimensionless numbers (Archimedes, Froude, Eötvös and Weber) and the group (d_o/D_C) for the prediction of gas holdup in homogeneous regime is proposed. The average error between the correlation predictions and experimental values remains under ±10%.The proposed correlation is compared with the published data and found to be in fairly good agreement.     

Supercritical CO2 generator using bubble collapse by water hammer

We have developed a new apparatus to dynamically generate supercritical CO₂ (scCO₂) bubbles in water using a water hammer facility by efficiently concentrating water energy. We measured the internal and external pressures of a CO₂ bubble covered with a rubber membrane using pressure transducers, and observed the bubble's oscillations by a high-speed video camera. We evaluated the maximum duration of the scCO₂ for conditions 60 μs in experiments. We performed numerical simulations using the Rayleigh–Plesset equation by substituting the experimental external pressure profiles of the bubble and confirmed that numerical results agreed with the experimental internal pressure. Moreover, in the minimum external pressure condition where we experimentally achieved the condition of scCO₂ in the bubble for 16 μs by water hammer, we obtained the maximum duration of scCO₂ conditions up to 55 μs by numerical simulations assuming isotropic compression.     

Consideration of the dynamic forces during bubble growth in a capillary tube

Single bubbles were generated from a capillary tube in quiescent water and the bubble formation process was studied in detail using high-speed video at two pressures, 1.38 and 0.93 kPa. The bubble equivalent spherical radius, r, was derived from the data sets of 100 bubbles: (1) the a and b semi-axes values for an ellipsoidal model and (2) a (more accurate) cylindrical integration of the bubble image in 1-pixel layers. The two methods were compared and showed a significant improvement with the more accurate integration approach. Based on the time trends in the derivatives of a and b, three growth phases were identified and the different forces acting on the bubble were calculated. The analysis showed significant differences between the two cases, despite similar times from appearance to detachment. For the 0.93 kPa case, the bubble shape detachment is described by Cassini oval while for 1.38 kPa it is a lenmiscate. For the study conditions, the momentum force was negligible for both cases; however, the viscous drag force, added mass, and surface tension forces were not. The bubble eccentricity exhibited an oscillatory behavior, which we propose arose from highly non-linear wave-phenomena. Finally, only the low-pressure case was in agreement with the predictions of Oguz and Prosperetti (1993); for flows less than the critical flow, the high pressure was not in agreement, indicating a smaller value for the transition to super-critical flow for the study conditions.     

Synthesis of large plate-like gypsum dihydrate from waste gypsum board

This paper reports the synthesis of large plate-like gypsum (calcium sulfate) dihydrate generated from waste gypsum board via a wet process. Gypsum hemihydrate was formed by dehydrating waste gypsum board in a sodium sulfate solution at 100 °C for 1 h to form needle-like crystals. The large gypsum dihydrate was formed by lowering the temperature of the gypsum hemihydrate suspension below 80 °C. The large plate-like gypsum dihydrate was obtained by adding seed crystals to a suspension cooled at 80 °C. When 40 μm seed crystals were added at 0.5 mass%, the large plate-like gypsum dihydrate crystals obtained from waste gypsum board had average dimensions of 250 μm length × 100 μm width × 35 μm thickness.     

Porous alumina ceramics prepared with wheat flour

It is shown that wheat flour can be used as a pore-forming and body-forming agent in ceramic technology. In contrast to pure native starch, however, the pores do not result from the swelling starch granules alone but are mainly due to protein-assisted foaming. Therefore the porosity is significantly higher and the pore size larger than that resulting from the starch granules alone, and the wet milling time applied for homogenizing the ceramic suspensions becomes the most critical process parameter. Alumina suspensions with 70 wt.% alumina and 20–30 vol.% wheat flour with different initial particle size (fine grade and semolina, respectively) have been prepared using milling times of up to 8 h. Porosities of up to approx. 60% can be achieved with only 20 vol.% of flour or semolina after 8 h of milling time, with the cell sizes (diameters of pore cavities resulting from foam bubbles) being essentially independent of the milling time (median diameters of 120–240 μm). Effective pore throat sizes (i.e. diameters of cell windows or channels between cells), measured via mercury porosimetry, are 1–2 μm for short milling times (2–3 h), but for long milling times (8 h) they change by more than one order of magnitude to median sizes of 20–30 μm, closely corresponding to the median size of wheat starch granules (approx. 20 μm).     

Liquid dispersion,gas holdup and frictional pressure drop in a packed bubble column at elevated pressures

The gas holdup, frictional pressure drop and liquid dispersion have been investigated in a packed bubble column at elevated pressures for the air–water system. The bubble column, which had an internal diameter of 0.15 m and which was packed with 15 mm plastic Pall rings was operated in the semibatch mode. The operating pressures ranged from 0.1 to 0.66 MPa. It was found that increasing the pressure increases both the gas holdup and the dispersion coefficient. In contradiction to the results obtained from packed bubble columns fed with a continuous net flow of liquid, a maximum point of the frictional pressure drop was observed at the transition point between bubble and pulse flow region.     

Effect of sodium on the creep resistance of yttrium aluminium garnet (YAG) fibres

Aligned YAG fine fibres had been made previously from an aqueous sol–gel process, but it was suspected that sodium contamination in one of the starting materials lessened the creep resistance of the final product. Therefore, sodium-free gel precursor fibres were made using the same process, and upon firing in air at 900 °C these formed pure phase YAG fibres, of 4 μm diameter. When steamed over 200–500 °C for 3 h, the amorphous gel fibre formed single phase nanocrystalline YAG between 400 and 500 °C, an extraordinarily low temperature for this material to crystallise. Upon postfiring up to 1550 °C, grains averaging 0.5 μm and pores of 0.17 μm had formed, but despite this porosity and smaller grain size, the sodium-free fibre exhibited superior creep resistance than the sodium contaminated fibres reported previously by the authors, typically creeping at temperatures 50 °C higher. This demonstrates that even small levels of sodium contamination are harmful to creep resistance in YAG.     

Polymethylmethacrylate (PMMA) sub-microparticles produced by Supercritical Assisted Injection in a Liquid Antisolvent

A recently developed supercritical assisted process, called Supercritical Assisted Injection in a Liquid Antisolvent (SAILA) is proposed to produce polymer micro and nanoparticles in water stabilized suspensions. Polymethylmethacrylate (PMMA) has been selected as the model polymer for a systematic study of the influence of the SAILA operating parameters on particle morphology and diameter. The effect of expanded liquid injection pressure on particle size and distribution was studied and different expanded liquid temperatures and compositions were also explored. Successful precipitation of the polymer in a water stabilized suspension was obtained and narrow particle size distributions were obtained using 70 and 90 bar injection pressures. PMMA particles controlled diameter were produced ranging between 0.2 ± 0.04 μm and 0.9 ± 0.2 μm. Particles are formed from the expanded liquid solution as a consequence of very fast supersaturation produced by spraying it the liquid antisolvent.     

Local void fraction and bubble size distributions in cold-gassed and hot-sparged stirred reactors

Vertical distributions of local void fraction, bubble size and gas–liquid interfacial area in air–water dispersions at 24 and 81 °C have been measured with a dual electric conductivity probe in a fully baffled dished base stirred vessel of 0.48 m diameter holding 0.145 m³ liquid. The agitator was a hollow blade dispersing turbine below two up-pumping hydrofoils. The vertical distribution of the void fraction in the hot conditions is similar to that at ambient temperature though the void fraction is significantly lower in the hot system. The vertical distributions of bubble size show maxima with large bubbles above the bottom impeller, near the top impeller and close to the free surface. With given operating conditions, the overall Sauter means bubble size in the hot systems appears to be about 21% greater than when cold. Estimates of the local interfacial area show a maximum just above the level of the top impeller.     

Particle size effects in flash sintering

The study of 3 mol% yttria stabilized zirconia (3YSZ) with different particle sizes provides new insights into flash sintering. Four powders, all with the same crystallite size but various particle size were investigated: described as nominally 1 μm (D80 = 0.51 μm, meaning 80 vol% has a size less than 0.51 μm), 2 μm (D80 = 0.90 μm), 5 μm (D80 = 2.11 μm) and 10 μm (D80 = 3.09 μm). While the furnace temperature for flash sintering, at a field of 100 V cm^?1, increased from 920 °C to 1040 °C with particle size, the specimen temperature in all instances remained at ～1200 °C. The quantum increase in density decreased with larger particles. The grain size distribution of conventionally and flash sintered specimens remained similar, with some evidence of a preponderance of nanograins in the flash sintered specimens. Joule heating was well below the temperatures that would have been required for sintering in a few seconds. An explanation based upon the nucleation of Frenkel pairs is proposed.