Prediction of droplet size distribution in sprays of prefilming air-blast atomizers

Droplet size distribution is a crucial parameter of atomization process besides droplet mean diameter. In this paper, the finite stochastic breakup model (FSBM) of prefilming air-blast atomization process has been proposed according to the self-similarity of droplet breakup. There are four parameters in FSBM, which are the initial droplet diameter D₀, the maximum stable droplet size D_c, the minimum mass ratio of a sub-droplet to the mother droplet a, and the droplet breakup probability P(D). The simulation results of droplet size distribution with this model agreed well with the experimental results of prefilming air-blast atomizers. With this model, the nonlinear relationship between the mean droplet diameters and droplet size distribution of the air-blast atomization process can be predicted exactly.     

CHARACTERISTICS OF MICROMETER AND SUBMICROMETER SIZE O/W EMULSIONS PRODUCED BY RAMOND SUPERMIXER®

A novel motionless mixer named the Ramond Supermixer® (RSM®) was employed to produce O/W emulsions composed of micrometer and submicrometer-size droplets. Liquid paraffin as dispersed phase, aqueous sucrose solution as continuous phase, and anionic sodium dodecyl sulfate as emulsifying agent were used as the model emulsification system. Pressure drop, droplet size distribution, Sauter mean diameter (d ₃₂), and geometric standard deviation of the droplet size distribution (σ_g) were investigated under the various combinations of process variables; superficial liquid velocity, number of mixing units, number of passages through RSM®, dispersed phase viscosity (η_d), continuous phase viscosity (η_c), and dispersed phase volume fraction. Different modes of droplet size variations with process variables were obtained, with respect to micrometer- and submicrometer-size ranges, and theoretical explanations are forwarded. For the micrometer-size range, maximum droplet diameter (d _max) was proportional to d ₃₂. For the submicrometer-size range, d _max varied with d ₃₂ in the range of 1.53–2.19-fold, and a correlation is proposed with K (=η_d/η_c); d ₃₂ and σ_g were well correlated with the process variables. Furthermore, a semi-empirical mechanistic model was developed for the formation of droplets obtained under inertial sub-range to interpret the effect of process variables.     

Dispersed phase characteristics in three phase (liquid-liquid-solid) fluidized beds

Two ideal droplet length (l,) distributions have been derived for two different droplet shapes. The dispersed phase holdup (?_d) increases with increasing dispersed phase velocity (U_d), but decreases with increasing continuous phase velocity, (U_c) in three-phase fluidized beds. In the droplet-coalescing flow regime, l_v and the droplet rising velocity (v_d) increase, but the spherical droplet fraction (k) decreases with increasing U_d and u_c. In the droplet-disintegrating flow regime, the effects of u_d and U_c on l_v and k are insignificant, but v_d increases with increasing U_c. Maximum values of l_v, occur in the bed containing 1.7 mm diameter particles and l_v has an uniform length of around 2.0 mm in beds with particle size larger than 3.0 mm.     

Hydrodynamics and droplet size distribution of liquid–liquid flow in a packed bed reactor with orifice plates

A packed bed reactor with orifice plates (PBR@OP) was designed by adding orifice plates periodically in packed beds. Hydrodynamics and droplet size distribution in PBR@OP were experimentally investigated using fatty acid methyl esters (FAME)/water as the model liquid–liquid system. In PBR@OP, the flow pattern was close to plug flow. Droplets with Sauter mean diameter (d₃₂) of 150–550 μm were generated. The pressure drop of orifice, flow velocity and plate spacing were key parameters to control the droplet size. The reactor performance was evaluated by analyzing a FAME epoxidation process. At the same d₃₂ and residence time, the length and total pressure drop of PBR@OP were about 1/3 and 1/4 of those of PBR without orifice plates, respectively. Furthermore, a semi-empirical correlation describing the d₃₂ change in PBR@OP was developed, revealing a relative mean deviation of 8.64%. PBR@OP presents a cost-effective option for the intensification of liquid–liquid medium rate reactions.     

Analysis of droplet expulsion in stagnant single water-in-oil-in-water double emulsion globules

Double emulsions created by phase inversion can be used for fast liquid–liquid separation; therefore, the coalescence behaviors of these types of multiple emulsions need to be predictable for different physical properties and drop size ratios. The aim of this study is to determine the influence of the effective overall drop diameter and the internal droplet size on the coalescence time and the coalescence behavior. Experimental investigations on the physical stability of single stagnant water-in-oil-in-water (W₁/O/W₂) double emulsion globules are performed. For this investigation, a formation device to inject one water droplet into an oil drop inside a water bulk phase is developed. The coalescence process of the sole internal water droplet floating on the O/W₂ interface with the water bulk phase, often termed droplet expulsion or external coalescence, is recorded with a high speed camera. Based on image analysis, the diameters of the effective overall drop D, containing the oil and entrapped water volume, and the internal water droplet d are determined. Additionally, the coalescence time τ, including the time from the first contact of the internal droplet and the drop-bulk interface to the film rupture is measured. A large increase in coalescence time with increasing water droplet diameters is found. For the investigated paraffin oil–water system and initial drop sizes, partial coalescence occurs. In this case, the diameter ratio of daughter-to-mother droplet ψ is determined.     

Spray characterization: Numerical prediction of Sauter mean diameter and droplet size distribution

A simplified equation of the Nukiyama-Tanasawa type for droplet size distribution in sprays is obtained from the synergetic concept of entropy information, assuming spherical droplets and zero and infinity as their limit sizes. The introduction of Sauter mean diameter (SMD) definition in that equation yields a new distribution function dependent solely on SMD, which is calculated from available correlations for pressure-jet and pre-filming airblast atomizers. For plain-jet airblast atomizers a new and dimensionally consistent correlation is determined. Several droplet size distributions are then predicted. Experimental data are compared with predictions of SMD; the agreement is satisfactory.     

Predicting the sizes of toluene-diluted heavy oil emulsions in turbulent flow Part 2: Hinze-Kolmogorov based model adapted for increased oil fractions and energy dissipation in a stirred tank

The applicability of several classical size prediction models to predict the Sauter mean diameters d₃₂ of four volume fractions of toluene-diluted heavy oil in water emulsions was studied. Energy dissipation in a fully baffled cylindrical tank stirred at varied speeds by a Rushton turbine was related to the d₃₂ measured with a Mastersizer 2000 laser light scattering instrument after 75 min of agitation. At low oil fractions of 0.01 droplet breakage behaviour was in accordance with the breakage dominance assumption of Hinze-Kolmogorov (H-K) theory. At intermediate oil fractions of 0.05 and 0.1, the system behaved similarly with lower rates of breakage than at 0.01. At high oil fractions of 0.3, the droplet sizes were the largest. Here, coalescence may have played a larger role due to increased collision frequency, while turbulence dampening caused larger eddies and reduced breakage. The resilience to breakage due to the surface elasticity was assumed to be constant for all oil fractions. The experimental diameters compared with the diameters calculated from several size-predictive models showed unsatisfactory predictions at higher oil fractions of 0.05-0.3. By modification of existing constants and coefficients of a H-K-based model that accounts for breakage and coalescence and by using iterative techniques, specific predictive equations for each mixing speed and a new averaged equation were developed. Good agreement between measured and predicted sizes was achieved.     

Auslegung von Einstoff-Druckdüsen

Design of single substance pressure jets . The single substance pressure jet most commonly used produces liquid lamellae. These then degenerate to drops in the surrounding gas as a consequence of aerodynamic wave formation. The size of the droplets generated decreases according to Δp^?1/3. Calculation of droplet size can be unified by defining a lamella number. The lamella number expresses the drop in lamella thickness with increasing jet distance relative to the diameter of the jet. The smaller the geometry-dependent lamella number, the smaller the droplets relative to the diameter of the jet. The hollow cone jet produces the smallest lamella number and thus generates the smallest droplets. However, the liquid in the twist chamber of hollow conical jets at higher viscosities are decelerated much slower than in fan-shaped jets. Measuring methods for determining the principal characteristics of jets are demonstrated for hollow conical and fanshaped jets. Turbulence jets, such as solid conical jets, produce droplets some 5 to 10 times larger than those in hollow conical jets at the same jet diameter and pressure. Droplet formation occurs mainly by turbulence in the jet. It is important to minimize the danger of blockage when operating jets. For this reason, preference is often given to that jet which gives the smallest droplet size for a given pressure and a given narrowest orifice diameter. This is particularly significant at droplet sizes d < 20 μm. Comparison and choice of jets is accomplished with the aid of a diagram representing the relative droplet size d₃₂/D in terms of the jet pressure number Δp = ΔpD/σ and the Ohnesorge number. However, a characteristic field representation containing only characteristic quantities formed with the characteristic droplet size, apart from the relative droplet size, is suitable for determining the jet dimensions.     

An electrochemical model for reduction of MnO2 with Fe2+ ions

The reduction kinetics of MnO₂ by Fe²⁺ ions in acidic solution have been studied. The effects of stirring rate, particle size, temperature, Fe²⁺ and H⁺ concentrations have been investigated. Diffusional resistances are negligible above 900 r.p.m. and the rate is controlled by electrochemical reaction. A mixed-potential model developed to describe metallic corrosion has been used in combination with the shrinking core model to explain the reaction kinetics. The overall reaction has been written in terms of cathodic and anodic half-cell reactions. The Tafel equation has been used as a starting point to derive a rate equation. A value of 0.5 has been obtained for charge transfer coefficients, which implies the existence of symmetrical charge barriers. The kinetics of the cathodic reduction reaction are first order with respect to the proton concentration.Nomenclature D diffusion coefficient (cm² s^–1) - D _i impeller diameter (cm) - D _T reactor diameter (cm) - d ₀ initial particle diameter (m) - E e.m.f. between platinum and saturated calomel electrode (V) - F Faraday constant - e electrode potential (V) - e _j liquid-junction potential (V) - k _a rate constant of anodic half-cell reaction - k _c rate constant of cathodic half-cell reaction - k ₁ Mass transfer coefficient (cm s^–1) - m ₀ amount of MnO₂ charged in the reactor (g) - M Molecular weight of MnO₂ - N Stirring speed (r.p.m.) - Re _s Reynolds number for stirring (ND _i ²/) - Re _p Reynolds number for particle (d ^4/31/3/) - Sc Schmidt number (/D) - Sh Sherwood number (k ₁ d/D) - V Reaction volume (dm³) - X Conversion factor Greek letters _i stoichiometric coefficient of reactant i in Equation 1 - stirring energy per unit volume - kinematic viscosity (cm² s^–1) - time required for complete conversion (min)     

Combustion characteristics of droplets composed of light cycle oil and diesel light oil in a hot-air chamber☆

Development of gas turbines fueled with light cycle oil (LCO) and oil mixture of LCO and diesel light oil (LO) requires an understanding of the droplet burning and vaporization characteristics of those oils. The present study is devoted to comparing the burning characteristics of isolated fuel droplets composed of an LCO and an LO. The tests were conducted in an atmospheric hot-air chamber preset at 1173 K, and the examined LCO had a lower cetane number but higher volatility and aromatics content compared to LO. It was demonstrated that the burning of the LCO droplet was sootier, while that of the LO droplet was more disruptive. At the tested temperature, coke formation was indistinct for both the oils, whereas slightly higher ignition delay time was shown for the LO droplet. The microexplosive burning more or less complicated the time-series droplet size d, an explicit burning rate constant, however, was still definable according to the d²-law to show the overall regression speed of the droplet surface area d² with burning time t. The rate constant exhibited little difference for smaller LCO and LO droplets but was greater for LO when the droplet was larger. The rate constant also gradually increased with increasing the initial droplet diameter d₀, which caused the relative size d/d₀ to be unified (normalized) into a single curve by a burning time t/d₀ⁿ (1.0<n<2.0). Analysis revealed that this unification resulted from the respective overlaps of the unsteady and quasi-steady burning phases for differently sized droplets. Further, it was clarified that the unification and analysis are generally valid to isolated liquid fuel droplet burning in hot ambiences.     

Prediction of dispersed phase drop diameter in polymer blends: The effect of elasticity

This paper discusses the prediction of the dispersed phase drop diameter in polymer blends considering the viscoelastic properties of polymers. The prediction is based on a simple force proportionality. Polymers are viscoelastic, and thus the elasticity of the matrix and the elasticity of the dispersed phase affect the drop size. The forces that deform a polymer droplet in a polymer matrix are the shear forces, η_mγ, and the matrix first normal stress, T_11,m. This deformation is resisted by the interfacial forces, 2 Γ/D and the drop's first normal stress, T_11,d. As a first approximation, the forces were balanced to predict the particle size in polymer blends. The diameter of the dispersed phase was predicted reasonably well for several systems at different operating conditions. It was observed for some systems (PS/PP, PS/EPMA, PS/PA330) that, as the shear rate increased, the diameter of the dispersed phase initially decreased. At a critical shear rate, the diameter reached a minimum value, and beyond it, the diameter increased with shear. This critical value was found to be between 100 to 162.5 s⁻¹ for a PS/PP system. The force balance predicts this minimum drop diameter at a similar critical shear rate. The specific energy input (the amount of energy input into the blend) could not explain the phenomenon of a minimum drop diameter with increase in shear. This minimum is not observed for the high concentration systems, such as the 20% PP dispersed in PS, since the effects of coalescence become significant. In reactive blends, the predicted drop diameter was closer to the experimentally determined diameter, and there was less variation in diameter with changes in shear rate.     

A Comparative Study of Drop Sizing Equipment for Agricultural Fan-Spray Atomizers

In an attempt to standardize the classification of agricultural fan-spray atomizers by spray quality, the British Crop Protection Council formulated a standard test procedure for 110° fan-spray atomizers to be carried out by as many different drop-size analyzers as possible. The methods used ranged from in-flight laser and optical imaging systems to analysis of spray samples on collecting surfaces. Almost all systems showed a consistent increase in D_v.0.5 with increasing atomizer size for a given pressure and all showed a decrease in the percentage volume of spray below 100 μm. Although there were some variations between results given by analyzers of the same type but different models or operating systems, in general, the variability of results for specific machine types was not significant. However, except for the results from the Malvern equipment analyzed using two different programmes, the results from the other analyzers showed no consistent agreement. By aggregating data from similar equipment, a difference between spatially and temporally derived data was observed.     

Preparation of poly(vinyl acetate) microspheres with narrow particle size distributions by low temperature suspension polymerization of vinyl acetate

To estimate influences of suspension polymerization conditions including conversion, polymerization temperature, stirring rate, initiator concentration, monomer concentration, and suspending agent concentration on the volume average diameter (D_avg) and particle size distribution (PSD) of poly(vinyl acetate) (PVAc) microspheres, vinyl acetate (VAc) was suspension‐polymerized at low temperature using 2,2′‐azobis(2,4‐dimethylvaleronitrile) as an initiator. The effects of each condition, on D_avg of PVAc microspheres, were expressed as follows, D_avg = [conversion]^a[temperature]^b[rpm]^c[ADMVN]^d[VAc]^f [suspending agent]^g. Logarithms of D_avg were linearly proportional to those of polymerization conditions, and their exponents, a, b, c, d, f, and g were calculated as 0.27, ?13.7, ?1.37, ?0.21, 0.58, and 0.29, respectively. Variations of PSDs, according to polymerization conditions, were examined by considering polymerization rate, droplet or suspension viscosity, and droplet break‐up/coagulation equilibrium. From these results, PVAc microspheres with various sizes and narrow PSDs were obtained effectively under carefully controlled polymerization conditions, which can be used as promising precursors of novel PVA microspheres through heterogeneous surface saponification. © 2006 Wiley Periodicals, Inc. J Appl Polym Sci 101: 4064–4070, 2006     

Droplet splitting in multi-furcating microchannel: A three-dimensional numerical simulation study

This work investigates the splitting of a droplet in a multi-furcating microfluidic channel for a two-phase system employing 3D simulation. The simulations were performed using an explicit volume of fluid (VOF) method and have been validated using experimental data taken from the literature. The width ratio of the branch channel to the main channel is set to 0.25 for five branches of the multi-furcating microchannel, as it is the width ratio at which multiple splitting takes place. Simulations have been carried out at different oil velocities (V_o) ranging from 0.12 to 0.22 m/s and at different water velocities (V_w) ranging from 0.002 to 0.10 m/s. Oil fraction data in the main channel has been recorded and compared with the homogenous model. The average difference between the homogeneous model and the 3D simulations is 22.68%. Analysis of dimensionless droplet length in ±0°, ± 40°, and 90° branch channels has been done. α (length of the droplet in branch channel/width of the main channel) increases up to a flow rate ratio of 0.38, and then decreases, whereas β (length of the droplet in the main channel/width of the main channel) increases with an increase in flow rate ratio. A flow pattern map has been developed to identify the various droplet breakup regimes at the junction. Frequency (counts per unit time) of droplet generation increases with capillary number for all the branch channels except for the 0° branch channel, where the regime is that the droplet passes through three branch channels. The volume distribution ratio (λ) decreases at first, then increases with an increase in capillary number for 0°:90° and 40°:90° angle branch channels for the regime where the droplet passes through five branch channels. For the regime where the droplet passes through three branch channels, the trend is likely linear with λ = 0.3 ± 0.04. The dimensionless mother droplet length increases with an increase in capillary number for V_o = 0.13 and 0.16 m/s, but for V_o = 0.19 and 0.22 m/s, the dimensionless mother droplet length becomes constant after capillary number = 0.26 and 0.30 respectively. The droplet breakup time (t) for regime (a), where the droplet passes through three branch channels, is 0.002 s; for regime (b), where the droplet passes through five branch channels, it is 0.001 s; and for regime (c), where multi-furcation and coalescence of the droplet occurs, it is 0.0005 s. Multiple splitting is a topic covered in this paper that can be applied to upcoming microfluidic platform-based devices.     

Electrodeposition from a fluidized bed electrolyte. II. Effects of bed porosity and particle size

Mass transfer from a fluidized bed electrolyte containing inert particles has been found to depend on bed porosity and particle size. The optimum porosity was found to vary from 0.52 – 0.57 with decreasing particle size but mass transport increased with particle size.A mass transfer entry length effect was observed on the cylindrical cathode but its position within the bulk of the bed was found not to be critical, thus indicating that the hydrodynamic entry length was small. The limiting current density was found to vary as (d _e/L _e)^0.15 whered _e is the annular equivalent diameter andL _e the electrode length.List of symbols Re_I modified Reynolds No. =U _o d _p /v(1–) - Re_II particle Reynolds No. =U _o d _p /v - Re_O sedimentation Reynolds No. =U _i d _p v (constant value) - Re_t terminal particle Reynolds No. =U _t d _p /v - Sc Schmidt No. =v/D - St_I modified Stanton No. =k _L /U _o - C _b bulk concentration, M cm^–3 - D diffusion coefficient, cm² s^–1 - d _t tube diameter, mm - d _e electrode equivalent diameter, mm - d _p particle diameter, mm - bed porosity - zF Faradaic equivalence - cd current density - i _L limiting current density, mA cm^–2 - i _LO limiting current density in the absence of particles - k _L mass transfer coefficient, cm s_–1 - L _e electrode length, mm - m, n constants or indices - v kinematic viscosity, cm² s^–1 - U _o superficial velocity, cm s^–1 - U _i sedimentation velocity, cm s^–1     

Effect of bundling on the π plasmon energy in sub-nanometer single wall carbon nanotubes

Due to their unusual electronic and vibrational properties, single walled carbon nanotubes (SWCNTs) with sub-nanometer diameters d ∼ 0.5–0.9 nm have recently gained interest in the carbon community. Using UV–Vis–NIR spectroscopy and ultra-centrifugation, we have conducted a detailed study of the π plasmon energy (present at∼5–7 eV) in sub-nm SWCNTs as a function of the size of the bundle. We find that the energy of the π plasmon peak E varies with the bundle diameter D_h as E = (-0.023 eV)^∗ln(D_h/d_o) + 5.37 eV, where d_o = 0.5 nm and corresponds to the smallest tube diameter.¹ This is compared with the same data for HiPCo and Carbolex SWCNTs of larger diameter (1–1.4 nm) confirming a clear dependence of E on the bundle size, which is present in addition to the previously reported dependence of E on SWCNT diameter d.     

Morphological and fractal studies of polypropylene/poly(ethene‐1‐octene) blends during melt mixing using scanning electron microscopy

BACKGROUND: Polymer blending creates new materials with enhanced mechanical, chemical or optical properties, with the exact properties being determined by the type of morphology and the phase dimension of the blend. In order to control blend properties, morphology development during processing needs to be understood. The formation and evolution of polypropylene/poly(ethylene‐1‐octene) (PP/POE) blend morphology during blending are qualitatively represented by a series of time‐dependent scanning electron microscopy (SEM) patterns. The area diameter and its distribution of dispersed phase domains are discussed in detail. In order to characterize the formation and evolution of phase morphology quantitatively, two fractal dimensions, D_s and D_d, and their corresponding scaling functions are introduced to analyze the SEM patterns. RESULTS: The evolution of the area diameter indicates that the major reduction in phase domain size occurs during the initial stage of melt mixing, and the domain sizes show an increasing trend due to coalescence with increasing mixing times. The distribution in dispersed phase dimension obeys a log‐normal distribution, and the two fractal dimensions are effective to describe the phase morphology: D_s for dispersed phase dimension and D_d for the distribution in it. CONCLUSIONS: The fractal dimensions D_s and D_d can be used quantitatively to characterize the evolutional self‐similarity of phase morphology and the competition of breakup and coalescence of dispersed phase domains. It is shown that the fractal dimensions and scaling laws are useful to describe the phase morphology development at various mixing times to a certain extent. Copyright © 2007 Society of Chemical Industry     

Mass transfer in agitated vessels: energetic aspects

This paper deals with the experimental determination of the mass transfer rates between the liquid of an agitated vessel and a spherical particle immersed in a reactor. The spatial distribution of the mass transfer coefficients is obtained using an electrochemical method and the influence of the most pertinent hydrodynamic parameters (impeller speed and fluid residence time inside the vessel) is deduced from experimental results. The study considers the two limiting cases of mechanical agitation alone and agitation induced by the liquid jets generated by the feed nozzles. It is shown that knowledge of the specific power dissipated per unit mass of fluid can be useful for the theoretical prediction of the mass transfer rates.Nomenclature D molecular diffusion coefficient - d _p particle diameter - D _i impeller diameter - D _v vessel diameter - E hydrodynamic parameter defined in Equation 8 - H vessel height - k mass transfer coefficient - l characteristic length in Equation 2 - M mass of liquid in the vessel - N rotational speed of agitator - N _p specific power number - P specific power delivered - Q _v volumetric flow rate of the feed fluid - S area for fluid injection - (Sc) =v/D = liquid Schmidt number - (Sh) =kd _p /D kl/D = Sherwood number - U local fluid velocity - V vessel volume - x vertical distance between the bottom of vessel and the measuring point - coefficient in Equation 2 - kinematic viscosity - fluid density - =V/Q _v = residence time in vessel     

Study of the drying behavior of high load multiphase droplets in an acoustic levitator at high temperature conditions

Experimental data on the drying behavior of suspension droplets is limited, despite its importance in industrial applications for material processing, chemical or the food industry involving spray dryers. This fact is particularly significant for high load and temperature conditions, as found in such industrial applications. In this work, the drying behavior of acoustically levitated multiphase droplets has been experimentally investigated. The acoustic tube levitator has been modified in order to allow experiments to be performed at high temperature conditions. The flow rate, temperature and relative humidity of this air stream can be controlled by an air conditioning system. A CMOS camera and a back-light illumination system are used to measure the droplet cross-sectional area and vertical position of the droplet during the drying process. The experiments have been performed using water–glass particle suspensions. The glass particles have a mean particle size and relative density of 13 μm and 2.5, respectively. The effect of the air temperature (60 °C<T<120 °C), initial volume of the droplet (0.05 μl<V₀<0.7 μl), initial solid mass load (0.01<Y_S<0.5) and relative humidity of the air (0.05<HR<0.45) on the mean porosity of the grain, first drying period duration and liquid evaporation rate has been analyzed by means of a parametric screening matrix and also by means of a central composite design (CCD) experimental design. The most important parameters to be considered for the porosity and the drying behavior in the range of variables analyzed are the initial solid mass load and the initial droplet volume. The relative humidity of the air exerts a moderate influence on the drying behavior of the droplet and the temperature has only a very low impact on the mean porosity. In addition, particular attention should be given to the drying behavior of small droplets, which result in a very low mean porosity values for high solid mass loads. The CCD confirms that the initial droplet volume, the solid mass load and their interaction exert significant influence on the three responses.     

Plate wettability and droplet diameter in a pulsed perforated-plate extraction column

The liquid-liquid-plate contact angles of several plate materials were measured using a “preferential wetting method” proposed by the authors. The adhesional work (I_SL), calculated on the basis of contact angles was utilized to evaluate plate wettability. The experimental results indicate that different plate materials affect differently plate wettability. The plate wettability of stainless steel in an aqueous phase was observed to be different from that in an organic phase. The droplet diameter in the mixer-settler region of operation of a pulsed perforated-plate extraction column using several different plate materials, was measured. An empirical correlation of mean droplet diameter: d_p = 0.222(a_f/l_SL^1/3)^?0.13 is proposed except for alumina plates with dispersed aqueous phase operation.