Application of the Maximum Entropy Formalism on Sprays Produced by Ultrasonic Atomizers

A recent application of the Maximum Entropy Formalism on liquid atomization problems led to the development of a mathematical volume‐based drop‐size distribution. This function, which depends on three parameters, is a reduction of the four‐parameter generalized Gamma function. The aim of the present work is to investigate the relevance of the three parameters in the characterization of liquid atomization processes. To achieve this, a variety of experimental drop‐size distributions of ultrasonic sprays were analyzed with the mathematical function. Firstly, it is found that the mathematical drop‐size distribution is very suitable to represent the volume‐based drop‐size distribution of ultrasonic sprays. Furthermore, it is seen that when considering the three parameters introduced by the function, one of them is constant for all the situations investigated, and the other two are linked to a non‐dimensional group that includes the main parameters controlling the drop production. These results are very important, since they suggest a possible development of physical models of primary atomization based on the M.E.F., which would allow for the prediction of the spray drop‐size distribution. Thusfar, such a model does not exist.     

A New Formulation of the Maximum Entropy Formalism to Model Liquid Spray Drop‐Size Distribution

Considering the differences and disagreements involving the previous application of the Maximum Entropy Formalism to modeling the drop‐size distribution of liquid sprays, a new formulation is suggested. The constraints introduced in this formulation are based on characteristic features common to any liquid atomization process, i. e., the production of large and small drops is always limited. These limitations are a consequence of the action of both destabilizing and stabilizing forces such as aerodynamic and surface tension forces, respectively. The solution resulting from this approach, which makes use of statistical mechanics, is a three‐parameter Generalized Gamma Distribution, which can treat any type of distribution. It is shown that this solution is identical to a Nukiyama‐Tanasawa distribution that should no longer be regarded as an empirical distribution. Although this new formulation clearly answers the question concerning the amount of information required to describe a spray drop‐size distribution, it raises the problem of the mathematical form to be given to this information, and is discussed here.     

Deconvolution with Maximum Entropy Solution to Determine Local Extinction Coefficient and Local Volume Concentration Values from Laser Diffraction Data

The traditional use of the laser diffraction technique provides line‐of‐sight liquid spray drop‐size distribution. However, deconvolution of the measurements can be performed for axisymmetric spray in order to determine local spray characteristics. In a previous publication, a new deconvolution technique making use of the maximum entropy principle was established and applied to determine the local drop‐size distributions. The entire approach was experimentally validated. In this work, the technique is employed to determine local extinction coefficient values. As in the previous investigation, the measurement procedure consists of scanning a laser beam through the spray cross‐section from the center to the edge of the spray. By use of the transmittance theory, the local extinction coefficients allow the local volume concentrations to be calculated. This theory introduces the mean scattering coefficient. The results show that this coefficient must be determined as a function of the Sauter mean diameter in order to avoid overestimation of the volume concentration. Although no proper validation is presented, the coherence of the overall approach is discussed in detail and solutions for improving the spatial resolution are presented. Finally, the local volume concentrations are combined with the local drop‐size distribution to provide local volume‐weighted, drop‐size distributions. These distributions provide information on the localization of the drops according to their diameter as well as on the spatial liquid distribution. This work illustrates applications and performances of laser diffraction technique that are rarely used.     

Particle Size Distribution of Nanospheres by Monte Carlo Fitting of Small Angle X‐Ray Scattering Curves

A method for recovering the size distribution of spherical particles from small angle scattering data by using a Monte Carlo interference function fitting algorithm is presented. The method is based on the direct simulation of the small angle scattering data upon the assumption of non‐interacting hard sphere ensembles (“dilute” solution approximation). The algorithm for retrieving the particle size distribution does not require any additional parameters apart from the input of the scattering data. The fitting strategy necessarily implies positive particle size distributions, while preserving the advantage of the indirect transformation method for data desmearing. Furthermore, the present approach does not use any regularisation procedures of the best fit solution and favours smooth particle size distributions. The Monte Carlo procedure has been tested against several simulated cases with various types of mono‐ and bi‐modal size distributions and different noise levels. In the special case of non‐interacting spheres, the Monte Carlo fitting algorithm had the same retrieving ability as the well assessed indirect transformation, structure interference and maximum entropy methods. Finally, the algorithm was applied to retrieve the distribution of spherical nanopowders produced by gas‐to‐particle conversion both as free powder and as reinforcing second‐phase agent in polymer nanocomposites.     

Droplet size and velocity distributions for spray modelling

Methods for constructing droplet size distributions and droplet velocity profiles are examined as a basis for the Eulerian spray model proposed in Beck and Watkins (2002,2003) [5], [6]. Within the spray model, both distributions must be calculated at every control volume at every time-step where the spray is present and valid distributions must be guaranteed. Results show that the Maximum Entropy formalism combined with the Gamma distribution satisfy these conditions for the droplet size distributions. Approximating the droplet velocity profile is shown to be considerably more difficult due to the fact that it does not have compact support. An exponential model with a constrained exponent offers plausible profiles.     

A Predictive Model for the Initial Droplet Size and Velocity Distributions in Sprays and Comparison with Experiments

The distribution of sizes and velocities of droplets initially formed in sprays is an important piece of information needed in the spray modelling, because it defines the initial condition of the spray droplets in the predictive calculations of the downstream two‐phase flow fields. A predictive model for the initial droplet size and velocity distributions in sprays is formulated in this study. The present model incorporates both the deterministic and the stochastic aspect of spray formation process. The deterministic aspect takes into account of the unstable wave motion before the liquid bulk breakup through the linear and nonlinear instability analysis, which provides information for the liquid bulk breakup length, the mass‐mean diameter and a prior distribution for the droplet sizes corresponding to the unstable wave growth of various wavelengths. The stochastic aspect deals with the final stage of droplet formation after the liquid bulk breakup by statistical means through the maximum entropy principle based on Bayesian entropy. The two sub‐models are coupled together by the various source terms signifying the liquid‐gas interaction, the mass mean diameter and the prior distribution based on the instability analysis. The initial droplet size and velocity distributions are measured experimentally by phase‐Doppler interferometry for sprays generated by a planar research nozzle and a practical gas turbine airblast nozzle. For the two nozzles, the liquid bulk sheet is formed before its breakup in a coflowing air stream. It is found that the model predictions are in satisfactory agreement with the experimental data for all the cases measured. Hence the present model may be applied to a variety of practical sprays to specify the initial conditions for the spray droplets formed in practical spray systems.     

Experimental and Theoretical Study of Sprays Produced by Ultrasonic Atomisers

The atomisation of liquids by means of low-frequency ultrasonic atomisers results from unstable surface waves generated on the free surface of a thin liquid film. These unstable waves are obtained from the tuning of the amplitude and the frequency of an imposed oscillation. The thin liquid film develops as the liquid spreads over the atomising surface of the atomiser. This paper focuses on a systematic experimental analysis of the sprays produced by low-frequency ultrasonic atomisers. The thickness of the liquid film was measured and its effects on the drop diameter were studied together with the effects of both the liquid's physical properties and the ultrasonic atomiser's characteristics. The relationship between the mean drop diameter and the surface wave wavelength was accurately determined and introduced into a mathematical approach based on the maximum entropy formalism to predict the drop size distribution of the spray. Within the range of working conditions tested, the application of this formalism is successful and provides a procedure for the prediction of spray drop size distributions from calculations only.     

Statistical Analysis of Pellet Size Variation in Commercial Catalysts

The size of commercial catalyst pellets is distributed over a range of values. The suitability of the normal, lognormal and Weibull distributions to model pellet size variation was judged by three quality‐of‐fit criteria: the coefficient of determination, Akaike information criterion and Kolmogorov‐Smirnov test, respectively. It is concluded that the normal distribution is most appropriate to represent the diameter distribution of the spherical pellets, though the lognormal distribution also gives a satisfactory fit. The Weibull distribution is the best one to represent the length distribution of the catalyst tablets. It is elucidated that a downward tail of the tablet length distribution results from the fluctuating characteristics of feeding for the individual tablets. It is also pointed out that pellet size distributed in a narrow window is beneficial to the industrial application of solid catalysts.     

Modelling of Drop Size Distribution of Rain from Rain Rate and Attenuation Measurements at Millimeter and Optical Wavelengths

This paper describes a technique for modelling of rain drop size distributions at Calcutta in terms of negative exponential function, from the measurements of rain rate and attenuation over a dual wavelength LOS link at millimeter and optical frequencies. The DSD model obtained is then used to determine the attenuation at 94 GHz, for comparison with experimentally obtained attenuation at 94 GHz. This is also compared with the attenuation calculated by considering other experimentally obtained DSD models. The best fit negative exponential distribution function (modified M-P model) is presented along with some other experimentally obtained and reference models.     

The hot nonperturbative gluon plasma is an almost ideal colored liquid

We study properties of a gluon plasma above the critical temperature Tc in a generalized quasiparticle approach with a Lorentz spectral function. The model parameters are determined by a fit of the entropy s to lattice QCD data. The effective degrees of freedom are found to be rather heavy and of a sizable width. With the spectral width being closely related to the interaction rate, we find a large effective cross section, which is comparable to the typical distance squared of the quasiparticles. This suggests that the system should be viewed as a liquid as also indicated by an estimate of the plasma parameter Gamma. Furthermore, within the quasiparticle approach we find a very low viscosity to entropy ratio, eta/s approximately 0.2 for T > 1.05 Tc, supporting the recent conjecture of an almost ideal quark-gluon liquid seen at RHIC.     

Kinetics study of a Diels‐Alder reaction in mixtures of an ionic liquid with molecular solvents

The second‐order rate constants for cycloaddition reaction of cyclopentadiene with naphthoquinone were determined spectrophotometrically in various compositions of 1‐(1‐butyl)‐3‐methylimidazolium terafluoroborate ([bmim]BF₄) with water and methanol at 25 °C. Rate constants of the reaction in pure solvents are in the order of water > [bmim]BF₄ > methanol. Rate constants of the reaction decrease sharply with mole fraction of the ionic liquid in aqueous solutions and increase slightly to a maximum in alcoholic mixtures. Multi‐parameter correlation of logk₂ versus solute–solvent interaction parameters demonstrated that solvophobicity parameter (Sp), hydrogen‐bond donor acidity (α) and hydrogen‐bond acceptor basicity (β) of media are the main factors influencing the reaction rate constant. The proposed three‐parameter model shows that the reaction rate constant increases with Sp, α and β parameters. Copyright © 2008 John Wiley & Sons, Ltd.     

About the modified Gaussian family of income distributions with applications to individual incomes

In a recent paper in this journal [Q. Guo, L. Gao, Distribution of individual incomes in China between 1992 and 2009, Physica A 391 (2012) 5139–5145], a new family of distributions for modeling individual incomes in China was proposed. This family is the so-called Modified Gaussian (MG) distribution, which depends on two parameters. The MG distribution shows a satisfactory fit for the individual income data between 1992 and 2009. However, for the practical use of this model with individual incomes, it is necessary to know its probabilistic and statistical properties, especially the corresponding inequality measures. In this paper, probabilistic functions and inequality measures of the MG distribution are obtained in closed form, including the normalizing constant, probability functions, moments, first-degree stochastic dominance conditions, relationships with other families of distributions and standard tools for inequality measurement (Lorenz and generalized Lorenz curves and Gini, Donaldson–Weymark–Kakwani and Pietra indices). Several methods for parameter estimation are also discussed. In order to illustrate all the previous formulations, we have fitted individual incomes of Spain for three years using the European community household panel survey, concluding a static pattern of inequality, since the Gini index and other inequality measures remain constant over the study period.     

Stochastic modeling of atomizing spray in a complex swirl injector using large eddy simulation

Large-eddy simulation of an atomizing spray issuing from a gas-turbine injector is performed. The filtered Navier–Stokes equations with dynamic subgrid scale model are solved on unstructured grids to compute the swirling turbulent flow through complex passages of the injector. The collocated grid, incompressible flow algorithm on arbitrary shaped unstructured grids developed by Mahesh et al. (J. Comp. Phys. 197 (2004) 215–240) is used in this work. A Lagrangian point-particle formulation with a stochastic model for droplet breakup is used for the liquid phase. Following Kolmogorov’s concept of viewing solid particle-breakup as a discrete random process, the droplet breakup is considered in the framework of uncorrelated breakup events, independent of the initial droplet size. The size and number density of the newly produced droplets is governed by the Fokker–Planck equation for the evolution of the pdf of droplet radii. The parameters of the model are obtained dynamically by relating them to the local Weber number and resolved scale turbulence properties. A hybrid particle-parcel is used to represent the large number of spray droplets. The predictive capability of the LES together with Lagrangian droplet dynamics models to capture the droplet dispersion characteristics, size distributions, and the spray evolution is examined in detail by comparing it with the spray patternation study for the gas-turbine injector. The present approach is computationally efficient and captures the global features of the fragmentary process of liquid atomization in complex configurations.     

Absorption and Scattering of Light by Highly Concentrated Two‐phase Flows

The spray cone emerging during an extended metal atomization process (called spray forming) has been investigated in order to quantify the influence of highly concentrated multiphase flows on phase‐Doppler‐anemometry (PDA) measurements. Using this non‐intrusive, optical measurement technique not only the local particle size and velocity distributions of the spray can be obtained but also additional information about the mass flux in the multiphase flow. Since standard phase‐Doppler systems can be easily applied to low concentrated particle systems (spherical particles with smooth surfaces and an optical transparent continuous phase taken for granted) the application of this measurement technique to highly concentrated multiphase flows is more complex. Both the laser light propagating from the PDA device to the probe volume and the scattered one going backward to the PDA receiving system are disturbed by passing the highly concentrated multiphase flow. The resulting significant loss in signal quality especially concerns the measurement of the smaller particles of the spray because of their reduced silhouette (in comparison with the bigger ones). Thus, the detection of the smallest particles becomes partially impossible leading to measurement of a distorted diameter distribution of the entire particle collective. In this study the distortions of the measured distributions dependent on the particle number concentration as well as on the path length of the laser light are discussed.     

Fusion hindrance in the neck evolution of symmetric nuclear reactions

Fusion hindrance in the radial degree of freedom for massive nuclear reactions is known for a long time. However the present work shows that the fusion hindrance also exists in the neck evolution. We calculate the potential at different distances and different neck parameters by the two-center liquid drop model and then check whether fusion hindrance exists in the neck evolution by examing the sign of slope of potential vs. the neck parameter. The area of fusion hindrance in the neck evolution is shown.     

PDA and Neural Network Investigation of Swirl Spray Interaction Phenomena

Experiments are performed to investigate the atomization characteristics of mixed‐interaction regions of sprays of two swirl injectors installed side by side. Both droplet size and velocity distributions on a plane perpendicular to the axes of the injectors are measured using a PDA system. As a result of the interaction phenomenon, a region of secondary atomization is identified that differs significantly from the hollow region spray of a single swirl injector. A neural network algorithm is used to reconstruct the entire spray field for both droplet size and velocity distribution in extrapolation regimes for injector spacing as well as three dimensional spatial coordinates. Excellent agreement between the predicted values and the measurements is obtained. It is observed that points on the extrapolation regime of the neural network can be predicted with an accuracy of 93 % using a training data set with less than 50 % of the number of data points to be predicted. The results indicate the capability of performing design‐ and optimization studies for pressure‐swirl injectors, with sufficient accuracy, by applying a modest amount of data in conjunction with an overall optimized value for the width of the probability.     

Mean Particle Diameters. Part VIII. Computer Program to Decompose Mixtures of (Truncated) Lognormal Particle Size Distributions Using Differential Evolution to Generate Starting Values for Nonlinear Least Squares

Multimodal size distributions can result from a mixing of two or more component distributions and arise in quite different application areas. Physical and statistical approaches are described for decomposition of a multimodal particle size distribution into a number of lognormal components. These approaches, incorporated in the Fortran computer program FitDist, use a nonlinear least‐squares (NLLS) optimization, requiring initial parameter estimates. A hybrid deconvolution method has been developed. Differential Evolution (DE), is used for generation of initial parameter values, followed by an NLLS optimization to derive precise parameter values at the local optimum. The DE algorithm is required to decide on the proper number of modes to be fitted. Mathematical relationships have been derived to convert the parameter values of one multimodal lognormal moment distribution, e.g., a number distribution, to those of another moment distribution, e.g., a volume distribution. Moreover, mathematical relationships have been derived to compute mean diameters (Moment‐Ratio notation) from the parameters of a multimodal lognormal size distribution. Fitting a 1.5th moment distribution, being just in between a number and a volume distribution, has been introduced as an instrument to balance inaccuracies in both tails of a distribution due to sampling inaccuracies or truncation of these tails. The program fits a truncated size distribution by fitting its frequency density distribution, whereas a complete size distribution is fitted by fitting the cumulative distribution. Some guidelines are given for fitting Number, Diameter, Surface area, and Volume distributions to measured size distributions. Although fitting of multimodal normal distributions is an option, higher moment distributions will not be fitted as these distributions are not normally distributed. Practical examples demonstrate the validity of the method to decompose multimodal particle size distributions by use of DE.     

On the structure of the velocity field under the free spheroidal surface of a viscous liquid drop oscillating in an electrostatic field

An asymptotic analytical solution to an initial boundary-value problem considering (i) the time evolution of the capillary oscillation amplitude as applied to a viscous spheroidal liquid drop placed in a uniform electrostatic field and (ii) the liquid flow velocity field inside the drop is found. The problem is solved in an approximation that is linear in two small parameters: the dimensionless oscillation amplitude and the dimensionless field-induced constant deformation of the equilibrium (spherical) shape of the drop. Terms proportional to the product of the small parameters are retained. In this approximation, interaction between oscillation modes is revealed. It is shown that the intensity of the eddy component of the oscillation-related velocity field depends on the liquid viscosity and the external uniform electrostatic field strength. The intensity of the eddy component decays rapidly with distance from the free surface. The depth to which the eddy flow (which is caused by periodical flows on the free surface) penetrates into the drop is a nonmonotonic function of the polar angle and increases with dimensionless viscosity and field strength.