SEDIMENTATION OF MULTISIZED PARTICLES by AKALANKKUMAR C. KOTHARI, B.E. A THESIS IN CHEMICAL ENGINEERING Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN CHEMICAL ENGINEERING Approved August 1981 I 7-- ACKNOl/LEDGMEflTS The author wishes to express his deep aopreciation and thanks to Dr. Sami Selim for his encouraqement and advice throughout this work. The author wishes also to thank Dr. H. R. Heichelhei'^, Dr. Lizi Mann, and Dr. R. W. Tock for serving on the thesis committee a^^d for giving valuable suggestions. Financial support provided by NSF grant - ENG78-27000 for carrving out the literature survey and for ourchasing the glassbeat-^s is ar^^tefully acknowledged. Financial suoport in the form of a teac'^in^ assis- tantship from the Chemical Engineering Department at Texas Tech University is gratefully appreciated. Special thanks is expressed to classmate David L. Roberts for helping in the computer work and for reading the manuscript. 11 TABLE OF CONTENTS PAGE ACKNOWLEDGMENTS ii ABSTRACT v LIST OF TABLES vi LIST OF FIGURES I. II. III. viii INTRODUCTION 1 SEDIMENTATION OF EQUISIZED PARTICLES 11 Dimensional Analysis 17 Published Correlations 23 SEDIMENTATION OF MULTISIZED PARTICLES 31 Sedimentation of Suspensions of Uniform Particles . . . . 35 Sedimentation of Particles in a Binary Suspension. . . . 36 Sedimentation of Suspensions of Multisized Particles. . . 39 Sedimentation of Suspension with Continuous Particle Size Distribution IV. V. 47 PROCEDURE AND EXPERIMENTS 48 Glass Spheres 48 Coloring Glass Particles 50 Suspending Media 50 Experimental Setup 52 Experimental Procedure 52 RESULTS AND DISCUSSION 54 Comparison of Models for Binary Suspensions 55 Comparison of Models for Ternary Suspensions 101 Behavior of Models at High Reynolds Numbers 106 i11 PAGE Comparison of Models for Suspensions with. VI. Continuous Size Distribution 108 CONCLUSIONS AND RECOMMENDATIONS 116 Conclusions 116 Recommendations for Further Work 117 NOMENCLATURE 119 BIBLIOGRAPHY 122 APPENDICES A. Experimental Data and Comparison with Predictions from Proposed Model B. 126 Computer Program for Prediction of Interface Velocities in Multisized Particle Suspensions . . . 1 V 139 ABSTRACT A new model is developed for the sedimentation of multisized particles. Unlike previously published models, the present model takes into account interparticle interactions by taking into consideration the buoyancy effect induced by the smaller size particles on the terminal falling velocities of large size particles. The new and previously published models are tested against published and newly collected data on suspensions with discrete size distribution. It is shown that the present model is the most accurate and represents the experimental data satisfactorily. Moreover, the model satisfactorily represents the ex- perimental data on binary countercurrent operations where the Reynolds number is as high as 546. It is also shown that the proposed model can be used to predict the sedimentation velocities of suspensions with continuous size distribution. LIST OF TABLES Table 2.1 Page Published Correlations for the Solid-Fluid 24 Vertical Flow Operations 4.1 Properties of the Glass Spheres 49 4.2 5-1 Properties of Suspending Media Comparison of the Proposed and Previously Published Models with the Experimental Data on Binary Suspensions 52 57 Comparison of the Proposed and Previously Published Models with the Experimental Data on Ternary Suspensions 102 Comparison of the Proposed and Previously Published Models with the Experimental Data of Lockett and Al-Habbooby on Countercurrent Operations 107 Comparison of the Proposed Model with the Experimental Data for Binary Suspension in Ethylene Glycol 127 Comparison of the Proposed Model with the Experimental Data for Binary Suspension in Diethylene Glycol 129 Comparison of the Proposed Model with the Experimental Data for Binary Suspension in Aqueous Glycerol 130 Comparison of the Proposed Model with the Experimental Data of Smith (1965) on Binary Sedimentation 131 Comparison of the Proposed Model with the Experimental Data of Mirza and Richardson (1979) on Binary Sedimentation 132 Comparison of the Proposed Model with the Experimental Data for Ternary Suspension in Diethylene Glycol 133 5.2 5.3 A.l A.2 A.3 A.4 A.5 A.6 vi Table A.7 A.8 A.9 Page Comparison of the Proposed Model with the Experimental Data of Lockett and Al-Habbooby (1973) on Countercurrent Solid-Liquid Vertical Flow 134 Comparison of the Proposed Model with the Experimental Data for Suspensions with Continuous Size Distribution 137 Comparison of the Proposed Model with the Experimental Data for Suspensions with Continuous Size Distribution 138 vn LIST OF FIGURES Figure 2.1 2.2 Page Relation between rate of sedimentation and voidage of suspension. 22 Slope n as a function of d/D for various values of Re . 25 00 2.3 2.4 3.1 3.2 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 General correlation for the relative velocity in solid-liquid systems, based on extended definitions of Re. and Cp.. 28 A growth curve showing relationship between Ug/U^ and Re at several values of c. 29 Segregation and zone-formation during sedimentation of a binary suspension. 32 Formation of zones in sedimentation of multisized particle suspension. 40 Comparison of experimental results with model for dL=0.0460 cm, d^= 0.0194 cm, and C^= 0.0696. 58 Comparison of experimental results with model for dL= 0.0460 cm, d^= 0.0194 cm, and C^= 0.119. 59 Comparison of experimental results with model for dL= 0.0460 cm, d^= 0.0194 cm, and C,_= 0.168. 60 Comparison of experimental results with the model for d,= 0.0460 cm, d^= 0.0194 cm, and C^= 0.217 61 Comparison of experimental results with the model for d,= 0.0460 cm, d^= 0.0137 cm, and 0^= 0.0595. 62 Comparison of experimental results with model for d,= 0.0460 cm, d^= 0.0137 cm, and C^= 0.119. 63 Comparison of experimental results with model for dL= 0.0460 cm, d^= 0.0137 cm, and C^= 0.168. 64 Comparison of experimental results with model for dj_= 0.0460 cm, d^= 0.0137 cm, and C, = 0.217. 65 Comparison of experimental results with model for d^= 0.0326 cm, d^= 0.0137 cm, and 0^= 0.0797. 66 vm Figure 5.10 5.11 5.12 5.13 5.14 5.15 5.16 5.17 5.18 5.19 5.20 5.21 5.22 5.23 5.24 5.25 Page Comparison of experimental results with model f o r d^= 0.0326 cm, d^= 0.0137 cm, and 0^= 0.139. 67 Comparison of experimental results with model f o r dj_= 0.0326 cm, d^= 0.0137 cm, and 0^= 0.126. 68 Comparison of experimental results with model f o r d^= 0.326 cm, d^= 0.0137 cm, and C^= 0.200. 69 Comparison of experimental results with model f o r d|_= 0.0326 cm, d^= 0.0081 cm, and C^= 0.0624. 70 Comparison of experimental results with model f o r dj_= 0.0460 cm, d^= 0.0194 cm, and 0^= 0.0612. 71 Comparison of experimental result:- with model f o r d(_= 0.0460 cm, d^= 0.0194 cm, and 0^= 0.122. 72 Comparison of experimental results with model f o r d,= 0.0460 cm, d^= 0.0194 cm, and 0^= 0.173. 73 Comparison of experimental results w i t h model f o r d^=0.0460 cm, d^= 0.0194 cm, and 0^= 0.225. 74 Comparison of experimental results with model f o r d^=0.0460 cm, d^= 0.0137 cm, and 0^=0.0612. 75 Comparison of experimental results with model f o r d,= 0.0460 cm, d^= 0.0137 cm, and 0^= 0.120. 76 Comparison of experimental results with model f o r d,= 0.0460 cm, d^= 0.0137 cm, and C^= 0.188. 77 Comparison of experimental results with model f o r d, = 0.0460 cm, d^= 0.0137 cm, and C^= 0.225. 78 Comparison of experimental results with model f o r d,= 0.0326 cm, d^= 0.0137 cm, and C^= 0.0612. 79 Comparison of experimental results with model f o r d.= 0.0326 cm, d^= 0.0137 cm, and C^= 0.122. 80 Comparison of experimental results with model f o r d, = 0.0326 cm, d^= 0.0137 cm, and Cj_= 0.174. 81 Comparison of experimental results w i t h model f o r d, = 0.0326 cm, d^^^ 0.0137 cm, and C^= 0.0615. 82 \x Figure 5.26 5.27 5.28 5.29 5.30 5.31 5.32 5.33 5.34 5.35 5.36 5.37 5.38 5.39 5.40 Page Comparison of experimental results with model for d|_= 0.0326 cm, d^= 0.0137 cm, and 0^= 0.123. 83 Comparison of experimental results with model for dj_= 0.0326 cm, d^= 0.0137 cm, and C^= 0.174. 84 Comparison of experimental results with model for dj_= 0.0326 cm, d^= 0.0137 cm, and 0^= 0.225. 85 Comparison of experimental results with model for dj_= 0.0326 cm, d^= 0.0081 cm, and C^= 0.0615. 86 Comparison of experimental results with model for d^= 0.0326 cm, d-= 0.0081 cm, and C^= 0.123. 87 Comparison of the proposed model with experimental data of Smith. 88 Comparison of the proposed model with experimental data of Smith. 89 Comparison of the proposed model with experimental data of Smith. 90 Comparison of the proposed model and MizraRichardson model with experimental data of Mizra and Richardson. 91 Comparison of the proposed model and MizraRichardson model with experimental data of Mizra and Richardson. 92 Comparison of the proposed model and MizraRichardson model with experimental data of Mizra and Richardson. 93 Comparison of models with the new experimental data for the lower interface. 94 Comparison of models with the new experimental data for the lower interface. 95 Comparison of models with the new experimental data for the lower interface. 96 Comparison of models with the new experimental data for the lower interface. 97 Figure Page 5.41 Comparison of models with the new experimental data for the lower interface. 5.42 Comparison of models with the experimental data for the upper interface. 99 Comparison of models with the experimental data for the upper interface. 100 Comparison of the proposed model with the experimental data on ternary suspensions. 103 Comparison of the proposed model with the experimental data on ternary suspensions. 104 Comparison of the proposed model with the experimental data on ternary suspensions. 105 Comparison of the proposed model with experimental data on suspensions with continuous size distribution. 110 Comparison of the proposed model with experimental data on suspensions with continuous size distribution. Ill Comparison of the proposed model with experimental data on suspensions with continuous size distribution. 112 Comparison of the proposed model with experimental data on suspensions with continuous size distribution. 113 Comparison of the proposed model with experimental data on suspensions with continuous size distribution. 114 Comparison of the proposed model with experimental data on suspensions with continuous size distribution. 115 5.43 5.44 5.45 5.46 5.47 5.48 5.49 5.50 5.51 5.52 XI 98 CHAPTER I INTRODUCTION Chemical engineers often come across systems in which relative motion takes place between a fluid and suspended solids. The last fifty years of research have witnessed an increased importance of solidliquid operations and have resulted in the need for a greater understanding of their underlying characteristics. Most of the investigations so far have been confined to systems of equisized spherical particles moving relative to a fluid. Unfortunately, such idealized systems are rarely found and attention has turned towards systems of multisized spherical or non-spherical particles. Sedimentation, fluidization, and co- or counter-current solid-liquid operations are among the major processes that fall under relative motion between solids and fluids. Such pro- cesses are widely employed in a large number of chemical and allied industries as listed below. (1) Various gravity settling equipment. (a) thickeners, (b) dust collectors, (c) gravity settlers, (d) spray dryers. This includes: (2) Fluidized-bed and moving-bed processes. (3) Pneumatic transport of solids in vertical or inclined pipe- lines using forced draft. (4) Settling and particle-growth in crystal 1izers. (5) Electrostatic separators. 1 (6) Erythrocyte sedimentation, a standard clinical test. (7) In mining and paint industries where relative motion between solid and fluid is usually found. The first important theoretical study of the forces acting on an immersed body moving relative to a viscous fluid was made by Stokes (1851). Although the theory can be used to predict terminal falling velocities, it is confined to a spherical particle moving in an infinite fluid medium at low relative velocity so that the inertia of the fluid could be neglected. Predicted terminal velocities are accurate within 1.5% for a particle Reynolds number (Re) less than 0.1. It has been observed that the container walls of the solid-fluid system oppose the motion of a settling particle and so the terminal falling velocities in a finite fluid medium are always lower than those in an infinite medium. Several theoretical and empirical correlations have been sug- gested for the wall correction of terminal falling velocities. Filderis and Whitmore (1961) reported an extensive experimental study of the wall effect for spheres over the Reynolds number range between 0.05 and 2 X 10 . Happel and Brenner (1965) discussed various theoretical equations describing the wall effect in the viscous region (Re < 0.2). For Reynolds number less than 0.2, an equation developed by Francis (1933) appears to be the most satisfactory. A simple and quite accurate equa- tion was reported recently by Garside and Al-Dibouni (1977) over the Reynolds number range between 3 and 1200. Stoke's Law overpredicts the terminal falling velocity by about 3.0"-^ when the particle Reynolds number is around 0.2. A closer ap- proximation was achieved by Oseen (1910) and by Goldstein (1929) up to Reynolds number of 20. Terminal falling velocities at higher Reynolds numbers can be calculated using an experimental plot of drag coefficient versus Reynolds number. This method involves a trial and error computation as both coordinates contain the terminal velocity as an unknown variable. The terminal velocity can be eliminated from one o 2 of the coordinates by using the Galileo number (d p^(p -P^)/LI.|: ). Zenz (1957), Jottrand (1958) and Zenz and Othmer (1960) have constructed "I/O 7 "[ I '\ plots of (ReCj^)^^ versus (Re Cpj)'^"^ to avoid trial and error tions. calcula- Equations developed by Davies (1945) and by Turian et al. (1971) can be used to predict terminal velocities with little loss in accuracy. No satisfactory treatment has been developed to predict terminal falling velocities of non-spherical particles but the problem has been considered by several investigators. The first important development was reported by Heywood (1938), who published results covering the range of particle Reynolds number between 0.01 and 1000. Non-sphericity was taken into account by introducing a volume coefficient. Pettyjohn and Christiansen (1948) made careful measurements of the settling rate of a number of well defined isomeric bodies in the region of streamline flow and showed that Stokes' Law could be extended to cover nonspherical particles. Kunkel (1948), on the other hand, investigated the magnitude of errors introduced by calculating the settling rate of dust particles from Stokes' Law. Gurel et al. (1955) carried out an extensive investigation by gradually changing the shape of particles from spherical to cubic and from cubic to cylindrical. They claimed that their equation predicts the terminal falling velocities within an error of + 2%. A theoretical work was reported by Becker (1959). who claimed that his equation showed a minimum dependence on particle shape and Reynolds number. Hottovy and Sylvester (1979) conducted a study to measure the terminal falling velocity of roundish but irregularly shaped particles over the particle Reynolds number range between 7 and 3000. Particles passing through a screen, but retained on the next size smaller screen, were assumed to have an average particle diameter equal to the average of the opening width of the two screens. Settling veloci- ties agreed with those of spherical particles for Reynolds number less than 100. For Reynolds number ranging from 100 to 3000, the settling velocities were lower than those of spherical particles and the error introduced was as high as 50% at Reynolds number of 3000. Sedimentation velocities of particles in concentrated suspensions are considerably lower than their terminal falling velocities under free settling conditions. as hindered settling. Such behavior of settling is primarily known The decrease in rate of sedimentation of parti- cles is due not only to the interference of neighboring particles but also to the appreciable upward flow of displaced fluid. An earlier study on sedimentation was carried out by Coe and Clevenger (1916), who concluded that a concentrated suspension may settle in two different steps. In the first, after an initial brief accelera- tion period, the interface between the clear liquid and the suspension moves downward at constant rate and a layer of sediment builds UD at the bottom of the container. When the interface is closer to the sedi- ment, the rate of fall of the interface decreases rapidly until a direct interface between sediment and clear liauid is formed. Further sedimentation then results solely from a consolidation of the sediment, with liquid being forced upwards around the solids which are then forming a loose bed with particles in contact with one another. A number of attempts were made to predict the apparent settling velocities of a concentrated suspension. Robinson (1926) suggested a modification of Stokes' Law and used the density and viscosity of the suspension in place of the properties of the fluid. Steinour (194^), who studied the sedimentation of sm.all uniform particles, adopted a similar approach, using the viscosity of the fluid, the density of the suspension and a function of suspension voidage to account for the shape and size of the flow spaces. Hawksley (1950) also gave an ex- pression for the rate of sedimentation, based on the assumption that lateral forces produce a more or less uniform spacing in a horizontal plane, and that the particles were presumed to arrange themselves in such a way that they offered the minimum resistance to the upward flow of displaced fluid. Many investigations were reported in the last three decades and several theoretical, semitheoretical, and empirical correlations were suggested. Most of the correlations could predict the sedimentation velocities over a small range of Reynolds number and the predicted velocities showed fair agreement with experimental results. Jottrand (1952) developed a simple empirical correlation which could predict sedimentation velocities for particle Reynolds number up to 0-4 but showed significant lower velocities at all voidage. Lewis and Bower- man (1952) gave two simple empirical equations, one over the particle Reynolds number range between 2 and 500 and the other for the particle Reynolds number greater than 500. Both equations showed ooor agreement with experimental results. Until the work of Richardson and Zaki (1954), no correlation was available to predict sedimentation and fluidization velocities over an entire range of Reynolds number. They carried out an extensive investi- gation on sedimentation and fluidization of various sizes of equisized particles using several suspending media. Unlike Robinson (1926), Steinour (1944), and Hawksley (1950) who used the density and viscosity of the suspension, Richardson and Zaki used the density and viscosity of the suspending fluid in their correlation. The latter consists of six equations which cover the entire range of Reynolds number. These equations are perhaps the most widely used and predict the correct trend in behavior for all flow regimes. They do, however, predict signifi- cantly higher velocities for all flow regimes and suffer from discontinuities at the transition Reynolds numbers. Happel (1958) developed a mathematical treatment on the basis that two concentric spheres can serve as a model for a random assemblage of spheres moving relative to a fluid. The inner sphere comprises one of the particles in the assemblage and the outer sphere consists of a fluid envelope with a "free surface". Velocities predicted from this model were substantially low at low particle concentrations. A theoretical relationship between the concentration and the sedimentation velocity of a suspension of particles was developed independently by Maude and Whitemore (1958) and by Zuber (1964). Similar but somewhat nore comolex equations for the low Reynolds number range were suggested by Loef^ler and Ruth (1959) and by Oliver (1961). Wen and Yu (1956) suggested a correlation which could predict velocities over the Reynolds number 4 range between 0.01 and 10 . At high concentrations, predicted velocities were significantly high at all values of Reynolds number, and at low concentrations, they were high at low Reynolds number and low at high Reynolds number. Lapidus and Elgin (1957), Struve et al. (1958) and Price et al. (1959) carried out an extensive investigation on countercurrent operations, cocurrent operations and fluidization. They showed that the data for cocurrent and countercurrent flow were in excellent agreement with the operating diagram determined from the holdupslip velocity (relative velocity between solid and fluid) relationship obtained from the batch fluidization experiments. They concluded that all those systems in which relative motion takes place between fluid and suspended solids showed similar operating behavior. Barnea and Mizrahi (1973) collected published experim.ental data from twelve different sources and developed a correlation for predicting sedimentation velocities in all flow regimes. Although it was complex and difficult to use, the Barnea-Mizrahi correlation gave the lowest error compared with all previous correlations. Letan (1974) extended the work of Zuber (1964) for higher Reynolds numbers and developed a semitheoretical correlation which showed good agreement with the Richardson and Zaki correlation. The latter, as mentioned earlier, predicts significantly higher velocities at all Reynolds number. On the other hand, the equation developed by Wen and Fan (1974) predicts consistently low values of velocities at all voidages. Bedford and Hill (1976) provided a theoretical justification for the use of the slip velocity in transient one-dimensional particulate 8 sedimentation. This fact has not always been recognized in previous sedimentation investigation and, even when recognized, has often been included empirically. Garside and Al-Dibouni (1977) collected a large number of published data from various sources and developed a correlation based on a logistic curve, which, in their case, is a Graphical presentation of slip velocity versus Reynolds number with voidage as a parameter. Their correlation showed a significant improvement over the Richardson-Zaki equation and it is also easier to use than the Barnea-Mizrahi equation. Velocities predicted from the Garside- Dibouni correlation showed an absolute average error of 8.5%, which is the lowest among all correlations. In practice, most solid-fluid operations involve solid particles of various sizes. In the limit, any suspension can be regarded as being composed of a very large number of closely-sized fractions. In order to develop a sedimentation correlation for such systems, mixed suspensions consisting of two or more distinct measurable particle sizes rather than a spectrum of sizes could be used. No systematic study of such systems had apparently been made until the work of Hoffman et al. (1960), who studied the fluidization of binary and ternary mixtures of glass beads. They observed sharp size segregation in bi- nary mixtures of size ratios from 1.58 to 2.23 and partial segregation in mixtures of size ratio 1.24. Smith (1965, 1966, 1967) presented a theoretical analysis on sedimentation of binary particle mixtures in very slow flow. He developed a physical model for the sedimentation of binary mixtures by extending the spherical f^uid envelope model develooed by Happel (1958) for the sedimentation of single size particles. The settling velocities predicted from this model were, however, significantly lower compared with the experimental results. Using glass spheres (regular shape) and quartz powder (irregular shape), Davies (1968) provided an extensive set of experimental data on the sedimentation of binary and ternary particle mixtures. In an at- tempt to reproduce Davies' experiments, Lockett and Al-Habbooby (1973) found that it was not possible to achieve an initially uniform suspension due to the high settling rates encountered. As a result, Davies' experimental data on the sedimentation of binary and ternary particle mixtures appear to be suspect. Finkelstein et al. (1971) carried out an experimental study on mixed sized particles in a fluidized continuous system. Their studies illustrated the validity of the unique characteristic holdup-slip velocity relationship formulated by Lapidus and Elgin (1957) for systems of single-size particles. Lockett and Al-Habbooby (1973) carried out an extensive experimental study on countercurrent opev^ations and sedim.entation of binary particle mixtures of two distinct particle sizes. Using the Richardson and Zaki equation (1954) for single-size particles, they developed a physical model for sedimentation of binary mixtures assuming that a particle settles only according to the local voidage fraction areund it, irrespective of whether its neighbors are particles of the same or of another size and whether they are moving relative to it or not. The model showed good agreement with the experimental results for the initial settling velocities. Since initial settling velocities are considerably higher than average settling velocities, their correlation cannot be used to predict average settling velocities. Recently '*irza 10 Richardson (1979) carried out an experimental study on the sedimentation of suspensions of two distinct particle sizes. Their experimental sedimentation velocities were compared with those predicted from the Lockett and Al-Habbooby model. It was found that the predicted velocities for both interfaces were overpredicted in almost all cases by between 5 and 50%. Introducing a correction factor into the Lockett and Al-Habbooby equations, they were able to match their experimental and predicted results to within j;^ 10%. Clearly, such an approach is highly empirical and cannot be applied over a wider range of conditions than those used in their experimental work. This being the case, Mirza and Richardson suggested that a further investigation on the sedimentation of binary suspensions is required. The purpose of the present study is fourfold: (1) to develop a physical model for the sedimentation of binary suspensions of two distinct particle sizes, (2) to compare the model with experimental results obtained here and elsewhere, (3) to extend the model to suspensions of multisized particles (suspensions with three or m.ore distinct particle sizes) and to test the model developed against experimental results, (4) to extend the analysis to systems with a continuous particle size distribution. CHAPTER II SEDIMENTATION OF EQUISIZED PARTICLES There are wide discrepancies between the prediction of the many published correlations for the velocity-voidage relationship observed during fluidization and sedimentation in solid-liquid systems. Barnea and Mizrahi (1973) have discussed some of the difficulties involved in describing the characteristics of multiparticulate systems and in particular in consolidating into one correlation results from batch and continuous sedimentation, fluidization and co- or counter-cur-rent vertical two-phase flow. In analyzing sedimentation operations, most investigators have concentrated on developing expressions for the settling velocity of the particles in terms of the properties of the particles and fluid and the volumetric concentration of particles. Not all, however, have recognized the following fundamental features: Hydrostatic Effect: Richardson and Meikle (1961) have shown ex- perimentally that the effective buoyancy force acting on an individual particle in a suspension is greate>^ than the buoyancy force exerted by the fluid alone. If p , p^ and p represent the densities of particles, fluid and suspension, respectively, the following expression for the driving force may be written Driving force per unit mass of ^ /, _ ^ ^ particle in the fluid alone -^^''p ' ' f^ (? ]) \ - i Driving force per unit mass of / \ -c(o-c) particle in the suspension -^ p s = gc(o_. - p j P 11 I (2.2) 12 where p can be expressed as P3 = Pp(l-£) + p^z (2.3) £, the voidage, is the volume fraction of the fluid in the suspension. It is clear then that the density of the suspension p , should be used in the expression for the driving force in particulate sedimentation. However, as Equation (2.2) indicates the final expression for the driving force in a suspension is related to (p - p^). It is instruc- tive to note that the dynamic pressure relevant to the inertial forces, should still be related to the fluid density p^, since the particles are moving all together relative to the fluid and not relative to the suspension. Thus p^ should be used in the expression of the Reynolds number and the drag force. Wall Hindrance: While it is reasonable to neglect wall effects in large-scale equipment, it might not be acceptable in laboratoryscale equipment. Significant wall effects are detectable even when a single particle is settling in a vessel whose diameter is larger than the particle size by one or two orders of magnitude. Since the settling velocity of a particle in suspension is related to the terminal free-falling velocity, wall effects in a sedimenting suspension can be taken into account by applying wall correction to the terminal velocity of the free-settling particle. Slip Velocity: In particulate sedimentation, particle drag is governed by the velocity of the particle relative to the ""luid, U , and not by U , the velocity of the particle relative to the container. While this effect may not be important in a very dilute suspension 13 due to the low velocity of the displaced fluid, it is highly significant in concentrated suspensions. This fact has not always been re- cognized in previous sedimentation research and, even when recognized, has often been included empirically. The extensive study of Meters and Rhodes (1955) and the work of Lapidus and Elgin (1957) have shown experimentally that only U^ can be used to compare and correlate data in sol id-fluid operations. A theoretical justification for the use of the slip velocity U was eventually given by Bedford and Hill (1976) Consider the sedimentation under gravity of solids particles of uniform composition, size, and shape through a fluid held in a container of constant cross section. The downward direction is indicated by the positive x axis. Regarding the particles and fluid as two superimposed continua, in the absence of reaction, dissolution, or other mass transfer processes, each constituent must staisfy the usual one-dimensional conservation of mass equation ^'h%V-^ (2.4) ^-y^fV-° (2.5) where p = mass of solids per unit volume of suspension P P^ = mass of fluid per unit volume of suspension U c = velocity of particles relative to the container U^ = velocity of fluid relative to the container The partial densities p and Cr can be expressed in terms of the den- sities of the solid and fluid and the volumetric concentration as 14 where p ^p = CPp (2.6) Pf = (l-C)p^ = cp^ (2.7) = density of the solids p^ = density of the fluid C = volumetric concentration of particles e = voidage Substitution of Equations (2.6) and (2.7) into Equations (2.4) and (2.5) gives h^%^' h^%V - ° (2.8) |t(£Pf) + |^(£PfUf) = 0 (2.9) Assuming solids and liquids to be incompressible. Equations (2.8) and (2.9) can be written as i^ll(%> = ° (2-10) || + |^(.V (2.11) =0 For a single particle moving in a suspension, the forces acting on the particle in the x direction are the drag, gravity and buoyancy. Let f. be the drag force which is a function of the relative velocity between the particle and fluid U -U^, and the concentration of the suspension C. Then the net force F on the particle can be expressed as F = -fd(C, U^-U^) + p Vg - p^Vg, (2.12) 15 where p s = density of suspension which is (p p + 6^) f V = volume of a single p a r t i c l e g = gravitational acceleration The force on the p a r t i c l e s per u n i t volume of mixture is ( Y ) F . Applying Newton's Second Law to p a r t i c l e s in a unit volume of suspension, PpSp = - ( v ) f d ' C , U^-U^) + Cg(pp - p^) (2.13) where the particle acceleration a is given by 9U 3U Equations (2.10), (2.11) and (2.13) with Equation (2.14) provide three equations in the three variables C, U and U^. E is not an independent variable as e = 1-C. Here note that Adding Equation (2.10) and (2.11) |^(C.e).|3^(CU^.cU,) = 0 (2.15) Since C + e = 1, the first term on the left-hand side of Equation (2.15) drops out giving |-(CU + cU.) = 0 3x c f (2.16) The term (CU + eUr) is the volume of the mixture passing a point x per unit time per unit area. For sedimentation, this volume is zero, so that Equation (2.16) yields ' - - ^ ^ (2.17) 16 Using Equation (2.17) to eliminate U^ from Equation (2.13) and using Equation (2.14) we get 3U^ 5U ^ P ^ a r ^^cW-^ = -(v'^d'C' ^c 'iTc^c^ ^ ^9(P. - 0^) (2.18) Simplifying using Equation (2.2), we obtain 9U 3U ^p^9t- ^^cW-^ , = - Y ^d(C' T ^ ^^c' ^ 9(Pp - P^)(l-C) (2.10) Equations (2.10) and (2.19) have two unknowns U and C, Getting f , experimentally or empirically, these two equations can be solved tor transient sedimentation problems. 3C For steady sedimentation ^ becomes zero and Equation (2.10) leads to dU dx ^ = 0 (2.20) Substituting Equation (2.20) into Equation (2.19) and noting that aU /8t = 0, we get ^df^' 1 ^ ^c' ^ '^9(PD - P^)(l-C) (2.21) This merely indicates that the draa and buoyance ^orces are balanced by the weight. This equation clearly indicates that particle dnaq in suspension is governed by the oarticle velocity relative ^.o the fluid U, - ^f-T^^ (2.22) Use of this slip velocity is essential when various sol id-liquid operations are to be out on a common base. 17 Dimensional Analysis Dimensional analysis is of prime importance in any transport operation. It reduces a large number of independent variables to a few number of dimensionless groups. Dimensional analysis of sedimentation was first studied by Richardson and Zaki (1954). The following assump- tions are common in the study of sedimentation of equisized spherical particles: (1) The suspension consists of spherical particles, of common den- sity, with a relatively narrow size distribution which can be reduced to an average characteristic size. (2) There are no interactions of any kind between the particles, except hydrodynamic effects through the fluid. Flocculation and aggrega- tion are assumed not to be present. (3) The relative positions of the particles in the suspension are completely random, without any segregation. This assumption comes closer to reality with narrow size distributions and more concentrated suspensions. The drag R. per unit projected area of a spherical particle settling at its terminal falling velocity U^^ is a function of the density and viscosity of the fluid, the diameter of the particle d, the terminal falling velocity U. , and the ratio of the particle diameter d to the container diameter D. R^ = Thus, f(p^, y^, U^^, d, ^) For the isolated spherical particle. (2.23) 18 RjTr/4)d^ = (7T/6)d^(Pp - p^)g (2.24) For the sedimentation of a particle in a suspension, R. will alter to R| because the upthrust is equal to the weight of displaced suspension. Thus R[(Tr/4)d2 = (TT/6)d^(Pp - p^)g 3 = (7T/6)d £(p - p^)g (2.25) From Equations (2.24) and (2.25) it follows that R; = cR^ (2.26) R' is a function of the v e l o c i t y of p a r t i c l e r e l a t i v e to f l u i d U , and is also a function of the voidage, e, which determines the flow patterns and the area available f o r the flow of the displaced f l u i d . K f(p.f:> Ufy U^, d , £ , ^) Thus, (2.27) Equations (2.23) and (2.27) may be rearranged to give the f o l l o w i n g exp l i c i t equations f o r U. U t^' U^ = and U f(R^, P^> y.p» d, ^) (2.28) f ( R [ , p^, y^, d, e, -^) (2.29) Since R' = sR., Equation (2.29) may be written as ^s ^ ^'^t' ^f ^f ^' ^' D^ Dividing Equation (2.30) by Equation ( 2 . 2 8 ) , we ootain (2.30) 19 u, u (2.31) f ( R ^ , P^, 1-1^, d , £ , -^) t°° Since U^/U^^ is dimensionless, the right-hand side of Equation (2.31) should also be a function of dimensionless groups. The quantities R. , P.p, y^ and d can be arranged to form the dimensionless group 2 2 R^cl p^/y^ . Therefore U. = f ( ^ y ^ , c, §) U (2.32) t^' Now, R^d2p. -)(^^-^)2 ( (2.33) y^r P-pU? f t°o y. Here R./p^U. U. d p - ^ R, is a resistance c o e f f i c i e n t which is a unique function of the Reynolds number U. dp^/y^ f o r a spherical p a r t i c l e . U. = Ut«' U^ cip. , f(- t"' f ' ^' d D' y. Thus (2.34) The above derivation does not take into account the nature of the flow past the particles in suspension. Further simplification can be obtained by considering various flow regimes. The Viscous Regime: In the Stokes' Law range the relative velocity is sufficently low for inertial effects to be negligible so that the whole resistance may be attributed to skin friction. R: t oc y u d ^f s Under this condition, 20 R'^ being independent of p^, the density of the fluid, Equation (2.31) becomes U^ = f(yf, d, R^, e, ^) (2.35) Here y^, d, and R^ cannot be arranged to form a dimensionless group and therefore U^/U^^ must be independent of these variables under condition of streamline flow. This leads to i r - = f(e,~) t°o The Turbulent Regime: (2.36) At high velocities the effect of the inertia of the fluid becomes important and the viscous forces are negligible. Under this condition Newton's law is applicable and R; - PfU^^ (2.37) The resistance per unit area is thus independent of the viscosity of the fluid and the diameter of the particle except insofar as the latter influences the effect of the wall. Equation (2.31) therefore becomes ^s d (/- = f(Pf, R^, e, §) (2.38) Again p^ and R. cannot be arranged to form a dimensionless group and therefore ^s U^= d f{e, §) (2.39) 21 Thus, when either skin friction or form drag is predominant, U /U. is a function of e and -^ as shown in Equations (2.36) and (2.39), respectively When both drag and skin resistances are of comparable magnitude, the Reynolds group is significant in addition to e and -^ as indicated in Equation (2.34). In sedimentation experiments the velocity of the particles relative to the container U is measured. This velocity is related to the slip velocity according to the expression U^ = eU^ (2.40) Substitution of U in place of U into Equation (2.34) gives U, U. dp. t°o . f Experimental results on sedimentation showed a linear relationsnip between log U and log e. A typical plot is illustrated in Figure 2.1 which shows a straight line with a slope n and intercept U. corresponding to infinite dilution (£=1.0). Thus log U^ = n log £ + log U. (2.42) or U^ = £" (2.43) ^• In Equation (2.43) U- is the sedimentation velocity of suspension in a tube of diameter D at infinite dilution. Thus U^ can be taken as a terminal falling velocity of a particle in a tube. Therefore, — 1 1 : O ' 1 O7 i1 ' ' 1i t t t 1 u o o d 1 r i i O1 y 1 1 1 •'! 1 i ' 1 ' i ^ X1 y^ : ' ' i IX 1 : : 1 M O "^ X i ^ 1 ! 1 1 i ' ^:• :1 \ -O i* -O 22 -C 2 -OI8 -C i<3 -0"S -Oi2 -O O C Zs-OCb- ZZ'*-'^Cl Log £ 'igure 2 . 1 . Relation between rata of sedir';entation and voidage of suspensions. (2.ia) f(§) 'J. 00 From Equations ( 2 . 4 1 ) , (2.43) and (2.44) n .>'t-^^f _ d. (2 ^S'' Noting that n is independent of £, '^.- 11 00 n = f( d^ i..^D •' D^ •f In the viscous and turbulent "'lew regimes, U^/U^^^ 'n —uaticn i_.-:! ::ecomes indepencent of Reynolds number 'J, (2.46) simplifies to ' - s/ -. As a r e s u l t Ecuaficn n = f(^) (2.47) This behavior has been observed by most investigators in t h e i r e x p e r i mental work. For low and high Reynolds numbers, n was found to be func- t i o n of -p as shown in Equation (2.47). For intermediate Reynolds numbers i t was found to be a function of Reynolds number and ^ as shown in Equation (2.46). In some instances, however, i t was found that n varies i n - s i g n i f i c a n t l y with -p so t h a t n becomes independent of ^ . Sedimentation v e l o c i t i e s may be predicted from Equation (2.43) in conjunction with Equation (2.46) or (2-47). The v e l o c i t y U. in Equation (2.43) may be obtained from the terminal f a l l i n g v e l o c i t y U. a f t e r applying awal1 cor- r e c t i on. Published Correlations There are many correlations in the l i t e r a t u r e that aim to predict the sedimentation or f l u i d i z a t i o n v e l o c i t i e s in s o l i d - f l u i d systems. The m a j o r i t y of these apply over a r e s t r i c t e d range of Reynolds number although some have been suggested as being s a t i s f a c t o r y f o r a l l regimes. Table 2 . 1 . flow Most of the published correlations have been summarized in Predicted v e l o c i t i e s by most of these correlations have been compared by Garside and Al-Dibouni (1977). The correlations suggested by Richardson and Zaki (1954), Barnea and Mizrahi (1973), and Garside and Al-Dibouni (1977) are by f a r the most important. Richardson and Zaki (1954) were the f i r s t to develop a c o r r e l a t i o n which was able to predict sedimentation and f l u i d i z a t i o n v e l o c i t i e s in a l l flow regimes. They c a r r i e d out sedimentation and f l u i d i z a t i o n exper- iments in various suspending media using d i f f e r e n t - d i a m e t e r tubes, and — 3 01 tu I o to o at o i I ^1 u <w u .^ .^J — . «.* Ol IB u — — >a *J u V. k o E 01 o a; u •a u u — 01 ... >. u .^ >o ..-1 o I. *J Ot i. .c o o .9— 01 JZ. t— Of <u u o u •o u o o 01 •— v> l/l ^— •fl o 0< u u .*I. •^ k .c • — • • . ~ •^ .." k .^ a. l/> UJ =- at •« u "^ *J 01 k ^" >o U L. ..— B Si ^IQ .^«> 0; 01 f 4.J *J H- * r^ 1. 01 *J a. L^ LLJ i- / o <n o O m m CO a. o vl u eo o o V u O O V CVI at OJ O Of V o o M 01 QC I o o — CM ot 8 V o o o V V at o to o O K V o o O V evj ^ 8 CM — « 8 9 ec o m o V Q: A A 8 V CM 0) o: (VI o • V V O o at CM ce — V in O 01 r— . V | A 01 01 o e • — QC 01 a: V CM CM V V v | 0 o — «»o ee >— <vj O O X en at ^ O tn o. ^ o• V V V 01 ot ex oc at a£ ^— •e01 a oc V V V V at tn in oc n1 o u> . ^- r— o ..^ o V ^^ o o• ae V •— V a; A 03 r— .... ^^ <o o CM o CO in CO SO CM O O o o o — o I 8> at o O 01 o CM 00 .— o} >— + + + at S- 001 so I ^«, (it 1 n c 1 o \r% . o» CM r o 1 1 -o 1... >.^ Q. a. X </1 • O at X at ••• •I J3 m «o —>-- 8 u 3 *.* w m • z ^— u ^^u 1 «.> o<.• ^^ O \o u *J 3 =3 a\ CM r,. o. o. c X — en u at on in . — < • - CM C CM tn >» CM at ON ce. 5 e OC VO 01 CM I J . «• Ot |ac ^^ 'in ^ '-r' _. 8 o ** n ^• ce. in O CO o vO 01 C£ o Ol N 3 a CO I3 I/) 3 o </) j/1 o «/» 3 i/» l/> (J o •9O (J f-1 O ^- CM OJ -Q rt3 CM tn CO m m a» ^- " ' • ^ 01 1. rsi o» o» ^ in 91 CM at cr> o m .— o 1. O l/l c ot B CO X at as T3 01 l/l C u iQ X o -3 I TO C 00 •a o lO l/l •o L. c i/i at _J u *-"^ 00 ^.* .^. JZ kn X a\ r^ ^•m^ ^^ at a. Q. •a 3: •o c •a Ol •C3 3 £ ut •eu 10 00 t«Ot K ^ u r,. CM .c <?;^ in o 00 I 8 a* o 10 CM 3E at i. at £i CM ^.. o\ V <j «n • o in «o X en a: c o +-> «o CM I 8 at </t => u OUI I 5 .— ^ CT» P-) r>» Ol 1 m a* ^.c *1— 0* Of 10 <T> ^.^ 3 •o c ....^ V0 ON « k ot ^^ 1*. lb. 01 o .~J ^ ^ H . i. 01 > .—..o ..-^ ^• v0 ... ^-.* 3 o> >- . o * ^ •o ^— C b. at >a 3 0) 3 .a »>.l c ^^— .f— ^-. ^ o> ^ • <a u N .^ z •a c •a <a at c L. m CO ...... TT r^ Ol ^— ..—.. c n 4.J 91 _J - ^ la u. •o c (Q C at 3 o <o cpCM >a o n c( 25 various spherical particles of different size and density. Id.':-^' set of data when plotted as logU^versus log £p qave a straight ''ne wit.' slope n. Tne exper''mental values of n were plotted against ^ witn Reynolds number as a parameter as shown in -igure 2.2. A single curve represents the data for all values of Reynolds numbers less than 0.2. The figure also shows another sinqie curve ~jr all values of Reynolds numbers greater than 500. "or in^^ermediate values 3^ Reynolds number, n was found to be a function of Re^. is in agreement with Equations (2.46) and (2.^7). This analysis From analysis o* t'^e curves, the following equations were obtained n = 4.55 + 19.5 ^ 0 < Re < 0.2 X) 0.2 < Re < 1.0 n = (4-35-17.5 ^)Re;'^-^^ (2.^9) 1.0 < Re < 200 n = (4-45 + 18 ^)Re;^-^ V- •0) !-4 :oi :c4 d/3 ^icure 2.2. S>pe n as a ^"unc'icn r/1 26 200 < Re^ < 500 n = 4.45 Re^^'^^ (2.51) Re^ > 500 n = 2.4 (2.52) These values of n may be used in Equation (2-43) to predict the settling velocity. This set of equations is perhaps the most widely used and predicts the correct trend in behavior for all flow regimes. However, predicted velocities are consistently high and show an absolute average deviation of about 20%. Also, the predicted relationship between velocities and Reynolds number shows discontinuities at Reynolds numbers of 0.2, 1.0, 200 and 500. Barnea and Mizrahi (1973) carried out an extensive investigation on sedimentation and fluidization. It was found that the transition points between the laminar and intermediate and between the intermediate and turbulent regions were dependent not only on Reynolds number but also on the concentration of the particles. Poth effects were taken into account by introducing a modified Reynolds number Re^ which includes the slip velocity U and viscosity of the suspension y^. Experimental data were collected from twelve different sources and were plotted as (Re^/Cp,^)^'^^ versus (Re^^C^^)^'^^, where Re, = ^ _ 1 = ReJ-__^|^-^) •(D U3 and 4d(p * - p.)g 3y^ U. .r 1+C (2.53) 27 The plot is shown in Figure 2.3. For convenience this relationship was also made available in an algebraic form. For a single particle moving in a fluid C^ was found to be Cn ^ = (0.63 +-~±^)^ / R ooe ^ (2.55) By analogy, a similar equation was obtained from the curve for multiparticle systems. This equation may be written as Cnd, = from which U^/U. Re or U (0-63 + - M ) ^ /R^ (2.56) can be obtained algebraically as a function of C and as a function of p^, u^, p , d and C. This correlation predicts better results than the Richardson and Zaki equations but it is complex and difficult to use. regimes although the predicted values of U It covers all flow are consistently low for Reynolds numbers less than about 300 and consistently high for Reynolds number above this value. This trend can be seen in Figure 2.3. Garside and Al-Dibouni (1977) collected a large number of published experimental data and plotted them as Reynolds number Re = (dU.po/iJ^) against U /U. with the voidage £ as a parameter. Here U. is the terminal falling velocity of a single particle in a finite fluid medium. The curves obtained had the shape of a 'logistic' or growth curve (Wolfenden, 1942) as shown in Figure 2.4. These curves may be described by the equation U -^ - A ^t -^—[j- '•i = K Re- (2.57) 23 •r1 .r ' 5 , ^ ' v^»^-v." tr^ y 'J* i7i <m o / ' ^ W O) 7 /' I .L ZD I A 10 0.1 . >^V.y "^ .^ — .'^ •aur? J. .^enerai cor^-eia'ion "c tne re i a t i ve ve ' oc: -.y 1igui a systems, basec on extended ce-'ni:^:cns ana Cr>,. 29 .J ' (/) Re "igure 2.4. A growth curve showing 'Relationship between 'J /'J s' ^t and Re at several values of £. A and B are the asymtotic values of U /O^ at low and high /alues of Re, r e s p e c t i v e l y , while the position and rata of increase of U^/0, in the in termediate region were determined by the constants .< and : . ~ne oest values of A, B, K and z were deterTiined for voidages of 3.5, 0.-:, 0 . ^ , C.3, 0.9 and 0.95. a function of £. Equations were then develooec for - , 3, K anc : ss These equations iv.ay be l i s t s o as "" 3 = ^ 4.14 ~ . J ^ 0. 8£ 9 n^ '?'=l c o f o r -: <_ 0.35 f o r £ > 0.35 (' \ - 55) 30 K = 0.06 (2.61) z = £ + 0.2 (2.62) Substituting K and z into Equation (2.57), we get U. r- - A - ^ p = 0.06 Re^"0-2 (2.63) When compared with the experimental results, Equation (2.63) showed an absolute average deviation of 8.6%. Since U^^/U^ = £ is perhaps the most widely used equation for predicting veolcities of solids in fluidization and sedimentation, an equation for predicting values of n as a function of Re was also developed. This equation may be written as ^'1 2 y = 0.1 Re°-^ (2.6^) where 5.1 and 2.7 v/ere the asymtotic values of n at low and high values of Re, respectively. Although Equation (2.64) represents ex- perimental results with an accuracy comparable to that achieved with Equation (2.63), the behavior at high voidage (greater than about 0.9) is not well represented. Compared to all correlations. Equations (2.63) and {2.6^) 'vere found to be the most accurate and are, therefore, recommended for use in further investigation on sedimentation and fluidization. CHAPTER III SEDIMENTATION OF MULTISIZED PARTICLES There seems a little prospect at the present time of dealing with the hydrodynamics of binary particle-liquid suspensions in a fundamental way. Even when the particles are all identical little progress has been made; except for the very slow flow regime, and for higher values of Reynolds number we are dependent upon empirical correlations such as that of Richardson and Zaki or of Garside and Al-Dibouni. It is therefore unrealistic to hope for an analytic solution to the problem for a binary particle mixture except perhaps for the very slow flow regime and this has been dealt with by Smith (1965, 1966, 1967) with a little success. The problem becomes even more complex when more than two sizes of particles are sedimenting in a suspension. Segregation of particles has been observed during the sedimentation of suspension of two or more sizes cf particles. Distinct sedi- menting zones are formed due to complete segregation of particles when the size ratio of the two closest sizes of particles is greater than 1.6. Partial segregation without distinct zone formation has been observed even at a particle size ratio of 1.19. When a uniform suspension of two distinct sizes of particles starts settling, the segregation gives rise to tv;o sedimenting zones; a lower zone in which particles of both sizes are settling and an upper zone in which only smaller size particles are settling (see Figure 3.1). Sedim.entation of a suspension of three sizes of particles gives rise to three distinct zones of sedimentation, the lowest where all the three sizes of 31 32 E <v "=i -o <v oo E o CO E CD C CO E c/1 CD OJ C N •r— •^ e (/) sro •1— <X3r^ -M 1 — CO E fO O) O E p— U CO (J S_ -1- ^3 Z3 <V CT C_J 03 +-> o ^vl I— O C Ol <v 1>^-M .sz E— SfC E o C". c/^ E CD •r— 1 — o. U r <zr> 1 — cr.^ <D <V e •^- fT3 r— c E U •r— in n3 p ^ +J r— CO C fC O) -l-J fO +-> -o sE E fC O <o Q . •1^ T3 S- -rfT3 Z3 <V c r r— "r— O I— CO N •^ CO c c •^- 1 — u 1 — u OJ > ) 4-) c; -a CD CO -a • I — o E CO O rvi +-> • 1 — -(-> +J CD sCO <T3 £2. CD 4-> CD <D CT) 1 — i. C 03 1 c — O O 03 -M E CD e o s- 03 Q. u QJ CO OJ ^-/. O) E S- N fO • 1 — o ^sj p — CO +j 4-) E CD CO CJ) CD E E OJ •r- O 00 E SO M- M cr E c: %C 4 I CD C o Ni "O E 4-) C E • , - OJ -M E 03 fO • - - O cr-T3 CD CD s- -c en CD OO E O s 03 cm CD S- cr CD OO '^ E o o r03 •rJ-> •-E E CD (_) E O CJ 33 particles are settling, the middle where medium and small sizes of particles are settling, and the top where only small particles are settling. The samie physical phenomenon occurs in a suspension of m sizes of particles, segregating and giving rise to m sedimenting zones Interpretation of experimental behavior on sedimentation of binary particle mixtures was first suggested by Lockett and Al-Habbooby (1973). They developed a physical model which is based on the assump- tion that a particle settles only according to the local voidage fraction around it, irrespective of whether its neighbors are particles of the same or of other size and whether they are moving relative to it or not. In developing the model, they used the equation U^ = U ^ d - O " (3-1) which was given by Richardson and Zaki (1954) for uniform spherical particles. Rather than using the equations developed by Richardson and Zaki for predicting the exponent n in Equation (3.1), Lockett and Al-Habbooby measured n experimentally for each size of particles. Mirza and Richardson (1979) used the sam.e model for binary systems and extended it to multisized particle systems. They used Equation (3.1) in their model with n values given by the Richardson and Zaki equations. Two remarks should be mentioned regarding this m.odel. These are: (1) Values of the exponent n in Equation (3.1) can be predicted using generalized correlations. Instead, Lockett and Al-Habbooby measured n values experimentally as -^.64, 5.41, 5.14 and 5.07 for 34 four different sizes. Corresponding values given by the Richardson and Zaki equations are 4.75, 4-77, 4.82, and 4.92, respectively. It has been reported by several investigators that values for the exponent n remain essentially constant rather than varying drastically as measured by Lockett and Al-Habbooby. In general, it is not recommended to mea- sure n values by taking few experimental observations since reliable correlations are available for their prediction. Mirza and Richard- son predicted n values using Richardson and Zaki (1954) equation which predicts consistently lower values of n and thus considerably higher values of settling velocities are obtained. (2) The assumption that the settling velocities of larger par- ticles are unaffected by the sizes and velocities of surrounding particles is physically unacceptable and results in higher settling velocities of the larger particles. It has been observed when a large particle is settling in a suspension of smaller size particles that it is displacing not only a fluid but also the smaller size particles. Thus sedimentation velocities of large particles in a multisized particle system are lower compared to their velocities in a suspension of their own kind at the same total concentration. It is clear, therefore, that inter-particle interactions cannot be ignored in multisized particle systems. These interactions may be accounted for by taking into consideration the buoyancy effect of all particles with sizes less than i on the terminal falling velocity of particles with size i. It is instructive to note that the use of Equation (3.1) implies that the settling rates are not influenced by interparticle collisions which become increasingly important as the velocities of the particles increase. 35 Sedim.entation of Suspensions of Uniform Particles The settling velocity of particles in a suspension of uniform spherical particles is given by Equation (3.1). The exponent n may be predicted from the equation given by Garside and Al-Dibouni (1977) ~ ^ = 0.1 Re°-^ (3.2) d U p. where Re = The upward velocity of the fluid U^ can be calculated by equating the flow rate of solids and the flow rate of fluid at any cross section of the settling zone. This gives U^£ = U^C (3.3) U,C U. = - ^ (3.4) or f £ The r e l a t i v e v e l o c i t y between s o l i d and l i q u i d U^ is given as U = U + U. s e t (3.5) Substituting for U. from Equation (3.4) into Equation (3.5) and simplifying, U^ U = -^ S (3.6) £ Combining Equations (3.1) and (3.6), we obtain n-1 U^ = U^(l-C)' = U/"^ (3.7) 36 Use of Equation (3.7) is required in the analysis of systems with two or more sizes of particles. Sedimentation of Particles in a Binary Suspension In general, a suspension of two distinct sizes of particles will give rise to four zones during the course of sedimentation (see Figure 3.1). From the top downwards these will consist of: clear licuid, suspension of smaller particles, suspension of particles of both sizes with concentration equal to initial concentration, and finally, sediment. Consider sedim.entation in the zone containing both sizes of particles. Let suffixes S and L apply to the small and large particles respectively, and suffixes 1 and 2 apply to upper and lower sedimenting zones. In order to account for particle interactions in the lower sedimenting zone, the terminal falling velocity for the larger particles must be calculated as if they were settling in a suspension of smaller size particles. In the Stokes' Law range the terminal falling velocity of a large particle U^^^j_^2 ""^ S^ven as U, , , =-Ko^r where o ^ (3-8) is the density of a suspension consisting only of the small s size particles. The density p^ may be written as 37 The terminal falling velocity for the small size particles U^ c- ^ may be written as U _ ^S (Pp - Pf>9 t-,S,2 18 y. (3.10) Equations (3.8) and (3.10) may be corrected for wall effect using equation given by Francis (1933). t,i,2 ^ t-,i,2 1-0.475 d./D -4 where i = L,S 1 - d./D (3.11) Equation (3.7) may be written for the small and large size particles in this zone to give ^s,L,2 " ^t,L,2 ^2 U \X^ (3.12) ^,2-^ s,S,2 U. . o t,S,2 (3.13) £o The n. ^ and n^^ ^ may be computed from Equation (3.2). In sedimenta- tion, the upward flow rate of liquid is equal to the downward flow rate of particles at any cross section. This gives ^f,2 ^2 " ^c,L,2 ^L,2 "^ ^c,S,2 ^S,2 (3.14) Using Equation (3.5), Equation (3.14) may be written as U f , 2 ^ 2 = (Us,L,2- U f , 2 ) ^ , 2 ^ (^^^3^2- Uf,2^S,2 (^l.S) Equation (3.15) may be rearranged to give ^f,2 Now, ^s,L,2 ^L,2 ^ ^s,S,2 ^S,2 (3.16) 38 ^c,L,2 = "s.L,2 - ^f,2 = "s.L,2 - ("s,L,2 \,2 ' Us,S.2 ^ . 2 ' = ^s,L,2(l - C L , 2 ) -^3,3,2^3,2 (^-H) Substituting Equations (3.12) and (3.13) into Equation (3.17) "c,L.2 = Ut,L,2 =2"''^ (1 - C L , 2 ) - 'Jt,S,2 ^ 2 " ' ' ' S,2 ^^-'^^ In the same manner, ^c,S,2 " ^s,S,2 " ^f,2 = ^'s,S,2 • (^'s,L,2 \ , 2 "" ^s,S,2 S , 2 ^ " ^"s,S,2'^ " ^S,2^ " ^s,L,2 ^L,2 = Ut,S,2 ^ 2 " ' ' ' (^ - S , 2 ^ - Ut,L,2 ^2"'''''^L,2 ^''''^ Since all the terms on the right-hand side of Equations (3.18) and (3.19) are known, U , ^ and U c; o can be calculated. Here, U , ^ corresponds to the observed rate of fall of the interface between the two sedimenting zones. The concentration of the particles in the upper zone is not directly known but can be calculated using a mass balance. The volume- tric rate at which the small particles pass from the lower zone to the upper zone ^'^ (U < 2 " ^c s 2^^S 2 '^t' ^''^^'^- "^t ^"^ cross sectional area of the container. The rate of increase in the volume of the upper zone is (U , o - U ^^ J A . . Thus the concentration of partiC,L,^ C,o,! C cles in upper zone C^ , is given as 39 '^c,L.2 - "c.S,2)S,2 - u c,S,l T (3.20) WZ: 'S,l Since this zone contains only uniform particles. Equation (3.1) can be directly applied to give U U c,S,l 'S,l = U. c,S,l t,S,l (1 -<^s.i) (3.21) where U^^j,, = U^,s,2 ^"d so nj^, = n^^^ Equations (3.20) and (3.21) can be solved simultaneously to calculate U 3 , and C^ ,. ^^ c i corresponds to the observed rate of fall of the interface separating the suspension and the clear liquid. Sedimentation of Suspensions of Multisized Particles Consider a suspension of m different size particles (a,b,...m, a being the smallest); this will give rise to M(=m) zones of settling suspensions, (1,2...M counting from top), with clear liquid above and a sediment layer at the bottom. All zones will exist at the beginning of the sedimentation (Figure 3.2) and each zone will disappear in turn during the sedimentation process as its upper boundary coincides with that of the sediment layer. The model developed for binary suspensions may be extended to the sedimentation of multisized particle systems. Three suffixes will be used for the particle velocity 'J; the first suffix indicates the kind of velocity (settling, slip, terminal), the second indicates the size of particle (size a, size b, ..., size m ) , and the third indicates the sedimenting zone in which the velocity is being 40 Settling velocities of interfaces Clear liquid U c,a,l Contains only the smallest size oarticles U c,b,2 U c,m-l,M-1 Zone M-1 U c, m, M Zone M • Contains particles of : sizes a,b,...m-1 Contains particles of all sizes a,b,...m Sedimen Figure 3.2. Formation of zones durina sedimentation of a multisized particle suspension. considered (zone 1, zone 2, ..., zone M ) . The notation for the concentration C has two suffixes, the first suffix indicates the size of the particle and the second indicates the zone in which this concentration is being considered. Velocity of Particles in the Lowest Zone (M) This zone contains all sizes of particles and their velocities can be written using Equation (3.7) ^'s,i,M = ^t,i,M ^M '" T = a, b, ..., m (3.22) As there is no material crossing the sediment zone, the upward flow rate of the liquid must equal the downward flow rate of the settling particles. Writing this in velocity terms ^c,a,M^a,M ^ ^c,b,M^b,M ^ • • • "^ ^c,m,M^m,M ^f,M ^M m = Z U . ^.C. ^, . ^ c,i,M 1,M (3.23) The slip velocity of any size particle may be written as "s.i.M^'^f.M^^Ci.M Substituting U i = a , b, ...,m (3.24) . ^ from Equation (3.24) into Equation (3.23) m U f . M ^ M ^ .? (Us,i,M- ^f,M'S,M 1 —a Equation (3.25) may be rearranged to give (2-25) 42 m ^f,M " .1^ ^s,i,MS-,M I —d m n • = Z U t,i,M ^M i=a M-1 (3.26) ^i,M Substitution of U^^.^^ from Equation (3.22) and U^ ^^ from Equation (3.26) into Equation (3.24) gives n • M-1 U U c,i,M £ 1 ,M ^t,i,M ^M m Z U t,j,M^M j=a n- M-1 C J,m J,M i = a, b, ..., m (3.27) Since concentrations in this zone are the same as the initial concentrations, all the terms on the right-hand side of Equation (3.27) are known and the settling velocity of each size particle can be calculated directly. U ,, corresponds to the rate of fall of the interface separating the zones M and M-1. However, it should be kept in mind that the terminal falling velocity of a particle of size i is calculated as if it were settling in a suspension consisting only of particles of sizes smaller than size i. Ut«3,i,M where p _ ^- g'Pp - Ps,i,M> 18 u, Accordingly we may write i = b, c. (3.28) , m • M is the density of a suspension consisting of all particle S , I ,n sizes smaller than size i in zone M, and may be written as i-1 m J=a -^' s,i,M m (1 - .^C ^ j^a ^^ ^) i = b, c , , m (3.29) ^3 For the smallest size particles (i=a), the terminal falling velocity may be written as d g(p - p_) ^t°o,a,M " 18 y. (3.30) Equations (3.28) and (3.30) must be corrected f o r wall e f f e c t s . WaT corrected terminal v e l o c i t i e s can be obtained by using the equation ^t,i,M ^ " ^'^'^^ ^ - / ^ '^ 11 — - [—T H~Tn ] ^t«',i,M ' " ^V^ i = a , b, . . . , m (3.31) Equation (3.27) in combinations with Equations (3.28), (3.29), (3.30), and (3.31) may be used to compute the settling velocities U • M of C , I ,11 all different size particles in zone M. Velocities of Particles in Zone M-1 The particles present in this zone are of sizes a, b, .... m-1. Concentrations and settling velocities of particles of all sizes present are unknown and must be calculated. The rate per unit area at which particles of size i cross the interface from zone M to zone M-1 (U M - U . .JC. ^ ^ c,m,M c,i,M' 1,M i = a, b, ..., m-1 (3.32) ^ ' The rate per unit area at which the volume of zone M-1 is increasing is clearly the difference in rate of fall of the interfaces forming this zone. This is, therefore, given by "c.m,M- Uc,m-1,M-1 (3-33) d^ Dividing the quantity in (3.30) by that in (3.31), we get the concentration C. ,, , as 1, M- I C ^i,M-i = ^^c,m,M ' ^ c , i , M ) ^ - , M Tu ~nr ^ c,m,M ,• _ , r m u ^ i ^ - ^^ b, ..., m-i (7 7A\ (3.34) c,m-l,M-1^ For t h i s zone, equations analogous to Equations (3.27) through (3.31) may be l i s t e d as follows n,- M 1-1 ^c,i,M-l = U t , i , M - l ^ M - l " m-1 "j,M-l' - .^ ' t , j , M - l ^ M - l J~a S,^M i = a, b, . . . , m-1 (3.35) ' ~- '^' ^ ' • • • ' ^ - ^ (2-26) ^- 9(Po " ^s i M-P ^too,i,M-l = s, i,M-1 18 M^ ' ' m-1 i-1 (12 C . . . T)P^ + P L -1=. j^,' M - r ^ f "^p .^^ j=a ^ j=a m-1 C... T J,M-1 i = b, c,. . . , M-1 ^ " j ^ i S\M-1 (3.37) d g(p - Pf) U.„ o - ^y . t " , a,, MM- l, = ^ - T18 U. . ,, , U 7-— too,1,M-1 (3.38) 1 - 0.^75 d / D - L 1 . d./D -4 ^ ' " ^ ' ^'' • • • ' " " ' ^-^-"^^ l' The terminal v e l o c i t i e s U, . .. ^ ( j = a, b, . .. , m-1) are f i r s t computed from Equations (3.36) and (3.38). required in Equation (3.36). The densities p • M •. are These are obtained from Equation (3.37) 45 An iterative scheme is, however, required since Equation (3.37) contains the unknown concentrations C. ., , (j = a, b, ..., m - 1 ) . The J j''-' iteration scheme proceeds as follows: (1) An initial set of concentrations C. |^_i (j = a, b, ..., m-1) is assumed and are used as starting values in Equation (3.37) to compute p^ . j^_, (i = b, c, ..., m - 1 ) . A reasonable guess for this set is the corresponding concentrations in the lower zone. (2) The terminal falling velocities U.^ . .,_-. (i = a, b, ...,m-l) are computed using Equations (3.36) and (3.38), and are then corrected for the wall effect using Equation (3.39). (3) Equations (3.34) and (3.36) are then solved sim.ultaneously for the concentrations C. .. . and the velocities U^ . .. , I , rl- I C , 1 ,r!- I (i = a, b, ..., m - 1 ) . (4) The values obtained in step (3) for the concentrations C. ^, -, (i = a, b, ..., m-1) are used in step (1) to recalculate I 5 I I"" I p . f^, -, (i = b, c, ... , m-1) and the iteration cycle is repeated. 5 , I , I 1— I In the calculation the iteration cycle is checked for completion by the requirement that (n+1) C. i,M-l (n) 'i,M-l < 6. and (n+1) U c,m-l,M-1 (n) U c,m-1,M-1 < 0, hold for all values of i (i = a, b, .... m - 1 ) , where o-j and 5,^ a re predetermined error values. 46 Tne simultaneous solution of Equations (3.34) and (3-35) can be obtained in a direct manner by first writing Equation (3.35) for the largest particle size in that zone. For the present zone (zone M - 1 ) , the largest particle size is m-1. The resulting equation takes the form ^c,m-1,M-1 " •^'^a,M-r S , M - r •'•' ^m-l,M-l' '^•'^°' which indicates that U^ ^_^ j,^_^ is a function only of all concentrations present in zone M-1. As Equation (3.32) indicates, each of these concentrations is a function of U T ... alone; i.e. c,m-1,M-I ^a,M-l "^ "^a'^c,m-l,M-l^ ^b,M-l " "^b'^cm-UM-l^ (3.41) ^m-l,M-l " Vl^^cm-UM-l^ When expressions for these concentrations are substituted from Equation (3.41) into Equation (3.40), there results a nonlinear equation for U^^ ^_i ^_y The latter may be solved for U^ ^_.^ ,,_^ using the Newton-Raphson method or the Reguli-falsi method (Lapidus, 1962). With U _i M 1 known, all unknown concentrations may be obtained directly from Equation (3.34). Substitution of the resulting values for the concentrations into Equation (3.35) gives the r-^naining unknown velocities and thereby the solution is com,plete. Similar equations can be written for zones M-2, ii-O, ..., etc.; and zone by zone computations can be carried out to calculate the concentrations and the settlinc velocities in all zones. 47 The simultaneous solution of Equations (3.34) and (3.35) may also be obtained in an iterative manner. An initial value of U , ,, . is assumed as U^ ^__^ ^, the settling velocity of the same size particle in the zone below the present zone. Substitution of U , ., -, in Eouation c ,m- I ,:1- I (3.34) gives the concentrations of all particle sizes in this zone. These new values of concentrations are used to calculate U Equation (3.35). , ., , from This value of U^ ^ , .. , may be used to calculate new c,m-I ,M-1 -^ concentrations from Equation (3.34). The iteration cycle continues until the difference between two successive velocity values is less than a oredetermined error value. A computer program which carries out tnese computations for an arbitrary preassigned number of zones was written. A sample program is shown in Appendix B. Sedimentation of Suspension with Continuous Particle Size Distribution Partial segregation has been observed in suspensions with continuous particle size distribution. Unlike segregated suspensions, no distinct settling zones are observed. Hov/ever, the overall physical effects may be assumed identical to those in segregated suspensions. In general, any suspension with a broad particle size distribution can be regarded as a mixture of several fractions, each with a narror particle size distribution. Thus the rate of fall of the interface sepa- rating the suspension and the clear liquid in a partially senregated suspension may be predicted usino the theory developed -^or senregated suspensions. CHAPTER IV PROCEDURE AND EXPERIMENTS There are significant inconsistencies in the experimental observations on sedimentation of suspensions and usuallya large number of data are required to evaluate any correlation on sedimentation. It is important to evaluate experimentally the proposed model and to compare it with other models as there are wide discrepancies among the published correlations. Although data were available in the litera- ture , they are not sufficient and more data are essential to verify the model. A large number of experimental data on sedimentation of multisized particles were taken using various sizes of spherical glass particles and different suspending media. Details of the ap- paratus and experimental procedure used for the sedimentation experiments is given below. Glass Spheres Ten sizes of glass particles were used in the sedimentation experiments. These were spherical in shape and were obtained from Cataphote Division of Ferro Corporation. Each particle size was a fraction between two consecutive sieves with at least 95% of the parti cles having a diameter between the openings of the two sieves. The density of each particle size was measured using a pycnometer. A density of 2.43 gm/cm was taken as an average for all sizes since only small variations were observed in the densities of the various 48 49 sizes. The average density supplied by the manufacturer was 2.42 3 gm/cm . An arithmetic average of the sieve-openings was taken as the diameter of each particle size. This assumption was acceptable since the ratio of the diameters of the largest particle to that of the smallest particle in each fraction was 1.19. identification, For easy reference and each particle size was assigned a number. Table 4-1 lists the identification number, the sieve numbers, the size range of the particles, the average diameter, and the density of each sizefraction. Table 4.1 Properties of the Glass Spheres P a r t i c l e size i n sieve numbers No. Diameter range cm Mean Diameter cm Density gm/cm3 1 -35 + 40 0.0500-0.0420 0.0460 2.45 2 -40 + 45 0.0420-0.0354 0.0387 2.42 3 -45 + 50 0.0354-0.0297 0.0326 2.45 4 -50 + 60 0.0297-0-0250 0.0274 2.44 5 -60 + 70 0.0250-0.0210 0.0230 2.43 -70 + 80 0.0210-0-0177 0.0194 2.47 7 -SO + 100 0-0177-0.0149 0.0163 2.4'! 8 -100 + 120 0.0149-0.0125 0.0137 2.^6 9 -120 + 140 0.0125-0.0105 0.0115 2.39 10 -170 + 200 0.0088-0.0074 0.0081 2.34 6 ' 50 Coloring of Glass Particles Segregation of particles can be observed visually if the larger size particles are colored. It is almost impossible to carry out ex- perimental work on sedimentation of multisized particle suspensions without coloring all but the smallest size particles. Particles can be colored by a process originally proposed and patented (U.S. Patent, 1941) for dyeing glass fibers. The fibers are heated with various ionic solutions and supposedly the ions are adsorbed on the glass by a base exchange of the solution ions with the alkali and alkaline earth metals in the glass. It was found that blue and orange colors were easy to produce on the glass particles. Particles were treated with a 3% solution of ferrous sulfate at 85°C for about 30 minutes with occasional stirring, and were subsequently treated with either a warm dilute solution of sodium carbonate for 10 minutes to color them orange, or with a warm 1% solution of potassium ferrocyanide acidified with HCl to color them Prussian blue. Density measurements after coloration showed that the coloring process had no significant effect upon the particle density. Suspending Media Ethylene glycol, diethylene glycol and 60'o aqueous glycerol were used as suspending media. Selection of these liquids depended upon the following requirements: (1) number. High viscosity, which helps in lowering the Reynolds 51 (2) Transparency, which permits observation of the particles while settling and hence monitoring the positions of the interfaces. (3) Chemical inertness, which is essential to prevent any chem- ical attack on the glass particles as well as on the Plexiglas tube in which sedimentation is taking place. (4) Physical inertness: No physical properties other than den- sity and viscosity should affect the rate of sedimentation. Liquids with flocculating effects or those supplying positive or negative ions to the particles cannot be used. (5) Non-hygroscopicity: Suspending media should not show a tendency to absorb moisture from the atmosphere, which would otherwise result in significant density and viscosity changes. Aqueous glycerol has little tendency to absorb moisture from air but this was prevented by keeping it in an air-tight vessel. Viscosity mea- surements taken before and after an experimental run showed no significant change. The densities of the liquids were measured at 24 j^ 1°C using a pycnometer. The viscosities were measured in the temperature range of 20°C to 26°C using a Brookfield Viscometer. The resulting smooth temperature-viscosity curves were used to obtain the viscosities o^ the liquids at 24°C. ties at 24°C. Table 4.2 lists the densities and the viscosi- 52 Table 4.2 Properties of Suspending Media QMcnoo^-;.,^ m«^-,-..m Suspending ^ ^ medium Density at o/ior / 3 24 C, gm/cm^ Viscosity at onor r-r. 24 C, cp Ethylene glycol 1.108 0.184 Diethylene glycol 1.115 0.302 60% Aqueous glycerol 1.165 0.135 Experimental Setup Sedimentation experiments were carried out in a v e r t i c a l bottomed Plexiglas tube 3.2 cm in diameter and 76 cm long. flat- During an experimental run, the tube was held in a v e r t i c a l position using a metal stand which was kept in a controlled temperature chamber. scales graduated to one m i l l i m e t e r were fixed to the f l a t Paper vertical metal plate j u s t behind the tube so that the positions of the i n t e r faces could be e a s i l y read. The controlled temperature chamber was used to keep the v i s c o s i t y and density of the l i q u i d constant and to avoid convection currents. A l l experiments were carried out at a constant temperature of 24 j ^ 0.5°C. Experimental Procedure A s l u r r y of glass p a r t i c l e s was prepared in the sedirr^entation tube by mixing known weights of two or three sizes of p a r t i c l e s with a known volume of the f l u i d . Total volume of the suspension in a l l o experimental runs was 596 cm . A rubber cork was used to plug the 3 top opening of the tube. About 8 cm of air was deliberately allowed to be trapped between the suspension and the cork. When tre tube 53 was tilted upside down and then brought back to the upright position, the trapped air helped in agitating the suspension. The suspension was placed in the constant temiperature chamber until its temperature reached the desired temperature of 25°C. Since the thermal conduc- tivity of plexiglas is low, it took generally more than one hour for the suspension to reach the desired temperature, but was easy to maintain thereafter. The tube was then taken out and the suspension was agitated for several minutes to achieve a uniform concentration of solids throughout the suspension. No significant temperature change was noted during the agitation period. During sedimentation the temperature of the chamber was kept between 23.5°C to 25°C. positions of the interfaces were recorded with time. The At the end of an experimental run, the solids and liquids were recovered by filtering out the liquid through a ceramic filter using a vacuum pump. The particles were then washed with distilled water and were dried in a drying oven. Varying proportions of each particle size were used to prepare 39 different suspensions. These included 30 suspensions of two dis- tinct sizes of particles, 3 suspensions of three distinct sizes of particles, and 6 suspensions with a continuous particle size distribution. Sedimentation data were taken at five different concentra- tions for each suspension. Total concentrations of the solids ranged from 12% to 45% by volume. In all, 195 data points were collected. Each experimental data point was reproduced twice and reproducibility was within 3'.. The arithmetic average of the two measurements was used in the final calculations. These data are tabulated in Appendix A. \ CHAPTER V RESULTS AND DISCUSSION A new model for the sedimentation of multisized particle suspensions was developed in Chapter 3. In order to test the validity of the model, it is necessary to compare results from the model with experimental data. For binary suspensions, the only data available were due to Smith (1965) on sedimentation, Lockett and Al-Habbooby (1973) on sedimentation and countercurrent operations, and due to Mirza and Richardson (1979) on sedimentation. The data by Smith totalled 85 points, out of which 43 data points were used in testing the new model. The remaining data points were excluded since they were only for fine particles of approximately 60 micron in diameter. In this range of small particle size, reliable correlations for sedimentation are not available. Lockett and Al-Habbooby collected data on the initial sedimentation rate for binary suspensions, and also obtained a total of 89 data points for a countercurrent solidliquid system. Only the latter data were included since the present model predicts only the average settling rates. All 45 data points collected by Mirza and Richardson on binary suspensions were used. For ternary systems, only five data points were available from Smith (1965). These were excluded because of the absence of the concentration variable. The data on sedimentation of binary and ternary suspensions collected by Davies (1968) were excluded because of the extremely 54 55 high settling rates encountered in his experiments. Also, the physical properties of the solids and the suspending medium were not reported in his published paper. Because the published experimental data on sedimentation of multisized particle systems were found to be inadequate, a total of 195 new experimental data points were collected. These included 150 data points for binary suspensions, 15 for ternary suspensions and 30 for suspensions with continuous particle size distribution. The published and new experimental data were compared with predictions from the proposed model, the Lockett and Al-Habbooby model and the Mirza and Richardson Model. The results of the comparative analysis are presented below. Comparison of the Models with Experimental data for Binary Suspensions The new experimental data on binary suspensions are listed in Tables A.l through A.3. Tables A.4 and A.5 contain the data acquired from Smith (1965) and Mirza and Richardson (1979), respectively. These tables compare the observed values of velocities of the lower and upper interfaces with those predicted from the present model. The first and second columns of each of these tables give the concentrations of the large and small size particles, respectively. Columns three through six list the observed and predicted interface velocities. The last two columns show the percentage deviation of the predicted values of velocities from the observed values. For each suspension the results are expressed by plotting the observed and predicted velocities of the interfaces versus the voidage of the 56 suspension. The new experimental results for 30 binary suspensions are plotted in Figures 5.1 through 5.30. Figures 5-31 through 5.33 compare the proposed model with the data acquired from Smith (1965). Figures 5.34 through 5.36 compare the present model and MirzaRichardson model with the experimental data provided by Mirza and Richardson (1979). The proposed model and the two previously pub- lished models, those of Lockett and Al-Habbooby (1973) and of Mirza and Richardson (1979), are compared graphically in Figures 5.37 through 5.43. It is clear from these tables and figures that the proposed model is in good agreement with the experimental data for both interfaces. Figures 5-37 through 5.43 show clearly that the Lockett and Al-Habbooby model significantly overpredicts the velocities for both interfaces. Although these figures show that pre- dictions from the Mirza and Richardson model are not unsatisfactory for the upper interface, significant deviations from the experimental data are observed for the lower interface. In order to compare the success of the proposed and previously published models in representing experimental results. Table 5-1 gives the average percentage deviation of predicted values of the interface velocities from the measured values. The results of Table 5.1 are based on a total of 238 data points for the lower interface and an equal number for the upper interface. It is clear from this table that the proposed model shows a substantial improvement over the previously published models. The mean-error and standard deviation shown in Table 5.1 also emphasize the superiority of the proposed model. Though not shown in the 57 o E (D <_> en »* I CD U 03 lO -l-> E E O .»— tn • 1 — • O CL 03 <>j * ~ •r— 03 -(-» o1—1 N > "O CD s03 -o O s: -o •o E 03 4-> OO CD -E CO O s - ^<v sCD Q.4J CL r) <y^ E • 1—1 cvj CXI +-> C L 03 X •o CD CD O CO I O o ZD 1 O CD S-. SC L CD rD - Q ^—' w E =3 c^ S CD _J E I o CD U 03 U- M CD S- I o o LD 1^ <T> X Ln • I — 1 — J3 Z3 a. >i r— CO Z3 O • 1 — > CD S-. CI. CO E O •r^ CO E CD Q. CO 3 (/^ >5 i03 E •r" O CQ CD CO E O Q. O 03 i- -f-> Q . 03 Q CD x : r— +j 03 •a E 03 +E-> CD E •r— S</) CD •r— CL i. X 03 .O QE CD O JE C_> + J E O Ln CD 03 cn CD CJ 03 S_ M CD S2 CD O •4-> Ln Ln CT. , ro c^ CO CO O Ln Ln CTi o I o sLLJ I E 03 CD o 4O CD C_> 03 S- <4CD SCL CD Q.4-> E O .f— -M 03 •n- > CD Q CD CJ 03 i- M CD SQ . CD CL-M =D E cn <,o CD C71 03 4J E CD CJ CD CJ 03 CD Q_ CD CD 03 SCD CX) CTi Ln Ln s.. <+- CD SS CD O +J> _J E CO > -D E rc >, -C O O 4-3 J2! 4-5 X : .—^ CD 03 ro CD "O O r^ -i.: rn U 1 ^ —1 <=c ' — O r— en f ^ E -c o E CO "C S03 03 N .E S- CJ 4_: ra '—V CTi r-^ CTi 1 — • 1 — - 1 — ^ cc: - — f— OJ CO <D SQ_ r\ 53 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^ = 0.0460 cm 00 d3 = 0.0194 cm C3 = 0.0696 Suspending Medium: Ethylene Glycol H CM u OJ o CM CO CJ u o CD nter1^ace <v CNJ >—• CO CD rr o o y ^ 1 O.AQ 0.50 0.6G U /L Voidage, £ w. -ru Figure 5.1. Comparison of experimental results with model for d^=0.0460 cm, d3=0.0194 cm and 03=0.0696- 'v 59 Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from Model d. = 0.0460 cm d3 = 0.0194 cm 00 CM C3 = 0.119 CO Suspending Medium: Ethylene Glycol C3 CJ OJ CO O CM O E o «« >> +-> •r— U CD o ^— CD >• a CD u <a MSCD -!-> CM E _ r . I—< O CO o ^ Q.'AQ 0.50 v_v.w*J '— . ' w w.^_^— W . ^ Voidage, £ Figure 5.2. Comparison of experimental results with model for d =0.0460 cm, d3=0.0194 , 03=0.119. Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model 0.0460 cm OO CM ^S = 0.0194 cm C. = 0.168 Suspending Medium: Ethylene Glycol ZT CM • / sec .20 / / E u ^-^ m >•> •r— u CO o •^-<v • j > I r^- •I o <v u <a s_ CU c OvJ o / 00 o (3 c O.UO 0.50 0.60 0.70 Voidage, c Figure 5.3. Comparison o f experimental r e s u l t s w i t h model f o r d|^=0.0460 cm. d3=0.0194 cm and C, =0.163- ^n 61 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^_ = 0.0460 cm oc d3 = 0.0194 cm CM C^_ = 0.217 Suspending Medium: Ethylene Glycol CM O o U CM CO O 1 a CO o •~^ CU > • o • a; u fO Mt- cu rsi OO CD. c: CD 0.40 0.50 0.60 n ''n Voidage, £ •1 -n 0-20 Figure 5.4. Comparison of experimental r e s u l t s w i t h model f o r d, =0.0460 cm, dc-=0.0194 cm and C, =0.217. 62 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model = 0.0460 cm 00 ':M = 0.0137 cm = 0.0595 Suspending Medium: Ethylene Glycol o <v O CSJ CO ^-^ E u u o CO r— * (U u M- &. (U •M CM o y OO CD o a CD 0.'40 0.50 Voidage, e Figure 5.5. Comparison of experimental results with model for d^_=0.0460 cm, d3=0.0137 and 03=0,0595. 63 Experimental Results (Upper I n t e r f a c e ) Experimental Results .(Lower I n t e r f a c e ) Predicted from Model 0.0460 cm 00 d . = 0.0137 cm 0.119 Suspending Medium: Ethylene Glycol o a O CU CO CM O E CJ o ^-. ^ % CU u <T3 scu CM C — CO c CD o CD o.^o 0.50 O.SO 0.70 0.30 0. 30 Voidage, e Figure 5.6. Comparison of experimental results with model for d,=0.0460 cm, d^=0.0137 cm and C^=0.119. 6^ ~i Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model CO CM 0.0460 cm d . = 0.0137 cm 0.168 Suspending Medium: Ethylene Glycol CM CD 'sec CD CD E u 9k 0C11 J..) O 'ace CU > CO CU CM C3 / CO / ^'O 40 0.50 Voidage, £ Figure 5.7. Comparison of experimental r e s u l t s with model f o r d^=0.0460 cm, d3=0.0137 cm and C^_-0.168. 5V 65 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d|_ = 0.0460 cm oc d^ = 0.0137 cm CM Cj_ = 0.217 Suspending Medium: Ethylene Glycol H CM o a U CD CO ' M _ O 'E u I u ^ CU CJ 03 scu CM •CO CD I 5j CD O .40 ;o ;u ^n Voidage, £ Figure 5.8. Comparison of experimental results with model for d,=0.0460 cm, d3=0.0137 cm and C^=0.217. 66 o • Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^^ = 0.0326 cm zr d3 = 0.0137 cm ft o C3 = 0.0797 Suspending Medium: Ethylene Glycol CD C o CU CO u o CO c <v <v u 03 scu 4J c CO c CD, CD X CNJ CD CO CO 0.40 J . :.5o •:!„ Voidage, £ Figure 5.9- Comparison of experimental results with model for d^=0.0326 cm, d3=0.0137 cm and 03=0.0797- 67 o • — -1 Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from Model d^ = 0.0326 cm d3 = 0.0137 cm C3 = 0.139 Suspending Medium: Ethylene Glycol CM o CU to "E o o •,/ ft o >> •r— u o CD CD CU CU u <o <*- i- <v 4J E CD CD fM (D « O CD 0.40 0 50 n m 0. :C Voidage, £ Figure 5.10- Comparison of experimental results with model for d, =0.0326 cm, d3=0.0137 cm and 03=0.139- 0 ^0 68 O Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^ = 0.0326 cm 7 d3 = 0.0137 cm n / C^ = 0.126 Suspending Medium: Ethylene Glycol CM ^ u <v o CO • / E CJ •« >» -»-> .f— CJ o 1 — CU > CO CD . o CU u 03 MiCU V CO / •-H O CD >^ ,^ u . cG _ . /o '^. z'^ _ . rl Voidage, £ Figure 5.11. Comparison of experimental results with model for d|_=0.0326 cm, d3=0.0137 cm and C^=0.126 0.40 0-^0 69 o Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^ = 0.0326 cm d3 = 0.0137 cm C^ = 0.200 Suspending Medium: Ethylene Glycol 'sec CM CD E CJ A +-> LOO O Face CU >• CD s_ CU CO c / • / _J / /• CNJ ^ ap Figure 5.12 o ,r n ^ / J Voidage, £ Comparison of experimental r e s u l t s with model f o r dj^=0.0326 cm, d3=0.0137 cm and C,=0.200. 7Q> o Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower Interface) Predicted from Model d^^ = 0.0326 cm d3 = 0.0081 cm oo CM C3 = 0.0624 Suspending Medium: Ethylene Glycol '^g 0 CU CO CM 0 E CJ A >1 -M •r— CJ 0 ^ t— CU • / CD 0 CU CJ 03 M- s_ CU +-> E ^—^ CM _ — 0 00 CD CD CD CD c: -n u. 60 Voidage, £ Figure 5.13. Comparison of experimental results with model for d(_=0.0326 cm, d3=0.0081 cm and 03=0.0624. w . -tC- ^ . / • 71 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d|_ = 0.0460 cm d3 = 0.0194 cm oc CM C3 = 0.0612 Suspending Medium: Diethylene Glycol zr 0.20 'sec CM E CJ M LOO CD nte rface Vel •*-) 0 / I CM J / / CO / c 0.40 0.60 Voidage, £ Figure 5.14. Comparison o f experimental r e s u l t s with model f o r d|_=0.0460 cm, d3=0.0194 cm and 03=0.0612. J . DU '\ 72 O Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lov/er I n t e r f a c e ) Predicted from f-lodel d, = 0.0460 cm d3 = 0.0194 cm C3 = 0.122 Suspending Medium: Diethylene Glycol CNJ CD U CU CO O CD I o o oc CU o CD ft :=» CU o ra t+- scu •M E CJ c o o r^j Voidage, £ Figure 5.15. Comparison of experim.ental results with model for d^=0.0460 cm, d3=0.0194 and 03=0.122. \ 73 O 9 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model dj_ = 0.0460 cm ZT d3 = 0.0194 cm ft O C^ = 0.173 Suspending Medium: Diethylene Glycol CM CD 0. 10 /sec / E u m >> •r- U OO o o o nte race <U > ft s- CD CD / / CD C^J 0.40 0.50 n ,q n w. ru Voidage, e Figure 5.16. Comparison of experimental results with model for d,=0.0460 cm, d3=0.0194 cm and 0^=0.173. 74 o Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d|_ = 0.0460 cm d3 = 0.0194 cm C^ = 0.225 Suspending Medium: Diethylene Glycol CM (Di o CU CO "o o CD — o OO o CU CU o <a ^S- cu CD CD / rj CD V.^.-iCJ l _ . - w l _ . _ w _ . ^ ^ . <w ^ « Voidage, £ Figure 5.17. Comparison of experimental results with m.odel for d^_=0.0460 cm, d3=0.0194 cm and C,=0.225. 75 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e Predicted from Model d^ = 0.0460 cm d3 = 0.0137 cm M C3 = 0.0612 Suspending Medium: Diethylene Glycol 'M CJ CD CO CJ o CU o CM CD ft o • / CU o 03 I I &CU ,M CD o ,1 J L n "n w'. -TL Voidage, z Figure 5.18. Comparison of experimental results with model for dL=0.0460 cm, d3=0.0137 cm and 03=0.0612. 76 ~i i o • Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model • / d|_ = 0.0460 cm / d3 = 0.0137 cm C3 = 0.120 Suspending Medium: Diethylene Glycol CNJ ft o O <U CO 'E o ft o u u o CO o CU CU o ^i- <u CD CO ft o rr CD CM CD 0.40 0, 50 0.70 0 60 Voidage, £ -( on Figure 5.19. Comparison of experimental results with model for d, =0.0460 cm, d3=0.0137 cm and 03=0.12a 77 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model ^L = 0.0460 cm ^ s - 0.0137 cm C. = 0.188 Suspending Medium: Diethylene Glycol CM ft o a CU CO OCl u CU > OO c a ft CU u ca • i_ CU •M CD CD zr CO 3/ CM CD r y o c n 0.60 90 Voidage, £ Figure 5.20. Comparison of experimental results with model for d|^=0.0460 cm, d3=0.0137 cm and C,=0.188. ^.^^ 0.50 78 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^ = 0.0460 cm IT d3 = 0.0137 cm CD Cj_ = 0.225 Suspending Medium: Diethylene Glycol CM CD a CJ . CU o CO E CJ «ft >^ +-> o o r— •r— CO CD •:—1 CU >• CU CJ 03 MSCU -l-J E CD CD ^ zr CD CM C2 CD ^ .0 0.50 - CU Voidage, £ Figure 5.21. Comparison of experimental results with model for d^=0.0460 cm, d3=0.0137 cm and 0^^=0.225. 79 o Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model dj_ = 0.0326 cm d3 = 0.0137 cm C3 = 0.0612 Suspending Medium: Diethylene Glycol C\J ft CD u CU CO o ft a "E" CJ -M •r— CJ o CU 00 CD a CU u <T3 S- cu 4-> E CO CD CD ft / CM CD C u 40 C.5L Figure 5.22 'n Voidage, £ Comparison of experimental results with model for d, =0.0326 cm, d3=0.0137 cm and 03=0.0612- 80 o ^ Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted w i t h Model d^_ = 0.0326 cm / / d3 = 0.0137 cm (•>„ CD ft C3 = 0.122 o Suspending Medium: Diethylene Glycol CD CD CO CD "E CJ CO O O CU CD — •—^ ft CU o u 03 cu CD ft O CM CD 0 40 en Voidage, £ Figure 5.23- Comparison of experimental results with model for d^=0.0326 cm, d3=0.0137 cm and 03=0.122. 81 o o Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lov;er I n t e r f a c e ) Predicted w i t h Model d|_ = 0.0326 cm d3 = 0.0137 cm CD C^ = 0.174 Suspending Medium: / Diethylene Glycol / / CO (3 u CU CO m CO E u +-> CJ o CD CD CU U <0 <+- s_ CU ^ -n .~i / / CM / dP zn Voidage, £ Figure 5.25- Comparison of experimental r e s u l t s with model f o r d^=0.0326 cm, d3=0.0137 cm and C^=0.174. 82 o o Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted w i t h Model d^ = 0.0326 cm d3 = 0.0137 cm X Csj C3 = 0.0615 CD Suspending Medium: Aqueous Glycerol ^j ft o CJ CU CO O cu a CD D- CD CU u 03 cu CM CO CD CD CD CD CD 0 40 U . OL o un Voidage, £ Figure 5.25- Comparison of experimental results with model for d|_=0.0326 cm, d3=0.0137 cm and 03=0.0616. 83 Experimiental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model "1 0.0326 cm 0.0137 cm Co = 0.123 Suspending Medium: Aqueous Glycerol CM CD u CU CO O ft o -M •r- u o <v CO CD CU U 03 MS- CU CD CD ."-J CD CD n 40 Figure 5.26 ;U n cr n :"n t;n Voidage, Comparison of experimental results with model for d^=0.0326 cm. d3=0.0137 cm and 03=0.123. 84 O <y — Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from M.odel d, = 0.0326 cm d3 = 0.0137 cm C^ = 0.174 Suspending Medium: Aqueous Glycerol ft O O CJ CU CO ft o u -^ CJ o CO CD CU CU u «3 cu -M CD (D; CD 3 / ; / / I CD_^ / ^J CD CD 0.40 0.50 en Voidage, £ Figure 5.27. Comparison of experimental results with model for dj_=0.0326 cm, d3=0.0137 cm and Cj_=0.174. 85 Experimental Results (Upper I n t e r f a c e ) Experimental Results (Lower I n t e r f a c e ) Predicted from Model d^_ = 0.0326 cm d3 = 0.0137 cm Cj_ = 0.225 Suspending Medium: Aqueous Glycerol CM ft o CJ CU Q CO ^-^ CJ •r— CJ o ^- 'CO CD CD cu >• CU CJ 03 M- s_ CU -M CD CD CD / CD 1 aI 0.5i w . DU ~n ;u Voidage, £ Figure 5.28- Comparison of experimental r e s u l t s with model f o r d, =0.0326 cm, d3=0.0137 cm and Cj_=0.225. 86 O Experimental Results (Upper ExperifTiental Results (Lower Predicted from Model Interface) Interface) 0.0326 cm 0.0081 cm CD OJ c . = 0.0615 Suspending Medium: Aqueous Glycerol OJ o CJ CU CO o CM CD U G CO O —' CU CU CJ sCU CM OO CD CD ft CD CD CD 0.40 0.50 0.60 0.70 Voidage, £ ,^ zn Figure 5.29- Comparison of experimental r e s u l t s with model f o r d,=0.0326 cm, d3=0.0081 cm and 03=0.0615. n 87 o • Experimental Results (Upper I n t e r f a ce) Experimental Results (Lower I n t e r f a ce) Predicted from Model / dj_ = 0.0326 cm •I I I d3 = 0.0081 cm C3 = 0.123 Suspending Medium: Aqueous Glycerol OJ ft o CD CJ < u CO — 1 O , E CJ «ft >> 4-> -r— CJ o OO CD CU o r"" z=> ft <v u 03 l l Scu +J cc E CD zr CD CM C CD 0.40 0.50 Voidage, £ Figure 5.30. Comparison of experimental r e s u l t s with model for d^_=0.0326 cm, d3=0.0081 cm and 03=0.123. 88 OO Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from the Proposed Model d^ = 0.0252 cm o d3 = 0.0131 cm C^_:C3 = 2:1 o o u CD CO 5 <5 o CJ o CD CD CJ fC <+s- CM O CD OQ O. « o 3* o o o. 0.40 0.50 T 0.60 T 0.70 0.80 Voidage, £ Figure 5.31. Comparison of the proposed model with the experimental data of Smith. 0.90 89 Experimental Results (Upper Interface) 00 •' Experimental Results (Lower Interface) o Predicted from the Proposed Model d^_ = 0.0252 cm d3 = 0.0187 cm o C,_:C3 = 4:1 o u CD CO (O CJ u o <v <v CJ 03 4- CM o sCD 00 o, o o. o o o 0.40 0.50 T 0.60 Voidage, T 0.70 0.80 c Figure 5.32. Comparison of the proposed model with the experimental data of Smith. 0.90 90 OO CM Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from the Proposed Model dj_ = 0.0252 cm d3 = 0.0187 cm O CJ OJ CO 'E" CJ <o -t-> u o CD CM CD U 03 4S- o CD 00 0.40 0.50 ""! 0.60 T" 0.70 0.80 Voidage, £ Figure 5.33. Comparison of the proposed model with the experimental data of Smith. 0.90 91 LO cn CD Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from the Models d^ = 0.0463 cm O CO d3 = 0.0115 cm C3 = 0.156 LO CM CJ CD CO o . o CJ O CD CD U 03 4S- lO o cu 4-> o Present Model o Mirza-Richardson Model LO O ilirza-Richardson Model Present Model o o 0.40 0.50 0.50 0.70 0.80 0.90 Voidage, £ Figure 5.34, Comparison of the Mirza-Richardson model and the proposed model with the experimental data of Mirza and Richardson. 92 in on O Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from the Models d^ = 0.0463 cm o on d3 = 0.0115 cm C,_ = 0.308 CM CJ CD CO o CJ O CD CD U 03 <+s_ un 4 o CD Present Model !'1irza-Richardson Model o ft o Mirza-Richardson Model o Present Model a o 0.40 0.50 Figure 5.35. 0.60 0.70 Voidage, £ Q.8G 0.90 Comparison of the Mirza-Richardson model and the proposed model w i t h the experimental data of Mirza and Richardson. 93 O Experimental Results (Upper Interface) Experimental Results (Lower Interface) Predicted from the Models d^_ = 0.0231 cm CO o d3 = 0.0115 cm Cj_ = 0.315 o u <v CO o u o CD en OJ CJ 03 <+S- ci; C50 o Mirza-Richardson Model -Present Model Mirza-Richardson Model O « o — Present Model a o 0.40 0.50 I i O.SO G.7G G.ao 0.90 Voidage, £ Figure 5.36. Comparison of the Mirza-Richardson model and the proposed model with the experimental data of Mirza and Richardson. 94 ^ Experimental Results (Lower Interface) Predicted from the Models d, = 0.0460 cm s_ d3 = 0.0194 cm C3 = 0. Suspend « _ U CD CO u 8- U o CD tf) CU U ro MSCD -l-J E o bbooby Model 8_ hardson Model odel 8. 0.40 0.50 ~[ 0.60 r~ 0.70 T 0.80 0.90 Voidage, c Figure 5.37. Comparison of models with the new experimental data for the lower interface. 95 Experimental Results (Lower Interface) •' o Predicted from the Models d|_ = 0.0460 cm d3 = 0.0194 cm W-. C,_ = 0.217 Suspending Medium: Ethylene Glycol ft CJ CD CO CJ (O u o CD CD CJ CM 03 ^s_ CD tt-Habbooby Model -Richardson Model nt Model —[ 0.40 0.50 0.60 r" 0.70 0.80 0.90 Voidage, £ Figure 5.38. Comparison of models with the new experimental data for the lower interface. 96 » -I • Experimental Results (Lower Interface) Predicted from the Models dj_ = 0.0460 cm d3 = 0.0137 cm Cj_ = 0.217 Suspending Medium: Ethylene Glycol s_ CJ CD CO u (O u o CD r> CD CJ O* * ^ . ^ /-J M- O SCD 8. tt-Habbooby Model S_ irza-Richardson Model Present Model s. 0.40 O.SO •~T 0.60 r~ 0.70 0.80 0.90 Voidage, £ Figure 5.39. Comparison of models with the new experimental data for the lower interface. 97 8 Experimental Results (Lower Interface) Predicted from the Models d|_ = 0.326 cm d3 = 0.0137 cm R- C^_ = 0.126 Suspending Medium: Ethylene Glycol 8 CJ CU CO CJ (O >, u o CU CD CM CJ 03 4iCD -M E s. ockett-Habbooby Model irza-Richardson Model Present Model s. 0.40 0.50 Figure 5.40. •n 0.80 Voidage 0.70 0.80 0.90 Comparison of models with the new experimental data for the lower interface. 98 n «. • Experimental Results Predicted from the Models d^_ = 0.0326 d3 = 0.0137 J5_ C3 = 0.123 Suspending Medium: 60% Glycerol 8. CJ CD CO CJ (O u o CD CD CJ 03 4SCD CM S. Lockett-Habbooby Model s_ lirza-Richardson Model Present Model 0.40 0.50 T 0.60 T 0.70 0.80 0.90 Voidage, £ Figure 5.41. Comparison of models with the experimental data for the lower interface. 99 Experimiental Results Predicted from the Models d|_ = 0.0460 d3 = 0.0194 C3 = 0.119 8. Suspending Medium: Ethylene Glycol O 8 u CU CO S. CJ +j CJ o CD >• CD CJ 03 SCD -M E cn O •' O s. LOckett-Habbooby Model Mirza-Richardson Model Present Model 0.40 0.50 —[ 0.60 T" 0.70 0.80 0.90 Voidage, £ Figure 5.42. Comparison of models with the experimental data for the upper interface. 100 Experimental Results Predicted from the Models d^ = 0.0326 d3 = 0.0137 (O C^ = 0.200 Suspending Medium: Ethylene Glycol in CJ CD CO F CJ «ft >, 4-> •r— U O r— CD s • o >• 0) CJ 03 M- <- 23 o• o CD -l-J E CM tt-Habbooby Model -Richardson Model nt Model s. "~r~ 0.40 0.50 •"1 0.60 r" 0.70 0.80 I 0.90 Voidage, £ Figure 5.43. Comparison of models with the experimental data f o r the upper i n t e r f a c e . 101 table, the proposed model shows a deviation of 20'. or more for just 12 data points, while such a deviation is observed for 77 data points in the case of Mirza and Richardson model and 218 data points in the case of Lockett and Al-Habbooby model. Comparison of the Models with the Experimental Data for Ternary Suspensions: The new experimental data for ternary suspensions are listed in Table A.6. The table compares the observed values of velo- cities of the lower interface, the middle interface, and the upper interface with those predicted from the proposed model. The first three columns of the table give the concentrations of the large, medium and small size particles, respectively. Columns four through nine list the predicted and observed interface velocities. The last three columns show the percentage deviation of the predicted values from the observed values. For each suspension the results are ex- pressed by plotting the observed and predicted velocities of the interfaces versus the voidage. through 5.46. These plots are shown in Figures 5.44 In general, these figures show that the agreement of the proposed model with the experimental data is satisfactory, though the predicted velocities are underpredicted for some interfaces. Using the same data points, the three models are compared in Table 5.2. The results in this table indicate that the present model shows a slight improvement over the Mirza and Richardson model. Because of the lack of sufficient experimental data (only 15 data points are available), it is not possible to carry out a conclusive comparative study and, obviously, more experimental data for terr.ary 102 «!r <U j t ''J 1 !<J c OJ U | l/l JZ. 4-> i l^ c o S fe •r. i n CO 1 •." > CM OJ —» u-> ^r => K "D Sn3 fO .4-1 1/1 ''r"— o. o s La£ ^-r ^vi ^ '—' l z Lrt CO O ^— X '0) en ro O o • ^— -^ r^ CVJ 1 0) CJ XI ro ro 1 1 W <4_ r^ o O <u ^ X X X yD «3- cz: ITI 3 <U O ro —1 c CO E 00 ro t u <4- •o -O CM m <U — ro Ol C OJ TS ' ° -a c JZ CO i rfl a . IT3 Q o s: -a CD Z2 Q- IQ. 1—1 X o " io «5 O CD "O =3 \-r. 4-1 ^^ X a. (u ., •r— o 1^ i . <*<U i- un ' • — O > ^ T 1— CO (J O •r- m a. a; a.-t-» Q- =3 lO > Z3 CD <J-i 1 (U s- i_ 03 E S_ CD o s<1u - a 1— rt3 CD CO E O O Q. O 03 S- + J Q03 Q CD - E 1— -t-> 03 -M ME O CD S- X C-J - M • un •r- 4-1 o o t — z = .—. <u un o p _ IB O un CM CO 4J '—' 01 CJ .T5 ' 3 3 01 C 0) o 00 ' CM r«» un un .1-1 "3 — .>1) ^ rs • —c• OJ 0) cn «3 4-> ! c OJ u i '^ 1 CVI 0) 0) s> 3 01 03 LLJ C3. E CD O -E CM -3 L. M - •!- CD C i*- 0) o s- in •1- 1—. 1 — z E E •vO c a; (O "O E 03 un o 5.4 13 CO E Z3 CD i r— 1^— CM -3 -3 S0) UO •f— 4—1 y • uo c •— 1 OJ CJ j1 1 CM iti r— i. <40) 1 . CM CO • ro 3 OJ ro ro O 4-J LO CD JZl <o -3 O •o ~r 1) -c w I -A <.—' fT3 -a o 103 Experimental Results (Upper Interface) Experimental Results (Middle Interface) cn Experimental Results (Lower Interface) Predicted from the Proposed Model Suspending Medium: 60% Glycerol d^ = 0.0460 cm a d^^ = 0.0194 cm o » o d^ = 0.0081 cm C^ = 0.0518 C^ = 0.0518 in CM CJ CD CO o o CM 4-> O o CD CD U in 03 ^s- CD o « o in o o o 0.40 0.50 Figure 5.44- ^1 \— 0.60 0.70 Voidage, £ 0.80 Comparison of the proposed model with the experimental data on ternary suspension. 0.90 r. 1 -«- Experimental Results (Upper Interface) ^ • Experimental Results (Middle Interface) • Experimental Results (Lower Interface) — Predicted from the Proposed Model Suspending Medium: 60% Glycerol d^ = 0.0460 cm a d^ = 0.0194 cm d^ = 0.0081 cm C^ = 0.104 C = 0.0518 in CM U CD CO (J c^ >, o CJ o CD CD CJ 03 MS_ CD in o T 0.40 0.50 0.60 0.70 0.80 Voidage, £ Figure 5-45- Comparison of the proposed model with the experimental data on ternary suspension. 0.90 105 Experimental Results (Upper Interface) Experimental Results (Middle Interface) R Experimental Results (Lower Interface) Predicted from the Proposed Model Suspending Medium: 60% Glycerol d^ = 0.0460 cm d « d|^ = 0.0194 cm d^ = 0.0081 cm C^ = 0.207 (M . O C = 0.0518 u CD CO F CJ 0\ ^ -l-> •r— (O ^^ . o CJ o r^ CD :> CD CJ 03 (M ^^ <+- o i- CD -M E a s • o o 0.40 0.50 T T 0.60 0.70 Voidage, c 0.80 Figure 5-46. Comparison of the proposed model v.'ith the experimental data on ternary suspension. 0.90 10' suspensions are needed. Nevertheless, the present model does appear to be quite staisfactory in describing the sedimentation of ternary particle systems. Behavior of the Models at High Reynolds Numbers: Of particular importance is to study the validity of the model for a wide range of Reynolds numbers. Sedimentation experiments provide data only in the low Reynolds number range. Data at higher values of Reynolds number may be obtained from co- or counter-current solid-liquid vertical flow operations. The data collected by Lockett and Al-Habbooby (1973) from countercurrent flow experiments covered a range of Reynolds number of about 79 to 546. These data are 1 isted in Table A.7. Unlike sedimentation, continuous countercurrent operations of binary suspensions do not give rise to two zones. However, the physical system is similar to the lower zone of a binary sedimenting suspension and slip velocities of each size particles are calculated using Equations (3.12) and (3.13) developed in Chapter three. Slip velocities predicted from the present model are compared with the observed values as shown in Table A.7. The first three columns of the table list the flowrates of the liquid, the large size particles, and the small size particles, respectively. Columns four and five give the concentrations of the large and small size particles. Columns six through nine list the predicted and observed slip velocities for each size. The last two columns show the percentage de- viation of the predicted velocities from the observed values. Table 5.3 shows the average percentage deviation, the m.ean-error. 107 u CD CO o LD CO E O •r— -M 03 CD %. - E CD -M Q. O -E +J -M •I— E 2 CD SCO r— : 3 CD CJ s_ -a so CD s: +EJ 13 -o CD O E O CO -i-> E cn o o +J o 03 • r - C\J > c\j I — CD 03 N E -f<.r) (Ji Q. ^ CD Q 03 Q. X CD S03 "O E 03 Z t/) KI I 03 -o O CU i(ZL SCD SO E ZS CD CD CD SM 03-1_ J <>0 CM cn - E CJ CO "r— E 1 O -O Z3 > > Q . JO O >^ 1 — o o CM CM o <J^ JO CO SO Zi 03 3: •r— 1 > 1— CD c C SQ_ • u E • o 03 E 03 -f-J 4-> CD CD J^ CO CJ O Q 1 O s_ ^ Q- LO r— 03 o SUJ I E 03 CD E -r- I <y) (J-) CD CD CD SM _i C>0 I— 03 CD N c o cr> CD N 00 CO o o CM I CM CO CM -o o o <V 03 -E +j -i-J 03 Q MO r— 03 E +J E O CO CD •r— E S- • 1 — 03 SQ. CD f= C L O X CJ LU ro LO E O +J 03 > CD Q CM E -r- C/1 OO CM <m CM CM CM CD CD 03 +J E CD CJ idJ QCD CD r3 S_ CD > CTi C30 O CD c n CD SN 03 - r _ J U^ o CO O CM CD ^— JZi 03 -D E 03 CD >^ -O O O -M XJ + J - Q ..—s O) 03 c^ ^^ or r^ CJ O E X3 O E CO ro " O S- - ' - ro 03 CTi IN - E 1 ^ S- CJ CTi 1 o .— —1 < : CT> r— •r— T— • 1 2: cr —' o CJ i/"; r~^ f V -—\ ^ Cl. 108 and the standard deviation for all three models. The results in Table 5.3 show that predictions from the published models do not compare well with the experimental data. On the other hand, the pre- sent model represents the experimental data fairly well. The super- iority of the present model over the published models and its validity over a wide range of Re is thereby established. Comparison of the Proposed Model with the Experimental Data on Suspensions with Continuous Size Distribution: There is no reason for the progressive settling analysis, which is apparently successful in the case of discrete size distribution, not to be successful for the progressive settling of continuous size distribution. In fact, any suspension can be regarded as being composed of a large number of closely sized fractions. In making the progressive settling computa- tions with the distribution broken into various numbers of size species. Smith (1966) found that it is sufficient to break down the distribution such that the ratio of the species-to-species average diameter is 1.2. In the U.S. sieve series, the sieve interval, which is the ratio of two successive sizes of screen openings in the series, is V2( = 1.19). Accordingly, two kinds of suspensions were prepared; the first kind consisted of a mixture of two successive size-fractions, and the second of three successive size-fractions. Tables A.8 and A.9 show the experimental data and further analysis for the two-species and three-species suspensions, respectively. The first two columns of Table A.8 list the diameter of each size-fraction. Columns three and four of the table show the concentrations of the two ^4 -<^»^',' 109 fractions. Columns five and six list the predicted and observed interface velocities. The last column list the percentage deviation of the predicted velocities from the observed values. Similarly, the first three columns of Table A.9 list the diameter of each sizefraction. Columns four through six show the concentrations of the three fractions. Columns seven and eight list the predicted and ob- served interface velocities. The last column shows the percentage deviation of the predicted velocities from the observed values. Figures 5.47 through 5.49 display the plots of the observed and predicted velocities versus the voidage for the two-species suspensions. Similarly, Figures 5.50 through 5.52 show the plots of observed and predicted interface velocities versus the voidage for the threespecies suspensions. It is clear from these tables and figures that the proposed model satisfactorily represents the experimental data. Error analysis for the two-species suspension gives an average percentage deviation of 8.23 and a mean-error of -3.24. The average percentage deviation and the mean-error for the three species suspensions are 16.24 and -16.24, respectively. This indicates that the predicted velocities are underpredicted in the latter case. These errors are of the same order of magnitude as those of the top interface for binary suspensions and ternary distinct sized suspensions. Thus suspensions with continuous size distribution show the same settling pattern as those with discrete size distribution. Furthei-- more, the proposed model can be used to predict the sedimentation velocities of suspensions with continuous size distribution. no OO CM • -n Experimental Results Predicted from the Present Model Particles: Fractions 1 and 2 Suspending Medium: Diethylene Glycol <M^ O O CM. u CD (/) CJ o +J CJ o CD CM CD CJ 03 MSCD 4-> E OO o. « o 3* o o. 0.40 0.50 T 0.80 T 0.70 0.80 0.90 Voidage, £ Figure 5.47. Comparison of the proposed model with experimental data on suspensions with continuous size distribution. Ill O • Experimental Results — Predicted from the Present Model Particles: Fractions 5 and 6 Suspending Medium: Diethylene Glycol CO o in o CJ CD CO CJ rr o* — » - _ -i-> •r- u o CD > en CD ^ ro ^s- O, rA O CD CM O « o o o 0.40 0.50 0.60 0.70 0.80 0.90 Voidage, £ Figure 5.48. Comparison of the proposed model vyith experimental data on suspensions with continuous size distribution. 112 o Experimental Results o Predicted from the Present Model Particles: Fractions 8 and 9 Suspending Medium: Diethylene Glycol OJ o o CD o o u CD CO u 'JO o o CJ o CD CD O 03 M- o c s_ <v o CM o o o. 0.40 0.50 Figure 5.49. —1 I 0.60 0.70 Voidage, £ 0.80 0.90 Comparison of the proposed moael with experimental data on suspensions with continuous size distribution. 113 ZT —•_ « O CM • Experimental Results Predicted from the Present Model Particles: Fractions 1, 2, and 3 Suspending Medium: Diethylene Glycol O o o a CD CO CJ OO o CJ o OJ CD CJ 03 CO o t ° CD +J 3* O CM O O o 0.40 0.50 ~~1 0.60 \— 0.70 Voidage, £ 0.80 0.90 Figure 5.50. Comparison of the proposed model with experimental data on suspension with continuous size distribution. 114 o Experimental Results Predicted from the Present Model Particles: Fractions 3. 4, and 5 Suspending Medium: Diethylene ^ilvcol CO o in o u CD CO CJ o^ ' — j • O « >^ -M •r— O O CD CD U 03 MSCD CO o CM O • o o o 0.40 0.50 Figure 5.51. 0.60 0.70 Voidage, £ 0.80 0.90 Comparison of the proposed model with experimental data on suspension with continuous size distribution. 115 Experimental Results Predicted from the Present Model Particles: Fractions 5, 6, and 7 Susoending Medium: Diethylene Glycol OO O u CD CO a O 4-> O o CD OJ U 03 <4SCD +J o, o o o 0.40 0.50 O.SO 0.70 0.80 0.90 Voidage, £ Figure 5.52. Comparison of the proposed model i.'ith experimental data on suspension with continuous size distribution. CHAPTER VI CONCLUSIONS AND RECOMMENDATIONS Conclusions A physical model for sedimentation of binary suspensions was first developed by Lockett and Al-Habbooby (1973). This model was based on the assumption that a particle settles only according to the local voidage fraction around it, regardless of whether its neighbors are particles of the same or of other sizes and whether they are moving relative to it or not. This model, therefore, does not take into account particle collisions and interactions arising from the relative velocities of the particles. When tested experimentally, the model was found to overpredict the sedimentation velocities by between 5 and 70-o. Mirza and Richardson (1979) extended the Lockett and Al-Habbooby model for multisized particle suspensions. and collisions were neglected. Again, interparticle interactions In order to match the predicted velo- cities with the observed values, the predicted velocities were multiplied by a factor equal to the voidage raised to the power 0.4. By introducing such a correction factor, they were able to represent the data reasonably well for binary and ternary suspensions. However, no 0 d physical explanation was provided for the correction factor £ ' . Clearly, such an empirical correction cannot be aoplied over a wider range of conditions than those used in their experimental work without further investigation. 116 117 In the present study, a model was developed for the sedimentation of multisized particle suspensions which takes into account interparticle interactions. These interactions were accounted for by taking into consideration the buoyancy effect induced by the smaller size particles on the terminal falling velocities of large size particles. The new model and the two previously published models, those of Lockett and Al-Habbooby and of Mirza and Richardson, were tested against a large number of published and newly collected experimental data on binary and ternary suspensions. It was found that the proposed model is the m.ost accurate and represents the experimental data satisfactorily. Qy suc- cessful representation of the Lockett and Al-Habbooby data on binary countercurrent operations, the model was proven to be valid over a wide range of Reynolds number. Furthermore, the model was found to apply for suspensions with continuous size distribution. In making computations for continuous size distributions, it was found sufficient to break down the distribution into closely sized fractions such that fraction-tofraction average diameter ratio is 1.19. It is recommended that the present model be used to predict the sedimentation velocities of suspensions with discrete or continuous size distributions. Also, the model can be used to predict particle velocities in multisized co- or counter-current solid-liquid operations. Recommendations for Further Work 1. The model was shown to agree favorably well with the experi- mental data for ternary suspensions and those with continuous size distribution. However, because of the meager data base used, more 118 experimental data and further verification of the model may be complimentary. 2. The model should be extended to include the sedimentation of multicomponent mixtures of particles of different densities and different sizes. 3. Sedimentation data for fine particles of diameter less than 100 micron is required and a new correlation for the sedimentation of equisized fine particle suspensions needs to be developed. 4. An attempt should be made to develop the shape factor or equi- valent diameter in the case of sedimentation of non-spherical particle suspensions. 5. An experimental technique for monitoring the position of the solid-liquid interface during the sedimentation of opaque suspensions needs to be developed. In this connection, the method introduced by Raffle (1976) to study the settlement and consolidation of concentrated suspensions may be used. By continuously monitoring the pressure vari- ations at various depths during sedimentation, the position of the interface may be determined. NOMENCLATURE A = Asymptotic value of U^/U^ at low Re (Equation 2.57) A^ = Cross-sectional area of vessel or tube ap = particle acceleration B = Asymptotic value of U^/U^ at high Re (Equation 2.57) C = r Cp Fractional volumetric concentration of particles n (TT/6)d (p_-p.p)g = Drag coefficient = ^—i 1/2 p^(Tr/4 d^)U^ % = [4d(Pp-p^)g/(3, U j ) ] [ ( l - C ) / ( U C ^ / ^ ) ] D = Diameter of vessel or tube d = diameter of spherical particle F = f. = drag force on a particle g = Acceleration due to gravity K = Constant in Equation (2.57) n = Exponent of £ in Equation (2.57) R, = Drag force per unit projected area of isolated spherical Force on a single particle in suspension particle R' = Drag force per unit projected area of spherical particle in suspension t = time U = S e t t l i n g v e l o c i t y of p a r t i c l e in suspension U^ = Velocity of displaced f l u i d U. = S e t t l i n g v e l o c i t y at i n f i n i t e U = Superficial v e l o c i t y of f l u i d 119 dilution 120 U^ = Terminal velocity of a single particle in finite fluid medium ^too " Terminal velocity of a single particle in an infinite fluid medium V = Volume of a single particle X = distance z = constant in Equation (2.57) Greek Letters £ = voidage (fluid fraction in suspension) <^^ = [1 - 1.21(l-£)2/^]"'' y^ = Viscosity of f l u i d ]i = Viscosity of suspension Pr = density of f l u i d pr = Mass of f l u i d per unit volume of f l u i d - p a r t i c l e mixture p = density of p a r t i c l e s p = Mass of p a r t i c l e s per u n i t volume of f l u i d - p a r t i c l e mixture r p = density of suspension Dimensionless Groups 3 ' 2 d p^(p -p.p)g/p.p Ga = Galileo number Re = Particle Reynolds number dU,p-,/u^ Re^ = Fluid Reynolds number dU p^p/y^ Re^ = Particle Reynolds number dU^^pVy^p Re, (p = U /U, Modified Reynolds number Re ( - r r / o M r\- ) -^ 00 exp^. 5 L / 3 ( 1 - C ) ,, 121 Subscripts L = Large particle S = Small particle a-m = Sizes of particles (smallest to largest) 1-M = Settling zones (top to bottom) BIBLIOGRAPHY 1. Barnea, E. and J- Mizrahi, "A Generalized Approach to the Fluid Dynamics of Particulate Systems Part 1. General Correlation f o r F l u i d i z a t i o n and Sedimentation in Solid M u l t i p a r t i c l e Systems," Chem. Enq. J . . Vol. 5, pp 171-189 (1973). 2. Becker, H. A., "The Effects of Shape and Reynolds Number on Drag in the Motion of a Freely Oriented Body in an I n f i n i t e F l u i d , " The Can. J . Chem. Eno. . Vol. 37, pp 85-91 (1959). 3. Bedford, A. and 0. D. H i l l , "A Mixture Theory Formulation for P a r t i c u l a t e Sedimentation," A.I.Ch.E. Journal, Vol. 22, pp 938940 (1976). 4- Brinkman, H. C., "A Calculation of the Viscous Force Exerted by a Flowing Fluid on a Dense Swarm of P a r t i c l e s , " Appl. Sci. Res., Vol. A l , pp 27-34 (1947). 5. Coe, H. S. and G. H. Clevenger, "Methods for Determining the Capacities of Slime S e t t l i n g Tanks," Trans. Am. I n s t . Min. .Met. Eng., Vol. 55, pp 356-384 (1916). 6. Davies, C. N. , " D e f i n i t i v e Equations for the Fluid Resistance of Spheres," The Proceedings of the Physical Society, Vol. 57, pp 259-270 (1945). 7. Davies, R., "The Experimental Study of the D i f f e r e n t i a l S e t t l i n g of P a r t i c l e s in Suspension at High Concentrations," Powder Technology, Vol. 2, pp 43-51 (1968/69). 8. F i l d e r i s , V. and R. L. Whitmore, "Experimental Determination of the Wall Effect f o r Spheres F a l l i n g A x i a l l y in C y l i n d r i c a l Vessel," B r i t . J . Appl. Phys., Vol. 12, pp 490-494 (1961). 9. F i n k e l s t e i n , E., R. Leton, and J . C. E l g i n , "Mechanics of Vertical Moving Fluidized Systems with Mixed P a r t i c l e Sizes," A.I.Ch.E. Journal, Vol. 17, pp 867-872 (1971). 10. Francis, A. W., "Wall Effect in Falling Ball Method for V i s c o s i t v , Physics, Vol. 4, pp 403-406 (1933). 11. Garside, J . and M. R. Al-Dibouni, "Velocity-Voidage Relationship f o r F l u i d i z a t i o n and Sedimentation in Solid-Liquid Systems," Ind. Enq. Chem., Process Des. Dev., Vol. 16, pp 206-214 (1977). 12. Goldestein, S., "The Steady Flow of Viscous Fluid Past a Fixed Spherical Obstacle at Small Reynolds Number," Proc. Roy. Soc. , Vol. 123 A, p 225 (1929). 122 123 13. Gurel, S., S. G. Ward, and R. L. Whitmore, "Studies of the Viscosity and Sedimentation of Suspensions Part 3. - The Sedimentation of Isometric and Compact Particles," Brit. J. Apol. Phys., Vol. 6, pp 83-87 (1955). 14. Happel, J., "Viscous Flow in Multiparticle Systems: Slow Motion of Fluids Relative to Bed of Spherical Particles," A.I.Ch.E. Journal, Vol. 4, pp 197-201 (1958). 15. Happel, J. and H. Brenner, "Low Reynolds Number Hydrodynamics," Prentice-Hall, Englewood-Cl iffs, N.J., 1965. 16. Hawksley, P. G. W., "The Effect of Concentration on the Settling of Suspensions and Flow Through Porous Media," Inst, of Phvs. Symposium, p 114 (1950). 17. Heyv/ood, H., "Measurements of the Fineness of Powdered Materials," Proc. Inst. Mech. Engrs. (London), Vol. 140, pp 257-308 (1938). 18. Hoffman, R. F., L. Lapidus, and J. C. Elgin, "The .Mechanics of Vertical Moving Fluidized Systems," A.I.Ch.E. Journal, Vol. 6, pp 321-324 (1960). 19. Hottovy, J. D. and N. D. Sylvester, "Drag Coefficients for Irregularly Shaped Particles," Ind. Eng. Chem., Process Des. Dev., Vol. 18, pp 433-436 (19807: 20. Jottrand, R., "An Experimental Study of the Mechanism of Fluidization," J. Appl. Chem., Vol. 2, pp S^^-S^g (1952). 21. Jottrand, R., "Calculations of Terminal Falling Velocity of Particles in Fluids," Brit. Chem. Enq., Vol. 3", pp 143-146 (1958). 22. Kunkel, W. B., "Magnitude and Character of Errors Produced by Shape Factors in Stokes' Law Estimates of Particle Radius," J. Appl. Phys., Vol. 19, pp 1056-1058 (1948). 23. Lapidus, L., "Digital Computation for Chemical Engineers," McGraw-Hill, New York, NY, 1962. 24- Lapidus, L. and J. C. Elgin, "Mechanics of Vertical-Moving Fluidized Systems," A.I.Ch.E. Journal, Vol. 3, pp 63-68 (1957). 25. Lewis, E. W. and E. W. Bowerman, "Fluidization of Solid Particles in Liquid," Chem. Enq. Progr., Vol. 48, pp 603-619 (1952). 26. Lewis, W. K., E. R. Gilliland, and W. C. Bauer, "Characteristics of Fluidized Particles," Ind. Enq. Chem., Vol. 41, pp 1104-1117 (1949). 124 27. Letan, R., "On Vertical Dispersed Two-Phase Flow," Chem. Eng. S c i . , Vol. 29, pp 621-624 (1974). 28. Lockett, J . M. and H. M. Al-Habbooby, " D i f f e r e n t i a l S e t t l i n g by Size of Two P a r t i c l e Species in a L i q u i d , " Trans. I n s t , of Chem. Enq., Vol. 5 1 , pp 281-292 (1973). 29. L o e f f l e r , A. L. and B. F. Ruth, " P a r t i c u l a t e F l u i d i z a t i o n and Sedimentation of Spheres," A.I.Ch.E. Journal, Vol. 5, pp 310314 (1959). ~ 30. Maude, A. D. and R. L. Whitmore, "A Generalized Theory of Sedimentation," B r i t . J . Appl. Phys., Vol. 9, pp 477-482 (1958). 31. Meters, T. S. and H. B. Rhodes, " L i q u i d - P a r t i c l e Behavior," Chem. Enq. Progr., Vol. 5 1 , pp 429-432 and pp 517-522 (1955). 32. Mirza, S. and J . F. Richardson, "Sedimentation of Suspension of P a r t i c l e s of Two or More Sizes," Chem. Eng. Sci. , Vol. 34, pp 447-454 (1979). 33. O l i v e r , D. R., "The Sedimentation of Suspension of Closely Sized Spherical P a r t i c l e s , " Chem. Eng. S c i . , Vol. 15, pp 230-242 (1961). 34. Oseen, C. W., Arkiv f. Mat., Astron. Och Fysik, Vol. 6, No. 29 (1910). 35. Pettyjohn, E. S. and E. S. Christiansen, "Effect of P a r t i c l e Shape on Free S e t t l i n g Rates of Isomeric P a r t i c l e s , " Chem. Eng. Progr., Vol. 44, pp 157-172 (1948). 36. P r i c e , B. G., L. Lapidus, and J . C. E l g i n , "Mechanics of V e r t i c a l Moving Fluidized Systems," A.I.Ch.E. Journal, Vol. 5, pp 93-97 (1959). 37. R a f f l e , J . F. , "Pressure Variation Within Concentrated S e t t l i n g Suspensions," J . Phys. D: Appl. Phys., Vol. 9, pp 1239-1252 (1976). 38. Richardson, J . F. and W. N. Zaki, "Sedimentation and F l u i d i z a t i o n Part I , " Trans. I n s t . Chem. Eng., Vol. 32, pp 35-53 (1954). 39. Richardson, J . F. and R. A. Meikle, "Sedimentation and F l u i d i z a t i o n Part I I I . The Sedimentation of Uniform Fine Particles and of Two-Component Mixtures of S o l i d s , " Trans. I n s t . Chen. Eng.. Vol. 39, pp 348-356 (1961). 40. Smith, T. N., "The D i f f e r e n t i a l Sedimentation of P a r t i c l e s of Two D i f f e r e n t Species," Trans. I n s t . Chem. Eng., Vol. 43, pp 769-773 (1965). 125 41 Smith, T. N., "The Sedimentation of Particles Having a Dispersion 0^ Sizes," Trans. I n s t . Chem. Eng., Vol. 44, pp T153-T157 (1966). 42. Smith, T. N., "The D i f f e r e n t i a l Sedimentation of Particles of Various Species," Trans. I n s t . Chem. Enq., Vol. 45, T311-T313 43. Steinour, H. H., "Rate of Sedimentation," Ind. Eng. Chem., Vol. 36, pp 618-624 (1944). 44. Stokes, G. G., "On the Effect of the Internal F r i c t i o n of Fluids on the Motion of Pendulum," Trans. Cambridge P h i l . S o c , Vol. 9, p 8 (1951). ^ 45. Struve, D. L., L. Lapidus, and J . C. E l g i n , "The Mechanics of Moving Vertical Fluidized Systems," The Can. J . Chem. Eng., Vol. 36, pp 141-152 (1958). 46. Turian, R. M., T. F. Yuan and G. Mauri, "Pressure Drop Correction f o r Pipeline Flow of Solid-Liquid Suspensions," A.I.Ch.E. Journal, Vol. 17, pp 809-817 (1971). 47. U.S. Patent 2,245,783 (1941). 48. Wen, C. Y. and Y. H. Yu, Chem. Eng. Progr. Symp. Ser. No. 62, Vol. 62, p 100 (1966). 49. Wen, C. Y. and L. S. Fan, "Some Remarks on the Correlation of Bed Expansion in L i q u i d - S o l i d Fluidized Beds," Ind. Enq. Chem., Process Des. Dev., Vol. 13, pp 194-197 (1974). 50. Wolfenden, H. H., "The Fundamental Principles of Mathematical S t a t i s t i c s , " pp 87, 238, MacMillen, Toronto, 1942. 51. Zenz, F. A . , "Calculate F l u i d i z a t i o n Rates," Petroleum Refiner, Vol. 36, pp 147-156 (1957). 52. Zenz, F. A. and F. D. Othmer, " F l u i d i z a t i o n and F l u i d - P a r t i c l e Systems," Reinhold, New York, 1960. 53. Zuber, N., "On the Dispersed Two-Phase Flow in the Laminar Flow Regime," Chem. Eng. S c i . , Vol. 19, pp 897-917 (1964). APPENDIX A EXPERIMENTAL DATA AND COMPARISON WITH PREDICTIONS FROM PROPOSED MODEL 126 127 Table A . l . Comparison of the Proposed Model with the Experimental Data for Binary Suspension in Ethylene Glycol ! I n t e r f a c e V e l o c i t y , cm/sec Concentration of P a r t i c l e s voluTC f r a c t > o n Laroe Small 1 j Lower i n t e r f a c e Predicted Exoerimental i, = 0.0460 cin Upper i n t e r f a c e E.xoeriniental Predicted dj ' a t i on Percent Devi ^ C » 1 «* w ' \ * • ' ! LoweInterface uppeInter-ace -1.45 -5.72 -4.55 -4.01 - c . oS \SE -. . - 3.47 -1 SC - 3. ^' 0.0194 CT C.199 0.239 0.268 C.29a 0.33& 0.0696 0.0696 0.0696 0.0696 0.0696 0.163 0.123 0.0957 0.078C 0.0560 G. 170 0.124 0.100 0.0R05 0.0582 0.0725 0.0625 0.0547 0.0465 0.0359 G.0735 0.0663 :.0573 0.0454 o.o:':- -4.12 -1.13 -1.3'; -3.06 -2.7C 0.139 0.179 0.219 0.259 0.295 0.119 0.119 0.119 0.119 0.119 0.170.131 0.0959 0.0706 O.C504 0.181 0.133 0.0960 0.0692 0.0494 0.0570 0.045= C.0416 0.0336 0.0260 C.0559 0.0^93 0.414 0.0342 0.0270 -1.8' 0.97 2.01 ' , 53 0.158 0.168 0.168 0.166 C.16B 0.0696 0.201 0.151 0.111 0.0804 0.0566 0.212 0."'56 0.113 0.0826 0.0581 0.0791 0.0558 0.0390 0.0271 0.0183 0.0769 0.0530 0.0373 0.0271 C.0187 -:. 0.217 0.217 0.217 0.21? 0.217 0.0696 0.144 0.106 0.0758 0.0536 0.0365 0.141 0.''03 0.0759 0.0537 0.0362 0.C681 0.0459 0.03G6 0.0204 0.0132 0.0705 0.0439 0.0311 C.0210 0.0140 :.9G 2.60 -G. ' 4 -0.26 0.69 -3.35 •1.4' -•' .56 -2.84 -e.s: o.no 0.151 0.191 0.232 O.no 0.151 0.191 C.232 d^ = 0.0460 cir i-i £ . C' • -3.21 -1.38 -2.62 -2.64 5. 3c - . - J. COG -1 .93 _ • ^Ci d<; = 0.0137 cir 0.197 0.238 0.27S C.319 0.359 0.0595 0.0595 0.0595 0.0595 0.0595 C.179 0.134 0.0999 0.0723 0.0517 0.180 0.132 0.101 0.0743 0.0521 0.0^50 0.0418 0.0373 0.0327 0.0270 0.0483 0.0445 C.040S 0.0348 0.0289 -0.63 ^ ^C -1.12 -2.6S -0.84 G. 139 0.179 0.219 0.259 0.299 0.119 0.119 0.119 0.119 0.119 0.176 0.133 0.0985 0.0719 0.0513 0.173 0.129 0.0948 0.0684 0.0476 0.C313 0.02':'.5 0.0254 0.0219 0.0181 3.0327 0.0285 0.0251 0.0217 0.0182 1.5C 2.81 3.9^ 5.04 7.77 0.168 0.168 C.16c 0.158 0.168 0.0696 0.110 C.151 0.191 0.232 0.203 0.153 0.113 0.0827 0.0587 0.201 0.149 0.11c 0.0322 0.0530 0.0433 0.0315 0.0225 0.0159 0.0109 0.0433 0.0306 0.0218 0.0160 D.011C 0.88 2.75 2.S7 0.59 1.13 0.00 3.03 3.12 -0.72 0.21'' C.217 0.21"" o.:i" 0.0695 0.110 O.iSi 0.191 0.145 0.107 0.0775 0.0552 0.147 0.109 0.0762 0.0520 C.0398 0.C278 0.0185 0.0123 0.0421 0.0287 C.0193 -1.40 -1.59 -5.3c -3.3' -1.34 -Z.59 d^ = C.0326 cn •J .J\ JC 1 .67 6.13 -6.08 -7.37 -6.01 -o.c5 . 1 -^ c.co 1 1£ C.S^ -G.'io .1', 01 •» Of - CO'. 3 0.0993 0.159 0.21° 0.278 0.337 0.0797 0.0797 0.0797 0.C797 0.0797 C.'i47 G.lOO C.C563 0.0421 0.':257 0.149 C.102 C.0686 0.0402 C.C257 0.0422 0.0372 0.0308 0.C232 0.0159 C.0131 0.0335 C.G320 0.0237 0.016"' -1.3" -", .93 0,0993 0.139 0.179 0.229 0.256 0.139 0.139 0.139 0.139 0.139 0.0938 0.0752 C.C552 0.0380 0.0274 0.10^ 0.C743 0.0555 0.C355 0.0252 0.0274 0.C242 C.C2S4 0.0246 C.C21C 0.0162 0.0129 -2.13 ;.13 1 .19 4.2c e.9i 0.126 C.126 C.126 0.126 0.126 0.0595 0.119 G.17e C.238 0.297 0.142 0.0950 C'.Ooi:' C.0381 0.0226 C.146 G.."93:J.0C3E 0.0392 0.0225 0.04c& 0.0296 0.2184 0.J2C7 0.0162 0.0126 o.onc 0.00635 - 3 . 35 4.6C O.OC _ •" u.j3or 0.0^9. ''.0112 3.00660 CC " - • ** -2.91 - te . •' ' 3.45 -3.40 - 0. ^ 1 •* C - 1 . .3 -\5f -'• 7£ -6.^9 - - , ; J " r•- " . u- ^ "^ 7 ^ 128 Table A . l . —r 1 (continued) I In : e r f a c e Ve l o c i t y , c-n/se: 1 1 t 1 Concentrat i o n o f P a r t i c l e s volume f r a c t i o r 1 Large S.Tiall P-edictt'd £ x p e n Tie;" I d 1 Predicted 0.200 0.200 0.200 0.200 0.200 C.C72S 0.120 0.159 J.199 U.C239 ..0793 C.055G 0.0406 0.0238 0.ei5V C.C695 0.0515 C.0382 0.0270 0.0179 0.335;: G.C225 0.0155 0.010^ 0.03^^92 LOwer I n t . : r f a c e d, 0 .0326 cm 0.0993 0.0149 0.209 0.268 0.328 0.0524 0.0624 0.06J4 0.0624 0.0524 0.165 0.121 0.0311 0.0528 C.C32B •^ r.^Zi'•i 0.122 0.0752 0.0523 0.0326 •Jcoar '^s - :3ev:afior inte-'ace Experi-'^rita 1 " , •^ * ~ 0.:22b 0.0137 c.oi:^ C.GC7jt :.ower Inter-'oce 13.4: 7 ~C 6.40 c . .'c 11 .^' •.DV9'^ Interface 4 '7 -".33 1 •^', G.03 -1.9/ 0.0081 cr, 0.0174 C.0i67 0.0157 0.0142 0.012C 0.0155 0.0175 0.017: 0.0155 0.G130 2 . •:•? 3 . 4» C'.5v 0.52 .-.12 - C . DC .-.'' Q' - -,. ; r - • . T>~ 129 Table A.2. Comparison of the Proposed Model with the Experimental Data for Binary Suspension in Diethylene Glycol I I n t e r f a c e Veloci t y , cm/sec Concentration of P a r t i c l e s volume frac; t i o n Large Small t ! Lower ! i t e r f a c e Predicted Upper In t e - f a c e Experimental Predictec - 0.0460 cm d^ = 0.0194 err. i Experimenta1 ^ ^s . . ^ A n ^ re'^N-cn. Deviation ucwer Interface Upper Interface C.102 0.153 0.204 0.255 0.306 0.0612 G.0612 0.0612 0.0612 0.0612 0.194 0.141 0.100 0.0696 0.0469 0.196 0.142 0.0931 0.0695 0.0477 0.0579 0.0531 0.0467 0.0385 0.0294 G.0525 0.0499 G.C442 C.G35'. 0.0313 -C.3G -0.5£ 2.41 0.00 1G.2" 6.42 r.l'l - 1.76 -6^0^ 0.102 0.153 0.204 0.255 0.306 0.122 0.122 0.122 0.122 0.122 0.130 0.0922 0.0635 0.0425 C.0274 0.131 0.0901 0.0602 0.0425 0.0280 0.0374 0.0321 0.0262 0.0200 0.0143 0.C343 0.0298 C.0250 0.0203 0.0151 -0.39 2. 35 5.5C O.OC -2.C3 8.9C 7.65 4.59 -\33 0.173 0.173 0.173 0.173 0.173 0.0714 0.112 0.153 0.194 0.235 0.115 0.0861 0.0630 0.0451 0.0315 0.112 0.0847 0.0620 0.0447 0.0317 0.0465 0.0326 0.0226 0.0155 0.0104 0.0438 0.0314 0.0216 0.0147 0.0106 2.9S ' .57 1.53 O.G' -0.63 6.07 3. 71 / ^^ 3^21 0.225 0.225 0.225 0.225 0.225 0.0714 0.112 0.153 0.194 0.235 0.0802 0.0586 0.0417 C.0290 0.0196 0.0761 0.0569 0.0409 0.0284 0.0188 C.0392 0.0261 0.C173 0.0113 0.00724 0.0379 0.0257 0.0173 0.0119 0.00811 = 0.0460 cn d - = 0.0137 cm 5.4C 2.91 3.00 2.11 4.22 «r ^ -- - T: 1 . ^ 3.36 1.55 ,-1 f; I-, -5.09 -Vz.h 0.102 0.153 C.2C4 0.255 0.3C6 0.0612 0.0612 0.0612 0.0612 0.0612 0.196 0.142 0.101 0.0701 0.0473 0.230 0.151 0.0992 0.0691 0.0472 0.0300 0.0285 0.0266 0.0239 0.0204 0.0348 0.0306 0.0268 0.0252 0.0228 -14.9C -5.84 1.S9 1.47 0.15 -13.91 -6.35 -C.91 -5.16 -10.73 0.0986 0.149 0.199 0.249 0.299 C.120 0.120 0.120 0.120 0.120 0.137 C.0977 0.0681 0.0462 0.0303 0.153 0.101 0.0657 G.C441 0.0301 0.0202 0.0184 0.0162 0.0136 0.0103 0.0211 0.0192 0.0162 0.0136 0.0113 -10.34 -3.28 2.08 4.68 0.6£ -4.13 -4.26 0.00 0.188 0.188 G.ISS 0.188 0.188 0.0707 0.110 0.150 0.189 0.229 0.106 0.0796 0.0587 0.0428 0.0303 0.103 0.0851 0.0569 0.0421 0.0300 0.0250 0.0182 0.0129 0.00901 0.00614 0.0264 0.0192 0.0127 0.00904 0.00632 2.50 -2.34 3.14 1.63 G.90 -4.50 -5.27 1.24 -0.28 -2.83 0.225 0.225 0.225 G.225 0.225 0.0714 0.112 0.153 0.194 C.235 0.0810 0.0595 0.0427 0.0299 0.0204 0.0803 0.0582 0.0414 0.0285 0.0194 0.0233 0.0161 0.0109 C.00720 0.00467 G.0248 0.0163 0.0111 0.00764 0.00503 0.81 2.19 3.09 4.99 -6.00 -1.50 -2.19 - 5 . 72 - 0.0326 cm d . - 0.0137 cm 1.46 1.34 3.66 £. 39 1C.62 - C. 6£ 0.00 Z. Ic - . .c \ 0.102 0.153 0.204 0.255 0.30C 0.0612 0.0612 0.0612 C.0612 0.0612 0.0983 0.0713 0.0506 0.0351 0.0236 G.0969 0.0688 0.0465 0.0317 0.0223 0.0290 0.0266 C.0234 0.0193 C.0148 0.0236 0.0263 :.0234 0.0189 0.0153 0.102 0.153 0.204 0.255 0.306 0.122 0.122 0.122 0.122 0.122 0.0659 0.0466 0.0320 0.0214 0.0138 0.0670 0.0435 0.0302 0.0202 0.C139 0.0187 0.0161 0.0131 0.0101 0.00720 0.0192 C.0161 0.0135 0.0105 C.00772 -1.59 0.17^ C.174 0.174 0.174 0.174 0.0714 C.112 0.153 0.194 0.235 0.0579 0.0432 0.0315 0.0225 0.0157 0.0536 0.0398 0.0288 0.0212 G.02S2 0.0163 G.0H3 0.00771 C.00519 0.0233 0.0162 n.0112 C.00813 C.005r£ 7.93 £.43 U. wI3 ^ -4.57 - J . ^ I 7 • r\: . w 6. G' 5.55 -C.£: 5.44 f. 3 3 4.55 -I. 53 G. OG -3.54 -"*..:.-6.7: -b . iC 0.45 G.5t * - . . C -£.23' 130 Table A.3. Comparison of the Proposed Model with the Experimental Data for Binary Suspension in Aqueous Glycerol 1 1 Interface Veloci t y . cr,/se: Concentration o'f' Particles volume •fraction Large Small 1 Lower Interface Predicted Experimental 1 ; 'Jpoer Interface Predicted C ^ ^ ^ A ^^ ^ *'M r L TI Deviation | Experimental Lower Interface upper Interface d. = 0.0326 cm dj = 0.0137 0.102 0.154 0.205 0.256 C.307 0.0615 0.0515 0.0165 0.0C15 0.0615 0.212 0.153 0.109 0.0756 0.C510 0.0237 0.0152 0.0111 0.0759 0.0521 0.0623 0.0571 0.0502 0.0515 C.031S 0.0678 0.0632 C.0561 G.0461 0.0343 -10.41 -5.36 -1.3G -1.5£ -2.15 0.102 0.154 C.205 0.255 0.307 0.123 0.123 0.123 0.123 0.123 0.142 0.0997 0.0687 0.0459 0.0297 0.152 0.101 0.0680 0.0459 0.0297 0.0400 0.0343 G.02S0 0.0214 G.C153 0.042S C.0359 C.0293 0.0229 0.0168 -6.59 -1.31 0.97 G.c: C.OC -6.-" .4.45 -4.58 -5.45 -£.56 0.174 0.174 0.174 0.174 0.174 0.0717 0.112 G.153 0.194 0.235 0.125 0.0938 0.0586 0.0491 0.0343 0.129 0.0962 0.0697 0.0494 0.0342 0.0500 0.0351 0.0244 0.0167 0.0112 0.0560 0.0385 0.0261 0.0180 0.0119 -2.2' -2.50 -1.50 -0.53 G.41 -13.76 -5.3" -6.53 -7.37 --..•; 0.226 0.226 0.226 0.226 0.226 0.0717 0.112 0.152 0.194 0.235 0.0872 0.0538 0.0455 0.0316 0.0214 0.0372 0.0538 0.0463 0.0312 0.0221 0.0422 0.0282 0.0187 0.0122 C.00781 0.0459 0.0306 0.02G0 0.0131 0.00861 G.OG O.OC -1.78 1.35 -3 31 -£.05 0.102 0.154 0.205 0.256 0.307 C.C614 0.0615 0.0515 0.0615 0.0515 0.214 0.155 0.0764 0.0515 0.0228 0.C219 0.0207 0.0191 0.0170 0.0253 0.0229 0.0219 0.02C4 0.0179 -C.35 1.12 7.91 5.96 4.55 -9.5^ O.no 0.216 0.153 0.102 0.0721 0.0493 0.102 C.154 0.205 0.256 0.307 C.123 0.123 0.123 0.123 0.123 0.145 0.102 C.0702 0.0470 0.0304 0.148 0.0978 0.0548 0.0427 0.0277 0.0151 0.0139 0.0124 0.0107 0.00364 0.0151 0.0132 0.0119 0.0102 0.0C845 -2.26 3.57 10.04 9.77 -o. 1 - -5.64 - . J . . . - -IC.09 - c . 72 -6.73 -6.9^ -5.2i -4* . ^ U -5.56 -0.35 -5.25 O.OC 5.16 4.24 4.43 2.30 131 Table A.4. Comparison of the Proposed Model with the Experimental Data of Smith (1965) on Binary Sedimentation 1 1 — — Interface Veloci t y , cm/sec 1 Concentra tion of P a r t i c l e s , __ volume f r a c t i o n 1 1 1 Large Predicted Small 0 Lower Interface Experimental ' Expe^-i mental Percent Deviat"on Upper Lower Inte'-'^ace Interface Upper Int erface Predicted = 0.G252 cm d . = G.0130 cm 0.099S 0.150 0.200 0.250 0.300 0.0250 0.0374 0.0500 0.0624 0.0749 0.208 0.142 0.0941 0.0602 0.0358 0.212 0.15S 0.0819 0.0545 0.0293 0.0915 0.0765 C.0580 0.0400 0.0254 0.0771 0.0714 G.0552 C.0400 C.020C -1.88 -15.37 14.92 1G.33 25.53 15.7: 0.0874 0.131 0.175 0.218 0.262 0.0437 0.0655 0.0874 0.109 0.131 0.200 0.133 0.0852 0.0529 0.0308 0.232 0.138 0.0334 0.0530 0.0290 0.0811 0.0630 0.0449 0.0295 0.0178 0.0740 0.0615 0.0531 0.0310 0.0191 -14.C7 ^3.57 2 21 = 3.24 6.25 26.52 9.55 2. 53 -15.53 -4.73 -£.?3 0.0624 0.093C 0.125 0.156 C.0624 0.0935 0.125 0.155 0.205 0.140 0.0915 0.0580 0.215 0.186 0.101 0.0655 0.0740 0.0560 0.0397 0.0265 0.C67C 0.0559 0.0465 C.0274 --.42 -24^79 -5.21 -11.38 IG.-C. 12 -1^64 -3.15 0.0437 0.0655 0.0874 0.109 0.131 0.0874 0.131 0.175 0.218 0.262 C.197 0.130 0.0825 0.0504 0.0289 0.248 0.138 0.105 0.463 0.250 0.0650 0.0462 0.0311 0.0199 0.0118 0.0635 0.0490 0.0346 0.0186 0.0104 -20.65 -5.67 -22.36 8.94 15.62 2.23 -10.04 6.72 13.08 <^L 7.•'6 ' 0.0252 dm d^ = 0.0187 cm 0.0998 0.15C 0.200 0.25C 0.500 0.0250 0.0374 0.0499 0.0624 0.0749 0.207 0.140 0.0924 0.0585 0.0355 0.246 0.191 0.0975 0.0620 0.0332 0.165 0.118 0.0792 0.0506 0.0307 0.153 0.119 0.0333 0.0521 0.0312 -15.92 -26.51 -5.26 -5.54 6.89 3.32 -0.72 -4.92 -2.97 -1.55 C.0S74 0.131 0.175 0.218 0.262 0.0437 0.0555 0.OS74 0.109 0.131 0.197 0.130 0.0823 0.0503 0.0286 0.235 0.168 0.0823 0.0455 0.0246 0.147 0.102 0.0657 0.0407 0.0235 0.137 0.105 0.0701 0.0409 0.0290 -16.75 -22.72 0.00 10.61 17.51 7.27 -3.31 -6.21 -0.59 -19.05 0.0624 0.0936 0.125 0.156 0.0624 0.0936 C.125 0.156 0.202 0.135 0.0370 0.0541 0.214 0.136 0.104 0.0555 0.139 0.0976 0.0647 0.0411 0.158 0.126 0.0720 0.0546 -5.69 -0.70 -16.61 -2.59 -11.73 -22.72 -11.28 -24.75 0.0437 0.0655 0.0S74 0.109 0.131 0.0874 0.131 0.175 0.218 0.262 0.191 0.124 0.0765 0.0454 0.0249 0.222 0.135 0.0771 0.0506 0.0286 0.125 0.0849 0.C545 0.0337 C.0195 0.131 0.0351 0.0601 0.0446 0.0195 -13.96 -3.45 -0.82 -10.37 -12.96 -4.47 -1.39 -5.32 -24.51 0.00 0.0250 0.0374 0.049? 0.0624 0.0749 0.0993 0.150 0.200 0.250 0.300 0.197 0.129 0.0817 0.0492 0.0280 0.206 0.139 0.0934 0.0563 0.0338 C.124 0.0851 0.0564 0.0353 0.0219 0.128 0.0980 0.0C07 0.0391 0.0268 -4.58 -6.£5 -12.59 -12.57 -"7.04 -13!21 -7.17 -5. 25 -18.44 132 Comparison of the Proposed Model with the Experimental Data of Mirza and Richardson (1979) on Binary Sedimentation Table A.5 1 r/\f%/»(a»*^fc-a • i rtrt f\^ Da»«*irlAc volume f r a c t i o n Large Small j I nterface Veloci ty, cm/sec i Percent D e v i a t i o n Lower In terface Predicted Experimental UoDer Inte'-'^ace Predicted 1 Experimental Lower Interface upper Interface * 0.0115 cm '^L • 0.0462 cm d. 0.201 0.246 0.289 0.335 0.380 0.0579 0.0679 0.0579 0.0579 C.0679 0.0288 0.0208 0.0145 0.0102 0.00685 0.0320 C.0227 0.0164 0.0114 0.00^38 0.00537 0.00498 0.00451 C.00387 C.00313 0.00523 0.00500 0.00434 0.00392 0.00328 -10.12 -S.35 -C.8S -1C.2C -7.05 2.63 -'". 35 3.53 -i'.24 -4.65 0.217 0.246 0.272 0.296 0.324 0.156 0.156 0.156 0.155 0.155 0.0130 0.0102 0.00813 0.00641 0.00499 C.0147 0.0111 0.0911 0.0630 0.0488 0.00241 0.00216 0.00152 0.00168 0.00144 0.00255 0.00223 0.00213 0.C0180 0.00156 -11.45 -£.00 -10.71 1.77 -5.35 -3.23 -9.31 - c . 7" -:.£1 C.308 0.308 0.308 0.308 C.308 0.0396 0.0652 0.0946 0.124 0.153 C.0161 0.0131 0.0102 0.00788 0.00519 0.0173 0.0141 0.0111 0.00336 0.00516 0.00581 0.00439 0.00318 0.00229 0.00138 0.00585 0.00449 0.00323 C.00254 0.C0158 -6.73 -7.01 -7.86 -5.79 -' T: '- z-> -'.'.. 0 5 - 2.2 C ,' c .* -9.7: ':Z.8i = 0.0327 cm d<; = 0.0115 cm •^ "\:: 0.183 0.234 0.285 0.333 0.385 0.0588 C.C688 0.0688 0.0638 0.0688 0.0163 0.0113 0.00769 0.00457 0.00326 0.0190 0.0134 C.00870 0.00590 0.00370 0.00509 0.00447 0.00365 0.00279 0.00201 0.00520 0.00433 0.00337 0.00315 0.00236 -14.37 -15.36 -11.59 -15.34 0.178 0.211 0.243 0.275 0.306 0.141 0.141 0.141 0.141 0.141 0.00992 0.00769 0.00553 0.00451 0.0C341 0.0117 0.00872 0.00664 0.00480 0.00374 0.00277 0.00243 0.00209 0.00175 0.00143 0.00296 0.00253 0.00233 0.0C193 0.00155 - 15. 20 -11.80 -10.66 -6.01 -6.82 -6.32 -7.57 -10.39 -9.56 -13.48 0, 303 0.303 0.303 0.303 0.3C3 0.0388 0.0676 0.0977 0.128 0.157 0.00850 0.00672 0.00520 0.00396 0.00301 C.0105 0.00344 0.00600 0.00460 0.00370 0.00438 0.00344 0.00242 0.00170 0.00120 0.00545 G.00375 0.00273 0.00209 0.00159 -19.05 -11.99 -13.33 -13.92 -18.74 -10.38 -8.55 -11.44 - 18. 74 -24.33 • — . - ' - -7.39 -4.57 -11.55 -14.72 = 0.0231 cm d<; = 0.0115 0.192 0.245 0.297 0.346 0.3S6 0.0563 0.0563 0.0568 0.0568 0.0558 0.00828 0.00559 0.00383 0.00255 0.0C153 0.00939 0.00595 0.00437 0.00282 0.00179 0.00470 0.00371 0.00273 0.00193 0.00128 0.00503 0.00381 0.00311 0.0C219 0.00153 -11.31 -4.31 -12.35 -5.31 -S.73 -7.53 -2.54 -12.17 -12.03 -16.G1 0.205 o!243 C.271 0 301 o'.327 0.135 0.135 0.135 C.125 0.135 0.00414 0.00305 0.00240 0.00184 0.00144 0.00442 0.00330 0.00247 0.00195 0.00163 0.00208 0.00164 0.00135 0.00108 0.00087 0.00230 0.00188 0.00154 0.00132 0.00112 -6.36 -7.61 -2.6" -7.55 -11.55 -9.^" • -12.59 -12.14 -18.^8 -21.60 0.315 0.315 0.315 c'.315 0.315 C.0402 C.0690 0.0953 0.124 0.152 0.00381 0.00298 0.00236 0.00180 0.00136 0.00413 0.C0333 0.00262 0.00173 0.00134 0.00296 0.00209 0.00153 0.00109 0.00078 0.00323 0.00241 0.00131 0.00126 C.00094 -7.68 -10.44 -10.04 1.07 1 .£5 -8.2; -13.28 1 C 1" -15 . .^' - 1 . ; . 2u -16.71 133 ^SlS' ~ c : ^ ~ r - . 0 ^ CNJ C C tN; — ^-. r^ - P»» cr~ n " -. - r^. —~. —-. o _ ^- oc ^c ^ ^ ^- ^s —- — C^ — i- O I I I I I I c u u a; <o 4-> OJ u C- 0> r r m r-> O^ CNJ CNJ o^ • • • «3- r n — «T ^ . . Q 2 C CNJ r o L n o - CNJ O .'^ U" r** ^£> O" t <o i l l OJ O) CL OJ 0)1 • - Til a' i-1 CJ +-> I— o -C u -a m <v 0) c • o CD 1 Ol'^ 1 — 1 (•) ^^ >^ •o sz <v a» to • ^ o Q u. C o •p— i- Q. CU c o 4-> CO MO c a; a. CO e o =5 OO •f— >1 «-SS^ &.5S?3i OOC5CC O O O d o o o o c o ci d> di o d S O C C ^ <6 t6 d d d ^ 2^'t; * m 00 CO i n CNJ o o o o o o o o o c S o o o o o i 5 ^',m ^^^22 s: \ ^ 1 >! { —) ! ij ' OJI lOI d c; kC ^r- f^ I— o O c O o s d d d d d CO o o 01 i 0)1 u; d d d d cv CC . CD a. X LU 00 ro c^ O II .. 01 r^^ u^ ca o ro r^ p^ c n o^ u") ^y. V.C i n o o r^ O O r^ ^ * roCT^cvj ^ . oc LT) ^ ^ ro CNJ O O O O C UJ -D ^ ' o o o o c O O O O O 1^ ^r CO O ^ r u o CNJ f ^ i n r^ o^ ^ CO LO LO «3- . - • r o r j CO CC r o L n O O .— CO 0 0 f— r ^ LO r o CNJ CNJ c o o o c c o d d d c 1 o OJ! s 0) 4.J , :>' ^•Pl 1 ! ai'i 1 !j: i., u •o S 0) u ^ 1 i.. as • ^ o ! CO •c 0), O CNJ VO CO VO VO vn >w >"» vD Ln r r O O O O o o o o c o ^ ^ r LD LO CC O i n ro . y r» VO <3- r o CNJ I— O O O O C O O O O O O C O O C 0 0 l"^ LO . 3 - OsJ 'O- T CO r o c r o CN; . — r — C "S- * ! • ^ O ' — UO — CO LD C-d _ _ O C CNi - ^ o VO CNJ O o c o o o o o o o c O C O O C — VO r*. r o o ^ LO O r— LD I— CO LO c\j . — — O C ro C ro . ^ i^ Lf) CNJ C> P^ L O Ol O O O O O '-D cr> '-- uo oil U) •a CJ i. sfO <o C L sz E io <v aji 31 Ol o XI O) o uo O^ LO r*^ .— LO C r o r>» r o O l r o CNJ — .— O c o o o o o o o o o ro Ln C — — c o e o o c c o d II OO GO CO CO CO <: = uo CD LT) UO i n UO un CC CO 00 CO oc uo in uo i n uo o o o o c Lf) uo i— Ln «o O o in o c O O O O c c c o o T "-• . 3 - «S- .?• o o o o c O O O O O O CO CC X 00 ra JD <o c u- c O •'c u O 00 00 00 IC 11 O O cn 00 u o u o LO u o o^ O O O O C O C O O C O C O O C o o o o o T i n LO CO CT^ — — « r u o LO CO a ^ C . a - CO <Nj LO — :— — (Nj rLj O r r CO r\j vo — CsJ CM d d d d d O O O O C CO o> .— CNJ T u o LD uo « — -^ r^ C O —— — o o o c c r^ r^ r^ r^ r^ O C O O C CNJ r ^ CNJ CNJ rv.' o c c c c 134 «T3 (V .^ 01 u o (/I c o _J ,M- 5o 4.1 — a v « v o u i u i r > » a 3 c v j ^ r > » « r v O v o r > . p n v o o » . — ^ • « \ j r o i o o r » » r ^ - - i O . — csiwrt.— o r * - ' — r ^ r ^ ' ^ i ^ < » ' ' ^ v ^ ' ' 0 ' ^ ' ^ S T 2 S — S'^?'^S^^F*'**''''^'*'^o'^'^'^cdtvJcodp^Ln'»^»'^f^dvoevJcvjvr>CNi<jvvrnro , T . ' - - < ^ < * ' ^ c v j — c v i i n V ^ «r CNJ ir> V i « Lrt ^ c n . — — — i — , ^ t^j ^ _ m — ro . - — — t^ O 1— ' > Ll. ' ' I I I I I I I I I I i I I • i 01 Q <o +J 1— 03 <T3 c Q O 0) '.J r- •r4-> a. ^ 03 i +-» QJ C > <u <jt —_ — SX ! 5. f l^^;SSSS, '' ^^; ^5 1 2 2 f i S «' ~ — ' ^ ' '• "' •^I o r '^— cO N fj ^ ro ™j rr ^o > o i r iv. s rr >> .our 3> 0. u0 3v 0O0O O«a'avo>(nvn«rcNj _ J ^ VSXS«ri3^ '^ ' xO^'j-aoO... o—.v«-n.o> ^ r >^ a p^ ^L 3 u n" r^t \"^* ^' — -c.N . > . . .— . . r ^— . ^o / . .S . .S ^ .^ . ^ v— '— c N j u ^ L r t v O O ^ r o O " ^"^ ~^' ^^0 '® .C N^j 'L ^ ' *f i. u*'?*''^. - ^.t..«...^ ' ' .-.= .— ' * * — --<^"'<«^'o S _' /S. . o <u r ^ c n o ^ ' V i S ' - o ^ ^ ' f o ' C T l - ^ u j c N i — r ^ l d i n c ; — — v o < ^ j d L o • u ^ • « T d a d — a J i n ' i ').— T l— — CNJCNJ— I I _ , I I — CNlCMll l l l l •a ..J Orn^vO^rOvOO OJ E -O i. O) 3 cr •r— " P - 4J s s Q..,- X LU _l u 1/1 e 00 u 1 « y» "O <V - 1 - -C •M « N .^l U _ .— 3 r— O CO +-> 4-> u C *J i. •r- 2 <V i- I— <V c 3 in o ^4 •o 01 (•) ^^, 01 Ol iITJ OO dII aOO>vovoaOvOLrtvor^vovovoroooovovor^'*«<^^|^c0^^fl00^r«.LOOVLO.-•CTlr^^o^^neo u rs -a X LjJ a. c O f^-» 0) u iQ U W — >^ c ja o o CO o o^ <^ .^ o o o > cxzz i. -J IB u 1 •.- a . 01 • i • i 1 11 CU 03 va — «rr^'^Ln>.C0 rOLf)«TCOo^CNJCNJOOvocorocOl^^CNJvoaoo^"TvovOO^voc^r^aOr^^ovoaoCNJ^«?^cOvcaocM^co."nmcoo^ ^ C N i C N i O ' ^ T T V O O J f — rOffiLOvnrOCNirOiOLnvO — O l T i L O T ' ^ " — O O ^ r O O N L T . rrO>'— C N i 0 0 O C T « T C 0 "^corooiooor^coroooovoroocNjoir.. — voo^*CT>*r^oooocMromaovocNio%ooO'—O — ^ C N I L O C O O <3^LOvovor«.r^u^^^aOl*»oo^^ro,^oovof>•f^f^C'^'^'^r^<3^^>»Lno^LnOO^<NJroo^vrtaO^—LOOVOLOVO^* LnvOi—Ovoi^oivoi^-CNJuni—<Ni«-p^LOr^roo>LO<nLn<y>0^'Of—CNj<3vfNivnoocNi«rvO'—0^u^—vfivoi^oo ^r^<j<<yicvj<y>p^(yi<— — o > v O v o O ^ O i L n r o i y > r o o > r o < j \ o v o r ^ r o L O O v r O L o a o r > . L O C N j O % L O L n v O L O r o . ^ O C N J O O ^ - C M O O i — C N i ^ - O O — CNJCNIOO — O C N J O ' — O — 0 0 < — C N J O C N I O — I — O i — o o . — O — • — — — ddddddddddddddddddddddddddddddddddddddddddd iITJ CNjmomcMOOr^Ovocriro^'OOCNjLOTf^.cvjvOf—vooooOLomLn^Lf)CMOLnOLC^rr>ro.TO^Or^^O'"»O^O^OCNJl^<NJ^o^-vooO^OrocOcNJ<rvOO^CNIOroo^(3^^r^r».CNl^-0-oOLa«T•u^«»ro.^r^rOLnro^ .— O O r o c N j q - — CNj(Njroco.T — •— — « r r r r n r o ^ - « s - r o n r o o o o . — — — O — O O O O O O O O O O O ddddddddddddddddddddddddddsdddddd^^ddddddod d (U s — 1 o u 1 u . ..> 1 -Q VI C U OJ lU .-a £ .1 a . 3 s- j a < _ f " S «+->.—' •^ •rocNjCNie\jCNi«rcNievjCNicNjoor».vo^-vo<a"COLOvoaOvOP»covOOCLO>0'^r«. 01 *J a. 0) 11 E 1 O <— c_3 < : ^r^OCNirocNjvororocvjCNiojLfi O m e o c N j C N i o o o > o r o r o o > ' — f — r o r o v o a O v o ^ ' C O o o o ^ - c \ i o c N i r o o % . — vr>p*.^Ovr)Cvjvouir..^-C»>—o^cvi e v i v o w ^ . — LO<J>^—OcvimLOroro^-,.— r v . L O ^ L n - ' ^ 0 0 > ' * » v n L n ^ r o i r ^ O O L f > ^ < n ' — ( B C N j — cOQLnooo^ r > » r O i ^ r o . — O ^ T L O O O C N i L O r O L o r o r r c o u i . — CNj>—rooo^Lrt — r o o v o e N i a o a » v o ^ ^ o o o r * . w ' ) u i a O O O ^ o » *»—^ a; CO -C 1 ^ + j cn 03 •£2'^'T;2'~!^'^'^®"''^^**"'^°0^*'*'^OCNir^r>o — o^o«romLooiaoo^Ln«TvornmvooLnrnnLO ^ v n o o p v p c o i — v O T i g o w c y i v o c n c N j r o ^ L o c o ^ r c o o o r ^ w . — — co<r>vcuo^^oio%Lna»«—c-»r«.Lpcivo i w ' ~ o % f ^ c s i ^ L O ^ » ^ o o o ^ L n r * ^ r ^ ^ ^ " v O v O i ^ * ri v oO vc vo ^O vo v0 o ^0 f^~r»^r roor o m^^— —^^ccnrOi O O T^<i ^ Oov vv v OO vO ^L vo OC i p^i^oCOro^ G o lU o 0) 03 II 5a • f " (J i D ror«,voropofnvorororo^-^r^'ro^-LO«-^if>mLnuiLnvouovfiLO'!rTLor>.Lr>Lni*»LOvOLOLnvouOLO>OP^ iq 4.i >j 4J C O ..— s J v ; ^ i P S ! 5 ' 2 ' * 5 ? ^ 2 * ' ' ~ ' » " < ' ^ ^ 0 0 « T . — ^ ( ^ ^ C N j O O ^ O O r o c y i r o . — a > v r o — saco — C M L S r o r o c M o d VO v/ QL. o ^S5l^.^5SS^2S2'5i5^'^^^!I!S<'''*~^"^"*o^Loov"^<^voooo<Tir>.«-vocviooLO<T>ao3>co — r* 03 <u i. a. 3 oQ . o O S- V Q. X LU •5 Vh. -o u o sz: <v 4-> OJ u a. i. Zi "O u 1 1— . w .^ r ^ O C N j v O O ' — O v o r o c N i O C N i r o o O s y i G ^ r O O o ^ r o a r ^ , — rncNjCNj00r~»Oi— C N i c o r o ( n < v j f ^ C f N i . — a i y i c o roc>ocNlCNif^.^r^r«.OLnLOCnvOf^rocNjCO"—OOLOCNICO^OLO'— OL.OC\JCNir^"3-rOTTro^O<r^COCNiCvj"oroo •2 L n L O v n r n n . ^ r r C N j ^ - — .— O ^ - r o u o c s j o ^ L i r — I — .-NJ.— — . — c M c v i r o c ^ r O . ' N j r y i . — CNICNJCNJI— C - — C C < — — ^ l/l ~ s 3 4) LJ ^ - 4-1 O 'O => a£ V 3> •a _j .— «y IJO D O 3^-POro.^^uO ori^ro-^ c o r o r o c N j i n a o c s i o ^ Ul c O i ^ O ' — c O « r c M L n r ^ c N l C O O ^ ^ r ' ~ « ^ ^ > 3 • c O L O ' ^ ^ ^ c c i r ^ o c - x o u o - o . — o ^ - s '—(TiLn^«^LOO<Njrn.a"LOvoro..— .^LOLnLnLOCO^vnuo-oGO — O L O ' ^ c o o > o r ^ X f ^ v O O ^ L n ^ r u o _ o u o u ••- it T w1 . —' US .- u j j a : :3 OJ e "" l/» 3 3 L ^ - O - r o^* o u—- •* C3 r> U o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o c o o c o o o o o — e M m ^ . ^ - . 3 - ^ t O L o r » . C 0 3 3 C 0 C T i < n c O ' — c N j c N j < r r o ( n i — >— > ' o . 3 • . J ^ ~ - a 3 X • T > a ^ c • — - o 3 % c — ' o m . - x j i " * ' ^ ' — — ^ . — ^— — •— .— .— ,— ^ 1 . ^ . — ^— . . • v j c N J C N i C M C N j c N j C N j c N j r o r o r o r o r o . ^ r n r o ^ o r o r j ' . x . T ^ u 0 . . o u o ^ L O 135 « e o ^» ^— ..— <S« *j •a </l — O eo <— 00 o% r o VO O O ^^OOvcNj.— Or^.CMCO^'^^OV'—COW"— OOO/^roocovovovovorovooooivriLrt CMO'^rOCVjCNJvOCM'— O f ^ O C V J ^ ' ' ^9V ^ coro^«»roro«TC^O^CNlO^00c^J^-p>•'—>— ir> r o r*. Ln fNi o» CM ,— r o »— '^^vrSvrt.^uoOr—rororoooc^^vOOvOU* C N i c M O C N i r ^ v r i T ' — c v i c v j c v i O v o c M ^ v o r * I , —. CM I — I I (M ov VO vo OS a% CO r^ GO CO CM •— ' ~ « . v o m a o c N j o v n ' 3 - — c o r > » < j > r o o o O vnr.. — L O r ^ o > ( N i c o u o ^ r ^ » — 0 O ' » ' ^ » « ' r—rocM<T>r»-oo«rvo — o v O ' — o < ^ o o o m ^ o o ^ r « . , O v o r o a 0 « T « a • 0 0 O V ' — c O « T i N j « - r o VO O >— «»• O rorororoO'—CNiOLn^-CMCM.3-rorof^ • I I I I I I I I I ro*?" I— — ...> at a t j c 01 <J u at a. 01 N . f v/> OJ o> u Lrt CO v o c*> U l OJ Q. X 0 9 0 \ O 0 « CM LO r— ov <n CM VO CO VO CM r^ — VJ m rv at o f^voocoro^'—vocMOOOOr>».^oovo CMaoo^r—r^O^rroooorOLO^—OrocM CM«^OirovOCOr^<nvo«TOuo«*--ro.— U O L n L O v O V O v O v O P ^ r ^ O O C O C O c T i O O C M — — — «• r>.» «3- Ul ^ CM OS CO CM O CM r^ O^ 00 VO -o <v H </» iDvoW'—oOo^'*»CNlr«.^...3•vocoo^np>» •O 0) a. C O o CO CO f^ CO 00 vO Lrt-—rouir.. C O U ^ L O C O O V C M ' ^ ^ T ^ ^ ^ — r ^ O N O O r ^ f M U O O O u o r o — . a r o ^ - u o o o o v aOCNj.TCOCOiXiOf*^^— O^OO^COO^^— - - 0 > r o ^ ^ r r o ^ - r o ^ L n ^ y u o u o - o u O L n P ^ i O i / ^ O d o 41 vj CM CO O v O . — O O r « . * C M C 0 O v c M v n u ^ < ^ ^ i / > 1 — 1 CM r— CM I '— r o c o r v O O ^ r f N i O L O O ' — f ^ L n c v n o rOr—LO^Or>.LOvoo^^vLncoocMOO •3"vovnLor^LOi^r»»r».oo<3iooo.— II LO •o L n c M r * . - L O L n o s - — C M C O v o r o O v n - — r o o o r v vOO^ro — r o o c o c v r ^ — v o ^ ^ r ^ X v O ^ r o CT\^-.TP^CM^C0^TO'—-"iTi'^f^cor^^ ro^»rrOLn^»rLOLr>LOLnLnvOLOvOP«.vo CNI^ s o s0) Q. X CO CTI T r o f»» CO ^ - 0 0 >— — CM r o VO v / i . — O O O — • CM CM CM CM CM CM u O LO Ol <— rvf>»00O'— CNjr^cocMCMLO'—TCOOoO pv r v CM 0> vn ^ * r o r o r o r v c-i r o CM r o L O r o vO.»roLO«rOCMvoooocNicMvnLnrvai vo^vcoo^OO^—^--—^yvn^vococo^.— . — — r— . — . — — . — . — — , — CM o < O' u <v so ro q00 00 »— ro rv. TT O CM .— ro CM 00 '— CM O LO « • »3- 0 0 O O^ CNJ . — —I -O Ln ^ CMvncO'—corof^^coo.— ^vncMro.— c o ^ voO^TCMCMuOvOLn^TCMCOcoro^uoovO roLOrooorvLnrvo»rooi'—00>«TfMLO»r Lnvnvo^'rvLnvo'^^^ooiPvovo^aoooov r~ d d OJ E (j II CMvni—00>vnvOCTir«.cMCMCMLn.— VOCM •^o>o%LnvovoooroooLO — CMr—ococn T — LncMrocMTvn^''3-oo<?»f^o»^r— VO 1 ^ 00. O i O O " — — —.f—^-.— CMro^rrovnrvcu.™ ^ - — .— — — _ 'O LOCMCNjinoivn'—CM — cocTvpvf^cM^-cvjrv. r v ^ r O C M O v x o m v O T O v P v . — O r v r o o 0 « T r o v o u o v o v n L n r v o x e M v n < n r o — o> LOvOvOvn^vOrvoOrvovrvOiOCOOCMO^ — ^-cvj 03 c e o -•<- O •— vn «.» O) ( J o » CO r v . ^ I O Ov LO O ^ * CO o o —o • J .— UJ o O v O O O - — r o , —O O L O I — rvLn^OvocMf—..— CNJLOCNJr— — .— vocovovnvn-—otvn CMrOCMrOCMCMCM<Mr— .— — ,— — ^ — O O i ^ ^ o e o o O v o — cM^eo.— ^ c o o ^ O f ^ p^LOmcO — rvu^cviLOrvLOr—ovrororv^— r - r — CM — CM<— ' ^ - — .— , . Q , _ — . O ^ - d ddd d dddddd<d<ddddddddd ddddddddddddddddd L n o ^ » r ^ r o L n L n « r o o c M o o a o o i r o o o C M a > r v r - » e O r o o o > C M r v ^ j - O O C M C M C M ' — C M O — O O i — — o > — O O O O O O O CMCMCM^rr^^vn^—oOCMLnvo'^rouoaO'— CM*—OCMCOaof^*LOLnvouorooNroLnvOO CMCM.— C M O . — - - — ' O i O, r - O O . — ddddddooodoooooo ddddooooooooooooo CM 1 ^ C ^ O vO r— r o o PO O rorvLnvorOLnvn^TGOOOrOr.— r v r o o v n <j>rv.^^—OirvCTirooCM-—CO^'vOOrv ^— C O r v o —o m o LnoDTrvuirocMrocM.'^eM'—'— r - — . o ro.TLO'— L n O C M C M O < T O ' ^ — r ^ . — ^ . l O CO L O V O O — vOOuOLnrvLOrv.— -ncMOO''—0'OLn^*cocMroi*vLOO^Lnrv .yrOLOrOPv^pv-suof^^r^vnuoCMroro -o o «C i. « V IO E U EX 3 01 o» ' — . — CM « • . — CM r o CM CM CM O vt- lO o o o o o (d <d <d d d 01 I . — O (J 2 U IQ U r- a. at ^ l O O i C O O V O O C M O v O V . T ' L n r v ^ O ' — • O O O ^ C M r v ^ — r o L n r O L n r o O v o r— <j\ Ln O r O O O • J n . -^^ 0^ o r o S B 3 o; (J 01 Ol r— 4.1 CM cj< . y o vr) o <a u ..- rouo^CNi l O C M O r v ^ r O C O O L O O N U i r v L O L r t r — CM C 0 C 7 V < T ^ T T V O r v . ^ T T ^ - ' O L n . T r o ^ ^ . C0^rC0^rvnr**LOvO^vLn^vO<MCMCMr^ Ol 1- 4J -O V oc •s at a .— 1/) 3 I u . ~.. — o ro O — "•- S > L I . O VJ o o o o o O O O O O O O O O O O O O O O O ocooocoocccoooooc vO VO 0 0 0 > CO IO v o vO vO r ^ Ol'—ONOUoovrnfv.rOGOCMLnOPvLnvo '— C N j c M r o r o r o ^ T ' ^ u O L n v o v o o o o o o t ' ^ rruo.v.- — .— .— . — C M . n L C O O r o r v c M - . o o r v o % 0 C M C M C M C N j r O r O r O r O ^ T T . y . r r ^ T L n I 136 III = ... o r— IO '—. *J «0 (/) > <^ *J .— O O r o c s i ^ - r v r r ^•CMOvo*J-ro^-»T ^.TLO^rro^yr^LO OJ IM c 01 u L. .^ VI OJ o 0) cn^-CMCMrv^^-in .— mr^vocMr^Poro &. ^ IO rv^runrocM—-om at >—CMrvosfv..— O r ^ mcoo> — Orv.cMf^ OvOVVOCM^..— ^ 4 T rvvovooorvr«.aoao a. X •o 01 •o at "O <v i. ro^no^'—CMrv.ocM vor»«oicM«-oiincM OCMO^rvp— O O GOrvrvotCOGOO^Ov Z3 (0 4.) o u c Lnm«j-cocMrooro "s-vocM.— Oini~»ON vorvcNimcM^O^ u ^ - 0 0 « » ^ - — ^ro 01 1/0 a. X r>v 0) -o <: Ol I.' 01 u ... <v •a JQ 03 moooroentnco^ rvtoo«ri^a*i—»— ^ • C M G O O G O r v . . — CM CMr— O U O r O C M ^ T T a. voroo^CM^rcMO^'^• cootr—ro^rvLOuo lO O vrt 4-1 •— CJ u 4-> — (O IO u ^ E I/) OO — o o o o o dddddddd c t- u OGO^lOf^ONi— fv ^•vom^m^cMro O O O O O O O O OJ IO £ U O. 3 C r— o>*- o u o > dddddddd a> s — o u O^rocM.— nri*v^Ln oiinorv^-OCMO — O O O O O — O ( J l e VJ • - Q . 01 it/i 4.> > ^ > v . cy oro_ £ s 5 0) u 01 .— C M G O r o o i ^ m m Ujro^CMr^^Pv^- — 4.1 O IO IO CMfOCM^r^CM^-ro 9> o : u • . - OJ S- -tJ - o 4 J 10 •— Oj QC 3 3 — U U S i Z V.V O ro O — «*- E ^ I— o u O O O O O O O O oicMrovoo^—.— pv vnvovovorvr^rvGO 137 c •!-> o •r- c 0) 4J o &(U 03 Q_ ' O 1— r^ r*. 1 — V D CTt •r" • • <LO • • • cy» crt CO CO p*» > o I I I T ) r— V O L O cr* VO 1*^ ix> 00 CNJ CXD 1— 1— 1 1 CO L n L O VO VO r*. CO o 1 — CM 1 — «do 1 1 1— r»v un o^ o O 00 CO r^ 1— ro 4-> C CU «v icit >> CO) to > o (U 'f- CJ 03 E CU •1i- O) 4-> Q . =3 X XI LU -r- S. CU •*-> sz CO 4J Mi- -r- -a =3 o =3 C -r- •a <U -M CO c o o CLC_) o s- . c +J d a •r— 1—H X3 CU CO CU 03 E £ c: O CU CO CL CO •r- S- 3 03 OO CL 00 •r- +-) c s- o 03 • ^ CL +J o £ CO r— CJ CU S s- o o ••- 1 — n— •r- (/) C r^ r^ 1 ^ cri o 1— CD in ^ 1 — f ^ CO i r ^ CO • ^ CTi O CO 0 0 ^ CM CM . — O f— I— r— o o O O O O O IT) o o o o o o •r+J o o o -D SZ <o 1 ^ cr> VO •o £ 03 r— "vl- L O r ^ o o o o o r— r— r— r— i — O O O O O LD o o o o o <;^ "vT •?!- ^ ^ scu JZi E rj 03 1— r— o o o o o o o o o o o o o o o o o o 00 HO .£= 4-> VO 1 — C\J ca_ 4-> o r^ si- cTv 1— VO VO CO cn «;j- O r^ un CO ^ •M ^ S o CU <v 00 r—" <U CO o ^ o o o o c Qu CU N O O O O O IT) O VO CM CT> CO Q^ O ^ 0> ^ CM CM 1 — O -o o 4-> •1- O O O O O 1— CO o cyv f— 1— o CO CO COCT^r^ 1 ^ CO 1— o o o o o o o o o o o c» Cy> CX3 VO «^ o o CJ -r- .c OJ CM O CO r^ CO o^ r - 1— o CU ^sec u Exp rim 03 4J 03 (Tl LD o un o LO O 1— f— CM CM -c «:*• «v^ ^ £ 03 o o o o o CO o o o o o i- S- <v JD E zs £ cu x^ E Z3 £ s_ £ £ £ £ CU o •r— O o •r- +J U •M M- E Z2 03 r— S_ o +-> > c <v u c o CJ> •r— +J CU Cl s- o 03 su_ u <0 <o i- 1 — <;:f L D r ^ CJ^ t n o LO o in O S- U- CX3 r— 1 — C\J CM o o o o o r— CO r^ o or— or— or— or— 1— >vi- un r-v cT". un o un o un O I — 1 — CM CM o CJ >» r— • • • • • o o o o o <3 o o o o o -t^f ^ - "vf «;*• " ^ i - 03 _l «d- -cf u_ <v £ E So o CU CJ) <+- sz • -(-> • <c •r— 00 CO 03 E OO CU E 03 •r— Q CO C>-) CO CO CO o o o o o o o o o o <^ 'd- «::^ ^ «^ cn C3^ cn CTi 0"i o o o o o o o o o o <u un un un un LD , , r— 1 Q 1 •• o o o o o o o o c c 5 *•— Suspending med XJ fo 1— ;:: ter of Particle CU E o •^ r«v r«« r»v rv. r^ <X) CO ex: CO cx) <u cn %. <o -J o o o o o VO ( O V O CO «:!- «!i- 'si- « ^ ILO ' ^ O O O O O o o o o o o o o o o r>v r^ r>v r>v r>». CO CO CO CO C>0 CM CM CM CM CM CO CO CO CO c^ 1 1 o o o o o o o o o o O O O O O 1 o o o o c 1 n— 1 ^— •K 138 £ i +-> o £ ^• CU 4-> 03 u i- ^ <v > CU C O C M C O "!d- VO 0 ^ 0 ^ VO CO c o CO 0 0 r ^ cyi VO r>v cn un • • • • VO r— CO CTv O CM c n CO CX3 o ^ VO CTi O VO 1— 1— CM CM l l l l ' ^ r»v c T i C M V O r— r— f — C M C M 1 1 1 1 1 r»« 0 0 r ^ r ^ CM 1 1 1 1 1 &(U « * CM CTI CJ^ L n CO CM o 0 0 VO CO 00 o 00 cn r» o en CO o ^ ^ v n CO CM CM 1— CO L n o r ^ CM 1— r— C Q. X LU 1— o c c » • • • c o c o o o o o o c o o o o o c c c c o o o o c CVJ c n CM 1— r— CO L n "5? a > VO CO CM CM CO •^ C Q. Q , i- o It- <o 4-J £ CU #1 OS 4-> 03 Q >, +J •r— o o r"~ CU CJ 03 > £ <U E •riCU-K Q. £ X O UJ •!— -M <U Z3 x : jzi 4-> • ! - CU CJ 03 4SO) +J £ cn 00 E •r- CU CO i^v Ln c c "^ O Vv^ E u T3 CU +-> O •f— "O CU iQ. 1—1 c o c o • • • • c o c o CO CO u n r-^ CO r^ LO CO 1— ^ - CTi VO ' ^ CM r— C C C 1— 1— 1— CO CM CO VO o CO CM f— 1— O O O O o c c c c o c c c c c o o o o &o 1— CU (U M "O • ! O OO I^ r— 03 1/1 <V r— E OO CM u •r— CO X J 13 CU O CO :3 O £ CL-fO -M i- SZ 4-> sz S- o 03 • I — Cl. + J a. o c_> <u £ O CU •r— E 4-> ZJ <0 r— i- o 4-> > £ CU J£ . £ + J 4-' •r— MS O CO £ £ O O CO •!— • 1 - CO S£ 03 CU CL CL E (/5 4O X3 £ 03 o 03 i4- u E Z3 • f— -a CU s: CU cn J03 _J £ o C_3 CO 0 0 0 0 CO r— p ^ r— r— m w) in IT) o o o o • • • • o o o o m CO 0 0 0 0 CO CO -o Ln Ln Ln un un c o o o o c o o o o £ 03 ^ r^ CO -o 1— -^a- un r-s en LO c un c LO C 1— 1— CM CM r— r— 1 — 1 — 1 — £ 03 VO n r\ r— CO un scu sz i- S_ CU XJ 1 — n— r— 1 — E un un un un Z3 £ £ O •r— 4-> CJ 03 SLL. CO 0 0 oc CO o o o o • • • • o o o o c: Z5 £ £ O «N ^ " ^ ^ ^'^ <^ c o o o o c o o o o "vT un I ^ r— o un o 1— 1— r— CM CO • • • * o o o o u_ <v E JD Z3 £ £ o +-> u « ^ <=i- <vr " d - ^ c o o o o o o o o c •r— •1— -i-J CJ fO S- o c c c c 00 r— <vf un r-v cn un c un o un C r— I— CM CM c o o o o <o iLi_ CO 0 0 CO CC 0 0 p— r— r^ o 1 ^ ^— CJ un un un un un Vfc_ o o o o o o c c c c -^ o o :3 f— C_) O O • cr» • <: CU r-" J3 03 1— E u r» ^_ _ 03 E C>0 VO CO CO V O CM C M C M C M CO CO CO CO o o o o • • • • o o o o c o o o o CO CO CO CO CO CM CM CM CM CM O C O O C ^— n— t— r— 1 — O C O O o c c c c o c c c c r ^ r~v r ^ !>-« «a- "51- ' ^ «vf < ^ •vT - ^ " ^ ' ^ "=d- 0 0 0 0 0 0 CO pvv r^ r^ pv. r^ cn cn cn cn en ^— r— r— 1— r— o o o o . . . . o o o o o c c c c c o o o o O o o o VO VO v c VO «::t «d- " ^ ^ v c VO VO VO v c CM CM CM CM CM CO CO CO rr> CO o o o o • • * • o o o o o c c c c o c c c c O OO CM O r^ u U(- E 3 •r— -o <v s: CU 4-5 •r— Q . § 1 -r— 1 -a 1 m cn cn en CM C M C M C M CNI C C C C o c c c c O a' E 03 Q C (/) CU •r— +J S03 Q- 4J CU CO CO CO CO CO VO VO VD VO v c <u cr. s. 03 _l o C C C C CO CO CO CO CM CM CM CM C C C C o c c c c OJ —cn £ ispend 2 CO 1 — lene J £ 4-) + J CO •r— 'r— Z2 OO •¥• APPENDIX B COMPUTER PROGRAM FOR PREDICTION OF INTERFACE VELOCITIES IN MULTISIZED PARTICLE SUSPENSIONS 139 1-^0 c c c c c c c c c c c c c c c c c c c c c c c c r c c c c c c c c c c c c c c c c c c c c **:^*^ LIST PROGRAM SYMBOL C D DT EPSLON MP NZ PN REo ROHF ROHP ROMS SCS TCOMC uc UT UTI VISF OF c>KlNCIPAL VARIABLES ***•• DEFINITION CONCENTRATION OF PARTICLES PARTICLE DIAMETER TUBE DIAMETER VOIDAGE NUMBER OF PARTICLE SIZES NUMBER OF ZONES INDEX N PARTICLE REYNOLDS NUMBER DENSUITY OF THE FLUID DENSITY OF THE PARTICLES DENSITY OF THE SUSPENSION SUM OF CONCENTRATIONS OF SMALLER SIZE PARTICLES TOTAL PARTICLE CONCENTRATION SETTLING VELOCITIES OF =>A.^TICLES TERMINAL VELOCITY OF A PARTICLE IN A FINITE FLUID MEDIUM TERMINAL VELOCITY Or A PARTICLE IN AN INFINITE FLUID MEDIUM VISCOSITY OF THE FLUID COMPUTER PROGRAM **:4ca)c;Jc:^:ic ^^ ^ ^^^^ DIMENSION D(10),C{10»10),UTI(10),UT(10),REP(10), ilPNdG) ,UC(10,10)tC0LD{10)tRaHS(10),CNE'/(10) *«* READ DATA **• READ NUMBER OF PARTICLE SIZES READ(5,1G0)NP NZ=NP READ PARTICLE DIAMETER AND CONCENTRATION DO 1 1=1,NP READ(5,101)D(I)tC(ItNZ) CONTINUE READ PHYSICAL PRQC^ERTIES READ(5,102)R0HP,R0HF,VISF,DT ^^"^ WRITE DATA WRITE(6,201) WRITE(6,202) WRITE{6,203)ROHF WRITE(6,204)KQHP WRITE{6,2C5 )VISF *** Ul WRITE(6,206)DT WRITE{6,207) WRITE(6,208) WRITE(6,211) DO 7 I=1,NP WRITE(6»2 09)I,0(I),C(1,NZ) C C C C C CONTINUE *** COMPUTATIONS FOR THE LOWEST Z J N E ^** CALCULATE TOTAL CONCENTRATION TCONC=0.0 DO 2 I=ltNP TCONC = TCONC'»-C{I,NZ) CONTINUE CALCULATE VOIDAGE C c C EOSL0N=1.0-TC0NC C C * CALCULATE TERMINAL VELOCITIES AND THE N-INDICES * FOR PARTICLE SIZE KSMALLEST SIZE) c c c UTI(1)=D(1)*0(1)*981.0*(ROHP-ROHF)/(13.0*VISF) UT(1)=UTI(1 )*( ( 1 .C-0.^73*D( l)/DT)/( 1.0-D(1)/DT) )*'^(-4) REP( 1)=R0HF*UT(1 )=)'D ( 1 )/VI SF RE9=REP(1)**0.9 PN( !)=( 5. 10+0.27*RE9)/(1.0-»'0.1*RE9) C c C DO 4 FOR P A R T I C L E 1=2,NP SIZES 2,3,...tNP IS=I-1 SCS=0.0 DO 3 J = 1 , I S SCS = SCS-»-C( J , N Z ) CONTINUE ROHS( I ) = ( EPSLCN*ROHF+RQri?^SCS)/{EPSLC!N-»-SCS) UTI(I)=D(I)*u(I)*931.0«{R0HP-R0HS(I))/(18.0*VlSF) U T U ) = U T I ( I )=^( ( 1 . 0 - 0 . 4 7 5 * D ( I ) / D T ) / ( 1 . 0 - D ( 1 ) / D T ) ) ' « ( REP(I)=ROHF*UT(I)*D(I)/VISF RE9 = REt>( I )=!=*0.9 PN(I) = ( 5 . l O + 0.27='RE9)/(1.0-»-C.l*KE9) CONTINUE * CALCULATE S E T T L I N G V E L O C I T I E S * C C c 5 6 DO 6 1 = 1 , N P SUM=0.0 DO 5 J = 1 , N P SUM = SUM + UT( J)*EPSL0N*=5^( °N ( J ) - 1 . 0 ) ^C ( J , N'Z ) CONTINUE UC( I , N Z ) = U T ( I )=^E^SLGN«=*{PN( I ) - l . 0 ) - S U M CONTINUE WRlTE{o,21Q) WRITE(6,212) 142 WRITE(6,200 )NZ,UC(NP,NZ) C C «** COMPUTATIONS FOR THE REMAINING ZONES ITER=NZ-1 DO 24 K=1,ITER NP=NP-1 NZ=NZ-1 c C C ASSUME INITIAL CONCENTRATIONS AS CCRRESOONOING CONCENTRATIONS IN THE ZONE LOWER TO THE PRESE.NT ZONE c c c c c c c c c c c c 8 9 10 EPSLON=1.0-TCONC CALCULATE TERMINAL VELOCITIES AND N-INDICES 12 r c c c c c c DO 8 1=1,NP C( I,NZ)=C{I ,NZ + 1) COLD(I>=C{I,NZ) CONTINUE CALCULATE TOTAL CONCENTRATION TCONC=0.0 00 10 1=1,NP TCONC=TCONC+C(I,NZ) CONTINUE CALCULATE VOIDAGE 11 c c *^* 13 00 12 1=2,NP IS=I-1 SCS=0.0 DO 11 J=1,IS SCS = SCS-^C( J,NZ) CONTINUE ROHSC I )=( E^ SLON'J'ROHF+ROHP*SCS)/{EPSLGN-^ SCS) UTI(I)=D(I)*D(I)*931.G*(RCHP-R0HS(I))/{18.0*VISF) UT(I)=UTI(I)*((1.0-0.^7 5-D(I)/DT)/(1.0-0(I)/DT))**( R E P d )=ROHF^UT(I )*D(I )/VISF RE9 = REP(I)^^0 .9 PN(I)=(5.10+0.27«RE9)/(1.0+C.1*RE9) CONTINUE ASSUME VELOCITY OF THE LARGEST SIZE ^ARTICLE TO BE THAT OF IT IN THE ZONE LOWER TO THE PRESENT ZONE UC(NP,NZ) =UC (N«^,NZ-»-l ) UCOLD=UC(Nt>,NZ) CALCULATE CONCENTRATIONS ACCORDING TO EQATION(3.34) DO 14 1=1,NP C ( I ,NZ) = (UC(NP-H,NZ-»-l )-UC{I ,NZ-H) )^C( I,NZ-H )/ C(UC(NP-^l,NZ-»-l )-UC(N^,NZ) ) 143 C C 14 CONTINUE 15 CALCULATE TOTAL CONCENTRATION TCONC=0.0 DO 15 1=1,NP TCONC = TCONC-^C( I ,NZ) CONTINUE c c c c c c c c CALCULATE VOIDAGE EPSL0N=1.0-TC0NC 16 c c c c c c c 17 18 19 c c c c 20 21 22 24 CALCULATE THE SETTLING VELOCITY OF THE LARGEST SIZE PARTICLES SUM=0.0 DO 16 J=1,NP SUM=SUM>UT(J)*£PSLON*«(PN(J)-1.0)*C(J,NZ) CONTINUE UC(NP,NZ)=UT(NP)*EPSL0N^*IPN(NP)-1.0)-SUM COMPARE THE NEW SETTLING VELOCITY WITH THE OLD SETTLING VELOCITY FOR THE LARGEST SIZE PARTICLE UCNEW=UC(NP,NZ) ERROR=(UCNEW-UCOLD)/UCNEW IF(A3S( ERROR) .LE.O.ODGO TO 17 UCOLD=UCNEW GO TO 13 COMPARE THE NEW CONCENTRATIONS WITH THE OLD ONES OERROR=0.0 DO 18 1 = 1,NP CNEW(I)=C(I,NZ) ERR0R=(CNEW(1)-COLD(I))/CNEW(I) ERROR=ABS(ERROR) IF(ERROR. GT.OERROR)OERROR = ERROR CONTINUE IF(0ERR0R.LT.0.01)G0 TO 20 DO 19 I=1,NP COLD(I)=CNEW(I) CONTINUE GO TO 9 CALCULATE SETTLING VELOCITIES FOR ALL PARTICLE SIZES DO 22 1 = 1 ,NP SUM=G.O DO 21 J=l,r4P SUM=SUM+Ut( J)*EPSL0N**(PN(J)-1.0)=*C(J,NZ) CONTINUE UC(I,NZ)=UT(I)*EPSLON-*(PN(J)-1.G)-SJM CONTINUE WRITE(6,200 )NZ,UC(NP,NZ) CONTINUE 14: C C c 100 lOi 102 200 201 202 203 204 205 206 207 208 209 210 211 212 FORMAT STATEMENT FOR READ AND WRITE FORMAT( I D FORMAT( 2F10.4 ) F10.3,F10.3,F10.2) FORMAT( F10.2, 12,21X,F7.3) FOR^AT( /,15X, X,»»** INPUT DATA ^^**) FORMAT( ///,19 FORMAT( / / / / / , 3X,'PHYSICAL PROPERTIES: ) •FLUID DENSITY, GM/CUBIC CM = FORMATt F5 .3 ) FORMAT( //,5X, PARTICLE DENSITY, GM/CUBIC CM F4 2) ?F5 F 0 R M A T ( /,5X,» FLUID VISCOSITY, POISE = •,F5.3) TUBE DIAMETER, CM = •,F4.2) FURMAT( /,5X, •3Xf'PARTICLE DIAMETER AND CONCENTRATION:* FORMAT( /f5X,» DIAMETER CONCENTRATION' FORMAT( / / / / / , ,»SIZE NUMBER FORMAT( ///,5X 2,9X,F6 4,8X,F6.4) X,'ZONE NUMBER FOKMAT( INTERFACE VELOCITY') CM VOL. FRACTION') FORMAT( /,9X,I FRACTION 1H1,10 FORMAT( 38X, ) 20X. 'CM/SEC • STOP END i; **• INPUT DATA ••• PHYSICAL PROPERTIES: FLUID DENSITY, GM/CUBIC CM = 1.165 PARTICLE DENSITY, GM/CUBIC CM = 2.43 FLUID VISCOSITY, POISE = 0.136 TUBE DIAMETER, CM = 3.14 PARTICLE DIAMETER AND CONCENTRATION: SIZE NUMBER DIAMETER CONCENTRATION CM VOL. FRACTION 1 0.0081 0.0518 2 0.0194 0.0518 3 0.0460 0.1040 146 *** ZONE NUMBER OUTPUT *** INTERFACE VELOCITY CM/SEC 3 0.32009 2 0.08517 1 0.02056