To simulate dispersed two-phase flows CFD tools for predicting the local particle number density and the size distribution are required. These quantities do not only have a significant effect on rates of mixing, heterogeneous chemical reaction rates or interfacial heat and mass transfers, but also a direct relevance to the hydrodynamics of the total system, such as the flow pattern and flow regime. The Multiple Size Group (MUSIG) model available in the commercial codes CFX-4 and CFX-5 was developed for this purpose. Mathematically, this model is based on the population balance method and the two-fluid modeling approach. The dispersed phase is divided into N size classes. In order to reduce the computational cost, all size groups are assumed to share the same velocity field. This model allows to use a sufficient number of particle size groups required for the coalescence and breakup calculation. Nevertheless, the assumption also restricts its applicability to homogeneous dispersed flows. We refer to the CFX MUSIG model mentioned above as the homogeneous model, which fails to predict the correct phase distribution when heterogeneous particle motion becomes important. In many flows the non-drag forces play an essential role with respect to the bubble motion. Especially, the lift force acting on large deformed bubbles, which is dominated by the asymmetrical wake, has a direction opposite to the shear induced lift force on a small bubble. This bubble separation cannot be predicted by the homogeneous MUSIG model. In order to overcome this shortcoming we developed an efficient inhomogeneous MUSIG model in cooperation with ANSYS CFX. A novel multiple velocity multiple size group model, which incorporates the population balance equation into the multi-fluid modeling framework, was proposed. The validation of this new model is discussed in this report.

A pool boiling experimental facility has been designed and built to investigate nucleate pool boiling in water under sub atmospheric pressure. Liquid crystal thermography, a non intrusive technique, is used for the determination of surface temperature distributions. This technique uses encapsulated liquid crystals that reflect definite colors at specific temperatures and viewing angle. Design of the test section is important in this experimental study. Since a new TLC is required for every new set of test conditions, a permanently sealed test section is not an option. The real challenge is to design a leak proof test section which is flexible so that it can be taken apart easily. A plexiglass test section, including a top chamber with an internal volume of 60.9 x 60.9 x 66.4 mm and a bottom plate of 5.5mm thickness is designed and assembled together using quick grips. In the test section, water is boiled using 85.0mm x 16.0mm and 0.050mm thick Fecralloy® as the heating element. The TLC sheet is attached to the bottom plate and the heating element is placed on top of TLC so that the temperature distribution of the heating element during boiling can be interpreted from TLC. A camera system fast enough to capture the thermal response of the TLC and an arrangement to capture both hue of the TLC and growth of the bubble on the same frame has been designed and successfully used. This system allowed recording of position, size and shape of the bubble with synchronized surface temperature. In order to get hue vs. temperature relation, in-situ calibration of the TLC is performed for each test condition with the present experimental setup and lighting conditions. It is found that the calibration curve of the TLC at atmospheric pressure is different from the calibration curve of the same TLC at subatmospheric pressures. The maximum temperature difference between the two curves for the same hue is found to be only 0.6°C. The experiment is run at four different test conditions of subatmospheric pressure and low heat flux. It is run at system pressures of 6.2kPa (0.89Psi) and 8.0kPa (1.16Psi) with a constant heat flux of 1.88kW/m2 and 2.70kW/m2, and a constant heat flux of 2.70kW/m2, 3.662kW/m2 and 4.50 kW/m2 respectively. Analysis of nucleating surface temperatures using thermochromic liquid crystal technique is performed for these test conditions and the bubble dynamics is studied. The temperature distribution is quite varied in each case and the temperature is at its maximum value at the center of the bubble and it decreases radially from the center. The dry spot observed during the experiments indicates that the process of evaporation of the microlayer is dominant at subatmospheric pressures. It is observed that at very low pressure and heat flux the bubble growth is accompanied by the neck formation. Boiling parameters such as bubble frequency, bubble size and contact are also analyzed and a summary of these results for four different test conditions is presented and the relevant differences between the cases are discussed and the effect of increase in pressure and heat flux is noted.

La simulation de réacteurs à bulles en régime industriel est un grand défi. L'objectif principal de ce travail est la prédiction de la taille des bulles à l’aide d’un modèle numérique de bilan de population, basé sur la modélisation des phénomènes de brisure et de coalescence, et pouvant être couplé aux conditions hydrodynamiques présentes dans les réacteurs. Différentes données expérimentales sont obtenues pour valider le modèle. La taille des bulles est mesurée à l'aide d'une technique innovante de corrélation croisée. Les essais, réalisés en eau du réseau (partiellement contaminée) et en eau déminéralisée avec ajout éventuel d'éthanol, montrent que les additifs réduisent la coalescence et diminuent la taille moyenne des bulles. Deux distributeurs du gaz différents sont utilisés pour découpler l'étude de la brisure et de la coalescence. Les données expérimentales sont utilisées initialement pour valider des simulations CFD 3D transitoires Eulériennes-Eulériennes. La loi de traînée est corrigée par un facteur de swarm pour intégrer l’effet d’une fraction de gaz élevée. Différents modèles de turbulence sont testés. La contribution de la turbulence induite par les sillages de bulles au mélange de scalaires est évaluée. Enfin, pour prédire la taille des bulles, un bilan de population est couplé au modèle hydrodynamique préalablement validé et est résolu par la méthode de quadrature des moments (QMOM). Un set original de kernels de brisure et coalescence est proposé, capable de prédire la taille des bulles pour différentes conditions opératoires. Le comportement du modèle lors de l’extrapolation des réacteurs est également examiné / The simulation of bubble column reactors under industrial operating conditions is an exciting challenge. The main objective of this work is to predict the bubble size, in turn interconnected to the reactor hydrodynamic conditions, with computational models, by modelling bubble breakage and coalescence. Experimental data is collected for model validation, including bubble size measurements with an innovative cross-correlation technique. Experiments are carried out with tap water and demineralized water, with or without the addition of ethanol, and gathered results show that additives reduce coalescence and lower the mean bubble size. Two different spargers are used, in order to decouple the investigation of breakage and coalescence. The experimental data set is used to validate out unsteady three-dimensional Eulerian-Eulerian CFD simulations. A drag law for oblate bubbles is considered, together with a swarm factor, that accounts for the swarm effect. Several turbulence models are tested. The contribution of bubble induced turbulence (BIT) to scalar mixing is assessed. To predict bubble size, a population balance model is coupled to the hydrodynamic model and is solved with the quadrature method of moments. A set of breakage and coalescence kernels is proposed, capable of predicting the bubble size for different operating conditions. Scale-up effects are also investigated

Les colonnes à bulles sont largement utilisées dans les domaines du génie chimique et biologique, grâce à leur configuration simple, exempte de toute partie mobile. Néanmoins, leur extrapolation aux échelles industrielles engendre des modifications de l’hydrodynamique globale (vitesse du liquide, taille des bulles) qui sont encore difficile à prédire avec les outils numériques disponibles.La thèse a pour objectif d’établir une base de données sur l’évolution radiale et axiale de l’hydrodynamique locale (taux de vide, taille de bulles, vitesse liquide), dans différentes tailles de colonnes allant de 0.15 à 3 m de diamètre, pour des vitesses superficielles gaz comprises entre 3 et 35 cm/s, générant des taux de vide atteignant les 35%. Les mesures de taux de vide local, de vitesse de bulles et de la taille verticale des bulles sont réalisées à l’aide d’une sonde optique 1C. De plus, une nouvelle méthode pour mesurer la taille horizontale des bulles à fort taux de vide et en écoulement fortement multidirectionnel est proposée dans cette thèse. Cette méthode est basée sur la corrélation croisée spatiale de signaux provenant de deux sondes optiques placées parallèlement à la même élévation, et à une distance l’une de l’autre devant nécessairement être plus faibles que les bulles les plus petites présentes dans l’écoulement. Les mesures de taille de bulles sont validées en les comparant à un traitement d’images par endoscopie. Pour des vitesses superficielles de gaz supérieures à 9 cm/s, un bon accord est trouvé entre les trois méthodes (sonde optique 1C, corrélation croisée et endoscopie). La taille des bulles augmente légèrement lorsque la vitesse superficielle gaz augmente, par contre elle n’est pas impactée de manière significative par le diamètre des colonnes. Une plus grande ségrégation radiale est tout de même visible dans les plus grandes colonnes testées.Un modèle 1D radial développé pour un écoulement invariant le long de l’axe de la colonne est utilisé pour tester différents formalismes de forces de trainée, utilisant les données expérimentales de taille moyenne de bulles. Les simulations montrent que pour prédire correctement le flux gazeux expérimental, il est nécessaire d’introduire un « swarm factor » (Simonnet et al, 2008) diminuant le coefficient de trainée à fort taux de vide. De plus, des simulations 3D URANS avec Fluent® sont réalisées avec la loi de trainée validée par le modèle 1D précédemment cité. Un bon accord est observé entre les valeurs expérimentales et simulées des profils radiaux de taux de vide et de vitesse liquide, pour des diamètres de colonne allant de 0.4 m à 3 m, et pour des vitesses superficielle gaz de 3 à 35 cm/s. / Bubble columns reactors are widely used in chemical and biological engineering due to its simple configuration without mobile parts. However, the scale-up prediction of a bubble columns reactors is still a challenging process, due to the lack reliable experimental data and models.The present work aims to construct detailed database of the radial and axial evolution of local hydrodynamics properties (gas hold-up, bubble size and velocity, liquid velocity) acquired in several bubble columns in a scale factor of 20 (from 0.15 to 3 m in diameter), for a superficial gas velocity from 3 to 35 cm/s, yielding gas hold-ups up to 35 %. Measurements of local gas hold-up, bubble velocity and bubble vertical size are performed by a 1C mono-fiber optical probe. Moreover, a novel method to measure mean horizontal diameter of bubbles at high void fraction and in a multi-directional flow is proposed. This method is based in the spatial-cross correlation of signal of two optical probes placed parallel side by side, at a given distance from each other, at the same elevation in the column. The validation of the bubble size measurements are performed through a comparison of the results with an endoscopic imaging method. For superficial gas velocities higher than 9 cm/s, a good agreement is found between the three methods (1C mono-fiber optical probe, cross-correlation and endoscopic imaging). A slight increase is registered with the increase of the superficial gas velocity, however there is no significant variation with the column diameter.A 1-D radial model of a bubbly flow (Ueyama and Miyauchi, 1979) developed for a invariant flow along the column axis, is used to benchmark several classic formalisms of the drag force, using experimental average bubble size. Results show that to correctly predict the experimental gas flowrate, it is necessary to use a Swarm factor (Simonnet et al, 2008) that reduces the drag coefficient for high gas hold-up values. Moreover, Fluent® 3D URANS simulations are performed using the previously validated drag force formalism. A good agreement is found between experimental and simulated radial profiles of gas holdup and liquid velocity for column diameters ranging from 0.4 m up to 3 m in diameter in a range of superficial gas velocities from 3 cm/s to 35 cm/s.

The use of nanofluids as heat transfer fluids has received a lot of attention from the heat transfer research community. Due to the increased thermal conductivity of nanofluids over their base fluids, the number of nanofluids scientific publications increased significantly in the past decade. The effects of the heated surface roughness, nanoparticles and surfactant concentrations on pool boiling of nanofluids have been thoroughly investigated. However, contradicting findings have been observed under what appeared to similar test conditions. In this experimental investigation, two boiling surfaces have been prepared with an average surface roughness of 6 and 60 nm using high precision machining. Alumina Oxide-Water based nanofluids have been used in this investigation. The initial nanoparticle size reported by the manufacturer is 10 nm. The nanoparticles concentration has been kept at 0.05 wt. %. A Sodium Dodecylbenzenesulfonate (SDBS) surfactant has been added to the nanofluids in order to improve its stability. Results showed that the nanofluids boiling performance depended on the boiling surface roughness. The heat transfer coefficient (HTC) obtained in the case of the smooth, mirror finished surface showed an enhancement of 205% with respect to pure water. This trend was reversed in the case of the rough surface which is believed to be due to significant nanoparticles deposition. The HTC obtained with the rough surface was 12% lower than that of pure water. The effect of the surfactant concentration on nanoparticles deposition has been investigated by changing the surfactant concentration from 0.1 to 1.0 wt. %. In the case of the rough surface, the increase of surfactant concentration was found to reduce the formation of the nanoparticles deposition layer. The HTC obtained with the higher surfactant concentration was increased by 46 %. The effect of nanoparticles concentration on the smooth surface shows an unexpected trend of 20 % reduction of the transfer rate of the nanofluids coupled with the increase of the nanoparticle concentration from 0.05 to 0.1 wt. %. However all concentrations showed heat transfer enhancement with respect to pure water. The minimum heat transfer coefficient ratio enhancement was 11 % using 0.1 wt. % nanofluids with respect to pure water. Since nanoparticles deposition has been observed and attributed to micro-layer evaporation, an investigation has been carried out to examine the nucleation process during the pure water and nanofluids pool boiling. The bubble growth rate in both cases was analyzed at different wall degrees of superheat ranging from 104.3 to 105.9 ºC. In addition, the bubble departure diameter and frequency have been measured and compared for both cases. The nanofluid bubble size was about 80 % smaller than that of pure water. The nanofluid bubble departure had almost constant frequency of 500 Hz over the range of wall superheats whereas the maximum bubble frequency in the case of pure water was 22.72 Hz. / Thesis / Master of Applied Science (MASc)

The collisions frequency of dispersed phases (particles, droplets, bubbles) in a turbulent carrier phase is a fundamental quantity that is needed for modeling multiphase flows with applications to chemical processes, minerals flotation, food science, and many other industries. In this dissertation, numerical simulations are performed to determine collisions frequency of bi-dispersed particles (solid particles and bubbles) in homogeneous isotropic turbulence. Both direct numerical simulations (DNS) and Large Eddy simulations (LES) are conducted to determine velocity fluctuations of the carrier phase. The DNS results are used to validate existing theoretical models as well as the LES results. The dissertation also presents a CFD-based flotation model for predicting the pulp recovery rate in froth flotation machines. In the direct numerical simulations work, particles and bubbles suspended in homogeneous isotropic turbulence are tracked and their collisions frequency is determined as a function of particle Stokes number. The effects of the dispersed phases on the carrier phase are neglected. Particles and bubbles of sizes on the order of Kolmogorov length scale are treated as point masses. Equations of motion of dispersed phases are integrated simultaneously with the equations of the carrier phase using the same time stepping scheme. In addition to Stokes drag, the pressure gradient in the carrier phase and added-mass forces are also included. The collision model used here allows overlap of particles and bubbles. Collisions kernel, radial relative velocity, and radial distribution function found by DNS are compared to theoretical models over a range of particle Stokes number. In general, good agreement between DNS and recent theoretical models is obtained for radial relative velocity for both particle-particle and particle-bubble collisions. The DNS results show that around Stokes number of unity particles of the same group undergo expected preferential concentration while particles and bubbles are segregated. The segregation behavior of particles and bubbles leads to a radial distribution function that is less than one. Existing theoretical models do not account for effects of this segregation behavior of particles and bubbles on the radial distribution function. In the large-eddy simulations efforts, the dissertation addresses the importance of the subgrid fluctuations on the collisions frequency and investigates techniques for predicting those fluctuations. The cases studied are of particles-particles and particles-bubbles collisions at Reynolds number Re_λ = 96. A study is conducted first by neglecting the effects of subgrid velocity fluctuations on particles and bubbles motions. It is found that around Stokes number of unity solid particles of the same group undergo the well known preferential concentration as observed in the DNS. Effects of pressure gradient on the particles are negligible due to their small sizes. Bubbles as a low inertia particles are very sensitive to subgrid velocity and acceleration fields where the effects of pressure gradient in the carrier phase are dominant. However, particle-bubble radial distribution functions from LES are not as low as that from DNS. To account for the effects of subgrid field on the dispersion of particles and bubbles, a new multifractal methodology has been developed to construct a subgrid vorticity field from the resolved vorticity field in frame work of LES. A Poisson's solver is used to obtain the subgrid velocity field from the subgrid vorticity field. Accounting for the subgrid velocity fluctuations (but neglecting pressure gradient) produced minor changes in the radial distribution function for particle-particle and particle-bubble collisions. We conclude from this study that for accurate particle tracking in LES the subgrid velocity fluctuations must be dynamically realizable field (temporally and spatially correlated with the large scale motion). Adding random SGS velocity fluctuations is not enough to capture the correct radial distribution functions of dispersed phases especially for bubbles-particles collisions where the pressure gradient term ( or acceleration Du_f′/Dt) is responsible for particle-bubble segregation around particle Stokes number near one. A CFD-based model for minerals flotation machines has been developed in this dissertation. The objective of flotation models is to predict the recovery rate of minerals from a flotation cell. The developed model advances the state-of-the-art of pulp recovery rate prediction by incorporating validated theoretical collisions frequency models and detailed hydrodynamics from two-phase flow simulations. Spatial distributions of dissipation rate and air volume fraction are determined by the two-phase hydrodynamic simulations. Knowing these parameters throughout the machine is essential in understanding the effectiveness of different components of flotation machine (rotor, stator or disperser, jets) on the flotation efficiency. The developed model not only predicts the average pulp recovery rate but also it indicates regions of high/low recovery rates. The CFD-based flotation model presented here can be used to determine the dependence of recovery rate constant at any locality within the pulp based on particle diameter, particle specfic gravity, contact angle, and surface tension. / Ph. D.

Experimente mit neuartigen Messverfahren lieferten Daten über die Struktur von transienten Flüssig-keits-Gas-Strömungen für die Entwicklung und Validierung von mikroskopischen, d.h. geometrieunabhängigen Konstitutivbeziehungen zur Beschreibung des Impulsaustauschs zwischen Flüssig-phase und Gasblasen sowie zur Quantifizierung der Häufigkeit von Blasenkoaleszenz und -zerfall. Hierzu wurde eine vertikale Testsektion der Zweiphasentestschleife MTLoop in Rossendorf genutzt, wobei erstmals Gittersensoren mit einer Auflösung von 2-3 mm bei einer Messfrequenz von bis zu 10 kHz angewandt wurden. Dabei wurde die Evolution von Gasgehalts-, Geschwindigkeits- und Bla-sengrößenverteilungen entlang des Strömungsweges und bei schnellen Übergangsprozessen aufge-nommen und so die für die Modellbildung erforderlichen Daten bereitgestellt. Für den Test der Mo-dellbeziehungen wurde ein vereinfachtes Verfahren zur Lösung der Strömungsgleichungen entlang des Strömungswegs erstellt. Es basiert auf der Betrachtung einer größeren Anzahl von Blasengrö-ßenklassen. Die erhaltenen numerische Lösungen haben erstmals gezeigt, dass der bei Erhöhung der Gasvolumenstromdichte stattfindende Übergang von einer Blasenströmung mit Randmaximum zu einem Profil mit Zentrumsmaximum und anschließend zu einer Pfropfenströmung ausgehend von einem einheitlichen Satz physikalisch begründeter und geometrieunabhängiger Konstitutivgleichun-gen modelliert werden kann. Die Modellbeziehungen haben sich in einem abgegrenzten Gebiet der Volumenstromdichten als generalisierungsfähig erwiesen und sind für den Einbau in CFD-Modelle geeignet. Weiterhin wurden Arbeiten zur Kondensation durchgeführt, die direkten Bezug zu den Kon-densationsmodellen haben, die in Thermohydraulik-Codes enthalten sind. Die Untersuchung liefert darüber hinaus experimentelle Daten für die Modellvalidierung hinsichtlich des Verhaltens und des Einflusses nichtkondensierbarer Gase. Hierfür wurden spezielle Sonden für die Bestimmung der Konzentration und für die Lokalisierung von Pfropfen nichtkondensierbarer Gase entwickelt und bei transienten Kondensationsversuchen in einem leicht geneigten Wärmeübertragerrohr eingesetzt.

In gas-solid fluidized beds, the generation of electrostatic charges due to continuous contacts between fluidizing particles, and the particles and the fluidization vessel wall, is unavoidable. Industrial operations, such as the production of polyethylene, are susceptible to significant operational challenges caused by electrostatics including reactor wall fouling, a problem known as “sheeting”. The formation of particle sheets can require shutdown periods for clean-up which results in significant economic losses. To gain a better understanding of the underlying mechanisms of electrostatic charging in gas-solid fluidized beds, in an attempt to eliminate or minimize this problem, a pilot-scale pressurized gas-solid fluidization system was designed and built, housing an online electrostatic charge measurement technique consisting of two Faraday cups. The system permits the study of the degree of particle wall fouling at pressures and temperatures up to 2600 kPa and 100°C, respectively, and gas velocities up to 1 m/s (covering a range including turbulent flow regime). The system also allowed, for the first time, the measurement of the fluidizing particles’ mass, net charge and size distribution in various regions of the bed, especially those related to the wall coating under the industrially relevant operating conditions of high pressures and gas velocities. Experimental trials were carried out using polyethylene resin received from commercial reactors to investigate the influence of pressure and gas velocity on the bed hydrodynamics and in turn, the degree of bed electrification. Mechanisms for particle charging, migration and adherence to the column wall were proposed. The size distribution of the gas bubbles shifted towards smaller bubbles as the operating pressure was raised. Thus, higher pressures lead to greater mixing within the bulk of the bed and resulted in a higher degree of particle wall fouling. Moreover, the extent of wall fouling increased linearly with the increase in gas velocity and as the bed transitioned to turbulent regime, due to the increase in particle-wall contacts. Bipolar charging was observed especially within the wall coating with smaller particles being negatively charged. Overall, particle-wall contacts generated negatively charged particles resulting in a net negative charge in the bed, whereas particle-particle contacts generated positively and negatively charged particles resulting in no net charge when entrainment was negligible. The formation of the wall layer and its extent was influenced by the gravitational and drag forces balancing the image force and Coulomb forces (created by the net charge of the bed and the metallic column wall as the attraction between oppositely charged particles).

Experimente mit neuartigen Messverfahren lieferten Daten über die Struktur von transienten Flüssig-keits-Gas-Strömungen für die Entwicklung und Validierung von mikroskopischen, d.h. geometrieunabhängigen Konstitutivbeziehungen zur Beschreibung des Impulsaustauschs zwischen Flüssig-phase und Gasblasen sowie zur Quantifizierung der Häufigkeit von Blasenkoaleszenz und -zerfall. Hierzu wurde eine vertikale Testsektion der Zweiphasentestschleife MTLoop in Rossendorf genutzt, wobei erstmals Gittersensoren mit einer Auflösung von 2-3 mm bei einer Messfrequenz von bis zu 10 kHz angewandt wurden. Dabei wurde die Evolution von Gasgehalts-, Geschwindigkeits- und Bla-sengrößenverteilungen entlang des Strömungsweges und bei schnellen Übergangsprozessen aufge-nommen und so die für die Modellbildung erforderlichen Daten bereitgestellt. Für den Test der Mo-dellbeziehungen wurde ein vereinfachtes Verfahren zur Lösung der Strömungsgleichungen entlang des Strömungswegs erstellt. Es basiert auf der Betrachtung einer größeren Anzahl von Blasengrö-ßenklassen. Die erhaltenen numerische Lösungen haben erstmals gezeigt, dass der bei Erhöhung der Gasvolumenstromdichte stattfindende Übergang von einer Blasenströmung mit Randmaximum zu einem Profil mit Zentrumsmaximum und anschließend zu einer Pfropfenströmung ausgehend von einem einheitlichen Satz physikalisch begründeter und geometrieunabhängiger Konstitutivgleichun-gen modelliert werden kann. Die Modellbeziehungen haben sich in einem abgegrenzten Gebiet der Volumenstromdichten als generalisierungsfähig erwiesen und sind für den Einbau in CFD-Modelle geeignet. Weiterhin wurden Arbeiten zur Kondensation durchgeführt, die direkten Bezug zu den Kon-densationsmodellen haben, die in Thermohydraulik-Codes enthalten sind. Die Untersuchung liefert darüber hinaus experimentelle Daten für die Modellvalidierung hinsichtlich des Verhaltens und des Einflusses nichtkondensierbarer Gase. Hierfür wurden spezielle Sonden für die Bestimmung der Konzentration und für die Lokalisierung von Pfropfen nichtkondensierbarer Gase entwickelt und bei transienten Kondensationsversuchen in einem leicht geneigten Wärmeübertragerrohr eingesetzt.

Operações de produção de petróleo, que utilizam sistemas de bombeamento centrífugo submerso, constantemente encontram a presença de gás livre nos poços, o que pode gerar ou agravar problemas como cavitação e falhas dinâmicas, quando o gás é succionado pela bomba. O separador gravitacional do tipo shroud invertido é uma solução possível para este problema nos casos de operação em poços direcionais de petróleo. O trabalho tem como objetivo apresentar um estudo do diâmetro das bolhas que ocorrem no interior de um separador gravitacional gás-líquido do tipo shroud invertido. A finalidade é o aprimoramento de um modelo fenomenológico do funcionamento deste tipo de separador, aplicado à indústria petrolífera, sendo que o modelo fenomenológico garante total separação de gás, fornecidas determinadas condições. O modelo prevê, através do cálculo da energia cinética turbulenta, o tamanho médio das bolhas carregadas para o seio do líquido por aeração no duto anular. Partindo de estudos anteriores, uma verificação do modelo fenomenológico foi feita utilizando um aparato experimental com misturas bifásicas ar-água e ar-óleo em três diferentes inclinações. O aparato possui dimensões radiais reais de um poço de petróleo offshore. Foi utilizado um sensor 3D ORM para a medição do tamanho médio (sauter) das bolhas arrastadas pelo líquido até a entrada do tubo de produção, em diversas combinações de vazões da mistura água-ar. Esta medição permite o ajuste das correlações que regem o modelo fenomenológico, no que diz respeito às equações dependentes do diâmetro teórico de bolhas arrastadas pela fase líquida. As descobertas provenientes deste estudo foram implementadas em um código computacional que será utilizado pela PETROBRAS, financiadora do projeto, para suas operações de bombeamento. / Oil mining operations powered by centrifugal submersible pumping systems suffer constant setbacks due to the presence of free gas in wells. Decompression in the reservoir liberates this gas in the form of bubbles that, upon reaching the suction end of the pump, cause cavitation and dynamic failures resulting in production and equipment losses. The Inverted-shroud gravitational separator is a possible solution to this problem. This work presents a study on diameters of bubbles that occur inside this separator. The goal is the improvement of the understanding of this kind of separator as well as the enhancement of a previously reported phenomenological model, which ensures total gas separation when the separator is installed in directional wells and under specific operational conditions. Empirically adjusted correlations are used to ensure that all entrained bubbles do not reach the pump. The model was tested for two-phase flows of water-air and oil-air mixtures using three different inclinations. Tests were performed with an experimental apparatus that simulates a pilot-scale well casing with an inverted-shroud separator installed. A 3D ORM particle-size sensor was employed in order to measure the average (sauter) diameter of entrained bubbles that are dragged by the liquid flow towards the end of the production tube. This investigation will be used to improve the reliability of the phenomenological model and reduce its dependency on a theoretical prediction of the bubble size. The findings were incorporated to the final version of an in-house gas separator design software developed at the request of PETROBRAS, the project funder and Inverted-shroud patent holder, for usage on its oil mining operations.