La diminution de la qualité des collectes des papiers récupérés ainsi que l’accumulation de substances dissoutes dans les eaux de procédés affecte l’efficacité des lignes de désencrage industrielles et contamine davantage les effluents liquides. Dans ce contexte, le LGP2 a développé depuis quelques années un procédé innovant de désencrage, la flottation réactive à l’ozone, afin de dégrader chimiquement les polluants dissouts en parallèle de l’élimination de l’encre. Afin de mieux comprendre les mécanismes mis en jeu, des essais de flottation à l’air et à un mélange ozone/oxygène, sur trois contaminants modèles, sélectionnés après une étude bibliographique préalable, ont été réalisés dans un milieu diphasique gaz/liquide, en l’absence de fibres cellulosiques. Les expérimentations ont été conduites sur deux pilotes de laboratoire instrumentés : une colonne à bulles fonctionnant avec de l’air uniquement, pour l’étude du comportement hydrodynamique (taille et distribution de bulles, rétention gazeuse) en présence des contaminants dissouts, et une deuxième colonne à bulles, similaire mais conçue avec des matériaux résistants aux gaz corrosifs, dédiée à l’étude des réactions d’oxydation en présence d’ozone. L’examen du comportement hydrodynamique montre que les conditions de débit de gaz et d’injection retenues conduisent à des tailles de bulles optimales pour une flottation efficace, que ce soit en présence ou absence de contaminants. Ces conditions obtenues avec de l’air ont été transposées en première approximation au système ozone/oxygène. L’étude du transfert de l’ozone et de sa réactivité avec les trois contaminants modèles, à différentes températures et concentrations en ozone, a conduit à la détermination des constantes cinétiques de réaction et a montré que les contaminants étaient, selon leur nature, oxydés ou dépolymérisés. Bien que la DCO des solutions traitées diminue très peu après la flottation réactive à l’ozone, la qualité des effluents est améliorée sur le plan de leur biodégradabilité. / The decrease of the recovered paper collection quality and the accumulation of dissolved substances in process water affect the deinking line efficiency and contaminate more and more the liquid effluents. In this context the LGP2 has developed an innovative deinking process, the ozone reactive flotation, to chemically degrade dissolved pollutants in parallel with ink removal. To better understand the mechanisms involved, air and ozone/oxygen flotation trials have been conducted on three model contaminants selected in a preliminary bibliographic review, in a two-phase gas/liquid system, in the absence of fibers. Experiments have been carried out on two instrumented laboratory pilots: a bubble column operating only with air for the study of the hydrodynamics of the reactor (bubbles size and distribution, gas hold-up) in the presence of dissolved contaminants, and a second one, similar in its conception but built using materials resistant to corrosive gas, dedicated to the study of the oxidation reactions with ozone. The evaluation of the hydrodynamics related to gas flow and injection system selected, studied with air but supposed to be the same with ozone/oxygen gas mixture, shows that the bubble size, with or without contaminants, is optimal for an efficient flotation process. The study of ozone mass transfer and reactivity with the three model contaminants, for several temperatures and ozone concentrations, leads to the calculation of kinetic constants and shows that the contaminants, depending on their nature, have been oxidized or depolymerized. Although the COD of the treated solutions does not decrease a lot after the ozone reactive flotation, the effluent quality has been improved in terms of biodegradability since contaminants are partially degraded.

Atualmente, o tipo de propelente para produtos em aerossol mais usado no mundo é uma mistura de hidrocarbonetos leves (butano e propano - sendo o primeiro em maior proporção). Parte de seu processo de produção é a dessulfurização do butano líquido, através de adsorção em leito fixo usando peneiras moleculares 13X. A literatura científica não apresenta muitas publicações sobre o tema e considerando que a técnica de adsorção é fortemente dependente de dados experimentais para seu maior entendimento, é objetivo deste trabalho estudar este processo através do equilíbrio de adsorção, das curvas de ruptura e do desenvolvimento de um modelo matemático que represente o funcionamento de uma coluna de adsorção de leito fixo a temperatura constante. Obtiveram-se dados sobre o equilíbrio de adsorção por meio de ensaios em banho finito comparando-se a interação de diferentes compostos de enxofre com a peneira molecular 13X e o efeito da temperatura. Utilizou-se a técnica de cromatografia gasosa como método analítico para obtenção dos teores de compostos sulfurados e de hidrocarbonetos. O modelo de Langmuir apresentou bom ajuste aos dados experimentais. Avaliando-se a interação dos componentes sulfurados com a zeólita 13X, identificou-se diferenças significativas, sendo a maior em ordem decrescente: etil-mercaptana, n-propil-mercaptana e terc-butil-mercaptana, respectivamente. A dinâmica do processo de adsorção foi estudada através da obtenção das curvas de ruptura em leitos em escala laboratorial e piloto. Avaliaram-se as influências da variação da concentração de entrada da n-propil-mercaptana e da velocidade do fluido em leito fixo recheado com zeólitas 13X, mantendo os demais parâmetros constantes. Conforme esperado o aumento da concentração inicial reduz o tempo de ruptura, aumenta a quantidade total adsorvida pelo leito e não altera o comportamento da zona de transferência de massa (ZTM). Analisando-se a elevação da velocidade, nota-se também uma diminuição no tempo de ruptura e um aumento da ZTM. O modelo matemático apresentado considera os balanços de massa microscópicos aplicados ao leito, os fenômenos de transporte de massa com modelo de dispersão axial e transporte por convecção da fase líquida para a superfície da partícula e as isotermas de adsorção. As equações diferenciais parciais resultantes foram adimensionalizadas e resolvidas empregando-se o método de diferenças finitas, implementado por código Matlab®. A partir da simulação matemática das condições experimentais obtiveram-se os parâmetros de dispersão axial e de transferência de massa que possibilitaram uma boa reprodução do tempo de ruptura quanto do perfil da zona de transferência de massa para os experimentos em escala laboratorial e piloto. / Currently, the kind of propellant for aerosol products most widely used in the world is a blend of light hydrocarbons (butane and propane - the first in greater proportion). Part of the production of propellant is the desulfurization of liquid-phase butane by molecular sieves 13X in a fixed bed. The scientific literature concerning the adsorption of the mercaptans using zeolite are scarce and considering that the adsorption technique is strongly dependent on experimental data for its better knowlegde, the aim of this work was to study this process by adsorption equilibrium, the breakthrough curves and the development of a mathematic model, simulation and comparison with the operation of a fixed-bed pilot and laboratory-scale column. Adsorption equilibrium parameters were obtained using finite bath experiment and comparing the interaction of different sulfur compounds with molecular sieve 13X and the temperature effect. It was used the gas chromatography as an analytical method in order to obtain the levels of hydrocarbons and sulfur compounds. The Langmuir model well fit the experimental data. Significant differences were identified in the interaction of the sulfur components with zeolite 13X. The major interactions of the sulfur components are in decreasing order: ethyl-mercaptan, n-propyl-mercaptan and terc-butyl-mercaptan. The dynamic of the adsorption process was studied by obtaining the breakthrough curves in laboratory and pilot scale. It was investigated the influence of the initial concentration of n-propyl mercaptan and the fluid velocity in a fixed bed packed with zeolite 13X keeping the other parameters constant. As it was expected, as the inlet sulfur concentration increases the break point time decrease, and enhances the total amount adsorbed by the bed. Analyzing the increase in velocity on the breakthroug profile, it was noted that also decreases the break point time and causes a greater decline of the curve resulting in greater ZTM and anticipating the bed saturation time. The model equations account the mass balance applied in the flowing liquid phase in the column, transport phenomena as the effect of axial dispersion and convection from liquid phase to the adsorbent surface and adsorption isotherm. Finite difference method was used to solve the dimensionless general partial differential equations and it was implemented by Matlab® software. From mathematical simulation of experimental conditions it were obtained the axial dispersion parameters and mass transfer coefficient which allowed a fair agreement in predicting break point time and the mass transfer zone profile.

Há a necessidade de prevenir formação de gelo nas asas e nos estabilizadores de aeronaves, pois as formas de gelo podem causar a degradação do desempenho aerodinâmico, o aumento de peso, bem como dificuldades de controle e manobra que, em casos críticos, leva a uma diminuição da margem de segurança operacional. Quando as aeronaves atravessam nuvens com gotículas de água sub-resfriadas, ou seja, em equilíbrio metaestável, o crescimento de gelo ocorre nas superfícies não protegidas. Usualmente, os sistemas antigelo térmicos de aerofólios são projetados, desenvolvidos e certificados com o auxílio de programas de simulação numérica. O presente trabalho visa desenvolver e implementar um modelo matemático para prever a transferência de calor e massa no escoamento bidimensional bifásico em torno de aerofólios de uso aeronáuticos, equipados com sistema de antigelo térmico operando em regime permanente. Em condições de formação de gelo, é necessário aquecer o bordo de ataque e controlar a temperatura da região protegida para que não ocorra formação de gelo. O sistema de aquecimento compensa os efeitos do resfriamento imposto principalmente pelos mecanismos acoplados de evaporação e transferência de calor por convecção, que são causados pelo escoamento do ar carregado de gotículas sub-resfriadas e pelo escoamento da água líquida residual. O modelo deverá estimar a distribuição de temperaturas de superfície e o coeficiente de transferência de calor com precisão ao uso em aplicações aeronáuticas. O presente trabalho implementou novos submodelos para: 1) estimar a molhabilidade da superfície do aerofólio por meio de um modelo matemático para caracterizar o escoamento da água líquida residual na padrão de filme e de filetes; 2) avaliar o comportamento dinâmico e térmico da camada-limite laminar e turbulenta por meio de análises integral e diferencial, que considera efeitos do gradiente de pressão, da transição laminar-turbulenta, da transpiração e da não uniformidade de temperatura da superfície e 3) estimar o início e o término da região de transição laminar-turbulenta. O presente trabalho seguiu um processo de desenvolvimento de código numérico que: verificou os resultados de cada submodelo separadamente para depois implementados no modelo do antigelo; validou os resultados da simulação de desempenho do sistema antigelo com os novos submodelos implementados. Os resultados obtidos foram considerados satisfatórios para o modelo do antigelo que utilizou os submodelos de ruptura de filme e formação de filetes pelo critério da Energia Mecânica Total Mínima, de camada-limite diferencial compressível e de previsão da transição laminar-turbulenta por correlações algébricas, que consideraram efeitos do gradiente de pressão e do nível de turbulência ao longe. / It is required to prevent ice accretion on wings and horizontal stabilizers because it may cause aerodynamic performance degradation, weight increase, flight control difficulties and, in critical cases, may lead to operational safety margins reduction. When aircraft flies through clouds containnig supercooled water droplets, which are in metastable equilibrium, ice will form in all non-protected surfaces. Usually, anti-ice protection systems are designed, developed and certified with a support from a numerical tool. The present describes the development and implementation of a mathematical model for prediction of heat and mass transfer in two-phase flow around airfoils, which are equipped with thermal anti-ice system and operating in steady state regime. Under icing conditions, it is necessary to heat and control the temperature of the airfoil surface at leading edge region to prevent ice formation. The heating system balances the evaporative cooling effects, which are caused by the coupled heat and mass convection transfer, imposed by the air flow loaded with supercooled water droplets and the runback water flow around the airfoil. The present work implemented submodels to: 1) estimate airfoil surface wetness factor by adopting a liquid water film flow model as well as a rivulet formation and flow model; 2) evaluate laminar and turbulent boundary layers with pressure gradient and laminar-turbulent transition over nonisothermal and permeable airfoil surface by implementing differential boundary layer analysis and 3) predict the onset position and length of laminar-turbulent transition region. The present paper followed a validation and verification process during the numerical code development. All sub-models results were verified separately against experimental data before their inclusion in anti-ice model.The results of anti-ice model with selected submodels were validated against reference cases. The results were considered suficiently accurate when solving the film breakdown and rivulets formation by total mechanical energy method, compressible boundary layer by differential analysis and laminar-turbulent transition prediction by algebraic correlations, which considered pressure gradient and freestream turbulence level.

Neste trabalho foi realizada a modelagem fenomenológica do desempenho de uma torre de resfriamento de água e de um sistema de duas torres de resfriamento em série, com temperatura de água de entrada de até 65 ºC. Verificou-se a validade do modelo através de comparação dos resultados previstos com os obtidos em ensaios em uma unidade piloto. Em seguida, através de simulação matemática, a partir do modelo desenvolvido, foi feito o estudo de diversos casos de aplicação. Estudou-se a influência das principais variáveis de operação (vazão de ar, vazão de água e temperatura de bulbo úmido) no desempenho de torres de resfriamento acopladas (em série e em paralelo) e no custo operacional. Verifica-se que a variável de maior influência é a vazão de água que circula pela torre. / This is a study about the phenomenological modeling of the performance of a water cooling tower and a system of two cooling towers in series, with water temperature input up to 65°C. The validity of model was verified by comparing the expected results with those obtained in tests on a pilot plant. After this, through a mathematical simulation, based on the model developed, several cases of application were analyzed. It was studied the influence of main operating variables (air flow, water flow and wet bulb temperature) on the performance of couples cooling towers (in series and parallel) and on operational cost. It was verified the most influential variable is the water flow rate through the tower.

No presente trabalho foi implementado um modelo matemático para simular o sistema antigelo eletrotérmico de um aerofólio. Por meio do programa ONERA2D simulou-se o escoamento potencial completo com velocidade 44,7 m/s (100 mph) e 89,4 m/s (200 mph) em torno de um aerofólio perfil NACA0012 de corda 0,914 m (3 pés) com ângulo de ataque de 0°, e calculou-se a eficiência de coleta local de gotículas de água com diâmetro mediano volumétrico de 20 μm. Foram simuladas quatro condições de teste com diferentes distribuições de fluxo de calor nos aquecedores elétricos do sistema antigelo. O modelo previu a distribuição de temperaturas na superfície sólida do aerofólio e no filme de água líquida, e as distribuições de fluxo de água líquida sobre a superfície do aerofólio (\"runback water\") e de coeficiente de transferência de calor por convecção de calor entre a superfície do aerofólio e o escoamento gasoso. Os resultados da simulação obtidos com o modelo foram comparados com resultados experimentais da NASA e os resultados numéricos dos programas LEWICE/ANTICE (EUA) e CANICE (Canada). Para as regiões molhadas pelo filme de água líquida, obteve-se um desvio máximo de temperatura de 2,6°C entre os resultados do presente modelo e o resultados experimentais. Para as regiões secas, onde não existe o filme de água líquida sobre a superfície do aerofólio, obteve-se um desvio de máximo de temperatura de 8°C. As previsões para distribuição de vazão de \"runback\", posição do término do filme de água líquida foram comparadas com os resultados do programa LEWICE/ANTICE. O modelo desenvolvido simula com adequada aproximação os efeitos da transferência de calor e de massa por convecção entre a superfície não-isotérmica do aerofólio ou do filme de água líquida e o escoamento gasoso, bem como os efeitos da transição entre o escoamento laminar e o turbulento na camada limite dinâmica e térmica e ainda a influência do escoamento do filme de água líquida sobre o desempenho do sistema de antigelo do aerofólio. / An electro-thermal anti-ice system was simulated with a mathematical model developed in the present work. A 44.7 m/s (100 mph) and 89.4 m/s (200 mph) full potential flow around a 0.914 m (3 ft) chord NACA0012 airfoil with 0° angle of attack and the local water catch efficiency of 20 μm median volumetric diameter droplets impingement were calculated by the numerical code ONERA2D. Four test conditions were simulated with four different heat flux distributions of the anti-ice system according to the experimental work developed at NASA. The model predicted distributions of solid surface and liquid water film temperatures, runback water flow and convection heat transfer coefficient between airfoil or water surface and gaseous flow. The simulated results obtained by the mathematical model developed were compared to NASA experimental results and the ones predicted by the numerical codes LEWICE/ANTICE (US) and CANICE (Canada). For the regions wetted by the water film, the present model provided 2.6°C maximum temperature deviations between the predicted results and experimental data. For the dry regions, where there is no liquid water on the airfoil surface, an 8°C maximum temperature deviation was obtained. The runback flow and water film ending point position were compared to LEWICE/ANTICE numerical results. The developed model predicts adequately the convection heat and mass transfer effects between the non-isothermal airfoil or liquid water film surface and the gaseous flow, as well the effects of laminar to turbulent flow transition within dynamic and thermal boundary layer and the influence of the liquid water film flow on the anti-ice system performance.

Despite the central importance of porous electrodes to any advanced electrochemical system, there is no clear answer to “<i>How to make the best electrode</i>?”. The source of ambiguity lies in the incomplete understanding of convoluted material interactions at smaller – difficult to observe length and timescales. Such mesoscopic interactions, however, abide by the fundamental physical principles such as mass conservation. The porous electrodes are investigated in such a physics-based setting to comprehend the interplay among structural arrangement and off-equilibrium processes. As a result, a synergistic approach exploiting the complementary characteristics of controlled experiments and theoretical analysis emerges to allow mechanistic insights into the associated mesoscopic phenomena. The potential of this philosophy is presented by investigating three distinct electrochemical systems with their unique peculiarities.

Le couple zéolite/H2O, qui présente une densité énergétique importante et remplit les conditions d’innocuité requises pour un système de stockage de chaleur pour l’habitat, est mis en œuvre dans réacteur modulaire à lit fixe. Un modèle monodimensionnel de transferts couplés de masse et de chaleur dans un lit fixe de grains de zéolite parcourus par un flux d'air humide été élaboré. Ce dernier a été conçu de façon à pouvoir intégrer rapidement des données sur de nouvelles générations de matériaux et coupler le réacteur à d’autres modèles : bâtiment/sous-station/quartier. L'étape de validation expérimentale montre que le modèle permet une estimation satisfaisante de l'autonomie, la durée d'amorçage et la puissance moyenne fournie en phase de décharge ainsi que la durée de charge. Ce modèle est donc un bon outil de dimensionnement et de pilotage du réacteur. L'analyse de sensibilité a montré que l'amélioration des prévisions du modèle requiert une évaluation plus précise de la chaleur complémentaire de sorption et de la porosité du lit. Après avoir estimé les besoins de chauffage d'une maison BBC deux dimensionnements ont été proposés afin d'effacer soit l'hyper-pointe de 18h - 20h soit la semaine la plus froide. Si la première stratégie aboutit à un système de stockage plus compact, la seconde permet de réduire le nombre de cycles marche/arrêt. Pour un îlot de 50 maisons BBC, la notion de foisonnement est considérée lors de l'estimation des besoins en chauffage pendant la semaine la plus froide. La phase de charge du système de stockage se ferait par le biais de la chaleur fatale récupérée dans l'industrie. Pour des stratégies d'effacement similaires (semaine la plus froide en hiver), un volume équivalent de 544 litres par maison dans l'îlot suffit pour répondre aux besoins de chauffage à Nancy contre 580 litres pour une maison BBC seule. / Heat storage systems for residential house heating could contribute to smoothing the load curve and would help prevent the use of the most polluting power plants or electricity imports during consumption peaks. Thermochemical heat storage systems are suitable for the intended application since they have high energy densities and low thermal losses. This thesis focuses on the design of an adsorption heat storage system that would be used to shed the load curve of the heating device of a house or residential district during the winter peak consumption periods. The zeolite/H2O pair, which has interesting features such as a high energy density and meets the conditions of safety required for a heat storage system for housing, is implemented in a modular fixed bed reactor. A 1D pseudo-homogeneous model was developed in order to simulate the performance of a fixed bed of zeolite during the adsorption and desorption of water. The latter was designed so as to facilitate the integration of data on new generations of materials and model couplings. The need to obtain data on the sorption properties of the zeolite/H2O pair to have reliable simulation results has been demonstrated, particularly at low partial pressures of water vapor and under the operating conditions selected. The experimental validation phase shows that the pseudo-homogeneous model provides a satisfactory estimate of criteria such as the autonomy, the responsiveness and the average power delivered during the discharging phase and the charging time. The model is thus a good sizing and management tool of the reactor. A sensitivity analysis, with the method of Morris, showed that improved model estimates require a more accurate assessment of the additional heat of sorption and porosity of the bed. After assessing the heating needs of the LEB house with a thermal model of the latter in cold climate conditions, two heat storage reactors were sized in order to shed the heating system's load curve either between 6 and 8pm or during the coldest week of the year. While the first strategy results in a more compact storage system, the second makes it possible to reduce the number of on/off cycles. The need for predictive control for monitoring the storage system was highlighted. As for the residential district of 50 LEB houses, diversity is considered when estimating the heating needs of the latter during the coldest week in Nancy. The heat source during the charging phase of the container would be industrial waste heat. During the coldest week, two sizings are suggested. For similar load shedding strategies, the comparison of the equivalent storage volume per house in the district with the storage volume for a single house serves highlights the importance of taking into account diversity. In order to meet the heating needs in Nancy, an equivalent volume of 544 liters per house in the district is sufficient whereas 580 liters are needed for a LEB house.

Ce travail est axé sur l'étude de l'évaporation lente d'un liquide confiné dans un tube capillaire de section carrée, en lien avec l'étude du phénomène de séchage. Dans un tel capillaire, si le liquide est suffisamment mouillant, des films liquides se forment par capillarité le long des coins internes. L'évaporation du liquide en sommet de film engendre un pompage capillaire et l'espèce volatile est alors transportée, sous phase liquide, au plus près du sommet du capillaire. Ce mode de transport dépend de la compétition entre les effets capillaires et les effets visqueux et de gravité qui s'opposent tous deux au mouvement du liquide vers le sommet du capillaire. Ces films liquides sont étudiés en adoptant une approche expérimentale. Le principe des expériences est de laisser un liquide volatil s'évaporer dans un tube capillaire carré. Plusieurs expériences d'évaporation sont réalisées en faisant varier la nature du liquide, la taille du capillaire et son orientation (horizontale et verticale). Une méthode de thermographie infra-rouge permet de mesurer le profil de température le long du capillaire. Le refroidissement induit par le changement de phase liquide-vapeur ainsi que sa position sont alors mesurables. A partir d'une méthode de visualisation par ombroscopie, plusieurs grandeurs sont mesurées : position du ménisque principal, taux d'évaporation et épaisseur relative des films. En s'appuyant sur une analyse simple du transfert de masse, on montre alors que les cinétiques d'évaporation obtenues expérimentalement se divisent en trois principales phases caractéristiques, ressemblant fortement aux trois périodes de la cinétique classique de séchage des milieux poreux capillaires. L'analyse de l'hydrodynamique des films montre qu'il est indispensable de prendre en compte l'arrondi interne des coins des capillaires dans la modélisation de l'écoulement au sein des films. On montre notamment que le phénomène étudié est très sensible à ce paramètre, qui limite l'extension des films. Ce travail expérimental a permis de développer une modélisation du transfert de masse dans la configuration étudiée, couplée à une modélisation de l'écoulement des films, et finalement de proposer un modèle de séchage d'un capillaire carré quantitativement satisfaisant. / In connection with the study of the phenomenon of drying, this work focuses on the study of slow evaporation of a liquid confined in a capillary of square cross section. In such a capillary, if the liquid wetting contact angle is low enough, liquid films are trapped by capillary forces along the capillary inside corners. Evaporation of the liquid at the film top creates a capillary pumping. The volatile species is then transported in liquid phase to the top of the capillary. This efficient mode of transport depends on the competition between the effects of capillarity and the effects of viscosity and gravity both opposing to the liquid flow towards the top of the capillary. In this work, the liquid films are studied experimentally.The principle of the experiments is to leave a volatile liquid evaporate in a square capillary tube. Several evaporation experiments are conducted, varying the liquid, the capillary tube size and its orientation (horizontal and vertical). An infrared thermography method allows to measure the temperature profile along the capillary. The cooling induced by the liquid-vapor phase change and its location is then measured. Owing to an ombroscopy visualization method, the location of the bulk meniscus, the evaporation rate and the relative thickness of the films can be measured. The experimental results show that the evaporation kinetics is similar to the drying kinetics of capillary porous media. This finding allows to study evaporation in a square capillary by analogy with the study of drying of capillary porous media. Based on a simple analysis of mass transfer in the system, it is then shown that the evaporation kinetics obtained experimentally can be divided into three main characteristic phases. The analysis of the hydrodynamic of the films shows that it is essential to take into account the roundeness of the capillary tube inside corners in the modelling of the flow in the films. We show that the phenomenon studied is very sensitive to the degree of roundedness of the tube internal corners, which limits the extension of the films. Modelling of the mass transfer coupled with modelling the film flow lead to a quantitatively satisfactory model of the drying of a square capillary tube.

No presente trabalho foi implementado um modelo matemático para simular o sistema antigelo eletrotérmico de um aerofólio. Por meio do programa ONERA2D simulou-se o escoamento potencial completo com velocidade 44,7 m/s (100 mph) e 89,4 m/s (200 mph) em torno de um aerofólio perfil NACA0012 de corda 0,914 m (3 pés) com ângulo de ataque de 0°, e calculou-se a eficiência de coleta local de gotículas de água com diâmetro mediano volumétrico de 20 &#956;m. Foram simuladas quatro condições de teste com diferentes distribuições de fluxo de calor nos aquecedores elétricos do sistema antigelo. O modelo previu a distribuição de temperaturas na superfície sólida do aerofólio e no filme de água líquida, e as distribuições de fluxo de água líquida sobre a superfície do aerofólio (\"runback water\") e de coeficiente de transferência de calor por convecção de calor entre a superfície do aerofólio e o escoamento gasoso. Os resultados da simulação obtidos com o modelo foram comparados com resultados experimentais da NASA e os resultados numéricos dos programas LEWICE/ANTICE (EUA) e CANICE (Canada). Para as regiões molhadas pelo filme de água líquida, obteve-se um desvio máximo de temperatura de 2,6°C entre os resultados do presente modelo e o resultados experimentais. Para as regiões secas, onde não existe o filme de água líquida sobre a superfície do aerofólio, obteve-se um desvio de máximo de temperatura de 8°C. As previsões para distribuição de vazão de \"runback\", posição do término do filme de água líquida foram comparadas com os resultados do programa LEWICE/ANTICE. O modelo desenvolvido simula com adequada aproximação os efeitos da transferência de calor e de massa por convecção entre a superfície não-isotérmica do aerofólio ou do filme de água líquida e o escoamento gasoso, bem como os efeitos da transição entre o escoamento laminar e o turbulento na camada limite dinâmica e térmica e ainda a influência do escoamento do filme de água líquida sobre o desempenho do sistema de antigelo do aerofólio. / An electro-thermal anti-ice system was simulated with a mathematical model developed in the present work. A 44.7 m/s (100 mph) and 89.4 m/s (200 mph) full potential flow around a 0.914 m (3 ft) chord NACA0012 airfoil with 0° angle of attack and the local water catch efficiency of 20 &#956;m median volumetric diameter droplets impingement were calculated by the numerical code ONERA2D. Four test conditions were simulated with four different heat flux distributions of the anti-ice system according to the experimental work developed at NASA. The model predicted distributions of solid surface and liquid water film temperatures, runback water flow and convection heat transfer coefficient between airfoil or water surface and gaseous flow. The simulated results obtained by the mathematical model developed were compared to NASA experimental results and the ones predicted by the numerical codes LEWICE/ANTICE (US) and CANICE (Canada). For the regions wetted by the water film, the present model provided 2.6°C maximum temperature deviations between the predicted results and experimental data. For the dry regions, where there is no liquid water on the airfoil surface, an 8°C maximum temperature deviation was obtained. The runback flow and water film ending point position were compared to LEWICE/ANTICE numerical results. The developed model predicts adequately the convection heat and mass transfer effects between the non-isothermal airfoil or liquid water film surface and the gaseous flow, as well the effects of laminar to turbulent flow transition within dynamic and thermal boundary layer and the influence of the liquid water film flow on the anti-ice system performance.

Neste trabalho foi realizada a modelagem fenomenológica do desempenho de uma torre de resfriamento de água e de um sistema de duas torres de resfriamento em série, com temperatura de água de entrada de até 65 ºC. Verificou-se a validade do modelo através de comparação dos resultados previstos com os obtidos em ensaios em uma unidade piloto. Em seguida, através de simulação matemática, a partir do modelo desenvolvido, foi feito o estudo de diversos casos de aplicação. Estudou-se a influência das principais variáveis de operação (vazão de ar, vazão de água e temperatura de bulbo úmido) no desempenho de torres de resfriamento acopladas (em série e em paralelo) e no custo operacional. Verifica-se que a variável de maior influência é a vazão de água que circula pela torre. / This is a study about the phenomenological modeling of the performance of a water cooling tower and a system of two cooling towers in series, with water temperature input up to 65°C. The validity of model was verified by comparing the expected results with those obtained in tests on a pilot plant. After this, through a mathematical simulation, based on the model developed, several cases of application were analyzed. It was studied the influence of main operating variables (air flow, water flow and wet bulb temperature) on the performance of couples cooling towers (in series and parallel) and on operational cost. It was verified the most influential variable is the water flow rate through the tower.