We employ molecular dynamics simulations to investigate the evaporation process of nanosized droplets adsorbed on a substrate. Beads interacting with each other via Lennard-Jones (LJ) potentials are used to construct the simulation systems. The solid substrate contains 6 layers of beads forming a face-centered-cubic lattice. The bottom 3 layers are held rigid while the rest is kept at a constant temperature with a Langevin thermostat. A liquid droplet, consisting of LJ beads as well, is placed on top of the substrate. An appropriate amount of vapor beads are also supplied to the simulation box to help establish liquid-vapor equilibrium. To ensure adsorption, a stronger attraction is rendered between the droplet and a circular patch of 3 layers of beads at the center of the substrate surface while the rest of the substrate is made non-sticky for the fluid beads. During equilibration, the droplet and vapor are maintained at the same temperature as the thermalized substrate. After relaxation, the droplet adheres to the attractive patch as expected. Then a deletion zone is introduced into the top part of the vapor region. Fluid beads in this zone are removed at a given rate. To ensure that the evaporation dynamics and kinetics are properly captured, only the thermalized substrate is kept at the constant temperature during droplet evaporation. To carry out steady-state evaporation, the removed beads are reintroduced into a channel through the substrate and right below the droplet's center. These beads are then supplied to the droplet, compensating for the evaporation loss at the droplet surface. When the evaporation rate and the insertion rate are balanced, the system enters a steady state with the droplet undergoing continuous evaporation and its contact line pinned at the boundary of the attractive patch on the substrate. A one-to-one correspondence is found between the evaporation rate and the total number of fluid beads in the simulation box, as well as the contact angle of the droplet. Using this steady nonequilibrium system, we have mapped out the flow, temperature, and density fields inside and around the evaporating droplet as well as the local evaporation flux along the droplet surface with unprecedented resolutions. The results are used to test the existing theories on sessile droplet evaporation. / Master of Science / Droplet evaporation is a widespread natural phenomenon with numerous applications across various fields. While there has been extensive research on droplet evaporation, it remains a challenge to characterize the interior of the droplet and the local evaporation behavior on the droplet surface. Here we employ molecular dynamics (MD) simulation to model a nanosized droplet adsorbed on a substrate, which evaporates continuously while maintains a constant shape. This is realized by supplying the evaporated fluid back to the bottom of the droplet through an in-silico approach. Such a steady-evaporation system allows us to accurately map out the internal capillary flow of the evaporating droplet with a pinned contact line, where the droplet, vapor, and substrate meet. We find that local evaporation occurs faster near the contact line than at the apex of the droplet.

This thesis presents a review of the theory of evaporation of pure substances with respect to kinetic and mechanistic models. These concepts are applied to multi-component evaporation and a model for the evaporation of solute atoms from a solvent is postulated. The evaporation experiments were performed on Fe 1% Mn melts at a constant pressure of approximately 10 microns over a temperature range of 1320° to 1810°C. The correlation between the experimental results and the expected theoretical results is discussed thoroughly in terms of surface control and diffusion control. / Thesis / Master of Engineering (ME)

No description available.

The use of an ejector to improve the performance of a vapor compression refrigeration system was investigated. The ejector harnesses kinetic energy developed during the throttling process and returns it to the compressor in the form of a pressure boost. An ejector was designed taking into account the expected nonequilibrium behavior in the motive nozzle, and assuming equilibrium conditions in the other components. A standard vapor compression system using R-12 was modified to test the ejector. Included in the resulting ejector cycle was a hot gas bypass arrangement which was found to effectively control the condenser and evaporator flowrate. Hot gas bypass was used instead of a throttling valve, which would have undermined the benefit of the ejector. Experiments were conducted by running the ejector cycle at varying amounts of hot gas bypass flow. Performance results for the ejector and the overall cycle were obtained which showed that while the ejector recovered some of the kinetic energy of the expansion process, its design needed optimization. Recommendations were made for a more in-depth study of the ejector cycle. / Master of Science

This thesis is on molecular dynamics studies of drying suspensions of bidisperse nanoparticle mixtures. I first use an explicit solvent model to investigate how the structure of the dry film depends on the evaporation rate of the solvent and the initial volume fractions of the nanoparticles. My simulation results show that the particle mixtures stratify according to their sizes when the suspensions are quickly dried, consistent with the prediction of recent theories. I further show that stratification can be controlled using thermophoresis induced by a thermal gradient imposed on the drying suspension. To model larger systems on longer time scales, I explore implicit solvent models of drying particle suspensions in which the solvent is treated as a uniform viscous background and the liquid-vapor interface is replaced by a potential barrier that confines all the solutes in the solution. Drying is then modeled as a process in which the location of the confining potential is moved. In order to clarify the physical foundation of this moving interface method, I analyze the meniscus on the outside of a circular cylinder and apply the results to understand the capillary force experienced by a spherical particle at a liquid-vapor interface. My analyses show that the capillary force is approximately linear with the displacement of the particle from its equilibrium location at the interface. An analytical expression is derived for the corresponding spring constant that depends on the surface tension and lateral span of the interface and the particle radius. I further show that with a careful mapping, both explicit and implicit solvent models yield similar stratification behavior for drying suspensions of bidisperse particles. Finally, I apply the moving interface method based on an implicit solvent to study the drying of various soft matter solutions, including a solution film of a mixture of polymers and nanoparticles, a suspension droplet of bidisperse nanoparticles, a solution droplet of a polymer blend, and a solution droplet of diblock copolymers. / PHD / Drying is a ubiquitous phenomenon. In this thesis, I use molecular dynamics methods to simulate the drying of a suspension of a bidisperse mixture of nanoparticles that have two different radii. First, I use a model in which the solvent is included explicitly as point particles and the nanoparticles are modeled as spheres with finite radii. Their trajectories are generated by numerically solving the Newtonian equations of motion for all the particles in the system. My simulations show that the bidisperse nanoparticle mixtures stratify according to their sizes after drying. For example, a “small-on-top” stratified film can be produced in which the smaller nanoparticles are distributed on top of the larger particles in the drying film. I further use a similar model to demonstrate that stratification can be controlled by imposing a thermal gradient on the drying suspension. I then map an explicit solvent system to an implicit one in which the solvent is treated as a uniform viscous background and only the nanoparticles are kept. The physical foundation of this mapping is clarified. I compare simulations using the explicit and implicit solvent models and show that similar stratification behavior emerge in both models. Therefore, the implicit solvent model can be applied to study much larger systems on longer time scales. Finally, I apply the implicit solvent model to study the drying of various soft matter solutions, including a solution film of a mixture of polymers and nanoparticles, a droplet of a bidisperse nanoparticle suspension, a solution droplet of a polymer blend, and a droplet of a diblock copolymer solution.

The development and evaluation of a fuel cell test stand incorporating various methods for controlling the temperature and humidity of fuel cell reactants is described. The test stand is capable of accurately metering gas flows, controlling the temperature and humidity of the gases, and delivering the gases to the fuel cell in a safe manner. Additionally, the test stand can measure the voltage and current produced by the fuel cell during operation. Two test stands were constructed and evaluated, one using steam injection for fuel cell stacks and the other using flash evaporation for individual fuel cells. Both test stands were shown to provide adequate control at the upper end of the design range. The flash evaporation test apparatus was used to investigate the effect of inlet gas humidity on fuel cell performance. The results from this investigation showed that, for a fuel cell and reactant temperature of 75°C, the best performance was achieved with a high relative humidity (90%RH) for the hydrogen and a comparatively low relative humidity (60%) for the air. / Master of Science

Il est démontré que dans de nombreux procédés de transformation de la matière, les dégradations entropiques (et les consommations énergétiques qui en découlent) peuvent être minimisées en répartissant les flux d’énergies dans le volume plutôt qu’aux bornes du système (exemple de la distillation diabatique). Cependant la réalisation et la gestion de profils thermiques contrôlés (tels que des gradients thermiques et des étagements de température maîtrisés) dans les appareils compactes ne sont pas très souvent réalisées parce que souvent complexes et coûteuses (batterie d’échangeurs, gestion des fluides). Cette difficulté technologique affecte non seulement les performances énergétiques et les efficacités de transformation des appareils classiques de grandes tailles, mais aussi les réacteurs microstructurés dont les faibles dimensions internes associées à la grande conductivité thermique du matériau constituant les parois du contacteur entraînent souvent l’homogénéité thermique de l’appareil au détriment de la gestion des gradients thermiques. Par ailleurs, les garnissages utilisés dans les séparateurs sont de formes très complexes dans lesquels il est difficile de faire à la fois de la structuration thermique et hydrodynamique. Les systèmes microstructurés basés sur des géométries à plaques semblent offrir une possibilité intéressante de structuration thermique (contrôle et modulation de flux énergétiques pour l’obtention d’un profil thermique spécifique) et hydrodynamique (contrôle des transferts de matière, des temps de séjour, des pertes de charges etc.). Dans cette thèse, les réflexions menées sur ces verrous technologiques ont conduit à la conception d’un microcontacteur à film tombant. Ce microprocédé a été caractérisé d’un point de vue thermique et hydrodynamique. Une étude de faisabilité sur le potentiel de ce microcontacteur à séparer un mélange binaire d’alcools (ethanol/n-propanol) a été menée, ses performances ont été expérimentalement évaluées d’un point de vue qualitatif (pureté du distillat et du soutirat) et quantitatif (rapports des débits entrant et sortant), mais aussi en intégrant la notion d’exergie compositionnelle, également appelée puissance de séparation / In many processes of mass transformation, entropic degradations (and energy consumptions which results from) can be minimized by distributing heat flows in all process volume rather than boundaries (example: diabatic distillation). However it is difficult to control and impose thermal gradients on small scales because of their complexity and high costs (exchangers, fluids). This technological difficulty affects not only the performances and energy efficiencies of conventional devices (macro scales) but also small devices (e.g. microstructured reactors). Indeed, compact equipments with small dimensions generate some difficulties. First, the driving force inducing liquid flow by gravity is very small. Second, small size and high thermal conductivity of the material induce thermal homogeneity instead of managing temperature gradients in the system. In many separators, the used packing material has complex forms that make difficult thermal and hydrodynamic structurations. Microstructured devices, based on microchannel plate technologies, offer an interesting possibility of thermal structuration (control and modulation of energy fluxes) and hydrodynamics control (e.g. mass transfer, residence times, pressure drop etc). In this thesis, many discussions on these technological barriers led to the design of a new microstructured falling film contactor. The aim of the present work is to study heat transfer phenomena and liquid hydrodynamics in this device and, investigates the separation feasibility of a binary mixture of ethanol and n-propanol. Microcontactor performances were experimentally evaluated in terms of quality (bottom and top concentrations) and quantity (bottom and top flow rate ratios). To fully characterize contactor performances, the separation power concept (also called compositional exergy) was integrated

L’évaporation d’eau chargée en sels fait partie des processus de dégradation d’un milieu poreux. Lors de l’évaporation, les sels vont s’accumuler à l’interface liquide-gaz, pouvant aller jusqu’à la cristallisation. Lors de la cristallisation, des contraintes importantes sont exercées sur la matrice solide du milieu poreux, ce qui à terme peut la détériorer.Les travaux présentés ici portent sur l’évaporation d’une solution de NaCl en situation de mèche et se focalisent sur les phénomènes de transport jusqu’à la cristallisation. Nous avons analysé le lien entre le transport de la vapeur, l’écoulement induit dans la solution par l’évaporation, le transport de sel depuis le réservoir de solution saline vers l’interface, et la cristallisation. Plusieurs études expérimentales ont été réalisées pour différentes configurations de mèche. Par ailleurs, nous avons aussi mis en place divers modèles numériques (approche continue 1D et 2D, réseaux de pores 2D et 3D).Une première étude sur une mèche saturée met en évidence l’influence de la cristallisation sur les différents transports. Les cristaux forment un nouveau milieu poreux, favorisant l’évaporation et générant un effet de pompage sur la solution saline. Une deuxième étude sur des mèches saturées a permis d’analyser l’influence de l’évaporation et des propriétés du milieu poreux sur la localisation et le temps d’apparition de la cristallisation. Les expériences montrent une cristallisation discrète à la surface des mèches, se formant préférentiellement dans les zones où l’évaporation est la plus intense. Dans le cas des milieux hétérogènes, la localisation de la cristallisation dépend des propriétés des milieux poreux formant les mèches (porosité et perméabilité).Finalement, la situation d’évaporation en milieu partiellement saturé est étudiée et montre aussi une cristallisation discrète. Nous avons constaté qu’une approche continue classique ne permet pas de prédire correctement la cristallisation en raison des hétérogénéités des fronts. Pour palier ce manque, des modèles de réseaux de pores ont été développés. Les résultats obtenus indiquent que pour une évaporation insuffisante, la cristallisation n’a jamais lieu à l’interface. Lorsque l’évaporation augmente, la proportion de fronts amenant à la cristallisation augmente. Lorsque l’évaporation devient suffisamment intense, la totalité des fronts atteignent la cristallisation. Les zones de cristallisation préférentielles le long des fronts sont identifiées et caractérisées. / The evaporation of water with dissolved salt is a main source of degradation of porousmedia. As water evaporates, dissolved salts accumulate under the liquid-gas interface, possibly reaching crystallization. As crystals grow, stresses can be generated andmay deteriorate pore walls. In this context, our study focuses onNaCl transport and crystallization processwhich result from evaporation inside or at the surface of the porousmedium. The link between vapour transport, brine flow, salt transport and crystallization, is studied with both experiments and numerical simulations (continuummodels and pore network models). A firstwork on saturatedwicks shows howthe growth of efflorescences affects the different transports occurring during evaporation. Efflorescences create a new porous medium which increases evaporation, and consequently salt transport through the wick, generating a "pumping effect". The influences of evaporation rate distributions and porous medium properties on crystallization are also analysed. Results showthat crystallization occurs in a discrete way over the surface of the saturated wicks, due to the porous medium disorder. In addition, it is found that efflorescences tend to grow preferentially in strong evaporating areas. For heterogeneousmedia, results show that crystallization occurs over the less permeable and the less porousmedium. A study of evaporation inside partially saturatedwicks also indicates discrete crystallization at the front. Classic continuum models can not predict accurately the crystallization over this kind of heterogeneous interface. Pore network models are more suitable to simulate transports with these large scale heterogeneities. Results show that depending on the global evaporation rate at the front, crystallization never occurs, may occur with a certain probability or always occurs. The relation between fronts structures, evaporation rate distribution and transports in the liquid phase, is analysed in order to understand and predict crystallization localization. These crystallization spots are then identified and characterized.

Peut-on modéliser de manière fiable les effets collatéraux (en termes de quantité ou concentration de produits éjectés) et les conséquences d’une explosion en milieu clos, et en présence de produits chimiques liquides ? Pour répondre à cette vaste question, qui soulève spontanément de nombreux sous-problèmes, les travaux de thèse se sont déroulés en trois temps : 1/une étude qualitative et semi-quantitative du scénario général, afin de comprendre le déroulement chronologique des évènements, et d’émettre les hypothèses nécessaires à une modélisation, 2 /L’étude systématique des effets d’une explosion en milieu clos, en présence ou non de produits liquides. Avant de modéliser la dégradation du produit liquide soumis à une température et à une pression élevées, les expérimentations préliminaires ont en effet fait apparaître la nécessité de quantifier dans le temps et dans l’espace, les effets thermiques et mécaniques d’une explosion à volume constant, 3/L’élaboration de modèles 0D(t) à partir des conclusions précédentes afin, d’une part, d’estimer la durée de vie d’une phase liquide dans un environnement thermodynamique contraint et, d’autre part, de démontrer la possibilité de modéliser le problème global de manière réduite. En effet, le terme source d’un tel évènement ne peut être modélisé par une libération ponctuelle de produit : il s’agit d’une libération étendue dans le temps, par le biais d’un écoulement chaud a priori diphasique et de débit variable. Les couplages des phénomènes, observés expérimentalement, rendent nécessaires : 1 - Une modélisation instationnaire de l’évolution de la pression et une estimation du niveau de température atteint dans le volume d’étude, après détonation d’une charge explosive, 2 - Une modélisation de la libération de la phase liquide dans l’enceinte, sous forme de gouttes millimétriques ou de gouttelettes microniques 3 - Une modélisation instationnaire des transferts couplés de masse et d’énergie entre la phase liquide et la phase gazeuse en présence et prise en compte d’éventuelles réactions chimiques à haute pression et haute température 4 - Et enfin, une modélisation instationnaire des rejets à la brèche. L’étude d’une explosion à volume constant a montré qu’il est possible de modéliser de manière simple la montée continue en pression de l’enceinte par une fonction exponentielle croissante. Pour une configuration de référence donnée – explosion d’une sphère d’explosif dans un parallélépipède – la valeur maximale de pression est directement proportionnelle au taux de chargement en explosif, sur l’intervalle [0,01 – 0,6] kg/m3. Le passage à une géométrie différente ou plus complexe demande l’introduction d’un coefficient correctif pour traduire l’amplification (ou l’atténuation) de la combustion turbulente des produits de détonation avec l’air ambiant. En ce qui concerne le champ de température par contre, notre analyse a montré qu’il coexiste des zones chaudes et des zones dites « froides » et que la valeur de température homogène finale calculée à partir d’un code thermochimique ne peut constituer qu’une simple indication. Seule une estimation du volume respectif de ces zones a été proposée ici. Nous avons établi que les propriétés physico-chimiques des produits stockés sont un point clef du problème et on suppose ces données connues pour une gamme de produits chimiques liquides à pression ambiante, communément utilisés dans l’industrie. Seul le phénomène d’évaporation a été développé dans ce mémoire. L’introduction de réactions chimiques entre constituants se traduirait dans les modèles par des termes sources supplémentaires liés à l’apparition ou la disparition d’espèces. / Is it possible to model collateral effects due to an explosion (on a chemical facility for instance) occuring in a closed volume containing liquid chemical products storage units ?This thesis deals with a zerodimensionnal modelisation of such a 3D complex problem to asses the final thermodynamic state of chemical products released in the atmosphere. Developped sub-models take into account:- the unsteady time histories of the internal overpressure and temperature,- the unsteady liquid ejection (droplets sizes)- the unsteady modelisation of the local heat and mass transfers between the gas phase and the liquid phase- the unsteady ejection process of the resulting multiphase mixture in the environment.Models and sub-models are validated thanks to many experimental results.

Após a reunião de Kyoto ficou estabelecido o sistema de \"crédito carbono\" no qual as indústrias que reduzissem a emissão de gás carbônico para o ambiente aufeririam algumas vantagens. Neste sentido já há uma procura por parte de algumas empresas em reduzir esta emissão. Na maioria dos casos esta emissão é reduzida por meio da absorção do CO2 através de uma base, por exemplo, soda, amônia, cal, aminas, entre outras. O caso da amônia e da soda é interessante para empresas que possuem outros efluentes com estas bases em concentrações bem baixas, resultando assim soluções bem diluídas dos sais correspondentes. Soluções muito diluídas ou são descartadas no meio ambiente, que atualmente é proibido, ou são concentradas. A concentração, dessas soluções, usando vapor como meio de aquecimento torna-se um contra-senso por exigir a emissão de CO2 decorrente da queima de óleo combustível em caldeiras. Técnicas bem mais compatíveis com a preservação ambiental serão bem vindas daqui por diante, principalmente aquelas que não emitem CO2. Observando este fato este trabalho visa desenvolver um evaporador com promotor de película, em escala de laboratório para concentrar soluções diluídas, empregando energia solar como meio de aquecimento. O procedimento proposto não emite CO2, sendo mais compatível com a preservação ambiental. O equipamento construído consta de: coletor solar tipo placa plana com inclinação ajustável, promotor de película (aderido ao coletor),distribuidor de líquido, coletor de concentrado e acessórios. Foram estudadas as influências das variáveis: inclinação do coletor, vazão de alimentação e condições meteorológicas, na taxa de evaporação. As condições meteorológicas não podem ser controladas, mas foram constantemente monitoradas. Obtiveram-se maiores eficiências, quando a inclinação do coletor foi ajustada mensalmente, com valores até 36,4% maiores do que quando o coletor permanece fixo. / The system of Carbon Credits established by the 1997 Kyoto Protocol benefits companies that reduce their emissions of carbon into the environment with some advantages. Since the Protocol was signed, many companies have sought new ways to reduce their emissions. In most cases, these emissions are reduced through CO2 absorption by a base, e.g., soda, ammonia, lime and amines, among others. Ammonia and soda are interesting bases for companies that produce other effluents containing these products in much lower concentrations, resulting in highly diluted solutions of the corresponding salts. Highly diluted solutions are either discharged into the environment, which is forbidden today, or they are concentrated. Concentrating these solutions using vapor, as a means of heating is unfeasible since that would involve the emission of CO2 from burning oil in boilers. Therefore, from now on, attention will focus increasingly on more environmentally friendly techniques, especially techniques that do not cause CO2 emissions. A solar energy powered falling film evaporator with film promoter was developed for concentrating diluted solutions (industrial effluents). The procedure proposed here does not emit CO2, making it a viable alternative to the method of concentrating solutions. This novel device consists of the following components: a flat plate solar collector with adjustable inclination, a film promoter (adhering to the collector), a liquid distributor, a concentratecollector, and accessories. The evaporation rate of the device was found to be affected both by the inclination of the collector and by the feed flow. The meteorological variables cannot be controlled, but were monitored constantly to ascertain the behavior of the equipment in response to the variations occurring throughout the day. ) Higher efficiencies were attained when the inclination of the collector was adjusted monthly, showing up to 36.4% higher values than when the collector remained in a fixed position.