La vidange d'un réservoir à travers un orifice a été décrite par Torricelli il y a presque 400 ans. Son modèle ne prévoit aucun effet du mouillage du fluide qui s’écoule sur la plaque percée. Cette thèse montre expérimentalement que pour un orifice de taille comparable à la longueur capillaire, pour des fluides newtoniens peu visqueux, l’effet du mouillage sur le débit est important. Un modèle calculant la variation d'énergie cinétique dans le ménisque à la sortie du trou rend compte des observations expérimentales. Une instabilité, inconnue, du jet en sortie du trou apparaît également ; l'oscillation de la ligne triple du ménisque en est la cause. Les relations de dispersion de la fréquence d’excitation et des fréquences secondaires apparaissant le long du jet ont été établies.Cette étude a été complétée par des écoulements de fluides visqueux et viscoélastiques. Pour les fluides visqueux, le modèle d’écoulement du fluide parfait est corrigé à partir de nos expériences. Pour les fluides viscoélastiques, les expériences montrent qu'il y a compétition entre les dissipations visqueuses et les effets élastiques tout le long de l'écoulement. La prédiction de ces deux effets n'est pas aisée. Nous montrons des situations où les effets élastiques prédominent, permettant à une solution de polymère de couler plus vite que de l'eau.Pour finir, un second système expérimental a été construit permettant l'écrasement de fluides complexes entre deux plaques de verre parallèles. La visualisation, la mesure de la position et de la force normale, permettent de mieux comprendre le comportement de systèmes tels que les mousses, les émulsions, les gels sous une contrainte normale. / The draining of a tank through an orifice was described by Torricelli almost 400 years ago. His model does not provide for any wetting effect of the flowing fluid on the drilled plate. This thesis shows experimentally that the effect of wetting on the flow rate is important for Newtonian fluids with low viscosity in the case of an orifice the size of which is comparable to the capillary length. A model calculating the kinetic energy variation within the meniscus at the outlet of the hole allows us to account for experimental observations. Unknown jet instability also appears at the outlet of the hole; this is the oscillation of the meniscus triple line that is causing it. The relations of dispersion of the excitation frequency as well as that of the secondary frequencies appearing along the jet have been established.This investigation was supplemented by flows of both viscous and viscoelastic fluids. For viscous fluids, the perfect fluid model is corrected based on our experiments. For viscoelastic fluids, experiments show that there is competition between viscous dissipations and elastic effects throughout the flow. The prediction of both effects is challenging. We show situations where elastic effects dominate, allowing a polymer solution to flow faster than water.Finally, a second experimental set-up was build for compressing complex fluids between two parallel glass plates. Visualization, both position and normal force measurements, allow a better understanding of the behavior under normal stress of systems such as foams, emulsions, gels.

Le développement récent de la nanofluidique pose de nombreuses questions concernant les lois et longueurs caractéristiques qui régissent l’hydrodynamique et le mouillage à l’échelle du nanomètre. Pour aborder ce sujet, nous avons utilisé des techniques de microscopies avancées en sondant l’interface liquide/air à l’aide de pointes non conventionnelles. L’AFM utilisé dans le mode oscillant modulation de fréquence (FM-AFM) donne accès, de manière indépendante, à la force exercée par le liquide pendant une approche-retrait, et aux composantes conservatives et dissipatives de l’interaction pointe-ménisque. Des expériences complémentaires menées en microscopie électronique (MEB) permettent de visualiser le nanoménisque créé et de mesurer la force capillaire résultante. La couche visqueuse entraînée par l’oscillation de la pointe est d’abord étudiée. Le coefficient de friction et la masse ajoutée sont mesurés par FM-AFM en fonction de la viscosité des liquides et de la fréquence d’excitation. Un modèle basé sur une description classique rend compte quantitativement de l’ensemble des résultats expérimentaux permettant ainsi une évaluation du champ de vitesse entraîné par la nanosonde.Les méthodes développées ont permis d’étudier les caractéristiques de l’interface liquide. Le profil du nanoménisque est modélisé et validé grâce aux observations MEB. La raideur du ménisque mesurée expérimentalement par FM-AFM et décrite théoriquement démontre une dépendance logarithmique avec l’extension latérale du ménisque.Des résultats préliminaires sont également obtenus avec des pointes de carbone sur lesquelles glisse la ligne de contact, donnant accès à la dissipation dans le nanoménisque et à la ligne de contact, ainsi qu’à l’ancrage sur des défauts uniques, un des problèmes ouverts de la physique du mouillage.Cette étude démontre que le FM-AFM et le MEB sont des outils pertinents pour sonder quantitativement les propriétés des liquides à l’échelle nanométrique, ouvrant la voie à des études systématiques sur le mouillage à l’échelle nanométrique / The recent development of nanofluidic raises many issues about laws and characteristic lengths governing hydrodynamics and wetting at the nanometer scale. To address this issue, we used advanced microscopy techniques to probe the liquid/air interface with unconventional tips. The oscillating frequency modulation mode (FM-AFM) of the Atomic Force Microscope (AFM) gives independent access to the force applied by the liquid during an approach-withdrawal ramp, and to the conservative and dissipative components of the tip-meniscus interaction. Additional experiments conducted by electron microscopy (SEM) helped visualizing the shape of nanomeniscus to measure the resulting capillary force.The viscous layer set in motion by the oscillation of the tip is studied first. The friction coefficient and the added mass are measured by AFM-FM as a function of the viscosity of the liquid and of the excitation frequency. A model based on a classical description reflects quantitatively all experimental results enabling an evaluation of the velocity field caused by the nanoprobe.The developed methods also served to study the properties of the liquid interface. Nanomeniscus profile is modeled and validated through SEM observations. The stiffness measured experimentally by FM-AFM and described theoretically shows a logarithmic dependence with the lateral extension of the meniscus.Preliminary results are also obtained with carbon tips on which the contact line slides, giving access to the energy dissipation in the nanomeniscus and at the contact line, as well as to the anchoring of single defaults, an open issue of wetting physics.This study demonstrates that FM-AFM and SEM are relevant tools to probe quantitatively the properties of liquids at the nanoscale, opening the way for systematic studies on wetting at the nanoscale

Abstract Printed electronics offer unique possibilities for the development of devices and manufacturing methods. A prime example of printed electronics where the production volume can be significantly increased are solution-processed organic solar cells. Roll-to-roll (R2R) technology has made it possible to print solar cells almost as fast as newspaper. Unfortunately, the performance of printed devices depends strongly on film morphology, which is affected by the behaviour of the used ink on the confining surface - wetting. Key parameters that influence the wetting behaviour include surface energy, ink formulation, surface roughness, solvent properties, processing temperature and pre/post-treatments (heat, acid or plasma) and chemical heterogeneity. Importantly, a precise control of wetting and, consequently, film morphology is emphasized by many authors as an important factor for the commercialization of printed solar cells. This research focuses on measuring and analysing the influence of substrate processing temperature as well as plasma and UV pre-treatments on the wettability of various inks and substrates used in Organic Solar Cell (OSC) fabrication. It also explores the application of interesting novel materials, such as nanocellulose, in solar cell manufacture. The main tool applied here is the contact angle measurement method, since it is commonly used to obtain quantitative data describing the behaviour of ink droplets on substrate surfaces. Chief among the achieved results is the finding that the three factors mentioned above significantly influence ink-substrate interactions. Therefore, manipulation of plasma and UV treatments as well as substrate processing temperature, allow us to control wetting properties and, in consequence, the printing process. Another important result shows that the degree of control is strongly dependent on ink formulation and material composition and must, therefore, be taken into account in process development. These findings will contribute to a faster development of printed solar cells and their manufacturing conditions and requirements. / Tiivistelmä Painettava elektroniikka tarjoaa uusia mahdollisuuksia elektronisten laitteiden ja niiden valmistusmenetelmien kehittämiseen. Liuoskäsitellyt orgaaniset aurinkokennot ovat hyvä esimerkki painetun elektroniikan tuotteesta, jonka tuotantomäärää on voitu kasvattaa merkittävästi. Rullalta rullalle (engl. Roll-to-roll, R2R) -menetelmän avulla aurinkokennoja voidaan painaa lähes samalla nopeudella kuin sanomalehteä. Painettujen laitteiden suorituskyky riippuu suuresti tuotetun kalvon morfologiasta, johon vaikuttavat tuotantoprosessissa käytettyjen materiaalien kostumisominaisuudet. Tärkeimmät kostumiskäyttäytymiseen vaikuttavat parametrit ovat pintaenergia, pinnan karheus, musteen koostumus, liuotinominaisuudet, käsittelylämpötila, esi- ja jälkikäsittely (lämpö, happo tai plasma) sekä kemiallinen heterogeenisyys. Kostumisen, ja sitä kautta kalvon morfologian, tarkka säätely on tärkeää painettujen aurinkokennojen kaupallisen hyödyntämisen kannalta. Tässä väitöskirjatyössä mitataan ja analysoidaan käsittelylämpötilan sekä plasma- ja UV-esikäsittelyiden vaikutuksia orgaanisten aurinkokennojen valmistuksessa käytettyjen musteiden ja alustojen kostumisominaisuuksiin sekä tarkastellaan aurinkokennoliuoksissa käytettäviä uusia, mielenkiintoisia materiaaleja, kuten nanoselluloosaa. Työssä eniten hyödynnetty menetelmä on kontaktikulman mittaus, joka on yleisesti käytetty tapa hankkia kvantitatiivista tietoa mustepisaroiden käyttäytymisestä erilaisilla pinnoilla. Keskeisin saavutettu tutkimustulos on se, että kaikilla yllämainituilla kolmella käsittelyllä on huomattava merkitys musteen ja alustan vuorovaikutuksiin. Näin ollen plasma- ja UV-käsittelyillä sekä alustan käsittelylämpötilan säätelyllä voidaan hallita kostumisominaisuuksia ja sitä kautta koko painatusprosessia. Toinen tärkeä löydös on, että musteen koostumus ja alustan materiaali vaikuttavat siihen, kuinka voimakkaasti kostumista voidaan hallita. Näin ollen ne täytyy ottaa huomioon painatusprosessin suunnittelussa. Työssä saavutettuja tuloksia voidaan käyttää painettujen aurinkokennojen sekä niiden tuotantomenetelmien kehittämiseen.

Trickle bed reactors (TBRs) exhibit complex hydrodynamics and this study is aimed at giving insight into whether liquid-solid mass transfer and wetting are influenced by bed depth in a 5 m trickling column using 4 mm glass spheres as random packing. Measurements were made using the novel electrochemical technique developed by Joubert and Nicol (2013). Using this technique the wetting and mass transfer could be measured simultaneously. The study proves that the liquid-solid mass transfer and wetting efficiency do not stabilise at a minimum bed depth. The parameters were found to continue decreasing until the bottom of the bed. For the upper branch of the hydrodynamic envelope, the rate of decrease for the wetting efficiency was slow at the top of the bed and decreased rapidly closer to the bottom. However, only the wetting efficiency decreased significantly as a function of bed length; the liquid-solid mass transfer exhibited only a slight decrease of 14%. This compared well with the results of Du Toit et al. (2014), who found an 11% decrease in the liquid-solid mass transfer in a column with an x/D value of 29,4. The lower branch of the hydrodynamic envelope showed a linear decrease with respect to bed length for both wetting and mass transfer. The liquid-solid mass transfer decreased by 50% from the top of the bed to the bottom. These results are also in agreement with those of Du Toit et al. (2014)1 who found a decrease of 30% for a 1,6 m column. The wetting efficiency for the Levec mode decreased by 52%, whereas Du Toit et al. (2014)2 found a decrease of 20%. / Dissertation (MEng)--University of Pretoria, 2014. / lk2014 / Chemical Engineering / MEng / Unrestricted

Untersuchungen der Hochtemperaturbenetzung und Grenzfächeninteraktionen in Flüssigmetall/Übergangsmetall-Diborid-Systemen werden durch die technologische Nachfrage nach qualitative hochwertigen Metall-Matrix-Kompositen und verläßlichen Verbindungen von Ultrahochtemperaturkeramiken für aggressive chemische und/oder thermische Umgebungen angetrieben. Die physikalischen und chemischen Charakteristika der Metall/Keramik-Grenzflächen (z.B. die Benetzungskinetik, Grenzflächenreaktionen und die Phasenbildung) sind unerläßlich um die fundamentalen Mechanismen, die kontrollierenden technologischen Parameter, sowie die definierenden Eigenschaften und die Qualität des Endprodukts zu verstehen. Die Methode des liegenden Tropfens (engl. sessile drop method) ist die am häufigsten verwendete Verfahren für die quantitative Charakterisierung der Benetzungseigenschaften und für die direkte Untersuchung von Grenzflächeninteraktionen zwischen Flüssigmetall und festen Substraten unter Zuhilfenahme von Mikroskopie. Der Fokus dieser Arbeit liegt auf der Hochtemperaturbenetzung und den Grenzflächeninteraktionen von geschmolzenen reinem Al und Ti-Al, Ni-Al und Ni-B Schmelzen mit der TiB2 Ultrahochtemperaturkeramik. Die über die Methode des liegenden Tropfens hergestellten Metall/Keramik Verbunde werden hauptsächlich mittels Rasterelektronenmikroskopie gekoppelt mit energiedispersiver Röntgenspektroskopie sowie Röntgenbeugung untersucht. Die temperatur- und zeitabhängige Benetzung von flüssigem Al auf der TiB2 Keramik wurde mittels der klassischen Technik des liegenden Tropfens und der sogenannten dispensed drop technique (engl.) über einen weiten Temperaturbereich untersucht. Die Ergebnisse zeigen, dass mit ansteigender Temperatur die Benetzung bedeutend schneller abläuft. Die beiden Methoden liefern einen Unterschied in der Benetzungstemperatur von etwa 300°C aufgrund der nativen Oxidschicht auf der Al-Oberfläche bei dem klassischen Sessile-drop Versuch. Beginnend bei 1000°C füllt das flüssige Al entweder die intergranularen Poren auf oder dringt entlang der Korngrenzen in das TiB2 Substrat ein. Es wurde keine Reaktion im Al/TiB2 System beobachtet. Die Grenzflächeninteraktionen zwischen Ti-Al Schmelzen und der TiB2 Keramik wurden mit der klassischen Technik des liegenden Tropfens untersucht, da ein passender Tiegel für flüssiges Ti und Ti-haltige Schmelzen nicht vorhanden war. Für reines Ti auf TiB2 setzt das Schmelzen bei etwa 120 °C unter seinem Schmelzpunkt ein, wie aus Untersuchungen der Form und Struktur der erstarrten Ti/TiB2 Proben hervorgeht. Dies wird durch Festkörperdiffusion von B aus dem Substrat in die Ti Probe hinein und einer Verschiebung der Zusammensetzung von reinem Ti zu einer Ti-B Legierung in der substratnahen Region verursacht. Die Rolle von Al scheint für das Eindringen von Ti-Al Schmelzen entlang von Krongrenzen in das Keramiksubstrat von größerer Bedeutung zu sein als die Rolle der Temperatur. Die Benetzung und Grenzflächeninteraktionen zwischen Ni-Al Schmelzen und TiB2 wurden mit dem dispensed drop Verfahren untersucht. Flüssige Ni-Al Legierungen zeigen deutlich unterschiedliche Benetzungsverhalten auf dem TiB2 Substrat in Abhängigkeit von Legierungszusammensetzung und Testbedingungen. Das Verhalten von Ni-Al Schmelzen auf TiB2 Keramik verändert sich von einem auflösenden, reaktiven Benetzen (engl. dissolutive, reactive wetting) auf der Ni-reichen Seite zu einem nicht-reaktiven Benetzen auf der Al-reichen Seite. Die Ergebnisse deuten darauf hin, dass der Ni-Gehalt in Ni-Al Legierungen eine Hauptursache für die Veränderungen der Substratauflösung und der geometrischen Konfiguration an der Metall/Keramik Grenzfläche ist. Um den Einfluss des Ni-Gehalts auf die Auflösung von TiB2 zu verstehen wurde das Schmelzen und Benetzen von Ni83B17 und Ni50B50 Legierungen auf der TiB2 Keramik mittels der klassischen Technik des liegenden Tropfens im Hinblick auf mögliches fügen von TiB2 Keramiken untersucht. Basierend auf den Benetzungstest wurden Zwischenschichten der Ni50B50 Legierung verwendet um TiB2 Keramiken zu Verbinden. In dieser Arbeit wurde die Technik des liegenden Tropfens erfolgreich angewandt um die Hochtemperaturbenetzung und die Grenzflächeninteraktionen zwischen flüssigen Al, Ti-Al, Ni-Al und Ni-B Legierungen und keramischem TiB2 zu untersuchen. Die in dieser Arbeit erhaltenen Ergebnisse ermöglichen ein besseres Verständis der Interaktionsmechanismen zwischen flüssigen Al, Ti-Al, Ni-Al und Ni-B Legierungen und TiB2 keramik in diesen Systemen ermöglichen und Erstellung von Richtlinien für die Herstellung von Metall-Matrix-Verbunden und/oder Keramik-Matrix-Verbunden sowie für die Verbindung von keramischen TiB2 Teilen für strukturelle Hochtemperaturanwendungen dar. / Investigations of high-temperature wetting and interfacial interactions in liquid metal/transition-metal diboride systems are driven by technological demand in obtaining high-quality metal matrix composites and reliable joining of ultrahigh-temperature ceramics for aggressive chemical and/or thermal environments. The physical and chemical characteristics of metal/ceramic interface (e.g. wetting kinetics, interfacial reactions and phase formation) are indispensable for understanding the fundamental mechanisms, controlling technological parameters, and defining the properties and quality of final products. The sessile drop method is the most commonly used for quantitative characterization of the wetting properties and direct investigations of the interfacial interactions between a liquid metal and a solid substrate with the help of microscopy. This thesis is focused on the high-temperature wetting and interfacial interactions of molten pure Al and Ti-Al, Ni-Al and Ni-B alloys with TiB2 ultra-high-temperature ceramic. The metal/ceramic couples after the sessile drop tests are mainly characterized using scanning electron microscopy coupled with energy dispersive X-ray spectroscopy and X-ray diffraction. The temperature- and time-dependent wetting between the liquid Al and TiB2 ceramic over a wide temperature range was investigated using the classical sessile drop and dispensed drop techniques. The results showed that the wetting was significantly accelerated with increasing temperature. A difference of the wetting temperature by these two techniques was about 300 °C, due to the native oxide film present on the Al surface in the classical sessile drop tests. Starting from 1000 °C, liquid Al either filled the inter-grain pores or penetrated along the grain boundaries of the TiB2 substrate but there was no reaction observed in the Al/TiB2 system. The interfacial interactions between Ti-Al melts and TiB2 ceramic were studied by the classical sessile drop technique due to the absence of appropriate crucible for liquid Ti and Ti-containing melts. Pure Ti on TiB2 exhibited an incipient melting at about 120 °C below its melting point in view of the shape and structure of the solidified Ti/TiB2 couple. It was caused by the solid state diffusion of boron from the substrate into the Ti sample and a composition shift from pure Ti to a Ti-B alloy in the near-substrate region. In comparison to pure Ti, the role of Al in the penetration of Ti-Al melts penetration along the grain boundaries in the ceramic seemed to be more important than that of temperature in this study. The wetting and interfacial interactions between Ni-Al molten alloys and TiB2 were investigated using the dispensed drop technique. Liquid Ni-Al alloys showed a strong dependence of the wetting behavior on the TiB2 substrates, both on the alloy composition and testing conditions. It changed from a dissolutive, reactive wetting on the Ni-rich side to a non-reactive wetting on the Al-rich side. The results suggest that Ni content in Ni-Al alloys plays a major role in the changes of substrate dissolution and geometrical configuration at the metal/ceramic interface. To understand the effect of the Ni content on TiB2 dissolution, the melting and wetting of Ni83B17 and Ni50B50 alloys on TiB2 ceramic were investigated using the classical sessile drop technique in view of possible joining of TiB2 ceramics. Based on the wetting tests, TiB2 ceramics have been joined using Ni50B50 melt-spun ribbon as an interlayer. The results obtained in this work provide a better understanding of the interaction mechanisms in between liquid Al, Ti-Al, Ni-Al and Ni-B alloys and TiB2 ceramic and make basis for development of guidelines for the preparation of metal matrix composites and/or ceramic matrix composites and joining of TiB2 ceramic parts for high-temperature structural applications.

Trickle-flow is traditionally modeled by means of hydrodynamic parameters such as liquid holdup, two-phase pressure drop and wetting efficiency. Several studies showed that these parameters are not only a function of flow conditions and bed properties, but also of the flow history and morphology of flow. These can have a major influence on the distribution in the bed. The effect of flow morphology on liquid holdup and pressure drop is widely discussed in literature, but little attention is paid to its effect on wetting efficiency. Trickle-bed reactor models suggest that not a only bed-averaged but also the distribution of wetting efficiency may be of importance for reactor performance. Both the average wetting efficiency and the distribution of wetting are probably a function flow history and morphology. The distribution of wetting efficiency for different flow morphologies were investigated by means of a colorometric method that was developed for this purpose. Representative wetting distributions could be obtained. Flow morphologies and liquid distributions were manipulated by means of the pre-wetting procedure that was performed prior to flow. Pulse and Levec pre-wetted beds were investigated. These distributions were explained in detail in terms of flow morphology. It was found that the average wetting efficiency in pulse pre-wetted beds are much higher than in Levec pre-wetted beds. All particles in the pulse pre-wetted beds at all investigated flow conditions were contacted by the flowing liquid. This was not the case for the Levec pre-wetted beds. It was found that the flow in Levec pre-wetted beds become similar to that in pulse pre-wetted beds at high liquid flow rates. It was investigated how these distributions can affect reactor modeling, based on popular particle-scale models that relate reactor efficiency to wetting efficiency. According to these models, the wetting efficiency distribution in pulse pre-wetted beds can be characterised by means of only its average value. This is not the case for Levec pre-wetted beds. These results are however a strong function of the models that were employed. Finally, some recommendations are made in terms of the preferred pre-wetting method or flow morphology for different types of reactions. These recommendations are also based on models and have not been verified with experiments. / Dissertation (MEng (Chemical Engineering))--University of Pretoria, 2007. / Chemical Engineering / unrestricted

The biggest challenge with magnesium alloy biodegradable implants is the rapid corrosion at the earlier stages of the healing process after implantation. In this research, the impact of surface roughness generated by different means on the corrosion rate of AZ31 magnesium alloy in a simulated biologic environment is investigated. In order to perform accurate experimentation, an in vitro setup is assembled that simulates the human body environment accurately has been prepared using Schinhammer’s in vitro immersion testing setup and Kokubo’s Simulated Body Fluid (SBF). For the immersion test of Mg in SBF, several surface texture groups of Mg have been prepared and submerged into the in vitro tank. The Mg samples’ comparative analysis has been made in terms of corrosion rate, total weight loss and hydrogen gas evolution within a span of 7 days for the first experiment to narrow down the scope and 14 days for the follow up experiment. After 14 days of in vitro immersion test with varying roughness and hydrophobic modifications such as Cytop coating and stearic acid modification, it has been observed that the roughness group created by etching in aqueous NaCl solution for three minutes, shows better corrosion resistance compared to the polished control group. Hydrophobic modifications on the surfaces did not affect the corrosion behavior significantly.

Cette thèse est consacrée à l’étude du drainage dans des micromodèles de milieu poreux. Les techniques classiques de microfluidique (verre, PDMS) sont utilisées pour la réalisation de poreux modèles 2D de type grille. Des techniques de traitement de surface permettent de faire varier les conditions de mouillage. Le mouillage total, pseudo-partiel et partiel est étudié. Des méthodes d’analyse d’images sont développées afin de quantifier les vitesses locales lors de l’invasion du milieu poreux, vitesses qui sont ensuite comparées aux vitesses imposées. Ces données mettent en évidence avant percolation des comportements très différents selon la nature du mouillage. Un modèle permet de rendre compte des phénomènes observés. Nous expliquons pourquoi l’évolution après la percolation diffère selon la nature du mouillage. Enfin, l’influence de la rhéologie du fluide pousseur est abordée, celle d’un balayage initial à l’eau à très faible nombre capillaire également. / Drainage experiments in model porous media are performed. Our 2D micromodels consist of a regular network; they are made in glass or in PDMS. The wetting properties of the chip vary from total, pseudo-partial to partial wetting. Experiments are performed under flow rate control. Taking advantages of microfluidic devices, local velocities of the injected fluid are measured. The average of all these local velocities is compared to the velocity imposed by the syringe pump. Before percolation, the invasion percolation process for all wetting cases is studied. Depending on the wetting properties, several behaviours are observed. We develop a model to explain our experimental data. After percolation, the effects of wetting are huge; we explain why oil could remain trapped or not. Finally, we consider the influence of rheology by injecting non Newtonian liquids as pushing fluids, and also the effects of preferential paths.

Trickle flow is encountered in a variety of process engineering applications where gas and liquid flow through a packed bed of stationary solid. Owing to the complexities of three interacting phases, a fundamentally exhaustive description of trickle flow hydrodynamics has not been achieved. A complicating factor in describing the hydrodynamics is the fact that the hydrodynamic state is dependent not only on the present operating conditions but also on their entire history, including fluid flow rate changes and pre-wetting procedures. This phenomenon is termed hydrodynamic multiplicity and is the subject of this work. Hydrodynamic multiplicity greatly complicates both the experimental investigation into the behaviour of a trickle flow column and the theoretical modelling of the observed behaviour. Broadly speaking, this study addresses hydrodynamic multiplicity on three levels. First, a conceptual framework is proposed that can be used to study hydrodynamic multiplicity with limited resources. It is based on the absolute limiting values that the hydrodynamic parameters can adopt for a certain set of conditions, and encompasses both flow rate hysteresis loops and pre-wetting procedures. There are 5 such hydrodynamic modes. When the existing literature is critically evaluated in light of this framework, it is established that the reported experimental studies have not addressed all the issues. Previous modelling attempts are also shown to be unable to qualitative explain all the existing data. Moreover, authors have suggested different (and often contradictory) physical mechanisms responsible for hydrodynamic multiplicity. Secondly, an experimental investigation intended to supplement the existing literature and illustrate the utility of the proposed framework is launched. This includes bed-scale measurements of liquid holdup, pressure drop and gas-liquid mass transfer for a variety of conditions including different flow rates, pressures, particle shapes, particle porosity and surface tension. The second part of the experimental effort uses radiography and tomography in new ways to visualise the temporal and spatial characteristics of the different hydrodynamic modes. The tomographic investigation incorporates advanced image processing techniques in order to culminate in a pore-level evaluation of the hydrodynamic modes that reveals additional features of hydrodynamic multiplicity. Thirdly, the experimental insights are condensed into a set of characteristic trends that highlight the features of hydrodynamic multiplicity. A pore-level capillary mechanism is then introduced to qualitatively explain the observed behaviour. The mechanism shows how the differences in advancing and receding contact angles and the characteristics of the packed structure (or pore geometries) are ultimately responsible for the observed hydrodynamic multiplicity behaviour. Lastly, the effect of hydrodynamic multiplicity on trickle bed reactor performance is discussed. It is established experimentally that depending on the reaction conditions, different modes yield optimal performance. The idea of optimizing the performance by manipulating the hydrodynamic state is introduced. In totality, this work advances the understanding of trickle flow hydrodynamics in general and hydrodynamic multiplicity in particular. / Thesis (PhD (Chemical Engineering))--University of Pretoria, 2008. / Chemical Engineering / unrestricted

Nous étudions à travers plusieurs expériences la dynamique spéciale engendrée par des objets non mouillants. Un liquide en état Leidenfrost est autopropulsé lorsqu’on le pose sur un substrat texturé avec des rainures formant un motif à chevrons: les textures canalisent l'écoulement de vapeur dans une direction bien définie de sorte que ces aéroglisseurs liquides sont entraînés par la vapeur sous-jacente. Ces objets déformables subissent très peu de friction sur une surface plane. Toutefois, sur des substrats crénelés, les impacts sur les textures créent une friction spéciale qui est également étudiée. Nous étendons ce scénario d'entraînement visqueux dans d'autres situations où le liquide est remplacé par une plaque solide. Pour permettre la lévitation, on le place sur un substrat poreux à travers lequel de l'air est soufflé. Une fois de plus, l’écoulement est rectifié par des textures permettant l’entraînement d’une lamelle de verre dans un mouvement de translation ou même de rotation. Si nous augmentons la profondeur des textures, le confinement est perdu et on observe un mouvement dans la direction opposée dû à l'effet fusée. Nous nous intéressons également à une situation de non mouillage particulièrement simple: la goutte en chute libre. Nous abordons le problème de l’issue de cette chute: l'impact. Nous étudions d'abord l'impact d'une goutte sur un tamis. Dans cette situation, le liquide passe à travers les trous ou est arrêté par les sections bouchées. Nous nous concentrons ensuite à la force d’impact subie par le substrat. Nous la mesurons et la calculons en fonction des caractéristiques du liquide, de l’impact, et de la nature du substrat. / We investigate through several experiments the special dynamics generated by non-wetting objects.On a substrate textured with grooves forming a herringbone pattern, a Leidenfrost levitating liquid is propelled: the textures channel the vapor flow in a well-defined direction so that the slider above is driven by vapor viscosity. These deformable objects undergo very little friction on flat surfaces. However, on crenelated substrates, impacts on the texture sides greatly enhance dissipation. We extend this entrainment scenario to other situations where the liquid (and its deformable nature) is not involved anymore. A solid plate can levitate over a porous substrate through which air is blown. Again, escaping flow can be rectified by the textures and entrain the plate, leading to translation movement or even to rotation. If we create deeper channels (hence losing flow confinement), we observe motion in the opposite direction due to “rocket effect” (conservation of momentum). We are also interested in an extreme non-wetting situation: the falling drop. Indeed, all along the fall, the drop only experiences air drag friction, easily reaching high speeds. We tackle the problem of the dramatic issue of this fall: the impact. We first study the impact of a drop on a sieve. In this situation intermediate between a solid wall and no obstacle at all, mass either passes through the holes or gets stopped by the closings. We then focus on the impact force experienced by the substrates and characterize the force as a function of the drop and impact properties, but also of the nature of the solid on which impact takes place.