Orientador: Elaine Maria Cardoso / Resumo: A intensificação da transferência de calor por meio de alterações na morfologia da superfície aquecida vem sendo estudada no meio científico, a fim de suprir a crescente demanda de resfriamento de dispositivos com alta capacidade de processamento e dimensões cada vez menores. O presente trabalho apresenta o estudo experimental do efeito de superfícies nanoestruturadas e do espaçamento do canal de confinamento durante a ebulição em piscina da água deionizada, à temperatura de saturação na pressão atmosférica, sobre o coeficiente de transferência de calor, HTC, e fluxo crítico de calor, CHF. As superfícies nanoestruturadas foram obtidas pelo processo de ebulição do nanofluido de Al2O3-água deionizada em duas diferentes concentrações más-sicas: 0,03 g/l (“baixa” concentração, LC) e 0,3 g/l (“alta” concentração, HC). Foram realizados testes livres, com espaçamento, entre a superfície aquecida e a superfície adiabática, de 30 mm (correspondendo a Bo = 12), e testes sob condições de confinamento, com espaçamento de 1,0 mm (Bo = 0,4). As superfícies de teste foram caracterizadas por meio de medição da rugosidade média (Ra), do ângulo de contato estático (molhabilidade), e imagens MEV. Foi observado um aumento médio de 45% no HTC do teste com superfície lisa nanoestruturada em baixa concentração de nanofluido, em relação à superfície lisa sem deposição. Esse ganho está relacionado com o aumento do número de sítios ativos de nucleação causado pela deposição das nanopartículas sobre a ... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The intensification of heat transfer through changes in the heated surface morphology has been studied in the scientific community to meet the increase demand for cooling of devices with high processing power and smaller dimensions. This work presents the experimental study of the effect of nanocoated surfaces and gap size during nucleated boiling of deionized water, in saturation temperature at atmospheric pressure, about heat transfer coefficient, HTC, and critical heat flux, CHF. The pool boiling process of Al2O3-water based nanofluid at two different mass concentrations: 0.03 g/l (“low” concentration, LC) and 0.3 g/l (“high” concentration, HC), produced nanostructured surfaces. Unconfined tests were analyzed, with gap size between the heated surface and the adiabatic surface of 30 mm (corresponding to Bo = 12), and tests under confinement conditions, with gap size of 1.0 mm (Bo = 0.4). The tested surfaces were characterized by means of surface roughness (Ra) measurement, static contact angle (wettability), and SEM images. An average increase of 45% in HTC of the test with nanocoated smooth surface in low nanofluid concentration was observed in relation to smooth surface without deposition. This enhancement is related to the increase in the number of active nucleation sites caused by the nanoparticle’s deposition on the smooth surface. For all tests with rough nanocoated surfaces and nanocoated smooth one with high nanofluid concentration, there was degradation of the HTC ... (Complete abstract click electronic access below) / Mestre

Submitted by Jessica Martha Nunes (je.nunes25@gmail.com) on 2018-10-04T01:31:03Z No. of bitstreams: 1 Dissertação_NunesJM.pdf: 7307066 bytes, checksum: 4cd6a3e84b8e10900f14961caf7df4e5 (MD5) / Approved for entry into archive by Cristina Alexandra de Godoy null (cristina@adm.feis.unesp.br) on 2018-10-08T14:23:35Z (GMT) No. of bitstreams: 1 nunes_jm_me_ilha.pdf: 7307066 bytes, checksum: 4cd6a3e84b8e10900f14961caf7df4e5 (MD5) / Made available in DSpace on 2018-10-08T14:23:35Z (GMT). No. of bitstreams: 1 nunes_jm_me_ilha.pdf: 7307066 bytes, checksum: 4cd6a3e84b8e10900f14961caf7df4e5 (MD5) Previous issue date: 2018-08-10 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / A intensificação da transferência de calor por meio de alterações na morfologia da superfície aquecida vem sendo estudada no meio científico, a fim de suprir a crescente demanda de resfriamento de dispositivos com alta capacidade de processamento e dimensões cada vez menores. O presente trabalho apresenta o estudo experimental do efeito de superfícies nanoestruturadas e do espaçamento do canal de confinamento durante a ebulição em piscina da água deionizada, à temperatura de saturação na pressão atmosférica, sobre o coeficiente de transferência de calor, HTC, e fluxo crítico de calor, CHF. As superfícies nanoestruturadas foram obtidas pelo processo de ebulição do nanofluido de Al2O3-água deionizada em duas diferentes concentrações más-sicas: 0,03 g/l (“baixa” concentração, LC) e 0,3 g/l (“alta” concentração, HC). Foram realizados testes livres, com espaçamento, entre a superfície aquecida e a superfície adiabática, de 30 mm (correspondendo a Bo = 12), e testes sob condições de confinamento, com espaçamento de 1,0 mm (Bo = 0,4). As superfícies de teste foram caracterizadas por meio de medição da rugosidade média (Ra), do ângulo de contato estático (molhabilidade), e imagens MEV. Foi observado um aumento médio de 45% no HTC do teste com superfície lisa nanoestruturada em baixa concentração de nanofluido, em relação à superfície lisa sem deposição. Esse ganho está relacionado com o aumento do número de sítios ativos de nucleação causado pela deposição das nanopartículas sobre a superfície lisa. Para todos os testes com superfícies rugosas nanoestruturadas e lisa nanoestruturada com alta concentração, houve degradação do HTC, devido ao efeito de preenchimento das cavidades e formação de uma resistência térmica adicional. Para baixos fluxos de calor, houve um aumento no HTC para os casos confinados em comparação aos livres, como consequência da evaporação do filme líquido presente entre a superfície aquecida e a bolha de vapor. Porém com o aumento do fluxo de calor, o fenômeno do dryout é antecipado em relação aos testes livres, o que compromete o desempenho de componentes sob essas condições. Nos testes sob confinamento foram observados ganhos no fluxo de calor de início do dryout para todas as superfícies nanoestruturadas testadas, chegando a 52% para a superfície lisa nanoestruturada em alta concentração, em comparação à superfície lisa sem nanoestrutura. Isso mostra que a nanoestruturação, apesar de não promover ganho no HTC, auxilia no ganho do fluxo de calor de início do dryout, que é o limite operacional de sistemas que trabalham sob confinamento. / The intensification of heat transfer through changes in the heated surface morphology has been studied in the scientific community to meet the increase demand for cooling of devices with high processing power and smaller dimensions. This work presents the experimental study of the effect of nanocoated surfaces and gap size during nucleated boiling of deionized water, in saturation temperature at atmospheric pressure, about heat transfer coefficient, HTC, and critical heat flux, CHF. The pool boiling process of Al2O3-water based nanofluid at two different mass concentrations: 0.03 g/l (“low” concentration, LC) and 0.3 g/l (“high” concentration, HC), produced nanostructured surfaces. Unconfined tests were analyzed, with gap size between the heated surface and the adiabatic surface of 30 mm (corresponding to Bo = 12), and tests under confinement conditions, with gap size of 1.0 mm (Bo = 0.4). The tested surfaces were characterized by means of surface roughness (Ra) measurement, static contact angle (wettability), and SEM images. An average increase of 45% in HTC of the test with nanocoated smooth surface in low nanofluid concentration was observed in relation to smooth surface without deposition. This enhancement is related to the increase in the number of active nucleation sites caused by the nanoparticle’s deposition on the smooth surface. For all tests with rough nanocoated surfaces and nanocoated smooth one with high nanofluid concentration, there was degradation of the HTC due to the filling effect of the cavities and the formation of an additional thermal resistance. For low heat fluxes, the HTC increased for confined cases compare to unconfined ones, as consequence of the liquid film evaporation present between the heated surface and the vapor bubble. However, with heat flux increase, the dryout phenomenon incipience is precipitated in relation to unconfined tests, which compromises the performance of components under these conditions. In the confined tests, enhancement in dryout incipience heat flux were observed for all nanocoated surfaces tested, reaching 52% for the nanocoated smooth surface in high concentration, compared to the smooth surface without nanostructure. This shows that nanostructure, while not promoting HTC enhancement, helps to delay the dryout incipience heat flux, which is the operational limit of systems that work under confinement.

Suite à un défaut de contrôle de la réaction nucléaire, un accident d’insertion de réactivité (RIA) peut survenir dans une centrale. Un pic de puissance se produit alors dans certains crayons de combustible, suffisamment important pour entraîner l’ébullition en film du réfrigérant qui les entoure. Ceci provoque la chute du refroidissement des crayons et donc une rapide et importante augmentation de la température de la gaine qui les entoure. L’évaluation du risque de rupture de la gaine est un sujet d’étude de l’Institut de Radioprotection et de Sûreté Nucléaire. Ces échanges de chaleur transitoires ne sont toujours pas compris et modélisés. Pour comprendre ces phénomènes, une boucle expérimentale a été construite à l’Institut de Mécanique des Fluides de Toulouse. Du HFE7000 circule de bas en haut dans une section d’essais verticale de géométrie semi-annulaire. Le demi-cylindre intérieur est une feuille de métal chauffée par effet Joule. Sa température est mesurée par une caméra infrarouge, couplée avec une caméra rapide pour la visualisation de l’écoulement. La courbe d’ébullition entière est étudiée en régimes stationnaire et transitoire : convection, déclenchement de l’ébullition, ébullition nucléée, passage en film, ébullition en film et remouillage. Les régimes stationnaires sont bien modélisés par des corrélations de la littérature. Différents modèles sont proposés pour représenter les transferts de chaleur transitoires : l’évolution de la convection et de l’ébullition nucléée se font de manière auto similaire pendant un palier de puissance. Ce constat permet de modéliser des évolutions plus compliquées telles des rampes de température. Le modèle de Hsu instationnaire prédit bien le déclenchement de l’ébullition. Pour des créneaux de puissance, le passage en film se fait à une température constante et le flux critique augmente avec la puissance, tandis que pour des rampes de puissance la température augmente mais le flux critique diminue avec l’augmentation de la puissance. Quand la paroi est chauffée, les flux de chaleur en ébullition en film sont beaucoup plus importants qu’en stationnaire mais ce régime est encore mal compris. Le refroidissement en ébullition en film et le remouillage sont bien caractérisés par un modèle à deux fluides. / A failure in the control system of the power of a nuclear reactor can lead to a Reactivity Initiated Accident in a nuclear power plant. Then, a power peak occurs in some fuel rods, high enough to lead to the coolant film boiling. It leads to an important increase of the temperature of the rod. The possible risk of the clad’s failure is a matter of interest for the Institut de Radioprotection et de Sûreté Nucléaire. The transient boiling heat transfer is not yet understood and modelled. An experimental set-up has been built at the Institut de Mécanique des Fluides de Toulouse (IMFT). Subcooled HFE-7000 flows vertically upward in a semi annulus test section. The inner half cylinder simulates the clad and is made of a stainless steel foil, heated by Joule effect. Its temperature is measured by an infrared camera, coupled with a high speed camera for the visualization of the flow topology. The whole boiling curve is studied in steady state and transient regimes: convection, onset of boiling, nucleate boiling, criticial heat flux, film boiling and rewetting. The steady state heat transfers are well modelled by literature correlations. Models are suggested for the transient heat flux: the convection and nucleate boiling evolutions are self-similar during a power step. This observation allows to model more complex evolutions, as temperature ramps. The transient Hsu model well represents the onset of nucleate boiling. When the intensity of the power step increases, the film boiling begins at the same temperature but with an increasing heat flux. For power ramps, the critical heat flux decreases while the corresponding temperature increases with the heating rate. When the wall is heated, the film boiling heat transfer is higher than in steady state but it is not understood. A two-fluid model well simulates the cooling film boiling and the rewetting.

O estudo presente constitui uma análise da influência do acabamento superficial no coeficiente de transferência de calor na ebulição nucleada de refrigerantes halogenados. Dados para três superfícies distintas (cobre, latão e aço inoxidável), dois fluidos refrigerantes (R123 e R134a) e pressões reduzidas entre 0,023 e 0,26 são analisados com o intuito de verificar a influência da rugosidade nestes três parâmetros. O efeito da rugosidade foi avaliado com três acabamentos distintos (massa polidora, lixa e jato de areia) cobrindo uma faixa de rugosidades médias variando desde 0,03 até 10,5 micrômetro. Uma análise de diversas publicações da literatura foi levada a cabo, tendo sido particularmente investigadas algumas correlações que consideram o efeito do acabamento superficial em sua estimativa do coeficiente de transferência de calor. As tendências destas correlações são comparadas entre si e com os dados experimentais. A análise dos resultados permitiu levantar tendências inéditas na literatura consultada. A superfície em ebulição recebeu especial atenção com a obtenção de microfotografias e o cálculo de diversos parâmetros de rugosidade. Foram, ainda, investigados efeitos de envelhecimento da superfície, caracterizado pela diminuição do coeficiente de transferência de calor ao longo do tempo de ebulição. / The present research has been focused in an analysis of the effect of surface finishing on nucleate boiling heat transfer coefficient of halocarbon refrigerants. Experimental data for three different surface material (cooper, brass and stainless steel), two refrigerants (R123 and R134a) and reduced pressures between 0.023 and 0.26 have been analyzed aiming to verify the roughness effects on these three parameters. Three different finishing processes (polishing, emery papering and shot pining) have been used to result in an average roughness range from 0.03 to 10.5 micrometer. An analysis of varied publications and some correlations, particularly those which estimate the effect of surface roughness in heat transfer coefficient, has been done. The tendencies from these correlations have been compared with themselves and with experimental data. These results have shown some effects still unpublished. The boiling surface has received an especial attention, micro-photography has been taken and various parameters have been evaluated. Ageing effects, characterized by the reduction of heat transfer coefficient, have been verified and analyzed.

This study experimentally investigated the effect of hydrophobic and hydrophilic surfaces characteristics on nucleate boiling heat transfer performance for the application of direct immersion cooling of electronics. A dielectric liquid, HFE – 7100 was used as the working fluid in the saturated boiling tests. Twelve types of 1-cm2 copper heater samples, simulating high heat flux components, featured reference smooth copper surface, fully and patterned hydrophobic surface and fully and patterned hydrophilic surfaces. Hydrophobic samples were prepared by applying a thin Teflon coating following photolithography techniques, while the hydrophilic TiO2 thin films were made through a two step approach involving layer by layer self assembly and liquid phase deposition processes. Patterned surfaces had circular dots with sizes between 40 – 250 μm. Based on additional data, both hydrophobic and hydrophilic surfaces improved nucleate boiling performance that is evaluated in terms of boiling incipience, heat transfer coefficient and critical heat flux (CHF) level. The best results, considering the smooth copper surface as the reference, were achieved by the surfaces that have a mixture of hydrophobic/hydrophilic coatings, providing: (a) early transition to boiling regime and with eliminated temperature overshoot phenomena at boiling incipience, (b) up to 58.5% higher heat transfer coefficients, and (c) up to 47.4% higher CHF levels. The studied enhanced surfaces therefore demonstrated a practical surface modification method for heat transfer enhancement in immersion cooling applications.

Nucleating boiling is a repeating cycle of bubble initiation, growth and departure at many nucleation sites at the heated wall. Thereby, the bubble growth process significantly affects the dynamics of bubble departure. Experiments were performed to study the influence of heater surface characteristics, such as wettability and roughness, on single bubble growth and departure dynamics for natural circulation and upward flow boiling conditions. Self-assembled monolayer (SAM) coating, wet-etching and femtosecond pulsed laser treatment were used to alter the surface wettability and produce nano- and microstructures on stainless steel surfaces with a roughness in the range of micrometers. These surface preparation techniques allowed to separately quantify the effect of surface wettability and roughness on the bubble dynamics. The surface wettability and roughness are represented by the liquid contact angle hysteresis (θhys) and root mean square roughness of the surface (Sq). Boiling experiments were conducted at atmospheric pressure with degassed deionized water at low-subcooling. Stainless steel heater surfaces were vertically oriented during natural circulation boiling. In the experiments, bubbles were generated from an artificial nucleation cavity on the treated stainless steel heater surfaces. High-resolution optical shadowgraphy has been used to record the bubble generation, departure, sliding, detachment and inception of the next bubble. Higher bulk liquid velocity yielded smaller bubble departure diameters and slower bubble growth rates for all heater surface types. The effect of surface wettability on single bubble dynamics was studied for smooth surfaces with different liquid contact angle hysteresis. Low wetting surfaces yielded a greater bubble growth rate and departure diameter. The bubble growth rate and departure diameter were found maximum for an intermediate surface roughness Sq between 0.108 and 0.218 m. The corresponding roughness height is referred to as the ‘optimal roughness height’ in this work. Surface roughness was found very influential to the bubble growth and departure, which can be explained by considering its interaction with the microlayer underneath a bubble. The role of the heater surface parameters for the bubble growth was qualitatively assessed by evaluating the microlayer thickness constant C2. Hence, an improved bubble growth model was derived in this work. The bubble growth model was formulated on the basis of the evaporation of the microlayer beneath a bubble with the dryout area, inertia and heat diffusion controlled bubble growth and condensation at the bubble cap. The model can also predict the superheated liquid layer around a bubble which helps to determine the portion of a bubble that is in contact with the subcooled liquid. As bubble growth Abstract is highly dependent on the effective interactions of heater surface roughness and microlayer, a term Ceff was introduced in the bubble growth model. The effective microlayer thickness constant Ceff incorporates the impact of heater surface characteristics on the bubble growth process until the departure of a bubble. The bubble growth model was utilized in the analysis of high-resolution experimental data of steam bubble growth and the values of Ceff were calculated for different heater surface characteristics. The value of Ceff was found to decrease with the increase of bubble growth rate. A simplified model for the bubble departure criterion was derived from the expressions of forces which act on a nucleating bubble throughout its growth cycle. It was found that 90% of the departing bubbles satisfy the bubble departure criterion model with ±25% deviation. The knowledge gained from this work shall be particularly useful to improve nucleate boiling models for numerical simulations. The findings are also useful for designing heater surfaces in the future.:Abstract v Kurzfassung vii Acknowledgements xiii Abbreviations and Symbols xv Chapter 1: Introduction and Motivation 1 1.1 General overview 1 1.2 Theoretical background 3 1.3 Objectives and outline of the thesis 7 Chapter 2: Fundamentals of Bubble Dynamics in Nucleate Boiling 9 2.1 Bubble growth in nucleate boiling 9 2.2 Bubble growth models 12 2.3 The physical process of bubble departure 16 2.4 Experimental investigations of bubble dynamics 20 2.4.1 Effects of heater surface characteristics 21 2.4.2 Effects of bulk liquid velocity 24 2.5 Chapter conclusion 26 Chapter 3: Heater Surface Preparation and Characterization 27 3.1 Surface properties 27 3.2 Surface preparation 29 3.2.1 Self-assembled monolayer coating 30 3.2.2 High-power pulsed laser irradiation 31 3.2.3 Wet-etching 32 3.3 Surface cleaning 32 3.4 Surface characterization 32 3.4.1 Wettability measurement 32 3.4.2 Roughness measurement 33 3.4.3 Analysis of surface characteristics 34 3.4.4 Uncertainty of surface parameters 38 3.5 Artificial cavity preparation 38 Chapter 4: Experimental Setup and Procedure 41 4.1 Natural circulation boiling (NCB 41 4.1.1 Experimental procedure and measurement techniques 41 4.1.2 Uncertainty analysis 44 4.2 Upward flow boiling (UFB) 45 4.2.1 Experimental procedure and measurement techniques 45 4.2.2 Uncertainty analysis 48 4.3 Image processing 50 Chapter 5: Experimental Results 53 5.1 Introduction to the analysis of the bubble dynamics 53 5.1.1 The bubble life cycle 53 5.1.2 Calculation of the bubble equivalent diameter 55 5.1.3 Bubble dynamics with the increase of heat flux 57 5.1.4 Qualitative assessment of the bubble dynamics for different parameters 60 5.2 Bubble dynamics 61 5.2.1 Effect of heater surface wettability 61 5.2.2 Effect of heater surface roughness 65 5.2.3 Effect of bulk liquid velocity 70 5.3 Bubble departure 76 5.3.1 Effect of heater surface wettablity 76 5.3.2 Effect of heater surface roughness 76 5.3.3 Effect of bulk liquid velocity 78 5.4 Chapter conclusion 79 Chapter 6: Analysis and Model Development 81 6.1 Numerical evaluation of the role of heater surface characteristics 81 6.1.1 Derivation of an improved bubble growth model 86 6.1.2 Calculation of Ceff 82 6.2 Effect of liquid velocity on the bubble growth 93 6.3 Improved modeling of bubble departure 95 6.3.1 Analysis of important parameters 95 6.3.2 Formulation of a bubble departure criterion 100 6.4 Chapter conclusion 102 Chapter 7: Summary and Outlook 105 Bibliography 109 List of Figures 121 List of Tables 127 Appendix: Surface Parameters and Profile 129 / Der Blasenabriss von einer Keimstellenkavität ist ein komplexer Ablösemechanismus und spielt eine wichtige Rolle beim Wärmetransport. Zur Beschreibung der Blasendynamik sind Kenntnisse über den Blasenwachstumsprozess sowie die Vorhersage eines Kriteriums für die Blasenablösung erforderlich. In den existierenden Blasenwachstums- und Blasenablösungsmodellen wird die Oberflächencharakteristik des Heizers bisher nicht berücksichtigt. Im Rahmen dieser Promotion wurden Experimente durchgeführt, um den Einfluss der Heizeroberfläche und der Hauptströmungsgeschwindigkeit auf diese Parameter für eine vertikale Heizfläche zu untersuchen. Hierbei wurden das Naturkonvektionssieden und das aufwärtsgerichtete Strömungssieden betrachtet. Die Experimente wurden mit vollentsalztem Wasser bei einer Unterkühlung zwischen 1,68 und 4,00 K bei Atmosphärendruck und einem aus Edelstahl gefertigten Heizer durchgeführt, dessen Oberfläche anhand der Parameter Oberflächenrauigkeit und Benetzbarkeit charakterisiert ist. Unterschiedliche Oberflächenbearbeitungstechniken, wie Beschichtung durch Self-Assembled Monolayer (SAM), Nass-Ätzen und Hochleistungspuls-Laserbestrahlung wurden genutzt, um die Oberflächenbenetzung und –rauigkeit zu modifizieren. Der Unterschied zwischen dem gemessenen Fortschritts- (θadv) und Rückzugskontaktwinkel (θrec) der Flüssigkeit wird als Flüssigkeitskontaktwinkelhysterese (θhys) bezeichnet und beschreibt die Oberflächenbenetzbarkeit. Die Oberflächenrauigkeit wurde durch ein Konfokal-Mikroskop bestimmt und durch das gemittelte Quadrat der Rauigkeit (Sq) und den Maximalwert der Rauigkeit (St) definiert. Insgesamt wurden 18 unterschiedliche Heizoberflächen mit einer Größe von 130 x 20 mm² untersucht. Davon kamen jeweils die Hälfte für das Naturkonvektionssieden bzw. aufwärtsgerichtetes Strömungssieden zur Anwendung. Der Einfluss der Oberflächenbenetzbarkeit auf die Blasendynamik wurde für polierte Oberflächen (Sq  0,01 μm) analysiert. Die Wirkung der Oberflächenrauigkeit auf die Blasendynamik wurde für konstante Flüssigkeitskontaktwinkelhysteresen von 40,05°±1,5° und 59,97°±1,5° für Naturzirkulation und Strömungssieden untersucht. Eine künstliche zylindrische Kavität mit einer Fläche von 1963,5 m² und einer Tiefe von 50 m wurde mittels Mikrolaser in die Heizoberflächen eingebracht, um die Blasen in einer spezifischen Position zu erzeugen. Während des Naturkonvektionssiedens betrug die Wärmestromdichte 19,22 bis 30,29 kW/m². Bei den Experimenten mit aufwärtsgerichtetem Strömungssieden wurde die Hauptströmungsgeschwindigkeit im Bereich von 0,052 bis 0,183 m/s variiert und eine Appendix: Surface Parameters and Profile Wärmestromdichte zwischen 39,41 und 45,47 kW/m² aufgeprägt. Daraus resultierten insgesamt 87 Experimentalserien. Um den Blasenlebenszyklus zu erfassen, wurde hochauflösende Bildgebungstechnik verwendet. Mit der Bildverarbeitungssoftware ImageJ wurden die erfassten Videos weiterverarbeitet. Die Temperatur der Hauptströmung wurde mit Typ-K Thermoelementen gemessen. Die zeit- und ortsgemittelten Heizerwandtemperaturen wurden für die Naturzirkulation durch Infrarotthermografie und für das aufwärtsgerichtete Strömungssieden durch Typ-K Thermoelemente erfasst. Die mittlere Flüssigkeitsgeschwindigkeit wurde bei der Naturzirkulation mittels Particle Image Velocimetry (PIV) und beim Strömungssieden mittels Coriolis-Durchflusszähler bestimmt. Eine hochauflösende optische Schattenbildtechnik diente zur Aufzeichnung der Hauptphasen des Blasenlebenszyklus: Blasenerzeugung, Blasenwachstum, Blasenablösung, Blasengleiten und Blasenabriss. In dieser Arbeit wurden die der Blasenablösung vorrausgehenden Phasen untersucht. Blasenhöhe, Blasenbreite, Blasenbasisdurchmesser und Schwerpunkt der Blase wurden mit Hilfe der Bildverarbeitung ermittelt. Der blasenäquivalente Durchmesser wurde mittels des geometrischen Mittelwertes, der Blasenbreite und der Blasenhöhe berechnet. Basierend auf den Messdaten können folgende Erkenntnisse für das Blasenwachstum und den Blasenablösemechanismus postuliert werden: (i) Eine höhere Wärmeströmedichte führen zu größen Blasen und kürzeren Wachstumsperioden. Der Einfluss der Oberflächenbenetzbarkeit und der Oberflächenrauigkeit auf die Blasendynamik zeigt ähnliche Tendenzen für Naturkonvektion und aufwärtsgerichtetes Strömungssieden. (ii) Eine höhere Flüssigkeitskontaktwinkelhysterese führt zu einer schnelleren Expansion der Blasenbasis und zu einem schnellern Blasenwachstum. Für gut benetzbare Oberflächen bewegt sich der Blasenschwerpunkt schneller entlang der Strömungsrichtung. Für Oberflächen mit geringer Benetzbarkeit ist die Blasengröße vor der Blasenablösung größer und die Ablöseperiode länger. Der mittlere Blasenablösedurchmesser für unterschiedliche Hauptströmungsgeschwindigkeiten der Flüssigkeit erhöht sich von 0,75 auf 1,75 mm bei zunehmender Flüssigkeitskontaktwinkelhysterese von 42,32° auf 62,30°. (iii) Eine, bezogen auf die Mikrogrenzschichtdicke, optimale Oberflächenrauigkeit erhöht die Blasenwachstumsrate und die Blasengröße. Dieses Ergebnis ist bisher einzigartig bei der Untersuchung der Einzelblasendynamik beim Blasensieden. Die Expansion der Blasenbasis und der Blasenwachstumsrate erreicht ein Maximum für das gemittelte Quadrat der Rauigkeit (Sq) im Bereich zwischen 0,156 und 0,202 m für Naturzirkulation. Für aufwärtsgerichtetes Strömungssieden war die Expansion der Blasenbasis und die Blasenwachstumsrate für Sq-Werte zwischen 0,108 und 0,218 m maximal. Der Blasenablösedurchmesser wurde für einen großen Bereich der Hauptströmungsgeschwindigkeiten und Wärmestromedichte gemittelt. Das Maximum des mittleren Ablösedurchmessers wurde für die Oberfläche mit einem Wert von Sq = 0,218 m erreicht. Die Oberflächenrauigkeit erweitert die Wärmeübertragungsoberfläche neben der Blasenbasis. Der Einfluss der Oberflächenrauigkeitshöhe auf die Blasen hängt von der Mikrogrenzschichtdicke sowie vom Blasenbasisradius ab. Das Modell der Mikrogrenzschichtdicke von Cooper und Lloyd [1] und die konzeptionelle Idee zur Störung der Mikrogrenzschicht durch die Rautiefe von Sriraman [2] wurden analysiert. Es wurde nachgewiesen, dass die Oberflächenrauigkeit die effektive Mikrogrenzschichtdicke und die dazugehörige Wärmeübertragung beeinflusst. (iv) Es wurden geringere Blasenwachstumsraten für höhere Hauptströmungs-geschwindigkeiten gemessen. Weiterhin reduzieren sich der Blasenablösedurchmesser sowie Ablöseperioden mit zunehmender Hauptströmungsgeschwindigkeit bei unterschiedlichen Wärmeoberflächencharakteristiken. Bei niedrigen Hauptströmungs-geschwindigkeiten im Bereich zwischen ungefähr 0,052 und 0,16 m/s reduziert sich der durchschnittliche Blasenablösedurchmesser deutlich. Die experimentellen Ergebnisse zeigen einen wesentlichen Einfluss der Oberflächenbeschaffenheit auf das Blasenwachstum und den Ablöseprozess beim Blasensieden. Um diesen Einfluss numerisch zu charakterisieren, wurde ein neues Blasenwachstumsmodel entwickelt. Existierende Blasenwachstumsmodelle berücksichtigen den umfangreichen Einfluss der Oberfläche des Heizers bisher nicht. Das vorgeschlagene Model bezieht die plausibelsten Mechanismen des Blasensiedens mit ein. Dazu zählen: Mikrogrenzschichtverdampfung im Bereich der Austrocknung, trägheits- und wärmediffusionskontrolliertes Blasenwachstum und Kondensation an der Blasenoberseite. Das Modell berücksichtigt, dass die überhitzte Flüssigkeitsschicht an der Heizerwand durch die wachsende Blase nach außen verdrängt wird und die so gestreckte Flüssigkeitsschicht einen Teil der Blase einhüllt. Kondensation erfolgt an der Blasengrenze, die in Kontakt mit der unterkühlten Flüssigkeit steht, und demzufolge mit der überhitzen Flüssigkeitsschicht nicht in Kontakt kommt. Das vorgeschlagene Blasenwachstumsmodel arbeitet mit drei Konstanten für die beschriebenen Wärmeübertragungsmechanismen beim Blasenwachstum. Dabei handelt es sich um eine Konstante für die effektive Mikrogrenzschichtdicke (Ceff ), eine weitere Konstante 𝑏 ́ für die Wärmediffusion hin zur Blase und der Trägheit sowie letztendlich einer Konstante S zur Abbildung des Kondensationswärmeübergangs, anhand der Beschreibung des Anteils der Blase, welcher in Kontakt mit der unterkühlten Flüssigkeit steht. Die effektive Mikrogrenzschichtdickenkonstante (Ceff) definiert den Einfluss der Heizoberflächencharakteristik auf die Verdampfung der Mikrogrenzschicht und somit die Blasenwachstumsrate beim Blasensieden. Die numerisch berechnete und experimentell gemessene Blasengröße wurde verglichen, um die Mikrogrenzschichtdickenkonstante Ceff zu definieren. Der Einfluss der Kondensation auf Ceff wurde geprüft.:Abstract v Kurzfassung vii Acknowledgements xiii Abbreviations and Symbols xv Chapter 1: Introduction and Motivation 1 1.1 General overview 1 1.2 Theoretical background 3 1.3 Objectives and outline of the thesis 7 Chapter 2: Fundamentals of Bubble Dynamics in Nucleate Boiling 9 2.1 Bubble growth in nucleate boiling 9 2.2 Bubble growth models 12 2.3 The physical process of bubble departure 16 2.4 Experimental investigations of bubble dynamics 20 2.4.1 Effects of heater surface characteristics 21 2.4.2 Effects of bulk liquid velocity 24 2.5 Chapter conclusion 26 Chapter 3: Heater Surface Preparation and Characterization 27 3.1 Surface properties 27 3.2 Surface preparation 29 3.2.1 Self-assembled monolayer coating 30 3.2.2 High-power pulsed laser irradiation 31 3.2.3 Wet-etching 32 3.3 Surface cleaning 32 3.4 Surface characterization 32 3.4.1 Wettability measurement 32 3.4.2 Roughness measurement 33 3.4.3 Analysis of surface characteristics 34 3.4.4 Uncertainty of surface parameters 38 3.5 Artificial cavity preparation 38 Chapter 4: Experimental Setup and Procedure 41 4.1 Natural circulation boiling (NCB 41 4.1.1 Experimental procedure and measurement techniques 41 4.1.2 Uncertainty analysis 44 4.2 Upward flow boiling (UFB) 45 4.2.1 Experimental procedure and measurement techniques 45 4.2.2 Uncertainty analysis 48 4.3 Image processing 50 Chapter 5: Experimental Results 53 5.1 Introduction to the analysis of the bubble dynamics 53 5.1.1 The bubble life cycle 53 5.1.2 Calculation of the bubble equivalent diameter 55 5.1.3 Bubble dynamics with the increase of heat flux 57 5.1.4 Qualitative assessment of the bubble dynamics for different parameters 60 5.2 Bubble dynamics 61 5.2.1 Effect of heater surface wettability 61 5.2.2 Effect of heater surface roughness 65 5.2.3 Effect of bulk liquid velocity 70 5.3 Bubble departure 76 5.3.1 Effect of heater surface wettablity 76 5.3.2 Effect of heater surface roughness 76 5.3.3 Effect of bulk liquid velocity 78 5.4 Chapter conclusion 79 Chapter 6: Analysis and Model Development 81 6.1 Numerical evaluation of the role of heater surface characteristics 81 6.1.1 Derivation of an improved bubble growth model 86 6.1.2 Calculation of Ceff 82 6.2 Effect of liquid velocity on the bubble growth 93 6.3 Improved modeling of bubble departure 95 6.3.1 Analysis of important parameters 95 6.3.2 Formulation of a bubble departure criterion 100 6.4 Chapter conclusion 102 Chapter 7: Summary and Outlook 105 Bibliography 109 List of Figures 121 List of Tables 127 Appendix: Surface Parameters and Profile 129

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

Approved for public release; distribution in unlimited. / Spray cooling is a promising means of dissipating large steady state heat fluxes in high density power and electronic systems, such as thermophotovoltaic systems. The present study reports on the effectiveness of spray cooling in removing heat fluxes as high as 220 W/cm2. An experiment was designed to determine how the parameters of spray volumetric flow rate and droplet size influence the heat removal capacity of such a system. A series of commercially available nozzles were used to generate full cone water spray patterns encompassing a range of volumetric flow rates (3.79 to 42.32 L/h) and droplet Sauter mean diameters (17.4 to 35.5 micrometers). The non-flooded regime of spray cooling was studied, in which liquid spreading on the heater surface following droplet impact is the key phenomenon that determines the heat transfer rate. The experimental data established a direct proportionality of the heat flux with spray flow rate, and an inverse dependence on the droplet diameter. A correlation of the data was developed to predict heat flux as a function of the studied parameters over the range of values tested in this. / Lieutenant, United States Navy

Microchannel heat exchangers present many advantages, such as reduced size, high thermal efficiency and low fluid inventory; and are increasingly being used for heat transfer in a wide variety of applications including heat pumps, automotive air conditioners and for cooling of electronics.However, the fundamentals of fluid flow and heat transfer in microscalegeometries are not yet fully understood. The aim of this thesis is to contribute to a better understanding of the underlying physical phenomena in single-phase and specially flow boiling heat transfer of refrigerants in small channels. For this purpose, well-characterized heat transfer experiments have been performed in uniformly heated, single, circular, vertical channels ranging from 0.64 to 1.70 mm in diameter and using R-134a, R-22 and R-245fa as working fluids. Furthermore, flow visualization tests have been carried out to clarify the relation between the two-phase flow behavior and the boiling heat transfer characteristics. Single-phase flow experiments with subcooled liquid refrigerant have confirmed that conventional macroscale theory on single-phase flow and heat transfer is valid for circular channels as small as 640μm in diameter. Through high-speed flow boiling visualization of R-134a under non adiabatic conditions seven flow patterns have been observed: isolated bubbly flow, confined bubbly flow, slug flow, churn flow, slug-annular flow, annular flow, and mist flow. Two-phase flow pattern observations are presented in the form of flow pattern maps. Annular-type flow patterns are dominant for vapor qualities above 0.2. Onset of nucleate boiling and subcooled flow boiling heat transfer of R-134a has been investigated. The wall superheat needed to initiate boiling was found as large as 18 ºC. The experimental heat transfer coefficients have been compared to predictions from subcooled flow boiling correlationsav ailable in the literature showing poor agreement. Saturated flow boiling heat transfer experiments have been performed with the 640 μm diameter test section. The heat transfer coefficient has been found to increase with heat flux and system pressure and not to change with vapor quality or mass flux when the quality is less than ∼0.5. For vapor qualities above this value, the heat transfer coefficient decreases with vapor quality. This deterioration of the heat transfer coefficient is believed to be caused by the occurrence of intermittent dryout in this vapor quality range. The experimental database, consisting of 1027 data points, has been compared against predictions from correlations available in the literature. The best results are obtained with the correlations by Liu and Winterton (1991) and by Bertsch et al. (2009). However, better design tools to correctly predict the flow boiling heat transfer coefficient in small geometries need to be developed. Dryout incipience and critical heat flux (CHF) have been investigated in detail. CHF data is compared to existing macro and microscale correlations. The comparison shows best agreement with the classical Katto and Ohno (1984) correlation, developed for conventional large tubes. / QC 20101101

To diversify their genetic material, allowing adaptation to environmental disturbances and colonization of new ecological niches, bacteria use various evolutionary processes, including the acquisition of new genetic material by horizontal transfer mechanisms such as conjugation, transduction and transformation. Electrotransformation mediated by lightningrelated electrical phenomena may constitute an additional gene transfer mechanism occurring in nature. The presence in clouds of bacteria capable of forming ice nuclei that lead to precipitations and are involved in the triggering of lightning, such as the global phytopathogen Pseudomonas syringae, led us to postulate that natural electrotransformation in clouds may affect bacteria, by contributing to increase their adaptive potential. We first determined if the ice nucleator bacterium P. syringae could survive when in clouds and acquire exogenous genetic material through lightning shock-simulating in vitro electroporation. In comparison to two other bacteria, P. syringae appears to be best adapted for survival and for genetic electrotransformation under these conditions, which suggests that this bacterium would be able to survive and evolve whilst being transported in clouds. Secondly, we evaluated the impact of lightning shock-simulating in vitro electroporation on the survival, the electrotransformation potential and the diversity of bacteria collected from rain samples. These isolates better resisted lightning than the laboratory strains and some were able to electrotransform exogenous DNA. The rain bacteria we isolated were of different origins and were representative of life modes of the various sources of bacterial emissions on Earth. Our study suggests that bacteria aerosolized from diverse terrestrial ecosystems can spread to new habitats through clouds whilst also being able to acquire new genetic material via lightning-based electrotransformation, thereby potentially enhancing their genetic diversity. The final part of our work consisted of evaluating whether electrotransformation could be applied to the engineering of indigenous soil bacteria in order to develop a tool for the bioremediation of lindane, a once widely used pesticide. Optimized experiments revealed that both natural and electrotransformation contributed to the incorporation of a plasmid harboring a gene encoding the first lindane dechlorination steps by indigenous soil bacteria. In conclusion, we showed that natural electrotransformation mediated by electrical discharges such as those occurring in clouds or reaching soils can be involved in the horizontal gene transfer process among bacteria and, considering the importance of lightning worldwide, may play a role in the adaptation and evolution of these organisms.