The problem under consideration is that of obtaining a representation of the permeability of a porous medium which is heterogeneous and anisotropic from limited information. To solve this inverse problem we propose the use of two different pieces that work together. A simulated annealing algorithm is presented and coupled with an homogenization technique; together these solve the problem which was posed. Further, numerical simulation results are presented illustrating the use of the simulated annealing algorithm as well as a coupling with the homoginization technique. This study illustrates that the performance of the annealing algorithm is enhanced with usage of homogenization.

In some configurations of compact, biofilm based photobioreactors, algae grow on a porous substrate that acts as the support system for the cells providing them with the necessary water and nutrients as well as carrying away their secreted products. The flow in these porous media can be driven by evaporation, mimicking the function of a synthetic leaf. The surface properties of the porous medium as well as the presence of a second immiscible phase in the fluid transported can significantly alter the transport capability and evaporative performance of the porous medium. The focus of this study is to investigate these effects through an experimental study. A dilute, 1% emulsion of lauric acid (chemical formula: C₁₂H₂₄O₂) in water was prepared using Tween® 80 surfactant. Evaporation driven flow of deionized water and the emulsion through two porous media, a hydrophilic glass fiber membrane and a less hydrophilic poly(vinylidene) fluoride (PVDF) membrane were studied. Experiments were conducted to determine the effect of porous medium and fluid properties on the rate of evaporation. The parameters investigated were the hydrophilicity of the porous medium and the area of the porous medium available for evaporation for both water and emulsion. During the experiment, the mass flow rate of the fluid as well as the temperature and the relative humidity of the ambient air were monitored. The results showed that for dilute emulsions, the rate of evaporation observed was the same as that for water and was dictated by the governing laws of convection applicable to the situation based on the geometry of the setup and the ambient conditions. The response of the porous medium to flow of dilute emulsion showed that the highly hydrophilic glass fiber porous medium rejected any accumulation of the oil phase in the pores, and ejected it out, whereas the lesser hydrophilic PVDF porous medium allowed the pores to be clogged by the oil phase, resulting in change in the properties of the medium. However, the dependence of this observation solely on surface properties of the medium cannot be ascertained as the glass fiber medium had a larger pore diameter than the PVDF medium, and this factor could be of effect. The relative humidity of ambient air affected the rate of evaporation, which implied that the flow was limited by evaporation rather than by the viscous losses in the porous medium. The response of change in rate of evaporation to change in relative humidity showed a high time lag. Also, it was seen that there was a maximum area over which evaporation occurred which was dictated by the capillary pressure generated by the porous medium and the viscous losses for the fluid flow through the medium. Any excess area available for evaporation did not have any effect on the rate of evaporation. Electrospinning, as a simple and effective process for generating fibrous porous media was presented and a sample porous medium was prepared using this method. A parametric analysis of the effect of the potential difference applied between the syringe tip and the collector electrode, and the distance between the tip and the collector on the diameter of fibers produced, was performed. / text

Made available in DSpace on 2014-06-11T19:30:10Z (GMT). No. of bitstreams: 0 Previous issue date: 2000-01Bitstream added on 2014-06-13T19:06:14Z : No. of bitstreams: 1 franca_ac_me_guar.pdf: 603778 bytes, checksum: d86c030bdbc65869a9f5aa8f9efc6b33 (MD5) / É comum serem identificados trincas em alvenaria de tijolos, não somente depois de um certo tempo da construção, mas também por vezes aparecem estas trincas desagradáveis durante a própria construção. Depois de alguns estudos é possível verificar que, dentre outros fatores, um deles é devido à umidade absorvida pelos elementos da parede (tijolos e argamssa). Es umidade pode vir do ar (dia únmido), de chuvas, do solo e até do excesso da quantidade de água contida na argamassa de assentamento. Ela em contato com os elementos da parede, porosos, concorrerá para a percolação capilar dando origem a esforços internos consideráveis, que virão provocar deslocamentos, que via de regra são diferentes para o tijolo e para a argamssa(deltaLtijolo#deltaLargamassa), daí então o aparecimento de trincas. Portanto, com vistas a esses fenômenos, resolveu-se estudar este assunto de forma inicial, sem grande e sofisticado equipamento de medição do escoamento capilar, mas conduzindo a análise para obtenção de valores médios de pressão e velocidade de fluído, medindo as deformações gerais dos elementos tijolos. / Abastract: It is common to find cracks in bricks masonry not just after some time of the construction but, the appearance of these unpleasant cracks, are also common during the contruction work. After some research it is possible to verify that, among other factors, one of its cause is due to the humidity absorbed by the elements of the wall (bricks and mortar. Now, this humidity may come from the air (e. g. a humid day), from rain, from the ground and from an exaggerated amount of the water mixed in the mortar laying. These when in contact with theporous elements of the wall will induce a capillary percolation that creates considerable internal stress that will give origin to displacements. These displacement are usually different for the brick and for de mortar (deltaLBrick#deltaL mortar) and because of that cracks will appear. With that in mind, it was decided to study this subject initially without bulky and sophisticated equipment for the measurement of capillary flow. Our approach was to make the analysis by obtaining medium values for the pressure and fluid speed, measuring the general deformations of the bricks.

No description available.

No description available.

An experimental research program was designed to study laminar flows through and over models of porous media with or without inertial effects. The models used were made up of circular or square rods arranged to cover solid volume fraction ϕ ranging from 0.03 to 0.49, and filling fraction h / H ranging from 0.34 to 1 of the test channel. In this way, the ratios of the depth of the test section to the porous medium pore H / l ranged from 5.75 to 18.25. Three types of model porous media were tested: (1) two-dimensional ‘horizontal’ models, having rod axes aligned along the span of the channel in a staggered or non-staggered fashion; (2) three-dimensional ‘vertical’ models with rod axes aligned in the transverse direction; and (3) three-dimensional ‘mesh’ models with rod axes aligned along both transverse and spanwise directions. Using a pressure-driven viscous fluid, the bulk Reynolds number Rebulk was varied from 0.1 to 10.3. Velocity measurements were obtained using particle image velocimetry at various streamwise-transverse planes of the test section. Differential pressure measurements were also obtained using electronic transducers. These measurements were used to determine relevant governing equations for the flow through the porous media; to characterize the effects of ϕ rod shape and arrangement, h / H, H / l, porous media dimensionality, and Rebulk on the flow; and to predict the flow at the porous medium-free flow interface. The Izbash and quadratic Forchheimer equations were respectively found to describe well the flow through two- and three-dimensional porous media. Penetration of the free flow into the porous medium varied with ϕ and rod arrangement, but was nearly independent of the rod shape. At the interface between the porous medium and the free flow, h / H and H / l effects were found to be counteractive. Penetration was highest for the vertical models compared with the mesh and horizontal models. Inertial dependence of interfacial flow was weak when porous medium conditions were considered. The interfacial flow was found to follow a dose response formulation with a predictable slip coefficient.

An experimental research program was designed to study laminar flows through and over models of porous media with or without inertial effects. The models used were made up of circular or square rods arranged to cover solid volume fraction ϕ ranging from 0.03 to 0.49, and filling fraction h / H ranging from 0.34 to 1 of the test channel. In this way, the ratios of the depth of the test section to the porous medium pore H / l ranged from 5.75 to 18.25. Three types of model porous media were tested: (1) two-dimensional ‘horizontal’ models, having rod axes aligned along the span of the channel in a staggered or non-staggered fashion; (2) three-dimensional ‘vertical’ models with rod axes aligned in the transverse direction; and (3) three-dimensional ‘mesh’ models with rod axes aligned along both transverse and spanwise directions. Using a pressure-driven viscous fluid, the bulk Reynolds number Rebulk was varied from 0.1 to 10.3. Velocity measurements were obtained using particle image velocimetry at various streamwise-transverse planes of the test section. Differential pressure measurements were also obtained using electronic transducers. These measurements were used to determine relevant governing equations for the flow through the porous media; to characterize the effects of ϕ rod shape and arrangement, h / H, H / l, porous media dimensionality, and Rebulk on the flow; and to predict the flow at the porous medium-free flow interface. The Izbash and quadratic Forchheimer equations were respectively found to describe well the flow through two- and three-dimensional porous media. Penetration of the free flow into the porous medium varied with ϕ and rod arrangement, but was nearly independent of the rod shape. At the interface between the porous medium and the free flow, h / H and H / l effects were found to be counteractive. Penetration was highest for the vertical models compared with the mesh and horizontal models. Inertial dependence of interfacial flow was weak when porous medium conditions were considered. The interfacial flow was found to follow a dose response formulation with a predictable slip coefficient.

Natural convection with high cooling effects is of increasing interest in cold region geotechnical engineering. To study natural air convection in a highly-permeable porous medium, convective and conductive heat transfer experiments were carried out using an insulated cylindrical tank filled with styrofoam chips. Convection and conduction were caused by controlling the temperatures at the top and bottom of the tank, and a series of cross-sectional conductive and convective isotherms were generated from collected temperature data. Additional convective patterns were obtained from tests by centrally localized heating below or cooling above. Flow velocities were measured at the center of the tank. Results showed that convective heat transfer rate was higher than thermal conduction. Convective isothermal patterns varied with various boundary conditions and could be influenced by small temperature perturbation. Given appropriate environmental conditions, efficient convective cooling effects can be used to enhance ground freezing or to protect permafrost from degradation. / Geotechnical Engineering

Ce travail propose une étude expérimentale de l'ébullition en masse dans une structure poreuse modèle. L'objectif est d'approfondir la compréhension des transferts de chaleur dans un écoulement diphasique avec changement de phase liquide-vapeur en milieu poreux, en liaison avec la problématique de la gestion des accidents graves dans les réacteurs nucléaires. A la suite d'un dysfonctionnement sur le circuit de refroidissement d'un réacteur nucléaire, l'augmentation de la température au sein du cœur provoque l'effondrement des tubes contenant le combustible. Il en résulte la formation d'un lit de débris chaud, assimilable à un milieu poreux dégageant une puissance thermique importante, qui peut être refroidi efficacement par renoyage avec de l'eau. Cela engendre des mécanismes d'ébullition intenses qu'il convient de modéliser proprement pour estimer les chances de succès du renoyage. Notre étude vise à caractériser les échanges de chaleur à l'échelle du pore en fonction des caractéristiques de l'écoulement local. Une partie importante du travail a été consacrée à la mise au point du dispositif expérimental. Le cœur du dispositif est un milieu poreux bidimensionnel formé de cylindres disposés aléatoirement entre deux plaques de céramique. Chaque cylindre est une sonde à résistance de platine, utilisée non seulement pour fournir la puissance thermique désirée mais aussi pour mesurer la température de l'élément : chaque élément chauffant est contrôlé individuellement ou en groupe à l'aide d'un système d'asservissement temps réel. La plaque supérieure étant transparente, la distribution des phases au sein du poreux est obtenue par visualisation haute vitesse. L'acquisition d'images et les mesures thermiques permettent de caractériser l'échange de chaleur effectif local en fonction du régime d'ébullition. Deux configurations principales ont été étudiées. Dans la première, le milieu est initialement saturé en liquide et chauffé jusqu'à l'apparition de la vapeur et l'obtention de différents régimes d'ébullition. Ceci a notamment permis d'établir des courbes de Nukiyama en milieu confiné. Dans la seconde, dite de renoyage, le liquide est injecté dans le milieu sec et surchauffé initialement. Ceci a permis de caractériser la dynamique du renoyage et de visualiser les régimes d'écoulement rencontrés. Les résultats sont discutés en relation avec le modèle macroscopique à non-équilibre thermique local actuellement le plus avancé pour l'étude de ces différentes situations d'ébullition. / This manuscript presents a pore-scale experimental study of convective boiling heat transfer in a two-dimensional porous medium. The purpose is to deepen the understanding of thermohydraulics of porous media saturated with multiple fluid phases, in order to enhance management of severe accidents in nuclear reactors. Indeed, following a long-lasting failure in the cooling system of a pressurized water reactor (PWR) or a boiling water reactor (BWR) and despite the lowering of the control rods that stops the fission reaction, residual power due to radioactive decay keeps heating up the core. This induces water evaporation, which leads to the drying and degradation of the fuel rods. The resulting hot debris bed, comparable to a porous heat-generating medium, can be cooled down by reflooding, provided a water source is available. This process involves intense boiling mechanisms that must be modelled properly. The experimental study of boiling in porous media presented in this thesis focuses on the influence of different pore-scale boiling regimes on local heat transfer. The experimental setup is a model porous medium made of a bundle of heating cylinders randomly placed between two ceramic plates, one of which is transparent. Each cylinder is a resistance temperature detector (RTD) used to give temperature measurements as well as heat generation. Thermal measurements and high-speed image acquisition allow the effective heat exchanges to be characterized according to the observed local boiling regimes. This provides precious indications precious indications for the type of correlations used in the non-equilibrium macroscopic model used to model reflooding process.

Orientador: Luiz Roberto Carrocci / Banca: José Luz Silveira / Banca: Mauro Pereira Otero / Resumo: É comum serem identificados trincas em alvenaria de tijolos, não somente depois de um certo tempo da construção, mas também por vezes aparecem estas trincas desagradáveis durante a própria construção. Depois de alguns estudos é possível verificar que, dentre outros fatores, um deles é devido à umidade absorvida pelos elementos da parede (tijolos e argamssa). Es umidade pode vir do ar (dia únmido), de chuvas, do solo e até do excesso da quantidade de água contida na argamassa de assentamento. Ela em contato com os elementos da parede, porosos, concorrerá para a percolação capilar dando origem a esforços internos consideráveis, que virão provocar deslocamentos, que via de regra são diferentes para o tijolo e para a argamssa("delta"Ltijolo#"delta"Largamassa), daí então o aparecimento de trincas. Portanto, com vistas a esses fenômenos, resolveu-se estudar este assunto de forma inicial, sem grande e sofisticado equipamento de medição do escoamento capilar, mas conduzindo a análise para obtenção de valores médios de pressão e velocidade de fluído, medindo as deformações gerais dos elementos tijolos. / Abastract: It is common to find cracks in bricks masonry not just after some time of the construction but, the appearance of these unpleasant cracks, are also common during the contruction work. After some research it is possible to verify that, among other factors, one of its cause is due to the humidity absorbed by the elements of the wall (bricks and mortar. Now, this humidity may come from the air (e. g. a humid day), from rain, from the ground and from an exaggerated amount of the water mixed in the mortar laying. These when in contact with theporous elements of the wall will induce a capillary percolation that creates considerable internal stress that will give origin to displacements. These displacement are usually different for the brick and for de mortar ("delta"LBrick#"delta"L mortar) and because of that cracks will appear. With that in mind, it was decided to study this subject initially without bulky and sophisticated equipment for the measurement of capillary flow. Our approach was to make the analysis by obtaining medium values for the pressure and fluid speed, measuring the general deformations of the bricks. / Mestre