Ο σκοπός της διατριβής αυτής είναι η γενίκευση και η μελέτη των προβλημάτων ελέγχου του δυαδικού οριακού στρώματος σε ένα γενικότερο πρόβλημα στο οποίο το μαγνητικό πεδίο και η έγχυση ξένου ρευστού συνυπάρχουν και αλληλεπιδρούν, ενώ ταυτόχρονα η ροή χαρακτηρίζεται ως ροή ολίσθησης, λόγω της αραίωσης του μέσου. / --

Heat transfer analysis of two-dimensional, incompressible, constant property, hydrodynamically developed, thermally developing, single phase laminar flow in microtubes and microchannels between parallel plates with negligible axial conduction is performed for constant wall temperature and constant wall heat flux thermal boundary conditions for slip flow regime. Fully developed velocity profile is determined analytically, and energy equation is solved by using finite difference method for both of the geometries. The rarefaction effect which is important for flow in low pressures or flow in microchannels is imposed to the boundary conditions of the momentum and energy equations. The viscous dissipation term which is important for high speed flows or flows in long pipelines is included in the energy equation. The effects of rarefaction and viscous heating on temperature profile and local Nusselt number are discussed. The results of the numerical method are verified with the well-known analytical results of the flow in macrochannels (i.e. Kn =0, Br =0) and with the available analytical results of flow in microchannels for simplified cases. The results show significant deviations from the flow in macrochannels.

En raison du développement croissant des MEMS (Micro Electro Mechanical Systems), l'étude des écoulements de liquide ou de gaz et des transferts thermiques dans des conduites, chauffées ou non, dont le diamètre hydraulique est de l'ordre de quelques microns, a suscité un nombre considérable de travaux au cours des vingt dernières années. À cette échelle, le nombre de Knudsen peut être important (Kn>0,01), ce qui donne naissance à des phénomènes de glissement, de saut de température et de pompage thermique qui s'ajoutent aux effets de compressibilité, telles que la puissance due aux forces de pression et à la puissance des contraintes visqueuses et aux variations des propriétés du fluide avec la température. Dans les modélisations de la littérature, ces phénomènes sont rarement pris en compte simultanément et sont souvent partiellement négligés, sans justification. Notre démarche consiste à proposer une modélisation des micro-écoulements gazeux se rapprochant au mieux de la réalité en prenant en compte tous les phénomènes et à étudier les effets de chacun d'entre eux. L'étude est, en premier lieu, menée en utilisant un code commercial, résolvant les équations de conservation par la méthode des volumes finis et adapté par le biais de sous-programmes développés au cours de cette thèse. Des validations ont été effectuées pour des problèmes allant des plus simples (incompressibles, non glissants) aux plus complexes (compressibles, glissants). Cette étude a permis de mettre en évidence les problèmes liés à la modélisation quand les nombres de Péclet des écoulements sont inférieurs à l'unité (Pe < 1). Dans ce cas, les effets de diffusion inverse sont dominants et l'utilisation d'extensions à l'amont de la conduite devient incontournable. Les effets de compressibilité qui se traduisent par des détentes du gaz près de la sortie de la conduite (accélération + refroidissement) ont été analysés. Enfin, des comparaisons ont été effectuées avec des solutions analytiques d'écoulements compressibles et glissants, supposés isothermes en imposant de faibles variations de pression. Nous avons pu montrer que ces solutions restent valables, même lorsque les variations de pression sont importantes parce que les détentes ne sont localisées que près de la sortie de la conduite et n'influencent donc pas les propriétés globales de l'écoulement. La suite du travail a été réalisée à l'aide d'un code de calcul développé au laboratoire et validé pour les écoulements à grandes échelles. Des conditions aux limites de glissement ont été introduites afin de l'adapter à la problématique de ce travail de thèse. Compte-tenu de ses performances (précision et rapidité des calculs en particulier), ce code a permis de réaliser une étude paramétrique sur une large gamme de pressions d'entrée et de sortie, de telle sorte à balayer tous les types d'écoulements : de peu à très compressibles et de peu à très glissants. Les résultats sont d'abord présentés pour des écoulements quasi-isothermes puis comparés aux solutions analytiques afin de tester ces dernières sur une plus large gamme de pression. Enfin, de nouveaux résultats ont été obtenus pour des écoulements chauffés. Des corrélations, en fonction des paramètres adimensionnels caractéristiques de ces écoulements, ont été obtenues pour les modèles complets à l'aide d'un logiciel de statistiques et de plans d'expériences. Des comparaisons à des modèles simplifiés ont été effectuées pour évaluer les erreurs commises lorsque certains termes sont négligés / These phenomena are rarely taken into account all at once, at least one of them is neglected and often without justification. Our approach is to get as close as possible to reality by taking into account all the phenomena that appear at once and then to study the effect of the phenomena most often overlooked. First, the study is conducted using a commercial code for solving the conservation equations by the finite volume method. Validations were performed for problems ranging from the simplest (incompressible, non-slip flow) to the more complex (compressible, slip flow). This study highlighted the problems associated with simulations when the flows Peclet numbers are less than unity (Pe <1). In this case, the inverse diffusion effects are dominant and the use of extensions at the upstream becomes unavoidable. In addition, compressibility effects were identified; they have resulted in expansions of the gas near the exit of the pipe (acceleration + cooling). Finally, comparisons were made with analytical solutions of compressible slip flows assumed isothermal by imposing small variations of pressure. We showed that these solutions remain valid even if the pressure variations are important because the detents are located only near the exit of the pipe. In this case, they do not affect the properties of the flow. Further works were carried out using an in-house computer code, previously developed and validated in the laboratory for flows with large scales and for which slipping limits conditions have been added so that it can properly resolve slip flows. In view of its accuracy and performances in terms of CPU-time, the code allowed us to achieve a parametric study on a wide range of input and output pressures, so as to sweep all runoff from few to very compressible and few to very slip flow. The results were first presented for quasi-isothermal flow, which subsequently were compared to analytical solutions to test these ones on a wider range of pressure. Finally, the results were made for heated flows. Correlations have been obtained for a complete model using a statistical based software and design of experiments. Comparisons to simplified models were performed to assess the inaccuracies linked to the omission of terms often overlooked in the literature

Nous présentons dans cette thèse une adaptation de la technique de vélocimétrie par marquage moléculaire (MTV) aux micro-écoulements gazeux. Les effets de luminescence de l’acétone gazeuse excitée par un rayonnement UV, mis en jeu dans la MTV, ont été analysés en vue d’une application en vélocimétrie à basse pression. La phosphorescence de l’acétone diminue fortement avec la pression, pour devenir non détectable à des pressions de l’ordre du kPa. En revanche, la fluorescence reste détectable à des pressions de l’ordre de la dizaine de Pa. L’analyse des déplacements de molécules luminescentes au sein de l’écoulement porteur a montré que les effets de la diffusion moléculaire sont importants et augmentent avec la raréfaction du gaz. Une méthode de reconstruction, basée sur l’équation d’advection diffusion, a été développée. Elle permet d’extraire le profil de vitesse à partir de l’analyse du déplacement et de la déformation de la zone marquée, en prenant en compte la diffusion des molécules luminescentes. L’analyse d’un écoulement de Poiseuille dans un canal de section rectangulaire et de dimensions millimétriques, sous des pressions de l’ordre de 100 kPa, a montré la capacité de la MTV à extraire avec précision la vitesse locale en écoulements confinés, lorsque la méthode de reconstruction est appliquée. Une feuille de route pour l’analyse future par MTV des écoulements raréfiés, notamment dans le régime glissant, est finalement proposée. / In the present thesis we present an adaptation of Molecular Tagging Velocimetry (MTV) to gas microflows. The photo-luminescence effects of gaseous acetone excited by UV light, implemented in MTV, have been analyzed in various pressure conditions. The acetone phosphorescence shows a drastic decrease with pressure and becomes non measurable for pressures lower than 1 kPa. On the other hand, fluorescence shows a slower decrease and remains clearly detectible at pressures as low as 10 Pa. The motion of tracer molecules within the carrier flow has been studied. The analysis of the displacement of the tagged molecules has shown the strong influence of molecular diffusion, this influence being increased with the gas rarefaction. A reconstruction method based on the advection-diffusion equation has been developed. It allows to extract the velocity profile from the analysis of the displacement and deformation of the tagged region, taking into account the diffusion of tracer molecules. This reconstruction method has been successfully implemented to analyse a Poiseuille flow in a rectangular millimetric channel, under atmospheric pressure conditions, and the capability of MTV to accurately extract the local velocity in confined gas flows has been demonstrated. Finally, some short term perspectives have been proposed with the aim to help achieving rarefied flows analysis by MTV.

Certaines applications industrielles nécessitent des niveaux d’étanchéité exceptionnels pour permettre la réalisation d’un vide poussé ou pour répondre à des enjeux de sécurité radiologique par exemple. Ces niveaux de haute étanchéité statique sur des assemblages démontables sont obtenus à l’aide de joints entièrement métalliques. La fuite résultante de l’assemblage n’est due qu’à la persistance d’un champ des ouvertures à l’interface entre le joint d’étanchéité et la bride d’assemblage, conséquence d’un contact imparfait entre les deux surfaces rugueuses. Le champ des ouvertures à l’interface de contact est assimilable à une fracture rugueuse hétérogène, de nature multi-échelle, et peut en principe être obtenu par un calcul de déformations mécaniques préalable. Dans ce travail, on s’intéressera plus particulièrement à l’écoulement gazeux raréfié dans le régime glissant au sein de cette fracture. Pour les régimes modérément raréfiés,l’écoulement est modélisé par l’équation de Reynolds faiblement compressible avec correction de glissement de premier ordre aux parois que l’on développe. On effectue ensuite un changement d’échelle par la méthode de la prise de moyenne volumique, permettant d’établir un modèle macroscopique d’écoulement à l’échelle d’un élément représentatif, où le débit massique est relié au gradient de pression via le tenseur de transmissivité. Celui-ci, caractéristique de l’élément représentatif de fracture, est calculé par résolution d’un problème de fermeture et est dépendant de la microstructure ainsi que du libre parcours moyen représentatif sur l’élément. Pour remonter à l’écoulement dans l’ensemble de la fracture, hétérogène à cette échelle, celle-ci est subdiviséeen pavés sur chacun desquels est calculé un tenseur de transmissivité local par la méthode sus-citée. Ensuite, l’écoulement dans ce champ de tenseurs est résolu par une méthode des éléments finis de frontière, donnant la transmissivité apparente glissante du joint dans son ensemble. Cette approche à deux échelles, vue comme outil d’aide à la conception, permet une réduction de la complexité de calcul par rapport à une simulation directe, rendant possible une analyse plus efficace du comportement d’un système d’étanchéité. Pour valider l’utilisation du modèle de glissement d’un point de vue macroscopique et s’affranchir des incertitudes sur le calcul de déformation mécanique, des puces nanofluidiques de type réseau hétérogène de canaux droits sont fabriquées par photolithographie par niveaux de gris. Des essais expérimentaux de mesure de fuite sont réalisés sur ces géométries modèles, représentant des joints idéalisés. Ces essais sont effectués en appliquant une forte différence de pression d’hélium par utilisation d’un spectromètre de masse mesurant la fuite, produisant une condition de vide en sortie de puce.Selon les puces, les régimes de raréfaction atteints vont alors du régime glissant au régime moléculaire. Le débit de fuite mesuré est alors supérieur à celui prédit par le modèle de premier ordre, l’écart restant inférieur à un ordre de grandeur quel que soit le régime / Some industrial applications require exceptional sealing levels to maintain ultra-high vacuumconditions or for radiological safety concerns for example. Such high performance static sealingconditions on mechanical assemblies are reached using entirely metallic gaskets. The resultingleak-rate is only due to the persistence of an aperture field at the seal-flange interface,consequence of a non-ideal contact between the two rough surfaces. This aperture field can beviewed as a rough and heterogeneous fracture, of multi-scale nature, and can be obtained by aprior contact mechanics computation. In this work, we are interested on the rarefied flow of a gasin this fracture, drawing our attention to the slip regime. For such moderately rarefied regime, theflow is described by the slightly compressible Reynolds equation with a first-order slip-flowcorrection at the walls, which we develop. Using the method of volume averaging, an upscalingprocedure is performed to derive the macroscopic flow model at the scale of a representativeelement, and where the mass flow rate is related to the pressure gradient by the transmissivitytensor. This latter is characteristic of the representative fracture element and is obtained by solvingan auxiliary closure problem which depends on the micro-structure as well as the representativemean free path on the element. To compute the flow in the whole fracture, heterogeneous at thisscale, it is subdivided in tiles on which a transmissivity tensor is locally computed by theaforementioned method. Then, the flow problem in this tensor field is solved using a boundaryelement method, leading to the apparent slip-corrected transmissivity of the entire aperture field.This two-scale approach is a conception tool which reduces the overall complexity with respect toa direct numerical simulation, allowing a more efficient analysis of the behavior of a sealingassembly. To validate the use of slip models at the macroscopic level and to eliminate theuncertainties of the contact mechanics computation, nanofluidic chips composed ofheterogeneous network of straight channels are fabricated using a grayscale photolithographytechnique. Experimental measurements of the leak-rate are performed on these idealizedgeometries that mimic a seal assembly. They are realized by applying a strong helium pressuredifference on the chip using a mass spectrometer to measure the leak, which produces a nearvacuum condition at the outlet. Depending of the chip, the rarefaction regime ranges from slip tofree-molecular. The measured leak-rate is greater than predicted by the first order model, thoughbeing of the same order of magnitude whatever the regime

La perméabilité est l'un de paramètres importants pour tous les procédés faisant intervenir des transferts couplés de chaleur et de masse. Sa valeur est directement liée à la morphologie du réseau de pores, clairement double échelle dans le cas du bois. Il existe plusieurs outils d’investigation 3-D par exemple la micro-tomographie voire nano-tomographie pour décrire la morphologie des pores. Néanmoins, ces investigations 3-D échouent pour les plus petits pores dans le bois, qui peuvent avoir une taille largement inférieure au micromètre. Ce travail de thèse utilise l'effet du libre parcours moyen du gaz sur la perméabilité apparente pour identifier la taille des pores utilisés par le cheminement du fluide.Une approche équilibrée entre expérimentation et modélisation est proposée. Dans la première partie du travail, nous avons développé un dispositif original destiné à mesurer la perméabilité apparente sur une large plage de niveau de pression moyenne. Ce dispositif ne comporte pas de débitmètre : le flux massique est simplement obtenu par la relaxation de la différence de pression entre deux réservoirs de volume connu.Ce dispositif a permis de mesurer la perméabilité apparente en fonction de la pression moyenne pour différents matériaux. Les valeurs obtenues de perméabilité intrinsèque sont en bon accord avec les publications précédentes.L'effet de la pression moyenne sur la perméabilité apparente a été analysé pour identifier des paramètres structuraux des milieux poreux. Partant d'un agencement série/parallèle de pores, les paramètres structuraux sont identifiés par méthode inverse en tirant profit du changement de régime d’écoulement avec le nombre de Knudsen (du régime de Darcy vers le régime de glissement moléculaire lorsque le nombre de Knudsen passe de zéro à l’infini).Cette approche a été validée avec des membranes à pore monodisperse grâce à une observation en microscopie électronique à balayage. Les paramètres ont ensuite été identifiés sur des échantillons de bois, selon différentes directions et sur des matériaux à base de bois. Ils confirment certaines données de la littérature, et apportent un éclairage nouveau, notamment sur les chemins de migration transverse chez le hêtre. / Permeability is one of the important parameters for all processes involving coupled heat and mass transfer. Its value is directly related to the morphology of the pore network, clearly a dual scale organisation in the case of wood. Nowadays, several 3D investigative tools exist, such as micro-tomography or nano-tomography. However, these 3-D investigations fail for the smallest pores active in fluid flow in wood, due to their submicron size. The present work takes advantage of the effect of the mean free path on the apparent gaseous permeability to identify the pore size of the pore network.A balanced approach between experimentation and modelling is proposed. In the first part of the work we have developed an original device for measuring the apparent permeability over a wide range of average pressure. This device was conceived to work without flowmeter: the mass flow is obtained by the relaxation of the pressure difference between two tanks. This device was used to measure the apparent permeability as a function of the average pressure for different materials.The intrinsic permeability values obtained are in good agreement with literature data.The effect of the average pressure on the apparent permeability was analysed to identify structural parameters of the porous media. Starting from a serial / parallel arrangement, the structural parameters are identified by inverse method taking advantage of the dependence of the flow regime with Knudsen's number (from Darcy's regime to pure molecular slip regime when Knudsen's number shifts from zero to infinity).This approach was validated with a monodisperse membrane, whose pore size was determined using a scanning electron microscopy. Structural parameters were then identified on wood samples, measured along different directions and on wood-based materials. They confirm literature data and bring new outcomes, namely regarding the fluid pathway in beech in radial and tangential directions.

Peristalsis is a mechanism of pumping fluids in ducts when a progressive wave of area contraction or expansion propagates along the length of a distensible tube containing fluid. It induces in general propulsive and mixing movements and pumps the fluids against pressure rise. Physiologically, peristaltic action is an inherent property of smooth muscle contraction. It is an automatic and vital process that drives the urine from the kidney to the bladder, food through the digestive tract, bile from the gall-bladder into the duodenum, movement of ovum in the fallopian tube and many other situations. A major industrial application of this principle is in the design of roller pumps, which are used in pumping fluids without being contaminated due to the contact with the pumping machinery. Even though peristalsis is a well-known mechanism in biological system, the first theoretical and experimental analysis of its fluid dynamics aspects were given four decades ago. In reality, the peristaltic flow problems are unsteady moving free boundary value problems where the shape of the wave on flexible tube wall is not known apriori. But the mathematical models on peristaltic transport considered in the literature deal with a prescribed train of waves moving with constant speed on the flexible boundaries and they are studied in either a fixed frame or a wave frame moving with constant velocity of the wave. In a wave frame the moving walls become stationary wavy walls. Further the motion could be treated steady under the assumptions that the peristaltic wave train is periodic, the pressure difference across the length of the tube is constant and the tube length is an integral multiple of the wavelength. Some mathematical models of peristaltic flows representing some physiological situations are studied using a wave frame of reference in this thesis. The important characteristics of these flows namely pumping (variation of time averaged flux with difference in pressures across one wavelength), trapping (splitting of streamlines enclosing a bolus which moves as a whole along with the wave), reflux phenomena (the presence of some fluid particles whose mean motion over one cycle is against the net pumping direction) are discussed in detail. A brief general introduction to the peristaltic transport and their application in physiological fluid dynamics is presented in chapter one. In the second chapter, the peristaltic transport of an incompressible viscous Newtonian fluid in an asymmetric channel is studied under long wavelength and low-Reynolds number assumptions. Choosing the peristaltic wave train on the walls to have different amplitudes and phase produces the channel asymmetry. This study is motivated by the intra-uterine fluid flow induced by uterine wall contractions which represent a peristaltic flow in an asymmetric channel and this flow is responsible for embryo transport to a successful implantation site. The solution for the stream function is obtained by neglecting inertia and curvature effects. The streamlines are plotted in both fixed and wave frames. The effects of different geometric parameters causing asymmetry like phase difference; varying channel width and wave amplitudes are investigated on the pumping characteristics, streamline pat-tern, trapping and reflux phenomena. It is observed that the pumping against pressure rise, trapping and reflux layer exists only when cross-section of the channel varies along the axis. The limits on the time averaged flux for trapping and reflux are obtained. The peristaltic waves on the walls with same amplitudes propagating in phase produce zero flux rate as the channel cross-section remains the same through out. The trapping and reflux regions reduce for asymmetric channels compared to symmetric channels. The flow of an incompressible viscous fluid driven by the traveling waves along the boundaries of an asymmetric channel is studied in the chapter three, when inertia and streamline curvature effects are not negligible. It was well documented that the inertial forces cannot be ignored in the pharyngeal phase of bolus transport. Choosing the wave train on the walls to have different amplitudes and phases produces the channel asymmetry here. An asymptotic solution is obtained in powers of a geometric parameter £, the ratio of the channel width to the wavelength, giving curvature and inertia effects. A domain transformation is used to transform the channel of variable cross section to a uniform cross section, and this facilitates in easy way of finding closed form solutions at higher orders. The solutions are presented upto second order in 6. It has been found that, the relation connecting the pressure gradient and time average flux rate is a cubic leading to a non-unique of flux for a given pressure gradient. A uniqueness criterion is derived which restricts the parameters to get a unique flux for a prescribed pressure difference. The effects of inertia and curvature on peristaltic pumping, trapping and shear stress are discussed for various parameters governing the flow for symmetric and asymmetric channels and compared with the existing results in the literature. Even under a favourable pressure gradient the possibility of fluid flow in a direction opposite to the direction of the waves propagating on the walls is detected as in the case of some non-Newtonian fluids. It is noticed that the Reynolds number and asymmetry of the wall geometry may play an important role in producing mixing. The appearance of a second trapped bolus near the down streamside of the channel for some Reynolds number is a new feature. Further, the non-zero curvature produces three trapped boluses for high Reynolds number in symmetric channel as well as for inertia free flow in an asymmetric channel. Another interesting phenomena is that the shear stress distribution on the walls vanishes at some points but it does not indicate any flow separation as the MRS criteria is not satisfied. The gastrointestinal tract is surrounded by a number of muscle layers having smooth muscle. The most important smooth muscle layers in gastrointestinal tract are submucosa and a layer of epithelial cells and these- are responsible for the absorption of nutrients and water in the intestine. These layers consist of many folds and there are pores through out the tight junctions of them. Thus a study of peristaltic transport with porous peripheral layer and porous boundaries of a duct are important. Motivated by this the flow in gastrointestinal tract is mathematically modeled by a peristaltic flow of two fluid system in a two-dimensional channel with a porous peripheral layer and a Newtonian fluid core layer, in chapter four. The fluid flow is investigated under the assumptions of long wavelength and low Reynolds number in a wave frame of reference. Brinkman extended Darcy equation is utilized to model the flow in the porous peripheral layer. A shear stress jump boundary condition of Ochoa-Tapia and Whitaker is used at the interface between porous and fluid regions together with continuity of velocity and normal stress conditions. Here one needs an extra assumption that the fluid interface and the peristaltic wave on the boundary have the same period in addition to the constant pressure difference at the ends of channel and the length of the channel to be an integral multiple of the wavelength, to consider the flow to be steady. The interface is determined as a part of the solution using the conservation of mass in both the porous and fluid regions independently. Matlab packages are used to solve the transcendental equation governing it. An interval of critical time averaged flux Q is obtained for the existence of a unique solution for the interphase. The physical quantities of importance in peristaltic transport namely, pumping, trapping and reflux are discussed for various parameters of interest governing the flow like Darcy number Da, porosity 6, shear-stress jump constant /3, viscosity ratio /i. It is observed that the peristalsis works as a pump against greater pressure rise with a porous medium in the peripheral layer than a viscous fluid. The limits on the time averaged flux Q for trapping in the core layer are obtained. The existence of reflux near the axis is observed for small values of Darcy number and large values of /?. Chapter five deals with the peristaltic transport in a tube with a poroflexible wall and having a porous material layer in the peripheral region and a Newtonian fluid in the core region. Flow in tube may be more realistic to model a flow in gastrointestinal system. At the poroflexible wall, a slip boundary condition of Saffman.type is used. The fluid flow is studied in a wave frame of reference under lubrication approach. Brinkman extended Darcy equation in cylindrical polar coordinates is considered for the porous medium with a shear-stress jump boundary condition of Ochoa-Tapia and Whitaker at the interface of porous and fluid regions together with the continuity of velocity and normal stress. The interface is found as a part of the solutions using the conservation of mass in both the regions of deformable porous medium and fluid medium independently. The interface equation turns out to be a transcendental equation involving modified Bessel functions and it is solved by using Matlab packages. The uniqueness criterion of the solutions for the interface equation in the flow region is determined for certain values of time averaged flux Q. Pumping characteristics, trapping and reflux phenomena are discussed for various parameters of interest governing the flow like, wall slip constant fc, Darcy number Da, viscosity ratio /x. shear stress jump constant f) and peripheral layer thickness 7. The slip condition at the boundary, arising due to the poroflexible nature of the wall, enhances pumping. The trapped bolus volume in the core layer is observed to decrease with a decrease in Da, /i and k and an increase in /?. The reflux phenomena is discussed in detail. The trapping limits on the flux rate Q in the core region are obtained. As the behaviour of most of the physiological fluids is known to be non-Newtonian, the peristaltic flows of power-law and micro polar fluids are investigated in the next two chapters. In chapter six, the peristaltic transport of a power-law fluid in an axisymmetric tube having poroflexible wall is studied. The power-law model of Ostwald-de Waele type is considered, which accommodates the study of both shear thinning and shear thickening fluids. The flow characteristics are studied in wave frame analysis under lubrication approach. The slip boundary conditions of Beavers-Joseph and Saffman type are considered at the wall in obtaining solutions for the flow and resulting pumping characteristics are investigated with a straight section dominated (SSD) wave form other than the sinusoidal one. It is observed that the time mean flow becomes negative in free pumping for a shear thickening fluid with a SSD expansion wave and the same is observed for a SSD contraction wave in the case of shear thinning fluid. The pressure rise increases with increasing of Darcy number Da against which the peristalsis works as a pump and decreases for an increase in Beaver-Joseph constant a. Peristalsis works as a pump against a greater pressure rise for a shear thickening fluid and the opposite happens for a shear thinning fluid compared with Newtonian fluid. Trapping and reflux phenomena are discussed for various parameters of interest governing the flow like Da, a and the fluid behaviour index n. The trapping limits on Q are derived. The trapped bolus volume for sinusoidal wave is observed to decrease as the fluid behaviour index decreases, i.e as the fluid behaviour changes from shear thickening to shear thinning, where as it increases with an increase in Darcy number. The rheological properties of fluid, wave shape and porous nature of the wall play an important role in the peristaltic transport and may be useful in understanding the transport of chyme in small intestine. The chapter seven contains the study of peristaltic transport of a micropolar fluid in an axisymmetric tube. Micropolar fluids exhibit some microscopic effects arising from the local structure and micro motion of the fluid elements. Further, they can sustain couple stresses. It is speculated that, in microcirculation, peristalsis may be involved as well in the vasomotion of small blood vessels which change their diameters periodically. Therefore, modelling blood by a micropolar fluid may be more appropriate. The closed form solutions are obtained for velocity, microrotation components, as well as the stream function under the assumption of long wavelength and low Reynolds number. The solution contains new additional parameters namely, N the coupling number and m the microploar parameter. In the case of free pumping (pressure difference Ap = 0) the difference in pumping flux is observed to be very small for Newtonian and micropolar fluids but in the case of pumping (Ap > 0) the characteristics are significantly altered for different N and m. It is observed that the peristalsis in micropolar fluids works as a pump against a greater pressure rise compared with a Newtonian fluid. Streamline patterns which depict trapping phenomena aie presented for different parameter ranges. The limit on the trapping of the center streamline is obtained. The effects of N and m on friction force for different Ap are discussed. The nomenclature of symbols in each chapter is independent of the other. Each of the chapter has its own appendix and they are numbered with the corresponding roman number of the chapters. The purpose of the study here is not to represent exactly the functioning of various physiological applications, but rather to understand the fluid-mechanical aspects inherent in the problems of peristaltic transport.

Ce travail de thèse porte sur le développement de la technique de Vélocimétrie par Marquage Moléculaire (Molecular Tagging Velocimetry - MTV) pour l’étude expérimentale des micro-écoulements gazeux internes. Les micro-écoulements gazeux sont des écoulements raréfiés, caractérisés par un nombre de Knudsen non négligeable. L’analyse de la littérature montre un besoin crucial de données expérimentales de grandeurs locales relatives aux micro-écoulements gazeux. Ces données permettraient une discussion pertinente de la précision et des limites d’applicabilité des différents modèles théoriques proposés dans la littérature pour l’étude du régime de glissement, régime raréfié le plus souvent rencontré en microfluidique gazeuse. Dans cette optique, un banc d’essais expérimental a été développé pour la mesure de champs de vitesses par MTV. La technique consiste à suivre des molécules traceuses d’acétone introduites dans le gaz en écoulement et qui deviennent phosphorescentes lorsqu’elles sont excitées par une source lumineuse UV. Les différents compromis pris en compte pour le développement de ce banc (choix du traceur et du matériau, conception du canal instrumenté,…), ainsi que les techniques d’acquisition et de traitement de signal sont détaillés dans le manuscrit. L’analyse expérimentale commence par une étude du signal de phosphorescence de l’acétone. Ensuite, la technique de vélocimétrie par marquage moléculaire est validée par la mesure de champs de vitesses dans des écoulements laminaires confinés en régime non raréfié. Les résultats obtenus sont comparés à des profils de vitesse théoriques d’un écoulement de Poiseuille à pression atmosphérique. Enfin, les premiers résultats obtenus à basse pression sont présentés et commentés. La détection du signal à un niveau de pression de 1kPa est encourageante et offre de nombreuses perspectives pour l’exploration d’écoulements en régime raréfié / This thesis focuses on the development of Molecular Tagging Velocimetry (MTV) technique for the experimental analysis of internal microflows of gases. Gaseous microflows are rarefied flows characterized by a non-negligible Knudsen number. A literature review highlights a crucial need of experimental data on velocity fields within gaseous microflows. These data are required for a relevant discussion on the validity and limits of applicability of the different boundary conditions proposed in the slip flow, which is a regime often encountered in gaseous microsystems. An experimental setup has been designed for analyzing by MTV the velocity distribution in microchannels. The technique consists in detecting the displacement of acetone molecules, introduced as tracers in a gas flow; these molecules exhibit phosphorescence once excited by a UV light source. The various compromises taken into account for the setup design (choice of tracer, laser, channel material and design, camera and intensifier…), as well as the acquisition and processing techniques are detailed in the manuscript. The experimental analysis starts with a study of the acetone phosphorescence signal. Then, the MTV technique is validated by velocity field measurements in internal laminar flows through a rectangular minichannel in non-rarefied regime. The obtained results are successfully compared to the theoretical velocity profile of a Poiseuille flow. Finally, preliminary results obtained at lower pressures are presented and commented. The signal detection at a pressure level as low as 1 kPa is encouraging and draws various perspectives for the exploration of rarefied regimes