An ascending aortic aneurysm is a pathologic enlargement of the ascending portion of the aorta. Comorbidities of dilation include aortic valve disease and connective tissue disorders. If the ascending aorta exceeds a threshold diameter, open heart surgery is recommended. This is a traumatic procedure and the recovery is demanding. As our population ages and with improved technologies to diagnose the disease, the number of cases will increase every year. Understanding the mechanics of ascending aortic tissue will help cardiac surgeons make timely decisions on when to intervene. There are many ways to characterize the mechanical properties of aortic tissue. In this study, we used biaxial and uniaxial tensile testing with an optical tracking system to record the Green-Lagrangian (Green strain) strain. Engineering and true stiffness values were calculated and compared along with patient characteristics. Aortas were classified by valve type as healthy, tricuspid, bicuspid type 1 and bicuspid type 2. The results show that diseased tissue does behave differently than healthy tissue indicating that a local remodeling does occurs to the aortic wall. There are also differences in the mechanics between the types of diseased valves suggesting that valve type also affects the way the aortic wall responds to the disturbed hemodynamic environment. Correlations between stiffness and patient characteristics show that no matter which experimented technique or method of stiffness calculation is used, relationships are generally conserved. The only difference is the magnitude of the elastic modulus. The conclusions drawn from the data would not change whether biaxial or uniaxial experiments were performed. However when comparing engineering and true stiffness, only 7/12 covariances were similar and therefore the conclusions are inconsistent. / Un anévrisme de l'aorte ascendante est un agrandissement pathologique de l'aorte ascendante. Les comorbidités de dilatation incluent les maladies de la valve aortique et du tissu conjonctif. Si l'aorte ascendante dépasse un diamètre seuil, la chirurgie à cœur ouvert est recommandée. Il s'agit d'une procédure traumatique et la récupération est exigeante. Comme notre population vieillit et que les technologies pour diagnostiquer la maladie se sont beaucoup améliorées, le nombre de cas décelé va augmenter chaque année. Comprendre les mécanismes du tissu aortique ascendant aidera les chirurgiens cardiaques à prendre les bonnes décisions au bon moment. Il y a plusieurs façons de caractériser les propriétés mécaniques du tissu aortique. Dans cette étude, nous avons utilisé les essais de traction biaxiale et uniaxiale avec un système de suivi optique pour enregistrer la souche Green-Lagrange (souche verte). Valeurs de rigidité d'ingénierie et de vrai ont été calculées et comparées avec les caractéristiques des patients. Les aortes ont été classées en 4 types de valves : bonne santé, tricuspides, bicuspides de type 1 et bicuspides de type 2. Les résultats montrent que les tissus malades sont différents des tissus en bonne santé qui indiquent qu'un remodelage local se produit sur la paroi aortique. Il existe aussi des différences dans le mécanisme des différents types de valvules qui suggèrent que le type de valve affecte également la façon dont la paroi aortique répond à l'environnement perturbé hémodynamique. Les corrélations entre les caractéristiques de rigidité et le patient nous montrent que quelque-soit la technique ou la méthode de calcul utilisée pour la rigidité, les relations sont généralement conservées. La seule différence est la grandeur du module d'élasticité. Les conclusions tirées de ces données ne changeraient pas, peu importe le type d'expérimentation effectué; biaxiale ou uniaxiale. Cependant, lorsque l'on compare la rigidité d'ingénierie et de vrai, seulement 7 covariances sur 12 sont similaires et les conclusions sont contradictoires.

One of the goals in catalyst research for proton exchange membrane fuel cells (PEMFCs) is to find a cost-efficient alternative to platinum. Due to sluggish kinetics, the major requirement of the platinum comes from the catalyst layer used for the oxygen reduction reaction (ORR). Functionalized carbon nanomaterials present themselves as good candidates for the replacement of platinum due to their low cost, excellent electrical conductivity, and chemical resistance to acidic and basic environments. In this work, graphene nanoflakes (GNFs), which are nanopowders consisting of stacked graphene sheets, were used to support atomic iron as a non-noble metal catalyst. In the first stage of the study the iron-based catalyst was synthesized. Synthesis steps include the production of GNFs in methane plasma, adsorption of ferric acetate, and pyrolysis in ammonia-rich atmosphere. The catalyst structure was characterized at various stages throughout the synthesis steps and it was found that 0.28 atomic percent of iron could successfully be incorporated onto the surface. However, the synthesis method employed caused a general decrease to all calculated crystallinity parameters: purity decreased by 28%, crystallite size decreased by a factor of 2, and the average length of graphene plane decreased by a factor of 4. Characterization was also performed on the catalyst layer after it had been exposed to the PEMFC environment, revealing that the crystallinity parameters actually improved with respect to exposure time: after 100 hours purity increased by 32%, crystallite size increased by 25%, and the average length of graphene plane increased by 107%. Exposure to the PEMFC environment repairs the damage done to the original GNFs during the synthesis steps. The synthesized catalyst was used in the catalyst layer for the ORR of a PEMFC with a 1 cm2 active surface. A current of 150 mA/cm2 was observed at an applied voltage of 0.5 Volts with a catalyst loading of 1 mg. When the current is normalized with respect to the amount of metal present, the result of 11.8 A/mg of metal catalyst from the present catalyst out-performs most platinum-based catalysts being used in industry; current platinum catalyst have values ranging from 3 to 14 A/mg of platinum. In stability experiments, no losses were observed at the end of 100-hours long experiments performed at an applied voltage of 0.5 Volts. This represents a great improvement over comparable iron-based catalysts, which show a 45% loss under identical test conditions. The increased stability of the catalyst support structure demonstrates the advantage of the high crystallinity and large crystalline lengths of the GNFs in comparison to other commercial carbon blacks. / Un des objectifs de la recherche sur les catalyseurs pour les piles à combustible à membrane électrolyte polymérique (PCMEP) est de trouver une alternative moins coûteuse au platine. En raison d'une cinétique lente, le platine est surtout utilisé dans la couche de catalyseur au niveau de la cathode pour la réaction de réduction de l'oxygène (RRO). Les nanomatériaux de carbone fonctionnalisés se présentent comme de bons candidats pour le remplacement du platine en raison de leur faible coût, d'une excellente conductivité électrique et d'une résistance chimique aux milieux acides et basiques. Dans ce travail, les nanoflocons de graphène (NFG) constitués en moyenne d'une dixaine de plans de graphène empilées, ont été utilisés comme support aux atomes de fer pour créer un catalyseur métallique non noble. Lors d'une première étape, le catalyseur à base de fer a été synthétisé. Les étapes de synthèse comprennent la production des NFG dans le plasma de méthane, l'adsorption de l'acétate ferrique, et la pyrolyse dans une atmosphère riche en ammoniac. La structure du catalyseur a été caractérisée tout au long des étapes de synthèse, et il a été constaté qu'un pourcentage de 0,28 % en atomes de fer ont été incorporé aux structures NFG. Cependant, la méthode de synthèse utilisée a provoqué une baisse générale de tous les paramètres cristallins calculés: la pureté a diminué de 28%, la taille des cristallites a diminué d'un facteur 2, et la taille moyenne des plans de graphène d'un facteur 4. La caractérisation a été également effectuée sur la couche de catalyseur après avoir été exposée à l'environnement PCMEP, révélant que les paramètres cristallins sont effectivement améliorés avec la durée d'exposition. Au bout de 100 heures, la pureté a augmenté de 32%, la taille des cristallites de 25%, et la taille moyenne des plans de graphène de 107%. L'exposition à l'environnement de PCMEP a réduit les dommages causés aux NFG pendant les étapes de synthèse. Le catalyseur synthétisé a été utilisé pour la RRO dans un PCMEP avec une surface active de 1 cm2. Un courant de 150 mA/cm2 a été observé pour une tension appliquée de 0,5 volts et une masse de catalyseur de 1 mg. Lorsque le courant est normalisé par rapport à la quantité de métal présent, le résultat de 11,8 A/mg de métal surpasse les catalyseurs à base de platine les plus utilisés dans l'industrie. Les catalyseurs au platine ont des valeurs allant de 3 à 14 A/mg de platine. Dans les expériences de stabilité, pour une tension appliquée de 0,5 Volts, aucune perte de courant n'a été observée à la fin des 100 heures de l'expérience. Cela représente une grande amélioration par rapport aux autres catalyseurs à base de fer, qui montrent une perte de 45% dans des conditions expérimentales identiques. La stabilité accrue de la structure du catalyseur démontre l'avantage d'utiliser des NFG par rapport à d'autres nanomatériaux de carbone, grâce à leurs cristallinité élevée et leurs grandes longueurs cristallines.

A capillary pressure curves provides important information about the porous system such as porosity, breakthrough pressure, fluid–solid wettability and capillary hysteresis. The measurement of capillary pressure curves is one of the most commonly adopted methods of measuring wettability in a porous system. However, to date these measurements are done only at room temperature because of the limitations of utilising existing equipment at higher temperatures. The capillary behaviour of liquid in porous system can be a different when subjected to harsh condition such as high temperature and pressure. An experimental setup has been designed and developed to measure air-water capillary pressure curves for gas diffusion layers (GDLs) used in polymer electrolyte membrane fuel cells at elevated temperatures. Experiments were conducted at various temperatures in the direction of increasing and decreasing temperatures for Toray 120 and Toray090 with varying PTFE loadings. Notable shifts in the capillary behavior were seen and in some cases the shifts were of the same size as PTFE addition. In untreated samples the results suggest that GDL wettability changes with temperature once the temperature dependence of the surface tension has been accounted for. In treated samples, however, the wettability remains more or less constant with temperature. This behavior is in agreement with the temperature dependent values of water contact angle on PTFE and graphite. Observed shift with temperature for treated samples are more or less reversible during decreasing temperatur / A courbes de pression capillaire fournit des informations importantes sur le système poreux tels que la porosité, la pression de fuite, le liquide-solide mouillabilité et l'hystérésis capillaire. La mesure des courbes de pression capillaire est une des méthodes les plus couramment adoptée pour mesurer la mouillabilité dans un système poreux. Cependant, à ce jour, ces mesures sont effectuées uniquement à la température ambiante en raison des limitations de l'utilisation des équipements existants à des températures élevées. Le comportement capillaire de liquide dans le système poreux peut être un différent lorsqu'elle est soumise à l'état dur, comme la température et une pression élevées. Un dispositif expérimental a été conçu et développé pour mesurer air-eau courbes de pression capillaire pour les couches de diffusion de gaz (GDL) utilisés dans les cellules à électrolyte polymère à combustible à membrane à des températures élevées. Expériences ont été réalisées à différentes températures dans le sens d'une augmentation et la diminution des températures de Toray120 et Toray090 avec différents chargements PTFE. Changements notables dans le comportement capillaire ont été observées et, dans certains cas, les changements étaient de la même taille que l'addition de PTFE. Dans les échantillons non traités les résultats suggèrent que les changements de mouillabilité GDL avec la température une fois que la dépendance en température de la tension de surface a été pris en compte. Dans les échantillons traités, cependant, la mouillabilité reste plus ou moins constant avec la température. Ce comportement est en accord avec les valeurs dépendantes de la température de l'angle de contact de l'eau de PTFE et de graphite. Décalage observé avec la température pour les échantillons traités sont plus ou moins réversible pendant la température diminue.

Model Predictive Control (MPC) has become one of the dominant methods of chemical process control in terms of successful industrial applications. A rich theory has been developed to study the closed loop stability of MPC algorithms when the plant model is perfect, referred to as the nominal stability problem. In practical applications, however, the process model is never perfect and nominal stability results are not strictly applicable. The primary disadvantage of the current design techniques for MPC is their inability to deal explicitly with the plant model uncertainty. In this thesis we develop a new framework for robust MPC synthesis that allows explicit incorporation of the plant uncertainty description in the problem formulation. Model uncertainty is parameterized by ellipsoid bounds on the Finite Impulse Response (FIR) parameters of the plant model. Robust stability is achieved through the addition of constraints that prevent the sequence of the optimal controller costs from increasing for the true plant. The framework developed here can also be used for constant output disturbance rejection. Ward input and soft output constraints can be easily added to the algorithms without affecting the closed loop stability properties.

In this thesis we develop a Nonlinear Robust Model Predictive Control (NRMPC) algorithm for nonlinear plants modeled by second order Volterra series. Robust stability is achieved through the addition of cost function constraints that prevent the sequence of the optimal controller costs from increasing for the true plant. Model uncertainty is parameterized by ellipsoid bounds on the plant parameters. The same approach is used to reject constant output disturbances. If some certain assumptions concerning the disturbance and plant parameters are satisfied, the plant is guaranteed to reach an offset free steady state.

The purpose of this study is to determine the effect of oscillating flow in a conduit in the absence of secondary flow on the rate of mass transfer to the wall, and the parameters that influence the magnitude of the change in the flux. Flow in an artery is pulsatile in nature and it will be shown in a later section that the model discussed in this paper, that of pulsating flow in a distensible tube is applicable to such flows. Nutrition of the arterial wall depends on the rate of diffusion of luminal plasma to and through the wall. This research is of importance in understanding the functioning of the cardiovascular system. It also finds application in the study of certain disorders of the artery. Another area of interest is in the study of diffusion controlled heterogeneous reactions in tubular reactors. The majority of the heterogeneous reactions and especially those of industrial importance are diffusion controlled. If pulsations bring about a change in the rate of diffusion as has been postulated, then a change in the rate of the diffusion controlled reaction will result. Two physical models are discussed in this paper. The first refers to flow in a rigid conduit with a periodic disturbance superimposed on a steady mean flow. The second case applies to flow in a distensible tube which is longitudinally constrained, and the steady flow is disturbed by a pulse wave progressing in the direction of flow.

The Burnett method was used to determine the volumetric behavior of helium, nitrogen, and six evenly spaced helium-nitrogen mixtures at low temperature and high pressure. Compressibility isotherms were determined for both of the pure components and all of the mixtures at 0, -50, -90, -115, -130, and -140°C. Two series of expansions were made in order to accurately define each isotherm. The maximum pressure for the first series of expansions was maintained at 500 atm for each isotherm, and sufficient expansions were carried out to reduce the final pressure to 2--3 atm. The results were expressed in terms of the compressibility factors at the experimental pressures, the Leiden reciprocal volume series coefficients, and the compressibility factors at forty-three even values of pressure ranging from 1 to 500 atm. A total of approximately twelve hundred experimental compressibility factors were determined. From an error analysis developed specifically for the Burnett method, the maximum error in the experimental compressibility factors was estimated to be about one-tenth percent. The results of the experimental investigation were used to calculate the second virial coefficients, the interaction second virial coefficients, the third virial coefficients, the interaction third virial coefficients, the Lennard-Jones (12:6) potential parameters for nitrogen, the quantum mechanical Lennard-Jones (12:6) potential parameters for helium-nitrogen, and the fugacity coefficients for both of the pure components and all of the mixtures. The values of the second virial coefficients for the pure components were found to be in excellent agreement with similar values obtained from other methods by different investigators. Thus the Burnett method was established as an accurate way to determine second virial coefficients at low temperature. The existing data for gaseous mixtures were combined with the present results to study the mixing rules which have been proposed to predict the interaction parameters of the unlike species from known data for the pure components. The geometric mean and average rules were found to predict the correct values of the interaction second virial coefficients when the difference in the molecular size of the interacting species was small; however, for systems such as He-N2 and He-A, they predicted values which were fifteen to twenty percent too small. The mixing rule proposed by Srivastava and Madan was investigated where sufficient data were available over an extended range of temperature. In such cases one experimental value of the interaction second virial coefficient was used to supplement their rule. The predicted values of the interaction second virial coefficients were in good agreement with the experimental values for the six systems studied. An empirical rule was also developed for use with the Srivastava and Madan rule. Second interaction virial coefficients were predicted with these rules and compared with similar predictions based on the geometric mean and average rules for twelve systems. The average deviation between the experimental and predicted values was 5.1 cm3/mole for the empirical correlation and 8.8 cm3/mole for the geometric mean and average rules.

The purpose of this investigation was to study the effect of vapor velocity, temperature difference and fluid properties upon the heat transfer coefficient of a vapor condensing within a horizontal tube. Condensing heat transfer coefficients for propane were measured over a wide range of conditions. The data would not correlate using the existing equations for condensation. A new mechanism was postulated of a laminar condensate film, forming an annulus next to the tube wall, surrounding a turbulent vapor core. On this basis, the following equation was derived: hDkPr -13=B DGamL rLrV 12 n The data correlate to approximately 20% using this equation. The value of 'n' was 0.2 at low Reynolds numbers and approached a constant value of 0.8 at a Reynolds number of 105. The effect of the liquid loading and the temperature difference across the condensing film on the heat transfer coefficient was negligible over the range of the investigation.

Abstract Not Available.

The two phase system of water and kerosene in cocurrent stratified turbulent flow in a closed rectangular channel was studied. Profiles of mean velocity were measured in both phases by means of a hot-wire anemometer, for various combinations of flow rates of the two phases. The range of Reynolds numbers covered was from about 1900 to 4000 for the kerosene phase and from 2400 to 6200 for the water phase. This Reynolds number is defined as the mass flow rate per unit channel width divided by the fluid viscosity. The velocity profiles in each phase for all runs could be represented by an empirical equation of the form: U/U*=a'+b' h+c'h2+d' h2n+e'h2 n'+1. This form of the equation was derived from a combination of the expressions given by Pai for the two cases of turbulent Poiseuille and Couette flow between parallel plates. The parameters necessary to define the profiles by this expression could be correlated as functions of the water and kerosene Reynolds numbers. The wall friction coefficient as determined from the profiles with the aid of the above expression was also correlated as a function of the water and kerosene Reynolds numbers. The same correlation equation was found to be valid for both the kerosene and water walls merely by interchanging the Reynolds numbers in the equation. The interfacial shear coefficient, however, was found to be primarily a function of the Reynolds number of the upper (kerosene) phase. Furthermore, the function apparently depends upon the sign of the interfacial shear (i.e. which phase is moving the faster). A difference in shear across the interfacial region was found, which could be explained on an order of magnitude basis by a consideration of the rate of change of interfacial wave momentum.