In this thesis, a novel and efficient numerical method is presented for the computation of the L² optimal mass transport mapping in two and three dimensions. The method uses a direct variational approach. A new projection to the constraint technique has been formulated that can yield a good starting point for the method as well as a second order accurate discretization to the problem. The numerical experiments demonstrate that the algorithm yields accurate results in a relatively small number of iterations that are mesh independent. In the first part of the thesis, the theory and implementation details of the proposed method are presented. These include the reformulation of the Monge-Kantorovich problem using a variational approach and then using a consistent discretization in conjunction with the "discretize-then-optimize" approach to solve the resulting discrete system of differential equations. Advanced numerical methods such as multigrid and adaptive mesh refinement have been employed to solve the linear systems in practical time for even 3D applications. In the second part, the methods efficacy is shown via application to various image processing tasks. These include image registration and morphing. Application of (OMT) to registration is presented in the context of medical imaging and in particular image guided therapy where registration is used to align multiple data sets with each other and with the patient. It is shown that an elastic warping methodology based on the notion of mass transport is quite natural for several medical imaging applications where density can be a key measure of similarity between different data sets e.g. proton density based imagery provided by MR. An application is also presented of the two dimensional optimal mass transport algorithm to compute diffeomorphic correspondence maps between curves for geometric interpolation in an active contour based visual tracking application.

Der Bericht beschäftigt sich mit spontaner Grenzflächenkonvektion und -turbulenz beim Stoff- und Wärmeübergang an fluiden Phasengrenzen zwischen zwei nicht mischbaren Phasen. Solche Effekte sind von großer industrieller Bedeutung, da die erzielten Stoffübergangsraten um ein Vielfaches über den bei gewöhnlicher Diffusion auftretenden liegen. Zwei unterschiedliche Mechanismen sind der "Motor" für die Instabilitäten: Marangoni-Instabilität: Die Grenzflächenspannung ist eine Funktion der Temperatur und der Grenzflächenkonzentration des ausgetauschten Stoffes. Schwankungen der Temperatur und der Konzentration entlang der Phasengrenze führen folglich zu Grenzflächenspannungsgradienten. Grenzflächenspannungsgetriebene Instabilitäten äußern sich durch rollenförmige oder polygonale Konvektionszellen, Eruptionen oder Turbulenz an der Phasengrenze. Schwerkraftgetriebene Instabilität: Die Dichte ist ebenfalls eine Funktion der Temperatur und der Konzentration des gelösten Stoffes. Der Transport eines Stoffes über eine fluide Phasengrenze verändert die Zusammensetzung und die Dichte der angrenzenden Flüssigkeitsschichten, sodass instabile Dichteschichtungen auftreten können. Temperaturgradienten entstehen dabei durch Freisetzung von Reaktions- und/oder Lösungsenthalpie. Auftriebsbewegungen haben die Form von Thermiken (engl. plumes, thermals). Die Phänomene der Grenzflächenkonvektion werden in einer vertikalen Kapillarspaltgeometrie untersucht. Neben Stoffsystemen mit reaktivem Stoffübergang (Neutralisation von Karbonsäuren, Hydrolyse und Veresterung von Alkanoylhloriden) kamen auch solche mit reaktionsfreiem Stoffübergang (Karbonsäuren, Tensid) zur Anwendung. Die instabile Dichteschichtung, die durch den Konzentrationsgradienten infolge der Stoffdiffusion erzeugt wird, führt zu Auftriebskonvektion in Form von Thermiken. Die Anwesenheit einer exothermen Reaktion bewirkt eine Vergrößerung des Längenwachstums der Thermiken in der oberen Phase durch Aufprägung eines zusätzlich destabilisierenden Temperaturgradienten. In der unteren Phase kommt es dagegen zum Entstehen des doppeldiffusiven Fingerregimes bei Überlagerung des destabilisierenden Konzentrationsgradienten durch den stabilisierenden Temperaturgradienten. Beim Übergang eines Tensids konnten die für diese Stoffklasse charakteristischen Rollzellen, die durch Grenzflächenspannungsgradienten angetrieben werden, beobachtet werden. Diese Konvektionsstrukturen bleiben auf einen schmalen Bereich ober- und unterhalb der Phasengrenze beschränkt. Die Transportgleichungen für Impuls, Stoff und Wärme wurden in ihrer 2-dimensionalen Form in einen Rechenkode umgesetzt und der Übergang einer einzelnen Komponente simuliert. Die hydrodynamischen Bedingungen an der Phasengrenze wurden so formuliert, dass lokale Änderungen der Zusammensetzung und der Temperatur zu Grenzflächenspannungsgradienten führen und die Phasengrenze damit dem Marangonieffekt unterliegt. Die Stoffeigenschaften wurden mit Ausnahme der Dichte im Volumenkraftterm der Impulsgleichung als konstant angenommen, sodass dichtegetriebene Konvektionen simuliert werden können. Die verschiedenen Konvektionsformen werden durch die Simulation qualitativ gut wiedergegeben. Bei Marangonikonvektion kommt es zu einer Verschiebung des steilen Konzentrationsgradienten von der Phasengrenze in die Kerne der Phasen, was zum schnellen Absterben der Marangonikonvektion führt. Die Wiedergabe des Längenwachstums der Thermiken durch Simulation eines realen Stoffsystems ist zufriedenstellend. Ebenso gibt die Simulation eine realistische Abschätzung zu erwartender Stoffströme bei Anwesenheit hydrodynamischer Instabilitäten. Größere Abweichungen zwischen Simulation und Experiment sind jedoch bei der horizontalen Größenskala der Fingerstruktur festzustellen, die wahrscheinlich auf die Boussinesq-Approximation zurückzuführen sind.

Includes vita. / Title from PDF of title page. / Document formatted into pages; contains 185 pages. / Thesis (Ph.D.)--University of South Florida, 2003. / Includes bibliographical references. / Text (Electronic thesis) in PDF format. / ABSTRACT: Seagrass communities are composed of numerous organisms that depend on water-column nutrients for metabolic processes. The rate at which these organisms remove a nutrient from the water column can be controlled by physical factors such as hydrodynamic regime or by biological factors such as speed of enzyme reactions. The impact of hydrodynamic regime on rates of nutrient uptake for seagrass (Thalassia testudinum) communities and for organisms that comprise the community (seagrass, epiphytes, phytoplankton, and microphytobenthos) was quantified in a series of field flume experiments employing the use of 15N-labeled ammonium and nitrate. Rates of ammonium uptake for the entire community and for seagrass leaves and epiphytes were significantly dependent on bulk velocity, bottom shear stress, and the rate of turbulent energy dissipation. / ABSTRACT: Relationships between uptake rates and these parameters were consistent with mass-transfer theory and suggest that the effect of water flow on ammonium uptake is the same for the benthos as a whole and for the organisms that form the canopy. In addition, epiphytes on the surface of T. testudinum leaves were shown to depress leaf uptake by an amount proportional to the area of the leaf covered by epiphytes. Water flow influenced rates of nitrate uptake for the community and the epiphytes; however, uptake rates were depressed relative to those for ammonium suggesting that uptake of nitrate was also affected by biological factors such as enzyme activity. Epiphytes reduced uptake of nitrate by the leaves; however, the amount of reduction was not proportional to the extent of epiphyte cover, which provided further evidence that nitrate uptake by T. testudinum leaves was biologically limited. / ABSTRACT: As an additional component of the research, hydrodynamic regime of a mixed seagrass and coral community in Florida Bay was characterized using an acoustic Doppler velocimeter. Hydrodynamic parameters estimated from velocity data were used in mass-transfer equations to predict nutrient uptake by the benthos over a range of water velocity. Measured rates of uptake from field flume experiments conducted in the same community confirmed that hydrodynamic data could be used to accurately predict nutrient transport to the benthos under natural flow conditions. / System requirements: World Wide Web browser and PDF reader. / Mode of access: World Wide Web.

Thesis (doctoral)--Freie Universität Berlin, 2002. / "January 2003"--P. [2] of cover. Lebenslauf. Includes bibliographical references (p. 137-142). Also available via the World Wide Web.

This work evaluates stripper performance for CO₂ capture using seven potential solvent formulations and seven stripper configurations. Equilibrium and rate models were developed in Aspen Custom Modeler (ACM). The temperature approach on the hot side of the cross exchanger was varied between 5 - 10°C. The results show that operating the cross exchanger at a 5°C approach results in 12% energy savings for a 7m MEA rich solution of 0.563 mol/mol Alk and 90% CO₂ removal. For solvents with [Delta]H[subscript abs] < 60 kJ/gmol CO₂, stripping at 30 kPa is more attractive than stripping at 160 kPa. Normal pressure (160 kPa) favors solvents with high heats of desorption. The best solvent and process configuration, matrix with MDEA/PZ, offers 22% and 15% energy savings over the baseline and improved baseline, respectively, with stripping and compression to 10 MPa. The energy requirement for stripping and compression to 10 MPa is about 20 % of the power output from a 500 MW power plant with 90% CO2 removal. Rate model results show that a 'short and fat' stripper requires 7 to 15% less equivalent work than a 'tall and skinny' one. The optimum stripper design could be one that operates between 50% and 80% flood at the bottom. Stripping at 30 kPa and 160 kPa require 230 s and 115 s of effective packing volume to get an equivalent work 4% greater than the minimum. Stripping at 30 kPa with [Delta]T = 5°C was controlled by mass transfer with reaction in the boundary layer and diffusion (88% resistance at the rich end and 71% resistance at the lean end) and mass transfer with equilibrium reactions (84% resistance at the rich end and 74% resistance at the lean end) at 160 kPa. The model was validated with data obtained from pilot plant experiments at the University of Texas with 5m K⁺/2.5m PZ and 6.4m K⁺/1.6m PZ under normal pressure and vacuum conditions using Flexipac AQ Style 20 structured packing. Foaming was experienced during tests. The effective packing height was 5.09m for 5m K⁺/2.5m PZ and 6.47m for 6.4m K⁺/1.6m PZ. / text

The overarching goal of this work is to address the design of next-generation, high torque density electrical machines through numerical optimization using an integrated thermal-electromagnetic design tool that accounts for advanced cooling technology. A parametric thermal model of electric machines was constructed and implemented using a finite difference approach incorporating an automated, self segmenting mesh generation. A novel advanced cooling technology is proposed to improve thermal transport in the machine by removing heat directly from the windings via heat exchangers located between the winding bundles. Direct winding heat exchange (DWHX) requires high convective transport and low pressure loss. The heat transfer to pressure drop tradeoff was addressed by developing empirically derived Nusselt number and friction factor correlations for micro-hydrofoil enhanced meso-channels. The parametric thermal model, advanced cooling technique, Nusselt number and friction factor correlations were combined with a parametric electromagnetic model for electric machines. The integrated thermal-electromagnetic model was then used in conjunction with particle swarm optimization to determine optimal conceptual designs. The Nusselt number correlation achieves an R² value of 0.99 with 95% of the data falling within ± 2.5% similarly the friction factor correlation achieves an R² value of 0.92 with 95% of the data falling within ± 10.2%. The integrated thermal-electromagnetic design tool, incorporating DWHX, generated an optimized 20 kW permanent magnet electric machine design achieving a torque density of 23.2 N-m/L based on total system volume.

This thesis examined the biological treatment of gas streams containing benzene, toluene, ethylbenzene and o-xylene (BTEX) using solid-liquid two-phase partitioning bioscrubbers (SL-TPPBs). SL-TPPBs consist of a cell containing aqueous phase and a polymeric solid phase that sequesters poorly water soluble and/or toxic substrates, mitigating substrate toxicity in the aqueous phase and improving the gas mass transfer during treatment of VOC contaminated gases. An initial investigation of oxygen transport determined that the polymers in a stirred-tank SL-TPPB enhance gas-liquid mass transfer. In addition, a study on biodegradation kinetics of BTEX by a bacterial consortium identified and quantified substrate interactions such as inhibition, enhancement and cometabolism. The stirred-tank SL-TPPB was then experimentally investigated for treatment of BTEX gas streams during steady-state and dynamic step-change operation to determine performance of the system relative to other biotreatment methods. A mathematical model was developed to predict system performance, which included the microbial kinetic model structure and parameters estimated during kinetic and oxygen mass transfer studies. As a less energy intensive alternative, an airlift SL-TPPB was operated and characterized. The airlift SL-TPPB was compared to an airlift liquid-liquid TPPB (silicone oil as sequestering phase) and a single phase airlift over dynamic step-change loadings, which showed that the airlift SL-TPPB outperformed the single phase airlift by >30% and had similar performance to the liquid-liquid airlift. However, the airlift SL-TPPB performance was lower relative to the stirred-tank SL-TPPB by >15%. Steady-state operation of the airlift SL-TPPB identified a range of operating conditions that provided maximum performance and conditions that were not oxygen limited. This prompted a study of oxygen mass transfer and hydrodynamics in the airlift system, which identified that the addition of polymers to an airlift does not cause physical enhancement of the gas-liquid mass transfer coefficient, but improves aqueous phase mixing and enhances overall oxygen transfer rate. A tanks-in-series mathematical model was formulated to predict performance of the airlift SL-TPPB, wherein the number of tanks-in-series to describe mixing in the airlift was obtained from a residence time distribution analysis of the airlift system completed during the hydrodynamic investigation. This thesis contributes a low-energy solution for the effective treatment of gases contaminated with BTEX. / Thesis (Ph.D, Chemical Engineering) -- Queen's University, 2009-08-18 16:16:22.598

No description available.

Worldwide demand for energy is increasing rapidly, partly driven by dramatic economic growth in developing countries. This growth has sparked concerns over the finite availability of fossil fuels and the impact of their combustion on climate change. Consequently, many recent research efforts have been devoted to the development of renewable fuels and sustainable energy systems. Interest in liquid biofuels, such as ethanol, has been particularly high because these fuels fit into the conventional infrastructure for the transportation sector. Ethanol is a renewable fuel produced through the anaerobic fermentation of sugars obtained from biomass. However, the relatively high energy demand of its production process is a major factor limiting the usefulness of ethanol as a fuel. Due to the dilute nature of the fermentation product stream and the presence of the ethanol-water azeotrope, the separation processes currently used to recover anhydrous ethanol are particularly inefficient. In fact, the ethanol separation processes account for a large fraction of the total process energy demand. In the conventional ethanol separation process, ethanol is recovered using several distillation steps combined with a dehydration process. In this dissertation, a new hybrid pervaporation-distillation system, named Membrane Dephlegmation, was proposed and investigated for use in ethanol recovery. In this process, countercurrent vapour-liquid contacting is carried out on the surface of a pervaporation membrane, leading to a combination of distillation and pervaporation effects. It was intended that this new process would lead to improved economics and energy efficiency for the entire ethanol production process. The Membrane Dephlegmation process was investigated using both numerical and experimental techniques. Multiphase Computational Fluid Dynamics (CFD) was used to study vapour-liquid contacting behaviour in narrow channels and to estimate heat and mass transfer rates. Results from the CFD studies were incorporated into a simplified design model and the Membrane Dephlegmation process was studied numerically. The results indicated that the Membrane Dephlegmation process was more efficient than simple distillation and that the ethanol-water azeotrope could be broken. Subsequently, a pilot-scale experimental system was constructed using commercially available, hydrophilic NaA zeolite membranes. Results obtained from the experimental system confirmed the accuracy of the simulations.

The purpose of this research was to determine the solubility of apricot (Prunus armeniaca L.) oil in supercritical carbon dioxide (SC-CO2), effects of parameters (particle size, solvent flow rate, pressure, temperature and co-solvent (ethanol) concentration) on extraction yield and to investigate the possibility of fractionation. Solubility, increased with pressure and increased with temperature above the crossover pressure, which was found between 200 and 300 bar, and decreased with temperature below the crossover pressure. Appropriate models were fitted to data. Extraction of apricot kernel oil occurred in two extraction periods as fast and slow extraction periods. Most of the oil was extracted in the fast extraction period and the oil recovered in the slow extraction period was negligible. Extraction yield increased with decrease in particle size and recovery of more than 99 % of the oil was possible if particle diameter decreased below 0.425 mm. Extraction rate increased with increase in flow rate, pressure, temperature and ethanol concentration. The volume mass transfer coefficient in the fluid phase changed between 0.6 and 3.7 /min, whereas the volume mass transfer coefficient in the solid phase changed between 0.00009 and 0.00048 /min. Extraction yield at 15 min for particle diameter smaller than 0.85 mm was formulated as a function of solvent flow rate, pressure, temperature, and ethanol concentration by using Response Surface Methodology. According to the model yield was highest (0.26 g /g) at 4 g/min flow rate, 60 oC, 450 bar and 3 % ethanol concentration. Fractionation was not possible at significant levels.