<p>Polymer electrolyte fuel cells (PEFC) convert chemicalenergy into electrical energy with higher efficiency thaninternal combustion engines. They are particularly suited fortransportation applications or portable devices owing to theirhigh power density and low operating temperature. The latter ishowever detrimental to the kinetics of electrochemicalreactions and in particular to the reduction of oxygen at thecathode. The latter reaction requires enhancing by the verybest catalyst, today platinum. Even so, the cathode isresponsible for the main loss of voltage in the cell. Moreover,the scarce and expensive nature of platinum craves theoptimisation of its use.</p><p>The purpose of this thesis was to better understand thefunctioning of the porous cathode in the PEFC. This wasachieved by developing physical models to predict the responseof the cathode to steady-state polarisation, currentinterruption (CI) and electrochemical impedance spectroscopy(EIS), and by comparing these results to experimental ones. Themodels account for the kinetics of the oxygen reduction as wellas for the transport of the reactants throughout the cathode,i.e. diffusion of gases and proton migration. The agglomeratestructure was assumed for the description of the internalstructure of the cathode. The electrochemical experiments wereperformed on electrodes having a surface of 0.5 cm2 using alaboratory fuel cell.</p><p>The response of the cathode to various electrodecompositions, thickness, oxygen pressure and relative humiditywas experimentally investigated with steady-state polarisation,EIS and CI techniques. It is shown that a content in thecathode of 35-43 wt % of Nafion, the polymer electrolyte, gavethe best performance. Such cathodes display a doubling of theapparent Tafel slope at high current density. In this region,the current is proportional to the cathode thickness and to theoxygen pressure, which, according to the agglomerate model,corresponds to limitation by oxygen diffusion in theagglomerates. The same analysis was made using EIS. Moreover,experimental results showed that the Tafel slope increases fordecreasing relative humidity. For Nafion contents lower than 35wt %, the cathode becomes limited by proton migration too. ForNafion contents larger than 40 wt %, the cathode performance athigh current density decreases again owing to an additionalmass transport. The latter is believed to be oxygen diffusionthroughout the cathode. The activity for oxygen reduction ofcatalysts based on iron acetate adsorbed on a carbon powder andpyrolysed at 900°C in ammonia atmosphere was alsoinvestigated. It was shown that the choice of carbon has atremendous effect. The best catalysts were, on a weight basis,as active as platinum.</p><p><b>Keywords:</b>polymer electrolyte fuel cell, cathode, masstransport, porous electrode, modelling, agglomerate model,electrochemical impedance spectroscopy, current interrupt,transient techniques, non-noble catalysts</p>

Oxygen gain is the difference in fuel cell performance operating on oxygen-depleted and oxygen-rich cathode fuel streams. Oxygen gain experiments provide insight into the degree of oxygen mass-transport resistance within a fuel cell. By taking these measurements under different operating conditions, or over time, one can determine how oxygen mass transport varies with operating modes and/or aging. This paper provides techniques to differentiate between mass-transport resistance within the catalyst layer and within the gas-diffusion medium for a polymer-electrolyte membrane fuel cell. Two extreme cases are treated in which all mass transfer limitations are located only (i) within the catalyst layer or (ii) outside the catalyst layer in the gas diffusion medium. These two limiting cases are treated using a relatively simple model of the cathode potential and common oxygen gain experimental techniques. This analysis demonstrates decisively different oxygen gain behavior for the two limiting cases. For catalyst layer mass transfer resistance alone, oxygen gain values are limited to a finite range of values. However, for gas diffusion layer mass transfer resistance alone, the oxygen gain is not confined to a finite range of values. This analysis is then extended to evaluate ionic effects within the catalyst layer. / text

Submarine landslides retrogressively fail from intact material at the headwall and then become fluidized by strain weakening; the final deposits of these flows have low porosity, which controls their character in seismic reflection data. Submarine landslides occur on the open slope and also localized areas including margins of turbidite channel-levee systems. I develop and quantify this model with 3-D seismic reflection data, core and log data from Integrated Ocean Drilling Program Expedition 308 (Ursa Basin, Gulf of Mexico), flume experiments, and numerical modeling. At Ursa, multiple submarine slides over the last 60 ky are preserved as mass transport deposits (MTDs). Retrogression proceeded from an initial slope failure that created an excavated headwall, which reduced the horizontal stress behind the headwall and resulted in normal faults. Fault blocks progressively weakened until the gravitational driving stress imposed by the bed slope exceeded soil strength, which allowed the soil to flow for more than 10 km away from the source area. The resulting MTDs have lower porosity (higher bulk density) relative to non-failed sediments, which ultimately produces high amplitude reflections at the base and top of MTDs. In the laboratory, I made weak (low yield strength) and strong flows (high yield strength) from mixtures of clay, silt, and water. Weak flows generate turbidity currents while moving rapidly away from the source area. They create thin and long deposits with sinuous flow features, and leave behind a relatively smooth and featureless source area. In contrast, strong flows move slowly, do not generate a turbidity current, and create blocky, highly fractured source areas and short, thick depositional lobes. In Pleistocene turbidite channels of the Mississippi Fan, deep-seated rotational failures occurred in the flanking levees. The rotational failures displaced material into the channel from below where it became eroded by turbidity flows. This system achieved a delicate steady state where levee deposition and displacement along the fault into the channel was balanced by erosion rate of turbidity flows. This work enhances our understanding of geohazards and margin evolution by illuminating coupled processes of sedimentation, fluid flow, and deformation on passive continental margins. / text

No description available.

This thesis explores ideas from transport theory and optimal control to develop novel Monte Carlo methods to perform efficient statistical computation. The first project considers the problem of constructing a transport map between two given probability measures. In the Bayesian formalism, this approach is natural when one introduces a curve of probability measures connecting the prior to posterior by tempering the likelihood function. The main idea is to move samples from the prior using an ordinary differential equation (ODE), constructed by solving the Liouville partial differential equation (PDE) which governs the time evolution of measures along the curve. In this work, we first study the regularity solutions of Liouville equation should satisfy to guarantee validity of this construction. We place an emphasis on understanding these issues as it explains the difficulties associated with solutions that have been previously reported. After ensuring that the flow transport problem is well-defined, we give a constructive solution. However, this result is only formal as the representation is given in terms of integrals which are intractable. For computational tractability, we proposed a novel approximation of the PDE which yields an ODE whose drift depends on the full conditional distributions of the intermediate distributions. Even when the ODE is time-discretized and the full conditional distributions are approximated numerically, the resulting distribution of mapped samples can be evaluated and used as a proposal within Markov chain Monte Carlo and sequential Monte Carlo (SMC) schemes. We then illustrate experimentally that the resulting algorithm can outperform state-of-the-art SMC methods at a fixed computational complexity. The second project aims to exploit ideas from optimal control to design more efficient SMC methods. The key idea is to control the proposal distribution induced by a time-discretized Langevin dynamics so as to minimize the Kullback-Leibler divergence of the extended target distribution from the proposal. The optimal value functions of the resulting optimal control problem can then be approximated using algorithms developed in the approximate dynamic programming (ADP) literature. We introduce a novel iterative scheme to perform ADP, provide a theoretical analysis of the proposed algorithm and demonstrate that the latter can provide significant gains over state-of-the-art methods at a fixed computational complexity.

Devido à proximidade de grandes cidades, ecossistemas aquáticos são alvos de despejos industriais, esgotos domésticos e resíduos sólidos. Modelos matemáticos são largamente utilizados para entender os padrões de fluxo e o transporte de substâncias nestes ecossistemas. Neste trabalho, desenvolveu-se um modelo hidrodinâmico e de transporte de poluentes bidimensional para corpos d’água rasos que utiliza grades não estruturadas. O modelo foi desenvolvido em dois módulos: hidrodinâmico e de transporte de massa. O módulo hidrodinâmico utiliza as equações completas de Saint Venant para quantificação da circulação em corpos d’águas. No módulo hidrodinâmico aqui desenvolvido é feito um aperfeiçoamento de um esquema semi-implícito de diferenças finitas utilizado no modelo IPH-ECO para um esquema semi-implícito de volumes finitos. O módulo de transporte de massa, por sua vez, representa o transporte de um escalar conservativo sob influência dos processos físicos do meio. Dois esquemas foram empregados e comparados para a solução numérica da equação de transporte: um esquema de diferenças centrais e um esquema com limitador de fluxo. O modelo foi aplicado em um estudo de caso simplificado formado por uma lagoa circular e no rio Guaíba. O módulo hidrodinâmico foi avaliado através de testes de conservação de volume feitos comparando com os resultados encontrados nno modelo IPHECO na aplicação realizada na lagoa circular e na solução analítica no rio Guaíba. O módulo hidrodinâmico do modelo de grades não estruturadas (IPH-UnTRIM2D) mostrou-se mais conservativo que o modelo IPH-ECO em todos os cenários simulados na lagoa circular. A solução hidrodinâmica e de transporte do modelo de grades não estruturadas no estudo de caso idealizado no rio Guaíba representou meandros, sulcos, estreitamentos e alargamentos de canais com comprometimento com uma situação real. Quanto à simulação de transporte de escalares e poluentes, foi comprovada a eficiência do esquema com limitador de fluxo sobre os esquemas de diferenças centrais. Os testes de conservação de massa para diferentes cenários no rio Guaíba indicaram erros pouco significativos em ambos os esquemas quando comparados com a ordem de grandeza do ecossistema. / Aquatics ecossistems around large towns are always exposed waste domestic and industrial and solid residues. Mathematical models are widely used to understand flow patterns and transport of mass in these ecosystems. In this work a model was developed on unstructured two-dimensional complex dynamic model for shallow lakes. The model was divided in two modules: (a) hydrodynamic module, describing quantitative flows and water level; (b) a mass transport module. The hydrodynamic module solves Saint Venant equations to quantify water circulation in a water body. Futhermore, the hydrodynamic module was improved to semiimplicit finite volume scheme. The mass transport module represents transport of a conservative scalar under the physical processes influence. Two schemes were used and compared to the transport equation numerical solution : a central difference scheme and a flux limiter scheme. The model was applied in the circular lake and river Guaíba. The hydrodynamic module was validated by evaluating tests of volume conservation made comparing with results of the model IPH-ECO and analytical solution. It proved to be more conservative than the model IPH-ECO in all scenarios simulated in the circular lake. The hydrodynamic and transport solution in river Guaíba study case represented bends, grooves, and channels widening and narrowing. In the transport simulation, the flux limiter scheme was more efficient than central differences scheme. Futhermore, mass conservation tests for different scenarios in Guaíba indicated minor errors in both schemes when compared with ecosystem magnitude.

Devido à proximidade de grandes cidades, ecossistemas aquáticos são alvos de despejos industriais, esgotos domésticos e resíduos sólidos. Modelos matemáticos são largamente utilizados para entender os padrões de fluxo e o transporte de substâncias nestes ecossistemas. Neste trabalho, desenvolveu-se um modelo hidrodinâmico e de transporte de poluentes bidimensional para corpos d’água rasos que utiliza grades não estruturadas. O modelo foi desenvolvido em dois módulos: hidrodinâmico e de transporte de massa. O módulo hidrodinâmico utiliza as equações completas de Saint Venant para quantificação da circulação em corpos d’águas. No módulo hidrodinâmico aqui desenvolvido é feito um aperfeiçoamento de um esquema semi-implícito de diferenças finitas utilizado no modelo IPH-ECO para um esquema semi-implícito de volumes finitos. O módulo de transporte de massa, por sua vez, representa o transporte de um escalar conservativo sob influência dos processos físicos do meio. Dois esquemas foram empregados e comparados para a solução numérica da equação de transporte: um esquema de diferenças centrais e um esquema com limitador de fluxo. O modelo foi aplicado em um estudo de caso simplificado formado por uma lagoa circular e no rio Guaíba. O módulo hidrodinâmico foi avaliado através de testes de conservação de volume feitos comparando com os resultados encontrados nno modelo IPHECO na aplicação realizada na lagoa circular e na solução analítica no rio Guaíba. O módulo hidrodinâmico do modelo de grades não estruturadas (IPH-UnTRIM2D) mostrou-se mais conservativo que o modelo IPH-ECO em todos os cenários simulados na lagoa circular. A solução hidrodinâmica e de transporte do modelo de grades não estruturadas no estudo de caso idealizado no rio Guaíba representou meandros, sulcos, estreitamentos e alargamentos de canais com comprometimento com uma situação real. Quanto à simulação de transporte de escalares e poluentes, foi comprovada a eficiência do esquema com limitador de fluxo sobre os esquemas de diferenças centrais. Os testes de conservação de massa para diferentes cenários no rio Guaíba indicaram erros pouco significativos em ambos os esquemas quando comparados com a ordem de grandeza do ecossistema. / Aquatics ecossistems around large towns are always exposed waste domestic and industrial and solid residues. Mathematical models are widely used to understand flow patterns and transport of mass in these ecosystems. In this work a model was developed on unstructured two-dimensional complex dynamic model for shallow lakes. The model was divided in two modules: (a) hydrodynamic module, describing quantitative flows and water level; (b) a mass transport module. The hydrodynamic module solves Saint Venant equations to quantify water circulation in a water body. Futhermore, the hydrodynamic module was improved to semiimplicit finite volume scheme. The mass transport module represents transport of a conservative scalar under the physical processes influence. Two schemes were used and compared to the transport equation numerical solution : a central difference scheme and a flux limiter scheme. The model was applied in the circular lake and river Guaíba. The hydrodynamic module was validated by evaluating tests of volume conservation made comparing with results of the model IPH-ECO and analytical solution. It proved to be more conservative than the model IPH-ECO in all scenarios simulated in the circular lake. The hydrodynamic and transport solution in river Guaíba study case represented bends, grooves, and channels widening and narrowing. In the transport simulation, the flux limiter scheme was more efficient than central differences scheme. Futhermore, mass conservation tests for different scenarios in Guaíba indicated minor errors in both schemes when compared with ecosystem magnitude.

Com base em um modelo fenomenológico e a partir de dados experimentais obtidos numa planta piloto, foi obtida uma correlação entre o desempenho de uma torre de resfriamento em função das principais variáveis de processo: fluxos mássicos do gás e da água pela torre, e temperatura de entrada da água. Os resultados apresentaram boa consistência, comparados com os da literatura. A metodologia desenvolvida pode, com relativa facilidade, ser aplicada para torres de resfriamento industriais, pois se baseia em medidas de variáveis, factíveis em termos práticos. Efetuou-se também um estudo paralelo com base em modelagem e simulações matemáticas do comportamento de uma torre de resfriamento de água em condições severas, com temperatura da água de alimentação superior a 50°C. Constatou-se que o coeficiente de transporte de massa na torre de resfriamento aparentemente não é afetado. / Cooling towers are widely used in many industrial and utility plants and its thermal performance is of vital importance. In the present work, using a phenomenological model and by experiments carried on over a pilot installation, the mass transfer coefficient dependence of air and water flow rates and inlet cooling water temperature is determined. The approach proposed may be useful in addition for characterization of industrial cooling towers since it depends on temperature and flow rate measurement usually available in typical plants. A parallel study concerning high mass transfer rate theory is accomplished. Through mathematical modeling and simulations based on this study no influence is detected on the mass transfer coefficient in the cooling tower, operating under harsh conditions with inlet water temperature up to 90°C.

Devido à proximidade de grandes cidades, ecossistemas aquáticos são alvos de despejos industriais, esgotos domésticos e resíduos sólidos. Modelos matemáticos são largamente utilizados para entender os padrões de fluxo e o transporte de substâncias nestes ecossistemas. Neste trabalho, desenvolveu-se um modelo hidrodinâmico e de transporte de poluentes bidimensional para corpos d’água rasos que utiliza grades não estruturadas. O modelo foi desenvolvido em dois módulos: hidrodinâmico e de transporte de massa. O módulo hidrodinâmico utiliza as equações completas de Saint Venant para quantificação da circulação em corpos d’águas. No módulo hidrodinâmico aqui desenvolvido é feito um aperfeiçoamento de um esquema semi-implícito de diferenças finitas utilizado no modelo IPH-ECO para um esquema semi-implícito de volumes finitos. O módulo de transporte de massa, por sua vez, representa o transporte de um escalar conservativo sob influência dos processos físicos do meio. Dois esquemas foram empregados e comparados para a solução numérica da equação de transporte: um esquema de diferenças centrais e um esquema com limitador de fluxo. O modelo foi aplicado em um estudo de caso simplificado formado por uma lagoa circular e no rio Guaíba. O módulo hidrodinâmico foi avaliado através de testes de conservação de volume feitos comparando com os resultados encontrados nno modelo IPHECO na aplicação realizada na lagoa circular e na solução analítica no rio Guaíba. O módulo hidrodinâmico do modelo de grades não estruturadas (IPH-UnTRIM2D) mostrou-se mais conservativo que o modelo IPH-ECO em todos os cenários simulados na lagoa circular. A solução hidrodinâmica e de transporte do modelo de grades não estruturadas no estudo de caso idealizado no rio Guaíba representou meandros, sulcos, estreitamentos e alargamentos de canais com comprometimento com uma situação real. Quanto à simulação de transporte de escalares e poluentes, foi comprovada a eficiência do esquema com limitador de fluxo sobre os esquemas de diferenças centrais. Os testes de conservação de massa para diferentes cenários no rio Guaíba indicaram erros pouco significativos em ambos os esquemas quando comparados com a ordem de grandeza do ecossistema. / Aquatics ecossistems around large towns are always exposed waste domestic and industrial and solid residues. Mathematical models are widely used to understand flow patterns and transport of mass in these ecosystems. In this work a model was developed on unstructured two-dimensional complex dynamic model for shallow lakes. The model was divided in two modules: (a) hydrodynamic module, describing quantitative flows and water level; (b) a mass transport module. The hydrodynamic module solves Saint Venant equations to quantify water circulation in a water body. Futhermore, the hydrodynamic module was improved to semiimplicit finite volume scheme. The mass transport module represents transport of a conservative scalar under the physical processes influence. Two schemes were used and compared to the transport equation numerical solution : a central difference scheme and a flux limiter scheme. The model was applied in the circular lake and river Guaíba. The hydrodynamic module was validated by evaluating tests of volume conservation made comparing with results of the model IPH-ECO and analytical solution. It proved to be more conservative than the model IPH-ECO in all scenarios simulated in the circular lake. The hydrodynamic and transport solution in river Guaíba study case represented bends, grooves, and channels widening and narrowing. In the transport simulation, the flux limiter scheme was more efficient than central differences scheme. Futhermore, mass conservation tests for different scenarios in Guaíba indicated minor errors in both schemes when compared with ecosystem magnitude.

Orientador: Jose Tomaz Vieira Pereira / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-07-21T10:26:59Z (GMT). No. of bitstreams: 1 Martins_Gilberto_D.pdf: 7571642 bytes, checksum: acd686af9af32ba4282365b2381bcb66 (MD5) Previous issue date: 1996 / Resumo: Este trabalho apresenta uma simulação do escoamento, transferência de calor e massa da mistura gasosa amônia-hidrogênio em convecção natural em um sistema de refrigeração por absorção com gás inerte. Os fenômenos de transporte envolvidos em cada um dos equipamentos do circuito foram descritos através das equações de conservação de massa, quantidade de movimento, energia e espécies químicas.A utilização das simplificações da camada limite tornou possível a resolução acoplada dessas equações através do método das diferenças finitas marchando na direção do fluxo. São apresentadas soluções para cada um dos principais componentes do circuito, o evaporador e absorvedor, para diversas condições de contorno, de operação e configurações geométricas. Finalmente foi proposta uma configuração para o circuito de circulação de gás e, para esta configuração, resolveram-se os sistemas de equações para cada um dos equipamentos do circuito acoplados. Os resultados mostram que os fatores geométricos são os que mais influenciam o comportamento do sistema. O modelo de simulação desenvolvido possibilita o projeto e a otimização do circuito de circulação de gases de novos sistemas sob diversas condições de operação / Abstract: This work presents a simulation of the flow, heat and mass transfer of a gaseous ammonia hydrogen mixture in natural convection inside an absorption reftigeration system with inert gas. The transport phenomena involved in each equipment ofthe circuit were described by the conservation equations of mass, momentum, energyand chemicalspecies. By using the boundary layer simplifications, it was possible to use the finite-differences method marching downstream to solve the coupled equations. Solutionsfor the two main components ofthe system, the evaporator and the absorber, are presented for various boundary and operating conditions and geometric configurations. A geometric configuration for the whole gas circulation system is then proposed and, for this configuration, the solution ofthe equations for each component ofthe system is performed in a coupled way. The results obtained show that the geometric configuration is the most influencing parameter in the system behavior. The simulation model presented in this work can be used for the design and optimization of the gas circulation circuit of new reftigeration system sunder various operating conditions / Doutorado / Termica e Fluidos / Doutor em Engenharia Mecânica