Lorsqu'un phénomène de détonation survient dans des milieux confinés comme un bâtiment présentant un système de galeries, l'onde de choc générée se propage et peut engendrer des dégâts matériels et humains considérables. En effet, l’onde de choc en se propageant dans ces systèmes de canaux confinés va interagir avec les obstacles qu’elle rencontre (humains et matériels), et va provoquer de très fortes élévations de la pression localement. Ce phénomène peut être très destructeur pour les structures et létale pour les humains. L'objectif de cette étude, expérimentale et numérique, est donc d'étudier la propagation d'une onde de choc dans différents systèmes de canaux afin de comprendre son comportement et de proposer des solutions pour en atténuer les dégâts. L’étude s’articule autour de deux grands axes. Le premier étudie de manière très académique la propagation d’une onde de choc au travers de cinq configurations différentes ; deux sans variation de section (un coude à 45$^{\circ}$ et un coude à 90$^{\circ}$), et trois avec variation de section (un divergent, un élargissement brusque et une bifurcation en « Y »). L’influence de la singularité sur la pression réfléchie en fond de configuration est étudiée ainsi que les mécanismes physiques complexes qui se produisent derrière l’onde de choc le long de la singularité. Le deuxième axe étudie l’atténuation de la pression réfléchie par l’ajout de piège (cavités) le long du système étudié. Une étude paramétrique montre que la position ainsi que la taille et la forme du piège ont un impact sur l’atténuation de l’onde de choc. / In the search for protection from explosions phenomena, a variety of underground shelters were studied to minimize the risks related to the propagation of shock waves in closed areas. Indeed, the blast effect could be really destructive for equipment and humans encountering the shock wave propagation, because of the high elevation of the local pressure it generates. Thus, the propagation of a shock wave through various canal systems is both experimentally and numerically studied. This study is based on two topics. The first part is focused on the study of the propagation of a shock wave through five various configurations ; two whithout aera change (a 45$^{\circ}$ bend duct and a 90$^{\circ}$ bend), and three configurations whith a sectional enlargement (a divergent, an abrupt area change and a « Y » bifurcation). The impact of the geometry on the end wall reflected pressure is studied as well as some complex physical mechanisms that occur behind the incident shock wave. The second part explores the mitigation of the pressure level in the device with the addition of traps (cavities) along the studied configuration. A parametric study based on the shape and size of the cavities, shows that these parameters have a considerable impact on the pressure level in the duct system.

Os mecanismos de instabilidade hidrodinâmica têm um papel importante no processo da transição do escoamento de laminar para turbulento. A análise da instabilidade hidrodinâmica em uma cavidade com tampa deslizante foi realizada através da decomposição em modos globais (biglobal) para avaliar o efeito da compressibilidade neste fenômeno. O escoamento base foi obtido através de simulação numérica direta (DNS). Para tal, foi desenvolvido um código DNS compressível com discretização espacial por diferenças finitas compactas de alta resolução espectral e capacidade de processamento paralelo, com um método de decomposição de domínio que mantém a precisão das diferenças finitas compactas. O escoamento base é usado para montar o problema de autovalor oriundo das equações de Navier-Stokes linearizadas para a perturbação, discretizadas por diferenças finitas explícitas. O uso de diferenças finitas em conjunto com a implementação em matrizes esparsas reduz sensivelmente o uso de memória. Através do algoritmo de Arnoldi, a ordem do problema de autovalor é reduzida e os autovalores de interesse são recuperados. Os resultados indicam o efeito estabilizante da compressibilidade nos modos dominantes da cavidade e revelam modos inerentes ao escoamento compressível, para os quais a compressibilidade tem efeito desestabilizante. Dentre estes modos compressíveis, estão presentes modos de propagação sonora em dutos e modos relacionados à geração de som na cavidade. / Hydrodynamic instability mechanisms play an important role in laminar to turbulent transition. Hydrodynamic instability analysis of a lid-driven cavity flow was performed by global mode decomposition (biglobal) to evaluate compressibility effects on this phenomenon. The basic flow was calculated by direct numerical simulation (DNS). A compressible DNS code was developed with spectral-like compact finite difference spatial discretization. The code allows parallel processing with a domain decomposition method that preserves the compact finite difference accuracy. The basic flow is used to form the eigenvalue problem associated to the linear Navier- Stokes equations for the perturbation, which were discretized by an explicit finite difference scheme. The combination of sparse matrix techniques and finite difference discretization leads to a significant memory reduction. The order of the eigenvalue problem was reduced using the Arnoldi algorithm and the eigenvalues of interest were calculated. Results show the stabilizing effect of compressibility on the leading modes and reveal some modes intrinsic to compressible flow, for which compressibility has a destabilizing effect. Among these compressible modes, there are some related to sound propagation in ducts and to sound generation inside the cavity.

O presente trabalho tem por objetivo o estudo da otimização multiobjetivo aplicada ao projeto de perfis aerodinâmicos em regime transônico, analisando comparativamente diferentes formas de definir as funções objetivo. A otimização é efetuada pelo algoritmo genético NSGA-II. Os resultados são avaliados utilizando métricas de diversidade da população e otimalidade das soluções, das quais duas são propostas. As funções objetivo são constituidas de diferentes parametrizações da geometria e diferentes técnicas de simulação numérica. A parametrização da geometria é feita utilizando a paramentrização Parsec ou a parametrização baseada em pontos de controle. A discretização do domínio espacial é feita utilizando malha estruturada conformada ao perfil e suavização por EDP elíptica. As duas técnicas de volumes finitos com diferentes modelos para o cálculo do fluxo na face do volume implementadas foram o método de Jameson (esquema centrado) e o método de Roe (esquema upwind). As comparações feitas são as seguintes: utilização de modelo viscoso e invíscido, com o uso do código Mses com a parametrização por ponto de controle; a utilização da parametrização por ponto de controle e parametrização Parsec usando o método de Jameson; e a comparação entre o método centrado e o upwind, utilizando a parametrização Parsec. Conclui-se dos resultados obtidos que a utilização da parametrização por pontos de controle é melhor. Entretanto, ainda é necessária a utilização de uma parametrização que garanta maior suavidade ou a imposição de restrições sobre a suavidade da solução. A utilização do modelo viscoso torna os resultados da otimização melhores do ponto de vista da otimalidade. Na utilização de modelos de correção viscosa, como no caso do Mses, é necessária a utilização de métodos invíscidos que forneçam resultados com maior representatividade física / The objective of present study is analyze the multi-objective optimization applied to transonic airfoils project comparing different ways to define the objective functions. The optimization is evaluated by the genetic algorithm NSGA-II. The results is analyzed using metrics of diversity and optimality for multi-objective problems, which two are proposed. The objective functions are defined by different parametrizations of geometry and different techniques of numerical simulation. The geometry parametrization was made by two distinct forms: using Parsec parametrization; and the control points based parametrization. The space domain discretization was made using structured body-fitted mesh with elliptical PDE smooth. A finite volume code with two different techniques for calculations of flux interface had been implemented: the Jamesons method (centered); and the Roes method ( it upwind). For viscous model usage analysis was used the Mses code that has implemented a finite volumes technique with viscous model correction. The following comparisons has been made: viscous and inviscid model using the Mses code with the control points parametrization; the control points and Parsec parametrizations using the Jamesons method; and the comparison among the centered method and upwind using the parametrization Parsec. From the results, it is concluded that the used of control points parametrization is interesting. Although, is still needed the used of a parametrization that guarantees a better smoothness or the imposes of a geometrical or property distribution restriction. The uses of viscous model gives better optimizations results in optimality requirement. It is needed the uses of inviscid method that forces better physical representation when using viscous correction model

O comportamento do escoamento supersônico no interior de bocais convergente-divergente retangulares é investigado numericamente, comparando-se quatro bocais com diferentes seções divergentes, com a mesma razão de aspecto AR=1.14 e mesma relação áreas da saída e da garganta dos bocais NAR=1.43. Os bocais são submetidos a diferentes pressões de admissão do fluido de trabalho, mantendo-se a relação entre a pressão de admissão e de descarga constante NPR=5. As simulações consideram o escoamento em regime permanente, compressível, viscoso, com abordagem baseada na massa específica (abordagem acoplada) , juntamente com o modelo de turbulência − /SST. A qualidade dos resultados é medida empregando-se três níveis de refino da discretização do domínio computacional, observandose a ordem de convergência e o índice de convergência de malhas GCI. Os resultados numéricos mostram que o número de Mach e a temperatura do fluido de trabalho independem da pressão de admissão, ao contrário do comportamento da pressão local e da massa específica. As propriedades do escoamento são fortemente dependentes da variação da geometria, e a variação do ângulo da seção divergente provoca uma mudança direta do número de Mach e inversa da pressão, da temperatura e da massa específica do escoamento no interior dessa seção. As simulações são comparadas com os resultados da teoria isentrópica e mostram que a linha sônica é deslocada do centro geométrico da garganta dos bocais para cada geometria simulada. A comparação com a teoria e com dados experimentais mostra desvios inferiores a 6x10-3 %. O uso do modelo de turbulência − / SST é capaz de resolver o escoamento com boa precisão, prevendo bem seu perfil de velocidades, as ondas de expansão de Prandtl-Meyer, juntamente com as interações dessas ondas com a camada limite. / The behavior of the supersonic flow inside rectangular convergent-divergent nozzle is investigated numerically by comparing four nozzles with different divergent sections, with a common aspect ratio AR=1.14, and the same nozzle exit-to-throat area ratios NAR=1.43. Nozzles are subject to several working fluid inlet pressures, maintaining a constant pressure ratio NPR=5. Simulations assume the flow in steady state, compressible, viscous, using a coupled approach with the turbulence model − /SST. The quality of results is measured by employing three refining levels of the computational domain discretization, observing the order of convergence and the grid convergence index GCI. Numerical results show that the Mach number and the temperature of the working fluid are independent of the inlet pressure, unlike the behavior of local pressure and the density. Flow properties are strongly dependent on the geometry variation, and the change on the angle of divergent section causes a direct effect on the Mach number and inverse on the pressure, the temperature and the density of the flow in this section. Simulations are compared to the results of the isentropic theory and show that the sonic line is offset from the geometric center of the throat nozzle, for each simulated geometry. Results from this work are compared to experimental and theoretical data and show deviations below 6x10-3 %. The − / SST turbulence model is able to solve the flow with good accuracy, and predicts its velocity profile, Prandtl-Meyer expansion waves, and their interactions with the boundary layer.

Os mecanismos de instabilidade hidrodinâmica têm um papel importante no processo da transição do escoamento de laminar para turbulento. A análise da instabilidade hidrodinâmica em uma cavidade com tampa deslizante foi realizada através da decomposição em modos globais (biglobal) para avaliar o efeito da compressibilidade neste fenômeno. O escoamento base foi obtido através de simulação numérica direta (DNS). Para tal, foi desenvolvido um código DNS compressível com discretização espacial por diferenças finitas compactas de alta resolução espectral e capacidade de processamento paralelo, com um método de decomposição de domínio que mantém a precisão das diferenças finitas compactas. O escoamento base é usado para montar o problema de autovalor oriundo das equações de Navier-Stokes linearizadas para a perturbação, discretizadas por diferenças finitas explícitas. O uso de diferenças finitas em conjunto com a implementação em matrizes esparsas reduz sensivelmente o uso de memória. Através do algoritmo de Arnoldi, a ordem do problema de autovalor é reduzida e os autovalores de interesse são recuperados. Os resultados indicam o efeito estabilizante da compressibilidade nos modos dominantes da cavidade e revelam modos inerentes ao escoamento compressível, para os quais a compressibilidade tem efeito desestabilizante. Dentre estes modos compressíveis, estão presentes modos de propagação sonora em dutos e modos relacionados à geração de som na cavidade. / Hydrodynamic instability mechanisms play an important role in laminar to turbulent transition. Hydrodynamic instability analysis of a lid-driven cavity flow was performed by global mode decomposition (biglobal) to evaluate compressibility effects on this phenomenon. The basic flow was calculated by direct numerical simulation (DNS). A compressible DNS code was developed with spectral-like compact finite difference spatial discretization. The code allows parallel processing with a domain decomposition method that preserves the compact finite difference accuracy. The basic flow is used to form the eigenvalue problem associated to the linear Navier- Stokes equations for the perturbation, which were discretized by an explicit finite difference scheme. The combination of sparse matrix techniques and finite difference discretization leads to a significant memory reduction. The order of the eigenvalue problem was reduced using the Arnoldi algorithm and the eigenvalues of interest were calculated. Results show the stabilizing effect of compressibility on the leading modes and reveal some modes intrinsic to compressible flow, for which compressibility has a destabilizing effect. Among these compressible modes, there are some related to sound propagation in ducts and to sound generation inside the cavity.

This work deals with the design of a hybrid time integrator that couples spatially explicit and implicit time integrators. In order to cope with the industrial solver of Ariane Group called FLUSEPA, the explicit scheme of Heun and the implicit scheme of Crank-Nicolson are hybridized using the transition parameter : the whole technique is called AION time integration. The latter is studied into details with special focus on spectral behaviour and on its ability to keep the accuracy. It is shown that the hybrid technique has interesting dissipation and dispersion properties while maintaining precision and avoiding spurious waves. Moreover, this hybrid approach is validated on several academic test cases for both convective and diffusive fluxes. And as expected the method is more interesting in term of computational time than standard time integrators. For the extension of this hybrid approach to the temporal adaptive method implemented in FLUSEPA, it was necessary to improve some treatments in order to maintain conservation and acceptable spectral properties. Finally the hybrid time integration was also applied to a RANS/LES turbulent test case with interesting computational time while capturing the flow physics.

Batch sedimentation is a method that enables us to understand the mechanism of compaction and compression of sedimenting slurry. However, batch settling behaviour is a very complex phenomenon that is not easily described fully by a mathematical model. This causes unrealistically large empirical calculations when the thickener size estimations are required. Channelling, reverse concentration gradients and the initial concentration of the slurry have large effects on batch settling. Existing procedures do not provide clear relationships involving these three significant variables. In this study, batch sedimentation phenomena are examined in detail and possible explanations are given to clarify the complex behaviour using recent theories. Modern research has shown that channelling is an unwanted formation because channels can change the concentration at the bottom and top of the bed by carrying a great amount of flocs upwards. Batch sedimentation tests were performed using flocculated slurry of Calcium Carbonate at various initial concentrations such as 250 g/l, 500 g/l, 750 g/l and 1000 g/l to observe channelling and reverse concentration gradients. Flux plots for the batch system reveal behaviour which can be attributed to the upward flow of solids. In addition, photographic methods were used to observe settling processes, channelling mechanisms and flocs in the channels. One of the purposes of this work was to examine the phenomenological solid-liquid separation theory of Buscall and White (1987), which employs the material properties of the local volume fraction, compressive yield stress Py ()ö and hindered settling function R()ö to identify the material behaviour in batch sedimentation. Stepped-pressure filtration and batch settling tests were used to measure the material characteristics for the flocculated CaCO3 suspension. Experimental data were demonstrated using Height versus Time and Height versus Concentration graphs and displayed the possible region of reverse concentration gradients and channelling in the settling bed. Mathematical predictions adopted from Usher (2002) were performed employing material characteristics of the material and graphical documentations were presented. The results of mathematical predictions were compared to the experimental results and the modes of sedimentation explained by Lester et al. (2005). Fundamental theoretical models and experimental observations highlight that the main driving force for channelling is the high-pressure gradient at the bottom of the bed and the most important factors that cause channelling are high initial concentration of slurry and settling time. The predictions also show that the material and flocculant used for the batch settling tests demonstrate important effect on the settling process. The knowledge and information gained from this study is valuable to maximize the thickening process.

On simule l'écoulement de gaz avec arc électrique au moment d'une interruption du courant dans un disjoncteur haute ou moyenne tension. Les équations de Navier Stokes compressible sont résolues pour un gaz réel non visqueux. L'arc électrique est du gaz chaud et conducteur. Le rayonnement est pris en compte. L'équation d'état et les propriétés du gaz sont tabulées. On présente une analyse du problème physique et des équations a résoudre. Deux modélisations numériques monodimensionnelles, par une methode d'éléments finis (avec upwind et capture de choc) et par une methode a pas fractionnaires séparant le traitement des phénomènes de diffusion et de convection, ont été implémentées. Elles ont permis d'affiner le modèle et de mettre en évidence l'importance du rayonnement. La dernière partie est relative au développement d'un code de calcul industriel 2 d plan ou axisymetrique. Seule la methode a pas fractionnaires a été implémentée. Le code permet de résoudre le problème sur des maillages non structures et mobiles. Des tests numériques valident les schémas mis en œuvre

Dans cette étude nous considérons l'ébullition en paroi sous les angles de la modélisation et de la simulation numérique. Dans un premier temps nous proposons une revue bibliographique au sujet du régime d'ébullition nucléée à fort ux de chaleur pariétal et analysons plus particulièrement la compréhension du phénomène de crise d'ébullition de type caléfaction. Nous en déduisons une motivation pour l'étude de la dynamique de croissance de bulle au moyen de la simulation numérique. L'essentiel du travail concerne alors le développement d'un modèle de type champ de phase pour l'étude des écoulements liquide-vapeur avec changement de phase. Nous proposons une fermeture thermodynamique quasi-compressible dont les propriétés sont adaptées aux simulations envisagées. Le système d'équations du mouvement qui s'en déduit constitue une régularisation thermodynamiquement cohérente de la description discontinue du système diphasique, ce qui est l'avantage des modèles à interfaces diffuses. Nous démontrons que la formulation retenue permet de définir l'épaisseur de la zone régularisée indépendamment de la description thermodynamique des phases, ce qui est intéressant numériquement. Nous établissons la relation cinétique et analysons ainsi la modélisation champ de phase des mécanismes dissipatifs. Finalement nous étudions la résolution numérique du modèle à l'aide de simulations de transition de phase en configuration simple et de dynamique de bulle en isotherme.

This thesis deals with flow measuring techniques applied to steady and pulsating gas flows relevant to gas exchange systems for internal combustion engines. Gas flows in such environments are complex, i.e. they are inhomogeneous, three-dimensional, unsteady, non-isothermal and exhibit significant density changes. While a variety of flow metering devices are available and have been devised for such flow conditions, the performance of these flow metersis to a large extent undocumented when a strongly pulsatile motion is superposed on the already complex flow field. Nonetheless, gas flow meters are commonly applied in such environments, e.g. in the measurement of the air flow to the engine or the amount of exhaust gas recirculation. The aim of the present thesis is therefore to understand and assess, and if possible to improve the performance of various flow meters under highly pulsatile conditions as well as demonstrating the use of a new type of flow meter for measurements of the pulsating mass flow upstream and downstream the turbine of a turbocharger. The thesis can be subdivided into three parts. The first one assesses the flow quality of a newly developed flow rig, designed for measurements of steady and pulsating air flow at flow rates and pulse frequencies typically found in the gas exchange system of cars and smaller trucks. Flow rates and pulsation frequencies achieved and measured range up to about 200 g/s and 80 Hz, respectively. The time-resolved mass flux and stagnation temperature under both steady and pulsating conditions were characterized by means of a combined hot/cold-wire probe which is part of a newly developed automated measurement module. This rig and measurement module were used to create a unique data base with well-defined boundary conditions to be used for the validation of numerical simulations, but in particular, to assess the performance of various flow meters. In the second part a novel vortex flow meter that can measure the timedependent flow rate using wavelet analysis has been invented, verified and extensively tested under various industrially relevant conditions. The newly developed technique was used to provide unique turbine maps under pulsatile conditions through time-resolved and simultaneous measurements of mass flow, temperature and pressure upstream and downstream the turbine. Results confirm that the quasi-steady assumption is invalid for the turbine considered as a whole. In the third and last part of the thesis, two basic fundamental questions that arose during the course of hot/cold-wire measurements in the aforementioned high speed flows have been addressed, namely to assess which temperature a cold-wire measures or to which a hot-wire is exposed to in high speed flows as well as whether the hot-wire measures the product of velocity and density or total density. Hot/cold-wire measurements in a nozzle have been performed to test various hypothesis and results show that the recovery temperature as well as the product of velocity and stagnation density are measured. / QC 20120510