La Simulation aux Grandes Echelles (SGE) est de plus en plus utilisée dans les processus de développement et la conception des réacteurs aéronautiques industriels. L'une des raisons pour ce besoin résulte dans la capacité de la SGE à fournir des informations instantanées d'un écoulement turbulent augmentant la quantité des prédictions de la composition des gaz d'échappement. Ce manuscrit de thèse aborde deux sujets récurrents de la SGE. D'une part, les schémas numériques pour la SGE nécessitent certaines propriétés, notamment une précision élevée avec une diffusivité faible pour ne pas nuire aux modèles de turbulence. Afin de répondre à ce pré requis, une famille de schémas d'intégration temporelle d'ordre élevée est proposée, permettant de modifier la diffusion numérique du schéma. D'autre part, la SGE étant intrinsèquement instationnaire, elle est très consommatrice en temps CPU. De plus, une géométrie complexe prend beaucoup de temps de simulation même avec les super calculateurs d'aujourd'hui. Dans le cas particulier d'intérêt et souvent rencontré dans les applications industrielles, l'approche bas-Mach est constitue une alternative intéressante permettant de réduire le coût et le temps de retour d'une simulation LES. L'impact et la comparaison des formalismes compressible et incompressible sont toutefois rarement quantifiés, ce qui est proposé dans ce travail pour une configuration représentative d'un brûleur swirlé industriel mesuré au CORIA / Large-Eddy Simulation (LES) becomes a more and more demanded tool to improve the design of aero-engines. The main reason for this request stems from the constraints imposed on the next generation low-emission engines at the industrial development level and the ability for LES to provide information on the instantaneous turbulent flow field which greatly contributes to improving the prediction of mixing and combustion thereby offering an improved prediction of the exhaust emission. The work presented in this thesis discusses two recurring issues of LES. For one, numerical schemes for LES require certain properties, i.e. low-diffusion schemes of high order of accuracy so as not to interfere with the turbulence models. To meet this purpose in the context of fully unstructured solvers, a new family of high-order time-integration schemes is proposed. With this class of schemes, the diffusion implied by the numerical scheme become adjustable and built-in. Second, since fully unsteady by nature, LES is very consuming in terms of CPU time. Even with today's supercomputers complex problems require long simulation times. Due to the low flow velocities often occurring in industrial applications, the use of a low-Mach number solver seems suitable and can lead to large reductions in CPU time if comparable to fully compressible solvers. The impact of the incompressibility assumption and the different nature of the numerical algorithms are rarely discussed. To partly answer the question, detailed comparisons are proposed for an experimental swirled configuration representative of a real burner that is simulated by LES using a fully explicit compressible solver and an incompressible solution developed at CORIA

Cette thèse est consacrée à l'étude de la dynamique des gaz à faible nombre de Mach. Le modèle étudié provient des équations de Navier-Stokes complètes lorsque le nombre de Mach tend vers zéro. On cherche à montrer que le problème de Cauchy correspondant est bien posé. Les cas visqueux et non visqueux sont tous deux considérés. Les coefficients physiques peuvent dépendre de la densité (ou de la température) inconnue. En particulier, nous prenons en compte les effets de conductivité thermique et on autorise de grandes variations d'entropie. Rappelons qu'en absence de diffusion thermique, la limite à faible nombre de Mach implique la condition d'incompressibilité. Dans le cadre étudié ici, en introduisant un nouveau champ de vitesses à divergence nulle, le système devient un couplage non linéaire entre une équation quasi-parabolique pour la densité et un système de type Navier-Stokes (ou Euler) pour la vitesse et la pression. Pour le cas avec viscosité, on établit le résultat classique, à savoir qu'il existe une solution forte existant localement (resp. globalement) en temps pour des données initiales grandes (resp. petites). On considère ici le problème de Cauchy avec données initiales dans des espaces de Besov critiques. Lorsque les coefficients physiques du système vérifient une relation spéciale, le système se simplifie considérablement, et on peut alors établir qu'il existe des solutions faibles globales en temps à énergie finie. Par un argument d'unicité fort-faible, on en déduit que les solutions fortes à énergie finie existent pour tous les temps positifs en dimension deux. Pour le cas sans viscosité, on montre d'abord le caractère bien posé dans des espaces de Besov limites, qui s'injectent dans l'espace des fonctions lipschitziennes. Des critères de prolongement et des estimations du temps de vie sont établis. Si l'on suppose la donnée initiale à énergie finie dans l'espace de Besov limite à exposant de Lebesgue infini, on a également un résultat d'existence locale. En dimension deux, le temps de vie tend vers l'infini quand la densité tend vers une constante positive. Des estimations de produits et de commutateurs, ainsi que des estimations a priori pour les équations paraboliques et pour le système de Stokes (ou d'Euler) à coefficients variables, se trouvent dans l'annexe. Ces estimations reposent sur la théorie de Littlewood-Paley et le calcul paradifférentiel / This thesis is devoted to the study of the dynamics of the gases with small Mach number. The model comes from the complete Navier-Stokes equations when the Mach number goes to zero, and we aim at showing that it is well-posed. The viscous and inviscid cases are both considered. The physical coefficients may depend on the unknown density (or on the unknown temperature).In particular, we consider the effects of the thermal conductivity and hence large variations of entropy are allowed. Recall that if there is no thermal diffusion, then the low Mach number limit just implies the incompressibility condition. In the framework considered here, by introducing a new solenoidal velocity field, the system becomes a nonlinear coupling between a quasi-parabolic equation for the density and an evolutionary Stokes (or Euler) system for the velocity and the pressure. For the case with viscosity, we establish classical results, namely the strong solutions exist locally (resp. globally) in time for big (resp. small) initial data. We consider the Cauchy problem in the critical Besov spaces with the lowest regularity. Under a special relationship between the two physical coefficients, the system recasts in a simpler form and one may prove that there exist weak solutions with finite energy. In dimension two, this implies that strong solutions with finite energy exist for all positive times. In the inviscid case, we first prove the well-posedness result in endpoint Besov spaces, which can be embedded into the set of Lipschitzian functions. Continuation criterions and estimates for the lifespan are both established.If we suppose the initial data to be in the borderline Besov spaces with infinite Lebesgue exponent and to be of finite energy, we also have a local existence result. In dimension two, the lifespan goes to infinity when the density tends to a positive constant. Estimates for products and commutators, together with a priori estimates for the parabolic equations and the Stokes (or Euler) system with variable coefficients, are postponed in the appendix. These estimates are based on the Littlewood-Paley theory and the paradifferential calculus

O foco principal do presente trabalho é estender uma metodologia numérica embasada no uso de uma técnica de refinamento adaptativo de malha (AMR - Adaptive Mesh Refinement) e no uso de esquemas temporais multipasso implícitos-explícitos (IMEX) a aplicações envolvendo escoamentos reativos com baixo número de Mach. Originalmente desenvolvida para escoamentos incompressíveis, a formulação euleriana daquela metodologia emprega as equações de Navier-Stokes como modelo matemático para descrever a dinâmica do escoamento e o Método da Projeção, baseado no divergente nulo da velocidade do escoamento, para tratar o acoplamento pressão-velocidade presente na formulação com variáveis primitivas. Tal formulação euleriana original é estendida para acomodar novas equações agregadas ao modelo matemático da fase contínua: conservação de massa, fração de mistura (para representar as concentrações de combustível e oxidante), e energia. Além disso, uma equação termodinâmica de estado é integrada ao modelo matemático estendido e é empregada juntamente com a equação de conservação de massa para produzir uma nova restrição (não nula desta vez) ao divergente do campo de velocidade. Assume-se que o escoamento ocorre a baixo número de Mach (hipótese principal). O Método de Diferença Finita é empregado na discretização espacial das variáveis eulerianas de estado, empregando-se uma malha AMR. As vantagens e dificuldades desta extensão são cuidadosamente investigadas e reportadas. Pela importância, do ponto de vista de aplicações práticas, alguns estudos numéricos preliminares envolvendo escoamentos incompressíveis turbulentos com sprays são realizados (as gotículas compõem a fase dispersa). Num primeiro momento, apenas sprays com gotículas inertes são considerados. Embora ainda apenas iniciais, tais estudos já se mostram importantes pois identificam com clareza, em primeira instância, algumas das dificuldades inerentes a serem enfrentadas ao se tratar dentro desta nova metodologia um conjunto relativamente grande de gotículas lagrangianas. No caso de escoamentos incompressíveis turbulentos com sprays, a integração temporal se dá com métodos IMEX para a fase contínua e com o Método de Euler Modificado para a fase dispersa. A turbulência, em todos os casos que a envolvem, é tratada pelo modelo de Simulação das Grandes Escalas (LES - Large Eddy Simulation). As simulações computacionais se dão em um domínio tridimensional, um parelelepípedo, e empregam uma extensão (resultante do presente trabalho) do código AMR3D, um programa de computador sequencial implementado em Fortran90, oriundo de uma colaboração de longa data entre o IME-USP e o MFLab/FEMEC-UFU (Laboratório de Dinâmica de Fluidos da Universidade Federal de Uberlândia). O processamento foi efetuado no LabMAP (Laboratório da Matemática Aplicada do IME-USP). / It is the main goal of the present work to extend a numerical methodology based on both the use of an adaptive mesh refinement technique (AMR) and the use of a multistep, implicit-explicit time-step strategy (IMEX) to applications involving low Mach number reactive flows. Originally developed for incompressible flows, the Eulerian formulation of that methodology employs the Navier-Stokes equations to model the flow dynamics and the Projection Method, based on the vanishing divergence of the velocity field, to tackle the pressure-velocity coupling present when using primitive variables. That Eulerian formulation is extended by adding a new set of equations to the original mathematical model, describing the various properties of the continuous phase: mass conservation, mixture fraction (to represent concentrations of fuel and oxidizer) and energy. Also, a thermodynamic equation of state is included into the extended mathematical model which is employed, along with the equation for the conservation of mass, to derive a new restriction (this time, different from zero) to the divergence of the velocity field. It is assumed that one is dealing with a low Mach number flow (the main hipothesis). The discretization in space employs the Finite Difference Method for the Eulerian variables on a AMR mesh. Advantages and difficulties of such an extension of the previous methodology are carefully investigated and reported. For its importance in the real-world applications, few preliminary numerical studies involving incompressible turbulent flows with sprays are performed (the droplets form what it is called the dispersed phase). Only sprays formed by inert droplets are considered. Even though initial yet, such studies are most important because they clearly identify, first hand, certain difficulties in handling relatively large sets of Lagrangian droplets in the context of this new AMR methodology. In the context of turbulent incompressible flows with sprays, the overall time-step scheme is given by IMEX methods for the continuous phase and by the Improved Euler Method for the dispersed phase. In all the cases in which it is considered, turbulence is modeled by the Large Eddy Simulation (LES) model. The computational simulations are held in a tridimensional domain given by a paralellepiped and all of them employ the extention (resulting of the present work) of the AMR3D code, a sequencial computer program implemented in Fortran90, whose origin is the collaborative work between IMEUSP and MFLab/FEMEC-UFU (Fluid Dynamics Laboratory, Federal University of Uberlândia). Computations were performed at LabMAP (Applied Mathematics Laboratory at IME-USP).

A body in a hypersonic flow field will experience very high heating especially during re-entry. Conventionally this problem is tackled to some extent by the use of large angle blunt cones. At the cost of increased drag, the heat transfer rate is lower over most parts of the blunt body, except in a region around the stagnation point. Thus even with blunt cones, management of heat transfer rates and drag on bodies at hypersonic speeds continues to be an interesting research area. Various thermal protection systems have been proposed in the past, like heat sink cooling, ablation cooling and aerospikes. The ablative cooling system becomes extremely costly when reusability is the major concern. Also the shape change due to ablation can lead to issues with the vehicle control. The aerospikes themselves may become hot and ablate at hypersonic speeds. Hence an alternate form of cooling system is necessary for hypersonic flows, which is more feasible, cost effective and efficient than the conventional cooling systems. Injection of a mass of cold fluid into the boundary layer through the surface is one of the potential cooling techniques in the hypersonic flight corridors. These kinds of thermal protection systems are called mass transfer cooling systems. The injection of the mass may be through discrete slots or through a porous media. When the coolant is injected through a porous media over the entire surface, the coolant comes out as a continuous mass. Such a cooling system is also referred as “transpiration cooling system”. When the fluid is injected through discrete slots, the system is called as “film cooling system”. In either case, the coolant absorbs the incoming heat through its rise in enthalpy and thus modifies the boundary layer characteristics in such a way that the heat flow rate to the surface is less. Injection of a forward facing jet (opposite to the freestream direction) from the stagnation point of a blunt body can be used for mitigating both the aerodynamic drag and heat transfer rates at hypersonic Mach numbers. If the jet has enough momentum it can push the bow shock forward, resulting in reduced drag. This will also reduce heat transfer rate over most part of the body except around the jet re-attachment region. A reattachment shock impinging on the blunt body invariably increases the local heat flux. At lower momentum fluxes the forward facing jet cannot push the bow shock ahead of the blunt body and spreads easily over the boundary layer, resulting in reduced heat transfer rates. While the film cooling performance improves with mass flow rate of the jet, higher momentum flow rates can lead to a stronger reattachment leading to higher heat transfer rate at the reattachment zone. If we are able to reduce the momentum flux of the coolant for the same mass flow rate, the gas coming out can easily spread over the boundary layer and it is possible to improve the film cooling performance. In all the reported literature, the mass flow rate and the momentum flux are not varied independently. This means, if the mass flow rate is increased, there is a corresponding increase in the momentum flux. This is because the injection (from a particular orifice and for a particular coolant gas) is controlled only by the total pressure of injection and free stream conditions. The present investigation is mainly aimed at demonstrating the effect of reduction in momentum of the coolant (injected opposing a hypersonic freestream from the stagnation point of a blunt cone), keeping the mass flow rate the same, on the film cooling performance. This is achieved by splitting a single jet into a number of smaller jets of same injection area (for same injection total pressure and same free stream conditions). To the best of our knowledge there is no report on the use of forward facing micro-jet array for film cooling at hypersonic Mach numbers. In this backdrop the main objectives of the present study are: • To experimentally demonstrate the effect of splitting a single jet into an array of closely spaced smaller micro-jets of same effective area of injection (injected opposite to a hypersonic freestream from the stagnation zone), on the reduction in surface heat transfer rates on a large angle blunt cone. · Identifying various parameters that affect the flow phenomenon and doing a systematic investigation of the effect of the different parameters on the surface heat transfer rates and drag. Experimental investigations are carried out in the IISc hypersonic shock tunnel on the film cooling effectiveness. Coolant gas (nitrogen and helium) is injected opposing hypersonic freestream as a single jet (diameter 2 mm and 0.9 mm), and as an array of iv micro jets (diameter 300 micron each) of same effective area (corresponding to the respective single jet). The coolant gas is injected from the stagnation zone of a blunt cone model (58o apex angle and nose radius of 35 mm). Experiments are performed at a flow freestream Mach number of 5.9 at 0o angle of attack, with a stagnation enthalpy of 1.84 MJ/Kg, with and without injections. The ratios of the jet stagnation pressure to the pitot pressure (stagnation pressure ratio) used in the present study are 1.2 and 1.45. Surface convective heat transfer measurements using platinum thin film sensors, time resolved schlieren flow visualization and aerodynamic drag measurements using accelerometer force balance are used as flow diagnostics in the present study. The theoretical stagnation point heat transfer rate without injection for the given freestream conditions for the test model is 79 W/cm2 and the corresponding aerodynamic drag from Newtonian theory is 143 N. The measured drag value without injection (125 N) shows a reasonable match with theory. As the injection is from stagnation zone it is not possible to measure the surface heat transfer rates at the stagnation point. The sensors thus are placed from the nearest possible location from the stagnation point (from 16 mm from stagnation point on the surface). The sensors near the stagnation point measures a heat transfer rate of 65 W/cm2 on an average without any injection. Some of the important conclusions from the study are: • Up to 40% reduction in surface heat transfer rate has been measured near the stagnation point with the array of micro jets, nitrogen being the coolant, while the corresponding reduction was up to 30% for helium injection. Considering the single jet injection, near the stagnation point there is either no reduction in heat transfer rate or a slight increase up to 10%. · Far away from stagnation point the reduction in heat transfer with array of micro-jets is only slightly higher than corresponding single jet for the same pressure ratio. Thus the cooling performance of the array of closely spaced micro jets is better than the corresponding single jet almost over the entire surface. • The time resolved flow visualization studies show no major change in the shock standoff distance with the low momentum gas injection, indicating no major changes in other aerodynamic aspects such as drag. · The drag measurements also indicate that there is virtually no change in the overall aerodynamic drag with gas injection from the micro-orifice array. · The spreading of the jets injected from the closely spaced micro-orifice array over the surface is also seen in the visualization, indicating the absence of a region of strong reattachment. · The reduction in momentum flux of the injected mass due to the interaction between individual jets in the case of closely spaced micro-jet array appears to be the main reason for better performance when compared to a single jet. The thesis is organized in six chapters. The importance of film cooling at hypersonic speeds and the objectives of the investigation are concisely presented in Chapter 1. From the knowledge of the flow field with counter-flow injection obtained from the literature, the important variables governing the flow phenomena are organized as non-dimensional parameters using dimensional analysis in Chapter 2. The description of the shock tunnel facility, diagnostics and the test model used in the present study is given in Chapter 3. Chapter 4 describes the results of drag measurements and flow visualization studies. The heat transfer measurements and the observed trends in heat transfer rates with and without coolant injection are then discussed in detail in Chapter 5. Based on the obtained results the possible physical picture of the flow field is discussed in Chapter 6, followed by the important conclusions of the investigation.

O foco principal do presente trabalho é estender uma metodologia numérica embasada no uso de uma técnica de refinamento adaptativo de malha (AMR - Adaptive Mesh Refinement) e no uso de esquemas temporais multipasso implícitos-explícitos (IMEX) a aplicações envolvendo escoamentos reativos com baixo número de Mach. Originalmente desenvolvida para escoamentos incompressíveis, a formulação euleriana daquela metodologia emprega as equações de Navier-Stokes como modelo matemático para descrever a dinâmica do escoamento e o Método da Projeção, baseado no divergente nulo da velocidade do escoamento, para tratar o acoplamento pressão-velocidade presente na formulação com variáveis primitivas. Tal formulação euleriana original é estendida para acomodar novas equações agregadas ao modelo matemático da fase contínua: conservação de massa, fração de mistura (para representar as concentrações de combustível e oxidante), e energia. Além disso, uma equação termodinâmica de estado é integrada ao modelo matemático estendido e é empregada juntamente com a equação de conservação de massa para produzir uma nova restrição (não nula desta vez) ao divergente do campo de velocidade. Assume-se que o escoamento ocorre a baixo número de Mach (hipótese principal). O Método de Diferença Finita é empregado na discretização espacial das variáveis eulerianas de estado, empregando-se uma malha AMR. As vantagens e dificuldades desta extensão são cuidadosamente investigadas e reportadas. Pela importância, do ponto de vista de aplicações práticas, alguns estudos numéricos preliminares envolvendo escoamentos incompressíveis turbulentos com sprays são realizados (as gotículas compõem a fase dispersa). Num primeiro momento, apenas sprays com gotículas inertes são considerados. Embora ainda apenas iniciais, tais estudos já se mostram importantes pois identificam com clareza, em primeira instância, algumas das dificuldades inerentes a serem enfrentadas ao se tratar dentro desta nova metodologia um conjunto relativamente grande de gotículas lagrangianas. No caso de escoamentos incompressíveis turbulentos com sprays, a integração temporal se dá com métodos IMEX para a fase contínua e com o Método de Euler Modificado para a fase dispersa. A turbulência, em todos os casos que a envolvem, é tratada pelo modelo de Simulação das Grandes Escalas (LES - Large Eddy Simulation). As simulações computacionais se dão em um domínio tridimensional, um parelelepípedo, e empregam uma extensão (resultante do presente trabalho) do código AMR3D, um programa de computador sequencial implementado em Fortran90, oriundo de uma colaboração de longa data entre o IME-USP e o MFLab/FEMEC-UFU (Laboratório de Dinâmica de Fluidos da Universidade Federal de Uberlândia). O processamento foi efetuado no LabMAP (Laboratório da Matemática Aplicada do IME-USP). / It is the main goal of the present work to extend a numerical methodology based on both the use of an adaptive mesh refinement technique (AMR) and the use of a multistep, implicit-explicit time-step strategy (IMEX) to applications involving low Mach number reactive flows. Originally developed for incompressible flows, the Eulerian formulation of that methodology employs the Navier-Stokes equations to model the flow dynamics and the Projection Method, based on the vanishing divergence of the velocity field, to tackle the pressure-velocity coupling present when using primitive variables. That Eulerian formulation is extended by adding a new set of equations to the original mathematical model, describing the various properties of the continuous phase: mass conservation, mixture fraction (to represent concentrations of fuel and oxidizer) and energy. Also, a thermodynamic equation of state is included into the extended mathematical model which is employed, along with the equation for the conservation of mass, to derive a new restriction (this time, different from zero) to the divergence of the velocity field. It is assumed that one is dealing with a low Mach number flow (the main hipothesis). The discretization in space employs the Finite Difference Method for the Eulerian variables on a AMR mesh. Advantages and difficulties of such an extension of the previous methodology are carefully investigated and reported. For its importance in the real-world applications, few preliminary numerical studies involving incompressible turbulent flows with sprays are performed (the droplets form what it is called the dispersed phase). Only sprays formed by inert droplets are considered. Even though initial yet, such studies are most important because they clearly identify, first hand, certain difficulties in handling relatively large sets of Lagrangian droplets in the context of this new AMR methodology. In the context of turbulent incompressible flows with sprays, the overall time-step scheme is given by IMEX methods for the continuous phase and by the Improved Euler Method for the dispersed phase. In all the cases in which it is considered, turbulence is modeled by the Large Eddy Simulation (LES) model. The computational simulations are held in a tridimensional domain given by a paralellepiped and all of them employ the extention (resulting of the present work) of the AMR3D code, a sequencial computer program implemented in Fortran90, whose origin is the collaborative work between IMEUSP and MFLab/FEMEC-UFU (Fluid Dynamics Laboratory, Federal University of Uberlândia). Computations were performed at LabMAP (Applied Mathematics Laboratory at IME-USP).

Ce manuscrit porte sur le développement d’une méthode de calcul du bruit d’écoulement à faible nombre de Mach en conduite en présence d’obstacles. Elle consiste en une simulation numérique de l’écoulement dont les données sauvegardées servent dans un deuxième temps à l’estimation du rayonnement acoustique. Le calcul de mécanique des fluides est réalisé avec un modèle de turbulence LES incompressible. Un code a été développé pour le calcul acoustique qui comprend la transformée de Fourier des données de l’écoulement, le calcul par méthode intégrale et les post-traitements. L’originalité de la méthode réside dans le fait qu’elle nécessite comme informations sur l’écoulement uniquement des données surfaciques. Aucune donnée volumique n’a besoin d’être sauvegardée durant le calcul de mécanique des fluides. Dans une conduite obstruée par un obstacle, les principales sources de bruit se situent à proximité de l’obstacle et le rayonnement diffracté sur l’obstacle domine sur le rayonnement direct des sources. C’est la diffraction du rayonnement incident qui rend négligeable les données volumiques dans le calcul acoustique. Elle est étudiée et illustrée à travers le cas simplifié du rayonnement d’une source placée à proximité d’un obstacle en conduite. Des comparaisons calculs/essais favorables valident la méthode et le code de calcul développé. Plusieurs configurations sont étudiées : l’insertion dans un conduit rectangulaire droit d’un diaphragme, d’un agencement de deux diaphragmes et d’un volet. En présence d’un diaphragme, une théorie de la similitude est introduite ; en présence d’un double diaphragme, des phénomènes d’interaction entre les deux obstacles apparaissent et sont analysés ; et en présence d’un volet, des résonances particulières sont examinées. / This manuscript deals with the development of a calculation method of low Mach number flow noise in ducts in the presence of fixed obstacles. lt consists of a numerical simulation of the flow during which data are saved and used in a second time to estimate the acoustic radiation. The fluid calculation is performed with an incompressible LES turbulence model. A code has been developed for the acoustic computation which includes the Fourier transform of the flow data, the integral computation and the post-processing. The originality of the method lies in the fact that it requires as information on the flow only surface data. No volume term needs to be saved during the fluid calculation. ln a duct obstructed by an obstacle, the main sources of noise are located near the obstacle and the scattered field on the obstacle dominates on the direct radiation of the sources. lt is the diffraction of the incident radiation which makes the volume data negligible in the acoustic calculation. This is studied and illustrated through the simplified case of the radiation from a source located near a ducted obstacle. Favorable comparisons with measurements validate the method and the developed code. Several configurations are studied: the insertion into a straight rectangular duct of a diaphragm, an arrangement of two diaphragms and a flap. ln the presence of a diaphragm, a theory of similarity is introduced; in the presence of a double diaphragm, interaction phenomena between the two obstacles appear and are analyzed; and in the presence of a flap, particular resonances are examined.

L'instabilité de Rayleigh-Taylor (IRT) est notamment rencontrée lors des expériences de Fusion par Confinement Inertiel, et son développement est un obstacle à la réussite de ces expériences. L'objet de cette thèse est d'étudier la croissance de l'IRT pour différents régimes de compressibilité, au moyen de simulations numériques directes réalisées à l'aide d'un code pseudo-spectral multidomaine de type Chebyshev-Fourier-Fourier.La méthode du développement asymptotique permet d'établir des modèles à bas nombre de Mach pour lesquels la contribution acoustique est éliminée. L'implantation dans le code de simulation du modèle anélastique, qui met en jeu des fluides stratifiés et capture les effets thermiques, est améliorée. Le modèle de Boussinesq est ajouté au code. La précision de la méthode numérique est étudiée pour différents découpages en sous-domaines. Plusieurs éléments de validation sont présentés, dont la comparaison avec une expérience.La première simulation présentée, réalisée avec le modèle de Boussinesq, s'intéresse à la croissance auto-semblable de l'IRT. Les lois d'échelle de la vorticité et de la dissipation sont dégagées. La structure de la turbulence et du mélange entre les deux fluides est discutée. Certaines propriétés de la turbulence homogène et isotrope sont retrouvées, mais on note la persistance d'anisotropie aux petites échelles. Les premières simulations 3D de l'IRT avec le modèle anélastique sont présentées. L'influence des effets de compressibilité sur les premières phases de la croissance est étudiée. En outre, une couche de mélange anélastique en faible stratification est analysée et présente des effets de compressibilité non négligeables. / The Rayleigh-Taylor instability (RTI) is especially observed in inertial confinement fusion experiments, and its development prevents the success of these experiments. The purpose of this work is to study the growth of the RTI for different compressibility regimes by using a multidomain pseudospectral Chebyshev-Fourier-Fourier simulation code. The asymptotic expansion method allows to establish several low Mach number models which do not contains acoustics. The implantation of the anelastic model, which deals with stratified fluids and captures thermal effects, has been improved. Moreover, the Boussinesq model is added to the simulation code. The accuracy of the entire numerical method is studied, as a function of the subdomain separation, and several validation elements are shown, including a comparison with an experimental study. The first simulation to be analyzed is achieved with the Boussinesq model. We focus on the self-similarity of the RTI growth. The temporal scalings of vorticity and dissipation are displayed, and the structures of turbulence and mixing are discussed. Some properties of isotropic and homogeneous turbulence are observed, however some anisotropy remains at small scales. The first three-dimensional anelastic simulations are presented. The influence of compressibility effects on the first stages of the growth is studied. Finally, a developed anelastic mixing layer involving weakly stratified fluids is described and was found to display non-negligible compressibility effects.

This work represents the initial steps in a wider project that aims to map out the sensitive areas in fuel injectors and combustion chambers. Direct numerical simulation (DNS) using a Low-Mach-number formulation of the Navier–Stokes equations is used to calculate direct-linear and adjoint global modes for axisymmetric low-density jets and lifted jet diffusion flames. The adjoint global modes provide a map of the most sensitive locations to open-loop external forcing and heating. For the jet flows considered here, the most sensitive region is at the inlet of the domain. The sensitivity of the global-mode eigenvalues to force feedback and to heat and drag from a hot-wire is found using a general structural sensitivity framework. Force feedback can occur from a sensor-actuator in the flow or as a mechanism that drives global instability. For the lifted flames, the most sensitive areas lie between the inlet and flame base. In this region the jet is absolutely unstable, but the close proximity of the flame suppresses the global instability seen in the non-reacting case. The lifted flame is therefore particularly sensitive to outside disturbances in the non-reacting zone. The DNS results are compared to a local analysis. The most absolutely unstable region for all the flows considered is at the inlet, with the wavemaker slightly downstream of the inlet. For lifted flames, the region of largest sensitivity to force feedback is near to the location of the wavemaker, but for the non-reacting jet this region is downstream of the wavemaker and outside of the pocket of absolute instability near the inlet. Analysing the sensitivity of reacting and non-reacting variable-density shear flows using the low-Mach-number approximation has up until now not been done. By including reaction, a large forward step has been taken in applying these techniques to real fuel injectors.

L'objectif de cette thèse est l'accélération des solveurs de Gradient Conjugué avec déflation utilisés pour la résolution de l'équation de Poisson pour la pression, dans le cas de la simulation d'écoulements à faible nombre de Mach sur des maillages non structurés. Une méthode de redémarrage basée sur une estimation de l'effet des erreurs numériques a été mise en œuvre et validée. Par la suite, une méthode à trois niveaux de maillage a été créée, et deux techniques ont dû être développées pour réduire le nombre d'itérations sur les niveaux grossiers : l'une permet la création de solutions initiales grâce à une méthode de projection adaptée, l'autre consiste en une adaptation du critère de convergence sur les niveaux grossiers. Les résultats numériques sur des simulations massivement parallèles montrent entre autres une réduction considérable du temps de calcul global. D'autres pistes de recherche sont introduites, notamment concernant l'équilibrage dynamiques de charge de calcul. / The main objective of this thesis is to accelerate deflated Conjugate Gradient solvers used for solving the pressure Poisson equation, for the simulation of low-Mach number flows on unstructured meshes. A restart method based on an estimation of the effect of numerical errors has been implemented and validated. Then, a three-level deflation method has been created, and two techniques are developed in order to reduce the number of iterations on the coarse levels : one of them is the creation of initial guesses thanks to a well-suited projection method, the other one consists in adapting the convergence criterion on the coarse grids. Numerical results on massively parallel simulations show, among others, a drastic reduction of the computational times of the solver. Other lines of research are introduced, especially regarding dynamic load balancing.

L'objectif de cette thèse est de développer un nouveau schéma numérique du type volumes finis pour la dynamique des gaz. Dans deux articles, F.Berthelin, T.Goudon et S.Minjeaud proposent de résoudre le système des équations d'Euler barotrope en dimension 1 d'espace, avec un schéma d'ordre 1 fonctionnant sur grilles décalées et dont la conception des flux est inspirée des schémas cinétiques. Nous proposons d'enrichir ce schéma afin qu'il puisse résoudre le système des équations d'Euler barotrope ou complet, en dimension 2 d'espace sur maillage cartésien ou non structuré, possiblement à l'ordre 2 et le cas échéant à bas nombres de Mach. Nous commencerons par développer une version 2D du schéma sur grilles cartésiennes (ou MAC) à l’ordre 2 via une méthode de type MUSCL, d'abord pour les équations barotropes puis pour les équations complètes. Ces dernières demandent de traiter une équation d’énergie supplémentaire et l’un des problèmes -résolu- est de trouver une définition discrète convenable de l’énergie totale telle qu'elle satisfasse une équation conservative locale. Dans un troisième chapitre nous étudierons le passage à la limite du compressible vers l'incompressible et nous verrons comment utiliser les atouts de notre schéma afin de le modifier et d'en faire un schéma Asymptotic Preserving pour des écoulements à bas nombres de Mach. Dans un quatrième temps nous proposerons une adaptation du schéma sur des maillages non structurés. Notre approche sera fortement inspirée des méthodes DDFV et pourra présenter des avantages dans les régimes à faibles nombres de Mach. Cette thèse se termine par un cinquième chapitre issu d’une collaboration lors du CEMRACS 2017, où le point de vue considéré n’est plus macroscopique mais microscopique. Nous commencerons par étudier un modèle micro/macro idéalisé auquel un processus stochastique a été ajouté puis nous tenterons d'en déduire un modèle à grande échelle pour un système fortement couplé, qui soit consistant avec la description micro/macro sous-jacente du problème physique. / The objective of this thesis is to develop a new numerical scheme of finite volume type for gas dynamics. In two articles, F.Berthelin, T.Goudon and S.Minjeaud propose to solve the barotropic Euler system in dimension 1 of space, with a first order scheme that works on staggered grids and of which fluxes are inspired by kinetic schemes. We propose to enhance this scheme so that it can solve the barotropic or complete Euler systems, in dimension 2 of space on Cartesian or unstructured grids, possibly at order 2 and at Low Mach numbers where appropriate. We begin with the development of a 2D version of the scheme on Cartesian (or MAC) grids, at order 2 via a MUSCL type method, for the barotropic equations at first and then for the complete equations. The latter require to handle with an additional energy equation and one of the -solved- problems is to find a suitable discrete definition of the total energy such that it satisfies a local conservative equation. In a third chapter we study the transition from the compressible case to the incompressible limit and we shall see how to use the advantages of our initial scheme in order to make it an Asymptotic Preserving scheme at low Mach numbers. In a fourth chapter we propose an adaptation of the scheme on unstructured meshes. Our approach is strongly inspired by the DDFV methods and may have advantages in low-Mach regimes.This thesis ends with a fifth chapter issued from a collaboration during CEMRACS 2017, where the considered point of view is no longer macroscopic but microscopic. We begin by studying a simplified micro/macro model with an added stochastic process and then we attempt to deduce a large-scale model for a strongly coupled system which has to be consistent with the underlying micro / macro description of the physical problem.