L’objectif de cette thèse était la modélisation numérique de la torche à plasma à couplage inductive. (ICP). Nous avons établi les bases de données nécessaires : composition, propriétés thermodynamiques et de transport appliqués aux mélanges d’argon, d’acide nitrique et d’eau. Le modèle hydrodynamique de la torche ICP (7 spires) considère le plasma à l'équilibre thermodynamique et couple les équations de Navier-Stokes pour décrire l'écoulement du plasma aux équations de Maxwell pour décrire l'évolution du champ électrique et du champ magnétique. / The purpose of the work was the numerical modeling of the inductive coupling plasma torch. (ICP). We have established the necessary databases: composition, thermodynamic and transport properties applied to argon mixtures of nitric acid and water. The hydrodynamic model of the ICP torch (7 coils) considers the plasma at thermodynamic equilibrium and uses the Navier-Stokes equations to describe the plasma flow and the Maxwell equations to describe the evolution of the electric field and the magnetic field.

A numerical model for inductive plasma wind tunnels is developed. This model provides the flow conditions at the edge of a boundary layer in front of a thermal protection material placed in the plasma jet stream at the outlet of an inductive torch. The governing equations for the hydrodynamic field are derided from the kinetic theory. The electromagnetic field is deduced from the Maxwell equations. The transport properties of partially ionized and unmagnetized plasma in weak thermal nonequilibrium are derived from the Boltzmann equation. A kinetic data base of transport collision integrals is given for the Martian atmosphere. Multicomponent transport algorithms based upon Krylov subspaces are compared to mixture rules in terms of accuracy and computational cost. The composition and thermodynamic properties in local thermodynamic equilibrium are computed from the semi-classical statistical mechanics. The electromagnetic and hydrodynamic fields of an inductive wind tunnel is presented. A total pressure measurement technique is thoroughly investigated by means of numerical simulations.

La production d'électricité dépend étroitement de la disponibilité d'une source froide. C'est la raison pour laquelle la plupart des centrales de grande puissance dans le monde sont construites près d'une source d'eau. Le problème de la source froide a été soulevé à plusieurs reprises en France, notamment après les canicules de 2003 et de 2006. Le refroidissement à l'air sec est une des options possibles. Cependant, étant donné le besoin de surface d'échange plus important, le changement de la source froide pour l'air ambiant n'est pas, dans la majorité des cas, viable économiquement.Une des solutions à ce problème imaginées à EDF était de changer l'architecture du cycle de production en considérant un cycle de production composé de deux cycles de Rankine en cascade, le premier fonctionnant avec de la vapeur d'eau et le deuxième fonctionnant avec de l'ammoniac dont la vapeur à basse pression est beaucoup plus dense que celle de l'eau. Cette solution permet de faciliter l'utilisation d'un aérocondenseur et de réduire la taille de la salle machine.En raison de la nature toxique et corrosive de l'ammoniac, il est intéressant d'étudier la possibilité de remplacer ce dernier par d'autres fluides plus adaptés, notamment en envisageant de nouveaux fluides pour lesquels peu ou pas de données sont disponibles. Nous comparons les fluides sur le plan énergétique et en terme de taille des composants de l'installation.Cette thèse illustre la démarche des différentes étapes de notre travail : la recherche de nouveaux fluides de travail, l'évaluation de performance du système en régime nominal et non-nominal, le dimensionnement des principaux composants du cycle ainsi que l'évaluation de coût et de gain économique éventuel. / This work considers a two stage Rankine cycle architecture slightly different from a standard Rankine cycle for electricity generation. Instead of expanding the steam to extremely low pressure, the vapor leaves the turbine at a higher pressure then having a much smaller specific volume. It is thus possible to greatly reduce the size of the steam turbine. The remaining energy is recovered by a bottoming cycle using a working fluid which has a much higher density than the water steam. Thus, the turbines and heat exchangers are more compact; the turbine exhaust velocity loss is lower. This configuration enables to largely reduce the global size of the steam water turbine and facilitate the use of a dry cooling system.The main advantage of such an air cooled two stage Rankine cycle is the possibility to choose the installation site of a large or medium power plant without the need of a large and constantly available water source; in addition, as compared to water cooled cycles, the risk regarding future operations is reduced (climate conditions may affect water availability or temperature, and imply changes in the water supply regulatory rules).The concept has been investigated by EDF R&D. A 22 MW prototype was developed in 70s using ammonia as the working fluid of the bottoming cycle for its high density and high latent heat. However, this fluid is toxic. In order to search more suitable working fluids for the two stage Rankine cycle application and to identify the optimal cycle configuration, we have established a working fluid selection methodology. Some potential candidates have been identified. We have evaluated the performances of the two stage Rankine cycles operating with different working fluids in both design and off design conditions. For the most acceptable working fluids, components of the cycle have been sized. The power plant concept can then be evaluated on a life cycle cost basis.

A numerical model for inductive plasma wind tunnels is developed. This model provides the flow conditions at the edge of a boundary layer in front of a thermal protection material placed in the plasma jet stream at the outlet of an inductive torch. The governing equations for the hydrodynamic field are derided from the kinetic theory. The electromagnetic field is deduced from the Maxwell equations. The transport properties of partially ionized and unmagnetized plasma in weak thermal nonequilibrium are derived from the Boltzmann equation. A kinetic data base of transport collision integrals is given for the Martian atmosphere. Multicomponent transport algorithms based upon Krylov subspaces are compared to mixture rules in terms of accuracy and computational cost. The composition and thermodynamic properties in local thermodynamic<p>equilibrium are computed from the semi-classical statistical mechanics.<p>The electromagnetic and hydrodynamic fields of an inductive wind tunnel is presented. A total pressure measurement technique is thoroughly investigated by means of numerical simulations.<p> / Doctorat en sciences appliquées / info:eu-repo/semantics/nonPublished