Ce travail de thèse traite des modifications de l’écoulement et des transferts de chaleur en convection naturelle. La configuration retenue est celle de la cavité différentiellement chauffée de rapport de forme 4. Les modifications sont réalisées au moyen d’une perturbation locale de faible étendue spatiale positionnée sur la paroi chaude pour une valeur du nombre de Raleigh se situant autour de la valeur critique correspondant à la transition d’un état stationnaire à une état instationnaire. Deux études complémentaires ont été menées. Une étude numérique préliminaire a été réalisée, dans laquelle une perturbation d’origine thermique a été introduite sous la forme d’un changement local de la température imposé sur la paroi isotherme. Il a été montré que l’on peut changer le régime d’écoulement depuis un état stationnaire vers un état instationnaire. De plus, les transferts de chaleur peuvent être modifiés de façon significative par un choix approprié des paramètres de la perturbation (fréquence, amplitude et position). Une étude complémentaire sur un dispositif expérimental a été conduite avec le positionnement d’un obstacle cylindrique sur la paroi chaude. Deux types d’obstacles ont été choisis : isolant et conducteur. L’impact sur la thermique et la dynamique de la longueur de l’obstacle,de sa position verticale, du nombre de Rayleigh de l’écoulement ou encore du nombre d’obstacles introduits a été analysé dans le plan vertical à mi- profondeur. / This thesis deals with the modifications of heat transfer and fluid flow in natural convection. The configuration chosen is a differentially heated cavity of aspect ratio equal to 4. The modifications are achieved using a small spatial extent disturbance located at the hot wall at a Rayleigh number value close of the critical value corresponding to the transition from steady to unsteady state. Two complementary studies were conducted. A preliminary numerical study was carried out, in which a thermal disturbance was introduced by means of a local change of the imposed temperature at the isothermal wall. It was shown that the flow regime can be changed from a steady state to an unsteady one. In addition, the heat transfers can be significantly modified by an appropriate choice of disturbance parameters (frequency, magnitude and location). An additional study, using an experimental device, was conducted by positioning a cylindrical obstacle on the hotwall. Two types of obstacles were chosen : insulating and conductive. The influence on heat transfers of the length of the obstacle, its vertical location, the Rayleigh number ofthe flow as well as the number of obstacles was analyzed in the vertical mid-depth plane.

This work treats the thermal and mechanical performances of gas-filled, flat plate solar collectors in order to achieve a better performance than that of air filled collectors. The gases examined are argon, krypton and xenon which all have lower thermal conductivity than air. The absorber is formed as a tray connected to the glass. The pressure of the gas inside is near to the ambient and since the gas volume will vary as the temperature changes, there are potential risks for fatigue in the material. One heat transfer model and one mechanical model were built. The mechanical model gave stresses and information on the movements. The factors of safety were calculated from the stresses, and the movements were used as input for the heat transfer model where the thermal performance was calculated. It is shown that gas-filled, flat plate solar collectors can be designed to achieve good thermal performance at a competitive cost. The best yield is achieved with a xenon gas filling together with a normal thick absorber, where normal thick means a 0.25 mm copper absorber. However, a great deal of energy is needed to produce the xenon gas, and if this aspect is taken into account, the krypton filling is better. Good thermal performance can also be achieved using less material; a collector with a 0.1 mm thick copper absorber and the third best gas, which is argon, still gives a better operating performance than a common, commercially produced, air filled collector with a 0.25 mm absorber. When manufacturing gas-filled flat plate solar collectors, one way of decreasing the total material costs significantly, is by changing absorber material from copper to aluminium. Best yield per monetary outlay is given by a thin (0.3 mm) alu-minium absorber with an argon filling. A high factor of safety is achieved with thin absorbers, large absorber areas, rectangular constructions with long tubes and short distances between glass and absorber. The latter will also give a thin layer of gas which gives good thermal performance. The only doubtii ful construction is an argon filled collector with a normal thick (> 0.50 mm) aluminium absorber. In general, an assessment of the stresses for the proposed construction together with appropriate tests are recommended before manufacturing, since it is hard to predict the factor of safety; if one part is reinforced, some other parts can experience more stress and the factor of safety actually drops.

Les écoulements de convection naturelle en espace confiné sont généralement turbulents (grandes dimensions et/ou écarts de température importants). L'approfondissement de la connaissance de ce régime et des transferts qui y sont associés semble donc essentielle. Cette étude expérimentale concerne les mécanismes de transferts thermiques turbulents en s'appuyant sur des mesures couplées vitesse/température dans un écoulement de convection naturelle turbulente à haut nombre de Rayleigh (Ra1-1= 1,2x1011) au sein d'une cavité différentiellement chauffée de rapport de forme vertical (hauteur/largeur) égale à 4. Une chaîne de mesure PIV permet les acquisitions de vitesse tandis que la température est mesurée par micro-thermocouple de type K (0=12,7 μm). Les deux mesures étant synchronisées via un générateur de pulses type BNC®. Une attention particulière a été portée à la détermination des conditions aux limites en température des parois adiabatiques et au post traitement des données de PIV à l'aide d'une décomposition orthogonale basée sur l'énergie cinétique des champs de vitesse (POD). Cela a rendu possible, d'une part le calcul expérimental de quantités liées à la turbulence comme les flux d'enthalpie et des nombres adimensionnés pour la turbulence (Prandtl, diffusivité ... ) et, d 'autre part, l'évaluation de la contribution des autres termes des équations de Navier-Stokes (gradient de pression et terme source volumique). Enfin, une discussion sur les échelles caractéristiques de l'écoulement est menée et une comparaison avec des résultats de simulations numériques est apportée. / Natural convection flows in confined spaces are often turbulent (large dimensions and/or temperature difference ). Improving the knowledge of this type of regime and of the associated heat and mass transfers seems, therefore, to be essential. This experirnental work studies heat transfer mechanisms using coupled and synchronized measurernents of temperature and velocity in a turbulent natural convection flow at high Rayleigh number (Ra1-1=I.2 x1011) in a differentially heated cavity with a vertical aspect ratio (height/width) of 4. Velocity measurernents are acquired by PIV and temperature is measured with K-type micro-thermocouple (0=12.7 μm). Both measurementsystems are synchronized using a BNC® pulse generator. Thermal boundary conditions of the adiabatic walls and post treatment of PIV data using an orthogonal decomposition based on the kinetic energy of the flow - POD - have been considered with caution. It makes possible, on the one hand, the experimental assessment of turbulent quantities like enthalpy fluxes and dimensionless nurnber (Prandtl, diffusivity ... ) and, on the other hand the estimation of the contribution of the other terms of Navier-Stokes equations (pressure gradient and source term). Finally a discussion on characteristic scales of the flow and a comparison with numerical simulations are provided.

The simulation of physical systems in the early stages of conceptual designs has shown to be a key factor for adequate decision making and avoiding big and expensive issues downstream in engineering projects. In the case of aircraft cabin design, taking into account the thermal comfort of the passengers as well as the proper air circulation and renovation can make this difference. However, current numerical fluid simulations (CFD) are too computationally expensive for integrating them in early design stages where extensive comparative studies have to be performed. Instead, Zonal Models (ZM) appear to be a fast-computation approach that can provide coarse simulations for aircraft cabin flows. In this thesis, a Zonal Model solver is developed as well as a geometry-definition and meshing framework, both in Matlab®, for performing coarse, flexible and computationally cheap flow simulations of user-defined cabin designs. On one hand, this solver consists of a Fractional Step approach for coarse unstructured bi-dimensional meshes. On the other, the cabin geometry can be introduced by hand for simple shapes, but also with Computational Aided Design tools (CAD) for more complex designs. Additionally, it can be chosen to generate the meshes from scratch or morph them from previously generated ones. / <p>The presentation was online</p>