No description available.

Boron carbide is produced in a heat resistance furnace using boric oxide and petroleum coke as the raw materials. In this process, a large current is passed through the graphite rod located at the center of the cylindrical furnace, which is surrounded by the coke and boron oxide mixture. Heat generated due to resistance heating is responsible for the reaction of boron oxide with coke which results in the formation of boron carbide. The whole process is highly energy intensive and inefficient in terms of the production of boron carbide. Only 15% charge gets converted into boron carbide. The aim of the present work is to develop a mathematical model for this batch process and validate the model with experiments and to optimize the operating parameters to increase the productivity. To mathematically model the process and understand the influence of various operating parameters on the productivity, existing simple one-dimensional (1-D) mathematical model in radial direction is modified first. Two-dimensional (2-D) model can represent the process better; therefore in second stage of the project a 2-D mathematical model is also developed. For both, 1-D and 2-D models, coupled heat and mass balance equations are solved using finite volume technique. Both the models have been tested for time step and grid size independency. The kinetics of the reaction is represented using nucleation growth mechanism. Conduction, convection and radiation terms are considered in the formulation of heat transfer equation. Fraction of boron carbide formed and temperature profiles in the radial direction are obtained computationally. Experiments were conducted on a previously developed experimental setup consisting of heat resistance furnace, a power supply unit and electrode cooling device. The heating furnace is made of stainless steel body with high temperature ceramic wool insulation. In order to validate the mathematical model, experiments are performed in various conditions. Temperatures are measured at various locations in the furnace and samples are collected from the various locations (both in radial and angular directions) in the furnace for chemical analysis. Also, many experimental data are used from the previous work to validate the computed results. For temperatures measurement, pyrometer, C, B and K type thermocouple were used. It is observed that results obtained from both the models (1-D and 2-D) are in reasonable agreement with the experimental results. Once the models are validated with the experiments, sensitivity analysis of various parameters such as power supply, initial percentage of B4C in the charge, composition of the charge, and various modes of power supply, on the process is performed. It is found that linear power supply mode, presence of B4C in the initial mixture and increase in power supply give better productivity (fraction reacted). In order to have more confidence in the developed models, the parameters of one the computed results in the sensitivity analysis parameters are chosen (in present case, linear power supply is chosen) to perform the experiment. Results obtained from the experiment performed under the same simulated conditions as computed results are found in excellent match with each other.

Le but de cette thèse était de modéliser la réponse mécanique de nanotubes de carbone à des champs électriques. Nous avons commencé par utiliser le potentiel AIREBO dans des simulations de dynamique moléculaire afin d'étudier l'élasticité non-linéaire et la limite de déformation en torsion de divers nanotubes, en fonction de leur longueur, rayon et chiralité. Nous trouvons notamment que le module d'Young effectif des tubes décroît d'autant plus vite que la chiralité est faible. D'autre part, nous montrons que la limite de l'énergie stockable par atome lors de la torsion d'un tube est d'autant plus grande que le diamètre est petit.<br><br>Nous modélisons ensuite, de façon atomistique, la distribution surfacique de charge électrique sur des nanotubes de carbone possédant une charge nette. Nous retrouvons notamment l'effet de pointe classique avec un très bon accord quantitatif avec des résultats expérimentaux obtenus par microscopie à force électrostatique.<br><br>Par combinaison des méthodes utilisées dans les études précédentes, nous simulons la déflection de nanotubes semi-conducteurs et métalliques par un champ électrique extérieur, dans une configuration de type interrupteur moléculaire. L'effet des caractéristiques géométriques des tubes et du champ sur cette déflection ont été systématiquement étudiés.<br><br>En outre, nous avons vu que des simulations de dynamique moléculaire avec le potentiel AIREBO permettent de retrouver quantitativement les énergies expérimentales d'adsorption du benzène, du naphtalène et d'anthracène sur le graphite. Ce type de simulation nous permet d'avancer sur la voie de la compréhension de la sélectivité de l'adsorption de certaines molécules surfactantes à plusieurs cycles benzéniques sur des nanotubes de chiralité donnée.