O carbeto de silício (SiC) é um material que apresenta baixa expansão térmica, altas resistências mecânica e ao choque térmico e alta condutividade térmica. Em razão disto é empregado na confecção de mobília de fornos de sinterização. O SiC no entanto sofre degradação a altas temperaturas quando submetido a atmosferas agressivas. A utilização de recobrimentos protetores evita a exposição direta da superfície do material à atmosfera dos fornos; a mulita pode ser um recobrimento protetor apropriado em razão de sua alta estabilidade em temperaturas elevadas e seu coeficiente de expansão térmica compatível com o do SiC (4x10-6/°C e 5,3x10-6/°C, respectivamente). No presente trabalho foi estudada a obtenção de recobrimento de mulita, para refratário de SiC, a partir da utilização de polímero precursor cerâmico e alumínio particulado. Foram preparadas composições com 10, 20, 30 e 50% (vol.) de alumínio adicionado ao polímero, sendo utilizados pós de alumínio de diferentes distribuições de tamanhos de partículas. As composições foram submetidas a diversos ciclos térmicos para determinação da condição mais adequada à obtenção de alto teor de mulita. A composição que apresentou melhor resultado foi a contendo 20% do pó de Al de menor tamanho de partículas. A partir desta, foi preparada e aplicada suspensão para ser aplicada sobre o refratário de SiC. A suspensão aplicada, após seca, reticulada e tratada termicamente a 1580°C, originou um recobrimento de mulita. Foram realizados ciclos de choque térmico em amostras com e sem recobrimento para comparação, num total de 26 ciclos. As condições foram 600°C/30 min. seguida de resfriamento ao ar até a temperatura ambiente. Após cada choque térmico, as amostras foram caracterizadas por microscopia óptica e eletrônica e determinado o módulo de elasticidade. Os recobrimentos apresentaram boa adesão e não foram detectados danos significativos após os choques térmicos. / Silicon carbide (SiC) presents low thermal expansion, high strength and thermal conductivity. For this reason it is used as kiln furniture for materials sintering. On the other hand, SiC degrades at high temperature under aggressive atmosphere. The use of protective coatings can avoid the right exposition of SiC surface to the furnace atmosphere. Mullite can be a suitable material as protective coating because of its high corrosion resistance and thermal expansion coefficient matching that of SiC (4,7 x10-6/°C e 5,3 x10-6/°C, respectively). In the present work a mullite coating obtained from ceramic precursor polymer and aluminium powder was studied to be applied over SiC refractories. Compositions were prepared with 10, 20, 30 and 50% (vol.) of aluminium powder added to the polymer. They were used aluminium powders with different distributions sizes These compositions were heat treated at different thermal cycles to determine a suitable condition to obtain a high mullite content. The composition with 20% of the smaller particle size Al powder was selected and used to be applied as a suspension over SiC refractory. The applied suspension, after dried, crosslinked and heat treated, formed a mullite coating over SiC refractory. Cycles of thermal shock were performed in coated and uncoated SiC samples to compare each other. They were carried out 26 cycles of thermal shock, in the following conditions: 600°C/30 min. and air cooling to room temperature. After each thermal shock, samples were analised by mean of optical and electron microscopy, elastic modulus was also determined. After thermal shock cycles the coating presented good adhesion and no significant damage were observed.

O carbeto de silício (SiC) é um material que apresenta baixa expansão térmica, altas resistências mecânica e ao choque térmico e alta condutividade térmica. Em razão disto é empregado na confecção de mobília de fornos de sinterização. O SiC no entanto sofre degradação a altas temperaturas quando submetido a atmosferas agressivas. A utilização de recobrimentos protetores evita a exposição direta da superfície do material à atmosfera dos fornos; a mulita pode ser um recobrimento protetor apropriado em razão de sua alta estabilidade em temperaturas elevadas e seu coeficiente de expansão térmica compatível com o do SiC (4x10-6/°C e 5,3x10-6/°C, respectivamente). No presente trabalho foi estudada a obtenção de recobrimento de mulita, para refratário de SiC, a partir da utilização de polímero precursor cerâmico e alumínio particulado. Foram preparadas composições com 10, 20, 30 e 50% (vol.) de alumínio adicionado ao polímero, sendo utilizados pós de alumínio de diferentes distribuições de tamanhos de partículas. As composições foram submetidas a diversos ciclos térmicos para determinação da condição mais adequada à obtenção de alto teor de mulita. A composição que apresentou melhor resultado foi a contendo 20% do pó de Al de menor tamanho de partículas. A partir desta, foi preparada e aplicada suspensão para ser aplicada sobre o refratário de SiC. A suspensão aplicada, após seca, reticulada e tratada termicamente a 1580°C, originou um recobrimento de mulita. Foram realizados ciclos de choque térmico em amostras com e sem recobrimento para comparação, num total de 26 ciclos. As condições foram 600°C/30 min. seguida de resfriamento ao ar até a temperatura ambiente. Após cada choque térmico, as amostras foram caracterizadas por microscopia óptica e eletrônica e determinado o módulo de elasticidade. Os recobrimentos apresentaram boa adesão e não foram detectados danos significativos após os choques térmicos. / Silicon carbide (SiC) presents low thermal expansion, high strength and thermal conductivity. For this reason it is used as kiln furniture for materials sintering. On the other hand, SiC degrades at high temperature under aggressive atmosphere. The use of protective coatings can avoid the right exposition of SiC surface to the furnace atmosphere. Mullite can be a suitable material as protective coating because of its high corrosion resistance and thermal expansion coefficient matching that of SiC (4,7 x10-6/°C e 5,3 x10-6/°C, respectively). In the present work a mullite coating obtained from ceramic precursor polymer and aluminium powder was studied to be applied over SiC refractories. Compositions were prepared with 10, 20, 30 and 50% (vol.) of aluminium powder added to the polymer. They were used aluminium powders with different distributions sizes These compositions were heat treated at different thermal cycles to determine a suitable condition to obtain a high mullite content. The composition with 20% of the smaller particle size Al powder was selected and used to be applied as a suspension over SiC refractory. The applied suspension, after dried, crosslinked and heat treated, formed a mullite coating over SiC refractory. Cycles of thermal shock were performed in coated and uncoated SiC samples to compare each other. They were carried out 26 cycles of thermal shock, in the following conditions: 600°C/30 min. and air cooling to room temperature. After each thermal shock, samples were analised by mean of optical and electron microscopy, elastic modulus was also determined. After thermal shock cycles the coating presented good adhesion and no significant damage were observed.

Les céramiques de type non-oxyde à base de silicium (SiC, Si3N4, Si-C-N) ont été très largement étudiées comme matrice dans le domaine des Composites à Matrices Céramiques (CMCs) en raison de leurs propriétés thermostructurales généralement très supérieures à celles des matériaux plus conventionnels comme les métaux et les céramiques de type oxyde. Ces matériaux proposent par ailleurs des propriétés mécaniques (dureté, résistance au fluage et à la rupture) et une résistance à l’oxydation de premier plan. Comme matrices, ils sont généralement produits en voie gazeuse par la méthode CVI (Chemical Vapor Infiltration). Dans le cadre de la présente thèse, nous nous intéressons à leur élaboration en voie liquide à travers la méthode PDCs (Polymer Derived Ceramics), qui pourra être éventuellement couplée à terme à la méthode CVI, dont la mise en œuvre est plus aisée et les coûts de production des CMCs plus faibles. L’objectif principal est de travailler la chimie de polymères précéramiques commerciaux afin, d’une part, d’optimiser les étapes d’imprégnation des préformes fibreuses et de pyrolyse des composites ‘crus’ obtenus (Polymer Infiltration and Pyrolysis (PIP)) et d’autre part d’améliorer les propriétés thermostructurales des composites SiC et Si-C-N pour un fonctionnement à des températures de l’ordre de 1500°C. Après une étude bibliographique sur la thématique abordée (chapitre 1) et un chapitre 2 dédié à la partie expérimentale et à la description des outils de caractérisation, les travaux de thèse se sont orientés dans les chapitres 3 et 4 vers la modification de polymères précéramiques commerciaux comme l’allyhydridopolycarbosilane (AHPCS, précurseur SiC) et le poly(vinylméthyl)-co-(hydridométhyl)silazane (HTT1800, précurseur Si-C-N) par l’élément bore. L’idée générale est de diminuer les températures de gélification de ces polymères tout en augmentant leur rendement céramique et d’obtenir, après pyrolyse, des céramiques amorphes de type Si-B-C-(N) avec une meilleure stabilité thermique à haute température. Dans le chapitre 5, les travaux se sont dirigés vers la préparation de mélange HTT1800-perhydridopolysilazane (PHPS, précurseur Si3N4) pour s’affranchir de la présence de carbone libre dans les matériaux finaux. Une caractérisation complète, allant de la structure chimique des polymères jusqu’à l’évolution de la microstructure des matériaux finaux traités à haute température, a été conduite dans chacun des chapitres. La fabrication de pièces denses, par la méthode dite de casting, à partir des polymères sélectionnés a permis d’accéder aux propriétés mécaniques des matériaux. Des essais préliminaires de fabrication de composites sont présentés en fin de chaque chapitre. / Non-oxide Si-based ceramics (SiC, Si3N4, Si-C-N) have been extensively studied as matrices in Fiber-Reinforced Ceramics Matrix Composites (CMCs) because of their thermostructural properties which are generally significantly higher than those displaying by more conventional materials such as metals and oxide ceramics. These materials also offer superior mechanical properties (hardness, resistance to creep and rupture) and excellent resistance toward oxidation. As a matrix, they are produced in gas phase by the well-known Chemical Vapor Infiltration (CVI) process. Within the framework of the thesis, we focus on their synthesis in liquid phase through the PDCs (Polymer Derived Ceramics) route because of its easier access and lower production cost. The main objective is to focus on the chemistry of preceramic polymers to 1) optimize each step of the PIP (Infiltration and Pyrolysis Polymer) process and 2) improve the thermostructural properties of SiC, Si-C-N matrix composites prepared from commercially-available preceramic polymers. After a state-of-the art part (Chapter 1) on the targeted topic and an experimental part completed by the description of the characterization tools (Chapter 2), the manuscript focused on the modification of commercial preceramic polymers such as allylhydridopolycarbosilane (AHPCS, SiC precursor) and poly(vinylmethyl)-co-(hydridomethyl)silazane (HTT1800, Si-C-N precursor) with boron elements. The idea behind this work was to reduce the gelification temperature of these polymers while to increase their ceramic yield. Thus, after pyrolysis, we obtain amorphous Si-B-C-(N) ceramics with better thermal stability at high temperature. This work is described in the chapters 3 and 4 of the manuscript. In chapter 5, the work concerned the preparation of polymer blends based on HTT1800 and perhydridopolysilazane (PHPS, precursor Si3N4). The idea behind the chapter 5 was to avoid the presence of free carbon in the final materials. A complete characterization, ranging from the chemical structure of the polymers to the evolution of the microstructure of the final materials, is done in each chapter. Dense pieces were prepared by the casting method from the selected polymers and their mechanical properties have been investigated. Composite materials have been also prepared to evaluate the quality of interface between the matrix and the surface of the fibers which is presented at the end of each chapter.

Le développement de protections balistiques toujours plus légères et performantes reste un sujet de recherche très actif. Malgré de très hautes performances, la difficulté de mise en forme du SiC conduit généralement à l’utilisation d’aides au frittage en quantité importante, favorisant la formation de phases secondaires pouvant fragiliser le matériau. De plus, les hautes températures de mise en forme induisent la présence de phase α, conférant au matériau des propriétés mécaniques anisotropes et inférieures à celles de la phase cubique β.Dans ces travaux de thèse, l’objectif a été d’élaborer un matériau SiC cubique de très haute pureté, avec une densité de 100% et une stœchiométrie Si/C idéale afin d’optimiser les performances de cette céramique. Deux types de précurseurs ont été envisagés : une poudre commerciale et une poudre issue de la conversion d’un précurseur polymère précéramique.Dans un premier temps, une étude paramétrique de frittage par SPS a permis d’atteindre des densités de 95% pour les 2 précurseurs, tout en conservant la phase cubique seule. Ces résultats, bien qu’encourageants mais n’étant pas suffisants pour l’application visée, l’étude s’est tournée vers l’ajout d’aides au frittage. Des densités de 100% ont ainsi été obtenues sur des échantillons préparés à partir de poudre commerciale, même pour de très faibles teneurs en additif. Un second aspect de ces travaux a permis de mettre en évidence une dépendance de la température de transition β -> α du SiC vis-à-vis de la pression de frittage mais également vis-à-vis du type de précurseur, l’utilisation du précurseur polymère étant plus favorable à la stabilité de la structure cubique. Enfin des mesures de dureté ont été réalisées sur les meilleurs échantillons et ont permis de souligner le rôle prépondérant de la densité sur cette propriété. / The development of light and high performance ballistic protections is currently a sensitive subject of research. Despite promising mechanical characteristics, the complexity of SiC shaping generally leads to the use of high content of sintering aids, favouring secondary phases formation which could weaken the material. Nevertheless, high sintering temperatures induce the presence of the α form of SiC, conferring to the material anisotropical and lower mechanical properties than the one obtained with the cubic β phase.The goal of this PhD work is the development of high purity cubic SiC, with density close to 100% and perfect Si:C stoichiometry to optimize the performances of this ceramic. Two kinds of precursors were considered: a commercial powder and a powder from the conversion of preceramic polymer precursor.Firstly, the parametric study of SPS sintering allowed to reach densities of 95% for both precursors, while conserving only the cubic phase. These encouraging results being not sufficient, this study switched to the use of sintering aids. Densities close to 100% were thus reached on samples sintered with prepared mixtures from commercial powder, even for very low content of additive. The second subject of this thesis highlighted a dependence of the β -> α transition temperature of SiC as a function of sintering pressure, but also according to the kind of precursor. Indeed, the use of polymer precursor is favourable to cubic structure stability. Then, hardness measurements were performed on the most promising samples and allowed to highlight the major role of density on this property.