This dissertation examines the academic and social goals that three second-grade English language learners in a U.S. urban school bring to their blog-mediated academic writing practices, and the interrelated nature of those goals. This study aims to bridge the dichotomy between approaches to studying computer-mediated language and literacy development that are oriented toward academic goals inside school, and those that are oriented toward social goals outside school. The study also aims to investigate connections between language use and language development by highlighting linguistic features of semiotic choices that the students made for their texts. This builds upon recent research studies of literacy practices that focus only on situated uses of literacy in various social and cultural contexts (Christie & Martin, 2007). In this study, learning is defined as appropriation and language is defined as a semiotic system, from sociocultural perspectives that capture the transformative nature of tool-mediated practices (Bakhtin, 1981; Halliday, 1985; Kress, 1998; Vygotsky, 1978). Ethnographic data collected over the course of a year include students’ texts, blog comments, videotaped classroom interactions, interviews, instructional materials, and school documents. Analysis of the data examines student goals, semiotic choices employed by the students, and roles adopted by the students, in the social processes of learning academic genres. Systemic functional linguistics is used to analyze register variables across texts and blogging comments, to examine changes in the students’ uses of linguistic resources. The findings demonstrate that students appropriate blogging for both academic and social goals, and compose their texts by drawing on linguistic features appropriate for goals related to the audiences reading their blog posts. Writing for meaningful goals and for wider audiences encourages ELLs to become more invested in learning, and to use linguistic patterns in context-dependent ways. The study concludes with a discussion of the significance of social goals in developing critical academic literacies (Gebhard, Harman, & Seger, 2007), and implications for K-12 educators who are attempting to open up curricular spaces in which all stakeholders collaboratively work toward transformative learning experiences for ELLs (Willett & Rosenberger, 2005).

Two new homologous series of amphiphiles–five long-chain, three-headed amphiphiles [3CCb14, 3CCb16, 3CCb18, 3CCb20, 3CCb22; CH3(CH2)n-1OCONHC(CH2CH2COOH)3, n = 14, 16, 18, 20, 22], and six branched-chain, three-headed amphiphiles [3CCb1(7,7), 3CCb1(8,8), 3CCb1(9,9), 3CCb1(10,10), 3CCb1(11,11), 3CCb1(12,12); (CH3(CH2)n-1)2CHOCONHC(CH2CH2COOH)3, n = 7, 8, 9, 10, 11, 12]–were synthesized. The synthesis of the 3CCbn series was accomplished in two steps from Weisocyanate™ and the long chain alcohol in good yields of chromatographed products (65–81%). The 3CCb1(n,n) series was similarly synthesized from Weisocyanate™ and the two-tailed symmetric alcohol (produced from a reaction of alkyl magnesium bromide and ethyl formate) in good yields of chromatographed products (71–84%). CMC data were collected by pendent-drop technique for the 3CAmn, 3CCbn, 3CUrn, and 3CCb1(n,n) series of amphiphiles to establish the concentration required for detergency. The triethanolammonium salt provided better solubility and higher CMCs of these amphiphiles than the potassium salt. All amphiphilic series tested lowered the solution surface tension from ~ 72 mN/m to ~ 50–55 mN/m, indicating that these amphiphiles are less surface active than typical surfactants such as sodium dodecyl sulfate. The CMCs for the 3CAmn series were found to decrease in value from 2 × 10⁻² M (3CAm15) to 2 × 10⁻³ M (3CAm21) in a linear fashion. The CMCs for the 3CCbn series were found to decrease in value from 7 × 10⁻³ M (3CCb16) to 0.4 × 10⁻³ M (3CCb22) in a linear fashion. The CMCs for the 3CUrn series were found to decrease in value from 2 × 10⁻³ M (3CUr18) to 1 × 10⁻³ M (3CUr22) in a linear fashion. Due to discrepancies in several of the IFT vs. log concentration plots for the previous homologous series of amphiphiles, the CMC data was collected using a pyrene fluorescence measurement technique. The data from the pyrene fluorescence technique seems likely to be more accurate, indicating that surface tension may not be the most reliable method for determining the CMC of these amphiphiles. The CMCs (as determined by pyrene fluorescence) for the 3CAmn series were found to decrease in value from 2 × 10⁻² M (3CAm15) to 2 × 10⁻³ M (3CAm21) in a linear fashion. The CMCs for the 3CCbn series were found to decrease in value from 7 × 10⁻³ M (3CCb16) to 0.3 × 10⁻³ M (3CCb22) in a linear fashion. The CMCs for the 3CUrn series were found to decrease in value from 7 × 10⁻³ M (3CUr16) to 0.2 × 10⁻³ M (3CUr22) in a linear fashion. In both the surface tension and the pyrene fluorescence techniques, the shortest chain length homologues (3CAm13, 3CCb14, and 3CUr14) did not show a break up to the limits of solubility. The CMCs as determined by surface tension for the 3CCb1(n,n) series were found to decrease in value from 0.5 × 10⁻³ M (3CCb1(9,9)) to 0.02 × 10⁻³ M (3CCb1(12,12)) in a linear fashion. The 3CCb1(8,8) and 3CCb1(7,7) amphiphiles did not show a CMC break up to the limits of solubility. The 3CCb1(12,12) showed an unusually steep decrease in surface tension over a very narrow range of concentration. There is considerable doubt as to the accuracy of the 3CCb1(11,11) data, and the CMCs for these two-tailed amphiphiles needs to be measured by a second method as was done for the single-tail series to verify the CMCs of all the two-tail homologues. Activity (minimal inhibitory concentrations, MICs) for the 3CAmn, 3CCbn, 3CUrn, 3CCb1(n,n), 2CAmn, and 2CCbn series was measured against several different bacteria, mycobacteria, yeast, and fungi. Additionally, anti-HIV and cytotoxicity data was collected for the 3CAmn, 3CCbn, and 3CUrn series. Greatest inhibition was typically seen from the 18–20 carbon tail length homologues of each series (3CAm19–3CAm21, 3CCb18–3CCb20, 3CUr18–3CUr20, 2CAm19–2CAm21, and 2CCb18–2CCb20). Inoculum density affected the activity of our earlier studies, and selected organisms were retested to obtain the intrinsic activity. 3CUr18 and 3CAm19 proved most effective against Mycobacterium smegmatis, with MIC99 = 6.3 μM @ 10⁵ CFU/mL inoculum density. 3CCb20 was most effective against Mycobacterium marinum with MIC99 = 16 μM @ 10⁵ CFU/mL inoculum density. 3CAm19, 3CCb18, and 3CUr18 all showed equivalent activity against Mycobacterium chelonae with MIC99 = 17 μM @ 10⁵ CFU/mL inoculum density. Against Staphylococcus aureus, the 2CAm21 was most effective, with MIC90 = 2.0 μM @ 10⁵ CFU/mL inoculum density. 3CCb20 was most effective against MRSA with MIC90 = 2.9 μM @ 10⁵ CFU/mL inoculum density. The two-tailed analogs (3CCb1(n,n), 3CUr(n,n), and 3CUr1(n,n)) typically showed little to no activity against the tested microorganisms. Comparison of MIC to CMC is a relative measure of safety of a drug candidate. All single-tail amphiphiles showed ratios of MIC/CMC of 16–126, with a ratio of 100 or better being optimal. The ratios for the two-tail amphiphiles ranged from 0.39 to 2.9. / Ph. D.

This study examines collective identity, a concept that is used in social movement theory to understand why people are motivated to participate in social movements and social movement groups. Collective identity is a social-psychological process that links the individual to the group through a series of group interactions that revolve around social movement activity. This is a qualitative study that examines collective identity in an online social movement group and an offline social movement group. Reports from the two groups are compared to see what variation exists between these two different groups. This research is one of the first examinations of collective identity outside of conventional face-to-face group settings. The research presented in this thesis demonstrates the difficulty a social movement group that exists online might have in generating a collective identity because of an absence of face-to-face interaction. / Master of Science

[ES] El cambio de co'digo esta' sujeto a una gran variedad de factores que dependen del medio de comunicacio'n y de la situacio'n comunicativa. En general, segu'n la lingu¿i'stica, el cambio de co'digo ocurre cuando un hablante alterna entre dos o ma's lenguas o variedades de una lengua en una misma conversacio'n. Estas pra'cticas comunicativas se han tratado en muchos contextos, lenguas y contacto entre culturas. Sin embargo, no hay estudios sobre el tema en el contexto eslovaco, de ahi' la relevancia del trabajo recogido en esta tesis doctoral que tiene como fin u'ltimo paliar esta escasez de estudios lingu¿i'sticos. En concreto, aqui' analizamos el cambio de co'digo exclusivamente en comunicaciones realizadas por correo electro'nico en un entorno laboral multilingu¿e y multicultural. El estudio se centra en la eleccio'n de lengua, las actitudes hacia una lengua y la mezcla de lenguas en las comunicaciones entre los compan¿eros de trabajo de una multinacional hotelera radicada en Eslovaquia. El ana'lisis examina u'nicamente los correos escritos en eslovaco que presentan cambios de co'digo al ingle's. Debido a la naturaleza multidisciplinar de este proyecto de investigacio'n, asi' como a su cara'cter dual, es decir, el examen de las actitudes hacia una lengua por una parte y las pra'cticas de cambio de co'digo por otra, esta tesis plantea varias preguntas de investigacio'n y tiene una serie de objetivos que pasamos a detallar. El objetivo principal del estudio cuantitativo basado en el cuestionario disen¿ado es examinar hasta que' punto los encuestados son conscientes de los cambios de co'digo al ingle's durante sus comunicaciones, en particular durante sus conversaciones electro'nicas, y determinar las razones por las que se lleva a cabo este cambio de co'digo, sacando a la luz las actitudes hacia este feno'meno lingu¿i'stico. Adema's, dado que los conocimientos que aportan los cuestionarios son generalmente limitados, se ha llevado a cabo un ana'lisis del discurso para observar ma's detalladamente el alcance de los cambios de lengua, los tipos y las funciones que presentan. La metodologi'a empleada, que sigue el me'todo mixto de investigacio'n, se utiliza para analizar las motivaciones y las razones por las que nuestros participantes prefieren usar el ingle's en lugar de su lengua nativa. Asi' pues, esta tesis doctoral recoge el primer ana'lisis completo de este tipo sobre el cambio de co'digo eslovaco/ingle's en la comunicacio'n electro'nica que examina conversaciones aute'nticas por Internet dentro de una corporacio'n. Palabras clave: cambio de co'digo, comunicacio'n electro'nica, correo electro'nico, comunicacio'n en entornos laborales, actitudes hacia la lengua / [CA] El canvi de codi esta¿ subjecte a una gran varietat de factors que depenen del mitja¿ de comunicacio' i de la situacio' comunicativa. En general, segons la lingu¿i'stica, el canvi de codi ocorre quan un parlant alterna entre dues o me's llengu¿es o varietats d'una llengua en una mateixa conversa. Aquestes pra¿ctiques comunicatives han estat tractades en molts contexts, llengu¿es i contacte entre cultures. No obstant aixo¿, no hi ha estudis sobre el tema en el context eslovac, d'aqui' la relleva¿ncia del treball recollit en aquesta tesi doctoral que te' com a finalitat u'ltima pal.liar l'escassetat d'estudis lingu¿i'stics sobre el tema. En concret, aci' analitzem el canvi de codi exclusivament en comunicacions realitzades per correu electro¿nic en un entorn laboral multilingu¿e i multicultural. L'estudi se centra en l'eleccio' de llengua, les actituds cap a una llengua i la mescla de llengu¿es en les comunicacions entre els companys de treball d'una multinacional hotelera radicada a Eslova¿quia. L'ana¿lisi examina u'nicament els correus escrits en eslovac que presenten canvis de codi a l'angle's. A causa de la naturalesa multidisciplina¿ria d'aquest projecte d'investigacio', aixi' com al seu cara¿cter dual, e's a dir, l'examen de les actituds cap a una llengua per una part i les pra¿ctiques de canvi de codi per altra, aquesta tesi planteja diverses preguntes d'investigacio' i te' una se¿rie d'objectius que detallarem a continuacio'. L'objectiu principal de l'estudi quantitatiu basat en el qu¿estionari dissenyat e's examinar fins a quin punt les persones enquestades so'n conscients del canvi de codi a l'angle's durant les seues comunicacions, en particular durant les seues converses electro¿niques, i determinar les raons per les quals es duu a terme aquest canvi de codi, traient a la llum les actituds cap a aquest fenomen lingu¿i'stic. A me's, ate's que els coneixements que aporten els qu¿estionaris so'n generalment limitats, s'ha realitzat una ana¿lisi del discurs per a observar me's detalladament l'abast dels canvis de llengua, els tipus i les funcions que representen. La metodologia emprada, que segueix el me¿tode mixt d'investigacio', s'utilitza per a analitzar les motivacions i les raons per les quals els nostres participants prefereixen fer u's de l'angle's en comptes de la seua llengua nativa. Per tant, aquesta tesi doctoral recull la primera ana¿lisi completa d'aquest tipus sobre el canvi de codi eslovac/angle's en la comunicacio' electro¿nica que examina converses aute¿ntiques per Internet dins d'una corporacio'. Paraules clau: canvi de codi, comunicacio' electro¿nica, correu electro¿nic, comunicacio' en entorns laborals, actituds cap a la llengua. / [EN] Code-switching (CS) is subject to the wide range of interrelations between medium and situation factors. Generally, from a linguistic point of view, CS occurs when a speaker alternates between two or more languages, or language varieties, in the course of a single conversation. The practice has been noticed all around the world in many contexts, language and culture contact situations. Hence, based on earlier studies of CS phenomenon, but shifting towards a more specific environment, the workplace, the present study aims to fill a considerable gap in scholarly knowledge about the online/ written CS practices of Slovak native speakers in the context of workplace email communication. Therefore, the study focuses on language choice, language attitudes and CS practices among colleagues in a multilingual workplace environment of a multinational hospitality company in Slovakia, focusing solely on the participants' workplace interactions, in particular their email messages written in Slovak (the national language) with switches to English. Due to the interdisciplinary nature of this research project, as well as its dual focus on language attitudes on the one hand and actual CS practices on the other, this thesis addresses a number of research questions and provides a series of analyses centring around the following objectives. The main focus of the quantitative, questionnaire- based study is to examine the participants' metalinguistic awareness of the extent of switching to English during their communication (particularly focusing on their CMC interactions) and to determine their reasons for doing so, while uncovering the attitudes they hold towards this phenomenon. Furthermore, as the depth of knowledge obtained through questionnaire survey is limited, the corpus analysis of email interactions is conducted in order to investigate more closely the extent of switching and the types, forms and functions of CS involved. Employing a mixed method approach in the process, motivations and reasons why our participants choose English over their native language are examined. Hence, the study represents the first comprehensive analysis of its kind on Slovak-English CS in CMC using authentic naturally-occurring computer-mediated corporate interactions. Keywords: code-switching, CMC, email, workplace communication, attitudes / The traineeship was partly funded by Erasmus+ grant, for which I am grateful. / Lengyelová, A. (2019). Associates or zamestnanci? Language choice, attitudes and code-switching practices: The case of workplace email communication in Slovakia [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/124352

<p>Syftet med uppsatsen är att undersöka om det finns negativa konsekvenser av att använda sig för mycket av email istället för av andra typer av kommunikation inom organisationen.</p><p>Det skickas stora mängder email varje dag inom organisationer, men många organisationer saknar policy för att styra denna kommunikation. Därför genomförs denna undersökning av emailvanor och hur email upplevs på Sörman Information AB.</p><p>För att besvara syftet har en kvalitativ datainsamlingsmetod använts. Vi har genomfört tio intervjuer på en av avdelningarna på fallföretaget Sörman Information AB.</p><p>Undersökningen har visat att email är den vanligaste kommunikationsvägen för arbetsrelaterad kommunikation mellan anställda på det undersökta företaget.</p><p>Slutsatsen blir därmed att email till viss del har negativa konsekvenser för en organisation men de flesta problem går att minimera genom att behandla email på rätt sätt. Som ett hjälpmedel för detta har vi utformat ett utkast till en emailpolicy.</p> / <p>The purpose of this thesis is to investigate the possibility of negative consequences caused by over use of email instead of other means of communication in an organization.</p><p>A very large amount of emails are sent every day within organizations. Most organizations lack a proper email policy to govern this communication. Hence this study is conducted to investigate the email habits and the general perception of email use of the employees at Sörman Information AB.</p><p>The investigation has shown that email is the most preferred choice of communication regarding work related communication between colleagues within the studied organization. T</p><p>hus the conclusion is that email to some degree can cause negative consequences for an organization. However most of these problems can be minimized by correct email use. As a complementing tool for this we have developed a first draft of an email policy.</p>

L’industrie aéronautique a exprimé un besoin en matériaux pour des zones fortement chargées thermiquement et mécaniquement. L’objectif est d’optimiser grâce à ces matériaux, notamment en termes de poids et de rendement, le coeur des turboréacteurs. Un procédé par voie liquide appelé «Reactive Melt infiltration» est industriellement envisagé pour fabriquer ces matériaux. La densification est tributaire de la compétition entre la montée capillaire et la réaction chimique entre le silicium liquide et la poudre préalablement introduite. Cette concurrence peut conduire à des phénomènes de «choking off» qui doivent être évités. Dans ce travail, l’approche numérique est à la fois macroscopique et microscopique. Des validations numériques, des applications dans le cadre de géométries modèles et des analyses physiques sont présentées pour les deux échelles. Les simulations d’imbibition réactive à l’échelle de la pièce sont effectués avec un outil développé au sein de SAFRAN alors que les simulations de montées capillaires dans des géométries modèles à l’échelle du pore sont réalisées avec le code de calcul Thétis (développé à l’I2M, Bordeaux). Une méthodologie expérimentale pour le suivi et l’analyse de l’imprégnation capillaire réactive d’un milieu granulaire constitué de matériaux modèles est également présentée. / The development of ceramic matrix composites (CMCs) to replace certain metal components ininternal hot parts of aircraft engines is an active research field for the aeronautical industry. Theseadvanced components may be manufactured with a fluid processing called ReactiveMelt Infiltration(RMI). The densification step is the capillary rise of a molten metal such as silicon within a wovenpreform. The molten metal can react with the previously introduced ceramic powder and lead to arefractory matrix. The competition between capillary rise and the reaction between molten siliconand introduced carbon powder has to be managed. The ambition is to prevent choking off effectsand closed pores created by preferential paths. Simulations of two-phase flows at macroscopic scaleare undertaken with a homemade code developed in SAFRAN whereas simulations of two-phaseflows atmicroscopic scale were undertaken with CFD code Thétis (developed at I2M, Bordeaux). Anexperimental methodology for analysis of reactive or not reactive capillary impregnation in granularmedia is described.

Les composites à matrice céramique (CMC) sont destinés à remplacer les superalliages comme constituants de l'architecture des chambres de combustion des moteurs aéronautiques. Dans ces conditions, la durée de vie de ces matériaux diminue fortement du fait de leurs dégradations par oxydation. Pour pallier à ce problème, des CMC à matrice autocicatrisante sont élaborés. Ils possèdent la particularité de s'auto-protéger vis-à-vis de l'oxydation par la formation d'oxyde passivant limitant la diffusion des espèces oxydantes au sein des fissures matricielles. Dans le cadre de ces travaux de thèse, la durabilité d'un composite SiC/[Si-B-C] est évaluée. Son comportement en oxydation/corrosion est alors étudié entre 450 et 1000°C sous air à des pressions partielles d'humidité variables. Une approche multi-échelle (échelle constituants et composite) est envisagée pour comprendre les différents mécanismes mis en jeu lors de la non-cicatrisation/cicatrisation du matériau. / Ceramic matrix composites are potential candidate to replace the nickel-based alloys in advanced aeronautic engines as in civilian ones. These composites display cracks due to their elaboration process but also due to mechanical loading in use. These matrix cracks become an extended network for oxygen diffusion, and cause the premature damage of the material. To avoid this process, composites with a sequenced self-healing matrix have been developed and investigated. To resume, the self-healing process consists to consume part of incoming oxygen by sealing the matrix cracks with an oxide phase. In this work, a SiC/[Si-B-C] composite is investigated and a multi-scale approach is used. In a first time, the oxidation rate of each constituent is evaluated and in a second, composite specimens are aged between 450 and 1000°C in different gas mixtures and total pressures. This approach is a good way to better understand the contribution of each element under oxidizing environment and thus the self-healing process.

How do people make use of emojis in their daily computer-mediated communication? In order to obtain data, a sampling methodology was implemented in which data from 15 female and 15 male participants was used to provide reports on the differences in preference and functions of emoji usage across genders. The study found that in variety and in total, males tend to select a greater amount of emojis, in comparison to females. The participants’ reasons for using emojis in instant messages varied across the two gender groups, females focused on illustrating a state of emotion while the males focused on emphasizing a message. This partially corresponded with previous studies. The study also found that the males replace words with emojis because it is easy, fun or quick, and in order to illustrate a state of emotion, while females do not prefer to replace words with emojis. In accordance with previous findings, this study shows that in general, people prefer using emojis to complement rather than to replace words. All of the males in this study tend to make use of emojis in order to complement words, while there was a low amount of females who do not use emojis for the purpose of complementing.

Syftet med uppsatsen är att undersöka om det finns negativa konsekvenser av att använda sig för mycket av email istället för av andra typer av kommunikation inom organisationen. Det skickas stora mängder email varje dag inom organisationer, men många organisationer saknar policy för att styra denna kommunikation. Därför genomförs denna undersökning av emailvanor och hur email upplevs på Sörman Information AB. För att besvara syftet har en kvalitativ datainsamlingsmetod använts. Vi har genomfört tio intervjuer på en av avdelningarna på fallföretaget Sörman Information AB. Undersökningen har visat att email är den vanligaste kommunikationsvägen för arbetsrelaterad kommunikation mellan anställda på det undersökta företaget. Slutsatsen blir därmed att email till viss del har negativa konsekvenser för en organisation men de flesta problem går att minimera genom att behandla email på rätt sätt. Som ett hjälpmedel för detta har vi utformat ett utkast till en emailpolicy. / The purpose of this thesis is to investigate the possibility of negative consequences caused by over use of email instead of other means of communication in an organization. A very large amount of emails are sent every day within organizations. Most organizations lack a proper email policy to govern this communication. Hence this study is conducted to investigate the email habits and the general perception of email use of the employees at Sörman Information AB. The investigation has shown that email is the most preferred choice of communication regarding work related communication between colleagues within the studied organization. T hus the conclusion is that email to some degree can cause negative consequences for an organization. However most of these problems can be minimized by correct email use. As a complementing tool for this we have developed a first draft of an email policy.

L’utilisation des composites à matrice céramiques (CMC), et particulièrement les matériaux SiC/SiC, à la place des alliages métalliques dans les moteurs aéronautiques peut permettre de réduire leur consommation. Ils possèdent en effet des propriétés physiques et mécaniques très intéressantes à haute température : faible densité, résistance élevée aux chocs thermiques et rupture non-fragile. Dans ces matériaux, une fine couche est insérée entre les fibres et la matrice : l’interphase. Le nitrure de bore pyrolytique est le matériau d’interphase de choix pour les applications visées. La bibliographie souligne bien la nécessité et la difficulté de « contrôler » l’intensité des liaisons interfaciales fibres/matrice (F/M) grâce à l’interphase. Mais l’influence exacte de la cristallinité et de l’épaisseur des interphases de type BN sur son contrôle, et par conséquent sur le comportement mécanique final du CMC industriel est encore mal connue.Une première problématique abordée dans cette thèse est l’échelle du CMC de laboratoire à utiliser. En effet jusqu’ici, les matériaux modèles les plus couramment employés sont les mini- et les micro- composites 1D. Ils peuvent être élaborés facilement et rapidement par dépôt chimique en phase vapeur mais ne rendent pas compte de phénomènes inévitablement présents au sein du composite industriel. C’est pourquoi l’utilisation de nouveaux matériaux modèles 2D comme les « monostrates » comprenant un pli de tissu, l’interphase BN et une matrice de SiC dont la porosité peut être comblée par du Si comme dans le cas des CMC industriels est plus pertinente et est proposée. Cependant, de par la faible épaisseur des éprouvettes, les protocoles de caractérisation et de tests mécaniques ont dû être revus. Il s’agit notamment de caractériser la liaison F/M par deux paramètres : la contrainte de cisaillement de la liaison interfaciale (τi) et le module de cisaillement du matériau (G12).Pour la partie mécanique, des essais de traction monotone et cyclée dans l’axe des fibres du renfort (exploités à l’aide de modèles micromécaniques), des essais de cisaillement Iosipescu, ainsi que des essais de push out ont été mis au point et exploités. Des analyses de micro caractérisation par microscopie électronique (MEB, FIB-MEB, MET) ont été réalisées avant et après essais mécaniques avec des interphases de différentes configurations afin de relier les différences de microstructures et d’épaisseurs aux chemins de fissuration matricielle et aux comportements mécaniques macroscopiques des composites. La liaison F/M la plus forte est notamment obtenue quand le degré de cristallisation et l’anisotropie structurale du BN sont peu élevés, pourvu que l’épaisseur de l’interphase soit suffisante. / The use of ceramic matrix composites (CMC), and particularly SiC/SiC materials, in place of metal alloys in aircraft engines has the potential to reduce their fuel consumption. They have very interesting physical and mechanical properties at high temperatures: low density, high resistance to thermal shock and non- brittle failure. In these materials, a thin layer is inserted between the fibres and the matrix: the interphase. Pyrolytic boron nitride is the interphase material of choice to achieve the desired applications. The bibliography clearly highlights the need and difficulty of "controlling" the intensity of fibre/matrix interfacial bonds (F/M) thanks to the interphase. But the exact influence of the crystallinity and thickness of BN-type interphases on its control, and consequently on the final mechanical behaviour of the industrial CMC, is still insufficiently known.A first issue addressed in this thesis is the scale of the CMC to be used in the laboratory. Indeed, so far, the most commonly used model materials are 1D mini and micro composites. They can be easily and quickly prepared by chemical vapour deposition but do not account for phenomena inevitably present in the industrial composite. Therefore, the use of new 2D model materials such as "monostrates" comprising a single-ply woven, the BN interphase and a SiC matrix in which the porosity can be filled with Si as in the case of industrial CMCs is more relevant and is proposed. However, due to the thinness of the specimens, the characterization and mechanical testing protocols had to be reviewed. This involves characterizing the F/M bond by two parameters: the shear stress of the interfacial bond (τi) and the shear modulus of the material (G12).For the mechanical part, monotonic and cycled tensile tests in the fibre axis of the reinforcement (operated using micromechanical models), Iosipescu shear tests, as well as push-out tests were developed and used. Micro characterization analyses by electron microscopy (SEM, FIB-SEM, TEM) were performed before and after mechanical tests with interphases of different configurations in order to link the differences in microstructures and thicknesses to the matrix cracking paths and macroscopic mechanical behaviours of the composites. The strongest F/M bond is obtained notably when the degree of crystallization and structural anisotropy of the BN are low, provided that the interphase is thick enough.