No description available.

No description available.

Le marché des matériaux composites à matrice organique (CMO) pour la réalisation de pièces structurales "froides" (-55°C < T < Tamb) arrive à saturation et l’industrie aéronautique vise à utiliser les CMO tissés 3D dans les pièces structurales dites "chaudes" (50°C < T < 300°C) des avions (nacelles,turbomoteurs). Ces conditions environnementales peuvent entrainer des phénomènes de dégradation à long terme. L'action de l'environnement et la complexité de la microstructure peuvent conduire au développement de gradients de propriétés dans les matériaux. À haute température, la matrice polymère peut présenter un comportement complexe dépendant du temps. Il est donc nécessaire de mettre au point une technique expérimentale capable de caractériser le comportement du matériau en fonction du temps à l'échelle locale, pour saisir les gradients des propriétés. Ce travail propose la mise en place d'un essai cyclique d'indentation instrumentée pour répondre à ces questions.L'indentation instrumentée est une technique expérimentale qui a rencontrée un grand succès au cours des dernières années. Dans sa forme classique, développée pour les matériaux à comportement élasto-plastique, elle consiste à réaliser un seul cycle de charge/décharge pour en déduire la dureté du matériau et le module élastique d’indentation. Cette analyse n'est pas appropriée pour les matériaux présentant un comportement dépendant du temps, ce qui nécessite de développer une nouvelle méthode optimisée pour les matériaux polymères. Les méthodes d'indentation pour les matériaux polymères proposées dans la littérature (fluage, dynamique), sont limitées à certaines charges. L'essai cyclique d'indentation proposé dans ce travail vise à mettre en évidence la complexité du comportement du polymère. Il est dérivé des essais cycliques macroscopiques et emploi une méthode d’analyse similaire. La technique est développée sur un polymère thermoplastique, le PEHD, pour lequel la réponse cyclique macroscopique en traction et cisaillement est connue. En suivant l’évolution au cours du temps des principaux indicateurs du comportement cyclique (le module d'indentation,l’aire de la boucle d'hystérésis et l'accumulation du déplacement) pour différentes fréquences, il est possible de mettre en évidence la réponse du matériau et d'effectuer une comparaison qualitative avec le comportement macroscopique. Le protocole d'indentation cyclique est ensuite utilisé pour étudier les gradients de propriétés dans la résine époxy thermodurcissable PR520 soumise à un vieillissement thermique à 150°C sous air à pression atmosphérique (jusqu'à 1000h), sous 2 bar d’O2et de N2 (pour 400h). Il est montré que la cinétique d’évolution du module d'indentation et du déplacement n'est pas affectée par le vieillissement. Cependant, leurs valeurs absolues varient de la surface au cœur du polymère, ce qui indique la présence de gradients. L'hystérésis du premier cycle est différente à travers le gradient, mais à partir du deuxième cycle, l'hystérésis est similaire pour toutes les conditions de vieillissement et les distances de la surface exposée. Les résultats obtenus sur des échantillons vieillis sous 2 bar d’O2 et de N2 permettent de conclure que la cinétique de vieillissement n'est pas de la pure thermo-oxydation. La méthode d'indentation cyclique est ensuite appliquée pour caractériser le comportement de la matrice époxy PR520 dans un composite tissé 3D,à l'état vierge et vieilli à 150°C sous air à pression atmosphérique. L'étude du composite à l'état vierge révèle que le comportement de la matrice polymère à proximité de la surface externe est différent de celui situé dans les zones internes du composite et du polymère pur. La comparaison entre le polymère pur et la matrice à l’état vieilli montre que les gradients de propriétés induits par l'environnement sont similaires. / The market of organic matrix composite (OMC) materials for the realisation of “cold” (-55°C < T < RT) structural parts is going towards saturation and aircraft manufacturers foresee the employment of 3D reinforced OMC in “warm” (50°C < T < 300°C) aircraft structural parts (nacelles, turbo-engines). These environmental conditions may lead to degradation phenomena over long time. The action of environment and the complexity of the material microstructure may lead to the development of material property gradients. At high temperature, the organic polymer matrix may exhibit complex time-dependent behaviour. Therefore, there is a need to develop an experimental technique able to characterise the material behaviour at local scale and to capture material gradients and time-dependent behaviour. The present work proposes the development of a cyclic instrumented indentation test to tackle all these issues. Instrumented indentation is a popular testing technique: its basic version, appropriate for elasto-plastic materials, includes the realisation of a single loading/unloading test, the measurement of the material hardness through the analysis of the indentation print, and the analysis of the unloading curve based on the assumption of elastic unloading behaviour to obtain the indentation modulus. This analysis is inappropriate for materials exhibiting time-dependent behaviour, which leads to the need of a new method optimized for polymer materials.Several indentation methods are available for polymer materials (indentation creep, nanoDMA), but are limited to some specific loadings. The instrumented indentation cyclic test proposed in this work tries to emphasize the whole complexity of the polymer behaviour, is inspired by macroscopic cyclic tests and is analysed similarly. The technique is first set up and developed by testing a HDPE thermoplastic polymer, for which the response to macroscopic cyclic tension and shear loading isknown. By following, at different frequencies, the evolution with time (with cycles) of the principal indicators of the cyclic behaviour – that is, the indentation modulus, the hysteresis loop area and the indentation depth accumulation - it is possible to highlight the time-dependent response of the material and to perform a proper – though qualitative - comparison with the macroscopic behaviour. The cyclic indentation protocol is then employed to study the material gradients in a thermoset PR520epoxy resin subjected to thermal aging at 150°C under air at atmospheric pressure (up to 1000h), 2 barO2 (for 400h) and N2. It is shown that the evolution with cycles of the indentation modulus and the cyclic creep is not affected by thermal aging. However, their absolute values vary from the surface to the core of polymer indicating the presence of gradients. The hysteresis of the first cycle is different through the gradient: from the second cycle, however, the hysteresis is similar for all aging conditions and distances from the exposed surface. Moreover, the time-dependent behaviour stays unchanged. Results from samples aged under 2 bar O2 and 2 bar N2 allow to conclude that the aging kinetics is not related to pure thermo-oxidation phenomena. The cyclic indentation method is finally applied to characterise the behaviour of PR520 epoxy matrix within a 3D interlock textile composite, in its virgin state and thermally aged at 150°C under air at atmospheric pressure (up to 1000h). The polymer inlarge matrix pockets between the fibrous reinforcements is studied in this work, so the constrainingeffect coming from the reinforcement is negligible. The study of the composite in virgin state revealsthat the behaviour of polymer matrix close to the external surface is different from that located ininternal zones of the composite and from the neat polymer. The comparison between the thermallyaged neat polymer and matrix in internal zones of the composite shows that the environment-inducedproperty gradients are similar.

Dans un contexte général de réduction de la pollution atmosphérique, l’industrie automobile cherche à augmenter le rendement des moteurs thermiques pour en limiter la consommation et les émissions. Pour satisfaire cet objectif, les températures de combustion sont en constante augmentation, ce qui entraîne une augmentation de la sollicitation thermique de la face d’échappement de ces moteurs. Ces pièces n’étant pas refroidies, elles sont exposées à des températures toujours plus élevées, ce qui nécessite une durabilité à haute température accrue des matériaux. Cette thèse a pour objectif d’évaluer la durabilité des pièces de fonderie de la face échappement telles que les collecteurs, corps de turbine ou turbo collecteurs réalisées en fonte GSSiMo+ et en acier inoxydable moulé 1.4826Nb à des températures supérieures à leur température maximale actuelle d’utilisation. Une première étape a porté sur l’effet de la vapeur d’eau sur l’oxydation de la surface pendant des traitements thermiques continus et cycliques et sur l’évolution associée de la microstructure. La seconde étape a porté sur l’effet de ces traitements thermiques sur les propriétés mécaniques à température ambiante. Enfin, un montage de fatigue thermomécanique par dilatation différentielle entre un support en carbure de silicium et une éprouvette métallique en V a été conçu pour être utilisé sur un banc d’oxydation cyclique sous atmosphère contrôlée afin d’étudier les mécanismes de fissuration et d’endommagement en fatigue thermomécanique ainsi que l’effet de l’environnement sur l’initiation de fissures dans ces conditions. L’augmentation de la température maximale des cycles thermiques appliqués de 700 à 800°C pour la fonte GS SiMo+ et de 850 à 950°C pour l’acier 1.4826Nb a pour principale conséquence une accélération de la perte de section par oxydation et une diminution des propriétés mécaniques du fait de l’évolution de la microstructure. Cette perte de section paroxydation est fortement accélérée en présence de vapeur d’eau. Les essais préliminaires réalisés avec le montage de fatigue thermomécanique ont permis de valider son dimensionnement et de mettre en avant les cycles thermomécaniques pour lesquels l’initiation et la propagation de fissures étaient les plus rapides sur des cycles 300-800°C pour la fonte et 300-950°C pour l’acier.Une plus ample campagne d’essai reste à réaliser pour évaluer les mécanismes de fissuration et l’effet de l’environnement en fonction de la température maximale / Due to tightening environmental standards, the automotive industry is constantly trying to improvethe efficiency of the internal combustion engines in order to increase their fuel economy andreduce their carbon emissions. The main way to meet this goal on a turbocharged engine is toincrease the combustion temperature. This leads to increasing exhaust gas temperature andincreasing thermal loads on exhaust manifolds and turbine housings. These components beinguncooled, their maximum temperature is increasing and require better durability at hightemperature. The main objective of this thesis is to evaluate the durability of SiMo spheroidalgraphite iron (SGI) and 1.4826Nb cast stainless steel (CSS) at higher temperature than the actualmaximum operating temperature in order to evaluate their use on engines with higher power. Thefirst part of the study focuses on the effect of water vapor on high temperature oxidation duringcontinuous and cyclic heat treatments and on the underlying microstructure evolution. The secondpart focuses on the effect of these heat treatments on the tensile properties at room temperature.At last, a thermomechanical fatigue test setup has been designed to study the effect ofenvironment on crack initiation during thermomechanical fatigue (TMF). This setup uses thedifference in thermal expansion between silicon carbide and metallic materials to generatemechanical strain and stress in a V-shaped specimen during thermal cycling in a controlledatmosphere cyclic oxidation test bench. Increasing maximum temperature during thermal cyclingfrom 700 to 800°C for SiMo SGI and from 850 to 950°C for 1.4826Nb CSS leads to acceleratedwall thickness loss due to increased oxidation and to a drop in mechanical properties duemicrostructure evolution. The wall thickness loss is further accelerated in humid atmosphere.Preliminary tests performed with the TMF setup demonstrated the design is valid and enabled toidentify the thermomechanical loads leading to faster crack initiation and propagation for 300-800°C thermal cycling of SiMo SGI and 300-950°C thermal cycling of 1.4826Nb CSS. More testsare necessary to identify the crack initiation mechanisms according to the maximum temperatureof the thermal cycle and the atmosphere

Solute clustering is known to play a significant role in the embrittlement of reactor pressure vessel (RPV) steels. When precipitates form they impede the movement of dislocations, causing an increase in hardness and a shift in the ductile-brittle transition temperature. Over time this can cause the steel to become brittle and more susceptible to fracture. Thus, understanding precipitate formation is of great importance to the nuclear industry. The first part of this thesis aims to isolate and better understand the thermal aging component of embrittlement in low copper, model RPV steels. Currently, relatively little is known about the effects of Ni and Mn in a low copper environment. Therefore, it is of interest to determine if Ni and Mn form precipitates under these conditions. To this end, hardness measurements and atom probe tomography were utilized to link the mechanical properties to the microstructure. After 11,690 hours of thermal aging a statistically significant decrease in hardening was observed. Consistent with hardness measurements, no precipitates were present within the matrix of the thermally aged RPV steels. The local chemistry method was then applied to investigate the very early stages of solute clustering. Association was found to be statistically significant in both the thermally aged and as-received model RPV steels. Therefore, no apparent trends regarding the changes in solute association between the as-received and thermally aged RPV steels were identified. Small, non-random clusters were observed at heterogeneous nucleation sites, such as carbide/matrix interfaces and grain boundaries, within the thermally aged material. The clusters found at the carbide/matrix interfaces were all rich in Mn and approximately 90-150 atoms in size. The clusters located along the observed low-angle grain boundary, however, were significantly larger (on the order of hundreds of atoms) and rich in Ni. Lastly, copper-rich precipitates (CRPs) and Mn- and Ni-rich precipitates (MNPs) were observed within the cementite phase of a high copper and low copper RPV steel, respectively, following long term thermal aging. APT was used to characterize these precipitates and obtain more detailed chemical information. The presence of such precipitates indicates that a range of precipitation can take place within the cementite phase of thermally aged RPV steels. The second part of this thesis aims to investigate the effects of ion irradiation on the microstructure of low copper RPV steels via APT. These steels were ion irradiated with 6.4 MeV Fe³⁺ ions with a dose rate of 1.5 x 10^-4 dpa/s at 290°C. MNPs were observed in all five of the RPV steels analyzed. These precipitates were found to have nucleated within the matrix as well as at dislocations and grain boundaries. Using the maximum separation method these MNPs were extracted and characterized. Precipitate composition, size, volume fraction, and number density were determined for each of the five samples. Lastly, several grain boundaries were characterized. Several emerging trends were observed within the samples: Ni content within the precipitates did not vary significantly once a threshold between 30-50% was reached; bulk Mn content appeared to dictate Si and Mn content within the precipitates; and samples low in bulk Ni content were characterized by a higher number density of smaller precipitates. Additionally, by regressing precipitate volume fraction against the interaction of Ni and Mn, a linear relationship was found to be statistically significant.

Les matériaux isolants des câbles électriques des centrales nucléaires sont constitués d’élastomères ayant de bonnes capacités de résistance au vieillissement, comme les copolymères d’éthylène et de propylène (EPR, EPDM, …), car leurs conditions d’utilisation les soumettent à de faibles radiations et à des températures pouvant atteindre 50°C dans le bâtiment réacteur. L’objectif de la présente thèse est de contribuer à affiner la compréhension du vieillissement des isolants des câbles, en étudiant l’évolution de la structure et du comportement de matériaux modèles. Ces matériaux sont composés d’un EPDM amorphe réticulé au peroxyde et comprenant différent taux de charges micrométriques, constituées de trihydrate d’aluminium (ATH), utilisées pour leurs propriétés anti-feu. Aucun anti oxydant supplémentaire n’a été ajouté. L’approche utilisée dans cette étude consiste à réaliser des vieillissements thermiques accélérés dans une étuve ventilée à 130°C durant 20h, 30h et 40h, puis à caractériser l’évolution de la structure des matériaux, leur résistance à la traction et à la fissuration. La caractérisation expérimentale indique que l’oxydation, induit par le vieillissement, provoque un phénomène de rupture de chaînes prédominant, dont résulte la formation d’une quantité considérable de fraction soluble. La diminution de la densité de chaînes actives associée à l’accroissement de la partie soluble du matériau, constituée de chaînes libres devenant de plus en plus courtes durant le vieillissement, entraine une perte des propriétés mécaniques et de la résistance à la fissuration. / Insulating materials of electric cables in nuclear power plants are constituted by rubbers having good ageing resistance properties, like ethylene and propylene copolymers (EPR, EPDM …), because in service these cables are submitted to low irradiation and temperatures that could reach 50°C in the reactor building. The objective of this thesis is to contribute to improve the understanding of the ageing of cables insulating materials, by studying the evolution of the structure and the behavior of models materials. These materials are constituted an amorphous EPDM crosslinked with peroxide and including different ratio of micrometric fillers, constituted of aluminum trihydrate (ATH), used for its fire retardancy properties. Any antioxidants have been added. The approach used in this study consist to realize accelerated thermal ageing in air ventilated oven at 130°C during 20h, 30h and 40h, then characterizing the evolution of the structure of the materials, their tensile and crack propagation resistance. The experimental characterization show that the oxidation, resulted from the ageing, provoke a predominant chain scission phenomenon which leads to the formation of a large quantity of soluble fraction. The decrease of the density of active chains associated with the increase of the soluble part, constituted of free chains becoming shorter during the ageing, leads to a loss of tensile strength and crack propagation resistance.

The dissertation thesis submitted deals with the study of dielectric properties of epoxy nanocomposites containing nanoparticles of inorganic oxides. These nanocomposites may have a promising technologic application for electric insulations in view of their higher resistance against partial discharges; yet information about their behavior in the course of ageing is not available. If at least a partial mass replacement of the currently used epoxy insulation with nanocomposite-based insulations is due to occur, the knowledge of the changes of their dielectric properties in the course of their operation will become indispensable. Within the framework of this dissertation, ensembles of samples of epoxy resins without fillers and with Al2O3, WO3, TiO2 and SiO2 fillers in the form of nanopowders, in concentrations up to 12 wt %, have been prepared. These ensembles have been measured prior to ageing and exposed to long-time (up to 5000 hours) ageing at increased temperatures 200, 250 and 300 °C and in a few cases also at 330 and 360 °C. Samples were measured in the course of ageing roughly in a logarithmic time series after 1, 2, 5, 10, 20, 50, 100, 200, 500, 1000, 2000 and 5000 hours. The measured quantities included complex permittivity , internal resistivity i and loss factor tan at temperatures ranging from -153 °C to +167 °C and in the frequency range 10-2 – 106 Hz. Changes in nanocomposites have been investigated using not just dielectric spectroscopy measurements, but other methods, too, namely Fourier-transformed infrared spectroscopy (FTIR), thermogravimetric analysis (TGA), differential scanning calorimetry (DSC) and scanning electron microscopy (SEM). The experiments have proved that materials with different fillers respond to the same concentrations of various fillers in different ways. The addition of nanoparticles, without the addition of microparticles, at a relatively low concentration (max 12 wt %), was not sufficient for reaching fundamental changes in dielectric spectrum; only smaller changes of dielectric strength and shifts of relaxations and in relaxation maps have occurred. Out of more pronounced changes, increase of concentration of the SiO2 filler in the epoxy matrix brings about a decrease of electrical conductivity in the resulting nanocomposite. The TiO2 filler had a different impact. Different TiO2 concentrations make their marked appearance in the region between the relaxation and relaxation. The TiO2-filled nanocomposites do not exhibit the unambiguous dependence of electrical conductivity on nanofiller concentration. It can be concluded that the mere addition of nanoparticles, without the addition of established microparticles, does not change the dielectric spectrum substantially. Generally, a serious problem was the production of the nanocomposite with a uniform distribution of nanoparticles. The preparation of such a nanocomposite was not trivial and, in industrial applications, this issue will require a specific focus, so as to avoid the formation of undesirable aggregates. Within the framework of this research, a methodology for the production of an epoxy nanocomposite has been developed with as high as possible uniformity of nanoparticle distribution.

Le polydicyclopentadiène (PDCPD) est un polymère thermodurcissable synthétisé à partir de la polymérisation par métathèse par ouverture de cycle, qui présente l'avantage d'un temps de polymérisation court du fait de la présence de doubles liaisons et de l'utilisation de catalyseurs organométalliques. Dans cette thèse, PDCPD est considéré comme un candidat pour des applications en milieu marin, pour lesquelles la question de la durabilité doit être évaluée. L'objectif de ce travail est d'établir un modèle cinétique non empirique pour la prédiction de la durée de vie du PDCPD au cours de l'oxydation thermique. Tout d'abord, la caractérisation par spectroscopie Infrarouge et dosages des peroxydes au cours de l’oxydation a été réalisée afin de mettre en place un modèle cinétique pour le PDCPD sans stabilisant. En particulier, les constantes de vitesse d'initiation ont été identifiées par une étude de la thermolyse des hydroperoxydes; les constantes de vitesse de terminaison ont été obtenues par des expériences spécifiques sous différentes pressions d'oxygène. L'effet du catalyseur sur la cinétique d'oxydation a également été étudié dans ce travail. La consommation de doubles liaisons du polymère entraîne une augmentation significative de la concentration de nœuds de réticulation, puis une augmentation de la contrainte de rupture de PDCPD. Le modèle cinétique créé prédit ces tendances ainsi que l'épaisseur de la couche oxydée. L'effet des stabilisants visant à ralentir la vitesse d'oxydation a également étudié expérimentalement puis modélisé. En particulier, les amines encombrées (HALS) se sont révélées très efficaces et compatibles avec les catalyseurs de la ROMP. Dans la dernière partie, l’effet de l’oxydation thermique du PDCPD sur le comportement de sorption de l’eau a été étudié. Le PDCPD vierge présente une très faible absorption d'eau à cause de sa structure apolaire, alors que le PDCPD vieilli est plus hydrophile en raison de la formation de groupes polaires ce qui peut avoir une influence sur sa durée de vie en milieu marin. / Polydicyclopentadiene (PDCPD) is a thermoset polymer synthesized from ring opening metathesis polymerization with the advantage of short processing time due to the presence of double bonds and the use of organometallic catalysts. In this thesis, PDCPD is for example considered as a candidate for the field-joint in a marine environment. The question of the durability of PDCPD has to be assessed. The objective of this work is to establish a non-empirical kinetic model for the lifetime prediction of PDCPD during thermal oxidation. Firstly, the characterization by Infrared spectroscopy and peroxide titration of PDCPD during its thermal oxidation was conducted to set up a kinetic model for additive-free PDCPD. In particular, the initiation rate constants have been identified especially through hydroperoxide decomposition rates; the termination rate constants have been obtained by specific experiments under various oxygen pressures. The consumption of double bonds in the polymer causes a significant increase in crosslink concentration then increasing yield stress of PDCPD. The kinetic model predicts those trends as well as the thickness of the oxidized layer. The effect of stabilizer aimed at slowing down the oxidation rate was also experimentally studied and modeled. In particular, hindered amine stabilizers were shown to be promising candidates with good stabilizing properties and compatible with ROMP catalysts. Lastly, the effect of thermal oxidation of PDCPD on water sorption was investigated. Virgin PDCPD presents very low water absorption due to its apolar groups. Whereas, aged PDCPD shows increased hydrophilicity associated with polar groups build-up (which can probably impact its lifetime in marine conditions.

In Pressurised Water Reactors, the reactor pressure vessel (RPV) is considered a life limiting component due to the degradation of its mechanical properties. Nano-scale Cu-enriched precipitates are known to cause embrittlement in the form of increases in hardness and the ductile-to-brittle transition temperature. The effect of irradiation on the RPV is the dominant contributor to this embrittlement. This is due to the increased mobility of Cu from the high number of vacancies, and the matrix damage providing many heterogeneous nucleation sites. However, there are also thermal effects which may be difficult to separate from the irradiation effects. To understand the contribution of the long term thermal ageing to RPV embrittlement a series of weld and plate materials containing systematic variations of Ni and Cu has been thermally aged for times up to 100,000 hrs at 330 degrees C, 365 degrees C and 405 degrees C. Microstructural characterisation using Atom Probe Tomography has been performed. Complimentary Monte-Carlo simulations have been used to investigate the early stages of formation of Cu-enriched precipitates. Thermal ageing produces a high number density of nano-scale Cu-enriched precipitates. These nanometre precipitates have a Ni-Mn-Si rich interface which was found to be wider with increased precipitate size, lower ageing temperature and higher bulk Ni content. This interface reduces the interfacial energy of the Cu-enriched precipitates through a combination of the minimising of unfavourable Fe-Cu bonds and reduction in lattice strain. The matrix Cu levels after ageing for 90,000 - 100,000 hrs were found to be around 0.06 - 0.07 at.%, close to the expected solubility limits for Cu in Fe. The Fe content of the precipitates has been characterised and found to be higher at lower ageing temperature and for smaller precipitate sizes. Cu precipitation and solute segregation at dislocations were observed, particularly in the SG steels aged at higher temperatures where the supersaturation is lowest. Movies were produced from the Kinetic Monte-Carlo modelling (see accompanying DVD) and along with other analyses indicated that sub-critical Ni and Mn clusters may be active in the formation of Cu-enriched precipitates. This mechanism appears to occur in both the high and low Ni steels. Thus, the higher number density of larger precipitates observed with increased bulk Ni content is thought to be a consequence of the greater number of sub-critical Ni-Mn clusters providing more nucleation sites. A small number of grain boundaries were examined. Segregation of many solute species to them was observed, which is thought to result from a multi-element co-segregation process. Ni-Mn-Si precipitates were observed at grain boundaries and dislocations in the high Ni steels with high and low Cu levels. These are consistent with similar phases in found in irradiated high Ni steels. In the high Cu steels these particles were much larger and associated with Cu-enriched precipitates.

Les aciers ODS (Oxide Dispersion Strengthened) sont envisagés comme matériau de gainage du combustible pour les futurs réacteurs nucléaires au sodium de Génération IV. Ces matériaux présentent une excellente résistance au fluage à haute température et au gonflement sous irradiation, mais des interrogations subsistent quant à leur propriété de résilience. L'objectif de ce travail de thèse est alors de comprendre l'effet de différents paramètres (composition chimique, texture, vieillissement thermique…) sur le comportement à l'impact des aciers ODS et sur l'apparition de la rupture de type fragile. L'objectif à terme est d'appréhender l'apparition de ce type de rupture afin d'assurer l'intégrité des composants en acier ODS en conditions normales ou incidentelles.Dans un premier temps, cette étude a permis d'évaluer la stabilité de deux nuances d'acier ODS contenant 14%Cr et 18%Cr dans des conditions de vieillissement thermique. La nuance à 18%Cr a été écartée à cause de la formation d'une phase intermétallique fragilisante de type sigma à 600°C. Par contre, la nuance à 14%Cr est plus prometteuse puisque son comportement est resté stable entre 400°C et 600°C pour une durée de vieillissement maximale de 10000 heures, et ce même si des précipités de phases alpha', de Laves Fe2W et des carbures de chrome ont été observés. D'autre part, la texture morphologique, caractérisée par des grains allongés selon la direction de filage, est propice à la propagation de fissures intergranulaires selon cette direction. La texture cristallographique régit, quant-à elle, les micromécanismes de clivage. En effet, les entités microstructurales qui contrôlent la propagation du clivage sont des groupes de grains faiblement désorientés les uns des autres d'un point de vue cristallographique, et qui sont associés à la notion de grains effectifs. Enfin, le comportement en traction et en flexion d'une nuance d'acier ODS à 14%Cr a été modélisé, et un critère de rupture fragile basé sur une valeur de contrainte critique a été proposé. Ce modèle nous a alors permis de simuler des essais sur différentes géométries et de prédire l'apparition de la rupture fragile sur cette nuance. / The Oxide Dispersion Strengthened (ODS) steels have been identified as potential materials for fuel cladding in Generation IV nuclear reactors. They are characterized by a very good resistance to swelling under irradiation and to high temperature creep, but questions still remain about the impact toughness of these materials. The first aim of this work is to understand the effects of different parameters (chemical composition, texture, thermal ageing...) on the impact behaviour of ODS steels. The final objective is to predict the occurrence of the brittle failure on the ODS steel components in normal or incidental conditions.Firstly, this study assesses the stability of two ODS steel grades containing either 14%C or 18%Cr during thermal ageing. The 18%Cr steel grade has been discarded due to the growing of the brittle intermetallic sigma phase at 600°C. However, the 14%Cr steel grade showed a promising behaviour given the stability of its mechanical properties after thermal ageing between 400°C and 600°C, for a maximum duration of 10000 hours, and regardless of the presence of chromium carbides, alpha' and Laves phases. Then, the morphological texture characterized by elongated grains along the extrusion direction enhances the propagation of intergranular cracks along this direction. Meanwhile, the crystallographic texture controls the cleavage micromechanisms. Indeed, the microstructural entities that control the propagation of the cleavage crack are groups of grains which are characterized by low crystallographic misorientations and referred to as effective grains. At last, the tensile and bending behaviour of a 14%Cr ODS steel grade has been modelled. A brittle failure criterion based on a critical stress value had also been suggested. This model allowed us to simulate mechanical tests on different geometries and to predict the occurrence of the brittle failure.