Search

The performance of insulating refractory materials is determined by the thermo-physical properties. The first part of this thesis is dedicated to the evaluation of the thermal conductivity, which is an input parameter in modelling studies of heat transfer through a steel ladle lining. Comparison of different methods for its determination leads to some variation in the experimental results. However, taking into account effects such as heat losses, humidity, anisotropy and heterogeneity, the discrepancy could be reduced to within 10% when the same heat flow direction through the material is considered. Thermal conductivity is significantly modulated by the microstructure. Given that refractories are multi-phase materials, model materials with similar microstructures were studied to identify the role of factors such as porosity, grain boundaries, grain size and presence of second phases. Then three real cases of industrial interest were examined. The ladle is a vessel made by different bricks. Therefore, joints play an important role. The interface thermal resistance was estimated to be in the range 10-4 to 10-3 m2·K·W-1 useful as a first approximation in computer modelling. In fact the lining is subjected to hydrostatic pressure when the ladle is filled with hot liquid steel implying the materials require sufficient compressive strength. For compression loads up to 0.4 MPa on the thermal expansion and up to 0.04 MPa on the thermal conductivity of insulating lining materials, no significant effect was observed. Finally, one of the main problems for steelmaking industry is represented by corrosion, which can require replacement of the materials in order to guarantee the same operation condition for all the campaign. The role of carbon contamination for the thermal conductivity of the insulating lining was evaluated.; Les performances des matériaux réfractaires isolants sont fortement dépendantes de leurs propriétés thermophysiques. La première partie de cette thèse est consacrée à l’évaluation de la conductivité thermique, qui est un paramètre important dans la modélisation du transfert de chaleur à travers le revêtement d’une poche d’acier. Les méthodes expérimentales utilisées ont révélé des différences entre les valeurs mesurées, qui peuvent être réduit d’environ 10% pour une direction du flux de chaleur considérée, dès lors que les pertes de chaleur, l’humidité, l’anisotropie et l’hétérogénéité sont prises en compte. La conductivité thermique est donc significativement influencée par la microstructure. Étant donné que les réfractaires industriels sont des matériaux multi-phasés complexes, une approche permettant de mieux appréhender leur comportement consiste à étudier des matériaux modèles de microstructure voisine mais simplifiée, afin d’identifier la contribution de paramètres tels que la porosité, les joints de grains, la taille des grains et la présence de secondes phases. Puis trois cas d'intérêt industriel en lien avec les réfractaires précédents ont été examinés. Comme une poche d’acier est constituée de différentes briques, les joints jouent un rôle majeur. Leur résistance thermique interfaciale a été estimée entre 10-4 et 10-3 m2·K·W-1, ce qui constitue une valeur expérimentale de haute importance pour améliorer la fiabilité de la modélisation. De plus, le revêtement est soumis à des contraintes de compression induites par la pression hydrostatique de l’acier en fusion, ce qui implique que les matériaux requièrent une résistance mécanique suffisante. Pour des valeurs de contrainte appliquée de 0.4 MPa et de 0.04 MPa, aucun effet significatif n'a été observé respectivement sur la dilatation et la conductivité thermique des matériaux isolants étudies. Enfin, l'un des principaux problèmes rencontré dans l’industrie sidérurgique concerne la corrosion, qui conduit souvent au remplacement prématuré des matériaux afin de garantir des conditions de service optimales. Le rôle de la contamination par le carbone sur la conductivité thermique du revêtement isolant a ainsi été évalué.

The origins of the variability of hemp concrete material properties are numerous. They include among others those related to the properties of its constituents and material itself as well as the methods used for their characterizations. This thesis is interested in the study of the variability of hemp concrete properties taking into account these different parameters and in particular the type of hemp particles used. The litterature review corried out allowed to present the hemp concrete material, the properties of its constituents, their methods of characterization and also the extend of properties variability and sensitivity due to various parameters. Furthermore, it allowed to identify the parameters to be considered in the context of this thesis. Therefore, this thesis is devided into four chapters in addition to the literature review. In the second chapter, a multi-criteria study on the properties of hemp particles is presented. After an experimental study of characterization for 13 types of hemp particles, a two level analysis of result is performed. It takes into account the density, water absorption capacity and particle size distributions. First, the results of the single characteristic study are presented; it is about a comparison of a given characterstic for all hemp particles at the same time. Secondly, the results of the multi-criteria study are presented. In this last case, the analysis is corried out by taking into account all hemp particles and characteristics, both at the same time. The outcome of this study allowed to classify hemp particles into three groups from which, low, medium and high mechanical peformances are expected, respectively. Chapter 3 is a preliminary study taking into account various parameters as sources of variability for hemp concrete properties, such as the testing laboratory (equipments), the batch, the specimen size and hemp particles type. The obtained results highlight the need for further investigations about the impact of hemp particles type. Moreover, the considerable dispersion in the results of Young’s modulus is likely associated with its calculation method. It then becomes important to deepen the study of its impact on the values of Young’s modulus obtained. The chapter 4 aims to answer the problem found on the method for calculating the Young’s modulus. Various methods from literature are used to analyse the stress-strain curves from samples manufactured under the same conditions. The variability observed in results with respect to used method allowed to highlight their impact and to provide the floating method as the most pertinent since it presents less variability. In addition, a model to describe the mechanical behavior law of hemp concrete is proposed. It allows to determine the enveloppe curve which corresponds to experimental curve from the monotonic loading. It can also allow to reproduce the experimental curve from a cyclic loading. In the last chapter, based on the results of chapter 2, nine types of hemp particles were selected for the preparation and manufacturing of specimens of the study. Under the same conditions (manufacturing and test), it has been demonstrated experimentally the impact of hemp particles on mechanical properties with a factor 10 between low and high values from obtained results. The mechanical response characterized by low (5% and 8% and, Les origines de la variabilité des propriétés du matériau béton de chanvre sont nombreuses. On distingue celles liées aux propriétés des ses constituants et du matériau lui-même auxquelles il faut ajouter les dispersions qui résultent des méthodes utilisées pour caractériser ce béton. Ce travail de thèse, s’intéresse à l’étude de la variabilité des propriétés du béton de chanvre en tenant compte de ces différentes sources et en particulier au type de chènevotte utilisée. L’étude bibliographique réalisée permet de comprendre le matériau béton de chanvre, notamment ses constituants, les méthodes de caractérisations en usage ainsi que l’ampleur de la variabilité et de la sensibilité de ses propriétés vis-à-vis des différents paramètres. Cet état de l’art, permet en outre d’identifier les paramètres à considérer dans le cadre de cette thèse. Notre travail de thèse est alors subdivisé en 4 chapitres en plus de l’étude bibliographique. Dans le deuxième chapitre, une étude multicritère sur les propriétés des chènevottes est présentée. Après une campagne expérimentale de caractérisation de 13 types de chanvre, une analyse des résultats à deux niveaux est réalisée. Elle prend en compte, la masse volumique, la capacité d’absorption d’eau et la granulométrie. Dans un premier temps on présente les résultats de l’étude mono-caractéristique ; il s’agit de la comparaison d’une caractéristique donnée pour l’ensemble des chènevottes. Dans un second temps, on présente les résultats de l’étude multicritère. Il s’agit d’analyser l’ensemble des chènevottes en prenant en comptes les différentes caractéristiques à la fois. A l’issue, les chènevottes sont classées en trois groupes. Le chapitre 3 constitue une étude préliminaire tenant compte des différents paramètres sources de variabilités des propriétés du béton de chanvre, comme le laboratoire d’essai (équipements), la gâchée, la taille d’éprouvette et le type de chanvre. Les résultats obtenus montrent la nécessité d’étudier l’impact du type de la chènevotte. Par ailleurs, la dispersion considérable obtenue pour les résultats du module d’Young est vraisemblablement associée à sa méthode de calcul. Il devient alors important d’approfondir l’étude de son impact sur les valeurs du module d’Young obtenues. Le chapitre 4 vise justement à étudier les méthodes de calcul du module d’Young. Les différentes méthodes de la littérature sont alors utilisées pour exploiter les courbes contrainte-déformation dont les essais ont été réalisés dans les mêmes conditions. Les variabilités observées sur les résultats vis-à-vis de chacune des méthodes permettent alors de mettre en évidence leur impact et de proposer la méthode « flottante » comme étant la plus pertinente. Par ailleurs, un modèle permettant de décrire la loi de comportement mécanique du béton de chanvre est proposé. Il permet de déterminer la courbe enveloppe correspondant à la courbe expérimentale issue du chargement monotone, et permet également de reproduire la courbe expérimentale issue d’un chargement cyclique. Dans le dernier chapitre, en se basant sur les résultats du chapitre 2, neuf types de chanvre ont été sélectionnés pour la confection des éprouvettes de l’étude. Dans les mêmes conditions (fabrication et essai), il a été mis en évidence expérimentalement l’impact de la chènevotte sur les propriétés mécaniques avec un facteur égal à 10 entre les valeurs faibles et élevées. La réponse mécanique caractérisée par des faibles (5% et 8% et

International audience; In the last decade, many finite element (FE) models of the spine were used as surrogate experiments to provide relevant knowledge on the biomechanics of spinal trauma [1-2]. One of the major concerns in the development of such model is the modeling and validation of the spinal components’ mechanical behavior in dynamic loading conditions. For instance, very few FE models have considered the strain rate dependency shown by the intervertebral disk (IVD) when submitted to high strain rates [3], thus limiting their ability to represent traumatic conditions. Such limitations arise from the lack of appropriate material properties available in the literature. Thus, the purpose of the current study was to determine suitable material properties of the IVD for its FE modeling under dynamic compressive loads. Lumbar spine; Material properties; Finite element analysis; Reliability study, High strain rate.

An investigation was undertaken of the creep of columnar-grained ice under constant compressive stresses of 98 and 59 kPa in order to extend the observed relationship between strain-rate and stress further into the low-stress region. Stress was applied perpendicular to the long direction of the grains. Observations were made for temperatures between -5 and -40°C. The steady-state creep rate for secondary creep was not yet attained for strain of 1.4% and stress of 98 kPa. An initial yield was observed at that stress in the strain range of 0.2 to 0.3%, similar to that seen at higher stress. The observations showed that the strain-rate tended to a linear dependence on stress below 49 kPa and that more than one dislocation process with different values for the stress exponent may contribute to the strain at higher stresses. An activation energy of 63 kJ mol−1 was consistent with the observed temperature dependence of the strain-rate. Straining ice to a given strain under a stress of about 0.3 MPa and then reducing the load may be a convenient way to study the stress, strain and temperature dependence of the strain-rate at low stress.

Observations are reported on the cracking activity during compressive creep of columnar-grained ice with the axis of hexagonal symmetry of each grain tending to lie in the plane perpendicular to its long direction. Because of the nature of the ice and the direction of the stress relative to the columnar grains, the deformation behaviour and cracking activity were two-dimensional in nature. The cracks that formed were of the cleavage type, long and narrow, and usually involving only one or two grains. Information is presented on the strain, stress and temperature dependence of the crack density, and cracking rate (i.e. number of cracks per cm[2] per unit strain). It was found that crack formation takes place when the compressive stress exceeds about 6 kg/cm[2]. For stress between 6 and about 10 kg/cm[ 2], most of the cracking activity occurs during primary creep. If the stress is greater than about 12 km/cm[2], deterioration of the structure due to the cracking activity causes the primary stage of creep to be transformed directly to the tertiary stage. The strain dependence of the crack density indicates the existence of two independent families of cracks, each obeying Weibull statistics. The implications of this for the ductile to brittle transition in ice are discussed.

Previously undeformed columnar-grained ice exhibits a period of increasing or constant creep rate during the transient creep stage when loaded in simple compression perpendicular to the long axis of the columns. It is shown that this behavior is associated with the formation of small-angle boundaries and internal cracks. Creep strain beyond 0.25% for first load tends to a power-law dependence on time. On reload, specimens exhibit a normal transient creep behavior and have, initially, a lower resistance to deformation than for first load. With deformation, this resistance increases so that the reload creep curves cross the first load curves at about 0.2% creep. For reload, creep strain less than about 0.25% and greater than about 0.2% appears to have a power-law dependence on time with exponent about equal to that for first load.

Observations have been made at two temperatures on the time to formation of the first internal cracks in columnar-grain ice subjected to a constant compressive load applied perpendicular to the long direction of the grains. Crack formation was observed to be a thermally activated process, with the time to formation of the first large crack (2 or more mm wide and 20 or more mm long) following the general law t=t(0)exp[(U[0]-alpha rho)divided by kT]. The value of U(0), the activation energy for zero applied stress rho, was found to be about 25 kcal/mole, t(0) about 4.6 X 10(-17) sec and alpha about 1.4 X 10(-20) cm(3)/molecule. The amount of strain associated with the formation of the first crack was observed to increase with decreasing stress, but to be independent of temperature to within the accuracy of the experiments. The results were found to be consistent with the models for crack nucleation by pile-up of dislocations discussed by Stroh (1954) and Bullough (1964). It was observed that dislocation cracks could initiate Griffith cracks if the stress exceeded about 7 kg/cm(2), and no cracks were observed when the applied stress was less than about 5 kg/cm(2).

The apparent activation energy for creep of columnar-grained ice subjected to a compressive stress of 9.8 X 10[4]N m[-2] ( 1kgf/cm[2]) is found to be 15.5 kcal mole[-1] from measurements made on a single specimen over the temperature range -5 degrees C to -40 degrees C. The analysis indicates that the deformation is controlled by diffusion processes, and that the difference between the measured activation energy and that for self-diffusion in ice can be explained by the temperature dependence of Young's modulus.