The compressive behavior of an annular rubber seal constrained in a pipe and the interaction between the pipe and the seal was studied experimentally using a specially designed test fixture that allowed the concentric alignment of the seal within the pipe and its axial compression using an electro-hydraulic Instron load frame. The hoop strain introduced in the pipe wall, due to the constraint of lateral expansion of the seal, displayed a parabolic distribution with a maximum value at the mid-height of the seal similar to the parabolic shape of the lateral expansion of the seal. The magnitude of the pipe strain increased with the friction coefficient of the interface between the seal and the compression rings, strain rate, and shape factor for a constant gap between the seal and the pipe wall. The relationship between the apparent compressive modulus and the shape factor (beyond experimental range) was studied using FEA. / October 2016

A polymer laminate consisting of multiple layers of proprietary blends of Ethylene Vinyl Alcohol (EVOH) and Thermoplastic Polyurethane (TPU) and used in the construction of air bladders was evaluated for hygroscopic effects driving delamination and multiple layer fragmentation. These air bladders are observed to suffer delamination during a manufacturing process that involves immersion in an alcohol/water solution. A plausible underlying mechanism is differential swelling and absorption by the laminate constituents. Both room and elevated temperature swelling tests were carried out to find the absorption and swelling coefficients of the constituents. These coefficients served as input into a Finite Element Analysis (FEA) model used to predict laminate failure. The diffusion coefficient for EVOH could not be obtained because the material did not reach saturation within the available timeframe of the experiment. The diffusion coefficient of the TPU was found to be 4.09E-12 [m^2/s] at room temperature and 1.26E-11 [m^2/s] at 40°C. The diffusion coefficient for TPU was an order of magnitude larger at elevated temperature and the TPU reached saturation much quicker than the EVOH, suggesting that the diffusion coefficient for the TPU was significantly greater than that of the EVOH. The swelling coefficients were 1.03E-3 m^3/kg and 9.97E-4 m^3/kg for the EVOH at room temperature and 40°C respectively, and 1.16E-3 m^3/kg and 1.09E-3 m^3/kg for the TPU at room temperature and 40°C respectively. The swelling coefficient was very close across materials and within the margin of error across temperatures. These results are required for future FEA simulations to confirm that differential swelling is the driving mechanism behind debonding and laminate failure. Tensile testing was done on laminated sheets used in production to identify cracking and layer separation at strains of 20%, 40%, and 100%. Scanning Electron Microscope (SEM) imaging was used to understand damage initiation and accumulation in the layers.

Les explosifs sont des matériaux qui, bien que potentiellement sensibles, sont conçus pour être stables en conditions normales, ainsi que lors de sollicitations mécaniques, chimiques ou thermiques « faibles ». Pourtant, sous sollicitations mécaniques de faible intensité, comme les impacts basse vitesse, ils peuvent réagir de manière intempestive. Les propergols, et plus particulièrement la butalite, objet de notre étude, présentent ce caractère : on observe des « réactions » pour des vitesses d'impacts inférieurs à 100 m.s-1, dont l'origine est probablement liée à l'endommagement microstructural du matériau.Dans ce contexte, le but ultime du CEA2 Gramat est d'obtenir un outil de prédiction de la vulnérabilité des matériaux énergétiques pour les impacts à basse vitesse de type tour de chute. Pour ce faire, il est essentiel de disposer de données sur la morphologie et le comportement (thermo)mécanique macroscopique du matériau considéré, de ses phases constitutives à l'échelle mésoscopique et de ses interfaces.Ainsi l'objectif de la thèse est de déterminer le type et le niveau de(s) endommagement(s) apparaissant(s) dans une « butalite inerte » suite à un impact mécanique dit « à basse vitesse » (i.e., inférieure à 100 m.s-1) réalisé à l'aide d'un dispositif de type tour de chute modifié, associant un suivi par vidéo numérique rapide et une analyse microtomographique ante- et post-essai, en étudiant le ou les phénomènes physiques à l'origine des réactions sous « faibles » sollicitations, leur évolution et leur(s) origine(s) physique(s). Les grains sont modélisés par une loi de comportement purement élastique et la matrice en PBHT est décrite par une loi visco-hyper-élastique (couplage d'une série de Prony et du modèle de Mooney-Rivlin). / Although they are potentially sensitive, energetic materials are designed to be stable under normal conditions, as well as “weak” mechanical, chemical or thermal loadings. However, under low mechanical loadings, such as low velocity impacts, they may react untimely. Propellants and especially the butalite, object of our study, show "reactions" to impact velocities below 100 m.s-1, whose origin is probably related to the material microstructural damage.In this context, the ultimate goal of CEA2 Gramat is to obtain a predicting tool for the vulnerability of energetic materials for low velocity impacts as drop weight test. So it is essential to have data on the morphology and macroscopic (thermo)mechanical behavior, its component phases at the mesoscopic scale and its interfaces.Thus, the objective of the thesis is to determine the type and the damage(s) level(s) generating in an "inert butalite", during a low velocity mechanical impact (i.e., less than 100 m.s-1), using a fast camera recording and ante- and post-test microtomographic analysis, or by studying the physical phenomena which are at the origin of reactions, their evolution and physical origin(s). Grains are represented by a purely elastic model and HTPB matrix is described by a visco-hyper-elastic model (coupling a Prony serie and Mooney-Rivlin model).