A combination of fractography, microstructural analysis and finite element modelling was used to investigate several topics relating to the fatigue of nuclear grades of austenitic stainless steel operating in both air and simulated PWR water environments. The work is broadly separated into four main categories. The first two involved analysing specimens from standard fatigue endurance tests using a wide range of microscopic techniques. The relevance and uses of a modern laser scanning confocal microscope are presented and the benefits of using such a technique are discussed. Methods for the automation of both striation counting procedures and hysteresis data analysis are described and the results are demonstrated. Finite element analyses were performed in order to develop the understanding of fatigue crack growth within standard cylindrical endurance specimens. A variety of different crack tip parameters were used in order to develop expressions for crack growth rates in terms of the strain intensity factor and the J-integral. The derived expressions were compared to the results of striation spacing measurements from multiple endurance specimens that were tested in both air and water environments. The expressions were used to perform back-fitting calculations on standard endurance curves in order to produce alternative curves representing the number of loading cycles to cause the initiation of short cracks with depths in the range of 0.25-0.5 mm. The effects of hold-times on the fatigue life of stainless steel endurance specimens were explored as part of the international AdFaM research programme. Results from the programme partners are presented which demonstrate the beneficial effects of static hold-times on extending the fatigue lifetime of specimens. A range of microstructural analyses were performed on test specimens and results are presented. No significant effects of hold-times on microstructure, crack growth rates or material hardness were found. Analysis of hysteresis data demonstrated an increase in the cyclic hardening and a decrease in the plastic strain range after a hold. From an analysis of the fatigue test results, it was concluded that hold-times affect the earliest stages of fatigue (nucleation and initiation), most likely due to the effects of strain ageing. Several possible explanations for the observed phenomenon of specimen shrinkage during static holds are presented and discussed, however no conclusive explanation was identified. Further work is identified that could lead to future improvements in the understanding of all areas of investigation that have been reported. Overall, the work reported here has helped to develop the understanding of fatigue behaviour and mechanisms in the materials of interest. This was done through investigations using a synergistic combination of microscopy and numerical modelling techniques.

Dimmuborgir is thought to be a former rootless shield volcano, which was fed with lava from a nearby crater row, 2170 ± 38 calendar years before present. In this study, the orientation of striations on the sides of lava channels, collapse structures and lava pillars were measured to find out how the enigmatic ~2 km by 2 km volcanic structure at Dimmuborgir was drained. During one week of field work 149 striations were found and measured, with respect to their dip angle, dip direction and elevation. Their locations were recorded with a GPS (Global Positioning System) receiver. The orientations of the striations were visualized on Google Earth satellite images and on images from a Digital Terrain Model (DTM) of Dimmuborgir. Resulting visualizations show that Dimmuborgir was drained radially and in multiple stages. It is concluded that Dimmuborgir was drained towards the west, the northeast and the southeast. The drainage towards the west was channeled. The drainage towards the northeast and the southeast was radially inwards, towards the collapsed parts of Dimmuborgir.

Cette thèse comporte une étude expérimentale sur la dynamique de croissance par fatigue d’une fissure dans un caoutchouc naturel renforcé avec des nanoparticules (silice, noir de carbone). Elle s’inscrit dans un cadre industriel visant à mieux comprendre si l’amélioration des performances apportée par la silice dans les caoutchoucs synthétiques est transposable au caoutchouc naturel qui cristallise sous contrainte. L’objectif est de comprendre comment la rupture en fatigue dépend de la nature du matériau et des paramètres de contrôle de l’expérience. La dynamique de croissance de la fissure est suivie à l’aide d’une caméra optique et une caméra infrarouge permet d’évaluer l’auto-échauffement. La morphologie des faciès de rupture est caractérisée par observation post mortem (MEB, profilométrie optique). L’influence de la température, de la fréquence, de la nature et du taux de charge a été étudiée. Dans tous les cas, la croissance de la fissure devient instable à partir d’un certain niveau de déformation. Cela se traduit sur les courbes de l’énergie de déchirure en fonction de la vitesse de la fissure par l’existence de deux branches distinctes : une branche à vitesse basse où une morphologie de rupture très rugueuse, pouvant présenter des signes de cavitation, est observée ; une branche à haute vitesse où la surface de rupture présente des stries dont la taille croît avec la vitesse. En contraste avec les données de la littérature, nous démontrons qu’il faut au minimum deux cycles de fatigue pour former une strie. Le seuil d’apparition des instabilités et le seuil de rupture catastrophique dépendent de la nature des renforts et des conditions expérimentales / In this thesis, we present an experimental study on fatigue crack growth dynamics in a natural rubber filled with silica or carbon black nanoparticles. This work has been developed in an industrial context aiming to transpose the qualities of silica filler on synthetic rubber to natural rubber, which has a strain induced crystallization behavior. The main research objective is to understand the influence of material and test parameters on fatigue fracture of rubber. Crack growth dynamics is followed by video tracking using optical and thermo-graphic cameras. The later allows us to measure the heating build-up of specimens due to cyclic loading. The morphology of rupture surfaces is characterized by post mortem observation (SEM, optical profilometry). The influence of temperature, frequency, kind of filler and filler rate has been studied. Whatever the compound or test conditions, crack growth becomes unstable starting at a certain strain level. This behavior is clearly visible on the curves of tear energy vs. crack growth rate, where two branches of crack velocity appear. In the low velocity branch, the roughness of rupture surfaces is very important because of cavities formation. In the high velocity branch, the morphology of rupture surfaces is characterized by sawtooth striations; their size increase with velocity. We demonstrate that, in contrast with literature data, at least two fatigue cycles are needed to form one striation. The instability occurrence and catastrophic rupture thresholds depend on filler type and test conditions