Computer simulations are widely used in the truck industry in order to provide assistance in the product development. Bolt joints are common in trucks. A bolt fracture usually has a great influence on how a truck structure will behave in a crash. Therefore, when simulating truck crashes it is important to be able to predict when bolt fracture occurs. A material model for 10.9 bolts has been calibrated and validated by using the finite element software LS-DYNA. The material model consists of a failure strain surface, which depends on the triaxiality, Lode parameter and the element size. In this thesis, the calibrated material model is referred to as the bolt model. A good agreement to predict the force at fracture in bolts between simulation model results and physical test results has been obtained. Still, further validation is needed to evaluate the bolt model completely.

Glassy polymers are extensively used to make all kinds of structural components. Polymers, such as epoxies, are often chosen as matrices in polymer matrix composites (PMC). Ever since the 1960s, these types of composites have been gaining importance in aerospace and automotive advanced applications due to their high sti ness and weight saving potential. In order to provide clues on the dependence of the fracture behavior upon the stress triaxiality, a series of tensile tests on epoxy (Epon862) round notched bars were carried out at NASA Glenn Research Center. Using state-of-the-art non-contact dig- ital image correlation measurement technique, the mechanical quantities of interest were extracted in order to understand how the fracture behavior responds when sub- jected to various levels of stress triaxiality induced by varying the notch radius of the specimens. E ects of aging on the fracture behavior were also investigated. A physics-based macromolecular constitutive model that accounts for temperature and pressure sensitivity as well as small-strain softening and large-strain hardening was used to model the deformation behavior. Good correlation between experiments and numerical simulations was achieved. To predict fracture, a pressure-sensitive model motivated by previous work is introduced. Based on the experimental and numerical results, the relation between the mean strain to failure versus the stress triaxiality was de ned and it was shown that the fracture response of the material is strongly a ected by the level of stress triaxiality.

The mechanism of metal material failure due to inelastic cyclic deformations is commonly described as Low-Cycle Fatigue (LCF). Fracture in steel structures caused by earthquakes can be associated with this mechanism. Mathematical expressions describing the material deterioration due to LCF are often referred to as LCF laws. The accurate determination of the safety of steel structures against earthquake-induced failure requires the use of LCF laws which have been sufficiently validated with experimental test data. The present study combined experimental testing and computational simulation to enhance the understanding of structural steel fracture due to LCF. The experiments were conducted in specimens extracted from the flat and corner regions of two rectangular steel hollow sections with different thickness. A total of 60 cylindrical specimens with a circumferential notch were subjected to different combinations of axial and torsional loading. The loading protocols and notch geometry were designed to produce different stress states at the location of fracture initiation. Finite element analyses were conducted to obtain the stress state and inelastic strains at the fracture initiation location. This information was then used for the calibration of five existing LCF laws. The calibration also allowed the comparative evaluation of the capability of the different laws to capture fracture initiation for different stress states, with a single set of values for the various parameters. The accuracy of the calibrated LCF laws to predict fracture initiation in a large-scale test was also investigated. To this end, a test was conducted on a rectangular steel tube subjected to cyclic axial loading. A finite element analysis of this test was conducted, and predictions of the instant and location of fracture initiation using the calibrated LCF laws were compared with the experimental observations. / Doctor of Philosophy / The mechanism of material failure due to repeated cycles of large deformations is denoted as Low-Cycle Fatigue (LCF); this failure mechanism can occur in steel structures subjected to loading conditions such as those induced by earthquakes. Mathematical expressions that evaluate the material deterioration due to LCF are often used to predict the instant and location of fracture initiation in small-scale and large-scale tests. An experimental program was conducted for the study of fracture associated with LCF. A total of 60 specimens were fabricated with material extracted from the flat and corner regions of two rectangular steel tubes; the applied loads elongated and/or twisted the specimens until they ruptured. Computational simulations of these tests were conducted to obtain key information at the location of the observed fracture initiation. This information was used to adjust five mathematical expressions suggested by previous researchers that could predict the same instant of fracture initiation observed in the experiments. The accuracy of the predictions from each of these mathematical expressions was evaluated. The accuracy of these mathematical expressions to predict fracture initiation in a large-scale test was also investigated. To this end, an experiment was conducted on a rectangular steel tube subjected to repeated cycles of deformation. A computational simulation of this test was also developed, and predictions of the instant and location of fracture initiation were compared with the experimental observations.

Le présent travail de doctorat vise à proposer une méthode de post-traitement de la plasticité, à partir de deux calculs éléments finis en élasticité, afin de déterminer la durée de vie de structures avec concentration de contraintes, comme les disques ou les aubes de turbines. Leur dimensionnement en fatigue avec des calculs éléments finis en plasticité conduit à des temps de calcul longs, voir prohibitifs. Les méthodes de types Neuber et l’analyse par changement d’échelles sont des alternatives beaucoup moins couteuses en temps de calcul. En effet, ces méthodes permettent d’estimer la plasticité localisée uniquement à l’aide de calculs éléments finis en élasticité. Néanmoins, lorsque l’état de contrainte est multiaxial, le post-traitement de la plasticité est très imprécis. Cette imprécision provient d’une mauvaise estimation de la triaxialité des contraintes. Les méthodes de type Neuber ont été d’abord enrichies par une description plus précise de la triaxialité des contraintes. Elles ont ensuite été validées sur des structures soumises à des chargements monotones. Cependant, cette méthodologie n’est pas adaptée pour les chargements multiaxiaux alternés. Pour remédier à ce problème, une approche fondée sur les lois de changement d’échelles a été développée. Les zones de plasticité confinée sont considérées comme des inclusions. Cette approche est également enrichie par une meilleure description de la triaxialité des contraintes. Le post-traitement de la plasticité obtenue sur les bords libres de structures soumises à des chargements alternés est précis. Cette méthode a finalement été appliquée sur les cas d’un disque et d’une aube de turbine. / This work aims to propose a method to estimate plasticity from two structural analyses in elasticity in order to determine the lifetime of structures with stress concentration, such as turbine discs or turbine blades. Fatigue analyses computed by the finite element method in plasticity leads to costly computations not always suitable. Energetic methods such as Neuber’s and the scale transition approach are much less expensive alternatives. Indeed, these methods are able to estimate localized plasticity at free edges only from both computations in elasticity. However, when the stress state is multiaxial, the estimation of the plasticity is inaccurate. This inaccuracy comes from a wrong estimation of the stress triaxiality. Expressions for the stress triaxiality at free edges of elastoplastic structures are established. These expressions are derived thanks to a new heuristic that unify uniaxial tension (plane stress condition) and plane deformation in a single case. The proposed expressions for the stress triaxiality are then used to enhance energetic methods (as Neuber and Glinka rules). Notched specimens subjected to monotonic loadings are studied. However, this methodology is not able to handle alternating loadings. To deal with this problem, a new scale transition approach is developed. Localized plastic areas are considered as inclusions. This approach is enhanced by the new heuristic of multiaxiality. The elastoplastic notch tip strains and stresses of structure subjected to complex alternating loadings are accurately estimated from both elastic analyses. This method is finally applied to the case of a turbine disk and a turbine blade of a jet engine.

Le mode d’excitation collective exotique appelé chiralité a été étudié et des isomères de longue durée de vie ont été recherchés dans les noyaux ¹³⁶Nd et ¹³⁵Nd. Cinq paires de bandes Δ I=1 presque dégénérées et de même parité ont été identifiées à des spins élevés dans ¹³⁶Nd Les bandes observées ont été étudiées en utilisant la théorie de la fonctionnelle de densité covariante contrainte et rotation autour d'une axe incliné, et par un nouveau modèle de rotor plus particules développé pour décrire le couplage de particules dans quatre couches j à un rotor triaxial. Les propriétés de ces bandes doublet sont en bon accord avec les résultats des calculs théoriques. Par conséquent, l'existence de bandes chirales multiples (MχD) dans le noyau ¹³⁶Nd a été établie. Il s'agit de la première preuve expérimentale de l'existence de bandes chirales dans des noyaux pairs. Les cinq paires de bandes chirales constituent le plus grand ensablé de bandes chirales observé jusqu'à présent dans un seul noyau. De plus, le schéma de niveaux de ¹³⁶Nd a été considérablement étendu aux spins bas, moyens et très élevés. Les configurations possibles de toutes les bandes rotationnelles ont été attribuées á l'aide du modèle Nilsson-Strutinsky en rotation. La structure de ¹³⁶Nd a été clarifiée et les divers types d'excitations uni-particule et collectives ont été bien compris. Une nouvelle paire de bandes chirales á parité positive a été identifiée dans. Les caractéristiques des bandes soutiennent leur interprétation en termes de bandes chirales. Des bandes chirales à parité négative ayant été identifiées précédemment, la présence de multiple bandes chirales dans ¹³⁵Nd est ainsi établie. Les bandes chirales observées ont été étudiées en utilisant la théorie avec fonctionnelle de densité covariante contrainte et avec le modèle rotor plus particules; elles reproduisent fidèlement les données expérimentales, confirmant ainsi le phénomène MχD dans ce noyau. Les bandes chirales nouvellement observées dans ¹³⁵Nd représentent une étape importante dans la confirmation de l'existence du phénomène MχD dans les noyaux. La recherche d'états isométriques à longue durée de vie dans ¹³⁵Nd et ¹³⁶Nd a été aussi effectuée, mais n'a pas permis d'identifier de nouveaux isomères. Cependant, nous avons pu confirmer la présence des isomères déjà identifiés dans les noyaux ¹³⁸Nd, ¹³⁴Ce, ¹³⁶Pr, et ¹³⁷Pr dans nos données. / The exotic collective excitation mode called chirality has been investigated and long-lived isomers have been searched for in the ¹³⁶Nd and ¹³⁵Nd nuclei. Five pairs of nearly degenerate Δ I=1 bands with the same parity have been identified at high spins in ¹³⁶Nd. The observed bands were investigated by the constrained and tilted axis cranking covariant density functional theory and a new developed four single-j shells particle-rotor model. It was found that the properties of these doublet bands are in good agreement with results of the model calculations. Therefore, the multiple chiral doublets (MχD) phenomenon in the nucleus ¹³⁶Nd was confirmed. This was the first experimental evidence for the MχD bands in even-even nuclei. The five pairs of chiral doublet bands is the largest observed in a single nucleus until now. In addition, the level scheme of the ¹³⁶Nd has been extended significantly at low, medium, and very high spins. Possible configurations of all rotational bands have been assigned by using the cranked Nilsson-Strutinsky model. The band structure of ¹³⁶Nd was clarified and the various types of single-particle and collective excitations were well understood. A new pair of positive-parity chiral doublet bands has been identified in ¹³⁵Nd. The characteristics of the doublet bands support the chiral interpretation. Together with the previously reported negative-parity chiral doublet bands show the presence of MχD bands in ¹³⁵Nd. The observed doublet bands were compared with constrained covariant density functional theory and particle-rotor model calculations which nicely reproduce the experimental data, confirming the MχD phenomenon in this nucleus. The newly observed MχD bands in ¹³⁵Nd represent an important milestone in supporting the existence of MχD in nuclei. The search for long-lived isomeric states in ¹³⁵Nd and ¹³⁶Nd has also been performed. This did not lead to the identification of new isomers, but we could confirm the existence of the previously reported isomers in the nuclei ¹³⁸Nd, ¹³⁴Ce, ¹³⁶Pr, and ¹³⁷Pr in our data.

When developing an electric vehicle, it is essential to evaluate the deformation in and around the battery box for different crash scenarios, and it is necessary to develop a more advanced model that would take into account all the stress modes. Thanks to the excellent properties of Advanced High Strength Steel (AHSS) combine with high strength for more safety and weight reduction for less exhaust emission, AHSS is more and more commonly used in automobile industry. The material employed in this project is DOCOL 900M and it is a martensitic steel with yield strength higher than 700MPa.  The focus of the current work is to describe the experimental setup for the GISSMO model used in LS-DYNA. A number of experimental methods and theories have been reviewed. Different geometries of the test specimens under different stress triaxialities have been discussed. The study also compares the accuracy and robustness of each of the testing methods and setups. The effect of anisotropy of materials on the mechanical properties was studied. Some summaries about how to reduce errors in the experiment under the conditions of low costing and high efficiency have been discussed. According to the stress-strain response of ductile materials, the parameters of plasticity model can be calibrated. The material can be implemented in finite element software to calibrate the parameters of damage and the prediction of material failure can be achieved. The experiment and simulation are always good to be used together in the research.

Cette thèse s'intéresse à l'amélioration de la prévision de l'éclatement des pièces tournantes des turbomachines. L'axe de recherche principal a été celui de l'identification du comportement du matériau jusqu'à rupture pour des sollicitations multi-axiales représentatives et l'identification des conditions de rupture locale elles mêmes. Pour cela une approche basée sur la corrélation d'image numérique intégrée a été suivie. L'autre axe abordé dans cette thèse a été celui de la prévision objective et robuste de la rupture par la mécanique de l'endommagement dans le cadre de simulations explicites utilisant des techniques de scaling pour diminuer les temps de calcul. Deux matériaux ont été étudiés, un alliage titane TA6V et un alliage nickel Udimet 500. Des éprouvettes ont été définies pour permettre d'avoir des sollicitations proches de celles en service et pour permettre d'utiliser des techniques de mesure de champs. Un code commercial (ABAQUS) est utilisé dans une approche intégrée de la corrélation des images numériques (CIN). Le principe de ces méthodes est d'optimiser la corrélation directement à partir des paramètres du modèle sans passer par une procédure intermédiaire demandant de reconstruire préalablement les champs de déformations. Cette technique, qui peut-être vu comme très régularisante, confère à la méthode une grande robustesse, ce qui permet d'obtenir des informations même en présence de mouchetis dégradés lors des phases ultimes de chargement. L'ensemble des paramètres constitutifs des modèles peut alors être identifié avec un seul essai hétérogène. Dans l'étude un accent particulier a été mis sur l'analyse des champs de triaxialité, dont l'importance sur les conditions de rupture est suspectée, en utilisant des éprouvettes fines et des éprouvettes épaisses. Des lois de comportement ont pu être identifié pour des niveaux de déformations plastiques jusqu'à 3 fois supérieurs à ceux atteints dans des essais de traction uni-axiaux et ceci en présence d'adoucissement global. L'intégration étroite des essais et des simulations a également permis d'analyser l'état de déformation et contraintes des éléments proches de la surface de rupture juste avant celle-ci et ainsi de tester des critères de rupture potentiels avec un grand nombre de données. Pour la seconde partie de la thèse, plus exploratoire faute de temps, une étude préliminaire a porté sur la possibilité d'utiliser des modèles à taux d'endommagement limités combinés avec des techniques de scaling pour utiliser des simulations explicites dans le cadre de sollicitations centrifuges quasi-statiques. Les premiers résultats montrent qu'une accélération est possible avec des niveaux d'erreurs relativement bas sur les vitesses de rotation conduisant à la rupture et ceci dans des temps comparables avec des simulations implicites quasi-statique, l'avantage étant la robustesse des analyses en dynamique explicite. / This thesis addresses the amelioration of the burst prediction of turbo-engines rotating parts.The principal axis of this study was the identification of the material behaviour up fracture under multi-axial loading and the identification of the local condition at failure. To accomplish this, an integrated digital image correlation approach was employed.The other thematic axis of this thesis was the objective and robust prediction of rupture using the damage mechanics in the framework of explicit simulations and the study of the influence, on the prediction, of various scaling techniques used to decrease the computation time.Two materials were studied, the titanium alloy TA6V and the nickel alloy Udimet 500. The samples were defined to guarantee similar loading conditions as those in the disks in service and to allow the use of field measurement techniques.A commercial code (ABAQUS) is used in an integrated approach to Digital Image Correlation (DIC). The principle of this method is to optimize the correlation of the test directly with the model parameters without passing by an intermediate procedure requiring the reconstruction of the deformation field beforehand. This technique, which can be seen as highly regularising, grants a considerable robustness to the method, which allows obtaining information from degraded speckle patterns encountered at the end of the test. The entire set of constitutive parameters can be identified with only one heterogeneous test. Within this study, a particular accent was laid on the analysis of triaxiality fields by using thin and thick samples. The importance of these fields for the fracture conditions is suspected.Constitutive models were identified for plastic strains that were about three times higher than the ones achievable in uniaxial tensile test on smooth sample.The close integration of the experiments and simulations allowed the analysis of deformation and stresses of the elements near the failure surface at the instant just before failure. This allowed the testing of different criteria with a large amount of data.For the second axis of the thesis, preliminary studies examined the possibility to use limited damage rate models combined with scaling techniques. These latter allow performing explicit simulations in context of a centrifugal quasi-static loading. The results show that acceleration is possible. The errors of the rotational speeds, which lead to disk fracture, are relatively low. The simulation times are comparable to implicit quasi-static simulations, while the main advantage is the maintained robustness in explicit simulations.

No description available.

Le présent manuscrit de thèse regroupe les études par spectroscopie γ des noyaux 84−88Se produits par une réaction de fission induite par neutrons thermiques, sur une cible de 235U. Cette expérience a été réalisée à l’Institut Laue-Langevin à Grenoble après du multi-détecteurs EXILL. Les données brutes recueillies par EXILL ont été traitées et analysées afin de déterminer la structure des noyaux étudiés. Grâce à cette étude spectroscopique détaillée, un premier schéma de niveaux a été établi pour la première fois au 88Se et de nouveaux états excités ont été assignées à d’autre isotopes de sélénium (84-87Se). De plus, l’analyse de la corrélations angulaires a été faite pour confirmer l’assignation de spins des premières états excités. Des calculs théoriques utilisant le modèle en couches (SM), le modèle collectif algébrique (ACM) et le modèle quasiparticule plus phonon (QPM), ont été effectués pour interpréter la structure nucléaire et la forme du noyau d’intérêt 88Se. Selon ces modèles, le noyau 88Se est γ-instable et peut être interprété comme un noyau transitoire et sa structure nucléaire du 88Se a été bien reproduite par le modèle en couches et le modèle ACM. Les calculs du modèle QPM indiquent que le degré de collectivité de l’état 22+ du 88Se est faible parce que une composante neutronique reste dominante dans la structure de la bande γ. / The present thesis manuscript includes the nuclear spectroscopy studies of 84−88Se. A more favourable fission reaction for their study is 235U(nth, f). The measurement was performed at the PF1B cold-neutron beam facility of the Institut Laue-Langevin (ILL), Grenoble. The fission reaction has been used in conjunction with the EXILL Ge array to study excited states in the neutron-rich Se nuclei. Thanks to this spectroscopic study, a first level scheme of 88Se was established and new excited states were assigned to other selenium isotopes. Furthermore, the angular correlation studies were carried out to confirm the spin assignment of the first excited states in the 84-88Se nuclei. The theoretical calculations using the shell model (SM), the algebraic collective model (ACM) and the quasiparticle plus phonon model (QPM) were performed in order to interpret the nuclear structure and the shape of the nucleus of interest 88Se. According to these models, the nucleus 88Se is γ-unstable and can be interpreted as a transitional nucleus. The shell-model calculations using a 78Ni core and the ACM calculations reproduce the decay scheme of 88Se well. The QPM calculations show that the collectivity of second state 2+2 of 88Se is weak and that this state contains a strong quasiparticle component.

The aim of this thesis is the construction of a diagram of fracture strain for high strength steel OCHN3MFA and its application for the estimation of fracture toughness of this material. The first part of the thesis contains the necessary theoretical framework for numerical modelling of tensile tests of various specimens - smooth specimen, specimen with a notch and specimen with a circumferential crack, it describes the influence of triaxiality on the fracture of bodies and the possibilities of construction of the diagram of fracture strain. Subsequently, a numerical model of these specimens is created using the finite element method (FEM). A non-linear, elastoplastic calculation is performed. Fracture is identified by means of comparing the true stress versus the strain obtained from tests with the finite element analysis results. Stress–strain states of specimens at the moment of fracture are obtained from a numerical model. A diagram of fracture strain is constructed, and it is used to estimate the fracture toughness of a cracked body. Finally, a local approach, which uses the diagram of fracture strain, and a classical approach of fracture mechanics, especially the stress intensity factor, are compared.