La récupération d’énergie thermique peut être réalisée par de nombreuses techniques de transduction d’énergie. Les techniques directes de conversion d’énergie thermique en énergie électrique sont généralement les technologies les plus utilisées. Lorsque des générateurs miniaturisés son requis, des méthodes directes de conversion présentent des difficultés, y compris la nécessité des dissipateurs de chaleur volumineux ou la forte dépendance aux fluctuations de température rapides. Donc, les méthodes de conversion indirecte, comme la conversion d’énergie thermique à mécanique et puis mécanique à électrique sont présentées comme des alternatives aux récupérateurs d’énergie. Cette technologie ouvre une nouvelle ligne de recherche pour surmonter les contraintes des récupérateurs d’énergie à petite échelle. Même si leur rendement est relativement faible en raison des pertes liées aux étapes de conversion d’énergie, les générateurs d’énergie basés sur l’effet thermomagnétique présentent une densité de puissance élevée lors de leur miniaturisation. Néanmoins, peu de recherches sur la récupération d’énergie thermomagnétique à petite échelle ont été menées et aucune étude de faisabilité industrielle n’a été signalée jusqu’à présent. Ces travaux présentent la conception d’un générateur capable de convertir de faibles et de lentes fluctuations de température ambiante en électricité. L’effet thermomagnétique d’un matériau magnétique doux, à savoir l’alliage de fer et de nickel (FeNi) ainsi que la piézoélectricité sont la base de fonctionnement du dispositif. Cette thermo-magnétisation entraîne la conversion d’énergie thermique, sous la forme de fluctuations temporelles, en vibrations mécaniques d’une structure. La structure consiste en un bimorphe piézoélectrique (PZT). Le générateur a deux positions stables; la position ouverte et celle fermée. En modifiant la température de FeNi, l’interaction entre deux forces du système (forces magnétique et mécanique) amène le générateur à l’une de ses deux commutations. La température de Curie du FeNi étant proche de la température ambiante, des applications comme des dispositifs connectés portables peuvent être ciblées. Un modèle analytique est développé. Donc, une conception rapide du générateur est réalisée pour répondre aux cahiers des charges tels que: la température d’opération, la plage de températures, la réponse thermique, les capacités de conversion piézoélectrique, etc. De plus, des règles de conception ont été dérivées envers la réduction de la taille du générateur. Des modélisations par éléments finis sont développés sous ANSYS afin de valider notre modèle analytique simplifié. Ces modèles permettent aux concepteurs d’explorer d’autres matériaux et de faire des améliorations en utilisant des processus d’optimisation de la conception. Des prototypes des récupérateurs d’énergie atteignent des densités de puissance de 0.6μWcm^−3 pendant des commutations d’ouverture à 40°C et 0.02μWcm^−3 pendant des commutations de fermeture à 28°C. En réduisant la taille du générateur, des commutations d’ouverture à 31°C et des commutations de fermeture à 27°C, sont atteints. La distance initiale de séparation entre l’aimant permanent et l’alliage magnétique doux est identifiée comme une clé pour augmenter la capacité de conversion d’énergie du générateur. Un modèle équivalent électrique du générateur est développé afin de concevoir un circuit d’extraction d’énergie ainsi qu’un module de gestion d’énergie. Ce circuit est développé sous PSpice, permettant de mettre en œuvre des pertes liées aux matériaux (pertes mécaniques et diélectriques). Par le biais d’ajustement de courbe, ce modèle est capable de calculer des valeurs de pertes. Une analyse de la variabilité de la conception est réalisée afin d’explorer la faisabilité industrielle d’un tel générateur. Ainsi, la récupération d’énergie thermomagnétique peut concourir, pour la première fois, avec les thermo-générateurs les plus modernes. / Thermal energy harvesting can be realized by numerous techniques of energy transduction. Direct conversions of thermal to electrical energy are typically the most popular technologies used. When miniaturized generators are required, direct conversion methods present difficulties, including the need of bulky heat sinks or the strong dependence to rapid temperature fluctuations. Therefore, indirect conversion methods, like thermal-to-mechanical-to-electrical energy are presented as an alternative to thermal energy harvesters towards powering autonomous sensors. This disruptive technology opens up a new approach to overcome the limitations of miniaturized thermal energy harvesting systems. Even if having a relatively low efficiency due to losses linked to energy conversion steps, energy harvesters based on thermo-magnetic effect show a large power density upon miniaturization. Nevertheless, little research on thermo-magnetic energy harvesting at miniature scale has been conducted and no competitive electrical output has been reported until now.This work presents the design of a generator able to convert small and slow ambient temperature fluctuations into electricity. It exploits the thermo-magnetic effect of a soft magnetic material, namely, iron nickel alloy (FeNi) and piezoelectricity. Thermo-magnetization of FeNi is driving the conversion of thermal energy, in the form of temporal fluctuations, into mechanical vibrations of a structure. The structure consists in a piezoelectric bimorph (PZT) cantilever beam. The generator has two stable positions; open position and closed one. Curie temperature of FeNi being near to ambient temperature, applications like wearable connected devices may be targeted. By changing the temperature of the soft magnetic alloy, the interaction between counterbalance forces (magnetic and mechanical forces) leads the generator to one of its two commutations.Analytical model is developed in order to predict generator performance. Making use of this model, a rapid design of generator is conducted to fit custom requirements such as: temperature of operations, temperature range of operation, thermal response, piezoelectric energy conversion capabilities, etc.Additionally, main design rules were derived from the design parameters of the generator. Special attention was paid on how scaling down size affects the generator performance by using the analytical model.Finite element models are developed through ANSYS software in order to validate the analytical simplified model. They couple the thermal to magnetic field and then mechanical to electrical energy conversion is solved. This model allows designers to explore other materials and do improvements by using design optimization processes.First generation energy harvesting demonstrators achieve power densities of 0.6µWcm^-3 during opening commutations around 40°C and 0.02µWcm^-3 at closing commutations around 28°C. By reducing the generator’s size opening commutations at 31°C while closing commutations at 27°C are achieved. By modifying design parameters such as initial distance of separation between the permanent magnet and soft magnetic alloy is identified as a key to boost the energy conversion capability of the generator. Finally, electrical equivalent model of this thermo-magnetically activated piezoelectric generator is developed to design an energy extraction circuit and power management module. This circuit is developed in a unique software PSpice, to implement losses linked to materials (mechanic and dielectric losses). Making use of curve fitting processes, this model is able to find losses values. A variability analysis of the design is conducted by using the analytical model through Matlab in order to explore the feasibility of producing such a generator industrially. Thus, thermo-magnetic energy harvesting can compete for the first time with the state-of-the-art thermos-electrics.

Development in urban areas around the world has steadily increased in recent years. This rapid development has not been matched by the ever decreasing open space commonly associated with urban centers. Vertical construction, thus, lends itself a very useful solution to this problem. Deep excavation is often required for urban construction. Unfortunately, the ground movements associated with deep excavation can result in damage to adjacent buildings. Thus, it is critically important to accurately predict the damage potential of nearby deep excavations and designing adequate support systems. A new design method is proposed, as an attempt, to address the problem. The method is semi-empirical and directly links excavation-induced distortions experienced by nearby buildings and the components of the excavation support system. Unlike, the traditional limit equilibrium approach, the method is driven by the distortions in adjacent buildings. It goes further to propose a preliminary cost chart to help designers during the design phase. The benefit is that initial cost is known real time and will help speed up making business decisions. A new design flowchart is proposed to guide the designer through a step-by-step procedure. The method is validated using 2D Plaxis (the finite element program) simulation. Though the nature of deep excavation is three-dimensional, a plane strain condition is valid when the length of the excavation is long. Hence, two-dimensional finite element simulation was considered appropriate for this effort. Five hypothetical cases were compared and the model performed very well. The lack of available literature on this approach made verification difficult. It is hoped that future case histories will be used to ascertain the veracity of the deformation-based design method.

Le récent développement de la Fabrication Additive de pièces métalliques permet d'élaborer directement des structures à partir de modèles 3D. En particulier, la technologie "Electron Beam Melting" (EBM) permet la fusion sélective, couche par couche, de poudres métalliques. Elle autorise la réalisation de géométries très complexes mais apporte de nouvelles contraintes de fabrication.Ce travail se concentre sur la caractérisation géométrique et mécanique de structures treillis produites par cette méthode. Les pièces fabriquées sont comparées au design initial à travers des caractérisations par tomographie aux rayons X. Les propriétés mécaniques sont testées en compression uni-axiale. Pour les poutres de faibles épaisseur, la différence entre la structure numérique et celle fabriquée devient significative. Les écarts au design initial se traduisent pour chaque poutre par un concept de matière mécaniquement efficace. D'un point de vue modélisation, ce concept est pris en compte en remplaçant la poutre fabriquée par un cylindre avec un diamètre mécaniquement équivalent. Ce diamètre équivalent est utilisé dans des simulations et optimisations "réalistes" intégrant ainsi les contraintes de fabrication de la technologie EBM.Différentes stratégies sont aussi proposées pour réduire la proportion de volume "inefficace" et améliorer le contrôle de la taille des poutres, soit en jouant sur les paramètres procédé et les stratégies de fusion, soit en effectuant des post-traitements. / The recent development of Additive Manufacturing for the fabrication of metallic parts allows structures to be directly manufactured from 3D models. In particular, the "Electron Beam Melting" (EBM) technology is a suitable process which selectively melts a powder bed layer by layer. It can build very complex geometries but brings new limitations that have to be quantified.This work focuses on the structural and mechanical characterization of lattice structures produced by such technology. The structural characterization mainly rely on X-ray tomography whereas mechanical properties are assessed by uni-axial compression. The geometry and related properties of the fabricated structures are compared with the designed ones. For small strut size, the difference between the designed structure and the produced one is large enough to impact the desired mechanical properties. The concept of mechanical efficient volume is introduced. For the purpose of simulation, this concept is taken into account by replacing the struts by a cylinder with a textit{mechanical equivalent diameter}. After validation, it has been used into "realistic" simulation and optimization procedures, thus taking into account the process constraints.Post-treatments (Chemical Etching and Electro-Chemical Polishing) were applied on lattice structures to get rid of the inefficient matter by decreasing the surface roughness. The control of the size of the fabricated struts was improved by tuning the process strategies and parameters.

No presente trabalho, a conformabilidade do aço inoxinoxidável AISI 201 foi comparada com o aço inoxidável AISI 304. O aço inoxidável AISI 201 é uma liga baixo níquel ligado com manganês e nitrogênio. Nesse estudo a conformabilidade dos dois materiais foi examinada por meio de ensaios tecnológicos como ensaio de tração, determinação da curva de escoamento, determinação do índice de anisotropia, ensaio Erichsen, dureza, composição química, simulação computacional do processo de estampagem e estampagem dos blanks. O aço inoxidável AISI 201 possui propriedades como limite de escoamento e tensão de ruptura superior as do inoxidável AISI 304. Os dois aços possuem uma similaridade na anisotropia. A máxima altura alcançada no momento da fratura pelo ensaio Erichsen também é semelhante para os dois materiais. O aço inox AISI 201 apresenta uma dureza maior que o inoxidável AISI 304. Na composição química os dois aços apresentam elementos fora do especificado, caracterizando um problema de qualidade na fabricação desses aços. A simulação computacional do processo de estampagem apresentou uma redução de espessura na região mais critica, sem comprometer a estampagem do componente. Uma vez estampadas, obteve-se peças sem indícios de trincas, conforme previsto pela simulação computacional. Os resultados desse trabalho sugerem a possibilidade de utilização do aço inoxidável AISI 201 como opção para substituição ao inoxidável AISI 304, tendo uma observação a ser feita quanto aos cuidados na qualidade durante a fabricação do aço para que atenda os padrões exigidos. / In this study, the formability of stainless steel AISI 201 was compared to stainless steel AISI 304 stainless steel AISI 201 is a low alloy nickel alloyed with manganese and nitrogen. In this study, the formability of the two materials was examined by means of technological tests such as tensile test, determination of the flow curve, determining the anisotropy index, Erichsen test, hardness, chemical composition, computer simulation of the stamping process and stamping of the blanks. Stainless steel AISI 201 has properties such as yield strength and higher breakdown voltage of the stainless steel AISI 304. The two steels have a similarity in anisotropy. The maximum height reached at the time of fracture by Erichsen test is also similar for the two materials. Stainless steel AISI 201 has a hardness greater than the stainless steel AISI 304. In chemistry the two steels have elements outside the specified, featuring a quality problem in manufacturing these steels. A computer simulation of the printing process showed a reduction in thickness in the most critical region, without compromising the component stamping. Once stamped, gave no broken pieces of evidence as provided by the computer simulation. The findings suggest the possibility of use of stainless steel AISI 201 as an option to replace the stainless steel AISI 304, with a point to be made about the care as during the manufacture of steel that meets the required standards.

Les traitements de surface mécaniques sont appliqués dans la plupart des secteurs industriels comme procédé de finition afin de renforcer les propriétés des composants métalliques. Le grenaillage de précontrainte est probablement l’un des plus répandu. Ce procédé introduit des contraintes résiduelles de compression en générant un gradient de déformation plastique dans la profondeur de la pièce traitée. L’objectif de ce travail est de comprendre et prédire l’effet de la géométrie des composants sur la redistribution des contraintes résiduelles post-grenaillage. En effet, même lorsqu’elle est maîtrisée, l’opération de grenaillage peut générer un champ de contraintes résiduelles complexe qui dépend fortement de la géométrie de la pièce. Par suite, parmi les paramètres influents sur le comportement en fatigue des composants grenaillés, le paramètre géométrique peut donc avoir un rôle majeur. Puisque les approches conventionnelles de modélisation ne sont pas transposables aux géométries non planes, et ne sont pas conformes aux contraintes industrielles en termes de temps de calcul, une méthodologie basée sur la Méthode de Reconstruction des Eigenstrains est proposée. L’approche développée est construite à partir de relations analytiques pour des massifs plans traités de façon homogène. La principale contribution est la comparaison entre modélisation et expérimentation. Les données expérimentales sont obtenues à partir d’analyses de la microstructure et par diffraction des rayons X réalisées sur des échantillons d’un superalliage base nickel, pour plusieurs géométries complexes élémentaires (plaques minces, formes convexes et concaves). Par ailleurs cette étude vise à prendre en compte l’effet des contraintes résiduelles équilibrées sur la durée de vie en fatigue. A partir du critère de fatigue multiaxial de Crossland, la méthodologie complète est appliquée à des démonstrateurs industriels à géométrie complexe. / Most manufacturing industries perform mechanical surface treatments at the end of the manufacturing chain to reinforce relevant working parts. Shot peening is probably the most common of those processes. This treatment induces compressive residual stresses by generating in-depth plastic strains. The objective of this work is to understand and predict the effect of the geometry on the redistribution of residual stresses into shot peened mechanical parts. Indeed, even when properly controlled, shot peening treatment may induce a complex residual stress field depending on the geometry of the treated part. Hence, among the variables which affect the fatigue behaviour of shot peened components, the geometry could play a major role. Because the traditional approaches for the modelling of residual stresses are not convenient for complex non-flat geometries and not consistent with industrial constraints in terms of computing time, a methodology based on the Eigenstrains Reconstruction Method is proposed. The developed approach is built with analytical relationships for massive and plane geometries homogeneously treated. The main contribution lies in the capacity to provide a comparison between modelling and experiment. Experimental data are obtained by microstructural observation and by X-ray diffraction analyses, which are carried out on Ni-based superalloy samples with elementary complex geometries (thin sheets, convex and concave shapes). In addition, this study aims to take into account the effect of the rebalanced residual stresses for fatigue life prediction. Thus, using a Crossland criterion for high cycle fatigue regime, the complete methodology is applied on industrial demonstrator samples with complex geometry.

No presente trabalho, a conformabilidade do aço inoxinoxidável AISI 201 foi comparada com o aço inoxidável AISI 304. O aço inoxidável AISI 201 é uma liga baixo níquel ligado com manganês e nitrogênio. Nesse estudo a conformabilidade dos dois materiais foi examinada por meio de ensaios tecnológicos como ensaio de tração, determinação da curva de escoamento, determinação do índice de anisotropia, ensaio Erichsen, dureza, composição química, simulação computacional do processo de estampagem e estampagem dos blanks. O aço inoxidável AISI 201 possui propriedades como limite de escoamento e tensão de ruptura superior as do inoxidável AISI 304. Os dois aços possuem uma similaridade na anisotropia. A máxima altura alcançada no momento da fratura pelo ensaio Erichsen também é semelhante para os dois materiais. O aço inox AISI 201 apresenta uma dureza maior que o inoxidável AISI 304. Na composição química os dois aços apresentam elementos fora do especificado, caracterizando um problema de qualidade na fabricação desses aços. A simulação computacional do processo de estampagem apresentou uma redução de espessura na região mais critica, sem comprometer a estampagem do componente. Uma vez estampadas, obteve-se peças sem indícios de trincas, conforme previsto pela simulação computacional. Os resultados desse trabalho sugerem a possibilidade de utilização do aço inoxidável AISI 201 como opção para substituição ao inoxidável AISI 304, tendo uma observação a ser feita quanto aos cuidados na qualidade durante a fabricação do aço para que atenda os padrões exigidos. / In this study, the formability of stainless steel AISI 201 was compared to stainless steel AISI 304 stainless steel AISI 201 is a low alloy nickel alloyed with manganese and nitrogen. In this study, the formability of the two materials was examined by means of technological tests such as tensile test, determination of the flow curve, determining the anisotropy index, Erichsen test, hardness, chemical composition, computer simulation of the stamping process and stamping of the blanks. Stainless steel AISI 201 has properties such as yield strength and higher breakdown voltage of the stainless steel AISI 304. The two steels have a similarity in anisotropy. The maximum height reached at the time of fracture by Erichsen test is also similar for the two materials. Stainless steel AISI 201 has a hardness greater than the stainless steel AISI 304. In chemistry the two steels have elements outside the specified, featuring a quality problem in manufacturing these steels. A computer simulation of the printing process showed a reduction in thickness in the most critical region, without compromising the component stamping. Once stamped, gave no broken pieces of evidence as provided by the computer simulation. The findings suggest the possibility of use of stainless steel AISI 201 as an option to replace the stainless steel AISI 304, with a point to be made about the care as during the manufacture of steel that meets the required standards.

No presente trabalho, a conformabilidade do aço inoxinoxidável AISI 201 foi comparada com o aço inoxidável AISI 304. O aço inoxidável AISI 201 é uma liga baixo níquel ligado com manganês e nitrogênio. Nesse estudo a conformabilidade dos dois materiais foi examinada por meio de ensaios tecnológicos como ensaio de tração, determinação da curva de escoamento, determinação do índice de anisotropia, ensaio Erichsen, dureza, composição química, simulação computacional do processo de estampagem e estampagem dos blanks. O aço inoxidável AISI 201 possui propriedades como limite de escoamento e tensão de ruptura superior as do inoxidável AISI 304. Os dois aços possuem uma similaridade na anisotropia. A máxima altura alcançada no momento da fratura pelo ensaio Erichsen também é semelhante para os dois materiais. O aço inox AISI 201 apresenta uma dureza maior que o inoxidável AISI 304. Na composição química os dois aços apresentam elementos fora do especificado, caracterizando um problema de qualidade na fabricação desses aços. A simulação computacional do processo de estampagem apresentou uma redução de espessura na região mais critica, sem comprometer a estampagem do componente. Uma vez estampadas, obteve-se peças sem indícios de trincas, conforme previsto pela simulação computacional. Os resultados desse trabalho sugerem a possibilidade de utilização do aço inoxidável AISI 201 como opção para substituição ao inoxidável AISI 304, tendo uma observação a ser feita quanto aos cuidados na qualidade durante a fabricação do aço para que atenda os padrões exigidos. / In this study, the formability of stainless steel AISI 201 was compared to stainless steel AISI 304 stainless steel AISI 201 is a low alloy nickel alloyed with manganese and nitrogen. In this study, the formability of the two materials was examined by means of technological tests such as tensile test, determination of the flow curve, determining the anisotropy index, Erichsen test, hardness, chemical composition, computer simulation of the stamping process and stamping of the blanks. Stainless steel AISI 201 has properties such as yield strength and higher breakdown voltage of the stainless steel AISI 304. The two steels have a similarity in anisotropy. The maximum height reached at the time of fracture by Erichsen test is also similar for the two materials. Stainless steel AISI 201 has a hardness greater than the stainless steel AISI 304. In chemistry the two steels have elements outside the specified, featuring a quality problem in manufacturing these steels. A computer simulation of the printing process showed a reduction in thickness in the most critical region, without compromising the component stamping. Once stamped, gave no broken pieces of evidence as provided by the computer simulation. The findings suggest the possibility of use of stainless steel AISI 201 as an option to replace the stainless steel AISI 304, with a point to be made about the care as during the manufacture of steel that meets the required standards.

Never before has the car industry been as challenging, interesting, and demanding as it is today. New and advanced techniques are being continuously introduced, which has led to increasing competition in an almost ever-expanding car market. As the pace and complexity heightens in the car market, manufacturing processes must advance at an equal speed. An important manufacturing process within the automotive industry, and the focus of this thesis, is sheet metal forming (SMF). Sheet metal forming is used to create door panels, structural beams, and trunk lids, among other parts, by forming sheets of metal in press lines with stamping dies. The SMF process has been simulated for the past couple of decades with finite element (FE) simulations, whereby one can predict factors such as shape, strains, thickness, springback, risk of failure, and wrinkles. A factor that most SMF simulations do not currently include is the die and press elasticity. This factor is handled manually during the die tryout phase, which is often long and expensive. The importance of accurately representing press and die elasticity in SMF simulations is the focus of this research project. The research objective is to achieve virtual tryout and improved production support through SMF simulations that consider elastic die and press deformations. Loading a die with production forces and including the deformations in SMF simulations achieves a reliable result. It is impossible to achieve accurate simulation results without including the die deformations. This thesis also describes numerical methods for optimizing and compensating tool surfaces against press and die deformations. In order for these compensations to be valid, it is imperative to accurately represent dies and presses. A method of measuring and inverse modeling the elasticity of a press table has been developed and is based on digital image correlation (DIC) measurements and structural optimization with FE software. Optimization, structural analysis, and SMF simulations together with experimental measurements have immense potential to improve simulation results and significantly reduce the lead time of stamping dies. Last but not least, improved production support and die design are other areas that can benefit from these tools. / Aldrig tidigare har bilindustrin varit så utmanande, intressant och spännande som idag. Ny och avancerad teknik introduceras i en allt snabbare takt vilket leder till ständigt ökande konkurrens på en, nästan ständigt, ökande bilmarknad. Den ständigt ökande komplexiteten ställer även krav på tillverkningsprocesserna. En viktig process, som denna licentiatuppsats fokuserar på, är pressning av plåt. Tillverkningstekniken används för att forma plåtar till dörrpaneler, strukturbalkar, motorhuvar, etc. Plåtar formas med hjälp av pressverktyg monterade i plåtformningspressar. Plåtformningsprocessen simuleras sedan ett par decennium tillbaka med Finita Element (FE) simuleringar. Man kan på så sätt prediktera form, töjningar, tjocklek, återfjädring, rynkor, risk för försträckning och sprickor m.m. En faktor som för tillfället inte inkluderas i näst intill alla plåtformningssimuleringar är elastiska press- och verktygsdeformationer. Detta hanteras istället manuellt under, den oftast långa och dyra, inprovningsfasen. Detta projekt har visat på vikten av att representera press och verktygsdeformationer i plåtformningssimuleringar. Detta demonstreras genom en analys av ett verkligt pressverktyg som belastas med produktionskrafter. Det är inte möjligt att uppnå bra simuleringsresultat utan att inkludera verktygsdeformationer i simuleringsmodellen. Uppsatsen beskriver även numeriska metoder för att optimera och kompensera verktygsytor mot press och verktygsdeformationer. För att dessa kompenseringar ska stämma är det viktigt att man representerar både verktyg och press på ett korrekt sätt. Förslag på en metod för att mäta och inversmodellera pressdeformationer har utvecklats, metoden är baserad på mätningar med DIC-systemet ARAMIS och optimering i FE-mjukvaror. Optimering, strukturanalys, och plåtformningsanalys tillsammans med experimentella mätningar har en stor potential att förbättra plåtformningssimuleringar samt reducera ledtiden för pressverktyg. Sist men inte minst, andra positiva effekter är en enklare och smidigare konstruktionsprocess och förbättrad produktionssupport.

Les directives ROHS et WEEE banniront, dans les années qui viennent, le plomb de l’industrie électronique. Seulement, les assemblages électroniques de Schlumberger destinés à des applications hautes températures, tels que les ceux faisant intervenir des composants montés en surface, font intervenir des brasures à forte teneur en plomb. C’est pourquoi, Schlumberger investit énormément afin de trouver de nouvelles brasures sans plomb pour les remplacer. Ce projet, qui s’inscrit dans ce cadre, a pour objectif de prédire la durée de vie d’assemblages utilisant ces nouvelles brasures avec un substrat et des composants en céramique. Cette prédiction se fait en deux étapes. La première est expérimentale. Les assemblages sont soumis à des vieillissements accélérés thermomécaniques et isothermes. En plus de la durée de vie, ces tests apportent des connaissances sur les effets du vieillissement, sur les mécanismes et les zones de défaillances, sur l’interaction de ces brasures avec les finitions du substrat et des composants et enfin sur l’évolution de la microstructure et des phases d’intermétalliques créées lors du vieillissement.La seconde étape est la modélisation de ces assemblages afin de comprendre leur comportement sous sollicitations thermomécaniques. Les simulations aident à comprendre les phénomènes locaux qui apparaissent dans l’assemblage et à extraire des paramètres de fatigue pour diverses conditions thermomécaniques. Enfin, une corrélation entre les résultats de défaillance expérimentaux et la fatigue calculée grâce à la simulation va permettre d’estimer la durée de vie des assemblages pour différentes sollicitations thermomécaniques. Les simulations permettent donc de diminuer le nombre d’essais expérimentaux souvent chers et longs. Seulement, pour faire des simulations fiables, il est nécessaire de connaitre les paramètres mécaniques de tous les matériaux. Pour la brasure, cela veut dire le comportement élastique, plastique et en fluage. Donc, un bénéfice supplémentaire pour Schlumberger est la détermination de ses paramètres pour les nouvelles brasures. / Because of ROHS or WEEE directives, in a close future, lead materials will be banned from electronicindustry. Unfortunately, Schlumberger is using high-lead content solders for surface mount devices forhigh temperature applications. Considering this issue, Schlumberger puts in place high amount of investments to replace these solders by lead-free solders. The topic of the work is to study lead free candidates destined to support Schlumberger high temperature mission profiles. The device under test chosen for this project is a surface mount device composed of a passive component connected to a ceramic substrate by solder joints. The predictive study of reliability of these new assemblies for high temperature applications needs two complementary analyses. The first study is to characterize, experimentally, the life time of surface mount assemblies using these new lead free solders submitted to accelerated thermomechanical and isothermal aging tests. Hence, the first benefits for Schlumberger are knowledge on thecompatibility of these new alloys with their current finishes with the microstructure and intermetallic compounds evolution. More over, the main effects due to aging are investigated like failure sites and mechanisms. The second goal of the project is to perform thermo-mechanical simulations of surface mount assembly under thermal cycling. Simulations help to understand local phenomena and estimatefatigue parameters under other thermal conditions. Then, a correlation between experimental results about failure and calculated fatigue leads to an estimation of the life time of the assemblies. Thus, simulations have the advantage to reduce the number of time-consuming and expensive thermo-mechanical agingtests. To perform a simulation, the physical parameters of each solder material are needed like elastic,plastic and creep data. Additional benefits for Schlumberger involve mechanical properties which are, at the moment, unknown for these new high temperature lead free materials.

The level of residual stresses in welded components is known to have a significant influence on their failure behaviour. It is, therefore, necessary to understand the combined effect of mechanical loading and residual stresses on the ductile fracture behaviour of these structures in order to provide the accurate structural safety assessment. Recently, STYLE (Structural integrity for lifetime management-non-RPV component) performed a large scale bending test on a welded steel pipe containing a circumferential through-thickness crack (the MU2 test). The purpose of this test is to study the impact of high magnitude weld residual stresses on the initiation and growth of cracks in austenitic stainless steels. This research presents the simulation part of the STYLE project which aims to develop the finite element model of MU2 test in ABAQUS to enhance the understanding and ability to predict the combined influence of mechanical loading and residual stresses on the ductile fracture behaviour of nuclear pressure vessel steels. This research employs both fracture mechanics principles (global approach) and Rousselier damage model (local approach) to study this behaviour including crack initiation and growth. In this research, the Rousselier model was implemented into ABAQUS via the user defined subroutines for ABAQUS/Standard and ABAQUS/Explicit modules, i.e. UMAT and VUMAT. The subroutines were developed based on the integration algorithm proposed by Aravas and Zhang. The validation of these subroutines was checked by comparing the FE results obtained from the implementation of these subroutines with the analytical and other benchmark solutions. This process showed that UMAT and VUMAT provide accurate results. However, the UMAT developed in this work shows convergence problems when the elements start to fail. Hence, only VUMAT was used in the construction of the finite element model of the MU2 test. As mentioned above, the results obtained from both fracture mechanics approach and Rousselier model are compared with the experimental data to validate the accuracy of the model. The results shows that both fracture mechanics approach and the Rousselier model predict similar final crack shapes which correspond closely to the test results in south direction. The other conclusions about the influence of residual stress on ductile fracture obtained from this work are also summarized in this thesis.