No description available.

The behaviour of colloidal particles in two-dimensional (2D) systems is addressed in real space and time using magnetic fields, optical tweezers and optical video microscopy. First, the fluctuations of a grain boundary in a 2D colloidal crystal are analysed. A real space analogue of the capillary fluctuation method is derived and successfully employed to extract the key parameters that characterise the grain boundary. Good agreement is also found with a fluctuation-dissipation based method recently suggested in simulation. Following on from analysis of the interface fluctuations, the properties of the individual grain boundary particles are analysed to investigate the long standing hypothesis that suggests that grain boundary particle dynamics are similar to those in supercooled liquids. The grain boundary particle dynamics display cage breaking at long times, highly heterogeneous particle dynamics and the formation of cooperatively moving regions along the interface, all typical behaviour of a supercooled liquid. Next, the frustration induced by confining colloidal particles inside a pentagonal environment is investigated. The state of the system is adjusted via two separate control parameters: the inter-particle interaction potential and the number density. A gradual crystalline to confined liquid-like transition is observed as the repulsive inter-particle interaction potential is decreased. In contrast, re-entrant orientational ordering and dynamical effects result as the number density of the confined colloidal particles is increased. Finally, the dynamics of colloidal particles distributed amongst a random array of fixed obstacle particles is probed as a function of both the mobile particle and fixed obstacle particle number densities. Increasing the mobile and the obstacle particle number density drives the system towards a glass transition. The dynamics of the free particles are shown to behave in a similar way to the normal glass transition at low obstacle density and more analogous to a localisation glass transition at high obstacle density.

For centuries, scientists have dreamed of creating materials by design. Rather than discovery by accident, bespoke materials could be tailored to fulfill specific technological needs. Quantum theory and computational methods are essentially equal to the task, and computational power is the new bottleneck. Machine learning has the potential to solve that problem by approximating material behavior at multiple length scales. A full end-to-end solution must allow us to approximate the quantum mechanics, microstructure and engineering tasks well enough to be predictive in the real world. In this dissertation, I present algorithms and methodology to address some of these problems at various length scales. In the realm of enumeration, systems with many degrees of freedom such as high-entropy alloys may contain prohibitively many unique possibilities so that enumerating all of them would exhaust available compute memory. One possible way to address this problem is to know in advance how many possibilities there are so that the user can reduce their search space by restricting the occupation of certain lattice sites. Although tools to calculate this number were available, none performed well for very large systems and none could easily be integrated into low-level languages for use in existing scientific codes. I present an algorithm to solve these problems. Testing the robustness of machine-learned models is an essential component in any materials discovery or optimization application. While it is customary to perform a small number of system-specific tests to validate an approach, this may be insufficient in many cases. In particular, for Cluster Expansion models, the expansion may not converge quickly enough to be useful and reliable. Although the method has been used for decades, a rigorous investigation across many systems to determine when CE "breaks" was still lacking. This dissertation includes this investigation along with heuristics that use only a small training database to predict whether a model is worth pursuing in detail. To be useful, computational materials discovery must lead to experimental validation. However, experiments are difficult due to sample purity, environmental effects and a host of other considerations. In many cases, it is difficult to connect theory to experiment because computation is deterministic. By combining advanced group theory with machine learning, we created a new tool that bridges the gap between experiment and theory so that experimental and computed phase diagrams can be harmonized. Grain boundaries in real materials control many important material properties such as corrosion, thermal conductivity, and creep. Because of their high dimensionality, learning the underlying physics to optimizing grain boundaries is extremely complex. By leveraging a mathematically rigorous representation for local atomic environments, machine learning becomes a powerful tool to approximate properties for grain boundaries. But it also goes beyond predicting properties by highlighting those atomic environments that are most important for influencing the boundary properties. This provides an immense dimensionality reduction that empowers grain boundary scientists to know where to look for deeper physical insights.

Thesis (Ph. D.)--Ohio State University, 2007. / Title from first page of PDF file. Includes bibliographical references (p. 164-176).

The objective of this research is to use atomistic simulations to investigate dislocation nucleation from grain boundaries in face-centered cubic aluminum and copper. This research primarily focuses on asymmetric tilt grain boundaries and has three main components. First, this research uses molecular statics simulations of the structure and energy of these faceted, dissociated grain boundary structures to show that Σ3 asymmetric boundaries can be decomposed into the structural units of the Σ3 symmetric tilt grain boundaries, i.e., the coherent and incoherent twin boundaries. Moreover, the energy for all Σ3 asymmetric boundaries is predicted with only the energies of the Σ3 symmetric boundaries and the inclination angle. Understanding the structure of these boundaries provides insight into dislocation nucleation from these boundaries. Further work into the structure and energy of other low order Σ asymmetric boundaries and the spatial distribution of free volume within the grain boundaries also provides insight into dislocation nucleation mechanisms. Second, this research uses molecular dynamics deformation simulations with uniaxial tension applied perpendicular to these boundaries to show that the dislocation nucleation mechanisms in asymmetric boundaries are highly dependent on the faceted, dissociated structure. Grain boundary dislocation sources can act as perfect sources/sinks for dislocations or may violate this premise by increasing the dislocation content of the boundary during nucleation. Furthermore, simulations under uniaxial tension and uniaxial compression show that nucleation of the second partial dislocation in copper exhibits tension-compression asymmetry. Third, this research explores the development of models that incorporate the resolved stress components on the slip system of dislocation nucleation to predict the atomic stress required for dislocation nucleation from single crystals and grain boundaries. Single crystal simulations of homogeneous dislocation nucleation help define the role of lattice orientation on the nucleation stress for grain boundaries. The resolved stress normal to the slip plane on which the dislocation nucleates plays an integral role in the dislocation nucleation stress and related mechanisms. In summary, the synthesis of various aspects of this work has provided improved understanding of how the grain boundary character influences dislocation nucleation in bicrystals, with possible implications for nanocrystalline materials.

In this dissertation, an innovative approach combining continuum mechanics and atomistic simulations is exposed to develop a nanomechanics theory for modeling and predicting the macroscopic behavior of nanomaterials. This nanomechanics theory exhibits the simplicity of the continuum formulation while taking into account the discrete atomic structure and interaction near surfaces/interfaces. There are four primary objectives to this dissertation. First, theory of interfaces is revisited to better understand its behavior and effects on the overall behavior of nanostructures. Second, atomistic tools are provided in order to efficiently determine the properties of free surfaces and interfaces. Interface properties are reported in this work, with comparison to both theoretical and experimental characterizations of interfaces. Specifically, we report surface elastic properties of groups 10 11 transition metals as well as properties for low-CSL grain boundaries in copper. Third, we propose a continuum framework that casts the atomic level information into continuum quantities that can be used to analyze, model and simulate macroscopic behavior of nanostructured materials. In particular, we study the effects of surface free energy on the effective modulus of nano-particles, nanowires and nano-films as well as nanostructured crystalline materials and propose a general framework valid for any shape of nanostructural elements / nano-inclusions (integral forms) that characterizes the size-dependency of the elastic properties. This approach bridges the gap between discrete systems (atomic level interactions) and continuum mechanics. Finally this continuum outline is used to understand the effects of surfaces on the overall behavior of nano-size structural elements (particles, films, fibers, etc.) and nanostructured materials. More specifically we will discuss the impact of surface relaxation, surface elasticity and non-linearity of the underlying bulk on the properties nanostructured materials.

The aim of this study is to develop a theoretical basis and to perform computational experiments for understanding the grain boundary (GB) grooving in polycrystalline thin film metallic conductors (interconnects) by anisotropic surface diffusion due to capillary, electromigration and elastostatic forces. To this end, irreversible thermo&ndash / kinetics of surfaces and interfaces with triple junction singularities is elaborated, and the resulting well-posed moving boundary value problem is solved using the front&ndash / tracking method. To simulate the strain conditions of the interconnects during service, the problem is addressed within the framework of isotropic linear elasticity in two dimensions (plane strain condition). In the formulation of stress induced surface diffusion, not only the contribution due to elastic strain energy density (ESED) but also that of the elastic dipole tensor interactions (EDTI) between the stress field and the mobile atomic species (monovacancies) is considered. In computation of the elastostatic and electrostatic fields the indirect boundary element method (IBEM) with constant and straight boundary elements is utilized. The resulted non&ndash / linear partial differential equation is solved numerically by Euler&rsquo / s method of finite differences. The dynamic computer simulation experiments identify well known GB groove shapes and shed light on their growing kinetics. They also allow generating some scenarios under several conditions regarding to the applied force fields and/or physicochemical parameters. The destruction of groove symmetry, termination of the groove penetration with isotropic surface diffusivity, ridge/slit formations with anisotropic diffusivity and the role played by the wetting parameter are all identified for electromigration conditions. The kinetics of accelerated groove deepening with an applied tensile stress is examined in connection with GB cavity growth models in the literature and a diffusive micro-crack formation is reported at the groove tip for high stresses. On the other hand, the use of EDTI provided a means to dynamically simulate GB ridges under compressive stress fields with surface diffusion. An incubation time for hillock growth and a crossover depth over which GB migration becomes energetically favorable is defined and discussed in this context.

No description available.

L'objectif de cette thèse est de produire une avancée dans la compréhension du phénomène de fragilisation par les métaux liquide (FML), en nous appuyant sur l'étude expérimentale et numérique du couple cuivre/mercure. La fragilisation du cuivre pur OFHC (Oxygen Free High Conductivity) par le mercure liquide est mise en évidence et quantifiée par des mesures de ténacité. En outre, un procédé d'ingénierie des joints de grains est appliqué afin d'augmenter de façon importante la proportion de joints de grains spéciaux Σ3 dans le cuivre. Des essais de FML sont alors réalisés et permettent d'établir le rôle de ces joints de grains dans la fragilisation du cuivre par le mercure liquide. En parallèle, des modélisations de joints de grains spéciaux Σ3 et Σ5 sont réalisées par calcul basés sur la théorie de la fonctionnelle de la densité (DFT). Ces modélisations permettent à la fois de mettre en évidence une réduction des propriétés mécaniques de ces joints de grains en présence d'atomes de mercure, ainsi que de comprendre l'immunité des joints Σ3 observée expérimentalement. Cependant, ces modélisations ne permettent pas de rendre compte quantitativement des observations expérimentales. Pour améliorer cette description atomique de la FML, une contribution non locale est ajoutée, via l'utilisation d'un modèle de zone cohésive nourri par calcul DFT. Il est montré que le confinement du métal liquide en extrême pointe de fissure engendre une force normale aux parois de la fissure (l'origine physique de cette force est discutée), et que l'introduction de cette nouvelle composante permet de rendre compte des observations expérimentales de façon beaucoup plus quantitative. Ce dernier modèle est appuyé par la réalisation d'expériences de FML sous pression hydrostatique. / The aim of this thesis is to make an advance in the liquid metal embrittlement (LME) understanding, based on the experimental and numerical studies of the copper/mercury system. OFHC (Oxygen Free High Conductivity) copper embrittlement by liquid mercury is studied and quantified by toughness measures. Moreover, grain boundary engineering (GBE) is implemented in order to increase the proportion of special Σ3 grain boundaries. LME tests are performed and allow to establish the particular behaviour of the Σ3 grain boundaries in the copper embrittlement by liquid mercury. At the same time, modelling of special Σ3 and Σ5 grain boundaries based on density functional theory are performed. This allows to show the weakening of mechanical properties of both grain boundaries containing mercury atoms, and also to understand the immunity of Σ3 grain boundaries as observed experimentally. However, experimental observations can not been qualitatively explained by these modelling. In order to improve this description, a non-local contribution is introduced by means of a cohesive zone model. It is shown that the confinement of the liquid metal at the very crack tip produces a force normal to the surface of the solid (the origin of this force is discussed), and that the consideration of this force allows to describe more accurately experimental results. This model is supported by LME experiments under hydrostatic pressure.

La diffusion de l’hydrogène dans les métaux est un paramètre clef pour la compréhension des mécanismes de base de la fragilisation par l’hydrogène. Néanmoins, le rôle des joints de grains dans le processus de diffusion de cet élément reste à éclaircir. Avec une approche numérique, nous avons étudié les effets d’un réseau hétérogène de joints de grains sur la diffusivité macroscopique de l’hydrogène. Pour cela, des essais de perméation ont été simulés par la méthode des éléments finis, en modélisant le matériau comme un composite formé d’une phase intra-associée aux grains et deux phases intergranulaires, avec des propriétés de diffusion différentes. Nous avons montré, en caractérisant la topologie et la connectivité du réseau des joints de grains, qu’il existe des fortes corrélations entre les paramètres de connectivité et le coefficient de diffusion effectif. Il a été démontré également que ces corrélations sont plus importantes dans les matériaux nanocristallins. De plus, en utilisant une approche d’homogénéisation, il a été mis en évidence que le coefficient de diffusion effectif est contrôlé par le caractère percolant du réseau des joints de grains, sans pour autant avoir le même seuil de percolation que ces derniers. Une seconde étude, utilisant des microstructures de nickel obtenues par cartographies EBSD, a permis d’évaluer l’écart entre la connectivité des modèles simplifiés et les structures réelles. Cette étude a permis également de confronter les coefficients de diffusion effectifs obtenus par la simulation à des données de la littérature. En parallèle, une étude expérimentale a été conduite sur le nickel pour analyser l’influence des joints de grains sur la distribution locale de l’hydrogène. Les résultats ont montré des corrélations importantes entre la nature du joint de grains et le profil de concentration de l’hydrogène autour de ce joint, qui ne peuvent pas être expliquées uniquement par le processus de diffusion. / The diffusion of hydrogen in metals is a key factor for understanding the basic mechanisms of hydrogen embrittlement. However, the contribution of grain boundaries to the hydrogen diffusion is not well established. In this this work, we first investigated the effects of a heterogeneous grain boundary networks on the effective diffusivity in polycrystalline materials, using finite elements modeling. To do so, hydrogen diffusion through heterogeneous materials, modeled by a ternary continuum composite media, was simulated. We showed, by characterizing the grain-boundary connectivity, that there are strong correlations between the grain-boundary connectivity parameters and the effective diffusivity. It was found also that these correlations are more significant for nanocrystalline materials. Moreover, by using a homogenization method, it was evidenced that the percolation behavior of the effective diffusivity is controlled by the grain-boundary network evolution, without exhibiting the same percolation threshold than the latter. A second approach, using EBSD-based microstructures, was conducted to evaluate the effect of microstructural constraints on the grain boundary connectivity and to compare the effective diffusivity numerically obtained with experimental data on polycrystalline nickel from literature. In parallel, experimental analyses were performed to analyze the effects of the grain boundaries on the local hydrogen concentration. This highlighted the significant impact of grain-boundary character on the hydrogen distribution around grain boundaries, which can not be explained by the only diffusion process.