[pt] Métodos de aprendizado de máquina vêm sendo bastante utilizados na área de mecânica dos sólidos devido ao grande volume de dados disponíveis na literatura. A motivação deste trabalho foi o estudo do acúmulo de deformação plástica na escala de grãos, pois o uso do aprendizado de máquina pode ser uma significativa contribuição para criar modelos capazes de prever o acúmulo de deformações. O objetivo deste trabalho foi aprimorar a previsão do acúmulo de deformação plástica propondo um novo método de previsão de acúmulo de deformações plásticas em contornos de grãos de um material policristalino, usando modelos de aprendizado de máquina. Este trabalho utilizou-se de dados experimentais da literatura para estruturar três bancos de dados, os que consideraram somente os contornos de grãos. Nas previsões foram utilizados os seguintes métodos: Decision Tree, Random Forest, Stochastic Gradient Descent, K-Nearest Neighbors, Gradient Boosting Regressor e Análise de Componentes Principais (PCA). Na avaliação dos modelos foram empregados os métodos de validação cruzada e reamostragem de Monte Carlo. As métricas de erro aplicadas foram o coeficiente de determinação (R2) e o coeficiente de correlação de Pearson (R). Os resultados apontaram que as previsões foram coerentes e de boa qualidade, melhorando os valores médios do coeficiente de Pearson em aproximadamente 30 por cento comparados aos valores da literatura. Para o R(2) a média de valores alcançada foi de 0.85. Conclui-se que o uso do método de aprendizado de máquina se mostra confiável na previsão do acúmulo de deformação plástica no contorno do grão de um material policristalino. / [en] Machine learning methods have been widely used in the area of solid mechanics due to the large volume of data available in the literature. The motivation for this work was the study of the accumulation of plastic strain at the grain scale. Because the use of machine learning can be a significant contribution to creating models capable of predicting the accumulation of deformation. The objective of this work was to improve the prediction of plastic strain accumulation by proposing a new method for predicting the accumulation of plastic strains in grain boundaries of a polycrystalline material, using machine learning models. This work uses experimental data from the literature to structure three databases, which only consider grain boundaries. The following methods were used in the predictions: Decision Tree, Random Forest, Stochastic Gradient Descent, K-Nearest Neighbors, Gradient Boosting Regressor, and Principal Component Analysis (PCA). Monte Carlo crossvalidation and resampling methods were used to evaluate the models. The error metrics applied were the coefficient of determination (R2) and the Pearson correlation coefficient (R). The results indicate that the predictions were coherent and of good quality, improving the average Pearson coefficient values by approximately 30 percent compared to literature values. For R(2), the average value achieved was 0.85. It is concluded that the use of the machine learning method proves to be reliable in predicting the accumulation of plastic strain at the grain boundary of a polycrystalline material.

<p>Premelting is the formation of a thin, thermodynamically stable, liquid-like film at an interface for temperatures below the equilibrium melting temperature. Using a Monte Carlo technique, the underlying short range structural forces for premelting at the grain boundary can be directly calculated. This technique is applied to a (i) Σ9 ⟨115⟩ 120^o twist boundary and a (ii) Σ9 ⟨011⟩ {411} symmetric tilt boundary in an embedded atom model of Aluminum-Magnesium alloy. Both grain boundaries exhibit disordered structures near the melting point that depend on the concentration of Magnesium. The behavior is described quantitatively with sharp interface thermodynamics, involving an interfacial free energy that depends on width of the grain boundary, referred to as the disjoining potential. The disjoining potential calculated for boundary (i) displays a decreasing exponential dependence on width of the grain boundary, while the disjoining potential of (ii) features a weak attractive minimum. This work is discussed in relation to a previous study using pure Nickel, results of which can be useful to the theoretical study of thermodynamic forces underlying grain boundary premelting in an alloy.</p> / Master of Science (MSc)

Nanocrystalline metals exhibit a high strength/hardness but generally poor ductility during deformation regardless of their crystal structure which is often called the strength-ductility trade-off relationship and generally appears in most ultrafine-grained metals. The ultrafine-grained (UFG) high manganese austenitic twinning-induced plasticity (TWIP) steels have been found to overcome the strength-ductility trade-off but their underlying mechanism of discontinuous yielding behavior has not been well understood. In this study, our systematic TEM characterization suggests that the plastic deformation mechanisms in the early stage of deformation, around the macroscopic yield point, show an obvious association with grain size and nucleation of deformation twin was promoted rather than suppressed in UFG. More specifically, the main mechanism shifts from the conventional slip in grain interior to twinning nucleated from grain boundaries with decreasing the grain size down to less than 1 m. We also provide insights into the atomistic process of deformation twin nucleation at 3{111} twin boundaries, the dominant type of grain boundary in the UFG-TWIP steel of interest. In response to the external tensile stresses, the structure of coherent 3{111} twin boundary changes from atomistically smooth to partly defective by the grain boundary migration mechanism thus the "kink-like" defective step can act as a nucleation site for deformation twin, which deformation process is different from the one induced by dislocation pile-ups in coarse-grained counterparts and explain why UFG TWIP steel can retain the moderate ductility. In addition to the effect of grain size on deformation twin nucleation, grain boundary character was also taken into account. In coarse-grained TWIP steel, we experimentally reveal that deformation twin nucleation occurs at an annealing twin () boundary in a high-Mn austenitic steel when dislocation pile-up at boundary produced a local stress exceeding the twining stress, while no obvious local stress concentration was required at relatively high-energy grain boundaries such as or  A periodic contrast reversal associated with a sequential stacking faults emission from boundary was observed by in-situ transmission electron microscopy (TEM) deformation experiments, proving the successive layer-by-layer stacking fault emission was the deformation twin nucleation mechanism. The correlation between grain boundary character and deformation behavior was discussed both in low- and high-sigma value grain boundaries. On the other hand, localized strain concentration causes the nucleation of deformation twins at grain boundaries regardless of the grain boundary misorientation character in UFG TWIP steel. The invisibility of stacking fault (zero contrast) was also observed to be emitted at 3{111} boundaries in the coarse-grained TWIP steel, which deformation twin nucleation mechanism is found to be identical to UFG Fe-31Mn-3Si-3Al TWIP steel. / Doctor of Philosophy / High manganese (Mn) twin-induced plasticity (TWIP) steel is a new type of steels which exhibit pronounced strain hardening rate so that offering an extraordinary potential to adjust the strength-ductility relationship. This key advantage will help implement the current development of lightweighting components in automobile industry due to a considerable reduction of material use and an improved press formability. Such outstanding ductility can be contributed by the pronounced strain hardening rate during every such deformation processes, which is highly associated with several different controlling parameters, i.e., SFE, grain orientation, grain size, and grain boundary characters. In this study, we take particular attention to the effect of grain size and grain boundary characters on deformation twinning behavior besides well-known parameters such as SFE and grain orientation. The effect of grain size on deformation twinning behavior was found to be deeply associated with the yielding behavior in TWIP steel, i.e., a discontinuous yielding behavior with a unique yield drop was observed in ultrafine-grained TWIP while a continuous yielding behavior was observed in coarse-grained counterpart. Our TEM characterization indicates that the microstructural features of grains >10 m are different from the microstructural features in grains < 1 m. In over-10 m grains, normal dislocation slips and the formation of in-grain stacking faults are the main deformed microstructure. However, in the under-1 m grains, the in-grain dislocation slip is inhibited, but the deformation twinning is promoted at grain boundaries. This deformation transition from in-grain slip to twinning at grain boundary appears to be responsible for the discontinuous yielding behavior observed in stress-strain curve. The effect of grain boundary character on deformation twinning was examined in both coarse- and ultrafine-grained TWIP steels. In coarse-grained TWIP steel, we found that deformation twinning behavior varies as the function of boundary structure, i.e., different atomic configuration. Coherent twin boundary can act as a nucleation site for deformation twin as a localized strain concentration was introduced by dislocation pile-ups. On the other hand, incoherent boundaries can act as a deformation twin nucleation site by a boundary relaxation mechanism, i.e., grain-boundary dislocations can dissociate into partial dislocations to both side of boundary to accommodate the misfit between grains. In UFG TWIP steel, we found that the coherent twin boundary can act as a deformation twin nucleation site without presence of dislocation pile-ups. Alternatively, twin boundary becomes defective with a "kink-like" step by boundary migration. As a result, this defective step would progressively accumulate localized strain field thus stimulate the nucleation of deformation twin. Such study provides a novel insight into the UFG TWIP steel and a roadmap toward controlling TWIP effect.

During the past decade, Shape Memory Alloys (SMAs), particularly Nickel-Titanium (NiTi) alloys, have received increasing attention mainly because of their promising role to be integrated into multifunctional systems for actuation, morphing, and sensory capabilities in a broad variety of applications including biomedical, aerospace and seismological engineering. The unique performance of all the novel devices developed by SMAs relies on either the shape memory effect or pseudoelasticity, the two distinctive properties of SMAs. Both these unique properties are based on the inherent capability of SMAs to have two stable lattice structures at different stress or temperature conditions, and the ability of changing their crystallographic structure by a displacive phase transformation between a high-symmetry austenite phase and a low-symmetry martensite phase, in response to either mechanical or thermal loading. These properties make them a superior candidate for using as damping materials under high-strain-rate loading conditions in different engineering fields. SMA materials used in the most applications are polycrystalline in nature. In polycrystalline SMAs at the bulk-level, in addition to the phase transformation at the lattice-level, the thermomechanical response is also highly sensitive to the microstructural properties. In this work, the microstructure, as well as defects, such as dislocations and the stacking faults, are studied in the NiTi crystalline structure. In addition, the performance of NiTi under shock wave loading and vibrations, and their energy dissipation capabilities are examined using computational modeling, globally and locally. The effect of graphitic and metal structures, as reinforcements, on the performance of NiTi matrix composites under static and shock stress wave loading conditions is also investigated at the atomistic scale. / PHD / During the past decade, Shape Memory Alloys (SMAs), particularly Nickel-Titanium (NiTi) alloys, have received increasing attention mainly because of their promising role to be integrated into multifunctional systems for actuation, morphing, and sensory capabilities in a broad variety of applications including biomedical, aerospace and seismological engineering. The unique performance of all the novel devices developed by SMAs relies on their ability of changing their crystallographic structure by a displacive phase transformation between a high-symmetry austenite phase and a low-symmetry martensite phase, in response to either mechanical or thermal loading. These properties make them a superior candidate for using as damping materials in different engineering fields. In this work, the microstructure, as well as defects are studied in the NiTi crystalline structure. In addition, the performance of NiTi under shock wave loading and vibrations, and their energy dissipation capabilities are examined using computational modeling, globally and locally. The effect of graphitic and metal structures, as reinforcements, on the performance of NiTi matrix composites under static and shock stress wave loading conditions is also investigated at the atomistic scale.

Increasing energy demand leads to two crucial problems for the whole society. One is the economic cost and the other is the pollution of the environment, especially CO2 emissions. Despite efforts to adopt renewable energy sources, fossil fuels will continue to dominate. The temperature and stress are planned to be raised to 700 °C and 35 MPa respectively in the advanced ultra-supercritical (AUSC) power plants to improve the operating efficiency. However, the life of the components is limited by the properties of the materials. The aim of this thesis is to investigate the high temperature properties of materials used for future power plants. This thesis contains two parts. The first part is about developing creep rupture models for austenitic stainless steels. Grain boundary sliding (GBS) models have been proposed that can predict experimental results. Creep cavities are assumed to be generated at intersection of subboundaries with subboundary corners or particles on a sliding grain boundary, the so called double ledge model. For the first time a quantitative prediction of cavity nucleation for different types of commercial austenitic stainless steels has been made. For growth of creep cavities a new model for the interaction between the shape change of cavities and creep deformation has been proposed. In this constrained growth model, the affected zone around the cavities has been calculated with the help of FEM simulation. The new growth model can reproduce experimental cavity growth behavior quantitatively for different kinds of austenitic stainless steels. Based on the cavity nucleation models and the new growth models, the brittle creep rupture of austenitic stainless steels has been determined. By combing the brittle creep rupture with the ductile creep rupture models, the creep rupture strength of austenitic stainless steels has been predicted quantitatively. The accuracy of the creep rupture prediction can be improved significantly with combination of the two models. The second part of the thesis is on the fatigue properties of austenitic stainless steels and nickel based superalloys. Firstly, creep, low cycle fatigue (LCF) and creep-fatigue tests have been conducted for a modified HR3C (25Cr20NiNbN) austenitic stainless steel. The modified HR3C shows good LCF properties, but lower creep and creep-fatigue properties which may due to the low ductility of the material. Secondly, LCF properties of a nickel based superalloy Haynes 282 have been studied. Tests have been performed for a large ingot. The LCF properties of the core and rim positions did not show evident differences. Better LCF properties were observed when compared with two other low γ’ volume fraction nickel based superalloys. Metallography study results demonstrated that the failure mode of the material was transgranular. Both the initiation and growth of the fatigue cracks were transgranular. / <p>QC 20160905</p>

Austenitic stainless steels are frequently used in the cooling circuits of nuclear reactors. It has been found that cold-worked 304 stainless steels can be particularly susceptible to stress corrosion cracking at the operating conditions of such reactors. Despite more than 130 years of research underlying mechanisms are still not properly understood. For this reason, the effects of cold-work and applied stress on the oxidation behaviour of 304SS have been studied in this thesis. A set of samples with/without prior cold-work, and with/without stress applied during oxidation, were oxidized in autoclaves under simulated pressurised water reactor primary circuit conditions. Atom-probe tomography and analytical transmission electron microscopy were used to investigate the local chemistry and microstructure in the different samples tested. Regions containing grain boundaries, deformation bands, and matrix material in contact with the environment, were extracted from the coupon specimens with a focused ion beam machine. Cross-sections of crack tips were studied with secondary ion mass spectrometry and electron backscatter diffraction. The compositions of oxides grown along the surface and the different microstructural features were analysed. Fe-rich spinels were found at the surface and Cr-rich spinels were observed along fast diffusion paths. Ni-enrichment was found at the metal/oxide interfaces and a Ni-rich phase was detected in precipitates ahead of grain boundary oxides. Li was observed in all oxidised regions and B segregation, originating from impurities in the alloy, was observed in grain boundaries and crack tip oxides. Cavities and hydrogen associated with Ni-rich regions were found ahead of the bulk Cr-rich oxide in some of the samples. The implications of these findings for the understanding of SCC mechanisms are discussed. It is suggested that Ni precipitation as well as the presence of deformation bands may play an important role in controlling SCC susceptibility in 304 stainless steel. A modification of the film-rupture model including internal oxidation and fast diffusion along H-stabilised vacancies in strain fields at the crack front is proposed.

Approved for public release, distribution is unlimited / This study evaluated dome test samples of a superplastic AA5083 aluminum alloy deformed at nominally constant strain rates under biaxial strain conditions. Dome test samples resulted from gas-pressure forming of sheet material; for this study, samples were deformed at strain rates corresponding either to grain boundary sliding or dislocation creep control of deformation. Orientation Imaging Microscopy was utilized to determine texture development, grain size and grain-to-grain misorientation angle distributions for locations located along a line of latitude of the dome samples. The goal was to identify the location of the transition from grain boundary sliding to dislocation creep. Grain boundary sliding, which dominates at lower strain rates, can be recognized by a randomized texture and a higher concentration of high disorientation angles. Dislocation creep, which dominates at higher strain rates, is characterized by fiber texture formation and development of a peak at lower angles in the grain-to-grain misorientation angle distribution. / Ensign, United States Navy

Une approche couplée prenant en compte l’interaction de la plasticité cristalline et de la diffusion d’hydrogène a été établie et utilisée pour étudier le transport de l’hydrogène dans les agrégats polycristallins synthétiques de l’acier 316L avec des géométries de grains and des orientations cristallographiques différentes. Les champs mécaniques calculés à l’aide du code ZeBuLoN sont transférés dans un code de diffusion développé dans le cadre de ce travail. Une nouvelle formulation associée à un nouveau schéma numérique permet un calcul qui présente une bonne convergence. Les résultats des simulations montrent la redistribution de l’hydrogène dans les polycristaux due à la présence des hétérogénéités des contraintes hydrostatiques à l’échelle intragranulaire. L’effet de la vitesse de déformation a été quantitativement obtenu. Afin d’enrichir l’approche continue, un intérêt particulier est porté sur le rôle des joints de grains. Des simulations numériques d’un modèle atomique plan par plan ont été développées et appliquées aux bicristaux et aux structures de type ”bambou”. Les effets de puits ou de barrière induits par la présence des joints de grains sont clairement démontrés dans le cas du nickel pur. Pour reproduire ces effets dans les simulations de diffusion avec le modèle continue, une approche originale de simulation”multi-échelles” de la diffusion au joint de grain a été développée, et un nouveau régime de diffusion au joint de grain a été modélisé. / In the present work, we establish a one-way coupled crystal plasticity – hydrogen diffusion analysis and use this approach to study the hydrogen transport in artificial polycrystalline aggregates of 316L steel with different grain geometries and crystallographic orientation. The data about stress/strain fields computed at the microstructure scaleutilizing the crystal plasticity concept are transferred to the in-house diffusion code which was developed using a new numerical scheme for solving parabolic equations. In the case of initial uniform hydrogen content, the heterogeneity of the mechanical fields is shownto induce a redistribution of hydrogen in the microstructure. The effect of strain rate is clearly revealed. In the second part, hydrogen transport across grain boundaries is investigatedconsidering the specific diffusivity and segregation properties of these interfaces. Using a discrete atomic layer model, the retarding impact of grain boundaries is demonstrated on bicrystals and bamboo type membranes with and without external mechanical loading. To reproduce the effects observed in the atomistic simulations into the crystal plasticity – hydrogen diffusion model, a new physically based multi-scale method is proposed. Using this new approach we study the effect of grain boundary trapping kinetics on hydrogen diffusion and reveal a new grain boundary diffusion regime which has notbeen reported before.

With the recent development of in situ transmission electron microscopy (TEM) characterization techniques, the real time study of property-structure correlations in nanomaterials becomes possible. This dissertation reports the direct observations of deformation behavior of Al2O3-ZrO2-MgAl2O4 (AZM) bulk ceramic nanocomposites, strengthening mechanism of twins in YBa2Cu3O7-x (YBCO) thin film, work hardening event in nanocrystalline nickel and deformation of 2wt% Al doped ZnO (AZO) thin film with nanorod structures using the in situ TEM nanoindentation tool. The combined in situ movies with quantitative loading-unloading curves reveal the deformation mechanism of the above nanomaterial systems. At room temperature, in situ dynamic deformation studies show that the AZM nanocomposites undergo the deformation mainly through the grain-boundary sliding and rotation of small grains, i.e., ZrO2 grains, and some of the large grains, i.e., MgAl2O4 grains. We observed both plastic and elastic deformations in different sample regions in these multi-phase ceramic nanocomposites at room temperature. Both ex situ (conventional) and in situ nanoindentation were conducted to reveal the deformation of YBCO films from the directions perpendicular and parallel to the twin interfaces. Hardness measured perpendicular to twin interfaces is ~50% and 40% higher than that measured parallel to twin interfaces, by ex situ and in situ, respectively. By using an in situ nanoindentation tool inside TEM, dynamic work hardening event in nanocrystalline nickel was directly observed. During stain hardening stage, abundant Lomer-Cottrell (L-C) locks formed both within nanograins and against twin boundaries. Two major mechanisms were identified during interactions between L-C locks and twin boundaries. Quantitative nanoindentation experiments recorded during in situ experiments show an increase of yield strength from 1.64 to 2.29 GPa during multiple loading-unloading cycles. In situ TEM nanoindentation has been conducted to explore the size dependent deformation behavior of two different types (type I: ~ 0.51 of width/length ratio and type II: ~ 088 ratio) of AZO nanorods. During the indentation on type I nanord structure, annihilation of defects has been observed which is caused by limitation of the defect activities by relatively small size of the width. On the other hand, type II nanorod shows dislocation activities which enhanced the grain rotation under the external force applied on more isotropic direction through type II nanorod.

In this work, ultrafine-grained (UFG) Al-5wt%Zn alloy was produced by equal channel angular extrusion (ECAE). The microstructure evolution during ECAE and the mechanical properties of the UFG Al-Zn alloy were investigated. In order to identify the effect of Zn in the Al-Zn alloy, pure aluminum (4N, 99.99%) was also studied for comparison. The grains of the Al-Zn alloy could be refined effectively by increasing the ECAE passes. However, as the ECAE passes increased, the microhardness increased initially but maintained constant after 4 ECAE passes. The dislocation density within grain interior was decreased gradually with increasing ECAE passes. After being processed to twelve ECAE passes, the UFG Al-Zn alloy exhibited 53.7% of the grain boundaries being high angle grain boundaries (HAGBs). The UFG Al-5wt%Zn alloy exhibits superior tensile strength and elongation as compared with pure aluminum fabricated by the same ECAE process. Experimental results indicated that adding Zn in aluminum alloy could provide solid-solution strengthening and considerable enhancement in tensile ductility which might be related to an improved post-uniform elongation (PUE). The strain rate sensitivity (SRS) of the UFG Al-Zn alloy also increased with increasing the ECAE passes, which might be related to the fine grain size and the contribution of grain boundary sliding. The activation volume of the UFG Al-Zn alloy was in the range of 32b3~76b3, and the pure aluminum was in the range of 57b3~122b3. Because of the small value of the activation volume, it is suggested that the controlling mechanism for dislocation glide in the UFG Al-Zn alloy might be related to the generation and absorption of dislocations in grain boundary, as well as the interaction between dislocations and solute Zn atoms in the grain boundary.