The goal of this research work is to further the understanding of crystal plasticity, particularly at reduced structural and material length scales. Fundamental understanding of plasticity is central to various challenges facing design and manufacturing of materials for structural and electronic device applications. The development of microstructurally tailored advanced metallic materials with enhanced mechanical properties that can withstand extremes in stress, strain, and temperature, will aid in increasing the efficiency of power generating systems by allowing them to work at higher temperatures and pressures. High specific strength materials can lead to low fuel consumption in transport vehicles. Experiments have shown that enhanced mechanical properties can be obtained in materials by constraining their size, microstructure (e.g. grain size), or both for various applications. For the successful design of these materials, it is necessary to have a thorough understanding of the influence of different length scales and evolving microstructure on the overall behavior. In this study, distinction is made between the effect of structural and material length scale on the mechanical behavior of materials. A length scale associated with an underlying physical mechanism influencing the mechanical behavior can overlap with either structural length scales or material length scales. If it overlaps with structural length scales, then the material is said to be dimensionally constrained. On the other hand, if it overlaps with material length scales, for example grain size, then the material is said to be microstructurally constrained. The objectives of this research work are: (1) to investigate scale and size effects due to dimensional constraints; (2) to investigate size effects due to microstructural constraints; and (3) to develop a size dependent hardening model through coarse graining of dislocation dynamics. A discrete dislocation dynamics (DDD) framework where the scale of analysis is intermediate between a fully discretized (e.g. atomistic) and fully continuum is used for this study. This mesoscale tool allows to address all the stated objectives of this study within a single framework. Within this framework, the effect of structural and the material length scales are naturally accounted for in the simulations and need not be specified in an ad hoc manner, as in some continuum models. It holds the promise of connecting the evolution of the defect microstructure to the effective response of the crystal. Further, it provides useful information to develop physically motivated continuum models to model size effects in materials. The contributions of this study are: (a) provides a new interpretation of mechanical size effect due to only dimensional constraint using DDD; (b) a development of an experimentally validated DDD simulation methodology to model Cu micropillars; (c) a coarse graining technique using DDD to develop a phenomenological model to capture size effect on strain hardening; and (d) a development of a DDD framework for polycrystals to investigate grain size effect on yield strength and strain hardening.

[pt] A técnica de correlação digital de imagens (Digital Image Correlation, DIC) é uma técnica óptica de campo global que consiste na análise de imagens da superfície de um espécime antes e após seu carregamento para determinação de campos de deslocamentos e deformações. Neste trabalho foram usados um sistema DIC estereoscópico convencional 3D, para aplicações em campos de visão (macro) da ordem de 200mmx200mm, e um sistema DIC estereomicroscópico 3D, para aplicações em áreas tão pequenas quanto 1mm quadrado. Para a análise de deformações em campos com dimensões tão diferentes precisou-se desenvolver uma metodologia para controlar parâmetros da técnica DIC, tais como dimensões de pontos (speckles) impressos no espécime e tamanhos de subsets e steps usados para a correlação das imagens captadas. A seleção dos problemas estruturais para aplicação da técnica levou em conta não só a diversidade destes, como também o pioneirismo das aplicações de DIC a problemas nos quais podem ser gerados resultados que melhor ajudem na compreensão de suas facetas específicas. Assim, a técnica foi aplicada à medição de deformações em tubos com defeitos por suas vantagens de medição de grandes deformações e de visualização de regiões de interesse, sob a forma de campos globais de deformações. Na área de determinações de propriedades mecânicas de materiais compósitos, foram realizadas medições em espécimes não convencionais de resina epóxi reforçados por fibras de carbono ou fibras de vidro para determinação de suas propriedades elásticas. Por fim, foram realizadas medições em um espécime Compact Tension Specimen CTS de aço grau API 5LX60 contendo uma trinca e em espécimes com entalhes profundos (concentrações de tensões) constituídos por materiais distintos (policarbonato e alumínio), para obtenção de campos de deslocamentos e de deformações elastoplásticas. Nestas aplicações os resultados obtidos com a técnica DIC foram comparados com aqueles obtidos com modelos de elementos finitos (EF), com medições com extensômetros de resistência elétrica (strain gages) e com resultados analíticos publicados na literatura. Levando-se em conta as comparações feitas, a grande maioria dos experimentos realizados pode ser considerada satisfatória. Ao longo da tese foram descritas as adaptações, considerações e boas práticas consideradas necessárias para obtenção de bons resultados nas diferentes medições e para os diferentes aparatos experimentais utilizados. Estas recomendações serão bastante úteis para medições futuras ou mesmo para auxiliar na avaliação de confiabilidade de alguns resultados apresentados na literatura especializada. / [en] The digital image correlation (DIC) technique is a global field optical technique that consists in the analysis of images taken from the surface of a specimen before and after being subjected to a load, in order to determine displacement and strain fields. In the current work, both conventional 3D stereoscopic and micro-stereoscopic DIC systems were used, the former for applications in macroscopic fields of view (of the order of 200mm x 200mm), and the latter for applications in surface areas as small as 1mm square. For strain analysis in fields with such varied dimensions, the development of a methodology to control some parameters of the technique was required. Among the parameters to be controlled were dimensions of speckles printed on the specimen and size of the subsets and steps used in the image correlation procedure. The selection of structural problems to be analyzed by the DIC technique took into account not only diversity, but also the pioneering aspect in terms of DIC application to problems which can generate results that lead to a better comprehension of its specific issues. Hence, the technique was applied to the measurement of global strain fields in defective tubes, due to the advantages in measuring large strains and visualizing the regions of interest in such cases. For determination of the mechanical properties of composite materials, measurements were performed in non-conventional specimens made of epoxy resin and reinforced by carbon or glass fibers for evaluation of its elastic properties. Finally, measurements in a cracked Compact Tension Specimen (CTS) with degree of steel API 5LX60 and in specimens with deep notches (stress concentration) consisting of different materials (polycarbonate and aluminum) were carried out for estimation of displacement fields and elastoplastic strains. For those applications, the results were compared with those obtained from finite element models, from strain gages, and also with analytical results from the literature. Taking into account such comparisons, it can be said that the great majority of the experimental measurements was satisfactory. Throughout the current manuscript, the required adaptations, good practices needed to achieve reliable results from the different types of measurement and experimental apparatus, as well as other considerations, were carefully described. These recommendations will be quite useful for future measurements, or even to assist in the evaluation of the reliability of certain results presented in the specialized literature.