The isotropy of a glass can be broken by the application of a mechanical stress giving rise to a phenomenon of birefringence. Some lead-containing glass compositions are known to prevent this phenomenon and they are called zero-stress optic glass. Mueller’s theory of photoelasticity attempts to explain the structural origin of the photoelastic response in glass and crystal. Zwanziger’s empirical model is able to predict the photoelastic response of a glass based on its composition and the crystal structure of its constituents. Lead-, tin-, antimony-, zinc-, and cadmium-containing glasses were investigated in the binary silicate, borate, and phosphate systems. The stress optic coe?cient of these binary glasses was measured experimentally using the S´enarmont method or found in the literature. Solid-state Nuclear Magnetic Resonance spectroscopy and M¨ossbauer spectroscopy were mainly used to investigate the local environment of the cations. The photoelastic response of a glass and its structure were correlated, and the results were compared with the expectations arising from Mueller’s theory and Zwanziger’s empirical model. The theory and the model were both tested and their reliability was discussed. Zero-stress optic glasses are of technological interest, but new environmental regulations forbids the use of lead in materials, including glass. From experimental results and literature, a global strategy to design new zero-stress optic glasses was established. New lead-free zero-stress optic glasses were discovered with properties similar to the lead-containing zero-stress optic glass (high index of refraction, transparency, no coloration). The study of the structural dependence of the photoelastic response of oxide glass contributed to identify new parameters in?uencing the photoelasticity, such as covalency, polarizability and natural deformation of the additive.

Granular materials are composed of solid, discrete particles and exhibit mechanical properties that range from fluid to solid behavior. Some of the complexity exhibited by granular systems arises due to the long-range order that develops due to particle-particle contact. Inter-particle forces in granular materials often form a distributive network of filamentary force-accommodating chains (i.e. force chains), such that a fraction of the total number of particles accommodates the majority of the forces in the system. The force chain network inherent to a system composed of granular materials controls the macroscopic behavior of the granular material. Force transmission by these filamentary chains is focused (or localized) to the grain scale at boundaries such as the granular flow substrate. Recent laboratory experiments have shown that force chains transmit extreme localized forces to the substrates of free surface granular flows. In this work we combine analog and numeric experimental approaches to investigate the forces at the bed of a simplified granular flow. A photoelastic experimental approach is used to resolve discrete forces in the granular flows. We also conduct discrete element method (DEM) simulations, using input parameters derived from measureable physical material properties of experimental and natural materials, which successfully reproduce the analog experimental results. This work suggests that force chain activity may play an unexpected and important role in the bed physics of dense granular flows through substrate modification by erosion and entrainment, and that DEM numerical methods effectively treat force chain processes in simulated granular flows.

Thesis (Ph. D.)--Virginia Polytechnic Institute and State University, 1997. / Vita. Abstract. Includes bibliographical references (leaves 148-155). Also available electronically via Internet.

Thesis (M.S.)--University of Michigan, 1953.

Thesis (M.S.)--University of Michigan, 1976.

Thesis (M.S.)--The University of Michigan, 1976.

[pt] O presente trabalho visa determinar, experimentalmente,Fatores de Intensidade de Tensões, KI e KII em barras e tubos, submetidos à tração, com trincas situadas: I) Junto ao pé do cordão de solda em barras planas soldadas ortogonalmente com relações comprimento de trinca, a, largura da barra, B, a/B igual a 0,1 e a/B igual a 0,2. II) Próximo à raiz da solda em tubos unidos a 90 graus com relações comprimento de trinca, espessura do tubo principal, t, a/t igual a 0,1; a/t igual a 0,2; a/t igual a 0,3. III) Barras e tubos simples (sem uniões soldadas) tendo as relações a/B igual a 0,1; 0,2; 0,3 e a/t igual a 0,1; 0,2; 0,25 e 0,3. Nos modelos planos, foi feito um estudo comparativo entre espécimes que possuíam entalhes usinados para simular as trincas e espécimes onde forma propagadas trincas de fadiga. Os resultados para as barras forma obtidos utilizando-se a técnica fotoelástica bidimensional e para os tubos, o método tridimensional de congelamento de tensões e corte em fatias. Nas barras os dados coletados foram submetidos aos programas CRACK e FLAW, e nos tubos, apenas ao programa FLAW, para determinação dos valores de KI e KII. / [en] This work aims to determine Stress Intensity Factors KI and KII for bars and tubes loaded in tension with the following geometry: I) T Join bars with cracks located near the weld toe. Relations between crack length, a, and bar width, B, were: a/B equal 0.05; a/B equal 0.1 a/B equal 0.2. II) Tubes with T joints with cracks located at the weld toe. The relations between crack length and tube thickness, t, were: a/t equal 0.1; a/t equal 0.2 and a/t equal 0.3. III) Plains bars with ratios a/B equal 0.05; a/B equal 0.1 and a/B equal 0.2. IV) Plains tubes with ratios a/t of 0.1, 0.2, 0.25 and 0.3. A comparative study was made between plane models that had artificial machined cracks and those that had natural fatigue cracks. Results for bars were obtained employing two-dimensional Photoelasticity technique and for tubes, the three-dimensional stress freezing and slicing method was used. To generate KI and KII, the collected data were submitted to FLAW and CRACK computer programs in the case of bars, while for tubes, only the FLAW program was used.

[pt] Este trabalho apresenta uma nova técnica para análise de tensões em modelos fotoelásticos planos. Essa técnica baseia-se na determinação de coeficientes para funções de tensão aproximadas, através da coleta dos dados extraídos de ordens de franjas isocromáticas e de condições de contorno de modelos fotoelásticos planos. O modelo tem como vantagens a determinação de fatores de concentração de tensões com boa precisão, além de permitir a separação de tensões em pequenas ou grandes regiões de modelos fotoelásticos planos, sem requerer a leitura de franjas isoclínicas. Uma outra grande vantagem está na possibilidade imediata de acoplamento desse método com um analisador de imagens para determinação automática de Kt. Foram encontrados excelentes resultados em problemas onde haviam concentradores de tensão e no estudo de regiões próximas a cargas concentradas. A determinação dos coeficientes das funções foi feita através de um programa escrito em BASIC, que utiliza o método de Newton-Raphson em conjunto com o ajuste dos dados por mínimos quadrados. Vários exemplos foram estudados com o objetivo de avaliar sua performance. Entre eles estão: carga concentrada sobre uma superfície semi-infinita; placa infinita e finita com futo circular sujeita à tração; disco circular submetidos à cargas diametrais compressivas; viga reta carregada à tração constante e a momento constante; e barra com duplo entalhe em U submetida a momento constante. / [en] This work presents a new technique for improving stress analysis in photoelastic plane models. This technique is based on the determination of coefficients of approximated stress functions through the collection of isochromatic fringe data, and by the applications of boundary conditions to photoelastic plane models. The technique has as advantages the accurate determination of stress concentration factors and the stress without the need for collecting isoclinic fringe data. Another great advantage, is the immediate use of this method coupled with an image analyzer device for automatic extraction of Kt. Excellent results were found in problems where stress concentration or concentrated loading were present. The coefficients of the stress functions were determined by a BASIC program, using a Newton-Raphson technique coupled to a least square algorithm. Many example were tested to evaluate its performance. Among them are: concentrated force at a point of a straight boundary; infinite and finite plate with circular holes subjected to uniaxial tensile loads; circular disk under compressive diametral forces; bar subjected to pure tensile and pure bending loads; and a double U-notched bar under pure bending.

[pt] Este trabalho apresenta um método para determinação de fatores de intensidade de tensão, KI e KII, baseado nas técnicas de Newton-Raphson e mínimos quadrados onde o termo Teta 0x, referente ao campo de tensões distante, é tomado como um polinômio completo em x e y para ajustamento do campo de tensões não singular. Um análise do número e da localização dos pontos utilizados na coleta de dados é feita com intuito de melhor poder-se avaliar sua influencia sobre os valores de KI e KII determinados pelo método proposto. Apresenta-se uma comparação entre as distribuições de franjas isocromáticas obtidas experimentalmente e aquelas plotadas com auxílio de um computador utilizando-se para isto os valores determinados de KI e KII e, também, uma comparação entre os resultados obtidos e aqueles existentes na literatura. Finalmente, o método é utilizado na determinação de KI e KII para o problema da existência de trincas nas proximidades de junções soldadas de placas a 90 graus e 45 graus como exemplo de aplicação. / [en] This thesis presents a photoelastic method to determine stress intensy factors, KI and KII, base don the Newton-Raphson and least square techniques, where the term teta 0x refers to the far field stress distribution is substituted by a complete polynomial in x and y for the better adjustment of the nonsingular stress field. Analyses are made of the number and location of points used to obtain the experimental data with the purpose of estudying their influence on the values of KI an KII determined by the proposed method. The experimentally determined isochromatic fringe patterns are compared with computer plotted fringe patterns generated with the aid of the determined values of KI and KII. Finally, the proposed method is employed for the determination of stress intensity factors in problems simulating cracks near joints of welded plates subjected to tension and beding.

Made available in DSpace on 2014-10-09T12:50:08Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:58:46Z (GMT). No. of bitstreams: 1 10458.pdf: 6292161 bytes, checksum: 035d670a36d319ca420fee75d85f96a6 (MD5) / Tese (Doutoramento) / IPEN/T / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP