This thesis primarily focused on the development of novel composite tubes by braiding. The objective was to use hierarchical scale technique, i.e., micro, meso and macro scales, with the transfer of information from one scale to another to develop novel braided composite tubes. This research was conducted and reported in three journal papers. The aim of the first paper was to predict plane elastic properties for E-glass/epoxy braided composite structures at different braid orientations, by analytical and finite element techniques. The lenticular shape has been used to describe the geometry of the tow. Modified lenticular geometric model was developed to improve an existing geometric model, in terms of tow parameters, thereafter, plane elastic properties from Chamis micromechanical model for E-glass fibre and epoxy matrix without any knockdown effects were used as benchmark to develop predictive models, namely; Lekhnitskii's methodology and braided unit cell meso-scale finite element model to account for the effects of tow geometry, undulations/crimp, cross-over and braid orientations on the plane elastic properties of E-glass/epoxy composite. The results showed agreement in trend between the predictive models, Chamis micromechanical model, and a similar existing model. However, the plane elastic properties were knocked down in predictive models by 30% in the E11 direction and 32% in the E22 direction, when compared with Chamis micro-mechanical model for largest ±65° braid angle, among the braid angles, considered. The aim of the second paper was to manufacture E-glass/epoxy braided tubes at different braid orientations by vacuum bag infusion technique, conduct internal pressure tests, and determine the hoop and axial moduli of the infused tubes. Lekhnitskii's methodology was also used to develop plane elastic moduli by experiment using microscopy results, and by calculation. The experimental elastic moduli of the infused tubes and the experimental elastic moduli from Lekhnitskii's methodology were used to compare the predictive elastic moduli for E-glass/epoxy braided structures by Chamis micro-mechanical model, and the braided unit cell meso-scale finite element model. The two were from another paper. Results showed a perfect agreement in trend between the experimental results and the predictive results. However, the values of the experimental results were close but lower than the predicted results. Optical microscopy was performed on braided tube cross-section to evaluate the level of crimp or undulation. This was done by the determination of tow centreline crimp angle and aspect ratio. Results show that when compared with the predicted crimp, there was an agreement in trend, although the experimental results were lower than the predicted. Also, the knockdown factor was evaluated and used to quantify the reduction in experimental elastic moduli when compared with the predicted. Results showed that the absences of crimp in the Chamis model caused a tremendous difference between it, other predicted models and the experiment results. The elastic moduli of Chamis were by far higher than all others, including other predictive models. The purpose of the third paper was to manufacture E-glass/epoxy braided tube at ±31°, ±45°, ±55°, ±65° braid orientations using vacuum bagging and resin infusion technique, to design and manufacture a rig for tube internal pressures experiment, to determine the hoop and axial stress performances of the tubes by internal pressure experiment, to compare experimental results with laminate analysis predictions to evaluate the effect of crimp on the internal pressure performance of the braided tubes. To use E-glass braided tow meso-scale unit cell finite element model to predict the tow critical stresses, and the optimum braided tube architecture, using tube hoop and axial failure stresses or strains. The tubes were manufactured and subjected to internal pressure test (2:1), to failure. Failure mode was by weeping and bursting. Hoop stress was twice the axial stress. The highest value of hoop stress was at the ±65° braid angle, higher than the hoop stresses at the ±31°, ±45°, and ±55 ° braid angles by 50%, 39%, and 28% respectively. Hoop stress increased with increase in braid angle. The experimental results were validated by laminate analysis predictions by Chamis micro-mechanical model and Lekhnitskii's methodology, and the trend of the laminate analysis prediction matched that of the experimental results. However, the predicted values were higher than the experimental results by 21%, 14%, 11%, 10% for the ±31°, ±45°, ±55°, ±65° braid angles for the Chamis micro-mechanical model and 5%, 7%, 7%, 5% for the ±31°, ±45°, ±55°, ±65 braid angles respectively for the Lekhnitskii's model, showing the severe effect of crimp in the experimental tube, mostly when compared with Chamis micro-mechanical model. Braided tow unit cell finite element model prediction, showed that tow axial stresses increased with increase in braid angle, while the tow transverse stresses decreased with increase in braid angle. The predictions showed that the tow critical stresses and the tube optimum braided architecture lie between the ±65° and 90° braid angles. The tow critical stresses are the stresses at which the tow decreasing transverse stress and the tow increasing axial stress causes the tube to fail.

Submitted by JÚLIO HEBER SILVA (julioheber@yahoo.com.br) on 2017-05-25T20:22:07Z No. of bitstreams: 2 Tese - Juliana Simião Ferreira - 2013.pdf: 2071806 bytes, checksum: 8202c30edb9d0103d57d3b98ce204f2c (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-05-26T14:46:31Z (GMT) No. of bitstreams: 2 Tese - Juliana Simião Ferreira - 2013.pdf: 2071806 bytes, checksum: 8202c30edb9d0103d57d3b98ce204f2c (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-05-26T14:46:31Z (GMT). No. of bitstreams: 2 Tese - Juliana Simião Ferreira - 2013.pdf: 2071806 bytes, checksum: 8202c30edb9d0103d57d3b98ce204f2c (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2013-05-20 / Headwater streams biodiversity is very important to the maintenance of the biological integrity found in the whole hydrological basin. Therefore, understanding the processes generating and/or influencing biodiversity in such environments may contribute to future aquatic integrity monitoring programs and to the strategic planning for both species and community conservation found within the every stream networks. The main goals of the current study were to: i) identify landscape and spatial environmental predictors responsible for the structure of aquatic insects communities (Trichoptera); ii) identify which hierarchical scale within the hydrologic basins is responsible for the highest biodiversity variation, what may provide subsidies for future studies and planned conservation of lotic environments. In order to achieve such goals, we did different studies with specific objectives directed to (i) measure the contribution of environmental and spatial variables while determining species richness and composition within Trichoptera communities, separately considering species with wide and narrow spatial distribution; (ii) measure the influence of landscape attributes such as vegetation cover in different spatial scales on species richness and bet diversity, as well on the different trophic functional groups found within the Trichoptera community; (iii) evaluate the effects of nested headwater streams community structure embedded within superior hierarchical levels (e.g. microbasins, subbasins, and the whole catchment), as well as the effects of the environmental and spatial predictors on alpha, beta and gamaTrichoptera diversity considering different spatial scales. All studies were performed with a set of biological data sampled in 48 headwater streams within two hydrographic basins locate within the state of Goiás, Brazil, with different land use regimes. Within those streams, local environmental variables (physical-chemical variables and habtat integrity), landscape attributes (vegetation cover, and land use classification), and also the biological data regarding the Trichoptera insects were sampled. In the first study, we observed that compositional changes within the Trichoptera communities was determined by local processes, especially for genus with wider distribution, while the communities of those species with narrower distribution were probably structured by other factors not measured here. This study reinforces that the use of community deconstruction related to specific species attributes (e.g. spatial distribution) contributes to a better comprehension of the processes determining the metacommunities structure. In the second study, we verified that landscape attributes within the hydrologic basin, such as vegetation quality around the streamlet and vegetation cover up to 200m from the streamlet, are important predictors to the structure and functioning of Trichoptera communities. In the third study, results indicate that the variation in alpha, beta and gama diversities depend on the spatial scale considered and that the environmental degradation affects distinctly such patterns. We were able to verify that the structure of Trichoptera communities is determined by environmental factors, mainly in the local scales, such as water physical-chemical parameters and vegetation cover. Additionally, we observed that that the spatial scale within which the diversity is maximized is on catchment scale, reinforcing the need to increase the spatial scales considered in such studies in order to obtain a higher diversity for the Trichoptera in headwater streams. Finally, these studies confirm the vulnerability of Trichoptera communities to the variability on water quality and landscape changes, supporting the use of these organisms as indicators and reinforcing the need of conservation strategies which consider the aquatic diversity within each streamlet embedded in hydrologic basin elsewhere. / A diversidade biológica em riachos de cabeceira é muito importante para a manutenção da integridade de toda a bacia hidrográfica. Então, entender os processos que geram e/ou influenciam a diversidade nestes ambientes pode contribuir para futuros programas de monitoramento da integridade aquática e para o planejamento estratégico para a conservação de espécies e de comunidades das redes hidrográficas. O presente estudo teve por objetivo principal identificar os preditores ambientais locais, de paisagem e espaciais em múltiplas escalas responsáveis pela estruturação das comunidades de insetos aquáticos (Trichoptera). Além disso, objetivou verificar a escala hierárquica de bacias hidrográficas responsável pela maior variação da diversidade e desta forma, subsidiar futuros estudos e planejamento para conservação de ambientes lóticos. Para tanto, foramrealizados diferentes estudos com objetivos específicos direcionados a(i)medir a contribuição relativa de variáveis ambientais e espaciais na determinação da riqueza e composição taxonômica das comunidades de Trichoptera, considerando separadamente os táxons com distribuição ampla e restrita em riachos de nascente do bioma Cerrado, (ii)medir a influência dos atributos da paisagem como cobertura vegetal em diferentes escalas espaciais sobre a riqueza e diversidade beta, bem como sobre diferentes grupos tróficos funcionais da comunidade de Trichoptera, (iii) avaliar o efeito da estrutura aninhada de pequenos riachos de cabeceira inseridos dentro de níveis hierarquicamente superiores (microbacias, sub-bacias e bacias), bem como o efeito dos preditores ambientais e espaciais na diversidade alfa, beta e gama de insetos aquáticos. Todos os estudos foram realizados com o conjunto de dados amostrados em 48 riachos de nascente em duas bacias hidrográficas em Goiás, Brasil, com padrões de uso do solo distintos. Nestes riachos amostramos variáveis ambientais locais (físico-química e integridade de habitat) e de paisagem (índice de cobertura vegetal e uso do solo), assim como os insetos aquáticos da ordem Trichoptera.No primeiro estudo, observamos queamudança na composição das comunidades de Trichoptera foi determinada por processos locais, principalmente para os gêneros de distribuição mais ampla, enquanto os restritos provavelmente foram estruturados por outros fatores aqui não investigados. Este estudo reforça que o uso da abordagem de desconstrução da comunidade em função dos atributos das espécies contribui para uma melhor compreensão dos processos determinantes das metacomunidades. No segundo estudo, verificamos que as características da paisagem dentro da bacia de drenagem, como qualidade da mata em torno do riacho e da cobertura vegetal em escalas de até 200m do riacho, são importantes preditores da estrutura e funcionamento das comunidades de Trichoptera. No terceiro capítulo, os resultados indicam que a variação na diversidade alfa, beta e gama é dependente da escala espacial, e que o grau de degradação regional afeta distintamente estes padrões. Desta forma, podemos verificar que a estrutura das comunidades de Trichoptera é determinada por fatores ambientais, principalmente, na escala local, tanto as características da água como de cobertura vegetal. Além disso, observamos que a escala espacial em que a diversidade é maximizada é na escala de sub-bacia, reforçando a necessidade de ampliar as escalas espaciais para se obter a maior diversidade deste grupo em riachos de cabeceira. Por fim, estes trabalhos confirmam a vulnerabilidade das comunidades de Trichoptera à variações na qualidada de água e de modificações na paisagem como intensificação do uso do solo, o que suporta o uso destes como indicadores, além de reforçar a necessidade de estratégias para conservação que considere a diversidade aquática em riachos pertencentes a redes hidrográficas.