In this study, we examine strain localization and failure initiation in cylindrical aluminum tubes subjected to radial impulse loading. This thesis is divided into two parts: the first part explores strain localization in aluminum 6061-T6 tubes undergoing radial expansion. An electromagnetically induced body force is used to expand a total of six tube specimens with varying body force intensities. The experimental response of the tube specimen is obtained by capturing images at different times of the expansion via a high-speed camera and conical mirror. A time history of the tube radius and global hoop strain is acquired by fitting the images of the tube with best fit circles using an iterative least-squares approach. For one specimen digital image correlation is implemented in addition to the circle fitting as a means of tracking the local true strain with time. Postmortem observations of the tube specimens reveal a localization pattern different than those found in previous studies that examined aluminum 6061-O tubes. An argument is made that Joule heating and thermal softening contribute to the unique localization pattern. Two numerical models, with differing magnitudes of thermal softening, are constructed to understand the observed localization. At the end of the chapter, proposals for future investigations are presented. The second chapter is devoted to buckling and fracture of an aluminum 7075-T73 tube exposed to a compressive impulse. A numerical model is constructed to analyze the development of fracture nucleation and propagation. Element failure is modeled with the Johnson-Cook damage criterion. Fracture in the walls of buckled cylindrical tubes appears to be a topic not addressed in previous literature. Nodal data and contour plots of the model are used to construct a general case for the response mechanism leading to fracture. Numerical results are compared to previous experimental investigations. Finally, proposals for future experimental examinations of fracture in buckled cylindrical tubes are provided. / text

Absolutely pure modules act in ways similar to injective modules. Therefore, through-out this document we investigate many of these properties of absolutely pure modules which are modelled after those similar properties of injective modules. The results we develop can be broken into three categories: localizations, covers and derived functors. We form S1M, an S1R module, for any Rmodule M. We state and prove some known results about localizations. Using these known techniques and properties of localizations, we arrive at conditions on the ring R which make an absolutely pure S1Rmodule into an absolutely pure Rmodule. We then show that under certain conditions, if A is an absolutely pure Rmodule, then S1A will be an absolutely pure S1Rmodule. Also, we dene conditions on the ring R which guarantee that the class of absolutely pure modules will be covering. These include R being left coherent, which we show implies a number of other necessary properties. We also develop derived functors similar to Extn R (whose development uses injective modules). We call these functors Axtn R, prove they are well dened, and develop many of their properties. Then we dene natural maps between Axtn(M;N) and Extn(M;N) and discuss what conditions on M and N guarantee that these maps are isomorphisms.

<p>Ova doktorska disertacija uvodi nov pristup ispitivanju stabilnosti dinamičkih<br />sistema, korišćenjem teorije pseudospektra. Na taj način se postojeći pojam<br />stabilnosti profinjuje pojmom robusne stabilnosti, koji mnogo adekvatnije<br />opisuje realnu ekološku stabilnost. Razvijen je nov matematički alat za<br />izračunavanje indikatora stabilnosti, koji je zatim ilustrovan na primeru dva<br />ekosistema tla, sa po četiri uzorka, u četiri različita stadijuma razvoja.</p> / <p>This doctoral dissertation establishes a novel approach to the stability analysis of<br />dynamical systems, in terms of matrix pseudospectrum. In that manner, the existing<br />concept of stability has undergone essential refinement so as to give birth to the<br />concept of robust stability, which has the ability to capture the ecological stability at a<br />more adequate level. Additionally, within the framework of the dissertation, a new<br />mathematical tool for the stability indicators computation has been developed, which<br />has then been used to illustrate theoretical results in form of two soil ecosystems,<br />each of them sampled four times, all of them observed in four distinct stages of<br />evolution.</p>

Une des préoccupations de la société est de trouver un compromis entre le développement des territoires et la protection de la faune sauvage. La prise en compte des déplacements des animaux lors de projets d’aménagement nécessite de connaître les comportements des différentes espèces et de définir ce qui a une influence sur leur localisation et sur la sélection de leur lieu de vie. Notre objectif est de pouvoir représenter les déplacements d’animaux sur un espace géographique précisément décrit afin ensuite de simuler et d’évaluer les conséquences des aménagements.Nous avons commencé par analyser l’influence des éléments du paysage sur les déplacements à partir de localisations connues d’animaux comme des traces GPS (suivis menés par l’ELIZ, l’ANSES, l’ONCFS, l’INRA)et de données de description de l’espace notamment la BD TOPO®. Les cas d’étude correspondent à des milieux différents et à trois espèces : renard, chevreuil et cerf. Nous avons pu confirmer le rôle de certains caractères de l’espace selon les cas d’étude. Par exemple les préférences spatiales lors des déplacements des renards en milieu périurbain semblent se porter sur la végétation arborée et des lieux peu occupés par les hommes pendant certaines parties de la journée (parcs, zones d’activités, le long des voies ferrées). Concernant les cervidés en milieu forestier, la pente et le type de peuplement paraissent avoir le plus d’influence sur les déplacements.À l’aide des connaissances extraites par les analyses et de la littérature, nous avons défini et implémenté dans la plateforme GeOxygene un modèle de simulation de déplacements d’animaux. Les trajectoires sont construites par une approche agent reprenant le comportement spatial selon l’espèce et l’influence des éléments du paysage favorables ou faisant obstacle. Nous effectuons une analyse critique de notre modèle puis nous proposons des pistes d’enrichissement à l’aide de la comparaison avec les observations et le retour des écologues. Enfin des scénarios d’aménagement sont testés dans le but de mettre en évidence leur impact et leur efficacité. / Finding compromises between human development and wildlife protection is one concern of society.Taking into account animal movements in planning projects requires some knowledge on species behaviours and on what determines their localizations and their habitat places. Our goal is to be able to represent animal movements on an accurate geographical space in order to simulate and to evaluate the consequences of planning decisions. We first analysed how the features of the landscape influence movements from collected localizations on animals, for example GPS tracks (studies of ELIZ, ANSES, ONCFS, INRA) and from data describing spacesuch as BD TOPO®. The studied cases are about several types of environment and three species: red fox,roe deer and red deer. We found some results that confirm the role played by the spatial features,depending on the studied cases. For instance in a periurban environment, foxes seem to be more inwooded patches and in places with few human activities during some parts of the day (squares, areas with industrial or commercial activities, sides of railways). In a forested environment, deers are more likely to be influenced by slope and forest stands. Thanks to knowledge from data analyses and to literature, we defined a simulation model for animalmovements. We implemented it in the GeOxygene platform. The trajectories are built with an agent approach by taking into account the spatial behaviour of the species and the influence of elements that favour or hinder movements. We proposed a critical view of the modelling choices and some improvements from the comparison with observations and experts advices. Then, scenarios within frastructures are defined so that to identify their impact and their efficiency.

Made available in DSpace on 2014-07-29T14:49:10Z (GMT). No. of bitstreams: 1 Heloisio_Caetano.pdf: 4628443 bytes, checksum: b488772923cf9872b1b42391f8d89610 (MD5) Previous issue date: 2011-08-08 / The analyses of the dynamics of the crops composition in the State of Goiás was first started by Yokoyama (1988). Here, is made a continuity of this work and methodology enlargement elements like maps and graphics to better scrutinize the phenomenon throw the use of shift-share rates. Further socio economic inferences are based in agricultural frontier´s theory. The mathematical methods and uses, of the shift-share rates, are discussed. Using this model, the area effects are decomposed into scale and substitution effects, and it shows the geographic location effects for the selected microrregiões, and for this also the production effects. The agricultural frontier model is presented in basics ideas. The indices are presented for individual and the set of crops in the microrregiões and so are fluctuation of production of the set and location of crops in the State of Goiás. The results point changes in the crops composition in the State of Goiás. Staple food production in the new context, despite changes of territory of crops, it maintains important role in National context. It presents results about the important discussion that involves sugar-cane expansion, relations to soy-beans and cattle and conclusions about crop relations to ´cerrados´ area. The crop period goes is from 1990 to 2009. / A análise da composição das culturas no Estado de Goiás foi um trabalho iniciado por Yokoyama (1988). Neste realiza-se uma continuação aliada a uma atualização metodológica ao utilizar elementos como mapas e gráficos que permitiram uma visão mais ampla dos fenômenos evidenciados pelos índices Shift-Share. Neste trabalho também ocorre uma renovação da análise, onde o conceito de fronteiras agrícolas fundamenta o estudo da ação dos sujeitos. Pelo modelo decompõe-se o efeito área em efeito escala e substituição, e apresenta também o efeito localização geográfica, apresentando estes índices também em porcentagem, para as culturas selecionadas. Os efeitos são apresentados também para as microrregiões selecionadas, sendo que para estas foi possível discutir o efeito composição do produto. O modelo de fronteiras agrícolas é apresentado em suas idéias fundamentais. São apresentadas as análises a partir do cálculo dos índices, para as culturas e conjunto de lavouras, e a localização das culturas no Estado. Os resultados apontam para ocorrências na produção de alimentos frente a este novo panorama, onde apesar de o Estado sofrer consideráveis alterações na área plantada destas culturas a sua importância mantém-se no contexto nacional. Apresenta dados e resultados sobre a discussão que envolve expansão da cana-de-açúcar, relações com a soja e o gado e conclusão sobre as relação entre as culturas e as áreas de cerrado. O período analisado foi 1990 a 2009.