La plupart des problèmes d'ingénierie des deux derniers siècles ont été résolus grâce à des modèles structuraux pour poutres, plaques et coques. Les théories classiques, tels que Euler-Bernoulli, Navier et de Saint-Venant pour les poutres, et Kirchhoff-Love et Mindlin-Reissner pour plaques et coques, ont permis de réduire le problème générique 3-D, dans le problème unidimensionnel pour les poutres et deux dimensionnelle pour les coques et les plaques. Théories raffinés d'ordre supérieur ont été proposées au cours du temps, comme les modèles classiques ne consentez pas à d'obtenir une complète domaine des contraintes et des déformations. La Carrera Unified Formulation (UF) a été proposé au cours de la dernière décennie, et permet de développer un grand nombre de théories structurelles avec un nombre variable d'inconnues principales au moyen d'une notation compacte et se référant à des nuclei fondamentales. Cette formulation unifiée permet de dériver carrément des modèles structurels d'ordre supérieur, pour les poutres, plaques et coques. Dans ce cadre, cette thèse vise à étendre la formulation pour l'analyse des structures fonctionnellement gradués (FGM), en introduisant aussi le problème thermo-mécanique, dans le cas des poutres fonctionnellement gradués. Suite à la formulation unifiée, les variables génériques déplacements sont écrits en termes de fonctions de base, qui multiplie les inconnues. Dans la deuxième partie de la thèse, de nouvelles fonctions de bases pour la modélisation des coques, qui représentent une approximation trigonométrique des variables déplacements, sont pris en compte / Most of the engineering problems of the last two centuries have been solved thanks to structural models for both beams, and for plates and shells. Classical theories, such as Euler-Bernoulli, Navier and De Saint-Venant for beams, and Kirchhoff-Love and Mindlin- Reissner for plates and shells, permitted to reduce the generic 3-D problem, in onedimensional one for beams and two-dimensional for shells and plates. Refined higher order theories have been proposed in the course of time, as the classical models do not consent to obtain a complete stress/strain field. Carrera Unified Formulation (UF) has been proposed during the last decade, and allows to develop a large number of structural theories with a variable number of main unknowns by means of a compact notation and referring to few fundamental nuclei. This Unified Formulation allows to derive straightforwardly higher-order structural models, for beams, plates and shells. In this framework, this thesis aims to extend the formulation for the analysis of Functionally Graded structures, introducing also the thermo-mechanical problem, in the case of functionally graded beams. Following the Unified Formulation, the generic displacements variables are written in terms of a base functions, which multiplies the unknowns. In the second part of the thesis, new bases functions for shells modelling, accounting for trigonometric approximation of the displacements variables, are considered.

Text in English / As an external cause of death, strangulation represents an extreme and particularly pernicious form of violence. Following the evidence gap in the extant literature, the current research examined the incidence, distributions, individual and situational predictors, and structural determinants of homicidal strangulation in the City of Johannesburg for the period 2001-2010. The thesis is structured around four discrete but interrelated studies, which collectively offer an initial contribution to the body of scholarship on homicide generally, and on the characteristics and patterns of strangulation homicide specifically. The research drew on data from the National Injury Mortality Surveillance System and the South African National Census. Study I is a descriptive study that quantifies the extent of homicidal strangulation in the City of Johannesburg and describes its distribution by characteristics of person, time, place and alcohol consumption. The remaining studies are analytical in focus, and are aimed at explaining homicidal strangulation in the City of Johannesburg in terms of its determinants. These studies are differentiated by their focus on individual-level and neighbourhood-level risks. Study II assesses overall homicide strangulation risk in relation to all the other leading causes of homicide. Study III undertakes further disaggregation to investigate homicidal strangulation risk by gender specifically. Study IV considers the socio-structural correlates and geographic distributions of fatal strangulation. The study engages select micro-level and macro-level theories that focus on the intersection between vulnerability and routine activities, gender and neighbourhood derivatives of violence to explain the social ecology of lethal strangulation. The research findings demonstrate that homicidal strangulation in the City of Johannesburg is a unique phenomenon that is distinct from overall homicide. As the fourth leading cause of homicide in the City of Johannesburg, fatal strangulation exhibits a marked female preponderance in victimisation and distinctive socio-demographic, spatio-temporal, sex-specific and neighbourhood-level variation in risk. The study is aligned with the increasing trend towards disaggregating overall homicide into more defined and conceptually meaningful categories of homicide. The study may represent one of the first empirical investigations that also attempts to offer theoretically-derived explanations of homicidal strangulation in South Africa. Fatal strangulation is a multi-faceted phenomenon that requires multi-dimensional and multi-level interventions directed at several points of its social ecology. / Psychology / D. Phil. (Psychology)

Text in English / As an external cause of death, strangulation represents an extreme and particularly pernicious form of violence. Following the evidence gap in the extant literature, the current research examined the incidence, distributions, individual and situational predictors, and structural determinants of homicidal strangulation in the City of Johannesburg for the period 2001-2010. The thesis is structured around four discrete but interrelated studies, which collectively offer an initial contribution to the body of scholarship on homicide generally, and on the characteristics and patterns of strangulation homicide specifically. The research drew on data from the National Injury Mortality Surveillance System and the South African National Census. Study I is a descriptive study that quantifies the extent of homicidal strangulation in the City of Johannesburg and describes its distribution by characteristics of person, time, place and alcohol consumption. The remaining studies are analytical in focus, and are aimed at explaining homicidal strangulation in the City of Johannesburg in terms of its determinants. These studies are differentiated by their focus on individual-level and neighbourhood-level risks. Study II assesses overall homicide strangulation risk in relation to all the other leading causes of homicide. Study III undertakes further disaggregation to investigate homicidal strangulation risk by gender specifically. Study IV considers the socio-structural correlates and geographic distributions of fatal strangulation. The study engages select micro-level and macro-level theories that focus on the intersection between vulnerability and routine activities, gender and neighbourhood derivatives of violence to explain the social ecology of lethal strangulation. The research findings demonstrate that homicidal strangulation in the City of Johannesburg is a unique phenomenon that is distinct from overall homicide. As the fourth leading cause of homicide in the City of Johannesburg, fatal strangulation exhibits a marked female preponderance in victimisation and distinctive socio-demographic, spatio-temporal, sex-specific and neighbourhood-level variation in risk. The study is aligned with the increasing trend towards disaggregating overall homicide into more defined and conceptually meaningful categories of homicide. The study may represent one of the first empirical investigations that also attempts to offer theoretically-derived explanations of homicidal strangulation in South Africa. Fatal strangulation is a multi-faceted phenomenon that requires multi-dimensional and multi-level interventions directed at several points of its social ecology. / Psychology / D. Phil. (Psychology)