No description available.

This thesis develops novel statistical methodology for estimating the incidence and the prevalence of Human Immunodeficiency Virus (HIV) using routinely collected surveillance data. The robust estimation of HIV incidence and prevalence is crucial to correctly evaluate the effectiveness of targeted public health interventions and to accurately predict the HIV- related burden imposed on healthcare services. Bayesian CD4-based multi-state back-calculation methods are a key tool for monitoring the HIV epidemic, providing estimates of HIV incidence and diagnosis rates by disentangling their competing contribution to the observed surveillance data. Improving the effectiveness of public health interventions, requires targeting specific age-groups at high risk of infection; however, existing methods are limited in that they do not allow for such subgroups to be identified. Therefore the methodological focus of this thesis lies in developing a rigorous statistical framework for age-dependent back-calculation in order to achieve the joint estimation of age-and-time dependent HIV incidence and diagnosis rates. Key challenges we specifically addressed include ensuring the computational feasibility of proposed methods, an issue that has previously hindered extensions of back-calculation, and achieving the joint modelling of time-and-age specific incidence. The suitability of non-parametric bivariate smoothing methods for modelling the age-and-time specific incidence has been investigated in detail within comprehensive simulation studies. Furthermore, in order to enhance the generalisability of the proposed model, we developed back-calculation that can admit surveillance data less rich in detail; these handle surveillance data collected from an intermediate point of the epidemic, or only available on a coarse scale, and concern both age-dependent and age-independent back-calculation. The applicability of the proposed methods is illustrated using routinely collected surveillance data from England and Wales, for the HIV epidemic among men who have sex with men (MSM).

The Diploma Thesis deals with the technology of manufacturing the titan exhaust for a two-stroke engine. It includes a description of the manufacturing procedure. It analyzes in detail the forming issue of sheet-metal parts in order to eliminate the spring-back effect, and it finds possible solutions to this problem.

The geological disposal of nuclear waste, in underground openings and the long-term performance of these openings demand a detailed understanding of fundamental rock mechanics. A full scale field experiment: Äspö Pillar Stability Experiment was conducted at a depth of 450 m in sparsely fractured granitic rock to examine the rock mass response between two deposition holes. An oval shaped tunnel was excavated parallel to the σ3 direction to provide access to the experiment and also provide elevated stress magnitudes in the floor. In the tunnel floor two 1.75-m diameter 6-m deep boreholes were excavated so that a 1-m thick pillar was created between them. In one of the holes a confinement pressure of 700 kPa was applied and in the other displacement transducers were installed. The pillar volume was monitored by an Acoustic Emission System. Spatially distributed thermocouples were used to monitor the temperature development as the pillar was heated by electrical heaters. The excavation-induced stress together with the thermal-induced stress was sufficient to cause the wall of the open borehole to yield. The temperature-induced stress was increased slowly to enable detailed studies of the rock mass yielding process. Once the rock mass loading response was observed, the rock mass was unloaded using a de-stress slotting technique. This thesis focuses on the in-situ study of the rock mass response to coupled mechanical thermal loading and thermal-mechanical unloading. The experiment, its design, monitoring and observations are thoroughly described. An estimate of the yielding strength of the rock mass is presented and compared with laboratory test and results from other rock mass conditions reported elsewhere in the open literature. General conclusions about the effect of the confining pressure and the observations from the unloading of the pillar are also presented. Important findings are that the yielding strength of the rock mass has been successfully determined, low confinement pressures significantly affects the onset of yielding, the primary mode of fracture initiation and propagation is extensional, no significant time dependency of the yielding process was observed. The unloading studies also indicated that what appeared to be shear bands likely was a propagating zone of extensile failure that weakened the rock so that displacements in the shear direction could occur. / <p>QC 20100622</p>

A model is presented to use normalized multi-linear tension and compression material characteristics of strain-hardening textile reinforced concrete and derive closed form expressions for predicting moment-curvature capacity. A set of design equations are derived and simplified for use in spreadsheet based applications. The model is applicable for both strain-softening and strainhardening materials. The predictability of the simplified model is checked by model calibration and development of design charts for moment capacity and stress developed throughout the cross section of a flexural member. Model is calibrated by predicting the results of Alkali Resistant Glass and Polyethylene fabrics. A case for the flexural design of Glass Fiber Reinforced Concrete (GFRC) specimen as a simply supported beam subjected to distributed load is used to demonstrate the design procedure.

La température de l'eau joue un rôle majeur dans le cycle de vie des poissons. Dans un contexte de changement climatique global, le réchauffement peut avoir un impact fort sur la croissance, la fécondité et la survie. L'enjeu de cette thèse est la modélisation mathématique de l'influence de la température sur les traits d'histoire de vie d'une population de chabot (Cottus gobio) afin de faire de la prédiction à la fois au niveau individuel et populationnel. Les données expérimentales qui permettront de calibrer les modèles sont issues du bassin de la Drôme (France) et plus particulièrement du sous-bassin du Bez. Dans une première étape, il s'agit de développer un modèle de rétrocalcul qui peut être utilisé pour calculer les longueurs individuelles des chabots aux âges précédents à partir des données mesurées à la capture. Il s'agit, dans un deuxième temps, de développer un modèle de croissance dépendant de la température de l'eau qui sert à prédire la longueur moyenne des chabots à un âge donné. Enfin, il s'agit de passer de l'échelle de l'individu à celle de la population en prenant en compte tous les traits d'histoire de vie et leurs dépendances vis-à-vis de la température. Plus précisément, un modèle matriciel de type Leslie, à la fois dépendant du temps et de la température, structuré en classe d'âges est développé et utilisé pour prédire la dynamique de population sous différents scénario du réchauffement climatique / Water temperature plays a key role in the life cycle of fish. Therefore, increasing temperatures due to the expected climate change may have a strong impact on growth, fecundity and survival. The goal of this thesis is to model the impact of temperature on the life-history traits of a bullhead population (Cottus gobio) in order to make predictions both at individual and at population level. The models developed here are calibrated on experimental field data from a population living in the Bez River network (Drôme, France). First, a new back-calculation model is derived that can be used to compute individual fish body lengths at earlier ages from capture data. Next, a growth model is proposed that incorporates the water temperature and can be used to predict the mean length at a given age and temperature. Finally, the population is modelled as a whole by linking all life-history traits to temperature. For this purpose, a spatialised time- and temperature-dependent Leslie matrix model structured in age classes was used to predict the population dynamics under different temperature scenarios

A model is presented to use normalized multi-linear tension and compression material characteristics of strain-hardening textile reinforced concrete and derive closed form expressions for predicting moment-curvature capacity. A set of design equations are derived and simplified for use in spreadsheet based applications. The model is applicable for both strain-softening and strainhardening materials. The predictability of the simplified model is checked by model calibration and development of design charts for moment capacity and stress developed throughout the cross section of a flexural member. Model is calibrated by predicting the results of Alkali Resistant Glass and Polyethylene fabrics. A case for the flexural design of Glass Fiber Reinforced Concrete (GFRC) specimen as a simply supported beam subjected to distributed load is used to demonstrate the design procedure.

No description available.