A new testing method was introduced and developed to determine Mode I fracture toughness of disc type rock specimens. The new method was named as Straight Notched Disc Bending and it uses disc specimens under three-point bending. 3D Numerical modeling was carried out with a finite element program ABAQUS to find stress intensity factors for both well-known Semi-circular Bending specimen models and Straight Notched Disc Bending specimen models for varying disc geometries. Both specimen types included notches where a crack front is introduced at the tip of the notch to compute the stress intensity factors. For stress intensity analysis, crack front-upper loading point distance and span length between the two roller supports at the bottom boundary of the specimens were changed. Fracture toughness testing was carried on Ankara G&ouml / lbaSi pink colored andesite for both specimen types / crack front-upper loading point distance and span length between the two roller supports at the bottom boundary of the specimens were changed during the tests. For both specimen geometries, notch lengths changing from 5 mm to 20 mm were used. For each notch length, two different roller supports with span lengths 60 mm and 70 mm were used. For both methods, fracture toughness values determined by using numerically computed stress intensity factors and failure loads obtained from the experiments were very close / the new method was verified by comparing the results. The new method had advantages of lower confining pressure at the crack front and lower stress intensities with a possible smaller crack tip plasticity region.

Interval uncertainty affects our decision making activities on a daily basis making the data structure of intervals of real numbers more and more popular in theoretical models and related software applications. Natural questions for people or businesses that face interval uncertainty in their data when dealing with cooperation are how to form the coalitions and how to distribute the collective gains or costs. The theory of cooperative interval games is a suitable tool for answering these questions. In this thesis, the classical theory of cooperative games is extended to cooperative interval games. First, basic notions and facts from classical cooperative game theory and interval calculus are given. Then, the model of cooperative interval games is introduced and basic definitions are given. Solution concepts of selection-type and interval-type for cooperative interval games are intensively studied. Further, special classes of cooperative interval games like convex interval games and big boss interval games are introduced and various characterizations are given. Some economic and Operations Research situations such as airport, bankruptcy and sequencing with interval data and related interval games have been also studied. Finally, some algorithmic aspects related with the interval Shapley value and the interval core are considered.

A challenge posed by future computer architectures is the efficient exploitation of their many and sometimes heterogeneous computational cores. This challenge is exacerbated by the multiple facilities for data movement and sharing across cores resident on such platforms. To answer the question of how systems software should treat heterogeneous resources, this dissertation describes an approach that (1) creates a common manageable pool for all the resources present in the platform, and then (2) provides virtual machines (VMs) with multiple `personalities', flexibly mapped to and efficiently run on the heterogeneous underlying hardware. A VM's personality is its execution context on the different types of available processing resources usable by the VM. We provide mechanisms for making such platforms manageable and evaluate coordinated scheduling policies for mapping different VM personalities on heterogeneous hardware. Towards that end, this dissertation contributes technologies that include (1) restructuring hypervisor and system functions to create high performance environments that enable flexibility of execution and data sharing, (2) scheduling and other resource management infrastructure for supporting diverse application needs and heterogeneous platform characteristics, and (3) hypervisor level policies to permit efficient and coordinated resource usage and sharing. Experimental evaluations on multiple heterogeneous platforms, like one comprised of x86-based cores with attached NVIDIA accelerators and others with asymmetric elements on chip, demonstrate the utility of the approach and its ability to efficiently host diverse applications and resource management methods.

The Internet of Things (IoTs) has triggered rapid advances in sensors, surveillance devices, wearables and body area networks with advanced Human-Computer Interfaces (HCI). Neural Networks optimized algorithmically for high accuracy and high representation power are very deep and require tremendous storage and processing capabilities leading to higher area and power costs. For developing smart front-ends for ‘always on’ sensor nodes we need to optimize for power and area. This requires considering trade-offs with respect to various entities such as resource utilization, processing time, area, power, accuracy etc. Our experimental results show that there is presence of a network configuration with minimum energy given the input constraints of an application in consideration. This presents the need for a hardware-software co-design approach. We present a highly parameterized hardware design on an FPGA allowing re-configurability and the ability to evaluate different design choices in a short amount of time. We also describe the capability of extending our design to offer run time configurability. This allows the design to be altered for different applications based on need and also allows the design to be used as a cascaded classifier beneficial for continuous sensing for low power applications. This thesis aims to evaluate the use of Restricted Boltzmann Machines for building such re-configurable low power front ends. We develop the hardware architecture for such a system and provide experimental results obtained for the case study of Posture detection for body worn cameras used for law enforcement.

Increasing atmospheric reactive nitrogen (Nr), as consequence of human activities, has generated accumulation of nitrate (NO3-) in Arctic regions. The Arctic has fragile nitrogen limited ecosystems that can be altered by increases of dry or wet deposition of Nr. Ice cores have shown increments of twofold in nitrogen deposition over Greenland and Svalbard during the 20th century. Ice core NO3- stable isotopes, (δ15N-NO3- and δ18O-NO3-), have the potential to serve as proxy of nitrogen oxides (NOx) sources and atmospheric oxidation pathways. NO3- is difficult to interpret in ice since it has several sources and experiences post-depositional processes, e.g. photolysis, relocation, evaporation and diffusion. The present work shows the results of NO3- and NO3- stable isotopes analyses of ice cores, snow and precipitation from Svalbard, in order to obtain records of natural and anthropogenic sources of NO3-. In addition, meltwater percolation effects on the snowpack ion content were also studied. A comparison between NO3- records from different Svalbard ice cores and NOx and SOx emission profiles from different regions shows that the major source regions affecting Svalbard are Western Europe and North America, followed by Central Europe and former USSR. Post-1950s δ15N-NO3- measured at Lomonosovfonna is influenced mainly by fossil fuel combustion, soil emissions, and forest fires. There is an east-west gradient in snow NO3- stable isotopes during 2010/2011, with lower δ15N-NO3- and higher δ18O-NO3- values at eastern sites; the results evidence differences in the origin of air masses arriving at Svalbard sites, mainly Eurasia, to the eastern sites, and Northern Europe to the western sites. The effects of post-depositional change on the ice core chemistry were studied, finding that 45% of annual snowpack suffers melt at Lomonosovfonna during the last 60 years. Percolation lengths were estimated as ≈1 m for most of ions; therefore, it is expected that the atmospheric ionic signal is preserved at annual or bi-annual resolution within the last 60 years at the Lomonosovfonna summit. The results presented here suggest that NO3- stable isotopes from Svalbard ice cores are useful to describe different sources and source regions of NOx, contributing to the assessment of nitrogen enrichment for this region. / Halten av reaktivt kväve (Nr) har ökat i atmosfären som en följd av mänskliga aktiviteter och har lett till en anrikning av nitrat (NO3-) nedfallet i Arktis. Ekosystemen i de Arktiska områdena är mycket sköra och extremt anpassade till den låga tillgången på tillgängligt kväve, ett förhållande som drastiskt kan ändras genom en ökad våt och torr deposition av Nr. Studier av iskärnor från både Grönland och Svalbard har visat på en fördubbling av kvävenedfallet under de senaste hundra åren. Potentiellt så kan analyser av kvävets och syrets stabila isotoper (δ15N-NO3- and δ18O-NO3-) i nitrat från iskärnor ge information om de kväveoxidkällor, oxidations- och transportvägar som påverkar nitrat halten i Arktis. Dock är nitrat sammansättningen i is svårtolkat eftersom det nitrat som återfinns i isen kan härstamma från flera olika källor och där även flera post-depositionsprocesser har beskrivits, t.ex. fotolys, evaporation, diffusion och omlokalisering som följd av snösmältning. Denna doktorsavhandling använder såväl nitrat halter som des stabila isotoper från iskärnor, snö- och nederbördsprover från Svalbard, för att fastställa de naturliga och antropogena källor av nitrat som influerar denna del av Arktis. I tillägg till detta så har även snösmältningsförsök gjorts för att undersöka hur nitratet påverkas av perkolationsprocesser. En jämförelse av nitratkompositionen mellan iskärnor och snöprofiler från olika delar av Svalbard visar att de huvudsakliga källregionerna för Nr till denna del av Arktis är Västeuropa och Nordamerika, följt av Centraleuropa och forna Sovjetunionen. För perioden 1950 och fram till i dag så påvisar δ15N-NO3- värden mätt från en Lomonosovfonna (centrala Svalbard) iskärna en stark influens från förbränning av fossila bränslen följt av utsläpp från åkermarker samt för de sista tio åren en influens av Ryska skogsbränder. Generellt så finns det även en öst-västlig gradient bland de stabila isotoperna av nitrat, med lägre δ15N-NO3- och högre δ18O-NO3- värden på Svalbards östsida i 2010/2011. Resultaten visar på en skillnad i ursprungskällan av luftmassor, där östra Svalbard påverkas mest av luft från Eurasien medan västra Svalbard påverkas med av Nordeuropa. När det gäller perkolationens effekt på jonkoncentrationen, så påvisar 60 års iskärnedata att 45% av den årliga vinterackumulationen smälts bort under sommaren på Lomonosovfonna. Detta medför en förflyttning av de flesta joner nedåt i snön, med en perkolationslängd på ca 1 m, och den ursprungliga nitrat sammansättningen förväntas därför bevaras på årlig eller vartannat års basis. Resultaten som presenteras här tyder på att iskärnedata, vad gäller nitrat och dess stabila isotoper, från Svalbard är representativa för att beskriva de olika källor och regioner som bidrar till kvävetillförseln i denna del av Arktis.

Remarkably, the central regions of galaxies are very important in shaping and influencing galaxies as a whole. As such, galaxy cores can be used for classification, to determine which processes may be important in galaxy formation and evolution. Past studies, for example, have found a dichotomy in the inner slopes of early-type galaxy surface brightness profiles. Using deprojections of the galaxies from the ACS Virgo and Fornax Cluster Surveys (ACSVCS/FCS), we show that, in fact, this dichotomy does not exist. Instead, we demonstrate that the brightest early-type galaxies tend to have central light deficits, a trend which gradually transitions to central light excesses – also known as compact stellar nuclei – as we go to fainter galaxies. This effect is quantified, and can be used to determine what evolutionary factors are important as we move along the galaxy luminosity function. The number of stellar nuclei that we observe is, in fact, an unexpected result emerging from the ACSVCS/FCS. Being three times more common than previously thought, they are present in the vast majority of intermediate and low-luminosity galaxies. Conversely, it has been known for over a decade that there is likely a supermassive black hole weighing millions to billions of solar masses at the center of virtually every galaxy of sufficient size. These black holes are known to follow scaling relations with their host galaxies. Using the ACSVCS, along with new kinematical data from long-slit spectroscopy, we measure the dynamical masses of 83 galaxies, and show that supermassive black holes and nuclei appear to fall along the same scaling relation with host mass. Both represent approximately 0.2% of their host’s mass, implying an important link between the two types of central massive objects. Finally, we extract elliptical isophotes and fit parameterized models to the surface brightness profiles of new Hubble Space Telescope imaging of the ACSVCS galaxies, observed in infrared and ultraviolet bandpasses. Taken together, the two surveys represent an unprecedented collection of isophotal and structural parameters of early-type galaxies, and will allow us to learn a great deal about the stellar populations and formation histories of galaxy cores. / Graduate

Este trabalho apresenta a medida do espectro de energia dos nêutrons no núcleo do Reator IPEN/MB-01. Para tal, foram inseridos detectores de ativação na forma de diminutas folhas metálicas no núcleo do reator, na região moderadora, utilizando um dispositivo articulado que permite que as folhas fiquem posicionadas na posição central do núcleo do reator, garantindo que todas as folhas sejam irradiadas na mesma posição. Os detectores de ativação foram selecionados de forma a cobrir grande parte do espectro de energia dos nêutrons no Reator, para tanto foram utilizadas folhas de Au197, Mg24, Ti48, In115, Sc45 entre outras. Após a irradiação os detectores de ativação foram submetidos a espectrometria gama num sistema de contagem de Germânio hiper-puro, o que possibilitou a obtenção da atividade de saturação por núcleo alvo, um dos principais dados de entrada do código computacional de desdobramento de espectro unfolding SANDBP, que através de um processo iterativo, ajusta o espectro que melhor se adequa ao conjunto de dados de entrada do código, composto principalmente pelas taxas de reação por núcleo alvo obtidas experimentalmente e um espectro inicial de entrada, neste caso obtido à nível celular pelo código Hammer-Technion para a célula representativa do núcleo do Reator, fornecendo assim a solução do espectro. / This work presents the neutron spectrum measurements in the Reactor IPEN/MB- 01 using very thin activation detectors in the metallic form, in reactor core, in moderator region. An articulated device allows that the foils are inserted in the central position of reactor core, ensuring that all the foils are irradiated in the same position. The activation detectors of different materials such Au197, Mg24, Ti48, In115, Sc45 and others, were selected to cover a large range of neutron spectrum. After the irradiation, the activation detectors were submitted to a spectrometry gamma by using a system of counting with high purity Germanium, to obtain the saturation activity per target nuclide. The saturation activity is one of the main data of input of unfolding code SANDBP, that through an iterative adjustment, modify the spectrum that better agree with the dataset of code input, composition mainly for measure reaction rate per target nuclide and a initial input spectrum, calculated for Hammer-Technion code, supplying a solution spectrum.

Made available in DSpace on 2014-10-09T12:27:12Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:56:51Z (GMT). No. of bitstreams: 0 / Dissertacao (Mestrado) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN-CNEN/SP

Made available in DSpace on 2014-10-09T12:42:41Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:57:18Z (GMT). No. of bitstreams: 1 05028.pdf: 9521463 bytes, checksum: 8a386f78270b1c225d259433bbd3167d (MD5) / Dissertacao (Mestrado) / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP

Made available in DSpace on 2015-07-22T20:49:17Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-05-26. Added 1 bitstream(s) on 2015-08-11T03:26:24Z : No. of bitstreams: 1 Publico-251a.pdf: 353518 bytes, checksum: e03912dc2d4dcadf8fed24a5e27962a1 (MD5). Added 1 bitstream(s) on 2015-08-11T03:26:25Z : No. of bitstreams: 2 Publico-251a.pdf: 353518 bytes, checksum: e03912dc2d4dcadf8fed24a5e27962a1 (MD5) Publico-251b.pdf: 370261 bytes, checksum: 21113711cf007d18765e8fd06ce0b74d (MD5). Added 1 bitstream(s) on 2015-08-11T03:26:25Z : No. of bitstreams: 3 Publico-251a.pdf: 353518 bytes, checksum: e03912dc2d4dcadf8fed24a5e27962a1 (MD5) Publico-251b.pdf: 370261 bytes, checksum: 21113711cf007d18765e8fd06ce0b74d (MD5) Publico-251c.pdf: 708805 bytes, checksum: c133f7b882649d8ce6a6e41d70d8127e (MD5). Added 1 bitstream(s) on 2015-08-11T03:26:25Z : No. of bitstreams: 4 Publico-251a.pdf: 353518 bytes, checksum: e03912dc2d4dcadf8fed24a5e27962a1 (MD5) Publico-251b.pdf: 370261 bytes, checksum: 21113711cf007d18765e8fd06ce0b74d (MD5) Publico-251c.pdf: 708805 bytes, checksum: c133f7b882649d8ce6a6e41d70d8127e (MD5) Publico-251d.pdf: 1551671 bytes, checksum: 4b389eb595b13be306af85e44e227d6c (MD5). Added 1 bitstream(s) on 2015-08-11T03:26:25Z : No. of bitstreams: 5 Publico-251a.pdf: 353518 bytes, checksum: e03912dc2d4dcadf8fed24a5e27962a1 (MD5) Publico-251b.pdf: 370261 bytes, checksum: 21113711cf007d18765e8fd06ce0b74d (MD5) Publico-251c.pdf: 708805 bytes, checksum: c133f7b882649d8ce6a6e41d70d8127e (MD5) Publico-251d.pdf: 1551671 bytes, checksum: 4b389eb595b13be306af85e44e227d6c (MD5) Publico-251e.pdf: 1825649 bytes, checksum: 913727f9a754c440b9d3b488b707f7a9 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Contexto. A diferenciação entre os nódulos benignos e malignos da tireoide é possível na maioria das vezes pela citologia, obtida por punção aspirativa com agulha fina (PAAF). No caso das neoplasias foliculares (NFs), é necessária a remoção cirúrgica de ao menos parte da glândula, pois somente a análise histológica da peça permite diferenciar nódulos benignos de malignos. Há indícios na literatura de que os parâmetros de ultrassonografia com Doppler colorido (UDC) podem indicar maior ou menor probabilidade de malignidade nesses nódulos. Objetivo. Avaliar, nesta revisão sistemática, a acurácia diagnóstica da UDC em predizer malignidade em NFs da tireoide. Métodos. Foram pesquisadas as seguintes bases de dados: MEDLINE, Web of Science, EMBASE, Cochrane Library e LILACS. As referências de estudos relevantes foram verificadas para adicionais citações de interesse. Não houve restrições de linguagem. Foram incluídos estudos nos quais nódulos de tireoide com padrão histológico compatível com NFs, confirmado por biópsia de peça cirúrgica, haviam sido previamente submetidos à UDC e cujos padrões de fluxo ao mapeamento colorido foram descritos detalhadamente. Dois revisores realizaram independentemente a avaliação da qualidade e a extração de dados. Resultados. Foram incluídos quatro estudos, somando 457 nódulos, sendo que 67 foram considerados malignos com base na biópsia cirúrgica. A presença à UDC de padrões de fluxo no interior do nódulo considerado como moderado, rico, predominante ou exclusivo foi indicativa de malignidade com uma sensibilidade média de 85 % (Intervalo de confiança [IC] 95%: 74% a 93%) e especificidade média de 86% (IC 95%: 82% a 89%). Para uma prevalência média de 14,7%, os valores preditivos positivo e negativo são respectivamente 51% e 97%. A razão de verossimilhança positiva é 6,07, e a razão de verossimilhança negativa 0,18. Conclusão. A UDC apresenta boa acurácia para identificar malignidade em NFs da tireóide. Fluxo interno predominante à UDC está associado a maior risco de malignidade nessas lesões. / Background. In most cases it is possible to differentiate between benign and malignant thyroid nodules through cytologic analysis of the fine-needle aspiration biopsy (FNAB) samples. However, in the case of follicular neoplasms (FNs), to determine whether such nodules are benign or malignant, it is necessary to perform a surgical biopsy, which requires the removal of at least part of the thyroid gland. There are clues in the literature that the parameters of color Doppler ultrasonography (CDU) may indicate a greater or lesser likelihood of malignancy in these nodules. Objective. The purpose of this systematic review was to obtain summary estimates of the diagnostic accuracy of CDU in predicting malignancy in thyroid FNs. Methods. We searched Medical Subject Headings together with the search terms “follicular,” “thyroid,” and “Doppler” in the MEDLINE, Web of Science, Cochrane Library and Excerpta Medica databases as well as the Latin American and Caribbean Health Sciences Literature database, after which we performed manual searches of the reference lists to locate additional studies. There were no language restrictions. We included studies that assessed the diagnostic accuracy of CDU in identifying malignancy in thyroid FNs. The assessments of the quality and extraction of data were performed by 2 independent reviewers. Results. We included 4 studies, which collectively evaluated 457 thyroid FNs, 67 of which had been classified as malignant based on the evaluation of surgical biopsy samples. Moderate, rich, predominant, or exclusive internal flow on CDU of thyroid FNs was considered indicative of malignancy. The overall sensitivity of CDU was 85% (95% confidence interval [CI], 74%–93%), with an overall specificity of 86% (95% CI, 82%–89%). The overall prevalence was 14.7%, and the positive and negative predictive values were 51% and 97%, respectively. The positive likelihood ratio was 6.07, and the negative likelihood ratio was 0.18. Conclusion. CDU has good accuracy for identifying malignancy in thyroid FN. Predominant internal flow seen on CDU is associated with malignancy of thyroid FN. / TEDE / BV UNIFESP: Teses e dissertações