Made available in DSpace on 2015-03-26T13:54:53Z (GMT). No. of bitstreams: 1 texto completo.pdf: 443172 bytes, checksum: 25810212202085897d024a4831c5557d (MD5) Previous issue date: 2008-07-30 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The present study involved two experiments that aimed to evaluate nutrient intake and total digestibility, pH and ruminal ammonia concentration (first experiment) as well as the productive performance of beef cattle (second experiment) that were fed with diets containing soybean (SS) 0; 25; 50; 75; 100%. The rate forage:concentrate was 60:40, based on the dry matter. In the first experiment, five HxZ crossbred animals fistulated in the rumen and abomasum that had an initial weight of about 390kg were used. The animals were distributed in a 5x5 Latin square. Chromc oxide was used to estimate the fecal excretion. The intakes of dry matter (DM), organic matter (OM), crude protein (CP), neutral detergent fiber (NDF), nonfiber carbohydrates (NFC), total carbohydrates (TC) and total digestible nutrients (TDN), expressed as kg.dia-1, showeda decreasing linear behavior while the intakes of EE and CP adjusted to the linear response plateau (LRP) model, thus estimating maximum values for the rates of 84 and 57% of corn silage in the forage, respectively. Diet effects were not detected (P>0.05) on the total apparent digestibility of DM, OG, CP, NDF and TC they revealed and average of 66.96; 69.49; 63.49; 57.60; 82.34 and 71.10%, respectively. On the other hand, EE aparent digestibility increased linearly following the corn silage rates. Quadratic effect of time collection for the ammonia concentration and the ruminal pH was observed where the maximum and minimal values 17.42 mg/dL and 6.23 mg/dL at 2,54and 4,27 hours were respectively estimated after feeding. In the second experiment, 35 anerolados and emasculated animals initially weighing 370 kg were used.The animals were distributed into a randomized block design with five treatments and seven repetitions. The experiment, which lasted 99 days long, was divided into a 15-day adaptation period and three 28-day periods. Five animals that were used as references were killed after the adaptation period in order to estimate the carcass gain. For estimating fecal excretion, ingestible neutral detergent fiber (iNDF) was used as an indicator, after the 240 hours "in situ" incubation. The intakes of NFC and CHO decreased linearly (P<0.05) according to the soybean silage rates in the diets, while the EE intake adjusted to a LRP model, with maximum value at the 89.55% rate of soybean silage in the forage. Lower intakes (P<0.05) of NCF and TC were observed for the diets that had higher rates of soybean silage than corn silage. Intakes for the other nutrients, expressed as kg/day, were not influenced by the diets. Weight and carcass gain, carcass yield and food conversion were not influenced by the diets either, they showed average values of 1.12 kg.day-1, 0.67 kg.day-1, 53.00% and 8.05, respectively. Soybean silage may be used as the single forage source in diets for anerolados cattle since it does not compromise the cattle productive performance or the ruminal fermentation patterns. However, its use may become and interesting alternative when associated to corn silage. / O presente estudo envolveu dois experimentos e foi realizado objetivando-se avaliar o consumo e a digestibilidade total dos nutrientes, o pH e a concentração de amônia ruminal (experimento1), bem como o desempenho produtivo de bovinos de corte (experimento 2) alimentados com dietas contendo 0; 25; 50; 75; 100% de silagens de soja (SS). A relação volumoso:concentrado foi de 60:40, com base na matéria seca. No primeiro experimento, foram utilizados cinco animais mestiços Holandês × Zebu, fistulados no rúmen e no abomaso, com peso inicial médio de 390 kg, distribuídos em um quadrado latino 5 × 5. Usou-se o óxido crômico para estimar a excreção fecal. Os consumos de matéria seca (MS), matéria orgânica (MO), proteína bruta (PB), fibra em detergente neutro (FDN), carboidratos nãofibrosos (CNF), carboidratos totais (CHOT) e nutrientes digestíveis totais (NDT), expressos em kg.dia-1, apresentaram comportamento linear decrescente, enquanto os consumos de EE e PB ajustaram-se ao modelo linear com resposta platô (LRP), estimando-se valores máximos para os níveis de 84 e 57% de silagem de soja no volumoso, respectivamente. Não foi detectado efeito (P>0,05) das dietas sobre as digestibilidades aparentes totais de MS, MO, PB, FDN, CNF, CHOT, cujas médias foram de 66,96; 69,49; 63,49; 57,60; 82,34; 71,10%, respectivamente. Por sua vez, a digestibilidade aparente do EE aumentou linearmente com os níveis de silagem de soja. O tempo decoleta teve efeito quadrático sobre a concentração de amônia e o pH ruminal, que apresentaram valores máximo e mínimo de 17,42 mg/dL e 6,23 mg/dL às 2,54 e 4,27 horas após a alimentação, respectivamente. No experimento 2, foram utilizados 35 animais anerolados, castrados, com peso inicial de 370 kg, distribuídos em delineamento de blocos casualizados, com cinco tratamentos e sete repetições. O experimento teve duração de 99 dias, divididos em três períodos de 28 dias, após 15 dias de adaptação. Cinco animais- referência foram abatidos após o período de adaptação, para estimativa do ganho de carcaça. A estimativa da excreção fecal foi feita utilizando-se a fibra em detergente neutro indigestível (FDNi) como indicador, após incubação in situ por 240 horas. Os consumos de CNF e carboidratos totais (CHOT) decresceram linearmente (P<0,05) com os níveis de silagem de soja nas dietas, enquanto o consumo de EE ajustou-se ao modelo LRP, com valor máximo para o nível de 89,55% de silagem de soja no volumoso. Os menores consumos de CNF e CHOT (P<0,05) foram observados nos animais mantidos com as dietas com maiores proporções de silagem de soja. Os consumos dos demais nutrientes, em kg/dia, não foram influenciados pelas dietas. O ganho de peso, ganho de carcaça, rendimento de carcaça e a conversão alimentar também não foram influenciados pelas dietas e apresentaram médias de 1,12 kg.dia-1, 0,67 kg.dia-1, 53,00% e 8,05, respectivamente. A silagem de soja pode ser utilizada como fonte única de volumoso em dietas para bovinos anelorados, uma vez que não compromete o desempenho produtivo e o padrão de fermentação ruminal desses animais. No entanto, proporciona melhores resultados se utilizada em associação à silagem de milho.

Une technique de Contrôle Non Destructif (CND) a été développée, permettant la génération et la détection d’ondes de Lamb guidées le long de grandes plaques constituées de différents matériaux (métal, polymère ou composite renforcé par des fibres). Basée sur l’emploi de nombreux éléments étroitement couplés à la plaque, cette technique d’inspection diffère du Structural Health Monitoring (SHM) classique car tous les éléments émetteurs ou récepteurs sont regroupés dans une zone très localisée, définie par la surface active d’une sonde multiélément matricielle, et ne sont pas fixés de manière permanente et distribuée au sein, ou en surface, de la structure testée. De plus, le principe (connu) du déphasage entre éléments est appliqué à la sonde pour la génération et la réception d’un mode de Lamb pur dans (ou provenant) de multiples directions le long de la plaque. Les lois de retards appliquées à ces éléments, aussi bien lorsque la sonde fonctionne en mode émission qu’en mode réception, prennent en compte la nature dispersive de l’onde de Lamb. Enfin, un traitement de signal spécifique est appliqué pour compenser la dispersion subie par les ondes guidées au cours de leur propagation le long de la pièce testée. Un prototype expérimental et sa modélisation par éléments finis sont présentés, ainsi que les mesures et les résultats simulés de ses performances en termes de sélectivité modale et de directivité angulaire. Concernant les applications de CND, la construction d’images, représentatives de toutes les parties de la pièce testée qui diffractent le mode guidé (bord de pièce, défauts, trous, raidisseurs, etc.), a permis de démontrer le potentiel (et quelques limites) de cette technique vis-à-vis d’une inspection rapide de grandes structures, y compris de zones éloignées de la sonde ou encore de zones difficiles d’accès. / A technique of Non-Destructive Testing (NDT) was developed for the generation and detection of Lamb waves propagating along large plates made of different materials (metal, polymer or fibre-reinforced composite). Based on the use of many elements closely coupled to the plate, this inspection technique differs from the classic Structural Health Monitoring (SHM) because all the transmitters or receivers are grouped in a very localized area, defined by the active surface of a phased array matrix probe, and are not permanently attached and distributed within or on the surface of the test structure. In addition, the principle (known) of the phase shift between the elements is applied to the probe for generating and receiving a pure Lamb mode in (or from) multiple directions along the plate. The delay laws applied to these elements, in transmit mode or receive mode, take into account the dispersive nature of the Lamb wave. Finally, a specific signal processing is applied to compensate the dispersion suffered by the guided waves during their propagation along the test piece. An experimental prototype and its finite element modeling are presented, as well as measurements and simulation results of its performances in terms of modal selectivity and angular directivity. For NDT applications, the construction of images, representing all parts of the test piece, which diffract the guided mode (edges, defects, holes, stiffeners, etc.), has demonstrated the potential (and some limits) of this technique for a quick inspection of large structures, including areas remote from the probe or areas difficult to access.

This thesis deals with design of an eddy current transducer which enables non-destructive inspection of composite aircraft structures primarily carbon fiber reinforced plastic (CFRP) in areas of manufacture and maintenance. The design of the transducer is based on analytical-experimental approach and its electrical and mechanical parameters were optimized to ensure a good signal to noise ratio at the six composite samples. These samples contain artificial discontinuities in the form of various types of defects. These defects are simulating the various types of damage created in the aircraft structure, especially delamination or thickness changes of composite materials. The experimental measurements, data collection and non-destructive evaluation were performed during the period. The result of this work is functional eddy current probe, which is reliably able to detect some damage of the carbon composite structures to the depth of 3,9 mm.

Aerospace structural components made from polymer matrix composites (PMCs) offer numerous advantages. Their high stiffness and high strength combined with low densities enable lower fuel consumption coupled with higher payloads. As a result, PMCs provide an important economic advantage over typical metallic airframes. Textile reinforcements for PMCs are made by assembling reinforcement fibres, typically carbon. Then, the textile reinforcements are typically cut into smaller pieces, stacked, draped and assembled into a dry assembly called a preform, the shape of which generally approaches that of the PMC part to be made. This manufacturing process is labour intensive and expensive. Novel thick, net-shape, drapable, high vf textile reinforcements used toward manufacturing aerospace PMCs are being developed at the University of Ottawa. The technology enables the manufacturing of flat, drapable multilayered near net-shape preforms. The bending and in-plane shear behaviours of such novel thick reinforcement textiles was investigated to understand and define the behaviour of such thick fabric reinforcements when formed into required shapes. A bending apparatus was developed for investigating the bending behaviour of these novel thick reinforcement fabrics and an articulated frame shear rig was used for investigating the in-plane shear behaviour. A non-destructive inspection method using infrared imaging was used for investigating and identifying flaws and defects in these thick, dry textile reinforcements, aiming at increasing the quality and reproducibility of the final PMC parts made from these reinforcements.

Cette thèse s’est déroulée dans le contexte de l’évaluation non destructive du collage de patches composites employés pour la réparation de structures aéronautiques. Le besoin a été exprimé par la DGA Techniques Aéronautiques (Toulouse). Le travail effectué a consisté en l’exploitation des ondes guidées, de type SH puis Lamb, pour mesurer des caractéristiques mécaniques d'un joint de colle et tenter d’évaluer, de manière non destructive, la qualité du collage. Dans un premier temps, les propriétés viscoélastiques des matériaux assemblés sont déterminées. Elles servent alors de données d’entrée à des modèles, basés sur la méthode SAFE (Semi Analytical Finite Element method), permettant d’obtenir les courbes de dispersion ainsi que les champs des différents modes guidés se propageant le long de l’assemblage collé. Ces simulations numériques permettent alors d’évaluer le potentiel des ondes guidées à caractériser un adhésif. Des mesures expérimentales sont ensuite menées pour confirmer les prédictions numériques. Finalement la résolution d’un problème inverse, qui consiste à évaluer les modules de rigidité de la couche de colle se situant entre une plaque d’aluminium et un patch, est proposée pour établir des pistes d’évaluation non destructive d’un collage répondant au besoin industriel. Cette étude a été financée par la DGA, France. / Repairing of metallic structures using composite patches bonded with an adhesive layer is more and more common in the aeronautic field, and particularly at DGA Techniques Aéronautiques. In order to guarantee the quality of this bond, the sensitivity of ultrasonic guided waves is investigated. Shear-horizontally polarized (SH) waves, and Lamb waves are investigated to infer the stiffness of the adhesive layer. The SAFE method is used to predict the dispersion curves and mode shapes of the different wave modes propagating along the three-layer assembly. Numerical simulations are run for selecting the most appropriate wave modes, i.e. with higher sensitivity to the stiffness of the bond than to other components properties. Experiments are also made for generating-detecting pre-selected SH or Lamb wave modes in order to confirm the numerical predictions. Finally the resolution of an inverse problem, consisting in the evaluation of the stiffness modulus of the bond layer between an aluminium plate and a carbon-epoxy composite patch, at different curing time, is proposed as a contribution to the establishment of possible strategies for bonds testing. This work was supported by DGA, France.

Målet med projektet är att undersöka om det med matematiska modeller eller datorsimulationer går att förutsäga hur fibermängd mellan 20 kg/m³ och 60 kg/m³ i stålfiberarmerad sprutbetong med tjocklek mellan 5 cm och 15 cm påverkar attenuering av RFID signaler. Detta är intressant att undersöka eftersom RFID är en möjlig teknik att använda för mätning av betongtjocklek. Stålfibrernas påverkan på RFID signalen tillsammas med de möjligheter som finns att förutsäga densamma bestämmer i stor grad RFID-teknikens lämplighet att användas för mätning av betongtjocklek. I rapporten undersöks homogenisering med hjälp av Maxwell-Garnetts blandningsformel tillsammans med en analytisk metod som jämförs med en datorsimulation i CST Microwave studio. Utöver det utförs ytterligare en datorsimulation i CST där stålfiberarmerad betong modelleras som inhomogen bestående av stålfiber och betong. Mätdata från experiment som utförts på stålfiberarmerade betongblock presenteras och jämförs med de simulerade och beräknade fallen. Av de metoder som undersöks i projektet har ingen validerats. Beräkningarna för varje metod ger ofta olika resultat, varav inga är direkt jämförbara med den data som uppmätts experimentellt. Fibrernas påverkan på signalen verkar däremot vara signifikant och bidra med stora mätosäkerheter och förluster. Enligt den inhomgena simulationen är Maxwell-Garnetts blandningsmetod olämplig för syftet. Den tagg som testas verkar inte vara lämpad för detta användningsområde. Svårigheter med radiokommunikation genom betong identifieras och förslag ges på hur en tagg kan konstrueras för att fungera bättre i miljön.

For autonomous mine vehicles that perform the ”load-haul-dump” (LHD) cycle to operate properly, calibration of the sensors they rely on is crucial. The LHD cycle refers to a vehicle that loads material, hauls the material along a route and dumps it in an extraction point. Many of these vehicles are waist-actuated, meaning that the front and rear part of the machines are fixated at an articulation point.   The focus of this thesis is about developing and implementing two differ- ent frameworks to distinguish patterns from routes where calibration of the hinge-angle sensor was needed before and try to predict when calibrating the sensor is needed. We present comparative results of one method using ma- chine learning, specifically supervised learning with support vector machine and one optimization-based method using scan matching by implementing a two-dimensional NDT (Normal Distributions Transform) algorithm.   Comparative results based on evaluation metrics used in this thesis show that detecting incorrect behaviour of the hinge-angle sensor is possible. Evaluation show that the machine learning classifier performs better on the data used for this thesis than the optimization-based classifier.

This thesis deals with design of an eddy current transducer which enables non-destructive inspection of composite aircraft structures primarily carbon fiber reinforced plastic (CFRP) in areas of manufacture and maintenance. The design of the transducer is based on analytical-experimental approach and its electrical and mechanical parameters were optimized to ensure a good signal to noise ratio at the six composite samples. These samples contain artificial discontinuities in the form of various types of defects. These defects are simulating the various types of damage created in the aircraft structure, especially delamination or thickness changes of composite materials. The experimental measurements, data collection and non-destructive evaluation were performed during the period. The result of this work is functional eddy current probe, which is reliably able to detect some damage of the carbon composite structures to the depth of 3,9 mm.

This diploma thesis deals with the evaluation of the process of performing non-destructive testing in the Nuclear energy division of the ČEZ, a. s. and the design of changes for this process and the information systems involved. The first part of the diploma thesis defines the basic terminology and the theoretical basis for further use. The second part includes an analysis of the current state of the process of performing non-destructive testing and the relevant systems that are involved in the process. The third part contains a proposal to improve the process of conducting non-destructive testing, further proposals, and economic evaluation of proposals

This thesis deals with non-destructive testing (NDT). It Describes two basic thermographic methods, lock-in and pulse thermography. Different types of measurements using Fluke Ti55, points to the possibility of applying these methods in practice. Pulse thermography issue is a substantial part of this work, due to its good applicability. Also, it compares visual testing methods (VIS) and pulse thermography due to measurement with die-casting alloy AC-AlSi9Cu3 (Fe).