Fast, sensitive, quantitative, and low-cost analysis of environmental pollutants is highly valuable for environmental monitoring. Due to its single-molecule sensitivity and fingerprint specificity, surface-enhanced Raman spectroscopy (SERS) has been widely employed for heavy metal, organic compound, and pathogen detection. However, SERS quantitation is challenging because 1) analytes do not stay in the strongest enhancing region ("hot spots") and 2) SERS reproducibility is poor. In this dissertation, gold nanoparticle/bacterial cellulose (AuNP/BC) substrates were developed to improve SERS sensitivity by increasing hot spot density within the laser excitation volume. Environmentally relevant organic amines were fixed at "hot spots" by lowering solution pH below the analyte pKa and thus enabling SERS quantitation. In addition, a new SERS internal standard was developed based upon the electromagnetic enhancement mechanism that relates Rayleigh (elastic) and Raman (in-elastic) scattering. Rayleigh scattering arising from the amplified spontaneous emission of the excitation laser was employed as a normalization factor to minimize the inherent SERS signal variation caused by the heterogeneous distribution of "hot spots" across a SERS substrate. This highly novel technique, hot spot-normalized SERS (HSNSERS), was subsequently applied to evaluate the efficiency of SERS substrates, provide in situ monitoring of ligand exchange kinetics on the AuNP surface, and to reveal the relationship between the pKa of aromatic amines and their affinity to citrate-coated AuNPs (cit-AuNPs). Finally, colloidally stable stable pH nanoprobes were synthesized using co-solvent mediated AuNP aggregation and subsequent coating of poly(ethylene) glycol (PEG). These nanoprobes were applied for pH detection in cancer cells and in phosphate buffered aerosol droplets. The latter experiments suggest that stable pH gradients exist in aerosol droplets. / PHD / Traditional analytical methods, such as gas chromatography/mass spectroscopy, liquid chromatography/mass spectroscopy, etc., cannot meet the demand for rapid screening of target environmental pollutants in drinking water. This issue arises due to the requirements for time-consuming sample pre-treatment, well-trained experts, complex instrumental parameter optimization, and scale challenges that limit onsite measurement. Surface-enhanced Raman spectroscopy is a promising approach to overcome these limitations. To improve SERS quantitation, surface-enhanced elastic scattering was developed as a novel internal standard to account for the SERS signal variation caused by substrate heterogeneity (“hot spot” normalization). Compared with traditional SERS internal standards, using scattered light as an internal standard reduces cost, time, interference, and experimental complexity for SERS detection. With this novel approach, the kinetics of adsorption/desorption of guest ligands/citrate onto/from the AuNP surface were quantified in situ and in real time. In addition, the SERS intensities of organic amines acquired at different solution pH values were differentiated using “hot spot” normalization, which revealed the relationship between aromatic amine pK_a and their affinity to the AuNP surface. Finally, the chemistry in confined aqueous environments, such as aerosol droplets, membrane channels, and cells, is challenging to probe using conventional analytical tools due to their inaccessible small volumes. To address this problem, SERS pH nanoprobes were synthesized and used to detect the pH inside cancer cells and micrometer-sized aerosol droplets.

Although several studies have proven the accuracy of using a non-contact, air-coupled receiver in nonlinear ultrasonic (NLU) Rayleigh wave measurements, inconsistent results have been observed when working with narrow specimens. The objectives of this research are first, to develop a 3D numerical finite element (FE) model which predicts nonlinear ultrasonic measurements and second, to apply the validated model on the narrow waveguide to determine causes of the previously observed experimental issues. The commercial FE-solver ABAQUS is used to perform these simulations. Constitutive law and excitation source properties are adjusted to match experiments conducted, considering inherent effects of the non-contact detection, such as frequency dependent pressure wave attenuation and signal averaging. Comparison of “infinite” and narrow width simulations outlines various influences which impair the nonlinear Rayleigh wave measurements. When the wave expansion is restricted, amplitudes of the fundamental and second harmonic components decrease more significantly and the Rayleigh wavefronts show an oscillating interaction with the boundary. Because of the air-coupled receiver’s finite width, it is sensitive to these edge effects which alter the observed signal. Thus, the narrow specimen adversely affects key factors needed for consistent measurement of material nonlinearity with an air-coupled, non-contact receiver.

We examine the Rayleigh–Bénard convection as modelled by the Boussinesq equation. Our aim is at deriving bounds for the heat enhancement factor in the vertical direction, the Nusselt number, which reproduce physical scalings. In the first part of the dissertation, we examine the the simpler model when the acceleration of the fluid is neglected (Pr=∞) and prove the non-optimality of the temperature background field method by showing a lower bound for the Nusselt number associated to it. In the second part we consider the full model (Pr<∞) and we prove a new upper bound which improve the existing ones (for large Pr numbers) and catches a transition at Pr~Ra^(1/3).

Σκοπός αυτής της εργασίας είναι η μελέτη τεχνικών κατανεμημένου προσανατολισμού σε πραγματικές συνθήκες. Πιο συγκεκριμένα σε αυτά στα συστήματα θεωρείται ότι ο κόμβος-πομπός δεν έχει καλή σύνδεση με το δέκτη και κατά συνέπεια δεν μπορεί να επικοινωνήσει απευθείας με τον κόμβο-δέκτη χωρίς δραματική αύξηση της ενέργειας μετάδοσης. Παρόλα αυτά η χρήση κατανεμημένου προσανατολισμού δίνει τη δυνατότητα να βελτιωθεί σημαντικά η κατανάλωση ενέργειας. Το σχήμα που θα χρησιμοποιηθεί είναι ενίσχυση και προώθηση (AF) 2 βημάτων, με το οποίο οι συνεργατικοί κόμβοι απλώς ενισχύουν και στην συνέχεια επαναμεταδίδουν το μήνυμα. Συνεπώς, ζητούμενο είναι η εύρεση των μιγαδικών βαρών με τα οποία πρέπει ο κάθε συνεργαζόμενος κόμβος χωριστά να ενισχύσει το σήμα. Οι τεχνικές που θα χρησιμοποιηθούν έχουν ως κριτήρια την ελαχιστοποίηση της ενέργειας μετάδοσης με ταυτόχρονη ικανοποίηση του SNR, μεγιστοποίηση του SNR με περιορισμένη ολική ενέργεια μετάδοσης και μεγιστοποίηση του SNR με περιορισμένη ενέργεια μετάδοσης ανά συνεργαζόμενο κόμβο. Το πρώτο κριτήριο θα εξεταστεί, επίσης, και σε συστήματα με πολλαπλούς πομπούς και δέκτες. Λόγω της φύσης του προβλήματος, ο κατανεμημένος προσανατολισμός αναμένεται να έχει μεγάλη απήχηση σε συστήματα με πολλούς διασκορπιστές και εμπόδια, όπως σε ένα αστικό περιβάλλον, και, επομένως, είναι λογικό να θεωρηθεί ότι τα κανάλια του συστήματος είναι Rayleigh, δηλαδή ασυσχέτιστα χωρίς οπτική επαφή (LOS). Για να προσομοιωθεί το σύστημα σε πραγματικές συνθήκες οι μέθοδοι που θα υλοποιήσουμε στην εργασία χρησιμοποιούν τα στατιστικά του καναλιού. Επιπλέον, η εκτίμηση καναλιού εφόσον θεωρούμε ότι έχουμε Gaussian λευκό θόρυβο θα γίνει με την χρήση του βέλτιστου γραμμικού εκτιμητή (BLUE). Η επίδραση της εκτίμησης του καναλιού θα μελετηθεί για δύο περιπτώσεις: με αμοιβαία και χωρίς αμοιβαία κανάλια. / The purpose of this thesis is the study of methods of distributed beamforming under real circumstances. More specifically, these systems are considered that the transmitter must increase tremendously the required transmit energy to communicate with the receiver. However the use of the distributed beamforming allows the system to improve the energy consumption. The scheme that is used from relays is amplify and forward of two steps, where the relays only amplify and then forward the message to the destination. That is, the purpose is to find the complex weights to be used by the corresponding relay so as to amplify the message of the transmitter. The methods that are implemented have as criterions the minimization of transmit energy while satisfying the SNR, maximization of SNR while limiting the system's transmit energy and maximization of SNR while limiting transmit energy of each relay individually. The first criterion is also studied at systems with more than one pair transmitter-receiver. Due to the nature of the problem, distributed beamforming is expected to be used at environments with many obstacles and scatterers, like urban environment, and so it is rationale to suppose that the channels should be Rayleigh, meaning uncorrelated without line of sight. To simulate the system under real circumstances the methods that we will implement shall use the second order statistics of the channels. Moreover, due to Gaussian white noise, channels are estimated using the Best Linear Unbiased Estimator. The impact of channel estimation is studied in two cases: "reciprocal" and "not reciprocal".

L’agrégation érythrocytaire est le principal facteur responsable des propriétés non newtoniennes sanguines pour des conditions d’écoulement à faible cisaillement. Lorsque les globules rouges s’agrègent, ils forment des rouleaux et des structures tridimensionnelles enchevêtrées qui font passer la viscosité sanguine de quelques mPa.s à une centaine de mPa.s. Cette organisation microstructurale érythrocytaire est maintenue par des liens inter-globulaires de faible énergie, lesquels sont brisés par une augmentation du cisaillement. Ces propriétés macroscopiques sont bien connues. Toutefois, les liens étiologiques entre ces propriétés rhéologiques générales et leurs effets pathophysiologiques demeurent difficiles à évaluer in vivo puisque les propriétés sanguines sont dynamiques et fortement tributaires des conditions d’écoulement. Ainsi, à partir de propriétés rhéologiques mesurées in vitro dans des conditions contrôlées, il devient difficile d’extrapoler leurs valeurs dans un environnement physiologique. Or, les thrombophlébites se développent systématiquement en des loci particuliers du système cardiovasculaire. D’autre part, plusieurs études cliniques ont établi que des conditions hémorhéologiques perturbées constituent des facteurs de risque de thrombose veineuse mais leurs contributions étiologiques demeurent hypothétiques ou corrélatives. En conséquence, un outil de caractérisation hémorhéologique applicable in vivo et in situ devrait permettre de mieux cerner et comprendre ces implications. Les ultrasons, qui se propagent dans les tissus biologiques, sont sensibles à l’agrégation érythrocytaire. De nature non invasive, l’imagerie ultrasonore permet de caractériser in vivo et in situ la microstructure sanguine dans des conditions d’écoulements physiologiques. Les signaux ultrasonores rétrodiffusés portent une information sur la microstructure sanguine reflétant directement les perturbations hémorhéologiques locales. Une cartographie in vivo de l’agrégation érythrocytaire, unique aux ultrasons, devrait permettre d’investiguer les implications étiologiques de l’hémorhéologie dans la maladie thrombotique vasculaire. Cette thèse complète une série de travaux effectués au Laboratoire de Biorhéologie et d’Ultrasonographie Médicale (LBUM) du centre de recherche du Centre hospitalier de l’Université de Montréal portant sur la rétrodiffusion ultrasonore érythrocytaire et menant à une application in vivo de la méthode. Elle se situe à la suite de travaux de modélisation qui ont mis en évidence la pertinence d’un modèle particulaire tenant compte de la densité des globules rouges, de la section de rétrodiffusion unitaire d’un globule et du facteur de structure. Ce modèle permet d’établir le lien entre la microstructure sanguine et le spectre fréquentiel du coefficient de rétrodiffusion ultrasonore. Une approximation au second ordre en fréquence du facteur de structure est proposée dans ces travaux pour décrire la microstructure sanguine. Cette approche est tout d’abord présentée et validée dans un champ d’écoulement cisaillé homogène. Une extension de la méthode en 2D permet ensuite la cartographie des propriétés structurelles sanguines en écoulement tubulaire par des images paramétriques qui mettent en évidence le caractère temporel de l’agrégation et la sensibilité ultrasonore à ces phénomènes. Une extrapolation menant à une relation entre la taille des agrégats érythrocytaires et la viscosité sanguine permet l’établissement de cartes de viscosité locales. Enfin, il est démontré, à l’aide d’un modèle animal, qu’une augmentation subite de l’agrégation érythrocytaire provoque la formation d’un thrombus veineux. Le niveau d’agrégation, la présence du thrombus et les variations du débit ont été caractérisés, dans cette étude, par imagerie ultrasonore. Nos résultats suggèrent que des paramètres hémorhéologiques, préférablement mesurés in vivo et in situ, devraient faire partie du profil de risque thrombotique. / The aggregation of erythrocytes is the main determinant of blood non Newtonian behaviour under low shearing flow conditions. When red blood cells (RBCs) aggregate, they form « rouleaux » and complex tridimensional structures that increase blood viscosity from a few mPa.s to a hundred mPa.s. The reversible RBC aggregation phenomenon is attributed to weak adhesive links between erythrocytes that are readily broken by increasing flow shearing. Blood bulk rheological properties have been comprehensively studied. However, the in vivo physiological impacts of abnormal clustering of RBCs are more difficult to assess. Clinical studies have identified altered hemorheology as a risk factor for thrombosis, but a clear etiological relationship between abnormal aggregation and thrombosis has not yet been established, in part because clinical conclusions were derived from correlative findings. It is to note that cardiovascular diseases such as deep venous thrombosis generally occur at specific locations within the vascular bed, suggesting a hemodynamic contribution to the development of this disease. Consequently, it is postulated that in vivo hemorheological characterization may help shed some light on the role of RBC hyper-aggregation on cardiovascular disorders. Ultrasound imaging, a non-invasive method relying on the propagation of mechanical waves within biological tissues, is sensitive to RBC aggregation. Indeed, the study of backscattered waves allows characterizing blood microstructure in vivo and in situ under physiological flow conditions. The work described in this thesis is based on prior simulation studies, performed at the Laboratory of Biorheology and Medical Ultrasonics of the University of Montreal Hospital Research Center, in which the backscattering of ultrasound from aggregating RBCs was modeled by considering a particle scattering strategy. In this approach, each RBC is a weak ultrasound scatterer (Born assumption) and the backscattering coefficient is modeled as the product of the RBC number density, the RBC backscattering cross section and a structure factor. This model relates variations in the backscattering coefficient to the RBC spatial organisation through the structure factor, which is the only parameter that changes during the aggregation process. A second order expansion in frequency of the structure factor was used to describe blood microstructure in terms of a packing factor W and an ensemble averaged aggregate diameter D. The model was first presented and validated by considering a homogenous shear flow condition using three broadband mono-element transducers. It was then extended in 2D to allow computation of parametric images in tube flow. An extrapolation based on the assumption that viscosity is related to the level of aggregation was used to compute local viscosity maps. Finally, a last contribution was the demonstration that a sudden increase in aggregation tendency directly promoted the formation of venous thrombosis in an experimental animal model. In that study, RBC aggregation, thrombus formation and flow variations were monitored longitudinally for two weeks using ultrasound. The results reported in this thesis suggest that rheological parameters on RBC clustering, ideally assessed in vivo and in situ, should be included in thrombosis risk profiling.

Le présent travail de recherche concerne l'étude de la convection de Rayleigh-Bénard-Marangoni dans un récipient cylindrique doté d'un fond en substrat solide. Le substrat solide est chauffé sur sa face inférieure par un flux de chaleur localisé. L'étude comporte deux parties : La première partie du travail consiste en une modélisation physique du problème associée à des simulations numériques. Les équations de Navier-Stokes et de l'énergie sont résolues en 3D par une méthode de volumes finis. Un transfert de chaleur conjugué solide-liquide est considéré. Des morphologies originales de cellules (type et nombre) sont observées, elles dépendent des conditions géométriques, des nombres adimensionnels qui régissent la physique de l'écoulement (nombre de Prandtl, de Rayleigh et de Marangoni ainsi que du rapport des conductivités thermiques du substrat solide et du fluide). Les transferts de chaleur sont aussi évalués pour chaque cas d'étude. Dans la deuxième partie, nous allons détaillons une étude expérimentale de la convection de Rayleigh-Bénard-Marangoni dans la même configuration que celle étudiée numériquement. Les structures convectives et leurs évolutions sont étudiées à partir d’images relevées par thermographie infra-rouge. Différents modes d'organisation des cellules convectives ont pu être mis en évidence pour ce type de chauffage à flux thermique imposé non uniforme. / The present research work concerns the study of Rayleigh-Bénard-Marangoni convection in a cylindrical container with a solid substrate base. This solid substrate is heated by a localized heat flux on its underside. The study is divided into two parts : The first part of the work consists of a physical modelling of the problem associated with numerical simulations. The Navier-Stokes and energy equations are solved by using a 3D finite volume method. A conjugate solid-liquid heat transfer is considered. Original morphology of cells (type and number) are observed, they are linked to the geometrical conditions, the dimensionless numbers which govern the physical problem (Prandtl, Rayleigh and Marangoni numbers and the ratio of solid substrate to liquid thermal conductivities). The heat transfers are also evaluated in each case. In the second part of the work, we present an experimental study of Rayleigh-Bénard-Marangoni convection in the same configuration as that studied numerically. Convective structures and their evolutions are studied from images recorded by infrared thermography. Different modes of organization of convective cells have been highlighted for this type of heating with imposed non-uniform heat flux.

[pt] O presente trabalho apresenta um estudo numérico da transferência de energia por radiação e sua interação com as propriedades radiantes cinzas e espectrais dos gases produtos da combustão e da fuligem que são formados em um processo de combustão turbulenta. Assim, utilizam-se sistemas de forno/- queimador que operam em regime de chama não pré-misturada de maneira a avaliar, através da dinâmica dos fluidos computacional (CFD), a influência que exercem os diversos modelos de propriedades radiantes sobre a representação da termoquímica do escoamento reativo. Com o objetivo de identificar as principais características e deficiências que apresentam cada um destes modelos, foram considerados dois cenários. O primeiro, correspondente a um problema de radiação unidimensional de um sistema homogêneo e não isotérmico onde são estudados, modelos de propriedades radiantes (i) disponíveis em um software comercial de CFD e (ii) aqueles que foram implementados neste trabalho. Além disso, foi empregado um código numérico que determina as propriedades radiantes espectrais de gases produtos da combustão e da fuligem através de uma abordagem de banda estreita. Para este fim, este código foi acoplado com o software de CFD. Em seguida, dois queimadores de porte laboratorial são empregados de forma a avaliar a capacidade preditiva dos modelos de propriedades radiantes: o primeiro queima propano gasoso e ar enriquecido com oxigênio e o segundo utiliza querosene líquido e oxigênio como reagentes. Dados experimentais de fluxo de calor radiante e de fração volumétrica da fuligem são utilizados para comparação com os resultados obtidos da simulação. Para ambas as configurações de queimador foi também estudado o modelo de Moss-Brookes para previsão da formação/consumo da fuligem. Os resultados obtidos demonstraram o bom desempenho da maioria dos modelos de propriedades radiantes estudados. Em particular, a abordagem de banda estreita foi o que melhor previu a radiação térmica. Além disso, a sua utilização com o modelo de Moss-Brookes levou à melhor previsão da fração volumétrica da fuligem. / [en] This work presents a numerical study of radiation heat transfer and its interaction with gray and spectral radiation of combustion products and soot that are formed in a turbulent combustion process. Different burner/furnace systems operating in a non-premixed combustion regime were used in order to evaluate, through computational fluid dynamics (CFD), the influence of several radiant properties models. Aiming to identify the key features and shortcomings that exhibit each of these models, two scenarios were considered. The first corresponds to a 1-D radiation problem where radiative properties models of a homogeneous non isothermal system are studied as (i) available CFD commercial software and (ii) those implemented in this work. Moreover, a numerical code was used in order to determine, through a narrow band approach, the spectral radiative properties of soot and combustion products. For this purpose, this code was coupled with the CFD software. Then, two laboratory-scale burners are used to assess the predictive capacity of radiative properties models: the first, burning propane and enriched air oxygen, and the second uses kerosene and oxygen as reactants. Measurements of radiant heat flux and soot volumetric fraction are used for comparison with simulation results. For both configurations, the performance of the Moss-Brookes model for predicting the soot production was also studied. The results of this study demonstrated the good performance of the majority of the radiant properties models studied. Particularly, the narrow band approach was the model that provided the best thermal radiation prediction. Moreover, the combination of the narrow band approach with the Moss-Brookes model lead to the best prediction of soot volume fraction.

Compostos orgânicos constituem uma classe interessante de materiais para aplicações em óptica por apresentarem boa processabilidade, relativa facilidade para integração em dispositivos e, principalmente, pela possibilidade de otimização de suas propriedades ópticas através da engenharia molecular. Várias estratégias têm sido empregadas para sintetizar moléculas orgânicas, que exibam singificativos efeitos ópticos não lineares. Nesta direção, nos últimos anos moléculas multi-ramificadas vêm sendo produzidas com o objetivo de intensificar efeitos não lineares, já que estas podem exibir um forte efeito cooperativo entre seus ramos. Nesta tese estudamos a relação da absorção de dois fótons (A2F) e da primeira hiperpolarizabilidade com a estrutura molecular, para um conjunto de nove derivados de trifenilamina com diferentes grupos aceitadores de elétrons arranjadas em geometrias dipolar, quadrupolar e octopolar. O processo A2F foi estudados através da técnica de Varredura-Z, enquanto que a primeira hiperpolarizabilidade foi caracterizada pela técnica de espelhamento hiper Rayleigh. Os dados experimentais para a absorção de dois fótons revelaram espectros bem definidos, com valores razoáveis de seção de choque na região do visível e infravermelho próximo. Observamos ainda um engrandecimento para a seção de choque de A2F para as moléculas quadrupolares. Os resultados obtidos para a primeira hiperpolarizabilidade (&beta;) mostraram que moléculas quadrupolares apresentam maior &beta; do que as dipolares e octopolares, portanto, nossos resultados permitem concluir que o acoplamento eletrônico entre os ramos contribuem fortemente para a seção de choque por A2F e &beta; nas moléculas quadrupolares, não sendo este processo relevante nas moléculas octopolares. Todos os resultados foram interpretados por meio de estudos teóricos empregando a teoria do funcional da densidade (DFT). / Organic compounds constitute an interesting class of materials for optical applications due to their excellent processability, easy integration into devices and, mainly, the possibility of optimizing its optical properties through molecular engineering. Several strategies have been employed to synthesize organic molecules, which exhibit significant nonlinear optical effects. In this direction, in the last few years multi-branched molecules have been obtained aiming at intensifying nonlinear optical effects, since they may exhibit a strong cooperative effect among their branches. On this thesis we have studied the relationship of two-photon absorption (2PA) and first hyperpolarizability with the molecular structure of a group of nine triphenylamine derivatives attached to distinct electron acceptor groups arranged in dipole, quadrupole and octopolar geometries. The 2PA process was studied by Z-scan technique, while the first hyperpolarizability was characterized by the hyper-Rayleigh scattering technique. The experimental data for two-photon absorption revealed well-defined spectra with reasonable cross section magnitude in the visible and near infrared range. We also observed an enhancement of the 2PA cross-section for the quadrupolar molecules in comparison to the dipolar and octopolar ones. The results obtained for the first hyperpolarizability (&beta;) shown that the quadrupolar molecules present higher &beta; than the dipolar and octopolar, suggesting that the electronic coupling between the branches strongly contribute to the 2PA cross-section and &beta; in quadrupolar molecules, being not relevant in the octopolar molecules. All results were interpreted through theoretical studies based on the density functional theory (DFT).

Cette thèse s’inscrit dans le contexte des contrôles non destructifs du béton par ultrasons. Nous focalisons notre étude sur les mesures en acoustique non linéaire qui permettent d’ausculter le béton à l’échelle mésoscopique où les ondes interagissent avec les microfissures et le réseau de porosité. Les paramètres associés sont connus comme étant beaucoup plus sensibles que ceux issus des mesures linéaires. Le béton est un matériau hétérogène et complexe, ce qui présente un comportement fortement non linéaire croissant avec l’état endommagé.Nous développons dans cette thèse un type de mesure non linéaire : Dynamic Acousto-Elastic Testing (DAET). Cette technique fondée sur le principe d’une excitation dynamique du matériau, utilise les ondes ultrasonores pour suivre la variation du comportement élastique en fonction de l’amplitude d’excitation. Nous focalisons notre étude sur le problème du béton d’enrobage qui tient un rôle essentiel dans la durée de vie d’une structure de génie civil. Nous étudions l’interaction des mesures non linéaires par DAET avec les ondes de Rayleigh qui se propagent dans le béton d’enrobage. Nous montrons la sensibilité importante de l’évolution de paramètres non linéaires en fonction de l’endommagement thermique et de la carbonatation.Par la suite, nous proposons une nouvelle méthodologie de la mesure DAET, dans laquelle la vibration transitoire est générée par un impact et les ondes sont générées en continue. Nous présentons différentes applications de la méthode proposée, incluant le cas des éprouvettes de grandes dimensions. Cette approche ouvre de larges possibilités de transposer les mesures pour une application sur site. / This thesis is related to the field of nondestructive evaluation of concrete by ultrasound. We focus our study on nonlinear acoustic-based measurements that allow the concrete auscultation at mesoscopic scale where waves interact with microcracks and porosity network. The nonlinear parameters are known to be much more sensitive than those from linear measurements. Concrete is a heterogeneous and complex material. Its behavior is highly nonlinear with increasing damaged state.We develop in this thesis a type of nonlinear measurement: Dynamic Acousto-Elastic Testing (DAET). This technique is based on the principle of a dynamic excitation of the material, using ultrasounds to follow the variation of the elastic behavior depending on the excitation amplitude. We focus our study on the problem of concrete cover which holds a key role in the life of a civil engineering structure. We study the interaction of the DAET measurement with the Rayleigh waves which propagate in the concrete cover. We show the high sensitivity evolution of non-linear parameters in function of thermal damage and carbonation.Subsequently, we propose a new methodology of DAET measurement, in which the transient vibration is generated by an impact and ultrasounds are generated continuously. We present different applications of the proposed method including the case of large specimens. This approach opens broad possibilities of transposing measurements for on-site application.

Le thème central de cette thèse est l'étude des résonances pour le problème intérieur en élastodynamique (géométries elliptique et sphéroïdale), et pour le problème de diffusion en acoustique (géométrie elliptique). On s'intéresse en particulier à la levée de dégénérescence des résonances liée à la brisure de symétrie de l'objet lors de la transition du disque circulaire vers le disque elliptique (2D), et de la sphère vers le sphéroïde (3D). Ce phénomène est étudié et interprété d'un point de vue théorique en prenant en compte les symétries de l'objet à l'aide de la théorie des groupes. Cette approche est complétée par une modélisation numérique et une partie expérimentale. En 2D, nous étudions le problème intérieur pour un disque elliptique élastique (étude des modes résonants) et le problème de la diffusion acoustique par des cylindres elliptiques élastiques. Ils sont traités à partir du formalisme modal combiné à la théorie des groupes dans le contexte vectoriel de l'élastodynamique. La levée de dégénérescence est observée théoriquement mais aussi expérimentalement en diffusion. La méthode simplifie considérablement le traitement numérique des problèmes étudiés, fournit une classification des résonances selon les 4 représentations irréductibles du groupe de symétrie C2v (associé à la géométrie elliptique) et donne une interprétation physique de la levée de dégénérescence en termes de brisure de symétrie. Une partie expérimentale en spectroscopie ultrasonore vient compléter l'étude théorique du problème de diffusion. Une série d'expériences en cuve est menée dans le cas de cylindres elliptiques de différentes excentricités en aluminium immergés dans l'eau, dans la bande de fréquence 0 ≤ kr ≤ 50, où kr est le nombre d'onde réduit dans le fluide. Les résultats expérimentaux présentent un très bon accord avec les résultats théoriques, la levée de dégénérescence est observée expérimentalement sur des fonctions de forme et mise en évidence sur des diagrammes angulaires. Le problème intérieur en 3D est traité expérimentalement à partir de la génération et la détection optiques d'ondes élastiques. Une série d'expérimentations sur des objets tridimensionnels (sphère, sphéroïdes oblates et prolates de différentes excentricités) en aluminium est réalisée. Ils sont mis en vibration par impacts laser et les mesures de vitesse et de fréquence s'effectuent par vibrométrie laser. On réalise ainsi une comparaison qualitative entre la théorie 2D et l'expérience 3D. Les mesures sont menées à la fois dans les domaines temporel et fréquentiel pour mettre en évidence la levée de dégénérescence d'une part, et l'onde de Rayleigh qui se propage sur la surface des objets d'autre part. Nous identifions deux trajets pour cette onde en géométrie sphéroïdale, l'un circulaire et l'autre elliptique.Enfin, dans le cadre des problèmes intérieurs 2D et 3D, on donne une interprétation en termes de rayons à travers la dualité entre le spectre des résonances et le spectre des longueurs des orbites périodiques (OPs), avec la mise en évidence du phénomène de conversion de mode et l'identification de l'onde de Rayleigh. Un phénomène, nouveau à notre connaissance, vient s'ajouter au phénomène de bifurcation de certaines orbites. Au cours de la déformation vers le disque elliptique, les orbites avec conversion de mode du disque circulaire se séparent en deux orbites dont les longueurs sont associées aux trajets minimal et maximal qu'elles parcourent. Cette observation s'interprète comme une conséquence du théorème de Fermat. Dans le cas du sphéroïde, on retrouve les orbites du disque circulaire dans le plan équatorial et celles du disque elliptique dans le plan méridien. Nous mettons également en évidence les pics associés aux deux trajets parcourus par l'onde de Rayleigh sur le spectre des OPs. / Resonances for the interior problem in elastodynamics and the acoustic scattering problem are considered in elliptical and spheroidal geometries. Interest is focused on the splitting up of resonances which occurs when the symmetry is broken in the transition from the circular disc to the elliptical one (2D), and from the sphere to the spheroid (3D). From the theoretical point of view, this physical phenomenon is studied and interpreted taking into account the symmetries of the object with the help of group theory. This approach is completed by a numerical modeling and an experimental part. As far as the two dimensional problems are concerned, the interior problem for an elliptical elastic disc (study of resonant modes) and the acoustic scattering problem for infinite elliptical elastic cylinders are studied combining modal formalism and group theory in the vectorial context of elastodynamics. The splitting up of resonances due to the circular symmetry breaking is not only theoretically observed but also experimentally for the scattering problem. The method significantly simplifies the numerical treatment of the problems studied, provides a full classification of resonances over the 4 irreducible representations of the symmetry group C2v (associated with the elliptical geometry) and gives a physical interpretation of the splitting up in terms of symmetry breaking of the symmetry group O(2) (invariance under rotation). An experimental part based on ultrasonic spectroscopy complements the theoretical study of the scattering problem. A series of tank experiments is carried out in the case of aluminum elliptical cylinders immersed in water, in the frequency range 0 ≤ kr ≤ 50, where kr is the reduced wave number in the fluid. The experimental results provide a very good agreement with the theoretical ones, the splitting up is observed on experimental form functions and the split resonant modes are identified on angular diagrams. The interior problem in 3D is studied by means of an experimental approach based on the optical generation and detection of elastic waves. A series of experiments is performed on three-dimensional objects in aluminium. These objects (sphere, prolate and oblate spheroids of various eccentricity) are excited by laser impacts, and the velocity and frequency measurements are carried out by laser vibrometry. Theory and experiments are qualitatively compared. The measurements are performed in both the frequency and time domains to highlight the splitting up of resonances on one hand, and the Rayleigh wave propagating on the surface of the objects on the other hand. We emphasize two paths for this surface wave in spheroidal geometry: a circular one in the equatorial plane and an elliptical one in the meridian plane. Finally, in the context of the interior problems in 2D and 3D, a physical interpretation of resonances in terms of geometrical paths is provided. Mode conversion is highlighted and the Rayleigh wave is identified on the periodic orbits lengths spectrum.In addition to the bifurcations of some periodic orbits, a phenomenon, new to our knowledge, appears. The orbits with mode conversion of the circular disc split in two orbits when the transition to the elliptic disc occurs. The lengths of these orbits are associated with the minimal and maximal travel paths. This observation is interpreted from Fermat's theorem.For the spheroid, orbits of the circular disc and those of the elliptical disc are recovered in the equatorial and meridian planes respectively. We also emphasize the peaks associated with the travel paths of Rayleigh wave in spheroidal geometry appearing on the periodic orbits spectrum.