Two structural models which can be used to predict the buckling, post buckling and vibration behavior of flat and curved composite panels under thermomechanical loadings are developed in this work. Both models are based on higher-order transverse shear deformation theories of shallow shells that include the effects of geometric nonlinearities and initial geometric imperfections. Within the first model (Model I), the kinematic continuity at the contact surfaces between the contiguous layers and the free shear traction condition on the outer bounding surfaces are satisfied, whereas in the second model (Model II), in addition to these conditions, the static interlaminae continuity requirement is also fulfilled. Based on the two models, results which cover a variety of problems concerning the postbuckling behaviors of flat and curved composite panels are obtained and displayed. These problems include: i) buckling and postbuckling behavior of flat and curved laminated structures subjected to mechanical and thermal loadings; ii)frequency-load/temperature interaction in laminated structures in both pre-buckling and post buckling range; iii) the influence of a linear/nonlinear elastic foundation on static and dynamic post buckling behavior of flat/curved laminated structures exposed to mechanical and temperature fields; iv) implication of edge constraints upon the temperature/load carrying capacity and frequencyload/ temperature interaction of flat/curved structures; v) elaboration of a number of methodologies enabling one to attenuate the intensity of the snap-through buckling and even to suppress it as well as of appropriate ways enabling one to enhance the load/temperature carrying capacity of structures. / Ph. D.

Hyperlordosis or anterior pelvic tilt is a common non-neutral spinal posture associated with weak core stability, low back pain and altered lumbopelvic muscle activation patterns. Yet the effects of altered lumbopelvic posture and core stability on muscle activation patterns have not been evaluated during a functional movement. The main purpose of this study was to determine the relationship between pelvic tilt, core stability and muscle activation patterns during non-weighted squats in U/19 university-level rugby union players. A total of 49 rugby union players participated in this study. Pelvic tilt (dominant side) was measured from a digital photo with clear reflector markers on the anterior superior iliac spine (ASIS) and posterior superior iliac spine (PSIS) using the Kinovea video analysis software programme (version 0.8.15). Flexibility of the hamstrings, hip flexors and knee extensors was assessed with goniometry. Core stability was assessed using the pressure biofeedback unit and muscle onset times during the ascent phase of non-weighted squats. The onset times of the transverse abdominis (TrA), erector spinae (ES), gluteus maximus (GM) and biceps femoris (BF) muscles were measured using electromyography (EMG). Players were then grouped according to pelvic tilt (anterior and neutral) and by playing position (forwards and backs). The between group differences were evaluated for the abovementioned variables using p-value (statistical significance) and d-value (practical significance) measures. Muscle activation patterns and firing order were determined using descriptive statistics. The mean pelvic tilt of all participants (N=49) was an anterior tilt of 15.35°. When grouped by pelvic tilt, the anterior tilt group showed a mean pelvic tilt of 17.83° (n=27) and the neutral pelvic tilt group showed a mean pelvic tilt of 11.75° (n=22). Despite the differences in pelvic tilt, there was no significant difference in flexibility between the groups. Another controversial result is that the anterior tilt group showed practical significantly better core stability (d=0.54) than the neutral tilt group (46.93° vs 56.3°). During the double leg squat the muscle activation patterns were consistent between the groups. TrA activated first, followed by ES. Thereafter, the BF muscle activated, followed by the GM. The first place activation of TrA is consistent with the literature stating that the deep abdominal stabilisers of individuals with good core stability activate before the movement is initiated. The early onset of muscle activity of ES points to a focus on back extension during the ascent of the squat. Because the pelvic tilt was measured during static standing only, it is unclear whether the players in the neutral tilt group were able to hold the neutral pelvic tilt posture throughout the movement. Research has shown that there is an increased focus on trunk extension during the ascent phase of the squat which is not present during the descent. Future research should focus on assessing the pelvic tilt at the beginning of the ascent phase of the squat to ensure accurate results. The delay in GM activation during the ascent of the squat is concerning. GM acts as a lumbopelvic stabilizer, and its slow activation points to a decrease in lumbopelvic stability. This is very important in weight training, because weight training increases the strain on the lumbar spinal structures, which decreases performance and increases the risk of injury. / MSc (Biokinetics), North-West University, Potchefstroom Campus, 2014

Hyperlordosis or anterior pelvic tilt is a common non-neutral spinal posture associated with weak core stability, low back pain and altered lumbopelvic muscle activation patterns. Yet the effects of altered lumbopelvic posture and core stability on muscle activation patterns have not been evaluated during a functional movement. The main purpose of this study was to determine the relationship between pelvic tilt, core stability and muscle activation patterns during non-weighted squats in U/19 university-level rugby union players. A total of 49 rugby union players participated in this study. Pelvic tilt (dominant side) was measured from a digital photo with clear reflector markers on the anterior superior iliac spine (ASIS) and posterior superior iliac spine (PSIS) using the Kinovea video analysis software programme (version 0.8.15). Flexibility of the hamstrings, hip flexors and knee extensors was assessed with goniometry. Core stability was assessed using the pressure biofeedback unit and muscle onset times during the ascent phase of non-weighted squats. The onset times of the transverse abdominis (TrA), erector spinae (ES), gluteus maximus (GM) and biceps femoris (BF) muscles were measured using electromyography (EMG). Players were then grouped according to pelvic tilt (anterior and neutral) and by playing position (forwards and backs). The between group differences were evaluated for the abovementioned variables using p-value (statistical significance) and d-value (practical significance) measures. Muscle activation patterns and firing order were determined using descriptive statistics. The mean pelvic tilt of all participants (N=49) was an anterior tilt of 15.35°. When grouped by pelvic tilt, the anterior tilt group showed a mean pelvic tilt of 17.83° (n=27) and the neutral pelvic tilt group showed a mean pelvic tilt of 11.75° (n=22). Despite the differences in pelvic tilt, there was no significant difference in flexibility between the groups. Another controversial result is that the anterior tilt group showed practical significantly better core stability (d=0.54) than the neutral tilt group (46.93° vs 56.3°). During the double leg squat the muscle activation patterns were consistent between the groups. TrA activated first, followed by ES. Thereafter, the BF muscle activated, followed by the GM. The first place activation of TrA is consistent with the literature stating that the deep abdominal stabilisers of individuals with good core stability activate before the movement is initiated. The early onset of muscle activity of ES points to a focus on back extension during the ascent of the squat. Because the pelvic tilt was measured during static standing only, it is unclear whether the players in the neutral tilt group were able to hold the neutral pelvic tilt posture throughout the movement. Research has shown that there is an increased focus on trunk extension during the ascent phase of the squat which is not present during the descent. Future research should focus on assessing the pelvic tilt at the beginning of the ascent phase of the squat to ensure accurate results. The delay in GM activation during the ascent of the squat is concerning. GM acts as a lumbopelvic stabilizer, and its slow activation points to a decrease in lumbopelvic stability. This is very important in weight training, because weight training increases the strain on the lumbar spinal structures, which decreases performance and increases the risk of injury. / MSc (Biokinetics), North-West University, Potchefstroom Campus, 2014

Résumé : Une coordination appropriée entre les pattes antérieures et postérieures chez les mammifères terrestres est essentielle pour maintenir une stabilité pendant la locomotion quadrupède. Il a été fortement suggéré que les voies propriospinales et le retour sensoriel pourraient jouer un rôle important dans la coordination antéropostérieure. Cependant, les mécanismes neurophysiologiques impliqués dans la coordination entre les membres antérieurs et les membres postérieurs pendant la marche demeurent méconnus. Suite à une lésion partielle de la moelle épinière, plusieurs déficits de la coordination antéropostérieure ont pu être dénotés chez l’animal non-humain et chez l’humain. Malgré tout, les effets d’une lésion partielle sur la coordination antéropostérieure n’ont pas encore été clairement caractérisés. Au regard du nombre considérable de blessés médullaires présentant des troubles de l’équilibre et de la coordination entre les bras et les jambes après avoir récupéré la marche, il semble essentiel de mieux comprendre ces mécanismes. L’objectif du projet était d’offrir une meilleure caractérisation de la coordination antéropostérieure chez le chat intact et suite à une hémisection latérale de la moelle épinière. Pour ce faire, un tapis roulant partitionné transverse permettant de dissocier les vitesses de locomotion entre les pattes antérieures et postérieures a été utilisé. Huit chats implantés chroniquement pour réaliser des enregistrements électromyographiques ont été entrainés dans diverses conditions de locomotion partitionnée et non-partitionnée. Parmi ces 8 chats, 6 ont subi une hémisection latérale du côté droit de la moelle épinière entre la 5ème et la 6ème vertèbre thoracique (T6). Des analyses cinématiques et électromyographiques ont été réalisées chez les chats intacts et 8 semaines après la lésion partielle de la moelle épinière. Les résultats obtenus démontrent que la coordination antéropostérieure est contrôlée par des influences bidirectionnelles et asymétriques entre les pattes antérieures et postérieures. De plus, lorsque la vitesse de locomotion des pattes antérieures était plus élevée que celle des pattes postérieures, une dissociation du rythme était observée avec les pattes antérieures réalisant davantage de pas chez le chat intact. Suite à la lésion, cette dissociation était également observée lors de la locomotion non-partitionnée. Cependant, que ce soit avant ou après la lésion, quand une telle dissociation apparaissait, une nouvelle forme stable de coordination antéropostérieure apparaissait consistant à réaliser deux pas des pattes antérieures pendant un pas des pattes postérieures (coordination 2-1). Finalement, la coordination antéropostérieure pouvait être modulée par l’utilisation du tapis roulant partitionné et une coordination 1-1 pouvait être restaurée suite à la lésion en faisant marcher les pattes postérieures à une vitesse plus élevée. À partir de ces résultats, nous avons proposé un nouveau modèle théorique du contrôle neurophysiologique de la coordination antéropostérieure. Qui plus est, un raffinement des échelles d’évaluation de la récupération locomotrice suite à des lésions de la moelle épinière s’avère indispensable afin d’inclure une caractérisation détaillée de la coordination antéropostérieure. Finalement, d’un point de vue clinique, ces résultats suggèrent que de nouvelles stratégies thérapeutiques basées sur la coordination antéropostérieure pourraient être envisagées afin de renforcer la récupération locomotrice suite à des lésions de la moelle épinière. / Abstract : An appropriate coordination between the forelimbs and the hindlimbs in terrestrial mammals is essential to maintain stability during quadrupedal locomotion. It is thought that propriospinal pathways and sensory feedback contribute to the control of forelimbhindlimb coordination. However, the neurophysiological mechanisms involved in this coordination during locomotion remain poorly defined. After a partial spinal cord injury, several impairments of interlimb coordination have been observed in non-human animal models and human patients. Despite this, the effects of a partial lesion on forelimb-hindlimb coordination have not been clearly characterised. Patients with spinal cord injury have pronounced deficits with their equilibrium and a deficient control of interlimb coordination could be a main contributor. The purpose of these studies was to better characterize forelimb-hindlimb coordination in intact cats and following a lateral hemisection of the spinal cord. A transverse split-belt treadmill was used to independently control the speed of the forelimbs and of the hindlimbs. Eight cats were chronically implanted for e;ectromyography and trained to perform various tied-belt and transverse split-belt locomotor conditions. Among these 8 cats, 6 were hemisected at the 6th thoracic segment of the spinal cord on the right side. Electromyographic and kinematic analyses were performed in the intact state and 8 weeks post-hemisection. The results demonstrate that interlimb coordination is controlled by bidirectional and asymmetrical influences between the forelimbs and the hindlimbs. Moreover, when the forelimbs stepped faster than the hindlimbs, dissociation of the forelimb and hindlimb rhythms occurred, with the forelimbs taking more steps. After the lesion, this dissociation was observed, even during tied-belt locomotion. However, in both intact and injured cats, when such dissociation occurred, a new stable form of forelimb-hindlimb coordination appeared, consisting of two forelimb steps for one hindlimb step (2-1 forelimb-hindlimb coordination). Finally, the transverse split-belt treadmill could modulate forelimb-hindlimb coordination and 1-1 coordination could be restored after the lesion during transverse split-belt locomotion with a faster hindlimb speed. From these results, we propose a theoretical model of the neurophysiological control of interlimb coordination. Moreover, a refinement of performance scales evaluating the locomotor recovery after spinal cord injury is necessary to include a detailed characterisation of interlimb coordination. Finally, from a clinical perspective, these results suggest that new therapeutic strategies based on interlimb coordination could be used to strengthen locomotor recovery after spinal cord injuries.

La connaissance précise des champs électromagnétiques rayonnés par les dispositifs hyperfréquences nécessite des outils instrumentaux permettant la mesure directe ou indirecte de ces champs. La technique du scan champ proche fait partie de ces outils. Ce manuscrit décrit les travaux de caractérisation des sondes électromagnétiques pour une plate-forme de scan champ proche développée au laboratoire LAPLACE. L’accent a été mis sur la simulation électromagnétique des dispositifs de test utilisés dans le travail de déconvolution du facteur d’antenne des sondes de champs magnétiques ou électriques. Ces dispositifs comprennent aussi bien des structures planaires telles que des interconnexions en ligne microruban que des composants en guide d’ondes tels que des guides ouverts de formes rectangulaire ou circulaire. Pour analyser ces structures des logiciels commerciaux basés sur la méthode des différences finis ont été utilisés. Dans le cas des structures rayonnantes un programme basé sur la méthode de l’opérateur transverse a été développé, permettant la détermination de l’admittance de rayonnement et les champs rayonnés en zones proche et lointaine. Les résultats obtenus ont été validés par des simulations avec des outils commerciaux, et par des mesures réalisés au laboratoire. / The accurate knowledge of electromagnetic fields radiated by microwave devices requires instrumental tools for direct or indirect measurement of these fields. Near-field scan technique is one of those tools. This manuscript describes the work done to characterize electromagnetic probes using a near field scan platform developed in the laboratory LAPLACE. We focused our work on the electromagnetic simulation of test devices that are used in the deconvolution of antenna factor of magnetic or electric probes. These devices include both planar structures such as microstrip line and also waveguide components such as rectangular or circular open-ended waveguides. To analyze these structures, commercial software based on finite element method was used. In case of radiating structures, a program based on transverse operator method was developed. It allows the determination of the admittance of radiation and the radiated electromagnetic fields in near-field and far-field regions. The results were validated by simulations with commercial tools, and measurements made in the laboratory.

L'objet de cette thèse est l'étude de la sélection des structures transverses stationnaires dans les lasers. Nous nous attachons à décrire expérimentalement et à expliquer numériquement à l'aide d'une méthode de type Fox--Li les observations de modes transverses qui peuvent être réalisées dans un laser à état solide pompé longitudinalement. Ainsi, nous sommes amenés à explorer en profondeur le comportement des cavités non--dégénérées dans différentes situations de pompage, ainsi que les notions de dégénérescence exacte et de dégénérescence partielle. Nous sommes également amenés à étudier, bien que de manière assez qualitative, le rôle important des effets thermiques dans la formation des modes transverses dans les zones de dégénérescence de la cavité. Ces questions ont fait l'objet de multiples investigations depuis l'avènement du laser, mais nous mettons toutefois en évidence que de sérieux problèmes d'interprétations d'observations expérimentales persistent dans la littérature scientifique contemporaine. Une idée fausse assez répandue stipule par exemple qu'il est possible d'exciter un mode Laguerre--Gauss ou Ince--Gauss d'ordre élevé en utilisant un faisceau de pompe suffisamment petit afin qu'il recouvre parfaitement le lobe principal du mode visé. Nous montrons analytiquement, numériquement et vérifions expérimentalement que ceci est impossible dans une situation non--dégénérée, et expliquons comment certaines observations réalisées proches d'une dégénérescence peuvent contribuer à propager cette idée erronée. Bien que le modèle que nous utilisons n'élude pas les questions importantes concernant le rôle de la dynamique temporelle dans la formation des modes transverses dans un laser, sa pertinence réside dans sa simplicité et sa capacité à reproduire de manière très satisfaisante toutes les observations expérimentales. Ainsi, toutes les observations de modes transverses qu'il est possible de réaliser dans un laser peuvent être comprises à travers ce modèle, qui ne nécessite comme ingrédients principaux qu'une cavité, un gain saturé, des effets thermiques et éventuellement des effets de diaphragme. De plus, le modèle simple que nous présentons peut s'étendre sans difficulté à l'étude de cavités instables ou de cavités à miroirs asphériques qui offrent des possibilités nouvelles pour l'excitation de modes transverses originaux. / The main concern of this thesis is the selection of stationary transverse structures in lasers. Throughout the manuscript, we get involved into describing the observations that can be realized in a longitudinally pumped solid--state laser, from both experimental and numerical perspectives using Fox--Li simulations. We explore in detail, on the one hand, the particular behavior of non--degenerate cavities in various pumping situations, and on the other hand the exact and partial degeneracy situations. We also investigate the very influential role played by thermal effects concerning the formation of transverse structures in degeneracy regions, although this study is limited to a more qualitative description. These problems have been of great interest since the early development of laser, however we show in this manuscript that the contemporary scientific litterature still suffers from misinterpretations concerning experimental observations of transverse modes that can occur in lasers. For instance, there still remains the widespread belief that it is possible to excite a high-order Laguerre--Gaussian or Ince--Gaussian mode by strongly focusing the pump beam inside the active medium so that it matches the principal lobe of the targeted mode. We demonstrate analytically, numerically and experimentally that this idea is unfounded and show how some experiments that occur near a cavity degeneracy are misleading and can contribute to spread this misconception. Even if the model we use is unable to deal with some very significant issues concerning the mode formation in the time domain and the build--up of the laser oscillation, we believe that its strength lies into its simplicity and its ability to reproduce every experimental observation in a very accurate way. Hence, we managed to explain and interpret all the experimental observations we have realized given a very few assumptions including the presence of a cavity, a saturated gain, thermal effects and eventually diaphragm effects. Moreover, the simple model that we present can be extended without any difficulty to the case of unstable resonators, or to resonators built from graded--phase mirrors which offer new possibilities towards the excitation of original transverse modes.

Cette thèse concerne les homéomorphismes de surfaces.Soit f un difféomorphisme d'une surface M préservant l'aire et isotope à l'identité. Si f a un point fixe contractile isolé et dégénéré z0 avec un indice de Lefschetz égal à 1, et si l'aire de M est finie, nous prouverons au chapitre 3 que z0 est accumulé non seulement par des points périodiques mais aussi par des orbites périodiques au sens de la mesure. Plus précisément, la mesure de Dirac en z0 est la limite en topologie faible-étoile d'une suite de probabilités invariantes supportées par des orbites périodiques. Notre preuve est totalement topologique et s'applique au cas d'homéomorphismes en considérant l'ensemble de rotation local.Au chapitre 4, nous étudierons des homéomorphismes préservant l’aire et isotope à l’identité. Nous prouverons l’existence d'isotopies maximales particulières: les isotopies maximales à torsion faible. En particulier, lorsque f est un difféomorphisme ayant un nombre fini de points fixes tous non-dégénérés, une isotopie I joignant l'identité à f est à torsion faible si et seulement si pour tout point z fixé le long de I, le nombre de rotation (réel) ρ(I,z), qui est bien défini quand on éclate f en z, est contenu dans (-1,1). Nous démontrerons l'existence d'isotopies maximales à torsion faible, et nous étudierons la dynamique locale de feuilletages transverses à l'isotopie près des singularités isolées.Au chapitre 5, nous énoncerons une généralisation d'un théorème de Poincaré-Birkhoff local au cas où il existe des points fixes au bord. / This thesis concerns homeomorphisms of surfaces.Let f be an area preserving diffeomorphism of an oriented surface M isotopic to the identity. If f has an isolated degenerate contractible fixed point z0 with Lefschetz index one, and if the area of M is finite, we will prove in Chapter 3 that z0 is accumulated not only by periodic points, but also by periodic orbits in the measure sense. More precisely, the Dirac measure at z0 is the limit in weak-star topology of a sequence of invariant probability measures supported on periodic orbits. Our proof is purely topological and will works for homeomorphisms and is related to the notion of local rotation set.In chapter 4, we will define a kind of identity isotopies: torsion-low isotopies. In particular, when f is a diffeomorphism with finitely many fixed points such that every fixed point is not degenerate, an identity isotopy I of f is torsion-low if and only if for every point z fixed along the isotopy, the (real) rotation number ρ(I,z), which is well defined when one blows-up f at z, is contained in (-1,1). We will prove the existence of torsion-low maximal identity isotopies, and we will deduce the local dynamics of the transverse foliations of any torsion-low maximal isotopy near any isolated singularity.In chapter 5, we will generalize a local Poincaré-Birkhoff theorem to the case where there exist fixed points on the boundary

La conception de réservoirs composites sans liner pour les lanceurs spatiaux nécessite d’étudier la relation entre endommagement et perméabilité dans les matériaux composites stratifiés pour proposer des solutions répondant à la fois aux critères fonctionnels de résistance et de taux de fuite. L’objectif de cette étude est de fournir une description pertinente de l’évolution de l’endommagement à l’échelle mésoscopique (i.e. à l’échelle du pli) et des réseaux de fissures en résultant afin de prédire l’apparition de fuites à travers les composites stratifiés. Pour ce faire, la démarche mise en œuvre combine modélisation de l’évolution de l’endommagement et essais mécaniques et de perméabilité. Dans un premier temps, des procédures expérimentales spécifiques s’appuyant sur des observations par microscopie optique et par microtomographie sous chargement de traction sont proposées pour caractériser les interactions et l’agencement entre les endommagements des différents plis en termes de seuil de fissuration, de longueur et de position relative des fissures mésoscopiques. Ensuite, pour caractériser la percolation des fissures qui transforme des réseaux de fissures en chemins de fuite, une méthode de mesure de la perméabilité et deux dispositifs expérimentaux sont proposés pour réaliser des mesures de perméabilité sur des tubes sous chargement mono-axial d’une part et bi-axial proportionnel ou non d’autre part. Pour les deux chapitres expérimentaux, des résultats obtenus pour différents stratifiés carbone époxy mis en œuvre par placement de fibres automatisé sont présentés. Enfin, un modèle d’endommagement multi-échelles basé sur les énergies de fissurations est introduit pour décrire les cinétiques des endommagements mésoscopiques, et des pistes de réflexion pour le développement d’un méso-modèle de prédiction des densités de points de fuite sont données. / The design of liner-less composite pressure vessels for spatial launchers requires studying the relation between damage and permeability in laminates in order to offer solutions which fulfill strength and leak rate requirements. This study aims at providing a relevant description of damage growth and the resultant network at the meso-scale for leakage prediction. To do so, damage growth modelling is coupled with mechanical and permeability characterization. Two mains issues are addressed: crack network formation, by characterizing and modeling the growth and organization of cracks, and percolation and its effects on permeability, which is addressed by permeability measurement and leak path quantification. Several experimental methods based on optical microscopy and micro-tomography observations under tensile loading are proposed. These methods aim at evaluating the damage interaction and pattern of different plies in terms of cracking threshold, cracks length and relative location. Besides, two experimental setups are proposed to assess permeability evolution and percolation of cracks in pipes submitted to mono-axial or bi-axial loading. The results obtained for several carbon epoxy laminates manufactured by Automated Fibre Placement (AFP) are presented. At last, a multi-scale damage model based on energy release rates and finite fracture mechanics is introduced and key issues for developing a damage meso-model dedicated to the prediction of leak path density are established.

The mixing performance of multiple transverse jets has been evaluated experimentally. Measurement techniques included laser Doppler velocimetry and planar laser induced fluorescence. Basic findings are consistent with results presented in literature for single jet mixing behavior. Mixing performance has been compared to literature for the single jet case and the Holdeman parameter has been re-evaluated for effectiveness at low jet numbers. A single jet in a confined crossflow was found to have a local minimum at B(d⁄D) = 0.721. Results for two jets indicate monotonically decreasing unmixedness for the range of conditions tested, with no local optimum apparent. Data for three jets indicate a local optimum at B(d⁄D) = 0.87and relatively flat range of mixing performance in the range of 0.75 < B(d⁄D) < 1.5. Six jets indicate a minimum unmixedness near B(d⁄D) = 0.5, but exhibited poorer mixing performance than all other configurations at the highest values of B(d⁄D)tested. The most optimum configuration tested was six jets at B(d⁄D) = 0.5, resulting in an unmixedness of 0.0192. This value was 76% lower than the next lowest configuration (three jets) at the same B(d⁄D).Total momentum was found to collapse the data well, as configurations more closely matched a historical correlation for second moment of a single confined jet more closely.

No description available.