Not available / text

Thesis (Ph. D.)--University of Texas at Austin, 2004. / Supervisor: D.E. Winget. Vita. Includes bibliographical references. Also available from UMI.

Thesis (Ph. D.)--University of Florida, 1982. / Description based on print version record. Typescript. Vita. Includes bibliographical references (leaves 95-97).

Thesis (Ph. D.)--University of Texas at Austin, 2007. / Vita. Includes bibliographical references.

Les étoiles Miras sont des étoiles froides et évoluées (étoiles AGB), caractérisées par une variation de luminosité régulière et de longue période liée à une pulsation radiale. Cela se traduit par une atmosphère froide, étendue et faiblement liée gravitationnellement. Dans le spectre de ces étoiles, la détection variable d'émissions intenses dans les raies de Balmer de l'hydrogène est usuellement reliée à la propagation périodique d'une onde de choc radiative hypersonique dans l'atmosphère. Mon travail de thèse a eu d'abord pour objectif de confirmer l'existence d'une forte signature de polarisation linéaire accompagnant ces émissions, d'origine inconnue et déjà observée auparavant, et de caractériser son évolution temporelle. Je me base sur un suivi spectropolarimétrique inédit et effectué dans le cadre de ma thèse avec l'instrument NARVAL pour plusieurs Miras afin d'établir cette caractérisation. Ces suivis ont été réalisé sur tous les paramètres de Stokes (polarisation linéaire et circulaire) et surtout à des phases autour du maximum de luminosité.Un résultat majeur de ces observations est la détection systématique de signatures polarisées (polarisation surtout linéaire mais aussi parfois circulaire) liées aux émissions en intensité dans les raies de Balmer. L'utilisation d'un code de simulation d'atmosphère dynamique m'a permis de confirmer le lien entre une onde de choc et la présence d'émission dans ces raies, chose qui n'avait jamais été montrée par un modèle jusque là. Ces éléments me font affirmer que le mécanisme de polarisation est intrinsèque à l'onde de choc. La discussion sur l'origine de cette polarisation comporte deux grands axes : l'asymétrie globale de l'onde de choc amenant à une polarisation nette du rayonnement et la production locale dans le front du choc d'un champ magnétique responsable d'une polarisation par impact dans la zone de production des photons Balmer, c'est-à-dire le sillage du choc. D'autre part, j'invoque la possibilité d'une instabilité de Parker comme deuxième facteur de polarisation par impact et je discute le rôle potentiel de l'effet Hanle. / Mira stars are cool and evolved stars (AGB stars), characterized by a regular luminosity variation and a long period linked to a radial pulsation. All this imply a cool and extended atmosphere which is weakly linked by gravity. In the spectra of those stars, the variable detection of intense emissions in the Balmer lines of hydrogen are usually linked to the periodical propagation of a hypersonic radiative shock wave in the atmosphere. My thesis work's first objective was to confirm the existence of a strong polarimetric signature accompanying those emissions, whose origin is still unknown and already observed before, and to characterize its evolution through time. I rely on a novel spectropolarimetric survey, done in the framework of my thesis with the NARVAL instrument for several Mira stars in order to establish this characterization. Those surveys were done on all Stokes parameters (linear and circular polarization) and mainly during phases around the maximum of luminosity.A major result of these observations is the systematic detection of polarized signatures (mainly linear polarization but also circular sometimes) linked to the emissions in intensity in the Balmer lines. The use of a dynamical atmosphere simulation code allowed me to confirm the link between a shock wave and the presence of emissions in those lines. Such a result has never been been produced by a model until now. These elements make me state that the polarization mechanism is intrinsic to the shock wave. The discussion on the origin of such polarization consists of two main axes : the global asymmetry of the shock wave leading to a net polarization of the radiation and the local production in the shock's front of a magnetic field responsible for an impact polarization in the area of production of Balmer photons, namely the shock's wake. Besides that, I mention the possibility of a Parker instability as a second factor of impact polarization and I discuss the potential role of the Hanle effect.

No description available.

Automotive catalytic converters are used extensively in the automotive industry to reduce toxic pollutants from vehicle exhausts. The flow across automotive exhaust catalysts is distributed by a sudden expansion and has a significant effect on their conversion efficiency. The exhaust gas is pulsating and flow distribution is a function of engine operating condition, namely speed (frequency), load (flow rate) and pressure loss across the monolith. The aims of this study are to provide insight into the development of the pulsating flow field within the diffuser under isothermal conditions and to assess the steady-state computational fluid dynamics (CFD) predictions of flow maldistribution at high Reynolds numbers. Flow measurements were made across an automotive catalyst monolith situated downstream of a planar wide-angled diffuser in the presence of pulsating flow. Cycle-resolved Particle Image Velocimetry (PIV) measurements were made in the diffuser and hot wire anemometry (HWA) downstream of the monoliths. The ratio of pulse period to residence time within the diffuser (J factor) characterises the flow distribution. During acceleration the flow remained attached to the diffuser walls for some distance before separating near the diffuser inlet later in the cycle. Two cases with J ~ 3.5 resulted in very similar flow fields with the flow able to reattach downstream of the separation bubbles. With J = 6.8 separation occurred earlier with the flow field resembling, at the time of deceleration, the steady flow field. Increasing J from 3.5 to 6.8 resulted in greater flow maldistribution within the monoliths; steady flow producing the highest maldistribution in all cases for the same Re. The oblique entry pressure loss of monoliths were measured using a one-dimensional steady flow rig over a range of approach Reynolds number (200 < Rea < 4090) and angles of incidence (0o < α < 70o). Losses increased with α and Re at low mass flow rates but were independent of Re at high flow rates being 20% higher than the transverse dynamic pressure. The flow distribution across axisymmetric ceramic 400 cpsi and perforated 600 cpsi monoliths were modelled using CFD and the porous medium approach. This requires knowledge of the axial and transverse monolith resistances; the latter being only applicable to the radially open structure. The axial resistances were measured by presenting uniform flow to the front face of the monolith. The transverse resistances were deduced by best matching CFD predictions to measurements of the radial flow profiles obtained downstream of the monolith when presented with non-uniform flow at its front face. CFD predictions of the flow maldistibution were performed by adding the oblique entry pressure loss to the axial resistance to simulate the monolith losses. The critical angle approach was used to improve the predictions, i.e. the oblique entry loss was limited such that the losses were assumed constant above a fixed critical angle, αc. The result showed that the perforated 600 cpsi monolith requires the entrance effect to be restricted above αc = 81o, while the losses were assumed constant above αc = 85o for the ceramic 400 cpsi monolith. This might be due to the separation bubble at the monolith entrance being restricted by the smaller hydraulic diameter of the perforated monolith thus limiting the oblique entry loss at the lower incidence angle.

In this work, the effect of rotation on the oscillation frequencies of some radially and nonradially oscillating Delta-Scuti stars have been explored. Rotation has been considered as a perturbation and treated up to the second order. Series of evolutionary models have been calculated for the oscillating stars in question and compared with the observational parameters. Three stars are considered: V350 Peg with no rotation, CC And with a rotational velocity Vsini=20 km/s, and BS Tuc with Vsini=130 km/s. We find that splitting in the oscillation frequencies are conspicuous especially in fast rotating stars, with a considerable contribution from the related terms due to second order effect.

We search for planets and brown dwarves around white dwarves (WDs). Finding extra-solar planets is the first step toward establishing the existence and abundance of life in the Universe. The low mass and luminosity of WDs make them ideal stars to search for low mass companion objects. Theoretical predictions generally agree that a star will consume and destroy close-in, low mass planets as it ascends the red giant and asymptotic giant branch evolutionary tracks, but larger mass objects and those further out will survive. The matter ejected from the star as it evolves into a white dwarf may also be accreted onto daughter planets, or may coalesce into a disk from which planets can form. We employ two techniques to search for planets and brown dwarves (BDs) around WDs. A subset of pulsating white dwarf stars have a pulsational stability that rivals pulsars and atomic clocks. When a planet is in orbit around a such a star the orbital motion of the star around the centre of mass is detectable as a change in arrival times of the otherwise stable pulsations. We search for, and find, a sample of suitable pulsators, monitor them for between three and four years, and place limits on companions by constraining the variation consistent with a 2.4M[subscript J] planet in a 4.6 year orbit. We also observe a large sample of WDs to search for a mid-infrared excess caused by the presence of sub-stellar companions. We present evidence for a potential binary system consisting of a WD and a BD on the basis of an observed excess flux at near and mind-infrared wavelengths. We also place limits on the presence of planetary mass companions around those stars and compare our results to predictions of planetary survival theories. Our findings do not support suggestions of planet formation or accretion of extra mass during stellar death.

The reduction of fuel consumption and the reduction of toxic emissions are the main goals of research and development in the area of internal combustion engines. The use of exhaust gas recirculation (EGR) to come further in that direction is today an established method for diesel engines. EGR reduces the emissions of nitrogen oxides with a low penalty in fuel consumption. The increasingly hard regulations on emissions put high pressure on the manufacturers to improve these systems. The present work aims at increasing the knowledge in the area of EGR. Two of the main challenges when applying EGR are addressed, efficiency and mixing. The efficiency of the EGR-system is analyzed, focusing on keeping the fuel penalty low for a given EGR-rate. Different layouts of the EGR system are studied and compared regarding their stationary and transient properties. Exergy analysis is used to show the potential for improvement in different system components. In the same time, exergy analysis as a tool is introduced and compared to energy analysis of a system. The usefulness of exergy analysis of the entire gas exchange is shown by the example of a heavy-duty diesel engine. The problem of EGR and air mixing is approached by a detailed study of the mixing process in a heavy-duty diesel engine. Different methods for the measurement of EGR distribution are presented and compared. Additionally, the possibility to predict the mixing effects by 1-D and 3-D simulation is assessed. It is shown that the mixing between air and EGR is highly dependent on the pulsating nature of the flow. The EGR is shown to be transported in packets in the air flow. This leads to the conclusion that mixing not only at the mixing point, but also mixing in flow direction needs to be optimized, as the distribution of EGR between the cylinders is dependent on the timing between the passage of the EGR packets and the valve opening time. / <p>QC 20140203</p>