Dans le cadre de cette thèse, nous avons réalisé l’étude de systèmes amorphes moléculaires chiraux en évaluant leur vieillissement, leur mobilité ainsi que leur capacité à recristalliser en fonction de la composition énantiomérique du matériau. Pour limiter les facteurs additionnels à la chiralité, ce travail s’est concentré sur des systèmes modèles formant des conglomérats stables : N-acetyl-α-methylbenzylamine (Nac-MBA) et 5-ethyl-5-methylhydantoin (12H). De ces travaux il ressort que l’impact le plus spectaculaire de la chiralité est exprimé dans la propension à la cristallisation ou l’habilité à former un verre (qui augmente de façon inversement proportionnelle de l’excès enantiomérique (ee)). Les cinétiques de vieillissement sont implicitement impactées par l’ee : Celles-ci sont plus lentes pour les ee importants. Enfin, il semble que les processus de relaxation ainsi que les temps associés soient identiques quelle que soit l’ee, bien que le comportement à la cristallisation soit lui significativement impacté. A noter que la signature la plus manifeste de la chiralité dans l’état amorphe du Nac-MBA s’exprime dans l’intensité diélectrique des processus D et α. / In the framework of this thesis, we carried out the study of amorphous chiral molecular systems by evaluating their molecular mobility, the evolution of physical properties during aging and the recrystallization behavior as function of the initial enantiomeric excess (ee). In order to avoid factors additional to chirality itself, we focused on enantiomeric systems forming stable conglomerates (full chiral discrimination in the solid state) by choosing two model compounds: 5-ethy-5-methylhydantoin (12H) and N-acetyl-α-methylbenzylamine (Nac-MBA). From this thesis it was shown that the most spectacular effects of chirality in the amorphous state is expressed in the GFA or the crystallization propensity. The GFA increases as the ee decreases. The kinetics of physical aging is implicitly impacted by chirality. Glassy pure enantiomer requires more time to reach equilibrium than that of an intermediate composition. This situation is hypothetically due to constraints effects mostly resulting from a strong nucleation behavior in the glass state at high ee. Furthermore, the time scale of all the processes (D, α, βJG, γ) and the evolution of their temperature dependency are approximatively identical even though the crystallization behavior is highly impacted by ee. it seems that molecular mobility would not be a key parameter in the crystallization behavior of Nac-MBA. The main expression of chirality in amorphous Nac-MBA is evidenced in the signature of the dielectric strength of both D and α processes.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20257号 / 医博第4216号 / 新制||医||1020(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 宮本 享, 教授 村井 俊哉, 教授 高橋 淳 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Dynamic simulation of a large-scale electric power system involves solving a large number of differential algebraic equations (DAEs) every simulation time-step. With the ever-growing size and complexity of power grid, dynamic simulation becomes more and more time-consuming and computationally difficult using conventional sequential simulation techniques. This thesis presents a fully distributed approach intended for implementation on High Performance Computer (HPC) clusters. A novel, relaxation-based domain decomposition algorithm known as Parallel-General-Norton with Multiple-port Equivalent (PGNME) is proposed as the core technique of a two-stage decomposition approach to divide the overall dynamic simulation problem into a set of sub problems that can be solved concurrently. While the convergence property has traditionally been a concern for relaxation-based decomposition, an estimation mechanism based on multiple-port network equivalent is adopted as the preconditioner to enhance the convergence of the proposed algorithm. The algorithm is presented in detail and validated both in terms of accuracy and capability

Ziel der Arbeit war es die Quantifizierung funktioneller bzw. mikrostruktureller Parameter des Herzmuskels mit Hilfe T1-basierter Methoden zu verbessern. Diese Methoden basieren darauf, die gewünschte Information durch eine geeignete Präparation der Magnetisierung bzw. durch die Gabe von Kontrastmittel in den Zeitverlauf der longitudinalen Relaxation zu kodieren. Aus der Änderung der Relaxationszeit läßt sich dann die gewünschte Information bestimmen. Dafür sollte sowohl der Einfluß der Anatomie als auch derjenige der Meßmethodik auf die Bestimmung der longitudinalen Relaxationszeit und damit auf die Quantifizierung der Funktion bzw. Mikrostrukturparameter untersucht werden. Speziell der Einfluß der Bildgebungssequenz führt dazu, daß nur eine scheinbare Relaxationszeit gemessen wird. Während dies keinen Einfluß auf die T1-basierte Bestimmung der untersuchten Mikrostrukturparameter hatte, ergab sich für die Perfusionsquantifizierung eine deutliche Abhängigkeit von den Parametern der verwendeten IRLL-Sequenz. Um diesen Einfluß gerecht zu werden, wurden an die Meßmethodik angepaßte Gleichungen zur Bestimmung der Perfusion gefunden mit denen die systematischen Abweichungen korrigiert werden können. Zusätzlich reduzieren die angepaßten Gleichungen die Anforderungen bezüglich der Inversionsqualität im schichtselektiven Experiment. Dies wurde in einem weiteren Projekt bei der Bestimmung der Nierenperfusion im Mausmodell ausgenutzt. Neben der Untersuchung der Auswirkungen der Meßmethode wurde auch der Einfluß der anatomischen Besonderheiten des Blutkreislaufs am Herzen auf die Parameterquantifizierung mittels T1-basierter Methoden untersucht. Es konnte gezeigt werden, daß auf Grund der Anatomie des Herzens bei typischen Orientierungen der Bildgebungsschicht, auch bei der schichtselektiven Inversionspräparation der Magnetisierung des Herzmuskels ein Anteil des Blutpools invertiert wird. Daraus folgt, daß die vereinfachende Annahme, nach welcher bei schichtselektiver Präparation in Folge von Perfusion nur Blut mit Gleichgewichtsmagnetisierung den Herzmuskel erreicht, nicht erfüllt ist. Es konnte gezeigt werden, daß dies bei Perfusion zu einer deutlichen Unterschätzung der berechneten Perfusionswertes führt. Um mit diesem Problem umgehen zu können, wurde aufbauend auf einem vereinfachten Modell der zeitlichen Entwicklung der Blutmagnetisierung eine Korrektur für die Bestimmung der Perfusionswerte gefunden welche den Einfluß der anatomischen Besonderheiten berücksichtigt. Das für die Perfusionskorrektur eingeführte Model prognostiziert ebenso, daß auch bei schichtselektiver Inversion die T1-basierte Bestimmung der untersuchten Mikrostrukturparameter von der Perfusion abhängig wird und eine systematische Überschätzung der quantifizierten Werte verursacht. Da die Perfusion im Kleintier deutlich höher ist als im Menschen, ist dieser Einfluß besonders in der präklinischen Forschung zu beachten. So können dort allein durch verminderte Perfusion deutliche Änderungen in den bestimmten Werten der Mikrostrukturparameter erzeugt werden, welche zu einer fehlerhaften Interpretation der Ergebnisse führen und somit ein falsches Bild für die Vorgänge im Herzmuskel suggerieren. Dabei bestätigt der Vergleich mit experimentellen Ergebnissen aus der Literatur die Vorhersagen für das Rattenmodell. Beim Menschen ist der prognostizierte Effekt deutlich kleiner. Der prognostizierte Fehler bspw. im RBV-Wert liegt in diesem Fall bei etwa 10% und wird üblicherweise in der aktuellen Forschung vernachlässigt. Inwieweit dies in er klinischen Forschung gerechtfertigt ist, muß in weiteren Untersuchungen geklärt werden. Den untersuchten Methoden zur Bestimmung von funktionellen und mikrostrukturellen Parametern ist gemein, daß sie eine exakte Quantifizierung der longitudinalen Relaxationszeit T1 benötigen. Dabei ist im Kleintierbereich die klassische IRLL-Methode als zuverlässige Sequenz zur T1-Quantifizierung etabliert. In der klinischen Bildgebung werden auf Grund der unterschiedlichen Zeitskalen und anderer technischer Voraussetzungen andere Anforderungen an die Datenakquisition gestellt. Dabei hat in den letzten Jahren die MOLLI-Sequenz große Verbreitung gefunden. Sie ist eine Abwandlung der IRLL-Sequenz, bei der mit einer bSSFP-Bildgebungssequenz getriggert ganze Bilder während eines Herzschlages aufgenommen werden. Die MOLLI-Sequenz reagiert dabei empfindlich auf die Wartezeiten zwischen den einzelnen Transienten. Um mit diese Problematik in den Griff zu bekommen und gleichzeitig die Meßzeit verkürzen zu können wurde eine neue Methode zum Fitten der Daten entwickelt, welche die Abhängigkeit der scheinbaren Relaxationszeit von der Wartezeit zwischen den einzelnen Transienten, sowie der mittleren Herzrate fast vollständig eliminiert. Diese Methode liefert für das ganze klinisch Spektrum an erwarteten T1-Zeiten, vor und nach Kontrastmittelgabe, stabile Ergebnisse und erlaubte ein deutliche Verkürzung der Meßzeit, ohne die Anzahl der aufgenommenen Meßzeitpunkte zu reduzieren. Dies wurde in einer initialen klinischen Studie genutzt, um ECV-Werte in Patienten zu bestimmen. Ein Nachteil der Verwendung der MOLLI-Sequenz ist, daß nur die scheinbare Relaxationszeit aus den Fit der Meßdaten bestimmt wird. Die standardmäßig genutzte Korrektur benutzt aber dem gefitteten Wert der Gleichgewichtsmagnetisierung um den wahren T1-Wert zu bestimmen. Somit ist es für die Bestimmung des T1-Wertes notwendig, die Qualität der Inversionspräparation zu kennen. Auf Basis der neuen Fitmethode wurde eine Anpassung der MOLLI-Sequenz demonstriert, welche die Bestimmung der Gleichgewichtsmagnetisierung unabhängig von der Qualität der Inversionspräparation erlaubt. Dafür verlängert sich die Meßdauer lediglich um einen Herzschlag um in geeigneter Weise ein zusätzliches Bild aufnehmen zu können. Abschließend wurde in dieser Arbeit der Signal-Zeit-Verlauf der MOLLI-Sequenz eingehend theoretische untersucht um ein besseres Verständnis der getriggerten IRLL-Sequenzen zu entwickeln. In diesem Zusammenhang konnte eine einfache Interpretation der scheinbaren Relaxationszeit gefunden werden. Ebenso konnte erklärt werden, warum die für ungetriggerte IRLL-Sequenzen abgeleitete Korrekturgleichung auch im getriggerten Fall erstaunlich gute Ergebnisse liefert. Weiterhin konnten Fehlerquellen für die verbleibenden Abweichungen identifiziert werden, welche als Ausgangspunkt für die Ableitung verbesserter Korrekturgleichungen genutzt werden können. / The goal of this work was to improve T1-based methods for quantification of functional and microstructural parameters of the heart muscle. These methods encode the desired information in the longitudinal relaxation by a dedicated magnetization preparation or by due to the administration of contrast agents. Hence, the alteration of the longitudinal relaxation time can be used to determine the desired information. To accurately quantify these parameters, the influence of the anatomy as well as the data acquisition on the longitudinal relaxation time and hence the quantification of the functional and micro structural parameters is investigated. It is known, that the choice of imaging sequence may influence the recovery of the magnetization and only an apparent relaxation time can be measured. While this had no effect on the T1-based quantification of the investigated microstructural parameters, the calculated perfusion value showed a strong dependence on the parameters of the used IRLL sequence. To take the influence of the imaging sequence into account, adapted equations for perfusion quantification were found. Hence, it was possible to correct for the systematic deviation by the IRLL sequence. Additionally, it could be shown that these adapted equations relax some of the requirements on the slice selective inversion experiment which could be utilized in the quantification of renal perfusion in a mouse model. Beside the influence of the imaging sequence also the influence of cardiovascular anatomy of the heart on the T1-based quantification methods was investigated. It was shown that for typical orientations of the imaging slice, also for the slice selective preparation a part of the blood pool magnetization is inverted. This violates the assumption that in the slice selective case only magnetization in equilibrium state enters the heart muscle and leads to a drastic underestimation of the quantified perfusion value. Based on a simplified model of the evolution of the blood magnetization the effects of the partial blood pool inversion were derived for perfusion quantification. The same simplified model was used, to investigate the influence of the imperfect slice selective inversion preparation in the T1-based quantification of the investigated micro structural parameters. It was shown, that the inflow of partially inverted blood into the capillary bed results in a perfusion dependent overestimation of the investigated microstructural parameters. As perfusion in small mammals is higher than in humans, the resulting bias has to be considered particularly in pre-clinical studies. In these animal models a reduced perfusion can result in a strong variation of the microstructural parameters which could be misinterpreted and hence may lead to a wrong understanding of the processes in the heart muscle. The predicted bias was compared with residual errors in the literature neglecting the partial inversion and found a good agreement in a rat model. For humans the expected bias is much smaller due to the lower perfusion values. The predicted bias for the RBV value is approximately 10% and hence, the effect is neglected in the current literature. However, if this justified must be investigated in further studies. All investigated methods for parameter quantification require the exact knowledge of the longitudinal relaxation time T1. For small animals the usual choice is an IRLL sequence, which have been established and demonstrated to be reliable and robust. Due to the different timescales and other technical aspects, however, the requirements in clinical imaging are different for data acquisition. In recent years the MOLLI sequence has become popular for T1 quantification. The MOLLI sequence is modification of IRLL sequence with a single shot bSSFP imaging module triggered usually to the end diastolic heart phase. However, the MOLLI sequence shows a strong dependence on the waiting times between the inversion prepared transients. To overcome this problem and provide a robust quantification of the apparent relaxation time with reduced the overall measurement time a new fitting procedure was developed. Thus, it was able to almost completely eliminate the dependence on the waiting time between the transients as well as the mean heart rate. The new method provided robust quantification over the complete range of clinical relevant longitudinal relaxation times (pre and post administration of contrast agents). Additionally, it was possible to reduce the measurement time without reducing the number of acquired data. This method was used in a pilot study to measure ECV in patients. A disadvantage of the MOLLI sequence is that in only provides an apparent relaxation time from the data fit and a correction for the real relaxation time is necessary. To calculate $T_1$, the common correction requires the knowledge of the equilibrium as well as the steady state magnetization. Hence, the quality of the inversion preparation is important and must be determined. Exploiting the properties of the new fitting method an adaption of the MOLLI sequence was proposed which allows the measurement of the equilibrium magnetization independent from quality of the inversion preparation by extending the measurement time for only a single heart beat to acquire a single additional image before the first inversion preparation. The final part of this work was dedicated improve the understanding of triggered IRLL sequences as the MOLLI. Hence, the signal evolution of these triggered sequences was investigated theoretically. Hence, a simple interpretation of the apparent relaxation time could be found from the results. Furthermore, a better understanding was reached for the surprisingly good results of the commonly used correction which was derived from the untriggered continuous case. Additionally, sources of the remaining deviations were identified and can be used for subsequent investigations to find better correction equations which allow for a more accurate quantification of T1.

Using the Theory of Alternative Behavior (Jaccard, 1981), we examined the relationship of warmth sensuousness, physical sensuousness, and sensation seeking to individuals' tanning salon behaviors, intentions, and attitudes among undergraduates at a Southeastern university. Females, high sensation seekers, those high in warmth sensuousness, and those with darker skin types were more likely to tan. Females were more likely to intend to tan in the next year. Those higher in warmth sensuousness were more likely to intend to tan more than 10 times in the next year. Females and subjects higher in warmth sensuousness had more positive attitudes toward tanning. Significant interactions were found between warmth sensuousness and sensation seeking in the predictions of intention of tanning within the next year, and intentions of tanning more than 10 times within the next year. For both interactions, as sensation seeking increased, the relationship between warmth sensuousness and intentions strengthened.

The advancement of both electronics and instrumentation technology has fostered the development of multi-physics platforms that can probe the earth’s subsurface. Remote, non-destructive testing techniques have led to the increased deployment of electromagnetic waves in sensor technology. Electromagnetic wave techniques are reliable and have the capacity to sense materials and associated properties with minimal perturbation. However, meticulous data analyses and mathematical derivations reveal inconsistencies in some formulations. Thus, revisiting the fundamental physics that underlies both electrical impedance experimental setups and electromagnetic properties are paramount. This study aims to unravel inherent limitations in the understanding of the relationships between electromagnetic and non-electromagnetic properties that are relevant to the characterization of fluids in porous media. These correlations pervade porosity, permeability, specific surface, pore size distribution, tortuosity, fluid discrimination, diffusion coefficient, degree of saturation, viscosity, temperature, phase transformation, miscibility, salinity, and the presence of impurities. The focus is on the assessment of liquids, soils, rocks, and colloids using broad spectral frequency complex permittivity, conductivity, magnetic permeability, and nuclear magnetic resonance relaxometry. Broadband electrical properties measurement for saturated porous media can provide multiple physical phenomena: Ohmic conduction, electrode polarizations, Maxwell-Wagner spatial polarizations, rotational, and segmental polarizations. Liquids dominate the electromagnetic signatures in porous media as dry minerals are inherently non-polar and non-conductive. Results reveal that voltage drops due to the discontinuity of charge-carrier at the electrode-electrolyte interface named electrode polarization inherently affect the low-frequency electrical measurements both in two- and four-probe configurations. Rotational polarizations that occur in MHz-GHz ranges are defined by the electrical dipole moment and effective molecular volume. Both viscosity and effective molecular volume govern the NMR transverse relaxation time. An engineered soil suspension with ferromagnetic inclusions exhibits excellent characteristics for drilling fluid application. Overall, the study highlights the complementary nature of conductivity, permittivity, and NMR relaxation for the advanced characterization of fluid saturated geomaterials.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

We investigate the properties of two distinct disordered systems: the two-species predator-prey Lotka-Volterra model with rate variability, and an elastic line model to simulate vortex lines in type-II superconductors. We study the effects of intrinsic demographic variability with inheritance in the reaction rates of the Lotka-Volterra model via zero-dimensional Monte Carlo simulations as well as two-dimensional lattice simulations. Individuals of each species are assigned inheritable predation efficiencies during their creation, leading to evolutionary dynamics and thus population-level optimization. We derive an effective subspecies mean-field theory and compare its results to our numerical data. Furthermore, we introduce environmental variability via quenched spatial reaction-rate randomness. We investigate the competing effects and relative importance of the two types of variability, and find that both lead to a remarkable enhancement of the species densities, while the aforementioned optimization effects are essentially neutral in the densities. Additionally, we collected extinction time histograms for small systems and find a marked increase in the stability of the populations against extinction due to the presence of variability. We employ an elastic line model to investigate the steady-state properties and non-equilibrium relaxation kinetics of magnetic vortex lines in disordered type-II superconductors. To this end, we developed a versatile and efficient Langevin molecular dynamics simulation code, allowing us to do a careful study of samples with or without vortex-vortex interactions or disorder allows us to disentangle the various complex relaxational features present in this system and investigate their origin. In particular, we compare disordered samples with randomly distributed point defects versus correlated columnar defects. We extract two-time quantities such as the mean-square displacement, the height and density correlations, to investigate the relaxation kinetics of the system of flux lines. Additionally, we compare the steady-state mean velocity and gyration radius as a function of an external driving current in the presence of point-like and columnar disorder. We validate our simulation algorithm by matching our results against a previously-used Monte Carlo algorithm, verifying that these microscopically quite distinct methods yield similar results even in out-of-equilibrium settings. / Ph. D.

Nonsmoothness appears in various applications in chemical engineering, including multi-stream heat exchangers, nonsmooth flash calculation, process integration. In terms of numerical approaches, convex/concave relaxations of static and dynamic systems may also exhibit nonsmoothness. These relaxations are used in deterministic methods for global optimization. This thesis presents several new theoretical results for nonsmooth sensitivity analysis, with an emphasis on convex relaxations. Firstly, the "compass difference" and established ODE results by Pang and Stewart are used to describe a correct subgradient for a nonsmooth dynamic system with two parameters. This sensitivity information can be computed using standard ODE solvers. Next, this thesis also uses the compass difference to obtain a subgradient for the Tsoukalas-Mitsos convex relaxations of composite functions of two variables. Lastly, this thesis develops a new general subgradient result for Tsoukalas-Mitsos convex relaxations of composite functions. This result does not limit on the dimensions of input variables. It gives the whole subdifferential of Tsoukalas-Mitsos convex relaxations. Compare to Tsoukalas-Mitsos’ previous subdifferential results, it does not require additionally solving a dual optimization problem as well. The new subgradient results are extended to obtain directional derivatives for Tsoukalas-Mitsos convex relaxations. The new subgradient results and directional derivative results are computationally approachable: subgradients in this article can be calculated both by the vector forward AD mode and reverse AD mode. A proof-of-concept implementation in Matlab is discussed. / Thesis / Master of Applied Science (MASc)