The increases in complexity of industrial process simulations necessitate faster methods and more accurate models for analysis. In many cases automation of industrial plants using real-time control cannot be solved using present computer methodologies. This thesis introduces a parallel methodology that outlines the steps to design and implement a distributed continuous simulation. Implementing the system in parallel reduces the simulation time to comply with predetermined time constraints. The methodology determines the structure of the simulator, organization of the distributed tasks and management of the variable memory allocation. / A dynamic model for transient, incompressible pipe flow is used to test the parallel methodology. The generalized flow equations are developed for two applications: a tank-pipe and pipe network model. The models are translated into computer algorithms using parallel techniques developed in the simulation methodology. The individual subroutines of the simulation are directly implemented on a parallel processing and a multi-tasking computer. The flow algorithms are tested on both computers to demonstrate the benefits obtained by using the parallel methodology on a multi-processor computer. Model sizes are varied to provide a performance scale comparing multi-processor results with that of a single processor.

This thesis reports on an investigation of unstable detonations near the limits of gaseous mixtures, which has been carried out in a continuous loop of 5 cylindrical tubes, of 11 m length each, of decreasing diameter (from 152, to 38 mm). The experiments were carried out for different mixtures of hydrogen, ethylene and propane in air, and also hydrogen-oxygen with argon as diluent. It is reasonable to seek a dependence of detonability limits on the geometry length scale of the system, i.e. the tube diameter d, and on the characteristic of the cell width $ lambda$. A d/$ lambda$ criterion has been proposed to define the limits in circular tubes. However, the present results show that such a correlation is not sufficient. For a ratio $ lambda$/d less than 1, the transition back to detonation occurred relatively easily within the length of the tubes. However, for $ lambda$/$d >$ 1 the phenomena observed are highly unstable. Also mixtures of different components exhibit a different behaviour for the same $ lambda$/d ratio.

The quasi-steady analytical approach was used to develop a mathematical model for investigating instability using a rotatable five cylinder cluster with one "movable" cylinder. In this model, the important time retardation factor is taken into account. First, the equations of motion for the "movable" cylinder were written in dimensional form, and then non-dimensionalized and written in the final matrix form. Secondly, the force coefficients and their spacial derivatives with respect to cross-flow and in-flow displacements were measured on the "movable" cylinder. Also, the current direction was measured and the approach velocity was calculated. Finally, using this model and the experimental input data, the critical flow velocity for the onset of fluidelastic instability was calculated using a computer program developed for this purpose. / It is shown that, depending on the current direction, riser clusters may suffer from fluidelastic instability. The theoretical findings are in good agreement with experimental data. (Abstract shortened by UMI.)

This thesis considers the transverse vibrational control of a single flexible beam under-going large planar rotational motion. The beam is modelled using a cubic spline technique, which approximates the linearly elastic, continuous beam with a finite number of nodal points. It is shown that kinetic boundary conditions such as those associated with tip loads can be included in the cubic spline model. This spatial discretization method provides a useful linear relationship between displacement and curvature, which allows the use of strain gages to measure curvature along the beam. An optimal control strategy is used to suppress the transverse vibrations while forcing the end tip to follow a prescribed trajectory. A Kalman filter is employed to optimally estimate state variables which are not obtained through direct measurement. These state variables can be classified into two groups: (1) state variables corresponding to the time rate of change of curvature, which cannot be measured using any existing sensor and (2) state variables associated with curvature at certain nodal points which results in a smaller number of sensors needed to control the entire beam.

The goal of this thesis is to develop tools that will enable the design of the components of the Rankine micro-engine in order to satisfy performance and reliability requirements for high temperature operation. The Ashby approach was used to identify a set of candidate materials – zirconia, silica, titanium alloy, nickel-cobalt alloy, silicon carbide and silicon – for the insulating static structure and rotor. Subsequently, the stresses in, and deformations of, the insulating static structure were analyzed using analytical models and the method of finite elements. Subsequently, an idealized impact experiment was designed and performed to assess reliability of the rotor in the event of an accidental high-speed impact with the static sidewalls. The threshold velocity for initiating damage in alumina projectiles (balls) impacting on alumina targets (discs) was found to be 30 m/s, while the corresponding value for zirconia projectiles on alumina target was found to be around 65–70 m/s. These findings pave the way for developing detailed designs for the rotor and static sidewall of the Rankine micro-engine. / L'objectif de cette thèse est de développer des outils qui permettront la conception de composants de moteur, afin de satisfaire les exigences de performance et de fiabilité de l'appareil dans des conditions extrêmes. Le méthode Ashby a été utilisée pour identifier un ensemble de matériaux - la zircone, la silice, de l'alliage de titane, de l'alliage nickel-cobalt, de silicium et de carbure de silicium – pour la structure statique isolante du rotor, ce qui révèle un compromis entre la performance, la fiabilité et la fabricabilité . Ultérieurement, les contraintes, et les déformations de la structure statique d'isolation ont été analysées par la méthode des éléments finis.Par la suite, une expérience idéalisée d'impact a été conçue et réalisée pour évaluer la fiabilité du rotor dans le cas d'un accident d'impact à grande vitesse avec la paroi latérale de la structure statique. La vitesse seuil d'ouverture des dommages dans des projectiles d'alumine (balles) un impact sur les objectifs de l'alumine (disques) a été retrouvé à 30 m/s, alors que le projectile en zircone sur la cible d'alumine a été jugé dans la région de 65 - 70 m/s. Ces résultats ouvrent la voie à la formulation de matériaux de conception pour le rotor et la paroi latérale statique du micro-moteur Rankine.

In mineral beneficiation, grinding is used to reduce the size of the ore particles, generate larger surface area, and liberate valuable components. Grinding represents the most energy consuming stage of the process. Continuous research has aimed to increase the rates and efficiencies of milling by predicting the behaviour of large industrial mills, for example, through an examination of energy-size relationships and by using population balance models. The input for these models comes from data extracted from small laboratory-scale mills or impact tests. However, these breakage-testing techniques do not properly describe the behaviour of industrial mills, since these techniques cannot achieve the necessary modes of breakage or the correct input forces. Since the steel wheel abrasion test (SWAT) produces particle size reduction through abrasion breakage under a wide range of controlled input forces, this thesis investigates the use of the SWAT for ore characterization. This test is a variation of the dry and wet sand/rubber wheel abrasion test, a standard for measuring mass loss in metallic materials subjected to scratching abrasion. This thesis studies three different materials tested in the SWAT under the t10 ore characterization scheme. The results show that the t10 model can be properly fitted to the data collected from the tests. Two of the ores also were analyzed by using the drop weigh test (DWT), and the resulting parameters were compared to those from the SWAT, which showed that a relationship between the resulting parameters could be established. However, the results suggested that it seems more promising to use the SWAT to determine the abrasion breakage component and to calculate the breakage function of the small particles (below the inflection point), a function that otherwise would be unattainable in the DWT. / En minéralurgie, le broyage est l'opération durant laquelle le minerai est fragmenté afin d'en augmenter la surface libre et de le libérer de sa gangue. C'est l'étape la plus énergivore du processus de traitement de la matière première. Elle a constamment été l'objet d'études visant à augmenter le débit des grands moulins industriels et à en améliorer le rendement par la modélisation de leur comportement, en examinant par exemple les relations énergie-taille et en utilisant des modèles basés sur des bilans de population. Les paramètres d'entrée de ces modèles proviennent des données mesurées sur de petits moulins de laboratoires ou lors de tests d'impact. Cependant, ces expériences de broyage ne décrivent pas précisément les mécanismes de fragmentation ayant lieu dans les moulins industriels car elles ne mettent pas en jeu les mêmes forces et modes de rupture.Le test d'abrasion par roue d'acier (Steel Wheel Abrasion Test, SWAT) fragmente les particules de minerai par rupture abrasive sous une large gamme de forces en entrée. Notre thèse étudie la caractérisation du minerai par le test SWAT. Ce test est une variante du test standard d'abrasion par roue de caoutchouc sablée, humide ou à sec, utilisé pour mesurer la perte de masse de matériaux métalliques sujets à une usure pas frottements.Nous testons trois matériaux par le test SWAT dans cette étude et utilisons le procédé de caractérisation t10. Les résulats montrent que le modèle t10 peut s'accorder correctement avec les données collectées lors des tests. Deux matériaux ont également fait l'objet d'essais de choc (Drop Weight Tests). Les paramètres en résultant ont été comparés à ceux issus du test SWAT et une relation a été établie. Cependant, les résulats obtenus suggèrent que le test SWAT est plus adapté pour déterminer les composantes de la fragmentation par abrasion et pour calculer la

This thesis investigates the mechanics of lubricant starvation in rolling element bearings. A hydrodynamic study is made which assumes incompressible, steady, isoviscous flow with body forces negligible. A cylinder on plate mathematical model is utilized to establish an analytical solution for low capillary numbers in sliding contacts. A complete load and stability analysis is obtained for this flow regime. / The case of general capillary number is undertaken via a numerical method. A finite volume approach using curvilinear orthogonal coordinates is implemented. A complete grid generator and flow solver are developed and used to determine the velocity and pressure fields for general capillary number. Results of the numerical and analytical models are compared for low capillary numbers and show good correlation.

The effect of the ratio of the velocity field integral scale at the injection point to the scalar injection scale (the injection ratio) on the mixing of a passive scalar (temperature) is studied in the wake of a circular cylinder. Hot-wire anemometry and cold-wire thermometry are used to measure simultaneously the velocity and thermal fields, respectively. A mandoline is used to (slightly) heat the wake. / The injection ratio is varied in two ways. The velocity integral scale at the injection point is varied by changing the mandoline downstream position. Increasing the velocity scale, ℓ₀ (i.e., increasing the mandoline downstream position), increases the rate at which the scalar variance decays and reduces the integral length scale of the scalar. The time scale ratio (r = (〈q²〉/ε)/(〈θ²〉/εθ)) decreases with increasing ℓ₀, in agreement with experiments done in homogeneous and isotropic turbulence. / The scalar injection scale is varied by changing the width, w, of the mandoline. Increasing w decreases the decay rate of the scalar variance, which is analogous to increasing the wire spacing in grid-generated turbulence. The integral length scale of the scalar field (both longitudinal and transverse) and the time scale ratio are, however, not affected by w, in contrast with experiments done in homogeneous and isotropic turbulence.

The simulation of mathematical models of mechanical systems with closed kinematic chains involves the solution to a system of highly coupled differential-algebraic equations. The numerical stiffness of these systems calls for small time steps in order to insure accuracy. Real-time and interactive forward simulations tend to be difficult to achieve for such systems, especially for large multi-body systems with multiple links and many kinematic loops. One way to overcome the time constraint is to distribute the load onto several processors. / The modular formulation of mathematical models is attractive because existing models may be assembled to create different topologies, e.g. cooperative robotic systems. Conversely, a given robotic topology may be broken into smaller topologies with simpler dynamics. / Moreover, parallel-kinematics machines bear inherent spatial parallelism. This feature is exploited in this thesis, in which we examine the formulation of such modular and distributed models and evaluate their performance as applied to mechanical systems with closed kinematic chains. Three general undistributed formulation methods are specialized to cope with distribution and modularity and applied to a three-degree-of-freedom planar parallel manipulator to generate distributed dynamics models. / Finally, the results of case studies are reported, and a comparison is made to highlight the salient features of each method.

The behavior of the shear layers separating from a rotationally oscillating cylinder was investigated using hot-wire anemometry and smoke-flow visualization in a wind tunnel at a Reynolds number of 3700. The cylinder was oscillated at small peak-to-peak amplitudes and frequencies lower than and near the natural Kannan frequency. The shear layers responded in a way such that their characteristics were modified depending on the frequency and amplitude of oscillation. Results showed that as the oscillation frequency increased, the shear layers were highly disturbed, the shear-layer instability was promoted upstream and its frequency increased, the shear-layer vortices formed antisymmetrically and closer to the cylinder, the vortex formation region characteristics changed and the vortex formation process was enhanced. Consequent effects on the near-wake flow structure were observed; these ranged from variations in the wake width to reduction in the drag on the cylinder. The wake width increased to a maximum at subharmonic forcing, only to decrease again to values close to that of the stationary cylinder case at harmonic forcing. Further increase in the oscillation frequency resulted in the narrowing of the wake. Reduction in the drag on the cylinder as well as an indication of an increase in the vortex strength was observed at several oscillations frequencies. These effects suggest a potential and promising means of controlling the vortex formation mechanism and the near-wake flow structure by controlling the behavior of the shear layers through rotational oscillations.