This thesis consists of two major research topics: (i)the thermodynamics of dilute platinum-hydrogen solid solutions and (ii)the diffusivity of hydrogen in nickel-vandium solid solutions. For the thermodynamics of dilute platinum-hydrogen solutions, the solubility of hydrogen in platinum is measured. The equilibrate-quench-analyze technique is employed. The partial molar enthalpy and excess entropy of hydrogen are calculated. It is concluded that this type of solution behaves in excellent agreement with the quasi-regular model. For the diffusivity of hydrogen in Ni-V solutions, the diffusivity is determined by using the permeation time-lag technique. It is found that (a)the hydrogen diffusivity obeys the Arrhenius relationship, and (b)the effect of V-additions is remarkably small, even for the highest V-composition (13.4 At.%). Since the addition of the vanadium caused the Ni-V matrix lattice to dilate, these observations lead us to the following conclusion: The decrease of the hydrogen mobility caused by the trapping sites adjacent to the vanadium atoms is compensated by an increase of the hydrogen mobility created by the dilation of the Ni-V matrix lattice. For a better understanding of metal-hydrogen systems, a brief review of their basic properties and applications is also included.

An electrochemical technique has been used to measure the diffusivity of hydrogen in high purity noble metals (Ag, Au, Cu, Pd and Pt) in the well-annealed state at low temperatures (283-323K). The data obtained have been considered in conjunction with those for as-quenched and aged Au samples in order to draw conclusions concerning the interactions between dissolved H atoms and lattice vacancies. It has been concluded, at least, in Pd, Cu and Au, that the H-vacancy interaction energy is too low to render diffusion anomalies observable at high temperatures. This is in accord with theoretical calculations based on the effective medium technique. The resulting diffusivities for quenched and aged Au are consistent with a simple statistical model in which a fraction of the H atoms saturate traps formed by quenching in vacancies, but the mobility is primarily determined by the excess of nontrapped H atoms diffusing through "normal" sites in FCC metals.

In inhomogeneously stressed solid solutions, a force is exerted on each solute atom. The magnitude and direction of the force depend on the solute atom's chemical potential gradient. The chemical potential, in turn, depends on the local solute concentration and local stress state. The forces exerted on the solute atoms cause them to be redistributed within the stressed body. This diffusion of solute atoms is described by Fick's laws. Near a tensile crack, the primary component of interstitial diffusion is directed radially inward, toward the crack tip. Solute atoms accumulate there. Such accumulation is necessary in order for embrittlement to occur. Embrittlement is most acute at a particular temperature, the one at which the radial component of solute mobility is at a maximum. By introducing a zone of plastic deformation around the crack tip and by assuming that solute atoms do not interact, one obtains a critical material strength below which no embrittlement can occur. This critical strength has no explicit dependence on microstructure, applied stress or shape of the cracked body. Calculated values of the critical material strength and the temperature of most acute embrittlement are in reasonable agreement with experimental values, for the case of hydrogen in steel. Solute accumulation within a particular microstructural feature leads to fracture of that feature, followed by repeated solute accumulation and fracture. An average crack growth rate can be calculated for this discontinuous process. The calculated growth rate curves, as a function of applied stress intensity, exhibit three distinct regions that are also observed in experimentally measured growth rate curves.

This study investigates the thermodynamics and kinetics of hydrogen in a binary palladium-based metal solution using classical jump rate theory applied to the cell model approach. An electrochemical technique is used to measure the diffusivities of hydrogen in palladium alloys containing Er concentrations up to 8 atomic percent at temperatures ranging from 273 to 340 K. The activation energy for hydrogen diffusion was found to increase with Er content to a maximum at five atomic percent, then decrease. A second-order model is presented which reconciles the thermodynamic behavior of the system against the observed diffusivities in the Pd-Er-H system. Related topics for cell models and interstitial systems are also presented. Solubilities for high nitrogen concentrations in liquid iron-based solutions are explained using the cell model approach. The thermodynamic and kinetic behavior of interstitial-vacancy interactions are examined in light of experimental data for carbon in iron, cobalt, and nickel. In these studies, vacancy-interstitial binding energies as high as 1 eV for Fe-C are found to have no perceptible influence in the solubility data. Enthalpy data suggest an upper interstitial-vacancy binding energy limit of 0.4 eV in the Co-C system. When extended to second-order terms and lattice dilation, the cell model is in agreement with observed non-linearities for solvent diffusion for the fcc Fe-C and Ni-C systems. Finally, high temperature deviations in the Arrhenius behavior for interstitial diffusion are explained in terms of magnetic disordering for the Fe-C system, and dual interstitial site occupation for the Co-C system.

The mechanisms responsible for microplasticity in silicon and germanium subjected to low-load contact damage have been investigated. Microstructural characterization has been carried out primarily by electron microscopy and electron diffraction. The particular contact damage processes that have been studied are indentation, scratching, and single-point diamond turning (ductile-regime machining). The substantial plasticity that results from the initial stages of low-load indentation in silicon and germanium is not controlled by a traditional mechanism such as dislocation activity. Evidence of plasticity includes a permanent indentation impression and ductile extrusions emanating from the indentation interior. Rather, the plasticity is a result of a ductile metallic phase that forms under the indenter from a pressure-induced metallization of the semiconductor (Mott transition). A metastable amorphous phase is formed when the pressure on the metallic phase drops as the indenter is lifted from the surface. This amorphous phase makes up the indentation interior and the ductile extrusions. The extent of the transformed zone in plan-view equates with that of the contact area of the indenter. Observed dislocation activity results from high stress fields ahead of crack tips. Plasticity induced by low-load scratching of silicon is likewise the result of a pressure-induced metallization of the semiconductor. Scratches show the same microstructural features as indentations, namely, a permanent plastic impression with a ductile morphology, ductile extrusions, and an amorphous remnant. The extent of the phase transformed zone in plan-view is exactly that of the contact area. The extent of the transformed zone in cross-section is between 200 to 300 nm depending on load. Minor dislocation activity is and/or cracking is found under the transformed zone. The "ductile regime" in silicon and germanium that allows for ductile material removal in precision machining processes is also a result of the same pressure-induced metallization. The chips from a germanium machining experiment are completely or largely amorphous and show machining marks and shear instability morphology. The ductile morphology and amorphous structure of the chips lead to the conclusion that the metallization and back-transformation is again responsible for plasticity. Morphological and crystallographic analysis of the machining chips allow for the location on the original workpiece, from which certain chips originated, to be known.

Nondestructive testing of concrete for maintenance purposes is the focus of this thesis. The application of a new nondestructive testing technique known as impact-echo was examined. The technique is based on transient stress wave propagation for the detection of defects in concrete, as well as measuring the setting time, early age strength, and elastic properties. Impact-echo functions by a mechanical impact, where stress pulses are generated in the test subject. The stress pulses undergo multiple reflections between the top and bottom of concrete layers. The surface displacements are recorded and the frequency of the successive arrivals of the reflected pulses is determined. Thus, knowing the thickness of a given layer, together with the measured frequencies, P- and S-wave velocities can be calculated. If on the other hand, the thickness is unknown, the time-distance graph of the primary surface wave arrivals could be used to calculate the thickness. The position of the impact source relative to the receiver must be selected in such a way as to detect P- and S-waves at their maximum reflection amplitude. In this study, defects were detected using the reflected P-waves from their top surfaces. In addition, the change in elastic moduli of various types of concrete mixes was monitored for a 30-day period by measuring P- and S-wave velocities. The impact-echo elastic moduli were then compared with static and dynamic values obtained by standard methods in order to assess their accuracy.

The processing conditions for carbothermal reduction of silica to silicon nitride was found to be sensitive to several key processing parameters: namely the intimacy of mixing of carbon and silica, the temperature, the specific high surface area of carbon, the nitrogen gas purity and the action of the nitrogen gas passing through the reactants. / Sol-gel processing was found to provide superior mixing conditions over dry mixing, which allowed for complete conversion to silicon nitride at optimum carbon:silica ratios of 7:1. The ideal reaction temperature was found to be in the range of 1500$ sp circ$C to 1550$ sp circ$C. Suppression of silicon oxynitride and silicon carbide was achieved by ensuring that: (a) the nitrogen gas was gettered of oxygen, and (b) that the gas passed through the reactants. Thermodynamic modelling of the Si-O-N-C system showed that ordinarily the equilibrium conditions for the formation of silicon nitride are very delicate. Slight deviations away from equilibrium leads to the formation of non-equilibrium species such as silicon carbide caused by the build-up of carbon monoxide. Reaction conditions such as allowing nitrogen gas to pass through the reactants beneficially moves the reaction equilibrium well away from the silicon carbide and silicon oxynitride stability regions. / The particle size of silicon nitride produced from carbon and silica precursors was of the order of 2-3 $ mu$m and could only be reduced to sub-micron range by seeding with ultra-fine silicon nitride. It was shown that the mechanism of nucleation and growth of unseeded reactants was first nucleation on the carbon by the reaction between carbon, SiO gas and nitrogen (gas-solid reaction), and then growth of the particles by the gas phase reaction (CO, SiO, N$ sb2$).

A new process was developed to fabricate particulate metal matrix composites (MMCs). The process involves two steps: (1) forming a particulate porous compact, and (2) introducing molten magnesium or magnesium alloy (AZ91) into the channel network by a spontaneous infiltration technique. / A uniform distribution of SiC particulates in magnesium matrix was achieved. Microsegregation existed in composites when the particle size of reinforcement was 38 mum and 22 mum, however, when the particle size of reinforcement was 12 mum, the microsegregation was eliminated. / The interfacial reaction between SiC and Mg studied by Scanning Electron Microscopy (SEM), microanalysis, and X-Ray Diffraction (XRD) techniques showed that, Mg reacted with Si or SiO2 to form Mg2Si, Mg 2Si was present at the surface of SiC after precipitation, resulting in improved wettability between Mg and SiC. As a result of this interfacial reaction, the infiltration process became spontaneous. The infiltration process was related to the infiltration temperature, SiC particle size and matrix chemistry. Increasing the infiltration temperature, decreasing the particle size of SiC resulted in more successful infiltration. / Mechanical testing conducted on composites revealed that the hardness and Ultimate Tensile Strength (UTS) of composites increased with the decrease of particle size of SiC and corresponded to an increase of the volume fraction of SiC.

Knelson concentrators are the most widely used semi-continuous centrifuge separators for the recovery of gold and platinum minerals by gravity methods. Bench scale characterization studies on these units provide information about the occurrence of gold in ore samples (e.g. gold particle size distribution, amount of gold recoverable by gravity) but not about the effect of operating variables for full-scale units such as top size of particle, feed rate, fluidization flow rate and rotation speed. Such work is not easily performed online on full-scale units owing to the inevitable variations in feed quality and to the impossibility of varying operating parameters systematically in the face of production requirements. To attack the problem, a pilot plant comprising a 12-in CD Knelson concentrator, a feed screen and tailing sump-pump arrangement was installed in the grinding-B circuit of Dome Mine, Porcupine Joint Venture (PJV), now Porcupine mine, Goldcorp Inc. Timmins, Ontario. The pilot plant received a bleed from the feed to the full-scale units. The pilot facility was extensively sampled in two campaigns. Fifteen tests were conducted in the first campaign and another sixteen in the second. In all 31 pilot tests, twenty six 30-minute recovery cycle tests, called "short tests" and five, 90-minute recovery cycle tests, dubbed "long tests", were conducted. Measuring recovery was the focus of the "short tests"; measuring the deterioration of recovery over time was the focus of the "long tests". The sampling protocols were designed accordingly. / Detailed metallurgical balances were made to analyze the effect of operating and design variables on the performance of 12-in pilot Knelson Concentrator as a step towards understanding full-size units and to study the mechanism of concentrate bed erosion. To gain some fundamental information about the recovery mechanism of the Knelson concentrator, percolation of dense particles in a gangue bed was investigated using a fluidized bed column in the gravitational field. / Metallurgical results indicate that operating conditions including feed rate, rotation velocity, fluidization water flow rates and top feed particle size have little impact on the shape of the recovery compared to feed size distribution. A particle size hypothesis was tested using relevant industrial Knelson concentrator data. The analysis showed that a relatively coarse feed would impact negatively on the recovery between 106 and 425 µm. On the other hand, it would make it easier to recover particles between 25 and 106 µm. A finer feed would have a bigger impact on recovery around 25 to 106 µm and would yield a GRG recovery that decreases monotonically with the decreasing of particle size. This would be linked to the natural resistance offered by the gangue particles to the percolation of gold particles, which is significant at a particle size where the gangue is most abundant. The flowing slurry may be compared to a dynamic screen, with openings roughly the order of magnitude of the dominant particle size. This finding is useful for the simulation of the Knelson units, which uses the typical recovery curve "decreasing recovery with decreasing particle size" for estimating gravity recovery and it was thought that the shape of the curve had no impact on the estimation. Now, with this finding, either the fine or coarse recovery curve will be used depending on the size distribution of the gravity circuit feed. For example, for a coarse target grind, the coarse curve could be used and, for a fine target grind, the fine curve could be used. / Les concentrateurs Knelson sont les séparateurs centrifuges semi-continus les plus utilisés pour la récupération par gravité des minéraux d'or et de platine. Des études de caractérisation par banc d'essai sur ces unités fournissent de l'information sur l'occurrence d'or dans les échantillons de minerai (ex. la distribution de dimension des particules d'or, la quantité d'or récupérable par gravité) mais non sur l'effet des variables opérationnelles sur les unités à pleine échelle, comme la dimension supérieure, la vitesse d'alimentation, le débit de fluidisation et la vitesse de rotation. Ces travaux ne sont pas faciles à effectuer en ligne sur des unités à pleine échelle en raison des inévitables variations dans la qualité de l'alimentation et l'impossibilité de varier les paramètres d'opération à cause des contraintes de production. Pour résoudre le problème, une usine pilote comprenant un concentrateur Knelson avec un diamètre du cône (DC) de 12 pouces, un tamis d'alimentation et une pompe à résidus ont été installés sur le circuit de broyage-B de Dome Mine, Porcupine Joint Venture (PJV), maintenant Porcupine mine, Goldcorp Inc., Timmins, Ontario. L'usine pilote recevait une purge de l'alimentation des unités à pleine échelle. L'installation pilote a été échantillonnée de façon intensive lors de deux campagnes. Quinze tests ont été effectués durant la première campagne et seize autres dans la deuxième. En tout 31 tests, soit vingt-six tests ayant un cycle de récupération de 30 minutes (appelés « tests courts »), et cinq tests ayant un cycle de récupération de 90 minutes (appelés « tests longs ») ont été faits. Le focus était mis sur la mesure de la récupération pour les tests courts, le focus était mis sur la détérioration du temps de récupération pour tous les tests. Le protocole d'échantillonnage a été conçu conséquemment. / Les résultants métallurgiques indiquent que les conditions l'opération telles que la vitesse d'alimentation, la vélocité de rotation, le débit de l'eau de fluidisation et la dimension des particules grossières ont eu de l'impact sur la récupération comparativement à la distribution des dimensions de l'alimentation. Une hypothèse quant à la dimension des particules a été testée en utilisant les données industrielles reliées au concentrateur Knelson. L'analyse a montré qu'une alimentation relativement grossière aurait un impact négatif sur la récupération de particules entre 106 et 425 µm. D'un autre côté, elle faciliterait la récupération des particules entre 25 et 106 µm. Une alimentation plus fine aurait un plus grand impact sur la récupération de partie autour de 25 à 106 µm et apporterait une récupération par gravité de l'or (« GRG ») qui diminuerait de façon monotone avec la diminution de la dimension des particules. Ceci serait lié à la résistance naturelle des particules de gangue à la percolation des particules d'or, ce qui est significatif à une dimension de particules où la gangue est plus abondante. Le liquide chargé s'coulant peut être comparé à un tamis dynamique, avec des ouvertures environ de l'ordre de grandeur de la dimension dominante des particules. Cette découverte est utile à la simulation des unités Knelson, qu'utilise habituellement une courbe de récupération dans laquelle « la récupération diminue avec la diminution de la dimension des particules », pour permettre l'estimation de la récupération par gravité; il était industriel de penser que la forme de la courbe n'avait pas d'impact sur l'estimation. Avec cette découverte, dorénavant la courbe de récupération, fine ou grossière, sera utilisée en fonction de la distribution des dimensions de l'alimentation du circuit gravitationnel.

To establish the effect of reducing the temperature mismatch between the titanium hydride decomposition temperature and the aluminium melting point on the foams morphological features and their mechanical compression behavior, a nickel coating on titaniun hydride powders was used as a hydrogen diffusion barrier and the size of titanium hydride powders was controlled to modify the hydrogen evolution temperature. / The nickel diffusion barrier was produced by an electroless deposition technique and the hydrogen evolution behavior of coated powders was investigated by thermogravimetrical analysis. The effect of particle size was determined with powders of five particle size fractions along with powders of different particle size obtained from a supplier. Foamable precursors were obtained by hot pressing a mix of aluminium powders with 1 wt.% of titanium hydride powders and foams were fabricated at 750 and 800 °C. The foams mechanical strength was investigated by uni-axial compression on foam cylinders with and without outer skin. / Coating produced a continuous and homogeneous deposit of 96.5 wt.% nickel and reduced the initial temperature mismatch by approximately 70°C. Additionally, the coating adhesion proved to be good enough to withstand the mixing and compaction processes. Nickel-coated titanium hydride powders generated foams with a more homogeneous and reproducible pore structure than foams produced with powders in the as-received and passivated condition. On the other hand, the hydrogen evolution onset of titanium hydride shifted towards higher temperatures as the particle size increased. The particle size influenced the foam expansion and the porosity features. Powders of larger particle size produced foams with a more uniform pore distribution and size. Finally, compression tests on skinless foams containing nickel displayed quasi-horizontal energy regimes with longer stroke lengths than the rest, however the final energy absorption efficiencies (above 7.2 kJ/kg) were not remarkably increased. / Avec le but de réduire l'écart de température entre la température de fusion de l'aluminium et la température de décomposition de l'agent moussant (hydrure de titane) utilisé lors de la fabrication de mousses d'aluminium et d'étudier cet effet lors de tests mécanique en compression, un revêtement métallique déposé sur l'agent moussant a été utilisé pour former une barrière de diffusion de l`hydrogène. De plus, la distribution granulométrique des particules de l'hydrure de titane fut utilisée comme variable pour contrôler la modification de la température d'évolution de l'hydrogène. / La barrière de diffusion de nickel a été déposée en utilisant la technique de dépôt autocatalytique et la modification de la désorption de l'hydrogène des poudres avec le revêtement a été mesurée par thermogravimétrie. L'effet de la taille des particules a été déterminé en utilisant des poudres commerciales séparées en 5 classes granulométriques bien définies. Les précurseurs aux mousses ont été obtenus via la compression uniaxe d'un mélange de poudre d'aluminium contenant 1% poids de hydrure de titane, et le moussage subséquent fut effectué à 750 et 800°C. Le comportement mécanique des mousses a été mesuré par compression sur des échantillons cylindriques machinés. / Le revêtement de nickel a produit des dépôts uniforme et homogène contenant 96.5% poids en nickel. La présence du revêtement a augmenté avec succès la température de dégagement de l'hydrogène et réduit la disparité initiale de la température d'environ 70°C. De plus, l'adhérence du revêtement fut suffisante pour résister aux étapes de mélange des poudres et de compaction. Les mousses d'aluminium faites à partir des particules revêtues de nickel montrent une distribution des pores plus uniforme et une meilleure reproductibilité du contrôle de la porosité par rapport aux mousses fabriquées avec les poudres sans revêtement. Il fut aussi démontré que la température de dégagement de l'hydrogène du l'hydrure de titane augmente lorsque la taille de particules augmente. La taille des particules a influencé le niveau d'expansion et de porosité. Les mousses fabriquées avec des poudres plus grossières possédaient une distribution de taille des pores plus uniforme. Les essais de compressions sur les mousses contenant du nickel ont montré un régime d'absorption d'énergie quasi-horizontale avec une déformation allongée. Par contre, le coefficient de rendement d'absorption de l'énergie (plus grand que 7.2 kJ/kg) n'a pas été remarquablement augmenté.