Collector interaction with mineral surfaces has long been studied. Little work has been done, however, on directly quantifying reagent adsorption, certainly under industrial process conditions. The use of a novel surface analysis technique, Headspace Analysis Gas-phase Infrared Spectroscopy (HAGIS), is suggested for quantification of adsorbed reagents in mineral processing. / As a first exercise, a test system of xanthate adsorption onto lead sulphide minerals was studied. A survey of possible calibration standards (pure xanthate, a synthetic lead-xanthate, galena (PbS) and a lead sulphide ore conditioned with xanthate) resulted in linear curves for all four cases. The quantification of isopropyl xanthate adsorption onto batch flotation products (concentrate and tail) was used to determine that ore standards gave the most accurate results. / The technique was also tested for quantification of adsorbed amines. Two collectors, dodecylamine and diphenylguanidine, and a depressant, triethylenetetramine, were studied. A common calibration curve was prepared using diphenylguanidine adsorbed on Inco matte. Results show that the HAGIS technique can easily be used to quantify adsorbed amines. / It is concluded that the HAGIS technique is a powerful new tool for the quantitative determination of adsorbed reagents. The xanthate study showed the use of ores as standards produces the best calibration. The amine study introduced the possibility of analyzing reagent mixtures.

Among the apparently simplest methods to determine bubble size in flotation systems are photographic techniques, ranging from photography through transparent walls to imaging of extracted bubbles. All capture images, which to varying degrees include overlapping, touching or out of focus bubbles. As manual counting limits the total number of bubbles, image analysis software is used to automate the process. Accuracy is thus dependent on image treatment, including counting method and filters. / The McGill bubble size analysis method yields single plane, backlit images and utilises software that filters by shape factor. Proven effective for bubble size distributions ranging from approximately 0.5 to 3 mm, regular trends are observed when number (D10) and Sauter (D32) mean diameters are compared. When the method was extended to wide distributions typical of jetting spargers (e.g., 0.2--15 mm), no similar trends were evident. Revision of the analysis process for these two-phase systems included counting by number of holes, which reduced dependence on bubble shape. This allowed for inclusion of small and large bubbles, while excluding bubble clusters. A diameter assignment protocol reflecting individual bubble shape was also developed. Revised output distributions showed increased symmetry, and the D32 vs. D10 trend was recovered. / Impact of sample tube diameter on the output bubble size distributions, and types of bias introduced were also investigated. A means of selecting an appropriate sample tube diameter for a given bubble population is presented.

The effects of Ni, Pd and Pd77Ag23 alloy membrane microstructures on hydrogen permeation have been investigated using the electrochemical permeation method. Deformation, annealing and electrodeposition have been used to modify the membrane microstructure. / The delay for hydrogen permeation in palladium is increased when the degree of deformation increases. The annealing of deformed palladium and Pd 77Ag23 at 250ºC or 850ºC accelerates hydrogen diffusion. The annealed metals release hydrogen more rapidly than the deformed metals. The Pd77Ag23 alloy needs more time to release hydrogen. The solubility increases in both palladium and Pd77Ag23 when deformation increases. A hydrogen-trapping model has been developed, which allows the prediction of trends in the experimental data. / The nickel electrodeposits exhibit fiber texture with two components (100) and (110). Low current densities yield a dominant (100) texture. High current densities develop a strong (110) fiber texture. There is a texture transition region between 30 and 50A/dM2. After annealing at 800ºC for lh, the (100) texture is transformed to (111) texture. Diffusion coefficients increase as texture is transformed from (001) to (011) and (111). The diffusion coefficient of a sample without a dominant texture is lower than the values measured for the samples with (001), (110) and (111) textures. Diffusion coefficients of (001) and (011) polycrystalline membranes are higher than (001) and (011) single crystal membranes. However, in both cases the (011) diffusion coefficient is greater than the (001). By using absorption energies of (001), (011) and (111), the three basic planes in nickel, adsorption energies for other orientations have been estimated using the weighted average method. Absorption energies for different orientations have been shown in the form of an inverse pole figure. / The diffusivity of hydrogen increases six times from single crystal membranes to polycrystalline membranes, and increases six times again from polycrystalline membranes to nanocrystalline membranes. When nanocrystalline nickel membranes have been annealed between 200ºC and 900ºC, they yield a texture change in the nickel membrane from (100) fiber texture to a mixture of (100) fiber texture and (111) texture, and then only one recrystallization (111) texture component is observed. During annealing from 200ºC to 900ºC, a slight increase in the number of high angle (45º--65º) grain boundaries occurs and a percentage of the low (0º--15º) and middle angle (30º--45º) grain boundaries decreases slightly. After annealing at 800ºC and 900ºC, changes in the grain size do not affect the rate of diffusion of hydrogen. / There are four major microstructural factors: deformation defects, grain size, texture and grain misorientation, which affect hydrogen permeation in metals. In the polycrystalline materials studied, the importance of crystal defects, grain size, texture and grain misorientation on hydrogen permeation has been established. / In the developed multilayer membranes when hydrogen charging is from the nano-Ni side of the nano-Ni and poly-Ni composite membrane, the permeation current rises rapidly. When hydrogen charging is from the poly-Ni side of the same composite membrane, the permeation current gradually rises and it takes longer to reach a steady state. The permeability of the nano-poly-Ni membranes charging from the nano-Ni side is eight times higher than that of the same composite membrane charging from the poly-Ni side. The diffusivity of the nano-poly-Ni membranes charging from the nano-Ni side is two times higher than that of the same composite membrane charging from the poly-Ni side. The nano-poly-Ni membranes charging from the nano-Ni side release hydrogen faster than the same composite membrane charging from the opposite side. The diffusivity and permeability of the nano-poly-Ni membrane are smaller than those for a single nano-Ni membrane, but larger than those for a single poly-Ni membrane. The hydrogen permeation model for a bilayer membrane is built to simulate diffusion in the nano-Ni layer and the poly-Ni layer. The model can generate concentration profiles of hydrogen. The experimental data are in good qualitative agreement with the model.

Five steels were used in the present work to investigate the dynamic, static and metadynamic recrystallization behavior of hypereutectoid steels of 1% carbon, alloyed with high silicon and microalloying levels of vanadium. Compression tests were performed using single and double hit schedules at temperatures between 875--1100ºC, strain rates of 0.01s-1 -1s-1, and inter-pass times of 0.1--500 seconds and 0.1--30 seconds for static recrystallization and metadynamic recrystallization, respectively. For dynamic recrystallization, it was found that an increase in carbon and vanadium content led to smaller grain sizes. Equations were generated that can be used to predict the critical strain for dynamic recrystallization. Of interest is the finding that there is an activation energy for deformation specifically associated with dynamic recrystallization (i.e. peak strain). This activation energy associated with the peak strain is lower than that associated with the steady state stress. This is contrary to Sellar's original observation that the peak strain is a function of the activation energy for deformation according to the Zener-Hollomon relationship. The static recrystallization kinetics were measured, and the appearance of 'plateau' regions where softening is arrested for a period of time was used to quantify the kinetics of strain-induced precipitation and generate the precipitation temperature time diagrams for the three steels. Despite the wide range of V levels, the steels exhibited very similar precipitation kinetics at all temperatures tested, being particularly close below 950ºC. A kinetic equation for static recrystallization is proposed which takes the V and Si concentrations into account. The metadynamic recrystallization results show that there is a transition strain region in which both static and metadynamic recrystallization take place during the inter-pass time. The results also revealed that V and Si have a strong solute dra

Thesis (Ph. D.)--University OF MICHIGAN.

DISSERTATION (PH.D.)--THE UNIVERSITY OF MICHIGAN

No description available.

No description available.

No description available.

No description available.