In this study a technique--that almost completely eliminates shrinkage microporosity--has been developed which uses small additions of strontium together with risers and chills to produce porosity-free AZ91C (Mg/Al/Zn) magnesium casting alloy castings. With the optimum level of Sr addition (0.01% to 0.02% Sr), shrinkage microporosity was removed from the casting and concentrated in the riser. The castings were already grain-refined by carbon inoculation. The effect of strontium is explained by a further reduction in the grain size of the castings from 250$ mu$m to 120$ mu$m. This effect produces a much denser casting due to increased mass feeding. In addition, a slow grain growth rate, caused by Sr addition, may keep the liquid intergranular-channel radius large during the final stages of solidification, and results in improved intergranular feeding. Another effect of strontium addition may possibly be attributed to a decrease in the surface tension and viscosity of the liquid AZ91C alloy, which in turn increases capillary feeding of the liquid metal. (Abstract shortened by UMI.)

Iron and silicon are the major impurities in the 1000 series of wrought aluminum alloys. As the aluminum is recycled, it picks up more and more of these impurities which cannot economically be refined out of the aluminum. When the concentration of these impurities reaches a certain limit (maximum limit in 1000 series is 1 weight percent (Fe+Si)), the aluminum must be downgraded. The Fe and Si form brittle intermetallic phases in these alloys. The two main phases are the plate-like $ beta$-AlFeSi (Al$ sb5$FeSi) and $ alpha$-AlFeSi (Al$ sb8$Fe$ sb2$Si) which has a Chinese Script morphology. The mechanical properties of these alloys are believed to depend largely on the nature of these intermetallics. / In the first part of this study, the conditions at which the intermetallics form, along with the ability of strontium to modify them were investigated. The second part consisted of determining how the morphology of the Al-Fe-Si phases affects the mechanical properties of the worked product. It was found that the formation of the Chinese Script morphology is promoted with increasing cooling rates, Fe/Si ratios and additions of strontium. However, the relative amount of Chinese Script was found to decrease with increasing (Fe+Si) levels. Tensile testing and formability testing (Erichsen ball punch deformation test) revealed that the presence of a Chinese Script morphology of Al-Fe-Si intermetallics (as opposed to the plate-like morphology) imparts no significant beneficial effect on the formability of the final rolled sheet.

The solution chemistry and statistical modelling of the O$ sb2$-CaCl$ sb2$-HCl leaching of a high-grade arseniferrous uranium ore (Key Lake, N. Saskatchewan) were studied with the objective of optimizing the co-extraction of uranium and radium. To this end, two types of experiments were carried out: (a) those designed to study arsenic speciation under simulated U-leaching conditions, with the use of model arsenide minerals, namely niccolite (NiAs), rammelsbergite (NiAs$ sb2$) and gersdorffite (NiAsS); and (b) statistically designed tests involving actual leaching of samples of the Key Lake ore. / Niccolite and rammelsbergite were found to yield H$ sb3$AsO$ sb3$ in solution as oppossed to gersdorffite, which was found to yield H$ sb3$AsO$ sb4$. H$ sb3$AsO$ sb3$ was found to be metastable in the Fe$ sb2$(SO$ sb4) sb3$ leaching system, while in the FeCl$ sb3$ and pressure (O$ sb2$) leaching systems, further oxidation to H$ sb3$AsO$ sb4$ was observed to take place at relatively slow rates. A mechanism has been advanced to account for the latter slow oxidation of H$ sb3$AsO$ sb3$ to H$ sb3$AsO$ sb4$. Finally, it was concluded that the formation of As(III)/As(V) in oxidative acid leaching is independent of the prevailing suspension redox potential which, in turn, renders impractical the possible stabilization of As(III) via potential control. / The solubility of RaSO$ sb4$ in HCl-MeCl$ sb{ rm x}$ leaching media was quantitatively established by employing the "reduced activity coefficient" method. The estimated solubility profiles clearly demonstrated the effectiveness of CaCl$ sb2$ in solubilizing radium as oppossed to other monochloride salt additions. / Finally, optimization of the co-extraction of U and Ra was attained with the use of a computer-aided response surface methodology. An incomplete three-level, four-variable factorial design (Box-Behnken) was applied in the region: 0.33 kg/L $ leq$ d$ sb{ rm s}$ $ leq$ 0.99 kg/L; 0.5N $ leq$ (HCl) $ sb{ rm o}$ $ leq$ 1.5 N; 1.5 M $ leq$ (CaCl$ sb2 rbrack sb{ rm o}$ $ leq$ 3.5 M; and 2 hrs $ leq$ Time$ leq$ 6 hrs. Temperature and oxygen pressure were kept constant at 68$ sp circ$C and 510 kPa (75 psi), respectively. Uranium and radium were found to dissolve simultaneously and rapidly (99% co-extraction in 2 hrs). Effective reduction of radioactivity levels from 10,000 pCi/g Ra$ sp{226}$ down to 160 pCi/g has been achieved in a single-stage batch process, thus making possible the production of environmentally innocuous tailings.

Nickel constitutes the major impurity encountered during copper electrorefining (Cu ER) operations. Continuous nickel build-up in the electrolyte is controlled by withdrawing a fraction of solution from the tankhouse and separate nickel as crude NiSO4.2H2O via Evaporative Crystallisation . The energy required to evaporate water is high, increasing thus the operation costs; moreover, the product thus obtained is of poor quality in terms of purity and crystal size. / As part of an ongoing research project launched at McGill University, the Solvent Displacement Crystallisation (SDC) technique is investigated as an attractive alternative to the conventional method. SDC involves the addition of low-boiling point, watermiscible organic solvents (MOs) to aqueous solutions to cause salt precipitation based on the "salting out" effect. The crystals are separated by filtration whereas the solvent is subsequently recovered for reuse by low-temperature distillation. / The present work describes the successful application of SDC method to the precipitation of NiSO4.6H2O from Cu ER spent electrolytes. Tests have been performed on synthetic and industrial electrolytes (courtesy of Canadian Copper Refineries), using Isopropanol as precipitant. By maintaining a low supersaturation (i.e. controlled addition of the solvent to the electrolyte) and heterogeneous crystallisation conditions (use of seed/product recycling), crystal growth is favoured while impurity uptake/contamination is minimised.

This research investigates the possibility of recovering the metallic content of the dust generated by stainless steelmaking operations in the Electric Arc Furnace (EAF) at ATLAS Stainless Steels by recycling it to the EAF. The physical and chemical properties of the dust were determined. The pelletization of dust-carbon fines was optimized in order to obtain strong pellets capable of carbothermic self-reduction. The high temperature behavior of dust-carbon pellets was investigated in a Thermogravimetric Analyzer (TGA) coupled with a Fourier-Transform Infrared Spectrometer (FTIR). The mechanisms involved in metal recovery were identified. The feasibility of direct recycling was verified by testing pellets in a scaled-down simulation of the EAF process. As a result of this research practical recommendations concerning the application of the direct recycling process at ATLAS were formulated.

A 1/12 scale tank was constructed to physically simulate the metallurgical skimming process in order to gain a better insight of the skimming of fluid slag from copper anode furnaces. 1-decanol and water were used to represent the slag and the underlying liquid metal respectively. It was observed that slag skimming was comprised of two steps, a transportation step to carry slag to the skimming lip, and a removal step to move the slag across the lip. The efficiency of the transportation by bath surface circulation was strongly affected by the method of creation of the surface movement. Submerged gas injection failed to carry slag to the lip effectively because of the uncontrollable circulation pattern generated on the bath surface. On the other hand, three impinging gas jets in glancing contact with the bath surface were able to carry slag to the lip effectively. The Re of the jets (I.D. 0.01m) inclined 50° to the horizontal was 6100. The jet configuration was two of the jets were placed close and parallel to the rear wall of the model and faced each other. The third jet was parallel to the model short axis and was directed at the skimming mouth. With the same amount of time, 70% of the slag originally charged to the tank was skimmed with the assistance from the impinging gas jets in comparison to 20% when no jet was used. Detailed analysis of the skimming process revealed that the underlying liquid bath level continuously decreased as the slag was skimmed and resulted in continuous reduction of the skimming rate. In order to compensate the reduction of skimming rate due to bath level reduction, a control strategy was proposed for rotating the furnace continuously to maximize the skimming effectively.

Minor element (As, Sb, Bi, Pb, Ni) behavior, control and removal techniques in the conventional copper smelting/electrorefining process was studied. The analysis was based on the information collected from an exhaustive literature review, the visit of 23 smelters, 14 electrorefineries and consultations at 15 different institutes specialized in the field. / Data was collected for several types of matte smelting furnaces such as Outokumpu flash, INCO flash, Mitsubishi S-furnace, Teniente Converter, Noranda Process Reactor, Isasmelt and reverberatory. Behavior in Peirce-Smith converting furnaces, Mitsubishi C-furnace, Noranda Continuous Converter and Kennecott-Outokumpu Flash Converting Furnace was also discussed. / The effect of various matte smelting furnace operating parameters such as matte grade, oxygen enrichment, concentration in feed, other minor constituents and temperature on minor element partition to gas and distribution coefficient (wt% matte/wt% slag) was analyzed theoretically and validated with industrial data when possible. Because fewer data were available, only a brief comparison between the elimination in traditional batch converting and new continuous converting processes was performed. The behavior of minor elements in electrorefining was described from a theoretical viewpoint. (Abstract shortened by UMI.)

Interactions among sulphide minerals (sphalerite, galena and pyrite) and sphalerite and silica were studied using electrophoresis, an automated settling apparatus and, in the sphalerite/silica case, atomic force microscopy (AFM). / Significant changes in zeta-potential were observed in some cases as a result of contact (conditioning) with a second mineral. As a function of pH when sphalerite was conditioned with pyrite, the zeta-potential increased to ca. pH 9 and decreased above this pH. For sphalerite conditioned with galena, the zeta-potential decreased below pH 5, increased between pH 5 to 9 and decreased above pH 9. Galena conditioned with sphalerite and pyrite resulted in a decrease in zeta-potential below pH 5 and an increase above this pH. The results are interpreted on the basis of galvanic interactions. The mineral with the lower rest potential preferentially oxidizes and the ions released migrate and influence the zeta-potential. Measurements made in solutions of various suspected released ions generally supported the galvanic model. The zeta-potential results for mixed sulphide minerals were correlated with settling velocity. For the pyrite/galena, sphalerite/pyrite and sphalerite/galena systems, the pH of maximum settling rate corresponded to the zeta-potential of both minerals approaching zero. When the minerals were oppositely charged the conditions remained (relatively) dispersing. In the case of silica and sphalerite/silica at around pH 2 and 8.5 sphalerite homocoagulated and some silica was dispersed; from ca. pH 3 to 7, the system heterocoagulated; and above pH 9.5 the suspension was dispersed. This behavior did not correlate with the behaviour of the minerals alone. In the presence of calcium ions, from ca. pH 2 to 7, the sphalerite homocoagulated and the silica was dispersed while above pH 7 heterocoagulation was observed. This behavior did correlate more closely with that of minerals alone. The zeta-potential measurements provided only a partial interpretation: a force of attraction between sphalerite and silica appears to exist at moderately alkaline pH. An electrostatic origin to the attractive force due to the large difference in zeta-potential for sphalerite and si / The settling results for the sphalerite/silica system were Correlated with AFM measurements. For silica alone, the interaction was repulsive at pH 6.2 and 8.3. This behavior did correlate with the settling tests. In the presence of calcium ions, the interaction was repulsive at pH 6.1 and 9.6 and was attractive at pH 11.4. This behavior also correlated with the settling data. For the mixed system, the interaction was repulsive at pH 7.2, 8.4 and 10.6 and was attractive at pH 9.6. The behavior at pH 8.4 and 10.6 correlated with the settling tests but not at pH 7.2 and 9.6. The lack of correlation is discussed.

In this research the dynamics of a non-reactive and a reactive gas bubble under various pressure fields are studied. Comprehensive mathematical models have been developed to simulate the dynamic behavior of different gas bubbles. On the basis of these models several features of thermo-fluid mechanical behavior of gas bubbles are then discussed and their applications for the metals processing operations are described. The mathematical models form a set of coupled, highly nonlinear and stiff differential equations, which have been solved numerically by a modified Gear method. / The first part of this study is concerned with the transient dynamic behavior of a hydrogen gas bubble in a solidifying aluminium-3.4 wt pct copper alloy melt under various ultrasonic pressure fields. During the process of bubble collapse, the melt pressure surrounding the bubble is seen to increase very rapidly. The variations of pressure and supercooling in the melt surrounding the bubble can cause bulk crystallization. If the pressure in the vicinity of the dendrites exceeds a threshold value, dendrite fracturing takes place. Dendrite fragments become nuclei during metal crystallization in an ultrasonic field, which lead to the refined crystalline structure of the metal. The results show that adjacent to the bubble surface, the peak pressure generated in the melt is in the order of several hundreds to thousands atmospheres depending on the initial bubble size, pressure of undisturbed melt and the ultrasonic's specifications. / The second part of this study is related to the dynamics of a stable bubble under various ultrasonic pressure fields. When the imposed pressure field is beyond a threshold value, dissolved gas in the liquid flows into the gas bubbles by rectified diffusion. In this case, the bubbles grow sufficiently to float to the surface due to the hydrodynamic buoyancy force. The results show that the air bubble grows and the time-averaged bubble volume reaches several times of its initial value when the ultrasonic pressure amplitude is more than the threshold pressure. / The final part of this study consists of the complex dynamic behavior of a reactive gas bubble immersed in a nonreactive liquid under various liquid impulse pressures. Numerical investigations are conducted into the collapse and explosion of an isolated oxygen-hydrogen bubble immersed in water and in glycerin. The results show that, if the imposed pressure field is strong enough, the bubble is seen to explode. (Abstract shortened by UMI.)

A comprehensive methodology to physically and numerically model upsetting in cold heading was developed. / The physical model was a Drop Weight Test with a guided pocket-die set capable of approximating industrial cold heading conditions. The results show that the test is sensitive to the critical parameters for cold heading. These include surface quality, residual element level, nitrogen content, microstructure, decarburization, and specimen geometry. The test is capable of assessing the fracture behavior of cold heading materials. / One goal of the study was to reveal differences in fracture behavior with varying steel sources. Accordingly, the matrix of test materials consisted of grade 1038 steels from three different steel sources. / Material preparation and conditioning of test materials approximated industrial procedures for cold heading materials. These procedures included hot rolling, controlled rod cooling, descaling, straightening, lime coating and lubricating, and wire drawing. Spheroidization of test specimens was performed in an industrial batch furnace using an industrial heat treatment cycle. / A finite element program (FEM) enabled the simulation of upsetting in cold heading. The inputs required to model the cold heading process include flow stress behavior and friction conditions representative of cold heading. These inputs were obtained using the CANMET Cam Plastometer and the Friction Ring Test. / The Cockcroft and Latham fracture constants for an as-rolled and a spheroidize annealed 1038 material were computed by FEM modeling and the critical values were calibrated using the Drop Weight Test. The fracture criterion constant was found to be independent of strain path for upsetting in cold heading and thus is material-related.