No description available.

No description available.

High manganese steels offer an outstanding combination of high strength and ductility owing to their high sustained strain hardening rates. The strain-induced deformation products in these steels (mechanical twins and ε-martensite) increase the work hardening rates by acting as barriers for dislocation motion. A significant determinant of the deformation products in these steels is the value of stacking fault energy (SFE), which is in turn strongly dependent on the alloy manganese and carbon contents. The main objective of present work is to investigate the microstructural evolution and mechanical behaviour of both homogenous and compositionally graded high-Mn steels, where C compositional gradients were introduced into the latter. The starting materials in this work were cold rolled Fe-22Mn-0.6C and Fe-30Mn-0.6C steels. For both starting alloys, decarburization and annealing heat-treatments were used to obtain four homogenous alloys with carbon contents of approximately 0, 0.2, 0.4 and 0.6 wt. % with similar grain sizes. Decarburization heat treatments were also applied to obtain three carbon graded Fe-22Mn-C alloys (G1, G2, G3) and one carbon graded Fe-30Mn-C alloy. Microstructural observations determined the deformation products to be mechanical ε-martensite for the 22Mn-0C and 22Mn-0.2C alloys and mechanical twins for the 22Mn-0.6C, 30Mn-0.2C, 30Mn-0.4C and 30Mn-0.6C alloys. For the 22Mn-0.4C and 30Mn-0C alloys, both mechanical twins and ε-martensite were observed during deformation. For all the carbon graded Fe-22Mn-C alloys, the dominant deformation products changed from mechanical ε-martensite at the near-surface layer to a mixture of mechanical twins and ε-martensite or mechanical twins only at the inner cross-section layers. In the case of carbon graded Fe-30Mn-C alloy, the deformation products altered from a combination of mechanical ε-martensite and twins at the near-surface layer to mechanical twins at the inner cross-section layers. For all the homogenous alloys, the ultimate tensile strength and uniform elongation increased with increasing alloy carbon content. The work hardening behaviour of these steels was successfully modelled using a modified Kocks-Mecking model, in which the work hardening was the sum of the dislocation glide contribution and the phase transition contribution – mechanical twinning and/or mechanical ε-martensite formation – as dictated by the formation kinetics of both deformation products. For both alloy systems, the mechanical properties of the carbon graded alloys were not as good as the monolithic 22Mn-0.6C and 30Mn-0.6C alloys due to their lower sustained high work hardening rates. Both the mechanical ε-martensite and twin formation were found to follow a sigmoidal kinetic with strain. In the case of twin formation homogenous alloys, the saturated volume fraction of twins was directly proportional to the alloy SFE. For the ε-martensite formation homogenous alloys, the ε-martensite volume fraction at fracture was found to be strongly dependent on alloy SFE, where it declined sigmoidally with increasing alloy SFE. It was also found that the ε-martensite volume fraction at fracture – approximately 0.7 – was independent of SFE for SFE  6 mJ/m2. This indicated that the critical damage mechanism was determined by the kinetics of the ε-martensite formation, which was in turn dictated by the alloy SFE. Finally, it was found that the stress for the onset of mechanical twinning – and consequent increase in the work hardening rate – for the higher SFE, twinning dominated alloys was linearly proportional to the alloy SFE. / Thesis / Doctor of Philosophy (PhD)

No description available.

No description available.

No description available.

No description available.

No description available.

Mn is a geochemically important element that contributes significantly to the cycling of heavy metals. During precipitation, Mn oxyhydroxides scavenge many heavy metals, including Pb, in a variety of natural environments. Because of this phenomenon, the precipitation of Mn oxyhydroxides may provide a remediation technique for removing Pb from contaminated aqueous solutions. Therefore, this study was undertaken to provide a quantitative understanding of the coprecipitation of Pb with Mn oxyhydroxides to demonstrate their capacity to remove Pb permanently from contaminated solutions. To accomplish this, a series of factorial experiments with varying initial Mn and Pb concentrations were run in the presence of a borate buffer or a bicarbonate buffer. All experiments were run in batch reactors, in the presence of a quartz substrate, at 25 degrees celcius, at pH 8.5, and were continuously stirred. Initial Mn and Pb concentrations were varied by half log units from 100 to 0 mg/L and from 3 to 0 mg/L, respectively. Solutions were analyzed for Mn using the formaldioxime colorimetric method and for Pb using AA. Precipitates on quartz surfaces were analyzed by SEM, XPS, and XRD for precipitate identification and morphology. The amount of Mn and Pb associated with the quartz sand was determined by dissolving the precipitates from selected quartz samples using concentrated nitric acid. Finally, a different set of precipitate-coated quartz grains were leached in pH 5 acetic acid solution to assess the metal retention capacity of the precipitated material. Mn oxyhydroxides precipitated onto the quartz sand in both the borate and bicarbonate buffered experiments. SEM and XPS data revealed tiny crystallites in etch pits on the quartz surfaces that contained predominantly Mn3+. XRD analysis did not produce an X-ray pattern for these Mn oxyhydroxides but did identify the suspended Pb precipitates as hydrocerrusite and Pb(HBO3)2 in the borate buffered experiments and hydrocerrusite in the bicarbonate buffered experiments. Much more Mn and Pb are associated with the quartz surfaces in the borate buffered experiments, but no Pb was associated with quartz surfaces initially (< 6 hrs. of reaction time). Leaching of precipitates resulted in extracted Mn in both experiments but Pb was extracted in only the bicarbonate buffered experiments. The Mn precipitation rate was greater in the borate buffered experiments and higher initial Mn and Pb concentrations appear to increase the precipitation rate in both sets of experiments. These results indicate that Mn oxyhydroxides nucleated onto suspended Pb precipitates. The growing Mn oxyhydroxide particles were attracted to the quartz sand, carrying along the Pb precipitates. Further precipitation of Mn oxyhydroxides on the quartz surfaces trapped the Pb. This process was much more significant in the borate buffered experiments where much more Mn precipitated. The greater amount of Mn oxyhydroxide growth acts as a barrier protecting the Pb from the pH 5 leaching solutions. As a result, Pb was retained by the sand grains from the borate buffered experiments during leaching while significant amounts of Pb (35-100%) was extracted from the sand produced by the bicarbonate experiments. These results strongly suggest that coprecipitation of Pb with Mn oxyhydroxides in the presence of a borate buffer and a quartz substrate may be a remediation tool for Pb contaminated aqueous solutions. Not only will this process remove aqueous Pb2+ from solution but it appears it will also substantially incorporate colloidal Pb particles as well. / Master of Science

Soluble manganese (Mn) is a common contaminant in drinking water sources. High levels of Mn can lead to aesthetic water quality problems, necessitating removal of Mn during treatment to minimize consumer complaints. Mn may be removed during granular media filtration by the "natural greensand effect," in which soluble Mn adsorbs to manganese oxide-coated (MnOx(s)) media and is then oxidized by chlorine, forming more manganese oxide. This research builds on a previous model developed by Merkle et al. (1997) by either neglecting the empirically determined available fraction of sorption sites (referred to as the "simple" model), which took into account the fact that some adsorption sites in the porous media were inaccessible, or by explicitly accounting for the transport and reaction processes within the porous structure of the MnOx(s) coating (referred to as the "mechanistic" model). Both models were applied to experimental data and used to evaluate the oxidation rate constant, which was the only unknown parameter. An inverse relationship between the fitted reaction rate constant and chlorine concentration was observed, showing that the oxidation reaction does not depend on chlorine concentration for the experimental conditions considered. In a sensitivity analysis, the adsorption isotherm and reaction rate were found to have the greatest impact on predicted Mn removal. The simple model should prove useful for designing contactor units for manganese removal, provided its limitations are clearly understood, while the mechanistic model should be able to resolve differences in the various types of oxide coating (internal porosity, surface area and coating thickness) and will allow a more fundamental and mechanistically-consistent evaluation of the appropriate form of the oxidation rate expression. However, further research is needed to more completely characterize the adsorption and reaction mechanisms over the range of conditions commonly encountered in water treatment plants. / Master of Science