<p>The nature of reactions between sodium sulphide slag and carbon-saturated-iron and -copper alloys were studied to obtain a more clear understanding of slag/metal reactions as well as to enhance the knowledge with respect to decopperization and desulphurization reactions. A model has been developed for the kinetics of reactions, taking into consideration the chemical reactions at the gas/slag and metal/slag boundaries as well as diffusion of sodium vapour in the gas phase. The rate constants have been determined by comparing the experimental results with the proposed model of the system using a CDC 6400 computer. Some understanding of the thermodynamics of sulphide solutions, never studied before, was accomplished through calculations. Suggestions were made for the most efficient use of sulphide slags for removing copper and sulphur from carbon-saturated iron melts. It is hoped that this work will open a new area of research on the fundamentals of slag/metal reactions involving sulphides. Such investigations may be helpful in understanding the slag/metal systems involving oxides.</p> / Doctor of Philosophy (PhD)

<p>Experimental investigations of the attack of carbon-free and carbon-bearing magnesite refractories by synthetic steelmaking slags at 1600°C and relevant theoretical analysis are documented in this thesis.</p> <p>The attack of carbon-free magnesite brick by steelmaking slags was studied by immersing cylindrical specimens in molten slag contained in a noble metal crucible. The major variable of study was the composition of the slag, i.e., the alumina and magnesia contents and the lime-silica ratio.</p> <p>The reacted samples are examined macroscopically in terms of elongation, slag climb and penetration. Microscopic examination by petrographic techniques and microprobe analysis further illustrate the macroscopic observations associated with the chemical reactions taking place between the refractories and the penetrating slag.</p> <p>The role of carbon in extending the lining life of BOF carbon-bearing refractories is the main objective of this research. Test crucibles were reacted with slags. The reacted samples were microscopically studied. The formation and destruction of a dense MgO layer in reacted specimens was studied. Theoretical analysis of the formation of the dense MgO layer was undertaken based on the assumption, that the formation of MgO is the resul t of a gaseous reac.tion between magnesium vapor and carbon dibxide.</p> <p>This work has been carried out under conditions relevant to refractory problems involved in a few major steelmaking vessels. The results obtained here have shed light on the kinetics and mechanism of the attack of magnesite refractories by steelmaking slags.</p> / Doctor of Philosophy (PhD)

<p>An ironmaking process has been proposed in which composite iron ore-coal pellets are smelted to produce liquid iron. Studies have determined the manner in which individual composite pellets react, their compositional limitations and the chemical compositions of the resultant slag and iron.</p> <p>Laboratory reduction experiments were performed on various composite pellets bonded with bentonite, cement or calcium carbonate. Ore-coal pellets bonded with bentonite reacted to form a fayalite-type slag high in iron oxide content and an iron melt low in carbon (2%) but high in sulphur (0.3%) contents. Ore-coal pellets which were made self-fluxing by being bonded with cement or calcium carbonate produced a basic slag low in iron oxide content and an iron melt containing over 3% carbon and below 0.05% sulphur. The off-gas from reaction of ore-coal pellets was profound to be mostly carbon monoxide and hydrogen which in combustion with oxygen is to supply the heat required by the process.</p> <p>A computer model of the process was constructed using heat and mass balance equations which illustrated the effects of changes in process variables on input material requirements. Indications from a process analysis are considered to be positive for an ironmaking process based on composite pellet smelting. However, the ultimate success of the process will depend on whether sufficient heat can be retained within the reaction vessel where it is needed. To this end, hot-model "pilot plant" experimental trials are recommended.</p> / Doctor of Philosophy (PhD)

<p>The role of elastic, thermal and interfacial properties in the toughening of brittle particulate composites was studied. Three different systems were designed for these studies: 1) glass - partly oxidized Ni, where the thermal expansion coefficient of the glass is lower, equal to and higher than that of Ni; 2) glass - partly oxidized Al, in which both thermal expansion coefficients and elastic moduli are equal; and 3) the LiF-MgF₂ system, where the thermal expansion coefficient of LiF is 3.4 x that of MgF₂, and the elastic moduli of the two phases is similar. The first two systems were fabricated by mixing and hot-pressing the powders. In the LiF-MgF₂ system, second-phase particles were produced via a precipitation reaction.</p> <p>It was found that when the thermal expansion coefficient and elastic modulus of the second phase particles are higher than that of the glass matrix, a crack has a tendency to bypass the particles, propagating consistently in the matrix material. When the thermal and elastic stresses are eliminated and satisfactory bonding is achieved, a toughness increase up to 60 x that of the matrix results.</p> <p>In the LiF-MgF₂ composite, a 15 - 20 x toughness increase was achieved and is attributed to the precipitation of incoherent rod-shaped MgF₂ precipitates.</p> / Doctor of Philosophy (PhD)

<p>Before utilizing the two phase mixture of β- and α-AI₂O₃ as solid electrolytes in oxygen probes that can monitor very low oxygen potentials, the chemical stability of β-AI₂O₃ (Na₂O, 11Al₂O₃) was studied by determining the sodium oxide activity in α-AI₂O₃, β-Al₂O₃ coexistence using electrochemical cells of the type</p> <p>Pt⎮W(s), WS₂(s), Na₂S(s)⎮β- and α-Al₂O₃⎮M(s), MᵪO(s)⎮Pt</p> <p>where M stands for Cu, Ni or Fe. The over-all cell reaction for these cells can be written as</p> <p>W(s) + 2Na₂S(s) + MᵪO(s) = 2Na₂O(β) + WS₂(s) + xM(s)</p> <p>By considering the following reaction</p> <p>Na₂O (β) = 2Na(g) = ½O₂(g)</p> <p>the measured values of Na₂O activity in α-Al₂O₃, β-Al₂O₃ coexistence were used to evaluate the variation of sodium vapor pressure over this coexistence, PNa(α-β), with temperature and oxygen pressure. The results indicate that PNa (α-β) reaches one atmospheric pressure at 1076°C in environments with AI, Al₂O₃(α) imposed-oxygen potentials, and at 1834°C in atmospheres with Fe,FeO imposed-oxygen potentials.</p> <p>The activities of Na₂O in Al,Na,β-Al₂O₃ coexistence and β-Al₂O₃, β"-Al₂O₃ equilibria were also determined using the following cells</p> <p>Ta⎮Al(s or ℓ), Na(ℓ)⎮β-Al₂O₃⎮NiO(s),Ni(s)⎮Pt⎮Ta,</p> <p>and</p> <p>Pt⎮W(s), WS₂(s), Na₂S(s)⎮β- and β"-Al₂O₃⎮Ni(s), NiO(s)⎮Pt,</p> <p>respectively. The results obtained as well as the available thermodynamic data in the literature for the Na-AI-O system were used to partially represent the equilibrium oxygen pressure diagram of Na-Al-O system at 1000K.</p> <p>To establish the reliability of using the two phase mixture of β- and α-Al₂O₃ as a solid electrolyte in oxygen probes, galvanic cells using solid electrolyte tubes fabricated from this mixture with electrodes of fixed oxygen chemical potentials were developed. Experimental results indicated that this solid electrolyte responded reversibly to oxygen potentials as high as those of Cu, Cu₂O equilibria between 600°C and 1000°C, and as low as those of Al-α-Al₂O₃, β-Al₂O₃ coexistence between 550°C and 800°C.</p> <p>The thermodynamics parameters and phase compositions in the Ni-Al-O and Fe-Al-O systems at temperatures in the range 850°C-1150°C have been also investigated using electrochemical technique, electron microprobe analyses, and X-ray powder diffraction studies. Results showed that, about one ppm of Al in Ni or Fe is sufficient to stabilize α-AI₂O₃. Electrochemical cells with calcia stabilized zirconia as solid electrolyte and working electrodes of the type Ni (or Fe)-AI alloy Al₂O₃(α) had shown electrical instability which was not amenable to correction by coulometric disturbance or by temperature cycling. When β- and α-Al₂O₃ solid electrolyte was used in combination with Ni (or Fe)-AI,AI₂O₃(α) electrodes, steady and reproducible potentials were measured at 940°C which enabled the equilibrium oxygen pressures over these electrodes to be calculated. The values obtained for the oxygen pressures were outside the electrolytic domain of calcia stabilized zirconia which explain the instability problems of cells with this electrolyte.</p> <p>Quantitative results have been obtained for the phase compositions in the ternary systems Ni-AI-O, and Fe-AI-O. These resuIts have been used to construct the 1000°C Ni-AI-O isotherm, and to modify and extend the NiO-Al₂O₃ quasibinary system given previously in the literature. The equilibrium oxygen pressure diagrams of the Ni-AI-O and Fe-AI-O systems have been constructed at 1000°C based on the results of this investigation as well as the available data in the literature.</p> / Doctor of Philosophy (PhD)

<p>An understanding of the chemical reaction kinetics at a slag/metal interface requires knowledge of both transport and interfacial properties. However, fundamental interfacial reaction rate data are largely unavailable. Consequently, the ion exchange current density and interfacial capacitance have been measured with the double pulse galvanostatic method for the anodic dissolution of an iron electrode in liquid CaF(,2) and CaF(,2)-CaO slags. Both the exchange current and capacitance are strongly increasing functions of concentration. The measured exchange current densities are large (ranging from 4.5 to 400 amperes/cm('2)), so that a non-linear extrapolation of the experimental data is necessary. The kinetics of oxygen transfer at the Fe/CaF(,2) interface was studied in independent experiments. The reaction was driven by an FeO activity gradient. The kinetic information obtained in this manner compares favourably with that derived from the electrochemical technique. With suitable information on the relative magnitudes of transport and reaction rates, it is possible to predict the kinetic regime for the entire system (i.e. transport, reaction, or mixed control). The experimental data indicate that mixed control of the reaction kinetics will prevail at the Fe/CaF(,2) interface in the absence of forced stirring. If the kinetic parameters in more complex metallurgical systems are similar to those measured at the Fe/CaF(,2) interface, the influence of interfacial charge transfer reactions may be greater than is normally supposed.</p> / Doctor of Philosophy (PhD)

<p>The austenite-niobium carbonitride equilibrium has been accurately characterized between 1000°C and 1250°C for commercial steel compositions. A closed capsule technique using hydrogen as carrier gas has been employed to equilibrate Fe-Nb alloys with Fe-C alloys and determine the Υ/Υ+NbCx phase boundary in the temperature range 1000-1250°C. The results have been used to define the solubility of niobium carbide in austenite with improved accuracy. Solubility limits have also been determined by chemical analysis of carbides extracted from a laboratory melted steel. A defect carbide NbC₀.₉ and a carbonitride NbC₀.₉₁N₀.₀₄ have been identified as the equilibrium precipitate compositions at 1000°C in two steels.</p> <p>To test the data for thermodynamic consistency detailed analysis of the thermochemical data of the defect carbides of niobium has been made. The thermodynamics of the binary Fe-C, Fe-Nb and Nb-C systems have been utilized to determine the Υ/Υ+NbCx equilibrium over relevant composition and temperature ranges and this is in good agreement with the experiments. Solubility product relations are given for the carbide NbC0.87 as this carbide composition has been shown to be a justifiable average equilibrium composition over the temperature range in this study. The carbide solubility relations have been used along with the reported carbonitride solubilities, to estimate the solubility relation for a metastable cubic nitride of composition NbN₀.₈₇.</p> <p>An accurate method for calculating the composition and solubility of nonstoichiometric carbonitride precipitates in steel has been devised using the above carbide and nitride solubility relations and treating the carbonitrides as compounds with a fixed nonstoichiometry NbCxN(₀.₈₇-x). The effects of Mn, Si, Cr, Ni and Mo on the solubilities of carbonitrides have also been evaluated and it is shown that the carbonitride solubility in a multi-component austenite can be satisfactory predicted. It is therefore possible to calculate, for example, precipitate fractions and theoretical maximum temperatures required in slab reheating to completely dissolve precipitates.</p> / Master of Engineering (ME)

<p>A high temperature electrochemical technique has been developed, to continuously monitor the oxygen activity in iron based melts.</p> <p>The electrochemical cell used, can be represented as:</p> <p>Mo | Cr(s) -Cr₂O₃(s) | Calcia stabilised zirconia | Fe-M-O(ℓ) |Mo-ZrO₂, Mo cermet</p> <p>The electrolyte was in the form of a crucible to hold the metallic melt.</p> <p>The standard free energy change for the reactions:</p> <p>FeNb₂O₆(s) ⇆ 2NbO₂(s) + Fe(ℓ) + 2O lw/o</p> <p>FeNb₂O₄(s) ⇆ Al₂O₃(s) + Fe(ℓ) + O lw/o</p> <p>and the influence of manganese content of the melt on the oxygen activity for the equilibrium:</p> <p>(Fe1-x, Mnx)O.Al₂O₃(s) ⇆ Al₂O₃(s) + (1-x)Fe(ℓ) + xMn lw/o + O lw/o</p> <p>were determined using the technique at a temperature range of 1550 - 1600ºC.</p> <p>The results can be summarized as:</p> <p>(a) ΔGºT[FeNb₂O₆(s) - NbO₂(s)] = 105870 - 42.64ºT (±600) cal/mole FeNb₂O₆ [1550ºC - 1600ºC]</p> <p>(b) ΔGºT[FeAl₂O₄(s) - Al₂O₃(s)] = 49910 - 21.50 T(±300) cal/mole FeAl₂O₄ [1550ºC - 1600ºC]</p> <p>(c) Increasing the manganese content in the melt decreases the value of logh₀ in equilibrium with the transition reaction.</p> <p>The nature of the deoxidation inclusions of the liquid systems Fe-Nb-O, Fe-Al-O and Fe-Al-Mn-O were studied by X-ray analysis and their morphology by scanning electron microscopy. In the Fe-Nb-O system the inclusions were identified as NbO₂ and FeNb₂O₆ and appeared to form agglomerates of small crystallites.</p> <p>In the Fe-Al-O and Fe-Al-Mn-O the inclusions were identified as Al₂O₃ clusters along with FeAl₂O₄ or (Fe,Mn) Al₂O₄ respectively. The thermodynamic data obtained by the present technique are believed to be of high accuracy and the method significantly contributes in the studies of complex deoxidation systems at steelmaking temperatures.</p> <p>An attempt was made to use the present electro-chemical technique in studying the liquid Fe-Ti-O system at high oxygen potentials. Further research is suggested to fully resolve the results.</p> / Doctor of Philosophy (PhD)

<p>The thermodynamics of rare earth-oxygen-sulfur systems was studied at high temperatures using the following four types of oxygen concentration cells.</p> <p>Type I Cells:</p> <p>[equation removed]</p> <p>Type II Cells:</p> <p>[equation removed]</p> <p>Type III Cells:</p> <p>[equation removed]</p> <p>Type IV Cells:</p> <p>[equation removed]</p> <p>Type I and II cells were used to yield the standard free energy changes corresponding to the reaction</p> <p>[equation removed]</p> <p>at two different sulfur partial pressures. Type III cells were used to determine the standard free energy changes for the reaction.</p> <p>[equation removed]</p> <p>Type IV cells were used to determine the standard free energy changes for the reaction</p> <p>[equation removed]</p> <p>Combining the experimental results for the above three-reactions with the values of ΔGfº for Cu₂S and Ag₂S, the standard free energies of formation of La₂O₂SO₄, Pr₂O₂SO₄, Nd₂O₂SO₄, Sm₂O₂SO₄, Eu₂O₂SO₄, Gd₂O₂SO₄, La₂O₂S, Pr₂O₂S, Nd₂O₂S, Sm₂O₂,S, Eu₂O₂S, Gd₂O₂S, Y₂O₂S and Ce₂O₂S were determined at high temperatures. In addition, the standard free energy changes for oxidation of Tb₂O₂S and Dy₂O₂S to Tb₂O₂SO₄ and Dy₂O₂SO₄ were determined.</p> <p>The above data were used to construct the phase stability diagrams for the rare earth-oxygen-sulfur systems at high temperatures.</p> / Doctor of Philosophy (PhD)

<p>This thesis presents the results of a study of two aspects of peritectic alloy solidification: 1. an experimental<br />and theoretical study of the phenomenon of banding (alternate<br />deposition of layers of primary (α) and peritectic (β) phases)<br />during plane front solidification and 2. a theoretical study<br />of microsegregation during the non-isothermal diffusion<br />controlled peritectic transformation in an alloy system where<br />the solute diffusivities differ greatly, ie. Fe-C-Mn.</p> <p>Plane front solidification experiments using Sn-Cd<br />alloys produced specimens which solidified as metastable α<br />phase. The β phase was observed in only those specimens which<br />were purposely disturbed during solidification; the α phase<br />did not re-nucleate at the interface once β phase<br />solidification was established. Banding was observed in one of<br />these specimens, however, the α phase never entirely<br />disappeared from the interface. Mathematical model predictions<br />of the growth transient of the β phase and nucleation<br />considerations showed that the α phase would not be renucleated<br />at the β/liquid interface. It is thus expected that<br />when banding occurs in Sn-Cd peritectic alloys, α is never<br />completely removed from the growth front.</p> <p>The treatment of microsegregation during the non-isothermal<br />diffusion controlled peritectic transformation used<br />in this work exposed several aspects of the problem which are<br />obscured by some of the more sophisticated mathematical<br />treatments of this problem; the transformation was modelled as<br />a series of non-isothermal steps. For the binary Fe-C system,<br />these calculations showed that, while cooling rate and solute<br />diffusivity are important, closure of the δ-ferrite/austenite<br />two phase field ultimately determines the temperature at which<br />the transformation ends and what phases exist at that point.<br />In the ternary case, Fe-C-Mn, the combined influence of<br />constitutional and diffusional solute interactions promotes<br />the following: 1. the concentration gradient of the fast<br />diffusing solute in austenite is minimized with a<br />corresponding greater segregation of the slow diffusing solute<br />in all phases, compared to predictions based on no diffusion<br />in the austenite or complete diffusion in all phases. 2.<br />diffusional solute interaction may increase or decrease the<br />effect of constitutional solute interaction on<br />microsegregation. 3. At the liquid/austenite interface, the<br />possibility of metastable austenite solidification or<br />nucleation of δ-ferrite exists throughout the transformation.</p> / Doctor of Philosophy (PhD)