This thesis is the result of a comparative investigation of two alloy systems from which useful permanent magnet properties can be achieved by suitable heat treatment. Alnico alloys, and in particular a Fe-Al-Ni-Co alloy, are compared with Fe-Cr-Co alloys. The most striking observation throughout this work has been the great similarity in magnetic behaviour and metallurgical structure of the Alnico type alloys and the Fe-Cr-Co alloys. The alloys studied, although they display additional idiosyncrasies peculiar to the individual alloy, all, with the support of TEM observations, behave magnetically (specifically the properties of magnetic anisoiropy and coercivity) in a manner consistent with a common transformation mechanism. The evidence supports the view that this mechanism is spinodal decomposition, which initially produces an homogeneous, fine shape-anisotropic particle microstructure, and by subsequent diffusion during heat treatment achieves a highly ferromagnetic state; all of which are necessary conditions for good permanent magnet properties. The results indicates that some contribution to the magnet properties may ensue from further elongation of the particles with prolonged heat treatment. However the evidence indicates that elongated particles aligned, to an extent allowed by crystallographic conditions, with the direction of the applied field during the early stages of heat treatment are produced as the initial spinodal decomposition product, consistent with the theoretical predictions for this mechanism. The rotational hysteresis loss and rotational hysteresis integral specify curling as the mode of magnetisation reversal in a Fe-Cr-Co alloy after heat treatment to produce the optimum magnetic state. The behaviour of remanence and coercivity for Alcomax 111 and an Fe-Al-Ni-Co alloy suggests a possible proportional relationship between coercivity and the saturation magnetisation of the particles during tempering at 600°C. Coherent magnetisation reversal of the particles will give such a relationship at intermediate coercivity levels if it is assumed that some of the matrix magnetisation can reverse prior to the particle magnetisation.

A slice reactor has been constructed to study stirring patterns in jetted liquid/liquid reaction systems of the type found in top blown oxygen steel making processes. The two liquids used in the study have been mercury and highly viscous glycerol water mixture. The reaction has involved the oxidation of metallic sodium dissolved in the mercury by hydrochloric acid produced in the glycerol water mixture by jetting the liquids with a nitrogen hydrogen chloride gas mixture. The reaction takes place between droplets of the mercury amalgam and the acidified glycerol water solution, this reaction modelling the reaction that takes place in top blown steel making processes between droplets of iron carbon alloy and the oxidising slag. Gaseous hydrogen is produced by the reaction that has been studied, and it is bubbles of hydrogen that produce and maintain the foam within the slice reactor. The slice reactor has been constructed to represent a vertical section cut across the diameter of a centrally jetted axi-symmetric reactor of the BOS type, so that the jet is provided by a slit nozzle set at right angles to the sides of the reactor. The symmetry of gas and liquid flow patterns in the slice reactor is thus essentially two dimensional as opposed to the axi-symmetric flow patterns that exist in actual BOS reactors. The interactions between the two-dimensional jet and the liquids have been studied in the absence of any reaction in order to establish the extent to which the two dimensional system behaves in the same way as the axi-symmetrical system. Particular attention was paid in this study to the depth of the depression produced in the liquid surface by the impact of the jet and the conditions under which liquid was splashed out from the impact region. These studies showed that the characteristics of the two-dimensional jetting system are very similar to those of the axi-symmetric systems that have been extensively studied previously. The behaviour of the slice reactor in the presence of the reaction that produces and sustains the foaming slag has been studied by injecting a colour tracer into the foam and studying its movement. Two distinct flow zones have been observed in the foam, and the effect of changes in operating variables, such as lance height and gas flow rate, on these two flow zones has been studied. The presence of the tracer has allowed the metal droplets sprayed out from the jet impact region to be visible during their path through the foam. The size, velocity and spatial distribution of these droplets have been studied and their response to changes in the operating variables ascertained. As well as studying the behaviour of the reactor with a vertical jet produced from a single slit nozzle, a series of experiments have been carried out with two slit nozzles in which two jets of gas are produced at different angles of divergence. The behaviour of the system with these two slit nozzles has been compared with its behaviour with the single slit nozzle. The results of the investigation have been discussed in relation to the operation of industrial oxygen steel making processes.

Considerable controversy exists concerning the mechanism involved in the softening of the burden within the iron blast furnace and in particular the role played by volatile species in this process. This laboratory investigation was undertaken in order to develop a better understanding of the fundamental processes involved in the softening of the iron bearing components of the blast furnace burden and in particular the contribution of sulphur and alkali metal vapours in this process. The experimental technique developed involved the encapsulation of the test material inside a sealed silica tube containing a gas phase of known composition. The test material was moulded into the form of a Seger cone and the behaviour of the cone at elevated temperatures was used to monitor the softening of the material. This technique was used to study the effect of sulphur and alkali metals on the softening behaviour of both chemical reagents and a commercial iron ore. High temperature observation of the cones was supplemented by optical and scanning electron microscopy after testing. The results showed that the softening behaviour of the ferrous constituents of the blast furnace burden was influenced by the mass, fraction of liquid formed within the sample and by the viscosity of this liquid. Vapour borne sulphur plays an important role in the softening process in the blast furnace through the formation of ferrous sulphide which melts and acts as a lubricant between the solid particles and the more viscous liquid oxide mixtures formed. Alkalis play a less important role in the chemistry of softening in the blast furnace and they may delay softening at low oxygen potentials.

The use of coated iron-nickel powders in the production of sintered components of low-alloy steels has certain advantages over blended mixtures of the elemental powders. In particular, the former produce a more homogeneous material after a specific sintering treatment. There is a limited amount of evidence that the improvement in the degree of homogeniety produces an improvement in the tensile strength of quenched and tempered sintered compacts. The present investigation has aimed to develop a technique by which nickel coated iron powder may be produced and to determine the modulus of rupture of various sintered iron-nickel alloys produced from such powders. The coated powder was produced in a three phase fluidised bed, which contained iron powder, nickel plating solution and a fluidising gas. Coating of nickel on the iron powder particle was achieved by electro-deposition, using cathode rods suspended in a fluidised bed. The central anode consisted of nickel shot contained inside an anode bag. The apparent deposition efficiencies frequently exceeded 100%, which indicated that a process other than electro-deposition was also involved. The second process was shown to be due to a cementation reaction, in which the iron was replaced by nickel. An examination was made of the effect of temperature, anode bag fabric and type of anode material on both types of deposition process. It appears that a nickel coating may be readily prepared by the use of the electroless process alone. 'The modulus of rupture of compacts prepared from both plated and blended materials was strongly dependent on porosity while the degree of homogeneity had a significant effect at the higher nickel contents only. An increase in nickel content improved the modulus of rupture after a specified sintering and hardening treatment, although the magnitude of this effect became smaller as the porosity of the compacts was increased. A comparison of the results obtained from plated and blended powder compacts showed that the former possessed higher values of modulus of rupture, although once again the magnitude of this effect was diminished as the porosity of the compacts increased.

The cause of intergranular brittleness in a number of alloys based on Fe-8%Mn has been investigated. The intergranular brittleness in as-water quenched alloys was mainly because of segregation of N to austenite grain boundaries, but some specimens did show the presence of P as well. A desegregation of carbon in the vicinity of the grain boundaries was also observed which may have also contributed to embrittlement. This embrittlement was further enhanced on air cooling, with segregation of N and Mn to prior-austenite grain boundaries. On ice-brine quenching the alloys, the DBTT was lowered. Intergranular fracture was avoided and the brittle fracture mode changed to cleavage. However, preliminary examination by AES indicated a rise in N content on the cleavage facets. Aging quenched alloys at 450°C resulted in a rapid rise in DBTT and further embrittlement, with further segregation of Mn, N and at a later time P, to prior austenite grain boundaries. Reverted austenite was not detected until after 60h at 450°C, and the hardness rose < 12HV30 on aging. These factors were thought to have little effect on embrittlement. Analysis of segregation kinetics at 450°C showed that the segregation of Mn and P was consistent with bulk diffusion of Mn and P in a-Fe, while measurement of the kinetics of embrittlement by tensile ductility tests indicated that diffusion of N in a-Fe was the main rate controlling factor in isothermal embrittlement. Above 450°C reverted austenite formed rapidly on aging and was thought to be responsible for de-embrittlement. A thermal cycling treatment was devised to overcome embrittlement. Such treatment caused improvement in impact toughness through refinement of grain size and introduction of Y and e phases into the microstructure. The alloys showed resistance to embrittlement at 450°C and evidence of deformation induced transformation on tensile testing at -78°C.

The cementation reaction [chemical formula] although theoretically possible, is not normally considered to be of significance since the negative electrode potential of the reaction is very small and the rate of deposition is usually extremely low. However, a previous investigation of the corresponding reaction between nickel and iron suggests that such reactions may occur at a relatively rapid rate, provided the specific surface area of the iron is very high, as in the case of a fine powder. In the present work the cementation reaction has been followed in a three phase fluidized bed, which maintained the powder in suspension in the solution and ensured that all particles had an equal chance of contact with cobalt ions. An investigation was made of the effect on the reaction of certain process variables including temperature, PH, type of anion in the solution and concentration of iron powder in the cell and cobalt ions in the solution. This has allowed, under optimum conditions, cobalt deposition at a relatively high rate. In view of the relative cost of iron and cobalt this reaction may be a suitable way of extracting cobalt from waste solutions. An examination of the morphology of the deposit suggested that deposition always took place irregularly, so that some parts of the iron powder remained in contact with the solution. Thus the cementation reaction did not come to a rapid conclusion, on account of the formation of a continuous layer of cobalt on the iron surface. This irregular deposit was maintained even when electrodeposition occurred simultaneously with the cementation reaction.

No description available.

No description available.

No description available.

No description available.