Abstract
The strength and structural changes accompanying the reduction of hematite, magnetite and wustite have been determined in the temperature range 400 - 1100°C. The investigation was conducted on relatively dense polycrystalline specimens, in which the iron oxide grains were both directly bonded and those in which the grains were bonded with synthetic slags. These specimens were produced by compacting and sintering fine powders of the oxides and oxide/slag mixtures. Reduction was effected isothermally in controlled hydrogen/steam atmospheres to each of the lower oxides and iron. The changes in specimen volume, microstructure and compressive strength resulting from each reduction step were measured as a function of the reduction temperature. The rate of reduction in hydrogen of each material was also measured at 550, 750 and 1000°C. The results indicate that, within the experimental conditions employed, the reduction of hematite involved an expansion of the specimen, pore formation and extensive internal cracking. The major changes occurred during the hematite/magnetite transformation and were all influenced by the reduction temperature. These structural changes have been related to the crystallographic aspects of this transformation and also to the mechanical properties of hematite at high temperatures. Mechanisms have been proposed to explain the pore formation and cracking based primarily on the results of a microscopic examination of partially reduced specimens. It has been conclusively established that the microscopic and macroscopic changes in the hematite/magnetite transformation are directly responsible for the poor reduction strength and high reducibility of hematite. The addition of calcium silicate and calcium mono-ferrite slags to hematite decreased the volume expansion and internal cracking involved in the reduction. These improvements have been related to the mechanical properties of the slags at the reduction temperatures concerned. The reductions of magnetite and wustite were shown to involve only minor volume changes and usually no internal cracking. It is suggested that the difference between the reduction behaviour of these oxides and that of hematite is related to the similarity in the crystal lattices of magnetite, wustite and iron, and to the marked plasticity of wustite and iron at the reduction temperatures. The less extensive structural damage involved in the reduction of magnetite and wustite has been shown to be responsible for the superior reduction strength and inferior reducibility of these oxides compared with hematite. The addition of olivine slag to magnetite and wustite had little effect on the reduction structures, but raised the reduction strengths and lowered the reducibility. The latter has been attributed to the inherent high strength and poor reducibility of the slag. The probable influence of the structural and strength changes of iron oxides during reduction on the behaviour of ores and agglomerates in the blast furnace has been suggested.