Ironmaking & Steelmaking: Processes,Products and Applications

The morphology of alumina was studied in laboratory experiments and in samples from the steel plant. Varying morphologies of alumina were observed after deoxidation and reoxidation but a clear discrimination between alumina formed by deoxidation or reoxidation was not observed. Local concentration gradients during the addition and dissolution of aluminium can explain different growth mechanisms which lead to the observed varying morphologies of alumina.

Inside a silica tube, a piece of Al was brought in contact for a very short time (1–60 s) with liquid Fe containing different dissolved oxygen (O) levels (170–1800 ppm) at 1600°C. The purpose was to investigate the inclusion formation during the initial deoxidation period. Intermetallic compounds of the Fe–Al system appeared due to the interdiffusion of Fe and Al, and the reaction between AL and O at the interface gave rise to the formation of Al₂O₃ inclusions. These inclusions showed different characteristics according to the initial O level in liquid Fe and the interaction time. With lower O content, the amount of inclusions decreased and faceted particles were dominant among the morphologies. While increasing the O content to 780 and 1800 ppm, the inclusion number increased drastically and numerous small spherical inclusions were observed. Further away from the reaction front, clusters of fine particles were formed through a collision–coagulation mechanism. The change in Al₂O₃ morphology with the initial O content was interpreted as a consequence of different supersaturation degrees with respect to AL and O.

Ferroalloys are added during secondary steelmaking to impart special properties to the steel. Depending upon the ferroalloy quality this may lead to the formation of inclusions. The present knowledge lacks in the exact content of the individual elements and the nature of inclusions dispersed in the ferroalloys. In order to broaden the knowledge concerning ferroalloy quality, eight different ferroalloys (i.e. FeMo, FeNb, HCFeMn, LCFeMn, FeTi70, FeTi35, FeSi75 and FeP) were characterised for their impurity content. The samples were investigated for chemical analysis (inductively coupled plasma atomic emission spectroscopy and Leco combustion technique) and microstructural analysis (SEM energy dispersive spectroscopy). These impurities are linked to the ferroalloy manufacturing route. The inclusions observed in the microstructure are in good agreement with the inclusions extracted by the dissolution technique. In the present manuscript, the possible influence of ferroalloy quality over steel cleanliness is evaluated in the context of the impurities extracted and observed in the ferroalloys.

The purpose of this study was to investigate the wetting characteristics of three different oxide substrates (alumina, magnesia and dolomite) in order to examine the work of adhesion of Fe–Cr alloys on these substrates. The contact angle of a liquid drop (Fe–Cr alloy) on the oxide substrates was investigated at 1823 K under a H₂ atmosphere by the sessile drop technique. The contact angle slightly decreased on the alumina and dolomite substrates with increasing chromium concentration, but did not change as much on the magnesia substrate. Among the tested substrates, alumina exhibited the highest contact angle and dolomite exhibited the lowest one. Accordingly, dolomite had the strongest work of adhesion, which was in agreement with the observations in the practical process.

The dissolution of Ti additions in liquid steel during secondary steelmaking occurs in a two step process. In the first step, a steel shell solidifies around the initial cold addition, whereas in the second step, after this shell has remelted, the Ti dissolves directly in the steel bath. The initial presence of this steel shell modifies the position of dissolution and influences the local concentration and thus the inclusion formation. Further complications arise from the fact that part of the Ti will dissolve while enclosed by the steel shell, altering the alloy composition first released in the ladle and effectively shortening the subsequent free dissolution period. The duration of the steel shell period and the fraction of predissolved Ti have been investigated using a conservative one-dimensional sharp interface model solving the coupled heat and mass transfer in a cylindrical shell/addition composite. The influence of the convection conditions and the original Ti radius was evaluated in a parametric study. A pronounced effect of the convective heat transfer on the shell melting time was found. It is thus concluded that the dissolution behaviour is strongly dependent on the local flow conditions, which is determined by factors such as stirring conditions and addition characteristics.

It is known that the oxide composition in stainless steel changes during heat treatment; however, the precise conditions to cause change are not clear. Therefore, the conditions necessary to change the oxide composition by heat treatment have been investigated using steels with different concentrations of Si, Mn, Ni and Cr. The MnO–SiO₂ type inclusions changed into MnO–Cr₂O₃ type inclusions in steel containing 18%Cr and 8%Ni after heat treatment with a low silicon concentration. At high silicon content, the MnO–SiO₂ type inclusion was stable after heat treatment. The change in the composition of oxide inclusions owing to heat treatment depended on the concentrations of Cr in steel. The critical Si content, below which the oxide composition change was observed after the heat treatment, decreased with the decrease in Cr content. In steel containing 1%Cr, the oxide composition change was not observed.

Foamy slags are widely practiced in electric arc furnace (EAF) steel production, reducing energy consumption, improving yield, reducing furnace noise, and increasing refractory service life. Control of slag chemistry within a specific range is necessary to achieve maximum benefits from slag foaming. The 'optimum' slag during EAF steel production is MgO saturated, containing a suspension of magnesium wüstite (MgO.FeO) particles. A thermodynamic program, FACTSAGE, was utilised to study a simplified steel slag system containing four slag oxides (MgO–CaO–FeO–SiO₂), to predict the dual saturation level of CaO and MgO phases and to predict the MgO slag saturation level with variations in oxygen partial pressure, temperature and slag basicity. Results from thermodynamic calculations indicate that there are linear relationships among slag oxides and slag basicity which can be used to predict and control slag chemistry for foaming. Model data were compared with experimental data and models developed by other researchers.

The present work is aimed at the development of a new sustainable method for the utilisation of valuable elements from the steelmaking slag. In the framework of the innovative concept earlier proposed by the authors for utilisation of steelmaking slags, oxidation of iron mono-oxide in the liquid slag was studied experimentally in different atmospheres and in the temperature range 1623–1823 K using thermogravimetry technique. Synthetic (binary and ternary slag system) as well as industrial steelmaking slags were used in the experiments. Analysis of the reaction products was carried out using X-ray diffraction method. A possibility to transform the non-magnetic iron bearing compounds to magnetite in the steelmaking slag by oxidation has been confirmed.

Previous work in the authors' laboratory has shown that an understanding of droplet swelling is essential in predicting droplet residence times in basic oxygen steelmaking. Several workers have observed swelling of high carbon droplets when exposed to oxidising slags. In the present work, the authors present data on swelling rates measured by X-ray fluoroscopy and compare this with the total volume of gas evolved during the reaction. It is shown that only a small percentage of the gas generated is retained in the droplet to contribute to swelling. The gas generation rate is shown to be controlled by the rate of nucleation of CO bubbles inside the droplet. The critical supersaturation pressure for nucleation is found to be two orders of magnitude less than predicted from theory, which is in keeping with many other studies on nucleation of gases in liquids. However, the effect of surface tension, temperature and saturation pressure shows quantitative agreement with theory.

Detailed comments on governing equations and boundary conditions on which numerical results of gas flow in porous media should rely are presented. Numerical results are shown for the gas flow through a large blast furnace with designed layered structure of burdens. Not only the effects of radial and vertical distributions of resistance to flow but also the effects of different top profiles of burden, such as V-, M-shapes and V-shape with a side terrace, on the resultant gas flow are discussed in relation to the properties of burden, such as repose angle, particle size, voidage and different charging sequence. For the gas flow in a coaxial annulus bed, which is expected to appear in the lower part of blast furnace during a start-up or a blow-out operation, numerical predictions are compared with analytical ones based on a simple yet convenient gas flow model.