In the present paper three different methods to calculate the equilibrium oxygen activity have been compared with measured oxygen activities and oxygen activities based on sulphur equilibrium between slag and steel during ladle treatment at Scana Björneborg.

Three slag models were used to estimate the oxide component activities of Al2O3 and SiO2 in the top slag and in the equilibrium calculations the dilute solution model for the liquid steel phase was used. The results show significant discrepancies between the calculated and measured oxygen activities and the reasons for the differences are discussed.

The investigation focused on the mixing of the metal and slag phases during ladle refining from the point of tapping the EAF to casting. Steel droplet distributions were determined for slag samples taken at different stages in the ladle refining process at two different steel plants in Sweden. The droplet distributions were determined using light optical microscopy and classification according to the standard SS111116. Sample analysis results showed the slag samples taken before vacuum degassing to contain the greatest concentration of steel droplets. The total interfacial area between the steel droplets and slag was determined to be 3-14 times larger than the projected flat interfacial area between the steel and slag. The effects of slag viscosity and reactions between steel and slag on metal droplet formation in slag were also considered.

The dispersion of metal droplets in slag was investigated through analysis of slag samples taken during ladle refining at Scana Steel in Bjorneborg, Sweden. The chemical composition of steel droplets found in the ladle slag was determined for five industrial-scale heats. Possible reactions occurring between the steel droplets and slag were identified, as were differences in steel-droplet and steel-bulk composition. Three different slag models were used to calculate the activities of oxide components (Al2O3 and SiO2) in the slag. These results were then used in the dilute-solution model, whereby oxygen activities in the steel droplets were calculated and compared with measured oxygen activities in the steel bulk. Significant differences were found in the comparison of both the calculated and measured oxygen activities and the steel bulk and droplet compositions.

Focus in this work was to study dissolved oxygen content and oxygen activities in different cast irons. Plant trials were performed at three occasions for lamellar, compacted and nodular iron melts. The results show that at temperatures close to the liquidus temperature the oxygen activities ranged from 0.03-0.1 ppm for LGI, around 0.02 ppm for CGI and 0.001ppm for SGI. In addition, it was found that as oxygen activities increase with time after an Mg treatment, the ability to form compacted graphite or nodular graphite in Mg-treated iron melts was lowered. Also, oxygen activity differences up to 0.07 ppm were found for different hypoeutectic iron compositions for lamellar graphite iron at the liquidus temperature. Overall, the observed differences in the dissolved oxygen levels are believed to influence how graphite particles are incorporated into the austenite matrix and how the graphite morphology will be in the cast product.