Ceramically bonded doloma-based bricks may be used without problems in the freeboard of the ladle lining, for which average temperatures are limited.
Accretions on the inner walls of submerged entry nozzles (SENs) can be formed or modified by interactions between molten steel and refractory materials. Thermochemical reactions that qualitatively explain these interactions were modeled using Facility for the Analysis of Chemical Thermodynamics (FACT) thermodynamic software. The predicted phases were compared with results from high temperature static experiments. In these experiments, steels with and without aluminum were melted in refractory crucibles fabricated from various oxide refractories with and without graphite. Samples of the refractory-metal interface were characterized using cathodo-luminescence (CL) microscopy, reflected light (RL) microscopy and scanning electron microscopy (SEM) to understand the interactions and establish the effects of refractory and steel composition on accretion formation. Both microscopy data and thermodynamic predictions indicated that graphite-containing refractories, especially magnesia-graphite and alumina-silica-graphite, resulted in the greatest degree of interaction with steels. All of the carbon-free and impurity-free materials resulted in significantly fewer interactions.
Quantitative values for the sensitivity to dimensional variations are calculated and can be used to optimize the tolerances at rolling.
An experimental study on Al deoxidation of 18Cr-8Ni stainless steel under the presence of the top slag with CaO-Al 2 O 3 -MgO-F system was performed to understand the formation mechanism of alumina, spinel, magnesia and calcium aluminate inclusions. Alumina or spinel inclusions were formed immediately after the addition of Al. MgO and CaO in the slag phase were reduced by Al to generate soluble Mg and Ca in steel. This tendency was enhanced when Al was added to the slag surface. Thereafter, soluble Mg reacted with the inclusions, which formed early to form MgO inclusions, due to its higher activity than Ca in the molten 18Cr-8Ni stainless steel. The behavior of MgO.Al 2 O 3 and MgO inclusions agreed with the phase stability diagram calculated with thermodynamic data. This result validates the assumption of maintaining the local equilibrium between the inclusions and the molten steel. Calcium aluminate inclusions, which have proved to be the most stable with the thermodynamic calculation, formed at the last stage of deoxidation when the molten stainless steel contained Ca of at least 1 mass ppm and Al ranging from 0.1 to 1 wt. %.
Samples of solidified mold flux were collected from below the mold of a continuously casting machine. Fluxes used for the casting of medium carbon, low carbon and super ultra low carbon (SULC) steel grades were sampled to compare properties. Compositional analysis for all flux specimens showed little variation from samples taken from the top of the mold and bulk powder compositions. X-ray diffraction (XRD) and micrograph image analysis indicated the flux for medium carbon steel had between 80 and 90 percent crystalline material in its structure at the bottom of the mold. This compares to approximately 65 percent crystallinity for the low carbon steel flux and 45 percent for the SULC steel flux. XRD phase analysis indicated the crystalline proportion of all fluxes consisted almost entirely of the mineral phase cuspidine (3CaO.2SiO 2 .CaF 2 ). SEM microanalysis of the fluxes suggested that the medium carbon steel flux may also contain nepheline (Na 2 O.Al 2 O 3 .2SiO 2 ). Photomicrographs of the sample cross sections showed different crystal morphologies across the film thickness. Fine crystallites, distinct dendrites and amorphous regions were evident in the low carbon and SULC steel fluxes. Little amorphous material was observed in the medium carbon flux, with large distinct crystals dominating the structure instead.
The first step that must occur for refractory wear is penetration of the refractory by slag. This occurs through a capillary force sucking the slag down through the pores of the refractory.
An examination of the clogged nozzles used for stainless steels containing rare earth metal (REM) was performed. The accretion mainly consisted of solidified metal in which Ce and La oxides, which were considered to promote nucleation of delta ferrite, were observed at the nozzle/metal interface. To prevent nozzle clogging a study has been further conducted to determine the effects of Ce and Al for deoxidation on inclusion composition in 25Cr-6Ni stainless steel. The rate of the decrease in oxygen content is lower when both Ce and Al were added than when only Al was added. The compositions of the inclusions after deoxidation were analyzed as an Al 2 O 3 -Ce 2 O 3 system. The quantitative analysis revealed that the Ce 2 O 3 content increased with increasing Ce content in the molten steel. The composition of deoxidation products took only compounds of Al 2 O 3 , CeAl 11 O 18 , CeAlO 3 and Ce 2 O 3 . As a result, an appropriate timing of addition and ratio of Al and Ce was determined to prevent nozzle clogging, avoiding the formation of Ce 2 O 3 inclusions.
By means of surface characterization and subsequent hot-dip galvanizing tests, both Al-alloyed TRIP grades showed good galvanizability.
The possible causes for the poor foamability of stainless steelmaking slags were examined in this research. Specifically, the foam index of a simulated stainless steelmaking slag containing Cr 2 O 3 (CrO and Cr 2 O 3 ) was measured. The rate at which carbon reacts with Cr 2 O 3 , CrO and FeO was also measured. Experimental results show that the foam index of stainless steelmaking slags is comparable to that of carbon steelmaking slags, provided that the amount of solid chrome oxide particles or complexes is not excessive. However, the experimental results demonstrate that the reaction rate of carbon with CrO or Cr 2 O 3 dissolved in the slag is slow. Hence, this generation of CO is significantly slower than for the reaction rate of carbon with FeO dissolved in slags. Therefore, the lack of FeO in stainless steelmaking slags is a primary reason for poor foamability. Also, exploratory research indicates that waste oxides containing carbon and iron oxide produces high rates of CO evolution and consequently could improve foaming.
The spontaneous evaporation of gases from liquid slags containing fluorine is a major industrial and experimental problem at high temperatures (> 1,273 degrees Kelvin). The kinetics of evaporation of gaseous species from fluorine-containing slags was investigated to develop an understanding of the factors controlling the rate of fluoride evaporation in dry atmospheres at high temperatures. The objective of this study was to quantitatively evaluate the effects of slag chemistry and temperature on the rate of fluoride evaporation and to clarify the rate-controlling mechanism. In dry argon atmospheres, CaF 2 -SiO 2 -CaO slags are unstable and SiF 4 gas is released. In commercial mold fluxes, which contain Na 2 O, the gaseous product was NaF(g) and SiF 4 (g). The rate of evaporation depended on the temperature and slag composition. Mold fluxes that contained a high Na 2 O content exhibited the highest rate of weight loss among all specimens tested. Mathematical modeling of the kinetic results suggested that the rate of SiF 4 evaporation from the CaF 2 -SiO 2 -CaO slags is in a mixed control regime where gaseous diffusion and either surface reaction kinetics or mass transfer within the liquid slag both contribute to the overall evaporation rate.
A mathematical model of a nozzle used in AOD converters was developed in a previous study. In the work covered in this article, an attempt was made to verify velocity predictions from the nozzle model through a comparison of measurements. Experiments were performed at AGA AB's laboratories using a laser doppler anemometer (LDA) to measure axial velocities at the outlet of a nozzle having the same dimensions as nozzles used in industrial AOD converters. The model predictions agreed well with measured data. An attempt was also made to compare turbulent kinetic energy predictions from the k-e turbulence model with turbulence predictions calculated using an analytical expression that uses measured velocity fluctuation values. The positive results from this comparison indicate that the model can produce fairly reliable predictions of turbulence at the nozzle outlet. The results from this study are promising regarding use of the nozzle model to calculate boundary condition input values pertaining to the nozzle inlet that will be required for an AOD converter model currently being developed.
This paper is a review of the fundamentals of inclusion removal from liquid steel by attachment to the rising bubbles. Based on the water/particle study carried out in the field of mineral processing, silica inclusion and argon gas are taken as the bubble/inclusion phases. The mechanism of bubble/inclusion interaction, bubble size and rising velocity, liquid film formation and rupture between the bubble and inclusion, collision time, film drainage time, sliding time, collision and adhesion probabilities are discussed. A simple mathematical model of inclusion removal by attachment to the rising bubbles is described, and the effects of gas flow rate, bubble size and time on the inclusion removal are discussed. Based on the results of the models, it is concluded that the optimum bubble diameter for the inclusion removal is 1 to 5 mm. The shroud from the ladle to the continuous casting tundish is a good place to inject gas to remove inclusions by attachment to the rising bubbles.
Stantec Global Technologies Ltd., (Stantec) has implemented the Goodfellow EFSOP (TM) control system at Deacero, Saltillo. The Goodfellow EFSOP (TM) system includes analysis of off-gas chemistry and closed-loop control of freeboard oxygen injection, both through the existing burners and the CoJet (TM) oxygen injection system. The system has been on-line since August of 1999 with operating data from more than 4000 heats.In addition to energy savings from the control of furnace combustion, Deacero and Stantec have also been working on overall control of chemical energy sources in the furnace. Elements such as: decarburization rate, carbon usage, scrap mixture, burner firing rate and yield have been analyzed to determine the least cost operating practice for different operating conditions. The off-gas analysis system, in combination with the Goodfellow EFSOP (TM) data acquisition and analysis system have been key tools in this analysis.The results of the Goodfellow EFSOP (TM) system have been very impressive. Through a combination of process changes and closed loop control of freeboard oxygen, monthly average electricity consumption has been; reduced from 320 kWh/tonne to 285 kWh/tonne on a charge basis. Stantec and Deacero have also been able to identify the least cost operating scenarios for different production requirements. Depending on the level of productivity required, cost savings of $3 to $5 / billet tonne have been achieved which includes a savings from improved yield and productivity.This paper will include practical plant experience to demonstrate the benefits of the Goodfellow EFSOP (TM) system.
Improvements in direct-reduced iron,metallization, and carbon content have mainly contributed to increased productivity and are essential in controlling refractory life.
The VOF model provides a useful tool for predicting the cavity depth for an oxygen jet. it can also provide simulations for liquid flow and surface profiles in an EAF.
Accurate lining wear monitoring is an effective method of maximizing the lifetime of ladle-lining, improving safety and cutting refractory costs at the same time.
Use of the PCA system at Dofasco has proven successful in online fault detection. in addition to detecting the onset of abnormal casting, the PCA monitoring system permits the interpretation of alarms and has enabled operating personnel to gain process insight.
In the present study, the equilibria of manganese and sulfur between CaO sat -SiO 2 -Fe t O-MgO sat and molten iron were studied in the temperature range of 1,550 to 1,650°C. The decrease of CaO solubility by the addition of MnO was compensated by increasing MgO solubility. The addition of MgO to CaO sat -SiO 2 -Fe t O slag slightly increased the activity coefficients of Fe t O and MnO, and the properties of SiO 2 as an acidic oxide were clarified in this slag system. The oxygen content of metal in equilibrium with (375 to 41.1 mass percent) CaO-(5.5 to 6.6 mass percent) SiO 2 -(41.6 to 43.7 mass percent) Fe t O-(5.9 to 8.3 mass percent) MgO slags doubly saturated with CaO and MgO is expressed by the following equation: log[mass % O]=-5680/T + 1.87 The addition of MgO to CaO sat -SiO 2 -Fe t O slag also increased the sulfide capacity of the slag. The approximate validity of the regular solution model was examined to formulate the thermodynamic properties of complex slags containing MnO.
The evolution of the carbon distribution between bulk, grain boundaries and dislocations during both the continuous annealing and the strain aging (e.g. paint baking) of ULC BH steels is numerically simulated. The calculations are successfully fitted to strain aging experiments and internal friction measurements. An increase of the grain size together with a higher cooling rate from the annealing temperature significantly increases the bake hardenability of ULC BH steels. This is contrary to the situation in low carbon BH steels due to the presence of the cementite particle distribution between grain boundaries and matrix in the latter steel.
