A one-dimensional static model for simulation of the blast furnace process has been developed. This simulation model is a part of the KTH Blast Furnace Process Model. The internal state of the furnace is described by the calculated distribution of fourteen process variables along the furnace axis which include temperature of the gas and condensed phases, composition, volumetric flowrate, density and pressure of the gas phase, degree of iron ore reduction, limestone decomposition and coke solution loss.
The behaviour of these process variables in the blast furnace is described by a set of ordinary differential equations together with three algebraic equations. In order to solve these fundamental equations and facilitate the treatment of operational data from blast furnace plants, the following submodels have been built up: 1) submodel for data base creation, 2) mass balance submodel for the whole furnace, 3) heat balance submodel for the whole furnace and 4) submodel for determination of the reaction rates.
This simulation model has been used for the simulation of two Nordic blast furnaces. The calculated profiles of gas temperature and iron ore reduction degree show similar aspects to those measured by shaft sensors in working blast furnaces. The model is useful for obtaining a deeper understanding of the blast furnace process.

View full abstract

A prediction model has been developed based on the simulation model presented previously. These models are the two components of the KTH Blast Furnace Process Model. The concept of this prediction model is to use the same fundamental equations as in the simulation model and to use some output of a simulation of a blast furnace. Certain assumptions should be made for an individual change in operational conditions in order to build up the mass balance and heat balance submodels for the determination of the boundary conditions in a prediction. The ore to fuel ratio and the CO utilization are adjustable parameters in the model. The furnace internal state as well as furnace productivity and fuel consumption in the last iteration are considered to be the predicted results.
The prediction model has been designed for the following five cases: 1) increased blast temperature, 2) oxygen enrichment of the blast, 3) coal injection, 4) coal injection combined with oxygen enrichment and 5) changed coke quality. Moreover, this model can also be used for analysis of the thermal conditions in a blast furnace when an operational parameter, such as blast temperature, coke moisture and iron content in the ore, fluctuates.
The predicted operational indices were compared to the ones from industrial tests. The validity of the KTH model is indicated by this comparison.

View full abstract

The oxidation kinetics was studied thermogravimetrically with three kinds of cement-bonded natural ilmenite pellets. Appreciable temperature rise was found inside the pellets during their oxidation with pure oxygen. In order to suppress the temperature rise, a series of oxidation experiments was carried out with diluted oxygen (9.8mol%O₂-N₂) in the temperature range from 1073 to 1273 K. The experimental data were analyzed in the light of the unreacted core shrinking model resulting in that the oxidation rate was mainly controlled by the intrapellet diffusion of oxygen and the intrinsic chemical reaction. The temperature dependence of the kinetic parameters was also determined.

View full abstract

The hydrogen reduction kinetics of ilmenite was studied thermogravimetrically from 1073 to 1273 K with three kinds of cement-bonded natural ilmenite pellets. The kinetic behavior of the reduction of preoxidized ilmenite was researched with the same technique from 973 to 1273 K. The reduction of the pellets proceeded topochemically, but, the behavior of the grains constituting the pellets was more complex. The reduction kinetics of preoxidized pellets was analyzed according to a two-interface kinetic model and the kinetic parameters of the two main reactions Fe³⁺→Fe²⁺ and Fe²⁺→Fe°, which mainly constitute the overall reduction of preoxidized ilmenite pellets, were determined separately. The experimental data of the reduction of unoxidized pellets were interpreted in terms of a modified one-interface kinetic model derived from the aforementioned two-interface kinetic model. Diffusion of gaseous species through product layer and intrinsic chemical reaction were found to be the main rate controlling factors during reduction.

View full abstract

Using a half section three-dimensional model of a blast furnace, a tuyere-injection experiment of coke fines was conducted. The following findings were obtained from the analysis of the influence of the injection rate and particle size of the fines on the burden descent and gas permeability and the study of the circulation and deposit behaviors of the fines in the blast furnace. With an increase in the injection rate of fines, the frequency of slips, the gas pressure, and the frequency and extent of the raceway shape variation increased. But the influence of the fines injection was limited when their particle size was much smaller than the size of the charged materials. The injected fines are to be deposited at the surface of the dead-man and near the furnace wall where the burden descent velocity is low. When the high-concentration regions of fines are formed at the surface of the dead-man and near the furnace wall, the burden descent region in the lower part of the furnace reduced and the retention time of the burden decreased. With the V-shaped stock and inadequate central gas flow, fines are likely to be deposited in the center and they descend with the burden.

View full abstract

A general equation of the concentration distribution under laminar conditions was proposed. On the basis of this equation an integral method was derived. This method is based on setting a balance of flux gradients of the transfer phenomena that occur. It was shown that the flux gradient is identical to the characteristic quantity N which defines the state at the cross-section perpendicular to the surface, the definition of which enables the formulation of the proposed general equation.
The general approximative integral method was applied to the phenomena on the metal surface, mass transfer during a heterogeneous chemical reaction in the laminar boundary layer over a flat plate–measuring foil. A model of the mass transfer coefficient was presented as a function of the limiting heterogeneous chemical reaction rate constant, characteristic time relation and surface function of the appropriate gas atmosphere.
The proposed model is a theoretical basis for the development of a measuring device of non-equilibrium gas carburizing processes.

View full abstract

The meniscus shape of molten steel under the alternating magnetic field was calculated in the two dimensional case by means of finite differential method. Penetration of magnetic flux line into the conductive charge was also taken into account. The detailed method of calculation was described and the validity of calculation was checked by theoretical analysis.
The height of meniscus becomes large as the magnetic flux density and the frequency of the magnetic field increases. And further, it is cleared that the meniscus shape, under the magnetic field caused by the short coil, is more favorable than the uniform magnetic field or the one which is caused by the long coil, in terms of supporting meniscus.

View full abstract

The deoxidation system in the ESR process is a complex balance between atmosphere/slag/metal reactions involving the alloy composition, the electrode inclusion content and additions of metallic deoxidants during melting. We describe the relative contributions of slag/metal reactions and added deoxidants to the inclusion-forming system. It is concluded that the oxide compositions found in ESR ingots of low alloy steel can be explained using the slag FeO content as a monitor of the reaction scheme. Following this finding we also conclude that it is not possible in presently-developed ESR systems to produce oxide (or sulphide) inclusions with a Ca-content sufficiently high for shape control.

View full abstract

The rate of the dissolution of hematite and wustite pellets in bath smelting type slags was investigated at 1723 K. The rates were measured by the change in their diameter and the rate of gas evolution from the pellet. The diameter of the pellets, as a function of time, was obtained from recorded X-ray radiograph pictures. For unstirred slags it was found that the rate of the dissolution of hematite pellets was faster than that of wustite pellets due to gas evolution from the hematite pellets which increases heat and mass transfer. The rate of the dissolution of wustite and hematite pellets were much faster in slags stirred with argon and, in this case, wustite dissolved faster. The rate of gas evolution during the dissolution of hematite with an inert gas over the slag was also measured by using a mass flowmeter. The rate of gas evolution was also determined under different CO-CO₂ gas mixtures by analyzing the off gas with a mass spectrometer. It was found that the amount of oxygen evolved increases with decreasing the CO₂/CO ratio. From the experimental results the heat transfer coefficient for the present conditions was estimated. These values were extrapolated to the actual operating conditions of a bath smelter and the melting time and the amount of undissolved pellets at the steady state in the slag were calculated. It was estimated that the pellets melt in less than 10 sec and the fraction of unmelted pellets, in the slag in a typical practice, is less than 1% of the volume of the slag. The rate of production in a bath smelter depends on the amounts of Fe²⁺ and Fe³⁺ resulting from the dissolution of the pellets. The present results indicate that for Fe₂O₃ pellets the rate of iron production will be about 25% less than for FeO.

View full abstract

In order to investigate the effect of hold time on low cycle fatigue behavior of NT (normalized and tempered) 2 1/4Cr-1Mo steel at 723 K (0.4 Tm), strain controlled low cycle fatigue tests with three different tension-hold times, i.e. 900, 1800 and 3600 sec, were carried out. Fracture surface and microstructural change were examined to investigate the effect of hold time on the fracture process using optical, scanning electron and transmission electron microscopes (OM, SEM and TEM). Fatigue life was decreased and cyclic softening became larger as hold time increased. Judging from the examination by OM, SEM and TEM, hold time dependence of the cyclic softening and fatigue life is thought to be caused by the enhanced accumulated damage level during tension holding, not by creep effect.

View full abstract

In the production of Zn hot dip galvanized steel sheets, the concentration of Al in the galvanizing bath must be maintained within a limited range to ensure optimum performance. For this purpose, an accurate and rapid determination of Al by X-ray fluorescence spectrometry was studied. The results were as follows:
(1) The AlKα line intensity increased with the lapse of time after sample preparation because of the enrichment of Al near the surface of the sample.
(2) An accurate and rapid analysis was performed by preparing a sample with an SiC abrasive paper of coarse grit size 40 and by starting the X-ray intensity measurement after a brief 3 min lapse of time following the sample preparation.

View full abstract