Blast furnace based steel production will remain important in the world in future, according to analyses of the trend in steel production in over the last 30 years.
An outline of the trends in Japanese steelmaking technology during the past 10 years from the viewpoint of (1) improving the quality of steel, (2) greater productivity and (3) responding to environmental problems is discussed.
Based on a recognition of these conditions, future technological tasks are reviewed and comments offered on new aspects such as iron making processes and continuous casting in near net shape. A brief comment on the role of the Japanese steel industry in the future are described.

View full abstract

Some examples of basic studies related to steelmaking by physical methods, which have been carried out by the author and his group, are described on the estimation of some physical properties of liquid metals, a new concept of network parameter for slags and fluxes and a new viscosity measurement for metallurgical melts.

View full abstract

Improvements in steel quality are strongly dependent on elimination of "unsought variations" in chemistry, temperature and fluid flow of metal, slag and gas phases within, and during transfer between, the ladle, tundish and mold. As we move toward the next century it is imperative that we not only prevent "unsought variations" during processing but also develop on-line sensor technologies which will permit us to continuously monitor and thus detect at an early stage, the occurrence of any "unsought variations" due to perturbations or imperfections within the processing system.

View full abstract

In the Japanese steelmaking industry, continuous casting technologies for manufacturing defect-free slab or bloom with high productivity are becoming increasingly important. The surface and internal defects of steel products attributable to steelmaking conditions originate in the entrapment of inclusions or gas bubbles by the solidifying shell in the casting mold or strand, and in the formation of center-line segregation at the final stage of solidification. To suppress such defects and ensure high quality in continuous casting process, many advanced technologies have been developed and applied in industrial plants as follows:
(1) To promote the removal of inclusion from molten steel in the tundish, centrifugal flow tundish (CF-tundish) or H-shaped tundish was developed.
(2) To prevent the entrainment of inclusion or gas bubbles into molten steel in the mold, flow control devices using the electromagnetic brake and electromagnetic stirrer have been adopted. Flow control mold is one example which is characterized by two static magnetic fields imposed on the entire width of a slab. Surface and internal defects in cast slabs are remarkably reduced by this newly designed tool.
(3) To reduce the degree of center-line segregation, the solft-reduction process with a plane reduction unit or short-pitched roll segment for cast slab and continuous forging process with an anvil for cast bloom were applied to practical operations.

View full abstract

There have been recent developments in EAF-steelmaking in Japan in the quantity and the quality of steel scrap. Generation of scrap is gradually increasing the quantity, but the deterioration of market scrap and the higher requirements for EAF-steel products is making quality more of a problem. Scrap pre-treatment techniques to meet the quality requirements of expanded EAF-steel products, such as blenbing in a busket to control the tramp-elements, shredding to remove non-ferrous contamination etc., have been developed and are being used. The technological interests of most mini-mill are focussed on the better preparation and better utilization of scrap.

View full abstract

Some of the phenomena linked to inclusion formation by slag entrapment and further transformation in liquid and solid steel have been investigated and interpreted. Slag entrapment in mould is very easy to provoke and may lead to a wide range of inclusion sizes. Some TTT diagrams for inclusion crystallization have been determined. The driving force for crystallization can be estimated by thermodynamic evaluation using slag models.

View full abstract

Isothermal phase relations for the MnO-SiO₂-Ti₂O₃ system were investigated at 1 773 and 1 573 K by a chemical equilibration technique. The solubility of Ti₂O₃ was found to be 30 to 40 mass% and 20 to 33 mass%, respectively. In the MnO-Ti₂O₃ binary system, the existence of 2MnO·TiO_x was confirmed by X-ray diffraction. Isothermal phase relations for the MnO-SiO₂-TiO₂ system were also investigated at 1 773 K, as was the dependence of the SiO₂-TiO₂ doubly saturated composition in the MnO-SiO₂-TiO₂ system on temperature from 1 573 to 1 773 K. The dependence of Ti⁴⁺/Ti³⁺ ratio in 2MnO·TiO_x on oxygen partial pressure (P_O2=3.37×10^-18 to 1.94×10^-13 atm) is discussed.

View full abstract

The activities of calcium in the CaO_satd.-Ca-CaX₂ fluxes (X: Cl, Br, I) were measured as a function of the calcium composition at 1 473 K by equilibrating these fluxes with a solid zirconium plate. The activities decreased in the order CaCl₂, Cal₂ and CaBr₂ at the same calcium fraction of these fluxes. The observed activities are compared with those estimated using binary phase diagrams of the Ca-CaX₂ system. The possibility of the removal of phosphorus and the tramp elements such as arsenic, antimony, bismuth, lead and tin from carbon saturated iron using calcium-calcium halide fluxes is discussed.

View full abstract

The kinetic behavior of aluminum oxidation in molten iron by CaO-SiO₂-Al₂O₃, CaO-Al₂O₃-FeO, CaO-Al₂O₃-MnO and CaO-SiO₂-Al₂O₃-FeO-MnO slags was investigated at 1 600°C under an argon atmosphere. The oxidation rate of aluminum by SiO₂ is slower than those by FeO and MnO in CaO-Al₂O₃ slag. No difference in the oxidation rate could be observed between slag containing FeO or MnO. More insoluble alumina inclusions were observed in the metal for slag containing FeO or MnO than in that containing SiO₂.

View full abstract

The turbulence structure of a round air-water vertical bubbling jet generated in a cylindrical bath was investigated using a two-channel laser Doppler velocimeter. In the central part of the jet, turbulence production was much greater than for a single-phase water jet. This difference was mainly attributable to additional turbulence production in the wake of bubbles rising upward. Turbulent motions were classified into four distinct categories. The contributions of each category to the turbulence kinetic energy and the Reynolds shear stress were determined based upon a conditional sampling method.

View full abstract

Measurement of interfacial tensions under industrially significant conditions are diffcult due to the reactivity of industrial slags with refractory containers and due to non-equilibrium conditions between liquid steel and liquid slag. These two problems necessitated the development of a new experimental technique to allow steady state values of interfacial tension to be measured. This technique has shown that actual interfacial tensions in real systems can be variable with time and are either higher or lower than values predicted from equilibrium studies.

View full abstract

Model experiments were made of the floating and dispersing characteristics of buoyant particles in a gas-agitated water vessel to learn the behavior of inclusion particles in a turbulent flow field of liquid steel. Hollow glass particles were put into the vessel, nitrogen gas was blown from the bottom of the vessel, and the particle concentration was measured by a silicon photodiode. It was found that the concentration of particles in the bulk water increased abruptly at a certain gas-flow rate and saturated at higher rates. This critical gas-flow rate was correlated well with the terminal velocity and vessel geometries. Numerical simulation of the particle dispersion was also carried out using the k-ε model and the effect of turbulence on the particle behavior was discussed.

View full abstract

Cold model investigations were performed in two and three phase systems, including water-, slag- and gas-phase. The following nozzle-arrangements were chosen: single central nozzle, centric row, eccentric row and eccentric triangle. The investigations show an intensive vortex formation in the bath as a result of a coupled variation of parameters like blowing rate and existence of a slag simulation phase. A clear-cut vortex is evident at an undercritical blowing rate. The buoyant free jets do not describe a stable path to the bath surface. Two liquid zones with two mutually distinct directions of circulation are present. In the case of the eccentric arrangements of the nozzles, a vortex occurs in the smaller partial zone; however, this vortex is not always stable. In the larger partial zone, a stable vortex with circular motion occurs. The tests with eccentric arrangements yield the best mixing results as a whole. The flow conditions in the bath at a supercritical volume flow rate effect the phenomenon of sloshing, that is an oscillating motion of the bath. The critical flow rate depends on the arrangement of the nozzles. The inception of sloshing occurs at a relatively low blowing rate with the use of a single central nozzle. Turbulent regions are observed in the bath here. In the case of eccentric nozzle arrangements, sloshing occurs only at higher blowing rates.

View full abstract

Detailed information on the fluid flow around a bubble is required for better understanding of metallurgical reactions in the steelmaking processes agitated by gas injection. In the current research a model has been used for a cold simulation experiment of water-air two-phase flow. Previous studies have focused mainly on the time-averaged flow structure around bubbles. However, it is well known that the movement of a bubble is transient and three-dimensional, and hence, the flow around the bubble is also transient and three-dimensional. This study was carried out to elucidate the flow around a single bubble rising in a cylindrical water vessel.
3-D PTV (Particle Tracking Velocimetry) is used to analyze the movement of gas-bubbles in the water bath. The water flow around a single bubble is investigated by the technique of PIV based on visualization and image processing. An experimental apparatus is set up to generate a bubble of known volume rising from bottom. Velocities of the bubble and the surrounding fluid are obtained simultaneously by a three consecutive time picture cross-correlation 3-D PTV. A new technique that can measure the fluctuating velocity field and ensemble averaged kinetic energy from the time series of particle imaging is developed in data processing.

View full abstract

In general, the refining reactions in steelmaking are well understood and most elements can be controlled by the proper slag practice. However, the nitrogen reaction is complex and its content difficult to control. A recently developed model for computing the removal of nitrogen in OSM is presented along with results quantifying the effect of scrap, stirring gas, and oxygen purity. For the EAF, nitrogen is also removed by CO formation by the carbon-oxygen reaction and a model for the rate is presented. However, recent work shows that the CO evolved from DRI comes off in the slag and is not effective in actually removing nitrogen from the atmosphere. Nitrogen control in ladle processing is discussed, including a new model for vacuum tank or ladle degassing for nitrogen.

View full abstract

The weak oxidizing powder blowing method (VOD-PB) was applied to remove copper and tin from molten iron with decarburization in a 1.0-1.5 ton scale experiment. The results demonstrated that the technique was applicable to removal of these metals from iron melted from scrap under practical reduced pressure of about 130 Pa. SiO₂ powder blowing was the most appropriate to eliminate copper and tin among SiO₂, MgO, iron ore and oxygen gas. In case of SiO₂ powder blowing, removal percentages of copper and tin from molten iron containing less than 0.01 mass% sulfur were 30-40 and 20-30%, respectively, in 7.2×10³ sec treatment. About 80% of the tin was removed from the melt containing 0.1 mass% sulfur as a volatile substance like SnS.

View full abstract

The paper presents a study of the physical and chemical processes that occur when treating liquid steel by inert gas and reactive powders, and a mathematical model using FEM to simulate the stirring state of steel as a result of the injection treatment. The mathematical model is based on the turbulent flow theory and it generates information about velocity conditions, the space distribution of the whirl energy and the current line distribution. The operating parameters are also determined with using this mathematical model.

View full abstract

In the KCB-S process—a modified AOD process—process gases are blown into a metal bath through bottom tuyeres and onto the bath surface by a lance. Depending on the distribution of the gas volume flow rates to the two blowing devices, differing reaction spaces and reaction areas are formed.
The change in the reaction areas for nitrogen absorption and their ratio in the metal bath and on the bath surface, including metal droplets, has been determined from an investigation into the reaction kinetics of nitrogen absorption under modified blowing conditions.
The investigation was carried out in the final stage of lance use with carbon contents of between 0.5 and 0.2%. The composition and volumetric flow rate of the process gases was varied during the decarburization of AISI 304. The reaction areas increase as the volumetric gas flow rates rise. The reaction area of the gas bubbles is considerably greater than that of the metal droplets and the bath surface. Owing to the higher reaction rate of oxygen with the metallic melt, rising gas bubbles have a smaller reaction area for oxidation than for nitrogen transfer. Using the test technique presented here it is possible to optimize partial reaction areas and thus process control.

View full abstract

Removal of P and Cr, which are impurities in a Fe-36%Ni alloy, was investigated by conducting experiments 30 kg scale on oxidation refining of a 36% Ni melt containing 0.1 mass% P and 0.4-1.5 mass% Cr with a CaO-SiO₂-CaF₂-Fe_tO slag. FeCr₂O₄ was identified of X-ray diffraction analysis of water-cooled slag sample just after the experiments. Therefore, the oxygen potential of the dephosphorization reaction was taken to be controlled by the formation reaction of FeCr₂O₄. The removal reaction of P and Cr occurred along the P-Cr equilibrium (Eq. (A)), and consequently high temperature treatment was effective in lowering the P content in the range of the present Cr contents.

2[P]+5/4FeCr₂O₄(s)=(P₂O₅)(l)+5/4Fe(l)+5/2[Cr] ......................(A)

The phosphorus distribution ratio (L_P) decreased with increasing Cr content in the metal. This was considered because the oxygen partial pressure was lowered by increasing the content of Cr. In addition, L_P increased with increasing in Ni content in the metal. It was concluded that this was because the oxygen partial pressure was increased with increasing Ni content mainly.

View full abstract

For the production by medium size bloom, some trials were held, at first on the test bloom CC, next on the proper CC, then the calcium treatment technologies for special steel bars and wire rods was established. The appropriate calcium treatment condition for each steel grade shows various features, because the equilibria of the alumina modification and the calcium sulfide formation by calcium addition are directly affected by the contents of elements in the molten steel concerning these reactions. It was found that these features can be well explained by the thermodynamic data reported recently.

View full abstract

Recently, continuously cast steel production is subject to complex user requirements and also to small lot sizes. To meet these requirements, a new tundish technology has been adopted in No. 4 slab caster in which a hot tundish is used repeatedly. As a result, refractory, energy and labor costs have been reduced, and higher productivity and higher quality have been achieved.
These quality improvements have been achieved during the start and end of casting and also during the ladle exchange in unstable casting conditions. The countermeasures taken for quality, especially for inclusions, in unstable casting conditions are reported in this paper.

View full abstract

White spots are defects that occur in VAR melted ingots which appear as non-etched bright areas especially in the superalloy macro-structure. In the present study, 'dirty' discrete white spots which are associated with inclusions were surveyed. It was made clear that the 'shelf' is an important source of this type of white spot. It was thought that the large fill ratio may allow the arc to reach the circumference of the molten metal pool enough to prevent the growth of a 'shelf'. To investigate this, VAR ingots of Alloy 718 were remelted at various fill ratios. All the materials made from these ingots were then tested by ultrasonic inspection and the results showed that the larger the fill ratio, the fewer the number of defects in each ingot.
Moreover, we found two kinds of inclusions associated with this type of white spots: one, made of fine Al₂O₃ particles and TiN particles mixed and agglomerated to form a large scale cluster; and the second, made of a relatively small number of TiN particles. The reason for the difference of these morphologies was discussed.

View full abstract

To improve cleanliness in steel requires a reduction in total oxygen content. The combined EAF-LF-RH-CC process of manufacturing utilized several technologies which reduce total oxygen content. In high carbon chromium bearing steels, total oxygen content has decreased annually and is now 5 ppm on the average.¹⁾ Along with reduced oxygen conent, reduction of the number of oxide inclusions larger than 10 μm has further extended the fatigue life of bearing steel, which was once considered close to the completion. The fatigue life in bearing steels with these smaller inclusions is 5 times longer than that of conventional steels.
A recent study showed that large inclusions can be reduced in size by preventing ladle contamination, air oxidation and flow out ladle slag.

View full abstract

Twin shell type DC arc furnace with scrap preheating system was supplied by Daido to Takunan Steel in Okinawa. This furnace is equipped with one power source and two furnace bodies and also high efficiency scrap preheating function. The advantages, functions and operation results are reported in this paper.
The summary of the operation results obtained up to now are as follows;
1) Reduction of tap to tap time by 11 min comparing with single furnace,
2) Improvement of maximum 35 kWh/t in electric power consumption by scrap preheating, and
3) Reduction of 18 kWh/t through effective utilization of post combustion.
It should be noted that these operation figures are under the restriction of electric power source. Therefore, not less than the improvement figures described above are expected when normal electric power is inputted in near future.

View full abstract

Low hydrogen content in the steel is required to steels, such as special bar quality, crank-shaft, railway wheel, structural steel etc. By using the conventional hydrogen analyzer, quick response and necessary accuracy can not be obtained. Long time RH treatment was essential. Eventually, RH vessel life has been shorter. Therefore, "dynamic-dehydrogenation model" has been developed to attain both an optimum operation of RH-degasser and a security for dehydrogenation of hydrogen content after treatment. Followings are features of the dynamic dehydrogenation model:
1) Prompt adjustment of dehydrogenation by using direct hydrogen measurement system.
2) Correspondence to hydrogen pick-up from added alloys.
3) The K_H depends on many factors, for example, RH conditions, chemical component and other characteristics. It is well known that Q depends on the RH operating conditions.
4) The K_H clearly shows that maintenance of the snorkel diameter is important to keep the enough dehydrogenate rate. Also K_H is decreased sharply in the low [H] region lower than 1 ppm.
With this model, treatment time was reduced by 6 min and the hydrogen content was controlled within the accuracy of ±0.2 ppm.

View full abstract

To control the composition control of inclusion in Mn-Si killed steel, elucidation of the thermodynamics of the reaction between trace amounts of Al and manganese-silicate inclusions is very important. SIMS (Secondary Ion Mass Spectrometry) was adopted for the microanalysis of dissolved Al in the steel. Based on the results of fundamental experiments, the thermodynamic behavior of the slag-metal-inclusion reaction was discussed, using the analytical value obtained by SIMS. Trace amounts of dissolved Al in the steel could be measured using SIMS without influencing the inclusion. The inclusion compositions in the molten steel were predicted by trace amounts of dissolved Al in Si-Mn killed steel.

View full abstract

The use of thermodynamic data in describing calcium modification of aluminium oxide inclusions has been summarised and reviewed. The majority of the published Al-S equilibrium diagrams, based on the following reaction:
3CaO+3S+2Al=3CaS+Al₂O₃,
vary significantly due to different sources for thermodynamic data, especially for the activity of CaO and Al₂O₃. Using ThermoCalc, the activities of CaO and Al₂O₃ have been calculated and compared to the published data. Calculations in the present work pertaining to the molten range of the system are in good agreement with the most recently published experimental data of Fujisawa et al.
Based on the assessed thermodynamic data and observed phenomena during calcium modification of inclusions in steel melts of moderate sulphur content, a model for alumina modification by calcium treatment has been developed. According to the model, the reactions progress by the following sequence until the activity of Al₂O₃ becomes so low that precipitation of CaS occurs:
Al₂O₃⇒CA₆⇒CA₂⇒CA⇒CA_x(liquid)

View full abstract

The conditions of formation of inclusions in steel during ladle refining in MgO-C lined ladles were investigated. From the experimental results as well those of a thermodynamic study conducted in parallel, the following conclusions can be made about the reaction mechanism between the MgO-C refractory and aluminum deoxidised molten steel:
- an internal oxidation-reduction occurs in the MgO-C refractory and aluminum deoxidised molten steel during ladle refining;
- the formation of a thin oxide layer at the interface is due to the reaction between magnesium vapour and aluminum dissolved into the molten steel and the CO(g) generated by the reaction between MgO and C in the crucible walls;
- the oxide inclusions formed in the steel have been shown to consist mainly of MgO, Al₂O₃ or a mixture of them;
- some of the finest inclusions are believed to be formed as secondary inclusions during cooling and solidification of the steel, and are connected only with the diffusion of magnesium from the crucible to the molten steel.
- thermodynamic calculations indicate that during vacuum refining, as the pressure decreases, the chemical compatibility of carbon-bonded magnesia for ladle lining decreases.

View full abstract

The inclusions in stainless steel are of different shapes on a micrograph: circle, triangle, square, bar and so on. A new method to identify the shape of an inclusion was developed with the aid of image processing equipment. Two morphological parameters of y=L²/(4Aπ) and x=πD²/(4A) were selected. On a two dimensional diagram of x and y, the respective area of each shape was determined theoretically and experimentally. The inclusions in specimen taken from molten steel, CC slab and rolled sheet were examined on optical micrographs of 1000 magnification and classified using this diagram. The inclusions in molten steel were more than 96% circular, those in CC slab, about 90% circular and those in rolled sheet, 80% circular. In the last specimen, squares accounted for about 10% and bars, 7%. The inclusions were analyzed by EPMA. The circular inclusions were mostly of oxide, whereas the elongated inclusions were of MnS based complex. Sizes of the circular inclusions were also determined, and the mean diameter in molten steel, CC slab and rolled sheet was 1.4, 2.0 and 1.0 μm, respectively.

View full abstract

Model experiments on the particle coagulation in NaCl-aqueous solutions were made in an agitated vessel to learn the coagulation rate of inclusion particles in liquid steel. Polystyrene-latex particles, silica particles and alumina particles were tested at various agitation speeds, and coagulation coefficients were obtained by comparing the observed coagulation rate with the predicted rate by the Saffman-Turner model on turbulent collision. The Higashitani theory which considered the effect of van der Waals force and viscous resistance force on the coagulation coefficient was applied to obtain the Hamaker constant from the measured coefficients. Based on the measured Hamaker constants, the coagulation coefficients of inclusion particles in liquid steel were predicted for several steelmaking processes with different stirring intensities.

View full abstract

Laser microprobe mass spectrometry (LAMMS) was used in a state analysis of nonmetallic inclusions and precipitates in a high purity steel. At the same time, SEM-EDX and XPS were applied for morphological, elemental, and state analyses, respectively. As an example of a pure steel sample, Ti-added ultra low carbon steel (interstitial free steel: IF steel) was selected for the analysis. Inclusions in IF steel were prepared by an acid extraction method for oxide analysis, an electron beam melting method for in situ analysis, and non aqueous solvent selective potentiostatic electrolytic etching method for in situ analysis on the electrolyzed surface for oxide and nitride. LAMMS analysis detected Ti oxide and Zr oxide as trace elements in addition to alumina which was determined as a main component of inclusions by EDX analysis, and, it was clear that each oxide existed independently. LAMMS analysis of TiN gave only a spectrum of Ti oxide related peaks but not of Ti nitride, while Ti existed as nitride thermodynamically and analytically by XPS. Hence, in LAMMS analysis, careful interpretation of spectra is needed for such inclusions as nitride, which does not show its specific spectra.

View full abstract

Chemical state of inclusions in IF steel was studied by EPMA and Auger electron spectroscopy (AES). The chemical state of Al and Ti was mostly Al₂O₃ and TiN, respectively. X-ray images of Al₂O₃ inclusions in IF steel (as cast) appearing on the mechanically polished sample surface were taken by EPMA. The inclusions could be classified into two types. One is a small inclusion (1 μm) having a small Al₂O₃ core completely covered with TiN, and the other is a large one which consists mostly of Al₂O₃ and is partially covered with TiN. To estimate the state of inclusions in a bulk sample quickly, sample was melted by electron beam and floated inclusions (raft) were analyzed (EB method). Floated inclusions of IF steel were Al₂O₃ and those containing Ti inclusions were rarely found. Observed inclusions containing Ti showed very vague X-ray images and some of them showed a melted appearance. Care must be taken to examine the state of inclusions in IF steel by EB method.

View full abstract

In order to produce clean stainless steel, the inclusion softening technology and the large inclusion removal technology are important. Then, to elucidate the origin of inclusion in slabs and billets, experiments were conducted, adding a tracer to AOD slag and tundish slag. To control the composition of inclusions, a thermodynamic calculation model was developed. And, to elucidate the characteristics of floating separation of large inclusions in the tundish, tests using the water model were conducted. The following results were obtained:
(1) The inclusions in stainless steel were caused by the AOD slag suspended in the molten steel which is not removed in the tundish but carried over into the cast shapes.
(2) The thermodynamic calculation model in which the suspended slag is considered was developed. The composition of inclusions could be predicted by this model. In the actual process, the precipitation of MgO-Al₂O₃ spinel in inclusion was prevented by reducing the MgO and Al₂O₃ contents of the AOD slag.
(3) For the removal of large inclusions in the tundish, an increase in residence time is particularly effective. In the actual casting, the quality of stainless steel products were improved by controlling the weight of steel and the residence time in the tundish.

View full abstract

The effect of impurities such as Al and Ca on the type and amount of inclusions formed during deoxidation of liquid steel has been studied on a laboratory scale. The microstructure of different grades of ferrosilicon and the occurrence of impurities (Al, Ca, Ti and P) were also studied. The results show that impurities are present primarily in intermetallic phases in the ferrosilicon and that oxide inclusions are very rare. The grade of ferrosilicon has a direct impact on the composition and amount of inclusions in the liquid steel. An increased content of aluminium results in the formation of alumina inclusions, while high purity ferrosilicon results in the formation of silicates which are removed more slowly from the melt.

View full abstract

In order to understand the reoxidation behavior by slag in the steelmaking process, the oxygen potential of the slags in the ladle was estimated quantitatively adopting the regular solution model by Lumsden. The activity of iron oxide in slag was determined by an automatic facility employing an electro-chemical technique which incorporates stabilized zirconia as the solid electrolyte. The conversion equation for the activity of Fe_tO has been deduced from the correlation between the measured values of Fe_tO activity and the calculated ones by the regular solution model. The activity coefficient of Fe_tO for the slag in refining process was estimated to be 2.308 to the mole fraction of Fe_tO, and 0.0228 to the mass% total Fe on average.

View full abstract

Inclusions of CaO-SiO₂-Al₂O₃-MgO-TiO₂ in type 304 stainless steel were found to be formed by Al and Ti deoxidation on nucleation sites of suspended AOD slag droplets in molten steel. The crystallized phases in inclusions, i.e., MgO·Al₂O₃ spinel, CaO·TiO₂ (perovskite) and TiO₂ (rutile), showed very high melting temperature and formed in parallel to the decrease in molten steel temperature. The major factors affecting MgO·Al₂O₃ spinel crystallization are aluminum content in steel and weight fraction of MgO in AOD slag. Also, decreasing the AOD slag basicity (%CaO/%SiO₂) is the most effective way to reduce the crystallization of the harmful CaO·TiO₂ phase in inclusions.

View full abstract

The cooperative research was conducted on conventional and new estimation technique of inclusions to understand their merits and demerits. Three kinds of samples examined here were Al-killed ultra low carbon steel, high carbon bearing steel and stainless steel. They were analyzed by conventional extraction and two new methods: electron beam melting and photo scattering method.
The results were evaluated from the viewpoint of quantity and size distribution of inclusions to determine both the applicability and shortcomings of each technique.

View full abstract

The rate of shape control of alumina inclusions with calcium in line pipe steel for sour service is determined by a 30 kg high frequency induction furnace. The shape control of alumina inclusions proceeds with the increase in time after calcium addition. The degree of shape control is increased with the increase of stirring energy supplied and the decrease of the time from aluminum addition to calcium addition. Applying the unreacted core model to experimental results, the rate determining step in the shape control of alumina inclusions is believed to be calcium diffusion in the calcium-aluminate product layer.

View full abstract

A major objective of calcium treatment in steelmaking is to improve product performance through inclusion modification. To establish inclusion modification, shape and composition of inclusions were monitored using SEM during and after addition of calcium wire into 2 kg heats of steel under various conditions of Ca consumption and Ca addition pattern. A mathematical model of the kinetics considering the evaporation of calcium and the reaction between inclusions and melt was developed to thoroughly examine these phenomena in calcium treatment. The calculated results were compared with the observed ones which were obtained in experiments.
(1) Shape and composition of inclusions changed from a spherical SiO₂-MnO system to almost lumpy Al₂O₃ with the addition of aluminum, and to spherical CaO-Al₂O₃ including CaS after addition of CaSi.
(2) The process and consumption of calcium were found to affect the change of CaO and CaS contents of inclusions.
(3) A mathematical model was developed assuming that vaporization rate of calcium from the melt and rate of reaction between melt and inclusions can be described by equations of the first order reaction, and that the size and number of inclusions remain constant during treatment.
(4) The calculated [Ca] and CaO, CaS contents of inclusions showed good agreement with observed ones regardless of the conditions of CaSi addition.

View full abstract

A mathematical model of the reaction between mold flux and molten steel has been newly developed to estimate the change in composition of the mold flux during the continuous casting of steels and to design an optimum flux composition. The model takes into account mass transfer in the molten steel and molten flux and equilibrium reaction at the interface. On the reaction, the activities of the practical mold flux components like CaO, SiO₂, Al₂O₃, CaF₂, Na₂O are estimated by a statistical thermodynamic model of slag, called cell model. Comparison of calculated composition change with data obtained in the practical casting experiment of high titanium steel showed that the mathematical model is useful for the estimation of the reaction between the mold flux and the molten steel and for the design of the mold flux composition to minimize change in composition due to that reaction.

View full abstract

Stress-strain data of continuously cast (CC) steel at high temperatures available in the past are based on uni-axial tensile tests in a uniform temperature field, the data being utilized for thermal stress analysis to simulate crack formation on a CC-slab surface. New equipment for hot tensile and bending tests has now been developed by the authors that allows determination of more realistic stress-strain behavior of the solidifying shell subjected to temperature gradients similar to those in a CC-mold. Key factors affecting crack formation have been studied with this equipment. Fine cracks were found to form at the hot side of the tensile test-piece during straining. These cracks are initiated at grain boundaries which are located on the previous interdendritic solute-segregated region. The critical strain for the crack formation is independent of the strain rate within the range of 5–50×10^-4s^-1, depending on the carbon content, measuring 2% for peritectic carbon steels, 1.5% for ultra low carbon steels and about 1% for low, middle and high carbon steels.

View full abstract

Initial shell formation and surface quality in continuous casting are affected by thermal and mechanical stresses which may lead to crack formation during solidification. Thus, detailed knowledge of high temperature strength properties as function of steel composition is required under given conditions. To simulate shell straining by tensile force perpendicular to the main dendrite growth axis, the "Submerged Split Chill Tensile" (SSCT) test was (and still is) further developed. The present report details updated results for ferritic and austenitic iron, and compares different chill materials. The determined effect of P-content on shell strength is quantified. Primary dendrite arm spacing without and with load appears to indicate that tensile elongation is not uniformly distributed over the test length.

View full abstract

Engulfment behavior was observed in situ using transparent organic materials and the critical growth velocities were determined. The critical growth velocity of the SCN-acetone alloys increased with increasing temperature gradient and decreasing alloy concentration, while that of the salol-acetone alloys was independent of temprature gradient. Not only the interfacial energies but also the interface shape and morphology contribute to the engulfment/pushing behavior. The present result suggests that conventional alloys show the engulfment behavior of inclusions/insoluble particles as well as the organic alloys.

View full abstract

The effect of carbon content of steel melt on the heat transfer in a continuous casting mold was investigated in a laboratory scale mold by simultaneous measurements of the heat flux during solidification of the melt and shrinkage of the resulting shell during solidification and cooling. For peritectic medium carbon steels, the heat flux on the mold surface was subject to an anomalous decrease, since the surface roughness of the shell was substantial, providing an air gap at the shell/mold interface. For ultra low carbon steels, however, the heat flux was large in spite of the fact that surface roughness was also significant. For peritectic medium carbon steels, the deformation of the shell caused by the shrinkage due to δ/γ transformation during and just after solidification is responsible for the formation of the surface roughness, resulting in the heat transfer anomaly. On the contrary, the surface roughness for ultra low carbon steels is formed later on the shell after solidification during cooling, and hence it does not decrease the heat flux at the initial stage of solidification.

View full abstract

A three-dimensional fully coupled turbulent fluid flow, heat transfer and solidification model for a twin-roll caster of steel has been developed. The model enables different metal delivery configurations and their effect on solid shell thicknesses to be evaluated. A typical tubular nozzle with horizontal outlets directed towards the side dams, as well as a vertical slot nozzle, have been simulated. The results show that the tubular nozzle leads to a non-uniform thickness of the twin solidifying shells along the width of the rolls. These results are in agreement with measurements of heat fluxes to the roll surfaces. A similar model has been constructed for a horizontal single belt caster with extended nozzle delivery systems, with and without flow modifiers. It is shown that the flow modifier has a positive effect on the flow patterns and on steel shell formation.

View full abstract

Conductive thermal resistance from solidifying steel shell through mold flux to continuous casting mold was investigated in bench tests. Thermal diffusivities of six mold fluxes used in continuous casting of steels were determined in glassy, crystalline and molten states with a laser flash method. Interfacial thermal resistance between the solidifying mold fluxes and copper mold was also investigated by pouring the molten fluxes onto copper mold and measuring transient heat transfer to the mold. The surface morphology of solidified fluxes in contact with the mold was observed with a stylus type profile meter and related to the interfacial thermal resistance. Thermal diffusivity of glassy state (ca. 4.5×10^-7 m² s^-1) and molten state (ca. 5×10^-7 m² s^-1) was insensitive to chemical composition and temperature, while that of crystalline state was higher (5.8–7.8×10^-7 m² s^-1 at room temperature) and decreased with increasing temperature, reaching similar value to that for glassy and molten state at high temperatures. The surface morphology consisted of large cells on which small cells were superimposed. Observed interfacial thermal resistance increased with increasing height of the large cells, contributing more than 50% of the total conductive heat transfer resistance and being a significant factor to control the heat transfer between the solidifying steel shell and the mold.

View full abstract

A computer software package has been developed to simulate temperature, shell growth and phase transformations in continuous casting of steel. The package includes two earlier developed models, a heat transfer model and a thermodynamic–kinetic phase transformation model. The phase transformation model takes into account the effect of alloying and cooling on the phase transformation temperatures and on this basis, on important, solidification related thermophysical properties, enthalpy and thermal conductivity, generated by the model. Hence, while affecting these properties, the phase transformations have a special influence on the heat transfer in the strand. The output of these coupled calculations are temperature distribution, phase distribution (liquid, delta ferrite and austenite, and approximately, proeutectoid ferrite, pearlite, bainite and martensite) and volume distribution through the strand, and also, hardness distribution on the cross-section of the strand at room temperature. Characteristics of the coupled model and some results of calculations are presented.

View full abstract

In the continuous casting process argon gas is usually blown into the tundish to suppress the formation of non-metallic inclusions which cause blocking of the nozzle. The gas which comes into the mold with the molten steel through the tundish nozzle, however, is sometimes trapped as small bubbles at the solidification front in the mold. It is well known that these trapped bubbles cause surface defects of the products. Many simulation studies have examined the trapping mechanisms of gas bubbles in the subsurface of the cast, however, there have been very few experimental studies. In this paper, we used paraffin liquid instead of molten steel to observe the bubble behavior as it is a more convenient substance.
The testing conditions were as follows: the temperature of paraffin liquid was 329K (the melting point is 325K), the tilting angle of solidification front in the direction of bubble flow was 30 degrees, measuring time was 30 sec, the average diameter of a bubble was about 1 mm and frequency of bubble flow was about 18 sec.
It was determined that decrease in the temperature of cooling water, which means an increase in the solidification speed of paraffin, causes an increase in the number of trapped bubbles.

View full abstract

The control of molten steel flow in continuous casting molds is essential to increase the casting speed and produce clean steel. This requires accurate calculation of the velocity of the molten steel flow. Since there is no reliable velocimeter, water model experiments and numerical simulations have been carried out to predict this velocity, however, no experimental confirmation of such predictions has been made in a practical situation. In this study we developed a probe capable of measuring the velocity of a molten steel flow. The measurement principle is based on the linear relation between the shedding frequency of Kármán's vortex streets generated behind a circular cylinder and the velocity of fluid flow approaching the cylinder. The fluid flow velocity can be determined by measuring the oscillation frequency of the cylinder because the cylinder oscillates at the same frequency as the shedding frequency of Kármán's vortex streets.

View full abstract