The oxidation rates of silicon in molten high carbon iron by synthetic slags containing FeO, pure FeO and LD slag were examined at 13001500°C. The results were analyzed with a coupled reaction model. The oxidation rate of silicon was more rapid with higher FeO content in slag, because the interfacial oxygen activity and the transport rate in slag were increased with increasing FeO content in slag. For the LD slag, the silicon oxidation proceeded simultaneously with the reduction of MnO in slag, whereas the dephosphorization of the metal was not observed with this slag. As an oxide CO was more thermodynamically stable than silica under the present experimental conditions, but the oxidation of silicon proceeded prior to the decarburization at the early stage of reaction. This result suggested that the decarburization was controlled by the chemical reaction, while the oxidation of silicon was controlled by the mass transfer. After the silicon was oxidized to certain contents, which were depended on the experimental conditions, the decarburization predominantly proceeded.This behavior was well explained by the coupled reaction model. The analysis of experimental results with reaction model suggested that the oxidation rate of silicon was controlled by the transport rate of FeO in slag.
The flow pattern of air-water two-phase flows in a vertical, circular pipe was experimentally investigated. The wettability of the pipe was changed by surface treatment. The original acrylic pipe had a contact angle θc of 77°, while two pipes after the surface treatment had θc of 36°and 104°, respectively. The flow patterns in the pipes having θc of 36° and 77° were the same and agreed favourably with previously published results for pipes of good wettability (θc<90°). On the other hand, different types of bubbly and slug flows appeared in the pipe of θc=104°, being caused by attachment of bubbles and slugs to the wall of the pipe. The bubbly flows for θc of 104° were classified into two categories and slug flows for the same contact angle were classified into three categories. The boundary between the bubbly and slug flows, however, was hardly dependent on the wettability of the pipe.
The carburization reaction effected the rapid reduction and the separation process by means of the rapid heating of carbon composite iron ore pellets at the high temperatures. Therefore, it was considered that the carbon content of the pellet effects the behavior of this reaction significantly. While carbon content of the pellet was increased, the melting and separation temperature was decreased. But when atmospheric temperature reached less than 13301380°C, the melting and the separation of metallic iron was repressed though the carbon content was increased. It was considered that this phenomenon was caused by decrease of the carburization speed, which accompanied with the stagnancy of the solution loss reaction and the generation of CO gas.
In order to examine the effect of several parameters on steel scrap melting in molten steel, the heat and mass transfer coefficients were evaluated. The change in radius of steel scrap with time was obtained by the experiment with 30kg steel bath. The simultaneous heat and mass transfer equations were numerically solved with the proper parameters which were obtained by adjusting the computed radius change to the experimental one. The results are summarized as follows; (1) The heat transfer coefficient was evaluated as 27.7 to 77.2 kW/m2K, and the mass transfer coefficient, as 0.83 to 2.08×10-4m/s. (2) The heat transfer coefficient decreases with an increase in radius. (3) Both coefficients are independent of carbon content in steel bath. (4) Both coefficients increase with an increase in temperature. (5) The dimensionless correlation for heat and mass transfer are given as Nu=0.017Re0.8Pr1/3, Sh=0.017Re0.8Sc1/3
The optimization of measuring condition for the determination of carbon and nitrogen in steels by means of the glow discharge mass spectrometry (GDMS) has been studied. A double focusing glow discharge mass spectrometer, Fisons Instruments (UK) VG 9000, was used. A disk sample were dry-polished with 120-grit zirconium oxide endless-paper and preliminary exposed to glow discharge for 30 min in the discharge cell. The optimum values of discharge parameters and the suitable size of tantalum front mask were examined. The relative sensitivity factor (RSF) of each analyte was evaluated by analyzing the five standard reference materials: IARM CRMs and the high alloy steels prepared in our Institute. The average RSF-values obtained were 2.493 for carbon and 26.34 for nitrogen. The regression analysis for nitrogen showed that the slope of regression line for each alloy types was different. The concentration of matrix element required was determined by the fundamental parameter/X-ray fluorescence spectrometry. The GDMS analytical value of each analyte in high tensile steel HT80 was in good agreement with chemical analytical values.
In order to produce Tapered Leaf Spring (TLS), a new Controlled Taper Rolling (CTR) method is developed. This process is a combination of the edge rolling used grooved rolls and the flat rolling. It is very important in this method to predict the deformation properties such as spread precisely. For the first pass, quadratic equations which relate the reduction in width and the mean thickness increment, is obtained. For the second pass, it was clarified that Ekelund's equation was good agreement with the experimental values even if the rolling diameter changed from 125mm to 720mm and the rolling temperature changed from 1073K to 1223K. The actual pass schedule to produce TLS is designed by using the data obtained from these steady sate rolling. It is afraid that there is large difference in the deformation and loading properties between the steady state rolling and the non-steady state rolling such as tapered rolling. However, it was clarified that the spread is not effected and projected contact length is effected in tapered rolling. The pass schedule based on the steady state rolling can be applicable and the rolling load and the rolling torque might be predicted by concerning the change of the projected contact length.
The authors developed a compact continuous mill named satellite mill. In the satellite-mill rolling, longitudinal compressive stress decreases the elongation significantly and promotes transverse metal flow, so that this mill can produce more complex profiled wires efficiently than conventional mill. In this study, the influence of pass schedules is studied in rolling of U-shaped and H-shaped wires.In addition, the effect of roll surface roughness is studied, with two central rolls, which have different roughness on the groove. The obtained main results are as follows, 1) In U-shaped wire rolling, the elongation decreases with increase in the reduction at the first stage. In H-shaped wire, the influence of pass schedule on elongation is not obvious. 2) The rolling characteristics are sensitive to the surface roughness of the groove. Smooth surface finish of the groove increases the elongation and decreases the rolling force considerably. 3) The hardness in cross section of products with smoother roll is lower and more uniform than that with rougher roll.
Steel specimens containing 0.01, 0.1 and 0.6 mass% Si were coated with 110μm nickel layer by means of electroless plating and then galvanized in molten Zn at 733K for various periods of time. Formation and growth kinetics of alloy layer on the surface of the specimens was examined. Ni-Zn alloy layer composed of δ, γ and γ1, phases was formed on the surface of the electroless Ni deposit. Its thickness increased with dipping time according to a paraObolic rate law. After the Ni deposit was consumed by the diffusion with molten Zn, Fe-Zn alloy layer composed of δ1, phase or δ1 and γ phases appeared between steel substrate and Ni-Zn alloy layer. As Fe-Zn alloy layer grew, thickness of γ and γ1 phase layers decreased and δ phase layer peeled off after disappearance of γ and γ1 phase layers. Subsequently, in the case of the specimens containing 0.1 and 0.6 mass% Si, δ1 phase layer cracked along the boundaries of palisade structure and molten Zn penetrated into the layer near the substrate, while ζ phase layer appeared on the surface of δ1 phase layer of the specimen containing 0.01 mass% Si. The elctroless Ni plating was found to be effective to prevent the abnormal structure during hot dip galvanizing of silicon-containing steels.
Recently challenge of further grain refinement has been made, which aims to refine the ferrite grain size down to 1μm. It has been reported independently by two differing research groupsu1, 2) that a heavy deformation after accelerated cooling can successfully produce ultra fine grains as small as 1μm in diameter for relatively low cost steels. The aim of present study is to reveal how chemical composition of steels affects the grain refinement by such thermomechanical process. Compression experiments are carried out at various temperatures (down to 530°C) and at various cooling rates (up to 50K/s) before deformation. The interplay between chemical composition of steels and deformation condition on microstructures is discussed with focussing on CCT diagrams. The process window required for ferrite grain refinement down to 1μm must satisfy the following conditions; (1)steels with relatively low quench-hardenability, (2) accelerated cooling before deformation, (3) deformation at deeply supercooled austenite region, (4) heavy deformation.
Effect of Cu addition on microstructure was investigated in Fe-Cr-Cu martensitic alloys by means of optical and transmission electron microscopy, and X-ray diffractometry. Lath-martensitic single structure can be obtained in 9mass%-Cu steels water-quenched from the solution treatment temperature of 1273K. The morphology of martensite-laths is independent of Cu content in the steels with 4 mass% Cu or less. Dislocation density in martensitic structure, however, is enhanced with increasing Cu content. On the solution treatment, Cu atoms markedly segregate at austenite grain boundary, and the Cu content increases to about 50 at% (53 mass%) in the case of 9 mass% Cr-4 mass% Cu steel. This results in the retardation of grain growth of austenite at the solution treatment temperature probably due to the grain boundary dragging effect by Cu atoms and/or the decrease in grain boundary energy. The austenite grain refining by Cu addition also causes the refining of sub-structure within prior austenite grains; martensite-block and -packet structure.
The influences of the gas jet spouted from the atomizing nozzles, gas and metal flow rate ratio and gas pressure on the median diameter and particle size distribution were investigated. The median diameter of atomized powder varies in proportion to the square root of Ml/Mg value when the atomized powder was formed at various melt or gas flow rate under the same other conditions. The median diameter varies inversely proportional to the mean gas velocity when the powder was atomized using various atomizing nozzles which have different dimension i.e. the apex angle and gas jet nozzle configuration. Approximately the same median diameter powder was obtained at the same gas flow rate regardless of the gas jet pressure. An experimental formula was obtained as follows; d50p/Dl=63.6√(Ml/Mg)·(σl/(Dl·ρl·V2g))·(vl/vg) where, d50p: Median diameter of atomized powder (m), Dl: Diameter of melt delivery tube (m), Ml: Mass flow rate of molten metal (kg/s), Mg: Mass flow rate of atomizing gas (kg/s), σl: Surface tension of molten metal (N/m), ρl: Density of molten metal (kg/m3), Vg: Gas velocity (m/s), vl: Kinematic viscosity of molten metal (m2/s), vg: Kinematic viscosity of atomizing gas (m2/s). The fraction of coarse powder increased when the atomization was performed with a concave gas velocity distribution compared to a convex one which is the ordinary case. The calculated particle size distribution coincides roughly with the experimental results. It was shown that the particle size distribution depends on the gas velocity distribution.