Molten oxide containing iron oxide was reduced with coal, char as well as graphite at temperatures 1300, 1350 and 1400°C. It is revealed that volatile matter in coal accelerates reduction rate of the oxide. Apparent rate constants can be expressed empirically as follows : for coal containing volatile matter : In k1 = -5.00 × 103/T-4.08 (s-1) for char and graphite : In k3 = -11.0 × 103/T -1.73 (s-1) When the surfaces of coal, char as well as graphite are surrounded with foam of the molten oxide, the reduction comes to a halt while carbonaceous material and molten iron oxide remain in the reaction system.
Two kinds of swirl motions are observed in a cylindrical bath with centric bottom gas injection. One is induced byapproximately periodical generation of bubbles at the nozzle exit. The other is caused by hydrodynamic instability of a large scale ring vortex enclosing the bubbling jet. The two swirl motions occur for HL/D_??_1 and HL/D_??_2, respectively, where HL is the bath depth and D is the bath diameter. Air-water, mercury-nitrogen and Wood's metalnitrogen systems are used in the present cold model experiments. The following three subjects for the former type of swirl motion are treated. (1) Time from the start of gas injection to the initiation of swirl motion, being called the starting time of swirl motion. (2) Time required by the swirl motion to damp out after the stopage of gas injection, being termed the damping time of swirl motion. (3) Amplitude of swirl motion at the side wall of the vessel. Empirical correlations of the starting time and damping time are derived. The amplitude of swirl motion exhibits a peak value at HL/D between 0.4 and 0.5.
A new method of melting and holding of a material with low thermal and electrical conductivities is proposed toovercome the problems of long heating period and localized overheat in a charge. In this method, a water cooled induction coil is directly used as a crucible, and the elements heated by a high frequency magnetic field are submerged into the charge. Experiments have been conducted on glass under various operating conditions, i.e. the coil current, the frequency and the radius of the heating elements. A mathematical model is developed to calculate the process variables such as the heat generation rates in the coil and the heating elements, and the temperature distribution in the charge. The calculated process variables are compared with the observed ones. The considerably good agreement between the theoretical and experimental results indicates the validity of the model. The heat generation rate in the heating elements is determined by the ratio of the radius of the heating element to the electromagnetic penetration depth and has the maximum value at the ratio of 1.8.
Kinetic studies have been made on SiO2 inclusion removal from molten Cu to Li2O-SiO2-Al2O3 slag under Ar gas injection stirring condition. The effect of gas injection on the inclusion removal rate has been investigated. Experiments were done at 1523K. The gas flow rate of injected Ar was in the range of 360cm3/s under the conditions of 1523K and 1.013 × 105Pa. The crucible diameters were 30, 40, 60, and 75mm. During each experimental run, the frequency of bubble formation was measured by means of a pressure pulse technique. The bubble size ranged from 7 to 18mm. From the [mass%O]T-time relation, the rate constant of inclusion removal ko, was obtained. It is shown that koincreases with increasing Ar gas flow rate, Vg. The rate of inclusion remove increases with increasing gas-metal interfacial area. A mathematical model is developed to explain the experimental result. In the model, the inclusion particles are assumed to adhere to the gas bubble-metal interface during the bubble formation and ascent periods. The apparent mass transfer coefficient, km, for inclusion removal obtained experimentally is independent of the gas flow rate and the crucible diameter. Thus, it is considered that the inclusion particles are removed from the melt mainly through adhesion to the gas bubble-metal interface.
Study was made on the desiliconization and dephosphorization reaction kinetics at various size equipments. Coupled reaction model was employed to analyse the rate determining step of desiliconization and dephosphorization reaction. By model calculations, the rate determining step of desiliconization reaction is the mass transfer process in the hot metal and mass transfer both in hot metal and slag for dephosphorization reaction. The mass transfer coefficient in hot metal km proportionally increased with 0.7 power of parameterε d/2c, independently on the equipment size, where, ε, deis the stirring energy of hot metal and the diameter of vessel, respectively. Apparent rate constant of dephosphorization kp' =ln ([%P]i/[%P]f)/tf increased from 0.10min-1 to 0.33min-1 when stirring energy ε' and the rate of oxygen supply Vo2 was increased under the optimum relationship between ε' and Vo2 which is predicted by the model calculations, where [%P]i and [%P]f is initial and final phosphorus content and tf, refining time. This optimum relationship was expressed as the empirical equation, ε'=1.51 Vo22 +3.31Vo2 under the condition of this study.
A new decarburization model for the vacuum degasser was constructed. In this model, the mass transfer of carbon andoxygen in the liquid phase, the mass transfer of CO in the gas phase and the chemical reaction rate at the interface were taken into account for the rate controlling steps. Also, as the decarburization sites, the Ar bubble surface, bath surface, and the CO bubble formation at inner sites were considered. This model was verified by the correspondence of the calculation results with the experimental results of the small scale tests and applied to the various RH degassers. The following results were clarified : 1) Decarburization at inner sites mainly occurs in the initial stage of the decarburization process (Stage I ), and decarburization at the bath surface becomes predominant in the final stage of the decarburization process (Stage II). 2) The reaction in Stage I is mainly governed by the circulation rate and evacuation rate. 3) The evacuation rate has a smaller influence on the reaction in Stage II. In this stage, it is essential to increase the circulation rate and to increase the effective reaction area for the decarburization at the bath surface by inducing violent surface agitation.
In order to develop of mold powder for high speed slab type continuous casting, viscosity of molten flux and melting rate are investigated. The results can be summarized as follows; 1) Viscosity of molten flux is estimated from the equation as a function of anion cation interaction parameter and net work parameter. 2) Melting rate is estimated from the equation as a function of carbon content per unit volume and carbonate content. 3) The improved mold powder succeeded in casting at 5m/min speed.
We investigated on the shear adhesion strength of galvannealed steel (GA). High adhesion strength of GA with less than 7 mass% Fe, 11 mass% Fe or more was obtained. A structure of GA coating/substrate interface on adhesion strength was aimed. The observations were carried out by transmission electron microscope (TEM) and scanning electron microscope (SEM). Γ phase remarkably grew at the coating/substrate interface of GA with over 11 mass% Fe. The ledge with pitch of approximately 100nm was formed at Γ/αFe interface. It seems that Γ phase grows by lateral growth mechanism. On the other hand, epitaxy between αFe and Γ phase was not found out. From the above mentioned results, it was proposed as a model that interface adhesion strength of GA was varied with geometrical shape of the interface.
Formation and growth kinetics of Fe-Zn intermetallic phases on surfaces of Fe specimens in molten Pb containing 10 mass% Zn at 613 to 693 K were investigated. ζ phase layer was first formed on the surface and then palisade δ1, Γ, Γ1 and compact δ1 phase layers were formed successively with the lapse of annealing time. Incubation times of these phases decrease with an increase in temperature. Growth of the intermetallic phases does not follow the parabolic rate law, although their thickness increases with annealing time. The formation order, incubation times and growth kinetics of the intermetallic phases are much the same as those in Fe-Zn diffusion couple. The reason why such a result was obtained can be explained on the basis of the local equilibrium concept in reactive diffusion, because value of Zn chemical potential in the molten Pb-Zn is nearly equal to that in solid Zn.
The mechanism of descaling at high temperature was investigated from the analysis of operation factors in plate rolling shop and the study of scale properties oxidized at high temperature. The descalability of Si-killed steels was inferior to semi-killed steels, because their sensitivity of scale cracking on air cooling was more than on semi-killed steels. The state of cracks in scale on air cooling controlled the descalability of steels. The oxide scale of Si-killed steels including large pores cracked easily but these cracks usually stopped at large pores which were at nearly half thickness of scale. At descaling by the hydraulic scale breaker, the cracks propagated in the scale parallel to the iron substrate and therefore the lower parts of scale thickness remained on the steel slab surface. On Si-killed steels, the fayalite, Fe2SiO4, was formed in layers at the interface of scale/iron. Under 1170°C, which was eutectic point of Fe2SiO4 and FeO, the fayalite became solid, and then the descalability became worth owing to the increase in adhesion of scale to iron substrate.
Medium carbon steels added 2.0% Si or 1.5% Si-30ppm Ca with 1400N/mm2 tensile strength show high resistance to delayed fracture. It is important to clarify the hydrogen occlusion behavior to delayed fracture of steels with high resistance to delayed fracture. The hydrogen diffusion coefficient and hydrogen content were measured by electrochemical permeation technique, and hydrogen evolution content after accelerated delayed fracture test was directly measured by hydrogen thermal analysis. It is found that high resistance to delayed fracture of 2.0% Si steel is due to the high hydrogen content needed for fracture, in spite of its high hydrogen occlusion rate. Also the high resistance to delayed fracture of 1.5% Si-30ppm Ca steel is found to be due to the low hydrogen occlusion rate and the low hydrogen content needed for fracture.
Static recrystallization kinetics and high temperature deformation properties of austenitic stainless steel SUS304 (AISI 304) have been investigated. It was established that the grain size of reversely transformed (α'→γ) and statically recrystallized microstructures varied directly with the volume of initial martensite phase which was induced by cold working. It was concluded that such recrystallization processes with reverse transformation could be divided into four stages. Tensile tests in this alloy sheets with initial grain sizes of about 1.0μm have been conducted at different temperatures, below the static recrystallization temperature, and at strain rates ranging from 3.0×10-5 to 1.0×10-2s-1. The elongation to failure exceeded 300pct and strain-rate sensitivity index (m) of about 0.4 were obtained at a test temperature of 998K at an initial strain rate of 3.0×10-5. The influence of high temperature deformation on grain coarsening has also been investivated in this material.