An experimental device which enables to measure the electromagnetic force working toward the metallic ball in the cold crucible was invented to obtain the basic data on the installation and the operation for the optimum levitation melting. The levitation force at the arbitrary position of metallic ball could be measured directly under supplying the electric power. By this device, levitation forces were measured using balls made of different kind of metals under various shapes of cold crucibles. The results obtained are as follows. (1) Levitation forces depended upon the electric resistivity of metallic ball. Low resistivity metals such as copper could be levitated easily but ferromagnetic substance such as iron could not be levitated below Curie temperature. (2) Levitation forces varied considerably with metallic ball position and coil position, and these behavior were influenced by the number of slits that regulated both magnetic flux and eddy current.
The reduction rate of dense wustite plates to iron with CO-CO2-Ar gas mixtures has been measured at 1203, 1273 and 1323K by using a thermobalance. The effects of partial pressures of CO and CO2/CO ratios in the reducing gas on the reduction rate and the morphology of the iron formed on the wustite were investigated. The reduction proceeded in two consecutive stages. In the initial stage, no formation of metallic iron was observed and the reduction rate was slow. The iron ion concentration at the wustite surface, estimated from the lattice parameter, increased with increasing reduction time and reached eventually to the concentration equilibrated with iron. In the next stage, metallic iron was formed and the reduction rate became markedly higher than that of the initial stage. The reduction rate was affected by the morphology of formed iron and the deposited carbon. On the conditions that porous iron layer was formed uniformly and carbon deposition was negligible, the reduction rate was controlled by the chemical reaction process at the interface. The reduction rate r (kg-oxygen/m2·s) was expressed by the following equation, r=kc·(1-α/Ke)·Pco where Ke and α are the equilibrium constant for the chemical reaction and Pco2/Pco ratio in the gas phase, respectively. The temperature dependence of the apparent chemical reaction rate constant kc (kg-oxygen/m2·s·Pa) was expressed by the following equation kc=3.84×10-6·exp(-9.42×104/RT).
Theoretical and experimental studies were carried out to quantitatively understand the effect of reaction sites on the decarburization rate of ultra low carbon steel. The decarburization rate was studied in a 30kg vacuum induction furnace by varying the areas of three reaction sites such as bath surface, steel melt-crucible interface and inner part of molten steel. With the oxygen content of 400ppm, the ratio of the decarburization rate at the bath surface to the overall one is largest for [C] ≤27ppm, whereas the largest is that of the inner part of molten steel for [C] ≥27ppm. The fluiddynamic calculation of the flow pattern of CO bubbles nucleated inside the molten steel of RH vacuum vessel made clear that a part of CO bubbles near the down-leg was sucked into the ladle and dissolved into the steel melt without decarburization. The effect of each reaction site on decarburization rate in RH was calculated to be the same tendency as that in the 30kg induction furnace.
In order to improve surface quality of continuously cast steel, the investigation applying alternating magnetic field on initial solidification shell was carried out. The capacity of electric generator is 200kW, and frequency is 20kHz. Mold size is 150mm and 180mm square, and the upper part of mold is divided into 28 segments and surrounded with coil. 0.12% carbon steel which has high crack sensitivity, was cast to verify the effect of alternating magnetic field on the billet surface quality. Casting speed ranges from 0.4 to 1.3m/min. At the casting, a mold flux was used. Surface roughness caused by mold oscillation and meniscus fluctuation can be remarkably decreased by applying optimum magnetic flux density distribution. In case of no magnetic field, average surface roughness caused mainly by mold oscillation is about 600 to 700μm and the deepest one reaches to 1300μm. However, when the electric power of 150kW is applied, the average roughness can be improved less than 300μm. When too much power is applied, the surface roughness worsens. It can be also cleared that there must be an optimum condition of the distance between coil and meniscus. The relation between the coil and meniscus position is an important factor to realize a soft contact situation between the solidified shell and mold. Shell formation in the electromagnetic mold was also examined.
Casting tests of stainless steel type 304 was carried out by a twin-roll type strip caster, which is one of possible technology to produce thin strips from molten metal directly. The main problem to be solved was surface defects of strips, such as longitudinal cracks, transverse cracks and small depressions. And these defects were overcome in this study. Uniform growth of solidified shell is most important to reduce longitudinal cracks and small depressions. By water model experiments, the best pouring conditions were obtained. They present uniform feeding rate along the width ofstrips and small meniscus level fluctuations by optimizing pouring position, ratio of the molten steel weight on the twin rolls to the pouring rate and fall height of molten steel from a tundish to twin rolls. On the other hand, small reduction of the cast strips is important to reduce transverse cracks. First, in order to control this, heat transfer coefficient between rolls and solidified shell was estimated by measuring temperature of strips, rolls and cooling water of rolls. And the best casting condition, where solidification ends just at the closest point of twin rolls, was calculated by using this heat transfer coefficient.
The effects of heating, rolling and normalizing conditions on creep rupture strength (CRS) have been examined in a 9Cr-1Mo-VNbN steel. When the steel was rolled after heating at 1, 250°C, normalized at temperature as high as 1, 100°C, and then tempered, the CRS was improved without the coarsening the prior austenite (γ) grain size. It is supposedly the reasons for the prevention of γ grain growth that the heating at 1, 250°C dissolves Nb (C, N) into solution completely, and the dissolved Nb, then, precipitates as Nb (C, N) during normalizing and acts fully as the inhibitor against grain growth. The CRS was hardly influenced by the prior γ grain size, possiblly because the sub-grain size is an effective unit for the CRS, and the sub-grain size was not varied by the coarsening of prior γ grain size. Improvement of the CRS is attained through optimizing the precipitation of complex carbo-nitride of V and Nb by the control of rolling and heat treating processes. High temperature heating makes the inter-precipitate distance short, and changes the CRS along with the yield strength. Higher normalizing temperature makes the precipitate larger. In this case, the CRS is higher for larger precipitate, and the yield strength is unchanged.
Effect of B addition on hot workability of Fe-36%Ni alloy has been studied by means of hot tensile tests, micro hardness tests and physical analyses of fractured surface. Addition of 0.0020.004%B was effective on improving hot workability as well as preventing slightly grain boundary oxidation. It was clarified by Auger Electron spectrometer that hot workability was greatly affected by S segregation at grain boundaries, which could be decreased by segregation of B. From the view of diffusion of B and S, grain boundary cohesion by B segregation and grain boundary oxidation, the mechanism of improvement in hot workability of Fe-36%Ni alloys by B addition has been discussed.