Flow and concentration fields in bench scale gas-injected molten steel baths were numerically studied to investigate dependency of flow structure and mixing characteristics on bath aspect ratio and gas injection rate. The problem was formulated as a two-phase flow on the basis of Eulerian approach for both gas and liquid phase transports. The effects of vessel aspect ratio and gas injection rate on mixing efficiency were characterized from the predicted time evolution of a tracer distribution in the melt and also from the liquid circulation rate. Results demonstrated that both vessel aspect ratio and gas injection rate play crucial roles on flow structure in the bath and hence on the mixing efficiency. It was observed that there is an optimum vessel aspect ratio (1.5) for minimum mixing time for the systems considered. There also exists an optimum gas injection rate for maximum mixing intensity (minimum mixing time) in a given system.
Aluminium-oxygen equilibria between CaO-Al2O3 melts and liquid iron were studied in the temperature range of 1 823 to 1 923 K, using an alumina and lime crucibles. The equilibrium constant K for the reaction: 2Al+3O= Al2O3 was discussed with reference to the supersaturation for nucleation of alumina, and compared with the previous experimental and thermodynamic data. The supersaturation was observed in the experiments in which an Fe-Al alloy with low concentrations of initial oxygen was used under the condition of no stirring. Supersaturation increased with a decrease of oxygen content in liquid iron. Two types of alumina inclusions with a size of 2-5 μm and below 0.1 μm were observed by SEM.
Nitrogen and sulfur distribution ratios between CaO-Al2O3 slags and Fe-0.0001-33mass%Al alloys were measured in the temperature range of 1823 to 1923 K, using an Al2O3 and CaO crucibles. The distribution ratios under a CaO crucible were higher than those under an Al2O3 crucible, and increased with an increase of Al content. Nitride and sulfide capacities defined by CN3-=(mass%N)·PO23/4/PN21/2 and CS2-=(mass%S)·PO21/2/PS21/2, respectively, which were obtained by using the reported values for activity of Al2O3, agreed well with those directly measured in a gas-slag experiment under an Al2O3 crucible, but disagreed with those under a CaO crucible. Activities of Al2O3 in the CaO-Al2O3 melts saturated with CaO were evaluated based on the results in the gas-slag experiments.
Slags containing CaO, CaF2, SiO2 and Al2O3 have found wide application with refining processes in ironmaking and steelmaking practice and ladle metallurgy. A study of sulphur equilibrium partition between CaO-CaF2-SiO2-Al2O3 slags and carbon-saturated iron in temperature range 1 450-1 600°C was carried out in present work. The sulphur equilibrium distribution increases with replacing SiO2 by CaO and SiO2 by CaF2, while it decreases with substitution of CaO by CaF2 at 1 500°C, PCO=0.1 MPa if other variables like wt% CaF2, wt% Al2O3 or (wt%Al2O3+wt%SiO2) are kept constant. Based on experimental data the equation of temperature dependency was found to be: log LS=8.2–(9700/T) The activity coefficient of sulphur (fS) in carbon saturated iron was determined at 1 500°C as 6.9. Based on obtained equilibrium data for LS, sulphide capacity of investigated quaternary slag system was calculated.
A test on soft reduction employing conventional one-piece rolls was conducted to analyze the influence of soft reduction and reduction timing on center segregation. As a result, the following findings were obtained. Center segregation was improved when soft reduction was applied only in the final stage of solidification in which the fraction of solid at the slab center is larger than 0.25. When soft reduction was applied employing one-piece rolls over the wide range of solidification stage including the solidification stage with the fraction of solid at the slab center below 0.25, center segregation deteriorated with increasing amount of reduction and the amplitude of variations in center segregation in the casting direction was also increased. It is estimated that the deteriorating tendency of center segregation by soft reduction on the upper stream side is attributable to the new flow of molten steel caused by uneven reduction due to mechanical factors such as roll bending. For the improvement of center segregation, therefore, the optimization of reduction timing is indispensable especially when mechanical factors of uneven reduction are present.
The influence of roll bending on center segregation in continuously cast slabs was studied by conducting the test in which roll bending was controlled during casting. The behavior of roll bending and center segregation in sequential casting were also investigated. As a result, it was clarified that (1) variations in center segregation in the casting direction are attributable to roll bending and center segregation deteriorates with increasing roll bending, (2) the amount of roll bending increases up to 1.7 mm during casting, as a result of which variations in center segregation in the casting direction are increased and segregation deteriorates with increasing number of cast heats in sequential casting, and (3) the use of divided rolls in the final stae of solidification is effective in decreasing roll bending and keeping center segregation at lower level. Of all mechanical factors, roll bending was estimated to be the largest factor for deterioration of center segregation in the case of casting employing one-piece rolls of conventional type.
An advanced mathematical model for the fluid flow coupled with heat transfer confined by a free surface and a solidification front in a cold crucible was developed. Validity of the model was confirmed through measurements of a solidification front, surface velocity of the melt and surface temperatures. Effects of operation parameters, such as coil current, dome height and casting velocity, on the fields of velocity and temperature were investigated. Generally, two kinds of recirculation flow are expected to appear in the melt. They make a collision slightly above the contact point between the melt and the crucible wall. An appropriate geometrical configuration of the dome height and the position of a coil exists to promote melting of scraps fed onto dome. A titanium ingot consisting of completely melted scraps was continuously cast aided by the proposed operation obtained through the numerical calculation.
The use of the Diercks equation, which was developed for assessing the creep-fatigue lives of AISI 304 (JIS SUS 304), for predicting the creep-fatigue lifetimes of Cr-Mo steels was examined in comparison with the linear cumulative damage rule and the strain range partitioning method. Following conclusions were reached: (1) The differences in the fatigue and the creep-rupture strengths existing between the SUS 304 and the Cr-Mo steels can be accounted for by modifying two factors in the original Diercks equation: for the fatigue lifetime ratio α, the relative lifetime ratio, αr, which is a ratio of the "pure" fatigue life of SUS 304 to that of the Cr-Mo steel concerned, should be used, while for the temperature TC, the equivalent temperature, Te, which is a temperature that will give rise to the same creep-rupture lifetime for the Cr-Mo steel as for 304 for the stress concerned, should be used. (2) The Diercks equation, modified as in (1) above, predicts the creep-fatigue lives of Cr-Mo steels to a factor of 2, i.e., no more than twice if overestimated and no less than one half if underestimated. This is the same accuracy that the strain range partitioning method features. (3) The linear cumulative damage rule should not be applied to Cr-Mo steels not only because it can give mutually contradicting evaluations for different strain waveforms, but because the accuracy of prediction is unacceptably large. (4) Since the method proposed herein for formulating the modified Diercks equation is free of the complexities in experimentation and in data analysis that characterize the strain range partitioning method, it should be taken as a better means, to the first approximation at least, of assessing the creep-rupture properties of Cr-Mo steels.
Reversion mechanism from deformation induced martensite (α') to austenite (γ) has been investigated in two metastable austenitic stainless steels, 15.6%Cr-9.8%Ni (the 16Cr-10Ni) and 17.6%Cr08.8%Ni (the 18Cr-9Ni) steels, by means of magnetic analysis and transmission electron microscopy. Metastable γ almost completely transforms to lath α' by 90% cold rolling, and the α' again reverts to γ during annealing at temperatures above 700 K. Deformation induced α' in the 16Cr-10Ni steel undergoes a martensitic shear reversion during heating to 923 K annealing, while that in the 18Cr-9Ni steel does a diffusional nucleation-growth reversion on 923 K annealing. Grain refining processes are greatly influenced depending on the reversion mechanism. Martensitically reversed γ has a high density of dislocations immediately after the reversion and the γ grains are refined through recovery and recrystallization process just like that taking place in a deformed γ. On the other hand, diffusionally reversed γ is characterized by the nucleation of equiaxed γ grains within the α' matrix and the γ grains gradually grow during annealing. The reversion mechanism significantly depends on the chemical compositions of steels and annealing temperature. An increase in the Ni/Cr ratio causes an increase in the Gibbs free energy change between fcc and bcc structure, leading to a fall-down of austenitizing temperature for the martensitic shear reversion. The critical driving force required for the complete martensitic shear reversion is about -500 J/mol. To obtain the critical driving force in the 18Cr-9Ni steel, it should be heated to a high temperature above 1 023 K. However, the diffusional reversion can easily occur because the martensitic shear reversion temperature is too high in the 18Cr-9Ni steel. The 16Cr-10Ni steel also undergoes the diffusional reversion when it was annealed at low temperatures below the martensitic shear reversion, 923 K.
The solubilities of Al2O3, SiO2 and Cr2O3 in molten Fesatd.-FeS and Fesatd.-MnSsatd.-FeS systems have been measured between 1 273 and 1 773 K. The oxide solubilities were observed to decrease with raising temperature and decreasing FeS content of the sulfide melts. The solubilities of Al2O3 and SiO2 in the FeS-MnS system were also measured at 1 573, 1 673, and 1 773 K. The mechanism of the formation of oxysulfide complex inclusions is discussed based n the present findings.
Transient temperature distributions in solid steel during common heat treatment operations have been investigated. Towards these, a mathematical model has been developed and a large number of experiments were performed in the laboratory on different grades of steel samples. Four different approaches have been considered in the mathematical model (viz., procedures based on the TTT characteristics, Fe-C equilibrium diagram, variable specific heat and negligible heat effect) to embody the heat effect (e.g., generation/dissipation) of the phase transformation reaction and to assess the associated contribution to the overall time-temperature history. Numerical predictions as well as experimental measurements appear to indicate that the procedures adopted to model the heat source/sink are not critical, as far as prediction of the transient temperature profiles are concerned.