Due to mitigation of CO2 emission, low reducing agent operation of blast furnace is actively investigated. For blast furnace operation, low reducing agent indicates decreasing consumption of coke or PC with maintaining appropriate heat balance in the furnace. Accordingly, O/C of input burden increases with decreasing reducing agent rate. Coke particle acts as structural material for holding burden and making void in the packed bed for maintaining gas flow. Therefore, decreasing gas flow resistance is an issue in low reducing agent rate operation. In 1970's, dissection study of blast furnaces had been carried out, and that uncovered existence of cohesive zone. Solid and liquid phases coexist in the cohesive zone. Gas flow resistance increases with a decreasing void fraction in the cohesive zone. Where, coke slit composed with coke layer between half molten iron ore layers act as gas vent slit. However, low reducing rate operation that causes thin or no coke slit. How to increase gas permeability in the cohesive zone with thin coke slit would be a serious issue with a decreasing coke rate for blast furnace operation. In the present paper, investigations on cohesive zone and its permeability are reviewed and softening and melting behaviors of iron ore layer in the blast furnace are thermodynamically discussed.
This paper is concerned with controller design for the automatic gauge control system. In order to produce steel plates of good quality with high reliability, stochastic optimal control for variance suppression is adopted in this paper. The probability distributions of the physical parameters of the plant system may be time varying. Whereas the conventional stochastic optimal control methods can treat only time invariant probability distributions of the parameters, this paper derives an optimal control law to take care of time varying probability distributions. We apply the proposed method to gauge control in order to suppress the variation of the plate thickness. In addition, a numerical example exhibits the effectiveness of the proposed method.
Beverage cans have being produced mainly with draw and ironing (DI) process since 1971. The cans in the process are formed in wet condition, generally followed by being washed and lacquer-coated. In 1992, a dry forming process using polyester film laminated steel, called Stretch and Draw forming, was developed. Application of Polyethylene terephthalate (PET) film laminated steel to DI process was expected to be beneficial but had a problem (hereinafter referred to as PET-hair) of film shaving or breaking at formed can edge due to no flange making in process. The PET-hair generated on the condition of over 15% in ironing rate, indicating it was difficult to make can body thinner and the laminated material could not be applied for DI process. In this background, we investigated the influences of film thickness, film properties and substrate yield point on PET-hair generation, and obtained the following findings. Firstly, thinner films tend to get PET-hair reduced. Secondly, film properties impact little on PET-hair due to the temperature increase by process heat generation. Thirdly, the substrates with lower yield point tend to get PET-hair reduced.
Thermal histories in multi-bending process after hot rolling were clarified by numerical analysis. Temperature vibrates cyclically at strip surface, whereas relatively small vibration can be seen at center in thickness. Transferring speed and cooling condition characterize temperature distributions in strip thickness. In industrial production, water-cooling equipments would be necessary in order to cancel heat generation during bending deformation and maintain strip temperature in appropriate range.
The surface carbon concentration (Cs) of carburized Nb-bearing steel (SCM420Nb) was found to decrease substantially when carburized after severe machining. It was observed that the effect of machining condition is small in normal case hardening steel (SCM420), in which the Cs was in the range of 0.9~1.1%. However, the Cs in carburized SCM420Nb was reduced to 0.61% by mild machining, and 0.38% by severe machining. Since the Cs of severely machined SCM420Nb recovered to > 0.8% after the removal of surface layer by polishing, it was concluded that the abnormal decrease in Cs of SCM420Nb is mainly due to severely deformed surface layer produced by machining. Discussion was made on the abnormally low Cs observed in severely machined SCM420Nb steel with regard to the effect of solute Nb on the carbon activity and the effect of surface roughness. It was found that these effects are small compared with the observed reduction in Cs. Comparing the experimental and calculated carbon concentration profiles, Cs was expected to be reduced immediately after the start of carburizing. Finally, it was proposed that the oxide layer created on the surface during carburizing might be the cause of the abnormally low Cs.
A numerical model to quantitatively predict the cleavage fracture initiation in ferrite-cementite steel is proposed. The model is based on the microscopic fracture process of the three stages; Stage-I: formation of fracture origin by cementite cracking, Stage-II: propagation of the cementite crack into ferrite matrix and formation of a cleavage crack, and Stage-III: propagation of the cleavage crack across ferrite grain boundary. Influencing factors on Stage-I is quantified by using tensile testing with circumferential notched round bar specimens. The specimens are made from steels with various ferrite and cementite sizes. Probability of cementite cracking is formulated based on the results of SEM observation and measurements. Notched three point bend interrupted tests with fractography shows the importance of Stage-III because a number of arrested micro-cracks are observed on the fracture surface. In the proposed model, fracture conditions are formulated by the ratio of cementite cracking based on the experimental results on Stage-I and the concept of fracture stress of ferrite matrix on Stage-II and Stage-III. Active zone is divided into finite volume elements. Ferrite grains and cementite particles are assigned based on their distributions into each volume element. Applied plastic strain and stress of each volume element are calculated by macroscopic finite element analysis. Cleavage fracture is assumed to initiate at the time when the fracture conditions of the all stages are satisfied in any one of the volume elements.
The prediction model of cleavage fracture toughness proposed in the authors' previous study is validated by its application to three point bend testing of notched specimens. Experiments are carried out by using steels with various ferrite and cementite sizes. The numerical predicted results of fracture toughness show good agreement with the experimental ones under all the conditions of test temperatures of the all steels. The bottleneck process of cleavage fracture is then evaluated by the number of arrested micro-cracks until all of the fracture process is satisfied. A lager number of the arrested micro-cracks are formed at higher temperature. It means that the bottleneck process is changed from Stage-II to Stage-III with increasing temperature. Grain size at fracture initiation site is evaluated. The scatter of the grain size in the case of the steels with coarse ferrite grains is larger than that of the steels with fine grains. It is therefore shown that the developed model can reproduce the size effect of cleavage fracture. Dependence of the fracture stress on temperature was clarified. In particular, lower temperature makes larger scatter of the fracture stress both on Stage-II and Stage-III. This tendency is remarkable in the cases of the steel with fine grains. Based on the aforementioned results, the validation and the effectiveness of the proposed model are found out.
Tensile tests of 16mass%Cr bearing ferritic steels were performed in order to investigate the effect of the mean interparticle spacing (λ) on the local deformation energy. λ of Cr precipitates was changed from 3.3 to 5.1 μm. The amounts of Cr precipitates in the specimens tested were controlled to be the same level. The local deformation energy was increased with increasing λ in the specimens without a notch, while the local deformation energy does not depend on λ for the notched specimens. AsB (Angle selective Backscattered Electron Detector) images of voids beneath the fracture surface showed that the difference of the local deformation energy dependence on λ between notchless and notched specimens was due to the void coalescence processes. This finding reveals that the local deformation energy is influenced by the stress triaxiality through the void coalescence behavior.