Understanding of the properties of slags is a pre-requisite in optimizing their functions towards the making and refining of steel. Important contributions towards this objective have been made over the past half a century, in different parts of the world, especially in USA, UK, Germany and most of all in Japan. Knowledge of the slag properties enables in understanding the slag structure as well. The present review paper summarizes the contributions made in this field by the Division of Materials Process Science, Royal Institute of Technology, Stockholm, Sweden. The paper deals with the measurement and modeling of (a) thermochemical properties of slags, (b) thermophysical properties and (c) inter-property correlations. Some important contributions during recent years, such as the determination of the valence states of Cr and V in slags, wetting characteristics related to hot-metal desulphurization, diffusion of sulphide ions in slags, partition of phosphorus between slag and metal phases and studies on process phenomena such as foaming are highlighted. The research work has led to the evolution of a new basicity concept. Dynamic physical property measurements are pointed out to be an experimental tool towards understanding of reaction mechanisms. Developments with respect to slag/metal interfacial phenomena, viz. the concept of surface velocity and surface viscosity and quantification of these properties are presented.
Pelletizing of Indian chromite ores is more challenging due to their high refractory nature. High Cr/Fe ratio and high MgO content in these ores demand high firing temperatures and longer firing cycles but often result in low strength fired pellets. Aim of this study was to develop cold bonded chromite pellets for smelting in submerged arc furnace (SAF) from chromite fines using suitable binder that induce less gangue into the pellets but cures quickly. Different binders were studied through laboratory pelletizing experiments for their suitability for cold bonding the pellets. As result, a composite binder comprising dextrin and bentonite, was found to be suitable and pellets made from the same were tested for their low and high temperature behavior. Electron and optical micro structural studies with image analysis were carried out to find out the type and amount of phases formed in the chromite pellets during high temperature reduction. High temperature reduction studies revealed that pellets were resistant to disintegration up to 1200°C. Pilot scale arc furnace trials were also carried out to compare the performance of cold bonded pellets (CBPs) with sintered chromite pellets and found that for a constant power input, smelting rate was faster for CBPs than sintered pellets. Loss of Cr to slag was reduced in case of cold bonded pellets usage.
Surface tensions of low carbon slabs and 16 mass%Cr stainless steel were estimated using a surface thermodynamic model proposed by Mukai et al. As an application of the model, an index to evaluate the driving force for the fine bubble entrapment by the solidifying shell, the Mukai-value, M, was calculated from the surface tension values. The relationship between Mukai-value and number of entrapped bubbles was discussed. A linear relationship was found between the number of captured bubbles and Mukai-value. In the previous work, the Mukai-value was used as a relative scale to evaluate the driving force for the movement of bubbles. However, by calculating the M from the surface tension values by the surface thermodynamic model, physically reasonable Mukai-values could be obtained.
The compositional effect of NaF and CaO/SiO2 on the hydrogen solubility in the NaF–CaO–SiO2–FeOt welding flux system at 1823 K is presented. At a CaO/SiO2 of 1.3, higher NaF decreased the hydrogen solubility and at a CaO/SiO2 of 1.5, higher NaF had relatively little effect on the hydrogen solubility in the flux. The hydrogen solubility with CaO/SiO2 showed a parabolic behavior showing a minimum near unit basicity and increasing with higher basicity. FTIR (Fourier transform infra red) and Raman spectroscopy of as-quenched flux samples showed the addition of NaF and higher CaO/SiO2 depolymerized the flux structure. At low and intermediate basicities from 0.9 to 1.3, higher NaF decreased the incorporated hydroxyl sites of Si(OH) and FeO(OH), which correlates well to lower hydrogen values.
In order to increase the permeability of the sintering bed for sinter ore productivity, RF-MEBIOS (Return Fine - Mosaic Embedding Iron Ore Sintering) process, in which return fine as dry particle is added on granulated raw materials and then they are charged into sintering machine, is proposed. In RF-MEBIOS, it is demonstrated by pot tests that productivity increases at the same moisture content in sinter mixture at charging. This productivity increase is caused by higher permeability in sinter packed bed due to two major phenomena. One is increasing the pseudo-particle size at granulation and the other is decreasing the bulk density of sinter packed bed after charging. The former is achieved by a higher moisture content in the raw materials at granulation, which has the role of decreasing small size of pseudo-particle (–0.25 mm). The latter is achieved by higher friction in the packed bed composed of dry and wet particles compound, which has a role of decreasing bulk density. In the development of RF-MEBIOS, return fine was chosen as the dry particle because it is dry when produced by the sintering machine. The sinter productivity increases with the increase of the quantity of the return fine added after granulation stage. The effect of pseudo-particle (–0.25 mm) ratio and ε on flame front speed were evaluated as 55% and 41% to increase of frame front speed, respectively. Effect of RF-MEBIOS on sinter productivity is confirmed in No.3 sinter plant in Kashima Steel Works. Under the condition of constant moisture content in sinter mixture at charging, this improvement degree is proportioning to the ratio of bypass return fine which is added to granulated the other sinter materials without granulation. It means granulation at higher moisture has superiority compared to increase of fine material in bypass return fine. Finally, RF-MEBIOS method is installed on three commercial sintering machines (Kashima, Wakayama, and Kokura) belonging Sumitomo Metals. In all three sinter plants, productivity increase has been confirmed. Therefore, introducing RF-MEBIOS has been demonstrated to cause a universal improvement of sinter productivity.
The reduction of MnO in slag by blends of coke with high density polyethylene (HDPE) was investigated by the sessile drop method at 1500°C in this study. The results show improved wettability and extents of reduction are realised with the use of an HDPE/coke blend in this system by comparison to reduction by pure coke, whereby increasing HDPE content resulted in further improvement in extent of reduction and increased wettability. The extensive devolatilisation from HDPE samples is the primary cause for these improvements, whereby the gasified HDPE created both CH4 and H2 reducing gases. Additionally, increased sample porosity allowed for improved wetting, and thus improved reduction capabilities. The dynamic contact angle between the carbon substrate and the slag varied, with HDPE samples ranging between 140°–60°, whilst the coke samples ranged between 160°–120°. The addition of HDPE allowed for the near complete reduction of MnO and partial reduction of SiO2 from the slag with distinct metallic regions of Mn–Si formed in the sample; regions containing pure Si were also found.
Numerical modeling has been used to investigate the influence of electromagnetic stirring on melting of a single piece of scrap in an eccentric bottom tapping (EBT) electric arc furnace (EAF). The heat transfer and fluid flow in the melt for both conditions with and without electromagnetic stirring were studied. The buoyancy and electromagnetic forces were considered as the source terms for momentum transfer in the studied conditions. The enthalpy-porosity technique was applied to track the phase change of a scrap piece defined in the EBT region of the furnace. Different scrap sizes, preheating temperatures, stirring directions and force magnitudes were considered, and the heat transfer coefficient was estimated from the heat transfer rate at the melt-scrap interface. The results showed that electromagnetic stirring led to a reduced melting time and an increased heat transfer coefficient by a factor of four. The results for Nusselt number versus Grashof number for natural convection and Reynolds number for electromagnetic stirring were compared with those obtained through correlations from previous studies.
The present work aims to understand the mechanism and rate of iron oxide formation in molten mould flux initially having no iron oxide to discuss how to keep the mild cooling performance of the flux. The concentration change of iron oxides in mould fluxes was measured at 1891 K as a function of time in variations of three physico-chemical conditions: (i) oxygen partial pressures in atmospheres (air and argon), (ii) SiO2 activities in mould fluxes and (iii) oxygen concentrations in molten irons, the last being provided by Al-killed and non-killed operations on electrolytic iron samples. The following findings have been obtained: (i) iron oxide concentrations increase with holding time and reach ca 3.5 mass% within 1.8 ks, independently of atmospheres, (ii) higher SiO2 activity leads to higher viscosity of mould flux and lower iron oxide concentrations and (iii) iron oxide concentrations decrease to ca 0.5 mass% due to lower oxygen concentrations in molten Al-killed iron. Thus, the following mechanism has been proposed: oxygen dissolved in molten iron is primarily oxygen source and reacts with iron to form iron oxides at the metal/flux interface, which oxides diffuse into molten flux phase. A kinetic discussion has given the total reaction rate constant as k = 7.5 × 10–6 cm·s–1 and suggested that the rate be dominated by iron oxide transfer through the boundary layer. To suppress iron oxide formation, additions of reducing agents would be more efficient than controls of oxygen partial pressures.
Today, the demands for Advanced High Strength Steels (AHSS) have gradually increased due to their ability to reduce vehicle weight as a means to save energy, reduce the environmental impact while simultaneously improving passenger safety. However, AHSS often require the addition of large amounts of alloying elements such as aluminum and this can make it difficult to cast sound slabs without surface defects. When casting high aluminum AHSS, due to the reaction between aluminum in steel and silica in mold flux, the viscosity and crystallization characteristics of the mold slag changes drastically, and deteriorates mold lubrication. Therefore, it is critical to limit the reaction between Al in steel and mold slag and at the same time to provide consistent and adequate mold slag in-use properties. This paper describes the development of non-traditional lime-alumina based mold fluxes which have the potential to reduce slag-steel interaction during casting of high aluminum TRIP steel. Several trial casts of 1.45% Al TRIP steel have been conducted on a pilot caster to examine the performance of mold fluxes developed. When the lime-alumina based mold fluxes were applied successfully, alumina pickup was reduced to less than 5% as compared to 15% alumina pickup for corollary trial casts using conventional lime-silica mold fluxes. The developed lime-alumna mold fluxes showed improved in-mold performance as indicated by enhanced lubrication and stable mold heat transfer, again compared to lime-silica fluxes. Cast slabs from the trials using these lime-alumina fluxes have periodic and sound oscillation marks and minimized defects.
Intricate physical phenomena such as boiling heat transfer, free surface flow, and the moving of a cooling object (running hot plate) are included in the ROT (Run Out Table) cooling process of hot rolling. Mixed reciprocal actions are also included. Therefore, analyzing the process with considering all relevant phenomena simultaneously has been an important objective in this field for a long time. In the ROT process, cooling performance could be affected by the upstream process, and by various environmental factors such as refining, reheating, and the ambient temperature and humidity. In this study, with such disturbances numerically fixed, cooling capacities are compared by varying the arrangement of the cooling water supply nozzle. Generally, the cooling heat flux is known to vary in the plate running direction. Thus, clustering the nozzle in the upstream or downstream direction could affect the average cooling capacity of the facilities. The cooling histories, the heat fluxes on the plate surface, and the thickness of the Leidenfrost steam layer are compared for various nozzle arrangements.
A novel gray-box model is proposed to estimate molten steel temperature in a continuous casting process at a steel making plant by combining a first-principle model and a statistical model. The first-principle model was developed on the basis of computational fluid dynamics (CFD) simulations to simplify the model and to improve estimation accuracy. Since the derived first-principle model was not able to estimate the molten steel temperature in the tundish with sufficient accuracy, statistical models were developed to estimate the estimation errors of the first-principle model through partial least squares (PLS) and random forest (RF). As a result of comparing the three models, i.e., the first-principle model, the PLS-based gray-box model, and the RF-based gray-box model, the RF-based gray-box model achieved the best estimation performance. Thus, the molten steel temperature in the tundish can be estimated with accuracy by adding estimates of the first-principle model and those of the statistical RF model. The proposed gray-box model was applied to the real process data and the results demonstrated its advantage over other models.
The present paper describes a facile method for the determination of trace impurities in iron metals. This method is based on the selective removal of the matrix element as hydroxide precipitate followed by the determination of desired trace elements left in the solution. To an iron sample solution were added ethylene glycol bis(2-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA) and diethylenetriamine. The iron(III) matrix was almost completely (>99%) precipitated at pH 9.5, while trace elements [Mn(II), Co(II), Ni(II), Cu(II), Zn(II), Cd(II), and Pb(II)] were left in the solution due to the complexation with EGTA and diethylenetriamine. The sufficient removal of the matrix element and negligible loss of trace elements allowed the sensitive determination at ng g–1 levels by inductively coupled plasma-mass spectrometry.
Numerical simulations of the hydraulic bulge test are carried out by the implicit static finite-element method. The sheet specimen is characterized as a rate-independent elastoplastic material with a power-law hardening rule. The stress and strain relationship of the specimen is evaluated from the internal pressure and nodal coordinates obtained in the finite-element simulation of the hydraulic bulge test. Varying the gauge lengths of the spherometer and extensometer and the ratio of the initial thickness to the diameter of the specimen, their influences on the estimated stress and strain are investigated. By comparing the estimated stress-strain relationship with that of exact input data, the stress and strain measurement accuracy is assessed. Furthermore, the stress state at the apex is examined for orthotropic specimens and is found to deviate by 1–5% from the equi-biaxial stress state.
The influences of different normalizing heat treatments on microstructure, non-metallic inclusions and impact toughness in MAG Ti-bearing weld metal of HSLA steel has been studied. It has been shown that for the Ti-bearing weld metal the impact toughness after normalizing treatment decreases significantly against prolonging holding time and increasing normalizing temperature. The Mn-depleted zone forms around the Ti-bearing phase (MnTiO3) precipitated on Mn–Si oxide. Proeutectoid ferrite preferentially nucleates at the Mn-depleted zone and the interface between austenite and proeutectoid ferrite becomes the nucleation sites for pearlite thereafter. Mn-depleted zone formation increases the pearlite nucleation sites, and makes the pearlite fine. The dissolve of Ti-bearing precipitate causes disappear of Mn-depleted zone at strong normalizing processes (longer normalizing time and higher normalizing temperature), and the number of ferrite nucleation sites decreases, then the pearlite become coarser, which causes the deterioration of impact toughness.
This paper aims at a systematic comparison of effect of single, double and triple pass welding on heat affected zone and tensile strength of AISI 304 stainless steel and chrome-manganese austenitic stainless steel. Degree of sensitization (DOS) increased with increase in number of passes and highest DOS (35.53%) was obtained for triple pass welding of chrome-manganese austenitic stainless steel. The decrease in tensile strength is relatively more in chrome-manganese austenitic stainless steel as compared to AISI 304 SS. The mode of failure for AISI 304 SS was ductile fracture, whereas chrome-manganese austenitic stainless steel failed due to intergranular brittle fracture.
To improve vehicle fuel economy and crash worthiness the automotive industry has been redesigning parts from advanced high strength steels such as dual-phase and martensitic steels. These steels have high strengths with the higher formability characteristics when compared to lower strength conventional steels of similar ductility. These steels derive their unique properties from their complex microstructures containing ferrite and martensite. During assembly welding, the martensite within the sub-critical region of the heat-affected zone tempers, which locally reduces mechanical properties. Although this phenomenon is well studied, it has yet to be quantified. The present work proposes a technique to measure the softening kinetics of dual-phase and martensite steels using rapid isothermal tempering. The resulting model was then validated by predicting the heat-affect zone softening that occurs in laser and resistance spot welds as well as by comparing the microstructures of the rapid tempered samples to the microstructures found in the sub-critical heat-affected zone of welded samples.
Curtain coating, which can achieve very high speed coating and smooth film without defects, is recently used to produce pre-painted steel sheets for home appliances and utensils. One of the difficulties limiting curtain coating operation arises from a low limit in flow rate, below which the falling liquid sheet breaks apart and we can not operate. We measured the minimum flow rate for curtain formation for polyester resin solutions in very low Reynolds number (Re) experimentally, and found that the minimum flow rate decreased with decreasing surface tension and density, and with increasing viscosity of the liquids. Relationship between the physical properties of the liquids and the minimum flow rate for curtain formation (curtain stability) was analyzed by using Reynolds number (Re) which represented the flow rate and Physical property number (Ka) which included surface tension, viscosity and density. As a result, the calculating formula by which the minimum flow rate can be estimated was given. Further more, effects of surfactant on the minimum flow rate were studied. It is suggested that homogeneity of the surface is very important to stabilize the curtain, and if we use surfactants which decrease surface tension of the liquids uniformly, the curtain stability increases. In the range of this study, curtain stability is mainly dominated by liquid properties, and neither edge guides nor disturbances like bubbles in the liquids seem key factors for curtain stability.
In order to understand how exact Goss-oriented grains form in electrical steel, we studied Goss orientation and microstructure during primary recrystallization using electron back-scattered diffraction. In particular, we examined the effect of heating rate (20°C/s and 150°C/s) on the primary recrystallization behavior. In a cold rolled sheet, the Goss fraction in the outer layer is more than twice that in the inside layer, particularly because the fraction of Goss±5° grains is much greater at the surface compared with the inner part. When primary recrystallization is completed, the fraction of grains that have Goss±5° and ±10° orientations in the outer and inner layer appear to be similar regardless of the heating rate, and the fraction of Goss±15° grains appears to be higher when the sheet is subjected to a rapid heating rate than when it is subjected to a typical heating rate. In addition, the number of Goss±5° grains corresponding to a rapid heating rate is twice that corresponding to a typical heating rate, with smaller grain sizes and a more uniform distribution.
A computer program was developed to simulate competitive hydrogen trapping and carbon segregation to the trap site which are often referred to site competition, and the spectra of medium carbon martensitic steels, which were analyzed previously assuming a single trap energy, were re-examined. The McNabb-Foster equations in which carbon segregation was treated as trapping to the defect site were solved simultaneously incorporating a phenomenological interaction coefficient between hydrogen and carbon within the trap site. Assuming that the primary trap site of hydrogen was dislocation, experimental TDA peaks, 50–100°C lower than those of heavily deformed pure iron, were reproduced well both in height and width with a narrow range of the interaction coefficient no matter hydrogen was charged at room temperature or high temperature, i.e. prior to martensitic transformation. Due to relatively faster segregation kinetics of carbon the peak temperature does not appear to be sensitive to the carbon content or the carbon occupancy prior to thermal desorption analysis in medium carbon steels.
In order to understand the mechanism of isothermal transformation of Fe–N alloy, the isothermal transformation microstructure that forms in a wide temperature range below Ae1 was investigated in Fe–2.6 mass%N hypereutectoid alloy by means of the electron back scatter diffraction method in addition to the conventional microstructural observation methods. High-nitrogen austenite fully decomposed to ferrite and Fe4N over the entire temperature range, and the time-temperature-transformation (TTT) diagram had a C shape with a nose temperature around 700 K. The hardness linearly increased with decreasing transformation temperature because the microstructure became finer, but the morphology of the (ferrite + Fe4N) structure changed discontinuously at around 800 K. From the microstructural and crystallographic analyses, it was concluded that the microstructure formed at higher temperature is a lamellar eutectoid structure, braunite, while the other is an upper bainitic structure containing bainitic ferrite formed through a displacive mechanism and Fe4N formed by concentration and ordering of the nitrogen. Since Fe4N is a counterpart of the cementite in Fe–C alloy, the respective structures are similar to pearlite and upper bainite in carbon steel.
The 63 flow curves published previously in ISIJ Int. (2012) are re-analyzed for the presence of departures from simple work hardening (dynamic recovery) behavior. These curves were determined on a 0.019%C plain C, a 0.11%C Nb microalloyed, and a 1.56%Mn-1.56%Si Nb-modified TRIP steel by means of hot compression tests carried out from 900 to 1150°C and at strain rates up to 1 s–1. Two sets of second derivative minima are shown to be associated with all 63 curves, in contrast to the single minima reported earlier. It is shown that double minima can only be obtained when the polynomial order of the fitting procedure on the whole flow curve is 11 or higher. By contrast, when only the ascending portion of the flow curve is fitted, double minima are found as long as the polynomial order is at least 8. The first set of minima corresponds to the initiation of dynamic transformation (DT), as shown in an earlier analysis of torsion flow curves. The second set of minima is associated with the nucleation of dynamic recrystallization (DRX); these were the only minima reported earlier. The temperature dependences of the two sets of minima intersect at higher temperatures, indicating that the only softening mechanism operating at high austenite deformation temperatures (in addition to dynamic recovery) is DRX.
It is known that high r-value hot-rolled steel sheets can be produced by hot rolling in the ferrite region with lubrication, whereas dynamic recovery easily occurs during ferritic rolling, in which texture formation is quite different from that in cold rolling, when the quantity of recovery increases. In this study, one pass hot rolling in the ferrite region was conducted at higher temperatures, using various rolling temperatures and rolling reductions, with two types of ULC steels, 0.016% Nb and 0.023% Ti, and their recrystallization behaviors immediately after hot rolling were investigated. The Ti-bearing steel easily recrystallized; recrystallization occurred even at 1323 K with a low rolling reduction of 30%. The γ-fiber strength reached its maximum at around 50% rolling reduction at 1273 K with the Nb-bearing steel and 1323 K with the Ti-bearing steel. On the other hand, in high temperature rolling of the Ti-bearing steel, the γ-fiber did not develop, independent of rolling reduction. These changes corresponded to the recrystallized fraction, in that the strength of the γ-fiber decreased when recrystallization occurred immediately after rolling. The deformation microstructures were different in each grain and even in each part of the same grain. New recrystallized-like grains were produced in the domain where distortions were particularly concentrated. Recrystallization seemed to be the result of various mechanisms, as some recrystallized grains were formed by a bulging mechanism, whilst others were surrounded by high angle grain boundaries.
EN H320 LA steel sheets were boronized with four different boron sources (Ekabor 1, Ekabor 2, Ekabor 3 and Ekabor HM) by the powder pack method. The boronizing process was carried out at 973, 1073 and 1173 K for 3 h. The borided samples were investigated by means of X-ray diffraction, optical microscopy, microhardness measurements and tensile tests. XRD studies showed that the boride layer consisted of phases FeB and Fe2B. The microhardness of the boride layers ranged between 1450 and 1651 HV, depending on the boronizing source. The thickness of the boride layer increased with increasing particle size of the boronizing powders and with increasing boronizing temperature. Tensile properties were also affected by the boronizing process and the particle size of the boronizing powders. Microhardness, tensile strength and elongation values were improved with the boronizing treatment. Extensive cracking was observed at the interface between the Fe2B and FeB layers.
The X-ray diffraction technique for measuring the microscopic-scale distribution of internal stress in grains of polycrystalline metal was developed in order to investigate the mechanism of the intergranular stress corrosion cracking in non-sensitized stainless steel. This technique utilized the energy dispersive X-ray diffraction with white X-ray micro beam. It enabled to evaluate elastic strain occurring in each grain distorted by cold rolling. The materials were cold-rolled in one dimension to produce 20% reduction in thickness at room temperature. The internal stress of individual grains in 20% cold-rolled SUS316 stainless steel was evaluated applying external tensile stress up to 380 MPa over elastic limit. The result indicated the inhomogeneous microscopic-scale distribution of internal stress in grains. The residual stress was distributed dispersedly to each grain in the range from compressive stress of about –500 MPa to tensile stress of about 400 MPa. The distribution of applied tensile stress to each grain was inhomogeneous.
The substitutional effect of SiO2 with Al2O3 content on the viscosity in the calcium-silicate melt-based system containing 12 mass% Na2O and 5 mass% Li2O was studied. Additions of Al2O3 increased the viscosity and higher Al2O3 concentrations modified the dominant silicate network into complex alumino-silicates with the aluminate structure becoming particularly prevalent at an Al2O3 content of 20 mass% and higher. FTIR (Fourier transform infra red) and Raman analysis showed increased amounts of symmetric Al–O0 and Si–O–Al stretching vibrations with higher Al2O3. Compared to the effect of Al2O3, structural changes and the viscosity with higher CaO/(SiO2+Al2O3) were not as significant. However, the symmetric Al–O0 stretching and the Si–O–Al in the [AlO4]5–-tetrahedral units seems to decrease with increased apparent basicity. The apparent activation energy for viscous flow increased with higher Al2O3 due to the increased polymerization of the melt and varied from 129 to 156 kJ/mol.
The influence of Al2O3/SiO2 ratio on viscosity of CaO–Al2O3–SiO2 melt was investigated by the rotating cylinder method in this study. Three groups of compositions with different CaO contents of 0.35, 0.4 and 0.45 were studied. From the experimental results, it was concluded that viscosity always first increases and then decreases as gradually increasing Al2O3/SiO2 ratio. The appearance of viscosity maximum may be resulted from the opposite variation tendencies of mean bond strength and the degree of polymerization of melt with Al2O3/SiO2 ratio. When substituting SiO2 by Al2O3, the mean bond strength decreases because of the weak Al–O bond relative to Si–O bond, which decreases viscosity; while the degree of polymerization is enhanced by the charge compensation effect of Al3+ ion, which increases viscosity. Our new proposed viscosity model can well describe the viscosity variation of CaO–Al2O3–SiO2 melt.