The oxidation behavior of various reactive elements, such as Al/Ti, B/Si, and rare earth metals (REM), by electroslag remelting (ESR) type slag has been investigated utilizing the systematic thermodynamic analysis based on the calculated activity in the fluoride-containing slags by the ion and molecule coexistence theory (IMCT). The results indicate that the IMCT model can be reliably applied to calculate the activity of each component in the ESR type slag. The oxidation behavior of the various reactive elements is completely different by changing the same component in the slag, such as CaO and Al2O3, during the ESR process. Therefore, it is indispensable to find a key parameter to control the homogeneity of the reactive elements in remelted ingots by comparing the effects of the activity of each component and temperature on the equilibrium content of the various reactive elements in alloy melts. The results demonstrate that the oxidation loss of Al/Ti, B/Si, and REM (Ce and La) can be effectively prevented by employing TiO2, B2O3/SiO2, and CaO during the ESR process. Oxidation behavior of B/Si and Ce(La)/Al is weakly susceptible to temperature fluctuation compared with that of Al/Ti in alloy melts, which can be controlled by adding TiO2.
The evolution of non-metallic inclusions in liquid steel involves a series of processes, including nucleation, growth by diffusion and Ostwald ripening, growth by collisions, floating up and removal of relatively large inclusions, as well as pushing and engulfment of remaining inclusions during solidification. All the evolution processes occur uniquely at the interface between inclusions and liquid steel. Therefore, interfacial properties between inclusions and liquid steel, such as interfacial energy and contact angle, play a crucial role in the evolution of inclusions, thus determining the inclusion characteristics. To effectively control the inclusion characteristics, the role of interfacial properties in the evolution processes of non-metallic inclusions is systematically reviewed in this work, based on theoretical analysis and published experimental results. In the early and middle stages of deoxidation, inclusions should have as high interfacial energy or contact angle as possible to enhance inclusion removal. In the later stage, however, the interfacial energy should be decreased as much as possible to weaken the clustering and pushing of inclusions, favoring the formation of small-sized and uniformly-distributed inclusions. To optimize the characteristics of Al2O3 inclusions, which are the most common in steel, several control strategies are proposed.
To reduce the phosphorus content in steel, the CaO–SiO2–FexO–MgO-based dephosphorization slag was designed to clarify the microscopic reaction behavior of phosphorus. In the present study, Raman spectra of final slag were measured and deconvoluted. The spectral analysis showed that P5+ ions removed from liquid iron existed in the form of Q0(P) and Q1(P) groups. According to the result, phosphate capacity and phosphorus distribution ratio (LP*) were redefined. The results showed that lgLP* was proportional to molar ratio of Q0(P)/Q1(P), indicating that Q0(P) exerted a significant influence on LP* compared with Q1(P). As the increases of CaO/SiO2 from 0.15 to 1.25 and FeO from 7.2 to 21.9 mass% in the final slag, lgLP* gradually increased, whereas lgLP* showed a downward trend when FeO increased to 29.2 mass%. This was because the increase of O2− ions generated by the dissociations of CaO, MgO and FeO continuously destroyed the network structure and formed more Q0(P) units, which were compensated by the increasing Ca2+, Mg2+, and Fe2+ cations to form stable groups. Meanwhile, the Q0(Si) units formed by slag depolymerization further played a role in fixing Q0(P). However, due to the stronger polarization of Fe2+ than Ca2+ and Mg2+, Q0(P) and Q1(P) units were easily deformed and decomposed by Fe2+. The excessive Fe2+ diluted the proportion of Ca2+ and Mg2+, and made Q0(P) and Q1(P) lose stability. The P5+ ions in the Q0(P) and Q1(P) units were reduced to liquid iron, and the rephosphorization phenomenon occurred, resulting in a decrease of lgLP*.
According to the spreading dynamics of a single melt droplet on the heated substrate, the mathematical model of the melt droplet contacting hot dry wall is established, and the transport process of the melt droplet is numerically simulated by the VOF method. The effects of the wall temperature profile on the spreading behavior and the heat transfer characteristics of the melt droplet are elaborated. With the increase of the wall temperature difference, the Marangoni convection is strengthened, the spreading diameter and the spreading height of the melt droplet become large and the spreading rate slows down periodically. The wall heat flux also increases and decays periodically. The droplet temperature increases periodically in the early stage of spreading, and increases linearly in the later stage. The melt droplet moves right under the Marangoni force, which results in an obvious difference in the velocity field and the phase interface between the right half and the left half of the melt droplet. The melt droplet temperature distribution shows layered and radial for uniform and linear wall temperature, respectively. This would help to reveal the spreading behavior and the heat transfer mechanism between the melt droplet and the horizontal hot dry wall, make up for the lack of experiment, and provide theoretical guidance for enhancing iron and steel smelting.
The reaction mechanism of the carbothermic reduction of Fe3O4 powder heated by radiating multi-mode microwave with 2.45 GHz and 2.8 kW was clarified by means of gas analysis and changing the contact state of mixed powder. The reaction process was classified in three stages. In the stage I, the powder was heated by the magnetic loss of Fe3O4 and the induction current loss of graphite (Gr) and reached peak temperature of 1000°C to 1200°C in 30 s to 40 s after irradiation, and then dropped rapidly at 600°C to 700°C. Fe3O4 was reduced to FeO. In the stage II, the powder was heated by the induction current loss of FeO and Gr, and the entropy production was maximum after about 100 s. In the stage III, molten pig iron was produced with light emission. The reaction rate was always larger for the Boudouard reaction than the reduction of Fe3O4 or FeO. This indicates that the Boudouard reaction is preferentially promoted by microwave near the contact points between Gr particles. The dominant reactions were determined to be Fe3O4+CO→3FeO+CO2, FeO+CO→Fe+CO2 and the Boudouard reaction from the smaller negative value of the Gibbs energy change of reactions. In the stage I and III, the pre-exponential factor of reaction rate increased with time. In the stage II, the apparent activation energy of reduction rate of iron oxide heated by microwave was lower than conventional heating.
The synthesis and characterisation of SFCA-I compounds has been performed. SFCA-I is one of the key bonding phases in iron ore sinter, influencing physical properties such as strength, reducibility, and strength degradation during reduction. The results reported herein extend the understanding of the solid solution limits of the SFCA-I phase, with the synthesis of single-phase SFCA-I materials containing 2.71, 3.23 and 3.73 mass% Al2O3. These materials contain less Al2O3 than any previously reported SFCA-I compounds. The results also confirm the low tolerance of the SFCA-I structure for SiO2, and a low tolerance for the variation of the CaO concentration. Finally, the results confirm that the SFCA-I structure accommodates more FeO than the related sinter phase SFCA.
The phase evolution during heating of two Al2O3-free mixtures designed to form the SFC iron ore sinter bonding phase was investigated using in-situ X-ray diffraction over the temperature range 293–1623 K. Results showed that during heating at an oxygen partial pressure of 5 × 10−3 atm, SFC formation was preceded by the formation of γ-Ca4Fe14O25 and γ-CFF (nominal composition Ca3.0Fe14.82O25) intermediate phases. On further heating the SFC melted to form Fe3O4 in a Fe2O3–FeO–CaO–SiO2 melt. During heating of one of the mixtures in air, it was not possible to distinguish between SFCA-I and α-CFF (nominal composition Ca3.43Fe14.39O25) as intermediate phases in the formation of SFC. The improved signal-to-noise ratio and angular peak resolution afforded by synchrotron-based XRD experimentation would be required to possibly distinguish between these two phases. Regardless of whether SFCA-I or α-CFF formed as intermediate phases during heating in air, this work shows that the oxygen partial pressure has a significant effect on phase formation mechanisms in the Al2O3-free SFC system.
This paper proposes a method of preparing high-strength formed coke from woody biomass without binder. Chipped and pre-dried Japanese cedar was heat-treated in an inert atmosphere (i.e., torrefied) at 225–325°C (Tt), pulverized to sizes in three different ranges, molded into briquettes (in the form of thick disk with diameter/thickness ≈ 2.5) at temperature up to 200°C by applying mechanical pressure of 128 MPa. The torrefied/briquetted cedar (TBC) was then converted into coke by heating to 1000°C in an inert atmosphere at normal pressure. This process sequence enabled to prepare coke having indirect tensile strength (St) of 8–32 MPa, which was much higher than that without torrefaction, below 5 MPa. The torrefaction greatly improved pulverizability of the cedar, which was further promoted by increasing Tt. St of TBC and that of coke both increased as the particle sizes of TBC decreased, but this explained only a minor part of significant effect of Tt on St of the coke. St was maximized at Tt = 275°C regardless of the degree of pulverization. The Tt effects on physicochemical properties of TBC and coke were investigated in detail. The difference in St of coke by Tt was mainly due to that in the increment of St along the carbonization at 500–1000°C. Fracture surfaces of the coke had particular morphologies that had been inherited from the original honeycomb structure of the cedar.
An intelligent case-based hybrid converter model has been established to predict the converter endpoint and process operations. The case-based reasoning (CBR) technique combined with the similarity calculation system was used in the intelligent converter model. The actual production data during the converter process in a time span from 2018 to 2020 was collected in the model database. The influencing factors during converter stage were divided into five categories, including steel composition, steel temperature, total processing time, steel cleanliness and total steelmaking cost of converter process in the model. All these factors were used in the similarity degree calculation, and then the historical cases in the model database were retrieved to guide the actual operation by analyzing its similarity degree to new case. The prediction accuracy was improved through self-learning with constant supplement of massive data. The prediction results based on 120-ton industrial converter showed a reasonable operation selection, the mean deviation of endpoint [C] content and [P] was 9.66% and 6.24%, respectively. Moreover, the difference between the predicted endpoint temperature and the measured endpoint temperature is within 10°C. Accordingly, the intelligent converter model was stable and suitable for large-scale prediction in converter endpoint and process operations.
With the improvement of the requirements for lightweight automobiles, the research and development of a higher Al content of TRIP steel have received extensive attention. However, Al in molten steel is easy to react with SiO2 in mold slag, which is one of the restrictive factors of high Al-TRIP steel continuous casting quality control. Applying the double-film mass transfer theory, combined with the slag-steel reaction equilibrium test, a dynamic equilibrium model is established, and the crystalline phases composition before and after the reaction is analyzed. The results show that the melting process of mold flux samples A and B is uniform and meets the continuous casting control temperature. The main components of the two mold flux samples changed rapidly within 10 minutes of the initial reaction. When the reaction progressed to 20 minutes, the composition of the two mold flux samples had no obvious change trend, and the slag-steel reaction reached equilibrium. The main precipitated crystalline phases of the mold flux sample are LiAlO2, Ca3(BO3)2, NaAlO2, ZrO2, and CaZrO3 crystals. Mold slag A identified the mass transfer of Al in the steel as the limiting part of the reaction, while mold slag B identified the mass transfer of SiO2 in the slag as the limiting part of the reaction.
Carbon materials play an important role in controlling the sintering and melting behaviors of mold flux. However, the reactions between carbon and mold fluxes and corresponding effects on the properties of mold fluxes are ignored in previous studies. The effects of carbon black on the sintering and melting behavior of mold flux during the heating process were systematically investigated by TG-FTIR, DSC, XRD, TEM, SEM, XPS, Raman spectroscopy, and viscosity measurement. The results showed that: carbon black between the mineral particles was visualized by TEM, and Ca4Si2O7F2 as sintering phase was not formed until 1300°C in CaO–SiO2–CaF2–Na2O slag on account of the insulation of carbon black. Residual carbon black absorbed on the surface of slag droplets suppressed the melting by preventing the aggregation of droplets. Besides, SiC as production of the carbon-slag reaction significantly increased the slag viscosity by improving the polymerization degree of molten slag.
Solidification simulations are effective in designing a casting process to improve the quality of steel slabs and ingots. In this study, a new method was developed to efficiently estimate multiple heat transfer coefficients to improve the accuracy of three-dimensional (3-D) solidification simulation for a casting process. The heat transfer coefficients for the two heat transfer directions—side and bottom—of a prismatic mold were independently estimated by data assimilation using one-dimensional solidification simulations near the boundaries. The optimum values of the heat transfer coefficients at the side and bottom boundaries were elucidated by comparing the 3-D solidification simulation and experimental cooling curves. The maximum and average errors between the cooling curves of the 3-D solidification simulation with the optimum values and those of the experiment were less than 1.8% and 0.2%, respectively.
The cellular automata (CA) method has been widely used in microstructure simulation during solidification, but it is time-consuming when used for multiple orientations and large-scale computation. In this work, a novel high-efficiency virtual submesh cellular automata (VISCA) method is proposed to reduce the mesh-induced artificial anisotropy. This method conceptually divides the mesh into a series of submeshes by giving a unique index to each CA cell. The proposed VISCA method results in a good accuracy in all orientations using uniform mesh, while an aggressive time stepping can be used. Compared with the previous CA models, the VISCA method exhibits a significant reduction in computational cost and improvement in computational efficiency. The nucleation is simulated based on a continuous nucleation model, and the Kurz-Giovanola-Trivedi (KGT) model is used to simulate the growth of the grains nucleated in the bulk material. The proposed VISCA method is comprehensively validated using an analytical solution, a casting simulation, and a directional solidification experiment. A quantitative comparison with the experiment in the directional solidification example shows that the difference in position of Columnar-to-Equiaxed Transition (CET) between the simulation and the experiment is within 1 mm while the length of directional solidification is 60 mm.
The end-point control technology of basic oxygen furnace (BOF) steelmaking is a very important production process in the later stage of smelting, and it has become a research hotspot in the smelting industry. In this paper, the carbon content and temperature prediction models were established based on a twin support vector regression (TSVR). In order to ensure the stability of the model, the input variables of the model were determined by using grey relational analysis (GRA) and partial correlation analysis (PCA). The results of the prediction models demonstrate that the hit rate of carbon content error bound within 0.005 wt% is 90%, and the hit rate of temperature error bound within 10°C is 94%, respectively. Moreover, a double hit rate reached 84%. By comparing with the back propagation (BP) neural network model, the hit rates of the proposed prediction model are higher than those of the BP model. Based on prediction models, a whale optimization control model was established for the calculations of auxiliary materials and oxygen blowing volume. The simulation results illustrate that the proposed control model has the mean absolute error of 9.3114 tons for scrap weight, mean absolute error of 2.5791 tons for lime weight, 0.7919 tons for dolomite weight and 537.8215 Nm3 for oxygen blowing volume, which are better than the other two control models in the aspects of the calculation accuracy.
Crack detection for iron ore green pellet is an essential step in the measuring process of drop strength, which is one of the important quality metrics of green pellet. However, current method for crack detection of green pellet is manual inspection, which is rather laborious, tedious and subjective. Although various deep network-based methods are proposed to automatically detect cracks in tunnel, pavement and wall, little effort has been made on pellet crack detection. Therefore, it is still unknown whether the current deep network-based methods can solve the crack pellet detection problem. In the present work, we perform comparison study to evaluate the performance of six state-of-the-art deep networks, using our green pellet dataset with various crack types and complex background. Comprehensive comparatives are conducted to evaluate the performance and computing efficiency of six deep networks on pellet crack detection. Moreover, task-driving comparison is performed to show what to extent the six deep networks affect the measuring accuracy of drop strength. Our experimental analyses demonstrate that CrackSegNet achieves better crack detection accuracy than other five networks (DeepCrack-Z, DeepCrack-L, U-net, CrackSegNet, GCUnet), and thereby performs better in the task of drop strength measurement. However, computing time needed by CrackSegNet (0.26 seconds per image) is longer than other networks (0.05–0.20 seconds per image) in processing one image with the size of 512×512. In future work, the performance of deep networks needs to be improved in crack detection accuracy as well as computing efficiency to ensure more accurate and fast measurement of pellet quality.
Soluble and insoluble O contents ([S. O] and [I. O], respectively) in Al-killed steels of various C contents (0 to ~2 mass pct.) were simultaneously measured using “two-stage” inert gas fusion infrared absorptiometry, where majority of the inclusions was alumina. Several steel specimens were used for the O content analysis where the specimens were either taken from a steel plant or prepared in this laboratory. In addition to this, several pretreatment methods of the specimens were tested in order to provide reliable analysis results. It was found that mechanical grinding of the specimen’s surface should be carefully carried out in order not to induce unwanted surface oxidation. This resulted in overestimation of the O content of the specimen. [S. O] and [I. O] in the steel specimens were successfully analyzed using the “two-stage” gas fusion method in the context of inert gas fusion infrared absorptiometry. Considerable portion of total O content ([T. O] = [S. O] + [I. O]) was the [S. O], which was higher than the equilibrium O content of the known Al deoxidation equilibria. Therefore, the supersaturation in the Al deoxidation was confirmed. The analyzed [I. O] was independently validated by measuring area of the exposed alumina inclusions on the polished section of the specimen using SEM. It was also found that increasing C content in the steel lowered the [S. O] while [I. O] hardly changed. It is concluded that C in the steel, mechanical stirring, and remelting, could relieve the supersaturation.
The effect of weld parameters on the microstructure and mechanical properties of AA5052/S45C dissimilar joints fabricated by a novel pressure-controlled joule heat forge welding (PJFW) was systematically investigated. The deformation behaviors of the materials during the welding process were able to be controlled by regulating the electrical resistance which affects the heating rate. Since the welding temperature decreased with the increasing applied pressure, the formation of the intermetallic compounds layer at the weld interface was suppressed. Thus, a peak joint efficiency of 85% was reached by applying a high pressure of 180 MPa.
Modelling of nitride precipitation during isothermal heating was conducted in duplex stainless steel which was rapidly cooled from high temperature before isothermal heating. Nitride precipitation is frequently harmful for the properties such as toughness or corrosion resistance in duplex stainless steels. The heat affected zone (HAZ) in the vicinity of fusion boundary in weldments tends to have more remarkable nitride precipitation because of rapid cooling from the ferrite phase enriched temperature. It was taken into considered in modelling of nitride precipitation phenomena that the growth of austenite phase from overcooled ferrite phase occurs concurrently in the HAZ in weldment of duplex stainless steel. Because the austenite phase has extremely higher solubility of nitrogen than the ferrite.
Employing 25%Cr duplex stainless steels containing the levels of nitrogen of 0.1% and 0.3%, the measurement and observation of nitride precipitation were conducted in the specimens isothermally heated at 873 K to 1073 K after rapidly cooled from 1653 K. The model to describe the change of nitride precipitation during isothermal heating considering the effect of growth of austenite phase was proposed and that was confirmed to be good fit to the experimental results.
The effect of blasting on hydrogen analysis was investigated with the aim of establishing a hydrogen analysis method for precisely measuring hydrogen that entered steel in a corrosive environment. The hydrogen existing states of the specimens blasted under various conditions were analyzed using thermal desorption analysis and the hydrogen visualization method by secondary ion mass spectrometry. The phenomenon of hydrogen entry into steel by blasting was demonstrated for the first time. It should be noted that the effect is remarkable in the case of a specimen with a large specific surface area, and the blasting becomes an inhibitory agent in the measurement of the hydrogen content in steel. The hydrogen source for increasing the hydrogen content due to blasting is mainly the water contained in the abrasive. The mechanism of increasing the hydrogen content in steel by blasting is that the fresh surface of the steel exposed by blasting reacts with the water in the abrasive, which results in the hydrogen generation and entry into steel. Additionally, the water in the abrasive remaining on the steel surface reacts with steel during the thermal desorption analysis to release hydrogen. To suppress the increase of hydrogen content by blasting, it is effective to use abrasive with low water content and to remove rust by repeating a short blasting time in order to suppress the temperature rise of the specimen.
Micro-tensile tests were employed to clarify the post-plastic-instability behavior in the shear fractures of specimens with the dimensions of 18×25×50 µm3 made from iron-based amorphous (AM) and ultrafine-grained (UFG) alloys. The AM specimen yielded by localized shear bands with an inclination angle of ~52° with respect to the loading axis, followed by sliding off almost throughout the entire specimen thickness. Micro-tensile and micro-shearing tests revealed that the Mohr failure envelope of the AM specimens could be described by a quadratic equation rather than a linear equation. Therefore, the sliding-off process is assisted by the applied normal stress, which suggests that it is caused by free-volume coalescence. For the UFG specimen, yielding set in by shear band formation with an inclination angle of ~45° with respect to the loading axis, following the Tresca criterion. Necking after shear band diffusion formed a triaxial stress state, which resulted in a final shear fracture plane via void coalescence in the UFG specimens. Voids formed along the intersection of the primary shear bands with secondary shear bands during the necking process. This indicates that the deviation of the shear fracture plane in the UFG specimen was determined by the strain development process. A comparison of the post-plastic-instability behavior between the AM and UFG specimens suggests that the external control of triaxial stress conditions is key to improving the formability of AM specimens.