To reduce the reducing agent ratio and CO₂ emissions in blast furnace operation, it is important to control the material structure of sintered ore, which affects its metallurgical and mechanical properties. Multicomponent calcium ferrites (also called CF or SFCA (silico-ferrite of calcium and aluminum)), which is a type of melting and solidification structure, has attracted considerable interest recently, and the chemical composition and crystal structure of each CF have been researched. Although the crystal structure of CF has conventionally been analyzed mainly by XRD, the atomic arrangement could not be observed directly. Therefore, in this study, CF was investigated at the atomic level by scanning transmission electron microscopy (STEM). This research revealed that acicular CF, which was previously understood to be SFCA-I, has a SFCA (≠ SFCA-I) structure. It was also found that columnar CF had a non-periodic SFCA structure induced with a magnetite-like structure. Furthermore, a CF in which SFCA and SFCA-I were alternately stacked repeatedly was also discovered. This research clarified the fact that CF has a non-periodic structure at the atomic level.

The quasi-particle structure significantly affects the combustion heat and mass transfer process. To investigate the influence of the solid fuel existence state in iron ore sintering on the combustion behavior and CO and NO emission characteristics, four kinds of quasi-particles with different structures were prepared by using pure analytical reagents and coke for non-isothermal combustion experiments and CO and NO emission intensity detection. The following results were obtained. The S-type adhesive layer increases the diffusion resistance of the gas phase, resulting in the prolonged residence time of CO and NO in quasi-particles. The strong exothermic reduction reaction of CO–NO was promoted, the combustion efficiency was significantly improved, and the CO and NO emission concentrations were greatly reduced. The P-type had fines cluster structure, and the gas diffusion channel was easily blocked by the formed high-temperature liquid phase. The gas phase diffusion resistance was increased, the gas phase reaction between CO and O₂ and NO was promoted, the CO and NO generation rates were reduced, the heat release was increased. The all-around performance of the S’-type and the C-type was the worst. The gas phase products CO and NO generated by combustion were easy to escape quickly with the external airflow, and the emission concentration was higher.

Sintering, as one of the major sources of carbon emissions in the steel industry, is in urgent need of innovation in carbon reduction technology under the targets of carbon peaking and carbon neutrality in China. We propose the carbon free sintering process which is similar to the belt roaster, where heat is provided by gas fuel combustion for the layer. In order to achieve the process, this study uses the micro-sintering method to simulate it by varying the sintering temperature, sintering time and binary basicity of the mixture. It was investigated the process of mineralisation solidification of carbon-free sintering and to provide relevant guidance for future carbon-free sintering process experiments. As the sintering temperature increased, the strength and calcium ferrite of the samples first increased and then decreased, reach the maximum value at 1300°C and 1275°C respectively. However, the porosity of the samples first decreased and then increased, reach the minimum value at 1275°C. As the sintering time increased, the strength and calcium ferrite first increased and then decreased, reach the maximum values at 12 min and 6 min respectively, and the porosity of the samples first decreased and then increased, reach the minimum values at 6 min. As the binary basicity increased, the strength of the samples first decreased and then increased, reach the minimum value at basicity 1.8, and the calcium ferrite and porosity of the samples gradually increased. The carbon-free sintering process does not change the consolidation process of iron ore, and it has theoretical feasibility.

Because the CO gas is usually used as the reduction gas in the blast furnace process, a huge CO₂ gas has been emitted during the ironmaking process. Therefore, H₂ reduction gas has been proposed as a potential alternative to the CO gas for achieving carbon neutrality. However, the diffusion behaviors of CO and H₂ gases inside the iron-oxide particle are markedly different due to the higher gas diffusivity of H₂ gas. The reaction surface is observed in the CO reduction whereas the H₂ reduction has a broadly-reaction area. The conventional reduction analysis models were suitable for use in the CO reduction, as they assumed an exponential gas diffusion behavior through the certain reaction surface inside the particle. However, exponential diffusion is not sufficient to analyze the broad diffusion aspect of H₂ gas. In this study, the H₂-based reduction reactions is applied to the 3D diffusion model, which can accurately analyze the broad H₂ diffusion behavior. The gas components considered were the CO–CO₂–H₂–H₂O–N₂, considering the conditions of the blast furnace. The necessity of the 3D diffusion model was analyzed by comparing the H₂ reduction distributions with those obtained using the shrinking core model. The intra-particle distribution for reducing iron oxide particles, which have pellet and sintered ore shapes, were analyzed in CO–H₂ and CO–CO₂–H₂–H₂O–N₂ gas to clarify the impact of H₂ on reduction behavior. As the results, the presence of H₂ gas affected the effective gas diffusivity of the gas mixture, the reduction rate increased with the H₂ content.

High-chromium vanadium-titanium magnetite (HVTM) is a critical polymetallic ore resource, and its large-scale utilization is considered feasible by smelting flux pellets in blast furnaces. This study investigated the softening-melting properties of pure HVTM pellets (100% HVTM pellets) and HVTM with 30% conventional iron pellets (70% HVTM pellets) with different basicity (R). The results indicated that the softening-melting properties of 100% HVTM pellets deteriorated when R increased from 0.2 to 1.8, and the properties of 70% HVTM pellets first deteriorated and then improved. Moreover, 100% HVTM pellets exhibited superior softening-melting properties for R<1.0, while 70% HVTM flux pellets were better at R>1.6. The gas permeability of 100% and 70% HVTM pellets decreased with increasing basicity, although the permeability of 70% HVTM pellets was higher. Notably, the key components of acid slag are Mg-bearing anosorite and pyroxene, whereas those of basic slag are perovskite and melilite. With the increase of basicity, the content of chromium in dropping iron decreased and that of vanadium increased.

In recent years, the gangue component in iron ore deposits has increased owing to the increasing scarcity of high-grade iron ore. Such high levels of gangue causes a loss in strength of the granules used in the sintering process and a decrease in operating efficiency. In this study, the granule strength is compared by applying two types of samples, namely commercial iron ore and the gangue-free iron ore prepared from reagents, to both nuclear ore and fine ore granules. The experimental results confirm that gangue components have little effect on the wet compressive strength. However, the dry compressive strength increases with the addition of gangue in the fine ore. Granules of gangue-free fine ore exhibit significantly lower compressive strength. These results are attributed to the difference in the Coulomb repulsion force caused by the zeta potential of the fine ore.

Hydrogen-enriched blast furnace (BF) operation is currently being assessed to mitigate greenhouse gas emissions while the steelmaking industry transitions to low carbon emission technologies. Increasing the usage of lump ore in the BF also presents opportunity to decrease carbon emissions, as it can be directly charged to the furnace without agglomeration. Use of lump ore in modern blast furnace operations is facilitated by high temperature interactions with sinter. With more emphasis on hydrogen enrichment in BF operations, the behaviour of lump and sinter mixed burdens must be characterised under new conditions. In this study, 15% hydrogen is added to the standard gas conditions of a Softening and Melting (S&M) apparatus (replacing nitrogen). Analysis of auxiliary reactions such as the Boudouard Reaction and the Water-Gas Shift Reaction is presented and their impact on burden reduction and performance assessed. Results indicate that with the inclusion of hydrogen, the performance of sinter burden deteriorates, while lump burden shows significant improvement. Interaction between sinter and lump still occurred with the inclusion of hydrogen in the gas, and the mixed burden behaviour of 20% lump and 80% sinter fell between that of the individual burdens. From interrupted experiments, it is noted at high degrees of reduction, the lump burden forms a solid metallic layer which maintains its interparticle voidage at high temperatures, supressing exudation of liquid slag.

Since the 20th century, the pressure on carbon emission reduction has increased gradually in blast furnace iron-making. Using the high reducibility sinter is considered as the way to further reduce carbon emissions. In this study, the effect of sinter reducibility on the softening-melting-dripping behavior is investigated based on sinter-coke-sinter-coke-sinter layered charging. It provides theoretical guidance for optimizing the burden structure. The results show that the softening-melting behavior is determined by the upper layer sinter under experimental conditions. With increasing sinter reducibility, softening start temperature drops from 1150°C to 1046°C, softening end temperature drops from 1270°C to 1185°C, melting start temperature decreases from 1270°C to 1197°C, dripping temperature increases from 1477°C to 1516°C. The cohesive zone of high reducibility sinter widens. Gas permeability index increases from 1042.20 Kpa·°C to 1199.56 Kpa·°C. Under 1150°C, gas utilization ratio increases with increasing sinter reducibility, and then the gas utilization ratio decreases as the melting begins. Therefore, the packing structure has a significant impact on the use of high reducibility sinter. High reducibility sinter can only be used to reduce emissions under a reasonable burden structure.

In iron ore sintering plants, the emissions of nitrogen oxides (NO_x) must be reduced. Previous studies have shown that calcium ferrites could significantly reduce NO_x emissions. However, the reaction paths between calcium ferrites and NOx have not been clarified yet. In this study, NO_x reduction tests were conducted at 1073 K–1373 K with a CO/CO₂ ratio of 0.03–0.10. Accordingly, the phase compositions of calcium ferrites before and after NO_x reduction were investigated. CaO·FeO·Fe₂O₃ (CWF), which had a significant effect on NO_x reduction, was formed with a CO/CO₂ ratio of 0.03–0.10. Based on the phase compositions of the calcium ferrites before and after NO_x reduction, we determined that the CWF reduced the amounts of NO_x, and CaO·Fe₂O₃ (CF), 2CaO·Fe₂O₃ (C2F), and Fe₃O₄ are produced from the CWF. The NO_x reduction behavior during coke combustion tests was also investigated. By coating a mixture of iron ore and quicklime around the coke, NOx formation was decreased during coke combustion. This phenomenon may have been caused by the formation of reduced calcium ferrites from a mixture of iron ore and quicklime on the coke surface.

In this paper, the cause of bending fracture of high-strength fasteners after electroplating and the effect of T.Ca, T.O, and T.S contents in Al-bearing high carbon steel on the type, size, and shape of inclusions were studied. The fracture surface of fastener exhibited stepwise cracks and a lot of intergranular facets associated with ductile tearing, which were typical of hydrogen embrittlement. The agminated elongated MnS inclusions in the center of fasteners could aggravate the susceptibility of hydrogen embrittlement. Thermodynamic calculation showed that the ideal modification of MnS inclusions needed to meet the conditions that the inclusions in molten steel before solidification should be C₁₂A₇, and the mass ratio of CaS/MnS at the end of solidification was 23.7. Based on the above conditions, the quantitative relationships between the contents of T.Ca, T.O, and T.S were determined. Among them, the T.O content in steel increased with the increase of S content. Therefore, to suppress the formation of type B-stringer shaped inclusions and improve the cleanliness of steel, the sulfur content in steel should not be high.

In the high purification refining processes, a CaO-containing flux or slag is used commonly for the desulfurization of molten steel. Because the slag generated during the actual process is not always a homogeneous liquid, the quantitative relation between the liquid fraction of the slag and its mass transfer coefficient must be clarified. Therefore, to elucidate the effects of the liquid fraction on the desulfurization rate and the mass transfer coefficient in the slag phase, k_slag, desulfurization experiments were performed on molten steel using a resistance heating furnace with solid CaO-liquid coexisting slag corresponding to the CaO–Al₂O₃ and CaO–CaF₂ systems. The overall mass transfer coefficient and sulfur partition ratio at the slag-metal interface were evaluated based on the regression analysis of the experimental data using the reaction rate equation, and k_slag could be determined. It was found that k_slag decreased slightly with a decrease in the liquid fraction of the slag for both the CaO–Al₂O₃ and CaO–CaF₂ systems. This is because the effective diffusion coefficient decreased with the decrease in the liquid fraction. Two empirical equations for k_slag for the solid-liquid coexisting compositions were formulated; these were based on a previously reported equation for the effective diffusion coefficient and the regression analysis of the experimental data obtained in this study. The threshold value of the liquid fraction between the equations for k_slag was determined to be 0.87. This is possibly because the state of the solid in the slag changes because of the percolation transition.

Ladle bottom-blowing is one of the most important ways to remove the inclusions in the molten steel during steel making process. Many works have been reported to optimize the bottom blowing parameters by numerical simulations. However, it is a challenge to capture the collision between the inclusions and the bubbles during a discrete time step, which may lead to the underestimation of the inclusion removal rate. In this work, a program based on the user-define-function (UDF) in ANSYS FLUENT was developed to solve the problem above. The bottom blowing parameters of the ladle were optimized based on the developed novel model.

The microstructure and mechanical properties of high-speed steels are sensitive to some surface-active additives. In this work, four M42 high-speed steel ingots containing tellurium (Te) in concentrations ranging from 0 to 500 ppm were prepared with a vacuum induction furnace, and their microstructure and mechanical properties were systematically investigated. Te combines with Mn and possibly forms MnTe phase located in the interdendritic regions together with eutectic carbides. Te obviously refines the primary dendrite stem and decreases the secondary dendrite arm spacing of as-cast M42 steel. Te hardly affects the fraction of eutectic mixtures, but greatly refines the size of eutectic mixtures. The presence of 163 ppm Te strongly promotes the formation of M₆C eutectic carbides. This trend, however, is reversed at higher Te contents. 163 ppm Te increases the red hardness of M42 steel from 58.51 HRC to 60.8 HRC by suppressing the dissolution of secondary carbides. Further increasing Te content has no significant effect on the red hardness. Moreover, Te slightly deteriorates the tensile strength of as-cast M42 steel, while substantially improves the compressive and bending strength of tempered M42 steel. The reasons why Te affects the formation tendency of M₆C eutectic carbides as well as the mechanical properties of M42 steel were discussed.

During peritectic steel continuous casting, mold flux properties and flux film structures play significant roles in controlling slab quality. In this study, mold fluxes and flux films for casting peritectic steel slab were obtained using mineral raw materials such as quartz, wollastonite, fluorite, soda ash and others. The effects of mineral raw materials on mold flux properties and flux film structures were investigated through the measurement of melting point, viscosity, crystallization temperature, critical cooling rate, crystallization ratio and crystalline phase content. The results indicated that with increasing the quartz addition (16 to 24 mass%) and the wollastonite addition (11 to 19 mass%) in mineral raw materials, the melting point, viscosity and wollastonite content of flux film increased, while the crystallization temperature, critical cooling rate, crystallization ratio and cuspidine content of flux film decreased. The melting point, viscosity and wollastonite content of flux film reduced with increasing the fluorite addition (8 to 16 mass%) and soda ash addition (10 to 18 mass%) in mineral raw materials. Furthermore, compared with soda ash, the fluorite predominantly enhanced the crystallization temperature, critical cooling rate, crystallization ratio, cuspidine content of flux film. In addition, it was showed that the heat transfer performance and the slab quality might be primarily attribute to the crystallization ratio and cuspidine content of flux film. These results provided a theoretical foundation for optimizing the mold flux of the peritectic steel and were vital to improving the slab quality.

The hearth liquid level is essential for optimizing the tapping strategies. Considering that the tapping strategy is crucial for the stable operation and longer campaign life of a blast furnace, a hearth liquid level monitoring (LLM) system is developed by using the strain gauge method together with finite element analysis. Since 2020, the real-time LLM system has been successfully implemented at Taiwan’s China Steel Corporation (CSC) on three blast furnaces (BFs): two (BF3, BF4) have the spray-cooled hearths with four tapholes, and the other (BF2) has a stave-cooled hearth with two tapholes.

The liquid level history is visualized with other tapping information; therefore, the tapping strategies can not only be optimized in regular casts but are also more flexible in coping with unexpected incidents. With the aid of a real-time LLM system, the tapping strategies has been significantly improved by (a) making data-driven decisions to open and close a cast, (b) making quick decisions to interrupt a cast, (c) optimizing the gap time of one-sided casting, (d) using the optimal remedy for rapidly increased liquid level, (e) improving the hearth state during the blast furnace operation, etc.

The statistical analysis on the completed casts shows that the LLM system does significantly contribute to a more smooth and stable tapping operation: for BFs 2, 3, and 4, the deviations of cast duration are decreased by 31, 26, and 22%, and the mean values of slag-iron ratio are increased by 8.3, 3.5, and 3.8%, respectively.

Aiming at the problem of complex internal environment of blast furnace and the difficulty of gas flow distribution (GFD) detection and prediction, an intelligent prediction and real-time monitoring system of blast furnace roof gas flow distribution based on long short-term memory neural network (LSTM) and fuzzy C-mean clustering (FCM) is proposed, which solves the problems of poor stability of traditional GFD model, low accuracy of multi-step prediction and “black box”. The system consists of a GFD identification model and a multi-step prediction model. The system consists of GFD identification model and cross-beam temperature measurement (CBTM) multi-step prediction model. The GFD identification model first uses the blast furnace CBTM information to establish temperature field of blast furnace burden surface, calculates the center and edge gas flow development index, and adopts the FCM algorithm model to identify the blast furnace GFD pattern; the CBTM multi-step prediction model uses the LSTM model to predict the temperature of 29 CBTM points respectively; Finally, the system returns the predicted CBTM data to the GFD identification model to complete the prediction of the GFD model in the future moment. The experiment shows that the system can effectively predict the GFD development pattern within the next 7 hours, the model prediction accuracy reaches 99%, and the correct rate of all kinds of GFD pattern recognition is above 95%, which can achieve better intelligent prediction and real-time monitoring effect than other traditional GFD models, and provide effective help and support for blast furnace operators to analyze the furnace condition.

As a challenging problem in industrial scenarios, few-shot steel plate surface defect detection aims to detect novel classes given only few defect samples. Most existing few-shot object detection (FSOD) methods usually cannot accurately detect the complex and diverse steel plate surface defects, especially when the defects share similar appearance. To solve this problem, we propose a novel meta-learning based few-shot detection method with multi-relation aggregation and adaptive support learning strategy. Our method follows the training paradigm of dual-branch meta learning and tries to exploit the implicit relationships between query and support images. More specifically, we design a Multi-Relational Aggregation (MRA) module to aggregate query and support feature from three different perspectives: the attention relation, the depth-wise convolution relation, and the contrastive relation. MRA module is used to guide the subsequent classification and regression by mining the commonalities and differences between the query and support images in a category-independent manner. Besides, we propose an Adaptive Support Learning (ASL) module to dynamically adjust the weights of support representations in the learning process. We evaluate our method on three datasets of steel plate surface defect (F-SSD), NEU-DET, TianChi Aluminium profile surface defect (F-TCAL) and thorough experiments we demonstrated that our model outperforms existing state-of-the-art methods by a large margin on multiple settings. Our work provides a promising direction for the field of few-shot defect detection and can be generalized to other industrial scenes.

The fatigue crack propagation properties of newly-developed SM490 grade steels were investigated in comparison with a conventional steel of the same grade. The fatigue crack propagation rate of the developed steel in Region II of the da/dN-ΔK relationship was suppressed to about 1/2 that of the conventional steel, and its ΔK_th value was more than twice as large as in the conventional steel. However, fatigue crack resistance for long crack propagation does not necessarily improve the fatigue life in a condition of increasing ΔK from a small defect, which is usually detected in practical fatigue damage in actual structures in service. The developed steels were subjected to surface crack propagation tests using specimens with artificial small defects to examine their potential under more practical conditions. The fatigue life of the developed steel was about three times longer than that of the conventional steel. A detailed analysis of the surface crack propagation revealed crack propagation below ΔK_th only in the developed steels, which suggested the so-called “short crack regime” in a fatigue crack. The crack propagation from a surface defect that deviated from long crack behavior was convincingly explained by the corrected threshold using the R-curve model of a short crack proposed in the previous literature. Based on the experimental fatigue life improvement and its analytical estimation of the propagation resistance in the short crack regime, the effect of the ΔK_th value for a long crack in the initial propagation stage just after fatigue crack initiation was discussed.

Inhomogeneity in microstructures along the thickness direction in the stir zone of duplex stainless steel (SUS329J4L) welded at rotational rates of 275, 400, and 800 rpm was investigated using the electron backscattered diffraction method. Changes in the volume fractions and average grain sizes of ferrite and austenite along the thickness direction may reflect the temperature gradient along the thickness direction. However, near the top surface, significant grain refinement occurred, presumably owing to the introduction of additional strain by the shoulder. The kernel average misorientation (KAM) values in the stir zone were higher in austenite than in ferrite along all thickness directions, which is inferred to be related to the difference in dynamic recrystallization behavior governed by stacking fault energy, which is lower in the austenite phase than in the ferrite phase. The layer thickness per unit length of the layered structure became smaller than that of the base metal as the rotational rate of friction stir welding (FSW) was reduced to 275 rpm, which implies that new grains nucleated during FSW. Furthermore, some ferrite grains nucleated at the austenite/austenite grain boundaries, satisfying the Kurdjumov-Sachs orientation relationship. FSW is assumed to promote the nucleation of new grains with different phases, probably because of the stirring effect of the elements by FSW. In a duplex structure formed in the stir zone of FSW, a linear relationship between the ferrite and austenite grain sizes was found to hold irrespective of the rotational rate.

The large heat input of submerged arc welding (SAW) usually leads to coarse grains, reducing the mechanical properties of weldments. In this work, double-pulsed (DP) current waveform modulation technology was innovatively applied in SAW. The SAW experiments with/without DP current were performed to investigate the comprehensive effect of low frequency on the welding process stability, weld formation, microstructure and mechanical properties of the Q235 weldments. The results demonstrated that DP current significantly improved welding stability, and high-quality weldments with slight undercuts and spatters, without welding collapse, cracks or hump were obtained by DP-SAW. The low frequency had a significant effect on the spacing between each fish scale pattern. In addition, the microstructure of weld metal mainly consisted of proeutectoid ferrite, acicular ferrite, fine-grained ferrite and slight pearlite, while the microstructure of the heat-affected zone consisted of proeutectoid ferrite and widmanstatten. The stirring action of DP-SAW on the molten pool refined the weld microstructure, which improved the mechanical properties of weldments. With the optimal low frequency of 4 Hz, the microhardness, tensile strength and impact toughness of DP weldment were 188.2 HV, 537.9 MPa and 81.1 J/cm² respectively, which were enhanced by 11.1%, 8.1% and 9.8% compared with that of SP weldment respectively. The obtained results provide a new idea for improving the weld quality of submerged arc welding.

For ZnMg production feasibility, the adhesive strength of ZnMg alloy coatings must be raised to the level of commercial EG or GI steels. In this study, ZnMg/Zn bilayer coatings with various void contents in the Zn interlayer were synthesized using electromagnetic heating deposition, and the adhesion strength of ZnMg/Zn bilayer coatings on TRIP steels was investigated. As the input current during deposition decreased, the density of the Zn interlayer decreased from 83.2% to 93.2%, and the preferred orientation of the Zn interlayer changed from (101) to (002). During the ZnMg deposition on top of the Zn interlayer, a similar preferred orientation was observed in the ZnMg/Zn bilayer coating. A lap shear test result showed that the adhesion strength of ZnMg/Zn bilayer coatings increased from 19.1 MPa to 21.7 MPa as the (002) texture became dominant with decreasing void contents in the Zn interlayer. These adhesion strength results above 19 MPa were higher than those of the Zn coatings in the commercial EG and GI steels, suggesting that an additional improvement in the adhesion strength of the Zn–Mg/Zn bilayer coatings was possible by controlling the void contents in the Zn interlayer.

The cyclic fatigue, dwell fatigue and room temperature creep properties were evaluated in three types of Ti-6Al-4V forged bar samples having different micro-texture-regions (MTR) and tensile properties in the loading direction. In the S-N curve where the stress (σ_nor) was normalized by 0.2%-proof-stress, the fatigue lives of all samples were almost the same, whereas the dwell fatigue lives were not the same. So the ratio of the cyclic fatigue life to dwell fatigue life (dwell debit) changed to 2–60. In cyclic fatigue the initiation site was a facet of 1–2 α grains, and the fracture surface was typical. In dwell fatigue and creep, on the other hand, facet and dimple regions were confirmed. In addition, the facet region consisted of initiation facets of 1–2 α grains and the propagation facets which were the majority of the facet region. Initiation facets in dwell fatigue occurred earlier than 25% of the life ratio, and the angle between the c-axis of the α grains with the initiation facets and loading direction was 15–55°. The propagation facets were the MTR in which the angle between the c-axis of the α grains and loading direction was 30° or less. The lengths of the facet regions were proportional to the MTR size. In dwell fatigue, the larger the σ_nor or MTR size, the larger was the dwell debit. Therefore, the MTR size was considered the dominant factor determining the dwell fatigue life.

We have attempted to predict creep rupture time for a wide range of ferritic heat resistant steels with machine learning methods using the NIMS Creep Data Sheets (CDSs). The datasets consisted of commercial-steel data from 27 CDSs, including those on various grades of carbon, low- alloy, and high-Cr steels. The prediction models were constructed using three methods, namely, support vector regression (SVR), random forest, and gradient tree boosting with 5132 training data, to predict log rupture time from the chemical composition (19 elements), test temperature, and stress. Evaluation with 451 test data proved that all three models exhibited a high predictivity of creep rupture time. In particular, the performance of the SVR model was the highest with a root mean squared error as low as 0.14 over the log rupture time; this value means that, on average, the prediction error had a factor of 1.38 (=10^0.14). The high predictivity achieved without using microstructure information was presumably due to the fact that the data used were for commercial steels in which there should be a correlation between the chemical composition and the microstructure. We confirmed that the prediction did not work exceptionally well for two heats having the same composition but different microstructures with and without stress-relief annealing. The predictivity could be markedly increased by adding the 0.2% proof stress at the creep test temperature as one of the explanatory variables. As a demonstration of the prediction model, the effect of elements was evaluated in modified 9Cr–1Mo steels.

To develop CO₂ utilization technology in converter process, it is necessary to study the chemical reactions of the CO₂–O₂ mixed gas in the cavity. ¹³CO₂–¹⁸O₂ dual isotope gas was used to distinguish the complex reactions and analyzed the effect of injection height. The results show that as the injection height increased from 10 to 30 mm, the influence of CO₂ participating in the reaction is relatively unobvious; regarding O₂, the flowrate of uninvolved reaction increases; the flowrate of post combustion increases first and then decreases; the flowrate of O₂ directly involved in decarburization is basically stable in the range of 20–30 mm after decreasing. Maintaining the injection height of 20 mm, the O₂ flowrates of post combustion in the ranges of 0–5 mm, 5–10 mm, and 10–15 mm are 4.95 mL·min⁻¹, 13.85 mL·min⁻¹ and 6.73 mL·min⁻¹, respectively. It can reduce the waste of mixed gas due to post combustion and improve the quality of furnace gas if the injection height is not in the range of high post combustion ratio.

https://doi.org/10.2355/isijinternational.ISIJINT-2023-087

The authors have found that the numerical values in Equation (13) reported in the above paper are incorrect. The authors would like to correct the parameters as follows:

     Original values (incorrect)     Corrected values

Eq. (13) (MnO) + [C] = (MnO) + [Fe].   (MnO) + [C] = {CO} + [Mn]

The error was due to a transfer mistake of the values after the derivation of Equation 13. This correction does not affect the article’s Tables, Figures, other equations, discussion, and conclusions. The authors sincerely apologize for causing the readers inconvenience by this mistake.