The main task in the production of steel in the basic oxygen furnace (BOF) is dephosphorization Therefore, the prediction and control of the end-point phosphorus content of molten steel is of great significance. Four machine learning regression models (Lasso, Random Forest, Xgboost, and Neural Network) were established to predict the end-point phosphorus content of molten steel in the BOF based on raw and auxiliary material data, process parameters, and production quality data. The prediction effect of the four models was further compared, and their prediction results were interpreted based on the interpretability of the models and the permutation importance method. The results showed that compared with linear regression and neural network regression model, two types of ensemble tree model have higher prediction accuracy, better stability with small data sets, and lower data preprocessing requirements. The factors influencing the end-point phosphorus (P) content in BOF were ranked in order of importance as: Tapping temperature > Turning down times > Steel scrap quantity> Operation habits of different working groups > Amount of oxygen injection> Sulfur and phosphorus content of molten iron > Addition amount of lime, limestone, and lightly burnt dolomite in the slag > Slag-splashing amount.

The production of silicon steel involves complex metallurgical processes, where the kind, composition, size, and quantity of the inclusions generated affect the silicon steel properties. This article is based on the smelting process for W350 non-oriented silicon steel produced by a certain factory. By systematically sampling, at key nodes of the converter–RH refining–tundish smelting process, the change in cleanliness of molten steel in the whole smelting process, the evolution of typical inclusions, and the transformation rules for the precipitated phase were analyzed by means of SEM-EDS, ASPEX, and Thermal-Calc. The results indicate that the total oxygen mass fraction in the steel decreases by more than 95% after deoxidation alloying, and the average oxygen mass fraction in the RH outbound steel is 0.0012%. While the nitrogen mass fraction shows a rising trend as a whole, the average nitrogen mass fraction in the tundish steel reaches approximately 0.0014%. Before RH refining, large Al₂O₃–CaO–SiO₂ and Al₂O₃–CaO–SiO₂–MgO composite inclusions are the main inclusions. MnO and Al₂O₃–SiO₂–MnO inclusions are the main inclusions after RH inlet and RH decarburization. After RH deoxidation with aluminum, the inclusions were almost entirely transformed into Al₂O₃ inclusions. After RH alloying, with the content of Si and Mn increased, the inclusions transformed into Al₂O₃–SiO₂–MnO inclusions. The number of inclusions from RH desulfurization to the RH outbound stage declined significantly, and composite inclusions containing CaS and precipitates such as AlN and MnS began to appear. The inclusions’ main types were Al₂O₃–MgO–CaS, AlN–MnS, AlN, and Al₂O₃–MgO. The inclusions inside the tundish were the same, but the numbers were slightly increased due to the secondary oxidation of molten steel. More than 80% of the oxide inclusions in the whole process were between 1 μm and 5 μm in size. The average size and the number of inclusions per unit area reached 5.45 μm and 63.1 per mm², respectively, after RH deoxidation, and respectively decreased to 3.71 μm and 1.9 per mm² during the RH outbound stage, but both increased slightly in the tundish. Thermodynamic calculation shows that Al₂O₃–MgO inclusions are formed when w([Mg]) > 0.0033% in molten steel at 1873 K. Under the actual temperature of 1828K and w([Al]s) = 0.6515%, the range of w([Mg]) corresponding to the stable existence of Al₂O₃–MgO is between 0.0053% and 0.1676%. The liquidus temperature of W350 non-oriented silicon steel is 1489 °C. MnS and AlN inclusions are precipitated successively with the solidification of molten steel, and the precipitation temperatures are 1460.7 °C and 1422.2 °C, respectively. As the temperature decreases, the sequence of inclusion precipitation calculated in liquid was as follows: Al₂O₃–CaO → 2Al₂O₃–CaO + MnS → 6Al₂O₃–CaO → Al₂O₃ + AlN + MnS + CaS.

This article uses a water model with a ratio of 1:5.75 to study the influence of factors such as the position and flow rate of top and bottom composite argon blowing on the mixing time of molten steel in a 300t ladle at a certain factory. Using engine oil to simulate steel slag, the mass transfer velocity of molten steel under different bottom and top blowing positions and flow rates of the ladle was compared. At the same time, numerical simulation was used to analyze the changes in the flow field of molten steel under different ladle blowing modes. The optimal ladle composite bottom argon process was proposed and industrial experiments were conducted on site. The research results show that the stirring effect of top–bottom composite argon blowing in the ladle is significantly better than that of the pure bottom blowing mode. When the top blowing gun is located 300 mm at the bottom of the ladle, the mixing time of the molten steel is shortest and the stirring efficiency is highest. The higher the insertion depth of the top blowing gun, the faster the flow rate of the molten steel, and the smaller the proportion of dead zones. Top and bottom blowing can improve the mass transfer rate between steel slag and promote the formation of refined slag. Through industrial experiments, it was found that the S content in the molten steel decreased by approximately 22.3% and the total oxygen content decreased by 25% before and after 10 min of composite argon blowing at the top and bottom of the ladle.

Given the increasing demand for low-phosphorus molten iron in high-value-added steel production and the rising phosphorus content in raw materials caused by the use of high-phosphorus ores in blast furnaces, the traditional converter single-slag process faces challenges such as high dephosphorization pressure, high slag consumption, and unstable endpoint control. This study systematically investigates the process principles and key influencing factors of the converter double-slag method (MURC process) as an efficient pretreatment technology for molten iron. Through thermodynamic analysis combined with industrial tests, the core process parameters affecting dephosphorization efficiency were identified, including temperature, slag basicity (R), iron oxide (T.Fe) content, and bottom-blowing stirring intensity. The results show that the optimal temperature during the dephosphorization stage is 1350–1400 °C, with slag alkalinity controlled at 1.6–2.0 and T.Fe content maintained at 19–23%. During the decarburization stage, the optimal temperature is 1620–1640 °C, and the final slag alkalinity should be increased to above 3.5. After applying the optimized “low-high-low” oxygen supply pattern and enhanced bottom-blowing stirring (0.04–0.20 Nm³/(t·min)), significant improvements were achieved in industrial practice on 180-t and 60-t converters. Lime consumption was reduced by more than 30%, the average endpoint phosphorus content decreased by approximately 0.005%, the phosphorus removal rate remained stable at above 90%, and the oxygen content in molten steel at the endpoint decreased by 50–100 ppm. This study provides a systematic theoretical basis and practical guidance for efficient and stable dephosphorization using the converter double-slag process.

Weathering steels have been widely used in transportation, energy facilities, and civil infrastructure because of their good atmospheric corrosion resistance and long service life. Their corrosion resistance is generally achieved by adding alloying elements such as Cu, P, Cr, and Ni, which promote the formation of a stable and protective rust layer. Among these elements, however, Ni and Cr are relatively expensive and their resources are limited. Therefore, reducing the use of Ni and Cr while maintaining good corrosion resistance has become an important issue in the development of weathering steels. Recent studies have demonstrated that rare-earth elements can markedly improve the corrosion resistance of steels through multiple mechanisms, including optimization of rust-layer structure, microstructural refinement, inclusion modification, and regulation of grain-boundary characteristics. Rare-earth microalloying therefore provides an effective approach to reducing the consumption of conventional alloying elements in weathering steels. This paper systematically reviews the corrosion mechanisms of weathering steels, the evolution of rust-layer structures, and the roles of alloying elements, with particular emphasis on the effects of Cu, P, Cr, and Ni on the densification and stabilization of rust layers. The microalloying effects of rare-earth elements, their inclusion-modification behavior, and their deoxidation and desulfurization mechanisms are also summarized, and their roles in reducing pitting susceptibility and promoting the formation of stable protective rust layers are clarified. In addition, the occurrence states of rare-earth elements in steel and their effects on performance stability are discussed. Key challenges encountered in the industrial application of rare-earth steels, including fluctuations in rare-earth yield and continuous-casting nozzle clogging, are also summarized. The available evidence indicates that the effective application of rare-earth elements in weathering steels remains constrained by molten-steel cleanliness, inclusion control, and the stability of continuous-casting operations. Future research should focus on technologies for the stable addition and controlled occurrence of rare-earth elements under low-oxygen and low-sulfur smelting conditions, thereby enabling the efficient utilization and stable industrial application of abundant rare-earth resources in weathering steels.

To optimize the hydrodynamic conditions of the converter vanadium extraction process, a three-dimensional multiphase flow model of the top-blowing system in a 200 t vanadium extraction converter was established based on computational fluid dynamics (CFD). Conventional 3-hole and “3+1”-hole (with a central nozzle) oxygen lance schemes were designed to investigate the effects of nozzle design Mach numbers and inter-nozzle flow distributions on jet characteristics, impact cavity morphology, and internal molten bath flow fields, aiming to select the optimal oxygen lance nozzle structure. The results indicate that for oxygen lances without a central nozzle, lowering the design Mach number mitigates shock wave energy dissipation and enlarges the gas–liquid reaction surface area; however, due to the lack of longitudinal penetrating force from a central jet, a large stagnation dead zone is prone to forming at the bottom. For oxygen lances with a central nozzle, the inter-nozzle flow distribution governs the impact cavity morphology and the evolution of the flow field. Excessive central flow causes the impact cavity to exhibit a “deep and narrow” profile and exacerbates surface kinetic energy dissipation, whereas insufficient central flow results in a “shallow and wide” cavity that makes it difficult to drive deep circulation. Based on a multi-objective evaluation, the “3+1” configuration adopting a 1:1 balanced flow ratio between the central nozzle and a single peripheral nozzle (Case 2#) effectively balances the allocation of jet momentum between radial expansion and longitudinal penetration, achieving synergistic optimization of the gas–liquid reaction interface expansion and deep-bath stirring, thus serving as the optimal scheme. The findings of this study provide an important theoretical reference for the engineering design and industrial trials of oxygen lance nozzles in large-tonnage vanadium extraction converters.