Electromagnetic stirring(EMS) technology plays an important role in the cooling solidification process of molten steel for high efficiency continuous casting process, which can effectively improve the internal and external defects of continuous casting billets. Specifically, the EMS technology takes the benefit of the generation of Lorentz force to optimize the molten steel flow, so as to decrease flow instability, mold powder entrapment, and surface defects and the bloom quality is improved obviously. For this purpose, to gain a clear understanding of the effect of EMS technology on the continuous casting process, the principle of electromagnetic stirring is briefly expounded. Also, the latest research progress of mold electromagnetic stirring(M-EMS), secondary cooling zone electromagnetic stirring(S-EMS), final electromagnetic stirring(F-EMS) and electromagnetic stirring combination technology in continuous casting is analyzed in detail. The problems encountered in the application of electromagnetic stirring in continuous casting and the improvement ideas are pointed out. Finally, the future development trends of electromagnetic stirring are prospected.

With the development direction of ultra-wide and ultra-thick design of caster, it is necessary to re-evaluate the flow field characteristics of casters under extreme conditions. Based on the probable future designs of ultra-wide slab caster in China, the numerical simulation is established for the flow of molds with large cross-section of 3 300 mm×180 mm, and the flow characteristics of molds with three typical slab continuous casting nozzles were studied. The results show that the flow field of the two-orifice nozzle mold is a classical double-circulation flow pattern, while the five-orifice nozzle mold increases the disturbance of the molten steel at the nozzle and the central axis. The flow field of the CSP-type water port casting mold is a large bottom-up circulation. The fluctuation of the liquid level of the five-orifice water orifice not only ensures the fluctuation of the liquid level within the safe value range, but also increases the disturbance of the flow dead zone of the meniscus, whose optimization effect of the flow field is remarkable. The typical structural parameters of the five-hole water outlet are as follows: the angle of the main side hole is 10°-20°, the insertion depth is 150-160 mm, and the distance between the main and secondary holes is 10-20 mm.

Continuous casting process occupies a “central” position and plays a connecting role in steel production process. The development of continuous casting technology has significantly promoted high quality, high efficiency and green production in steel industry. By reviewing the award-winning projects in the field of continuous casting in the “Metallurgical Science and Technology Award” from 2020 to 2024, The new progress made in continuous casting theory, production efficiency, billet homogenization, products development, key process equipment and intelligent control technology in China were summarized. Billet homogeneity, stable production, and high efficiency are the main indicators of high-quality continuous casting process. The collaborative optimization of mold metallurgical functions, secondary cooling control, electromagnetic stirring, and end pressing are the main technical measures to improve the quality of continuous casting billets. From the award-winning projects in recent years, it can be seen that the intelligent control technologies, such as digital twin factory, mold liquid level fluctuation control, billet defect detection and prediction, automatic steel casting, etc. have made remarkable progress. Continuous casting equipment is developing towards the direction of high efficiency, stabilization, flexibility, long service life, and more intelligence, and at the same time meeting the production requirements of stable, high quality and energy saving. Reference for readers to understand the whole development of continuous casting technology in China will be provided.

Continuous casting is one of the most important processes in steelmaking. There are many factors affecting the breakout accidents in continuous casting process, among which the sticker breakout is the most common, accounting for about 70% of the total breakout. In order to better solve the sticker breakout accident, clarify the cause and provide a theoretical basis for establishing a more comprehensive and accurate prediction model, the occurrence mechanism, influencing factors and prediction model of sticker breakout in continuous casting were summarized and analyzed. The influence of mold power performance, molten steel conditions, crystallizer parameters, personnel operation and other factors on the sticker breakout accident law was mainly discussed. For the current commonly used logical judgment breakout prediction model and machine learning-based breakout prediction model, the current research status was analyzed, and the current shortcomings were put forward. It is predicted that the breakout prediction system will continue to develop in the direction of intelligence and high-end in the future, which provides a reference for the research of sticker breakout and how to effectively prevent it.

Taking 42CrMo4 round billet with a section size of ϕ800 mm as the research object, a large round billet model was established with ProCAST finite element method and CAFE method. The effects of superheat, drawing speed and specific water volume in secondary cooling zone on solidification behavior and grain structure were compared. The results show that the influence of the drawing speed on the solidification temperature, solidification end point and the thickness of the solidified billet is greater, while the influence of superheat on the grain growth is greater. When the drawing speed is increased by 0.02 m/min, the solidification end point moves back about 3 m, and the change of superheat and specific water volume in the secondary cooling zone has no obvious effect on the temperature and central solid fraction. When the superheat is increased from 10 ℃ to 55 ℃, the equiaxial crystal rate decreases from 43.2% to 9.2%, while the influence of pulling speed and secondary cooling zone specific water volume on the equiaxial crystal rate is not obvious, and the change is less than 8%. Based on the electromagnetic stirring position of the end, an optimization scheme was proposed for the existing process. The superheat was reduced to 10 ℃, the specific water volume was reduced to 0.16 L/kg, which could effectively improve the equiaxed crystal rate and refine the grains.

Inclusions in the steel have a significant impact on the performance. It is necessary to detect and analyze non-metallic inclusions to improve the quality and performance of steel products. A comprehensive introduction to the detection and characterization methods for inclusions in the steel was reviewed from three aspects: the characterization of inclusion features, the extraction of inclusions by dissolution, and non-destructive three-dimensional detection of inclusions. Different characterization methods were compared comprehensively from different aspects including size analysis range, advantages and disadvantages. Optical microscope and other methods can directly observe the morphology of inclusions with a wide analysis range, but their magnification is relatively low, making it difficult to distinguish the morphology characteristics of small inclusions. However, scanning electron microscopy and other methods can clearly observe the two-dimensional morphology and the composition of inclusions, but their analysis cost is relatively high. The erosion method can characterize the three-dimensional morphology of inclusions, but its operation is relatively complex and time-consuming. The application of non-destructive characterization methods such as ultrasonic testing can characterize inclusions without damaging the specimen, which is convenient and fast.

In order to solve the problems of low metal yield rate and insufficient modification of inclusions in high Al steel with traditional feeding calcium wire technology, feeding composite calcium aluminate cored wire was fed into molten steel. The melting behavior of composite cored wires with different diameters in molten steel as well as the effect of feeding cored wire on the flow field of molten steel and the removal ratio of inclusions were studied by the numerical simulation and water model test methods. The numerical simulation results of the cored wire melting showed that the composite cored wire with a core powder diameter of 11 mm and a skin thickness of 1.07 mm satisfied the ideal condition for feeding speed of 2-4 m/s. The water model experiment results of feeding composite cored wire showed that when the feeding depth was 100-300 mm, the removal ratio of inclusions ≤ 50 μm, 50-100 μm and 100-150 μm increased by 3.29%, 2.71% and 2.56%, respectively. When the feeding depth was 100 mm from the bottom of the ladle, with the composite core powder size increasing per 50 μm, the removal rate of inclusions ≤50 μm, 50-100 μm and 100-150 μm increased by 4.58%, 3.01% and 2.42%.

Mold fluxes play important role in ensuring the smooth casting and slab quality, which is closely related to the properties of mold fluxes. The viscosity and melting temperature of mold fluxes are key parameters in the design and production of mold fluxes; however, the determination of property is time-consuming and labor-intensive. In view of the development of machine learning technology, three machine learning algorithms including KNN, KRR and RF were employed to establish the prediction model to accurately predicting viscosity and melting temperature of mold fluxes, providing a guide to the fast and convenient design of mold fluxes. The experimental results show that compared to other two models, KRR-based model has the best performance in the viscosity prediction, in which the coefficient of determination (R²) is 0.983, the root mean squared error (RMSE) is 0.023, and the mean absolute error (MAE) is 0.014. The RF-based model is more reliable in predicting the melting temperature with R² of 0.823, RMSE of 14.004, and MAE of 8.974. The ranked importance analysis of the input features shows that the order of influence of viscosity is SiO₂, F, and Al₂O₃ and Na₂O has the greatest influence on melting temperature. Compared with the widely used viscosity and melting temperature prediction models, the mean relative errors (MRE) of the KRR-based viscosity prediction model and the RF-based melting temperature prediction model are 6.26% and 0.83%, respectively, which are much lower than MRE of the widely used prediction models, indicating that the machine learning-based models have a high reliability. Moreover, the viscosity and melting temperature distribution maps established based on the models could provide intuitive reference for the design of mold fluxes.

In order to solve the problem of poor steel flow consistency inside the four-stream tundish of a steel plant, the guide hole structure of the tundish baffle is designed, and turbulence suppressors are added to form a combined structure of the flow control device to optimize the flow field of the tundish. Orthogonal tests are designed for the structure of baffle deflector holes, and the residence time distribution (RTD) curves of each scheme are obtained through water modeling to determine the optimal structural parameters of the baffle deflector holes as well as the degree of influence of factors by combining with the polar analysis method and the actual situation. On this basis, turbulence suppressors of different structures were designed and combined with the baffle for water modeling, and the optimal production scheme was finally obtained. The results show that the scheme U2 is the optimal structure of the baffle. The parameters of the guide hole are 25° of vertical inclination, 140 mm in diameter, 400 mm in height, and 10° of horizontal inclination for the optimal structure. The dead zone ratio of this scheme is 17.83%, which is 2.12% less than that of the original scheme, and the flow consistency is improved by 84.61% compared with that of the original scheme. The flow behavior of the steel is optimal after the polygonal turbulence suppressors are applied in combination with U2. The dead zone ratio of this scheme is 16.1%, which is 3.85% less than the prototype tundish, and the flow consistency is improved by 71.54%. Compared with a single baffle structure, the combined structure of U-type baffle and polygonal turbulence suppressors can effectively reduce the dead zone ratio and the standard deviation of the stagnation time, significantly reduce the peak concentration of the water outlet, extend the peak time, enhance the consistency of the water outlet, and achieve the optimization of the flow field inside the tundish, which provides the theoretical basis for the subsequent development of the optimization of the flow field of the combination of flow control devices.

With the advancements in thin slab direct rolling and efficient continuous casting, the issue of funnel-shaped cracks in copper molds resulting from high-speed casting has become increasingly severe. Researchers had explored the formation mechanism and control of these cracks in thin slab casting by means of theoretical analysis, industrial experiments and numerical simulation methods, aiming to prolong the lifespan of copper molds and reduce production costs. The macroscopic morphology and hazards of funnel-shaped copper mold cracks were elucidated, and the mechanisms of crack formation, including thermal fatigue, creep fatigue, and elemental erosion, were summarized. A comprehensive review of the development of thermal/mechanical analysis models for copper molds was provided, along with an overview of measures for controlling copper mold cracks, new research directions were proposed for the mechanism of copper mold crack formation, thermal/mechanical analysis models, and control strategies.

Defects such as slab shape, surface cracks, center segregation, and center porosity are key issues limiting the production and development of ultra-thick slab continuous casting. A domestic steel plant has newly built a 460 mm thick slab casting machine. By studying the characteristics of the continuous casting production process for thick slabs and addressing the main defects of thick slabs, a 460 mm thick slab continuous casting production process technology has been developed, including high casting speed, secondary cooling dynamic amplitude cutting control process, high alkalinity and viscosity process of protective slag, low superheat control, and large pressing process at the solidification end of the slab. This has solved the problem of bulging of thick slabs, reduced the occurrence rate of surface cracks on thick slabs, and improved the center looseness and center segregation of thick slabs.

A transient three-dimensional numerical model coupling multiple physical fields in the slab mold was developed to investigate the effects of electromagnetic stirring on molten steel flow, initial solidification, and the non-steady-state fluctuation behavior of the steel/slag interface in a 1 500 mm×230 mm slab mold. The accuracy of the electromagnetic stirring model was validated by measuring the magnetic field intensity and electromagnetic forces within the mold during actual production. The results show that, compared to conditions without electromagnetic stirring, the traveling wave magnetic field generated by increased stirring transforms the flow pattern in the mold from an upward circulating flow to a horizontal circulating flow. This enhances the horizontal flow of molten steel, reducing the velocity of the solidification front along the casting direction and thinning the solidified shell. For instance, when the current is 500 A, the solidified shell thickness at the mold outlet decreases by approximately 3 mm. Additionally, the weakening of the upward circulation leads to a more uniform velocity distribution at the steel/slag interface. With the increase in current intensity, the horizontal flow of molten steel is further promoted, reducing the downward circulation velocity and effectively decreasing the impact depth. To quantitatively characterize the overall level fluctuation and flow of the steel/slag interface in the mold, evaluation criteria for the average level fluctuation and velocity uniformity were proposed. The results indicate that the level fluctuation and molten steel flow at the steel/slag interface are optimal under the electromagnetic stirring parameters of 500 A and 3.2 Hz.

Carbon emissions from the iron and steel industry account for a large proportion of all industries in China, and how to realize carbon emission reduction and CO₂ resource utilization in the steelmaking process under the background of "double carbon" has attracted much attention. Taking the resource utilization of CO₂ in the iron and steel industry as an entry point, it introduces the current situation of CO₂ emission and resource utilization in the iron and steel industry, and focuses on the analysis of the technical status of applying CO₂ as reaction gas, stirring gas and protective gas in the steelmaking process. Through thermodynamic calculations, it is concluded that when CO₂ oxidizes 0.1 mass% of the metal elements in the iron liquid [M], the comprehensive exothermic heat release is reduced, and it plays a role of controlling the temperature. The analysis of CO₂ as a stirring gas shows that CO₂ is used in the steelmaking process to reduce carbon emission. As a stirring gas,CO₂withthe utilization rate of about 85% can provide twice as much stirring effect as Ar. CO₂ can be used as a kind of resource gas gradually applied in the steelmaking process, with energy saving and emission reduction and cost-effective effect, but needs to further clarify the reaction characteristics of CO₂ in the high-temperature melting pool and the role of the form, in order to accurately and efficiently realize the resourceful use of CO₂.

The production of high-speed continuous casting is the achievement and realization ways of the high efficiency, green and intelligent development of continuous casting. Aiming at realizing high-speed continuous casting, the mold is studied on the process control and equipment design, and the mold model is established. Based on the heat transfer analysis of mold, it is proposed that the key to optimize high-speed mold is to improve the homogeneity of Cu plate with the premise of increasing the cooling efficiency. By the multi-objective optimization, a new method to optimize the water seam of mold is proposed. After optimization, the hot surface temperature and the temperature difference of Cu plate of the mold are reduced significantly, which can obviously reduce the risk of breakout caused by the uneven cooling of the slab at high speed casting. This is of great significance for the promotion and application of high speed production technology of caster.

During the continuous casting process of steel, the microsegregation and macrosegregation will caused by the uneven distribution of solute elements. It should be noted that the macrosegregation in the billet cannot be removed by the subsequent heat treatment or rolling process, and it affects the product properties. Therefore, it is of great significance to simulate the macrosegregation by considering the microsegregation behavior, which aims at enchancing the simulation accuracy of macrosegregation, and improving the quality of billet and the performance of products. In this work, a macrosegregation model that couples microsegregation behavior in order to compare macrosegregation behavior with different microsegregation models was established. The best solute microsegregation will be chosen, and accurate prediction of solute macrosegregation behavior in continuous casting billet will be realized. Besides, the Won-Thomas microsegregation model was used to investigate the effects of secondary dendrite arm spacing on the solidification end point and macrosegregation defects for 82B cord steel billets, when the spacing of the secondary dendrite arms is reduced from 200 mm to 40 mm, the solidification end point is shifted forward by 1.4 m, and the carbon mass fraction in the center of the continuous casting billet decreases from 0.154% to 0.138%, indicating that the larger the spacing of the secondary dendrite arms, the slower the cooling and the more serious the segregation, and the rationality and reliability of the model were verified through a carbon sulfur analyzer.

The effect of electromagnetic stirring, degree of superheat and secondary cooling intensity on the homogeneity of 200 mm×200 mm continuous casting SWRH82B billet was studied and the influence of the homogeneity of casting billet on the rating of the central grain boundary cementite of wire rod was investigated through central and semi-macro segregation tests and macrostructure tests on the continuous casting billet. The key continuous casting process of 200 mm×200 mm SWRH82B steel was developed. The low segregation control of the continuous casting billet was achieved, and an important homogeneous foundation for the control of microstructure and performance was laid. The test results showed that the homogeneity of continuous casting billet can be optimized by using the technologic parameters of M-EMS current intensity of 300 A, F-EMS current intensity of 200 A, superheat degree of (25±2) ℃ and weak secondary cooling process. With these process technologic parameters, the central carbon segregation degree of the billet can be decreased to 1.08 in average, the proportion of grain boundary cementite of as-rolled wire rod rating not higher than 2.0 reached 96%, and the average wire breaking rate was decreased to 2.3 times per 100 tons in average.

Based on the full process control of steelmaking-hot rolling-cold rolling (heat treatment) process, a digital platform for full process quality control was built using modern data communication and database (data cloud), and an online quality rating system for continuous casting slab was developed based on this platform. This digital platform for quality control throughout the entire process monitors and tracks the parameters of raw materials, production, processes, equipment, and other aspects in the steelmaking process in real time. The tracking results are fed back to each furnace and slab in the form of process events. By formulating corresponding rules, continuous casting slabs are graded and evaluated to achieve online real-time prediction and evaluation of continuous casting slab quality, in order to determine the grading and repair treatment method of continuous casting slabs. The quality inspection and judgment results of the subsequent rolling process show that the quality rating system can classify and screen continuous casting slabs with different quality risk levels more intuitively and accurately, reducing the proportion of slab defects and improving the quality of continuous casting slabs.

The inhomogeneous heat transfer, solidification, thickness of primary shell, temperature and stress of secondary cooling zone directly affect the surface and internal quality of slab. The temperature field and stress field of the whole process of Q235B slab in a steel plant were simulated by ANSYS Finite Element Software, the surface temperature, equivalent stress, shell inhomogeneity, heredity and crack sensitivity were analyzed. The results show that the non-uniformity of slab solidification thickness in mold is inherited to secondary cooling zone, and the non-uniformity of equivalent stress on slab surface is less inherited, the non-uniformity of surface equivalent stress wasgreatly improved. In the case study, the temperature at the left corner of the inner arc was the lowest (892 ℃), the equivalent stress was the highest (12.2 MPa), and the temperature at the deflection corner was the highest (1 263 ℃), the equivalent stress was the lowest (2.54 MPa). The sensitivity of slab surface crack is 1/4 center > 1/8 center > 1/2 center.The sensitivity index of slab surface crack reaches the maximum value of 0.52 in mold. In the early stage of secondary cooling, the crack sensitivity is increased.And in the late stage of solidification, the shell is thicker, the crack sensitivity is lower and the change range is small.

The star cracks are hidden in the subsurface of continuous casting slab, which are difficult to be detected in manual on-line inspection, and it is easy to cause a large number of peeling defects when the burning loss of heating furnace is insufficient. Through metallographic analysis of rolled piece and slab, it is clear that the cause of defects comes from star cracks of the slab. In view of the star crack of the slab using the full-arc continuous casting machine, the influences of typical elements (C, Al, N, V), cooling intensity of mold, slag in the continuous casting, secondary cooling strength and others on the crack defects of casting slab were analyzed. The causes of the cracks in casting slab were clarified that deep vibration marks and a large amount of precipitates form the crack initiation, and the slab is located in the brittle zone in the caster and suffers excessive stress. The effective measures to control the cracks are put forward, including reducing the contents of typical elements, increasing and stabilizing the casting speed, employing improved continuous casting slag, weakening the mold strength and secondary cooling strength, improving the quality of caster, etc. After a large number of actual applications, the crack defects of low alloy steel slab are reduced, and the peeling defect rate of hot-rolled plate is reduced from 21.84% to 0.47%.

The flow of liquid steel in continuous casting mold is a complex three-dimensional turbulent flow, which includes many complex phenomena, such as turbulence in the SEN and mold, the movement of bubbles and inclusions, and the influence of multiphase flow and electromagnetic force on liquid steel flow.The fluctuation of the liquid level of the mold is an important reason for the surface quality of the casting slab.In the actual production process, the fluctuation of liquid level generally only detects a single point or a single area, which cannot reflect the overall fluctuation state of the liquid level. However, there is no direct detection method for the flow field, which is generally analyzed and evaluated by a certain proportion of water model. The characteristic parameters (liquid level difference, liquid steel flow rate) at the mold meniscus are evaluated by numerical simulation andnail board test. The results of numerical simulation and nail board test are basically consistent. Numerical simulation results show that the EMBR current changes by 50A, the maximum surface flow rate changes by 0.01-0.02 m/s,while the maximum liquid level difference changes little. When the SEN insertion depth increases by 50 mm, the maximum surface flow rate decreases by 0.03-0.06 m/s, and the maximum liquid level difference decreases by 2 mm. The pulling speed increases by 0.3 m/min, the maximum surface flow rate increases by 0.04 m/s, and the maximum liquid level difference increases by 2 mm.SEN insertion depth and casting speed have great influence on liquid steel flow rate and liquid level difference, while EMBR current has little influence. Under the condition of casting speed 5.2 m/min, EMBR current 175A, SEN insertion depth 170 mm, the slab quality is better.

In order to evaluate the influence of baffle wall structure on the flow field, improve the flow behaviour of liquid steel in the middle tundish of a steel plant, select the optimal solution in line with the current production conditions and improve the quality of special steel grades, the guide holes on the basis of the original baffle wall structure were designed and optimized.Orthogonal experiments are designed considering the number of guide holes, mounting height, and inclination angle factors.The flow field of the tundish under the prototype retaining wall structure is evaluated by the water simulation experiment method for conducting range analysis. Thus, considering the influence of various factors under the indexes of dead zone ratio and standard deviation of peak time, the optimal diversion hole structure can be found.Numerical simulation is carried out using Ansys Fluent to simulate the flow and temperature fields of the tundish before and after optimization. The experimental results show that the flow field of the prototype tundish has problems such as poor consistency at the mouth and poor flow characteristics of the internal liquid steel, with a dead zone ratio of 29.26% and a peak time standard deviation of 61.52 s.Under the water model experiment, the optimized tundish dead zone ratio is 27.92%, which is 1.34% higher than that of the prototype tundish.The standard deviation of peak time is 87.27% higher, and the water inlet temperature difference is reduced to 0.2 K.The flow field is very uniform as a whole, and the consistency of the two streams is significantly improved.

In order to improve the internal quality of 42CrMoA steel in a factory and reduce the casting of billet center segregation, the ProCAST software was used to simulate the casting of billet solidification structure, and the effect of continuous casting parameters on the distribution of solidification organization, the rate of equiaxial crystal and equiaxial crystalline area density was studied. When the casting speed was increased from 0.8 to 1.1 m/min, the proportion of equiaxial crystal zone has increased by 5.26%; When the cooling water flows increases from 0.144 to 0.192 L/kg, the proportion of equiaxial crystal zone has decreases by 3.34%; The superheat increases from 5 to 35 ℃, and the proportion of isometric crystal zone decreases by 7.98%. It is concluded that it is possible to reduce the macroscopic segregation of 42CrMoA steel by appropriately increasing the casting speed, decreasing the amount of water in the second cooling ratio and reducing the superheat of the steel.

Aiming at the problem of the slag entrapment in the thin slab mold of a steel mill, the funnel mold was used as the research object, and the VOF method was established to model and calculate the steel-slag interface, and the 1:2 physical model and 1:1 numerical model were constructed to analyze the effects of casting speed, viscosity of mold flux and immersion depth on the flow field of the mold and the fluctuation behavior of the steel-slag interface. The surface velocity, the phase distribution at the steel slag interface, and the height of the liquid level fluctuation of the mold were chosen as the evaluation standard. The results show that the liquid level is relatively stable when the flow rate of steel-slag interface is 0.12-0.26 m/s and the viscosity of the slag is greater than 0.253 Pa·s. With the increase in immersion depth, the impact depth of the molten steel decreases, the range of the upper turning zone becomes larger, and the fluctuation height of the slag interface decreases to 5.4 mm. When the immersion depth is greater than 190 mm, the viscosity of the mold flux is 0.324 Pa·s, and the surface flow rate of the steel slag is less than 0.26 m/s, the appearance of slag rolling can be effectively inhibited, which provides a reference for improving the quality of the casting billet and the safety and stability of industrial production.

In the critical process of metallurgical production, the continuous casting is closely linked to the quality of the cast slab products. Among various parameters, the liquid level fluctuation in the mold is one of the most crucial during the continuous casting process. Therefore, accurate monitoring of the liquid level fluctuation inside the mold is particularly critical. An artificial intelligence model aimed at real-time accurate prediction of the mold liquid level fluctuations was introduced. The model employs a genetic algorithm (GA) optimized random forest (RF) technique, referred to as the GA-RF model. The model optimizes the parameters of random forest network through genetic algorithm, aiming at finding the optimal parameters of the model and obtaining the model with the best prediction performance. Experimental results demonstrate that the GA-RF predictive model achieves a mean absolute error (MAE) of 0.534 and a mean squared error (MSE) of 0.73, with a high prediction success rate of 96%. Compared with CNN model, BP model and SVM model, it is found that GA-RF model MAE and MSE are superior to other models, confirming the model’s high precision and its ability to meet the stringent requirements of practical production applications. Furthermore, through sensitivity analysis, the influence of different production parameters on the model is also discussed.

A full-structure physical model was established based on a 30-ton six-strand rectangular billet continuous casting tundish in a domestic steel plant, and the fluid-solid-thermal coupling numerical simulation was carried out. The flow field, temperature field and stress distribution behavior in the tundish were studied. Based on the von Mises stress considering the thermal effect, the damage of the refractory was evaluated. A damage model of the turbulence inhibitor-porous baffle integrated flow control element was established, and the flow and heat transfer behaviors in the tundish before and after the damage of the flow control element were compared and analyzed. The results show that the average volume temperature of each fluid domain and solid domain in the six-strand tundish is quite different. The maximum temperature difference between the working layer and the slag layer and the air layer can reach 312 and 578 K, respectively. The stress of the wall surface reaches 40-52 and 52-73 MPa, respectively, which is higher than that of the wall surface in contact with the molten steel(32-40 MPa).In the integrated flow control element, the easily damaged area is mainly three parts: a turbulence suppressor, a retaining wall diversion hole, and the connection between the corner of the porous retaining wall and the side wall as well as the bottom. The stress in these parts reaches 35-52 MPa. After the failure of the integrated flow control element, the response time of No.1 and No.2 outlets decreased by 26 and 17 s respectively, the volume fraction of the piston zone decreased by 25.5%, the volume fraction of the dead zone increased by 9.4%, and the consistency of each flow was obviously weakened. The minimum temperature of molten steel at the outer corner of No.3 outlet decreased by 14 K. The field industrial test shows that the failure morphology of the integrated flow control element obtained by numerical simulation in this paper is in good agreement with the actual failure situation.

In the steel production process, continuous casting billets are cut according to length; rolled steel is rolled by weight; and the inconsistency of steel rolling implementation standards directly causes fluctuations in the total length of rolled products. These all resulted in the generation of a large number of short length materials and waste materials, which has become an industry pain point that needs to be solved urgently. The basic principles of fixed-weight cutting technology, influencing factors of fixed-weight cutting and related models of continuous casting billet are expounded. The intelligent fixed-weight system is developed to make continuous casting billets cut by weight to realize steel-rolling coupling, which has been successfully applied to Hebei Tangyin Iron and Steel Co., Ltd. The results of the model show that the difference between the theoretical weight and the actual weight is ±0.2%, the fixed weight pass rate reaches more than 90%, the actual weight yield reaches more than 97.5%, and the fixed size rate is stable at more than 99.5%. This greatly improves the negative difference stability and yield level of small bar size products while maintaining high yield, and provides a reference for the development of fixed weight cutting technology for related enterprises.

Oriented silicon steel is known as “art” in iron and steel materials for the reason of complex production process and strict manufacturing technology. In order to improve the casting speed of conventional slab of oriented silicon steel and improve the profit level of enterprises, the solidification end test and the bulging behavior of casting slab under different casting speeds were carried out, and the following results could be drawn. At the casting speeds of 0.9 and 1.0 m/min, the solidification end of oriented silicon steel slab is located in first 7 sections, while at high casting speed of, the solidification end is in the 8th section. When the casting speed is increased from 0.9 to 1.1 m/min, the thickness of the billet shell decreases. Under the same static pressure, the bulge volume and bulge deformation rate increase, with the maximum bulge deformation rate of 0.107%. When the casting speed increases by 0.1 m/min, the bulge volume of the billet at the same position increases by about 3.7%-16.6%. Combining the experimental study of liquid level fluctuation of mold, bulging behavior of billet and solidification end position, the casting speed of conventional slab of oriented silicon steel can be increased from 0.9 to 1.1 m/min.

In order to solve the problem of internal cracking of vanadium-containing steel in the continuous casting process, the solidification process analysis, organization prediction and surface temperature control of HRB500E rebar steel were carried out by using thermodynamic calculations, solidification organization simulation and the method of second-cooling water distribution control. It was found that the thermodynamic precipitation temperature of V(C,N) in the rebar steel was 1118 ℃ and the maximum precipitation amount was 0.06 wt%, which indicated that vanadium had the dual raction of precipitation strengthening and solid solution strengthening; the organization analysis of the as-cast billet showed that when vanadium was increased from 0 wt% to 0.089 wt%, the grain radius was reduced by 2 times, the number of grains was increased by 2.8 times, and the central equiaxial crystal was increased by 2.2%. It shows that the addition of vanadium can refine the grain, improve the organizational properties and gain metallurgical properties; at the same time, it is found that the reason for cracks and other defects is mainly due to the unreasonable water distribution of the second cold, resulting in the surface of the billet return temperature is too large. Based on this, a corresponding optimization of secondary cooling water distribution process has been proposed, along with the introduction of a three-section program for 2.9 m length of secondary cooling process, and ultimately designed a 1.4-2.7 m/min pulling rate of the water table, the surface of the billet return temperature was below 100 ℃ and the return temperature met the metallurgical guidelines. Industrial production has proved that the internal quality of billet has been greatly improved by the optimization of water distribution in the secondary cooling.

The clogging behaviors of sub-merged nozzle (SEN) in three types of Ti-baring Al-killed steels including ultra-pure ferritic stainless steel, austenitic stainless steel and interstitial-free steel were studied by cross section observation and acid dissolution method. The results show that Al₂O₃ inclusion is the main inclusion comprising the transition layer, and MgO·Al₂O₃ caused by poor calcium treatment is the main factor for the growth of nozzle clogs. TiO₂ formed by the reaction between liquid steel with refractory material is the main inclusion in the transition layer, and subsequently the TiN precipitation caused by poor control of titanium-nitrogen product and steel cleanliness combined with rapid heat transfer of transition layer deteriorates castability. Metallic iron comprises a major amount of the transient layer of interstitial-free steel and Al₂O₃inclusion plays a dominant role in the clogging layer. TiN inclusions have less influence on clogging behaviors during continuous of ultra-pure ferrite and interstitial-free steels. The good wettability of titanium-containing liquid steel and titanium oxide will accelerate the clogging process in interstitial-free steels. The personalized strategy should be adopted for the mitigation of SEN clogging according to the steel characteristics and refining equipment.

To address the significant erosion observed in the tundish dam with jet hole, a 65 t billet tundish from a domestic steel plant is researched. It employs numerical simulation to examine the distribution of impact stress on the upstream surface of the dam. The findings clarify the factors affecting the impact stress on tundish dam. Results indicate that the impact stress along the upper edge of the upstream surface is negative. Additionally, a specific impact stress exists at the lower section of the central axis, with peak stress recorded around the jet hole. The peak impact stress on the dam’s upstream surface exhibits a positive correlation with increasing casting speed. Variations in the distance between the weir and long nozzle show minimal effects on the peak impact stress, yet significantly alter the stress distribution around the jet hole. The impact stress distribution shows non-uniform characteristics at 1 400 mm, but demonstrates significantly improved uniformity at 1 800 mm.

In the continuous casting process, tundish plays an important role in controlling the number of inclusions and improving the quality of the casting billet. A three-dimensional mathematical model was established by using the open-source computational fluid dynamics software OpenFOAM by numerical simulation, and the effects of particle size, molten steel flow rate and shape of non-metallic inclusions on the removal rate of non-metallic inclusions in tundish were analyzed. The results show that with the decrease of the particle size of the inclusions, the removal rate of non-metallic inclusions gradually decreases. When the flow rate of molten steel is 0.47 m³/min, the removal rate of non-metallic inclusions with a particle size of 200 μm reaches 97.64% and the removal rate of non-metallic inclusions with a particle size of 10 μm is only 1.298%. When the particle size of the inclusions is less than 100 μm, the removal rate increases significantly with the increase of particle size; and when the particle size of inclusions exceeds 100 μm, the increase of particle size has no obvious effect on the removal rate. With the increase of molten steel flow, the removal rate of non-metallic inclusions showed a decreasing trend, and the molten steel flow rate had a greater effect on the removal rate of non-metallic inclusions with particle sizes of 50 μm, 75 μm and 100 μm, but had little effect on the removal rate of non-metallic inclusions at 200 μm, 10 μm and 25 μm. With the decrease of the shape factor, the removal rate of non-metallic inclusions showed a decreasing trend, and the removal effect of spherical non-metallic inclusions (shape factor 1) was the best. And when the equivalent diameter of non-metallic inclusions was less than 100 μm, the influence of shape factor on the removal rate of inclusions gradually decreased with the decrease of the equivalent diameter of non-metallic inclusions.

During the continuous casting process of producing high aluminum steel, at the liquid level of the crystallizer, there is a strong interface reaction between the Al element in the molten steel and the mold slag, which leads to serious denaturation of the mold slag and directly endangers the quality of the billet and the smooth operation of continuous casting production. The mold slag layer was dynamically sampled during the continuous casting process of high aluminum steel, and analyzed the changes in the composition and physicochemical properties of the slag layer during the continuous casting process. The reaction behavior between steel slag was studied through thermodynamics and kinetics. The results showed that during the pouring process, the Al₂O₃ content in the slag increased from 1.40% to 24.75%, and the SiO₂ content decreased from 32.26% to 17.69%. The average viscosity of the liquid slag increased from 0.38 Pa·s of the original slag to 0.58 Pa·s, and the average melting temperature increased from 1 112 ℃ of the original slag to 1 189 ℃. The trend of changes in the composition of mold slag conforms to the thermodynamic law of steel slag reaction, and the mass transfer behavior of Al₂O₃ in the liquid slag layer is a limiting link in the interface reaction of steel slag and the degree of liquid slag denaturation.

The technology of feeding steel strip to mold in continuous casting process is an effective method to control the temperature field distribution of molten steel and restrain or reduce the internal defects such as central segregation and central porosity. However, the continuous casting process is invisible at high temperature, so it is impossible to observe and measure the evolution of solidification structure in the mold online. A visual solidification experiment system was set up to investigate the effects of no cold strip, fixed cold strip and vibrational strip feeding on the microstructure evolution of NH₄Cl-70% H₂O solution during solidification. The results show that: firstly, feeding cold strip can significantly reduce the superheat in the center of mold, fuse dendrites, effectively increase the number of free grains, and then increase the volume ratio of equiaxed grains in solidification structure. Secondly, the superheat in the mold center decreases more rapidly after the cold strip is vibrated, and the nuclei formed by fusing are more and finer, and the proportion of equiaxed crystal volume increases further. The proportion of equiaxed crystal volume under three conditions of no cold strip, fixed cold strip and vibrational feeding strip is 26%, 31% and 36%, respectively.

Abnormal large austenite grains are considered a significant cause of transverse cracks in continuous cast billets. Thermal simulation methods combined with in-situ liquid quenching experiments are employed to investigate the growth process of surface austenite grains in nine types of microalloyed steel during the bending-type continuous casting process. Results show that at the straightening starting point, where corner transverse crack defects are likely to occur, the size of abnormally coarse austenite grains increases with rising austenite start growth temperature T_γ, exhibiting an extremum near the peritectic point. As the absolute difference |ΔT| between T_γ and the precipitation temperature of TiN increases monotonically, it indicates that the austenite start growth temperature and the TiN precipitation temperature are key factors influencing abnormal austenite growth. High-temperature tensile tests demonstrate that steels with higher austenite start growth temperatures and larger |ΔT| show greater crack sensitivity, which is consistent with the pattern of abnormal austenite growth.

At present, ingot casting has obvious advantages in producing special steel, manufacturing large-weight ingots and making small-lot castings. However, steel ingots generally have many internal defects, low yield, and high mold loss. How to optimize the ingot structure has become the fundamental method to improve the internal quality of steel ingots. Based on the multiphase solidification model, it took a 5 t cubic ingot as the research object and focuses on the effects of different feeding efficiency ratios, slow cooling process and strong cooling process on the macro segregation behavior of steel ingots. The results showed that increasing the rising capacity ratio or using the strong cooling process has a significant effect on improving the macroscopic segregation, while the slow cooling process has little effect. For this ingot, increasing the rising capacity ratio to more than 11.5% can restrain the positive segregation of the head above the riser line. It can save the design time of the ingot pattern for the designer and at the same time provide a data reference for the actual production on site.

In order to optimize the metallurgical effect of four-strand tundish for extra-large round billet, the flow control devices were explored and optimized through physical simulation experiments firstly, and the influence of U-shaped baffles with different parameters, turbulence inhibitor and dam on flow field was analyzed. The optimal flow control device was selected. And then the flow field, temperature field and concentration field of tundish before and after optimization were simulated and verified by CFD/FLUENT software. Through the physical and numerical simulation, it is found that the flow field and temperature field of optimized tundish are greatly improved. The actual average residence time, stagnation time and peak time of each outlet are extended and the consistency is increased to more than 80%. The proportion of dead zone volume decreases from 32.18% of bare tundish to less than 10%, and the proportion of plug flow volume increases from 9.52% to more than 20%. The maximum temperature difference of tundish decreases from 39.01 K of bare tundish to 16.22 K, and the temperature difference of each outlet decrease from 3.15 K of bare tundish to 0.31 K.

Based on the process parameters of 304 stainless steel production by ϕ350 mm billet mold, a three-dimensional electromagnetic stirring model of billet mold was established. The effects of M-EMS on the flow, temperature distribution and solidification behavior of molten steel in billet mold under different current intensities were studied, and the current intensity for optimal stirring was obtained. The results show that after the application of electromagnetic stirring intensity, the molten steel in the mold undergoes swirling motion, which produces periodic eccentricity with the stirring time. The tangential velocity of the molten steel at the solidification front increases with the increase in current intensity. The swirling molten steel promotes a more uniform temperature distribution of molten steel in the billet mold, the high-temperature area of molten steel expands, and the surface temperature of the billet with M-EMS increases. With appropriate current intensity, the increase in solidified billet shell caused by increasing the heat dissipation of molten steel is greater than the decrease of solidified billet shell caused by swirl scouring of molten steel, resulting in the increase in solidified billet shell thickness, and finally the optimal current intensity parameter of 295 A was selected.

The systematic traceability analysis was conducted on the issue of linear defects (“sliver”) affecting the product qualification rate during the production of IF steel automotive panels using BOF-RH rimming steelmaking/vacuum decarburization/aluminum deoxidation process on a domestic production line. Through sampling, testing, and process data analysis of the entire smelting casting rolling process, the influencing factors and laws of the upstream process conditions of steelmaking and continuous casting on the “sliver” defects on the surface of the strip steel were systematically evaluated, and the process parameters were further improved. The research has shown that the “sliver” defects on the surface of IF cold-rolled strip steel are mainly caused by large-sized Al and O deoxidation products in the molten steel, and slag containing elements such as Si, Ca, Na, Mg, which are captured by the solidification of the casting slab and extended and exposed on the surface during the rolling process of the strip steel. The tapping temperature of the IF steel converter was controlled between 1 600 ℃ and 1 630 ℃, which was shown to reduce the necessity of oxygen blowing and aluminum heating operations due to low RH arrival temperatures, thereby effectively decreasing both total aluminum consumption and the quantity of Al-O inclusions in the steel liquid. The tapping temperature of the IF steel converter was controlled between 1 600 ℃ and 1 630 ℃, which was shown to reduce the necessity of oxygen blowing and aluminum heating operations due to low RH arrival temperatures, thereby effectively decreasing both total aluminum consumption and the quantity of Al-O inclusions in the steel liquid. Measures such as controlling the superheat of the ladle in the continuous casting process to 20-30 ℃ were implemented to alleviate the pressure on the heating value requirements of the upstream process. Additionally, the protection of the casting process was further strengthened, and the water nozzle insertion depth was optimized, which effectively promoted the upward floating of oxides in the steel to the powder slag. “sliver” defect coils are mostly located in the later stage of casting, mainly distributed from the inner arc to the center of the billet. Immersion nozzles are used to ensure active steel tapping, with a frequency of nozzle replacement less than 6 ladles and small steps adjustments to the cast speed. These measures can achieve stable control of the crystallizer flow field and slag/steel interface, effectively reducing the probability of slag rolling in the crystallizer.

In the steel continuous casting process, center porosity and shrinkage cavity are two common quality defects that can significantly affect product performance. Accurately predicting these defects is essential, as it not only avoids destructive testing but also supports technicians in adjusting process parameters, thereby ensuring the smooth progression of subsequent rolling operations. However, due to the nonlinear dynamics, strong coupling characteristics, and multiple external disturbances inherent in the continuous casting process, predicting center porosity and shrinkage cavity levels remains highly challenging. Existing approaches typically rely on single-task learning, which limits their ability to predict multiple defects simultaneously. This constraint not only reduces prediction efficiency but also often results in suboptimal accuracy that fails to meet production requirements. To overcome these limitations, this study proposes a novel defect prediction method based on a multi-task learning framework, which integrates multi-task learning with the TabNet deep learning model designed for tabular data. This approach enables simultaneous prediction of both center porosity and shrinkage cavity levels by leveraging the intrinsic relationship between the two defects, thereby improving both efficiency and predictive accuracy. Extensive experiments using real-world production data from a steel plant demonstrate that the proposed method achieves outstanding performance in industrial defect prediction, with prediction accuracies of 100% for center porosity and 99.9% for shrinkage cavity, and a 53.47% reduction in inference time, fully validating the effectiveness and practical value of the proposed multi-task learning strategy.