From a cyber-physical system (CPS) perspective, this paper reviews the current status and evolutionary trends of continuous casting process control systems in steelmaking shops (hereinafter referred to as Level-2 systems), with emphasis on their structural functions within hierarchical automation architectures, key models, and data-utilization paradigms. A synthesis of representative literature and engineering practices indicates that Level-2 systems serve both as process coordinators and data hubs. The main technical routes include mechanism-based process optimization, data-driven quality evaluation, and the integrated application of "process rule base-mechanism-based models-data-driven models". Solidification heat transfer, secondary cooling, soft reduction, and defect prediction models have established relatively mature engineering frameworks; however, their performance is still widely constrained by data fragmentation, weakened closed-loop verification due to insufficient execution feedback and delayed return of quality results, and limited cross-process coordination capability. To address these bottlenecks, a CPS-oriented layered architecture and implementation guidelines are proposed, together with a systematic pathway in terms of unified data modeling, full-lifecycle model management, and the machine-readable construction of the process rule base. The development trends suggest that Level-2 systems will take the triadic integration of "process rule base-mechanism-based models-data-driven models" as the core and, supported by platform-based and service-oriented architectures, gradually evolve from stand-alone automation tools into intelligent decision-making platforms enabling whole-process collaborative optimization across steelmaking-continuous casting-rolling.
The rapid growth in global liquefied natural gas (LNG) demand has posed mounting cost challenges for conventional storage tank materials. High manganese steel, which combines excellent low-temperature properties with a distinct cost advantage, has become a promising alternative. However, longitudinal cracking defects in continuous casting have hindered its large-scale application. In this study, the solidification characteristics of high manganese steel slabs were systematically investigated through thermodynamic simulation and experimental testing. The results show that the experimental steel possesses a wide solid-liquid phase region. The as-cast macrostructure consists predominantly of 64% columnar grains and 30.66% equiaxed grains, a grain-zone configuration that tends to induce stress concentration during solidification shrinkage. Microstructural analysis further reveals that the dendrite arm spacing in the columnar zone increases with solidification depth, and carbides enriched in Mn and Cr precipitate along grain boundaries, indicating pronounced elemental segregation. Thermophysical measurements demonstrate that the steel exhibits low thermal conductivity but a high coefficient of thermal expansion, which results in considerable thermal stress during solidification.
The content and distribution of high-temperature ferrite (δ-ferrite) have a crucial impact on the mechanical properties and corrosion resistance of austenitic stainless steels, and are key factors for realizing precise performance regulation of the materials. Taking 304 austenitic stainless steel as the research object, this paper systematically investigates the effect of pulsed current treatment with different intensities applied during the nucleation stage of δ-ferrite on the microstructural evolution of 304 stainless steel. Various characterization methods, including optical microscopy, scanning electron microscopy, electron backscatter diffraction and transmission electron microscopy, were employed to analyze the changes in grain distribution, as well as the morphology, content and distribution of ferrite in the treated stainless steel. The research results show that pulsed current treatment can significantly refine the solidification dendrite structure and effectively promote the nucleation and precipitation of δ-ferrite, with its content increasing from 1.21% to 1.73% as the current density rises. Pulsed current reduces the nucleation barrier by altering the interfacial free energy between ferrite nuclei and the surrounding melt, thereby facilitating the formation of ferrite. This study provides a theoretical basis and experimental foundation for regulating the microstructure and improving the properties of stainless steels by means of pulsed current.
Marine engineering steels (such as FH420) are essential materials for offshore infrastructure; however, challenges persist in achieving uniform microstructure and mechanical properties across the thickness of ultra-thick plates. Conventional uniform temperature rolling (UTR) often leads to insufficient strain and delayed recrystallization in the core, whereas differential temperature rolling (DTR) creates a gradient temperature field (with lower surface temperatures and higher core temperatures) by surface cooling, significantly enhancing deformation penetrability and microstructural uniformity. This study focuses on FH420 steel, utilizing data from hot tensile experiments and thermophysical parameters calculated by JMatPro software to establish a thermo-mechanical coupled three-dimensional finite element model. A comparative analysis was conducted on the temperature evolution, stress, and strain fields under both UTR and DTR processes. Industrial rolling experiments validated the model′s accuracy, with the simulation and measured surface temperature deviation remaining below 5%. The mechanical properties of the plate core under DTR were superior to those under UTR in all aspects. The results indicate that the rapid cooling process can generate a temperature difference of up to 300 ℃, enhancing the equivalent strain in the core. The optimization of reduction rate further improved deformation uniformity. This study quantitatively reveals the coupling mechanism between the temperature gradient and stress field, providing theoretical support for the optimization of ultra-thick marine steel rolling processes.
This study systematically reveals the regulation mechanism of electromagnetic induction heating on the slag/metal interface behavior and molten steel flow in the tundish by establishing an electromagnetic-flow field-multiphase coupling model. A decoupled model of Joule heat supply and electromagnetic force transport was constructed based on the VOF method. Numerical simulation results under three power conditions (1 600, 1 200, and 800 kW) show that electromagnetic force dominates the flow field structure and interface stability. At 1 600 kW, the peak velocity at the centerline of the slag/metal interface reaches 0.185 m/s, which is 97.2% higher than that at 800 kW; Joule heat only reduces the viscosity of molten steel through temperature rise, with a maximum temperature rise of 37.04 K, and has a weak effect on the flow field. 1 600 kW is the critical power for slag eye formation, under which the slag layer on the wall of the distribution chamber is completely lost within 36 s, and the slag eye lasts for 8 s; the 1 200 kW condition achieves the optimal balance between flow enhancement and interface stability. The coupling of electromagnetic force and Joule heat significantly improves temperature uniformity, with an average temperature gradient increase of 6.09 K, and the interface fluctuation characteristics are highly consistent with those under pure electromagnetic force conditions. Based on the above research conclusions, a zonal optimization scheme based on power-spatial mapping is proposed, which provides a theoretical basis for the control of slag entrainment risk and efficient thermal management in the continuous casting process.
In the steelmaking process, the liquid level fluctuation of continuous casting molds exerts a crucial influence on slab quality and production safety. Excessive mold level fluctuation beyond the preset threshold will induce slag inclusions, slab surface defects and even steel breakout accidents, which severely degrade the operational stability and production efficiency of the continuous casting process. To overcome the limitations of traditional Proportion Integration Differentiation (PID) control in mold level fluctuation modeling, this paper proposes a hybrid deep learning prediction framework integrating wavelet transform and Bidirectional Gated Recurrent Unit (BiGRU). The proposed framework adopts wavelet transform to conduct multi-scale decomposition of mold level time-series data and extract discriminative multi-scale time-frequency features; meanwhile, the BiGRU module is employed to capture the bidirectional temporal dependency of mold level variation. A high-precision mold level fluctuation prediction model is established by incorporating key technological parameters, including stopper rod opening, casting speed and slab width. Experimental results verify that the proposed hybrid model outperforms the conventional GRU model and other baseline comparison methods in mold level fluctuation prediction. The model achieves remarkable improvements in prediction accuracy and operational robustness, which provides a feasible technical solution for intelligent optimal control of continuous casting processes and possesses promising prospects for industrial engineering applications.
Reduction efficiency directly quantifies the capacity of surface reduction to be transmitted to the mushy zone during the soft reduction process of continuous casting, and is a key factor governing the control of centerline segregation and centerline porosity. To quantitatively investigate the effect of reduction amount on reduction efficiency from a theoretical perspective, this work developed a thermo-mechanically coupled numerical model of the soft reduction behavior of continuous casting slabs based on the incremental thermo-elastoplastic theory and proposed a method for calculating reduction efficiency through the area-integral ratio of equivalent plastic strain. Five solid fractions, namely 0.1, 0.3, 0.5, 0.7, and 0.9, were considered, and the evolution of reduction efficiency with reduction amount and solid fraction was systematically analyzed over a total reduction amount of 9 mm. The results show that the stress and strain within the slab exhibit a distinct inverse distribution: the solidified shell is subjected to relatively high stress but undergoes limited plastic strain, whereas the central mushy zone experiences considerable plastic deformation despite its low stress level. The reduction efficiency first decreases and then tends to stabilize with increasing reduction amount. The results obtained by the proposed method are highly consistent with those of previous studies, effectively verifying the rationality and reliability of the proposed calculation method.
In order to improve the center carbon segregation of high carbon steel SWRH82B from 160 mm×160 mm section billet produced by a steel plant, a full-process three-dimensional model of flow, solidification, heat transfer, component transport coupled with electromagnetic stirring in the continuous casting process was established. The accuracy of the model was verified by means of measurement and experiment. Based on the established verified three-dimensional model of the whole process, the influence of mold and final electromagnetic stirring on the center carbon segregation of billet was studied. The method of combining numerical simulation and experiment is used to optimize the optimal stirring process parameters under the actual working conditions of the continuous casting machine and carry out industrial application. Combined with the actual situation of the electromagnetic stirrer of the casting machine, the low frequency alternating mode of Lake Shore Gauss meter was adopted. Considering the shielding effect of the copper tube and the shell on the magnetic field, the optimum stirring frequencies of the mold and final electromagnetic stirrer were determined to be 4 Hz and 7 Hz, respectively. The segregation simulation results show that when the mold electromagnetic stirring frequency is 4 Hz and the electromagnetic stirring current increases from 200 A to 500 A, the center carbon segregation index decreases from 1.11 to 1.08. When the final electromagnetic stirring frequency is 7 Hz and the stirring current increases from 300 A to 600 A, the center carbon segregation index decreases from 1.16 to 1.07, respectively. The optimum parameters of mold and final electromagnetic stirring are 400 A/4 Hz and 500 A/7 Hz, respectively. The corresponding center carbon segregation index of slab is 1.08 calculated by using this stirring parameter model. After adopting the optimized process, the center equiaxed crystal rate of the slab increased from 38.4% to 47.2%, the center carbon segregation index decreased from 1.18 to 1.08, and the core network cementite of the corresponding 14 mm hot rolled wire rod decreased from grade 3-4 to grade 0-1, and the drawing fracture phenomenon was greatly improved.
The secondary cooling process serves as a core control procedure in slab continuous casting production. Improper secondary cooling water flow easily induces internal defects such as central porosity and surface longitudinal cracks in cast slabs. Given the differences in chemical composition and solidification characteristics among various steel grades, their requirements for secondary cooling water supply vary significantly, which necessitates targeted adaptive optimization. The conventional secondary cooling water control mode suffers from empirical reliance in water adjustment and delayed parameter optimization. Moreover, it lacks a unified digital platform for centralized management and dynamic invocation of secondary cooling optimization parameters and slab temperature field data for multiple steel grades. This study develops an integrated digital software for heat transfer simulation and secondary cooling optimization in slab continuous casting, adopting the particle swarm optimization (PSO) algorithm for parameter optimization and Qt/C++ for program development. The software integrates intelligent secondary cooling water optimization, online slab temperature field simulation, and digital management of process and temperature field data. Equipped with a database and heat transfer simulation optimization module, the system supports parameter input, rapid retrieval of multi-steel-grade secondary cooling processes, and visualized monitoring of slab solidification temperature field, featuring favorable portability and scalability. Taking AH32 ship plate steel as the research object, relevant steel grade parameters were imported into the software database to realize systematic management of slab temperature field data under different casting speeds. The zone-based secondary cooling water flow was adaptively optimized according to the temperature field characteristics under varied casting speeds. With the slab temperature field quality fully guaranteed, the total secondary cooling water flow of AH32 steel at a casting speed of 1.05 m/min was reduced from 2 256.90 L/min to 2 078.75 L/min, effectively cutting the secondary cooling water supply cost. Built on the Qt/C++ framework, the software supports cross-platform packaging and mobile terminal adaptation, possessing prominent industrial practicability and engineering promotion value.
The composition and distribution of inclusions in Al-killed Ca-treated steels during continuous casting and rolling processes have a significant influence on steel properties. However, the evolution mechanism of inclusions during solidification, cooling, and heat treatment processes was still unclear. In this study, systematic sampling was conducted on tundish, continuous casting slab, and rolled plate samples of typical Al-killed Ca-treated steels. Automated inclusion analysis techniques and FactSage thermodynamic calculations were employed to analyze the composition, quantity, size, and distribution characteristics of inclusions. It was indicated that during solidification and cooling, inclusions transform from CaO-Al2O3 with high CaO content to composite inclusions of CaO-Al2O3 and MgO-Al2O3-CaS with lower CaO content. Inclusions at the surface layer of the slab exhibit higher CaO and lower CaS content, while the opposite is true for inclusions in the interior. After the heat treatment at 1 200 ℃, the composition distribution of inclusions in the rolled plate becomes more uniform. Thermodynamic calculations confirmed that temperature change was the main driving force for the transformation of inclusions. This study revealed the evolution mechanism of CaO-Al2O3-based and MgO-Al2O3-CaS composite inclusions throughout the continuous casting-rolling process of Al-killed Ca-treated steel. It provided insights for improving the uniformity control of inclusions in continuous casting slabs and rolled plates.
To improve the internal quality of DC01 slabs and reduce center segregation of the cast billets, the solidification process of the cast billets was simulated using ProCAST software. The effects of continuous casting parameters on the solidification endpoint, solidification structure distribution and equiaxed grain ratio were studied to further enhance the internal quality of the cast billets. Key process parameters were obtained via pin shooting tests and on-site investigations, and the established heat transfer model was calibrated based on these parameters. The control variable method was adopted to study the influences of different processes on the internal quality and solidification process of the slabs. The simulation results indicate that under the current process parameters, casting speed has the greatest impact on the solidification endpoint. A 7.1% change in casting speed has a slightly greater effect on the solidification endpoint than a 20% change in water flow rate, and a greater effect than a 33% change in superheat. Increasing the casting speed, reducing the water flow rate and lowering the superheat degree are conducive to expanding the equiaxed grain zone of the cast billets and refining the grain structure to a certain extent. It is concluded that the macro segregation of DC01 steel can be reduced by appropriately increasing the casting speed, decreasing the secondary cooling water flow rate and lowering the superheat degree of molten steel. Moreover, the theoretically optimal production conditions for improving internal quality are obtained as follows: a cast speed of 1.5 m/min, a water flow rate of 0.78 L/kg and a superheat of 20 ℃.
Through the analysis of the morphological characteristics of the leaked slab shell, the temperature curves of thermocouples on the mold copper plate, the flow field in the mold, the thickness of the liquid slag layer of the mold powder, and the nodules on the inner wall of the submerged entry nozzle (SEN), it is revealed that the nozzle nodules are TiO2+Al2O3 composite inclusions. The main cause of steel leakage during continuous casting of high-titanium steel is that the local nodules on the inner wall of the nozzle fall off and are captured by the slab shell, which eventually leads to the cracking of the slab shell and the overflow of molten steel at the mold outlet under the action of molten steel static pressure. By adopting measures such as using a tundish with argon blowing for both the stopper and the upper nozzle, improving the cleanliness of molten steel and promoting the full floatation of inclusions, strictly controlling the flow field in the mold and the modification of mold powder, and strengthening protective casting to prevent secondary oxidation, the problems of inner wall nodulation of the submerged entry nozzle and steel leakage in high-titanium steel have been effectively solved, ensuring the stability and continuity of continuous casting production.
To address the problems of slab corner transverse cracks occurring during the high-speed continuous casting of thin slabs and subsequent edge cracks after rolling, the formation mechanism, macroscopic structural characteristics and microscopic structural features of transverse crack defects in high-speed thin slab casting were analyzed via metallographic microscopy, scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS). The influencing factors including molten steel composition, mold flux, mold flow field and mold vibration parameters were systematically investigated. The results show that molten steel with carbon mass fraction w([C])>0.05% is more susceptible to transverse crack formation, and mold flux with low basicity and high viscosity aggravates the occurrence of such defects. Adopting vibration parameters with high frequency and low amplitude, and optimizing the submerged entry nozzle (SEN) structure of the tundish to improve the flow field distribution in the mold can effectively control transverse cracks. After the implementation of the above targeted control measures, the defect rejection rate of slab edge breakage and edge cracks induced by transverse cracks decreased from 1.46% to below 0.09%, achieving a significant improvement in edge crack defects. Meanwhile, the proportion of steel breakout accidents caused by transverse cracks dropped from 44.93% to zero, which greatly improved the production stability of thin slab continuous casting under high casting speed conditions.
Abnormal perforation repeatedly occurred above the slag line of submerged entry nozzles (SENs) during casting on a thin slab continuous caster at a plant. Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) were adopted to characterize the microstructure and compositional evolution of refractories at the perforation region. Obvious color stratification was observed on the cross-section of the SEN, which was divided into an outer corroded layer and an inner layer contacting molten steel. The corroded layer contained prominent characteristic elements of mold flux, including Na, Mg and Ca. Combined with practical casting conditions, the root cause of SEN abnormal perforation was identified. The mold flux was fed into the mold in a circular diffusion pattern through the nozzle of the slag feeder, and part of the mold flux adhered to the region above the SEN slag line. At high temperatures, Na2CO3 in the mold flux powder reacted with Al2O3 in the SEN refractory, deteriorating the corrosion resistance above the slag line and eventually inducing perforation. Countermeasures including reducing the SEN immersion depth (from 121-221 mm to 101-201 mm), optimizing the dimension of the slag feeder nozzle, and spraying ZrO2 coating on the refractory surface above the slag line were implemented. As a result, the residual thickness of the SEN increased from 13.3 mm to 21.5 mm, and the abnormal perforation above the slag line was successfully eliminated.