Vacuum degassing (VD) refining, characterized by strong reactions between slag and steel within a vacuum atmosphere, serves functions such as degassing, desulfurization, and the control of non-metallic inclusions. This method is of great importance for producing high-quality clean steel. This work presented an overview of the development history and industrial applications of VD refining technology, described in detail the research and production of VD refining in degassing, desulfurization and inclusion control, and gave some suggestions for this process at the end. In VD refining, researches on vacuum degassing (including O, N, H) mainly focused on optimizing process parameters and enhancing equipment performance. It could control [H] in steel within (1-2)×10^-6 and [N] within (40-60)×10^-6, and denitrogen in VD had some advantages. Regarding desulfurization, VD refining benefitted from favorable thermodynamic and kinetic conditions, enabling control of [S] in steel at relatively low levels. In terms of inclusion control, VD process could decrease the quantity of inclusions effectively and modify inclusions through optimizing slag compositions and process parameters. The related research results had certain guiding significance for the research and industrial production of VD refining, offering favorable references for reducing impurity element content, controlling non-metallic inclusions, as well as improving steel cleanliness.

According to the actual production data of 120 t converter of Zhongtian Iron and Steel Group Co., Ltd., a prediction model of phosphorus content in molten steel at the end of converter blowing based on Stacking algorithm was established. The main factors influencing dephosphorization were determined through the thermodynamic analysis of dephosphorization, and then the input variables of the model were determined. After the data preprocessing was completed, six machine learning algorithms (RF, ET, XGBoost, LightGBM, CatBoost, and NN) were used to establish the models, and then the prediction results of these six models were used for Stacking ensemble modeling by multiple linear regression algorithm. By comparing the prediction results of these seven models, it could be concluded that the Stacking ensemble model had the best prediction effect, and the hit rates of the predicted endpoint phosphorus mass fraction were 90.59 % and 97.56 % when the error interval was ±0.004 % and ±0.005 %, respectively.

In order to reduce converter alloy cost, the converter alloy model was developed using linear programming as the core algorithm. The model through designing carbon redundancy and phosphorus redundancy, determining main parameter such as alloy yield and carbon increment of LF treating process and allocating alloy amount of converter based on preferentially using lower price alloy. The cost of alloy was reduced under the premise of meeting with qualified component of liquid steel, and the economic benefits was 3.27 yuan/t.

Converter smelting process control and end-point control are key to affecting steelmaking production efficiency, product quality, production cost and other indicators. However, converter smelting control of most steel companies mainly relies on manual experience. Zhongtian Iron and Steel Group Co., Ltd. and Hunan RAMON Technology Co., Ltd. took the lead in researching and developing the non-sublance converter intelligent steelmaking system based on sonar slagging, flame analysis and flue gas analysis, realizing the whole process of “one-click” steelmaking without manual intervention under the circumstance of non-sublance. The system has achieved good results in the automation control of the smelting process. At the first turning down, the hit rate of carbon mass fraction (≤0.15 %) prediction within the error range of ±0.02 % was 90 %, the hit rate of temperature prediction within the error range of  ±15 ℃ was 87 %,and the hit rate of phosphorus mass fraction prediction within the error range of ±0.005 % was 90 %. In addition, the accuracy rate of splash recognition reached 99 %. These achievements play a positive role in promoting the intelligent steelmaking development of iron and steel enterprises.

Under industrial conditions, the Mg treatment test of Nb-Ti micro-alloyed steel was carried out. The evolution of nonmetallic inclusions in micro-alloyed steel treated with Mg was analyzed. The results show that, Al₂O₃ inclusion is the main product of aluminum deoxidation in micro-alloyed steel. After Ca treatment, Al₂O₃ inclusions are reduced and transformed into CaO-Al₂O₃ inclusions, and part of MnS inclusions are transformed into CaS inclusions. After Ca treatment is changed to Mg treatment in the refining stage, Al₂O₃ inclusions of aluminum deoxidation product can be denatalized to MgO-Al₂O₃ inclusions. Before Mg treatment, the inclusions  in LF are mainly Al₂O₃ inclusions. After Mg treatment and Ti alloying, the main inclusions in steel become MgO-Al₂O₃ inclusions, MgO inclusions and TiN inclusions. The final inclusions in the billet are MgO-Al₂O₃ coated with CaS/MnS and (Nb, Ti) N complex inclusions. The size of inclusions treated with Ca and Mg was compared, and the size of inclusions mainly distributed within 5 μm. The proportion of inclusions within 2 μm formed by Mg treatment is higher than that by Ca treatment, which indicates that magnesium treatment can fine inclusions better. The formation of a large number of dispersed small inclusions is helpful to improve the comprehensive mechanical properties of steel.

To investigate the controlling effect of magnesia inclusions on the sulfide morphology in sulfur-containing gear steel, this study performed magnesium treatment on 40Cr steel bearing sulfur. Combining thermodynamic analysis, the evolution behavior of inclusions during smelting and rolling processes was analyzed. The research findings revealed that after introducing metallic magnesium into the molten steel, the Al₂O₃ inclusions in the steel could be transformed into dispersed MgAl₂O₄ inclusions. During the solidification process, the transformed MgAl₂O₄ inclusions could serve as the precipitate core of sulfide manganese, forming multiphase inclusions. During the rolling process, the core MgAl₂O₄ further played a skeletal role, inhibiting the extension of surrounding sulfide manganese. This effectively reduced the aspect ratio of sulfide manganese and is expected to improve the cutting and mechanical performance of gear steel. The research results also indicated that to fully utilize the skeletal role of MgAl₂O₄, it was necessary to ensure a sufficiently high dissolved magnesium content in the steel for the full transformation of Al₂O₃, and to promote suitable particle size and quantity of MgAl₂O₄ to serve as an effective core for controlling sulfide manganese precipitation and deformation.

Aiming at the problem of the over dependence of LF refining operation on manual experience，based on metallurgical mechanism，temperature model, alloy model, argon blowing model, and slag making model were developed on the 120 t LF of the Long Product Division of Maanshan Iron and Steel Co., Ltd.  And a control model was established based on the mutual unity of molten steel temperature，composition and slag.The big data technology and self-learning function were used to optimize the control model，the collaborative integration of various models and the whole process intelligent control of LF was realized，and good application results were achieved．The proportion of LF automatic control reached 80 %，the hit rate of endpoint target temperature ±5 ℃ was more than 95 %，the hit rate of narrow range of endpoint component (w(Si)±0.02 %、w(Mn)±0.02 %、w(S)±0.001 %、w(Al_s)±0.005 %) was above 97 %，the power consumption of LF refining was reduced by about 4 kWh/t，the refining treatment time was reduced by about 5 min，and the production efficiency and molten steel quality were improved．It played an important role in reducing cost and increasing efficiency during steelmaking process．

At the end of ladle teeming, the molten steel will produce confluence vortex under the action of the Coriolis force and its own initial state, which becomes a bottleneck for controlling ladle slagging and enhancing molten steel yield. Taking a 150 t single-nozzle ladle in a steel plant as a prototype, a physical model of the ladle with a similarity ratio of 1∶3.5 was established based on the principle of similarity, and the influence degree and mechanism of the nozzle eccentricity on the confluence vortex at the end of teeming were first explored, then the effects of the ladle standing time, the anti-vortex device and the bottom stirring process, and other flow-control measures on the confluence vortex in the ladle with an eccentricity of 0.83 were investigated. The results show that the formation of asymmetric flow field in the ladle can inhibit the development of vortex. When the standing time of the ladle is less than 45 min, the change of the formation height and penetration height of vortexis are not obvious with the extended standing time, but when the standing time is more than 45 min, the formation height of vortex increases significantly and the penetration height decreases obviously. In addition, at an eccentricity of 0.83, neither the installation of anti-vortex device nor the addition of bottom blowing stirring operation can suppress the vortex, but closing the nozzle briefly during vortex initiation can delay the subsequent development of the vortex.

The production of ultra-low carbon automobile plate DC04 steel by scrap management-EAF-RH vacuum treatmentslab continuous casting process was introduced. Aiming at the problem that C, N and residual elements were difficult to control in short-process smelting process, the reaction principle and production process were emphatically analyzed. The targeted process improvement schemes and technical measures were adopted from scrap management, EAF, vacuum refining to continuous casting, including establishing digital scrap management mode, optimizing EAF endpoint control and scrap preheating process, accurate control of RH vacuum degree and optimization of slag system, revising raw materials standards for continuous casting and optimizing process parameters.The mass fraction of carbon and nitrogen elements in steel was reduced to below 0.002 5 % and 0.003 0 % respectively, and the mass fraction of residual elements of molybdenum, nickel, copper and boron in steel were reduced to below 0.02 %, 0.05 %, 0.08 % and 0.000 6 % respectively, which met the standard requirements of DC04 steel for ultra-low carbon automobile plates and improved the product quality.

The defects caused by high hydrogen content were obtained in the flaw detection of ultra large scale continuous casting round billets for wind power tower with a diameter not less than 800 mm. The variation law of hydrogen content in steel throughout the entire process was studied by detecting the moisture content of raw and auxiliary materials and using the Hydris hydrogen analyzer. The results showed that moisture in raw and auxiliary materials, a damp furnace gas atmosphere, and incomplete decomposition of refractory materials are the main sources of hydrogen content in castings. To reduce the hydrogen content of the casting billet, the following process improvement measures were proposed. After 8 minutes of blowing in the converter, raw materials should not be added or added baked one. In the LF refining process, less raw and auxiliary materials are input. Before entering the VD process, a small amount of slag should be controlled. After breaking the vacuum, low-speed wire feeding should be carried out to control the number and type of wire feeding within a small range. Bake the continuous casting tundish at low heat for 60 minutes and clean the air holes in the tundish regularly. After process optimization, the hydrogen mass fraction of molten steel was not exceed 3.6×10^-6 when arriving at the VD station, and not exceed 0.77×10^-6 when leaving the VD station. In the molten steel of the continuous casting tundish, the hydrogen mass fraction of the first furnace was not exceed 2.0×10^-6, and not exceed 1.6×10-⁶ in the second furnace, and not exceed 1.1×10^-6 in the continuous casting furnace. Therefore, the hydrogen content of the continuous casting round billets for ultra large scale wind power tower was reduced successfully, ensuring the quality of the finished forgings. 

Hangang New Area is an integrated relocation project of Hangang East District. Limited by the mountainous terrain and land area, the project adopts compact layout, the nearest distance from blast furnace to converter is 450 m, and the area of tons of steel is only 0.3 m². The iron and steel interface adopts the “one tank technology” mode using vehicle for the transportation of hot metal, and the hot metal tank can be pulled to steelmaking production in 10 minutes after receiving hot metal. And the use of digital, Internet of Things and other new technologies to assist, after the improvement and optimization since the operation, the new district hot metal tank turnover rate reached 7.0 times/day, the steelmaking hot metal temperature of 1 440 ℃, the proportion of “one tank technology” 99 %, which respectively increased 4.0 times/day, 98 ℃, 39 percentage points higher than before the relocation.Hot metal temperature decreased to 82 ℃, 81 ℃ lower than that before the relocation. The significant reduction in interface heat loss has created favorable conditions for increasing steel and reducing consumption in the steelmaking process, which is in line with the national concept of energy conservation and emission reduction, and has certain reference significance for the “dual carbon” strategy of the industry.