Aiming at the problem of coadjustment of blast furnace raw materials and operation parameters, this paper proposes a cooptimization model of blast furnace batching that integrates Random Forest and NSGA-III (Non-dominated Sorting Genetic Algorithm III) algorithm. First, blast furnace field data were collected for a two-year time span, and a predictive model for CO² emissions and blast furnace permeability was constructed using the Random Forest algorithm; taking the goodness of fit (R²), mean square error (MSE) and mean absolute error (MAE) as the evaluation indexes, the R² of the two prediction models obtained reached 0.93 and 0.96 respectively, and the MSE and MAE tended to be close to the zero value. Then, NSGA-III was used to establish the blast furnace batching optimization model to optimally solve the batching scheme and the corresponding blast furnace operating parameters by taking the lowest batching cost, the lowest carbon dioxide emission and the maximum blast furnace permeability as the objective function, and the composition requirement of raw materials and the range limitation of operating parameters as the constraints; finally, the model was validated using the actual on-site data, and the application results showed that the output of the model conformed to the Finally, the results show that the model output meets the composition requirements and obtains a lower-cost dosage scheme than the original dosage ratio; moreover, this scheme corresponds to a blast furnace with less carbon dioxide emission, better blast furnace permeability and less slag. Therefore, the model can provide an effective reference for field operators to optimize blast furnace batching and operation.

The silicon content of hot metal is a key index for the determination of blast furnace status, and accurate prediction of the silicon content of hot metal is crucial for blast furnace ironmaking. First, 10992 sets of blast furnace data obtained from the site of an iron and steel enterprise were preprocessed. Then, 22 important feature parameters related to the silicon content of hot metal were screened by feature engineering. Finally, the hyperparameters of the Gradient Boosting Decision Tree (GBDT) algorithm model were optimized with the help of the Optuna framework, and the Optuna-GBDT model was established to predict the silicon content of hot metal. The experimental results show that compared with the Bayesian algorithm and the traditional stochastic search method, the Optuna framework can achieve better hyperparameter optimization with fewer iterations and smaller errors. The Optuna-GBDT model performs better in predicting the silicon content of hot metal compared with the optimized Random Forest (RF), Decision Tree and AdaBoost models, and the prediction results are basically in line with the actual values, with the mean absolute error (MAE) of 0.0094, the root mean square error (RMSE) of 0.0152, and the coefficient of determination (R2) of 0.975. The experimental results verified the validity and feasibility of establishing the Optuna-GBDT model to predict the silicon content of hot metal, which provides a reliable tool for iron and steel enterprises and helps to optimize the ironmaking process, improve production efficiency and product quality.

The O₂ and N₂ mixing top-blowing method could effectively improve the mixing degree and suppress the temperature increase rate of the molten bath in vanadium extraction converter. In this paper, four kinds of top-blowing lances designed by an extra N₂ flow rate and various Mach numbers have been investigated by a series of water experiments and numerical simulations. On the basis of result, the mixing time was first increased and then decreased with the increase of lance height, and the lance height of 1400 mm obtained the longest mixing time. There were two high-velocity regions generated by impaction of top-blowing jets and stirring of bottom-blowing bubbles. Simultaneously, there were two low-velocity regions formed by the block of furnace wall, and one low-velocity region formed by the local eddy. Comparing with the current top-blowing lance, all three new kinds of top-blowing lances obviously improved the kinetic condition and impaction cavity area of molten bath, which would further be improved with a larger design Mach number. Therefore, an appropriate top-blowing lance had been selected in the industrial application research, which achieved a shorter melting time and a faster vanadium extraction rate, in contrast to the current lance.

In the current study, a novel laboratory experiment and a kinetic calculation were proposed to analyze the modification behavior of alumina inclusions in the molten steel. To obtain the shape and composition of a single Al₂O₃ inclusion at different times during the modification process, confocal scanning laser microscopy experiments were conducted to track the evolution of the Al₂O₃ inclusion particle on the surface of Ca-treated steel. Then, the composition evolution of the Al₂O₃ inclusion particle during the modification process was predicted using a kinetic model. It was assumed the product layer was homogeneous. The diffusion of dissolved [Ca], [Al], and [O] crosses through the inclusion-steel interface was considered. Experimental results agreed well with kinetic calculated results. Meanwhile, the kinetic model was used to analyze the modification behavior of Al₂O₃ inclusions in steel with various influence factors including the [Ca] content in steel, the [Al] content in steel, and the initial size of inclusions.

All 61 sticker breakouts and 183 false sticker breakouts were obtained based on the on-line mould monitoring system during the conventional slab continuous casting. The 16-dimensional temperature characteristics and temperature velocity characteristics of the sticker breakout were extracted. The sticker breakout recognition based on the XGBoost forward iterative model was developed and optimized by the mean square error algorithm. The results show that the prediction probability of the sticker breakout after optimization is in the range of 0.72–1.00. The smallest output value is 0.5 higher than that before optimization. When the threshold is set to 0.65, the optimized XGBoost model can correctly predict all sticker breakouts and has a 99.5% accuracy rate. The XGBoost model has a stronger generalization ability and higher prediction accuracy, which promotes the intelligent production of continuous casting.

3D vision technologies have been widely used in metallurgy industry to measure particle size distribution (PSD) of green pellets on conveyor. However, 3D camera only captures the point clouds of surface pellets, and algorithms measure the surface PSD. To what extent the measured surface PSD can reflect whole PSD is a question that hasn’t been answered yet. In the present work, a simulation method is proposed to analyze the PSD measurement error of green pellets. First, the motion process of green pellets on conveyor is simulated by discrete element method to obtain PSD of whole pellets; then, a transformation method is proposed to generate point clouds of simulated surface pellets, and region growing-based method is adopted to measure the PSD of surface pellets; finally, the PSD measuring error can be obtained by comparing surface PSD and whole PSD of pellets. Error analysis of green pellet size distribution measurement on conveyors is conducted, in aspects of camera location, patch number of point clouds, thickness as well as size distribution of pellet bed. Results illustrate that although the PSD measuring error (up to 12.3%) cannot be neglected when camera is installed above conveyor, it can be effectively reduced by increasing the patch number of captured point clouds (reduced by more than 7.4%) or installing camera near discharge of conveyor (reduced to less than 3.1%).

Accurately predicting the end temperature of molten steel is significant for controlling ladle furnace (LF) refining. This paper proposes an error correction method called EC-CBR based on case-based reasoning (CBR) to reduce errors in the prediction models caused by discrepancies between actual production data and training data. The proposed method combines the incremental learning advantage of CBR with the ability of other models to fit nonlinear relations. First, a prediction model is established, and historical heats similar to the new heat are retrieved by CBR. Then, the model error of the new heat is calculated by employing the errors of similar heats. The prediction result is calculated by subtracting the error from the predicted value. Testing and comparison are conducted on the models (support vector regression, backpropagation neural network, extreme learning machine and mechanism model) and general CBR using actual production data. Results show that the EC-CBR is effective for both data-driven and mechanism models, with an increase of approximately 5% in hit rate within the range of ±5°C for data-driven models and an increase of 21.73% for mechanism model. Moreover, the corrected data-driven models show higher accuracy than the general CBR, further proving the effectiveness of the proposed method.

Gas jet cooling is widely used because the device is simple, oxidation can be prevented, and a uniform cooling capacity can be obtained with thin steel sheets. Because the gas jet cooling ability is affected by the physical properties of the gas such as the mixed gas ratio, a quantitative evaluation of the influence of these factors is very important. However, few studies concerning prediction of the cooling capacity of mixed gas jets in atmospheres with different concentrations have been published.

In this research, the results of experiments and a fluid analysis with an air-helium gas jet in an air atmosphere were compared with the results obtained with Martin’s non-dimensional empirical equations. As the nozzle condition, a single round nozzle with a tapered shape was examined. The helium concentrations with respect to air were 0, 20, 50, and 100 vol%, and the pressure conditions were 3 and 5 kPa.

Compared with the experimental results, Martin’s equations overestimated the improvement of cooling performance with increasing helium concentrations. In the analysis in the present study, it was found that mixing with ambient air increased as the helium concentration decreased.

The trend of divergence between the experimental and predicted cooling capacity was clearly presented in this research. The results of this study will make it possible to improve the accuracy of predictions of the cooling capacity of impinging gas jets with different concentrations of the atmosphere and the gas jet.

Austenitic Stainless Steel (ASS) and Duplex Stainless Steel (DSS) are joined to optimize the Resistance Spot Welding (RSW) process parameters and to predict the parametric influence on the response of Tensile Shear Fracture Load (TSFL). The Response Surface Methodology (RSM) is an optimization technique is used in this research to develop the satisfactory quadratic mathematical model and to predict the response. The optimal parameters and their levels are found and reported as follows: welding current = 9 kA, welding time = 0.18 seconds and electrode tip radius = 3 mm. The actual and predicted values of TSFL for the optimized parameters are 17.6 kN and 17.9 kN respectively. The developed quadratic model is efficiently predicted the response with an average error percentage of 2.18. The significant and insignificant terms in the models has been identified by 95% confidence level using ‘p’ test. The insignificant terms are removed from the model and the ANOVA table is formulated only with the significant terms. Significance or effect of each term in the ANOVA table is identified by calculating the percentage of contribution and noticed that welding current has the highest significance (46%) on TSFL. The macroscopic examination confirmed that the larger nugget is observed during the maximum welding current due to the high heat generation. Also, the variation in TSFL against the process parameters are observed as same as nugget length, because, TSFL and nugget length are perfectly correlated.

A thermomechanical simulator Gleeble 3800 was used to simulate the thermal cycles experienced by various heat-affected zones (HAZ) during the welding process. The influence of peak temperature (T_p, 500°C–1320°C) on the hardness, microstructure, precipitates, and properties of complex steel 780FB with microalloyed elements Ti, Nb, and V was systematically studied. The contributions of dislocation strengthening, precipitation strengthening, fine grain strengthening, and phase transformation strengthening increments to strength changes of samples after different thermal cycles were quantified, and the calculated results were found to be consistent with the experimental data. Compared with 780FB, there was little change in microstructure and properties when T_p was 500°C. When T_p was 650°C, the increase in VC density from 43/µm² to 288/µm² caused the enhancement of hardness and strength. The precipitation strengthening increment (49.84 MPa) played a dominant role in strength improvement. As partial bainite in the microstructure of 780FB transformed into ferrite at T_p of 800°C, the weakening of phase transformation strengthening (−57.5 MPa) became the main factor in strength change. The softening and strength reduction further increased when T_p was up to 980°C, as 780FB completely recrystallized and transformed into ferrite and MA islands. The phase transformation strengthening further reduced by 74.75 MPa. When T_p was 1320°C, the VC density decreased from 43/µm² to 13/µm², and the (Ti,Nb)C density decreased from 34/µm² to 14/µm², leading to severe grain growth (2.24 µm to 19.89 µm) and bainite transformation. The decrease in precipitation strengthening (−26.86 MPa) and fine grain strengthening (−87.91 MPa) counteracted with the increase in phase transformation strengthening (51.62 MPa), resulting a slight decrease in hardness and strength.

The agglomeration of droplets dispersed in an immiscible liquid is often an issue in metallurgical processes. To determine the effect of wettability on droplet agglomeration, we conducted (1) wettability and (2) agglomeration experiments using immiscible liquid paraffin and glycerin aqueous solution. In the wettability experiment, a droplet of one liquid was settled on the other liquid and its shape was observed. The glycerin droplet was wrapped by liquid paraffin, while the paraffin droplet spread on the surface of the glycerin solution. Therefore, liquid paraffin wetted the glycerin droplet, while glycerin solution did not wet the paraffin droplet. In the agglomeration experiment, after the droplet settled or floated in the other liquid layer to arrive at the boundary between the two liquid layers, we measured the time required for droplet agglomeration in its liquid layer. The agglomeration of paraffin droplets from the glycerin solution was faster than that of glycerin droplets from liquid paraffin, indicating that non-wettability of droplets accelerated agglomeration.