Search Results (10)

In steel production, the blast furnace is a critical element. In this process, precisely controlling the temperature of the molten iron is indispensable for attaining efficient operations and high-grade products. This temperature is often indirectly reflected by the silicon content in the hot metal. However, due to the dynamic nature and inherent delays of the ironmaking process, real-time prediction of silicon content remains a significant challenge, and traditional methods often suffer from insufficient prediction accuracy. This study presents a novel Multi-Scale Fusion Convolutional Neural Network (MSF-CNN) to accurately predict the silicon content during the blast furnace smelting process, addressing the limitations of existing data-driven approaches. The proposed MSF-CNN model extracts temporal features at two distinct scales. The first scale utilizes a Convolutional Block Attention Module, which captures local temporal dependencies by focusing on the most relevant features across adjacent time steps. The second scale employs a Multi-Head Self-Attention mechanism to model long-term temporal dependencies, overcoming the inherent delay issues in the blast furnace process. By combining these two scales, the model effectively captures both short-term and long-term temporal dependencies, thereby enhancing prediction accuracy and real-time applicability. Validation using real blast furnace data demonstrates that MSF-CNN outperforms recurrent neural network models such as Long Short-Term Memory (LSTM) and the Gated Recurrent Unit (GRU). Compared with LSTM and the GRU, MSF-CNN reduces the Root Mean Square Error (RMSE) by approximately 22% and 21%, respectively, and improves the Hit Rate (HR) by over 3.5% and 4%, highlighting its superiority in capturing complex temporal dependencies. These results indicate that the MSF-CNN adapts better to the blast furnace’s dynamic variations and inherent delays, achieving significant improvements in prediction precision and robustness compared to state-of-the-art recurrent models. Full article

Silicon content plays an important role in determining the operational efficiency of blast furnaces (BFs) and their downstream processes in integrated steelmaking; however, existing sampling methods and first-principles models are somewhat limited in their capability and flexibility. Current data-based prediction models primarily rely on a limited set of manually selected furnace parameters. Additionally, different BFs present a diverse set of operating parameters and state variables that are known to directly influence the hot metal’s silicon content, such as fuel injection, blast temperature, and raw material charge composition, among other process variables that have their own impacts. The expansiveness of the parameter set adds complexity to parameter selection and processing. This highlights the need for a comprehensive methodology to integrate and select from all relevant parameters for accurate silicon content prediction. Providing accurate silicon content predictions would enable operators to adjust furnace conditions dynamically, improving safety and reducing economic risk. To address these issues, a two-stage approach is proposed. First, a generalized data processing scheme is proposed to accommodate diverse furnace parameters. Second, a robust modeling pipeline is used to establish a machine learning (ML) model capable of predicting hot metal silicon content with reasonable accuracy. The method employed herein predicted the average Si content of the upcoming furnace cast with an accuracy of 91% among 200 target predictions for a specific furnace provisioned by the XGBoost model. This prediction is achieved using only the past shift’s operating conditions, which should be available in real time. This performance provides a strong baseline for the modeling approach with potential for further improvement through provision of real-time features. Full article

For many complex industrial applications, traditional attribute reduction algorithms are often inefficient in obtaining optimal reducts that align with mechanistic analyses and practical production requirements. To solve this problem, we propose a recursive attribute reduction algorithm that calculates the optimal reduct. First, we present the notion of priority sequence to describe the background meaning of attributes and evaluate the optimal reduct. Next, we define a necessary element set to identify the “individually necessary” characteristics of the attributes. On this basis, a recursive algorithm is proposed to calculate the optimal reduct. Its boundary logic is guided by the conflict between the necessary element set and the core attribute set. The experiments demonstrate the proposed algorithm’s uniqueness and its ability to enhance the prediction accuracy of the hot metal silicon content in blast furnaces. Full article

Super-reduced phases (SRPs), such as silicon carbide (SiC) and metal silicides, have increasingly been reported in various geological environments. However, their origin remains controversial. SRP inclusions (e.g., metal silicides and metallic silicon (Si⁰)) within SiC are commonly believed to indicate a natural origin. Here, we identified an unusual SRP assemblage (SiC, (Fe,Ni)Si₂, and Si⁰) in situ in an H5-type Jingshan ordinary chondrite. Simultaneously, our analysis showed that the SiC abrasives contain (Fe,Ni)Si₂ and Si⁰ inclusions. Other inclusions in the artificial SiC were similar to those in natural SiC (moissanite) reported in reference data, including diverse metal silicides (e.g., FeSi, FeSi₂, Fe₃Si₇, and Fe₅Si₃), as well as a light rare earth element-enriched SiO phase and Fe-Mn-Cr alloys. These inclusions were produced by the in situ reduction of silica and the interaction between Si-containing coke and hot metals during the synthesis of the SiC abrasives. The results demonstrate that the SRP assemblage in the Jingshan chondrite originates from abrasive contamination and that the SRP inclusions (with a low content of Ca, Al, Ti, and Zr) cannot be used as a conclusive indicator for natural SiC. Additionally, the morphologies, biaxiality, and polytypes (determined by Raman spectroscopy) of SiC abrasives bear resemblance to those reported for natural SiC, and caution must be exercised when identifying the origin of SRP in samples processed by conventional methods using SiC abrasives. At the end of this paper, we propose more direct and reliable methods for distinguishing between natural and synthetic SiC. Full article

This study investigated the influence of high silicon (4.2 wt%) and varying aluminum (3.5–4.8 wt%) content on the high temperature oxidation behavior and thermophysical properties of SiMoAl vermicular graphite cast iron for hot-end exhaust components. Isothermal oxidation tests at 800 °C and nonisothermal oxidation tests in a dry-air atmosphere were conducted on SiMo nodular iron, along with two SiMoAl vermicular graphite cast iron variants alloyed with 3.5 wt% Al and 4.8 wt% Al. The investigations revealed the formation of a thin duplex layer of oxide scale, consisting of an iron-rich external oxide layer and continuous aluminum oxide at the metal/oxide interface. Although aluminum oxide acted as a protective barrier by impeding the solid-state diffusion of oxygen, severe subsurface oxidation was observed due to the interconnected vermicular graphite covered by aluminum oxides after decarburization. Furthermore, based on nonisothermal oxidation experiments, the effective activation energy of oxidation was found to be significantly increased by the addition of aluminum, even though the oxidation activation energies of SiMoAl samples exhibited small changes in comparison to each other. Additionally, thermophysical analysis demonstrated a substantial decrease in the thermal conductivity and a slight increase in the thermal expansion with the addition of aluminum. Full article

In the process of blast furnace smelting, the stability of the hearth thermal state is essential. According to the analysis of silicon content in hot metal and its change trend, the operation status of the blast furnace can be judged to ensure the stable and smooth operation of the blast furnace. Based on the error back-propagation neural network (BP), the flower pollination algorithm (FPA) is used to optimize the weight and threshold of the BP neural network, and the prediction model of silicon content is established. At the same time, the principal component analysis method is used to reduce the dimension of the input sequence to obtain relevant indicators. The relevant indicators are used as the input, and silicon content in the hot metal is used as the output, which is substituted into the model for training and utilizes the trained model to predict. The results show that the hit rate of the prediction model is 16% higher than the non-optimized BP prediction model. At the same time, the evaluation indicators and operation speed of the model are improved compared with the BP prediction model, which can be more accurately applied to predict the silicon content of the hot metal. Full article

Arrays of aligned carbon nanotubes (CNTs) are anisotropic nanomaterials possessing a high length-to-diameter aspect ratio, channels passing through the array, and mechanical strength along with flexibility. The arrays are produced in one step using aerosol-assisted catalytic chemical vapor deposition (CCVD), where a mixture of carbon and metal sources is fed into the hot zone of the reactor. Metal nanoparticles catalyze the growth of CNTs and, during synthesis, are partially captured into the internal cavity of CNTs. In this work, we considered various stages of multi-walled CNT (MWCNT) growth on silicon substrates from a ferrocene–toluene mixture and estimated the amount of iron in the array. The study showed that although the mixture of precursors supplies evenly to the reactor, the iron content in the upper part of the array is lower and increases toward the substrate. The size of carbon-encapsulated iron-based nanoparticles is 20–30 nm, and, according to X-ray diffraction data, most of them are iron carbide Fe₃C. The reasons for the gradient distribution of iron nanoparticles in MWCNT arrays were considered, and the possibilities of controlling their distribution were evaluated. Full article

The silicon content of the molten iron is an important indicator of the furnace temperature trend in blast furnace smelting. In view of the multi-scale, non-linear, large time lag and strong coupling characteristics of the blast furnace smelting process, a dynamic prediction model for the silicon content of molten iron is established based on the analysis of comprehensive furnace temperature characterization data. The isolated forest algorithm is used to detect anomalies and analyze the causes of the anomalies in conjunction with the blast furnace mechanism. The maximum correlation-minimum redundancy mutual information feature selection method is used to reduce the dimensionality of the furnace temperature characterization data. The grey correlation analysis with balanced proximity is used to obtain the correlation between the furnace temperature characterization parameters and the silicon content of the molten iron at different time lags and to integrate the furnace temperature characterization data set. The GRA-FCM model is used to divide the parameter set of the integrated furnace temperature characterization and construct a parameter directed network from multiple control parameters to multiple state parameters. The GWO-SVR model is used to predict the state parameters of each delay step by step to achieve dynamic prediction of the silicon content of the molten iron. Finally, the control parameters are adjusted backwards according to the prediction results of the state parameters and the silicon content of the molten iron and expert experience to achieve accurate control of the furnace temperature. Starting from the actual production situation of a blast furnace, the characteristic parameters are divided into control parameters and state parameters. This model establishes a multi-step dynamic prediction and closed-loop control model of “control parameters-state parameters-silicon content in hot metal-control parameters”. Full article

As electronics become more portable and compact, the demand for high-performance thermally conductive composites is increasing. Given that the thermal conductivity correlates with the content of thermally conductive fillers, it is important to fabricate composites with high filler loading. However, the increased viscosity of the composites upon the addition of these fillers impedes the fabrication of filler-reinforced composites through conventional methods. In this study, hexagonal-boron-nitride (h-BN)-pattern-embedded aluminum oxide (Al₂O₃) composites (Al/h-BN/Al composites) were fabricated by coating a solution of h-BN onto a silicone-based Al₂O₃ composite through a metal mask with square open areas. Because this method does not require the dispersion of h-BN into the Al₂O₃ composite, composites with high filler loading could be fabricated without the expected problems arising from increased viscosity. Based on the coatability and thixotropic rheological behaviors, a composite with 85 wt.% Al₂O₃ was chosen to fabricate Al/h-BN/Al composites. The content of the Al₂O₃ and the h-BN of the Al/h-BN/Al-1 composite was 74.1 wt.% and 12.8 wt.%, respectively. In addition to the increased filler content, the h-BN of the composite was aligned in a parallel direction by hot pressing. The in-plane (k_x) and through-plane (k_z) thermal conductivity of the composite was measured as 4.99 ± 0.15 Wm⁻¹ K⁻¹ and 1.68 ± 0.2 Wm⁻¹ K⁻¹, respectively. These results indicated that the method used in this study is practical not only for increasing the filler loading but also for achieving a high k_x through the parallel alignment of h-BN fillers. Full article

The carburization of molten iron is close to saturation in the blast furnace process, while that in the flash ironmaking process is uncertain because there is no pressure from solid charge and no carburization reactions occurring between the deadman and hot metal. Some experiments were conducted to reveal the kinetic mechanism of coke dissolving in carbon-iron melts. Reduced iron powder, electrolytic iron as well as chemical pure graphite were used as experiment materials. With high-purity argon injected as the protective gas, the specimens were heated up to 1873 K in a tubular resistance furnace to study the isothermal mechanism. The results show that the composition of the ferrous sample affects the dissolution rate. When the FeO content of the iron-bearing material rises from 0% to 4.76%, the apparent dissolution rate constant, k_t, falls from 7.98 × 10⁻⁶ m/s to 5.48 × 10⁻⁶ m/s. There are some differences amongst the dissolution rate coefficients of different cokes despite interacting with similar carbon-iron melts, with coke 1 of 7.98 × 10⁻⁶ m/s, coke 2 of 5.17 × 10⁻⁶ m/s, and coke 3 of 3.77 × 10⁻⁶ m/s. Besides, this index decreases with the increase of the dissolution time and solely depends on the procedure of the mass transfer. A negative correlation is demonstrated between k_t and the sulfur content in the iron bath as well. The content of silicon dioxide in the coke has a significant influence on k_t. Additionally, the dissolution rate coefficient increases with the increase of the graphitization degree of coke. Full article