Search Results (10)

To address the limited data availability and low predictive accuracy of nickel matte grade models in the early stages of facility operation, this study introduces a unique stepwise prediction methodology that integrates data augmentation and ensemble learning, specifically tailored for limited industrial datasets. Predicting matte nickel grade accurately is critical for nickel sulfate production, a key precursor in cathode manufacturing. However, in newly adopted facilities, operational data are scarce, posing a major challenge for conventional machine learning models that require large, well-balanced datasets to generalize effectively. Moreover, the nonlinear dependencies between raw material composition, operational conditions, and metallurgical reactions further complicate the prediction task, often leading to high errors in traditional regression models. To overcome these challenges, this study introduces an innovative approach that integrates feature engineering, Gaussian noise augmentation, SMOTE regression, and a stacking ensemble model, using XGBoost (2.0.3) and CatBoost (1.2.7). First, input variables were refined through feature engineering, followed by data augmentation to enhance dataset diversity and improve model robustness. Next, a stacking ensemble framework was implemented to mitigate overfitting and enhance predictive accuracy. Finally, SHAP, an XAI technique that quantifies the impact of each input variable on the model’s predictions based on cooperative game theory, was employed to interpret key process variables, offering deeper insights into the factors influencing nickel grade. The experimental results demonstrate a substantial improvement in prediction accuracy, with the R² coefficient increasing from 0.3050 to 0.9245, alongside significant reductions in RMSE, MAE, and MAPE. The proposed methodology not only enhances predictive performance in data-scarce industrial environments but also provides an interpretable framework for real-world process optimization. These findings validate its applicability to nickel matte operations, offering a scalable and explainable machine learning approach for metallurgical industries with limited data availability. Full article

Due to the scarcity of modeling samples and the low prediction accuracy of the matte grade prediction model in the copper melting process, a new prediction method is proposed. This method is based on enhanced generative adversarial networks (EGANs) and random forests (RFs). Firstly, the maximum relevance minimum redundancy (MRMR) algorithm is utilized to screen the key influencing factors of matte grade and remove redundant information. Secondly, the GAN data augmentation model containing different activation functions is constructed. And, the generated data fusion criterion based on the root mean squared error (RMSE) and the coefficient of determination (R²) is designed, which can tap into the global character distributions of the copper melting data to improve the quality of the generated data. Finally, a matte grade prediction model based on RF is constructed, and the industrial data collected from the copper smelting process are used to verify the effectiveness of the model. The experimental results show that the proposed method can obtain high-quality generated data, and the prediction accuracy is better than other models. The R² is improved by at least 2.68%, and other indicators such as RMSE, mean absolute error (MAE), and mean absolute percentage error (MAPE) are significantly improved. Full article

The duration of the high-grade matte converting process is short, the amount of slag is small, and it is difficult for the original impurity removal operation in the low-grade matte converting process to meet the current production demand. Because the removal method for impurity elements during high-grade matte converting is unclear, the phase transformation of impurity elements during this process is investigated in this study. The results show that arsenic exists mainly in the form of FeAsO₄ and As₂O₅, antimony in the form of Sb and Sb₂O₅, and lead in the form of PbS and PbO in high-grade matte. During the converting process, arsenic and antimony mainly exist in the melt in the form of oxides and gradually aggregate into large particles with increasing copper content in the melt. Lead exists in matte in the form of PbS until the end of the converting process, and PbS is not completely oxidized until the matte converted to blister copper phase. The phase transformation characteristics of copper, iron, sulfur and impurity elements in the process of high-grade matte converting were revealed. This study provides a theoretical reference for the formulation of an efficient impurity removal scheme for the converting process. Full article

Waste electrical and electronic equipment (WEEE) contains various valuable metals, making it a potential secondary resource for sustainable metal usage. Pyrometallurgical smelting is an efficient technique to recycle WEEE by extracting precious metals into copper matte and removing impurities into slags. The impact of WEEE impurities such as CaO and Al₂O₃ on the phase compositions of the smelting products attracts great attention for industrial metal recovery. This study clarified the impact of CaO and Al₂O₃ on the equilibrium phase compositions of copper matte and SiO₂-saturated FeO_x-SiO₂-Al₂O₃-CaO slags. The high-temperature smelting experiments were taken at a controlled p(SO₂) of 0.1 atm and 1300 °C, followed by quenching and electron probe microanalysis. The results showed that the copper and sulfur in the smelting system were highly deported into copper matte, and their distribution in matte was enhanced by increasing CaO and Al₂O₃ concentrations introduced by WEEE. The chemical copper dissolution in slags increased with increasing matte grade but decreased by adding CaO and Al₂O₃. The iron was preferentially concentrated in slags, and higher matte grades improved the iron distribution in slags. The current experimental results enrich fundamental thermodynamic data and help optimize WEEE smelting operations for efficient recovery of valuable metals. Full article

The results of studies on the distribution of rare elements among the products of distillation processing of polymetallic mattes are present in this article. Schemes of the developed technological equipment for the implementation of the extraction processes of rare elements via the vacuum distillation of mattes are presented. Technological tests were performed with a matte of lead, copper, and antimony plants at 1100–1250 °C and a pressure of up to 700 Pa. It was established that As, Cd, Bi, In, and Ge, by more than 90% in total, are extracted into condensate and dust in the distillation process of volatile components from mattes of lead production. At the same time, antimony is distributed between the distillate residue and condensate. Antimony by 90.47%, arsenic by 78.83% and cadmium by 98.72% are distributed into sulfide condensate and dust in the distillation of copper production matte. From the matte of the antimony plant, Sb and Bi (90.76% and 89.78%, respectively) are transferred into the condensate and cyclone dust. Arsenic is distributed between the liquid and vapor phases. Based on calculations, Se and Te will be mainly concentrated in the distillation residue. High-grade copper mattes obtained in processing mattes from lead and copper plants can be further used to obtain metallic copper by converting. The condensate and dust can be processed separately or with the dust of the mainline production for rare metal extraction. Antimony matte processing condensate containing more than 70% Sb can be directed to the process of crude antimony refining. Full article

In recent years, with the development of the coating industry and the increasing awareness of environmental protection, the modification of waterborne wood coatings has become the focus of research. Generally, the system composed of silica sol modification and UV curing can make up for the defects of poor mechanical properties, low hardness, and slow curing speeds of waterborne wood coatings. Herein, we used silica sol-reinforced UV-curable waterborne acrylic wood coatings and tested the related physical properties of the coatings. FT-IR analysis showed that the Si-O-Si bond appeared, indicating that the silica sol was successfully grafted onto the waterborne acrylic molecular chain. The results showed that the mechanical properties of the UV-curable waterborne acrylic wood coating film reached their optimum when the content of silica sol was 1 wt%, the number of UV lamps was 3, and the drying time was 20 min. The corresponding values for wear resistance, hardness, adhesion, and impact strength were 0.106 g (high level), grade 3, and 90 kg·cm, respectively. However, when the content of silica sol is greater than 1 wt%, the related physical properties of the coatings will decrease. The results showed that the gloss of the coating decreased with increasing silica sol content. When the silica sol content was 2 wt%–6 wt%, the coating showed a matte gloss. This present work shows that the modification process is simple, controlled, inexpensive, and meets the demand for UV-curable waterborne acrylic wood coatings in daily life. Full article

Bottom-blowing copper smelting is a bath smelting technology recently developed in China. It has the advantages of good adaptability of raw materials, high oxygen utilization and thermal efficiency, and flexible production capacity. Plume eye is a unique phenomenon observed in the bottom-blowing copper-smelting furnace where the slag on the surface of the bath is pushed away by the high-pressure gas injected from the bottom. The existence of plume eye was first confirmed by analyzing the quenched industrial samples collected above the gas injection area and then investigated by laboratory water model experiments. Combining the plant operating data and the smelting mechanism of the copper concentrate, the role of the plume eye in bottom-blowing-enhanced smelting is analyzed. It reveals that the direct dissolution of copper concentrate as a low-grade matte into the molten matte can significantly accelerate the reactions between the concentrate and oxygen. The productivity of the bottom-blowing furnace is therefore increased as a result. The effects of the gas flow rate and thickness of the matte and of the slag layer on the diameter of the plume eye were studied using water-model experiments. It was found that increasing the gas flow and the thickness of the matte and reducing the thickness of the slag can increase the diameter of the plume eye. This work is of great significance for further understanding the copper bottom-blowing smelting technology and optimizing industrial operations. Full article

Annually, hundreds of thousands of tons of slags are involved in the reverberator and flash smelting as well as converting operations of Cu-Fe sulfide concentrates to produce matte in the Sar Cheshmeh copper smelter plant, Iran, disposed in the landfill and cooled in air. Due to their relatively high average copper content (about 1.5 wt%), a mineral processing plant based on the flotation process has recently been established to produce thousands of tons of Cu-sulfide concentrate after slag crushing and fine grinding operation. In order to make the flotation process more efficient, more knowledge is required on the form and origin of the copper losses in the slag. To achieve this, mineralogical studies of the slags using optical microscopy, X-ray diffraction (XRD), and scanning electron microscopy (SEM) methods have been carried out. Mineralogical analyses showed the main part of copper losses into the semi- to fully-crystallized magnetite-rich reverberator and flash slags characterized by crystal–glass matrix ratio ≤ 1 is moderate to coarse particles of Cu-Fe sulfides, i.e., chalcopyrite (CuFeS₂) and bornite (Cu₅FeS₄), that are mainly chemically entrapped. In contrast, the mechanically entrapped fine- to coarse-grain (from 20 up to 200 µm) spherical-shaped of high-grade matte particles with chalcocite (Cu₂S) composition containing droplets or veinlets of metallic copper (Cu⁰) are the dominant forms of copper losses into the converter slags characterized by crystal–glass matrix ratio > 1. From the value recovery point of view, our result show that the fully crystallized slags containing moderate- to coarse-grain copper-bearing particles will result in efficient recovery of a significant amount of entrained copper due to better milling response compared to semi-crystallized ones due to locking the fine- to moderate-grain copper particles in the silicate glassy matrix. Laboratory-scale grinding experiments showed that normal (≤74 μm) to fine (≤44 μm) grinding of high- Cu grade slags lead to a significant increase in the liberation degree of copper particles. in contrast, the increase in fine particle fractions (<37 μm) due to re-grinding or ultra-fine grinding of the originally low-Cu grade slags does not lead to the liberation of copper particles, but it will reduce the efficiency of the flotation process. This study suggests that the highest rate of copper recovery of the slag by the flotation process will be obtained at particle size 80% passing 44 µm which has also reached the optimal liberation degree of copper-bearing particles. Full article

The slags generated in the conventional copper conversion process are mainly composed of Cu₂O–Fe₂O₃–SiO₂ with CaO, Al₂O₃, and MgO compounds—in concentrations up to 10 wt %. The present work contributes to the knowledge of the conversion process, generating experimental data for the phase diagrams of the Cu₂O–Fe₂O₃–SiO₂–Al₂O₃ and Cu₂O–Fe₂O₃–SiO₂–CaO systems. The experiments were carried out in a tubular furnace at temperatures of 1150 °C and 1200 °C, under a condition of saturation with tridymite and spinel. Once the equilibrium was reached, the samples were immediately quenched in water. The phases in the samples were observed through a scanning electron microscope (SEM) and the elemental composition of the phases were analyzed by means of energy-dispersive X-ray spectroscopy (EDS) detectors. The addition of Al₂O₃ and CaO into the Cu₂O–Fe₂O₃–SiO₂ system resulted in an appreciable displacement of the liquidus lines, corresponding to an expansion of the liquid in the tridymite primary phase field. The addition of CaO and Al₂O₃ combined was evaluated on industrial slags and from samples obtained in a Peirce–Smith furnace, with increasing amounts of CaO in the flux. Full article

The control of arsenic, a toxic and carcinogenic element, is an important issue for all copper smelters. In this work, the reaction mechanism and distribution behavior of arsenic in the bottom blown copper smelting process (SKS process) were investigated and compared to the flash smelting process. There are obvious differences of arsenic distribution in the SKS process and flash process, resulting from the differences of oxygen potentials, volatilizations, smelting temperatures, reaction intensities, and mass transfer processes. Under stable production conditions, the distributions of arsenic among matte, slag, and gas phases are 6%, 12%, and 82%, respectively. Less arsenic is reported in the gas phase with the flash process than with the SKS process. The main arsenic species in gas phase are AsS (g), AsO (g), and As₂ (g). Arsenic exists in the slag predominantly as As₂O₃ (l), and in matte as As (l). High matte grade is harmful to the elimination of arsenic to gas. The changing of Fe/SiO₂ has slight effects on the distributions of arsenic. In order to enhance the removal of arsenic from the SKS smelting system to the gas phase, low oxygen concentration, low ratios of oxygen/ore, and low matte grade should be chosen. In the SKS smelting process, no dust is recycled, and almost all dust is collected and further treated to eliminate arsenic and recover valuable metals by other process streams. Full article