PurposeConventional manufacturing (CM) methods for complex forgings involve high energy consumption and substantial carbon emissions, posing significant environmental challenges. This study aims to evaluate the emissions of hybrid deposition and micro-rolling (HDMR) technology compared with CM in producing aviation forgings, assessing its feasibility as a sustainable alternative.Design/methodology/approachUsing an industrial metabolism model based on carbon sources, this study conducts a carbon accounting analysis of the manufacturing process for an aircraft landing gear external cylinder. The carbon emission performance of HDMR and CM methods is evaluated and compared along two key dimensions: raw material consumption and energy usage.FindingsThe results show that the total CO2 output of the HDMR method for manufacturing the external cylinder are reduced by a coefficient of 32 compared to the CM method, and the metal raw material consumption is reduced by 85.40%.Originality/valueThis study proposes a novel industrial metabolism model to assess carbon emissions in HDMR processes. The results confirm the potential of HDMR to significantly reduce carbon footprint and material consumption in forging manufacturing, thereby supporting the transition of industry to low-carbon production.

To fulfill diverse performance requirements in various industrial fields, it is essential to precisely control the inclusion characteristics in 316L stainless steel (316L SS), including chemical composition, size, and quantity and so on, thus ensuring the steel with high cleanliness and quality. Solidification experiments with different cooling methods were conducted using an Al2O3–CaO–SiO2–MgO refining slag system to investigate the influence of cooling methods upon the characteristic variation and growth behavior of inclusions in 316L SS during the solidification process. Thermodynamic mechanisms underlying the inclusion composition transition during the solidification process were revealed, and a kinetic model for the prediction of the solute element segregation and the inclusion growth in the steel was developed and verified. Results indicated that a nonlinear fitting curve equation for the correlation between the average size of inclusions d and the cooling rate Rc was obtained, which was: d = 3.728 · R_c^ - 0.134 . As the cooling rate went down, Al2O3 content in the inclusions correspondingly decreased from 69.43 to 30.54 mass

Recently, the utilization of deep reinforcement learning (DRL) for solving combinatorial optimization problems, specifically in improving heuristic algorithms for routing problems, has garnered significant attention. However, most methods often depend on a single operator to perform local search, which limits their solving ability across various types of problems. This paper introduces a novel adaptive heuristic for solving routing problems by lever aging DRL, where a policy network is designed to facilitate decision-making on both the selection of the operator and the node position. Firstly, a self-attentive mechanism is employed to extract sequence features, thereby fa cilitating the identification of the specific position for the search operator. Secondly, a convolutional layer is used to transform a single output into multiple outputs, thus enabling adaptive operator selection. By integrat ing self-attention and convolution within the same policy network, this approach can simultaneously output the appropriate search operator and the corresponding node pair required to modify the solution. Furthermore, a simple fine-tuning strategy is introduced for model training. Experimental results from both randomly generated and real-world datasets demonstrate that the proposed method outperforms learning methods based on a single operator in terms of solution quality, especially in traveling salesman problem (TSP) instances and small-to-medium-scale capacitated vehicle routing problem (CVRP) instances. This method also has strong generalization ability and is applicable to various problem distributions and scales.

A combined approach of numerical simulation and water model experiments was employed to investigate the steel–slag–air multiphase flow behavior and initial solidification characteristics in a 200 mm × 1248 mm slab mold with varying submerged entry nozzle (SEN) inclination angles and argon-blowing rates. The results demonstrated that as the argon-blowing rate increases from 0 to 15 L/min, the fluctuation amplitude at the steel–slag interface expands from 6.4 to 14.3 mm. When the argon-blowing rate ranges between 0 and 10 L/min, the shell thickness at both narrow and wide faces increases from 6.68 and 23.1 to 12.38 and 27.11 mm, respectively, but excessive blowing rates lead to shell thinning. The steel exposure behavior is eliminated as the SEN is inclined downward. As the SEN inclination angle decreases from upward 12° to downward 15°, the relationship between the SEN inclination angle and shell thickness growth exhibits nonlinear characteristics. The shell thickness at mold out ranges from 9.71 to 20 mm at the narrow face and 12.49 to 27 mm at the wide face. The superior parameters are with 10 L/min argon flow and 12° downward nozzle inclination, delivering no slay layer while achieving mold exit shell thicknesses of 11.63 mm at the narrow face and 26.5 mm at the wide face.

The global steel industry’s shift toward green manufacturing is intensifying the need for efficient scrap utilization. A three-dimensional mathematical model coupled turbulent flow, solidification/melting behavior, and species transfer was established to investigate the melt flow and scrap melting characteristics in a 210 t bottom-blown hot metal ladle. The study systematically evaluated the effects of argon blowing rate, scrap carbon content, preheating temperature, and specific surface area. Key results indicated that increasing the argon blowing rate to 200 L min−1 can reduce dead zones and shorten scrap melting time by up to 77 pct. Scrap carbon content emerged as a critical factor, with high-carbon scrap (1.836 and 3.672 wt pct) forming a thin solidified layer and melting completely within 120 seconds, whereas low-carbon scrap (0.183 to 0.612 wt pct) formed a thicker solidified layer (volume increase to 158 pct of the original volume), prolonged melting time to 200 to 375 seconds. Preheating the scrap to 1273 K can reduce the initial solidified layer volume to 107 pct of its original volume and shorten the initial melting time by about 38 pct, though its effect on the subsequent melting stage was limited. Geometrically, cylindrical scrap (239 m2 m−3) melting is 16.7 pct faster than a cubical pieces due to superior heat exchange. The present work can provide a validated strategy for enhancing scrap melting efficiency in various iron-carbon vessel.

In copper smelting, the traditional ladle slow-cooling process is inefficient for waste-heat recovery and poses environmental risks. Dry centrifugal granulation offers a promising solution, where granulation efficiency dictates heat-recovery performance. To this end, this study established a three-dimensional simulation model for the centrifugal granulation of copper slag, systematically investigating the influence of the granulator's process parameters and structural configurations on the granulation behavior. The results indicate that increasing the rotational speed from 600 rpm to 1200 rpm effectively reduces the average particle diameter by 20%. Surface roughness had a limited effect on the average diameter. However, higher roughness led to a more dispersed particle size distribution, adversely affecting granulation uniformity. Increasing the granulator size alters the liquid film breakup mechanism. When the diameter reached 300 mm, the particle size distribution tended to disperse, and its trend exhibited an inflection point, suggesting the existence of an optimal size range. Compared to disc granulators without vanes, the Sauter mean diameters of the two types of granulators equipped with guide vanes-the vane-disc and the vane-cup-increase by approximately 82% and 37%, respectively. Furthermore, the innovative application of the single rotating frame model effectively enhanced the reliability of simulations for complex-structured granulators. This study thereby offers crucial guidance for optimizing industrial copperslag centrifugal granulation and presents a new methodology for designing complex granulators.

Phosphogypsum (PG), a byproduct of the phosphate fertilizer industry, is a promising candidate for CO2 mineralization due to its high calcium content. However, the direct mineralization of PG typically yields calcite rather than the metastable vaterite phase of CaCO3, which holds superior industrial value. This study investigates the influence of impurities in PG on vaterite formation through experimental and density functional theory (DFT) approaches. Through the analysis of diverse impurity types in PG, fluorine was identified as the dominant element influencing vaterite formation. Mineral carbonation experiments were conducted using pure CaSO4 & centerdot;2H2O and PG with varying fluorine species (NaF, Na2SiF6, and CaF2) and concentrations. It revealed that soluble fluorine significantly promotes calcite formation, even at a low concentration of 0.1 wt %, while sparingly soluble fluorides exhibit minimal effects. DFT calculations demonstrated that F- adsorption on vaterite (110) surfaces involves strong chemical interactions, destabilizing the structure and facilitating its transformation to calcite. In that case, a selective flotation process was used to remove soluble fluorine in PG, indicating that flotation pretreatment effectively reduced fluorine content in PG from 0.62 to 0.13%. Flotation PG enables the synthesis of high-purity vaterite with 70% content via direct mineralization. These findings elucidate the critical role of fluorine in governing CaCO3 polymorphism and provide a practical pathway for enhancing the value of PG-derived products through impurity control.

The complex oolitic structure of high-phosphorus iron ore (HPIO) poses a major challenge to dephosphorization, restricting its high-value utilization. This study proposed a carbothermal smelting reduction process to simultaneously extract and enrich Fe and P from HPIO into a valuable Fe-P alloy. Thermodynamic calculations demonstrate that Fe and P are preferentially reduced and the increase of C/O ratio contributes to the enrichment of P, C, and Si elements in molten iron phase. Reduction experiments confirm that the oolitic structure begins to be destroyed at 1000 degrees C and a high temperature above 1400 degrees C is essential for the separation of iron and slag. After smelting reduction at 1600 degrees C and C/O = 1.0, a Fe-P alloy containing 1.33 wt% P was obtained with high recovery rate of 96.24% for Fe and 86.94% for P. Kinetic studies derived from thermogravimetric analyses reveal that the carbothermal reduction of HPIO is governed by rate-limiting steps: (1) interfacial chemical reaction dominates below 725 degrees C; (2) gas diffusion becomes limiting at 835 degrees C-890 degrees C; and (3) interfacial chemical reaction reemerges as the limiting step above 1000 degrees C. These findings provide fundamental insights into the reduction of HPIO and offer a promising route for the high value utilization of HPIO.

Large flexible appendages of spacecraft are prone to prolonged vibration under external excitation, which degrades on-orbit attitude control precision and stability. Model reduction of the finite element model for flexible appendages is essential for efficient dynamic and control analysis of spacecraft with complex configurations. The interfacial dynamic stiffness matrix for flexible appendages, with statically determinate interfaces and weak damping, is investigated to establish a symmetric order-reduced model through theoretical derivation. A demonstration case study of a spring-mass system with multiple degrees of freedom is performed to illustrate the model reduction process and to verify the accuracy of the proposed method. A finite element model of a full-scale solar array is further presented to verify the practical feasibility of this model reduction approach in real engineering applications.

To enhance the calcium yield during calcium treatment of non-oriented silicon steel, a novel Si-Fe-Ca alloy is designed and prepared using commercial FeSi, Si-Ca alloy, and silicon steel scrap. The influence of Fe content on elemental yield, phase characteristics, physicochemical properties, and calcium-treatment efficiency is systematically investigated. The obtained results show that the yields of Si and Ca elements exceeding 85% in the prepared Si-Fe-Ca alloy. Low-Fe alloys (S1, S2) primarily consist of Si, FeSi2, and CaSi2 phases. With increasing Fe content (S3-S6), the dominant phase changes from Si to FeSi, accompanied by refined Si phase, coarsened FeSi and finely dispersed CaSi2 in the matrix. Thermodynamic calculations indicate that although higher Fe content raises Ca activity, it remains below 0.0008 (<50 wt% Fe), and the resulting vapor pressure is much lower than that of pure Ca, effectively suppressing Ca volatilization. The activity and vapor pressure of Si in the alloy also decrease with Fe. Alloy density increases as the Fe content increases, dynamic viscosity decreases, and the liquidus temperature exhibits a non-monotonic trend, first decreasing and then increasing. Calcium treatment experiments in silicon steel demonstrate that alloys with higher Fe content promote the modification of Al2O3 inclusions into low-melting-point Al-Ca-O and Al-Ca-O-S complex inclusions because increased alloy density prolongs residence time in steel, and refined CaSi2 phase lowers Ca release rate, and reduced local Ca activity suppresses violent vaporization of Ca. This study provides theoretical and technical insights for designing highCa-yield alloys and optimizing calcium treatment in non-oriented silicon steel production.

The non-linear influence of hot metal composition and scrap ratio on flux addition requires accurate predictive models to optimise charging operations. A random forest (RF) model was developed using multi-heat industrial data to predict lime and light-burned dolomite additions. The model effectively captured non-linear interactions between key process variables and flux inputs but showed slight systematic deviations under fluctuating operating conditions. To improve prediction accuracy and physical consistency, metallurgical mechanisms were embedded into the data-driven framework. Empirical features for light-burned dolomite were obtained by polynomial fitting, while theoretical lime additions were derived from quaternary basicity theory and used as mechanistic constraints. This hybrid model retained the non-linear learning capability of RF while enhancing interpretability and robustness. After feature enhancement, all performance indicators improved markedly: for dolomite, the coefficient of determination (R2) increased from 0.4801 to 0.5675, the mean absolute error (MAE) decreased from 123.17 kg to 116.52 kg and the root mean square error (RMSE) from 152.90 kg to 139.46 kg; for lime, R2 rose from 0.5843 to 0.7553 and MAE and RMSE dropped by 24.1% and 23.3%, respectively. The proportion of samples within +/- 5% error increased significantly, confirming improved reliability for basic oxygen furnace (BOF) charge prediction and physically consistent steelmaking control.

Amorphous magnetic powders (AMPs), an emerging soft magnetic material, are highly applicable to high-frequency electronic devices. However, commercial FeSiBCCr AMPs require remelting and alloying of high-purity pre-alloys, giving rise to complicated fabrication process and elevated production cost. In this study, a novel strategy was proposed to directly prepare AMPs from laterite nickel ore (LNO) rich in Fe, Ni, Cr and Si elements. Thermodynamic calculations coupled with carbothermal reduction experiments were conducted to investigate the phase evolution and element migration during the reduction of LNO. FeNiCrSiCB alloy ribbons and powders were then prepared from the reduced alloy to verify the feasibility of the developed strategy. The results reveal that the complete slag-iron separation is achieved at 1600 °C, where high-temperature carbothermal smelting reduction of LNO enables simultaneous enrichment of Fe, Ni, Cr, Si, C elements. Of practical significance, the composition of reduced alloy can be tailored by controlling the C/O ratio, which refers to the molar ratio of carbon in reducing agent to oxygen in iron oxides of LNO. At C/O = 0.95, the resulting reduced alloy reaches a high Fe recovery rate of 97.03% with a nominal composition of Fe91.4Ni2.3Cr4.8Si0.5C1.0 (wt.%). Supplementary additions of 6.80 wt% Si and 2.70 wt% B further yield FeNiCrSiCB amorphous ribbons and powders featuring homogeneous element distribution. These findings verify the feasibility of directly preparing FeNiCrSiCB amorphous alloys from LNO, which is of paramount importance for efficient production of AMPs and comprehensive utilization of LNO.

Current research focuses on analyzing the melting mechanism of scrap in laboratory settings, with limited reports on scrap melting within actual converter blowing environments. A three-dimensional full-scale converter mathematical model coupling turbulent, multiphase, heat transfer, and mass transfer was established to examine the melting behavior of scrap with varying initial weight and carbon concentration in a combined blowing converter. And, the mathematical model was verified by a water model experiment and a thermal experiment. The results showed that the time required for complete melting of 100 kg scrap was 495 s, exceeding that of 150 and 200 kg scrap by 105 and 30 s, respectively, due to changes in the specific surface area of scrap during the melting process. The melting time for scrap with a carbon concentration of 0.612 wt.% is notably shorter at 465 s compared to 1575 s for 0.326 wt.% carbon concentration and 2475 s for 0.183 wt.% carbon concentration. Additionally, secondary solidification occurs during the melting of low-carbon steel and medium-carbon steel. For scrap with a carbon concentration of 0.612 wt.%, the carburizing time was 134 s, significantly lower than 300 s for 0.326 wt.% carbon concentration and 445 s for 0.183 wt.% carbon concentration, respectively.

Spring steel is widely used in high-stress components due to its excellent mechanical properties, but its fatigue performance is often compromised by high-melting-point, brittle Al2O3–SiO2–CaO–(MgO) system inclusions, especially for inclusions containing spine phases. This study investigates the modification of these inclusions in 60Si2Mn spring steel using mixtures of Fe powder and Na2CO3 with various ratios. Results show that Na2O significantly expands the low-melting-point region of the inclusions. After Na2CO3 addition, both the proportion of Na2O‑containing inclusions and the Na2O content within them reach maximum levels after a reaction duration of 20 to 40 minutes. Adding a mixture of Fe powder and Na2CO3 suppresses Na2CO3 volatilization, with an optimal Na2CO3:Fe ratio of at least 1:1. Thermodynamic analysis reveals that Na2O enters inclusions by displacing CaO in the Al2O3–SiO2–CaO–(MgO) system. Moreover, the presence of Na2O in the inclusions affects the activities of Al2O3, SiO2, and CaO, thereby promoting the corresponding reactions. Moreover, Na2O lowers the activity of the spinel phase, suppressing its precipitation from the inclusions during the solidification process. When the Na2O content in inclusions exceeds 10 pct, no spinel precipitation occurs during solidification at temperatures above 1200 °C. The modified inclusions exhibit markedly improved deformability during hot-rolling simulations, thereby reducing the risk of fatigue crack initiation. This work elucidates the modification mechanism and provides both theoretical insights and practical guidance for optimizing the smelting process of spring steel to enhance its service reliability.

As global environmental awareness grows, reducing energy consumption have become imperative in the manufacturing sector. Hybrid additive manufacturing (HAM) offers an innovative approach to making complex parts compared to traditional manufacturing processes. To achieve sustainable development in this field, accurate predictions of energy consumption and efficiency prior to manufacturing are essential for optimizing energy utilization. This study applies industrial metabolism analysis to characterize the energy metabolism framework of hybrid deposition and micro-rolling (HDMR), and integrates this with G-code parsing to propose a G-code-driven dynamic industrial metabolism model for predicting the energy efficiency of manufacturing processes. Based on the analysis of two case studies, the energy efficiency prediction deviations for the total manufacturing system and for the HDMR process range from 5.33

This research examines the impact of varying solidification methods on the behavior of nonmetallic inclusions in 304 stainless steel during the solidification process. Five sets of cooling experiments are carried out to systematically characterize and analyze the inclusion features. The findings indicate that when the cooling rate is reduced from 45.33 to 1.21 K s(-1), the average inclusion diameter increases from 1.64 to 3.31 mu m, while the number density declines from 31.97 to 16.47 mm(-2). Specifically, the percentage of small-sized inclusions (<3 mu m) drops sharply from 93.8% to 41.4%. In terms of composition, as the cooling rate diminishes, the average contents of MnO and SiO2 in the inclusions rise, whereas the Cr2O3 content decreases. Morphologically, the inclusions undergo a gradual transition from spherical or ellipsoidal homogeneous-phase particles to irregular multiphase structures. Building upon a solute segregation model, a kinetic model for inclusion growth is developed. Calculations performed with this kinetic model reveal a notable inverse relationship between the size of the inclusions and the rate of cooling, as described by the fitted equation: . The predictions generated by the model show a strong correlation with the experimental results.

To address the issues of interfacial stress concentration and nozzle clogging induced by hard Al2O3 inclusions in St12 cold-rolled steel, this study systematically investigated the effects of various calcium-containing alloys (CaSi, CaSiFe, AlCa10, and AlCa75) on the evolution behavior and modification mechanism of inclusions. Laboratory refining experiments were conducted at 1600 °C, and the evolution of inclusion morphology, size, number density, and composition at varying durations (1 to 16 minutes) post-calcium treatment was characterized utilizing ASPEX and SEM-EDS. Additionally, predominance area diagrams and Liquid Phase Window under different conditions were calculated using the FactSage 8.1 thermodynamic software. The kinetic evolution process of inclusions was evaluated based on the established kinetic equations. The results demonstrate that the alloy type and composition directly govern the calcium release behavior and the phase evolution trajectory of inclusions. CaSi, CaSiFe, and the low-calcium alloy (AlCa10) exhibited superior modification efficiency, transforming 85 to 93 pct of inclusions into low-melting-point, spherical complex liquid inclusions after 16 minutes of treatment. Conversely, the high-calcium alloy (AlCa75) resulted in an extremely low calcium yield due to intense vaporization at high temperatures, leaving up to 51 pct of unmodified spinel phases in the melt. Furthermore, thermodynamic calculations revealed that fluctuations in the calcium content of molten steel affect the size of the Liquid Phase Window and shift the critical calcium concentration for CaS formation, thereby exacerbating the risk of sulfide precipitation. The compositional evolution of inclusions is governed by interfacial mass transfer kinetics. Among the investigated alloys, the AlCa10 alloy exhibited the highest apparent modification rate while achieving the maximum theoretical extent of transformation. Furthermore, by employing a “three-stage” melting kinetic model, this study elucidates the primary physical mechanism responsible for the superior calcium release stability of the CaSiFe alloy: the local high-silicon concentration field effect significantly enhances calcium solubility, driving micro-calcium bubbles to rapidly dissolve into the liquid steel matrix prior to their escape. This study provides valuable theoretical and experimental foundations for alloy selection and inclusion control during the calcium treatment of St12 steel in industrial production.

This study addressed the issues of uneven molten steel flow and low inclusion removal rate in a three-strand induction heating tundish (IHT) by evaluating four optimized designs. The optimizations involved modifying the port diameter and upward angle of the double-branch port channel (DPC) and were assessed based on flow patterns, residence time distribution (RTD) curves, temperature uniformity, and inclusion removal behavior. The results indicated that a design without diameter reduction and with a 15 degrees upward inclination at Port 1 effectively eliminated short-circuit flow, increased the average residence time by 10.57 s, reduced the dead volume by 1.1%, and ensured excellent flow consistency. The mass flow ratio among the strands remained stable at approximately 1.67:1, which is close to the theoretical value. In terms of thermal performance, the optimized DPC achieved the best interstrand uniformity, with outlet temperature differences of only 0.339 K without and 0.422 K with induction heating, respectively. It also delivered the highest inclusion removal rates: 55.88% for 10 mu m particles and 60.63% for 20 mu m particles, representing improvements of 9.39% and 8.72% over the original structure. Moreover, industrial trials confirmed that the optimized DPC enhances steel cleanliness by apparently reducing inclusion exceedance rates in the casting bloom.

To tackle the challenges of liquid steel reoxidation and mixing control during grade changes, a three-dimensional mathematical model incorporating turbulence, multiphase flow, and mass transfer is developed for a single-strand tundish–mold system. The model is validated through water modeling experiments and systematically examines the effects of tundish emptying time and refilling strategies on the behavior of the steel–slag–air multiphase system and the mixing of steel grades. The results demonstrate that a liquid-level drop of 371.5 mm during the 5-minute tundish emptying process prior to introducing the new grade is optimal. Under this condition, a 90 pct concentration of the new grade is achieved at the mold outlet in approximately 2400 seconds, without prolonged steel exposure caused by an excessively low liquid level. Furthermore, with this emptying strategy, maintaining a low casting level of 620 mm without a tundish refilling operation further shortens the mixing time to 1948 seconds, with a transition slab length of 35.7 m, while effectively controlling surface fluctuations and steel exposure, and completely eliminating secondary exposure. This study offers valuable insights for the integrated control of the tundish–mold system during transient casting operations.