In this work, a novel type of short-process deformation technology of Mg alloys, bifurcation-equal channel angular pressing (B-ECAP), was proposed to refine grain and improve the basal texture. The cylindrical billets were first compressed into the die cavity, then sequentially flowed downward through a 90° corner and two 120° shear steps. The total strain of B-ECAP process could reach 3.924 in a single pass. The results of microstructure observation showed that DRX occurred at upsetting process in the die cavity and completed at position D. The grains were refined to 6.3 µm at being extruded at 300 °C and grew obviously with the extrusion temperature increase. The shear tress induced by 90° corner and two 120° shear steps resulted in the basal poles of most grains tilted to extrusion direction (ED) by ±25°. Compared with the original billets, the extruded sheets exhibited higher yield strengths (YS), which was mainly attributed to the grain refinement. The higher Schmid factor caused by ED-tilt texture resulted in a fracture elongation (FE) more than that of the original bar in ED, while was equivalent to that in transverse direction (TD). As the extrusion temperature increased, the variation of UTS and YS in ED and TD decreased gradually without ductility obviously decrease.

In this study, the microstructure and mechanical properties of a multi-layered 316L-TiC composite material produced by selective laser melting (SLM) additive manufacturing process are investigated. Three different layers, consisting of 316L stainless steel, 316L-5 wt% TiC and 316L-10 wt% TiC, were additively manufactured. The microstructure of these layers was characterized by optical microscopy (OM) and scanning electron microscopy (SEM). X-ray diffraction (XRD) was used for phase analysis, and the mechanical properties were evaluated by tensile and nanoindentation tests. The microstructural observations show epitaxial grain growth within the composite layers, with the elongated grains growing predominantly in the build direction. XRD analysis confirms the successful incorporation of the TiC particles into the 316L matrix, with no unwanted phases present. Nanoindentation results indicate a significant increase in the hardness and modulus of elasticity of the composite layers compared to pure 316L stainless steel, suggesting improved mechanical properties. Tensile tests show remarkable strength values for the 316L-TiC composite samples, which can be attributed to the embedded TiC particles. These results highlight the potential of SLM in the production of multi-layer metal-ceramic composites for applications that require high strength and ductility of metallic components in addition to the exceptional hardness of the ceramic particles.

In this study, Mg-based composites, by the addition of ZnO, Ca₂ZnSi₂O₇, Ca₂MgSi₂O₇, and CaSiO₃ as bioactive agents, were fabricated using friction stir processing. The microstructure and in vitro assessment of bioactivity, biodegradation rate, and corrosion behavior of the resultant composites were investigated in simulated body fluid (SBF). The results showed that during the immersion of composites in SBF for 28 d, due to the release of Ca²⁺ and PO₄ ³⁻ ions, hydroxyapatite (HA) crystals with cauliflower shaped morphology were deposited on the surface of composites, confirming good bioactivity of composites. In addition, due to the uniform distribution of bioceramic powders throughout Mg matrix, grain refinement of the Mg matrix, and uniform redistribution of secondary phase particles, the polarization resistance increased, and the biodegradation rate of composites significantly reduced compared to monolithic Mg matrix. The polarization corrosion resistance of Mg-ZnO increased from 0.216 to 2.499 KΩ/cm² compared to monolithic Mg alloy. Additionally, Mg-ZnO composite with the weight loss of 0.0217 g after 28 d immersion showed lower weight loss compared to other samples with increasing immersion time. Moreover, Mg-ZnO composite with the biodegradation rate of 37.71 mm/a exhibited lower biodegradation rate compared to other samples with increasing immersion time.

Preparation of high purity ruthenium nitrosyl nitrate using spent Ru-Zn/ZrO₂ catalyst was studied, including melting and leaching to obtain potassium ruthenate solution, reduction, dissolving, concentrating and drying to obtain ruthenium trichloride, nitrosation and hydrolysis to obtain ruthenium nitrosyl hydroxide, removing of K⁺ and Cl⁻, and neutralization with nitric acid. The effects of temperature, concentration, time and pH on the yield and purity of intermediates and final product were studied, and the optimum process conditions were obtained. The yield of ruthenium nitrosyl nitrate is 92%, the content of ruthenium in high purity product is 32.16%, and the content of Cl⁻ and K⁺ are much less than 0.005%. The reaction kinetics of ruthenium nitrosyl chloride to ruthenium nitrosyl hydroxide was studied. The reaction orders of Ru(NO)Cl₃ at 40, 55 and 70 °C are 0.39, 0.37 and 0.39, respectively, while those of KOH are 0.16, 0.15 and 0.17, respectively. The activation energy is −2.33 kJ/mol.

The high-temperature requirement for liquid iron smelting via molten oxide electrolysis presents significant challenges. This study investigates the electrochemical reduction of Fe(III) in a novel low-temperature electrolyte, Na₂SiO₃-SiO₂-Fe₂O₃, utilizing cyclic voltammetry and square wave voltammetry techniques. The results show that Fe(III) reduction occurs in two steps: Fe(III)+e⁻→Fe(II), Fe(II)+2e⁻→Fe, and that the redox process of Fe(III)/Fe(II) at the tungsten electrode is an irreversible reaction controlled by diffusion. The diffusion coefficients of Fe(III) in the molten Na₂SiO₃-SiO₂-Fe₂O₃ in the temperature range of 1248–1278 K are between 1.86×10⁻⁶ cm²/s and 1.58×10⁻⁴ cm²/s. The diffusion activation energy of Fe(III) in the molten salt is 1825.41 kJ/mol. As confirmed by XRD analysis, potentiostatic electrolysis at −0.857 V (vs. $\mathrm{O}_{2}/\mathrm{O}_{(\text{complex})}^{2-}$) for 6 h produces metallic iron on the cathode.

In this work, the boron phosphide (BP) was synthesized and used for the adsorptive removal of methylene blue (MB) dye from aqueous solutions. To determine the optimum adsorption conditions, studies were performed by varying parameters of temperature (298–328 K), pH (2–12), contact time (0–120 min), adsorbent dose (0.01–0.20 g/50 mL), and dye concentration (10–50 mg/L). Different isotherm and kinetic models were applied to the adsorption data. The linear correlations coefficient showed that the Langmuir isotherm best fits (R ²=0.9996). The maximum adsorption capacity of BP was obtained as 555.56 mg/g at 55 °C and the removal rate reached up to 84.11%. Additionally, the pseudo-second-order kinetic model described the adsorption process best (R ²=0.9998). The thermodynamic studies represented that the adsorption occurred spontaneously (ΔG _A ^Θ = − 24.90 kJ/mol) and endothermically (ΔH _A ^Θ=16.67 kJ/mol). The results showed that BP is an efficient adsorbent for removing cationic dyes from aqueous solutions.

Massive amounts of low-grade tin middlings have been produced from tin tailings, in which arsenic and tin are worthy to be recycled. Owing to high sulfur content in these tin middlings, a novel self-sulfurization roasting was proposed to transform, separate and recover arsenic and tin in this research. There was no extra curing agent to be added, which decreased the formation of pollutant S-containing gas. The self-sulfurization process involved a two-stage roasting of reduction followed by sulfurization. First in reduction roasting, FeAsS decomposed to FeS and As and the As then transformed to As₄(g) and As₄S₄(g), via which the arsenic was separated and recovered. The arsenic content in the first residue could be decreased to 0.72 wt.%. Accompanied with it, the FeS was firstly oxidized to Fe_1−xS and then to SO₂(g) by the coexisted Fe₂O₃, and finally reduced and combined with the independent Fe₂O₃ to form Fe_1−xS. In the followed sulfurization roasting, the Fe_1−xS sulfurized SnO₂ to SnS(g), due to which tin could be recovered and its content in the second residue decreased to 0.01 wt.%. This study provided an efficient method to separate and recover arsenic and tin from low-grade tin middlings.

Fluidized reduction roasting is an efficient metallurgical technique. However, its application to nickel laterite ore has rarely been reported. In this paper, the effects of reduction temperature, reduction time, CO concentration, and material particle size on the roasting characteristics of ferronickel fluidization reduction were investigated. Combined with X-ray diffraction, scanning electron microscopy-energy dispersive spectrometry (SEM-EDS) characterization, the mineral phases and microscopic morphology of nickel laterite ore and its roasted ores were analyzed in depth. The results indicated that under the condition of a CO/CO₂ ratio of 1:1, a reduction temperature of 800 °C, and a reduction roasting time of 60 min, a nickel-iron concentrate with a nickel grade of 2.10% and an iron content of 45.96% was produced from a raw material with a nickel grade of 1.45%, achieving a remarkable nickel recovery rate of 46.26%. XRD and SEM-EDS analysis indicated that nickel in the concentrate mainly exists in the form of [Fe, Ni], while the unrecovered nickel in the tailings is primarily present in the form of [Fe, Ni] and Ni₂SiO₄ in forsterite. This study established a theoretical foundation for further exploration of fluidized reduction roasting technology.

The electrochemical behavior of Al(III) in urea-1-butyl-3-methylimidazolium chloride-aluminum chloride (urea-BMIC-AlCl₃) ionic liquids, and the effect of potential and temperature on the characterization of cathode products, current efficiency and energy consumption of aluminum electrorefining have been investigated. Cyclic voltammetry showed that the electrochemical reduction of Al(III) was a one-step three-electron-transfer irreversible reaction, and the electrochemical reaction was controlled by diffusion. The diffusion coefficient of Al(III) in urea-BMIC-AlCl₃ ionic liquids at 313 K was 1.94×10⁻⁷ cm²/s. The 7075 aluminum alloy was used as an anode for electrorefining, and the cathode products were analyzed by XRD, SEM and EDS. The results from XRD analysis indicated that the main phase of the cathode products was aluminum. The results from SEM and EDS characterization revealed that the cathode product obtained by electrorefining −1.2 V (vs. Al) was dense and uniform, and the mass fraction of aluminum decreased from 99.61% to 99.10% as the experimental temperature increased from 313 K to 333 K. In this work, the optimum experimental conditions were −1.2 V (vs. Al) and 313 K. At this time, the cathode current efficiency was 97.80%, while the energy consumption was 3.72 kW·h/kg.

The oxidation behavior of ferrovanadium spinel (FeV₂O₄), synthesized via high-temperature solid-state reaction, was investigated using thermogravimetry, X-ray diffractometry, and X-ray photoelectron spectroscopy over the temperature range of 450–700 °C. The results revealed that the oxidation process of FeV₂O₄ can be divided into three stages with the second stage being responsible for maximum weight gain due to oxidation. Three classical methods were employed to analyze the reaction mechanisms and model functions for distinct oxidation stages. The random nucleation and subsequent growth (A₃) kinetic model was found to be applicable to both initial and secondary stage. The third stage of oxidation was consistent with the three-dimensional diffusion, spherical symmetry (D₃) kinetic mode. Both the model-function method and the model-free method were utilized to investigate the apparent activation energy of the oxidation reaction at each stage. It was found that the intermediates including Fe₃O₄, VO₂, V₂O₃, and Fe_2.5V_7.11O₁₆, played significant roles in the oxidation process prior to the final formation of FeVO₄ and V₂O₅ through oxidation of FeV₂O₄.

Gypsum sludge, a hazardous waste generated by the non-ferrous smelting industry, presents a significant challenge for disposal and utilization. To investigate the feasibility of substituting gypsum sludge for limestone as a flux for smelting, the effects of calcium sulfate (CaSO₄) and smelting conditions on oxygen-rich smelting of lead concentrate were studied. The interaction between CaSO₄ and sulfides facilitates the conversion of CaSO₄ into CaO, which is crucial for slag formation. The order of the influence of sulfide minerals on the conversion of CaSO₄ is pyrite > sphalerite > galena. When using gypsum sludge exclusively as the calcium source, under optimal conditions with a CaO/SiO₂ mass ratio of 0.8, an FeO/SiO₂ mass ratio of 1.2, a melting temperature of 1150 °C, an oxygen flow rate of 1.3 L/min, the recovery rates of Pb and Zn in the lead-rich slag reached 85.01% and 95.69%, respectively, with a sulfur content of 2.65 wt%. The As content in the smelting slag obtained by reduction smelting was 0.02 wt%. Resource utilization of gypsum sludge in lead smelting is a feasible method.

The failure characteristics of thermal treated surrounding rocks should be studied to evaluate the stability and safety of deep ground engineering under high-ground-temperature and high-ground-stress conditions. The failure process of the inner walls of fine-grained granite specimens at different temperatures (25–600 °C) was analyzed using a true-triaxial test system. The failure process, peak intensity, overall morphology (characteristics after failure), rock fragment characteristics, and acoustic emission (AE) characteristics were analyzed. The results showed that for the aforementioned type of granite specimens, the trend of the failure stress conditions changed with respect to the critical temperature (200 °C). When the temperature was less than 200 °C, the initial failure stress increased, final failure stress increased, and failure severity decreased. When the temperature exceeded 200 °C, the initial failure stress decreased, final failure stress decreased, and failure severity increased. When the temperature was 600 °C, the initial and final failure stresses of the specimens decreased by 60.93% and 19.77% compared with those at 200 °C, respectively. The numerical results obtained with the software RFPA3D-Thermal were used to analyze the effect of temperature on the specimen and reveal the mechanism of the failure process in the deep tunnel surrounding rock.

To investigate the complex macro-mechanical properties of coal from a micro-mechanical perspective, we have conducted a series of micro-mechanical experiments on coal using a nano-indentation instrument. These experiments were conducted under both dynamic and static loading conditions, allowing us to gather the micro-mechanical parameters of coal for further analysis of its micro-mechanical heterogeneity using the box counting statistical method and the Weibull model. The research findings indicate that the load – displacement curves of the coal mass under the two different loading modes exhibit noticeable discreteness. This can be attributed to the stress concentration phenomenon caused by variations in the mechanical properties of the micro-units during the loading process of the coal mass. Consequently, there are significant fluctuations in the micro-mechanical parameters of the coal mass. Moreover, the mechanical heterogeneity of the coal at the nanoscale was confirmed based on the calculation results of the standard deviation coefficient and Weibull modulus of the coal body’s micromechanical parameters. These results reveal the influence of microstructural defects and minerals on the uniformity of the stress field distribution within the loaded coal body, as well as on the ductility characteristics of the micro-defect structure. Furthermore, there is a pronounced heterogeneity in the micromechanical parameters. Furthermore, we have established a relationship between the macro and micro elastic modulus of coal by applying the Mori-Tanaka homogenization method. This relationship holds great significance for revealing the micro-mechanical failure mechanism of coal.

Landfill leaks pose a serious threat to environmental health, risking the contamination of both groundwater and soil resources. Accurate investigation of these sites is essential for implementing effective prevention and control measures. The self-potential (SP) stands out for its sensitivity to contamination plumes, offering a solution for monitoring and detecting the movement and seepage of subsurface pollutants. However, traditional SP inversion techniques heavily rely on precise subsurface resistivity information. In this study, we propose the Attention U-Net deep learning network for rapid SP inversion. By incorporating an attention mechanism, this algorithm effectively learns the relationship between array-style SP data and the location and extent of subsurface contaminated sources. We designed a synthetic landfill model with a heterogeneous resistivity structure to assess the performance of Attention U-Net deep learning network. Additionally, we conducted further validation using a laboratory model to assess its practical applicability. The results demonstrate that the algorithm is not solely dependent on resistivity information, enabling effective locating of the source distribution, even in models with intricate subsurface structures. Our work provides a promising tool for SP data processing, enhancing the applicability of this method in the field of near-subsurface environmental monitoring.

Cable-stayed bridges have been widely used in high-speed railway infrastructure. The accurate determination of cable’s representative temperatures is vital during the intricate processes of design, construction, and maintenance of cable-stayed bridges. However, the representative temperatures of stayed cables are not specified in the existing design codes. To address this issue, this study investigates the distribution of the cable temperature and determinates its representative temperature. First, an experimental investigation, spanning over a period of one year, was carried out near the bridge site to obtain the temperature data. According to the statistical analysis of the measured data, it reveals that the temperature distribution is generally uniform along the cable cross-section without significant temperature gradient. Then, based on the limited data, the Monte Carlo, the gradient boosted regression trees (GBRT), and univariate linear regression (ULR) methods are employed to predict the cable’s representative temperature throughout the service life. These methods effectively overcome the limitations of insufficient monitoring data and accurately predict the representative temperature of the cables. However, each method has its own advantages and limitations in terms of applicability and accuracy. A comprehensive evaluation of the performance of these methods is conducted, and practical recommendations are provided for their application. The proposed methods and representative temperatures provide a good basis for the operation and maintenance of in-service long-span cable-stayed bridges.

The breakage and bending of ducts result in a difficulty to cope with ventilation issues in bidirectional excavation tunnels with a long inclined shaft using a single ventilation method based on ducts. To discuss the hybrid ventilation system applied in bidirectional excavation tunnels with a long inclined shaft, this study has established a full-scale computational fluid dynamics model based on field tests, the Poly-Hexcore method, and the sliding mesh technique. The distribution of wind speed, temperature field, and CO in the tunnel are taken as indices to compare the ventilation efficiency of three ventilation systems (duct, duct-ventilation shaft, duct–ventilated shaft-axial fan). The results show that the hybrid ventilation scheme based on duct-ventilation shaft–axial fan performs the best among the three ventilation systems. Compared to the duct, the wind speed and cooling rate in the tunnel are enhanced by 7.5%–30.6% and 14.1%–17.7%, respectively, for the duct-vent shaft-axial fan condition, and the volume fractions of CO are reduced by 26.9%–73.9%. This contributes to the effective design of combined ventilation for bidirectional excavation tunnels with an inclined shaft, ultimately improving the air quality within the tunnel.

This paper explores the deformation mechanism and control technology of roof pre-splitting for gob-side entries in hard roof full-mechanized longwall caving panel (LTCC). The investigation utilizes a comprehensive approach that integrates field monitoring, theoretical analysis, and numerical simulation. Theoretical analysis has illuminated the influence of the length of the lateral cantilever beam of the main roof (LCBM) above the roadway on the stability of the gob-side entry behind the panel. Numerical simulations have further revealed that the longer LCBM results in heightened vertical stress within the coal pillar, developed cracks around the roadway, and more pronounced damage to the roadway. Moreover, numerical simulations also demonstrate the potential of roof pre-splitting technology in optimizing the fracture position of the hard roof. This technology significantly reduces the length of the LCBM, thereby alleviating stress concentration in the coal pillars and integrated coal rib while minimizing the destruction of the gob-side entry. Therefore, this manuscript first proposes the use of roof pre-splitting technology to control roadway deformation, and automatically retain the entry within a hard roof LTCC panel. Field implementation has demonstrated that the proposed automatically retained entry by roof pre-splitting technology effectively reduces gob-side entry deformation and achieves automatically retained entry.

This study focused on exploring the specificity of mechanical behavior for completely weathered granite, as a special soil, by consolidated drained triaxial tests. The influences of dry density (1.60, 1.70, 1.80 and 1.90 g/cm³), confining pressure (100, 200, 400 and 600 kPa), and moisture content (13.0%, that is, natural moisture content) were investigated in the present work. A newly developed Duncan-Chang model was established based on the experimental data and Duncan-Chang model. The influence of each parameter on the type of the proposed model curve was also evaluated. The experimental results revealed that with varying dry density and confining pressure, the deviatoric stress–strain curves have diversified characteristics including strain-softening, strain-stabilization and strain-hardening. Under high confining pressure condition, specimens with different densities all showed strain-hardening characteristic. Whereas at the low confining pressure levels, specimens with higher densities gradually transform into softening characteristics. Except for individual compression shear failure, the deformation modes of the specimens all showed swelling deformation, and all the damaged specimens maintained good integrity. Through comparing the experiment results, the strain-softening or strain-hardening behavior of CWG specimens could be predicted following the proposed model with high accuracy. Additionally, the proposed model can accurately characterize the key mechanical indicators, such as tangent modulus, peak value and residual strength, which is simple to implement and depends on fewer parameters.

Engineering shallow, large-span rock tunnels challenges deformation control and escalates construction costs. This study investigates the excavation compensation method (ECM) and its associated technologies to address these issues. Utilizing five key technologies, the ECM effectively modulates radial stress post-excavation, redistributes stress in the surrounding rock, and eliminates tensile stress at the excavation face. Pre-tensioning measures further enhance the rock’s residual strength, establishing a new stability equilibrium. Field tests corroborate the method’s effectiveness, demonstrating a crown settlement reduction of 3–8 mm, a nearly 50% decrease compared to conventional construction approaches. Additionally, material consumption and construction duration were reduced by approximately 30%–35% and 1.75 months per 100 m, respectively. Thus, the ECM represents a significant innovation in enhancing the stability and construction efficiency of large-span rock tunnels, marking a novel contribution to the engineering field.

Laser technology holds significant promise for enhancing rock-breaking efficiency. Experimental investigations were carried out on sandstone subjected to laser radiation, aiming to elucidate its response mechanism to such radiation. The uniaxial compressive strength of sandstone notably decreases by 22.1%–54.7% following exposure to a 750 W laser for 30 s, indicating a substantial weakening effect. Furthermore, the elastic modulus and Poisson ratio of sandstone exhibit an average decrease of 33.7% and 25.9%, respectively. Simultaneously, laser radiation reduces the brittleness of sandstone, increases the dissipated energy proportion, and shifts the failure mode from tensile to tension-shear composite failure. Following laser radiation, both the number and energy of acoustic emission events in the sandstone register a substantial increase, with a more dispersed distribution of these events. In summary, laser radiation induces notable damage to the mechanical properties of sandstone, leading to a substantial decrease in elastic energy storage capacity. Laser rock breaking technology is expected to be applied in hard rock breaking engineering to significantly reduce the difficulty of rock breaking and improve rock breaking efficiency.

In order to study and analyze the stability of engineering rock mass under non-uniform triaxial stress and obtain the evolution mechanism of the whole process of fracture, a series of conventional triaxial compression tests and three-dimensional numerical simulation tests were carried out on hollow granite specimens with different diameters. The bearing capacity of hollow cylindrical specimen is analyzed based on elasticity. The results show that: 1) Under low confining pressure, the tensile strain near the hole of the hollow cylindrical specimen is obvious, and the specimen deformation near the hole is significant. At the initial stage of loading, the compressive stress and compressive strain of the specimen are widely distributed. With the progress of loading, the number of microelements subjected to tensile strain gradually increases, and even spreads throughout the specimen; 2) Under conventional triaxial compression, the cracking position of hollow cylinder specimens is concentrated in the upper and lower parts, and the final fracture mode is generally compressive shear failure. The final fracture mode of complete specimen is generally tensile fracture. Under high confining pressure, the tensile cracks of the sample are concentrated in the upper and lower parts and are not connected, while the cracks of the upper and lower parts of the intact sample will expand and connect to form a fracture surface; 3) In addition, the tensile crack widths of intact and hollow cylindrical specimens under low confining pressure are larger than those under high confining pressure.

In this work, the flow surrounding the train was obtained using a detached eddy simulation (DES) for slipstream analysis. Two different streamlined nose lengths were investigated: a short nose (4 m) and a long nose (9 m). The time-average slipstream velocity and the time-average slipstream pressure along the car bodies were compared and explained in detail. In addition to the time-averaged values, the maximum velocities and the pressure peak-to-peak values around the two trains were analyzed. The result showed that the nose length affected the slipstream velocity along the entire train length at the lower and upper regions of the side of the train. However, no significant effect was recognized at the middle height of the train along its length, except in the nose region. Moreover, within the train’s side regions (y=2.0–2.5 m and z=2–4 m) and (y=2.5–3.5 m and z=0.2–0.7 m), the ratio of slipstream velocity U _max between the short and long nose trains was notably higher. This occurrence also manifested at the train’s upper section, specifically where y=0–2.5 m and z=4.2–5.0 m. Similarly, regarding the ratio of maximum pressure peak-to-peak values Cp – p _max, significant regions were observed at the train’s side (y=1.8–2.6 m and z=1–4 m) and above the train (y=0–2 m and z=3.9–4.8 m).

The complex structure of the bottom of a high-speed train is an important source of train aerodynamic drag. Thus, improving the bottom structure is of great significance to reduce the aerodynamic drag of the train. In this study, computational fluid dynamics (CFD) based on three-dimensional steady incompressible Reynolds-average Naiver-Stokes (RANS) equations and Realizable k-ε turbulence model were utilized for numerical simulations. Inspired by the concept of streamlined design and the idea of bottom flow field control, this study iteratively designed the bogies in a streamlined shape and combined them with the bottom deflectors to investigate the joint drag reduction mechanism. Three models, i.e., single-bogie model, simplified train model, and eight-car high-speed train model, were created and their aerodynamic characteristics were analyzed. The results show that the single-bogie model with streamlined design shows a noticeable drag reduction, whose power bogie and trailer bogie experience 13.92% and 7.63% drag reduction, respectively. The range of positive pressure area on the bogie is reduced. The aerodynamic drag can be further reduced to 15.01% by installing both the streamlined bogie and the deflector on the simplified train model. When the streamlined bogies and deflectors are used on the eight-car model together, the total drag reduction rate reaches 2.90%. Therefore, the proposed aerodynamic kit for the high-speed train bottom is capable to improve the flow structure around the bogie regions, reduce the bottom flow velocity, and narrow the scope of the train’s influence on the surrounding environment, achieving the appreciable reduction of aerodynamic drag. This paper can provide a new idea for the drag reduction of high-speed trains.

Mill vibration is a common problem in rolling production, which directly affects the thickness accuracy of the strip and may even lead to strip fracture accidents in serious cases. The existing vibration prediction models do not consider the features contained in the data, resulting in limited improvement of model accuracy. To address these challenges, this paper proposes a multi-dimensional multi-modal cold rolling vibration time series prediction model (MDMMVPM) based on the deep fusion of multi-level networks. In the model, the long-term and short-term modal features of multi-dimensional data are considered, and the appropriate prediction algorithms are selected for different data features. Based on the established prediction model, the effects of tension and rolling force on mill vibration are analyzed. Taking the 5th stand of a cold mill in a steel mill as the research object, the innovative model is applied to predict the mill vibration for the first time. The experimental results show that the correlation coefficient (R ²) of the model proposed in this paper is 92.5%, and the root-mean-square error (RMSE) is 0.0011, which significantly improves the modeling accuracy compared with the existing models. The proposed model is also suitable for the hot rolling process, which provides a new method for the prediction of strip rolling vibration.