A beneficiation experiment for super enrichment was carried out with the low-grade niobium ore from western zone of Bayan Obo mine. In the experiment, a process of magnetic separation followed by gravity concentration for tailings discarding was firstly adopted for pre-concentration, and the obtained rougher concentrate grading 2.61% Nb₂O₅ was subjected to a reverse flotation for desulfuration, followed by a direct flotation with CYN-7 as a collector and CYD-2 as a depressant for recovering Nb minerals, resulting in efficient separation between Nb and other gangue minerals. It is found that the raw ore grading 0.06% Nb₂O₅, with a grinding fineness of－0.074 mm 85% is treated by magnetic separation and gravity concentration for tailings discarding, followed by a reverse flotation for desulfuration, and a direct flotation consisting of one stage of roughing and three stage of cleaning. As a result, a Nb concentrate grading 22.37% Nb₂O₅ at 33.64% recovery can be finally obtained.

Effects of quenching and tempering at different temperatures on the microstructure and hardness of annealed W6Mo5Cr4V2 steel were studied by means of OM, SEM, XRD analysis and hardness test. The results show that the microstructure of quenched W6Mo5Cr4V2 steel is mainly composed of martensite, MC carbide rich in V, W and Mo, and M₆C carbide rich in W and Mo, as well as residual austenite. As tempering temperature rises, the carbide in the microstructure of W6Mo5Cr4V2 steel grows in size, and the hardness decreases after an initial increase. After quenching at 1 160 ℃ plus three times of tempering at 540 ℃, W6Mo5Cr4V2 steel has its carbide become smaller and evenly distributed, and its residual austenite almost eliminated, showing the hardness up to 64.2HRC.

To improve the accuracy of fracture recognition in borehole images, a borehole fracture recognition approach for open-pit mine was proposed. First, borehole images of an open-pit mine is obtained with an intelligent borehole inspection camera, and then data augmentation is performed by using random cropping and image flipping, while median filtering is used for noise reduction and images are converted to grayscalere, so as to eliminate noise and reduce computational complexity. Next, spatial attention and channel attention mechanisms are integrated into the U-Net model to improve the semantic segmentation model for fractures, forming an AU-Net model, which can enhance the model′s ability to extract features from both overall and local image information. Experimental results show that compared to the original U-Net model, the AU-Net model can achieve lower loss and higher accuracy in the fracture recognition dataset by borehole imaging. Specifically, the mean intersection over union is improved by 4.38 percentage points, up to 82.34%, bringing better image segmentation effect.

For reducing energy consumption while ensuring weld joint quality, the welding process parameters for electron beam welding of TU2 oxygen-free copper were optimized with SYSWELD simulation software, and a finite element model with appropriate boundary conditions was established for a 30-mm-thick TU2 copper plate. A heat source model was determined based on the actual situation of electron beam welding, and simulation analysis was made by imputing varied energy consumption value in the welding process, with the obtained simulation results compared with actual welding results. It is shown that with a double-ellipsoidal heat source input of 1260 J/mm and a 3D Gaussian heat source input of 5 000 J/mm, a full penetration weld with the lowest energy consumption can be achieved. The simulation results show that the obtained morphology of molten pool is almost the same with that in the actual welding, proving that the welding parameters set in the simulation analysis are feasible for TU2 welding. It is concluded that this study can provide a reliable reference for optimizing the parameters in welding of TU2 oxygen-free copper.

A process mineralogy study was conducted for a niobium-bearing ore in eastern Africa. Based on the study results, the raw ore was ground to－0.10 mm (100% passing size), and then pretreated with a process consisting of desliming, iron removal and zircon flotation for reducing the influence of slime, magnetite, and zircon on flotation performance. Under weakly acidic conditions, a closed-circuit niobium flotation test was carried out, with water glass and aluminum sulfate as a combined depressant, lead nitrate as an activator, and a self-developed chelating collector GYX₂, producing a niobium concentrate grading 26.87% Nb₂O₅ at 88.31% recovery, with a corresponding recovery of 80.86% from the raw ore. It is concluded that the niobium minerals in the ore can be effectively recovered.

A new short-process preparation technique for battery-grade iron phosphate was explored with iron powder and phosphoric acid as main raw materials. The effects of iron powder dissolution mechanism, iron phosphate reaction conditions, and mother liquor recycle on product indicators were investigated. The results show that the dissolution rate of iron powder can reach 97.92% at a temperature of 70 ℃, with phosphoric acid at a concentration of 20%, iron and phosphorus at a ratio of 1/3, and iron powder at a size of 150 μm；after 1 h precipitation at 100 ℃, with an addition of hydrogen peroxide at 110% of the theoretical amount, an iron phosphate with D₅₀ of 2 μm was prepared with the precipitation rate of 98.86%. The obtained iron phosphate has stable crystal form with uniform crystal grains, and the contents of elemental impurities are significantly lower than the requirements specified in the HG/T 4701—2021 standard. The precipitated mother liquor in the process can be recycled, leading to actualization of low-carbon and green production.

To improve the surface quality of high-end plates by descaling and cleaning process, an experimental study was carried out with different process routes that are determined based on the combination of Steel Eco Descaling (SED) technology and other surface pretreatment processes. Based on the analysis of surface appearance, surface roughness, surface micromorphology and surface energy spectrum of the plates treated by different processes, an optimal descaling process is finally determined, consisting of SED, soaking with roller brush cleaner, polishing with scouring pad buffering wheel, and high-pressure flushing. After cleaning treatment with this process, the plate sample is overall in a uniform metallic color, with significantly reduced surface roughness and oxygen content. It is concluded that after treatment with this combined process, surface quality of the third-generation advanced high-strength steel (AHSS) plates for automotive can be effectively improved.

With tartaric acid as a leaching agent, an experiment was conducted to investigate the effects of different leaching conditions on the leaching rates of sodium and titanium from red mud. The process parameters were optimized by orthogonal experiments, and the leaching mechanism was then analyzed in combination with the study of leaching kinetics and the properties of leaching residues. It can be seen from the leaching test and orthogonal experiment that with tartaric acid at a concentration of 1.00 mol/L, leaching temperature of 90 ℃, liquid-solid ratio of 8 mL/g, and leaching time of 60 min, the leaching rates of sodium and titanium from red mud can reach 93.50% and 76.66%, respectively. XRD analysis of red mud before and after tartaric acid leaching shows that the sodium in the leaching solution mainly comes from cancrinite and sodium hydroxide in the Bayer process, while the titanium mainly comes from perovskite and secondary titanium that dissolves and re-precipitates on the surface of red mud during the Bayer process. The leaching kinetics analysis shows that both alkali removal and titanium leaching from red mud conform to an unreacted shrinking core model, and the processes of alkali removal and titanium leaching are mainly controlled by interfacial chemical reactions.

Under the current production conditions, the fluorite concentrate obtained from the associated fluorite resources in a mine of Hunan Province has its processing indicators undesirable and fluctuating widely. In view of this technical bottleneck, a new processing technique was developed to recover such kind of fluorite ore for improving the quality of resources. In this technique, a principle of alkaline environment followed by acidic environment was adopted in the cleaning stage, and also the middlings obtained from the cleaning stage are returned to be processed again. As a result, a full-process closed-circuit test yielded 13.67% high-grade fluorite concentrate grading 96.11% CaF₂ at 60.13% recovery, and 3.93% low-grade fluorite concentrate grading 85.08% CaF₂ at 15.34% recovery. The comprehensive concentrate has a CaF₂ grade up to 93.65%, with total recovery of 75.47%. It is concluded that this low-grade associated fluorite ore can be efficiently and comprehensively recovered using this technique.

An intermittent sedimentation test was performed for vanadium-titanium magnetite tailings to obtain settling curves of flocculated tailings, and then the solid flux in sedimentation by flocculation was calculated by Coe-Clevenger equation and material balance calculation. An innovative Φ38 m heavy-duty and efficient thickener is designed to have a unit area processing capacity of 1 000 kg/(m²·h). The industrial application of this thickener in dewatering of vanadium-titanium magnetic tailings show that with dry ore processing capacity of 1 100 t/h and the feed concentration of 20%－25%, its underflow has concentration of 45% and the overflow has solid content less than 200 mg/L. It is concluded that such thickener can solve the technical problems in the production faced by enterprises, and also bring enterprises enhanced economic and environmental benefits.