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.

To address problems of accuracy and efficiency in automatically reading of float flowmeter at low flow rates under a complex environment in sintering kilns for new energy materials, an innovative YOLOFFM algorithm was proposed. The core improvements include reconstructing the backbone network, enhancing the neck structure, introducing an asymmetric compression decoupling head, and optimizing the loss function, which can significantly improve the efficiency and detection performance of the algorithm. The results show that this YOLOFFM algorithm has accuracy up to 99.15% and a recall rate of 98.69%, significantly reducing the number of model parameters and computational costs. Compared with various advanced algorithm models, YOLOFFM can improve accuracy while reducing the computational cost by more than 90%, fully demonstrating its high efficiency and reliability for new energy materials in a complex environment of sintering kilns.

An experimental study was carried out for mineral processing of a Cu-S ore with grade of MgO at 11.86%, grade of S at 6.81%. In the experiment, a flotation process, consisting of preferential flotation of copper, Cu-S separation and reverse flotation of copper concentrate for removal of magnesium, was adopted, with SG-2 as a copper collector in combination with HD as a depressant for magnesium-containing gangue, producing a standard copper concentrate grading 23.47% Cu and 2.99% MgO, with copper recovery at 81.10%.

The south slope of the Jingxi-Barak mining area of Xinjiang Jinchuan Mining Industry was taken for study. As for the issues of weakening mechanical parameters of slopes after an earthquake and the stability assessment of slopes under aftershocks, a BP neural network model optimized by the crow search algorithm (CSA-BP) was proposed for the inversion of mechanical parameters of slopes after an earthquake. The stability of slopes with dangerous rock mass under aftershocks was evaluated by using the discrete element method. The results show that the CSA-BP model can quantitatively reveal the weakening characteristics of rock masses through the inversion of mechanical parameters. Under aftershock of a magnitude 5 earthquake, significant displacement occurs in the tuffaceous sandstone in the middle and upper parts of the slope, with horizontal (x-direction) displacement far exceeding vertical (z-direction) displacement, indicating that slope instability is dominated by horizontal sliding. The CSA-BP model can accurately identify high-risk zones through a parameter-dynamic coupling mechanism, providing theoretical support for slope protection after earthquakes.

Based on process mineralogy study, a processing technique consisting of atmosphere roasting and leaching with dilute sulfuric acid was proposed to separate a kind of mixed rare earth concentrate containing 48.23% REO. It is found that the phase composition of such concentrate is mainly composed of bastnaesite, monazite, apatite and fluorite, with rare earth elements primarily existing in bastnaesite and monazite. Under air/argon atmosphere roasting, phases of monazite, apatite and fluorite can remain stable, and phase transition is mainly attributed to bastnaesite reaction. In an air atmosphere, bastnaesite is transitioned in the following steps, CeFCO₃→Ce₇O₁₂→Ce₁₁O₂₀→CeO₂, while its transition steps in an argon atmosphere include CeFCO₃→Ce₇O₁₂, Th_0.5Ce_0.5O_1.84→CeOF. After roasting in an air atmosphere at 500 ℃ followed by leaching with dilute sulfuric acid, the leaching rates of F-REO and P-REO can reach 86.67% and 1.70%, respectively, while the process of roasting at 700 ℃ in an argon atmosphere can lead to the leaching rates of F-REO and P-REO at 75.97% and 13.33% respectively. And the leaching residues obtained from above-mentioned two different processes have F-REO and P-REO contents of 6.92%, 26.02% and 11.86%, 21.81% respectively, all dominated by monazite. It is concluded that this processing technique of atmosphere roasting followed by dilute sulfuric acid leaching can effectively achieve separation between bastnaesite and monazite.

In a Hunan polymetallic ore concentrator, the tailings with CaF₂ grade of 26.56% exhibit certain economic value. However, due to the influence of residue reagents in the upstream processing of main tungsten-molybdenum-bismuth minerals and the interference of calcium-bearing gangue, it is extremely difficult to recover fluorite from the tailings. A process consisting of selective regrinding, tailings discarding by high intensity magnetic separation, and fluorite flotation was adopted with a new reagent NH-1 as activator of fluorite, CY-63 as collector of fluorite, SWG+SZY as depressant of silicate and micro-fine gangue, and CYAB as depressant of calcareous gangue. As a result, fluorite concentrate with CaF₂ grade of 90.50%, CaCO₃ content of only 1.35% and CaF₂ recovery of 79.50% was finally reclaimed in the closed-circuit test. It is concluded that such tailings resources can be comprehensively utilized by adopting this process.

Extreme high-speed laser cladding (EHLA) technology was used to prepare Al₂O₃-316L coatings on Ti6Al4V rods. The effects of scanning speed, powder feeding rate on single-track coating quality, and 316L powder content on multi-track coating quality and corrosion resistance were explored. Results show that with the following optimum process parameters, including scanning speed of 250 mm/s and powder feeding rate of 8 g/min, the coating can present superior quality with the 316L powder in a mass fraction of 20%. It is shown that all Al₂O₃-316L coatings exhibit superior corrosion resistance compared to the substrate. As 316L content increases, the corrosion resistance of coatings increases followed by decline. With 316L powder in a mass faction of 20%, the electrochemical impedance of the coating is 2.27 times that of the substrate.

As for the molybdenum ores from Yichun Luming Mine, an m-order grinding kinetics model was used to explore the effects of four grinding aids, such as triethanolamine (TEA), triisopropanolamine (TIPA), polyacrylic acid (PAA) and polycarboxylic ether (PCE), on grinding kinetics of molybdenum ores. The results show that before addition of grinding aids, the proportion of the milled product in the size range of -0.074 mm increases significantly with the increase of the medium charge ratio, and the grinding kinetic parameters of m and k all increase; with the same medium charge ratio, the parameter of m increases and the parameter of k decreases as the grain size of milled product decreases. Comparatively, as for the samples milled with grinding aids of TEA, TIPA, PAA and PCE, the milling products in the size range of -0.074 mm increase by 6.19, 2.40, 3.62 and 1.89 percentage points, respectively. Compared to the system without grinding aids, an addition of TEA can lead to higher values of parameters of k and m, and larger particle size can result in great decline in the k value, while small particle size can bring a gentle k-grain curve, indicating the k value plays a dominant role at the initial grinding stage. After addition of grinding aids, the grains present improved dispersibility with significantly less agglomeration.

A chrysoberyl-type beryllium ore from Xinjiang, containing 0.102% Be with quartz as the dominant gangue mineral, was taken in an experimental study on sulfuric acid leaching. According to the Box-Behnken experimental design principle, a multivariate regression equation was established by using response surface methodology (RSM) for three influencing factors of leaching effect, including liquid-solid ratio, leaching time and leaching temperature, as well as their interactions. Based on the test of significance of regression equation, the optimum process conditions were determined as follows: roasting temperature of 900 ℃, roasting time of 3 h, sulfuric acid concentration of 50%, liquid-solid ratio of 3.2∶1, leaching time of 4.8 h and leaching temperature of 90 ℃. Under these conditions, the Be leaching rate is predicted to be 89.73%, and an average value is 89.10% based on three parallel experiments, indicating that the established model is accurate. A kinetic model of leaching revealed that the leaching process was controlled by interfacial chemical reactions, with an apparent activation energy at 50.42 kJ/mol.

Surfactants of Span80 and Tween80 were adopted to emulsify diesel in an experiment. Based on the studies on effects of diesel microemulsion on the concentrate recovery rate by graphite flotation and fuel-saving efficiency, the optimal conditions for preparing diesel microemulsion were determined as follows: emulsifier's hydrophilic-lipophilic balance (HLB) value of 9, n-butyl alcohol as a co-emulsifier, and emulsifier and co-emulsifier in a mass ratio of 2∶1. Under these conditions, the diesel microemulsion can be prepared with an average particle size of 78.82 nm and uniformly dispersed, presenting excellent stability. Flotation experiments showed that the usage of diesel microemulsion could save 28.66% of oil. And the mechanism of interaction between diesel microemulsion and graphite reveals that diesel microemulsion can significantly reduce the surface tension of the pulp, thus increasing the contact angle of graphite, leading to a higher flotation recovery rate compared to the adoption of conventional diesel.