The project of transition from open-pit to underground mining in Panzihua Iron Mine was taken for study. Based on the analysis of relationship between storage and discharge in the stope, it is predicted that the accumulated water in open pit will be maximally 5 m deep in the event of a 50-year extreme rainfall, which has little influence on the stability of crown pillar. Numerical simulation of seepage in the case of fault occurring at the bottom of pit shows that fault zones are the main leakage channel of crown pillar. Some special anti-seepage measures can be taken to effectively reduce the leakage of crown pillar.

Based on an established three-dimensional analytical model for arching stress of cemented backfill, a calculation formula for vertical load of cemented backfill was obtained. According to a model of four-sided fixed roof and a criterion for ultimate tensile strength of rock mass, a method to determine the thickness of horizontal pillar for safe extraction was obtained. The application practice of horizontal pillar extraction in 612 stope of a lead-zinc mine shows that the vertical stress of cemented backfill calculated with this analytical model is basically consistent with the numerical simulation results, and no plastic zone occurs in the roof in numerical simulation, which verifies that the reserved roof thickness determined by calculation meets the requirement of safe production. The results of stoping process show that the maximum cumulative displacement near the monitoring point is lower than the displacement limit, and the reserved roof tends to be stable. It is concluded that the determined thickness of horizontal pillar is reliable for safe extraction.

Experimental studies on comprehensive recovery of valuable resources in iron-dressing tailings with TFe grade of 15.70% were conducted by adopting processes consisting of Cu-S bulk flotation, high intensity magnetic separation to reclaim Fe minerals, roasting reduction of Fe minerals, and shaking table gravity separation and flotation separation to recover barite and mica. As a result, the experiment produced an iron concentrate with TFe grade of 60.17%, S content of 0.11% and iron recovery of 66.74%, a sulfur concentrate grading of 47.86% S at 66.58% recovery, a copper concentrate grading 19.79% Cu at 31.24% recovery, a barite rough concentrate with BaSO₄ content of 86.96% and recovery of 9.15%, and a mica concentrate with K₂O content of 6.66% and recovery of 6.63%, respectively. In this way, comprehensive utilization of tailings can thus be realized.

Mineral processing tests were carried out for a low-grade copper ore from South America containing 0.027% Mo and 0.45% Cu. After optimization of the reagent regime, a close-circuit test by adopting a flowsheet of Cu-Mo bulk flotation plus Cu/Mo separation resulted in a copper concentrate grading 26.93% Cu at 81.96% recovery, and a molybdenum concentrate grading 25.32% Mo at 76.24% recovery. Therewith, copper and molybdenum resources in the lean ore can be effectively recovered.

To solve problems of high power-consumption and difficulty in load prediction of a ball mill, a comprehensive detection system based on non-magnetic sensing was proposed. The non-magnetic coupling sensing technology is used to monitor the mill load without being affected by the magnetism of materials therein. Combined with the low-power cylinder vibration detection unit, which can more accurately reflect the actual load characteristics, the vibration data can be transmitted to the signal acquisition unit via wireless transmission, through which the difficulty for signal acquisition unit to take electric power from cylinder can be solved. By integrating the mill sound collection data, the loading characteristics of a mill can be predicted adopting the edge algorithm unit aided by in-depth learning. Then, the prediction can be fed back to the PLC controller for real-time adjustment of ore feeding and ball-adding, so as to realize the closed-loop control of the system. This system can greatly improve the milling efficiency, stabilize the production indicators, and reduce the energy consumption. The trial test result in an ore dressing plant has verified that this system demonstrates a satisfying effect in detecting mill load.

Flotation technique was introduced to reclaim copper resource from a copper smelting slag with Cu grade of 2.70%. With sodium sulfide as the regulator, the scavenger middling was collected for flotation and its tailings were classified into a size range of +20 μm to return, together with the cleaning concentrate of middling, to the quick flotation stage. With this optimized flowsheet, a stage of scavenging can be dislodged and the mud content in the middling can be reduced. Finally, a closed-circuit test produced a copper concentrate grading 20.59% Cu at 91.75% recovery, and the tailings with Cu grade reduced to 0.26%.

The kinetics and thermodynamics of oleate ion adsorption on the surface of fluorapatite and dolomite were studied by adopting TOC measurement and molecular simulation. The study on kinetics of adsorption shows that with HEDP as the depressant, the adsorption rate of oleate ions on fluorapatite is greater than on dolomite, and the adsorption on both minerals conforms to the quasi-second-order kinetic model. Furthermore, according to the study on adsorption thermodynamics, the characteristics of oleate ion adsorption on the inner pore surface of both fluorapatite and dolomite, on (001) surface of fluorapatite and (104) surface of dolomite nearly conform to Langmuir model. For identical specific surface area, the adsorption amount of oleate ions on dolomite is higher than that on fluorapatite, indicating that there are more active sites per unit surface area on dolomite than on fluorapatite.

Flotation tests were carried out for a fluorite-calcite symbiosis ore containing 31.09% CaF₂ and 53.23% CaCO₃, with the predominately valuable mineral of fluorite, and the dominantly gangue minerals of calcite, quartz and feldspar. After being milled to a fineness of -0.074 mm 75%, the ore was processed with acidized sodium silicate and EM-318 as depressants and EM-OL-3 as collector. Adopting a close-circuit flowsheet consisting of two stages of roughing, one stage of scavenging and eight stages of cleaning, a fluorite concentrate grading 97.26% CaF₂ at 86.55% recovery can be obtained. Qualified as an acid-grade fluorite powder (with grade FC-97A) according to the Industry Standard for Ferrous Metallurgy (YB/T 5217—2005 (Fluorspar)), this fluorite concentrate can be used as the raw material in high-end fluorine chemical industry.

The crystal structure and surface properties of minerals and the mechanism of interaction between minerals and flotation agents were systematically analyzed by means of quantum chemical calculation and molecular dynamics simulation. Meanwhile, the flotation mechanism of kyanite minerals and quartz was discussed. The results show that the differences in the surface properties of kyanite, andalusite, sillimanite and quartz determine the differences in floatability, and citric acid can increase the floatability difference between kyanite minerals and quartz. The simulation result of computational chemistry is consistent with the experimental result of flotation.

Aiming at problems of high mud content in tin tailings from Russia, finely disseminated particle size of cassiterite, and influence of sulfur and iron content on tin recovery rate, a combined process consisting of sulfide ore flotation, cassiterite preconcentration, and cassiterite flotation was adopted in a beneficiation test. It is found that the dominant tin mineral in tailings is cassiterite, and sulfide minerals therein include pyrite, pyrrhotite among others, while the dominant gangue includes quartz and tourmaline. The effect of sulfide ore and iron ore on subsequent cassiterite flotation can be reduced by adopting cassiterite preconcentration process for waste discarding. With the cassiterite grading 0.31% Sn, the beneficiation test with a combined process consisting of sulfide ore flotation, cassiterite preconcentration and cassiterite flotation can produce a concentrate grading 5.60% Sn at 34.01% recovery.