Kinetic Studies On Differential Floatability And Floating Rate In Flotation Of Huize Lead-Zinc Sulphide Ores

Through batch flotation tests in laboratory, the floatability and floating rate of galena, sphalerite and pyrite from Huize lead-zinc ores are studied in this subject. The rule of differential floatability and floating rate in flotation process is summarized. Asynchronous flotation technology is further discussed, and the flotation process is simulated by methods of regression analysis and artificial intelligence technology.The floatability and floating rate of sulphide ores with no collector and using kaium ethyl xanthate or sodium diethyldithiocarbamate as collector are studied systematically through batch flotation. In addition, regulator influences of Na2S, Pb(NO3)2, ZnSO4, CUSO4and CaO on mineral flotation behavior are studied. Results show that the floatability of pyrite is similar to galena, and different from sphalerite significantly in the condition of pH=7.5, KEX2.4×10-4mol/L and ZnSO41.5×10-4mol/L. From the general rule, galena usually floats fast, with floating rate gradually fall in the flotation process. However, sphalerite and pyrite float slow, with floating rate fall first and then rise in the flotation process.Mineral multi-rate model is proposed combined with the asynchronous flotation technology features of Huize lead-zinc sulphide ores, which is based on classical first-order kinetic model. The classic first-order model and second-order kinetic model in traditional flotation models and multi-rate model are chosen to establish flotation kinetic models. Results show that the correlation coefficient between fitted value and test data of cumulative recovery are all0.999for galena, sphalerite and pyrite. Thus, multi-rate flotation model can simulate the flotation process well.The distribution of flotation rate constant for Huize sulphide ores is analysed according to method of integral rehabilitation. Results show that K value is distributed in a band of0.75～1.27min-1for most galena,0.83～1.33min-1for most sphalerite and0～2.07min-1evenly for pyrite.The effect of particle size on galena flotation kinetics is studied through using fuzzy logic. Tests show that flotation recovery increases initially, and then decreases with increasing particle size, while flotation rate decreases gradually. Variance analysis showed that only flotation rate was influenced by particle size significantly. Simulation analysis shows that fuzzy logic system designed by MATLAB can simulate flotation process of galena with different particle size well.A new method is presented to evaluate mineral flotation characters of floatability and floating rate comprehensively, which is based on membership and membership function in fuzzy math. In this way, mineral flotation character is quantized. Applying new method, flotation characters of galena, sphalerite and pyrite are analysed in condition of pH=7.5, KEX2.0×10-4mol/L and ZnSO41.5×10-4mol/L. Results show that the floatability order is galena> pyrite＞＞sphalerite.Asynchronous flotation tests of mixed minerals of galena-pyrite and galena-sphalerite-pyrite are carried out based on individual mineral flotation. Results of galena-pyrite flotation using KEX as collector show that the concentrate I is mainly composed of galena(73.42%) with less pyrite (37.50%). However, the concentrate II is mainly composed of pyrite (51.93%) with less galena (22.95%). Results of galena-sphalerite-pyrite flotation using KEX as collector and ZnSO4as modifier show that the recovery of galena, sphalerite, pyrite in the first process is84.00%,3.48%,43.71%, respectively. And in the second process, the recovery of galena, sphalerite, pyrite is14.92%,5.05%,43.85%, respectively. The technology realizes that galena and pyrite float synchronously and galena-pyrite and sphalerite float asynchronously.Methods of regression analysis and artificial intelligence technology are used to simulate flotation process of sulphide ores. Results show that regression model based on SPSS could analyze influences of operating variables on flotation index, but the prediction accuracy is poor. Flotation index is predicted by neural network model well, but it could not express the mechanism between operating variables and flotation index. On this basis,"platform of flotation dynamics research for Huize lead-zinc sulphide ores" and "platform of flotation process simulation for Huize lead-zinc sulphide ores" are designed to achieve a friendly exchange of human-machine information.