【作者】 王晓；

【导师】 李济顺；

【作者基本信息】 河南科技大学 ， 机械制造及其自动化， 2022， 博士

【摘要】 矿石破碎加工是磨矿过程的重要环节,磨矿工艺决定了矿石的生产效率和加工质量。矿石自身的破碎特性直接影响矿石破磨效果,是磨矿工艺参数制定的关键因素。矿石破碎试验和磨矿过程有相当大的区别,致使采用试验得到的矿石破碎特性制定的磨矿工艺难以取得良好效果,是磨矿工艺优化和磨机选型急需解决的难题。离散元仿真可以模拟矿石破碎过程,其准确性取决于矿石模型参数与实际矿石破碎特性的符合程度。提高离散元建模精度始终是矿石破碎仿真的重点及难点。颗粒黏结模型是矿石破碎离散元仿真的基础,但目前关于颗粒黏结模型参数的确定尚缺少系统而有效的方法,难以实现对矿石破碎行为的准确模拟。因此,深入分析矿石破碎特性、磨矿工艺与离散元仿真模型之间的关系,研究离散元颗粒黏结模型的精准建模及仿真方法,对提高磨矿工艺水平、磨矿效率及磨矿粒度控制具有重要意义和工程实用价值。本文针对磨矿仿真中矿石破碎模型参数确定的难点问题,采用离散元方法构建以基础颗粒和黏结键为关键参数的颗粒黏结模型,从矿石强度、矿石粒级、冲击能量3个维度深入系统地研究矿石破碎行为,并利用矿石落重试验对颗粒黏结模型的适用性和精度进行分析及验证,探讨可准确反映矿石破碎特性的颗粒黏结模型建模方法。论文的主要研究内容和创新性成果如下:（1）研究了用于离散元破碎仿真的矿石颗粒黏结模型建模技术。应用EDEM软件建立了颗粒黏结模型,进行了不同冲击能量下的落重破碎仿真。通过对比落重试验和仿真得到的矿石破碎粒级质量分数,确定了模型的适用性。（2）完成了3种粒级铜矿石、金矿石和铁矿石的落重试验。分析了落重试验得到的破碎粒级质量分数及其分布特征,讨论了不同矿石强度变化的特点;探讨采用对t₁₀进行冲击能量归一的矿石破碎强度指数t_n10表征矿石破碎强度,而t_n10与冲击能量无关,找到了矿石破碎强度和颗粒黏结模型黏结键之间关系的表达方法。（3）研究了颗粒黏结模型基础颗粒种类和粒径的确定方法。分析了矿石落重试验的破碎质量分数分布,讨论了基础颗粒的种类和粒径确定的原则。选用3种基础颗粒即能满足建模需求,基础颗粒粒径的选择和矿样粒级相关;以3种基础颗粒建立的模型具有灵活性,且建模效率较高。建模时采用3种直径的球形基础颗粒、颗粒直径均匀分布、最大颗粒直径小于落重试验质量分数最小粒级的基础颗粒,可较好地解决建模精度与建模效率的矛盾。（4）研究了颗粒黏结模型基础颗粒数量的确定方法。设计了5组基础颗粒数量进行建模和破碎仿真,以落重试验和仿真破碎质量分数的方差作为优化目标,得到了最优的基础颗粒体积比,基础颗粒的最优体积比和矿样粒级密切相关。应用基础颗粒体积比确定基础颗粒数量,并通过优化得到了不同尺寸颗粒黏结模型的最优基础颗粒体积比,可降低破碎仿真质量分数的误差,提高颗粒黏结模型的精度。（5）探讨了颗粒黏结模型黏结键的特征及表征方法。以3种粒级的铜矿石为例进行落重破碎仿真,得到了与落重试验质量分数一致的黏结键参数。通过分析不同黏结键参数的破碎仿真结果,提出了应用等效临界变形表征颗粒黏结模型黏结键的临界刚度和临界应力。黏结键强度决定了颗粒黏结模型的精度,应用黏结键等效临界变形可以准确表征黏结键强度。等效临界变形随模型尺寸的增大而减小,与实际矿石强度的变化趋势相同。等效临界变形能够有效表征颗粒黏结模型黏结键的破碎特征,按照等效临界变形设置黏结键参数可以提高建模精度。（6）研究了颗粒黏结模型黏结键的确定方法。以铜矿石破碎仿真为基础,研究了矿石破碎强度和黏结键等效临界变形的关系,给出了颗粒黏结模型黏结键等效临界变形和矿石破碎强度指数t_n10的线性方程,提出了基于矿石破碎强度和基础颗粒尺寸的黏结键确定方法。再以金矿石和铁矿石为例,构建了3种粒级金矿石和铁矿石颗粒黏结模型,仿真表明破碎仿真粒级质量分数与落重试验完全一致,验证了颗粒黏结模型关键参数确定方法的正确性。可见,依据矿石强度指数t_n10确定黏结键等效临界变形,能够保证黏结键参数的准确性,显著提高建模效率和模型精度。本文采用离散元仿真和落重试验相结合的方法,系统研究了颗粒黏结模型基础颗粒参数和黏结键参数对矿石破碎离散元仿真精度的影响。给出了基础颗粒的种类和粒径确定的原则,探讨了矿石破碎强度和黏结键等效临界变形的关系,提出了基于矿石破碎强度指数t_n10的黏结键确定方法,为建立矿石颗粒黏结模型的通用建模方法奠定了理论基础,为实现矿石破碎过程的精准仿真提供了方法支撑。更多还原

【Abstract】 Ore crushing is an important part of the grinding process,and the grinding process determines the production efficiency and processing quality of the ore.The crushing characteristics of the ore directly affect the crushing and grinding,which is the key factor in the formulation of grinding process parameters.There is a great difference between the ore crushing test and the grinding process,so that it is difficult to achieve good results by the grinding process based on the ore crushing characteristics obtained from the test,which is an urgent problem to be solved in grinding process optimization and mill selection.The ore crushing process can be simulated by discrete element method,but its accuracy depends on the agreement between the ore model parameters and the actual ore crushing characteristics.Improving the accuracy of discrete element modeling is always the key and difficult point of ore crushing simulation.Bonded particle model is the basis of ore crushing discrete element simulation,but at present,there is still a lack of systematic and effective methods to determine the parameters of bonded particle model,so it is difficult to simulate ore crushing behavior accurately.Therefore,the relationship between ore crushing characteristics,grinding process and discrete element simulation model is deeply analyzed,and the accurate modeling and simulation method of discrete element bonded particle model are studied.It is of great significance and practical value of engineering to improve the grinding process,grinding efficiency and control of grinding particle size.In this paper,in order to solve the difficult problem of determining the parameters of ore crushing model in grinding simulation,the discrete element method was used to construct a bonded particle model with basic particles and bonds as key parameters.The ore crushing behavior was studied deeply and systematically from the three dimensions of ore strength,ore size and impact energy,and the drop weight test was used to verify the applicability and accuracy of the bonded particle model.The modeling method of bonded particle model was discussed which can accurately reflect the crushing characteristics of ore.The main research contents and innovative achievements of this paper are as follows:（1）The modeling technology of ore bonded particle model for discrete element crushing simulation was studied.The bonded particle model was modeled by using EDEM software,and the drop weight crushing simulation under different impact energy was carried out.By comparing the mass percentage of broken ore of different size obtained by drop weight test and simulation,the applicability of the model was verified.（2）The drop weight tests of three sizes of copper ore,gold ore and iron ore had been completed.The mass percentage and distribution characteristics of broken ore obtained from drop weight test were analyzed,and the characteristics of strength changes of different ores were discussed.In this paper,the ore crushing strength index t_n10,which normalizes the impact energy of t₁₀,was used to characterize the ore crushing strength.The expression method of the relationship between the ore crushing strength and the bond of the bonded particle model was found.（3）The method of determining the basic particle type and particle size of the bonded particle model was studied.The crushing mass percentage distribution from ore drop weight test was analyzed.The types of basic particles and the principle of determining particle size were discussed.The selection of three kinds of basic particles can meet the needs of modeling,and the selection of basic particle size is related to the size of ore samples.The model based on three kinds of basic particles has flexibility and high efficiency of modeling.The contradiction between modeling accuracy and modeling efficiency can be well solved by using spherical basic particles of three diameters,uniform distribution of particle diameter,and the maximum particle diameter is less than the minimum particle size in the drop weight test.（4）The method of determining the number of basic particles in the bonded particle model was studied.Five groups of basic particles were designed for modeling and crushing simulation.Taking the variance of mass percentage from drop weight test and simulation crushing as the optimization objective,the optimal basic particle volume ratio was obtained.The optimal volume ratio of basic particles is closely related to the size of ore samples.The number of basic particles is determined by the basic particle volume ratio,and the optimal basic particle volume ratio of the bonded particle model with different sizes is obtained by optimization,which can reduce the error of the mass percentage of crushing simulation and improve the accuracy of the bonded particle model.（5）The characteristics and characterization methods of bonds in bonded particle model were discussed.Taking three kinds of copper ores as examples,the drop weight crushing simulation was carried out,and the bond parameters were obtained,which are consistent with the mass percentage of the drop weight test Based on the analysis of the crushing simulation results of different bond parameters,the equivalent critical deformation was proposed to characterize the critical stiffness and critical stress of the bond in the bonded particle model.The bond strength determines the accuracy of the bonded particle model,and the bond strength can be accurately characterized by the equivalent critical deformation of the bond.The equivalent critical deformation decreases with the increase of the model size,which is the same as the change trend of the actual ore strength.The equivalent critical deformation can effectively characterize the crushing characteristics of the bond of the bonded particle model,and setting the bond parameters according to the equivalent critical deformation can improve the modeling accuracy.（6）The method of determining the bond of particle bonding model was studied.Based on the crushing simulation of copper ore,the relationship between ore crushing strength and bond equivalent critical deformation was studied.The linear equations of bond equivalent critical deformation and ore crushing strength index t_n10 of bonded particle model were given.A bond determination method based on ore crushing strength and basic particle size was proposed.Taking gold ore and iron ore as an example,three sizes of bonded particle models of gold ore and iron ore were constructed.The simulation results show that the particle size mass percentage of the crushing simulation is completely consistent with the drop weight test,which verifies the correctness of the method for determining the key parameters of the bonded particle model.It can be seen that determining the equivalent critical deformation of the bond according to the ore strength index t_n10 can ensure the accuracy of the bond parameters and significantly improve the modeling efficiency and model accuracy.In this paper,the effects of basic particle parameters and bond parameters of bonded particle model on the accuracy of ore crushing discrete element simulation were studied systematically by means of combining discrete element simulation and drop weight test.The types of basic particles and the principle of determining particle size were given.The relationship between ore crushing strength and equivalent critical deformation of bond was discussed,and a bond determination method based on ore crushing strength index t_n10 was put forward,which lays a foundation for establishing a general modeling method of ore bonded particle model and provides method support for accurate simulation of ore crushing process.更多还原