Numerical Simulation Of The Hydraulic Overflow Classification Process Of Silicon Carbide Micron Particles

Currently the development trend of modern silicon carbide particle industry is towards fine particle size and narrow size distribution, and the demand of classifier in industry is increasingly high along with this trend. Hydraulic overflow classifiers are wet grading equipments widely used in silicon carbide particles'classification. It is of great significance to take study on hydraulic overflow classifiers and find out their principles and how equipment parameters and operating parameters effect on classification efficiency, which is of greatest importance on structure optimization of the classifier and improvement of classification efficiency.Numerical simulation methods are used in this paper. The Fluent software is utilized to simulate the classification processes of silicon carbide particles in the continuous feed-type, semi-continuous feed type and inner-flood hydraulic overflow classifiers. It is obtained that the flow field information inside the classifier and the influential rules about how equipment parameters and operation parameters could effect on classification performance,which provides a certain theoretical basis on structural optimization of classifiers.1. The k-ωmodel is used to simulate the liquid flow inside the continuous feed hydraulic overflow classifier. The overflow surface is modeled using the volume of fluid multiphase (VOF) model, and the particle classification is described by the stochastic Lagrangian model. Those models above are used to simulate the grading process of the silicon carbide particles (size range 0.5-100μm) in the 1.8m (diameter) continuous feed hydraulic overflow classifier. Then the topics are studied by the calculation cover the effect law of the diameter, the cylindrical part height, the conical part height and the water flow rate on the continuous feed hydraulic overflow classifier's classification performance. The results are as follows:(1)The rules and data about liquid flow field and the particle motion inside the classifier are obtained, and the result show that serious backflow occurs in classifies; the particles'movements in classifiers include up-flow and overflow. Besides, classification of particles occurs mainly in the overflow segment.(2) The results prove that increasing the diameter and reducing the inlet liquid velocity are helpful in particles'settlement and decrease of cut size; with the variation of the height of cylindrical and conical parts, the cut size changes so small that can be ignored.(3) Stokes theory and the simulation results are used to derive the cut size formula of continuous feed hydraulic overflow classifiers. Then, dimensional analysis is conducted to obtain a non-dimensional equation:Meanwhile, the value of m is calculated by Plitt's model. It turns out that the value of m initially increases and then decreases along with the increase of body size, cylindrical part and conical part length, and inlet velocity. That means the four variables above have the optimal value to make the highest classification accuracy.2. The k-ωand Mixture models are utilized to simulate the classification processes of silicon carbide powders (size range 2.47-17.75μm) in the semi-continuous feed hydraulic overflow classifier and the inner launder hydraulic overflow classifier. The diameter of both classifiers is 1.8m. Meanwhile it is calculated that the size distribution of settled particles and overflowed particles of the two classifiers at different classification time. In addition, the influence of water flow rate, the diameter and the cone angle on the semi-continuous feed hydraulic overflow classifier's classification performance are studied. All the results are as follows:(1) Rules about volume fraction spatio– temporal distribution of particles with different sizes in the two classifiers are obtained. It can be found that the volume radial distribution of particles in inner launder hydraulic classifier's overflow section is more uniform than that in outer launder hydraulic classifier's.(2) The results prove that increasing the water flow and reducing classifier's diameter, can reduce the required classification time. However, they will result a wider size distribution of settled particles and overflowed particles, and a smaller recovery of settling particles. The smaller the cone angle is, the greater the capacity is, and the narrower the size distributions of settled and overflowed particles are. It, however, leads to longer required classification time. The inner overflow cone can greatly reduce the classification of the classification time, but make classification performance worse.