Simulation and modelling of screening processes

Abstract

Screening or sieving is an important unit operation to separate particles on the basis of size by a screening surface which will either accept or reject particles. In spite of its wide application, the understanding of this process is still limited, particularly considering the large number of variables which are related to the screen geometries, operational conditions and particle properties. A better understanding of the process can lead to improvements on the performance and has a great economic significance.

Generally, a screening involves two processes: stratification (or percolation) and the passage of free particles through the apertures, in which a particle always experiences complicated particle-particle and particle-wall interactions. Therefore, it is essential to study screening processes at a particle scale. In this work, numerical models based on the Discrete Element Method (DEM) are developed, which readily provides information at a particle scale that can help understand the process better but hardly be obtained from physical experiments.

Using DEM, different screening processes have been studied. Firstly, particle percolation in a vibrated screen is investigated. The effects of vibration amplitude and frequency on the percolation behaviours are studied in terms of percolation velocity, residence time distribution and radial dispersion. Further analysis of the diffusion coefficient and porosity shows that the percolation behaviours can be related to the dimensionless vibrational velocity (Vb) for different amplitudes and frequencies. Interestingly, it is found that the percolation velocity can be linked to the dimensionless vibrational velocity, and the critical percolation velocity can always be achieved when Vb equals 2, which corresponds to the onset of the fluidized state.

Secondly, a numerical study of particle flow and sieving behaviour in a sieve-shaker is presented, with the sieve-shaker consisting of three sieves in one column. The effects of variables like vibrational amplitude and frequency, volume ratio and sliding and rolling friction coefficients are studied through a series of controlled numerical experiments. Such a process is modelled by reaction kinetics. The sieving performance is then analysed in terms of sieving rate constant (the kinetic rate) for different sized particles at the top, middle and bottom screens, and a mathematical model is proposed to link sieving rate constant with the considered variables, which can satisfactorily estimate the process kinetics.

Finally, the effect of aperture shape, sliding friction coefficient, moisture content and feed rate in different vibration conditions in a vibrated screening process are studied through a series of controlled numerical experiments. The sieving performance is analysed in terms of overall percentage passings of different sized particles and the distribution of percentage passing along the screen deck. The performance of such a screening process is shown to be related to particle flow on the screen such as the structure of particle bed, particle velocities and particle-screen interactions. The sieving behaviour of individual particles is analysed based on the microdynamics information, particularly the particle-screen interactions. On this basis, the probability of a single attempt and the number of attempts for a particle to pass an aperture are modelled which are linked to the macroscopical sieving performance.

The results are useful for developing a fundamental understanding of the effect of operational conditions, particle properties and screen geometries on sieving, which will help design, control and optimise practical processes.