A comparative analysis of particle tracking in a mixer by discrete element method and positron emission particle tracking

Characterisation of particle flow using Positron Emission Particle Tracking (PEPT) is based on tracking the position of a single particle in a dynamic system. Recent developments in PEPT have facilitated tracking multiple particles aiming at improvements in data representation. Nevertheless for systems with a wide residence time distribution and/or dead zone, the conditions for getting representative data which could reflect the bulk behaviour of the powders need to be analysed and specified. In the present work, an attempt is made to simulate PEPT experiments for a paddle mixer using Discrete Element Method (DEM), with a view to investigate the effect of increasing the number of tracers on their time-averaged velocity distribution and whether it can represent the data on whole population of particles. The time averaged velocity distribution of the individual tracer particles (resembling simulated PEPT) is obtained and compared with the time averaged data on entire particle population. The DEM results indicate that for the investigated paddle mixer, it takes 251. s for one tracer to travel adequately in all the active space of the system. The instantaneous tracer velocity fluctuates around the average value obtained for all the particles, suggesting that the average tracer velocity is adequately representative of the average particle velocity in the system. The data of the PEPT experiment with one tracer with those of DEM with one tracer are in good agreement; however, DEM simulation suggests that increasing the number of tracers in the paddle mixer system does not influence the average velocity distribution. Furthermore, the velocity for all particles in the DEM shows a smooth distribution with a peak frequency of the velocity distribution that is lower than PEPT and DEM tracer. When tracking a single tracer in DEM or PEPT, it may not be detected to have zero velocity at any instant of time, whilst the data for all particles show that about 0.3% of particles are stagnant.