Numerical Study of Particle–particle Collision in Swirling Jets: A DEM–DNS Coupling Simulation
Nan Gui,Jian Ren Fan,Song Chen
DOI: https://doi.org/10.1016/j.ces.2010.02.026
IF: 4.7
2010-01-01
Chemical Engineering Science
Abstract:The particle–particle collisions in swirling jets are studied by a coupling method of discrete element method (DEM, a hard-sphere approach) and direct numerical simulation (DNS). The characteristics of distribution of collision in configuration and velocity spaces are investigated in detail through probability density functions (PDFs) in the generalized coordinates. The dependency of particle–particle collision on turbulence characteristics, such as turbulent kinetic energy (TKE), dissipation rate (TDR), fluctuation, and correlated fluctuations, is studied by exploring the PDFs and the correlations between them. The results show that the spatial distribution of particle–particle collision in swirling jets is highly dependent on the Stokes numbers. For small particles, collision is dominated by the enclosure of bubble vortices whereas for large particles it is mainly determined by the configuration of the flow domain. The distribution of collision in velocity space has corresponding features of dependency on the particle property. Small particles are most probable to collide with each other near zero streamwise velocity within the recirculation zone, whereas large particles are most probable to take collision with their axial velocities close to the inflow velocity of fluid. The dependency of collision on TKE and TDR is fairly complicated. For example, for Stokes number slightly less than unity and far larger than unity, collision is relatively well-correlated to TKE, resulting in an augmented effect of turbulence modulation. It is investigated in detail and the physical mechanisms are well interpreted. Finally, the correlation between the PDF of collision and fluctuations of turbulence indicates that collision probability is positively correlated to the normal components of Reynolds stress tensor, but negatively correlated to the shearing components of Reynolds stress tensor.