Effect of particle diameter on agglomeration dynamics in multiphase turbulent channel flows

The present work uses a fully coupled direct numerical simulation-Lagrangian particle tracking solver in conjunction with an interaction energy-based deterministic agglomeration algorithm to determine the effect of particle diameter on the aggregation properties of a wall-bounded, particle-laden channel flow at shear Reynolds number, \u1d445\u1d445\u1d452\u1d452\u1d70f\u1d70f= 180. Three primary particle diameters are considered of relevance to the nuclear industry resembling 200µm - 400µm calcite particles dispersed in water, with a Hamaker constant of 3.8×10-20 J. The simulations are initialized with randomly dispersed particles of numbers calculated to ensure a constant volume fraction ΦP = 10-3. Analysis is focused on elucidating the collision and agglomeration behaviour throughout the channel flow over time. A statistically steady state for collision and agglomeration rate is observed 10 non-dimensional time units after the particles have been injected which persists until at least \u1d461\u1d461∗=50. Results indicate a decrease in particle agglomeration efficiency as diameter is increased, which provides for a reduction in agglomeration rate at large time scales as the particles begin to aggregate and the mean agglomerate diameter increases. Further to this, the normalized number of collisions is similar in all simulations, with the smallest particles showing a slightly increased collision rate. Arguments associated with energy dispersed in collisions are presented to substantiate these findings. Collision rates across the channel are approximately constant with an increase close to the walls which, when normalized by the total number of primary particles, are actually favoured by smaller particles. Finally, agglomeration outcomes after a collision are shown to be more likely towards the channel centreline, since the particle dynamics in this region favour collisions with low relative velocity