Numerical behavior of a multiscale aggregation model—coupling population balances and discrete element models

The aggregation-rate kernel of shear-induced aggregation introduced by Smoluchowski (1917) is accurately reproduced by a discrete element model to investigate the behavior of a multiscale approach that couples discrete element simulations concurrently to a population balance model. The analyzed model comprises nucleation, growth and aggregation. From the analysis of the discrete element model an estimation for the coefficient of variation of the simulation results is derived and validated. The results of the multiscale simulations show that, in general, only a certain fraction of the total particle volume can be predicted and also that, dependent on the growth term, a biased deviation of the total volume exists. With the investigated multiscale-modeling framework the need for an analytical aggregation-rate kernel is at least in principle obsolete. Hence, the limitations of the modeling assumptions necessary to derive analytical kernels may be avoided in the future.

► An analytical aggregation-rate kernel is accurately reproduced by DE simulations.
► The discrete element model is concurrently coupled to a population balance model.
► The errors that occur in the resulting multiscale framework are analyzed in detail.
► Additionally, we derived an estimate for the accuracy of the underlying DE simulation.