Numerical simulations of two-component granulation: Comparison of three methods

We compare three numerical methodologies for the solution of the population balance in two-component granulation. These methods are (a) direct solution of the discrete bicomponent aggregation equation; (b) constant-number Monte Carlo (cNMC); and (c) direct quadrature method of moments (DQMOM). We apply these methodologies to bicomponent aggregation with a kernel that depends both on size and composition using various initial conditions. We find that the cNMC method is in excellent agreement with the direct discrete solution in all cases. The DQMOM method is highly accurate when the kernel is independent of composition. With kernels that depend on the composition of granules, the accuracy of DQMOM drops and appears to be sensitive to the details of the seed distribution.