Investigation of an FFT-based solver applied to dynamic flowsheet simulation of agglomeration processes

Abstract

The growth of particles due to agglomeration is often mathematically described by population balance equations. The numerical evaluation of these equations and applying new methods to their solution is an area of increasing interest. In this contribution, a new approach for solving the agglomeration population balance model based on a separable approximation of the agglomeration kernel and a fast Fourier transformation is investigated. Its applicability within a dynamic flowsheet simulation of continuous agglomeration processes with complex structures is analysed. A simulation framework Dyssol is used to study the new method and compare it to the well-known fixed pivot technique. Studies have shown that the new approach can provide a more efficient solution if certain constraints on the number of classes and on the separation rank of the agglomeration kernel are met.