A framework for the simulation of mass crystallization considering the effect of fluid dynamics

The behavior of large-scale crystallizers is strongly affected by the fluid dynamic characteristics of the apparatus like e.g. local supersaturation and velocity profiles. The simulation of this effect is a complex multi-scale and multi-phenomena problem. This contribution presents an approach to solve the coupled problem of crystallization and fluid dynamics by means of software integration. Existing specialized software tools (Fluent and Parsival) are employed for the solution of specific subproblems, namely the solution of population balance models and fluid dynamics. To reflect the phenomena on their characteristic scales different grids are used during the simulation of the respective subproblems. The population balance and crystallization kinetics are formulated in the coarse scale compartments while the fluid dynamics are solved on the fine CFD grid. The problem decomposition needed for the formulation of the subproblems and the proper selection of the compartments are discussed. The approach is validated with a simple model of a crystallization in a tubular reactor which can be solved using reduced methods without introducing a systematic error.