In recent years, the variety of chocolates has steadily increased, often leading to multiple chocolates being produced on one line. As a result, the frequency of so-called flushing processes, where one chocolate is replaced by the other has increased. These processes are often designed inefficiently due to a lack of understanding of the process itself and the influencing process parameters. This thesis focuses on developing numerical models which enable a better understanding of the process, a way of modeling the influence of important process parameters, and a way of optimizing them in the future.
The most detailed model is validated using experimental data. It is also used to determine a dimensionless approach for evaluating the flushing process. The latter enables an efficient comparison of processes using different flushing process parameters. The influence of temperature, mass flow rate and chocolate pairing is investigated.
Insights from the most detailed model are used to develop models with subsequently lower computational requirements. By reducing the considered domain to a straight pipe without the influence of gravity, an analytical solution and an algebraic model are developed. Both enable the simulation of flushing processes within seconds.