Modelling particle deagglomeration in a batch homogenizer using full CFD and mechanistic models

Modelling of titanium dioxide deagglomeration in the mixing tank equipped with a high shear impeller is presented in this study. A combination of computational fluid dynamics with population balance was applied for prediction of the final particle size. Two approaches are presented to solve population balance equations. In the first one, a complete population balance breakage kinetics were implemented in the CFD code to simulate size changes in every numerical cell in the computational domain. The second approach uses flow field and properties of turbulence to construct a mechanistic model of suspension flow in the system. Such approach can be considered as an attractive alternative to CFD simulations, because it allows to greatly reduce time required to obtain the results, i.e., the final particle size distribution of the product. Based on experiments shattering breakage mechanism was identified. A comparison of the mechanistic model and full CFD does not deviate from each other. Therefore the application of a much faster mechanistic model has comparable accuracy with full CFD. The model of particle deagglomeration does not predict a very fast initial drop of particle size, observed in the experiment, but it can predict, with acceptable accuracy, the final particle size of the product.