Skip to content Skip to footer

Influence of the dispersed phase fraction on experimental and predicted drop size distributions in breakage dominated stirred systems, 2012

— Authors: Maaß, S., Paul, N. und Kraume, M. —

Chemical Engineering Science, 76: 140-153 (link)

The effect of the dispersed phase fraction on the evolving drop size distribution in different low viscous agitated liquid/liquid systems was investigated. The analysis focused on the drop breakage phenomena by hindering the coalescence completely. Therefore, polyvinyl alcohol concentrations were used around three times higher than the critical micelle concentration.

The measured drop sizes were increasing with increasing dispersed phase fraction. As coalescence was completely hindered and also the measured dispersion viscosity showed no influence on the dispersed phase hold-up, the size increase is proposed to be a result of turbulence hindering.

The influence of the dispersed phase fraction on the drop sizes in breakage dominated systems was well reproduced with population balance equation (PBE) simulations. The used breakage models require a turbulence damping factor (1+φd), which is used in most of the common models. Summarizing the various PBE simulations we can conclude that drop sizes in systems with different dispersed phase fractions can be easily predicted, if the model parameters are fitted to one set of experiments studying the same physical system. The change of the solvent was successfully simulated with outstanding results for two of the three further investigated organics.

The used Weber correlations were also able to reproduce the linear interdependency between the drop size and the dispersed phase fraction. Unfortunately, every change in the dispersed phase needed new parameter estimation. As at least three out of four different liquid/liquid systems were predicted with excellent results, the PBE is proposed as a more robust tool which gives additionally information about the transient behavior of the system. Therefore, PBE should be used rather than the classical correlations widely used in academics and industries.