Characterization and modelling of dispersion systems with variable viscosity

Promovendus/a: Pavel Krýsa

Promotor(en): Prof. dr. ir. Simon Kuhn, Prof. Miroslav Šoóš

Multiphase systems are ubiquitous in both natural phenomena and industrial processes – from biochemical reactors and chemical manufacturing to food production and pharmaceutical formulation. Understanding how these systems behave is critical for process efficiency, product quality and safety. Such systems may consist of a combination of liquid, gas or solid phases that interact via the exchange of momentum, heat and mass, while phenomena like droplet or bubble breakage and coalescence are of great importance as they strongly influence overall performance. Thanks to the rapid growth of computer power, we can now simulate many of these complex systems on a computer instead of relying only on physical experiments, which can be costly in both time and money.

The primary objective of this thesis was to develop a general and robust computational framework capable of accurately describing multiphase systems with variable dispersed phase viscosity. Such situations occur, for example, in biochemical reactors with living cells, in the mixing of two immiscible liquids or in processes where crystallization starts to occur within a droplet injected into another liquid. To achieve this, a comprehensive model was constructed by combining the Euler-Lagrange approach, which tracks individual bubbles or droplets, with the volume of fluid method, which accounts for the dynamics of the free surface. This combination enabled a detailed description of droplet and bubble motion, while taking into account breakage and coalescence, as well as the influence of the vessel geometry and operating conditions on mass transfer. Extensive experimental work was performed in parallel to the simulations which consisted of measurement of bubble and droplet sizes, as well as volumetric mass transfer coefficients under various process parameters such as impeller speed, dispersed phase viscosity and concentration. The combined model demonstrated strong agreement with experimental results.

As an extension of the liquid-liquid system, the process of spherical crystallization, an attractive technique for producing spherical particles, was also investigated. Experiments revealed the critical parameters governing successful particle formation, including stabilizer concentration, temperature gradients and minimum droplet loading. These findings can serve as an essential guidance for extending the computational model to more complex processes in the future.

In summary, this research establishes a versatile and validated computational framework for predicting the behavior of multiphase systems. By combining rigorous modelling with comprehensive experimental validation, the work provides a foundation for more reliable design, optimization, and scale-up of industrial processes, ranging from biochemical production to advanced materiál synthesis. The results contribute to a deeper understanding of multiphase flow dynamics and offer practical tools for improving efficiency, safety and process control in a wide range of applications.