Numerical and experimental study of turbulent drop breakup in high­-pressure homogenizers

Sammanfattning: Emulsification is the process of making stable mixture of two or more immiscible substances (typically oil and water). The product of this process, the emulsion, is found in various applications and industries such as food (milk and other dairy products, non-dairy alternatives, mayonnaise, salad dressing), cosmetics and pharmaceuticals (creams, lotions, intravenous medications), chemical (paints, cleaning products), etc. To have a stable mixture of the disperse phase (substance with lower volume fraction) inside the continuous phase (substance with higher volume fraction), one should use a third substance called an emulsifier. The emulsifier forms a layer around the disperse phase drops decreasing their tendency to coalesce with their neighbors. Furthermore, breaking the disperse phase into smaller drops increases the stability of the emulsion. The emulsifier decreases the energy needed for this breakup. High-pressure homogenizers (HPHs) are one of the most effective ways to create more stable emulsions through breaking the disperse phase drops. Drop breakup mechanisms are still debated. But, the majority of the literature agrees on turbulence to be a dominating factor. Therefore, investigating the turbulence inside an HPH geometry is an essential step to improve the understanding of breakup mechanisms in these devices. Investigations of the flow field and drop breakup phenomenon inside HPHs have been limited to experimental (PIV and drop visualizations) and CFD turbulence models. Direct numerical simulation (DNS) resolves the turbulence to the smallest scales (Kolmogorov-scales), providing information which none of the other tools are able to achieve. In this study, a simplified scale-up HPH geometry is designed to provide similar flow conditions as in a real HPH and DNS is performed on this geometry to describe flow field and turbulence properties to the smallest possible spatial and temporal scales. The results of industrially-favored CFD tools are also compared to the DNS results to provide best-practice recommendations. Finally, drop breakup studies are carried out using experiments and numerical simulations and the results are compared. The validation of the numerical results through experiments provided the opportunity to use high-resolution data (particularly the dissipation rate of turbulent kinetic energy field as an important parameter in drop breakup) for further investigations on understanding turbulent drop breakup mechanisms.