(310f) Turbulent Droplet Breakage in a Von Kármán Flow Cell

Breakage of dispersed phase liquid droplets immersed in a second immiscible continuous liquid phase is a common phenomenon in the production of petrochemicals, polymers, metals, foods, and pharmaceuticals.[2, 5] It is also vital for environmental management, such as cleaning up oil spills and cleaning leaking underground storage tanks. Moreover, droplet breakage - and the resultant evolution of the liquid-liquid interfacial area - has a substantial impact on interphase mass transfer and plays an important role in the performance of liquid-liquid reactors and liquid extraction process separation efficiency. Hence, a fundamental understanding of the role of the hydrodynamic phenomena that lead to drop breakage is essential to predict and control interphase mass transport in these systems of interacting liquids.

The development and validation of mathematical models of breakage based upon mechanistic physical understanding of droplet breakage in turbulent flow is challenging not only because of the numerous variables that impact breakage events, but also because of the difficulty acquiring detailed and statistically significant experimental data sets for breakage events carried out under well-controlled conditions.[2, 3, 4] In particular, since droplet breakage strongly depends on the local hydrodynamic environment, droplet breakage experiments should ideally be performed in a homogeneous flow field. Nevertheless, most existing experimental studies of turbulent drop breakage have been performed using flow devices known to produce heterogeneous, non-isotropic mixing environments, such as stirred tanks[1, 6].

In this study, we describe attempts to overcome some important limitations of previous droplet breakage experiments by using high-speed photography to capture thousands of droplet breakage events in a von Kármán swirling flow device designed to (a) generate homogeneous isotropic turbulence in a region surrounding a droplet injection port, and (b) carefully control the size of parent droplets that are introduced one at a time into the breakage zone, which allows control over two of the most important factors impacting droplet breakage, i.e., the turbulence dissipation rate and parent droplet size. By introducing droplets one at a time, large data sets were gathered using canola, safflower, and sesame oils for the droplet phase and water as the continuous phase.

Custom image analysis software was used to determine breakage time, breakage probability, and child droplet size distributions for various turbulence intensities. These droplet breakage characteristics extracted from the analysis of thousands of individual breakage events demonstrate that droplet breakage occurs as a result of competition between stresses that cause and resist droplet deformation, and it is shown that when these competing stresses are equal, the droplets have an equal chance of breaking or retaining their integrity.

For the fluid pairs tested, the Ohnesorge number was small, and the viscous resistance to drop deformation was negligible when compared to the disruptive inertial stresses and cohesive interfacial stress. Such conditions satisfy some assumptions of the classic Coulaloglou-Tavlarides droplet breakage model, and therefore, the experimental results for breakage time and breakage probability were used to compute a breakage rate coefficient, which in turn was compared to the model predictions. For the fluids tested, these model predictions compared well with the experimental results.

Both the breakage probability and the breakage time were found to increase with an increase in parent droplet size, consistent with CT model predictions, particularly for the fluids with high interfacial tension and viscosity (canola and safflower oils), which both exhibit binary breakage with a monomodal breakage distribution function. In contrast, for experiments carried out using sesame oil, which has both low viscosity and low interfacial tension, droplets were observed to undergo greater deformation and the breakage events produced bimodal breakage distribution functions.

Reference:

1. Chatzi, E.G., A.D. Gavrielides, and C. Kiparissides, “Generalized Model for Prediction of the Steady-State Drop Size Distributions in Batch Stirred Vessels,” UTC, (1989).

2. Herø, E.H., N. la Forgia, J. Solsvik, and H.A. Jakobsen, “Single drop breakage in turbulent flow: Statistical data analysis,” Chemical Engineering Science: X, 8, (2020).

3. Karimi, M., and R. Andersson, “Dual mechanism model for fluid particle breakup in the entire turbulent spectrum,” AIChE Journal, 65 (8), (2019).

4. Maaß, S., and M. Kraume, “Determination of breakage rates using single drop experiments,” Chemical Engineering Science, 70, pp. 146–164 (2012).

5. Nachtigall, S., D. Zedel, and M. Kraume, “Analysis of drop deformation dynamics in turbulent flow,” Chinese Journal of Chemical Engineering, 24 (2), pp. 264–277 (2016).

6. Zaccone, A., A. Gäbler, S. Maaß, D. Marchisio, and M. Kraume, “Drop breakage in liquid-liquid stirred dispersions: Modelling of single drop breakage,” Chemical Engineering Science, 62 (22), (2007).