Rising droplets in a centrifugal field: A way to avoid interfacial contamination in liquid-liquid flow

Investigations of the settling of liquid droplets in another liquid phase have to deal with the potential contamination of the interfaces with residual impurities. It is shown in this Letter that the use of a centrifugal field to generate the drop settling may overcome this issue: the drop terminal velocity and shape are reached at much shorter time scales than that taken by contaminant adsorption from the bulk to the interface, so it is possible to obtain data for clean interfaces. In this study, using a centrifuged cell, drop formation, velocity, and shape have been characterized in the inertial regime. By comparing these quantities to theoretical models and by performing complementary direct numerical simulations with clean interfaces, it is proved that interfacial contamination can be disregarded in the experiments. Such an experiment can therefore be used to establish results of reference for clean interfaces. For instance, a correlation for the drop deformation of settling droplets in clean liquid-liquid systems is obtained for a wide range of parameters, as a function of the sole Weber number. It emphasizes that clean droplets are less distorted than clean bubbles.

Figure 1

Drop formation and settling in the rotating cell, for (a) a nearly spherical and (b) an ellipsoidal droplet.

Figure 2

Scaling law, Eq. (1), and experimental points [20] for the drop size at detachment; the four arrows shift the $B{o}_D$ of cases $E_{21-24}$, containing surfactants, by considering the same surface tension as for a clean interface.

Figure 3

Cases [20] (a) $E_{25}$, (b) $E_{16}$, (c) $E_5$, and (d) $E_{14}$. Evolution of the Reynolds number $Re$ and aspect ratio $\chi$ as a function of the distance traveled by the droplet in the experiments, compared to the predictions from the force balance model of Eq. (2), and DNS results without surfactants.

Figure 4

Drop aspect ratio $\chi$ versus $We$: several numerical [20, 40, 41] and experimental results [20] (without surfactants), compared to the correlation proposed, Eq. (3), for clean liquid-liquid droplets. Inset: same plot including experiments with surfactants at large concentration (cases $E_{21-22}$ [20]) and previous experimental results [47, 48, 49].