Mechanical Inhibition of Foam Formation via a Rotating Nozzle
Entry by Emily Redston, AP 225, Fall 2011
Work in progress
Mechanical Inhibition of Foam Formation via a Rotating Nozzle, by W. D. Ristenpart, A. G. Bick, E. A. van Nierop, and H. A. Stone, J Fluid Eng-T Asme 133 (4) (2011).
bubbles, capillarity, confined flow, drops, foams, nozzles, rotational flow, two-phase flow
Many industrial processes involve a step where two or more liquid streams are combined in one container. If one of the liquids is poured, sprayed, or dripped into liquid already in the container, air is oftentimes entrained upon impact, and the consequent bubbles form a foamy layer. Generally these foams are an unintended and unwanted by-product of the process. Foams have many deleterious effects because they can (1) interfere with unit operations, (2) decrease process efficiency, (3) increase process time, and (4) lead to additional process defects. Therefore, a great number of commercial chemical additives have been developed to minimize the impact of foams; for example, anti-foaming agents are added to prevent foam formation. Unfortunately, these additive chemicals have several drawbacks: they may contaminate the final product, pose environmental disposal problems, and increase the overall process cost and complexity. Non-chemical strategies are thus desirable. Although some work has suggested that mechanical or ultrasonic vibrations help disrupt foams after they form, there had been no demonstration of a non-chemical technique that prevents foam formation until this paper.
In this paper, the authors present a simple mechanical apparatus that, for appropriate flow rates, significantly reduces the amount of foam generated when a liquid is sprayed into a container. Specifically, they demonstrate a technique to substantially prevent bubble entrainment due to what they refer to as “multidrop” impacts. Multidrop bubble entrainment occurs when two successive drops impact a liquid-air interface within a critical time interval. In earlier work, they demonstrated that the critical time interval is proportional to the time required for an impact crater, formed by the first drop, to close by capillarity (approximately 5 ms for millimeter scale water droplets). The key implication here is that bubble formation, and hence foam formation, can be minimized in the multidrop regime simply by ensuring that no two droplets impact the air-liquid interface at the same location within the critical time interval. Building on this observation, the authors report a design for a rotating nozzle that prevents successive collocated impacts, thereby minimizing bubble entrainment. They demonstrated that a lab-scale prototype can reduce the volume of foam formed by as much as 95% for a given flow-rate, provided the angular velocity of the nozzle is sufficiently high.
The prototype apparatus is shown in Fig. 1. Two plastic circular gears were placed together, one of which contained the nozzle for fluid delivery while the other was attached to a rotating shaft powered by a motor. When the motor was activated, the nozzle thus traced out a circular trajectory. Note that the flow rates used here (q >> 1 ml/min) caused the fluid to exit the nozzle as a jet with velocity on the order 0.1 m/s; the jet rapidly broke up via the Rayleigh-Plateau instability into discrete droplets. Although rotation of the nozzle imparted some angular momentum to the droplets, the large vertical component of the velocity ensured that the droplets impacted the bottom surface rather than the container walls.
The foam-suppressing effect for a sufficiently high angular velocity is demonstrated qualitatively in Fig. 2. Dawn dish soap was chosen as a trial fluid because of its well known foamability, and as expected, a significant volume of foam was generated for a stationary nozzle [Fig. 2(a)a]. In contrast, almost no foam was generated when the nozzle rotated at omega = 7.6 rad/s [Fig. 2(b)b]. We emphasize that both the chemical composition of the fluid and the volumetric flow rate were identical in each trial; the only difference was in the angular velocity of the nozzle. This result demonstrates that rotating the nozzle during fluid addition can suppress the formation of foam.
To quantify the transition between these two limits, we systematically varied the rotation rate of the nozzle for two different flow rates and measured the approximate volume of foam formed after dispensing 30 ml of 2% SDS solution. The normalized foam volumes are plotted versus time for different angular velocities and flow rates in Fig. 3. Several features of the data are noteworthy. First, for zero angular velocity (i.e., a stationary nozzle), the foam is formed at an approximately constant rate and collects at the top of the container. Qualitatively different behavior occurs for non-zero angular velocities, with the dynamics highly sensitive to the magnitude of the angular velocity. For a very slow angular velocity, the rate of foam generation actually increased [Fig. 3(b)b, omega = 1.9 rad/s]. In this case, the rotating nozzle served to more rapidly distribute the foam across the entire liquid/area interface; in contrast, foam was generated by the stationary nozzle in only one location and hence filled the entire area more slowly. At a slightly larger angular velocity, however, the amount of foam generated decreased compared to the stationary nozzle [Fig. 3(a)a, omega = 3.2 rad/s]. Note that the slight oscillations apparent in the curve stem from transient capillary waves that “sloshed” the foam around, resulting in artifactual deviations in the foam volume as obtained via image analysis. The main result, however, is that the total amount of foam formed at a fixed flow rate was reduced by 50% compared to the stationary nozzle even at the slow angular velocity of 3.2 rad/s.
Even more dramatic reductions in the foam volume were observed at higher angular velocities: a 75% reduction was observed for omega = 14.3 rad/s at a flow rate of 10 ml/min. Likewise, at 7.8 ml/min, an 85% reduction was observed for omega = 13.4 rad/s, while a 95% reduction was observed at omega = 20.9 rad/s. We note that the distinct increase in foam suppression at higher angular velocity is consistent with our previous findings 6 because successive drops dispensed from a nozzle moving less than 3 rad/s have increased likelihood of interacting with the craters formed by previously impacting drops, while a nozzle moving at 13 rad/s is less likely to do so.