From air-sea gas exchange, oil pollution, to bioreactors, the ubiquitous fragmentation of bubbles/drops in turbulence has been modeled by relying on the classical Kolmogorov-Hinze paradigm since the 1950s. This framework hypothesizes that bubbles/drops are broken solely by eddies of the same size, even though turbulence is well known for its wide spectrum of scales. Here, by designing an experiment that can physically and cleanly disentangle eddies of various sizes, we report the experimental evidence to challenge this hypothesis and show that bubbles are preferentially broken by the sub-bubble-scale eddies. Our work also highlights that fragmentation cannot be quantified solely by the stress criterion or the Weber number; The competition between different time scales is equally important. Instead of being elongated slowly and persistently by flows at their own scales, bubbles are fragmented in turbulence by small eddies via a burst of intense local deformation within a short time.

No abstract

No abstract

A vertical water tunnel facility has been constructed to study the dynamics of turbulent multiphase flow. The new system features several unique designs that allow us to study bubble deformation and breakup in strong turbulence: (i) The mean flow can be adjusted to balance the rising velocity of buoyant bubbles/droplets so that they can stay in the view area for an extended period of time. (ii) Turbulence is generated and controlled using a 3D-printed jet array that can fire 88 random high-speed momentum jets with the individual jet velocity of up to 12 m/s. This component allows us to attain turbulence with a high energy dissipation rate (≥0.1 m2/s3), which is orders-of-magnitude higher than most of the existing turbulent multiphase flow facilities. (iii) Turbulence generated in the test section is nearly homogeneous and isotropic, and the turbulent fluctuations are also decoupled from the mean flow. The resulting turbulence intensity can be varied between 20% and 80% with the speed of the mean flow at around 0.2 m/s. (iv) This system has an octagonal test section that allows six cameras to image and reconstruct the 3D shape of deforming bubbles/droplets in turbulence. The same set of cameras was also used for tracking tracers in the surrounding turbulent flow. Both the reconstruction and particle tracking were completed using our in-house codes that were parallelized to run on high-performance computing clusters efficiently.

No abstract