Breakup cascade in gas filament

TL;DR

This study uncovers the physics of gas filament breakup, revealing a self-similar fragmentation process that produces a characteristic bubble size distribution relevant to natural and industrial processes.

Contribution

It introduces a deterministic model for gas filament breakup, demonstrating a universal power-law bubble size distribution distinct from liquid ligament fragmentation.

Findings

Filament breakup produces a $d^{-3/2}$ bubble size distribution.

The shape at breakup determines the initial bubble size distribution.

Turbulence influences initial conditions but not size distribution.

Abstract

Despite its importance in both geophysical and industrial contexts, the inertial fragmentation of gas filaments has received much less attention than their liquid counterparts. Yet, gas filaments produce the smallest bubble sizes, which drive gas dissolution, critical to ocean-atmosphere exchange such as carbon dioxide and oxygen, as well as marine aerosols emission, serving as nuclei for cloud condensation and ice particle production. Here, we unravel the fundamental physics governing the splitting of a single filament in a model geometry by combining numerical simulations, laboratory experiments and theory. We show that the splitting of a single filament generates a power-law bubble size distribution following $d^{-3/2}$ with $d$ the volume equivalent bubble diameter, suggesting the existence of a self-similar breakup mechanism, absent in liquid ligament fragmentation. We propose a deterministic model, based on the capillary fragmentation of a filament with power-law shape, which quantitatively captures the bubble size distribution. We demonstrate that the filament shape at breakup sets the size distribution of a first generation of bubbles. This distribution is then reproduced at smaller and smaller scales by latter breakups in a self-similar manner. The $d^{-3/2}$-distribution coincides with the size distribution of small bubbles observed in dilute turbulent flow, such as below breaking waves. We argue that the turbulent bubble size distribution observed in nature arises as the superposition of many individual filament splittings. The turbulence nature of the flow only sets the initial conditions of each splitting dynamics, and play no role in the bubble size selection.