Floating soap bubbles are often a sight to behold, particularly due to the plethora of colors they demonstrate as light gets reflected from the soap film. However, soap bubbles (or films) are inherently ephemeral, and burst soon after they are formed – often to the disappointment of the mesmerized observer. Interestingly, this bursting phenomenon itself is rich in physics, and has fascinated fluid dynamicists for nearly two centuries. Despite the long history of studying this bursting phenomenon, several key questions are still unanswered; in particular, what role, if any, the ambient plays in the physics of bursting. In this talk, I will revisit this age-old problem, with an emphasis on studying the influence of the ambient on the physics of bursting soap bubbles. I will show that high-speed imaging reveals new scalings in the dynamics of bursting films in an ambient, which can be further explained through scaling analysis and numerics. The findings of this talk are not only of fundamental importance, but are also relevant in understanding, and thus mitigating, oil spills in oceans and the spread of respiratory diseases via aerosols.

In many applications, such as airlifts or bubble columns, bubbles are injected into a liquid to enhance the transfers between a gas and a liquid. The buoyancy-driven motions of the bubbles generate an intense agitation of both phases. Various experimental investigations involving isolated bubbles and homogeneous swarms of rising bubbles allow us to determine the statistical properties of this agitation. On the one hand, the fluctuations of the bubble velocity are controlled by the instability of the individual bubble wakes and therefore weakly depend on the gas volume fraction. On the other hand, the fluctuations of the liquid velocity are almost uncoupled from those of the bubbles and combine two contributions of different natures: the first corresponds to the local spatial inhomogeneities of the flow in the vicinity of the bubbles while the second results from the flow instabilities that develop when the Reynolds number is large enough. Separating these two contributions is a very hard task with moving particles, it becomes however practicable with fixed ones. For that reason, we carried out experimental investigations of the flow trough a random array of spheres, which was shown to produce the same velocity fluctuations as a swarm of rising bubbles. The liquid velocity can thus be decomposed by using both time and spatial averagings. The statistical and spectral description of each contribution makes clear their respective roles in the dynamics of the streamwise and the transversal fluctuations. These two contributions, which have to be account for in the modeling of bubble-induced agitation, also leads to the existence of two distinct mixing mechanisms that will be described in the last part of this talk.