Acoustic cavitation at audible frequencies

Abstract

The dynamics of bubble collapse are of great interest in a variety of fields across science and engineering. Amongst these, sonochemistry has shown to be a particularly promising technology in a number of applications, ranging from the degradation of pollutants to the synthesis of valuable chemicals. The ensemble of studies concerned with sonochemistry carried out up to the present day suffer from a lack of exploration of the lower end of the frequency spectrum. In fact, almost no study at all is available at frequencies lower than ∼20 kHz. This has led many researchers to believe the existence of a set of reasons why sonochemistry is practically feasible only in the ultrasonic range of frequency. However, such reasons do not seem to be backed up by solid evidence. The scarce interest in the lower frequency range has had an impact on the number of fundamental studies on the single bubble dynamics, which still remains little.

In an attempt to reduce the gap in the theory, this thesis concerns itself with the dynamics of a single spherical bubble driven by audible sound. Although the ultimate aim is the study of the chemical effect of bubble collapsing at low frequency, much is done in order to elucidate the mechanism of mass and heat transfer first. In fact, lowering the driving frequency has the dramatic effect of causing very large bubble expansions with consequent much higher amounts of the surrounding liquid's vapour intruding in the gas phase. This apparently innocuous effect has a significant impact on the bubble collapse at low frequency, which has been studied only using approximated models. It is here argued that the predictions offered by existing studies are incorrect because of flaws in the basic assumptions of the used theories. A novel model is developed in order to provide more accurate predictions maintaining a reasonable level of computational efficiency. The validity of any discussed reduced order model is inferred by comparison with the results of the accurate equations of the bubble dynamics that, although being the most accurate tool available for simulations, are very time-expensive.

The new proposed model is extended in order to take into account chemical reactions of the gas phase and a large parametric study is carried out across many acoustic amplitudes, frequencies and bubble ambient radii. The production of ammonia is chosen as a case study, and the effect of the driving frequency is explored in relation to the number of molecules produced during one acoustic cycle. All the studies carried out show that for acoustic amplitudes lower than 1.2 atm, production of ammonia is observed only in the lower end of the frequency spectrum.

Some of the theoretical predictions are backed up by an experimental validation that has been carried out on a custom made apparatus designed to take high-speed pictures of inertially collapsing bubbles.