When is a bubble a bubble?: The subtleties of cavitation threshold detection at MHz frequencies

Abstract

Detecting the onset of cavitation induced by short-pulse ultrasound at MHz frequencies is tricky business, for one must be positioned to observe the genesis and evolution of a single micron-sized bubble lasting as short as one acoustic cycle. The observation is rarely accomplished directly. Rather, investigators often rely on indirect evidence of cavitation activity manifested by physical effects such as noise (a.k.a. passive cavitation detection), light emission, and mechanical damage. These techniques are noise limited, and investigators frequently measure a threshold for ‘‘detectable’’ cavitation. Another approach is to actively probe the medium with a high-frequency acoustic beam, and watch for anomalous backscatter associated with bubble activity. This technique, often termed active cavitation detection, can be quite sensitive, which is part of the problem; transient disturbances in the medium can be mistaken for bona fide cavitation activity. Such errors are often associated with, say, the motion of solid particles in a test liquid that is put into motion by acoustic radiation forces or entrained by acoustic streaming. In all cases, the ability to detect and monitor minute, spurious cavitation activity is limited by noise, clutter, or reverberation. These issues are discussed, examples given, and potential solutions proposed. [Work supported by DARPA.]