Controlled Multibubble Surface Cavitation

Heterogeneous bubble nucleation at surfaces has been notorious because of its irreproducibility. Here controlled multibubble surface cavitation is achieved by using a hydrophobic surface patterned with microcavities. The expansion of the nuclei in the microcavities is triggered by a fast lowering of the liquid pressure. The procedure allows us to control and fix the bubble distance within the bubble cluster. We observe a perfect quantitative reproducibility of the cavitation events where the inner bubbles in the two-dimensional cluster are shielded by the outer ones, reflected by their later expansion and their delayed collapse. Apart from the final bubble collapse phase (when jetting flows directed towards the cluster’s center develop), the bubble dynamics can be quantitatively described by an extended Rayleigh-Plesset equation, taking pressure modification through the surrounding bubbles into account.

Figure 1

(a) Electron microscopy image of an individual hydrophobic microcavity (diameter $4\mu \mathrm{m}$) etched on a silicon plate acting as gas trap. (b) Sketch of the spatial distribution of the artificial nuclei set on a hexagonal lattice with a pitch $d=200\mu \mathrm{m}$. The nuclei fall into six different classes denoted by the numbers 1 to 6. The distance of the nuclei of class $i$ to the center nucleus “1” is denoted as $r_{1i}$.

Figure 2

Pressure generated by the piezoelectric transducer and recorded with a fibre optic probe hydrophone [18]; the dark curve represents the low pass filtered signal. The pressure pulse is used for triggering the expansion of gas pockets trapped on the patterned substrate.

Figure 3

Time sequence showing the expansion and collapse of a hexagonal cluster of bubbles emerging from a patterned solid surface underwater and triggered by a negative pressure pulse as shown in Fig. 2. The lattice step $d$ is $200\mu \mathrm{m}$. The origin of time is set when the bubble at the center starts to grow. Each snapshot corresponds to an experiment run, the phenomenon being lit by a flash lamp for limiting blurring motion. The slight asymmetry of the bubbles size along the vertical axis in the first picture is due to a nonuniform nucleation time. Indeed, the time needed for the pressure wave traveling from bottom to top (about 1 mm) at the speed of sound (about $1500\mathrm{m}/\mathrm{s}$) is about $0.7\mu \mathrm{s}$. The initial wall velocity of the bubbles being a few meters per seconds, this delay is enough for inducing this small but visible effect.

Figure 4

Time evolutions of the bubble radius for the six spatial locations pointed out in Fig. 1. The size is averaged over all bubbles in the corresponding class. Each data point on one curve has been measured for one experiment, the substrate being removed from the water and dried between successive runs.

Figure 5

Bubbles dynamics predicted by a Rayleigh-Plesset approach accounting for bubble-bubble interaction [Eq. (1)] and corresponding to the same conditions as in Fig. 4.

Figure 6

Comparison between experimental and theoretical evolution of the radius of the bubble at the center ● (class 1) and the most far away ▵ (class 6).