physics0111133/PLA.tex
1: % M. Ida
2: 
3: \documentclass{elsart}
4: \usepackage{epsfig}
5: 
6: \begin{document}
7: 
8: \begin{frontmatter}
9: 
10: \title{A characteristic frequency of two mutually interacting gas bubbles in an 
11: acoustic field}
12: \author{Masato Ida}
13: \address{Satellite Venture Business Laboratory, Gunma University, \\
14: 1--5--1 Tenjin-cho, Kiryu-shi, Gunma 376--8515, Japan 
15: %\\E-mail: ida@vbl.gunma-u.ac.jp \\
16: %TEL : +81-277-30-1126 \quad FAX : +81-277-30-1121
17: }
18: 
19: \begin{abstract}
20: Transition frequencies of two spherical gas bubbles interacting in an 
21: acoustic field are discussed theoretically. In the present study, 
22: transition frequency is defined as the frequency of 
23: external sound for which the phase difference between a bubble's pulsation 
24: and the external sound is $\pi / 2$. It is shown by a linear theory that a 
25: bubble interacting with a neighboring bubble has three (or fewer) 
26: transition frequencies but only two natural frequencies. This result means that 
27: the bubble has a characteristic frequency besides the natural frequencies.
28: \end{abstract}
29: 
30: 
31: \begin{keyword}
32: Two-bubble dynamics \sep Radiative interaction \sep Natural frequency 
33: \sep Phase reversal
34: \PACS 43.20.+g \sep 47.55.Bx \sep 47.55.Dz
35: \end{keyword}
36: 
37: \end{frontmatter}
38: 
39: It is known that radiative interaction between gas bubbles varies their 
40: natural (or resonance) frequencies, around which the bubbles indicate 
41: resonance response. This variation in the natural frequencies has been the 
42: subject of many studies 
43: \cite{ref1,ref2,ref3,ref4,ref5,ref6,ref7,ref8,ref9,ref10,ref11,ref12,ref13}. 
44: In a case where two spherical gas 
45: bubbles of different radii interact acoustically, a bubble has two (or one) 
46: natural frequencies (see, e.g., \cite{ref2,ref5,ref6,ref8}).
47: 
48: The aim of this Letter is to show that a gas bubble interacting with a 
49: neighboring bubble has an alternative characteristic frequency; we now call 
50: this a transition frequency. In the present work, a transition frequency is 
51: defined as the frequency of an 
52: external sound (the driving frequency) for which the phase difference 
53: between a bubble's pulsation and the external sound becomes $\pi / 2$. It is 
54: well known that, in a single-bubble case, the phase difference between the 
55: bubble and an external sound becomes $\pi / 2$ when the driving frequency is 
56: equal to the bubble's natural frequency \cite{ref14}. However, it is shown in 
57: this Letter that, in a double-bubble case, the number of 
58: transition frequencies differs from the number 
59: of natural frequencies. This result means that the bubbles in the latter 
60: case have a characteristic frequency that differs from the natural 
61: frequency.
62: 
63: First, we derive a mathematical model to use for the present investigation. 
64: This model describes the pulsation of two coupled bubbles. Since we are 
65: interested only in the qualitative natures of the transition frequencies, 
66: we use a classical 
67: theoretical model, called a coupled-oscillator or self-consistent model 
68: \cite{ref10,ref11,ref12}, although models with greater accuracy have been 
69: proposed \cite{ref4,ref6,ref7,ref9}.
70: 
71: A gas bubble immersed in a liquid pulsates when a sound wave is applied. The 
72: sound pressure at the bubble position drives the pulsation. When other 
73: bubbles (named ``bubble 2'' $\sim $ ``bubble $N$'', where $N$ is the total 
74: number of bubbles) exist near the bubble (bubble 1), the sound waves scattered 
75: by the other bubbles also drive the pulsation of bubble 1. Namely, the driving 
76: pressure acting on bubble 1, $p_{{\rm d}\,1}$, is expressed as
77: %
78: \begin{equation}
79: \label{eq1}
80: p_{{\rm d}\,1} = p_{{\rm ex}} + {\sum\limits_{j = 2}^{N} {p_{{\rm 
81: s}\,1j}} } {\rm ,}
82: \end{equation}
83: %
84: where $p_{\rm ex}$ and $p_{{\rm s}\,1j}$ are the sound pressures 
85: of the external sound field and the scattered wave emitted by bubble $j$, 
86: respectively, at the position of bubble 1. When the wavelength of the 
87: external sound wave is much larger than the bubbles' radii and the distances 
88: between the bubbles, $p_{{\rm ex}}$ can be considered spatially uniform. 
89: By assuming that the 
90: sphericity of the bubbles is thoroughly maintained and that the liquid 
91: surrounding the bubbles is incompressible, the scattered pressure can be 
92: estimated with (see, e.g., \cite{ref15})
93: %
94: \begin{equation}
95: \label{eq2}
96: p_{{\rm s}\,1j} \approx \frac{\rho }{{r_{1j}}}\frac{d}{{dt}}(R_j^2 \dot R_j),
97: \end{equation}
98: %
99: where $\rho$ is the density of the liquid, $r_{1j}$ is the distance between 
100: the centers of bubble 1 and bubble $j$, $R_j$ is the radius of 
101: bubble $j$, and the over dot denotes the time derivative.
102: 
103: The radial oscillation of a spherical bubble in an incompressible viscous 
104: liquid is described by the RPNNP equation \cite{ref16}:
105: %
106: \begin{equation}
107: \label{eq3}
108: R_1 \ddot R_1 + \frac{3}{2}\dot R_1^2 - \frac{1}{\rho }p_{w\,1} = - \frac{1}{\rho }p_{{\rm d}\,1} ,
109: \end{equation}
110: %
111: \[
112: p_{w\,1} = \left({P_0 + \frac{{2\sigma }}{{R_{10} }}} \right)\left( {\frac{{R_{10} }}{{R_1 }}} \right)^{3\kappa } - \frac{{2\sigma }}{{R_1 }} - \frac{{4\mu \dot R_1 }}{{R_1 }} - P_0 ,
113: \]
114: %
115: where $P_0 $ is the static pressure, $\sigma$ is the 
116: surface-tension coefficient at the bubble surface, $\mu $ is the viscosity of 
117: the liquid, $\kappa$ is the polytropic exponent of the gas inside the bubbles, 
118: and $R_{10}$ is the equilibrium radius of bubble 1. Substituting 
119: Eqs.~(\ref{eq1}) and (\ref{eq2}) into Eq.~(\ref{eq3}) yields
120: %
121: \[
122: R_1 \ddot R_1  + \frac{3}{2}\dot R_1^2  - \frac{1}{\rho }p_{w\,1}  =  - \frac{1}{\rho }[p_{{\rm ex}}  + \sum\limits_{j = 2}^N {\frac{\rho }{{r_{1j} }}\frac{d}{{dt}}(R_j^2 \dot R_j )} ] .
123: \]
124: %
125: When $N = 2$, this equation is reduced to
126: %
127: \begin{equation}
128: \label{eq4}
129: R_1 \ddot R_1 + \frac{3}{2}\dot R_1^2 - \frac{1}{\rho }p_{w\,1} = - \frac{1}{\rho }[p_{{\rm ex}} + \frac{\rho }{D}\frac{d}{{dt}}(R_2^2 \dot R_2 )],
130: \end{equation}
131: %
132: where $D = r_{12}$ ($ = r_{21}$). Exchanging subscripts 1 and 2 yields the 
133: model equation for bubble 2:
134: %
135: \begin{equation}
136: \label{eq5}
137: R_2 \ddot R_2 + \frac{3}{2}\dot R_2^2 - \frac{1}{\rho }p_{w\,2} = - \frac{1}{\rho }[p_{{\rm ex}} + \frac{\rho }{D}\frac{d}{{dt}}(R_1^2 \dot R_1 )] ,
138: \end{equation}
139: %
140: \[
141: p_{w\,2} = \left( {P_0 + \frac{{2\sigma }}{{R_{20} }}} \right)\left( {\frac{{R_{20} }}{{R_2 }}} \right)^{3\kappa }  - \frac{{2\sigma }}{{R_2 }} - \frac{{4\mu \dot R_2 }}{{R_2 }} - P_0.
142: \]
143: %
144: By assuming that the time-dependent bubble radii can be represented as 
145: $R_1 = R_{10} + e_1$, $R_2 = R_{20} + e_2$, and 
146: $\left| {e_1} \right| \ll R_{10}$, $\left| {e_2} \right| \ll R_{20}$, 
147: Eqs.~(\ref{eq4}) and (\ref{eq5}) are reduced to the following linear formulae:
148: %
149: \begin{equation}
150: \label{eq6}
151: \ddot e_1 + \omega _{10}^2 e_1 + \delta _1 \dot e_1 = - \frac{{p_{{\rm ex}} }}{{\rho R_{10} }} - \frac{{R_{20}^2 }}{{R_{10} D}}\ddot e_2,
152: \end{equation}
153: %
154: \begin{equation}
155: \label{eq7}
156: \ddot e_2 + \omega _{20}^2 e_2 + \delta _2 \dot e_2 = - \frac{{p_{{\rm ex}} }}{{\rho R_{20} }} - \frac{{R_{10}^2 }}{{R_{20} D}}\ddot e_1,
157: \end{equation}
158: %
159: where
160: %
161: \[
162: \omega _{j0}  = \sqrt {\frac{1}{{\rho R_{j0}^2 }}\left[ {3\kappa P_0  + (3\kappa  - 1)\frac{{2\sigma }}{{R_{j0} }}} \right]} 
163: \quad {\rm for} \;\; j = 1,2,
164: \]
165: %
166: are the partial natural (angular) frequencies of the bubbles, and
167: %
168: \[
169: \delta _j  = \frac{{4\mu }}{{\rho R_{j0}^2 }}
170: \quad {\rm for} \;\; j = 1,2,
171: \]
172: %
173: are the damping coefficients. (In general, the damping coefficients are 
174: determined by the sum of viscous, radiation, and thermal damping \cite{ref17}; 
175: we however use the above setting to simplify the following discussion.)
176: 
177: The system of equations (\ref{eq6}) and (\ref{eq7}) corresponds to that 
178: derived by Shima when $\sigma = 0$, $\delta _1 = 0$, and $\delta _2 = 0$ 
179: \cite{ref2}, and to that by Zabolotskaya when $\sigma = 0$ \cite{ref5}. 
180: This kind of system of differential equations, called the 
181: coupled-oscillator model or self-consistent model, has been used repeatedly 
182: to analyze acoustic properties of coupled bubbles 
183: \cite{ref2,ref5,ref8,ref10,ref11,ref12,ref13}, and has 
184: been proved to be a useful model that can provide a qualitatively accurate 
185: result despite its simple form.
186: 
187: Let us analyze in detail the transition frequencies of the linear coupled 
188: system. By assuming that 
189: the external sound pressure at the bubble position is written in the form of 
190: $p_{{\rm ex}}  =  - P_a \sin \omega t$, a harmonic 
191: steady-state solution of the linear coupled system is given as
192: %
193: \begin{equation}
194: \label{eq8}
195: e_1  = K_1 \sin (\omega t - \phi _1 ) ,
196: \end{equation}
197: %
198: where
199: %
200: \begin{equation}
201: \label{eq9}
202: K_1  = \frac{{P_a }}{{R_{10} \rho }}\sqrt {A_1^2  + B_1^2 } ,
203: \end{equation}
204: %
205: \[
206: \phi _1  = \tan ^{ - 1} \left( {\frac{{B_1 }}{{A_1 }}} \right) ,
207: \]
208: %
209: with
210: %
211: \begin{equation}
212: \label{eq10}
213: A_1 = \frac{{H_1 F + M_2 G}}{{F^2  + G^2 }} ,
214: \end{equation}
215: %
216: \begin{equation}
217: \label{eq11}
218: B_1 = \frac{{H_1 G - M_2 F}}{{F^2  + G^2 }} ,
219: \end{equation}
220: %
221: \[
222: F = L_1 L_2  - \frac{{R_{10} R_{20} }}{{D^2 }}\omega ^4  - M_1 M_2 ,
223: \]
224: %
225: \[
226: G = L_1 M_2  + L_2 M_1 ,
227: \quad
228: H_1  = L_2  + \frac{{R_{20} }}{D}\omega ^2 ,
229: \]
230: %
231: \[
232: L_1  = (\omega _{10}^2  - \omega ^2 ) ,
233: \quad
234: L_2  = (\omega _{20}^2  - \omega ^2 ) ,
235: \]
236: %
237: \[
238: M_1  = \delta _1 \omega ,
239: \quad
240: M_2  = \delta _2 \omega .
241: \]
242: %
243: Here the solution for only bubble 1 is shown. The phase difference 
244: $\phi _1$ becomes $\pi /2$ when
245: %
246: \begin{equation}
247: \label{eq12}
248: A_1  = 0 .
249: \end{equation}
250: %
251: It should be noted that a case in which both $A_1$ and $B_1$ become zero does 
252: not exist. From Eqs.~(\ref{eq10}) and (\ref{eq11}), one obtains
253: %
254: \[
255: A_1^2  + B_1^2  = \frac{{H_1 ^2  + M_2 ^2 }}{{F^2  + G^2 }}\left( {\frac{{P_a }}{{R_{10} \rho }}} \right)^2 .
256: \]
257: %
258: The numerator of this equation always has a nonzero value, since 
259: $M_2  > 0$; this result denies 
260: the existence of a case where both $A_1 = 0$ and $B_1 = 0$ are true. Also, 
261: it should be noted that $F^2 + G^2$ appearing in the denominator of 
262: Eq.~(\ref{eq10}) always has a nonzero value. When $G = 0$, $F$ is reduced to
263: %
264: \[
265: F = - \frac{{M_2 }}{{M_1 }}L_1^2 - \frac{{R_{10} R_{20} }}{{D^2 }}\omega ^4 - M_1 M_2 .
266: \]
267: %
268: This has a nonzero, negative value because $M_2 L_1^2 /M_1 \ge 0$, 
269: $R_{10} R_{20} \omega ^4 /D^2 > 0$, and $M_1 M_2 > 0$. This result means that 
270: no case exists where both $F = 0$ and $G = 0$ are true. As a consequence, 
271: Eq.~(\ref{eq12}) is reduced to
272: %
273: \begin{equation}
274: \label{eq13}
275: H_1 F + M_2 G = 0 .
276: \end{equation}
277: %
278: In the following, we analyze this equation.
279: 
280: When the damping terms in Eq.~(\ref{eq13}) are negligible (but exist), one can 
281: easily obtain the solution for this equation. By assuming that 
282: $M_1 \approx 0$ and $M_2 \approx 0$, one obtains
283: %
284: \[
285: H_1 F \approx 0 .
286: \]
287: %
288: This equation can be rewritten into two independent equations:
289: %
290: \begin{equation}
291: \label{eq14}
292: F \approx L_1 L_2  - \frac{{R_{10} R_{20} }}{{D^2 }}\omega ^4 = 0
293: \end{equation}
294: %
295: and
296: %
297: \begin{equation}
298: \label{eq15}
299: H_1 = L_2  + \frac{{R_{20} }}{D}\omega ^2 = 0 .
300: \end{equation}
301: %
302: Equation (\ref{eq14}) is the same as the theoretical formula given in 
303: Refs.~\cite{ref2,ref5} to derive the natural frequencies of a double-bubble 
304: system; 
305: namely, the transition frequencies given by this equation correspond to the 
306: natural frequencies. This equation predicts the existence of two natural 
307: frequencies,
308: %
309: \begin{equation}
310: \label{eq16}
311: \omega _{1 \pm }^2  = \frac{{\omega _{10}^2  + \omega _{20}^2  \pm \sqrt {\left( {\omega _{10}^2  - \omega _{20}^2 } \right)^2  + 4\omega _{10}^2 \omega _{20}^2 \frac{{R_{10} R_{20} }}{{D^2 }}} }}{{2\left( {1 - \frac{{R_{10} R_{20} }}{{D^2 }}} \right)}} ,
312: \end{equation}
313: %
314: and is symmetric; namely, it exchanges subscripts 1 and 2 (or 10 and 20) in 
315: this equation to reproduce the same equation. This means that the two 
316: bubbles have the same two natural frequencies. One of the natural 
317: frequencies converges to $\omega _{10}$ and the other to $\omega _{20}$ for 
318: $D \to \infty$, and the higher natural frequency rises and the lower one falls 
319: as $D$ decreases \cite{ref2,ref18}.
320: 
321: The solution of Eq.~(\ref{eq15}) given for the first time is
322: %
323: \begin{equation}
324: \label{eq17}
325: \omega _1^2  = \frac{{\omega _{20}^2 }}{{1 - R_{20} /D}} .
326: \end{equation}
327: %
328: This converges to $\omega _{20}^2$ for $D \to \infty$, and increases 
329: as $D$ decreases. In contrast to Eq.~(\ref{eq14}), Eq.~(\ref{eq15}) is 
330: asymmetric ($H_1 \ne H_2$); namely, this serves to break the symmetry of 
331: the natural frequency mentioned above. The transition frequency given by 
332: Eq.~(\ref{eq17}) is not the 
333: natural frequency because it cannot be given by the natural frequency 
334: analysis performed in Refs.~\cite{ref2,ref5}. (Even the other models used in, 
335: e.g., \cite{ref6,ref8} give only two natural frequencies.)
336: 
337: The results given above show that, when the damping effect is negligible and 
338: the radii of bubbles are not equal, the bubbles have three asymmetric 
339: transition frequencies; one of these, for $D \to \infty$, converges to the 
340: partial natural frequency of a corresponding 
341: bubble, while the remaining two converge to that of a neighboring bubble. 
342: (In the following, the former transition frequency is called 
343: ``fundamental transition frequency'' (FTF), and the latter two are called 
344: ``sub transition frequencies'' (STFs).) One of the STFs always increases as 
345: bubbles approach each 
346: other. The other STF decreases (increases) and the FTF increases (decreases) 
347: when the partial natural frequency of the bubble is higher (lower) than that 
348: of the neighboring bubble.
349: 
350: Figure \ref{fig1} shows the transition frequencies for $\mu \approx 0$ as a 
351: function of $l = D/(R_{10} + R_{20})$. Other parameters are set to 
352: $\rho = 1000$ kg/m$^3$, $P_0 = 1$ atm, $\sigma = 0.0728$ N/m, and 
353: $\kappa = 1.4$. Equations (\ref{eq16}) and (\ref{eq17}) are used in plotting 
354: those graphs. The radius of bubble 1 is fixed to 
355: 10 $\mu$m and that of bubble 2 varies from 5 $\mu$m to 10 $\mu$m. As 
356: discussed above, three transition frequencies changing with $D$ appear in 
357: each graph, except for the case of $R_{10} = R_{20}$, where only one 
358: decreasing transition frequency appears. (When the two bubbles have the same 
359: radii and pulsate in phase, Eq.~(\ref{eq6}) is reduced to
360: %
361: \[
362: \left( {1 + \frac{{R_{10} }}{D}} \right)\ddot e_1  + \omega _{10}^2 e_1 + \delta _1 \dot e_1  =  - \frac{{p_{{\rm ex}} }}{{\rho R_{10} }} .
363: \]
364: %
365: This equation has only one transition frequency \cite{ref5,ref10} of
366: %
367: \begin{equation}
368: \label{eq18}
369: \omega _1^2  = \frac{{\omega _{10}^2 }}{{1 + R_{10} /D}} ,
370: \end{equation}
371: %
372: which converges to $\omega _{10}^2$ for $D \to \infty$ and decreases 
373: with decreasing $D$.) It is interesting to point out that the larger bubble, 
374: whose partial natural 
375: frequency is lower than that of the smaller bubble, has the highest 
376: transition frequency among all the bubbles.
377: 
378: Now we present numerical solutions of Eq.~(\ref{eq13}) to examine the 
379: influences of viscosity on transition frequencies. Figures 
380: \ref{fig2}$-$\ref{fig4} show the results obtained by using 
381: $\mu = 1.137 \times 10^{-3}$ kg/(m s), which corresponds to the viscosity of 
382: water at room temperature. From those figures, we can observe that, as the 
383: viscous effect grows strong, i.e., the bubbles' radii become smaller, the 
384: STFs vanish gradually from the large-distance region (and sometimes from the 
385: small-distance region), and only the FTF remains. In the case of 
386: $R_{10} = 1$ $\mu$m, the STFs of a larger bubble disappear, and, when 
387: the bubbles are of similar size to each other, the FTF and the higher STF of a 
388: smaller bubble vanish in the small-distance region. In the case of 
389: $R_{10} = 0.1$ $\mu$m, it is difficult to distinguish the STFs from the FTF 
390: since only a smooth curve, decreasing almost monotonically with decreasing $l$, 
391: appears. (The transition frequency of the smaller bubble remaining in 
392: the small-distance region may be the lower one of the STFs.) Those results 
393: show that the influence of the mutual interaction of the bubbles on the 
394: transition frequencies depends on the viscosity of the surrounding material, 
395: and that this 
396: interaction weakens as the viscous effect strengthens, i.e., the threshold 
397: of the distance for the appearance of the STFs is shortened.
398: 
399: Lastly, we briefly discuss the pulsation amplitudes of the bubbles. Figure 
400: \ref{fig5} shows $\alpha _j  \equiv (P_a /R_{j0} \rho )A_j$ and 
401: $\beta _j  \equiv (P_a /R_{j0} \rho )B_j$ for $R_1 = 10$ $\mu$m, 
402: $R_2 = 8$ $\mu$m, $l = 3$, and $\mu = 1.137 \times 10^{-3}$ kg/(m s) as 
403: functions of $\omega /\omega _{10}$, where we set $P_a = 0.1 P_0$. The present 
404: setting is equivalent to that used for one of the graphs shown in 
405: Fig.~\ref{fig2} except for $l$. In each graph shown in Fig.~\ref{fig5}, we 
406: observe three reversals of the sign of $\alpha _j$ and only two 
407: resonance responses, as expected. At the transition frequency that does not 
408: correspond to the natural frequency (the highest one of bubble 1 and the 
409: second highest one of bubble 2), no resonance response is obtained. This 
410: result confirms that the characteristic frequency given by Eq.~(\ref{eq15}) is 
411: not the natural frequency.
412: 
413: In summary, it was predicted in this Letter that a gas bubble interacting 
414: acoustically with a neighboring bubble has three transition frequencies that 
415: make the phase difference between the bubble's pulsation and an external sound 
416: be $\pi /2$, while readymade 
417: theories predict only two natural frequencies. This present result shows 
418: that, in a double-bubble case, the phase reversal of a bubble (e.g., from 
419: in-phase to out-of-phase with an external sound) can take place not only at 
420: the bubble's natural frequencies but also at some other frequency. The 
421: present theory for transition frequencies may be useful for understanding the 
422: reversal of the sign of the secondary Bjerknes force, which has so far been 
423: interpreted using only the natural frequencies \cite{ref5,ref19,ref20,add01}. 
424: (This subject is discussed in a separate paper of ours \cite{ref21}.) 
425: Furthermore, the present results might also 
426: affect understandings of some related subjects such as acoustic localization 
427: \cite{ref22,ref23} and superresonances \cite{ref10}, previous investigations of 
428: which have employed systems containing only identical bubbles. The present 
429: theory for a double-bubble system will be verified \cite{ref24} by the 
430: direct numerical simulation technique proposed in Ref.~\cite{ref25}, and will 
431: be extended to a theory for an $N$-bubble system \cite{ref26}, where $N$ 
432: denotes an arbitrary positive integer.
433: 
434: 
435: 
436: 
437: \begin{thebibliography}{}
438: \bibitem{ref1}
439: M. Strasberg, J. Acoust. Soc. Am. \textbf{25}, 536 (1953).
440: \bibitem{ref2}
441: A. Shima, Trans. ASME, J. Basic Eng. \textbf{93}, 426 (1971).
442: \bibitem{ref3}
443: G. N. Kuznetsov and I. E. Shchekin, Sov. Phys. Acoust. \textbf{21}, 147 
444: (1975).
445: \bibitem{ref4}
446: J. F. Scott, J. Fluid Mech. \textbf{113}, 487 (1981).
447: \bibitem{ref5}
448: E. A. Zabolotskaya, Sov. Phys. Acoust. \textbf{30}, 365 (1984).
449: \bibitem{ref6}
450: Yu. A. Kobelev and L. A. Ostrovskii, Sov. Phys. Acoust. \textbf{30}, 427 
451: (1984).
452: \bibitem{ref7}
453: S. T. Zavtrak, Sov. Phys. Acoust. \textbf{33}, 145 (1987).
454: \bibitem{ref8}
455: H. Takahira, S. Fujikawa and T. Akamatsu, JSME Int. J. Ser. II 
456: \textbf{32} (1989) 163.
457: \bibitem{ref9}
458: A. S. Sangani, J. Fluid Mech. \textbf{232}, 221 (1991).
459: \bibitem{ref10}
460: C. Feuillade, J. Acoust. Soc. Am. \textbf{98}, 1178 (1995).
461: \bibitem{ref11}
462: Z. Ye and C. Feuillade, J. Acoust. Soc. Am. \textbf{102}, 798 (1997).
463: \bibitem{ref12}
464: C. Feuillade, J. Acoust. Soc. Am. \textbf{109}, 2606 (2001).
465: \bibitem{ref13}
466: P.-Y. Hsiao, M. Devaud, and J.-C. Bacri, Eur. Phys. J. E \textbf{4}, 5 
467: (2001).
468: \bibitem{ref14}
469: T. G. Leighton, \textit{The Acoustic Bubble} (Academic Press, London, 1994), p.293.
470: \bibitem{ref15}
471: R. Mettin, I. Akhatov, U. Parlitz, C. D. Ohl, and W. Lauterborn, Phys. 
472: Rev. E \textbf{56}, 2924 (1997).
473: \bibitem{ref16}
474: W. Lauterborn, J. Acoust. Soc. Am. \textbf{59}, 283 (1976).
475: \bibitem{ref17}
476: C. Devin, J. Acoust. Soc. Am. \textbf{31}, 1654 (1959).
477: \bibitem{ref18}
478: M. Ida, e-Print, physics/0108067 (not for submission).
479: \bibitem{ref19}
480: A. A. Doinikov and S. T. Zavtrak, Phys. Fluids \textbf{7}, 1923 (1995).
481: \bibitem{ref20}
482: A. A. Doinikov and S. T. Zavtrak, J. Acoust. Soc. Am. \textbf{99}, 3849 
483: (1996).
484: \bibitem{add01}
485: A. Harkin, T. J. Kaper, and A. Nadim, J. Fluid Mech. \textbf{445}, (2001) 377.
486: \bibitem{ref21}
487: M. Ida, (submitted); e-Print, physics/0109005.
488: \bibitem{ref22}
489: Z. Ye and A. Alvarez, Phys. Rev. Lett. \textbf{80}, 3503 (1998).
490: \bibitem{ref23}
491: A. Alvarez and Z. Ye, Phys. Lett. A \textbf{252}, 53 (1999).
492: \bibitem{ref24}
493: M. Ida, e-Print, physics/0111138.
494: \bibitem{ref25}
495: M. Ida and Y. Yamakoshi, Jpn. J. Appl. Phys. \textbf{40}, 3846 (2001).
496: \bibitem{ref26}
497: M. Ida, (submitted); e-Print, physics/0108056.
498: \end{thebibliography}
499: 
500: 
501: 
502: \newpage
503: 
504: \begin{figure}
505: \begin{center}
506: \epsfxsize = 12 cm
507: \epsffile{fig1.eps}
508: \caption{
509: Transition frequencies of bubbles 1 ($\omega _1$) and 2 ($\omega _2$) for 
510: $R_{j0} \sim 10$ $\mu$m and $\mu \approx 0$, normalized by $\omega _{10}$. 
511: The dashed line denotes $\omega _j /\omega _{10} = 1$. The highest 
512: transition frequency of bubble 1 and the second highest one of bubble 2 are 
513: given by Eq.~(\ref{eq17}).}
514: \label{fig1}
515: \end{center}
516: \end{figure}
517: 
518: \newpage
519: 
520: \begin{figure}
521: \begin{center}
522: \epsfxsize = 12 cm
523: \epsffile{fig2.eps}
524: \caption{
525: Transition frequencies for $R_{j0} \sim 10$ $\mu$m normalized by 
526: $\omega _{10}$. The viscous effect is taken into account. The dashed line 
527: denotes $\omega _j /\omega _{10} = 1$.}
528: \label{fig2}
529: \end{center}
530: \end{figure}
531: 
532: \newpage
533: 
534: \begin{figure}
535: \begin{center}
536: \epsfxsize = 12 cm
537: \epsffile{fig3.eps}
538: \caption{
539: Same as Fig.~\ref{fig2} except for $R_{j0} \sim 1$ $\mu$m.}
540: \label{fig3}
541: \end{center}
542: \end{figure}
543: 
544: \newpage
545: 
546: \begin{figure}
547: \begin{center}
548: \epsfxsize = 12 cm
549: \epsffile{fig4.eps}
550: \caption{
551: Same as Fig.~\ref{fig2} except for $R_{j0} \sim 0.1$ $\mu$m.}
552: \label{fig4}
553: \end{center}
554: \end{figure}
555: 
556: \newpage
557: 
558: \begin{figure}
559: \begin{center}
560: \epsfxsize = 8 cm
561: \epsffile{fig5.eps}
562: \caption{
563: $\alpha _j  = (P_a /R_{j0} \rho )A_j$ and 
564: $\beta _j = (P_a /R_{j0} \rho )B_j$ for $R_1 = 10$ $\mu$m, $R_2 = 8$ $\mu$m, 
565: and $l = 3$.}
566: \label{fig5}
567: \end{center}
568: \end{figure}
569: 
570: 
571: \end{document}
572: