Numerical Simulation of Cavitation Bubble Dynamics Based on

Different Frame Rayleigh-Plesset Equation

Hong Tian, Chen Yang, and Zhengzhu Liao

Abstract—In order to examine the dynamics characteristic of in the attempt to understand cavitation processes it is quite
cavitation bubble in terms of different dynamics parameter of natural to start with single cavitation bubbles to learn about
radius and volume respectively. So we consider the dynamics their dynamics first. Acoustic cavitation is characterized by
factor of viscosity, the surface tension, the vapour pressure and the dynamical evolution of clouds of bubble. The
the adiabatic compress. To use the Gear Stiff algorithm to investigation of single bubble dynamics is not only as a
simulate the famous traditional Rayleigh-Plesset equation in starting point towards a muti-bubble cavitation, but also as a
terms of radius frame and use the Runge-Kutta algorithm to
basic for a more complete comprehension of the whole
simulate the Rayleigh-Plesset equation in terms of volume frame
derived by T. G. Leighton according to the Navier-Stokes phenomenon of ultrasonic cavitation.
equation and other conditions. Based above conditions, we In 1917, Rayleigh [4] developed the collapse of an empty
discuss the influence of different simulation modeling dynamics spherical cavity under a static pressure, and first put forward
characteristic by change the ultrasonic pressure amplitude, the model of cavitation bubble collapse under the infinite
bubble radius or volume and ultrasonic frequency. The medium and incompressible liquid, and this model neglecting
simulation research of the two different dynamics parameter of surface tension and viscosity of liquid, and the bubble stay
Rayleigh-Plesset modeling indicate that the results have the spherical at all times during the pulsation. The governing
resemble variety tendency and the two model may be used to equation of this model is issued from the integration of a
research the ultrasonic cavitation bubble dynamics. The results
combination of the continuity and momentum balance
also supply the different research ways of cavitation bubble
dynamics and the theory base of practice engineer applications equation describes the motion of the bubble boundary and
of ultrasonic cavitation effect. neglect the liquid compressibility and the mass transfer at the
interface [5], [6]. This is the famous Rayleigh bubble
I. INTRODUCTION dynamic equation using the bubble radius R as the dynamic
parameter:
I t is well-known that the mechanical interaction between
ultrasonic waves and bubbles in liquids leads to the
phenomenon of ultrasonic acoustic cavitation. The
•• 3 •
R R+ R 2 =
2
1
ρL L
[ p ( R ) + p∞ (t )]  (1)

cavitation bubble in a liquid undergoes cycles for growth, Where dot denotes the time derivative, R the instantaneous
rapid collapse, and damped rebounds in response to a driving radius, ȡL the liquid density, pL(R) the liquid pressure on the
pressure of ultrasonic sound waves. During cavitation, the external side of the bubble wall, and p’(t) the pressure in the
collapse of bubbles produces intense local heating and high liquid far away from the bubble which is the sum of the
pressures, with very short lifetimes. In clouds of cavitating undisturbed liquid pressure p0 and the acoustic pressure
bubbles, these hot-spots have equivalent temperatures of PA(t)=ˉPAsinȦt.
roughly 5000K, pressures of about 50MPa, and heating and After that, many researchers [7]–[18] developed the
cooling rates above 109 K.s [1]–[3]. In single bubble cavitation bubble dynamic equation based on the Rayleigh


cavitation, conditions may be even more extreme. Ultrasonic equation. The forepart theories suppose the cavitation bubble
cavitation creates extreme physical and chemical conditions is spherical symmetry and model the water as an
which are responsible for all observed effects of ultrasound. incompressible, or an almost incompressible, fluid. The
However, Cavitation bubbles normally come in swarms or bubble dynamics equations were described by others research
clouds, in acoustic cavitation also in highly structured such as Nolting & Neppiras, and Leighton and Gilmore. All
filaments. In these circumstances it is difficult to follow the these theories are the basis of most theoretical analyses of
life of individual bubbles. Moreover, their motion is nonlinear bubble dynamic.
complicated by the interaction with other bubbles. Therefore, According to the theory developed by Rayleigh and others,
the radius, R(t), of the bubble is governed by what is now
Manuscript received June 10, 2008. often called the most popular equation for describing the
Hong Tian is with the Simulation Engineering Graduate School, College nonlinear cavitation bubble is the Rayleigh–Plesset equation:
of Power Engineering, University of Chongqing, Shapingba district,
Chongqing 400044, China ([email protected]). Hong Tian is also with
••
3 •2 1 2σ R
R R+ R = (( p0 + − pv )( 0 ) 3κ
the College of Mechanical and Electronic Engineering, Maoming University, 2 ρ0 R0 R
Maoming 525000, China. •
Chen Yang is with the Simulation Engineering Graduate School, College
of Power Engineering, University of Chongqing, Shapingba district,  + pv − 2σ − 4η R − p0 − P(t ))  (2)
Chongqing 400044, China ([email protected]). R R
Zhengzhu Liao is with the College of Mechanical and Electronic where ȡ0 is the unperturbed liquid density, Ș is the shear
Engineering, Maoming University, Maoming 525000, China viscosity of the liquid, and p0 is the static pressure, and p0 and
([email protected]).

1312

978-1-4244-1787-2/08/$25.00 
c 2008 IEEE

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DA BAHIA. Downloaded on August 25,2023 at 17:56:21 UTC from IEEE Xplore. Restrictions apply.
 P(t) is applied acoustic field P(t), and the ı is the flow across the bubble wall. V is cavitation bubble volume.
contributions of surface tension, and pv is the vapour pressure Where the bubble has volume V(t) and wall volume velocity.
in the bubble, and R0 is the unperturbed bubble radius, and the In the following derivation, using of the function Z
ț is the polytropic index in order to adjust the gas stiffness for substitute the wall volume velocity, then we can get the
reversible heat flow across the bubble wall. following first order differential equation groups:
All the above these theories most are used as the bubble
radius R as the dynamic parameter. So there is few reporter •
(4)
V =Z
reported the acoustic cavitation bubble dynamics simulation 1 1
in terms of volume frame. But the T. G. Leighton [12] – [15]
• •
4π 3V 3 4π 3 V
Z = f (t ,V ,V ) = ( ) (( p0 + 2σ ( ) − pv )( 0 )κ
put forward the volume V is used as the dynamic parameter, ρ 0 4π 3V0 V
and derived the Rayleigh–Plesset equation in terms of the 1 1
• •
4π 3 4π 3V 3 4η V V 2 (5)
bubble volume according to the Navier–Stokes approach and + pv − 2σ ( ) − p0 − P(t )) − ( ) +
other approaches. 3V ρ 0 4π 3V 6V
The paper is organized as follows: we literature the
development of famous Rayleigh–Plesset equation, then we The original values:
are presenting our work to simulate the cavitation bubble •
V (t = 0) = V0 , Z (t = 0) = V (t = 0) = 0 (6)
dynamics. We adopt the Runge–Kutta algorithm to simulate
Rayleigh–Plesset equation in terms of volume frame and
adopt the Gear’s Stiff algorithm to simulate traditional We use the Runge–Kutta algorithm to compute the
famous Rayleigh-Plesset equation in terms of radius frame one-order constant differential equation group, supposing that
according to the different volume values, acoustic field and the time space is ǻt, then for n=0, 1, 2, 3…, for every time
frequency respectively. We discuss the simulation results that space:
can analysis the collapse of transient cavitation bubbles and
pulsation of stable cavitation bubbles. Finally, we compare • Δt 2 (7)
V n+1 = Vn + Δt V n + (M 1 + M 2 + M 3 )
the two difference kinds of Rayleigh–Plesset equations using 6
different simulation algorithm respectively. • • Δt (8)
V n+1 = V n + ( M 1 + 2 M 2 + 3M 3 + M 4 )
6
•
M 1 = f (t n , V n , V n ) (9)
II. METHOD AND MODEL
Δt Δt • • Δt (10)
T.G Leighton [12], [14] derives a ‘volume frame’ version M 2 = f (t n + , V n + V n , V n + M 1 )
2 2 2
of the Rayleigh–Plesset equation from first principles in the
Δt Δt • Δt 2 • Δt (11)
volume frame which incorporates viscous damping and this M 3 = f (t n + ,Vn + V n + M 1 ,V n + M 2 )
has been done using the Navier-Stokes approach to generate 2 2 4 2
the equation refer to “(3),”: Δt Δt • Δt 2 • Δt (12)
M 4 = f (t n + ,Vn + V n + M 2 ,V n + M 3 )
2 2 2 2
•
1 1
••V 2 4π 3V 3 4π 3 V We use the Runge–Kutta algorithm to simulate the ‘volume
V− = ( ) (( p0 + 2σ ( ) − pv )( 0 )κ
6V ρ 0 4π 3V0 V frame’ version of the Rayleigh–Plesset equation refer to
1 1
• “(3),”. Because using the Runge–Kutta algorithm to simulate
4π 3 4π 3V 3 4η V (3) the ‘radius frame’ version of the Rayleigh–Plesset equation
+ p v − 2σ ( ) − p 0 − P(t )) − ( )
3V ρ 0 4π 3V refer to “(2),”will bring the question of stiff, so we use the
Gear's stiff algorithm to simulate the ‘radius frame’ version of
This equation is under the following common assumptions the Rayleigh–Plesset equation refer to “(2),”. We use the
[12], [14]: The bubble exists in an infinite medium. The Visual Fortran to simulate the equation refer to “(2),” and
bubble stays spherical at all times during the pulsation. “(3),” and get the simulation results. All the general original
Spatially uniform conditions exist within the bubble. The parameters as the following: the liquid is water, the
bubble radius is much smaller than the wavelength of the environment temperature is 298.15K, liquid density
driving sound field. There are no body forces acting (e.g. ȡ0=1000kg.m-3, viscosity Ș=0.001Pa.s, surface tension
gravity). Bulk viscous effects can be ignored. The density of ı=0.0725N.m-1, hypothesis adiabatic process, the polytropic
the surrounding fluid is much greater than that of the internal index ț=1.33, velocity of sound in water c=1481m.s-1, the
gas. The gas content is constant. The ‘volume frame’ version static pressure p0=101.3kPa, the vapour pressure
of the Rayleigh–Plesset equation in terms of the bubble pv=3.2718kPa.
volume, including viscous damping. where ȡ0 is the
unperturbed liquid density, Ș is the shear viscosity of the
liquid, and p0 is the static pressure, and p0 and an applied III. RESULTS AND DISCUSS
acoustic field P(t), and the ı is the contributions of surface
tension, and pv is the vapour pressure in the bubble, and R0 is A. Ultrasonic Pressure Amplitude
the unperturbed bubble radius, and the ț is the polytropic
index in order to adjust the gas stiffness for reversible heat

th
2008 Asia Simulation Conference — 7 Intl. Conf. on Sys. Simulation and Scientific Computing 1313

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DA BAHIA. Downloaded on August 25,2023 at 17:56:21 UTC from IEEE Xplore. Restrictions apply.
 effect than the steady cavitation, so we should enlarge the
pressure amplitude and consider the increase scale of volume
or radius. When the ultrasonic pressure amplitude is increase,
the change quantity of volume and radius are decrease.
B. Bubble Radius or Volume
The original dimension is very important to the effect of
cavitation. In graphic (a), the curve 3 the volume max
expands the 17 times. The max radius expands the nearly 3
times in graphic (b). From the Fig.2 we know that the
cavitation bubble not generate the violence cavitation effect,
this is a steady cavitation. The cavitation bubble persists
many periods time. So this kind of cavitation should be
(a) Bubble volume variation with time avoided in the use of ultrasonic acoustic cavitation. From
above analysis we can draw the following conclusion that the
selection of cavitation original dimension of radius and
volume is very important, and we should use simulation way
to calculate the suit cavitation original dimension and adopt
some way make the solution full of the suit cavitation
dimension. From the graphic (a) and graphic (b) we know
they have same variable tendency and the two algorithms are
suitable.

(b) Bubble radius variation with time

Fig.1. Numerically simulated cavitation bubble dynamics results based

on the Runge–Kutta algorithm in terms of volume frame (Fig. 1a) and
the results based on the Gear’s Stiff algorithm in terms of radius frame
(Fig. 1b). Ultrasound frequency fa =100 kHz, cavitation bubble original
radius R0=20.0ȝm, cavitation bubble original volume V0= (4ʌR03)/3,
Curves for different acoustic pressure field parameter:
1-PA=6.5×102MPa, 2-PA=1.0×103MPa, 3-PA=2.2×103MPa.

From the above two graphics (a) and (b), we can draw the
conclusion that the two graphics have the same variable trend (a) Bubble volume variation with time
and have the same acoustic cavitation bubble dynamics
characteristic. From the Fig. 1, curve 3 and curve 2 we know
the cavitations finish the expand, compress and collapse in
more than one period. But the curve 1 in two periods finish
the expand, compress and collapse, also the bubble appear
two peak values and have part spring. In graphic (a) and (b),
the volume cavitation bubble max expands 948 times and the
radius max expands 9.7 times at the same pressure amplitude
2.2MPa. But in order to get the violence transient cavitation
effect, we should select the propriety pressure amplitude. If
the pressure amplitude is larger, the more time of bubble
collapse will need, this is the over larger pressure amplitude
will make the cavitation bubble volume and radius become
large and which can not complete compress and collapse in a (b) Bubble radius variation with time
period time, so that the bubble will enter the another compress
period. The acoustic cavitations have steady cavitation and Fig.2. Numerically simulated cavitation bubble dynamics results based
transient cavitation. Steady cavitation is a long life-span on the Runge–Kutta algorithm in terms of volume frame (Fig. 2a) and
the results based on the Gear’s Stiff algorithm in terms of radius frame
cavitation take place in low pressure amplitude. The transient
(Fig. 2b). Ultrasound frequency fa =100 kHz, cavitation bubble original
cavitation produces in higher pressure amplitude and radius PA =1.5×102MPa, cavitation bubble original volume V0=
accomplish the growth, rapid collapse in two period’s time. (4ʌR03)/3, Curves for different volume and radius parameter:
The transient cavitation will bring the violence cavitation 1-R0=0.5ȝm, 2-R0=0.8ȝm, 3-R0=50.0ȝm.

th
1314 2008 Asia Simulation Conference — 7 Intl. Conf. on Sys. Simulation and Scientific Computing

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DA BAHIA. Downloaded on August 25,2023 at 17:56:21 UTC from IEEE Xplore. Restrictions apply.
 frequency and make the bubble generate the violence
cavitation effect in much short time.

D. Compare of Two Models and Two Algorithms

The traditional cavitation bubble dynamics equation is the
famous Rayleigh–Plesset equation in the radius frame. Many
researchers simulate the bubble dynamics according to the
famous Rayleigh–Plesset in the radius frame. But there are
few researchers simulate the bubble dynamics of
Rayleigh–Plesset in the volume frame. From above
simulation results we know that the two models may be used
to simulate the cavitation bubble dynamics, their difference is
the change times of volume and radius. This is maybe because
(a) Bubble volume variation with time the radius dimensionless value is nearly 1012 times of the
volume dimensionless values. The two equations refer to
“(2),” and “(3),” in different frame of radius and volume
encompasses exactly the same physics [12], and the some
discrepancy between the two equations maybe the difference
in the damping and the greater numerical precision. To use
the Runge–Kutta algorithm to simulate model of
Rayleigh–Plesset in the radius frame will bring the stiff
questions so we use the Gear’s Stiff algorithm to simulate the
model refer to “(2),” of Rayleigh–Plesset in the radius frame.
We use the Runge–Kutta algorithm to simulate model refer to
“(3),” of Rayleigh–Plesset in the volume frame. The two
algorithms all may be used to simulate the cavitation bubble
dynamics according to different simulation model.
(b) Bubble radius variation with time

Fig.3. Numerically simulated cavitation bubble dynamics results based IV. CONCLUSION
on the Runge–Kutta algorithm in terms of volume frame (Fig. 3a) and
the results based on the Gear’s Stiff algorithm in terms of radius frame
The two ultrasonic cavitation bubble dynamics simulation
(Fig.3b). Cavitation bubble original radius PA=1.5×102MPa, models of Rayleigh–Plesset equation in the radius frame and
R0=1.5ȝm, cavitation bubble original volume V0= (4ʌR03)/3, Curves for Rayleigh–Plesset equation in the volume frame are all used to
different ultrasound frequency: 1-fa=100 kHz, 2- fa =200 kHz, 3-fa=300 simulate the cavitation bubble dynamics and the result have
kHz.
the same variety trend. The two algorithms all may be used to
simulate the cavitation bubble dynamics according to
C. Ultrasonic Frequency different simulation model. When the ultrasonic pressure
amplitude is increase, the change quantity of volume and
In graphic (a), fa=100 kHz the volume max expands the 134 radius are decrease. In order to generate the violent cavitation
times, but in graphic (b) the radius max expands the 5 times. effect, the suited original dimension of radius and volume is
In graphic (a) and in 2ȝs, fa=100 kHz the bubble accomplish also very important. Ultrasonic frequency determine the
two cycles, fa =200 kHz the bubble accomplish four cycles, cavitation bubble cycles for growth, rapid collapse, and
fa=300 kHz the bubble accomplish six cycles. In graphic (b) damped rebounds in response to a driving pressure of
and in 2ȝs, fa=100 kHz the bubble accomplish one cycles, fa ultrasonic sound waves. We should control the fitness
=200 kHz the bubble accomplish two cycles, fa=300 kHz the ultrasonic pressure amplitude, suited original dimension and
bubble accomplish three cycles. In graphic (a) the curve 1 has ultrasonic frequency of cavitation bubble so as to generate the
the violence cavitation effect in 1ȝs because of the bubble violent cavitation effect during the cavitation bubble
volume expand the 134 times and accomplish the cavitation compress and collapse.
bubble growth, rapid collapse, and damped rebounds, so the
collapse of bubbles produces intense local heating and high REFERENCES
pressures in 1ȝs, but its frequency is smaller in the three [1] K. S. Suslik, D. A. Hammerton, and R.E. Cline, “The sonochemical hot
curves, this means that the ultrasonic frequency and the point,” J.Am.Chem.Soc, vol.108 (17), pp. 5641–5642, 1986.
cavitation effect is not proportioned. In Fig.3, the two [2] K. S. Suslick, “Ultrasound: It’s Chemical, Physical, and Biological
algorithms appear the different result, but they have same Effects,” VCH, New York, 1988.
[3] K. S. Suslick, Science, vol.247, pp. 1439, 1990.
variable tendency, this because the influence of frequency [4] Rayleigh, “on the pressure developed in a liquid during the collapse of a
have more important than the radius and volume. The more spherical cavity,” Philos Mag, pp. 34–94, 1917.
frequency is larger, the more period times of cavitation [5] E. A. Neppiras, Phys. Rep, vol.61, pp.159, 1980.
bubble will appear. So we should select the suitable [6] F. R. Young, Cavitation, McGraw-Hill, London, 1989.

th
2008 Asia Simulation Conference — 7 Intl. Conf. on Sys. Simulation and Scientific Computing 1315

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DA BAHIA. Downloaded on August 25,2023 at 17:56:21 UTC from IEEE Xplore. Restrictions apply.
 [7] Sascha Hilgenfeldt, “Analysis of Rayleigh-Plesset dynamics for
sonoluminescing bubbles,” J. Fluid Mech, vol.365, pp. 171–204, 1998.
[8] J. B. Keller and I. I. Kolodner, J.Appl. Phys, vol.27, pp. 1152, 1956.
[9] J. B. Keller and M. Miksis, J.Acous.Soc. A, vol.68, pp. 628, 1980.
[10] H. Lamb, Hydrodynamics, 6th edn, Cambridge University Press, 1932.
[11] M. S. Plesset, “The dynamics of cavitation bubbles”, ASME J, Appl.
Mech, vol.16, pp. 228–231, 1949.
[12] F. G. Blake, Technical Memo No. 12 Acoustics Research Laboratory,
Harvard University, Cambridge, MA, USA (1949).
[13] T. G. Leighton, “The Rayleigh–Plesset equation in terms of volume
with explicit shear losses,” Ultrasonics, vol.10, pp. 1–6, 2007.
[14] T. G. Leighton, the Acoustic Bubble, Academic Press, London, 1994.
[15] T. G. Leighton, “Derivation of the Rayleigh–Plesset Equation in Terms
of Volume,” ISVR Technical Report No. 308, 2007.
[16] T.G. Leighton, S. D. Meers, and P. R. White, “Propagation through
nonlinear time-dependent bubble clouds, and the estimation of bubble
populations from measured acoustic characteristics,” Proc. Royal Soc,
vol.460, pp. 2521–2550, 2004.
[17] M. S. Plesset and R. B. Chapman, “Collapse of an initially spherical
vapour cavity in the neighbourhood of a solid boundary,” J. Fluid
Mech, vol.47, pp. 283–290, 1971.
[18] F. R. Glimore, “The growth and collapse of a spherical bubble in a
viscous compressible liquid,” Hydro Lab Calif Inst Tech Report, pp.
26–40, 1952.

th
1316 2008 Asia Simulation Conference — 7 Intl. Conf. on Sys. Simulation and Scientific Computing

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DA BAHIA. Downloaded on August 25,2023 at 17:56:21 UTC from IEEE Xplore. Restrictions apply.