1.9 2.

Article

An Analysis of Acoustic Cavitation

Thresholds of Water Based on the
Incubation Time Criterion Approach

Ivan Smirnov and Natalia Mikhailova

Special Issue
Cavitating Flows
Edited by
Dr. Xavier Escaler

https://doi.org/10.3390/fluids6040134
 fluids
Article
An Analysis of Acoustic Cavitation Thresholds of Water Based
on the Incubation Time Criterion Approach
Ivan Smirnov * and Natalia Mikhailova

Mathematics and Mechanics Faculty, Saint Petersburg State University, Universitetskaya nab. 7/9,
199034 St. Petersburg, Russia; [email protected]
* Correspondence: [email protected]

Abstract: Researchers are still working on the development of models that facilitate the accurate
estimation of acoustic cavitation threshold. In this paper, we have analyzed the possibility of using
the incubation time criterion to calculate the threshold of the onset of acoustic cavitation depending
on the ultrasound frequency, hydrostatic pressure, and temperature of a liquid. This criterion has
been successfully used by earlier studies to calculate the dynamic strength of solids and has recently
been proposed in an adapted version for calculating the cavitation threshold. The analysis is carried
out for various experimental data for water presented in the literature. Although the criterion
assumes the use of macroparameters of a liquid, we also considered the possibility of taking into
account the size of cavitation nuclei and its influence on the calculation result. We compared the
results of cavitation threshold calculations done using the incubation time criterion of cavitation
and the classical nucleation theory. Our results showed that the incubation time criterion more
qualitatively models the results of experiments using only three parameters of the liquid. We then
discussed a possible relationship between the parameters of the two approaches. The results of our



study showed that the criterion under consideration has a good potential and can be conveniently

 used for applications where there are special requirements for ultrasound parameters, maximum
Citation: Smirnov, I.; Mikhailova, N. negative pressure, and liquid temperature.
An Analysis of Acoustic Cavitation
Thresholds of Water Based on the Keywords: acoustic cavitation; ultrasound; threshold amplitude; incubation time criterion
Incubation Time Criterion Approach.
Fluids 2021, 6, 134. https://doi.org/
10.3390/fluids6040134

1. Introduction
Academic Editor: Xavier Escaler
A local decrease in pressure of a liquid due to the propagation of an acoustic wave can
lead to the formation of vapor–gas bubbles. This effect is called acoustic cavitation. Cavitation
Received: 28 January 2021
Accepted: 22 March 2021
processes are accompanied by powerful hydrodynamic disturbances with microflows and
Published: 1 April 2021
microshockwaves. The physical and chemical effects of ultrasonic cavitation can be used to
perform many tasks, such as initiating and accelerating chemical reactions [1–3], enhanced
Publisher’s Note: MDPI stays neutral
oil recovery [4], emulsification [5,6], surgical and medical procedures [7,8], sterilization of
with regard to jurisdictional claims in
liquids [9], isolation of biologically active substances from plant cells [10,11], etc.
published maps and institutional affil- Cavitation bubbles are formed in a liquid when negative pressure in the rarefaction
iations. phase of an acoustic wave reaches a certain threshold value. This threshold value varies
depending on a combination of different parameters, such as hydrostatic pressure level,
the temperature of the liquid, and rate of change in local pressure. One of the ways to
generate an acoustic wave with a given rate and duration of local pressure change is by
Copyright: © 2021 by the authors.
using ultrasonic exposure. The combination of frequency and amplitude of ultrasound
Licensee MDPI, Basel, Switzerland.
allows selecting the required cavitation threshold for specific external conditions and
This article is an open access article
liquid parameters.
distributed under the terms and Although numerous researchers have studied cavitation threshold theoretically, a
conditions of the Creative Commons reliable and generally accepted approach to assessing acoustic cavitation threshold has
Attribution (CC BY) license (https:// not yet been formed. For example, classical nucleation theory (CNT) [12–14] is widely
creativecommons.org/licenses/by/ used to calculate the threshold of inception and the rate of nucleation of vapor–gas nucleus
4.0/). in liquids [15–17]. According to CNT, the vapor–gas phase in a liquid can be nucleated

Fluids 2021, 6, 134. https://doi.org/10.3390/fluids6040134 https://www.mdpi.com/journal/fluids

Fluids 2021, 6, 134 2 of 14

in two conditions. One is when the pressure of the liquid drops below the pressure of
the vapor–gas phase (cavitation). The second is when the liquid temperature rises above
the saturated vapor temperature (boiling). The first method is more commonly used in
research. Theoretical models of cavitation based on the CNT theory provide the best results
that are qualitatively, and sometimes quantitatively, consistent with experimental data.
However, CNT is still only an approximate theory.
Another common approach to calculating cavitation threshold is by modeling the
dynamics of bubble boundaries with a given initial radius under acoustic influence [18,19].
Critical negative pressures for the inception of transient cavitation are estimated using
Blake-type criteria [20] with various changes, taking into account the characteristics of the
liquid or the thermodynamic process involved. This approach also produces acceptable
qualitative and quantitative results. However, it still needs improvement.
The topic of cavitation threshold continues to be the focus of many studies with
various purposes. For instance, experimental study of the cavitation pressure in D2 O,
ethanol, heptane, and DMSO subjected to tension in an acoustic wave was presented
in [21]. The results obtained were compared with those of the CNT. It is discussed that
an appropriate microscopic model of liquids with high surface tension is necessary for
accurate prediction of the cavitation threshold. In addition, in [22], a theoretical study of
the thresholds for inertial cavitation from acoustic radiation force impulse (ARFI) imaging
was carried out for different durations of acoustic periods in water, urine, blood, heart and
skeletal muscle, brain, kidney, liver, and skin. The effects of positive and negative pressure
on the nanodroplet-mediated histotripsy (NMH) cavitation threshold were investigated
separately in [23]. The experimental results showed good agreement with the theoretical
calculation by the CNT. Acoustic cavitation thresholds for two ultrasound contrast agents
(UCA) exposed to 120 kHz ultrasound were measured over a range of duty cycles in [24].
It was shown that the thresholds occurred at approximately the same acoustic pressure
amplitude for all duty cycles and for pulses longer than 20 cycles. In addition, it was
discussed that nucleation of cavitation occurs via rupture of the UCA. In [25,26], the
influence of lipids on the cavitation of liquids was considered. The thermodynamic state
of lipid interfaces during shockwave-induced cavitation in water with sub-microsecond
resolution was analyzed in [25]. It was shown that the cavitation threshold is lowest
near a phase transition of the lipid interface. On the other hand, combining molecular
dynamics simulations with kinetic modeling in [26] revealed that that the presence of
lipid aggregates imposes an upper stability limit for the magnitude of negative pressures
in biological liquids that contain lipid bilayers. The problem of the cavitation inception
threshold has been addressed also by numerical works [27–30]. In [27,28], a numerical
study was performed to analyze the effect of multifrequency excitation on the inertial
cavitation threshold. It was found that the introduction of additional frequencies into
the excitation can significantly decrease the cavitation threshold. A numerical study of
cavitation in a liquid using a free-energy lattice Boltzmann simulation based on the van der
Waals equation of state was carried out in [29]. It was shown that the local bulk pressure is
not sufficient to explain the phenomenon of cavitation inception: it is necessary to take into
account the viscous stress, interfacial contribution to the local pressure, and the Laplace
pressure. In a theoretical work [30], classical nucleation theory was applied to investigate
the interrelation of high-intensity focused ultrasound (HIFU) pressure and temperature
fields with the energetic requirements of bubble nucleation. The thermal and mechanical
effects of HIFU during bubble nucleation for timescales of a few milliseconds were found
to be inextricably linked.
Common to all of the above works is the need to improve existing approaches to assess
the cavitation threshold, taking into account a wide range of parameters, including both
external conditions of excitation on a liquid and the liquid, itself. As discussed in [31–33],
one of the reasons for the discrepancy between the calculations and experiments may be
the use of macroparameters to describe processes at the microlevel. For example, it is
not obvious that the macroscopic parameter of surface energy can be used for a nucleus
Fluids 2021, 6, 134 3 of 14

with a radius of several tens of molecules. Another problem is that the dimensions of the
“defective” structure of the liquid (vapor–gas bubbles, solid particles) have to be known
in advance, which is not always acceptable in real-life situations. Moreover, there is little
evidence to prove that existing models can be used to predict the cavitation threshold
when external conditions change. For example, if the frequency of an ultrasonic wave or
hydrostatic pressure changes, then using model parameters obtained for specific conditions
may not provide an accurate calculation of the cavitation threshold. Thus, at present,
there is still a need to develop existing models or produce new models that are capable of
predicting the inception and development of acoustic cavitation.
A new approach to analyzing the cavitation threshold was proposed in [34,35]. This
approach is based on the application of the incubation time criterion, which was first
applied to assess the dynamic strength of solids [36]. In short, the core problem is that
the strength of a material depends on the strain rate. In the case of a quasistatic load, the
ultimate stress that the material can withstand can be considered as a constant. However,
with a gradual increase in the strain rate, an increase in the ultimate stress is observed. At
high strain rates, the ultimate strength can increase several-fold. With that, the time to
failure decreases to a certain value and remains fairly constant at high strain rates. This
forms two branches on the strain rate dependence of the time to failure: static and dynamic.
The incubation time approach allows one to calculate both branches, and, accordingly, to
predict the strength of materials for any type of loading; it has already been successfully
used to calculate the strength of materials under dynamic loads [35,37–39]. An example
of such modeling can be found in [37], where the problem of spall fracture at nanosecond
pulses was considered. Spall fracture is a widespread method of creating a tensile dynamic
load in materials. This problem does not find a solution within the framework of traditional
concepts of material strength. Using the incubation time criterion, the authors calculated
the strain rate dependence of spall strength. This can be done, because of the assumption
that loading to a critical value of stress is insufficient for failure under dynamic loads, and
the force impulse must reach a critical value within a certain time.
A similar concept can be applied to cavitation. As in problems of strength of material,
according to the concept of incubation time criterion, reaching the critical value of negative
pressure is not sufficient to start cavitation. The negative pressure pulse must reach a
critical value within a certain time. The results presented in [34,35,40] indicate that it is
possible to apply the incubation time criterion to cavitation problems in liquids to predict
the transient cavitation threshold, depending on the acoustic pulse duration or ultrasonic
wave frequency. Additionally, the criterion was applied to calculate the dependence
of ultrasound threshold amplitude on the liquid temperature at different hydrostatic
pressures [40]. However, the effect of combined changes in the parameters of ultrasonic
action, hydrostatic pressure, and liquid temperature on cavitation threshold has not yet
been analyzed. Moreover, calculations based on the incubation time criterion of cavitation
were not compared with calculations based on other models.
In this paper, we have considered the possibility of applying the incubation time
criterion to estimate the threshold amplitude of ultrasound for the onset of cavitation in
water under various experimental conditions. We determined the efficiency of the criterion
by analyzing the experimental data known from existing literature. This criterion uses
only parameters that are assumed to be macroparameters of a liquid. In this study, we
have additionally considered the possibility of taking into account a microparameter in
the size of a cavitation nucleus. In addition, we have also made a comparison between the
calculations made using the criterion and the classical nucleation theory. The results of
our study showed that the incubation time criterion of cavitation allows one to select the
necessary threshold for the onset of cavitation for a given hydrostatic pressure and water
temperature by calculating the combination of frequency and amplitude of ultrasound.
This approach is convenient for applications in which there are special requirements for
the parameters of ultrasound, maximum negative pressure, and temperature of the liquid.
Fluids 2021, 6, 134 4 of 14

2. Theoretical Approach
When studying the tensile strength of liquid, the question arises of determining the
conditions that lead to cavitation. Since the continuity of liquid is violated during cavitation,
we can draw an analogy with failure in solids.
An increase in strength with a decrease in loading duration or an increase in strain rate
observes in the fracture of solids. This effect indicates the need to preserve the magnitude
of the fracture force impulse. The magnitude of the critical stress cannot remain at the
same level, otherwise infinitely short impacts can lead to fracture with infinitesimal energy
being supplied to the place of fracture. In addition, the critical stress must be withstood for
some time, since the reaction of the medium and microprocesses cannot occur instantly
and require a certain amount of time. Based on this reasoning, the incubation time criterion
of fracture was proposed [36,41]. In a generalized form, the criterion can be represented by
the following integral condition [35]:

Zt  α
1 P(t′ )
dt′ ≥ 1, (1)
τ Pc
t−τ

where P(t’) is the intensity of the local force field; Pc is the static limit of the local force field;
τ is the incubation time associated with the dynamics of structural processes preceding
the observed event and characterizing and therefore the time to this event; and α is the
sensitivity of the medium to a level of the local force field. The time and place of the event
are defined as the moment and the point (in the material volume) of the fulfilment of the
condition (Equation (1)).
In order to determine acoustic cavitation, it is convenient to represent the variables
and parameters of the criterion (Equation (1)) in the following form:

Pc ( T ) = Pst − Pph ( T ), (2)

P(t) = A sin(2π f t), (3)

W
τ ( T ) = τ0 e kT , (4)
where Pc is the cavitation threshold at low frequencies, wherein a change in the frequency
does not cause a significant change in the cavitation threshold; T is the liquid temperature;
Pst is the hydrostatic pressure; Pph (T) is the pressure described by the line of the phase
boundary between liquid and vapor; P(t) is the acoustic pressure, A and f are the amplitude
and the frequency of acoustic oscillations; τ 0 is the incubation time typical for a given
spatial scale level (can be considered as a timescale); k is the Boltzmann constant; and W
is the fraction of energy required to start cavitation in a representative volume of a given
scale level.
If we substitute Equations (2)–(4) in Equation (1), we can obtain the expression for cal-
culating the threshold pressure of the beginning of liquid cavitation for different ultrasound
frequencies, at different liquid temperatures, and hydrostatic pressure:

− 1
α
1
R tz
max |sin z|α dz , λ( T ) ≤ π,

 Pc ( T )·

λ( T ) t ∈(0;π ) tz − λ( T )
Pth ( T ) = z (5)
− α1
Pc ( T )· λ(1T ) 0 |sin z|α dz
 Rπ
,

 λ( T ) > π,

where λ( T ) = τ ( T )2π f , z = 2π f t, and t ∈ (0, π/2π f ).

The application of criterion (Equation (1)) and expression in form (Equation (5))
was proposed for calculating the onset of acoustic cavitation in seawater [42] or water,
depending on its temperature and hydrostatic pressure [40]. However, it should be noted
that the parameters τ 0 and W are determined using a semiempirical method, fitting the
calculated curves to the experimental points with the smallest deviation. Despite the
Fluids 2021, 6, 134 5 of 14

clear physical evidence of these parameters, the question of whether they can be directly
determined from experiment or calculation still remains open. Possible ways to determine
these parameters and potential future research on this matter will be considered in the
discussion section.
One more remark needs to be made for Equation (3). Here, we assume a sinusoidal
acoustic waveform. However, in a real-life situation, this form is not always realized.
The mathematical idealization of the acoustic waveform can lead to calculation errors. In
addition, studies of acoustic cavitation are often considered impulse waves. Therefore,
in this case, it is necessary to use a different expression for the acoustic wave profile in
Equation (3) as, for example, it was done for pulse pressure in [34].
In the following sections, we will analyze the results of calculations of the onset of
acoustic cavitation in water for various external conditions and compare them with various
experimental data. Equation (5) was solved using numerical calculations in MATLAB; the
optimal approximation of the calculated curves to experimental points was determined by
using the least squares method.

3. Results
Although numerous studies have analyzed cavitation problems, not many experi-
ments have been conducted to study the dependence of acoustic cavitation threshold on the
frequency of ultrasound, liquid temperature, and hydrostatic pressure. Therefore, we have
collected some experimental data together to evaluate the efficiency of utilizing the concept
of the incubation time criterion to predict the onset of acoustic cavitation on the example of
water. In these works, the authors investigated the dependence of the cavitation threshold
on one external parameter, with the exception of the work of Bader et al. (2012) [43], where
both temperature and static pressure were changed in the tests. In this work, transient
cavitation threshold was studied in ultrapure water. A radially symmetric standing wave
field was generated in a spherical resonator at frequency of 25.5 kHz. Static pressure
varied from 1 bar to 300 bar, and the temperature range was 18–34 ◦ C. In the study by
Vlaisavljevich et al. (2016) [16], distilled degassed water underwent focused ultrasonic
pulse with a frequency of 1 MHz. In the experiments, passive cavitation detection and
high-speed photography were used. Cavitation threshold was measured from the pressure
amplitude dependency on the probability of cavitation. Experiments were carried out
within a temperature range of 10–90 ◦ C. Similar work was done by Herbert et al. (2006) [44].
Authors presented experimental and theoretical study of cavitation threshold for different
temperatures at a frequency of 1 MHz. Threshold pressure was defined from the statistics of
cavitation. Connolly and Fox (1954) [45] also studied cavitation threshold in a temperature
range of 0–30 ◦ C. Ultrasound was applied as plane progressive and focused waves driven
at a frequency of 1 MHz. Barger (1964) [46] measured cavitation thresholds at a wide range
of frequencies. Water was placed in a spherical glass to which radially symmetric resonance
modes were applied. As in previous works, ultrasonic action was focused on a center of
the sphere.
The experimental conditions (ultrasound frequency f, water temperature T, and hy-
drostatic pressure Pst ) and the parameters required for the calculations by Equation (1)
are given in Table 1. The results of calculations, together with the experimental data, are
presented in Figures 1 and 2. In all cases, the parameter α was equal to 1/2 [47].
The calculation was carried out in such a way that the calculated curves obtained
by Equation (5) approximate the experimental points that meet certain conditions in
water. Since, in the experimental data, the onset of cavitation was dependent either on
water temperature or on ultrasound frequency, while the other conditions of exposure
to water were fixed, the parameters for the incubation time criterion of cavitation were
selected for one of the cases. After that, another case was calculated, even if there were no
corresponding experimental data. Figure 1 shows the calculated dependencies of the onset
of acoustic cavitation on water temperature at fixed ultrasound frequency and hydrostatic
pressure. Figure 2 shows the calculated dependencies of the onset of acoustic cavitation on
Fluids 2021, 6, 134 6 of 14

ultrasound frequency at fixed liquid temperature and hydrostatic pressure. The calculated
curves indicated by the same number in both figures were obtained for the same parameters
Pc , τ 0 , and W. Thus, we assumed that the parameters are invariant with respect to the
experimental scheme.
It is clearly seen that the experimental data differs greatly in assessing the cavitation
strength of water. However, good qualitative and quantitative agreement between the
calculated curves and experimental points was observed. The corresponding coefficient
of determination R2 is presented in Table 1. Here, the higher the cavitation threshold, the
longer was the incubation time of cavitation τ 0 and the lower was the energy fraction for
activation of cavitation W. Nevertheless, their value depends on the accuracy of determin-
ing the static cavitation threshold Pc . The results obtained are discussed in more detail in
the nextτsection.

Table 1. The experimental conditions and parameters for the calculation of the onset of acoustic cavitation. Each ordinal line
number corresponds to a specific calculated curve in the figures. R2 is the coefficient of determination, which corresponds
′

to the calculated curve (with the parameters τ 0 and W) approximated to the experimental points (with the parameters f, T,
Pst , and Pc ). τ

f (kHz) T (◦ C) Pst (kPa) Pc τ 0 (ns) W ·10−23 (J) Rc (nm) R2

τ ∙ −
Experimental Conditions Calculated Data
1 Connolly and Fox, 1954 [45] 1000 0–30 101 Equation (2) 29.5 1220 1.17 0.944
2 Barger, 1964 [46] 183–1160 21 101 Equation (2) 2.53 2700 1.74 0.9279
3 Barger, 1964 [46] 183–1160 21 101 300 * kPa 1.46 2700 1.74 0.9549
4 Bader et al., 2012 [43] 25.5 18–34 101 Equation (2) 59.0 2810 1.77 0.821
5 Bader et al., 2012 [43] 25.5 18–34 5000 Equation (2) 48.0 2585 1.70 0.8917
6 Bader et al., 2012 [43] 25.5 18–34 10,000 Equation (2) 26.5 2790 1.77 0.9545
7 Bader et al., 2012 [43] 25.5 18–34 101 Equation (6) ** 75.8 2390 1.64 0.7911
8 Bader et al., 2012 [43] 25.5 18–34 5000 Equation (6) ** 76.3 2380 1.63 0.7866
9 Bader et al., 2012 [43] 25.5 18–34 10,000 Equation (6) ** 23.6 2830 1.78 0.9536
10 Bader et al., 2012 [43] 25.5 18–34 101 Equation (6) *** 47.1 2840 1.78 0.8231
11 Bader et al., 2012 [43] 25.5 18–34 5000 Equation (6) *** 58.8 2500 1.67 0.8840
12 Bader et al., 2012 [43] 25.5 18–34 10,000 Equation (6) *** 32.2 2710 1.74 0.9557
13 Vlaisavljevich et al., 2016 [16] 1000 10–90 101 Equation (2) 6190 20 0.15 0.727
14 Vlaisavljevich et al., 2016 [16] 1000 10–90 101 Equation (6) **** 152 505 0.75 0.9225

* specified in [35] as a cutoff for transient cavitation; ** R0 = 0.1 µm; *** R0 = 1 µm; **** R0 = 2.5 nm.
μ μ

108
13

107
5 6
Pth(Pa)

10 6 4 2
1
5
10
[45] [46] [16] [44]
[43] [43] [43]
104
0 10 20 30 40 50 60 70 80 90
T (oC)

Figure 1. Dependencies of the threshold ultrasound amplitude for the onset of cavitation in water on
temperature. The points are experimental data. The curves were obtained using Equation (5). The
experimental conditions and parameters used for the calculations are presented in Table 1, according
to the designation of the curves in the figure.
Fluids 2021, 6, 134 7 of 14

109

108
5
4
7
10
Pth (Pa) 13 2
106
1
105
[45] [46] [43] [43] [16] [44]
104
104 105 106
f (Hz)

Figure 2. Dependencies of the threshold ultrasound amplitude for the onset of cavitation in water
on ultrasound frequency. The points are experimental data. The curves were obtained using
Equation (5). The experimental conditions and parameters used for the calculations are presented in
Table 1, according to the designation of the curves in the figure.

4. Discussion
The results shown in Figure 2 were obtained based on assumption that the “qua-
sistatic” strength of water Pc in all experiments is the same and can be calculated by using
Equation (2). However, for example, as shown by curve 2, this approach can significantly
τ
deviate the calculated curve from the experimental points. In general, a liquid may contain
cavitation nuclei of various sizes. This can lead to significant differences in the cavitation
strength of the liquid. Therefore, it is necessary to take into account these features and
determine the quasistatic cavitation threshold for using the incubation time criterion more
accurately. The best way to do this is through direct experimental measurement. Never-
theless, as this is not always possible, let us consider another possible theoretical way of
estimating the “quasistatic” threshold of cavitation Pc .
Following Laplace′ s equation, the critical negative pressure should be proportional to
the pressure caused by the surface tension of the initial bubble. As shown in [46,48,49], the
proportionality constant can be calculated based on the values of hydrostatic and acoustic
pressure and the critical radius of the bubble. It follows that the acoustic pressure required
to exceed the tensile strength of a small air or vapor bubble at a sufficiently slow pressure
change can be estimated as

2σ( T )
Pc ( T ) = Pst + − Pph ( T ), (6)
R0
′ tension and R0 is the initial radius of the bubble. The temperature
where σ(T) is the surface
dependence of the surface tension of pure water in contact with its vapor can be represented
using the following formula [50]:

T 1.256
    
T
σ ( T ) = 0.2358 1 − 1 − 0.625 1 − , (7)
Tc Tc

where Tc is the critical temperature 2σ(𝑇)

𝑃 (𝑇) = (647.096
𝑃 + K). We − 𝑃used a reduced surface tension of
(𝑇),
27.5%, because, as discussed by various authors 𝑅 [31–33,44], it is not accurate to use the
bulk macroscopic
σ surface tension value for calculating the surface tension of a cavitation
nucleus with a nanoscopic radius.
The results of calculations by Equation (5) using Equation (6) instead of Equation (3)
are presented in Figures 3 and 4. The conditions for calculations are presented in Table 1,
.
with the corresponding designation of the 𝑇 curves 𝑇 initial bubble radii
in the figures. The
𝜎(𝑇) = 0.2358 1 − 1 − 0.625 1 − ,
were chosen according to their estimates 𝑇given in the corresponding𝑇 papers. The addition
Fluids 2021, 6, 134 8 of 14

of pressure due to surface tension significantly changes the calculated dependencies of

acoustic threshold for the onset of cavitation on ultrasound frequency. However, as can
be clearly seen in Figure 3, the initial size of the cavitation nucleus and the hydrostatic
pressure also affect the calculation results. We used two different initial bubble radii to
model the results of the experiments from [43]. To simplify the calculation, we assumed
that the change in the bubble radius, as a result of the temperature change, is negligible.
When the hydrostatic pressure is equal to atmospheric pressure (curves 7 and 10), a clear
influence of the bubble radius was observed, especially at liquid temperatures above 40 ◦ C
or ultrasound frequencies below 20 kHz. When the hydrostatic pressure was above 5 MPa
(curves 8 and 11 or 9 and 12), the effect of the bubble radius was almost indistinguishable.

108 12 9
14

11 8
107
Pth (Pa)

3
106

[43] [43] [43] 7

[46] [16] [44] 10
105
0 10 20 30 40 50 60 70 80 90
T (oC)
Figure 3. Dependencies of the threshold ultrasound amplitude for the onset of cavitation in water on
temperature calculated, taking into account the assumed radius of the cavitation nucleus. The points
are experimental data. The curves were obtained using Equation (5). The experimental conditions
and parameters used for the calculations are presented in Table 1, according to the designation of the
curves in the figure.

109

108
9 14

107 12
Pth (Pa)

7
106
10
3
105
[46] [43] [43] [16] [44]
104
104 105 106
f (Hz)
Figure 4. Dependencies of the threshold ultrasound amplitude for the onset of cavitation in water on
ultrasound frequency calculated, taking into account the assumed radius of the cavitation nucleus.
The points are experimental data. The curves were obtained using Equation (5). The experimental
conditions and parameters used for the calculations are presented in Table 1, according to the
designation of the curves in the figure.
Fluids 2021, 6, 134 9 of 14

Another regularity can be identified by comparing the results in [43] and [16,44]. Since
there are no data regarding hydrostatic pressure in [16], we assumed it to be equal to
atmospheric pressure. In [44], the authors note that the threshold negative pressure was
not dependent on the hydrostatic pressure in the range of 1–9 MPa. However, the threshold
negative pressures in [43] were reported to have depended significantly on hydrostatic
pressure and reached the level of the threshold negative pressures reported in [16,44] only
at the hydrostatic pressure of 10 MPa. On the other hand, these values were obtained at
different ultrasound frequencies (25.5 kHz in [43] and 1 MHz in [16,44]). The calculation
based on the incubation time criterion (Figure 4, curves 7,9,14) shows that the threshold
negative pressure at the frequency of 1 MHz for the hydrostatic pressure of 10 MPa and
0.1 MPa pertaining to the liquid in [43] should significantly exceed the threshold negative
pressure at the same frequency of acoustic oscillations pertaining to the liquid in [16,44].
Note that the energy required for the activation of cavitation W obtained for curves 7 and
9 is significantly higher than the energy for curve 14. The opposite is observed for the
characteristic timescale τ 0 .
In [16,44], the experimental data were compared with calculations based on the
classical nucleation theory (CNT) [12–14]. Let us consider how the calculations of the
threshold acoustic amplitude according to the CNT are consistent with the calculations,
according to the incubation time criterion. According to CNT [12,13,16,44,48], the cavitation
pressure for an experiment carried out with a volume V during time ξ at the cavitation
probability of 1/2 can be determined using the formula

 12
16πσ3 1

PCNT = Pth − , (8)
3kT ln( J0 Vξ/ ln 2)

where k is the Boltzmann constant, and J0 can be approximated by [48]:

r
2σ
J0 = n . (9)
πm
Here, n is the number density of the liquid, and m is the mass of the molecule. Note
that there are no generally accepted expressions for V, ξ and J0 yet, and ways to define them
are still being discussed. However, these parameters are inside the logarithm, and their
change has little effect on the cavitation threshold PCNT . The surface tension provides the
main contribution to the value of the cavitation threshold. Since the dependence of surface
tension on the frequency of acoustic oscillations has not been established, in order to plot
the dependence of PCNT on frequency, let us assume that volume V and time ξ depend on
frequency. For example, let the volume V be equal to 4/3π(c/f )3 (where c is the speed of
sound), and the time ξ be equal to 1/(10f ). Thus, volume V is determined by the length of
the acoustic wave. The choice of ξ allows us to assume that the variations of P and T are
minimal and that their values could be considered almost constant over such timescale.
Figures 5 and 6 show the results of the calculations made using Equation (8). The
use of CNT allowed us to obtain the dependence of cavitation threshold on temperature,
which quantitatively coincides with the experimental results [16,44]. To obtain this result,
we had to reduce the surface tension by almost 4 times. The same correction was used
in [16]. Note that the calculation of the experimental results from [43] required a decrease
in the surface tension value to 5%. However, the calculation did not correspond well to
the dependence of the cavitation threshold on temperature. The results of calculating the
cavitation threshold according to CNT (Equation (8)) and the incubation time criterion
(Equation (5)) were close to each other, when considering experiments from [16] (curves
14 and CNT0.275σ in Figure 5). In other cases, the incubation time criterion provided a
better result.
Fluids 2021, 6, 134 10 of 14
σ

108
14

107 CNT(0.275σ)
Pth (Pa) 9
3 CNT(0.05σ)

106
7
[46] [43] [43]
[16] [44] CNT CNT
105
0 10 20 30 40 50 60 70 80 90
T (oC)
Figure 5. A comparison of the results of calculating the dependencies of the threshold ultrasound
amplitude for the onset of cavitation in water on temperature, calculated using Equation (1) and
classical nucleation theory (CNT) (Equation (8)). The points are experimental data. The curves show
the results of the calculations. The experimental conditions and parameters used for the calculations
are presented in Table 1, according to the designation of the curves in the figure.

109
9
108 CNT(0.275σ)

107
14 CNT(0.05σ)
Pth (Pa)

7
106
3
5
10
[46] [43] [43]
[16] [44] CNT CNT
104
104 105 106
f (Hz)
Figure 6. A comparison of the results of calculating the dependencies of the threshold ultrasound
amplitude for the onset of cavitation in water on ultrasound frequency, calculated using Equation
(1) and CNT (Equation (8)). The points are experimental data. The curves show the results of the
calculations. The experimental conditions and parameters used for the calculations are presented in
Table 1, according to the designation of the curves in the figure.

According to CNT [12,13], there is the critical radius for the cavitation nucleus Rc . For
the nucleus of radius Rc , the minimum work required to create a spherical bubble of radius
R filled with vapor will reach its maximum. This maximum is an energy barrier that must
be overcome for the bubble to grow spontaneously. Since the incubation time criterion
assumes the presence of such a liquid parameter as the energy required for the onset of

/
2𝑊
𝑅 = .
4𝜋𝜎
Fluids 2021, 6, 134 11 of 14

cavitation in a representative volume, it is interesting to estimate to which critical size of

the nucleus it can relate. The critical radius of the nucleus can be defined as:

2W 1/2
 
Rc = . (10)
4πσ

Such estimates of the critical radius are presented in Table 1. For all calculations, the
critical radius exceeds the average distance of intermolecular bonds by only 3–6 times
(~2.7 Å; see, for example, [51,52]). Therefore, it can be assumed that the cavitation activation
energy W introduced into the incubation time criterion should be associated with the
formation of nuclei at the intermolecular level. In this case, the incubation time τ 0 can be
considered as the time required to prepare the onset of cavitation at the given spatial scale.
However, it remains unclear how this parameter can be defined. In continuation of the
analysis of the experimental data, let us consider a way for estimation of τ 0 and W based
on the CNT.
In the classical theory of nucleation, it is customary to consider the nucleation rate [12]
W∗
J = NBe − kT , (11)

where N is the number of molecules per unit volume of the liquid, B is the kinetic factor, and
W* is the work of forming a critical nucleus. The factor N exp( − W ∗ /kT ) is the average
number of critical nuclei per unit volume of the liquid, and the factor B corresponds to the
average rate of transition of the nucleus through the critical size Rc . Work W* characterizes
the height of the Gibbs free energy barrier, which must be overcome by the system due to
atomic fluctuations in order to form a nucleus. The ratio W ∗ /kT is called the Gibbs number
and is considered as a dimensionless measure of the stability of a liquid to the formation of
cavitation nuclei. In this case, the height of the activation barrier, i.e., the Gibbs potential
difference for the two phases, is related to the average energy of thermal motion per degree
of freedom. Herewith, the more the critical size of the nucleus is exceeded, the less the
growth of the bubble is subject to fluctuations, i.e., the time of the “settled” position of the
atoms around the nucleus increases. Thus, the average time of the appearance of a nucleus
in a given volume of a liquid V is inversely proportional to the rate of nucleation. If we
assume that the incubation time (Equation (4)) is the time characterizing average waiting
time for irreversible expansion of the cavitation nucleus depending on temperature, and
W = W* , then we can obtain an expression for determining τ 0 :

τ0 = ( NBV )−1 . (12)

The molecules that oscillate around the temporary equilibrium positions then jump
to another position, to the bottom of a potential well formed by the particles of the new
environment. The magnitude of the jumps determines the structural scale d, and the
average settling time τ 0 is the characteristic timescale for a given liquid. The kinetic factor
does not yet have a universal expression. Its classical estimates are discussed in detail
in [12,53]. The ambiguous variants of the factor B provide ample opportunities for a large
set of τ 0 values. Therefore, additional research is needed on this issue. Nevertheless, the
incubation time can be calculated, taking into account the thermal and hydrodynamic
conditions near the nucleus. The accuracy of its calculations will be determined by the
accuracy of the estimation of the CNT parameters.
Note again that, in this work, we estimate the threshold for the onset of cavitation
based on the approach that is used to estimate the strength of solids. Since failure is a
process that develops over time, it is natural to characterize it either by a certain rate of
accumulation of defects or by the time during which the process develops to a certain
stage of failure. Therefore, according to the kinetic concept of failure, the durability of a
solid is a fundamental characteristic of the mechanical strength of a material. It can be
considered as a value inversely proportional to the average rate of the failure process. In the
Fluids 2021, 6, 134 12 of 14

classical expression, this formula was presented by Zhurkov [54,55] and has a general form
(Equation (4)). The proportionality of τ to the factor exp(−W/kT ) indicates the thermal
fluctuation nature of failure. Thus, the application of the criterion for the incubation time
of failure of solids, considering the thermal fluctuation nature of failure for the problem
of failure of water under tensile pressures, gave results that qualitatively describe the
experimental data, at least within the framework of the considered experimental conditions.
It is noteworthy that similar ideas are embedded in the classical theory of nucleation.
Postulating the equality of the activation energies of failure in Equations (4) and (11), we
can propose a way for estimating the parameters of the incubation time criterion based on
the CNT principles. However, this requires more detailed studies, including experimental
ones, which we plan to implement in future works. Moreover, the results obtained need to
be considered for other liquids, including multiphase ones.

5. Conclusions
Our study proved that the incubation time approach can be used to calculate the
dependence of the threshold amplitude of ultrasound for the onset of cavitation in water on
the frequency of ultrasound, liquid temperature, and hydrostatic pressure. The results of
numerical calculations qualitatively, and, in some cases, also quantitatively, corresponded
to the experimental results obtained for various exposure conditions and fluid quality.
Meanwhile, these experimental results cannot be predicted using the classical theory of
nucleation. The calculation of the threshold amplitude using the incubation time criterion
was carried out on the basis of three parameters of the liquid. All of them imply the
consideration of cavitation on the same spatial scale. The first is the cavitation threshold at
a slow change in the negative (tensile) pressure. This can be determined experimentally or
through direct calculation. However, the method of calculation determines the accuracy
of further modeling. The second is the incubation time of cavitation, which characterizes
the dynamics of the preparatory processes of the onset of cavitation. The third one is the
fraction of energy required to start cavitation. The last two parameters are still determined
semiempirically by approximating the calculated curves to experimental data. Neverthe-
less, they can potentially be considered as physical quantities. For example, a comparison
with cavitation threshold calculated according to the classical nucleation theory shows
that, in certain cases, both approaches give similar results. Taking into consideration the
fact that the nucleation theory also assumes the presence of a characteristic volume and
timescale, the combination of the two theories opens up new possibilities for analyzing the
physical essence of the parameters of the incubation time criterion. This can be the subject
of further research about the development of the incubation time criterion of cavitation for
various liquids and external conditions.

Author Contributions: Conceptualization, I.S.; methodology, I.S.; validation, I.S. and N.M.; formal
analysis, I.S. and N.M.; investigation, I.S. and N.M.; data curation, I.S.; writing—original draft
preparation, I.S.; writing—review and editing, I.S.; visualization, N.M.; supervision, I.S.; project
administration, I.S.; funding acquisition, I.S. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was supported by the Grant of the President of the Russian Federation for
young scientists (MK-2269.2019.8). Section 4 contains the results obtained within the grant from the
Russian Science Foundation, grant number 20-79-10078.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Suslick, K.S.; Eddingsaas, N.C.; Flannigan, D.J.; Hopkins, S.D.; Xu, H. The Chemical History of a Bubble. Acc. Chem. Res. 2018, 51,
2169–2178. [CrossRef] [PubMed]
2. Rashwan, S.S.; Dincer, I.; Mohany, A. An investigation of ultrasonic based hydrogen production. Energy 2020, 205, 118006.
[CrossRef]
3. Kerboua, K.; Hamdaoui, O. Oxygen-argon acoustic cavitation bubble in a water-methanol mixture: Effects of medium composition
on sonochemical activity. Ultrason. Sonochem. 2020, 61, 104811. [CrossRef] [PubMed]
Fluids 2021, 6, 134 13 of 14

4. Luo, X.; Gong, H.; He, Z.; Zhang, P.; He, L. Recent advances in applications of power ultrasound for petroleum industry. Ultrason.
Sonochem. 2021, 70, 105337. [CrossRef] [PubMed]
5. Zhao, S.; Yao, C.; Dong, Z.; Liu, Y.; Chen, G.; Yuan, Q. Intensification of liquid-liquid two-phase mass transfer by oscillating
bubbles in ultrasonic microreactor. Chem. Eng. Sci. 2018, 186, 122–134. [CrossRef]
6. Tiong, T.J.; Chu, J.K.; Lim, L.Y.; Tan, K.W.; Hong Yap, Y.; Asli, U.A. A computational and experimental study on acoustic pressure
for ultrasonically formed oil-in-water emulsion. Ultrason. Sonochem. 2019, 56, 46–54. [CrossRef] [PubMed]
7. Brennen, C.E. Cavitation in medicine. Interface Focus 2015, 5, 20150022. [CrossRef] [PubMed]
8. Stride, E.; Segers, T.; Lajoinie, G.; Cherkaoui, S.; Bettinger, T.; Versluis, M.; Borden, M. Microbubble Agents: New Directions.
Ultrasound Med. Biol. 2020, 46, 1326–1343. [CrossRef] [PubMed]
9. Matafonova, G.; Batoev, V. Dual-frequency ultrasound: Strengths and shortcomings to water treatment and disinfection. Water
Res. 2020, 182, 116016. [CrossRef]
10. Alarcon-Rojo, A.D.; Carrillo-Lopez, L.M.; Reyes-Villagrana, R.; Huerta-Jiménez, M.; Garcia-Galicia, I.A. Ultrasound and meat
quality: A review. Ultrason. Sonochem. 2019, 55, 369–382. [CrossRef]
11. Kumar, K.; Srivastav, S.; Sharanagat, V.S. Ultrasound assisted extraction (UAE) of bioactive compounds from fruit and vegetable
processing by-products: A review. Ultrason. Sonochem. 2021, 70, 105325. [CrossRef] [PubMed]
12. Skripov, V.P. Metastable Liquids; J. Wiley: New York, NY, USA, 1974.
13. Debenedetti, P.G. Metastable Liquids: Concepts and Principles; Princeton University Press: Princeton, NJ, USA, 1996.
14. Karthika, S.; Radhakrishnan, T.K.; Kalaichelvi, P. A Review of Classical and Nonclassical Nucleation Theories. Cryst. Growth Des.
2016, 16, 6663–6681. [CrossRef]
15. Caupin, F.; Arvengas, A.; Davitt, K.; Azouzi, M.E.M.; Shmulovich, K.I.; Ramboz, C.; Sessoms, D.A.; Stroock, A.D. Exploring water
and other liquids at negative pressure. J. Phys. Condens. Matter 2012, 24, 284110. [CrossRef] [PubMed]
16. Vlaisavljevich, E.; Xu, Z.; Maxwell, A.D.; Mancia, L.; Zhang, X.; Lin, K.-W.; Duryea, A.P.; Sukovich, J.R.; Hall, T.L.; Johnsen, E.;
et al. Effects of Temperature on the Histotripsy Intrinsic Threshold for Cavitation. IEEE Trans. Ultrason. Ferroelectr. Freq. Control
2016, 63, 1064–1077. [CrossRef] [PubMed]
17. Baidakov, V.G.; Vinogradov, V.E.; Pavlov, P.A. Homogeneous nucleation in liquid nitrogen at negative pressures. J. Exp. Theor.
Phys. 2016, 123, 629–637. [CrossRef]
18. Rozenberg, L.D. High-Intensity Ultrasonic Fields; Springer US: Boston, MA, USA, 1971; ISBN 978-1-4757-5410-0.
19. Neppiras, E.A. Acoustic cavitation thresholds and cyclic processes. Ultrasonics 1980, 18, 201–209. [CrossRef]
20. Blake, F.G. The Onset of Cavitation in Liquids; Technical Memo No.12; Acoustics Research Laboratory, Harvard University:
Cambridge, MA, USA, 1949.
21. Arvengas, A.; Herbert, E.; Cersoy, S.; Davitt, K.; Caupin, F. Cavitation in heavy water and other liquids. J. Phys. Chem. B 2011, 115,
14240–14245. [CrossRef] [PubMed]
22. Church, C.C.; Labuda, C.; Nightingale, K. A Theoretical Study of Inertial Cavitation from Acoustic Radiation Force Impulse
Imaging and Implications for the Mechanical Index. Ultrasound Med. Biol. 2015, 41, 472–485. [CrossRef] [PubMed]
23. Vlaisavljevich, E.; Aydin, O.; Lin, K.W.; Durmaz, Y.Y.; Fowlkes, B.; Elsayed, M.; Xu, Z. The role of positive and negative pressure
on cavitation nucleation in nanodroplet-mediated histotripsy. Phys. Med. Biol. 2015, 61, 663–682. [CrossRef] [PubMed]
24. Gruber, M.J.; Bader, K.B.; Holland, C.K. Cavitation thresholds of contrast agents in an in vitro human clot model exposed to
120-kHz ultrasound. J. Acoust. Soc. Am. 2014, 135, 646–653. [CrossRef] [PubMed]
25. Shrivastava, S.; Cleveland, R.O. Thermodynamic state of the interface during acoustic cavitation in lipid suspensions. Phys. Rev.
Mater. 2019, 3, 1–13. [CrossRef]
26. Kanduč, M.; Schneck, E.; Loche, P.; Jansen, S.; Jochen Schenk, H.; Netz, R.R. Cavitation in lipid bilayers poses strict negative
pressure stability limit in biological liquids. Proc. Natl. Acad. Sci. USA 2020, 117, 10733–10739. [CrossRef] [PubMed]
27. Wang, M.; Zhou, Y. Numerical investigation of the inertial cavitation threshold by dual-frequency excitation in the fluid and
tissue. Ultrason. Sonochem. 2018, 42, 327–338. [CrossRef] [PubMed]
28. Suo, D.; Govind, B.; Zhang, S.; Jing, Y. Numerical investigation of the inertial cavitation threshold under multi-frequency
ultrasound. Ultrason. Sonochem. 2018, 41, 419–426. [CrossRef] [PubMed]
29. Kähler, G.; Bonelli, F.; Gonnella, G.; Lamura, A. Cavitation inception of a van der Waals fluid at a sack-wall obstacle. Phys. Fluids
2015, 27, 123307. [CrossRef]
30. de Andrade, M.O.; Haqshenas, S.R.; Pahk, K.J.; Saffari, N. The effects of ultrasound pressure and temperature fields in millisecond
bubble nucleation. Ultrason. Sonochem. 2019, 55, 262–272. [CrossRef] [PubMed]
31. Kashchiev, D. Thermodynamically consistent description of the work to form a nucleus of any size. J. Chem. Phys. 2003, 118,
1837–1851. [CrossRef]
32. Bruot, N.; Caupin, F. Curvature Dependence of the Liquid-Vapor Surface Tension beyond the Tolman Approximation. Phys. Rev.
Lett. 2016, 116, 056102. [CrossRef]
33. Oxtoby, D.W. Homogeneous nucleation: Theory and experiment. J. Phys. Condens. Matter 1992, 4, 7627–7650. [CrossRef]
34. Besov, A.S.; Kedrinskii, V.K.; Morozov, N.F.; Petrov, Y.V.; Utkin, A.A. On the similarity of the initial stage of failure of solids and
liquids under impulse loading. Dokl. Phys. 2001, 46, 363–365. [CrossRef]
35. Petrov, Y.V. Incubation time criterion and the pulsed strength of continua: Fracture, cavitation, and electrical breakdown. Dokl.
Phys. 2004, 49, 246–249. [CrossRef]
Fluids 2021, 6, 134 14 of 14

36. Petrov, Y.V.; Utkin, A.A. Dependence of the dynamic strength on loading rate. Sov. Mater. Sci. 1989, 25, 153–156. [CrossRef]
37. Petrov, Y.V.; Smirnov, I.V.; Utkin, A.A. Effects of strain-rate strength dependence in nanosecond load duration range. Mech. Solids
2010, 45, 476–484. [CrossRef]
38. Petrov, Y.V.; Karihaloo, B.L.; Bratov, V.V.; Bragov, A.M. Multi-scale dynamic fracture model for quasi-brittle materials. Int. J. Eng.
Sci. 2012, 61, 3–9. [CrossRef]
39. Petrov, Y.V.; Gruzdkov, A.A.; Bratov, V.A. Structural-temporal theory of fracture as a multiscale process. Phys. Mesomech. 2012, 15,
232–237. [CrossRef]
40. Volkov, G.A.; Petrov, Y.V.; Gruzdkov, A.A. Acoustic strength of water and effect of ultrasound on the liquid-vapor phase diagram.
Technol. Phys. 2015, 60, 753–756. [CrossRef]
41. Petrov, Y.V.; Morozov, N.F. On the Modeling of Fracture of Brittle Solids. J. Appl. Mech. 1994, 61, 710–712. [CrossRef]
42. Volkov, G.A.; Gruzdkov, A.A.; Petrov, Y.V. The incubation time criterion and the acoustic strength of sea water. Acoust. Phys. 2007,
53, 119–122. [CrossRef]
43. Bader, K.B.; Raymond, J.L.; Mobley, J.; Church, C.C.; Felipe Gaitan, D. The effect of static pressure on the inertial cavitation
threshold. J. Acoust. Soc. Am. 2012, 132, 728–737. [CrossRef]
44. Herbert, E.; Balibar, S.; Caupin, F. Cavitation pressure in water. Phys. Rev. E 2006, 74, 041603. [CrossRef]
45. Connolly, W.; Fox, F.E. Ultrasonic Cavitation Thresholds in Water. J. Acoust. Soc. Am. 1954, 26, 843–848. [CrossRef]
46. Barger, J.E. Threshold of Acoustic Cavitation; Technical Memo No.57; Acoustics Research Laboratory, Harvard University: Cam-
bridge, MA, USA, 1964.
47. Gruzdkov, A.A.; Petrov, Y.V. Cavitation breakup of low-and high-viscosity liquids. Technol. Phys. 2008, 53, 291–295. [CrossRef]
48. Blander, M.; Katz, J.L. Bubble nucleation in liquids. AIChE J. 1975, 21, 833–848. [CrossRef]
49. Apfel, R.E. The Role of Impurities in Cavitation-Threshold Determination. J. Acoust. Soc. Am. 1970, 48, 1179–1186. [CrossRef]
50. Petrova, T.; Dooley, R.B. Revised Release on the Surface Tension of Ordinary Water Substance. Proc. Int. Assoc. Prop. Water Steam 2014.
51. Hakala, M.; Nygård, K.; Manninen, S.; Pettersson, L.G.M.; Hämäläinen, K. Intra- and intermolecular effects in the Compton
profile of water. Phys. Rev. B 2006, 73, 035432. [CrossRef]
52. Fu, L.; Bienenstock, A.; Brennan, S. X-ray study of the structure of liquid water. J. Chem. Phys. 2009, 131, 234702. [CrossRef]
53. Frenkel, J. Kinetic Theory of Liquids; Dover: New York, NY, USA, 1955.
54. Zhurkov, S.N. Kinetic concept of the strength of solids. Int. J. Fract. 1984, 26, 295–307. [CrossRef]
55. Regel’, V.R.; Slutsker, A.I.; Tomashevskiı̆, É.E. The Kinetic Nature of the Strength of Solids. Sov. Phys. Uspekhi 1972, 15, 45–65.
[CrossRef]