Chapter 3

Physics of cavitation: GAS CONTENT

AND NUCLEI

Objective: General description of the pressure)as a function of temperature was

behavior of gas bubbles which act as cavitation obtained at 10 degrees Celcius and amounted
nuclei to 277 bars (27.7 MPa)! Although this value
was reduced to some 20 bars near the freezing
point, it remained far away from the equi-
librium vapor pressure. Similar experiments
The beginning of cavitation is called cavi- were carried out much earlier by Berthelot
tation inception. Inception is a special topic, (1950) and Dixon(1909), See [5].It required
because in the initial formation of cavitation careful attention of cleaning and purification
nuclei are involved and once cavitation occurs of the water to obtain these high tensions,
these nuclei loose their meaning. In this chap- and still the theoretical tensile strength of
ter the need of nuclei will be explained and water molecules is even much higher [5]. In
the presence of nuclei in fluids investigated. practical circumstances one may conclude
that pure water does not cavitate. It
should be kept in mind that it is very difficult
to make ”pure” water [7]. Contaminations
3.1 Cohesive Forces of some sort are unavoidable in practice, and
these contaminations determine the tensile
The initial formation of vapor (cavitation strength of water.
inception) is complicated, because pure water The cohesive forces between water molecules
can withstand much lower pressures than the are much greater than the forces between
equilibrium vapor pressure without vaporiza- water molecules and other materials, the
tion. This is due to the mutual attraction or adhesive forces. It means that the next weak
cohesionof the water molecules. A classical spot in water is the wall of the container
experiment dating back to 1950 is that of or the surface of a wing in the low pressure
Lyman J.Briggs [4], a soil physicist who after region. So when a low pressure occurs on a
his retirement (at age 77!) did investigate surface, cavitation will start at the surface.
the tension strength of fluids. He used a However,the tensile strength due to these
capillary Z-shaped tube rotating horizontally adhesive forces between e.g. metals and water
in a centrifuge. This created a calculable is still orders of magnitude higher than the
pressure in the center of the tube. The water equilibrium vapor pressure.
remained inside the tube as long as no vapor
was formed. The maximum tension (negative

9
10 G.Kuiper, Cavitation in Ship Propulsion, January 15, 2010

In practice the main mechanism of inception in which µ is the dynamic viscosity of water
is the growth of small gas bubbles called nu- (related to the kinematic viscosity ν by ν = µρ .
clei,which are present in the water. These nu- The rise velocity of a bubble can then simply
clei are small bubbles, filled with air (or other be written by equating the rise force FD with
gases) and vapor. The diameter can vary from the bubble drag, which results in Stokes’ law:
a few microns to nearly visible bubbles of the
order of 1 mm. That seems queer, because g ∗ D2
Vb = (3.1)
in standing water at atmospheric conditions 18 ∗ ν
gas bubbles will generally disappear due to two This equation is useful to estimate the time
mechanisms: bubble rise and dissolution. To needed for a bubble to rise out of the measure-
understand the behavior of nuclei it is useful ment volume. Consider e.g. a nucleus of 100
to describe these two phenomena first. microns (100 ∗ 10− 6 m)in diameter. As we will
see later in chapter 4, such a nucleus is suf-
ficient in many cases to bring the cavitation
3.2 Rising bubbles inception pressure close to the vapor pressure.
In standing water a bubble will rise to the sur- Such a nucleus reaches a rising speed (from
face due to the difference between the weight eq. 3.1of
of the bubble and the weight of the displaced 9.81 ∗ 10− 8
water. This difference causes a buoyancy force = 0.55 ∗ 10− 2m/sec
18 ∗ 10− 6
on the bubble. Neglecting the weight of the
air in the bubble this force F is equal to the or approximately 5 mm/sec. In a towing tank
weight of water displaced by the bubble: the upper layer of 1 meter depth will be free
of such nuclei after a few minutes. Nuclei of
ρw ∗ π ∗ Db3 ∗ g 10 microns take 100 times as much time, so a
F = few hours.
6
At sea the turbulent motions in the upper
, in which ρw is the specific mass of water and layer of the water are often in sufficient
Db is the bubble diameter. motion to keep nuclei mixed up forever. But
The subsequent rise velocity is determined by after a storm the nuclei content of the upper
the drag Db of the bubble. That is a compli- layer of the sea can contain much more larger
cated matter, but the simplest solution is a so- nuclei. The storm itself creates these nuclei
lution of the Navier-Stokes equation in which by breaking waves and the increased wave
the kinematic forces are negligible relative to motions prevent the rise to the surface or at
the viscous forces. The drag of the bubble least delay it considerably. However, over
is thus completely due to friction along the time they will rise to the surface.
bubble wall wilt a laminar boundary layer and
this condition is expressed in non-dimensional
terms as a low Reynolds number Reb = VbνDb ,
in which nu is the kinematic viscosity of wa- 3.3 Dissolving bubbles
ter and Vb is the bubble velocity. In this equa-
tion it is also assumed that the bubble remains Another mechanism for gas bubbles to disap-
spherical, which is applicable for small bubbles pear is dissolution. In that case the gas bub-
at low velocities. The drag of a bubble can bles disappear because the air dissolves in the
then be written as surrounding water due to diffusion and the va-
por in the bubble condensates because of the
Db = 3πµDb Vb surface tension.
 January 15, 2010, Nuclei11

Water can absorb a lot of air and it is very pressure of that gas in equilibrium with
important to distinguish between dissolved that liquid. The coefficient is called Henry’s
air and free air. Free air is present in the form constant, but this ”constant” is temperature
of bubbles. The majority of the air in water is dependent.
dissolved and dissolved air has no significant When the water is not saturated, there will
effect on the cohesive forces in the water. be a mass flow through the bubble wall and
At 15 degrees Celcius in atmospheric condi- the bubble will disappear because the free
tions the amount of air that can be dissolved air goes into solution. Even when the water
in one cubic meter of water is about 25-30g is saturated, small bubble will disappear
(25-30 liters in atmospheric conditions). because the pressure inside a bubble is higher
This is expressed as the gas concentration than in the surrounding fluid due to the
C = 3 ∗ 10− 2kg/m3 . In that case the water surface tension, as will be discussed later.
is saturated. In that condition there is no
net transfer of air molecules through the free Contrary to evaporation, diffusion is a
surface. The air content of water is often relatively slow process. Evaporation takes
expressed as the fraction of the maximum place at the surface of a bubble only, but
amount of gas that can be absorbed: the diffusion extends over a much thicker layer
saturation rate. For the calculation of the of fluid around the bubble. At the bubble
air content in in various units see Appendix wall the saturation rate of the water will
A. Note that the saturation rate can be immediately be such that the partial pressure
exceeded considerably without gas coming of the dissolved air is equal to the partial
out of solution immediately. In that case the pressure of the air in the bubble. When the
water is oversaturated or supersaturated and saturation rate of the bulk of the water is
this can amount to two or three times the different, air molecules will migrate to or
saturation rate. from the bubble. Over a certain distance a
concentration gradient will form.
As mentioned bubbles in water contain air
The mass flow m in kg/m3 through the wall
and vapor. Dalton’s law states that the total
of a bubble with radius R can be written as
pressure exerted by a gaseous mixture is equal
to the sum of the partial pressures of each
individual component in a gas mixture. This dm dCg
= 4πR2 Dg ( )R (3.2)
empirical law was observed by John Dalton in dt dr
1801. So the pressure inside of a bubble is the in which Dg is a diffusion coefficient
sum of the partial pressures of air and vapor. (m2 /sec) and Cg is the gas concentration in
The partial pressure of vapor is the equi- (kg/m3 ). The gradient ( ∂C g
∂r R
) is the gradient
librium vapor pressure. In many conditions of the gas concentration at the bubble wall.
the vapor pressure can be neglected and a When the vapor is neglected and the gas den-
bubble can be considered as a pure gas bubble. sity remains the same the mass change is
dm d(4/3πR3 )
Transport of gas through the bubble wall = ρg
dt dt
is governed by Henry’s law (formulated by
and from eq.3.2 the bubble radius changes as
William Henry in 1803), which states that
at a constant temperature, the amount of a
given gas dissolved in a given type and volume dR Dg dCg
= ( )R (3.3)
of liquid is directly proportional to the partial dt ρg dr
12 G.Kuiper, Cavitation in Ship Propulsion, January 15, 2010

Eppstein and Plesset [9] gave an approxi- (πDt)1/2 is 5 ∗ 10− 4, which is still an orders
mate solution for the concentration gradient: of magnitude larger than the initial radius
R0 = 5−5 . It is therefore justified to neglect
this term in eq.3.4.
dCg 1 1
( )R = ( + )(Cg∞ − CgR )
dr R (πDg t)1/2
Surface tension will reduce this time for a
(3.4) bubble to dissolve, but these approximations
can be used to determine a time scale. It fol-
in which the length (πDg t)1/2 can be re- lows that the diffusion rate is low when com-
garded as a diffusion boundary layer thickness. pared to the time scale of bubble dynamics in
For small bubbles this thickness is much larger a flow. However, when the existence of gas
than the radius R, so eq. 3.4 can be written bubbles in a fluid is considered, small bubbles
as will tend to go into solution.
dCg (Cg∞ − CgR )
( )R = (
dr R
and eq.3.3 can be integrated and written as 3.4 Mechanisms to sustain
nuclei
2Dg (Cg∞ − CgR )
R2 = R0 2 + t (3.5) From the foregoing it is still unexplained
ρg why there are so many nuclei in water.
The time to go completely into solution is One mechanism that can keep nuclei in the
when R=0 in eq.3.5 and when Cg∞ − CgR is water without rising to the surface is the
negative: combination of gas and solid particles. For
that mechanism we have to consider the
surface tension of a fluid. The difference in
R0 2 ρg bond between cohesion between the water
t= (3.6)
2D(CgR − Cg∞ ) molecules and adhesion between water and
air molecules causes surface tension. The
The diffraction coefficient Dg is close to cohesion attracts the molecules at a free sur-
2 ∗ 10−9 m2 /sec. face and prevent these molecules to mix with
the air. The result is also a tension on the
An example can give a rough idea about surface molecules, which curves the surface in
time scales. Consider a bubble with a diame- the neighborhood of a solid wall. The surface
ter of 100 microns (R = 0.5 ∗ 10− 5m) in water tension makes that a small gas bubble tends
at atmospheric pressure with a saturation to be spherical. The surface tension increases
rate of 50%. The saturation rate of the the pressure inside a spherical gas bubble
water at the bubble wall will be 100 %. The with an amount of 2s/Rb Pa, in which s is
concentration gradient is then estimated as the surface tension and Rb the bubble radius.
1.5 ∗ 10−2 kg/m3 and with ρg = 1 this means The surface tension of water in contact with
(C −C )
that g∞ρg gR = −1.5 ∗ 10−2. From eq.3.6 air at 15 degrees Celcius is 0.075 Nm and
it follows that the time for the bubble to the sensitivity to the temperature is not very
dissolve is 41 seconds. When submerged at strong.
1 m below the surface, such a bubble will
not reach the surface (at 5mm/sec), but go Consider again a nucleus of 0.1mm diame-
into solution. At a time t = 41 the term ter in a fluid of 1 bar (105 P a). The pressure
 January 15, 2010, Nuclei13

Figure 3.1: Contact Angle

inside the bubble is higher than the surround-

ing pressure due to the surface tension. The Figure 3.2: Gas Pocket in a Crevice
difference is 0.15/10− 4 = 1500P a. Note that
this is of the same magnitude as the vapor
pressure. When the bubble is only 10 microns working on bioluminescence and interested in
in diameter the surface tension increases the nuclei in blood.
pressure with 15000 Pa!. The surface tension Diffusion can also be prevented by a molec-
will increase the diffusion of gas into the fluid ular skin on the surface of the nucleus, as was
and thus decrease the time required to go into already assumed by Fox and Herzfeld in 1964
solution and this is increasing with decreasing [14]. A variation on this assumption was used
bubble size. by Yount [50], who assumes a skin of surface
active molecules that make the nucleus stable.
Surface tension also causes a contact angle Morch [38],[39] showed that gas could not only
at the interface of water-gas-solid. This con- be present in crevices, as Harvey assumed, but
tact angle depends on the molecular forces in also in slightly concave surfaces on solid parti-
the water, the gas (air) and the material. Two cles and that the effective radius of these thin
types of material can be distinguished. On hy- gas remains could be larger than the size of
drophobic surfaces the contact angle is larger the solid particle.
than 90 degrees, and air bubbles have the ten- Experiments have shown that solid particles
dency to stick to the surface. In hydrophilic are important for the tensile strength of water.
surfaces the contact angle is smaller than 90 Filtering of water delays cavitation inception
degrees and air bubbles have the tendency to and increases the tensile strength of the water
detach from the surface (Fig. 3.1). [24]. But also pre-pressurization can strongly
When the surface is not flat but has the increase the tensile strength of water by driv-
shape of a crevice, the schematic model is as ing the nuclei attached to the solid particles
in Fig. 3.2. into solution before the low pressure occurs.
The material in Fig.3.2 is hydrophobic. The Nuclei with an organic skin seem to be crushed
pressure inside the gas bubble will be lower at some pressure and thus can go into solution
than in the water due to the surface tension. also.
When the water is not saturated an equilib-
rium is possible where no net gas diffusion
takes place and the bubble is stable. The den- 3.5 Mechanisms to gener-
sity of the particle and bubble together can ate nuclei
in principle be equal to water and the nucleus
is stable. Note that this model was first used As discussed in previous sections, diffusion and
by Edmund Newton Harvey [16], a biologist gravity will remove nuclei from water. How-
14 G.Kuiper, Cavitation in Ship Propulsion, January 15, 2010

ever, in general there are still ample nuclei in

regular tap water and also in sea water. At
sea, the breaking of waves is a major source
of bubbles in the water and the mixing effect
of these waves and the related turbulence will
prevent small nuclei to rise to the surface. Bi-
ological life in the water can also produce free
gases and in combination with biological mat-
ter it can be a source of nuclei. Actually the
tensile strength of sea water in prototype con-
ditions is close to zero and cavitation inception
will thus take place at the vapor pressure. //
In test facilities this is different. First of all
the amount of organic or solid matter may be
less in clean test facilities, but also the size
of the nuclei required for inception is larger.
And these larger nuclei are not always avail-
able in sufficient numbers. So in test facilities
like cavitation tunnels and depressurized tow-
ing tanks the nuclei content is always a point Figure 3.3: Nuclei densities
of concern. Locations where nuclei are formed
are low pressure regions in a flow, especially
when sharp edges cause local flow separation.
clei per cubic meter or, more imaginable, 50
These edges may occur on a very small scale,
nuclei per cubic centimeter. The average dis-
such as on surface roughness.
tance between the nuclei is then 2.5mm.
Fig. 3.3 also shows that this amount can eas-
3.6 Nuclei measurements ily vary with a factor 100 depending on the
circumstances. Even when there are 50 ∗ 106
It is not easy to measure the size distribution nuclei of in average 75 microns in diameter in
of nuclei. These techniques will be discussed 1m3 of water, this means that there is a volume
later. Measurements indicate that there are of 1.1 ∗ 10−5 m3 of free gas per m3 of water. In
many small nuclei in water. Gates [15] com- atmospheric conditions this is also 1.1∗10−5 kg
piled such measurements in his thesis and and this is the free air content in 1m3 . This is
shown in Fig. 3.3. The size range of the nuclei still only a third of the dissolved air content of
measured is from 5 microns to 500 microns. C = 3 ∗ 10− 2kg/m3 . In general the amount of
The number of nuclei is expressed as the num- free air will be (much) smaller then the amount
ber density N d, which is the number of nuclei of dissolved air. So when there is a mechanism
in a small small size range divided by the size to bring air out of solution and create bubbles,
range. Its unit is therefore m− 4. this will not immediately deplete the dissolved
The range of number densities is between 1010 air content. It is also important to note that
and 1014 . It helps to get a feeling for those the dissolved air content has little or no effect
numbers. When the number density is 1012 on cavitation inception.
and we consider nuclei in the range of 50-100 From Fig.3.3 it can be seen that the nuclei
microns (a size range of 50∗10∗−6m) it means are predominantly small. These small nuclei
that there are 50 ∗ 10−6 ∗ 1012 = 50 ∗ 106 nu- can easily be transported through the fluid by
 January 15, 2010, Nuclei15

currents and water motions such as waves. But

they would still dissolve over time . This seems
to be significantly reduced by the presence of
an organic skin on the surface of a nucleus.
When the nuclei dissolve, the reduction of the
radius makes the skin thicker and this ulti-
mately stops the diffusion, making the nuclei
stable.
Since in sea water ample nuclei are available
not much attention has been given to their ori-
gin and presence. However, for model tests the Figure 3.4: Path of a Nucleus in the Pressure
required nuclei size is larger, as explained in Field near the Leading Edge of a Foil
chapter refchincept, and these nuclei are often
scarce or not available in test facilities. This
boundary layer (the pressure in a thin bound-
makes that nuclei distributions dominate the
ary layer is constant perpendicular to the
problem of scaling of cavitation inception and
wall). This pressure gradient perpendicular
the mechanisms of dissolving and rising nuclei
to the wall is is especially strong near the
have to be considered when test facilities are
stagnation point and, in opposite direction,
used.
near the minimum pressure location. Near
the stagnation point the pressure gradient
will move the nuclei away from the wall. In
3.7 Bubble Screening the minimum pressure region the nuclei are
A phenomenon which is often observed is moved towards the wall. As a result the path
that cavitation does not take place in the of the nuclei will not follow the streamlines,
minimum pressure location with a minimum but follow a path away from the wall, as
pressure significantly below the vapor pres- sketched in Fig 3.4. The path of the nuclei
sure. This occurs even in conditions with may thus avoid the minimum pressure close
ample free nuclei in the incoming flow. Since to the wall. The pressure gradient thus has a
this is observed spedifically in cases when screening effect and this effect will be greater
the boundary layer is still laminar in hte with increasing size of the nuclei.
minimum pressure region, it was suspected Calculations of the path of nuclei were first
that this delay in inception was caused by made by Johnsson and Hsieh [22] in 1966,
a local lack of nuclei. It means that the assuming a certain drag coefficient of nuclei
incoming flow contains enough nuclei , but and assuming that they remained spherical.
that they are inhibited to enter the low
pressure region. The interpretation of what Although bubble screening is a plausible hy-
happens is only tentative because it is very pothesis, experimental evidence is difficult to
difficult to verify experimentally. A possible obtain.
explanation is The common explanation is
that bubble screening occurs. In a pressure
gradient nuclei experience a force, similar to
the rising force in a gravitational field. Near
the leading edge of a foil there is a strong
pressure gradient along the foil surface, but
also perpendicular to the wall outside the
Bibliography

[1] Arakeri,V.H., Acosta, A.J., 1979, Vis- [9] Epstein, P.S., Plesset, M.S. 1950 On the
cous Effects in the Inception of Cavita- stability of Gas Bubbles in Liquis-Gas
tion, A.S.M.E. Int. Symposium on Cavi- Solutions, Journal of Chemical Physics,
tation, New York, USA. Vol.18., pp1505-1509.

[2] Arndt, R.E.A., 1976, Cavitation on Model [10] , Feindt, E.G., 1956,Untersuchungen
Propellers with Boundary Layer Trips, uber die Abhangigkeit des Umschlages
A.S.M.E.Conference on Polyphase Flow, Laminar-Turbulent von der Oberflachen-
New Orleans, USA. rauhigkeit und der Druckverteilung,
Jahrbuch STG, Bd.50, pp180-205
[3] Arndt, R.E.A., Ellis, C.R., Paul, S. 1995,
Preliminary Investigation of the Use of [11] Flynn, H.G., 1964, Physics of
Air Injection to mitigate Cavitation Ero- Acoustic Cavitation, W.P.Mason ed.
sion, A.S.M.E. Journal of Fluids Engi- Vol1,PartB,Academic Press New York-
neering, Vol.117. London.
[4] Briggs, L.J., 1950 Limiting negative
[12] Foeth, E-J., Kuiper, G. 2004, Ex-
pressure of water., J. Applied Phys.
ploratory experiments to determine flow
21:721722.
and structure borne noise of erosive
cavity implosions, A¿S¿M¿E¿ Fluids
[5] Brennen, C.E., 1995 Cavitation and Bub-
Eng. Summer Conference, HT-FED2004-
ble Dynamics, Oxford University Press,
56789, Charlotte, NC. USA.
ISBN 0-19-509409-3. Also available on in-
ternet.
[13] Foeth, E-J., The Structure of Three-
[6] Burrill, L.C., 1951, Sir Charles Parsons Dimensional Sheet Cavitation, Thesis
and Cavitation, Transactions of the Insti- Technical University Delft.
tute of Marine Engineers.
[14] Fox, F.E., Herzfeld, K.F., 1954 Gas Bub-
[7] Borkent, B.M., 2009, Interfacial Phe- bles with Organic Skin as Cavitation Nu-
nomena in Micro- and Nanofluidics: clei, J.Acoust. Soc. Am. Vol.26, pp 984-
nanobubbles,cavitation, and wetting, Thi- 989.
sis Twente University, The Netherlands.
[15] Gates, E.M., 1977 ( The influence of Free-
[8] Coutier-Delgosha, O., Devillers, J-P., Stream Turbulence, Free Stream Nuclei
Leriche, M., Pichon, T., 2005, emphEf- Populations and a Drag-reducing Poly-
fect of Roughness on the Dynamics of Un- mer on Cavitation Inception on Two Ax-
steady Cavitation, Journal of Fluids Eng. isymmetric Bodies California Institute of
, vol 127. Technology, Rep. No. Eng 183-2.

65
66 G.Kuiper, Cavitation in Ship Propulsion, January 15, 2010

[16] Harvey, E.N., McElroy, W.D., Whiteley, [25] KLEBANOFF, P.S., SCHUBAUER,
A.H., 1947 On Cavity Formation in Wa- G.B., TIDSTROM,K.D., 1955, Measure-
ter, Journal of Applied Physics, Vol.18, ments of the Effect of Two-dimensional
pp162-172. and Three-dimensional Roughness Ele-
ments on Boundary Layer Transition, J.
[17] Hoekstra, M., Vaz, G. The Paartial Cav- Aeron. Sciences
ity on a @D Foil Revisited, 7th Int. Con-
ference on Cavitation CAV2009, Ann Ar- [26] Knapp, R.T., Hollander, A., 1948, Lab-
bor, Michigan, U.S.A. oratory Investigations of the Mechanism
of Cavitation, Trans. A.S.M.E., Vol.70,
[18] Holl, J.W., 1960, An Effect of Air Con- pp.419-435.
tent on the Occurrence of Cavitation,
Trans. A.S.M.E., Journal of Basic Eng. [27] Knapp, R.T., Daily, J.W., Hammitt,
Vol.82, pp941-946. F.G., 1970, C avitation, New York :
McGraw-Hill.
[19] Holl, J.W., Carrol, J.A., 1981, Obser-
[28] Knapp, R.T., Hollander, A., 1948,
vations of the Various Types of Lim-
Laboratory Investigations of the Mecha-
ited Cavitation on Axisymmetric Bodies,
nism of Cavitation, Trans. A.S.M.E., Vol
PUB A.S.M.E. Journal of Fluids Eng.,Vol
70, pp419-435.
103,pp415-423
[29] Korkut, E., Atlar, M., 2000, On the Im-
[20] Huang, T.T., 1981, Cavitation Inception
portance of Effect of Turbulence in Cav-
Observations on Six Axisymmetric Head-
itation Inception Tests of Marine Pro-
forms, A.S.M.E. Journal of Fluids Engi-
pellers, Proceedings of Royal Society of
neering, Vol 103,PP273-278
London A: Mathematical, Physical and
Engineering Sciences, Vol.458 , pp.29-48.
[21] ITTC, 1978,Proceedings 15th ITTC, The
Haque, Report of the Performance Com- [30] Kreider, W., Crum, L., Bailey,M., Mat-
mittee. ula, T.,Khoklova, V., Sapozhnikov,O.,
2006, Acoustic Cavitation and Medical
[22] Johnsson, C.A., Hsieh,T., 1966, The In- Ultrasound,Sixth International Sympo-
fluence of Trajectories of Gas Nuclei on sium on Cavitation CAV2006, Wagenin-
Cavitation Inception, 6th Symposium on gen, The Netherlands.
Naval Hydrodynamics, Washington D.C.,
pp163-178. [31] Kumar, S., Brennen, C.E., 1996, A Study
of Pressure Pulses generated by Travelling
[23] Takagi,K., Kato,H.,Kato,D.,Sugimoto, Bubble Cavitation, Journal of Fluid Me-
A., 2006, Destruction of Plankton by chanics Vol 255 pp541.
Two-Dimensional Cavitating Jet, Sixth
International Symposium on Cavitation [32] Hwansung Lee, Tomonori Tsukiya,
CAV2006, Wageningen, The Netherlands. Akihiko Homma, Tadayuki
Kamimura,Eisuke Tatsumi, Yoshiyuki
[24] Keller, A.P., 1974 Investigations Con- Taenaka, Hisateru Takano, Observation
cerning Scale Effects of the Inception of of Cavitation in a Mechanical Heart
Cavitation, Proc. I.Mech.Eng. Conference Valve, Fifth International Symposium on
on Cavitation, Edinburgh, 109-117. Cavitation (cav2003),Osaka, Japan.
 January 15, 2010, Leading Edge Roughness67

[33] Kuiper, G., 1978, Scale Effects on Pro- Standard Bodies, St. Anthony Fall Hy-
peller Cavitation, 12th Symposium on draulic Lab, Univ. of Minnesota, report
Naval Hydrodynamics, Washington D.C., 118.
USA.
[44] Schlichting, H., 1968, Boundary Layer
[34] Kuiper, G. 1981, Cavitation Inception on Theory, McGraw-Hill, 6th edition.
Ship Propeller Models, Thesis Technical
University Delft. [45] Terwisga, T.J.C., Fitzsimmons, P.A.,
Li,Z., Foeth, E-J. Cavitation Erosion-A
[35] Kuiper, G. 2008, Fundamentals of Ship review of Physical Mechanisms and Ero-
Resistance and Propulsion, Course Lec- sion Risk Models, 7th International Sym-
tures Technical University Delft. posium on Cavitation, CAV2009, Ann
Arbor, Michigan, U.S.A.
[36] Landa, E.R., Nimmo, J.R.,2003 The Life
and Scientific Contributions of Lyman J. [46] Jin Wang, 2009, Nozzle-geometry-
Briggs, Journal of the Soil Science Soci- dependent breakup of diesel jets by
ety of America, Vol 67 no 3, pp 681-693. ultrafast x-ray imaging: implication of
http://soil.scijournals.org/cgi/reprint/67/3/681
in-nozzle cavitation, Seventh Int. Symp.
on Cavitation: CAV2009, Ann Arbor,
[37] Ligtelijn, J.T., van der Kooij, J., Kuiper, Michigan, U.S.A.
G., van Gent,W., 1992, Research on
Propeller-Hull Interaction in the Depres- [47] Watanabe, S., Furukawa, A., Yoshida, Y.,
surized Towing Tank, Hydrodynamics, Tsujimoto, Y., 2009, Analytical investiga-
Computations,Model Tests and Reality tions of thermodynamic effect on cavita-
(Marin ), Elseviers Science Publishers. tion characteristics of sheet and tip leak-
age vortex cavitation, Seventh Int. Symp.
[38] Morch E.M., 2000, Paper on Cav2003. on Cavitation: CAV2009, Ann Arbor,
Michigan, USA.
[39] Moerch, K.A. 2009, C avitation Nuclei:
Experiment and Theory,Journal of Hy- [48] Williams,M., Kawakami, E., Amromin,
drodynamics Vol21 p176. E., Arndt,R. Effects of Surface Charac-
teristics on Hydrofoil Cavitation, Seventh
[40] Neppiras, E.A., Noltink, B.E., 1951, Cav- Int. Symp. on Cavitation: CAV2009, Ann
itation produced by Ultrasonics, Proc. Arbor, Michigan, USA.
Phys. Soc. London, pp 1032-1038.
[49] Yoshimura,T., Kubota, S., Seo, T.,
[41] Ohl, C-D., Arora, M. ,Roy, I., Delius. Sato,K., 2009, Development of Bal-
M., Wolfrum, B.,2003, Drug Delivery Fol- last Water Treatment Technology by
lowing Shock Wave Induced Cavitation , Mechanochemical Cavitations,Seventh
Fifth International Symposium on Cavi- Int. Symp. on Cavitation: CAV2009,
tation (cav2003),Osaka, Japan. Ann Arbor, Michigan, USA.

[42] Plesset, M.S., 1949, The Dynamics of [50] Yount, D.E. 1979, S kins of Varying Per-
Cavitation Bubbles, ASME Journal of meability: A Stabilization Mechanism
Appl. Mech. 1949, pp 277-232. For Gas Cavitation Nuclei,A. ACCOUST.
SOC. AM. 65(6).
[43] Schiebe, F.R., 1972, Measurement of the
Cavitation Susceptibility of Water using
Appendix A

Air Content of Water

The amount of air dissolved in water α can with saturated water therefore contain 36 per-
be expressed in many ways. The most common cent oxygen. But nuclei which are generated
ways in literature are from the air above the water contain 21 per-
cent oxygen. Since the ratio between oxygen
• the gas fraction in weight ratio αw and nitrogen is not fixed, it is difficult to re-
late measurements of dissolved oxygen (by os-
• the gas fraction in volume ratio αv
mose) to measurements of dissolved air (from
• the molecule ratio e.g a van Slijke apparatus).
The amount of oxygen dissolved in water at
• the saturation rate
atmospheric pressure at 15 degrees Celcius is
• the partial pressure of air approximately 10 ∗ 10− 6kg/kg. For nitrogen
this value is about 15 ∗ 10− 6, so the solubility
of air in water is the sum of both: 25 ∗ 10− 6.
A.1 Solubility Here the dissolved gas contents are expressed
as a weigth ratio αw .Air is very light relative
Air is a mixture of 21 percent oxygen, 78 per- to water and the weight ratio is very small.
cent nitrogen and one percent of many other This ratio is therefore often expressed as parts
gases, which are often treated as nitrogen. The per million (in weight), which is 106 ∗ αw .
specific mass of gases involved in air are:

Oxygen (O2 ) 1.429 kg/m3

Nitrogen (N2 ) 1.2506 kg/m3
Air 1.292 kg/m3 A.2 The Gas Fraction in
Volume Ratio
The maximum amount of gas that can be
dissolved in water, the solubility), depends on The volume of gas dissolved per cubic meter
pressure and temperature. It decreases with of water depends on temperature and pres-
increasing temperature and increases with in- sure. Therefore this volume ratio is expressed
creasing pressure. The solubility of oxygen in in standard conditions of 0 degrees Celcius and
water is higher than the solubility of nitrogen. 1013 mbar (atmospheric conditions). The de-
Air dissolved in water contains approximately pendency of the volume of water on tempera-
36 percent oxygen compared to 21 percent in ture and pressure is neglected. The volume of
air.The remaining amount can be considered the dissolved air is then described by the law
as Nitrogen. Nuclei which are in equilibrium of Boyle-Gay-Lussac:

69
70 G.Kuiper, Cavitation in Ship Propulsion, January 15, 2010

be defined using the ratio of oxygen and nitro-

p ∗ V ol gen of 21/79 this virtual molar weight of air is
= constant (A.1) about 29.
273 + T
The volume fraction at (p,T) can be related
to the volume fraction in standard conditions: A.4 The saturation rate
273p The saturation rate is the amount of gas in so-
αv = αv (p, T ) (A.2) lution as a fraction of the maximum amount
(273 + T )1013
that can go in solution in the same conditions.
The gas fraction in volume ratio is dimen- Since the saturation rate is dimensionless. It is
sionless (m3 /m3 ). Be careful because some- independent of the way in which the dissolved
times this is violated by using cm3 /l (1000∗αv ) gas or the solubility is expressed. The satura-
or parts per million (ppm) which is 106 ∗ αv . tion rate is important because it determines if
αv is found from αw by: and in which direction diffusion will occur at
a free surface. The saturation rate varies with
ρwater temperature and pressure, mainly because the
αv = αw (A.3) solubility of gas changes with these parame-
ρair
ters.

in which ρ is the specific mass in kg/m3 . At

15 deg. Celcius and 1013 mbar pressure the
A.5 The partial pressure
specific mass of water ρw = 1000kg/m3 and Sometimes the amount of dissolved gas is ex-
the specific mass of air is 1.223kg/m3 , so for pressed as the partial pressure of the gas (mbar
air αv = 813αw . or even in mm HG). This is based on Henry’s
law, which states that the amount of gas dis-
A.3 The Gas Fraction in solved in a fluid is proportional to the partial
pressure of that gas. In a van Slijke appa-
Molecule ratio ratus a specific volume of water is taken and
subjected to repeated spraying in near vac-
The dissolved amount of gas can also be ex- uum conditions (a low pressure decreases the
pressed as the ratio in moles(Mol/Mol). Mo- solubility). This will result in collecting the
lar masses may be calculated from the atomic dissolved in a chamber of specific size. By
weight in combination with the molar mass measuring the pressure in that chamber the
constant (1 g/mol) so that the molar mass of a amount of dissolved gas is found. Note that
gas or fluid in grams is the same as the atomic this pressure is not directly the partial pres-
weight. sure. A calibration factor is required which
The molar ratio αm is easily found from the depends on the apparatus.
weight ratio by

M( water)
αw = αm (A.4)
M( gas)
in which M is the molar weight, which is 18
for water, 16 for oxygen(O2 ) and 28 for Nitro-
gen (N2 ). For air a virtual molar weigth can
Appendix B

Standard Cavitators

A standard cavitator is a reference body

which can be used to compare and calibrate
cavitation observations and measurements.
Its geometry has to be reproduced accurately
and therefore an axisymmetric headform has
been used as a standard cavitatior.

Such an axisymmetric body has been in-

vestigated in the context of the ITTC (In-
ternational Towing Tank Conference).This is
a worldwide conference consisting of towing
tanks (and cavitation tunnels) which have the
goal of predicting the hydrodynamic behavior
of ships. To do that model tests and calcula- Figure B.1: Contour and Pressure Distribu-
tions are used. They meet every three years tion on the ITTC Headform [31]
to discuss the state of the art and to define
common problem areas which have to be re-
viewed by committees. The ITTC headform headform was used to compare inception
has a flat nose and an elliptical contour [22]. measurements in various cavitation tunnels.
Its characteristics are given in Fig B.1. However, it was realized later on that the
This headform has been used to compare boundary layer flow on both the ITTC and
cavitation inception conditions and cavitation on the hemisperical headform was not as
patterns in a range of test facilities. The simple as the geometry suggested. In most
results showed a wide range of inception cases the Reynolds numbers in the investi-
conditions and also a diversity of cavitation gations was such that the boundary layer
patterns in virtually the same condition, over the headform remained laminar and the
as illustrated in Fig B.3. This comparison pressure distribution was such that a laminar
lead to the investigation of viscous effects on separation bubble occurred, in the position
cavitation and cavitation inception. indicated in Fig. B.1. This caused viscous
effects on cavitation inception and made the
The simplest conceivable body to investi- headform less suitable as a standard body.
gate cavitation is the hemispherical headform. Note that the location of laminar separation is
This is an axisymmetric body with a hemis- independent of the Reynolds number. When
pere as the leading contour. Its minimum the Reynolds number becomes high transition
pressure coefficient is -0.74. The hemisperical to turbulence occurs upstream of the sep-

71
72 G.Kuiper, Cavitation in Ship Propulsion, January 15, 2010

Figure B.2: Contour and Pressure Distribu-

tion of the Schiebe body [31]

aration location and separation will disappear.

To avoid laminar separation another head-

form was developed by Schiebe ([43]) and
this headform bears his name ever since.
The contour and pressure distribution on
the Schiebe headform are given in Fig. B.2.
This headform has no laminar separation and
transition to a turbulent boundary layer will
occur at a location which depends on the
Reynolds number.

Many other headform shapes have been

investigated with different minimum pres-
sure coefficients and pressure recovery
gradients.(e.g.[20])
 January 15, 2010, Standard Cavitators73

Figure B.3: Comparative measurements of cavitation inception on the ITTC headform

source:ITTC
Appendix C

Tables

T pv
Celcius N/m2
0 608.012
2 706.078
4 813.951
6 932
8 1069
10 1226
12 1402
14 1598
15 1706
16 1814
18 2059
20 2334
22 2638
24 2981
26 3364
28 3785
30 4236
32 4756
34 5315
36 5943
38 6619
40 7375

Table C.1: Vapor pressure of Water.

75
76 G.Kuiper, Cavitation on Ship Propellers, January 15, 2010

Temp. kinem. visc. kinem. visc.

deg. C. fresh water salt water
m2 /sec × 106 m2 /sec × 106
0 1.78667 1.82844
1 1.72701 1.76915
2 1.67040 1.71306
3 1.61665 1.65988
4 1.56557 1.60940
5 1.51698 1.56142
6 1.47070 1.51584
7 1.42667 1.47242
8 1.38471 1.43102
9 1.34463 1.39152
10 1.30641 1.35383
11 1.26988 1.31773
12 1.23495 1.28324
13 1.20159 1.25028
14 1.16964 1.21862
15 1.13902 1.18831
16 1.10966 1.15916
17 1.08155 1.13125
18 1.05456 1.10438
19 1.02865 1.07854
20 1.00374 1.05372
21 0.97984 1.02981
22 0.95682 1.00678
23 0.93471 0.98457
24 0.91340 0.96315
25 0.89292 0.94252
26 0.87313 0.92255
27 0.85409 0.90331
28 0.83572 0.88470
29 0.81798 0.86671
30 0.80091 0.84931

Table C.2: Kinematic viscosities adopted by

the ITTC in 1963
 January 15, 2010, Tables 77

Rn Cf × 103
1 × 105 8.333
2 6.882
3 6.203
4 5.780
5 5.482
6 5.254
7 5.073
8 4.923
9 4.797
1 × 106 4.688
2 4.054
3 3.741
4 3.541
5 3.397
6 3.285
7 3.195
8 3.120
9 3.056
1 × 107 3.000
2 2.669
4 2.390
6 2.246
8 2.162
1 × 108 2.083
2 1.889
4 1.721
6 1.632
8 1.574
1 × 109 1.531
2 1.407
4 1.298
6 1.240
8 1.201
1 × 1010 1.17x

Table C.3: Friction coefficients according to

the ITTC57extrapolator.
78 G.Kuiper, Cavitation on Ship Propellers, January 15, 2010

Temp. density density

deg. C. fresh water salt water
kg/m3 kg/m3
0 999.8 1028.0
1 999.8 1027.9
2 999.9 1027.8
3 999.9 1027.8
4 999.9 1027.7
5 999.9 1027.6
6 999.9 1027.4
7 999.8 1027.3
8 999.8 1027.1
9 999.7 1027.0
10 999.6 1026.9
11 999.5 1026.7
12 999.4 1026.6
13 999.3 1026.3
14 999.1 1026.1
15 999.0 1025.9
16 998.9 1025.7
17 998.7 1025.4
18 998.5 1025.2
19 998.3 1025.0
20 998.1 1024.7
21 997.9 1024.4
22 997.7 1024.1
23 997.4 1023.8
24 997.2 1023.5
25 996.9 1023.2
26 996.7 1022.9
27 996.4 1022.6
28 996.2 1022.3
29 995.9 1022.0
30 995.6 1021.7

Table C.4: Densities as adopted by the ITTC

in 1963.
Appendix D

Nomenclature
kg
ρ density of water m3
See TableC.4
3
Cg gas concentration kg/m see Appendix A
Dg diffusion coefficient m2 /sec representative value 2 ∗ 109
D diameter m
Fd drag N
m
g acceleration due to gravity sec2
Taken as 9.81
−
N d number density of nuclei m 4
pg gas pressure f racN m2
f racN m2
pv equilibrium vapor pressure
R radius m
kg
µ dynamic viscosity of water m∗sec
m2
ν kinematic viscosity of water sec
(ν = µρ )See Table C.2
s surface tension Nm for water 0.075

79