To be published in Journal of Ship Research 2008

Numerical Study of Cavitation Inception due to

Vortex/Vortex Interaction in a Ducted Propulsor
C.-T. Hsiao and G. L. Chahine
DYNAFLOW, INC.
10621-J Iron Bridge Road, Jessup, MD 20794
Email: [email protected], [email protected]
http://www.dynaflow-inc.com
ABSTRACT

Cavitation inception in a ducted propulsor was studied numerically using NavierStokes computations and bubble dynamics models. Experimental observations of the
propulsor model and previous numerical computations using Reynolds-Averaged
Navier-Stokes (RANS) codes indicated that cavitation inception occurred in the region
of interaction of the leakage and trailing tip vortices. The RANS simulations failed
however to predict correctly both the cavitation inception index value and the inception
location. To improve the numerical predictions, we complemented here the RANS
computations with a direct Navier-Stokes simulation in a reduced computational domain
including the region of interaction of the two vortices. Initial and boundary conditions in
the reduced domain were provided by the RANS solution of the full ducted propulsor
flow. Bubble nuclei were released in this flow field, and spherical and non-spherical
bubble dynamics models were exercised to investigate cavitation inception. This resulted
in a solution in much better agreement with the experimental measurements than the
original RANS solution. Both the value of the cavitation inception index and the location
of the cavitation inception were very well captured. The characteristics of the emitted
acoustic signals and of the bubble shapes during a cavitation event were also computed.

To be published in Journal of Ship Research 2008

1. Introduction
Prediction of vortex cavitation inception on marine propulsors is of great interest
to naval applications and has been the subject of many studies seeking scaling laws for
the prediction of cavitation inception. Existing scaling laws are typically formulated
using data from open propellers and are not applicable to ducted propellers. Unlike most
open propellers, which generally form a single trailing vortex, a ducted propeller
typically exhibits two well-defined vortices: a trailing vortex formed near each blade tip
trailing edge, and a stronger leakage vortex generated in the gap region between the blade
and the duct wall. These two unequal co-rotating vortices introduce small-scale unsteady
motions during vortex merging that add to the fluctuations resulting from upstream
turbulence and vortex wandering1,2
Recent experimental observations of cavitation inception on a ducted propulsor3
have indicated that the interaction between the tip-leakage vortex and the trailing-edge
vortex cause cavitation inception to occur in the region where the two vortices merge.
However, predictions of cavitation inception using the pressure field obtained by
Reynolds-Averaged

Navier-Stokes

(RANS)

computations5,6,7

or

inferred

from

experimental velocity measurements4 were in poor agreement with the experimental

observations in terms of both cavitation inception number and inception location.
A preliminary controversial conclusion made by Chesnakas and Jessup3 was that
cavitation inception does not necessarily occur in the minimum pressure region. This
conclusion, however, was drawn based on the inferred pressure field obtained from the
measured average tangential velocities, using an assumed Rankine vortex profile. This
not only neglects the effect of the axial velocities on the pressure field, but also relies on

To be published in Journal of Ship Research 2008

time-averaged tangential velocities which could be significantly smaller than the
instantaneous values, especially at downstream locations where vortex wandering may be
significant. RANS computations with inadequate turbulence models and grid resolution
can also result in over-diffusion and dissipation of the vortex flow8,9. This usually leads
to a significant under-prediction of the velocities in the vortex core at the downstream
locations. In a combined numerical and experimental study of a tip vortex flow, DaclesMariani et al.8 used the measured flow field to specify the inflow and outflow boundary
conditions of the vortex flow and investigated vortex preservation in the wake region.
With the turbulence model turned off and significant grid refinement they were able to
match the numerical solution results to the experimental measurements.
In the previous RANS simulations of the present propulsor flow, not only the flow
field was not well resolved, also the effects of bubble dynamics on cavitation inception
were not included. In our previous studies10,11,12 we have shown that inclusion of bubble
dynamics can significantly affect the predictions of cavitation inception for both steadystate tip vortex flows and unsteady vortex/vortex interaction flow fields.
In the present study we aim at improving the numerical predictions of cavitation
inception for a ducted propulsor in two ways: refinement of the flow field solution and
inclusion of bubble dynamics. We consider a reduced computational domain which
includes the region of interaction of the two vortices but excludes for simplicity the
propulsor solid surfaces, and a direct Navier-Stokes simulation is conducted using initial
and boundary conditions provided by RANS computations of the full ducted propulsor
flow field. Then, a corrected spherical bubble dynamics model developed by Hsiao and
Chahine10,11 and a two-way flow-coupled non-spherical bubble dynamics model

To be published in Journal of Ship Research 2008

developed by Hsiao and Chahine12,13 are applied to study bubble dynamics and to predict
cavitation inception.

2. Numerical Approach
2.1 Flow Configuration

We consider the three-bladed NSWCCD Propeller 5206, a rotating ducted

propulsor with tip diameter of 0.8503m and a constant chord of 0.3812m from hub to tip.
This propeller operates in a duct having a diameter of 0.8636m. Detailed propulsor
geometry and experimental observations are described by Chesnakas and Jessup3.
Numerical studies5,6,7 using several RANS codes, described the flow field of this
propulsor when operating under an advance coefficient, J=0.98, with an inflow velocity,
U = 6.96 m/s . The corresponding Reynolds number based on the blade tip radius and
the inflow velocity was Re = 3 106 . Although the numerical studies all gave results in
reasonable agreement with the experimental measurements for the averaged local flow
field quantities, the predicted cavitation inception number, i , varied between 6.5 and
8.0, and the inception location, xi, was about 0.1 chord length downstream from the tip
trailing edge. However, the experimental observations gave much larger values for both
inception number and location: i11.5, and xi0.5.
To improve on the RANS numerical solution, we constructed a reduced
computational domain behind the trailing edge of the propulsor blade that included the
region of interaction of the two vortices and excluded propulsor solid surfaces. The basic
computational domain had a square cross area of 0.094m 0.094m and extended from

To be published in Journal of Ship Research 2008

the tip trailing edge to 0.34 m downstream from the tip trailing edge. Figure 1 illustrates
the location of this reduced computational domain relative to the ducted propulsor.

2.2 Navier-Stokes Computations

In the reduced domain, the flow is obtained via direct numerical solution of the
Navier-Stokes equations without turbulence modeling. Since the present computation is
conducted in a rotating frame attached to the propeller blade, source terms, i.e. the
centrifugal force and the Coriolis force terms, are added to the momentum equation. The
resulting unsteady incompressible continuity and Navier-Stokes equations written in nondimensional vector form and Cartesian notations are given as
u = 0 ,

Du
1 2
= p +
u + 2r 2 u ,
Dt
Re

(1)

(2)

where u = (u, v, w) is the velocity, p is the pressure, r is the radial position vector, is
the angular velocity, Re = u * L * / is the Reynolds number, u* and L* are the
characteristic velocity and length, is the liquid density, and is the liquid dynamic
viscosity.
To solve Equations (1) and (2) numerically, a three-dimensional incompressible
Navier-Stokes solver, DF_UNCLE, initially developed by the Mississippi State
University and modified by DYNAFLOW is applied. DF_UNCLE is based on the artificialcompressibility method14 in which a time derivative of the pressure is added to the
continuity equation as
1 p
+ u = 0 ,
t

(3)

To be published in Journal of Ship Research 2008

where is an artificial compressibility factor. As a consequence, a hyperbolic system of
equations is formed that can be solved using a time marching scheme. The method can
be marched in pseudo-time to reach a steady-state solution. To obtain a time-dependent
solution, a Newton iterative procedure is performed at each physical time step in order to
satisfy the continuity equation. In the present study the time-accurate solution is obtained
when the maximum normalized velocity divergence is less than 1.010-3. Detailed
descriptions of the numerical scheme can be found in Ref. 15.

2.3 Initial and Boundary Conditions

Since the unsteady Navier-Stokes computations were conducted in the reduced

domain, appropriate initial conditions are required. These conditions were obtained from
the steady state solution of the full ducted propulsor flow obtained by Yang7 using a
RANS computation. The original coarser RANS solution was interpolated for the present
finer reduced domain unsteady computations. Figure 2 shows the interpolated RANS
pressure contours at different streamwise locations to indicate the position of the main
vortex in the reduced domain. Figure 3 shows the pressure contours and the velocity
vectors at the inlet x-r plane boundary. The two co-rotating vortices (the tip-leakage
vortex and the trailing edge vortex) can be readily seen. The tip-leakage vortex is much
stronger than the trailing-edge vortex.
Figure 4 shows the reduced computational domain and its boundaries. In addition
to the inlet and outlet boundaries, there are three side boundaries in the liquid domain and
one boundary bounded by the duct wall. At the inlet boundary, the method of
characteristics was imposed with the three components of velocities specified based on

To be published in Journal of Ship Research 2008

the RANS solution. At the outlet boundary, all variables were extrapolated from the inner
domain except for the grid points on the duct wall where the initial RANS value of the
pressure was fixed. For the three side boundaries, both velocities and pressures were
imposed as given by the RANS solution.

2.4 Bubble Dynamics Models

Two bubble dynamics models, a corrected spherical model and a non-spherical

model, were used in this study. The spherical bubbles were tracked using a Lagrangian
scheme in a flow field combining the reduced domain and the RANS domain outside of
it. The grids of the reduced domain were overset to the overall propulsor grids. As a
bubble was released upstream of the propeller, it was subjected to the flow field given by
the RANS solution. Once the bubble entered the reduced domain, the flow field obtained
by the direct simulation of the Navier Stokes equations was applied. Bubble transport
was modeled via the motion equation described by Johnson and Hsieh16 while the bubble
dynamics was simulated by solving the Surface Averaged Pressure (SAP) spherical
bubble model equations developed by Hsiao and Chahine10,11.
The non-spherical bubble dynamics model was embedded in the unsteady NavierStokes solver, DF-UNCLE, with appropriate free surface boundary conditions and a
moving Chimera grid scheme12,13. Since unsteady Navier-Stokes computations are timeconsuming, this non-spherical model was combined with the spherical model mentioned
above. The spherical model was used to track the bubble during its capture by the vortex
and the non-spherical model was turned on only after the bubble size exceeds a preset
limit value. When the non-spherical model was turned on, the flow field due to the

To be published in Journal of Ship Research 2008

spherical bubble motion and volume change is superimposed on the liquid phase flow
field solution to provide an initial condition for the unsteady viscous computations. This
model allowed the bubble to deform non-spherically and a full two-way interaction
between the bubble and the flow field could be obtained. Detailed description of this
model and numerical implementations can be found in Refs 12, 13 and 17.

3. Results and Discussion

3.1 Grid and Domain Size Independence Study

The simulation of the vortex/vortex interaction in the reduced computational

domain was conducted with no turbulence modeling. The mean RANS flow solution was
specified at the inlet boundary and different grid resolutions and computational domain
sizes were considers. Unsteady turbulent fluctuations from upstream are being simulated
in on-going efforts and are the subject of a second publication18.
For the basic computational domain size described in Section 2.1, we generated a
4-block grid with a total of 101 grid points in the streamwise direction and three different
density cross flow grids: 61 61 , 121121 and 181 181 , in order to study the influence
of grid resolution. All grid points were evenly distributed without stretching. This
resulted in a uniform grid size of 3 mm in the streamwise direction and
0.5 mm in both cross directions for the finest grid. At least 34 grid points were within the
vortex core in each direction for the finest grid since the vortex core size had a diameter
at the trailing edge of about 17mm.
Figure 5 shows a comparison of the resulting pressure coefficients, Cp, along the
vortex center line for the three grid densities. We can also observe convergence of the

To be published in Journal of Ship Research 2008

simulations as the grid is refined. The maximum differences in Cpmin between the
121121 and the 181181 grids are within 3%. We can however observe a major
difference between the RANS solution and the present computations as one moves
downstream. While the RANS solution gives for the minimum pressure coefficient,
Cpmin, a value of about -8.5 at a distance of about 0.1 chord length, the present
computations yield a value of -11.2 at a location 0.35 chord length downstream from the
tip trailing edge. These values are much closer to the experimental results.
To further examine the grid dependence, we doubled the number of grid points in
the streamwise direction with the cross plane having 121121 grid points. As shown in
Figure 6 , doubling the grid resolution in the streamwise direction only slightly changed
the solution.
In addition to the grid resolution, we also studied the effect of the computational
domain size on the solution by constructing two larger computational domains while
maintaining the same grid resolution as the 121121101 case. One of the
computational domains was constructed by doubling the domain length in the azimuthal
direction and extending it 50% of length in radial direction. Note that the top boundary
in the radial direction was always bounded by the duct wall. The second computational
domain was constructed by extending 50% of the length in the streamwise direction. The
resulting pressure coefficients along the vortex center line obtained with these two larger
computational domains are compared to the basic domain and shown in Figure 7. It is
seen that all three cases produced practically the same solution (less than 1% difference).
Since increasing the grid and domain size from the basic computational domain
(121121101 grid points) only yielded small differences in the pressure coefficients,

To be published in Journal of Ship Research 2008

this basic domain size and grid resolution were used for subsequent bubble dynamics
computations in order to minimize CPU time.

3.2 Prediction of Cavitation Inception for Single Phase Flow

In Figure 5 the pressure coefficient along the vortex centerline obtained by the
RANS computation is compared to the present direct Navier Stokes unsteady
computations in the reduced domain. Major differences are seen between these two
results. The RANS computation predicts Cpmin=-8.2 at s/C=0.1 where s is the helical arc
distance from the tip trailing edge and C is the chord length, while the present
simulations show Cpmin=-11.2 at s/C=0.35 for the 181181101 grid, which is much
closer to the experimental observations. This indicates cavitation inception at 11.5 at
about 0.5 chord length downstream of the trailing edge. The failure of the RANS
computations is probably due to excessive vortex diffusion and dissipation.
Another qualitative comparison can be made between the two solutions by
displaying the same iso-pressure surfaces obtained in each case as shown in Figure 8.
This is similar to visualizing the cavitating vortices at different cavitation numbers. The
present results agree with the experimental observations which show a long-extended
fully cavitated vortex core at = 5.6 (see Figure 9 ).
It is important to examine the flow field near the location where the pressure
reaches its minimum value. Hsiao and Chahine19 have shown that the two co-rotating
vortices periodically approach each other and get closer during the vortex merger. The
present computations show that as they move closer, the flow in the axial direction is
accelerated and results in further decrease of the pressure at the vortex center. The

To be published in Journal of Ship Research 2008

computed Cp and axial velocity along the vortex center line are shown together in Figure
10. We can observe that the pressure correlates very well with the axial velocity and is
in phase opposition. As the velocity reaches a maximum value, the pressure in the vortex
center drops to a minimum. The axial velocity variations are, however, mostly ignored
by other investigations.
We would like to point out that, although the computations in the present study
eas unsteady, the simulation actually converged to a quasi steady-state solution since the
input solution to the inlet boundary was steady. The instability due to strong
vortex/vortex interaction as shown in Ref. 19 was not observed in the present problem.
This is probably due to the relatively weak trailing-edge vortex in the problem. Indeed,
Hsiao and Chahine19 have shown that as the relative strength of the main vortex is
increased, the merger of the two co-rotating vortices occurs faster and the unsteady
interaction becomes weaker.

3.3 Area of Bubble Capture: Window of Opportunity

The window of opportunity through which a nucleus needs to enter to be

captured by the vortex and generate strong acoustic signals has been studied for a finitespan tip vortex flow by Hsiao and Chahine11. It is also important to determine this
window of opportunity for the current flow field because with the knowledge of the
location and size of this small window, we are able to distribute and follow nuclei more
efficiently. In addition, this concept provides physical understanding of the flow field
and of the nuclei that are captured by the vortex. Near inception, the size of the window
of opportunity is directly related to the probability of cavitation event occurrence.

To be published in Journal of Ship Research 2008

To establish the window of opportunity a rectangular release area was
specified upstream of the tip leading edge of the propulsor in the x-r plane where bubble
nuclei were released from a 1511 grid points array. Figure 11 illustrates the location of
the release area relative to the propulsor blade. A high arbitrary value for the cavitation
number was selected (=12.0) such that each nucleus was tracked and the minimum
pressure it encountered during its travel was recorded and assigned to the release point.
The corresponding bubble size increased during its path in the propeller flow field by
less than 10 %. This enabled us to plot a contour of the minimum encountered pressure
coefficient for the release grid array and to obtain a contour plot of the cavitation
probability in the window of opportunity for each case. This selection was to obtain a
bound of the area where bubbles passing through this area would be affected by the
vortex field. The details on the encountered pressures vary with the selection of but
the area geometric bounds do not. The capture area concept enabled minimization of the
number of bubbles that needed to be tracked in the propeller flow domain.
Figure 12 shows contours of the minimum encountered pressure coefficients for
three different nuclei sizes, R0 = 5, 10, 20 m. The contours are blanked out for the
release points where the nuclei collided with the propeller surface. It is seen that the size
of the window of opportunity becomes smaller and its location shifts closer to the
propeller pressure side surface as the nucleus size decreases.

This brings out an

important selection concept on the probability of cavitation for waters containing very
small nuclei. Not only much lower pressures are needed to get such waters to cavitate
(smaller bubbles require much lower pressures to grow explosively11) but also fewer of

To be published in Journal of Ship Research 2008

them can participate into the cavitation event because of the reduced window of
opportunity.

3.4 Single Bubble Dynamics for Prediction of Cavitation Inception

Experimental observations conducted by Chesnakas and Jessup3 using a high

speed video camera and a hydrophone were able to capture the bubble and its emitted
acoustic signal during a cavitation event. According to the duration of the acoustic signal,
the cavitation events were categorized into popping and chirping events. They stated
that the popping event had a very short duration of noticeable acoustic signal less than
0.3 ms and that the bubble virtually remained spherical when its size was less than 0.1
mm in diameter. The chirping event had a much longer duration ranging from 0.3 to 10
ms, and the bubble had an elongated shape. They found that all the cavitation inception
events occurred near or behind a location 0.5 chord length downstream of the tip trailing
edge.
To simulate different cavitation events, we investigated the bubble behavior and
the emitted acoustic signal during bubble dynamics for different initial nuclei sizes at
different cavitation numbers. We found that popping cavitation events can be observed
for small bubbles. Figure 13 shows the bubble size variation and the emitted acoustic
signal for an initial nucleus size, R0=6 m at =10.47. It is seen that the maximum
bubble size achieved is about 0.1mm in diameter and that the noticeable acoustic signal
only lasts about 0.3ms. Figure 14 shows an example for a chirping event in which the
bubble grows to a much larger size (~1.7 mm) and the duration of the acoustic signal is
much longer (~2.5 ms) for an initial nucleus size, R0=20m at =10.75

To be published in Journal of Ship Research 2008

Figure 13 and 14 also show the pressure encountered by the bubble during its
journey. There is a small time delay for the bubble to grow to its maximum size after
encountering the minimum pressure. This delay significantly increases when a cavitation
event is produced by a small size nucleus. An example of such a cavitation event is
shown in Figure 15 for R0=5 m at =10.3.
To illustrate where the cavitation events occur in the flow field, the bubble
trajectory and size variations are plotted with the propulsor blade and some iso-pressure
surfaces as shown in Figure 16. It is seen that for larger initial nuclei radius, R0, the
cavitation events occur at a location slightly earlier than the experimental observations
while for smaller R0 the bubbles grow to a maximum size near the location 0.5 chord
length downstream of the tip trailing edge.
Bubble dynamics was also studied using our non-spherical models12. Figure 17
compares the bubble shapes obtained with the spherical and the non-spherical model for
R0=20m at =10.75. It is seen that both models predicted almost the same maximum
growth size. In non-spherical model the bubble elongated in the axial direction and taook
a quasi-cylindrical shape as it grew. However, for R0=20 m at =10.85 the bubble
remained almost spherical at its maximum size as shown in Figure 18. For both cases the
bubble started to collapse after reaching its maximum size. The non-spherical
computations, however, failed here to continue once strong deformations developed over
the bubble surface during the collapse.

3.5 Multiple Bubble Dynamics for Prediction of Cavitation Inception

To be published in Journal of Ship Research 2008

In order to simulate a realistic nuclei flow field as exists in nature or in the waters
of a cavitation tunnel, Hsiao and Chahine11 used a statistical nuclei distribution model
and showed that the nuclei size distribution has a strong influence on the prediction of
cavitation inception. Nuclei size distribution studies in water tunnels, lakes and oceans20
show a power-law distribution for the number density distribution function,

n ( R ) 1 / R , where R is the nucleus radius and the exponent lies between 3.5 and
4.0. Since the nuclei size distribution is not available for the experiments conducted on
the ducted propulsor investigated here, we have selected two very different nuclei size
distributions, with =3.0 and 4.5, which encompass the nuclei size distributions reported
for natural waters. We then compared their effect on the prediction of cavitation
inception. The first nuclei size distribution contained relatively larger nuclei sizes ranging
from 2.5 to 25 m while the other one contained smaller nuclei sizes ranging from 2.5 to
10 m. In both cases we randomly released the nuclei from a 0.02m0.03m window in
the computational domain. A total of 600 nuclei were released within a 0.4 second
period. The nuclei size number density distribution and the total number of nuclei
released for both cases are shown in Figure 19.
As the nuclei traveled through the computational domain, the emitted acoustic
signals were monitored. The acoustic pressure was monitored on the duct wall at a
location 0.5 chord length downstream of the tip trailing edge. A series of computations
were conducted at different cavitation numbers for both nuclei distributions to obtain
acoustic signals for conditions above and below cavitation inception.

Figure 20

illustrates the acoustic signals for the larger size nuclei distribution at three different
cavitation numbers.

To be published in Journal of Ship Research 2008

From the results shown in Figure 20, we can define a cavitation inception number
based on the number of cavitation events per unit time exceeding a certain value. Here a
cavitation event is defined arbitrarily as occurring when an oscillating bubble emits an
acoustic signal higher than 100 Pa. The curve of the number of cavitation events per
second versus cavitation number is shown in Figure 21. It can be seen that there is a
critical cavitation number above which no cavitation events occur. For nuclei size
distribution No. 1 (larger bubbles) an abrupt rise in the number of cavitation events is
seen when the cavitation number is below the critical cavitation number. Based on these
curves one can determine the cavitation inception number for both cases by defining a
criterion. For example, if one defines that 10 events per second are needed for cavitation
inception, then we obtain a cavitation inception number i=10.9 for the larger nuclei size
distribution and i=10.6 for the smaller nuclei size distribution can be deduced from
Figure 21. Actually, the selection of the amplitude of the threshold peak and number of
peaks per unit time should be a function of the application and of the detection capability
of the observer. Therefore, we do not intend to propose a criterion here but to illustrate
that the use of a bubble dynamics model can provide the researchers with a tool to predict
the cavitation inception number similar to what is actually done in experimental
observations. Chesnakas and Jessup (2003) in their experimental study defined the
cavitation inception criterion as one event per second and obtained a cavitation inception
number about 11.5. This inception number is very close to the critical cavitation number
(10.9) for the larger nuclei size distribution, but these results depend on the two criteria
selected: amplitude of the peak and number of peaks per unit time.

To be published in Journal of Ship Research 2008

4. Conclusions
In order to address grid resolution issues in RANS computations numerical
simulations of the two interacting vortices in a ducted propulsor flow field were
conducted in a reduced computational domain. It was found that numerical vortex
diffusion and dissipation were significantly reduced with grid refinement. The resulting
solutions illustrated with iso-pressure surfaces agree much better than the RANS
computations with experimental observations for fully developed cavitation in the vortex
core.

The value of the cavitation inception number and the location of cavitation

inception also agree significantly better with the experimental observations. No

instability was seen in the vortex/vortex interactions due to the large strength of the tipleakage vortex relative to the trailing-edge vortex. This leads to rapid merging of the
two vortices. Further grid refinement and perturbed input conditions may be required to
resolve any such instability.
The location and size of the window of opportunity through which a nucleus
needs to enter to be captured by the vortex was identified for different nuclei sizes.
Bubble dynamics studies showed that the characteristics of the acoustic signals
and bubble shapes as well as the location of cavitation inception resemble those observed
experimentally. A bubble dynamics model tracking multiple bubbles was also applied to
study the effect of nuclei size distribution and to predict cavitation inception in real flow
field conditions. Different nuclei size distributions and various definitions of the
cavitation inception criterion were found to influence the cavitation inception number.
However, the range of cavitation inception numbers (i10.8-11.0) was found to agree

To be published in Journal of Ship Research 2008

much better than previous studies (i 6.5-8.2) with the experimental measurements
(i11.5).

ACKNOWLEDGMNETS
This work was conducted at DYNAFLOW, INC. (www.dynaflow-inc.com) and was
supported by the Office of Naval Research under contract No. N00014-04-C-0110
monitored by Dr. Ki-Han Kim. Dr. Kims support is gratefully appreciated. The RANS
solution of the full propulsor flow was provided by Dr. C. I. Yang from NSWCD and the
experimental data was provided by Dr. Christopher J. Chesnakas and Dr. Stuart D. Jessup
from NSWCD. Their cooperation and help are greatly appreciated.

REFERENCES
1

A. L. Chen, J. D. Jacob, and O. Savas, Dynamics of co-rotating vortex pairs in the

wakes of flapped airfoils, J. Fluid Mech. 382, 155 (1999).

W. J. Devenport, C. M. Vogel, and Zsoldos, Flow structure produced by the

interaction and merger of a pair of co-rotating wing-tip vortices, J. Fluid Mech. 394,
357 (1999).

C. J. Chesnakas and S. D. Jessup, Tip-vortex induced cavitation on a ducted

propulsor, Proceedings of the ASME Symposium on Cavitation Inception,
FEDSM2003-45320, Honolulu, Hawaii, July 6-10, 2003.

To be published in Journal of Ship Research 2008

4

G. Oweis, S. Ceccio, C. J. Cheskanas, D. Fry, and S. D. Jessup, Tip leakage vortex

(TLV) variability from a ducted propeller under steady operation and its implications
on cavitation inception, 5th International Symposium on Cavitation, Osaka, Japan,
November 1-4, 2003.

W. H. Brewer, D. L. Marcum, S. D. Jessup, C. J. Chesnakas, D.G. Hyams, and K.

Sreenivas, An unstructured RANS study of tip-leakage vortex cavitation Inception,
Proceedings of the ASME Symposium on Cavitation Inception, FEDSM2003-45311,
Honolulu, Hawaii, July 6-10, 2003.

J. Kim, Sub-visual cavitation and acoustic modeling for ducted marine propulsor,
Ph.D. Thesis, The University of Iowa, 2002.

C. I. Yang, M. Jiang, C.J. Chesnakas, and S.D. Jessup, Numerical simulation of tip
vortices of ducted-rotor, NSWCCD-50-TR-2003/46, 2003.

J. Dacles-Mariani, G. G. Zilliac, J. S. Chow, and P. Bradshaw, Numerical/

experimental study of a wingtip vortex in the near field, AIAA J., 33, 1561 (1995).

C.-T. Hsiao and L. L. Pauley, Numerical study of the steady-state tip vortex flow
over a finite-span hydrofoil, ASME J. Fluids Eng. 120, 345 (1998).

10

C.-T. Hsiao, G. L. Chahine, H. L. Liu, Scaling effects on prediction of cavitation

inception in a line vortex flow, ASME J. Fluids Eng. 125, 53 (2003).

11

C.-T. Hsiao, G.L. Chahine, Scaling of tip vortex cavitation inception noise with a
statistic bubble dynamics model accounting for nuclei size distribution, ASME J.
Fluids Eng. 127, 55 (2005).

To be published in Journal of Ship Research 2008

12

C.-T. Hsiao, and G.L. Chahine, Prediction of vortex cavitation inception using
coupled spherical and non-spherical models and Navier-Stokes computations, J.
Marine Science and Technology, 8, 99 (2004).

13

C.-T. Hsiao, and G.L. Chahine, Numerical simulation of bubble dynamics in a

vortex flow using moving Chimera grid and Navier-Stokes computations, Fourth
International Symposium on Cavitation CAV2001, Pasadena, CA, June 20-23, 2001.

14

A. J. Chorin, A numerical method for solving incompressible viscous flow

problems, J. Computational Physics, 2, 12 , 1967.

15

K. Vanden, D. L. Whitfield, Direct and iterative algorithms for the threedimensional Euler equations, AIAA-93-3378, 1993.

16

V.E. Johnson, T. Hsieh, The influence of the trajectories of gas nuclei on cavitation
inception, Sixth Symposium on Naval Hydrodynamics, pp. 163, 1966.

17

G.L. Chahine, Nuclei Effects on Cavitation Inception and Noise, 25th Symposium
on Naval Hydrodynamics, St. John, Canada, 2004.

18

C.-T. Hsiao, and Chahine, G.L., Effect of Unsteady Turbulent Fluctuations on

Vortex/Vortex/Nuclei Interaction, 26th Symposium on Naval Hydrodynamics,
Rome, Italy, 2006.

19

C.-T. Hsiao, and G.L. Chahine, Effect of vortex/vortex interaction on bubble

dynamics and cavitation noise, Fifth International Symposium on Cavitation
CAV2003, Osaka, Japan, November 1-4, 2003.

20

R.E. Franklin, A note on the radius distribution function for microbubbles of gas in
water, ASME Cavitation and Mutliphase Flow Forum, FED-Vol. 135, pp.77, 1992.

To be published in Journal of Ship Research 2008

To be published in Journal of Ship Research 2008

Figure 1. A view of the reduced computational domain in the three-bladed NSWCCD

Propeller 5206 used for the current computations.

Figure 2. Pressure field in the reduced computational domain interpolated from the RANS
solution.

To be published in Journal of Ship Research 2008

Figure 3. Pressure contours and velocity vectors imposed at the inlet boundary of the
reduced domain in the x-r plane.

Duct Boundary

Outlet
Boundary
Inlet
Boundary

Side
Boundaries

Figure 4. Geometry of the reduced computational domain bounded by six boundaries with
different imposed boundary conditions.

To be published in Journal of Ship Research 2008

Pressure coefficient along vortex center

-4

DNS 61x61 grid

DNS121x121 grid
DNS 181x181 grid
RANS

Cp

-6

-8

-10

0.1

0.2

0.3

s/C

0.4

0.5

0.6

0.7

Figure 5. Comparison of the variation of the pressure coefficient along the vortex

centerline between the RANS solution and the present Navier-Stokes numerical
simulations with three different grid resolutions.

Pressure Coefficient along Vortex Center

-4
Basic Domain, 121x121x101
Basic Domain 121x121x181

Cp

-6

-8

-10

0.1

0.2

0.3

s/C

0.4

0.5

0.6

Figure 6. Comparison of the pressure coefficient variation along the vortex centerline for
two streamwise grid resolutions.

To be published in Journal of Ship Research 2008

Pressure Coefficient along Vortex Center

-4
Basic Domain, 121x121x101
50% increase in streamwise direction, 121x121x151
100% and 50% increase in azimuthal and radial directions respectively, 241x181x101

Cp

-6

-8

-10

0.1

0.2

0.3

0.4

0.5

0.6

s/C

0.7

0.8

0.9

1.1

1.2

Figure 7. Comparison of the pressure coefficient variation along the vortex centerline for
three computational domain sizes.

Figure 8. Iso-pressure surfaces at various cavitation numbers as obtained by the RANS

solution and the present Navier-Stokes numerical solution with the 121121x101 grid. This
is equivalent to showing cavitation extent at these cavitation numbers.

To be published in Journal of Ship Research 2008

Figure 9. Fully developed cavitation in the vortex core at = 5.6 (Chesnakas and Jessup
2003).

3.5

-8

Cp
Axial velocity

-9

3.3

Vs

Cp

3.4

3.2

-10

3.1

0.1

0.2

0.3

0.4

s/C

0.5

0.6

0.7

Figure 10. Pressure coefficient and axial velocity as a function of the distance from the tip
trailing edge.

To be published in Journal of Ship Research 2008

Figure 11. The location of the release area for establishing the window of opportunity.

To be published in Journal of Ship Research 2008

Figure 12. Contours of the bubble encountered Cpmin for released nuclei with R0=5, 10, and
20 microns.

To be published in Journal of Ship Research 2008

Figure 13. Bubble radius, emitted acoustic pressure signal, and encountered pressure
during a cavitation event for R0= 6m at =10.47.

Figure 14. Bubble radius, emitted acoustic pressure and encountered pressure during a
cavitation event for R0= 20 m at =10.75

To be published in Journal of Ship Research 2008

Figure 15. Bubble radius, emitted acoustic pressure and encountered pressure during a
cavitation event for R0= 5 m at =10.3.

R0=20m
=10.85

s/C=0
Cp=-10.9

R0=5m
=10.3

R0=20m
=10.75

Cp=-5.6

Cp=-10.9

s/C=0

s/C=0

Cp=-5.6

Cp=-5.6

Cp=-10.9
s/C=0.5

blow up

s/C=0.5

s/C=0.5

blow up

Figure 16. Bubble trajectories and size variations during a cavitation event for three
different initial radii and cavitation numbers. Also shown are some iso-pressure contours.

To be published in Journal of Ship Research 2008

Spherical Model

Non-Spherical Model

Figure 17. Computed bubble sizes and shapes using both spherical and non-spherical
models shown with two levels of iso-pressures for R0= 20 m at =10.75.

Figure 18. Computed bubble sizes and shapes of non-spherical modes for R0= 20m at
=10.85.

To be published in Journal of Ship Research 2008

1.00E+10

1.00E+09
Nuclei Distribution 1
Nuclei Distribution 2
d N /d R (N /m )

1.00E+08

1.00E+07

1.00E+06

1.00E+05

1.00E+04
1.0E-06

1.0E-05

1.0E-04

Nuclei size (m)

600
500

Nuclei Distribution 1
Nuclei Distribution 2

400

300
200
100
0
2.5E-06

5.0E-06

7.5E-06

1.0E-05

1.5E-05

2.0E-05

2.5E-05

Nuclei Size (m)

Figure 19. Nuclei size distribution and number of nuclei released versus nuclei size for two
different nuclei size distributions considered in this study.

To be published in Journal of Ship Research 2008

Cavitation Number =10.90
Void fraction =9.0e-10

Acoustic Press (pa)

600

400

200

0.1

0.2

Time (Sec)

0.3

0.4

0.3

0.4

0.3

0.4

Cavitation Number =10.85

Void fraction =9.0e-10

Acoustic Press (pa)

600

400

200

0.1

0.2

Time (Sec)

Cavitation Number =10.80

Void fraction =9.0e-10

Acoustic Press (pa)

600

400

200

0.1

0.2

Time (Sec)

Figure 20. Acoustic signals for the large size nuclei distribution case at three different
cavitation numbers.

Number of events per second

80
70

Nuiclei Distribution 1
Nuclei Distribution 2

60
50
40
30
20
10
0
10.2

10.3

10.4

10.5

10.6

10.7

10.8

10.9

11

Cavitation Number

Figure 21. Number of events per second versus cavitation number for the nuclei size

distributions shown in Figure 19.