SUPERCAVITATION RESEARCH AND DEVELOPMENT

SUPERCAVITATION RESEARCH AND DEVELOPMENT
Ivan N. Kirschner, Neal E. Fine, James S. Uhlman,
Thomas A. Gieseke, Robert Kuklinski,
David C. Kring, and Benjamin J. Rosenthal
and Abraham N. Varghese
Anteon Corporation – Engineering Technology Center
Naval Undersea Warfare Center Division
One Corporate Place
1176 Howell Street
Middletown, RI 02842-6277 USA
Newport, RI 02841-1708 USA
David R. Stinebring, John E. Dzielski, Jules W. Lindau, and Robert F. Kunz
Applied Research Laboratory / Penn State University
PO Box 30
State College, PA 16804 USA
ABSTRACT
Supercavitating bodies can achieve very high speeds under water by virtue of reduced drag: with proper design, a cavitation bubble is generated
at the nose and skin friction is drastically reduced. Depending on the type of device under consideration, the drag coefficient can be an order of
magnitude less than that of a fully wetted vehicle.
For several years, the United States Navy has supported basic research and development involving supercavitating high-speed bodies. Under
this program, the theory of high-Mach-number underwater flows has been investigated and first-principles modeling of cavity development and
body dynamics have been addressed. To complement these analytical efforts, a sophisticated experimental program has matured.
This presentation provides an overview of recent research in this topic, including the results of selected experiments with supercavitating
projectiles at supersonic speeds in water, experiments involving systems suitable for self-propelled vehicles, and several topics in computational
methods and simulation.
INTRODUCTION
Supercavitation is a hydrodynamic process in which an undersea body is almost entirely enveloped in a layer of gas
initiated at a cavitator mounted at the forward end. A supercavity can be maintained in one of two ways: (1) by achieving
such a high speed that the water vaporizes near the nose of the body; or, (2) by supplying gas to the cavity at nearly
ambient pressure. The first technique is known as vaporous cavitation; the second is termed ventilation, or artificial
cavitation.
Supercavitating bodies can achieve very high speeds under water by virtue of reduced drag: with proper design, a
cavitation bubble is generated at the nose and skin friction drag is drastically reduced. Because the density and viscosity
of the gas are dramatically lower than that of seawater, skin friction drag can be reduced significantly. If the body is
shaped properly, the attendant pressure drag can be maintained at a very low value, so that the overall body drag is also
reduced dramatically: by roughly eighty percent for a self-propelled vehicle capable of executing maneuvers, to an order
of magnitude for free-flying gun-launched projectiles. However, because the center of pressure is typically located well
forward with respect to the center of gravity, control and maneuvering present special challenges. Also, whereas a fullywetted vehicle develops substantial lift in a turn due to vortex shedding off the hull, a supercavitating vehicle does not
develop significant lift over the gas-enveloped surfaces. This requires a different approach to effecting hydrodynamic
control.
For most of the last decade, the United States Navy has sponsored basic research and exploratory development programs
addressing the physics and engineering of supercavitating high-speed bodies. This paper highlights some results of the
Navy-sponsored activities at the Naval Undersea Warfare Center (NUWC) Newport Division, at the Applied Research
Laboratory of The Pennsylvania State University (ARL/PSU), and at the Engineering Technology Center Division of
Anteon Corporation (Anteon/ETC). Selected results of experiments and of physics modeling and simulation are briefly
presented.
EXPERIMENTS WITH SUPERCAVIT ATING HIGH-SPEED BODIES
Over the course of the Navy program, many tests with captive models have been performed – both in various water
tunnels and in the Langley Tow Tank located in Hampton, Virginia – building on the extensive cavitation research
performed in many countries throughout the twentieth century (see, for example, the excellent review in May, 1975).
Beginning circa 1993, the Navy began testing free-flying supercavitating projectiles launched directly into water, again
building on the results of previous research using launchers located above the water surface. One of the immediate future
objectives of the Navy program is the demonstration of safe, self -propelled flight of a supercavitating high-speed body
capable of prescribed maneuvers. As with the previous aspects of the program, this activity will build on previous tests,
in this case, with small, uncontrolled, underwater rockets.
Captive Model Tests
Photographs of cavities produced by disk and cone cavitators are shown in figures 1a and b. The models used in these
experiments at ARL/PSU can incorporate systems for introducing ventilation gas, cavitator force balances, pressure
measurement systems, and other instrumentation. Just downstream of the cavitator in figure 1b is a gas deflector that
redirects the gas flow to minimize disturbances to the cavity. Care must be taken to ensure that the design for the
introduction of ventilation gas does not have a significant effect upon the force measurements. Short exposure times,
such as with stroboscopic lighting, should be used when investigating unsteady cavity dynamics. Strobe lights can be
synchronized with the scanning rate of video systems to provide exposure times of a few microseconds for each video
1
frame. The video record of the cavit y can then be compared with transient measurements, such as forces, by use of a
master clock. The photograph in figure 1a shows a stable cavity, while cavity instabilities are apparent in figure 1b.
(a)
(b)
Figure 1. Photographs of Water Tunnel Cavitation Experiments: (a) Disk Cavitator at an Angle of Attack; (b) Conical Cavitator
with Stroboscopic Lighting
Cavity-piercing supercavitating hydrofoil-type control surfaces such as those shown in figure 2a have been tested in the
ARL/PSU 12-in diameter water tunnel. The hydrofoil is attached to a force balance and test fixture that mounts through
the water tunnel test section wall and is designed so that there is a small gap between the base of the hydrofoil and the test
section wall. The hydrofoil angle of attack can be adjusted by rotating the test fixture. A two-dimensional cavity is
created upstream of the hydrofoil by ventilating downstream of a wedge mounted to the tunnel wall. The hydrofoil then
pierces the tunnel wall cavity as shown. The meas ured forces in the water tunnel have been used to validate
computational codes, such as the results of boundary-element modeling using an Anteon/ETC program known as LScav
(Kirschner, et al, 2001, 2000; Fine and Kinnas, 1993; Fine, 1992) shown in figure 2b.
0.3
Lift Coefficient
0.25
Fin #4, Lift Coefficient
LScav Data
ARL Data
0.2
0.15
0.1
Supercavitating
0.05
(a)
Partially cavitating
0
0
3
6
9
12
Angle of attack (degrees)
15
(b)
Figure 2. Cavity-Piercing Hydrofoils: (a) Photograph from Water Tunnel Test; (b) Results of Boundary-Element Modeling
Another type of captive model test has been conducted by NUWC at the Langley Tow Tank located in
Hampton, Virginia, to investigate the physics of ventilated cavities on large-scale models. The tests were conducted on a
freely rotating model that was equipped with an actuated cavitator. The ventilated cavity boundary is a free surface,
which may be susceptible to instabilities that could grow and affect the areas of wetted contact on a supercavitating
vehicle. This undesirable change in wetted contact could produce large hydrodynamic forces leading to catastrophic loss
of control. The stability of the cavity may be affected by changes in the ventilation rate, changes in depth, or unintended
coincidence of system natural frequencies with the cavity oscillation frequency. The tow tank tests were performed to
measure the amplitude and frequency of the cavity oscillations. Figure 3 shows a photograph of the test model, along
with selected cavity dynamics data. High-frequency, solid-state pressure transducers were used to determine the behavior
of the cavities. The tests confirmed that the dominant cavity frequency was correlated with cavity length and towing
speed.
2
The data shows the frequency
and amplitude of the pressure
fluctuations in the ventilated
cavity as a function of time. As
the ventilation gas supply was
increased at 43 s, the cavity
grew and the cavitation number
decreased. A longer cavity at a
constant
forward
speed
corresponds
to
a
lower
fundamental
frequency
of
oscillation.
This run clearly
shows two different frequencies
of oscillation at the initial gas
flow rate and then a single
dominant frequency at the
higher gas flow rate, for which
the cavity was longer. Similar
data was obtained for over 100
test conditions. Such tests will
mitigate risk in the design of
supercavitating vehicles by
allowing
prediction
of
ventilation gas requirements
and avoidance of cavity states
that are likely to lead to failure.
3.0
2.0
Pressure
Reduced noise –
ventilated
cavitation
High-frequency
pressure signal
1.0
0.0
-1.0
High-frequency noise – wetted flow
-94
35.0
Change in dominant
frequency with
cavitation number
and cavity length
20.0
-109
-129
10.0
Sliding
transform
0.0
-154
Velocity or
Flow rate
60.0
2.5
40.0
20.0
2.0
Flow
conditions
0.0
1.5
Carriage speed
-20.0
0
10
20
1.0
Vent gas flow rate
Cavity pressure
30
60
40
50
70
Pressure
Frequency
30.0
Level (dB relative)
-2.0
0.5
80 85
Time (s)
Figure 3. Tow Tank Testing of Captive Models: Sting-Mounted Model (Inset) and Cavity
Dynamics Data
Free-Running Model Tests
Captive models are valuable for testing components and subsystems. However, many questions related to vehicle control
and cavity-body interactions are best answered using free-running models with on-board inertial measurement units and
other instrumentation. An example of an ARL/PSU model is shown in figure 4a. Another class of free-running models,
known as Adaptable High-Speed Undersea Munitions (AHSUM), is represented in figure 4b. This device was designed
by NUWC and personnel now at Anteon/ETC for high speed launch from a fully-submerged gun. The high-speed film
frames of figure 4c show AHSUM in supersonic flight under water. The shock wave and cavity are clearly visible.
Frame 1
(a)
Shock
Frame 2
+100 µs
Projectile
base
(b)
Figure 4. Examples of Free-Running Supercavitating Models: (a) SelfPropelled Vehicle; (b) Gun-Launched Supercavitating Munitions Package;
(c) Supercavitating Projectile in Supersonic Flight Underwater
Shock
Frame 3
+200 µs
Shock
3
(c)
The
supersonic
tests
were
performed on an indoor range at
NUWC that was customized for
such scientific experimentation. In
order to assess the range
performance of such devices, the
gun was installed by NUWC for
field testing at the Army Research
Laboratory’s SuperPond located in
Aberdeen, Maryland. A schematic
of the test set-up, along with an
aerial view of the SuperPond and a
view looking down range from the
launch area are shown in figure 5.
Measured velocity data collected at
several stations along the range are
also plotted, along with a theoretical
prediction based on the Leduc
ballistic equation and the best
estimate of the projectile drag
coefficient.
Target
Barge-mounted
gun launcher
Velocity / Launch Velocity
1.0
0.8
0.6
0.4
0.2
0.0
0.0
Shot 510
Shot 604
shot 605
Leduc
0.2
0.4
0.6
0.8
1.0
Range / Maximum Observed Range
Figure 5. Supercavitating Projectile Open Range Tests
MODELING AND SIMULATION OF SUPERCAVITATING HIGH-SPEED BODIES
An important element of the Navy program involves the development of theoretical and computational models that
facilitate the investigation and understanding of the physics of supercavitating high-speed bodies. Such tools are also
being incorporated into resources for simulating the behavior of such systems, allowing assessment of their utility. Such
simulations are expected to play an important role in planning experiments with free-running, self-propelled vehicles, in
assessing the safety of such exercises, and in analyzing and applying test data.
Boundary-Element Modeling
Boundary-element modeling has proven successful in cost-effectively predicting many important features of
supercavitating flows, including the cavity shape, the cavity drag, and even such details as the re-entrant jet at the cavity
closure point (Uhlman, et al, 1998; Savchenko, et al, 1997; Kirschner, et al, 1995). An example of such predictions was
discussed above in connection with the lift on supercavitating hydrofoils. Figure 6 shows an application by NUWC of
boundary-element methods to predict flows past partially-cavitating axisymmetric bodies of various geometry (Varghese,
et al, 2001). The generic geometry of the flow system is depicted in the upper left-hand corner of the figure. The
predicted cavity shapes for fixed cavity lengths are shown in the lower left-hand graph. The relationship between the
cavitation number and cavity length is presented in the lower right-hand graph for two selected body radii, and for the
case of supercavitation. It can be seen that the presence of the body plays an important role. The upper right-hand graph
shows a prediction of the drag budget for partially-cavitating flows at high Reynolds number. It can be seen that,
although the total drag of partially-cavitating bodies of the selected form is largely dominated by the friction and base
drag components, a significant decrease in drag occurs in the regime where the cavity length approximately equals the
forebody length. However, it should also be noted that non-feasible solutions are predicted for which the cavity intersects
the body profile at the forebody-cylinder junction. A similar effect leading to hysteresis has been observed in
experiments (Savchenko, et al, 2000). Such methods are being extended by Anteon/ETC to predict fully-threedimensional, time-dependent flows with lift (Kring, et al, 2001; Kring, et al, 2000). Note that the forebody geometry is
not the same for the various cases presented in figure 6.
Large-Scale Computational Fluid Dynamics (CFD) Modeling
UNCLE-M is an implicit, time-accurate, pre-conditioned, multi-phase, Reynolds-averaged Navier-Stokes (RANS) solver,
with high-order-accurate discretization, flux limiters, and a three-dimensional, fully generalized multi-block, parallel
structure. Mixture volume and constituent volume fraction transport and generation for liquid, condensable vapor, and
non-condensable gas fields are solved. Mixture momentum and two-equation turbulence model equations are also solved.
Finite rate mass transfer modeling is employed to account for exchange between the liquid and vapor phases. The code
can handle buoyancy effects and the presence and interaction of condensable and non-condensable fields. This level of
modeling complexity represents the state of the art of CFD cavitation analysis. The code is bein g adapted and applied at
ARL/PSU to model the physics associated with high-speed maneuvers, controls, control surface actuation, body-cavity
and fin-cavity interactions, viscous effects such as flow separation, and compressibility effects associated with rocket
exhaust, ventilation, bubbly mixtures, and very high speed projectile motion.
4
Although axisymmetric boundary-element models facilitate an understanding of key aspects of the flow physics and are
useful in the design and analysis of supercavitating high-speed bodies, vaporous cavitation over an axisymmetric body is
known to be a highly nonlinear and three-dimensional event. This is clearly illustrated in figure 7a, which is a photograph
obtained during water tunnel testing of a blunt cavitator at zero angle-of-attack. The cavitation number was σ≅0.35 and
the Reynolds number was Red≅150,000. The cavity behavior compares favorably with the CFD result shown in figure 7b.
The modeled conditions were nearly the same as those in the experiment. Approximately 1.2 million computational
nodes were used. The cavity boundary in the CFD result is characterized by a void fraction isosurface at α l=0.5. Selected
streamlines are also shown, and the cylinder surface has been colored by volume fraction. Clearly in neither the model
result nor the photograph is the flow field in and around the cavity axisymmetric. There is little likelihood of obtaining
purely axisymmetric conditions in even the most well-controlled environments. This is compounded by the influence of
highly nonlinear turbulent flow dominated by phase transition, et cetera. It is suggested that here, via an enhanced level
of modeling, the real flow has been well captured. An in-depth understanding of the detailed causal mechanisms is a goal
of ongoing research.
2.0
Drag Coefficient
Reynolds number: 3.0E7
Total
Pressure
Friction
Base
1.5
Forebody angle: 9.55º
Base diameter: 2.4
1.0
0.5
0.0
0
2
4
6
8
10
Dimensionless Cavity Length
1.2
Body Radius=2.0
1.2
1.0
Body
1.0
Cavitation Number
Dimensionless Cavity Radius
1.4
0.5000
1.0000
0.8
1.5000
0.6
Dimensionless
cavity
length
0.4
0.2
2.0000
Non-feasible
solutions
4.0000
5.0000
Supercavitation
0.8
0.6
Forebody angle: 6.96º
0.4
0.2
6.0000
Forebody angle: 15.92º
Base diameter: 2.4
0.0
Body Radius=1.5
0.0
7.0000
0
0
2
4
6
1
2
3
4
5
8
Dimensionless Cavity Length
Dimensionless Axial Coordinate
Figure 6. Boundary-Element Modeling of Partially-Cavitating Axisymmetric Flows (Body Length is 80 Cavitator Diameters)
Figure 7. Three-Dimensional Vaporous Cavitation over a Blunt Circular Cylinder: (a) Photograph at σ ≈ 0.35 ; ( b) CFD
Prediction at σ ≈ 0.4
A sample of the results obtained by three-dimensional time-domain modeling of vaporous cavitation over a blunt cylinder
at zero angle of attack is presented in figure 8. These results appear to agree with both significant qualitative and
quantitative experimental observations. As in the experiment, the modeled reentrant flow has been observed to follow a
helical pattern. This helical flow revolves around the circumference of the cylinder, driven by a complex reentrant flow.
5
At the same time, other aspects of the flow tend to cause a cavity cycle that is largely axial. This axial cycle fits the
typical observations of reentrant flow (Stinebring, et al, 1979, 1983; May 1975). This axial motion is observable in the
snapshots and is also well captured by the drag coefficient history given in figure 8a of the figure. Here the drag history
has been given over a cycle as defined by the three-dimensional flow. Clearly the drag coefficient is insufficient, by
itself, to provide the true model cycle. However by examination of figure 8b in conjunction with figure 8a, it is possible
to deduce the model cycle. In figure 8c, the cycling frequency data based on a three-dimensional analysis is compared
with experimental results and those of an axisymmetric analysis. Due to the obser ved helical nature of the reentrant
region, it was necessary, experimentally, to use high speed movies to determine the period cavity cycling (Stinebring
1976, 1983, and personal communications). Generally two consecutive observed cycles were required to determine the
reported cycle. This would then coincide with the cycle determined by a complete revolution of the reentrant jet. It is
apparent that, particularly in comparison to the axisymmetric UNCLE-M results, the three-dimensional UNCLE-M results
com pare favorably with the experimental data.
(a)
(b)
Figure 8. Naturally Cavitating Flow over a Blunt Cylinder: (a) Drag History;
(b) Sequence showing Surface Pressure and Isosurface of Liquid Volume
Fraction; (c) Comparison of Measured and Computed Cavity Cycling Frequencies
(c)
Figure 9 shows an underwater supersonic projectile. Both computational results and a corresponding photograph of an
actual test are included in the figure. The flow Mach number for the case shown is 1.03 and the liquid-to-gas density ratio
is nominally 1000. The experiments and the computations show the presence of a bow shock upstream of the nose. Also,
because of the high velocity, the cavitation number is about 10 -4. Consequently, most of the flow immediately adjacent to
the body is completely vaporized as is the downstream wake portion. Field contours of mixture density and surface
contours of pressure are shown. The single phase water shock system, vaporous cavity, and vaporous wake are
observable in the simulation.
Figure 10 shows the flow field of an underwater rocket exhaust. The plume is supersonic and is slightly under -expanded.
It is surrounded by a co-flowing secondary subsonic gas stream, that in turn is surrounded by a liquid water free-stream
flow. The nominal liquid-to-gas density ratio is 1000. Figure 10a shows the density field. Figure 10b shows the shock
function field, which exhibits the classic expansion pattern. In particular, the interaction of the compressible gas stream
with the incompressible liquid is demonstrated first by the contraction and then by the expansion of the gas stream. In
addition, the interface between the liquid and gas phases is comprised of a two-phase mixture that is also fully supersonic
due to the low magnitude of the mixture sound speed.
Flight Simulation of Supercavitating Vehicles
Large-scale simulation has also been applied in the time domain to begin to address vehicle maneuvering. In figure 11, a
set of preliminary results for a notional self-propelled, ventilated, supercavitating vehicle is presented. Figure 11a
illustrates a view of the vehicle geometry, which has a relatively blunt cavitator and several annular ventilation ports with
aft oriented gas deflectors. A cavity gas ventilation rate was prescribed sufficient to enshroud the entire vehicle during
steady flight. A gas propellant flow rate was also specified at the exhaust nozzle. For this analysis, ARL/PSU applied the
incompressible UNCLE-M model.
6
1000kg/m3 Liquid Free Stream
Density
1kg/m3 Supersonic Jet
1kg/m3 Subsonic Gas
Cavity
Supersonic interface (not shock) due to
low mixture sound speed.
a
Shock Function V • ∇ρ m
Projectile
Projectile Base
Bow
Bow Shock
Shock
b
Figure 9. Supersonic Supercavitating
Projectiles: (a) High-Speed Film
Frame; (b) Computational Result
Figure 10. Supercavitating Vehicle Rocket Plume Modeled as a 3-Stream Wake
Figures 11b-h show cavity shapes corresponding to different instants during the prescribed pitch cycling maneuver shown
in figure 11i. A non-dimensional time step of ∆t/tref = 0.09473 was specified, where tref = Lvehicle /U∞. A grid consisting of
over 1.2 million vertices was used. The simulation was run on 48 processors of a Cray T3E. The snapshots of the
evolving cavity during the maneuver are characterized by isosurfaces of liquid volume fraction at a value α l=0.5.
Figure 11i shows the predicted lift history for the vehicle during the maneuver, in addition to the prescribed angle-ofattack. A three-field simulation was carried out. Vaporous cavitation occurs upstream of the first gas deflector. The
significant perturbation of the cavity geometry for this maneuver is clearly evident. Indeed, the cavity impinges on the
body at t/t ref = 47.4. Also, natural cavitation near the leading edge is not sufficient to keep the first injection port
unwetted.
Figure 11. Three-Dimensional Time-Domain Simulation of a Prescribed
Maneuver
7
Such large-scale simulation and the related
simulation
using
fully
three-dimensional
unsteady boundary element models allow
investigation of the detailed interaction between
the body and the cavity during the course of a
maneuver. In some cases, however, basic albeit
important aspects of the vehicle dynamics can be
more easily captured using simpler models.
ETC_Supercav and ETC_Simplecav are two
such tools being developed by Anteon/ETC.
ETC_Supercav is a six-degree-of-freedom
underwater flight simulator that incorporates
simple hydrodynamics models of the cavitator,
the fins, the afterbody, and the cavity itself into a
numerical simulation resource that allows the
investigation of various physical aspects of the
vehicle (such as fin sweep-back under constanttorque actuation), maneuvering strategies (such
as a banked turn), and the control system itself
(including such aspects as state estimation to
account for noisy hydrodynamic excitation, as
discussed above, and robustness to uncertainty).
Example output from the flight simulator is
presented in figure 12, which shows a
comparison of vehicle response with the two
different fin configurations indicated.
-3
Pitch (rad)
8
-3
x 10
8
7.5
7.5
7
7
6.5
6.5
6
6
5.5
5.5
5
0
150
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
0
0
Planing Force (N)
x 10
-200
-400
-400
-600
-600
-800
-800
-1000
-1000
-1200
-1200
Fin Sweep-Back Angle (deg)
-200
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
1
-1400
-1400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 2 15 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
2 5
2 0
2 0
1 5
1 5
1 0
1 0
5
5
0
0
0
0.2
0.4
0.6
Time (s)
0.8
1
0
0.2
0.4
0.6
0.8
1
An
even
simpler
tool,
ETC_Simplecav, has been found
to be very useful in mapping the
basic parametric space of such
vehicles. This tool is used to
simulate the model problem
presented in figure 13a.
The
simplified system consists of a
mass and a cavity-like surface
that is deformed via motion of
the mass. A system of nonlinear
springs is used to represent the
forces acting on the mass via the
fins and afterbody planing on the
cavity boundary.
The key
features of both the full and the
simplified systems are the
discontinuous forces associated
with the rather sharp boundary
between the gas and liquid
phases of the ambient fluid, and
the memory effects associated
with advection of disturbances
downstream from the cavitator.
This highly nonlinear behavior
leads to a system map with the
fractal character shown in
figure 13b, which presents the
Lyapunov exponent as a function
of the time delay (as set by the
cavitator position and the vehicle
speed) and of a characteristic
system frequency (as set by the
mass and spring constants).
Time (s)
Figure 12. Sample Results from ETC_Supercav
Because of the presence of a time delay and non-linear coupling, it appears that even this simplified system satisfies the
necessary conditions such that chaotic behavior might be encountered (Baker and Gollub 1991). This may be related to
the fractal nature of the plot shown in figure 13b.
6
Dimensionless
l/dTime Delay
5.5
5
4.5
S uperCav Test Bed
4
3.5
3
2.5
2
(a)
1.25
1.5
1.75
Figure 13. Simplified Simulation with ETC_Simplecav: (a) Simplified System; (b) Parametric System Map
8
2
w d/U
Dimensionless System Frequency
Parameter
(b)
Simulation capabilities are also being developed at ARL/PSU to support the development and testing of supercavitating
high-speed bodies. Analytical and computational models have been developed describing the full six-degree-of-freedom
dynamics of a supercavitating body. In addition to the dynamics of the body, forces on the cavitator, fins, and planing
contact point are modeled. These forces are modeled using a combination of empirical, analytical, and experimental data
in a highly modular simulation environment that has been specifically developed over the last 10 years to support the
integration and testing of prototype vehicles. A novel scheme for tracking the cavity has been implemented that allows
the full simulation model to run in real-time. This simulation capability is being used in the hydrodynamic design,
optimization, and analysis of supercavitating vehicles; in the development and bench shakedown of practical control
systems for in-water testing; and in the analysis of in-water data.
The vehicle simulation model has been coupled with the immersive visualization capabilities of the ARL/PSU SEALab.
This immersive environment is a 4-wall cave automated virtual environment (CAVE) that provides a user with a
stereoscopic view of a scene that may include just a vehicle or imagery of a specific geographical location. Figure 14
shows a snapshot of a visualization of the simulation described above. Various vehicle simulations run in a distributed
fashion on a network and are tied to the visualization environment via a distributed simulation/high-level architecture
(DIS/HLA)-compliant system.
Cavity
Bearing
Vehicle (aft view
showing planing)
Figure 14. Dynamics Visualization Tool in SEALab
The user can interact with a simulation as it
progresses to change a viewpoint, issue a
command to a player, change the initial
location of players and their characteristics,
and control player responsiveness to
environmental stimuli.
This capability is
being used to evaluate different vehicle
concepts in operational environments and to
educate the technical community in the
capabilities
of
supercavitating
vehicle
technology. The hydrodynamic performance
of supercavitating vehicles in maneuvers is of
particular interest.
Relevant quantities
include the transient forces and moments, as
well as transient cavity behavior, each of
whic h is important in the design of vehicle
control systems and gas ventilation schemes.
SUMMARY AND CONCLUSIO NS
This article has provided an overview of recent basic research and exploratory development results in the topic of
supercavitating high-speed bodies. Both experimental and modeling aspects of the Navy program have been discussed.
The focused effort by a nationwide team of scientists and engineers over the last decade has resulted in an improved
understanding of the physics of these devices, as well as an enhanced capability to predict their performance and assess
their utility. The confidence gained in overcoming the many technical challenges associated with vehicle and projectile
operation at very high speeds underwater has positioned the Navy to set more challenging objectives for the next few
years, including field testing of free-flying vehicles in maneuvers. Continued development of the resources established to
date will help ensure that suc h activities can be carried out safely and cost-effectively.
ACKNOWLEDGEMENTS
This research was sponsored, in part, by the Office of Naval Research, High-Speed Supercavitating Undersea Weaponry program, Dr. Kam Ng,
Program Officer, who also sponsored the preparation, publication, and presentation of this article. Continuing exploratory development of the
AHSUM concept and adaptation of the supercavitation flight simulator for analysis of supercavitating projectiles is sponsored by the Office of
Naval Research Counterweapons and Countermeasures program, Dr. Teresa McMullen, Program Officer, who also sponsored the AHSUM field
tests. Certain results presented herein were achieved under projects jointly sponsored by the Defense Advanced Research Projects Agency,
Dr. Theo Kooij, Program Officer, and by the Naval Undersea Warfare Center In-House Laboratory Independent Research and Bid and Proposal
Programs (Messrs. Richard Philips and Bernard Myers, Program Managers, respectively). The gun launchers used in the supercavitating
projectile tests were supplied by General Dynamics Armament Systems. The author’s would like to thank Dr. Ann W . Stokes, Mr. Jeff O’Brien
and Ms. Jamison L. Miller for their contributions to the vehicle dynamics and control systems simulation discussed in this article.
REFERENCES
Baker, G.L., and J.P. Gollub (1991) Chaotic Dynamics, Cambridge University Press, New York.
Fine, N.E., (1992) Nonlinear Analysis of Cavitating Propellers in Non-Uniform Flow, PhD thesis, Massachusetts Institute of
Technology, Cambridge, MA.
Fine, N.E., and S.A. Kinnas, (1993) “A Boundary Element Method for the Analysis of the Flow Around 3-D Cavitating Hydrofoils,” J.
Ship Research, 37(1).
Kirschner, I.N., (2001) “Results of Selected Experiments involving Supercavitating Flows,” Lecture Notes for the RTO AVT/VKI Special
Course on Supercavitating Flows, von Karman Institute for Fluid Dynamics, Rhode Saint Genèse, Belgium.
9
Kirschner, I.N., D.C. Kring, A.W. Stokes, N.E. Fine, and J.S. Uhlman (2000) “Control Strategies for Supercavitating Vehicles ,”
accepted for publication in the Journal of Vibration and Control. (Also presented at the Eighth International Symposium on Nonlinear
Dynamical Systems, Blacksburg, VA.)
Kirschner, I.N., J.S. Uhlman, Jr., A.N. Varghese, and I.M. Kuria (1995) “Supercavitating Projectiles in Axisymmetric Subsonic Liquid
Flows,” Proceedings of the ASME & JSME Fluids Engineering Annual Conference & Exhibition, Cavitation and Multiphase Flow
Forum, FED 210, J. Katz and Y. Matsumoto, editors, Hilton Head Island, SC.
Kirschner, I.N., N.E. Fine, D.C. Kring, and J.S. Uhlman (2001) “Numerical Modeling of Supercavitating Flows,” Lecture Notes for the
RTO AVT/VKI Special Course on Supercavitating Flows, von Karman Institute for Fluid Dynamics, Rhode Saint Genèse, Belgium.
Kring, D.C., N.E. Fine, and I.N. Kirschner (2001) “The Application of a Fast Solver in Marine Hydrodynamics,” Proceedings of the
First MIT Conference on Computational Fluid and Solid Mechanics, 2, K.J. Bathe, editor, Elsevier Science Ltd., Oxford, England.
Kring, D.C., N.E. Fine, J.S. Uhlman, and I.N. Kirschner (2000) “A Time-Domain Cavitation Model using a Three-Dimensional,
Boundary-Element Method,” Applied Hydromechanics, 2 74-3, Institute of Hydromechanics, Ukrainian National Academy of Sciences,
Kiev, Ukraine.
May, A., (1975) “Water Entry and the Cavity-Running Behavior of Missiles,” SEAHAC TR 75-2, Naval Sea Systems Command,
Arlington, VA.
Savchenko, Y.N., V.N. Semenenko, S.I. Putilin, Y.D. Vlasenko, V.T. Savchenko, and Y.I. Naumova (2000) “Theory of Stable Model
Motion with Ventilated and Unventilated Supercavities,” IHM Technical Report, National Academy of Sciences of Ukraine, Institute
of Hydromechanics, Kiev, Ukraine.
Savchenko, Y.N., V.N. Semenenko, Y.I. Naumova, A.N. Varghese, J.S. Uhlman, Jr., and I.N. Kirschner (1997) “Hydrodynamic
Characteristics of Polygonal Contours in Supercavitating Flow,” Proceedings of the Third International Symposium on Performance
Enhancement for Marine Applications, T. Gieseke, editor, Naval Undersea Warfare Center Division, Newport, RI.
Stinebring, D.R., (1976) Scaling of Cavitation Damage, MS Thesis, Pennsylvania State University, State College, PA.
Stinebring, D.R., and J.W. Holl (1979) “Water Tunnel Simulation Study of the Later Stages of Water Entry of Conical Head Bodies:
Phase II - Effect of the Afterbody on Steady State Ventilated Cavities,” TM 79-206, Applied Research Laboratory, Pennsylvania State
University, State College, PA.
Stinebring, D.R., M.L. Billet, and J.W. Holl (1983) “An Investigation of Cavity Cycling for Ventilated and Natural Cavities,” TM 8313, Applied Research Laboratory, Pennsylvania State University, State College, PA.
Uhlman, J.S., A.N. Varghese, and I.N. Kirschner (1998) “Boundary-Element Modeling of Axisymmetric Supercavitating Bodies,”
Proceedings of the 1st Symposium
on Marine Applications of Computational Fluid Dynamics, Hydrodynamics/Hydroacoustics
Technology Center, Naval Surface Warfare Center Carderock Division, Carderock, MD.
Varghese, A.N., J.S. Uhlman, and I.N. Kirschner (2001) “Partially Cavitating Axisymmetric High-Speed Bodies ,” (in progress).
10