This is a reproduction of a library book that was digitized

by Google as part of an ongoing effort to preserve the

information in books and make it universally accessible.

https://books.google.com
HYDROBALLISTICS

MODELING

by John C Waugh

and G.W Stubstad

.
HYDROBALLISTICS MODELING

By
John G. Waugh
and

G. W. Stubstad

Naval Undersea Center

San Diego, California

For sale by the Superintendent of Documents , U.S. Government

Printing Office , Washington , D.C. 20402
5
 ENGINEERING LIBRARY

623.57
FRONTISPIECE W35 h Engin
Missile heads used in water-entry pitch modeling.
 FOREWORD

Our sincerest appreciation is extended to the many persons who assisted in the preparation of
this volume. We should like to express our thanks specifically to the late Harold A. Eggers of the
Bureau of Naval Weapons and to Dr. Thomas E. Peirce of the Naval Ordnance Systems Command for
their encouragement and sponsorship of this work , and to the following Naval Undersea Center per
sonnel: Dr. Jack W. Hoyt for his careful and painstaking review of this work and valuable suggestions ,
Dr. Russell A. Nielsen , who also reviewed the work , and Mrs. Margery S. Ross for her careful and
efficient editing. Finally we should like to express our appreciation to Dr. William B. McLean,
Technical Director of the Naval Undersea Center and formerly of the Naval Ordnance Test Station ,
under whose direction this work was undertaken .

J. G. W.
G. W. S.
 PREFACE

Realization of the ocean's importance to man has grown rapidly during the last decade , and
with it has come increased activity in undersea research and development. Vehicles and weapons are
appearing that may operate at speeds ranging from a mere creep to higher than presently in use , along
trajectories having any orientation to the water surface, and at increasingly deeper submergence. All
degrees of cavitation will be encountered , in addition to cavities composed partly of air or of
highspeed exhaust gases..
The need for small -scale modeling techniques arises from the inadequacy of the three-phase ,
three -dimensional hydrodynamic theory and from the difficulties inherent in testing large missiles in
the uncontrolled environment of the open sea. Expenditures of time and money are bound to be
large in prototype testing , and the acquisition of data of adequate resolution and precision can be a
formidable task.
The ultimate purpose of modeling is , of course, the practicable prediction of prototype missile
behavior , an ideal that cannot always be achieved, particularly when the physical phenomena under
investigation are not fully understood. Yet model studies can provide insight into little-known
phenomena and prove useful in improving prototype design or in controlling the events that
influence prototype behavior.
Experimental investigation of water entry began at least 75 years ago , but most of the work
reported here has been done since World War II. Many specific shapes and some general
configurations have been observed, usually at low velocities (500 fps or less ) entering still water.
These missiles, together with their entry cavities and the surrounding water, represent only aa small
portion of the hydrodynamic-flow regime that may be encountered in the ocean. Ambientconditions
typical of the open sea and the new factors introduced by high-speed propulsion systems have not
been thoroughly investigated . If modeling is to continue as a useful tool , it must widen its capabilities
to keep pace with oceanography's growing needs.
It is hoped that the studies reported here , most of them conducted at the U.S. Naval Ordnance
Test Station, now the Pasadena Laboratory of the Naval Undersea Research and Development
Center, will provide a basis for assessing the present state of the modeling art, a springboard for
future experiments , and a guide for extending the modeling technique to include other phases of
hydroballistics .

J. G. W.
G. W. S.

Pasadena, California , 1972

vii
1
 CREDITS FOR ILLUSTRATIONS AND TABLES

Figures 1.1 , 7.1 , 7.2 , 7.3 , 7.4 , 7.5 , 7.6 , and 7.7 taken from “ Water -Entry Pitch Modeling,” by
John G. Waugh . Journal of Hydronautics, Vol . 2 , No. 2 ( 1968 ) , pp . 87-92 . Reproduced by
permission of the American Institute of Aeronautics and Astronautics. (Reference 96 )
Figures 3.7 and 4.1 taken from “ A Study of Splashes,” by A. M. Worthington. New York, N.Y., the
MacMillan Co., 1963. Reproduced by permission. (Reference 112 )
Figure 4.2 and Tables 4.1 , 4.2 , and 4.3 taken from “ The Impact of aa Solid on a Liquid Surface ,” by
E. G. Richardson . Proceedings of the Physical Society of London , Vol. 61 , Part 4 ( 1948 ) , pp . 352-66 .
Reproduced by permission of the Institute of Physics and the Physical Society. ( Reference 68 )
Figure 4.3 taken from “ Fluid Flow Patterns, ” by Garrett Birkhoff and Thomas E. Caywood. Journal
of Applied Physics , Vol. 20 , No. 7 ( 1949 ) , pp . 646-59 . Reproduced by permission of the American
Institute of Physics . (Reference 7 )

Figures 17.1 , 17.2 , 17.4 , 17.5 , 17.6 , 17.7 , 17.8 , and 17.9 taken from “ The Variable-Angle
>

Variable -Pressure Launching Tank ,” by J. G. Waugh . Published in “ Cavitation Research Facilities and
Techniques,” New York , N.Y., American Society of Mechanical Engineers , 1964. pp . 118-126 .
Reproduced by permission . (Reference 89 )
Figures 17.3 , 17.10 , 18.10 , 18.11 , 18.12 , and 18.13 taken from “ Techniques for Metric
Photography,” by J. G. Waugh , A. T. Ellis and S. B. Mellsen . Journal of the Society of Motion
Picture and Television Engineers , Vol. 75 , No. 1 ( 1966 ) , pp . 2-6 . Reproduced by permission of the
Society of Motion Picture and Television Engineers . (Reference 94 )
Figures 17.11 , 17.12 , 17.13 , 17.18 , 17.19 , 17.20 , 17.21 , 17.22 , 17.23 , 17.24 , and 17.25 taken from
“ The Hydrodynamics Laboratory of the California Institute of Technology,” by R. T. Knapp , Joseph
Levy , J. P. O'Neill and F. B. Brown. Transactions of the American Society of Mechanical Engineers ,
Vol. 70 , No. 5 ( 1948 ), pp. 437-57 . Reproduced by permission of the American Society of
Mechanical Engineers . ( Reference 30 )

Figure 19.4 taken from “ Cavitation Research Capabilities of the Lockheed Underwater Missile
Facility,” by Robert L. Waid . Published in “ Cavitation Research Facilities and Techniques,” New
York, N.Y., American Society of Mechanical Engineers, 1964. pp. 127-131 . Reproduced by
permission . (Reference 87 )

Figures 19.6 , 19.7 , and 19.8 taken from “ The Variable-Atmosphere Wave Tank , ” by J. G. Waugh and
A. T. Ellis. Published in “ Cavitation Research Facilities and Techniques,” New York, N.Y. , American
Society of Mechanical Engineers , 1964. pp . 114-117 . Reproduced by permission . (Reference 93 )

ix
 CONTENTS

FOREWORD . V

PREFACE ... vii

CREDITS FOR ILLUSTRATIONS AND TABLES . ix

I THEORETICAL MODELING 1

1 Hydroballistics Modeling Theory 1

Use of small-scale models to simulate prototype behavior ; the scaling laws
involved ; scaling parameters : Froude, Reynolds number, Weber number,
cavitation number, and gas density .

2 Underwater -Launch Modeling 6

Modeling techniques ; scaling constraints ; limitations imposed by physical
properties of gas ; selection of gases ; partial scaling and selection of partial-scaling
techniques .
II EXPERIMENTAL MODELING 13

3 Water-Entry Phenomena 13
Factors influencing water entry ; airflight; impact ; flow formation ; the open
cavity ; cavity closure ; the closed cavity .
4 Evolution of Experimental Hydroballistics Modeling 51
Contributions of early investigators ; water entry of spheres and missiles with
hemisphere , ogive , cone , and plate head configurations ; cavity and splash studies .
III PROTOTYPE AND MODELING EXPERIMENTS 69

5 Full-Scale Whip and Deceleration at Water Entry 69

Importance of whip and water-entry deceleration to full -scale missile behavior;
effects of head shape , angle of attack and velocity .
84
6 Whip and Underpressure
Effect of underpressure on whip ; its importance in missile behavior ; tests of
vented and unvented full -scale missiles and models ; effects of varying the
atmospheric pressure and surface tension.

7. Water-Entry Pitch Modeling 113

One-to -one Froude and cavitation-number scaling ; the addition of partial
gas-density scaling ; prototype and model water penetration.
xi
 8 Underwater Trajectory Modeling . 123
Importance of boundary-layer flow in model water-entry behavior; artificial
production of turbulence ; effect of a turbulent boundary layer on model and
cavity behavior; effects of entry angle of attack, atmospheric pressure, and entry
velocity .
9 Water-Entry Cavity Modeling 145
Role of cavity in missile performance ; sensitivity of cavity to Froude and
cavitation-number and gas-density scaling ; vertical and oblique cavities ; vertical
and oblique missile water-entry in-cavity behavior.
10 Exploratory Earth-Entry Studies 178
Launching of models into clean sand ; observations of trajectory, whip, and
cavitation .

11 Water-Exit Studies 186

Comparison with water entry ; perturbation of missile attitude at water exit ;
effect of trajectory angle, cavitation number, unstable cavity flows, missile
shape, and Reynolds number ; design of water-exiting service missiles .
IV PHENOMENOLOGICAL STUDIES . 217

12 Water-Entry Splash Contours . 217

Vertical and oblique entry of spheres at atmospheric condition ; description of
splash contours.
13 Wake Formation 222
Fully wetted flow behind a sphere and a plate-cylinder missile ; development of
the wake ; vortex formation and behavior; cavitating flow behind a sphere and a
plate-cylinder missile ; cavity , vorticity , and re-entrant jet formation and
behavior.

14 Wake Effects 244

Shedding of vortices from base of missile under fully wetted flow ; correlation
with missile velocity perturbations; flow over fins.
15 Water-Entry Cavities of Spinning Spheres 251
Cavity character; separation point ; sphere drag coefficient ; boundary layer
effect .

16 Effects of Gravity on the Entry Cavity 257

Removal of gravitational effects by launching in a free-fall system ; vertical
water-entry cavities of spheres and rods ; cavity closure ; effects of ambient
conditions ; splashes.
V SPECIAL EXPERIMENTAL TECHNIQUES AND HYDROBALLISTICS FACILITIES 267

17 Facilities and Equipment 267

Variable-Angle Variable-Pressure Launching Tank ; Controlled-Atmosphere
Launching Tank, Fixed Angle Launcher; Variable Angle Launcher ; Open
Launching Tank ; Liquid Interface Tank ; Vertical Drop Tanks ; Sectional Tank.

xii
 18 Experimental Techniques 303
Preparation and maintenance of launching tank water ; creation of desired
atmosphere ; control of Reynolds effects; preparation of models ; photography ;
flow -marking techniques .

19 Hydroballistics Facilities Design .. 314

Facility design ; instrumentation ; tank construction ; tank materials ;
photography; advanced design ; recent developments in tank design ; NUC
Variable Atmosphere Tank ; Lockheed Underwater Missile Facility ; NOL
Hydroballistics Facility; CIT Variable Atmosphere Wave Tank ; ARL Controlled
Atmosphere Tank.
VI THE CHANGING STATUS OF MODELING 327

20 Present day partial knowledge of modeling ; need to establish or confirm actual

scaling conditions in many areas ; need for further study of environmental effects
on missile behavior ; hydroballistics problems requiring further study .
NOMENCLATURE .
331

BIBLIOGRAPHY 335

SUBJECT INDEX . 343

xiii
 PART 1
THEORETICAL MODELING

CHAPTER 1
HYDROBALLISTICS MODELING THEORY

The use of small- scale models to simulate of both gravitational (Froude ) and viscous
prototype conditions is a common practice in (Reynolds) scaling. A hydroballistic modeling
aeronautics , ship design , and other areas of fluid system must be designed to scale one force ratio
mechanics . Modeling occurs when a model fol (i.e. , one scaling or modeling parameter) and
lows a trajectory geometrically similar to that of minimize the other.
a prototype and reproduces the attitude of the Because the relative importance of Froude
prototype at corresponding positions along the and Reynolds forces can be adjusted through
trajectory. Modeling is accomplished by apply. selection of missile velocity , hydroballistics
ing a force system to the model that is similar to modeling is possible. At the lower, laminar -flow
the system acting upon the prototype velocities , the Reynolds effects cannot be
(Reference 61 ) . ignored. Either corrections must be made for the
The problem to be resolved in modeling is viscous forces in a Froude -scaled system or a
the identification of each of the involved forces specially restricted system must be devised. The
as it affects the motions of the missiles . In former technique is a long-established practice in
hydroballistic phenomena of practical interest, extrapolating towing-tank results to ship design .
the inertial , gravitational, and viscous forces are An example of a restricted system is the use,
of primary concern . Since inertia is always with Froude scaling, of models sufficiently large
important in the velocity range to be considered , that both model and prototype velocities are
the other forces are compared to it. Either the above the critical Reynolds number, and viscos
gravitational or the viscous force can be scaled ity effects can be neglected. Another example
simultaneously with the inertial force, but not (Reference 39 ) is the modeling of high-speed
both, unless different liquids are used in the maneuvers of deeply submerged submarines in
model and in the prototype systems . Unfortun the vertical plane . Here Froude scaling can be
ately, when the prototype liquid is water, there neglected and the model operating speed (if it is
is at present no known liquid for use in the above the critical Reynolds number ) becomes a
model system that will satisfy the requirements matter of free choice .

11t
It is possible that aqueous polymer solutions will permit simultaneous scaling of these parameters .
1
 Froude scaling can be successfully utilized negligible. The flexure number must be con
in the higher-velocity regime because the. sidered only if the missile becomes distorted
Reynolds forces become essentially constant under the load of impact or of the hydrody
once turbulent flow has been established . Model namic forces.
ing cannot be expected in the transitional regime The Froude number , F , is the ratio of the
between laminar and turbulent flow , although it inertial and gravitational forces and is defined as
can sometimes be achieved by triggering turbu
lent flow artificially. The upper velocity limits F = =
V
( 1.1 )
of Froude scaling occur when the gravitational (dg ) /2
forces become negligible compared to those of
inertia, or when the forces of compressibility ,
either in the missile or in the surrounding fluid , The cavitation number , 0, is the ratio of the
become significant. Then Mach scaling is used . static and dynamic pressures and is given by
Models and prototypes that satisfy both
the Froude and the Reynolds scaling laws are -

geometrically similar in external shape and in 0 = = Pa + Pwgh - (pg + pv ) ( 1.2 )

center of gravity (CG ) location. Their masses 12 p .,w v2
/ 2P
scale as the third power and the moments of
inertia as the fifth power of the ratio of missile The gas-density coefficient, p ', is defined as the
linear dimensions (i.e., model diameter/ ratio of the gas density in the model system to
prototype diameter ) . If different liquids are used the density of air at standard conditions of
in the model and in the prototype systems , the temperature and pressure . 2
masses and moments of inertia must also be
scaled directly as the respective liquid densities. Pm
The Froude and the Reynolds scaling laws, e' ( 1.3 )
however, require the models to operate at Pstd
different velocities . Froude scaling requires
smaller-than-prototype models to operate at
lower-than-prototype velocities ; Reynolds scal The Reynolds number, R, is the ratio of the
ing calls for higher-than -prototype velocities inertial and viscous forces.
with the smaller models .
The theory of modeling evolved from basic vd
=

concepts of dimensional analysis, which is no R ( 1.4 )

V
more than an attempt to relate physical
phenomena through ancient Greek ideas of
The Weber number, T , is the ratio of the inertial
geometrical similarity, ratio, and proportion.
The extensive literature that describes both and surface tension forces .
dimensional analysis and the theory of modeling V
has been outlined by Garrett Birkhoff (Refer T = ( 1.5 )
ences 5 , 6 , and 55 ) and discussed by F. S. Burt (S /Pd)1/2
(Reference 10 ) .
The important hydroballistics modeling
parameters in a Froude-scaled system are the The Mach number is the ratio of the missile
Froude and cavitation numbers , gas-density velocity and the velocity of sound in the fluid to
coefficient, and Weber numbers. be penetrated .
The Reynolds and Mach numbers are also
V
of interest because the validity of the modeling Mach -

( 1.6 )
technique depends upon these parameters being с

2In the NUC studies, the reference density was that of air at 20°C and 740 torrs, i.e. , the average ambient density at
the Morris Dam Torpedo Range near Azusa, Calif. , where prototype launchings were made.
2
The flexure number, f, is the ratio of inertial to and therefore the fluid temperature scaling
flexural forces. constraint is

f
v2 Iwm = T
Tw p ( 1.14 )
( 1.7 )
n? l?
Since the density of water varies only slightly
where no is the fundamental transverse fre with temperature over the usual temperature
quency of the missile . range , the constraint of Equation 1.14 is prob
In water entry or exit, it has been custom ably not important for water-density scaling ;
ary to define the parameters of the model over a temperature range of 0 - 25°C, the
system arbitrarily in terms of the velocity of the density of water varies only about 0.3 %.
missile at the first contact with the air-water The corresponding points along the geo
interface, i.e., at the moment of entry or exit . metrically similar trajectories are identified by
At the interface , the static head , pwgh , is zero equal distances in diameters:
and , in water entry , the cavity has not yet
formed . The cavitation number of the water
entry system thus reduces to Sm
=

Asp ; h m
= the ( 1.15 )

o
Pa - (pg + pv) Pa The time similitude is
2 2
( 1.8 )
1/2PWV We . V
1/2 pWe
t =

if Pg + Py is small compared to pa :
Im ( smm ) = vt (sp); tmm (hm) = Vītoche)
im р р (hp р

( 1.16 )
The modeling constraints and the model
prototype relationships derived from the asp
( ) = As,
=

Sm (tm (t ); hm ( tmm ) = Ahр (5р )

Froude -scaling law can be developed on the basis
of geometrically and dynamically similar models ( 1.17 )
and water as the fluid for model and prototype.
Let the ratio of the model diameter to the where tm and tp are the time intervals required
prototype diameter (modeling scale factor) be by the model and prototype to reach corre
given by sponding points .
Other model-prototype relationships are
flos

λ ( 1.9 ) =

d.
V
Ym (tm ) = vīvp (top) р ( 1.18 )

The modeling constraints for dynamic similarity =

( 1.19 )
are thus am ( tm ) = 2p(to )
l
m
= λΩ no ( 1.10 )
$m (tm ) = $ (t )
m ( 1.20 )

=
m
m 18mр ( 1.11 )
vas(tm ) = $ (tp) ( 1.21 )

I =
‫ܢ‬5
m р ( 1.12 )
aj(tm ) = šp( top) ( 1.22 )
The fluid (water ) density scaling constraint
is where v and a are the missile velocity and
Pwm
acceleration , and ſ represents any angle , i.e.,
Pwp ( 1.13 ) pitch angle, 0, trajectory angle, š , and angle of
3
attack, a (as ' shown in Figure 1.1 ), as well as Scaling of this type is commonly referred to as
yaw, y , and roll , 0. The Š and Š are angular Froude and cavitation -number scaling. It scales
velocity and angular acceleration , respectively. static pressure on the water and cavity surfaces
to the extent that pressure is constant over
CG
connected regions of the air, but it does not
scale the vapor pressure of water, nor does it
scale the air (gas) dynamic pressure,Yapvz. It
a has been found experimentally that this dynamic
pressure acts to close the water-entry cavity;
hence , reducing the dynamic pressure delays
ur
है

Horizontal
cavity seal and increases cavity size by allowing
a larger volume of air to enter. Modeling the
ө water -entry cavity size requires that the gas
density also be scaled :
Trajectory of CG

Pm Pp ( 1.24)
Figure 1.1 . Attack , pitch , and trajectory angles; all angles
are negative. ( From Reference 96. )

It is evident that cavity size and shape This extension of the Froude law is called
affect the interfacial and underwater trajectory Froude, cavitation -number, and gas-density
of the missile and that successful modeling of scaling.
missile motion within a cavity necessarily Scaling the Weber number5 requires that
includes modeling of the cavity . Cavity modeling the surface tension of the water be scaled as
requires that the ratio of the pressure forces
acting to collapse the cavity to the dynamic
pressure acting to open the cavity be preserved .
Thus , the cavitation number must be scaled
Sm = 125 ( 1.25 )

according to the Froude law ; this is accom

Some reduction of surface tension in the model
plished3 by scaling the atmospheric pressure4 in
the model system as system can be achieved, without significantly
altering the properties of the water, by adding
Pam AP ap ( 1.23 ) minute quantities of Aerosol to the water. A

0
(Pa - Pdp + Pwpsh р =
AP, -Pep*Pwp8 Ah. ,
р p

2
Y2Pwor?
wp p р
12p
wpp
2v2 m

4In the NUC and NUC -CIT studies, 1 atmosphere pressure is defined as 740 torrs, i.e., the average ambient pressure at
the Morris Dam Torpedo Range, Azusa, Calif., where prototype launchings were made.

14112
5

T =
PVPwede
VVPwpp
' =
р Newpool р

✓Ssр p /p Vs.p 12 m

4
solution of 0.1 % Aeroso16 by weight results in a model systems significantly affected by surface
lowering of the surface tension from 76 dynes/ tension forces.
cm at 20°C to 40 dynes /cm , a reduction of 47%. Scaling of the flexure number requires that
This is the reduction required to scale the Weber the bulk modulus of elasticity of the prototype
number when 1 is 1 /1.38 , a higher scale factor be equal to that of the model7 divided by the
than is ordinarily used. Thus , only partial scaling square root of 1. If the missile is so flexible that
of Weber number is possible in many Froude this scaling is required , materials are available
scaled model systems. Fortunately , this partial that allow scaling over a restricted range of 2.
scaling has been adequate in the water-entry

6
6 Aerosol OT , dioctylsodiumsulfosuccinate (Reference 2 ) .

?7( From the flexure number vă/ne2,nom = nop/11/2. For a uniform free-free beam (Reference 61,pages 40-42) , no =
11.2 /7 ) (cK/L2) where c is the velocity of sound in the missile material , K is the radius of gyration of the beam
cross -section with respect to the neutral axis, and l is the length of the beam . Then cm = 11 /2cp where c =
(E /PM ) 1/2, and E and PM are the modulus of elasticity and density of the material , respectively. Another form of
the flexure number is l2v2/K2c2 .

5
CHAPTER 2
UNDERWATER - LAUNCH MODELING

.
2
Interest in the problems arising when a Gas expansion over the launching and
missile emerges from water created a need for gas-bubble expansion phases is free
water-exit modeling and for the modeling of and adiabatic .
underwater-launch systems ( Reference 88 ). 3. The constraints of Froude and
Examination of the physical aspects of the exit cavitation -number scaling (necessary
problem leads to the conclusion that it should for missile -trajectory modeling ) apply.
be possible to model the underwater flight, exit,
and subsequent air trajectory of a missile in a 4. The effects of viscosity, surface ten
Froude cavitation -number scaled system , sion , and water compressibility are
and also that the possible need for the additional negligible.
constraint of gas-density scaling must be recog It will be shown that the additional pre
nized . launching constraints

TECHNIQUES OF MODELING Pgm AP SPO ( 2.1 )

o
In underwater-launching systems where
water is in contact with the missile during part where pg is the gas pressure ,
or all of its launching phase , the character of the
Tg =

( 2.2 )
flow and the missile motion must be modeled. If mo TP .
gas is ejected into the water as the missile base
emerges from the launcher muzzle, the gas where Tg is the gas temperature,
bubble may affect the behavior of the missile
and therefore it must be modeled . V =
13V P ( 2.3 )
m. o
In the following discussion a compressed
gas launching system is assumed in which water where V is the gas volume, and
is in contact with the missile over its launching im
phase and a gas bubble is formed on missile base Үр (2.4)
emergence. The following assumptions are made : where y is the gas specific -heat ratio , will be
1 . Both prototype and model under sufficient for modeling missile motion over the
water launching systems are geomet launching phase. Later it is shown that these
rically scaled. constraints are also sufficient for modeling the

1
Base emergence is considered to take place when the lower cylindrical edge of the missile clears the launcher muzzle ,
allowing free ejection of launcher gas into the water.

2 Free gas expansion will not obtain over a brief regime following missile-base emergence from the launcher muzzle
due to the throttling effect of the aperture between base and muzzle , and the gas bubble will not be modeled by the
scaling techniques discussed. It is assumed that this regime is sufficiently brief that significant deviations in
gas-bubble modeling will not be introduced.
6
gas bubble provided gas expansion is adiabatic or the friction -generating area scales as 12. If it is
the same operating gas specific heat ratio can be assumed that the pressure on the friction
assumed for both systems . If gas density is also generating area scales as 1, and the coefficient of
scaled, the constraint of Equation 1.24 must be sliding friction is the same for both systems ,
included . then the frictional force scales as 13. The force
The constraint of Equation 2.2 is satisfied accelerating the missile must therefore scale as
by holding the prototype and model- launching 13. Since the mass of the missile scales as 13 ,
systems at the same temperature (Equation 1.14 missile accelerations scale directly and velocities
and subsequent discussion ) and the constraint of scale as 11/2 . In summary, Equations 1.18 , 1.19 ,
Equation 2.3 is satisfied by geometric scaling of 2.8 , 2.9 , and 2.10 obtain at missile-base emer
the launching systems . Using Equations 2.1 , 2.2, gence from the launcher muzzle .
2.3 , and the isothermal equation of state For gas bubbles, neglecting fluid vapor
pressure , the cavitation number o assumes the
R' form
nR'T ;
EVV
PSV
g
=
; Pg M
PT g ( 2.5 )
P ++ Pwgh - Pg
W
gives o ( 2.11 )
n (2.6) 12P w v2
m
= 14np W

It can be seen that o is scaled at the instant of

where n is the quantity of gas (in moles) in the
launcher system . missile -base emergence from the launcher muz
At ordinary pressures , the adiabatic equa zle. Using gas -pressure scaling, modeling of the
tion of state of a gas is given by bubble will obtain over the regime for which gas
expansion can be assumed adiabatic.
=
c; TV7-1 = c'=

( 2.7 ) The scaling constraints above can readily be

PgV g
extended to include modeling of underwater
where c and c'are constants . From the pre launching systems that have a diaphragm over
launching constraints (Equations 2.1 , 2.2 , 2.3 , the launcher muzzle and gas in the tube before
and 2.4 ) , the assumed geometric similarity of launching. As initial boundary conditions, it is
assumed that the same kind of gas is used, and
the prototype and model- launching systems, and its volume, temperature and pressure (hydro
the adiabatic equation of state (Equation 2.7 ), it
is evident that for geometrically scaled positions static pressure at the launcher muzzle ) are scaled
of the missiles in the launching systems in the prototype and model launching systems,
but that diaphragm rupture takes place at first
V
m
=

= 23vp (2.8 ) missile motion. Hence Equations 2.8 , 2.9 , and

2.10 are satisfied .
Pom AP gp ( 2.9 ) A similar analysis shows that missile accel
erations and velocities are properly scaled (Equa
T
gm
=
T
gp (2.10) tions 1.18 and 1.19 ) ; the cavitation number
(Equation 2.11 ) is scaled also . Therefore, model
It will be shown that missile accelerations ing of the nose bubble will obtain over the
for geometrically scaled missile positions scale regime for which gas expansion can be regarded
directly as Equation 1.19 and hence (since initial as adiabatic.
missile velocities are zero ) missile velocities scale It is evident that after the initial gas-bubble
as Equation 1.18 . Since the missiles are geomet expansion phase, heat transfer between the
rically similar, from Equation 2.9 gas pressure bubble gas and the surrounding water may cause
forces on the missile scale as 13. From Equa deviations in the bubble modeling, but the
tions 1.13 , 1.15 , and 1.23 , the sum of the gas-bubble behavior may have little effect on
atmospheric and hydrostatic pressures acting on missile behavior after the initial expansion
the missile scales as I and hence the force due to phase , and thus the previously discussed scaling
these pressures scales as 13. Assuming similarity constraints will be adequate.
of flow in both systems, the force required to The scaling of the gas specific-heat ratio
produce the same fluid acceleration in both (Equation 2.4 ) without regard to gas-density
systems scales as 13. For geometrical similarity, scaling offers no problems because it may be
7
accomplished by using the same gas in the model where y is assumed to be constant over the range
and in the prototype systems. The additional of temperatures and pressures encountered
constraint of gas-density scaling (Equa within the system . It should be noted, however,
tion 1.24 ), however, involves the use of different that y generally increases with increasing pres
gases in the two systems. sure and decreases with increasing temperature.
The adjustment of gas density and specific The following values of y for air (taken from
heat ratio in the model system to their scaled Reference 16 ) give an indication of this
values as given by Equations 1.24 and 2.4 variation .
generally entails the use of gas mixtures. Let Mp
be the molecular weight of the gas used in the Lb/in2abs
prototype system (if air, Mp = 29 ) and let M1
Mi ,
t, °F 0 2,000
M2 .. Mr be the molecular weights of the 1,000 4,000
components in the model-system gas mixture. 0 1.40 1.58
Let nl , n2 .. nr be the respective mole 200 1.39 1.47 1.53 1.62
fractions of the components . Then
400 1.38 1.43 1.46 1.51
r

600 1.37 1.39 1.42 1.45

»

n. 1
(2.12 )
VA

and
It is apparent that a significant variation of y
occurs as pressures approach the order of
1000 psi . For examp
e le , a 1/10 scale model of a
M
m
Ση
i= 1
n.M.
1 ( 2.13 ) prototype operating in the 1000-psi range would
require pressures only on the order of 100 psi.
where Mm may be regarded as the “molecular Thus , using air in both systems would not
weight” of the model-system gas mixture . From suffice to scale in this hypothetical modeling
Equations 1.24 , 2.5 , 2.9 , 2.10 , and 2.13 problem.
Mm 1
The partial pressure Pgi of each gas com
ponent is given by
Mh
рe
M Ση1;M;
,Μ, = 5λ ( 2.14 )
p
P i= 1 Pgei n PS ( 2.16 )
The quantity of heat necessary to raise the
temperature of 1 gram molecular weight of gas where pg is the gas-mixture pressure. That is,
by 1 °C at constant volume is the molecular heat
at constant volume, Cy. The quantity of heat r
necessary to raise the temperature of 1 gram ps
Ση,P , = =
W?

molecular weight of gas by 1 °C at constant

pressure is Cp. Now let Cvi and Cpi be the
Σ
i= 1
i
i= 1
९६
i= 1
ni Pg (2.17 )
molecular heats of the gas component of molec
ular weight Mi and mole fraction ni. If the usual It should be noted that the vapor pressure
assumption is made that the internal energy of of the gas usually decreases with increase in
the mixture is equal to the sum of the internal
energies that the component gases possess sepa molecular weight. If the vapor pressure of the
rately at the same temperature , the specific-heat gas exceeds the partial pressure at which it is to
ratio of the gas mixture in the model system is be used , it will liquefy. Also , gases of increased
given by molecular weight become increasingly poly
atomic and the specific-heat ratio decreases
rapidly . Thus the number of available gases
Ση,C , С.
suitable for simultaneous scaling of gas density
i= 1 and the specific-heat ratio decreases with
im ( 2.15 )
11

r decrease in the modeling scale factor 1. Addi

tional restrictions are non-toxicity , non
C

<n;ov;
i= 1
inflammability , and chemical inertness under
modeling conditions.
8
SELECTION OF WORKING GASES scaling would still be good : complete scaling of
Of the gases that satisfy the constraints of the specific -heat ratio and 93.4% scaling of the
gas-density and specific-heat ratio scaling where gas density would be attained .
air is used in the prototype system , a mixture of Other heavy gases (Reference 67 ) may be
xenon and Freon C-318 (octafluorocyclobutane) applicable to modeling studies. The investigator
gives one of the smallest modeling scale factors. must determine a compromise between the
The value of the scale factor obtainable with this desired degrees of scaling and the cost and
mixture can be determined from the table availability of the gases suitable to attain them.
below . Xenon, for example, is expensive .
Air Xenon Freon C-318
1.059
Partial Scaling
Specific-heat ratio , y 1.4 1.666
(Ref. 40 ) (Ref. 67 ) Scaling of one factor at the expense of
Cp, cal/mole °C 4.97 38.353
others may cause large differences between the
(Ref. 25 ) ( Ref. 67 ) model and prototype behavior ; it is then neces
Cy, cal/mole °C 2.98 36.216
sary to consider a compromise partial scaling of
several factors in order to obtain model data of
Molecular weight, M (29 ) 131.3 200

(Ref. 25 ) ( Ref. 67 )
acceptable accuracy. Partial scaling can be
achieved through the selection of physical
Let n be the mole fraction of xenon in the properties , and the underwater-launch system
binary gas mixture. Then from Equation 2.12 , scaling can also be adjusted by altering the gas
the mole fraction of Freon C-318 is 1 - n. From expansion during the launching phase.
Equation 2.15 and the table above for specific It is assumed , as before , that the prototype
heat ratio scaling, and model systems are geometrically and
dynamically scaled, and, furthermore, that the
=
4.97n + 38.353 ( 1 - n)
-

base -emergence velocity of the missile is scaled.

Yp (air) = 1.4 =
2.98n + 36.216 ( 1 - n) (Since only launcher gases are involved, the
( 2.18 ) subscript g is omitted from the symbol denoting
gas pressure in the ensuing equations . )
From Equation 2.18 , n = 0.9393 . Therefore, the The work performed by the gas in launch
mole fraction of xenon is 0.9393 , and the mole ing the missile is the sum of the kinetic energy
fraction of Freon C-318 is 0.0607 . From imparted to the missile and water and the work
Equation 2.14, expended in overcoming friction and expansion
against hydrostatic pressure . If W is the work
Mm 1 performed by the gas , then
M 29
[ 131.3 ( 0.9393) + 200 ( 0.0607 ) ]
р Vi1
Sv. pdV
=
W ( 2.20 )
= 4.67 ( 2.19 )
λ
where V. and Vi are the initial and final
The vapor pressure of Freon C-318 at
launcher-gas volumes. From Equation 2.7 and
ambient temperature ( 70 ° F ) is 40 psia (Refer the initial boundary conditions p = Po when V =
ence 67 ), which is not likely to be exceeded by V. ,
scaled model-system launching-gas pressures
with consequent danger of gas liquefaction . 7-1
With this gas mixture, complete scaling is W = P. V. 1 |
possible with a modeling scale factor as small as
=

1 = 1 /4.67 . For 1 /5-scale modeling (a = 1/5 ), the

7-1
[6 ] ( 2.21 )

3Heavy gases (and liquids that would form heavy gases at reduced pressures for scaling gas density over the water
surface) are available from : Union Carbide Corporation, Linde Division, New York, N.Y.; E.I. du Pont de Nemours
and Co., Inc. , Freon Products Division , Organic Chemicals Department, Wilmington , Delaware; Halocarbon Products
Corporation, Hackensack, New Jersey; and Minnesota Mining and Manufacturing Co. , Chemical Division, St. Paul,
Minnesota .

9
If y = 1 ( isothermal expansion ) If it is assumed that the energy imparted to
the water and the flow pattern is negligibly
Vi1 affected by small variations in the missile
W
W
=

= p. V. In V
‫܂‬ o
( 2.22 ) position – time relation during the launching
phase (provided that the constraint of scaled
base-emergence velocity is satisfied ), Equa
From Equation 2.20 , the work performed by gas tions 2.25 and 2.26 may be equated.
expansion in the prototype and model systems That is ,
would be given by
no -1 n -1
р m

Pp.'P P Pm Vim m

and
W

[ ]p
Үр -1
-
1

1P1)
V
( 2.23 )
[
[0 ]
Ym - 1 V
m
1

n
р
Pm.m m

=
Υ .m -1

W
m
Pm o V m o

Ym - 1
1

9 ]
vm

V m
1
( 2.24) [8 ] Yp - 1
-

and the initial and final temperatures of the

V
m
1
( 2.27 )

launcher gas are obtained from the second form

It is obvious that under the launching of the adiabatic equation of state (Equa
constraints of Equations 2.1 , 2.2 , 2.3 , 2.4 , and tion 2.7 ). For properly scaled model-launching
=

2.6 , Wm = 14Wp. With the assumption that the conditions

work done on (i.e. , energy imparted to ) the
model and prototype missiles is in the same Tn - 1
р
proportion , and applying the constraint of Equa m

tion 1.11 , it follows that Equation 1.18 obtains T

gm 1
= T
gm o ( 2.28 )
and the missile velocities are Froude-scaled . V
m
Now assume a properly scaled model sys
tem where Ym = Yp
A variety of combinations of non -scaled launch
r
ing conditions that satisfy Equation 2.27 may
PmoV m
p -1. now be used to approximate scaled launching
o
conditions. The elements of the combinations
W
m

Yp -
-
1
[ )
1
V
m
1
(2.25 ) depend upon the parameters considered by the
investigator to be of greatest importance for
scaling: the gas density , the quantity of launcher
gas , its initial or final temperature , or some
and the same model system where ym + compromise combination thereof.
It is apparent that perfect scaling of both p
and Υy is possible only under limited circum
Υ .m
stances ; it will usually be necessary to select a
P'm o
V'
m m -?7 partial -scaling technique . The importance of
W
m
im - 1
[A m
( 2.26 ) gas-density scaling is demonstrated by existing
experimental evidence (e.g. , Chapters 7 and 9 ) ,
but experimental scaling of the gas specific -heat
ratio is virtually unknown . A theoretical evalua
The primed quantities in Equation 2.26 are not tion must be made. Inspection of Equa
necessarily scaled . The final gas volume at tions 2.14 and 2.15 shows that the specific
missile-base emergence, Vm ], is scaled because heat-ratio scaling will diverge rapidly with only
the same launcher is used in both systems. slight improvement in gas-density scaling.
10
 When the dimensions of an actual launcher scaling. Gas density appears to be of lesser
and assumed initial launching conditions are importance in the underwater-launch system (no
substituted into Equation 2.27 , various pres surface closures occur ) . It is possible to achieve
sures , volumes , and temperatures within the sufficient gas-density scaling to evaluate this
system can be calculated with the aid of premise experimentally . It is probable that the
Equation 2.5 or 2.7 . The importance of scaling v partial gas-density scaling available simultane
can be assessed numerically by assigning differ ously with a y of at least 1.3 will be adequate to
ent values in Equation 2.27 and comparing the model the underwater-launch system under
results . The importance of scaling y during a free consideration .
expansion can be explored further by consid A preliminary analysis such as this is based
ering the effect of change in y with the upon an estimate of the physical phenomena
expansion of a simple spherical bubble . By involved, and some of the assumptions made in
examining this parallel physical system, it is formulating the theory are necessarily “ ideal.'
possible to observe the effects of y in the The derived scaling laws should be used as
absence of the assumptions used in formulating guidelines for initial experimentation and should
Equation 2.27 (Reference 35 ) . be continually reassessed in the light of accumu
The modeling equations indicate that a y of lating physical data. Thus physical parameters
less than about 1.3 in the model system will within the system can be measured instead of
cause significant variation between the model assumed ; the prototype gas expansion, for exam
and prototype gas expansions . Fortunately , the ple , will in reality be neither isothermal nor
gas-density -scaling data suggest that partial scal adiabatic, and the specific heat ratio and gas
ing of p may be adequate. In the water-entry density necessary for scaling can be established.
system described in Chapter 9 , where gas density A continuous feedback from experiment to
is one of the primary physical parameters theory tends to evolve a scaling technique that
influencing the phenomena, good modeling has both sound theoretical basis and physical
resulted from 80% scaling and generally usable reality .
results were obtained with only about 70%

11
 PART II .
EXPERIMENTAL MODELING

CHAPTER 3
WATER-ENTRY PHENOMENA

Knowledge of the physical forces to be where the large , dominant forces of gravity and
scaled is a requisite in modeling. It is often inertia may nearly cancel each other. By deter
necessary, however, to begin a modeling pro mining the occurrence of cavity separation or
gram before this knowledge is complete , and closure , a small, seemingly unimportant force
then the investigation becomes a dual-purpose can greatly influence cavity behavior and , hence ,
study that seeks a basis for greater under missile behavior. The mechanisms influencing
standing as well as a determination of the cavity behavior have been extensively observed
relative importance of the forces involved . It has but are not fully understood .
been thus with water entry . The scaling laws The factors that might influence the water
outlined in Chapter 1 evolved as experimental entry phenomena, condensed from Refer
knowledge grew. ence 61 , are listed in Table 3.1 and their dimen
In water entry , cavity formation is irrevo sions in the plv system in Table 3.2 .
cably blended with missile performance. The When an airborne missile hits the water, it
physical properties of missile, gas, and liquid creates a cavity that persists and grows during
together establish the forces that combine with the underwater trajectory, as seen in Figure 3.1 .
gravity and inertia to determine the entry (This sequence of exposures, showing progres
phenomena. It is reasonable to assume that the sive stages of a missile entering the water
force exerted by the missile nose upon the water obliquely , was printed on a rotating disk of film ,
opens a cavity that continues to widen by virtue using a camera with a sector-shaped slot and
of its own inertia until gravity causes the water stroboscopic flash lamp . The camera is described
to return to its undisturbed position. Yet such in detail in Part V) .
comparatively small forces as gas flows within Initially , the cavity extends back to the
the cavity , minor physical properties of the surface of the water. As the missile travels along
liquid ( surface tension , dissolved gas nuclei , its trajectory, the cavity becomes longer and
etc. ) , and even the surface of the missile are thinner until it closes from the surface in either
important in determining cavity behavior. Small of two ways : the surface of the water can close
forces become significant at critical points of over the cavity or the cavity can “ neck down”
cavity closure and at missile-cavity separation and break off at some point beneath the surface
13
 TABLE 3.1 . Factors Influencing Water Entry

Amenable
Parameter to Scaling

Properties of the liquid

Density . . Yes
Kinematic viscosity . No
Compressibility No
Vapor pressure No
Surface tension No
Imperfections in fluidity
Cohesion No
Ability to support tension stress (momentarily ) No
Stress-time relaxation characteristics No
Dissolved particles No
Gas nuclei No
Adhesion . Yes
Properties of the gas
Density Yes
Kinematic viscosity No
External impressed pressure Yes
Thermodynamic properties . No
Properties of the missile
Density distribution
Position of CG Yes
Values of moments of inertia, total mass, etc. . . Yes
Size, geometric form, and imperfections in shape Yes
Surface condition
Smoothness . No
Wetting and adhesion characteristics . No
Elastic properties
Local stresses and deformations (e.g. , on nose at impact) No
General flexural deformations (e.g., those due to nose and tail
lift forces) No
Miscellaneous influences
Impressed gravitational field Yes
Wall effects in model tanks . Yes
Initial conditions at water impact
Angle & between trajectory CG and surface . Yes
Aspect of missile at entry values of its pitch (0 ) , roll (º ) , and yaw ( y ) . Yes

( Figure 3.2 ) . The abruptness of deep closure is dummy missile more than 100 diameters from
shown in Figure 3.3, in which a 2-inch-diameter entry. Deep closure and sudden cavity slippage
torpedo model is photographed at 1 / 100-second occur during the launching of the 2-inch
intervals. 1 After closure, the missile continues to diameter torpedo model shown in Figure 3.4 .
travel in the remaining cavity until the cavity Water entry is, then, a series of transient
slips away or is completely entrained by the phenomena occurring after the quasi-steady
flowing water. Remnants may persist about a state airflight.
1A net prevents possible missile injury from contact with the wall.
14
 TABLE 3.2. Dimensions of Quantities Occurring in Water-Entry Problems
( from Reference 61 )

ply System Associated

Property Symbol of Units Dimensionless Number

Impose No Conditions on Scale Factors

a
Density ρ р Gas density coeff,a P model /Pstd
Length ľ e

Time t ly -1

Velocity u V

Angle Dimensionless

Angular velocity elyV

Dynamic fluid pressure

in perfect fluid р pv2
Temperature T

Impose Special Conditions on Scale Factors

Externally impressed
pressure , vapor
pressure , stress
(adhesion,
cohesion) р pv2 Pressure no . po /42pv2
Bulk modulus, Young's
modulus , shear
modulus pv2
Compressional wave
velocity in fluid C V Mach no . v /c
Kinematic viscosity V ly Reynolds no. lv/v
Surface tension т elv2 Weber no. n /p &v2
Flexibility of specimen 1 /n22 v -2 Flexure no. v2 /n22
Impressed gravitational
acceleration l -1,2 v2 /ge
Froude no. v-
an

a Adjusting externally impressed pressure in model system changes density of atmosphere gas ;
1 -to- 1 scaling of density does not hold if same gas is used in model and prototype systems.
To preserve 1 -to- 1 scaling of gas density, a heavier gas must be used in model system .
15
Figure 3.1 . The water -entry cavity .

16
 Surface closure

129 11

Deep closure

Figure 3.2. Water-entry cavities.

17
 >

Figure 3.4 . Cavity behavior : deep closure ,

sudden strippage, and attrition .

properties of the missile, determine both the

inertia of the missile at impact and the orienta
tion of the missile with respect to the water
surface. Missile orientation , of course , deter
mines the specific shape that first contacts the
water.

IMPACT
Impact initiates a short-duration , high
velocity shock wave that radiates at approxi
mately the velocity of sound in the water
( Reference 10 ) . Although the impact force is
very large and may cause water -entry damage, its
short duration (a fraction of a millisecond)
appears to preclude its greatly influencing sub
sequent hydrodynamic flows in the low-velocity
regime . Impact does, however, alter the missile
Figure 3.3 . Sudden occurrence of deep closure.
velocity and produces a splash that may ulti
mately affect cavity closure . Order-of-magnitude
The phenomena of water entry tradi calculations show that the compressibility of the
tionally have been divided into four phases: fluid medium should have little effect on the
1. Impact dynamic behavior of the missile.
2. Flow formation The sequence of photographs in Figure 3.5
.
3

Open cavity (Reference 60) shows the growth of a splash

4. Closed cavity during the vertical entry of an 8.6-inch sphere
To a certain degree, these phases have been entering water at 24.4 fps. ( The raised water
studied separately , although no phase of missile profile in the vicinity of the sphere is made
trajectory is independent of preceding ones. visible by talcum powder on the water surface.)
The velocities of the missile , both linear Figure 3.6 ( Reference 60) shows in silhouette
and angular, established during airflight have the splash contour of aa 4.5-inch-diameter sphere
much to do with the entry phenomena because entering water at 24.4 fps. The splash sheath ,
these velocities , coupled with the physical which may travel four or five times missile speed
18
Figure 3.5a. Growth of a splash during vertical entry of a sphere .

19
1
Figure 3.5b . Growth of a splash during vertical entry of a sphere .

20
Figure 3.5c. Growth of a splash during vertical entry of a sphere .

21
Figure 3.6. Shadowgraph of vertical entry of a sphere .

22
 20
Figure 3.7 . Rise of liquid surface in beaker after impact of sphere . 12
( From Reference 112.)

(Reference 55 ) , leaves the spherical surface suspended in air, leading to the conclusion that
tangentially and each individual water particle the striae were generated by missile vibration
continues to travel in a straight line . The initiated at impact, as previously suggested by
curvature of the sheath does not represent the Birkhoff ( Reference 8).
path of the individual water particle, but rather To estimate the magnitude of the impact
the integrated paths of many. force, it is assumed that the missile is rigid, and
A slight rise in the otherwise undisturbed that during the first instant of contact between
liquid surface beyond the splash , which occurs the missile and the water, the water does not
immediately after impact, has been attributed to have time to flow away and, therefore, acts as if
the entry shock wave ( Figure 3.7 , from Refer it were solid. Then the elastic peak pressure on
ence 112) .2 the missile caused by the elastic deformation of
Another interesting manifestation of the the water is on the order of
impact force was found during the vertical
entries of a right circular cylinder (Refer P = Pw cv sin § = (E Pw ) 1 /2v sin Ğ ( 3.1 )
ence 101 ) . During these launchings , circular
striations normal to the direction of missile To scale the elastic forces, model velocity
motion formed in the surface of the cavity must equal prototype velocity (Mach scaling) if
( Figure 3.8 ) . The striae were generated at the the same liquid is used in both systems . Elastic
same frequency as the natural fundamental forces in the Froude-scaled systems currently in
longitudinal vibration frequency of the missile use are not scaled because they are too small by
2
A facsimile of Worthington's earlier edition published by Longmans Green , London , 1908 .
23
 O

Figure 3.8 . Cavity with circular striations.

the factor of a 1/2 . Failure to scale these forces FLOW FORMATION

has not been detrimental to modeling in the Immediately after impact the water is set
relatively low -velocity range thus far considered into motion by the missile , and a flow -forming
because their effect on the trajectory appears to regime ensues during which fluid pressures on
be negligible, but at higher velocities this will the missile are considerably lower than in the
not be the case. When the elastic forces become impact regime, but considerably greater than in
significant, Froude scaling must be modified by the later quasi-steady-state flow regime . This
or replaced by Mach scaling. flow -formation regime is extremely complex ; a
24
 Figure 3.9. Cavity beginning to form immediately after impact.

complete theoretical description of it is not yet the vertical water entry of a sphere over the
available. The first theoretical treatment of impact , cavity -flow , and cavity -collapse phases
impact upon water was initiated by von Karman using submerged hydrophones (Reference 26) ,
(Reference 86) . Bibliographies of work in hydro and missile water-entry decelerations using a
dynamic impact and water entry are included in telemetry system (Reference 57 ) .
References 52 and 83. Studies associated with or After impact, the water flows away from
conducted at NUC are given in Chapter 8 of the missile and a cavity begins to form ( Fig.
Reference 61 (P.Y. Chou ) and in References 14, ure 3.9 ). The wetted surface of the missile
15, 20, 21 , 50, 58 , and 85. Recently, hydro increases rapidly until a relatively stable line of
dynamic pressure measurements were made of cavity separation is established ( Figure 3.10) .
25
Normal separation Clinging flow

Figure 3.10 . Growth of wetted surface to establish stable line of cavity separation.

26
This separation line may be sharply defined, ir angular pitch velocity (whip ) occurs when the
regular, malformed , or fluctuating ( Figure 3.11 ) , resultant of the hydrodynamic force on the
giving rise to clear, obscured , irregular, or distort missile nose does not pass through the missile
ed cavities, respectively. If the line of sepa CG. Whip also arises from the phenomenon
ration is irregular, the distance of water flow over known as underpressure.
the missile varies, producing small variations in Hydrodynamic whip occurs during oblique
the velocity of the water as it leaves the missile . entry because the lower side of the missile
Small velocity variations cause striations in the contacts the water first. The hydrodynamic
wall of the cavity parallel to the direction of pressure is normal to the wetted surface of the
flow , rendering the cavity translucent or even missile and, unless the contacting surface is a
opaque. When the line of separation is so portion of a sphere centered on the missile CG,
irregular as to be distorted , the velocity variation the resultant force produces a moment about
is so great as to cause large bulges in the cavity the CG (Figure 3.17 from Reference 55 ) . The
wall .
whip from this moment causes a flat head shape
Irregular separation is frequently associated to nose down and a convex head such as a
with erratic trajectory behavior ( Figure 3.12 hemisphere or ogive to nose up. After the nose
from Reference 60) . When the missile is of a of the missile is fully wetted , the pitching
smooth, streamlined configuration, the fluid moment from this hydrodynamic-pressure force
flow nearly conforms to the body shape , and becomes zero as long as the missile is traveling
only a thin wedge-shaped film of gas separates with a zero angle of attack.
the liquid from the solid body (Figure 3.13 ).3 Another hydrodynamic factor influencing
This film is sometimes swept away by the water, missile trajectory, underpressure , can best be
distorting the line of cavity separation (Fig explained in terms of the pressure distribution
ure 3.14 ). A blunt missile shape , on the other about the missile nose. The steady -state pressure
hand, produces large clearances between missile on a completely wetted body is highest at the
and cavity ( Figure 3.15 ) ; its separation line is stagnation point. Farther along the body, the
sharp and clear. pressure drops to a minimum and then recovers
The unstable condition characterized by to a nearly constant value somewhat less than
irregular separation may be influenced by stagnation pressure ( Reference 73 ).
secondary forces ( those arising from liquid Now if the body is partially enveloped in a
surface tension , missile-surface condition , gas cavity, the pressure normally drops from stagna
nuclei in the water, gas viscosity , etc. ) , and tion pressure to cavity pressure at the point of
modeling of this condition is uncertain . It is separation (Reference 73 ) . The shape of the
sometimes possible to trigger a normal separa body and its attitude in the cavity will alter the
tion artificially, by , for example , grooving the pressure distribution, but for small angles of
missile nose at the zone of normal separation , inclination with respect to the cavity , the
but great care must be exercised to assure that character of the pressure distribution is
the cavity-tripping mechanism does not other unchanged ( Reference 72 ) . Under some entry
wise alter the flow . conditions , a nonsymmetric low-pressure area
A large exchange of energy occurs during can thus exist ahead of the actual cavity
the flow-forming phase of entry. Some con separation point.
ception of the energy ultimately imparted to the A low-pressure area on one side of the nose
water can be gained from Figure 3.16 , which and cavity pressure on the other produces a
shows a cavity ( formed by a rifle bullet) piling pressure differential across the nose equal to , at
up on the bottom of the tank, probably due to most, the difference between atmospheric pres
cavity momentum, as discussed in Reference 8 , sure and the vapor pressure of water. A large ,
after the missile was stopped . A change in the long-persisting underpressure area can cause the

3 A mirror inclined at 45 degrees shows the undersides of missile and cavity.

27
 1

Irregular separation, obscured cavity

Sharp separation ,
clear cavity

Malformed separation following minor fluctuations

Figure 3.11 . Cavity separation .

28
Figure 3.1 2a. Erratic trajectory caused by irregular separation .

29
Figure 3.1 2b . Erratic trajectory caused by irregular separation.

30
 Figure 3.13 . Gas film separating liquid from body .

Figure 3.14. Line of separation distorted by flow clinging to streamlined body .

31
 1

Figure 3.15 . Flow sharply separated from blunt missile body.

32
 Figure 3.16. Cavity formed by rifle bullet.

missile to whip so violently that diving or

broaching trajectories occur (Figure 3.18 from
Reference 60). Reducing the atmospheric pres
CG
F ||
X CG
sure reduces the maximum pressure differential
that can occur across the missile nose .

Underpressure-induced whip is modeled when

the atmospheric pressure in the model system is
scaled .
Nose up impulse Nose down impulse
Delayed separation , and thus the possibility
of underpressure effects, will occur whenever
the clearance between missile and cavity wall
becomes so small that the water carries the gas
- CG
in the film away faster than it can be replaced
x11 from the cavity . These narrow clearances are
inherent in the flow about a streamlined body
but they may also occur during oblique launch
ings of less streamlined shapes when the missile
No whip impulse is oriented asymmetrically in its cavity.
The gas flow in such a narrow film will occur at
Figure 3.17. Whip impulse. relatively low Reynolds numbers, increasing the

33
 =

Figure 3.18a. Diving trajectory of torpedo model.

34
Figure 3.18b . Diving trajectory of torpedo model .

35
 T
S

u r
Fo

Figure 3.18c. Diving trajectory of torpedo model .

36
 Figure 3.18d . Diving trajectory of torpedo model.

importance of the viscosity of the gas in Birkhoff's model are shown to correspond very
determining the flow within the film . Although well to observed cavity shapes for spheres of
underpressure areas are more prevalent in small various sizes and materials (References 1 and
missiles , they have also been measured on 22.42 46) . Scale effects in water-entry cavity behavior
inch-diameter prototypes (Chapter 6 ) . are discussed in References 5 , 6 , and 55 .
In addition to Froude scaling, pressure During the third or open -cavity phase of
density or gas-density scaling is necessary for water entry , the cavity remains open to the
proper modeling of whip forces . Modeling of atmosphere at the surface and grows longer as
the underwater trajectory cannot be expected the missile travels along its trajectory. Gas feeds
unless significant whip forces are properly in from above to fill the void behind the missile ,
scaled, because the whip affects missile orienta giving rise to a cavity that is filled partly with
tion at the beginning of the underwater trajec gas and partly with water vapor. Cavity width is
tory . determined by the rate at which energy is
imparted to the water, the wider cavity being
THE OPEN CAVITY associated with the larger energy input (Refer
A complete mathematical description of ence 46) . The growth of the cavity , then, is
transient water -entry cavity behavior is not yet largely determined by the shape and velocity of
available. Birkhoff and Isaacs ( Reference 8 ) dis the missile , by the density and pressure of the
cuss various hydraulic flow models and propose atmospheric gas, and by the position of the
a method for predicting ( from a knowledge of missile in the cavity. The surface tension of the
its mass and cavity drag coefficient alone ) the liquid, the viscosity of the liquid and the gas ,
behavior of a cavity formed under vacuum and the surface condition of the missile are
conditions by any axially symmetric missile ordinarily of secondary importance .
traveling with no yaw . For vertical water entries Several conditions alter the texture
that result in the formation of open cavities (no appearance of the cavity . Turbulence in the
surface seal), the cavity shapes as predicted by boundary layer ahead of cavity separation
37
increases with missile velocity and sometimes and by the physical properties of the missile, the
produces noticeable roughness when the liquid, and the gas. Closure is an event of great
boundary layer has separated to become the importance because it terminates the flow of
cavity wall. Turbulent gas flows within the external gas into the cavity, although it is
cavity produce pressure variations that roughen possible that fluid flow inertia may produce a
the walls. The cavity walls may be distorted by slight enlargement of the cavity after closure .
blowing splash. Re -entrant or closure jets some The conditions determining deep cavity
times strike the cavity wall directly or impinge closure are relatively simple , being primarily the
upon the missile and then splash into the wall, inertia of water and gravity. Momentum
and the simple cavity is often distorted by a imparted by the missile to the water causes the
secondary cavity formed by the missile tail cavity to open . The hydrostatic force upon the
( Figure 3.19 from Reference 111 ). All these cavity and the surface tension of the liquid
cavity disturbances have been observed experi acting within the cavity walls tend to close the
mentally, and are discussed in detail in Part IV. cavity. When the closure forces become
The orientation of the missile in the cavity dominant , the cavity will begin to narrow , form
influences trajectory shape. A straight trajectory a neck, and finally close (Figure 3.23 ). Since the
approximating the continuation of the air trajec surface tension forces are usually small with
tory occurs when the missile travels with only its respect to the hydrostatic (i.e. , gravity ) forces,
nose in contact with the water or oscillates with Froude scaling of deep closure is reasonably
its tail bounding between top and bottom cavity successful.
surfaces (Figure 3.20 ). If the tail remains in The more complex phenomenon of surface
contact with the cavity top or bottom, a curved closure , also shown in Figure 3.23 , has been
trajectory (either diving or upturning) results observed for many years without being com
( Figure 3.21 from Reference 60 ). A fine-nosed pletely explained. The force system at the cavity
missile with large underpressure area can dive mouth arises from both water and gas flows. A
with its tail riding in the cavity bottom large hydrodynamic force can occur in the
(Chapter 8) . If the missile is unstable, it may cavity lip where the masses of water lie . The
broadside ( Figure 3.22 ) . hydrodynamic forces in the splash are small
If the resultant of the forces determining because little water is involved. However, the
the orientation of the missile in the cavity is splash cannot be ignored because it can modify
small , minor differences in ambient or launching the size of the cavity orifice , thereby signif
conditions may change the orientation and alter icantly influencing the gas flow into the cavity.
the missile trajectory . It is in these instances that Thus, three separate flows – cavity -lip , splash,
-

and gas
the art of modeling is most uncertain. interact to produce cavity closure.
The presence of the cavity makes it diffi Empirical theory regarding cavity closure and
cult to measure missile position . If the cavity closed cavity behavior is included in
wall is clear, the interface acts as a reducing lens . Reference 8.
The distortion of the missile image varies with The major hydrodynamic flow in the cavity
both time and space because the curvature of lip is at first directed outward as the water flows
the cavity wall is not constant. If the cavity wall away from the impact. Then the surface tension
is rough , the missile image is blurred, or even the liquid , together with the hydrostatic
obscured . forces , will act to restore the disturbed surface
The open -cavity phase of water entry con to its normal condition . The pressure forces
tinues until the disturbed water comes together created by the flow of the gas rushing through
and seals the cavity from the atmosphere . the cavity mouth produce a region of low
pressure on the underside of the cavity lip . The
CAVITY CLOSURE higher atmospheric pressure acting from above
Cavity closure occurs either at the surface, causes the cavity to close.
some point below, or both. The nature of the The pressure produced by the gas flow is
closure is determined by launching conditions proportional to the atmospheric density. Thus ,
4The resultant force from these flows determines whether or not the cavity will form . If the closure force is dominant
at impact , the flow merely slips around the missile and no cavity occurs .

38
 a

Air pressure 1 atmosphere

b
Air pressure 1/6 atmosphere

Figure 3.19 . Secondary cavity formed when missile tail contacts cavity wall.

39
 Only missile nose contacting water

Tail oscillating

Figure 3.20 . Straight water-entry trajectories.

40
 S
W

129 g

Figure 3.21 . Examples of diving and up-turning trajectories when tail contacts cavity top or bottom.

41
 Figure 3.22. Missile broadsiding after water entry .
۱
۲
_

‫با‬

Figure 3.23 . Deep and surface cavity closure.

42
gas pressures sufficiently low to cause cavity
closure are not produced on the side walls when
the atmospheric density is low. Then the cavity
mouth continues to widen until the hydrostatic
and surface tension forces restore the surface to
its normal rest condition (Figure 3.24) . Under
these circumstances, cavity growth is terminated
by deep closure.
An indication of the pressures and forces
created by the gas flow can be gained from
Figure 3.25 ( Reference 111 ), which compares
the surface contours generated at two different
gas pressures during two otherwise equivalent
launchings of a hemisphere-head torpedo model
with a flared cone tail. The 2 -inch-diameter
missile was 15 inches long with a trajectory
water - entry angle of about 19 degrees and an
entry velocity of about 120 fps. At the lower
pressure, the cavity surface retains its original Figure 3.24 . Termination of cavity growth by deep closure
(right to left).
contour and grows rapidly at first and then more
Horizontal lines denote undisturbed water surface
Vertical line denotes the point of entry

0.002

0.02

0.04

0.06

0.08

0.12
entry
Time

0.17
s,from
ec

0.19

0.21

0.29 Model

0.52

0.72

Air pressure 1 atmosphere , Ve = 120 fps, š = -18.7 deg

Figure 3.25a. Water -entry cavity contours.

43
slowly until about 0.2 second after water entry . At some height, the sheath closes completely to
Then, presumably under the influence of gravity form a dome . Above , the surface-tension forces
and surface tension , the water moves in to fill shape the remaining ring of water into a circle of
the void. At the higher pressure , the motion of droplets. The closure dome is a fragile thing that
the surface is more complex. The contour begins may be destroyed by gas flow into the cavity or
to deviate from the original less than may be blown apart by the jet formed by
0.06 second after entry. The general motion of closure of the cavity lips . The irregular surface
the cavity wall is inward but a large rapidly of the splash sheath following vertical entry
growing protuberance forms beneath the cavity indicates that the splash is of variable thickness
mouth and continues to elongate . This protuber ( Figure 3.27 ).
ance is probably caused by a closure jet, by air An irregular splash spoils the symmetry of
and entrained water rushing into the cavity , or surface closure and usually tends to delay it.
by a combination of the two effects. Since water in the splash is drawn from the
The sequence of photographs in Fig surface, disturbance of the surface before entry
ure 3.26 shows how the cavity mouth changes as is likely to produce irregularity.
the splash and cavity grow following the vertical Oblique entry inherently produces an !

entry of a sphere . Above the surface, a ring of asymmetric splash, as can be seen in Figure 3.28
splash rises rapidly. With the passage of time , the from Reference 60 and also in Figure 3.29 . In
splash falls back, thickens toward the base , and general, surface closure follows deep closure
the surface - tension forces act to effect closure. during such launchings.

Horizontal lines denote undisturbed water surface

Vertical line denotes the point of entry
0.002
0.02

0.04

0.06

0.08

0.12
entry
Time
s,from
ec

0.17

0.19

0.21

0.29 Model

0.52

♡
0.72 to
mo

Air pressure 1/6 atmosphere , Ve = 122 fps , g = -19.0 deg

Figure 3.25b . Water-entry cavity contours.

44
 S

Figure 3.26. Cavity and splash growth after vertical entry of sphere.

Numerous forces influence surface closure Fortunately, the cavity lip is usually
and make it difficult to model. Among these dominant in determining surface closure . Its
forces are surface tension , the viscosity of the behavior is largely influenced by the hydro
liquid, the viscosity of the atmospheric gas, the dynamic force in the main body of the liquid
adhesive forces between the missile and the and the aerodynamic pressure force in the gas
liquid (determined by the surface condition of flow . Both forces are modeled by extending the
the missile ), gas nuclei present in the liquid , and Froude scaling laws to include gas-pressure and
cleanliness of the liquid surface.5 gas-density scaling. 6
5Closures are also discussed in Chapters 9 and 16.
6lf the resultant of these two forces happens to be nearly zero, the minor forces in the splash and in the cavity lip will
control cavity formation. Failure to scale these may cause the model missile to enter fully wetted under conditions
that produce a cavity-running prototype.

45
 soort

Figure 3.27 . Irregularity of splash sheath thickness following vertical entry of sphere.

Jets of water, directed either up or down, powerful in the absence of surface closure, and
are produced by cavity closure . Data indicate becomes less powerful the earlier surface closure
that the up -jets formed at deep closure are occurs. The down -jets following deep closure
vertical or nearly so , even for non -vertical can be strong; they can deform the cavity wall
entries. According to Reference 22 , this is true or strike the missile, deflecting its trajectory ,
because the up -jet is caused by water masses and have been known to knock the missile from
colliding at deep closure as the result of the the cavity. In less spectacular instances, the
restoring action of hydrostatic pressure. For this impinging of the jet upon the missile has
reason , up -jets have no opportunity to form at decreased the apparent coefficient of drag.
oblique , low-angle entries. Down -jets also occur
at both surface and deep cavity closure . When THE CLOSED CAVITY
the water- entry angle is less than about The final transient effects of water entry
45 degrees, the down -jet at surface closure occur during the closed-cavity phase , when the
creates a dipole field that travels downward missile- cavity system is no longer in direct
away from the cavity. When the entry angle is contact with the atmosphere . The cavity slowly
greater than 45 degrees, the down -jet is diminishes until it is entirely dispersed by the
apparently dissipated in the cavity wall entraining action of the water (Figure 3.30 from
( Reference 8 ) . Reference 60 ) , or until it slips from the missile
Closure jets usually are stronger than the ( Figure 3.31 ). Sudden asymmetric shedding
re-entrant jets formed by closed cavities. As sometimes causes gross changes in the direction
pointed out in Reference 22 , the up -jet has been of missile motion (Figure 3.32 ) . With the
observed to have a velocity at deep closure that disappearance of the cavity, the fourth and final
is higher than that of the missile. It is most phase of water entry is terminated and the fully
46
 Figure 3.28a. Surface closure delayed by irregular splash. (Top photographs are double exposures.)

wetted or purely cavitating flows normally On rare occasions, the fourth phase of
encountered in hydrodynamics occur. If gaseous water entry does not occur. The missile may
exhausts from an internal power plant are broach while it is still enveloped in a cavity or
present , the missile can travel under quasi the open cavity may strip suddenly from the
steady-state conditions in a cavity artificially missile soon after it is formed, causing a marked
maintained by the exhaust gases. change in the missile trajectory.
47
Figure 3.28b . Surface closure delayed by irregular splash .

48
Figure 3.29 . Rear view of cavity produced by oblique water entry.

Figure 3.30 . Cavity slowly dispersed by entraining action of water.

49
 Figure 3.31 . Cavity slipping from missile .

Figure 3.32 . Change in missile motion caused by sudden

asymmetric shedding of cavity .

50
 CHAPTER 4
EVOLUTION OF EXPERIMENTAL HYDROBALLISTICS MODELING

Experimental hydroballistic modeling and The experiments made with liquid drops
theoretical modeling are interdependent ; they (water or milk ) into milk or water enabled him
have grown together in a climate of mutual to watch the growth of the splash from the
support. Theoretical studies provide guidelines surface of the liquid and, at higher entry
for experiment; discrepancies arising in the velocities , to observe that a cavity was also
resulting data often lead to refinement of the formed. The motion of the liquid originally
theory. within the falling drop and the contours of the
Investigations of flow and cavity forma- splash revealed much information on the flow
tions at water entry began in the 19th century , pattern and forces involved. Worthington's
but the behavior of a missile during water entry explanation of splash formation has remained
was no more than a scientific curiosity until largely unchanged for nearly 85 years.
malfunction of World War II torpedoes indicated His work with solid spheres was chiefly
a deficiency in design information . Research concerned with splash formation . He inves
received an impetus, and water -entry modeling tigated the effects of entry velocity and the
technique grew slowly from study of the sphere's surface condition on the splash and, by
physical factors involved. using an oil and various glycerine mixtures
The opposite phenomenon , water exit , was instead of water, learned something of the
discussed and a few experiments were performed effects of liquid density , surface tension , and
at the Ordnance Research Laboratory, State viscosity upon the entry splash. Within the range
College, Pa. , about 1950. Decisions to develop of his experiments, he detected no difference
an underwater-launched missile created an between splashes formed in air and those formed
urgent accompanying need for information on in vacuum .
the phenomena of water exit. Preliminary Worthington observed important general
studies showed water exit to be not simply characteristics of water entry such as the small ,
water entry in reverse but a process beset with rapid rise of the liquid surface at impact . He was
problems uniquely its own (Chapter 11 ) . able to explain the formation of closure jets
Modern water- entry research began with from the flow about the cavity . Cavity flow was
A. M. Worthington in the 1890's. His detailed rendered visible in his photographs by streams of
photographs of spheres entering the water small bubbles from electrolysis. So far as is
vertically (Reference 113 ) were unequaled for known , his photographs of a jet penetrating a
40 years; his experimental setup was beautiful in closed splash dome ( Figure 4.1 ) have yet to be
its simplicity . His work was first published in duplicated in the modern laboratory.
book form in 1908 and a facsimile was pub- Worthington's experience with the water
lished in 1963 (Reference 112 ). entry of a smooth sphere typifies the per
Worthington's contribution to water-entry plexities so frequently encountered in modeling.
studies included observation of liquid drops and For his series of drops , all test conditions were
solid spheres entering still liquids under various identical except for the two handkerchiefs used
conditions, and analysis of the experimental to wipe the sphere between tests. After having
phenomena in terms of the physical forces been wiped with one handkerchief, the sphere
manifested . consistently entered with a separating splash ; use
51
 Figure 4.1 . Closure jet penetrating splash dome. ( From Reference 112. )

of the other handkerchief resulted in a clinging and found that the flow pattern at the time of
sheath-type splash . Yet microscopic examination deep closure was in reasonable agreement with
showed the two pieces of cloth to be “ extremely the steady-state flow pattern predicted by
similar.” As modern investigators have also potential theory . He then substituted zinc
learned (Reference 43 and Chapter 8 ) , even chloride solution (sp. gr. = 1.75 ) , petrol (sp .
=
minute variations can cause marked differences =
gr. = 0.71 ), and glycerine (sp. gr. = 1.26) for
in modeling results. water; the flow pattern was independent of the
Interest in the water entry of projectiles liquid density in this range. Deep closure ,
was limited until 1942 , when E. G. Richardson however, occurred significantly later in the
of the Royal Aircraft Establishment in England highly viscous glycerine.
began the study of water-entry cavity formation Richardson measured transient cavity pres
( Reference 68 ) by dropping spheres ranging in sures and related them to cavity closure (one of
diameter from 1/5 inch to 1 inch from heights the few such records that have been made ) . A
up to 126 feet above the water. Richardson steel disk 0.004 inch thick and 1.5 inches in
introduced a fine network of bubbles to delin diameter, mounted flush with the flat base of a
eate the flow about the missile and the cavity , hollow steel missile, transmitted cavity pressure
52
to the stylus of a scratch recorder carried within entry are influenced by change in the air
the missile. Another stylus inscribed records of pressure above the liquid surface . Davies was the
the ambient hydrostatic pressure and the time . first to show experimentally the importance of
The sensitivity and response of this recorder gas density in surface closure (Reference 55 ,
were carefully determined to make certain that page 170) .
the cavity pressure records were valid. Results In the late 1940's, David Gilbarg and
showed that cavity pressure was essentially Robert Anderson of the Naval Ordnance Labora
atmospheric until surface closure ; it then oscil tory , White Oak, reported a high-speed motion
lated sharply before asymptotically approaching picture study of the effects of atmospheric
the ambient hydrostatic pressure (Figure 4.2 ) . pressure upon the splash and entry cavity of
Richardson also measured drag coefficients spheres (Reference 22 ). Cavity closure at the
of spheres entering water at 1 and 1/5 atmos surface was clearly a factor in determining cavity
phere and entering other liquids ( Table 4.1 ) , and formation . These investigators suggested that
he measured the water-entry impact force and atmospheric density rather than atmospheric
the added mass of a number of shapes pressure largely determined the surface closure.
( Table 4.2 ). He noted that the increased added Vertical jets were observed at the sites of both
mass obtained at higher entry speeds was surface and deep closures, and velocities as high
reflected in the amount of water that was able as 1.5 times projectile velocity were measured in
to “get away ” after impact in the splash . a jet arising from deep closure.
Table 4.3 summarizes the results of his ricochet Gilbarg and Anderson evaluated the
studies, which indicate that the critical angle of Froude-scaling technique in entry cavity and
ricochet was independent of velocity in the splash modeling. Failure to model the cavity was
range studied , but that it varied with missile attributed to lack of gas-density scaling. Splash
density . studies , made both with and without adjusting
In the early 1940's , R. M. Davies showed the cavitation number in the model systems to
that the cavity and splash accompanying water equal that of the prototype, showed that the
pPressur e

2
, si

0
0.1
AM Time , sec
1
0.2

5
pPressure

4
, si

Cavity pressure
Local hydrostatic pressure

0.4 0.6 0.8 1.0

Time, sec

Figure 4.2 . Cavity pressure as a function of time. ( From Reference 68. )

1The importance of surface closure in cavity formation was first pointed out by B. D. Blackwell (Reference 55 ,
page 170 ) .
53
 TABLE 4.1 . Drag Coefficients of Spheres Entering Liquids
( from Reference 68 )

Sphere Ср
Mean Fully
Medium Diameter , cm Sp. Gr. Reynolds No. In Cavity Immersed

Steel in water 1 to 2.5 8 3 X 105 0.30 0.4

Steel in water 1 to 2.5 8 3 X 105 0.30 0.4

( low pressure)
Lead in water 7 12 106 0.28 0.4

Steel in zinc 1 to 2.5 4.5 3 X 104 0.30 0.4

chloride
( sp. gr. 1.75 )
Steel in petrol 1 to 2.5 +1 3 X 105 0.34 0.4
(sp. gr. 0.71 )
Steel in glycerine 1 to 2.5 6.5 3 X 102 0.50 0.7
( sp . gr. 1.26 )
Steel in watera 1 to 1 8 5 X 103 0.32 0.5

a
Reference cited: Ramsauer, C. ( and Dobke, G. ) , 1927. (Ann. Phys. Lpz. , 84, 697. )

splash from the smaller spheres, scaled up to modeled over this low-velocity range , the model
prototype size in accordance with the Froude ing failures possibly being caused by lack of
law , decreased in height with model size . When a viscosity or surface-tension scaling.
Freon-air mixture was used to maintain the gas E. N. Harvey, J.H. McMillen , and their
density at the lower pressures, the splash height colleagues at Princeton University studied small
still did not scale, suggesting that surface tension steel spheres entering water at speeds ranging
was important in determining the size and shape from 800 to 3000 fps. They found that the
of the splash . entry cavities formed at high velocities
The experimental studies begun by Gilbarg resembled those occurring at low velocities. In
and Anderson were continued at NOL under the addition , elastic shock waves formed by a sphere
direction of Dr. Albert May . Cavity formation entering water at velocities much below sonic
and closures, hydrodynamic drag, and virtual (with respect to water) were visible in photo
mass effects were extensively investigated (Ref graphs by Harvey ( Reference 55 , page 160) .
erences 43 , 44, 46 , 47 , and 48). Dr. May's data Later McMillen and Albert May extended these
pertain to the cavity formed during the vertical studies at NOL to small steel spheres entering
entry of spheres ranging from 1/4 to water vertically at supersonic velocities of about
1 1/2 inches in diameter. Entry velocities were 7000 fps. Pressure oscillations within high-speed
between 26 and about 80 fps. He assessed the cavities caused noticeable pulsations in cavity
cavity growth and closures in terms of both the size ; they became more marked with increase in
physical phenomena and Froude scaling. Cavity entry velocity (Reference 55 , page 170 ) , in
>

growth and closures are not satisfactorily agreement with the oscillation in cavity pressure
54
 TABLE 4.2 . Water-Entry Impact and Added Mass Measurements
of 3-Inch-Diameter Missiles ( from Reference 68 )

Entry Added
Mass, Specific | Angle, a
Mass, a at à at
Head gm Gravity deg vo/v1 gm 40 m/secb 4 m/secb
Hemisphere 725 0.73 90 1.2 145 2.7 1.8

64 1.25 180 3.4

35 1.3 220 4.1

Disk 620 0.57 90 1.5 310 6.0 4.0

•••
64 1.4 250 4.7

35 1.3 185 3.6

22 1 /2-deg semi-angle cone 620 0.57 90 1.15 95 1.8 1.5

64 1.15 95 1.8

35 1.2 125 2.3

45-deg semi-angle cone 550 0.60 90 1.3 165 3.2 3.0

64 1.2 110 2.0

35 1.15 85 1.6

1 : 1 ogive 710 0.75 90 1.0 0 0.0 0.0

64 1.05 35 0.6

:
35 1.1 70 1.30
:
a ==

V., V1 = velocities just before and just after impact, respectively.

b
= m/c3 = an added mass parameter , where m (the added mass) is expressed in grams, and c
( the model radius) in centimeters.
с
This result appears anomalous.

measured by Richardson at lower velocities and Caywood (Reference 7 ) improved the

( References 8 and 68 ) . bubble-marking technique and determined
Garrett Birkhoff (Reference 6) and numerical velocity distributions for several flow
Birkhoff and Zarantonello (Reference 9 ) used fields. Of particular interest are their photo
experimental evidence to substantiate theoret graphs showing a dipole moving down from the
ical predictions, providing sound guidelines for point of surface seal and the flow accompanying
exploration . They refined theory to keep pace the up -jet that forms at the surface at the time
with advancing experimental evidence. Birkhoff of deep closure ( Figure 4.3 ) .
55
 TABLE 4.3 . Ricochet Studies
( from Reference 68 )
Velocity in Angle to Water
Path, ft/ sec Surface, deg Path Max. Mean Mean
a а b Length , Depth, с с
V
via Bo' Bib
1 in. diam . CD CLC
3-in. Dural Sphere, sp. gr. 2.7
100 50 7.5 4.5 50 0.50 0.15 0.045
120 33 10.0 6.0 60 0.50 0.23 0.055
100 53 9.5 4.0 35 0.75 0.20 0.076
130 65 7.5 6.0 55 0.75 0.14 0.042
100 65 7.0 5.0 42 0.11 0.052
147 52 9.0 10.5 84 0.75 0.13 0.048
120 30 5.5 4.5 120 1.00 0.12 0.015
100 25 6.0 9.0 90 1.00 0.17 0.028
135 67 8.5 0.0 60 3.50 0.12 0.027

1 -in. Steel Sphere, sp. gr. 7.8

205 55 7.0 0.0 72 2.00 0.19 0.013
210 140 8.0 4.0 66 1.00 0.06 0.033
310 140 7.5 4.0 60 0.50 0.14 0.035
30

330 140 6.5 5.0 0.75 0.30 0.062

150 50 11.5 54 0.21
270 80 6.5 0.0 100 1.80 0.13 0.012
300 65 6.0 72 1.50 0.22
360 55 5.0 0.0 76 1.50 0.25 0.012

3-in. Ebonite Sphere, sp. gr. 1.1

135 35 12.5 10.0 160 1.25 0.04 0.011
155 58 17.0 11.0 55 1.25 0.08 0.036
155 37 14.0 14.0 48 1.50 0.13 0.043
142 56 11.5 3.5 42 0.70 0.10 0.027
135 16 6.0 6.0 54 0.80 0.17 0.018
•

...

142 12.0
140 12.5
::
::

a =

O entry velocity ; V1 = exit velocity.

b ßo = entry angle; B1 = exit angle.
Bo
с
CD and Cl = mean drag and lift coefficients over the path
56
 Surface closure

Figure 4.3. Flow patterns at closure. ( From Reference 7.)

57
 The experiments of Mason and Slichter inability of the air in the cavity to maintain the
( References 60 and 61 ) , characterized by the gas film between the model and the cavity wall
type of ingenuity that marked Worthington's (i.e. , to maintain the line of normal cavity
work, were in two categories : studies of the separation ) , and ( 2 ) that failure to maintain
water entry of air-launched missiles, and special normal separation caused the model to dive .
studies (also using missile shapes) of the water The existence of the low-pressure area was
entry phenomena. The latter group of experi proved by the following experiment Slichter
ments comprise a classic example of identifica made with a 2-inch-diameter model of a torpedo
tion and evaluation of the physical forces launched at an entry velocity of 117 fps.
influencing a phenomenon .
Slichter listed the forces that might con “ A 0.2-in. diameter detachable disk (weight
ceivably be important (Chapter 3 ) and devised a 40 mg) was inserted near the separation
series of experiments to evaluate them. He did area, flush with the surface of the model.
not fall into the obvious trap of describing a This disk was held in place by evacuating
force as being a magnitude or more smaller than the model. If the disk came off during the
others and, therefore, insignificant. trial run , the easily measured internal pres
Slichter's work was given impetus by World sure was taken as greater than the external
War II. The Froude -scaled model of a standard pressure on the disk , and less if the disk
U.S. aircraft torpedo ( Figure 4.4, top ) repro stayed on. In ten tries with an internal
duced the trajectory of the prototype to a pressure of 0.02 atmosphere ( absolute ), the
usable degree of accuracy ; but with a slightly disk came off eight times and stayed on
finer head ( Figure 4.4, bottom), the prototype twice . With various higher internal pres
broached and the small ( 1- and 2-inch diameter) sures the disk invariably came off.” See
models dived with violence . When the water in Figure 4.6 from Reference 61 .
the model tank was of sufficient depth , the
model actually turned through 180 degrees to The tests were repeated with a light helical
head in a direction up -range instead of down spring holding the disk in place , and sub
( Figure 4.5 ) . The underpressure phenomena stantially the same results were obtained .
(Chapter 3 ) that caused the model with the finer
head to dive were first identified through “ It wa
wass concluded that the absolute
Slichter's experiments , which proved ( 1 ) that a external pressure on this area almost com
low-pressure region was present, caused by the pletely vanishes during entry.”

NCHES

Hemisphere ( standard) head

SPORTS

Finer head

Figure 4.4 . Torpedo models.

58
Figure 4.5 . Diving trajectory , 2 - inch -diameter model with finer head .

Figure 4.6 . Separation of pressure disk during launching.

59
 The cause of the low-pressure area was was cold enough to condense the water vapor in
established by a series of tests with “ vented” the film as fast as it formed. This approach was
models. Small, symmetrically spaced holes were abandoned without trying a heated model2
drilled in the nose of the model , saltshaker because the Controlled Atmosphere Launching
fashion ( Figure 4.7 ) , to allow air from the cavity Tank at the California Institute of Technology
to flow through the model and into the narrow would soon be available.
space between the cavity wall and the model, Slichter investigated the underpressure phe
thus preventing a low-pressure area. The differ nomena further with the finer head model
ences between the lines of separation and the and established relationships between separation
cavity contours of the models with and without angle, gas film pressure and thickness, and under
venting are shown in Figures 4.8 and 4.9 . The pressure forces. He recognized the importance
crease in the bottom of the cavity caused by the of gas viscosity , liquid surface tension, model sur
flow clinging to the bottom of the unvented face finish , and the adhesion between liquid and
model as the cavity was generated is noticeable missile in determining separation when the sepa
in Figure 4.9 . The vented model torpedo did not ration angle was small; when the entry velocity
dive (Figure 4.10) . was also small , he recognized that liquid viscosity
The importance of maintaining a gas film might influence separation. Slichter tried to eval
between the model and the cavity wall was uate the importance of surface tension by adding
established by showing that when cavity separa 0.01 % Aerosol OT to the water to reduce
tion was artificially forced with rings of wire the surface tension to one-third of its origi
( 0.005 and 0.010 inch in diameter) or with nal value. He notes only that change in surface
cellophane tape ( 0.003-inch thick ) at the line of tension did not alter the trajectory of the hemi
normal cavity separation, the model with the sphere head torpedo model.
finer head followed a normal upturning trajec Time and funds did not allow Slichter to
tory. investigate thoroughly all the forces involved, but
It was apparent that the underpressure he examined the various elastic and compres
force would be greatly reduced by scaling the sibility forces in the system . As before, his test
cavitation number. Launching facilities capable vehicle was a 1 -inch-diameter finer head torpedo
of operating at reduced atmospheric pressure model. The compressibility of the water was al
were not then available; Slichter attempted to tered by introducing a carefully spaced network
maintain the gas pressure beneath the model by of fine bubbles ( 3 -mm diameter) into the water.
raising the water temperature to 99 °C , thereby The air content (by volume ) was approximately
increasing the vapor pressure to nearly 1 atmos 1 part per thousand , reducing the bulk modulus
phere. The expected results did not ensue, and of thewater by about 1 / 16.3 Slichter noted only
the model dived, probably because the model that the diving trajectory of the model was not

Figure 4.7 . Vented model with finer head.

2Worthington experimented with smooth heated spheres and noted that flow separation was inhibited by heating. He
concluded that the hot missile surface was more dust -free because dust does not settle upon surfaces heated above
room temperature (Reference 112 ) .
3
E
Ē
1 - q + qE /YP.
=
where E = average effective bulk modulus of the mixture , E = bulk modulus of water, q = relative air content by
volume, y = a constant varying between 1.0 and 1.4 ( depending upon the degree of heat interchange during
compression of the air ), and p . = atmospheric pressure .

60
 Figure 4.8 . Normal cavity generated by vented model .

altered by this modification or by changing the covered by a layer of acetone “ several inches
elasticity of the model with a nose made of a thick ” . Although several other properties in
1 / 16-inch-thick layer of soft rubber supported addition to the elasticity were changed by this
by a metal frame. He felt that further investiga procedure,44 the trajectory was not noticeably
tion of the elasticity of the missile and of the affected .
effects of a frangible nose cap might prove Slichter established the importance of mis
valuable in solving missile- entry problems . sile flexure by showing that the trajectory of the
Slichter also tried altering the elasticity of hemisphere head model was markedly altered by
the liquid by launching the model into water deliberately offsetting the afterbody and shroud

4some properties of acetone with respect to those of water are as follows : surface tension 1/3 , compressibility 2.1 ,
kinematic viscosity 0.38 , density 0.79 , and velocity of sound 0.81 .
61
 Figure 4.9 . Crease along bottom, distorting cavity of unvented model .

ring by only 1.5 degrees. He demonstrated a sphere and several other mathematically trac
that the oblique trajectory of aa torpedo generally table shapes upon the water surface
turns upward as missile density is decreased. A ( Figures 4.11 through 4.14) .
similar effect was obtained when missile density After Slichter laid the foundations of
was held constant and entry velocity increased. understanding and demonstrated several
In Slichter's experiments with spheres, he techniques , experimenters under the direction of
measured the rotation of an obliquely entering R. T. Knapp at the Controlled Atmosphere
sphere to see whether the shear force at impact Launching Tank at CIT noted that the trajectory
would perturb a small missile ; he concluded that modeling of a fine-nosed shape was greatly
the shear force was insignificant. Using bubbles improved by reducing atmospheric pressure and,
to mark the flow field , he studied the impact of furthermore , that the best modeling was
62
 Figure 4.10. Nondividing vented model.

obtained at the atmospheric pressure that scaled were obtained on air and underwater trajec
the cavitation number one - to - one tories, entry whip , and the time of first contact
(References 37 and 38 ).5 between missile tail and cavity. The missile
Another interesting study at the CIT tank attitude during underwater flight was also
(Reference 36 ) showed the effects of variation recorded for a few launchings. Some underwater
in Froude and cavitation numbers upon the drag data were taken, and the effect of missile
behavior of a 2-inch-diameter torpedo model. shape was studied by successively varying nose
The data showed the trajectory to be very and tail contours , shroud ring, fin , propeller
sensitive to change in cavitation number ; Froude design and placement, and the overall length of
number influence was more complex. Some the missile (References 56 and 59 ).
times the trajectory was almost independent of Part III of this book presents more recent
Froude number; at other times aa small change in studies. The water-entry phenomena from initial
Froude number would change a diving trajectory whip through cavity attrition have been
into a broach. In general, sensitivity to Froude inspected in detail, and many of the secondary
number was noted when a small resultant force forces originally listed by Slichter have lent
determined model orientation in the cavity. interest and an aura of the unexpected to the
Concurrently with the early model studies , experimental results. Preliminary studies of
prototype missile launchings were made at the water exit and earth entry are also included.
NUC Morris Dam Torpedo Range (Part V) . Data

5Extensive studies of water entry were made at the Alden Hydraulic Laboratory (now Alden Research Laboratories ),
Worcester Polytechnic Institute , Worcester, Mass . , and by the British Admiralty at its laboratories in Teddington ,
Middlesex , U.K.

63
.

| |

*
*
*
*
*
||||| \\\ \
|| ||

********

11
I
\

Figure 4.11 . Flow field immediately after impact of 8.6-inch-diameter sphere.

64
‫نهبه‬

Figure 4.12 . Flow field established by vertical entry of 90 -degree cone.

65
Figure 4.13. Flow field established by vertical entry of 2 -inch
diameter sphere.

66
 !!!

Figure 4.14. Flow field established by oblique entry of model.

67
 PART III .
PROTOTYPE AND MODELING EXPERIMENTS

CHAPTER 5
FULL-SCALE WHIP AND DECELERATION AT WATER ENTRY

The value of modeling lies in the assistance determining the underwater trajectory of a
it gives to the prediction of full-scale missile missile. Whip influences the position of the
behavior . Various physical influences encount missile in the water-entry cavity , which in turn
ered when a torpedo (prototype or model) hits largely determines whether the underwater tra
the water require scrutiny, measurement, and jectory will be upturning, downturning, or
analysis in order that they may be applied to the straight . Deceleration of the missile in the cavity
modeling process and so increase its worth to not only influences the underwater trajectory
the engineering development of underwater and angular momentum but may be so severe as
missiles . to cause damage to the shell and to the missile's
Whip and deceleration depend upon the internal mechanism .
water-entry conditions and upon the missile Two separate mechanisms contribute to
shape and dynamic parameters. The studies whip during oblique entry . The more important
discussed here ( first reported in References one results from the lower part of the nose being
91 , 97 , and 98 ) are concerned with the effect of wetted first , giving rise to unbalanced forces that
head shape on the whip and water-entry decelera may be quite large. The forces come not only
tion of full - scale missiles . The program was from fluid pressure but also from reduced
planned primarily to obtain data that could be pressure in the nose undercavity (underpressure
directly applied to service-missile design, and it effect ). The torque thus applied about the CG
was also anticipated that it could be used later produces the change in pitch velocity known as
to investigate whip modeling . whip. The other whip -producing mechanism
The ultimate trajectory of a missile under occurs when the missile enters the water with an
going water entry is markedly influenced by its attack angle , a. Then the drag forces producing
behavior in the cavity stage. Whip , A , the the change in the axial velocity contribute a
change in missile pitch velocity while the tip of whip moment about the CG.
the nose is being wetted , may be attributed to The whip is linearly dependent on the
the high shock at water impact and to the forces water -entry velocity , v, because the forces on
associated with the establishment of flow . It is the head are proportional to v2 , and the time
probably the most important single factor in during which these forces are unbalanced (acting
69
on the lower side of the head only ) varies as 1 /v. 57° 70°
The impulse thus will vary as v. For spherically
tipped heads (assuming small attack angles) , Ł
there should be a linear dependence of whip
upon angle of attack. Calculation of the whip Head a Head b Head c
requires a knowledge of the instantaneous pres
sure distribution over the missile nose during R /R70 R/Rp = 0.615 R /Rp = 0.872
water entry ; this, at present, is unavailable. ( 70.25
After water impact and the establishment 1.5-diam ogive 1.5-diam ogive 1.5-diam ogive
57 ° half-sphere 70 ° half-sphere
of flow , the missile ordinarily travels several angle angle
lengths beneath the water surface with only its
nose in contact with the water. During this time , 65 ° 70°
the drag force does not usually act through the *
CG ; the moment, depending upon the head
configuration and the amount of angle of attack,
is directed toward either increasing or decreasing Head Head e Head f Head g
the pitch angle of the missile . The angular d

acceleration produced by the drag moment thus R/ Rp 0 R /Rp = 0.379 R /Rp = 0.615 R /Rp = 1
either adds to or subtracts from the whip . Q =0.125 Q=0.5
3.5 - diam 3.5 - diam 3.5 - diam Hemisphere
To implement the investigation of whip ogive ogive ogive
and deceleration at water entry , 14 full -scale 65-deg 70-deg
heads (Figure 5.1 ) were attached to an aircraft half-sphere half-sphere
torpedo afterbody. The geometry of the ogive, angle angle
spherogive, and plate-ogive heads is shown in
Figure 5.2 . The ogival portion of the sphero
give and the plate ogive heads is designated by
expressing the radius of the generating curve in Head h Head i Head j Head k
projectile diameters. The spherical portion is
expressed in terms of the half-sphere angle, i.e. ,
the angle subtended by the generating curve of
the spherical nose measured from the tip of the
nose to the point of junction with the ogival
curve . R/Rp = Rp = 6.50
R /Ry R /Rp = 00 R /Rp = 0.576
In the nomenclature used to designate the
0.5 - diam 0.5 - diam spher 0.5- diam 0.576- diam
plate -ogive heads , the flat-plate radius , P, is given plate on ical segment plate on sphere on
by 8.4-deg ( center at CG ) 8.4 -deg 8.4-deg
cone on 8.4 -deg notched cone
cone cone
P R.o
= 0.5 -
=
(5.1 )
2R 2RP
р

Headl Head m Head n

where Ro is the ogival radius and Rp is the
projectile radius. Thus, the plate 0.1 -diameter
ogive head has a plate radius of 0.4 diameter.
The tolerances on the ogive, spherogive, and
plate-ogive head dimensions were held to +0.001
OO
diameter of the nominal values . R /Rp R /Rp = R /Rp
Heads h , i , j , and k (shown in Figure 5.1 ) Q=0.667 Q=0.833 Q= 1.00
are modifications of a cone of 8.4 + 0.05 -degree
Plate 0.25 Plate 0.1 Plate cylinder
generating angle. Head h is a truncated cone diam ogive diam ogive
with a plate of 0.500 +0.002 diameter. The flat
plate may be regarded as a segment of a sphere Figure 5.1 . Full- scale head shapes used in underwater
of infinite radius . Head i differs from Head h in trajectory studies.

70
 2.0
APR
Ro E
o 5.0 -diam spherogive
с. 1.9
R D A x 2.3-diam spherogive
a Hemi
B
1.0 -diam spherogive
R I Head shapes selected sphere
head g

Maximum
,diameters
1.8 for study

diameter
D А

cavity
FRp
1.0-diam ogive

of
- Rp 1.7
(estimated ) Head /c , 70°

А Center of curvature of ogive Head a 1.5-diam ogive Head b,

1.6
B Center of curvature of sphere 57 °
C Junction of sphere and ogive Х
D Junction of ogive and cylinder 1.5 2.3 - diam ogive
E Junction of ogive and plate
P Plate radius (estimated )
R Sphere radius 1.4
Ro Ogival radius
R. Headd 3.5 -diam ogive
Rp Body or projectile radius I. Head f, 70°
a Half-sphere angle 1.3
Head e, 65

5.0- diam ogive /

Figure 5.2 . Geometry of ogive, spherogive ,
and plate-ogive heads. 1.2

1.1
that a segment of a sphere centered at the
missile CG was substituted for the flat plate . The
spherical radius is 6.5 times the projectile radius. 1.0 ‫ار‬
Head j differs from Head h in that an annular 0 0.100 0.200 0.300 0.400 0.500
groove was added just aft of the nose flat. Radius of sphere, R , diameters
Head k has a spherical-cap nose with a radius of Figure 5.3 . Radius of sphere versus cavity diameter.
0.576 +0.002 of the projectile radius and meets
the conical section tangentially.
SELECTION OF HEAD SHAPES size dividing the two regimes varies with differ
It was expected that water entry would ent spherogive families. There appears to be a
take place under conditions of well -developed linear relationship (shown as a solid line in
cavitation and that cavity separation would Figure 5.3 ) between sphere size and maximum
affect missile performance. Since water-entry bubble diameter when cavitation occurs on the
cavity separation should approximate steady spherical portion of the nose .
state cavity separation , the results of steady The 1.5- and 3.5-diameter ogive families
state cavity studies on 1.0., 2.3-, and were selected for extensive water-entry behavior
5.0 -diameter ogives and spherogives ( Figure 5.3 ) studies . The maximum diameters of the cavities
were used as a guide in the selection of head forming on the ogive were estimated (by para
spheres.1 These data indicate that when the bolic interpolation of those forming on the 1.0 ,
spherical segment is large in comparison with the 2.3 , and 5.0 -diameter ogives ) to be 1.60 and
ogive, cavity separation takes place on the 1.34 diameters . When these dot -dash lines are
sphere, and cavity growth is controlled by extended to meet the solid line, the points of
sphere size. When the spherical segment is small junction give the critical sphere radii of 0.40 and
in comparison with the ogive, cavity separation 0.30 diameter, respectively, for these ogives. The
occurs on the ogive and cavity growth is corresponding half-sphere angles are approxi
independent of sphere size . The critical sphere mately 66 and 70 degrees.

lFigure 5.3 is taken from Reference 33. It shows the original data; the dot-dash lines for the 1.5- and 3.5- diameter
ogives and lettered points were added subsequently .

71
 Cap sizes below and above the critical value Center of

ant
(57 and 70 degrees ) were selected in the gravity

ult
e
forc
Res
1.5 -diameter sphere-ogive family in order to Ax
compare water-entry performance in the pres Res
ult
is

ence of both spherical and ogival separation. The ant force Moment
critical sphere size (70 degrees ) , where variation
in angle of attack would result in varying degrees
of separation on both the spherical and ogival Center of

ant
portions , and the subcritical 65 -degree cap sizes gravity

ult
were chosen for the 3.5 -diameter family . The Center of

Res
points representing these heads are identified in A pressure

ce
Rexsis

for
ult Center of pressure
Figure 5.3 by their letter designations. a ntſ force
The plate-ogive heads were selected in such Moment

a manner as to extend the ogive-head data over

the entire range O < Ro/Rp < . Because pure

ant
ogives are not possible in the range 0 < Ro /Rp < 1, Center of gravity

ult
Center of
Center of pressure

ce
a modification was introduced in the form of a Ax
is

Res
for
pressure
flat plate . Moment

In order to present the ogive - plate -ogive

data graphically , the arbitrary parameter.

1
Q = ( 5.2 )
1 + R / RR Center of pressure
Center of pressure

was used. This parameter has the more conveni Figure 5.4. Water entry , Head h .
ent range 0 < Q < l. The value of Q for the
hemisphere head is 0.5 and lies in the middle of
the range .
Test shapes corresponding to Q values of Interest in the configuration designated
0.667 , 0.833 , and 1.000 were selected : the plate Head j was prompted by studies of water
0.25-diameter ogive , the plate 0.1 -diameter running bodies . A sharp -edged separation ring or
ogive , and the plate cylinder, respectively . groove in the head of a Henschel body causes
Cones of 8.4 -degree generating angle were the flow to separate from the body at the
selected primarily because they approximated discontinuity , surrounding it with a cavity. The
the contours of rocket ordnance . Head h, a size of the cavity is controlled by choice of head
truncated cone , was of interest because it was shape . A conical midsection with aa flared skirt at
likely to experience low whip and reasonably the rear causes the body to oscillate back and
low water -entry deceleration. Low whip was forth in the cavity, thereby stabilizing the
expected because nose-down forces generated as underwater trajectory. Of particular interest in
the disk was wetted would tend to be canceled Head j was its underwater trajectory perform
by nose-up forces on the conical section ( Fig ance ; its water-entry behavior was expected to
ure 5.4 ) . Head i was chosen in order to ascertain be similar to that of Head h .
the effect аa curved nose would have on the whip
of a missile if it were so designed that the nose EXPERIMENTAL CONDITIONS
forces passed through the CG of the torpedo, The full -scale dummy torpedoes used in
thereby producing no moment. Since the shear this study of whip and deceleration at water
forces on the curved nose should be very small entry were launched into the fresh water of the
in comparison with the inertial forces, it was NUC Morris Dam Torpedo Range (Part V). The
hoped that the contributions of the flat plate trajectory angle at water entry was about 20
and conical section of the head to the whip degrees . Nominal entry velocities were both 400
could be estimated from a comparison of the and 500 fps for all except the two bluntest
whips of Heads h and i . heads (m and n were restricted to 400 fps to
72
avoid damage from the high water-entry deceler Table 5.1 lists the ballistic parameters of
ation forces ) . At these velocities, o< 0.03 and torpedoes with the different head configura
comparison of this low cavitation number with tions . All were 22.42 inches in diameter, and the
the results of other investigations (Refer CG was always within 0.01 inch of the longi
ences 31 , 32 , and 33 ) indicated that water entry tudinal axis of symmetry . Because of the con
would take place under conditions of well siderable variation in head shape and the
developed cavitation . consequent variation in weight and weight

TABLE 5.1 . Ranges of Ballistic Parameters of Torpedo Dummies

Moment of
Length Weight Distance Distance Buoyancy inertia of
Length torpedo torpedo CG from CB from of torpedo torpedo
of head , a dummy,b dummy , nose , nose , dummy, dummy ,
с

Head in. ft lb in . in. lb slug-ft2

a 51.7 12.8 1,480–1,520 71.1 70.5–71.1 168-208 748–756

b 51.7 12.8 1,514–1,663 71.1-72.9 69.6 65--214 719-851

с 51.7 12.8 1,526–1,549 71.1-72.1 68.9–69.0 204-227 789

d 51.7 12.8 1,461-1,565 71.1 75.3 21-125 693–766

e 51.7 12.8 1,457–1,494 71.1 68.1–71.9 160–192 711-719

f 51.7 12.8 1,614–1,680 71.1 70.2–70.3 70-98 803-821

600

g 51.7 12.8 1,526–1,527 69.5–70.0 67.5 268-269 771

h 51.7 12.8 1,397–1,470 71.1-73.3 72.1-72.9 136-209 682–737

i 51.7 12.8 1,496–1,555 73.0 73.2 59-118 752–764

j 51.7 12.8 1,537–1,559 71.1 68.0–71.1 49-77 766-814

k 51.7 12.8 1,430–1,513 71.1-72.9 72.9–73.4 124-176 707_736

e 45.6 12.2 1,582–1,590 63.7 63.7 150–158 743

m 42.3 12.0 1,544–1,558 63.7 68.3 182-196 694

n 42.3 12.0 1,547–1,552 63.8 68.2 153–193 694

a Distance from nose to joint ring.

b Distance from nose to tail plug.
CAbout a transverse axis through CG.
73
distribution , it was impossible to adjust all the 390

body parameters to the same nominal values. In Water entry

order to compare the data, corrections were

‫ مم‬,
y
380 9066

fVelocit
necessary; the actual range of values is shown in
Table 5.1 .
esolo Plotted for a constant

, ps
370 CD = 0.25
The whip data were corrected to a moment
of inertia of 800 slug-ft2 and an entry velocity

‫و وه‬
‫د‬

‫ع‬
360
of 500 fps. The water-entry deceleration was

‫عمم‬
corrected to a torpedo weight of 1,500 pounds.2

‫مه‬

‫ه‬‫م‬
‫ده‬
Because the linear velocity of the missile at the 350
end of the whip phase was within 2.5 percent of
the entry velocity, the entry velocity was used in
reducing the whip data. Torpedo yaw at water 0 10 20 30 40 50 60

entry was small and therefore was neglected. Time, ms

Whip , pitch, velocity, and deceleration at
Figure 5.5 . Deceleration during water entry .
water entry was recorded with a flare camera
(described with other equipment and facilities in

attack
Angle
3
Part V) . The slope of the velocity-time curve

,deg
of
obtained from the range position data gives the
deceleration ( Figure 5.5 ) . The angle made by 2
--- gna
lines joining synchronous flare-image pairs with ced
the torpedo line of flight gives the angle of or
attack. Since the line of flight does not change
perceptibly , the rate of change in angle of attack 0 1

60
is equal to the rate of change in pitch. There 0 10 20 30 40 50
fore , the pitch velocity can be obtained from the Time, ms
slopes of the angle -of- attack -versus-time curves
( Figure 5.6 ) . Water-entry angles of attack of Figure 5.6 . Pitch velocity during water entry .
varying degrees were obtained by attaching to
the muzzle of the launcher tube steel extensions
which induced asymmetric blast, thereby all head shapes , it was found that the whip is a
imparting a pitch velocity to the torpedo. linearly increasing function of the angle of
Whip data for Heads a and d are shown in attack for the range of angles studied. Regres
Figure 5.7 together with regression lines fitted sion lines fitted to the data for all heads are
to the data by the method of least squares. For shown below, where a is in degrees and A ) is in

2To correct whip to the nominal entry velocity of 500 fps (Reference 106) :
500
Δθ .
, I
- ‫فح‬
V

To correct whip to the nominal moment of inertia of 800 slug-ft2 :

I
Do2 z =
‫فح‬
800

To correct water-entry deceleration to the nominal missile weight of 1,500 pounds:

m
a
a1 1500

Measured values of entry velocity, moment of inertia, whip, weight, and deceleration are v, 1, , m, and a,
respectively . The desired corrected whips and deceleration are A01 , Að 2 ,and aj . These relationships are discussed in
References 91 and 106 .

74
 500 compare whip at -4, 0 , and +5 degrees and at
x 400 fps
500 -fps velocity , the whips and attack sensitivi
400
ties of the ogives and spherogives are plotted in
/s,dWhip

. 500 fps
Aec eg
ė

Figure 5.8 as functions of the ratio of spherical

300
cap radius to body radius .
For the same value of R /Rp, the
200 t
tte 1.5 -diameter spherogive family (Heads a, b, c)
100
shows greater whip than the 3.5-diameter family
Steep Flat (Heads d, e, f) at both zero and - 4-degree entry
0 angle of attack and, except for smaller values of
-6 -4 -2 0 2 4 6
R /Rp, the + 5 -degree curves follow the same
Angle of attack , a , deg trend. The high attack sensitivity of the
Head a . ΔΘ 13.9a + 175 3.5 -diameter ogive causes the observed reversal
in the whips of the pure ogive heads . Head e,
with only a small spherical cap , also shows a
400
high attack sensitivity.
300
X 400 fps The whips and attack sensitivities of the
/s,dWhip

o 500 fps ogive, plate -ogive series (Heads d, a, g, l, m, n)

aec eg
ė

200
are shown in Figure 5.9 as functions of the
parameter Q (Equation 5.2 ). A hypothetical
o

= oo
100
to “ needle ” ogive head of infinite radius ( Ro = 0o)
should have zero whip , and this whip was
0 included in the 0-degree curve to provide aa limit
Steep Flat point for Q = 0. The distances of the CG from
-100 I the nose were about 70 inches for the hemi
-6 -4 -2 0 2 4 6
sphere (Head g) and the ogives and 64 inches for
Angle of attack, a, deg the plate ogives. (The CG position affects the
Head d . a ) = 22.5a + 151
air
whip and attack sensitivity. ) The data could not
be corrected to a standard CG position because
Figure 5.7 . Angle of attack dependence of whip . the exact relationships were unknown ; the ogive
Data corrected to 500 fps entry velocity and and plate -ogive data should be compared only
800 slug- ft2 moment of inertia. for trends .
The general shape of the whip functions is
degrees per second. It is difficult to evaluate similar for the attack angles studied. The attack
whip performance from these data because the sensitivity of the plate-ogive series decreased
change in whip with with increasing Q (decreasing R) to a minimum
Head a . Δό 13.9a + 175 value of essentially zero for the plate cylinder
Head b Δθ = 19.0α + 150 (Head n ), and its whip can be assumed to be
Head c . A 0 = 15.8Q + 113
=
independent of the angle of attack within the
Head d. A0 = 22.5~ + 151 range of entry angles investigated. For the
Head e . A j = 21.2a + 134 hemisphere head (Head g) , the contour contact
Head f. Δθ = 16.3α + 88
ing the water remains unchanged by the angle of
Head g . ΔΘ
.

=
10.2a + 133
attack, and only the whip-producing drag force
Head h . ΔΘ = 12.7 ~ + 62
lever arm differs. A theory developed on this
basis for the attack sensitivity of the hemisphere
Head i . Δθ = 13.9α + 57
Head j . Δθ = 13.0α + 49 head shows good agreement with the experi
mental evidence (Reference 91 ) .
Head k.
. Δό 16.40 + 81 The cone heads h , i , and j show small
Head l . Δθ = 5.0α + 65
differences in whip and, as might be expected
Head m . A0 = 2.3a -24
A 0 = 1.02a -48
AO
from the behavior of the hemisphere and plate
Head n. heads, the whip of Head k is greater than that of
entry attack angle (i.e. , attack sensitivity ) varies the other cone heads ( Figure 5.7 ) . The contribu
greatly among the different head shapes. To tion of the conical section of the plate-cone
75
 Angle of attack , deg

sensitivity
140
280 0 +5 30

/),(daChange
attack ttack
d

angle
whip
0 0 g
Attack sensitivity

eg
28

sec
of
A -4

in
a b Deceleration
240 с 120 See Equation 5.4 26
d

24

ation
200 с 100 b 22

Tatian

20
g
sWhip

Deceler
a
eg
/,dec

k k f
of

80

-
160 18

,g
d b

e 16

bus
120
a
60 bo , 14

bus
g 12

I
k
bus
k g
80 40 10
b
d d
8
е

20

-
6

il
40
k yf 2
A
0 0
0
0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0.8 1.0
0.8 1.0

Ratio spherical cap radius to body radius, R/Rp Ratio spherical cap radius to body radius, R/Rp
Whip graph Attack sensitivity and deceleration graph

Figure 5.8 . Whip , attack sensitivity , and deceleration of water entry as functions of spherical cap radius, spherogive
heads. Entry velocity 500 fps.

Head d Head a Head g Head Head m Head n

40

sensitivity
),Change
(attack
whip
300 +5 deg 30 attack
angle
in
eg
/d)s( ec
of
20

200
Whip
/sWhip

Xt
,deceg

o deg
100 10

Attack sensitivity
-4 deg
0 0
1

-100 1 1 1 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Ratio, Q = 1 /( 1 + Ro/Rp)
Figure 5.9 . Whip at --4 ., 0- , and + 5 - degree angle of attack and attack sensitivity at 500- fps entry velocity .

76
heads to the whip can be estimated by compar 120
110
ing the data of Heads h and j with results 100
obtained from Head i, where forces acting on
90
the spherical portion of the head have no

deceleration
80
moment. The similarity in performance indicates 70
that the flat plate had little effect on the whip ;

Entry
60
the whip of these heads was produced largely on

,g
the conical surface . 50
Mean water-entry decelerations were meas
ured during the time from nose contact to tail 40

submergence, and the corresponding mean entry

velocities were also determined for use in drag 30
coefficient calculations and in the subsequent
deceleration curves .

25
The deceleration , corrected to a missile
weight of 1,500 pounds , was plotted on a 20

logarithmic scale as a function of the water

entry velocity. On the hypothesis that the 15
deceleration would be of the form 100 150 200 300 400 500 600

Mean entry velocity , fps

a =
cCVy2 ( 5.3 )
Head a. a = 2.47 x 10-4 v2 ; Cp = 0.14
(where a is the deceleration , v the water-entry Figure 5.10. Deceleration as a function of mean velocity
velocity , and c a constant characteristic of the at water entry .
head ) , parabolic regression curves of the type
given by Equation 5.3 were fitted to the data by
the method of least squares. These regression
curves appear in the logarithmic plots as straight
lines with a slope of two. The good agreement of coefficients indicates that drag forces other than
the data with these lines shows the validity of those acting on the spherical cap are small. The
drag coefficients for the 1.5- and 3.5 -diameter
the assumption that the cavity drag is propor
tional to v . ogive heads show good agreement with results
A typical plot of water-entry deceleration obtained from integration of experimentally
(Head a) is shown in Figure 5.10 ; the formulas obtained pressure distributions (Reference 56 ).
Head a . a = 2.47 x 10-4v2 ; CD = 0.14 =
for all 14 heads are given below, where v is in
feet per second and a is in gravitational units. Head b . a = 2.50 x 10-4y2 ; CD = 0.14
= =

The drag coefficient for the hemisphere (Head g) Head c . a = 3.93 x 10-4v2 ; CD = 0.22 =

was 0.30 , the same as that measured for a Head d. a = 1.50 xX 10-4v2 ; CD = 0.09

hemisphere-head dummy aircraft torpedo at Head e . a == 1.47 xX 10-4v2 ; CD = 0.08

entry velocities of about 400 fps (Reference Head f. a= -
1.90 x 10-4v2 ; CD = 0.11
106 )3 . It is in good agreement with the results of
.
Head g. a= 5.33 x 10-4y2 ; CD = 0.30
water- tunnel tests (Reference 32 ) and with Head h . a= 3.84 x 10-4v2 ; CD = 0.22
integrated experimentally obtained pressure dis Head i . a= -
3.55 x 10-4y2 ; CD = 0.20
tributions (Reference 56 and 73 ) . Other investi Head j . a == 3.65 x 10-4v2 ; CD = 0.21 =

gations showed that for Head k , cavitation Head k . a = 1.89 xx 10-4v2 ; CD = 0.11
occurs on the spherical cap , and its drag coeffi Head l. a = 7.85 x 10-4v2 ; CD = 0.44
a
-

cient based on the cross -sectional area of the Head m . -

a = 11.05 x 10-4v2 ; CD = 0.62
sphere is 0.32 . The agreement between the drag Head n . a = 14.71 x 10-4y2 ; CD = 0.83

3The Cp of the aircraft torpedo, based on its maximum cross-sectional area, was 0.25 . The radius of the nearly
hemispherical cap of the torpedo was 10.30 inches . The value of 0.30 was obtained when the CD was calculated on
the basis of the projected area of the sphere .

77
 If it is assumed that water contact is The drag coefficients of the ogive and
restricted to the spherical cap of the spherogives plate -ogive heads and drag coefficients deter
and that the area of contact is proportional to mined by integrating the pressure distributions
the square of the cap radius, then from the basic of ogives and plate ogives for o = 0 and 0 = 0.05
drag equation and from Newton's Second Law (Reference 73 ) are plotted as a function of Q in
(Reference 91 ) , Figure 5.11 . The deviation of the Head d
2
(3.5 -diameter ogive) drag coefficient from the
.
a
R
dotted curve indicates that skin friction is more
a =
important in determining the drag coefficient of
-- Bp
R
р
( 5.4)
finer noses . The contribution of the skin
friction to the Head d drag coefficient was
estimated as 0.0085 , which would account for
where ao is the deceleration of the hemisphere about half the discrepancy .
head . The drag coefficient of the plate -cylinder
This parabolic relationship is shown as a head is 0.83 , in good agreement with the value
dot-dash curve in the righthand graph of Fig of 0.82 derived from free -streamline theory
ure 5.8 . Heads c , f, and k lie close to the (Reference 65 ) and with the value 0.80 calcu
parabola ; Heads a, b , d , and e (which are either lated from pressure distributions measured in
pure ogives or have small spherical caps ) show the California Institute of Technology water
deviation . Evidently, cavitation occurs on the tunnel (Reference 32 ).
ogival portion of these heads as originally Heads h , i , and j showed similar water-entry
predicted from the water-tunnel data. For cap decelerations , in keeping with the High-Speed
sizes above the critical, the deceleration curves Water-Tunnel data (References 31 , 32 , and 33 )
for both spherogive families appear to follow the that indicate cavity separation on the plate or
parabolic curve . For cap sizes below the critical, spherical cap . The drag coefficients of the
there is a departure from the curve, the plate-cone Heads h and j , expressed on the basis
1.5 -diameter spherogive family having higher of the plate area , were 0.88 and 0.84 , respec
deceleration than the 3.5 - diameter family , as tively. These corrected drag coefficients show
would be expected from the data of Figure 5.3 . good agreement with that of 0.83 obtained for a

Head d Head a Head g Head ! Head m Head n

1.0
coefficient

0.8
Drag
,CD

0.6

0.4

0.2 o Water -entry data

Oo = 0.001 Integrated pressure
Xo = 0.05 ) Distribution ( Ref. 73 )
0.0 1 1
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Ratio, Q = 1 /( 1 + Ro/Rp)
-

Figure 5.11 . Drag coefficients, ogive and plate-ogive heads.

78
full-diameter plate head on another dummy The data for Head j (Henschel type ) at the
aircraft torpedo launched under similar condi nominal 400- and 500 -fps entry velocities are
tions, and with the value 0.82 derived from the taken from Reference 95 as examples of stable
free-streamline theory (Reference 65 ). The underwater trajectories. The relative launching
annular groove of Head j evidently has little conditions are shown in Table 5.2 , elevation and
effect on water-entry deceleration. The cor plan views of the trajectories in Figures 5.13 and
rected drag coefficient of 0.80 for Head i indi 5.14 , and vertical and lateral deviation from the
cates that the slight curvature of the spherical extended air trajectory after 100 feet of under
nose had little effect on the water-entry deceler water travel in Figures 5.15 and 5.16 . Additional
ation and that other drag forces acting on the data are available in Reference 95 .
nose are small. Some general trends were apparent in the
Underwater trajectories, recorded while trajectory data obtained with the 14 head
water -entry whip and deceleration data were shapes . The first 100 feet of underwater trajec
being obtained , were determined from the tory was not appreciably affected when (other
hydrophone array shown in Figure 5.12 and entry conditions being equal) the water-entry
described in Part V. Hydrophones recorded the velocity was increased from 400 to 500 fps.
position of the missile at several known times Apparently , gravitational and buoyancy forces
during underwater flight, and nets were sta are of minor consequence over this velocity
tioned from 530 to 780 feet from the entry range. (A detailed explanation is available in
point at intervals of 50 feet and supported at the Reference 95. ) Farther along, the trajectory was
surface by wire cable floated by corks . The nets influenced by gravity and buoyancy forces and
were light enough not to impede the flight of these forces became more predominant after the
the missiles, and the location of the holes made torpedo had lost most of its velocity .
in them provided additional data.

ur

ny
isi

Ker TE

1100
5 100'
3
17

16 9 100'
Station 7. 10
525'
68 11
50

625 ' 412

°
50

725'
%

Hydrophones 1 , 3 , 5 , 7 , 9 , 11 are 4 ft below surface

Hydrophones 2 , 4 , 6, 8 , 10, 12 are 54 ft below surface 775'
'50

• Marker buoys
D - D = Centerline of range
- 825'

Figure 5.12 . Fixed-angle launcher site , hydrophone locations.

79
 TABLE 5.2 . Launching Conditions, Head j

Angle Water
Curve No. of Pitch Entry
( Fig. 5.13 Attack, Whip, Velocity, Yaw , Velocity ,
and 5.14) deg deg/sec deg /sec aa deg fps

1 5.8 94 100 1.4 394

2 4.4 79 96 -1.0 400

3 1.7 0 -1.9 395

4 0.7 -1.7 414

5 -1.7 399

1 5.6 161 176 -1.0 507

2 3.5 105 7ob -1.0 507

3 -2.5 507

4 2.7 77 92 0.0 504

5 1.1 54 40 0.5 502

6 0.8 63 72 2.1 500

7 0.6 58 91 -1.0 506

8 0.2 -0.5 471

9 -0.2 42 24 -0.7 506

10 -0.8 40 28 0.7 511

11 -2.2 33 20 0.4 508

aPitch velocity just after whip has taken place.

bPitch velocity subject to error because of lack of information on roll.

80
 0

400 fps

2
.
20
1
5
3
Depth

40
,ft

20-deg line

60

4
80

100
0 20 40 60 80 100 120 140 160 180 200

Horizontal distance, ft

500 fps
20

2
Depth

1
40 4
+ſ
f, t

20- deg line

or

5 3
7 -
1

60 6
8
9

80 10

100 1
0 20 40 60 80 100 120 140 160 180 200
Horizontal distance, ft

Figure 5.13 . Elevation, Head j underwater trajectory .

With the ogive and spherogive configura maximum depth of dive with increase in attack
tions , the trajectory rapidly became more angle .
shallow with increase in entry angle of attack. Several of the full-scale missiles used in this
The data indicated “ critical ” angle of attack4 study of whip and deceleration at water entry
values, and there was fair agreement between the also served as prototypes in subsequent studies
experimental and calculated values of critical of water-entry pitch modeling (Chapter 7 ). The
pitch. For plate-cone configurations, the data data obtained , although not directly related to
were less defined , although there was a sugges the modeling of their behavior , could have direct
tion of critical angle of attack. With the plate application to missile design and could also
ogive configurations, there was evidence of a point up areas in missile water-entry behavior
critical angle of attack and , in contrast to the that should be studied through correlation with
ogive-spherogives, evidence of an increase in models .

4By “ critical” is meant that angle of attack at which the trajectory changes from upturning to downturning or vice
versa .

81
 Left
60

400 fps
deviation
Lateral 40

20
,ft

2
0 4
5
1
3

20
U

40
Right

60 1 1 1 1
0 20 40 60 80 100 120 140 160 180 200
Horizontal distance, ft
Left

60

500 fps
40
deviation
Lateral

20

7
,ft

2
5
0 8

3
20

9
6
10
11

40
Right

60 1
20

0 40 60 80 100 120 140 160 180 200

Horizontal distance, ft

Figure 5.14. Plan view, Head j underwater trajectory .

82
 deviation
Vertical

40

Head j
,ft

20

o
0
‫ܩ 'ܩ‬
O
-20
o 400 fps
-40
Steep
o 500 fps Flat
-60 1
-8 -4 0 4 6

Pitch , deg

Figure 5.15. Vertical deviation at 100 feet, Head j .

Left

40
deviation

20 Headj
Lateral

0 og 8 . do
,
ft

-20
O
400 fps
Right

-40 500 fps

Right Left
-60
-4 -2 0 2 4

Yaw, deg

Figure 5.16. Lateral deviation at 100 feet, Headj .

83
CHAPTER 6
WHIP AND UNDERPRESSURE

During early modeling studies of oblique entry of a 2-inch-diameter plate-cylinder-head

water entry of missiles , it was observed that model (entry velocity 128 fps, trajectory angle
Froude scaling alone did not accomplish model 20 degrees, and atmospheric pressure over the
ing, the model generally tending toward a more water surface ) . An undercavity of air and water
downward trajectory than the prototype ; in vapor has been sealed off by the splash lip
some instances the model dived while the impinging on the lower side of the cylinder.
prototype turned upward and broached. L. B. According to the hypothesis stated above , the
Slichter (Reference 61 ) investigated this anom cavity expanded after it was sealed off from the
alous behavior and established that portions of outside air, thereby reducing the gas pressure
the underside of the model may be subjected to within it to below atmospheric pressure. The
less than atmospheric pressure during water splash lip is even more apparent in Figure 6.2,
entry . In that event, atmospheric pressure acting which shows a flat -plate model entering water
on the upper side of the model nose could, in
opposition to the reduced pressure areas beneath
it, generate a moment large enough to overbal
ance the water-pressure forces completely and
produce a nose-down whip and diving trajectory.
Slichter's studies were confined to small
models, but the inference is plain : if this
reduced - pressure or under-pressure effect
obtains for models, it must also obtain for
full -scale missiles . The magnitude of the under
pressure effect was not known at that time, but
it was clear that the effect would have to be
scaled if correlation was to occur, and that
Froude scaling alone was insufficient .
A hypothetical explanation of the genera
tion of underpressure areas is that a cavity forms
beneath the missile nose when it enters water
obliquely . While it is first forming, the cavity is
open to the atmosphere and is composed of
water vapor and air, but as the missile penetrates
farther into the water, the cavity is sealed off
from outside air but continues to grow , with the
result that the gas pressure inside it is lowered,
and the pressure on the lower side of the nose is
reduced to less than atmospheric.
Supporting evidence for the hypothesis can
be seen in Figure 6.1 , which shows photograph
ically and diagrammatically the oblique water Figure 6.1 . Oblique water entry of plate-cylinder head.

84
 underpressure effect. Such venting can be
accomplished by modifying the missile head,
and if the other water-entry characteristics of
the head are changed only slightly, vented and
unvented heads can then be launched under
similar conditions to determine whether aa signi
ficant underpressure exists at service-missile
scale. Model heads with different degrees of
venting can be launched under similar conditions
at atmospheric pressure to provide information
on the correlation of the underpressure effect
with undercavity behavior. Finally, reducing the
air pressure over the water surface should also
reduce the underpressure effect. The launching
of an unvented model head at different ambient
air pressures but with otherwise similar condi
Figure 6.2 . Oblique water entry of flat plate .
tions should not only provide information on
the correlation of the underpressure effect with
obliquely. The sides of the flat plate are in cavity behavior but should also indicate whether
contact with the glass walls of the tank, there is reduced air pressure could be used to scale the
no flow around the sides of the plate, and the underpressure effect. The NUC studies described
flow is essentially two-dimensional . The splash below include published (References 99 and
lip that would form the undercavity seal if a 100 ) and unpublished data.
rectangular-head configuration were used Prototype missiles selected for study con
appears as a cusp behind the plate . sisted of vented plate and unvented plate
If the assumptions of the hypothesis are cylinder heads attached to a torpedo after body
correct, venting the undercavity to the atmo (Figure 6.3 ) . Duralumin 2-inch-diameter models
sphere should equalize cavity and ambient air were built with different degrees of venting
pressure , thereby reducing or nullifying the achieved by varying the cylinder length. The

7
143
8
42
21
CG
-

63.7
Plate -cylinder head

5
127
8
1
- 54 26

CG

后 Vented-plate head
60.4

Figure 6.3 . The 22.42-inch-diameter prototype. ( Dimensions are in inches. )

85
head shapes (designated A through H) and their attempt was made at correlation ; the sets of
dimensions are shown in Figure 6.4 . ( The appli tests were independent of each other.
cation of Head H is described later , in connec Whip was arbitrarily defined as the change
tion with Figure 6.20 . ) These heads , as well as a in pitch velocity between water contact and the
plate-cylinder head, were attached to model complete submergence of the nose plate ; this
afterbodies , as shown in Figure 6.5 . definition proved to be physically reasonable
Prototype and model parameters are shown upon inspection of the pitch -velocity time func
in Table 6.1 . Variations in head configuration tions ( Figure 6.6 ) . Whip so defined is equal to
and the consequent weight and weight the corrected pitch velocity at the time of
distribution differences made it impossible to complete nose-plate submergence because all
adjust the body parameters to bring them into data are corrected to zero pitch velocity at water
complete agreement within each set of tests. contact .
However, the distance of the CG from the nose To investigate the effects of venting on
of the missiles was held constant (by adjustment whip, full -scale dummy torpedoes and 2-inch
of internal weights ) within each set , and in all diameter models were subjected to test launch
instances the CG was located on the longitudinal ings. The prototypes were launched into fresh
axis of symmetry . Corrections were applied for water at the NUC Morris Dam Torpedo Range
the small variations in moment of inertia. 1 No ( Figure 5.12 ) . The torpedo with the plate
cylinder head was launched 12 times from the
Fixed -Angle Launcher (FAL) , the torpedo with
3 169 in .
3
0.5 in . the vented head 21 times from the Variable
typical
4
1
in. - 人 45 45 typ Angle Launcher (VAL) . The trajectory angles at
water entry ranged from 20 to 22 degrees ; the
1. in.typ nominal launching velocity was 400 fps. To
Head A , fully -vented plate Head B , plate study the correlation of the underpressure effect
0.5- in. cylinder with cavity behavior and the degree of venting,
1.0 in , 1.5 in . the model afterbody shown at the top of
ㅕㅑ ៗ Figure 6.5 was launched with Heads A to G into
the Open Launching Tank at the NUC Hydro
Head C, plate Head D , plate
ballistics Laboratory. The entry angle was
1.0-in. cylinder 1.5- in. cylinder 20.35 degrees, the entry velocity was 137.5 fps,
and the atmospheric pressure was ambient. To
2.0 in .
study the effects of atmospheric pressure and
water surface tension on whip , launchings using
only the fully vented model (Head A with the
afterbody shown at the top of Figure 6.5 ) and
Head E, plate 2.0-in. Head F , unvented the unvented model (shown at the bottom of
cylinder platc-cylinder Figure 6.5 ) were made later in the NUC
Variable-Atmosphere Variable-Pressure (VAVP)
3 in.t 23
Launching Tank (described in Part V) . The entry
yp in .
8 32
3
5 angle was 20 degrees, pressures ranged from 0.1
in .
forsine 16 to 1.5 atmospheres, and launching velocities
were 80 to 81 fps. Velocities were measured
SHO 1
140° with a photocell timer and whips were measured
with the optical whip recorder (Part V) .
22 -in,. hole Actual launching conditions for all missiles
Head G , vented plate Head H , fully-vented are listed in Table 6.1 . The effects of yaw at
2.0- in. cylinder spherical sector water entry were neglected because in most
cases the yaw was very small (less than a degree)
Figure 6.4. Head shapes of 2 -inch - diameter models. and the maximum angle recorded was

Isee footnote 2 , Chapter 5.

86
 14.0
Head

CG

6.2

For venting studies in Open Tank.

8.4

2.25 6.15 Head

十
CG

5.68

For atmospheric pressure and water surface tension studies in VAVP Tank .

Figure 6.5 . Two - inch -diameter model afterbodies with plate-cylinder heads. (Dimensions in inches. )

0 1.8 degrees. For comparison of results, proto

type data were corrected to a nominal water
entry velocity of 400 fps and moment of inertia
Corrected

-50
velocity

of 800 slug-ft2 ; model data in the first set were

pitch

corrected to 137.5 fps and 21.3 lb -in 2. All data

eg
/s,dec

-100
were corrected to zero pitch velocity at water
Curve faired into data points contact .
-150
Plate submergence was defined in terms of
-200 Time of complete the geometry of the static model system ( Fig.
plate submergence ure 6.7 ) . The water -contact velocity was used to
--250 calculate the time of plate submergence and was
also used in the whip calculations because
-300 experimental evidence showed that the velocity
decreased only 2.5% between contact and total
-350 plate submergence. The whip was measured
Whip
-400 1
from the fitted curve .
I
0 1 2 3 4 5 6 7 8
Prototype whip as a function of attack
Time from water contact, ms
angle is shown in Figure 6.8 . In spite of the large
scatter in the data, it is apparent that the
Figure 6.6. Typical water- entry pitch velocity - time missile's negative whip is reduced by venting, as
plot, 2-inch plate cylinder, Head E. predicted by the underpressure hypothesis. The
87
 Whip-attack regression equations

‫هر‬ Half submergence

o
X
Head A
Head BAė = 11.6a - · 139.1
A --Û.84u -- 128.2

x
0
Undisturbed water level
0 o Head C Δή 17.5a 220.7
Head D Aė = 19.8a 292.9
+ Head E AO = 26.9a - 335.8 -

Head F 40 = 21.3a - 354.3

Head G A) = 25.6a - 198.5

sWhip
Full submergence -100

eg
Undisturbed

/d,Aec
ė
Head B
water level A
b
A
B XXX
Figure 6.7 . Plate at half and full submergence. G

a
g
-200

oa
GO

DO

o
o
o
С
Vented plate

O
Plate cylinder D
/s,dWhip

Δή 0.1la - 30.8
Aec eg
ė

1
-10 Δθ == -0.6la -- 42.4
Aė

+
E
ΔΔ
-300 D

+
+
-20 F

+
e .

+
a

A
-30 E
OO
A

O
QA
-40
-400 1 1 1
A 0 0.5 1.0
-50 -1.0 -0.5 1.5

Steep Flat
Angle of attack, a , deg
-60 III
-6 -4 -2 0 2 4 6 8
Figure 6.9. Whip versus angle of attack , plate and
Angle of attack, a , deg plate-cylinder heads.

Figure 6.8. Dependence of whip on angle of attack. Water

entry velocity 400 fps, 20- degree water-entry angle . vented missiles ceased at about one-half plate
submergence regardless of cylinder length ; the
test data also prove that the phenomenon of whip function of the unvented head indicated
underpressure exists with full-scale missiles as that underpressure continued almost until total
well as with models . plate submergence .
The whip of the plate-cylinder model Duplicate launchings of models with
becomes increasingly negative as the cylindrical Heads A through G were photographed with the
section is lengthened (Figure 6.9 ) , which can rotating disk camera ; side views of Heads A, E,
reasonably be attributed to underpressure and F launchings are shown in Figure 6.11 . It is
because actual impact of the flow upon the apparent that half-plate submergence corre
underside of the cylinder would naturally result sponded approximately to the complete entry of
in positive whip-producing forces . The vent hole the 2 -inch cylindrical body section and that the
in the 2-inch cylinder head reduced the whip CG crossed the interface at about the time of
considerably , but venting was not complete, The photographs of Figure 6.12 show that
approximating instead that of a cylinder some small secondary cavities formed as the midsec
what less than 1 inch in length. tions of the vented models crossed the interface
The whip as a function of distance traveled but did not form during the launching of the
beyond water contact is plotted in Figure 6.10. unvented missile . The cylindrical section of the
These averaged curves show that evidence of unvented head remained in contact with the
underpressure in the whip functions of the water as the upper half of the plate face was
88

1
 TABLE 6.1 . Ballistic Parameters and Launching Conditions,
Plate-Cylinder and Vented-Plate Missiles
Full scale ( 22.42 in. ) Small scale ( 2 in. ) Small scale ( 2 in. )
FAL VAL Open tank VAVP tank
Plate Vented Plate Vented Plate Vented
cylinder a a b с Ven d
Parameter cylinder plate plate cylinder plate

Ballistic

Weight, lb 1547 1481 1.081 1.170 1.100 1.172

1552 1502 1.174

Length of 143.9 127.6 14.0 14.0 12.0 14.0

missile , in .
Distance of CG 63.7 60.1 6.20 6.20 5.68 6.20
from nose, in. 60.6

Moment of 694 503.5 20.22 20.22 18.35 20.22

inertia about slug slug lb 21.90 lb -in ? lb-in2
2
transverse axis ft4 ft? in 2 lb -in?
through CG

Launching
Trajectory entry 20-22 20-22 20.35 20.35 20.0 20.0
entry angle , deg 20.5 20.5

Angle of attack, -5.9 to -4.1 to -0.75 to -1.35 to +0.25 +0.25

deg +6.8 +4.3 +1.15 +1.15

Velocity, fps 400 400 137.5 137.5 80-81 80-81

Atmospheric Ambient Ambient Ambient Ambient 0.1 to 0.1 to

pressure, atm 1.5 1.5

Water surface Normale Normale Normal Normal Normale Normale

tension , (also
dynes/ cm 28.7 )

a
Afterbody at top of Figure 6.5 with Head F.
b
Afterbody at top of Figure 6.5 with vented head ( except Head H ).
C
Plate cylinder head model at bottom of Figure 6.5 .
d
Afterbody at top of Figure 6.5 with Head A.
e
e Normal = 75.6 dynes/cm.
=

89
 0 The experimentally obtained whip for the full
scale vented-plate head is -30.8 deg/sec. The
difference between experimental and scaled
-50 F whip is within the scatter of the data.
E Atmospheric pressure also affects the whip
of the plate-cylinder model, being about five
-100 G
C
times as great at 1.5 atmosphere as at 0.1
Head
atmosphere ( Figure 6.13 ) . It increased steadily
/s,dWhip

A
Aec eg
ė

with time until the plate was approximately half

-150 B submerged, then decreased for a millisecond or
two before increasing again.
G Photographs of cavity growth ( Figure 6.14 )
-200 reveal two interesting features . First, the size of
D
с the undercavity increases with decrease in pres
sure, as illustrated in Figure 6.15 , in which the
-250
increase in cavity size with decrease in pressure
D for comparable distances of missile penetration
-300
is apparent. Second, small protuberances are
E created along the bottom of the undercavity
Plate half
submerged
Plate sub
F
with increase in pressure. They reduce to almost
merged
-350
nothing at 0.1 atmosphere , indicating that the
0 1 2 3 4 5 6 7 8 9 10 undercavity vents , at least partially soon after
Distance traveled after water contact, in . the missile enters the water. The small differ
ences can be considered important because
Figure 6.10. Water -entry whip , plate and plate-cylinder heads. duplicate cavities resulted from the same entry
conditions ( Figure 6.16 ) . It is possible that
wetted underpressure areas may exist that are
not within the undercavity .
wetted , and the missile was fully engulfed by the
cavity soon after the CG crossed the interface. Figure 6.17 shows that change in ambient
It is interesting to compare the whips of air pressure does not significantly affect the
the vented full -scale and model heads . The CG's whip of the fully vented plate model (Head A).
are situated on the axis of symmetry of both The ballistic parameters and conditions are given
missiles , and the hydrodynamic forces on the in the last column of Table 6.1 . Compared with
plate must act normally to the plate ( except for the unvented plate-cylinder model, this configur
slight shearing forces ). Thus, the moments of the ation experienced a relatively small whip at air
pressures ranging from 0.1 to 0.5 atmosphere,
hydrodynamic forces about the CG will not be
indicating that the underpressure effect in that
influenced by the distance of the CG aft of the environment is negligible .
nose. According to the hypothesis , these fully Launchings made with the vented model
vented missiles should experience no moment were repeated with an unvented hemisphere
from underpressure and, moreover, the whip head model , using approximately the same
should also be independent of the lateral CG ballistic parameters. The negative whip of this
location. Because the whip varies as the fourth configuration , like that of the unvented plate
power of the plate diameter (Reference 97 ) , cylinder, increased greatly with atmospheric
directly as the entry velocity , and inversely as
the moment of inertia, the whip at zero pitch pressure, being +250 deg/sec at 0.1 atmosphere
and ranging to - 400 deg/sec at 1.5 atmospheres.
-

scaled from the 2 -inch model to the 22.4 -inch

prototype would be The whip at 1.5 atmospheres was not repro
ducible. A similar variance in whip was noted at
22.4141400 1.0 atmosphere , a lesser variance at 0.7 , and
21.3
none at 0.4 or 0.1 atmosphere. At the higher
( 11 37.5M
-128.2
2 / \800 X 32.2 X 144 pressure conditions, a change in angular velocity
comparable to whip sometimes occurred in the
= -33.7 deg /sec (6.1 ) yaw plane.
90
 Head A , fully vented plate cylinder

Figure 6.11a. Water entry cavities as affected by venting.

91
 C

Head E , plate 2 -inch -diameter cylinder

Figure 6.1lb . Water entry cavities as affected by venting.

92
 Head F , unvented plate cylinder

Figure 6.11c. Water entry cavities as affected by venting.

93
 Head A, fully vented plate cylinder

Figure 6.12a. Secondary water -entry cavities as affected by venting.

94
 Head E , plate 2 -inch -diameter cylinder

Figure 6.12b . Secondary water-entry cavities as affected by venting.

95
 Head F , unvented plate cylinder

Figure 6.12c. Secondary water -entry cavities as affected by venting.

96
 -480 pressures ( Figure 6.18 ). Considerable irregularity
-440
Water surface tension and lack of reproducibility were apparent during
(normal) = 76.5 dynes /cm 1.5 atm launchings at 1.0 atmosphere (Figure 6.19 ) . The
-400 stable, symmetric cavity formed at reduced
-360
pressure (0.1 atmosphere ) is illustrated in
1.0 atm Figure 6.20 .
sA/,dWhip

-320
A spherical sector (Figure 6.4 , Head H) was
eg
ec
ė

-280 0.7 atm

installed on the afterbody shown at the top of
Figure 6.5 and this missile was launched at 0.1
- 240
and at 1.0 atmosphere . Stable , reproducible
-200 0.4 atm cavities were formed at both pressure conditions
and the two cavities looked much alike
-160 ( Figure 6.21 ) .
-120
The effect of surface tension upon the
0.1 atm whip of the plate-cylinder model was also
-80
recorded (Figure 6.22 ) . Launchings were made
-40
Plate half Plate under the same conditions as those shown in
submerged submerged Figure 6.13 , except that the water surface ten
0
0 1 2 3 4 5 6 7 8 9
sion was reduced to 28.7 dynes /cm by Aerosol.2
Time from water contact , ms Comparison of the two sets of runs shows little
difference in the magnitude of the whip func
Figure 6.13 . Whip versus time , unvented plate cylinder tions, but their reproducibility in similar launch
(Head F ), at varied air pressures. ing conditions was less at the lower surface
tension .
Cavity photographs of the unvented It was concluded from these tests that
hemisphere head model at different atmospheric underpressure significantly affects the water
pressures made simultaneously with whip entry whip of both prototype service missiles
records show marked variation in cavity shape and small models and would have to be scaled if
with air pressure. Reproducible whip was asso water-entry trajectory modeling is to occur. For
ciated with a stable cavity and normal separa such an accomplishment, it would be necessary
tion . Unstable cavities , sometimes with erratic to scale the gas pressure over the water surface
separation , occurred at the higher atmospheric in the model systems .

2Less than 0.1 % by weight of Aerosol OT (Reference 2 ) ; hence not significantly affecting the water density or the
vapor pressure .

97
 1.5 atmospheres

Figure 6.14a. Cavity growth as affected by atmospheric pressures, unvented plate cylinder, Head F.
98
 잇

1.0 atmosphere

Figure 6.14b . Cavity growth as affected by atmospheric pressures, unvented plate cylinder, Head F.

99
 1

GE

0.1 atmosphere

Figure 6.14c. Cavity growth as affected by atmospheric pressures , unvented plate cylinder, Head F.

100
 1.5 atmospheres

Figure 6.15a. Enlargements of water- entry cavities of Figure 6.14.

101
 1

1.0 atmosphere

Figure 6.15b . Enlargements of water-entry cavities of Figure 6.14.

102
 0.1 atmosphere

Figure 6.15c. Enlargements of water -entry cavities of Figure 6.14.

103
 Run No. 1

Figure 6.16a. Identical cavities formed under equal entry conditions , unvented plate cylinder,
Head F , 1.0-atmosphere air pressure .

104
 Run No. 2

Figurę 6.16b . Identical cavities formed under equal entry conditions, unvented plate cylinder,
Head F , 1.0 -atmosphere air pressure.

105
 -120

Water surface tension

sWhipeg
/d,Aec
ė

( normal ) = 76.5 dynes/cm

-80

-40
Plate half submerged Plate submerged

0
0 1 2 3 4 5 6 7 8 9 10 11

Time from water contact, ms

Figure 6.17 . Whip versus time, fully vented plate ( Head A) , at varied air pressures.

106
Figure 6.18 . Unstable cavity separation , unvented hemisphere head model entering water at
1.0 atmosphere air pressure .

107
 ^ !!

bor .

Figure 6.19. Unreproducibility of unstable cavities under similar launching conditions, unvented
hemisphere head model , 1.0 -atmosphere air pressure.

108
 i !!

Figure 6.20. Stable , symmetrical cavity, unvented hemisphere head model ,

0.1 -atmosphere air pressure.

109
 !

1.0 atmosphere

Figure 6.21a. Stable cavity , fully vented spherical sector, Head H.

110
 0.1 atmosphere

Figure 6.21b . Stable cavity , fully vented spherical sector, Head H.

111
 -400

Water surface tension = 28.7 dynes/cm

1.5 atm
-360

-320
1.0 atm

-280
/s,dWhip

- 240
Aec eg
ė

0.7 atm

-200

0.4 atm
-160

-120

0.1 atm

-80

Plate submerged
-40 Plate half submerged

0
0 1 2 3 4 5 6 7 8 9
Time from water contact , ms

Figure 6.22 . Whip versus time , unvented plate cylinder (Head F ) ,

at varied air pressures and reduced surface tension .

112
 CHAPTER 7
WATER-ENTRY PITCH MODELING

Water-entry theory (Chapter 2 ) indicates Prototype and model parameters are listed
that water-entry pitch behavior may be properly in Table 7.1 . The prototypes were launched into
modeled with one-to-one Froude and cavitation fresh water at the NUC Morris Dam Torpedo
number and gas-density scaling. The experi Range and the models were launched into the
mental investigation described in this chapter NUC Variable -Angle Variable-Pressure Launch
was designed to test the theory . The oblique ing Tank. These facilities and their equipment
water-entry pitch behavior of fully cavitating are described in detail in Part V.
missiles over approximately one missile length of Models were launched into fresh water at a
water penetration was studied (Reference 96).1 nominal water-contact velocity of 119.6 fps and
One-to-one Froude scaling alone was not evalu with 1/11 atmosphere air or gas pressure . The
ated because earlier tests (Chapter 4 ) had nominal trajectory angle at water contact was
demonstrated this technique to be inadequate. - 20.5 degrees, and the accompanying angles of
Its failure was traced, at least in part, to attack and yaw were so small as to be taken as
nonmodeling of the pressure distribution in the nominally zero. Water-entry pitch and yaw data
separation area beneath the missile nose were obtained at intervals of 0.8 ms if the
immediately after water impact (underpressure change in water-entry pitch was low and 0.4 ms
effect, Chapter 6 ) . if the change was high . Simultaneous sideview
Full-scale water-entry pitch data from six camera data were also obtained, permitting
of the missiles used in the studies discussed in analysis of the missile water-entry pitch , water
Chapter 5 (Heads a, d, g, l , m
m , and n of penetration, and cavity development .
Figure 5.1 ) were compared with results from For the plate-cylinder-head model, data
2-inch-diameter models using one-to -one Froude could be recorded over only about 4 diameters
and cavitation-number scaling. Partial gas of water penetration . Different programming of
density scaling was also investigated. the launching equipment allowed the range of
The 2-inch models were geometrically and penetration to be extended to 5 or 5 1/2
dynamically similar to the prototype missiles diameters, but water-contact pitch and pitch
except that a cylindrical afterbody with a simple velocity data could not be obtained for these
tail ( Figure 7.1 ) was used because the studies launchings. Since the water-contact conditions
were to be made , at most , over one missile were nearly constant for successive launchings,
length of water penetration in which the tail did the values obtained previously were used to
not contact the cavity wall. All the heads were correct the extended-range launching data. Pro
3.6 inches long ; all the missiles were 12 inches in totype and model launching conditions are
overall length . The models had movable internal summarized below :
weights that allowed the masses , CG's, and Model
moments of inertia to be adjusted as required by Prototype
dynamic similarity . Velocity 400 fps ( nom .) 119.6 fps

1Spatial limitations on the model launching tank precluded modeling the underwater trajectory .
113
 101.4

31.25 31.75 38.40

E
18.3
22.42 22.42

Prototype

8.4

2.25 6.15

个
2.000
Model

Figure 7.1 . Experimental configurations with Head n . Dimensions are in inches. ( From Reference 96. )

Prototype Model for the purpose of assessing the modeling tech

nique and not to overcome the limitations of the
Ambient air pressure Atmospheric 1/11 atm. equipment . These corrections are derived in
Trajectory angle -20.5 to -20.9 - 20.5 deg Reference 96 .
deg The gas density could not be scaled com
Angle of attack -1.6 to +1.4 -0.20 to pletely (one to one ) because aа sufficiently heavy
deg +0.55 deg
gas was not available ; a partial scaling of 0.7 to
Yaw angle +2.3 deg +0.5 deg 0.8 was achieved .
Freon 114B2 dibromotetrafluoroethane)
Table 7.2 gives further data ; the angles (Reference 67 ) with a density nine times greater
were defined in Chapter 1 ( Figure 1.1 ). The than that of air under the same conditions of
effect of yaw on the water-entry pitch behavior temperature and pressure, was used to achieve
appears to be small, and this parameter was not partial gas-density scaling. Were it not for the
considered in the modeling studies . presence of water vapor in the tank,, it should be
It was not always possible to scale proto theoretically possible to obtain a gas-density
type water-contact velocity and pitch velocity coefficient of 9/11 at 1/11 atmosphere. From
accurately, and corrections for scaling discrep Dalton's law of partial pressures , the gas-density
ancies were made in order to compare model coefficient for equilibrium gas-water-vapor mix
and prototype data, the corrections being made tures at 1/11 atmosphere in the presence of
114
 . rototype
7.1
TABLE
Pand
Parameters
Model

Distance Inertia
of
Moment
Diame
, ter ,
Length Ma
, ss from
Nose about
Transverse Contour
Tolerance
Parameter .
in .
in lb. ,i. n
CG
to CG
through
Axis Radius
.on
in

dHead
-,3.5
iameter
ogive
( igure
F)7.2
+0.020 +5slug
745
ft2
Prototype 22.42 1+154
/8 +|7/4
/2
1.1
11,560–1,565 0.040
-0.010
Model
Froude
scaling
by
Required 2.0000 1.108 6.343 lb
2in 51
19.
used
Actually +
2.0000
0.0002 12.0 0.005
+
1.112 6.35
0
+ .01 2lb
in
+ .1564
019. +0.000
0
-.001

-d1Head
a,.5
ogiveiameter
)( igure
F7.3
+0.020 1,4958–1,49671.1 ft2lug
s5756
Prototype 22.42 1+154
/8 +1/2 1+ /4 0.040
-0.010
Model
scaling
Froude
by
Required 2.0000 1.061 lb
219.80
in6.343
used
Actually 0.0002
+
2.0000 12.0 0.005
+
1.096 lbin29.99
+| .15
.01
016.35 0.001
–
+0.000

h,gHead

115
emisphere
)(Figure
7.4
+0.020 ft2
5lug
s771
Prototype 22.42 1154
+/8 11,526
=/2 169.5
+/4 0.040
-0.010
Model
scaling
Froude
by
Required 2.0000 1.083 inlb
220.19
6.200
used
Actually +0.0002
2.0000 12.0 0.005
+
1.084 in .15
0.12
2 .01
2/6.20
+0lb 0.001
–
+0.000

l late
0.25 iameter
dp-Head
(F7.5
)ogive
igure
+0.020 ft2 lug
s5+743
Prototype 22.42 +/8
1,583–1,58
1=154
/2 +/4
6163.7 0.040
-0.010
Model
scaling
Froude
by
Required 2.0000 1.124 5.683 in
lb
2 46
19.
used
Actually 2.0000
0
+ .0002 12.0 0.005
+
1.130 0.01
+
5.68 in2
lb
+ .1546
019. 0.001
–
+0.000
 7.1
TABLE
rototype
.Pontinued
(and
Parameters
c)Model

Distance Inertia
of
Moment
Diamet
, er ,
Length Mass
, from
Nose about
Transverse Tolerance
Contour
Parameter .
in .
in lb. .in
CG
to Axis
through
CG Radius
oin
.n

, late
m
dp0.1
-Head
iameter
7.6
)( igure
Fogive
+0.020 ft2
5+lug
s694
Prototype 22.42 |144,544–1,545
+163.7
/8
/4 0.040
-0.010
Model
scaling
Froude
by
Required 2.0000 1.097 5.683 2in
lb
18.18
used
Actually +0.0002
2.0000 12.0 0
+ .005
1.103 0.01
+
5.68
2lb
in
+.157
018.3 0.001
-
+0.000

pHead
,nlate
cylinder
)(F7.7
igure
+0.020 ?5lug
fts694
Prototype 22.42 | +1|144
6.37
/8
/2
/4
,551-1,552
-0.010
Model
scaling
Froude
by
Required 2.0009 1.100 5.683 in2lb
18.18
used
Actually 0.0002
+
2.0000 12.0 0.005
+
1.109 0.01
+
5.68 in2
lb
18. 35
0+ .15

a1,480
for
lb
prototype

116
а launching
.one
 TABLE 7.2. Launching Conditions, One-to-One Froude and Cavitation -Number
Scaling ( F , o Model Data)

Prototype Model

Trajectory Angle of Trajectory Angle of

Velocity, Angle , Attack , Velocity, Angle, Attack ,
fps deg deg fps deg deg
Head d, 3.5 -Diam. Ogive ( Fig. 7.2 )

416 -20.6 -1.4 117.5 -20.30 0.42

418 -20.7 -0.3 119.2 -20.30 0.22
416 -20.7 0.7 118.9 -20.30 0.34
419 -20.7 -1.0 120.6 -20.30 0.19

Head a, 1.5-Diam. Ogive ( Fig . 7.3)

446 -20.7 -0.2 120.3 -20.30 0.47

430 -20.9 0.2 120.3 -20.30 0.55
425 -20.8 0.8 120.5 -20.30 0.53
430 -20.7 -0.5 121.8 -20.30 0.37

Head g, Hemisphere ( Fig. 7.4)

415 -20.7 -0.2 120.8 -20.30 0.07

411 -20.7. -0.6 120.2 -20.30 0.01

412 -20.6 0.2 119.0 -20.30 0.15
118.2 -20.30 -0.20

Head l , Plate 0.25-Diam. Ogive ( Fig. 7.5 )

430 -20.5 -0.2 119.8 -20.30 0.17

428 -20.6 0.2 120.5 -20.30 0.13

417 -20.7 -0.9 120.6 -20.30 0.16

421 -20.6 0.2 120.9 -20.30 0.07

Head m , Plate 0.1 -Diam. Ogive ( Fig. 7.6 )

394 -20.9 -0.9 122.4 -20.30 -0.17

398 -20.8 1.4 119.6 -20.30 0.23
398 -20.8 0.9 118.9 -20.30 0.41

119.3 -20.30 0.33

117
 TABLE 7.2 . Launching Conditions, One-to-One Froude and Cavitation -Number
Scaling ( F , o Model Data) (Continued)

Protype Model

Trajectory Angle of Trajectory Angle of

Velocity, Angle, Attack , Velocity, Angle, Attack ,
fps deg deg fps deg deg

Head n, Plate-Cylinder ( Fig. 7.7 )

400 -20.8 1.0 117.8 -20.30 -0.04

401 -20.8 0.4 120.6 -20.30 -0.18

403 -20.8 1.4 121.0 -20.30 -0.06
401 -20.8 -1.6 119.7 -20.30 -0.13
119.6 -20.20 ( a)

)
a
(
120.0 -20.20
122.1 -20.20 ( a)

)
a
(
122.4 -20.20

a
Angle of attack not known but probably between -0.5 and 0.5 deg.

water is 0.69 at 20°C and 0.79 at 10 °C ;2 an five of the six head configurations. For the
appreciable gain in partial scaling could be sixth , plate 0.1 -diameter ogive head
the plate
realized by chilling the water . Therefore , the (Head m ), the model's change in nose -down
tank water was chilled to 10 °C for the launch pitch was about half that of its prototype. A
ings that involved gas-density scaling. Experi smaller whip is experienced by this head, and
mental gas -density coefficients differed from the consequently the pitch -inducing force, which is
theoretical value because of air leakage and the resultant of hydrodynamic and underpres
because of difficulties in metering the heavy gas sure forces , must be smaller. The hydrodynamic
and controlling tank pressure . component of the pitch-inducing force on the
Prototype data (Chapter 5 ) were corrected Head m configuration is also comparatively
to a nominal water -contact velocity of 400 fps. small and in the same direction as the underpres
Model data involved such small deviations in sure force ( Chapter 5 ) . This head would thus
water - contact velocities that no correction was naturally be more sensitive to non-modeling of
made from the scaled nominal value ( 119.6 fps ). the underpressure force because that force
Both model and prototype water-entry pitch contributes a more significant fraction of the
data were corrected to zero water- contact pitch total whip .
velocity and are shown in Figure 7.2 to 7.7 as The dispersion of the model data curves
the change in pitch after water contact. The also indicates the sensitivity of Head m , and this
model and prototype time scales are in the ratio conclusion seems to be substantiated further
of 1 : 711.2 . With one-to-one Froude and when gas-density scaling is added.
cavitation-number scaling (F, o model data ), One-to -one Froude and cavitation -number
water- entry pitch correlation to within the and gas-density scaling (F , 0 , p model data ) was
accuracy of the prototype data was obtained for used for three configurations – Heads g, m , and

2Water-vaporpressures are 17.54 and 9.20 torrs at 20°C and 10°C, respectively.
118
 2.8 2.8

Head d , 3.5-diam ogive head Head a, 1.5 - diam

2.6 2.6
ogive head

2.4 20.5° 2.4

2.2 2.2

2.0 2.0 20.5°

Change
Change

1.8

pitch
1.8
pitch

,dineg
,deg
in

77
1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0
0

0
1

5
2 3
Trajectory angle ,
deg
Angle of attack,
deg
Velocity , fps
Gas pressure , atm

4 5

Prototype time from water contact , ms

1.0 2.0 3.0

Model and prototype water penetration ,

diameters
6
Prototype data
F, o model data

Model time from water contact , ms

1
10
1
15 20

4.0

Figure 7.2 . Prototype and model water-entry pitch data ,

Head d . ( From Reference 96. )
Proto
type

7
20.5

5.0
400

25
8
Model

20.5

120
1/11

30
10
1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0
1

1.0
2 3

10

2.0
Trajectory angle,
deg
Angle of attack,
deg
Velocity , fps
Gas pressure , atm

3.0
15
5

Model and prototype water penetration,

diameters
20

Prototype time from water contact , ms

6

4.0

Figure 7.3 . Prototype and model water-entry pitch data,

Prototype data
F, o model data

Proto
type Model

20.5

7
Model time from water contact , ms
0
400
1

5.0
25
8
20.5

120
0

1/11

30
10

Head a. ( From Reference 96. )

n (Table 7.3 and dot-dash curves on Figures 7.4, was small unless gas-density scaling was also
7.6 , and 7.7 ) . Water-entry correlation to within used .
the accuracy of the prototype data was obtained Model and prototype water penetration
for all heads when 0.7 gas -density scaling was was measured, and plotted (in diameters ) for
used. It is significant (although expected on the comparison with the respective model and pro
basis of theory ) that modeling of the Head m totype time scales . In no case did the penetra
configuration was obtained without any reduc tion data for corresponding scale times deviate
tion in the correlation obtained for the other more than 7% in a penetration range of about
two configurations in the absence of gas-density 5 diameters . Missile penetration is indicated in
scaling. Figures 7.2 to 7.7 by aa third scale (the other two
The tests indicate that water-entry pitch being model and prototype times ) constructed
modeling with one-to-one Froude and cavitation by plotting the averages of the model and
scaling was adequate for the head configurations prototype penetration in diameters for corre
that showed considerable change in water-entry sponding scaled times. The deviation of experi
pitch , but correlation failed when the change mental data from the averaged penetration scale
119
 2.8 2.8

Head g, hemisphere Head l , plate 0.25 -diam ogive head

2.6 2.6

2.4 2.4
20.5°

2.2 2.2

2.0 20.5 ° 2.0

dChange
Change

Proto

pitch
1.8
pitch

1.8 Model

,ineg
type
Prototype data
,dineg

F, o model data Trajectory angle,

1.6 1.6 deg 20.5 20.5
F, 0, p model
data
Angle of attack ,
1.4 1.4 deg 0 0
Velocity, fps 400 120
1.2 1.2 Gas pressure , atm 1 1/11

1.0 1.0
1

0.8 0.8
Proto
type Model
0.6 0.6
Trajectory angle ,
deg 20.5 20.5
0.4 Angle of attack, 0.4
deg 0 0

0.2 Velocity, fps 400 120 0.2

Gas pressure , atm 1 1/11 Prototype data
0 0 F, o model data
0 1 2 3 4 5 6 7 8 9 10
-0.2
Model time from water contact , ms 0 1 2 3 4 5 6 7 8 9 10

Model time from water contact , ms

0 5 10 15 20 25 30

Prototype time from water contact, ms 0 5 10 15 20 25 30

Prototype time from water contact, ms
0 10 2.0 3.0 4.0 5.0
0 1.0 2.0 3.0 4.0 5.0
Model and prototype water penetration ,
diameters Model and prototype water penetration ,
diameters
Figure 7.4. Prototype and model water- entry pitch data,
Head g. ( From Reference 96. ) Figure 7.5 . Prototype and model water-entry pitch data ,
Head l . ( From Reference 96. )

is , at most, 3.5 %. The data for extended-range

launchings made with the plate-cylinder head are It is of interest to compare the change in
shown in Figure 7.7 for a range of 4 to about model pitch resulting from the addition of
5.5 diameters of water penetration . gas-density scaling. With Head m , pitch correla
One-to-one Froude and cavitation-number tion did not occur without gas density scaling.
and gas-density scaling models some of the Head n's pitch change became slightly more
initial boundary conditions that are necessary in negative but still lay within the scatter of the
order that the subsequent underwater trajectory prototype data ; that of Head g was little altered
can be modeled ; other studies (Chapters 8 and except for becoming slightly more positive
9 ) indicate that this technique models water during the first diameter of underwater travel.
entry cavities for at least 20 diameters of under The exact mechanism by which gas density
water Thus,, this scaling technique
flight. Thus alters pitch is unknown beyond the fact that
simultaneously models the water-entry and change in gas density probably affects the
underwater trajectory of a cavity-running undercavity separation . Extrapolation of cavity
missile. studies (Chapter 9 ) suggests that the undercavity
120
 0.2 0

0 -0.2

-0.2 -0.4

-0.4 -0.6

-0.6 -0.8 Prototype data

Proto F , o model data

Change
type Model + F, o , p model data

pitch
-1.0
Change

-0.8

,deg
Trajectory angle ,
pitch

in
deg 20.5 20.5
,deg

-1.0 -1.2
in

Angle of attack , Prototype Model

deg 0 0 Trajectory angle,
-1.2 400 120 -1.4 20.5
Velocity , fps deg 20.5
Gas pressure, atm 1 1/11 Angle of attack ,
-1.4 -1.6 deg 0 0
Velocity , fps 400 120
-1.6 -1.8 Gas pressure , atm 1 1/11
Head m , plate 0.1 -diam ogive head
-1.8 -2.0 Head n , plate cylinder
20.5
-2.0 -2.2
20.50
-2.2 -2.4

-2.4 -2.6
Prototype data
F , o model data -2.8
-2.6
++ F , 0 , p model data. 0 1 2 3 4 5 6 7 8 9 10
-2.8
Model time from water contact, ms
0 1 2 3 4 5 6 7 8 9 10

Model time from water contact , ms

0 5 10 15 20 25 30

0 5 10 15 20 25 30 Prototype time from water contact, ms

Prototype time from water contact, ms
0 1.0 2.0 3.0 4.0 5.0

0 1.0 2.0 3.0 4.0 5.0 Model and prototype water penetration,
diameters
Model and prototype water penetration,
diameters
Figure 7.7 . Prototype and model water-entry pitch data,
Figure 7.6. Prototype and model water-entry pitch data , Head n . ( From Reference 96. )
Head m. ( From Reference 96. )

pressure of 0.1 atmosphere was maintained

might close earlier, be smaller, or last longer, but during all launchings. The pitch change of the
those suppositions cannot be verified because model in the helium atmosphere was just per
the cavity photographs taken simultaneously ceptibly more positive than that of the model in
with the pitch records lack sufficient detail. air ; the change in the gas density coefficient
Three launchings were made with the 6 ' = 0.1 to p '== 0.014 ) was insignificant. How
=

0.1 -diameter plate ogive (Head m ) missile to ever, the pitch data from the helium launchings
investigate the effect of decreasing the gas showed more scatter, suggesting that the under
density in the model system. An atmosphere of cavity was less stable in helium. The relationship
helium (density approximately one seventh that between pitch change and undercavity forma
of air) was substituted for air. An ambient tion warrants further investigation .

121
 TABLE 7.3. Model Launching Conditions,
One-to-One Froude and Cavitation
Number and Gas-Density Scaling
(F , 0 , p Model Data)a

Angle of
Velocity , Trajectory Attack , Gas- Density
fps Angle, deg deg Ratio

Head g, Hemisphere ( Fig. 7.4)

119.8 -20.30 -0.33 0.75

119.8 -20.30 -0.41 0.76

119.9 -20.30 -0.49 0.68

122.9 -20.30 -0.40 0.65

Head m , Plate 0.1 -Diam. Ogive ( Fig. 7.6 )

122.2 -20.30 -0.62 0.59

121.9 -20.30 -0.67 0.80

122.1 -20.30 -0.62 0.84

121.7 -20.30 -0.80 0.82

Head n, Plate-Cylinder ( Fig. 7.7 )

120.8 -20.30 -0.82 0.71

120.3 -20.30 -0.80 0.82

120.9 -20.30 -0.90 0.80

119.6 -20.30 -0.83 0.82

a Prototype data are in Table 7.2

122
 CHAPTER 8
UNDERWATER TRAJECTORY MODELING

Water-entry pitch modeling, discussed in sensitive to small secondary forces. Prototype

the previous chapter, proved feasible, auto and model missiles are shown in Figure 8.1 and
matically satisfying the initial conditions for their physical properties listed in Table 8.1 . The
modeling the subsequent missile underwater model was scaled in accordance with the Froude
trajectory. It thus becomes necessary to deter law to be dynamically and geometrically similar
mine to what extent additional conditions such to the prototype .
as cavity behavior and the character of the flow The prototype was launched from the NUC
( laminar or turbulent ) affect model behavior. It Variable -Angle Launcher at nominal entry veloc
is evident that the model must be completely ities of 200 and 400 fps. The model was
geometrically scaled if the missile underwater launched in the CIT Controlled-Atmosphere
trajectory is to be modeled. Launching Tank at velocities of 60 and 120 fps,
The experimental program detailed in this corresponding to Froude scaling of the proto
chapter was specifically planned to test the type, and at 60 fps (scaled ) and 80 fps (non
effect of one-to-one Froude scaling in conjunc scaled ) in the NUC Variable -Angle Variable
tion with varying prototype velocity , cavitation Pressure Launching Tank . All tests were made
number , boundary-layer condition, and water with a nominal air trajectory angle of
surface tension in water -entry modeling (Refer 22 degrees. The angle of attack of the prototype
ence 110 ) . Less attention was paid to the details at water entry varied between 0 and +6 degrees
of water-entry behavior and more emphasis was (nose up with respect to the trajectory ) ; the
placed on the underwater trajectory and angle of attack of the model varied between - 3 -

phenomena related to it. Launching conditions (nose down ) and +6 degrees (nose up ). All
and missile shape were selected to include prototype tests were made at full atmospheric
regions where modeling failures might be pressure ; the model tests were made at air
expected (References 36 , 37, and 38 ) , and the pressures of 1 , 1/11 , and 1/21 atmosphere. For
range of entry attack angles was wider than any several of the launchings, the nose of the model
that had been used before . was artificially roughened with grains of sand.
Gas-density scaling was not included . Its Five launchings were made in the smaller VAVP
importance was not known at the time , and the launching tank , with the surface tension reduced
large volume of the CIT Controlled-Atmosphere to 40 dynes /cm . Entry conditions of the
Launching Tank, in which model water-entry individual launchings are shown in Tables 8.2 ,
and underwater trajectory tests were conducted, 8.3 , and 8.4 .
would have made such scaling difficult. Initial tests demonstrated the significance
Head f, the 3.5-diameter 70 -degree of boundary-layer flow on water-entry behavior,
spherogive shown in Figure 5.1 , was chosen for a factor previously hoped to be negligible since
this study because it had a fine nose shape that only a small portion of the missile surface is
was known to be extremely sensitive to variation wetted during the cavity phase of the trajectory.
in entry condition. As discussed in Chapter 5 , It was found that Froude and cavitation-number
water-tunnel tests indicated that cavity separa modeling did not result in correlation without a
tion might occur either on the spherical portion turbulent boundary layer. Both the Reynolds
or on the ogive, tending to make this shape number and the roughness were sufficient to
123
 +

2-inch-diameter model

1
WT
ch

22.42- inch - diameter prototype

0.285
3.500R 70°

Ti
1.000 1.000
CG
1.000
0.305R

I 1.399
3.174
4.016
6.036
6.470
6.827

Dimensions in diameters

Figure 8.1 . Model and prototype with a 3.5-inch-diameter , 70 -degree spherogive nose .

124
 TABLE 8.1 . Model and Prototype Dimensions

Model

Required by Actually
Parameter Prototype a Froude Scaling Used

Diameter , in. 22.42 2.000 2.001

+0.020
-0.010 +0.0002

Length , in. 152-1 / 8 13.654 13.645

+1/8 +0.005

Weight, lb between 1640 1.186 1.190

and 1675 +0.005
+1/2

Distance from between 71 and 6.522 6.514

CG to nose tip, in. 73-1 / 8 +0.02
# 1 /4

Moment of inertia between 770.9 0.1436 0.1436

about transverse and 798.1 lb ft2 +0.001 lb ft2
axis through CG +5 slug ft2
Contour Tolerance , +0.040 +0.001
in . on radius

Surface roughness, 40 – 60 12
u in .

a Internal mechanism not the same for all prototype launchings; weight , CG location , and
moment of inertia therefore varied slightly.

insure a turbulent boundary layer on the proto required (0.030 = 0.002 inch ) was estimated
type, but the model surface was smooth and the from the relationship (Reference 74, page 40 ) :
Reynolds number relatively low , the transition
from laminar to turbulent boundary layer 8 vl )-1/5
occurring at some velocity between 60 and 0.37
120 fps, probably above 80 fps. Consequently , a
& (3) V
= 0.37 (Re ) -1/5 ( 8.1 )

turbulent boundary layer was induced artifi

cially for the 60 -fps launching velocity by where for a smooth , flat plate at zero incidence,
roughening the surface of the model nose with s is the boundary -layer thickness in turbulent
grains of sand cemented to it. The degree of flow , l is the length of the plate , and Re is the
surface roughness cannot be scaled from that of Reynolds number related to the length of the
the prototype because boundary -layer thickness plate.
is a function of Reynolds number and cannot be Equal Froude and cavitation-number
scaled as an ordinary linear dimension of a scaling produced valid correlation when a turbu
Froude-scaled system . The sand-grain diameter lent boundary layer was artificially induced ; the
125
 TABLE 8.2. Prototype Launchings, TABLE 8.3. Model Launchings, Variable-Angle
.

Variable -Angle Launcher a Variable-Pressure Launching Tank a

Air Trajectory Entry Entry Entry Air Roughness, Surface

Angle , Pitch Angle, Velocity, Velocity, fps Pressure, Type Tension,
deg deg fps +1/2 fps atm ( Figure 8.3) dynes /cm
20.2 -0.7 414 80.8 1/11 B 76
81.2 1/11 76
23.0 +1.1 200 (nom )

:
81.4 1 76
22.8 +1.0 193 61.4 1 76
22.5 -0.3 197 61.2 1/11 76

:
20.0 -0.1 200 (nom)
61.4 1/11 B 76

da
en:
20.0 +0.1 195 61.8 1 B 76
22.5 +0.8 201 80.8 1 B 76
22.2 +0.4 202 60 1/11 76
22.2 +2.1 200 (nom)
61.2

..
22.2 +1.7 202 1/11 76

61.7 1 76
22.1 +2.2 207
61.4 1 B 76
22.2 +1.6 209
61.7 1 B 76
21.6 +1.9 213
61.4 1 B 76
21.5 +0.3 206
81.0 1 B 76
22.6 +2.8 204
81.2 1 B 76
22.7 +2.7 198 81.7 1 B 76
23.7 +2.5 183 81.0 1/11 76
23.7 +2.4 198 80.7 1 76
23.5 +5.8 189 81.0 1/11 B 76
:oo

23.3 +3.0 187

81.3 1/11 B 76
23.2 +2.7 214 61.2 1/11 B 76
20.0 +5.9 194 61.4 1/11 B 76
61.5 1 B 39.8
a 60 1 B 39.8
Angles measured to 1 deg,
velocity measured to + 5 fps (nom )
62.4 1 B 41.8
62.1 1 . B 39.2
smooth model did not reproduce prototype
behavior. When the entry velocity of the model 61.7 1/11 B 40.1
was increased until a turbulent boundary layer
a
occurred naturally , the presence of the rough Nominal entry angles: 30-deg air
ness made no difference in model behavior; the trajectory , -0.5-deg entry pitch.
model with and without roughness reproduced Actual entry angles, +0.5 deg of
the prototype trajectory . nominal angles.
126
 TABLE 8.4. Model Launchings, Controlled Atmosphere Launching Tank
Entry Entry
Air Angle of Tolerances Entry Pitch Air Roughness,
Trajectory Attack, on Angles, Velocity , Velocity, Pressure, Type
Angle, deg deg deg fps +1/2 fps deg/sec atm ( Figure 8.3 )

1
23.3 +2.2 59.3 11

***
22.4 +1.3 1 61.3 -5 1/11
22.7 +0.2 1 60.3 -5 to -10 1/11
22.4 +0.2 1 60.4 +10 1/11
23.1 +0.4 1 59.6 +20 1/11

22.3 -2.4 1 60.0 -10 1/11

22.6 -1.0 1 60.5 -10 1/11
1

22.4 +0.2 61.0 +15 1/11

22.2 +1.4 1 60.6 +10 1/11
22.5 -12.5 1 60.6 +25 1/11

:
1

22.3 +2.1 60.2 0 1/11

21.4 +3.5 1 61.1 +10 1/11
21.3 +4.2 1 60.6 -10 1/11
21.2 +4.2 1/2 60.1 -5 1/11
21.3 +5.0 1/4 60.1 0 1/11
20.8 +1.0 1/2 60.6 0 1/21
20.8 +1.1 1/2 60.9 +10 1/21
21.3 +5.4 1/4 60.3 -5 1

20.6 +4.4 1/4 60.6 -10 to -15 1

20.6 +2.6 1/4 60.0 0 1

1

20.6 +0.8 1/2 60.6 0

21.1 +2.1 1 60 ( nom ) 1/11 A

20.9 +1.9 1/2 60.1 0 А

1/11
21.0 +1.2 1/4 60.8 +5 1/11 B

22 ( nom ) +1.5 60 ( nom) 1/11 B

(approx)
20.8 +1.6 59.9 +10 B
1/2 1/11
20.0 0.0 1/4 120.2 -10 1/11
19.9 -0.3 1/4 120.0 -10 to -15 1/11
17.7 -2.9 1/4 122.4 -25 1/11
21.0 +1.5 1/2 60.0 -10 1/11 B

127
 TABLE 8.4. Model Launchings, Controlled Atmosphere Launching Tank ( continued)
Entry Entry
Air Angle of Tolerances Entry Pitch Air Roughness,
Trajectory Attack , on Angles, Velocity , Velocity, Pressure , Type
1
deg deg /sec atm
Angle, deg deg fps +1/2 fps ( Figure 8.3)

21.0 +1.6 1/2 60.4 -10 1/11 B

21.2 +5.4 1/2 60.2 -10 1/11 B

21.2 -0.5 1/2 120.5 1/11 B

21.1 -1.0 1 120.0 -30 1/11 B

21.6 +6.2 1/4 60.7 -5 1/11 B

21.3 -0.2 1/2 119.8 -10 1/11 B

21.5 -0.4 1/4 121.0 -5 1/11

21.5 -0.6 1/2 120.6 -10 1/11

:On
21.6 +1.9 1/4 60.2 +5 1/11
21.5 +5.4 1/2 60.3 -15 1/11 D

21.1 +5.2 1/2 60.1 -10 1 D

21.3 +5.2 1/4 60.2 -25 1/11 E

19.0 -1.0 1 60.8 -15 1 B

1 1

19.9 +0.1 60.2 -20 1 B

19.0 -0.9 60.2 -15 1 B.

19.1 -1.0 1 60.8 -15 1 B

20.1 -0.2 1/2 120.5 -15 1 B

20.2 -0.1 1/2 121.0 -20 1 B

...

20.0 -0.1 1/2 120.9 1 B

Because of the uncertainties associated model ceases to approximate the behavior of the
with causing a turbulent boundary layer artifi prototype after 35 diameters of underwater
cially , it is preferable to conduct modeling travel. Furthermore, the prototype trajectories
experiments at entry velocities that will cause are concave upward ; the model trajectories are
the turbulent boundary layer to occur naturally. convex .

Entry velocities : 200 fps (prototype) and The effect on trajectory of artificially
60 fps (model). Figure 8.2 compares the traject inducing a turbulent boundary layer was inves
ories of the smooth model at air pressures of tigated by roughening the model noses in several
1 , 1/11 , and 1/21 atmosphere with the proto different ways ( Figure 8.3 ). Figure 8.4 compares
type trajectories during launchings having several prototype trajectories with those of
approximately the same entry pitch angles. 1 / 11 -atmosphere models having variously
Although the model trajectories occurring at roughened noses ; entry angles of attack were
1/11 atmosphere ( equal cavitation number in comparable in the range from -0.3 to
model and prototype systems ) are in best +2.8 degrees. Roughness A did not alter the
agreement with those of the prototype, the model trajectory appreciably . The increased area
128
 Water surface
, iameters
0
dDepth

Model: 2 launchings, 1/21 atm, + 1 °

Prototype: 9 launchings,
entry attack angles + 1.7 °
20

to + 2.8°
Model : 2 launchings,
1 atm, + 0.8º and + 2.2°

40 Model : 3 launchings, 1/11 atm , + 1.3 ° to + 2.5°

1 1
0 20 40 60 80 100 120 140

Distance from water entry , diameters

Figure 8.2 . Underwater trajectories . Entry velocities 60 fps for Froude-scaled

smooth- nosed model, 200 fps for prototype.

Type A Type A Type B

A. Circular cap of sand on spherical tip

B. Larger circular cap of sand on spherical tip

C. Two rings of sand on ogive

D. Three rings of sand on ogive

E. Three rings of sand plus band of grains placed at random on ogive

Type C , D and E

Figure 8.3 . Location of roughness ( 0.030-inch -diameter regularly spaced grains of sand )
applied to model nose.

129
 , iameters
Water surface

dDepth
0
Model : roughness type C,
1 launching, + 1.9°
Prototype : 8 launchings, entry
attack angles + 1.7 ° to + 2.8°
20

Model : roughness type A,

-2 launchings, +1.90 and + 2.1 ° Prototype: 3 launchings, -0.3°
to + 2.2°
Prototype : 1 launching, -0.1°
40 1
0 20 40 60 80 100 120 140

Distance from entry , diameters

Roughness types A and C

, iameters

Water surface
dDepth

0
Model: roughness type B, 6
0
launchings, + 1.2 ° to + 1.6°
Prototype : 88 launchings,
entry attack angles + 1.7 °
to 2.8°
20
-

Prototype: 3 launchings,
Prototype: 1 launching, -0.1 ° -0.3 ° to + 2.2°
1 1
40
0 20 40 60 80 100 120 140
Distance from entry , diameters

Roughness type B

Figure 8.4 . Underwater trajectories. Entry velocities 60 fps for Froude- scaled
roughened model , 200 fps for prototype .

of roughness Bresulted in model trajectories to 80 fps (Figure 8.6 ) . Roughnesses A and did
that curved slightly upward, and, within the not affect the drag as long as the model
accuracy of the prototype data, reproduced the remained in the cavity. Roughnesses C, D , and E
prototype trajectory. Roughness C did not increased the drag (Figure 8.7 ) . Thus, the rough
permit valid modeling. Data on roughnesses D ness that produced valid modeling (type B ) and
and E were not obtained in this range . the roughness that failed to perturb the trajec
Although no large differences occurred in tory of the smooth model (type A) did not alter
the cavities of the smooth model and in those of the drag of the model.
the models with various roughnesses , detailed When the entry attack angles were
photographs of the cavities show that the increased to +5.2 to +6.2 degrees , the model
bottom of the cavity made by the smooth model trajectories curved downward, the types B- and
was scalloped and the cavity of the roughened D -roughened models deviating from the proto
(type B ) model was not (Figure 8.5 ). The scal type trajectory after 40 diameters of underwater
loping was probably caused by an irregular travel and the type E -roughened model deviating
fluctuating separation of the flow from the after 60 diameters (Figure 8.8 ).
bottom of the model ; failure to model flow The failure of the roughened model to
separation could cause the observed discrepancy reproduce the prototype trajectories having large
between the prototype and the smooth model in nose-up entry pitches may be attributed to
the Froude and cavitation -number -scaled several factors. First, it was necessary to induce
system . The scalloping almost disappeared when a nose-up pitch velocity of about 50 deg/sec
the velocity of the smooth model was increased during the prototype airflight in order to obtain
130
 Smooth model

Model with type B roughness

Figure 8.5 . Cavities of smooth and roughened models compared 47 ms after

water entry . Entry velocity 60 fps , air pressure 1/11 atmosphere.

the flat entry attack angles, but the model had a greater influence of nose underpressure at
nose-down pitch velocity of 10 to 21 deg /sec. reduced velocity .
The behavior of this sensitive nose shape is likely Third , at the greater nose-up angles of
to be affected by that amount of difference in attack, where separation occurs on the ogive , the
entry pitch velocity . separation angle is small, making this region
Second , it is difficult to induce aa turbulent particularly sensitive to scaling parameters. It is
layer artificially without significantly altering possible that the correlation failure was caused
some other aspect of the flow about the missile.. by improper cavity-separation modeling due to
When the entry angle of attack was so flat that non-scaling of gas density .
the ogival portion of the nose entered the water The pitch of the torpedo with respect to
before the spherical tip , satisfactory correlation the horizontal was recorded as a function of
could not be induced with either roughness on time during several of the prototype launchings .
the tip or roughness on the ogive. Sand on the The only comparable model launching
nose tip apparently failed to trigger a turbulent ( 1 / 11 -atmosphere air pressure, type Brough
boundary layer, and sand on the ogive increased ness ) was compared to the prototype results.
the drag, thereby decreasing the model velocity The model reproduced the orientation of the
more rapidly. The lower velocity caused the prototype as well as its path ( Figure 8.9 ).
boundary layer flow to revert to laminar flow Several tests were made with the type B
earlier in the trajectory and also produced a roughened-nose model at full atmospheric pres
steeper trajectory, probably because of the sure and O -degree entry attack angle to see
131
 .

Smooth model
1

Model with type B roughness

Figure 8.6 . Cavities of smooth and roughened models compared 38 ms after water entry . Entry
velocity 80 fps, air pressure 1/11 atmosphere.

whether Froude scaling alone would be occurred ( Figure 8.11 ) . The early loss of the
adequate. During half of the launchings , the cavity was associated with a marked decrease in
entry cavity stripped from the model and trailed drag (Figure 8.12 ) , but as long as the cavity did
from the shroud ring after only 1 1/2 or not strip from the roughened model, the rough
2 lengths of underwater travel (Figure 8.10 ). ness made no detectable difference in the cavity
The model trajectory deviated sharply from that at
full atmospheric pressure (Figure 8.13 ).
of the prototype when the model remained in its Evidently , Froude scaling alone is inadequate.
cavity and it was erratic when early stripping
132
 80

, iameters
dDistance
3 Launchings, smooth model
entry
from 2 Launchings, roughness type A
60 2 Launchings, roughness type B

1
Entry attack angles + 1.4° to + 2.5°

40 1 Launching, roughness type C

1 Launching, roughness type D
Entry attack angles +1.90 and +5.4
20

1 Launching, roughness type E ,

1

entry attack angle + 5.2°

0 1 1

0 0.04 0.08 0.12 0.16 0.20 0.24 0.28 0.32 0.36

Time from entry , sec

Figure 8.7 . Effect of model roughness on drag. Entry velocity 60 fps, air pressure 1/11 atmosphere .
,diameters

Water surface
Depth

0
+ 5.9°
Entry attack angle + 5.8°
Roughness type E, + 5.2°
20

Roughness type B, +5.4

Roughness type B , +6.2
Roughness type D, + 5.4 ° Prototype
40
Model, 1/11 atm

I 20 40 60 80 100 120 140

Distance from entry , diameters

Figure 8.8 . Effect on trajectory of increasing entry angle of attack. Entry

velocity 60 fps for Froude-scaled roughened models, 200 fps for prototype .
upse
No

40
Pitch
dangle
, eg

0
down
Nose

Prototype
Model , 1/11 atm , roughness type B ,
40 entry attack angle + 1.6 °

0 0.04 0.08 0.12 0.16 0.20 0.24 0.28 0.32 0.36

Time from entry , sec ( model )

Figure 8.9 . Pitch of model and prototype compared as a function of time . Entry velocity
60 fps for Froude-scaled model , 200 fps for prototype . Equal cavitation numbers.
133
Figure 8.10. Entry cavity stripping from model early in
underwater travel. Entry velocity 60 fps , air pressure 1
atmosphere , roughness type B.

134
 Water surface

,diameters
0

Depth
Entry attack angle + 0.4 °

+ 0.3 °
-0.3 °
20

-1.0° , no cavity

-0.1 ° + 0.1 °
T

+ 0.1° , no cavity Prototype

40 Model, 1 atm, roughness
2 Launchings, -1.0° and -0.9 °, cavity type B

0 20 40 60 80 100 120 140

Distance from entry , diameters

Figure 8.11 . Effect of early cavity stripping on model underwater trajectory . Entry velocity
60 fps for Froude-scaled model, 200 fps for prototype. Air pressure 1 atmosphere .
, iameters

80
dDistance
entry
from

60 Model without cavity,

U

2 launchings
-

40

Model with cavity , 1 launching

( curve coincident with curves
from smooth model and model
20

with roughnesses A and B, Fig. 8.7 )

0
0 0.04 0.08 0.12 0.16 0.20 0.24 0.28 0.32 0.36
Time from entry , sec

Figure 8.12 . Effect of early cavity stripping on model drag. Entry velocity 60 fps,
air pressure 1 atmosphere .

Entry Velocities: 400 fps ( prototype) and with the results from the model launchings. The
120 fps (model ) . Only one launching was made prototype curve was aligned with the model
at 400 fps. Figure 8.14 compares its trajectory curves by sliding it horizontally until the first
with both smooth and type B-roughened model recorded pitch angle equaled the pitch of the
trajectories at 120 fps. All model tests were model . Then the pitch of the prototype as a
made with an air pressure of 1/11 atmosphere. function of time differed 6 degrees or less from
At this velocity , the models with the rough nose that of the model during the ensuing
and the models with the smooth nose repro 0.27 second (0.9 second in the prototype sys
duced the trajectory of the prototype to within tem ) . The scatter of pitch data for several
3 diameters during 70 diameters of underwater prototype runs having nearly identical trajec
travel. tories is evident in Figure 8.9 .
Figure 8.15 compares the fragmentary When the rough -nosed model was launched
pitch record obtained during the prototype test at full atmospheric pressure, the model remained

135
 Smooth model

Model with type B roughness

Figure 8.13 . Similarity of smooth and roughened model cavities at full atmospheric pressure.

in the cavity and the trajectory deviated from model launched at 1/11 atmosphere. When the
that of the prototype, proving that Froude roughened model was launched at full atmos
scaling alone was inadequate . pheric pressure ( eight times at 60 fps, four times
When type B -roughened models shed their at 80 fps, and three times at 120 fps ) , the cavity
cavities after only 1 1/2 or 2 lengths of under stripped during half of the 60- and 80 -fps
water travel during full -pressure launchings, launchings and did not strip at all at 120 fps.
further investigation was instituted. The smooth To investigate its possible contribution to
model did not leave its cavity during entries at cavity stripping, the surface tension of the water
60 , 80 , and 120 fps ; neither did the roughened was reduced from 76 to 40 dynes /cm by adding

136
 Water surface

Depth rs
0

diamete Roughness type B , entry attack angle -0.2°

20 드 .
Smooth , -0.4°

-0.7° Roughness type B, -1.0°

Smooth , -0.6°
Prototype, 400 fps
40 Model , 1/11 atm , 120 fps

0 20 40 60 80 100 120 140

Distance from entry , diameters

Figure 8.14 . Underwater trajectory of prototype compared with trajectories of

Froude -scaled smooth and roughened models. Equal cavitation numbers .
upse
No

40
le

Smooth
egch
d,ang
Pit

Roughness type B

0 =====
wn
Nose

Roughness type B
do

Prototype, 400 fps

Model , 1/11 atm , 120 fps Smooth
40
0 to -1 entry attack angle
1 1
0 0.04 0.08 0.12 0.16 0.20 0.24 0.28 0.32 0.36

Time from entry , sec ( model )

Figure 8.15 . Pitch of prototype compared with pitch of Froude- scaled smooth
and roughened models.

Aerosol Otl ( Reference 2 ) to obtain partial the cavity with change in surface tension
scaling (a reduction of 1/121 is required for (Figure 8.17 ) .
modeling, but no known liquid is capable of At the trajectory angle used in these tests,
meeting this requirement ) . During four launch the spherical portion of the nose contacts the
ings of the roughened model at an entry velocity water first if the entry attack angle is steeper
of 60 fps and full atmospheric pressure , none of than 0 to -1 degree; if the angle of attack is
the cavities stripped, indicating that surface flatter than +3 or +4 degrees, the ogive contacts
tension forces were at least partially responsible first. In the intermediate attack -angle range, the
for the phenomenon . Figure 8.16 shows a contact is at the junction between sphere and
stripped cavity and compares normal entry ogive. At an air pressure of 1/11 atmosphere , the
cavity in water with one in the Aerosol solution ; entry attack angle appears to influence the
there is very little difference. trajectory of the model (Figure 8.18 ). Available
One launching into the Aerosol solution data are insufficient to show whether or not a
with the tank air pressure reduced to 1/11 similar trend exists at full atmospheric pressure.
atmosphere showed no change in the shape of

1solution concentration less than 0.1 % Aerosol by weight.

137
 Water surface tension 76 dynes/cm , stripped cavity.

Water surface tension 76 dynes/cm, normal cavity .

Dilute Aerosol solution, surface tension 40 dynes/cm.

Figure 8.16. Entry cavities of roughened models in water and in Aerosol solution . Entry velocity
60 fps, full atmospheric pressure, roughness type B.

138
 Water surface tension 76 dynes/cm

Dilute Aerosol solution , surface

tension 40 dynes/cm

Figure 8.17 . Entry cavities of roughened model in water and in Aerosol solution. Entry velocity
60 fps, air pressure 1/11 atmosphere , roughness type B.

Model velocity as a function of time was approximately +2 degrees, the trajectory became
not significantly affected by change in entry flatter as the air pressure decreased ; at angles
attack angle, indicating that model orientation from +4 to +6 degrees, the trajectory became
in the cavity and the degree of contact between steeper with decrease in air pressure (Fig
model and water had little effect upon drag. ure 8.20 ). Photographs of four of the Fig
This was observed at both the 1- and ure 8.20 launchings are shown in Figure 8.21 .
1 / 11 - atmosphere pressure conditions The entry cavity and the orientations of the
(Figure 8.19 ) . model in the last three parts of Figure 8.21 are
The effect of change in air pressure on similar at the outset , but the subsequent trajec
underwater trajectory depended upon model tories show considerable variation, probably as
angle of attack at water contact. At an angle of the result of differences in flow separation about
139
 Watcr surface

,diameters
0

Depth
+ 1.3 °
Entry attack angle + 3.5 °
+ 4.2°
20
+2.1 °
+ 5.0 °
+ 0.2° + 2.0 °
+ 1.3 °
40
-2.9°
+0.4° -1.0°

0 20 40 60 80 100 120 140

Distance from entry , diameters

Figure 8.18. Influence of entry angle of attack on trajectory of smooth model.

Entry velocity 60 fps, air pressure 1/11 atmosphere .

the nose. The lines of separation around the prototype (entry velocity of 400 fps) during the
nose are not apparent. first length of underwater travel.
Tail contact with the bottom of the cavity In drawing a conclusion from the results of
was not sufficient to prevent a downturning these underwater trajectory studies , it must be
trajectory, indicating that nose forces were remembered that the experiments with this
stronger than tail forces . The deceleration of the particular missile were carried out under rather
model was almost the same whether the tail restricted conditions . Care should be exercised
initially contacted the water at entry or 70 ms in extrapolating from its behavior.
later (Figure 8.22 ) . It was found that one-to -one Froude and
The air pressure in the model system also cavitation -number scaling is sufficient to repro
made little difference in the velocity-time func duce the trajectory of the prototype and its
tion ( Figure 8.23 ). The 1- and 1 /11 -atmosphere attitude in space if the velocity of the model is
curves are the averages of the individual tests great enough to cause a turbulent boundary
plotted in Figure 8.19 compared with the results layer in the flow . The addition of gas-density
of two launchings at 1/21 atmosphere. scaling may be necessary for the modeling of
Increasing the entry velocity flattened the some types of missiles .
trajectory (Figure 8.24 ) and affected the decel The effect of surface tension on one-to-one
eration of the model. The average drag Froude and cavitation number scaling is insignif
coefficient during the first 26 diameters of icant . The uncertainty of inducing a turbulent
underwater travel was 0.17 for an entry velocity boundary layer by artificial means makes it
of 60 fps and 0.12 for an entry velocity of preferable to conduct modeling experiments at
120 fps. The value of 0.12 is in good agree entry velocities that will naturally result in
ment with the value of 0.11 measured for the turbulence in the boundary layer.

140
 2.0
Instantaneous
velocity

1.0
velocity
Entry

0.8

0.6
20 Launchings, smooth model , entry attack
angles -3.0° to +6.0°
0.4

0.2 2 Launchings, roughness type A

7 Launchings, roughness type B
Entry attack angles + 1.0 ° to +6.5°

0.1 I LUI 1 IIIII

10-3 2 4 6 8 10-2 2 4 6 8 10-1 2 4 6 8 1.0
Time from entry , sec

Air pressure 1/11 atmosphere

2.0
TTT
Instantaneous
velocity

1.0
velocity
Entry

0.8

0.6

0.4
4 Launchings, smooth model , entry attack
angles + 1.0° to + 4.5°

0.2

0.1
10-3 2 4 6 8 10-2 2 4 6 8 10-1 2 4 6 8 1.0

Time from entry , sec

Air pressure 1 atmosphere

Figure 8.19. Ratio of instantaneous velocity to entry velocity as a function of time.

Entry velocity 60 fps.

141
 , iameters

Water surface
dDepth

1/21 atm

1 atm
20

1 atm
1/11 atm
40
0 20 40 60 80 100 120 140

Distance from entry , diameters

Entry attack angles +1 to +2.5 degrees

, iameters

Water surface
dDepth

0
1 atm

1/11 atm
20

40 1
0 20 40 60 80 100 120 140
Distance from entry , diameters

Entry attack angles +4 to +6 degrees

Figure 8.20. Effect of change in air pressure on trajectory of smooth model .

142
 1/11 atmosphere , entry attack angle +2.5 degrees

1 atmosphere , entry attack angle +2.2 degrees

1/11 atmosphere, entry attack angle +5.0 degrees

1 atmosphere , entry attack angle +5.4 degrees

Figure 8.21 . Entry cavities of smooth model at various air pressures and entry attack angles ( 31,52 , and 167 ms after water entry ) .

143
 2.0

Instantaneous
velocity BBS
pogreb
velocity
Entry 1.0
‫ ه ه ه‬pada
0.8

0.6

0.4

O Tail contact 70 ms after entry

0.2 A Tail contact at entry

0.1 1 LUI 1
10-3 2 4 6 8 10-2 2 4 6 8 10-1 2 4 6 8 1.0
Time from entry , sec

Figure 8.22 . Effect on velocity - time function when tail of smooth model contacts bottom
of cavity. Entry velocity 60 fps, air pressure 1/11 atmosphere .

2.0
Instantaneous
velocity

1.0
velocity
Entry

0.8

0.6 1/11 atm

0.4

1 atm

1/21 atm
0.2

0.1 UL Ulul
10-3 2 4 6 8 10-2 2 4 6 8 10-1 2 4 6 8 1.0
Time from entry , sec

Figure 8.23 . Effect of air pressure on velocity- time function of smooth model .
Entry velocity 60 fps.
, iameters

Water surface
dDepth

120 fps
20

60 fps
60 fps

40 1
0 20 40 60 80 100 120
Distance from entry , diameters

Figure 8.24. Effect of increased entry velocities on trajectories of smooth model .

Entry attack angles -0.6 to +0.4 degrees, air pressure 1/11 atmosphere.

144
 CHAPTER 9

WATER-ENTRY CAVITY MODELING

The behavior of a missile is noticeably 2-inch-diameter model to represent the proto

affected by the size and shape of the cavity type missile, and 1 -inch and 1 /2-inch-diameter
enveloping it during high -speed water entry. models geometrically and dynamically similar to
Modeling of the cavity is an essential part of the prototype. The relatively uncomplicated
modeling missile behavior. vertical cavity was investigated first (Refer
As mentioned in the first chapter, one-to ence 101 ) ; the knowledge and experience so
one Froude and cavitation-number scaling does gained were then focused on the more complex
not scale air density . Static pressures on the problems of the oblique cavity (Reference 102 ).
water and on cavity surfaces are scaled only to For the vertical-entry modeling studies , a
the extent that pressure is constant over con missile with a simple shape, a hemisphere head, a
nected regions of air ; the air (gas ) dynamic cylindrical body section, and a spool-shaped tail
pressure, 12pvž is not scaled. Therefore, the ( Figure 9.1 ) was selected because previous inves
pressure tending to close the cavity is reduced tigations ( References 22 and 46 ) had shown that
and cavity seal is delayed, possibly permitting change in atmospheric density caused large
more air to enter and the cavity size to be differences in the vertical water-entry cavity of a
increased. For these reasons, the modeling of sphere. Thus gas-density effects could be
cavity size may require gas-density scaling. expected with the blunt-nosed hemisphere head.
Although one-to-one scaling of only the This configuration also has the advantage of
Froude and cavitation numbers had been reason relative stability at water entry , which reduces
ably successful in modeling oblique water-entry the chance of the density effect being obscured
motion (Chapter 7 ) and underwater trajectory by small fluctuations of the cavitation number
behavior (Chapter 8 and Reference 38 ) , no or by differences in missile attitude at water
systematic data on the size and shape of the entry . The spool-shaped tail was designed to
cavity were obtained, nor were the few instances exert a stabilizing effect on the missile during its
of failure to model the trajectory fully underwater trajectory . Nominal water-entry
explained . Thus the need arose to examine velocities were 120 fps for the 2 -inch-diameter
scaling techniques for modeling the water-entry prototype missile, 85 fps for the 1 -inch model,
cavity in some detail , so that both the role of and 60 fps for the 1 /2-inch model. An arbitrary
the cavity in determining missile performance length-to-diameter ratio of six was used .
and the sensitivity of the cavity to scaling For the oblique-entry modeling studies,
conditions might be established . two missile shapes were required to generate the
two kinds of cavities that seemed to be poten
EXPERIMENTAL PROGRAM tially sensitive to gas-density scaling. A narrow
The experimental program conducted at cavity that tends to close quickly at the water
the NUC Hydroballistics Laboratory in the surface was generated by a missile consisting of a
Variable -Angle Variable -Pressure Launching truncated cone having a 10.5 -degree generating
Tank sought to investigate the importance of angle and a 0.313 -diameter plate (Figure 9.2 ).
gas -density scaling in conjunction with Froude The second cavity is one that forms when the
and cavitation -number scaling in water-entry missile enters the water at a shallow angle and
cavity modeling. Three models were used : a quickly contacts the top cavity lip , the lip
145
 0.531

1.062

0.500

1.000

1.062 -
5.500
6.000

Figure 9.1 . Hemisphere missile for vertical water-entry modeling. Dimensions in diameters.

having a narrow cross section and relatively low The oblique-entry cavities were investigated
inertia. The missile fabricated to generate this under the following conditions :
cavity had a simple plate-cylinder nose shape,
with a cylindrical body and tail (Figure 9.3 ). Froude Cavita- Gas-Density
Missile dimensions and parameters are shown in Scaling Condition No. ,, F tion No., o Coefficient, p '
Table 9.1 .
1. One -to -one 51.8 0.073 0.8
For the initial study, the simplest of entry Froude-number ,
cavities was preferred - an axially symmetric
-

cavitation -number,
cavity formed when the trajectory and axis of and gas -density
the missile are vertical at water entry . The scaling
effects of four scaling conditions on the 2. One-to -one 51.8 0.073 0.41
Froude -number
vertical-entry cavity modeling of the hemisphere and cavitation
missile were studied .
number scaling
The 2 -inch -diameter hemisphere prototype
3. One-to-one 51.8 0.146/1 0.8
was tracked at least 25 diameters after water
Froude -number
entry , the 1 -inch -diameter model for
and gas-density
35 diameters and the 1 / 2 -inch - diameter model scaling
for 45 diameters . The prototype was photo
graphed at 0.0025 -second intervals . The smaller The truncated -cone missiles were photo
models were photographed at intervals of graphed during the first 20 diameters of their
0.00167 and 0.00125 second, respectively , underwater trajectory .
The plate-cylinder
which scale approximately as 2 1/2 . missiles were photographed either during the
146
 7.47

10.5

1.000 1

T
0.313

1.062

0.531

Figure 9.2. Truncated -cone missile for oblique water-entry modeling. Dimensions in diameters.
-

1.000

5.31

Figure 9.3 . Plate -cylinder missile for oblique water- entry modeling. Dimensions in diameters.

147
 Parameters
.Prototype
9.1
TABLE
Model
and

Missile
Hemisphere Truncated
Missile
-Cone Missile
- ylinder
CPlate
odel
1iniam
-D
.M M
D iam
.-i1/2nodel 1inodel
-D
.M iam 1inodel
-D
.M iam

Required
by Required
by Required
by by
Required
Froude Froude Froude Froude
Scaling
SD niam
caling
.-2iActual
iam
Parameter a = =
4.999
Parameters
1=/1Prototype
Prototype
Parameters
.012
=
|Parameters

,in
.Diameter 1.998 0.999 0.999 0.498 0.498 1.998 0.999 0.999 1.998 0.999 0.999

.,in
Length 12.016 6.011 6.005 2.995 3.007 14.95 7.48 7.46 10.61 5.31 5.31

,lb
Weight 1.128 0.139 0.141 0.017 0.017 1.304 0.163 0.167 1.045 0.131 0.130

CG
of
Distance
from
,in
.nose 5.90 2.95 2.92 1.47 1.47 7.97 3.99 3.93 5.35 2.70 2.65

Moment
of
in
2,alb
inertia

148
15.61 0.489 0.49 0.015 0.015 21.44 0.672 0.672 12.59 0.394 0.389

Cone
angle 3'10
°0 °30
'10 °24
'10

---
Diamet
of
nose er
iflat
,. n 0.619 0.310 0.318 1.998 0.999 0.999

.
---

---
a*Aransverse
the
tCG bout
through
.axis
first 15 diameters or during the interval between pressure of the model system , to that of a
20 and 30 diameters of their underwater travel. standard , defined as the density of air at 20°C
Whip data were obtained when the first 15 or and 740 -torr pressure . The desired densities were
20 diameters of travel were recorded . obtained by mixing air with two heavy gases
As in the vertical water-entry studies , the Freon 12 (dichlorodifluoromethane) and
models were photographed at shorter time Freon 114B2 (dibromotetrafluoroethane ) which
intervals than the prototypes, the ratio approxi have densities , respectively , four and nine times
mating 11/2. The truncated-cone prototype was greater than air under the same conditions of
photographed at 0.0033 -second intervals and the temperature and pressure (Reference 67 ). A
plate-cylinder prototype at 0.005 - second gas-density coefficient of 0.8 was used as the
intervals . Some tests were conducted at even experimental baseline atmospheric density .
shorter intervals to permit more detailed studies Helium was added to air to reduce its gas-density
of cavity behavior. Water entry was defined as coefficient to 0.8 for the full atmospheric
the first contact between the missile and the pressure launchings .
water surface . All events were timed from water To determine the extent to which proto
entry and all positions measured with respect to type cavities having a gas-density coefficient of
it . 1.0 could be predicted from a modeling system
Figure 9.4 is a typical print from the having a gas-density coefficient of 0.8 , vertical
rotating -disk camera (described in Part V) launchings of the hemisphere missile were also
illustrating the vertical water entry of the made at atmospheric pressure and a gas-density
coefficient of 1.0 .
1 -inch-diameter hemisphere model launched
with a water contact velocity of about 85 fps, To determine whether the air content of
gas pressure of 1/4 atmosphere, and gas-density the water altered the entry cavity , it was varied
coefficient of 0.8. To preclude any ambiguity in for several launchings of the prototype hemi
the terms “ surface closure” and “ deep closure ,” sphere missile. Cavity size and shape were not
they are identified in Figure 9.4 . Surface closure affected , but the cavity wall became roughened
is the closure of the water surface itself and not when the air content of the water exceeded the
the dome of the splash ; deep closure is any equilibrium concentration with air at atmos
permanent break in the cavity beneath the water pheric pressure . Since well-defined cavity con
surface. tours were important to modeling, this equili
The following measurements were made brium concentration was carefully maintained
from the photographs: ( 1 ) the time and position during all tests.
of cavity closure, (2 ) the cavity diameter at the
missile tail, and (3 ) the projected cavity area in a EXPERIMENTAL RESULTS : VERTICAL
vertical plane containing the cavity axis of CAVITY
symmetry. The missile positions and the cavity The three sizes of hemisphere head missiles
parameters listed and illustrated in Figure 9.5 formed reproducible cavities as long as the
were measured for each oblique launching and gas-density coefficient was 0.8 or higher. Surface
obtained as functions of time . Lack of informa closure always occurred before deep closure.
tion on yaw made it impossible to determine When the gas-density coefficient was decreased
missile underwater pitch angles; angles of by lowering the atmospheric pressure, cavity
inclination (see Nomenclature ) as shown in closure became erratic under presumably equal
Figure 9.5 were measured. For coincident trajec entry conditions. Surface closure was late, and
tory and object planes , sideview photographs in some instances followed deep closure or failed
show missile inclination . In vertical trajectories, to occur during the recorded trajectories.
the indeterminate trajectory plane is taken to be Cavity closure is extremely important in
the object plane chosen for photography. The determining cavity history because ( 1 ) the mass
model data were scaled up to prototype size, of gas in the cavity cannot appreciably increase
making the curves and photographs of the small after the cavity closes , and (2 ) the persistence of
and large missiles directly comparable. the cavity about the missile is largely determined
Gas densities are described in terms of the by the time and place of deep closure.
gas-density coefficient p ', which is the ratio of As shown in Figure 9.4 , the seal that
the density of the gas , at the temperature and formed at the top of the cavity was always flat

149
 DEEP
CLOSURE

SURFACE
CLOSURE

Figure 9.4 . Vertical water -entry cavity of 1 -inch -diameter hemisphere missile.

150
 Without gas-density scaling, modeling does not
occur .
To find out what happens when only gas
density or only gas pressure is varied, an
additional series of launchings was made with
b
the 2-inch prototype missile. Figure 9.6 com
pares gas-density coefficients of 0.4 and 0.8.
Figure 9.7 compares gas pressures o = 0.146 and
o = 0.073 . Increase in gas density at constant
a = Width of cavity mouth
b = Diameter of cavity at
pressure accelerates cavity surface closure , with
missile tail a reduction in projected cavity area. Reduction
c , d = Cavity lips in gas pressure at constant gas density only
e = Cavity cusp slightly affects surface and deep closure and
1 = Inclination of missile axis projected cavity area .
Truncated -cone missile
The ratio of the difference in depth to the
difference in time between the surface and the
first deep closure (Figures 9.6 to 9.8 ) provides
a an index of cavity modeling. If the ratio for the
2-inch-diameter missile with one-to -one Froude
and cavitation -number and gas-density scaling is
taken as the prototype modeling condition, an
empirical closure parameter w which compares
d cavity -closure modeling under the different
e
a = Width of nose cavity mouth scaling conditions, can be defined as follows:
b Width of tail cavity mouth
c = Length of cavity
d = Diameter of nose cavity λ - 1 ΔΗ ) ΙΔh,
Ahpр
e = Diameter of tail cavity
1 = Inclination of missile axis

Plate -cylinder missile

W =

)
λ- 1 /2 Δ1 Дteр *

Ah
Ah р

Δt
(9.1 )
Figure 9.5 . Oblique water-entry cavity parameters ( 1/2 Δt
measured from photographic data.

where Ah is the difference in depth and At is the

and essentially parallel to the water surface. The difference in time between surface and the first
depth of the cavity top was measured and deep closure. Depth and time in the numerator
plotted as a function of time after water entry. of Equation 9.1 have been scaled to prototype
In instances , several
many deep closures conditions, and the denominator (Ahp /Atp )*
occurred during a single launching. represents the ratio for the prototype modeling
Figures 9.6 to 9.8 compare the closures condition. Subscripts are not used in the numer
obtained with one -to -one Froude- and ator because ratios for both prototype and
cavitation -number scaling or Froude-number and models are used. For cavity closure modeling,
gas-density scaling with the closures obtained w = 1. From Figure 9.9 , where w is plotted as a
when all three – Froude and cavitation numbers function of gas pressure and scaling conditions,
and gas density – are scaled simultaneously . it is evident that one-to -one gas -density scaling is
With gas-density scaling, both surface closure necessary for modeling the closure parameter
and deep closure modeling are obtained, and that one-to-one cavitation-number scaling is
whether or not the cavitation number is scaled ; of relatively little importance.
surface-closure modeling is slightly improved The magnitude of the cavity diameter at
when cavitation -number scaling is included. the missile taill appeared to be little changed by

1Measured only when the tail was not in contact with the cavity wall.
151
 F = 51.8 p = 0.073
=

Missile Gas Gas density

diam , pressure , coefficient,
180
e' Symbol
Projected

in atm
160
scaleotype
cavity

2 1/2 0.4
,iarea
ofn.2

140 1 1/4 0.2

prot

1/2 1/8 0.1 1980081

120
2 1/2 0.8
100
parte
80
60

40
prototype

20
Diameter

scale
missile

0
cavity
tail
of

.,in

6
at

Surface closure Water surface

0 Q! No surface
- ( 1 of 3 only) closure with
10 1 /2-in.-diam
model
inrototype
20
pcavity

First deep
Depth
scale

closure
,.top

Second deep
of

closure 7
30

40
First deep
closure
50

60 Second
deep
closure
70

80

0 10 20 30 40 50 60 70 80 90 100 110
Time from water entry , ms, prototype scale

Figure 9.6. Cavity parameters of hemisphere missile with one-to-one Froude

1
and cavitation -number scaling. g = -90 degrees.

152
 F = 51.8 p = 0.8
Missile Gas Cavitation
180
iProjected

diam , pressure , number,

in.
scaleotype

160 atm Symbol

cavity

o
area
, n.2
of

140
prot

2 1 0.146
120 1 1 0.292

1/2 1 0.584 8692881

100
2 1/2 0.073
80
1.A
60
40
prototype

20
Diameter

0
scale
missile
cavity

6
of

tail
,.in
at

Surface closure Water surface

0
10
inrototype

First deep closure

.,Depth
cavity
pscale

20
top
of

30
UB
40

Second
50 deep
closure
Third
60

deep
closure

70

80

0 10 20 30 40 50 60 70 80 90 100 110

Time from water entry, ms, prototype scale

Figure 9.7. Cavity parameters of hemisphere missile with one - to -one Froude
and gas-density scaling. { = -90 degrees.

153
 F = 51.8, o = 0.073, p = 0.80
Missile Gas
180 diam, pressure,
Projected

in. atm Symbol

scale type
cavity

160
area

2 1/2
,in.2
of
proto

140 1 1/4
120 1/2 1/8
100

80
60

40
prototype

20
Diameter

missile

0
scale
cavity
tail

6
of

i,. n
at

Surface closure
Water surface
0
nrototype

10
Depth
pcavity
scale
.,itop
of

20

First deep closure

30

40
50

Second deep
closure
60

70

80

0 10 20 30 40 50 60 70 80 90 100 110
Time from water entry, ms, prototype scale

Figure 9.8. Cavity parameters of hemisphere missile with one- to -one Froude
and cavitation -number and gas-density scaling. & = -90 degrees.

154
 5.0

Scaling
Symbol Condition F 0 e'
F, o , p' 51.8 0.073 0.8
4.0
,parameter

Х 51.8 0.146/1 0.8

F, e'
Closure

+ 51.8 0.146 /1 1.0

I
w

F, o 51.8 0.073 0.4λ

0
3.0

o
2.0
1
Modeling line

1.0

0
0 1/8 1/4 1/2 3/4 1.0

Gas pressure in tank , atm

Figure 9.9. Importance of gas-density scaling in modeling cavity closure of hemisphere missile .
= -90 degrees.

the various scaling conditions ; the cavity diam Figures 9.6 to 9.8 show that fair modeling
eter curves never deviated more than one missile of cavity volume was achieved by one-to-one
diameter from the prototype condition . scaling of Froude number and gas density, and
Small oscillations in the cavity diameter that excellent modeling resulted when the cavi
time curves were not modeled by any scaling tation number was also scaled . Successful model
condition. However, it was observed that : ing of cavity volumes, with 1 -inch and 1 /2-inch
1. Oscillation occurred only after the diameter spheres in a one-to-one Froude and
cavity was closed from the atmos cavitation-number and gas-density scaled system,
phere above the water surface. was reported in Reference 46 .
2. The period of the oscillation increased Equations 1.9 , 1.16 , and 1.17 indicate that
with the underwater trajectory (here , water penetra
a. Increase in missile diameter at tion distance ) will be modeled if the distance,
constant Froude number expressed in diameters, is the same when the
b. Decrease in gas density corresponding times of penetration scale as
C. Decrease in gas pressure 1 1/2 . Measurements of penetration distances
For geometrically similar cavities, the corresponding to 0.05 second in the prototype
cavity volume ratio is equal to 1 times the scale (about 24 diameters of penetration, the
projected area ratio . After deep closure , the greatest distance for which modeling comparison
cavity area remaining about the missile was could be made ), indicate that good underwater
measured . When the gas-density coefficient was trajectory modeling was obtained for all model
0.8 or higher , the cavities made by the three ing conditions .
hemisphere missiles were sufficiently similar
(Figure 9.10 ) that the relationship between EXPERIMENTAL RESULTS : OBLIQUE
cavity volume and area ratios could be assumed . CAVITY
When the gas-density coefficient was not scaled ,
the cavities were initially similar but became Truncated -Cone Missile
quite different in shape later in the trajectory The narrow water-entry cavity produced by
( Figure 9.11 ) ; no direct comparison of volume the truncated - cone missile was found to be
could be made . extremely sensitive to gas-density scaling. The
155
 1 - in . -diam model 1 / 2-in.-diam model
Scaling condition 2-in.- diam prototype ( enlarged) (enlarged )

F = 51.8
p = 0.8

Cavity length scaled to prototype

0.146
0 0.292 0.584

dimensions, in. 15.0 14.4 13.6

L / Pmax 3.13 3.18 3.15

F = 51.8
e ' = 1.0
=

o risici 0.146 0.292 0.584

Cavity length scaled to prototype
dimensions, in . 14.1 13.2 14.4
L /Dmax 3.18 3.08 3.29

F = 51.8
o = 0.073
' = 0.8
-

Cavity length scaled to prototype

dimensions , in . 14.8 14.6 14.4
L / D max 3.06 3.05 3.14

Figure 9.10. Shape of hemisphere missile vertical water-entry cavity , gas density scaled.

cavities that formed at low and at high densities of the two pictures in each figure shows clearly
in otherwise equal systems were entirely differ that change in gas-density coefficient and not
ent in character, as illustrated in Figure 9.12
=
change in cavitation number is responsible for
(cavitation number o = 0.073 ) and in Fig. the change in cavity character.
ure 9.13 (cavitation number o = 0.146 ) . At low Shifting the gas -density coefficient from
densities , the cavities are extremely large and 0.11 to 0.195 made a difference in the surface
still growing at the end of the photographic closure of the truncated-cone missile cavities, as
records ; at high densities the cavities close at the shown in Figure 9.14 . Increasing the gas -density
surface and soon begin to disappear. Comparison coefficient retarded the growth of the cavity
156
 1 - in .-diam model 1 / 2-in .- diam model
Scaling condition 2 - in .- diam prototype (enlarged ) ( enlarged )

F = 51.8
0.73

' 0.4 0.2 0.1

Cavity length scaled to prototype
dimensions, in . 13.7 14.8 16.0
L/ Dma ; 2.61 2.78 2.78

F = 51.8
0 = 0.073

e'
Cavity length scaled to prototype
dimensions, in.
L / Dmax
AVE 0.4

36.7
8.54
0.2

40.5
4.54
0.1

38.4

Figure 9.11 . Shape ofhemisphere missile vertical water-entry cavity , gas density not scaled.
3.73

mouth at the water surface. The growth rate is in Figure 9.14 . This parameter is not so depend
extremely sensitive to gas-density changes in the ent upon gas density as is surface closure . In the
range of 0.4 < p < 0.8 ; above and below this range 0.8 <p < 1.95 , the cavity diameter
range, the sensitivity is less. No variation in decreases 17 %, but below 0.8 the diameter is
missile attitude occurred during the range of test almost independent of p ' ( Figure 9.15 ) . The
variables. time between water entry and the missile tail's
The cavity diameter at the missile tail as a first contact with the cavity wall indicates a
function of time is another measurement shown dependence upon gas density similar to that
157
 p = 0.11 p = 1.95
37 ms after water entry 37 ms after water entry

Figure 9.12. Effect on truncated-cone prototype cavity of change in gas-density coefficient

when o = 0.073 , g = -39º .

U
p' = 0.17 b ' = 3.48
40 ms after water entry 42 ms after water entry

Figure 9.13. Effect on truncated -cone prototype cavity of change in gas-density

coefficient when o = 0.146, 8 = -39º .

158
 Inclinat ion
6

,dl eg
-40
Inclination , nose down
5

Diameter
-20
p ' = 0.98 p = 0.78

cavity
1.08 4

tail
0.22

of
.,iatn
Cavity diameter
b' = 0.11
0.78 0.40
1.95 0.90
10 1.89 3

0.43
Cavity 0.50
i.,mouth

-
cavity
Width

8 mouth 2

1
of
n

0.59
0.64

1
6

0.63
1.08
4 0

0.67
0.71
2 0.78
0.78

1.95 1.89 0.90

0.98
0
0.00 0.01 0.02 0.03 0.04 0.05 0.06
Time from water entry , sec

Figure 9.14. Effect on truncated-cone prototype cavity of variation in gas-density

coefficient from 0.11 to 1.95. F = 51.8 , 0 == 0.073, š = -39 degrees.

shown by the cavity diameter : tail contact much alike, and cavity-mouth curves showed
occurs earlier with decrease in cavity width. that an excellent degree of modeling was
The response of the 1 -inch model water obtained . Closure curves and photographs
entry cavities to change in gas density (Fig illustrating the conditions p ' = 0.1 and p ' = 0.8
ures 9.16 and 9.17 ) was similar to that of the are shown in Figures 9.18 and 9.19 , respectively.
prototype. No significant change in the diameter For the scaling that included only Froude
at the missile tail was observed for a gas-density number and gas-density coefficient, a good
range of 0.1 to 0.8 and cavitation numbers degree of modeling also was obtained
0.073 and 0.292 . (Figure 9.20 ) .
The narrow oblique cavity responded in Figure 9.21 shows the effect of one-to-one
much the same way as the vertical cavity to the scaling of the Froude and cavitation number
various scaling techniques ; i.e., one-to-one upon the truncated -cone cavities . Here p ' = 0.4
scaling of the Froude and cavitation numbers for the prototype and 0.2 for the model . Both
and the gas-density coefficient modeled the the photographs and the cavity-mouth curves
cavity to a high degree of accuracy. Failure to show that the prototype cavity mouth is
scale the cavitation number did not prevent beginning to close at the end of the photo
good modeling, but when the gas-density coef graphic record , while the mouth of the model
ficient was not scaled , modeling did not occur. cavity is still growing. The data show that
Figures 9.18 to 9.21 are a series of cavity satisfactory modeling of surface closures does
photographs paired with surface-closure curves not result when p'is not scaled .
for the respective cavities . These are discussed The change in missile attitude at water
below according to the scaling conditions they entry, measured from the whip records, was
represent . found to be independent of scaling conditions.
When all three factors were scaled , cavities The change in missile yaw angle was negligible.
that formed under equal conditions looked very From the data available, the undercavity seems
159
 Diameter
cavity 4.5
missile
water
tions. Further, it was sensitive to changes in
entry
.after
0.02
of
at
tail
sec
,in cavitation number ; the truncated-cone missile
4.0 text so google was not .
Cavity diameter The pitch of the plate-cylinder prototype
3.5 increased with both gas-density coefficient and
Average cavity diameter at cavitation number, as shown in Figure 9.22,
p < 0.8 is 3.95 in. where the change in pitch is plotted as a
3.0
function of time, and in Figure 9.23 , where the
change in pitch after 5 ms of underwater trajec
tory is presented as a function of the gas-density
coefficient. Approximately equal increases in
between

40
negative pitch were caused by changing the
water
entry
,contact
Time

gas-density coefficient from 0.1 to 0.8 at a

and

constant cavitation number of 0.073 , and by

tail
m s

00

Xo

30 changing the cavitation number from 0.073 to

Tail contact 0.146 at a constant gas -density coefficient of
0.8 . The time between water entry and first
20
-

contact between missile tail and top cavity wall

0 0 = 0.073
decreased at higher gas-density coefficients and
-

10
X0 = 0.146
cavitation numbers, in keeping with the concur
rent increase in negative pitch ( Figure 9.24 ).
0 Increase in negative whip (i.e. , pitch ) with
0.0 0.4 0.8 1.2 1.6 2.0 cavitation number had been observed in the
Gas-density coefficient, oblique water entry of a 2-inch-diameter model
with a plate-cylinder nose ( Chapter 6 ) , and some
Figure 9.15 . Effect of gas-density coefficient indication of increase in negative pitch with
on truncated-cone prototype cavity diameter. increase in gas-density coefficient also appears in
F = 51.8 , 8 = -39 degrees.
the plate-cylinder data shown in Figure 7.7 .
Flow separation is probably influenced by
the density of the ambient gas . Because cavity
to be insensitive to scaling condition. The slope closure is accelerated by increasing gas density,
of the cone appears to prevent elongation of the it is possible that undercavity sealing also occurs
undercavity with decreasing atmospheric pres earlier , causing the undercavity to be smaller.
sure, thereby preventing a reduction in whip This would increase the wetted underpressure
similar to that experienced by the plate-cylinder area and explain the observed increase in nega
missile (Chapter 6 ) . The whip of this head shape tive pitch with gas-density coefficient.
might profitably be studied over a wide range of During the first 15 diameters of underwater
entry conditions ; it is unusual to find whip travel , the trajectory of the plate-cylinder proto
independent of cavitation number. type was independent of launching conditions.
Reference 102 indicates that the first After approximately 20 diameters , the trajec
20 diameters of trajectory are independent of tory became flatter with decreasing gas-density
gas density in the range 0.11 < p < 1.95 but coefficient or cavitation number, resulting in a
may be affected by gas densities above p ' = 2.0. vertical spread of three diameters in the trajec
The inclination of the missile axis was also tories after 28 diameters of horizontal travel.
found to be independent of scaling conditions Figure 9.25 shows the relationship between
during the first 20 diameters of underwater missile trajectory and the inclination of the
travel. missile axis , which changes 9 degrees after
15 diameters of travel , although the trajector
ies at this point are very nearly the same.
Plate Cylinder Missile After 28 diameters of travel (Figure 9.26 ) , the
In contrast to the truncated -cone missile, trajectories deviate only three diameters but the
the plate-cylinder missile showed greater trajec spread of the inclination curves is 40 degrees.
tory and attitude sensitivity to gas-density varia The extreme sensitivity of the missile inclination
1
160
 =
' = 0.81
49 ms after water entry

o = 0.292

=
p ' = 0.10 p ' = 0.21
52 ms after water entry 52 ms after water entry

P = 0.47 p ' = 0.81

50 ms after water entry 50 ms after water entry

o = 0.073

=
Figure 9.16 . Typical truncated -cone model cavities . F = 51.8 , = -39° ; time in prototype scale.

161
 Inclinat ion
6

,l
Inclination, nose down
-40

deg

ofnrototype
Diameter
5

ipcavity
.scale
p = 0.12

tail
-20
Cavity diameter

,at
0.10
p = 0.22
inrototype
12 4
0.10
mouth
cavity

0.21
pWidth

= 0.81 0.47
,.scale
of

- 0.82 0.81 0.22

10 3
0.56
Cavity 0.21
mouth
8 2

4 0
0.81
-O = 0.073 d' = 0.54
--

-O = 0.292 0.82 0.75

2

p = 0.81 10.56 0.47

0 1
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Time from water entry , prototype scale, in.

Figure 9.17 . Effect on truncated-cone model of variation in gas-density coefficient

from 0.10 to 0.82 . F = 51.8 , 8 = -39 degrees.

to changes in launching conditions is apparent in gas-density coefficient was changed from 0.1
Figure 9.27 . to 0.8 .
The cavities formed by the plate-cylinder Both trajectory and cavity of the plate
prototype under the various launching condi cylinder prototype were independent of launch
tions were similar in appearance, all so wide that ing conditions during the early portion of the
surface closure did not occur during the photo trajectory, where model data were also obtained.
graphic record . Typical cavities, formed under All scaling techniques investigated resulted in
various scaling conditions during the second modeling both cavity and trajectory.
series of launchings, are shown in Figure 9.28. Figure 9.31 presents typical photographs of
Three photographs from each launching show the model and prototype cavities, taken at the
that the cavities were still similar in appearance same time after water entry . All the cavities are
after 20 diameters of travel, and that differences almost identical in appearance, except for the
appeared only as the cavities began to close differences in cavity texture and the varying
about the missile, evidencing variations that degrees of tail contact. The textures of the
probably affected missile position. Even rather cavity wall are similar when the gas-density
small differences in the cavity assume signifi coefficient is scaled . At a constant cavitation
cance because the cavities from duplicate number of 0.073 , the walls of both model and
launchings were so similar that their photo prototype cavities are opaque when p '= 0.8 and
=
graphs at the same times after water entry transparent when p ' = 0.1 ; at a constant gas
appear to be two prints of the same negative density coefficient of 0.8 , the cavities were
(Figure 9.29 ) . opaque at all cavitation numbers investigated.
Figure 9.30 reproduces average parameter The clearing of the cavity wall is apparently an
curves for each launching condition. The width effect of gas density rather than of cavitation
of the cavity mouth is the only parameter number.
significantly affected by launching conditions, Most of the launchings of the l-inch
showing a maximum variation of 10% when the diameter plate-cylinder model were made with
162
 Prototype Prototype
33 ms after water entry 32 ms after water entry

Prototype Model
47 ms after water entry 44 ms after water entry

Figure 9.18a. Shape of truncated-cone missile oblique -entry cavities with F = 51.8 ,
o = 0.073 , p = 0.1 , $ = -39º . Time in prototype scale.
=

163
 dInclination
,i eg
Inclination, nose down
-40
5

pnrototype
,iDiameter
14
nrototype p = 0.12
-20 Cavity diameter
ニー

cavity
0.10

.scale
ニニ
mouth
Width
pcavity

tail
.,iscale

of
12 4

at
U
of

10 3
Cavity
mouth
8 2
d ' = '0.11

Prototype

Model
2

0
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Time from water entry , prototype scale , sec

Figure 9.185. Curves for Figure 9.18a cavity parameters.

some initial nose-down pitch (Reference 102 ), launched truncated-cone missile were sensitive
which tended to make tail contact earlier and to variation in gas density . Gas-density scaling is
more violent. This in turn exaggerated the size necessary to obtain modeling. Modeling of the
of the tail cavity and frequently caused the tail hemisphere cavity is somewhat improved by the
bulge to break through the water surface (Fig. addition of cavitation-number scaling. Accuracy
ure 9.32 ) . When this occurred , both the nose in prediction of full -size cavities of this type is
and tail cavities were measured. Cavity param greatly improved by gas-density scaling. Good
eter curves from three launchings at the condi = =
water - penetration-distance modeling was
tion F = 51.8 , o = 0.073 , and p ' = 0.8 , presented obtained for all modeling conditions studied.
in Figure 9.33 , show that the behavior of the Cavitation number may prove more impor
nose cavities was independent of tail-cavity tant in scaling cavities formed by other missiles
formation . or cavities formed by these missiles under
Average curves , presenting only model different entry conditions. Furthermore , model
nose -cavity data from each scaling condition, are ing of later missile motion requires that the
compared with each other in Figure 9.34 ; they cavitation number be scaled (Chapter 8 ). The
can be compared with the corresponding proto general insensitivity of the truncated-cone
type curves in Figure 9.30 . The cavities formed missile to change in cavitation number is prob
by model and prototype under equal scaling ably related to the constant whip at water entry.
conditions differ less than the cavities formed by The plate-cylinder prototype data show
the same missile under various launching condi that both cavitation number and gas-density
tions . coefficient influence the cavity , trajectory, and
attitude of this obliquely launched missile
DISCUSSION OF RESULTS beyond 15 diameters of underwater travel ;
This investigation showed that narrow hence , modeling cannot be expected unless both
water-entry cavities formed by the vertically parameters, in addition to the Froude number,
launched hemisphere missile and the obliquely are scaled .

164
Prototype Model
35 ms after water entry 33 ms after water entry

Prototype Model
49 ms after water entry 48 ms after water entry

Figure 9.19a. Shape of truncated - cone missile oblique -entry cavities with F = 51.8 ,
=
o = 0.073 , = 0.8 , š = -39º . Time in prototype scale.
=

165
 Inclinat
i eg ion
Inclination, nose down

,d
-40

-20
Cavity diameter
4

,Diameter
pe
cavity
prototy
tail
of
at
-
3

scale
.i, n
2

Cavity Prototype
pe

mouth
mouth
Width
cavity
prototy

Model

-
6
iscale
of
,
., n

-
' = 0.77

0.78 0.81
0
0.00 0.01 0.02 0.03 0.04 0.05 0.06

Time from water entry , prototype scale , sec

Figure 9.19b. Curves for Figure 9.19a cavity parameters.

The studies demonstrated that gas-density Froude-number scaling is required for modeling
scaling, in conjunction with Froude and missile oblique water-entry motion, cavitation
cavitation -number scaling, is generally necessary number scaling and gas-density scaling in con
for water-entry cavity modeling, which is neces junction with Froude-number scaling are
sary for modeling the behavior of the missile in probably required for simultaneous modeling of
its underwater trajectory in the cavity . Since missile behavior during water entry and during
cavitation -number scaling in conjunction with underwater trajectory in the cavity .

166
 Prototype Model

37 ms after water entry 35 ms after water entry

Prototype Model
47 ms after water entry 46 ms after water entry

Figure 9.20a. Shape of truncated -cone missile oblique -entry cavities with F = 51.8 , o = 0.146, for
the prototype and 0.292 for the model , p = 0.8 , š = -39º . Time in prototype scale.
-

167
dInclination
,l eg

Inclination , nose down

-40

-20
Cavity diameter
4

Diameter
type
cavity
tail
,of
at
proto
-
3

.i,scale
n
Cavity
mouth

-
8 2

Prototype
type

.
mouth

Model
cavity
,Width

1
proto

6
iscale
of
., n

p' = 0.81
0.79
2

0.80
0.82
0 ‫ܛܐ‬
0.00 0.01 0.02 0.03 0.04 0.05 0.06

Time from water entry , prototype scale, sec

Figure 9.20b. Curves for Figure 9.20a cavity parameters.

168
Prototype Model
35 ms after water entry 35 ms after water entry

Prototype Model
49 ms after water entry 49 ms after water entry

Figure 9.21a. Shape of truncated -cone missile oblique -entry cavities with F = 51.8 , o = 0.073,
=

and p ' = 0.4 for the prototype , and 0.2 for the model , g = -39º . Time in prototype scale .

169
 1

Inclination
,i eg

-40
Inclination , nose down

pnrototype
d

‫ ܥܢ‬-

,iDiameter
cavity
14 5

of
at.scale
tail
-20
Cavity diameter
.inrototype

p = 0.21
-
mouth

4
,pWidth

12
cavity

0.22
scale

====
of

10 3
0.40

0.43
8 Cavity I 2
mouth
S
6 1

2 Prototype
Model

0
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Time from water entry , prototype scale , sec

Figure 9.21b. Curves for Figure 9.21a cavity parameters.

170
 between
water
entry
Time
contact
80

and
-2.4

tail
s
D' = 0.80 0.76

,m
0.78 60 *
-2.2

40

1
-2.0
lla Oo = 0.146
-

20
0.42

.
Ill 0.79
-Change

-1.8 X 3 = 0.073
1
pitch

ll
d-eg
,000
in

// 0 1
-1.6
0 0.2 0.4 0.6 0.8 1.0 1.2

-1.4 Gas-density coefficient, e'

0.11
-1.2
Figure 9.24. Time between water entry of plate
cylinder prototype and first contact of tail with
-

-1.0
cavity wall. F = 51.8, 8 = -20 degrees.

-0.8

ion
==

dInclinat
-0.6 -60

i, eg
0 = 0.073
-

-0.4 0.8
-
0 = 0.146 20.8
-40
p ' as indicated 0.4
-0.2 0.1
-20
0 o = 0.073
0 8 16 24 32 40 48 56 64 72 ' = 0.8
o = 0.073
Time from water entry , sec X 104 0
' = 0.4
O = 0.073
Figure 9.22. Change in pitch after water entry of plate-cylinder e ' = 0.1
prototype, with F = 51.8, & = -20 degrees. o = 0.146
Vertical
diameters

= 0.8
-
travel

0
,

-2.4
Change

5
0-pitch
00,5
ms

-2.0
in
water
entry
,-
after

10 1
,deg

*
-1.6 -5 0 5 10 15 20 25

Horizontal travel , diameters

-1.2
Figure 9.25. Trajectory of plate-cylinder prototype
-0.8 during first 15 diameters of underwater travel.
F = 51.8, 8 = -20 degrees.
0 0 = 0.146
-0.4
X 3 = 0.073

0
0 0.2 0.4 0.6 0.8 1.0 1.2

Gas-density coefficient, p

Figure 9.23 . Change in pitch 5 ms after

water entry of plate-cylinder prototype.
F = 51.8 , 8 = -20 degrees.

171
 -100

ion
0.8

dInclinat
-80
0.8

i, eg
1 " 0.8
-60 0.2
0.1

-40

-
o = 0.037
-20
e' = 0.8
-

0 = 0.073
0 p = 0.8
0 = 0.1
,diameters

0.146
Vertical

p = 0.8
travel

2
p = 0.2

4
0.8

6 0.1
1

8
0.2

10
0.8 0.8

12 1 1
20 22 24 26 28 30 32
Horizontal travel , diameters

Figure 9.26. Trajectory of plate -cylinder prototype

between 20 and 28 diameters of underwater travel.
F = 51.8 , 8 = -20 degrees.
>

-100

0 = 0.073 p = 0.79
-90
0.146
0 0.037
-80 0.79
i, eg nation

-70 0.83
0.18
dIncli

0.17
-60

-50
' 0.83
0.82
-40
0.11

-30 0.12

-20

-10

0
0.00 0.02 0.04 0.06 0.08
Time, sec

Figure 9.27 . Effect of scaling condition on inclination of

plate-cylinder prototype. F = 51.8, x = -20 degrees.
-

172
 ree .

F = 51.8
o = 0.146
p' = 0.17

F = 51.8
F - 51.8
o== 0.036
p = 0.82 o = 0.073
' = 0.11

F = 51.8 F - 51.8
o = 0.073 o = 0.146
p ' = 0.83 p ' = 0.79

Figure 9.28 . Effect of scaling condition on plate-cylinder prototype cavity , 8 = -20° . Photographed
about 120 , 170 , and 190 ms after water entry .

173
 F - 51.8
o = 0.146
p = 0.79

F = 51.8
o = 0.146
p ' = 0.79

Figure 9.29 . Cavity similarity of two plate -cylinder missile launchings

under the same conditions. 8 = -20° .

174
 -4
Diameter

Entry point
cavity

0
from
lips

0.073
of

4
' = 0.8
-
.i,point
entry

o = 0.073
n

p = 0.4
'
8
0 = 0.146
p = 0.8
12
o = 0.073
16 b ' = 0.1
-

20

,diameter
mouth
-

Width
40 20
Cavity
of
Length
cavity

mouth
32 16
cavity

Cavity
of
.,in

length
.of
,in

24 12

Cavity
-

16 diameter 8

8 4

0 0
0.00 0.02 0.04 0.06
Time from water entry , sec

Figure 9.30. Effect of launching conditions on plate

cylinder prototype cavities. F = 51.8 , š = -20 degrees.

175
 Prototype Model
57 ms after water entry 57 ms after water entry

p ' = 0.8

Prototype Model
57 ms after water entry 57 ms after water entry

p ' = 0.1

Figure 9.31 . Effects of scaling conditions on plate-cylinder missile cavities,

F = 51.8, o = 0.073 , 8 = -20° . Time in prototype scale.
= =

176
 70 ms after water entry

Figure 9.32 . Tail cavity of plate-cylinder model breaking water surface ,

with & = -20° . Time in prototype scale.
-
., nrototype

Distance

-4

-4
Distance

.i, nrototype
cavity
entry
from
point
pentry
iscale

lips
cavity

of
from

Entry point Entry point

ppoint
lips

scale

0 0
of

0 = 0.073
4 P = 0.8
O = 0.073
8 p = 0.2
-

0 = 0.292

12
p = 0.8
-
-

12 0 = 0.073
nrototype

28 p = 0.1

diameter
pmouth

16 28
Width

16
.,iscale

mouth
Width
cavity
of

24
24

20 Tail cavities
,of
F 20
n eotype
1
,pofrototype
20

20

40
Length

prot
iprototype

i.,scal
cavity

Width of 40
cavity mouth Cavity
,Length

length
cavity

32 16 Cavity
scale

32 16
Tail, cavities mouth
., n
of

24 12
12
.scale

Cavity length
24
Cavity
,in

diameter 8
16
16 8
Note : Scale of cavity length Cavity diameter
8 4
curves 1/2 scale of cavity
8 4
diameter curves
0 1 0
0.00 0.02 0.04 0.06 0 0
0.00 0.02 0.04 0.06
Time from water entry, prototype scale , sec
Time from water entry , prototype scale, sec
Figure 9.33 . Plate-cylinder model nose-cavity behavior Figure 9.34. Effect of scaling condition on plate-cylinder model
-
versus tail-cavity behavior, with F = 51.8, o = 0.073 , nose-cavity parameters, with F = 51.8, š = -20 degrees.
p' = 0.8 , š = -20 degrees.

177
CHAPTER 10
EXPLORATORY EARTH -ENTRY STUDIES

The major forces affecting a missile in fairly homogeneous composition , which yielded
earth -entry would presumably be similar to data on missile sand -entry whip (change in pitch
those in water entry and consist of deceleration velocity ) and “ splash .”
forces , cross-forces on the head and stabilizing A box of sand was placed in the Open
forces on the body and tail. It was of interest to Launching Tank ( Figure 10.1 ) . Two nested
apply the techniques developed in water- entry boxes were used ( Figure 10.2 ) , the outer one
studies to earth-entry studies in the NUC Hydro constructed of 3 /4 -inch plywood over a 2-inch
ballistics Laboratory. They consisted of launches angle -iron frame with a plate -glass window on
into clean ungraded beach sand, a simple soil of one side. The lid had an opening in the upper

Launcher

Soil Box
08

Optical Whip
Rotating Disk
Camera
Recorder

Figure 10.1 . Laboratory setup for earth -entry studies.

178
 Figure 10.2 . Nested test boxes .

front to allow missile entry and to admit light The missile was propelled into the sand at a
for the optical whip recorder (Part V) . The nominal velocity of 100 fps from a catapult
interior was painted white to reflect light for launcher (Part V) . The velocity was measured by
photography. The inner box was made of sheet an electronic chronograph actuated by two
metal , 24 inches long, 12 inches wide , and photocell pickups ; the whip of the missile at
12 inches deep . entry was recorded by the optical whip recorder.
The sand was sifted through ordinary Tests were made at 20- and 30 -degree entry
window screen . For dry sand tests , the inside angles into both dry and wet sand. For each
box was filled to the top , packed down , and the condition , three launchings were made : two for
excess sand scraped off to leave a smooth level whip records and one for splash . Prior to
surface . For wet sand tests , the hard-packed launching, the missile was coated with a thin
sand was thoroughly soaked with water, allowed layer of machinist's Prussian blue ; any color
to stand for some time, and the surface rewetted remaining on its surface at recovery was consid
just before firing. This procedure was followed ered evidence of cavity formation. After each
so that the condition of the sand would remain shot , the sand was removed with great care in
reproducible. order not to disturb the missile because its final
The missile was a cylinder 6 inches long position in the box supplied information on
and 1 inch in diameter with a 1 -diameter ogive depth of penetration , approximate trajectory,
nose ( Figure 10.3 ) . It was made of case and average deceleration .
hardened steel hollowed out to reduce the Similar launchings were made into water to
weight to 0.5 pound, with a small rear mirror for allow direct comparisons between water- and
the whip records (Part V) . The center of gravity sand-entry splash characteristics.
was 3.28 inches from the tip of the nose ; the Figures 10.4 , 10.5 , and 10.6 show the
moment of inertia was 1.86 lb in 2 . launching of the missile along a 30 -degree
179
Mirror

Figure 10.3. One-inch -diameter missile with mirror for whip records.

trajectory into wet sand, dry sand , and water, After all launchings into sand, the Prussian
respectively . Similar splash contours occurred blue was gone from the ogive nose and the lower
when the missile entered wet or dry sand , and rear of the model , but the color on the
there was little difference between them and the remaining portions was practically undisturbed,
water-entry splashes. Figure 10.7 shows the indicating that the model must have been
surface of wet sand after the missile was fired enveloped in a cavity during most of the flight.
into it ; Figure 10.8 shows the final position of The color pattern remaining after a 30-degree
the missile after the sand had been removed . launching into dry sand is shown in Figure 10.9 .

180
181
Figure
M10.4
issile
entering
wetsand
30
1degrees
00
.,at
fps
182
3010.5
Figure
at
sand
dry
entering
.Mfps
,1issile
degrees
00
183
10.6
Figure
issile
water
1,entering
degrees
30
Mfps00
.at
 Figure 10.7 . Surface of wet sand after missile impact Figure 10.8 . Sand removed to show missile .
at 30 degrees.

Figure 10.9 . Erosion of 1 -inch-diameter missile bottom by dry sand .

Measurements indicated that sand separation on water can be assumed to leave the surface of the
the head took place about 0.8 projectile diam missile tangentially at separation, it is reasonable
eter aft along the underside of the missile , where to expect the forward portions of the sand and
lateral motion of the sand was restricted . Water water cavities to be similar in shape.
tunnel tests show separation 0.85 diameter aft The missile experienced a positive (nose
when a cavity forms on this missile head shape up ) whip at earth entry . In both wet and dry
in water (Reference 33 ). Since both sand and sand, the whip was greater at the 30-degree than

184
at the 20-degree trajectory angle ( Figure 10.10 ) . the missile tail mirror of the optical whip
Calculations of the expected whip of this recorder (Part V) moved off the camera lens
missile, based on tests of large-scale missiles in before the whip-producing regime terminated.
water and corrected for the density of the sand, However , the pitch velocity curves of Fig
gave a whip of 450 deg/sec . ure 10.10 indicate that the whip is more than
Complete whip records were not obtained 1,500 deg/sec for the 20 -degree entry and
because the trace of the light ray reflected from 2,000 deg/ sec for the 30-degree entry .

Launching angle 20 deg Launching angle 30 deg

2500
Dry sand Wet sand Dry sand Wet sand

2000
sdvelocity
Pitch
eg
/,oec

1500

1000

500
1

0 1 2 0 2 0 1 2 0 1 2

Time, ms

Figure 10.10. Whip measurements, Earth Entry Studies.

185
CHAPTER 11
WATER-EXIT STUDIES

The phenomena that occur when a missile EXPERIMENTAL PROGRAM

exits from water into air differ from those that The studies (References 103 , 104 , and 105 )
occur in water entry , and water exit cannot be were made at the NUC Hydroballistics Labora
regarded as water entry in reverse . The hydro tory in the Variable -Angle Variable -Pressure
dynamic flow pattern , varying as it does during Launching Tank. Three missiles were used . The
exit , causes torques and forces that have no first (used most extensively ) was a simple body
counterpart in water entry . The water-air inter having a grooved hemisphere head with a probe,
face " sees ” a combined missile -cavity shape, a cylindrical midsection, and a spool-shaped tail.
causing even the missile configuration to be a The second missile had a truncated - cone nose
variable . When water-exit is oblique, the force with a probe and a torpedo -type tail. The third
pattern is likely to be asymmetrical , causing a missile, a composite of the other two, combined
refraction of the missile trajectory and changes grooved hemisphere head with torpedo tail.
in orientation and angular velocity . Thus the Probes concentric with the missile axes were
water-exit behavior of a missile depends on installed to permit good attitude measurements
many factors such as missile shape, ballistic when the missiles themselves were obscured by
parameters, velocity , trajectory angle, orienta cavitation .
tion , and the cavity. The missile configurations are shown in
The structural-strength requirements of a Figure 11.1 ; their parameters are given in 1

water-exit missile also differ from those of a Table 11.1 .

water-entry vehicle . Instead of impact as at A specially designed underwater launcher
entry , there is an abrupt reduction of forces on was used to propel the missile without imparting
the missile ; these are likely to be asymmetrical any angle of attack or yaw or angular velocity,
and cause perturbations. and, in addition, without introducing air or any
Studies were made of the ways in which extraneous substances such as hydraulic fluid or
change in trajectory angle and cavitation number grease (Part V).
affect the flight of 2 -inch -diameter missiles In order to remove any effects of abnor
exiting from still water into air. Repeatable mally high air content in the water on cavity
perturbations , however, did not occur under behavior, the air content was reduced to below
presumably equal exit conditions ; evidently that of water at 20°C in equilibrium with air at
some factors must have shown random fluctua 1 -atmosphere pressure. De-aeration was accom
tions. Since studies of random phenomena must plished by bubbling air through the water for at
be statistical in nature, the number of launchings least an hour at 0.1 atmosphere. Tests with a
needed to predict missile performance depends Van Slyke manometric gas analyzer (Refer
upon the confidence level desired. Rather than ence 62 ) showed that the air content of the
concentrating on a few test conditions for de-aerated water was reduced to at least two
statistical analysis, a wide range of conditions thirds that of water at 20°C in equilibrium with
was investigated less thoroughly in order to air at l - atmosphere pressure.
obtain a picture of overall missile performance.

186
 1.0625
1.000

2.125 5.500

1/16 X depth
2.000
0.030
CIG

78 °

2.125 - 0.0625
+ 0.0005
11.000

12.000 1.500

Hemisphere

Truncated cone

Composite

Figure 11.1 . Missile configurations. Dimensions in inches.

187
 TABLE 11.1 . Missile Parameters

Truncated
Parameter Hemisphere Cone Composite

+0.000
Diameter, in . 2.000
-0.001 2.000 +0.000
-0.001
2.000
+0.000
-0.001

Length , aa in . 12.004 + 0.040 12.845 12.90

Mass , lb 1.275 +0.001 1.327 - 1.332 1.268 - 1.275

Distance from CG to
nose , a in.. 5.50 5.00 5.38

Moment of inertia, bb lb-in2 12.018 18.33 15.104

Excluding probe.
b
About aa transverse axis through the CG.

Sideview water -exit data were obtained One-to -one Froude- and cavitation -number
with arotating-disk camera (Part V ) and scaling was assumed for modeling water exit in
Edgerton-type stroboscopic flash lamps adjusted
inclina

to give multiple exposures at a frequency of 250

Missile

70
per second, i.e. , at an interval of 4 ms between
dtion
,i eg

frames. The rotating-disk camera was positioned

65
so that its axis was normal to the plane of the
tank window and passed through the point of
O
,fps ty

60
veloci

intersection of the water surface and the

Missile

100
projected tank centerline. The range of missile
80
positions over the exposure sequence was сооо
о
Tail exit

adjusted by means of a time-delay apparatus 60

40
00400 O O O O TO
o to ‫ܘ ܘ‬
that triggered the flash lamps at an appropriate
distance

3
time after launcher actuation .
travel
Total
nose
surface

Several launchings were made at trajectory

of
water
from

2
angles of 75 and 90 degrees, using two cameras. 0 0
,ft

Perturbations occurred in both planes and were

-

generally of comparable magnitude and

randomness .
The photographic data were reduced and
O

plotted as functions of time from water exit. o o o

1
-

Missile position was measured at the junction of

the probe and the missile nose . The time of
water exit is defined as the instant at which the -2 h
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
junction of the probe and the missile intersects Time from water exit, sec
the plane of the undisturbed water surface.
Figure 11.2 presents data from a typical launch Figure 11.2 . Data from typical underwater launching
ing. of hemisphere missile .
1

188
the turbulent-flow regime where Reynolds cavitation . The test conditions consisted of all
number scaling could be ignored. Gas-density combinations of the following parameters:
scaling, although required for some conditions in
water -entry modeling (Chapters 7 and 9 ) , could Trajectory launching 90 , 75 , 60 , 45 , 30 , 15
angle, & , deg
be ignored here because, at exit , the system of
Cavitation number1 o 0.576 , 0.431 , 0.289 ,
forces acting upon the missile is made up of 0.106 , 0.045
large water forces established during the under
Water-exit velocity , 53 to 63
water trajectory . The density of the water is on Ve, fps
the order of 103 to 104 times that of the tank
atmosphere . Thus , inertial and aerodynamic
pressure-reduction forces that are directly At least three launchings were made at each
test condition .
dependent upon the atmospheric density are not
apt to be significant in altering a force system EXPERIMENTAL RESULTS
already established by the water. It was found
experimentally that variations in the gas-density Hemisphere Missile
coefficient , p ', from 0.20 to 1.18 made no The hemisphere missile trajectories were
significant difference in the water-exit perturba
tion . not greatly perturbed by water exit. A maxi
An exit velocity of 60 fps, the limit of the mum spread of about 1/6 foot (one missile
launcher mechanism, was used . This velocity was diameter ) occurred after about 2 1/2 feet of air
slightly bel ow
below that necessary produce
travel ( Figure 11.3 ) .
to
Perturbations in missile attitude occurring
turbulent flow over a hemispherical missile nose
(Reynolds number 9.7 x 105 , Reference 23 ) . at water exit were observed under all degrees of
The consistent results obtained when a 2 -inch cavitation and at all trajectory angles inves
model with its nose artificially roughened tigated. At first glance, the magnitude and
direction of the perturbations appeared random,
(Chapter 8 ) was operated in a 50- to 60-fps but they assumed a pattern when interpreted in
water -entry velocity range prompted the terms of flows about the missile . The system of
decision to use a missile with a roughened nose
for these tests . water-exit forces is complex , and the net effect
However, preliminary tests of aa hemisphere
missile without groove or probe showed that
distance
Vertical

Š = 90°
surface

75
wo

60 -fps exit velocity did not produce turbulent 3

water
from

60
flow ; instead, it gave rise to sporadic disturb
,ft

45
ances ahead of the zone of cavity separation that
-

2
caused erratic cavities in some instances and 30

hence large exit perturbations.

1

Turbulent flow was triggered artificially by 15

providing a concentric groove cut in the missile
nose at the approximate zone of cavity separa 0
tion, as shown at the top of Figure 11.1 . The
groove helped to stabilize the cavity and prevent
-1
erratic cavity formation .
Launchings of the hemisphere missile with
-2

groove and probe were made over a wide range :

the trajectory angle was varied from 90 degrees
(vertical ) to 15 degrees above the horizontal in -3 -2 -1 0 1 2 3

15-degree increments , and the cavitation num Distance from reference line , ft
bers were selected to give flow regimes varying
from nearly fully wetted to fully enveloping Figure 11.3 . Trajectories of hemisphere missile.

1Cavitation numbers taken at water surface and based on 60-fps water -exit velocity and 22°C water temperature.
189
of these forces upon the missile is delineated by enveloping cavitation occurred during all under
the shape of the cavity, wake, water surface, and water flights. At nose emergence , a small
splash . From study of these, it is possible to amount of water exited on top of the missile
predict the nature of the perturbation nose and accelerated ahead of the missile (Fig.
(Reference 105 ) . ure 11.4 ). At the water surface, most of the
The water-exit perturbation of the hemi cavity was wiped from the missile and stayed
sphere missile was dependent upon both the behind in the water at all cavitation numbers
cavitation number and the trajectory angle , the except one : o = 0.289 (Figure 11.5 ). At this
cavitation number being by far the more impor cavitation number, the cavity was rapidly
tant variable . Typical flow patterns could be collapsing forward just prior to exit , the forward
associated with each cavitation number. They motion apparently giving it sufficient
were modified by the superimposed variation of momentum to break through the water surface
trajectory angle, but they were recognizable at and exit with the missile. This condition was
all trajectory angles. noted at all trajectory angles ( Figure 11.6 ) and
A general similarity was observed in the was generally associated with a small exit
flow patterns of the hemisphere missile at all perturbation. Also , during all launchings, a
test conditions. Cavitation of varying degrees portion of the tail cavity followed the missile
ranging from nearly fully wetted flow to fully through the interface to disintegrate violently

Trajectory Angle , š , deg

90

75 60 45 30

15

Full cavity , o = 0.045

Trajectory Angle , , deg

90 75 60 45 30

15

Partially wetted flow , o = 0.576

Figure 11.4. Cavitation variations, water exit of hemisphere missile.

190
 -1.1 +2.9 +6.9 +10.9 +1 4.9 +18.9

-0.1 +3.9 +7.9 +11.9 +15.9 +19.9

Time From Water Exit , ms

=
Figure 11.5 . Hemisphere missile at water exit , g = 90° . Top : cavity left behind (o = 0.106) .
=
Bottom : cavity exits with missile (g = 0.289 ) .

after exit (Figure 11.7 ). The violence of cavity The water-exit perturbations observed can
disintegration appeared to be about the same at be divided into two types : ( 1 ) those arising
all cavitation numbers ( Figure 11.8 ) . The differ when the missile exits in a fully developed cavity
ing aspects of the flow pattern occurring at each (top of Figure 11.5 ) and (2 ) those arising when
cavitation number are discussed in the missile exits under conditions of partially
Reference 105 . wetted flow (bottom of Figure 11.5 ). The latter
191
 Trajectory angle
90° 75° 60 ° 45°

30°

15 °

Figure 11.6. Hemisphere missile nose -cavity exit ( o = 0.289).

n9
16.2 20.2 24.2 28.2 32.2 36.2 40.2 44.2 48.2

Time From Water Exit, ms

Figure 11.7 . Disintegration of hemisphere missile tail cavity at exit , o = 0.576, g = 75° .
=

192
 Cavitation No.

0.576 0.431 0.289 0.106 0.045

About 28 ms after water exit

Cavitation No.

0.576 0.431 0.289 0.106 0.045

About 42 ms after water exit (one missile length of airflight)

Figure 11.8 . Effect of cavitation number on disintegration of hemisphere missile

=
tail cavity , g = 90° .

193
type of perturbation is modified by the degree continued wide at the 30- and 15 -degree condi
of partial wettedness. When exit occurred in a tions. The perturbation curves for the oblique
full cavity (o = 0.106 and 0.045 ) , the perturba trajectory angles were all of similar shape and
tion was predominantly positive (nose-up ); the tended to become more negative as the trajec
partially wetted condition (o = 0.576 , 0.431 , tory angle became more shallow .
and 0.289 ) was more apt to be associated with a The photographic records reveal the nature
negative down ) perturbation . At o
(nose--down
(nose of the mechanisms causing water -entry perturba
0.431 , where the cavity was particularly tions . When exit occurred in the presence of
unstable and tended to disrupt the water surface fully developed cavitation, the perturbation was
during exit, the scatter in the perturbation was influenced by the flow ahead of the zone of
so great that the perturbation ranged from cavity separation and by the degree of isolation
extremely positive to extremely negative during maintained by the cavity between the missile
presumably duplicate launchings. and the water surface . When the base of the
Figure 11.9 summarizes the perturbations missile was partially wetted, the flow pattern
observed at the various trajectory angles at nose was more difficult to interpret. However, the
exit , tail exit, and 45 ms after nose exit as flow was delineated to some extent by a ring of
functions of cavitation number. The range of vortex cavitation shedding from the missile base.
perturbation as a function of cavitation number It is possible that the perturbation was caused
occurring at any given trajectory angle increased by the shedding of the vortex (Reference 80).
slowly as the trajectory became more oblique The position of the shed vortex indicates the
until the 45 -degree trajectory angle was reached. direction of missile perturbation. If the vortex
Then the range became noticeably larger and ring trailed from the left of the missile tail

Nose exit Tail exit 34 ms after nose exit

90 90 90

80 80 80
inclination

70 70 | 70
Missile
i,deg

60 60 60

50 50 50

40 40 40

30 30 30
20

20 20

10 10 10

0 1 0 1 0 1
0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8
Cavitation No. , oO Cavitation No. , o Cavitation No. , O

Figure 11.9 . Summary of hemisphere missile water-exit perturbations at various trajectory angles.

194
during exit, the perturbation was negative ; a The large scatter typical of the data at o =
right-trailing vortex caused a positive perturba 0.431 for the steeper oblique trajectory angles
tion (Figure 11.10 ). did not occur at the 30 -degree trajectory angle
The great randomness in perturbation char ( Figures 11.19 and 11.20 ). An extremely large
acteristic of the oblique launchings made at o = nose-up perturbation occurred at o = 0.045 , but
=

0.431 occurred because the weak cavitation at o = 0.106 the growth of the perturbation was
system produced unrepeatable hydrodynamic checked by the missile dragging in the water
flow patterns during both the underwater trajec during exit .
tory and the water exit . At the 15 -degree trajectory angle, the
At the 90 -degree trajectory angle , the obliqueness of the trajectory was most impor
perturbation was small and seemed to grow tant in determining the perturbation (Fig
predominantly from conditions established prior ures 11.21 and 11.22 ) . The partially wetted
to exit. The trends and behavior patterns typical missile dragged heavily in the water during exit ;
of the oblique trajectory angle did not appear. even the full cavity did not completely protect
Figure 11.11 shows parts of five launchings, one the missile from water contact. The nose-up
=
at each cavitation number, made at the motion of the missile was arrested at o = 0.045
90 -degree trajectory angle. The missile attitude and reversed at o = 0.106 , and a nose -down
as a function of time and approximate distance perturbation occurred at all the higher cavitation
from water exit is shown in Figure 11.12 . numbers .
At the 75 -degree trajectory angle ( Fig. The randomness in water exit observed
ures 11.13 and 11.14 ) , the scatter at the two during this study may have been caused by
higher cavitation numbers was so great that the ( 1 ) unstable cavity flows, (2 ) missile shape, or
perturbations overlapped those at the three (3 ) Reynolds-number effects either in the form
lower numbers. At o = 0.289 (partially wetted) of unstable cavity separation or in cavitating
the perturbation was negative, and at o = 0.045 vortices that shed into the wake and thereby
and o = 0.106 ( fully cavitating) , the perturba increase its activity. Each of these three condi
tion was positive. Two launchings from each of tions was examined in detail and further launch
the partially wetted conditions are included in ings were made to illuminate the last two of the
Figure 11.13 to show how variation in perturba three problem areas.
tion grows from differences in the underwater The blunt missile base used in these studies
and exit-flow patterns. was expected to have a large underwater drag,
Detailed data from the 60 -degree trajectory but the large and random exit perturbations
angle are presented in Figures 11.15 and 11.16. were not foreseen . The possibility that the probe
The scatter at the o = 0.431 condition was so exerts an adverse effect upon cavity formation
great that theseOtherwise
perturbations overlapped the by feeding sporadic vortices into the boundary
other data. , the perturbation layer ahead of the cavity separation zone,
occurring under fully developed cavitation was altering the pressure distribution sufficiently to
positive with respect to the partially wetted flow distort the line of cavity separation and changing
perturbation . the cavity configuration, was studied. Without a
The range of perturbations observed at probe ( first part of Figure 11.23 ), the ungrooved
45 -degree trajectory angle was noticeably wider hemisphere missile was enveloped in a clear
than at the steeper angles ( Figures 11.17 and cavity that did not fluctuate significantly with
11.18 ) . Scatter still occurred at o = 0.431 and to = =
time (ve = 56.7 fps, o = 0.052 ) . With a probe
a lesser degree at o = 0.289 , but the difference (second and third parts of Figure 11.23 ) , the
between the fully cavitating perturbation and cavity was either rough and opaque, fluctuating
= =
that of the partially wetted flow condition was slightly with time (ve = 63.3 fps, o = 0.038 ) or
great enough that the data from o = 0.431 no
=
erratic and flapping, with large protuberances in
longer overlapped the other results. The flow the wall and a shifting line of separation.
pattern determining the perturbation at It had been hoped that the 60-fps exit
0.431 was particularly well delineated (Fig velocity would be adequate to prevent transition
ure 11.17 ). Strong differences occurred both in from laminar to turbulent flow and thereby
the wake and in the ring of cavitation shedding avoid Reynolds-number effects, but full turbu
from the nose cavity . lence apparently did not occur in the flow ahead
195
 is
Perturbation
.
left
trail
to
right
from
separating
vortex
Second -down
nose
left
from
separating
Vortex and
is
during
exit
of
most
persists
vortex
This
trail
to
right ,
perturbation
affect
to
position
in

man
2.4 6.4 10.4 14.4 18.4 22.4 50.4
54.4
Time
From
Water
Exit
,m s

=
fps 5
05=,v°,$4e,514
5.0
Perturbation
is
nose
u
- p
.
right
trail
to
left
from
separating
vortex
Second

196
from
separating
Vortex in
is
and
exit
persists
of
most
during
vortex
This
left
trail
to
right .
perturbation
affect
to
position

2.4 6.4 10.4 14.4 18.4 22.4 50.4 54.4

From
Water
,Time
Exit
m s

== =
fps 5
.470
=5v,0°$4e7.5

11.10
convention
Light
-rFigure
eft
identifying
patterns
vortex
hemisphere
in
launchings
.missile
 0=
o.289
0
=.576
o
201
-13.4 -1.4 +10.6 +22.6 +34.6
+46.6 -12.1 -0.1 +11.9 +23.9 +35.9 +47.9

0.045
=
o

197
0
=
o.431 0
=
o.106

-16.2 -4.2 +7.8 +43.8

+31.8
+19.8 -13.1 -1.1 +10.9 +22.9 +34.9 +46.9 -13.1 -1.1 +34.9
+22.9
+10.9
+46.9

Time
From
Water
Exit
,ms Time
mFrom
Water
,Exit
s Time
From
Water
Exit
m
,s

hemisphere
11.11
Figure
in
perturbations
.S°missile
,g=9mall
exit
water
0
 1

105

o
Atm

0.1 0.045
100 0.2 0.106
dinclination 0.5 0.289
0.75 0.431
Missile

1.0 0.576
,i eg

95

90
a7
85

80

75 1

-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06

Time from water exit, sec

-1 0 1 2

Approximate distance from water exit , ft

Figure 11.12 . Plots of hemisphere missile water exits of Figure 11.11 , š = 90 degrees.
-

!
of the zone of cavity separation. The viscosity ef velocity decreases along the trajectory . More
fects manifested themselves either in the form of over, a fairing would defeat the original purpose
unstable cavity separation and shape or in large, of the probe , which was to obtain accurate
sporadic vortices shedding into the wake. The missile-attitude data.
addition of the nose probe, which was necessary
if adequate missile attitude data were to be Truncated-Cone Missile
taken , seemed to increase the instability of the When the truncated-cone missile was

hydrodynamic flow still further. launched (Figure 11.1 and Table 11.1 ) , its nose
The flow over the missile nose was made shape provided an immediate discontinuity for
visible with nigrosine dye paste introduced cavity separation , and its faired torpedo -type tail
through slots in the probes . During launching stabilized the flow about the missile afterbody.
and subsequent underwater trajectory, the dye The missile was launched with ( 1 ) the standard
streamed from the slots, marking the flow along 1 1 /2-inch probe , (2 ) the standard probe
the probe and over the missile nose. Minute shortened to 1 inch, and (3 ) no probe. All
vortices were visible in the dye trace and a zone launchings were made at a 45 -degree trajectory
of separation seemed to occur at the base of the angle, some at 70 fps and some at the maximum
probe ( Figure 11.24 ) . Adding a nylon fairing to obtainable velocity of about 80 fps. Typical
the probe to fill the approximate zone of photographs of missile launchings with the
separation tended to stabilize the cavity ( Fig. standard probe were made at constant Froude
ure 11.25 ) , suggesting that the erratic cavity number and several different cavitation numbers
formation was being caused by disturbances in (Figure 11.26 ). Missile water-exit perturbations
the flow introduced by separation of the probe can be seen in Figure 11.27 .
=
boundary layer. Using a fairing to eliminate Ato = 0.195 , perturbations from 12 launch
erratic cavities is not feasible because the shape ings ranged from 0 to -1/2 degrees , a perturba
of the separation zone varies as the missile tion of low magnitude. The photographs showed
198

1
 o = 0.576 ( nom. ) o = 0.576 ( nom.)

-15.8 -3.8 +8.2 +20.2 +32.2 +44.2 -15.7 -3.7 +8.3 +2.03 +32.3 +44.3

o = 0.431 ( nom.) o = 0.431

-13.7 -1.7 +10.3 +22.3 +34.3 +46.3 -16.5 -4.5 +7.5 +19.5 +31.5 +43.5

o = 0.289 o = 0.289

-14.9 -2.9 +9.1 +21.1 +33.1 +45.1 -13.4 -1.4 +10.6 +22.6 +34.6 +46.6

o = 0.045
o = 0.106

-14.8 -2.8 +9.2 +21.2 +33.2 +45.2 -12.6 -0.6 +11.4 +23.4 +35.4 +47.4

Time From Water Exit , ms Time From Water Exit , ms

Figure 11.13. Perturbations in hemisphere missile water exit, g = 75° .

199
 :

90

Atm

6
0.1 0.045
85
inclination
-
0.2 0.106
* 0.5 0.289
Missile

0.75 0.431
,ideg

1.0 0.576
80

75 TA
70

65

60
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06
Time from water exit , sec

-1 0 1 2 1

Approximate distance from water exit, ft

Figure 11.14. Plots of hemisphere missile water exits of Figure 11.13 , š = 75 degrees.

the nose cavity rapidly collapsing forward as the Composite Missile

missile approached the water surface, and, during The grooved hemisphere head was com
exit , bursting through the surface instead of re bined with the torpedo tail to make a composite
maining behind in the water. The velocity field missile ( Figure 11.1 and Table 11.1 ), and a few
generated by the collapsing cavity apparently launchings were made at the 45 -degree trajec
prevented interaction between the missile and tory angle, at water -exit velocities ranging from
the water surface . 60 to 75 fps, and at various cavitation numbers.
The large scatter occurring in the perturba Typical photographs of water exit are presented
tion at o = 0.078 appeared to be caused by in Figure 11.29 , and the missile attitude is
interaction between the water surface and the shown as a function of cavitation number in
tail fins or by the flat nose cutting into the Figure 11.30 .Figure 11.31 illustrates the
water surface and pushing water ahead. smooth separation of the vortices from the
The presence of the probe caused no tapered missile tail.
significant difference either in the perturbation The composite missile's behavior was
( Figure 11.27 ) or in the flow patterns about the similar to that of the truncated -cone missile .
missile (Figure 11.28 ), in contrast to the hemi Scatter occurred in the exit perturbations, but it
sphere missile, which experienced a change in was of lesser magnitude and the randomness was
flow pattern when the probe was added. not detectable until late in exit, suggesting that
Apparently , any disturbance in flow introduced it was not nose-induced .
by the probe was overridden by the large
disturbance created by the nose flat.
A 10% variation in exit velocity did not SUMMARY
cause any significant variation in either perturba The following conclusions can be drawn
tion or flow pattern . from the water-exit study.
200
 o = 0.576 ( nom. ) o = 0.576 ( nom.)

-16.2 -4.2
h +7.8 +19.9 +31.8 +43.8 -16.2
w124
-4.2 +7.8 +19.8 +31.8 +43.8

o = 0.431 o = 0.431

-14.2
IT
-2.2 +9.8 +21.8 +33.8 +45.8 -13.6 -1.6 +10.4 +22.4 +34.4 +46.4

o = 0.289 o = 0.106

AT
-15.4 -3.4 +8.6 +20.6 +32.6 +44.6 -13.7 -1.7 +10.3 +22.3 +34.3 +46.3

Time From Water Exit , ms

o = 0.045

wali
-13.9 -1.9 +10.1 +22.2 +34.1

Time From Water Exit , ms

+46.1

Figure 11.15 . Perturbations in hemisphere missile water exit , $ = 60° .

201
 75

Atm o

0.1 0.045
70
0.2 0.106
inclination
Missile 0.5 0.289
0.75 0.431
i,deg

65 1.0 0.576

60

55

50

45 1
-0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 0.06

Time from water exit, sec

-1 0 1 2
Approximate distance from water exit , ft

-
Figure 11.16 . Plots of hemisphere missile water exits of Figure 11.15 , g = 60 degrees.

1.
The trajectories of the missiles were perturbations of the standard missile
not greatly disturbed at water exit. A measured during this study are sum
maximum deviation of about marized in Figure 11.9 . The perturba
1/2 diameter occurred after tion range at any given trajectory
15 diameters of air travel . angle was less when the trajectory was
2
.

2. Water -exit perturbations in missile steeper than 45 degrees.

attitude were observed under con 4. Repeatable water-exit perturbations
ditions of flow ranging from nearly did not arise from presumably dupli
fully wetted to completely enveloping cate launchings.
.
5

cavitation and at all trajectory angles A faired torpedo -type tail greatly
investigated ( 15 to 90 degrees with reduced the magnitude and random
respect to the water surface ) . ness of perturbations occurring when
3. Cavitation number was more impor the missile base was wetted .
tant in determining the magnitude and 6. An artificially maintained cavity, or a
direction of the perturbation than was cavity collapsing under controlled
the trajectory angle . When the missile conditions , or both , may be a way to
exited under conditions of fully devel limit the magnitude and randomness
oped cavitation , the perturbation was of the exit perturbation .
predominantly positive (nose up ) ; 7. The basic design rule for minimizing
when exit occurred under conditions water -exit perturbations is that care
of partially wetted flow , the perturba must be taken to avoid anything –
tion tended to be negative (nose missile shape or ambient condition -

down ). Maximum perturbations of that tends either to produce an

+13 degrees were recorded after about unstable hydrodynamic flow pattern
13 diameters of air travel . The
or to magnify one that has occurred.
202
 o=.576
0).(n om o=.289
0.)(n om

-13.5 +10.5
-1.5 +22.5 +34.5 +46.5 -3.4
-15.4 +8.6 +20.6 +32.6 +44.6

=
n om
0.)(o=.431
LU o=.289
0.)(n om
2

203
-13.6
-1.6 +10.4 +22.4 +34.4 +46.4 -14.5 -2.5 +9.5 +21.5 +33.5 +45.5

(o=.431
0).n om 0o=.045
).(n om
=
o=.106
0.)(n om

-13.6
-1.6 +22.4
+10.4
+34.4 +46.4 -14.4
-2.4 +9.6 -14.6
-2.6
+21.6 +33.6 +45.6 +9.4 +21.4 +33.4
+45.4

Time
mFrom
Water
,Exit
s Time
mFrom
Water
,Exit
s Time
From
Water
Exit
,m s

Figure
P11.17. erturbations
hemisphere
missile
water
exit
°4,&=in5
 60

Atm

la *
0.1 0.045
55
0.2 0.106
inclination

0.5 0.289
Missile

0.75 0.431
50

1.0 0.576
i,d eg

45

40

35

30 1 1

-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06

Time from water exit, sec

-1 0 1 2

Approximate distance from water exit , ft

Figure 11.18. Plots of hemisphere missile water exits of Figure 11.17 , 8 = 45 degrees.

204
 o = 0.576

-13.4 -1.4 +10.6 +22.6 +34.6 +46.6

o = 0.431

-15.9 -3.9 +8.1 +20.1 +32.1 +34.1

o = 0.289
=

-15.3 -3.3 +8.7 +20.7 +32.7 +44.7

o = 0.106

-13.5 -1.5 +10.5 +22.5 +34.5 +46.5

o = 0.045

-14.7 -2.7 +9.3 +21.3 +33.3 +45.3

o = 0.045

-15.0 -3.0 +9.0 +21.0 +33.0 +45.0

Time From Water Exit , ms

Figure 11.19 . Perturbations in hemisphere missile water exit , g = 30° .

205
 45
저
Atm 0

b!
40 0.1 0.045
0.2 0.106
dinclination

0.5 0.289
Missile

.
0.75 0.431
,l eg

35 1.0 0.576

30
न्यान

25

20

15 1 1

-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06

Time from water exit, sec

-1 0 1 2

Approximate distance from water exit, ft

Figure 11.20. Plots of hemisphere missile water exits of Figure 11.19 , 8 = 30 degrees.

206
 0
0
=
o.289
=.576
O

-18.3 -6.3 +5.7 +17.7 +29.7 +41.7 -16.6 -4.6 +7.4 +19.4 +31.4 +43.4

076
=.5
o 0.106
=
o

-15.6 -3.6 +8.4 +20.4 +32.4 +44.4 -16.2 -4.2 +7.8 +19.8 +31.8 +43.8

0
=.431
o 0
=.106
o

-17.8 -5.8 +6.2 +18.2 +30.2 +42.2 -16.4 -4.4 +7.6 +19.6 +31.6 +43.6

207
0
=.045
o
0
=
o.431
-

-15.2 -3.2 +8.8 +20.8 +32.8 +44.8 -15.5 -3.5 +8.5 +20.5 +32.5 +44.5

0
=
o.045
0
=
o.289

-16.8 -4.8 +7.2 +19.2 +31.2 +43.2 -17.9 -5.9 +6.1 +18.1 +30.1 +42.1

Time
From
Water
Exit
,ms Time
From
Water
Exit
m
,s

missile
,&=
1inexit
5
P°.water
hemisphere
erturbations
11.21
Figure
 30

Atm 19

25 0.1 0.045
inclination

0.2 0.105
O 0.5 0.289
Missile

0.75 0.431
,i eg

20 ----
1.0 0.576
d

15

10

0 1
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06

Time from water exit , sec

1
-1 0 1 2

Approximate distance from water exit, ft

Figure 11.22 . Plots of hemisphere missile water exits of Figure 11.21 ,

Š = 15 degrees.

208
 Without probe
Ve = 56.7 fps
o = 0.052

With probe With probe

ve = 63.3 fps ve = 63.3 fps
=

o = 0.038 o = 0.045

Figure 11.23 . Effect of probe on ungrooved hemisphere missile cavity configuration , g = 30° .
=

209
i

Figure 11.24. Dye trace from slotted probe marking large zone of separation on
ungrooved hemisphere missile nose.

210
Figure 11.25 . Cavity stabilized by nylon fairing on ungrooved hemisphere missile nose.

211
 ir
o = 0.451

JI
o = 0.078
ir

o = 0.195

o=
= 0.033

Figure 11.26. Effect of cavitation number on flow pattern and water exit of truncated-cone missile ,
t = 45° , ve = 68 to 73 fps.

212
inclination
Missile

55 Without Probe
,

o Tail oriented with

50

one pair of fins nor

8əр,“

mal to water surface

* Tail unoriented
45
inclination

40
E
Missile

55
With 1 - in. Probe
,

o Tail oriented with one

50
pair of fins normal to
8əр“

water surface
45
inclination

40
Missile

55
With Probe
,

Random orientation
i,deg

50
of tail
45

40
0.0 0.2 04 006 0.8 1.0 1.2 1.4 1.6

Cavitation number , o

Figure 11.27 . Truncated - cone missile perturbations com

pared with hemisphere missile perturbation range ( shaded
areas ) . & = 45 degrees.

213
 Standard probe ( 1 1/2 in. )

24 1 -in. probe

GA No probe

o = 0.195

Standard probe ( 1 1/2 in. )

Wi*L

!! 1 -in . probe

No probe

No probe

o=
= 0.078

Figure 11.28 . Effect of probe length on flow pattern and water exit of
truncated -cone missile, g = 45° , ve = 70 to 73 fps.
=

214
 Ve = 60 fps
o = 0.045

Ve = 64 fps
ve
o = 0.091

=
ve = 73 fps
o = 0.194

Ve = 77 fps
o = 0.261
Ve = 75 fps
0 = 0.369

=
Figure 11.29. Effect of cavitation number on flow pattern and water exit of composite missile, š = 45° .

215
 65

60

dinclination

ох
Missile
R = 9.8 to 12.8 x 105
,i eg = 24.4 x 105
R -
55 Х

50

45

40

35

0.0 0.2 0.4 0.6 0.8 1.0

Cavitation No. , O

Figure 11.30 . Comparison of missile inclination about 242 feet

after water exit, composite missile versus hemisphere missile
( shaded area) as affected by cavitation number . g = 45 degrees.

wah
Figure 11.31 . Smooth separation of vortices from tapered tail of
composite missile. g = 45° , 0 = 0.369 , v e. = 75 fps.

216
 PART IV.
PHENOMENOLOGICAL STUDIES

CHAPTER 12
WATER-ENTRY SPLASH CONTOURS

The contour of a missile's water-entry The spark flash timing sequence

splash can be helpful in interpreting water-entry initiated by a time-delay generator that was
phenomena and in developing water-entry actuated when the falling sphere interrupted a
theory. However, the transparency of water photocell beam above the water ; it was termi
sometimes prevents clear photographs of the nated immediately after the first half of the
splash contours by obscuring the outline of the sphere crossed the water surface. Any amount of
splash. To avoid this difficulty, a study using delay between 0 and 1 second could be intro
shadow photography was undertaken. duced into the sequence in increments of 0.0001
The splash contours formed in the vertical second, allowing the sphere to be photographed
and oblique entires of 1- and 2-inch stainless- at various penetrations and thus making it
steel ball bearings were studied in a NUC possible to follow the growth of the water-entry
Hydroballistics Laboratory tank about 10 by 20 splash .
by 5 inches deep and filled with water until the In the first series of tests, the steel sphere
surface rose above the edges to form a negative was held above the water by an electromagnet
meniscus. Shadowgraphs of the sphere water and dropped vertically by opening the magnet
entries ( Figure 12.1 and 12.2 ) were made by circuit. The height of the drop was varied to give
recording the images cast by a collimated spark several entry velocities. In the second series, the
beam upon a photographic film placed 5 inches 2-inch diameter sphere entered the water at
behind the entry point and parallel to the entry oblique angles. The sphere was suspended by a
trajectory. The spark , oriented so that its beam fine wire 13.5 feet long and swung into the water
was normal to the plane of the film , was in front like a pendulum ( Figure 12.3 ) . Entry velocity
of the entry point and precisely aligned with the was determined by measuring the distance
water surface. The spark was generated between traveled between two exposures recorded on a
two carbon electrodes with an 0.030-inch gap single film at an interval of 0.001 second. The
and had a duration of about 1 us . deviation of the arc trajectory from a straight

1Data taken from Reference 108..

217
 LA 1

Figure 12.1 . Vertical water -entry shadowgraph of 2 -inch sphere ; Ve = 22 fps.

line was considered to be negligible over such a entry of a sphere at a low velocity . The splash
short trajectory. Oblique entry angles of 7 , 12 , resembles the one obtained by Worthington
18, 24 , and 47 degrees were used, with entry ( Reference 112 ) under similar circumstances in
velocities varying from 16 to 30 fps. his Series X photographs. The appearance of the
The typical basket splash generated by the splash at closure with its spray of droplets
vertical entry of a sphere is shown in Fig. resembles the one in Figure 3.7 (Chapter 3 ) ,
ure 12.1 . The splash has two distinct parts : the which also was reproduced from Worthington's
base, which extends from the original water level work .
up to the zone of maximum splash contour When a sphere enters water at an oblique
constriction , and the sheath, which includes the angle ( Figure 12.4 ) , the forward splash is small
layer of water about the sphere above the zone and thin, the forward surface rise is less , and
of maximum constriction and the spray of water there is a relatively large rear surface rise that
leaving the sphere. The pressure of the sheath is develops into a marked hump at low entry
practically atmospheric, and it is believed that angles. The rear hump , as such , had not been
the sheath has little effect upon the sphere since noted previously, but can be seen vaguely in
it is beyond the point of maximum pressure. In Figure 3.28 in Chapter 3 (taken from Reference
Reference 108 , the curvature of the splash 60 ) . Photographs of the rear splash made from a
contour was found to approximate a parabola point directly behind the sphere show two
with its axis inclined to the horizontal. The distinct types of rear-surface rise. At the lower
equation of the parabola and associated entry velocities ( 16 fps), the hump rises as a
parameters were experimentally derived as func single column at the first instant of water
tions of the distance of penetration . contact and persists until the sphere is com
The sequence in Figure 12.2 shows a pletely immersed (top of Figure 12.5 ). At higher
typical “ sheath ” splash produced by the vertical entry velocities ( 30 fps ), two separate humps
218
Figure 12.2 . Vertical water -entry shadowgraph of 1 -inch sphere ; Ve = 5.6 fps.

219
 form (bottom of Figure 12.5 ) , move toward
Entry each other, coalesce , and finally rise together in
Side view angle
a single column as at the lower velocity .
Photographs of the front of the sphere
( Figure 12.6) show that very little water was
thrown off to the sides during water entry and
that no discontinuity existed in the front surface
Sphere Film
of the splash contour. The sideview shadow
graph thus is seen to represent a true cross
Tank section of the splash through the center of the
sphere . All shadowgraph records are sequences
of single exposures, each obtained from a
different launching .
Reference 108 compares experimental
vertical-entry splash contours and theoretical
Front view approximations. A comparison is also made of
experimental forward splash contours obtained
in oblique water entry of spheres and approxi
Spark mate theory developed for the oblique entry of
000 cylinders.
Collimated beam
Film The shadowgraph method is more advanta
0 Tank geous than conventional photography in the
ele recording of actual splash contours . However, it
should not be assumed that a central cross
sectional view is being obtained without
Figure 12.3 . Oblique water-entry shadowgraph apparatus supporting evidence from front and rearview
( not to scale ). photographs.

-
Figure 12.4. Oblique water -entry shadowgraph of 2 -inch sphere ; Ve = 20 fps, g = -7 ° .
=

220
 =
Ve = 16 fps

Ve = 30 fps

Figure 12.5 . Two distinct types of rear -surface hump formed during oblique water entry , š = -10° . Photographed by view camera
directly behind 2-inch sphere.

Figure 12.6. Oblique water entries photographed by view camera in front of 2-inch sphere ;
Ve = 20 fps, š = -14 °.

221
CHAPTER 13
WAKE FORMATION

When a body moves through a fluid under effects governed by the Reynolds number are
fully wetted flow , a wake of turbulent fluid is concerned .
formed . The turbulence arises from vortices The experimental technique not only
generated in fluid that has passed near the delineates the flow at the interface but marks
surface of the body, i.e. , the boundary layer, as the flow during the underwater flight by
well as from vortices that may be formed at the entrained oil . Photographs of vortex and jet
after end of the body , particularly if it is blunt. formation obtained by this technique add to the
Since the wake contains momentum that is not understanding of these complex phenomena.
recovered by the body, there will be a body The studies described here were conducted
drag, and the extent of the wake is an index of at the NUC Hydroballistics Laboratory in the
the drag. Skin friction forces and pressure Liquid Interface Tank (Part V) . A description of
reduction at the after end of the body cause the the apparatus and experimental techniques
drag. In addition, vortex shedding may cause the involved is given in Reference 90.
flow about the body to become perturbed , A 1 -inch-diameter steel sphere and a
resulting in fluctuations in the pressure distribu 0.75 -inch - diameter steel plate cylinder with
tion , which in turn can cause fluctuations in the circular cross section were launched in the
missile direction of travel, velocity, and water-filled tank topped with a layer of high
orientation . grade mineral oil having a kinematic viscosity of
Extensive studies have been made of the approximately 75% of that of air. The oil was
wake under conditions of fully wetted flow , but dyed red for contrast.
very little is known regarding its formation and The physical properties of the two liquids
behavior as a missile crosses an air-water inter were as follows:
face . The speed of most air-to-water missiles is Oil
Water
sufficiently high that water entry takes place Relative density 0.996 0.837
under conditions of cavitation . The shape of the
cavity is known to affect missile performance, Viscosity , lb/ (ft sec) 0.54 x 10-3 6.4 x 10-3
and in some cases re -entrant jets are known to Kinematic viscosity , 0.87 x 10-5 12.2 x 10-5
have formed, which could impinge on the after ft2 /sec
end of the missile, thereby deflecting it and Vapor pressure , torrs 31.8 0.5
possibly damaging it. Thus, the character of the Temperature, °C 30 30
water- entry cavity can be influenced by wake
formation and behavior. The missiles started from rest beneath the
The hydrodynamic flow about aa missile can surface of the oil and traveled through the
be studied to advantage through the use of an interface. Velocities ranging from 30 to 244
oil-water interface . The presence of oil elimi in/sec at the interface and ambient pressures of
nates the water-entry splash and renders the 737 and 25 torrs absolute resulted in both fully
flow more visible. The kinematic viscosity of the wetted and cavitating flows.
oil can be adjusted to that of air so that the flow As a solid body traverses a liquid, at all
in the neighborhood of a missile will simulate except creep velocities, i.e. , at Reynolds num
that obtained in an air -water interface insofar as bers much less than 1 , the flow separates behind
222
the moving body and a wake is formed, the bounding the wake narrow abruptly. At these
nature of which is determined by the Reynolds higher Reynolds numbers, the wake is charac
number of the flow as well as by body shape . terized by completely turbulent flow ;
The change in wake with increase in Reynolds periodicity or pattern has vanished. Further
number follows the same general pattern for increase in Reynolds number causes little
cylinders, spheres , and other blunt bodies, the apparent change in the wake until the onset of
actual wake pattern associated with a particular cavitation , although a slight and gradual
Reynolds number being influenced by body broadening of the wake may occur if the
shape , general turbulence level of the flow , etc. separation zone creeps forward.
The first appearance of the wake occurs as In the quasi-two-dimensional wake formed
the streamlines broaden about the body , separat by long, thin bodies (cylinders , airfoils , etc. )
ing to allow a volume of disturbed fluid to form oriented with their long axes normal to the flow ,
immediately behind it. Separation begins near the vortex pattern within the wake is quite
the point where the velocity of flow around the regular. At the onset of wake formation, the
body is decreasing most rapidly . In the case of circulatory flow forms a vortex pair. The sta
the circular cylinder ( Figure 13.1 ) , separation tionary vortex pair is permitted to exist behind
begins near the rear stagnation point A. As the the body by unstable equilibrium ; any dis
Reynolds number increases, the separation zone turbance that spoils the symmetry of flow will
BB' moves forward to an almost stationary cause the vortex pair to shed. As the Reynolds
position, allowing the streamlines BC and B'C number increases, the vortex pair elongates in
that bound the wake to broaden further. The the direction of motion until it becomes
flow within the wake then becomes apparent . unstable . Then a series of vortices is shed from
An inflow exists along the axis of sym alternate sides of the body , forming the well
metry of the wake and an outflow exists in the known Kármán vortex street in the wake down
other portions. The circulatory motion produces stream (References 4 and 70) . The frequency of
a vortex configuration in the wake. The fluid in vortex shedding increases with Reynolds number
the wake is isolated from the main flow by until the wake becomes so disturbed that the
layers of vorticity generated in the streamline pattern is lost. For a circular cylinder normal to
bounding the body. The vorticity diffuses par the flow , periodicity disappears from the wake
tially into the main flow but feeds primarily into above Reynolds numbers of approximately 105 .
the wake, causing it to elongate as the Reynolds Extrapolation of the two -dimensional case
number increases until the wake becomes suggests that a vortex ring will form within the
unstable. Vorticity is then discharged periodi wake of an axisymmetric body. Experimental
cally into the main stream, where it is dissipated data substantiate this, but the vortex ring either
by turbulent mixing. When the Reynolds num within the wake itself or in aa shed array appears
ber has increased sufficiently to cause a turbu to be much less stable than the two-dimensional
lent boundary layer on the body, the separation vortex pair or Kármán vortex street (References
zone shifts to the rear and the streamlines 24, 35 , and 70) . Observation of the wake
formed by a circular disk ( Reference 24) proved
it to be composed of distorted loops of vorticity
arranged with some symmetry . The flows
reported in Chapter 14 also indicate a
periodicity in the discharge of vorticity in the
wake of aa three -dimensional body.
B
The growth of a vortex configuration in a
A С
wake can be seen behind the sphere in Fig
ure 13.2 . Rotation of the fluid within the wake
B'
is indicated by arrows in Figure 13.3 . The vortex
appears to be a somewhat irregular ring sta
tionary within the wake.
The vortex soon elongates in the direction
Figure 13.1 . Wake behind aa circular cylinder of the main stream, probably because of the
at a low Reynolds number ( R < 50 ). abrupt change in Reynolds number that occurs
223
 35.0 39.2

43.3 47.5

Approximate time , ms

Figure 13.2 . Growth of vortex in wake. Conditions : v (crossing oil-water interface ) = 30 in/sec ; Pa = 737 torrs;
F == 1.5 ; R x 10-4 = 0.17 for oil and 2.4 for water ; o = 406 for oil and 326 for water .
S =

224
 a Sheath splash of oil narrows as
sphere drops . Flow within splash
appears to impinge on surface.
Splash finally separates ; upper
portion flows back into oil layer,
Jower portion is shed from sphere .

b Sphere is followed by turbulent wake

of oil that lengthens and is finally shed
in strands and droplets. Wake appears
to move upward in outer area and down
ward in inner area .

Oil -water interface

1
b

Motion of sphere

Diagram of flow Colored oil - water wake

Figure 13.3 . Direction of flow within wake .

at the interface . With the Reynolds number may increase the stability of the vortices. How
0.17 X 104 in oil and 2.4 X 104 in water, the ever, the growth , shedding, and subsequent
external configuration of the wake would be decay of a cavitation vortex formed on a sphere
changed , even though the flow in the oil started impulsively from rest in water ( Fig.
entrapped in the wake would remain at the ure 13.4 ) differs little from that of the oil
lower R. The Reynolds number of the flow vortex. Both initially approximate a torus and
where the two liquids meet is not well defined . subsequently decay into irregular loops. The
Since a vortex ring is unusual in a wake cavitation vortex , being of lesser mass and also
composed entirely of water , the presence of oil of vaporized fluid, decays more rapidly.
225
 Figure 13.4 . Cavitating vortex from a steel sphere started impulsively from rest under water.

More prominent vortices appear at higher Cavitation will occur in the region of separation
velocities (Figure 13.5 ). A large vortex con when the ambient pressure falls below the vapor
figuration forms in the wake and is subsequently pressure of the fluid medium.
shed, the momentum imparted by the shedding The first evidence of cavitation is not a
vortex causing the sphere to deviate noticeably visible change in the flow , but rather an abrupt
from its trajectory . increase in the flow noise. The first visible sign
Vorticity diffuses into the main flow from of cavitation (incipient cavitation ) is the tiny
the streamline bounding the wake, resulting in a bubbles along the streamline bounding the wake
" pineapple top ” wake configuration (Fig ( Figure 13.9 from Reference 28) , which are
ure 13.6). Fragments of fluid peel from the believed to occur at minute centers of vorticity ;
wake into the main flow soon after it separates a dye was used to delineate the shear layer. As
from the sphere, travel along the streamline the stream velocity increases further, cavitation
bounding the wake until, at the upstream end, grows in the wake until the entire wake area is
they separate and move away with the main occupied, as shown in CIT water-tunnel tests of
flow instead of recirculating in the wake. a torpedo model ( Figure 13.10 from Reference
Figures 13.2 and 13.7 , of sphere launchings 31 ). The size of the cavity is, of course,
at 30 and 158 in/sec, respectively, show the influenced by missile shape but is primarily
behavior of the interface soon after passage of determined by the cavitation number, which
the sphere and its wake. Studies with the thus replaces the Reynolds number as the index
plate-cylinder missile ( Figure 13.8) show that a of flow .
slight angle of attack introduces a marked At the downstream endl of a cavity, the
asymmetry into the flow . Vorticity can be seen flow in the streamlines in and near the cavity
in the larger wake area, and the ragged bits of oil wall may reverse direction and turn inward to
peeling from the wake appear to be small form a re - entrant jet through the cavity
fragments of vorticity trapped in the main flow . opposing the direction of the main flow (Fig
Once a fully turbulent wake has formed , ure 13.11 ) . Such a jet has been observed in the
further increase in Reynolds number causes little cavitation -number regime from 0.15 to 0.30
change in the wake until the onset of cavitation . (Reference 10) and has been known to exist for

Commonly called the base, a term that seems contradictory when referred to a vertical cavity .
226
 0 8.3

16.7 33.3

Approximate time, ms

Figure 13.5a. Vortex shedding from fully wetted sphere. Conditions : v ( crossing oil-water interface) = 96 =

in/sec; Pa = 737 torrs ; F = 4.9 ; R x 10-4 = 0.55 for oil and 7.7 for water ; o = 39 for oil and 32 for water.
= = =

227
 50.0 58.4

66.7 83.3

Approximate time, ms

Figure 13.5b . Vortex shedding from fully wetted sphere. Conditions : v (crossing oil-water interface) = 96
= =
in/sec ; Pa = 737 torrs; F = 4.9 ; R x 10-4 = 0.55 for oil and 7.7 for water ; 0 = 39 for oil and 32 for water.
228
 10.8 13.3

17.5 20.0

Approximate time , ms

=
Figure 13.6 . Diffusion of vorticity into free stream . Conditions: v ( crossing oil-water interface) = 158 in /sec ; Pa = 737 torrs ;
=
F = 8.0 ; R x 10-4 = 0.9 for oil and 12.6 for water ; o = 15 for oil and 12 for water.

229
 6.7 7.5

8.3 10.0

Approximate time , ms

Figure 13.7 . Change in interface with time. Conditions : v( crossing oil-water interface ) = 158 in/sec ; Pa = 737 torrs;
F = 8.0 ; R x 10-4 = 0.9 for oil and 12.6 for water ; o = 15 for oil and 12 for water.

230
 16.7 58.4

83.3 104.2

Approximate time , ms

Figure 13.8 . Fully wetted plate cylinder crossing oil -water interface. Conditions : v (crossing oil -water interface )
= = =
= 30 in /sec; Pa = 737 torrs; F = 1.8 ; R x 10-4 = 0.13 for oil and 1.8 for water ; o = 406 for oil and 326 for water.

231
 CAVITATION
DYE

V.

1 in .

Figure 13.9 . Shear layer and cavitation delineated by dye .

some time , but it is an experimentally elusive missile and cavity . In Figure 13.13 ( the cavitat
thing that has been little photographed. Either ing sphere crossing the interface ) , the jet had
the downstream end of the cavity is opaque or , formed while the sphere was still in the oil. At
if translucent, the contrast between the jet and about the time the base of the cavity crossed the
the main flow is too vague to resolve photo interface , the jet had begun to impinge upon the
graphically. At still lower cavitation numbers, sphere. A short time later, the impinging jet
the cavity has been observed to terminate in a splashed from the sphere and distorted the
pair of hollow vortex tubes trailing from the cavity wall. The distortion continued to grow
base of the cavity (Figure 13.12 , from radially after the sphere and impinging jet had
Reference 82 ). passed, but below the point of distortion the
During the NUC studies of a sphere and a cavity soon assumed normal diameter, indicating
plate cylinder crossing an oil-water interface, the that the jet was no longer splashing against the
difference in vapor pressures of the oil and water cavity wall. At approximately 50 ms after the
caused the cavitation number in the oil to be missile reached the interface, a deep closure
higher than that in the water; at any given began just beneath the portion disturbed by the
velocity and submergence , the cavity formed in jet. After deep closure , the re-entrant jet
oil was shorter and slimmer than the water re-formed at the base of the cavity .
cavity. The size difference became apparent In a second sphere launching at higher
when the base of the cavity , which had been velocity under conditions of cavitating flow
traveling downward through the oil , remained (Figure 13.14) , a re-entrant jet again formed
stationary in the interface while the cavity while the sphere was still in the oil. Later the jet
lengthened. After the cavity had elongated deviated from the vertical and struck the cavity
sufficiently to achieve equilibrium at the wall. A deep closure occurred at the contact
ambient cavitation number, the cavity base again point between the jet and cavity wall , the upper
traveled downward . Cavity elongation illustrates portion of the cavity collapsing about the jet
the response of the cavity to sudden changes in and rebounding.
cavitation number . Figures 13.15 and 13.16 show a similar
Photographs of sphere launchings made launching of the plate-cylinder missile. The
under conditions of cavitating flow show re sharp discontinuity at the leading edge produced
entrant jets , revealing not only their formation a cavity so clear that the jet could be photo
but the interaction between the jet and the graphed as it overtook and then engulfed the
232
 - o = 0.71

4112-20

o = 0.54

NO ANG

o = 0.35

SIZNO236

o = 0.24

CIT-NOX -A36

o = 0.22

CIT.NO2 -A370

o = 0.21

C17-N02-1974

o = 0.21

CO -ND2 A37

Figure 13.10 . Increase in cavitation with decrease in o . Torpedo model with

hemisphere head.

233
 preceding cavity collapse, a vortex ring may
form at the base of the cavity .
The delineation of the flow when a missile
crosses an interface from colored oil into water
(the oil having approximately the same kinematic

B
Figure 13.11 . Re- entrantjet formation in cavitating
flow behind a plate .
viscosity as that of air) provides information on
wake behavior in missile air-to-water entry. The
results show that wake formation and behavior
are very complex. In fully wetted flow at low
velocities, vorticity appears in the wake with sub
sequent vortex shedding and decay . At higher
velocities , more prominent vortices appear ; a
large vortex configuration forms in the wake
and is subsequently shed, the momentum
imparted by the shedding vortex causing missile
missile. The jet created a small disturbance near trajectory deviation. In cavitating flow , the flow
the interface before it struck the missile . In the in the streamlines in and near the cavity wall re
motion pictures that were taken , the jet flow verse and turn inward to form a re-entrant jet
appeared to be pulsing. The cavity produced by that passes through the cavity and opposes the
this shape was so broad that the jet had almost direction of the main flow . This jet either im
reached the leading edge of the missile before pinges on the rear of the missile and is deflected
the cavity became distorted. into the side of the cavity or it strikes the side of
A second disturbance in the cavity wall the cavity directly. Later, deep closure may
with its characteristic bulge appeared soon after occur, with final cavity collapse and rebound.
the first, again suggesting pulsing or some other The features of this study that may affect
irregularity in the jet flow . In Figure 13.16 the missile design are ( 1 ) vortex shedding, which
collapse and rebound of the cavity after the may perturb the missile trajectory and perma
missile had passed from the camera field can be nently deflect it, and ( 2 ) the re-entrant jet ,
seen. Study of the motion pictures of the which can not only cause missile deflection but
cavitating flow leads to the belief that in the structural damage and malfunctioning in missiles
higher cavitation-number range immediately which exhaust gases from the rear.
1

Figure 13.12 . Vortices trailing from cavity .

234
 0 8.3

10.0 11.7

Approximate time, ms

Figure 13.13a. Cavitating sphere crossing oil-water interface . Conditions : v (crossing oil-water interface) = 159 =

in /sec; Pa = 25 torrs; F = 8.1 ; R x 10-4 = 0.91 for oil and 12.7 for water ; o = 0.7 for oil and 0.1 for water.
=

235
 9

13.3 16.7

20.8 25.0

Approximate time , ms

=
Figure 13.13b . Cavitating sphere crossing oil-water interface . Conditions: v (crossing oil-water interface ) = 159
in/sec; Pa = 25 torrs; F = 8.1 ; R x 10-4 = 0.91 for oil and 12.7 for water; 0 = 0.7 for oil and 0.1 for water.
236
 0 8.3

33.3 37.5

Approximate time , ms

=
Figure 13.14a. Cavitating sphere crossing oil-water interface. Conditions : v ( crossing oil-water interface) = 244
=
in/sec ; pa = 22 torrs; F = 12.4 ; R x 10-4 = 1.4 for oil and 19.5 for water ; 0 = 0.3 for oil and 0.02 for water.
- =

237
 50.0 66.7

75.0 100.0

Approximate time, ms

Figure 13.14b . Cavitating sphere crossing oil-water interface. Conditions: v (crossing oil -water interface ) = 244
in /sec; Pa =
22 torrs ; F = 12.4 ; R x 10-4 = 1.4 for oil and 19.5 for water ; o = 0.3 for oil and 0.02 for water .
=

238
 0 4.2

8.3 10.0

Approximate time, ms

Figure 13.15a. Cavitating plate -cylinder missile crossing oil-water interface. Conditions: v ( crossing oil -water
= = =
interface) = 150 in/sec; Pa = 25 torrs; F = 8.8 ; R x 10-4 = 0.64 for oil and 9.0 for water ; 0 = 0.8 for oil and
0.1 for water.

239
 13.3 16.7

1
1

20.8 22.5

Approximate time , ms

Figure 13.15b . Cavitating plate-cylinder missile crossing oil-water interface. Conditions : v ( crossing oil -water
=
interface) = 150 in/sec; Pa = 25 torrs; F = 8.8 ; R x 10-4 = 0.64 for oil and 9.0 for water ; 0 = 0.8 for oil and
0.1 for water.

240
 25.0 29.2

Approximate time, ms

Figure 13.15c. Cavitating plate-cylinder missile crossing oil-water interface. Conditions: v (crossing oil-water
interface) = 150 in/sec ; Pa = 25 torrs ; F = 8.8 ; R x 10-4 = 0.64 for oil and 9.0 for water; 0o = 0.8 for oil and
= = =

0.1 for water.

241
 33.3 50.0

66.7 75.0

Approximate time , ms

Figure 13.16a . Collapse and rebound of cavity in latter part of launching shown in Figure 13.15 .

242
O

O
91.7 100.0
IC

116.7 133.3

Approximate time , ms

Figure 13.16b . Collapse and rebound of cavity in latter part of launching shown in Figure 13.15 .

243
CHAPTER 14
WAKE EFFECTS

The fine structure or rapid fluctuation of To determine whether vortex shedding

the flow around a missile tail is important in could have produced such a pitching motion,
wake studies because of the vibration it induces . tests were conducted in the large NUC Open
In a large vehicle, the result may be metal Launching Tank (Part V) . The experimental
fatigue and personnel discomfort. In a small procedure was as follows:
missile, the tail may be so perturbed that missile
orientation and velocity are somewhat affected . 1. The missile was trimmed to be nose
The large wake fluctuations reported in the heavy and its specific gravity was
sphere studies of Chapter 13 showed that aa shed
adjusted so that a sinking velocity
vortex resulted in trajectory deflection . range of 1 to 6 fps could be studied
Studies of the fully wetted flight of a when the missile was released verti
research missile ( Basic Finner, Reference 80 and cally at the water surface.
Figure 14.1 ) , revealed an additional small pitch
ing motion superimposed on the major motion 2. A thin layer of nigrosine (black
with a frequency of about 15 Hz at a missile water-soluble dye ) paste was applied
velocity of 7.5 fps. This frequency approximates either on the missile base near the
that predicted by theoryl for Karman vortex intersection of fins and body cylinder
street shedding ( 18 per second) . Since a
or on the after edges of the fins.
shedding vortex exerts a force normal to the
3. The missile was held vertically nose
missile motion, it was suspected that vortices
were shedding from the base of the missile at a down , partially submerged, with the
comparable frequency, and the observed small dyed area just above the water sur
pitching motion was caused by interaction face , and released . It was held at least
between the missile fins and the shedding 15 inches away from the tank sides to
vortices . avoid wall effects .

From
flow is
theory (Reference 24, page 570) , the frequency of vortex shedding behind a cylinder with axis normal to the
Nd 19.7
= 0.198 ( 1
U. 08 (1 - 10:)
R

where N is the number of vortices shed per unit time , U. is the velocity of the undisturbed fluid with respect to the
cylinder , d is the cylinder diameter , and R is the Reynolds number with respect to the cylinder diameter. For
two- dimensional plates of width d normal to the flow
Nd
= 0.16
U.

where 0.16 is an average of values found by several investigators . The Reynolds number based on a velocity of
7.5 fps, a missile length of 10 inches, and a water temperature of 20°C' is 5.79 x 105 , which corresponds to
transition on the missile body ; thus the flow about the fins was probably turbulent.
244
 1.000

0.08
3.00
To0.015
20°
1.000

10.00

Figure 14.1 . Basic Finner missile . Dimensions in inches.

4. The sinking missile and the dye trail missile base to suppress vortex formation by
were photographed with the rotating reducing or eliminating the surface discontinuity
disk camera (Part V). at the base. A small cone was installed first, but
it was completely enveloped in the disturbance
5. This procedure was repeated several at the missile base ; vortices still formed and
times to assure typical performance. there were corresponding velocity fluctuations.
A second afterbody consisting of an approxi
experimental results are illustrated in mate 3.5- to- 1 streamlined form was substituted .
Figures 14.2 to 14.6. Figure 14.2 shows dye With this fairing, which eliminated surface dis
trace photographs of vortices trailing from the continuity at the base, no large vortices were
missile base at velocities of 1.0 to 1.2 fps. shed and no velocity fluctuation was noted. The
Theory predicts that at 1.1 fps the vortex dye-trace data indicate that the minor pitch fluc
shedding frequency should be between two and tuations and velocity fluctuations were caused
three per second ; data measurements showed by vortex shedding at the missile base.
this frequency to be about three. The missile Figure 14.5 shows small high -frequency
acceleration over the velocity range studied was vortex shedding from the missile fins at approxi
about 0.2 ft/sec2 , which is small. It is doubtful mately 25 per second and an average velocity of
that acceleration significantly affected the 1.1 fps. This is in excellent agreement with 26
results . per second predicted by theory (footnote 1 of
A velocity fluctuation at the frequency of this chapter, second equation ) where a fin base
vortex shedding ( Figure 14.3 ) was observed thickness of 0.08 inch is used in the calculations.
from the data of Figure 14.2 . It could have been Enlarged photographs of fin vortices from the
caused by the fluctuating drag that resulted missile with and without the streamlined after
from vortex shedding (Reference 23 , pp . body are shown in Figure 14.6. The predicted
59-65 and Reference 24 , pp . 556-562 ) . Fig vortex shedding frequency at 1.0 fps missile ve
ure 14.4 illustrates tests made to determine locity was 24 per second ; the actual vortex
whether the velocity fluctuation was caused by shedding frequency was 25 per second. The pre
shedding vortices. À fairing was added at the dicted vortex shedding frequency at 2.8 fps was
245
Figure 14.2. Vortices shedding from base of missile (0.04 sec between photographs).

246
 1.2
fVelocity
' .

O
, ps

co
1.0

0.8 1 1
0.00 0.16 0.32 0.48 0.64 0.80 0.96

Time, sec

Figure 14.3 . Missile velocity versus time during vortex study of Figure 14.2.

67 per second ; the actual vortex shedding fre travel of a steel sphere in fully wetted flow
quency at 2.7 fps and 2.9 fps was 65 per second. after shedding a vortex. It is evident that in
The vortex frequencies are approximately pro designing a missile or vehicle , care should be
portional to the linear velocity and agree with taken to avoid or to suppress the generation of
the values predicted by theory. vortices that could give rise to undesirable
The effect of vortex shedding on missile be effects. In this connection , the elaboration of
havior is illustrated in Chapter 13 ( Figure 13.6) , theory for vortex generation and shedding in the
which shows the deviation in the direction of three -dimensional case would be helpful.

247
 ㅓ

Missile alone , Missile with cone , Missile with streamlined

v = 1.0 fps v = 1.2 fps afterbody, v = 2.7 fps

Figure 14.4. Disturbances in flow behind missile .

248
 v = 1.2 fps

v = 1.0 fps

Figure 14.5 . Vortices shedding from missile fins (0.04 sec between photographs).

249
Missile alone , Missile with streamlined Missile with streamlined
v = 1.0 fps afterbody, v = 2.7 fps afterbody , v = 2.9 fps

Figure 14.6. Enlargements of vortices shedding from fins.

250
 CHAPTER 15
WATER-ENTRY CAVITIES OF SPINNING SPHERES

When a body moves through a real fluid ,

the generated boundary layer gives rise to skin
friction drag. Under certain conditions , the flow
in the immediate neighborhood of a wall
becomes reversed, causing the boundary layer to
separate from it and eddies to be formed in the
wake of the body. Thus the pressure distribution
is changed and differs markedly from that in a
frictionless stream . This deviation is the cause of
form drag and is detrimental in application in
that it increases the body drag. In addition ,
boundary -layer growth as it affects pressure
distribution must also affect fluid separation and
hence cavity flow . Cavity flow , in turn , can give
rise to control problems in the case of a missile
in a completely enveloping water-entry cavity.
When a sphere is dropped vertically into
water with a sufficiently high velocity , a cavity
of air follows the body for some distance below
the surface ( Reference 112 ) . The cavity wall
appears to be tangent to the sphere at the point
0°
of separation, and is transparent. If the sphere is €
spun about an axis parallel to its direction of
motion, the character of the cavity is basically
unchanged for small spin velocities ; the principal
effect is merely to wrap any irregular striations
around the cavity with a pitch corresponding to
the forward speed and spin of the sphere
( Figure 15.1 taken from Reference 51 ). V
From Figure 15.1 it is apparent that
Figure 15.1 . Cavity surrounding a 2-inch sphere spinning at
=
24 rps, Ve = 24.5 fps.
V
tan E = ( 15.1 )
wr' =
€ = angle that striations make with a plane
normal to the direction of sphere
where travel

w = rate of spin of sphere in radians per r ' = radius of cavity bubble in region where
second about an axis parallel to its e is measured, ft
direction of motion v = velocity of sphere, fps
251
If the spin velocity is increased sufficiently, a from 10.2 and 5.3 feet above the water level ,
marked change in the character of the cavity corresponding to entry velocities of approxi
wall and the separation angle takes place. The mately 25 and 17 fps. A magnetic release was
wall becomes frothy and opaque and the direc used ; the steel ball was held to a conical seat in
tion of separation is no longer tangent to the the end of a steel rod by an electromagnet and
sphere at the separation point. For intermediate released by interrupting the current in the sole
velocities, complex spiral cavities occur noid. The bowling ball was suspended by a non
(Figure 15.2 ). twisting string and dropped by cutting the string.
Since the body drag in cavity motion is As the falling ball approached the water surface,
loosely correlated with the maximum diameter it interrupted a photoelectric light beam that
of the cavity, it should be expected that a tripped the timing circuit for flash -lamp
marked difference in drag will occur between photography.
low and high spin velocities. Studies made in the Before it was dropped , the steel ball was
NUC Hydroballistics Laboratory ( Reference spun on the rod by a variable-speed motor
107 ) dealt with the effect of spin on the type of mounted with its shaft in a vertical position. The
separation and on the drag. motor was brought up to the desired speed, the
The experiments were performed in the speed was checked stroboscopically, and the ball
Open Launching Tank (Part V) . Magnetic was released. A small dot painted on the ball
stainless-steel ball bearings of 1., 1.5- , 2- , 2.5-, made it possible to determine photographically
3-, and 4.5-inch diameters were used . A few the exact rotative speed. The lights were flashed
drops were also made with an 8.5 -inch-diameter and photographs taken at 0.005 -second inter
plastic bowling ball. The balls were dropped vals. Figure 15.3 is a typical photographic record .

zero spin 40 -rps spin 60 -rps spin 80-rps spin

Figure 15.2 . Types of cavity motion , 3-inch-diameter sphere, Ve = 25 fps.

252
 Figure 15.3 . Disk camera photography used for measurements.

Figure 15.2 shows the types of separation If the shift in character of the cavity and
that occurred as the spin velocity was increased . the separation point for water entry of a
A 3-inch-diameter sphere with zero spin and spinning sphere is due to a transition from
25 -fps entry velocity shows separation of the laminar to turbulent boundary layer, it should
cavity well forward of the equator, with trans be possible to correlate this transition with a
parent cavity walls. With a spin velocity of Reynolds number that takes into account the
40 rps, striations due to slightly irregular separa size of the sphere, entry velocity , and spin
tion are wrapped around the cavity, but the velocity. Luthander and Rydberg ( Reference 41 )
general character is unchanged. A further defined a “ Reynolds number ” velocity for any
increase in spin velocity to 60 fps shows two point on the sphere as the relative velocity
distinct regions of separation, with the section between that point and a fluid particle lying
of the larger cavity from the forward region just outside the boundary layer and hence
wrapped around the smaller cavity in the after moving with potential flow uninfluenced by
region. When the spin rate is increased to 85 rps , spin. The characteristic length is taken as
the separation takes place only aft of the the length of the path of the projection on
equator, and the cavity wall is completely the sphere surface of a point moving in the
opaque.. This shift in separation position appears water just outside the boundary layer from
closely analogous to laminar-turbulent transition the upstream pole along a meridian. In the
above the critical Reynolds number. Similar case of zero spin, this reduces to the dis
effects have been found with rotating spheres in tance along a meridian to the point at which
wind tunnels. R is being evaluated.
253
 separation
If the change in character of the separation 150 TTTT TU
is due to transition from a laminar to a turbulent

Angle
Turbulent

,deg
of
120
boundary layer, this transition must take place
ahead of the point of separation of the type
shown in the first panel of Figure 15.2 , which 90

was observed to occur between colatitude angles Laminar

60
of 70 and 90 degrees, measured from the
upstream pole of the sphere . By choosing a

30
colatitude angle of 78 degrees, it can be shown
that
0 null
1x 105 2 4 6 8 1x1062 4 6 8 1x107 2
4r r2w2 + ( 1.5ve)?
+
Reynolds number
RE ( 15.2 )
3 1.5ve "
Figure 15.4 . Separation angle as a function
of Reynolds number.
where
R = “ Reynolds number ” for spinning Reynolds numbers were obtained with 1.5- and
spheres 2-inch spheres at 17 fps and zero spin. No
r sphere radius, ft consistent trends were found to be associated
with ball size , height of drop , or spin velocity.
Ve = velocity of sphere at water surface just In corkscrew drops , two clearly defined
before entry , fps
types of separation (laminar and turbulent) are
= kinematic viscosity of water, ft2 sec- 1
V =
unmistakable. They are shown in Figure 15.5,
For nonspinning spheres , Equation 15.2 reduces along with some drops in which boundary -layer
to the usual definition of Reynolds number for a turbulence was induced by roughening the nose
sphere. (the upstream surface) of the sphere. The solid
lines of Figure 15.4, showing separation angle as
2rve a function of Reynolds number for non
R =
( 15.3 ) corkscrew cavities , are repeated in Figure 15.5
V for comparison .
The pairs of points designated by “ O ” and
Values obtained from this arbitrarily joined by dashed lines represent the upper and
defined Reynolds number for spinning spheres lower limits of separation for corkscrew cavities.
cannot be compared with those for non-spinning The upper values form a group distinct from the
spheres , or, in general , with each other; equal lower values, although both sets are below the
values of these Reynolds numbers do not neces solid line representing single separation at the
sarily indicate geometrical similarity of flow and same Reynolds number. Such a difference is not
therefore mechanical similarity on which the surprising, however, since the shape of the cavity
Reynolds number is based. This anomaly is is certain to affect the pressure distribution in
apparent in the extraordinarily high transition the surrounding water.
Reynolds number ( 1 to 2 x 106 ) obtained for Since all the turbulent separations had
spinning spheres. Transition for spheres usually occurred with spin, these data were insufficient
occurs at a Reynolds number of 2 x 105 . to separate centrifugal effects from boundary
Figure 15.4 shows separation points layer transition effects. To this end, a 25 -fps
plotted against the Reynolds number of Equa zero spin drop was made with the 3 -inch
tion 15.2 for 52 drops in which separation was diameter sphere with a 1 -inch-diameter patch on
well defined , ignoring such corkscrew cases as the nose artificially roughened by cementing
the one in Figure 15.2. Most of these data were sand to it. An irregular separation occurred, but
taken from drops with 2-, 2.5-, and 3-inch the upper and lower limits were well defined and
diameter spheres with entry velocities of 17 and are plotted as points “ e ” in Figure 15.5 . To
25 fps and spin velocities of approximately 0, reach higher Reynolds numbers without spin,
40, 80, 120, and 200 rps. The two lowest the 4.5 -inch - diameter steel sphere was also
254
 180

o Dual separation for corkscrew-type cavities

o 4.5-in. diameter steel ball
x 4.5-in. diameter steel ball with sand on the nose
150
A 3-in. diameter steel ball with sand on the nose
V 8.5- in. diameter bowling ball
separation

O
Angle

120
,deg
of

е Д a *b
90

--O
a

Dog
60
XXXd
CX
30

1x 105 2 3 4 5 6 7 8 9 1x 106 2 3 4 5 6 7 8 9 1x107 2

Reynolds number

Figure 15.5 . Separation angle as a function of Reynolds number for

corkscrew cavities and induced turbulence.

sanded . The pair of points “ a” represent the the same basic phenomenon as that obtained
upper and lower limits of separation for a patch with spin. The quantitative difference between
of sand about 2.3 inches in diameter. When still the turbulent separation point for a nonspinning
more sand was added , the lower separation sphere and for a spinning sphere at the same
occurred on the sand and the upper separation Reynolds number is probably due to the manner
limit agreed well with case a. Points b and c were in which R is defined . In order to improve this
obtained for a patch with a chord diameter of definition , the actual velocity distribution just
about 3.4 inches. Point d resulted when the outside the boundary layer should be ascertained .
diameter of the patch was increased to 3.7 In the water entry of spheres, the transition
inches and the separation took place entirely from laminar to turbulent boundary layer, with
from the sand. All the 4.5 -inch-diameter drops a consequent reduction in the size of the wake,
were made at 25 fps. is accompanied by a marked reduction in drag,
Still higher Reynolds numbers were as it is in air. The drag coefficient is plotted in
obtained with no spin by using an 8.5 -inch Figure 15.7 against Reynolds number for the
diameter bowling ball. Two 25 -fps drops were same set of drops that were used to determine
made, with results plotted as point f. The the separation point. A marked lowering of drag
character of the separation is shown at the left is apparent with the shift from laminar to
of Figure 15.6 . A sand patch about 4 inches in turbulent separation .
diameter was cemented to the ball ; the separa At the low speeds involved in these studies,
tion point moved sharply aft, as shown at the a turbulent boundary layer can inhibit separa
right of Figure 15.6 and plotted as points g for tion and hence reduce the cavity size and the
two drops in Figure 15.5 . The position of the drag, much as in ordinary flow . The transition
separation, the shape of the cavity, and the from laminar to turbulent flow is probably the
character of the cavity all indicate that this is most important factor in determining the nature
255
 !

Sand
=
Figure 15.6 . Effect of sand patch cemented to 8.5 - inch -diameter bowling ball, ve = 25 fps .

of the separation. The Reynolds number as viscosity. Further consideration of the actual
defined in these studies for spinning spheres is velocity distribution about spinning spheres is in
not an adequate criterion for mechanical order .
similarity under the action of inertia and

0.40
coefficient

TTTTI
Laminar
Drag
,CD

0.30

Turbulent
0.20

0.10
I

1
0
0 2 4 6 8 1X106 2 4 6 8 1x107 2

Reynolds number

Figure 15.7 . Effect of laminar and turbulent

separation on drag coefficient.

::

256
 CHAPTER 16
EFFECTS OF GRAVITY ON THE ENTRY CAVITY

Water-entry cavity behavior is a major fac

a A better understanding of the forces affect
tor in determining the underwater trajectory of ing surface closure and thereby control of cavity
a missile (Chapter 3 ) and , in turn, cavity closure formation could be advantageous in developing
affects cavity behavior because it terminates ex scaling techniques for hydroballistics modeling
ternal gas flow into the cavity , thereby limiting at low entry velocities. It is possible that the
its size. If the forces acting to close the cavity altered surface closure pattern resulting from the
are sufficiently large, the cavity is repressed. elimination of hydrostatic pressure may aid in
Cavity closure occurs either at the water the interpretation of these forces.
surface (surface closure) or below (deep clo Deep closure can be explained satisfactorily
sure ) , the nature of the closure being influenced in terms of forces generated by missile inertia
by both ambient and launching conditions and tending to open the cavity and hydrostatic
by the configuration and velocity of the missile forces tending to close it. It would be expected ,
(Chapter 3 ) . then , that deep closure would be inhibited in a
Surface closure appears to be the more zero -gravity system where the hydrostatic pres
complex and is the major factor in controlling sure tending to close the cavity does not exist.
cavity formation (Reference 22 ) . Interplay The hydrostatic pressure produced by
among the many factors influencing surface clo gravitational force can be nullified in the labora
sure has obscured its mechanism although it has tory by launching a missile in a free-fall system.
been studied at length (Reference 8 , 22 , 46 and The gravitational force acts equally upon all
112 ) . Modeling studies (Chapter 9 ) in which the elements of the system instead of only upon the
minimum entry velocities were 60 fps indicate missile and the disturbed liquid surface, reducing
that one-to-one Froude-number, cavitation the net effect of gravity to zero . Free-fall
number , and gas-density scaling are adequate for conditions can be achieved by releasing the
water-entry cavity modeling but this scaling does bottom of a tank filled with water immediately
not always model surface closure at lower entry prior to missile water entry. The NUC Vertical
velocities (Reference 45 ) , indicating that some Drop Tanks (Part V) were used for a series of
other force or forces may also be significant. such tests. Spheres and plate-cylinder missiles
During the early stages of missile penetration at were dropped vertically at low velocities in the
low velocities, the most prominent force is prob range from 12 to 22 fps. Since, in the case of
ably fluid adhesion to the missile surface (Refer spheres, it is known (References 46 and 112 )
ence 112 ) . Later, after the splash has formed, that at these low velocities the surface condition
surface tension may assume a more dominant affects cavity formation , the tests were made
role in splash seal and surface closure . The in with precision -ground, highly polished stainless
creased kinematic viscosity of the atmospheric steel spheres that had been given a thin durable
gas brought about by reduced pressure scaling hydrophobic coating of silicone to minimize
may also delay fluid separation from the missile water adhesion forces.1 The plate-cylinder mis
(Reference 8 ) . siles consisted of cylindrical steel rods of

1The spheres were cleaned and placed in polymethylsiloxane , DC-500, 70 centistokes, for two days at 150 to 160°C .
This produced a yellowish coating, the thickness of which is less than the wavelength of light.
257
different diameters that were undercut near the conditions 0.036 second after entry . The slope
leading end to provide a clean line of cavity of the splash as it separated from the sphere was
separation. measured from the photographic data of Figure
Figure 16.1 shows the vertical water entry 16.1 over approximately one sphere diameter of
of a 1 /2- inch - diameter sphere at atmospheric water penetration (0.007 second ) . For this pur
pressure traveling at 21 fps with respect to the pose, the first 20 frames after water contact,
water surface. The first sequence shows entry taken at 2900/second, were measured (the sphere
occurring in the normal gravitational field and is moved about one diameter in 17 frames) and the
typical of that observed many times. The splash angles of separation on either side of the sphere
is of the normal “ basket ” type, and the cavity is were determined with respect to coordinates
glassily smooth with striated walls. The cavity fixed in the water. The results are shown in Fig.
necks off and closes beneath the water surface in ure 16.5 . The difference in slope between the
a typical manner. In the second sequence, two splashes was more than 10 degrees.
gravity is nullified by free fall of the water and a Interplay between gravitational and surface
sheath -type splash occurs. The cavity has glassy , tension forces may initially determine the splash.
striated walls, but because of the absence of A sheath of water rises around a sphere at water
hydrostatic pressure , and therefore buoyancy entry . As noted by Worthington (Reference
forces, it follows the sphere down and there is 112 ) , whether the rising sheath leaves the sphere
no deep closure. or follows it depends upon the efficiency of fluid
The most striking differences in water-entry adhesion . For spheres with a hydrophobic
behavior associated with removal of gravitational coating (as used in these studies), adhesion forces
effects were noted in the cavity closures. Figure are small, but the surface tension of the sheath
16.2 shows water- entry cavities of aa 1 -inch will act at all times in favor of adhesion to the
diameter sphere at ambient conditions during the sphere (Figure 16.6 ) , and the resulting small
first 0.055 second after water entry . Within the capillary pressure may have some effect. Below
time span recorded, surface closure was almost the sphere equator, the gravitational force will
complete, and incipient necking indicated that tend to bring about separation of the sheath from
deep closure was in progress in the gravitational the sphere. Hence, removal of the gravitational
field . Without gravitational effects, the surface forces should delay separation, resulting in a
had completely closed and the cavity pulled more vertical splash slope and a sheath-type
away , but there was no tendency toward deep splash . Such a splash , having a smaller throat and 1

closure. greater curvature, should be more susceptible to

Figure 16.3 shows water-entry cavities of a surface tension and gas dynamic underpressure
1 -inch -diameter sphere at 16.1 fps entry velocity . forces, resulting in earlier cavity surface closure.
Freon 12 (dichlorodifluoromethane) at 2 atmos These observations are substantiated by the
pheres pressure was used over the water surface, photographic data.
thereby providing a gas density eight times Three vertical launchings of a 0.168-inch
greater than that which would obtain with am diameter plate-cylinder rod made with an entry
bient air. Thus earlier surface closure should be velocity of 14.9 fps show the effects of the
expected in both fields (Chapter 3 ) . In the grav gravitational field upon the cavity ( Figure 16.7 ).
itational field , a dome-like splash occurred and The first sequence shows a launching in the
surface closure took place with pullaway from normal gravitational field ; in the other two
the surface , but buoyancy prevented the cavity sequences, the field is gravitationless. The
from following the sphere downward. Without ambient pressures were close to the vapor
gravitational effects , cavity closure occurred ear pressure of water, ranging from 24 to 35 torrs.
lier, with a smaller cavity that did follow the The differences in deep closure are readily
sphere. As predicted , both cavities were marked apparent. In the normal gravitational field , the
ly smaller than the comparable cavities formed cavity closes about the rod ; without gravity it
at ambient conditions . remains open for the duration of the photo
In the gravitationless field the splash was graphic record. However, both zero -gravity cavi
more cusp-like in shape, as illustrated in Figure ties evidence necking in two places, indicating
16.4, which shows water-entry cavities of a 1 that two later deep closures will occur in spite of
inch -diameter sphere under ambient atmospheric the absence of hydrostatic pressure. Cavity
258
 IN

Normal gravitational field

Gravitationless field

Figure 16.1 . Vertical entry of sphere, Ve = 21 fps. Camera speed 290 frames/sec.
=

259
 Normal gravitational field , Gravitationless field , Ve = 16.1 fps
Ve = 14.4 fps

Figure 16.2 . Effect of gravity on cavity closure .

C)

UVOD
Normal gravitational field

o
100
Gravitationless field

Figure 16.3 . Effect of gravity on surface closure, ve = 16.1 fps. Time between photographs 0.004 second.

260
Normal gravitational field, Gravitationless field ,
Ve = 14.4 fps Ve = 16.1 fps

Figure 16.4. Effect of gravity on splash .

She
separation

25
splash
Slope
,8deg

• Normal gravitational field

of
at

X Gravitationless field
20

: :
-

15
XX

10
U

x
Х

5
x Š
xx
w

0 Х

8 10 12 14 16 18 20

Frames after water contact

Figure 16.5 . Effect of gravity on slope of sphere splash ;

ve = 21 fps.

261
 Surface
tension
force

Oc Gravitational
force

Figure 16.6. Action of surface tension and gravitational

forces on splash .

Normal gravitational field

Gravitationless field

Gravitationless field

Figure 16.7 . Effect of gravitational field on cavity, ve = 14.9 fps. Time between photographs 0.008 second .

262
 necking indicates the action of surface-tension
forces2 . An indication of deep necking was ob
served with some of the spheres tested , but their
diameters were greater than the plate diameters ;
therefore their cavity diameters would be larger
and surface tension effects less significant.
Enlarged frames of Figure 16.7 are pre
sented in Figure 16.9 to show details of the
splash. The zero -gravity splash formed by the
rods is less cusplike than that formed by the
spheres (Figures 16.1 and 16.4) , but is definitely
smaller and less domelike than the splash of the
rod in the normal gravitational field.
Launchings of a 0.252-inch-diameter plate
cylinder rod were made to study the effects
upon the zero -gravity cavity of change in atmos
pheric pressure. The tank pressures were 35 and
90 torrs ; the entry velocity was 14.9 fps. The
cavities increased in size with reduction in
atmospheric pressure (Figure 16.10) . The results
Figure 16.8 . Cavity necking in illustrate the sensitivity of cavity closure to
gravitationless field . small changes in gas dynamic pressure reduction.
Removal of gravitational force effects pro
vides a useful technique for studying water-entry
necking in the gravitationless field is also illus cavity phenomena. The complex interaction of
trated in Figure 16.8 (0.070-inch-diameter rod the forces affecting cavity behavior are not yet
entering water at 14.9 fps and 40-torr air fully understood, especially at lower water -entry
pressure ). velocities where the surface adhesion and ten
Consideration of the forces acting upon the sion forces that affect surface closure (and
cavities shown in Figures 16.7 and 16.8 suggests therefore cavity formation) become more signifi
that the observed necking is a surface tension cant. It is believed that further water-entry
effect. The pressure force due to surface tension studies that show splash and closure phenomena
is of the order 10-2 psi immediately behind the in fine detail will aid in interpreting the forces
plate ; the gas dynamic pressure reduction force affecting the water-entry cavity and may be
is considerably less, being of the order 10-4 psi. useful in extending hydroballistics modeling
In addition, the suggestion of periodicity in the techniques to lower water-entry velocities.

2
A fluid column is known to become unstable, neck, and break into droplets when surface -tension forces in the
column boundary exceed internal column pressure (Reference 112) . Here an air column is involved, but
surface - tension forces would tend to constrict and break the column into separate bubbles.

263
 Normal gravitational field , air pressure 30 torrs

Gravitationless field , air pressure 35 torrs

Gravitationless field , air pressure 24 torrs

Figure 16.9 . Details of splashes shown in Figure 16.7 .

264
 Air pressure 35 torrs

Air pressure 90 torrs

Figure 16.10 . Effect of atmospheric pressure on water entry cavity in gravitationless field, ve = 14.9 fps.

265
 PART V.
SPECIAL EXPERIMENTAL TECHNIQUES AND
HYDROBALLISTICS FACILITIES

CHAPTER 17
FACILITIES AND EQUIPMENT
Most of the data originating from this ratio, etc. ) in water-entry and underwater
study of hydroballistics modeling were obtained trajectory modeling, and to develop , if feasible ,
at the Naval Undersea Research and Develop scaling techniques that provide similitude within
ment Center (NUC ) 1 either in the Hydrobal the accuracies established as acceptable.
listics (Model) Laboratory in Pasadena, The steel tank is 3 feet wide by 3 feet deep
California, or at the prototype launching site at by 8 feet long, with windows of 1-1 /4-inch-thick
the Morris Dam Torpedo Range. The rest of the tempered clear glass on three sides, a dome 2
model data resulted from tests in the Hydro feet high on the top, and a well 1 foot in
dynamics Laboratory at the California Institute diameter and 1 foot deep below. The windows
of Technology (CIT) . Those test facilities and are 24 inches wide by 69 inches long. The dome ,
equipment that are individually unique are well, and tank centerlines are coincident. A pit
described below. below the tank provides space for underwater
launching equipment .
VARIABLE -ANGLE VARIABLE-PRESSURE The tank, mounted on curved tracks with a
LAUNCHING TANK 46-inch radius so that it can be rotated from 5
The VAVP tank at NUC ( Figures 17.1 and to 90 degrees with respect to the horizontal, is
17.2 and Reference 89 ) is a facility designed for driven by a hand-operated worm gear through a
studies of the water-entry , water-exit , and double sprocket and a 150-inch-long chain sys
underwater-trajectory behavior of momentum tem attached to the tracks .
propelled missiles for varied trajectory angles, Missiles up to 2 inches in diameter and
missile velocities , and conditions of cavitation. It 1-1 /4 pounds in weight can be launched in the
was built primarily for research into the tank with maximum velocities of 120 fps at
importance of scaling parameters ( Froude, water entry and 80 fps at water exit. The air (or
Reynolds, and cavitation numbers, gas -density gas) pressure in the tank can be varied from the

1at the time, the U.S. Naval Ordnance Test Station (NOTS) .
267
 Access Port

OOC

Figure 17.1 . Variable-Angle Variable - Pressure Launching Tank, oblique water -entry position.
( From Reference 89. )

Access Port no

Figure 17.2. Variable-Angle Variable -Pressure Launching Tank,

vertical water-exit position . ( From Reference 89.)

268
vapor pressure of water to 1.5 atmospheres the free airflight before water entry was at least
absolute. For metric photography, electronically 2 feet; the free water flight before water exit
operated Edgerton -type stroboscopic lamps of 1 was at least 1 foot, the actual length of the flight
to 2 -us flash duration illuminate the missile path being determined by the water level
airflight, water -entry, water-exit, and under selected. The system was designed to launch the
water trajectory, as well as the associated cavity missile with no yaw or angular velocity and to
phenomena. Missile water- entry velocities , minimize air (or water) flow due to launcher
accurate to 0.1 fps, are immediately obtained action .
from a photo -electric timer record taken during For water-entry launchings, the missile was
airflight. In addition , an optical whip recorder loaded through an access port ( Figure 17.1 ) in
the side of the tank dome . The missile was
(described below) can be used to obtain highly
accurate data on missile orientation during inserted into the carriage with the impelling
airflight and water entry. piston forward and the launcher was cocked by
The tank water as supplied from high pushing the piston back. For water-exit launch
pressure city mains is supersaturated with air. ings, the missile was loaded through a port in the
The water was deaerated to a repeatable equilib center of the dome (Figure 17.2 ) . The missile ,
rium condition for all launchings (Chapter 18) held in a close-fitting cylindrical carriage on the
so that diffusion of air into the missile cavity end of a long wooden pole, was pushed into the
was minimized , and water -vapor or very nearly launcher by a piston in the pole carriage; the
water- vapor cavitation obtained. The air content entire launcher system was then pushed down
of the water was determined with a Van into launching position. A flared skirt on the
Slyke-Neill manometric gas-analysis apparatus pole carriage simplified alignment with the
(Reference 62 ). The same water was used launcher carriage.
throughout a single series of launchings . The After prelaunching operations had been
high degree of clarity necessary for metric completed the pressure and density of the tank
stroboscopic photography was maintained by atmosphere were adjusted with either a large
filtering the water through diatomaceous earth vacuum pump or compressed air or gas as
for about an hour each morning before tests required to satisfy the scaling condition con
were begun . straints, and the density of the tank atmosphere
Water temperatures as low as 10°C (for was determined gravimetrically .
reduced water -vapor pressure studies ) can be At the same time , the air pressure in the
obtained in about 2 hours by circulating the pneumatic missile launcher was adjusted to give
tank water through a refrigerator, and water the predetermined missile velocity, and the
temperatures as high as 80°C (for reduced water cameras were brought up to speed. With the
kinematic viscosity studies ) can be obtained by laboratory in complete darkness, the missile was
circulating the tank water through a 24-kw launched , and photographs taken with electroni
electric heater for about 2 hours . cally timed Edgerton-type stroboscopic lamps,
A single launcher projected missiles along or, when the optical whip recorder (described
the tank axis and served for both water-entry below) was used, by electronically timed sparks.
and water-exit studies. For water entry, it was Sideview pictures of the missile and of the
mounted on the dome of the tank , and for water underwater cavity could be obtained in the pitch
exit, on the well cover. The missile was held by plane ( the vertical plane that contains the
slight friction in aa ventilated sleeve-type carriage trajectory) for all trajectory angles, but only 70
mounted on the end of a piston extending into to 90 -degree trajectories were visible in the
the tank through a watertight and gas-tight seal. vertical plane normal to the pitch plane.
Upon actuation , the piston was impelled by a Missile flight was terminated by a heavy
pneumatic system outside of, and sealed off nylon fabric backstop that is impervious to
from , the interior of the tank . After the moisture and to the action of bacteria or molds.
acceleration stroke, the piston was stopped by a After launching, the missile was retrieved
hydraulic buffering system; the momentum through the dome access port with a remotely
propelled missile emerged from the carriage to operated clamp on the end of aa pole.
continue in free flight. With missiles 1 foot long,

269
Rotating-Disk Camera properly gating the number of flashes, the
The camera ( Figure 17.3 ) and Reference 94 complete disk can be covered with exposures.
was designed and built at NUC for sideview The range of missile positions over the exposure
photography. It has some advantages over a strip sequence can be set with aa time-delay generator,
camera when only a few exposures are required : adjustable to l us , that triggers the flash lamps
at after launcher
the larger frames permit closer detail and greater the appropriate moment
metric accuracy and there is no film waste . It is actuation .
a 5-by- 7-inch view camera (with back removed)
mounted on a housing containing a rotating disk Optical Whip Recorder
with a circularly cut 11-inch-diameter photo The optical whip recorder, or whip camera
graphic film clamped to its surface . An ( References 54 and 89 ) , is used in water-entry
adjustable triangularslot in the disk cover plate studies to measure changes in the angular
limits the photographic exposure to a sector for velocity of a missile at water entry, especially in
each flash of the stroboscopic lamps. The disk is pitch , although some changes in yaw may also
rotated at any speed up to 3400 rpm + 0.5% of occur. Because the change in pitch velocity is
set speed with a General Electric Thymotrol often impulsive, it is referred to as the “whip ”
controlled-speed motor. The speeds are set with of the missile.
a revolution counter and a stop watch for low The apparatus consists of a light source and
speeds and with a stroboscope for high speeds. A a camera housed on the tank-dome lid (Fig
Bausch and Lomb Tessar f /4.5, 11-7/ 8-inch ure 17.1 ) . The light source ( Figure 17.4) is an
focal-length lens was used for the tests described intense spark electronically timed to synchro
here, but any lens suitable for a 5 -by-7-inch view nize with the stroboscopic flash lamps. The
camera can be used . position of the spark is stabilized by a small hole
Before launching, the field of photographic in a ceramic insulator that guides the spark from
coverage, the number of exposures, and the time a tantalum electrode to the aperture in a
interval between exposures were determined. disk -shaped phosphor bronze or brass electrode.
From this information, the slot opening and disk A collimating lens converts the light into a beam
speed were derived and the electronic apparatus of accurately parallel rays directed along the
was set to give the preassigned number of lamp expected path of the missile.
flashes (exposures) at the desired frequency. By Two small mirrors (0.08 inch square)
mounted on the missile tail ( Figure 17.5 ) inter
cept a bundle of collimated rays and reflect
them to the whip camera ( Figure 17.6) , which is
focused to receive such light and concentrate it

Brass

Aperture disc : brass or

phosphor bronze, 0.010
in. thick, aperture 0.006
to 0.010- in. diameter
Tantalum electrode
8,000 volts to ground
0.060-in . diameter Spark guide “Lavite,"
Brass case at ground hole 0.006- to 0.010
in. diameter
potential

Figure 17.3. Rotating -disk camera open for loading. Figure 17.4. Light source for optical whip recorder.
( From Reference 94.) ( From Reference 89.)

270
 Parallel light beam from collimator

1
IL

Reflected ray
奔 Left yaw

Camera lens

Photographic plate
Mirrors
Yaw

Reflected ray

Figure 17.5 . Mirrors on missile tail to reflect light to whip Parallel light beam
from collimator
camera. ( From Reference 89. )

Point light
source
DES A
Down pitch D

Collimating
Camera lens
lens
Pitch Photographic plate
Parallel light beam
from collimator

Missile
Missile trajectory Figure 17.7 . Operation of optical system.
( From Reference 89. )
Whip camera

Reflected ray from mirror that is normal to the axis is traveling in

missile tail mirror the path of the parallel beam from the light
source , the mirror orientation will remain con
Whip camera lens stant and all reflected rays will be parallel and
Photographic plate at focused upon the plate at the same spot, as long
focal plane of lens as the missile does not rotate . Such will be the
case regardless of transverse motion of the
Figure 17.6. Optical whip recorder system. reflected light rays across the face of the lens
( From Reference 89. ) caused by the mirror motion on an oblique path .
If the mirror is not normal to the axis and the
missile rotates or rolls, its orientation relative to
in a point image on the camera plate . A ray the lens will change and the position of the
striking any portion of the lens is focused at the point image will shift. The recorder is sensitive
same point as any other ray coming from the only to angular motion of the missile and not at
same direction . If the ray changes its direction, all to translation .
the spot image shifts on the camera plate . Both Since the reflected rays subtend small
the spark-source assembly and the whip camera angles with the camera axis , missile orientation
are sealed off from the tank interior by optically change is nearly proportional to point-image
flat plate -glass windows that do not affect the displacement. Change in missile pitch and yaw
direction of the rays and hence the image may thus be obtained directly from vertical and
position. horizontal point displacements of the precali
The operation of the optical system is brated plate. If the possibility of roll is sus
shown in Figure 17.7 . When a missile with a pected, a second mirror with slightly different

271
orientation can be mounted on the missile , and image from the second mirror rotates about that
roll can be determined from the relative position from the first. Hence, when lines are drawn
of synchronous dots on the two traces. connecting any two pairs of corresponding
Time measurement is supplied by flashing points (such as aa' and bb ', Figure 17.8 ), the
the spark at accurately controlled intervals as angle between these lines represents missile roll
small as 0.00025 second, resulting in a succes during the interval.
sion of dots on the photographic record that fall To calibrate the whip - camera plate, the
on top of each other if there is no angular missile is placed in a jig whose optical system
motion and are spread out if there is. The use of duplicates that of the whip recorder and then is
a low flash rate during airflight minimizes accurately oriented to a position corresponding to
overlapping of the dots when the angular zero pitch and yaw along the air trajectory. The
velocity is small. To facilitate identification of wax holding the mirrors is warmed and softened
synchronous dots from the two mirrors, the so that the mirrors can be oriented to give
flashes are so generated that every tenth flash is suitably positioned dot images on a white screen
omitted. corresponding in size and position to the whip
Figure 17.8 shows schematically a typical camera plate. The plate is then substituted for
record obtained from the optical whip recorder. the screen and calibrated by exposing it to point
At the left , each dot is separated from its images from the mirrors corresponding to zero
neighbor by an oblique distance whose vertical and 1-degree positive or negative missile pitch
component represents change in pitch angle and with zero yaw. ( Calibration spots are shown in
whose lateral component represents change in Figure 17.8 ) . The calibrated plate is then used to
yaw angle. These measurable changes occurred obtain whip data.
in the known time interval between the two Figure 17.9 reproduces typical records
dots; curves can readily be plotted to show from the optical whip recorder, the missiles
change of angle versus time in both pitch and having heads of slightly different shapes. The
yaw. record at the left shows a slight tail-up rotation
The trace at the right of Figure 17.8 is an of the missile in airflight, accompanied by
auxiliary record made to measure the roll of the considerable angular motion in yaw . Imme
missile. The dots are images reflected from a diately upon impact, the tail started down while
second mirror mounted on the missile tail at a the yaw continued in the original direction. The
different orientation , the rays differing slightly record at the right shows practically no angular
in direction and focusing at a different spot on motion prior to impact, with a sharp reversal in
the photographic plate. As the missile rolls , the direction from tail -down to tail-up shortly after
impact . Comparators were used to measure the
records, with a precision corresponding to a
Yaw at missile orientation of 0.0005 degree.
Calibrating water
Water
spot entry
entry
Simultaneous Whip and Sideview Camera Data
b Roll during Initially, whip and sideview camera data
airflight
could not be obtained simultaneously because
al the light used to illuminate the model tank for
1 deg pitch Airflight Record from the sideview camera fogged the whip -camera
cocked mirror record. Separate launchings to determine whip
(tail up )
Pitch at
water entry
and trajectory were necessary , and a one-to-one
correspondence between missile trajectory posi
Record from
primary mirror 'T tion and orientation was impossible to obtain .
Dot positions at
intervals of 0.0003 sed
The use of light of different wavelengths for the
Zero pitch and yaw ( in two cameras solved the problem (References 53
degrees from tangent and 89 ) . Blue filters were placed over the flash
to trajectory ) lamps that furnished illumination for the side
view cameras and a red filter was placed over the
Figure 17.8 . Typical record, optical whip recorder.
whip -camera lens. The whip camera picks up the
( From Reference 89. ) red component of light from the spark source,
272
 Optical Comparators
The NUC Hydroballistics Laboratory data
consist largely of photographic records. To
determine time-position, attitudes, cavity char
acteristics, etc. , the data were measured with
micro- or optical comparators that read directly
in microns (cm x 10-4 ) and could be adapted to
35-mm and 16-mm motion-picture film , slides,
or cut film .
One comparator used in these studies ( Fig.
ure 17.10) was made by David W. Mann of
Lincoln , Mass . (Reference 94). The stage is set
horizontally and the microscope is mounted at
an angle, with a prism to bend the line of sight.
The overall magnification obtainable is about 6
to 1. This instrument has a measurement capa
bility of 200 mm in the horizontal (abscissa) and
30 mm in the vertical (ordinate ) directions.The
stage and cross hairs may be rotated, and both
are provided with circular vernier scales that
read to 1 minute of angle.

CONTROLLED -ATMOSPHERE LAUNCHING

TANK
The Controlled-Atmosphere Launching
Tank ( Reference 30) at CIT is used primarily to
study the hydrodynamic problems involved as a
free -flying body enters water from air, and for

Figure 17.9 . Whip records of two missiles.

( From Reference 89.)

but its red filter does not permit the filtered

blue light of the flash lamps to fog the whip
film .
The system used a tantalum rod (whose
spectrum contains red lines ) for the spark
source , a Wratten No. 24 red filter over the
whip - camera lens, and Wratten No. 67A blue
filters over the General Electric FT 125 or EG
and G FX-2 flash lamps. Kodak Royal Pan or
Super Panchro -Press Sports Type B film ( sensi
tive to blue ) was used in the sideview camera
and Kodak Super Panchro-Press Sports Type Figure 17.10 . The Mann comparator.
film (sensitive to red ) in the whip camera. ( From Reference 94.)

273
associated studies, such as underwater Empty, it weighs about 40 tons ; with water to a
explosions . Figure 17.11 is an artist's overall depth of 10 feet, its weight is about 150 tons.
view and Figure 17.12 is a cutaway drawing To provide adequate distance between the
showing the launching end of the tank. For launching plane and the data-recording cameras,
normal operation, the tank is about three a section of aa smaller cylinder is attached to one
fourths filled with water, leaving an air space side . The large rectangular hatch opening is
above. A centrifugal device , mounted on the located off-center on the top near the far end of
underside of a large hatch cover, launches the the tank. The entire launching mechanism is
model at any trajectory angle from vertical mounted on the hinged cover, which has an
downward to horizontal, with any attack angle O -ring pressure seal and heavy C-clamp frames to
up to £ 10 degrees, and at any speed up to hold it in place during launchings. The tank was
180 fps. A battery of high -speed motion-picture designed to withstand an external pressure of a
cameras record the path of the model during full atmosphere and an internal pressure of
both airflight and underwater travel. The 40 psi. It provides a clear launching plane 25
cameras operate without shutters ; exposures are feet long with a water depth of 10 feet.
made by intermittent illumination of the tank A high degree of water clarity is essential
interior with Edgerton -type flash lamps. The for underwater photography. With both lamps
fields of view of adjacent cameras overlap by and cameras on the same side of the tank, the
60%, and with the stereoscopic span thus light must travel 24 feet from the lamps to the
obtained, the path of the model can berecreated model and back to the cameras. Even a slight
step by step with analyzing equipment. The amount of color, fine suspension , or micro
flash lamps that illuminate the tank interior are organisms in the water could absorb or scatter
installed in six Lucite tubes that pass through the light before it reached the cameras. The
the tank above and below the underwater interior of the tank was designed not only to
cameras ..
prevent corrosion of the steel tank , to provide a
The tank is a large horizontal welded steel dark background, and to minimize the pos
cylinder, 13 feet in diameter and 29 feet long. sibility of damage to models from striking tank

Cm

Wi
lk

Figure 17.11 . Controlled-Atmosphere Launching Tank at California Institute of Technology. ( From Reference 30.)

274
 R
ORTE
PRIP

Figure 17.12. Cutaway view of launching end of CIT tank. ( From Reference 30. )

walls, but to avoid any contamination of the the model from rotating about a transverse axis
water that might impair its optical properties. as it goes around with the wheel, and with a
To meet these diverse and stringent require mechanism for releasing the model at any
ments, a polyvinyl chloride plastic (Koroseal) , in predetermined point along a 90 - degree arc.
sheets 3/32 inch thick, was cemented over the The hatch cover is hinge-mounted so that
entire interior, the seams being heat-sealed with the launcher can swing down into the hatch
strips of similar material. Commercial sand and (shown covered in Figure 17.13 ) . A model is
alum filters remove suspended materials from shown in place in the chuck. Figure 17.14 is a
the water , a row of germicidal ultraviolet lamps multiple-exposure photograph showing the fixed
along the ceiling prevent bacterial growth , and a orientation of the model on the wheel. The
vacuum pump controls the atmospheric wheel, a heavy steel plate having sufficient mass
pressure . to insure uniform velocity , is supported on a
When the local water supply happens to stainless-steel shaft mounted on four preloaded
contain traces of yellow matter in solution that precision ball bearings , assembled in a quill to
cannot be removed by filtration , the water form an accurately aligned unit. The launcher is
absorbs the blue and violet light of the flash driven by a 10 -hp DC motor whose speed is con
lamps. For this reason, distilled water was used trolled electronically. The control is activated
in the tank . by a selsyn generator driven by chain from the
launcher shaft .
The Launcher The model is counterbalanced by a
The centrifugal launcher consists of a rotat movable weight on a screw in the plane of
ing wheel that carries the model near its rotation of the model and displaced from it by
periphery, with a planetary - gear system to keep 180 degrees. No provision is made for shifting

275
 Rio

Woad Lid

Figure 17.13. Launcher with model in chuck. Wood cover over hatch opening.
( From Reference 30.)

this counterbalance after the model is released

because the structure is massive , and a slight
imbalance is not critical once the model is free.
The chuck (Figure 17.15 ) is shown with a
steel ribbon in place. The visible end of the
ribbon is fitted into an adapter in the cylindrical
08

section of the model ( Figure 17.16) and holds

the model in place ( Figure 17.17 ). Launching is
accomplished by hitting the ribbon at the
notched end with a rapidly moving knife in the
chuck assembly, actuated by a release arm
tripper ( Figure 17.18). The model moves along a
tangential path , and the chuck , continuing its
circular motion, gradually lifts away from it.
The rigidity of the model seat reduces to a
minimum the energy stored in it that might
affect the motion of the model at release . The
entire chuck mechanism is mounted on ball
bearings.
Figure 17.14 . Multiple-exposure photograph Figure 17.18 also sketches the opposite
of model in chuck. face of the wheel with the planetary -gear system

276
 Ribbon

Figure 17.15 . Launcher chuck with steel ribbon suspension.

and launching controls. The gear train, com gear; the central layer and the other outer layer
posed of specially cut fine-pitch precision gears, engage the chuck gear. The hub and chuck gears
prevents the chuck from rotating around its own are equal in diameter.
axis . To insure smoothness of operation and to Two levers or arms are bearing -mounted on
prevent backlash, its idler gear is made in three the wheel shaft and prevented by clamps from
layers. The central layer is integral with the rotating. The trajectory -angle arm clamps to the
spokes and hub ; the two outer layers are ring trajectory -angle scale, has a 90 -degree adjust
gears, loaded against the central layer by small ment, and is integral with the release arm , which
tangential coil springs in the rim. The central carries the solenoid-operated tripper. The attack
layer and one outer layer mate with the hub angle arm , attached to the central gear of the
277
 Ribbon

Figure 17.16. Adapter in cylindrical section of model.

transmitted to the electronic interlocks that

synchronize the operation of the tripper with
that of the cameras and the lights .

Trajectory -Recording System

The cameras and flash lamps are described
in Reference 29. The cameras are used as .

precision measuring instruments ; many of their

special features were designed to meet this
requirement. The camera system should com
pletely cover the underwater volume of the tank
as well as the trajectory from launcher to water
surface . The duration of aa test run is 1 second or
less; photographs were taken at rates between
500 and 3000 per second, depending upon the
launching speed and the anticipated
Figure 17.17 . Model held in chuck .
accelerations .
The synchronized high -speed motion
picture cameras used standard 35 -mm film . The
planetary system, is clamped to the trajectory main bank of five recorded the underwater
angle arm with an adjustment of +10 degrees on trajectory; two or three additional cameras
the attack-angle scale. recorded the air trajectory. In the launching
When the trajectory -angle arm is set at plane, the adjacent camera fields had a 60%
zero, the release arm hangs vertically in position overlap; in the vertical direction the field of view
to release the model on a horizontal trajectory. covered the entire water depth ( Figure 17.19 ).
If, at the same time , the attack -angle arm is set at Such multiple coverage makes it possible to use
zero, the axis of the model will be horizontal a stereoscopic technique to measure six com
(i.e. , parallel to the tangent at the point of ponents of motion.
release ) and the model flies with zero attack Each camera was attached to a flanged I

angle. opening in the tank shell, looking in through a i

On the trajectory -angle arm , on the side small window . The front nodal point of the
facing the wheel, there is a light source and a underwater camera lens was at the center of cur
photocell . A small mirror mounted on the wheel vature of the optically ground spherical window.
reflects light into the photocell at every revolu All primary rays passed through the air-water in
tion when the chuck is about 20 degrees ahead terface at an angle of 90 degrees without refrac
of the launching point. A signal is thus tion ; there was no distortion or reduction of
278
 Indicator
support

50 40

70
Trajectory angle scale
10
80 Trajectory angle clamp
Trajectory angle indicator

Attack angle scale

Attack angle indicator
Drive wheel
Photocell (hidden )
Spring loaded idler gear
Attack angle clamp
Driving gear (hidden )
Photocell mirror
Driven gear
Release arm
Locking lever
do

Model chuck
Tripper

Figure 17.18. Launcher controls. ( From Reference 30. )

field. The curved water surface was a negative

Air trajectory camera
lens that increased the effective focal length of
Air trajectory window
the system and shifted the focal plane , causing a
Underwater field reduction of 4%, compared to a field
trajectory camera Water level
reduction of 29% with a plane window .
A continuously moving film was required
in order to provide the desired frame height and
to meet the design requirement of 3000 frames
per second, maximum. The selected speed of
31.25 fps gave a standard 35-mm frame height
of 3/4 inch at 500 frames per second, requiring
the extremely short exposures (on the order of
Underwater 2 us) that can be obtained from Edgerton-type
trajectory flash lamps.
correcting Figure 17.20 shows one of the underwater
window
cameras, with film magazine, spherical window,
Launching plane and mounting flange. The lens is a 1 -inch , f /2.3 ,
Transparent tube
Bausch and Lomb Baltar. To prevent friction,
the film is guided through the focal plane by
rollers instead of the usual pressure-plate
Multiflash lamp arrangement. The film magazine can be de
tached from the camera and tilted to a hori
Figure 17.19. Camera field of view in vertical direction . zontal position for loading. A 32-foot length
( From Reference 30. ) of film , sufficient for a single run , is stored in

279
 !

-Magazine 1
Keeper rollers
Film spools

Run-data printer
Continuous
path of film

Supporting flange
Rubber seal
Spherical correcting
window

Lens
0000

FH
Gate rollers

Drive shaft
-Drive sprocket

Magazine release
Magazine light lock

fo

Figure. 17.20. High-speed recorder camera. ( From Reference 30. )

the magazine in a number of passes over the two the housing brake is released , and power is
sets of idler spools. The ends of the film , which applied. The housing begins to rotate , comes up
extend through the light locks, are spliced to to speed, and is synchronized. The shaft brake is
form a slack loop about 2 feet long outside the then released and the housing brake is applied
magazine. gradually. As the housing slows down, the shaft
The loading operation is completed by takes up the difference between the syn
tilting the magazine into the vertical position chronous speed and the housing speed, and
and threading the loop over the guide rollers and when the housing stopped, the shaft was running
drive sprocket of the camera. With the film in a at synchronous speed. By means of time-delay
continuous belt, it is possible to bring it up to relays, this sequence of events proceeded auto
speed gradually, expose it, and slow it down matically when the motor power was turned on.
again without wasting film or using long leaders. To slow and stop the film, the sequence was
All cameras are driven through line shafting reversed .
by a single synchronous motor ( Figure 17.21 ) . The film travel was synchronized with the
To provide gradual acceleration and deceleration instant of launch so that the short strip of film
of the film , the motor housing rotates in exposed during loading did not pass through the
trunnion bearings; power is brought in through camera during the actual recording period. A
slip rings on the left-motor face. Two electrically microswitch , actuated by a cam driven by a
operated brakes stop the shaft and the housing. reduction gear from the camera shaft , made one
To start the motor, the shaft brake is clamped , contact for every pass of the film belt . In
280
 37

.O . 0000
ក ក ក

Figure 17.21 . Underwater cameras and driving motor. ( From Reference 30.)

operation, the camera shaft was first rotated

manually until the microswitch closed ; then all
the films were threaded into their respective
cameras , thus aligning the position of the splice
with the microswitch. When the camera drive
was running, the microswitch signaled each
passage of the splices through the cameras and
controlled the operation of the launcher release .
Two other switches, activated by the same cam,
controlled the operation of the flash lamps. To
insure that the splices remained synchronized
throughout the run, all the film belts had to be
of exactly the same length, i.e., having the same
number of sprocket holes . A film -loading device
measured out the required length .
The cameras are operated with lenses con
tinuously open, and exposures are made by
intermittent illumination of the tank interior
with flash lamps. A system consisting of from 30 Figure 17.22 . Lamp with Lucite reflector.
( From Reference 30. )
to 42 Edgerton-type flash lamps , all operated
simultaneously, was constructed. The individual
lamps are synchronized with each other within
less than 1/4 us. Each lamp is a quartz tube control panel. Each is operated through an
about 8 inches long, filled with xenon , hydro individual surge circuit that receives 4,000 volts
gen , and a trace of radium bromide , with two of DC power from a large rectifier, and operates
metallic electrodes sealed into the ends of the at twicc this value through a voltage doubler.
tube . Two types of lamps are used: one a The power consumption is approximately 0.8
straight tube with an aluminized Lucite reflector joules per flash ; at 3,000 flashes per second , the
( Figure 17.22 ) , the other a helical tube in a battery of 30 lamps requires a continuous input
headlight-type sealed -beam reflector with a of approximately 80 kw. The lights were operat
smooth lens (Figure 17.23 ) . ing only about 1/200 of the time , making the
The power for each light is carried through rate of power input during illumination more
an individual coaxial cable running to the than 16,000 kw. The heat generated in the tubes
281
 which case it would be impossible to launch the
model. Provision is made therefore to change the
phase relationship by slowly rotating the
camera -motor housing with a small electric
motor geared to the housing brake.
Data -Analyzing System
The data-analyzing system was essentially a
duplicate of the recording system. Projectors
took the place of cameras and a movable screen
replaced the model. All the films from one run
in the launching tank were placed in the
analogous projectors with the film strips
synchronized so that the corresponding frames
taken at the same time would be projected at
the same time . Once the film strips were
synchronized , they remained so during the
projection of the entire run because the film
drive of the projectors was a continuous shaft
Figure 17.23. Headlight-type lamp . ( From Reference 30. )
(Figure 17.24 ).
There was only one position in this space in
themselves limited the length of operation. The which the two images could coincide . The
maximum for successful high -speed operation is exploring screen of the analyzer was adjusted
3600 flashes per run. until the two images fused into one, the move
The steps in the photography of a typical a
ments being transferred to a battery of counters .
model launching were as follows. When the screen registered precise image fusing,
The camera magazines were loaded and the the counters indicated the projectile position in
model was installed in the launcher, which was space corresponding to the pair of photographs.
set for the desired trajectory and pitch angles The analyzer was built to a scale of one half that
before the hatch cover was closed . Then air of the tank .
pressure was adjusted, the launcher brought up The projectors ( Figure 17.25 ) are precision
to speed, and the camera motor was turned on . instruments. Lenses were procured in matched
The automatic launch sequence was started pairs , one for the camera and the other for the
by the operation of the first of three interlocks.
The second interlock was actuated by the
microswitch on the camera drive immediately Image from projector
No. 3
after the exposed film splices passed through the
focal plane of the camera. The third and final Exploring screen out of position Image from projector
No. 2
interlock , which trips the model and launches it ,
was actuated on the next revolution of the
launching wheel. The flash lamps began to Exploring screen
at model position
function simultaneously with the release of the
model and continued to flash until the entire Motion of image//
length of film was exposed, at which time the Moti
lamps cut off and the equipment was automati
cally shut down in the reverse order of the Photo in
starting sequence . dex head
Images from
The camera motor runs at a constant speed. projectors
When the launcher runs at any multiple of that No. 2 and
No.3 coincide
speed, the passage of the film splice through the
camera and the passage of the model past the Projector No. 1 2 3 4 5

launching point occur with a fixed phase rela

tion between them and may never coincide , in Figure 17.24 . Data-analyzing system. ( From Reference 30. )

282
 Exploring Screen

Figure 17.25 . Data-analyzing equipment . ( From Reference 30.)

corresponding projector. The gate mechanism the rear of the analyzer were aligned before each
held the film exactly in the focal plane. The measurement .
light source was kept as low in intensity as The exploring screen was a small disk with
feasible for accurate readings in order to elimi a half-model attached to it ( Figure 17.25 ) so
nate the heat that can affect both film and that the final projection was on a curved surface
optical-system dimensions. Water cells and air similar to the one photographed. The circular
cooling were also used to control temperature. exploring screen was carried on a mechanical
Reference marks on the back wall of the transport that provided three linear and two
launching tank and on a background screen at angular motions. A sixth motion was achieved
283
by rolling the half-model about its longitudinal district permits work throughout the year. A
axis. A carriage spanning the length of the room simplified map of the range area showing the
rode on overhead rails and transported a smaller principal facilities appears as Figure 17.26.
carriage on a pair of transverse rails. Both Projectiles launched from the VAL have a
carriages had rack-and-pinion drives. A pair of clear range 1000 yards long and at least 200
vertical guide tubes with a screw drive , sus yards wide into water that varies in depth from
pended from the transverse carriage , could be 150 to 190 feet. Surrounding mountainsides
rotated in azimuth to provide the second of provide excellent locations for detail and
three angular motions, the third being obtained general-view camera stations . Both the FAL and
by rotation of the circular screen in its own the VAL ranges are instrumented with hydro
plane. Selsyn repeaters and mechanical counters phones and nets to record underwater trajectory
functioned as position indicators to transmit the data. The short airflight precludes studies of
image-position information to the operator's aerodynamic behavior or air stabilization.
desk. Roll was read directly from a scale on the
screen . Fixed - Angle Launcher
The FAL is a blowgun , 22.5 -inch inside
PROTOTYPE LAUNCHING FACILITIES AT diameter and 300 feet long, that uses com
NUC pressed air as a motive power. The whole system
The NUC Morris Dam Torpedo Range near is mounted at an angle of 19 degrees with the
Azusa, Calif., provides facilities to launch proto horizontal. Plan and elevation views are given in
type missiles from air into water under con Figure 17.27 and the hydrophone array in
trolled conditions of attitude and velocity and Figure 5.12 in Chapter 5. The launcher is
to measure water entry and subsequent anchored rigidly at its breech end and is sup
behavior. The Fixed -Angle Launcher ( FAL ) and ported along its length in such a way that it is
the Variable-Angle Launcher (VAL) are installed free to expand lengthwise and yet is held in
on a peninsula approximately 1000 yards alignment. A keyway the length of the tube can
upstream from the dam. The mild climate of the be used to keep the torpedo from rolling. A

1450
1450
1950 1400
State Highway

1300
1300
1400
1350 N 1900
1350
1250
1450 1200
1150 1250
1400 1250
1200
1350 1200
1200 150 1150
1 200
1150 Side view camera car track
1100

1050
1000 뭉
950
1150
1200
Variable
angle T
Morris Dam Fixed angle launcher range
Morris launcher
small
Dam range
950 caliber
1000 1150
1050 range 1 200

1100
Propulsion
1150 laboratory ?
1200

Morris Dam
torpedo range 1 200
1250

1150
1200
12501

Im
0 400
Scale of feet

Figure 17.26. Morris Dam Torpedo Range facilities.

284
 Road

Paved work area

TE Bridge

Stairway Shore line

Loading
carriage

Air tank
in
y joint
Stiffleg derrick Valve Launching
PLAN tube
Loading carriage

Bridge Trench

Concrete channel
Trench anchorage

Launching tube

Rocket booster stations

Tube support

ELEVATION Trestle

Figure 17.27 . Fixed-Angle Launcher ( FAL) .

1-ton projectile can be launched into the water trajectory can be measured for almost 1000
at velocities of 400 to 500 fps. The FAL yards. Instruments inside the torpedo measure
(Reference 59 ) was extensively used until the missile attitude as a function of time or record
more versatile VAL was constructed. its operation and that of its components. The
projectile is recovered either by mechanical
Variable-Angle Launcher devices or by divers .
The VAL be adjusted to angles
can The main launcher ramp is constructed on
between 0 and 40 degrees with respect to the a natural 45 - degree slope on the downstream
horizontal ( Figure 17.28) and is designed to side of the peninsula; a counterweight track runs
launch a 1500-pound missile at velocities up to down a natural 30 -degree slope on the upstream
1000 fps. Tests can be carried out under side (Figure 17.29 ).
accurately controlled conditions over a range of The VAL is a blowgun, 22.5 -inch inside
velocities and entry angles. The missile entry diameter, and 300 feet long. It is supported
point can be predicted with sufficient precision within a movable bridge that was designed to
that cameras and other instruments can be used accommodate additional tubes2 ranging in
at short range, either from land -based or diameter from 8 to 21 inches. Figure 17.30
anchored stations , enabling missile orientation , shows the tube muzzle and the bays available for
angular and linear velocities , and deceleration to other tubes. The movable launcher bridge is
be measured during water entry . Detailed entry supported at the outer end on two pontoons
photographs show the establishment of flow connected by a 95 - foot-span and at the shore
around the missile nose ; broach and re-entry are end by a carriage traveling on a 45 -degree ramp
also recorded photographically. The underwater ( Figure 17.31 ). The inclination of the launcher

2The large tube can be used for subcaliber launchings; the subcaliber missile rests in a sectioned wooden sabot that
separates from the missile during airflight.

285
 Figure 17.28. Variable -Angle Launcher (VAL ).

Counterweight car

Figure 17.29. VAL operative equipment, including counterweight car.

286
 WANN

Figure 17.30. VAL launching tube and bays for additional tubes.

287
 TE

ERA

Figure 17.31 . VAL on supporting pontoons.

bridge is adjusted by shifting the carriage up or counterweight car runs on a track of standard
down the ramp (Figure 17.29 ) , the weight of the railroad rail laid over reinforced Gunite paving.
bridge and carriage being balanced by a counter A large concrete structure supporting the
weight car on the back slope of the hill . The launcher rails and the counterweight track was
bridge is moved by a motor drive acting on built over the ridge between the launcher and
cables connecting the bridge support carriage to counterweight car ( Figure 17.32 ) . The rails for
the counterweight car. The launcher bridge is the carriage and counterweight car are continued
noteworthy in that it was, when built, the to the top level of the cellular structure on slabs
longest all -welded steel span in the United States integral with the side walls and webs. The space
( 300 feet long, 22 feet wide, and 35 feet high ) . within this structure houses electronic control
The carriage supporting the inner end of and recording equipment, operating machinery ,
the launcher bridge moves on special rails etc. In all, there are seven levels ; the uppermost
weighing 1200 lb/yd , or about 10 times the carries the main-drive machinery by which the
weight of a standard railroad rail. The carriage is bridge is moved .
secured to the rails by four 7-inch-diameter pins A 40-foot steel tower was erected at the
to resist the recoil of the launcher. The 600-ton top of the concrete structure (Figure 17.32 ) to

288
 Overhead Camera Cable

Hoist Control House

Main Hoist
Upper Deck

Recording Room ,
Launcher Ramp
440 Volt Load Centers
Counterweight Car
2400 Volt Main Switch

Battery Chargers Hoist

Motor -Generator
Low Voltage Power Transformers

Lighting Transformers
Instrumentation Connection Room
Torpedo Loading Platform

O
R
T
5
1

Figure 17.32. Concrete structure of VAL.

provide a vantage point for photography and to on the inside of the launching tube breach door
support an overhead cableway for a camera car. was connected to a radio frequency oscillator,
Further details concerning the construction the launching tube itself acting as a cavity
and instrumentation of the VAL can be found in resonator. A 394-MHz signal caused resonance
References 27 and 78 . to occur every 24 inches as the projectile
progressed down the tube .
Instrumentation 3
Because a prototype launching needs pre Flare Camera
cision instrumentation for assessment of results, This camera is installed on the sideview
a variety of instruments and techniques was camera car that can be moved to a position
developed for the Morris Dam Torpedo Range . opposite the entry point ( References 12 and
Because a missile's underwater trajectory is 27 ) . The track for the camera car is parallel to
influenced by the motions it has experienced the center line of the range and 456 feet away
during airflight and water entry , accurate meas on the west bank (Figures 17.26 and 17.33 ) .
urements must be made of entry velocity , The entry velocity , angular velocity and
attitude , trajectory angle, angular velocity and acceleration in a vertical plane, and linear
acceleration in both pitch and yaw, and linear deceleration before the tail of the missile
deceleration during the early phase of water entered the water were measured precisely from
entry . the flare-camera records . Two brilliant pyrotech
The velocity and acceleration of the nic flares attached to the missile tail were
torpedo in the launching tube were determined photographed at a rate of 1000 exposures per
from data recorded by the torpedo position second at an exposure time of about 0.0003
interferometer (Reference 64 ), an electronic second. A solenoid -operated shutter controlled
device that records the time the torpedo passes the exposure time .
stations spaced 24 inches apart. A coupling loop

3Described more fully in References 27 , 59 , 78 , and 106.

289
 General rearview
camera, station

General sideview
camera stations

Sideview
camera car

Review camera
stations

Hydrophone 1

IL array

Figure 17.33 . VAL camera stations and hydrophone array .

The optical system prevented blurring of Because the path of the projectile, as seen
the image by compensating for the forward by the rotor, must be normal to the axis of the
velocity of the projectile (Figure 17.34) . The rotating spindle , two movable reflecting prisms
photographic plate was stationary in the camera were used to accommodate the various tra
and the image of the flares was focused on it by jectory angles instead of rotating the camera
a specially designed achromatic meniscus lens of about its optical axis. Adjustment of the prisms
70 -inch focal length and 1 -inch aperture . An also accommodates the camera to changes in
accurate plane-parallel plate mounted in a rotat lake level. The exact speed of the rotor, which
ing spindle between the lens and the photo must be known in order to determine the time
graphic plate rotated at 30,000 rpm, and interval between individual exposures of the
refraction of light held the image stationary flare images, was determined by comparing
during each flare exposure. The retainers holding pulses from a variable -reluctance pulse generator
the plane-parallel plate in the spindle were built into the spindle drive with a time scale
shaped to give an intermittent shutter action from the master time base . Both traces were
with light transmitted during 47 degrees of recorded on a single oscillograph record .
spindle rotation. Plate thickness was determined The flare images appear on the photo
by missile velocity and spindle speed. It is graphic plate as a series of regularly spaced
feasible to use a single plate during a series of marks approximately at right angles to the
11 launchings because satisfactory definition occurs missile trajectory. The spacing between succes
over a usable velocity range . sive images of the flares represents the speed of
290
 Flare
( 5,6 )
Photographic plate
Projectile Point images of future
elements of travel Without compensating
Element of travel for plate , image of flare
which record on plate is Rotating plane (3,4) would blur over this
parallel plate Position of image range during time in
to be a single spot ( 1,2 ) terval from T1 to T2
Flare position of T2 without com
Lens pensating plate
at T1
( 1,2 ))1
ki #
'I2 T3
TA Position Position
.T5 at T2 at T1
1.•716 30,000 rpm
Future elements of
travel to appear on ( Exposes plate when between +15 deg
plate as spots ( 3,4 ) of position normal to axis of camera ) Position of image of T1 without
( 5,6), etc. compensating plate

With compensating plate, image of

flare remains at this point during
time interval from T1 to T2

Figure 17.34. Optical system of flare camera.

the torpedo, providing an accurate determina other; time from the master time base was
tion of velocity during free flight and of recorded on each frame in each camera to
deceleration during water entry . A typical flare correlate the data.
camera photograph is shown in Figure 17.35 . Three 70-mm Mitchell cameras mounted at
The complement of the angle made by lines stations along the road bordering the west bank
joining synchronous flare-image pairs with the of the reservoir operated at 32 frames per
torpedo line of flight gives the angle of attack . If second, jointly covering the first 1000 feet of
the line of flight does not change perceptibly, the range. The last two cameras can be pivoted
the rate of change in angle of attack is equal to to follow the projectile still farther down range.
the rate of change in pitch. A carefully posi Two 35-mm Mitchell cameras operating at
tioned moored spar provides a horizontal refer 120 frames per second recorded the details of
ence for pitch determination as well as a scale water entry. One was set up on the sideview
factor for distance determination. The flare camera car (Figures 17.26 and 17.33 ) and the
images are straight line segments and syn other was suspended directly above the entry
chronous images lie in a superimposed straight point on an overhead cableway. A 16-mm
line. Mitchell camera operating at 64 frames per
The plates were measured on the compara second can be mounted on the connecting
tors that were used for the model data. The scale bridge below the muzzle of the launcher to give
of the plate was determined by fiducial marks a close-up rear view of the water entry , and a
on the spar. The range position of a missile can general rearview camera (Army K-25 operating
be measured to within 0.01 inch from a good at about 1 frame per second) on top of the
flare image and projectile attitude in the pitch tower supporting the camera cableway can also
plane can be measured to within 0.1 degree. photograph the trajectory and the wake.
Motion Picture Cameras Underwater Trajectory
High - speed Mitchell motion-picture Recording the underwater trajectory of a
cameras recorded the airflight, linear and angular missile is a difficult yet essential procedure. Data
water-entry motion , and splash of the missile . Its concerning the path of the missile and its
wake, broach, and re -entry were also recorded. velocity and orientation along the trajectory
The cameras were not synchronized with each were obtained in several ways.
291
 Figure 17.35 . Typical flare camera record.

The hydrophone array ( Figure 17.33 ) sabot striking the water , and reflected noises
recorded missile position at several known times from the launching. A magnetic loop marks the
during the underwater trajectory. Sound pulses, time at which a magnetized missile passed
produced by the timed explosion of blasting through it, but does not define missile position
caps inside the missile , were picked up by the or orientation .
hydrophones, amplified, and recorded on an The loops , each 25 feet square and consist
oscillograph. In an array 75 feet deep, 150 feet ing of 80 turns of copper wire , are hung
wide, and 750 feet long, 24 hydrophones were vertically at selected positions across the range.
held in position by buoys and anchor lines in The passage of the magnetized missile produced
five 150-foot-long bays. For one launching, 12 a pulse of current to operate the galvanometer
hydrophones recording on a 16-channel oscillo of a high -speed oscillograph. The pulses were
graph were usually selected. The cap explosion strong enough to require no amplification.
times were tallied with an accurate time scale on Nets (Reference 11 ) give aa direct means of
the oscillograph record. Torpedo positions were locating points on the trajectory. Twine nets
determined by time differences, reduced to (similar to fish nets but made with non-slip
equivalent distances in feet, and set into the knots) were used for this purpose . The hole torn
scale board of a computer (Figure 17.36 ) . The by passage of the missile provided the coordi
position of a missile could be determined within nates of a point on the trajectory, but could not
# 1 foot on each coordinate , and the average record time and missile orientation .
velocity between points on the trajectory to Electric nets have also been used . No. 20
within + 1 %. copper wires are stretched at 9-inch intervals
Magnetic loops were sometimes used horizontally across one face of a frame and
instead of hydrophones during the early part of vertically across the other, with a 1 -inch space
the underwater trajectory because of the high between the two arrays. Passage of the torpedo
noise level from the cavity, turbulence, pieces of forces one set of electrified wires against the
292
 Figure 17.36. Computer for reducing hydrophone range data.

other, thereby closing a circuit to record time of record a particular time interval. A master time
passage. The broken wires marked the position base correlates launching records from oscillo
of torpedo passage , but these positions were less graphs by supplying the same time indications to
accurate than those determined by nets made of all records, obviating the need for individual
twine. synchronized recording units. It was governed
Internal gyroscopic instruments were by a fork-controlled oscillator used in conjunc
developed to record missile attitude during air tion with frequency -divider networks. Con
and underwater flight (Reference 63 ). Ordi sistently high accuracy was achieved by
narily, two gyroscopes were used so that orienta compensating for temperature changes.
tion about three axes could be determined from Each motion-picture camera photographed
the recorded data. The gyroscopes were a cluster of seven 2-inch cathode-ray tubes
equipped with commutators that indicated ( Figure 17.37 ) that record the launching number
angular motion in 1 /2 -degree steps . A bank of and the time as generated by the master time
miniature neon lights connected to the commu base. A chronograph attachment on the Mitchell
tator flashed in sequence during angular motion motion-picture cameras focused the image of
of the gyro gimbals and exposed a continuously these tubes onto one corner of each frame.
moving 16-mm film to form a series of dashes Three of the cathode-ray tubes recorded launch
indicating direction and angle of rotation . ing number and four indicated the time by a
It is important to the success of hydrobal decade system ( Figure 17.38 ). The spot on the
listics testing that a precise time base be first tube moved clockwise around the tube face
recorded on each photographic and oscillograph in ten steps at 1 -second intervals, those of the
record for coordination of the results from the second and third tubes moved at 0.1 -second and
several instruments and for use in velocity 0.01 -second intervals, respectively , and the spot
calculations. Some instruments have to be on the fourth tube moved continuously in a
started at a predetermined instant in order to circle , completing one revolution every 0.01
293
 1

The launching-valve mechanism , the oscillo

graphs in the recording station , and the cameras
at the various remote stations can be operated
either manually or by an automatic-sequencing
system. The equipment performs three
3 functions:
2
1. A pulse generator and a group of
selector units start and stop the
launching controls and instruments at
any time during a 10-second interval
1
after the launching operator has
started the pulse generator (timing
from the pulse generator is accurate to
0.1 second ) .

.
2
Precision electronic timers accurate to
0.01 second are started by the initial
motion of the projectile to give time
delay and function durations of up to
2 seconds .

Figure 17.37 . Master time base unit . 3. Auxiliary motor - driven timers
sequence functions requiring more
than 10 seconds of operation .
Starting times were selected so that
motion-picture cameras and oscillographs were |

at operating speed prior to launching. The

sequencing system can control launching
mechanism and instruments; any desired com
bination of automatic or manual operations can
be achieved. It can also turn off each instrument
at the end of the test .
Instruments can also be started by switches
or break-rods actuated by passage of the torpedo
through the launching tube or by motion of
some part of the launching mechanism . A
channel of the sequencing system was usually
included in the circuit so that time delay and
duration of operation could be controlled, as
well.
Missile underwater photography in open
ranges (e.g. , the Morris Dam Torpedo Range) is
Figure 17.38 . Master time base record showing described in References 13 and 79 ; it is far more
time 1.533 seconds.
complicated than underwater photography in
model tanks.

OTHER MODEL FACILITIES IN NUC

second. The precision was 0.001 second when LABORATORY4
the trace on the fourth tube was read to 0.1
revolution . A time scale from the master time Open Launching Tank
base was also recorded on a channel of each The large model tank measuring 25 feet
recording oscillograph. long by 52 inches high by 54 inches wide , with

4Further details are in Reference 54 .

294
an inside width of 44 inches, is shown with its tank; canvas and felt blankets suspended above
catapult-launching arrangement in Figures 17.39 the water down range caught broaching models.
and 17.40. The glass section nearest the launcher The launcher is a 5-foot cross bow ( Figure
was 47 inches high by 66 inches long by 3/4 17.40 ) . The wire -rope bow string is attached to
inch thick so that the top rail did not obstruct the ends of two cantilever springs, each consist
the photographic view. Glass panes 3/4 inch ing of eight leaves of 3- by 3/ 8-inch spring steel.
thick, 40 inches high , and 66 inches long The combined spring constant is 443 lb/in ,
comprised the other three sections , separated by which, at the safe working stress of 100,000 psi,
U - frame columns 6 inches wide . No tie bars develops about 900 ft-lb of energy . At the
crossed the width of the tank ; the columns maximum 18-inch draw , the bow develops
supported the entire hydrostatic load (40 inches about 3,500 pounds of pull so that a 3/ 16-inch
of water). The section nearest the launcher was bow cable is required. A 2-inch-diameter model
also able to withstand shock loads arising from weighing 1.1 pounds can be launched at about
missile water entry . 140 fps .
The tank contained 2,200 gallons of water The whip camera and optical accessories
continuously filtered to remove particles as are mounted on the launcher carriage, together
small as 0.5 micron . Commercial sodium hypo with the pickup for a photoelectric device used
chlorite was added to the water ( approximately to determine launching velocity. A double rail
1 gallon to the tankful) to prevent the formation guides both the launcher carriage , which is
of algae. Sufficient clarity was achieved to allow attached to the draw rope , and the electrically
excellent photography. operated draw -screw trip mechanism . Scales and
Diving models were stopped by strips or adjustable -limit switches are mounted on the
rolls of rubber cushion along the bottom of the guide rails so that the bow can automatically be

Launcher

Figure 17.39 . Open Launching Tank.

295
 Figure 17.40. Catapult launcher for Open Launching Tank .

set for a predetermined velocity, and so that windows at the front and rear ( Figure 17.41 ).
timing sequences can be initiated by carriage The absolute pressure in the tank can be varied
movement . from 2 atmospheres to a value corresponding to
A cable lift rotates the launcher carriage the vapor pressure of the liquid within the tank. 1

about a horizontal axis through the water Mineral oil , slightly colored with a red dye
impact point, maintaining a constant free air for contrast with the water, is used for the
flight distance and constant entry point for all second liquid (Figure 17.41 ) . The interface
launching angles from 3.5 to 45 degrees. The between the two liquids can be cleaned by
launcheris raised or lowered by a hand crank on skimming and filtering. The tank is equipped
a double - drum hoist winch and can be clamped with fittings for thermometers and gages, and a
at any angle between the 3.5- and 45 -degree vertical notched scale mounted inside the tank
limits. It is possible to launch missiles with provides the necessary fiducial marks.
angles of attack up to 10 degrees. Projectiles can be dropped from the top of
the tank above the oil surface or started from
Liquid -Interface Tank beneath the surface by an impulse launcher
This tank permits the study of water-entry designed to impel hardened-steel balls of 1.5- or
phenomena as a projectile passes into water 2-inch diameter downward from a starting posi
from another liquid rather than from air tion in the oil layer. The launcher was developed
(Chapter 13 ). The presence of the second liquid to meet the need for a device that would not
eliminates splash and other conditions that entrain air in the liquid or impart turbulence to
hamper observation of water -entry phenomena. it and would not affect the pressure in the tank .
In this tank, a missile can also be propelled from The impulse launcher is mounted on top of
a standing start under the surface of the liquid , the tank ( Figure 17.41 ). A steel ball of high
thus eliminating the entry cavity. magnetic permeability is held lightly on the
The tank measures 46 inches high , 15 lower end of the impulse rod by the magnetic
inches square, and has 0.75 -inch -thick plate-glass field from a solenoid ( Figure 17.42 ) . A stud
296
 Falling weight
impeller ( steel)

Weight release solenoid

72 in .
LA *
Vatros

Holding
stud
Tube weight
guide ( brass
Variable height 1 1 /4-in. ID )
Steel sleeve

Air vents
2 in.
1/8 in.

Top of
tank
Gaskets

Impulse rod
( 1 / 2 -in . diam )
10 in.
Solenoid coil
Mounting plate support sleeve
silver soldered
Oil surface
to impulse rod
1/16 in .
1 1/2 in.
Solenoid
holding coil

Depression to
seat ball
Stainless -steel ball
Oil- water interface 1 1 /2- or 2-in. diam

Figure 17.42. Impulse launcher.

and to impel the falling weight at higher

velocities with compressed air.
Figure 17.41 . Liquid -interface tank .
Vertical-Drop Tanks
Two tanks with fallaway bottoms permit
restraining a weight at the top of the tube is study of water-entry phenomena in the absence
withdrawn magnetically and the weight falls, of gravity (Chapter 16) . One tank measures
imparting an impulse to the rod that overcomes 11-3/4 inches by 14 inches internal cross section
the slight magnetic restraint from the solenoid- and is 18-3/4 inches deep . Smooth internal walls
holding coil at the bottom of the plunger and and plate - glass windows have no internal
impels the ball downward . A velocity of 17 fps obstructions or projections to induce circulation
can be obtained at the interface . in the water . The bottom of the tank is held in
A second impulse launcher was designed to place with two clamps that can be released
launch spinning and nonspinning projectiles at simultaneously by powerful solenoids on the
higher velocities. In basic design , the launcher same energizing circuit. This vertical-drop tank
was similar to the first, but provision was made was mounted above the open launching tank
to rotate the impulse rod in an airtight bearing (Figure 17.43 ). The left part of Figure 17.44
297
 The other vertical-drop tank , measuring 7
by 7 inches internal cross-section and 20 inches
deep, was mounted in the liquid - interface tank
( Figure 17.45 ) for studies of water entry under
conditions of varying gas pressures and densities. .

Sectional Tank
This tank was constructed for study of
two -dimensional water-entry flow which would,
in theory , be produced by the oblique entry of
an infinitely wide flat plate. Two heavy sheets of
plate glass measuring 2 feet high by 3 feet long
are held accurately parallel and 2 inches apart by
spacers ( Figure 17.46) . A rectangular flat plate 2
inches wide is impelled obliquely into the water.
Continuous contact between the plate and the
tank walls prevents flow around the ends of the
plate and the effect , so far as flow is concerned ,
is similar to a sectional view of the flow induced
by water entry of a flat plate of infinite width
(Chapter 6, Figure 6.2 ) .
The plate is impelled into the water with a
rod driven by a pinion gear engaging a rack
mounted on the side of the rod. A heavy
Figure 17.43. Vertical drop tank. flywheel is brought up to speed and suddenly
coupled to the pinion gear by a quick-acting
electric clutch. The rod with the attached plate
is propelled forward at a relatively constant
shows a plate-cylinder projectile entering while velocity by the energy stored in the flywheel.
the water is held stationary, and the right part After about 1.5 feet of travel, the end of the
shows the same projectile entering while the rack runs off the pinion and the flywheel idles
water is freely falling. Studies indicated that the to a stop . The rod moves a short distance
acceleration of the falling water was very nearly forward under its own momentum.
gravitational, and circulation of the water due to The impeller may be adjusted to give
release or wall effect was negligible. water-entry angles from about 5 degrees to

Figure 17.44. Plate-cylinder projectile entering stationary water (left) ; freely falling water (right).

298
 vertical. Water-entry velocities of about 15 fps
have been attained .

Hydroballistics Range
A hydroballistics range (HBR) was estab
lished in the NUC laboratory to measure hydro
dynamic coefficients of fully wetted missiles
during free flight. The data-recording techniques
developed can be utilized in any facility where
fully wetted motion is to be observed, but the
technique of multiple-exposure photography is
not suitable for recording cavity motion.
For the deduction of usable hydrodynamic
coefficients by fitting an equation to the free
flight trajectory data, the missile position must
be determined to a high degree of accuracy
(References 80 and 81 ) . The simple but pre
cisely aligned and instrumented HBR ( Figure
17.47 ) was established in three bays of the open
launching tank in the NUC Hydroballistics
Laboratory. Here , horizontal and vertical dis
tances in the trajectory plane and angles of pitch
and roll can be measured (4 degrees of freedom) .
Distances can be measured to within +0.001
feet, the pitch angle to 12 minutes, and roll
Figure 17.45 . Vertical-drop tank in liquid interface tank. angle to within +4 degrees.

80
F

Figure 17.46. Sectional tank with plate-propulsion apparatus.

299
 Reference Reference

Launching plane

Launcher Backstop
t
Lamp Lamp Lamp Lamp

Camera Camera

Figure 17.47 . Hydroballistics range ( plan view).

The portion of the tank in which the HBR To estimate distortions caused by the HBR
was established is 4 by 4 by 18 feet. Three optical system , a precision stadia rod with 1 -inch
3 / 4 -inch - thick glass panes with clear spans of 3 graduations was photographed in several hori
by 5-1 /2 feet between the structural-steel sup zontal and vertical positions in the trajectory
porting ribs are mounted in the two sides. Panes plane, and the distances between images of the
can be adjusted with leveling screws to be graduations are measured with an optical com
vertical and coplanar to within less than 2 parator. In view of the variation in the hori
minutes of angle. The bulge in the glass caused zontal image distance, individual magnification
by hydrostatic pressure is such that the maxi factors were determined for each missile image.
mum deviation of the glass surface from its Since valid data can be obtained from the
reference plane is approximately 3 minutes of HBR only if the missile remains in the launching
angle. A dull black cloth hung over the glass wall plane, a backstop was hung normal to the
farthest from the cameras provides a suitable launching plane approximately 8 feet beyond
background for multiple-exposure photography. the launcher, with the intersection of the
The two precision cameras were positioned launching plane marked on it. The missile flew
as shown in Figure 17.47 ; 4-by-5 -inch ultraflat into the backstop and remained hanging there
glass photographic plates (Chapter 18) were until retrieved. Trajectories yawing more than
selected because film is too flexible for accurate 0.2 degree from the theoretical trajectory plane
positioning in the cameras. The cameras located were disregarded. A mirror mounted above the
the plates to within 0.001 inch of their focal water surface directly over the trajectory plane
planes and held the angular orientation of the was used to verify that the plan view of the
plates normal to their axes to within 1 minute of trajectory was a straight line, the position of the
angle. The plates themselves were flat to 0.0007 trajectory plane being established by the image
inch/linear inch. The cameras were rigidly of a white line on the bottom of the tank. The
mounted on surface plates equipped with mirror was included in the field of view of the
threaded position and leveling adjustments. cameras, allowing this information to be
Cross-hair telescope sights attached to the recorded simultaneously with the position of the
cameras were used for their accurate alignment missile. No roll in excess of 5 degrees was
with respect to precision markers on the tank . detected in any of the missile launchings.
Controlled development procedures minimized The HBR is illuminated by four Edgerton
migration of the exposed silver grains and type flash lamps that can be flashed at pre
bleeding of the images during processing of the selected frequencies and can be programmed to
photographic plate (Reference 49 and Chapter give several frequencies during a single launch
18 ) . ing, preventing overlap of images near the end of

300
the trajectory by decreasing the flash rate of the images of the white fiducial marks. The number
lights as the missile decelerates. The flash fre of marks should be kept to a minimum and they
quencies of these lamps are accurate to 0.01 % of should be so located that a minimum of inter
the set frequency. ference occurs among their traces. It is also
The superposition technique of recording advantageous to make each mark different in
photographic data requires that a flat-black shape or orientation so that each trace can be
missile bearing small, white fiducial marks be readily identified on the photographic records.
photographed against a flat -black background. Figures 17.48 and 17.49 show two different
The data then consist of traces left by successive types of photographic record from the HBR.

Traces of Fiducial
Marks

Figure 17.48. Photographic record of Basic Finner with fiducial marks as shown in Figure 14.1 , Chapter 14.

301
 IIII
IIII
HTTTTT +++ TI

302
Photographic
17.49
Figure
Basic
record
.of
inset
in
shown
as
marks
fiducial
with
Finner
 , mis

CHAPTER 18
EXPERIMENTAL TECHNIQUES

ENVIRONMENTAL FACTORS with ambient pressure. The tank water can be

Sharply defined underwater photography deaerated by maintaining reduced atmospheric
requires clear water. Ordinary tap water contains pressure within the tank while a stream of air is
suspended material that reduces light transmis bubbled through the water. The agitation of the
sivity and blurs the image if the light path is bubbling air removes excess air from the water
long . Furthermore, some light-absorbing and equilibrium is attained. This technique has
substances such as yellow -brown dyes resulting been found to ensure repeatability. Air content
from the accumulation of organic vegetable can be measured with a Van Slyke-Neill mano
matter in reservoirs are not removed by ordinary metric gas analysis apparatus ( Reference 62 ) .
filtration . Distilled water is preferable for under The maximum possible atmospheric
water photography, but cost prohibits its use in pressure within a facility is determined by the
large hydroballistics tanks . A technique employ strength of the pressure vessel, the lower pres
ing diatomaceous earth filters has been devel sure limit being set by the vapor pressure of the
oped to clarify water for long -distance photo liquid. The vapor pressure of lubricants or
graphy ( Reference 45 ) , and is now being used in sealing compounds used within the tank can also
the NOL Hydroballistics Facility add to the minimum pressure . The vapor pres
(Reference 75 ). sure of the liquid can be controlled to some
Both tap and distilled waters require almost degree by adjusting the liquid temperature ; e.g. ,
constant filtration and skimming to remove the vapor pressure of water varies from 9 to
dust; air and gases in contact with the water 760 torrs over the temperature range of 10 to
surface should also be filtered . Underwater 100°C, offering a possibility of adjusting water
equipment and the experimental tank itself vapor pressure for gas-density scaling in aа model
should be both noncorrosive and clean . Materials system.
that are to be left in contact beneath the water Froude number scaling requires that the
(steel and brass, for example) should be so model velocity scale according to Equation 1.18 ,
selected that electrolysis is not a problem. the addition of cavitation number scaling
Most tap and distilled waters carry an requires that the gas pressure scale according to
excess of dissolved or entrained air, which does Equation 1.23 , and the further addition of
not perceptibly affect the shape of a stable gas-density scaling requires that the gas density
cavity. If, however, the concentration of air scale according to Equation 1.24 . In the absence
exceeds its equilibrium value at the experimental of cavitation-number scaling, gas density is
air pressure , the cavity wall roughens. Excess air scaled simply by using air at the same pressure as
in the water influences an existing unstable in the model system, but the constraint of
cavity separation and alters the inception of cavitation -number scaling requires the use of
cavitation . In sufficient quantity , air can diffuse heavy gases for gas - density scaling. In this case ,
into the cavity and change the cavitation num metered quantities of one or several gases were
ber . Deaerated water was used in the NUC introduced into the tank after the air and water
experimental studies . vapor had been reduced to a predetermined
The amount of dissolved air in equilibrium amount. These quantities , calculated from Equa
with water at any given temperature decreases tions 2.12 , 2.13 , and 2.14 , were metered into

303
the tank with a manometer, making use of Canton , Mass.) might be more suitable . To
Equation 2.5. They were then mixed with a fan , achieve roughness in this manner, each grain is
and the tank atmosphere mixture brought to the placed individually in a preset pattern to insure a
desired launching density and pressure . constant model configuration, and the grain
Immediately before a test model was launched , a pattern is replaced or at least repaired after each
gas sample was drawn into a calibrated gas launching, which is a time-consuming procedure .
density balloon and the gas density determined Instead the grain pattern can be installed on the
gravimetrically . model and copper roughness caps electroplated
If a maximum or minimum density is from the impression ( Figure 18.1 ) . Duplicate
required, the air must be removed as completely caps can be installed on the model nose in a
as possible from the tank . The evacuation matter of minutes . Each cap lasts through
process is hastened and the degree of air removal several launchings .
possible with ordinary equipment is increased Turbulence can also be induced by
when the tank is “ rinsed” several times with the installing a wire ring around the missile nose at
working gas. the zone of separation , by cutting a groove in
The range of gas densities possible within a the model nose (as shown in Figure 11.1 ) , or by
tank can be extended by chilling the liquid to altering the model contour slightly to introduce
reduce its vapor pressure. For example , the a discontinuity in the curvature of the model
decrease in vapor pressure from 17.52 to surface where separation should occur. Valuable
9.2 torrs occurring when the water temperature test data have resulted from the use of these and
is reduced from 20 to 100C increases the other techniques to induce turbulence artifi
maximum gas-density coefficient obtainable at cially , but there is always the possibility that the
1/11 atmosphere with Freon 114B2 (dibromote system may be disrupted to the point where
trafluoroethane ) from 0.69 to 0.79 erroneous data occur.
(Reference 67 ). Turbulence can obviously be induced by
It is sometimes desirable or necessary to increase of the Reynolds number. If the velocity
operate models at velocities that do not cause within the model system is fixed , the Reynolds
turbulent flow in the boundary layer, thereby number can be increased only by changing the
violating the requirement (Chapter 1 ) that both liquid properties. Survey of the literature indi
model and prototype velocities be above the cates that no liquid exists that can practicably
critical Reynolds number for turbulent flow to be substituted for water in the model system ,
obtain and for viscosity effects to be neglected. but it is possible that in the future, aqueous
Turbulence can be induced artificially either by polymer solutions can be used advantageously.
devices attached to the model or by slight
modifications of the model shape. Devices used
to induce turbulence either increase the surface
roughness ahead of the normal transition zone
or introduce a discontinuity in the missile shape
where boundary -layer transition would normally
occur .
There are a number of techniques for
calculating boundary -layer parameters and the
critical height or critical roughness which causes
transition in the boundary layer, as , for
example, Equation 8.1 . Roughness can be
increased by attaching grains of graded sand to
the surface of the missile (as shown in Figure 8.3
of Chapter 8 ) but when the missile is forcibly
halted by collision with a backstop, the grains of
sand often dislodge and gouge the missile
surface . A softer substitute such as poppy seeds ,
hollow plastic Ecco spheres, or microballoons Figure 18.1 . Electroplated copper roughness cap
(obtainable from Emerson and Cuming, Inc. , on CIT model nose.

304
Their action is imperfectly understood at the
present time .
The density of water is almost constant
from 0 to 100°C , but the viscosity varies from
1.792 to 0.284 centipoises . Heating water offers
some increase in Reynolds number and under
some circumstances may be adequate to shift
the flow from the transitional regime to full
turbulence. At the higher temperatures , water
vapor pressure becomes increasingly important
in determining the cavitation number and gas
density.

PREPARATION OF MODELS Figure 18.2. NUC model CG balance.

Models should be accurately scaled to the
prototype , and if several models of the same
shape are used in a single test program , they
should be as nearly alike as possible ; for 2-inch
diameter models a tolerance of +0.001 inch on
external contours is desirable. Models made
from the same templates can be within a few
ten thousandths of an inch of each other in size .
Model design requirements, besides specify
ing the external contour and satisfying the
structural demands imposed by launcher acceler
ation and dynamic forces, should also include
the scaled dynamic properties of the prototype .
The CG and moments of inertia of all model
components should be calculated or estimated as
the design progresses because only minor adjust
ments can be made with the trim weights.
Designing a small model to be both strong
enough and of specified dynamic parameters can Figure 18.3 . CIT model CG balance.
be difficult because weight scales down as the
cube of the linear dimension and moment of
inertia as the fifth power (Equations 1.11 and that could be combined in many ways . Threaded
1.12 ) . internal weights allowed adjustment of total
Model weight can be determined with an weight , CG location, and moment of inertia. The
ordinary precision balance , but special proce adjustable internal arrangement of the models
dures are needed for determining CG and could be adapted to the different structural
moment of inertia. Two combination CG requirements imposed by the launchers in the
balance fixtures were used (Figures 18.2 and two laboratories.
18.3 ) . To determine the moment of inertia, the The NUC model's slip -fit joints were held
NUC system ( Figure 18.4 ) compared the angular together by an internal tie rod threaded along its
acceleration of the missile about a transverse length ( Figure 18.6). The trim weights on the tie
axis through the CG upon application of a rod permit large adjustments in model trim.
known torque to the angular acceleration of a Model materials included stainless steel, Dural,
calibration rod. In the CIT system, models and Lucite , and nylon.
calibration rods were clamped at the CG in a The gripping pressure of the chuck in the
torsional balance ( Figure 18.5 ) and oscillation CIT launcher required a stronger model (Fig
times were measured. ure 18.7 ) made with aluminum alloy nose and
Both NUC and CIT torpedo models were tail sections screwed into a heavy stainless-steel
made of hollow Dural or stainless-steel sections center section. The cycling acceleration force
305
With calibration rod With model

Figure 18.4 . NUC rotational inertia balance.

With calibration rod With model

Figure 18.5 . CIT torsional inertia balance .

306
 holes in the nose and tail. Trimming was more
difficult than it was with the NUC models
because less weight remained to adjust the trim
after the structural demands had been met, and
the tapped holes allowed less choice of weight
position than the tie rod did. A high moment of
inertia was also difficult to achieve with so much
of the weight concentrated near the CG.
Models cannot be permanently sealed
against water leakage because the trim weights
must be accessible for alteration and because it
is sometimes expedient to combine different
head and afterbody sections. The slip joints in
Figure 18.6. NUC model. the NUC models were sealed with dental wax
and had to be resealed several times during a
series of launchings. The CIT models were sealed
with special buttress threads coated with
Permatex; one section was heated and the other
chilled before the models were assembled . They
rarely leaked but were so tight they were
difficult to disassemble. The models were
weighed before and after each launching to
ascertain that no leakage had occurred .
The model's tail-fin assembly poses special
leakage problems. If the entire fin structure is
made of stainless steel , the joints can be sealed
with silver solder (top of Figure 18.8 ) , but these
joints tend to crack during usage. If the shell is
Figure 18.7 . CIT model. thick enough, the fins can be inserted into
external slots (bottom of Figure 18.8 ) , but this
method is not always feasible in the light of
produced by rotation of the launching wheel structural and weight requirements.
precluded the use of Lucite or nylon nose or tail The condition of the model surface is
sections. Trim weights were screwed into tapped known to affect cavity separation (Chapters 3 ,

t 1
Removable steel tail with silver -soldered joints Tail with embedded fins

Figure 18.8 . CIT model tails.

307
4, and 16). It was difficult to explain this normally clean hands during the routine pre
surface effect ( References 43 and 112 ) because launch procedure that includes model trim and
model surfaces were not sufficiently reproduc loading the model into the launcher. Oil from
ible. Tests in the NUC laboratory revealed that the skin provided a mild and uniform contami
different cavities could form during apparently nant on the model surface.
duplicate launchings of a polished steel sphere Merely washing the model in aa container of
that was ( 1 ) normally handled, (2 ) rinsed in solvent is not adequate because particles of
ether, or (3 ) greased. grease will be redeposited during withdrawal.
Washing the model in a solvent such as Rinsing minimizes this condition but manual
ether and then wiping it with a paper tissue does handling is still necessary to distribute the
not guarantee a clean surface. Stainless steel remaining isolated grease formations and insure
models, if chemically inert, can be further that a light film of grease covers the entire
cleaned by dipping them into a hot saturated surface .
solution of potassium dichromate in concen Some kind of visible marks should be
trated sulfuric acid and then rinsing them in applied to the models for contrast and clarity in
distilled water before they are dried . Hot water photography, especially when light is limited,
directly from the still is preferable. This tech when photographic images are small, or when
nique is not practical for use with chemically metric data are required. 2 A white model with
reactive metals or plastics. black marks provides the greatest contrast
The models should not be handled after against a black background.
being cleaned, and the water surface and the Black anodizing provides an excellent,
portions of the launcher that come in contact inexpensive surface finish for aluminum models.
with the model should also be cleaned. For Models of other materials can be given a
assurance of a completely grease -free water 0.001 -inch coat of lacquer or painted with
surface , the water must be allowed to overflow . enamel and baked. Because a glossy surface may
A reasonably hydrophobic and durable produce highlights that obscure fiducial marks,
coating was applied to stainless-steel spheres in the CIT models were lightly blasted with
the NUC laboratory by placing clean spheres 200-mesh sand before being coated.
(i.e. , solvent-washed) in polymethylsiloxane , Fiducial marks should be flush with the }

DC-500, 70 centistokes , for 2 days at 150 to model surface. At CIT, after the models were
160°C. The resulting coating caused a yellowish painted, masking tape was applied to the entire
discoloration but its thickness was less than the surface except for the fiducial marks and then
wavelength of visible light. The spheres were the models were sandblasted . The indentations
subsequently handled with paper tissues and made by sandblasting were filled with a contrast
cleaned with ether before being launched. ing color of paint and rubbed until smooth and
Such procedures are appropriate in study flush with the model surface (Figure 18.9 ) .
ing the effects of surface condition , but in Another method involves applying the fiducial
normal field work and in laboratory work, they marks over the paint and rubbing until the edges
are undesirable because completely hydrophobic fair into the surface. For accurate interpretation
or hydrophilic surfaces do not occur. A repro of the photographic data, the width of the
ducible, uniform surface that is neither hydro fiducial marks should be constant and the edges,
phobic nor hydrophilic is better for routine sharply defined .
model testing. Before each launching in the NUC Adequate fiducial marks were easy to
laboratory , the model was dipped in ether and achieve in the small Variable -Angle Variable
carefully rinsed by pouring etherl over it, Pressure facility at NUC . In the large tank at CIT
always wiped with the same brand of paper (Chapter 17 ) , however, the photographic image
tissue, and then handled for several minutes with was very small and the light was limited.

lEthyl and petroleum ether are highly flammable. The use of dichloromethane, a nonflammable solvent, is
recommended .

2Whip measurements made with the optical whip recorder (Chapter 17) depend upon images of one or more small
mirrors mounted on the missile tail instead of fiducial marks painted on the model .

308
 The brilliance of the fiducial marks is
important when several images are recorded on
the same film or plate ; the image must be
readable against the repeatedly exposed , yet not
overexposed, background.
Fiducial marks could not be used during
one series of water -exit tests in the Variable
Atmosphere Wave Tank at CIT ( Chapter 19 and
Reference 93 ). The motion of highly polished
1-inch-diameter precision -ground steel spheres
was measured over distances of a few inches and,
at the same time , photographic data on
associated hydrodynamic phenomena such as
water -surface deformation were obtained. Any
Figure 18.9 . CIT model with inset fiducial marks. rotation of the sphere would have introduced
error into the data obtained from a fixed
fiducial mark on the sphere surface . Sphere
positions measured from photographs by setting
Initially , fiducial marks were hung along the the cross hairs of a comparator (Chapter 17 ) on
center line of the tank between the cameras and
the curved edge of a sphere image lacked the
the launching plane, but they not only proved required degree of precision. A technique
time- consuming to maintain in proper position described in References 90 and 94, using the
but were a hazard to the models. virtual image or highlight produced in a sphere
New fiducial marks, consisting of Edgerton by a stroboscopic point-light source, obtained
type flash lamps bent into right-angle con accurate sphere position-time measurements .
figurations, were housed in watertight boxes and By suitable placement of the light and
bolted to the tank walls opposite the CIT camera with respect to the trajectory of the
cameras. Threaded adjusting screws allowed sphere, the projections of the highlights on the
these markers to be placed on the cameras' object plane can be made to coincide with the
projected optical axes. Power was supplied from positions of the sphere center except for a slight
the same units that operate the regular flash lateral offset perpendicular to the motion of the
lamps and, therefore, the markers flashed in sphere . Moreover, the rotation of a precision
unison with the tank lights. The marker images ground sphere does not measurably affect the
were so bright that several could be operated by position of the highlight.
aa single pulsing unit. The five underwater trajec Highlight photographs permit precise
tory cameras were spaced close enough (with centering of the cross hairs of a measuring
60% overlap at the launching plane) that three microscope, and thus precise measurement of
fiducial marks were visible in the fields of all highlight (i.e. , sphere) positions. The intensity of
except the two end cameras. As the metric the highlight permits many exposures to be
analysis became more refined, fiducial marks made on a single stationary film . Displacements
were added to the top and bottom of each of the sphere can be measured directly , and data
photographic frame for accurate positioning reduction involves fewer measurements than
during projection. when each exposure is recorded on a separate
At least one fiducial mark should always be frame.
visible to each data-recording camera and, if Use of the highlight technique in air is
possible, the vertical should be marked with a illustrated in Figure 18.10 . A camera and a
plumb bob. In a multiple-camera system, at least stroboscopic point-light source , L, were so posi
one of the fiducial marks seen by each camera tioned that the trajectory of the sphere center,
should also be seen by one other camera. If C, lay in the object plane of the camera; points
projected images are to be measured , several along the trajectory were equidistant from the
fiducial marks should be located near the light source and camera lens node , N. For a ray
extremes of the field of view so that the film can of light from the point-light source to be
later be aligned as it was during exposure . reflected from the sphere (assumed to be moving

309
 Commercial Ektar f /6.3, 8-1 /2-inch focal-length
Steel
sphere
lens and a Polaroid Land camera back were used
L2 C
with Polaroid 3000 Speed/Type 47 film to
obtain highlight data.
Li's
A Kodak Wratten No. 25 (A) red -gelatin
film filter cemented between circles of optical
glass mounted in front of the camera lens
C = center of sphere reduced the light wavelength range (and hence
L = light source light dispersion due to tank-wall prismatic
S = point of reflection
N = lens node effects), resulting in sharper highlight images.
Camera
L1 = high light Both the camera lens node and the lamp
L2 = projection of highlight N'
were 8 inches from the Lucite tank wall . The
on object plane
i = angle of incidence
trajectory of the 1 -inch-diameter sphere was
r = angle of reflection Lamp vertical , parallel to and 9 inches from the tank
i= r wall. The magnification factor for highlight
photographs was about 1.1 and the camera lens
Figure 18.10. Optical setup for highlight method
stop between f /32 and f / 45 .
of photography. ( From Reference 94. ) To focus the camera, a sphere was posi
tioned so that its center was in the object plane
at some point in the trajectory. A zirconium
concentrated arc lamp (Sylvania C2/DC/S ) was
vertically ) into the camera lens, it was necessary placed just in front of and in line with the
that the incident and reflected rays lie in the stroboscopic lamp and the sphere. The room was
plane LCN. Then the point of reflection , S , then darkened and the camera, with lens fully
would lie in LCN, as would the virtual image open, was focused on the highlight. The range
(highlight), L1 , which was midway between the and number of sphere positions to be observed
surface and the center of the sphere . The over the exposure sequence was adjusted by
projection of Li on the object plane is L2 and setting the frequency and by gating the number
would be at the same height as C , displaced of flashes. For data acquisition , the room was
slightly to one side of it.
darkened and a time-delay generator, adjustable
The highlight accurately represents the
sphere center and can be used for vertical
distance, velocity , and acceleration measure
ments . It can also be used to measure the EGG FX- 2 strobe lamp
with parabolic reflector
sphere's lateral deviation in the object plane , if
the deviation is small ; the relation of the Kodak Wratten Tracing paper for
+
projection of the highlight to the center of the No.58 filter light diffusion
sphere does not change significantly . Precision
ground spheres up to a diameter of 1-5/ 16 inch Sphere
with a sphericity tolerance of 0.000025 inch are
available commercially.
Kodak Wratten
The highlight technique for acquisition of No. 25( A) filter Lucite tank
data is also applicable to a sphere immersed in
water. Here, in addition to the optical setup PEK X- 81 xenon short
shown in Figure 18.10, it is necessary that the arc strobe lamp in
metal box
sphere trajectory be parallel to the plane surface
of the water-tank window and that the point Rotating-disk
light source and lens node be equidistant from camera
the window ( Figure 18.11 ) . For studies at CIT, a Graphic view
PEK X-81 xenon short-arc lamp of about 1 -js camera
flash duration (PEK Labs, Inc. , Sunnyvale ,
California) was used as the point-light source . A Figure 18.11 . Optical setup for rotating disk and highlight
Graphic View II camera fitted with a Kodak methods of photography. ( From Reference 94. )

310
to lus, triggered the flash sequence at the proper photographs of the water surface. A Kodak
instant. Calibration of the highlight system Wratten No. 58 green filter in front of the FX-2
consisted of photographing, without change in lamp and the Kodak Wratten No. 25 (A ) red
focus adjustment, a scale in the object plane. filter in front of the highlight camera lens
Figure 18.12 shows a typical highlight film prevented fogging of the highlight film by light
with 55 highlight images as indicated by an from the FX-2 lamp. The PEK X-81 lamp light
electronic counter. The images are clear and did not adversely affect rotating-disk camera
sharp , except those obtained during the initial photography and no additional light filtering
motion as the sphere slowly started from rest . was needed. Both stroboscopic lamps were
As many as 100 exposures have been taken pulsed simultaneously to obtain synchronous
without noticeable loss of contrast with the data.
background. The images are round and the cross A series of rotating-disk camera photo
hairs of a measuring microscope can be centered
a graphs using Kodak Royal Pan film is shown in
on them with a precision of +5 microns for good Figure 18.13 . The image of the X-81 lamp can
highlights and +2 microns for very good be seen as a highlight in the sphere .
highlights.
A combination of highlight and rotating PHOTOGRAPHY
disk camera techniques was used to obtain In metric photography, the densest image
synchronous photographs of the sphere position of the greatest contrast obtained with the fastest
and perturbation of the water surface as the film may not always be the most easily
sphere moved toward it. The rotating-disk measured. Under a microscope, a grainy image is
camera was so positioned that its lens axis was too amorphous to position accurately . The edge
normal to the water-tank window and lay in the of the image , which should be sharp and clear,
plane of the undisturbed water surface ( Fig. resembles the broken edge of a macadam high
ure 18.13 ) . An E G and G ( Edgerton , way. A dense image tends to bleed into the
Germeshausen , and Grier) FX- 2 stroboscopic surrounding background, thus adding to the
lamp of about 1-4s flash duration provided irresolution . A slower, fine-grained emulsion will
diffused background illumination for silhouette produce a thinner image that may appear less

Figure 18.12. Highlight film of sphere moving upward , 1 ms

between exposures. ( From Reference 94.)

311
O
Figure 18.13 . Rotating disk camera photograph, 1 ms between
exposures. ( From Reference 94.)
desirable to the naked eye but will be far more Worthington and Cole used a network of minute
readable under the microscope . bubbles generated by electrolysis. So long as the
Manufacturers' film -speed ratings do not water was still, the bubbles rose uniformly from
apply to film exposed with Edgerton-type flash the bottom of the tank. When the liquid was in
lamps; the reciprocity relationship that deter- motion, the bubbles were carried with it , the
mines film speed breaks down for extremely bubble velocity being equal to the sum of the
short, high -intensity exposures ( Reference 49 ) . water velocity and the velocity induced by
Individual emulsions should be tested in the buoyancy. Slichter added the elements of time
actual model system . and multiple exposure to the bubble-tracer
Special emulsions sensitive to narrow light- technique by photographing his experiments
frequency bands have been developed for with Edgerton flash lamps ( Chapter 4 ) , and in
astronomers' use. They are available coated on 1947 , Birkhoff and Caywood took high -speed
" ultra-flat” glass plates (i.e. , flat to 0.0007 inch Edgerton photographs of missiles traveling in a
per linear inch ) or even on optically flat glass metric-bubble field ( Chapter 4 ) .
plates for precise photographic measurements Care is required in the generation of a
( Reference 34). metric-bubble field . Ideally, the bubbles, tiny
Precision photography is dependent upon and of uniform size and spacing, lie within the
controlled developing. Minimum developing launching plane and have a specific gravity of
time , a fine- grain developer, and controlled just less than one (about 0.85 ) so that they rise
temperature and agitation are necessary , even slowly to the water surface. The surface tension
the position of the plate in the developer can be of the fluid comprising the bubble should be less
significant. Faulty developing can cause the than that of water, its index of refraction
exposed silver halide grains to migrate within the greater, and its compressibility approximately
gelatinous base, or unexposed grains adjacent to equal. Benzene (sp. gr. 0.69 ) can be mixed with
a heavily exposed area can be erroneously carbon tetrachloride (sp. gr. 1.60) to a density
developed, introducing error into the photo- giving the desired bubble velocity. The major
graphic measurement. problem lies in generating equally spaced
Color film enhances the value of cavity bubbles of equal size.
photographs, seeming almost to add an extra Birkhoff and Caywood worked with bubble
dimension to the experimental data. networks of both air and hexane. The air -bubble
Shadowgraphs have been successful for network arose from a punctured gum rubber
recording entry splash (Reference
( 112 and tube lying along the intersection of the launch
Chapter 12 ) . A tank , filled with water to the top ing plane with the tank bottom. The tiny holes
so that its surface formed a negative meniscus in the tube were of uniform size , and the rate of
with the sides, stood between the camera and the bubble emission was controlled by applying
spark light source. The spark and the camera tension to the tube . The hexane was introduced
were positioned so that the spark lay on the into the water through a line of hypodermic
extended optical axis of the camera, which needles. The rate of bubble generation was
crossed the water in the plane of the water determined by the pressure applied to the needle
surface. The moving missile was usually used to plungers ; uniformity was achieved by driving
trigger an adjustable, preset time delay in the them with a common fluid pressure source.
spark circuit. Flows have also been marked by soluble
dyes introduced into the flow by the model. The
FLOW - MARKING TECHNIQUES dyes can be coated on the model surface,
In 1893 , a French experimenter, Marey , implanted in grooves, or ejected from within the
used silvered wax -rosin spheres whose specific model through holes in the skin (Chapters 11 ,
gravity approximated that of water, to mark the 13, and 14). Dyes are particularly helpful in
flow of a liquid ( Reference 42) , in 1904 Prandtl studying boundary -layer flow but their use is
used flakes of mica ( Reference 66) , and in 1900 somewhat limited because the very presence of
Worthington and Cole (Reference 113 ) pub- the dye influences the flow to some unknown
lished a paper describing a bubble-marking tech- degree; the error is minimized when the specific
nique that rendered flow in water more visible. gravity of the marking fluid equals that of water.

312
 Flows on clean surfaces can be marked circular area in the center. If the powder
with powdered aluminum , as in the classic photo remained spread after being blown , the surface
graphs of Prandtl and Tietjens (Reference 66) , was sufficiently clean ; if the cleared circle
with talc, lycopodium powder, etc. Prandtl and closed, the surface was contaminated .
Tietjens tested the surface condition by blowing A thin coating of paraffin on stationary
directly down on a little powder that had been models that are partially submerged will prevent
sprinkled on the water surface. The powder capillary action from influencing the flow near
spread out in all directions, leaving a cleared the fluid boundary .

313
CHAPTER 19
HYDROBALLISTICS FACILITIES DESIGN

Many hydroballistics model facilities were the effect of waves, the wave channel must be
designed during World War II to model the rectangular in section. The testing area should be
water -entry behavior of air-launched torpedoes at least twice (preferably three times) the water
that entered the water at relatively low velocities depth away from the wave generator in order
and of mines that were dumped from moving that free or gravity waves can be achieved. The
ships or propelled from deck launchers. These wave channel should also include an artificial
early facilities were open to ambient air, but beach to absorb wave energy and thereby
theory and experiment indicated the need to prevent wave reflection from the channel end.
scale (control ) the gas pressure and possibly the The design of wave tanks, wave generators , and
gas density over the water surface in order to wave absorbers is thoroughly discussed in
model missile water-entry behavior. References 71 and 109 .
Postwar studies with pressure-vessel model The stringent requirement that the
ing facilities confirmed the need to scale gas launcher introduces no perturbations in the
pressure and density and also demonstrated that model system becomes increasingly difficult to
models larger than 2 inches in diameter (as used meet as the energy imparted to the model
at NUC and CIT) would be required in order to becomes greater. The mass of the model
minimize the effect of unscaled forces (viscosity, increases as the scale factor 13 , and in a
surface tension , etc. ) , and to provide adequate Froude -scaled system the velocity of the model
internal space for instrumentation and mech increases as 21/2 . Thus the needed energy
anisms for missile control , propulsion , etc. increases as 24 , and absorption of the energy
Interest in water-to-air missiles and the remaining in the launcher without disturbing the
effect of waves on both missile water-entry and modeling system becomes more difficult as the
water -exit behavior required facilities with model size increases ; the launcher itself may
underwater launching and wave-making capa become prohibitively large.
bilities. To meet the need for larger and more For some applications, a self-powered
versatile modeling facilities, the Variable missile may be more feasible than a momentum
Atmosphere Tank (VAT) at NUC, the Lockheed propelled missile . The small, high -speed turbine
Underwater Missile Facility ( LUMF ), and the engines and miniature storage batteries now
NOL Hydroballistics Facility were constructed. available make such an approach more attrac
Reference 45 provides an extensive descrip tive . For studies at supersonic velocities, smaller
tion of research into the problems involved in models are satisfactory because the inertial
the design of a large modeling facility . A basic forces are great enough to obscure the gravita
design requirement is that the facility be able to tional force, and the modeling thus becomes
be evacuated to the vapor pressure of water so independent of Froude scaling.
that the cavitation -number and gas-density Instrumentation of the model is becoming
scaling requirements can be met. It should also more important. Telemetry has been used with a
be able to withstand the positive pressure model 1.25 inches in diameter and 7.4 inches
incurred in the simulation of deep submergence . long ( Reference 57 ) to transmit data on water
If the modeling facility is expected to investigate entry accelerations. It appears feasible , if larger
314
models are used , to include apparatus that will facilities becomes a critical factor in photogra
transmit data on missile accelerations and phy; for a distance of 10 meters from light
attitude and on both cavity and missile pres source to missile to camera, the light intensity at
sures. Such data are difficult to obtain photo the camera in terms of that of the light source
graphically; in fact, photographic determination for various transmissivities is as follows :
of the missile attitude in completely enveloping
cavitation and missile pressures cannot at Percent light Light
present be accomplished, and estimation of transmissivity intensity
cavity pressure is questionable. Data verification per meter at camera
through comparison of missile velocities deter
mined photographically ( by differentiating 95 0.60
missile displacement with respect to time ) and 90 0.35
telemetry (by integrating missile accelerations 85 0.20
with respect to time ) should be possible . The 80 0.11
telemeter, although increasingly useful, cannot
entirely replace the camera to record the time It is evident that the light intensity at the
history of cavity formation and behavior. camera decreases rapidly with a slight decrease
The factors important to success in under in transmissivity. For light in the visible region
water photography are adequate light, low light of the spectrum, a transmissivity of 95 % /meter
attenuation, an efficient optical system, and corresponds to that of distilled water and
fast film response . Edgerton-type flash lamps 90 % /meter corresponds to that of very clear
(Chapters 17 and 18 ) can put out 1 joule /flash water. A transmissivity of 93%/meter can be
at repetition rates ranging from 1 to more than achieved by filtration through diatomaceous
5000 per second. They can be combined in earth filters (Chapters 17 and 18 and Reference
banks to illuminate wide areas , but their phys 45 ). An extensive transmission bibliography is
ical size limits the amount of light that can be given in Reference 18 .
made available at any given point. The laser Water contamination results from corrosion
now offers unlimited possibilities for superior of the tank and submerged equipment and from
illumination (Reference 19 ) . biological growths. The specific details of the
The attenuation of light is determined by corrosion problem depend upon whether distil
the length of the light path through the water led, tap, or seawater is used , the type of filter
and by the water clarity. As a beam of light system, and the composition of the tank and
passes through water, its forward intensity other surfaces in contact with the water. Sea
decreases according to Lambert's law water, because of its abundance of dissolved
( Reference 17 ) salts, is considerably more corrosive than distil
led or tap water. Materials are ordinarily selected
I = 1, e-kx
e ( 19.1 )
to prevent electrolytic action (i.e., corrosion ),
0 but electrolysis can sometimes be used to
advantage: magnesium plates are installed in the
where I is the light intensity at any position x NUC tanks (tap water filtered through diatoma
measured along the direction of the beam, k is ceous earth ) to protect materials of lower
the absorption coefficient, and Io is the inten electromotive force (e.g. , iron , steel) from elec
sity at x = 0. A measure of the transparency of trolytic action. Experience at the CIT tank with
the medium may be defined by assuming a unit distilled water filtered through aluminum
distance (i.e. , setting x = 1 ) . Let li be the hydroxide precipitated on graded sand showed
intensity at this distance. Then the followingresistance:
materials to18-8
be stainless
of satisfactory
corrosion steel

11/10 = e - k ( 19.2 ) ( Types 302 and 304 ) , brass (some nautical

brasses and Tobin bronze are suitable ), Monel,
nylon, Tygon, Tygon paint, polyvinyl chloride
The ratio 11/10 is the transmissivity of the water plastics, Lucite, and fiber glass. The following
and may be expressed in terms of a foot, a yard, materials were not satisfactory for use in the
or a meter. Water clarity in larger modeling tank : 18-8 stainless steel (Type 303 or 416 ) ;

315
galvanized steel; aluminum with or without unnecessary lighting of the tank to deter algal
standard anodizing; various wax and tar coat growth is recommended (Reference 45 ).
ings; cadmium ; copper; copper nickel ; and cop All exposed windows and ports should be
per nickel chromium plates . Corrosion also closed when the tank is not in use , and the use
affects water turbidity and hence light transmis of ultraviolet lamps to control biological
sivity; the corrosion effect of various materials is growths above the water should be considered.
discussed in Reference 45 . Good circulation was attained in the CIT tank
As a launching tank construction material, by locating the tank outlet in a standpipe
stainless steel has many advantages over ordinary running the entire depth of the tank. Ports were
steel, which requires down time and expense to spaced about a foot apart and sized to be
maintain linings, paints, and coatings. An epoxy inversely proportional to the ambient static
coating provides noncorrosive, biologically inert, head.
pinhole-free tank lining that will adhere to a Filters on all air inlets should be used to
steel vessel through repeated pressure and tem screen out airborne contaminants. Filters are
perature cycles , and is less expensive than particularly necessary in the inlet line that is
stainless steel. All external plumbing should be used to return the tank to normal atmospheric
as corrosion - free as the tank itself. pressure after evacuation .
Before installation , a proposed tank lining The moisture that condenses on instru
should be tested by subjecting it to the prevail ments mounted within enclosed launching tanks
ing conditions : distilled , tap, or seawater, plus when tank pressure is reduced can be avoided by
all possible additives such as chlorine, rust heating the instruments slightly above ambient
inhibitors, etc. The water should have been temperature.
circulated through the filter because some filters In designing an optical system for aa launch
introduce additional ions. The condition of the ing tank, the number of cameras should be
water should be tested photographically in the carefully weighed against the resultant image
actual test setup using the camera, film emul size . Too few cameras can be false economy ; the
sion, light source, and exposure time involved. If data's worth is measured by the resolution of
the lining material meets the photographic the photographic image. The size of the photo
requirements, it should be subjected to pressure graphic image should be as large as practicable.
and temperature cycling representative of the Provision should be made for both metric and
various scaling conditions to be used. view photography with both direct and silhou
Model-tank water must be continuously ette lighting.
filtered. Generous filter capacity should be The following are possible metric photog
provided; it is difficult to maintain a filter raphy techniques :
performance at the efficiency level assumed by 1. A metric camera installation similar to
manufacturers. The filter system should be
designed to prevent accidental dumping of the that developed for the CIT tank
filter bed or cake into the tank when power (Chapter 17) to record high -speed
failures or some other untoward conditions stereoscopic motion pictures of any
occur. An efficient skimmer should also be action in the launching tank. It can
included . measure six degrees of freedom and
The water- circulation system should pre can also record the details of the
vent, as far as is reasonably possible , stagnant hydrodynamic flow pattern as delin
regions from forming in the tank and should eated by wake, cavity, and splash.
provide a constant flow of water across viewing ( The use of 70-mm film should be
windows, camera ports, and other submerged considered. )
points struck by external light, where biological 2. An installation of precision -plate cam
growths are likely to occur. Chlorination of the eras similar to that developed for the
water or the addition of copper sulfate as hydroballistics range at NUC (Chap
algicides have been found to decrease the light ter 17 ) to record position data only. It
transmissivity markedly; therefore, avoidance of can easily measure four degrees of

316
 freedom and, if a device such as an also permit relatively inexpensive , rapid surveys
aerial-mapping machinel is used to of many areas of a test program , materially
measure the photographic plates, six reducing the number of expensive tests in the
degrees of freedom should be larger tank. Much useful information
measurable. obtained in preliminary research studies con
Both systems give time-position data of ducted in pilot tanks (Reference 45 ) which were
sufficient accuracy that excellent first dif built prior to the construction of the NOL
ferences are possible but second differences give Hydroballistics Facility (Reference 75 ) .
only a reasonable estimate of acceleration . Precision instruments and new materials
Several large plate -glass windows would be developed in other fields should be used when
useful to allow pictorial records to be made of applicable. All new techniques and devices
should be examined to make certain that loss of
events within the tank (e.g. , cavity fluctuations,
flow separation ) . The rotating-disk camera used data accuracy is not the price of greater speed or
in the NUC Hydroballistics Laboratory (Chap ease of operation.
ter 17 ) is especially valuable in cavity studies.
Plate cameras of several sizes, and high-speed , RECENT DEVELOPMENTS IN TANK DESIGN
portable motion-picture cameras should also be
provided (Chapter 17 and Reference 76) . The NUC Variable-Atmosphere Tank
Selection of photographic film is The Variable-Atmosphere Tank (VAT) at
influenced by the wave length and intensity of NUC, a steel pressure vessel designed for water
the light source , and by the duration of the light exit studies ( Figure 19.1 ) , is a cylinder 12 feet in
flash . The reciprocity relationship between diameter and 39 feet high. The bottom of the
exposure and image density provided by film cylinder rests at ground level and a steel skirt
manufacturers and determined for general extends 10 feet below to support the tank and
photographic work is not necessarily valid for to form a subterranean housing for a hydraulic
the short, high -intensity Edgerton flash . With lift and other machinery. At present, the VAT is
limited illumination , position data of greater instrumented to observe the water exit of
accuracy can be measured from the thin fine relatively large ( 11 -inch diameter) models
grained image of slower emulsions than from the ( Reference 77 ) .
darker, larger-grained image formed by faster The pressure within the tank can be varied
films ( Chapter 18 ). from approximately 0 to 1 atmosphere absolute.
The Polaroid process can be used to pro The vacuum pump can reduce the tank pressure
vide immediate verification of system operation to the vapor pressure of water in about
and to measure quantities important in establish 15 minutes .
ing the conditions for the subsequent launching. The water system, designed to maintain
In designing a new facility, existing ones clarity at all times, includes a noncorrosive lining
should be studied for valuable features that in the tank, continuous filtration of all water,
should be retained . The aim should be simple , and surface skimming. Algae growths are con
adaptable equipment with mechanized data trolled by window -washing devices and ultra
systems. Whenever possible, the tank should be violet lamps at the water surface .
bolted instead of welded together to facilitate A vertical row of five observation windows
future modification . measuring 28 inches wide and 76 inches high
A large facility could profitably be built gives complete coverage of the tank interior.
first in pilot size . Many design , instrumentation , There are 45 observation and illumination ports
and operation problems could be met and in vertical rows , 15 ports in each row. One row
solved, the operating personnel could be trained of ports is directly behind the observation
while the large facility is under construction , windows; the other two rows are 45 degrees to
and the small facility would remain a valuable each side (90 degrees to each other) , to accom
tool in its own right. It would allow close modate an orthogonal stereoscopic camera sys
observation of detailed phenomena and would tem for gathering metric data. The observation

1Wild-Heerbrugg Instruments, Inc. , 465 Smith St., Farmingdale, Long Island, N.Y.
317
 Observa
tion and
and outside the tank for handling such large
lighting loads as missiles and launchers ( Figure 19.3 ) .
ports When the tank is in operation, the monorail is
withdrawn and the door is closed . A flat circular
platform moved by a telescoping hydraulic lift
transports the launcher, launching equipment,
and missile between the access door and the
tank bottom, where it remains to support the
launcher during operation . The maximum depth
Funnel of water above the surface of the lowered
shaped net
platform is 24-1 /2 feet.
A cushion in the top of the tank stops
missiles and a funnel-shaped net hanging from
the padding and extending into the water has a
Observation catcher net at the bottom that is programmed to
window
close after an upward -traveling model has
passed. The net is fastened to the top of the
tank when it is not needed .
Underwater launchers designed to launch
Catcher net missiles 11 inches in diameter have been used in
the VAT. An entire launcher assembly can be
Underwater
launcher mounted on rails and driven across the platform
tank by a hydraulic system to simulate cross
flow . Lateral velocities up to 1 fps have been
used. In addition , the launcher can be tilted to
simulate the rolled condition of a submarine .
Murah

The internal instrumentation for models

used in the VAT varies with the objective of
do.

each test. Various equipment for model instru

mentation , such as pressure transducers,
accelerometers, and oscillographs, are available.
However, photographic coverage of a test in the
a
n
VAT is the primary source of data. The VAT
was expanded to include a vertical water tunnel
(Reference 77 ) , at the same time retaining its
hydroballistics capabilities.
Figure 19.1 . Variable-Atmosphere Tank (VAT) at NUC.
The Lockheed Underwater Missile Facility
The Lockheed Underwater Missile Facility
windows and parts are shown in Figure 19.2 . (LUMF ) at Sunnyvale , Calif., was constructed to
The entire interior of the tank can be covered study the water exit of 11 -inch-diameter missiles
photographically with cameras at the observa in the presence of waves operating in Froude
tion windows and at corresponding alternate cavitation -number scaled systems . The LUMF
port holes on each side of the main windows . ( Figure 19.4) was subsequently enlarged to give
Ports used for camera coverage can be equipped it towing capabilities.
with spherical windows to remove distortion The tank itself is made of reinforced
causedby light refraction. Either photoflood or concrete and is a wave channel 15 feet wide ,
Edgerton -type stroboscopic flash lamps provide 17 feet deep, and 180 feet long, with an operat
illumination . Silhouette photography is achieved ing water depth of 15 feet. It has a 17-foot-high
by lighting the field from the rear ports. piston-type wave generator capable of generating
A side access door about 25 feet above the waves 0 to 4 feet crest to trough , with periods
tank floor is equipped with a monorail and hoist from 1.07 to 5.37 seconds and a beach-type
of 1 /2-ton capacity that can be extended inside wave absorber. Two basins 25 and 37 feet deep
318
 C

Figure 19.2. Interior of VAT, with launcher at bottom.

provide underwater launching capabilities for speeds ( Figure 19.5 ) . Gun launchers both above
missile water-exit studies. An underwater and below the water surface propel the models
launcher is available for testing the 11-inch into the tank.
diameter missiles and a missile catcher is The tank occupies the central portion of a
installed above the water surface . nine-story building resting on a concrete mat
The tank's tow system provides controlled foundation, 36 inches thick, laid on solid rock .
speeds of 5 to 50 fps in both directions and Four stories of the building are below grade. The
continuous accelerations from 1 to 8 fps. A bottom of the tank is a prestressed concrete slab
pressure shell 27.5 ft in diameter is mounted over 24 inches thick; the top , a concrete slab
the tank, and the air pressure above the water 42 inches thick; and the walls, supported outside
surface can be varied from ambient to 0.73 psia. by vertical and horizontal reinforced concrete
Instruments installed inside the pressure shell ribs, are 14 inches thick .
include recording oscillographs, flight recorders, The structural ribs serve also as the parti
high -speed cameras, and transducers. A Van tions and floors of the observation cubicles
Slyke gas analyzer ( Reference 62 ) is used to ( Figure 19.5 ) , 23 of which are equipped for
determine the air content of the water. The remote operation of recording cameras and
design and operation of this versatile facility are instrumentation in the test tank . There are 124
described in Reference 87 . observation ports ( 16 inches in diameter and
1-1 /4 inches thick ) in the walls, top, and bottom
The NOL Hydroballistics Facility of the tank for coverage of missile launchings
A tank measuring 100 by 35 feet and and underwater trajectories.
75 feet deep was constructed in the Hydroballis Fifteen 36 -inch - diameter gunports in the
tics Laboratory at the Naval Ordnance Labora top , bottom, and north end of the tank accom
tory , White Oak , Md. ( References 69 and 75 ) to modate the launchers . Monorail hoists handle
allow study of relatively large models entering, the launcher equipment outside the tank, and an
leaving, and passing through the water at high overhead - travel crane of 6- ton capacity ,

319
 remotely controlled from the cubicles, handles
the equipment inside the tank .
Special water-treatment equipment supplies
the 1,750,000 gallons of clear water for high
speed photography. To eliminate all possible
sources of contamination , the entire tank is
lined with 1 /8-inch-thick stainless -steel plate,
CHECK WIT
PRIOR TO
and all accessories in the tank, including the
Net LAUNCHER overhead travel crane , are of stainless steel and
other noncorrosive materials. Water drained out
to change the level or to empty the tank is
discharged through a 36-inch-diameter stainless
steel pipe into a storage container having an
interior specially coated to prevent
contamination.
The tank is virtually airtight to allow
control of atmospheric conditions, being evacu
ated by steam - ejector equipment on the roof of
the building. The structure was designed to
withstand a differential pressure of 14.7 lb/in.2
with a water depth of not less than 35 feet.

The CIT Variable-Atmosphere Wave Tank

Small, pilot-type facilities are needed for
Figure 19.3 . VAT access doorway . exploration of the hydroballistic phenomena

Pressure shell
27.5' diameter

180 ft

Old tank New tank

Catcher

17 ' 25 ' 37 ' 27 '

Beach

Wave
Torpedo tube
generator launcher
Towing
carriage Vertical launchers

Figure 19.4 . Lockheed Underwater Missile Facility (LUMF ).

( From Reference 87. )

320
 1 WATER - ENTRY LAUNCHER 6 PERSONNEL ELEVATOR 9 PARKING LOT
2 UPPER LAUNCHER GALLERY 7 TEST TANK ( STEEL LINED) 10 LOADING AREA
3 PHOTO OBSERVATION CUBICLE 8 STORAGE TANK ( STEEL ) 11 WATER EXIT LAUNCHER
4 LABORATORY SPACES 12 MECHANICAL AND ELECTRICAL
5 MAIN ENTRANCE SERVICE EQUIPMENT

00 8

7 6

00

12
5

Figure 19.5 . Hydroballistics Facility at NOL.

predicted by hydroballistic theory and for of cavitation numbers within the tank . Water of
detailed observations that might be useful in the high clarity necessary for metric strobo
extending the theory itself. Such a facility is the scopic photography is maintained with a
Variable -Atmosphere Wave Tank at CIT (Ref Millipore filter .
erence 93 ), designed for the study of the water The 5.6 -foot-long test section ( Figures 19.6
entry, water-exit, and underwater -trajectory and 19.7 ) has a turret 2 feet high, with a
behavior of small momentum-propelled missiles removable Cover for missile loading and
for varied trajectory angles, missile accelerations recovery. In the floor of the test section there
and velocities, wave fields, and conditions of are three 1 -foot-diameter holes for mounting
cavitation . missile launchers. One hole is directly beneath
The wave tank is shown in Figures 19.6 and the turret for vertical or near-vertical launchings;
19.7. It is a reduced-pressure vessel constructed the other two are on either side for oblique
almost entirely of 2 -inch -thick Lucite sheeting, launchings into the turret. The missile may be
its three flanged sections bolted together with launched either into the advancing wave or away
nylon bolts and sealed with O-rings : a wave from it. A nylon fabric backstop is used to
generation section , a test section , and a wave terminate missile flight. When not in use , the
absorption section . The tank rests on four holes are covered from underneath by plates so
supports equipped with screws for height adjust constructed that the holes are filled and the
ment and leveling. The two inner supports are channel bed is smooth and uniform.
heavy laminated wood piers with fiber A major feature of the wave tank is its
impregnated plastic bases and adjusting screws . magnetic propulsion system that operates on the
The two outer supports are made of steel, and principle of an accelerating coil. A magnetically
the tank end-plate associated with the wave permeable sphere is placed on the axis of aa wire
generation equipment is made of stainless steel. coil whose radius is large compared both to the
The floor of the tank is about 3 feet from sphere and to the coil -winding cross section. A
ground level. current pulse passing through the coil when the
There are access ports with O-ring fitted sphere is on axis but not centered in the coil
covers in the wave -generation and absorption causes the sphere to be drawn toward the coil
sections. Tank air pressure can be lowered to the center with a force proportional to the spatial
vapor pressure of water, providing a wide range gradient of the energy in the magnetic field. The
321
 ww

C
Figure 19.6. Variable-atmosphere wave tank at CIT . ( From Reference 93.)

Launching-well field energy is proportional to the square of the

Wave -generating flap holes ( 3) current. If the current terminates just before the
11.57 / 4.0 +1.541.5+ -6.5 sphere reaches the coil center, the sphere con
T tinues in its path along the coil axis without
1.5
1 being decelerated by magnetic forces. Propulsion
-4.2 te 2.8-72.8- + 5.2
is thus achieved without introducing extraneous
15.0 hydrodynamic disturbances. An example of an
accelerating coil launching is shown in
Figure 19.8.
Access port Hoist
The wave tank and the coil holder are
with cover Missile back stop constructed of nonmagnetic and nonconducting
Motor - crank +2.04 Turret materials to obviate reaction force on the coil .
flap drive Nominal water
T Thus, the force on the coil is equal and opposite
level 2.0 to the propelling force applied to the missile. If
0
the coil is instrumented externally, it may be
1.5.12.
possible to measure the force directly without
Air -tight missile attachments or telemetering. The data so
piston seal 3.0 obtained can be combined with data from
Flap hinge
displacement-time photographs to provide infor
Launching well Missile acceler
mation on virtual mass and drag under condi
ating coil tions of suddenly accelerated motion
Missile
(Reference 92 ).
Figure 19.7 . Detail of variable -atmosphere wave tank. The propulsion force on the magnetic
( From Reference 93.) sphere, being a true body force rather than a
322
 speed gear-motor system, is used to generate
progressive waves from 1 to 3 feet in length and
up to 3 inches in amplitude. The flap is
1-1 / 2 feet from the end of the channel so that
fluctuation in water level behind it is moderate
and less than flap height for all flap amplitudes.
The distance from the wave generator to the
center of the test section is 5-1 /2 feet, almost
four times the maximum wave depth . Theory
indicates ( Reference 109 ) that the extraneous
disturbances of the water at the wave generator
become negligible in two or three wave depths
of travel; hence essentially free gravity waves
should exist in the test section .
This method of generating waves was

chosen because the longest wave contemplated

for study is about 3 feet in length , relatively
short in comparison with the water depth , and
theory shows (Reference 71 ) this type of genera
tor to be suitable for producing such waves.
Moreover, the inertia of the generator is not
large, the generator is easily regulated, and the
flap motion required to produce a desired wave
is calculable.
No wave absorber is presently used because
Figure 19.8 . Steel sphere launched vertically upward by the wave channel, extending 8 feet beyond the
submerged accelerating coil . ( From Reference 93.) center of the test section , is long enough for
four of the longest waves (3 feet) to pass the
contact force, prevents the generation of elastic center of the test section before the reflected
waves within the sphere itself that might inter waves return . It is believed that the wave train
fere with an analysis of the hydroelastic effects will stabilize and that tests can be completed
under study . during this interval. Tests of longer-duration
Using this magnetic propulsion system, a waves may require a wave absorber.
1 -inch - diameter stainless-steel sphere attained an
underwater velocity of about 50 fps in 2 inches ARL Controlled Atmosphere Tank
of travel from rest for an expenditure of A rectangular, glass-front, steel tank (Fig
20,000 watt -seconds of electrical energy when ure 19.9 ) was constructed at the Alden Research
an 8-inch ID coil was used. With a 1-1 /2-inch ID Laboratories,2 Worcester Polytechnic Institute,
coil, about 90 fps was attained for about Worcester, Mass., for water -entry , water-exit,
1,500 watt-seconds of energy. The reduction in and underwater trajectory studies of small
efficiency with increase in coil size was models (Reference 3 ). The rubber-lined interior
expected. The period of missile acceleration can of the tank measures 9 feet in length by 4 feet
be varied by adjustments in the electrical circuit high by 4-1 / 2 feet wide. There are three hatches,
and, if desired , the missile may even be made to two on top and one on the end, where a
describe oscillations . launcher may be mounted for studies of water
The tank wave channel (Figure 19.6) is entry, water exit at low velocities, and under
15 feet long, 1-1 /2 feet wide, and 2 feet high, water trajectories.
with a maximum undisturbed water level for For observation or photography, a
wave studies of about 1-1 / 2 feet. A rigid flap , 1-1 /4-inch-thick tempered glass plate 9 feet by
hinged at the bottom of the channel and 4 feet is mounted as one side of the tank. A
actuated by an adjustable crank and controlled 70-mm strip camera with Edgerton-type flash

2 Formerly the Alden Hydraulic Laboratory .

323
lamps is used. With the present electronic mounted on a hollow drum and can be rotated
equipment, these lamps have a maximum rate of to vary the launching angle. The missile ,
1100 flashes per second. The gas pressure in the attached to the plunger by means of small
tank can be varied from atmospheric pressure to springs, is directed by a set of guides during its
0.06 atmosphere. Evacuation of the tank to 0.06 launching phase and is in free flight after 2 feet
atmosphere, when operating at normal water of guided travel. With this arrangement, models
elevation, takes about 20 minutes. With heavy up to 2 inches in diameter can be launched at
gases, both the pressure and density of the velocities up to 150 fps. No variation in tank
atmosphere above the water surface can be pressure results from this launching technique .
varied. Missile launchings at entry angles from 0 to
The tank's carbon dioxide powered 90 degrees, exit angles up to 30 degrees, and
launcher of the plunger type (Figure 19.10) is underwater trajectories can be studied.

Figure 19.9 . Controlled atmosphere tank at Alden Research Laboratories.

324
 O
D
O
O
I

Figure 19.10. Launcher on Controlled Atmosphere Tank .

325
 PART VI .
THE CHANGING STATUS OF MODELING

CHAPTER 20
THE CHANGING STATUS OF MODELING

Present-day knowledge of modeling is at conditions are yet to be established or

best only partial and , certainly , the state of the confirmed .
art does not permit any dogmatic assertions The effects of sea state upon missile
regarding the reliability of existing techniques. behavior need to be investigated. Wave and wind
Modeling has persisted as an art rather than as a action of a highly variable nature exists at the
well-ordered science because there is no infalli air -water interface. Knowledge of the flow about
ble way to measure the relative significance of a missile operating immediately beneath the
the physical forces in a model system . The order surface is lacking, and little is known about the
of numerical magnitude usually used in engineer effects of deep submergence on the cavitation
ing as a measure of significance is inadequate characteristics of missiles and on the various
because the resultant of the large forces deter power plants that are dependent upon external
mining a phenomenon may be nearly zero. exhaust . Variations within the sea water as
Understanding of the phenomena to be modeled introduced by salinity, thermal layers , etc. , may
and past experience with similar modeling also trigger irregular cavity behavior . Effects
systems help in evaluating the physical forces such as these must be recognized before missile
involved, yet unexpected results sometimes behavior can be predicted with confidence.
occur. Such modeling failures can be of subtle It is almost certain that very high pressures,
origin. many times greater than stagnation pressure, are
Experimental studies of water entry have present at impact; but these pressures , never
produced аa set of empirical design criteria fully evaluated at lower entry velocities, are
applicable to a few specific problems but little virtually unexplored in the sonic regime (Refer
information of a general or systematic nature ences 52 , 83 , and 84 ) . They are important
has evolved and design data are lacking in many structural design criteria, particularly as the
areas . When scaling conditions are known, entry velocities approach the transonic (4000 –
experimental modeling studies can be pro 5000 fps) range. At these velocities , the pressure
grammed to obtain general or specific design surge will probably be important in determining
data, but in many areas the actual scaling the trajectory and may also affect the entry

327
cavity. Compressibility is a factor that requires Instead, it is shed or progressively entrained in a
consideration in all media. variety of ways by the flowing water. It is known
Mitigation of water-entry shock will be a that the flow at the base of the cavity sometimes
major problem at the higher entry velocities. terminates in re - entrant jets directed upstream,
Frangible nose caps have been used to cushion but the exact flow conditions producing these
water-entry shock, and a step -nosel has been jets are not known , even though the jets are
considered , but usually the issue has been sometimes of sufficient strength to interfere with
avoided by deploying a parachute to decelerate missile performance. The momentum exchange
the missile prior to entry. Study of the high between the missile and the water might be de
velocity regime should also include the modeling termined by observing the motion of a water
of ricochet characteristics. entry cavity after the generating missile has been
The scaling laws required to determine a stopped suddenly .
Mach -modeling system have not yet been formu The phenomena of water exit requires
lated. Were it possible to create an ideal system further study. Some theoretical study and a
of perfect fluids and frictionless, rigid solids, great deal of particular water-exit information
these scaling laws could be easily established ; in have been generated by the Navy's Polaris and
a system of real materials, the physical Poseidon programs, but it is recognized that
phenomena must be described and the signifi much more investigation is needed. Existing
cant forces identified first. experimental data show that the exit phenom
Extrapolation of the vast library of water ena are at least as sensitive to secondary forces
tunnel data to the entry cavity is attractive , but as are those of entry .
the steady-state cavity occurring in a water Pre-exit flight is of major importance in
tunnel differs from the transient cavity at water determining the linear and angular velocities
entry (Reference 101 ) . The importance of retained by the missile following water exit.
secondary forces in determining cavity separa Transient-flow phenomena arising from random
tion might profitably be studied in a water turbulence in the boundary layer and asymmet
tunnel , however, or a controlled , transient cavity ric flow patterns caused by irregular cavities2
might be generated there after greater know can produce a great variety of exit angles and
ledge is gained of the pressure variation in a velocities in successive launchings even though
typical water-entry cavity. the missiles have been launched under presum
A record of cavity pressure as a function of ably equal conditions. The possibility of
time, and a special pressure distribution to stabilizing performance by bolstering a weak or
describe gas flows within the cavity would fluctuating cavity with exhaust material from an
increase understanding of cavity growth , shape, internal propulsion system is attractive. Cavity
and dissipation ; knowledge of cavity pressure at life might also be extended artificially by intro
the critical points of missile-cavity separation ducing wetting agents into the boundary layer
would aid in evaluating the forces in equili ahead of the separation point, effectively reduc
brium . A meaningful study of the relationship ing the surface tension in the boundary layer,
between the surface condition of the missile and thereby reducing the cavity-closure force.
cavity formation might then be possible. The release of hydrodynamic forces occur
Little is known of cavity attrition – how ring as the missile becomes airborne will in
little might be measured by the fact that the dis general be asymmetric and can easily be large,
appearance of the water-entry cavity is often mis increasing the probability of significant pertur
takenly termed “ cavity collapse .” The water bation of missile motion at water exit.
entry cavity , being partially of a noncondens The first investigations of water exit should
able gas, does not indeed cannot collapse. be directed toward control of the pre-exit flight;

lEffective, because at entry it generates a series of progressively larger cavities, thus dividing the total entry shock into
several smaller shocks . The step-nose appears promising from the standpoint of shock mitigation , but other problems
will inevitably arise from its use .

2Observation of cavities formed by missiles exiting at low velocities suggests that the large vortex rings shed from the
base of the cavity must influence the exit perturbation (Chapter 11 ) .
328
then the effects of the launcher , the missile the cavity introduces a rudder effect that can
power plant , and the sea state should be alter both the trajectory and the missile orienta
examined . tion and can also affect the operation of the
If the problems of hydroballistics have a sensing equipment within the missile. Under
common denominator, it is the cavity ; inevi some circumstances, there might be some
tably, cavity behavior is examined first as a benefit in jettisoning the cavity artificially by
measure of any new effect. Undesirable cavity suddenly slowing the missile with a parachute
behavior can disrupt missile performance com device .
pletely , yet this same cavity offers a promising Many of the experimental techniques and
possibility of missile control. It is evident from design criteria discussed in Chapters 18 and 19
several experimental studies (Chapters 8 and 11 , are now in common use ; they represent the state
for example) that under some circumstances the of the art. This was not true when the tech
cavity imparts a highly desirable stability to niques and criteria were evolved , nor are they
missile performance. It is not now possible to necessarily familiar now to those not intimately
produce a stabilizing cavity upon demand ; to be concerned with experimental hydroballistics. It
able to do so would be of great advantage . is hoped that the studies reported here , most of
The shape of the cavity will affect missile them conducted at the Naval Undersea Research
control in many ways . Too large a cavity will and Development Center, Pasadena Laboratory,
not only increase drag but will make large will provide a basis for assessing the present state
changes in missile orientation possible within the of the modeling art, a springboard for future
cavity . The separation of too slender a cavity is experiments , and a guide for extending the
inherently unstable (Chapters 3 and 8 ). The modeling technique to other phases of
possibility of using the interaction between the hydroballistics .
jet exhaust and the cavity wall as a means of Much remains to be done to advance and
missile control offers an attractive way to refine the techniques of hydroballistics. Cam
influence the trajectory of a quasi-steady, eras , telemeters , pressure-sensing devices, and
cavity-running projectile that has only its nose in small power plants presently available place the
contact with the water . suggested studies within the realm of possibility,
The rate of cavity dissipation and the although challenging development of both
cavity shape as dissipation occurs can be critical instrumentation and experimental techniques
factors in missile performance; a sudden shift in still lies ahead .

329
 NOMENCLATURE

A Maximum cross-sectional area of missile , ft2

a Water-entry deceleration of missile , g

Drag coefficient of missile [CD = 1/(1/2)Pw v2 A]

‫یکبیھی‬

Ср W

Liftcoefficient of missile (C1 = L/( 1/2 ) P w v2 A ]

-
A

Molecular heat at constant pressure, cal (mole °C)-1

Molecular heat at constant volume,cal(mole °C )-1
с Velocity of sound in water, ft sec- 1
D Drag of missile, lb ; force exerted on missile in aa direction parallel to line of travel
d Diameter of missile , in .

E Bulk modulus of elasticity of water, lb ft- 1 sec--2

F Froude number ; ratio of inertial and gravitational forces [ F = v/(dg) 1/21

=

f
Flexure number; ratio of inertial and flexural forces (f = v2/n222)
=

g Gravitational acceleration, ft sec-2

h Depth of missile below water surface, in ., ft , missile diameters
I Moment of inertia of missile about any transverse axis through CG, slug ft? or lb
in2 ; light intensity in forward direction (Equation 19.1 )
k Light absorption coefficient (Equation 19.1 )
L Lift of missile , lb. Component in vertical plane of force exerted on missile in
direction normal to line of travel

331
 O

l Distance from missile nose to CG , in. (Equation 1.10 ) ; beam or missile length in
flexure number scaling, ft (Equation 1.7 ) ; generalized measure of length in
Table 3.2

M Molecular weight
m Mass of missile , lb
n Moles of gas

n
no Fundamental transverse frequency of missile , Hz (Equation 1.7 )
P.
Flat plate radius (P = R
=

Rp - Ry, Figure 5.2 )

o

Pa Atmospheric pressure (gas pressure over water surface), atm., torrs, lb in-2
Pc Sum of gas and vapor pressures in cavitation bubble acting to keep bubble open,
=

atm.,, torrs,, lb in-2 (Pc = Pg + Py)

Pg Gas pressure (exclusive of vapor pressure) in cavitation bubble acting to keep
bubble open ; gas pressure in missile underwater launching system , atm. , torrs, lb
in -2

Py Fluid vapor pressure , atm. , torrs, lb in-2

Q Parameter used to present ogive-plate -ogive data [Q = 1 /( 1 + R /Rp)]
R Reynolds number ; ratio of inertial and viscous forces (R = vd/ v) ; spherical cap
radius (Figure 5.2 )
R' ° ]
Gas constant [RR '= 1.987 cal (mole ºc)-1
Ro Ogival radius (Figure 5.2 )

Rp Missile radius ( Figure 5.2 )

r Sphere radius
r' Radius of cavity bubble in region where striation angle , € , is measured
(Figure 15.1 )
S m -1
Surface tension of fluid , dynes ccm
S Missile water penetration measured along trajectory from point of water entry,
in . , ft , missile diameters

Tg Temperature of gas , °C
IN
W Temperature of fluid , °C
t Time, sec ; measured from instant of missile water contact or emergence

V Volume of gas in missile underwater-launching system or cavity bubble

332
 V Missile velocity, ft sec-1
g Gas velocity , ft sec-1
eEao

Work performed by gas expansion in missile underwater-launching system

Missile angle of attack in pitch , deg ; angle in vertical plane between missile axis
and direction of motion , positive in sense of nose-up rotation (Figure 1.1 ) ;
half-sphere angle ( Figure 5.2 )
r Gas specific heat ratio (y = Cp /Cy)
дө Whip of missile, deg sec-?; change in missile pitch velocity at water entry over
impact and flow -formation phases
8
Slope (deviation from vertical) of water-entry splash at point of separation from
sphere , deg (Figure 16.5 )
E
Angle between water -entry cavity striations and plane normal to direction of
travel of sphere , deg (Figure 15.1 )
Š

Generalized angle ; used in Equations 1.20 , 1.21 , and 1.22 to represent pitch
angle , 0 , trajectory angle , š, angle of attack, Q, yaw, y , or roll , 0 .
>

n Surface tension of fluid, dynes cm -1 ;;used in Table 3.2

ө Missile angle of pitch, deg ; angle between missile axis and horizontal plane,
positive in sense of nose-up rotation (Figure 1.1 )
ec -1
Pitch velocity , deg ssec

Missile angle of inclination , deg ; projection of missile pitch angle on vertical

plane containing the trajectory; positive in sense of nose-up rotation (Figure 9.5 )
λ Modeling scale factor; ratio of the model and prototype missile diameters used
for convenience ( = dm/dp)
M Viscosity of fluid , lb ft- 1 sec-1
V Kinematic viscosity of fluid, ft2 sec- 1
६ Missile trajectory angle , deg ; path angle of missile CG with respect to horizontal
plane, positive in climb (Figure 1.1 )
p Density of gas , slug ft -3 or in terms of the density of a “ standard atmosphere”
(see definition ofp ' )
р Gas-density coefficient; ratio of gas density in the model system to density of air
at 20 °C and 740 torrs.

Pw Density of fluid , slug ft-3

333
o Cavitation number; ratio of the static and dynamic pressures [o = (Pa +
0
wgh -
Pg - Py)/(1/2) 2 wv2]
V

T
Weber number; ratio of the inertial and surface tension forces [7T = T =

v /(S /P wd)1/2]
Missile angle of roll , deg ; angle through which missile rotates about its
longitudinal axis , positive in clockwise direction as viewed from the rear
4

Missile angle of yaw, deg ; angle between missile axis and vertical plane
containing trajectory, positive when missile tail is to the right as viewed from
rear

Empirical cavity closure parameter (Equation 9.1) ; rate of spin of sphere about
3

an axis parallel to its direction of motion, rad sec-1 (Equation 15.1 )

Subscripts

e Water -entry or exit conditions

m Model conditions
p Prototype conditions
O Initial condition
1 Final or end condition

334
 BIBLIOGRAPHY

1. Abelson , H. I. Cavity Shapes at Vertical Water Entry : a Comparison of Calculated and Observed
Shapes. White Oak, Md., U.S. Naval Ordnance Laboratory, January 23 , 1967. (NOL
Technical Report 67-31 . )

2. Aerosol Wetting Agents. New York , N.Y. American Cyanamid Co. , 1946 .

3. Alden Hydraulic Laboratory. Controlled Atmosphere Tank. Worcester, Mass., Worcester

Polytechnic Institute, 1958-59 . (Contract NORD 15520. )

4. Birkhoff, Garrett. “ Formation of Vortex Streets , ” in Journal of Applied Physics , Vol. 24 ,

No. 1 ( 1953 ) , pp. 98-103 .
5. Hydrodynamics: a Study in Logic, Fact and Similitude. Princeton , N.J., Princeton
University Press for University of Cincinnati, 1950. 186 pp.
6. Hydrodynamics: a Study in Logic, Fact and Similitude. Princeton , N.J., Princeton
University Press, 1960. Revised edition, 184 pp .
7. Birkhoff, Garrett, and Thomas E. Caywood . “ Fluid Flow Patterns,” in Journal of Applied
Physics , Vol. 20 , No. 7 ( 1949 ) , pp . 646-59 .

8. Birkhoff, Garrett, and Robert Isaacs. Transient Cavities in Air -Water Entry . Washington, Navy
Department, January 1951. (NAVORD Report 1490. )

9. Birkhoff, Garrett, and E. H. Zarantonello . Jets, Wakes, and Cavities. Vol. 2 of Applied
Mathematics and Mechanics. New York , N.Y., Academic Press , 1957 , 353 pp .
10. Burt , F. S. “ New Contributions to Hydroballistics,” in Advances in Hydroscience. Vol. 1 , Ven
Te Chow, ed . New York , N.Y., Academic Press, 1964. Pp. 182-280 .

11. Christie, William H. Determination of the Underwater Trajectories of Missiles by the Use of
Nets . Inyokern , Calif. , U.S. Naval Ordnance Test Station, November 25 , 1949 .
(NAVORD Report 1200. )

335
12 . The Rotating Plane -Parallel Plate Flare Camera. Inyokern , Calif. , U.S. Naval
Ordnance Test Station , April 17 , 1950. (NAVORD Report 1209. )
13 . Underwater Photography , Its Application to High - Speed Missiles . China Lake ,
Calif., U.S. Naval Ordnance Test Station, December 15 , 1950. (NAVORD Report 1247. )
14. Cooper, E. P. Theory of Water Entry of Missiles with Flat Noses. Inyokern , Calif. , U.S. Naval
Ordnance Test Station, April 26 , 1949. (NAVORD Report 1154. )
15. Dergarabedian, Paul. Compressibility Effects During Water Entry . China Lake , Calif. , U.S.
Naval Ordnance Test Station, June 23 , 1955. (NAVORD Report 3523. )
16. Dodge , Barnett F. Chemical Engineering Thermodynamics. 1st ed. New York, N.Y. ,
McGraw -Hill Book Co., 1944. 680 pp .

17. Dorsey , N. Ernest. Properties of Ordinary Water-Substance . New York, N.Y. , Reinhold
Publishing Corp., 1940 , 673 pp .

18. DuPré, Elsie F., and Leo H. Dawson . Transmission of Light in Water : an Annotated
Bibliography. Washington, Naval Research Laboratory, April 1961. (NRL Bibliography
No. 20. )

19. Ellis, A. T., and M. E. Fourney. “Application of a Ruby Laser to High Speed Photography,” in
Proceedings of the Institute of Electrical and Electronics Engineers , Vol. 51 , No. 6
( 1963 ) , pp. 942-3 .
20. Fabula , A. G. “ Ellipse-Fitting of Two-Dimensional Normal Symmetric Impact of Rigid Bodies
on Water,” in Proceedings of the Fifth Midwestern Conference on Fluid Mechanics,
April 1 and 2 , 1957. Ann Arbor, Mich., the University of Michigan Press, 1957. Pp.
299-315 .

21. Fabula, A. G. , and I. D. Ruggles, Vertical Broadside Water Impact of Circular Cylinder :
Growing Circular-Arc Approximation. China Lake, Calif. , U.S. Naval Ordnance Test
Station , October 12 , 1955. (NAVORD Report 4947. )
22. Gilbarg, David , and R. A. Anderson . “Influence of Atmospheric Pressure on the Phenomena
Accompanying the Entry of Spheres into Water, ” in Journal of Applied Physics , Vol. 19 ,
No. 2 ( 1948 ), pp . 127-39 .
23. Goldstein , S. , ed. Modern Developments in Fluid Dynamics. Vol . I. Oxford , Great Britain ,
Oxford University Press, 1938. Pp. 1-330 .
24 . Modern Developments in Fluid Dynamics. Vol. II. Oxford, Great Britain. Oxford
University Press , 1938. Pp. 331-702 .

25. Handbook of Chemistry and Physics. 48th ed. Cleveland, Ohio, Chemical Rubber Publishing
Co. , 1967-68 .

26. Hoover, W. R., and V. C. D. Dawson . Hydrodynamic Pressure Measurements of the Vertical
Entry of a Sphere. White Oak , Md . , U.S. Naval Ordnance Laboratory , October 26 , 1966 .
a

(NOL Technical Report 66-70 . )

336
27. Jennison, James H. The Variable-Angle Launcher for Air-to-Water Missiles . Inyokern, Calif. ,
U.S. Naval Ordnance Test Station, May 15 , 1950. (NAVORD Report 1180. )
28. Kermeen , R. W., and B. R. Parkin . Incipient Cavitation and Wake Flow Behind Sharp-Edged
Disks. Pasadena, Calif. , California Institute of Technology , August 1957. (Engineering
Division Report No. 85-4. )

29. Knapp. R. T. “ Special Cameras and Flash Lamps for High-Speed Underwater Photography,” in
Journal of the Society of Motion Picture Engineers , Vol. 49 , No. 1 , ( 1947 ) , pp . 64-82 .
30. Knapp . R. T. , and Joseph Levy, J. P. O'Neill , and F. B. Brown . “ The Hydrodynamics
Laboratory of the California Institute of Technology,” in Transactions of the American
Society of Mechanical Engineers, Vol. 70 , No. 5 ( 1948 ) , pp . 437-57 .
31. Knapp , Robert T. Entrance and Cavitation Bubbles . Pasadena , Calif. , California Institute of
Technology , December 27 , 1944. (Hydraulic Machinery Laboratory Report No. ND-31 . )
32 . Hydrodynamic Force Resulting from Cavitation on Underwater Bodies. Pasadena,
Calif ., California Institute of Technology, July 21 , 1945. (Hydrodynamics Laboratory
Report No. ND-31.2 . )
33 . Nose Cavitation : Ogives and Spherogives . Pasadena, Calif., California Institute of
Technology , January 18 , 1945. (Hydraulic Machinery Laboratory Report No. ND-31.1 . )

34. Kodak Plates and Films for Science and Industry. Rochester, N.Y. , Eastman Kodak Co. , 1967 .
(Kodak Publication No. P-9 . )

35. Lamb, Horace. Hydrodynamics. 6th ed. New York, N.Y. , Dover Publications , 1945. 738 pp .

36. Levy, Joseph . On Modeling the Oblique Entry into Water of Air Launched Missiles . Pasadena,
Calif., California Institute of Technology , August 1956. (Hydrodynamics Laboratory
Report No. E- 12.19 . )

37. Levy , Joseph, and John Kaye. Investigations of Water Entry Phenomena : Preliminary Studies
of Effect of Atmospheric Pressure on Trajectory of 2-Inch Correlation Model of
Mark 13-6 Torpedo . Pasadena, Calif. , California Institute of Technology , March 1 , 1948 .
(Hydrodynamics Laboratory Report No. M-59 . )
38 . Studies of Water Entry : Effect of Atmospheric Pressure on Entry Behavior of
Models of Mark 13-6 Torpedo With Standard Head (Head F ) and One Finer Head
(Head I) . Pasadena,, Calif., California Institute of Technology , January 1949 .
(Hydrodynamics Laboratory Report No. N-59 . )
39. Levy, Submarines
Joseph , and Donald A. Price , Jr. Free-Body Modeling of the Stability and Control of
. Pasadena, Calif. , California Institute of Technology , May 1956 .
(Hydrodynamics Laboratory Report No. E-27.2 . )
40. Loeb , Leonard , The Kinetic Theory of Gases . New York, N.Y. , McGraw-Hill Book Co. , 1934 .
687 pp .

41. Luthander, S., and A. Rydberg. “ Experimentelle Untersuchung uber den Luftwiderstand bei
einer um eine mit der Windrichtung parallele Achse rotierenden Kugel,” in Physikalische
Zeitschrift , Vol. 36 , No. 16 ( 1935 ) , pp . 552-8 .
337
42. Marey , M. “ Le Mouvement des Liquides : Etudie par la Chronophotographie,” in Comptes
Rendus de l'Academie des Sciences, Vol . 116 ( 1893 ) , pp. 914-24.
43. May, Albert. "Effect of Surface Condition of a Sphere on Its Water-Entry Cavity,” in Journal
of Applied Physics, Vol. 22 , No. 10 ( 1951 ) , pp. 1219-22 .
44 . The Influence of the Proximity of Tank Walls on the Water-Entry Behavior of
Models . White Oak, Md., U.S. Naval Ordnance Laboratory, October 3 , 1951. (NAVORD
Report 2240. )
45 . Research for Design of a Hydroballistics Tank. White Oak, Md. , U.S. Naval
Ordnance Laboratory, August 30 , 1963. (NOL Technical Report 63-119 . )
46 . “Vertical Entry of Missiles into Water,” in Journal of Applied Physics , Vol. 23 ,
No. 12 ( 1952 ) , pp. 1362-72 .
47. May, Albert, and Jean C. Woodhull. “Drag Coefficients of Steel Spheres Entering Water
Vertically , ” in Journal of Applied Physics, Vol. 19 , No. 12 ( 1948 ), pp . 1109-21 .
>
48 . “ The Virtual Mass of a Sphere Entering Water Vertically ,” in Journal of Applied
Physics , Vol. 21 , No. 12 ( 1950 ) , pp . 1285-89 .

49. Mees , G. E. Kenneth. The Theory of the Photographic Process. 3rd ed., T. H. James , ed. New
York , N.Y., the Macmillan Co., 1966.591 pp .
50. Mosteller, G. G. Axial Deceleration at Oblique Water Entry of 2-Inch-Diameter Models with
Hemisphere and Disk-Cylinder Noses. China Lake, Calif. , U.S. Naval Ordnance Test
Station, February 12 , 1957. (NAVORD Report 5424. )
51 . -- Spin Rate of Spheres Under Water as Determined from the Direction of Striations
on the Cavity Formed at Water Entry . Inyokern , Calif. , U.S. Naval Ordnance Test
Station, April 1948. (NAVORD Report 1012. )
52 . Water -Entry Acceleration : an Annotated Bibliography. China Lake, Calif. , U.S.
Naval Ordnance Test Station , December 19 , 1956. (NAVORD Report 5397. )
53. Mosteller, G. G. , and H. I. Scribner . Improved Instrumentation for Water-Entry Modeling:
Simultaneous Whip and Sideview Photography. China Lake , Calif. , U.S. Naval Ordnance
Test Station , May 19 , 1955. (NAVORD Report 3499. )

54. Mosteller, G. G., and J. G. Waugh. Laboratory Facilities for Model Studies. Inyokern, Calif.,
U. S. Naval Ordnance Test Station, November 14 , 1949. (NAVORD Report 1198. )
55. National Defense Research Committee. Mathematical Studies Relating to Military Physical
Research . Vol. 1 of Summary Technical Report of the Applied Mathematics Panel,
NDRC. Washington , NDRC , 1946 .
56 . Torpedo Studies. Vol. 21 of Summary Technical Report of Division 6 , NDRC .
Washington , NDRC , 1946 .
57. Neale, Lawrence C., Alfred H. Barrett, and James A. Wentworth. Development of aa Telemetry
System for Water-Entry Missiles. Worcester, Mass . , Worcester Polytechnic Institute ,
December 1965. (Alden Hydraulic Laboratory Report No. 58. )
338
58. Nisewanger, C. R. Experimental Determination of Pressure Distribution on a Sphere During
Water Entry . China Lake, Calif. , U.S. Naval Ordnance Test Station , October 27 , 1961 .
(NAVWEPS Report 7808. )

59. Office of Scientific Research and Development. Aircraft Torpedo Development and Water
Entry Ballistics . Pasadena, Calif., California Institute of Technology , 1946. (OSRD
Report No. 2550. )
60 . Photographic Studies in Underwater Ballistics. Pasadena, Calif., California Institute
of Technology , November 15 , 1945. (CIT Morris Dam Report 153 ; CIT IPC 81. )
61 . Water Entry and Underwater Ballistics of Projectiles. Pasadena, Calif. , California
Institute of Technology, 1946. (OSRD Report 2551. )

62. Peters, J. P., and D. D. Van Slyke. Quantitative Clinical Chemistry . Vol. II, Methods . Baltimore ,
Md. , Williams and Wilkins, 1932. Pp . 267-74 .

63. Pitzer, R. L., J. A. DeJulio , R. A. Nielsen , H. C. Riggins, and Gardner Wilson. A Digital
Instrumentation System Developed at the Naval Ordnance Test Station. China Lake,
Calif., U.S. Naval Ordnance Test Station, May 25 , 1954. (NAVORD Report 3332. )
64. Pitzer, Robert L. The Torpedo Position Interferometer. Inyokern , Calif. , U.S. Naval Ordnance
Test Station, November 17 , 1949. (NAVORD Report 1217. )

65. Plesset, Milton S. , and Philip A. Shaffer, Jr. “Drag in Cavitating Flow , ” in Reviews of Modern
Physics, Vol. 20, No. 1 ( 1948 ), pp . 228-31 .
66. Prandt), L., and O. G. Tietjens. Applied Hydro- and Aeromechanics. New York, N.Y. ,
McGraw-Hill Book Co., 1934. 311 pp .

67. Properties and Applications of the Freon Fluorocarbons. Wilmington , Del . , E. I. DuPont de
Nemours and Co. , 1966. (Freon Technical Bulletin B-2 . )

68. Richardson, E. G. “ The Impact of a Solid on a Liquid Surface ,” in Proceedings of the Physical
Society of London, Vol. 61 , Part 4 ( 1948 ) , pp . 352-66 .
69. Richter, Victor J. “NOL's Hydroballistics Laboratory ,” in Naval Research Reviews , Vol . 18 ,
No. 7 ( 1965 ), pp . 23-25 .
70. Rosenhead, L. “ Vortex Systems in Wakes,” in Advances in Applied Mechanics. Vol. III. New
York, N.Y., Academic Press , 1953. Pp. 185-95 .
71. Ross, James, and C. E. Bowers . Laboratory Surface Wave Equipment. Minneapolis , Minn. ,
University of Minnesota, 1953. (St. Anthony Falls Hydraulic Laboratory Project Report
No. 38. )

72. Rouse, Hunter. Cavitation and Pressure Distribution : Head Forms at Angles of Yaw. Iowa City ,
Iowa, State University of Iowa. 1962. (Studies in Engineering Bulletin 42. )
73. Rouse, Hunter , and John S. McNown . Cavitation and Pressure Distribution : Head Forms at
Zero Angle of Yaw. Iowa City , Iowa, State University of Iowa, 1948. (Studies in
Engineering Bulletin 32. )

339
74. Schlichting, Dr. Hermann. Boundary Layer Theory. 6th ed. New York, N.Y., McGraw-Hill
Book Co., 1968. 744 pp .

75. Seigel , A. E. The Hydroballistic Facility at NOL. White Oak, Md. , U.S. Naval Ordnance
Laboratory , August 1 , 1966. (NOL Technical Report 66-125 . )

76. Shaftan , Kenneth , and Dean Hawley. Photographic Instrumentation : Techniques, Equipment,
Applications. Redondo Beach , Calif., Society of Photographic Instrumentation Engineers,
1962 , 330 pp .

77. Smith , H. D. The Hydroballistics Laboratory at NOTS . China Lake , Calif. , U.S. Naval Ordnance
Test Station , April 1966. (NAVWEPS Report 9035. )

78. Smith, J. A. Design and Construction of the Variable -Angle Launcher. China Lake, Calif. , U.S.
Naval Ordnance Test Station, January 16 , 1952. (NAVORD Report 1327. )
79. Stanton, J. S., and W. G. Weigel. Investigation of Certain Problems of High-Speed Photography
Under Water. Inyokern , Calif. , U.S. Naval Ordnance Test Station, January 20 , 1949 .
(NAVORD Report 1063. )
80. Stubstad, G. W. Pure Pitching Motion of the Basic Finner. China Lake, Calif. , U.S. Naval
Ordnance Test Station, October 1 , 1958. (NAVORD Report 6417. )
81. Stubstad , G. W., and J. G. Waugh. Drag Coefficient of Fully Wetted Basic Finner Missile . China
Lake , Calif ., U.S. Naval Ordnance Test Station , June 7 , 1956. (NAVORD Report 5269. )
82. Swanson , W. M. , and J. P. O'Neill. The Stability of an Air -Maintained Cavity Behind a
Stationary Object in Flowing Water. Pasadena, Calif ., California Institute of Technology,
September 5 , 1951. (Hydrodynamics Laboratory Memorandum Report No. M-24.3 . )

83. Szebehely, Victor G., and Michel K. Ochi. “ Hydrodynamic Impact and Water Entry ,” in
Applied Mechanics Surveys. Washington , Spartan Books , 1966. Pp . 951-57 .
84. Tate , Dr. Manford B., compiler. Proceedings of the Naval Ordnance Laboratory Water Entry
Symposium. White Oak, Md., U. S. Naval Ordnance Laboratory, December 1 , 1955 .
(NAVORD Report 4180. )
85. Trilling, Leon. “ The Impact of aa Body on a Water Surface at an Arbitrary Angle,” in Journal of
Applied Physics , Vol. 21 , No. 2 ( 1950 ), pp . 161-70 .
86. von Karmán , Th. The Impact of Seaplane Floats During Landing. Washington , National
Advisory Committee for Aeronatics, 1929. (Technical Note No. 321. )
87. Waid, Robert L. “ Cavitation Research Capabilities of the Lockheed Underwater Missile
Facility ,” in Cavitation Research Facilities and Techniques. New York, N.Y. , American
Society of Mechanical Engineers, 1964. Pp . 127-31 .

88. Waugh , J. G. Underwater Launch and Trajectory Modeling Scaling Techniques . China Lake,
Calif., U.S. Naval Ordnance Test Station, March 15 , 1960. (NAVORD Report 7029. )
89 . “ The Variable-Angle Variable -Pressure Launching Tank , ” in Cavitation Research
Facilities and Techniques. New York , N.Y. , American Society of Mechanical Engineers ,
1964. Pp . 118-26 .
340
 .

90 . “Water-Entry Study with Two Liquids,” in Proceedings of the Seventh Underwater

Ballistics Conference. Part I, Unclassified Sessions. Washington , Office of Naval Research ,
1948. Pp . 43-51 .
91 . Water- Entry Whip and Deceleration of Eight Full-Scale Torpedo Models with Ogive
and Spherogive Heads. Inyokern , Calif., U.S. Naval Ordnance Test Station, March 2 ,
1950. (NAVORD Report 1223. )

92. Waugh, J. G. , and A. T. Ellis. “Fluid-Free-Surface Proximity Effect on a Sphere Vertically

Accelerated From Rest , ” in Journal of Hydronautics, Vol. 3 , No. 4 ( 1969 ) , pp. 175-9 .
>
93 . “ The Variable-Atmosphere Wave Tank , ” in Cavitation Research Facilities and
Techniques. New York , N.Y., American Society of Mechanical Engineers, 1964. Pp.
114-17 .

94. Waugh, J. G., A. T. Ellis , and S. B. Mellsen . “ Techniques for Metric Photography,” in Journal
of the Society of Motion Picture and Television Engineers , Vol. 75 , No. 1 (1966),pp . 2-6 .
95. Waugh , John G. Underwater Trajectories of Fourteen Full-Scale Unpowered Torpedo Models
with Ogive, Spherogive, Plate -Ogive, and Modified Cone Heads . China Lake, Calif. , U.S.
Naval Ordnance Test Station, February 5 , 1953. (NAVORD Report 2012. )
96 . “ Water-Entry Pitch Modeling,” in Journal of Hydronautics, Vol. 2 , No. 2 ( 1968 ) ,
pp . 87-92 .
97 . Water-Entry Whip and Deceleration of Four Full- Scale Torpedo Models with
Modified Cone Heads . China Lake , Calif. , U.S. Naval Ordnance Test Station, October 9 ,
1950. (NAVORD Report 1262. )

98. Waugh , John G. , and Raymond W. Ager. Water-Entry Whip and Deceleration of Six Full-Scale
Torpedo Models with Ogive and Plate-Ogive Heads . China Lake, Calif. , U.S. Naval
Ordnance Test Station, April 20 , 1951. (NAVORD Report 1308. )
99. Waugh , J. G., G. G. Mosteller, and Raymond W. Ager. Effect of Underpressure of the Water
Entry Whip of 2-Inch-Diameter Missiles with Vented Disk-Cylinder Heads . China Lake,
Calif., U.S. Naval Ordnance Test Station , June 28 , 1954. (NAVORD 3342. )
100 . Effect of Underpressure on the Water-Entry Whip of a 22.4-Inch-Diameter Missile
with Plate-Cylinder Head. China Lake, Calif., U.S. Naval Ordnance Test Station ,
November 3 , 1952. (NAVORD Report 1995. )
101. Waugh , John G., and G. W. Stubstad . Water-Entry Cavity Modeling : Part 1 , Vertical Cavities .
China Lake, Calif., U.S. Naval Ordnance Test Station , October 8 , 1956. (NAVORD
Report 5365 , Part 1. )
102 . Water-Entry Cavity Modeling : Part 2. Oblique Cavities . China Lake, Calif. , U.S.
Naval Ordnance Test Station, December 4, 1957. (NAVORD Report 5365 , Part 2. )
103 . Water Exit Behavior of Missiles : Part 1 , Preliminary Studies . China Lake , Calif. , U.S.
Naval Ordnance Test Station , May 11 , 1961. (NAVWEPS Report 7735 , Part 1. )

341
104 . Water-Exit Behavior of Missiles : Part 2 , Effect of Cavitation and Exit Angle. China
Lake , Calif. , U.S. Naval Ordnance Test Station , April 1964. (NAVWEPS Report 7735 ,
Part 2. )
105 . Water-Exit Behavior of Missiles : Part 3. Macroscopic Effects of the Hydrodynamic
Flow Pattern . China Lake, Calif. , U.S. Naval Ordnance Test Station, April 1964.
(NAVWEPS Report 7735 , Part 3. )

106. Wayland, Harold. Scale Factors in Water Entry. Inyokern, Calif., U.S. Naval Ordnance Test
Station, April 18 , 1947. (NAVORD Report 978. )
107. Wayland, Harold, and Frank G. White. Boundary Layer Effects on Spinning Spheres . Inyokern,
Calif., U.S. Naval Ordnance Test Station, October 25 , 1949. (NAVORD Report 1190. )
108. White, Frank G. Photographic Studies of Splash in Vertical and Oblique Water Entry of
Spheres. Inyokern , Calif., U.S. Naval Ordnance Test Station, April 24, 1950. (NAVORD
Report 1228. )

109. Wiegel , Robert L. Oceanographical Engineering. Englewood Cliffs, N.J., Prentice-Hall , Inc. ,
1964.532 pp.

110. Wilcox, G. M. , and J. G. Waugh . Water-Entry Study of the Mk 25 Torpedo with 3.5 -Calibre
70 -Degree Spherogive Head. Pasadena, Calif. , California Institute of Technology, May
1953. (Hydrodynamics Laboratory Report No. E-12.12 . )
111. Wilcox , Genevieve M. The Effects of Atmospheric Pressures upon the Water Entry Behavior of
a Missile Having a Hemispherical Nose and Flared-Cone Tail. Pasadena, Calif. , California
Institute of Technology, July 15 , 1952. (Hydrodynamics Laboratory Memorandum
Report EM- 12.8 . )

112. Worthington , A. M. A Study of Splashes. New York, N.Y. , the Macmillan Co. , 1963 .
113. Worthington , A. M., and R. S. Cole. “ Impact with a Liquid Surface Studied by the Aid of
Instantaneous Photography ,” in Philosophical Transactions of the Royal Society
(London ) , Vol. 194A ( 1900 ) , pp . 175-200 .

342
 SUBJECT INDEX

Alden Research Laboratories (ARL) , 323 cavity : continued

ARL (see Alden Research Laboratories) closure, 4,13,38-46 ,
asymmetric cavity shedding, 46 54,149-164,
asymmetric splash , 44 257-263
attack sensitivity and deceleration collapse, 328
graph , . 75,76 collapse and rebound , 242,243
corkscrew, 254
down -jet, 46
flow noise, 226,232
formation and growth, 13,54-55 ,
226-233,328
Basic Finner Missile, 244-247 , missile orientation, 38
301,302 momentum , 27
boundary layer : motion, types of, . 252
flow , 131
mouth, 38,151,156,
transition, 304 159,162-165
turbulence, 123-140, necking , incipient water entry, 38,258,263
253-255
parameters, curves of, 152-155 ,
broaching trajectories, . 33 159-160 ,
bubble networks , 312 164-170,175-177
bulk modulus of elasticity , 5,60 pressure , . 52-54
re-entrant jets, 38,46,226,
232-234
California Institute of Technology role of, in missile performance,
(CIT) , . 267 separation , .25-37,71,72 ,
cavitating plate-cylinder (oil-water) , . 239-241 97,131,189-202 ,
cavitating sphere (oil-water), 235-238 251-263,329
cavitating vortices , .194,195,225 , shape . 184,329
226,232,234 stripping, .
18,46,47 ,
cavitation -number changes, 132,136
sensitivity to : trailing vortices, 234
plate-cylinder missiles, . 160 up -jet, .
46,51,52,55
cavitation -number, definitions of, .
2,7 cavity and splash growth , 19-22,45-48
cavity : cavity closure :
attrition, 63,328 38
cavity-lip flow ,
behavior, influencing mechanisms contours , 43,44
on , 13,327 deep and surface, .42,43,151-155 ,
circular striations, 23 257-263

343
 cavity closure : continued drag coefficients : continued
gas flow , . .
4,38,145 underwater , . 140
atmospheric density, 38,43-46 dynamic similarity, modeling
atmospheric pressure, .
38,43-46 constraints for 3

slippage, • .
14,50
splash flow , 38
water entry phenomena, 38-50 earth-entry studies , exploratory, 178-185
cavity -flow phase, . 25 entry cavity, effects of gravity , 257-265
cavity growth : environmental effects , modeling
atmospheric pressure, effect of, . 98-100 missile behavior , 327-329
unvented hemisphere , .
97,107-109 equal entry conditions, identical
unvented plate cylinder, . . 88-105 cavities , 104,105
vented hemisphere , . .
97,110,111 erratic trajectory , 29
vented -plate cylinders, 88-97 experimental techniques , . . 303-313
cavity -lip , splash and gas flows,, 38
cavity sensitivity :
Froude and cavitation number FAL (see fixed -angle launcher)
scaling, 145-177 fiducial marks, . 301,302 ,
gas-density scaling, 145-177 308-311
cavity striae, 23 fixed -angle launcher ( FAL),
cavity -tripping and stabilizing description of, . .. 284,285
mechanisms, .
27,129 , fixed -angle launcher site, hydrophone
130,189 locations , 79
cavity-wall clearing, gas density, flare camera , . 289
effect of, . .
162 flexure number, 3,5
CIT (see California Institute of flow and cavity formations, missile
Technology) water entry , 24-37
CIT modeling system : flow -marking techniques,
cameras, fiducial marks, and lights, 309,310 photographic, 312,313
CG balance , . . 305 flow pattern , studies of:
model fiducial marks, 309 glycerine, 52

2
model roughness cap , 304 petrol, . . 52
models , 305 zinc chloride solution , . 52
torsional inertia balance , . 306 flow pattern turbulence :
controlled atmosphere launching tank,. 273-284 asymmetric, . . 328
controlled atmosphere tank, ARL, 323-325 boundary layer, . 328
fluid medium , compressibility of, 18
frangible nose cap , missile water
entry, 61,328
data-analyzing system : frictional force scaling, 7.
analogous projectors, 282 Froude and Reynolds forces, 1

position indicators, . 284

deceleration, drag:
at water entry , 69,74-79 gas bubble expansion phase, 6
underwater , . . .
132-135 , gas density:
140,144 , scaling, 6,8,10,11 ,
245-247 113-122,145-177
deceleration -water-entry velocity sensitivity of cavity to , 145-177
regression equations, . 77 variations, sensitivity to :
diffusion of vorticity, . . . 223-229 plate-cylinder missile, . 160-162 ,
1
diving trajectories, . . 33-35 171,172
down -jet, cavity closure, 46 truncated-cone missile, 159-170
drag -coefficients: gas expansion, adiabatic, . 6-11
at water entry, 74,77,78 gas flow pressure reduction, 38

344
gas liquefaction, model-system , launcher, description of: continued
scaled, . 9 VAL, 285-289
gas metering, 303 launching facilities, prototype, NUC, 284-302
gravitational force scaling, Froude, .
1,2 liquid interface tank, 296,297
gravitationless field, water-entry Lockheed Underwater Missile
cavity, . 259-265 Facility ( LUMF) , .314,318-320
gravity, effects of, water-entry cavity, . 257-265 LUMF (see Lockheed Underwater
Missile Facility )

HBR (see hydroballistics range) missile flexure, 61,62

hemisphere missile, 146 missile head shape geometry ,
.
3,70-72
hemisphere missile cavities, 149-155 missile motion perturbation , water
high - speed recorder camera, exit, 328
trajectory, 278-282 missile underwater trajectory, 27,29,30,
high-velocity regime, study of, 328 34-38,40,41
hydroballistics modeling : missile underwater trajectory studies, 79-83 ,
changing status of, 327-329 123-144
advancements and refinements, missile water-entry in-cavity behavior,
need for , 329 155-172
.

oblique, . .

dimensional analysis , 2 vertical , 149-155

techniques, underwater launch Mitchell high-speed motion-picture
modeling, 6-11 camera, 291
theory, 1-11
.
model and cavity behavior, turbulent
constraints for dynamir similarity, 3
boundary layer, 123-131
parameters , . 1-11 model CG balances, NUC and CIT, . 305
hydroballistics modeling, model inertia balances, NUC and
experimental : CIT, 305,306
early investigators of, 51-63 model preparation , 305-309
evolution of, 51-67 model water -entry behavior, boundary
facilities and equipment for, . 267-302 layer flow , 123
facilities design for, . . 314-325 modeling gas : . 305-309
techniques and design criteria for, . 303-313 dibromotetrafluoroethane, . .
.114,149,304
hydroballistics range (HBR) , . 299-302 dichlorodifluoromethane, .
149,258
hydrodynamic whip , . 27 octafluorocyclobutane, . 9
hydrophobic coating, 257,308 requirements of, 8
hydrostatic force, 38 sources of, 9
hydrostatic pressures , .
7,46 , modeling status :
257,258 environmental effects on missile
behavior, . 328
hydroballistics problems, 329
instrumentation : present day knowledge, . 327
interferometer, torpedo position , 289 327,328
.
scaling conditions,
prototype launching, 289 modeling studies :
underwater trajectory, 123-144
water-entry cavity , 145-177
launcher, description of : water- entry pitch, 113-122
catapult, .
295,296 modeling theory :
centrifugal, . 275-278 constraints , 3-5
FAL, 284,285 1-5
hydroballistics, .

impulse, .
296,297 underwater-launch, . 6-11
magnetic, . 321-323 models, NUC and CIT . 305-309
plunger, .
324,325 motion-picture camera, high-speed,
pneumatic piston, 269 Mitchell, . 291

345
Naval Ordnance Laboratory (NOL) , 319 plate-cylinder missile cavities, . . 160-166,
Naval Undersea Research and Develop 171-177
ment Center (NUC), 267 polymethylsiloxane, DC -500,
NOL (see Naval Ordnance Laboratory) (Dow Corning Corp.), 257,308
NOL hydroballistics facility, 303,314 prototype and model experiments :
319-321 exploratory earth -entry studies, 178-185
NUC (see Naval Undersea Research and underwater trajectory modeling, . 123-144
Development Center) water-entry cavity modeling, 145-177
NUC prototype launching facilities, 284-294 water -entry pitch modeling, . 113-122
water-exit studies, 186-216
whip and deceleration at water
oblique cavity, experimental results, 84,85 , entry, 69-83
88-111 , whip and underpressure , 84-112
130-143,155-177
open launching tank , 294-296
optical comparators , 273 re- entrant jet formation and
optical whip recorder, . .270-272,308 behavior, . . .222,226,232 ,
234-241,328
Reynolds and Mach numbers, 2
Reynolds effects, 1
parameters, hydroballistics modeling Reynolds number :
theory, 1-11 boundary -layer thickness, 123-125
peak pressure (elastic) of water on effects in experimental studies, 188,189,195 ,
missile, 23 222-241,244-247 ,
phenomena, water -entry, 13-50 253-256
cavity closure, 17,38-46 model and prototype velocities, . 1
closed cavity, . . .
13,46-50 ricochet characteristics, modeling of, 328
factors influencing water entry . 14 ricochet studies, spheres, . . 53,56
flow formation 24-37 rotating-disk camera , . 270
impact, 18
open cavity, . . 37,38
18-23 scaling :
splash growth , .
underpressure, . . 27,33,37,38 cavitation -number, . 113-177
whip, 27,33,37 gas-density, .. .
113-122 ,
phenomenological studies : 145-177
effects of gravity on the entry gravitational and viscous, 1
cavity, . .
257-265 one -to -one Froude, . 113-177
wake effects, 244-250 prototype and model water -entry , 113-177
wake formation, 222-243 scaling conditions, modeling, . 327,328
water - entry cavities of spinning scaling constraints :
spheres, 251-256 fluid temperature , 3,6
water- entry splash contours, 217-221 fluid (water ) density , 3,6
photography : gas density , . . 4
color film, 312 gas pressure , 7

developing, . 312 gas specific heat ratio , . 6

emulsions, 312 gas temperature, . 6
flow -marking techniques of, .
312,313 gas volume, . 6

metric, . 311,312 scaling elastic forces, 23

shadowgraphs, 217,312 scaling laws, . 1-5
pitch-velocity : scaling parameters :
during water entry, . 74 cavitation number, . . 2-11
time functions, 86,87 flexure number , .
3,5
time plot, 74,87 fluid density , 3

346
scaling parameters : continued trajectory: continued
fluid temperature , 3. recording system of, 278-282 ,
Froude number , 2-11 291-294,299-301
gas-density coefficient, 2 transient flow , 328
gas specific heat ratio , . 6-11 truncated -cone missile , 147,148

NN
.

Mach number, . 2 truncated -cone missile cavities , . 155-170

Reynolds number, . 2 turbulence : 123-132,189 ,
Weber number, .
2,4 195,222-226 ,
sectional tank , . .
298,299 251-256,328
selection of working gases, . 9-11 induction of, 125,126 ,
shear layer and cavitation, 232 128-133,189,304
simultaneous whip and side view
camera data, 272,273
smooth and roughened models,
cavities of, , . .131-139,143 .27,33,37,38 ,
underpressure, .
special experimental techniques and 60,84,87,88,90,97,160
hydroballistics facilities: underwater cameras , 278-282
experimental techniques, 303-313
underwater-launch modeling, 6-11
facilities and equipment, 267-302 underwater trajectory :. 4,27-41 ,
hydroballistics facilities design , 314-325
79-83,123-144
sphere and plate-cylinder missile, underwater trajectory modeling,
oil-water flow studies of, . 222-243 effects of :
sphere ricochet studies, 53,56 atmospheric pressure , 137-144
spinning spheres, water-entry cavities, . 251-256 entry angle of attack , . 130-138
splash characteristics , water- and entry velocity , . . 139-144
sand- entry, 179-183 turbulent boundary layer, . 123-128 ,
splash contours, water-entry, 217-221 131,140
splash formations: unstable cavity :
experimental studies of, 38,44-46 , separation of, 107
51-54 stabilizing the, . 189,328
water-entry, effect of gravity on, 257-263 unreproducibility of, .
97,108
splash growth, .. 18-23 46-55
up -jet, cavity closure,
stable cavity, vented spherical sector, . 97,110,111
.

stroboscopic flash lamps, Edgerton

type , .
188,269,281 , VAL (see variable-angle launcher)
282,309,310,312,315 VAL camera stations , 290
surface tension , 4,5,37,38,43 , VAL hydrophone array , 290
45,60,61,97,112 , Van Slyke-Neill manometric gas
136-140,257 , analysis apparatus, 186,303
258,263,328 variable-angle launcher (VAL) ,
symmetrical cavity, stable, . 97,109 description of, . 285-289
variable -angle variable -pressure
launching tank (VAVP) :
tank design developments, 317,318 oblique water-entry , 267-269
test facilities and equipment, vertical water-exit, . 267-269
CIT , 273-284 , variable -atmosphere tank (VAT) , .314,317,318
305-311,320-323 variable-atmosphere wave tank , .
.309,320-323
NUC , 267-273 , VAT (see variable-atmosphere tank)
284-302,305-308 , VAVP ( see variable-angle variable
317,318 pressure launching tank)
time similitude, trajectory , 3
vertical cavity, experimental results , 149-155
trailing vortices , cavity , 234 vertical drop tanks, 297,298
trajectory: vertical entry of sphere, . 44-46
factors affecting, 27-38 viscous force scaling, Reynolds , .
1,2

347
vortex : water-entry, studies of: continued
cavitation shedding, 194,195 , Birkhoff, Garrett and
225,226,232 Caywood , T.E. , 55,57
generation and shedding, . 194,195 , Birkhoff, Garrett and
225-228,234 , Zarantonello , E.H. , . 55
244-250,328 Davies, R.M. , 53
vortices, cavity trailing, 234
Gilbarg, David and Anderson,
vorticity formation and behavior, .223-226,234 R.A. , 53,54
Harvey , E.N. , . 54
Knapp, R.T. , . 62
Mason, Max and Slichter, L.B. , 58-67
May, Albert, 54
McMillen , J.H. , 54
wake : McMillen, J.H. and May, Albert , 54
effects of, 244-250 Richardson, E.G. , 52-56
formation and behavior of, . 222-243 Slichter , L.B. , 58-67,84
water deaeration, 186,303 Worthington , A.M., 51,52
water, elastic peak pressure on water-entry whip and deceleration, . 27,37,69-78
.

missile, 23 water-exit studies : . 186-216

water filtration , 269,303 , composite missile , 186,187 ,
316,321 200,215,216
water , light transmission of, 315 hemisphere missile, . 186,187 ,
water -entry cavities: 189-198,203-211
of spinning spheres, . 251-256 summary ,, 200,202
venting effect , . 84-112 truncated-cone missile, 186,187 ,
water-entry cavity modeling: 145-177 198-200,212-214
oblique, 145-149 , Weber number , 2
155-177 wetted flow , 222,247
vertical , 145-157 , wetted sphere, vortex shedding, 227,228
164,166 whip and deceleration at water entry, 69-78
water-entry , full scale whip and ballistic parameters, . 73
deceleration , 69-83 experimental conditions, . . 72
water -entry, oblique , missile, selection of head shapes, . O 71
studies of : hip and underpressure, effects of: 84-112
flat plate , . 85 atmospheric pressure, 90,97,112
hemisphere and vented hemisphere surface tension, 97,112
heads, 97,107-111 underpressure on whip, 27,33 ,
plate -cylinder and vented -plate 84-86,88,90,97
heads , 84-106,112 venting, 84-112
water-entry phenomena : 13-50 whip -angle of attack, 75-77
cavity closure , 17,38-46 whip-attack regression equations , 75,88
closed cavity, 13,46-50 whip, definition of, . 86,87
factors influencing water entry, 14 whip dependence on water-entry
flow formation , 24-37 velocity, .. 69,70
impact , 18 whip, earth entry studies, 184,185
open cavity,. 37,38 whip measurements :
splash growth , 18-23 flare camera , 74,289-291
underpressure , 27,33,37,38 optical whip recorder, .
.270-273,308
whip , . 27,33,37 working gases, partial-scaling,
water-entry pitch modeling, 113-122 selection of, 9-11
water -entry splash contours, . 217-221
water-entry, studies of: 51-84
Birkhoff, Garrett, 55 yaw , effects of, water entry, 86,87

348
 1

C.
chien
De # .
ce
 UNIVERSITY OF ILLINOIS - URBANA

HY3.
62 51AL
DROB 5HSTICS MODELIC0NG
W3LI 01$WASHINGTON

46
3 0112 0083462