UNITED STATES NAVAL ACADEEIY
Annapolis, Maryland
ENGINEERING DEPARTMENT
Report E-68 - 5
THE CONCEPTUAL DESIGNOF A HIGH
PERFORMANCE TOWING TANK FOR THEU. S. NAVAL ACADEMY
Roger H.
Compton1
2
Thomas R. Dyer
Bruce
Johnson3
Revised and Reissued
18 December 1968
Presented to
15th American
Towing Tank Conference
Ottawa, 25 June 1968
Distribution of this
document is unlimited
Assistant Professor,
Naval Engineering Division
Lieutenant, USNR,
Project Officer
Associate Professor,
Director, Hydromechanics
TABLE OF CONTENTS Summary Introduction I. Background A. General Requirements Present Capabilities Present Limitations HydroacoustiC Requirement
B. Investigation of Existing. Facilities 1. Rail Systems
2. Carriage Configuration
3. Carriage Drive Systems
ii. Laboratory Plans 12
Hydr.omeChafliCS Laboratory 12.
The 380' Towing Tank . 13.
1.. Tank Length
.2. Tank.Depth and Ceiling Height
3. Tank Width ..
4. Architectural Features
III. Conventional Equipment
A. Conventional Carriage
Configuration
Speed. Control
3.. Wheels
4. Fast Return System
B. Seakeeping Capabilities
Wavemaker
Wave_damping Beach
C. Model Handling and Technical Support Instrument Modules Drydock. 3 Technical Support D. Rail System Rails Rail Supports 3 28 28 32 35 36
IV. Special Capabilities
39
A. High Performance Carriage and
Drive 39
Configuration
Velocity and Acceleration Control
Carriage Drive System
B. Data Transmission and Reduction Systems
45
C. Acoustic Provisions
47
V. Developmental Engineering
A.. Developmental Engineering Contracts
Faculty Studies . 49 Design Review 50 References . 53 Figure 1 54 Figure 2 55 Figure 3 56 Figure 4 57 Figure 5 . .
Appendix A Statistical Analysis of Towing
Tank Data
A-i
Figure A-I
SUNMARY
The proposed hydromechafliCs laboratory will include: an existing circulating water channel; the existing 5'
tank, extended to 120'; an existing ship stability and model ballasting tank; a new cavitation tunnel, and the 3O' high performance towing tank.
The proposed tank is designed for a maximum steady state speed of 30 knots, with a steady state run of five seconds duration. This speed is required to achieve Reynolds numbers of interest in the study of near field flow noise, while the length of run is required for the gathering of sufficient statistical data to allow
reasonable confidence levels. The required sample time
for low frequency statistical data is long; and casts doubt on the utility of testing ship models in random seas.
The tank is to be construdted with a width of 26
feet and a water depth of 16 feet, in order to minimize blockage and shallow water effects within the bounds of
architectural requirements. Other architectural
considerations will include water treatment, water
tem-perature control, air conditioning, underwater observation
and photography provisions, corrosion control, and a drydock and carriage rigging area.
The tank will be equipped with the conventional
equipment now found in most towing tanks. A large (16'x20') fifteen knot carriage is planned. This carriage will
incorporate an 11' square central well, and will provide
ample deck area for groups of midshipmen. Since this carriage
and seakeeping research, it will be of heavy box, girder
construction with unobstructed viewing, and will ride
on vibration damped steel wheels. A precise speed
control, such as a frequency based drive powering four d.c. or synchronous a.c. motors, will be required. A
system to allow rapid return runs, while Smoothing the
free surface, is desirable.
A wavernaker able to generate three foot high regular waves, thirty feet long, will be installed. It will be
capable of random sea demonstrations and operation with varying water depths. Th& rigid beach will be 'hinged and segmented to permit unobstructed runs, and to offer
various beach profiles for optimum damping and the study
of near shore wave phenomena. Experiments are also underway with an air bubble curtain beach.
Since a major purpose of the tank is the study of
flow noise, acoustics must play a large role in its
design. Noise due to the carriage, models, and towing struts
must be considered. Anechoic material must' line the tank, perhaps serving also as a wave and current damper and covering. 'Anechoic, noise free,
conditions are not
required, so long as an acceptable signal-to-noise 'ratio is achieved at the model.
The rails will be cantilevered from a structure independent of the tank, isolating the wheel noise 'and reducing the required span. The tank structure will be
isolated in similar manner from all other structures
within the 'building.
A high speed carriage will be specially designed for acoustics tests. Design Parameters will include velocity, acceleration, time ra'te of change of acceleration, and
vibrational levels'. The carriage will be of lightweight damped structure ridingon slick rubber tires-or air
11
bearings. Several power systems are under consideration: twin continuous loops of steel tape or cable, pulling
the carriage from external motors and clutches; very high
power to weight ratio induction motors aboard the
carriage a linear induction motor; or a catapult with
sustaining motors aboard the carriage.
Data transmission systems such as telemetry, trailing wires, inductive trolleys, or on-board tape recorders,
will need to be considered.. Vital in this consideration
will be the dynamic range of the equipment relative to that of the expected signals. The data reduction function-must consider real time statistical analysis, optimization
of the number of frequency resolution elements,
pre-processing for a digital computer, and the interfacing with a time sharing computer system.
A CKNOWLEDGEMEN TS
The authors wish to thank Captain
R. W. King, USN,
Associate Professor
R. F. Latham, and Associate
Professor P. R. Van Mater of
the Engineering Department
of the U. S. Naval Academy, for
their interest, their
support, and their ideas.
Dr. Karl Remmers of Keinpf
and
Rernmers, and Mr. W. F. Brownell,
Jr. and Mr. V. W. Miller
of the Naval Research and
Development Center have also
been of great help,
as have many people too
numerous to
name who have shared freely of their ideas
and experiences.
To Professor Thomas C. Gilitner
very special thanks are
in order, for his instrumental
role in making this entire
project possible.
Since the original issue
of this paper, the authors
have received numerous
comments from the delegates to the
15th American Towing Tank Conference.
Many of these ideas
will be incorporated into the
final design, and have
resulted in changes to this
paper..
To all the delegates,
and
especially toMr. S. T. Mathews
and,Dr.. D
Gospodnetic,
we extend grateful acknowledgement for
their generous
sharing of time and ideas,
and for an exceptionally fine
INTRODUCTION
Naval Architecture and Marine Engineering have long
been taught at the U. S. Naval Academy. Since 1956 an
85' towing tank has been in operation, both as a teaching
laboratory and. increasingly as a research tool for
midshipmen and faculty. As part .of the recently approved
modernization of the Naval Academy facilities, a new Engineering Department building is now being planned.
Included in this building will be a modern hydromechanics
laboratory centered about a 380' towing tank with hydracouStic capabilities.
The Naval Academy must educate and train a man for the Navy of the future, and for the technical problems he will encounter. The proposed tank will support this
mission by providing a dynamic atmosphere in which to
educate informed and inquisitive junior officers under the supervision of a first-rate faculty. As the Naval.
Engineering curriculum evolves, instructional use of the hydromechaflics facilities will grow above the already significant levels. But in addition to instruction, the
tank will be used for midshipman research, especially
under the Trident Scholar program. It will serve the vital function of attracting capable midshipmen and faculty.
The tank will also contribute greatly to research in hydro-acouSticS, ocean engineering, hydrodynamics, and naval architecture. Such research will promote faculty
involvement with current naval problems, will allow faculty advancement in the academic community, and will maintain the faculty's technical proficiency. The tank
itself should serve as a source of new ideas for towing tanks of the future.
The faculty of the Naval Academy has thoroughly
investigated the free world's existing towing tanks,
visiting most of the modern tanks in the United States, Europe and Japan. The successful features of
these tanks
have been examined for applicability in the proposed Naval Academy tank, and have led to a conceptual design
SECTION ,1
I. BACKGROUND,
A. General Requirements
1. Present Capabilities
The present 5 foot towing tank at the Naval.
Academy has received increasing use in recent years, both
in a teaching and research capacity. The increased
emphasis on independent student projects and the Trident
Scholar program frequently results in experimental projects
which exceed the capabilities of the present facility. iiidshipmen and their facu.lty advisOrs are naturally
interested in âuch areas as drag reduction, 'the interaction
between turbulence and flow noise, the hydrodynamics of various forms oI advanced marine vehicles from hydrofoils to winged submersibles, and various exotic projc,t in ocean
engineering, naval architecture, and marine engineering.
An attempt has been made to' upgrade the h'rdromechafliCS laboratOry in terms of modern electronic
instrumentation to demonstrate how on-line data acquisition
and analysis cn be 'applied to realistic 'engineering
problems. A new 0.01% regulatiOn speed control
system is under contract. It will automaticallY control a
new 8channel
digital 'data acquisitiOn which is tieddirectly to a 'time-shared computer'SySteflI. 1The data
acquIsition system conSistS of eight scalers controlled by
a single time base. The present analog devices for
statistical anaIysi'Of turbulence and acoustic data will
soon be joinedby
an on_l'inè,"real'time, signal processor. The facUlty is engaged in a major effort iti theexperi-mental statistical analysis of coherent signatures buried
in noise, especially as related to ASW.
Itis important
that the future officers, who will be directingvast
research and development programs in these areas, have experience in an experimental labàratory
whe±'e the' complex ASW problems can be isolated into speciflO projects
studies under controlled conditions. It is most effective
to learn first hand of the limitations and frustrations
of trying to solve a fleet problem through
research and
engineering efforts involving simplifying, models. In addition, the midshipmen and facul'y gain a creative
out-let for personal involvement in the. design, development,
test and evaluation of significant Navy-oriented
engineer-ing projects.
2. Present Limitations
Even with the improvements in instrumentation,
the existing towing tank
is subject to severe limitations.
These are largely imposed by its small size,
with resulting
lack of ability to achieve realistic. Reynolds Numbers,
to hold reasonable speeds for adequate steady state periods,
or to test models of adequate size.
It becomes necessary both in research 'and
instruction to examine isolated and
individual portions of the ship resistance problem.
Such examination, because of the difficulties of scaling and the limitations of the physical law of similitude, requires
testing of larger models at speeds more closely
related to
??full scale."
Much of the problem of model scaling is related to the modeling of a turbulent boundary layer to
correlate with that found on a full scale ship. This is impossible of course, Since Reynolds and Froude scaling
cannot be achieved simultaneously
experimental approximations may be obtained provided the
boundary layer remains turbulent, once stimulated.
The problems of achieving "natural transition"
in the boundary laye'r of ship models can best be analyzed
by looking at the influence of model size on the model
Reynolds Number at a given Froude Number.
Recalling that, for equal Froude Numbers
Rn
= V
1'fl , andV = V j
mit can be seen that
Rn = Vs
(L)3/2
tmr
V
Thus, a tank cross section and length which will enable testing of 12 foot ship models, as compared
to 3 foot models in the 5 foot tank, will increase the
model Reynolds Number by a factor of for the same Froude Number. For most6tests, this means increasing Reynolds
Numbers to the 10 range where turbulence stimulation is
quite easy, as compared to the l0 range, where effective stimulation is more difficult.
3. Hydroacoustic Requirement
For submerged models, Reynolds Number must be
increased by at least a factor of 20 compared to the present tank. (Speed by a factor of 5 and model length by a
factor of 4.) This conclusion is based on futile efforts
to make meaningful flow noise measurements in the
5'
tank. Because submarines cannot effectively "see" whatlies ahead at speeds greater than 20 knots because of the self-generated "flow noise", this is an especially important
problem area for the Navy.
As Patrick Leehey (1) ad others have noted, no experimental facility exists anhere in the world
to study flow noisearound submerged bodies under controlled
environmental conditions. Most flow noise facilities investigate internal flow in a pipe, where the radiative field is quite different from that found in external f low about a body. Buoyant. and gravity propelled
submerged bodies have been used with some success to study external flow in open lakes, but much remains to be
learned about the near-field "pseudo-sound" generated by
the interaction of the turbulent boundary layer with a
solid body. it is felt that a major cotitribution to
greater understanding of this phenomena could be made by
testing small well-instrumented sonar domes and other
objects at nearly fill scale Reynolds Numbers, at speeds
above 20 knots inthe controlled envIronment of a towing
basin designed to acoustic. specifications. The ambient
noise levels in all existing tanks preclude such testing.
Special care must therefore be taken to inéure. that the desired signal from the "pseudo-sound" is greater than
that of the acoustic radiation sources withinthe. new_=.
-tank. It must be clearly understood that anechoic, noise free, conditions are not required.; rather an acceptable
signal-to-noise xust be achieved at the model so that meaningful hydrophone data may be obtained..
4. Proposed Experimental Techniques
Qnce the, considerable technical problems of
achieving acceptably low noise and vibration levels in the tank are solved, several important new experimental
techniques can be used to study the problem of. "near-field
sound (as opposed to radiated sound, which will be .
Flush-mounted hot film anemometers, accelerometers,
wave-vector filters, and small hydrophones can be
mounted close together to try to separate the.influenCe
of the locally radiated sound, involving fluid momentum transfer on the microscopic scale, from "pseudo-sound",
which involves fluid momentum transfer on the macroscopic scale.
The wave-vector filter (2), under development at NSRDC, is sensitive to pressure fluctuations which
travel at the convection velocity of the turbulent eddies
in the boundary layer (this is what is meant by "pseudo-sound"). It is designed, however, to reject pressure fluctuations traveling at the local speed of sound in the
fluid (hence the term "filter"). Since acoustic
radiation is of the latter type, the wave-vector filter does not require anechoic conditions for successful flow noise measurements.
Because of the well established relationship between wall shear stress and convective heat transfer,
the hot film sensoris sensitive only to macroscoPiC,
momentum transfer, which in turn is related to the
instantaneous wall shear in the case of a flush-mounted
sensor. Acoustic radiation accompanied by no relative fluid motion produces no detectable signal in the
anemometer.
A. flush mounted hydrophOfle is sensitive to
both macrosCOPiC and microsCOPiC momentum transfer as well as to.vibratiOt1S in the structureto which it:is attached.. It is this complex sensitivity which makes hydrophone
measuremfltS difficult to interpret by themselves. Acce1eromete'S- are, of course, sensitive to model
By subtracting and analyzing the respective
signals of the four sensors in the amplitude, frequency, and time domains, it is hoped that greater insight can be gained into the. nature Of flow-noise. . By comparing
the results to similar experiments performed in air, it
could be determined whether the relative strengths of the macroscopic and microscopic momentum fields vary with
the medium according to similarity laws yet to be
determined. Until such measurements can be made,
significant differences of opinion will continue to exist as to the nature of the forcing function causing sonar
done vibrations.
B. Investigation of Existing Facilities,
In the summer and fall of 1966, Professor Johnson visited 26 towing tanks in the United States, Gerrnany,
Denmark, Holland, England, and apan, gathering design dataon basin construction, rail design, carriage
configuration and propulsion, instrumentation, and
wave-making capabilities. In observing both the.good
and bad features of towing tank design, the following innovations stood out and have been, considered in the conceptual
design of the Naval Academy Tank.
1. Rail Systems . . .. .
Many recently built
tanks.have.continuous.
welded rails, which eliminate the vibration and noise
associated with rail joints. So long as the rails are mounted directly on the sides, of the basin, however, even.
the quietest carriage (found at Feltham, England) transmits too much noise to the water for acoustic tests. The
cantilevered rail systems found at the Osaka University Tank, at one of the Wageningen tanks1 and at the NRc tank in Ottawa offers several advantages in addition to the'
possibility of good acoustic isolation. The carriage
spans less of the. tank width, and can consequently be made lighter; and the tank water level can be changed without disturbing the rail alignment.
2. Carriage Configuration
Many tanks favor a rectangular carriage with four sets of drive wheels. The center section is quite open so that interchangeable dynamomel.er and
instrumen-tation modules may be easily mounted as needed. The new carriages at Feltham, England; West Berlin; and the Ship
Research Institute, and the University Tank, in Tokyo all feature large central openings. The University of
Tokyo carriage is specifically designed for photo studies
of ship generated waves.
The Feltham. and Ship Research Institute
carriages are of tubular construction of a more open design than those at the U. S. Naval Ship Research and
Development Center. The Fukuoka, Japan, carriage is of
box girder construction, with a large open platform.
Aircraft-inspired monocoque construction is featured in
the Meguro (Tokyo) and new Berlin carriages.
The Feltham carriage is exceptionally vibration free.. This can be attributed to the.use of individual
drive units which are vibration isolated from the
carriage through a three-point rubber suspension system. The drive units make use of two-wheel bogeys with a
wheel spacing designed.to minimize the effects of rail
deflection between supports.
Rubber tires are used as drive wheels in a unique way on the Berlin carriage. Each rubber tire has a pivoted steel wheel on either side of it, so that
the relative carriage weight supported by either set
of wheels can be varied by changing the tire inflation. Precision seakeeping tests can be performed on the
steel wheels, and a softer ride can be utilized for high speeds.
3. Carriage Drive Systems
On Board Electric Motors
This is the most popular carriage drive system, with improvements being made in speed control through various feedback systems. One of the best
systems is found on the Feitham carriage, which has four
tachometer generators, mounted one to each drive motor.
As the carriage speed exceeds the range of the lower speed tachometers, they are disengaged through the use
of magnetic clutches. This is an excellent technique to give wide speed ranges. Another interesting speed
control is found on the Fukuoka carriage: a fast
responding tachometer generator analog control for
acceleration up to 99% of preset speed with takeover by a slotted disk digital device for very accurate steady
state speed control.
On Board Hydraulic
This is a very smoothly accelerating
propulsion device, but it is quite noisy. Preset steady
state speeds are difficult to attain if an accumulator
is used for acceleration. Better feedback controls for hydraulic drives are still needed.
On Board Water Jet Sustaining Engine
This has been proposed for the Wageningen high speed carriage asa minimum cost, high performance system, but its practicality is still tobe proven.
External Continuous Loop Steel Cable Drive Thisis the most popular external drive
motor can be either, electric or hydraulic with the
advantages and disadvantages previously, listed. The
carriage is made fast to a loop of cable pulled externally
by a drive drum located at one en4 of the tank. T1is,
drive makes possible the use of lightweight carriages for high acceleration rates with no on board drive vibrations; but cable stretch, due to e1asiCitY and untwisting, can cause serious velocity fluátuatiOn
problems and sometimes results in the slack side of the cable loop dragging in the water. One: posib1e improvement
would be the use of preformed non-rotating stainless cable
developed for ocean technology applications by U. S. Steel. External Continuous Loop Pretensioned,
Tape Drive
This scheme, utilizing relatively inelastic steel tape, was originated at M.I.T., and seems to offer
the best combination of low weight, quiet operation and
lack of cable elasticity induced speed oscillations.
Linear Induction Motor Drive
This is a' fascinating idea in which the rail acts as the "rotor" while the carriage is the "stator".
It is frequently suggested by those who would like to see
someone else try it first. It would represent a sizeable investment in research and development and might present a magnetic field problem for many types of sensors and
5ocjated cabling. This technique is currently being. developed in France for the French "Aerotràifl" by The:
Merlin-Germ Company of Grenoble, and for rapid
transit by the compagnie d'EriergtiqUe
Lineaire; itmar
ultimately become popular.Hydraulic or 'Pneumatic Catapult
This idea is under' 'development at both the
date. The problem of matching feedback control
sytems
between the catapult and the electric sustaining motorsrequires a great amount of additional development.
The
Berlin technique of using the arresting gear, to help recharge the catapult is very clever but mechanically complex.
h. Water Jet Catapult
This technique,in which a high energy jet impinges on a carriage mounted bucket,is used by
the
Langley, Virginia high speed test facility and has been
proposed for the Wageningen high speed tank.
This is a very economical system but it cannot maintain a constant
speed, is very noisy, and gets everything rather wet. It is spectacular to watch in operation.
j. Acceleration and Deceleration
Rails This novel concept is used on the small Meguro tank. An extra set of rails is located
outside the main rails in the acceleration and deceleration sections of the basin. An automotive tire
is mounted outboard of each of the carriage wheels to give greater
wheel traction for acceleration and deceleration.
The wheels are free spinning during the steady state portion
of the run.
SECTION II
II. LABORATORY PLANS
A. Hydromechanics Laboratory
The proposed hydromechanics
laboratory (Figure 1)
will include: the 3O foot towing tank; an existing
16" x 16" section, 75 hp, free surface circulating water channel; an existing 19? x 23' x 4' deep steel tank, currently used to study the static stability of large
new towing tank; a new cavitation tunnel,, proposed to
have a 12" x 12" test section with no free surface but
the capability of testing various hydrodynamic bodies and propellers; and the existing small towing tank. The
small towing tank which will be extended to 120' length now measures
85' x 6'.x
3.51
deep and is. of steelconstruction. The rail system, and perhaps the water
basin proper, will require replacement in the new building. The carriage, the speed control system, the wavemaker, the instrumentation, and the iodels from the small tank will be moved to the new building. It is felt that the utility
of a small tank, with easy and unobstructed model viewing, is a very great asset in the teachig of large groups of midshipmen. The tank has been especially useful in the
study of boundary layer additives which necessitate a complete change of tank water after their use, and in
testing small models such as yachts. If the water basin is replaced, it will be made a
1/3
scale model of the30'
tank, thus creating a "geosim" for blockage and scale effect studies.B. The 30' Towing Tank
The toiing tank will have a total usable run. of
30'; it will be 26' wide,
16'
deep, and will have435'
of rails. Two towing carriages will be used: a
conventional carriage designed for resistance, seakeeping,
and other Froude-scaled testing in the one to fifteen
knot speed range; and a high performance carriage, designed
for flow noise, boundary layer, submerged model, and other Reynolds-scaled testing in the two to 30-knot speed range. The basin will be equipped with a wavemaker and an
1. Tank Length
In designing a towing basin for hydroacoustics and turbulence research, 'thestatisticalaccuracy which can be obtained for a given data run is an important consideration, and must be viewed in the light of the
data analysis techniques which are tobe used. Turbulence
and flow noise data are generally analyzed for spectral
content (amplitude vs. ±'requency),In an attempt to
establsh the energy. content of the ariOus eddy' scales present in the turbulent boundary layer. (1) The most elementary measure. of the eddy scales is. the.
one-dimensional, wave number, k1, 'defined in. the. direction.of
the mean motion as follows:
k1 = 2rrf
where f=frequency
measured by the sensing device, and, mean convection velocity of the eddies past the sensor ' ..The convection velocity, as measured through cross correlation of several closely spaced sensors, or with th wave-vector
filter, varies with theeddy sizes, but for purposes of
this discussion an average mean value may be taken 'as
approximately O.3 times the free stream velocity () (môdel
velocity in the case of towing tests), i.e.
O.3 V.
Combining the above expressions it s seen that eddy scales
of a given wave number, k1.,appearas higher irequencies as the velocity past the sensing device is iicreased.
Table 1 shows the range of wve..numbers generally found.
number will appear when sensed by a single hot film
probe at velocities of 10 and 50 feet per second.
Table 1. Equivalent frequencies corre-sponding to wave number ranges found in
boundary layer turbulence. -.
Analysis of the spectral energy content (turbulence or acoustic energy at a given wave number) is usually done..
on a power spectral density analyzer (analog or digital). Since the data sample is finite. (and quite short in towing
basin tests), one must be concerned about the statistical
accuracy of the data, relative to the infinite ensemble
of a stationary random process which is statistically
equivalent to the phenomoneflabeiflg measured.. If one assumes the turbulence to be a stationary random process
(time invariant mean values, correlation functions, etc
which are difficult to achieve without very precise ca'riagespêed control) and that: the data smples are ergodic (every firiite..data sample exhitsthe same.
statisidal prOperties as an infinite sample) , 'the following analysis parameters may he defined:
-15 -Frequency Corresponding to V .1Ofps Frequency Corresponding to. V 50 fps (ctn_l)
(ft)
. ':z. , .()
10_i 1' 10 100 0.5 305.3050j
'
. if O 'J+O0 .4000 - '20 " 200 '2000 20000Lr Length of the steady state tank run,
Tr = Tirne.over which the sample is to be
analyzed, the maximum possible time
being dictated by the length of the
steady state run divided by the
carriage velocity,
f = frequency of the elecr.onic
signal ,p Hertz (Hz),
B = Bandwidth of the analyzer filter (Hz),
V = Steady state carriage velocity,
Q = f/B ??quality??
or analysis resolution,
n = "degrees of freedom?? (See Appendix A)
2/2,
= normalized mean square error (Appendix A). Now from Appendix A, the reliability of a statistical
analysis may be related to the statistical degrees of freedom, n, i.e.,
flmax2BTr2fTr21Lr
QV
but K1 and U
O.3
V,
thus nmax
2(O.3)UK1L
= KlLr
2rrTJQ .8Q
and rinaxQ KlLr,.
3.8
which shows that for a given wave number, a new dimensionless
parameter, max9' which is a measure of the statistical
accuracy and resolution of a spectral analysis, may be defined and is a function only of the length of the steady state
Typical analyzer Q's are as follows:
Previous experience has shown that the
minimum value of n which can yield meaningful, turbulence
results is about 50, with a value between 100 and
400
being desirable for good statistical accuracy. At lower
values of n, Figure A-1 in Appendix A shows confidence levels diverging severely so that the measured power spectral density function may no longer approximate the true power spectral density function. Thus for one-third
octave analysis, which represents a minimum resolution figure for most work, and. for a minimum value of n,
= (50) (4.43) =
222
Comparing the 50 foot steady state run in the present tank with the
250
foot steady state run(50
fpsfor 5 seconds) in the proposed tank, one obtains, for the expected values of turbulence wave numbers:
Table
2.
Comparison of dimensionlessstatistical parameter for wav number ranges found in boundary layer turbulence.
La1yz
1/3.
Octave 1/10 Octave 5% 1% Q4.43
14.5
20
100
max max .(cm)
(ft1)
Lr = 50 ft L = 250 ft101
3.05
40
200 130.5
400
x lO
10
3.054 x
io3
2. x
100
13050
4 x lO
2 x l0
Thus Table 2 shows that the proposed new tank length
just meets. .the minimum criteria at. the lowest wave
numbers of interest in turbulence work.
From a different point of view, increasing
the steady state runIenth from
50 to250 feet allows an increase offlm by a factor of five. This allows
corresponding reduction of the uncertainty interval at 90% confidence by a factor of approximately three. Reference to Figure 1-A, Appendix A, shows that for n = 10, the true mean 'square value lies between +lao% and -44% of the measured mean square value. For .j = 50, 'the true mean
square is between +40% and -27% of the measured mean
square value. Thus with a 250 foot steady state run, the
true power spectral density function may be estimated
with three times the accuracy at a 90% confidence level. This is a significant improvement.
Note that so long as the carriage speed is
steady'and repeatable, which requires precise speed control, all of the above results are independent of carriage
speed. Speed only determines the
frequency corresponding
to a given wave number. Changing the carriage
speed will not allow use of a shorter tank. Even though statistically significant. hot film measurements will be possible at
frequencies as low as 20 Hz, it is still expected that low
frequency hydrophone measurements win be meaningless.
Low frequency acoustic absorbtion in water is not possible with anechoicmaterjals of reasonable thickness. These facts, and thegerieralnecessjty of high Reynolds numbers,
requires high carriage speeds despite. the independence of velocity and
%ax
'From a more traditional naval architectural
viewpoint, the' p'rOpose.d basin length is. 'd.esirable for
the ability to run-high speed EHP tests
with a reasonable steady state data collection interval
and relatively mild accelerations,
the.ability to obtain a reasonable
period of steady state conditions in self-propelled model tests, and
c. the
long (and hence more without unreasonable It must
ing experiments in random seas involve extremely low.
frequencies, and thus require great run lengths to achieve good statistical accuracy. The proposed tank will use models ten feet long, and will operate with a maximum
wave lengthof
thirty feet. Therefore, by reasoning similar to that above, it is possible .to.calcu.latE.. aminimum steady state run as. a function of spectral. analysis techniques. It was, shown above.that
flmax=2fLr.
QV
where in this case f is the frequency of- encounter with the longest wave of .interest, . .
f =V.+.Vw
EThus,
=
'y)
imaxw
Therefore for the.followiflg assumed conditions: model
speed, V = 7 ft/sec.; maximum wave length, L = 30 ft;
maximum wave velocity,.V =12.4 ft/sec; and
minimum
desirable
-
"max
= 222;' the required steady state run
length, Lr will be 1200:feét. If the model proceeds at
only L1. ft/see, the length of run required to achieve the ability to, install a relatively
efficient) wave absorbing beach, sacrifice in useful model ran.
seakeep-necessary run time, and thus the seakeep-necessary will
be reduced to 00 feet. Thus, in the case of seakeeping
experiments, Lr is strongly speed dependent.
Since a 1200+ foot tank is not practical, seakeeping experiments must allow for the restricted
length. In the case sited, five
consective runs at exactly the same speed or thrust will meet the minimum criteria. This is obviously time consuming. In addition,
multiple runs require truly random seas, and exactly
repeatable carriage speeds. Violation of either of these
requirements adds to the uncertainty of the results. Some relief may be achieved by using smaller models, thus
raising the frequency of the longest important wave, and
lowering the model speed at a given Froude number.
However, several other factors make seakeeping
tests in random seas quite impractical. The uncertainty
interval quoted above for n = 50 is very large in the
context of steady state or low frequency naval architectural
problems. Bringing n to a really desirable value, such as
400, will require a steady state run of 9600 feet. For
high speed vessels, such as planing vessels and hydrofoils, the dependence of run length on speed will increase this
length even further. In the following sea case, where frequency approaches zero, the required length of run
will approach infinity. Therefore, except in cases of
highly non-linear phenomena, testing in regular seas and
employing linear superposition to predict random response appears not only to be more practical, but often more
accurate as well.
Thus ignoring the random sea problem, we see the minimum 30 knot run length is 250'. To this we must
(1 second at 30 knots) to.allow starting transients to damp out. The total usable run over which the carriage
can actually tow a model must therefore measure
30'.
Any less run will compromise the intended mission of
the proposed facility.
In addition to the 30' of run, the tank
room will require auxiliary space at either end of the basin itself for a wavemaker, a drydock, and a rigging
area. Th drydock is essential when operating frequently
with submerged models. A large pump will transfer water very rapidly from the d.rydock to the tank proper. Thus adjustments, repairs, and inspections may be made on a submerged model quickly and efficiently without removing
the model from the carriage.
Since the basin will employ two carriages, described below, parking space must be available for the
unused carriage. The high speed carriage will be parked over the wavemaker when not in use, while the conventional carriage will be parked over a small "finger" drydock
behind the main dock. In this manner either carriage may
use the main drydock the high speed carriage may accelerate from within the drydock; the high speed carriage may be
stopped over the wavemaker in the event of braking system failure; and the conventional carriage may be rigged with surface models in the "finger" dock while the high speed
carriage is in operation.
The entire length of rail required includes the two 17' drydock lengths, 21' for the wavemaker and
emergeny braking, and a run of 30', for a total rail
length of 435 feet.. The tank is nominally referred to
2. Tank Depth and Ceiling Height
The testing of self-propelled models and high speed surface and sub-surface craft impose strict restrictions on tank cross section. Since shallow water
wave effects will inevitably be strongly present whenever the model runs at the maximum design speed of 30 knots,
a compromise depth must be specified. A depth of'16 feet,
which seems maximum under architectural considerations, allows a model speed of 9.5 knots without shallow water
effects (4). This represents a 10 foot model operating
at a speed/length ratio of 3.0, or a 5 foot model at a speed/length ratio of 4.25. These model speeds are adequate and 16 feet was chosen as the required depth.
Most models operating at higher speeds will be:ubmerged and Froude number scaling will be of less importance,
thus shallow water waves will caus less difficulty.
The ceiling height is also of critical concern. As seen in Figure 3, the cantilevered rail system and
large flat carriage deck necessitàtesa carriage which
runs relatively high above the free surface. Nine foot
clearance above the carriage deck is essential for head-room, hoist clearance, and safety in the vicinity of the electrified trolley system. In addition to these
considerations, Inui and Tagori recommend an unobstructed height over the free surface of 1.5 times the model length for purposes of stereo-photo analysis of ship mOdel
wave patterns (5). This type of work wOuld normally be done using 10' models; thus requiring an unobstructed 15'
above the model. Inlight of theseconsiderations, it is
evident that a.total ceiling ôlearance of 16'-6" above
3. Tank Width
Tank width was determined after consideration of self-propelled model tests. A six inch diameter
propeller was considered the minimum useful size. A
variety of ships, with differing displacement_length and
L/T ratios, were considered operating through arange of Froude numbers. The resulting blockage data was
weighed against architectural contraintS and a width of 26 feet was chosen. The tank's designed cross sectional area is therefore 416 ft2. The largest self-propelled
model considered would have a midship sectional area of 2.1 ft2, or about 0.5% of the tank area. The maximum
increase in local flow velocity seen by any model due to blockage will be less than 2% depending on the
interaction of wave and blockage effects (6). Similar considerations applied to the existing 5' tank limit model lengths, in some cases, to as little as 24", which
is uncomfortably small when scale effects are considered.
Because a significant amount of testing is to done at high speeds with a submerged model, the
traditional 2:1 ratio of depth to width was not employed. It should be noted that an increase in either depth or width would be welcome, but because this tank is to be
constructed as part of a larger building it does not appear architecturally feasible. The chosen dimensions
represent the minimum cross section consistent with the tank's pràposed mission.
Another feature effecting cross section
design is the roposed hydroacousticS capability. In order to minimize noise reflection from the tank walls
and bottom it is anticipated thatthe tank will be lined with a sound absorbing material. The tank dimensions
L1.. Architectural Features
The detail design of the following items
will be intimately related to the design of the new
Engineering Department building, and thus will be the responsibility of building architect.
a. Water Basin
The basin is to be constructed of any
suitable material, probably reinforced concrete. If
reinforced concrete is to be used several questions must
be answered.. Is it possible that steel reinforcing will
cause electrical: problems in the towing tank instrumentation,,
for example acting as an antenna, collecting and
reradiating radio energy? Or perhaps can the
reinforcing
material be grounded and act as screened room, shielding
the tank from radio interference? Is it possible to use
post-tensioning to alter the natural frequency of the
basin itself, thereby improving its acoustic and vibrational
characteristics?
The basin walls must conform to dimensional
tolerances of about +1", and must be as smooth as
possible. Roughness elements,, such as exposed aggregate, seams,
non-flush fittings etc., will interact with waves generated
in the tank and lead to undesirable secondary wave patterns.
b. Basin Foundation
The water basin foundation must be
carefully considered. The strict acoustic specifications will undoubta.bly effect the choice of foundation and fill material. Sand, or other granular material, is a good
base with good acoustical properties. The foundation of
the water basin must be acoustically independent of all surrounding foundations.
c. Observation and Photography
Observational consideratiotis are of utmost importance and fall into three categories: above the
free surface, at the free surface, and below the free surface. For observation above the free surface, a walk way should be constructed along both sides of the basin. Open central well carriage will allow good viewing from above, while photographic subcarriages, such as developed
by Inui and Tagori in Japan, might someday be employed for overhead photography of wave systems.
Observation at the free surface will be
accomplished by the installation of 2' high windows over the mid 50 feet of the basin length. These windows, flush mounted to the inside of the basin, will be centered at
the free surface and will be spaced as closely together
as is structurally sound. A platform will be constructed below these windows so that the free surface is at a
short male observer's eye level. Several underwater
flood lights will be installed below these windows, so as to illuminate the field of vision.
Far underwater viewing a vertical shaft and a horizontal tunnel to a point beyond the tank
center-line will be constructed at the tank midlength. A series of flush mounted windows or port lights will be placed
along the shaft and tunhel to allow observation, television, and conventional photography, and flood lighting of models.
These windows or ports should be as closely spaced as structural considerations will allow. The ports will accommodate only several persons. However, the building
is to have an educational television system, and it i anticipated that this system may be employed to view the
d. Environment
The acoustic environment specification will be written as a result of acoustic studies, and will
be very stringent. Three other envi1onmenta1 features must also be considered: air conditioning,, water
treatment, and lighting.
(1) Air Conditioning
Air conditioning and humidity control of the towing tank room may prove impractical, however
provisions must be made to insure a very constant ±oom
temperature at some point between 7Q0 and O°F. In the event the constant temperature and humidity which are to be maintained are not suitable for electronics and office spaces, such spaces will be separately air conditioned. The presence of 1-1/4 million gallons of constant
temperature water will certainly assist in stabilizing conditions. Temperaturemustbe maintained high enough
to allow effective turbulence stimulation, but must be
lOw enough for comfort, for the electronic instrumentation, and for humidity control.
(2) Water Treatnient
The water itself must be very
care-fully considered. Water clarity is essential for high-spéedH
color motion picture and television photography through
13 feet of water. A good filtering system and a surface
skimmer are therefore required. But eén more than clarity, bacterial cleanliness is essential. Bacteria and algae
excrete substances which severely affect the frictional
drag of ship models. Thus organisms in the tank water
must be strictly controlled through chlorination, the control
of acidity (pH), and the contrOl Of sunlight in the tank
room. The water must be maintained at the same very
heating and filtration
must be accomplished without
establishing convection currents and/or thermal gradients.
(3)
Illumination
The tank room need not
be brightly
lighted. .Lights suitable
for corridors or passages
should be sufficient, with
loca.l illumination
pr.ovided.
on the large
towing carriage and
in the work areas..
Sunlight (and some types
of artificial
light) promotes the
growth of algae and is thus not, desired.
In way of the
observation areas illumination for photography,
bçth
below and above the
surface will be possible.
e.
Mono-rail Hoist
A mono-rail hoist system
will be installed
to service the
380' towing tank.
A low overhead electric
hoist will run on an
I-beam monorail over,
the centerline
of the basin, the
model storage racks, the'model rigging
area, the
stability/ballaSt1n
tank, and the loading
dock.
f
Pumps' and Pi,ing
A pumping capacity of
1000 gal/mm
will
be required to drain the.water basin, and to shift water
from the drydock to
the water basin. .This
wili,r?qUie
a pump notor
of.abOut 50 'HP, which
is substantial, but
required to drain the dry docki' 15 minutes.
A pump of
this ,size will draii
the basin in 2l.hotrs.
,,,Pumps .and piping will
also need be
supplied to operate the filtration, surface skimming, an
heating systems.
These.sySterflS should be co
cted tothe
hydrorneChafliCs and
oceanography laboratories, so ,that,
these laboratories
might use treated, filred
and.heated:
water.
Discharge.frOm these 1aborat0.rie5
riust not return
to the large tank;
therefore.all facilities
should have
independent drains to a sewer
or to the river.
Inlet water should be supplied to the
large tank at the maximum rate acceptable to the Naval Academy Public Works Department. It is estimated that an inlet rate of 250,000 gal/day will be acceptable,
filling the tank in five days. All inlet water should pass through the heating, treating, and filtration system, so as to be completely compatible with water already in the tank.
SECTION III
III. CONVENTIONAL EQUIP1VNT
The conventional equipment planned for the new tank will include a 15 knot resistance and seakeeping
carriage,
a wavernaker and beach, model and support facilities,and a rail system.
A. Conventional Carriage
1 Configuration
The conventional carriage (Figure 4) is envisioned as being quite similar to a number of existing box-girder central-well carriages. Model visibility will be good, since a box-girder offers no Obstructions.
Thus
the open central well offers advantages as a teaching aid. The carriage will accommodate approximately 22
nidshiprnen,
standing around the central well. The instructor may
stand on the catwalk within the well while explaining the experiment underway. One side of the carriage is
cantilevered away from the well to allow construction of
an air-conditioned cab containing instrument racks and a
desk, leaving the central well area unobstructed. The carriage will be used both to teach the Naval Engineering curriculum and to demonstrate research which is not
formally curriculum connected. The entire facility hs been designed with this in mind, and it is hoped that
Naval Architecture students will be shown all current research.
The carriage structure must be quite stiff, so as to allow the forced oscillation of models at wave encounter frequencies (0 to 7 Hz) with no resonant
carriage response. This will require a carriage structure
natural frequency of at least 14 Hz. This stiffness should apply to all vibrational modes, but especially
to the vertical vibrations of the cross tank span. If a reinforcing truss is required, model visibility must be
retained from every side of the center well.
2. Speed Control
Controls 'of undetermined type will be
located on the carriage.. In one system considered, the
carriage would be propelled by four synchronous motors,
driven by what is essentially an adjustable frequency
power supply. Such solid state rectifier-inverter 'drives
are now commercially available in the 5 to 500 horsepower range. In these drives, ordinary a.c. current is
rectified by silicon diodes to d.c. current. The inverter then employs silicon controlled rectifiers to invert the d.c'. current back to a.c. current at a predetermined frequency. The output frequency is set by very stable
oscillators which may be digitally controlled for
repeat-ability. The speed regulation of such a system is, from no load to full load, essentially 0%. The most serious
disadvantage of this system is a relatively slow response
time when speed changes are called for, and a limited
range of motor speeds (15:1).
Another promising control system is the
digitally controlled d.c. drive. In this system, control is accomplished through a frequency based feedback loop. Excellent speed regulation may be achieved through the
by Cutler-Hammer, is presently being installed on the
3 H.P. carriage of the existing Naval Academy tank. It will offer a speed regulation of O.Ol%.of.t..speed over a100:l speed range. The primary.advantage. of this system is the rapidspeed change which may be accomplished
through the use of regenerative 'braking and the d.c.
motor's inherently rapid response. The primary disadvantage of the.system is the complexity of control required. to.
achieve the required .speed regulation.
The speed control finally chosen must: achieve a steady speed regulation of .0.01% of set speed
through .a spee.d range, of at least 1 to 15 knots. . Equally
important is repeatability. If statistical da.ta is
being taken, either in waves or in the boundary .iayer,
.
runs made at different speeds may not e compared iith-out reduction of the level of confidence.. It is,.therefore
necessary that the speed control be pre-set, and that
the system repeatedly match that pre-.set speed within.the tolerances of the speed regulation. Because 'of the
inherent reproducibility of digital speed settings, they
are to be preferred. Repeatable pre-set speeds are also of very great convenience..in the stereo-photo analysis
of wave trains... . .
Other considerations in the design of a
speed control system are: .compatability with, or use. of. the high performance carriage .drive (eg.. sharing an external drive, or an M.G. set) electromagnet±c inter-, ference with the various .instrumentat-iôn; a tracking
capability for operation with self-propelled models; data acquisition controls which 'will initiate data collection
upon the. .carriage':attaining steady speed; adjustable .:
accelerationand braking in the 0.01 g to 0.2 g range
(4 ft/sec2 is an acceptable working acceleration for
steel wheels on steel rails);'and noise and vibratiOn
reduction. The new 3 H.P. system at Annapolis meets these requirements, as' they apply.
Wheels "
The recent world-wide increase in rapid' transit technology is a potential boon to towing tank
designers. An example of this is research being done to reduce wheel noise and vibration. Thè"Acoustic Flex"
railroad wheel, manufactured by the Standard Steel Division
of Baidwin-Lima-Hamilton, employs an aluminum center
bonded to a steel tire with an acoustically inhibiting
material. Such 'a wheel is. lightweight andis said to
substantially reduce iheel induced noise in rapid transit opera1ons; it is therefore under consideration for use.
on the proposed towing carriage. " V
Fast Return System
The faculty at the Naval Academy feels'that
a' system to allow fast return runs of the carriage would
greatly increase the efficiency of testing. . 'Such a
system should allow the carriage to proceed backwards at speeds greater than the speed at which the model should be backed downo ±f at the same time the residual waves in the tank could be "skimmed off," the tank should be ready for the next ,run shortly after the, carriage returns to the starting end. This might be accomplished, for' one-way tests, by a mechanically positioned plank towed beneath the carriage. The system must keep the model dry
during the return run, by either lifting it from the water
or by protecting it in some manner. The system used should put approximately the same amount of energy into
the water on the return run as the model put in on the
ahead run. In this way surface currents will tend to
be canceled. The fast return apparatus should also
maintain an evenly distributed high level of background
turbulence conducive to turbulence stimulation. For
calm water tests, a turn-table mounted dynamometer,
allowing precisely l00 of rotation, would make data runs
in both directions possible. It might also be'used for yawed model tests, as with yachts.
B. Seakeeping Capabilities 1. Wavemaker
A wavernaking capability is envisioned for
the new facility for the investigation of wave induced ship
motions and hull structural loadings. Regular, long-crested, waves will undoubtedly be the most highly used wave
configuration. The resulting responses will be used in
conjunction with the principle of linear superposition to predict vessel responses to irregular seas. The wavemaker will be able to generate regular waves from 2 to 30 feet in length, with wave heights of 2.0 to 36.0 inches
(i.e. 0.3< Lw = 3.0, and a maximum wave slope of 0.144
for wave lengths of 21 feet). For generation of the very
high waves, the tank water depth will be reduced.
The large wairemaker will allow testing of
reasonably large models of vessels and structures in
surf conditions. Coupled with the variable profile
beach (paragraph 2, below), this will allow the tank to
conveniently test a variety of. problems important in the design of amphibious and rescue craft, assault ships, and ocean structures. Few tanks exist today with this capability.
In addition to the regular long-crested
wave capability, the ability to generate irregular
long-crested waves, using a preselected irregular. signal on
a magnetic tape, is desired for investigations involving
such variables as slamming frequency, deck wetness,
forefoot emergence, or to simply demonstrate to students the principle of linear superposition, Fourier analysis techniques (generalized and periodic),etc. It is hoped
that real-time spectrum analysis can be made during
preparation of the magnetic tape such that, by filtering and component frequency amplitude control, a desired
spectrum shape and energy can be "sculpted" and presented
to the wavernaker control system. The wavemaker will be
controlled both from ashore, and from the conventional
carriage.
While vessel response to waves coming from directions other than directly ahead or astern are
important and are desirable to investigate, the problems
of wave reflection, refraction, orbital velocity
interaction, and consequently the unknown wave characteristic seen by a ship model as it traverses a long narrow basin
seem at this time to rule out the installation of a "snake-type" wave generator.
It appears at this point that a pneumatic
wavemaker, similar to the NSRDC installation (and the present USNA wavemaker), will be designed and built for the proposed. tank. However,, in the event a hydraulically
actuated plunger proves significantly less expensive than
the pneumatic device, it will be used. The.pneumatic,
although large and complex, will have fewer problems due to inertia, plunger buoyancy, natural frequency, and
mechanical alignment. The dome of such a wavemaker will
be adjusted vertically to accommodate changes in water level. Programming of the butterfly valves will allow
regular and irregular wave generation. It is also
conceivable that nodal wave patterns could be generated
by incorporating contiguous but independent valve-dome units across the tankwidth.
2. Wave-damping Beach
The generation of well formed waves of desired length and amp1itude is only part of the problem
encountered in a seakeeping test facility. In order to
know what waves the model "sees" and to reduce the time
interval between test runs, an effective wave absorbing beach (Figure 2) must be installed at the tank end
opposite the wavemaker. Tank sidewalls must be smooth and vertical to avoid diagonal wave generation. Two
beach types are being considered: a curtain of air bubbles, and a variable profile rigid beach. The bubble curtain
has reportedly been successful in Europe, and is presently
being tried at the Naval Academy.
The rigid beachwil]. be utilized in the simulation of near shore ship motions, amphibious craft
situations and surf structure.problems. Therefore, the
ability to roughly simulate a shelf or reef coastal
geometry is necessary. To accomplish this, a segmented,
adjustable beach is envisioned, constructed of six, or more, eight foot concrete segments, positioned with hoists
and held by chains. The beach must have great
rigidity, to maximize wave energy absorbtion, and a
roughened surface, to break up the waves and cause viscous
energy dissipation. Bubble curtains along the tank wall, in way of the beach, may allow sustained zero Froude Number surf testing, damping any model generated
transverse waves.
In regular seakeeping work, the beach
could be configured ma
roughly parabolic shape with a low (about 50) slope at the waterline fOr maximum wave energy absorption. It ma.y also be that for some tests a short Hammil-tYPe beach will be required; this couldbe rigged using existing beach segments. If the bubble
curtain beach is successful, the rigid beach will not be required for a majority of experiments.
In any case, the beach will be lowered
during cairn water tests to allow model passage into the drydock_outfitting area.
C. Model Handling and Technical Support Instrument Modules
The use of two carriages has inspired a
modular instrumentation concept. All dynamometers,
sub-carriages, and instrumentation packages will be rigged to fit the 11' well on both carriages. The well will have a set of machined ways on their fore and aft sides upon which modules up to 11' in width may rest. It will also
be possibleto run
models 5' off center, in order to allow clear photography and observation. The modules will be rigged in the shop area and transported to the towing tank for mounting and calibration, all while the second carriage is engaged in testing.Drydock
The proper layout of the drydock (Figure 2)
will be crucial to this system. It must be possible to attach a floating or immersed model to the carriage. Thus,
the drydock must allow the installation of floating models
on the conventional carriage, and the installation of submerged models on the high performance carriage. A drydock with a deep dock adjacent to the tank, and a
shallow finger dock extending behind should be satisfactory. Submarine models must be rigged
in thedeep dock
available to either carriage. Surface models could.be rigged in the finger dock, available only to the
conventional carriage. Thus the conventional carriage
will be placed on the rails aft of the high speed
carriage.
3. Technical Support
The new Engineering Department building is to house not only the proposed hydromechanics laboratory, but aerospace, mechanical, marine, materials, nuclear,
weapons, and systems engineering laboratories. What would normally be considered a school of engineering,
along with its laboratories and shops, will all be under
one roof. As such, the new building will contain
approximately 20,000 ft2 of technical support facilities
including: a machine shop, a model and pattern shop, an electronics shop, a mechanical instrument shop, a
c'alibration and standards laboratory, a photography
laboratory, welding, foundry, electrical and paint shops.
Approximately 35 to 40 men will be employed in support of
the research teaching laboratories, including model makers, a mechanical instrument maker, and electronic technicians. The hydromechanics laboratory will employ several technicians to set up, run, and maintain the
equipment.
D. Rail System 1. Rails
The rail system is a critical component in any towing tank, and must reflect the tank's overall
mission. Traditionally, the primary consideration in the design of rail systems is a.àcurate installation. Accuracy
is required both in maintaining a constant vertical 36
eight above the free surface, and in providing at
Least one rail with a very straight horizontal alignment.
kaii insta11aion tolerances should be between 0.001" nd 0.005", since they represent the reference plane ror ship motio.n measurements. and for straightline motion
ssential to most testing. it is interesting to note
hat alignment errors of as little as 0.001" in the bwo-mile length. of the Stanford .Linear Accelerator h.ave
een measured using gas lasers and special detection equipment, (7)
In recent years more emphasis has been laced on smooth rails, for the elimination of carriage
ibration. Thus, as mentioned above, several recent
tanks have utilized continuous welded and machined rails.
3ince the study of flow noise is a primary mission of he proposed tank, such rails are absolutely essential.
2. Rail Supports
The primary reason that flow noise studies are impossible in existing towing tanks is the presence of carriage generated, noise in the water. To. avoid this,
and to gain other previously mentioned advantages, a
pantilevered rail system has been chosen. This cantilevered
ystem will be structurally independent of the water basin,
thus preventing structure borne carriage noise from 3ntering the water.
The cantilevered rail system (Figure 3) is patterned after the Osaka University tank in Japan and the National Research Council tank in Ottawa. Reinforced concrete cantilevers will extend over the tank, steel
r concrete beams will be supported by the cantilevers and
ill run the length of the tank, supporting the actual ails. The deflection of reinforced concrete beams can