ºſº
- ， , ; , ！
<>， e=…=）=）

º
B
#
-7
M.
ºf
!:
º º
º

The Experiment. Ship Tank
OF
The University of Michigan
- liºt'ſ * *
BY
HERBERT CPSADLER, D.Sc.
PROFESSOR OF MARINE ENGINEERING
REPRINT FROM
JUNE MICHIGAN TECHNIC
| ? 06
PRINTERS
- THE ExPERIMENTAL SHIP TANK OF THE UNIVER
SITY OF MICHIGAN -
HERBERT C. SADLER, D. Sc. PROFESSOR OF MARINE ENGINEERING.
As the experimental tank and apparatus connected therewith is now
completed, a short account of the method of preparation and testing of
ship models may be of interest at this time.
In the first place it may be pertinent to describe in a general way the
object of the tank. The shape or lines of a vessel, together with the esti-
mate of the horse power necessary to drive her any given speed, is the
most interesting and at the same time most difficult part of the naval arch-
itect’s work. Fortunately or unfortunately the nature of the conditions to
be fulfilled are rarely the same for any two ships, so that although former
experience, properly and scientifically applied, may be of great value in
solving any new problem in ship propulsion that may arise, it seldom
happens that a new design is the exact counterpart of a former one. It
is, therefore, difficult to say beforehand what the effect of modifications of
a vessel’s form will be, and even more difficult to say which is the best
form to fulfil any given set of conditions.
The primary object of the tank is to perform experiments upon various
forms of ships and to determine the resistance to motion of these forms
at all speeds. In general the line of work that will be followed will be
of a Systematic character, that is to say, models of various forms will be
tried and the effect upon the resistance of modification of these forms,
such as variations of the ratios of length to beam or draft, determined.
Numerous other experiments such as the effect of bilge keels upon - the
rolling and speed of vessels will also receive attention. ,
The tank itself is 300 feet long, 22 feet wide, and Io feet deep, and
forms the basement to the east wing of the New Engineering building.
This length is the least that can be used in order to allow time for start-
ing, obtaining uniform speed and stopping. The breadth and depth are
necessary so that the effect of the sides and bottom will not have any
tmaterial influence upon the resistance of the model. Spanning the tank
is a traveling truck which runs on rails on either side of the tank. This
truck, shown in Plate 2, is driven by a 25 H. P. motor whose speed can
be so regulated as to give speeds to the truck varying from about ten feet
per minute up to eight hundred feet per minute. It is essential that the
speed of the truck should be uniform at any speed between these limits,
so that the resistance of the model may be determined accurately at any
\ , , ,
A . ~~
** *
s
t
•
}
.*

— 4 —
speed. The models are run at a series of different speeds and a curve
of resistance in terms of speed obtained. In order that the speed may be
uniform and not affected by changes of load in the power house, a special
motor generator set has been installed, and owing to exigencies of space
and also for convenience, has been placed upon the truck. The current
from the power house is taken to this set by trolleys and converted as re-
quired. The connections are such that if any fluctuation takes place upon
the line, this is compensated for in the installation and the speed of the
driving motor remains unaffected. The speed is regulated by a controller
with five main stops, and between each stop an auxiliary rheostat with
fifty stops may be thrown in, so that between the limits of speed given
above, two hundred and fifty different speeds may be obtained. The
driving motor is also fitted with a high and low speed gear. The switch-
board and connections are also mounted on the truck.
On the forward end of the truck is the dynamometer through which
the models are towed. This consists essentially of a vertical bar mounted
on Emery supports, that is to say, instead of the usual knife edge a thin
piece of spring steel is used, with about one-sixteenth of an inch, between
the supports. This gives a rigid but flexible and practically frictionless
bearing. Parallel rods are introduced so that both the pull upon the
spring and model are in a horizontal direction. The model is attached
to the lower end of the dynamometer and its resistance taken up by the
spring as shown in Plate 3. The amount of extension of the spring is
registered upon a revolving drum which is driven from the main shaft
of the truck. Upon this drum are two other pens, one of which is con-
nected to a clock and registers every half second, the other is connected
with contacts along the side of the tank and registers every ten feet. Thus
the time and distance and hence speed are determined. Two other pens
register the amount that the model rises or falls at the bow or stern.
when moving at different speeds. While not at present installed, a cam-
era will also be arranged so that a photograph of the wave formation may
be obtained. The above are the principal observations that are taken in
connection with the truck and dynamometer.
The preparations of the models themselves will now be considered. ,
The substance of which the models are made is parafine wax with a mi
ture of about four per cent bees wax. This material is the same as that
used in most laboratories of this kind, and was chosen for the following
reasons: It is very easy to handle, that is to say, it may be melted at
a low temperature and cast without difficulty; it is also easily cut, planed
or scraped; it enables us to have a uniform surface for all models, and
when a model is not required for further experimenting it may be broken
up and used for another of a different type. -
Before casting a model, a mould is first prepared. This mould is
e' i.
f
PLATE I.

"z 3 LV Ta
---- |-
----

— 7 —
made in ordinary modelling clay. Sections of the vessel at different points
in its length are first cut out of wood, about one-quarter of an inch larger
than the actual size required. These are placed in the bed and the clay
moulded and faired in until it conforms to the proper shape. As the
models have to be cast hollow, a core is next made. The forms as above
are cut out so as to allow a thickness of paraffin wax of about one and
one-half inches. These forms are then connected together by thin wooden
strips. and covered with canvass, so that the core forms as it were a cam-
vass canoe which is suspended inside the clay mould. The paraffin wax
is then melted in a tank provided with a steam coil and the mould poured.
While cooling considerable contraction occurs so that small quantities of
melted wax must be added continually. While the wax is being poured,
water is introduced into the inside of the core in order to overcome its
tendency to float and also to aid in cooling.
When cooled the core is withdrawn and the model floated from its bed
by introducing water between the walls of the mould and the model. The
model is now in its rough state and ready to be cut to the correct form.
It is next placed in the cutting machine (Plates 3 and 4). This machine
consists primarily of two tables, on one of which is placed the model
and on the other the drawing of the lines which it is desired to reproduce.
These two tables move together and are driven by a motor, the motion
of the driving table is, however, usually about one-half as fast as that
of the nodel table; but this ratio can be varied by introducing change
gears. The object of this is so that the drawings do not have to be made
unnecessarily large. In the middle of the machine is a cross piece upon
which are two traveling heads which move together inwards or outwards
and are operated by a right and left handed screw by means of the handle
shown on the right hand side immediately over the drawing. These heads
carry two vertical hollow spindles which have a screw thread cut on the
outside. By means of a worm gear these may be raised or lowered to
any desired extent, the amount of the vertical movement being measured
by a scale and vernier upon one of them. Inside each spindle is a shaft
which is driven by a vertical motor on the top, and to the bottom of which
is attached a two-bladed cutter. As the in and out motion of the cutters
must correspond with the breadth upon the drawing at various points, the
motion of the cutters is transferred to the drawing by means of a panto-
graph. One end of the pantograph is attached to one of the cutters and
the other moveable or center part to a bar which is carried over to the
drawing, and on the end of which is a pointer: If for example the draw-
ing is one-half the size of the model, the arms of the pantograph are set
in this ratio. When therefore the screw which operates the heads carrying
the cutters is revolved, the cutters move in or out a certain amount, and
the pointer on the drawing one-half this amount. As, however, the points
— 8 —
of the cutters describe circles, the pointer is also a circle but of one-half
the radius of the cutter circle. (In cases where a different ratio of length
to breadth from that as shown in the drawing, is being cut, the pointer
is an ellipse). In Plate I it will be noticed that an adjusting lever is at-
tached to the fixed arm of the pantograph, the reason for which is as fol-
lows: It sometimes happens that the drawing paper and hence the center
line of the drawing may become warped. Under these circumstances the
cutters would not be cutting the actual breadth from the center line, but
from some imaginary straight line. The end of this lever is therefore
attached to a frame which carries a roller, which bears against a batten
which is placed upon the center line of the drawing, no matter what form
this has taken. If the center line is slightly curved the roller follows this
curve and moves the fixed end of the pantograph a corresponding amount
and therefore compensates for this error.
From one drawing it is possible to cut any number of models of the
same form but varying in ratio of beam or draught to length. If a broader
or narrower model is desired all that has to be done is to alter the fulcrum
of the pantograph and the relative motions of the cutter and pointer on
the drawing are altered accordingly. If it is desired to change the ratio
of draught to length, the amount by which the vertical motion of the
cuters is changed for each waterline may be correspondingly increased or
decreased.
The method of operation is as follows: The model is placed upside
down upon its table carefully centered and clamped down, the drawing
is also placed upon its table and the center line adjusted. The cutters are
now run up to the bottom of the model and moved in so that they nearly
touch. They are then put into motion, as is also the table, and the base
line or top of keel cut from one end to the other. The cutters are now
lowered to a depth corresponding to the first waterline as shown on the
drawing. The cutting is usually started from amidships and worked both
ways toward the ends. The operator now runs in the cutters until the
pointer is exactly upon the line required and starts the two tables mov-
ing. By operating the handle on the right and left handed screw, the
cutters are gradually brought together and their motion transferred to the
pointer. He must so regulate this motion that the pointer remains tangent
to the line he desires to follow, while the drawing moves along. In this
way the cutters are made to follow exactly any desired line. When one
line has been cut they are lowered to the next and so on until all the lines
have been cut. When this part of the work is finished, the model has
the appearance as shown in Plate 4, that is to say, a series of longitudinal
grooves have been cut, which represent the correct shape of the various
waterlines. It is now taken to the finishing table and the superfluous
material between the grooves removed by a chisel, plane, or spokeshave
.-
PLATE 3.




wº £I —
until the grooves have almost disappeared. It is finally finished with a
scraper and when the grooves have just disappeared and the surface is
fair, it receives a final burnishing. The position of the desired load or any
other water line is now marked upon it at different points and check
measurements are taken to see if the model is correct to the drawings.
It is then carefully weighed and placed in the tank. The amount of bal-
last, which consists of shot bags, necessary to bring the model down to the
desired load line, is also calculated and weighed out. This is now put
into the model and moved until the latter floats upright and in her correct
trim. This is the final check that the model receives, as it should float
at its correct waterline when the calculated amount of ballast has been
added.
The attachment for the towing and steering rods are now made and
the model connected up to the dynamometer.
The models used are from ten to twelve feet long, and may repre-
sent a vessel of any size. In order to predict the resistance of a full
sized ship from that of a model, Froude's “law of comparison” is used.
The total resistance of a vessel consists of two parts, surface friction and
wave making resistance. The frictional resistance may be calculated for
any kind of surface whether parraffine, paint, copper, etc., from the knowl-
edge of the corresponding coefficients of friction of these surfaces. If
this be subtracted from the total, the remainder will be the wave making
resistance. According to Froude's law, this resistance varies as the cube
of the lineal dimensions at corresponding speeds, i. e., at speeds varying
as the square root of the lineal dimensions. For example, if a model were
one forty-ninth the full size and were towed at a speed of three knots per
hour, the corresponding speed of the full sized ship would be 7 × 3 = 21
knots. If at this speed of three knots the wave making resistance of the
model were say three pounds, the wave making resistance of the
full sized ship would be 3 × (49)* = 352,947 pounds (at 21 knots). To
this must be added the frictional resistance, calculated as explained above,
in order to obtain the total resistance of the vessel.
It is therefore possible to experiment with models at fairly low
speeds and yet obtain results which correspond with high speeds in the
full sized ship.
The subject of resistance of ship's forms is one which requires a
large amount of investigation, and it is expected that the work done at the
University will help towards the solution of many problems and prove a
benefit to the profession of Naval Architecture as a whole.
|
| | | |
||
|
CNI
LO
OO
O
<†
<†
OO
！=
O
LO
<！--
CD
O
cro

ſaeae
§§§§§）:
; : • • • § • ¶
：ſº