Chapter 8

PROPULSION MODEL EXPERIMENTS

8.1 Introduction

The performance of a propeller in a ship and the resulting hull-propeller

interaction are usually determined through model experiments. Experiments with full-

size ships and propellers are difficult to carry out and purely theoretical approaches are

as yet inadequate for predicting the behaviour of ships and propellers fully. The

conditions under which model experiments can be used to determine ship performance

have been considered in Chapter 4. It is not practicable to fulfil all these conditions

exactly, and therefore some empirical corrections are necessary in determining ship

performance from experiments with models.

The experiments that are normally carried out with ship models and model

propellers are :

(i) resistance experiments,

(ii) open water experiments,

(iii) self-propulsion experiments,

(iv) wake measurements, and

(v) cavitation experiments.

Model experiments in waves are not considered here. Resistance experiments are also

not really within the scope of the present work but are discussed briefly because they

are essential for analysing self-propulsion experiments. The resistance, open water and

self-propulsion experiments are normally carried out in towing tanks. Cavitation

experiments require a cavitation tunnel or a depressurised towing tank. Wake

measurements may be carried out in a towing tank or in a circulating water channel.

A towing tank is a long narrow tank in which a ship model is towed through

water at a steady speed usually by a towing carriage running on rails along the sides of

the tank. Instruments are provided to measure the speed and the resistance of the ship

model, and the thrust, torque and revolution rate of a model propeller if fitted. In a

circulating water channel, the ship model is stationary and the water is made to move

past it at a steady speed. A cavitation tunnel, which is discussed in more detail in Sec.

8.6, is similar except that the pressure in the tunnel water can be varied over a wide

range.

8.2 Resistance Experiments

Resistance experiments are carried out with ship models to determine the

resistance of the model and thereby of the ship in a given condition. Such experiments
are useful in optimising the hull form and for predicting the power required for

propelling a ship at a specified speed. Before discussing resistance experiments, it is

necessary to consider the relationship between the resistance of a ship and the resistance

of a geometrically similar model.

It can be shown in a manner similar to that described in Sec. 4.3 that the total

resistance RT of a ship can be expressed in the form :

(8.1)

where L is the length of the ship, V its speed, g the acceleration due to gravity, and

and the density and kinematic viscosity of water respectively. This may be written in

terms of non-dimensional parameters as follows :

(8.2)

where :

is the total resistance coefficient,

is the Reynolds number,

is the Froude number, and

S is the wetted surface of the ship, proportional to L2.

 Following W. Froude, it is usual to divide the total resistance into a viscous

resistance component which is a function of the Reynolds number and a residuary

resistance component which is a function of the Froude number. This is expressed in

terms of a viscous resistance coefficient CV and a residuary resistance coefficient CR :

(8.3)

The viscous resistance coefficient is determined with the help of a “friction line”, i.e. an

equation giving the resistance of a plane surface in terms of the Reynolds number.

Several friction lines have been proposed, based on theoretical considerations and

experimental measurements, but the standard formulation in use at present is the ITTC

(International Towing Tank Conference) 1957 friction line :

(8.4)

where CF is the frictional resistance coefficient. (The ITTC friction line is not strictly a

“two-dimensional” friction line.) The viscous resistance coefficient and the frictional

resistance coefficient are related by a “form factor” , which accounts for the

increase in viscous resistance due to the three-dimensional hull form of a ship :

(8.5)
One therefore has the following expression for the resistance of a ship :

(8.6)

It is then assumed that the residuary resistance, which for most ships is mainly wave

making resistance, obeys the Froude law : the residuary resistances of geometrically

similar ships are proportional to their displacements if their speeds are proportional to

the square roots of their lengths. This implies that :

if FnS = FnM (8.7)

where the suffix refers to the ship and to a geometrically similar model, and is

the Froude number. This allows the total resistance of a ship at a given speed to be

estimated from the total resistance of a model at the “corresponding” speed (the speed

for which the Froude numbers of the model and the ship are equal), as illustrated by the

following example.

Example 1

A ship of length 100 m and wetted surface of 2700 m 2 has a speed of 15 knots. A 4.0 m

long geometrically similar model of the ship has a total resistance of 25.0 N at the

corresponding speed. Determine the total resistance of the ship assuming a form factor of 1.05.

LS = 100 m SS = 2700 m2 VS = 15 k = 7.716 m s-1

LM = 4.0 m SM = m2

m s-1

RTM = 25.0 N 1 + k = 1.05

since

=
 =

= 250.7 kN

The total resistance RT or the total resistance coefficient CT determined in this

manner must be corrected for the differences between the conditions of the model

experiment and the operating conditions of a ship. A simple way to carry out this

correction is to add a “correlation allowance” CA to the total resistance coefficient of the

ship. The correlation allowance, which primarily accounts for the roughness of the ship

surface, is determined by correlating the performance of ships on speed trials and the

results of the corresponding model experiments. Other differences between the

operating conditions of the ship and the controlled conditions of the model experiments

are taken into account by adding a “service allowance” or a “trial allowance” to the

resistance of a ship determined through model experiments.

More detailed methods may also be used to correct the resistance for differences

between the model and the ship. Three such corrections are used in the standard ITTC

ship performance prediction method (ITTC 1978) :

(i) The wetted surface of the ship hull is normally rough whereas the model

surface is smooth. It is therefore necessary to add to the viscous

resistance coefficient of the ship a “roughness allowance” given by :

 (8.8)

where is the average amplitude of roughness of the wetted surface of

the ship. A standard value of m is used in the ITTC

method.

(ii) It is usually impracticable to reproduce bilge keels on a ship model, and

therefore a correction is required. This correction is made by increasing

the wetted surface of the ship hull by the surface area of the bilge

keels in calculating the viscous resistance of the ship

(iii) It is also necessary to account for the resistance of the above-water part

of the ship. This “air resistance” can be determined by model tests in a

wind tunnel. For a ship moving in still air, the ITTC performance

prediction method uses an air resistance coefficient as follows :

(8.9)

where is the air resistance, the air resistance coefficient and

the transverse projected area of the above water part of the ship.

After making these three corrections, the total resistance coefficient of the ship

is given by :
 (8.10)

Eqn. (8.10) gives the total resistance coefficient of a ship in ideal conditions. To allow

for the actual conditions during ship trials or in service one may multiply the total

resistance coefficient by a “load factor” (1+ x), the “overload fraction” x corresponding

to a “trial allowance” or a “service allowance”, ranging from 10 to 40 per cent

depending upon the type of ship and its service route.

For resistance experiments, it is first necessary to make an accurate

geometrically similar model to an appropriate scale, which should be large enough for

making the ship model and model propellers of sufficient accuracy and yet be within the

limits imposed by the size of the towing tank and the speed of its towing system. Large

models are also necessary to ensure turbulent flow around the model, and even then it is

necessary to fit artificial turbulence stimulators at the bow of the model. The model is

then accurately ballasted to the required waterline and towed at a number of steady

speeds in still water, the speeds covering slightly more than the normal operating speed

range of the ship. The tow force or resistance is measured by a resistance dynamometer

at each speed, and the results analysed in a manner similar to that given in Example 1.

In addition to speed and resistance, it is also necessary to measure the temperature of

the towing tank water, since the density and viscosity of water depend on its

temperature. Resistance experiments may be carried out on a model without appendages

such as rudders and shaft brackets, in which case one obtains the “naked hull”

resistance. The model may also be tested when fitted with appendages made to the same
scale. If only a naked hull resistance test is carried out, it is necessary to add an

empirical appendage resistance to the naked hull resistance.

8.3 Open Water Experiments

The object of an open water experiment is to determine the open water

characteristics of a propeller. A geometrically similar model is made of the propeller,

the size of the model propeller being governed by the size of the ship model if it is

intended to use the model propeller for self-propulsion tests also. If only open water

tests are to be carried out, when generating propeller methodical series data for

example, the model propeller is made somewhat larger, its size depending upon the

capacity of the propeller dynamometer available. In any case, model propellers should

be large enough to be made accurately and to have a sufficiently high Reynolds number

so that the flow is turbulent. A further measure to promote turbulent flow is not to give

the model propeller a highly polished surface. It is recommended that the model

propeller Reynolds number based on the resultant of the axial and tangential velocities

and the section chord at 0.7R be at least 3.2 105; otherwise turbulence should be

artificially stimulated, e.g. by roughening the leading edges of the blades.

For the open water experiment, the model propeller is attached to a propeller

dynamometer fitted in an “open water boat” as shown in Fig. 8.1. The propeller

dynamometer measures the thrust and torque of the propeller. The propeller shaft
extends a sufficient length forward from the boat to ensure that the flow around the

propeller is not disturbed by the boat. A fairing cap is provided at the forward end of the

propeller boss. The open water boat is ballasted so that the propeller shaft is horizontal

and its depth below the water surface is at least equal to the model propeller diameter.

The open water experiment is conducted by towing the open water boat at a

steady speed while running the propeller at a constant revolution rate. The speed of the

boat (i.e. speed of advance VA), and the revolution rate n, thrust T and torque Q of the

propeller are measured in each run. It is usual to run the model propeller for all the

speeds of advance at a constant revolution rate, this preferably being the highest that

can be attained at zero speed of advance at which the torque is maximum and which

must be within the capacity of the driving motor and dynamometer. The speed of

advance is varied in steps from zero to the value at which the propeller thrust just

becomes negative. (Four quadrant measurements involving negative speeds and

revolution rates are not considered here.) The measured thrust and torque are corrected

for the “idle” thrust and torque, i.e. the thrust and torque measured by the dynamometer

when the experiment is carried out with a dummy boss of equal weight replacing the

propeller.

The open water characteristics of the model propeller can be easily calculated

from the measured values of VA and n, and the corrected values of T and Q. The open

water characteristics of the ship propeller will be slightly different because of the

difference between the Reynolds numbers of the model propeller and the ship propeller.
The surface of the ship propeller may also be rough. A procedure for correcting the

open water characteristics for Reynolds number and roughness effects has been given in

the ITTC 1978 performance prediction method. This procedure, which is based on the

assumption that the blade section at 0.75R can be assumed to represent the propeller as

a whole, is summarized by the following equations :

= (8.11)

where :

= resultant velocity of the blade section at 0.75R

 = speed of advance of the model propeller

= revolution rate of the model propeller

= model propeller diameter

= expanded blade widths (chords) of the section at 0.75R of the

model propeller and the ship propeller respectively.

= kinematic viscosity of water for the model experiment

= Reynolds number of the model propeller

= drag coefficients of the blade section at 0.75R for the model

propeller and the ship propeller respectively

= thickness chord ratio of the blade section at 0.75R

= chord diameter ratio at 0.75R

Z = number of blades

= propeller surface roughness (standard value m)

 = thrust coefficients of the model propeller and the ship propeller

respectively

= torque coefficients of the model propeller and the ship propeller

respectively.

Example 2

In an open water experiment with a model propeller of 200 mm diameter, a thrust of

252 N and a torque of 9.250 N m are measured with the propeller running at 2400 rpm at a

speed of advance of 5.6 m per sec. The idle thrust and torque (to be subtracted from the

measured values) are – 4.0 N and 0.034 N m respectively. Determine the thrust, torque and

advance coefficients of the model propeller and its open water efficiency. The ship propeller has

four blades, a diameter of 5.0 m and a constant pitch ratio of 0.8. The blade section at 0.75 R has

a thickness of 0.0675 m and a chord of 1.375 m. If the roughness of the propeller blade surface

is 30 microns, determine the thrust and torque coefficients of the ship propeller and its open

water efficiency.

DM = 200 mm = 0.200 m TM = 252 – (- 4.0) = 256 N

nM = 2400 rpm = 40.0 s-1 QM = 9.250 – 0.034 = 9.216 N m

= m s-1 kg m-3 = m2 s-1

m = m at 0.75R = m

kp = 30 m = 30 m
 =

= m

= m2 s-2

VRM = 19.6638 m s-1

=
 =

8.4 Self-propulsion Experiments

Self-propulsion experiments are used to determine the performance of the ship

hull and propeller taken together. An analysis of the results of a self-propulsion

experiment allows one to predict the delivered power and the revolution rate of the ship
propeller at a given speed of the ship, and to determine the wake fraction, thrust

deduction fraction and relative rotative efficiency. The data of the resistance and open

water experiments are, however, required for analysing the data of the self-propulsion

experiment.

For a self-propulsion experiment, the model propeller is fitted in its correct

position at the stern of the ship model and connected to a propeller dynamometer for

measuring the thrust and torque of the propeller at various revolution rates. The ship

model should be fitted with all appendages as far as possible, particularly those lying in

the propeller slipstream, e.g. a rudder. The ship model is attached to a resistance

dynamometer, which in this test measures the force required to make the ship model

move at a constant speed with the propeller running. The ship model is accurately

ballasted so that it floats at the correct waterline. The model is then towed at a steady

speed with the propeller running at a constant revolution rate, and the thrust and torque

of the propeller and the force applied to the ship model through the resistance

dynamometer are measured. Fig. 8.2 illustrates the experimental set up.

The analysis of the data recorded in a self-propulsion experiment is based on the

following considerations. When a ship is moving at a steady speed VS with its propeller

producing a thrust TS, then :

(8.12)
where is the resistance of the ship at the speed VS and t the thrust deduction

fraction. If the model scale is , and DM and DS are the model and ship propeller

diameters :

(8.13)

The model must be tested at a speed VM such that the Froude numbers of the model and

the ship are equal, which implies that :

(8.14)

If during the self-propulsion experiment at the speed VM the model propeller

runs at a revolution rate and has a thrust TM and torque QM while the force applied

to the model through the resistance dynamometer is F, then :

(8.15)

being the resistance of the model at the speed VM. This may be written as :

(8.16)
where is the thrust coefficient of the model propeller in the behind condition, and

the density of water in which the experiment is carried out. If the model propeller

and the ship propeller are to fulfil the conditions of dynamic similarity :

(8.17)

where and are the advance coefficients of the ship propeller and the model

propeller, and the corresponding wake fractions, and the thrust coefficient

of the ship propeller in the behind condition. Using Eqns. (8.13) and (8.14) with the first

equation of the Eqns. (8.17), one obtains :

(8.18)

and then one may write Eqn. (8.16) with the help of Eqns. (8.12) and (8.13) as follows :

= (8.19)
In terms of the total resistance coefficients of the model and the ship, this becomes :

F =

= (8.20)

Eqns. (8.19) and (8.20) show that for dynamic similarity as indicated by Eqn. (8.17), the

ship model is not fully self-propelled, and the thrust of the model propeller must be

augmented by the force F applied through the resistance dynamometer. The condition

represented by Eqn. (8.19) or (8.20) is termed the “ship self-propulsion point on the

model” in contrast to the “model self-propulsion point” at which F = 0 and the model is

fully self-propelled by its propeller. It may be noted that whereas at the ship

self-propulsion point , at the model self-propulsion point is greater

than (at least for x = 0) so that the model propeller is overloaded in comparison to

the ship propeller and is therefore operating less efficiently. It is possible for F to be

negative if the overload fraction x is sufficiently large.

If the self-propulsion experiment is to be carried out for only one value of the

load factor (1+ x), then for each speed VM the value of the force F is calculated in

advance using Eqns. (8.19) or (8.20), and with the model moving steadily with the
speed VM the propeller revolution rate is adjusted until the calculated value of F is

achieved. The thrust TM and the torque QM are then measured along with VM, F and .

This procedure may be repeated for several speeds to cover a range corresponding to the

operating speed range of the ship. This method of conducting a self-propulsion

experiment is called the “constant loading” method or the “Continental” method

because of its use in European ship model tanks. If self-propulsion test data are required

for a different loading, the experiment is repeated for a new value of F at each speed.

In an alternative method of carrying out self-propulsion experiments, called the

“constant speed” method or sometimes the “British” method, tests are carried out for

several values of at each value of VM, and F, TM and QM are measured. This

allows the data to be analysed for different load factors (1+ x), which may be selected

after the experiment is over. In actual practice, the constant loading method for different

loadings and the constant speed method for different speeds are equivalent since both

lead to values of F, TM and QM as functions of both VM and .

The analysis of the data of a self-propulsion experiment requires the resistance

of the model and the resistance of the ship derived from it, as well as the open water

characteristics of the model propeller obtained through the open water experiment. The

resistance of the model should be for the same model condition as in the self-propulsion

experiment. This may involve correcting the model resistance data for any temperature

difference between the resistance experiment and the self-propulsion experiment. The

analysis then proceeds as described in the following.

 For each model speed VM and the corresponding ship speed VS, the required

force F is determined from the model resistance and the ship resistance

using Eqn. (8.19). The model propeller revolution rate for this value of F, and the

corresponding values of TM and QM are then obtained, directly in the Continental

method, and by interpolation if necessary in the British method. These are the values of

, TM and QM at the ship-propulsion point on the model for the speed VM. TM and

QM are converted to the thrust and torque coefficients and for the behind

condition :

= (8.21)

Referring to the open water characteristics of the model propeller (KTM, KQM and

as functions of J), one determines :

(a) the values of J, KQM and for , i.e. for thrust identity

(b) the values of J, KTM and for , i.e. for torque identity.

Then :
 w =

t =

= (8.22)

= (thrust identity)

= (torque identity)

There are usually small differences between the values of w, and obtained using

thrust identity and the values obtained using torque identity, and the two sets of values

are identified by the suffixes and , e.g. and are the wake fractions

determined by thrust identity and torque identity respectively. The values of t and

with thrust identity are the same as with torque identity.

One may also determine the revolution rate and delivered power of the ship

propeller from the following equations :

 =

= (8.23)

The values of obtained by Eqns. (8.22) and (8.23) are identical.

The analysis of self-propulsion experiment data may be carried out using non-

dimensional coefficients throughout. If the wake fraction, thrust deduction fraction and

relative rotative efficiency are not required, the open water data of the model propeller

are not required, and one may directly calculate the delivered power and revolution rate

of the ship propeller from the self-propulsion experiment data using the Eqns. (8.23).

The procedure for analysing self-propulsion experiments described in the

foregoing ignores the differences that exist between the model and the ship. These
differences may cause the values of the delivered power and propeller revolution rate of

the ship determined from the self-propulsion experiment to differ from the values

obtained during the speed trials of the ship. A simple way to allow for these differences

is to introduce correlation factors based on experience with previously built ships into

the values predicted from self-propulsion experiments :

(8.24)

where and are correlation factors for the propeller revolution rate and delivered

power respectively.

Example 3

The effective power of a single screw ship of length 100 m at a speed of 15 knots is

2150 kW. The resistance of a 4.0 m long model with appendages is 25.4 N at the corresponding

speed. Determine the tow force to be applied to the model in a self-propulsion test at this speed

to obtain the ship self-propulsion point. The model propeller of diameter 0.2 m is found to run

at 534 rpm for this tow force to be obtained, and the propeller thrust and torque are then 21.75

N and 0.682 N m. The open water data of the model propeller are :

J :

:
Analyse these data and determine the rpm of the ship propeller and the delivered power given

that the correlation factors for rpm and delivered power are respectively 1.02 and 1.00.

= 100 m = 2150 kW = = 7.716 m s-1

= 4m = 25.4 N = 534 rpm = 8.9 s-1

= 0.2 m = 21.75 N QM = 0.682 N m

 = = m s-1

DS = m RTS = kN

F = N

= N

t =

Thrust Identity Torque Identity

KT = KTBM = 0.1716 KQ = KQBM = 0.02691

J = 0.6500 KQM = 0.02801 J = 0.6678 KT = 0.1618

= 0.6338 = = 0.6390

= =

= 1 – 0.7497 = 0.2503 = 1 – 0.7703 = 0.2297

= =

= =

= 0.7039 = 0.7037

= rpm

= 3054.41 kW

Ship prediction:

= rpm
 = kW

This neglects the differences between the open water characteristics of the model and ship

propellers.

In the ITTC ship performance prediction method, a specific correction for the

difference between the wake fractions of the ship and the model is introduced :

= (8.25)

where and are the wake fractions of the ship and the model based on thrust

identity, and and are the viscous resistance coefficients given by :

(8.26)

The quantity 0.04 associated with the thrust deduction fraction t accounts for the thrust

deduction contributed by a rudder in the propeller slipstream, and should be omitted if

the rudder is not in the propeller slipstream, e.g. when there is a single rudder on the

centre line of a twin screw ship. The thrust deduction fractions of the ship and the
model are assumed to be equal. Should Eqn. (8.25) yield a value of greater than

, should be taken equal to .

The propeller loading at the ship self-propulsion point is then determined as

follows :

i.e.

which reduces to :

= (8.27)

where is the thrust coefficient of the ship propeller in open water, and :

The value of is first obtained from Eqn. (8.27). The values of , and

corresponding to this value of are then determined from the open water

characteristics of the ship propeller (i.e. after correcting the characteristics of the model

propeller for the difference in drag coefficients of the model propeller and the ship
propeller.) The revolution rate of the ship propeller and the delivered power are then

easily obtained :

(8.28)

These are values for the ship in conditions corresponding to the ideal conditions of

the model experiments. The propeller revolution rate and delivered power for a

ship during trials are obtained by applying correlation factors or allowances, for which

two alternative methods have been proposed :

Method 1: = = (8.29)

This is similar to the method described earlier.

Method 2 : In this, correlation allowances are applied to the resistance coefficient

and the wake fraction of the ship for calculating the propeller loading :
 = (8.30)

This value is used to determine , KTS and from the open water characteristics of

the ship propeller, and then :

(8.31)

The correlation factors CN and CP and the correlation allowances and are

obtained by analysing the data from previous model experiments and ship trials.

The analysis of self-propulsion experiments described in the foregoing is for

single screw ships. For twin screw ships, the analysis is identical except that the

resistance is shared equally by the two propellers and this must be taken into account.

Self-propulsion experiments for ships with more than two propellers are complicated

because there is an additional variable involved, viz. the distribution of loading among

the different propellers.

8.5 Wake Measurements

Although the wake fraction is determined indirectly through a self-propulsion test,

it is desirable to observe the flow past a ship model at the position of the propeller to

determine the distribution of wake velocity over the propeller disc directly. Wake

measurements are also carried out to determine the radial and circumferential

distribution of wake velocity. Such measurements are useful for propeller design and

for calculating unsteady propeller forces, and could indicate unacceptable flow

conditions at the stern, requiring changes in the hull form of the ship.

Direct methods of wake measurement make use of devices such as wake wheels

and pitot tubes, Fig. 8.3, with a ship model in a towing tank or circulating water

channel. These devices can only be used when the propeller is not fitted to the ship

model, and hence they yield only the nominal wake (Sec. 5.2). Other disadvantages of

these devices are that their presence affects the flow around the ship model and that

their response is slow.

The wake wheel consists of a number of small vanes attached to radial spokes

from a hub mounted on a shaft supported by very low friction bearings. The flow causes

the wake wheel to revolve at a speed proportional to the flow velocity, and hence the

average circumferential flow velocity at a given radius can be obtained. Repeated

measurements with the vanes set at different radii provide the radial variation of wake.

The pitot tube measures the total (hydrostatic plus hydrodynamic) pressure at a point,
and by subtracting the hydrostatic pressure the hydrodynamic pressure is obtained

which is proportional to the square of the flow velocity. Pitot-static tubes allow the total

and the hydrostatic pressures to be measured together. By having a “wake rake”

consisting of a series of pitot tubes mounted on a streamlined holder and connected to a

manometer bank, wake measurements at several points can be obtained simultaneously.

Pitot tubes however can be used only for the measurement of the axial velocity

component of the flow. For obtaining all the three components of flow velocity at a

point, it is necessary to use a five-hole spherical head pitot tube. Wake wheels, pitot

tubes and five-hole pitot tubes are calibrated by running them in undisturbed water at

known velocities. The disadvantages of these wake measurement devices are eliminated

in laser doppler velocimeters, which allow velocity measurements to be made without

disturbing the flow, are very accurate and have a quick response. Advanced laser

doppler velocimeters allow all three velocity components to be measured

simultaneously.

The results of wake measurements may be presented in various forms, Fig. 8.4.

Wake wheel measurements may be used to determine the wake fraction at various radii,

and the results presented in a form showing the wake fraction as a function of radius.

Pitot tube measurements may be presented in the form of diagrams showing the wake

fraction at various points of the propeller disc at equal angular and radial spacing, or

contours of constant wake fraction. If all the components of wake velocity are

determined, separate diagrams may be prepared for each component.

 Wake measurement data may be analysed to obtain the average wake fraction or

the mean circumferential wake fraction at each radius. If v(r, ) is the axial velocity at

any point (r, ) of the propeller disc, the average nominal wake fraction w is given by :

(8.32)

where V is the speed of the model, R the radius of the propeller, the boss radius, and

the axial velocity at the radius r averaged over the circumference (i.e. the value

that would be found by a wake wheel at the radius r). Alternatively, one may write :

(8.33)

where w(r,) is the local wake fraction at the point (r,) and the average

circumferential wake fraction at the radius r. The average wake fraction obtained in this
way is the “volumetric mean”, i.e. both sides of the Eqns. (8.32) and (8.33) represent

the volume of water flowing through the propeller disc per unit time.

For studying unsteady propeller forces and propeller excited vibration, the local

wake fractions w(r,) may be expressed in the form of a Fourier Series :

w(r, ) = (8.34)

the coefficients and being found by a least squares fit and the series being

truncated after, say, twenty terms.

Example 4

The local wake fractions obtained by measurements at various angular and radial

positions at the location of the propeller in a single screw ship model are as follows :

 deg : 0 30 60 90 120 150 180

0.2 0.5832 0.4690 0.3426 0.3013 0.2843 0.3524 0.4082

0.3 0.5622 0.4521 0.3303 0.2905 0.2741 0.3397 0.3935

0.4 0.5328 0.4285 0.3130 0.2753 0.2597 0.3219 0.3730

0.5 0.4950 0.3980 0.2908 0.2558 0.2413 0.2991 0.3465

0.6 0.4489 0.3609 0.2637 0.2319 0.2188 0.2712 0.3142

0.7 0.3942 0.3170 0.2316 0.2037 0.1922 0.2381 0.2759

0.8 0.3312 0.2663 0.1946 0.1711 0.1615 0.2001 0.2318

 0.9 0.2598 0.2089 0.1526 0.1342 0.1267 0.1570 0.1819

1.0 0.1800 0.1448 0.1058 0.0930 0.0878 0.1088 0.1260

Determine the average circumferential wake fraction at each radius and the average wake

fraction over the propeller disc.

For a single screw ship, the flow normally has port and starboard symmetry. The

average circumferential wake at any fractional radius x = r/R :

For x = 0.2, using integration by Simpson’s Rule :

SM f( )

___ ______ ___ ______

deg

0 0.5832 1 0.5832

30 0.4690 4 1.8760

60 0.3426 2 0.6852

90 0.3013 4 1.2052

120 0.2843 2 0.5686

150 0.3524 4 1.4096

180 0.4082 1 0.4082

 6.7360

where 1/3 is the common multiplier in Simpson’s Rule and radians (30 deg) is the

spacing between the values of .

The average circumferential wakes at the other radii are similarly obtained as :

The average wake fraction over the whole propeller disc is given by :

In the present case = 0.2, so that :

Using integration by Simpson’s Rule :

x SM f( )
 ___ ______ ________ ___ _____

0.2 0.3742 0.07484 1 0.07484

0.3 0.3607 0.10821 4 0.43284

0.4 0.3419 0.13676 2 0.27352

0.5 0.3176 0.15880 4 0.63520

0.6 0.2880 0.17280 2 0.34560

0.7 0.2529 0.17703 4 0.70812

0.8 0.2125 0.17000 2 0.34000

0.9 0.1667 0.15003 4 0.60012

1.0 0.1155 0.11550 1 0.11550

3.52574 .

8.6 Cavitation Experiments

Cavitation experiments with propellers are carried out to study propeller

cavitation and determine its effects on propeller performance. In addition to observing

cavitation patterns on propeller blades and the effect of cavitation on the open water

characteristics, such experiments allow one to study cavitation erosion and noise due to

cavitation. Although it is possible to study cavitation on full size propellers in ships

with the aid of underwater television cameras, cavitation experiments are usually

conducted with model propellers in cavitation tunnels. There is also a depressurised

towing tank at MARIN, the Netherlands, which is a unique facility for studying

cavitation.

A variable pressure recirculating water tunnel, usually called a “cavitation

tunnel”, Fig. 8.5, has a shell made up of a number of pieces bolted together, an impeller

to circulate the water in the tunnel, an air pump to control the pressure in the tunnel, and

the equipment associated with the model whose cavitation characteristics are to be

studied. The tunnel is normally erected in a vertical plane with the impeller at one end

of the lower horizontal limb, and the working section in which the model is placed in

the upper limb. The flow is guided around the corners at the junctions of the horizontal

and vertical limbs by guide vanes, and the rotation of the flow caused by the impeller is

reduced to a great extent by a “honeycomb” section. A contraction section just ahead of

the working section increases the flow velocity to the desired value. A diffuser after the

working section decelerates the flow and increases the pressure at the impeller position

to minimise the possibility of impeller cavitation. A pressure coaming at the highest

point of the tunnel is connected to an air chamber from which air may be evacuated by a

vacuum pump to allow the pressure of air above the water in the tunnel to be reduced.

The pressure can also be increased to a high value to simulate the operating conditions

of a submarine at great depth.

For cavitation experiments with propellers, a model propeller is placed in the

working section and attached to a propeller dynamometer, which can measure the thrust

and torque of the propeller while it runs at a steady revolution rate. The working section
has large airtight windows to allow cavitation to be observed with the help of

stroboscopic lighting. Instruments are provided to measure the speed of water in the

working section and the pressure. Non-uniform wake conditions may be simulated by

providing flow regulators, wire mesh screens or partial ship models ahead of the

propeller. The non-uniform velocity field is measured by a wake rake and the flow

regulating devices adjusted until the desired velocity field is obtained.

It has been seen that cavitation is affected by the amount of air dissolved in

water. An instrument to measure air content is therefore provided. However, air

liberated from water by cavitation does not easily go back into solution and the air

bubbles circulate with the water in the tunnel, interfering with the cavitation phenomena

and their observation. This problem is usually overcome by removing the air dissolved

in the tunnel water. Alternatively, the tunnel may be provided with a “resorber” in

which the tunnel water is circulated at low velocity and high pressure to force the air

back into solution. The reduced air content and the purity of the tunnel water in

comparison with sea water cause substantial differences between the cavitation

phenomena observed in a cavitation tunnel and those occurring with full scale

propellers at sea. Better correlation between the cavitation of a model propeller in a

cavitation tunnel and of the ship propeller at sea is obtained by reducing the cavitation

number of the model propeller in comparison with the ship propeller by as much as 15-

20 per cent when air is removed from the tunnel water. Cavitation nuclei may be

artificially introduced by electrolysis in the tunnel water to increase its similarity with

sea water in which microscopic particles act as nuclei.

 In carrying out propeller cavitation experiments in a cavitation tunnel, the thrust

and torque of the model propeller are measured for different speeds of advance,

revolution rates and pressures so that the complete range of advance coefficient and

cavitation number is covered. In addition, the temperature of the tunnel water (on which

vapour pressure depends) and its air content are also measured. Cavitation patterns are

observed and recorded using powerful stroboscopic lighting and high speed

photography. Noise measurements using hydrophones may also be carried out.

Cavitation erosion may be studied by covering the propeller blade surfaces with a

suitable coating that is easily eroded by the collapsing cavities.

The results of cavitation experiments are usually presented in terms of the thrust

and torque coefficients as functions of the advance coefficient for different cavitation

numbers, as shown in Fig. 6.4. Sketches showing the type and extent of cavitation for

different advance coefficients and cavitation numbers or photographs may also be

given. Zones of different types of cavitation may be shown in a diagram similar to Fig.

6.3. Cavitation noise and erosion parameters may be determined for given advance

coefficients and cavitation numbers.

In addition to cavitation experiments with model propellers, cavitation tunnels

are also used to study cavitation on other bodies such as hydrofoils using appropriate

instruments and equipment.

 The variable pressure towing tank of MARIN is 240 m long and 18 m wide, and

has water up to a depth of 8 m. The towing carriage is driven by a cable towing system.

The air pressure in the tank can be brought down to about 4 kN per m 2. The

performance of the model propeller can be seen through a periscope by observers in a

chamber in which normal atmospheric pressure is maintained and which is sealed off

when the pressure in the tank is reduced.

Example 5

A ship has a propeller of 4.0 m diameter with its centre line 3.0 m below the surface of

water. The propeller in its design condition has a speed of advance of 8.0 m per sec at 150 rpm

and 2500 kW delivered power. Experiments are to be carried out with a scale model

propeller in a cavitation tunnel in which the maximum speed of water in the working section is

10.0 m per sec, the maximum propeller rpm is 3000, the maximum torque is 45 N m and the

minimum static pressure at the centre line of the working section is 15 kN per m 2. Determine

the speed of water, the model propeller rpm and the static pressure at the propeller position if

the model propeller is to be run (a) at the correct Froude number, and (b) just within the limits

of the cavitation tunnel. The tunnel water is deaerated and the tunnel cavitation number must be

reduced by 15 per cent compared to the ship. What is the minimum cavitation number of the

tunnel based on the speed of water in the working section?

DS = 4.0 m nS = 150 rpm = 2.5 s-1 VAS = 8.0 m s-1 PDS = 2500 kW

hS = 3.0 m = 16 DM = DS / = 4.0/16 = 0.25 m

VAM max = 10.0 m s-1 nM max = 3000 rpm = 50 s-1 QM max = 45 N m

pTunnel min = 15 kN m-2

 =

= 3.9569

(a) At the correct Froude number

= m s-1

= s-1 J M = JS

= kN m-2

(which is below the minimum limit of pressure)

(b) At the limits of the cavitation tunnel

If = m s-1 :

= s-1

=
 = 0.05923 kN m = 59.23 N m, which is above the limit.

Hence, keeping QM = 45 N m, the limiting value,

= 43.5824 s-1

= m s-1

= 129.475 kN m-2

The minimum cavitation number of the tunnel is :

Problems

1. A resistance experiment is carried out on a 5.0 m long model of a ship of length

120 m and wetted surface 3200 m2. The following results are obtained :

Model speed, m per sec : 1.050 1.260 1.470 1.680 1.890

Model resistance, N : 12.13 17.78 25.45 36.29 52.02

 Determine the effective power of the ship at the corresponding speeds using the

ITTC friction line with a form factor of 1.06 and a correlation allowance of

0.0004.

2. A ship of length 150 m has a transverse projected area above water of 403 m 2, the

wetted surface of the hull being 6200 m 2. The bilge keels have a wetted surface of

260 m2. The roughness of the hull surface is 150 microns. The resistance of a

model of length 6.0 m as measured during a resistance test is as follows :

Model speed, m per sec : 1.235 1.440 1.646 1.852 2.058

Model resistance, N : 26.99 38.45 54.54 77.71 111.56

Determine the effective power of the ship at the corresponding speeds using the

ITTC performance prediction method with a form factor of 1.05.

3. An open water experiment covering the complete range of advance coefficient is to

be carried out on a model of a propeller of diameter 5.0 m and pitch ratio 1.0. The

effective pitch ratio of the propeller is expected to be about 1.08, and the torque

coefficient at zero speed of advance to be 0.0625. The maximum speed at which

the model propeller can be made to advance is 6.0 m per sec, and the maximum

rpm and torque of the propeller dynamometer are 3000 and 3.0 N m respectively.

Determine the maximum model propeller diameter if (a) the model propeller is to

run at a constant 3000 rpm for the complete range of advance coefficient, and (b)
 the model propeller is to run at 2400 rpm. If the minimum propeller diameter is to

be 150 mm, what is the maximum rpm ? Calculate for each case at zero advance

and zero slip the Reynolds number of the model propeller using the resultant

velocity at 0.75R, given that the chord diameter ratio is 0.28 for this radius.

4. An open water experiment is carried out with a four-bladed model propeller of

200 mm diameter and 0.8 pitch ratio. The blade section at 0.75R has a chord of

50 mm and a thickness of 2.5 mm. The model propeller is run at 2100 rpm over a

range of speeds and the following values of thrust and torque (corrected for idle

thrust and torque) are obtained :

Speed, m per sec : 0.000 0.700 1.400 2.100 2.800

Thrust, N : 666 612 551 484 412

Torque, N m : 15.29 14.54 13.57 12.38 10.98

Speed, m per sec : 3.500 4.200 4.900 5.600 6.300

Thrust, N : 333 247 156 59 -45

Torque, N m : 9.35 7.50 5.43 3.14 0.62

Determine the open water characteristics of the full size propeller of diameter

4.0 m, assuming that its surface roughness is 30 microns.

5. In a self-propulsion test, the model speed is 2.0 m per sec and at the ship self-

propulsion point the 0.2 m diameter model propeller runs at 900 rpm and the

thrust, torque and tow force are respectively 55.00 N, 1.33 N m and 2.55 N. The

resistance of the model at 2.0 m per sec is 49.03 N. The model propeller has open

characteristics as follows :

J : 0.300 0.400 0.500 0.600 0.700

KT : 0.210 0.182 0.152 0.120 0.086

10KQ : 0.253 0.225 0.195 0.164 0.132

Calculate the propulsive efficiency and its components on the basis of (a) thrust

identity, and (b) torque identity. If the model scale is 1 : 25, determine the

effective power, the delivered power and the propeller rpm of the ship at the

corresponding speed. Neglect the difference between the open water

characteristics of the model and ship propellers.

6. In a self-propulsion test with a twin screw ship model, the model is self-propelled at

a speed of 2.3 m per sec when the two propellers together have a total thrust of 34

N and a total torque of 1.18 N m at 720 rpm. In the open water test with either

propeller running at 720 rpm a thrust of 17 N is obtained at a speed of 2.0 m per

sec, the torque being 0.61 N m. The resistance of the model at a speed of 2.3 m

per sec is 29 N. Calculate the propulsive efficiency and its components. Are

these values applicable to the ship ?

7. A ship of length 100 m and a wetted surface of 2250 m 2 has a propeller of diameter

5.0 m. At its design speed of 20 knots, the ship has an effective power of 5000

kW. A self-propulsion test is carried out on a 1/25-scale model, which has a

resistance of 35 N at the corresponding speed. The following readings (after

correction) are obtained at this speed :

Propeller rpm : 500 550 600 650 700

Thrust, N : 27.15 30.84 35.83 42.22 50.81

Torque, N m : 0.75 0.93 1.42 2.08 3.10

Tow Force, N : 6.33 5.47 4.65 3.70 2.63

The open water data of the model propeller are as follows :

J : 0.700 0.800 0.900 1.000

KT : 0.255 0.224 0.191 0.153

10KQ : 0.508 0.468 0.420 0.368

Calculate the delivered power of the ship and its propeller rpm at 20 knots and the

propulsion factors using the ITTC ship prediction method with the following

values :
8. In a model of a single screw ship, the following average velocities over the

circumference at various radii at the propeller position are measured with the

model moving at a speed of 5.0 m per sec in undisturbed water :

Radius, mm : 20 30 40 50 60

Velocity, m per sec : 3.503 3.557 3.632 3.730 3.848

Radius, mm : 70 80 90 100

Velocity, m per sec : 3.988 4.150 4.333 4.538

Determine the nominal wake fraction, the model propeller having a diameter of

200 mm and a boss diameter ratio of 0.2.

9. In a twin screw ship model, the axial velocities measured at various radii r and

angular positions at the propeller disc when the model has a speed of 4.0 m per

sec are as follows :

0 30 60 90 120 150

r mm Axial velocity, v m per sec

15.0 3.600 3.680 3.792 3.880 3.960 3.960

22.5 3.552 3.636 3.788 3.880 3.956 3.956

30.0 3.488 3.596 3.764 3.876 3.952 3.948

37.5 3.440 3.552 3.760 3.876 3.952 3.936

 45.0 3.368 3.508 3.752 3.872 3.948 3.924

52.5 3.308 3.448 3.728 3.868 3.948 3.916

60.0 3.228 3.400 3.716 3.852 3.940 3.892

67.5 3.136 3.348 3.696 3.840 3.928 3.872

75.0 3.040 3.280 3.680 3.828 3.920 3.840

: 180 210 240 270 300 330

r mm Axial velocity, v m per sec

15.0 3.936 3.880 3.844 3.640 3.440 3.200

22.5 3.928 3.872 3.836 3.608 3.408 3.120

30.0 3.920 3.856 3.816 3.568 3.312 3.032

37.5 3.896 3.836 3.792 3.528 3.236 2.936

45.0 3.880 3.808 3.760 3.484 3.160 2.844

52.5 3.864 3.776 3.724 3.440 3.080 2.744

60.0 3.840 3.740 3.680 3.392 2.996 2.640

67.5 3.800 3.700 3.632 3.340 2.896 2.520

75.0 3.756 3.656 3.584 3.280 2.800 2.348

Determine the mean circumferential wake fraction at the various radii and the

nominal wake fraction. The model propeller is of 150 mm diameter with a boss

diameter of 30 mm.
10. A single screw ship has a design speed of 22 knots with its propeller running at 150

rpm and the delivered power being 12500 kW. The propeller has a diameter of

6.0 m and its centre line is 5.2 m below the water line. Model experiments are to

be carried out in a cavitation tunnel with a free surface working section using a

ship model to create a variable wake. The model propeller diameter is 150 mm.

Determine the speed of water in the working section ahead of the model, the

model propeller rpm and torque, and the pressure in the air above the water

surface in the working section for simulating the conditions in the ship propeller.

Assume that the cavitation number of the model propeller must be 10 per cent less

than that of the ship propeller.