Marine Propellers

and Propulsion
Second Edition

J S Carlton
Global Head of Marine Technology and
Investigation, Lloyds Register

AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD

PARIS SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO
Butterworth-Heinemann is an imprint of Elsevier
Butterworth-Heinemann is an imprint of Elsevier
Linacre House, Jordan Hill, Oxford OX2 8DP
30 Corporate Drive, Suite 400, Burlington, MA 01803, USA

First edition 1994

Second edition 2007

Copyright 2007, John Carlton. Published by Elsevier Ltd. All right reserved

The right of John Carlton to be identied as the authors of this work has been
asserted in accordance with the Copyright, Designs and Patents Act 1988

No part of this publication may be reproduced, stored in a retrieval system

or transmitted in any form or by any means electronic, mechanical, photocopying,
recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elseviers Science & Technology Rights
Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333;
email: permissions@elsevier.com. Alternatively you can submit your request online by
visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting
Obtaining permission to use Elsevier material

Notice
No responsibility is assumed by the published for any injury and/or damage to persons
or property as a matter of products liability, negligence or otherwise, or from any use
or operation of any methods, products, instructions or ideas contained in the material
herein. Because of rapid advances in the medical sciences, in particular, independent
verication of diagnoses and drug dosages should be made

British Library Cataloguing in Publication Data

Carlton, J. S. (John S.)
Marine propellers and propulsion. 2nd ed.
1. Propellers
I. Title
623.873

Library of Congress Number: 2007925588

ISBN: 978-07506-8150-6

For information on all Butterworth-Heinemann publications visit

our web site at http://books.elsevier.com

Printed and bound in Great Britain by MPG Books Ltd, Bodmin Cornwall

06 07 08 09 10 10 9 8 7 6 5 4 3 2 1

12.3.3 Direct model test
Model testing of a ship in the design stage is an import-
ant part of the design process and one that, in a great
many instances, is either not explored fully or is not
undertaken. In the authors view this is a false economy,
bearing in mind the relatively small cost of model testing
as compared to the cost of the ship and the potential costs
that can be incurred in design modication to rectify a
problem or the through life costs of a poor performance
optimization.
General procedure for model tests
Whilst the detailed procedures for model testing differ
from one establishment to another the underlying gen-
eral procedure is similar. Here the general concepts are
discussed, but for a more detailed account reference can
be made to Phillips-Birt (Reference 11). With regard to
resistance and propulsion testing there are fewer kinds
of experiment that are of interest: the resistance test, the
open water propeller test, the propulsion test and the
ow visualization test. The measurement of the wake
eld was discussed in Chapter 5.
Resistance tests
In the resistance test the ship model is towed by the
carriage and the total longitudinal force acting on the
model is measured for various speeds (Figure 12.22).
The breadth and depth of the towing tank essentially
governs the size of the model that can be used. Todds
original criterion that the immersed cross-section of the
vessel should not exceed one per cent of the tanks cross-
sectional area was placed in doubt after the famous Lucy
Ashton experiment. This showed that to avoid bound-
ary interference from the tank walls and bottom this
proportion should be reduced to the order of 0.4 per cent.
The model, constructed from parafn wax, wood or
glass-reinforced plastic, requires to be manufactured
to a high degree of nish and turbulence simulators
placed at the bow of the model in order to stimulate the
transition from a laminar into a turbulent boundary layer
over the hull. The model is positioned under the carriage
and towed in such a way that it is free to heave and pitch,
and ballasted to the required draught and trim.
In general there are two kinds of resistance tests:
the naked hull and the appended resistance test. If
appendages are present local turbulence tripping is
applied in order to prevent the occurrence of uncon-
trolled laminar ow over the appendages. Also the
propeller should be replaced by a streamlined cone to
prevent ow separation in this area.
The resistance extrapolation process follows Froudes
hypothesis and the similarity law is followed. As such
the scaling of the residual, or wave making component,
follows the similarity law
RWship = RWmodel 3 (s /M )
provided that VS = VM , where = LS /LM .
In general, the resistance is scaled according to the
relationship
 
S
Rs = [RM RFM (1 + k)]3 + RFs (1 + k) + RA
M
 
S
= [RM FD ]3 (12.30)
M
in which
2 M
FD = 12 M VM SM (1 + k)(CFM CFS ) RA /3
S
that is,
2
FD = 21 M VM SM [(1 + k)(CFM CFS ) CA ]
(12.31)
The term FD is known as both the scale effect cor-
rection on resistance and the friction correction force.
The term RA in equation (12.30) is the resistance com-
ponent, which is supposed to allow for the following
factors: hull roughness; appendages on the ship but not
present during the model experiment; still air drag of
the ship and any other additional resistance component
acting on the ship but not on the model. As such its
non-dimensional form CA is the incremental resistance
coefcient for shipmodel correlation.
When (1 + k) in equation (12.30) is put to unity, the
extrapolation process is referred to as a two-dimensional
approach since the frictional resistance is then taken as
that given by the appropriate line, Froude at plate data,
ATTC or ITTC 1957, etc.
The effective power (PE ) is derived from the resist-
ance test by the relationship
PE = RS VS (12.32)
Open water tests
The open water test is carried out on either a stock or
actual model of the propeller to derive its open water
characteristics in order to derive the propulsion coef-
cients. The propeller model is tted on a horizontal
driveway shaft and is moved through the water at an
immersion of the shaft axis frequently equal to the
diameter of the propeller (Figure 12.23).
The loading of the propeller is normally carried out by
adjusting the speed of advance and keeping the model
revolutions constant. However, when limitations in the
measuring range, such as a J -value close to zero or
a high carriage speed needed for a high J -value, are
reached the rate of revolutions is also varied. The meas-
ured thrust values are corrected for the resistance of the

Figure 12.22 Ship model test facility
hub and streamlined cap, this correction being deter-
mined experimentally in a test using a hub only without
the propeller.
The measured torque and corrected thrust are
expressed as non-dimensional coefcients KTO and KQO
in the normal way (see Chapter 6); the sufx O being
used in this case to denote the open rather than the
behind condition. The open water efciency and the
advance coefcient are then expressed as
J KTO
0 =
2 KQO
Ship resistance and propulsion 305
and
Vc
J =
nD
where Vc is the carriage speed.
Unless explicitly stated it should not be assumed that
the propeller open water characteristics have been cor-
rected for scale effects. The data from these tests are
normally plotted on a conventional open water diagram
together with a tabulation of the data.
306 Marine propellers and propulsion
Figure 12.23 Propeller open water test using towing tank carriage
Propulsion tests
In the propulsion test the model is prepared in much the
same way as for the resistance test and turbulence stimu-
lation on the hull and appendages is again applied. For
this test, however, the model is tted with the propeller
used in the open water test together with an appropri-
ate drive motor and dynamometer. During the test the
model is free to heave and pitch as in the case of the
resistance test.
In the propulsion test the propeller thrust TM , the pro-
peller torque QM and the longitudinal towing force F
acting on the model are recorded for each tested com-
bination of model speed VM and propeller revolutions
nM .
Propulsion tests are carried out in two parts. The rst
comprises a load variation test at one or sometimes more
than one constant speed whilst the other comprises a
speed variation test at constant apparent advance coef-
cient or at the self-propulsion point of the ship. The
ship self-propulsion point being dened when the tow-
ing force (F) on the carriage is equal to the scale effect
correction on viscous resistance (FD ), equation (12.31).
The required thrust TS and self-propulsion point of
the ship is determined from the model test using the
equation:
and 2 cm from the hull surface and the observation
TM 3 S
TS = TM + (FD F) (12.33)
F M
In equation (12.33) the derivative TM /F is determined
from the load variation tests which form the rst part
of the propulsion test. In a similar way the local vari-
ation test can be interpolated to establish the required
torque and propeller rotational speed at self-propulsion
for the ship.
In the extrapolation of the propulsion test to full scale
the scale effects on resistance (FD ), on the wake eld
and on the propeller characteristics need to be taken
into account. At some very high speeds the effects of
cavitation also need to be taken into account. This can
be done by analysis or through the use of specialized
facilities.
Flow visualization tests
Various methods exist to study the ow around the hull
of a ship. One such method is to apply stripes of an
especially formulated paint to the model surface, the
stripes being applied vertical to the base line. The model
is then towed at Froude identity and the paint will smear
into streaks along the hull surface in the direction of the
ow lines.
In cases where the wall shear stresses are insuf-
cient tufts are used to visualize the ow over the hull.
In general, woollen threads of about 5 cm in length
will be tted onto small needles driven into the hull
surface. The tufts will be at a distance of between 1
made using an underwater television camera. The inter-
action phenomenon between the propeller and ships
hull can also be studied in this way by observing the
behaviour of the tufts with and without the running
propeller.
 Ship resistance and propulsion 307

Model test facilities
Many model test facilities exist around the world almost
all of which possess a ship model towing tank. Some of
the model facilities available are listed in Table 12.7;
this, however, is by no means an exhaustive list of
Table 12.7 Examples of towing tank facilities around the
world (Reproduced with permission from Reference 55)
facilities but is included here to give an idea of the range
of facilities available.
Two-dimensional extrapolation method
This as discussed previously is based on Froudes ori-
ginal method without the use of a form factor. Hence
the full-scale resistanceis determined
 from

RS = (RM FD )3
S

Facilities Length Width Depth Maximum M

(m) (m) (m) carriage where
2
speed (m/s) FD = 12 M VM SM (CFM CFS CA )
European facilities and when Froudes friction data is used CA is set to zero,
Qinetiq Haslar (UK) 164 6.1 2.4 7.5 but this is not the case if the ATTC-1947 or ITTC-1957
270 12.0 5.5 12.0 line is used.
Experimental and 76 3.7 1.7 9.1 When the results of the propulsion test are either inter-
Electronic Lab. 188 2.4 1.3 13.1 polated for the condition when the towing force (F) is
B.H.C. Cowes (UK) 197 4.6 1.7 15.2 equal to FD or when FD is actually applied in the self-
MARIN propulsion test the corresponding model condition is
Wageningen (NL) 100 24.5 2.5 4.5 termed the self-propulsion point of the ship. The direct
216 15.7 1.25 5
220 4.0 4.0 15/30
scaling of the model data at this condition gives the con-
252 10.5 5.5 9 dition generally termed the tank condition. This is as
follows:  
MARIN 240 18.0 8.0 4
Depressurized S
PDS = PDM 3.5

Facility, Ede (NL) M

 

Danish Ship 240 12.0 6.0 14 

TS = TM 3 S

Research
M
Laboratories (12.34)
Ship Research 27 2.5 1.0 2.6 nS = nM /



Institute of 175 10.5 5.5 8.0 VS = VM
Norway (NSFI)  

S

SSPA. Gteborg, 260 10.0 5.0 14.0 RS = (RM FD ) 3
Sweden M
Bassin dEssais de 155 8.0 2.0 5 The power and propeller revolutions determined from
Carnes, Paris 220 13.0 4.0 10 the tank condition as given by equation (12.34) require
VWS West 120 8.0 1.1 4.2 to be converted into trial prediction gures for the ves-
Germany 250 8.0 4.8 20 sel. In the case of the power trial prediction this needs
H.S.V. Hamburg 30 6.0 1.2 0.00231.9 to be based on an allowance factor for the results of
West Germany 80 4.0 0.7 3.6 trials of comparable ships of the same size or alterna-
80 5.0 3.0 3.6 tively on the results of statistical surveys. The power
300 18.0 6.0 8.0 trial allowance factor is normally dened as the ratio of
B.I.Z. Yugoslavia 37.5 3.0 2.5 3 the shaft power measured on trial to the power delivered
23 12.5 6.2 8 to the propeller in the tank condition.
293 5.0 3.5 12
The full-scale propeller revolutions prediction is
North American Facilities based on the relationship between the delivered power
NSRDC Bethesda 845 15.6 6.7 10 and the propeller revolutions derived from the tank con-
USA 905 6.4 3.04.8 30 dition. The power predicted for the trial condition is then
NRC, Marine 137 7.6 3.0 8 used in this relationship to devise the corresponding
Dynamics and Ship propeller revolutions. This propeller speed is corrected
Laboratory, Canada
for the over- or underloading effect and often corre-
Far East Facilities sponds to around 12 per cent decrease of rpm for a 10
Meguro Model 98 3.5 2.25 7 per cent increase of power. The nal stage in the pro-
Basin, Japan 235 12.5 7.25 10
340 6.0 3.0 20
peller revolutions prediction is to account for the scale
effects in the wake and propeller blade friction. For the
Ship Research 20 8.0 01.5 2
trial condition these scale effects are of the order of
Institute, Mitaka 50 8.0 4.5 2.5 1
Japan 140 7.5 03.5 6 2 % for single-screw vessels
375 18.0 8.5 15 12% for twin-screw vessels
KIMM Korea 223 16.0 7.0
The allowance for the service condition on rotational
Hyundai Korea 232 14.0 6.0
speed is of the order one per cent.
308 Marine propellers and propulsion

Three-dimensional extrapolation method According to the ITTC 1987 version of the manual for
the use of the 1978 performance reduction method, the
The three-dimensional extrapolation method is based
relationship between the ship and model Taylor wake
on the form factor concept. Accordingly the resistance
fractions can be dened as
is scaled under the assumption that the viscous resist-
ance of the ship and its model is proportional to the w TS = (t + 0.04) + (wTM t 0.04)
frictional resistance of a at plate of the same length (1 + k)CFS + CF
and wetted surface area when towed at the same speed,
the proportionality factor being (1 + k) as discussed in (1 + k)CFM
Section 12.2. In addition it is assumed that the pres- where 0.04 is included to take account of the rudder
sure resistance due to wave generation, stable separation effect and CF is the roughness allowance given by
and induced drag from non-streamlined or misaligned    
appendages follow the Froude similarity law. ks 1/3
The form factor (1 + k) is determined for each hull CF = 105 0.64 103
LWL
from low-speed resistance or propulsion measurements
when the wave resistance components are negligible. In The measured relationship between the thrust and
the case of the resistance measurement of form factor torque coefcient is corrected for the effects of friction
then this is based on the relationship: over the blades such that
 
R KTS = KTM + KT and KQS = KQM + KQ
(1 + k) = lim
Fn 0 RF where the factors KT and KQ are determined from
In the case of the propulsion test acting as a basis for the ITTC procedure as discussed in Chapter 6.
the (1 + k) determination then this relationship takes The load of the full-scale propeller is obtained from
the form the relationship

KT S CTS
F T /(T /F) =
(1 + k) = lim J 2 2D (1 t)(1 wTS )2
2
Fn 0 (F|T =0/R)RF
The low-speed measurement of the (1 + k) factor can and with KT /J 2 as the input value the full-scale advance
only be validly accomplished if scale-independent pres- coefcient JTS and torque coefcient KQTS are read
sure resistance is absent, which means, for example, off from the full-scale propeller characteristics and the
that there is no immersed transom. In this way the form following parameters calculated:
factor is maintained independent of speed and scale in (1 wTS )VS
the extrapolation method. nS =
In the three-dimensional method the scale effect on JTS D
the resistance is taken as KQTS
PDS = 2D5 n3S 103
2
FD = 21 M VM SM [(1 + k)(CFM CFS ) CA ] R
(12.35)
KT 2
in which the form factor is normally taken relative TS = 2 JTS D4 n2S
to the ITTC-1957 line and CA is the shipmodel cor- J
relation coefcient. The value of CA is generally based KQTS 5 2
QS = D nS
on an empirically based relationship and additional R
allowances are applied to this factor to account for
extreme hull forms at partial draughts, appendages The required shaft power PS is found from the deliv-
not present on the model, contract conditions, hull ered power PDS using the shafting mechanical efciency
roughness different from the standard of 150 m, S as
extreme superstructures or specic experience with
previous ships. PS = PDS /S
In the three-dimensional procedure the measured
relationship between the thrust coefcient KT and the
apparent advance coefcient is corrected for wake scale
effects and for the scale effects on propeller blade fric-
tion. At model scale the model thrust coefcient is
dened as
KTM = f (Fn , J )M
whereas at ship scale this is
   
1 w TS
KTS = f Fn , J + KT
1 w TM