Shipboard acoustics proceedings of the 2nd international symposium on shipboard acoustics ISSA ’86

The first International Symposium on Shipboard Acoustics, held in Noordwijkerhout (The Netherlands) in 1976, was a meeting of invited experts, each having considerable expertise in ship acoustics. Many of the participants were dealing with research on various ship acoustical subjects, and it proved to be a good idea to discuss future investigations and new techniques. At that time acousticians learned to use realtime signalprocessing techniques and attempts were made to establish sound level prediction methods based on semifundamental considerations instead of the methods using empirically obtained data. Time was pressing as it was assumed that, in view of the adoption of Recommendation 141 of the International Labour Conference in 1970, authorities would soon make appropriate provisions to protect seafarers from the ill effects of noise. This resulted in several national recommendations followed by the IMO Code on noise levels aboard ships which was adopted by the IMO Assembly in 1981. After that, pressure on the authorities was increased further by the decision of the European Community to protect labourers against harmful noise at their workplaces, including ships. Legally enforceable noise limits will therefore become normal in the future.

Proceedings of the 2nd International Symposium on Shipboard Acoustics ISSA '86, The Hague, The Netherlands, October 7-9, 1986

Organized by the

Netherlands Organization for Applied Scientific Research (TNO)

TNO Institute of Applied Physics (TPD)

in co-operation with

Maritime Research Institute Netherlands (MARIN)

National Foundation for the co-ordination of Maritime Research

in the Netherlands

Royal Netherlands Navy

edited by

J BUITEN

TNO Institute of Applied Physics (TPD)

Delft, The Netherlands

a member of the KLUWER ACADEMIC PUBLISHERS GROUP lIiII

Drive, Assinippi Park, Norwell, MA 02061, USA

for the UK and Ireland: Kluwer Academic Publishers, MTP Press Limited,

Falcon House, Queen Square, Lancaster LAI lRN, UK

for all other countries: Kluwer Academic Publishers Group, Distribution Center,

P.O Box 322, 3300 AH Dordrecht, The Netherlands

ISBN -13:978-94-010-8070-5 e- ISBN -13 :978-94-009-3515-0 DOl: 10.1007/978-94-009-3515-0

Copyright

© 1986 by Martinus Nijhoff Publishers, Dordrecht

All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publishers,

Martinus Nijhoff Publishers, P.O Box 163, 3300 AD Dordrecht,

The Netherlands

Noordwijkerhout (The Netherlands) in 1976, was a meeting of invited experts, each having considerable expertise in ship acoustics Many of the

participants were dealing with research on various ship acoustical subjects, and it proved to be a good idea to discuss future investigations and new techniques At that time acousticians learned to use real-time

signal-processing techniques and attempts were made to establish sound level prediction methods based on semi-fundamental considerations instead

of the methods using empirically obtained data

Time was pressing as it was assumed that, in view of the adoption of

Recommendation 141 of the International Labour Conference in 1970,

authorities would soon make appropriate provisions to "protect seafarers from the ill effects of noise" This resulted in several national

recommendations followed by the IMO "Code on noise levels aboard ships" which was adopted by the IMO Assembly in 1981 After that, pressure on the authorities was increased further by the decision of the European

Community to protect labourers against harmful noise at their workplaces, including ships Legally enforceable noise limits will therefore become normal in the future

In many countries recommendations with respect to maximum allowable sound pressure levels in the crew accomodations and work area aboard ships were already taken into account by ship owners, long before the existence of the Recommendations

Shipyards are confronted with the problem of how to fulfil the requirements

at the lowest possible costs, a situation which does not essentially differ from that in 1976 In addition however, modern ships tend to be of a lighter construction and as a consequence, the acoustic countermeasures have to be light Yards are constructing ships that are more specialized and

technically more complicated than in the past; acoustical standard solutions are of little use for these vessels The yards are compelled to find

solutions that have' flexibility in the design stage and standardization in the production stage

The consultant is confronted with rapid and drastically changing demands of owners and yards, which require fast and highly precise answers He should have perfect knowledge of the results of research and has to transform these into practical tools Much more often than before, he has to combine

disciplines, which makes it difficult to be an expert in one of them

In the past ten years research workers have experienced a rapidly growing demand on research, followed by a sharp decrease of funds Their tools changed dramatically compared with preceding decennia, in capabilities as well as in the necessary investments, especially in computer hardward and

in software They discovered that solutions of rapidly increasing quality could be given, but that costs increased correspondingly Many developments

eXisting, or to establish new data bases

Increasing demands on research cause increasing specialisation and, for the research worker an increasing need for open discussions with colleagues Society however, requires methods that are simple to handle, and industry wants to protect its trade secrets

In spite of the conflicting demands of the different parties involved in design, construction and management of ships, and those in charge of the crew and passengers, much has been attained in the last decennium At a symposium on shipboard acoustics and vibrations only some recent

highlights and experiences can be emphasized in a limited number of papers The discussions, which will be added to this issue later, will certainly complete the papers presented We trust that the contributions to this book will give rise to an exchange of ideas on the application of research results, on subjects for future research and on the necessary

The editor would like to thank his colleagues of the TNO Institute of Applied Physics (TPD) and Dr A de Bruijn of SACLANT ASW Research Centre for their assistance, and the authors who made this task a pleasant one August 1986

J Buiten

PREFACE

1 A de Bruijn, W.H Moelker, and F G.J Absil

Prediction Method for the Acoustic Source Strength of Propeller

Cavitation

v

2 A Colombo, P Ausonio, L Grossi, and L Accardo 21 Propeller Induced Noise and Vibration Reduction: Acquired

Experience in Design and Testing Approach

Experimental and Analytical Aspects of Propeller Induced Pressure Fluctuations

Model and Full Scale Measurements of Propeller Cavitation Noise

on an Oceanographic Research Ship with Two Different Types of

Screw Propeller

A Study of Simulation and Elimination of Propeller Sinqing

6 B Bajic, J Tasic, A Dzubur, and I Jovanovic 91 Propeller Noise: Some Topics from the Activities of Brodarski

Institute

Some Experiments on the Transmission of Propeller Cavitation

Noise into the Ship's Structure

Quiet High-Speed Yachts and Water Jet Applications

Model-Scale Measurements of the Transmission and Radiation of

Hull-Borne Vibrational Energy Using Frequency/Wavenumber Analysis

Underwater Noise Radiation Due to Transmission through the

Cooling Water System of a Marine Diesel Engine

11 A.R Clark and P.S Watkinson

Measurements of Underwater Acoustic Intensity in the Nearfield of

a Point Excited Periodically Ribbed Cylinder

177

12 Zhu Xiqing 189 Sound Generation from a Moving Shell

Low Noise Levels as the First Task of a Vessel A Description

and Some Remarks about Acoustic Quieting Design Criteria and

Features of Saclant ASW Centre Oceanographic Research Vessel

Ship Noise Criteria Do They Reflect the Present Level of Noise

Reduction Technology?

A Literature Survey Concerning Propeller as a Noise Source and

Prediction Methods of Structure-Borne Noise in Ships

Applications o"f Two Mathematical Approaches to Predict Airborne

Noise Levels in Ship Superstructures

Some Observations on the Achievable Properties of Diesel

Isolation Systems

18 J.G van Bakel

Acoustic Transfer Functions of Flexible Shaft Couplings;

Measurement Method and Results

279

Random Vibration of Multiterminal Mechanical Systems

The Influence of the Internal Impedance on Vibration Reduction

Main Propulsion Diesel Generator Sets with Acoustic Enclosure

and Double Resilient Mounting for Low Noise Application

24 A Blanchet, G Chatel, and A Paradis

Study of Structure-Borne Noise Transmission Inside Cabins by

Sound-Intensity Measurements

377

25 M.J.A.M de Regt

Experiments on Sound Reducing Floors Including

Visco-Elastic-Damping on Board a Rhine Cruise Vessel

393

TNO INSTITUTE OF APPLIED PHYSICS (TPD) , DELFT, THE NETHERLANDS

1 INTRODUCTION

Ship propeller cavitation is considered to be one of the most tant sources of underwater noi se Furthermore it often contri butes con- siderably to the noise level aboard the ship Much research has been carried out in this field in order to reduce the extent of cavitation by a proper blade design Also through the use of model simulation techniques more insight into the cavitation performance has been gained An important overview about this subject has been presented by ISAY [1]

impor-Modern computer-based design techniques,(cf KERWIN [2], VAN GENT [3]) applying three-dimensional lifting surface theories, have improved con- siderably the insight into the hydromechanical aspects For example, the pressure distribution on the blades gives the position where cavitation probably will occur

The price to be paid to reduce the cavitation is mostly the decrease of propulsion efficiency This is the reason that in the design phase the con- sequence of noise reduction has to be known at a very early stage Mostly nothing more is known but the propulsion power, rotation speed, diameter

of the propeller etc., so the noi se and vi br,ation anal i st is 1 eft with empi ri ca 1 methods to predi ct the acoustic source strength or the hull- induced vibrations

Prediction of the hull-pressures induced by the cavitation has been the subject of considerable investigations NOORDZIJ [4J has attempted to deve- lop a calculation scheme for estimating the cavitation volumes at any blade position By calculating the volume variations he was able to determine the pressure amplitudes exerted on the hull The agreement with experiments is quite limited but useful in comparative studies

Up ti 11 now a cheaper, but st i 11 re 1 i ab 1 e, method is the semi -empi ri ca 1 approach The method is based on the compari son of a great number of experimental data taken from as many ships as possible By statistical regression it has been attempted to find the significant parameters which describe the induced hull-pressures A successful evaluation of this approach for hull-induced pressures at the two lowest blade-rate frequen- cies is given by HOLDEN et al [10]

* Currently employed at: SACLANT ASW Research Centre, La Spezia, Italy

Buiten, J (ed), Shipboard Acoustics ISBN 90-247-3402-7

© 1986.' Martinus Nijho!! Publishers, Dordrecht

a 1 arge number of hi gher harmon i cs [5] These are, however, part of the random spectrum The cavitation volume variations can be hardly called

"harmonic" The sheet is randomly vibrating, giving rise to random noise components Contrary to the general opi nion these random components are very significant They are extremely difficult to predict since no suitable mathematical model exists to calculate these random vibrations

This paper presents a method which is very much related to HOLDEN's method, but with the difference that it attempts to predict the source strength of propeller cavitation noise above the lowest blade-rate frequencies, viz between 30 and 500 Hz It makes use of many experimental data on propeller noise measured aboard a variety of Dutch ships By means of a statistical regression analysis a semi-empirical prediction formula has been found

2 INVENTARISATION OF PROBLEMS

As stated many times the noise aboard ships is a difficult problem for the des i gner as well as for the acoust i ci an The propeller des i gner is always interested in reducing the cavitation, although he knows that the price to be paid can be quite high: reduced efficiency, larger diameters, adaption of the wake He ends up with a compromise; acceptable cavitation performance (mostly in view of erosion) with acceptable hydromechanical properties Required acceptable noise levels in the accommodation are well established and give guidelines how far cavitation should be avoided Still he is left with an unresolved question: what is acceptable for cavi- tation in view of noise and vibrations or what is the relation between ca-vitation picture or performance and the perceived noise level in the accommodation From the point of erosion the solution is, however, not so difficult: avoid bubble cavitation, cloud cavitation or pressure-side cavi- tation With the present st~te of the design techniques this is feasible Mostly it is quite difficult to avoid both sheet-cavitation and tip-vortex cavitation These types are mainly responsible for the noise generated aboard the ship or underwater

The main question is now what is the relation between the cavitation ture and the noise generated An a-priori answer cannot be given, since there is no suitable mathematical sol ution on the relation between the noise and the picture

pic-Model experiments can give reliable solutions Cavitation has been gated traditionally in cavitation tunnels in which the static pressure in the fluid and the flow speed can be varied To improve the simulation of full-scale conditions, and thus the prediction of cavitation and its detri- mental effects (like propeller erosion, pressure fluctuations, noise) several special facilities have been built during the last two decades Among these the NSMB Depressurized Towing Tank occupies a special place This large tank is in operation since 1972 The original purpose was to perform propulsion tests with models of the (at the time) even increasing size of tankers, bulk carriers and container ships at a reasonable scale, say 1:30 By adding the possibility of controlling the air pressure in the tank, a unique facility was obtained for testing cavitation and its asso- ciated phenomena In the depressurized towing tank it is feasible to measure the radiated noise in an appropriate way, since the reflections to

investi-the walls can be kept quite reduced in contrast with investi-the cavitation tunnel The acoustic results taken from experiments in this tank converted to the full-scale have proven to be quite reliable, cf VAN DER KOOIJ & DE BRUIJN [8]

Sti 11 there is another factor that comp 1 i cates the matter cons i derab ly The noi se generated by the propeller is transmitted through the shi p' s structure This means that the noise levels do not yield unambiguous infor- mation on the source strength of the propeller cavitation proper

Comparison of levels either in different spaces or in different ships is meaningless without a correction for the transmission loss Transfer of sound through the steel structure is a most comp 1 i cated affai r, not very much understood or predictable, although the progress in this field is quite impressive It is of prime importance to eliminate the transfer func- tion before judging the source strength of various propellers

This paper concentrates on the determination and evaluation of the tion source strength This is a first step in the prediction scheme for the noise aboard the ship Transmission of the propeller noise should be dealt with separately We refer to another paper in this proceedings [9]

cavita-The source strength can be obtained by model experiments, as explained above This is a lengthy procedure and in fact not suitable in early design stage Another successful quick method is based on the evaluation of empirical data obtained from ships of the same kind HOLDEN [10] has been rather successful in predicting the hull-induced pressures His approach was in fact simple: measure the hull-pressures on a variety of ships and apply a statistical regression analysis so that the induced pressures can

be related to a number of hydrodynamical parameters: the wake field, geometry of the propeller, loading of the propeller etc By calculating or estimating this kind of parameters for all his ships and applying a sta- tistical regression method he was able to formulate an empirical rule for the amplitude of the hull-pressures at the blade-rate frequencies This method has been proven to be extremely fruitful in the selection of a pro- peller geometry in the design stage Additions or alternatives have been proposed (such as the influence of skew-back), but essentially this approach has been maintained We refer also to the work of HOLTROP [11] and LEENAARS & FORBES [12] on the prediction of hUll-vibrations

As stated above this method applies to the amplitudes of the blade-rate frequencies For the higher harmonics it seems not too accurate any more This means that the audible noise cannot be predicted by this method Also the concept of hull-pressures is not relevant to transfer of sound in the accommodat ion For thi s purpose we have developed the idea of the equi va- lent monopole This is based on the observation that the noise generated in the water is mainly C:ue to vol ume variations With respect to sound generation we could say that a monopole representation can be assumed On vessels with a sharp wake non-uniformity this dominant volume variation component of the pressure may be approximated by an oscillating monopole source fixed in a position below the ship's hull, radiating the free field pressure

where: U = volume velocity, f = frequency

r = distance from monopole to observation point

various prediction methods The best known are given by ROSS [6J and BROWN [7J Both methods are very much inaccurate at frequencies below 100 Hz as has been pointed out by measurements by THIELE/0DEGAARD [14J

3 DETERMINATION OF VOLUME VELOCITY

The source strength of propeller cavitation can be determined in either two ways Directly in the water or by means of the method based on the principle of reciprocity applied onboard the ship The first method is becoming more popular We refer to the work of SASAJIMA et al [13] and THIELEl0DEGAARD [14 J By measuri ng the noi se radi ated underwater at dif- ferent distances from the ship, the source level being the pressure at a distance of 1 m from a fictitious monopole source placed at the cavitation centre was obtained This source level is found by adding the estimated transmission loss to the measured sound pressure levels The transmission loss depends on the distance between the measuring hydrophone and the ship Mostly the underwater noise level was measured with the aid of an array of hydrophones, whi ch was suspended into the water from a pil ot boat The underwater noise recordings were performed continuously from the pilot, while the ship approached, passed and left the recording position A dif- ficulty in these measurements is the correct determination of the distance One investigator used a pistol sound as a distance indicator THIELE/ 0DEGAARD measured the transmission loss by means of small explosive charges and two hydrophones, one close to the explosion ("source hydrophone") and one placed at different distances from it ("receiver hydrophone") Other investigators assumed a simple square distance law in the transmission loss, neglecting the reflections from the sea bottom and surface These methods are very expensive and time-consuming Morever, these experiments are to be performed in deep water Thi s makes thi s method unwi e 1 dy for routine measurements

For the determination of the source level onboard another method can be adopted Place a large sound source in the propeller region and measure the response in the accommodation Since the acoustic source strength is known for such a transducer, the levels due to the propeller cavitation can be

di rect ly compared with those obtained from the auxil i ary source Sti 11 this method is only feasible when a transducer can be mounted in the pro- peller region A strong electro-dynamical underwater sound source is heavy and difficult to handle in the propeller region by divers To avoid these kinds of experiments we have proposed another variant of this method At low frequencies, say below 1 kHz, the sound source is much smaller than an acoustic wavelengt~ As has been observed during viewing trials the cavita- Hon is mostly concentrated on the blade in the upper-most position This means that the dimension and duration is quite limited so it is safe to consider the cavitation as a point source The measuring technique

is based in the reci procity pri nci p 1 e Each measurement cons i sts of two different experiments (Fig 1)

~. -.~, -,~~

"';];Xlectro-~ acoustical

U transducer

voZ·ume velocity of propeUer cavitation

In the first ("silent") experiment the transfer function is measured and it

is determined by a reciprocal technique for practical reas'ons In the second ("sailing") one the noise or vibrations due to the propeller cavita- tion at an arbitrary point in the accommodation are measured These results represent the combining effect of the source strength of the operating propeller and the transfer function describing the relation bet- ween the source strength and the vi brat ions in the selected poi nt in the accommodation In practice the procedure is as follows A reversible linear ,mechanical- or electrical-acoustical transducer is installed in the accommodation In principle the transducer can be placed anywhere in

the accommodat i on above the propeller, but inmost cases a 1 ocat ion close

to the propeller is chosen in order to obtain the highest signal-to-noise ratio In the first experiment the ship propeller is idle thus not rotat i ng A number of hydrophones is mounted onto the the blade in the upper pos it ion The transducer is used now as an exc iter, dri ven by an electric current i, while the average resulting pressure p is measured by means of the hydrophones In the second experiment the propeller is in operation and the output open voltage e of the transducer is measured

According to the reciprocity principle the volume velocity of the

cavita-tion U is given by:

U = e • i Ip (m3/s)

This equation holds for pure tones but can be easily extended to 1/3-octave

frequency bands, although there are some pitfalls The noise from the peller cavitation (represented by the measured quantity e) is mostly broadband, but the ratio ilp is mostly made up by the generation of sweep tones to obtain the best signal-to-noise ratio The result of the measure- ment depends heavily on the location of the hydrophone in the first experi- ment From extensive investigations it has been found that the best location is the centre of the sheet cavitation when the sheet has started its collapse stage To obtain a reasonably unique transfer function from the cavitation region to a point inside the ship or model it is absolutely necessary to take more measuring positions in order to average these transfer functions with respect to position It has become standard prac-

pro-rotation We obtain then 9 transfer functions, which can be easily averaged It has turned out that the pressure sometimes shows very sharp dips, so it is recommended to average p instead of i/p Frequency- averaging is also essential to obtain reasonably unique transfer functions

We prefer to take 1/3-octave frequency bands, narrow enough to observe the

i nfl uence of the bl ade-rate harmoni cs separately and wi de enough to get reasonably stable averages

4 SHIPS

Seven single-screw ships have been investigated The size of these ships and their trial speeds covered a fairly wide range The ships were investigated mostly in the harbour, especially with respect to the "silent" experiments The "sail i ng " experiments were, of course, performed at sea, preferably in deep water

a.o Weight Diameter(Dp) Blades(Z RPM(N) Speed(Vs )

TABLE 1: Characteristics of ships which have been investigated

Not in all cases the data of the ships could be obtained, especially cerning the propulsion Mostly the size of the ship is known and something about the geometry but little about the propulsion configuration The res i stance and the propul s i on power were cal cul ated accordi ng the method of HOLTROP [15J From these data the optimum blade area and pitch can be calculated It has assumed that for normal propellers a Wageningen B-design could be adopted and the MCR-power could be calculated These parameters were necessary to obtai n the relevant parameters, on whi ch the source strength depends

con-5 HYDRODYNAMICAL PARAMETERS

It is expected that a 1 arge number of hydrodynami ca 1 parameters wi 11 playa role in the generation of cavitation noise With the aid of a sta- tistical regression the relevance of these parameters has been analysed Geometrical variables, whose influence has been analysed are: prismatic coefficient, number of propeller blades Z, thrust coefficient KT, torque coefficient KQ, thrust coefficient KT/J2 (where J = ratio of advance VA/nDp, VA = effective flow velo~ity~, torque coefficient (KQ/J5)0.25, Froude number, Reynolds number, cavltatlon number 0, etc ThlS klnd

of parameters has been studied and from this statistical analysis it can

be concluded that only two parameters are important [19]:

The logarithm of the dimensionless power:

K'Q = 19 (KQ/J5)0,25

0,25 19 (2 'IT Ps n2/ P VA5)

VA = (1 - w).Vs (effective flow velocity)

Vs = ship speed

w = effective wake field value

and a parameter about the blade geometry: cf HOLDEN [10]

F = (AE/AO)N

(AE/AO)A where: (AE/AO)N reference blade area ratio according to HOLDEN [10J

1,9.(0,235.0+ 0,063) * (1,067 - 0,23.(P/Dp)0,7) (AE/AO)A = actual blade area ratio

(P/Dp)0,7 = pitch/diameter at 0,7 of the radius

o = cavitation number (Po - Pv) / {! P (V 2A + (0,7 'IT .n.Dp)2}

The factor F i ndi cates how far a cav i tat ion criteri urn has exceeded The ref erence blade area appears to be a funct i on of the thrust coeffi c i ent, cavitation number, the effective wakefield and the pitch The larger the value of F the stronger the propeller cavitates The most significant parameter K'a gives an indication about the quality of the propulsion con- figuration It is a measure for the propeller load The larger the value the more the propeller is loaded In the various speed conditions of the

sh i p K Q does not vary very much The quantity (KQ/ J5) 0.25 is related to

understandable that only these two parameters describe the noise generated

We would expect that the wakefield and the Reynolds number to be more significant parameters

6 VOLUME VELOCITIES

The values of the volume variations determined forthe various ships are frequency-dependent, but an average value with respect to a wide frequency band could be a meaningful quantity to compare various designs [18J It is evident that the smaller the propeller, the smaller the cavitation volumes

to be generated become Therefore it is necessary to make it dimensionless

by dividing the volume velocities by the factor n.Dp3 (n = rotation speed (s-I); Dp = propeller diameter This removes in the volume velocities, in any case factors wh i ch are related to the size of the prope 11 er and the rotation speed

pro-scales to one reference I/3-octave band in which all the blade-rate qiencies coincide.The transformation is such that the I/3-octave band with

fre-the lowest blade-rate frequency is assigned fre-the number 1

L'UBF(j) L'u(i)

where

j i - BF' +1 BF'= Ent(lO 19 fo + 0,5)

L'UBF = frequency shifted volume velocity level

j = band number of shifted 1/3-octave-band

i = band number of actual 1/3-octave-band BF'= band number of 1/3 octave band of blade-rate

fo = blade-rate frequency (= NZ/60)

It is an interesting question whether the volume velocity or the radiated acoustical power is the fundamental physical quantity which describes the acoustical process Acoustical power can be related to the mechanical power In thi s way a proper compari son between shi ps wi th different pro- pulsion powers can be justified Since the acoustical power contains a fac- tor f2 when related to tne volume velocity we have to include a frequency correction, since Wac = TI .p.U2.f2/c Because we have made a transfor-

mation with respect to the lowest blade frequency it is essential to include now a correction on the levels by the frequency shift

L"UBF(j) L'UBF(j) + 20 19 fo

where

L"UBF(j) = frequency-corrected volume velocity level

So the val ues of the vol ume velocities are transformed by three tions cf Fig 2:

opera-* make the volume velocities dimensionless;

* a frequency transformation to e 1 imi nate the infl uence of the blade rate frequency;

* corrections of the levels in view of similarity of radiated power

is very much ship dependent Therefore we have to include a weighting tor which corrects for the fact that sometimes more than one measurement has been obtained aboard one specific ship The weighting factor has been calculated so that every ship counts equally independent of the number of the observations of a particular ship

total number of observations in the

frequency band j

total number of ships in band j

total number of observations on ship k and frequency band j

For the determination of F the weighting factor is always one Every vation is equally dependent on this factor For the factor K'a is this dif- ferent since this factor is very much dependent of the type of ship

obser-8 PREDICTION FORMULA

The volume velocity in third-octave frequency bands of a fictitious equivalent monopole in the position of the upper part of the blade in the upper position can be approximated by the following equation:

The table presents results for frequency bands larger than the second monic of the blade rate frequency (j=4) It was possible to go lower, but

har-we did not have enough data to support the statistical results

9 COMPARISON OF MEASURED AND PREDICTED SOURCE STRENGTH

Taking the above mentioned formula it is possible to calculate again the volume velocities for a ship which has been part of the collection We can compare in this way whether the statistical regresssion analysis leads

to acceptable "prediction" (Figs 3) The agreement of the calculated data with the measured ones is satisfactory for most ships Of course we must realize that this kind of calculation has limited validity and proves nothing about the regression analysis itself The ro-ro container ship (No.7) is a typical example where the peak of 31,5 Hz is very outspoken

It is quite unusual for the fourth harmonic to be so strong In this way the model appears not to be so adequate Also the results for the oceanographic vessel (No.6) are disappointing The propeller was fairly large, 6-bladed with relatively thick blades Only a tip-vortex was visible, no sheet cavitation The noise levels are much lower than the calculated ones, especially for the low frequencies

An interesting question remains the sensitivity of the predictions Four variables are important: 20 19 (n.Dp'3), 20 19 (n.Z), K'Q, and F These variables show some spread, expressea in the standard deviation If this value for each variable is multiplied by the coefficient averaged for

a 11 frequency bands we obtai n the average devi at i on of the cal cu 1 ated volume velocities in dB as the results of the standard deviations of the various variables

variable SD coefficient SD x coefficient

TABLE 3: Mean deviation of the calculated volume velocities in dB

as the results of the standard deviation of the variables of the

prediction method

The interpretation of the data gives some information on the importance

of each vari ab 1 e The size of the propeller appears important Thi sis quite understandable, since it is the major factor for scaling The factor K'Q appears also to be significant This is less understandable Especially at the high frequencies this variable predicts the values of the volume velocities apparently very well There are no specific arguments to support this fact Maybe the physical background of the high-frequency noise generation has not been explained sufficiently yet The factor F is only important for the low frequencies This is most probably due to the fact that a high value of F agrees with an outspoken sheet cavitation with strong low-frequency deterministic components

field w Assume that the quantity takes values between 0,4 and 0,9 (mean value 0,6) As given above this factor is included in the effective velo- city VA and this quantity is also included in both the factors K'Q and F Calculations learn that the variation in volume velocity level is about 3

to 4 dB around the mean value in all frequency bands In view of the tainty of the value of the wakefield we may conclude that it still leads to

uncer-an acceptable spread in the volume velocities

The best'way to check the validity of the method is to measure the volume velocities aboard a number of other ships outside the collection and to compare the calculated data with the experimental ones We have chosen three dredging ships, which were measured after the period in which the present method was developed by J.A Wind

The following table gives the characteristics of the three ships

TABLE 4: Characteristics of three dredging ships

All three ships have a twin-propeller propulsion configuration The volume velocities have not been determined by the above-described reciprocal method, but with the aid of the hull-plate response This method is experi- mentally simpler and has recently been developed by TPD The results are in good agreement with those obtained by the reciprocal method as has been checked with a number of ships

The results for the dredging ships are presented in the Figs 4 The ship with the rather conventional propeller design gives the most satisfactory results The agreement with the "predicted" values is encouraging The next ship had a propeller design with some skeW-back The results are already somewhat disappointing in the low frequency region The last dredging ship gave experimental results much lower than the predicted ones It is evident that the skew-back propeller has a geometry which falls outside the collec- tion of propellers in the semi-empirical method The difference can be apparently more than 10 dB in a broad frequency-range

This was also observed on two sister ships: PASADENA and PATAGONIA [17J The first ship had a conventional 4-bladed propeller, while the other one had an advanced skew-back des i gn For both sh ips the pressure vari at ions

on the hull-plating up to the 6th harmonic have been measured Also in this case a remarkable difference in the pressures was been observed

10 DISCUSSION AND CONCLUSIONS

Although the number of ships in the statistical collection was quite limited we believe it was sufficient to develop a reasonably reliable pre- diction method A strong point in the analysis was that a wide range of ship types was taken into account This makes the method applicable to dif- ferent types of vessels The disadvantage is that it makes the prediction less reliable

The interpretation of the data is still difficult The best manner to start

a regression analysis is to take clear physical quantities which describes the process adequately This was not possible in the present investigation The noise generation by cavitation is such a complicated process with so many different, sometimes dependent, variables, that an a-priori

selection of important parameters is useless By a careful appl ication of the statistical regression method with many meaningful parameters we could select 2 significant quantities It is surprising that only a limited number of parameters appear to be significant Fortunately we could reduce the significant parameters to dimensionless numbers, which seem to have some physical significance One quantity is related to the propulsion con- figuration (K'Q), while the other one is related to the loading of the pro- peller

Skew-back of the propeller blade was not included in the present model and the first results on this point were quite disappointing Maybe it is possible in the future to correct it by reducing the factor Cj3 by about

10 dB

An attractive point of the present method is that it can also be used in model experiments All quantities involved can be made dimensionless, so the experimental results of this kind of tests can be used to update the present prediction technique

The following conclusions can be put forward:

pro-500 Hz Below that lower frequency and especially for the amplitudes

of the blade-rate frequencies we refer to HOLDEN's method

The present method holds for conventional propeller designs Skew-back designs need more attention and it is most necessary to collect more information about the cavitation noise performance of this kind of propellers

More measurements about the source strength of propeller cav i tat i on both on full-scale and model scale are recommended to refine the pre- sent semi-empirical approach

This study is the result of extensive programme aboard sea-going ship concerning propeller noise, which started already in 1973 Much of this research has been funded by the TNO Nati ona 1 Defence Research Council and by the Netherlands Maritime Institute/Netherlands Foundation on the Coordination for Maritime Research Most of the statistical work has been carried out by J.A Wind as a graduate research study for the Department of Marine Technology of the Delft University of Technology in cooperation with the Institute of Applied Physics TNO, Ship Acoustics Section

Pressure field induced by a cavitating propeller,

Int Shipbuilding Progress ~ (1976), pp 93 - 105

N.P.Tyvand,

Theoretical model for propeller

Paper Gin No i se Sources

Milj0vardserien 1981:2

noise prediction,

in Ships; I: Propellers, Nordforsk,

[6J D Ross,

Mechanics of Underwater Noise,

Pergamon Press, New York etc.,1976, p.276

[7] N Brown,

Cavitation noise problems and solutions,

Proceedings International Symposium in Shipboard Acoustics, 1976, Ed J.H Janssen, Elseviers Scientific Publishing Company, Amsterdam,

1977, pp 21 - 38

[8J J van der Kooij, A de Bruijn,

Acoustic measurements in the NSMB depressurized towing tank,

Int Shipbuilding Progress II (1984), pp 13 - 25

[9J J.H Janssen, W.H Moelker,

Some experiments on the transmission of propeller cavitation noise into ship's structure,

Present proceedings

[10J K.O Holden , O Fagerjord, R Frostad,

Early design-stage approach to reducing hull surface forces due to propeller cavitation,

Transactions of the Society of Naval Architects and Marine Engineers (SNAME), Vol 88 (1980), pp 403 - 442

[12] C.E.J Leenaars, P.E Forbes,

An approach to vibrations problems at the design stage,

RINA Symposium on Propeller Induced Ship Vibration, December 1979, London, Paper no 17

[13] T Sasajima, M Nakamura, A Oshima,

Model and full-sca)e measurements on propeller cavitation of an oceanographic research ship with two different propeller designs, Present proceedings

A statistical power prediction method,

Int Shipbuilding Progress 25 (1978), pp 253 - 256

[16] N.R Draper, H Smith,

Applied regression analysis,

John Wiley & Sons, New York, 1981, 2e edition

[17J G Bar.k, C.-A Johnsson,

Prediction of cavitation noise from model experiments in a large tation tunnel,

cavi-Paper F in Noise Sources in Ships;I: Propellers, Nordforsk, Miljovardserien, 1981:2

[18J A de Bruijn,

Acoustic source strength of propeller cavitation,

Proceedings INTER-NOISE 79, Warzawa, Paper 115 - C, pp 659-664, 1979 [19] J.A Wind

Investigation into the empirical relation between the acoustical source strength of cavitating ship propellers and ship hydrodynamical parameters (in Dutch)

Delft University of Technology, Department of Marine Technology, Laboratory of Ship Constructions, Report No 265, November 1983

_ Transformed f requency band nu.bfr

dimensionZess, frequen y- shi ted and orreeted

-b-in Hz

15

Fig 3a : Comparison of measured and calculated spectra

f or the volume velocities of a number of ships, which were part of the collecti on

Fig , 3b : Comparison of measured and calculated spectra

for the volume velocities of a number of ships, which were part of the collection ,

Ik

Hz

velocity for three dredging ships, which were

not part of the collection

PROPELLER INDUCED NOISE AND VIBRATION REDUCTION

IN DESIGN AND TESTING APPROACH

A Colombo, P Ausonio - FINCANTIERI CNI - Naval Ship Design Dept., Genoa (Italy)

L Grossi - CETENA - Italian Ship Research Centre,-Genoa (Italy)

L Accardo (CDR, IN) MARICONAVARMI, MDM - Rome

ABSTRACT

shipyard (CNI) and CETENA for propeller design are briefly summarized, as well as the methodologies adopted for forecasting and reducing undesired phenomena such as the inception and extension of cavitation, induced pressures and radiated noise

Some of the results achieved and experience acquired for both merchant and naval ships are presented and discussed

During the last two decades, designers and owners have devoted increasing efforts to studying the problems related to on board vibration and noise, for both naval and merchant ship applications

Already important from a general point of view, these problems are most

mer-chant vessels (Ro-Ro, container, multipurpose, etc.), due to their cial hull forms and structures, passenger vessels due to their high comfort

both on board noise and vibration levels, and underwater noise radiation For ships of this kind, noise and vibrations may represent a severe constraint to their operability

In effect, Ro-Ro and container ships are generally characterized by quite high design speeds and consequently high installed power, in most cases developed on a single-screw arrangement Due to space availability requirements, moreover, the stern counter is frequently wide, with reduced propeller-hull clearances to allow larger propulsor diameters for greater efficiencY These elements, combined with open main ship structures and short and high superstructures often lead to a high excitation level as well as to vibration resonance risks in the low frequency range, which may cause severe annoyance to the crew and/or structural damages

For passenger vessels, the problem arises from the need to ensure comrort

in all the living spaces of the ship, with very lmu vibration and noise levels on board, even at high speed and wi th very powerful propulsion engines For naval vessels, the problem is even more complex because of the higher propulsion performance requirements and greater installed power on board

Buiten, J (ed), Shipboard Acoustics ISBN 90-247-3402-7

© 1986 Martinus NijhoJJ Publishers, Dordrecht

a ship having a compulsorily lighter structure So in addition to the risk of structural damages, vibrations and noise may lead to dangerous drops in the operational efficiency of the crew, while operational problems may also arise for the sophisticated equipment and electronic devices installed

on board Moreover, from the point of view of underwater radiated noise, two aspects are of paramount importance, i.e detectability of the ship, and self-induced noise and interference problems in the active or passive sensors with which a naval ship is always equipped

Of course the problem of controlling and reducing the vibrations and noise

on board a ship involves the general design of the ship itself, from the hull and appendages forms to its propulsion pl~nt and auxiliary machinery layout and mounting, to its structure

As the general design problem does not fall within the scope of this paper, the studies and investigation carried out on one of the main sources of vibration and noise for both naval and merchant ships will be dealt with the propeller For several years now, CNI and CETENA, in cooperation with the Italian Navy, have been working in this field, acquiring experience in propeller design aimed specifically at reducing noise and vibrations

Due to the complexity of the various phenomena related to the working propeller and to propeller-hull interaction, the problem of design was not considered to be completely solvable by a theoretical approach only, but was backed up by extensive experimental work This experimental work was carried out both on model-scale in a towing tank and a cavitation tunnel, and also full-scale for various types of ship Besides propeller design and performance forecasting methodologies, the data base acquired includes cavitation surveys, hull-induced pressures and radiated noise measurements, both model and full-scale, for several merchant ships, such

as container carriers, Ro-Ro vessels, bulk carriers, tankers, and, to a greater extent, for naval vessels, from the large helicopter carrier to the small missile-gunboat

2 GENERAL DESIGN METHODOLOGIES

The propeller design development procedure in use in the FINCANTIERI Group can be divided up, from a logical standpoint, into five main phases, briefly described herebelow :

2.1 The first stage consists of preliminary optimization of the hydrodynamic and main geometrical characteristics, in compliance with design requirements and constraints, such as ship design speed, required thrust, maximum power, ship and superstructure natural frequencies, propeller diameter, number of blades and/or turns This phase is carried out by parametric calculations performed with the "lifting line" approach, using the "ELIPACK" program package These calculations are aimed at determining the most efficient propeller among those which comply with the basic resonant frequencies avoidance, ship performance and blade strength requirements, and give a sufficient margin against bubble cavitation of the blades The considered load distribution along the blade

a certain degree of tip and root unloading (from 20 to 100 %, depending on the propeller under design) is adopted, to delay the inception of tip and hub vortex cavitation The effects of axial and tangential wake distribution on hydrodynamic performances are also taken into account in a stationary manner, as well as the effects of the radial load distribution

on propeller efficiency

2.2 Detailed definition of the geometry for some propellers, meeting the general requirements~ and selected during the previous phase (with different blade configurations, loads, chord length and skew distribution): this work is carried out by the "steady lifting surface" program

"PESP" developed by CETENA (ref 1), based on the theories of Pien and Kerwin The "lifting surface" approach allows definition of the blasfe geometry taking directly into account the tridimensional effects and skew effects on pitch and camber distribution

Some different chordwise loading distributions have also been investigated and designed, using the PESP program, with various degrees of leading-edge unloading, in order to further delay sheet cavitation Again, this program allows exact forecasting of blade cavitation, during a revolution in the hull wake,· following a quasi-steady approach based on an equivalent bidimensional section theory (ref 2)

2.3 The third phase consists of theoretical estimates of the propeller operating in its non-uniform wake with assessment of fluctuations in thrust, torque and bearing forces, blade sections, extent of cavitation at various angular positions and hull/stern-induced pressure due to load and cavitation fluctuations These calculations are carried out by an

"unsteady lifting surface" program, "PRESS" (ref 3 )developed by CETENA on the basis of work by Tsakonas et al (ref 4), and Geurst's theories The effective wake distribution adopted in these calculations is obtained from the usual tridimensional wake surveys carried out on models in a towing tank, but corrected for both scale effect (Sasajima method) and propeller effect (from self-propulsion results)

The estimation of these latter effects has proved to be the most critical point in the forecasting of the undesired phenomena associated with a full-scale propeller operating in a strong wake field, as is the case of most single-screw ships

The results of this theoretical investigation of the overall behaviour of the various propeller solutions singled out in the previous phases are analyzed and compared For one or two of them, showing the best performances, the relevant model/s are defined in detail and accurately manufactured using numerical control machinery

2.4 The fourth phase is devoted to model experimental

aimed at verifying the total propeller performances,

cavitational behaviour, hull-induced pressures and

investigation, such as efficiency, underwater radiated

noise

After open-water propeller tests and model self-propulsion tests at the towing tank with the previously defined propeller alternatives, cav itation experiments are carried out, generally at the Italian Navy Cavitation Tunnel (CEIMM) During these experiments, visual observation of the cavitating behaviour of the propeller is used to discover the inception of each phenomenon at various propeller operating points and

full-scale operating points, deduced on the basis of self-propulsion tests, with suitable correlations, is also performed

At the same time, measurements of propeller-induced pressures in way of a few points on the stern and of propeller radiated noise are carried out and analyzed

All the tests are carried out simulating the ship's wake by arranging the propeller downstream from a "dummy model" hull, reproducing in a suitable fashion the stern of the ship and its appendages Implementation of the dummy model is studied from time to time on the basis of the towing tank

cavitation tunnel is then measured by the LDV technique

As far as concerns the methodology adopted at CEIMM for both cavitation and pressure and noise measurements, same is described in length in ref

summarized herebelow :

- Cavitation tests are carried out by implementing the identity of the full-scale propeller advance coefficient and cavitation index, but,

disregarding the Froude similarity, the propeller r.p.m is set as high

as possible in order to reach the highest Reynolds Number, due to its great importance for both the blade flow pattern and the inception of cavitation

- The cavitation inception curves obtained are then scaled to full-scale

"buckets", by taking into account the di fferences in the Reynolds

reliability of which has been proved by several model and full-scale experiments

- Due to the scale effect between model and full-scale cavitation inception diagrams, the pressure and noise surveys at the most significant propeller operating points are then carried out as described

cavitation index (as compared to the full-scale index), so as to better reproduce the blade cavitation phenomena and its extent as forecast in full-scale conditions Fig I illustrates qualitatively the above procedure for scaling of cavitation inception curves and for the definition of the propeller operating points during pressure and noise measurements at the cavitation tunnel

interesting improvements, mainly in the correlation between tunnel and

the Italian Navy are discussed

- The values measured for pressure induced at various points of the dummy model, above the propeller, are scaled to full-scale values, assuming

propeller speed and diameter respectively

- Propeller radiated noise measurements are carried out according to the methodology and set-up described in ref 5 The measurement chain is

essentially to that suggested by the ITTC Cavitation Committee

increased, and the final sound pressure levels derive from an averaging procedure The noise signal is analyzed in the 1/3 octave band, and both frequencies and sound pressure levels are scaled to full-scale

The resulbs obtained for the propeller models tested using the procedure outlined above are then compared and analyzed, leading to the choice of the best solution, if satisfactory, or to recommendations for further modifications in design, if necessary

2.5 Full-Scale Tests

The fifth phase, although it should not, strictly speaking, be defined as

a propeller design phase, is considered by us to be of paramount importance mainly for very demanding projects such as the design of propellers for naval units Indeed, these ships are generally built in a certain number (Class of ships), so that the full-scale results obtained

on the prototype can lead to further refinement and improvement, easily applicable to subsequent sister ships

But mainly, the collection of reliable data obtained from the full-scale product is essential as a feed-back referred to the preceding work phases and for increasing the data base and know-how of any Company which aims at optimizing its product capability and the quality of its production Without going too deeply into the various kinds of full-scale trials and into the instrumentation adopted for both measurement and data analysis, which are described in detail in several CETENA reports, listed below are the full-scale tests generally carried out at least on every new prototype ship built :

- Ship Speed Survey, using the Raydist System (over the entire speed range)

entire speed range)

- Hull Vibration as well as Local Vibration Survey

- Artifical Vibration Exciting Test (using the Vibrodyne method)

- Ambient Noise and Machinery Noise Tests

- Propeller Cavitation Observations (by strobo-lamp through portholes)

- Propeller-Induced Pressures Survey (using hull flush-mounted pressure transducers)

- Near-Field Propeller Radiated Noise Survey (by hull-flush-mounted hydrophones)

- Manoeuvrability tests (using the Raydist System for ship tracking)

- Far-Field Radiated Noise Measurements (at an acoustic range), only for naval units

3 APPLICATIONS OF THE METHODOLOGY DESCRIBED AND EXPERIENCE ACQUIRED The propeller design, verification and testing methodology described has been applied during the last decade by the most important Italian Shipyards, on a more or less consistent fashion, to a large number of newly built ships with different operational characteristics and performance requirements While some specific applications have already been reported and described in detail in several previously published papers (ref 8, 9, 10, 11) it was thought to be of a certain interest, in this paper, without focusing in detail on a specific case, to summarize very briefly some representative experiences acquired by our Companies on the subject of the propeller as a source of vibration and noise for different types of ships

To this end, some of the results obtained for both merchant and naval ships are illustrated and discussed in the following sub-paragraphs: 3.1 Bulk Carriers and Tanker Ships

Supported by our long experience in this most conventional type of ship,

it can be stated that the problem of propeller-induced vibrations may be considered of minor importance, assuming that a careful selection of the main propeller parameters (diameter, number of blades and r.p.m.) , based

on a thorough knowledge of the hull and superstructure natural frequencies has been made Indeed, the generally high propeller-hull clearance values and the relatively small stern surface above the propeller lead to a relatively low excitation level Nevertheless, the main results obtained

by the propeller veri fication procedure described above, applied to a medium-sized oil tanker, are reported as a typical example(ref 12) The ship, of which the main characteristics are given in Fig 3 , did not experience any classical vibration problem either at full-load or in ballast operating conditions, but a quite high airborne noise level was recorded in some crew accomodation areas (two cabins), in the aftermost part of the ship, below the weather deck

The calculations carried out during the propeller design phase are summarized in Fig 4a, b, showing the induced pressure distribution above the propeller plane and the induced vertical surface forces (longitudinal distribution) at blade rate and at twice the blade rate

The pressure pulses (1st blade harmonic) measured subsequently, during full-scale trials, are also reported and compared in the same figure During the design phase the ship structure was schematized and a complete 3D FEM calculation was carried out

On the basis of the calculated exciting surface forces, the forced response of the ship was derived, and for a comparative analysis both the theoretical and full-scale experimental results are shown in Table 1 , for

Ship Vibration Response - Calculated and Measured Amplitude Values

in Full-Load and MCR Ship Conditions Measurement points

1 74 x 10-3 cm

two significant points along the ship As can, :be seen, a very good correlation exists at the afterbody and an acceptable correlation at the navigation bridge

Furthermore, the complete vibration survey on the ship showed very low amplitudes throughout the classical low frequency range, as had been foreseen

The explanation for the relatively high noise levels in the aftermost part

of the ship can be found in Fig 4c, showing the harmonic content of the measured induced pressure above the propeller tip, at the M.C.R Power Even if of a relatively low absolute value, the harmonic content has an unpredictable spreading over at least seven blade harmonics, as was also confirmed by the measurements carried out by means of a hull-flush mounted hydrophone In fact, the sound spectrum measured showed a very high level

in the low frequency range (20 to 80 Hz), and a normal range at higher frequencies, as can be seen in Fig 5, where sound spectra are reported for different ship speeds

On the other hand, the measurements of the airborne noise inside the ship,

in spaces just above the propeller (after peak and steering-geear room) showed the same trend, as illustrated in the same figure with reference to the M.C.R ship operating condition

3.2 Ro-Ro Ship

As already stated, for this type of ship the propeller-hull induced excitation level may rise to critical values, leading to an unacceptable vibratory behaviour of the ship, and consequently to a quite high level of airborne noise on board In fact, the generally open main structure, the large stern counter (large surface above the propeller), and the very uneven wake field in which the propeller operates are all elements which contribute towards increasing the vibration and noise problems on board these ships

In this respect, application of the complete outlined propeller design methodology has proved its efficiency and reliability, leading to quite satisfactory results in some recent instances of propeller design and ship building

It was found, during the design of various propellers, that two factors

were of paramount importance : on one hand accurate assessment of the foreseen full-scale wake field, and on the othe hand the effectiveness of

the ship wake unevenness

As typical examples of the problems experienced in the past on this kind

of propeller induced pressure levels and the relevant ship stern vibratory behaviour as a result of a full-scale vibration survey on a 30,000 DW tons Ro-Ro ship The main characteristics of the ship in question are shown in Fig 6{see also ref 10)

The measured and calculated (by considering an estimated full-scale wake field) propeller-hull induced pressures on some significant points of the

harmonic contents are reported As it can be noted, the measured pressure

frequency of which was very close to the natural frequency of the

vibration and induced noise trouble in the accomodation areas of the ship The correlation between the harmonic pressure content, as measured above the propeller tip, and the vibration harmonic content measured at a stern

It must be pointed out that the experienced pressure harmonic content was not previously foreseen by theoretical calculation (by considering the ship model wake field) nor by the model-scale pressure survey

3.3 Container Ships

The methodology described was recently fully applied in redesigning the propeller of a container ship, when derating its propulsion plant (steam

interesting also under the "energy saving" aspects, which today form the basis of propeller and propulsion plant refitting for more and more ships

of different types

In this case, the decreased ship speed and engine output also prompted redesign of the propeller by adopting a different number of blades and a large reduction in the blade area ratio, aimed at increasing the propulsion efficiency while maintaining the good vibratory behaviour of the ship

The efficiency increase obtained can be seen in the following table, while the goal of a reduced excitation level was reached through accurate

3.4

Table 2

% power red with

ne~,1 propeller 12.8

11 9

9.9

% cons red with rating and new propeller 16.6

de-14.7 8.5

tunnel measurements The comparison of these values over the propeller tip

is shown in Fig 9 Comparative model tests of the levels of noise induced by the propeller were also performed, and are summarized in Fig

10, where it can be seen that the original and redesigned propeller generate a fairly similar sound level distribution

Full-scale measurements and cavitation observations confirmed satisfactory alignment with the forecast induced pressure values as well as cavitation behaviour of the propeller (see Fig 11), while the vibrations measured on board were maintained within an acceptable level, as shown in Fig 12 3.4 Passenger Vessel

For this kind of ship, generally devoted to pleasure cruises, the passengers' comfort is one of the most important requirements, and consequently very low onboard noise and vibration levels must be achieve~

Furthermore, refittings or changes in the operational conditions during the life of the ship may alter the original set-up, often generating a decrease in propeller efficiency and induced vibration and noise problems A passenger liner from the sixties was subsequently allotted to pleasure cruises, with some significant alterations, such as a lengthening of a bridge to create a dance hall and reduction of the cruising speed from 21

to 16 knots This speed reduction gave rise to a drop in the steam ·turbine propulsion plant efficiency The hull was not suited to the new operating speed, and in addition a troublesome vibration level on the new bridge and

in the surrounding cabins at speeds from 15 to 19 knots made it impossible to operate the ship'at-the~~peeds durIng dancing parties The solutions adopted to overcome this problem were redesign of a bulbous bow better suited to the new speed and redesign of the propeller, changing the number of blades (from 3 to 4) and the adoptionbfan adapted propeller load distribution and skew in order to minimize propeller induced pressure on the hull To increase the turbine plant efficiency, the propeller r.p.m at design speed was also slightly increased

The power reduction obtained by means of the adoption of the new bulbous bow and the new propeller was about 6 ~~ at design speed, and a comparison

of power and r.p.m versus speed is shown in Fig 13

But the most important result was achieved in the reduction of vibration with the change in blade frequencies and skew adoption, as shown in Fiq

14, where the vibratory behaviour (peak l~els) for both the original the refitted ship (stern point measurement) are compared The considerable reduction in the vibration peaks is very noticeable

and-3.5 Naval Vessel Applications

Al though design of any propeller entails an exacting study, propeller design for a naval ship is one of the most demandillg tasks for hydrodynamicists and ship designers This is because the designer must match a certain number of different design requirements and constraints,

such a very high propulsion efficiency throughout the entire speed range

of the ship (from endurance to maximum ship speed), proper fulfilment of strength demands, high cavitation-free ship speed, and low levels of hull-excited vibrations Moreover, a low propeller-radiated noise signature is of paramount importance, mainly in the lower speed range but under quite different operationg conditions, such as free sailing or under tow (different kinds of hydrophone arrays are usually towed by ASW naval vessels), as well as asymmetrical propulsive conditions, such as one-screw propelling willth the other windmilling or feathering (for the usual twin-screw arrangement)

The propeller design and verification methodology described in the previous paragraphs was set up at the beginning of the last decade mainly

as a specific requirement for improving naval propeller performances, and the feed-back from its applications has added considerable wealth to the general know-how and data base available to propeller designers Due to the mani fold operational requirements of naval vessels, C P Propeller applications proved to be the most reliable solution, so that great experience has been acquired by FINCANTIERI CNI in the design of both controllable-pitch propeller mechanisms and propeller blades and hubs, as demonstrated in recent extensive work carried out by CNI within the framework of the feasibility study for the NFR 90 (ref 13)

Of course a certain amount of Fixed-Pitch Propeller designing has also been done by CNI, and many recent newly built naval ships having less complex operational requirements are equipped with this kind of propeller Instead of reporting detailed results for speci fic instances of applications (to be found in other previously published papers - ref

S, 9, 14 ) , we shall report and discuss herebelow, on a comparati ve basis, some of the principal results obtained from theoretical as well as model-scale and full-scale investigations carried out on the C.P propellers of first and second generation Frigate Classes Some data gathered from model-scale and full-scale tests carried out on the fixed-pitch propeller of a large naval cruiser will also be given and discussed

- First and Second Generation Frigate Classes

Due to the paramount importance attributed to the subject of underwater radiated noise, First Generation Frigates have been subjected, during the seventies, to extensive acoustic surveys with the purpose of checking the overall ship design reliability from this point of view, as well as defining the contribution of the propellers to the total acoustic signature of the ships

At the same time, further investigation was carried out on the model propeller at the Italian Navy Cavitation Tunnel in order to improve testing methodologies for induced pressures and radiated noise, as well as

to investigate the geometri~al characteristics of new blade sections and their cavitational performances

As far as concerns the full-scale noise survey, besides tests carried out

in two acoustic ranges, a technique for propeller radiated noise

hydrophones in the vicinity of the propeller itself The results obtained

by applying this technique proved to be sufficiently representative of the total ship noise signature, especially for higher ship speeds, where the propeller is the dominant source of noise As a confirmation of this, Fig

15 compares the ship's total signature spectrum levels as ascertained at

hydrophones

Another interesting feature offered by the hull-flush mounted hydrophones consists of the possibility of investigating the most reliable and effect-ive blade pitch value, from the point of view of radiated noise, for low speed operation both in free sailing and under tow In Fig 16 the main results of an investigation of this kind are reported, referred to a free sailing speed ranging from 12 to 15 knots

As far as the investigation of new blade sections is concerned, full-scale application of a modified camber line and chordwide distribution of thickness based on cavitation tunnel investigations and results was tested

on a sister ship of this First Generation Frigate Class The main results may be summarized as follows :

- propeller efficiency was slightly decreased over a limited ship speed range

- the cavitation-free ship speed was increased by about two knots

consequently the propeller radiated noise levels dropped appreciably in the critical-ship speed range, as shown in Fig 17, which compares the two propeller noise spectra for a ship speed of 16 knots

Further investigation on this subject is currently under way at the cavitation tunnel

The Second Generation Frigate design benefited by the experience gained previously, mainly as far as concerned propeller design optimization, while improvements in cavitation tunnel model testing techniques have led

to a closer degree of correlation between model-scale and full-scale data The described propeller design methodology was applied, and particular

skew and load distributions, as well as the mean line and thickness of the modified sections

One of the most important improvements in this design was achieved in the

above the propeller tip (1st blade harmonic) by the propeller of this Second Generation Frigate are compared to those of the First Generation Frigate for very similar ship operating conditions

The correlation between model (cavitation tunnel) measurements and full-scale (with hull-flush mounted hydrophones) radiated noise levels can

be seen in Fig 19, where for the two most significant ship speeds the model and full-scale propeller noise spectra are compared

In the same obtained by

acoustic range measurements are also shown

technique adopted for model testing does need further veri fication and validation, so that other full-scale measurements have already been scheduled on some other naval ships

For the sake of completeness, it should also be considered that other means for reducing propeller radiated noise throughout the entire ship

spe~d range have been applied for the Second Generation Frigate One of the most important means is a propeller air blowing system, designed and applied in order to reduce radiated noise in that ship speed range where cavitation phenomena and associated noise are unavoidable Optimization of the air flow was obtained by thorough investigations carried out during the ship acoustic survey, at an acoustic range

reduction in the propeller radiated noi1le was achieved not

ship speeds where blade cavitation is well developed,

only at higher but also for

- As a final example of application of the outlined propeller design and testing methodology, some results obtained for a fast and high-powered

characteristics of which are described in ref 14 , are given below

During the exacting design of this propeller, the main efforts were devoted, as well as to increasing the propeller efficiency, to delaying the inception of cavitation as much as possible and to minimizing both the bearing fluctuating forces and the propeller induced pressures and surface forces

As a result of the extensive model-sca1e investigation carried out (three propeller models were made and tested in the towing tank and cavitation

degree of accuracy achieved is noticeable in Fig 20 where the various sketches are illustrated A similar comparison referred to the total

noise, again measuring by hull-flush mounted hydrophones, some results are illustrated in Fig 22, where the increase in the noise spectrum level as

a fun~tion of the ship speed can be seen In the last figure, Fig 23, the correlation between the full-scale and model-scale propeller radiated

excellent, this correlation may be defined as acceptable from a design point of view

4 CONCLUSIONS

In the previous paragraphs, propeller design verification and testing methodologies as developed and applied by CElENA and the main Italian