For Minimizing Vibration In: A Design

SSC-29 i
A DESIGN PROCEDURE
FOR MINIMIZING
PROPELLER-INDUCED
VIBRATION IN HULL
STRUCTURAL ELEMENTS
This document has been approved

for public release and sale; its
distribution is unlimited.
SHIP STRUCTURE COMMITTEE

1979
-,-
SHIP STRUCTURE COMMITFEE
The SHIP STRUCTURE COMMITTEE

is constituted to prosecute a research
program to improve the hull structures of ships and other marine structures
and methods of
by an extension of knowledge pertaining to design, materials
construction.
R4DM H. H. BELL (Chairman) Mr. M. PITFJIJ
Chief, Office of Merchant Marine Assis tant Administrator for
Commercial Deve 1ap ment
Safety Maritime Administration
U. S. Coast Guard
Mr. P. M. PALEF?MO Mr. R. B. XPAHL
Director Chief, Branch of Marine Oil
Hull Integrity Division and Gas Operations
Naval Sea Systems Command U. S. Geological Survey
Mr. C. J. WHITESTONE
Mr. W. N. HA.NNAN
Vice President Chief Engineer
American Bureau of Shipping Military Sealift Command
LCDR T. H. ROBINSON, U.S. Coast Guard (Secretary)
SHIP STRUCTURE SUBCOMMITTEE
The SHIP STRUCTURE SUBCOMMITTEE acts for the Ship Structure

Committee on technical matters by providing technical coordination for the
determination of goals and objectives of the program, and by evaluating
and
and interpreting the results in terms of structural design, construction
operation.
MILITARY SEALIFT COMMAND
U.S. COAST GUARD
Mr. T. W. CHAPMAN
CAPT R. L. BROWN (Chairman)
Mr. A. B. STA VOVY
CDR J. C. CARD
LC'DR J. A. SANIAL, JR. Mr. D. STEIN
C'DR W. M. SISON, JR. ARICAN BUREAU OF SHIPPING
NAVAL SEA SYSTEMS COMMAND
Dr. H. -Y. JAN
Mr. R. CHIU Dr. D. LIU
Mr. R. JOHNSON Mr. I. L. STERN
Mr. J. B. O'BPJEN ADMINISTRATION
MARITI
Mr. G. SORI'JN
U. S. GEOLOGICAL SURVEY Mr. F. J. DASHNAW

Mr. N. O. HAM]R
Mr. R. GIANGERELIJ Mr. F. SEIBOLD
Mr. J. GREGORY Mr. M. TOlIMA
INTERNATIONAL SHIP STRUCTURES CONGRESS
NATIONAL ACADEMY OF SCIENCES
SHIP RESEARCH COMMITTEE
Mr. S. G. STIANSEN - Liaison
Mr. Q. H. OAKLEY - Liaison
Mr. R. W. RLThfl - Liaison AMERICAN IRON & STEEL INSTITUTE
ThE SOCIETY OF NAVAL ARCHITECTS Mr. R. H. STERNE - Liaison

& MARINE ENGINEERS
STATE UNIVERSITY OF NEW YORK MARITIME COLLEGE
Mr. N. O. HA'R - Liaison Dr. W. R. PORTER - Liaison
WELDING RESEARCH COUNCIL
U. S. COAST GUARD ACADEMY
Mr. X. H. KOOPMAN - Liaison
CAPT W. C. NOLAN - Liaison
U. S. RCHANT MARINE ACADEMY
U. S. NAVAL ACADEMY
Dr. C. -B. KIM - Liaison Liaison
Dr. R. BRAT'TACIL4RYYA -
29
Member Agex: Addrem Corresponden to:
United Stat Ccxt Guard
Seretary, S111P Structure Con2mitt
C
N0vr11 Sei Systems Command
Military Sealift Command US. Ct Guard Headquarters, (G-M/82)
Marilzme Admistrctzan Wan&ngton, D.C. 2059G
United Striter Geo/ogicx2/ Survey
Amencxzn Bureau of Shippng Ship
Structure
Committee
An Interagency Advisory Committee
Dedicated to Improving the Structure of Ships
SR- 12 40
September 1979
The rapid advance in ship size and power and the trend
toward lighter hull scantlings prompted the Ship Structure
Committee to investigate the propeller-induced vibrations in the
hull and superstructure of the ship. High vibratory forces in
the ship can cause discomfort in the living
quarters, excessive
"panting" type deflection of tank bulkheads, and fatigue
cracks
in webs and plating.
The first phase developed a bibliography published as

SSC-281. It was made available for the October 15 - 16, 1978,
Ship Vibration Symposium, sponsored jointly by the Ship Structure
Committee and the Society of Naval Architects and Marine
Engineers.
The subsequent phases, including the development of a

design procedure for minimizing propeller-induced vibration in
hull structural elements, have been completed and are reported
here.
nry H. Bell
Rear Admiral, U.S. Coast Guard
Chairman, Ship Structure Committee
METRIC CONVERSION FACTORS
Approximate Conyersiono horn Metric Measures

Approximate Conversions to Met,ic Measures
Multiply by To lind Symbol
Symbol Whoe you know
When You Know Multiply by Ta Fryd Symbol

Symbol
LENGTH
dies IA
- noii milllnielers 0.04
LEN 6TH licItes in
9 cnt centimeters 0.4
3,3 leet It
nl motets
yards yd
nu meters 1,1
n inches 2.5 ce,,r,nmle,s lilIes in,
ce,,t,mouters cm kni ltiknnetnrs 0,5
net 30
ft iii
yards 0.9 linIers
yd bnl
mites 1.6 kilo.nn;ers
si, AREA
AREA squame inches n'
- cn'7 square centinletnes 0.16
1.2 squire yards ydt
cm2 9 m2 square meters
squatS inches 5.5 Square cent,melers sillmili 11111es te,'
1fl2 bn,2 square kilcinieters 0.4
0.09 Squalo rioters u2 acorro
It' Square lent .02 liii t,octa,es 110.000 rit2) 2.5
squire yards 0.9 square nu'tcrS
0d2 bn,2
square nulos 2.6 5qi..ivt kila,iet.'is
ill,' lucta,i,S lia
0.4
MASS (weight)
MASS (weight) ounces er
9l100s 0.035
9 jnrrid5 lb
kilogranis 2.2
ounces 28 aliolis bI) short tolls
on 1.1
0.45 kiloujran.0 kg tonnes (1000 kg)
lb pounds
sht toes 0.9
e
12000 lb)
VOLUME
VOLUME
0.03 Ilaid tullIdOs Il or
nil trmillilìturs
nil pulls fIl
teaspoons S nhilliliteru t litets 2.1
(sp nil gnarls ql
tablespoons 15 nijlluliters liters 1.06
tbsp iu,lluI,te,s ml gallons gut
lluni ounces 30 (,tors 0.26
II or I. Ir'
liters t 35 cubic lue!
cups 0.24 m3 cubic motels
liters 1.3 cub,c yards ydt
pr pints 0.41 ni3 cubic meterS
Ittets t
qUarts 0.95
gallons 3.8 liters
gal li3
cubic met 0.03 Cubic melero TEMPERATURE (exact)
lt3 o.3
Cubic yards 0.76 cubic nioters
yd3
Celsius 9/5 (thon Fahrenheit
TEMPERATURE (exact) n'
InnIperulutre add 32) temperature
Fohronheit 5/9 (alter Celsius

F subtracting temperature 012
temperature oF 32 986
lO 7190
32) 0 411 00 170
_fl 90 ira in lIti 40( 'ra 'n
60 po lOSt
nut nue &la,ie,l 101,1%. see NtIS Slut. P,diI. 2lt, '20 0 50 140
o 2.30 tn.rtliyl. t n
tutet eflaCt termins -40 3?
Ct3.ty286. E cc
Units cl 0u,0rit ant l,ae.rs,rcs. Poco sl.25. 00.1 Carrluii Nu.
Technical Report Documentation Page
1. Report No. 2. Government Accessfo No. 3. Recipient s Catalog N0.
SSC-291
4. Title and Subtitle 5. Report Date
A DESIGN PROCEDURE FOR MINIMIZING September 1979
PROPELLER INDUCED VIBRATION IN HULL STRUCTURAL 6. Porrrrg Organi:mttiort Code
ELEMENTS SWRI 02-4821
8. Performing Orgoni Ration Report No.
7. Author's)
O. H. Burnside, D. D. Kana, and SWRI-02-4821
F. E.. Reed
9. Perforrrrirrg Orgaiizo?ion Nome and Addiess IO. Womit Unit No. (TRAIS)
Southwest Research Institute
6220 Culebra Road, P.O. Drawer 28510 11. ContractorGrantNo,
San Antonio, TX 78284 DOT-CG-61907-A
13. Type f Report and Period Covered
12. Sponsoring Agency Nome ond Address
Fi nal Report
U. S. Coast Guard 2/4/77 thru 4/27/79
Office of Merchant Marine Safety
Washington, D.C. 20590 Sonsorirro Agertcy Code
G-M
15. Supplementary Notes
16. Abstract
A design procedure for minimizing propeller-induced vibration in hull struc-

tural elements is recoimnended. This procedure begins when the ship's vibra-
tion specifications are defined and continues through the design and con-
struction process until the vibration levels measured during sea trials are
compared with the specifications. Consideration is given to the hydro-
dynamic excitation and structural response of the propeller-induced vibra-
tion problem, with both analytical and experimental techniques being used
in the design process. The recommended procedure is presented and discussed
in the form of a flow diagram with 27 separate design steps. The process
also contains five evaluation milestones. At these points, the design is
assessed, and, if deficiencies are found, corrective action can be taken
before the design proceeds. The recommended complete procedure is presented
in this report for the first time. Many of the aspects of this procedure
are still being developed, in particular, the influence of propeller cavi-
tation on hull pressures and a simple but accurate treatment of water in-
ertia. These indefinite aspects have to be treated empirically using judg-
ment and experimental data. The portions of the procedure which are avail-
able are illustrated in an example using a single-screw, containerized and
unitized cargo ship.
17. Key Words 18. Distribution Statement

Propellers Structural Analysis Document is available to the U.S. Public
Vibration Hydrodynamic Forces through the National Technical
Ship Hull Structures Cavitation Information Service, Springfield, VA
19. Security C(asif. (of tb s report) 20. Security CIas f. (of this page) 21. No. of P ages 22. Price
Unclassified Unclassified 160
Form DOT F 7OO.7 )B-72) Reproduction of completed page authorized

LIST OF ILLUSTRATIONS (Cont'd)
Figure Page
17 Finite-Element Mesh for Two-Dimensional Model

of 370,000 dwt Tanker, from [75] 55
18 Forced Response Depending on the Applied Global

Damping Value. 370,000 dwt Tanker, Ballast
Condition, from [75] 55
19 Calculated Forced Response at the Top of Super-

structure in Longitudinal Direction. 370,000
dwt Tanker, Ballast Condition, from [75] 55
20 Position of Nodal Points on the Main Deck for

Forced Vibrations Calculations of the Hull
Girder. 370,000 dwt Tanker, Ballast Condition,
from [75] 55
21 Elasto Dynamic Model of Aft Part and Correlation

of Exciter Tests with Free Vibration Calculations,
from [94] 56
22 Preliminary Hydrodynamic Design Phase 61
23 MILESTONE I - Preliminary Hydrodynamic Design

Evaluation 62
24 Axial Wake Distributions for Original and

Modified Body Lines, from [97] 63
25 Final Hydrodynaniic Design Phase 65
26 MILESTONE II - Final Hydrodynamic Evaluation 66
27 Ship Substructure Design Phase 69
28 MILESTONE III - Ship Substructure Evaluation 70
29 Complete Shíp Structure Design Phase 71
30 MILESTONE IV - Complete Ship Structure Evaluation 72
31 Test and Evaluation Design Phase 74
32 MILESTONE V - Test and Evaluation Review 75
33 Outboard Profile 79
34 Propeller in Aperture 79
LIST OF ILLUSTRATIONS (Cont'd)
Figure Page
35 Frequency of Longitudinal Vibration Versus

Foundation Stiffness 80
36 Ratio of Measured Hull Pressures Where Cavitation

Exists to Calculated, Noncavitating Pressures 88
37 Vertical Harmonic Force and Transverse Bending

Moment Generated on Hull by the Propeller 89
38 Inboard Profile 92
39 Grid Points on Frame 170, X = 396 In. 93
40 Grid Points on Frame 181, X = O In. 94
41 Amplitude of Axial Motion at the Propeller 99
42 Amplitude of Fore and Aft Vibration at Propulsion

Shaft Thrust Collar and Thrust Bearing Foot 100
43 Vibratory Motion on Bridge, Frame 164 at Center-

line 101
44 Vertical Vibration on 36-ft Flat Generated by

Axial Propeller Force 102
45 Vertical Vibration on 26-ft Flat Excited by

Axial Harmonic Force at the Propeller 103
46 Vertical Vibration in Tank Top due to Longitudi-

nal Excitation at the Propeller at 12.43 Hz 104
47 Vertical Vibration in Tank Top due to Longitudi-

nal Excitation at the Propeller at 10.8 Hz 105
48 Double Amplitude of Sixth Order, Fore and Aft

Notion of Thrust Bearing Foundation 108
49 Double Amplitude of Twelfth Order, Fore and Aft

Vibration on Thrust Bearing Foot 109
50 Vibration on 26-ft Flat at About 100 RPM 111
51 Vibration on 36-ft Flat at About 100 RPM 112
52 Vibration of Bridge Deck at About 98.6 RPM 113
53 Vibration of Bridge Deck at About 102 RPM 114
-X-
LIST 0F ILLUSTRATIONS (Cont'd)
Figure Page
54 Experimental Deflection Patterns on Tank Top

at Q at 10, 11, 12, and 12.5 Hertz 115
55 Experimental Deflection Patterns on Tank Top

at Q at 12.9 14, and 1 6 Hertz 116
56 Response to Shaker Excitation 117
Fi Longitudinal, Tangential, and Axial Wakes at

0.335 Radius 152
F2 Longitudinal, Tangential, and Axial Wakes at

0.520 Radius 153

0.723 Radius 154

0.950 Radius 155

1.100 Radius 156
LIST OF TABLES
Table Page
1 Summary of Design Block 2--Establish General

Ship Design Data 12
2 Summary of Design Block 4--Estimate Longitudinal

Propulsion Frequencies 17
Summary of Design Block 5--Design Propeller 18

3
4 Summary of Design Block 6--Compute Propeller

20
Forces
5 Summary of Design Block 7--Compute Hull Pressures

22
Without Cavitation
6 Summary of Design Block 10--Direct Calculation of

Cavitation Pressures and Forces 27
30
7 Some Cavitation Test Facilities
8 Summary of Design Block 14--Determine Forced

Longitudinal Response of Shafting 34
9 Natural Frequencies of Double Bottom and Shaft

for Separated and Integrated Models, from [67] 36
Summary of Design Block 15--Determine Forced

Response of Machinery Space 39

Lateral Response of Shafting (Rigid Hull) 40

Lateral Response of Shafting (Flexible Hull) 43
Correlation Between Measured and Calculated

Superstructure Fundamental Resonant Frequency
for Different FiniteElement Models. 138,000 dwt
Tanker, Ballast Condition, from [76] 44
14 Summary of Design Block 18--Conduct Superstructure

45
Modal Analysis
47
15 Summary of Design Block 20--Design Local Plating
16 Summary of Design Block 21--Assemble Model of

53
Entire Ship
LIST OF TABLES (Cont'd)
Table Page
17 Summary of Design Block 22--Determine Vibration

and Stress Levels of Entire Ship 53
18 Summary of Design Block 24--Assess Location

Vibrations, Structural Damping, and Modeling
Techniques 58
19 Pressure Data at 16 Locations on Seabridge Ship 84
20 Computed Cavitation Factors for the Three Ships 87
21 Hull Forces and Moments Due to Cavitation Effects 88

TABLE OF NOMENCLATURE
A cross-sectional area
A ,A areas associated with grid points m and n

m n
a acceleration of piston m
In
a longitudinal clearance of propeller-hull, forward
a vertical clearance of propeller-hull
C factor reflecting density of fluid and proportions of pistons
c speed of sound in fluid
e critical damping coefficient
D propeller diameter
do derivative of ( ) with respect to time
E modulus of elasticity
F force on sphere in direction of vibration
F mutual force between areas Am and An

mn
F,F,F components of force in x, y, and z directions
f frequency
g hysteretic damping coefficient
I moment of inertia about transverse axis

y
I product of inertia relative to horizontal and vertical axes
yz
I moment of inertia about vertical axis
z
J torsíonal area constant about longitudinal axis

X
K thrust bearing foundation stiffness
Kp1 pressure coefficient for first harmonic
Kp2 pressure coefficient for second harmonic
K12 dimensionless pressure variable

thrust coefficient
KTB thrust block stiffness
KA
xy
shear area constant transverse plane
KA
xz
shear area constant vertical plane
L length of ship
xyz
M ,M ,N components of moment in x, y, and z directions
n number of propeller revolutions per second
p hull pressure
po static pressure at centerline of propeller shaft at propeller;

single amplitude of vertical component of propeller-exciting
force in pounds (at blade frequency)
vapor pressure
R radius
r distance between dipole's center and location of desired

pressure; radius from source
r minimum surface dístance between grid poínts m and n on ship's

hull
T propeller thrust
t time
V velocity
V axial velocity of water relative to propeller disc
V velocity of model
m
volume
W Taylor wake number
W maximum Taylor wake number

max
Y single amplitude in mils
y transverse coordinate of neutral axis
coordinates of shear center of section

z number of propeller blades
z vertical coordinate of neutral axis
angle of attack
displacement of ship in long tons
1,p2 amplitude of first (second) component of averaged pressure

fluctuation
change in angle of attack
e angle between a vector to measuring point and force vector
frequency in cycles per second
p mass density of fluid
a cavitation index
n
phase angle of first (second) harmonic
angular frequency
wt angular blade position (wt = O for blade in vertical, top

position)
V displaced volume, m3
xvi
I. INTRODUCTION
Overview of Program
At the outset of this study, it was recognized that various aspects

of the propeller-induced, ship hull vibration problem had received atten-
tion in many previous investigations. These ranged in character from
research-oriented technical papers (both theoretical and experimental),
to some papers and reports that dealt with certain parts of ship hull de-
sign. Thus, the primary objective of this program has been to conduct a
comprehensive survey and evaluation of these existing information sources
and methods for predicting the exciting forces and the response of the
hull structure to vibratory loads caused by the propeller system, and to
recommend for design those procedures which appear both practical and re-
liable. In meeting this objective, the previously fragmented information
is brought together into an overall design procedure which addresses the
complete design problem.
This program has been carried out under four tasks. Initially a
literature search was conducted to establish the state-of-the-art for
prediction methods currently available. The results of this task served
as the informational basis for this report, and have been previously pub-
lished in the form of the bibliography identified in Reference 1. Empha-
sis in that document has been placed on the current generation of large,
high-powered vessels, so that, of the approximately 550 bibliographic
entries, over 60 percent were published since 1970. The second task dealt
with evaluation of the design procedures. This has included principally
a judgment about the suitability of the data for design purposes. The
results of this task formed the basis for the third task, which required
the recommendation of an overall design procedure and associated detailed
parts. Application of this procedure to a specific ship was demonstrated
in the final task. In essence, the results of all but the literature
search are documented in this final report. However, emphasis is placed
on the recommended procedure and additional practices appropriate for
overall design, with very little explanation of why some previously used
methods may have been omitted.
Definition of Propeller-Induced Hull Vibration Design Problem
In view of the fact that ship hull vibrations can be excited by a

variety of energy sources, it is appropriate that some definition of the
propeller-induced vibration problem be established, along with what gen-
eral concepts will be included in the design procedures established for
its solution. For simplicity, the problem to be addressed is shown con-
ceptually in Figure 1. Thus, only those vibration responses excited by
the propeller and its associated shafting are to be considered. Further-
more, a conceptual diagram of a desired design procedure is identified
in Figure 2. Hence, the design procedure is to start with a given set
of specifications, and progress wíth both analyses and tests to where the
design has been validated by suitable model and full-scale sea trials.
HULL
FORCES
RUDDER
FORCES
FIGURE 1. CONCEPTUAL IDENTIFICATION OF

HULL VIBRATION SOURCES
PRELI MI NARY FINAL

SPECIFICATION
HYDRODYNANIC HYDRODYNAMI C
PEAS E
PHASE PHASE
TEST AND CO1PLETE SHIP

EVALUATION SHIP STRUCTURE SUBS TRUCTURE
PHAS E PHASE PEAS E
FIGURE 2. CONCEPTUAL DIAGRAM OF

DESIRED DESIGN PROCEDURE
2
3. Consideration of Interdisciplinary Requirements
In view of the previously stated objective, it is obvious that de-

velopment of a sufficiently general design procedure is a formidable task.
This is especially true if it is to be applicable to many classes of ships.
To be successful, the design process involves several different specialized
naval architecture and marine engineering disciplines, as well as some others
from traditional branches of engineering. Some areas included are:
Naval Architecture and Marine Engineering

Ship Form Design
Propeller Design
Propulsion System Design
Rudder Design
Theoretical Hydrodynamics
Propeller and Hull Loading
Cavitation
Boundary Layer and Potential Flow Theory
(wake survey interpretation)
Experimental Hydrodynamics
Wake Survey
Model Tests
Cavitation Tests
Hull Pressure Tests
Structural Analysis
Propeller Shafts
Substructures
Main Hull and Superstructure
Acoustics
Human Response
Equipment Response
Experimental Vibration Force and Stress Measurements
With Shaker
In Service
It is obvious that no one person, and few engineering organizations,

have complete expertise in all the above disciplines. However, a proper
vibration analysis requires an understanding of the interrelationships
between all of these factors and their coordination with the ship design
procedure. Therefore, the design of a ship having acceptable vibration
levels has been and will be established from the technical input of several
sources. This is an important point. One should realize that a realistic
recommended vibration design procedure must mesh with other ship design
and construction processes. It is also important to consider the proced-
ures in toto, and not merely in terms of one of its parts. For example,
hydrodynamicists should not view the problem only in terms of a hydrody-
namic solution; structural engineers should not view the problem only in
terms of detuning the response from the excitation. Each group needs to
realize the other!s potential contribution to a solution and the necessity
of incorporating input from all necessary sources.
3
II. DESCRIPTION OF RECOMMENDED OVERALL
DESIGN PROCEDURE
In order to establish a recommended general design procedure for mini-

mizing propeller-induced vibrations, it was necessary to adopt a philosophy
on which the procedure would be based. Hence, a five-part design philoso-
phy was formulated, as follows:
Vibration Specifications Should be Quantitatively Defined

with Attention Given to Human Exposure, Machinery and
Equipment, and Structural Strength.
Excitation Forces Should be Kept to a Minimum.
Structural Resonances with Propeller Excitations Should

be Avoided.
Vibration Response Levels Should be Measured During Sea

Trials to Ensure Specifications are not Exceeded.
Measured Vibration Levels Should be Compared wíth Pre-

dicted Values to Assess Design Procedures.
The significance of this design philosophy will become more apparent

when details of the design procedure are discussed. However, a few general
coliuLlents are in order at this time. It is clear that for a ship hull de-
sign to be successful, there must first be selected a reasonable set of
design criteria, or goals, on which the process is to be based. This is
the purpose of Item 1 of the design phílosophy. These specifications should
be established ín the ship's design contract and serve as a standard in
guiding the design process. Vibrations levels recorded during sea trials
can also be compared with the specifications to judge the ship's accepta-
bility from a vibration point of view.
It is almost axiomatic that excitation forces should be kept to a

minimum, as stated in Item 2. A propeller mounted far aft of a ship's
stern may induce very little ship vibration, but this solution is not very
practical in terms of propeller efficiency, propeller whirl, and other
associated structural problems. What is meant is that attention should be
given to those factors which can reduce the excitation, e.g., stern config-
uration, propeller geometry and clearances, propeller wake, and cavitation.
There are many components of a ship's structure which can be ex-

cited by the propeller-generated forces and pressures. They include vi-
brations associated with the lateral, longitudinal, and torsional response
of the propulsion system; overall vertical bending and coupled lateral-
torsional bending of the ship hull; vibrations of major substructures such
as the engine room, machinery spaces, and superstructures; and response of
local structures such as the rudder and local plating. To make matters more
complicated, each of the above systems is coupled to some degree to the
others. One of the primary objectives of the design procedure outlined in
4
this report is to be able to predict accurately the various structural res-
onances of the ship and determine if they will be excited by the propeller.
If so, these resonances should be avoided because they will result in large
amplifications. This is the reasoning behind the design philosophy pre-
sented in Item 3.
Item 4 is normally part of the sea trials for the acceptance of the
vessel by the ship owner. The only additional comment which should be made
in this report is that vibration levels should be measured at the critical
locations throughout the ship. For human exposure, these include living
quarters, watch stations, steering gear spaces, machinery spaces, and cargo
spaces. For machinery and equipment, the longitudinal, lateral, and tor-
sional vibration levels of the propulsion system should be measured along
with those of any other critical components. Finally, the vibrational
stresses in critical structural locations should be monitored to ensure
that fatigue endurance limits are not exceeded. Critical areas would in-
clude, for example, bottom framing over the propeller, rudder and rudder
horn, stern bearing support, vertical columns on intersecting bulkheads,
and masts and spars.
The purpose of Item 5, in which the measured vibration levels are

compared with the predicted values, is to assess the validity of the de-
sign procedure. It is extremely important to conduct this post-mortem anal-
ysis because it allows the entire design process to be critically reviewed
to determine its strong and weak points. For example, if unacceptable
vibrations were measured on the bridge where none were predicted by the
analysis, the fault probably lies with inadequate structural modeling tech-
niques of the superstructure. If the stern plating vibrates at the correct
frequency, but at greater amplitudes than predicted, the problem could be
traced to underestimation of the propeller-generated pressures by the
hydrodynamic computer code or the influence of cavitation.
Having the previously defined philosophy in mind, we now introduce

in Figure 3 a flow diagram of the recommended design procedures for mini-
mizing propeller-induced vibrations. The procedure consists of twenty-
seven individual blocks ranging in time from the establishment of vibration
specifications to after the sea trials are conducted. Each one of these
individual blocks will be discussed in detail in Chapter III. In discus-
sing these sub-procedures, it is the intent not only to give perspective
to the function and purpose of each block, but to present detailed infor-
mation on how each can be used in the design process. This can best be
accomplished using tables summarizing the pertinent information.
The overall procedure is divided into six design phases: (1) speci-
fication, (2) preliminary hydrodynamic, (3) final hydrodynamic, (4) ship
substructure, (5) complete ship structure, and (6) test and evaluation.
The purpose of these phase designations is simply to give a qualitative
description of the overall design process in accordance with the design
which was given in the Introduction. Figure 3 also shows five evaluation
milestones which are located approximately at the end of each of the last
five design phases. The purpose of these evaluation milestones is to pro-
vide a means of assessing the design integrity up to that point. If it
is acceptable, the design may continue on to the next phase; if not,
5
DESIGN
SPECIFICATION
+ PRELLMINARY
HThRODYNAMIC
FINAL HYDRODYNAMIC
£ G
ESTIMATE
CONDUCT
LONGITUDINAL
CAVITATION
PROPL'LSI ON
TESTS
FREQUENCIES
o o o
DEFINE ESTASLISH CONDUCT IEVALUATE EVALUATE COMPUTE TOTAL
DESIGN
VIbRATION CENE LAL -* VANE -P PROPELLER -.4 NULL PRESSURES -$4 PROPELLER PRESSURES AND
FORCES
SPECIFICATIONS SHIP DES CN SLR\tY CAVITATION CAVITATION CAVITATION
IPROPELLER
DATA £ FACTORS
G DIRECT CALCULATION
CORE UTE
OP CAVITATION
PROPELLER PRESSURE AND
FORCES FORCES
V
f
CONDUCT
MODEL
'I TEST S
I.
(.ATIoN (EVALUATION \
%MILESTONE ij MII.ESTONE 11/
FIGURE 3. FLOW DIAGRAM OF RECOMMENDED DESIGN
6
- _.-o .'.y-ts.
PHASES
SHIP SUBSTRUCTURE COMPLETE SHIP STRUCTURE TEST AND EVALUATION
EVALUATE
LOCAL PLATING
DESIGN
i
G
DETEPINE NATURAL
FREQUENCIES AND
FORCED RESPONSE
OF RUDDER
G
DETERMINE DETERMINE CONDUCT COMPARE
FORCED FORCED
.ONGLTIJDINAL -* RESPONSE OF
SHARER
TESTS VIBRATIONS
RESPONSE 0F MACHINERY
SHAFTING WITH
SPACE
SPECIFICATIONS
ASSEMBLE MEASURED
G G ©
DETERMINE f DETERMINE
l MODE. OF SETER.NINE
Gv ©
FORCUD lATERAL FORCED LATERAL VIBRATION ASSESS MEASURE
RESPONSE 0F -Ø RESPONSE OF ENTIRE - ALITUDES
ANO STRESS
LOCAL VIB.
STRUCT. DAMP.
VIBRATIONS
SHAFTING SHAFTING DURING SEA
(RIGID WiLL) LEVELS OF MODELING TECH. TRIALS
(FLEX. HULL) SHIP COMPLETE SHIP
G o
CONDUCT COMPARE
SUPERSTRUCTURE MEASURED
MODAL ViBRATIONS
ANALYSIS WITH
£ CALCULATIONS
F
EVALIJATION'\
MILESTONE
PROCEDURES TO NINIMIZE PROPELLER-INDUCED VIBRATIONS
7
corrective action should be taken before the process continues. The ad-
vantage in having these frequent evaluations is that potential problem
areas can be identified and corrected early in the overall process. This,
hopefully, will avoid the all-too-frequent problem of having a certain de-
sign fixed, with changes possible only through costly modifications. The
evaluation milestones will be discussed in Chapter IV.
8
III. DETAILED STEPS FOR SHIP VIBRATION DESIGN
The procedures associated with individual design steps identified

as blocks in Figure 3 will now be discussed in detail.
1. Define Vibration Specifications
It is the goal of the entire vibration design procedure to secure

a ship which has a minimum of vibration. This goal cannot be reached,
however, without clearly establishing what vibration levels are and are
not acceptable to the shipowner. These levels must not be arbitrarily
set, but must be within certain limits to ensure safe and efficient oper-
ation of the ship. Too stringent conditions impose an unwarranted burden
on the shipbuilder and high design and construction cost, while the ship
may vibrate badly if the specified levels are too high.
Undesirable levels of shipboard vibration manifest themselves in

one or more of the following areas:
Human Exposure and Habitability
Machinery and Equipment
Structural Strength
Of these three, experience w.ith the current generation of ships clearly

shows that the greatest problem lies with human exposure and habitability.
This is due to the increased size, horsepower and speed of the vessel, and
the greater attention paid to the health and comfort of the crew. Refer-
ence 2 discusses in some detail the problem of shipboard vibration and
its effect on habitability. It is important to note that acceptable lev-
els of vibration for human exposure can be different in different portions
of the ship. For example, continuous exposure levels must be maintained
in the living quarters and watch stations, while less stringent require-
ments would be imposed in areas such as the steering gear, cargo, and
machinery spaces.
The machinery which is affected by propeller-induced vibrations is

usually associated with the ship propulsion system. Included are the
longitudinal, lateral, and torsional vibrations of the shafting system
and vibration in the main power plant. Other items of equipment particu-
larly sensitive to vibration would include those associated with naviga-
tion, communication, or special cargo which the ship is carrying. How-
ever, all machinery and equipment should be able to withstand levels
which are acceptable to humans.
The final way in which vibrations can be detrimental to the per-

formance of a ship is by reducing its structural strength through fatigue.
This is particularly a problem at highly loaded areas in the ship which
experience many stress reversals. Such areas would include the bottom
framing over the propeller, the rudder and rudder horn, the stern bearing
support, vertical columns on intersecting bulkheads, and masts and spars.
9
If we turn now to the set of specifications, they should include as
a minimum the following items:
Limits of acceptable vibration for human exposure,

machinery and equipment, and structural strength.
The types of experimental and analytical studies

which must be performed to ensure the requirements
are met and the extent of the documentation for
these studies.
The test requirements and methods for vibration

measurements during the acceptance trials.
The responsibility for correcting vibration prob-

lems should they occur during the acceptance trials
or during the subsequent warranty period of the ship.
From the literature it appears that the limits of acceptable vibra-

tion in humans are well established. Reed [31, in a 1973 paper, discussed
the question of acceptable vibration levels and pointed out that the Inter-
national Standard ISO 2631, "Guide for the Evaluation of Human Exposure to
Whole-Body Vibration" [4], published by the International Organization for
Standardization, provides an excellent base for setting these levels on
ships. The standard permits vibration levels to be rated numerically as
percentages of the established standard of fatigue-decreased proficiency.
The standard is related to frequency, direction of motion, and the expo-
sure time at the different locations in the ship. Safe exposure limits
and reduced comfort limits are defined in terms of percentages of the
fatigue-decreased proficiency level. This ISO Standard also has been
adopted by the American National Standards Institute, and it appears the
Standard can be used to establish rational vibration limits for human ex-
posure.
Guidance for acceptable vibration of marine steam and heavy-duty gas-

turbine main and auxiliary machinery plants has been published recently
by The Society of Naval Architects and Marine Engineers (SNAME) [5]. It
was prepared by Panel M-20 (Machinery Vibrations) of the Ships' Machinery
Committee and was intended to serve as a reference standard in ship's speci-
fications and procurement documents for new marine equipment. This Code
C-5 presents in detail the vibration limits of the machinery plants as
well as those for the longitudinal, lateral, and torsional response of
the propulsion system. It also specifies what type of tests are to be
conducted and the instrumentation required to measure the vibration lev-
els.
SNANE also has two additional codes dealing with vibration measure-
ments. The first, Code C-1 [6], is concerned with (1) vibration of the
ship girder excited by the propulsion system at shaft frequency, harmon-
ics of the propeller-blade frequency, and frequencies associated with
major components of machinery; and (2) vibration caused by propeller ex-
citation of the propulsion shaft system. The second, Code C-4 [71, ad-
dresses local vibration of ship structural elements such as the deckhouse,
decks, bulkheads, masts, machines, foundations, or other appurtenant ele-
ments of interest. Both of these Codes can be referenced in the ship
specifications as to the manner in which vibration measurements will be
made.
The final item in the specification concerns structural strength.

Since the ship's structure is least affected by propeller-induced vibra-
tion, little attention has been received in this area. The specification
should state that the stresses in structural locations subjected to high
loadings should not exceed the fatigue stress endurance limit of the ma-
terial with an appropriate factor of safety. Typical critical areas were
mentioned earlier in this section; these include the bottom framing over
the propeller, the stern bearing support, and masts and spars. Stresses
in these locations could be measured with strain gages during the accep-
tance trials.
2. Establish General Ship Design Data
In any design process there must be a starting point at which basic

information is assumed to be known. For the complete ship design, noth-
ing more specific than the mission of the vessel would be given. This is
too early to seriously consider the problems associated with propeller-
induced vibrations. More information about the ship's size, configura-
tion, and operating requirements must first be established in the feasi-
bility studies.
The procedures presented and discussed in this report assume that

certain general ship design data are available. The amount of mf orma-
tion required is the minimum necessary to begin the design procedure. As
additional ship data become known from other segments of the total design
process, they will be used in the vibration study. One should also real-
ize that this information is preliminary and may be altered if the design
procedures show changes are necessary.
This study wíll assume that the preliminary design data necessary
for approval of the basic design by the Maritime Administration are avail-
able. Such data would include:
Preliminary Lines Plan
Preliminary Midships Section Based on ABS Rules
Prelimínary General Arrangements of Decks and Inboard

and Outboard Profiles
Preliminary Weight and Center of Gravity Estimates
Speed and Power Estimates (No Model Tests)
Preliminary Machinery Arrangements
Preliminary Capacity Plan
Preliminary Hydrostatic Curves
li
Preliminary Flooding Curves and Damaged Stability
Calculations
Preliminary Specifications Indicating Propeller RPM

and Diameter
This report will discuss those recommended design procedures which

can be used after the preliminary design has been completed. This is not
to imply that the preliminary design should not consider the problem of
propeller-induced vibrations. It is simply more difficult to quantita-
tively establish the vibration characteristics of a given ship because
all the necessary elements are not yet defined. Instead, the preliminary
design must rely heavily on the experience of the naval architects
and existing rules from the classification societies. Insight as to
whether a ship will develop vibration problems can often be inferred by
the excitation and response levels on ships having similar stern lines,
propeller RPM and power, machinery arrangement, and general structural
configuration.
Table 1 shows the data which are necessary to begin the design
procedure presented in Figure 3. As the entire ship's design progresses,
these data will be supplemented by additional information when it becomes
known.
TABLE 1. SUMMARY OF DESIGN BLOCK 2--

ESTABLISH GENERAL SHIP DESIGN DATA
Input Data To Develop Required for
Hydrodynamic Test Wake Survey

Model Definition
Ship Power
and Speed Estimate of Propeller Propeller Design
and Shaft Sizes and Shaft RPM
Evaluation of Thrust Longitudinal and

Bearing and Location Lateral Analysis
Preliminary of Bearings of Shafting
Scantling
and Shaft- Formulation of Struc- Entire Ship
ing Plans tural Model Mass and Vibration
Stiffness Analysis
Hydrodynamic Test Wake Survey

Model Definition
Ship Lines
Hydrodynamic Test Cavitation Tests
Model Definition Behind Ship Model
3. Conduct Wake Survey
After the general ship design data, including the ship's

lines, have
been established, the next task is the conduct of model tests.
These tests
are basically used to confirm data which were predicted in the
preliminary
hydrodynamic design. However, as seen from Figure 3,
the model tests are
continued in time until all hydrodynamic work is completed.
The latter
model tests are not only used to confirm predictions made in the
final hydro-
dynamic design phase, but also to obtain data not available through
current
analysis techniques. This is generally in the area of propeller cavitation
and cavitation-generated pressures.
Some of the data which can be obtained from a complete set of

model
tests include:
Resistance or EHP versus speed, with and without appendages

(usually done)
Sinkage and trim of the hull (usually done)
Wave profile and flow pattern around hull (sometimes done)
Shaft horsepower and RPM versus speed for fully appended hull.
Determined in early tests with stock propeller, then with
propeller designed for the hull (usually done).
Wavegoing performance of the hull (sometimes done)
Shallow-water and restricted-channel behavior (rarely done)
Dynamic stability, maneuvering characteristics, and controlla-

bility when backing (often done)
Wake vectors without appendages to serve as a guide for proper

appendage location and orientation (often done)
Wake vectors in way of propeller disc with appendages

located
(often done)
Open-water and cavitation data on propellers designed

especially for hull (usually done)
Nature and magnitude of the propeller vibratory forces

imposed on hull (rarely done)
These items cover the entire area of model testing, and as indicated,
not
all tests are conducted for every shíp. Items i, j, and k relate directly
to the propeller-excited ship vibration problem, and each will be discussed
in the appropriate design block.
As far as the wake survey is concerned, its primary

vide data necessary for the propeller design, the purpose is to pro-
computation of propeller
and hull forces, and an evaluation of the extent. of
propeller cavitation.
13
The wake, when determined in absence of the propeller, ís called the nominal
wake field. Van Oossanen [8] points out that it is becoming common practice
to no longer accept the measured wake behind a model in a towing basin as
representative of the full-scale wake field. Differences arise because this
nominal wake does not consider the effects of the propeller on the true wake
and because Reynolds number scaling is not included. In the last few years,
there have been attempts to include these effects by numerical calculations.
Hoekstra [9] at the Netherlands Ship Model Basin has developed a procedure
to account for both Reynolds number scaling and the effect of the propeller
on the nominal wake field.
Refeence 10 presents the results of a recent British Ship Research

Association project on propeller-excited vibrations in which methods of
wake quality assessment are discussed. The authors point out quite cor-
rectly that the main cause of unsteady cavitation and large propeller
bearing forces is the non-uniformity of flow into the propeller. It is
therefore of the utmost importance that this wake be measured and evalu-
ated accurately.
Until recently, only the axial component of the wake velocity was
measured. This was partly because the available experimental techniques
could only measure one component at a time and because existing hydrody-
namic computational techniques did not include tangential and radial veloci-
ties. However, at the present time, all three components can be measured
with a five-hole pitot tube. This has stimulated additional research into
the ways the information concerning the three-dimensional velocity field
can be used in analytical prediction techniques. Other experimental tech-
niques which are used to obtain the wake are the hot-wire anemometer and
the laser-Doppler anomometer. These are discussed briefly in a paper by
van Gent and van Oossanen [11].
4. Estimate Longitudinal Propulsion Frequencies
In general, to keep propeller and hull excitation forces low, it is

desirable to use many blades on the propeller. The number of blades chosen
is set primarily by the natural frequency of the shafting and propeller in
longitudinal vibration. To ascertaín the probable frequency that will be
found after the design of the propulsion system and its supports are de-
veloped, it is useful to have a plot of natural frequency versus foundation
stiffness such as shown in Figure 4. Using values of the probable range of
foundation and thrust bearing stiffness, the probable range of shaft longi-
tudinal frequency is determined. The number of blades for the propeller is
chosen so that, preferably, the excitation frequency is less than 80 percent
of a possible propulsion natural frequency. A less desirable, but sometimes
necessary, solution is to locate the longitudinal natural frequency f1 about
30 percent below the blade frequency of the minimum steady operating speed.
Figure 4 shows that, if the foundation stiffness K is estimated to be be-
tween 10 x 106 and 20 x iO6 lb/in, then a four-bladed propeller will satis-
fy the criterion over the entire stiffness range. The five-bladed propel-
lers natural frequencies all are above the excitation frequency but if
10b lb/in,
the actual foundation stiffness turns out to be close to lO x
undesirable vibrations could develop. A six-bladed propeller would not be
acceptable because the shaft would pass through resonance for the lower
14
¿'600
>--
C)>
w w
cr
w= 500
u-
z DESIGN RPM=
90rpm
-J
<
o: w
D L
zw2z
0< cx
I- - 300
- z-
200 I i
0 10 20
THRUST BEARING FOUNDATION STIFFNESS, K( lb/in x 106)
NAIN SHAFT SPEED (rpm)

K (lb/in ) f1 (VPM) 4 BLADES 5 BLADES 6 BLADES
10 x 106 494 124 99 82
15 x io6 542 136 108 90
20 x 106 571 143 114 95
FIGURE 4. FIRST MODE LONGITUDINAL NATURAL FREQUENCY

VERSUS THRUST BEARING FOUNDATION STIFFNESS
15
foundation stiffness and would lie dangerously close to resonance for the
remaining values.
Estimates for the thrust bearing foundation stiffness can be found

the
in the recent SNAME Technical and Research Report R-15 [12] and in
earlier work by Kane and McGoldrick [13].
For making these predictions, the power and machinery arrangements

of the plant must be defined. These, along with the propeller RPM or
diameter, will have been specified by the preliminary design data. From
this, the approximate propeller weight and water inertia associated with
longitudinal vibration can be established by the techniques given by Mott
of the
[14] and Lewis and Auslaender [15]. Also, the approximate diameter
established by rules of the various ship
tailshaft and lineshaft can be
classification societies.
is on
The simplest procedure for predicting the natural frequencies
system consisting of the propeller
the basis of a one-degree-of-freedom
the
and water inertia plus a portion of the shaft mass vibrating against
stiffness of the thrust bearing and its foundation. Since the shafting
weighs considerably more than the propeller and adds flexibility, this
procedure is not very good.
An improved procedure is to model the propeller and shaft as a series

for
of concentrated masses and elastic elements and use a Holzer process
it becomes de-
frequency computation. With this degree of complication,
These
sirable to use one of the many digital computer programs available.
finite-element or finite-difference methods,
programs are usually based on
With the high
and several of the programs are discussed in Reference 16.
degree of sophistication and accuracy found in commercially available struc-
of a particular code is governed by its
tural analysis programs, the choice
convenience and cost to the user.
Table 2 presents a summary of the purpose, the input and output

in-
for this design block. Its format is

formation, and pertinent references
intended to
typical of the tables for the remaining design steps and is
of the indivídual procedure. The
provide the reader with a concise summary
by no means exhaustive, but are partic-
references listed in the tables are
exhaustive
ularly useful for design purposes. Reference i provides a more
list.
5. Design Propeller
After the number of blades has been selected based on the results of
is to estab-
the longitudinal propulsion frequency analysis, the next step
The primary purpose in this step is to select
lish the propeller design.
propeller
the propeller geometry which will provide the ship the highest
The design of the pro-
efficiency for the specified operating conditions.
problems such as blade strength and
peller must also consider ancillary
deformation, as well as selection of propeller materials and coatings to
resist corrosion and erosion.
16
TABLE 2. SUMMARY OF DESIGN BLOCK 4--ESTIMATE
LONGITUDINAL PROPULSION FREQUENCIES
Purpose: To Establish the Number of Propeller Blades so

that the Blade Rate Frequencies are Removed from
Longitudinal Shafting Natural Frequencies
Input: Propeller RPM

Horsepower
Machinery Arrangements
Shafting Diameter
Approximate Propeller Diameter
Estimates of Propeller Weight and Water Inertia
Range of Thrust Bearing and Foundation Stiffness
Output: Recommended Number of Propeller Blades
References: 12 - 15
Propeller design is a highly specialized field, and selection is

usually based on the recommendations of a consultant or a company active
in propeller design. For these reasons, this report will not attempt to
make other than general cotiutients as to the propeller design process. Ref-
erence 17 provides excellent background information, while References 18-
20 discuss a few of the current techniques used for propeller design.
The propeller design primarily influences the ship's vibration levels

through the number of blades. It was for this reason that the longitudinal
shafting frequency analysis was conducted in Block 4. Generally, there is
little change in efficiency between, for example, a three-, four-, five-,
and six-bladed propeller, and the final selection may be based upon vibra-
tion considerations. As a general rule, increasing the number of
blades
usually lowers the excitation forces on the shafting and the fluid pressures
transmitted to the ship's hull. There is a tradeoff, however, because in-
creasing the number of blades also increases the possible number of reso-
nances with the hull and the propulsion system.
The amount of propeller skew also influences the vibration levels

experienced in the ship. Generally, as the skew back of the propeller
increases, the bearing forces as well as the surface pressures decrease.
The axial vibratory forces and torques generated by the propeller decrease
rapidly, and the vertical and lateral forces and moments generally, but
not necessarily, decrease. Surface pressures also decrease, sometimes
quite significantly. The decrease in surface forces comes from both the
contribution of cavitating and noncavítating propeller pressures. Refer-
ence 21 presents theoretical and experimental data which show the advantages
17
and disadvantages of highly skewed propellers as compared with conventional
propeller design. The design procedure and model evaluation techniques
used by the Naval Ship Research and Development Center for a highly skewed
propeller for a cargo ship are discussed in Reference 22.
Design of the propeller is really more a part of the ship design pro-
cess than the ship vibration analysis, although there is a small input from
the hull vibration process.
Table 3 shown below summarizes the data required for the propeller
design process.

DESIGN PROPELLER
Purpose: To Design a Propeller for the Given Ship Which Will

Produce the Highest Efficiency
Input: Power Requirements

Ship Speed
Propeller RPM
Wake Data
Propeller Diameter Limitation
Estimate of Propeller Skew
Number of Blades
Output: Geometric Form of Propeller
References: 17 - 22
6. Compute Propeller Forces
Prior to about 1960, the determination of propeller forces was by

measurements on models, primarily by Frank N. Lewis [23]. In the late
1950's estimates began to be made on a quasi-steady-state basis using the
procedures developed by Burrill [24] for evaluating the loading and effic-
the
iency of propellers whose circumferentially averaged wake varied along
propeller radiï. A computer program for calculating the harmonic forces
and moments generated by the propeller working in varying wakes based upon
this quasi-steady-state procedure was applied by Hinterthon [25]. A simi-
lar computer program, also based on Burrill's procedure, but including as
well the Theodorsen effects (i.e., the inertia of the fluid in responding
de-
to circulation changes resulting from changes in angle of attack),was
of harmonic
veloped by CONESCO [26]. The first tends to give high values
forces and moments and errors in their phase because the inertia effects
values of harmonic
are neglected. The latter program tends to give low
the
forces and moments because the steady-state solution assumes flow over
18
tip and interaction between blades that are not developed in the unsteady
flow. Both of these programs have been superseded by improved analysis of
the problem.
In l958.Ritger and Breslin developed a theory for the unsteady thrust

and torque of a propeller in a ship wake based upon unsteady airfoil theory.
This work has been continued by Tsakonas and Jacobs [27] and is now a fully
developed program for predicting the harmonic forces and moments exerted
by a propeller on its supporting shaft, when working in the wake behind a
ship. This program is based upon lifting surface theory. Although the com-
putations are long, they are easily handled by a computer. A description
of the program is given in Appendix A-1. This program is widely used both
in the United States and abroad.
The Department of Naval Architecture and Marine Engineering at the

Massachusetts Institute of Technology has also been active in the predic-
tion of the harmonic forces and moments transmitted by a propeller to its
supporting shaft. Using unsteady flow theory with the propeller blade
represented as a lifting line, Neal A. Brown developed relations for de-
termining the periodic propeller forces [28]. Several computer programs
based on this theory have been developed. They are presented in Appendices
A-2 and A-3.
More recently, Kerwin and Frydelund [29] have approached the unsteady
force problem with another procedure. It is a discrete element approach
for the computation of unsteady blade pressure distribution in the absence
of cavitation. The work is still ongoing, and plans
are to extend the pro-
cedure to include the effects of unsteady cavitation. A discussion of
Kerwin's computer program is presented in Appendix A-4.
Similar procedures to those developed in the United States have been

developed in Europe. M. T. Murray and J. E. Tubby [30] at the Admiralty
Research Laboratory developed a computer program for determining the
un-
steady shaft forces from propellers. Information on this is presented in
Appendix A-5.
Table 4 shows the basic information required to determine the hydro-

dynamic forces and moments acting on the propeller. Some of the input
data may vary slightly, depending on the particular computer program used.
For more detailed information, the reader is referred to the listed
ref-
erences. It should also be noted that all input information is available
at this point in the design stage, either from the ship's operating con-
dition, propeller design, or wake survey.
The results of the numerical computations provide the mean and har-
monics of the blade frequency forces and moments, usually in the longi-
tudinal, vertical, and lateral directions. These forces and moments can
be applied to a structural model of the ship to determine its forced
re-
sponse to propeller excitation. This procedure will be discussed
in the
ship substructure and complete ship structure design phases.
It would be very advantageous, at this point in the ship's design,

to be able to estimate the amplitudes of stern vibration based on the
19
TABLE 4 . SUI4NARY OF DESIGN BLOCK 6--
COMPUTE PROPELLER FORCES
Purpose: To Predict the Hydrodynamic Forces and Moments Act-

ing on the Propeller
Input: Propeller Blade Geometry

Propeller Geometry
Ship Speed
Results of Wake Survey (Usually Given in Terms of
the Fourier Coefficients of the Spatial
Variation of the Axial and Tangential Com-
ponents of Wake)
Output: Mean and Blade Frequency Force and Moment Components

(This Information is Usually Given for Three
Orthogonal Forces and Moments)
The Time Varying Blade Pressure Distribution at
Each Wake Harmonic
References: 27 - 32
computed propeller forces. This would allow the designer to predict if

excessive vibration levels are probable and to take corrective action be-
fore the design proceeds any farther. Unfortunately, no such generai es-
timation technique exists, but McGoldrick [31] does give empirical formulas
The
for the vertical displacement and torsional rotation of the stern.
above the
formulas are applicable only when the blade frequencies fall well
NcGoldrick gives the formula
range of significant hull mode frequencies.
for vertical vibration as
P
Y= o
3.4 x 10_6 x A x (cpm)2
where
Y = the single amplitude in mils (a mil is equal to 0.001 in.)
P = the single amplitude of the vertical component of

° propeller-exciting force in pounds (at blade frequency)
A = the displacement of the ship in long tons
cpm = the blade frequency in cycles per minute
3.4 x 10-6 which was

The empirical constant in the formulas is the factor
obtained by shaker tests conducted on the SS Gopher Mariner. McGoldrick
indicates that there is some reason to expect the empirical constants
20
could be used for different ships, but that much more experimental data are
needed to establish the constant for various classes of ships. To date,
this has not been done. It should again be emphasized that procedures for
estimating the stern vibration levels based only on the propeller forces
and gross ship properties are needed at this point in the design phase.
Development of such techniques is certainly worthy of further investigation.
7. Compute Hull Pressures Without Cavitation
It is the purpose of Design Blocks 7-10 to compute the excitation

pressures on the ship's hull, including the effects of cavitation. The
most direct method would involve calculating the type and extent of cavi-
tation on the propeller operating in a given wake and then computing the
fluid pressures generated on the hull. In the United States, no such direct
procedure is available, although research is being conducted in this area.
Van Oossanen reports [33] that procedures for calculating these hull cavi-
tation pressures are available at the Netherlands Ship Model Basin and are
used for design. They will be discussed in Block 11.
In this country the procedure is roundabout and consists of predict-

ing the hull forces generated by the propeller in the absence of cavita-
tion and modifying the results by an empirical factor to account for cavi-
tation effects. The factor is determined by (1) experience with full-scale
measurements, (2) model tests in a cavitation towing tank, and (3) model
tests in a cavitation tunnel of sufficient size to include modeling a por-
tion of the ship.
In a sense, the use of a factor applied to the pressures determined

in the noncavitating condition is theoretically unsound because the pres-
sures are generated by another mechanism other than that responsible for
the pressures generated in the noncavitating case. The hull pressures are
generated as the sum of three different processes. The first is the pres-
sure due to propeller loads, i.e., the difference in pressure on the face
and back of the blades. The second source of pressure generation is the
passage of the propeller blade bulk through the water. Generally the
pressures from these two sources are approximately equal in amplitude, but
can be quite different in phase. The third source, cavitation, is the
growth and decay of cavitation bubbles as the blade moves into high wake
regions. Since the growth and decay of a volume radiate pressure much
more effectively than moving a volume from one place to another or intro-
ducing a flow from a source to a sink, the pressures from small cavitation
volume changes can be large.
The determination of propeller-generated hull forces can be made by

two processes: (1) estimation of the hull pressure and (2) an integration
process involving Green's function which yields the total excitation force.
Either process involves many engineering approximations for a reasonable
solution. Generally the hull pressure process involves determining the
pressure that would be generated by the loading and thickness of the pro-
peller in a free field and multiplying this pressure by a factor to repre-
sent thepressure of the hull. This so-called boundary factor is 2 for an
infinite flat plate located adjacent to a working propeller because the
restraint on the flow caused by the plate can be simulated by an "image"
21
The free-
propeller working symmetrically opposite to the real propeller.
from both propellers, which is twice that
field pressure is that induced
of one. The value of 2 is normally used
with reasonable accuracy for
points on a ship's hull. References 34 and 35 indicate, however, that com-
parison between the calculated and measured hull pressures showed somewhat
large variations from the value of 2 for cavitating and noncavitating pro-
pellers. This process is entirely inadequate for estimating differential
The Green's
pressures across narrow surfaces such as skegs or rudders.
surface
function process requires an estimate of the added mass of the hull
and moment
for motions corresponding to each of the components of force
that are required [36]. Theoretical processes for predicting pressure dif-
but
ferences across wedge and cone shaped surfaces [37,38] are available,
Smith-Hesse procedure,
not yet programmed. A theoretical approach, the
is so
for predicting the hull pressure is available, but the calculation
long that it has not at the present time been programmed.
the
In Reference 39 some of the methods available for calculating
discussed and com-
pressure field around a propeller in a free stream are
The results are
pared to cases where experimental results are available.
the
only valid for the noncavitating propeller, and Reference 39 indicates
results are changed appreciably when extensive cavitation is present.
of
Table 5 presents the data generally required for the computation
depend-
the hull forces or pressures. Some of the information will vary,
Procedures for predicting
ing on the particular computer program used.
presented in
hull pressures by lifting line and lifting surface theory are
Appendices B-1 and B-2, respectively. Appendix B-3 presents the procedure
TABLE 5. SUNMARY OF DESIGN BLOCK 7--COMPUTE

HULL PRESSURES WITHOUT CAVITATION
Purpose: To Compute the Excitation Levels without Cavitation

Acting on the Hull at the Ship's Stern
Input: Propeller Blade Geometry

Wake Distribution
The Spatial Location of Points on the Stern where
Pressures are Desired
The Steady and Time-Dependent Blade Loading
Distributions
Output: Steady and Harmonic Components of the Pressure Field

(For the
Generated by a Noncavitating Propeller.
Green's function method, all components of the total
hull forces and moments at multiples of the propeller
blade rates can be computed)
References: 27, 32, 36, 40, 41, 42
22
for computing the total excitation force on the ship's stern by using the
Green's function approach.
Evaluate Propeller Cavitation
The evaluation of cavitation in this section refers to analytical

predictions of the type and extent of cavitation on the faces of the pro-
peller blades. Experimental techniques are considered in Section 12. The
purpose of these calculations is to determine how severe the cavitation
problem is in a given wake field. They will enable the designer to assess
the risk of damage to the propeller from erosion and bent trailing edges
and to estimate the magnification of propeller-generated hull pressure over
the noncavitating case. This report is concerned with the second of these
goals, and the next section discusses techniques for estimating these pres-
sure factors.
Most of the recent research on the theoretical prediction of cavi-

tation on propellers has been done in Europe, principally at the Netherlands
Ship Model Basin and at the Swedish State Shipbuilding Experimental Tank.
Van Oossanen [33] has reported results in late 1977 which show very good
agreement between calculated and observed full scale as well as calculated
and observed in the cavitation tunnel for the extent of cavitation. The
computer program used at NSMB for these predictions is further discussed
in Section 11.
Other predictions have been made by Johnsson [43] at the Swedish

State Experimental Tank, but they do not seem to agree as well with the
observed cavitation or with van Oossanen's predictions.
Evaluate Propeller Cavitation Factors
After the extent of propeller cavitation has been determined from

Design Block 8, the next step is to estimate the amount cavitation will
increase the pressures on the hull. Research into the problem of propeller-
induced forces has been ongoing for several decades, but it was only recog-
nized in the past ten years that transient cavitation influences the hull
pressures very strongly. In fact, Lewis and Kerwin stated in a recent
paper [44]:
While extensive work in noncavitating flows was not

entirely a wasted efforts it would seem clear now that
both analytical and experimental prediction of vibra-
tory forces are completely unrealistic without inclu-
sion of the influence of cavitation. What remains to
be seen is whether or not design decisions based upon
minimizing noncavitating propeller vibratory excitation
are optimum when cavitation is present.
Reference 45 reported to the 12th International Towing Tank Confer-

ence that the amplitudes of fluctuating pressures increased around 100
percent when cavitation was present between the propeller tip and about
0.85 radius. Other investigations of the variations in these pressure
fluctuations are given in References 46-48, and the reported cavitation
23
factors range from 2 to 40. Clearly, with this large variation in the
pressures generated between noncavitating and cavitating propellers, it
is not a simple task to estimate a reasonable value of the factor.
These factors must be obtained from experimental tests, either full

scale or model. The most straightforward approach is to make pressure
measurements on a ship with similar lines, wake, and operating conditions.
However, for a new design, this approach may not be feasible. Instances
in which such measurements were made are described in References 49-50.
Model tests must either be conducted in a cavitation tunnel or a depres-
surized towing tank facility. Por tests in the cavitation tunnel, the
wake field is simulated either by screens and an afterbody model of the
ship, or if the tank is sufficiently large, a complete ship model. These
direct test methods will be described in more detail in Block 12.
The evaluation of the propeller cavitation factors proposed in this

section is much less precise in that estimates based on previous model
experiments are used as a basis. Reference 48 contains results of cavi-
tation tests conducted in the large cavitation tunnel of the Netherlands
Ship Model Basin for a number of Wageningen B-series model propellers.
Figure 5 shows the geometry of the stern and the location of the pressure
pick-ups on the ship's centerline over the propeller. To use Reference 48
for estimating the cavitation factors, the investigator would first need
to evaluate the extent of cavitation on the propeller and compare it with
figures given in the reference. The corresponding cavitation index 0n can
now be estimated, and knowing the ship's thrust coefficient, KT, the ampli-
tudes of the first and second harmonic component of the average pressure
fluctuations (P1 and P2) can be estimated from the tables given in the
reference. Figure 6 shows the extent of cavitation on the front and back
faces of a Wageningen BB 4-70 propeller at cavitation indices of 2.5 and
5.0 along with the nondimensional pressure fluctuation corresponding to a
thrust coeffícient of KT = 0.075. The dimensionless pressure variable
Kp12 is defined as
Kp12 = Kp1 sin (zwt-41) + sin (2zut-2)
where
wt angular blade position (wt = O for a blade in the

vertical, top position)
z = number of propeller blades
Kp1 = 212 (pressure coefficient for first harmonic)

pn D
(pressure coefficient for second harmonic)

= 2 2
pn D
= amplitude of first (second) component of averaged

pressure fluctuation
24
¶010 UDO
A4L SIZES IN mm
900
FIGURE 5. STERN GEOMETRY FOR CAVITATION TUNNEL TESTS,

FROM [481
BACK SIDE CAVITATION

DIMENSIONLESS PRESSURE FLUCTUATIONS FOR KT = 0.075
First Harmonic Components (Frequency Equal to Blade Frequency)
place pressure
pick-up Kp1
an - 5 forward 0.0184 121

middle 0.0198 141
(N 1200 RPM) aft 0.0135 180
- 2.5 forward 0.0462 172

middle 0.0491 173
(N - 1200 RPM) aft 0.0395 182
Second Harmonic Components (Frequency Equal to Twice Blade Frequency)
FACE SIDE CAVITATION Xp2
C - 5 forward 0.0023 133

middle 0.0033 135
(N - 1200 RPM) aft 0.0025 150
- 2 forward 0.0085 33
middle 0.0067 12
(N - 1200 RPM) aft 0.0060 2
FIGURE 6. CAVITATION PROPERTIES OF MODEL

SERIES PROPELLER AT = 0.075,
a5 50 a5 = 25 FROM [48]
25
p = density of water
n number of propeller revolutions per second
= phase angles of first (second) harmonic
D = propeller diameter
The cavitation index 0n and thrust coefficient KT are defined as
P -P
o V
=
-
1 22
pn D
T
KT - 24
pn D
where
Po static pressure at the centerline of the propeller shaft

at the propeller
= vapor pressure
T = propeller thrust
The success of this procedure clearly depends on how closely the

stern configuration, propeller geometry, and wake match the test condi-
tions. The method will not yield exact results, but may provide valuable
information as to the magnitudes of the cavitation pressure factors. The
results are also limited in that they provide information only at the
measured locations and not at other points on the ship's hull.
10. Direct Calculation of Cavitation Pressures and Forces
After calculation of the pressure forces and moments as discussed

in Section 6, the next task is to compute the propeller-generated pres-
sures on the hull. These pressures come from three sources as discussed
in Section 7. The first results from a fluid element being displaced
from one point to another, the second is due to an expanding or contract-
ing fluid volume caused by cavitation, and the third is the pressure due
to the propeller loads. As discussed in the previous section, the cavi-
tation-induced pressures can increase the total hull pressure by several
factors and play a major role in producing propeller-induced vibrations.
Cavitation also reduces the propeller's service life in the form of erosion
and bent trailing edges. Consequently, accurate prediction of cavitation
would be a major step forward in improving ship design.
In recent years, with the increase in ship size, higher speeds and
power, the problem of computing cavitation directly has received increased
attention. Most of the work is being done in Europe, principally by van
Oossanen at the Netherlands Ship Model Basin. Reference 33 provides an
26
excellent state-of-the-art review in cavitation prediction techniques.
From work recently published by van Oossanen, it appears that his tech-
nique gives good correlation with observed cavitation patterns for lightly
and moderately loaded propellers. The predicted results are not as satis-
factory for a heavily loaded propeller in a very non-uniform wake. Van
Oossanen attributes this to lack of knowledge of the change in wake flow
due to the working propeller.
Reference 8 indicates that the Netherlands Ship Model Basin has a

computer program, designated as CAVANAL, for the computation of cavita-
tion on propeller blades. The input data required for the program are
listed in Table 6. Reference 33 gives comparisons of the extent of cavi-
tation as predicted by theory and observed in model tests.
TABLE 6. STJMMARY OF DESIGN BLOCK 10--DIRECT CALCULATION

OF CAVITATION PRESSURES AND FORCES
Purpose: To Determine the Extent of Cavitation and Pressure

Distribution on a Propeller
Input: Propeller Geometry Parameters

Propeller Diameter
Hub Diameter
Number of Blades
Expanded Blade Area Ratio
Radial Pitch Distribution
Radial Distribution of:
Distance of Leading Edge to Generator Line
Distance of Trailing Edge to Generator Line
Radius of Trailing Edge
Maximum Camber-Chord Length Ratio
Maximum Thickness/Chord-Length Ratio
Location of Maximum Camber with Respect to
Leading Edge
Angle Between Nose-Tail Line and Pitch Line
Wake Field at 25 Angular Coordinate Values for
Five or More Radii
Direction of Rotation
Propeller RPM
Ship Speed
Static Pressure at Centerline of Propeller
Fluid Minimum Vapor Pressure
Water Temperature
Fluid Density
Output: Pressure Distribution on Propeller Blade

Cavitation Index
Extent of Cavitation on Propeller Blades
References: 8, 33, 51
27
Recently, the program CAVANAL has been coupled with another NSMB
program which calculates the propeller-induced pressure field on the
hull. This last program, called HTJFO, is based on the theory developed
by Noordzij and described inì Reference 51. Van Oossanen reports [811 that
since CAVANAL and HUFO have been joined, a relatively large number of
successful calculations of hull surface pressures have been carried out.
These were performed to optimize propulsion configurations for a given
hull, given operational parameters, and a given wake.
Conduct Model Tests
The model tests discussed in this section refer to those associated

with noncavitating propeller-generated forces and pressures. Experimental
tests used to study the effects of cavitation will be discussed in the
next section. Tests related to other portions of the ship's design were
listed in Section 3, where the wake survey was first discussed. The pur-
pose of conducting these model tests is to measure the propeller, hull,
and rudder forces for comparison with the values calculated from the ana-
lytical techniques. A favorable comparison will confirm the wake survey
and give assurance that the excitation levels actually experíenced by the
ship will be close to those predicted. If the comparison is not within
acceptable limits, then the designer must determine what is causing the
discrepancies. Possible sources of errors are in the conduct of the wake
survey, the determination of the Fourier component of the wake as required
by the programs used to compute propeller forces and pressures, inaccu-
racies in these programs or associated data, or errors in the conduct of
model tests. Whatever the source of error, it is imperative that the
differences between analytical and experimental predictions be resolved.
There is little sense in making response predictions for the entire ship
and its subsystems if one does not have confidence in the accuracy of the
applied loadings.
The expense of these tests should only be a small portion of the

entire budget allocated to the vibration analysis. In fact, the same
model which was made for the wake survey can be used. Only a scale model
of the final propeller design need be constructed.
The method for conducting these tests was developed by F. M. Lewis

in 1936 [23] for the measurement of vertical and lateral forces plus a
longitudinal couple. More current techniques for conducting these tests
are described in References 52-53. It should be noted that the model
tests for measuring propeller forces require propeller dynamometers which
are located aft of the propeller on a separate measuring device. The
purpose of this arrangement is, of course, to measure directly the forces
and moments exerted by the propeller at its connection with the line
shafting. David Taylor Model Basin, Massachusetts Institute of Tech-
nology, and the Netherlands Ship Model Basin are three test facilities
which have propeller dynamometers.
Conduct Cavitation Tests
This report has already emphasized the ímportance of propeller cavi-

tation on the level of vibratory pressures imposed on the ship's hull. It
28
also has indicated that the theoretical prediction techniques of cavita-
tion and its effect are still being developed and are not available
throughout the world. For these reasons, various research organizations
have established experimental facilities to study this problem and assist
the shipbuilder to design a ship in which detrimental propeller cavita-
tion effects are minimized. Table 7 lists some of the cavitation test
facilities throughout the world.
These tests can be conducted by either one of two experimental tech-

niques: the cavitation tunnel or a facility capable of modeling the free
surface effects.
In a Cavitation tunnel of limited size, the wake is established by

screens or a combination of screens and a model of the ship's afterbody.
For larger tunnels such as at the Swedish State Experimental Tank, a com-
plete ship model can be used to establish the wake. References 44, 49, and
54 describe the testing techniques and results of several investigations
conducted in cavitation tunnels.
For tests conducted in a variable pressure towing tank, the dynamic

and kinematic similarities between the ship model and prototype are matched
as closely as possible. The usual procedure is then to maintain equal
cavitation and Froude numbers for the model and prototype. The largest
depressurized towing tank is operated by the Netherlands Ship Model Basin
in Ede, The Netherlands. A description of this facility along with some
recent experimental results is presented in Reference 55.
References 56 and 57 describe experimental programs for cavitation

studies in both tunnels and variable pressure towing tanks for ducted
propellers on large ships with deadweight tonnage over 200,000-dwt. Ref-
erence 58 shows that good agreement was obtained when results obtained
from the large cavitation tunnel at the Swedish State Experimental Tank
were compared with those from the Netherlands Ship Model Basin for a tanker.
13. Compute Total Pressures and Forces
The objective of the previous hydrodynamic investigations was to de-

termine the magnitudes and frquencies of the individual excitations act-
ing on the ship. As indicated in Figure 1, excitations arise from the
propeller operating in a given wake field and are transmitted to the ship
by three mechanisms: forces and moments exerted by the propeller on the
line shafting; pressures, both noncavitating and cavitating, transmitted
to the ship hull; and excitations introduced into the ship through hydro-
dynamic loading of the rudder in the vicinity of the working propeller.
All of these have been discussed in the previous sections. The purpose
of this design block is to assemble the results of the earlier work and
compute the total components of hydrodynamic excitation acting on the ship.
These components will then be used as input to determine the forced response
for each of the ship's structural subsystems and for the complete ship.
The response calculations will be discussed in the next two design phases.
Figure 7 shows a flow diagram for the computation of the total pro-
peller-induced pressures and forces acting on the ship. The reader can
29
*
TABLE 7. SOME CAVITATION TEST FACILITIES
Location Type of Facility

Organization
Cambridge, Massachusetts, Variable Pressure

Massachusetts Institute
USA Water Tunnel
of Technology
David W. Taylor Ship Carderock, Maryland, USA 2 Variable Pressure

Research and Development Water Tunnels
Center
Netherlands Ship Model Ede, The Netherlands Variable Pressure

Towing Tank
Basin
Ship Research Institute Trondheim, Norway Cavitation Tunnels

(2)
of Norway
G5teborg, Sweden Cavitation Tunnels

Swedish State Ship-
(2)
building, Experimental
Tank
West Berlin, Federal Free Surface

VWS
Republic of Germany Tunnel
This list is riot necessarily complete.

Block 6 Block 11
COMPUTE CONDUCT Block 13

PROPELLER MODEL
COMPUTE TOTAL PRESSURES
FORCES TESTS
AND FORCES
COMPUTE TO MODELS OF
PROPELLER
FORCES SHIP SUBSTRUCTURES
Block 7 Block 8 Block 9
COMPUTE HULL E VALUATE EVALUATE
PROPELLER COMPUTE HULL
PRESSURES W/O PROPELLER
CAVITATION PRESSURES
CAVITATION CAViTATION
FACTORS AND FORCES TO STRUCTURAL MODEL
-
OF COMPLETE SHIP
Block
COMPUTE RUDDER
PRESSURES AND
I
CONDUCT FORCES
CAVITATION
TESTS
J
Block 10
DIRECT CALCULATION
OF CAVITATION
+ PRESSURES AND
FORCES
FIGURE 7. FLOW DIAGRAM FOR COMPUTATION OF TOTAL

PROPELLER-INDUCED PRESSURES AND FORCES
see that all design blocks in the final hydrodynamic phase (Blocks 6-12)
are used in the process. The calculation of the propeller forces and mo-
ments is fairly straightforward and was discussed in Section 6 along with
available methods. It was pointed out in Section il that model tests
should also be carried out to confirm the wake survey used in the calcu-
lation of these moments and forces. The results of the model tests could
supplant the numerical calculation, but this is not recommended. In gen-
eral, the calculation procedures are much less time consuming and costly
than the experimental tests and give a broader picture of the excitation.
Therefore, it is probably desirable to carry them through even if detailed
model tests are run.
The computation of the hull pressures and forces is more difficult

than that for the propeller forces and moments. It is complicated even
more by the influence of cavitation, whose importance was only quantita-
tively recognized within the last ten years. At the present time, the
usual way of evaluating the total hull pressures is to compute the pres-
sures without cavitation, estimate the effect of cavitation, and apply
appropriate factors to determine the total excitation. These procedures
were discussed in Sections 7-9, where it was indicated this process is
roundabout and not very satisfactory from a theoretical point of view. A
direct calculation of the cavitation pressures and forces is certainly
more desirable, and research is active in this area. Section 10 discusses
the direct calculation method used by the Netherlands Ship Model Basin.
The other method of obtaining the cavitating hull pressure is through

model tests either in a cavitation tunnel or depressurized towing tank.
There are several organizations throughout the world which have experimen-
tal facilities especially designed for conducting these tests. These were
discussed in Section 12 along with references to some recently published
results. As in the case of the propeller forces, the cavitation model
tests could be used in place of analytical predictions of the extent and
pressure generated by cavitatíon. The word of caution expressed earlier
in this section still applies. Model tests are expensive to conduct, and
theoretical work needs to be continued so that the phenomenon of cavita-
tion is fully understood and its effect on the Ship'S design can be ana-
lytically predicted.
14. Determine Forced Longitudinal Response of Shafting
By previous calculations, the longitudinal exciting force at the

propeller will have been determined. The purpose of this study is to
find whether the vibration level generated by this excitation will be
acceptable. The propeller and the length and diameter of the shafting
will be known, and the thrust bearing will probably have been selected.
The unknown quantity will be the stiffness of the thrust bearing founda-
tion. The amplítude of motion at the thrust bearing as a function of fre-
quency for different values of foundation stiffness is required.
The stiffness of the thrust bearing foundation must be determined

to assure that a shaft longitudinal vibration resonance does not fall in
the operating range. For preliminary analyses, the foundation static
stiffness can be determined by the use of several methods. The simplest
32
process is to represent the foundation and bottom as a combination of
frustums of wedges and beams. This procedure is described in Reference
12. A process requiring less engineering judgment is to use finite-ele-
ment methods, assuming that the machinery double bottom is supported at
its edges. It is also possible to represent the machinery space double
bottom as an anisotropic plate.
Generally, it will be found that the natural frequency of the bottom

structure will not be far removed from the propeller blade frequency. If
coincidence occurs, the propeller through longitudinal vibration of the
shafting will excite engine room vibration even though the natural fre-
quency of the shaft in longitudinal vibration determined from static stiff-
ness considerations appears to be suitable. This aspect is considered in
the following section.
As in Design Step 4, in which the longitudinal propulsion frequen-

cies were estimated, the maximum excitation frequency should be less than
about 80 percent of the longitudinal resonance frequency. If this is not
possible, then the excitation frequency at the lowest steady operating
speed should be about 30 percent above the critical longitudinal frequency.
Several types of computer programs are suitable for this analysis.

The system can be broken down to a sequence of masses connected by springs.
This can be analyzed by a Holzer Table program, the kind developed for
torsional vibration, or by a standard finite element program such as NSYS,
MARC, STARDYNE, NASTRAN, SESAM, etc. However, the shafting, whose distrib-
uted weight is several times that of the propeller with its associated wa-
ter inertia, consists of long lengths of constant diameter. This charac-
teristic is encouraging to a program that represents the shaft as distrib-
uted mass and elasticity, and a few computer programs have been developed
which utilize this property. In such a case the system can be defined with
a minimum of input variables, thus saving time, improving accuracy, and
reducing the probability of erroneous inputs.
Table 8 shows the information necessary for conducting this forced

vibration analysis. All of the input data will be available either as
part of the basic ship design parameters or generated during the design
process.
The following appendices taken from Reference 59 indicate that the

Maritime Administration has a program (Appendix C-1) based upon the Holzer
Method for determining longitudinal vibrations; that J. J. McMullen has a
program (Appendix C-2) for determining longitudinal vibration where the
shaft is modeled as lumped masses; and that Newport News has a program
(Appendix C-3) that can represent the shaft as a distributed mass system.
A Littleton Research and Engineering Corp. program utilizing lumped and
distributed masses and elasticities is described in Appendix C-4.
15. Determine Forced Response of Machinery Space
The shafting system is connected in longitudinal vibration to the

machinery space double bottom through the thrust bearing. Thus, vibra-
tions of the shaft will be coupled with those in the machinery space, and
33
TABLE 8 . SUMMARY OF DESIGN BLOCK 14--DETERMINE FORCED
LONGITUDINAL RESPONSE OF SHAFTING
Purpose: To Compute the Resonant Frequencies and Response

Amplitudes of the Longitudinal Shafting Subjected
to Forced Excitations
Input: Stiffness of Thrust Bearing Foundation

Thrust Bearing Stiffness
Reduction Gear Mass and Stiffness
Length and Diameter of Shafting
Nass and Added Nass of Propeller
Estimates of Damping
Longitudinal Excitation Levels
Output: Longitudinal Shafting Resonant Frequencies

Amplitude of Motion at Thrust Bearing
References: 12-15, 59
vibrations in the machinery space bottom structure can be strongly coupled

with longitudinal vibration of the shafting. Although it might be desir-
able to model the double bottom as an anisotropic plate with variable in-
ertias for the same reasons that the distributed mass-elasticity procedure
is used for the shafting, this type of model has not been developed, and
it is necessary to use finite-element modeling. Reference 60 discusses
the modeling techniques used in a recent static and dynamic finite-element
analysis of the hull structure of a large surface-effect ship. The aniso-
tropic behavior of the hull plating due to longitudinal T-shaped stiffeners
was incorporated into the NASTRAN model by modifying the material property
matrix.
It is desirable that the computer system that is used be compatible

with that used for the complete ship. If the final ship is to be modeled
by finite-element procedures, the same system should be used for the ma-
chinery space, which can then be incorporated in the full model as a sub-
structure. If the complete ship is to be modeled as a Timoshenko beam
with sprung masses, any convenient finite-element model can be used for
the machinery space.
Figure 8(a) shows a finite element of the afterbody with the ma-
chinery space shaded. The model of the longitudinal shafting and thrust
bearing stiffnesses will be available from the analysis conducted in Block
14. The question which must be decided is to what extent and in what de-
tail must the surrounding machinery space structure be modeled to repre-
sent adequately its mass and stiffness properties. The answer depends a
great deal on the experience of the engineer conducting the analysis. The
model should extend in the transverse direction over the ship's half width
(symmetry can be assumed), in the longitudinal direction to at least several
34
#I
#1
Q
'g
's.
(a) Machinery space is shaded
trilli
437
4!..
j
4$
-'Q
(b) Aft bearing structure is shaded
FIGURE 8. FINITE ELEMENT MODEL OF A SHIP'S

AFTERBODY, FROM [61]
35
transverse floors on each side of the thrust bearing, and preferably the
whole length of the machinery space, and in the vertical direction from
the bottom of the ship's hull to about two decks above the thrust bearing.
The important point to remember is that the model must be accurate enough
to predict the lower natural frequencies and mode shapes.
If doubts exist in choosing a proper model size, the analyst should

err on the conservative side; i.e., pick a model larger than necessary.
First, the computed resonant frequencies and response will be closer to
the actual values. Second, the additional work required to generate the
larger model is not lost because the model will be used as a substructure
in the complete ship vibration analysis.
One of the earliest finite-element investigations into the natural

frequencies of the machinery space was made by Reed [62]. Other investi-
gators, especially in Europe, are using finite-element techniques on a
routine basis for the structural analysis of all critical locations in the
ship. G. C. Volcy at Bureau Ventas has been particularly active in this
area, and many techniques and results obtained by that organization in its
study of propeller-induced vibrations are published together in Reference
63. For particular papers dealing with machinery space vibrations, the
reader should see References 64-66.
Figure 9(a) taken from Reference 67 shows a finite-element model of

the shaft and double bottom for a large tanker. Table 9 gives the natural
frequencies for the separated and integrated models and indicates the
double bottom affects the shaft's frequency more strongly than the shaft
affects the double bottom's frequency. Figure 9(b) clearly indicates a
shaft resonance at 12.11 Hertz, and the corresponding mode shape is given
in Figure 9(c). Reference 67 indicated that the static stiffness of the
thrust block was 0.190 x 106 Np/rn (one Np equals 1000 kilograms of force),
which resulted in a first mode longitudinal shaft frequency of 12.83 Hertz.
When the double bottom model was coupled with the shaft, the frequency
TABLE 9. NATURAL FREQUENCIES OF DOUBLE BOTTOM AND

SHAFT FOR SEPARATED AND INTEGRATED MODELS,
FRON [67]
Double Bottom
Double Bottom Shaft and Shaft
ist mode of
9.20 Hz 9.15 Hz
double bottom
1st mode of
12.83 Hz 12.11 Hz
shaft
2nd mode of
14.06 Hz 13.99 Hz
double bottom
36
(a) Double bottom and shaft model of 172,000 dwt tanker
w
L)
z
zo
(I)
w
A N
0.6-
N u--- I N
ILn
0.5-
oJQ) o
OJ
0.4- S KA F T EKO
ui< Io
-iz
0.3- o ow
oc
U)
E-i
0.2-
oI
TOP OK TNKU5T
OUKOATON
BLOCK
0.1 - CF TTKUSTB
O
8 9 10 11 12 13 14 15
w(Hz)
(b) Resonant response of the center girder of 172,000 dwt tanker,

w = 12.110 Hz, thrust force of 10 Mp
MEPLITIJDE iN mtfl
(c) Response of shaft and thrust block of 172,000 dwt tanker,

thrust force 10 Mp at w = 12.110 Hz
FIGURE 9. FINITE ELEMENT MODEL AND RESPONSE OF THE DOUBLE BOTTOM

AND SHAFT FOR A 172,000 DWT TANKER, FROM [67]
37
decreased to 12.11 Hertz, and Figure 10 shows that this coupling decreases
the effective stiffness of the thrust block foundation by approximately 13
percent. These results indicate that a separated shaft model can only be
used in a vibration analysis where the natural frequencies are not close
to double bottom resonances.
Table 10 shows the information pertinent to the conduct of this analy-

sis. The only input data in addition to that required for the forced re-
sponse of Section 14 are estimates of the machinery space mass distribution
and the ship's scantling plans.
16. Determine Forced Lateral Response of Shafting (Rigid Hull)
The shaft responds laterally to the harmonic force and moment excita-
tions about axes normal to the rotational axis. If the lateral natural
frequencies of the propeller and shaft system coincide with the blade fre-
quency excitation, the input to the hull through the bearings can be strongly
amplified. Calculations of ship response generally show peaks associated
with lateral frequencies of the shafting. It is, therefore, desirable to
design the shafting system so that these resonances will not occur at the
normal operating speeds. As with the longitudinal vibrations, these studies
are successively made on models of increasing complexity. The first studies
are applied to the shaft simply supported at the bearings (either at the
forward and after edges or one-third of the distance from the rear of the
stern bearing). Since it is known that the bearings are relatively f lexi-
ble, this model will generally give a frequency that is high so that if the
lowest lateral frequency is less than, say, 30 percent above the full power
blade frequency, it will probably be wise to consider relocating the bear-
ings or modifying the shafting to raise the frequency. Bureau Ventas in
its Guidance Note [68] published in 1971 recommends, for example, that the
natural frequency be above 130 percent of the excitation frequency.
It is also important to ensure that all vertical bearing reactions

be positive, I.e., maintain contact with its support, over the range of
excitation. A negative reaction will result in hammering of the bearings.
This hammering is certainly undesirable from the standpoint of maintenance
and noise levels. However, it also changes the system's natural frequency
and could move it closer to the propeller-excitation frequencies. Large
lateral vibratory motions of the shaft would result, and these vibrations
would probably be transmitted to other portions of the ship. References
69 and 70 present results of investigations conducted at Bureau Ventas on
lateral shafting vibrations and the associated wearing on the bearings
caused by misalignment of the propeller shaft.
To determine the forced vibration response, the analyst can use any
of the commercially available computer programs such as ANSYS, ASKA, NAS-
TRAN, SESAM, STARDYNE, and STRUDL. There is no optimum choice among these
programs because all are highly developed and give essentially the same
type of results. The choice depends on the availability of the program
and the familiarity of the user with it. A discussion of the theoretical
aspects of the finite-element method as it pertains to transverse vibra-
tions of a ship's propulsion system is given in Reference 71.
38
L
15
14 -
13
12.11
12
11 -
10 -
9- L)I ol
8
0.05 0.10 0.15 0.20 0.25 0.30
kTB( 106 Mp/m
FIGURE lo. DEPENDENCE OF THE FIRST SHAFT NATURAL FREQUENCY ON

THRUST BLOCK STIFFNESS, KTB, 172,000 DWT TANKER, FROM [67]
TABLE 10. SUMMARY OF DESIGN BLOCK 15--DETERMINE FORCED

RESPONSE OF MACHINERY SPACE
Purpose: To Compute the Resonant Frequencies and Response

Amplitudes of the Longitudinal Shafting Including
the Mass and Stiffness of the Machinery Space
Input: Structural Plans

Machinery Space Mass Distribution
Thrust Bearing Stiffness
Reduction Gear Mass and Stiffness
Length and Diameter of Shafting
Mass and Added Mass of Propeller
Longitudinal Excitation Levels
Output: Longitudinal Shafting Resonant Frequencies

Amplitude of Motion at Thrust Bearing
References: 60 - 67
39
Two other methods of determining the transverse response should be
mentioned at this point. These are the transfer-matrix and finite-differ-
ence methods. Reference 72 discusses the transfer-matrix approach and
applies a particular computer program to the analysis of several ships.
The finite-difference method is also used in a fully operational computer
code available from the David W. Taylor Naval Ship Research and Develop-
ment Center. This program, called GBRP [73], can also treat longitudinal
and torsional vibrations of the shafting as well as the coupled lateral
and torsional vibration of a beam. This makes it particularly well suited
to study the bending response of a ship which can be idealized as a beam.
This application of the GBRP program will be discussed later in the report.
Since the analysis described in the next section is an extension of

the one assuming a rigid hull, it is wise to choose a computer program
which can be used for both. This will allow the analyst to use the al-
ready existing model of the shafting and only add the flexibility and mass
of the surrounding hull structure. With the ability of general purpose
computer programs to model longitudinal, lateral, and torsional vibration
of beams with the same finite elements, it may be possible for the analyst
to use the model already developed for the longitudinal vibration of the
shaft. Table 11 shows the ínformation necessary to carry out this analy-
sis of the shafting with rigid lateral supports.
TABLE il. SUMMARY OF DESIGN BLOCK 16--DETERMINE FORCED LATERAL

RESPONSE OF SHAFTING (RIGID HULL)
Purpose: To Compute the Resonant Frequencies and Forced

Response of the Lateral Shafting Along with Shaft
Bearing Reactions
Input: Length and Diameter of Shafting

Location of Bearings
Oil Film Stiffness
Vertical and Horizontal Excitation
Levels for Propeller Forces and Moments
Output: Resonant Frequencies

Response Amplitudes of Shafting Versus Excitation
Frequency
Static and Dynamic Bearing Reactions
References: 68 - 73
40
17. Determine Forced Lateral Response of Shafting (Flexible Hull)
The analysis conducted in Design Block 16 will ensure the response

of the shafting to lateral excitation will be acceptable if the bearings
are assumed rigid. This is the simplest type of analysis,
and the next
level of sophistication assumes some flexibility in the bearing
supports.
The information produced by this investigation is again the resonant fre-
quencies and forced response of the lateral shaft along with the bearing
reactions. The dynamic bearing reactions should
again be positive to avoid
slamming and gradual loss of contact.
If gyroscopic effects are neglected (they are important for whirl-

ing, but relatively unimportant at blade frequencies), and the
supports
are of equal stiffness in all directions, the natural frequencies and
response of the shaft will be the same in all directions. If the struc-
ture is not symmetrical, the fundamental normal modes may be in any pair
of orthogonal planes. The moment restraint at the bearings can also have
a significant influence on the shaft frequency.
The amount of structure to include in the stiffness calculation is

a matter for the analyst's judgment. The object is to evaluate the stiff-
ness to a region of a large hull mass. For a single-screw ship, this
may
involve the structure from the after peak bulkhead and up to the steering
gear flat. For shafts supported by struts, it will include the struts and
their backup structure. If the complete hull is analyzed using a fInite.-
element analysis, the validity of the modeling can be tested.
Since the structure supporting the shaft bearings is complicated,

the use of finite-element methods is the most feasible way of determining
the support stiffness. Figure 8(b) is a finite-element model of
the after-
body of a ship with the aft bearing structure shaded. When this model is
generated, consideration should be given to the fact that it will also be
used as a substructure in the complete ship analysis. The basic philosophy
is to have adequate representations of each subsystem available
so they can
be assembled as efficiently as possible. The stiffness between the shaft
and bearing of a stave bearing can be quite low if the staves
are rubber.
The stiffness of an oil film bearing is such that a bearing force intro-
duces a motion having a component perpendicular to the load.
If, as a result of the calculations of shaft response, it is found

that there are no shaft resonances within about 30 percent of the
operat-
ing speeds of the ship, the shafting can be considered satisfactory for
this level of refinement. Later analyses of the whole ship will confirm
its suitability. If, on the other hand, lateral resonances appear close
to the operating speed, then by changing one or more of the following, a
new propulsion system can be developed which has resonances properly lo-
cated:
The overhang of the propeller beyond the stern bearing.
The span between the last two bearings supporting the

propeller shaft.
41
The diameter of the propeller shaft.
The support of the propeller shaft bearing:
The skeg and stern tube structure for a single-screw

ship having a skeg supported bearing.
The angles, size, attachment to the bearing barrel of

the arms carrying a strut bearing for open-screw ships.
The structure supporting strut arms.
Other changes as indicated by the calculations.
Such changes are frequently required, and good judgment, often using anal-
yses of simple models, is required to discover the optimum solution rapidly
and inexpensively.
The references listed in the last section are all applicable to the
shafting analysis considering flexible lateral supports. In addition,
Reference 74 addresses specifically the problem of shaft vibrations in
elastically supported tail shafts.
Table 12 summarizes the information required to conduct this analysis

With the exception of the scantling plans necessary to model the mass and
stiffness of the hull structure surrounding the shaft, no input data in
addition to that used in Table 11 are required. Appendix D contains sev-
eral computer programs specifically written for the analysis of lateral
vibration of the shafting. However, any of the commercially available
finite-element computer codes could also be used.
18. Conduct Superstructure Modal Analysis
In addition to substructures of the shafting and machinery spaces,

subsystem since
it is desirable to make a study of the superstructure as a
Fi-
resonances in this region are a frequent cause of vibration troubles.
nite-element methods are generally most suitable for modeling this struc-
container ships and very
ture. With the high superstructures common on
long ships, some superstructures vibrate fore and aft as a cantilever beam.
On others the decks vibrate symmetrically within the sides, while on still
down) so
others the decks vibrate anti-symmetrically (port up, starboard
that the finite-element model should not be too coarse to suitably repre-
sent the complexity of possible modes.
Figure 11 taken from Reference 75 shows three levels of sophistica-

tion for the superstructure models. Generally, the beam and two-dimensional
models do not represent adequately the structure, and a three-dimensional
One of the most important considerations in the super-
model is required.
structure analysis is how it is connected to the rest of the ship. This
As a first approxima-
is shown schematically by the springs in Figure 11.
fixed and simply-supported conditions and
tion, the analyst could assume
natural frequencies and mode shapes. A bet-
determine their effect on the
ship extending
ter approach is to include in the model a portion of the
42
TABLE 12. SU11MARY OF DESIGN BLOCK 17--DETERMINE FORCED LATERAL
RESPONSE OF SHAFTING (FLEXIBLE HULL)
Purpose: To Compute the Resonant Frequencies and Forced

Response of the Lateral Shafting Along with Shaft
Bearing Reactions when Flexibility of the Surround-
ing Hull Structure is Considered
Input: Length and Diameter of Shafting

Location of Bearings
Oil Film Stiffness
Hull Scantling Plans
Vertical and Horizontal Excitation Levels for
Propeller Forces and Moments
Output: Resonant Frequencies

Response Amplitudes of Shafting Versus Excitation
Frequency
Static and Dynamic Bearing Reactions
References: 68 - 74
1. SUPERSTRUCTURE MODELS
1.1 BEAM 1.2 2-DIM. 1.3 3-DIM.
FIGURE 11. THREE LEVELS OF SUPERSTRUCTURE

MATHEMATICAL MODELS, FROM [75]
43
below the main deck. This is illustrated in Table 13 [76] which shows
calculated and measured fundamental natural frequencies of the superstruc-
ture as a function of the type of finite-element model. In this case, the
model containing the engine-room below the superstructure has an eight per-
cent error with the measured value. The error, of course, decreases as
more of the ship is included. Whichever type of model the analyst chooses,
he should realize that the representation will be used as a substructure
for the complete ship analysis.
On ships whose superstructures vibrate badly, it has been found that

the natural frequencies generally are in resonance with the propeller ex-
citation frequencies. For this reason, it is recommended that these fre-
quencies should not be within about 25 percent of each other. If this con-
dition is not met, then modifications must be made to the superstructure
or its surrounding foundation. Figure 12 shows the mode shape at 10.5 Hertz
and indicates that the superstructure vibrates essentially as a rigid body
on a flexible foundation. If the blade rate frequency of the propeller is
10 Hertz, then the superstructure should probably be stiffened at the points
shown to raise its fundamental frequency to about 12.5 Hertz. The adequacy
of these modifications can be confirmed when the analysis of the complete
ship is conducted.
Table 14 shows the information necessary to conduct the superstruc-

ture analysis.
TABLE 13. CORRELATION BETWEEN MEASURED AND CALCULATED

SUPERSTRUCTURE FIJNDANENTAL RESONANT FREQUENCY FOR
DIFFERENT FINITE-ELEMENT MODELS. 138,000 DWT
TAÎKER, BALLAST CONDITION, FROM [76J*
Model Size Number of Frequency Error

Finite Element Model
Elements Nodes- Supernodes DOT Hz
464 130 52 61 11.5 39

A
1190 400 68 165 9.0 8

B
510 134 315 8.8 6

C 1500
2150 810 185 445 8.2 -1

D
A - Superstructure clamped at main deck

B - Engine-room below superstructure included (mass free)
C - Complete model of afterbody
D - Total ship included (only starboard side is considered in all models)
*
Superstructure fundamental frequency measured value 8.3 Hz
44
SHAFT RPM = loo
LOWEST FREQUENCY
NO. BLADES = 6
f = 10.5Hz
BLADE RATE
FREQUENCY = lO Hz
FIGURE 12. FIRST-MODE SHAPE OF THREE-DIMENSIONAL

SUPERSTRUCTURE MODEL, FROM [75]
TABLE 14. SUNMARY OF DESIGN BLOCK 18--CONDUCT

SUPERSTRUCTURE MODAL ANALYSIS
Purpose: To Determine How Close Natural Frequencies of Free

Vibration of the Superstructure Are to the Harmonic
of Blade Rate Frequencies
Input: Scantling Plans

Estimates of Structural and Equipment Mass
Output: Superstructure Natural Frequencies

Superstructure Mode Shapes
References: 75 - 77
19. Determine Natural Frequencies and Forced Response of Rudder
Figure 1 indicates that excitations from the propeller can enter the
ship through the rudder as well as through the propulsion shafting and hull.
Although experience has shown that these rudder excitations are not gener-
ally as severe as the latter two, care must still be taken in the design
stage to avoid them. This can best be done by designing a rudder whose
natural frequencies are not close to the propeller's blade rate frequency
or its harmonics and which will be subjected to accepted levels of pro-
peller-generated pressures. Procedures for calculating the natural fre-
quencies of the rudder are further developed than those for calculating
the excitations.
A ship's rudder is essentially a flat plate immersed in a fluid. In

order to compute its natural frequencies, the stiffness and mass of the
system must be known. The system's stiffness is simply the stiffness of
the rudder with no surrounding fluid. The mass, however, is composed of
two quantities: the structural mass of the rudder and the so-called added
mass of the surrounding fluid. This added-mass effect reduces the natural
frequencies of the structure from their in-air values.
45
There have been many investigations to calculate the added masses
of different geometric shapes. For example, the added mass of a thin rec-
tangular plate was computed in Reference 78 using potential theory, and
experimental tests on the vibration characteristics of cantilever plates
in water were conducted in Reference 79. With the development of the fi-
nite-element technique, numerical methods are now available for computing
on a routine basis the free vibration natural frequencies and mode shapes
of an arbitrarily submerged structure. The results of one such recent
investigation on partially and fully submerged cantilever plates are re-
ported in Reference 80. The numerical calculations were made using the
NASTRAN computer code, with the plate being modeled by standard plate
finite elements and the fluid by isoparainetric three-dimensional solid
elements. Excellent comparison between numerical and experimental results
were obtained, and it appears this finite-element technique can be used
routinely on free vibration analyses of ship appendages such as rudders.
The computation of the excitation forces on the rudder is not as

simple as the calculation of the natural frequencies. A numerical pro-
cedure and computer program using unsteady-lifting-surface theory has
been developed at Davidson Laboratory [81] and subsequently modified to
take into account the effects of loading and thickness [82]. However, it
does not appear that this procedure is used at the present time for design
calculations.
There have also been experimental investigations of the interaction

between the rudder and propeller operating in a given wake field. The
major contributions in this field were made by Professor Frank M. Lewis
at MIT. References 52 and 53 indicate that the propeller-generated forces
are extremely sensitive to the axial clearance between the rudder and pro-
peller, and by proper placement of the rudder, low values of horizontal
force can be obtained. However, the rudder in the experimental tests is
essentially rigid, while the full-scale rudder is an elastic body. This
elasticity can play an important role in determining the response, and
the correct scaling parameters should be included in the model tests.
Currently, little attention is paid to propeller-rudder interaction

effects, and clearances are usually chosen on the basis of those recom-
mended in various classification rules. Additional research is required
in this area to fully understand this area of propeller-induced vibrations.
20. Evaluate Local Plating Design
Up to this point, the purpose of the proposed design procedure has

been to be able to compute the excitation and response of the major corn-
ponents of the ship. The overall objective will be realized in Design
Block 22, when the response of the entire ship will be calculated. Even
though the overall vibration characteristics of the ship's hull girder
and substructure may be acceptable, the possibility still exists that
local vibration problems may arise. Reference 7 defines local vibration
as the dynamic response of a structural element, deck, bulkhead, machinery,
or equipment elements which is significantly greater than that of the hull
girder at that location. However, hull girder vibration does not mean much
when the nodal length is of the order of the beam of a ship. The concept
46
of local vibration in this report is more restrictive in that local vibra-
tions are defined in terms of major structural elements rather than the
hull girder.
Since local vibration is simply an extension of the global vibration

problem, the finite-element techniques proposed in the earlier design
blocks would apply. This is not recommended for the entire ship, however,
because the analysis would become too complicated and expensive. A sim-
pler approach is taken in which the local plating natural frequencies are
removed from the excitation frequency or its harmonics. The response of
the plating is not computed. This approach allows the use of analytical
expressions for plate vibrations such as found in Reference 83. Reference
68, published by Bureau Ventas, contains formulas for the fundamental
natural frequencies of plate panels stiffened in one or two orthogonal
directions with various boundary conditions. The effects of the fluid
added mass on the vibration of plating adjacent to a fluid boundary are
also considered in Reference 68. These formulas are generally not easy
to apply because there are difficulties in defining the extent of the
structure taking part in the vibration, the geometric boundary conditions,
and the structural mass. In critical areas of the ship where the struc-
ture is complicated, these frequencies can be computed using the finite-
element method. Also, added-mass effects of external or internal liquid
can be included by the techniques explained in Reference 80.
Generally speaking, it would be too complicated to compute the nat-

ural frequencies of all local plating in the ship. By using an experienced
senior draftsman or designer, most of the structural details of the local
plating can be finalized in accordance with standard practice and exper-
ience. Only when a new structural arrangement or an unusually large un-
supported area is proposed, is a more detailed analysis warranted. If
local plating vibrations do occur, corrective measures are relatively easy
to implement before the shipts finish work is completed. These local vi-
brations can be found during the shaker tests which will be discussed in
Design Block 23.
TABLE 15. STJNMARY OF DESIGN BLOCK 20--

DESIGN LOCAL PLATING
Purpose: To Determine How Close Natural Frequencies of Free

Vibration of the Local Plating Are to the Harmonics
of the Propeller-Blade Rate Frequency
Input: Scantling Plans

Equipment Mass Estimates
Added Mass Estimates of Adjacent Fluid
Output: Local Plating Natural Frequencies
References: 68, 80, 83
47
21. Assemble Model of Entire Ship
have been designed so
When the subsystems discussed in Blocks 14-20
vibration resonances, it is
that it is expected that they will be free of
This analysis of
time to make a vibration analysis of the complete ship.
the full ship fulfills two important functions:
It checks and confirms the validity of the boundaries

assumed for the substructures.
By modeling the ship as a whole, it is possible, with the

in all
proper damping, to predict the vibration levels
parts of the ship as a function of frequency. Comparing
these predictions with established acceptable levels
allows an assessment of acceptability of the ship at a
changes to
point in construction where corrections and
and in-
overcome serious difficulties can be determined
corporated in the design.
for the
Until approximately ten years ago, the standard technique
using a
complete ship analysis was to model the ship as a beam structure
This approach is still satisfactory for cer-
Timoshenko beam as a base.
beam-like struc-
tain type ships which are compartmentalized and respond as
However, for ships
tures. Submarines and destroyers fall into this class.
do not give satisfac-
with wide beams and open deck structures, the models
the hull-girder modes and frequencies are affected by
tory results because bulk carriers are
vibrations wíthin these areas. Large containerized and
finite-element method
examples of such ships, and for these cases the
Even in the cases
should be used to define the structure accurately.
used, the results are generally not accurate
where beam models can be shear lag and
for above the fifth or sixth natural frequency because of
local vibration effects.
longitudi-
Reference 84 presents a review of the theory for bending,
them directly to the
nal, and torsional vibrations of a beam and applies
of all analytical arid numeri-
ship vibration problem. A more recent review presented in Reference 85.
cal methods used to calculate hull vibration is
beam, consider Figure
As an example of a ship modeled in terms of a
vibration model, and Figure 14, a coupled lateral-
13, an uncoupled vertical
torsional vibration model. The sprung masses shown in these figures rep-
For the analysis of structures
resent the various subsystems in the ship.
14, the computer program GBRP (Gen-
as complex as shown in Figures 13 and Ship Research and
eral Bending Response Program) developed at the Naval
In the analysis of the ship hull, it
Development Center f73] can be used.
lateral-torsional vibration. These
can treat vertical as well as coupled described in Appendix
capabilities and other features of the program are
E-l.
it is necessary
To define the elastic properties of the structure,
A program and procedure
to define the cross-sectional elastic properties.
for computing
48
MMMII
SYSTAM 56 - ENGINE ROOM
IIIIJM AuXILIARY NASSES
Ilulu lIIlIIllIIIUllITIlll huh. MihlUflhhihIhiUlllhl

ERS
(,o _,5 5] 60 o
ENGINE ROOM SYSTEM Il - luit).
L (o 55
SYSTEM #2 - STERN TARA AlIt) SEEK
I ç)
SYSTEM #3 - 'ROI'. AN)) SHAFT
FIGURE 13. UNITIZED AND CONTAINERIZED SHIP, VERTICAL VIBRATION MODEL
Ar. AUS] .105? 1)00?

iNK IRA S 6)11
S] ERS
IIIIIIIU ululi 'I 1111111 uu u uuui Ill

55 50 4._ENGINE
PIS_A)
SYS rEM DI - WS.l,.J
LI I 1J
/
SYSTEM # 3 - lo
pRor. SrLsrr
o
I j
SYSTEM RA - SEliG
FIGURE 14. UNITIZED AND CONTAINERIZED SHIP, TRANSVERSE VIBRATION MODEL
49
I (Moment of inertia about transverse axis)
y
(Moment of inertia about vertical axis)
I (Product of inertia relative to horizontal and

yz
vertical axes)
A (Cross-sectional area)
KxzA (Shear area constant vertical plane)
KxyA (Shear area constant transverse plane)
(Torsional area constant about longitudinal axis)
(Vertical coordinate of the neutral axis)
(Transverse coordinate of neutral axis)
y',z' (Coordinates of the shear center of the section)
are given in Reference 86. This program calculates the equivalent beam
parameters for the ship section properties using data tabulations obtained
from hull plans by a preestablished orderly procedure.
U. S. Steel Engineers and Consultants, Inc., developed a program

that represents a ship as a beam on an elastic foundation. Information
on this program is given in Appendix E-2. Other vibration programs based
upon modeling the hull as beams have been developed by Lloyd's Registry
of Shipping and by the Institut fur Schiffstechnik in Berlín, Federal Re-
public of Germany. These programs are described briefly in Appendix E-3.
The advantage of representing a ship by a beam model is that the

computer analysis is more direct and more easily interpreted and is con-
siderably less expensive than that with a finite-element analysis for a
structure that is as well defined. As mentioned earlier, the disadvan-
tages are that for many ship vibration problems, particularly where the
decks are open so that the vibration across the width of the ship is im-
portant, the beam representation of the ship is inadequate and a finite-
element process is required for satisfactory modeling.
The use of finite-element methods for predicting ship vibrations is

the method
becoming widespread. References 63, 75, 77, 87, and 88 discuss
application to ship structures. Finite--
and give specific examples of its
element applications in vibrations and structural dynamics problems of
ships and other marine structures are discussed in Reference 89, and an
extensive bibliography is presented.
Figures 15 and 16 show elevation and isometric views of a finite-

element model for a cargo ship. The forward portion of the vessel is
represented as a beam with offset masses. This is an acceptable practice
in this case because detailed knowledge of the vibration response in the
and
forward area is not required. Excessive vibrations in the habitable
50
uii
-II-
v4U
FIGURE 15. ELEVATION VIEW OF FINITE-ELEMENT MODEL
FIGURE 16. ISOMETRIC VIEW OF FINITE-ELEMENT MODEL
51
and machinery areas are of greater concern, and Figures 15 and 16 show
those portions of the ship are modeled in great detail.
Some organizations have developed finite-element programs specifi-

cally for ship applications. Among these should be mentioned SESAN-69
(Appendix E-3) developed by Det norske Ventas, the Norwegian classifica-
tion society [90], and DASH (Appendix E-3) developed by the Netherlands
Ship Research Center. Bureau Ventas, the French classification society,
also has made many-finite element analyses of ship structures. The Elec-
tric Boat Division of General Dynamics Corporation developed and maintains
a finite-element computer program CENSAN for use on submarine vibration
problems. Information on this is presented in Appendix E-3.
Since the cost of developing, maintaining, and updating a large fi-

nite-element computer program is high, it is common to apply general pur-
pose computer programs to ship vibratíon problems. In particular, NASTRAN
is used by Lloyd's, the American Bureau of Shipping, Littleton Research
and Engineering Corp., and probably by other organizations. This program
was developed for the analysis of large structures and readily applies to
ships. It is being updated continually to improve the representation of
structural elements and the processing efficiency. Because it was devel-
oped for large projects, NASTRAN carries high overhead structure. Most
ship finite-element studies involve models that are large enough to bene-
fit from the generality provided by the overhead, but for many small studies
that will not later be incorporated in the large model, it may be desirable
to use other finite element programs such as STARDYNE, ANSYS, MARC, STRUDL,
or SAP. Again, the choice should be dictated by the program's availability
and experience of the user.
This section has discussed the various ways of modeling the entire
ship and has presented computer programs which are used for this purpose.
A summary of this design block is given in Table 16. This model will be
used in the next design step to predict the vibration levels throughout
the ship.
22. Determine Vibration Amplitudes and Stress Levels of Complete Ship
At this point in the design process, all the ingredients necessary

to compute the forced vibration response of the entire ship are available.
The hydrodynamic excitations are available through the methods discussed
in Design Blocks 6-13, and a mathematical model of the entire ship struc-
ture can be assembled in Design Block 21. This step in the design process
simply applies the excitations to the model and determines the vibration
response. Table 17 provides a summary of this design block and indicates
the only information not yet discussed is the estimate of added mass and
damping.
The concept of added mass as it applies to the rudder was discussed in

Design Block 19. For the complete ship, it is generally assumed the added
mass underway is the same as for the ship at rest. At the present time,
there is little known about the differences.
Damping is, of course, very important in response calculations because

without damping, the resonant amplitudes of vibration have infinite values.
52
ASSEMBLE MODEL OF ENTIRE SHIP
Purpose: To Construct a Mathematical Model of the Entire Ship

Which Can Be Used to Predict the Response to
Propeller-Induced Excitations
Input: Substructure Models of Superstructure, Machinery

Space, Stern Structure, and Propulsion Shafting
Scantling Plans
Mass Distribution
Output: A Mathematical Model of the Entire Ship
References: 63, 75, 77, 84-90
TABLE 17. SUMMARY OF DESIGN BLOCK 22--DETERMINE

VIBRATION AND STRESS LEVELS OF ENTIRE SHIP
Purpose: To Compute the Forced Vibration Amplitudes and

Stress Levels of the Complete Ship Under Propeller-
Induced Excitations
Input: Mathematical Model of Entire Ship

Computed or Measured Propeller Forces and Moments
Computed or Measured Hull Forces or Pressures
Computed or Measured Rudder Forces
Locations on Ship Where Vibration Responses Are
Desired
Estimates of Ship's Added Mass
Output: Plots of Vibration Amplitude Versus Propeller RPM

at Selected Locations
Plots of Maximum Stress Versus Propeller RPM at
Selected Locations
Resonant Frequencies at Selected Locations
References: 63, 75, 77, 84-93
53
Investigations have shown that damping in ships is dependent on the vibra-
tion mode, the manner of construction and type of construction material, and
the type of cargo. It can be divided in the three forms
Structural, or hysteretic, damping
Coulomb, or dry friction, damping
Viscous, or fluid, damping,
and all three types are present in ship structures. Damping is usually
calculated on the basis of logarithmic decrement measurements obtained
either in vibration generator or anchor drop tests. Kumai [91] addressed
the question of how damping is affected by the higher modes of vibration,
and in Reference 92 values of the hysteretic damping coefficients, g, are
given for several United States, Japanese, and English ships in the verti-
cal modes of vibration. A value of g = 0.029 was recommended to be used
in the General Bending Response Program, at least until further data could
be obtained. However, each organization conducting analyses of this type
has its own recommended values of damping.
Specific results obtained from a particular forced vibration analy-

sis will be illustrated in detail in Chapter V when the application of
the recommended procedure is presented. At this point it is only necessary
to illustrate generally what type of results can be expected. Figure 17
shows a two-dimensional finite element model taken from Reference 75 for
a 370,000 dwt tanker. The forward portion of the ship is not represented
by a beam in this model. Let us assume that the forced response in the
horizontal direction at the top of the superstructure is desired for a
vertical force acting at the ship's stern. The assumed viscous dampíng
coefficient was taken to be 3.5 percent of critical damping. Figure 18
shows, however, that the coefficient is an important factor only in the
vicinity of the resonant peaks. The analysis presented in Reference 75
shows that the top of the superstructure has two resonant peaks at about
6.7 and 11.0 Hertz, as illustrated in Figure 19. The higher peak is due
to the natural frequency of the superstructure. Figure 20 reveals the
lower peak is essentially a rigid-body motion of the superstructure in-
duced by a hull bending.mode. At 6.67 Hertz a nodal point is located at
the aft position of the superstructure which will give extreme rigid-body
motion. This is the blade-rate frequency for a five-bladed propeller at
80 RPM service speed. If a six-blade propeller is used, this frequency
increases to 8.00 Hertz, and Figure 20 indicates that the horizontal re-
sponse at the top of the superstructure is reduced greatly. However, if
the service speed for the six-bladed propeller increases to 103 RPM, then
blade frequency will equal the natural frequency of the superstructure.
Reference 75 indicates the calculated vibration levels would be "very un-
pleasant" in this case.
23. Conduct Shaker Tests
Up to this point in the design process, all estimates of vibration

levels have been based upon analytical response predictions. Since it is
imperative the ship's actual response not exceed that established in the
specifications, experimental tests should be conducted to excite the ship
and measure the response before all finish work is completed. These tests
hopefully will indicate any areas of structural deficiency, and corrective
action can be taken before the ship's sea trials are begun. It should be
54
A
8
o
7
Y--
Io V
-I 6
VERTICAL FORCE IO Mp Il
CI)
5 /
I /
V) E.
C.O 6 /
I
E. 3
I-.
C
oz I)-. 2 \ REDUCED DAMPING

I.
/ I
o \c/c 0 02
o CI
o "Nc/Cc 0.035
o FREO
6 7 8 g Hz 11
FIGURE 17. FINITEELEMENT MESH FOR TWO- FIGURE 18. FORCED RESPONSE DEPENDING ON THE
DIMENSIONAL MODEL OF 370,000 DWT APPLIED GLOBAL DAMPING VALUE. 370,000 DWT
TANKER, FROM [75].
TANKER, BALLAST CONDITION, FROM [75].
4
8
X5-667 A5r800
VERTICAL FORCE IO Mp_
FREQUENCY Hz
HORIZONTAL FORCE IO
8O
---
70
FREQUENCY 60
O 11
SO POSITION OF NODAL POINTS IN MAIN DECK
Hz
FIGURE 19. CALCULATED FORCED RESPONSE AT THE TOP FIGURE 20. POSITION OF NODAL POINTS IN THE MAIN
OF SUPERSTRUCTURE IN LONGITUDINAL DIRECTION. DECK FOR FORCED VIBRATIONS CALCULATIONS OF THE
370,000 DWT TANKER, BALLAST CONDITION, FROM [751. HULL GIRDER. 370,000 DWT TANKER, BALLAST
CONDITION, FROM [75].
realized that vibration problems in the hull girder, engine room, or super-
structure may be very difficult and expensive to correct at this point.
The primary purpose of the shaker tests is to ensure the local plating and
other local structures are properly designed. The shaker test can also be
used to assess the accuracy of the structural modeling techniques and pro-
vide estimates of the ship's structural damping. These will be discussed
in Design Block 24.
The tests should be conducted with the ship sitting in the water and
the shaker mounted on a rigid foundation at the ship's stern. Surrounding
disturbances should also be minimized so that the resulting vibration
amplitudes can be related directly to the excitation forces. This may
require conducting the tests at night when work on the ship is at a mini-
mum and the water is calm. From experience in measuring vibrations gener-
ated in the bottom of a machinery space, it is believed that the stern
shaker need not be large if the above conditions are met. The rigid foun-
dation is necessary to excite global hull vibrations and not local vibra-
tions in the vicinity of the shaker. The shaker should also be positioned
as closely as possible over the propeller to more realistically simulate
the propeller excitations. Figure 21, taken from Reference 94, shows the
position of the exciter and some correlation of the experimental results
with the finite-element model calculations. However, excitations should
also be applied in the horizontal direction at the stern to assess the
lateral-torsional behavior of the ship. Reference 94 describes the test
procedures and results of shaker tests performed on a cargo-liner.
15th
Calculated free vibration mode form
of fully loaded ship.
aft draft 10.50 a.
6.74 Hz - f16 - 604 cpa.
\
Measured free vibration mode form.

F - 480 cpa.
aft draft - 6.40 a.
Equivalent beams of forward part.
4Hr -
i T1IJ_L li -
FIGURE 21. ELASTO DYNANIC MODEL OF AFT PART AND CORRELATION OF

EXCITER TESTS WITH FREE VIBRATION CALCULATIONS, FROM [94]
24. Assess Local Vibrations, Structural Damping, and Modeling Techniques
The shaker tests discussed in Design Block 23 provide a means of col-

lecting some very valuable information about the vibration characteristics
of the ship. By applying excitations at the stern and measuring the ampli-
tudes and phase angles of the response, these tests will show if local
resonances exist in the structure at the blade-rate frequencies or their
56
harmonics. The measurements should be taken at the points defined in the
ship's specifications, as well as other places where local plating vibra-
tion could be present. If deficiencies do exist and can be corrected by
local stiffening of the structure, these modifications are much easier
and less expensive to make now rather than waiting until all finish work
is completed.
The shaker tests also provide information on the amount of damping

present in the structure. These estimates are usually obtained by comput-
ing the width of the resonant peaks. Reference 92 discusses this method
along with the technique of impulsively exciting a ship and observing the
decay of the free vibration modes. Anchor drops or underwater explosions
are two methods of applying these impulsive loads. Although detailed know-
ledge of the damping will probably not help the design of the present ship,
it will be valuable in future designs. These tests will help assess the
influence that such variables as choice of materials, method of construc-
tion (e.g., welded or riveted), and type of framing have on the ship's
damping characteristics. However, the tests will not provide any infor-
Ination on the damping due to the cargo because they are conducted in an
unloaded condition.
The third way in which the shaker tests can aid the design process
is to help estimate the accuracy of the structural modeling techniques.
These tests will provide the measured response at various locations
throughout the ship as a function of a known excitation at the stern.
By applying excitations to themathematical model of the complete ship
developed in Design Block 21, the response can be calculated and compared
with the measured values. This should provide a reliable means of judg-
ing the accuracy of the structural model since the input forces are accu-
rately defined. The model can also use the damping computed earlier from
the shaker tests.
Table 18 summarizes the information in this design step.
25. Measure Vibrations During Sea Trials
After the construction of the ship has been completed, the next step
is to conduct the sea trials. The purpose of these trials is to determine
how well the ship meets its performance specifications and identify any de-
ficiencies in the design. This includes taking vibrations measurements at
specified locations throughout the ship under known conditions so that these
levels can be compared with the specifications defined in Design Block 1.
This report is not intended to serve as a guide for these vibration

tests. References 95 and 96 are specifically written for this purpose.
Reference 95 discusses vibration survey techniques, and an acquisition and
processing system for ship vibration which can simultaneously measure about
60 different locations is described in Reference 96. Vibrations measure-
ments to correlate with analyses should be much more detailed than those
required by most codes such as given in References 6 and 7. As a word of
caution, the vibration tests must follow a detailed test plan that ensures
that all required vibration levels be measured. As a minimum, the plan
should include:
57
TABLE 18. SUMMARY OF DESIGN BLOCK 24--ASSESS LOCATION VIBRATIONS,
STRUCTURAL DAIIPING, AND MODELING TECHNIQUES
Purpose: To Determine if Local Resonances Exist in the Ship

at Blade Rate Frequencies or Their Harmonics; To
Provide a Means of Collecting Data on Hull Damping
for Future Designs; To Estimate the Accuracy of the
Modeling Techniques
Input: Results of Shaker Tests

Structural Model of Entire Ship
Output: Location of Local Resonances in Structure

Estimates of Structural Damping
Quantitative Assessment of Accuracy of the Shipts
Structural Model
References: 91 - 94
Test Conditions
Instrumentation Required
Measurement Techniques
Measurement Positions
Data Reduction and Interpretation
Report Requirements
Since there is only a limited amount of time allotted to vibration measure-

ments during the sea trials, any confusion as to the test procedures will
result in an incomplete set of vibration measurements.
26. Compare Measured Vibrations with Specifications
The real test of how well the ship was designed to minimize propeller-
induced vibrations comes in this design block. This block is not related
to design per se, but is part of the overall recommended design procedure.
Its purpose is to assess quantitatively the acceptability of the shipTs
design with respect to propeller-induced vibrations.
All of the inputs necessary for making this comparison will be known
at this point. The vibration specifications were defined in Design Block
1, and the corresponding vibration levels were measured during sea trials
in Design Block 25. If the measured values are below those specified,
then the ship's design is acceptable from the vibration point of view.
If the measured values exceed the specifications, then the design is not
acceptable and corrective action must be made. Responsibility for this
58
action will be contained in the vibration specifications if they are writ-
ten correctly.
27. Compare Measured Vibrations with Calculations
Design Block 26 discussed how the acceptability of the ship was

judged by comparing the measured vibrations with the specifications. If
the vibrations levels are acceptable, then it could be assumed the recom-
mended design procedures were adequate and were applied successfully.
This may not be the case. There are instances of ships which should have
been vibration free and had severe vibration problems. There are also
examples where calculations predicted unacceptable vibration levels, but
none were present when the ship was built. These point out that rather
than the recommended procedures producing an acceptable design, the de-
signers were simply lucky that the ship had no vibration problems. If the
procedures are continued, then it is only a matter of time before a ship
which vibrates badly will be produced. The last step in the design pro-
cess is, therefore, to assess the process itself. The purpose of the
assessment is to judge the accuracy and acceptability of the established
design procedures and to identify areas in the design process in which im-
provements can be made. This can best be done by comparing the measured
vibration levels taken during the sea trials with those computed for the
entire ship assembly. A comparison of this type will provide the design
team a quantitative description of how successful it has been in producing
a ship with minimum propeller-induced vibrations. Let us assume, for
example, the measured mode shapes of the hull girder compare favorably
with those predicted, but the vibration amplituds do not agree. This
would indicate that the hull modeling procedures are correct, but the
methods of calculating the propeller-generated forces and pressures are
inaccurate and that improvements are needed in that area. This evaluation
and feedback process will be discussed in the next chapter.
The writers of this report would like to emphasize again a previous

point: the comparison between measured and calculated vibration levels
should be made even though the ship had acceptable levels. This is usu-
ally difficult because of financial and time constraints, but is a neces-
sary step in the design process if the recommended procedures are to im-
prove continually. As a final comment, the results of this comparison
should be disseminated to the maritime industry in some form. Again,
this is difficult to accomplish, especially for ships which vibrate badly,
because no organization wants to advertise it had part in such a design.
59
IV. DESIGN EVALUATION MILESTONES
In any design process, it is necessary to have checkpoints to deter-

mine if the design up to that stage is acceptable and should proceed or if
modifications must be made. For simple problems, this procedure is infor-
mal and is continuously being done in the mind of the responsible engineer.
However, in more complicated systems, such as a ship, where many individ-
uals from different organizations and technical backgrounds are involved,
the process needs to become more formalized. This is the purpose of the
evaluation milestones shown in Figure 3. There are five such points dur-
ing the ship's design procedure presented in this report, and they are
designated as follows:
MILESTONE I - Preliminary Hydrodynamic Evaluation

MILESTONE II - Final Hydrodynamic Evaluation
MILESTONE III - Ship Substructure Evaluation
MILESTONE IV - Complete Ship Structure Evaluation
MILESTONE V - Test and Evaluation Review
Although the evaluation milestones are shown approximately at the end of

each design phase, they are continuously being conducted throughout the
particular phase. Only at the end of a phase would a formal report of
the results be issued.
Since the evaluation steps are associated with a particular design

phase, they serve to divide the design process into smaller tasks accord-
ing to technical disciplines and make identification of problem areas
easier. They also serve to establish milestone points for reporting the
results and charting the progress of the entire project.
The five evaluation milestones discussed in the following sections

are first presented by showing the overall flow diagram for each of the
individual design phases. This diagram is then followed by a more de-
tailed presentation of the various phases showing the evaluation and f eed-
back processes. It is hoped that by utilizing this evaluation procedure,
deficiencies in the ship's design which could cause excessive vibrations
would be identified and corrected before the ship is built.
1. MILESTONE I - Preliminary Hydrodynamic Evaluation
Figure 22 shows the three blocks composing the preliminary hydro-

dynamic design. Block 2, in which the general ship design data are es-
tablished, is not part of this phase, but is shown for completeness. An
expanded description of the phase showing the feedback loops in the de-
sign process is shown in Figure 23. The reader should remember that the
three primary objectives of this portion of the analysis are:
(a) Choose the number of propeller blades so that the blade-

rate frequency is not near the longitudinal propulsion
natural frequency.
60
Block 4
ESTIMATE
LONGITUDINAL
PROPULS ION
FREQUENCIES
Block Block 3
ESTABLISH CONDUCT
DESIGN
I GENERAL SHIP WAKE
PROPELLE R
DESIGN DATA SURVEY
I
Block 5
FIGURE 22. PRELIMINARY HYDRODYNANIC DESIGN PHASE
Choose a stern configuration whose wake distribution does

not introduce large propeller and hull forces.
Design a propeller whose efficiency is as high as possible.
These three objectives are reflected in decision blocks shown ín Figure 23.
A very important part of this phase is to choose a suitable stern

configuration because it determines the nature of the flow into the pro-
peller. Since the propeller forces are dependent on the nonhomogeneous
flow into the propeller disc, their magnitudes and directions are directly
related to the wake distribution and hence the stern configuration. One
way of graphically representing the wake is in terms of the Taylor wake
number, W, defined as
V
a
W =
m
where
V = axial velocity of water to propeller disc
61
REEVALUATE
SPEED, POWER, NO.OF
RPM, OR NO BLADES CIVES
MACHINERY ACCEPT. PROP.
ARRANGEMENTS EFF. ?
REQUIREMENTS
A YES
Block 4
ESTIMATE RECOMMEND
LONGITUDINAL NUMBER OF
PROPULS ION PROPELLER
FREQUENCIES BLADES
Block 3
i_11
ESTABLISH CONDUCT ACCEPTABLE
YES
DESIGN FINALIZE
GENERAL SHIP WAKE WAKE DISTRI- PROPELLER * PROPELLER
[ESIGN DATAJ SURVEY BUTION ? DESIGN
Block Block 5
NO V
MODIFY WAKE
STERN
V
LINES GIVES ACCEPT.
YES
EFF. ?
NOKDISTRIBUTION
FIGURE 23. MILESTONE I - PRELIMINARY HYDRODYNANIC DESIGN EVALUATION

V = velocity of model
m
The wake number is usually plotted on a polar map showing lines of

constant W as shown in Figure 24 [97] for a particular stern configura-
tion. This figure was taken from the axial wake distributions obtained
in model tests for two aft body configurations. The zero degree angle is
located at the top of the propeller. Figure 24(a) indicates that the in-
flow velocity is essentially not moving relative to the ship at the top
position, and there is a steep wake gradient. From experience, this vari-
ation in wake would probably result in excessive excitations, and these
excitations can be reduced by altering the ship's stern lines in the de-
sign stage. Figure 24(b) shows the wake distribution with the aft body
modified, and the wake peaks and gradients are greatly reduced. Other
examples of how propeller forces can be reduced by proper selection of
the stern form and propeller type are given in a recent paper by Okamoto
[98]. Solumsmoen [97] points out that it is difficult to give criteria
for an acceptable wake because the excitations also depend on the pro-
peller geometry. Huse [99] does suggest some guidelines, for example,
Example of axial wake distribution

for an original hull giving rise
to great hydrodynamic forces
1ç 9O inca
12O
45 135
AlhS
150
65°
Axial wake distribution for modified

body lines
FIGURE 24. AXIAL WAKE DISTRIBUTIONS FOR ORIGINAL

AND MODIFIED BODY LINES, FROM [97]
63
For large tankers and other ships with high block coeffic-
ients, W should not exceed 0.75.
max
For ships with small block coefficients, e.g., below 0.6,
W should be below 0.55.
max
Wmax is the wake number measured at the center plane of the propeller in the
range 0.4 to 1.15 of the propeller's radius.
However, more research is needed in this area so that a systematic

correlation between wake distributions and hydrodynamic excitations and pro-
peller cavitation can be established. This correlation is dependent on the
propeller geometry, and an acceptable wake does not guarantee low propeller-
generated forces and no cavitation. But such information would provide the
designer with a rational means of assessing the suitability of proposed aft
body configurations.
2. MILESTONE II - Final Hydrodynamic Evaluation
After the design parameters in the preliminary hydrodynamic phase have

been established and deemed acceptable in its evaluation, the final hydro-
dynamic design may be initiated. This phase was discussed in Sections 6-13
of Chapter III and is presented in Figure 25. The overall objective of the
final hydrodynamic design is to be able to calculate all excitation forces
produced by the propeller and to evaluate if these forces are likely to
cause unacceptable levels of vibration in the ship.
The process by which the final hydrodynamic design is evaluated is

illustrated in Figure 26. This figure includes the Design Step 6-13 along
with the decision points and the feedback paths if redesign is deemed neces-
sary.
The process begins by evaluating if the propeller forces computed in

Block 6 are within acceptable limits. As discussed in Section 6 of Chapter
III, there are no accepted guidelines for making this judgment, and addi-
tional research is required in this area to aid the designer. References
31 and 100 do provide some simple criteria for maximum permissible excita-
tion, although they should be used with caution. If the magnitudes of the
propeller forces and moments are not acceptable, then changes in the design
must be made. These involve modifying the stern configuration to alter the
wake distribution and/or changing the propeller design. If the stern lines
are altered, then a new wake survey must be conducted to determine the new
nominal wake field. This new wake field would in turn require that another
propeller design be made. If the original wake is judged suitable, then
perhaps only changes in the propeller's geometry are necessary. It is cer-
tainly desirable to redesign the propeller without altering the number of
blades. However, if this is necessary, then a new longitudinal analysis of
the shafting (Block 4) must be conducted.
After the propeller forces are judged acceptable, then the calcula-
tions for hull pressures, including the effects of cavitation, can be made.
Chapter III indicated this could be done in either one of two ways. The
indirect method consists of computing the hull pressures without cavitation,
64
Block 12
CONDUCT
CAVITATION
TESTS
Block 7 Block 8 Block 9
EVALUATE
COFUTE HULL EVALUATE
PROPELLER
PRESSURES PROPELLER
CAVITATION
W/o CAVITATION CAVITATION
FACTORS
Block 13
A
* COMPUTE TOTAL
PRESSURES
AND FORCES
Block 10
DIRECT CALCUL.
COflUTE
OF CAVITATION
PROPELLER
PRESSURES AND
FORCES
FORCES
Block 6
Block 11
CONDUCT
MODEL
TESTS
FIGURE 25. FINAL HYDRODYNANIC DESIGN PHASE

Block 12
CONDUCT
CAVITATION
Block 7 Block S TESTS
Block 9
COMPUTE HULL EVALUATE

EVALUATE
PROPELLER
PRESSURES PROPELLER
CAVITATION Block 13
W/O CAVITATION CAVITATION V
FACTORS
HULL COMPUTE TO SHIP
PRESSURES WITH TOTAL
YES Block 10 CAVITATION EFFECTS PRESSURES -*
ACCEPTABLE ? AND FORCES SUBSTRUCTURE
YES DIRECT CALCUL. DESIGN PHASE
COMPUTE OF CAVITATION
FORCES NO
P ROPE LLER
ACCEPTABLE? PRESSURES AND
FORCES FORCES
Block 6 NO
V
RETURN TO REANALYZE
PROPELLER GEOM.
BLOCK 1
4- STERN CONFIG. Block il YES
CONDUCT RESULTS NO REANALYZE

MOD EL COMPARE WITH WAKE
TESTS ANALYTICAL ? SURVEY
RETURN TO
4 BLOCK 3
FIGURE 26. MILESTONE II - FINAL HYDRODYNANIC EVALUATION

evaluating the effects of cavitation, and estimating the cavitation factors.
This procedure was described in Blocks 7-9. The second method, Block 10,
involves direct calculation of the cavitation pressures. It is certainly
desirable to develop fully this direct process, and continuing research is
being conducted in this area. The acceptability of the calculated hull
pressures can now be judged. But as with the propeller forces, there is
no accepted criterion. Huse [101] in 1971 gave the value of acceptable
blade frequency pressure fluctuations p measured at a point directly above
the propeller as
= 13,000 (Pascals)
This figure was based on measurements from 12 ships. Some had vibration
problems, and some did not [102]. Reference 102 reports that the Swedish
State Shipbuilding Tank [103,104] has the following criterion for the
pressure fluctuations above the propeller:
/a
2P < 6250 (Pascals)
r - z
where
= peak value of pressure fluctuations at reference point
V = displacement, full load, m3
D = propeller diameter, m
a = longitudinal clearance of propeller-hull, forward
a = vertical clearance of propeller-hull
The criterion was based on results of measurements made on eleven ships of

different types. Reference 102 reports that the formula sometimes gives
too low values of permissible excitations on small ships and proposes the
following limit on single-screw merchant ships:
< (0.75 +
2allowed
where
L = length of ship in meters
= pressure amplitude calculated from previous formula
For detailed discussions of these criteria, the reader should see Refer-
ences 102-104.
The cavitation tests (Block 12) can also be conducted during this
time either in a cavitation tunnel or in a depressurized towing tank. The
67
purpose of these tests is to study the extent of cavitation on the propeller
blades and to measure the magnitudes of the pressure fluctuations on the
hull. If the measured or calculated hull pressures are not acceptable by
whatever criterion is chosen, then design changes must be made. Figure 26
shows that this again requires either modifying the stern configuration or
propeller geometry.
The model tests (Block 11) are also conducted during the final hydro-
dynamic design. The purpose of these tests is to confirm the validity of
the earlier wake survey by comparing measured and calculated propeller forces
and moments. If agreement is obtained, then more confidence can be placed
in the analytical hydrodynamic calculations of propeller-generated excita-
tions because all predictions required a wake survey as input. If agree-
ment is not obtained, then the original wake survey must be reanalyzed.
This process is also shown in Figure 26.
3. MILESTONE III - Ship Substructure Evaluation
The purpose of the ship substructure design phase is to allow for the
design, evaluation, and modification (if necessary) of each major subsystem
before the complete structure is assembled and analyzed. This process al-
lows the ship to be divided into smaller components for which the analysis
will be simpler than that for the ship as a whole. If each substructure is
properly designed, it is hoped the response will be acceptable when the com-
plete ship model is assembled. Figure 27 shows a flow diagram of the ship
substructure design phase, which was discussed in Sections 14-20 of Chapter
III. This diagram is expanded in Figure 28 to include the various evalua-
tion and feedback loops. These will now be discussed.
The first requirement is to determine if the resonant frequencies and

forced response of the longitudinal shafting as determined by DesIgn Block
14 are acceptable. As a guide, the resonant frequencies should not be with-
in 25 percent of the blade rate frequency or its harmonics. Reference 5
and the design specifications can be used as a guide to permissibla vibra-
tion levels. If the structure is deemed acceptable, then the stiffness and
mass of the machinery space can be added to the model. If not, modifica-
tions must be made to the thrust bearing and foundation stiffnesses. The
process is then continued with the machinery space included as shown in
Figure 28. The same acceptability criteria can be used, with any modifica-
tions being made in the stiffness of the thrust bearing and machinery space.
The lateral response of the shafting with both rigid and flexible hull
models is determined in Design Blocks 16 and 17. If the natural frequencies
are again within about 25 percent of the excitation frequencies or if the
bearing reactions are negative, it will be necessary to change the location
of the bearings or alter the stiffness of the lateral supports at the ship's
stern. This feedback process is illustrated in Figure 28.
The design of the superstructure is the next subsystem to be analyzed.

Since the superstructure is excited only through the ship's hull girder,
its evaluation is based on having the natural frequencies removed from
the excitation frequencies. If this criterion is not satisfied, then the
superstructure's natural frequencies must be changed, possibly by improving
its continuity with the main deck structure. The philosophy of separating
68
Block 20
EVALUATE
LOCAL PLATING
DESIGN
Block 19
DETERMINE FORCED
RESPONSE OF
RUDDER
Block 14 Block 15
DETERMINE FORCED TO COMPLETE

DETERMINE FORCED
LONGITUDINAL
RESPONSE OF
RESPONSE OF
MACHINERY SPACE SHIP STRUCTURE
SHAFT ING
DESIGN PHASE
Block 16 Block 17
DETERMINE FORCED DETERMINE FORCED

LATERAL RESPONSE LATERAL RESPONSE
OF SHAFTING
(RIGID HULL)
OF SHAFTING
(FLEX. HULL)
H
Block 18
CONDUCT
SUPERSTRUCTURE
MODAL ANALYSIS
H
FIGURE 27. SHIP SUBSTRUCTURE DESIGN PHASE
the natural and excitation frequencies also applies in the design of the
local plating. Modifications to the plating which did not meet the cri-
terion would be made by local external stiffening and/or changes in plate
thicknesses.
The final item in the ship substructure evaluation is the analysis

of the ship's rudder. Again the primary means of judging the rudder's
acceptability is based upon how close its natural frequencies lie to the
propeller excitation frequencies. Figure 28 also shows that a criterion
based upon forced response of the rudder should also be included. How-
ever, as discussed in Chapter III, the techniques for predicting propeller
69
NATURAL
EVALUATE YES
FREQUENCI ES
LOCAL PLATING
REMOVED FROH
DESIGN
BLADE FREQ.
YES
Block 20
Block 19
DETERMINE NATURAL ' NO CHANGE RUDDER
CHANGE THRUST FORCED FREQUENCIES AND CLEARANCES
BEARING AND RES PONS E FORCED RESPONSE OR GEOMETRY
FOUNDATION OF RUDDER \ACCEPTABLE 7/
STIFFNES SES
y to
ACCEPTABLE YES ACCEPTABLE CHANGE STIFF-
DETERMINE FORCED DETERMINE FORCED NO
RESONANT FREQ. RESONANT FREQ. NESS OF THRUST
RESPONSE OF RESPONSE OF
AND FORCED AND FORCED BEARING AND
LONG. SHAFTING MACHINERY SPACE MACHINERY SPACE
RESPONSE?,
RESPONSE? A
YES TO COIPLETE
Block 15
Q Block 14
YES
Block 17 SHIP
Block 16
ACCEPTABLE V
ACCEPTABLES RELOCATE BEARING STRUCTURE
DETERMINE LATERAL DETERMINE LATERAL RESONANT FREQ. NO
RESONANT FREQ YES POSITIONS OR DESIGN PHASE
FORCED RESPONSE FORCED RESPONSE FORCED RESPONSE
FORCED RESPONSE OF SHAFTING LATERAL SUPPORT
OF SHAFTING AND BEARING AND BEARING
(FLEX. HULL). STIFFNESS
(RIGID HULL) REACTIONS?
REACTIONS?
NO
Block 18 YES
y
NATURAL MODIFY
RELOCATE CONDUCT FREQUENCIES
BEARING SUPERSTRUCTURE SUPERSTRUCTURE
REMOVED FROM
POS IT IONS RODAL ANALYSIS DESIGN
BLADE FREQ.
FIGURE 28. MILESTONE III - SHIP SUBSTRUCTURE EVALUATION

excitations are not well developed at the present time. Modifications to
the rudder structure and its response must be made by changing its manner
of support, geometry, or clearances with the ship's hull.
4. MILESTONE IV - Complete Ship Structure Evaluation
After the various subsystems analyzed in the ship substructure phase

have been judged as acceptable, the next task is to conduct the complete
ship analysis. This process is fairly straightforward and is depicted in
Figure 29. The mathematical model of the ship is assembled from its compo-
nents, loaded with the propeller-generated forces and pressures, and the
response determined. The evaluation made in Milestone IV is based upon the
magnitudes of the vibration amplitudes and stress levels compared with those
set in the specifications. If the computed quantities are below the speci-
fications, then the complete ship design is acceptable, and the design pro-
cess may continue. If the specifications are exceeded, then the design is
not acceptable and modifications must be made.
Figure 30 shows the complete ship structure evaluation process. In

general, the vibration amplitudes are evaluated in each of the ship's major
subsystems such as the hull girder, superstructure, machinery space, and
propulsion shafting. It is again important to remember that the vibration
and stress levels must be computed at the locations designated in the speci-
fications. If deficiencies are found in the complete ship structure, then
Figure 30 shows that design changes must be made in the deficient structures
and a new analysis conducted. It is difficult to give any specific recom-
mendations as to the possible structural modification which will correct a
particular problem. If resonance occurs with the hull girder, it will be
almost impossible to make structural modifications to avoid this resonance
for all variations in ship mass and ranges of propeller RPM. Possible struc-
tural changes in the superstructure, shafting, and machinery space subsystems
could, from past experience, include:
Block 21
Block 22
AS SEMBLE
DE TE RNINE
MODEL OF VIBRATION TO TEST AND EVALUATION
ANFLITUDE S
ENTIRE AND STRESS DESIGN STAGE
SHIP
LEVELS OF
COMPLETE SHIP
FIGURE 29. COMPLETE SHIP STRUCTURE DESIGN PHASE
71
STRESS AMP.
VIBRATION YES YES
IN HULL GIRDER
AMPLITUDES IN
ACCEPT. FROM
HULL GIRDER
FATIGUE
ACCEPTABLE?
VI EWPOINT?
RECOMMEND DESIGN CHANGES NO $NO

IN DEFICIENT STRUCTURES
AND REFORMULATE
NO NO
APPROPRIATE MODEL
STRESS AMP. YES
VIBRATION YES IN SUPERSTRUC.
AMPLITUDES IN
ACCEPT. FROM
SUPERSTRUCTURE
FATIGUE
ACCEPTABLE?
VI EWI'OINT?
Block Block 22
V 21 TO TEST AND
DETERMINE VIBRATION EVALUATION
ASS EFLE
MODEL 0F AMPLITUDE AND
STRESS LEVELS OF
ENTIRE SHIP
COMPLETE SHIP
-*
DESIGN STAGE
VIBRATION STRESS AMP.

YES
AMPLITUDES IN YES IN MACHINERY
MACH INERY SPACE ACCEPT.
SPACE FROM FATIGUE
ACCEPTABLE? VIEWPOINT?
NO
RECOMMEND DESIGN CHANGES
IN DEFICIENT STRUCTURES
AND REFORMULATE
APPROPRIATE MODEL
VIBRATION
)YES
SHAFTING
ACCEPTABLE?
FIGURE 30. MILESTONE IV - COMPLETE SHIP STRUCTURE EVALUATION

Superstructure
Plate Thicknesses
Bulkhead Locations
Structural Continuity with Main Deck
Addition of Pillars
Strengthen Beams
Addition of Bulkheads
Machinery Space
Bulkhead Locations
Plate Thicknesses
Foundation Stiffness
Addition of Pillars
Strengthen Beams
Addition of Bulkheads
Shafting
Lateral Support Locations

Foundation Stiffnesses
Diameter of Shafting
This portion of the feedback process must rely heavily on the analysts' and
designers' correct assessment of the problem and possible corrective mea-
sures. Figure 30 is intended to show that structural changes should first
be made in the deficient subsystem, with the hope of not significantly alter-
ing the response of the rest of the ship. This may not be possible since
all systems are coupled together. In any case, structural changes must be
incorporated into the complete ship model and the analysis reconducted.
This process must continue until the structural response does not exceed
the limits set in the specifications.
5. MILESTONE V - Test and Evaluation Review
The final phase of the recommended design procedure is for test and
evaluation of both the ship and the design process. Figure 31 shows this
phase, which was discussed in Sections 23-27 of Chapter III. Based on the
input from the shaker tests, the local vibration levels, hull damping, and
structural modeling techniques are determined. Of these three, the assess-
ment of local vibration is the most important because it directly affects
the acceptability of the current design. Hull damping estimates and the
evaluation of the modeling techniques are also valuable, but more from the
viewpoint of future ship designs and design methods. This is illustrated
in Figure 32, which presents the review process for the Test and Evaluation
Phase. The figure shows that if the local vibration levels are not accep-
table as determined by the shaker tests, then the local structure must be
modified before the ship is finished and sea trials begun. These modifi-
cations would usually consist of stiffening the deficient structure by such
means as increased plate thicknesses or adding local stiffeners. However,
if the predicted response at selected locations does not agree with the
measured response to shaker excitations at the stern, then the structural
73
Block 23
CONDUCT Block 26
SHAKER
TESTS COMPARE MEASURED
VIBRATIONS WITH
SPE CIFICAT IONS
Block 25
y
ASSESS LOCAL VIB., MEASURE VIBRATIONS
HULL DAMPING, DURING
MODELING TECHNIQUES SEA TRIALS
Block 27
Block 24
COMPARE MEASURED
VIBRATIONS WITH
CALCULATIONS
-I
FIGURE 31. TEST AND EVALUATION DESIGN PHASE
model techniques are probably deficient and need to be improved. The de-
sign and construction process of the ship is not dependent on this review,
and after the ship is completed, sea trials will begin.
In the sea trials, vibration levels will be measured at locations and

under operating conditions given in the specifications. Figure 32 shows
that the acceptability of the ship from a vibration point of view is based
upon the measured levels meeting the specifications. If this is accomplished,
then the design can be considered adequate. If the specifications are not
met, then the design procedure has failed to produce an acceptable ship, and
corrective measures must be taken.
This report discussed in Section 28 of Chapter III the fact that even
though the vibration specifications were met, the design procedure may be
inadequate. It is necessary, therefore, to end the process with an assess-
ment of how well the complete design procedure works. This can best he done
by comparing the calculated vibration levels with those measured during the
sea trials. If sufficient agreement is obtained, then the recommended pro-
cedure can be considered adequate. If not, a review must be made to deter-
mine the source of errors. These errors arise essentially from two sources:
(1) calculation of the hydrodynamic excitations and (2) calculation of the
ship's response to these excitations. It is difficult to give any specific
rules to pinpoint deficiencies in the analysis techniques. However, if good
correlation is found between the calculated and measured response to the
shaker excitations, the discrepancies are probably caused by inaccurate
calculations of the propeller-generated forces and pressure rather than by
poor structural modeling techniques.
74
DETERMINE CAUSE FOR
INADEQUATE DESIGN.
RECOMMEND
MODIFICATIONS.
Block 23 L
NO
Block 26
CONDUCT MODIFY
SHAKER LOCAL COMPARE MEASURED YES SHIP' S
TESTS STRUCTURE MEASURED VIBRATION LEVELS VIBRATION
VIBRATIONS WITh MEET SPECIFICA-DESIGN IS
Block 25 SPECIFICATIONS TIONS?
+NO ADEQUATE
LOCAL YES MEASURE

VIBRATION VI BRATIONS
LEVELS ACCEPT- DURI NG
ASSESS LOCAL ABLE? SEA TRIALS Block 27
VIBRATIONS, HULL
DAMPING, MODELING COMPARE H EA SU RED DESIGN
TECHNIQUES MEASURED YES
SHAKER AND CALCULATED PROCEDURES
YES VIBRATIONS
RESPONSE AGREES VIBRATION LEVELS ARE
Block 24 WITH ANALYTICAL MITi! CALCULATIONS AGREE? ADEQUATE
MODEL?
REVIEW STRUC- REVIEW DESIGN

TURAL MODELING PROCEDURE FOR
TECHNIQUES SOURCES OF
ERRORS
FIGURE 32. MILESTONE V - TEST AND EVALUATION REVIEW

V. APPLICATION OF THE RECOMMENDED PROCEDURE
Overview
It was recognized during the conduct of this program that all the de-
tailed steps for vibration design recommended in Chapter III have not been
applied to a single ship. Most designs use a portion of the individual
blocks, but not necessarily in the sequence presented in this report.
In order to validate the recommended procedures to some degree, the

original scope of work for this project was extended to include application
to a specific ship. Detailed design information is difficult to obtain,
but fortunately vibration analyses which follow closely the recommended
procedures have been conducted by Littleton Research and Engineering Corp.
(LR&EC), a corporate coauthor of this report. Section 2 presents the pro-
cedure as it is applied to a containerized and unitized cargo ship. This
ship was built and underwent extensive vibration measurements. Appendix G
contains a summary of vibration investigations which were part of the de-
sign of a proposed large roll-on/roll-off ship. This information is pre-
sented to show the depth of the studies which were made to minimize pro-
peller-induced vibration. In both of these examples, the readers will
immediately see that not all the blocks in the recommended procedures have
been applied. Thus, only a partial example of application of the recom-
mended design procedure could be given.
Procedure as Applied to a Containerized and Unitized Cargo Ship
2.1 General
To provide a comparison of the reliability of the recouaiiended

procedure, it was proposed to apply certain elements of the program approach
to a ship built in 1972 by Ingalls Shipbuilding Division, Litton Systems,
Inc., for Farrell Lines, Inc. The ship had a longitudinal vibration of the
propeller and shafting system which required correction. Because of these
difficulties, an unusually complete set of vibration readings were made on
the ship. LR&EC had made quite extensive vibration studies of the ship in
the design stage, but the studies lacked the completeness and perspective
of the program presented in this report. The studies included a rough fi-
nite-element analysis of the machinery space, which indicated a double bot-
tom mode. This mode when coupled with the shafting system represented the
offending vibration. However, under the pressures of time and finances,
this difficulty was not recognized at the time. The following sections up-
date certain portions of the original analyses and compare the results with
test measurements.
2.2 Ship Description
The important ship characteristics are listed below:
Overall length 668.5 ft

Length between perpendiculars 625.0 ft
76
Beam 90.0 ft
Depth to main deck 53.0 ft
Design draft 29.0 ft
Displacement (at 31' 5" draft) 28,440.0 tons
Number of containers 872
The ship is divided by transverse bulkheads into eight holds

forward of the machinery space and one long hold aft. The quarters and
bridge are located over the machinery space. There are wide hatches port
and starboard of a centerline girder in which the containers are stowed
vertically in guides. The containers are stowed six high in the holds
and two high over the hatch covers.
Longitudinal strength and some torsional stiffness of the hull

are provided by a heavy box girder 9 ft deep by 9 ft 9 in. wide extending
between the second and main decks. In addition, there are three heavy H
beams on each deck, and portions of the ship obtain torsional stiffness
from wing tanks. Extensive use is made of high-strength steel.
The ship is driven by a steam turbine plant with a normal horse-

power rating of 26,000 and an ABS rated horsepower of 28,500. The corres-
ponding propeller shaft speeds are 103 RPM and 106.5 RPM, respectively.
The overspeed trip on the turbines is 115 percent of the maximum rated
speed, which corresponds to 123 RPM for the propeller. The normal operat-
ing speed of the ship is 23 knots. Reference 105 contains detailed infor-
mation on this class of containership.
The propeller is 23 ft 6 in. in diameter and has six blades.

It is made of Superston 40, ABS Grade 5, whose tensile strength is 90,000
psi. The propeller shaft is supported by oil-lubricated bearings. The aft
bearing has a diameter of 29-3/8 in. and a length of 74 inches. The for-
ward bearing is 29-1/2 in. in diameter and 25-1/2 in. long. The distance
between the bearings is 24 ft 9 in., and the distance from the center of
the aft bearing to the centerline of the propeller is 6 ft 9 inches.
The choice of clearances between the propeller tip and the hull
is determined by a balance of several factors. From a structural point of
view, it is desirable to support the propeller with little overhang and
minimum extension from the hull. The hydrodynamicist would prefer to have
the propeller work in the high wake boundary layer region close to the hull
for good propulsive efficiency. For low levels of propeller-excited vibra-
tion, however, it is desirable to have large propeller clearances.
The in-
tensity of the pressure field around the propeller falls off rapidly with
distance, and the propeller works in a more uniform wake. Consequently,
smaller harmonic propeller and surface forces are generated.
In the preliminary design, the designer normally defines the

propeller location in terms of clearances between the propeller and adja-
cent structure expressed as percentages of the propeller diameter. For
this ship the clearance to the hull over the tip of the blade is about
3.21 ft or 13.6 percent of the propeller diameter. The distance from the
propeller leading edge at 0.7 radius to the strut is about 6.85 ft or 29
percent of the propeller diameter. The distance from the propeller trail-
ing edge at 0.7 radius to the rudder horn is 4.09 ft or 18 percent of the
77
propeller diameter. These are adequate clearances for the high power.
The propeller is located 5 in. forward of Frame 223, or 616.58 ft from the
forward perpendicular, or about 95 percent of the waterline length from
the forward perpendicular.
An outboard profile of the ship is shown in Figure 33 and a

sketch of the propeller aperture in Figure 34.
2.3. Application of the Procedures
The following paragraphs discuss the individual blocks of the

Recommended Design Procedures which are covered in this application. The
last series of digits refer to Figure 3, the Flow Diagram of Recommended
Design Procedures to Minimize Propeller-Induced Vibrations, on pages 6 and
7 of this report.
2.3.1. Define Vibration Specifications. No quantitative

specifications for acceptable level of vibrations were established in the
ship construction contract. The contract specifications for the construc-
tion of the ship contained vibration clauses from the Maritime Administra-
tion Standard Specifications for Cargo Ship Construction. In connection
with its analytical predictions, Littleton Research and Engineering Corp.
developed some criteria. In analyzing the test results, the International
Standards Organization ISO 2631, "Guide for the Evaluation of Human Expo-
sure to Whole Body Vibration," was used. The reader is referred to Refer-
ence 4 for later revision of the standard.
2.3.2. Establish General Ship Design Data. The more signifi-

cant characteristics of the ship were presented earlier. Sketches of the
machinery space and the superstructure will be presented in the descrip-
tion of the model that is analyzed. The analysis is made for the ship at
trial draft conditions in order to compare with measurements.
The drafts and displacements on trials were:
Draft Draft
Forward Aft T r im Ton s
5TV 19,442
Official Trials 19' 9" 26' 6' 8"
Standardization
20' 7" 26' 1/2" 5' 5-1/2" 19,261
Trials
2.3.3. Conduct Wake Survey. A wake survey was conducted by

DTNSRDC in March 1968. The longitudinal, tangential, and axial wakes at
0.335 R (radius), 0.520 R, 0.723 R, 0.950 R, and 1.100 R are presented as
Figures Fl - F5 in Appendix F. These wakes are for 25,035 tons displace-
ment and 23 knots.
2.3.4. Estimate Longitudinal Propulsion Frequencies. Figure

35 plots the relationship between natural frequency of longitudinal vi-
bration and the stiffness of the thrust bearing foundation. Kane and
78
A.I'.
FIGURE 33. OUTBOARD PROFILE
230 225 220
FIGURE 34. PROPELLER IN APERTURE
79
Thrustshaf t Thrust Collar
Prop. &
Entrained 40" L - 27" D
Water
112,800 lb AE = 17.08 x iO9 lbf Bull Gear
n p 178 lbnliri.
I Hub & Shaft
20,000 lb
in
Tailshaf t Lineshaft
- 29" D 982" - 22 .64 D
A.E 19.7 x io lbf AE = 12.0 x lO9 lbf K, Thrust K, Foun-

Shoes, dation
p1871b/in.
n
pl2Olb/in. in Collar
Cas e - Thrust
26.6 lbf/in Bearing
33,000 lb
n
(a) Propeller and Shaft Mass-Elastic System
15
z
n 13
oz
00
BLADE- RATE F EQUENCY FOR
z
10.6
0' SIX-BL ADED PR( PELLER
0 5 10 15 20 25 30 35
FOUNDATION STIFFNESS - io6 LBF/IN.
(b) Estimate of Natural Frequency for

Propeller Shafting System
FIGURE 35. FREQUENCY 0F LONGITUDINAL VIBRATION VERSUS

FOUNDATION STIFFNESS
80
McGoldrick 1131 give some approximate ranges for K, the foundation stiff-
ness, for a thrust bearing located just abaft the gear and tied into gear
foundation, as 5 to 20 x 106 lb/in. On the basis of this, the designers
were informed early in the design process that if a rigid foundation were
designed (this should not be too difficult because the thrust bearing is
located in a narrow region of the ship), there should be an adequate margin
between the natural frequency of the shafting system and the operating fre-
quency of a six-bladed propeller turning at 106.5 RPM. The foundation was
well designed and of heavy construction, so by the general criteria applied
at the time, longitudinal vibration trouble should not be expected. Why
the trouble occurred will be explained later. The prediction of the trouble
required additional analysis of the engine room.
2.3.5. Design Propeller. The propeller design was furnished

to the ship builder along with the lines plan and general arrangement of
the ship as part of the contract documents.
2.3.6. Compute Propeller Forces. The blade order harmonic

forces and moments were computed using the LR&EC lifting line computer pro-
gram. It was not possible to compute the twice-blades order forces and
moments because only the first ten harmonics of the wake were furnished.
Computations of twice-blade order forces and moments for a six-bladed pro-
peller require the eleventh, twelfth, and thirteenth orders.
The values of the blade order harmonic forces and moments, with
their phase angles, excited by the propeller on the shaft at 99 RPM are:
Fx (longitudinal) 12,963 lb 65.53°

Fy (transverse 4,210 lb _19.700
Fz (vertical) 1,185 lb _l28.5l0
Mx (about rotational axis) 40,373 ft-lb 63.47°
My (about transverse axis) 99,363 ft-lb -11.42°
Mz (about vertical axis) 37,210 ft-lb -106.25°
All forces and moments are referenced to the coordinate system shown in
Figure 34, and all moments are computed about the origin. The angles are
measured from the top of the propeller, with the positive values in the
direction of propeller rotation.
2.3.7. Compute Hull Pressures Without Cavitation. The hull

pressures and the resulting hull force were computed using the LR&EC pro-
gram. This program computes the free-field pressures generated by the dis-
tribution of forces along the lifting levers of the propeller blades and
the thickness of the propeller blade and multiplies the result by a factor,
generally 2, to represent the solid boundary.
Because at the trial conditions the draft aft was between 26

ft and 26.5 ft, the waterline runs out within the propeller aperture be-
tween the propeller and the rudder. Therefore, there is little hull sur-
face below water in the region of the propeller. The predicted hull forces
generated by the propeller at 99 RPM, if the propeller is not cavitating,
are:
81
Fx 3,284 lb -166°
2,430 lb 700
Fy
Fz 4,310 lb -180°
Mx 36,706 lb-ft -111°
My 28,470 lb-ft 128°
730
Mz 19,112 lb-ft 0
The following blocks are combined:
2.3.8. Evaluate Propeller Cavitation
2.3.9. Evaluate Propeller-Cavitation Factors
2.3.13. Compute Total Pressures and Forces
LR&EC has developed a computer program which predicts the pres-

sures on the hull of a ship generated by the lift forces and the thickness
of the propeller moving through the water. It does this by computing the
location relative to
pressure that would be generated in open water at the
the propeller that represents a point on the hull and multiplying this free-
This is an
field pressure by two to account for the rigid-hull surface.
deal with the complexities of the reflec-
approximation because it does not
effects.
tions off the hull surface and does not include cavitation
The pressure radiation from cavitation is much more efficient

of the
than that from the harmonic force loadings and from the passage
In cavitation there is a harmonic change in the
blade through the fluid.
simple source with the
volume of the cavity. This transmits pressure as a
given by the equation
pressure amplitude at large distances from the source
(see Reference 106, page 313)
p
p 4rrr
dt2
where
d2V the second time derivative of the volume

.
-
2
dt
p = the mass density of fluid
r = the radius from the source.
blade through water

The harmonic forces on the blade and the passage of the
and the pressure
do not change the volume, but only displace the water,
from a source and sink, a dipole. The equation
field corresponds to that
large distance
for the amplitude of the pressure for a harmonic force at a
from the dipole is (see Reference 106, page 318)
3Fvcose
p 2cr
82
where
p = pressure
F = force on the sphere in the direction of vibration
= frequency in cps
e = angle between a vector to the measuring point and the

force vector
c = the speed of sound in the fluid
r = distance between the dipole's center and location

of desired pressure
It is clear that the pressure from a harmonic force, as a source and sink
in general, is a maximum in the direction of the force and zero in a plane
normal to the force direction. On the other hand, the pressure from the
changing volume has no directional properties.
At the present timethe procedures for predicting the hull

pressure from a changing cavitation bubble are not well developed and are
complicated. For this analysis the loads from pressures on the hull sur-
face, required as input to the hull vibration calculation,are estimated
by comparing measured pressures on two ships with propellers at times work-
ing in the cavitation region, with the calculated pressures without cavi-
tation determined by the procedure described earlier. These two ships are
an OBO ship of the San Clemente Class built by National Steel and Ship-
building Corp. and a RO/RO ship of the Seabridge Class built by Ingalls
Shipbuilding Corporation [107].
The OBO ship has the following characteristics:
Full Power - 24,00 HP

Ship Speed - 17.15 knots
Propeller RPM - 92
Propeller Diameter - 26 ft
Propeller Projected Area - 265.04 ft2
Propeller Tip Submergence - 18' 10-1/2"
Propeller Thrust - 420,000 lb
Mean Inflow Velocity Ratio at 0.8 R - 0.643
Inflow Velocity Ratio at 0.8 R, 00 - 0.306
Pressure readings were taken under steady ahead operation at

60 RPM and at 92 RPM. At 60 RPM the average ratio of 11 measured pressures
to calculated pressures was 1.592 and the standard deviation of this ratio,
0.718. At 90 RPM the average ratio of 11 measured pressures to calculated
pressures was 3.053 and the standard deviation of this ratio, 1.435.
Measured pressures and information on the harmonic forces are

reported in Reference 107.
83
The Seabridge Class ship has the following characteristics [22J:
Full Power - 30,000 HP

Ship Speed - 24.59 knots
Propeller RPM - 112.9
Propeller Diameter - 23 ft
Propeller Projected Area - 268 ft2
Propeller Tip Submergence - 7.9 ft
Readings of the hull-surface pressure were made under steady

ahead operation at 110, 108, 101.2, 95.8, 92, 80, and 70 RPM [108]. At
the higher speeds there was considerable hull vibration, so the pressure
readings were corrected for water inertia effects by assuming the hull an
infinitely long cylindrical body having a width corresponding to the water-
line width in the plane of the pickup. The water correction is signifi-
cant, but, on further reflection, the amount of the correction is consid-
ered to overcompensate, and half as much compensation would be more rea-
sonable. The pressures measured at 70 RPM were so small as to be consid-
ered unreliable (maximum calculated pressure 74 lb/ft2 = about 0.5 psi,
minimum calculated pressure 4.3 lb/ft2 = 0.03 psi). The remaining results
are given in the following table.
TABLE 19. PRESSURE DATA AT 16 LOCATIONS

ON SEABRIDGE SHIP
Maximum Maximum
Measured Calc. * *
Pressures Pressures Max. Min.
Avg. Standard Deviation
Accel. 2 2
RPM Correct. (lb/ft ) (lb/ft ) Ratio Ratio Ratio o
110.0 No 432 183.7 9.44 0.92 3.073 2.139

110.0 Yes 585 183.7 14.77 1.43 5.813 3.473
108.0 No 432 177.1 10.39 1.02 3.447 2.250
108.0 Yes 637 177.1 22.14 1.71 5.688 4.160
101.2 No 245 154.9 7.69 0.82 2.313 1.646
101.2 Limited 255 154.9 7.69 1.07 2.445 1.601
95.8 No 208 138.8 5.31 0.78 1.965 1.184
92.0 No 196 128.0 4.53 0.41 1.682 1.057
80.0 No 99 96.8 22.08 0 0.941 --
*MaxiTnum (minimum) ratio of measured to calculated pressures
84
The characteristics of the ship being analyzed are:
Design Shaft Horsepower - 26,000 HP

Ship Speed at 19,500 Ton (Model Test) - 25.53 knots
Corresponding Propeller RPM (Model Test) - 106.7
Propeller Diameter - 23.5 ft
Propeller Projected Area - 243.2 ft2
Propeller Tip Submergence - 2.71 ft
To utilize the experimental data for the new ship, it is neces-

sary to develop some method of comparison. It is presumed thatin general,
the pressures integrated over the surface of the ship to gíve the total
harmonic force are adequate for the noncavitating condition. These har-
monic forces and moments can be multiplied by a factor to account for the
cavitation. This is a simplification since, as shown earlier, the dis-
tribution of pressures due to cavitation is different from that due to
blade forces and displacement.
This simplification is worse for pressures than for forces.

In general, the forces due to loading the propeller act axially and tan-
gentially on the propeller and would, therefore, tend to give small pres-
sures where the plane of the propeller intersects the hull surface. This
is the region of peak pressures because of proximity to the propeller.
The cavitatíon effects being transmitted without any directional bias
would give higher ratios of (pressures with cavitation)/(pressure without
cavitation) in the regions close to the plane of the propeller disc. The
hull forces, being an integration over the surface, average the variations.
It is further presumed that some of the pressure available be-

fore cavitation occurs will be used by the steady lift of the propeller.
The remaining pressure will be available for pressure fluctuations, and
if these fluctuations are greater than that remaining, cavitation will
occur. Because the object is correlation of experimental data, hopefully
the procedures need not be precise. For example, average pressures rather
than peak pressures are considered. Because cavitation is a tip phenome-
non, 0.8 R is taken as the propeller blade section for study. The sub-
mergence plus atmospheric pressure at the propeller's top on the center-
line at 0.8 R is determined (neglecting thermal effects on vapor pressure).
The thrust per square foot of projected area is obtained by dividing the
thrust by the projected area of the propeller. The difference is avail-
able for handling fluctuations in wake. The pressure corresponding to
the major fluctuation, the high wake at the top of the propeller disc, is
determined very roughly by determining the change in angle of attack tc
at this location. The value is
= [mean axial wake velocity - local wake velocity I x

wR0
8
= (V
mean ratio at 0.8R
-Vratio at 0.8R 00)xVship
83
If the slope of the coefficient of lift curve at zero angle of attack is
taken to be 2rr, the average pressure on the section is
p = 2naxpV
1 2
1
= x- p [(wR0 8)2 + axial
We note that V2 is negligible compared with (wR0 8)2. Therefore, the

ial
change in pressure is
LPressure = {(V .
mean ratio at 0.8R - V .

ratio at 0.8 R, O ) x Vship
1 64 RPM
x 1.688] 32.2
x_gb_x 0.4D
= 0.01118 x AVelocity ratio x V .
ship
x RPM X D
where Vhi is the ship's speed in knots and D is the propeller diameter in
feet.
When these relations are applied to the full-power condition

with the OBO and RO/RO ships, the values shown in Table 20 are obtained.
The pressure magnifying factors for the RO/RO and the OBO ship
are plotted against the cavitation factor in Figure 36. There is clearly
no agreement between the OBO curve and the RO/RO curve, and although several
hypotheses were tried in an attempt to find a correlation, there appeared
to be no way to obtain a common relation compatible with both.
For the present, the unitized and containerized ship being

studied is similar to the Seabridge RO/RO ship, and so the pressure magni-
fication as a function of cavitation factor for this ship is applied to the
calculated hull forces and moments for the ship being analyzed. The verti-
cal hull forces and transverse moments calculated at 99 RPM are
Force Phase Angle
Fz 4,310 lb _1800
My = 28,470 ft-lb 128°
These values are small because most of the hull is out of the water. By
taking the forces and moments proportional to the square of the RPM, values
at other RPM's can be calculated. These are given in Table 21, along with
the forces and moments obtained by applying the cavitation factor. The
values corrected for cavitation are plotted in Figure 37, and these values
will be applied to the model developed and discussed in Sections 2.3.15 and
2.3.18.
86
TABLE 20. COMPUTED CAVITATION FACTORS
FOR THE THREE SHIPS
Seabridge Ship
Submergence Pressure at O.8R 2767 lb/ft2
Submergence Pressure,
Steady Pressure Minus Steady Pressure IPressure due to
Cavitation Factor
RPM Thrust, lb/ft2 lb/ft2 Wake, lb/ft2 Col. 4/Col. 3
112.9 1552 1215 329 0.273
110 1473 1294 312 0.241
108 1420 1347 301 0.223
101.2 1247 1520 264 0.174
95.8 1117 1640 237 0.145
92 1031 1736 219 0.126
80 779 1988 165 0.083
San Clemente OBO
Submergence Pressure at 0.SR= 3490 lb/ft2
92 1452 2038 155 0.076

60 618 2872 66 0.023
Containerized and Unitized Cargo Ship
Submergence Pressure at 0.8R 2441 lb/ft2
106.7 1459 982 341 0.348

100 1282 1159 300 0.259
95 1157 1284 270 0.211
90 1038 1403 243 0.173
85 926 1515 217 0.143
80 820 1621 192 0.118
The following three blocks were not accomplished:
2.3.10. Direct Calculation of Cavitation Pressures

and Forces
2.3.11. Conduct Model Tests
2.3.12. Conduct Cavitation Tests
2.3.14. Determine Forced Longitudinal Response of Shafting.

Although normally this improvement upon the estimate made earlier (Block 4)
should be made in the design procedure, this calculation will be included with
the response of the machinery space and superstructure (Blocks 14 and 18).
87
5
O 0.05 0.10 0.15 0.20 0.25
CAVITATION FACTOR
FIGURE 36. RATIO 0F MEASURED HULL PRESSURES WHERE CAVITATION

EXISTS TO CALCULATED, NONCAVITATING PRESSURES
TABLE 21. HULL FORCES AND MOMENTS DUE TO

CAVITATION EFFECT
F (lb) N (ft-lb)
x y Cavitation M (ft-lb)
F (lb)
RPM (Calculated) (Calculated) Factor x y
106.7 5006 33,071 3.00 15,000 100,000

100.0 4398 29,048 4.20 18,470 122,000
95.0 3969 26,216 4.25 16,870 111,420
90.0 3562 23,530 2.37 8,442 55,760
85.0 3177 20,987 1.93 6,132 40,505
80.0 2814 18,591 1.54 4,334 28,630
88
120,000
100,000 20,000
80,000 16,000
E-' o
z o
o
60,000 12,000
o z
z
o
40,000 8,000
20,000 4,000
O O
30 40 50 60 70 80 90 100 110
PROPELLER RPH
FIGURE 37. VERTICAL HARMONIC FORCE A1'TD TRANSVERSE BENDING MOMENT

GENERATED ON HULL BY THE PROPELLER
2.3.16. Determine Forced Lateral Response of Shafting
(Rigid Hull)
and
2.3.17. Determine Forced Lateral Response of Shafting (Flexible
Hull). Although these estimates should be made in the course of the design
development, ín this demonstration example the step is bypassed (as it was
in the original work), but is included in Block 21.
2.3.15. Determine Forced Response of Machinery Space

and
2.3.18. Conduct Superstructure Modal Analysis. These two steps
in the development program are combined in a modal analysis and forced re-
sponse calculation, which includes the machinery space and superstructure
as one common substructure. In the course of a design development, it would
be desirable to analyze these two sections separately and to combine them as
substructures in the final finite-element model. However, in this example,
it was considered advisable in the interests of obtaining good correlation
with experimental results to combine both the machinery space and the super-
structure into one large substructure since the two would be used in a beam
model rather than a large finite-element model.
In the original analysis of the ship in 1969 and 1970, the ma-
chinery space was analyzed with a coarse STARDYNE model to give inputs as
sprung masses to the general beam model of the ship. This model extended
to the main deck, but did not include the propulsion system. The model
gave indications of the type of difficulty that was experienced, but these
were not recognized at the time. In more recent finite-element analyses,
the modeling has been much more realistic, but there has never been a good
opportunity to compare predictions with measurements on the ship. To sup-
port the belief that good finite-element analyses will predict potential
vibration problems, it was decided to make the analysis on this ship to the
standard that has been applied on other ships that have been analyzed in
recent years. To further assure this, it was decided to include the super-
structure with the machinery space model since they are closely coupled.
0f course, the propulsion system should be included. Since the water in-
ertia values that have been used in past analyses have always been question-
able and are a significant part of the total mass, it was decided to include
a new representation of water inertia recently conceived and not developed.
Becauseamong the possible computer programs for analyzing the structure,
the ANSYS computer program has an element that appeared to be directly
applicable to the introduction of water inertia under the new concept, and
because of the good experience of SwRI with ANSYS, it was decided to use
this program for the analysis. This meant developing a familiarity with
a new program instead of NASTRAN, which LR&EC has customarily used for large
structural vibration problems. It has subsequently been determined that
the water-inertia matrix could be introduced into the NASTRAN program.
After a long period of frustration, it was discovered that the ANSYS ini-
tial introduction of water inertia was in error so that the choice of com-
puter program system for this example was unfortunate. The error in the
water-inertia matrix was subsequently corrected, however.
90
In subsequent sections, the following topics will be discussed:
The Structural Model

The Treatment of Water Inertia
The Natural Frequencies and Mode Shapes
The Response to Axial Excitation at the Propeller
General Discussion of the Substructure Model
(a) The Structural Model
The section of the ship modeled in this analysis is the machinery

space bounded by its forward and aft bulkheads, which are assumed to be
fixed. Included in the model are the complete superstructure, which gen-
erally falls within the length bounds of the bulkheads, and the propeller
and shafting, which is free to move axially in its bearings and is coupled
to the machinery space at the thrust bearing. An inboard profile of the
region is presented in Figure 38. A typical section is shown in Figure 39,
and Figure 40 shows the after engine room bulkhead at Frame 181. These
sections contain the grid points used to define its structure and mass load-
ing. In the machinery space length extending from Frame 157 to Frame 181,
there are five (5) of these web frames, which with the bounding bulkheads
give seven (7) longitudinal subdivisions. There are nine (9) deck levels
in the machinery space and seven (7) in the superstructure. A total of 577
grid points are used to define the structure. These carry masses and are
connected by membranes (plates whose bending is neglected), spars (columns
whose bending is neglected), and beams. A total of 2595 elements, includ-
ing masses, were used. The computer program determines the mass of the
structure, but additions must be made for equipment, furnishings, electri-
cal and piping fittings, machinery, water inertia, etc.
The computer program assembles all of these structural elements and

computes their assembled stiffness and mass matrices, which have generally
at least three times as many elements as the number of grid points. These
matrices are difficult to analyze, so the stiffness-mass matrix is custom-
arily reduced to an equivalent dynamic matrix referred to grid points
assigned by the analyst as capable of representing the dynamic behavior
of the structure. For thís model,there are 72 dynamic-degrees-of-freedom.
The choices of degrees of freedom represent what the analyst expects to be
good locations to represent the structure and define the vibration patterns.
To reduce cost and complexity, he likes to keep the number small, but by
so doing he may fail to reflect some particular vibration pattern that he
did not anticipate. Because all the vibration that was strongly observable
in the machinery space was symmetrical, i.e., the same on both the port and
starboard sides, only the symmetrical vibrations have been considered in
this study. Good practice in studying shIp hull vibration would dictate
that both the symmetric and anti-symmetric vibrations be analyzed. Once a
model for symmetric vibrations has been set up, it is only necessary to
change boundary conditions, repeat the last calculation, and interpret the
results to cover the anti-symmetric vibrations. Anti-symmetric vibrations
are not as common as symmetric, but they are far from unusual.
91
FALSE STACK
\12911 ABV
RSE TOP
109'S" ABV k.
WHEELHOUSE
KING / BRIDGE DK
100,11" A.BV
POST
EMERG. GEN.
FAN RM
SUN DK
92'2" ABV k.
GAMES
LOUNGE CABIN DK
AREA
fi 83'S" ABV k.
1ST ASSI
ENGR
BOAT 0K
74'8' ABV k.
MACHINERY UTILITY
CASING UPPER DIC
65'll' ABV k.
GALLEY
RAISED NND
57'2" ABV k.
CARGO HOLD
3RD 0K WING TANK

33'O" ABV F.O. SERVICE
i
26'O" AUX. MACBY FLAT WING TANK

F.O. SERVICE
MACHINERY SPACE
ENGINE ROOM FLOOR 3' 6" ABV
TANK TOP 5'6" ABV
D.B. MACHY-STABILIZER DUMP TANKS
FR 181 FR 157
FIGURE 38. INBOARD PROFILE
92
69 210 342.9 456
112
316 423 House Side
239
1104 11109
164 '100.5
197 . 6
13.0
32.8
FIGURE 39. GRID POINTS ON FRANE 170,

X = 396 IN.
93
239 342.9 456
-
112 214 316 423 5
MAIN DECK 57'_2" 1001 1029 1042 ..1043 1057 10 72 1085
686
I i I
I f I
901 929 I I 1943 957 971
lilt h îî îî t
I
2nd DI< 44'
ill liii
985
528
II 1iIIi
i f300
I
t
529.8
i
l?1HIIlII
i I i i I
I t
I 829 843
I
871
¿6' FL 1857
801
-tH+H ±bH
f
885
432
487.1
265
481
2 I I
I I i I
I67 I I
I
26' FL 629
i
685
t i i i
f312 ia 366.1
20' STR 501 I i 529 I i 543
585
f240
i I I ? t I t
273.1
--
13'-8" STR
FOUND 201 215

o
485
'a 202.9
14
TANK T. 157
125 . 2
6
BASE L.
34.4
FIGURE 40. GRID POINTS ON FRANE 181,

X = O IN.
94
The model was loaded at the propeller with longitudinal excitations.
Because the hull substructure did not extend aft to the propeller, no hori-
zontal or vertical excitations on the propeller or hull are included in
this analysis.
(b) Water Inertia
One of the present serious deficiencies of the analysis of struc-

tures in water by use of finite elements is the treatment of the effec-
tive mass of the water moving with the structure. This same difficulty
plagued early efforts to predict the natural frequency of ships as beams
until F. M. Lewis and J. Lockwood Taylor developed procedures for includ-
ing the water mass.
In a structure, it is desirable to assign one mass to each grid

point in contact with the water. However, this is not enough because
motion at one location will generate pressures at adjacent locations.
One way of representing the water accurately is to establish a volume of
grid points in the water and use fluid elements to develop the pressures
in the water and on the ship from the vibratory motion of the ship. There
are a limited number of cases where this has been done,and the results are
good. However, the procedure adds complexity to a computer analysis that
is already large.
For this analysis a procedure for computing the mass matrix of the
water inertia over a number of grid points on the bottom was applied.
It is based upon the pressure generated on a flat surface (an extension
of the pressure on the surface of a sphere) by a small vibrating piston
located in the surface. The pressure is proportional to the area and the
vibratory acceleration of the exciting piston and inversely proportional
to the separation between the exciting piston and the location of pres-
sure measurement. Thus, the mutual force between small areas Am and An
CA A a
F=
mn
n mm
r
mn
where
C = a factor reflecting the density of the fluid and the

proportions of the pistons
A ,A = the areas associated with grid points m and n
m n
r = the minimum surface distance between grid point m and n
mn .
on the ship s hull

a = the acceleration of the piston m
m
CA An
Clearly the quantity has the units of mass, and so the force!
r
acceleration interaction between grid points on a surface bounding a fluid

can be represented by a matrix of mass terms.
95
There are corrections for the physical dimensions of the sending and
receiving pistons. Where only half a surface is modeled, the interaction
with the other half can be represented for symmetrical and antisymmetrical
motions by dummy grid points. The influence of free water surface can also
be represented by using a mirror of grid points above the water surface.
If the dimensions of the pistons are small compared to the wave

length of the transmitted vibration, the pressure wave is transmitted
as a spherical wave and there is no significant correction required for
the location of the receiving piston relative to the sending piston.
It was found that the mutual interaction between pistons is sig-

nificant for considerable distance such as for the length of the machine-
ry space. Where the distance is long, there can be a phase shift between
the force at the remote piston relative to the acceleration at the trans-
mitting piston. The finite-element model of the fluid normally would not
include this.
A paper is being prepared for submission to the Journal of Ship

Research covering the background and details of this water-inertia treat-
ment. In the machinery space model, it was used in its simplest form with
minimal corrections for piston dimensions relative to the separation of
the sending and receiving pistons.
(c) Natural Frequencies and Mode Shapes
The modal analysis of the stiffness and mass matrices represent-

ing the structure, when reduced to equivalent matrices having 72 degrees
of dynamic freedom, gave the following frequencies and general mode shapes.
Mode 1: Frequency = 1.9 Hz
The maximum displacement is a vertical motion on the 26-ft

flat on the centerline at Frame 165. The motion drops rapidly in adja-
cent points. This grid point is on top of a portable plate and should
not have been selected as a degree-of-freedom point. The selection was
made because it was known that on trials the 26-ft flat vibrated heavily.
Although the 26-ft flat is a light structure and vibrates

noticeably, this particular vibration mode is probably spurious and
arises because the weights on the 26-ft flat were distributed evenly
among the grid points rather than concentrated at the particular items
of equipment. It will be shown later that this particular mode is not
excited by the longitudinal force at the propeller.
The maximum displacement is still on the 26-ft flat and still

at Frame 165, but in this case at a grid point 214 in. (17 ft 10 in.) off
the center. The structure at this location is permanent, but light, and
would not be loaded as heavily as is done by making it a dynamic degree-
of-freedom. The frequency is therefore probably spurious.
96
Node 3: Frequency 6.5 Hz
The maximum displacement for this mode is still on the 26-ft

flat at Frame 165, 214 in. off the center. However, the vibration is
much more widespread with amplitudes in the superstructure, along the
lower engine room stringer, and axial motions in the shaft one-eighth
as large as the maximum amplitude. This is probably the lowest real
natural frequency.
The maximum displacement for this mode occurs in the lower

engine room stringer at Frames 174 and 177. Reference to the forced
vibration response to the axial force from the propeller indicates that
this mode is not strongly excited by the longítudinal propeller force.
However, it might be excited by vertical and longitudinal forces and
moments at the propellcr. The kingpost on the after side of the super-
structure appears to show fore and aft resonances at this frequency.
Mode 5: Frequency 8.7 Hz
The maximum amplitude of vibration is at the main control

console on the 26-ft flat. There also appears to be some motion on the
36-ft flat in the same general region.
The maximum motion at this frequency is a lateral motion of

the kingposts This does not appear to be strongly excited by the axial
propelr force, but might be excited by lateral and vertical forces.
The maximum motion associated with this mode occurs on the

36-ft flat, Frame 170, centerline at the main control console. There
is also significant motion in the framing in the lower engine room at
Frame 170. This mode is quite strongly excited by the axial propeller
force.
This is the primary mode of the whole structure vibrating

vertically against its restraints at the forward and aft bounding bulk-
heads. The motion is accentuated at the 36-ft flat and in the kingposts.
It includes bending in the side frames of the machinery space. As will
be shown later, this mode is excited by the axial propeller force. It
might be more strongly excited by vertical forces and moments about a
transverse axis at the propeller and on the hull.
The location of the point of maximum motion is again on the

36-ft flat, Frame 170, centerline. The motion is vertical and although
97
the relative motion of the 36-ft flat is larger than for the previous
natural frequency (Node 8), the remaining motions in the ship are smaller.
This is the primary mode of interest in which the propeller

and shafting vibrate against the rest of the structure. This mode is
strongly excited by the propeller.
The mode shows the superstructure vibrating against the lower

ship side with small shaft motion.
Higher frequency modes are above 15 Hz and outside the range

of interest. These modes will be discussed further in section (e) after
consideration of the response of the structure to forced vibration from
the propeller.
Response of Substructure to Axial Harmonic Forces at Propeller
Using the substructure as modeled by 72 dynamic degrees-of-freedom,

the response to an axial force at the propeller was determined. To avoid
confusion from spurious resonances, a moderate amount of hysteretic damp-
ing was introduced (3 percent). Some of the results are plotted in Fig-
ures 41 through 47.
General Discussion of Substructure Model
In these calculations, the substructure is taken to be supported

rigidly at the forward and aft engine room bulkheads, which are Frames
157 and 181. Note from Figure 38 that the bulkhead at Frame 181 does
not extend to the tank top. Because the substructure in reality is sup-
ported by the elastic ship rather than the assumed rigid bulkheads, the
natural frequencies in which the whole structure participates will prob-
ably be lower than predicted by these calculations. Also because of the
flow of vibratory energy from the machinery space into adjacent struc-
ture (and also incomplete damping), the predicted amplitude of motions
arising from the longitudinal excitation from the propeller will be high.
Therefore, Figures 41-47 indicate relative vibration amplitudes throughout
the structure and are not the amplitudes predicted for the complete ship.
The excitation from the propeller and on the hull in the vertical and hori-
zontal planes is not considered since it acts outside the substructure.
Many modes which are weakly coupled to the propulsion shaft system will be
more important than the vibration excited by the axial forces at the pro-
peller. All of these effects are included and combined with the complete
ship in Blocks 21 and 22.
The vibration analyst would probably interpret the information that

has been generated as follows:
(1) There is a strong longitudinal vibration in the shafting sys-

tem. Its frequency at 12.4 Hz is about 17 percent above the maximum pro-
peller excitation at 10.6 Hz. However, it is a strong vibration with
98
I I I I
7 8 9 10 11 12 13 14 15
FREQUENCY, HZ
I
I I I I I I I I
J
60 70 80 90 100 110 120 130 140 150

PROPELLER, RPM
FIGURE 41. AMPLITUDE OF AXIAL MOTION AT THE PROPELLER

loo
Hl
80
AXIAL VIBRATION
OF THRUST COLLAR
60
40
20
FORE & AFT
VIBRATION OF
THRUST BEARING FOOT
6 7 8 9 10 11 12 13 14 15
FREQUENCY, HZ
I I I I I I
60 70 80 90 lOO 110 120 130 140 150
PROPELLER, RPM
FIGURE 42. AMPLITUDE OF FORE ANI AFT VIBRATION AT PROPULSION SHAFT

THRUST COLLAR AND THRUST BEARING FOOT
16
VERTICAL VIBRATION
FORE & AFT VIBRATION

12
/
i
- -- r- - i
6 7 8 9 lO 11 12 13 14 15
FREQUENCY, HZ
I I I I I
60 70 80 90 lOO 110 120 130 140 150

PROPELLER, RPM
FIGURE 43. VIBRATORY MOTION ON BRIDGE, FRAÌ 164 AT CENTERLINE

FRANE 170 Q , MULTIPLY ORDINATE BY lo
H
FRANE 165 Q , MULTIPLY ORDINATE BY 10
FRANE 170-210" (17.5') OFF Q , MULTIPLY ORDINATE BY
-- FRANE 165-276" (23') OFF Q , MULTIPLY ORDINATE BY 1
ZO
rIDO
HO
H
O
H
9 lO li 12 13 14 15
FREQUENCY, HZ
I I I I I I I I
I I
60 70 80 90 100 110 120 130 140 150
PROPELLER, RPM
FIGURE 44. VERTICAL VIBRATION ON 36-FT FLAT GENERATED

BY AXIAL PROPELLER FORCE
AT FRAME 170 - Q
AT FRANE 170 - 112" (9 '-4") OFF
Q
AT FRAME 165 - 214" (17'-lO") OFF (j
z
o
H 6
H
o 4
'-s
H
H
Q z
CI)
2
H
H
H
7 8 9 10 11 12 13 14 15
FREQUENCY, HZ
I I I I I I
60 70 80 90 100 110 120 130 140 150

PROPELLER, RPM
FIGURE 45. VERTICAL VIBRATION ON 26-FT FLAT EXCITED BY

AXIAL HARMONIC FORCE AT THE PROPELLER
50
40
AT THE CENTER LINE
30
THE SHELL
20 INTERSECTION
IAT
lo
I i i i i I i i i i I i i i i t I i i I
180 175 170 165 160

FRAME NUMBER
FIGURE 46. VERTICAL VIBRATION IN TANK TOP DUE TO LONGITUDINAL

EXCITATION AT THE PROPELLER AT 12.43 HZ
(Frequency of Longitudinal Shaft Resonance)
il
lo
AT THE CENTERLINE
/ AT THE MARGIN
(SHELL INTERSECTION)
T
\
I i i i i I i i i i I i i i I I i i i I i
's
i
180 175 170 165 160

FRANE NUMBER
FIGURE 47. VERTICAL VIBRATION IN TANK TOP DUE TO LONGITUDINAL

EXCITATION AT THE PROPELLER AT 10.8 HZ
(Frequency of Superstructure Resonance)
broad flanks, and its frequency will drop when introduced into the elastic
ship. It is lower than would have been expected from the preliminary cal-
culation of Block 4, which would índicate that the foundation structurehas
too low a stiffness. This is borne out by Figure 46, which indicates a
large slope in the deflection curve of the bottom. It would appear that
an effective way of improving the stiffness would be to make the transverse
structure between Frames 177 and 181 more rigid.
The natural frequency of the whole engine room and superstruc-

ture, Figure 43, falls at just about maximum speed. This frequency will
drop somewhat when incorporated in the elastic ship, but even so can be
expected to give some vibration difficulties. The amplitudes generated
by the longitudinal propeller excitation will probably be acceptable, but
this mode would be more strongly excited by propeller and hull force and
moments acting in the vertical plane. Raising the frequency above the
operating speed may be difficult when one recognizes that the predicted
value at this stage will drop when the substructure is incorporated in the
elastic hull. If raising the frequency appears desirable, it should be
accomplished by increased transverse stiffness at about Frame 165. See
Figure 47. The possibility of deepening the transverse bulkhead between
the 36-ft flat and the second deck to between the second and main decks
might be explored. It might also be well to make the web frame at Frame
165 heavier.
The natural frequency and mode-shape determinations show large

motions on the 26-ft and 36-ft flats. In particular, the 26-ft flat that
is the location of the Engine Control Console is a region where freedom
from vibration is critical. It would be advisable to improve the finite-
element model in these regions and to give some consideration to struc-
turai modifications to lessen the vibrations.
2.3.19. Determine Natural Frequencies and Forced Response of

Rudder. This study was not made in the original work and is not done in
this program. Generally rudders, even when in the air, have a low trans-
verse natural frequency and thus generally have little influence on the
hull vibration.
2.3.20. Evaluate Local Plating Design. This important aspect

of good vibration analysis has not been treated in this program. This
aspect of the vibration analysis can be handled most realistically if per-
formed after the general vibration level in different regions of the ship
is determined in Block 23.
2.3.21. Assemble Model of Entire Ship. It was the intention

of this program to carry the analysis through to a prediction of vibra-
tion in the ship and to compare the predictions with measurements. The
cost and time of running the finite-element analyses of the Machinery
Space and Superstructure have precluded this.
Under this itemthe finite-element analysis of the machinery

space and si'erstructure would be represented as a subsystem, System 4
in the ori,ínal ship model as shown in Figure 13. This subsystem would
be a beam with several sprung masses and would include the shafting sys-
tem and longitudinal excitation at the propeller. With this System 4
106
included, the whole ship would be properly represented as a damped, mass-
elastic system excited by five components of propeller force and moment
(torsional moments about the rotational axis are omitted) and six compo-
nents of hull force and moment.
Although this ship is too complex and the excitation frequency

too high to give good results from a beam model, this model was an im-
provement on the procedure in general use when it was developed in 1970.
It is expected that the results would be reasonably confirmed by the trial
measurements since the critical regions of the machinery space and super-
structure are accurately modeled. The connecting ship structure transmits
forces and motions to this region and dynamically interacts at the boun-
daries. The assumption is that the accuracy of these boundary reactions
will not too strongly influence the response in the critical regions.
2.3.22. Determine Vibration Amplitudes and Stress Levels of

Complete Ship. As discussed under Section 2.3.21, it has not been possible
to accomplish this.
2.3.23. Conduct Shaker Tests. A shaker test was made to de-

termine the response at the thrust bearing without the propeller and shaft-
ing connected. Although this test could be used to verify the analysis
procedures, the data obtained were limited, and a modified substructural
model would be necessary to make the analytical study for comparison with
the test data. These tests will be considered under Block 27.
2.3.24. Assess Local Vibrations, Structural Damping, and

Modeling Techniques. No work has been done in this design block.
2.3.25. Measure Vibrations During Sea Trials

and
2.3.27. Compare Measured Vibration with Calculations. During
sea trials, the ship was observed to have heavy vibrations of the following
types and locations:
Longitudinal vibration of shafting and thrust bearing
Vertical vibration on the 26-ft and 36-ft flats
Vertical vibrations in the cabin deck lounge over

the machinery casing
Vertical vibration on the bridge.
Graphs of the data obtained in exploring these vibrations are

presented in Figures 48 through 53. Figure 48 shows that the peak ampli-
tude of fore and aft vibration had probably not been reached at 102 RPM.
In Figure 49, the twice-blade frequency fore and aft vibration is plotted
in an effort to determine the natural frequency of longitudinal vibration.
The results are not very conclusive since the excitation at the low speed
required to define the resonance is low. However, the combination of
amplitude and phase shift (the phase is with respect to an arbitrary angle)
107
40 0 308°
o RECORD B (MAIN DECK AFT, FULL POWER)
0 304°
35 - ¿, ROLL 2, SHEET i (STANDARDIZATION)
D ROLL 2, SHEET 2 (VARIABLE SPEED)

<> ROLL 1 (VARIABLE SPEED)
OF PEAK TO PEAK VIBRATION
30 TRANCE
- ROLL 2, SHEET 2
282°
25
267°
1284: 278°
20 300°
249°
15
10
5 -
JI"
0242°
2OO°
o
236° 234°
29O0D
70° 180° 207°
O
IA I i i
80 90 100 110
40 50 60 70
RPM
FIGURE 48. DOUBLE AMPLITUDE OF SIXTH ORDER,

FORE AND AFT NOTION OF ThRUST BEARING
FOUNDATION
108
I I I I E I
O RECORD B (MAIN DECK AFT, F!JLL POWER)

2.8
á ROLL 2, SHEET i (STANDARDIZATION)
ri) D ROLL 2, SHEET 2 (VARIABLE SPEED)

z ROLL 1 (VARIABLE SPEED)
2.4
zo
H
c 55°
Q 66°
. 2.0
0 16°
o
1.6
'-I
o
0 86° o 348°
1.2 0 23°
¡, 70°
0.8 354° -
39°
347° 0042°
143° D
339°
¿ 261° 51° -
0.4
, 218°
o
40 50 60 70 80 90 100 110
PROPELLER, RPM
FIGURE 49. DOUBLE AMPLITUDE OF TWELFTH ORDER,

FORE AND AFT VIBRATION ON THRUST BEARING FOOT
109
might be construed to indicate a resonance at about 75 RPM or 15 Hz, or
possibly at 14 Hz since the excitation force increases with the square
of the RPM. Figures 50 and 51 confirm the vibration on the 26-ft and 36-
ft flats at the frequency of the primary structural resonance at about
10 Hz. This is reduced from 10.8 to 10 Hz by the flexibility of the sup-
ports. Finally, Figures 52 and 53 also show the primary vertical struc-
tural resonance of about 10 Hz and the motion on the bridge, which is less
than on the 26- and 36-ft flats.
On a subsequent ship, a shaker was mounted on a thrust-bearing

housing roughly bolted to its footing. The shafting was not connected.
The ship was incomplete, but had most of its machinery in place. It was
located in the fitting-out dock with little water under the keel. The
model patterns along the bottom at a number of frequencies are shown
in Figures 54 and 55. In Fígure 56 the experimentally determined responses
on the bridge (at Frame 164) and on the bottom at Frame 163 are shown.
The tests were run in the evening at the fitting-out dock under pressure
to complete the ship. It would have been helpful in interpreting the data
if the shaker had a circuit for providing a phase mark and if a few more
readings had been taken.
It appears that the fundamental 10.6 Hz superstructure reso-

nance is moved down to about lO Hz by the larger water inertia of the
shallow drafts. Since it is not strongly coupled with the shafting, its
omission is not strongly felt. On the other hand, the resonance which
appears at about 19.5 Hz probably corresponds in the bottom and the super-
structure to the 12.4 Hz longitudinal shaft vibration mode without the
shaft weight present.
2.3.26. Compare Measured Vibrations with Specifications.

Although the measured vibrations were compared to the ISO whole body vi-
bration standards, this comparison is not important to this evaluation
of the computation procedures.
3. Procedure as Applied to a Large RO/RO Ship Design
A portion of the procedure outlined in this report was applied to

the analysis of the vibration of a triple screw roll-on/roll-off trailer
ship designed by Sun Shipbuilding and Drydock Company. The studies were
made by Littleton Research and Engineering Corp., working closely with
Mr. Hector NcVey, project manager, and Charles Lof ft. The program was an
outstanding example of a good mutual interaction between ship vibration
analysis and ship design. The ship's dimensions are 945-ft overall length,
875-ft length between perpendiculars, 105.5-ft beam, 76.25-ft depth, and
30-ft design draft. It is driven by a center and two wing propellers,
each absorbing 50,000 rated horsepower at 155 RPM. Construction of the
ship has not been initiated, and so comparisons between predictions and
measurements are not possible. The correlation between the sequence of
studies recommended earlier in this report and those applied to the vibra-
tion studies on this ship will be recognized by the brief summary of the
reports generated in the course of the work. A listing of these reports
is contained in Appendix G.
110
+PORT
+VERT
t/ ' +D
TOILET
+ VECTOR PHASE 00 ± 900
- VECTOR PHASE 1800 ± 900
.DN AMPLITUDE IN MILS (0.001 INCH)
.' - O
' F°
26'-O" FLAT
I i i I I I i i i i i i I I t Iq"I I I i
180 175 i7Q ''.' IØo 160
4.. '-'
DN
uP
4, I0
FIGURE 50. VIBRATION ON 26-FT FLAT AT ABOUT 100 RPM

+PORT
+VERT
DISTILLED
WATER
\ TANK __
X
/
ON + VECTOR PHASE 00 ± 900
OT / 4
e WAT E R j
u-
VECTOR PHASE 1800 ± 900
\TANK
o z AHPLITUDE IN MILS (0.001 INCH)
I-
(80 75 160
'6P0T Z
WATER
°-
MACHINERY o TANK
X - 36_OT FLAT
/1 SPACE \\
/
ON
/
I
o
DISTILLED
WATER
\)AN K
FIGURE 51, VIBRATION ON 36FT FLAT AT ABOUT 100 RPM

+PORT
+VERT
- +FWD
120°
+ VECTOR PHASE 0° ± 900

- VECTOR PHASE 180° ± 90°
5.8
AMPLITUDE IN MILS (0.001 INCH)
d3.2 DQ
180° 0
I I
leo
IIL...Il
16
11.2
158°
8.8 oI
D 130°
12.5
1500
4.2
100° 53I6.I151e4I/
30°
3.9
127°
j
FIGURE 52. VIBRATION OF BRIDGE DECK AT ABOUT 98.6 RPM
+PORT
+VERT
k +FWD
6.1
1750
+ VECTOR PHASE 00 ± 900

- VECTOR PHASE 1800 900
±
AMPLITUDE IN MILS (0.001 INCH)
17.5
202° w
0,
110 2.0 DO
2020 210° 0
I I I II I I i
I i I III 16
l'i 160
180
lII) -ITO-
6.8y '.5 11.3
180° 225° 202°
I I 9.6 oJ
U 202°
218°
I6.9
5.O
950 8.29t9/
188° 202°
9.7
2020
5.2
80° 72°
I
FIGURE 53. VIBRATION OF BRIDGE DECK AT ABOUT 102 RPM
0.30
SHAKER MOUNTED ON THRUST HOUSING
SHAFTING NOT CONNECTED
z SHALLOW WATER AT FITTING OUT DOCK
0.20
Q
C)
Q 11HZ
/
0.10
z
r.
u, l.
z
I-4 FR .. FR FR
185 . 180 175
FR
170
/
I/I FR FR
165
.
._
-0.10 . ._
- -
-0.20 - 12.5 HZ 12 HZ
10 HZ
FIGURE 54. EXPERIMENTAL DEFLECTION PATTERNS ON TANK TOP

AT AT 10, 11, 12, AND 12.5 HERTZ
O . 30 SHAKER MOUNTED ON THRUST H0US1N(
SHAFIING NOT CONNECTED
SHALLOW WATER AT FITTING OUT DOCK
0.20
16 HZ
12.9 HZ
0.10
I I I I I I
I I I
0
I'I
FR FR FR FR FR FR
185 180 .. 175 170 165 160
-0.10
.
14 H Z
-0.20
-0.30
FIGURE 55. EXPERIMENTAL DEFLECTION PATTERNS ON TANK TOP

AT AT 12.9, 14, AND 16 HERTZ
ON BRIDGE AT ABOUT FRANE 164 (j
AT FOOT OF STANCHION FRAME 163 (j
SHAKER MOUNTED ON THRUST BRIDGE HOUSING
0.5 SHAFTING NOT CONNECTED
SHIP AT FITTING OUT DOCK
SHALLOW WATER
RESPONSE ON
0.4 BRIDGE
0.3
0.2
0.1
/RESPONSE AT
i FR 163 (j
I I I I I I I
I
10 12 14 16 18 20 22 24 26
FREQUENCY, HZ
FIGURE 56. RESPONSE TO SHAKER EXCITATION

Discussion of Example Problem
In this chapter. it has been demonstrated that,with regard to the

design of the Litton Industries, Ingalls Shipbuilding Division United
and Containerized Cargo Ship built for Farrell Lines, application of the
finite-analysis computer techniques now available would have forewarned
of the following vibration difficulties experienced on the ship.
Longitudinal vibration of the propeller-shafting system. The

analysis of the machinery space-superstructure subsystem indicated a
natural frequency of 12.4 Hz. The frequency could be expected to fall
when the subsystem is incorporated with the remaining ship structure and
is too close to the maximum operating speed. The difficulty occurs be-
cause of coupling between the engine room double bottom modes and the
vibration of the shafting system against the thrust bearing and founda-
tion and would probably not be discovered by a static analysis.
The fundamental frequency of the machinery space and superstruc-

ture substructure in vertical vibration is 10.6 Hz. This frequency can
also be expected to decrease when incorporated with the remaining hull
structure. It coincides almost exactly with the blade frequency at full
power and explains the heavy vibration on the 26- and 36-ft flats in the
engine room, the vibration in the wheelhouse, and that in the passenger
lounge.
The general vertical vibration levels in the ship are accentu-

ated on the 26-ft and 36-ft machinery space flats. This was observed on
trials. Since the main machinery control console is located on the 26-ft
flat, this is a particularly bad location for vibration.
Time and money limitations did not permit the substructure predic-
tions to be refined and quantified by being incorporated into a structure
model of the complete ship.
A listing of the reports presented to Messrs. McVey and Lofft of

Sun Shipbuilding and Dry Dock Company in the course of the design devel-
opment of a large RO/RO ship demonstrates how a portion of the design
procedure incorporating vibration analysis presented in this study was
applied to a highpowered ship in which vibrations might be critical.
Unfortunately, construction of this ship has not been initiated, so the
accuracy of the procedures cannot be assessed for this case.
Although the analysis of the machinery space and superstructure on

the Farrell Ship incorporates the best present techniques for predicting
vibrations, its proposed combination with the remainder of the hull does
not utilize the best presently available representations. It would he
desirable to complete the study with continuing emphasis upon making the
prediction as accurate as practically possible. All of the Farrell ships
built to this design have been increased in size, and so it is not pos-
sible to make additional vibration measurements applicable to the initial
studies.
118
Detailed calculations compared with detailed measurements under
operating conditions and with shakers, comparable to the program carried
on by McColdrick at DTMB in the late 1940's and early 1950's, would appear
desirable for developing confidence in vibration-prediction processes.
119
VI. CONCLUSIONS AND RECOMMENDATIONS
1. Conclusions
This report has presented a recommended design procedure for the

minimization of propeller-induced vibrations in hull structural elements.
The procedure is general in that it is not dependent upon the ship's size,
type, or service conditions, and is intended to mesh with the overall
ship's design process. The recommended procedure begins when the vibra-
tion specifications are defined and continues until after the vibration
levels are measured during the ship's sea trials.
As presented, this recommended procedure is quite comprehensive and

would be expensive to apply in its entirety. Therefore, the designer may
choose to combine or eliminate some of the steps for the sake of economy.
Such simplification can only be employed with great experience, and will
be accompanied by increased risk of vibration problems occurring in the
final design. A trade-off risk versus savings naturally results. Each
designer must decide what part of this trade-off he is willing to accept.
Many conclusions pertaining to specialized parts of the design pro-

cesses are contained in this report. Rather than reiterate them at this
time, it is felt that the following list of more general conclusions re-
garding the entire procedure would be more beneficial to the reader:
The technical state-of-the-art appears to be suf-

ficiently advanced so that ships can be designed to
have acceptable vibration levels.
Present-day ships are complex mechanical systems

whose successful design depends on extensive use of
current analytical and experimental analysis methods.
It is highly unlikely that "cookbook" type of design

procedures can be developed and used successfully by
the general maritime industry.
The design process includes many specialized areas

and requires the interaction of many individuals.
It should be approached from a systems point-of-view,
with equal emphasis given to propeller excitation and
structural response.
Any rational design procedure must, as a minimum,

contain the following:
A set of design specifications
A method of predicting or measuring propeller-

induced forces and pressures
A method of computIng structural response
120
A procedure for measuring actual vibration levels
during sea trials.
2. Recommendations
Although the set of general conclusions stated that ships can be

currently designed to have acceptable vibration levels, there are certain
areas which require additional research if the recommended procedures are
to be improved. They are
A study should be initiated to establish the relation-

ships between the ship stern form to the wake and its
variation, and the wake and wake variations to the
propeller forces, hull pressure forces, and propeller
cavitation.
A study should be made to more fully understand the

effects of a working propeller on the nominal wake
distribution.
Efforts should be increased to refine analytical

methods to predict hull pressures caused by cavi-
tation.
A study should be initiated to establish guidelines

for the acceptability of propeller-generated forces
and pressures based upon gross ship properties.
Additional work is needed in the area of propeller-

rudder interaction. This includes improved prediction
techniques for both the propeller-generated forces and
the response of the rudder.
It would be desirable to apply the procedures presented

herein to a ship that has been built and is operating.
The designer of a ship should use discretion in the application of

this procedure. This discretion should involve a review of each item of
the process as applied to the ship. For a ship of low power, where vi-
bration is rarely a problem unless there is a serious neglect in over-
looking important factors that generate vibration, this review should
give consideration to propeller clearance and the flow of water to the
propeller, adequate support for the thrust bearing and reasonable support
for outboard shafting. In ships that are similar to vibration-free ships
already in operation, the items in the procedure should be checked be-
tween the ships.
In ships of moderate power, say less than 25,000 shaft horsepower,

and of a design generally similar to other vibration-free ships, it may
be adequate to check propeller clearance, to ensure that the ship has
good lines for free flow to the upper portion of the propeller, that
the thrust bearing appears to be well supported, and that the super-
structure has good structural continuity.
121
With powers rísing above 30,000 shaft horsepower and with the
open structure characteristics of some of the recent roll-on/roll-off
ships, it is probably desirable to run through the complete program.
For balance between the different elements of the procedure, it is
probably equally important for a vibration-free ship to have good
structural studies as good hydrodynamic studies. If model studies
include measurements of the pressures on the hull from cavitating
propellers, then an equivalent degree of structural vibration analysis
in the form of a study of the dynamic response of the machinery space
and of the superstructure will be warranted.
The recommendations by the British Ship Research Association of

checking the vibration response of a ship under excitation by a shaker
at the stage where the structure is completed, but before the joiner
work is installed, would appear to have some value for catching local
resonances while it is inexpensive to correct them.
122
REFERENCES
Burnside, O. H., Kana, D. D., and Reed, F. E., "Bibliography for

the Study of Propeller-Induced Vibration in Hull Structural Ele-
ments," SSC-28l, Final Technical Report on Project SR-240, Depart-
ment of Transportation, U. S. Coast Guard Contract No. DOT-CG-
61907-A, July 1978.
Hagen, A., and Hammer, N. O., "Shipboard Noise and Vibration from
a Habitability Viewpoint," Narine Technology, Vol. 6, No. 1, Janu-
ary 1969.
Reed, F. E., "Acceptable Levels of Vibration on Ships," Marine

Technology, Vol. lO, No. 2, April 1973.
ISO Standard 2631, "Guide for the Evaluation of Human Exposure to

Whole-Body Vibration," International Organization for Standardiza-
tion, 1974.
The Society of Naval Architects and Marine Engineers, "Acceptable

Vibration of Marine Steam and Heavy-Duty Gas-Turbine Main and
Auxiliary Machinery Plants," Technical and Research Code C-5,
September 1976.
The Society of Naval Architects and Marine Engineers, "Code for

Shipboard Hull Vibration Measurements," Technical and Research
Code C-1, January 1975.
The Society of Naval Architects and Marine Engineers, "Code for

Local Shipboard Structures and Machinery Vibration Measurements,"
Technical and Research Code C-4, December 1976.
Van Oossanen, P., and Oosterveld, N. W. C., "Application of Theory

in the Optimization of Single Screw Container Ship," presented at
the International Symposium on Practical Design in Shipbuilding
(PRADS), Tokyo, Japan, October 1977.
Hoekstra, M., "Prediction of Full Scale Wake Characteristics Based

on Model Wake Surveys," International Shipbuilding Progress, Vol.
22, No. 250, June 1975.
Odabasi, A. Y., and Fitzsimmons, P. A., "Alternative Methods for

Wake Quality Assessment," International Shipbuilding Progress,
Vol. 25, No. 282, February 1978.
il. Van Gent, W., and van Oossanen, P., "Influence of Wake and Pro-
peller Loading and Cavitation," Second LIPS Propeller Symposium,
Drunen, Holland, Nay 1973.
12. The Society of Naval Architects and Marine Engineers, "Longitudi-

nal Stiffness of Nain Thrust Bearing Foundations," Technical and
Research Bulletin R-15, September 1972.
123
Kane, J. R., and McGoldrick, R. T., "Longitudinal Vibration of
Marine Propulsion-Shafting System," Transactions, SNANE, Vol. 57,
pp 195-252, 1949.
Mott, I. K., and Fleeting, R., "Design Aspects of Marine Propulsion

Shafting Systems," Transactions, Institution of Marine Engineers,
June 1967.
Lewis, F. M., and Auslaender, J.., "Virtual Inertia of Propellers,

Journal of Ship Research, Vol. 3, No. 4, March 1960.
Reed, F. E., and Burnside, 0. H., "Computer Techniques for Use in

Ship Hull Vibration Analysis and Design," Paper K, SSC and SNAME
Ship Vibration Symposium '78, Arlington, Virginia, October 1978.
Saunders, H. E.., Hydrodynamics in Ship Design, SNANE, New York, 1957.
Kerwin, J. E., "Computer Techniques for Propeller Blade Section

Design," International Shipbuilding Progress, Vol. 20, No. 227,
July 1973.
Van Gunsteren, L. A., and Pronk, C., "Propeller Design Concepts,"

International Shipbuilding Progress, Vol. 20, No. 227, July 1973.
Cox, C. G., and Morgan, W. B., "The Use of Theory in Propeller

Design," Marine Technology, Vol. 9, No. 4, October 1972.
Cummings, R. A., Morgan, W. B., and Boswell, R. J., "Highly Skewed

Propellers," Transactions, SNAME, Vol. 80, pp 98-135, 1972.
Boswell, R. J., and Cox, C. C., "Design and Model Evaluation of a

Highly Skewed Propeller for a Cargo Ship," Marine Technology, Vol.
il, No. 1, January 1974.
Lewis, F. M., "Propeller Vibration," Transactions, SNAME, Vols. 43

and 44, 1935-1936.
Burrill, L. C., "Calculation of Marine Propeller Characteristics,"

Northeast Coast Institute of Engineers and Shipbuilders, Vol. 60,
1943-1944.
Hinterthon, W. B., "Propeller-Excited Vibratory Forces for the Tanker

SS ESSO Gettysburg Calculated from Wake Surveys," Naval Ship Research
and Development Report 2870, August 1968.
Reed, F. E., and Bradshaw, R. T., "Ship Hull Vibrations II. Dis-
tribution of Exciting Forces Generated by Propellers," CONESCO
Report to Bureau of Ships, F-101-2, Index No. NS712-100ST2, June
1960.
Tsakonas, S., and Jacobs, W. R., "Documentation of a Computer Pro-

gram for the Pressure Distribution Forces and Moments on Ship Pro-
pellers in Hull Wakes," Stevens Institute of Technology, Davidson
Laboratory Report SIT-DL-76-l863.
124
Brown, N. A., "Periodic Propeller Forces in Non-Uniform Flow,"
Massachusetts Institute of Technology, Department of Naval Archi-
tecture and Marine Engineering Report 64-7, 1964.
Frydelund, O., and Kerwin, J. E., "The Development of Numerical

Methods for the Computation of Unsteady Propeller Forces," Sympo-
sium on Hydrodynamics of Ship and Offshore Propulsion Systems, Det
norske Ventas, Oslo, Norway, March 1977.
Murray, M. T., and Tubby, J. E., "Blade-Rate Force Fluctuations of

a Propeller in Non-Uniform Flow," Admiralty Research Laboratory,
ARL/M/P33A, June 1973.
McGoldrick, R. T., "Ship Vibration," David Taylor Model Basin Re-

port 1451, December 1960.
Breslin, J. P., "Techniques for Estimating Vibratory Forces Gen-

erated by Propellers," SNAHE Technical and Research Bulletin No.
l-34, January 1975.
Van Oossanen, P., "Theoretical Prediction of Cavitation on Pro-

pellers," Marine Technology, Vol. 14, No. 4, October 1977.
Huse, E., "The Magnitude and Distribution of Propeller-Induced

Surface Forces on Single-Screw Ship Model," Publication No. 100,
Norwegian Ship Model Experiment Tank, December 1968.
Frivold, H., "Solid Boundary Factors for the Afterbody of an LNG

Carrier," Norwegian Maritime Research, Vol. 4, No. 1, 1976.
Vorus, W. S., "A Method for Analyzing the Propeller-Induced Vibra-

tory Forces Acting on the Surface of a Ship Stern," Transactions,
SNANE, Vol. 82, 1974.
Yildiz, M., and Mawardi, O. K., "On the Diffraction of Multiple

Fields by a Semi-Infinite Rigid Wedge," Journal of the Acoustical
Society of America, Vol. 32, No. 12, December 1960.
Yildiz, M., and Mawardi, O. K., "Diffraction of a Dipole Field by

a Rigid Cone (Abstract)," Journal of the Acoustical Society of
America, Vol. 32, No. 11, November 1960.
Johnsson, C. A., "Pressure Fluctuations Around a Marine Propeller--

Results of Calculations and Comparison with Experiment," Swedish
State Shipbuilding Experimental Tank Report No. 69, 1971.
Breslin, J. P., and Eng, K. S., "A Method of Computing Propeller-

Induced Vibratory Forces on Ships," presented at First Conference
on Ship Vibrations, Stevens Institute of Technology, January 1965.
Tsakonas, S., Jacobs, W. R., and Ali, M. R., "Documentation for

the Computer Program for the Pressure Field Generated by a Pro-
peller in a Variable Inflow," Stevens Institute of Technology,
Davidson Laboratory Report 1910, May 1977.
125
Vorus, W. S., "Calculation of Propeller-Induced Vibratory Hull
Forces, Force Distributions, and Pressures; Free-Surface Effects,"
Journal of Ship Research, Vol. 20, No. 2, June 1976.
Johnsson, C. A., "Correlation of Predictions and Full Scale Obser-

vations of Propeller Cavitation," International Shipbuilding Pro-
gress, Vol. 20, No. 226, June 1973.
Lewis, F. M., and Kerwin, J. E., "Vibratory Forces on a Simulated

Hull Surface Produced by Transient Propeller Cavitation," Journal
of Ship Research, Vol. 22, No. 2, June 1978.
Takahashi, H., and Ueda, T., "An Experimental Investigation into

the Effect of Cavitation on Fluctuating Pressures Around a Marine
Propeller," 12th International Towing Tank Conference, Rome, 1969.
Van Manen, J. D., "The Effect of Cavitation on the Interaction Be-

tween Propeller and Ship's Hull," International Shipbuilding Pro-
gress, Vol. 19, No. 209, January 1972.
Sntvedt, T., "Propeller Induced Excitation Forces," Det norske

Publication No. 74, January 1971.
Van Oossanen, P., and van der Kooy, J., "Vibratory Hull Forces
Induced by Cavitating Propellers," Proceedings, Royal Institution
of Naval Architects, Vol. 115, 1973.
Johnsson, C. A., and Sntvedt, T., "Propeller Excitation and Re-

sponse of 230,000 TDW Tankers," Swedish State Shipbuilding Experi-
mental Tank Publication No. 70, 1972. Also presented at the Ninth
Symposium on Naval Hydrodynamics, Paris, August 1972.
Johnsson, C. A., et al., "Vibration Excitation Forces from a Cavi-

tating Propeller-Model and Full Scale Tests on a High Speed Con-
tainer Ship," Swedish Stat:e Shipbuilding Experimental Tank Publi-
cation No. 78, 1976.
Noordzij, L., "Pressure Field Induced by a Cavitating Propeller,"

International Shipbuilding Progress, Vol. 23, No. 260, April 1976.
Lewis, F. M., "Propeller-Vibration Forces in Single-Screw Ships,"

Transactions, SNAME, Vol. 77, 1969.
Lewis, F. M., "Propeller Excited Hull and Rudder Force Measure-

ments," Massachusetts Institute of Technology, Department of Ocean
Engineering Report 73-10, April 1973.
Huse, E., "Cavitation-Induced Hull Pressures, Some Recent Develop-

ments of Model Testing Techniques," Norwegian Maritime Research,
Vol. 3, No. 2, 1975.
Oosterveld, M. W. C., and van Oossanen, P., "Recent Results of

Tests in the NSMB Depressurized Towing Tank," First Ship Technology
and Research (STAR) Symposium, Washington, D. C., August 1975.
126
Norita, H., Kuritake, Y., and Yagi, H., "Application and Develop-
ment of a Large Ducted Propeller for the 288,000-dwt Tanker MS
Thorsage," Transactions, SNANE, Vol. 82, 1972.
Okamoto, H., et al., "Cavitation Study of Ducted Propellers on Large

Ships," Transactions, SNANE, Vol. 83, 1975.
Dyne, G., and Hoekstra, N., "Propulsion Cavitation and Propeller-

Induced Pressure Fluctuations of a Tanker (Comparative Tests in
SSPA Cavitation Tunnel No. 2 and NSMB Depressurized Towing Tank),"
SNAME, Spring Meeting, Philadelphia, Pennsylvania, June 1976.
The National Shipbuilding Research Program--Research on Computer

Applications to Shipbuilding VII, Catalog of Program Abstracts,
U. S. Department of Commerce, Maritime Administration, in coopera-
tion with Avondale Shipyards, Inc., May 1975.
Hananel, A. S., et al., "Static and Dynamic Analyses of the LSES Hull
Structure," Journal of Ship Research, Vol. 22, No. 2, June 1978.
Larsen, O. C., "Some Considerations on Shaft Alignment of Marine

Shafting," Norwegian Maritime Research, Vol. 4, No. 2, 1976.
Reed, F. E., "The Design of Ships to Avoid Propeller-Excited Vibra-

tions," Transactions, SNANE, Vol. 79, 1971.
Bureau Ventas Publications, Machinery Hull Interaction--Vibrations,

Paris, France, 1975.
Volcy, G. C., Gamier, H., and Masson, J. C., "Deformability of

the Hull Steeiwork and Deformations of the Engine Room of Large
Tankers," published in Machinery Hull Interaction--Vibrations,
Bureau Ventas, Paris, 1975.
Restad, K., "Investigation on Free and Forced Vibrations of an LNG

Tanker with Overlapping Propeller Arrangement," published in Ma-
chinery Hull Interaction--Vibrations, Bureau Ventas, Paris, 1975.
Volcy, G. C., Gamier, H., and Masson, J. C., "Chain of Static and
Vibratory Calculations of Propulsive Plants and Engine Rooms of
Ships," published in Machinery Hull Interaction--Vibrations, Bureau
Ventas, Paris, 1975.
Senjanovic, I., and Skaar, K. T., "Phenomena of Ship Vibration," Det

norske Ventas Publication No. 88, May 1975.
Bureau Ventas, "Recommendations Designed to Limit the Effects of

Vibrations on Board Ships," Guidance Note, N.I. 138 BM.3E, May 1971.
Volcy, G. C., "Forced Vibration of the Hull and Rational Alignment

of the Propeller Shaft," published in Machinery Hull Interaction--
Vibrations, Bureau Ventas, Paris, 1975.
127
Bourceau, G., and Volcy, G. C., "Forced Vibration Resonators and
Free Vibration of the Hull," published in Machinery Hull Interac-
tions--Vibrations, Bureau Ventas, Paris, 1975.
Hylarides, S., "Transverse Vibrations of Ship's Propulsion System,

Part I, Theoretical Analysis," International Shipbuilding Progress,
Vol. 21, No. 252, August 1975.
Vassilapoulos, L., and Bradshaw, R., "Coupled Transverse Shaft

Vibrations of Modern Ships," International Shipbuilding Progress,
Vol. 21, No. 239, July 1974.
Golden, M. E., and Henderson, F. M., "An Updated Guide to the Use of
General Bending Response Program (GBRP)," Computation and Mathematics
Department Research and Development Report 4601, Naval Ship Research
and Development Center, Bethesda, Maryland, April 1975.
Volcy, G. C., and Osouf, J., "Vibratory Behaviour of Elastically

Supported Tail Shafts," published in Machinery Hull Interaction---
Vibrations, Bureau Ventas, Paris, 1975.
Hansen, H. R., and Skaar, K. T., "Hull and Superstructure Vibrations

Design Calculation by Finite Elements," Det norske Ventas Publica-
tion 86, January 1975.
Johannessen, H., Skaar, K. T., and Sunogeli, H., "Dynamic Response

of Hull and Superstructure," presented at the International Sym-
posium on Practical Design in Shipbuilding (PRADS), The Society of
Naval Architects of Japan, Tokyo, October 1977.
Cansen, C. A., "A Parametric Study on Global Hull and Superstruc-

ture Vibration Analysis by Means of the Finite-Element Method,"
Paper Issued for Written Discussion, The Royal Institute of Naval
Architects, 1977.
Meyerhoff, W. K., "Added Masses of Thin Rectangular Plates Calcu-

lated from Potential Theory," Journal of Ship Research, Volume 14,
No. 2, June 1970.
Lindholm, U. S., Kana, D. D., Chu, W. H., and Abrainson, H. N.,

"Elastic Vibration Characteristics of Cantilever Plates in Water,"
Journal of Ship Research, Vol. 9, No. 1, June 1965.
Marcus, N. S., "A Finite-Element Method Applied to the Vibration of

Submerged Plates," Journal of Ship Research, Vol. 22, No. 2, June
1978.
Tsakonas, S., Jacobs, W. R., and Ali, M. R., "Application of the

Unsteady-Lifting-Surface Theory to the Study of Propeller-Rudder
Interaction," Journal of Ship Research, Vol. 14, No. 3, September
1970.
128
Tsakonas, S., Jacobs, W. R., and Ali, M. R., "Propeller-Rudder
Interaction Due to Loading and Thickness Effects," Journal of Ship
Research, Vol. 19, No. 2, June 1975.
Leissa, A. W., "Vibration of Plates," National Aeronautics and Space

Administration Publication NASA SP-160, Washington, D. C., 1969.
Robinson, D. C., "Ship Vibration Beam Theory: Part I - A Review of

the Theory," Report AVL-76-962, Acoustics and Vibration Laboratory,
Vibration Division, David Taylor Model Basin, Washington, D. C.,
August 1965.
Jensen, J. J., and Madsen, N. F., "A Review of Ship Hull Vibration.
Part I: Mathematical Models; Part II: Modeling Physical Phenomena;
Part III: Methods of Solution; Part IV: Comparison of Beam Models,"
The Shock and Vibration Digest, Vol. 9, Nos. 4-7, April-July 1977.
Leibowitz, R. C., and Harder, R. L., "Mechanized Computation of Ship

Parameters," David Taylor Model Basin Report 1841, June 1965.
Volcy, G. C., Gamier, H., and Masson, J. C., "An Analysis of the
Free and Forced Vibrations of Cargo Tank Structure by Finite-Element
Technique," HANSA, Heft 9, Hamburg, 1975.
Skaar, K. T., "On the Finite-Element Stress Analysis of Oil Tanker

Structures," Norwegian Maritime Research, Vol. 2, No. 2, 1974.
Kamel, H. A., and Liu, D., "Structural Dynamic Problems in Ships and
Other Marine Structures," contained in Finite-Element Applications
in Vibration Problems, edited by M. M. Kamel and J. A. Wolf, Jr.,
The American Society of Mechanical Engineers, New York, September,
1977.
Egeland, O., and Araldsen, P. O., "SESAM-69 - A General Purpose

Finite Element Program," Computers and Structures, Vol. 4, No. 1,
January 1974.
Kuniai, T., "Damping Factors in the Higher Modes of Ship Vibration,"

Reports of Research Institute for Applied Mechanics, Vol. VI, No.
21, Kyushu University, Fukuoka, Japan, 1958.
Woolam, W. E., "Research on Ship-Hull Damping Coefficients for Low-

Frequency Vertical Flexural Modes of Vibration," Acoustics and
Vibration Laboratory Research and Development Report 2323, Naval
Ship Research and Development Center, Nay 1967.
Volcy, G. C., and Nakayama, M., "Studies Leading to Vibration and

Noise Free Ships," Bulletin Technique Du Bureau Ventas (Special
English Issue), Bureau Ventas, Paris, June 1977.
Hart, H. H., "Hull Vibrations of the Cargo Liner Koudekerk,"

International Shipbuilding Progress, Vol. 18, No. 206, October 1971.
129
Knopf le, W. H., "Vibration Survey Techniques," Narine Technology,
Vol. 8, No. 1, January 1971.
Kuhii, K., "Ship Vibration Measurement and Analysis by Data Acqui-

sition and Processing System," IHI Engineering Review, Special Issue
'71 Ships, Ishikawajima-Harima Heavy Industries Co., Ltd., Tokyo,
Japan, July 1971.
Solumsmoen, 0. H., "Ship Vibration - Experience from Service Mea-

surements," Det norske Ventas Publication No. 96, January 1977.
Okamoto, H., "Reduction of Hull Vibration by Proper Selection of

Propeller Type and Stern Form," Bulletin of the Marine Engineers
Society of Japan, Vol. 5, No. 2, June 1977.
Huse, E., "Effect of Afterbody Forms and Afterbody Fins on the Wake
Distribution of Single Screw Ships," Norges Skipsforskning-institut,
Report R-3l.74, Trondheim, Norway, 1974.
Huse, E., "Uber propellererregte Vibrationen auf Schiffen," Hansa,

No. 12, 1970.
Huse, E., "Trykkimpulser fra Kaviterende propell," Nordiskt Skepps-

tekniskt Mate (NSTM), 1971. (In Norwegian)
Lindgren, H., and Johnssen, C. A., "On the Influence of Cavitation

on Propeller Excited Vibratory Forces and Some Means of Reducing
its Effect," presented at the International Symposium on Practical
Design in Shipbuilding (PRADS), Tokyo, Japan, October 1977.
Johnsson, C. A., "Stern Vibrations of High--Powered Ships," Final

Report No. 1763-14, Swedish State Shipbuilding Tank, G5teborg,
Sweden, August 1975.
Johnsson, C. A., "Hydrodynamically Generated Hull Vibrations--The

Propeller as Excitation Source," Nordiskt Skeppstekniskt Mate (NSTM),
Lyngby, Denmark, 1977. (In Swedish)
Santon, T. J., and Giblon, R. P., "Farrell Lines '85' - Class Con-
tainerships," Narine Technology, Vol. 16, No. 1, January 1979.
1948.
Morse, P. M., Vibration and Sound, McGraw-Hill Book Co., Inc.,
National Steel and Shipbuilding Company, "Highly-Skewed Propeller

Research Program, San Clemente Ore/Bulk/Oil (OBO) Carriers," Final
Report to Maritime Administration on Contract 2-36315, November 1974.
Littleton Research and Engineering Corp., "Measurement and Analysis

Pro-
of Harmonic Hull Pressures and Vibration Generated by Normal
pellers on the Seabridge Class RO/RO Ships," Report C-234-2 to
American Export Lines, November 1974.
130
APPENDIX A
COMPUTER PROGRAMS FOR COMPUTING

PROPELLER FORCES AND MOMENTS
A-l. Propeller Mean and Vibratory Forces Program

(Program developed by Davidson Laboratory, Stevens Institute of
Technology, under U.S. Navy contracts; widely used.)
Output Information
1. Steady and time-dependent blade loading distribution at multiples

of shaft frequency.
2. Mean and blade-frequency force and moment components in coefficient

form for:
(a)
24
Thrust/pn d
25
Torque/pn d
Transverse force/pn d
24
Vertical force/pn d
24
Transverse bending momentlpn2d5
Vertical bending moment/pn2d5
where
p = fluid density
n propeller rpm
d = propeller diameter
3. Blade bending moments about the pitch line at various radial positions
and for various orders of excitation.
4. Information for the study of cavitation inception.
5. Information for the study of blade stress analysis which is per-

formed by utilizing the STARDYNE-CDC finite element computer program.
Input Information
The propeller blade geometry.
The Fourier components of the spatial variation of the axial and tan-
gential components of the wake.
131
Basis for Calculation
The program uses unsteady lifting-surface theory and takes into con-
sideration all the relevant propeller geometry and the spatial nonuniformity
of the inflow field. The program is available through Davidson Laboratory
for $6,000. See Reference 27.
A-2. Harmonic Forces and Moments Generated by a Propeller in

Nonuniform Flow
(Program developed and used by Littleton Research and Engineering
Corp.)
Output Information
Magnitude and phase of the three components of harmonic propeller

force and the three components of harmonic propeller moment.
2. The steady vertical and horizontal forces and moments arising from
first-order wake action (thrust offset).
Input Information
Propeller drawing. The propeller drawings should show the following

information: propeller diameter, hub diameter, rake, number of blades
and propeller material; the variation with radius of chord, skewback,
and pitch; propeller sections at several radii showing the variation
of thickness along the chord. For propellers designed in Europe, the
variation with radius of the distance from the reference line to the
leading edge, trailing edge, and point of maximum thickness is accep-
table in place of the variation of chord and skewback.
Ship speed and corresponding shaft rpm.
Wake as measured in a model test. The results of a harmonic analysis

of the measured wake are required. If the harmonic analysis results
are not available, the measured inflow velocities specified at sev-
eral points along the radius and at frequent points around the cir-
cumference are acceptable, and a harmonic analysis will be performed.
If a measured wake is not available, it can be inferred from the

available wakes of other ships.
Basis for Calculations
Propeller forces are determined by lifting line theory. This is much

less complex than the Davidson Laboratory lifting surface theory, but is
considered adequate in view of uncertainties in the wake and the wide vari-
ation in service wake due to ship motions and sea action. The main reason
or continuing to use the lifting line theory calculation is that it is the
asis for the predictions of hull pressure and hull forces (see Appendix
B-l).
132
This is a proprietary program not developed for general distribution.
A-3. Calculation of Steady and Harmonic Propeller Forces

(Computer program used by the American Bureau of Shipping.)
Output Information
Mean and blade-frequency components of the three forces and three mo-
ments acting on the propeller.
Time-varying blade pressure distribution at each wake harmonic.
Input Information
Propeller blade geometry.
Fourier coefficients of the spatial variation of the axial and tan-

gential components of wake.
Basis of Calculation
The program employs an extended version of unsteady lifting line

theory as developed by Dr. Neal A. Brown at MIT [281. The extension in-
cludes the effects of propeller skew, which were not treated in the origi-
nal theory. The results of this program are used as partial input to the
"Surface Force" program described in Appendix B-3.

Nonuniform Flow
(A computer program under development at Massachusetts Institute of
Technology by Professor Justin E. Kerwin, Department of Ocean Engi-
neering.)
The program represents the propeller blade by grid points distributed

over the surface and the wake spatially definéd (cylindrical coordinates)
in three directions: longitudinal, tangential, and radial.
A distribution
of vorticity is assumed over the surface, and by successive iterations
is
refined to be compatible with the boundary of the propeller surface and the
laws of hydrodynamics, Kelvin's theorem, and the Kutta requirement for
flow
continuity at the trailing edge.
This discrete element approach appears to offer a number of advantages

as a starting point for the computation of unsteady, partially cavitating
flows:
It is capable of yielding accurate predictions of mean

loading, both at design and off-design conditions.
Being a numerical procedure, blade geometry can be incor-

porated exactly so that propellers with large skew, rake,
133
and varying pitch distribution can be accommodated. This
is considered essential, since it is through the variation
of these parameters that optimum propeller designs can be
evolved.
Since the procedure includes all three components of in-

duced velocity, there is no particular problem in including
tangential and radial wake field components.
Since no loading mode functions are employed, the modifi-

cations ultimately required to include the cavities would
appear to be feasible. Source elements presently included
to represent blade thickness can assume the further role
of representing the cavity volume.
A discrete element method lends itself naturally to a step-

by-step domain solution, which is also essential for the
subsequent inclusion of unsteady cavitation.
The procedures are still under development, but have been applied to
specific cases with good results. See References 18 and 29.

Nonuniform Flow
(Computer program developed and used by the Admiralty Research
Laboratory, Teddington, England.)
Output Information
The input data.
If wakes are given as velocity measurements, the harmonic values are

printed (to the 71 harmonic). If given as Fourier components, these
are listed.
The contribution to thrust, torque, vertical and horizontal forces,

and moments from each specified radial section.
The integrated thrust, torque, horizontal and vertical forces, and

moments for multiples of blade rate harmonics.
Input Information
Shaft speed.
Propeller geometry, including skew, chord length, blade pitch angle

at specific radii.
Wake, either in Fourier Series, amplitude-plus-phase form, or as

equally spaced measurements of wake at the radii where the propeler
geometry information is given. Only axial or both axial and tan-
gential wakes may be specified.
134
Calculations can be run for successive skew values.
Input radii may vary from 4 to 14.
As many as 20 skew configurations may be determined.
As many as 140 harmonics of the blade frequency forces may be cal-

culated, but generally the number is limited to 10.
As many as 100 wake harmonics and 200 wake measurements per radius
may be input.
The calculation of the fluctuating forces on a propeller falls into

three parts. The first part is the calculation of the variation of the in-
flow velocity to the blades; the next stage involves the calculation of the
fluctuating lift-distribution on a section of blade associated with this
fluctuating inflow; the final stage is the calculation of the propeller
forces and moments. The calculation of the fluctuating lift is based on
two-dimensional unsteady airfoil theory. It ignores blade-to-blade inter-
action and the variation with radius of the various significant parameters.
These approximations would be unacceptable for predicting the steady lift,
but are acceptable for the unsteady lift, probably overestimating the lift.
See Reference 30.
135
APPENDIX B
COMPUTER PROGRAMS FOR COMPUTING HULL

PRESSURES AND FORCES
B-l. Calculation of Harmonic Forces and Moments on the Hull

Generated by Propeller Action
Corp.)
Output Information
The harmonic hull surface pressure at blade-rate frequency generated

by the loaded, noncavitating propeller in the region of the propeller
(generally at grid points corresponding to underwater intersections
of buttocks and frames within four diameters of the propeller).
By integration of the above, the blade frequency harmonic hull forces

and moments acting on the hull due to noncavitating propeller action.
Input Information
The computed propeller lift distribution along the propeller blade

(see Appendix A-2).
The geometry of the propeller.
The hull coordinates at the points of pressure determination.
The free-field pressures (i.e., the pressures that would exist in

open water if the hull were not present) are calculated at each hull grid
point due to (1) the loading on the propeller blades (assumed to be concen-
trated at the forward quarter point of the blade chord), (2) the thickness
of the propeller blade. The sum of these two pressures, in their proper
phase, is multiplied by two to give the reported pressure on the hull sur-
face. The pressure from a harmonically varying force having x, y, and z
components involves the distance from the point to the location of the force.
Substituting steady and harmonic forces and distances as a function of shaft
angle yields values of the pressure. The resulting equations involve a se-
ries which under certain conditions converges slowly. Originally, only a
been developed,
few terms were developed. More recent1y. the general term has
allowing sufficient terms to assure convergence. This results in pressures
that correspond to measured values. The integration for blade thickness is
If the cavítation volume on the blade could be defined by a Four-
similar.
ier series, the same process could be applied. This has not yet been done.
This is a proprietary program not developed for general distribu-

tion. See Reference 26.
136
B-2. Calculation of Steady and Harmonic Pressure Fields Generated
by a Noncavitating Propeller
(Program developed by Davidson Laboratory, Stevens Institute of
Technology, under U.S. Navy contracts.)
Output Information
This program furnishes the steady and harmonic components of the pres-
sure field generated by a noncavitating ship propeller operating in a spa-
tially variable mf low.
Input Information
The propeller blade geometry.
The Fourier components of the spatial variation of the axial and

tangential variation of the axial and tangential components of the
wake.
The spatial location of the points where the pressures are desired.
The steady and time-dependent blade loading distribution at multiples

of any shaft frequency as produced by the program described in Appen-
dix A-l.
This program is a continuation of the one described in Appendix A-1

and requires data generated in that program. It is available through David-
son Laboratory for $5,000. See Reference 41.
B-3. Calculation of Propeller-Induced Hull Surface Forces

(Program developed and used by Professor William S. Vorus (University
of Michigan) and the American Bureau of Shipping.)
Output Information
This program computes all components of the hull force and moment at
multiples of the propeller blade rates. (In general, the vertical force
component is the only one desired.)
Input Information
Propeller geometry.
Wake distribution.
Stern lines and coordinates describing the sectional geometry of

approximately the aft one-third of the ship.
137
4. Lime-dependent geometry of propeller cavitation effects (optional). i
5. Time-varying blade pressure distribution at each wake harmonic (out-

put from program described in Appendix A-3).
Basis of Calculatíon
This program employs the method presented by Professor William S.

Vorus in Reference 36 . The conventional procedure of evaluating the hull
forces is to integrate the propeller-generated pressures over the hull sur-
face. These pressures are due to diffraction of the propeller-induced water
flow by the hull. The diffraction problem and hence the pressure integration
difficulties are avoided in the analysis and computer program by utilizing a
special application of Green's Theorem.
138
APPENDIX C
COMPUTER PROGRAMS FOR COMPUTING THE LONGITUDINAL

RESPONSE OF THE PROPULSION SHAFTING
C-l. Longitudinal and Torsional Shafting Vibrations

(Program used by Maritime Administration.)
Number and/or Name
**C_9_002
Category(s)
**Hull Shafting Calculations
Descriptive Program Title
**Shaft Vibrational Analysis using Holzer Method
Source Activity
**Offjce of Ship Construction

**Maritjme Administration
**Washington, D.C.
Engineer(s) Naine-Code-Phone
**Richard Siebert, 721.21, 254-7048
Programmer Name-Code-Phone
**NAVSEC
Program Status
**Pro duct ion
Classification (Security)
**Restricted - NAVSEC Program
Programming Language
**FORT IV
Computer Type Used
k*Control Data 6600
139
Special Hardware
**None
Special Hardware/Operation
**None
Program Size-Source Deck Cards
'c*258
Program Size-Object Core Words
**COOOO Octal Words
Average Running Time (Mm)
**257
Program Availability
**September 1970
Documentation Status
**Informal - Complete (15 pages)
Program Abstract
This program calculates torsional and longitudinal critical vibration

frequencies using the Holzer Method. It was originally developed by NAVSEC
for the IBM-7090 and subsequently converted to the CDC-6600 by the Maritime
Administratíon. Double precision requirements were eliminated. Input re-
quires hand calculation of all masses, inertias, and stiffness factors for
each component in the turbine-gear-shaft-propeller system. Damping factors
are not included in the calculation. Output consists of critical frequencies
in CPS and RPM for various numbers of blades.
C-2. Longitudínal Shafting Vibrations

(Program used by J. J. McMullen Associates, Inc.)
Number and/or Name
**F_8_0O8
Category(s)
**Machinery Shafting and Bearing Calculations
140
Source Activity
**John J. McMullen Associates, Inc.

**One World Trade Center-Suite 3047
**New York, New York 10048
Engineer (s) Name-Code-Phone
**Engineering Division
Program Status
**Product ion
**Unclassjfied
**FoRTpJ IV
Computer Type Used
**IBM 360/40 and IBM 1130
**Inforl - User's Guide
Program Abstract
Lumped mass system, using "1eve1' effect. Text description found in

NSRDC Report 3358, September 1970. Computes frequencies up to four modes.
C-3. Longitudinal and Torsional Shafting Vibrations

(Program used by Newport News Shipbuilding and Drydock Company.)
Number and/or Name
**F..0_016 9.5.0251 FORCE VIB
Category(s)
**Machinery Shafting and Bearing Calculations
**Longitudínal and Torsional Vibration in Propulsion

Shafting Systems
141
Source Activity
**Newport News Shipbuilding and Drydock Company

**Technical Systems Division
*!c4101 Washington Avenue, Newport News, Virginia 23607
(804) 247-7500
Engineer(s) Name-Code-Phone
**A. S. Pototzky
Programmer Name-Code-Phone
**F. E. Siegel
Program Status
**Actjve Production
**TJnclassifíed
*FORTPN IV
Computer Type Used
**1-Ioneywell 6080
**Not available for general distribution
**Incomplete
Program Abstract
FORCE VIB is a computer program to calculate the steady-state longí-

tudinal or torsional vibratory response of branched shafting systems, such
as propulsion systems. The system may have a maximum of 35 elements con-
sisting of masses, dampers, and springs, all with only one degree of free-
dom. The masses and springs may be lumped or distributed, and the dampers
may be viscous or solid. The program uses the mechanical impedance method
to calculate displacements, forces, and phase angles, which may all be fre-
quently dependent. The program also allows the varying of values to conduct
parametric studies.
142
C-4. Longitudinal Vibration of Shafting, II
Corp.)
Output Information
A plot of the blade-order harmonic force at the thrust bearing as a

function of rpm.
A plot of the amplitudes of axial motion at the propeller and at the

thrust bearing as a function of rpm.
Tabular data for above.
Input Information
Shafting arrangement, diameters, and lengths.
Propeller mass, diameter, number of blades, pitch, and developed

area ratio (or mean width ratio).
Harmonic thrust (can be determined by program described in Appendix

A-2).
The stiffness of the thrust bearing and its foundation.
Reduction gear weight.
The propeller is represented by its mass plus entrained water and

damping, estimated by Lewis and Auslaender's recommendations. The shaft is
represented by a distributed mass and elasticity and is assumed to have a
hysteretic damping (nominally 4%). The thrust bearing is represented as a
concentrated mass elastically connected to a rigid hull.
143
APPENDIX D
COMPUTER PROGRAMS FOR COMPUTING THE LATERAL

RESPONSE OF THE PROPULSION SYSTEM
D-l. Transverse Response of a Bean-i

(Program used by Newport News Shipbuilding and Drydock Company.)
Number and/or Name
*A_12_OO 9.5.0301 Beam Vibration
Category(s)
**Conceptual Design
**Ship Vibrations
**Vibration Analysis of Beams
Source Activity
**Newport News Shipbuilding and Drydock Company

**Production Computer Systems Division
**410l Washington Avenue, Newport News, Virginia 23607
(804) 247-7500
Program Status
**production Use
**Unclassjfjed
**FORT? V
Computer Type Used
**Honeywell 6080
Special Hardware
**None
**Not available for general distribution
144
Program Abstract
Computes the steady-state transverse vibratory response of a beam with

any number of intermediate flexible supports, with generalized end condi-
tions, section properties, and loading.
D-2. Transverse Vibration of Shafting and Propeller

Corp.)
Output Information
Plots of bearing forces, in two normal directions, as a function of

frequency.
Plots of shaft motions, in two normal directions, as a function of

frequency, at the propeller and at other critical locations.
Plots of the shaft deflection curves at each natural frequency with-

in the operating speed.
Plots of the steady plus harmonic bending moments in the shaft of

the aftermost bearing.
Computer tables for above.
Input Information
Shafting arrangement, diameters, and lengths.
Propeller weight and moment of inertia about its rotation axis, diame-
ter, number of blades, pitch, and developed area ratio.
Stiffness or flexibility matrices for each bearing about axes perpen-

dicular to the axis of rotation (force and rotation).
Horizontal and vertical harmonic forces and moments, and the steady
thrust (can be determined by the program described in Appendix A-2).
Basis for Calculations
The propeller is represented by its mass, its entrained water, its

moment of inertia about the rotational axis, its moment of inertia about an
axis perpendicular to the rotational axis, the moment of inertia of its en-
trained water, and the hydrodynamic damping in its several modes of motion.
The shaft is represented as a series of uniform beams having distributed
mass and bending stiffness and hysteretic damping. The bearings are repre-
sented by their stiffness in translation in two directions mutually perpen-
dicular to the shaft axis and by their stiffness in rotation about the same
two axes. The bearings are assumed to bend with the shaft; however, where
there is flexibility between shaft and bearing, e.g., rubber staves, this
flexibility, lateral and angular, is incorporated in the strut matrix. It
145
is generally acceptable to terminate the shaft at the after inboard line-
shaf t bearing.
The program computes the vibration in terms of coupled properties in

the horizontal and vertical directions. It includes the influence of the
steady thrust (small effect), but not that of the steady torque (very small).
146
APPENDIX E
COMPUTER PROGRAMS FOR COMPUTING THE RESPONSE

OF THE ENTIRE HULL GIRDER STRUCTURE
E-l. Cenerai Bending Response Program

(Program developed and used by Naval Ship Research and Development
Center.)
Name
**GB RP
Category(s)
**Lateral, Longitudinal, and Torsional Beam Vibrations

**Bending Coupled with Torsional Beam Vibrations
**Whirling Vibrations of Propeller Shafts
**General Bending Response Program
Source Activity
**Naval Ship Research and Development Center

**Bethesda, Maryland 20084
Engineer(s)
**Mjchael E. Golden
**Francjs M. Henderson
Program Status
**Actjve Production
**Unclas sif led
Prograrnmíng Language
**FORT IV
Computer Type Used
**CDC 6000 Series
Output Plotting
**SC 4020 Plots
147
Special Software/Operation
**Overlays for Program Subroutines

**Open Core to Optimize Storage
**Availability for general distribution through David Taylor Naval

Ship Research and Development Center
Documentation
**Complete
Program Abstract
The General Bending Response Program (CBRP) consists of the union of

three programs: General Bending Response Code i (GBRC1) for lateral, longi-
tudinal, and torsional vibrations; GBRC2 for vibrations involving bending
coupled with torsion; and GBRC3 for whirling vibrations of propeller shafts.
The latter two codes resulted from an extended application of the mathemati-
cal model used in the first code. The program formulates the finite-differ-
ence equations which approximate the boundary-value problem representing the
steady-state motion of a vibrating nonuniform mass-spring system such as a
ship hull or shafting in bending. The program calculates natural frequencies
and mode shapes and the response to specified harmonIc driving forces and
moments. The program can represent a ship hull connected elastically to other
systems such as the propulsion system and to sprung masses. Longitudinal or
torsional vibrations problems can also be solved by dividing each beam into
sections connected by springs, thus reducing the model to a mass-spring sys-
tem. See Reference 73.
E-2. Ship-Hull Vibratory Response

(Program used by USS Engineers and Consultants, Inc.)
Number and/or Name
**A_12_006 SHRVS
Category(s)
**Conceptual Design
**Shjp Vibrations
*Simuiated Ship-Hull Vibration
148
Source Activity
**USS Engineers and Consultants, Inc.

**600 Grant Street
**Pittsburgh, Pennsylvania 15230
Engineer(s) Name-Code-Phone
**F Ronald Griffith

(412) 433-6517
Program Status
**Produc t ion
**Unclassified
**F0RT IV
Computer Type Used
**CDC 6500
Special Hardware
* *N one
Special Software/Operation
* *No ne
Program Size-Source Deck Cards
**4,000 cards
Program Size-Object Core Words
**160K Octal Words
Average Running Time (Miri)
**3 minutes
**Tíme Sharing Service

Call F. R. Griffith
149
Cost
**Ne go t iab le
**]Informal Complete
Program Abstract
The purpose of SHRVS is the accurate prediction of the vibration

response of a ship hull to either steady-state or transient loads applied
in the vertical centerline plane of the hull. Factors considered include
cargo distribution, bulkhead location, machinery space location, as well as
the f lexural and shear stiffness of the main-hull girder and of the double-
bottom structure.
150
APPENDIX E-3
COMPUTER PROGRAMS AVAILABLE FOR COMPLETE HULL ANALYSIS
Institut für
Schiffetechnik Electric Boat
Lloyd's Register Netherlands Technische Division, Computas, Subsidiary of
of Shipping Ship Model Basin Universitat Berlin General Dynamics Corp. Det noruke Venta.
NANE OF PROGRAM Null Vibration DASH FREIS, ERZS CENSAN SESAM-69

LRS7P
PEPSON TO CONTACT Geoffrey H. Sole Dr. S. Hylaridee Dr. E. Metznieier Dr. Henna Allik B. Aamodt
TYPE OF ANALYSES Eigenvalues. Eigenvalues, Eigen- Eigenvalues, Eigen- Rigenvalues, Eigen- Forced and Free Vibration
Elgenvectors vectors, Steady-State, vectors, Steady-State, vectors, Steady-State, Analyses of Steady-State
and Transient Response and Transient Response Transient, and Random or Transient Vibrationa
Response
DATE OF INITIAL U/1976 1970/1975 1914/u 1967/1976 1969/1975
COI1PLETION/LATEST
REVISION
AVAILASILITY/ Service from Service from NSMB U

PRICE Proprietary - Avail- Service use available
Lloyd's Register ble on a case baci. for rent or cale
IJINCUACE FORTRAN IV M,GOL FORTRAN FORTRAN V USAS! FORTRAN

COMPUTER SYSTEM IBM 370/159 Irrelevant CDC 6500/ UNIVAC 1106, 1108, UNIVAC 1108. Other
Scope 3.4 1110 EXEC 8 Computers
BASIS OF PROGRAM U Finite Element Finite Difference Finit, Element Finite Element
TYPE OF STRUCTURES Represented by a Complete or Partial Ship Null General Structure. General Structure.
Free-Free Beam Ship
MANNER OF Beams Beams, Platea Beams
REPRESENTATION Beans, Platee, Beams, Platee, Shell.,
Shells, 2- and 3-D 2-D and 3-D Continua
Continua
CONCENTRATED OR Concentrated Concentrated Concentrated Both Concentrated
DISTRIBUTED MASS
SUBSTRUCTURING No U U Yea Ye.
COMPRESSION OF No No U Yes Ye.
DYNAMIC MATRIX
TREATMENT OF Direct Input Direct Input According to U

WATER INERTIA or Internal Direct Calculation
Lewis and Ruinai
Computation
TREP.TMENT OP No Damping Viecou. Viscous and
DAMPING User Supplied Damping Modal Damping
Hysteretic Array. Except Viscous
PROGRAM CAPACITY 121 Nodes Unlimited 41 DOF Long. Unlimited Unlimited
240 DO? 82 DOF Vert.
164 DO? Coup.
HOri.-Toraion
U - undefined
1.2 I I
vx/v (LONG.)
VT/V (TAN.)
1.0
VA/V (RADIAL)
0.8
0.6
0.4
0.2
-0.2
0 40 80 120 160 200 240 280 320 360

ANGLE IN DEGREES
V = 23 Knots , 0.335 Radius
FIGURE F1. LONGITUDINAL, TANGENTIAL, AND AXIAL WAKES AT 0.335 RADIUS

1.2
1.0
CI) 0.8
o
i!
0.2
-0.2
0 40 80 120 160 200 240 280 320 360

ANGLE IN DEGREES

1.2 I I
1.0
ocPDO00000000do c?SPo0oo0 oQ-
00 0
0.8
CI) o o
o o o o
o o
o
0.6 0°
H
z o - vx/v (LoNG.)
z A - VT/V (TAN.)
D - VA/V (RADIAL)
::
V..
J..
O a ' ¿4AAAAA1
D
D
-0 . 2
I I t
0 40 80 120 160 200 240 280 320 360
ANGLE IN DEGREES

1.2
1.0
00 0000
0.8 o o
cl) o o
o
H
ê o
0.6
o
H
z 00o o- VX/V (LONG.)
z
o D - VT/V (TAN.) 15)8
0.4
L) VA/V (RADIAL)
H
L)
o
Aa D
A,4ó1 a,':'
.
ci
-0 . 2
I i I i I I i i I i
0 40 80 120 160 200 240 280 320 360
ANGLE IN DEGREES

I I I I
I I I
I I
00-V
0.8 - o
o
o
o
O o - vx/v (LONG.)
0.6 %
-ò t -VT/V (TAN.)
O O A - VA/V (RADIAL)
O
0.4
00
o
Q o
0.2
A
o 110jgIIJ
D0000
-0.2
I I I I i I I
I i i
O 40 80 120 160 200 240 280 320 360
ANGLE IN DEGREES

APPENDIX G
SUMMARY OF VIBRATION STUDIES CONDUCTED BY LITTLETON RESEARCH

AND ENGINEERING CORP. FOR AN RO/RO TRAILER SHIP DESIGNED BY
SUN SHIPBUILDING AND DRYDOCK COMPANY
Rept.
No. Title Summary
Vibratory Propeller Forces, Report of computation of propeller

Hull Pressures and Forces, forces transmitted to shafts, hull
and Propeller Stresses of pressures, and hull forces for cen-
a Trailer Ship ter propellers and wing propellers
having 4, 5, and 6 blades.
II Longitudinal and Transverse Natural frequencies of and wing

Vibration of Shafts, Sun
Trailership, Hull 665 shafts in longítudínal vibration as
a function of combined thrust bear-
Ing foundation stiffness. Resonant
frequencies of and wing shafts
in vertical and transverse vibra-
tion. Hull assumed rigid, strut
stiffness in translation and rota-
tion included, bearing fluid film
assumed rigid.
III Transverse Vibration of Studied the effects of changes in

Shafts, Raked and Unraked shaft diameter on lateral frequency.
Struts, Sun Trailership Considered the influence of strut
Struts rake on vertical and transverse
resonant frequencies
IV Calculated Static Stiff- Determined the static stiffness of

nesses of Thrust Bearing the foundation of the and wing
Foundations and Longitudi-
nal Vibration of Shafts, thrust bearing foundations using a
Sun Trailership, Hull 665 finite element analysis of the
engine room structure. Using these
stiffnesses and a stiffness of the
thrust bearing, the natural fre-
quencies of GL and wing shafts were
computed.
V Frequencies and Mode Shapes Using a dynamic finite-element rep-

of Shafts and the Machinery resentation of the shafting and
Space, Sun Trailership, adjacent hull structure, the shaft
Hull 665 vibration frequencies of the
shaft system in vertical motion and
the wing shaft in vertical and
transverse motion were determined.
These are significantly lower than
157
Rept.
No. Title Summary
V those determined under the rigid-

hull assumption. Longitudinal vi-
bration frequencies of the and
wing shafts are determined. It is
found that the fundamental f requen-
cies are not much changed from those
obtained using static stíffness, but
that new frequencies of importance
are introduced.
VI Calculated Propeller In- The vibration levels throughout the

duced Vibration Levels, ship are predicted. Levels are
Sun Trailership, Hull 665 reasonable by ISO standards in the
machinery operating stations and in
the crew quarters low in the hull.
However, they are excessive in the
house. Frequencies of peak ampli-
tudes generally coincide with shaft
resonance frequencies.
VII Effect of Center Thrust Several thicknesses of the longi-

Bulkhead Stiffening on tudinal bulkheads supporting the
Longitudinal Shaft Fre- thrust bearing are studied to
quency, Sun Trailership, determine the influence on longi-
Hull 665 tudinal natural frequency.
VIII Effect on Relocating the In an effort to increase the nat-

Center Thrust Bearing on ural frequency of longitudinal vi-
the Longitudinal Shaft bration of the Cj propeller shaft
Resonance, Sun Trailership, system, the effect of moving the
Hull 665 thrust bearing aft is studied.
This modification is found to be
more effective than increasing the
scantlings of the thrust foundation
structure.
IX The Effect of Design The design changes are made in an

Changes Upon the Resonance attempt to bring the shaft reso-
Frequencies of Propulsion nance above the blade frequencies
Shafts, Sun Trailership, of a 5-bladed j propeller and 4-
Hull 665 bladed wing propellers. Structural
changes such as increasing thick-
ness of a longitudinal bulkhead,
adding a transverse bulkhead, in-
creases in shaft diameter, modifi-
cation of struts, increases in
thickness of shell plating and
plating of a flat, plus a change
in location of wing propellers, are
considered.
158
Rep t.
No. Title Summary
X Guide for Response Output Users' Guide
Interpretation, Sun Trailer-
ship, Hull 665
XI Effect of Stern Bearing Using the simple shafting model,

Location and Strut Rake on several alternative processes for
Center Shaft Vertical and changing the transverse frequency
Lateral Vibration, Sun without changing the vertical fre-
Trailership, Hull 665 quency are explored.
XII Center Shaft and Strut By reducing propeller overhang,

Vibration Study, Sun moving the struts aft on the bar-
Trailership, Hull 665 rel, stiffening the hull, and in-
creasing the strut thickness and
chord with tapering, it is possible
to locate the vertical and trans-
verse frequencies of the centerline
shaft at desirable values.
XIII Wing Shaft and Strut Vi- It is shown that a combination of

bration Study, Sun Trailer- scantlings and rake for the wing
ship, Hull 665 propeller struts can result in
acceptable resonant frequencies.
The addition of a transverse bulk-
head to avoid internal hull reso-
nance is recommended.
XIV Vibration Evaluation of Previous studies have been based

Built-up Struts, Sun on solid cast steel struts. Because
Trailership, Hull 665 of procurement difficulties, fabri-
cated struts are of interest. Suit-
able built-up struts are determined.
XV Vibration Evaluation of The design of the struts was shown
Alternate Strut Designs, in previous studies to have a strong
Sun Trailership, Hull 665 influence on the shafting resonant
frequencies. This study investi-
gated the effects upon natural fre-
quency of changing strut rake and
different rakes in each of the strut
arms.
XVI Static Deflections and Use is made of the finiteelement

Stresses, Sun Trailership, model, developed for dynamic analy-
Hull 665 sis, to compute the hull deflections
due to shaft thrust and torque.
These deflections are evaluated
against limits set by the reduction
gear manufacturer.
159
/
Rep t.
Title Summary
No.
Calculated Propeller Induced This is the final update of the

XVII
Vibration Levels, Sun whole hull vibration level predic-
Trailership, Hull 665 tion. As a consequence of the
attention given in the shaft reso-
nances, the vibration level at de-
sign power is at a minimum with
peaks below and above. Although
the vibration at the lower peak
would not be acceptable under con-
tinuous operation, this frequency
is at a speed which would be seldom
used.
U.S. GOVERNMENT PRINTING OFFICE: 1979-311586/251
160

For Minimizing Vibration In: A Design

Uploaded by

Copyright:

Available Formats

For Minimizing Vibration In: A Design

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

For Minimizing Vibration In: A Design

Uploaded by

Copyright:

Available Formats

SSC-29 i

This document has been approved

SHIP STRUCTURE COMMITTEE

SHIP STRUCTURE COMMITFEE

The SHIP STRUCTURE COMMITTEE

The SHIP STRUCTURE SUBCOMMITTEE acts for the Ship Structure

U. S. GEOLOGICAL SURVEY Mr. F. J. DASHNAW

ThE SOCIETY OF NAVAL ARCHITECTS Mr. R. H. STERNE - Liaison

The first phase developed a bibliography published as

The subsequent phases, including the development of a

Approximate Conyersiono horn Metric Measures

When You Know Multiply by Ta Fryd Symbol

Fohronheit 5/9 (alter Celsius

A design procedure for minimizing propeller-induced vibration in hull struc-

17. Key Words 18. Distribution Statement

Unclassified Unclassified 160

Form DOT F 7OO.7 )B-72) Reproduction of completed page authorized

17 Finite-Element Mesh for Two-Dimensional Model

18 Forced Response Depending on the Applied Global

19 Calculated Forced Response at the Top of Super-

20 Position of Nodal Points on the Main Deck for

21 Elasto Dynamic Model of Aft Part and Correlation

22 Preliminary Hydrodynamic Design Phase 61

23 MILESTONE I - Preliminary Hydrodynamic Design

24 Axial Wake Distributions for Original and

25 Final Hydrodynaniic Design Phase 65

26 MILESTONE II - Final Hydrodynamic Evaluation 66

27 Ship Substructure Design Phase 69

28 MILESTONE III - Ship Substructure Evaluation 70

29 Complete Shíp Structure Design Phase 71

30 MILESTONE IV - Complete Ship Structure Evaluation 72

31 Test and Evaluation Design Phase 74

32 MILESTONE V - Test and Evaluation Review 75

35 Frequency of Longitudinal Vibration Versus

36 Ratio of Measured Hull Pressures Where Cavitation

37 Vertical Harmonic Force and Transverse Bending

39 Grid Points on Frame 170, X = 396 In. 93

40 Grid Points on Frame 181, X = O In. 94

41 Amplitude of Axial Motion at the Propeller 99

42 Amplitude of Fore and Aft Vibration at Propulsion

43 Vibratory Motion on Bridge, Frame 164 at Center-

44 Vertical Vibration on 36-ft Flat Generated by

45 Vertical Vibration on 26-ft Flat Excited by

46 Vertical Vibration in Tank Top due to Longitudi-

47 Vertical Vibration in Tank Top due to Longitudi-

48 Double Amplitude of Sixth Order, Fore and Aft

49 Double Amplitude of Twelfth Order, Fore and Aft

50 Vibration on 26-ft Flat at About 100 RPM 111

51 Vibration on 36-ft Flat at About 100 RPM 112

52 Vibration of Bridge Deck at About 98.6 RPM 113

53 Vibration of Bridge Deck at About 102 RPM 114

54 Experimental Deflection Patterns on Tank Top

55 Experimental Deflection Patterns on Tank Top

56 Response to Shaker Excitation 117

Fi Longitudinal, Tangential, and Axial Wakes at

F2 Longitudinal, Tangential, and Axial Wakes at

F3 Longitudinal, Tangential, and Axial Wakes at

F4 Longitudinal, Tangential, and Axial Wakes at

F5 Longitudinal, Tangential, and Axial Wakes at