DAVID TAYLOR MODEL BASIN

WASHING TON. D. C.

SHIP VkbRATiQN

by

R. T. McGoldrick

Best Available Copy

December 1960
Report 1451
 TABLE OF CONTENTS
Page
P R E F A C E ........................................................................ ............................................. .. P- 1

N O T A T IO N .............................................................................................. ......................... ........ N. 1

CHAPTER 1 - HISTORICAL BACKGROUND .................................................................. 1- 1

CHAPTER 2 - QUALITATIVE DISCUSSION OF SHIP DYNAMICS ................. 2- 1

A. Introduction .................................................................................................................... 2- 1
B. Rigid Body Motions ....................................................................................................... 2- 1
C. Elastic Vibrations of the Entire Hull .......................................................................... 2- 3
D. Added Mass .................................................................................................................... 2- 6
E . L o c al E ffe c ts ............................................................................................................ 2- 8
F. Shallow WYater Effects .................................................................................................... 2- 8

CHAPTER 3 - BASIC BEAM THEORY OF SHIP VIBRATION ........................................ 3- 1

A. Introductioin ............................................................................... ................................ 3- 1
B. Basic Differential Equations for the Ship ................................................................. 3- 6
C. Methods of Calculating Natural Frequencies and Normal Modes
of Vibration of Ships .................................................................................................. 3-10

CHAPTER 4 - BEAM THEORY OF STEADY-STATE SHIP VIBRATION ...................... 4- 1

A . In tro du ctio n .................................................................................................................... 4- 1
B, Digital Method ................................................................................................................ 4- 3
C. Analog Method ................................................................................................................ 4- 5
D, Normal Mode Method ..................................................................................................... 4- 7
E. Method of Mechanical Impedance ................................................................................ 4-10

CHAPTER 5 - BEAM THEORY OF TRANSIENT SHIP VIBRATION .............................. 5- 1

A. Introduction ................................................................................................................... 5- 1
B. Normal Mode Method ..................................................................................................... 5- 2
C. Digital Method............................................................................................................. 5- 6
D. Analog Method ............................................................................................................... 5- 9

CHAPTER 6 -- EFFECT OF LOCAL FLEXIBILITY ON THE VIBRATORY

CHARACTERISTICS OF A HULL ............................................................ 6- 1
A. Introduction - The "Sprung Mass" Effect ................................................................ 6- 1
B. Local Elastic Structures ......... ........................................ 6- 2
C. Resiliently Mounted Assemblies ................................................................................. 6- 4
D. Summary of Effects of Sprung Masses on Hull Vibration ..................... 6-11

ii
 Page
CHAPTER 7 - PROPELLER-EXCITING FORCES ........................ ............................. T- I
A. Introduction .................................................................................................................. 7- 1
B. Pitch Unbalance Forces .................................................................................... .. .... 7- 2
C. Blade-Frequency Forces ......................................................................................... .. 7. 2
D. Experimental Force Data Available ................................. 7.-10

CHAPTER 8 - DAMPING OF HULL VIBRATION ........................................................ 8- 1

A. Introduction .................................................................................................................. 8- 1
B. Analytical Treatment of Hull Damping ............................... 8- a
C. Experimental Methods of Determining Damping ...................................................... 8- 5
D. Available Data on Hull Damping .............................................................................. 8- 7
E. Damping Action of Lifting Surfaces ........................................................................ 8- 7

CHAPTER 9 - ANTIVIBRATION DEVICES .............................. 9- 1

A. Introduction .................................................................................................................. 9- 1
B. The Tuned Vibration Neutralizer .............................................................................. 9- 1
C. Adjustable Rotating Eccentrics .................................... 9- 3
D. Synchronizing Devices ............................................................................................. 9- 5
E. Flexible Materials in the Vicinity of Propellers ........................ 9- 7
F . Summary ........................................................................................................................ 9- 7
CHAPTER 10 - DESIGN CONSIDERATIONS RELATING TO
STEADY-STATE HULL VIBRATION .................................................... 10- 1
A. Introduction ................................................................................................................ 10- 1
B. General Procedure ...................................................................................................... 10- 2
C. Reducing Propeller Forces ...................................................................................... 10- 5
D. Avoiding Hull Resonance ........................................................................................ 10- 6
E. Avoiding Local Resonance ........................................................................................ 10- 8
F . B alancing ........................................................................................................... ....... 10- 9

CHAPTER 11 - DESIGN CONSIDERATIONS RELATING TO

TRANSIENT VIBRATIONS ..................................................................... 11- 1
A. Introduction .................................................................................................................. 11- 1
B. The Hull Girder .......................................................................................................... 11- 2
C. Local Structures .......................................................................................................... 11- 4
D. Resiliently Mounted Assemblies .............................................. ...................... 11- 5

CHAPTER 12 -- DESIGN CONSIDERATIONS RELATING TO

VIBRATION OF THE PROPULSION-SHAFTING SYSTEM .............. 12- 1
A . Intro duction .................................................................................................................. 12- 1
B. Torsional Vibration ................................................................................................... 12- 2

iii
 Page
C. Longitudinal Vibration 12. 2
D. Lateral Vibration of Propeller Shafts ...................................................................... 12- 8

CHAPTER 18 - DESIGN CONSIDERATIONS RELATING TO

RESILIENT MOUNTINGS .................................................................. ... 18- 1

CHAPTER 14 - HYDROELASTICITY 14- 1

A. Introduction.14-1
B. R udder-H ull Vibration ................................................................................................. 14- 8
C. T he Singing Propeller ................................................................................................ 14- 7
D. Commentary ..................................................... 14- 8

CHAPTER 15 - SHIP VIBRATION RESEARCH ......................................................... 15- 1

A. Introduction..................................................... 15. 1
B. Vibration Generators .............................................. 15 2
C. Ship Vibration Instrum ents ........................................................................................ 15. 5
D. Experimental Techniques ......................................................................................... 15- 7
E. Correlating Theory and Experiment ........................................................................ 15-10

APPENDIX A - ILLUSTRATION OF A CALCULATION OF A VERTICAL

NORMAL MODE BY THE DIGITAL PROCESS .................................. A- 1

APPENDIX B - ILLUSTRATION OF A CALCULATION OF TRANSIENT

RESPONSE BY THE DIGITAL METHOD ............................................ B- 1
APPEND)X C - EMPIRICAL FORMULAS FOR HULL FREQUENCIES .................... C 1

APPENDIX D - EMPIRICAL FORMULAS FOR ESTIMATING THE LEVEL

OF STERN VIBRATION ................................... D- 1
APPENDIX E - SHIP VIBRATION SURVEYS .................................. E- 1

APPENDIX F - LEVELS OF SERVICE VIBRATION .................................................... .F- 1

APPENDIX G - SCALING CONSIDERATIONS IN MODEL VIBRATION

EX PER IMENTS .................................................................................... . 0- 1
APPENDIX H - MISCELLANEOUS INFORMATION ON VIBRATION OF
SHIPS IN SERVICE ............................................................................... . H- 1
GENERAL BIBLIOGRAPHY .............................................................................................. Bi- 1

INDEX .............................................................. In- 1

iv
 PREFACE
In treating the subject of ship vibration it is necessary to recognize that it falls in the
border region between exact science and empiricism. While it is true that intensive research
has been conducted in this field in several parts of the world and much progress has been
made, at r:' time of this writing (1960) it cannot be said that an adequate method of vibration
analysis of a ship exists.
It is true that one can consider the hull as an ideal mass-elastic system and can write
down a certain set of partial differential equations which govern the behavior of such a system,
but, in attempting to predict the level of service vibration of a ship in the design stage, one
must be well aware of the limitations of such a treatment.
It is also possible to present empirical data on the level of service vibration of ships
of various types together with the principal design features of the ships involved. This
approach is also inadequate since the level of vibration will vary with a number of parameters
simultaneously.
In the preparation of this book an attempt has been made to follow a path midway be-
tween the theoretical and empirical approaches. This leads to what may be called a "rational
theory of ship vibration." Ilse is made of th': properties of ideal free-free beams to obtain an
insight into the effects of various design changes on the vibratory response characteristics of
hulls. However, it is also attempted to guard the reader and the user against extending the
calculations into realms in which they have no validity. One aim is to show that the vibra-
tory characteristics are closely related to the structural strength characteristics.
While intended principally for the naval architect, the book has been prepared also
with the research worker and the student in mind. It has not been attempted, however, to
include a treatment of the fundamentals of mechanical vibration. It is presupposed that the
reader has or can acquire a background such as furnished by the courses in mechanical vi-
bration now given in practically all colleges of engineering. Specific references are included
at the end of each chapter and a general bibliography is given at the end of the book.
The contrast between the problem of avoiding serious steady-state vibration and with-
standing the effects of severe transient vibrations due to heavy seas is pointed out. The
problem of setting up design specifications with regard to vibration is also discussed.
However, no attempt is made to disguise the fact that the present state of the art of predicting
hull vibratory response characteristics is Primitive. Where controversial issues arise only
opinions can be furnished.
In the mathematical treatment of the subject and in the illustrations given in the
appendixes, the aim has been Lo emphasize the physical pri'nciples involved without burdening
the reader with too many details. It is assumed that the designer who makes use of the meth-
ods discussed in the book will assign the task of carrying out the actual vibration calculations
to a member of his staff who can consult the references when further details are needed.
Thus this book is not of the manual or handbook variety although concrete procedures for the

P-I
designer are suggested, With the rapid pace of development of computational aids today it is
clear that procedures that are written out in great detail may become obsolete almost over-
night, but the principles involved are durable.
Except as otherwise noted, the statements made apply to either surface ships or sub-
mersibles. Although most of the information given was obtained from research sponsored by
either the United States Navy or jointly by the Navy and the Society of Naval Architects and
Marine Engineers, the book is not intended specifically for naval designers, and problema
that are strictly naval are not discussed. Thus submarines are mentioned only because they
may become future commercial carriers and questions that relate to the detection of undersea
craft are omitted.
The relatively new field of hydroelasticity has been included since this is recognized
as a field of growing importance. In fact, in the broadest definition of hydroelasticity, the
subject of hull vibration itself would have to be included.
The book is based chiefly on the work of the U.S. Experimental Model Basin and the
David Taylor Model Basin and an exhaustive commentary on the work of other agencies is
not attempted here.
It is not overlooked that more elaborate analyses of the dynamical system comprising
the hull and the surrounding water than the beam-theory analysis presented in this book are
conceivable. However, it is felt that even after such analyses have been developed, the
designer will still be restricted to the methods discussed here in the preliminary design stage.
The data required for more elaborate analyses will, in general, be available only at a very
advanced stage of the design.
While this book is concerned chiefly with the problem of hull vibration, there has been
included among the chapters on design considerations, one dealing with the vibration of the
propulsion system itself. Here, however, the treatment is relatively brief and intended to
serve chiefly as a guide to. other sources of information on this subject in the technical
literature. In dealing with the hull itself, no attempt has been made to review all the avail-
able literature, but. to concentrate on the techniques that appear most fruitful.
In choosing a notation it was found impossible to adhere strictly Pither to standards
in naval architecture or in engineering since the subject involves both fields. In recent years
the American Standards Association has extended its sphere from acoustics into the field of
mechanical shock and vibration. Many of the symbols used conform to the ASA standards,
but the common symbols for the principal dimensions of ships used in naval architecture are
also retained. The common use of nondimensional notation in naval architecture has not been
followed here, as this has not found such wide acceptance in the field of mechanical vibration.
Nevertheless, it i pointed out in the chapter on hydroelasticity that the aeroelastician has
also fournd such notation preferable.
Finally it seems in order to point out that vibration theory plays a central role in ship
dynamics just as it does in mechanics in general. An acquaintance with the vibratory

P-2
characteristics of the hull not only assists the designer in avoiding serious vibration diffi-
culties when his product goes into service, but also gives him a deeper insight into many
factors involved in good structural design.

ACKNOWLEDGMENTS
Due to widespread interest in ship vibration Rear Admiral E.A. Wright, Director of the
David Taylor Model Basin at the time, authorized the preparation of a book on that subject
early in 1959. His enthusiastic support and encouragement made possible the preparation of
the present edition.
The author gratefully acknowledges constructive criticism and suggestions from the
following individuals during the preparation of the drafts from which the present manuscript
was prepared:
Dr. N.H. Jasper, Dr. E. Buchmann, Mr. R.C. Leibowitz,
Mr. N.L. Ficken, Jr., Dr. W.L. Haberman, and Mr. W.B.
Hinterthan of the David Taylor Model Basin; Mr. F.F.
Vane, Mr. 0.H. Oakley, Mr. D.L. Stevens, Mr. J. Vasta,
Capt. H.E. Saunders, USN (Ret), and Mr. L.K. Losoe of
the Bureau of Ships; and Mr. A.J. Tachmindji of the
Institute for Defense Analysis.

P 3
 NOTATION
As stated in the preface, the notation used in this book has been derived from the fields
of both naval architecture and mechanical engineering. It has not been feasible at this stage
to avoid using the same symbol with different meanings in different places. Hence, in listing
symbols here with specific meanings, the page on which the symbol first appears with this
meaning is given. Furthermore, in most cases the meanings of symbols are also given in the
text as they are used. As far as possible, symbols in naval architecture conform with Ref.
erence N-1; see page N-13. In formulas proposed by various authors, the original notation
has been converted in many cases to conform with that generally used in the book. In some
cases, however, it was considered preferable to retain the notation of the original author.
Just as it was found impossible to produce a wholly consistent notation, it was also
found necessary to use different systems of units in different places. The two principal sys-
tems used are the foo,-ton-second system and'the inch-pound-second system. A major excep-
tion is the frequent use of mils (thousandths of an inch) as a unit for displacement amplitude.
Where specific formulas for numerical computation are given, the units applicable to it are
given in the tex.
Since this book is not intended as an instruction manual for the use of those preparing
requests for vibration calculations to be made by the Applied Mathematics Laboratory of the
David Taylor Model Basin, the reader should make sure what units are currently in use with
the codings in operation at the time before initiating such requests. Reference N-2, page
N-13, will be found helpful in this respect.

Symbol Meaning P age

A Mechanical admittance based on displacement 4-11
A A ! t coefficient of a hydrofoil defined by the relation 8- 7
2
F, - A8 0
A Area of that portion of the cross section of a hull contributed by the A- 4
plating when plane is parallel to the direction of the shear load
(called the "web" area)
A Level of amplitude of vibration under calm sea conditions used as a H. 5
basis for comparison with rough sea conditions
Ad Mechanical admittance at driving point d (based on displacement) 4-12

a Acceleration 0- 2
ai Coefficient of the i th normal mode function in the series repre- 3-22
senting an arbitrary displacement pattern of hull vibration
B Beam of hull 8- 6
b Half-breadth of a ship section at the waterline A- 2

N-1
Symbol Meaning P age
b Semichord length of a hydrofoil or an aircraft wing 15- 8
C Schlick's empirical constant appearing in formula for the fundamental 8- 8
vertical frequency of a hull
C Lumped viscous damping constant of a hull (equal to oAx) 4. 3
C Viscous damping constant of-a vibratory system of a single degree 4-10
of freedom
0 Electrical capacitance 4-11

C Linearized (viscous) damping constant applicable to the translational 14- 4

degree of freedom of a control surface system of two degrees of free-
dom at zero velocity
C Lewis' two-dimensional added mass coefficient giving the ratio of the A- 2
added mass of a ship form (in vertical vibration) to that of a circular
form of the same beam
CM Moment coefficient of a spade rudder 14- 2
Cdi The effective viscous damping constant of a hull in its ith normal 4-10
mode of vibration and with respect to driving point d
0 Viscous damping constant equivalent to nonviscous dampihg on the 8- 2
basis of energy dissipation, damping force per unit velocity
o Linearized (viscous) damping constant applicable to the rotational 14- 4
degree of freedom of a control surface system of two degrees of
freedom at zero velocity
c Coefficient for the inertia effect of water appearing in Prohaska's C. 4
formula for the fundamental vertical frequency of a hull
0 Viscous damping constant, damping force per unit velocity in 3- 1
sense opposing the velocity
o Velocity of wave propagation 3- 5
C Distributed viscous damping constant of a hull, damping force per 4- 2
unit length per unit velocity
C, Angular viscous damping constant G- 2
eV Critical viscouo damping constant 8- 3
O/cc Ratio of damping to critical damping a- 2

cpm Blade frequency in cycles per minute 4-13

D Depth of hull 3- 6
d Draft A- 2
d A driving point in a mass-elastic system 4- 9
d Rectilinear displacement 0- 2

N-2
Symbol Meaning Page
d Propeller diameter 7- 5
E Young's modulus of elasticity 8- 2
E Electrical voltage D- 1
El Bending rigidity of a beam 3- 2
EVH Equivalent horizontal virtual inertia factor C- 6
/ effective added weight of water
for horizontal vibration
S+ displacement of ship

e Eccentricity of a rotating mass of a mechanical vibration generator 9- 4

F A concentrated force acting on a hull at an arbitrary point and treated 5-22
as constant over a short interval of time in the digital treatment of
transient response
F Force G- 2
F (t) A concentrated driving force (acting on a hull) which is an arbitrary 5. 4
function of time
F, Lift force acting on a hydrofoil 8- 7
FVP Rotating time vector representing the vertical component of the blade 9- 6
frequency force due to the port propeller
F,, Rotating time vector representing the vertical component of the blade 9- 6
frequency force due to the starboard propeller
Fo Area enclosed by the shell plating of the midship section of a hull C- 7
(not the area of the material)
f Frequency 7-11
G Shear modulus of elasticity 3- 4
0J Torsional rigidity of a shaft 3- 4
GJe Effective torsional rigidity of a hull with respect to its longitudinal 3- 8
axis
g Acceleration of gravity 8- 6
H Draft 7-11
H Impulse applied to a hull at driving point d 5- 5
h Distance from the axis of a control surface to the center of gravity of 14- 4
the rotating element (based on an allowance for added mass effect of
water) considered positive if the e.g. is downstream
h Distance from the top of a polemast to the elastic axis of a hull 11. 4
I Moment of inertia of the cross section of a beam with respect to its 8- 2
neutral axis (based on the area of the material)

N-8
Symbol Meaning P age
I Mass moment of inertia of the rotatable assembly of a control surface 14- 4
system of two degrees of freedom including the added mass moment of
inertia effect of the water
I Electrical current D- 1
I Mass moment of inertia of the entire ship with respect to the longitu- D- 8
dinal axis through its center of gravity without any allowance for the
inertia effect of the surrounding water
I Mass moment of inertia 0-2
1a Moment of inertia of an area 2-2
C- 6
IH Moment of inertia of the area of the midship section of a hull for
bending in the horizontal plane

1. Mass moment of inertia of a resiliently mounted assembly or "sprung 6-10

mass" with respect to an axis through its center of gravity
IV Moment of inertia of the area of the midship section of a hull for C- 6
bending in a vertical plane
!x, Iy, I, Mass moments of inertia of a resiliently mounted rigid assembly with 6- 8
respect to the X-, Y-, and 2-axes, respectively, with origin at the
center of gravity of the assembly
lXY, I.Z, Mass products of inertia of a resiliently mounted rigid assembly with 6- 8
Iy2 respect to axes X-Y, X-Z, and Y-Z, respectively, with origin at the
center of gravity of the assembly
LZ Mass polar moment of inertia of a beam or shaft per unit length with 3- 4
respect to its longitudinal axis
I Mass moment of inertia of a hull per unit length with respect to the 8. 9
x-axis including the allowance for the inertia effect of the surround-
ing water

lZ Rotary inertia of hull per unit'length (difference between the mass 3- 3

moment of inertia of the hull including the effect of added mass of
water and the value that would apply if all the mass were con-
centrated at tie longitudinal axis)

J Polar moment of inertia of the section area of a beam or shaft 3- 4

(based on the area of the material)

J Propeller advance ratio - inflow velocity 7- 6

nd

J Longitudinal coefficient applied by F.M. Lewis to values of A- 2

added mass of water in ship vibration to correct for departure from
two-dimensional flow

J80 Effective polar moment of inertia of the midship section area of a 3- 6

hull (based on the area of the material)

N-4
Symbol Meaning P age
j
/ JC (imaginary unit) 3-18

K Spring constant of a vibratory system of a single degree of freedom 4-10

K Translational spring constant of a control surface system of two 14- 4
degrees of freedom
K Shear rigidity factor for beam or hull such that the slope of the 8- 8
deflection due to shear is equal to the total shearing force at the
section divided by KAG, where A is the cross section area (of the
material) and G is the shear modulus of elasticity

K Pressure coefficient = 7- 4
pn2d 2

KT Thrust coefficient - T 7- 5
pn72 d-4

Kdt The effective spring constant of a hull in its ith normal mode of 4- 9
vibration and with respect to driving point d
Kuv Spring constant of an entire set of resilient mountings relating a 6. 7
displacement of the mounted assembly in the Y-direction with the
restoring force in the X-direction and conversely. A displacement
v in the positive Y-direction evokes a force - K. vtr in the X-
direction; if 'v and K "' are both positive the force is directed toward
-a. Similarly, a displacement u toward +x evokes a force -Ku,, u
in the Y-direction

K,, Ka, Spring constants of an entire set of resilient mountings defined by 6. 7

etc obvious extension of definitions of K.,, and Ku,. For Ku., K00 etC.,
the same axis is used twice
K,,p Spring constant of an entire set of resilient mountings giving either the 6- 7
restoring force in the X-direction due to a unit rotation of the mounted
assembly about the Y-axis, or the restoring torque about the Y-axis due
to unit displacement of the assembly in the X-direction. The sign
convention corresponds to that for KV
Kg Spring constant of entire set of resilient mountings installed between 6-12
the cradle and the hull in a compound isolation mounting system
KII Spring constant of an entire set of resilient mountings installed between 6-12
the assembly and the cradle in a compound isolationi mounting system
determined by holding the cradle fixed
Kn' Spring constant of an entire set of resilient mountings installed 6-12
between the assembly and the cradle in a compound isolation
mounting system determined by holding the assembly fixed
KAG Shear rigidity of a beam or hull a_ 8
k.E. Kinetic energy 4- 9

N-5
Symbol Meaning Page
KW Power G- 2
k Spring constant, restoring force per unit displacement 3- 1

k Horn's empirical coefficient appearing in formula for fundamental 3- 6

torsional frequency of a hull
k Torsional spring constant of a rudder-steering system 14- 2
k Torsional spring constant of a control surface system of two degrees 14- 4
of freedom
k' Angular spring constant G- 2
k. Axial spring constant of a resilient mounting 6- 6
ke Effective spring constant of a local ship structure referred to its
center 6- 8
of gravity

kI Generalized elastic constant of a mass-elastic system applicable to 4- 8

its i th normal mode of vibration
kr Radial spring constant of a resilient mounting 6- 6

L Length of a hull (usually assumed to be the distance between the 3- 3

forward and after perpendiculars)
L Electrical inductance 4-11

L Distance from the axis of a control surface to the center of lift, 14- 5
considered positive if the center of liUt is upstream
L A characteristic length or dimension of a ship 15- 8

M Bending moment 3- 2
H Mass of a vibratory system of a single degree of freedom 4-10
M That part of the mass of a control surface system of two degrees 14- 4
of freedom which can vibrate only in translation
N Total mass of a uniform bar C- 5
Al Imaginary component of the rotating time vector representing a 4- 4
vibratory bending moment
Mi Generalized mass of a mass-elastic system applicable to its ith 4- 8
normal mode of vibration
Mdi Effective mass of a hull in its ith normal mode of vibration and 4- 9
referred to the driving point d
meeg Effective mass of a local ship structure referred to its center of gravity 6- 3

MO Hydrodynamic moment acting on a spade rudder 14- 2

in Mass of a rigid body 3- 1

N-6
Symbol Meaning P age
1'I
M Mass of the rotatable element of a control surface system of two 14- 4
degrees of freedom including an allowance for added mass effect
of the water
MS Mass of a resiliently mounted rigid assembly or "sprung mass" G-10
flexibly supported in a hull
N Frequency of fundamental vertical flexural mode of a surface ship 8- 3
N Maximum rpm of a rotating member 10-10
N' Number of significant vertical flexural modes of a hull 4- 7
Ne Fundamental torsional frequency of a hull $_ 6

NE Frequency of the 2-node horizontal flexural mode of a hull C- 6

Na Predicted fundamental vertical natural frequency of a new ship C- 2
AIV Frequency of the 2-node vertical flexural mode of a hull C- 6
N0 Known fundamental vertical natural frequency of an old ship C- 2
n- Frequency of a simple harmonic vibration 3-11
n Revolutions per second (rps) 7- 5
n Rpm of a rotating member 10-10
n, Frequency of the fundamental mode of vibration of a system 3- 5
P Single amplitude of a simple harmonic driving force P sin Cal 4- 9
P0 Single amplitude of the vertical component of the propeller D- 2
oxciting force (at blade frequency)
P0 Single amplitude of a simple harmonic driving force 3- 1
P.E. Potential energy of a vibrating beam 4- 8
P (t) Concentrated driving force acting on an element of a hull of 4- 3
length Ax

P (x, t) Driving force per unit length acting on a beam-in a direction 3- 2

normal to the X-axis
Ph (t) Hurizontal component of the force produced by a rotating 9- 4
eccentric mass
Pv (t) Vertical component of the force produced by a rotating 9- 4
eccentric mass
p Pressure 7- 5
Q Resonance magnification factor 8- 4

Qdi(t) Generalized driving force on a hull applicable to the ith normal mode 5- 4
and referred to the driving point d

N-7
Symbol Meaning Page
Q,(t) Generalized driving force on a hull applicable to the ith normal mode 5. 3
but without reference to any specific driving point
Mass distribution coefficient appearing in Prohaska's formula for the C- 8
fundamental vertical frequency of a hull
q A generalized displacement in matrix notation 6- 9
q Number of cycles used in estimating the logarithmic decrement from a 8- 5
record of freely decaying vibration
q,(t) Generalized displacement of a vibrating beam in its ith normal mode 4- 8
R Electrical resistance 4-11
R Lever arm of weight unbalance of a rotating member 10-10
R Factor appearing in Prohaska's formula for the fundamental vertical C- 3
frequency of a hull
r J. Lockwood Taylor's shear correction factor C- 3
r1 Correction factor for variable inertia used in applying Prohaska's formula C- 3
for the fundamental vertical frequency of a hull
Correction factor for shearing force used in applying Prohaska's C- 3
formula for the fundamental vertical frequency of a hull
r3 Correction factor for transverse compression and dilitation used in C- 3
applying Prohaska's formula for the fundamental vertical frequency
of a hull
S Velocity of undisturbed water relative to a hydrofoil 8- 7
8 Distance from a fixed point measured along the shell plating of a hull C- 7
in a plane normal to the longitudinal axis of the hull
T Torque with respect to the longitudinal axis of a cylindrical shaft 3- 4
T Moment about the longitudinal axis of a hull due to all shearing 3-8
stresses in the cross section
T Torque G- 2
T Propeller thrust 7- 5
7O Single amplitude of blade-frequency driving torque with respect to the 4-13
longitudinal axis of a hull
T0 Single amplitude of blade-frequency exciting couple with respect to the D. 8
longitudinal axis of a hull
t Time 3- 1
t Tip clearance between propeller and hull (in the plane of the propeller) 7- 5
U Maximum allowable residual unbalance of a rotating member 10-10

N-8
 Symbol Meaning P age
u A displacement in the X-direction 6- 7

u, v, U Displacement in the X-, Y-, and Z-directions, respectively 6- 7

V The net shearing force in the direction of flexural vibration 3- 7

transmitted by one section of a hull to the adjoining section
V The real component of the rotating time vector representing 4- 3
a vibratory shearing force
V Velocity of an aircraft wing relative to the undisturbed air 15- 8
V Velocity of a ship 15- 8
V Volume 0- 2
V, The imaginary component of the rotating time vector representing 4- 4
a vibratory shearing force
Vv
(
Vertical virtual inertia factor
added weight of water for
vertical vibration
+ displacement of the
C- 6

ship
v A displacement in the Y-direction 6. 7

v Rectilinear velocity G- 2

W Energy dissipated per cycle in a simple harmonic vibration in the 8- 3

presence of damping
W Weight of a rotating member 10-10
WR Weight unbalance of a rotating member 10-10
w Displacement in the Z-direction 6- 7
X, Y,Z Rectangular coordinate axes'fixed in space 2- 2
X11 Y1
1 Z1 Rectangular coordinate axes with origin at the center of gravity of the 6-12
cradle in a compound isolation mounting system
X2' Y2) Rectangular coordinate axes with origin at the center of gravity of the 8-12
2,2 assembly in a compound isolation mounting system
Distance in the longitudinal direction forward of the plane of the
propeller
z Displacement in the X-direction 7- 6
X, The X-coordinate of a point r on a beam subject to vibration 4- 8
Y A rectangular coordinate axis fixed in space 2- 2
Y The displacement of points of a hull in the Y-direction when 2- 6
vibrating in one of its normal modes of vibration

N-9
Symbol Meaning P age
Y The single amplitude of a vibratory system of a single degree of freedom 4-11
Y The steady-state single amplitude of vibration at the stern of a ship due 4-12
to a simple harmonic driving force of single amplitude P0
Y Displacement in the Y-direction of the axis of rotation of a control 14- 4
surface member
Ydi The single amplitude of a hull in its ith normal mode of vibration at the 4- 9
driving point d
Y"(Z) Amplitude pattern assumed as a starting mode shape in the calculation 3-19
of a hull flexural mode by the Stodola method
ya,(-) Amplitude pattern used in the calculation of a hull flexural mode by the 3-20
Stodola method and obtained from Y.(-) by a parallel shift of the X-axis
Y.'(") Amplitude pattern used in the calculation of a hull flexural mode by the 3-20
Stodola method and obtained from Y,,(-) by a combination of a parallel
shift and a rotation of the X-axis
Y (W) An arbitrarily assumed normal mode pattern of vibration of a hull 3-22
Y (ir) Pattern of displacement in the Y-direction of a hull vibrating in its 3-22
i th normal mode
y Displacement in the Y-direction 3- 2
y Real component of the rotating time vector representing a vibratory 4- 3
displacement in the Y-direction
y Displacement in the Y-direction of the center of mass of an element 3- 8
of the hull of length Ax
y Imaginary component of the rotating time vector representing a vibratory 4- 4
displacement in the Y-direction
y Displacement in the Y-direction of the center of shear of the cross 3- 8
section of a hull

y Velocity in the Y-direction 3-16

y Acceleration in the Y-direction 3-12
Yog Single amplitude of vibration in the Y-direction of the center of 6- 3
gravity of a local ship structure

Yrs The amplitude of vibration at a point s of a beam due to a simple 4- 8

harmonic driving force applied at point r
yn;s Displacement in the Y-direction at the nth station of the hull at the s 5- 8
interval of time in the digital calculation of transient response of a hull
(z) Mode shape obtained in the calculation of hull modes by the Stodola2 3-20
method on a graph in which its magnitude differs by the factor i/6)
from that of the curve assumed in starting the calculation

N- 10
Symbol Meaning Page

Z Mechanical impedance based on displacement 4-11

Z Electrical impedance D- I
Z Electrical impedance of a circuit having resistance, inductance, and 4-11
capacitance all in series

ZV Mechanical impedance based on velocity 4-11

Zdi Mechanical displacement impedance of a hull in its ith normal mode of 4-11
vibration at driving point d
z Number of blades per propeller 7- 4

SZ-coordinate of the center of mass of an element of a hull of length 3- 7

Ax (including allowance for added mass of water)

SZ-coordinate of the center of shear of the cross section of a hull 3- 7

vibrating flexurally in the Y-direction
a Empirical constant appearing in impedance-type formulas for stern 7-11
vibration of a ship
a Angle of attack of a spade rudder 14- 2

aA Empirical constant in impedance-type formula for athwartship 7-11

vibration of hulls
Ca, Empirical constant in impedance-type formula for torsional 7-11
vibration of hulls
a Empirical constant in impedance-type formula for vertical 7-11
vibration of hulls
y, Angular displacements with respect to the A-, Y-, and Z-axes, 6- 7
respectively

Angular displacement with respect to the Y-axis 6- 7

/3 Section area coefficient A- 2
SEmpirical coefficient appearing in the formula of Todd and Marwood C- 5
for the fundamental vertical frequency of a hull

Component of the slope of the elastic line of a hull due to shearing 5- 2

only
, Real component of the rotating time-vector representing a vibratory 4- 3
angular displacement of the cross section of a hull with respect to a
Z-axis

Y Angular displacement with respect to the Z-axis 3- 8

Y Imaginary component of the rotating time-vector representing a 4- 4

vibratory angular displacement of the cross section of a hull with
respect to a Z-axis
A Displacement of a ship 3- 3

N-11
Symbol Meaning Page
AS A small distance along the shell plating in the plane of the section C- 6
Ax Length of a small element of a hull measured in the direction of its 3- 7
longitudinal axis
a Logarithmic decrement for a free vibration 8- 3
a Shell plating thickness C- 6
0 Angle of attack of a hydrofoil 8- 7
0 Angular displacement of a control surface from its equilibrium position 14- 4
6 Angular displacement 0- 2
IX Complex exponential term in the expression for vibratory motion whose 14- 5
real part indicates the rate of decay or buildup and whose imaginary
part indicates the circular frequency (A - + j )
A Scale factor by which a dimension of a ship is multiplied to obtain the G- 1
corresponding dimension for the ship modcl
X Wavelength 3- 5
Frequency of free vibration in the ith normal mode in the presence of 5- 5
damping
p Mass per unit length 3- 8
p Real part of complex exponential term in the expression for vibratory 14- 6
motion indicating the rate of decay or buildup of the vibration, that is,
the degree of positive or negative damping

p Radius of curvature of the elastic line of a deformed beam 3- 2

p Mass density of water 7- 5

Time at any instant between 0 and t 5- 5
A Rotation of the cross section of a beam or hull with respect to its 2- 4
longitudinal axis
Phase anglo by which the driving force leads the displacement in a 4- 9
simple harmonic vibration
Steady-state single angular amplitude of a hull at the stern and with 4-13
respect to its longitudinal axis due to a simple harmonic driving
torque of single amplitude TO
Empirical coefficient appearing in Burrill's formula for the C- 3
fundamental vertical frequency of a hull

Single amplitude in rotation about the longitudinal axis of a hull D- 3

95i (x) The ith function of a series of orthogonal functions of X 4- 8
Circular frequency of a simple harmonic time-varying quantity 3- 1

N- 1I
Symbol Meaning Page
I A column matrix 6- 9
[ A matrix 6- 9
The midship section of a hull A- 1
Equals approximately 4-38
Designates differenkiiation with respect to time, when over a symbol 8- 1
Designates double differentiation with respect to time when over a 8- 1
symbol

Equal to or less than 5- 7

REFERENCES

N-1. Saunders, H.E., "Hydrodynamics in Ship Design," published by SNAME (1957).

N-2. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).

N-18
 CHAPTER ]

HISTORICAL BACKGROUND

Although ship vibration phenomena were undoubtedly encountered much earlier, the
subject appears to have first attracted scientific study toward the end of the 19th century.
As might be expected, the impetus came from the occurrence of resonance which indicated
the need for a method of predicting the natural frequency of vibration of the hull.
In 1894 0. Schlick" 1' proposed a formula for the fundamental vertical hull frequency
and this provided the naval architect of that time with a guide for designing his propulsion
system so that the operating propeller shaft rpm would not coincide with this hull frequency.
Other evidence of interest in the subject in this early period is the French textbook
"Theorie du Navire," 1-2 published in 1894 which included a chapter on ship vibration among
its four volumes. In that work, examples were cited in which the rated speed of ships had to
be reduced to avoid hull vibration. It does not detract from the pioneering contribution of the
authors of that classical work that they were lhd astray ir, their speculations regarding hull
vibration by the observation that the ratios of natural frequencies of hulls to the fundamental
frequency corresponded more nearly to those of the string than to those of the solid bar with
free ends.
Another early investigator in this field was A.N. Krylov who recorded hull vibration on
a naval cruiser in 1900. 1-3 His work on both the theoretical and practical aspects of the sub-
ject led to a complete book on the subject of ship vibration published in 1936. 1-4 This work
is devoted chiefly to the fundamentals of mechanical vibration and the application of classi-
cal beam theory to the hull vibration problem.
Increasing interest in the subject is evident in the technical literature from about 1900
to World War II. A picture of the status of the development of the theory of hull vibration
around 1982 is given by the paper of E. Schadlofsky 1-5 where it is suggested that the fuln-
damental vertical frequency of the hull can be estimated by a beam-type analysis involving
graphical integration. This process, based on the method of Stodola, 1-6 is discussed in de-
tail in Chapter 3. As indicated in the bibliography on page Bi-1, numerous other authors have
explored the application of beam theory to the analysis of hull vibration.
About the time Schadlofsky's paper was published, considerable impetus was given
to the experimental phase of ship vibration research by the manufacture of machines capable
of vibrating entire hulls. This development took place in Germany where such machines had
been previously designed by the firm of Losenhausen in Dusseldorf for the dynamic testing
of riveted and welded bridges. These machines contained adjustable eccentric masses so
arranged that unidirectional sinusoidal forces and couples could be produced, as discussed
in Chapter 15.

-1References are listed at end of each chapter. For complete bibliography, see page Bi-L.

1-1
 The largest machine of this type was delivered to the U.S. Experimental Model Basin
in Washington, D.C. in 1981. Starting from that date, the U.S. Experimental Model Basin and
its successor, the David Taylor Model Basin, continued to maintain and develop machines of
this general type and to conduct experiments to verify theoretical predictions of hull vibration
characteristios.
Since World War I1 ship vibration research has been carried on at an expanding rate by
all the principal maritime nations of the world. This is evident from the bibliography. In the
United States, the Souiety of Naval Architects and Marine Engineers has done much to stim-
ulate interest in ship vibration and has cooperated closely with the Bureau of Ships of the
Navy Department in this field. Two of its research panels, in particular. have been directly
concerned with the ship vibration problem.
In recent years the development of analog and digital computers has contributed greatly
to the development of hull vibration analysis. 1,7 Simultaneously, experimental techniques have
been devised to determine the vibratory response characteristics of the hull 1-7 as well as the
forces tending to excite vibration in the hull. 1-8 So broad is the horizon that has been made
visible by modern developments in computing techniques that methods of vibration analysis
entirely independent of the beam theory of the hull are now under investigation. These
methodb are along the line suggested by Professor H.A. Schade in his discussion of Refer-
ence 1-7. In these "three-dimensional" analyses, the restriction that all points at the same
cross section of the hull partake of the lateral motion of the "hull girder" is removed. No
results of these investigations, however, are available at this time (1960).

REFERENCES

1-1. Schlick, 0., Series of Articles on Ship Vibration in TINA (1884, 1893, 1894, 1901,
and 1911).
1-2. Pollard, J. and Dudehout, A., "Theorie du Navire,"-Treizieme Partie, "Vibrations
des Coques des Navires a Helice," Vol. IV, Chapt. LXIX, "Causes et Periodes des Vibra-
tions," Paris (1894).
1-3. Babaev, N.N., "Contemporary Status of Development of the Problem of Ship Vibra-
tion," TM3 Translation 291 (Jul 1959) from Russian paper in Sudostroenie, No. 3 (161)
(Mar 1956).
1-4. Krylov, A.N., "Vibration of Ships," in Russian (!936).
1-5. Schadlofsky, E., "The Calculation and Measurement of Elastic Natural Frequencies of
Ship Hulls," STG, Vol. 33 (1932). Also EMB Translation 7 (Jun 1934). Supplement (Nov 1935).
1-6. Stodola, A., "Steam and Gas Turbines," (Dampf-und Gasturbinen). Translated by
Dr. Lewis C. Lowenstein, P. Smith, New York (1945).

1-2
 1-7. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 68 (1955). Also TMB Report 1060 (Jul 1956).
1-8. Pien, P.C. and Ficken, N.L., "The Measurement of Propeller Induced Vibratory
Forces on Scale Ship Models," Paper presented at ATTC Conference (Sep 1959).

1-3
 CHAPTER 2
QUALITATIVE DISCUSSION OF SHIP DYNAMICS

A. INTRODUCTION

Unlike the case of a body frpe in space, the dynamical system considered in this book
must include both the body itself and its surrounding medium. The density of the medium in
this case is comparable with the density of the material of which the vehicle is constructed;
this is contrary to the situation in aeronautics. It is therefore to be expected that the water
will have a marked effect on the dynamical behavior of the ship, and there is abundant evi-
dence that this is the case.
The forces exerted by the water on the hull arise either from pressure, which acts in a
direction normal to the hull surface at any point, or from friction or shear, which acts in a di-
rection tangential to the surface. As far as rigid body motions are concerned, when these two
sets of forces are integrated over the wetted surface of the hull, the entire system of forces
can be roduced to effective forces acting at the center of gravity in each of the three principal
directions (vertical, longitudinal, and athwartship) and effective moments about the three axes
through this point. In general, these forces and moments depend not only on the rectilinear
and angular displacements of the hull with respect to these axes but also on the rectilinear
and angular velocities and accelerations; or, in the case of rough seas, on the motion of the
water surface relative to the ship.
Concurrently with these rigid body motions the hull may execute elastic vibrations of
numerous types. Although these latter vibrations are the main subject of this book, they can-
not be considered as entirely independent of the rigid body motions. In fact, in rough seas
the rigid body motions frequently lead to vibrations accompanying large hydrodynamic impacts,
and, even in calm seas, the forward motion of the ship may generate hydrodynamic flow exci-
tations o? different types. It is shown in Reference 2-1 that, although the effect of buoyancy
may be detectable for the frequency of the fundamental mode of vibration of long, slender hulls,
A is 111 genirai ju;iLlable W neglect the effect.

B. RIGID BODY MOTIONS

When considered as a rigid body, a ship has six degrees of freedom, a id hence there
are six displacement-like quantities to be taken into account in completely specifying its
motion. The steady forward velocity, the only motion desired in the normal operation of the
ship, is not ordinarily considered in discussing its rigid body motions. They are the motions
superimposed on this steady forward velocity by the sea action, and always involve time-
varying velocities and accelerations. With reference to the axes shown in Figure 2-1, the
rigid body displacements in traislation in the X-, Y-, and Z-directions are called, respectively,
surge, sway (or sidling), and heave, whereas the angular displacements about the same axes

2-1
are called roll, pitch, and yaw. Of these six displacements, the three most important are roll,
pitch, and heave.
Although right-hand systems of coordinate axes are used throughout this book, it has
been found convenient, where elastic vibrations are dealt with, to orient the Y-axis in the
direction of the vibration. Thus when horizontal (athwartship) hull vibration is under discus-
sion the axes are oriented as in Figure 2-1, but when vertical hull vibration is discussed the
Y-axis is taken vertical and the Z-axis horizontal (with positive direction out of the paper.)

Z
,/Y
Figure 2-1 - Axes through the Center of Gravity of a Ship with Respect to
Which Forces and Moments Exerted by the Water May Be Defined

In the dynamics of rigid bodies, motions are defined in terms of the translation of the
center of mass and rotation about the center of mass. While this procedure is also applied
to the ship, it must be realized that, since the hull is elastic, these relatively slow motions
ara nhso accompanied by elastic deformations. These elastic deformations are not the ones
considered in the discussion of hull vibration. Furthermore, the hull by itself is not usually
considered as an isolated body because the component of the water forces due to acceleration
is usually accounted for by adding mass to the hull mass to take care of this inertia effect.
Since the rolling, heaving, and pitching motions, although slow, are still oscillatory and thus
have the essential characteristics of vibrations, it is important to distinguish them from the
elastic vibrations which, as has been stated, are the main subject of this book.
In the absence of an external alternating force, a body in free space could not execute
motions in which the displacement of the center of mass was oscillatory. The ship is sub-
ject to the constant force of gravity and to gravity mnments which vary with its angular dis-
placements about axes other than those through its center of gravity. The buoyancy moments
accompanying rolling, heaving, and pitching are due to buoyancy forces that vary with these
motions. Although elastic deformations in general accompany these motions, they are too
small to play an essential role in determining these motions, and the term "rigid body mo-
tions" is retained to distinguish them from the motions of the hull in which the elastic de-
formations do play an essential role.

2-2
 A ship stopped in a calm sea, if initially disturbed, will execute damped rigid body
motions and eventually again come to rest. In disturbed seas, the surrounding water provides
not only the restoring forces and moments necessary for oscillatory rigid body motions and the
forces assooiated with the added mass, but also the forces and moments required to maintain
these motions in the presence of damping.

C. ELASTIC VIBRATIONS OF THE ENTIRE HULL

When a ship is subjected to an impulsive load, such as occurs when a descending an-
chor is suddenly arrested, it will execute elastic vibrations in addition to whatever rigid
body motions are excited. Of these vibrations some are observed only locally and some are
observed throughout the hull. The latter, in general, are of the type that may exist in a beam
free in space and so are called "boamliko." Although the surrounding water plays an impor-
tant role in these vibrations, it does not destroy their beamlike characteristic and it is help-
ful to consider the vibrations of the ideal solid beam free in space. This is frequently spoken
of as the free-free beam (both ends free).
As emphasized in standard works on mechanical vibration, 2 2 , 2-3, 24 the two terms
"modes" and "nodes" are used repeatedly in the discussion of continuous systems and must
not be confused with each other in spite of the similarity in spelling. Thus the mode is the
pattern or nordiguration which the body assumes periodically while in the vibratory condition,
whereas the node is a point in the body which has no displacement when the vibration is con-
fined to one particular mode. "Normal mode" of vibration is another very common term. The
normal modes are the patterns in which the body can vibrate freely after the removal of ex-
ternal forces.
A beam free in space may undergo four principal types of elastic deformation designated
as bending, twisting, shearing, and extensional deformations. These may all occur simulta-
neously. In a solid beam, these same types of deformation may exist with respect to any of
the three principal directions even though the relative magnitudes of bending, shearing, and
torsion may be very different with respect to the different axes. In the case of the ship, the
elastic doformtions that play a significant role in its vibration arc limited to bending and
shearing in both the vertical and horizontal planes through its longitudinal axis, and to tor-
sion about the longitudinal axis. The identification of extensional (longitudinal) beamlike
vibrations of hulls has so far been inconclusive, and this type of vibration is ordinarily con-
sidered insignificant in ships although it may be quite significant in the propulsion systems
themselves, as shown in Chapter 12.
In a symmetrical beam the bending and shearing effects combine to produce what are
usually called the flexural modes, as illustrated ira Figure 2-2.
The curves plotted in Figure 2-2 indicate the displacements in the Y-direction of points
falling on the A-axis when the bar is at rest. Similar modes exist for displacements in the
Z-direction.

2-1
 Y

/X

S2- Node Mode

'4 3- Node Mode

a,

a 4-Node Madi

X.-

Figure 2-2 - Flexural Modes of a Free-Free Uniform Bar

Figure 2-3 illustrates the torsional modes in which a uniform beam

may vibrate, and
the curves plotted show angular displacement versus distance from
the end.

-Node Z
Mode

Y
0X
CL S3- Mods

M-ode

Mode

Figure 2-3 - Torsional M.odes of a Free-Free Uniform Bar

In both the flexural and torsional types of vibration, a natural frequency

is associated
with each pattern of vibration and the natural frequencies increase
as the number of nodes
(points at which the curves cross the X-axis) increases.

2-4
 I--
- x

2- Node Mode
L.6

3- Node Mode

E
* 4- Node Mode

0X

Figure 2-2 - Flexural Modes of a Free-Free Uniform Bar

Figure 2-3 illustrates the torsional modes in which a uniform beam may vibrate, and
the curves plotted show angular displacement versus distance from the end.

F"Node z
C Mode
.-
CY
5

40
S0 2-Node ' .

0
E

"4i-Node
Mode

Figure 2-3 - Torsional Modes of a Free-Free Uniform Bar

In both the flexural and torsional types of vibration, a natural frequency is associated
with each pattern of vibration and the natural frequencies increase as the number of nodes
(points Rt which the curves cross the X-axis) increases.

2-4
 If a free-free beam is unsymmetrical with respect to either the vertical or horizontal
plane through its longitudinal axis, it will be found that its natural modes of vibration involve
torsion, bending, and shearing simultaneously. Thus from Figure 2-4 it is clear that if the
left end of the bar accelerates in the V-direction, the bar will tend to twist because of the
inertia of the vertical member whose center of mass lies above the X-axis.

Figure 2-4 - Free-Free Bar Whose Normal Modes of Vibration Involve

Combined Torsion, Bending, and Shearing

The normal modes of vibration of the ideal free-free beam are independent of one
another, and, under an impact, the beam may vibrate in several of these modes simultaneously.
However, such a system has the property that, if it is initially deformed into a pattern cor-
responding to any of its normal modes of vibration, it will thereafter vibrate only in that mode
and at the ?requency associated with that more.
Patterns of two typical torsion-bonding modes of a hull are illustrated in Figure 2-5.
A hull, of course, is a much more complicated structure than a solid beam. It behaves
like the free-free beam only in its lower modes of vibration. Hence these modes are said to
be beamlike,and they may be excited by either transient or steady-state disturbancos. The
transient disturbances are due to wave or slamming impacts which induce trains of damped
vibrations in one or more of these modes simultaneously. Steady-state vibrations are caused
by rotating unbalanced engine or machine elements, unbalanced propellers, or unbalanced
shafting. Vibration may also be set up by nonuniformity of pitch among the blades of a
propeller and, above all, by the variation in load on the individual blades as they rotate in
the nonuniform velocity field in the propeller race. The propellers also cause pressure flue-
tuations on the surface of the hull wid appendages in their immediate vicinity. Propeller
blade excitation is the chief cause of steady-state ship vibration at this time (1960).
A common characteristic of the forced propeller-excited vibration of ships is that it is
concentrated in the stern. Since the beamlike modes of vibration involve large amplitudes at
both ends of the hull, this phenomenon obviously does not result from vibration in a single

2-5
 "71'
normal mode. It is apparently due to the resultant of the nonresonant responses in several
normal modes, as explained in Chapter 4.
F.P.

A.P. / '

"Y"
(+) '.(+).

Station
Y 22 20
Translation Translation X

Station 0 Station Station

z 02
y y
Rotation ttin Rotation
Rotattion

Figure 2-5 - Two Torsion-Horizontal Bending Modes of a Hull Having

Reversed Phase Relations between Rotation and Translation

D. ADDED MASS

Before proceeding with the qualitative discussion of ship dynamics, it is necessary to

give further attention to the inertia effect of the surrounding water as it relates to the elastic
vibrations of the hull. While it is true that the water can actually exert only normal pressures
and frictional forces on the hull, these forces may be broken down into components which have
different time rates of change. A component of the water forces that is proportional to the
acceleration of the hull at the point of interest and opposite to the acceleration in direction
yields here an effect of increased inertia. The relatively high density of water makes this
inertia effect of serious concern in the vibration of ships and underwater ship components.
Unfortunately, a variety of terms have been used in the technical literature for desig-
nating this water inertia effect. There are two distinct concepts responsible for some of the
confusion in terminology: (1) the apparent increase in mass of a bcdy vibrating in water; and
(2) the apparent total mass of the body (including the effect of the water). Such terms as
"added mass," "virtual mass," "apparent mass," and "apparent added mass" will all be
found in various publications on the subject. Although reference must inevitably be made to
papers in which such terms as "virtual mass" and "apparent mass" appear, only the term
"added mass" will be used hereafter in this book for the inertia effect of the water. This
conforms to Reference 2-5.

2-6
 Not on- ''..ero re&-h confusion about terminology regarding water inertia but there is
also a growiig .igU that 0 concept of added mass of bodies vibrating in water has out-
lived its u.qefuliess. This arise,:- from the fact that as the frequency increases, the assump-
tion of incompressibility of the water on which the added mass connept is based becomes
untenable and the vibrating body becomes, in effect, a source of une:,cwater sound.
In the theoretical treatment of the added mass effect, the flow pattern for vibratory
rigid body motion is considered the same as for steady or unidirectional motion. Moreover,
when a rigid circular cylinder with its axis lying in the plane of a free surface of water is
vibrating vertically, it is assumed that the flow pattern is the same below the surface as if
the cylinder were deeply submerged. This justifies the treatment as one for a circular
cylinder in infinite fluid and the subsequent discarding of half of the added mass that would
apply to the deeply submerged case; see Figure 2-6.
Y

Figure 2-6 - Pattern of Flow about a Circular

Cylinder Mo,'ing with Constant Velocity A
in the Positive Y-Direotion in an
Unbounded Fluid Medium

The theoretical derivation of added mass for vibrating hulls (by considering the flow
about ideal bodies in incompressible fluids) is based on the assumptions used in classical
hydrodynamics (potentin! flow). (See References 2-6 and 2-7.) The values derived on the
basis of two-dimensional incompressible flow are corrected for three-dimensional effects and
then applied as ordinary added masses in the hull vibration calculations as will be discussed
in the next chapter. It must be pointed out, hare, however, that, because of the complex form
of hull,, when motion of an underwater form takes place in a given direction, inertia effects
are developed not only in that direction but also in other directions. This phenomenon is
spoken of as an inertia coupling between the various degrees of freedom. As shown in Ref-
erence 2-8, in the most general case for a submerged rigid body having six degrees of freedom,
there are 21 such inertia terms. Fortunately, in practice most of these either vanish because
of symmetry or can be neglected.
It may seem surprising that the added mass effects for vibratory motion do not vary
widely with the ship's forward velocity, but the potential flow theory indicates no variation;

2-7
this has been borne out by experimental observations. For further discussion of this point,
see pages 59 and 60 of Reference 2-7.

E. LOCAL EFFECTS
In view of the complexity of a ship's structure, the number of "local" hull structures
is enormous. However, these structures may be divided into categories of different relative
importance. The most important distinction to be made as far as hull vibration is concerned
is between those structures that have an appreciable effect on the vibration characteristics
of the ship as a whole and tluboe that do not. The possibility of affecting the ship as a whole
obviously depends primarily on the mass of the local structure, but it also depends on its
location and its stiffness.
When a mass and spring are attached to a free-free beam, the "sprung mass" partic-
ipates in the normal mode vibrations of the combined system. It can introduce an extra mode
so that. as far as the beam itself is concerned, there may then exist two flexural modes with
the same number of nodes. Its effect on the previously existing modes depends on both its
mass and the proximity of the beam frequencies to the natural frequency of the mass-spring
combination, that is, to the natural frequency of the mass when the end of the spring is held
fixed.
When local floxibilities of ship structures produce a sprung mass effect the normal
modes of the hull tend to depart from beamlike form, and modes of vibration of the ship may
be found in which the local vibration is excessive, whereas at the ends of the ship the vi-
bration is well within tolerable limits.
When local structures are of relatively small mass in comparison to the mass of the
ship, their effect on the vibratory characteristics of the ship as a whole will be negligible.
However, because of resonance, they may themselves vibrate excessively. If their natural
frequencies coincide with the frequency of some source of excitation prevailing at the oper-
ating speed of the ship, these structures may respond to an imperceptible hull vibration so as
to produce an intolerable local condition. Obviously, the cure for such a condition is to change
the natural frequency of the local structure.

F. SHALLOW WATER EFFECTS

The vibration characteristics of ships are materially modified in passing from deep to
shallow water. In the first place, there is a marked increase in the added mass effect for
vertical vibration, and in the second place, the propeller exciting forces may be greatly
changed.
The change in added mass effect is due to the alteration of the noncirculatory flow
p 'Lern. A rough rule for the limit of depth at which this effect is no longer evident is six

2-8
times the draft. When the depth is less than this a lowering of the frequencies of the vertical
modes of the hull is observed.
The variation in propeller excitation in restricted waters arises from a modification of
the steady flow due to the restriction of the channel around the ship. As will be pointed out
in Chapter 7, any effect that disturbs the uniformity of flow into the propeller races will sot
up lateral forces which are transmitted to the hull through the propeller shaft bearings.
In addition to these two effects there will usually be a reduction in operating speed on
entering shallow water. Thus the vibratory level may vary because of any one of these sep-
arate effects.
Although it is not inconceivable under the circumstances that a particular ship might
experience less vibration while operating in shallow water, the chances are that it will ex-
perience more. Specific examples of increased vibration in shallow water are cited in Ref-
erences 2-9 and 2-5. If a particular hull happens to be subject to resonant vibration when
operating at its designed speed in deep water, then it is quite possible that resonance will
be avoided, at the speed assigned to shallow water operation.

REFERENCES
2-1. McGoldrick, R.T., "Buoyancy Effect on Natural Frequency of Vertical Modes of Hull
Vibration," JSR (Jul 1957).
2-2. Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York (1956).
2-3. Timoshenko, S., "Vibration Problems in Engineering," D. Van Nostrand Co., New
York (1955).
2-4. Rayleigh, Lord, "Theory of Sound," Second Edition, Dover Publications, New
York (1945).
2-5. Saunders, Il.E., "Hydrodynamics in Ship Design," published by SNAME (1957).
2-6. Lewis, F.M., "The Inertia of Water Surrounding a Vibrating Ship," Trans. SNAME,
Vol. 37 (1929).
2-7, Wendel, K., "Hydrodynamic Masses and Hydrodynamic Moments of Inertia," STG,
Voi. 44 (1950). Also TMB Translation 260 (Jul 1956).
2-8. Lamb, H., "Hydrodynamics," Dover Publications, New York (1945).
2-9. Noonan, E.F., et al., "Vibration Measurements on 6hip S6-2 during Builders Trials,"
prepared for Hull Structure Committee of SNAME by Noonan, Knopfle, and Feldman, Professional
Engineers (Nov 195 9 ).

2-9
 CHAPTER 3
BASIC BEAM THEORY OF SHIP VIBRATION

A. INTRODUCTION

The fundamental system considered in all texts, 3-1 3-2 on mechanical vibration is the
lumped mass-spring system of one degree of freedom shown schematically in Figure 3-1.

P0 sin cut -- b M

Figure 8-1 - Rectilinear Vibratory System k

of One Degree of Freedom

This system has mass m, spring constant k, viscous damping constant o, and in this case is
acted upon by a simple harmonic driving force P0 sin wt in the x-direction. The mass m is so
restrained that it can move only in the c-direction.
The differential equation govwrning this case; namely,

m, + ci + ko = P 0 sinet [w-1]

[where the dot denotes differentiation with respect to t (time)] is the most widely discussed
equation in mechanical vibration theory. Its steady-state solution yields the familiar res-
onance curves of forced vibration. These indicate that very large amplitudes of vibration of the
mass m will result when w0is close to the natural circular frequency of the system
,- , and the damping constant c is "small."

Also quite important in vibration theory is the solution of Equation [3-11 when the
driving force is absent (P0 = 0). This yields an exponentially decaying free vibration at a
frequency which approaches the undamped natural frequency as c-d0.
Just as the lumped system of one degree of freedom provides the basis for the under-
standing of the vibratory characteristics of many familiar mechanical systems (for example,
the pickup units; of many vibratio- instrunientd), Li the uniform free-free beam provides a
basis for an understanding of the essential vibratory characteristics of ships.
The free-free uniform bar or beam is, of course, a continuous system (as contrasted
with the lumped system of Figure 3-1), and, although it also has the properties of inertia

3-1
and elasticity possessed by the system of Figure 3-1, the differential equation governing its
vibratory motion is considerably more complicated.
The beam of Figure 3-2 is assumed to have a mass per unit length p and a bending
stiffness El in the xy-plane. In the terminology of the Euler-Bernoulli beam theory this means
that, if the beam is so slender that it can be bent into circular form in the XY-plane, with the
two ends joined together, certain simple relations exist. Thus the bending moment M, due to
the normal internal stresses acting at any cross section, will be related to the mean radius of
curvature p by the equation
El
M - [3-21
P

TP (a,) Figure 3-2 - Slender Beam Free in Space

Subjected to a Lateral Forcing
Function P(x, t)
0

When small deflections y of the beam of Figure 3-2 take place in the Y-direction, the approx-
imation that the curvature (reciprocal of the radius of curvature) is equal to the second de-
rivative of y with respect to -- can be used. The familiar equation relating bending moment
and deflection in simple beam theory is then

g = El--2Y[3-3]
aX2

From this relation it can be shown that, in contrast with the ordinary second-order
differential equation governing the system of Figure 3-1, the equation governing the system
of Figure 3-2 is a partial differential equation which is of the fourth order with respect to X
and the second order with respect to t (time). This equation is

El- + -i2 =P(x, t) j3-4]

004 9t2

where P (z, t) is the driving force per unit length in the Y-direction. This equation is widely
discussed in the literature, 31, 3-2 and in its homogenous form [P (x, 1) 0)Oleads to the well-known
formulas for the natural frequencies of uniform slender beams with various end conditions.
It was natural that, since the ship when advancing through waves is loaded in bending
and hence is essentially a beam, the early attempts to develop formulas for its natural fre-
quencies should be based on the formula for the natural frequency of the froo-free uniform
beam.
3-2
 The best known of such formulas is the Schlick formula for the fundamental vertical
frequency of a surface ship:

N
N
C 18-5]

This formula is given in mixed units for convenience in practical application; thus

N is the fundamental vertical frequency in cpm,

C is Schlick's empirical "constant,"
2 2
I is the area moment of inertia of the midship section in ft -in. units,
A is the displacement in long tons, and
L is the length in ft.
For ranges of values of C and further discussion of this formula, see Appendix C. Other
3 4 3
well-known formulas such as those of Burrill,3" Todd and Marwood,3- or Prohaska "s are
also discussed in that appendix.
Here the empirical formulas are contrived to account for the many ways in which the ship
departs from the free-free uniform beam. Aside from its nonuniformity, one of the chief re-
spects in which a ship departs from a slender beam in its vibratory characteristics is in the
relativoiy much greater shearing flexibility of the ship. This is because the ship is not, in
fact, as slender as the beams for which the Euler-Bernoulli assumptions are valid.
The modification of the Euler-Bemoulli uniform beam to allow for shearing flexibility
yields what is now generally referred to as the "Timoshenko beam." This is still a uniform
solid beam, but when it is deformed, the slope of its elastic line is considered to have one
component due to bending and another due to shearing. In the actual equation discussed by
Timoshenko,3" 2 there was included not only a term for shearing rigidity but also a term for
rotary inertia, neither of which appear in the equation for the Euler-Bernoulli vibrating beam.
The rotary inertia represents the increased inertia effect because the mass of the ship is not
concentrated along its longitudinal axis.
The homogeneous form of Timoshenko's equation in the notation adopted for this book
is:

El S4 )#y +F-
3ay IIz
-~
a4y pE! a4y =0 [3-6]
air2a& 2 at 2 KAG I'4 KAG d__2at 2

This equation has been discussed in many publications in addition to Reference 8-2.
Even when the term for rotary inertia (I z) is omitted, it has not been found possible to
derive from it a direct formula for the natural frequency of the free-free beam. However,
curves can be plotted showing how the frequencies of various modes vary with the ratios
pL4 El
-- and , as shown in Reference 3-6. The first of these ratios appears in the
Ei 2KAGL 2

3-3
formula for the slender uniform beam in which shearing flexibility is neglected. This formula
is
El
U-22.4 -[84]1

ILL4

The second ratio involves the relative magnitudes of the bending and shearing rigidities.
Many questions may be raised as to the interpretation of Equation [3-6] and of the
wave solutions to which it gives rise. Since the ship is not a uniform beam, and analytical
expressions cannot be given for the parameters El, A, lI., and KAG as functions of 0 huh
reader is referred to the literature for further discussion of this equation. (See the bibli-
ography at the end of the book.)
Before considering the equations that provide the basis for the rational theory of ship
vibration proposed in this book, it is necessary to discuss briefly the torsional vibrations of
the free-free uniform beam. As in the case of the flexural vibrations, in which the Euler-
Bernoulli assumptions provided an integrable equation, a simplified theory of torsional vibra-
tion of beams or hulls is based on the torsional equations for the uniform (cylindrizal) shaft.
Figure 3-3 shows a solid cylindrical shaft with axis coinciding with the X-axis. It is

IY

Figure 3-3 - Cylindrical Shaft with Axis

T, Coinciding with OX and Twisting
o X about OX

shown in texts on "strength of mat.erials"37 that the torsional rigidity of such a shaft is GJ,
where G is the shear modulus of elasticity and J is the polar moment of inertia of the area of
the cross section with respect to the X-axis. This means that, if one end is held fixed and a
torque T is applied to the other end, the resulting twist at the point of application of the
torque is given by the equation
9 = L [8-8]
GJ

where L is the length.

It can be shown that the torsional oscillations of such a shaft are governed by a par-
tial differential equation of the second order with respect to both x and time. This equation
is

GJeo - 7__ [3-91

ax 2 0e 2

where I is the mass polar moment of inertia of the shaft per unit length.

3-4
 This equation merits special attention because it has the form of the one-dimensional
wave equation
I~ 02U
a8- [1-0]
2 2
d2 a

In this well-known equation u may represent any displacement-like quantity, and c, the velocity
of wave propagation. Thus, if in Figure 8-2, u is the displacement in the X-direction and c
is the velocity of propagation of longitudinal waves along the beam, the longitudinal vibrations
of the beam are-governed by Equation [3-10]. In the torsional case of Figure 3-8 the velocity
of propagation is

As shown in texts on acoustics, 3 8 when the governing differential equation for the
vibration of the mechanical system can be expressed in the form of Equation [3-101, the
formulas for the natural frequencies can be readily derived from the relations among frequency,
wave velocity, and wavelength

c = n[-11]

where n is the frequency and A is the wavelength. The normal mode patterns in such cases
are sinusoidal, and the boundary conditions determine what fraction of a wavelength-is in-
cluded in the distance between the boundaries of the system.
For a cylindrical shaft free in space, the frequency of the fundamental torsional mode
which has one node at midlength is given by the formula:

ni= [3- 121

The mode shape shown in Figure 3-4 is a half cosine, it will be noted that in this case the

Figure 3-4 - Normal Mode Shape for Funda-

mental Torsional Mode of a Uniform Cylin-
drical Shaft with Free Ends 0 L

length of the shaft comprises a half wavelength (half of a ful! cosine) and that Equation
[3-12] conforins WoEquation [3-11] if

3-5
 Just as the empirical formula of Schlick for the fundamental vertical flexural mode of
a bull was based on the corresponding formula for the free-free uniform bar, so an empirical
formula for the frequency of the fundamental mode of torsional vibration of the hull resembling
the corresponding formula for the free-free cylindrical shaft was proposed by Horn.
Horn's formula is:

Ne 60k [3-18]
A(B 2 + D2 )L

This formula is given here in mixed units for convenience in practical application; thus:
Ne is the natural frequency in cpm,
g is the acceleration of gravity in ft/sec2 ,
Je0 is the effective polar moment of inertia of midship section area,
A is the displacement in long tons,
B is thebeam in ft,
D is thedepth in ft,
L is thelength in ft,
0 is theshear modulus of elasticity in tons/ft 2 , and
"k is Horn's empirical coefficient.
For discission of the evaluation of k and J. 0e see Appendix C.
Not only does the hull depart from the ideal free-free beam in the nonuniformity of mass
and stiffness distributions, but it also lacks symmetry, and this property complicates its vi-
bration characteristics to an extent that has so far been but little explored. Because of the
close approximation to symmetry of most ships with respect to a vertical plane passing
through the longitudinal centerline, the prototype beam for the equations to be derived here
is assumed to have one plane of symmetry. Here the term symmetry means mirror symmetry.
It is shown in Reference 3-10 that mirror symmetry is a sufficient condition for vibrational
symmetry, but not a necessary one.

B. BASIC DIFFERENTIAL EQUATIONS FOR THE SHIP

The "rational" theory of ship vibration proposed in this book is based on the assump-
tion that the hull may be considered as a free-free beam with three principal types of flexi-
bility; namely, bending, shearing, and torsional (or twisting). In this theory the inertia effect
of the water is treated as equivalent to mass added to the mass of the hull at suitable loca-
tions. The elastic axes for bending are assumed to be vertical and horizontal.

3-6
The longitudinal (X) axis is taken as passing through a point halfway between the main deck
and the keel at the midship section; this is considered to be parallel to the keel whether or
not the ship is actually at zero trim. The orientation of the Y- and Z-axes are dependent on
the modes of vibration considered.
This analysis provides for a coupling of horizontal vibration with torsional vibration,
but, because of the symmetry, no coupling of vertical vibration with either torsional or hori-
zontal vibration is considered here.
The basic equations are derived by treating the forces and moments acting on the ele-
ment shown in Figure 3-5 when it is taken as uniform with all parameters equal to the mean
value over the length Ax.

+ V/V + a1W

T +Al

+'tf M + AM

Figure 3-5 .- Element of Ship of Length Ax Subject to Forces and Moments Imposed by
Adjoining Elements while Executing Torsion - Horizontal Bending Vibration

The equations are derived by considering that the element of the ship of length AX
possesses a combination of the properties of the Timoshenko beam in flexure and the cylin-
drical shalt in torsion, but with account taken of the lack of both inertial and elastic sym-
metry. Thus, a term F is introduced to represent the vertical distance from the X-axis to the
center of mass, and a term 2 to represent the distance from the s-axis to the center of hori-
zontal shear (which is here assumed to be the centroid of the area of the cross section of the
hull),
The axes chosen form a right-hand set, and the sign convention is as indicated by the
plus signs at the left end of the element in Figure 3-5. Thuu the horizontal shear force V is
positive when the part of the hull to the left of the element exerts a force on the element in
the direction of positive Y.

3-7
 Application of Newton's law to motion in the Y-direction gives

fiAX a2Y-- AV [8-14]

where y' is the displacement in the Y-direction of the center of mass of the element. Since
V'- y - F , for small motions
----- 8- a [3-15]
a 2 aY72Y[

flence
a2 a
- AV-ptA
fAn-2 2 - aX [3-168
at 2 at 2

Newton's equation for moments about a Z-axis gives

AM = VAxn+ " AX .Y.: [3-17]

2
at

where M is the moment with respect to a Z-axis exerted on the element by the part of the
hull to its left. From the Timoshenko beam properties

Ay = MA [8-18]

where y is the rotation of the cross section with respect to a Z-axis.

VAX
Ay"= yAX - KAG [.3-l19

where y" is the displacement in the Y-direction of the center of shear and KAG is the shear
rigidity. The last equation also implies that the displacement of the center of shear depends
only on shearing and bonding action and not on torsion, that is, that the hull twists about the
center of shear.
The simple concept of torsional rigidity defines the quantity GJ,. by the torque required
to produce a given rate of twist with respect to the longitudinal axis. The complete set of
shear stresses in the cross section has a moment about the X-axis designated in Figure 3-5
as T. In the absence of a net vertical shear force, the torque with respect to the longitudinal
axis is obtained by subtracting from T the moment due to the net horizontal shear force.
This is - V F. Hence, from the definition of torsional rigidity GJ., and the sign conventions
indicated in Figure 3-5

8-8
 G- [8-20]

whence

(T + V)AX (3-21

Since y"- y - f.0, and the element of length Am is treated as uniform in the approximation
to the continuously varying hull, Ay*"= Ay - I AqS. So
2
AAY V TAY VV AX (8-22]
KAG 04 4J,
Since I x is specified with respect to the arbitrarily chosen X-axis, the equation for the time
rate of change of moment of momentum about the X-axis gives
-AT fi- pAei" .o2-y"~ 3.
2 2
at _t

where lax is the mass moment of inertia per unit length with respect to the X-axis, including
the allowance for the inertia effect of the water.
The first term in this equation is the rate of change with respect to the X-axis of the
moment of momentum associated with the center of mass; the second term is the rate of change
of moment of momentum about a parallel axis through the center of mass. Therefore

AT = pA070 a2Y -I Ax d2
a[8-24]
-k
0t 2 at2

If the foregoing equations are set in differential form with respect to x, the following
set of partial differential equations is obtained:
av 02y 6-5
-=-' - +us - [8-251
2
do~ at32 at

a v +I 2- [8-261
y at2
a, --
Oz - El [8-271

_95 (T + V 1) [8-28]
ax
e

8-9
 Sy-VY + - -
s ~ KAG GJ) 0J.

am OP Lx t2

In these equations a term involving A I is omitted since the element is treated as uni-
form over the length As. Also the X-axis is treated as a principal axis of inertia of the element.
The boundary conditions for free vibrations are

S= M -T -0 for
{aLJ
C. METHODS OF CALCULATING NATURAL FREQUENCIES AND
NORMAL MODES OF VIBRATION OF SHIPS
The set of partial differential equations given in the foregoing section cannot, in
general, be treated analytically for the nonuniform beam representing the ship since the coef-
ficients which are functions of x are not given in mathematical form. They are available only
in the form of graphs or tables. However, these equations furnish the basis for approximate
methods of calculation of considerable practical importance. These methods are classified
here as the digital, the analog, and the graphical methods.
In all three methods the problem is greatly simplified if only vertical hull vibration is
of concern, for here, unless the ship is of unusual design, the symmetry eliminates the cou-
pling effects. In this book, only the vertical case is used for illustration of the general
methods of solution. For the vertical modes, the Y- and Z-axes are then rotated 90 degrees
clockwise so that positive Y is upward and Z points outward from the plane of the paper.
The axes still form a right-hand set and the flexural vibration is still in the direction of Y'
so that the same equations are valid. Here, however, 7 and 1 are made zero and the equations
involving y are independent of q5. The reader is referred to Reference 3-11 for the treatment
of the horizontal and torsion-bending cases.

1. DIGITAL METHOD
The digital process that has shown the most promise so far in computing natural fre-
quencies and normal modes is referred to in David Taylor Model Basin reports as the Prohi-
Myklestad method. This method utilizes the equations in their finite difference form and is
closely related to the Holzer method widely usnd in torsional vibration analysis.
The process of treating the hull vibration problem by finite differences involves con-
sideration of the validity of various approximations. Naturally the aim is to obtain the great-
est possible accuracy with the minimum amount of computation. Some prefer to regard the
process as one of first converting the ship into an equivalent "lumped" system in which the

8-10
inertias are concentrated at certain points and the elastic properties are assigned to massless
elastic members joining these lumped elements. Others insist that no such concept is nec-
essary and that the representation of the continuous system by finite difference equations is
entirely mathematical. From either point of view it seems obvious that the greater the number
of lumps (or equations) used, the greater the attainable accuracy but sometimes, in the case
of the computing machines, the accumulation of roundoff errors gives less accuracy with a
large number of lumps.
Accuracy of computation, however, does not necessarily imply absolute accuracy. The
latter is limited not only by the uncertainty in the evaluation of all parameters appearing in
the equations, but also by the factors which cause a ship to depart from beamlike behavior.
Here are included such properties as local flexibility, structural discontinuities, concentrated
masses, hatch openings, and large superstructures.
The minimum number of sections for obtaining the accuracy warranted by the reliability
of the input data appears to be 20. The point of diminishing returns appears to be reached at
!0 secteions. The basic process is the same regardless of the number of sections used.
The case of vertical flexural vibration used here as an illustration requires only four
of the six equations given on pages 3-9 and 3-10. The parameters and variables dropping out are
*5, i, GJe, T, and h,
A lumped system having approximately the vibratory characteristics of the ship is
il1ustr'.ed in Figure 3-6. This is based on a 20-section breakdown which requires 21 mass
elements with one at each end.

020
Figure 3-6 - Illustration of Breakdown of a Ship for Digitai Calculation
of Vertical Modes of Vibration

3-11
 The mass elements are points of concentration of the translational inertia uAX and the
rotational inertia /I, Ax. The elastic properties of the hull are concentrated in the massless
connecting members. In this case there are the bending rigidity E1 and the shearing rigidity
KAG. Hence, two principal modifications are made of the equations given on pages 8-9 and 8-10
for this particular calculation. First, they are converted from differential form to algebraic form
with respect to time by using the fact that the desired solutions will be simple harmonic vi-
brations which can be expressed in the form

y(z, )t) Y(z) sin wt [3-81)

which gives

(z, t) ffi- ji (z) sin &, t -is y (m, t) [3-82]

where Y,(m) is the normal mode pattern for the ith normal mode. Second, they are expressed.
in finite difference form with respect to x.
The equations are then converted from differential form to finite difference form with
respect to x. As given here, they are in their simplest form.

2
AV = ,u Azw/) [8-88]

2 [y-34]
AM = VA_- -/I Acyo

A = !Axaas
El

VAx
=yAx - -- [8-361
Ay
KAG

As shown in Reference 3-11, a staggering system is usually used for increased accuracy of
calculation. This requires introducing half stations and considering forces and displace-
ments at staggered intervals with the boundary conditions V = M = 0 at each end. In this
simplified form the subscript notation required to identify a value at any particular station
along the length of the hull is unnecessary, as is also the use of half-station staggering of
certain terms for increased accuracy of computation. Each of the four equations indicates
the difference between two valups of the variable on the left side of the equation in advanc-
ing a distance Ax to the right when the hull is executing a simple harmonic vibration of
circular frequency w.
The computation is started by assigning an arbitrary value to o and unit value to y at
the left-end station. It is known that V0 m0 and M. - 0 but the value of y. must remain
undetermined temporarily. The set of four difference equations will then permit advancing
from Station 0 to Station 1 with V, 3M,y, and y all known in terms of " This proess can

8-12
 .~I

then be continued cyclicly until the right end of the hull t is known that if
the frequency w assumed was a natural frequency, it will be possible to find a value of yo
such that the boundary conditions, that V and A are both zero at this end, will be satisfied.
Once y0 is ditermined numerical values are then fixed for V, M, y, and y throughout the ship.
There are a number of ways in which this basic process can be carried out. Thus, use
can be made of the fact that the desired solution can be formed from a linear combination of
calculations made by starting at the left end, first with the values y = 1 and y - 0 and then
with the values y - 0, y - 1. A range of values of a) is then explored for which it is possible
to find a linear combination of these two calculations, which will make V and M both zero at
the right end of the hull.
Since the lumped system used here to represent the hull has only 21 degrees of freedom,
it can have only 21 natural frequencies and normal modes, whereas a continuous free-free beam
has a theoretically infinite number of modes. In pracdice, such a calculation would not be
carried beyond about the sixth or eighth mode, the number depending on the length to depth
ratio of the hull. If the method were valid beyond this number of modes it would be necessary
to use a larger number of elements in the breakdown in order to realize its potentialities.
Several points should be noted concerning this method of computation. The natural
frequencies are found by a process of searching for values of W at which the boundary con-
ditions can bp satisfied. At. each frequency thus found the calculation automatically gives a
normal mode pattern. There is no significance to the absolute amplitudes used in such a cal-
culation, but only to their relative values. Thus, the calculation shows what values V, M, y,
and y would have throughout. the hull if it were possible for it to vibrate with unit amplitude at
the left end at the particular frequency in question. In the foot-ton-second system of units
this would mean an amplitude of 1 ft at station zero, clearly an excessive value. However,
since the equations used are linear the result can be used to obtain the values associated
with amplitudes of practical magnitude, that is, fractions of an inch.
The Model Basin has for the present (1960) adopted the station designations as 0 at the
stern and 20 at the bow for the vibration calculations. This is the opposite to that ,1sed in the
naval architects' lines drawings but preserves the convention of having the bow to the right
in the ship's profile.
In spite of the increased number of variables and pm.ameters, the basic process of cal-
culating torsion bending modes by digital calculation is the same as for the simpler case just
illustrated. When a normal mode of this type is found by digital calcglaUion Lhero will be
obtained sets of values of 1/, y, 0, V, U, and T for each natural frequency on the basis of
unit amplitude in y at station zero. When y and 0 are plotted there may be found pairs of
torsion-horizontal bending modes in which there are the same number of torsional and flexural
nodes in each mode. However, there is reversal of phase between rotation and translation
as illustrated in Figure 2-5.

8-18
2 ANALOG METHOD

The subject of electrical analogs of mechanical systems is very broad and is of great
importance in vibration analysis in general, as is well known. The equations applicable to
electrical networks in which alternating currents are flowing are, ui general, of the same
form as the equations applicable to vibrating mechanical systems. This at once suggests
the use of electrical circuits as computing devices for studying the vibratory characteristics
of mechanical systems.
An important distinction is to be made between those electronic computing devices that
perform purely mathematical operations such as integration and differentiation, and those in
which there is a truly physical analogy between the currents or voltagev in various
branches of a circuit and the amplitudes of vibration or forces existing in the corresponding
mechanical system. The former are Frequently called operational analogs and the latter are
spoken of here as network analyzers. This chapter is concerned only with the latter.
Direct electrical analogs of vibrating mechanical systems are not unique and two such
analogs have been used at the David Taylor Model Basin in making vibration analyses on the
same network analyzer. These are known as the conventional or "classical" analog and the
"mobility" analog. The analogous quantities in these two systems are listed in Table 3-1.

TABLE 3-1
Analogous Quantities in Two Types of Electrical Analog of a
Vibrating Mechanical System

Conventional Mobility
Mechanical Values Analog Values Analog Values

Mass Inductance Capacitance

Reciprocal of Reciprocal of
Spring constant capacitance inductance
Viscous damping Reciprocal of
constant Resistance resistance
Force Voltage Current
t Integral -of
Displacement Charge voltage
Velocity Current Voltage
Frequency Frequency Frequency

In general, mechanical units and electrical units are of such magnitudes in practice
that neither of these analogs can be used without a scaling transformation. The electrical
circuit is driven at a much higher frequency than the mechanical frequencies met in practice.
This reduces the size of inductances and capacitances required in either analog.

a-14
 In dealing with the flexural vibration of hulls, the Model Basin has used a mobility
analog in the form of a pair of coupled transmission lines; in one the current represents the
shearing force in the hull, in the other the current represents the bending moment. This cir-
cuit is discussed in considerable detail in Reference 3-12. When a separate ground line is
drawn parallel to each of these ungrounded lines, one section of this network, which then
represents one of the sections into which the hull is broken down in the analysis, will have
eight terminals, four at each end. Such a section is shown in Figure 3-7. This circuit uom-
prises a "passive network," that is, one within which there are no sources of energy so that
the network by itself is dissipative because of internal losses. When external voltages are
removed, only transient oscillations remain which eventually die out just as the vibrations
of the mechanical system die out after the exciting force is removed.

Am Am
______KA G
V 00-00 v.+

izA x

As

El

00

Figure 3-7 - One Section of an Electrical Circuit Analogous

to a Section of a Vibraui'ug Ship

In Figure 3-7 currents flowing to the right in the upper line represent vertical shearing
forces in the hull, and currents flowing to the left in the lower ungrounded line represent bend-
ing moments in the hull. Voltages in the shear line are analogous to mechanical translational
velocities and voltages in the moment line are analogous to mechanical angular velocities.
It will be noted that inductances are proportional to the reciprocal of mechanical shearing or
bending rigidities whereas capacitances are analogous to mass or mass moment of inertia.
One winding of the transformer is in series with the shear line; the other winding is connected
between the moment line and ground.

3-15
 To find the natural frequencies of a hull experimentally, the usual procedure is to in-
stall a vibration generator that is operated over a wide range of speeds and to observe those
speeds at which resonance peaks of amplitude are found. Similarly, with the analog the nat-
ural frequencies of the circuit are found not by observing free oscillations but by applying
continuous excitation by means of an oscillator. In the case of the mobility analog the con-
dition corresponding to resonance in the mechanical system is actually an electrical anti-
resonance. This is advantageous in that little power is taken from the oscillator. A high
resistance is inserted between the ungrounded terminal of the oscillator and one end of the
shear line in the circuit. If this resistance is high enough, the current fed into the shear line
will remain practically constant regardless of the frequency since the change in impedance of
the circuit will have little effect on the total impedance. At each antiresonance the voltage
will rise at the ends of the shear line, and the normal mode pattern of the hull may then be
found by plotting the voltage measured along the shear line at the 20 stations.
To show that this electrical system is analogous to the vibrating hull it is required
merely to show that the application of Kirchhoff's laws for electrical circuits when applied to
the section shown in Figure 3-7 will yield the same set of difference equations as Equations
[3-33] through [3-36]. Thus, the difference in currents at the two ends of the shear line of
the section must equal the current flowing from thA shear line to ground. The latter current,
however, is the product of the voltage and the admittance jWo/Aw, where j = v In a-c cir-
cuit theory, differentiation with respect to time is effected by multiplying by the operator jeo,
and integration with respect to time by dividing by jf. Hence the current to ground is

Hence
V,3 - - pA X2: + Vn + I

or
AV = uAxyco 2 [3-38]

In the moment line Kirchhoff's current equation involves four currents: MdI M,,+ V,VAm
(through the transformer winding which is coupled to the shear line V), and * (jW1 JAX) through
the capacitance. When the directions of these currents are taken into account there results
the equation
AM = VA_-- I zAzy62 [8-34]

The remaining two equations follow from Kirchhoff's potential relations. Thus the voltage drop
in the inductance A is equal to the inductanue times the -..e of cOAnge of current, i.e.,
Ax El
-M.
El

3-16
Hence
Ax

or, on integrating with respect to time,

MAX
Ay - T&.85]

Finally, since the voltage drop in the winding of the transformer in the shear line is
Am times the voltage across the other winding (j) and since the voltage drop in the induct-
Az Az
ance - is equal to - V, the potential equation is
KAG KAG

On integrating this gives

VAz
Ay = yAm- -[8-86]
/tAG

It is seen that the boundary conditions of the free-free hull are satisfied in the analog
if the shear and moment lines are isolated at the ends since this makes V = M = 0. Of course,
when an oscillator is connected to one end, it gives an input current which is analogous to
applying an external simple harmonic driving force to one erld of the hull. The voltages de-
veloped at the ends correspond to the rectilinear and angular velocities at the ends of the hull.
A photograph of the TMB Network Analyzer used for vibration analysis by means of
the electrical analog is given in Figure 3-8. A complete description of this network analyzer
is given in Reference 3-13.

3. GRAPHICAL INTEGRATION - STODOLA METI;OD

Some success has been attained in computing the lower modes of hull vibration by
graphical integration. This is of practical importance since analog or digital computing fa-
cilities are not universally available. In this chapter the application of graphical integration
to the Stodola method is discussed, but it must be emphasized that numerical integration can
also be applied to this method. In applying this method the added mass must have been pre-
viously computed and added to the mass of the hull as a distributed mass. Here the hull is
considered as a beam froe in space and the added mass is assumed to remain constant in time.
The case of vertical vibration is again used to illustrate the method, since in this case
coupling with torsion is negligible. As a further simplification the effect of rotary inertia is
neglected here. The set of differential equations to be integrated is then:

8-17
 -474

Figure -8
8 Network Analyzer Used in Making Hull Vibration Calculations
by Means of Electrical Analogs

av [3-381

dy 141[3-38]

dy - [3-401
S7
Y KA G

With boundary conditions

V M 0 for
A

3-18
In these equations 1, El, and KAG all vary with z.
The curve of 1 versus x is first plotted from the weight curve of the ship and the esti-
mated added mass of water. This may be either a smooth or a stepped curve.

0 L

Figure 3-9 - Sample Plot of p versus x Required for Use of Stodola Method

A normal mode shape is then assumed for the 2-node vertical mode. The nearer this is
to the true mode shape the less labor is involved in the calculation, but, in the absence of a
basis for a more realistic curve, thc normal mode pattern Y. for the free-free uniform bar is
used as a starting curve. The value of Ya may be taken a6 unity at xs= 0. A set of values
for plotting the uniform bar curve is given under Case lA of Reference 3-14. The I values
are then multiplied by the corresponding values of Y. to give the curve 1Ya shown in Figure
3-11.

0L

Figure 3-10 - Sample Plot of 2-Node Pattern of Free-Free Uniform Bar

for Use in Stodola Method

If Y'. were the true normal mode pattern for the ship in question it would be found that

3-19
 L

Figure 3-11 -- Plot of tty. versus x for Use in Stodola Method

Lx
f igYaxdad o= 13-421

since the first integral is proportional to tho shearing force V (o) in the hull during vibration
(actually V (x) - fjLYn2 dx), and the second is proportional to the bending moment M (o). If
the first integral is not zero it may be made zero by a parallel shift of the X-axis in the plot
of YI0 (x) to obtain a second approximation Ya'(). If, after this modification, the second in-
tegral does not vanish, it may be made to vanish by a rotation of the X-axis about the point
whose abscissa is that of the centroid of the area under the A curve. The curve of ampli-
tudes, obtained after shifting and rotating the base of the Ya-curve and replotting on the
basis of unit value at Station 0, is labeled Ya" (,c). A mechanical integraph will draw the in-
tegral curve when the curve to be integrated is traced by the stylus.
Although the modifications that have been made so far in the starting mode shape of
the free-free uniform bar will aow make it satisfy the boundary condition on V and M, as yet
no account has been taken of the elastic properties of the hull.
The four differential equations on page 3-18 yield the following set of integral equations:

4 yw2 do . y [3-481

2
ffoy) do d -- N [3-44]

rdf
"TJ ffy'2 dd d y3

f ofdo 2O)
d do d- u2do=
O [3-46]
El ffpJKA GJf

3-20

-1
 II

After each integration the constant of integration must be found beforer proceeding to the
next. In the case of the first two integrations the constants are furnished by the boundary
conditions, and the value of y to be used in Equation [8-43] has already been adjusted so that
the first two constants of integration will be zero. Thus V and Mare zero for z - 0 and
x - L.
The last equation of the above set may be written

[d f8 dsx_
x _

Jyd do-fA.lJ yd 02 i-7

_E
IIJ

where it may be seen that, if all the constants of integration were zero, and, if the starting
mode shape were the true mode shape, the curve finally obtained by carrying out these inte-
grations would differ from the starting curve only by the scale factor --.2 Hence the nat-
ural frequency could be found directly from this scale factor.
As long as KAG and El remain finite, all integrations may be carried out starting with
zero values at x - 0. When the final integrations on the left side of Equation [8-47] are car-
ried out, the two curves plotted with zero value at x = 0 may be combined to give a curve for
F
2

0it

:L

Figure 3-12 - Combining of Components of Bending and Shearing Deflection

First the ends of each curve are joined by straight lines. Then the ordinates measured ver-
tically from the curve to the inclined line are combined. The frequency is found by comparing
the magnitude of this final curve with the similar curve obtained by joining the ends of the

8-21
curve Y " (o) used as a basis for the py curve. This curve was previously obtained from the
curve Y: (o) by a parallel shift and rotation of the X-axis. To find the normal mode pattern,
y curve. As in deriving the
however, it is still necessary to find the true base of the a)2
curve Y"' (o), this is again accomplished by a parallel shift of the X-axis and a rotation about
a point whose abscissa is that of the centroid of the area under the i curve. The true baue
must again be such as to make
L X

1 pda' and My
d oda

both equal to zero.

If the curve Ya"(x) is the true normal mode curve, the process will yield a - curve
02
differing from Ya" only in scale. It can be proven3"1 that the Stodola process is convergent
to the lowest mode; that 1s, that, if the whole procedure is repeated using the finally obtained
y curve as a second starting curve, the second calculation will come closer to giving the same
shape than previously. It can be shown also that no matter what shape is initially aS.u..od,
unless it coincides exactly with the shape of a higher mode, the process will eventually con-
verge to the 2-node mode shape.
If the Stodola method as previously outlined is applied to the calculation of the second
mode, the process will in general convergo to the first mode. To make it converge to the sec-
ond mnodc it is necessary to make use of the orthogonality relations applicable to normal
modes. 3 "1 As a result of this property, if Yi and Y. are two of the normal mode shapes,
then

flu Yj Y. do = 0
if i~j.
An arbitrarily assumed mode pattern can, in general, be resolved into a series of nor-
mal functions. This 5,

Y 0-) aj Yj (Z) [3-481

For the Stodola process to converge to the second mode the first mode component in the as-
sumed function Y (x) must be eliminated.
It has been shown that for a true normal mode the Stodola process gives a result that
Y (a')
differs from the initial pattern only in scale; that is, it gives for Y, (a) the value -,

S~2
Once the first mode shape is known, the curve assumed for the second mode can be corrected
fnr any first-mode component present by subtracting the function

3-22
 fL / Y Y1 do
0

It is seen that this quantity is equal to a1 Y1 (m)of the summation in Equation [8-48]
since
L L
f L uY 1d -f 0L A (a Y 1 + a2 Y2 + etc) Y, do [3-491
0

By the orthogonality relation all terms of the integration vanish except

L

f L a Y22 do

For further discussion of the Stodola method as applied to this problem, see Reference 3-6.

4. GENERAL
This chapter presents only the essential elements of the methods used so far at the
David Taylor Model Basin to calculate the natural frequencies and normal modes of vibration
of a hull. A recent TMB report 3"11 goes much further into the mathematical details of such
calculations than is feasible in a book dealing with the general subject of ship vibration.
This reference also goes into the extension of the beam theory to the cases in which part of
the mass contributing to the displacement of the ship is considered flexibly attached to the
hull girder. This question is also discussed in Chapters 5 and 13 of this book.

REFERENCES
2-1. Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York (1956).
3-2. Timoshenko, S., "Vibration Problems in Engineering," D. Van Nostrand Co.,
New York (1955).
3-3. Burrill, L.C., "Ship Vibration: Simple Methods of Estimating Critical Frequencies,"
NECI Trans, Vol. 51 (1934-35).
3-4, Todd, F.H. and Marwood, W.J., "Ship Vibration," NECI Trans, Vol. 64 (1948).
3-5. Prohaska, C.W., "The Vertical Vibration of Ships," Shipbuilder, Vol. 54 (Oct, Nov
1947).
3-6. Ormondroyd, J., et al., "Dynamics of a Ship's Structure," Engin Res Inst, Univ of

Michigan, Ann Arbor (Mar 1949) (out of print).

3-7. Boyd, J.E., "Strength of Materials," McGraw-Hill Book Co., New York (1924).

3-23

f~~
 3- 8. Rayleigh, Lord, "Theory of Sound," Second Edition, Dover Publications,
New York (1945).
3- 9. Horn, F., "Horizontal-und Torsions-1chiffschwingungen auf Frachtschiffen,"
WRH, Vol. 6, No. 18 (1925).
8-10. Vane, F.F., "A Guide for the Selection and Application of Resilient Mountings to
Shipboard Equipment," TMB Report 880, Revised Edition (Feb 1958).
3-11. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform
Beams Including Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).

3-12. Kapiloff, E., "Calculation of Normal Modes and Natural Frequencies of Ship Hulls
by Means of the Electrical Analog," TMB Report 742 (Jul 1954).
3-13. Konny, J.H. and Leibowitz, R.C., "Design Details and Operating Procedure for the
TMB Network Analyzer," TMB Report 1272 (Apr 1959).
3-14. McGoldrick, R.T., "A Vibration Manual for Engineers," TMB Report R-189,
Second Edition (Dec 1957).

3-24
 CHAPTER 4

BEAM THEORY OF STEADY-STATE SHIP VIBRATION

A. INTRODUCTION
As in the previous chapter the treatment of hulls given here is based on beam theory.
However, at this point, account must be taken of the limitations of this theory. This leads
to what is called here the "rational" theory of ship vibration. The methods presented in this
chapter are thus essentially heuristic and "quasi-mathematical."
When there is applied to the ideal Euler-Bernoulli beam (otherwise free in space) an
external forcing function, the differential equation applicable to he system is

El --y+ _2 t 41
aZ4 2e

where P (x, t) is the driving force in the Y-direction per unit length of the beam. It is to be
noted that in general P varies both with distance from the left end of the beam and with time.
When the forcing function is specified mathematically particular solutions of Equation [4-1]
can be given, as shown in References 4-1 and 4-2.
For the nonuniform Euler-Bernoulli beam the differential equation has the more general
form

a- 1
l(M+) '2Y=p(Zt) [4-2]
aM2 2 at 2

where El and p now vary with x. Even if the ship were of such construction that El(z) and
I (m) could be expressed mathematically it can be appreciated that Equation [4-2] would have
severe limitations in indicating the manner in which the hull would vibrate under a given ex-
citing force. In the first place, the inertia effect of the surrounding water is accounted for
simply by the added mass component of g; second, there is no dissipation or damping term in
the equation; and third, there is no provision for deflection due to shearing. Last, but not
least, there is nothing to indicate that. the equation is not equally valid regardless of the fre-
quency of the driving force. Thus, whether the driving force has a frequency of 1 cps or
10,000 cps, the patterns of vibratory response should be expected to be beamlike.
On the basis of experience the rational theory must recognize both the beanflike be-
havior of ships in their lower modes of vibration and the sharp departure from beamlike re-
sponse characteristics in the higher modes. The beam theory itself cannot automatically do
this. Although approaches to the vibration analysis of ships other than the beam theory ap-
proach have been suggested (see example in Reference 4-3), one of the eThief aimq in this

4-1.
book is to show that the beam theory can be combined with past experience to yield a pre-
diction of hull vibratory response characteristics of practical use to both the ship designer
and the ship vibration research worker. While a flexural mode of a hull may be excited by
either a force or a moment, only the lateral force is considered at this stage.
In the first method of calculating forced vibration to be discussed, namely, the digital
method, the ship is approximated by a lumped system exactly as in the digital method of cal-
culating normal modes, but now a simple harmonic driving force is applied at one station and
damping forces must be introduced. These damping forces limit the amplitude calculated at
resonance, that is, when the driving frequency coincides with one of the natural frequencies
of the system. The precise nature of the damping process in hull vibration is not well under-
stood at the present time. The subject is discussed at some length in Chapter 8. Here it is
only pointed out that in those calculations presented in this book which involve damping, the
damping action is visualized as produced by equivalent viscous dampers distributed along the
hull. These produce forces opposing the velocity at each point and proportional to that ve-
locity. This involves the use of distributed damping constants c having the dimensions of
force per unit length per unit velocity, and lumped damping constants C having the dimensions
of force per unit velocity. In this book the damping constant is also usually assumed to be of
C
the Rayleigh type, that is, viscous and proportional to mass (which makes - a constant.).
Howevor, there is also used in places a frequency-dependent damping defined by the relation
-- = constant.
Ila) The importance of damping in the attempt to calculate forced vibration of hulls
cannot
be overemphasized since it is the damping alone that limits the amplitude of a mass-elastic
system when resonance is encountered. This subject is discussed in more detail in Charter
8 where it is pointed out that experience has shown the utility of introducing equivalent vis-
cous damping constants based on the rate of energy dissipation even when the actual damping
process is believed to be much more complicated.
A characteristic of the forced vibration of mass-elastic systems in general is that, if
the driving frequency is steadily increased, the forced response exhibits a succession of res-
onance peaks which indicate that the driving frequency coincides with one of the natural fre-
quencies of the system. The minimum points between these peaks are designated as points
of antiresonance. Although the response at the driving point passes through a minimum at an
antiresonance, this is not necessarily true of the response at all points of the system. In
fact, at certain points, there may be a peak of response at the antiresonance frequency.
Just as in the case of free vibrations the differential equations for forced vibration of
ships cannot be integrnt- diraetly, and indirect methods must be devised. The methods dis-
cussed in this chapter are the finite difference or digital method, the analog method, and the
normal mode method. For further details of the theory of beam vibration other than discussed
in this chapter, the reader should see Reference 4-4.

4-2
 B. DIGITAL METHOD
The digital method of calculating forced vibration is illustrated here for the case of
vertical flexural vibration of the hull with no torsion involved. The lumped system to be used
for a 20-section !umping is shown in Figure 4-1. In this case, in addition to the elastic mem-
ber with the rigidities El and KAG connecting the mass elements, there is also inserted an
idealized dashpot between each element and a frame of reference fixed in space. The lumped
damping constant is C = cAz and, in addition, there is an external force P 0 ek"' acting on
one of the elements.

P0 sin wt (or PoeJc) El)

--tAxF

Figure 4-1 - Illustration of Lumped Approximation to a Ship for

Digital Calculation of Forced Vibration

In this problem the digital calculation uses rotating time vectors for all time varying
quantities. These vectors are resolved into real and imaginary components on an Argand
diagram as discussed in Appendix 2 of Reference 4-3. The driving force vector is used as
the reference vector for phase, and this is taken as falling instantaneously on the real axis.
The calculation requires evaluating the real and imaginary components of V, M, y, and y at
all stations to give the steady-state forced response pattern of the hull.
In real form the difference equations are the same as for the normal mode calculation
with the exception of the first (the shear force equation). The latter equations contain the
additional damping and driving force terms. Thus:

AV - C L.- + P (t) [4-3]

2~

AM = VAx + I Ax aY [4-4]
12 at2
t2

MAz
AV = Afx[4-51
El

4-3
 VAw
Ay - yAm - KAG
--G [4- 6]

In the rotating time-vector notation there must now be substituted for V the complex
value V + jV', and similarly for M, M + jlU ; etc. If these substitutions are made in Equa-
tions [4-8] through [4-6], and the differences in the real and imaginary parts are separated
out, there results the following set of eight difference equations:

AV - Axyo 2 + Ccoy' + PO [4- ']

AV'"- =/*Az' co2 _ Oy (4- 8]

AM = VAx - I/Lz Awy6e 2 [4- 9]

AM'- V'A~-- 02 [4-10]

Ax
A= M (4- 11]

El
Ay = M' - [4-12]
Ax

Ay- y- - [4-13]

A = - - [4-14]
KAG

The boundary conditions are

V-V' -M=M' 0o for

L

The linearity of the equations makes it convenient to set the driving force amplitude
P 0 equal to unity. Hence P0 will equal one at the driving point and zero at all other points.
Various schemes of coding this problem for a digital computer may be used as in the normal
mode problem. Here there are four unknown quantities at Station 0 (, = 0), namely, y, Y' Y
and y.: One scheme is to solve the complete set of equations successively with each of the
following sets of initial values and then to find a linear combination of these solutions which
makes V, V', M, and M' all zero at the right end of the system (Station 20).
The solution will give the real and imaginary components of V, Ml, y, and y at all
stations. The magnitudes (absolute values) will be the square root of the sum of the squares

of the real and imaginary components, and the arctan imaginary component will give the

-real comp

4-4
phase angle by which the vector representing the variable in question leads the driving force
vector. If the two vectors are in phase the imaginary component will be zero.

10 0 0

00 1 00
10 0 0, 1 0

If the damping values used are not too high, the calculation will show a relatively
large amplitude when the frequency of the driving force is set equal to or near to one of the
natural frequencies of the hull, as computed in the normal mode calculation. Furthermore, if
the amplitude at this frequency is plotted against z, the pattern obtained will be very close
to that obtained in the normal mode calculation. Further details of the calculation of forced
vibration by the digital method are given in Reference 4-5.

C. ANALOG METHOD
The analog method of calculating the forced vibratory response of a hull on the basis of
beam theory differs from that outlined in Chapter 3 for computing the normal modes only in
regard to the representation of damping. The method for finding normal modes by means of
the analog was based on setting up forced oscillations in the circuit since steady-state con-
ditions furnish a more practical method of electrical measurement than decaying oscillations.
When the analog is to be used to predict forced vibration per se, however, not only the steady-
state response patterns are important, but also the magnitudes of the voltages developed for
given input currents which yield the forced amplitude of the mechanical system.
Practical difficulties are encountered here in the use of a network analyzer for there
is always undesired dissipation in the inductances, capacitances, and transformers which
are treated mathematically as dissipationless. Hence, to set up circuits that are strictly
analogous to the mechanical systems defined by fixed parameters, it is necessary to take
these dissipation effects into account. However, those precautions in the use of the network
analyzer apply principally to predicting the resonant amplitudes of the hull under given ex-
citing forces. For off-resonance frequencies the forced response patterns are determined
chiefly by the elastic and inertia parameters. (See Reference .g.)
Figure 4-2 shows one section of a circuit analogous to a hull subjected to a simple
harmonic driving force Po eic"l This differs from the section of the circuit previously con-
sidered for the normal mode calculation (Figure 8-7) in two essential respects. First, a
1
resistance - is inserted between the shear line and the ground to allow for the mechanical
damping. In the ideal circuit shown here no allowance is made for the dissipation inherent in

4-5
 -pA ia+1

I Am

Figure 4-2 - One Section of an Electrical Circuit Analogous to a Ship

aAm Subjected to a Simple Harmonic Exciting Force

the inductances, capacitances, and transformers (as discussed in Reference 4-6). Second,
there is injected into the shear line at this section a current I equal to the mechanica] force
P0oei. The scaling factors applied in converting from mechanical to electrical values are
not considered here. This current injection device is also an idealization since it has the
property to maintain the current I0 e 1 going into the shear line at this point regardless of the
current flowing in the shear line due to any other sources of excitation present.
Of the four finite difference equations involved in the forced vibration calculations
Equations [4-8] through [4-6], the only one that differs from the equations involved in the free
vibration calculation is Equation [4-3] which involves the driving force P (t) and the damping
2-
term C -. Hence, to show that the circuit of Figure 4-2 is analogous to the ship in forced

vibration (just as the circuit of Figure 8-'7 is analogous to the ship in free vibration), it is
necessary only to show that the Kirchhoff equation for current~s in the shear line is identical
with Equation [4-8]. Equation [3-331 involves three currents, namely, the currents flowing
into and out of the section (V,, and Vn4+ respectively) and the current flowing from the shear
line t~o ground through the condenser (~/LiAaxw 2 y). In Figure 4-2 there are two additional cur-
prpetytominai tecuret 000 gin it te herlie t hi oit egrles-f h
rents, I Pej
P --P(t) and the current through the resistance1- -.1 The latter is
m cAz
equal t thF voltage - divided by the resistance, hence equal to go. Instead of Equation
[3-3], the current equation for Figure 4-2 becomes
currnt
ino uelowngo ay i thoher shar oures f ecitaionpreent

v/n= - iAwr2v/+ C . - P(t)+in+~ 1 [4-15]

at
which is identical with Equation [4-1] since
 Y
2 2 y and AV-Vn+, V,,
22

It must be emphasized that when the network analyzer is used to compute the forced
response of a continuous system, such as a hull, the circuit setup is actually the analog of a
lumped approximation to the hull. Hence, if a 20-section breakdown of the hull is used, the
circuit will in reality have only 20 degrees of freedom. Thus, while the off-resonance forced
response may agree very well with digital calculations, it is subject to the same limitations
as the digital calculations in giving a less and less realistic picture of the hull behavior as
the driving frequency is raised above the frequency range of the first few normal modes of
flexural vibration of the hull. The practical upper limit has been found to be the sixth vertical
mode for hulls of usual length to depth ratios, as indicated in the following section.

D. NORMA. ,MODE METHOD

The normal mode method of calculating the forced vibratory response of hulls presented
here is quasi-mathematical in that it combines only a limited number of terms of a series that
is not converging rapidly in the mathematical sense. The justification is that it has been
found by experience that the higher terms of the series are insignificant in the physical sense.
As a matter of fact, it can readily be shown that in many cases the terms that are discarded
are mathematically larger than the terms that are included.
First it is assumed (on the basis of experience only) that under a simple harmonic
driving force the hull can respond in only a limited number of normal mode components. As a
guide to the number of normal mode components to be used in forced vibration calculations,
the formula originally proposed by Baler and Ormondroyd4- 7 for the number of significant ver-
tical modes of a hull N'may be used as a guide. Assume

5 L L
9 D B

where the letters L, V, and B refer to ship length, depth, and beam, respectively. The ratio
of the number of significant horizontal modes to the number of significant vertical modes may
be taken as 2/3, and of torsional to vertical as 1/2.
While the method involves the selection of the number of terms to be used in the series
on the basis of experience, it still retains certain important properties of the normal modes
which are found in the theoretical analysis of ideal systems.
Three properties of the normal modes of ideal beams are of particular significance
(see Reference 4-8, page 6), namely, the properties of orthogonality, influence function,
and reciprocity.
Orthogonalicy of mathematical functions in general is discussed in Reference 4-9, and
if a series of functions of a single variable possess this property within certain limits of

4-7
this variable, then the integral of the product of any two is zero. Thus

b
f O (w) 01(.) do - 0

if i j. In the case of the ideal beam the orthogonality property involves the "weighting
function" p (z) and is

L
f IL(,r) Yj -)Y -- -

ifipj.
When a beam possesses this property its kinetic and potential energies can be ex-
pressed in terms of vibrations in its normal modes, and, in fact, as shown by Rayleigh, 4-1o
the kinetic and potential energies of mass-elastic systems in general can be expressed in
terms of the squares of generalized coordinates each involving the deformation or velocity
in one of the normal modes. Thus

and
P). E. k [4-17)

where K. E. is the kinetic energy,

P. E. is the potential energy,
jj is the generalized velocity in the ith normal mode,
n1 is the generalized displacement in the ith normal mode,
Sis a generalized mass applicable to the ith normal mode, and
ki is a generalized elastic constant applicable to the ith normal mode.
A more precise definition of these terms is not essential here.
The normal mode influence relation states that the normal mode pattern determines the
influence of the point of application of a simple harmonic driving force on the magnitude of
the amplitude excited in that mode. Thus, a given force will not excite a mode which has a
nodal point at its point of application, and will excite the maximum amplitude in this mode
when applied at the point at which the normal mode pattern is a maximum.
The reciprocity relation implies that, if a simple harmonic driving force applied at a
point xr of the ideal beam produces an amplitudo of vibration Yrs at T., then

Yr -" yar [4-18]

4-8
In other words, the same amplitude will be measured if the points of application of the driving
force and measurement of response are interchanged.
Without further discussion of thase very important properties of ideal beams it is merely
emphasized here that the great simplicity of dealing with mass elastic syste-u.1 in terms of
normal modes was demonstrated by Rayleigh in his "Theory of Sound."I It has been shown
in recent years that the free-free nonuniform beam with shearing and bending flexibility, a
close relative of the Timoshenko beam, retains these properties (see References 4-8 and 4-11).
In the case of the ship, which retains beamlike characteristics only within a limited
frequency range, the use of these concepts, the use of a limited number of terms in normal
mode response series summation, and the use of Rayleigh damping is largely intuitive. This,
however, does not impair its utility. Regardless of how the normal modes of the hull are de-
rived, if only flexural modes of vibration are involved, they may be used to compute the re-
sponse to a simple harmonic driving force by means of the equation derived from the beam theory

2]
i- N "P Yi ( m0)Yi ( )--sin ( 0 it - [41)
S+
22
2;[-9

where the force P sin wt is acting at x and N' is the number of significant normal modes.
This equation is discussed in detail in Reference 4-8.
The independent behavior of the normal modes of mass elastic systems permits the use
of the convenient concept of effective systems of one degree of freedom, each representing a
normal mode. There are various ways of defining such effective systems. They may be de-
fined with respect to a particular driving point, or without this restriction. When the hull is
represented by an equivalent lumped system, its effective mass at a driving point din the ith
normal mode is

y.my12
Mdi =
Ydi21
[4-20]

where Y, is the amplitude at any station in the ith normal mode pattern, Ydi is the normal
node amplitude at the driving point, and the summation includes the total number of lumps.
The effective spring constant of the one-dogrA, system is given by the equation

H-- wi22K [4-21]

4-9

A
and the effective damping constant is given for Rayleigh damping by the equation

Cdi - (constant) x Mdi [4-22)

The effective system can thus be visualized as the familiar one-degree-of-freedom sys-
tem usually considered in vibration theory and shown in Figure 4-3.

P sin cot -- V Mdi Figure 4-3 - Effective System of One Degree

of Freedom Representing a Hull in One

If each normal mode is thus treated, the component of driving point amplitude in each mode is
found from the familiar equation for the one-dvgrue system:

P sin (rit - [4-231

2
Y (K - M ) , c 2

where the effective values are used in each case for K, M, and C. To obtain the net response
the component of response at the driving point is multiplied by its normal mode function, and
the patterns are combined with account. taken of the algebraic signs. The result will be the
same as given by Equation [4-19] if the same number of terms is used in both cases.

E. METHOD OF MECHANICAL IMPEDANCE

The term impedance, well known in electrical engineering, has become familiar in me-
chanical vibration analysis only recently. It is most commonly used in conntction with the
rotating vector to represent a steady-state vibratory component. Although a mechanical im-
pedance may be defined on the basis of either displacement amplitude or vniocity amplitude,
the American Standards Association's Committee on mechanical shock and vibration prefers
the latter. This stems from the fact that this committee wav formerly affiliated with the ASA
Committee on Acoustics. In the latter field the acoustic impedance concept has been in wide
use for some time.
When the mechanical impedance is based on displacement it is the ratio of the ampli-
tude of the driving force to the displacement amplitude at the driving point. Thus

4-10
 P

In this equation all three symbols represent rotating time vectors and are complex numbers.
For the system of Figure 4-3 it then is found, if the subscripts are omitted, that

Z = K - Mro2 + jC [4-24]

Since, in complex notation, the vibratory velocity amplitude is obtained by multiplying the
displacement amplitude by j&), it follows that the mechanical impedance based on velocity
can be obtained from the mechanical impedance based on the displacement by dividing by
jc. Thus

K
Z" =C+jM- -( [4-251

The similarity between this and the well-known expression for the electric impedance of a
circuit having resistance R, inductance L, and capacitance C in series, namely,

j
Z = R + J (,,L - -[4-261

is noteworthy.
If the ship actually possessed only N-normal modes, in each of which its behavior
conformed with the beam theory, the displacement impedance in each mode referred to the
driving point could be found by computing

2
Zdi= Kdi - Mdi Cd + jPCdi [4-27]

To find the net impedance, the impedances in the individual modes must be combined recip-
rocallv qince the net amplitude is the vector sum of the amplitudes in the individual modes.
Hence

1
11 1 [4-28]
1- + I- +-.....
--
Zdl Zd 2 Zdn

Tho awkwardness of dealing with the reciprocals of complex quantities can be cir-
cumvented by resorting to the concept of mechanical admittance which is the reciprocal of
mechanical impedance. Thus

4-11
 be
1 Y
A .... [4-291

and

Y = PA [4-30]

The admittances in the various modes combine by direct addition to give the net driving point
admittance; that is,

Ad fI A A[4-31]
Adi

It is a common observation that, when the blade frequency of the propulsion device is
well above the range of significant hull mode frequencies, the forced vibration of the hull is
concentrated in the stern of the ship, and settles down to a filrly constant level regardless
of the speed, unless a local resonance of some structure in the stern is encountered.
From Equation [4-24] it can be seen that at high frequencies, the inertial component
is the major component of impedance vhen the damping is of the Rayleightype. If the elastic
1
and damping components are then neglected, the admittance becomes - and the net
admittance has the form Mdi w

Ad dw =2 27--"Md [4-321

This indicates that under forces increasing as the square of the frequency, the stern amplitude
will remain constant at shaft speeds above the range of significant hull Criticals.
On the basis of this reasoning, formulas have been proposed for estimating stern ampli-
tudes for given driving forces when the prescribed conditions are met; see Reference 4-3.
These formulas use empirical constants and the ship's displacement. It is hoped that data
accumulated in the future will indicate to what extent a single constant can be used for ships
of widely varying types. The formulas are as follows:
For vertical vibration

Y P [4-331
2
3.4 x 1O- 6 D x (cppm)

whore Y is the single amplitude at stern in mils,

P0 is the amplitude of the driving force in lb,

4-12
 D is the displacement of ship in long tons, and
cpm is the blade frequency in cycles per minute.
For horizontal vibration
P0
Y PO 2 [4-841
1.9 x 0-6 x D x (cpm)

For torsional vibration

r0
TO
S- 2 [4-35]
x (cpm)
0.46 x I

where q5 is the single amplitude at the stern in radians,

To is the blade frequency driving torque in lb-ft, (single amplitude),
I is the mass moment of inertia of entire ship about longitudinal axis through its
c.g. (with no allowance for added mass) in ton-sec 2-ft, and
cpm is the blade frequency in cycles per minute.
It is seen that in all three cases, if the driving force increases as the mquare of the frequency,
the formula gives a constant amplitude. These formulas are readily converted to formulas for
vibratory velocities in lieu of vibratory displacements.

REFERENCES

4-1. Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York (1956).
4-2. Timoshenko, S., "Vibration Problems in Engineering," D. Van Nostrand Co., New
York (1955).
4-3. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
4-4. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 131T (May 1961).
4-5. Adams, E.J., "'The Steady-State Respnese of a Ship Hull to a Simple Harmonic
Driving Force Computed by a Digital Process," TMB Report 715 (May 1950).
4-6. McCann, G.D. and MacNeal, R.H., "Beam-Vibration Analysis with the Electric-
Analog Computer," Journal of AppI Mech, Vol. 17, No. 1 (Mar 1950).
4-7. Baler, L.A. and Ormondroyd, J., "Vibration at the Stern of Single-Screw Vessels,"
Trans. SNAME, Vol. 60 (1952).

4-13
 4- 8. McGoldrick, R.T., et al., "Recent Developments in the Theory of Ship Vibration,"
TMB Report 739, Revised Edition (Oct 1958).

4- 9. Wylie, C.R., "Advanced Engineering Mathematics," McGraw-Hill Book Co., New

York (1951).
4-10. Rayleigh, Lord, "Theory of Sound," Second Edition, Dover Publications, New York
(1945).
4-11. Ormondroyd, J., et al., "Dynamics of a Ship's Structure," Engin Res Inst, Univ of
Michigan, Ann Arbor (Mar 1949) (out of print).

4-14
 CHAPTER 5

BEAM THEORY OF TRANSIENT SHIP VIBRATION

A. INTRODUCTION

Although the steady-state vibration of ships under normal operating conditions is a

very important consideration in naval architecture, it is clear that under rough sea conditions
much more severe vibrations of a transient nature are encountered. If quasi-mathematical
methods must be resorted to in dealing with steady-state ship vibration, this applies to a
much greater degree in dealing with its transient vibrations. It is to be noted at once that
the transient vibrations of ships are occasionally of such large amplitude as to deform the
hull girder beyond the linear range within which the beam theory of hull vibration is con-
sidered to have a fair degree of validity.
Obviously, just as in the case of steady-state vibration, the transient vibrations of
ships depend on both the exciting forces and the dynamical properties or response charac-
teristics of the hull. The forces, however, in this case are quite complex and cannot be
expressed in such simple mathematical terms as can the steady-state forces.
Much progress has been made in recent years in correlating the stress variations and
motiono encountered in ships in a seaway with the statistical data available on ocean waves;
see Reference 5-1. It is merely pointed out here that useful correlations have been discovered
between the statistical distributions of wave heights encountered at sea and the distributions
of motions and stresses in hulls. All that is attempted in this chapter is to indicate that,
within the linear range of deflections, the rational theory applied to steady-state forced vi-
bration in Chapter 4 should also be applicable to transient vibrations. -The linear range is
even lower than the range of deflections within which hull girder stresses reach the yield
point of shipbuilding steel, since buckling of members in compression will ordinarily occur
before this point is reached.
The value of the consideration here of the transient response of hulls to low magnitude
excitation is that it provides the naval architect with insight into the processes that take
place before damage actually occurs under slamming conditions in a seaway. Furthermore,
as far as possible damage to local structures and to equipment installed in the ship is con-
nerned, it points the way to avoid some of the alterations that frequently have to be made
after the builders' trials of a ship. There is no implication intended here that, up to this
time (1960), experimental verification has been obtained of the adequacy of the treatment of
the response of ship hulls to transient loading by the rational beam theory advocated in this
book, or in fact by any ether theory. This must await the verdict of investigations still
underway. Although it is shown in this chapter how the same general methods applied in
Chapter 4 to steady-state vibration are extended to the transient case, the presentation is
only analytical and does not encompass a practical evaluation of these methods.

f. I1
For simplicity and emphasis on basic principles, the discussion is restricted to flexural vi-
brations of the hull.

B. NORMAL MODE METHOD

The normal mode method of dealing with the transient response of ship hulls presented
here is discussed in more detail in Reference 5-2. Basically, it rests on the assumption that
in its flexural response, even when account is taken of its shearing flexibility and damping
characteristics, the hull satisfies the conditions of a Rayleigh system. This means that the
complex vibrations excited by transient loads can be analyzed by considering the separate re-
sponse in each of its normal modes in the general scheme applied to mass-elastic systems by
Rayleigh.5-3
The system comprising the hull and the surrounding water is idealized as an unre-
strained beam loaded by an arbitrary forcing function and governed by the following set of
partial differential equations:

at2 a~t ax
V ax
am
v -- -- [5-2]

-ax = P,+ [5-81

V = - KAG ' [5-41

M= a-y [5-5]

whore c is the viscous damping force per unit length per unit velocity,
P(r, t) is the external forcing function giving load per unit length varying arbitrarily
with respect to both x and t, and
is the component of slope of elastic line due to shearing only.
The other quantities have been defined previously.
As indicated previously, the modes of free vibrations of such a system (when c = 0,
and P (z, t) = 0) can be found only by graphical, finite difference, or analog methods if E1
and KAG vary with -. Nevertheless, it is shown in Reference 5-4 that the normal modes of
such a system retain the property of orthogonality; that is,

5-2
 L
FcY1 (x~) Y,(zdax= 05 Lu

where i .
It is shown in References 5-2 and 5-4 that, in accordance with the orthogonality rela-
tion, the dynamical behavior of this system can be treated in terms of series of responses in
each of its normal modes.
It is also shown in Reference 5-2 that even with damping (c 4 0), if it is of the Rayleigh
type, that is, viscous and proportional to mass (. = constant) ,the response to the arbi-
trary forcing function P(z, t) can still be treated in a series of normal mode responses.
With no pretense at mathematical rigor it has been assumed, on the basis of practical
experience, that under transient loading the hull will respond in only a limited number of
beamlike modes. Here it is presupposed that these modes have already been determined by
methods outlined in Chapter 3.
As shown in Chapter 4, in the dynamics of the ideal beam system representing the
hull, each normal mode of vibration is treated as a vibratory system of a single degree of
freedom having definite values of mass, spring constant, and viscous (amping constant
suitably derived.
Such systems have been derived and discussed on the basis of two slightly different
concepts which it is well to clarify at this point.
In one scheme (illustrated on page 21 of Reference 5-2), a generalized coordinate
q!(tMwith the dimension of length, is used to represent the displacement of the system in
its ith normal mode. The motion of the system in that mode is then given by multiplying
qj(t) by the normal mode function where the latter is considered dimensionless (that is,
merely a pattern of relative displacements at different distances from the left end of the sys-
tem). Then the response of the system is given in terms of its normal modes by the series
relation

Y(X,) 0- itN" q(t) (x) [5-71

where N' is the number of significant normal modes. The generalized force Qj(t), which in
this case also has the dimension of force, is defined by wiCe it&".Qi,

L
Qj (t) = f P(z, t) Y (z) do
0

In this scheme there are associated with each normal mode generalized or effective
masses, damping, and spring constants defined by the relations

5-3
 L

and

Ki M

where ci is the undamped natural frequency associated with the ith normal mode.
The other scheme, based on the same general theory, is more useful. Here the forcing
function is concentrated at one point (called the driving point d). In this case the behavior
of the system is derived by visualizing each normal mode of the system as presenting to the
driving force at d an effective inertia, spring constant, and damping constant, all based not
only on the normal mode pattern of each mode but also on the location of the driving point
itself.
Thus, if the external force acting at d is F(t), then

Qd(t) = F(t) Yi(Xd) [5-121

r . v. 2 t' _,, -

Mdi [lI3]
y2 (Xd)
L

f L(X) y.2 (_) d_

Cda - [5-14]
1 2 (xd)

and

Kdi ci M1 [5-15]

It then results that the response in each normal mode is governed by the same equaLion
as the vibratory system of one degree of freedom shown in Figure 3-1; that is,

MdiYdi + Cdiydi + kddidl - Qdit) [5-16]

It is shown in numerous text books (e.g., Reference 5-5) that when an external force
starts to act on such a system, if at rest at t- O, the response at time t is given by the
equat'ion

5-1
 d - (Cdi/ 2MOd)(t -

"edi fi 0ti i Mdi sin A (t-'r)dr [5-17]

where

/Cdi 2

Hence the general response of the system is given by the equation

I,

Y0 - [5-191

If a single impulse H is applied at the driving point d, the same theory yields the
equation for the response in the ith normal mode:

N Yi (0d) - Cdil/2 Mdi

Ydi(t) - xi e sin A. t [5-20]

As in the steady-state problem, the summation is to be carried out only for the number
of modes for which the beam theory is considered valid, Although the, normal mode patterns
are treated here as continuous functions, the same basic equations may be applied when the
normal modes have been computed by the digital method and are available only in tabular form.
In this case summations are substituted for integrations in evaluating the effective parameters.
Thus

M. y.ix 2
Mla = [5-21l
di

where the subscript x indicates the X-coordinate of an individual lumped mass.

It must be remembered that this transient analysis is based on the treatment of the hull
as a beam free in space. Such a beam has two rigid body modes of zero froqueney, namely,
the heaving mode and the pitching mode. If the components in these modes are included in the
series summation, the calculation will not be realistic if carried out over an interval of time
long enough for large rigid body rotations to develcn. In the actual ship case, heaving and
pitching are controlled by the buoyancy of the water and the effect of gravity, and the cor-
responding natural periods are finite. In practical transient problems, the principal vibrations
have been ex;ecuted by the hull before appreciable rigid body motions have had time to build up.

5-5
 C. DIGITAL METHOD

The digital method of treating the hull transient response problem requires no prior
computation of the normal modes. It uses finite differences in both time and space, and, in
the time domain, requires the establishment of stability criteria which will ensure that the
time steps are small enough.
The process, of course, involves the same basic concepts of the beamlike behavior of
the hull as discussed previously. As elsewhere in this book, the hull is treated here as a
beam free in space with a distributed mass added to allow for the inertia effect of the sur-
rounding water. Distributed viscous damping is also included. The physical picture of the
process is as follows.
The hull is approximated by the lumped system shown in Figure 5-1 as having 20
sections. As in the steady-state forced vibration problem, an ideal viscous damper is in-
serted between each lumped mass and a reference frame fixed in space.

fAG x

0 20

Figure 5-1 - Lumped System Used in Treating the Transient Response

of the Hull by the Digital Method

At the instant when the transient load is applied, all stations of the hull have
specified displacements and velocities relative to a set of axes fixed in space. In general,
both these displacements and velocities will be zero. However, there will be accelerations,
since external forces are now acting. The forces which actually may be varying continually
with respect to time are considered as held at fixed values for a short interval of time, and
then as changing instantaneously to another value, and so on. It is then possible, by the use
of the beam equations and the boundary conditions, to compute the acceleration at all stations
at t = 0. If then a step is taken in time during which these accelerations are assumed con-
stant, the velocities at all points can be calculated, and, a short time later, the displacements
due to these velocities can be calculated. Thus, as long as the external forces all along the

5-f]
hull are known at any instant of time, it is possible to repeat this cycle of operations indef-
initely and thus to compute displacement at all points of the hull as they vary in time. This
also involves computing the elastic deformations, and so the shearing stresses and bending
moments are also found in the process.
The mathematical problem of establishing the criteria for stability of the finite dif-
ference calculation involved here is discussed in Reference 5-6. The stability criteria
given there are:

At
0

and

(At)' < )
Ill,(KAG) + I(FI) +0.25 i (Am) 2 (KAG)

The finite difference equations in this case involve both differences with respect to 0
and differences with respect to t. The former are the same as in the forced vibration problem
except that simple harmonic motions and complex number notation are not required here. In
Figure 5-2 a section of the hull of length Ax, the distance between stations indicated in Fig-
ure 5-1, is shown uncompressed (not lumped at one point) so that its elastic properties may

//Ir

Figure 5-2 - Free-Body Diagram of an Element of a Ship of Length Ac Subject kA)

Forces and Moments Accompanying Transient Vibrations

5-7
be visualized. Here its mass could be conceived as concentrated at the midspan. It must be
emphasized that in any scheme of representation of the.hull as an equivalent lumped system,
for the purpose of setting up difference equations for a digital treatment, there will be con-
flicting requirements in the schematic representation. It is not attempted here to indicate the
notation which will yield the maximum accuracy for a given number of sections. Accordingly,
the equations in the differences with respect to x are given in simplified form as follows:

AV = - IzAxj"- Cy + F 15-22]

AN = VAX + IJZ AxjV [5-23]

MAaY
AV = M
El [5-241

Ay = VAX- VAX
KAG [5-25]

It should be noted that in Equation [5-23] the term for rotary inertia is included, where-
as in Equation [5-21 it is neglected. Experience so far has shown that its inclusion mnsakes
little difference in hull vibration calculations, and the demonstration of orthogonality of the
normal modes given in Reference 5-4 was based on the set of equations in which the term
was omitted. Also, although it is not exploited here, it might also be noted that the finite
difference method is not restricted to Rayleigh damping but can be applied to both nonlinear
and linear systoems.
The equations in the differences with respect to time (also in simplified form) are

A;- , At [5-26)

Ay s = ;n At [5-271

It is necessary in this case to use both subscripts and superscripts. The former
indicate the space coordinate, the latter the time coordinate.
If the calculation starts from the rest condition, the boundary conditions are

*0 0o 0= 0_

= H 0 0

At any instant of time there are sufficient equations to compute the accelerations at all points
along the beam. It is then possible to compute the velocities and displacements at the next
time step, and thus to continue the process indefinitely.

5-8

-I
 In this form of the problem, since the hull is considered free in space, no gravity or
buoyancy forces are acting. Al though the calculation will give the net result of the com-
bination of rigid body motions and elastic deformations, the rigid body motions will not be
realistic if it is carried out for a total time comparable with the pitching or heaving periods
of the ship because there are no restoring forces to limit the rigid body displacements.
Obviously, many variations of this problem are possible. Thus the damping forces
need not be restricted to the Rayleigh type and gravitational and buoyancy effects can be
included. As given here the virtual mass is constant in time and thus does not vary with
amplitude, frequency, or mode shape. While the method theoretically takes care of any elastic
wave effects that may exist in such a system, no use is made of the traveling wave concept
in setting up the problem.
Further discussion of this calculation is given in References 5-6 and 5-7 and a sample
calculation is given in Appendix B.

D. ANALOG METHOD
Both the classical (conventional) and the mobility analogs discussed in Chapter 3 are
applicable to the hull transient response problem as computing devices. It cannot be said,
however, at this time that their potentialities in this field have been more than superficially
explored.
While the networks representing the inertial and elastic parameters of the hull con-
sidered as a beam are the same as for the steady-state vibration problem considered in
Chapter 4, the techniques of exciting the network and measuring its response are obviously
radically different.
Instead of a simple oscillator capable of injecting a sinusoidal current (in the mobility
analog) at any frequency desired within a given range, the transient problem requires a special
transient injection circuit which can deliver a current of the desired waveform regardless of
the impedance characteristics of the network. Thus it may be desired to inject a single half.
sine pulse, a reotangular pulse, a triangular pulse, or a current pulse that rises "instantly"
to a given value and then decays exponentially.
The measurement of transient response is also more complicated than in the steady-
state case. In that case, an oscillogram is not really necessary if both an indicating a-c
voltmeter and a phasemeter are available. In the transient case, a record must be made on
an oscillograph or on magnetic tape of the transient signal at a number of points along the
network.
In lieu of a single transietr injection, it is often preferable to inject a series of input
signals at intervals far enough apart to permit the transient to die out between intervals. If
these intervals are not too far apart they will permit direct observation of the response at any
point in the network on a cathode ray oscilloscope since the pattern will be retained both by
the screen and the retina.

5-9
 Further details of the use of the electrical analog for such calculations will be found
in Reference 5-8. Obviously, at best, the analog predictions of hull response cannot be any
more reliable than the theory from which the analogous circuit is derived. Furthermore, the
analog may give a distorted picture of the true response as predicted by the mathematical
theory used. The advantages of the analog, however, in permitting the operator to vary the
inertia and stiffness parameters of the hull simply by turning dials make this method an
attractive one.
A recent proposal by Dr. N.H. Jasper considered by the David Taylor Model Basin is
the development of a transient computer for ship hulls which forecasts the stresses and vi-
brations encountered in seas of various wave heights. This is discussed in Reference 5-9.
The idea is to use an electrical network to represent the hull as a beam, but, instead of treat-
ing it as an ideal system free in space to which loads that are known with respect to time are
applied, the computer automatically applies the loads that are exerted by the sea, allowing
for the fact that the buoyancy and added mass effects vary with the rigid body motions of the
ship.

REFERENCES
5-1. Jasper, N.H., "Statistical Distribution Patterns of Ocean Waves and of Wave-Induced
Ship Stresses and Motions, with Engineering Applications," Trans. SNAME, Vol. 64 (1956).
Also TMB Report 921 (Oct 1957).
5-2. McGoldrick, R.T., et al., "Recent Developments in the Theory of Ship Vibration,"
TMB Report 739, Revised Edition (Oct 1953).
5-3. Rayleigh, Lord, "Theory of Sound," Second Edition, Dover Publications, New York
(1945).
5-4. Ormondroyd, J., et al., "Dynamics of a Ship's Structure," Engin Res Inst, Univ of
Michigan, Ann Arbor (Mar 1949) (out of print).
5-5. von Karman, T. and Biot, M.A., "Mathematical Methods in Engineering," McGraw-Hill
Book Co. (1940).
5-6. Polachek, H., "Calculation of Transiant Excitation of Ship Hulls by Finite Difference
Mothods," TMB Report 1120 (Jul 1957).
5-7. McGoldrick, R.T., "Calculation of the Response of a Ship Hull to a Transient Load
by a Digital Process," TMB Report 1119 (Mar 1957).
5.8. McCann, G.D. and MacNeal, R.H., "Bean-Vibration Analysis with the Electric-
Analog Computer," Journal of Appl Mech, Vol. 17, No. 1 (Mar 1950).
5-9. Hoyt, E.D., "Simulation of the Sea and Analog Computation of the Forces on a Ship
in Waves," Report to TMB on Contract Nobs 72376, T.O. 9, Project RR-1458-N by Reed
Research, Inc (Nov 1959).

5-10

j
 CHAPTER 6
EFFECT OF LOCAL FLEXIBILITY ON THE VIBRATORY
CHARACTERISTICS OF A HULL

A. INTRODUCTION - THE "SPRUNG MASS" EFFECT

While the chief aim in this book is to exploit the dynamics of a ship's hull when con-
sidered as a nonuniform free-free Timoshenko beam, it has been pointed out that in its vi-
bratory behavior, a ship departs from such a beam to an increasing extent as the driving
frequency rises. In the response to a steady-state simple harmonic driving force, this de-
parture frvm ideal beam behavior is evident in two respects of special significance to the
naval architect: First, at high driving frequencies the amplitudes of vibration at points in the
hull lying in the same transverse plane (normal to the longitudinal axis of the ship) cease to
be the same. Second, when the blade frequency is higher than the frequency of (roughly) the
sixth vertical flexural mode, the propeller-excited vibration is usually concentrated at the
stern of the ship.
When the amplitudes of vertical vibration at points lying in the same transverse plane
are not the same the situation is ascribed colloquially to "local flexibility." This local
flexibility, which at low frequencies may completely escape attention, may become so pro-
nounced at high frequencies as to fully control the vibratory response. For instance, if a
mechanical vibration generator were installed on the main deck of a ship in either the bow or
the stern, but in the middle of a panel of deck plating, it would be found that the lower fiexural
modes of the hull could be readily excited, but that beyond a certain running speed (driving
frequency), the machine would be quite incapable of exciting the hull. This is true in spite
of the fact that its driving force amplitude increases as the square of the speed. If, however,
there are installed under this deck heavy shoring members that transmit the load to points on
the deck below, where there are either vertical bulkheads or sections of shell plating, then the
range of frequencies over which the hull can be excited by the vibration generator is greatly
extended.
In dealing analytically with local flexibility, it has proven fruitful to'consider the
vibratory characteristics of an ideal beam having one or more masses elastically attached to
it. The attached mass has been designated a "sprung mass" and thus the effect of local
flexibility is often spoken of as the "sprung mass effect." In this book the term sprung mass
is applied tw local elastic structures themselves; to relatively rigid assemblies that are
supported in the hull by means of resilient mountings; and trileavy items of equipment that
are installed on foundations nominally rigid, but which in practice exhibit flexibility as a
consequence of the large mass attached to them. A discussion of the properties of the ideal
beam with an attached sprung mass is given in Reference 6-1.

t-i-
 B. LOCAL ELASTIC STRUCTURES
The chief structures falling in the category of local elastic structures are dock panels
(including plating and stiffeners supported only at the edges), longitudinal and transverse
bulkheads, panels of shell plating, stiffened shell plating panels supported only at the edges,
deck houses or superstructures, masts, docking keels or skegs, shaft struts, and control
surface members such as rudders or (in the case of submarines) diving planes.
The effect of these structures on the vibratory characteristics of the hull as a whole
depends chiefly on two properties, namely, the local natural frequency ard the local effective
mass. Both properties require special consideration here. By the "local natural frequency"
is meant the natural frequency that would be measured if the surrounding structure could not
move. The identification of local natural frequencies in practice is not always clear, and,
strictly speaking, any natural frequency observed may be considered as the frequency
associated with a mode of vibration of the entire hull system. Obviously, large local structures
cannot vibrate independently of the ship as a whole when the hull is unrestrained. When they
are excited by impact, the frequency measured locally is then definitely a frequency of one
of the modes of vibration of the hull considered as a beam, with the local structure acting
as a sprung mass. If, however, the test is made in drydock, in which case the hull is re-
strained, the observed frequency may be a true local natural frequency.
The mass of the local structure, or rather, the ratio of the sprung mass to the mass of
the ship, furnishes a criterion of the distinction between a local natural frequency and the
local manifestation of a hull natural frequency. Unfortunately, the precise determination of
such a criterion is not feasible. Moreover, it is not the actual mass of the local structure
but its effective mass that furnishes the criterion.
The concept of effective mass was discussed in Chapter 4 in connection with the
determination of the driving point mechanical impedance of the hull. It will be recalled that,
from the impedance point of view, at any desired driving point the hull may be considered to
present an effective mass for each of its normal modes of vibration. This mass has such a
value that, if vibrating with a given amplitude at the frequency corresponding to the normal
mode ini question, it will have the same kinetic energy as will the entire hull if vibrating in
this normal mode with the same amplitude at this driving point. The same concepts are
applicable to the effective mass of the local structure. Although it is usually the fundamental
mode of vibration of the local structure that is of concern, this is not always the case. Where
more than one mode of vibration of the local structure is of concern, an effective mass must
be evaluated for each mode.
In the mode involved the local structure presents to the hull the effect of a sprung mass
attached at a selected point. The point at which the equivalent sprung mass is considered
attached is somewhat arbitrary, and the value to be assigned to the effective mass depends
on the point selected. In general, this point may be taken at the center of gravity of the local
structure. The normal mode pattern and also the local natural frequency for the local structure

6-2
must be known approximately. The effective mass is then such a mass that, if vibrating at
this frequency and with unit amplitude, it will have the same kinetic energy as the local
structure would have when vibrating in the mode in question with unit amplitude of the center
gravity; the ship itself is considered at rest (restrained) in both cases.
In general, the determination of the effective mass of the local structure would have
to be carried out by an approximate method in which its volume was broken down into a large
number of elements; to each element an amplitude was assigned in accordance with the known
or assumed normal mode pattern. Then

me = -2 [6-i]
eg Y'9g2

where me is the effective mass of the local structure at its center of gravity for the
c9 local mode of vibration in question,
m is the mass of one of the elements into which the local structure is broken
down for the evaluation of effective mass,
y is the amplitude (in any arbitrary units) of the element m when the local
structure is vibrating in this mode, and
Y'9 is the amplitude in the same arbitrary units at the e.g. of the local structure
when vibrating in this mode.
Once the effective mass has been evaluated, the effective spring constant can be found from
the relation
,oo =[6-21
'g
cg

whore oo is the natural circular frequency of the local structure and ke is the effective
spring constant to be associated with m Cg.
Reference 6-2 discusses experiments conducted on a cargo ship in which the experi-
mental data indicated a considerable departure in the vibration characteristics of the hull
from those predicted by the beam theory. This reference indicates that there was reason to
believe that local flexibility accounted for the departure from beamlike vibratory response
characteristics. Exploratory calculations were later made by the electrical analog in which
a circuit was set up to represent this hull as a beam but with part of the mass flexibility
attached to the main hull girder. As indicated in Reference 6-3, these calculations showed
qualitatively that the sprung mass effect could account for the observed vibratory response
characteristics.
While sufficient information has not been obtained at this time to fix definitely the
magnitude required of flexibly attached masses before their effect on the normal modes of
hull vibration becomes perceptible, the following criteria are suggested here: (1) the effective

6-3
mass of the local structure is of the order of 1/2 percent or more of the displacement of the
ship, and (2) the local natural frequency is within the range of significant hull frequencies
(roughly below 800 cpm). When both of these conditions are met, it is not expected that the
usual normal mode pattern of beamlike form as discussed in Chapter 8 will apply to the ship
in question.
It is then necessary to redefine the problem in terms of a beam with sprung masses. In
doing this, account must also be taken of the fact that the local structure may react so as to
affect not only flexural modes but also torsion-bending modes of the hull. In the latter case,
the concept of the sprung mass representing the local structure must be generalized to take
care of both the translational and rotational effects. This extension of the sprung mass
concept is discussed in Reference 6-4.

C. RESILIENTLY MOUNTED ASSEMBLIES

The case of resiliently mounted assemblies requires special consideration here since
in recent years there has been a trend toward the resilient mounting of massive elements such
as diesel engines, auxiliary turbines, and turbogenerator sets. Since the isolation of such
assemblies from the hull requires a relatively "soft" mounting system, both of the conditions
previouly stated for effecting the vibratory response characteristics of the hull are met, That
is, the local item is significantly massive relative to the hull and ito natural frequency on its
resilient mountings falls below the upper limit of significant hull mode frequencies. A typical
resilient mounting for shipboard equipment is shown in Figure 6-1. Usually at least four such

Part of Assembly -X X Part of Mounting

/ ffcctive Point
of Attachruent

.Z Z

A!

Figure 6-1 - A BST-Type Resilient Mounting for Shipboard Equipment

mountings will comprise the set supporting a single assembly. Figure 6-2 shows schematically
a very common mounting arrangement. For further information on resilient mounting of shipboard
equipment see Reference 6-5,

A-4
 A resiliently mou,,ted rigid assembly has six degrees of freedom and thus there may be
six local sprung mass effects to consider. Fortunately, it is usually possible to design the
local system to have planes of "vibrational symmetry." A plane Is said to be a plane of
vibrational symmetry when vibrations in this plane produce no tendency for the system to
vibrate in translation in directions normal to the plane or in rotation about axes lying in the

Y-Assembly

Figure 6-2 - Schematic Illustration of a Resiliently Mounted Assembly with a

Typical Base-Mounting Arrangement

plane. As shown in, Reference 6-5, the presence of this condition greatly simplifies the
vibration analysis.
Even in the simplest type of ship vibration; namely, the vertical, which in general is
independent of horizontal and torsional effects, it is clear that a resiliently mounted assembly,
if sufficiently massive, can excito tha huiJ in more than. one way. If motion confined to a
vertical plane through the longitudinal axis of the ship is considered, it is clear that such an
assembly can affect the flexural hull modes when the ship is slamming. Moreover, if the
assembly is located at a node of a certain flexural mode, which therefore it cannot excite by
heaving, it may readily exciLe this mode by pitching or rocking (a combinatioL of pitching

6-5
 ZY "r f-r
Figure 6-3 - Right-Hand Coordinate System
Used in Calculation of Natural Frequencies
and Normal Modes of Vibration of a Re-
sillently Mounted Rigid Assembly
o
0 " -X'

about the c.g. with translation of the e.g.). As a matter of fact, in this case it is impossible
to find a location in the hull at which neither pitching nor heaving of the assembly would
induce flexural vibration of the hull.
In such a case, therefore, the effect of the local structure on the hull vibration charac-
teristics cannot be represented by a sprung mass of a single degree of freedom. Both the trans-
lational and rotational effects must be taken into account. To do this the mass mo.ment of
inertia of the assembly with respect to an axis through its e.g. normal to the vertical plane
through the centerline of the ship, as well as its mass, must be known.
To analyze the vibration of a rigid assembly supported by a single set of resilient
mountings, a fixed set of axes is taken with origin at the e.g. of the mounted assembly in its
rest position. A right-hand system is used and the X-axis is taken parallel to the longitudinal
axis of the ship; see Figure 6-8. The individual mountings will have principal elastic axes
that, in the most general cases, may not be parallel to the axes chosen for the calculation.
Since a very common type of resilient mounting has an axis of polar symmetry (such as that
shown in Figure 6-1), this is generally called the axis of the mounting; such a mounting is
said to have an axial spring constant k0 and a radial spring constant k,, each being the re-
storing force for unit displacement in the appropriate direction. If the mounting is displaced
into its nonlinear range under the gravitational load of the mounted assembly, then the values
of ka and kc to be used must be based on small displacements from the loaded position.
Mountings not having an axis of polar symmetry can be treated as combinations of mountings
having only axial stiffness.
If the mounting illustrated in Figure 6-1 is taken as an example, it is apparent that
an arbitrary motion of the assembly relative to the base (involving both translation and rota-
tion) will evoke not only restoring forces in the three principal directions, but also restoring
moments about the three axes. In deriving the dynamical equations for the assembly, a great
simplication results if two assumptions can be justified:
1. When the assembly moves relative to the base, the restoring forces developed in an
individual mounting in the axial and radial directions can be evaluated from the displacements
in these directions of a definite point within the mounting called the "effective point of
attachment," which remains fixed relative to the assembly.

6-6
 2. When the assembly moves relative to the base, the moments (with respect to axes with
origin at its c.g. in the rest position) of the forces developed by the mountings are so large
relative to any couples developed within the mountings themselves that the latter can be
neglected.
It is shown in Appendix 5 of Reference 6-5 that from the ko's and kr's of the individual
mountings, the coordinates of their effective points of attachment, and the orientation of their
axes relative to the fixed XYZ-axes, there can be derived a set of elastic constants of the
type KUV, K,,a, etc., which characterize the elastic properties of the entire set of mountings.
There will actually be a total of 21 distinct values of the K's, 15 of the form Kil where ij,
and 6 of the form Ki.
The dynamical equations which yield the six natural frequencies and six normal modes
of vibration of the system are Newton equations giving either (1) the relation between the rate
of change of rectilinear momentum in a given direction and the forces in that direction; or (2)
the relation between the rate of change of moment of momentum about a given axis and the
moments about that axis. Instead of expressing the dynamical equations in differential form
they are given here in algebraic form, as is commonly done in vibration theory. Thus, on the
assumption of simple harmonic vibrations, terms of the type -, mu 2 are substituted for terms
of the type ml.
The six dynamical equations for the resiliently mounted assembly, when converted to
algebraic form, are:

2
Ku u + v + Kw + K.a + K.f, + K. Y - maUC =0 [6-31

- [6-41]
KU.vu + KV to + KvW w + KVj a+ + Kvy Kvyy mv+o)2 = 0

lKuw u + Kvwv+ + Kww w + liOvaa+ Kwg 0 + KwYyy- mwwo2 -- o [6-5]

2 2
K.,u+Kpv +K.0w+Ka, a+Kp +Kayy -I ao 22+i
+1"o8 +l ry0
2 0= [6-6]

Kuf3 u + KVP v + Kwp U; + Kapl a+ Kpp 0 + Kp3),y- lyfpa/2 + IXY OW2 + iy,,YOJ2 -o [6-7r1

Kuyu + Kw+v + KwYw + K ac+Kp3yp + KIy,'y - 1 C')2 + 1,Z 0t&2 + i, pC02- 0 [6-8]

6-7
where m is the mass of the entire assembly;
u, v, and to are displacements of the e.g. of the assembly in the X-, Y-, and
Z-directions, respectively;
the K's are elastic constants for the entire set of mountings as defined in
Reference 6-5;
,.,
, and is are mass moments of inertia of the assembly about the X-, Y-, and
Z-axes, respectively, which have their origin at the c.g. of the
assembly and are not here restricted to principal axes of inertia;
/z)1 4.. and I., are mass products of inertia with respect to axes X-Y, X-2, and
Y-Z, respectively; and
cO is the circular frequency (2r times the frequency).
Such a set of equations, which reveals the vibratory characteristics of a mass-elastic
system, may be reduced to a single symbolic equation in which matrices are used to represent
entire sets of values in the initial set of equations. The use of matrix algebra in vibration
analysis is discussed in considerable detail in Reference 6-8. It is sufficient to point out
here that, in the present instance, there are six displacement coordinates involved, namely,
u, v, w, at, 1 , and y- These, if put in the form of a column matrix, can be represented by a
single coordinate q. Thus

LYJ
[ 0 Lq
[891

The array of quantities, which in the set of equations [6-4] through [6-8] represent
inertias, in this case yields a matrix, called the inertia matrix, which can also be repre-
sented by a single matrix symbol. Thus

m 0 0 0 0 0
0 m 0 0 0 0
0 0 M 0 0 0
0 o 00 / - lXy -1z z

0 0 0 - y 1'Y -IIyz
0 0 0 -Ixz -Iyz I1

Finally, the array of quantities, which in these equations represent elastic constants, in this
case yields a stiffness matrix which can be represented by a single matrix symbol. Thus

6-8
 KuU K V Kuw KuO KU6 Kuy

KuV KVV KVW K a Kvp Kvy

Kuw KVW Kww Kwa KwP K [K]

[wy

Kua Kva KWO Koo Ka Ky

Kuy Key KWY Ka y K1py KY

Then the entire set of six equations [6-3] through [6-8] can be represented by the
single matrix equation
- o2 [M I iq1 + [K] {i-
q6-12,
0

If Equation [6-12] is expanded by the rules for matrix multiplication (Reference 6-6), the set
of equations [6-3] through [6-8] will be reproduced. It should be noted that Equation [6-12] is
identical in form with the equation for the mass-spring combination having a single degree of
freedom.
As shown in Reference 6-6, the matrix representation of the dynamical equations
applicable to vibratory systems is not restricted to the free vibrations of undamped systems
but is applicable to damped systems and forced vibration as well.
The "dynamical matrix," which is obtained by combining the two terms on the left side
of Equation [6-12], is shown below.

Ku, - m& 2 KuV Kuw KuO KU,8 Ks

i ...

Ku Kvv -ma)2 KVW KV2 Kq3 KUp

Kwu KWV Kw - mO2 KwO KwB Kwy

Kau KVK KO-/o---o 2

Kp 3 + IxI,& 2 KaY + xzo2

K_13. KV K13 KKa +lIxy w2 K1913 Iy

- &) Kfgy + I, 22

K_,_ Kyv KKY 1 KKew+xz

Ka )2 K +yo
+ 02 K IZ

This matrix has the important property of symmetry with respect to the main diagonal;
also both the stiffness matrix and the inertia matrix are individually symmetrical. Since this
matrix yields the determinant of the coefficients of the variables of the six simultaneous
equations, it furnishes the so-called frequency equation since solutions of the simultaneous
equations are only possible for values of wLfor which the determinant vanishes. When the

6-9
determinant is expanded and set equal to zero, the resulting equation is of the sixth degree
--V:
in j2. The positive values of co for which the determinant vanishes will be the natural fre-
quencies of the assembly. Corresponding to each root there will be a set of relative values
of u,, V1 W,a, I3, and y constituting a normal mode pattern. These patterns are found by solv-
ing the simultaneous equations obtained when each of the natural circular frequencies is .ub-
stituted for co in the set of equations [6-3] through [6-8].
When planes of vibrational symmetry exist many of the K's and the mass products of
inertia may be zero. Then it will be found that all equations of the resulting set may be in-
dependent or will reduce to small groups of equations which are independent of the other
groups.
As an illustration of the manner in which the effect of a relatively massive item of
equipment, resiliently mounted in a hull, may be taken into account in the hull vibration
analysis, a special case, chosen for its simplicity, will be considered here. The case se-
lected is the analysis of the vertical vibration of a ship in which there is to be installed a
heavy resiliently mounted assembly whose center of gravity will fall on the longitudinal
centerline of the hull, and of such design that the vertical plane through the longitudinal
axis of the hull will be a plane of vibrational symmetry of the resiliently mounted system.
The assembly now has three normal modes of vibration in this vertical plane whon the
hull is hold fixed. In general, each of these modes will involve the displacements u, w, and
/3, and thus a combination of heaving, pitching, and surging motions is involved. Here, for
consistency with Reference 6-5, the Y-axis is taken as horizontal and not vertical, as in the
treatment of vertical hull vibraticn in Chapter 3.
In the finite difference equations used for finding the vertical normal modes of the hull,
the only equations requiring modification due to the presence of the resiliently mounted as-
sembly are those for the element of the hull of length Ax within which the assembly is mounted.
If the mass of the assembly is designated ms and its mass moment of inertia about the
athwartship axis through its e.g. (the Y-axis in this case) is designated Is, the finite differ-
ence equations (in simplified form) become for this element:

2
AV =O-xzAsj - Kww (o - as) + Kuw us -Kw (3-A,) [6-131

AM -I ,YyAzW2 + VAx + K us - K.8 0) (z- Zs) K- p (j693-f,,) [6-14]

MAx
AP - - Ef [6-151

VAx
f3
Ao P~- Ax- [6-161
KA G

6-10
Here the displacement amplitude of the assembly in the Z-direction is designated sa rather
than w. in conformity with the notation for the displacement amplitude of the hull in that di-
rection; whereas for the displacement amplitude at the assembly in the X-direction the des-
ignation u. is retained. The force and moment equations for the assembly are, respectively:

2
-msZSG =K.. (a - z.) - K Us + K.0 (Pf3,) [6-17]

and

- !8fc3 2 = - K,, u, + Kpw (a - a) + Kpp (9 - 9.) [6-18]

When these equations for the section containing the resiliently mounted assembly are
combined with the usual equations for the remainder of the hull, and the digital process is
carried out as described in Chapter 3, the solution then shows not only how the isolation
mounting affects the natural frequencies of the ship but also how it affects the normal modes
of the hull. It also shows how the resiliently mounted assembly vibrates in each of the nor-
mal modes of the entire system. From such a calculation it can be predicted whether the local
vibration of the resiliently mounted assembly will be excessive when the level of vibration of
the hull in its vicinity is within permissible limits. A more general treatment of the hull as a
beam with sprung masses is given in Reference 6-4. It can readily be seen that the additional
equations involved in the hull calculation to allow for the sprung mass effect of a resiliently
mounted assembly (Equations [6-13] through [6-18]) are directly derivable from the dynamical
matrix shown on page 6-9.
A further development in the isolation mounting of shipboard equipment has been the
use of a compound mounting system. Here the equipment to be isolated is supported by one
set of mountings f.ttached to a cradle. The latter, in turn, is supported in the hull by another
set of isolation mountings. Such a system is shown schematically in Figure 6-4. It has
twelve degrees of freedom, six for each body, and accordingly has twelve natural frequencies
and twelve normal modes of vibration.
A discussion of the calculation of the normal modes and natural frequencies of a com-
pound isolation mounting system is given in Reference 6-7. A treatment of the same problem
by means of the electrical analog, which in this case was derived from the Lagrangian equa-
tions, is given in Reference 6-8.

D. SUMMARY OF EFFECTS OF SPRUNG MASSES ON HULL VIBRATION

Both analytical studies and available experimental data indicate that the local flexi-
bility or the sprung mass effect can cause a considerable modification of the beamlike vibra-
tory response characteristics of a hull. The general effects of lowering hull natural frequen-
cies that are below the local natural frequency and raising hull frequencies that are above

6-11
 '92
Assembly

2 10
eaitKIIh (when ma is hold fixed) tl
KIp (when
ric is held fixed) ,w oes

Sml ,,, ~ 01 ' 1

Ship Hull
Figure 6-4 - Schematic Elevation of a Compound isolation Mounting System

this frequency are of considerable practical importance. it is also significant that the effects
on frequencies decrease the further they are from the local natural frequency itself. In gen-
eral, if the sprung mass is large enough to have a significant effect on the hull vibratory re-
sponse characteristics it will introduce an extra mode. Thus there may be fuund two modes

of the overall system showing the same number of nodes in the displacement pattern of the
hull girder proper. In such cases, the phase relation between the displacement of the sprung
mass and that of the hull in its vicinity will be reversed in these two modes.
Sprung mass effects may produce marked changes in the response of a hull to an ex-
ternal simple harmonic driving force since the sprung mass may act in the role of a vibration
neutralizer or dynamic vibration absorber. The properties of the latter are discussed in
Chapter 9.
Hence, if the designer knows in advance that a large mass is to be resiliently mounted
in the hull, he must take this into account in any vibration analysis that he attempts. The
means for doing this have been indicated in this chapter and are discussed in further detail
in Reference 6-4.

6-12
 REFERENCES

6-1. Kennard, E.H., "Forced Vibrations of Beams and the Effect of Sprung Masses,"
TMB Report 955 (Jul 1955).
6-2. McGoidrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARI.NR," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
6-3. Leibowitz, B.C., "Natural Modes and Frequencias of Vertical Vibration of a Beam
with an Attached Sprung Mass," TMB Report 1215 (Sep 1958).
6-4. Leibowitz, B.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 1817 (May 1961),
6-5. Vane, F.F., "A Guide for the Selection and Application of Resilient Mountings to
Shipboard Equipment," TMB Report 880, Revised Edition (Feb 1958).
6-6. Pipes, L.A., "Applied Mathematics for Engineers and Physicists," McGraw-Hill
Book Co. (1946).
6-7. McGoldrick, R.T., "Calculation of Natural Frequencies and Normal Modes of Vibra-
tion for a Compound Isolation Mounting System," TMB Report 1420 (Jul 1960).
6-8. Ryder, F.L. and Zaid, M., "Direct Electrical Analogs of Double 6-Degree-of-Freedom
Systems," Report on Contract Nonr-2952 (00) (X), Scientific Research Staff, Republic Aviation
Corp., Farmingdale, New York (Feb 1960).

6.18
 CHAPTER 7
PROPELLER-EXCITING FORCES

A. INTRODUCTION

This chapter is devoted chiefly to propeller-exciting forces, but it must be recognized

that propulsion devices other than screw propellers may also generate vibratory forces. It is
only the lack of information on the time-varying forces from paddle wheels and other propulsion
devices of infrequent use in naval architecture that necessarily confines the discussion to
screw propellers at this writing. A change in propulsion system design practice would neces-
sitate detailed study of the forces produced by the new propulsion devices. A further restric-
tion in this chapter is to hydrodynamic forces only. The forces arising from mass unbalance
are discussed in Chapter 10.
The generic term "forces" is used in the title of this chapter to cover any type of hy-
drodynamic excitation of the hull arising from propeller action whether this be a force or a
moment, and it may be recalled here that the significant flexural modes of the hull may be ex-
cited by either a force, a moment, or a combination of both.
The vibratory hydrodynamic forces arising from the operation of propellers may be sep-
arated into pitch-unbalance forces and blade-frequency forces. The former are due to irregu-
larities in the manufacture of the propeller and appear even when the flow into the propeller
is perfectly uniform.
The term "blade-frequency forces" is used here in a broad sense to cover the hydro-
dynamic forces which will exist regardless of the degree of perfection in the manufacture of
the propeller. These forces are usually understood when the term "propeller-exciting forces"
is used and their fundamental frequency is the blade frequency (rpm times the number of blades
per propeller).
It is clearly of great value to the designer to be 4h1P to predict whether, under a pro-
posed design, the propeller-exciting forces will be excessive. However, it is not the absolute
magnitude of these forces that is of prime significance but their magnitude relative to the
mechanical impedance of the hull at the "point" where these forces act. The vibratory re-
sponse characteristics of the hull are discussed in Chapter 4.
The designer would like to be able not only to predict whether the level of hull vibra-
tion will be excessive for a proposed design, but to say what changes should be made in the
design to reduce the level of service vibration to acceptable limits. This chapter, however,
is concerned only with the forces themselves.
The propeller hydrodynamic forces exciting hull vibration operate both directly on the
hull in the vicinity of the propeller as a fluctuating pressure and indirectly through the pro-
peller shaft bearing as a result of lift, drag, and moment on the individual blades, in the
latter case, all three components may vary in time even though the propeller maintains a con-
stant. angular velocity.

1-1
 B. PITCH UNBALANCE FORCES

The first-order hull vibration due to pitch unbalance is usually strikingly in evidence
when a propeller blade has been bent because of running into an obstruction. When this un-
balance is due merely to imperfection in manufacture of the propeller, it is of much smaller
amplitude than in the case of damage but is still of first-order frequency (frequency same as
the shaft rpm). When a perfectly formed screw propeller operates at constant angular velocity
in a uniform axial wake, its polar symmetry requires that it develop constant torque, constant
thrust, zero lateral force variation at its bearing, and zero moment variation about any axis
normal to its shaft axis.
The effect of nonuniformity of the blades may be seen qualitatively by considering a
one-bladed propeller. While the latter will still produce constant thrust and constant torque
if the velocity field is uniform, the bearing will now be subject to simple harmonic vertical
and athwartship forces and moments about both the vertical and horizontal axes. These
forces and moments will be of a frequency which is the same as the shaft rpm (one cycle per
revolution). Moreover, if the velocity field in the propeller race is not uniform there will be
a superimposed variation effect whose fundamental frequency is also of the first order, but
which may have harmonic components depending on the irregularity of the velocity field
(wake pattern). Thus, pitch unbalance gives rise to first-order hydrodynamic forces and
moments at the propeller shaft bearing and to harmonics of the first-order frequency.
Since, under current practice in the manufacture of propellers, forces due to pitch un-
balance are usually within acceptable limits, this phase of hydrodynamic propeller excitation
has not attracted much attention so far. However, when first-order hull vibration is encoun-
tered on the trials of a new class of ship the' naval architect, must always recognize that
either mass unbalance or pitch unbalance may be the culprit.

C. BLADE-FREQUENCY FORCES

1. FIRST PRINCIPLES
Although the term "blade frequency" is now in general use for the forces under dis-
cussion in this section, it must be emphasized that harmonics of this frequency may be im-
pai'ltaa, in rony hull vibration problems. Although it had been under investigation several
years before, the subject of blade-frequency excitation of hulls attracted increased attention
after World War H1; see References 7-1 and 7-2. In spite of this, the information available on
this subject at the present time must be considered relatively scant.
The forces that vary at blade frequency or harmonics of the blade frequency will exist
in spite of extreme precision in the manufacture of a screw propeller and are directly charge-
able to the hydrodynamics of the ship design. While they may be greatly magnified by non-
uniformity of the wake, it is important to recognize that a fluctuating pressure field will exist
forward of the propeller even in a uniform wake. In fact, the hull pressure forces (called the

7-2
"surface forces" in Reference 7-1) and the bearing forces are generally treated quite inde-
pendently. Thus, to predict the blade-frequency exciting forces, the strface forces and the
bearing forces must be combined, taking account of any relative phase shifts.
Although the theoretical treatment of the pressure fields in the vicinity of the propeller
(from which the hull forces must be deduced) is usually based on incompressible potentia.
flow, which implies an infinite'velocity of propagation, it must be recognized that such low
intensity fields cannot actually propagate at a velocitygreater than the velocity of sound.
Since the latter is in the neighborhood of 5000 ft/sec, it is clear that if the pressure field
extended over 100 ft from the propeller there could be appreciable phase shifts between the
forces transmitted through the propeller shaft bearing and the surface forces. However, be-
cause these pressure fields usually are not significant beyond a distance of one-half diameter
forward of the propeller (see Reference 7-8), and blade frequencies above 3000 cpm are rare
for ships of 2000 tons or more, this phase shift is not a serious consideration at present.
The phase relation between the bearing forces and the surface forces, however, in-
volves not only the consideration of the effect of finite velocity of propagation but the effect of
algebraic sign or direction in space as well. Thus, if a one-bladed outboard-turning propeller
is considered, when the blade is in the 3 o'clock position looking from astern, the bearing
force will be directed upward. The pressure field forward of the blade, however, will be neg-
ative (suction) and -AU peak when the blade is in the 12 o'clock position. For instantaneous
propagation, therefore, the bearing force and the vertical component of the surface force for-
ward of the propeller would be 90 dog out of phase (with the bearing force leading) if repre-
sented on a rotating time-vector diagram such as described in Chapter 4. Thus, it is obvious
that the separate determination of bearing and surface blade-frequency components is insuf-
ficient to predict the resultant hull driving force, and their phase relations must be taken into
account.
On multiple screw ships the phase relations between the blade-frequency-force com-
ponents are still further complicated by the fact that the forces from the different propellers
will continually shift in phase unless the propulsion system has a synchronizing device.
This effect and the pitching of the hull in a seaway are the principal causes of the fluctuation
in the amplitude of propeller-excited vibration at the stern of a ship.
When propellers with different numbers of blades are used on the same ship, there will
also be a "beating" at a frequency equal to the difference in the two blade frequencies. Thus,
if four- and five-bladed propellers were operating at 200 rpm, there would be a beating ata
frequency of 200 x (5 - 4) =.200 beats per min or 3.67 beats per see, and the beat frequency
would be of the same order of magnitude as the component frequencies. In such a case, the
usual form of beats, which appears when the difference in the two frequencies is very small
relative to their absolute values, would not be in evidence in the signal from a vibration
pickup. Since the shaft speeds would actually be varying, the hull vibration record would
appear quite irregular.

7-3

-n
 When the propeller blades are uniform, and the inflow to the propellers is uniform in
time (but not in space), the variations in thrust or torque that occur as the propeller rotates
will have a fundamental period equal to the time required for the propeller to rotate through an
angle equal to 860 deg/a, where a is the number of blades per propeller. This is obvious since
at intervalr of this duration the propeller always has the same orientation, or better, the same
"attitude" as seen from any point of the hull. The reciprocal of this interval is called the
blade frequency, and clearly this is equal to the propeller rpm times a. Whereas the blade
frequency is the fundamental frequency for propeller-exciting forces and moments, the number
of blades and the after body arrangement may be such that very large harmonic force and mo-
ment components exist. Hence the term "propeller-exciting forces" includes both blade-
frequency forces and forces whose frequencies are multiples of the blade frequency.

2. ANALYTICAL PREDICTIONS

At the present time there is available to the designer no purely analytical procedure
by which he can start with a given hull and screw propeller design and calculate the blade-
frequency forces and moments that would exist under service conditions. Reference 7-4
indicates the progress that had been made it. the analytical prediction of marine propeller
pressure fields up to 1959. Future progress in such analyses is to be expected, as indicated
by Reference 7-5.
In the problem of surface forces, the first aim in the theoretical attack is to derive the
free-space pressure field due to the operation of the propeller. To make use of theoretically
derived free-space pressure fields, the designer must then be able not only to correct for the
effect on the free-space pressure of the presence of the hull itself as well as its vibratory
motion, but to integrate these pressures over the curved surface of the hull in the vicinity of
the propeller. Finally, he must be able to compute the bearing forces and to combine these
with the hull surface forces, taking account of algebraic signs or phase shifts.. Thus, it is
not surprising that a purely analytical prediction of the blade-frequency exciting forces has
not as yet been achieved.
At the present stage of the art, such analytical predictions of blade-frequency exciting
forces as can be made are essentially comparative; see Reference 7-6. To make a prediction
the designer must have at hand force data on a previous ship of the same general type. He
can then make estimates of the relative magnitude of the surface forces for the new ship by
comparison of the axial and radial tip clearances from curves such as given in Figures 7-1
and 7-2 (from Reference 7-1j). The parameters involved in this process (also from Reference
7-3) are:

Pressure Coefficient- KP = _

pn 2d 2

7-4
 T
Thrust Coefficient: KT T
2 4
pn d
where p is the free-space oscillating pressure at blade frequency (single amplitude)
in psf,
n is the number of revolutions per sec,
p is the mass density of water in lb-sec 2/ft 4 ,
d is the propeller diameter in ft, and
T is the propeller thrust in lb.

0.20

Propeller 2714 0.18 -

0,0 -- 3 Blades Propeller 2714

y/d - 0 j 3 Blades
0.16 --
a,09 -a8y/d - 0
0 x/d - +Q153
0.0.04

0.07 o0.12

4.a 1 .,2
S0.00 o_ 3 .0 _ _o_ __ _

J .
.0 O46
0.0 'A 0.5 0.7-----
r _

J.0.5 0,04",
0.04 J. OL

-0.3~ ~ ~ ~~A
-02-.-0.102
l . 0.3 0.
0.0 0.0 0.0 0I 2b-

-0,3 -a,2 1-0. o 0i 0.2! 0,.3 0 0.05t 0.10 0.15 00: 25 60t

Figure 7-1 - Variation of Pressure Amplitude Figure 7-2 - Variation of Pressure Amplitude
with Axial Distance from Propeller with Radial Tip Clearance t (in a Plane
Ahead of the Propeller Plane)

7=5

-I
The coordinate system applicable to Figure 7.1 and 7-2 is indicated in Figure 7-8.

BY

Figure 7-3 - Rectangular Coordinate System

Applicable to the Pressure Amplitude
Data Presented in Figures
-

Aft Forward 7-1 and 7-2

Figures 7-1 and 7-2 were. plotted for a three-bladed propeller. The effect on the pressure
amplitude in going to increased numbers of blades may be seen from Figure 7-4 (also taken
from Reference 7-3) in which J is the propeller advance ratio J inflow veIocity.

0,35 IEstimated for Propellers Similar to

3192 for J - 0.7
0.30L Sk - -0 -'- Measured Values

0.25... , Estimated from I-Bladed Propeller

\ -- ,

0.20o - -

stimated from 2-Bladed Proeller \\ tim ted from 4-Bladed Propeller

oic I ' \'" -4 -.

1 2 3 4 5 6 7 8 9 to
Number of Riadeut Z
Figure 7-4 - Effect or Number of Propeller Blades on Free-Space Pressure Amplitude
(- 0O; - 0.IS)
d d

The "analytical" prediction of the bearing forces at present requires the availability
of a wake survey. There will be no thrust variation, torque variation, or bearing-force varia-
tion if the wake is uniform in the area of the propeller race and the propeller blades are uniform.

7-6
Thus, the process is similar to that of estimating the thrust and torque variation, the difference
being that the lift and drag on the individual blades are resolved into components normal to the
shaft axis, Details of the torque and thrust calculation are given in Reference 7-?. Although
the bearing forces will be zero if the torque and thrust variation are zero, it does not follow
that a wake pattern that gives high blade-frequency thrust variation will necessarily give
large bearing-force variations. This is because the components of force on the individual
blades may be additive as far as thrust is concerned while canceling as far as vertical or
horizontal forces in the plane of the propeller are concerned.
It is clear that, regardless of the complexity of the wake pattern, as long as it is con-
stant in time and the blades are uniform in pitch, such an analysis requires advancing the
propeller through an angle of only 3860 deg/a. The time taken for this angular displacement
is the fundamental period of the bearing-force variation.
Thus, for a four-bladed propeller the estimate can be started with one blade in the
12 o'clock position. The lift and drag on each blade for this position are then estimated
from the wake survey just as the thrust variation is estimated. Instead of combining the
blade forces to give the net thrust, however, they are now combined to give the net vertical
force (or horizontal force as the case may be). The propeller is then advanced a few degrees;
this process is repeated until the propeller (here assumed four-bladed) which was originally
in the 12 o'clock position has advanced to the 8 o'clock position. If the tabulated values of
net vertical force are then plotted against time, a smooth curve through the plotted points
gives one cycle of the bearing-force variation. The time interval will be the reciprocal of
the blade frequency and in that period the four-bladed propeller will have rotated 90 dog.
The total hull driving force is then estimated by combining the bearing and surface
forces. Although the estimate of their phase relation is uncertain, the maximum condition
will obviously be obtained if they are assumed in phase.

3. MODEL PREDICTIONS
In view of the difficulties involved in the analytical prediction of blade-frequency
exciting forces, it was inevitable that the possibilities of model predictions would be ex-
plored. The problem, however, also involves numerous difficulties which up to now have
not been completely resolved.
In the United States the attempt to determine blade-frequency exciting forces from
model tests was initiated by Professor F.M. Lewis under the auspices of the Society of
Naval Architects and Marine Engineers. The measurements were made at the U.S. Experi-
mental Model Basin and the results are discussed in References 7-8 and 7-9.
The basic idea of the method was to measure the overall or effective driving force
acting on a self-propelled model by nullification of the model vibration by means of a me-
chanical vibration generator installed in the model and geared to the propeller drive shaft.
Both the unbp1ance of the eccentrics of the vibration generator and their phase relative to
propeller could be varied while the model was underway.

7-7
 In experiments with this apparatus at the U.S. Experimental Model Basin, Professor
Lewis was able not only to determine the effective blade-frequency vertical driving force but

also to demonstrate that a large part of the effective force was due to the pressure pulsations
acting on the hull itself in the vicinity of the propeller. This was shown by installing in one
model a long propeller shaft which placed the propeller far enough astern for the pressure field
of this propeller to act on another model without a propeller and thus not subject to bearing
forces.
The model work on propeller-exciting forces was reactivated at the David Taylor Model
Basin in the post-World War II period, and the Society of Naval Architects and Marine Engineers
established two research panels in the field of hull vibration, one under its Hydrodynamics
Committee and one under its Hull Structure Committee. A comprehensive paper giving the re-
sults achieved up to that time was presented to the Society by F.M. Lewis and A.J. Tachmindji
in 1954.7" Although this paper reported marked progress in the model determination of propeller-
exciting forces, it emphasized the difficulties due to resonance effects in the wooden models
themselves in spite of the efforts made to reinforce them. Such a model with the test gear
installed in the stern is shown in Figure 7-5.

Figure 7-5 - Reinforced Wooden Model with Test Gear for the Determination
of Propeller-Exciting Forces
(From Reference 7-1)

In Reference 7-1 it was pointed out that, although the force system was known to be
considerably more complicatAd, for prantiftl purposes the model determination of blade-

7-8
frequency excitation was reduced to the prediction of a net vertical force, a net horizontal
athwartships force, and a couple about the longitudinal axis of the hull. The horizontal and
vertical forces are assumed to act at such locations as to cause no twisting moment on the
hull, whereas the couple does apply such a moment. It is to be noted that, although there is
no plane of symmetry for horizontal excitation, there can be a plane of symmetry for vertical
excitation on multiple-screw ships if all propellers are rotating at the same speed and in
phase. In the case of a single-screw ship, there will be no plane of symmetry for either
vertical, or horizontal excitation.
The latest development in the model determination of propeller-exciting forces up to
this time is described in Reference 7-10. In the technique described in this reference, the
difficulty with model resonance was circumvented by resorting to the use of a flexibly sup-
ported stern section designed to be nonresonant in the range of blade frequencies to be in-
vestigated. This is illustrated in Figure 7-6.

Figure 7-6 - Model for Propeller-Exciting-Force Determination Having

Flexibly Suspended Stern Eection

In the previous model technique a null method of fo; -,e determination was used. Thus
the propeller forces were neutralized by forces of known magnitude produced by a vibration
generator. The later technique did not employ the null method but was based on producing,
by means of a vibration generator, the same stern amplitude as produced by the propeller-
exciting forces.

7-9
 As in model testing for powering, the scaling laws are very important in determining
propeller-exciting forces by means of model tests. In such model tests the model shaft speed
is the same as in the ordinary self-propulsion test; that is, it is in accordance with the Froude
scaling relation
ilength of ship

model rpm = ship rpm length of moel

If the boundary-layer distribution is assumed to be the same between ship and model, the
experimentally determined model forces are then stepped up to full scale by multiplying by
the factor:

displacement of ship
displacement of model

The results of a series of model tests with the flexible stern technique are given in
Reference 7-11. As pointed out in this reference, in spite of the improvement in the model
technique, unknown scale effects make the model determination of propeller-exciting forces
still uncertain at this time.

D. EXPERIMENTAL FORCE DATA AVAILABLE

The experimental data on propeller-exciting forces at present are extremely meager in
spite of the effort made to obtain such information. Such data as now exist indicate blade-
frequency forces that are fairly large. To give a rough criterion of the order of magnitude of
these forces, it may be pointed out that for single-screw cargo ships the vertical and hori-
zontal forces have a single amplitude of the order of 10 percent of the steady thrust at the
normal operating speed.
In'ihe case of the Maritime C 4-Class dry cargo ship, both model and full-scale ex-
periments were conducted, as pointed out in Reference 7-1. The full-scale thrust of these
ships is 170,000 lb. Reference 7-1 gives the following values as determined from the model
(all in single amplitude):
Net vertical force on hull 6 percent of mean thrust
Net athwartship force on hull 16 percent of mean thrust
Net couple on hull 70 percent of mean shaft torque
Thrust variation in shaft 5 to 8 percent of mean thrust
Shaft horsepower 22,000

These ships have the following principal characteristics:

7-10
 L 525 ft
B 76 ft
D 44 ft 6 in.
H 31 ft6 in.
Top speed 22 knots at 102 rpm
Propeller blades 4
Single screw

A vibration-generator survey was conducted on SS GOPHER MARINER in the investi-

gation of the vibratory response characteristics of the hull .7-12 It was found during this
survey that a vertical force of about 7 percent of the mean thrust would produce the amplitude
in the stern observed at blade frequency under operating conditions. This would be about
12,000 lb.
In 1957 the David Taylor Model BasiA made both an underway-vibration survey and a
vibration-generator survey on a twin-screw naval destroyer, USS DECATUR (DD 936). These
surveys are discussed in detail in Reference 7-13. In addition to operating the vibration gen-
erator at the critical frequencies of the hull, it was run up into the operating blade-frequency
range (although not up to the maximum operating blade frequency). From the latter tests,
impedance-type expressions were derived for the relation between driving force and stern
amplitude at blade frequencies above the range of significant hull mode frequencies. As
given in Reference 7-13, the constants to be used in the impedance-type formulas were:

6 2
10- lb-mtn
01V =3.5 x mil-ton
2
lb-min
A-2.0 X10- ramil-ton

2
0.46
--0.46rad-ton-sec
'T aT lb-min 2

For the vertical and athwartship vibration these constants are to be used in the approximate
equation

Po
Yo 2
a Af
A=

where Y0 is the single amplitude at the after perpendicular in mils,

P0 is the single amplitude of the driving force in lb,
A is the ship's displacement in long tons, and
f is the frequency in cpm.

7-11
For torsional vibration the constant a T is to be used in the approximate equation

TO

where To is the single amplitude of the blade-frequency exciting couple in lb-ft, and I is as
defined for Equation [D-6] of Appendix D.
The following values of propeller-exciting forces were deduced for DD 986 from the
observed amplitudes in the stern in the underway-vibration survey in conjunction with the
foregoing data.

Net vertical blade-frequency force, at 310 rpm = 82,000 lb

Net athwartships horizortal foroe, at 310 rpm = 38,000 lb

The DD 981-Class destroyer has the following principal characteristics:

L 407 ft
B 45 ft
D 25 ft
Full load displacement 3800 tons
Thrust per shaft 220,000 lb
Blades per propeller 4
Propeller diameter 18 ft 3 in.
Twin screws
Twin rudders

From the initial model experiments made at the U.S. Experiment.al Model Basin in the
early 1930's, Professor F.M. LewisT"S estimated the net vertical blade-frequency Force for
SS PRESIDENT HOOVER to be 24,000 lb single amplitude. This was 12% percent of the
total thrust. This ship has the following principal characteristics:

L 630 ft
B 81ft
D 30 ft 3 in.
Displacement 29,000 tons
Bladeni per propeller 3
Twin Screws

The ship was provided with propeller shaft bossings.

7-12
 In Reference 7-14 the author suggested as a round number that the bearing forces
account for 25 percent of the total blade-frequency exciting force acting on the hull, and
that the remainder was due to the surface forces.
An example of reduction of blade-frequency vibration by the use of a fin to reduce the
nonuniformity of flow to the propeller is discussed in Reference 7-15.

REFERENCES
7- 1. Lewis, F.M. and Tachmindji, A., "Propeller Forces Exciting Hull Vibration,"
Trans. SNAME, Vol. 62 (1954).
7- 2. Breslin, J.P., "The Pressure Field Near a Ship Propeller," JSR, Vol. 1, No. 4
(Mar 1958).
7- 3. Tachmindji, A.J. and McGoldrick, R.T., "Note on Propeller-Excited Hull
Vibrations," JSR, Vol. 8, No. 1 (Jun 1959).
7- 4. Breslin, J.P. and Tsakonas, S., "Marine Propeller Pressure Field Including Effects
of Loading and Thickness," Trans. SNAME, Vol. 67 (1959).
7- 5. Reed, F.E. and Bradshaw, R.T., "The Distribution of Exciting Forces Generated
by Propellers," CONESCO Report F-101-2, Contract Nobs 77150 (Jun 1960).
7- 6. Breslin, J.P., "Ship Vibration. Part I - Propeller-Generated Excitations," Appl
Mech Rev, Vol. 18, No. 7 (Jul 1960).
7- 7. Yeh, H.Y., "Thrust and Torque Fluctuations for APA 249, TMB Model 4414,"
TMB Report 1364 (Jan 1960).
7- 8. Lewis, F.M., "Propeller Vibration," Trans. SNAME, Vol. 43 (1935).
7- 9. Lewis, F.M., "Propeller Vibration," Trans. SNAME, Vol. 44 (1936).
7-10. Pien, P.C. and Ficken, N.L., "The Measurement of Propeller Induced Vibratory
Forces on Scale Ship Models," Paper presented at ATTC Conference (Sep 1959).
7-11. Stuntz, G.R., et al., "Series 60 - The Effect of Variations in Afterbody Shape
upon Resistance, Power, Wake Distribution, and Propeller-Excited Vibratory Forces,"
Paper prepared for Annual Meeting of SNAME (1960).
7-12, McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARiNER," Trans. SNAME, Vol. 68 (1955). Also TMB Report 1060 (Jul 1956).
7-13. Paladino, A.R,, "Vibration Survey on USS DECATUR (DD 936)," TMB Report
1271 (Mar 1960),
7-14. Babaev, N.N., "Contemporary Status of Development of the Problem of Ship Vibra-
tion," TMB Trans. 291 (Jul 1959) from Russian paper in Sudosrwoenie, No. 3 (161) (Mar 1956).
7-15. Baier, L.A. and Ormondroyd, J., "Vibration at the Stern of Single-Screw Vessels,"
Trans. SNAME, Vol. 60 (1952).

7-13
 CHAPTER 8
DAMPING OF HULL VIBRATION

A. INTRODUCTION
Damping plays a very important role in ship vibration, as in most other areas of me-
chanical vibration, and it is unfortunate that so little is known about the actual damping
process at this time. It is not generally appreciated that the power required to maintain a
quite perceptible vibration in a structure as large as a ship is relatively minute. It is also
generally overlooked that an undamped mass-elastic system can be maintained in forced vi-
bration at a nonresonant frequency with zero net power input to the system. Even under
damped resonant conditions the power required to maintain noticeable hull vibration is amaz-
ingly small. Thus a ship of 5000-ton displacement can be maintained in vibration in its fun-
damental vertical mode with single amplitudes at bow and stern of the order of 10 mils with a
mechanical power input of only 100 watts, as follows from the driving force and driving point
amplitude data given in the first line of Table 5 of Reference 8-1.
While the designer will naturally aim to avoid resonance involving any of the signif-
icant hull modes with either the blade frequency, the shaft frequency, or the running rpm of
any maior piece of machinery, cases will inevitably arise in which this is not feasible. It
may well be necessary in many cases to operate a ship at a shaft speed at which the blade
frequency coincides with one of the natural frequencies of the hull. Furthermore, it must not
be overlooked that, even when the operating blade frequencies are well clear of the range of
significant hull frequencies, it will still be necessary to pass through critical speeds in
coming up to the operating speed.
As shown in Chapter 4, the resonant amplitudes of the hull depend on the magnitude
of the driving force, the location of its point of application relative to nodal points in the
mode in question, and the damping itself. Clearly, the importance of being able to predict
the damping values depends on the other factors determining the resonant amplitude. Thus,
if there is assurance that the driving forces will be negligible, the need for the damping pre-
dirtion is greatly diminished. If all the critical speeds with respect to hull vibration are
known to fall well below the operating speed range of the ship, it may not be necessary to
estimate the hull damping. In general, however, it is desirable to do so.
The difficulty in estimating damping values for the entire hull is due to the present
scant knowledge of the actual damping mechanism. To cite obvious sources of damping is
far easier than to assess their relative importance. Thus there is an immediate tendency to
assume, since the hull cannot vibrate without imparting motion to the surrounding water, that
this is an important source of hull damping. There is ample evidence, however, that in the
range of frequencies of significant or beamlike hull modes, the damping action of the water
is extremely small. Thus the decay rates for free vibration ol ships are no greater than for
steel structures in air.

8-1
 Aside from the surrounding water there are two other obvious sources of hull damping;
one is the internal friction in the hull structural material, the other is friction between slip-
ping or sliding surfaces within the hull, associated with equipment or cargo.
The experiments on GOPHER MARINER 8-2 show that at the levels of vibration excit-
ed by propeller action, the most important source of damping appears to be the cargo friction,
if by "cargo" there is included here everything movable inside the hull proper. The damping
source next in importance seems to be the hysteresis in the hull itself if the structure is weld-
ed. With a riveted construction, the working of the joints can cause a dissipation of energy
which is chargeable to the hull although not to the material. The least important source of

damping appears to be the water. This statement applies to practical conditions under which
the amplitudes are too small to set up appreciable surface waves, and to frequencies at which
acoustics radiation is not an important factor. Here it should be noted that acoustic power
letels are ordinarily extremely low as contrasted with the level of mechanical power required
to vibrate a hull. Present evidence (see References 8-1 and 8-2) indicates that hull damping
is actually dependent on both amplitude and frequency.

As in the case of the hull itself, "damping also determines the magnitude of resonant
vibration of local structures. Here both cargo and water damping effects are, in general, not
involved, and the damping source must be attributed to the material of which the local struc-
ture is fabricated, the type of joints. or the type of support which determines the transmission
of energy to other parts of the ship.
In harmony with the rational beam theory of hull vibration presented in Chapters 3 and
4, the treatment of damping in this book is also a rational or semiempirical treatment. Thus,
in the case of flexural vibration, the damping coefficient c is used which represents a damping
force per unit velocity, per unit length. This coefficient is further restricted to either of two
types: the Rayleigh type which is viscous and proportional to mass, so that - is constant
(where [Lis the mass per unit length of the hull including the allowance for added mass of
the surrounding water); or the type increasing with frequency so that - is constant, where
t is the circular frequency.
In this scheme of treating hull damping, the c values to be used in actual calculations
cannot be determined analytically from given hull design data but must be based on experi-
mentally determined values obtained on other ships.
The reader will find in the technical literature numerous treatments of beam vibration
with other types of damping; see, for example, References 8-3 and 8-4. In the belief that the
principal source of damping of ship vibration is the cargo, these methods are not evaluated
here. It is pointed out, however, that while the use of equivalent viscous damping constants
based on energy dissipation is a makeshift expedient, the viscous damping constant that is
almost universally used in the treatment of lumped vibratory systems in standard textbooks on
mechanical vibration also involves an idealization of the actual damping process.

8-2
 B. ANALYTICAL TREATMENT OF HULL DAMPING

The most widely used assumption in the analytical treatment of the damping of a me-
chanical system in vibration is that it is of the viscous type, as indicated in the previous
section. As applied to the elementary system of one degree of freedom, this is the type of
damping produced by a frictional force proportional to the velocity and having a direction
opposite to that velocity. This gives for the free vibrations the familiar differential equation,

my + 0i + ki= 0 [8-1]

where o is the viscous damping constant,

m is the mass, and
k is the spring constant.
In spite of the fact that mechanical damping is rarely of the true viscous type, an
"equivalent viscous" constant is widely used because the solutions of the resulting linear
differential equations are well known. The equivalent viscous constant is based on energy
dissipation per cycle. If this is designated W, then

W
7T~y2

where Y is the single amplitude and ra is the circular frequency. The viscous damping
concept is also commonly retained in establishing damping constants from the logarithmic
decrements deduced from observations of decaying free vibrations. Thus, in the elementary
system of one degree of freedom, the critical viscous damping constant is given by the equation

cc= 2m V/ [8-3]

and the logarithmic decrement is

cc
Cc [8-41

For small damping

8- 2- [8-51
cc

8-3
 A common criterion of the degree of damping is the resonance magnification factor.
This is frequently designated by the symbol Q, widely used in electrical circuit theory as an
index of dissipation for inductares. The lower the dissipation in the coil, the higher is its
Q. For viscous damping

Q - [8-6]
2-
cc

and for low damping

IT

Q 8-7]

In the case of the hull, the complexity and uncertainty regarding the actual damping
processes have forced an extension of these concepts. By assuming that the damping of the
flexural vibr-ation of the hull can be represented by a distributed viscous damping constant
proportional to the mass per unit length (including the added mass of surrounding water), the
ship can be treated as a "Rayleigh system," at least in dealing with vibration in its sig-
nificant flexural modes. The Rayleigh-type of damping is then of the type c/si = constant,
where c is the equivalent damping force per unit velocity per unit length (axial), and y is the
mass per unit length including Lhe added mass of water.
Under such assumptions, the free and forced vibrations of the hull may be treated in
terms of normal mode responses. This effects great simplification. In this procedure the
ith flexu-al mode of the hull is reduced to an effective system of one degree of freedom re-
ferred to a specific driving point d. According to the equations given on pages 4-9 to 4-11,
the steady-state amplitude is

P
i 2 [8-8]
Kdi - Mdi + J Cdi cO

This gives not only the magnitude of the displacement amplitude in the ith mode but also the
phase of its rotating vector in relation to that of the driving force of circular frequency wo.
Experience has indicated that the damping is actually dependent on both amplitude
and frequency. For frequency dependence the indication is that the relation

C
- = constant
JAWJ

is closer to reality than the relation

- constant

8-4
 In calculating steady-state forced vibration by the digital method discussed in Chap-
ter 4, it is possible to use values of a satisfying either relation. All that is required in the
former case is to adjust the value of c for each value of w for which the calculation is made.
However, this does not permit finding the response to an arbitrary excitation by the normal
mode method or by the digital method discussed in Chapter 5. Thus, where a method of cal-
culation requires the use of a true Rayleigh damping coefficient, its value must be based on
a mean value of the frequencies involved; but where the calculation requires only the response
of the hull at a single frequency, the damping coefficient can be selected according to that
frequency.
In the case of local structures, the same general principles used in treating the damp-
ing of the hull may be applied, but the determination of normal modes and the subsequent re-
duction to effective systems for each normal mode referred to a specific driving point cannot,
in general, be based on a beam-type analysis. Where the response in the fundamental mode is
of prime concern and where a reasonable guess can be made as to the fundamental normal
mode shape, the frequency may be estimated by the general Rayleigh method if the potential
energy can be evaluated for a deformation in this pattern. The Rayleigh method of finding
natural frequencies of systems is discussed in References 8-6 and 8-7 and requires the eval-
uation of both kinetic and potential energies. Thus, for a section of plating simply supported
at the edges but with stiffeners in both the fore and aft and transverse directions, the bending
energy may be evaluated in terms of the curvatures in the two principal directions and the
rigidity factors for the stiffened plate. The effective mass values are then derived from the
kinetic energy on the basis of the same concepts as applied to the hull girder.

C. EXPERIMENTAL METHODS OF DETERMINING DAMPING

Two standard methods of determining the damping of a vibratory system are (1) to ex-
cite the system by an impulse and to measure the rate of decay of the resulting free vibrations,
and (2) to measure the resonant magnification in forced steady-state vibration. Both methods
have been applied to the entire hull and also to local hull structures.
When the hull is excited by a vertical impulse at the bow, the predominant response is
usually in the 2-node vertical flexural mode, This impulse is most conveniently applied by
releasing an anchor and arresting it after a fall of a few feet. On one occasion it was found
possible to measure the fundamental vertical frequency of a large naval vessel during calis-
thenics of the crew on the forward main deck. A horizontal impulse may be applied at the
bow less conveniently by a bump applied by a tug boat. It is also possible to excite tran-
sient vibrations in the horizontal flexural modes by rudder maneuvering while underway.
The records from such tests usually show an initial complex vibration with high fre-
quency components which shortly settles down to a train of damped sine waves of constant
frequency. The logarithmic decrement is determined from the latter part of the record by
measuring two peak displacements q cycles apart and using the relation

8-5
 1 A
8i- -log [8- 91
q 7 -n+q

As applied to local structures, the impulse method of determining damping is similar,

but in this case it is often difficult to produce a decaying vibration in a single mode. Blows
with a heavy timber comprise the most common method of excitation.
The rational theory permits the evaluation of the damping constant to be used in the
calculation of forced vibration of ships directly from the experimentally determined logarithmic
decrement. Thus, suppose a calculation is to be made by the digital process outlined in Chap-
ter 4. Here the Rayleigh damping constant is retained and this means that c/p is treated as
constant for all points along the hull and at all driving frequencies. Let Mdi be the effective
mass at any arbitrary point for the mode of vibration in which the logarithmic decrement was
observed (here called the ith mode). Then the effective Rayleigh damping constant for this
mode and at this point of the hull is given by the equation

ITCdi
- [8-10]

The damping constants c used in the digital calculation are then evaluated from the relation

- - --
O di [8-111
p M/dI

Cdi
If hull damping were truly of the Rayleigh type, - would be independent of the mode
Mdi
and it would follow from Equation [8-10] that the logarithmic decrement would vary inversely
with the frequency of the mode. Since experience has shown it more feasible to assume that
the logarithmic decrement remains constant, it is expedient to assume

Cdi
- = constant [8-121
Mdi Wi

in any calculations dealing with response in a single mode.

Naturally, if experimental decrement values ake available for a previous ship of gen-
erally similar design these are the best to use, but as pointed out in Reference 8-12, an aver-
Cdi
age value of - obtained from experiments on a variety of ships is 0.08.
Mdi0 i
In deriving hull damping values from vibration generator tests, use can be made of the
fact that at resonance the mechanical impedance (defined in Chapter 4) in the mode in ques-
tion is solely the damping impedance CdOei. Hence, from the known exciting force of the

8-6
vibration generator and the measured resonant amplitude at the driving point, CJ, can be com-
puted from the equation

e 18.18]
' di Yi

D. AVAILABLE DATA ON HULL DAMPING

Although experimental data on the damping of ship vibration are still relatively scarce,
the designer is fortunate that such information is being accumulated at an accelerating pace.
Obviously, in making estimates for a particular design the naval architect should seek data
on ships of the same general type. Some typical data are presented in Tables 8-1, 8-2, and
8-3. Further data of this type can be found in Reference 8-5.
In Reference 8-8, the logarithmic decrement of an aircraft carrier in whipping following
slamming in a rough sea was reported as 0.037.

E. DAMPING ACTION OF LIFTING SURFACES

In general, the damping of a hull, just as its inertia, will not change with the ship's
forward speed. In the case of inertia, this is because the flow of water associated with the
added mass effect discussed in Chapter 2 is noncirculatory. In hull damping, it is because
the contribution of the water to the total damping effect is quite small in any case.
There is, however, a hydrodynamic damping effect which does depend on the ship's
forward speed and which cannot be considered negligible. This effect, discussed in more
detail in Chapter 14, involves the lifting surfaces such as rudders and submarine diving planes.
In the present chapter the effect is discussed only in the simplest possible terms.
The function of the lifting surface, of course, is to produce a lift force derived from the
flow whose moment will cause the rigid body rotation of the ship desired for a maneuver. This
lift force is proportional to a "lift coefficient," the angle of attack, and the square of the
relative velocity between the lifting surface and the water. In these simple terms the lift
force is given by the relation

F1 = AS2 O [8-14]

where F, is the lift force,

A is the lift coefficient,
S is the relative velocity, and
0 is the angle of attack.
As shown in Chapter 14, if the control surface acquires a componenL of velocity normal to the
direction of the velocity 8, there results a change in F, due to the change in apparent angle

8-7
of attack. A vibratory motion of the axis of the control surface, due to a vibration of the hull
as a beam, will therefore cause a variation in F, of the same frequency as that of the hull vi-
bration. If the entire hull is reduced to an effective M, K, 0 system of one degree of freedom
for the purpose of analysis, the governing equation for vibratory motion then becomes

M + C+ KY = F," [8-15]

where Ff"is the variation from the steady lift F1 due to any velocity of the lifting surface in
the Y-direction.
Under the simplifying assumptions made here and discussed further in Chapter 14, it
turns out that

=- 1
[YL.

Hence Equation [8-15] becomes

My+ (C +AS) i + KY= 0 [8-17]

This indicates that, when the vibration of the hull is accompanied onlyby a vibratory motion
in translation of the control surface (0 remainingconstant), the forward velocity of the ship
causes a damping action in addition to the damping that would otherwise exist. The latter
is represented by C in Equation [8-17].
It is shown in Reference 8-9 that the hydrofoil damping action of rudders at high ship
speed can reach the same order of magnitude as the ordinary hull damping action. For rud-
ders this, obviously, applies only to horizontal hull vibration. However, in the case of sub-
marines the diving planes can produce a similar damping action in vertical hull vibration.
When angular oscillations of the lifting surface are also present, the situation is radically
altered, leading to the possibility of flutter as shown in Chapter 14.

8-8
 sit~
TI
A
JS 1

CD-
AlE~

F3

CZ,~..

OD~o - ,

~4

F3 e

* IM

r tc ' ~ 5
 TABLE 8-2

Damping Factors Derived from Vibration Generator Tests on Ships Listed in Table 8-1
(Vertical Modes Only) 4i

Driving Point
Ship Mode 0 all, 0/11w Driving Force Single
hed/sec I/soc tons Amplitude
ft
NIAGARA 1st 11.5 0.49 0.043 0.51 0.0011
2nd 20.9 0.41 0.019 1.68 0.0021
3rd 30.5 0.83 0.027 3.57 0.0014
4th 37.1 2.8 0.067 5.27 0.0058
5th 46.8 2.20 0.047 8.44 0.0049
CHARLES R. WARE lst 8.2 0.11 0.021 0.30 0.0050
2nd 12.2 0.17 0.010 1.32 0.0070
3rd 27.3 0.31 0.014 3.32 0.003i
4th 37.6 1.3 0.035 6.29 0.0092
E.J, KULAS 5th 29.8 0.80 0.027 2.77 0.0006

C.A. PAUL Ist 4.71 0.029 0.006 0.16 0.0079

2nd 11.1 0.114 0.010 0.76 0.0064
PERE MARQUETTE 21 1st 11.7 0.168 0.014 0.89 0.0052

NORTHAMPTON 2nd 13.9 0.298 0.021 1.21 0.0008

3rd 21.4 0.512 0.024 2.86 0.0007
4th 30.2 0.722 0.024 5.71 0.0004
5th 37.6 1.33 0.035 8.84 0.0004
6th 45.8 2.55 0.056 12.95 0.0001
7th 52.4 7.80 0.149 17.19 0,D002
STATENISLAND eIt 29.3 0.976 0.033 2.81 0.0038

(av0) 0.034

TABLE 8-3
Expcrimontal Values of Logarithmic Decrements for Fundamental
Vortical Mode of Ships
Name of Type of Test Test Logarithmic
Ship Ship Displacement L B D Draft Decrement
tons ft-in ft-in fl-in ft-in
NIAGARA Transport 5,500 400 58 37 12.1 0.079
OCEAN VULCAN Freighter 13,750 416 56-11 37-4 25-11 0.053
HAMILTON aty 1,380 310 31 20.9 10-8 0.023

8-10
 REFERENCES

8-1. McGoldrick, R.T., "Comparison between Theoretically and Experimentally Determined

Natural Frequencies and Modes of Vibration of Ships," TMB Report 906 (Aug 1954).
8-2. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
8-3. Kumai, T., "Damping Factors in the Higher Modes of Ship Vibrations," European
Shipbldg, Vol. VII, No, 1 (1958).
8-4. Sezawa, K., "The Effect of the Distribution of Damping Resistance in a Ship on Her
Vibration," J Soc Nay Arch Japan, Vol. 66 (Jun 1940).
8-5. Vane, F.F., "Shock and Vibration in Ships," Handbook of Shock and Vibration Con-
trol, Harris, C.M. mad Credo, C.E., Editors, McGraw-Hill Book Co. (to be published in 1961).
8-6. Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York (1956).
8-7. Timoshenko, S., "Vibration Problems in Engineering," D. Van Nostrand Co.,
New York (1955).
8-8. Jasper, N.H. and Birmingham, J.T., "Strains and Motions of USS ESSEX (CVA 9)
during Storms Near Cape Horn," TMB Report 1216 (Aug 1958).
8-9. McGoldrick, R.T., "Rudder-Excited Hull Vibration on USS FORREST SHERMAN
(DD 931) - A Problem in Hydroelasticity," Trans. SNAME (1959). Also TMB Report 1431
(Jun 1960).

8-11
 CHAPTER 9
ANTIVIBRATION DEVICES

A. INTRODUCTION
The technical literature reports many developments in the field of antirolling devices
for ships, such as bilge keels, antirolling tanks, gyrostabilizers, and activated fin stabilizers.
Aside from the fact that the rolling frequencies of ships are of the order of 1/10 or less of the
lowest frequency normally associated with hull vibration, these devices perform the same
basic function as antivibration devices. In one way or another they set up moments opposing
the rolling motion of the ship. It may therefore seem surprising that relatively little attention
has been given to the development of antivibration devices for hulls. This is true in spite of
the fact that antivibration devices in the form of pendulum dampers have been used extensively
in large internal combustion engines to suppress torsional vibration in the crankshaft system.
The presumption in seeking any antivibration device is that the source of disturbance
cannot be eliminated. In the case of hull vibration (as shown in Chapter 7), while first-order
exciting forces can be reduced by improved methods of balancing propellers, shafting, and
machinery, or by closer tolerances for machining of propoller blades, the bl.de-frequency ex-
citing forces cannot be reduced without changing the number of blades per propeller, or al-
tering the stern configuratin. Thus, a practical antivibration device for hulls could find application
under the following situations: (1) when hull vibration develops unexpectedly on the initial
trials of a new class of ship; (2) when the design study indicates that an excessive lovel of
vibration will exist, but there are overriding advantages in the particular design adopted which
warrant its retention; and (3) when an unusual type of propulsion system known to produce
large vibratory exciting forces is required for special reasons.
In this chapter a brief discussion is given of certain antivibration devices that have
actually been used to a limited extent for the purpose of reducing the level of service vibra..
tion of ships. These devices are the tuned vibration neutralizer, often spoken of in the lit-
erature as the "dynamic vibration absorber"; the adjustable rotating eccentric; shaft synchro-
nizing devices; and flexible materials used in the vicinity of propellers. Both the tuned
vibration neutralizer and the adjustable rotating eccentric set up forces equal and opposite to
the external forces exciting the hull vibration. The synchronizing device, which is applicable
to multiple-screw ships, in effect reduces the external exciting force; the flexible material
attenuates the exciting force, that is, reduces the force transmitted to the hull.

B. THE TUNED VIBRATION NEUTRALIZER

The principle of the tuned vibration neutralizer is well known and is discussed rather
thoroughly in most textbooks dealing with the fundamentals of mechanical vibration; see

9-1
Reference 9-1. The device is often referred to in the literature as the "dynamic vibration
absrb
l'er.

The effect of a "sprung mass" on the vibratory characteristics of ship hulls is dis-
cussed in Chapter 6. The vibration neutralizer is in essence nothing but a sprung mass with
provision for tuning its natural frequency and possibly having adjustable damping. It might
seem that, if the designer is concerned about the prevention of a buildup of hull vibration in
a particular normal mode, he has only to install in the ship a sprung mass tuned to the fre-
quency of this mode. It follows from the discussion in Chapter 6, however, that the desired
objective may not be attained that easily. If the sprung mass is large enough to counteract
the external driving force, it will probably be large enough to substitute a pair of hull modes
for the single mode otherwise existing. Hence, for practical application, it is incumbent on
the designer to provide the vibration neutralizer with variable tuning so that its natural fre-
quency can be adjusted to the frequency of the driving force over a considerable range of ship
speeds.
Figure 9-1 shows a large vibration neutralizer that was actually installed on an Italian
motorship and reported to have eliminated 94 percent of the provicUsly existing vibration; see
Reference 9-2. In this case the vibration was caused by machinery and not by propeller action.
Of particular interest is the fact that.the apparatus had a gross weight of 12 tons which was
about 0.1 percent of the displacement of the ship. An interesting feature also was that the
inertia element was a tank divided into many cells that could be flooded with sea water. Thus
the tuning was adjusted by varying the mass, and, in service, an operator was required to ma-
nipulate the valves which flooded or emptied various cells.
The U.S. Experimental Model Basin conducted laboratory experiments with the vibration
neutralizer around 1938 with a view toward exploring its potentialities for use on naval vessels;
see Reference 9-3. While these experiments indicated that electronically controlled P-nrgizzing
devices could improve the performance of the vibration neutralizer under service conditions at
varying operating speeds, they did not appear promising enough to warrant the development of
a full-scale ship neutralizer at the time.
With regard to the feasibility of installing a full-scale shipboard vibration neutralizer,
one point brought out in References 9-2 and 9-3 deserves emphasis here. It may not be nec-
essary to install the apparatus in the immediate vicinity of the exciting source to obtain the
desired neutralizing action. It is true that setting up an equal and opposite force at the driving point
is the most direct method of neutralizing the driving force. However, it follows from the beam
theory of hull vibration that a mass-spring combination will produce an antiresonance of the
system at the frequency to which it is tuned when installed at either end. In the absence of
damping it will maintain whatever amplitude is necessary to hold motionless that point of the
system to which it is attached.
It will be of interest to the reader that, at a much later date than that of the experi-
ments discussed in Reference 9-3, it was discovered, on the trials of a naval destroyer with

9-2
 Figure 9-1 - Vibration Neutralizer Installed
on Italian Motorship MARIA

twin rudders, that the natural frequency in torsional oscillation about the rudder-stock axes
fell close to the frequency of the 3-node horizontal flexural mode of the hull; see Reference
9-4. The result was that the rudders acted as vibration neutralizers with respect to this
mode of the hull. and caused a peculiar forced response pattern when the hull was tested with
a large mechanical vibration generator. This was only one phase of the unusual vibratory re-
sponse characteristics observed on this particular class of ships. Further details are given
in Chapter 14.

C. ADJUSTABLE ROTATING ECCENTRICS

The fact that adjustable rotating eccentrics have been used in the experimental deter-
mination of propeller-exciting forces on model scale (see Chapter 7) indicates the possibility
of using such elements for the elimination of propeller-excited vibration on full-scale ships.

9-3
 When an eccentric of mass m and eccentricity e rotates with a shaft, the reaction trans-
mitted to the bearings in any one direction is

P meOa2 sin wt9-1]

if the angle wt is suitably specified. Thus, in Figure 9-2 the force in the vertical direction
(positive upward) is given by Equation [9-11, and the force in the horizontal direction (pos-
Rive to the right) is

P1 =-meW2 cos ct [9-2]

In Figure 9-2 the bearing is fixed to its supporting structure or base. It might seem
that if this base were part of a vibrating hull the expression for the bearing force would be
much more complicated. It is readily shown, however, that, if the mass of the eccentric is

Figure 9-2 - Rotating Eccentric (Schematic)

Exerting Sinusoidal Vertical and
Horizontal Forces on Its
Supporting Structure
Ph(t) Eccentric of

added to the mass of the base or to the effective mass of the entire system on which the
bearing force acts, then the effective forces acting on the system are still given by Equations
[9-i1 and [9-2], respectively.
If the rotating eccentric is to be used to cancel a force due to mass unbalance, it will
perform the function of an ordinary balancing head and may in fact be attached to the unbal-
anced shaft itself. The important case for consideration here is that in which the rotating
;s 1, bc -Fomd to cancel the blade-frequency exciting forces arising from propeller

9-4
action. In this case, if the propeller shaft has an angular velocity w, the propeller-exciting
forces have a circular frequency r, where s is the number of blades of the propeller. To cancel
these forces, any rotating eccentric device installed in the hull must therefore be driven at
an angular velocity z times that of the propeller shaft.

It is clear that, as contrasted with the vibration neutralizer which introduces a tuning
problem, the adjustable rotating eccentric, if driven through a suitable gear train by the pro-
peller shaft itself, will always synchronize in frequency with the blade-frequency exciting
forces.
In spite of this advantage over the tuned vibration neutralizer, a number of problems
need to be solved in designing a rotating eccentric device to cancel the blade-frequency pro-
peller forces. A single eccentric such as shown in Figure 9-2 yields a rotating force with
sinusoidal vertical and horizontal components. To obtain a pure sinusoidal force in one di-
rection, a pair of eccentrics rotating in opposite directions must be used. Phase control is
also necessary. Thus the device becomes essentially a vibration generator of the type used
to vibrate hulls in ship vibration research, as discussed in Chapter 15.
A complete rotating eccentric device should be designed for both components of the
blade-frequency force and for the couple with respect to the longitudinal axis of the ship as
well. When it is considered how limited the space inside the hull may be in the vicinity of
the propellers, it is apparent that a difficult design problem is involved with such a device.
The problem is discussed further in Reference 9.5.

D. SYNCHRONIZING DEVICES
Synchronizing devices used as antivibration devices are applicable only to multiple-
screw ships. Elementary considerations show that the scheme should be very effective in
certain cases.
If a twin-screw design is considered for illustration, and, if there exists perfect sym-
metry of the geometry of the two propellers and of the flow with respect to the vertical plane
through the longitudinal axis of the ship, then the vertical components of blade-frequency,
propeller-exciting force will be equal for the two propellers. These forces will then be rein-
forcing when the two propellers, rotating at the same angular velocity and in opposite direc-
tions, are so phased that a blade of each propeller reaches the 12 o'clock position at the same
instant. There is no impliuation here that the vertical force reaches its maximum value when
a propeller blade passes through the 12 o'clock position. This is merely a convenient ref-
erence for phase between the port and starboard force vectors shown in Figure 9-3.

9-5
 FPUP PO sin cot

Figure 9-3 -- Time-Vector Diagram for Vertical

Component of Blade-Frequency-Exciting
(Wt + ir) Forces When Canceling Each Other

F..s P. sin (w + r)

In the time-vector diagram (Figure 9-3), the blade position does not appear. This dia-
gram merely indicates that time is taken as zero when the vertical component is zero for the
port propeller. Whether a port propeller blade is in the 12 o'clock positign or not when lot =
1,/2, if the assumed symmetry exists, both vectors in the diagram should coincide when port
and starboard propeller blades pass through the 12 o'clock position simultaneously. If the
vertical forces are canceling, so that the time vectors are as shown in Figure 9-3, then the
reference blade of the starboard propeller should be advanced in the direction of its steady
rotation by an angle of It/s radians from the 12 o'clock position when the reference blade of
the port propeller is in its 12 o'clock position (where z is the number of blades per propeller).
In this simple illustration it follows that a synchronizing device that could maintain
the propellers in this phase relation would ensure the cancellation of vertical blade-frequency
exciting forces. The problemi, however, is never thi's simple. There are horizontal force com-
ponents to consider as woll. Under the ideal condition aesumed, since their vectors rotate at
the same rate, but the horizontal force components are equal and opposite when both propellers
have the reference vector in the 12 o'clock position, the phase condition for a cancellation of
vertical blade frequency forces im Lhte same sv the condition for reinforcement of the hori-
zontal components.
Thus, oven in the ideal situation considered here, the synchronizing adjustment would
have to be based on a compromise between canceling the vertical components and reinforcing
the horizontal components. In practice, moreover, the conditions are much more complicated
than this. Even if the synchronizing device can maintain a prescribed phase between

9-6
propeller rotations, the flow conditions change with speed, heading, rolling, pitching, and
trim of the ship. Nevertheless, if a synchronizing device can ameliorate one ouvere vibratory
condition on a particular class of ship, it may well justify its installation.
"Reference 9-6 discusses the synchronizing devices in detail and points out that the
scheme is most feasible in connection with electric drives. Reference 9-7 describes a syn-
chronizing gear applicable to diesel drives for which considerable success has been claimed.
In this case it was found that the optimum synchronizing angle could best be obtained exper-
imentally. It should also be noted that on multiple-screw ships it is invariably observed that
the blade-frequency vibration is rarely steady but has a beating characteristic. When the ship
is not pitching appreciably, this beating effect is due to the shifting phase accompanying the
slight changes in speeds of the shafts. Thus the degree of benefit to be derived from a syn-
chronizing device is indicated by vibration records that show this beating characteristic.
The synchronizing device, if performing its function, holds the hull vibration level at the
minimum value observed in the beating records.

E. FLEXIBLE MATERIALS IN THE VICINITY OF PROPELLERS

As pointed out in Chapter 7, a large fraction of the propeller-exciting forces acting on
the hull is associated with the pressure field in the vicinity of the propellers. Just as the
vibratory force due to an unbalanced piece of machinery may be attenuated by installing it
on isolation mountings, so, theoretically at least, the effect of the blade-frequency pressure
field .at the stern of a ship can be attenuated by the use of flexible material at the stern.
Elementary considerations suggest that the benefit of such an expedient is highly
frequency-dependent, and, by analogy with the simple problem of attenuating the force trans-
mitted by a machine having a single degree of freedom, the natural frequency must be well
below the frequency of the force that it is required to attenuate.
Blade frequencies are, in general, at the low-frequency extremity of the spectrum of
mechanical vibration. To produce a stern structure with a frequency well below the lowest
blade-frequency disturbance is a difficult design task. Nevertheless, according to Reference
9-8, this was accomplished on the survey ship NORD. In this case the pressure field direct-
ly over the propeller was attenuated sufficiently by the use of a rubber plate backed by an
airtight box to attain a marked reduction in the level of hull vibration.

F. SUMMARY
This chapter mentions only a few of the most promising antivibration devices that
have been suggested for ameliorating the effects of vibratory exciting forces acting on hulls,
A friction damper such as the Lanchester damper discussed in Reference 9-1 could con-
ceivably be developed for such a purpose and it has even been suggested that the propeller
bearing forces be isolated from the hull by flexibly supported struts. No doubt many other
schemes could be tried.

9-7
 The question naturally arises as to whether or not antivibration devices should be
considered in the early stages of the design of a ship. This is certainly a very important
question, for, if the designer could be assured that vibration difficulties could be circum-
vented with such devices, he need not concern himself with the problems of estimating the
magnitude of the propeller-exciting forces or of avoiding hull resonances. He could then con-
centrate on designing the afterbody and propellers strictly from the point of view of propulsive
efficiency.
The available information on the results obtained so far with antivibration devices for
ships and the difficulty of maintenance of such devices suggest that they should not be con-
sidered by the designer from the beginning. Rather, they should be regarded as a last resort
to be used when a design study indicates that large vibratory exciting forces are unavoidable
without alterations in the design which are otherwise inadmissible.

REFERENCES

9-1. Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York (1956).
9-2. Loser, U., "Vibration Neutralizer for Ships," Vol. 12, No. 1 (Jan-Feb 1984). EMB
Translation 12 (Jan 1986).
9-3. McGoldrick, R.T., "Experiments with Vibration Neutralizers," EMB Report 449
(May 1938).
9-4. McGoldrick, R.T., "Rudder Excited Hull Vibration on USS FORREST SHERMAN
(DD931) - A Problem in Hydroelasticity," Trans. SNAME (1959). Also TMB Report 1431
(Jun 1960).
9-5. MacNaught, D.F., Discussion of paper "Hull Vibration Investigation on
SS GOPHER MARINER," by R.T. McGoldrick and V.L. Russo, Trans. SNAME, Vol. 63 (1955).
9-6. Lewis, F.M.9 "Dynamic Effects," Chapter 2, Vol. 2 of Marine Engin, edited by
H.L. Seward, published by SNAME (1944).
9-7. Description of Schelde Synchronizing Gear in Private Communication to SNAME
(26 Aug 1957).
9-8. Waas, H. and Walter, H., "Technische Neuerungen auf dem Vermessungsschiff
'Nord' der Wasser-und Schiffahrtsverwaltung," Schiffstechnik, Vol. 3, No. 12/18
(Nov 1955).

9-8
 CHAPTER 10
DESIGN CONSIDERATIONS RELATING TO
STEADY-STATE HULL VIBRATION

A. INTRODUCTION

Up to the time of this writing, it cannot be said that the consideration of steady-state
hull vibration has been a major item in what may be termed the paper stage of a ship design.
Although, as shown by the existence of Reference 10-1, the subject could not escape the naval
architect's attention, the simple fact is that so little concrete information on ship vibration
has really been available to the naval architect that the question of hull vibration bas been
largely ignored. Only when serious vibration was encountered during the builder's trials of a
new ship did its consideration assume importance. Then measures were taken to modify the
hull or the propulsion machinery to reduce or eliminate the vibration.
While it is true that a large number of technical papers on the subject of ship vibration
were available to the naval architect (as may be seen from the general bibliography included
in this book), the practicing naval architect could not spare the time to digest the mass of
scattered inforrmation contained in these papers. He had to rely on such summaries as given
in Reference 10-1, and it is only because of the increasing tempo of research in this field
since World War II that it is at all feasible to go beyond the limits of Reference 10-1 in this
book.
There was, of course, for many years among naval architects a realization that unbal-
anced forces and moments set up by the propulsion machinery would cause hull vibration. In
fact, studies leading to improved balancing of rotating and reciprocating machinery had been
underway for over 50 years. It was indeed clearly recognized that, in avoiding serious first-
order hull vibration (frequency same as the propeller shaft rpm), the remedy lay chiefly in re-
ducing the forces rather than in trying to reduce the hull response to given forces.
Specifications for static and dynamic balancing of propellers and shafting as well as
the specifications for finishing of propellers were gradually introduced over the years as the
improvement in balancing and manufacturing techniques permitted.
The lack of information was most acute in the nrea of propeller hydrodynamic exciting
forces, and, fortunately, some progress has been made in that area in recert years. It is
obvious that, if the excitiug forces can be reduced to negligible magnitude, steady-state vi-
bration will not be a problem regardless of the natural frequencies of the hull. A possible
exception to this is the production of "flow-excited" vibration which, under special circum-
stances, may produce a fairly steady vibration above a certain critical speed. This point is
discussed further in Chapter 14. There it is also pointed out that control surfaces may induce
flutter phenomena. The latter are not considered flow-excited vibrations in the ordinary sense.
In general, the vibration due to sea action is of a transient nature and depends not only on the

lu-i
dynamical characteristics of the hull but also on the sea state and the seakeeping character-
istics of the hull.
In the design of ships to be driven by conventional screw-propellers, at least one of
the naval architect's chief concerns is the avoidance of serious vibration due to propeller-
exciting forces of blade frequency or integer multiples of the blade frequency.
As will be shown in the next chapter, a serious vibration problem may develop when
trraiisient vibrations of the entire hull (resulting from rough seas) are magnified in local hull
structures. This situation, however, is not chargeable to lack of attention to the items con-
sidered in the present chapter. Furthermore, the fact that transient vibrations may involve
hull amplitudes anywhere from ten to a thousand times those considered under steady-state
vibration does not give the steady-state hull-vibration problem a status of minor importance.
A ship may be slowed down temporarily in rough seas, and its heading may be changed to
minimize the effects of transient vibrations, but it must operate over long periods of time and
at full speed under normal sea conditions in the presence of steady-state vibration.
It is not attempted in this chapter to lay out a rigid design schedule for avoiding
steady-state hull vibration with conventional screw-propeller-driven ships. The subject of
hull vibration is still too obscure to permit this. In fact, too many of the ideas introduced
here are a matter of opinion to justify a manual-type presentation. It is attempted in this
chapter, however, to furnish the, naval. archhitect with specific recommendations based on both
experimental and analytical studies which, if followed, will reduce the "calculated risk" of
hull vibration that must still be taken by the ship designer.
Although this chapter deals only with considerations relating to hull vibration, it must
be pointed out that the selection of a propeller which is to be satisfactory from the hull-
vibration standpoint must be based also on considerations relating to vibration in the pro-
pulsion systemi, as discussd in Chapter 12.

B. GENERAL PROCEDURE
Since the level of steady-state hull vibration that will be encountered in service de-
pends on both the magnitude of the exciting forces and the response characteristics of the
hull, the designer naturally first must consider whether there are any unusual specifications
for the ship in question which would cause either abnormally large exciting forces or an un-
usual sensitivity of the hull to vibratory forces. It has been attempted to make it clear in
previous chapters that in neither category can accuracy of calculation be expected at present.
The only exception is the internal excitation due to mass unbalance. For this the designer
can fall back on the specifications for balancing of machinery, shafting, and propellers (see
Section F) to get an estimate of the first-order exciting forces and moments.
If the ship is to house some novel piece of machinery which develops large unbalanced
forces, then vibration trouble is to be anticipated unless hull natural frequencies are kept
well clear of the operating frequency of this machine.

10-2
 It is also obvious that, if hull natural frequencies can be kept cleaO of the range of
operating shaft rpm's, close attention to maz.s unbalance of propellers and shafting is less
urgent than when they fall in this range.
If the propulsion system is not to include the conventional screw-propeller, a vibration
problem for which little guidance is now available may be imminent. This applies to paddle
wheels of either the side or stern type. The exception is the shrouded propeller, for the
shrouding is intended to smooth out the flow in the propeller race.
If both the hull design and the design of the propulsion system are to be "conventional,"
the next consideration for the designer is the location of the range of operating blade fre-
quencies with respect to the range of significant hull critical requencies. When these two
ranges do not overlap, the likelihood of propeller-excited hull vibration difficulties is greatly
diminished. When they coincide, the reverse is true.
Presumably the designer will have on hand at this stage a design specification for a
propulsion system with a particular range of operating shaft speeds and a particular number
of blades per propeller. He should also have a general idea of the hull scantlings and an
approximate weight distribution plot.
The operating range of blade frequencies is equal to the operating range of shaft rpm's
multiplied by the number of blades per propeller s. The range of significant hull frequencies
can only be estimated on the basis of the hull scantlings and weight distribution data avail-
able at this stage. A simple estimate of the frequencies can be made in accordance with the
rational beam theory if: (1) all decks and expanses of shell plating are provided with longi-
tudinal and transverse stiffeners, so spaced that local natural frequencies are not abnormally
low; (2) no heavy pieces of equipment (heavier than 14 percent of the displacement of the ship)
are installed on resilient mountings; (3) the cargo will not have unusual flexibility such as
possessed by automobiles on inflated tires, large quantities of rubber or plastic material, or
large quantities of springs; (4) there are no abnormally large expanses of deck unsupported
by bulkheads or stanchions; and (5) the ship does not have abnormally large hatch openings
or unusual structural discontinuities.
If the designer knows that the foregoing conditions can be met, he can then make a
very simple estimate of the range of hull critical frequencies. First, he can estimate the
fundamental or 2-node vertical frequency by one of the empirical formulas given in Appendix
C. Better than this, he can use his own value of a constant to be substituted in the Schlick
formula. This formula is

N =C fT [1-1
VAL 3
where N is the frequency in cpm,
C is an empirical constant,
! is the area momient of inertia of the midship section in ft 2 -in. 2
units,

10-3
 A is the displacement of the ship in long tons, and
L is the length in ft.
In lieu of assuming some value of C within the range of values given in Appendix C, the de-
signer can devise his own value of 0 if he has available data on a previous ship of the same
general type. Thus, if he knows N, 1, D, and L for the other ship, he can solve Equation
[10-1] for C. Then from the values of I, D, and L for the proposed design, he can use the
same equation to find N for the new ship.
Unless the LID exceeds 18, the designer can then assume that there will be not more
than six vertical modes, four horizontal modes, or three torsional modes of significance. If
the L/1D exceeds 18, the formula suggested by Baler and Ormondroyd'0 3 can be used; namely

5L
N"-, --9 D [10-21

where N" is the number of significant vertical modes,

L is the hull length, and
D is the hull depth.
The number of horizontal and torsional modes considered significant would then have to be
increased proportionately.
At this stage the designer can make the assumption that the ratios of the frequencies
of the vertical modes fall in the series 1, 2, 3, etc., so that he has only to double the 2-node
frequency to estimate the 3-node frequency, and so on.
Next, the rough rule can be used that the 2-node horizontal frequency will be 1.5 times
the 2-node vertical frequency and that the horizontal frequencies Also follow the 1, 2, 3 rule.
It is not intended to furnish here any empirical rules for the frequencies of submarines but
suuih rules can be developed as information on such ships is accumulated.
Finally, an estimate can be made of the fundamental torsional frequency by Horn's
formula given in Appendix C. While there is little information on which to base the ratios
of the frequencies of the higher torsional modes to that of the fundamental, on the basis
of the information obtained on GOPHER MARINER, 10-4 it is suggested here that the de-
signer assume that the frequency of the third torsional mode will not be over 2.5 times the
fundamental torsional frequency (for GlOPHER MARINER the ratios were I : 1.6 : 2.2).
The next step depends on whether or not the range of hull critical frequencies is clear
of the range of operating blade frequencies. For large ships it is a common occurrence that
the highest significant hull critical frequency is lower than the lowest operating blade fre-
quency. In this case resonance with the blade-frequency forces is not to be anticipated, and
it is to be expected that the blade-frequency vibration will be concentrated in the stern. The
vibration level, of course, will depend on both the driving force and the driving moment. The
force and moment estimating is discussed in Chapter 7. The formulas for estimating the stern
amplitude for known forces and moments are given in Appendix D.

10-4
 15

When it is indicated that the fknqnAnny of amy of the significant. hull modes will fall in
the operating range of blade frequencies, the designer should attempt to predict the level of
re~owaut vibration that would exist it tle two frequencies should coincide. Again the level
will depend on the magnitude of the exciting force, but the response will depend also on the
damping. The force can be estimated in accordance with the information given in Chapter 7,
and the forced vibration calculation can be made by the methods given in Chapter 4.
If the design is such that the conditions specified on page 10-3 cannot be met, then
no vibration predictions are feasible in the early deaign stage and the designer must wait
until the design has advanced to the stage at which the parameters required for a more de-
tailed vibration analysis are available. A sample calculation of a vertical hull mode by the
digital process is given in Appendix A. For details of the calculation of hull modes when
heavy units are to be installed on resilient mountings, see Reference 10-5.
In lieu of attempting to estimate the propeller-exciting forces, the designer should con-
sider the possibility of negotiating for a model determination of these forces. The state of
the art of doing this at the present time is indicated in Chapter 7 as well as in Reference 10-6.

C. REDUCING PROPELLER FORCES

The first-order propeller forces (those due to mass and pitch unbalance of the propeller)
are considered in Section F of this chapter. The present discussion refer3 to the blade-
frequency forces and to forces whose frequencies are multiples of the blade frequency.
As pointed out in Chapter 7, these forces depend on both the pressure fluctuations at
the hull surfaces due to the individual propeller blades and the bearing forces. The latter
forces in turn depend on the uniformity of the flow into the propeller races. Obviously, moving
the propeller astern will usually reduce both components of the forces since the pressure field
at the hull will be weakened and the wake variation (which is aggravated by the boundary
layer when the propeller is close to the hull) will be reduced.
It is equally clear, however, that there are severe limitations to the process of moving
the propellers astern. Some limitations are the danger of fouling the propeller in docking,
the weakening of the support of the aftermost propeller shaft bearing, and the loss of thrust
due to the high wake near the hull. It is possible, however, to obtain increased propeller tip
clearance without moving the propeller by giving the propeller blades a rake in the aft direction.
The use of a propeller tunnel or shrouding is an obvious means of reducing propeller-
exciting forces. This is discussed in some detail in Reference 10-7. If absence of hull vi-
bration is an especially important requirement for the proposed ship, it may well warrant in-
stallation of a shrouded propeller for this reason alone.
In the past the most common expedient to ameliorate the effects of blade-frequency
hull vibration (when encountered on the initial trials of a ship) was to substitute for the
original propeller another of similar thrust and torque characteristics but with a different

10-5
number of blades. Obviously, this expedient may also be introduced in the design stage.
The trend has been from three to four-, five-, and six-bladed propellers.
Increasing the number of propeller blades will, in general, decrease both the forces
due to the pressure field acting on the stern and the bearing forces. The effect of increasing
the number of blades may be compared in a qualitative way to increasing the number of cyl-
inders in an internal combustion engine. The thrust per blade and the contribution to the lat-
eral force per blade are both reduced. Any given variation in wake will obviously cause less
variation in the lift and drag on a single blade when the blade area or angle of attack is re-
duced. Thus, in general, the net effect on the resultant force due to all blades will be re-
duced. The exception occurs when the wake pattern is such that there is greater reinforce-
ment of lateral force components with an increased number of blades. Hence, an analysis
such as discussed in Chapter 7 is really necessary before it is assured that increasing the
number of blades will reduce the exciting forces in the specific case in question. Thus,
when the supporting arrangement is such that two or more blades pass through thc wakez of
obstructions simultaneously, large blade-frequency thrust variations may be expected. Of
couise, the reinforcement of the thrust variations does not necessarily mean that the lateral
force variations will reinforce. The thrusts are unidirectional, whereas the lateral forces
have different directions for different blades. Details of the calculation of bearing forces
from the wake survey are given in Reference 10-8.
Hence, in seeking to reduce propeller-exciting forces without introducing unconven-
tional changes, the designer must, in general, compromise between the expedients of in-
creasing the propeller tip clearances and increasing the number of propeller blades.

D. AVOIDING HULL RESONANCE

The urgency of avoiding hull resonance depends, of course, on the magnitudes of the
exciting forces and the damping. When either the forces are small or the damping is large,
hull resonance is not intolerable and many a ship must operate under such a condition at
particular speeds. While it may be noted here that first-order disturbing forces will steadily
diminish as improved methods of balancing and machining are developed, the question at
issue here is: How can resonance be avoided?
It is, in general, feasible for the designer to avoid first-order resonance if the ship is
to operate at a fixed rpm. The reason is that, although propeller shaft speeds are relatively
low, there is a considerable spread in the frequencies of the lower hull modes. The advan-
tage of avoiding such resonance is obvious since, under such circumstances, a slight un-
balance in propellers, shafting, or machinery will not cause excessive hull vibration.
Keeping the blade frequencies clear of the range of significant hull critical frequencies
is usually relatively easy on large ships and relatively difficult on small ships. This is
because on large ships the operating blade frequencies are usually well above the range of
hull critical frequencies, whereas on small ships they are usually right in this range.

10-6
 When the blade frequencies fall in the range of significant hull critical frequencies, the
prospects of avoiding a hull resonance are not good. This is because first, as indicated in
the previous section, there are normally about a dozen hull modes that may be significantly
excited; and, second, the frequency of each mode will vary with hull displacement over a band
which depends on the variations in loading encountered in the operation of the ship in question.
Figure 10-1, taken from Reference 10-4, illustrates the difficulty the designer may face
in seeking to avoid hull resonance in the blade-frequency range. The ship in question (the

r Regoe
isD.9loemant

ItggvyA , 16.140 Tons

- eo Vibraflto IV)
Veort 6 Ble
---- Horizontal Vibration H)II Fr ty
600-gI

Figure 10-1 - Frequency of Hull Vibrations 400 V

versus Propeller RPM Based on Experi- --
mental Data, GOPHER MARINER, with IV,
Torsional Effects Neglected 3
(From Figure 1 of Reference 10-4.) .

200 .tm

0 20 1o0 0 60 80 20
Propeller RPM

MARINER class) was a cargo ship designed by the U.S. Maritime Administration, with the
following principal dimensions:

L 525 ft
B 76 ft
D 44 ft 6 in.
Displacement 17,000 tons
Mean draft 24 ft
Maximum shaft rpm 105
Propeller blades 4
Single screw

In this case, as in many other cases, the number of significant modes and the varia-
tion of natural frequencies with loading was such that resonance-free areas with propellers

10-7
of the usual number of blades were almost nonexistent. The designer might, however, by re-
sorting to a six-bladed propeller, have been able to boost the operating blade frequency above
the range of significant hull mode frequencies. This was not done in the case of the MAR-
INER Class, which operated satisfactorily with a four-bladed propeller, since the exciting
forces were not excessive for that ship.
When any prospect of keeping the operating blade frequencies above the range of hull
mode frequencies is indicated, the designer should attempt more refined calculations than are
given by the preliminary estimates to more definitely pinpoint the critical speeds. If, after
this, it is still indicated that the blade frequency will inevitably fall in the range of hull crit-
icals, then every effort must be made to reduce the forces.
Of course, blade frequencies can also be increased by using higher shaft speeds. If
the machinery design is not already frozen, this possibility should be considered. A faster
turning propeller can develop the same thrust with a smaller diameter; this also gives in-
creased tip clearance. A high-speed supercavitating propeller can bring the blade frequencies
well above the range of significant hull critical frequencies, and this might be tried if not
ruld oit by other considerations.

E. AVOIDING LOCAL RESONANCE

As in the case of vibration of the entire hull, the amplitude of steady-state local vibra-
tion will also depend on both the magnitude of the exciting force and the damping. Although
the classification "local structure" is somewhat arbitrary, here it means any structure that
can be excited by a local disturbance so as to vibrate without appreciable vibration of the
hull girder. Those are the structures considered in Chapter 6 and include masts, deck houses
of nhort longitudinal extent, panels of deck plating, bulkheads, heavy items of machinery on
nonrigid foundations, as well as items of equipment installed on resilient mountings.
Obviously, changing the natural frequency of a local structure to avoid a condition of
resonance is a possibility, whereas changing a natural frequency of the hull itself after it is
fabricated is extremely difficult.
One unfortunate aspect of the problem of local resonance is that the use of large
numbers of proi~nller blades to reduce the propeller-exciting forces often brings the blade-
frequency into the range of natural frequencies of many local structures whose frequencies
would otherwise be well out of range.
There are, of course, some relatively large local structures for which major alterations
would be required if it were considered necessary to raise their natural frequencies. In such
cases, however, it will usually be found that the local natural frequencies fall in the range of
significant hull mode frequencies, and, in fact, that these structures may even modify the vi-
bratory response characteristics of the entire hull. Under such circumstances calculations
should be made of the natural frequencies and normal modes of the hull with these local struc-
tures treated as equivalent sprung masses; see Chapter 6. The modes found in such

J0-S
calculations will indicate to what extent the amplitude of the local structure will exceed the
hull amplitude in the modes in which it plays a significant role. These modes must now be
considered modes of the combined system and the conditions which ensure that they will not
be dangerously excited are the same considerations that apply to resonant vibration of the hull.
The fortunate aspect of the local resonance problem is that trouble due to this can be
anticipated before the ship is ready for the builder's trials. The naval architect can use the
relatively small and portable vibration generators such as described in Reference 10-9 to de-
termine local natural frequencies. Even without these, such frequencies can be found by the
"bop" or impact test in which a heavy timber is used to excite the structure and the transient
vibrations following the impact are recorded with sensitive instruments. Some tests of this
type can be made even before launching, whereas those in which the bottom plating may have
an effect must be delayed until after launching.

F. BALANCING

The subject of balancing, in general, is a very important one for the naval architect
concerned with avoiding ship vibration. It is treated in. considerable detail in Reference 10-1
and the basic principles are discussod in standard textbooks on mechanical vibration; e.g.,
References 10-10 and 10-11. The naval architect (as contrasted with the marine engineer) is
not so much concerned with the techniques of balancing propulsion machinery as with the spec-
ifications that should be set for balance. He must have assurance that, if he establishes a
specification for maximum permissible unbalanced forces or moments on the basis of his
estimate of the hull response characteristics, these specifications can be met by the manu-
facturer.
Since the advent of steam turbine power plants the chief concern of the naval architect
with regard to balancing, at least for large ships, has been the specifications for balance of
propellers and shafting.
It has been pointed out that a screw propeller is subject to both mass unbalance and
pitch unbalance. The latter requires further consideration here. Pitch unbalance is caused
by lack of uniformity or symmetry in the geometry of the propeller, and may exist even when
mass unbalance is negligible.
If the pitch of one propeller blade is greater than that of the remaining blades, then,
under uniform flow conditions in the propeller race, the lift, drag, and moments on small
elements or strips of this blade are not the same as for the corresponding elements of any of
the other blades. To exaggerate the pitch unbalance effect, one can consider the situation
of a propeller having only one blade. The hydrodynamic force-system acting on the propeller
could then be reduced to a constant axial force, a force lying in a plane normal to the shaft
axis, a constant torque about the shaft axis, and a moment in a plane passing through the
shaft axis. The vectors representing the lateral force and the moment would be constant in
magnitude but would rotate with the propeller. It is seen that where the flow is uniform there
is no thrust variation and no torque variation but that the propeller bearing will be subject to

10-9
simple harmonic forces in both the horizontal and vertical directions as well as to simple
harmonic moments in both the vertical and horizontal planes. The frequencies are first order;
that is, the same as the rpm. When the velocity field is nonuniform, harmonies (integer
multiples of first order) will also, in general, be present.
When the flow in the propeller race is nonuniform, all forces and moments previously
considered for the one-bladed propeller will now be modulated by a "signal" due to the wake
pattern whose fundamental period will be the time taken for the propeller to make one revolu,
tion. In this case there will be thrust and torque variations at the signal frequency, bearing
forces in the vertical and horizontal directions, and moments in both planes of the first order,
All harmonics of the wake pattern signal will also contribute so that the spectrum of bearing
forces may have numerous frequency components. Thus, while pitch unbalance may be ex-
pected to produce first-order vibration, second and higher orders will also be present when
the inflow to the propeller is nonuniform. For further discussion of excitation due to pitch
unbalance, see Reference 10-8.
Military standards for mass balance of rotating members are given in Reference 10-2,
Specification 3.2.3.2. The following is quoted from page 6 of this reference:

"Balance limits - When balanced in accordance with 3.2.3.1 the residual

unbalance, in each plane of correction, of any rotating part shall not exceed
the value determined by:

4W
U - 4 for speeds in excess of 1,000 rpm

or
5680
U =- for speeds between 150 rpm and 1,000 rpm [10-4]
N2
or
V = 0.25 W for speeds below 150 rpm [10-5]

where U = maximum allowable residual unbalance in oz.-in.

W = weight of rotating part in lb
N = maximum operating rpm of unit."

The Westinghouse standard for naval equipment is expressed by the formula

4W
WR < 4 [10-6]

where n is the rpm and WR is the weight, unbalance.

10-10
 An intensive study of tolerance for shafting unbalance was made as a result of first-
order hull vibration on a recent class of naval destroyers. The results of this study are dis-
cussed in Reference 10-12.
Reference 10-13, which deals with relatively small units rotating at high speed, also
discusses the principles of dynamic balancing.
A frequent rule of thumb is that for the maximum running speed of the rotor

WRoA2 - W [10-41
9q 100

or that, hence

< 0.01 [10-81

9

in this formula a consistent set of units must be used. Thus, if R is in in. Y must be
in./sec 2 and c must be the angular velocity in radi/sec.
In the case of rp.cinrncatIing mschinry, if c-reaeilation cannot be obtained hy arranging
the crank angles suitably, it may be necessary to install auxiliary reciprocating masses
drivn by the machine to be balanced. This is discussed in Reference 10-1.

REFERENCES
10-1. Lewis, F.M., "Dynamic Effects," Chapter 2, Vol. 2 of Marine Engin, edited by
H.L. Seward, published by SNAME (1944).
10-2. "Military Standard - Mechanical Vibrations of Shipboard Equipment," MIL-STD-167
(Ships) (20 Dec 1954).
10-3. Baier, L.A. and Ormondroyd, J., "Vibration at the Stern of Single-Screw Vessels."
Trans. SNAME, Vol. 60 (1952).
10-4. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
10-5. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform
Beams Including Methods of Solution and Application to Ships," TMB Report 1317 (May
1961).
10-6. Stuntz, G.R., et. al., "Series 60 - The Effect of Variations in Afterbody Shape
upon Resistance, Power, Wake Distribution, and Propeller-Excited Vibratory Forces,"
Paper prepared for Annual Meeting of SNAME (1960).

10-7. Saunders, H.E., "Hydrodynamics in Ship Design," published by SNAME (1957).

10-8. Reed, F.E. and Bradshaw, R.T., "The Distribution of Exciting Forces Generated
by Propellers," CONESCO Report F-101-2, Contract Nobs 77150 (Jun 1960).

10-11
 10- 9. Robinson, Q.R., "Vibration Machines at the David W. Taylor Model Basin,"
TMB Report 821 (Jul 1952).
10-10. Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York
(1956).
10-11. Timoshenko, S., "Vibration Problems in Engineering," D. Van Nostrand Co.,
New York (1955).
10-12. Jasper, N.H., "Structural Vibration Problems of Ships - A Study of the DD 692
Class of Destroyers," TMB Report C-86 (Feb 1950) Unclassified.
10-13. Feldman, S., "Dynamic Balancing for Noise Reduction," Dept of the Navy,
Bureau of Ships (Code 871), R & D Report 871-V-24 (Apr 1955).

10-12
 CHAPTER 11
DESIGN CONSIDERATIONS RELATING
TO TRANSIENT VIBRATIONS

A. INTRODUCTION

The basic beam theory of ship vibration discussed in Chapter 3 is a general dynamical
theory and is applicable to the transient response characteristics of ships, as shown in Chap-
ter 5. However, this theory cannot be applied indiscriminately to problems such as the slam-
ming of ships in a seaway which may involve deflections so large that buckling and plastic
strain of structural mphnbers occur.
No sharp line of demarcation can be drawn between the field of structural seaworthiness
and the field of ship vibration. Certainly, in the structural strength of ships, the transient
vibration problem is of far more importance than the steady-state vibration problem. In fact,
the hull girder stresses due to steady-state vibration are almost negligible. Tolerances for
steady-state vibration must be based, not on hull girder stress levels, but on permissible
amplitudes determined from physiological factors, satisfactory operation of equipment in-
stalled in the ships, or by local stress and fatigue considerations.
All that is attempted in this chapter is to show that the rational beam theory of ship
vibration advocated in this book can be used as a guide in considering, at least qualitatively,
the transient response characteristics of a hull. Thus, for instance, after a severe slam in a
seaway, a ship will execute a complex pattern of decaying vibrations which eventually settle
down into vibrations in the normal modes considered in Chapter 3. Since these vibrations
often persist for a large fraction of a minute, and are much larger in amplitude than those
usually considered under steady-state vibration, they may cause intolerable vibrations of
local structures whose natural frequencies happen to coincide with their frequencies.
It is also appropriate to point out in this chapter that, although the bending moments
to which a hull is subjected in a seaway are actually of a transient nature., the components
having the frequency of wave encounter vary at such a slow rate that they are treated as
static in the standard strength calculation; see Reference 11-1. The static beam theory to be
used in this case, however, should logically coincide with the rational beam theory considered
to be valid in the low frequency range of hull vibration. Hence the designer should be able
to derive certain conclusions as to the hull strength in a seaway from its vibratory response
characterisaics. Therefore, it is suggested here that the hull girder characteristics as de-
termined from the vibration analysis are applicable regardless of whether the naval architect
continues to rely on the time-honored standard strength calculation which uses a static load
determined from an assumed wave profile, or undertakes to forecast the extreme loading con-
ditions in a seaway.

11-1
 The dynamical system treated in this book is actually an ideal system considered free
in space with mass added to account for the inertia effect of the water and viscous dampers
inserted to allow for the energy dissipation effects.
The transient loading in a seaway which imposes peak stresses in the hull involves
rigid body motions that cause both large variations in the buoyant forces (which are entirely
neglected in the vibration analysis) and large variations in the added mass effect of the sur-
rounding water (which for convenience has so far been treated as constant in time in the
vibration analysis).
Although this treatment of the hull as a system free in space for the purpose of the vi-
bration analysis may seem highly artificial, the designer should realize that the hull by itself
is actually a system free in space subject to a complex system of forces imposed by the sur-
rounding water. If all these forces could be predicted in advance as functions of time, then
the transient problem discussed in Chapter 5 would be valid, at least up to loadings that did
not produce strains exceeding the elastic limit.
It is especially interesting to recall that the hull proper is never permitted to execute
free vibrations, for, no matter how calm the sea may be, the pressures associated with the
innrtia of the surrounding water will always force it to vibrate at frequencies other than its
natural frequencies in free space. As pointed out in Reference 11-2, it was a common prac-
tVce around 1932 to compute the 2-node vertical frequency which a ship would have in air
(free space) and then to apply to this frequency (called the "theoretical frequency") a number
of correction factors. The largest of these was, of course, the factor for the inertia effect of
the surrounding water. Roughly, this effect lowers the free space fundamental vertical fre-
quency by 25 percent.
Although the free space natural frequencies of the hull are considerably higher than
the natural frequencies in water, the corresponding mode shapes are not very different be-
cause the inertia effect of the water conforms roughly with the weight and rigidity distribu-
tions. Since Lhe corresponding components of water pressure drive the hull at a frequency
below its natural frequency, they are in phase with the vibratory displacement, whereas the
vibratory buoyancy forcos are 180 deg out of phase with the vibratory displacement.

B. THE HULL GIRDER

Although the designer has not been accustomed to think of the structural design of the
hull girder as a vibration problem, he has been accustomed to regard the hull as a beam. If
the hull is designed structurally as a beam, then it may be expected to exhibit the vibratory
characteristics of a beam. In fact it is advocated here that the extent to which its vibratory
response characteristics are beamlike may be used as a design criterion. This statement, of
course, requires some clarification because it has been pointed out that if large items of
equipment are deliberately installed on flexible mountings, they can distort the normal mode
patterns considerably from beamlike form.

11-2
 Clearly, the fundamental vertical frequency of a hull increases with its bonding rigid-
ity or El. Although a thin-walled box girder will not be able to develop its nominal El, an
adequately stiffened thin-walled box girder can be made to approach this as a limit. As
stiffeners are added to increase the bending rigidity, the fundamental flexural frequency of
the free-free girder increases and thus the frequency is an indication of the flexural rigidity
in this simple case.
The theory also shows, however, that the frequency depends on the mass distribution
and hence it cannot be the sole criterion of bending rigidity. Furthermore, as suggested in the
case of GOPHER MARINER, 1 -3 8
nonbeamlike vibratory characteristics may be caused by the
cargo which a ship is carrying. The exception in this case proves the rule; it does not inval-
idate it.
If the hull is designed so that its vibratory characteristics are beamlike, in the absence
of heavy items of equipment resiliently mounted, they will then be predictable to a reasonable
degree. If they are predictable, then the propulsion system can be designed to avoid hull res-
onance. Even when it is known that a ship must carry cargo that will make the prediction of
its vibratory response characteristics difficult, there is still an advantage in producing a hull
that will have predictable vibration characteristics with a "normal" cargo. It may be possible
to operate at a slower speed whenever such abnormal cargo must be carried.
The transient loading of hulls in heavy seas can be treated only on a statistical basis
as pointed out in Reference 11-4. It is not intended to imply here that designing to produce
beamlike vibratory characteristics can reduce the slamming loads in a seaway. These depend
on seakeeping characteristics which involve, among other factors, the shape of the hull and
its speed. As a matter of fact, GOPHER MARINER experiments indicate that a cargo that
introduces nonbeamllike vibratory characteristics may introduce a desirable damping action.
The main contention here is that beamlike vibration characteristics for the hull proper are
desirable,
Another factor pertinent to the presenit discussion is the presence of structural dis-
continuities. These are usually associated with local stress concentrations and are con-
sidered undesirable in design. In a beam simply supported at its ends and uniformly
loaded statically, there would be an abrupt change of curvature in the deflection curve at
the point of discontinuity. Similarly, if such a beam were set vibrating in its fundamental
normal mode the pattern would also show a sharp change of cdrvature at this point.
Thus the rule (but a rule with exceptions) advocated here is that the vibratory response
characteristics should be beamlike. The exceptions occur when large units are to be installed
in the hull on resilient mountings or when unusual cargo having "springy" material is to be
carried. In such cases the hull, even if designed for beamlike characteristics without such
conditions, will not exhibit them in practice.
While it is not advocated here that the digital method of transient response calculation
given in Chapter 5 is applicable to severe slamming load conditions, it might be noted that

11-3
calculations made by the method discussed in Reference 11-5 predict that the vibratory re-
sponse to an impulsive load at the bow will be predominantly in the 2-node flexural mode.
This is generally confirmed by vibration measurements on ships in a seaway; see, for example,
Reference 11-6.

C. LOCAL STRUCTURES
The transient motions of local structures when the ship is operating in heavy seas are
of serious concern to the designer. Obviously, the vibratory motions of local structures which
do not carry an internal source of excitation (such as an unbalanced piece of machinery) will
depend on the motions of the hull girder in their vicinity. Even without considering the sta-
tistical problem of predicting the peak values of the transient motions which the hull will ex-
ecutA in a seaway, there can still be pointed out here the desirability of keeping the natural
frequencies of local structures clear of the frequencies of the significant hull modes.
If it were a simple matter to raise the natural frequencies of local structures, the an-
swer to this problem would be to reinforce them sufficiently so that all local natural frequen-
cies were above the range of hull mode frequencies. Since this not feasible with the larger
structures (such as masts), the designer must consider which hull modes are most effective
in exciting the particular local structure in question. As an illustration, a pole mast will
have relatively low cantilever natural frequencies in the fore-and-aft and athwartships direc-
tions. The fore-and-aft mode responds to hull girder rotation in the vertical plane passing
Lhrough the longitudinal centerline of the hull. In the vertical flexural modes of the hull the
rotation in the vertical plane is relatively large noar the nodal points and a minimum at the
antinodes. In the fundamental or 2-node mode the rotation is a maximum at the ends.
Figure 11-1 illustrates a situation in which a pole mast is located at a point of relatively
large rotation in the fundamental vertical flexural mode of the hull. On slamming in a seaway
a train of decaying vibrations of large amplitude in this mode will be excited. If the pole
mast were rigid, the fore-and-aft motions at the top of the mast due to this vibration of the hull
would be h y, where h is the distance from the top of the mast to the elastic axis of the hull
and y is the angular amplitude of the hull at the location of the mast. If, however, the fore-
and-aft cantilever natural frequency of the mast happens to coincide with the frequency of this
flexural mode of the hull, the fore-and-aft motion at the top of the mast will be greatly magni-
fied. Thus, in this case, a special effort should be made to ensure that the fore-and-aft can-
tilever frequency of the mast is clear of the 2-node vertical frequency of the hull. Similarly,
athwartship motion at the top of the mast is produced by torsional vibration of the hull at the
base of the mast and thus the frequency of the 1-node torsional mode of the hull should also
be avoided if possible.
It must be emphasized here again that when massive local structures have natural fre-
quencies (estimated on the assumption of a restrained hull) which fall near a frequency of one
of the modes which the hull would have if the local structure were rigid, the modes and

11-4
 Figure 11-1 - Illustration of Case in Which a Pole Mast Is Located at a
Nodal Point of the Fundamental Vertical Mode of a Hull

frequencies of the hull itself are modified by the "sprung mass" effect. Details of analytical
methods of dealing with this situation are given in Reference 11-7. Unfortunately, it is only
possible to carry out such analytical treatments when the hull design has reached a relatively
advanced stage. Only then can the parameters required in the equations be evaluated.

D. RESILIENTLY MOUNTED ASSEMBLIES

When all six of the natural frequencies of a rigidly mounted assembly (computed on the
assumption of an immovable hull) are well above the range of significant hull mode frequen-
cies, it may then be assumed that under transient excitation of the hull the assembly will
move with the hull. Under such circumstances the resilient mounting is installed for the pur-
pose of isolating the assembly either from blade-frequency vibrations that are well above the
hull natural frequency domain or from high-frequency vibrations accompanying shocks which
produue local vibrations in the area of the mounting.
As in the case of local structures, a massive assembly such as an engine, if installed
on isolation mountings, will affect the hull frequencies themselves, and then it must be con-
sidered as a sprung mass. For details of analysis in such cases, see Reference 11-7. In
the case of machinery, the resilient mounting is used chiefly to isolate the hull from the un-
balanced forces and moments due to the operation of the machine itself. Such isolation will
be achieved only when all six natural frequencies of the assembly are below the operating
frequencies (or speeds) of the machine.
A matter of chief concern in the selection of an isolation mounting system for an as-
sembly is whether the mountings will bottom under rough sea conditions. Should this happen
the equipment may suffer more severe damage than if it had been rigidly secured to the hull.
As a safeguard against this, shock tests have been devised for military equipment, as dis-
cussed in Reference 11-8. The damaging effect of bottoming can also be reduced by the use
of snubbers; see Reference 11-9.

11-5
 REFERENCES

11-1. Rossell, H.E. and Chapman, LTB., Editors "Principles of Naval Architecture,"
published by SNAMEO Vols. I and 2 (1989).
11-2. McGoldrick, R.T., "A Study of Ship Bull Vibration," EMB Report 895 (Feb 1935).
11-3. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 68 (1955). Also TMB Report 1060 (Jul 1956).
11-4. Jasper, N.H., "Statistical Distribution Patterns or Ocean Waves and of Wave-
Induced Ship Stresses and Motions, with Engineering Applications," Trans. SNAME, Vol. 64
(1956). Also TMB Report 921 (Oct 1957).
11-5. McGoldrick, R.T., "Calculation of the Response of a Ship Hull to a Transient Load
by a Digital Process," TMB Report 1119 (Mar 1957).
11-6. Jasper, N.H. and Birmingham, J.T., "Strains and Motions of USS ESSEX (CVA 9)
during Storms Near Cape Horn," TMB Report 1216 (Aug 1958).
11-7. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).
11-8. Vane, F.F., "A Guide for the Selection and Application of Resilient Mountings to
Shipboard Equipment," TMB Report 880, Revised Edition (Feb 1958).
11-9. Credo, C.E., "Vibration and Shock Isolation," John Wiley & Sons, Inc., New York
(1951).

11-6
 CHAPTER 12

DESIGN CONSIDERATIONS RELATING TO VIBRATION OF

THE PROPULSION-SHAFTING SYSTEM

"A.INTRODUCTION
Machinery vibration per ae is a subject which is outside the domain of hull vibration.
However, there are various types of vibration in the propulsion system of a ship which may be
excited by the propeller itself regardless of the degree of mass balance attained in the manu-
facture of the propulsion machinery. Since the designer is not free to select a propeller on
the basis of hull vibration characteristics alone but must also take such propeller-excited
machinery vibration into account, some consideration of the latter has been included in this
book. The vibration types in question are torsional vibration of the propulsion-shafting sys-
tem, longitudinal vibration of the propulsion-shafting system, and lateral vibration of the pro-
peller shaft.
Torsional vibration of propulsion systems is characterized by sinusoidal time vari-
ations in the angular velocity of the rotating members or, in other words, by the superposition
of angular oscillations on the steady rotation of these members.
Longitudinal vibration is characterized by fore-and-aft oscillations of the propeller,
the shafting, and 'he entire propulsion machinery, including both rotating and nonrotating
members.
Lateral vibration of the propeller shaft is frequently spoken of as whirling. Under this
type of vibration the shaft center describes a circular or elliptical path in a plane normal to
the shaft axis. In some cases the ellipse may be very narrow and in the limit yield a recti-
linear path.
In all three cases the basic phenomenon is now well understood and has been discussed
in the technical literature in recent years. This does not mean, however, that the critical
speeds can in all cases be predicted with high precision. As a matter of fact, in the evalua-
tion of the parameters used in predicting such vibration, difficulties exist similar to those
pertaining to the evaluation of the parameters used in the hull vibration calculations.
The analysis of all three types of vibration in the propulsion system follows the same
general line as in the case of the hull. Thus, an estimate is usually made of the natural fre-
quencies and normal modes of vibration of the system. Then an estimate is made of the ex-
citing forces. It must be noted here that in the case of diesel drives, torsional excitation ex-
ists not only at the propeller but also in the prime mover. Next, an estimate is made of the
damping in the system, and finally, a calculation is made of the forced vibratory response.
This is then compared with permissible levels established on the basis of past experience,
and, if predicted to be excessive, steps are taken to reduce it.

12-1
 B. TORSIONAL VIBRATION
The subject of torsional vibration of propulsion systems has been treated extensively
in the technical literature; see Reference 12-1. Historically, the phenomenon first became a
serious problem in internal combustion engine systems and the literature on the subject in
this connection is extensive; see Reference 12-2. The basic theory is the same as applied
to the pure torsional modes of the hull itself. In the propulsion system, however, the analy-
sis is complicated by the presence of reduction gears, high and low pressure turbine sys-
tems, or, in the case of diesel drives, by reciprocating members.
For the immediate problem of this book, the designer needs assurance that a selected
propeller will not create a torsional vibration condition which would not exist if the propeller
were replaced by a cylindrical mass of the same mass moment of inertia. The latter parameter,
incidentally, must include an allowance for water inertia effect and this also poses a problem
in evaluation of parameters; see Reference 12-8.
Fortunately, it is usually feasible to keep Lhe operating range of blade frequencies
clear of the range of torsional criticals. In the case of geared turbine drives, which at pre-
sent comprise the majority of ocean-going ships, a common practice is for the designer to
select a "nodal drive"; see Reference 12-1. In this case the first torsional critical speed is
so low compared to the normal operating speed of the ship that excitation from the propeller
is negligible. The second torsional mode is "tuned out." This is accomplished by design-
ing the two turbine branches so that they will have the same torsional natural frequency (the
bull gear being held fixed). This produces a mode such that the entire propulsion system from
the propeller to the bull gear is a node and hence, torque variations at the propeller will not
excite this mode. The third and higher torsional modes fall well above the maximum operating
blade frequency.
A general scheme for the torsional vibration analysis of propulsion systems developed
at the Bureau of Ships of the Navy Department is given in Reference 12-4. The principal aim
of the designer in applying this method is to guarantee that significant torsional vibration will
not exist in the propulsion system in normal operation of the ship.

C. LONGITUDINAL VIBRATION
Longitudinal vibration of propulsion systems has attracted general attention more re-
cently than torsional vibration although the phenomenon undoubtedly occurred long before it
was recognized. It was brcught strikingly to attention when it appeared unexpectedly on large
naval vessels during World War 11.12"5
In this case, the analysis involves equations quite similar to those involved in tor-
sional vibration analysis; the only difference is that forces and rectilinear displacements
are involved instead of torques and angular displacements.
The chief difficulty in the analysis is in the evaluation of the longitudinal stiffness of
the thrust bearing and machinery foundations, but, fortunately, in those propulsion systems in

12-2
which the longitudinal natural frequencies are most likely to fall in the operating range of
blade frequencies, the shafts are relatively long, and the frequencies of the lower longitudinal
modes do not vary widely for large variations in the foundation stiffness.
Here, as elsewhere, the designer is more concerned with obtaining assurance that the
longitudinal critical frequencies are clear of the operating range of blade frequencies than
with being able to predict the frequencies with precision.
SIn general, longitudinal and torsional vibrations of propulsion systems may be con-
sidered as independent of one another, but this is not always the case, as pointed out in
References 12-3 and 12-6. The propeller couples the longitudinal and torsional degrees of
freedom of the system to some extent under all conditions, but the coupling effect is signif-
icant chiefly when the critical frequencies that would exist without this coupling effect are
close to one another. In such cases the mode excited is actually a longitudinal-torsional
mode and the excitation involves a generalized force which combines the effect of torque
and thrust variations. However, for the designer, it is more important to ensure that the
longitudinal and torsional frequencies are kept far apart than to be able to predict the ampli-
tudo when they are close together.
It is obvious that, where short drive shafts are involved, as when the machinery is
installed far aft, the stiffnesses in the propulsion system are so high relative to the stiffness
which the hull presents to the thrust bearing foundation that analyses of longitudinal vibra-
tion that consider the hull as fixed are not realistic. Here, as in numerous other cases, the
longitudinal vibrations of the propulsion system merge into vibrations of the entire hull-
machinery system for which, as yet, reliable methods of prediction have not been developed.
The fortunate circumstance is that (at least in the case of surface ships) in such instances
the highest operating blade frequency is usually below the first longitudinal critical frequency.

D. LATERAL VIBRATION OF PROPELLER SHAFTS

Considerable impetus was given recently to the study of lateral vibration of propeller
shafts when it was suspected that the phenomenon of shaft whirl was the principal cause of
the failures of tailshafts that occurred at an alarming rate on single-screw merchant ships
shortly after World War II. Although it was later found that the chief stress variations in the
tailshafts were due to thrust eccentricity, 12-7 the investigation of the problem led to improved
methods of predicting tailshaft whirling speeds and stresses; see References 12-8 and 12-9.
As emphasized in Reference 12-9, the combination of a heavy propeller and a long
overhang from the tailshaft bearing may result in a low whirling speed of the tailshaft. The
Bureau of Ships procedure for ensuring against a whirling speed in the running range is out-
lined in Reference 12-10. A general discussion of design stage calculations for marine
shafting is given in Reference 12-11. Inertia effects of the surrounding water on propellers
are discussed in Reference 12-3.

12-8
 REFERENCES

12- 1. Lewis, F.M., "Dynamic Effects," Chapter 2, Vol. 2 of Marine Engin, edited by
H.L. Seward, published by SNAME (1944).
12- 2. Wilson, W.K., "Practical Solution of Torsional Vibration Problems," 2 Vols.,
John Wiley & Sons, Inc, New York (1956).
12- 3. Lewis, F.M. and Auslaender, J., "Vertical Inertia of Propellers," JSR (Mar 1960).
12. 4. Gariboldi, R.J., "Procedure for Torsional Vibration Analysis of Multimass Systems,"
Dept of the Navy, Bureau of Ships (CQde 371), R & D Report 37'1-V-19 (Dee 1953).
12- 5. Kane, J.R. and McGoldrick, R.T., "Longitudinal Vibrations of Marine Propulsion-
Sha'ting Systems," Trans. SNAME, Vol. 57 (1949). Also TMB Report 1088 (Nov 1956).
12- 6. Panagopulos, E.P., Discussion of Paper "Longitudinal Vibrations of Marine
Proulsion-Shafting Systems," by J.R. Kane and R.T. McGoldrick, Trans. SNAME, Vol. 57
(1949).

12- 7. Jasper, N.H. and Rupp, L.A., "An Experimental and Theoretical Investigation of
Propeller Shq-ft Fniuven," Trans.
12- 8. Neifert, H.R. and Robinson, J.-., "Further Results from the Society's Investi-
gation of Tailshaft Failures," Trans. SNAH.E (1955).
12- 9. Jasper, N.H., "A Design Approach to the Problem of Critical Whirling Speeds of
Shaft-Disk Systems," TMB Report 890 (Dec 1954).
12-10. Michel, R., "Interim Design Data Sheet DDS 4301., Propulsion Shafting," Dept of
the Navy, Bureau of Ships, S 1/1-2(415), DDS 4301 (1 May 1957).
12-11. Panagopulos, E.P., "Design-Stage Calculations of Torsional, Axial, and Lateral
Vibrations of Marine Shafting," Trans. SNAME, Vol. 58 (1950).

12-4

I~
 CHAPTER 13
71
DESIGN CONSIDERATIONS RELATING TO RESILIENT MOUNTINGS

It was inevitable that in the initial attempts at the isolation mounting of equipment in
ships it was found in many cases that the remedy was worse than the disease. For this reason
many such installations wore provided with locking arrangements so that, if the mounting
were found to make conditions worse, it could be in effect eliminated. It also became a com-
mon practice to provide snubbers to ease the- impact when the mounted assembly bottomed;
see Reference 13.1.
Although a decision to install major items of shipboard equipment on resilient mount-
ings is a very important one from the design standpoint, this chapter is concerned only with
the questions which arise after such a decision has been made. The subject of isolation
mounting of shipboard equipment, in general, is discussed in detail in References 13-1 and
13-2. The questions at issue here are: How does the use of resilient mountings in a ship
affect the behavior of the ship from the vibration standpoint? What account of this should be
taken in the hull design?
If the weight of an assembly to be supported by resilient mountings is of the order of
, percent or more of the displacement of the ship, and any one of its six natural frequencies,
computed on the assumption of fixed support of the ends of the mounts, falls within the range
of significant hull frequencies (as estimated with the assembly rigidly attached to the hull),
then it is advisable to consider the effect of the resilient itounting on the nurmnal modes of
vibration of the hull.
For simplicity it is assumed here that the mounted system can be designed to have
two planes of vibrational symmetry and that these planes are the XZ- and YZ-planes shown
in Figure 13-1 where 0 is at the c.g. of the
mounted assembly when in its rest position, Z
and the XZ-plane coincides with the vertical
plane through the longitudinal axis of the ship. - y
The concept of vibrational symmetry is dis-
cussed in Chapter 6.
In all cases considered in this book the
vertical hull modes are treated as independent
of the horizontal and torsional modes. How-
ever, it must be pointed out here that, if a
heavy assembly is flexibly mounted at a con-
siderable distance either to port or starboard Figure 13-1 - Right-Hand Coordinate System
Used in Calculation of Natural Frequencies
of the ship's longitudinal axis and above or and Normal Modes of Vibration of a
below it, there results a coupling of all three Resiliently Mounted Rigid Assembly

13-1

-i
(vertical flexural, horizontal flexural, and torsional) types of vibration. The equations for
the more general cases of hulls with sprung masses are given in Reference 18-3.
The designer first must carry out the calculation of the natural frequencies and normal
modes of the mounted assembly on the assumption that the hull is immovable. This process
is discussed in Chapter 6. Also, the estimates of natural frequencies and normal modes of
the hull must be made on the assumption that the mounted assembly is blocked so that it can-
not move relative to the hull.
If it is recalled that the effect of the sprung mass on the hull is significant only for
modes whose frequency is near the natural frequency of the sprung mass, and that the effect
is to lower the frequency of a mode below and to raise the frequency of a mode above, it is
apparent that the designer can derive some guidance from the two foregoing calculations even
without making the more elaborate calculations for the hull with the masses treated as sprung.
Thus, if the operating blade frequencies fall well above the range of significant hull mode
frequencies predicted with rigid mountings, no difficulty with steady-state, propeller-excited
vibration is indicated when the natural frequencies of the assembly fall well within the range
of hull mode frequencies. Of course, if an assembly has a natural frequency near that of the
highest significant hull mode, and the lowest operating blade frequency is only slightly above
this, a posmbility of resonance exists and the need for further calculations is indicated.
However, the designer is still faced with the possibility of excessive transient vibrations.
If the mountings have locking devices it may still be possible to avoid bottoming of the mount-
ings under rough sea conditions. When a sprung mass effect is due to the flexibility of a
massive local structure itself, this possibility, of course, does not exist.
One might argue that, if the resilient mountings of an engine were to be locked out under
rough sea conditions, they should not have been installed in the first place. Such reasoning,
however, is not sound. If the mountings are intended primarily to isolate the hull from the
relatively high frequency vibrations of the engine, then this will be accomplished under normal
operating speed conditions in relatively calm seas. When the sea gets rough, the engines
would be slowed down and thus their unbalanced forces and moments would be reduced.
Furthermore, if severe transient vibrations should be set up in the hull under these sea con-
ditions, the vibration induced by the engine would probably be relatively unimportant as long
as these conditions prevailed. Thus, if an isolation mounting system is designed mainly for
eliminating steady-state vi0-ation, it may be justified to lock it out under slamming conditions
in which it would be required to act as a shock mounting which should be designed for more
severe conditions.
Although the subject of mass-unbalanced rudders is discussed in some detail in con-
nection with hydroelasticity in the following chapter, it is in order to point out here that such
appendages can have a marked sprung mass effect on the horizontal vibration of a hull. At
the present time it is indicated that with a hydraulic steering geat L.Ie effective torsional
stiffness of the rudder system may be relatively low. Thus, it cannot always be assumed that

13-2
rudder torsional frequencies will be above the range of significant horizontal hull mode
frequencies. The vibration analysis of such systems is discussed further in References
13-4 and 13-5.
In summary, whenever the isolation mounting of large masses (of the order of %percent
or more of the mass of the ship) is under consideration, the designer should took into its
effect on the vibration characteristics of the ship as a whole.
It must be kept in mind that the discussion of isolation mountings in this book does
not extend into the range of acoustic frequencies in which wave effects, such as discussed
in Reference 13-6, take place.

REFERENCES

13-1. Credo, C.E., "Vibration and Shock Isolation," John Wiley & Sons, Inc., New York
(1951).

13-2. Vane, F.F., "A Guide for the Selection and Application of Resilient Mountings to
Shipboard Equipment," TMB Report 880, Revised Edition (Feb 1958).
13-3. Le~bc.ndtz, R.C. and Konriard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).
13-4. McGoldrick, R.T., "Rudder-Excited Hull Vibration on USS FORREST SHERMAN
(DD 981) - A Problem in Hydroelasticity," Trans. SNAME (1959). Also TMB Report 1431
(Jun 1960).
13-5. Leibowitz, R.C., "Modes of Vibration of Rudder of USS ALBACORE (AGSS 569),"
TMB Report C-952 (Feb 1959) CONFIDENTIAL.
13-6. Harrison, M., et al., "Wave Effects in Isolation Mounts," TMB Report 766 (Jan 1952).

13-3
 CHAPTER 14
HYDROELASTICITY
V
A. INTRODUCTION

At this time hydroelasticity must be regarded as a new science for the naval architect.
Although the basic phenomena dealt with in hydroelasticity are not new, for the most part their
occurrence has been rare in the operation of ships, and until recently they have been dealt
with individually as isolated phenomena.
What is really new is the recognition that these separate phenomena comprise a group
falling under the definition of the term "hydroelasticity." Although various definitions of
this term will be found in the literature (see References 14-1, 14-2, and 14-3), it seems suffi-
cient to state here that hydroelasticity is concerned with those problems in which water vehi-
cles are subjected to time-varying forces imposed by the water, but governed also by the elas-
tic properties of the hull or its appendages. In general, hydroelasticity in naval architecturo
parallels the well-established field of aeroelasticity in aeronautical engineering; see
Reference 14-4.
Note that the definition of hydroelasticity just given is really broad enough to include
propeller-excited hull vibration itself, which is the chief topic discussed in this book. How-
ever, the term is currently restricted to problems with hulls or appendages which involve the
water flow past the hull, with the exception of propeller-excited hull vibration, since the
latter is of such common occurrence as to warrant individual consideration.
Lateral vibrations of circular cylinders moving in fluids are frequently observed, and
this phenomenon has been found to be associated with the shedding of vortices. In particular,
such vortices are commonly called Karman vortices and the series of these observed in the
wake of a towed cylinder is called a Karman -,ortex street. The frequency of shedding such
vortices depends on both the towing velocity and the diameter of the towed cylinder, and
hence, in the design of appendages such as propeller struts, the use of a streamlined section
is important not only in reducing drag but also in preventing vibration. The Karman vortex
shedding phenomenon is discussed in further detail in Reference 14-I.
Another hydroelastic phenomenon that can be troublesome to the naval architect is the
so-called "Helmholtz resonator" phenomenon. It is illustrated very simply by the production
of a musical tone by blowing over an opening in an air-filled chamber. In naval architecture
the term is something of a misnomer since the chamber is an enclosure whose walls also may
vibrate and play an important role in determining the frequency. Nevertheless, the hydro-
elastic phenomenon is similar to the one in air in the sense that it is the flow over an open-
ing in a compartment that produces it.
Two other phenomena of prime importance in aeroe.asticity have their counterparts in
hydroelasticity although they occur less often in the latter case. These phenomena are

14-1
"flutter" and "divergence" and apply particularly to hydr -?, s, that is, to lifting surfaces or
control surface members.
The basic phenomena of flutter and divergence are dealt with in great detail in text-
books on aeroelasticity; see, for example, References 14-4, 14-5, or 14-6. They arise from
the variation in lift force and moment with velocity and angle of attack, and involve the ques-
tion of the static or dynamic stability of the lifting surface member under its normal range of
operating speeds.
As an illustration of divergence, consider the torsional stability of the spade rudder
shown schematically in Figure 14-1. In this figure the rudder stock is intentionally placed

Figure 14-1 - Schematic Plan and Elevation

of Spade Rudder Subject to Possible
Torsional Divergence
S X

near the mid-chord position to exaggerate the hydrodynamic instability effect. This axis
would normally be near the forward quarter-chord position. For simplicity, assume that the
water flow in this case is parallel to the ship's longitudinal axis and that the water velocity
relative to the rudder is equal to -S, where S is the ship's forward velocity. Further, assume
M'
that the flow imposes on the rudder a moment M. with respect to the rudder stock axis given
by the relation

met MCM/ s2a [14-1]

where a is the angle of attack and Cs is a moment coefficient for this particular rudder.
If the steering gear were locked, any change in the external steady moment applied
to
the rudder would evoke an angular displacement pa, and an elstipic restoring moment
equal
to -kA a, where k is the torsional spring constant of the rudder-steering system. Thus,
if
equilibrium exists at an initial angle of i a m , any additional angular displacement
Aa

evokes a restoring moment -kA a and an upsetting moment CM 52l.a, so that the net restoring
moment is

14-2
 - (k - CM 22) A a

Thus, in this case, the hydrodynamic moment produces a negative torsion spring effect which
varies with the speed of the ship. If the ship should attain such a speed that

k _CMS2 = 0 [14-2]

the torsional rigidity of the system would vanish, and, at any speed higher than this, the hydro-
dynamic moment would twist the rudder to failure. In this simplified example the value of S
obtained by solving Equation [14-21 would be called the speed of torsional divergence. Diver-
gence is thus a condition of static instability or instability under a nonoscillatory flow
condition.
Flutter is a condition of dynamic instability and is discussed in more detail later in
this chapter. As applied to a control surface member, the phenomenon may be explained in
simple terms also by reference to Figure 14-1. Suppose that, instead of steady conditions
prevailing, the stern of the ship is vibrating horizontally (in the athwartship direction).
Then, if the rudder lacks mass balance with respect to the rudder stock axis, angular oscil-
lations of the rudder will also be generated. The latter will evoke variations in the hydro-
dynamic lift force acting on the rudder. The effect of these variations in lift force on the
system will depend, of course, on their frequency, magnitude, and phase. If conditions are
such that the lift force variation is in phase with the vibratory velocity and in the same direc-
tion, it will exert a negative damping effect on the system. Since the lift force variation in-
creroses as the ship's speed increases, it may happen that it is suitably phased and of such
magnitude at a certain speed as to cancel out all positive damping effects in the system. A
speed at which this condition was reached would be called a critical flutter speed. At a
higher speed, if the phase did not shift, the system would have a net negative damping. Under
such a circumstance an oscillation once started would build up exponentially until damage or

Still another phenomenon falling under the category of hydroelasticity is the "singing"
propeller, also discussed in more detail later in this chapter. This is by no means a new
phenomenon to the naval architect (see Reference 14-7), but it is still not thoroughly under-
stood. It involves the development of vibration of a propeller blade when the shaft attains a
certain speed, the frequencies usually being in the audible range. It is clearly a hydroelastic
phenomenon since it depends on both the flow past the blade as it rotates and the mass-elastic
properties of tho propeller.

B. RUDDER-HULL VIBRATION

In spite of the fact that W. Ker Wilson (Reference 14-8) had cited rudder flutter as a
recognized phenomenon in 1954, few cases of rudder vibration had been reported in the

14-3
United States prior to its occurrence on a now class of naval destroyer. The United States
Navy encountered a case of hull vibration in 1956 which was traced to the behavior of a pair
of twin rudders. Although the ship involved was a naval craft, the investigation of the prob.
lem revealed nothing that would restrict the phenomenon specifically to naval types, and
hence it is cited here as a type of hull vibration problem other than propeller-excited with
which any naval architect might be faced in the future. It is a typical problem in hydro-
elasticity. Its investigation is discussed in detail in Reference 14-9 and only a synopsis is
given here.
On the initial trials of the ship in question, a 3-node horizontal vibration of the hull
developed in the upper speed range and persisted with a practically constant frequency
throughout this speed range. The steadiness of the frequency, of course, eliminated pro-
peller excitation as the cause. The vibration was finally traced to the twin rudders which
were initially set with the trailing edges "in" a few degrees to minimize power consumption.
On reversal of this setting, that is, by placing the trailing edges a few degrees "out," this
vibration diminished to a permissible level.
To make a long story short, the subsequent investigation showed that the torsional
natural frequency of the rudders fell close to the frequency of the particular hull mode excited,
and, more specifically, that the rudder-hull system had a normal mode in which the hull vi-
brated in 3-node flexure while the rudders executed torsional oscillations about the rudder
stock axes. The situation satisfied the basic conditions that would make control surface
flutter possible if the lift coefficient of the rudders and the speed were high enough.
The basic mechanism of control surface flutter can be explained in simple terms by
considering the "classical" flutter system of two degre~es of freedom illustrated schemat-
ically in Figure 14-2.
The control surface member or hydrofoil which may oscillate torsionally about a ver-
tical axis is supported by a structure which can oscillate only in translation in a direction
normal to the steady velocity. The latter structure of mass A is suspended by flexural
springs of combined stiffness K from a rigid member which moves at uniform velocity S above
the water surface. The rigid member also supports a surface plate (not shown) just above the
hydrofoil which moves with uniform velocity and eliminates surface disturbances. The dis-
tance from the axis to the center of gravity of the rotating element (based on an allowance for
added mass effect of water) is designated h which is considered positive if the c.g. is down-
stream.
The translational system has damping equivalent to a viscous damping constant C; the
rotational system has mass moment of inertia 1, which includes an added inertia for entrained
water; an elastic stiffness k; an equivalent viscous damping constant c; and a mass m, in-
cluding an allowance for added mass effect of the water. The displacement. of the axis of the
control surface in translation is designated Y and the rotation of the hydrofoil about the Z-
axis (not shown) is designated 0.

14-4
 M, C

, c.g.

A'A

Figure 14-2 - Schematic Plan View of a Control Surface Flutter

System of Two Degrees of Freedom

A simplified analysis of the oscillatory, motion of such a system may be based on the
following assumptions:
(a) The lift force during oscillatory motion may be taken from the lift coefficient for
steady flow which is a linear function of angle of attack.
(b) The hydrodynamic moment with respect to the axis of rotation of the foil is given by
multiplying the lift force by the distance L from the axis to the center of lift which is a con-
stant L. L is considered positive if the center of lift is upstream.
This yields the following pair of simultaneous ordinary differential equations in the
variables Y and 0:

I'+Ob + (k - ALS2) 0 - mAy + ALSi - o [14-81

-mhio- AS 2 0 + (H + ?m)Y + (C + AS1) i + KY- o [14-4]

Such a pair of differential equations can be solved by an analog computer, or they may
be converted to algebraic equations by the assumptions Y. - 0 . If the resulting
00el
equations are then divided by eAt, the result will be a pair of simultaneous homogeneous
algebraic equations in Y. and 0o with k appearing in the coefficients. The determinant of the
coefficients is called in this case the "flutter determinant." When it is set equal to zero the
fe~suking equation is a quattic in k.

14-5
 The stability of the oscillations then is determined by the four roots of this equation.
In general they will be complex. The real part of any complex root then indicates the exponen-
tial rate of decay or buildup of a possible mode of vibration. These roots, however, depend on
the value of speed S and, in general, it will be found that their characteristics change markedly
as S increases. If the roots of A are computed for a series of values of S, the damping of the
2
system may be plotted as a function of S. Thus, if X - p + jw, plots of // io., will show how
the rate of decay or buildup in a possible mode of vibration varies with speed for a given set
of values of the parameters of the system according to Equations [14-38 and [34-4].
Since the system has two degrees of freedom, it will, in general, have two normal modes
of vibration, and, if it is given an arbitrary displacement, its subsequent motion will be a com-
bination of oscillations in each mode. Any one root of A gives a frequency and rate of decay
or buildup in one of these modes. As flutter is approached, the motions in the separate modes
may merge into a "flutter mode" in which the phase relation between 0 and Y is such as to
give zero net damping to the system. At the value of $ corrAsponding to the critical flutter
speed, the calculations give a pure imaginary root (X = joe). This indicates that an oscillation
at this frequency (w) once started will persist indefinitely without damping. The analog com-
puter has the advantage that it automatically combines the responses in the separate modes
according to the initial conditions imposed by a given disturbance.
Of course the actual rudder-hull system is much more complex than the ideal system
just discussed. The hull has not just one normal mode of vibration in which the rudder stocks
move normal to the longitudinal axis of the ship, but several. The rudders also may oscillate
in more than one mode since the rudder system may bend, twist, and shear. The flow to the
rudders is not actually uniform since they lie in the outflow jet of the propellers. The lift
and moment coefficients of the rudders may not be linear functions of the anglo of attack even
without the occurrence of cavitation. If cavitation occurs, nonlinearities may be expected to
be much more pronounced.
In spite of all these complications it appears possible to use this simplified flutter sys-
tem as u guide in forecasting the possibility of flutter of rudder-hull systems or of control
surface-hull systems in general.
The investigation of the rudder-excited hull vibration on the DD 931-Class destroyer by
the David Taylor Model Basin (as supplemented by experiments in the towing basin with an
apparatus built especially for this purpose) suggested that in the field of naval architecture,
as contrasted with the aircraft field, a "subcritical flutter" condition is more likely than un-
stable flutter.
The subcritical condition is revealed by the same basic analysis and is essentially
part of the same general phenomenon. It is illustrated by a curve such as d in Figure 14-3
when the minimum fails to dip below the axis. Then the damping of the overall system is
reduced to an undesirable level by the condition which produces critical flutter, but

14-6
 0.5 Legend -
j Curve L mh
a ----- 2.8 4.3
- 0.4 2.. 2.0 -- -

4.3
42.0 -- - -d
1.0 0 4,3

0'.2 _, _

__'0.1 --

I -0.1-.. . .. . . . . .

-0.2 --
"0 2 4 6 8 1012 14 16 is 20
Speed (S) in knots

Figure 14-3 - Variation with Speed S in the Rates of Decay or Buildup of Possible
Modes of Vibration of the Ideal System Shuwn in Figure 14-2 According to
Equations [14-3] and [14-4]
Negative values of the damping ratio indicate oscillatory instability (critical flutter).

oscillatory instability is not actually produced. The objectionable feature of the situation
arises from the fact that there are always other sources of disturbance on a ship underway.
Hence, whenever the system is disturbed as by a wave impact, the vibrations in this par-
ticular mode of the rudder-hull system decay at a very slow rate. The other modes retain
their normal damping characteristic.
Further discussion of the phenomenon of control surface flutter will be found in Ref-
erences 14-4, 14-5, 14-6, and 14-10. The calculation of hull modes when the rudders are
treated as flexibly supported is discussed in Reference 14-11.

C. THE SINGING PROPELLER

The problem of the singing propeller is not usually considered a hull vibration problem,
but it is certainly a hydroelastic problem and one with which a general naval architect may
have to contend; see Reference 14-1.
The phenomenon has been investigated abroad both from t.he hydrofoil flutter point of
view and from the vortex shedding point of view. The former is discussed in Reference 14-7,
the latter in Reference 14-12. Much support for the latter point of view is furnished by the
fact that critical speeds for singing can frequently be boosted beyond the top operating speed
of the propeller shaft by sharpening the trailing edges of the propellers.

i4-7
 In the case of commercial ships, the chief objection to the singing propeller is the an-
noyance of the high frequency vibration to human ears rather than the absolute level of vibra-
tion amplitude produced. It is well-known that the levels of acoustic power that are intoler-
able from the physiological standpoint are extremely low from the mechanical vibration
standpoint. This means that singing propellers may become intolerable because of noise at
amplitudes far below those at which they are in danger of structural failure.

D. COMMENTARY
The naval architect will become increasingly aware of the science of hydroelasticity
in the future. This will require an acquaintance not only with the subject of mechanical vi-
bration but also with the hydrodynamics of flow about streamlined and unstreamlined forms.
Only the barest outline of the subject of hydroolasticity has been sketched in this chapter.
The subject is but little understood at this time. When one surveys the extensive literature
available in aeroelasticity, the amount of literature on hydroelasticity published so far ap-
pears meager indeed.
Studies in hydroelasticity up to the present time indicate that there are large discrep-
ancies between experimental results and analytical predictions based on the classical theory
used in aeroelasticity. This is emphasized in the discussions of Reference 14-9. The phys-
ical magnitudes involved in the two fields are, of course, quite different, and thus it might be
expected that effects that can be neglected in one field could play a dominant role in the
other.

In spite of the present lack of information in the field of hydroelasticity, certain areas
in which hydroelastic problems may be expected in the future can be pointed out here.
Control surface flutter has been discussed in this chapter; for surface ships the only
control surface members are the rudders, except when activated fin stabilizers are used.
However, should the submarine become a commercial ship, its diving planes are clearly sub-
ject to the same flutter mechanism as the rudders. Its periscopes, if it retains them, will of
course be subject to vortex shedding excitation, as will, in fact, any protruding member not
properly streamlined. Antipitching fins will introduoe hydruelatisic ot.Ls,.
The designer of hydrofoil boats will be concerned not only with the static stability of
the hydrofoils but also with their dynamic stability. Thus, they will warrant a hydroelastic
analysis as detailed as the aeroelastic analyses currently made in aircraft design.
For a discussion of panel flutter and further details of hydroelasticity, the reader is
referred to 14-13, 14-14, and 14-15. The classical treatment of flutter as used in aeroelastic-
ity is given in Reference 14-16. Hydrofoil flutter is discussed in Reference 14-17.

14-8
 REFERENCES

14- 1. Saunders, H.E., "Hydrodynamics in Ship Design," published by SNAME (1957).

14- 2. Haller, S.R., Jr. and Abramson, H.N., "Hydroelasticity: A New Naval Science,"
Journal ASNE (May 1959).
14- 3. Flax, A.H., "Aero-Hydro-Elasticity," Structural Mechanics, Proceedings of First
Symposium on Naval Structural Mechanics, Pergamon Press, New York (1960).
14- 4. Bisplinghoff, R.L., et al., "Aeroelasticity," Addison-Wesley Publishing Co. (1955).
14- 5. Fung, Y.C., "An Introduction to the Theory of Aeroelasticity," John Wiley
& Sons, New York (1955).
14- 6. Scanlan, R.H. and Rosenbaum, R., "Introduction to the Study of Aircraft Vibration
and Flutter," MacMillan Co., New York (1951).
14- 7. Taylor, J. Lockwood, "Propeller Blade Vibration," TINA (1943).
14- 8. Wilson, W. Ker, "A Review of Ship Vibration Problems," Mar Engr and Nay Arch
(Jul, Aug, Oct, Nov 1955). Also published in ASNE Journal (May 1954).
14- 9. McGoldrick, R.T., "Rudder-Excited H-ll Vibration on USS FORREST SHERMAN
(DD931) - A Problem in Hydroelasticity," Trans. SNAME (1959). Also TMB Report 1431
(Jun 1960).
14-10. McGoldrick, R.T. and Jewell, D.A., "A Control-Surface Flutter Study in the Field
of Naval Architecture," TMB Report 1222 (Sep 1959).
14-11. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).

14-12. Krivtsov, Y.V. and Pernik, AJ., "The Singing of Propellers," TMB Translation
281 (Oct 1958).
14-13. Abramson, H.N. and Chu, W.H., "Problems of HydroelasiLcity," JSR, Vol. 3,
No. I (Jun 1959).

14-14. Henry, C.J., et al., "Aeroelastic Stability of Lifting Surfaces in High-Density

Fluids," JSR (Mar 1959).
14-15. Jewell, D.A., "A Note on Hydroelasticity," JSR, Vol. 3, No. 4 (Mar 1960).
14-16. Theodorsen, T., "General Theory of Aerodynamic Instability and the Mechanism
of Flutter," NACA Report 496 (1984).
14-17. Abramson, H.N. and Chu, W.H., "A Discussion of the Flutter of Submerged
Hydrofoils." JSR, Vol. 3, No. 2 (Oct 1959).

14-9
 CHAPTER 15
SHIP VIBRATION RESEARCH

A. INTRODUCTION
Ship vibration research was initiated toward the end of the nineteenth century. Credit
for pioneering is usually given to 0. Schlick who proposed the now famous Schlick formula
for the frequency of the 2-node flexural mode of surface ships in 1894. Among other early
investigators was A. N. Krylov who reported on the recording of ship vibration in 1900.
Although the analogy to string vibration cited therein was unsound, it is of historical interest
that the subject was also discussed in a French treatise on naval arcbitecture published in
1894; see Reference 15-1.
As in so many other fields of research, progress has been slow but at an accelerating
pace, and the development of theory has proceeded hand in hand with the development of
experimental techniques. This is a field in which full-scale experimentation is costly. The
mere tying up of a ship for the length of time required for vibration experiments is often pro-
hibitive, to say nothing of the cost of installing and operating the necessary equipment. How-
ever, it must be emphasized that the correction of vibration unexpectedly encountered on the
trials of a new ship may be much more costly.
It is not attempted in this chapter to survey the ship vibration research work which is
now carried on by all the principal maritime nations abroad. This disoussion iq confined
chiefly to the ship vibration research in the United States which, for the most part, has been
carried on by the United States Navy either independently or in collaboration with the Society
of Naval Architects and Marine Engineers.
On the experimental side, this research has involved the following principal phases:
(a) The development of instrumentation for recording vibration of ships and ship models.

(b) The development of machines capable of vibrating ships and ship models.
(c) The development of apparatus for determining propeller-exciting forces on model scale.

(d) The conduct of full-scale experiments with vibration generators to determine the vibra-
tory characteristics of ships in service.
(e) Running of vibration surveys on ships in service and the systematic storage of data
so obtained.
(f) The conduct of model experiments for determining propeller-exciting forces.

(g) Model experiments for the determination of the added mass effect, of the surrounding
water.

15-1
 On the theoretical side, ship vibration research in the United States has involved:
(a) The development of analytical methods of predicting the normal modes and natural
frequencies of vibration of ships.
(b) The development of analytical methods of predicting the steady-state forced vibration
of a ship under given exciting forces.
(c) The development of analytical methods of calculating transient vibrations within the
elastic range.
(d) The development of analytical methods of predicting the vibratory exciting forces
which will act on a ship of given design.
(o) The development of computing techniques, either analog or digital, for carrying out
calculations by the analytical methods developed.
The research program, of course, has also included the comparison of analytical pre-
dictions with experimental results and the evaluation of the analytical methods on the basis
of these comparisons. Most of the theoretical research accomplishments have been covered
in previous chapters. Hence, the present chapter is devoted chiefly to the experimental phase
of ship vibration research and to the attempts to correlate theory with experiment.,

B. VIBRATION GENERATORS

The development of vibration generators for ship vibration research in the United States
.piuL Jevlopment of such machines for research on bridges by the
firm of Losenhausen in Dusseldorf, Germany, prior to World War II. Their successful use in
Germany led the U.S. Experimental Model Basin to purchase from Losenhausen in 1931 the
largest machine of this type when built. These generators consist of eccentric masses which
may be so unbalanced and so phased as to produce sinusoidal forces in one direction only,
sufficient to maintain a vibration of the entire hull of a magnitude permitting experimental
determinatibn of the normal-mode pattern. The machine had a force amplitude of 49,000 lb
and a deadweight of 25 tons. Its two eccentrics each weighed 6000 lb and could be offset
up to about 12 in. It was operated by two d-c motors, each of 15-kw capacity.
There was also purchased from Losenhausen at the same time a very small vibration
generator, weighing only about 140 lb, and having eccentric weights at each end of two
parallel shafts. This machine was not only suitable for vibration experiments on model scale
but was also capable of exciting local structures on full-scale ships and was used repeatedly
for many years. As originally furnished, it had a maximum force rating of 440 lb and a maxi-
mum speed of 3600 rpm.
The David Taylor Model Basin later undertook the development of vibration generators
for full-scale ship experiments of greater versatility than possible with the huge machine
originally purchased from Losenhausen. The Taylor Model Basin medium vibration generator

15-2
 Figure 15-1 - Ta-ylor Model Basin Medium Vibration Generator

Figure 15-2 -Taylor Model Basin Three-Mass 40,000-Pound Vibration Generator

shown in Figure 15-1 and the Model Basin three-mass 40,000-lb vibration generator shown in
Figure 15-2 were results of this effort.
The latter machine (shown in Figure 15-2) makes use of a schemio patented by Dr. R.
K. Bernhard who had been associated with the design of the large Losenhausen machine in
Germany. It has three parallel shafts; the two outer shafts rotate in one direction while the
inner rotates in the opposite direction. This permits the production of a sinusoidal exciting
force in any desired direction in a plane normal to the shaft axes.
The operation of large vibration generators on full-scale ships has been greatly facili-
tated by the progress that has been made since their inception in the development of speed.
control devices for direct-current motors. The difficulty in operation is due to the power
versus speed characteristic that must be met by a machine for this service. If there were no
damping, no power output would be required to maintain a constant amplitude of vibration in
a mass-elastic system such as a ship. Actually, even with damping, very little power is
absorbed except in the neighborhood of resonance. Since, on approaching a resonance speed
from below, the power required will increase, the speed is stable on the low speed side of
resonance. Thus, any change in line voltage which would tend to increase the speed will
also tend to load the motors which will limit the speed. Above the resonance speed, however,
the situation is reversed. As the speed increases, the motors are unloaded by the drop in
amplitude of the system and hence tend to run away. An amplidyne-type of speed control is
therefore beneficial.
For studying the vibratory characteristics of hulls, it is a prime requirement to main-
tain a selected speed of the'vibration generator. If this speed can be maintained the hull
may then be explored with portable instruments. This permits location of the nodal points
and the plotting of normal mode patterns. It also facilitates the step-by-step plotting of
cur'ves of displacement amplitude, velocity amplitude, acceleration amplitude, or power versus
speed.
Vibration generators for shipboard use must be able to produce not only sinusoidal
forces in either the vertical or horizontal directions but also couples about the longitudinal
axis for investigating torsional vibration of hulls. Some progress has been made in recent
years in investigating the coupling of hull torsion with horizontal hull flexure, but much re-
mains to be done. It is very difficult to design a single machine to produce a pure couple
large enough to excite the torsional modes of the hull. Proposals have been made to use
separate units synchronously driven, one installed on each side of the hull near the stern.
With large machines, however, the synchronizing becomes a problem comparable with that
of synchronizing propeller shafts. The design of the Taylor Model Basin three-mass, 40,000-
lb vibration generator is a compromise as far as pure couples are concerned. The maximum
couple attainable is 120,000 lb-ft.
Further details of vibration machines used by the Taylor Model Basin will be found in
Reference 15-2.

15-4
 C. SHIP VIBRATION INSTRUMENTS
Certain salient points in the design of vibration instruments for shipboard use are
discussed here. For further details, the reader should see Reference 15-8. The basic prin-
ciples used in the design of vibration instruments in general are discussed in standard texts
on mechanical vibration; e.g., References 15-4 and 15-5. It must be pointed out in the beginning that
instruments that perform satisfactorily in making a shipboard investigation with a vibration
generator may be quite inoperable at conditions under which underway ship vibration surveys
must be made. Vibration generator tests normally can be undertaken only under very calm
sea conditions, and with the ship either deaa in the water or advancing at a very slow speed.
The chief difficulty in the design and operation of vibration instruments for underway
vibration surveys arises from the fact that the rigid body motions of the ship are extremely
large in proportion to the displacement amplitudes in vibration which are to be measured and
the accompanying accelerations itre high. There is not only this problem to contend with but
also the fact that the vibratory amplitudes themselves cover a very large range, roughly from
a single amplitude of 0.001 in. to 1.0 in. This, of course, does not refer Lo attempts to meas-
ure vibration under slamming conditions in a seaway where amplitudes of a foot or more are
occasionally encountered.
As may bA seen from Reference 15-8, all the well-known types of vibration instruments
have been tried on board ship. In spite of all the effort expended in this direction, stand-
ardization of instrumentation for underway vibration surveys has not been attained up to the
present time. However, both the Bureau of Ships of the Navy Department and the Society of
Naval Architects and Marine Engineers are continuing their efforts in this area.
The principal types of instruments now in use are:
(a) Amplitude indicating: dial gage-type vibrometers, optical vibrometers, Cordero vibro-
meters, velocity- or accelerometer-type transducers with integrating amplifiers, cathode ray
or recording oscillographs, and rectifiers with d-c indicating meters.
(b) Frequency indicating: sets of Frahms reeds, single tunable reed, stroboscopes,
oscilloscope, and direct-indicating electronic frequency meter.
(c) Amplitude and frequency recording: mechanical pallographs, crystal accelerometers
with integrating amplifiers and oscillograph, velocity pickups with integrating amplifiers and
oscillograph, transducers with frequency-modulation tape recording using playback, optical
recording for film (mirragraph), and optical accelerometers.
(d) Phase measurement: electronic indicating and electronic recording.
A typical direct-recording instrument for shipboard use is illustrated in Figure 15-3
and a typical remote.recording system in Figure 15-4. In the former case, the instrument has
two seismic elements; one can move only in the vertical direction, the other only in the
horizontal direction. Each element is suspended in such a way as to have a natural frequency

15-5
 "2o.

Figure 15-3 - TMB Two-Component Pallograph

Direct Writing
Integrating Amplifier Oscillograph

To Other Stations 0 0
Velocity Pickup (electromagnetic Instruments at central recording

At remote station on ship) station on ship

Figure 15-4 - Typical Remote Recording System for Shipboard Vibration Measurements

considerably lowor than the vibration frequencies normally to be measured. Under such circum-
stances-the element remains practica!ly fixed in space whereas the rest of the instrument
vibrates with the deck of the ship on which it is placed. The relative motion between the
element and the base of the instrument is then magnified by a linkage and recorded on paper
together with a timing signal. It is shown in Reference 15-3, however, that, by introducing
controllable damping and with adequate calibration techniques, such instruments can be used
at ranges of frequencies within which their own natural frequencies fail.

15-6
 In the remote recording system shown in Figure 15-4, the pickup unit also contains
a seismic element. In this instance the element is a magnet whose field cuts across a coil
S: fixed to the base of the pickup. The voltage generated in the coil is proportional to the
relative velocity between the base and the element. An integrating amplifier converts the
signal to one proportional to vibratory displacement which is recorded on the oscillograph.
If desired, the velocity signal can be recorded directly without integration.
In general, instrumentation satisfactory for underway vibration surveys can be used
for vibration-generator surveys but the reverse does not hold. The limitation, if any, will
be in its sensifivitieVs and frequency ranges. However, although the earlier i,, nf viibrftion
generators on board ship was chiefly for determining the natural frequencies of the hull and
its normal mode patterns, recently it has been found practical to use such machines for de-
termining forced vibratory response characteristics over the range of blade frequencies on-
countered in the normal operation of the ship.
Under very calm water conditions and with very sensitive vibration-recording equip-
ment, vibration-generator surveys may be made with amazingly low amplitudes, as in the
case of PONTCHARTRAIN1' 5 ' 6 in which normal modes were plotted with amplitudes of the
order of .0.01 rpil.
The investigation of torsion-bending modes and pure torsional modes of hulls requires
special attention to instrumentation. Here it is required to separate the rotational and trans-
lational components of the hull vibration. This can be done with translational instruments
only by simultaneous recording of signals from pickups at opposite deck edges and obser-
vation of phase. A rocording angular accelerometer with integrating amplifier will give the
angular amplitude directly if its sensitivity to rectilinear vibration is sufficiently low.

D. EXPERIMENTAL TECHNIQUES

1. MODEL

Considerable model experimentation has been carried out in the United States in the
effort to clarify the added mass effect of the surrounding water on hull vibration character-
istics; see Reference 15-7. At this writing, however, the effort on model scale in the field
of ship vibration research is concentrated on the problem of determining model forces. As
stated in Chapter 7, this work is an extension of the pioneering work of F.M. Lewis.
Ia all model work one encounters the problem of similitude or scaling. This subject
is discussed further in Appendix G. It should be noted hero that it is not feasible (because
of conflicting scaling rules; see Appendix G) to build a complete self-propelled dynamic
model of a ship from which the propeller-excited vibration of the prototype can be predicted
from direct measurements of the vibration of the model. Quito aside from the difficulties of
fabricating structural models of ships, the research scientist is confronted with scaling rules
for ship vibration that do not conform with those for ship propulsion.

15-7
 In F.M. Lewis' original scheme, a continuous wooden model was used for the propeller-
exciting force determination and the line of reasoning was that, since the determination was
made by nullification of the forces by forces produced by a balancing machine inside of the
model, the dynamic characteristics of the wooden model itself were irrelevnt,t
In later experiments it was found that resonarnce of the wooden model was deleterious
and a scheme of flexibly suspending the stern portion from the rest of the model was adopted.
As shown in dotail in Reference 15-8, the present setup yields the following values:
a. Blade-frequency vertical force,

b. Blade-frequency horizontal force, and

c. Blade-frequency torsional moment about the longitudinal axis.
In the field of hydroelasticity (see Chapter 14), the question of the feasibility of model
experimentation is quite important. It must be decided whether it is feasible to predict a con-
trol surface flutter condition by model techniques. As shown in Reference 15-9, flutter model
testing has been found practical in aeronautical engineering. In naval architecture, however,
its use will have to be much more restricted. The criterion of similitude in flutter experiments
is the nondimensional Strouhal number or "reduced frequency." In aeroelasticity, this is

commonly expressed in the notation -w-, where b is the half-chord dimension of a wing, W,
is the circular frequency of the vibration, and V is the velocity of the wing relative to the
undisturbed air. This relation indicates that for a scaled model of the same material as the
prototype, the flutter speed should be the same as for the prototype.
In the hydroelastic field, if a dynamical model were manufactured of the same material
as the ship (since the natural frequencies vary inversely as the scale), the numerator in the
expression for the Strouhal number would be the same for model and prototype. Hence, the
Strouhai number itself would be the same if the model speed were the same as the speed of
the prototype. Obviously, for surface craft the flow patterns will not be similar if model and
prototype run at the same speed since the law for similitude of bow waves is the Froude law;
that is,
V2
- constant
Lg
where V is the velocity,
L is a characteristic length, and
g is the acceleration of gravity.
Although it has been pointed out that in the case of a deeply submerged submarine
model the flow pattern is independent of the speed unless flow separation develops, there
remains the problem of fabricating the dynamical model and of running it at such a high speed.
In the aircraft field, flutter model testing is well established and here the high speed re-
quired is feasible. It is also possible in wind tunnels to vary the density to reduce

15-8
discrepancies between model and full-scale Reynolds numbers. Furthermore, as pointed out
in Appendix (, in the field of ship slamming it is feasible to experiment with dynamic models
based on Froude's law since here the propulsion system itself is not involved. In such models
the elastic characteristics are not represented in the shell but in a special girder installed
along the longitudinal axis of the model.
In dealing with vibration problems of local structures it is easier to satisfy the con-
ditions of similitude than when the entire hull is involved, since hydrodynamic forces are
not involved. Thus, similitude for vibratory characteristics is theoretically satisfied if the
model is made of the same material as the prototype and duplicates it to scale. The practical
dirticulties here lie in the tabrication ot the model and in the simulation of boundtary re-
straints and damping characteristics. These difficulties, however, are so great that compara-
tively little has been accomplished to date in solving vibration problems of local hull struc-
tures by dynamic model experiments.

Z FULL SCALE
Full-scale experiments in the field of ship vibration are at present confined chiefly
to tests with mechanical vibration generators. These machines are installed at either end
of the ship and operated over a sufficient range of speed and driving force to excite the nor-
mal modes of vibration of the hull, one at a time. During the operation of the vibration gen-
erator, the vibration of the hull is recorded either by a multichannel oscillograph system or
tape recording system, or by exploring the hull with portable vibration instruments. The aim
is usually to record enough data to permit plotting the normal mode patterns for all modes of
the hull that can be excited. This includes the vertical, horizontal, torsional, or torsion-
bending modes. It is also possible in such tests to determine phase relations among the
displacements at stations along the hull. These show whether a standing or traveling wave
condition exists.
A useful adjunct to the vibration generator test is the anchor drop test. While such a
test usually excites only the 2-node vertical flexural mode of the hull, the recording of the
subsequent vibration permits a determination of the logarithmic decrement; that is, the damp.
ing. Furthermore, since the 2-node vertical flexural is the mode of the hull lowest in frequency,
it may be found in some cases that the vibration generator cannot produce sufficient force to
excite this mode even when the machiLd is set at its maximum eccentricity. The anchor drop
may then be the only means of determining the fundamental vertical frequency.
Although used less often for this purpose in the past, the vibration generator, if in-
stalled directly over the propellers, furnishes a means of determining the full-scale propeller-
exciting forces. Conceivably, it could be used as a balancing device to cancel out the pro-
peller forces when the ship is under way in the same manner as in the model determination
of propeller-exciting forces by the null method. This technique has not been used full scale

15-9
to any extent up to the present time, and synchronizing of the vibration generator with the
ship's shafts is not easily attained. However, if with the ship dead in the water, the vibra-
tion generator is operated at speeds (rpm) coinciding with operating blade frequencies (cpm),
the data so obtained may be used to estimate the underway propeller forces, provided the
underway hull vibration is also recorded. This, of course, involves the assumption that the
damping and added mass characteristics of the hull are the same underway as in still water.
In the case of vertical vibration, only the hydrofoil action of the propellers appears to cause
a variation in damping. In the case of horizontal vibration, the hydrofoil damping action of
the rudders is to be taken into account. In the case of submarines, the diving planes provide
a source of damping of vertical vibration. It should be noted, however, that where the opera-
ting blade frequency falls above the range of significant hull mode frequencies, the mechanical
impedance at the stern is chiefly an inertial impedance.
Aside from the full-scale vibration-generator experiments just described, both the U.S.
Navy and the Society of Naval Architects ahd Marine Engineers have initiated a systematic
series of underway vibration surveys on new classes of ships. Such surveys play an impor-
tant role in the overall ship vibration research program. Even if nothing unusual develops
during such a survey from the ship vibration standpoint, the data obtained are still extremely
valuable in establishing acceptable levels of vibration for various classes of ships. If a
severe vibratory condition is discovered, the investigation of this is likely to lead to a def-
inite advance in ship vibration theory. It is never desirable from the naval architect's point
of view to dismiss a case of serious ship vibration which has been corrected by the trial and
error process without really tracking down the phenomenon. Unfortunately, however, economic
considerations sometimes impose such an outcome.
With the advent of the application of statistical methods to the study of the performance
of ships in a seaway (see Reference 15-10), more attention has been given in recent years to
the collection of data on the transient vibrations of ships. As pointed out in Chapter 2, these
vibrations are much more important from the point of view of structural integrity of the hull
than the steady-state vibrations which are excited by the propeller.
While, in principle, the instruments already mentioned in this chapter can be used for
recording transient vibrations, it is in order to point out here that the severe sea conditions
undet which large transient hull vibrations are generated rule out an instrument of the type
shown in Figure 15-8 for such a purpose. The trend at this time is in the direction of remote-
recording systems using magnetic tape recording. This permits subsequent playback of the
record into vibration analyzers in the laboratory.

E. CORRELATING THEORY AND EXPERIMENT

The complexity of the ship vibration problem is such that exact theories are nonexist-
ent. While the theory of vibration of ideal mass-elastic systems can serve as a guide, it is
only by continuaiiy comparing theoretical predictions with experimental results that a rational
theory can be established.
15-10
 The theory advocated in this book is a beam theory with important modifications. The
idealized beam is considered to be free in space, but it carries added mass noinuniformly dis-
tributed to account for the inertia effect of the surrounding water. Thus it is nonuniform in
both its mass and rigidity diatributions along its length. It can bend, twist, and shear, but
does not execute significant longitudinal vibration and its torsion is only about its longitudinal
axis. If longitudinal vibrations of surface vessels, and especially submarines, are "signif-
icant" from some points of view, this arises from considerations not dealt with in this book
and involves amplitudes of a lower order of magnitude. The damping of the idealized system
-- .. u.uu viuuud daariping between itseif and axes fixed in space.
It may have one or several flexibly attached masses and these may move relative to it in both
rotation and translation. Finally, it has the unique, property (for a beam) that it can vibrate
in only a limited number of modes.
In checking this theory, the simplest criterion is the comparison between predicted and
measured natural frequencies. Such a comparison requires merely tabulating the frequencies
and identifying them by the node shapes, then listing the predicted frequencies and the per-
centage of errors in the predictions.
The second best criterion is the comparison of the experimental and computed normal
mode patterns. Even if the frequencies check, the theory will have less and less validity,
the greater the departure of the experimentally determined normal mode patterns from those
predicted. In this connection, beanlike modal patterns are not to be expected if there is
considerable local flexibility, especially when the local structure is massive. Hence, when
the observed modal patterns are not beamlike, calculations must be tried with various local
structures treated as flexibly attached, if correlation is to be improved.
The third area to be explored is that of the forced vibratory response of the hull. Here
much depends on whether the vibration is resonant or nonresonant. The forced resonant re-
sponse depends chiefly on the damping and the magnitude of the driving force. Since the
latter is known when a vibration generator is used, the forced resonant response in thiis
case yields an experimental value of damping.
In the range of frequencies above that of the significant hull modes, the observed forced
response in the stern provides a check on the proposed formula for mechanical impedance at
the stern of a .... p discut... n C;pn . I.
If the experimentally determined horizontal modes are observed to be torsion-free, or
very nearly so, then the same criteria for comparing theory and experiment in vertical vibra-
tion are applicable to horizontal vibration. When it comes to torsion-bending modes, it is
necessary to convert the mechanical system to a generalized system which combines the
effects of translation and rotation before the comparison can be made. The general scheme
used for doing thig is outinnd in Chapter 4.

15-11
 REFERENCES
15-1. Pollard, J. and Dudebout, A.:,"Theorie du Navire,"-Treizieme Pattie, "Vibrations
des Coques des Navires a He!.ce," Vol. IV, Chapt. LXIX, "Causes et Periodes des Vibra-
tions," Paris (1894).
15-2. Robinson, Q.R., "Vibration Machines at the David W. Taylor Model Basin," TMB
Report 821 (Jul 1952).
15-8. Allnutt, R.B. and Mintz, F., " Instruments at the David Taylor Model Basin for
Measuring Vibration and Shook on Ship Structures and Machinery,',' TMB Report 563 (Jul 1948).
15-4. Den Hartog. J.P.. "Meehanal Vibratinps," U91 .r.a..-illBook C, York..)
N.owJ..
15-5. Timoshenko, S., "Vibration Problems in Engineering,"' D. Van Nostrand Co., New
York (1955).
15-6. Allnutt, R.B., "Investigation of Hull Vibrations of USCGC PONTCHARTRAIN (WPG
70),"' TMB Report R-294 (Aug 1946).
15-7. Kaplan, P., "A Study of the Virtual Mass Associated with the Vertical Vibration of
Ships in Water," SIT, Davidson Lab Report 734 (Dec 1959).
15-8. Stuntz, G.R., et al., "Series 60 - The Effect of Variations in Afterbody Shape upon
Resistance, Power, Wake Distribution, and Propeller-Excited Vibratory Forces," Paper pre-
pared for Annual Meeting of SNAME (1960).
15-9. Bispliughoff, R.L., et al., "Aeroelasticity," Addison-Wesley Publishing Co. (1955).
15-10. Jasper, N.H., "Statistical Distribution Patterns of Ocean Waves and of Wave-Induced
Ship Stresses and Motions, -with Engineering Applications," Trans. SNAME, Vol. 64 (1956).
Also TMB Report 921 (Oct 1957).

15-12
 APPENDIX A

ILLUSTRATION OF A CALCULATION OF A VERTICAL

NORMAL MODE BY THE DIGITAL PROCESS

It is essential to point out here that the illustrations used in this or other appendixes
of this book are not to be construed as samples for use in preparing data for calculations to
be made by the Applied Mathematics Laboratory of the David Taylor Model Basin. The prob-
lems that have been coded for solution by that Laboratory are continually being studied for
improvement of the computing technique as new and improved corinputing racilities are installed.
Although the preparation of data for such a calculation as given in this Appendix is dis-
cussed in considerable detail in Reference A-1, the reader is urged to check directly with the
Applied Mathematics Laboratory for the latest instructions if a request for hull vibration cal-
culations is to be made of the Taylor Model Basin.
One of the simpler hull vibration cases (the calculation of the 2-node vertical mode of
a commercial cargo ship) is used for illustration in this Appendix to emphasize the basic el-
ements of the method. For the application of the digital method to the calculation of torsion-
bending modqs, including consideration of the effect of sprung inertias; see Reference A-1.
This illustration is based on the Maritime C-4 design cargo ship GOPHER MARINER whose
vibration characteristics were the subject of Reference A-2. The design of this class of Ship
is discussed in detail in Reference A-8.
The method is based on the finite difference equations given in their simplest form in
Chapter 3. The orientation of rectangular coordinate axes for the calculation is shown in
Figure A-1. It was assumed in this case that both i and a were zero because of the symmetry
of the ship with respect to the XY-plane.

Z 0

AP

I i I i I 1 I I
0 I 2 3 4 5 6 7 8 9 1O II 12 13 14 15 1 17 1 19 20

Figure A-i - Diagram Showing the Divizio-r:.f;A#" Ship lutc Z, .Z:uaf

Sections for the Vibration Calculations

A-i
 The hull was divided into 20 sections between the after and forward perpendiculars.
The ship was visualized as shown in Figure A-1; that is, with the stern to the left. It should
be noted that the station numbering here is from the after perpendicular forward for the reason
stated in Chapter 8. As this numbering system is purely arbitrary, the reader should check on
the numbering system in use at the time if a future request for calculations by the Taylor
Model Basin is contemplated.
The weight corresponding to the added mass of the surrounding water was added to the
weight of the vessel for the test loading condition. The weight added to each of the 20 sec-
tions was estimated from the formula:

I
Added weight per unit length - JCOrpb2 [A-1i
2

where J is the longitudinal coefficient applied to correct for the departure of the vibratory
motion of the water from two-dimensional flow in planes parallel to YZ and the C coefficients
are factors, based on two-dimensional flow, which give the ratio of the added mass for a
cylindrical form having the shape of a ship section to the added mass for a circular cylindrical
section of the same beam.
The C coefficients were obtained as follows:
(a) The section-area coefficient P for the section was estimated by inspection (P is the
ratio of the area of the underwater section to 2 bd, where d is the draft and b is the mean
half-breadth of the section at the waterline, in ft).
(b) Bid for the section was then computed (where B is the whole beam at the section).
(c) The value of C corresponding to these values of P and B/d was found from Figure A-2.
The coefficient J was obtained from Figure A-a after computing L/B (length over beam)
for the ship.

Combined Mass, mn
The values of the combined mass in used in the digital computation were derived as
follows: A continuous curve of combined weight per unit length was plotted as in Figure A-4
by adding to the ordinates of the weight curve of the ship the values of added weight of water.
The lumped masses used in the digital oalculation were then derived by concentrating the
weight indicated by the combined curve at the 20 stations after converting weight to mass by
dividing by g (the acceleration of gravity).

Area Moment of Inertia, I

The area moments of inertia were evaluated for a sufficient number of sections of the
hull to permit the plotting of a continuoue e,'.,. All the longitudinal structural members up
tn the uppermost continuous deck or weather deck were considered in the evaluation. If a
ship has decks with expansion joints, which hence are not designed to carry hull bending

A42
 4 5 6 7 8 9 10I i 12 L/B

0.9 0..00-- i -,,.oo

" 0.90

-- -
0.98 0.80

0.69

0.65 ,, 0.70 -- /

0.5 0.00
0.60 -
B/d
Figure A-2 - Curves for Estimating the Figure A-3 - Curve for Estimating the
Coefficient C Used in the Added Coefficient J Used in the Added
Mass Evaluation Mass Evaluation
(From Reference A-4) (From Reference A-4)

stresses, they should not be included.

Superstructures extending less than ee 1 1
80 percent of the length of the hull may 100
be completely omitted in computing the -- I
w igh- I
b Vil
area moment of inertia of sections, as
I 0
was done in this case. With longer
superstructures their contribution - 40 - -
should be neglected at. their ends but 20 -

added gradually so as to make them 2/ 4 *

fully effective in their midlength, Stem Sttion S I aow4
are continuous
provided theytiedint dekthemai and well
stuctre.Figure A-4 - Weight Curves for GOPHER
tied into the main dock structure. MARINER (Heavy Displacement) Used
Longitudinal bulkheads or longi- for Calculation of Vertical Modes
tudinel stiffeners terminating at or near
tho section in question should not be included, Where hatch openings fall at the section, the
hatch coaming section area should be added.
IL is important to note that both the 1 and I curves approach Aero aL the ends of the ship.
While the extreme stations (0 and 20) are taken at the perpendiculars, so that there is a pro-
jection of the hull beyond these stations at each end, the 1 and I curves may be plotted to
give excessive values of p/I near the ends of the ship. Such a condition can cause considerable

A-3
error in the digital calculation of the natural frequencies because it effectively makes the
ends of the ship very "flabby." If the ship actually had such a construction, the lowest
mode would be determined by this local condition at the ends. On the other hand, no amount
of overestimating of I at the ends can have much effect on the natural frequency since it
simply means that the overhang at each end is rigidly attached to the ship at the perpendic-
ular. Therefore, the rule is to check at the ends to make sure that the value of AI/I does not
rise abruptly. This precaution was followed in this illustration.

8 0 00
6, F- I I -4----" ' i

4 "00 Figure A-5 - Area Moment of Inertia Curve

Used for Calculation of Vertical Modes of
2 4 6000 1- GOPHER MARINER
0 2 4 6 8 0 12 14 S 10 20
Sten $totion Bow

Shear Rigidity Factor, K

For GOPHER MARINER the shear rigidity factor was evaluated by letting KA - A',
where A is the total area used in the evaluation of 1, and A' is the area of the vertical plating
only (such as side shell plating
,hp and continuous longitudinal bulkhead plating). The "web"
area A'was plotted for as many sections as were available (in this case six) and a smooth
curve was drawn as shown in Figure A-6.

7/

g 4/ -- Figure A-6 - Side Plating Area Curve Used

*- -- - - -in Evaluating Vertical Shearing Rigidity
-- of GOPHER MARINER

o
Stern
2 4 6 6 10
Station
12. 1441 8 20
BOW

The values actually furnished for the digital calculation of the vertical modes ?of
GOPHER MARINER for a heavy displacement loading are given in Table A-1. In this case,
the term for rotary inertia was omitted on the assumption that its effect would be negligible.
The reader is cautioned against using Table A-1 as a sample for furnishing data to the

A-4
 TABLE A-1

"InitialData Furnished for the Digital Calculation of Vertical Modes

of GOPHER MARINER for Heavy Loading Condition

x 10a Aw
El x 106
2 1 KAG
ton-sec
n ft ton-ft ft/ton

A AO
7.8 n 310 11 r

1 13.97 0.5205 8.56

2 22.31 0.3615 7.17
3 38.76 0.2880 6.75
4 56.51 0.2455 5.72
5 73.04 0.2185 5.67
6 69.14 0.2005 6.08
7 75.89 0.1880 6.70
8 85.47 0.1800 7.23
9 90.79 0.1760 7.62
10 95.11 0.1750 7.73
11 80.35 0.1750 7.63
12 75.76 0.1790 7.37
13 67.02 0.1885 7.05
14 50.61 0.2040 6.73
15 41.14 0.2320 6.33
16 25.77 0.2740 5.98
17 26.61 0.3270 5.52
18 23.80 0.3805 5.06
19 14.94 0.5265 4.96
20 5.53 0.3240 -

A-- 26.25 ft
L - 525 ft Displacement: 16,840 tons
B = 76.0 ft
D- 44 ft 6 in.

Applied Mathematics Laboratory of the David Taylor Model Basin for future hull vibration
calculations.The slightest changes made in recoding the problem for a new computing
machine may make it impossible for the machine to carry out the computation with the
initial data as presented, and current Applied Mathematics Laboratory instructions must be
followed,
The results obtained from the digital calculation of the 2-node vertical mode of
GOPHER MARIN4ER for the heavy displacement are given in Table A-2.
It wil! be noted that, although the calculation was made by means of the coding for
torsion-bending modes, and, since in this case the terms coupling flexure and torsion were
zero, the columns in Table A-2 for angular displacement 9S and twisting moment T are both
z ere.

A-5
 TABLE A-2

Data Furnished by Digital Computer for 2-Node

Vertical Mode of GOPHER MARINER

Station
No. VY
7 H

0 -1.00000 0.00641 0 0 0 0
1 -0.82620 0.00641 0 - 473 12,426 0
2 -0.49_17 0.0035 -12 - 44,02b 0
3 -0.47141 0.00619 0 -2120 - 99,686 0
4 -0.29428 0.00590 0 -3276 -185,700 0
5 -0.12648 0.00544 0 -4329 -299,341 0
6 0.02725 0.00479 0 -4913 -428,329 0
7 0.15974 0.00393 0 -4794 -554,186 0
8 0.26270 0.00289 0 -4027 -659,905 0
9 0.32635 0.00170 0 -2606 -728,322 0
10 0.34308 0.00042 0 - 731 -747,518 0
11 0.30958 -0.00088 0 1333 -712,506 0
12 0.23148 -0.00213 0 2908 -636,170 0
13 0.11606 -0.00326 0 4017 -530,701 0
14 -0.02779 -0.00426 0 4510 -412,310 0
15 -0.19169 -0.00511 0 4421 -296,255 0
16 -0.36871 -0.00579 0 3922 -193,302 0
17 -0.55466 -0.00632 0 3320 -106,134 0
18 -0.74303 -0.00667 0 2386 - 43,485 0
19 -0.92899 -0.00683 0 1267 - 10,214 0
20 -1.11188 -0.00689 0 389 0 0
20A -0.00689 0 0

The normal mode patterns of vertical displacement, angular displacement about a 2-

axis, shear force, and bending moment, for the 2-node vertical flexural mode of GOPHER
MARINER as determined by the digital calculation can then be plotted from Table A-2. The
absolute values given in this table have no significance. Only the relative magnitudes in a
normal mode pattern are indicated. It is to be noted, however, that the data were prepared in
the foot-ton-second system of units. Hence, although a single amplitude of 1 foot at the after
perpendicular would be an extremely large amplitude, the values of bending moment per foot
amplitude at this station are given in the column for N provided the hull is not deformed
beyond the elastic range. For smaller amplitudes the bending moments would be proportion-
ately smaller.
The normal mode pattern for displacement given by Table A-2 is plotted in Figure A-7
(with sign reversed).
For further details on the preparation of data for a calculation of this type, see Ref-
erence A-1.

A-6
Figure A-7 - Normal Mode Pattern for 2-Node
Vertical Flexural Mode of GOPHER 0
MARINER as Computed by Digital
Method for Heavy Displacement

0 2 4 8 $1 2 14MI!LIM=
AP Stations

REFERENCES

A-i. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including. Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).
A-2. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
A-3. Russo, V.L. and Sullivan, E.K., "Design of the MARINLER-Type Ship," Trans.
SNAME, Vol. 61 (1953).
A-4. Prohaska, C.W., "The Voetical Vibration of Ships," Shipbuilder, Vol. 54 (Oct, Nov
1947).

A-7
 APPENDIX B

ILLUSTRATION OF A CALCULATION OF TRANSIENT

RESPONSE BY THE DIGITAL METHOD

As in Appendix A, the calculation presented in this Appendix serves to illustrate the

basic principles involved, but it is not intended as a sample to be used in preparing requests
for transient hull vibration calculations to be carried out by the Applied Mathematics Labora-
tory of the David Taylor Model Basin. As a matter of fact, the recoding of this particular
problem with the addition of buoyancy forces is in progress at this tim,.
The analysis of transient vibration discussed in this book is confined to vibrations
within the elastic range, as pointed out in Chapter 5. For a discussion of the general prob-
lem of ship slamming in a seaway, see Reference B-1. The digital method of transient
analysis discussed in Chapter 5 was applied to GOPHER MARINER in the prototype example
used in coding the initial treatment of this problem on the UNIVAC. Further details of the
prototype calculation will be found in References B-2 and B-8.
The loading conditions for the calculation were as follows: a vertical force of 1 ton
was instantaneously applied amidships, held constant for 1 sec, then instantaneously re-
moved. No special significance is to be attached to the particular loading chosen for this
example, at least for simulating the loading encountered by a ship in a seaway. Mathemati-
cally, however, this is a standard type of forcing function sometimes described as a "rectan.
gular pulse load."
The reference axes were the same as indicated in Figure A-i, and an element of length
Ao is shown in Figure 5-2. The hull data used for this calculation were the same as used in
the calculation of the 2-node vertical normal mode given in Appendix A (same ship and same
displacement in both cases) with the exception that the parameters iMZ (rotary inertia) and
C (damping) were added. These additional parameters are given in Table B-1.
The rotary inertia values I., were estimated in this case by treating each lumped mass
m (which includes a value for added mass of water) as though it were uniformly distributed
throughout a rectangular parallelepiped bounded by the main deck, the bottom plating, -and the
hull side plating, and taking the mass moment of inertia of this parallelepiped with respect
to a Z-axis passing through its centroid.
The damping values C were based on a value of i for the ratio c/A. The experimental
basis for this value is discussed in Chapter 8. In this calculation the time steps selected
(At) were 0.02 sec. Hence, the load data was 1 ton at Station 10 fnr time steps 0 through
49 and zero at this station for all subsequent times. At all other stations the load was zero
at all times.
The calculation was carried out for 260 time steps or for a total duration of 5 sec and
the rezulu were printed for each step.

B-1
 TABLE B-1

Additional Parameters for GOPHER MARINER Used in the Calculation of Transient

Response by the Digital Method
These parameters wore not included in the normal mode calculation given in Appendix A.

rC
n
1 'mz
2
ton-sec -ft ton-sec /ft

0 1,068 7.48
1 4,663 i3.97
2 6,122 22.31
3 9,128 38.76
4 13,960 56.51
5 20,430 73.04
6 19,400 69.14
7 21,530 75.89
8 24,940 85.47
9 27,470 90.79
10 30,320 95.11
11 23,470 80.35
12 21,950 75.76
13 18,680 67.02
14 13,650 50.61
15 9,972 41.14
16 5,836 25.77
17 5,569 26.61
18 3,577 23.80
19 1,989 14.94
20 384 5.53

Figure B-1 shows the calculated displacement at Station 10 over the 5-sec interval.
Figure B-2 shows the instantaneous displacements and bending moments calculated
for all stations of the hull at the instant when the peak bending moment was reached.
At the stage of the development of the transient hull vibration calculation at which
this example was run, buoyancy and gravity forces were not included. Thus the hull was
treated as free in space as in the hull normal mode calculations. As shown in Reference
B-4, the coding was later modified to include buoyancy and gravity forces.
It must be realized that when the idealized nonuniform beam (with added mass and free
in space) is subjected to a unidirectional force there will ensue, in general, not only vibra-
tory elastic motions but also rigid body motions in both translation and rotation. Since the
gravity and buoyancy forces are not included in the treatment discussed in this Appendix,
the calculation will not be realistic if carried out long enough for large rigid body displace-
ments to build up. The revised coding discussed in Reference B-4 will include gravity forces,
buoyancy forces, an added mass that varies with heave and pitch, as well as provision for a
forcing function varying arbitrarily in space and time. Plans are also underway for handling
the hull slamming problem by means of an analog computer; see Reference B-5.

B-2
 0 _900
_ I _I_ I_-"
0.800C I I I -4 II

2 0.700
x
Z 0,60O0/
0.400
CL

2 a 0.300-
O.40C ...
0.200

0 0.5 1.0 I .5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

t in seconds

Figure B-1 - Transient Response of jGOPItER MARINER with Displacement

Calculated at Station 10

0-.

~ O.30
 1-34
-IOO

0 2 4 6 e 10 12 14 IS iS 20
Sltion

Figure B-2a - Bending Moment

60 0
S0-

0 2 4 6 a 10 12 14 16 18 20
Station

Figure 13-2b - Vertical Displacement

Figure B-2 - Transient Response of GOPHER MARINER to 1-Ton Rectangular Pulse

Load Amidships aL Instant of Peak Bending Moment by Digita] Calculation

REFERENCES

B-1. Jasper, N.H. and Birmingham, J.T., "Strains and Motions of USS ESSEX (CVA 9)
during Storms near Cape Horn," TMB Report 1216 (Aug 1958).

B-2. McGoldrick, R.T., "Calculation of the Response of a Ship Hull to a Transient Load
by a Digital Process," TMB Report 1119 (Mar 1957).

B-3. Polachek, H., "Calculation of Transient Excitation of Ship Hulls by Finite Difference
Methods," TMB Report 1120 (Jul 1957).

B-4. Cuthill, E. and Olsten, E., "A Routine for the Calculation of the Response of a
Ship to Transient Forces," TMB Report (in preparation).

B-5. Hoyt, E.D., "Simulation of the Sea and Analog Computation of the Forces on a Ship
in Waves," Report to TMB on Contract Nobs 72376, T.O. 9, Project RR-1458-N by Reed Ri-
search, Inc (Nov 1959).

B-4
 APPENDIX C
EMPIRICAL FORMULAS FOR HULL FREQUENCIES

A. FORMULAS FOR FREQUENCY OF 2-NODE VERTICAL MODE

It was pointed out in Chapter 10 and is emphasized in Reference C-1 that the designer
is fortunate in having a very simple rough rule for the ratios of the frequencies of the higher
vertical modes of vibration of a surface ship to the frequency of the fundamental mode. This
rule is that the ratios follow thA Aeries 1.. ate.
Some idea of the roughness or this ruie may be obtained from Table C-i, based on
experimental data obtained by the David Taylor Model Basin and summarized in Reference
C-2. The average values in Table C-1 conform quite closely with the rule although several
of the individual deviations are larger than 15 percent.
The availability of such a rule enhances the value of the empirical formulas for esti-
mating the 2-node vertical frequency of ships since these, in effect, yield an estimate of the
frequencies of the principal vertical modes.
The most widely known of the empirical formulas for the frequency of the 2-node verti-
cal mode of a surface ship is the formula of 0. Schlick:C- 3

A= C V [C-1]

TABLE C-1

Experimentally Determined Ratios of Frequencies of Higher Vertical Modes of

Ships to the Frequency of the Fundamental Vertical Mode
Ratios
Ship Type 1st Mode 2nd Mode 3rd Mode 4th Mode

NIAGARA Transport 1 1.83 2.65 3.23

CHARLES R. WARE Destroyer 1 2.08 3.30 4.55
C. A. PAUL Or Carrier 1 2.35 3.73 4.67
PERE MARQUETTE Car Ferry 1 2.00 3.10 4.57
OLD COLONY MARINER Dry Cargo 1 1.89 2.77 3.28
NORTHAMPTON Cruiser 1 2.11 2.99 4.23
STATEN ISLAND Ice Breaker 1 1,93 2.57
Average 1 2.03 3.01 4.09

C-1
where N is frequency in vibrations per mrin, I
C is Schlick's empirical "constant" ranging from 1.28 x 10s to 1.57 x 104,
I is the area moment of inertia of midship section in ft2 -in. 2 units,
A is the displacement of ship in tons (2240 lb), -and
L is the length of ship in ft.
Note that this formula is similar to the formula for the free-free uniform bar discussed
in Chapter a. The simplicity of the formula is naturally appealing but there remains the prob-
lem of selecting the best value of the empirical "constant." Table C-2, taken from Roference
-,gives experimentally determuined values of the Schlick constant for various types of
ships. Such a table can be used by the designer as a guide in selecting an appropriate value
of C to use in Equation [C-1]. It is recommended, -however, -that, -if the designer can obtain
more recent information on both the natural frequency and the design parameters of a ship of
a type more nearly similar to the Lype he i8 designing, he use this information to estimate
the fundamental vertical frequency of the new ship. Thus,
/I. A. Lo3,
N,, = I A L3[C-21
0 nl n

where the symbols have the same meaning as in Equation (C-1] except that the subscript a
applies to the new ship, and the subscript o applies to the old ship.
TABLE C-2
Empirical Values of Schlick Constants

Overall Inertia Measured Eu.

Author and Type of
3
Reference Ship Displacement Length Amidships T Frequency Constant
2
-- tons ft ft--in x 104 per min x 10-s

Todd (C-5) Tanker 1 15,190 440 476,000 6.08 78.9 1.30

Tobin (C-6) " 8,300 1350 233,890 8.11 112 1.38
Nicholls (C-7) Destroyer 1,378 310 33,000 8.96 120 1.34
Schmidt (C-8) Motorship 7,010 484 718,000 9.50 106 1.12
Schadlofsky Tanker 16,600 462 604,000 6.08 81 I 1.33
(C-9) Cable layer 834 181 15,600 17.8 207.5 1.17
Tanker 8,160 366 234,000 7.65 112 1.47
Freighter 8,360 371 264,000 7.88 105 1.33

Cole (C-10) Tanker 8,151 350 233,890 8.19 112 1.37

EMB Report
(C-ll) Destroyer 1,382 310 35,000 9.22 107 1.16
Roop (C.12) Tanker __15,430 475 447,000 5.20 60.3 1.16
_ Battleship 32,000 583 1,325,000 4.57 77.1 1.69

C-2
 3
L. C. Burrillc'l proposed a formula of greater flexibility than the Schiick formula
in that it includes the beam and the draft. Hence, -while it also included an empirical con-
stant, the latter was expected to be more stable than the Schliek constant. Burrill's formula
is

(1+ : (t+r)

where c is an empirical coefficient given by Burrill as 24 x 105,

N iz ihe Lundamenit.l vertical `&euency i" --

I is the effective moment of inertia of the midship section area in ft 4 ,

A is the displacement in tona,
L is the length between perpendiculars in ft,
B is the beam in ft,
d is the draft in ft, and
r is J. Lockwood Taylor's shear correction factor

[ 8.5D 2 (3a 3 +9a2 + 6a + 1.2)]

2
[-41
L(3a + 1)
B
whore a DD and D is the molded depth in ft. J. Lockwood Taylor's shear correction factor
is discussed further in Reference C-14.
Some idea of the constancy of Burrill's coefficient can be obtained from Table C-3,
derived from Reference C-15. Unfortunately, several of the values given in Table C-3 were
based on tests in shallow water, a condition known to increase the water inertia effect.
Two other empirical formulas for the frequency of the 2-node vertical mode of vibration
of ships are given here without further discussion; namely, -the formulas of Prohaska, 'and
Todd and Marwood.
Prohaska's formula (see Reference C-16) is

100/? /)I

where N is the 2-node vertical frequency,

1? is rIr2r3)
rI is the correction for variable inertia,
r.? is the correction for shearing force,
P3 is the correction for transverse compression and dilatation,
q is the mass distribution coefficient,

C-8
 TABLE C-8

Experimental Values of Burrill's Coefficient Derived

from Vibration Generator Tests of Ships

Design Test Displacement Depth of Burrill's

Name of Ship Type of Ship Displacement Design Displacement er Coefficient
tons under Keel
HAMILTON Destroyer 16.3 feet 22.7 x 101
SOUTH DAKOTA Battleship 42,500 0.90 Shallow 26.7 x 105
ALASKA Battle Cruiser 31,600 0.90 Shallow 24.3 x 105
SHILOH l anker 21,800 0.27 Shallow 33.0 x 10,
PHILIP SCHUYLER Cargo 14,200 0.36 14 feet 21.4 x 105
(full load)
PONTCHARTRAIN Coast Guard 1,970 0.75 Shallow 27.7 x 105
Cutter (full load)
MACKINAW Icebreaker 5,090 0.81 80 feet 27.5 x 10
2 feet 21.0 x 10-5
NIAGARA Transport 6,740 0.82 127 feet 25.5 x 10S
(full load)

c is the added mass of water/displacement of ship,

I is the moment of inertia of midship section,
A it the displacement, and
L is the length of ship.
The choice of units in this formula and the evaluation of the r's are discussed in Reference
C-16. It is noted here, however, that the formula, like the Schlick formula, is essentially the
formula for the free-free uniform bar. Since c, q, rF, .r , P3, and R are all dimensionless, one
is at liberty to apply Prohaska's factor
R

to the uniform bar formula in any coniotent set of units. The uniform bar formula is
22.4 1) '__
Cal 12 , p C6

where w,1 is the natural circular frequency of the 2-nodo flexural mode,
El is the bending rigidity, -and
I is the mass per unit length.

C-4
This gives

N1 - 8.58 V -8.58 1 [C-7]

where N, is the frequency of the uniform bar in cps,

E is Young's modulus in tons/ft 2 ,
I is the area moment of inertia in ft4 ,
M is the total mass of the bar in ton-sec 2 /ft units, and
L is the length of the bar in ft.

Hen.c these ..... bi-e .t,'d iplyvit Prnhfakfx'M Fnrmnla Fnr thA Rhin which becomes

8.58R EA
~g~~c) AL 3 C8
g

where A is the displacement in tons,

g is the acceleration of gravity in ft/sec2 , and
6
R, q, -and c are to be derived in nondimensional units from Prohaska's paper.CY'

The Todd and Marwood formula (see Reference C-17) is

D3
NA=. [C-91
AL 3

where N1 is the 2-node vertical frequency in cpm,

B is the moulded breadth in ft,
D is the moulded depth at side in ft,
A is the displacement in tons,
L is the length between perpendiculars in ft, and
0 is an empirical coefficient.
In Reference C-18 Todd gives values of fl ranging from 45,200 to 62,500.

B. FORMULA FOR RATIO OF HORIZONTAL TO VERTICAL FREQUENCIES

Experience has shown that the ratio of the 2-node horizontal frequency to the 2-node
vertical frequency is roughly 1.5, as pointed out in Reference C-19. In Reference C-20 A. J.
Johnson proposed the simple formula
N11 InXCV-

C-5
where NN is the 2-node horizontal frequency,
NV is the 2-node vertical frequency,
IR is the moment of inertia of midship section area for bending in horizontal plane,
1V is the moment of inertia of midship section area for bending in vertical plane,
Vy is the vertical virtual inertia factor =
added weight for vertical vibration
1+ displacement and

is the equivalent horizontal virtual inertia factor

effectivn adried woioht- _n. hrnv .:! 'ibrtic
1+ displacement

Although the data available on horizontal modes are more scanty at this time than the
data on vertical modes, the 1, 2, 8 rule for estimating the frequencies of the higher modes
from f-- freqney of the 2-~-ode moeo still Seem_ to have utility.

C. HORN'S FORMULA FOR THE 1-NODE TORSIONAL FREQUENCY

Horn's empirical formula for the fundamental torrsional frequency of surface ships (see
Reference C-21) is

N e =60k 2 [C-II]
T A(B + D2) L

If English units are used in this formula, then

Ne is the natural frequency in cpm,
g is the acceleration of gravity in ft/sec 2 ,
Je( is the effective polar moment of inertia of midship section area in ft 4 ,
A is the displacement in tons,
B is the beam in ft,
D is the depth in ft,
L is the length in ft,
0 is the shear modulus of elasticity in tons/ft 2 , and
k is the Horn's empirical constant.

Wt;O-hli,. ih , t wuia fur torsional rigidity proposed by L.

Gumbel. C-22
eo A2
[C-12]
8

C-6
where F0 is the area enclosed by the shell plating of the midship section (not the area
of the material),

a is the plating thickness, and

A8 is a small distance along the shell plating in the plane of the section.
This formula follows from the shear flow concept and does not take account of the effect of
inner decks and longitudinal bulkheads.
Horn assigned the value of 1.58 to k for a freighter (SS WASGENWALD). As in the
case of the Schlick formula, -the best procedure for the designer is to use Equation [C-11] in
n.... io....... ;th A th n1nal fmrffnincYv nf a shin nf the oeneral tvye he is designing,
if such data can be obtained. Moreover, in estimating Jeo he is not restricted to Equation
[C-12] but may apply any method of estimating torsional rigidity of hulls; see, for example,
Reference C-23.
In estimating the frequencies of higher torsional modes from the fundamental torsional
frequency, little guidance can be offered at this time. It may be pointed out that Horn C-2I
found for the first three torsional frequencies of a freighter the ratios 1 : 1.6 : 1.9, whereas for
GOPHER MARINER, if the apparent torsion bending modes are considered as flexure-free
torsional modes, the corresponding ratios were found to be I : 1.6 : 2.2.

REFERENCES

C- 1. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER

MARINER," Trans. SNAME, Vol 68 (1955). Also TMB Report 1060 (Jul 1956).
C- 2. McGoldrick, R.T., "Comparison between Theoretically and Experimentally Deter-
mined Natural Frequencies and Modes of Vibration of Ships," TMB Report 906 (Aug 1954).
C- 3. Schlick, 0., Series of Articles on Ship Vibration in TINA (1884, 1893, 1894, 1.901,
and 1.911).
C- 4. MeGoldrick, R.T., "A Study of Ship Hull Vibration," EMB Report 395 (Feb 1935).
C- 5. Todd, F.I., "Ship Vibration-A Comparison of Measured and Calculated Frequencies,"
NECI Trans (1933).
C- 6. Tobin, T.C., "A Method of Determining the Natural Periods of Vibration of Ships,"
TINA (1922).
C- 7. Nicholls, 1I.W., "Vibration of Ships," TINA (1924).
C- 8. Schmidt, F., "Die Berechnung von Schiffschwingungen," WRH, No. 6 (22 Mar 1930).
C- 9. Schadlofsky, E., "The Calculation and Measurement of Elastic Natural Frequencies
of Ship Hulls," STG, Vol. 33 (1932). Also EMB Translation 7 (Jun 1934).
C-10. Cole, A.P., "The Natural Periods of Vibrations of Ships," Trans Inst of Engrs and
Shipb Scotland, Vol. LXXII (1928-29).

C-7
 C-11. McGoldrick, R.T., "Vibration Tests on USS HAMILTON at Washington Navy Yard,"
EMB Report 372 (Dec 1933).
C-12. Roop, W.P., "Natural Frequencies of Hull Vibration," Trans. SNAME (1930).
C-13. Burrill, L.C., "Ship Vibration: Simple Methods of Estimating.Critical Frequencies,"
NECI Trans, Vol. 51 (1934-35).
C-14. Taylor, J.L., "Vibration of Ships," TINA (1930).
C-15. Hardy, V.S., "Vibration Studies of Ship Hulls by Means of Vibration Generators,"
TMB Report C-80 (Nov 1949) Unclassified.
C-16. Prohaska, C.W., "The Vertical Vibration of Ships," Shipbuilder, Vol. 54 (Oct, Nov
1947).
C-17. Todd, F.H. and Marwood, W.J., "Ship Vibration," NECI Trans, Vol. 64 (1948).
C-18. Todd, F.H., "The Fundamentals of Ship Vibration," Shipbuilder, Vol. 54 (May, Jun,
Jul 1947).
C-19. Tachmindji, A.J. and McGoldrick, ILT., "Note on Propeller-Excited Hull Vibrations,"
JSR, Vol. 3, No. 1 (Jun 1959).
C-20, Johnson, AJ., "Vibration Tests on All-Welded and All-Riveted 10,000 Ton Dry
Cargo Ships," NECI Trans, Vol, 67 (1551).
C-21. Horn, F., "Horizontal und Torsions-Schiffschwingungen auf Frachtschiffen,"
WRH, Vol. 6, No. 18 (1925).
C-22. Gumbel, L., "Torsional Vibrations of Elastic Shafts of Any Cross Section and
Mass Distribution and Their Application to the Vibration of Ships," TINA, Vol. LIV (1912).
C-23. Vedeler, G., "On the Torsion of Ships," TINA, Vol. LXVI (1924).

C-8
 APPENDIX D

EMPIRICAL FORMULAS FOR ESTIMATING THE

LEVEL OF STERN VIBRATION

It is pointed out in Chapters 2 and 4 that the propeller-excited or blade-frequency vi-

bration of ships is often concentrated in the stern and is practically imperceptible at points
further forward than 25 percent of the length from the after perpendicular. This condition pre-
vails when the operating blade frequencies are higher than the range of frequencies of signif-
icant hull modes.
An explanation of this phenomenon has been suggested in Chapter 4, and this Appendix
is concerned only with the empirical formulas for estimating the amplitude of this stern vi-
bration under such circumsLances. The formulas presented here were derived from the con-
cept of mechanical impedance which is the ratio of the driving force to the response in steady.
state vibration. Obviously, impedance formulas are useful in predicting levels of service vi-
bration only when the exciting forces themselves can first be predicted. The problem of esti-
mating the exciting forces is discussed in Chapter 7.
The concept of mechanical impedance, although not an essentially new concept, has
gained acceptance in naval architecture only quite recently. The preference of Committee
S-2 (Mechanical Vibration and Shock) of the American Standards Association has been to de-
fine mechanical impedance in terms of vibratory velocity; see Reference D-1. This preference
has been influenced by the concept of impedance in the field of acoustics in which driving
pressures and particle velocities are important quantities. When so defined, mechanical
impedance corresponds with eiectricai impedance in the classical analogy wherein electrical
current is analogous to mechanical velocity, electrical voltage is analogous to mechanical
force, and electrical impedance is analogous to mechanical impedance. Thus, the familiar
electrical equation of alternating current theory is

/ = E-- [D- 11

where I is the current,

E is the voltage, and
Z is the impedance.
This equation corresponds to the equation in mechanical vibration

P0
D-[D-2
z

D- 1
where 1' is the single amplitude of the velocity,
Po is the single amplitude of the driving force, and
2 is the mechanical impedance.
In naval architecture, however, the designer is much more familiar with levels of vibration
expressed in terms of displacement amplitude than in terms of vibratory velocity amplitude.
The formulas given here are based on experimental data on steady-state hull vibration
in the stern produced by known mechanical exciting forces. They are written in terms of
driving force and displacement amplitude and do not employ the mechanical impedance term
implicitly. Their generality, however, does actually depend on the scaling of mechanical
impedance. At this time it can only be said that, since hulls in general have been observed
to follow the same general pattern of stern vibration (once the frequency rises above the range
of significant hull natural frequencies) and, since at high frequencies mechanical impedance
is chiefly inertial, there is some logic in expecting that the same empirical constant will find
application to ships of different types.
Only when the designer has access to much more data than are now available will he
be in a position to decide whether he can use a universal constant in the formula for stern vi-
bration or whether he will have to establish separate constants for the various classes of ships.
The formula proposed in Reference D-2 for estimating vertical vibration at the after
perpendicular under the conditions just stated is

PO
P0 [D-31
3.4 x 10- 6 x A x (cpm) 2

where y is the single amplitude in mils,

Po J- the single amplitude of the vertical component of propeller-exciting force
in lb (at blade frequency),
A is the displacement of the ship in long tons, and
cpm is the blade frequency in cycles per minute.
The empirical constant in this formula is the factor 8.4 x 10-6. Should future experimental
data show a wide variation, then, in using the formula, the designer should select his own
factor from experimental data for the ship type nearest to the one he is designing.
It will be recalled that the rule for the number of significant vertical modes was
adopted from Reference D-3; namely

5L L
N'- - - -- [D-4]
9D B

where N'is the number of significant vertical modes,

L is the ship longth,
P is the ship depth, and
B is the ship beam-
ED-.2
 In considering the use of empirical formulas for horizontal and torsional vibration at
the stern, the designer must recognize that, just as in the vertical case, such formulas are
applicable only when the blade frequencies fall well above the range of significant hull mode
frequencies. This means that there must be no significant hull horizontal or torsional natural
frequencies in this range. It is also suggested that if in the approximate Equation [D-4], N'
is 6 for vertical, the limit for horizontal is to be taken as 4, for torsional as 3. Thus, if in
the case of a long slender hull the number of significant vertical modes is estimated to be
greater than 6, the number of horizontal modes would be considered greater than 4, and the
torsional modes greater than 3 in the same ratio.
There is, however- another element involved in dealing with horizontal and torsional
stern vihration.Iam..IJ,ha..p-ossibility of coupling of these two motions. As pointed ou. in
Chapter 7, the blade-frequency excitation at the stern is reduced for simplicity to a vertical
force, a horizontal force considered to act.at the center of twist, and a couple whose axis is
parallel to the longitudinal axis of the ship. It does not necessarily follow, however, that if
the couple is zero there will be no torsional vibration at the stern. Nor does it necessarily
follow that, if the horizontal force is zero, there will be no horizontal vibration at the stern.
This results from the fact that the effective center of mass may not fall on the axis of twist
(as pointed out in Chapter 3), so that flexural vibration and torsional vibration are coupled in
the hull.
At this stage of the development of ship vibration theory it is only feasible to propose
empirical formulas for horizontal and torsional vibration at the stern on the assumption that
this coupling effect is negligible. The possible coupling effects are discussed further in
Reference D-4.
The empirical formula for horizontal stern vibration is then similar to that for vertical;
namely, the approximate Equation [D-3]. The only difference is the factor to be used in the
denominator. At this time, a factor based on only a single test is available. This factor is
1.9 x 10- , and is based on a vibration generator test on the USS DECATUR (DD 936) de-
scribed in Reference D-5.
The similar empirical formula for torsional vibration in the stern is

70
[D-5]
2
0.46 x I x (Vpim)

where k is the single amplitude in radians,

To is the single amplitude of the blade-frequency exciting couple in lb-ft,
I is the mass moment-of inertia of the entire ship about the longitudinal axis
through its center of gravity. This does not include any allowance for virtual
mass moment of inertia of the surrounding water. it is u.&pressed in ton-sec"-ft
units,
cpm is the blade frequency in cycles per minute.

D-3
Here again the factor 0.46 is based on a single test made on DECATUR, described in Ref-
erence D-5. The designer should inquire for later information if available in the future.
A final word of caution in the use of the formulas given in this Appendix is necessary.
If any local stern structure or appendage has a natural frequency in the range of the driving
blade frequencies to be specified by the designer then, even though the conditions for freedom
from hull natural frequencies are met, the empirical formulas cannot forecast the stern ampli-
tude under a given excitation, and the designer must either resort to the more detailed analyses
discussed in Reference D-4 or assure himself that the exciting forces will be within safe limits.

REFERENCES
D-1. American Standards Association, "Proposed American Standard Nomenclature and
Symbols for Mechanical Impedance of Structures," (pending).
D-2. McGoldrick, R.T. and Ruso, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
D-8. Baier, L.A. and Ormondroyd, J.,. "Vibration at the Stern of Single-Screw Vessels,"
Trans. SNAME, Vol. 60 (1952).
D-4. Leibowitz, R.C. and Kennard, E.H., "Theory of Freely Vibrating Nonuniform Beams
Including Methods of Solution and Application to Ships," TMB Report 1317 (May 1961).
D-5. Paladino, A.R., "Vibration Survey on USS DECATUR (DD 936)," TMB Report 1271
(Mar 1960).

D-4
Calibration
(a) Frequency response - For vibration of propeller blade frequency, a minimum frequency
response of 3 cps is required. For shaft frequency (first order) a lower limit of 1 cps is re-
quired. This may be attained if the respective transducer or vibrograph is critically damped,
producing a linear response at or below resonance.
(b) Amplitude response of - 1/_mil to 100 mils.
(c) Laboratory calibration check - Prior to each test, or at least every three months, the
transducers and vibrographs should be calibrated. In the case of the electronic equipment,

(d) Shipboard calibration check - At intervals throughout the test checks should be made
on the electronic system by introducing known electrical signals. This will avoid error due
to possible drift in the system.

Test procedure
(a) Determine ship particulars:
(1) Ship dimensions.
(2 Propeller: type, dimensions, number, and number of blades.
(3) Propeller aperture clearance.
(4) After body configuration.
(5) Type of power plant.
(6) SHP, RPM, and speed.
(b) Determine test conditions:
(1) Displacement.
(2) Drafts forward and aft.
(3) Loading plan.
(4) Depth of water (120-ft minimum).
(5) Sea state (limit based on ship length).
(c) Take data during following conditions:
(1) Ear-h 5-rpm increment from one-half to full power.
(2) Hard turn to port.
(3) Hard turn to starboard.
(4) Crashback - Full ahead to full astern.
(d) Data-taking procedure:
(1) Permit ship to steady on speed for constant speed runs.
(2) Take sufficient length of tape to permit collection of maximum and minimum
values (about 80 sec for single-screw ships).
(3) For maneuvers, start recorder as throttle or wheel is moved. Allow to run until
maximum vibration is noted.

E-2
Data Analysis and Reporting
(a) The following data are to be evaluated for all runs:
(1) Maximum overall values of amplitude.
(2) Maximum first-order amplitude.
(3) Maximum amplitude of blade frequency.
(4) Maximum amplitude of blade-frequency harmonics.
(b) Data presentation should include the following curves plotted on a basis of shaft rpm:
(1) Vertical hull vibration.
(2) Athwartship hull vibration.
(3) VW.itical vibration at main thrust bearing.
(4) Athwartship vibration at main thrust bearing.
(5)/ Fore-and-aft vibration at main thrust bearing.
(6) Other curves as appropriate.
(c) Method of presentation of data:

(1) All curves should show single amplitude of displacement in mils plotted
against rpm.
(2) Maximum amplitudes obtained during maneuvering runs should be presented in
tabula form giving frequencies and amplitudes.

It will be noted that this draft of the Code makes no provision for measuring torsional
vibration of the hull. The reason for this is that trouble with hull torsional vibration is so
unusual that if a case develops a special investigation will probably be authorized.

REFERENCE
E-1. Noonan, Knopfle, and Feldman, a proposal for "A Code for Shipboard Hull Vibration
Measurements," in preparation for SNAME.

E-3
 APPENDIX F

LEVELS OF SERVICE VIBRATION

Even if hull vibration analysis were developed to the stage at which the designer could
accurately predict the level of service vibration for a ship of given design, he would still be
faced with the question of whether the level of vibration thus predicted was acceptable. This
brings up the vital problem of vibration norms and the definition of such terms as "normal,"
"tolerable," "permissible,", -"acceptable,"
. "severe, ""intolerable,"or "unacceptable,"
as applied to ship vibration.
Although at the prepent time it is safe to state that the naval architect is well aware
of the need for standards of comparison in this field, no definite standards have as yet been
established in the United States. As a matter of fact, there is a natural reluctance on the
part of both the shipbuilder and the ship operator to collaborate in any program for estab-
lishing such standards because of adverse effect that publicity regarding the vibration levels
on a particular ship might have on its earning capacity. Furthermore, in the case of naval
ships$ t`ss-i, the added hazard of revealing classified information in presenting data on the
levels of service vibration.
Accordingly, it seems feasible to include in this Appendix only certain data which have
been accumulated in the course of the ship vibration research carried on by the David Taylor
Model Basin, and which have already beAn disclosed elsewhere in unclassified reports or
papers. Such information will serve to familiarize the reader with some of the levels of hull
vibration encountered in practice. Much additional information on this phase of tho subject
will be found in the references listed in the general bibliography, and certainly much more
will be available in the future.
The following quotation is taken from Reference F-1:

"While it would be of great assistance to the naval architect if tolerances of vibration

amplitudes could be established over the entire range of frequencies encountered on ships
below which the vibration could be assumed acceptable, it appears premature to propose
such standards at present. Criteria based on vibration velocity, acceleration, and the rate
of change of acceleration have all been proposed and these have been based on physiological
effects as well as on engineering considerations. As a very rough indication of present con-
ditions, it may be stated that single amplitudes of 50 mils at 100 cpm may be considered
high as may also single amplitudes of 2 mils at 2000 cpm."

One of the chief points to keep in mind in connection with levels of service vibration
of ships is the enormous difference between the levels of steady-state vibration due to pro-
pellers or machinery, and the levels of transient vibration due to slamming in a seaway.
This point is brought out in the following quotation from Reference F-2 which deals with
shock and vibratioi instrumentation for ships.

F-1
 "If both hull and machinery are taken into account, instruments for measuring steady-
state vibration in ships must be able to cover the range of frequencies from 80 to 10,000
cycles per minute (cpm) and the range of single amplitudes from 0.0001 inch to 1.0 inch.
While desirable, it is not obligatory that these large ranges be covered by a single instrument.
in the case of shock measurement, frequencies range from about 80 cpm to at least 50,000
cpm if account is taken of the elastic vibrations of the entire hull on the one hand, and the
localized vibrations of component structures on the other. Shock amplitudes may range from
0.0001 inch to several feet, the larger amplitudes being due to whipping motions of the entire
hull."
In connection with this quotation the reader is cautioned that the term "shock loading"
here includes loading due to underwater explosion as well as to slamming in a seaway.
In Reference F-8 there are discussed criteria of acceptable levels of hull vibration
proposed by the Boston Naval Shipyard as a retuit uf many years of experience in making
underway vibration surveys on naval ships. A !criterion based on vibratory velocity was
recommended. Figure F-i shows three regions of dispiacement amplitude based on the Boston
figures. As this is written, this is not a Navy-wide standard.

20 .... .

200 300 400 500 Soo 700 800 900 1000 1i30
Frequency In pm

Figure F-1 - Range of Acceptable Amplitudes According to Boston

Criteria (Velocity 0.32 to 0.5- Inch/Second)

Clearly, in the question of acceptable levels of steady-state hull vibration, physic-

logical effects are a major consideration. The recognition of this has led to a number of
recent investigations on the physiological effects of mechanical vibration. in 1959 one of
the research panels of the Society of Naval Architects and Marine Engineers requested the
preparation of a summary of such data by one of its members. This led to the preparation of
a series of graphs by Mr. J.B. Montgomery of the Newport News Shipbuilding and Dry Dock
Company in 1959. These graphs were based on various criteria depending on the original

F-2
source. As an illustration, one of these charts, based initially on Reference F-4, is repro.
duced here as Figure F-2. It will be noted here that the criterion of comfort is acceleration
regardless of frequency. Some investigators have considered the rate of change of acceler-
ation or the third derivative of displacement with respect to time more significant than the
acceleration itself.

40 . .. .

20 Unbearable

Hardly Supportable

6 ~~Extremely Uncomfortoble-- '

C Very Uncomfortubie
z- 2

Slightly Uncomfortable
d 4=

10 20 40 60 100 200 600 1000 4000 10,000

Frequency in Cpm

Figure F-2 -Vibration Scandard Adopted by L'lnstitut de Recherches de la

Construction Navale (Based on Physiological Effects)

The Navy's "Military Stand ard-Mechanical Vibrations of Ship Equipment" (Reference

F-5) is intended as a guide for the vibration testing of equipment to be installed on board
ship. Table F-1 taken from Reference F-E, gives thfe Itc!Yel of vibration at which endurance
tosts on equipment are to be run. Although this is based on accumulated data on service vi-
11-0i'on of naval ships, it, is no. to be construed as a criterion of acceptable vibration levels
for future designs.
 TABLE F-1
Amplitudes of Vibration for Endurance Testing of Shipboard Equipment

Frequency Range Table Amplitude

cps single amplitude, inches
5 to 15 0.030 0.006
16 to 25 0.020 t 0.004
26 to 33 0.010 + 0.002

REFERENCES

F-i. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).
F-2. McGoldrick, R.T., "Shock and Vibration Instrumentation," Symposium aL Lhe 17th
Applied Mechanics Division Conference of the ASME held at Penn State University
(19-21 Jun 1952).
F-8. Gold, P.D., et al., "Vibration Engineering-Resume of Applications to Solution
of Marine Operational Problems Encountered by Naval Vessels," Trans. SNAME, Vol. 61
(1953).
F-4. Dieudonnt, J., "Vibration in Ships," Translation by W.C.S. Wigley, TINA (1968).
F-5. "Military Standard - Mechanical Vibrations of Shipboard Equipment," MIL-STD-167
(Ships) (20 Dec 1954).

F-4
 APPENDIX G
SCALING CONSIDERATIONS IN MODEL VIBRATION EXPERIMENTS

In any attempt to determine the vibratory response characteristics of a ship by means

of model experiments, the laws of similitude are of prime importance. It naturally occurs to
the research worker that it is far simpler and less costiy to conduct model experiments in the
field of ship vibration than to conduct full-scale experiments. However, as pointed out in
Chapter 7, aside from the difficulties of fabrication, it would not be feasible to run a self-
propelled model of a ship to determine the amplitudes of service vibration by direct obser-
vation of the amplitude of the model.
The reason for this situation is that significant quantities affecting the vibratory re-
sponse do not scale so as to give overall similitude in this case. This should not seem
surprising, and, in fact, it is hardly to be expected that overall similitude, regarding ship vi-
bration could he obtained when it is not attainable even for the fundamental problem of ship
resistanae. The general subject of model experimentation and similitude is treated in Ref-
erence 'G-1 and in many other publications. In Reference 0-1 it is shown how various dimen-
sionless parameters can be derived from the total number of physical quantities affecting the
phenomenon and inci number of fundamental units involved (such as mass, length, and time;
or force, length, and time). It turns out that dynamic similitude is attainable when the model
is constructed to scale of the same material and when external loads are applied which are
related to the full-scale loads both in magnitude and in timu variusiun i1L auuvrdwun with
Table G-1.
Thus, in accordance with Table G-1, any natural frequency of a solid body could be
determined by making a scaled model of the same material and measuring the frequency of the
desired mode of the model provided gravitational effects were negligible. The frequency of
the prototype would then be A-1 times the frequency of the model, where A is a number de-
termined by the first item of Table G-1. If the mode of oscillation is influenced by gravity,
the relation between the model and full-scale frequencies will not conform to Table G-1.
This relation could be fulfilled only if it were possible to modify the gravitational field
surrounding the model in such a way that the acceleration of gravity satisfied the 15th item
of Table G-1. A model of a simple pendulum would therefore not follow the rule for frequency
given in Table G-1.
It is thus readily seen that the rules of similitude for vibration are not satisfied in
self-propelled model tests of surface ships. In such tests the corresponding shaft speeds of
model and prototype are related as follows:

ship shaft rpm Lmhop

model shaft rpm =

G-i
 TABLE G-1

Similitude Relations in Dynamic Model Testing for Equal Stresses

Measured Quantity Prototype Model

Length L I XL
Angular Displacement 0
Area A A2 A
Area Moment of Inertia . k4 .
Volume V x3 V
Mass m Xm
Maas Moment of Inertia 1 ?h/
Mass Density p p
Modulus of Elasticity E E
Stress a a
Time t At
Natural Frequency N )C'-N
Displacement (rectilinear) d Ad
Velocity
Acceleration a -1 a
Force F A2F
Torque T A3 T
Spring Constant
(rectilinear) Ak
Spring Constant k"
(angular)
Damping Constant
(rectilinear viscous) C
Damping Constant A"0'
(angular viscous)
Ratio of Damping to
Critical Damping C/Ci
Power KW A2 KW
Mechanical Impedance z2z
(based on velocity)
Reduced Frequency bw
(Strouhal nLumber) -)

Since the blade frequencies would fall in the same ratio as the shaft speeds, the rule for
frequencies WiTable G-1 would not loo satisfied.

G-2
 2 although the scaling IwA
It is important to note that, used in ship resistance model
V
work (namely, Froude's law: = constant) does not conforza with the rules for model
-
r- Lg
testing involving elastic vibrations given in Table 0-1, it does furnish a basis for studying
rigid body oscillatory motions of ships. Thus, Froude's law is applicable to the determina-
tion of bending moments in ships in waves, and models miv be devised (not true scale models
in all respects) in which the frequencies of important mod,.s of elastic vibration are made to
conform with these rules. These rules are summarized in Table G-2,

TABLE G-2

S-inilitue RelaLions in Model Testing by Froude's Law

Measured Quantity Prototype Model

Length L AL
Velocity
Time t A~t
Acceleration (rectilinear) a a
Force F A3 F
Moment M X4 M
Pressure p p
Displacement (rectilinear) d Ad
Angular Displacemet 0 0
Mass Moment of Inertia I As/
Rigid Body Natural
Frequencies N ,7N

Flexural Natural Frequency

(required) I N -yNN

Another important dimensionless quantity in model testing involving flow is the

Lv
Reynolds number - where L is a characteristic lrngth, v is the fluid velocity, and v is the
V
kinematic viscosity. Where frictional resistance is predominant, as in the study of flow in
pipes, the Reynolds number is used as a basis for similitude. Important similitude relations
based on Reynolds number are given in Table G-3.
In Chapter 14 it was pointed out that in the aircraft field, model testing for flutter is a
well-established practice. Here the dimensionless quantity on which the rules of similitude

arc based is the reduced frequency or Strouhal number L where b is the semichord of the
V
airfoil, w !he circul ".f ,'vquency of vibration, and v is the velocity of undisturbed air
relative to the airfoil. The rules thus derived conform with those for dynanic model testing

G-.a
 TABLE G-8

Similitude Relations Based on Reynolds Number

(Assumizg the Same Fluid for Both Cases)

Measured Quantity Prototype Model

LePFth L X
kL

a I
[Fo~c..I F"

for equal stresses given in Table 0-1. In the case of wind-tunnel testing of aircraft flutter
models, it is possible to reduce the discrepancies between the model and prototype Reynolds
numbers by increasing the air density. In hydroelasticity this procedure is not feasible.
Although up to this time the extent of model work in hydroelasticity has been insignifi-
cant, the possibility of such developments should be noted. On the other hand, it must also
be noted that in hydroelasticity the only craft comparable to the vehicle in the aircraft case
is the submarine since the surface wave effect imposes the restriction that models of surface
vessels be propelled at velocities conforming not to the Strouhal number but to the Froude
number.
Further discussion of the use of models in vibration research will be found in Ref-
erences 0-2 and G-3.

REFERENCES

G-1. Buckingham, E., "Model Experiments and the Forms of Empirical Equations," ASME
Trans, Vol. 37 (1915).
G-2. Den Hartog, J.P., "Use of Models in Vibration Research," ASME Trans,
Vol. 54 (1932).
G-3. Hermes, R.M. and Yen, C.S., "Dynamic Modeling for Stress Similitude," Contract
Nonr-523, Dept of Appl Mech, Univ of Santa Clara, Santa Clara, Calif (Jul 1350).
 APPENDIX H

MISCELLANEOUS INFORMATION ON VIBRATION OF SHIPS IN SERVICE

Certain unclassified data with regard to vibration of ships in service are presented in
this Appendix without comment. A designer will often find such miscellaneous information
helpful in his efforts to guard against the occurrence of hull vibration in a ship of new desi 3n.
There is no attempt here to present a complete compendium of the information available at the
lime of this writing.
Table 11-1 presents observed values of natural frequencies of hulls reported in tech-
nical literature. Table H.2 gives the frequencies of vertical flexural modes of hulls as deter-
mined in vibration generator tests conducted by the David Taylor Model Basin. Table H-3
gives the frequencies of the horizontal flexural modes (considered as torsion-free) for the
ships listed in Table H-2.
Among the eight ships tested with a vibration generator (Tables H-2 and H-3), there
were a few for which a 1-node torsional mode was either identified or appeared likely. The
frequencies for these cases are given in Table H-4.
During and immediately after World War 11 the U.S. Navy conducted extensive investi-
gations of shock effects on naval ships. In connection with these investigations, data on the
natural frequencies of various local structures were obtained and summarized in an unclassi-
fied report; see Tables H-5 and H-6 taken from Reference H-15. Of special interest to the
designer are the data on the natural frequencies of panels of side plating, deck plating, and
transverse bulkheada which are reproduced here from that reference.
In connection with an investigation of environmental conditions at the location of radar
equipment on naval ships, the David Taylor Model Basin obtained information in 1959 as to
the relative magnitudes of various types of ship vibration in calm and rough seas; see Table
H-7 taken from Reference ii-i6. While these data are based on statistical analysis and wore
an innovation at the time they were presented, they may serve to give the designer some idea
of the augmentation of vibration levels which accompanies a change in the sea state.
The normal flexural modes of the free-free Euler-Bernoulli uniform beam are often used
as assumed modes in starting calculations of natural frequencies and normal modes of hulls
by the Stodola process. Values that may be used in plotting the first three modes are given
in Table H-8.
A sample test schedule for a vibration-generator survey on a commercial cargo ship is
given in Table H-9. The ship is GOPHER MARINER, which has been used as an example Ir
many places in this book. The survey is discussed in dueail in iirm , 11-17, Ml i.....'AU
,
be noted that this included, propeller-exciting force determination as weil as hull-vibratory
response characteristics.

H-1
 ':%." ;atur ;.[ ~' .,',: znci o , .x'..:l \HI,;!:-,ib
; , ,:,'t' e a(SL

,. - .. ,...-/,.z,,,

-- c .cmn.yeo
Ty,- re u( ir-

Author :,nd Yoeof O.,place entement Tre u .y

,efere ce ship - *,;ns ... r" > I..
:n'. Vibration

'.-odd (H-.1i 7:' __4 4er

2-.,o.v " -- I '"''; - ' -,.. e i ,

a- 13f41
43To377 -. .48
Cargo I G 90 /6,550 100
; ,. ..- ' 9.150 .,.89

11,
/ 7:?' 6,850 .. ' -.
Cargo. 1"0,07P .,, ,.!,0u . 3-n~ode ve, . 5
Tin'4 p, .,8'
0 -; .O 1
,., 13,5 ,'. 12,700 2-node hot.. "-"5

e .475 6-,850-

12.8' i ' 8,000 17

14I6 I "ldd (I i ) I r go
C ai 7 ,51 0 od e v i:'.
2 .n l0 0( V

a." .- '" lkei r v b I6 ".ker 15,1 7?

,,*n, I'I;4,377
I"C::rgo 6,590

Tpn.kei 8300 " 3,835 109

O J (H-4' Ir, Tanker 4, 11(' '-5

1,7P no(2 nor ' -5 4,317

i,30 3-node ' 1,871 *

S J-nod, r 260- 390g 10,070. 101
.Io .. n''",-.,
. oro r"-, 63T1anker
630- IDU-. '"'

U'-5) L! " 12..j 2-o", . . .i' ', I -ro 13,00

, Tanke! 11LA75
::- '.1. l -61 Tono,r 16,S. 1 ,*- Tanker.i1"6
L,h.le ;ayer 8 ,'," -' ,

Uestroye, ",4. 6,

FreTgtn I S!8,,0u
I 112 8,,"-

Lovett (;i.") Passenger ,,.err .Ro ( " ) i-ode v

84 I ;""
67 : .. j

.,,000 73.5

Tan6,' T 76 3attleship J,000

__... .... . . . . ......... .......... ill -..

. .. Cole (H-12) Tanke 5,500

Horn (H-13, f rei i-fr i 13,28'1 de ver. 83.1

L node
verL { 171.4

Best Available Copy

 TABLE H--2

Frequeneies of Vertical Flexural "d,)ds of Hulls in CPM

(From Reference i-li)

'

VI
"( GARA_
"
IName of Ship
I Type
Type
'Mode

Transport
1

110
2

200
3.....

2S?
4 i

355
5

448
6 7 I 8

SR L RE Destroyer 79 165 2V' 360

J, KULAS Ore Carrier 92 15') 200 246 285 304 30
.AUL
kP. Ore Carrier 45 106 16E 210 312 354 432
PEREMARQUETTE 21 Car Forry 112 224 34; 512
OflLD COOLONY MARINER Up, Cargo 82 155 221 270
. 'RI AMPTON Cruiser 68 133 204 288 359 437 500
..TTEN ISLAND Icebreaker 280 540 720

TABLE H1-3

' ,.envies of Horizontal Flexural Modes of Hulls (Considered Torsion-Free) in CPM

(From Reference H-14)

Type Mode
Name of Ship I - 2 _37 4 5
NIAGARA Transport 190 402 585
CHARLES R. WARE Destroyer 132 246
E. J. KULAS Ore Carrier 195* 320* 375*
C. A. PAUL Ore Carrier 180 300
PERE MARQUETTE 21 Car Ferry 220 390
OLD COLONY MARINER Dry Cargo 119 280 350 435
NORTHAMPTON Cruiser M03 183 276 327t 392
STATEN ISLAND Icebreaker 42010
*Experimental determination of number vf nodes not made; tabulation made
to yield beat agreement with calculated values.

tUncertainty as to whether this is a flexural or a torsional mode.

TABLE H-4
Frnquencies of 1-Node Torsional Modes of Hulls (A~sivmed Flexure-Free) in CPM
(Prom Reference H-14)

Ship Type Frequency

NIAGARA Transport 322
CHARLES R. WARE Destroyer 310
E. J. KULAS Ore Carrier 262*
NORTHAMPTON Cruiser 346*
Not positively identified as this mode.

If-3
 TABLE H-5

Dominant Frequencies LUf S;dde Plating and BUlkhes9 WV--- -s Naval SLips
(From Reference H-15)

Structural Units I ,I Method of

reOrcins
Flequercwos
JFrequen'
Ranges of-ie"irin'nt
s, cps Method of Exciting
Vibration

Destroyer CAMFJOC4l
Transverse bulkhead, Frame 29 Oscilograms K - 77 Depth-charge firing
Transverse bulkhead, Frame 99 Oscillograms 42-116 Depth-charge firing
Transverse bulkhead, Frame 152 Oscillograms 25-177 Depth-charge firing
Heavy Cruiser CANBERRA (CA 70)
Side plating at flag cabin and stateroom, Oscillogr -9f
main-deck level 51- 91 Gun firing
Panel of side plating at flag cabin Calculation 54-117
Panel of side plating at flag stateroom Calculation 38- 81 1
Light Cruiser
1 MIAMI. (CL89) _

Side plating and bulkhead at various locations IVibrograph recordss 10-'21 rning
Destroyer SUMNER (D, 692)
Oscillograms 7 _29- 49 G
Side plating of after deckhouse Calculation 34- 80 Gun firing
Side plating and bulkheads at various locations Vibrograph record, 52-108 Taoping
Submarine DRAGONET (SS293)
Periscope sheers Vibrograph records 504, 536 Ramming
Conning-lower coaming Vibrograph records 157 Ramming
Conning-tower fairing Vibrograph records 134 Ramming
Back side starboard running light Vibrograph records 60 Tapping
Bridge coaming __ Viorograph recordo 42 Tapping
__...... _____ F~attinship Wi, .!!RI (BR 63)11
Sde plP.,,g at senior staff officer's
cabin and stateroom Osciliograms 43- 86 Gun firing
Side plating at senior staff officer's Vibrograph records 61- 79 Tapping
cabin and stateroom
Side plating at senior staff officer's cabin Calculation 75-149
Side plating at senior staff officer's s'ateroom Calculation

R-4
 TABLE H1-6

-'.inant Frequencies of Decks on Various Naval Shipi

(rmRefernce 11.15)

Siluctiuil Unij Obann Ragso omnn ehdfEctn

Vocks at various locations$ 3 103 Ramming

Portion OfMain Oeck Osilogramss 26-46 Deptht-charge firing

Portions of dodck
- t vai.. ioilin Vbograph records 49-315 Topping
__________Submarine DRAGONET (SS293)
After te-nedo mom,over stitffen- Vroahrecords 1296 Ramming

TABLE H1-7

Factors for Converting Vibr tion Amplitudes in Calm Seas to Extreme Conditions

The vibration amplitude for c~almi sea operation is taken as A. (Prom Rteference H-16)

Vibration Amplitude for

Type -"'bration Extreme Conditions
Aircraft Carrier Destroyer
PropellIer-excitqd 4A 4A
Excited biunba~lance
of A A
prope~ier-shaft syste _________ A ____A_

Transient vibration A
A4
I A
during maneuvers

TABLE H-8

Normal Mo~-de 6 -ipos of who Free- Prov Euler-Bernoulli Beam

Mod2 Station No.

23 4 5 7 10 11 12 13 141is16 1718s19 20
2-node 0.768 0.544 0.312 0 098 i.099 0.172 0.414" I.21 V3.586
1.000 -0.608 0.586 0.621 044.220.099 0.098 0.312 0.544 0.768 1.0001
3-node ~ ~ 37FV w061 .30.263 0.0 0.1263
IQ00680281 0.483 065 .5 0.585 0.397 0.118 0.228 0.608 1.000
4.nodo 1.000 0.454 01052
F.'442 F.643 7.91. ,u.397 WI5 038.680.711 0.608 0.328 0.045 0O.397
0.621 0643 0.442 0.052 0.454 1.000

NOTM rne horizontal ii.. Dove the decimal point indicates a negative value.
 TABLE H-9

Schedule of Vibration Tests on GOPHER MARINER

Transverse Vibration

1st Day
6:00 a.m. - Check operation of equipment.
7:00 a.m. - Start testing immediately following the
clearance of the 30-fathom line.
7:15 a.m. - Operate ship at 5 knots.
Start running vibration generator.
Make frequency survey to determine transverse
natural modes,
10:15 a.m. - Continue frequency survey with increased eccen-
tricity - check important frequencies.
2:15 p.m. - Finish frequency survey.
Prepare for continuous acceleration.
2:30 p.m. - Build up from 5 knots to full power in period
of hour.
Record transverse vibration.
3:00 p.m. - Finish continuous acceleration test.
Prepare for propeller force measurements.
3:15 p.m. - Operate at 106 rpm.
Hold rudder 2 deglees.
Record transverse vibration.
3:30 p.m. - Operate at 100 rpm.
Hold rudder 2 degrees.
Record transverse vibration.
3:45 p.m. - Operate at 95 rpm.
Hold rudder t 2 degrees.
Record transverse vibration.
4:00 p.m. - Operate at 90 rpm.
Hold rudder 2 degrees.
Record transverse vibration.
4:15 p.m. - Finish propeller forces.
Build up to 95 rpm.
Prepare for turning tests.
4:30 p.m. - Execute turning tests at 95 rpm.
Record bending, torsional, and axial strains on rudder post.
Accomplish 90-degree change of heading with following
rudder angles:
5-degree left rudder 5-degree right rudder
10-degree left rudder 10-degree right rudder
15-degree left rudder 15-degree right rudder
Time between change of heading 3 minutes.
5:30 p.m. - Finish turning tests.
Build up to normal power.
Prepare for hull calibration test.
H-6
 TABLE H-9 (Continued)

Schedule of Vibration Tests on GOPHER MARINER

Transverse Vibration

2nd Day
6:00 a.m. - Calibrate hull for transverse propeller forces,
Operate at 106 rpm.
Run vibrator at equivalent of 90 and 95 rpm.
8:00 a.m. - Operate at 85 rpm.
Run vibrator at equivalent of 100 and 106 rpm.
10:00 a.m. - Finish hull vibration.
Shift generator for producing vertical force.
Operate at cruising speed.
Vertical Vibration
2:00 p.m. - Operate ship at 5 knots,
Start running vibration generator.
Make frequency survey to determine vertical
natural modes.
4:00 p.m. - Continue frequency survey with increased
eccet~tricity-check important frequencies.
7:00 p.m. - Finish frequency survey.
Prepare for continuous acceleration.
7:15 p.m. - Build up from 5 knots to full power in period
of h hour.
Record vertical vibration.
7:45 p.m. - Finish acceleration run.
Operate at cruising speed.

3rd Day
6:00 a.m. - Calibrate hull for vertical propeller forces.
Operate at 106 rpm.
Run vibrator at equivalent of 90 and 95 rpm.
8:00 a.m. - Operate at 85 rpm.
Run vibrator at equivalent of 100 and 106 rpm.
10:00 a.m. - Finish hull calibration.
Shift eccentrics for torque.
Prepare for propeller force measurements.
Record vertical vibration.
10:15 a.m. - Operate at 106 rpm.
Hold rudder t 2 degrees.
10:30 a.m. - Operate at 100 rpm.

10:45 a.m. - Operate at 95 rpm.

Hold rudder 2 degrees.

H-7
 TABLE H-9 (Continued)

Schedule of Vibration Tests on GOPHER MARINER

Vertical Vibration

3rd Day
11:00 a.m. - Operate at 90 rpm.
Hold rudder 2 degrees.
11:15 a.m. - Finish propeller forces.
12: Noon - Finish shift of eccentrics for torque.
Operate ship at 5 knots,
S."art running vibration generator,
Make frequency survey to determine
torsional natural frequencies.
3:00 p.m. - Continue frequoecy survey with increased
eccentricity.
Check important frequencies.
6:00 p.m. - Finish frequency survey.
Prepare for continuous acceleration.
"P6.15im. - ufrl'bUp from 5 knuts to full power in period
of hour.
Record torsional vibration.
6:45 p.m. - Finish continuous acceleration test.
Build up to cruising speed.
4th Day
6:00 a.m. - Calibrate hull for torque.
Operate at 106 rpm.
Run vibrator at equivalent of 90 and 95 rpm.
8:00 a.m. - Operate at 85 rpm.
Run vibrator at equivaient of 100 and 106 rpm.
10:00 a.m. - Finish hull calibration.
Prepare for propeller force measurements.
10:15 a.m. - Operate at 106 rpm.
Hold rudder 2 degrees.
10:30 a.m. - Operate at 100 rpm.
Hold rudder 2 degrees.
10:45 a.m. - Operate at 95 rpm.
Hold rudder 2 degrees.
11:00 a.m. - Operate at 90 rpm.
Hold rudder 2 degrees.
11:15 a.m. - Finish propeller forces.
Prepare for anchor test.
11:45 a.m. - Anchor test to occur in no less than 30 fathoms.
Arrest anchor by brake after fall of about 1 fathom.
Repeat anchor test.
12:45 p.m. - Finish anchor test.

H-8
 REFERENCES

H- 1. McGoldrick, R.T., "A Study of Ship Hull Vibration." EMB Report 895 (Feb 1985).
H- 2. Todd, F.H., "Ship Vibration-A Comparison of Measured and Calculated Frequen-
cies," NECI Trans (1988).
H. 3. Tobin, T.C., "A Method of Determining the Natural Periods of Vibration of Ships,"
TINA (1922).
H. 4. Nicholls, H.W., "Vibration of Ships," TINA (1924).
H- 5. Horn, F., "Horizontal-und Torsions-Schiffschwingungen auf Frachtschiffen,"
WRH, Vol. 6, No. 18 (1925).
H- 6. Schadlofsky, E., "The Calculation and Measurement of Elastic Natural Frequencies
of Ship Hulls," STG, Vol. 83 (1982). Also EMB Translation 7 (Jun 1984).
H- 7. Lovett, W.L., "The Vibration of Ships," Shipbuilding and Shipping Record (Jan
192T7).
H- 8. Schmidt, F., "Die Berechnung von Schiffschwingungen," WRH, No. 6 (22 Mar 1980).
H- 9. Taylor, J.L., "Vibration of Ships," TINA (1930).
H-10. Todd, F.H., "Some Measurements of Ship Vibration," NECI Trans (Nov 1981).
H-11. Roop, W.P., "Natural Frequencies of Hull Vibration," Trans. SNAME (1980).
11-12. Cole, A.P., "The Natural Periods of Vibrations of Ships," Trans Inst of Engrs
and Shipb Scotland, Vol. LXXII (1928-29).
H- 13. Horn, F., "li-ochseemessfahrt. Schwingungs-und Beschleunigungsmessungen,"
STO (1935).
H-14. McGoldrick, R.T., "Comparison between Theoretically and Experimentally Deter-
mined Natural Frequencies and Modes of Vibration of Ships," TMB Report 906 (Aug 1954).
11-15. Vane, F.F., "Natural Frequencies of Side Plating, Bulkheads, Decks, and Radar
and Radio Supports, Measured Aboard Three Cruisers, Three Destroyers, a Submarine, and a
Battleship," TMB Report 548 (Apr 1948).
H-16. Buchmann, E. and McConnel, J.D., "Environmental Conditions of Ship Motions and
Vibrations for Design of Radar Systems on Destroyers and Aircraft Carriers," TMB Report
.1298 (Jul, 11951).

11-17. McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER
MARINER," Trans. SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).

H-9I
 GENERAL BIBLIOGRAPHY

Abramson, H.N. and Chu, W.H., "Problems of Hydroelasticity," JSR, Vol. 3, No. 1 (Jun 1959).

Abramson, H.N. and Chu, W.H., "A Discussion of the Flutter of Submerged Hydrofoils,"
JSR, Vol. 3, No. 2 (Oct 1959).

Adams, F.J., "The Steady-State Response of a Ship Hull to a Simple Harmonic Driving
Force Computed by a Digital Process," TMB Report 715 (May 1950).

Allnutt, R.B., "Investigation of Hull Vibrations of USCGC PONTCHARTRAIN (WPG 70),"

TMB Report R-294 (Aug 1946).

Allnutt, R.B. and Mintz, F., "Instruments at the David Taylor Model Basin for Measuring
Vihration and Shock on Ship Structures and Machinery," TMB Report 563 (Jul 1948).

American Standards Association, "Proposed American Standard Nomenclature and Symbols

for Mechanical Impedance of Structures," (pending).
Arkenbout Schokker, J.C., et al., "Design of Merchant Ships," N.V. De Technischo
Vitgerei, H. Stam-Haarlem-Holland (1953).

Babacv., N.N., "Contemporary Status of Development of the Problem of Ship Vibration

(Sovremennoe Sostoianie I Hazvitie Voprosa 0 Vibratsii Sudov)," Sudostroenie, No. 3 (161)
(Mar 1956), pp. 4-10. TMB rranslation 291 (Jul 1959).

Baier, L.A. and Ormondroyd, J., "Vibration at theStern of Single-Screw Vessels,"

Trans SNAME, Vol. 60 (1952).

Bisplinghoff, R.L., et al., "Aeroelasticity," Addison-Wesley Publishing Co. (1955).

Boyd, J.E., "Strength nf Materials," McGraw-Hill Book Co., New York (1924).

Breslin, J.P., "The Pressure Field near a Ship Propeller," JSR, Vol. 1, No. 4 (Mar 1958).

Breslin, J.P., "Ship Vibration. Part I - Propeller-Generated Excitations," Appl Mech

Rev. Vol. 13, No. 7 (Jul 1960).

Breslin, J.P. and Tsakonas, S., "Marine Propeller Pressure Field Including Effects of
Loading and Thickness," Trans SNAME, Vol. 67 (1959).

Buchmann, L. and McConnell, J.D., "Environmental Conditions of Ship Motions and Vibra-
tions for Design of Radar Systems on Destroyeis and Aircraft Carriers," TMB Report 1298
(Jul 1959).

Buckingham, E., "Model Experiments and the Forms of Empirical Equations," Trans ASME,
Vol. 37 (1915).

Bi-1
 Bunyan, T.W., "Practical Approach to Some Vibration and Machinery Problems in Ships,"
Paper prosentod to Inst of Mar Engin and Inst of Nay Arch (Feb 1955).

Burrill, L.C., "Ship Vibration: Simple Methods of Estimating Critical Frequencies,"

NECI Trans, Vol. 51 (1984-85).

Burrill, L.C., "Calculation of Marine Propeller Characteristics," NECI Trans, Vol, 60

(1943-44).
Cole, A.P., "The Natural Periods of Vibrations of Ships," Trans Inst of Engrs and Shipb
Scotland, Vol. LXXII (1928-29).
Crede, C.E., "Vibration and Shock Isolation," John Wiley & Sons, Inc, New York (1951).
Csupor, D., "Methods for Calculating the Free Vibrations of a Ship's Hull (Methoden zur
Borechnung der Froion Schwingungen des Schiffskorpers)," Jahrb. d. STO, Vol. 50 (1956).
TMB Translation 288 (May 1959).
Cuthill, E. and Olsten, E., "A Routine for the Calculation of the Response of a Ship to
Transient Forces," TMB Report (in preparation).
Den Hartog, J.P., "Mechanical Vibrations," McGraw-Hill Book Co., New York, Fourth
Edition (1956).
Den Hartog, J.P., "Use of Models in Vibration Research," ASME Trans, Vol. 54 (1932).
Description of Schelde Synchronizing Gear in Private Communication to SNAME (26 Aug 1957).
Dieudonneo, J., "Vibration in Ships," Translation by W.C.S. Wigley, TINA (1958).
Feldman, S., "Dynamic Balancing for Noise Reduction," Dept of Navy, Bureau of Ships
(Code 371), R & D Report 371-V-24 (Apr 1955).
Flax, A.H., "Aero-Hydro-Elasticity,"Struc Moch, Proceedings of First Symposium on Nay
Strut Mech, Pergamon Press, New York (1960).
Fothergill, A.E., "Vibrations in Marine Engineering," Trans Inst of Engrs and Shipb
Scotland, Vol. 95, Part 4 (1951-52).
Francis, J.J. and Antkowiak, E.T., "Marine Vibration Diagnosis," Paper presented at
New England Section Meeting, SNAME (Oct 1960).
Fung, Y.C., "An Introduction to the Theory of Aeroelasticity," John Wiley & Sons, Inc,
New Yrirk (1955.
Gariboldi, R.J., "Procedure for Torsional Vibration Analysis of Multimass Systems," Dept
of Navy, Bureau of Ships (Code 371), R & D Report 371-V-19 (Dec 1953).
Gold, P.D., et al., "Vibration Engineering - Resume'of Applications to Solution of Marine
Operational Problems Engountered by Naval Vessels," Trans SNAME, Vol. 61 (1958).
Gumbel, L., "Torsional Vibrations of Elastic Shafts or Any Cross Section and Mass Dis-
tribution and Their Application to the Vibration of Ships," TINA, Vol. LIV (1912).

Bi-2
 Hardy, V.S., "Vibration Studies of Ship Hulls by Means of Vibration Generators," TMB
Report C-80 (Nov 1949) Unclassified.

Harrison, NM,,et al., "Wave Effects in Isolation Mounts," TMB Report 786 (Jan 1952).

Heller, S.R., Jr., and Abramson, H.N., "Hydroelasticity: A New Naval Science," Journal
ASNE (May 1959).

Henry, C.J., et al., "Aeroelastic Stability of Lifting Surfaces in High-Density Fluids,"

JSR (Mar 1959).

Hermes, R.M. and Yen, C.S., "Dynamic Modeling for Stress Similitude," Contract N8onr-
523, Dept of Applied Mechanics, Univ of Santa Clara, Santa Clara, Calif. (Jul 1950).

Hirowatari, T., "On the Natural Frequencies of the Vertical Flexural Vibrations of a
Ship's Hull (Part 3)," J Soc Nay Arch, Japan, Vol. 97 (Aug 1955).

Horn, F., "Horizontal-und Torsions--Schiffsschwingungen auf Frachtschiffen," WRH,

Vol. 6, No. 18 (1925).

Horn, F., "Tilgung Elastischer Schwingungen des Schiffskorpers," Schiffbau, Vol. 43 (1942).

Horn, F., "Hochseemessfahrt. Schwingungs-und Besch]eunigungsmessungen," STG, Vol.

S7 (1936).

Hoyt, E.D., "Simulation of the Sea and Analog Computation of the Forces on a Ship in
Waves," Report to TMB on Contract Nobs 72376, T.O. 9, Project RR-1458-N by Reed Re-
search, Inc. (Nov 1959).

Inglis, C.E., "A Suggested Method of Minimizing Vibration in Ships," TINA (1933).

Jacklin, II.M., "Human Reactions to Vibration," SAE Journal, Vol. 39 (1936).

Jaeger, H.E.: "The Determination of the Natural Frequency of Ship Vibrations," Journal
of BSRA, Vol. 5 (1950).

Jasper, N.H., "Structural Vibration Problems of Ships - A Study of the DD 692-Class of

Destroyers," TMB Report C-36 (Feb 1950) Unclassified.

Jasper, N.H., "Statistical Distribution Patterns of Ocean Waves and of Wave-Induced

Ship Stresses and Motions, with Engineering Applications," Trans SNAME, Vol. 64 (1956).
Also TMB Report 921 (Oct 1957).

Jasper, N.H., "A Design Approach to the Problem of Critical Whirling Speeds of Shaft-
Disk Systems," TMB Report 890 (Dec 1954).

13i-3
 Jasper, N. H. and Birmingham, J.T., "Strains and Motions of USS ESSEX (CVA9) during
Storms near Cape Horn," TMB Report 1216 (Aug 1958).
Jasper, N.H. and Rupp, L.A., "An Experirrontal and Theoretical Investigation of Propeller
Shaft Failures," Trans SNAME (1952).
Jewell, D.A., "A Note on Hydroelasticity," JSR, Vol. 8, No. 4 (Mar 1960).
Johnson, A.J., "Vibration Tests on All-Welded and All-Riveted 10,000 Ton Dry Cargo
Ships," NECI Trans, Vol. 67 (1951).

Johnson, A.J. and Ayling, P.W., "Graphical Presentation of Hull Frequency Data and the
Influence of Deck-Houses on Frequency Predictions," NECI Trans, Vol. 73 (1956-57).

Kapiloff, E., "Calculation of Normal Modes and Natural Frequencies of Ship Hulls by
Means of the Electrical Analog," TMB Report 742 (Jul 1954).

Kaplan, P., "A Study of the Virtual Mass Associated with the Vertical Vibration of Ships
in Water," SIT, Davidson Lab Report 734 (Dec 1959).

Kane, J.R. and McGoldrick, R.T., "Longitudinal Vibrations of Marine Propulsion-Shafting

Systems," Trans SNAME, Vol. 57 (1949). Also TMB Report 1088 (Nov 1956).

Kennard, E.11., "Forced Vibrations of Beams and the Effect of Sprung Masses," TMB
Report 955 (Jul 1955).

Kenny, J.H. and Leibowitz, R.C., "Design Details and Operating Procedure for the TMB
Network Analyzer," TMB Report 1272 (Apr 1959).

Kjacr, V.A., "Vertical Vibrations in Cargo and Passenger Ships," Acta Polytechnic?
Scandinavica, A.P. 244/1948; Mech Engin Series 2, Copenhagen (1958).
Koch, J.J. "Eine Experimertelle Methode zur Bestimmung der Reduzierten Masse des
Mitschwingenden Wassers bei Schiffscliwingungen," IngenieurrArchiv, Vol. IV, No. 2 (1933).
Krivtsov, Y.V. and Pernik, A.J., "The Singing of Propellers (Penie Grebnvkh Vintov),"
Sudostroeniye, No. 10 (Oct 1957). TMB Translation 281 (Oct 1958).
Krylov, A.N., "Vibration of Ships," in Russian (1936).
Kurdiumov, A.A., "Ship Vibration," in Russian (1953).
Kumai, T., "Some Notes on the Local Vibrations of Ships," Reports of Research Inst, for
Applied Mechanics, Kyushu Univ, Japan, Vol. III, No. 12 (1955).

Kumai, T., "Estimation of Natural Frequencies of Torsional Vibration of Ships," Reports

of Research Inst for Applied Mechanics, Kyushu Univ, Japan, Vol. IV, No. 13 (1955).

Kumai, T., "Shearing Vibration of Ships," Reports of Rosearch Inst for Applied Mechanics,
Kyushu Univ, Japan, Vol. IV, No. 15 (1955). In English.

Bi-4
 Kumai, T., "Coupled Torsional-Horizontal Bending Vibrations in Ships," European
Shipbldg, Vol. V, No. 4 (1956).

Kumai, T., "Damping Factors in the Higher Modes of Ship Vibrations," European Shipbldg,
Vol. VII, No. 1 (1958).

Kumai, T., "Some Measurements of Hull Vibration and Human Sensitivity to Vibration,"
European Shipbldg, Vol. VI, No. 3 (1957).

Lamb, H., "Hydrodynamics," Dover Publiuations, New York (1945).

Landweber, L. and do Macagno, M.C., "Added Mass of Two-Dimensional Forms Oscillating

in a Free Surface," JSR (Nov 1957).

Leibowitz, R.C., "Natural Modes and Frequencies of Vertical Vibration of a Beam with an
Attached Sprung Mass,' 1 TMB Report 1215 (Sep 1958).

Leibowitz, R.C., "Modes of Vibration of Rudder of USS ALBACORE (AGSS569)," TMB

Report C-952 (Feb 1959) CONFIDENTIAL.

Leibowitz, R.C. and Kennard, E.II., "Theory of Freely Vibrating Nonuniform Beams In
%ALudingMethods of Solution and Application to Ships," TMB Report 131.7 (May 1961).

Lewis, F.M., "The Inertia of Water Surrounding a Vibrating Ship," Trans SNAME, Vol. 37
(1929).

Lewis, F.M., "Propeller Vibration," Trans SNAME, Vol. 43 (1935).

Lewis, F.M., "Propeller Vibration," Trans SNAME, Vol. 44 (1936).

Lewis, F.M., "Dynamic Effects," Chapter 2, Vol. 2, Marine Engin, edited by H.L. Seward,
published by SNAME (1944).

Lewis, F.M., "Vibration and Engine Balance in Diesel Ships," Trans SNAME (1927).

Lewis, F.M. and Auslaender, J., "Virtual inertia of Propellers," JSR (Mar 1960).

Lewis, F.M. and Tachmindji, A., "Propeller Forces Exciting Hull Vibration," Trans
SNAME, Vol. 62 (1954).

Little, R.S., "Bibliography on the Longitudinal Strength of Ships and Related Subjects,"
Tecl and les Bulletin No. 2-3, 6NANiE (Sep 1958).

Loser, U., "Vibration Neutralizer for Ships (Neutralizzatore di vibrazioni per navi),"
Ric. di Ingegn., Vol. 12, No. 1 (Jan-Feb 19,4). EMB Translation 12 (Jan 1936). *

*EMB stands for U.S. Expermental Model Basin. These reports are av;ilable on reque*L to David Taylor
Model Basin (Code 141).

Bi-5
 Lovett, W.L., "The Vibration of Ships," Shipbldg and Shipping Record (Jan 192T).

MacNaught, D.F., Discussion of Paper "Bull Vibration Investigation on SS GOPHER

MARINER," by F.T. McGoldrick and V.L. Russo, Trans SNAME, Vol. 68 (1955).

Mathewson, A.W., "Preparation of Data for Computation of Vertical Flexural Modes of Hull
Vibration by Digital Process," TMB Report 632 (Sep 1949).

McCann, G.D. and MacNeal, R.H., "Beam-Vibration Analysis with the Electric-Analog
Computer," Journal of Appl Mech, Vol. 17, No. 1 (Mar 1950).

McGoldrick, R.T., "A Study of Ship Hull Vibration,"' EMB Report 395 (Feb 1935).

McGoldrick, R.T., "Experiments with Vibration Neutralizers," EMB Report 449 (May 1938).

McGoldrick, R.T., "Vibration Tests on USS HAMILTON at Washington Navy Yard," EMB
Report 372 (Dec 1933).

MlGuldrick, R.T., "Comparison brtween Tlz, recal lv ard Experimentally Dotorminod

Natural Frequencies and Modes of Vibration of Ships," TMB Report 906 (Aug 1U54).

McGoldrick, R.T., "A Vibration Manual for Engineers," TMB Report R-189, Second
Edition (Dec 1957).

McGoldrick, R.T., "Calculation of the Response of a Ship Hull to a Transient Load by a

Digital Process," TMB Report 1119 (Mar 1957).

McGoldrick, R.T., "Calculation of Natural Frequencies and Normal Modes of Vibration for
a Compound Isolation Mounting System," TMB Report 1420 (Jul 1960).

McGoldrick, R.T.. "Rudder-Excited Hull Vibration on USS FORREST SHERMAN (DD 931) -
A Problem in Hydroelasticity," Trans SNAME (1959). Also TMB Report 1.431 (Jun 1960).

McGoldrick, R.T., "Shock and Vibration Instrumentation," Symposium at the 17th Appl
Mecc Div Conf of the ASIME held at Penn State Univ (19-21 Jun 1952).

McGoldrick, R.T., "Buoyancy Effect on Natural Frequency of Vertical Modes of Hull Vi-
bration," JSR (Jul 1957).

McGoldrick, R.T. and Jewell, D.A., "A Control-Surface Flutter Study in the Field of Naval
ArchitetLure," TMB Report 1222 (Sep 1959).

McGoldrick, R.T. and Russo, V.L., "Hull Vibration Investigation on SS GOPHER MARINER,"
Trans SNAME, Vol. 63 (1955). Also TMB Report 1060 (Jul 1956).

McGoldrick, R.T., et al., "Recent Developments in the Theory of Ship Vibration," TMB
Report 739, Revised Edition (Oct 1953).

Bi-6
 Michel, R., "Interim Design Data Sheet DDS 4801, Propulsion Shafting," Dept of Navy,
Bureau of Ships, S1/1-2 (415), DDS 4801 (1 May 1957).

"Military Standard - Mechanical Vibrations of Shipboard Equipment," MIL-STD.167

(Ships) (20 Dec 1954).

Motora, S., "On the Effect of the Free Surfaces upon the Virtual Mass of Submerged Bodies
and Ships," J Soc Nay Arch, Japan, Vol. 86 (Mar 1954).

Neifert, H.R. and Robinson, J.H., "Further Results from the Society's Investigation of
Tailshaft Failures," Trans SNAME (1955).

Noonan, Knopfle, and Feldman, A Proposal for "A Code for Shipboard Hull Vibration
Measurements," in preparation for SNAME.

Noonan, E.F., et al., "Vibration Measurements on Ship S6-2 during Builder's Trials,"
prepared for Hull Structure Committee of SNAME by Noonan, Knopfle, and Feldman,
Professional Engineers (Nov 1959).

Nicholls, H.W., "Vibration of Ships," TINA (1924).

Okabe, T., ot al., "Vibration Measurements of 82,000 DW Super Tanker," Reports of Re-
search Inst for Applied Mechanics, Kyushu Univ, Japan, Vol. IV, No. 13 (1955). In English.

Ormondroyd, J., et al., "Dynamics of a Ship's Structure," Engin Res inst, Univ of
Michigan, Ann Arbor (Mar 1949). Out of print.

Paladino, A.R., "Vibration Survey of USS DECATUR (DD 936)," TMB Report 1271 (Mar
1960).

Panagopulos, E.P., Discussion of Paper "Longitudinal Vibrations of Marine Propulsion.

Shafting Systems," by J.R. Kane and R,T. McGoldrick, Trans SNAME, Vol. 57 (1949).

Panagopulos, E.P., "Design-Stage Calculations of Torsional, Axial, and Lateral Vibra-

tions of Marine Shafting," Trans SNAME, Vol. 58 (1950).

Pavlenko, G.E., "A Method of Calculating Ship Vibration Frequencies," Engineering

(25 Jun 1926).

Pien, P.C. and Ficken, N.L., "The Measurement of Propeller-Induced Vibratory Forces
on Scale Ship Models," Paper presented at ATTC Conference (Sep 1959).

Pipes, L.A., "Applied Mathematics for Engineers and Physicists," McGraw-Hill Book
Co (1946).

Pohl, K.H., "Unstationary Pressure Field in the Neighborhood of a Ship Propeller and the
Periodic Forces Produced in Its Neighborhood" (in German), Schiffsrechnik, 6, 32 (Jun 1959).

Bi-7
 Polachek, H., "Calculation of Transient Excitation of Ship Hulls by Finite Difference
Methods," TMB Report 1120 (Jul 1957).

Pollard, J. and Dudebout, A., "Theoric du Navire," Treizieme Partie, "Vibrations des
Coques des Navires a Helice," Vol. IV, Chapter LXIX, "Causes et Periodes des Vibrations,"
Paris (1894).

Prohaska, C.W., "The Vertical Vibration of Ships," Shipbuilder, Vol. 54 (Oct, Nov 1947).

Ramsay, J.W., "Aspects of Ship Vibration Induced by Twin Propellers," TINA (1956).

Rayleigh, Lord, "Theory of Sound," Second Edition, Dover Publications, New York (1945).

Reed, F.E. and Bradshaw, R.T., "Ship Hull Vibrations. Part II - The Distribution of
Exciting Forces Generated by Propellers," CONESCO Report F-101-2, Contract Nobs 77150
(Jun 1960).

Richards, J.E., "An Analysis of Ship Vibration Using Basic Functions," NECI Trans (1951).

Robinson, Q.R., "Vibration Machines at the David W. Taylor Model Basin," TMB Report
821 (Jul 1952).

Roop, W.F., "Natural Frequencies of Hull Vibration," Trans SNAME (1.930).

Ros~ell, H-1.and Cihpman, L.B., Edit. s,' .:-e .till_:

by SNAME, Vole. 1 and 2 (1939).

Russo, V.L. and Sullivan, U.K., "Design of the MARINER-Type Ship," Trans SNAME,
Vol. 61 (1953).

Rydnr, F.TL. and Zaid, M., "Direct Electrical Analogs of Double 6-Degree-of-Freedom
Systems," Report on Contract Nonr-29552 (00) (X), Scientific Research Staff, Republic
Aviation Corp, Farmingdale, New York (Feb 1960).

Saunders, H.E., "Hydrodynamics in Ship Design," published by SNAME (1957).

Scanlan, R.H. and Rosenbaum, R., "Introduction to the Study of Aircraft Vibration and
Flutter," Macmillan Co, New York (1951).
Schade, l1.A., Discussion of Paper "Hull Vibration Investigation on SS GOPHER
MAINNER," by R.T. McGoidrick and V.L. Russo, Trans SNAME, Vol. 63 (1955).
Schadlofsky, E., "The Calculation and Measurement of Elastic Natural Frequencies of Ship
Hulls (Uber Rechnung und Messung der elastischen Eigenschwingungen von Schiffskorpern),"
Jp.hrh. d. ST(,, Vol. 33 (1932). EMB Translation 7 (Jun 1934) Supplement (Nov 1935).
Schlick, 0., Series oi Articles on Ship Vibration in TINA (1884, 1893, 1894, 1901, 1911).
Schmidt, F., "Die Berechnung von Schiffschwingungen," WRH, No. 6 (22 Mar 1930).

Bi-8
 Schmitz, M.A. and Simons, A.K., "Man's Response to Low Frequency Vibration," ASME
Paper 59-A-200, presented at Annual Meeting (1959).

Srzawa, K., "The Effect of the Distribution of Damping Resistance in a Ship on Her Vi-
bration," J Soc Nay Arch, Japan, Vol. 66 (Jun 1940).

Spath, W., "Theory and Practice of Vibration Testing Machines (Theorie und Praxis der
Schwingungspr~unfmasehinen)," Julius Springer, Berlin (1934). EMB Translation 51 (Mar 1988).

Stodola, A., "Steam and Gas Turbines" (Dampf-und Gasturbinen), Translated by Dr. Lewis
C. Lowenstein, P. Smith, New York (1945),

Stuntz, G.R., et al., "Series 60 - The Effect of Variations in Afterbody Shape upon Re-
sistance, Power, Wake Distribution, and Propeller-Excited Vibratory Forces," Paper pre-
pared for Annual Meeting of SNAME (1960).

Suetsugu, I., "A Contribution to the Vibration of the Stern of Single-Screw Vessels,"
Internatl Shipbldg Progresqs, Vol. 2, No. 10 (1955).
Tachrmind,,t.j.A T and McGoldrick, ,".. "Note on Propeller-Excited Hull Vibrations,"

* N, o . .8 I- 11 L tlJ

Taylor, J. Lockwood, "Propeller Blade Vibration," TINA (1943).

Taylor, J.L., "Vibration of Ships," TINA (1930).

Taylor, J.L., "Ship Vibration Periods," NECI (1927-28).

Taylor, K.V., "Vibration of Ships," Journal of T3SRA, Vol. 6 (1951).

Theodorsen, T., "General Theory of Aerodynamic Instability and the Mechanism of Flutter,"
NACA Report 496 (1934).

Thurnton, D.... , 'Moch,,i:s Applied to Vibration and Balancing," Chapman & Hall, Ltd,
London (1951).
Timoshenko, S., "Vibratio 1 Problems in Engineering," D. Van Nostrand Co, New York
(1955).

Tobin, T.C., "A Method of Determining the Natural Periods of Vihration nf Shipsn," TINA
(1922).

Todd, F.tt., "Some Measurements of Ship Vibration," NECI (Nov 1931),

Todd, F.H., 'Ship Vibration - A Comparison of Measured with Calculated Frequencies,"

NECI (1933).

Todd, F.H., "The Fundamentals of Ship Vibration," Shipbuilder, Vol. 54 (May, Jun, Jul 1947).

Bi-9
 Todd, F.H. and Marwood, W.J., "Ship Vibration," NECI Trans, Vol. 64 (1948).

Tomalin, P.G., "Marine Engineering as Applied to Small Vessels," Trans SNAME, Vol.
61 (1958).

Vane, F.F., "A Guide for the Selection and Application of Resilient Mountings to Ship-
board Equipment," TMB Report 880, Revised Edition (Feb 1958).

Vane, F.F., "Shock and Vibration in Ships," Handbook of Shock and Vibration Control,
Harris, C.M. and Crede, C.E., Editors, McGraw-Hill Book Co. (to be published in 1961).

Vane, F.F., "Natural Frequencies of Side Plating, Bulkheads, Decks, and Radar and Radio
Supports, Measured Aboard Three Cruisers, Three Destroyers, a Submarine, and a Battleship,"
TIMI uiepori. 548 (Apr 1948).

Vedeler, U., "On the Torsion of Ships," TINA, Vol. LXVI (1924).

von Karman, T. and Biot, M.A., "Mathematical Methods in Engineering," McGraw-Hill

Book Co., Now York (1940).
"VuoiOi. H.. "Recent Findings andiric-l . +f. ln. Vhrwtic,
-.. , i' .he FinI1 4 Shin

(Einige neuere Erkenontnisse und Erfahrunger' bei Schirfavibrationen)," J thrb. d. STO, Vol. 47
(Feb 1953). TMBA Translation 268 (Feb 1958).

Waas, II., "Federnde Lagerung von Kolbenmaschinene auf Schiffen," ZVDI, Vol. 18,
No. 26 (1987).

Waas, H., "Erfahrungen bei der Untersuchung und Bekainpfung von Schiffschwingungen,"
Forschungshefte fur Schiffstechnik, No. 3 (1953).

Waas, H. and Walter, H., "Technische Neuerungen auf dem Vermessungsschiff 'Nord' der
Wasser-und Schiffahrtsvorwaltung," Schiffstechnik, Vol. 3, No. 12/18 (Nov 1955).

Wendel, K., "Hydrodynamic Masses and Hydrodynamic Moments of Inertia (Hydrodynamische

Massen und Hydrodynamische Massontragheitsmomente)," Jahrb. d. STO, Vol. 44 (1950). TMB
Translation 260 (Jul 1956).

Weinblum, G., "Uber Hydrodynamische Massen," Schiff und Hafen, Vol. 3, No. 12 (1951).

Whitaker, J., "Vibration in Small Ships," Paper read before the FAO International Fishing
Boat Congress, Paris (Oct 1952).

Wilson, W'.K., "Practical Solution of T-hrsional Vibration Problems," 2 Vols., John Wiley
& Sons, inc, New York (1966).

Wilson, W.K., "A Review of Ship Vibration Problems," Mar Engr & Nay Arch (Jul, Aug,
Oct, Nov 1955). Also published in ASNE Journal (May 1954).

Bi- 10
 Wright, E.A., "New Research Resources at the David Taylor Model Basin," Paper pre-
sented at Spring Meeting of SNAME (Jun 1958). Also TMB Report 1292 (Jan 1959).

Wylie, C.R., "Advanced Engineering Mathematics," McGraw-Hill Book Co, New York (1951).

Yeh, H.Y., "Thrust and Torque Fluctuations for APA 249, TMB Model 4414," TMB
Report 1364 (Jan 1960).

Yoshiki, M., "A Simple Method for Estimating the Natural Frequencies of Flexural Vi-
bration of Ships," J Soc Nay Arch, Japan, Vol. 78 (Jul 1951).

Bi-11
 INDEX
Accuracy of hull vibration calculations, P-i; Arbitrary forcing function, 5-8
3-11; 15-10
Assembly of six degrees of freedom, 6-4
Acoustic radiation, 8-2 to 6-11
Added mass, 2-6 to 2-8; 3-17; 4-1; 5-10; 15-1; Axes
A-2 coordinate, 2-1; 2-2
determination of by model experiments, coordinate for treatment of vibrating hull,
15-7 3-7; 8-10
effect, 11-2
shallow water effect on, 2-8 Axis, shift of in the Stodola method, 3-20
Admittance, mechanical, 4-11; 4-1.2 Baier and Ormondroyd, formula
for number
Advance ratio, propeller, 7-6 rnifirmnt vertiel modo... D-,
Aeroelasticity, 14-1; 14-8
model experimentation in, 15-8 Ba"er, L.A., 10-4
formula for number of significant vertical
Amplidyne speed control for vibration gen- modes of a hull, 4-7
orators, 15-4 Balancing, 9-1; 10-2; 10-9 to 10-11
Amplitude indicating instruments, 15-5 head, 9.4
Analog calculations, 15-2 of propellers and shafting, 10-1
Analog computers, 1-2
use hC in flutter calculations, 14-5 free-free, 2-3; 2-5; 3-1
Analog method radius of curvature of a. 8-9
as applied to transient vibration, 5-9; 5-10 theory, P-1; 1-2; 8-1 to 8-23; 4-1 to 4-18
of calculating forced vibration, 4-5 to 4-7 Beamlike modes, 15-11
of calculating hull modes, 3-10; 3-14 to Beamlike vibrations, 2-3
3-16 characteristics, significance of, 11-2
Analogy Bearing forces, 7-3; 7-4; 7-6; 10-5
conventional electrical, 3-14 Beating effect on multiple-screw ships, 9-7
mobility electrical, 3-14
mobiityeletricl,
-14Beats in propeller-exciting forces, 7-3
.Analysis of hull vibration data, E-3 Beasin opeller-e
Analytical methods of determining normal Bending, of a hull, 2-3
modes, 15-2 Bernhard, R.K., 15-3
Analytical prediction of propeller-exciting Bibliography, general, Bi-1 to Bi-li
forces, 7-4 to 7-7 Blade-frequency force, 7-2 to 7-9; 10-2
Analytical treatment of hull damping, 8-3 to 8-5 Blade-frequency range, effect of, 10-3; 10-4
Analyzers, network, 3-14 "Bop" test, 10-9
Anchor drop test, 8-5; 15-9 Boston Naval Shipyard criterion for vibratory
Angular accelerometer, 15-7 velocity, F-2
Antipitching fins, 14-8 Bottoming, of a resiliently mounted assembly,
Antiresonance, in mobility analog, 8-16; 11-5
4-2 Boundary conditions for freely vibrating
Antivibration devices, 9-1 to 9-8 hull, 3-10
satisfying of in Stodola method 3-21
ApptautIl 1 , `24 Boundary restraints, difficulty of :If!!-
Appendages, effect of, D-4 ment in model tests, 15-9
Applied Mathematics Laboratory of DTMB, A-1

In -1
 Builder's trials, 10-9; E-1 Conventional electrical analogy, 3-14
Bulkheads, flexibility of, 6-1 Convergence of the Stodola method, 8-22
longitudinal, eatment of, A-8 Cordero vibrometer, 1-5.
SBull gear, torsional, vibration of, 12-2
BsiCorrelation of theory and experiment, 15-10
Buoyancy, effect of on vertical modes of hull to 15-11
vibration, 2-1 Couple exciting hull, 7-9
SBuoyancy forces, 5-9; B-2 Coupling of torsional and flexural vibration
Bureau of Ships, U.S. Navy Department, 15-5 of hulls, 15-4
Burrill's formula, for 2-node vertical Coupling of torsional vibration with hori-
frequency, 3-8; C-8 zontal flexural vibration, D-8
Criteria for ratio of stiffness to mass at ends
C.A. PAUL, SS, 8-9 of the hull, A-8
Calculation of forced vibration of hulls, Criteria of validity of theory, 15-10 to 15-11
4-3 to 4-13; 15-2 Critical flutter speed, 14-8
Calculation of hull modes and natural Critical speeds, 10-8
frequencies, 3-1 to 3-24; A-i to A-7 of shafts, 12-1
Calibration of vibration instruments, E-2 Crystal accelerometer, 15-5
Cargo, damping action of, 8-2
Cavitation, effects of in enntrol surface Damping
systems, 14-6 action of cargo, 8-2
Center of mass, 2-2 action of control surface members, 8-7
t 8_8
CHARLES R. WARE, USS, 8-9
action of lifting surfaces, 8-7 to 8-8
Classical flutter theory, 14-8 action of water, 8-1
Coding of digital computer, possibility of analytical treatment of hull, 8-3 to 8-5
changing, A-5 constants, 4-2
Commercial ships, vibration surveys of, data, on hull, 8-7; 8-9 to 8-10
C -tom l s , veffect of diving planes, 8-7 to 8-8
effect of riveted construction, 8-2
Complex notation, 4-11 effect of ruddurs, 8-7 to 8-8
Complex numbers, use of in forced vibration equivalent viscous, 4-2; 8-4
calculations, 4-3 experimental methods of determining of
hulls, 8-5 to 8-7; 15-9 to 15-10
Compound isolation mounting system, 6- frequency dependence of hull, 8-4
Computational methods, rate of development full-scale determination of, 15-9 to 15-10
of, P-2 hydrofoil, 15-10
Computers, analog, 1-2 hysteresis, of hull, 8.2
in propulsion shafting systems, 12-1
Computers, digital, 1-2 in vibration instruments, 15-6
Constants of integration, in Stodola method, negative, 14-3
3-21 of hull vibration, 4-2; 8.1 to 8-11
Control surface Rayleigh, 4-2; 4-9; 4-12; 5-8; 8-2; 8-6
flutter, 8-8: 1fl.-; 14-3 tv 14-7 viscous, i-.2; 8-3
lift force of, 8.7 Data analysis, E-8
members, 6-2 Data, on bull damping, 8-7:8-9 to 8-10
members, damping action of, 8-7 to 8-8;
15-10

In-2
David Taylor Model Basin, P-2; 1-2; 7-8; Dynamics, ship, P-2; 2-1 to 2-9
7-11; 14-6; 15.2
DD9B1-Class Destroyer, 7-11; 7-12; 14-6; Eccentric, adjustable as antivibration device,
D-B; D-4 9-1; 9-8 to 9-4
DECATUR, USS, D-3; D-4 Effective mass, 6-2 to 6-8
vibration survey of, 7-fs
Decay (f free vibration, 8-5 Effective point of attachment of a resilient
mounting, 6-4; 6-6 to 6-7
Deck panels Effective system of one-degree-of-freedom, 8-8
experimental values of frequencies, H-4
flexibility of, 6-2 Effective systems representing normal modes,
4-9
Degrees of freedom, 2-1; 3-13
Design considerations relating to steady- E.J. KULAS, SS, 8-9
state hull vibration, 10-1 to 10-12 Elastic constants, of a set of resilient
Design criteria, hull, 11-2 mountings, 6-6; 6-7
Diesel drives, torsional vibration in, 12-2 Elastic range of hull vibration, B-I
Difference equations Electrical analogies, 3-10; 3-14 to 3-16; 4-5
for forced vibration, 4-8 to 4-7; 5-9; 5-10
for transient vibration, 5-8 as applied to compound imolation mounting,
for vibrating beam, 3-8 of 6-11
beam with sprung mass, 6-8
Digital computers, 1-2; 15-2
impedance, D-
DigitalElectrical
asgappliethod
as fEmpirical
applied to forced vibration, 4-2; 4-8 aspect of ship vibration problem, P-1
to 4-5 Empirical formulas for estimating the level of
as applied to hull with resiliently mounted stern vibration, D-1 to D-4
assembly, 6-11 Empirical formulas for hull frequencies, 3-8;
as applied to transient vibration, 5-6 C-1 to C-7
to 5-9
of calculating hull modes, 8-10 to 8-15 Energy dissipation per cycle, 8-8
of clcuatin
hul moes,3-10to .13 Endurance tests for shipboard equipment,
Dissipation effects in electrical analogs, 4-5 Edrn
F-B; F-4t s,
Divergence, torsional, 14-2 to 14-38- to Engines, resilient mounting of, 18-2
pane, oruc or
dunpng
Divig
Diving planes, damping effect of, 8-7 o Equation, differential, for vibrating beam, 3-2
8-8; 15-10
Driving force, 10-4 Equivalent viscous damping, 4-2; 8-4
Driving moment, 10-4 Euler-Bernouilli beam theory, 8-2; 4-1
Euler-Bernouilli uniform beam, normal mode
Driving point, 4-9 patterns of, 11-5
DTNII, P-2; 1-2; 7-8; 7-11; 14-6; 15-2 Exciting forces, propeller, 7-1 to 7-13
Dynamic balance, 10-11 analytical determination of, 7-4 to 7-7; 15-2
Dynamic instability, 14-3 Expansion joints, effect of, A-2
Dynamic vibration absorber, 9-1 to 9-3 Experimental methods of determining damping,
Dynamical equations, for resiliently mounted 8-5 to 8-7
assembly, 6-7 Extensional deformation of a hul', 2-3
Dynamical matrix, of a resiliently mounted
assembly, 6-9 Fatigue, due to vibration, 11-1
Dynamical theory of hull, general, 11-1

Ini-3
Fin, use of to reduce propeller-exciting Horn's, for frequency of 1-node torsional
force, 7-13 mode of a hull, 8-6; C-6 to C-7
Finite difference equations, for hull with a Ormondroyd, J., for number of significant
resiliently mounted assembly, 6-10; 6-11 vertical modes of a hull, 4-7
Prohaska's, for 2-node vertical mode of a
Finite differenceB method,
4hu A-1 to A-7; B-i to hull11,C-3;-3 C-4-
B-4 a s t -1.8 4 t Sehlick's, for frequency of 2-node vertical
of 4-i5
vibration analysis, to - mode of a hull, 3-8; C-1
Todd and Marwood, for frequency of 2-node
First-order hull vibration, 7-2 vertical mode of a hull, C-5
First-order resonance, avoiding of, 10-6 Forward velocity of ship, effect an damping
Flexible materials, use of to reduce vibration, action of control surface members, 8-8
9-1; 9-7 Frahms reeds, 15-5
Flow-excited vibration, 10-1 Free-free beam, 2-3; 2-5; 3-1
Flexural modes, 2-3 Free-free uniform bar, normal mode of, 3-19
Flexural rigidity, as indicated by natural Free space natural frequencies of hulls, 11-2
frequency, 11-3 Frequency dependence of hull damping, 8-4
Flutter Frequency equation, for
classical, theory, 14-8 assembly,
e 9 a resiliently mounted
critical, 14-8
determinant, 14-5 Frequency indicating instruments, 15-5
model testing for, 6-4; 15-8 Frequency modulation, 15-5
of control surfaces, 8-8; 14-3 to 14 Froude law, 15-8; G-8; U-4
panel, 14-8
sub-critical, 14-6 Full-scale experiments in field of ship
speed, 14-3 vibration, 15-9; 15-10
Forced vibration
analog method of calculating, 4-5 to 4-7 Generalized coordinates, 4-8
calculations, 15-2 Generalized force, 5-3
imagirary numbers, us6 of ir., 4-3 ceneraliz
eonccpt inforc 5-3longitudinal-torsioaal
coupl~d
complex numbers, use of in, 4-3 vibraton oi propulsion systems, 12-3
linear combination, use of in, 4-4
difference equations for, 4-3; 4-4 Generalized mass, 4-8; 5-4
digital method as applied in, 4-3 to 4-5 Generalized system for torsion-bending modes,
normal mode method of calculating, 4-7 15-11
to 4-9 Generalized velocity, 4-8
of hulls, 4-1 to 4-13
experimental determination of, 15-7 GOPHER MARINER, SS, 10-4; 11-3; A-1 to
phase, in, 4-3 A-7; B-i to B-3; C-7; H-6 to H-8
Forced vibratory response, 15-11 exciting forces on, 7-11
Grpicingr ati on, 1-2
Forces, propeller-exciting, 7-1 to 7-13; 15-2 integration, 1-2
Forcng uncton,4-1;5-2Graphical
Graphical methods of calculating hull modes,
Formula 3-10; 3-17 to 3-23
Baier, L.A., for number of significant Gravity forces, B-2
vertical modes of a hull, 4-7
Burrill's, for frequency of 2-node vertical
mode of a hull, C-3 HAMILTON, USS, 8-10
empirical, for hull frequencies, 8-8; C-I to Harmonics, of blade-frequency exciting forces,
C-7 7-2

In-4
Hatch openings, effect of, A-8 Hull design criteria, 11-2
Heave, 2-1; 2-2 Hull girder, structural design of, 11-2 to 11-4
Helmholtz resonator phenomena, 14-1 Hull vibration
Historical background, 1-1; 1-2 analysis, graphical method of, 8-17 to 3-28
Holzer method, 3-10 analysis,
ehdo,-1linear combinations in digital
method of, 3-18
Horizontal exciting force, 7-9 Hull vibration
Horizontal forces, blade frequency, 15-8 buoyancy, effect of on vertical modes of,
Horizontal modes, experimental values of 2-
frequencies, H-3 calculations, accuracy of, 3-11
Horizontal stern vibration, empirical characteristics, i-2
Hormuala for, v n eeffect of isolation mounting on, 6-4 to
formula for, D-3
6-13
Horizontal vibration, sample test schedule effect of local flexibility on, 6-1 to 6-18
for, H-6; 1-7 effect of water on, 2-1; 2-6 to 2-9
Horn's formula, 10-4 sample test schedule for, H-6 to H-8
for frequency of 1-node torsional mode damping of, 4-2; 8-1 to 8-11
of a hull, 3-6; C-6 to C-7 data, analysis of, E-3
Hullbending due to shock, F-2
of a, 2-3 effect of rudders on, 8-7 to 8-8; 9-3; 10-1;
14-3 to 14-7; 15-40
Hull(s) first order, 7-2
couple exciting, 7-9 levels of, due to slamming, 7-1
criteria for ratio of stiffness to mass at normal modes of, methods of calculating,
the ends of the ship, A-3; A-4 3-10
dynamical theory of, 11-1 propeller-excited, 2-5; 10-1 to 10-8
extensional deformation of a, 2-3 rudder maneuvering to excite, 8-5
forced vibration of a, 4-1 to 4-13 steady state, design considerations relating
free space natural frequencies of, 11-2 to, i0-1 to 10-11
hysteresis damping of a hull, 8-2 Hydroelasticity, P-2; 14-1 to 14-9, 0-8 to G-4
loading of hull, effect of on feasibility of model work in, 15-8; G-4
frequencies, 10-7
mechanical impedance of a, 7-1; 4-1.0 to Hydrofoil boats, 14-8
4-13; D-1 to D-4 Hydrofoil damping action, 8-7 to 8-8; 15-10
natural frequencies of a, 3-10; 15-11; Hysteresis damping of hull, 8-2
H-i1 H-2
normal modes of vibration of, 2-3 to 2-6;
3-10 Imaginary numbers, use of in forced vibration
power required to vibrate a, 8-1 calculations, 4-3
rigid body motions of the, 2-1; 2-2; Impact test, 10-9
11-2; 15-5
shearing of a, 2-3; 3-6 Impedance, electrical, D-1
steady-state vibration of, Impedance formulas, for stern vibration, 4-12;
structural design of, 11-2 4-13; 7-11; 7-12; D-1 to D-4
theoretical theoeticl
frql~eey
freqvenc-y oof a !-'2Impedance,
A, 11-2 meeharnicat, 4-10 to 4.-13; 7-1:
torsional modes of, 3-6; 15-7; C-6l D-1 to D-4
t~orsional vibration of, 4-13; 7-11; 7-12;
15-4; 15-7; D-3; E-3 Impedance-type Formula for stern vibrion,
transient vibration of, 5-1 to 5-10; 11-1 to 7-11; 7-12; D-1 to D-4; D-12; D-!3
11-6; 15-10; B-1 to B-4 Impulse excitation, 5-5; 8-5; 10-9
twisting of a, 2-3

In-S
Inortia mLt,4ix. of' - ri.'tiin-ly ri;.nnted v 10-2; 11-1; 11-4; 11-5
assembly, 6-8 vibratiin 'sting of, 15-2
Influence function, 4-7; 4-8 Local vibration, 2-8
Instruments for ship vibration measure- Locking arrangements for resilient mountings,
ment, 15-1; 15-5 to 15-7 18-1
Integral equations, in Stodola method, 8-20 Looking out of resilient mountings under slam-
Integrating amplifier, 15-5 ming conditions, 18-2
Isolation mounting, effects of on hull Logarithmic decrement, 8-3; 8-5; 9-6; 15-9
vibration characteristics, 6-4 to Longitudinal vibration
6-12; 11-5 of hulls, 15-11
of propulsion-shafting system, 12-1 to 12-8
Johnson, A.J., formula of, for ratio of hori- of propulsion systems, coupling with
zontal to vertical frequencies, C-5; C-6 torsional vibration, 12-8
Losenhausen, 1-1; 15-.2
Karma'n vortices, 14-1 Lumped system, 8-1; a-11; 4-3; 5-6
Kineticvorticgs, 14-1;
8approximating a vibrating hull, 8-11
Kinetic energy, 6-3; 8-5 Lmig -
Lumping, A-2
in terms of normal modes, 4-8
Kirchhoff's equations in the mobility analog,
3-16; 4-6 Machinery, propulsion, vibration in, 12-1 to
12-4
Krylov, A.N., 1-1; 15-1
Machines, vibration, 15-2 to 15-4

Lateral vibration of propeller shaft, 12-1; MARIA, SS, 9-3

12-8 Maritime Administration, U.S., 10-7
Levels of service vibration, F-1 to F-4 Mas_
added, 2-6 to 2-8
Levels of ship vibration, 15-10; F-1 to F-4 apparent, 2-6

Lewis, F.M., 7-7; T-1; 7-12; 15-7; 15-8 center of, 2.2
Lift force of control surface, 8-7 effective, 6-2; 6-3
Lifting surfaces, damping action of, 8-7 to generalized, 4-8; 5-4
sprung, 2-8; 64-0 to 6-12
8-8; 15-10 virtual, 9-6
Mass-spring system of one-degree-of-freedom,
Linear combination, use of in forced vibra-
tion calculations, 4-4 s-1 -
Linear combinations, in digital method of Mass-unbalance effect, 7-2
hull vibration analysis, 3-13; 4-4
Loading of hull, effect on frequencies, 10-7 Mass-unbalance of rudders, effect of, 14-8
Local flexibility, effect, of on hull vibra- Masts, flexibility of, 6.2; 11-4; 11-5
tion characteristics, 6-1 to 6-18 Masts, natural frequencies of, 11-4
Local natural frequency, 6-2 Matrix
Local resonance, avoiding o)f, 10-83 10-9 dyuiamical, of a resiliently mounted
assembly, 6-9
vocal ,*x,,ctures, 9-6; D-4 inertia, cr a resiliently mounted
feasibility of model testing of, 16-9 as'embly, 6.8
.r-quencic,3 of, experimental, 11-4 not,4iou, in tr.atment of resiliently
specification of, 10-8 mounted assernblies, 6-6
sprung mass effect of, 6-1 t C-18. 15-11

In-0
 ,ic

stiffness, of a set of resilient Model, wit. ,;- r .--

mountings. 6-8; 6-W . fcr hir .. r.i-,o :.'-'. , :

Measurement of hull vibration, E-1 to E-3 G-I tre G-44

Mechanical admittance, 4-11; 4-12 Mo.ent
o -E iertia of section "
Mechanical impedance, 4-10 to 4-13; 6-2; A-
7-1; 15-10; 15-11; D-1 to D-4 Mo-tfou:i-y, J.B., F-2
Ofaaihull, 7;of the5stern, of Motiuns, rigid body, of a ship, 2-1; ;:
Myklestud n'ithod, 3-10

Medium vibration generator, TMB, 15-2; At

15-3 Natural freqi~encies of hulls
comparisct o cIalculated and mea:'L,.:i, '
Method(s) 15-10; it"; 11 "
analog, of calculating hull modes,- 1p1et s, H'1
3-14 to 3-18 3-14 to 3-16methods of cr.ickilt,ding.
experimentaLt. %\'i..e5of, H1-1; 3-24-\
3-1 toH1-2 1.
digital, of calculating hull modes, m c g 3 -
3-10 to 3-13 Nqtural frequency cr local structure, post-
graphical, of calculating hull modes. bility of changing, 10-8
3-17 to 3-23 Navy, U.S., P-2; 15-10; U-1
Holzer, 3-10
Myklestad, 3-10 Negative dar-ng, 14-3
Prohl-Myklestad, of vibration Network analyzer-s, 3-14
analysis, 3-10 TMB, 3-17; 3-18
Stodola, 1-1; 3-10; 3-17 to 3-23 Neutralizer, .. -- '. 11 3
Mobility analog NIAGARA, USs, s -10
antiresonance in, 3-16 Nodal drive, uu of w avoid rorsiori 'bra-
Kirchhoff's equations in the, 3-16; 3-17 tion, 12-2
oscillator, use of in, 3-17
Nodes, 2-3; 2-4
Mobility electrical analogy, 3-14
Mode, normal. - Nonresonant fc.!-cud respons,, 15-i1
analog method of calculating hull, Nondimensionirnotatior. P-2
3-14 to 3-16 NORD, SS, 9-1T
digital method of calculating hull, Normal mode method
3-10 to 3-13
fiexural, 2-3 as applied to trtrs ien, , ibraion, 5-2 to
graphical methods of calculating hull, 5-5; 8-5
3-10; 3-17 to 3-23 0o calculating; forced vibration,. 4- i 0
of vibration, 2-3 orthogonalit4 -7f, 4-6: ;-2:
3-22; :-22; 4.7;
orthogonality of normal, 3-22; 3-23; 4-7;
4-8; 5-2; 5-3 Normal mode of' vibratio,_ ,
torsional, 2-3; 3-6; 10-4; 15-7; C-6; H-3 Normal mode p.r f Oln dm
,!iiV.l C:<tK itW.
torsion-bending, 2-6; 3-16 to 3-24; 15-7 A-7
Mode of .?ibration, ce.,-3 Normal modes
Model experiments in the fieli of ship vi- compa'ison :)f c,.cufatcd :indl ma:-urc,,i,
bration, 7-7 to 7-10; 15-7 to 15-9 15-12
experir-,e..t err..;in,,tin o' . ,-
Miode! predictiop of exciting fokc,,. 7. 7 ,r)
7-10; 10-5; 15-7 to 15-9 in control surr.<
Model prediction of propeller-exciting forces, of vibration ..;
7-7 to 7-10; 10-5; 15-7 to 15-9 3-1 to 3-24

In-7

Best Available Copy

 Polar symmetry, of a resilient mounting, 6-6
S:,,;l'.\M ),.fv VS. -9: -e-] Pole mast, natural frequencies of, 11-4
h <N-I1
,.. P--. PONTCHARTRAIN, USCGS, 15-7
:onl~r ioral, P..2 Portable vibration instruments, use of, 15-9
,. ufcl..i<. pro Potential energy, 4-8; 8-5
.: - - Potential flow, 2-8
" "m:,rofr
iropellor blades., e f foc t oA, 7 -, Power required to vibrate a hull, 8-1

PRESIDENT HOOVER, SS, propeller-

,'I; ,-------i exciting forces on, 7-12
.I.,-NI"
'OLj) -;ire coefficient. 7-4
t,-u i:.
propeiller, 7-4; I0-9 .' field (fluctuating), 7-2
Nibratory ef(,f'c" of, 7-2; 10-9 Pr..-sure, fluctuating due to water inertia,
,-, - of-fre
-. ,,re '"..~. .- i 2-6
,,,, a n . 4-Principal axes of inertia, of a resiliently
(,'...:i:;Iv:l arnahg3s, 3-14 mounted assembly, 6-8
Or!,on'troyd, J., 10-4 Prohaska's formula for 2-node vertical fre-
O:.mondro\d, J., formlc for nurnhm- rcl sig- quency, 3-3; C-3 to C-5
"tI vvti i )!O' !...oo f a hu1l, 4-7 Prohl-Myklestad method of vibration analysis,
Orthocri alit y of norral iiooes, 3-22: 3i-23 3-1n
4-7; 4-sK 5-92; 5-3 Propellers
Osci!lator Uw( in M!ohllity analog, 3-17 advance ratio, 7-6
' Oscil.ograp,,. t.--5 blades, number of, 10-5; 10-6
muIl'-hai:n [, 15-.) clearances, 7-5; 10-5
excited vibration, 2-5; 2-6; 10-1 to 10-9;
14-1
Pallograph, 15-.5 number of blades, effect of, 7-6; 10-5; 10-6
TMB two co0ipo:on;t, l5-6 one-bladed, 7-2; 10-9
Panel flutter, :,.-8 pitch unbalance of, 7-2; 10-9
e fshrouded, 10-3; 10-5
P-RE MARQUEYTE '., singing, 14-7
Phase oupercavitating, 10-z
in forced vibratioi, 4-3 tip clearances of, 7-5, 10-5
measurement, 15-9 unbalance of, 7-2; 10-9
meter, 15.5 vibratory effect of, 2-5; 2-6; 10-1 to 10-9
relations, 15-9 Propeller-exciting forces
shift, between exciting 'orces 'oP analytical prediction of, 7-4 to 7-7; 15-2
Sdifferent, propeiiers, 7-:-. beats in, 7-3
""shii, between, surfac-3.rd ,aring experimental data on, 7-10 to 7-13
.. '.s 7-3 fin, use of to reduce, 7-13
.. Pitc pr.",, .I . n,,.n fr ,,; ., of, Sv,-1 full-scale determination of, .-. 15_..10
'\_ :.. . of, ~ model determination of, 15-12
SIr.l::.e. -fi,:!. ofpr, 'F'li<.,s.7 - model prediction of, 7-7 to 7-10; 15-7;

Pitching ce ',- on PRESIDENT HOOVER, SS, 7-12

\Pl. syom'tryi synchronizing devices for reducing, 9-5

,,.. ting, experimental values of froquercic,., to 9-7

\B'est
9-7
\ , I-Best Available Copy
.%-
 Propeller shafls Resiliert rn',tIning
lateral vibration of, 12-1; 12-3 effective point of attachment of a, 6-4;
stresses in, 12-3 6-6; 6-7
whirling of, 12-1; 12-8 elastic constants, of a set of, 6-6; 6-7
Propulsion-shafting systems locking arrangements for, 13-1
Sdamping
damopsionghin,
12
in, 12-1 slocking 13-2 out of, under slamming conditions,
longitudinai vibration of, 12-1; 12-2; 12-3 1i2
torsional vibration of 12-28-2
polar symmetry of a, 6-6
vibration of, 12-1 to 12-4 snubbers, use of in, 11-5; 13-1
Propulsion system, vibration of, P-2; 10-2; Resonance, avoiding of, 8-1; 10-6 to 10-8
12-1 to 12-4Reoacaodnof8-;16to08
local. 10-8 to 10-9
Resonance, in wooden model, 7-9
Radit '. tuvature of a beam, 3-2 Resonance magnification, 8-4; 8-5
Range of blade frequencies, 10-8; 10-4; 10-7 Resonant forced response, 16-1
Range of significant hull frequencies, 10-4; Response to transient exvitation, 5-1 to 5-10
10-7
"Rational" theory of ship vibration, P-i; Reynolds number, 15-9; G-4
8M6; 4-1; 15-11 Rigid body motions of a ship, 2-1; 2-2; 11-!;
15-5
Ratios of frequencies of higher modes to the
fundamental, 10-4; C-i Riveted construction, damping effect of, 8-2
R'.tios of freauencies of hull modes, I0-4; C-1 Rolling, 2-2
Rayleigh, Lord, 4:; 5-'2 Rotary inertia, 3-17; B-1; B-2
damping, :-2; 4-9; 4-12; 5-8; 8-2; 8-6 Rotating vectors, 4-4; 4-10; 4-11; 8-4; 9-6
method of rinding natural frequencies, 8-5 Rudders
sysqtem, 5-2; 8-4 damping effect of, 8-7; 8-8
Reciprocity, 4-7; 4-8 effect of on hull vibration, 8-7; 8-8; 9-3;
Reduced frequency, 15-8; G-3 10-1; 14-3 to 14-7; 15-10
Reed tunable, 15-5 flexibility of, 6-2
hydrofoil damping action of, 8-7; 8-8; 9-3;
Resoarch, ship vibration, 1-2; 15-1 to 1 1;: 1.4-3 to 14-7; 15-10
15-!-
Resilient mounting 1lianeuvyuring to excite hull vibration, 8-5
asqe9Oer.bh, 6-, wo 6-13 torsional divergence of, 1412; 14-3
bottoming of a, 11-5 Rudder-hull vibration, 14-3 to- 14-7
dijtial meulod, appliid to hul'. 7.ith,
6-10 to 6-12
dyiiamical equations for, 6-10 to 6-11 Scsdipg considerations in model vibratioi
dynamical matrix of a, 6-9 experimetts, G. 1 to O-4
-fin;t.a difff-rone eq,,.fti!qo -or hull with Suhadlofsky, E., 1-1
a, 0-10
frequency equation for a, 6-9 Schlick, 0., 1-1; 15-1
inertia matrix of a, C-8 - -
matrix notation, ir, treatment of, 6-8 for frequency of 2-node vertical mode of
principal axes of inertia of a, 6-8 B hull, 3-3- 10-87; (>C1
snuibbers for, 11-5 Sea staLe, v'iridtion of vibration 1-orels with,
stiftnziss matrix of a s;et of, 6-8; e-9 H-5
design considerations relating to, 13-1 w, Seaworthiness, structural, 11-1
13-3
effect of. 11-2
effe, of,1 Service vibration,
to1-- miscellaneous information
offc: of slamming with, 13-2 on, H-1 to H-9

ITI-9
Shaft Similitude, laws of, G-1 to 0-4
cylindrical, torsional vibration of, Singing propeller, 14-7
8-4 to 8-6
cylinrl to torsional
cylindrical, rSix-bladed
rigidity of, 8-4 propeller, possible use of to avoid
speeds, use of higher, to avoid hull resonance, 10-8
resonance, 10-8 Skegs, flexibility of, 6-2
Shafting, unbalanced, 2-5; 10-11 Slamming, 2-5; 5-1 to 5-10; 11-1 to 11-6; 15-10;
B-2
Shallow water effects, 2-8; 2-9;
C-3 effect of with resilient mountings, 11-5; 18-2
levels of hull vibration due to, F-1
Shear deflection, 2-3; 3-3; 3-6; 4-1 SNAME, P-2; 1-2; 7-7; 15-1; 15-5; E-1
Shear rigidity, 3-8; A-4 Snubbers, for resiliently mounted assemblies,
Shearing force, 3-7; 3-20 11-5; 13-1
Shearing of a hull, 2-3 Society of Naval Architects and Marine
Shell plating, flexibility of, 6-2 Engineers, P-2; 1-2; 7-7; 15-1; 15-5; E-1
code for vibration surveys of commercial
Ship vibration, experiments, similitude in, codefr i
15-7; -1 04vessels, E-1
Ship vibration
Speed control of vibration generators, 15-4

full-scale experiments in field of, 15-7 Sprung mass effect, 2-8; 6-1 to 6-13; 9-1 to
to 15-10 9-3; 10-8; 13-2; 15-11
instruments for measurement of, 15-1; 15-5 Stability of finite difference calculation, 5-7
to 15-7 Staggering in finite difference calculation,
standardization of, 15-5 3-12
levels of, 15-11; H-1 to H-9
model experiments in field of, 7-7 to 7-10; Standardization of instrumentation for ship
15-7 to 15-9 vibration measurement, 15-5
norms of. F-1. Standing wave, 15-9
problem, em'pirical aspect of, P-1
"rational" theory of, P-1; 3-6; 4-1; 15-11
research, 1-2; 15-1 to 15-12 Statistical methods applied tc transient vi-
surveys, 15-1; 15-7; E-1 to E-8 brations, 15-10
commercial, code for, E-, Steady-state vibration, 4-1 to 4-14; 10-1 to
of USS DECATUR (ft 9), 7-11 10-12
Shock, H-1 hull, design consideration relating to,
hull vibration due to, F-2 10-1 to 10-12
tests for equipment, 11-5 Stern vibration, 4-12; 4-18; 7-12; 10-4; D-1 to
Shrouded propeller, 10-3; 10-5 D-4
Side plating in estimate of v.,srtical shear empirical formulas for, 4-12; 4-18; D-1 to
D-4
rigidity, A-4 impedance formula for, 4-12; 4-18; 7-12;
Sidling, -l- D-1 to D-4
Significant hull frequencies, number of, 4-7; Stiffness matrix, of a set of resilient mount-
" 100-4
Signi?cant. hull frequency range, effect of,
ings, 6-8; 6.9
Sooamto,11 -7t -2

Significant modes. 4-7; 4-9 boundary conditions, satisfying of in, 8-21

number of, 4-7; 4-9; 10-4 constants of integration, 8-21
Similitude in ship vibration experiments, convergence of the, 8-22
15-7 to 15-9 integral equations in, 8-20

II1
Stressee, hull girder, due to vibration, 11-1 Theoretical frequency of a hull, 11-2
Stresses in propeller shafts, 12-3 Theory
Stroboscope, i1-5 beari, P-1; 8-1 to 3-28; 4-1 to 4-18
Strouhal number, 0-8 Euler-Bernouilli beam, 8-2; 8-8
ur elimitations of, P-i; 4-1
use of in flutter experiments, 15-8 4"rational," of ship vibration, P-1; 8-6;
Structural design of hull from the vibration 4-1; 15-11
standpoint, 11-2 to 11-4 Timoshenko beam, 8-8

Structural integrity as related to transient Three-mass, 40,000-lb vibration generator,

vibration, 15-10 15.8; 15-4
Structural strength, relation to vibration Thrust coefficient, 7-5
characteristics, P-3; 11-1 to 14-4 Thrust variation, 7-7; 10-6
Structures, local ship, 2-8; 11-4; 11-5 Time steps in calculation of transient vibra-
Sub-critical flutter, 14-6 tion, B-I
Submarines, P-2; 14-8; 15-8; 15-11; G-4 Timoshenko beam theory, 8-8
diving planes, damping effect nf.. 8-7; Tip clearances o . ,er, 7-5; 10-5
8-8; 15-10
feasibility of model experiments with, TMB network analyzer, 3-11; 8-18
15-8 Todd and Marwood formula for 2-node vertical
Supercavitating propeller, 10-8 frequency, C-5
Superstructures, treatment of, A-3 Torque variation, 7-7
Surface forces, 7-4; 10-5 Torsional modes, 2-3; 3-6; 10-4; 15-7; C-6;
,H-3
Surge, 2-1 experimental values of frequencies, H-3
Surveys, vibration, of ships, 15-1; 15-7; E-1 of hulls, experimental investigation of,
to E-3 15-7
Sway, 2-1 ratLio of frequiencie., 10-4
Symmetry, in vertical modes, A-I Torsional moments, blade frequency, 7-4; 7-9;
7-12; 15-8
Symmetry, vibrational, 3-7; 6-5; 13-1
Synchronizing device, for propeller shafts, Tosnavirtn rigidity of cylindrical shaft, 3-4
Torsional
7.8 Torsional vibration
stern, empirical
at diesel formula f-r, 1;; .-
drives, 12-2
Synchronizing devices for reducing propeller- in
exciting forces, 7-3; 9-1; 9-5 to 9-7 in propulsion systm, coupling with longi-
Synchronizing vibration machines, 15-4 tudinal vibration, 12-3
Systemns of bull gear. 12-2
lumped, 3-1; 8-11; 4-3; 5-6 of free-free ;miform beam, 3-5
approximating a vibrating hull, 3-11 of hulls, 4-13; 7-11; 7-12; 15-4; 15-1; D-5;
mass-spring system of one-degree-of- BA
freedom, 3-1 experimentnl investigation of, 15-7
one-degree-of-freedom, 3-1 impedance formula for, 7-11; 7-12
promp)Nion, vihbrtin ,F, 'P,9.. in-,. ic-I of Propulsion-shafting system, 12-2
to 12-4 of turbine drives, 12-2
Rayleigh, 5-2; 8-4 use of nodal drive to avoid, 12-2
Torsion-bending modes, 2-6; 3-16 to 8-24;
15-7
Tachmindji, A.J., 7-8 experimental investigation of. 15-7
Test schedule for hull vibration charac-
teristics, sample of, H-6 to 11-8

In-11
Transient vibration of propulsion system, P-2; 10-2; 12-1 to
analog method as applied to, 5-9; 5-10 12-4
difference equations for, 5-8 propeller-excited, 4-1 to 4-18; 14-1; 15-11;
digital method as applied to, 5-8 D-1 to D-4; F-1
in impact test, 10-9 rudder-bull, 14-3 to 14-7
in resiliently mounted equipment in, 13-1 service, levels of, F-1 to F-4
instruments for, 15-10 steady-state, 4-1 to 4-18
normal mode method as applied to, 5-2 to stern, 4-12; 4-18; 7-12; 10-4; D-1 to D-4
5-6; 8-5 testing, of local structures, 15-2
of hulls, 5-1 to 5-10; 11-1 to 11-6; 15-10; transient, 5-1 to 5-10; 8-5; 10-9; 11-1 to
B-i to B-4 11-6; 18-1; B-1 to B-4
design considerations relating to, 11-1
to 11-6 VIbrsion analysis, Prohl-Myklestad mpthod
statistical methods applied to, 15-10 of, 3-10
structural integrity as related to, 15-10 Vibration analysis, finite difference method
time steps in calculation of, B-i of, 3-10 to 3-18; 4-3 to 4--
Trunsmission lines in electrical analog of a Vibration generator, 15-2 to 15-4; 15-11
vibrating hull, 3-15 mechanical, 3-16; 7-7; 7-9; 9-3 to 9-15;
10-9; 15-2 to 15-4; 15-11
Traveling wave, 16-9 difficulty of finding fundamental mode
Trials of new ships, 15-1 with, 15-9
Turbine drives, torsional vibration of. 12-2 TMB medium, 15-3
Twisting of a hull, 2-3 Vibration generator
three-mass, 40,000-lb, 15-3
speed control of, 15-4
Unbalance, limits for, 10-9 to 10-11 amplidyne, 15-4
US EMB, P-2; 1-2; 7-7; 7-12; 9-2 Vibration instruments, 15-5 to 15-7; E-1; E-2
U.S. Experimental Model Basin, P-2; 1-2; calibration of, E-2
?-7; 7-12; 92 damping in, 15-6
U.S. Maritime Administration, 107 for transient vibration, 15-10
U.S. Navy, Admni1r-10; 1. portable, 15-9
Vibration machines, 15-2 to 15-4;
15-11
Vibration measurements, E-1 to E-3
Vector, rotating, 4-4; 4-10; 4-11; 8-4; 9-6 Vibration neutralizer, 9.1 to 9-3
Velocity pickup, 15-5 Vibration surveys of ships, 15-7; E-1 to E-3
Vorticai excilting force, 7-9 sample test schedule, H-6 to H-8
Vertical forces, blade frequency, 15-8; 15-13 Virtual mass, 2-6 to 2-8; 3-17; 4-1; 5-10; 15-1;
Vertical vibration sample, test schedule for, A-2
H-8 shallow water effect on, 2-8
Vibration Viscous damping, 4-2; 8-3
beamlike, 2-3 Vortex shedding, 14-1; 14-7
flow excited, 10-1
hull girder stresses d~ie to, 11-1
in provulsion machinery, 12-1 to 12-4 Wake, effect of nonuniformity, 7-2; 7-6
lateral, of propeller shaft, 12-1; 12-3 Water, damping action of, 8-I
normal mode of, 2-8
of local structures, 10-2; 11-1; 11-3; 11-4 Water, effect of on hull vibration character-
of propulsion-shafting systems, 12-1 to istics, 2-1
12-4 Wave equation, one-dimensional, 3-5

In-12
Wave motion, 15.9
Westinghouse standard for balance of naval
equipment, 10-10
Whirling of propeller shafts, 12-1; 12-3
Wilson, W. Kor, 11-3
Wind tunnel, use of in aircraft flutter
testing, 19-8

Yaw, 2-2

In-13
 INITIAL DISTRIBUTION

Copies
Copies
34 CHBUSHIPS C NAVSHIPYD PTSMH
3 Tech Info Br (Code 335)
2 Tech Asst to Chief (Code 106) 5 NAVSHIPYD PUG
1 Lab Mgmt (Code 320)
2 Appl Sci Br (Code 342) 5 NAVSHIPYD SFRAN
10 Ship Noise, Meas, & Red (Code 345) 5 ADMIN, Maritime Adm
5 Prelim Des (Code 420)
2 Mach Des (Code 430) 5 SNAME
5 Hull Des (Code 440)
1 Conversion Des (Code 460) 5 DIR, DL, SIT, I:oboken
I Cruis Br (Code 521) 5 Head, Dept NAME, MIT
1 Aircraft Carrier Br (Code 522)
I Destroyer Br (Code 523) 2 DIR, Appl Mech, SW Res Inst
3 CHONR 5 Genl Dyn Cotp, Elect Boat Div, Groton
i Fluid Dyn Br (Code 432)
1 Struc Mech Br (Code 439) 1 Gibbs & Cox Inc, New York
2 CHBUWEPS (Code R) I J.G. Engineering Res Assoc, Baltimore
2 DIR, USNRL 2 DIR, Inst of Engineering Res, Univ
10 ODRASTIA of Calif, Berkeley
1 Prof J. OrCmondroyd,
Univ of Mich
I OTS, T 1 Dr. E.H. Kennard,' LaVerne, Calif.
2 SUPSHIP, Groton 2 DeLaval Steam Turbine Co., Trenton,
2 SUPSHIP, Newport News Now Jersey
5 SUPSHIP, Seattle I Geni Elec Co, River Works, West Lynn, Mass.
2 SUPSHIP, New York Attn: Mr. M.A. Prohl
2 SUPSHIP, Camden I Westinghouse Res Lab, Pittsburgh
2 SUP HIP, Quincy 1 CONESCO, Arlington, Mass,

2 SUPSHIP, Pascagoula
5 NAVSHIPYD BSN
5 NAVSHIPYD CHASN
5 NAVSHIPYD LBEACH
5 NAVSHIPYD MARE
7 NAVSHIPYD NYK
2 Material Laboratory
5 NAVSHIPYD NORVA
5 NAVSHIPYD PEARL
6 NAVSHIPYO PHILA
*'U.S. GOVERNi;Nr PRINTIN OP'IrE : 0R61
0 - 9271
* ~24
*- v

to> >

(n>

-U U) 0
0 0

CDZ -r

2
t U2
00 0 0

>o

0u
2 t--
u
-g 0 .0

0 a- 0
o~.0

U) '41~

*~e *00 0U 0

011 r- 0 03 U)

0 4).'Z 4
. Z . 0 o 0 0 P

00) a..
* 3 0 33 0 L)
0. - a
- .0 2

0~C
-~~

w CL.! 0>'

e0 oC .-.

;~ *;~co
 0 00

UM2 0
>00-

10IUD Co 1

* .. - ~ 101010,

CL W T-
* )* ~" ~tr.

> 0 0 - 0

0W I'D E oC

0-10.-Z

10 >1 CL. ,,
*~~~ >0.0

oc A

0 N c- In
- - 0
 * S!

*
.~C 2

.~ :'c

0 .Z4
.. C

.ai*
wS

.~

to
Sc~ 0

- a0
Vi

-2

0
o Lo
a.- 2

COP
Bes Ava~jab