AD-767 389

LOAD CRITERIA FOR SHIP STRUCTURAL DESIGN

WEBB INSTITUTE OF NAVAL ARCHITECTURE

PREPARED FOR
NAVAL SHIP ENGINEERING CENTER

MAY 1973

DISTRIBUTED BY:

KI
National Technical Information Service
U. S. DEPARTMENT OF COMMERCE
 ·•·

THIS DOCUMENT IS BEST

QUALITY AVAILABLE. THE COPY
FURNISHED TO DTIC CONTAINED
A SIGNIFICANT NUMBER OF
PAGES WHICH DO NOT
REPRODUCE LEGIBLYo

,,
 SSC-240

0*
00

LOAD CRITERIA FOR SHIP

STRUCTURAL DESIGN

This document has been approved for

public release and sale; its
distribution is unlimited.

SHIP STRUCTURE COMMITTEE

1973

«t-produced b

NATIONAL TECHNICAL
INFORMATION SERVICE
US Department of Commerce
Springfield, VA. 2215?

/
Y
 SHIP STRUCTURE COMMITTEE
AN INTERAGENCY ADVISORY
COMMITTEE DEDICATED TO IMPROVING
THE STRUCTURE OF SHIPS
MEMBER AGENCIES:
UN!HD STATES COAST GUARD ADDRESS CORRESPONDENCE TO:
"AVAI SHIP SYSTEMS COMMAND SECRETARY
Mil ITARY SEALIFT COMMAND SHIP STRUCTURE COMMITTEE
MARITIME ADMINISTRATION U.S. COAST GUARD HEADQUARTERS
AMERICAN BUREAU OF SHIPPING WASHINGTON, D.C.X96X& 20590

SR 198
1 8 JUL 1973
The development of a rational procedure for determining
the loads which a ship's hull must withstand is a primary
goal of the Ship Structure Committee program. In the last
several years, considerable research activity has been
devoted to theoretical studies on the prediction of hull
loads and to measurement of response both on models and on
ships at* sea.

This report describes a first effort into the synthesis of

the results of these diverse projects into a rational design
procedure.

Comments on this report would be

welcomed.

W. F. REA, III
Rear Admiral, U. S. Coast Guard
Chairman, Ship Structure Committee
 Unclassified
Sc< unty Cl.issifu niton

DOCUMENT CONTROL DATA -R&D

(Security f)»»il')fiiiliin nl tl'l-, /„,.»> „( nh-tra, I ant! hulrxitifi atvnotntinn mutt he entrred when fh« nveratl report It < M»»l/lerfl
iHNfi ACTIVITY ft'»rjinr<if* AUfhor) 2«. nrPORT ircuHiTY CLASSIFICATION

Unclassified
Webb Institute of Naval Architecture 3b GROUP

«I POUT TITLE

LOAD CRITERIA FOR SHIP STRUCTURAL DESIGN

« DFiCKiP Tivt MOTE» (Typ» of report and Irreffmlvm date*)

Final Report
3 Aii TMORiSi (Fit*1 name, mlddlm Initial, taut name)

Edward V, Lewis, Dan Hoffman, Walter M. Maclean, Richard van Hooff and
Robert B. Zubaly
« Bf. POB T O A IE 7«. TOTAL NO O' PAGES 7fc. NO OF Rfrj

May 1973 107 130

«a CONTRACT OB GRANT NO 9«. ORIGINATOR'« REPORT NUMBfRUI

NO0024-71-C-5372
b. C BOJr C T NO iSC-240

vh OTHER BTPOF-i- NOIS» (Arty other number* that may ha aaalfined

thla report)

DISTRIBUTION STATEMENT

The distribution of this document is unlimited.

II. SUPPLEMENTARY NOTES 12 SPONSORING MILI T ARV ACTIVITY

Naval Ship Systems Command

13 ABSTRACT

Consideration is given to the critical loads on ships' hulls, as indicated by

possible modes of structural damage and/or failure. An ultimate load criterion is
then set up involving the following bending moments:
Quasi-static wave-induced, vertical and lateral combined.
Still water, including effect of ship's own wave.
Dynamic loads, including slamming, whipping, and springing.
Thermal effects.
The determination of each of these loads is discussed in detail, and the need for
further clarification of dynamic loads is brought out. Methods of combining these
loads, all expressed in probability terms, are considered.
A criterion for cyclic loading is discussed, involving the prediction of the
expected number of combined loads of different levels, as well as the expected shifts
of mean value. A criterion for brittle fracture is also discussed.
Attention is given to estimating an acceptable probability of failure for use
in design. Finally, calculations of loads are carried out for a typical cargo ship,
the S.S. Wolverine State. The loads are then combined in accordance with the pro-
posed ultimate load criterion and compared with the standards under which the ship
was designed.

DD.rr..1473 Unclassified
Security Classification
Unclassified
4SIurity CUwtiflcaHon
KCV WORD»
ROLE I WT

Bending moments
Wave loads
Still vater loads
Thermal effects
Cvclic loads
Dynamic leads
Combining loads
Design of hull girder

Jinrl as.si f ipri

R Securit ' Classification
 SSC-240

Final Technical Report

on

Project SR-198, "Load Criteria"

to the

Ship Structure Committee

LOAD CRITERIA FOR SHIP STRUCTURAL DESIGN

by

Edward V. Lewis
Dan Hoffman
Walter M. Maclean
Richard van Hooff, and
Robert B. Zubaly

Webb Institute of Naval Architecture

under

Department of the Navy

Naval Ship Engineering Center
Contract No. N00024-71-C-5372

This document has been approved for public release and

sale; its distribution is unlimited.

U. S. Coast Guard Headquarters

Washington, D.C.
1973

I
 ABSTRACT

Consideration is given to the critical loac.« r-n .hips' hulls, as

indicated by possible modes of structural damage and/or ;af• ..re- I" is
recognized that of particular importance is the possibility of damage in the
form of compression buckling or plastic flow in tension of one or both
flanges which could lead to ultimate failure. Another mode of failure is
by fatigue, which is important because cracks may occur which must be re-
paired before they propagate to a dangerous extent. A third mode of fail-
ure is brittle fracture, which is particularly difficult to deal with but
can be minimized by control of material quality and use nf the customary
"fail-safe" approach by using crack arresters. Finally, the possibility of
shear and/or torsional buckling requires consideration.

Hence, an ultimate load criterion is set up involving the fol-

lowing bending moments:

Quasi-static wave-induced, vertical and lateral combined.

Still water, including effect of ship's own wave.

Dynamic loads, including slamming, whipping, and springing

Thermal effects.

The determination of each of these loads is discussed in detail, and the

need for further clarification of dynamic loads is brought out. Methods
of combining these loads, ill expressed in probability terms, are considered.

A criterion for cyclic loading is discussed, involving the pre-

diction of the expected number of combined loads of different levels, as well
as the expected shifts of mean value.

A criterion for brittle fracture is also discussed.

Attention is given to estimating an acceptable probability of

failure for use in design. Finally, calculations of loads are carried out
for a typical cargo ship, the S. S. WOLVERINE STATE. The loads are then
combined in accordance with the proposed ultimate load criterion and compared
with the stanoards under which the ship was designed.

-n-
 CONTENTS

Chapter Page

I. INTRODUCTION Edward V. Lewis ... 1

II. CRITICAL LOADS Edward V. Lewis and ... 8

Walter M. Maclean

III. STILL WATER LOADS Robert B. Zubaly ... 14

IV. WAVE LOADS Dan Hoffman ... 21

V. DYNAMIC LOADS Walter M. Maclean ... 32

VI. PHASING OF SLAM AND WAVE LOADS . . . Richard van Hooff ... 33

VII. THERMAL EFFECTS Robert B. Zubaly ... 47

VIII. COMBINING LOADS FOR DESIGN Edward V. Lewis ... 49

IX. SAMPLE LOAD CALCULATIONS Richard van Hooff and ... 63

Robert B. Zubaly

X. CONCLUSIONS AND RECOMMENDATIONS . . Edward V. Lewis ... 81

ACKNOWLEDGEMENTS 83

REFERENCES . 84

APPENDIX A - BIBLIOGRAPHY 92

-in-
 LIST OF TABLES

TABLE PAGE

I MEASURED FULL-SCALE - SLAM LOADS AND MIDSHIP STRESSES 37

II DETERMINATION OF ULTIMATE HULL GIRDER BENDING LOADS 54

III APPROXIMATE PROBABILITY OF FAILURE: "FOUNDERING" (126) . . 58

IV FRACTURES IN STRENGTH DECK AND SHELL PLATING . . . . 59

V STILL-WATER BENDING MOMENTS, 5. S. WOLVERINE STATE 65

VI DATA ON CLOUD COVER AND ESTIMATED TEMPERATURE DIFFERENTIALS, .... 67

NORTH ATLANTIC

VII CALCULATED CHANGES IN STRESS AND EQUIVALENT BENDING MOMENT, .... 61

3. S. WOLVERINE STATE IN NORTH ATLANTIC SERVICE

VIII WEATHER DATA FOR NORTH ATLANTIC 69

IX SUMMARY OF LIFETIME HULL LOADS 5. S. WOLVERINE STATE 73

X COMBINED LIFETIME HULL LOADS S. S. WOLVERINE STATE 74

XI CALCULATED TOTAL COST OF FAILURE 78

XII CALCULATED TOTAL COST OF FAILURE AND DAMAGE 80

-IV-
 LIST OF FIGURES
FIGURE PAGE
1 TYPICAL VOYAGE VARIATION OF MIDSHIP VERTICAL BENDING STRESS, ..... 3
S. S. ESSO MALAYSIA, LOADED CONDITION (3).
2 TYPICAL VOYAGE VARIATION OF MIDSHIP VERTICAL BENDING STRESS, 4
S. S. R. G. FOLLIS (3)
3 TYPICAL RECORD OF MIDSHIP VERTICAL BENDING STRESS, WITH SLAMMING, ... 6
M. V. FOTINI I.
4 HISTOGRAM OF STILL-WATER BENDING MOMENTS, CONTAINERSHIP NEW ORLEANS . . 19
5 TYPICAL STILL-WATER BENDING MOMENTS, TANKER ESSO MALAYSIA . . 19
6 TYPICAL STILL-WATER BENDING MOMENTS, ORE CARRIER, FOTINI L 19
EACH BOX REPRESENTS ONE VOYAGE
7 TRENDS OF STILL-WATER BENDING MOMENT, MAXIMUM VALUE BY ABS RULES .... 20
(1972) REQUIRING NO ADDITION TO SECTION MODULUS
3 COMPARISON OF WAVE STATISTICS: OBSERVED PERIODS AND HEIGHTS 31
9 PEAK-TO-PEAK SLAM STRESS DISTRIBUTIONS IN DIFFERENT WEATHER
CONDITIONS, S. S. WOLVERINE STATE
10 TYPICAL RECORD OF MIDSHIP STRESS VARIATION, M. V. FOTINI L> SHOWING . , 39
FILTERED WAVE-INDUCED AND DYNAMIC STRESSES (3)
11 DEFINITIONS OF STRESSES AND PHASE ANGLES INVOLVED IN SLAMMING 40
12 DISTRIBUTION OF SLAM PHASE ANGLES, 5. S. WOLVERINE STATE 42
13 DISTRIBUTION OF SLAM STRESS, 5. 5. WOLVERINE STATE 42
14 DISTRIBUTION OF WHIPPING STRESS, 5. 5. WOLVERINE STATE 43
15 A TYPICAL DECAY CURVE OF WHIPPING STRESS .... 43
16 HISTOGRAM OF SLAM STRESS ADDITIVE TO WAVE STRESS 45
17 HISTOGRAM OF THE RATIO OF SLAM STRESS TO WAVE BENDING STRESS 46
18 HISTOGRAM OF THE RATIO OF WHIPPING STRESS TO WAVE BENDING STRESS .... 46
19 PLOT OF SLAM STRESS vs. WAVE BENDING STRESS 46
20 TYPICAL LONG-TERM DISTRIBUTIONS OF WAVE BENDING MOMENT FOR SAG AND HOG . 55
21 TYPICAL LONG-TERM DISTRIBUTION OF WAVE BENDING MOMENT, SAG AND HOG, . . 56
WITH THERMAL STRESS SUPERIMPOSED
22 LONG-TERM DISTRIBUTION OF BENDING MOMENT OR STRESS, WITH REVERSED ... 62
SCALE SHOWING CYCLIC LOADING OR NUMBER OF CYCLES OF EACH STRESS
LEVEL IN 0N£ SHIP LIFETIME (108 CYCLES)
23 EXAMPLE OF APPLICATION OF CYCLIC LOADING CURVES TO STUDY OF FATIGUE . . 62
(128)
24 ESTIMATED DISTRIBUTIONS OF STILL-WATER BENDING MOMENTS, 64
25 CALCULATED THERHAL STRESSES, S. S. WOLVERINE STATE 68
26 LONG-TEW DISTRIBUTION CF BENDING MOMENT, LIGHT-LOAD CONDITION 70
27 LONG-TERM DIS.7IBUTI0N OF BENDING MOMENT, FULL-LOAD CONDITION 70
28 LONG-TERM DISTRIBUTIONS OF COMBINED BENDING MOMENTS: WAVE BENDING ... 73
(VERTICAL AND LATERAL) AND STILL-WATER BENDING
29 CYCLIC LOADING "SPECTRA", S. S. WOLVERINE STATE 80

-v-
 SHIP STRUCTURE COMMITTEE

The SHIP STRUCTURE COMMITTEE is constituted to prosecute a research

program to improve the hull structures of ships by an extention of knowledge
pertaining to design, materials and methods of fabrication.

RÄDM W. F. Rea, III, USCG, Chairman

Chief, Office of Merchant Marine Safety
U.S. Coast Guard Headquarters

CAPT J. E. Rasmussen, U^N Mr. E. S. Dillon

Head, Ship Systems Engineering Deputy Asst. Administrator
and Design Department for Operations
Naval Ship Engineering Center Maritime Administration
Naval Ship Systems Command

Mr. K. Mori and CAPT L. L. Jackson, USN

Vice President Maintenance and Repair Officer
American Bureau of Shipping Military Sealift Command

SHIP STRUCTURE SUBCOMMITTEE

The SHIP STRUCTURE SUBCOMMITTEE acts for the Ship Structure Committee
on technical matters by providing technical coordination for the determination
of goals and objectives of the program, and by evaluating and interpreting the
results in terms of ship structural design, construction and operation.

NAVAL SHIP ENGINEERING CENTER AMERICAN BUREAU OF SHIPPING

Mr. P. M. Palermo - Chairman Mr. S. Stiansen - Member

Mr. J. B. O'Brien - Contract Administrator Mr. I. L. Stern - Member
Mr. G. Sorkin - Member
Mr. C. H. Pohl er - Member
NATIONAL ACADEMY OF SCIENCES
U. S. COAST GUARD Ship Research Committee

LCDR C. S. Loosmore - Secretary Mr. R. W. Rumke - Liaison

CAPT H. H. BELL - Member Prof. R. A. Yagle - Liaison
CDR J. L. Coburn - Member
CDR W. M. Devlin - Member SOCIETY OF NAVAL ARCHITECTS & MARINE
ENGINEERS
MARITIME ADMINISTRATION
Mr. T. M. Buerman - Liaison
Mr. J. J. Nachts he im - Member
Mr. F. Dashnaw - Member BRITISH NAVY STAFF
Mr. A. Mai liar - Member
Mr. R. F. Coombs - Member Dr. V. Flint - Liaison
Mr. F. Sei bold - Member
WELDING RESEARCH COUNCIL
MILITARY SEALIFT COMMAND
Mr. K. H. Koopman - Liaison
Mr. R. R. Askren - Member
Mr. T. W. Chapman - Member INTERNATIONAL SHIP STRUCTURE CONFESS
CDR A. McPherson, USN - Member
Mr. A. B. Stavovy - Member Mr. J. Vasta - Liaison

-vi-
 I. INTRODUCTION

RATIONAL DESIGN

For many years the goal of truly rational design of ship structures has been
discussedv and a great deal of research bearing on this objective has been car-
ried out. The concept was describe.£, tor example, in an early planning document
of the Ship Structure Committee (1) , and since the establishment of the Inter-
national Ship Structures Congress (I.S.S.C.) in 1961 it has been regularly dis-
cussed on a worldwide basis by Comm-'ttee No. 10, Design Philosophy. Although this
report is intended only tc indicate progress to date, it is hoped that it will
assist in the advance toward the ultimate achievement of rational design of the
main hull girder.

The concept of rational design involves the complete determination of all loads
or; the basis of scientific rather than empirical procedures, in order that uncertain-
ties may be reduced to a minimum. This approach carries with it the idea that the
response of the stru:ture can also be accurately determined and that arbitrary
large factors of safety, or "factors of ignorance," can be avoided. The concept is
consistent with the modern approach to structural design that considers the "de-
mand" upon and "capability" of the structure. In short, instead of insuring that
a simple calculated design stress is below the ultimate strength of the material by
an arbitrary factor of safety, an attempt is made to determine the demand of all
loads acting on the structure and then the capability in terms of load-carrying
ability — the load the structure can withstand without failure. Of course, this
approach requires a definition of failure, which may be a serious buckle, a major
crack, complete collapse, or a tensile failure (Chapter II). The concept of ration-
al design of a ship hull is believed to be consistent with a probabilistic approach,
which has already been found to be essential for dealing with random seaway load-
ings. Both demand and capability can be expressed in terms of probabilities, and a
satisfactory design is then one in which the probability of failure is reduced to
an acceptably low value. The problem of determining local loads or stresses fo:
detailed structural design is much more complex and is not discussed here.

This particular report deals only with the demand — or loading — on the hull
girder, but an attempt has been made tc formulate it in a manner that is consistent
with the above approach. In due course, with the cooperation of the ship structural
designer, it is anticipated that a rational design procedure will evolve (2).

^Numbers in parentheses refer to References listed at the end of this report.

 »2-
It is not intended to minimize the importance of the conventional empirical
approach to ship structural design which has served the designer, builder and
operator well through the years. But there is currently a substantial need for
a fully rational approach because of such new maritime developments as larger
ships, faster ships, unusual hull configurations (such as the catamaran), and new
materials. Complete and comprehensive load criteria can facilitate the extrapola-
tion of ship design into new configurations, using new concepts and materials.

LONG-RANGE PATTERN OF LOAD VARIATION

It may be useful at this point to describe the typical long-range pattern of

load variations on typi. '. merchant ships as background for the detailed discussion
of the various types of xoads in subsequent chapters.

For completeness, we should perhaps begin with the construction of the ship on
the building berth. Strictly speaking the only loads present are those induced by
the weight of the structure itself. However, there are residual stresses in the
plating and locked-in stresses due to welding, often of considerable magnitude and
sometimes sufficient to lift the bow and/or stern off the keel blocks. The locked-
in stresses are of particular concern where they may exist in combination with other
stresses at a weld defect or notch and under certain conditions could help to pro-
duce a brittle fracture. For other types of failure it seems reasonable to consider
them to be of minor significance to longitudinal strength, since they tend to be
eliminated by "shakedown" or adjustment in service. That is, an occasional high
longitudinal wave bending load — in combination with other loads — may be expected
to cause local yielding in any of the high residual stress region. Upon determina-
tion of this high wave load the structure will tend to return to a condition of re-
duced residual stress.

During launching a high longitudinal bending moment may occur, but this is
usually calculated and allowed for by the shipyard. During outfitting a con-
tinual change of still water shear and bending moment can be expected as vari-
ous items of machinery and outfit are added. The longitudinal still water bend-
ing moment on the ship can always be calculated, but the midship stress will
probably not correspond exactly to this calculated value because of possible built-
in hog or sag residual stresses, and departures of the hull behavior from simple
homogeneous beam theory. In short, the ship is never in a simple no-load condition
nor even in a condition where the absolute value of even the longitudinal 'bending
moment is exactly known. Such a built-in bending moment will not be considered in
this report since it is believed that changes in load while the ship is in service
are of primary significance.

In general the still water hull loadings vary quite slowly. When a ship is in
port there are gradual changes in the bending moments, shears, and perhaps the tor-
sional moments as cargo is discharged and loaded, fuel oil and stores are tc?ken
aboard, etc. During the voyage there are even more gradual changes in mean loadings
as fuel is consumed, and ballast is added or shifted. Typical changes of this kind
are shown in Figs. 1 and 2(3). Finally, at the end of voyage changes resulting
from cargo discharging and loading, plus possible fuel oil and ballast rhanges, will
again modify the bending moments, shearing forces and torsional moments. The lead-
ing changes in port may be considerable and depend on the nature and quantities of
cargo carried on various legs of the voyage. These changes do not show up in Figs.
1 and 2 because the recording equipment zero was customarily readjusted at every

. mmm^tiKnmMHmimimimA^^
 -3-
-T—I—r
i

■
in
a;
S_
+->
00

CD
c
T3
C
OJ

03
CJ

-i-J

CD

■a

e o
o

T" (V)
i. •—-
03

o
CD T-

03 -r-

O C
> O
O

fO "O

■r- T3
CL 03
>> O

LÜ

ID
CD
 -4-

CO

i/1

QJ

XT-
C
QJ

QJ

in

c
o

QJ
en
ro
>^
O

ra
u
a.
>i

a:
CD
i—i
(ISdX) SMrtlS U_

hl&MiiMwntiinn if iiiiir i urnrfiliriin'Mii mmamtmtmMammim wmmm

 -5-
port visit. As explained in an SSC report, "This capability is necessary to pre-
vent the dynamic stress range from exceeding the limits of the instrumentation
system" (4).

When the ship gets under way to go to sea, the first new hull loading to be
experienced — especially if the ship is a high-speed vessel — is the sagging
bending moment induced by the ship's own wave train. This longitudinal bending
moment is a functi- a of ship speed, and will be superimposed with little change
onto other bending moments (5).

Another load variation results from diurnal changes in air temperature^and in

raaiant heating from the sun. The effect is clearly shown in Figs. 1 and 2 .
Such thermal stresses can be explained on the basis of irregular or uneven thermal
gradients, which can perhaps be considered as the "loads." In general, if a beam
is subject to heating that produces a uniform thermal gradient from top to bottom
It will deflect and there will be no resulting stresses. 3ut, if the gradient is
not uniform, stresses will be induced. In the case of a floating ship, the temper-
ature of all the steel in contact with the water will be at the nearly uniform
water temperature, and there will be very little change from day to night. But
the portion of the hull above water will usually be at a different temperature that
changes continually and depends on both the air temperature and the amount of sun
radiation (extent of cloudiness, duration of sunlight, altitude of sun at noon).
In respect to the latter factor, the color of the deck is important also. There is
usually a marked change in stress in the vicinity of the waterline. especially on
the sunny side of the ship, but from the point of view of longitudinal strength the
temperature change of the weather deck — in relation to the underwater hull tempera-
ture — is significant.

Another large load at sea is that induced by the encountered waves (Fig. 3).
This load usually varies in an irregular fashion with an average period of 5-10
seconds, depending on the ship. Not only is there irregularity in wave-induced
loads from one cycle to the next, but there is a pronounced variation in average
level with ship heading and with weather changes during a voyage and from one season
to another. The irregularity of these loads is, of course, due to the irregularity
of the waves at sea. However, the baffling irregularity of ocean waves has yielded
to modern analytical techniques. This was explained by Dr. Norbert Wiener, who
developed the necessary statistical techniques for another purpose. "How could one bring
to a mathematical regularity the study of the mass of ever shifting ripples and
wsves ..,.?," he wrote (6). "At one time ehe waves ran high, flecked with patches
of foam, whils at another, they were barely noticeable ripples What descrip-
tive language could I use that would portray these clearly visible facts without in-
volving me in the inextricable complexity of a complete description of the water
surface, i'his problem of the waves was clearly one for averaging and statistics . ."
In time Wieaer evolved his mathematical tcol, spectrum analysis --a means of break-
ing down complex patteins into a large number of measurable components.

In recent years wave-induced bending moments have been extensively studied,

so that a good statistical picture is beginning to emerge. Research over a number
of years (4) (7) has provided a bank of statistical stress data on four cargo ships
in several services. Using som of these, data it has been found (8) that two dif-
ferent mathematical models can bt used to extrapolate such results to much longer

*
Gages were temperature compensated.

.■■ünMif'-lifii im nin-tiiii - umgjttgüBji -fur. ,ämn\mim*iiJ*±**z^^ima**u*liM

 -6-

Fiaure 3 - lypical Record of Midship Vertical Bending Stress with Slamming,

■M.-V. FOTW'L.

term probabilities. Furthermore, it has been shown that using the same mathemati-
cal models — combined with model tests in regular waves and ocean wave spectra —
short-term (9) and long-term trends (10) can be predicted with a precision that
depends only on the reliability- of the data. At the same time, computer programs
have been developed for applying ship motion theory to the calculation of loads in
regular waves as a substitute for model tests.

Finally, o ceaugoing ships experience dynamic loads, the most troublesome

of which result from impact (slamming) and the vibratorv response (whipping)
that follows it (Fig. 3). In general these loads are transient and therefore are
difficult to de al with statistically. They are superimposed on the previously
mentioned loads Both full-scale measurements (11) and theoretical studies (12)
(13) have been carried out on slamming and whipping, and these have clarified
but not solved the problem. Shipping of water on deck and flare immersion are
other sources o f transient dynamic loading.

Recent attention has been focused on another dynamic phenomenon, springing,

which under certain conditions seems to be excited more-or-less continuously in
flexible-hulled ships, without the need for wave impact. Considerable progress
has been made toward solution by means of theoretical and experimental studies
(14).

All of the above loads will be discussed in detail in subsequent chapters.

HULL LOAD CRITERIA

In general treatises on structural design (15) two types of loading are usually
distinguished: controllable and uncontrollable. In the first case one can speci-
fy design loads with instructions to insure that these are never exceeded. An ex-
ample is a highway bridge designed on the basis of a posted load limit. In the
second case, usually involving natural forces, one must make a statistical analy-
sis and endeavor to design on the basis of the expected loads, with no limitation
on the structure or its operation.

.. .,.-^,-^ia«a^im
 -7-

In the design of ships, still water loads are generally controllable and wave
loads are not. If calculations of typical conditions of loading indicate that ex-
cessive still water bending moments might occur, specific operating instructions
may be issued to make sure that certain limits are not exceeded. The possibility
has been discussed of specifying limiting wave bending loads, as well — somewhat
in the same manner that wing loads on an aircraft are limited by requiring certain
performance restrictions. Such a limit on wave loads for ships could only be ap-
plied if special instrumentation were available to advise the officer on watch when
and if the limiting bending moment is reached, since there is no way for him to
judge this loading unaided. Furthermore, he must have guidance information at hand
that will enable him to take steps to reduce the bending moment if it should ap-
proach the safe limit.

Dynamic loads are partially controllable, since the vibratory espon.*« of the
hull girder can be felt by the Master on the bridge. By a change in ship speed
and/or course he can reduce the magnitude of the exciting forces and thus in-
directly reduce the loads to levels that he has found by experience to be acceptable,

In this report a compromise approach has been adopted regarding statistical

dynamic wave loads. An effort is made to determine all the loads acting on the
ship's hull to provide load cri eria from wnich a satisfactory but economical
structural design can be developed. However, to guard against the possibility of
some unforeseen extreme load condition, it is recommended that suitable stress in-
strumentation be provided as a warning device for added safety (16) .

A great deal of research has been done in recent years on the ship hull load-
ings mentioned in the previous section, much of it in the Ship Structure Committee
(SSC) program. Research under other sponsorship has also contributed to an under-
standing of hull loads, including particularly that supported directly by the U.S.
Navy, the Society of Naval Architects and Marine Engineers and the American Bureau
of Shipping in this country, and by various organizations in Great Britain, Nor-
way, the Netherlands, and Japan, as reported to the International Ship Structures
Congress (I.S.S.C). A partial bibliography is given at the end of this report
(Appendix A).

Some typical loads have received more attention than others, however, leaving
gaps in the overall picture. It is the purpose of this report to present a compre-
hensive and reasonably complete picture of the hull loads and hence load criteria
for ship design, with particular emphasis on dry cargo ships. Hence, considera-
tion will be given in the next chapter to identifying the critical loads of inte-
rest to the ship structural designer. In succeeding chapters each of the various
loads will be discussed in turn, and consideration of typical magnitudes and of pro-
cedures for detailed calculations will b*-. included. Finally the piobiems of combin-
ing these loads for hull girder design purposes will be taken up. Where important
j?;aps in our knowledge appear, they will be identified and recommendations made for
lurcher research. A numerical example for the S.S. Wolverine State will be pre-
sented .

A number of attempts have been made to consider how the available material on
loads can be combined and applied to the rational design of ships. Of these,
particular mention might be made of the work of Caldwell (17), Aertssen (18),
Abrahamsen, Nordenstr^m, and R^ren (19), and of Committee 10 of the I.S.S.C. (20).

1immmtmjmmmim0m
 II. CRITICAL LOADS

INTRODUCTION

Before discussing hull loads in detail it is necessary to consider the dif-

ferent ways that the structure can suffer damage or fail. The object is to in-
vestigate structural aspects of the problem only to the extent necessary to be
sure that all of the necessary information on loading — or demand — will be made
available to the structural designer. In short, we must ask, what are the criti-
cal loads and how do they combine? Meanwhile, it is hoped that work will continue
toward developing a completely rational approach to ship structura1 analysis and
determination of the capability of the structure.

Discussion of critical loads can be facilitated by defining structural failure.

Caldwell (17) considers ultimate failure as the complete collapse by buckling of
the compression flange and simultaneous tensile failure of the tension flange. How-
ever, it is clear that a considerably less severe damage would be a serious matter,
as indicated by such factors as necessity for major repairs, interference with rior-
mal ship operation and non-watertightedness. As pointed out by C. S. Smith in dis-
cussion of (17), "In designing a midship section, the designer should consider the
various levels of damage which a hull girder may experience between the limits of
initial yield and final collapse, and should attempt to relate each level of damage
to an applied bending moment."

Hence, for our purpose we may define damage as a structural occurrence that
interferes with the operation of the ship to the extent that withdrawal from service
for repair is required. Failure is then a severe damage that endangers the safety
of the ship.*

Further study of the subject of critical loads during this project has re-
sulted in no basic improvement in Gerard's analysis of specific ways in which tue
hull girder could f il, as given in "A Long-Range Research Program in Ship Structural
Design" (1), He considered overall damage by compressive buckling, overall ten-
sile yielding, low-cycle fatigue cracking and brittle fracture. To these should be
added combined normal and shear stress buckling, and it is possible to elaborate
somewhat on his scheme and in certain respects to obtain more definite statements.

The types of damage that should be considered then in connection with critical
loads might consist of any of the following:

Damage
•Excessive hull deflection associated with buckling and/or permanent
set.
•Fatigue cracking.
•Brittle fracture, minor or extensive.
•Shear or torsional buckling.

Failure
•Collapse and/or fracture of the hull giraer.

This is sometimes referred to as "collapse" (20), but we feel that this

term connotes buckling failure to the exclusion of tensile failure or perma-
nent set and therefore prefer "failure."

ttumam
iniiMi üir-
 -9-

Although only the last is considered to be structural failure, all of

these types of damage are important for a longitudinal strength criterion. Clari-
fying the nature of these potential damages will assist us in providing the neces-
sary information on loads.

Ine magnitude of elastic hull deflection is usually considered in the design

criteria of classification societies, and it will also be discussed in this chap-
ter.

Finally, consideration should be given to other minor effects, such as the

forces generated by rudders and anti-rolling fins. And other types of service
loading, such as berthing, drydocking, and grounding, which may have direct effects
or, primary hull girder structure, cannot be overlooked. Local damage to structure
that is not part of the main hull girder is excluded from consideiation.

An important consideration in structural design is corrosion. However, since

this is net a load it will not be considered in this report.

PERMANENT SET AND

ULTIMATE FAILURE

We may first consider overall static damage to one of the "flanges" (deck or
bottom) in either compression or tension, i.e., buckling or elasto-plastic yield-
ing. The effect of lateral as well as vertical longitudinal bending and torsion
must be included here. Consideration must be given to the combined effect of
still water bending, wave bending and thermal loads. In addition, a basic ques-
tion is whether or not the superimposed dynamic effects of high frequency "whipping"
following a slam and/or flare entry should be considered, as well as the effect
of wave impacts on the side of the ship and continuously excited springing. It is
quite possible that the short duration of dynamic bending moments — and stresses —
1"mits the amount of permanent set or buckling that they can produce. As noted by
Spinelli, "It should be borne in mind that the short time in which the wave mom-
ents due to slamming develop their maximum values, and the entity of the total de-
flection that would be consequent on them, make the probability of its realization
extremely scarce" (21).

And in referring to plastic deformation, Nibbering states, "In practice these

deflections will not develop the very first time an extreme load of the required
magnitude occurs. The time during which the load is maximum is too short, especi-
ally when a part of the load is due to slamming" (22). This is a problem in struc-
tural mechanics not within the scope of this project, and therefore we shall attempt
merely to identify and evaluate dynamic as well as static loads.

Finally, local loads (not due to longitudinal bending) on which all of the
above are superimposed must not be overlooked. These include deck loads, cargo
loads on innerbottom, liquid pressures within tanks, and external water pressures.

Although there seems to be general agreement on the importance of ultimate

strength, involving extensive plastic yielding and/or buckling, there seems to be
some doubt as to how to deal with it in design. From the point of view of the
present study, however, definite conclusions can be drawn regarding the load in-
formation needed for designing against potential damage of this type.

- 'nii.nriiiniM Tiii-^rMrMMrrMitifir-iiii1^MtTinitfiftiririiiiirM7a«MTliff]MiTrriiim-«iriiir I rini ■ i i i mfr mi ■ ■ irrf-ttTfni ii- mIBM - IHiMMIMJMM

 -10-
FATIGUE

Second is the possibility of fatigue cracking, which seldom constitutes fail-

ure but is important for two reasons: fatigue cracks can grow to the point that
they must be repaired, and fatigue cracks are notches that under certain circum-
stances can trigger rapid propagation as brittle fracture. Nibbering notes, "It
is a favorable circumstance that fatigue cracks propagate very slowly in ships'
structures'* (22) .

The possibility of fatigue cracking is increased by the presence of stress

concentrations — as for example, at hatch corners (23) (24), and it involves con-
sideration of the magnitude of still water bending — i.e., the shift of mean
value — as well as the range of variation of wave bending moments. For example,
a ship may operate with a large still water sagging moment (loaded) on its outward
voyage and with a large still water hogging moment (ballast) on its return, and
such a large variation in mean value needs to be considered in relation to fatigue.
As before, consideration of lateral as well as vertical bending must also be given,
Dynamic loads and vibratory stresses may be expected to contribute to the fatigue
loading.

It appears that the fatigue loading histories of actual ships show considera-
ble variety. Kence, the objective for this study is felt to be simply t:> obtain
clear statistical or probabiliscic pictures of each of the types of loading in-
volved :

1) Probability density of mean still water bending moments, which tenta-

tively and approximately appears to be two normal curves, one representing
outbound and the other inbound conditions.

2) Long-terra cumulative distribution of wave-induced bending moment, which

together with 1) can be interpreted as a low-frequency loading "spectrum."

3) Probability density of high-frequency bending moments associated with

dynamic loads (slamming, whipping an<^ springing). The combination of these
effects with low-frequency loads is a difficult problem, as discussed in
later chapters.

4) Thermal stress conditions, which cause a diurnal change in stress level.

At first glance it appears to be a hopeless task to collect all the necessary

statistical data on the various loads for ships of different types and to develop
ways of combining them that are not only sound by probability theory standards,
but are meaningful from the viewpoint of the mechanics of fatigue and of the
properties of the materials used. A short-cut answer, as proposed by Gerard (1)
would be simply to design to avoid overall combined loads as listed above that ex-
ceed the yield point of the material anywhere In the structure, including areas
of stress concentration. Design of the structure on this basis would virtually in-
sure the ship against low-cycle fatigue, but would possibly lead to heavier struc-
ture than in present designs. Since fatigue cracks can be detected and repaired, it
is not felt that It is necessary to limit stresses to yield point level in this way,
Attempts should be made to understand and evaluate all components of cyclic loading.

ii iiMfiiMrii
 -11-

BRITTLE FRACTURE

Third, is the possibility of failure by brittle fracture. This mode of failure

was common in early days of welded ship construction, but has been greatly reduced
in recently built ships. It cannot be overlooked in a comprehensive scheme, how-
ever. All of the above-mentioned loads apply, including residual and thermal
stresses and ehe notch effect of weld defects. It has been pointed out that a low-
cycle fatigue crack can be the initiation point for brittle fracture (24) .

It is generally recognized that the following factors are involved in brittle

fracture:

(1) Ambient temperature.

(2) Steal characteristics (transition temperature) ,
(3) Notches or stress raisers, including weld defects.
(4) Stress (or load) level.
(5) Strain rate.

Secondary factors include strain as well as stress fields, corrosion effects, metal-
lurgical effects of welding, structural details that introduce constraint, and
residual stresses .

Because of improvements in design and materials, brittle fracture now seldom

occurs in actual service. However, it is conceivable that if, as a result of more
rational approaches to design, working stress levels are increased we may again
have trouble with brittle fracture. Furthermore, it is important to recognize
that brittle fracture has been brought under control by careful attention to mate-
rial qualities, selection and control of fabrication techniques, and inspection
at all stages of construction. Diligence cannot be relaxed, especially as new
materials, new fabrication techniques, and more rational design procedures are
introduced.

Nibbering maintains "that 90% of all ships in the world move regularly and un-
damaged in conditions where the temperature is lower than the crack-arrest tempera-
ture of their steels .... The nominal stresses mostly are so low that with present
day quality of design and workmanship brittle fractures cannot initiate" (22) .

For design purposes the load information needed is generally the same as for
ultir.ci-;-- bending, as discussed in a preceding section, including all dynamic loads,
except that only tensile loads need be considered. Rate of application of dynamic
load.* and ambient temperature conditions should also be specified.

Of the various dynamic loads, it is believed that consideration should be giver

particularly to the midship stress following a bottom impact slam. Since higher
modes than the hull fundamental are involved, the strain rate may be quite high.
See Chapter VI for further discussion.

SHEAR AND TORSION

Fourth is the possibility of shear failure in the hull girder "web." Although
this is a problem in the design of light naval vessels, it has not been of much
concern in more heavily built merchant vessels. This is not to say that shear load-
ing on the side shell or longitudinal bulkheads is unimportant, but rather that
other types of side shell loadings probably constitute more severe criteria of
satisfactory design. Though there is a possibility that the side shell of merchant
ships is excessively heavy, safe reductions in these scantlings can only be made by

-
 -12-
developing more precise way,5 of determining the hull girder torsion and shear load-
ings, as well as lateral loadings due to such aspects of operation as bumping into
dock structures, being handled by powerful tugs, etc.

Another aspect of conceri here arises from the recent development of large
bulk carriers which are frequently loaded only in alternative holds with high den-
sity ores. The result of such loading is that large shear and moment variations ar :
experienced along the vessel's length which must be allowed for in the design of
hull girder structure. Further definition of this problem area is needed, since it
can be expected that large shear and moment, coupled with reduced structural effect-
iveness of the hull girder material, can lead to combined loadings of critical
magnitude.

Torsion is important in relation to both shear and deflection, especially in

wide-hatch ships (25). Excessive hatch distortion has become the major area of con-
cern as progressively larger hatch sizes have been employed. Hence, methods need
to be established for determining the magnitude of torsional loading as a basis for
rational design. In so doing, t;he influence of transverse shear on torsional de-
formation, resulting from the unsymmetric nature cf the ship's structure, must be
included. That is to say, the transverse shear loading must be lefined not only as
to magnitude but as to effective point of application as well, and it must also be
directly related to the torsional loading, since both are developed simultaneously
in any particular oblique wave condition.

To provide sufficient information on loads to carry out a satisfactory analysis

of torsional stresses it is necessary to know more than simply the torsional mom-
ents, since this implies a knowledge of the torsional axis. Hence, for example, in
model tests carried out in regular waves at the Davidson Laboratory for the SL-7
research program, the following measurements were made at the critical sections:

Vertical bending

Horizontal ^hear) About arbitrary but

Vertical shear > knevr. axps
Torque )

Since both amplitudes and phase angles were recorded, this provided the complete
information required for a general stress analysis — provided, of course, that
the number of sections for measurernenn was adequate. Such analysis would, in the
case cf cellular container ships, probably include the intersections of closed
cell systems as well as hatch corners.

It is concluded that shear and torsion need to be considered both as separate

load criteria and in combination with other criteria previously discussed.

DEFLECTION LIMITS

Overall hull girder design may be affected by elastic longitudinal deflection.

Some of the pertinent factors are:

1. Possible damage to shafting piping systems, etc.

2. Effects of deflection on drafts entering and leaving port.
3. Effects of hull flexibility on natural vibration frequency and hence
on springing and whipping stresses.

The question is whether some design criteria should be introduced to limit deflec-
tion in service, aside from the possibility of damage or failure of the structure.

rttHkfictt
 -13-
Direct effects of abnormal deflections on shafting, piping, etc., could no
doubt be provided for in design. Effects on drafts forward, aft and amidships —
hence on load line requirements, bottom clearances, etc, — could be dealt with by
special attention to loading condtions, perhaps with the help of additional
arrangements for ballast. But the effect of hull flexibility on dynamic structural
response requires further consideration. It has been established by the work of
Kline (26) and others that the increase in natural period associated with greater
hull flexibility is favorable from the viewpoint of slamming and the vibratory
stress, or whipping, that follows a slam. However, such may not be the case for
the more continuous vibratory response referred to as springing- Evidence to date
suggests t.'iat the latter phenomenon is increased by increasing hull flexibility.

In the past deflection has been limited by restrictions on length/depth ratio.

For example, the rules of the American Bureau of Shipping require that special con-
sideration be given to any design for ocean service in which L/D is greater than 13.
Whether or not such a severe limitation is necessary a&v never been clearly estab-
lished, but there can be no doubt it has prevented difficulties from deflection in
mild steel ships.

The question of deflection generally arises, therefore, only with considera-

tion of unusual ship proportions or when a material other than ordinary mild steel
is to be utilized. For example, a recent Ship Structure Committee report (27)
develops tentative criteria for aluminum alloy constrjetion of a bulk carrier, and
in addition to specifying section modulus requirements determined by strength con-
siderations it discusses the necessity for a midship moment of inertia value that
will limit deflection. It is stated that, "The only guidance in this area at
present is the ABS requirement that the hull girder deflection of an aluminum ship
shall not be more than 50 per cent greater than that of a 'Rules' steel vessel,
while Lloyd's and Bureau Veritas suggest no increase." The report itself does not
agree, however. "It is concluded that no limits should be placed on the hull
girder deflection of an aluminum bulk carrier, but that the effects of the deflec-
tion resulting from normal structural design should be considered in the areas
noted above." A study of the report indicates that no consideration was given to
the possibility that springing stresses would be aggravated by the increased flexi-
bility (and hence longer natural period of vibration). It is felt that any elimina-
tion of deflection limits should be qualified by a provision that a study be made
of the possibilities of serious springing.

A similar situation arises when extensive use of high strength steels is made.
If full advantage is taken of their higher strength, then greater flexibility and
hence the possibility of springing must be considered. A study of design procedures
for high strength steeds has 'been made (28), which accepts classification society
limits on deflection.

In the present report, in which dry cargo ships are under consideration, de-
flection is not often a problem. Such ships are volume limited, and hence L/D
ratios are quite low. This is especially true of container ships in which there
appears to be a trei i toward increasing depth in order to reduce the number of con-
tainers stowed on open decks. If high strength steels or aluminum is extensively
used, a check should perhaps be made of stiffness and vibration frequency. No
further detailed consideration of the problem is felt to be necessary for the pres-
ent purpose ot establishing hull load criteria.

A special cr.se of obiectionable deflection previously mentioned is the exces-

sive distortion of hatches, resulting from torsional hull moments, which may cause
loss of watertiphtness.
 -14-

SUMMARY

The consideration of critical loads and hull deflections leads to the conclu-
sion that information on the following loads is needed for rational longitudinal
strength design:

Statical still water bending loads, mean values and variation.

Thermal effects on hull girder.

Wave bending loads, both extreme values affecting hull girder damage (or
failure) and the cyclic loading picture affecting fatigue, including
shear and torsion.

Dynamic, loads, both extreme and cyclic, with phase relationships, dura-
tions and rates of application.

Each of the above will be discussed in turn in the succeeding chapters.

111. STILL WATER LOADS

INTRODUCTION

There were two aspects of the subject of still water bending moments studied
in this research project on load criteria. One was a review of available proced-
ures for calculating such leads and the other was a semi-statistical study of still
water loadings on typical ships, Both of the above studies will be reviewed in
this chapter.

CALCULATION OF STILL WATER BENDING MOMENTS

Broadly speaking, ehe techniques available for determining longitudinal static

still water bending moments in ship hulls can be classed in two categories:

1. Approximate methods, used primarily in the early design stage for

determining required scantlings before detailed lines and weight dis-
tributions are known, or used by ship's officers to determine changes
in the bending moment caused by variations in the distribution of cargo
and consumables aboard the ship.

2. Exact methods requiring detailed hydrostatic dat3 (Bonjean curves)

and weight distributions, used primarily by a design agent or shipyard
to determine the bending moments expected in service and to produce load-
ing guides for the ship's of cers.

Approximate Methods

Many variations of approximate calculation techniques have been presented in

the literature on longitudinal bending moments. Descriptions of the methods pro-
posed by W. J. M. Rankine (1866), John (1874), Vivet (1894), Alexander (1905),
Suyehiro (1913), and Foerster (1930) are given by Murray in his 1947 paper (29),
in which he also develops a simplified method for calculating the bending moment at
amidships for vessels of normal form. Mandelli (30) introduced the concept of "in-
fluence lines" which makes possible the quick tabular calculation of the midship
bending moment for any condition of loading in still water.

,___ ——nuiTiwriMinr
_J_^i_^-^»-:
 -15-

He also extended these approximate methods to include calculation of the com-

plete bending moment curve. It is worth noting that all of these methods, as well
as the exact calculations, may be used to determine static wave bending moments as
well as static still water bending moments. Indeed most of the techniques were
developed for the express purpose of determining bending moments in some standard
static wave profile.

Most of the tabular forms and graphs included in ship loading manuals and in
instructions for the guidance of ships' officers, which they may use to compute
midship bending moments, are based on the "influence lines" approach (30). Accu-
racy of the results is quite good (so long as the influence of trim on the buoy-
ancy distribution is accounted for in the graphical data furnished) because the
buoyancy and moment of buoyancy are known exactly for the completed ship. Less
accurate calculations using the same techniques ar? possible without detailed
hydrostatics by approximating nits influence lines and giving their equations as
functions of ship form (31) . For the complete still watei bending moment and shear
curves to be determined without detailed hydrostatics, the Faresi "integral factors
method" ^f approximation is also available (32).

Exact Method

Since the digital computer has come into general use in ship design, the de-
tailed methods customarily used in final ship design have become almost as easy to
use as the approximate ones. The latter remain useful only if,

(a) a computer is not available — as on shipboard (usually),

* (b) detailed data are not available — as in early design.

For other purposes, exact calculations of still water bending moments in various
conditions of loading are ehe rule. The basic metnod is well known and the princi-
ples involved are not at all complicated, but the numerical work is voluminous and
tedious if done by hand.

Briefly stated, the still water shear (or static wave shear) at any point in
the ship's length is calculated as fellows:

V = I W dx = I w dx —I b dx
o "o o

where V = shear force at distance x from bow (or stern),

w - weight per unit length

b = buoyancy per unit length,
W = *.; — b, load per unit length«

The integrations (summations) are performed from the bow (or stem) to station x.

For determining the bending moment at any point,

M = / / W dx dx * / /w dx dx - I' j b dx dx
J o JQ J0 J0 JQ JQ

'k<\mm*mi»taiaiämammäasi<m™.r~,^-^~-. , ....*,.-■*■■ , .... —.— «rtiMatimmmmt

 -16-
where M ■ bending moment at station x. The double integrals are the summations
of the moments of weight and buoyant forces forward (or aft) of station x, caken
about station x as an axis.

It is clear that the above equations can be easily evaluated numerically us-
ing a digital computer. In fact, the simplicity of the equations is the reason
for the fact that a large number of programs to calculate still water (and wave)
bending moment are available to the designer today. For example, nine firms and
organizations have furnished abstracts of their bending moment computer programs
to the SNAME index, T & R Bulletin No. 4-1.3 (33). In addition, various shipyards
and ship operators are known to have operating programs.

The first step in evaluating the integrals is the "balancing" of the ship for
a specific weight distribution, involving calculation of the displacement and the
longitudinal center of buoyancy by integrating the area under the sectional area
curve and taking first moments. The correct mean draft and trim can then be de-
termined by trial and error. The only difference between the process for still
water or for a static wave is the profile o_ the waterline, i.e., a straight line
in the first case or a specified mathematical wave shape in the latter. The sec-
ond step is evaluating the integrals for as many values of x as may be needed.

Another mathematically convenient program to calculate the above is one based

on a mathematical description of the hull which requires very little input informa-
tion and limits the numerical integration to the minimum necessary. Such a program
has been developed at Webb Institute based on the use of mapping coefficients to
describe the two-dimensional ship section (34). In this new method the computations
are programmed to give the bending moment and shear force anywhere along the hull
for any draft and trim (or for any mathematical"' /-defined wave profile). Hence,
comprehensive investigation of a wide range of still water loadings is possible with
a short computing time. The program, designated WTS 130, is presently being docu-
mented.

Electrical Methods

We have learned about two electric instruments for calculating ship longirudinal
stresses and/or bending moments:

Loadmaster - Kockums

Lodicator - Gotaverken

The former is particularly good because it shows visually a graph of the bending
moment distribution along the ship's length. One can see immediately the effect of
a change in load on the bending moment curve.

STILL WATER BENDING MOMENT TRENDS

A pilot study has be_n made of still water bending moments for three ships of
different types: a containership, a supertanker and a bulk ore carrier. The object-
ive was to obtain enough actual still water bending moments for each ship In the out-
bound and inbound loading conditions to evaluate their statistical distributions,
including mean values and standard deviations for outbound and inbound voyages sepa-
rately.
 -17-
Sources of Data

Ideally, the proposed program of calculations would require complete voyage

loading diagrams for many voyages of each ship, plus detailed hydrostatic data, in-
cluding Bonjean curves. The hydrostatic properties are always made available to
the ship owner by the designer or shipyard, and voyage loading diagrams of one kind
or another are in common use by ships' operating personnel. However, the availa-
bility and adequacy of such loading diagrams or tables for the purpose of calculating
bending moments varies considerably among different operators.

In many cases, especially for tankers and ore carriers in ballasted condition,
loading data are not sufficiently detailed tc permit accurate assessment of bending
moments. The total amount of ballast is usually recorded, but its actual distribu-
tion is left to the judgment of the ship's officers, who are not required to record
the quantities allocated to each ballast tank. Nor are records of ballast shifts at
sea during tank cleaning operations retained. Therefore, significant variations in
still water bending moment (SWBM) may actually occur which cannot be calculated from
recorded voyage data. To a lesser degree there are similar omissions of certain
items in the loaded conditions as well, usually for items whose influence on bending
moment is small. By contrast, rather complete loading information was obtained
for a containership for both outbound and inbound voyages.

Calculations Made to Date

The number of voyages for which loading information (complete or not) were ob-
tained is as follows:

Containership New Orleans (Seattle - Alaska)

59 outbound voyages
J ) „m.»i , , , -,
,n . , , * f SWBM's could be calculated.
60 inbound voyages '
Tanker Esso Malaysia

? l°***d™/fes „ } SWBM's could be calculated.

1 balxasted departure

Ore carrier Fotini L.

5 loaded voyages - incomplete data, but SWBM could be estimated.

7 ballasted voyages - data too incomplete to estimate SWBM.

Because of the paucity of reliable data on actual voyages of the latter two
vessels, their loading manuals were consulted for "standard" loading conditions. The
still water bending moments in these standard conditions were therefore included in
the results discussed below. Since these loading conditions were intended for the
guidance of shipboard personnel, they should be representative of actual practice.

Techniques and Results

As a first step toward a statistical description of the still water bending mom-
ents (SWBM's), histograms have been prepared showing the frequency of occurrence of
different values of SWBM's for outbound and inbound (or ballasted) voyages. The
"maximaMplotted in the histograms are the maximum values along the length of the ship,
which generally occur near, but not necessarily exactly at, amidships. The many-
peaked ore carrier bending moment curves required special treatment, as described be-
low. Additional actual voyage data would be required if we were uo proceed any fur-
ther in the analysis of the tanker ana o*e carrier. Special notes regarding the
 -18-
methods used in the calculations foilow.

Containership New Orleans. The owner furnished detailed calculations of mid-

ship bending moments In static L/20 wave profixes, hogging and sagging, for the
actual cargo loading conditions of 59 outbound (Seattle to Anchorage) and 60 in-
bound (Anchorage to Seattle) voyages. Two modifications were made to the calcula-
tions furnished in order to arrive at the SWBM's:

1) The still water bending moment was approximated as the mean of the static
wave bending moments in hog and sag, and was determined to be either hogging
or sagging.

2) The resulting approximate SWBM was adjusted because the consumables (fuel
and fresh water) assumed in the wave BM calculations were "burned out" for
the outbound voyages and "full" for the inbound voyages, rather than the con-
ditions actually listed for the given voyages. Actual tankages were therefore
substituted in the calculations to adjust them to actual conditions. Results of
the adjusted SWEM calculations are shown In Fig. 4.

Tanker Esso Malaysia. Two loaded voyages and one ballasted condition were
available, the SWBM's having been calculated by the owner using his own computer
program. Sufficient data to do the same for other runs were not obtainable. The
other three load and r.hree ballast conditions indicated in Figure 5 are "standard"
conditions from the leading manual. Each standard condition represents either a
departure or arrival condition, whichever has the larger SWBM.

Ore ship Fotini L. All SWBM's plotted in Figure 6 are from the loading manual,
since available data on actual voyage loadings were insufficient to calculate
actual SWBM's. When the vessel carried heavy ore in holds 1, 3, 5, 7, and 9, with
the remaining holds empty, the bending moment curve has many peaks, the highest
peak oftei occurring relatively far from amidships. Instead of one, there are
several ma, Lma in these cases. The upper plot of Fig. 3 shows the value of the high-
est peaks of the SWBM curves occurring within the midship 20% of length. The lower
plot shows the highest peaks occurring within the midship 40% of length. It is seen
that in a number of cases the peak value occurring outside of the midship 20% of
length is higher than that within. In other cases, there are no significant peaks
at all within the midship 20% of length.

It is tentatively concluded that for containerships a single distribution curve

for still water bending moments can be established in design for a particular ser-
vice. In the case of bulk carriers, two distribution curves are usually required —
one for loaded and one for ballast condition.

CLASSIFICATION SOCIETY LIMITS

It should be noted that although classification societies do not in general

base hull girder strength standards on a direct addition of still water and wave
bending moments, still water moments are taken into account.

For example, the current Rules for Building and Classing Steel Vessels of the
American Bureau of Shippingt. 1972, require an increase in deck section modulus if
the maximum still-water bendiig moment in the governing loaded or ballasted condi-
tion is greater than

s c f B (Cb + 0.5)

I,. -—r-"-
 -19-

STILL WATER B M IHOGGINGI. 10OO FT TONS

i ^J O'JTBOu. D VOYAGES |S9 VOV

^^J INBOUND VOYAGES (fiOVUV I

FIGURE 4 - Histogram of Still-Water

Bending Moments Container-
ship NEW ORLEANS

EACH SPIKE REPRESENTS ONE VOYAGE

LOADED BALLASTED

STILL WATER B.M.. MILLION FOOT TONS

FIGURE 5 - Typical Still-Water Bending

Moments, Tanker ESSO MALAYSIA

MIDSHIP 20% LENGTH

3+

_EJ3 EtLliQ
S.WB.M THOUSAND FOOT TONS

MIDSHIP 40% LENGTH

E_ 0 B

HEAVY ORE IN ALTERNATE HOLDS

COAL OR BAUXITE IN ALL HOLDS
BALLASTED

FIGURE 6 - Typical Still-Water Bending

Moments, Ore Carrier, FOTINI L.
Each box represents one voyage

—i iiiiiiinriüiMMM
 -20-

where s and f are coefficients depending on length (see below), B is breadth in

ft., C, is block coefficient (not less than 0.68), and c ■= 1.00, except for oil
carriers, where it is 1.03.

s s
Length, ft. £_ for oil carriers for others

600 594 3.78 A.25

800 1175 3.78 4.25
1000 1921 3.90 4.38
The trend of still water bending moments with ship length on the basis of the above
is shown in Fig. 71

INSTRUMENTS

Our study of still water loadings showed that in many ships a wide range of
bending moments can be experienced in service. Hence, a shipboard instrument for
quickly calculating still water bending moments can be an important adjunct to safe
ship operation. A recommendation is made by Lloyd's Register: "In order to guard
against high stresses being imposed through an unsatisfactory cargo or ballast load-
ing, the Society recommends that an approved instrument or other means of determin-
ing the suitability of loading be placed on board" (35) .

Various instruments of this type are available, as discussed at the beginning

of this chapter.

BENDING
MOMENT TYP ICAl
x10~6 TAN KERS

IT- /
TON-FT

sx TYPICAL C/l RGO

j
r ~
.... L ...
800
LENGTH. FT.

FIGURE 7 - Trends of Still-Water Bending Moment,

Maximum Velue by ABS Rules (1972)
Requiring no Addition to Section
Modulus

HI -MI mm am
in 11 ■■- ■■■ mmaimaämkilmlmämtametUiäe^
 -21-

SHIP'S OWN WAVE

When a ship proceeds at appreciable speed in calm water, as previously noted,

a wave pattern is generated which causes the ship's own bending moment. This may
be a hogging moment in the case of a full ship pushed to high speed, but in fine,
fast ships — where the effect is most pronounced — it is a sagging moment created
by the bow and stern waves.

Results of systematic model tests by Vossers give the trend of bending moments
due to a ship's own wave over a range of block coefficients and speeds (36).

Model tests with destroyers have shown (37) that when waves are encountered
the effects of the ship's own wave and of the ocean waves are superimposed with
very little interference.

SUMMARY

It has been shown that still water bending moments can readily be calculated
by available techniques in the design stage. But relatively little data in sta-
tistical form are available for actual service loadings of various ship types,
particularly in ballast conditions.

IV. WAVE LOADS

INTRODUCTION

The problem of wave-induced loads on a ship at sea is that of determining suc-

cessive conditions of dynamic equilibrium of forces and moments acting in and on
an elastic body moving in the irregularly disturbed interface of two different media.
This problem can be simplified by considering external loads only, on the underwater
part of the ship, which is considered to be a rigid body in an ideal fluid. Motions
and other ship responses in waves are regarded as linear functions of wave height,
and both the irregular waves and the irregular responses can be considered as the
sum of many sinusoidal functions. Hence, the analysis begins with the study of har-
monic oscillations of a rigid body, moving at forward speed on the surface of an
ideal fluid under the action of regular gravity waves.

Though in principle the ship motion problem has been solved for three-dimensional
cases (38)(39), the analytical solution is limited to forms such as a sphere or an
ellipsoid. In view of this, a less rigorous strip theory solution haj been devel-
oped which is suitable for long, slender bodies, where each cross section of the ship
is considered to be part of an infinitely long cylinder. Hence, a series of indi-
vidual two-dimensional problems can be solved separately and then combined to give a
solution for the ship as a whole. The idea was originally introduced by Korvin-
Kroukovsky (40) and has since been endorsed, criticized and improved by many authors
(41)(42)(A3).

The main drawback of the strip theory is that it neglects the mutual inter-
actions between the various cross sections, which are of particular importance
for certain frequency ranges, depending on the size of the body. Hence in waves
that are either very long or very short relative to a ship the theoretical justifi-
cation of strip theory is somewhat questionable. This statement is particularly
applicable to lateral motions, since the hydrostatic restoring force is small or
non-existent under these circumstances.

i iiiiiii liiTTHIliilll'
 -22-

In spite of the above reservations, the basic strip theory has been found to
be very satisfactory for heave and pitch motions and bending moments in head waves
(40), and it is the only suitable method for numerical computation. Modifications
have included the use of "close-fit" methods, which have led to a significant improve-
ment in the computation of the sectional added mass and damping coefficients for all
but the simplest sections. An additional major contribution to the theory has been
the inclusion of all the forward speed terms in the equations of motion in order to
satisfy the symmetry relationship proved by Timman and Newman (A4). All the modi-
fied strip theories developed in the past two to three years (45)(46) have practi-
cally identical forward speed terms. Extension of the theory to oblique waves, to
lateral motions, torsional moments, and lateral bending moments has also been
achieved, as shown in (39j(43). Finally the use of close-fit mapping techniques and
strip theory for determining the distribution of pressures over the hull has been
demonstrated (47)(48).

Since we are concerned with successive conditions of dynamic equilibrium, it

should be noted that a complete solution of the problem of wave loads and bending
moments cannot be obtained without first determining the motions.

LOADS IN REGULAR WAVES

Basic Theory

In order to evaluate the state of development of r,hip motion and load calcu-
lation in waves a short analysis of the basic approach of most investigators will
first be given. The mathematical formulation of the problem, i.e., a ship advanc-
ing at constant mean speed with arbitrary heading into regular sinusoidal waves,
can be presented in most general form by defining the velocity potential so as
to satisfy the assumptions of the ideal fluid, linearized theory. At this initial
stage no strip theory assumption is required. The time-dependent part of the poten-
tial can be decomposed into three components representing the potentials due to
incident wave, defraction and the mode of motion considered, as in the original
theory by Korvin-Kroukovsky (40). However, an additional time-independent term due
to steady forward motion of the ship has been added in more recent theories (43) .
Once the formulation of the component potentials is completed, the hydrodynamic
forces and moments acting on the hull can be determined. Using the Bernoulli equa-
tion the pressures in the fluid are defined and expanded in a Taylor series about
the undisturbed still water position of the hull. Ignoring steady pressure terms
at first,the linearized time-dependent pressure on the hull can be formulated and
integrated over the hull surface. The hydrodynamic forces and moments can be ob-
tained in two superposable parts: those associated with a wave passing a restrained
ship (excitation) and those acting on a body forced to oscillate in calm water.

In order to obtain a numerical solution the application of strip theory approxi-

mations are necessary for the integration of the sectional exciting and motion-
related forces over the length of the ship. These sectional forces involve two-
dimensional added mass, damping and displacement terms. The speed-dependent coef-
ficients are expressed in terms of a speed-independent variable, which is evaluated
by means of strip theory, and of a speed-dependent term which is obtained from a line
integral along the waterline as given by Stoke's theorem.

Hence, the main difference between the original strip theory in (40) and the
more recent "new" methods is in the formulation of the problem. In (40) the strip
theory assumptions were applied in the initial formulation, and the forward speed
effect was only introduced in certain terms. In the "new" theories the assumptions
with regard to strip theory were made after the general terms for the coefficients
in the equations of motion were determined, including the forward speed terms.
 -23-
In addition to the above, the theory presented in (A3) includes end terms in the
coefficients associated with the aftermost sections, which are not usually included
in the strip theory and are claimed to be important for bluff bodies. These terms
are independent of the strip theory assumptions.

Using either the old or the new approach, the formulation of hydrodynamic forces
and moTnonrs ro»rmirs the equations of motion to be solved and the amplitudes and phase
angles of motion determined. Then the longitudinal distribution of all forces —
including those that are dependent on the motions and forward speed — can be evalu-
ated and shearing forces and bending moments calculated for any instant in the motion
cycle, usually at midship. In general the solutions for two instants of time suffice
to determine the amplitudes and phase angles of these quantities.

In general, design calculations, full-scxle data collection, and model tests con-
centrate on conditions amidships. This is a sound procedure, particularly for col-
lecting statistical data on different ships at sea. But some consideration must also
be given to the longitudinal distribution of loads (48A). There are two questions to
be answered by suitable trial calculations or experiments:

1. Are bending moments ever significantly higher at any other section

than midships?
2. Over what ship length do midship bending moment values extend before
they begin to taper off significantly?

It is felt that these questions can be answered by a limited number of system-

atic calculations, using the technique just described, on the basis of regular waves
of various lengths, without the need for collection of long-term data or calculation
of irregular wave responses.
Computation in Regular Waves

The preceding section treated the overall problem of ship motions, shear forces
and bending moments in a seaway, indicating the need for a strip theory solution if
numerical computations are required. In the following, the specific computer pro-
grams available for the above computations and the theories associated with them
will be discussed.

When referring to an available computer program indication will be given of the

degree of availability in terms of whether the program is public property or whether
it is of proprietary nature.

The two programs in the first category are "SCORES," developed by Oceanics, Inc.
under SSC Project SR-174, and the MIT program developed under sponsorship of the
Maritime Administration. Though the basic equations of motion are identical for the
two programs, the scope is somewhat differant. SCORES calculates the vertical and
lateral motions and loads and torsional moment, while MIT is limited to the verti-
cal longitudinal plane only, but includes additional information such as approximate
mean added resistance in waves. SCORES is documented in a recent Ship Structure
Committee report (49) and the MIT program in two MIT reports (50)(51).

Although the basic equations of motion for pitch and heave are identical in
both of the above programs, there are slight differences in the coefficients of the
equations, as well as in the excitation forces and moments. Both programs are based
on coefficients originally derived in (40) and later modified slightly by (41). The
only difference in the coefficients is in the restoring force coupling terms which
are corrected in the MIT program to account for the fact that the origin of the ship
coordinate system is taken at midship rather than at the more conventional location
at the center of gravity. The programs are therefore virtually identical for

niriittTür" ■—
 -24-
vertical motion and loads when the center cf gravity is near midships. It should
be noted that the routines used to calculate the two-dimensional added mass and
damping coefficients are also identical in both programs and are based on (42).

Note should be made of the fact that the MIT program includes a special rou-
tine to handle bulb sections which cannot be properly mapped by the commonly used
"Lewis" routine (52). However, both programs lack a general routine to handle any
shape, such as can be obtained by "close-fit" techniques. SCORES includes the mo-
tion and loads in the lateral plane for which the two-dimensional properties are
calculated according to (46).

An additional program available to the public through the NSRDC is the "YF-17."
This program is limited to motions in the longitudinal vertical plane and is based
on the same equations and coefficients as the above two programs. The main feature of
this program is its close-fit subroutine which allows an exact mapping of any re-
quired section and therefore is of particular value for ships of non-conventional
section shapes. The YF17 is the most advanced available program at NSRDC. A new
program based on the theory in (43), which includes ship motions and loads in three
dimensions, as well as a new strip theory approach, is not yet available for public
use and is now being tested and modified. In the case of certain types of ships,
such as Naval destroyers or very full tankers, a combination of elements from two
or three of the above programs may prove to yield the best results.

It should be noted that all of the above programs supercede the original David-
son Laboratory program based on the theory in (40). In addition to the latter,
several less-known programs, generally not available to the public, also exist. The
University of California has a computer program called "SEALOAD" for the calculation
of dynamic loads (pressures) at discrete points on the hull of a ship for a range
of frequencies, speeds and headings. It is based on motion in the vertical and the
lateral directions. A similar program for calculating the transverse as well as
longitudinal pressure distribution on the hull due to motion in head seas is avail-
able at Webb Institute (47).

The above survey does not include programs presently in use outside the U.S.A.,
mainly because of lack of information with regard to the details. Such programs
exist in Japan (46), Germany (O), the Netherlands (41), Norway (43), and U.S.S.R.,
but are generally not available for distribution, although specific runs can usually
be purchased.

For most of the purposes of this project the SCORES program (Oceanics) is the
only suitable one available. Consequently, it has been adapted to the Webb computer
facilities, and is now regularly in use. It will be a simple matter to upgrade
this program in scope or in detail as new developments appear. Each section is sup-
posed to be amenable to a conformal transformation from which the hydrodynamic co-
efficients can be derived for vertical and lateral oscillations. The particular
coefficients used are:

Vertical oscillations: Grim (42)

Lateral " Tasai (46)

The hydrodynamic forces are then obtained for each strip after the motions are cal-
culated. The integration of the difference between hydrodynamic forces and gravity
forces yields the shear force and hence the bending moment for a particular position
of the wave along the length of the ship. The results for all wave lengths and
headings are the response amplitude operators as a function of the wave circular
frequency per unit wave height.

Kiiniiiirr- ■ - -■■rii 'in wiimiMMifmiii ~— " •"—"" ■""--**—«

 -25-
Comparison of Theoretical and
Experimental Bending Moment

A comprehensive comparison between calculations by program SCORES and model

basin test results in regular waves has been made by Oceanics, Inc. (53). Results
are presented graphically for the cases listed below, covering a range of wave head-
ings in all cases.
Load Bending Moment
Conditions Speeds Vertical Lateral Torsion Shear

Wolverine State (54) 2 2 x x

Series 60,
Block 0.80 (55) 1 1 x x x x

Series 60,
Block 0.70 (36) 1 4 x x

T-2 tanker (56) 1 5 x x x

The following conclusions were d^rawn from the graphical comparisons:

"The comparison between calculations of vertical and lateral bending

moments and the experimental results for the Wolverine State indicates
generally very good agreement. This holds for both loading conditions,
both speeds, and over the range of wave angle and wave length. The
experimental results shown for lateral bending moment in head and fol-
lowing seas, where lateral motions and loads should be zero as in the
calculations, are regarded as indicative of thä possible error, or
range of discrepancy, to be expected between calculations and experi-
mental results. These loads are believed to arise in the model tests
due to its free-running, but rudder controlled, condition. That is5
the model may undergo small lateral motions, with rudder corrections
to keep course, which leads to the measured lateral bending moments.

"The comparison for the Series 60, block 0.80 hull for vertical and
lateral bending moments indicates excellent agreement, in general ...
The agreement for torsicnal moments is only fair and indicates exces-
sive response at roll resonance conditions. The agreement for the
shear forces is quite good, in general, with the exception of some
deviation in lateral shear at 110° wave angle. However, the shear
forces are generally small at midships, and should really be investi-
gated at the quarter-length points."
For the Series 60, block 0*70 hull form and the T-2 tanker,

"In general, the agreement is fairly satisfactory, considering the

factors involved in the experimental comparison. With regard to
this point, consider the double peak calculated vertical bending
moment response for the T-2 i ,:aker at 120° wave heading and 1.65 fps
model speed. While the corresponding experimental data do not indi-
cate such a response, similar double peaked responses for vertical
bending are confirmed by experimental results for Wolverine State,
full load and the Series 60, block 0.80 hull. The greater resolution
of the test data due to testing at more wavelength conditions for
these latter cases tends to produce such results, thereby limiting the
utility of the experimental points for the T-2 tanker as a complete
measure of bending moment variation" (53).
 -26-

Bending moment calculations for the Wolverine State carried out by means of
Program SCORES are compared with experimental results in (53).

Another available comparison between theory and experiment is for the tanker
Universe Ireland (57) . Model tests were carried out at the Davidson Laboratory (58)
and calculation using a modified SCORES program were made at Webb Institute, both
under the sponsorship of the American Bureau of Shipping. Vertical bending moment
results for the head seas case in both full-load and ballast conditions indicate ex-
cellent agreement between theory and experiment over the range of wave lengths
tested. Similarly good results were found for 150°, 30°, and 0° headings, full load.

For the 60° and 120° headings, the comparison is not as good as for the above-
mentioned cases. Although similar trends are maintained, the differences in magni-
tude at some wave lengths are rather large. A possible explanation of the above
discrepancies is the variation between the actual and desired heading angle for
the free running model in oblique waves. It has been previously shown (59) that
deviations of up to 10° in the heading angle can occur for certain wave lengths.
This effect will be most pronounced for the 60° and the 120° cases, and any small
change in heading angle for these cases may result in a large shift in the response
curves, as proved to be the case.

An extensive comparison was made between Series 60 model results obtained at

Wageningen and strip theory calculations (60). The vertical bending moment compar-
isons for head seas covered block coefficients of 0.60, 0.70, and 0.80, and for
C„ * 0.70 three values of L/B, three values of L/H, three values of gyradius, four
forebody section shapes, and four heading angles (1 case only). Agreement with the
basic Korvin-Kroukovsky strip theory was excellent except for some higher speed
cases and cases of large gyradius and extreme V forebody.

Extensive comparisons are given by Faltinsen (61) for an 0.80 block coefficient
hull for which Wageningen model tests were available (55). Using the basic Korvin-
Kroukovsky theory (on which SCORES is based), agreement was good except at 50° and
130° heading angles — as in the above results for Universe Ireland. However, us-
ing the more refined procedure developed by Salvesen, Tuck and Faltinsen (43),
vertical bending moment results were good at all angles. The refined theory gener-
ally gave better results for vertical shear also.

Since comparisons between model tests and theory showed generally good agree-
ment in vertical bending and shear, the theoretical calculations are believed, to
be satisfactory for the present stage of development of load criteria, especially
when it is remembered that the integration of results over a range of ship-wave
headings averages out the result and reduces any errors considerably. But further
refinements in SCORES are needed along the line of Salvesen (43) and greater pre-
cision in determining lateral bending and torsional moments is needed.

Rudder Forces

Kaplan and Raff point out that "since the lever arm of the rudder forces is
large for moments at midships, it appears that rudder forces can significantly af-
fect the lateral bending and torsional moments, To the extent that the use of the
rudder affects the overall ship motion response in oblique seas, the vertical bend-
ing moment also can be influenced, but to a much smaller degree" (53). Rudder ef-
fects are not included in SCORES, but they are taken into account by Grim and
Schenzle (62). However, the rudder effect is not separated out so that its rela-
tive importance can be assessed. They also included the effect of anti-rolling
fins on torsional moment, which produced a significant reduction. It is felt that
 -27-
rudder effect can be neglected insofar as vertical wave bending moments are con-
cerned, but it should be given further attention with reference to lateral bending
and torsion.

LOADS IN IRREGULAR WAVES

The extension of regular wave results to short-crested irregular seas, by means

of the superposition principle, was accomplished by St. Denis and Pierson (63), on
the assumption that both the irregular waves and the ship short-term responses are
stationary stochastic processes. The computer calculation of ship response was
obtained some time ago, using model test data as inputs (64). In the case of all
the above-mentioned computer programs, however, irregular wave capabilities are in-
complete. The MIT and YF-17 programs have been extended to long-crested seas as
represented by the ° ^rson-Moskowitz one-parameter sea spectrum family, and sta-
tistical response parameters can be computed from which various seakeep ng events are
determined. The SCORES program has been extended by Oceanics, Inc., to include re-
sponse to short-crested seas, as veil as a limited option for the use of a two-
parameter spectrum fc .mulation (dependent on both wave height and period). These
are features which are considered to be absolutely essential in order to avoid erron-
eous results which may be obtained otherwise, such as near-zero longitudinal bend-
ing moment in beam seas. Furthermore, it has been previously shown (65) that super-
position of responses to irregular waves should not be limited to a single spectrum
for each sea condition (i.e., a one-parameter family), but a group of spectra should
be used in order that both the mean response and its standard deviation can be. de-
termined for each range of wave height. Such p'rograms are available using as in-
put spectra either ordinates of actual spectra obtained for the North Atlantic or
tables of probability of occurrence of certain combinations of wave height and pe- -
iod, which are substituted in the ISSC two-parameter spectrum formulations. These
programs, designated as WTS 120 and WTS 121, respectively, were recently used to
extend the SCORES program at Webb as additional options available to the user.

The present status of wave data will be reviewed in a subsequent section.

In the numerical exmmple given in Chapter IX, the Webb "H-family" of spectra,
based on wave spectra from (66), is used iti conjunction with WTS 120. For each
spectrum the standard spreading function was applied:

S(u) ~ — cos2 u
7T

where u represents the wave direction angl.. The RMS response corresponding to
each wave spectrum was obtained from,

(RMS): mJ J S(u) S(u) I R M\2 du dy

-TT/2

Then the mean m and standard deviation s. of RMS values within each wave height
group j were determined.

LONG-TERM COMPUTATIONS

The final step of ship loading response calculations is to extend the avail-

---■■-■■■ —--■ - -
.-,.,...,.^,.,J^.^,.^„: _ gaaaai—1t^jmmmaiam ,t tmttmätnmmimjlt^ rmwmn
 -28-
able calculations for limited periods of time in specific irregular sep conditions
to long-term, predictions, covering the lifetime of a ship or a fleet of ships. The
mathematical model and program WTS 62 were described in detail in (10) . The only
additional required input to this program is a distribution of sea conditions for
the particular ocean area of interest.

The mean m and the standard deviation s for each wave height group, together
with the probability of occurrence of that group'p,, give us the necessary paramo
eters to obtain the long-term distribution:' •*

P(xo) -V,:
*lV)ifoP(X°;
p j #p(x0; RMS) p(RMS; m , Sj) d RMS dxo

where p(x ; RMS) ■ Rayleigh distribution of x , for a particular value of RMS, and
p(RMS; m.° s.) * Normal distribution of RMS. °The technique is applied to the
shear forces, and to vertical, lateral and combined bending moments.

It is believed that the incorporation of the above programs, and the various
modifications to SCORES, have produced the most comprehensive program presently
available in this country for use in wave load aspects of ship design.

It should be noted that the long-term probability density function can be ex-
pressed as follows:

P (x0) = £ Pj f p(xo; RMS) p(RMS; mj, Sj) d RMS

"" o

This function is of importance, for two applications, in particular:

1. For combining with the probability density function for still water
bending moment. (See Chapter VIII).

2. For combining with the probability density function for structural

capability to determine the probability of failure (31>.

The present status of wave data is reviewed in the next section.

WAVE DATA

Wave data available for calculation of bending moments have been obtained In
two ways: by actual measurement of the surface elevation over a period time and
by observation of wave heights and periods. Each of these sources will be re-
viewed in turn.

Wave Records

The largest source of data available under the first category is the NIO wave
records collected by means of Tucker wave meters on British Weather Ships over the
past 18 years in the eastern part of the Atlantic Ocean. The records are available
in the form of strip charts. A typical example of one type of analysis which can
be performed on such data is given in (67). The analysis is limited to three basic

it mi»» Mm
 -29-
parameters, the peak-to-trough mean height, the mean zero-crossing period and the
mean-crest period. Other parameters, such as the significant wave height, or the
spectral width parameter can be approximated from the above measured parameters.

Strip chart wave data, though basically suitable for any type of further analy-
sis, ar' 'n practice rather expensive to transfer to the frequency domain. However,
in ordt ... extract maximum information out of such records a spectral analysis
should i.' performed in order that the results can be given in a form representing
the distribution of wave energy of the irregular wave components over frequency,
i.e., an energy spectrum. This constitutes the ideal form of wave data presenta-
tion, for ideally if one is to predict the performance of a ship under certain
environmental conditions the actual spectrum, or a family of spectra representing
different possible wave heights at that location, should be used.
Very few sources of wave records reduced uO spectrum form are available in
quantities adequate for statistical sampling. The main source of data is (66) and
(68). The former ?< a biased sample of 460 records selected from the North Atlantic
weather stations representing mostly fully-developed storm conditions. The latter
constitutes a sample of 307 records all taken at one station at midday over a period
of 12 years, using a stratified sampling procedure. These records were selected at
Webb Institute from National Institute of Oceanography (Great Britain) records, with
the assistance of Professor Pierson of New York University, and were spectrum analyzed
by the National Research Council (Canada). Correlation of results with Beaufort No.
and other anslyaes are being carried out currently at Webb Institute.

An area for which more measured wave data are needed is the North Pacific.
It is hoped that several projects now under way will produce some useful data.
Other sources of ocean wave spectra are Scctt (69), Fukuda (70), and Yamanouchi
(71).
An additional parameter often associated with spectra is the normalized average
period slope parameter, K. The value of is an indication of the deviation of the
actual average period of the spectrum from that of a fully developed Pierson-
Moskowitz spectrum and was originally introduced in (72). By defining K, as well
as the significant wave height and period, some further indication as to the. nature
of a spectrum is provided. A typical distribution of K values against wave height
was recently given in (73) based on 307 records collected at Station INDIA in the
North Atlantic.

A limited amount of data has been obtained by means of a buoy measuring wave
slope (74). Analysis of the records permitted the determination of directional
spectra that give the distribution of wave component directions as well as fre-
quencies .

It must be recognized tha*", shoaling water has a significant effect on wave pat-
terns. The waves tend to become steeper as they travel ir. from deep to shallow
water, Although this effect is noticeable only i.i water depths of magnitude compar-
able to wave length, there are a number of ocean areas where the effect is signifi-
crnt for large, modern ships. Of particular importance is the continental shelf at
the entrance to the English Channel. In general, the relatively shallow waters of
the North Sea, Irish Sea and English Channel are characterized by short, steep seas.
Some wave data including representative spectra are available from British sources
(75)(76).

Wave Observations

A much more common way of describing the world oceans, because ot its relative
simplicity and low cost, is by means of the observed heights and periods of the

»j i- -i--r: -iirr'iinfITifai-» •-^-■»'"■"—laiife—üaaJB

 -30-
waves and their frequency of occurrence. Both the wave height and period can be
estimated to varying degrees of accuracy by seagoing peronnnel. The most reliable
data of this type are those available in reports of trained observers on board
weather ships, as well as at various permanent weather stations around the world.

However, the most common source of observed data is the reports of various
ships sailing throughout the world's oceans. The most comprehensive such collec-
tion of data is given in (77), which covers most of the ocean areas. Other
sources are (78)(79)(80)(81)(82)(83)(84). Similar, more localized tables are avail-
able for specific areas of interest such as the Great Lakes (85), Cape of Good Hope,
(86) etc. The information is usually given in tabular form showing the probability
of occurrence of the various possible combinations of wave height and period. An
alternative form of presentation is by means of a histogram.

Figure 8 gives a comparison of the results given by various sources, mostly for
the North Atlantic. Local data for the vicinity of the Cape of Gooi Hope is also
included, in order to show how a local condition may differ from commonly used data.
This ocean area is of considerable interest because large tankers regularly traverse
it between the Persian Gulf and Europe, and long, heavy swells are known to occur
tb^re at times. More data are needed, but the new information helps to round out
the picture.

The interpretation.of observed wave height and period in terms of the signifi-
cant wave height and the average period is not easy. It has been found, however,
that for the untrained observer a simple approximate relation between the observed
wave height Hy and the significant wave height H-^/o can be established (87) . Sim-
ilar relationships can also be established for trained observers. When wave records
are available it is customary to refer to the period as the average zero-crossing
period.

Mathematical Formulations for Spectra

In order to make use of observed wave data for predicting ship motions or wave-
induced loads, it is essential to make use of some idealized spectrum formulation.
The description of the sea by means of two basic observed characteristics, i.e.,
wave height and period, can be achieved by means of a mathematical expression of
the spectrum in terms of these two parameters. The basic parameters are mathematic-
ally defined in terms of the area under the spectrum and the moments of the area.
Several such formulations are available, and the most common one is usually referred
to as the ISSC spectrum formulation (88).

For each pair of values for wave height and period a spectrum is defined in
terms of the. spectral ordinates at discrete values of the frequency, u>. Hence, for
a matrix of wave heights and periods a family of spectra can be obtained for which
the probability of occurrence of each spectrum can be defined and applied as a
weighting factor. Though the mathematical formulations do not completely describe_
the variability of spectral shape, they can be used for long-term predictions of
bending moments with reasonably satisfactory results (10).

Wind Parameter

In the absence of any of the above information, i.e., the actual spectra or
observed wave heights and periods, the only way to describe the sea is by means of
the wind speed, which can be considered to be the single most important factor in
generating waves. It is apparent that wind speed alone is not adequate for an ex-
act description of the sea, since other factors such as duration, fetch, tempera-

iMWMttftfilWfrlff'--*-■^-'-n illllliitlTM»lii"aMl,^aiJ^ua
 -31-

7 to 9 9 to 1' 11 to 13
VISUALLY OBSERVED PERIOD GROUPS T¥.SECONDS

'.

lb to 3 0 3.0 to 4.5 4 5 to 6 0 6.0 to 8.0 >8.0

VISUALLY OBSERVED WAVE HEIGHT GROUPS Hv, METERS

FIGURE 8 - Comparison of Wave Statistics:

Observed Periods and Heights

tures, etc., are important. However, wind speed or Beaufort No. is a relatively
easy parameter to measure and is commonly available in most weather charts. Rela-
tions between wind speed and wave height have been suggested in several references
such as (69) (71). It is always meaningful to relate the wind speed to a mean signi-
ficant wave height, so long as the standard deviation is also given. One should
realize, however, that in lakes or gulfs such data are less meaningful because of
the limited fetch conditions which prevail.

The relationship between wind speed and wave period is much less significant
and not so easy to define. However, if the significant wave height is available
some typical relationships between it and the wave period can be given for specific
ocean zones.

FUTURE DEVELOPMENTS

The short survey presented here indicates that a great deal of progress has
been made in recent years in the theory of ship behavior in waves, and a variety
of computer programs is available to apply this theory to specific problems. How-
 -32-
ever, these programs are only partly coordinated, and only a beginning has been
made in showing how different programs can be combined for design use in the area
of wave loads. It is felt that further work should be done in documenting, evalu-
ating and improving individual programs, in comparing results in specific cases —
among programs, and between programs and experimental results — and in determin-
ing how different programs can best be combined for various problems and for dif-
ferent types of ships.

An overall view of the problem of determining wave loads discloses that ex-
perimental and theoretical aspects have been brought to an advanced stage of
development, but that the biggest: gap is in our knowledge of ocean waves. Actual
wave records that can be subjected to spectral analysis are available for only a
limited number of locations on only a few routes, leaving a great uncertainty re-
garding all of the major trade routes of the world. The collection of more wave
records is believed to be of primary importance.

V. DYNAMIC LOADS

INTRODUCTION

Dynamic hull loads may be classed as:

a) steady-state,
b) transient.

Steady-state dynamic loads are the result of such exciting systems as propeller-
hull interaction, machinery-hull transmission and seaway component excitation of
hull girder response, as in "springing." Transient dynamic loads result from for-
ward bottom slamming, bow flare immersion and submergence, carrying green seas
aboard and wave impacts on structural members during rough sea operation,

The transient loadings will be considered first. There are a multitude of

possibilities for such transient loadings, and the generated impacts on the ship's
exposed surfaces often result in structural damage of considerable severity. The
impact loadings are generally most severe in the forward portions of a ship's
structure where relative velocities between ship and sea are highest. These impact
loads arise from the short-time exchange of momentum between ship and sea, and the
damage results when the local elastic energy absorption capability is inadequate.
Energy absorbed locally is distributed through the structure in the form of dynamic
response, "whipping," etc., and dissipated through damping, structurally, by cargo,
etc. The question of concern is what loading is to be used in design. There are
as yet no established criteria for assessing structural loadings arising from im-
pacts of the sea on ship structure. Consequently, a real need exists for the docu-
mentation of resulting damages and the conditions leading thereto, so that future
evaluations of probable loads would be possible.

Unlike quasi-static bending moments, dynamic loads acting on the hull girder
cannot be considered independently of the structural response of the hull. In the
first place, the dynamic magnification factor depends on the natural hull frequency
as well as the nature and duration of the exciting force. Secondly, the response
depends on the damping, which involves structural as well as hydrodynamic terras.
Thirdly, the extreme structural response of the hull which is of particular inter*-
est — plastic buckling or permanent set — requires time and the absorption of
large amounts of energy. As mentioned in Chapter II, the duration of a slam load
may be too short to cause buckling of a ship's deck or bottom amidships. This is
probably not true, however, of the longer-duration flare-immersion type of slamming
which in some ships may contribute significantly to hull girder damage or failure.

mmm i i HI .iiiM-m r»n^r-"

 -33-
Hence, the capability of the structure to wienstand a dynamic load or stress
may be increased by some factor which is dependent on the rate of application and
duration of the load, as well as on the structural properties of the hull. Pend-
ing resolution of this structural design problem, it is important that techniques
for determining and specifying the magnitude, duration and rate of application of
the dynamic loads be dealt with. The problem of phase relationships between dyn-
amic and static loads will be discussed in Chapter VI.

SHIPPED WATER

Consider first the miscellaneous loadings arising from taking aboard green
seas over the forecastle, poop and even at boat deck levels and higher. Such loads
have resulted in the shearing off and washing away of outfit items, deck houses and
bulwarks. More seriously, hatch covers have been stove-in, fo'c'sle decks collapsed
and deck cargoes torn free of their lashings and washed overboard (89) . There are
no known published evaluations of the sea forces associated with these actions,
though classification societies and cargo survey bureaus have revised scantling,
stowage and other requirements on the basis of reported damage. Most loadings
of this type are sufficiently random in occurrence, both in time and space, that
their importance is primarily local, even though the hull girder may on occasion
be excited in a manner which will superimpose significant stress variations upon
those arising from static and quasi-static hull girder flexural, torsional and
shear loads. Better documentation of shipping water damages and evaluation of the
loadings that cause them would allow assessment of such loads for purposes of set-
ting criteria.

Loading of the hull girder as well as local structure resulting from bow sub-
mergence into oncoming waves can cause, in addition to severe local damage, signi-
ficant dynamic loading of the hull girder and generate structural response of such
magnitude and duration that the stress generated can significantly augment the
levels experienced in main structural elements of the ship. The problem of green
seas impacting the ship and being carried over the bow has received attention by in-
vestigators in Europe, the U.S.A. and Japan (90). Hoffman and Maclean (91) have
shown that in order to predict the event, knowing the ship's relative motion opera-
tors and the sea condition is sufficient. Kawakami (92) has shown that carrying
green seas over the bow can generate whipping stresses equal in magnitude to as much
as 40%-70% of the sagging bending moment in regular waves. Thus, the prediction of
the event and the magnitude of the loading response of the hull girder appear amen-
able to treatment by combined experimental and theoretical means. There has been,
however, no full-scale confirmation of experimental findings, or vice versa. Conse-
quently, there is not at present an established procedure for evaluating the,magni-
tude and frequency of hull loading resulting from carrying green seas aboard at the
ship's bow. Though experimental work by Kawakami (92) showed that full-scale fore
deck loadings of 40-70 lb/in2 could be experienced, generalization of these findings
has not yet been attempted.

In view of the fact that these loadings can result in the generation of signifi-
cant hull girder responses which are of sufficient duration to cause superposition
on the quasi-static loading of the structure, it is recommended that further work be
carried out in this area with a view to establishing full-scale experience and com-
bined experimental-theoretical clarification for design.

FLARE IMMERSION

Slamming of the forward bottom and bow flare surfaces give rise to large loads
on the hull girder of sufficiently short duration that high-frequency hull girder

irr l -iiMiiMn Mini irifcn fciftt ■iimniiiiTiiiiiiMr «*—*>«^.

 -34-
dynamic stress variations are generated of such magnitude that they may signifi-
cantly augment the maximum stress levels experienced by main structural elements.
The slam-generated forces developed on the ship forward bottom surfaces are gener-
ally the result of high pressures acting over relatively small areas of the flatter
bottom plating; significant force duration is generally less than 100 milli-seconds.
The forces developed by the immersion of the bow flare are generally the resultant
of lower pressures acting on relatively larger surface areas of the hull. The dura-
tion of these forces is generally greater, but usually of less than a second with
significant magnitude. The^e latter forces are of importance primarily to vessels
with large bow flare, su^h as some aircraft carriers anJ some recently designed com-
mercial vessels with large bow flare and overhang.

The generated pressures and resulting forces developed on the bow f]are are
susceptible to computation on the basis of available theory. The initial effect
as the bow flare enters the water is a non-linear addition to the quasi-static wave-
induced bending moment. A step-wise evaluation of the exciting force for the cal-
culation of vibratory hull girder response, as discussed below, can be used to com-
pute the initial bow flare immersion effect. The duration of this bending moment is
sufficiently long that it should probably be added to the other moments involved in
ultimate strength.

The vibratory response or whipping following flare entry can be determined by

analog simulation (93)(94) or by deterministic calculation (95)(96). This short-
duration response, generally in the fundamental mode, should be considered in rela-
tion to fatigue and brittle fracture. However, there are no known published recom-
mendations concerning the magnitude of bow flare-generated forces that should be
considered as design loads either for the hull girder or the local structure for-
ward.

The basic theory of impact for a wedge entering the free surface of a fluid is
applicable to bow flare load generation. Such theory, developed by Wagner, von
Karman, Pabst and others has been modified by Szebehely, Ochi, and others for appli-
cation to ship forms in which the body is not stopped by the impact. This theory
has been the basic ingredient for all the work mentioned above. To use it for the
generation of load information, as per Kaplan's program (95) (96), requires a de-
terministic approach in which a mathematically defined hull is subjected to an ir-
regular wave pattern derived from a sea spectrum of interest. Kaplan's procedure
allows the prediction of ship loads on such a basis and, upon repeating the process
for a suitable number of sea conditions, a body of load data could be developed for
use in a suitable probability model.

The chief drawback seen in the above approach appears to be the fact that sink-
age and trim, as well as dynamic bow wave build-up, should be included if reliable
bow immersion results are to be expected. These non-linear effects have been identi-
fied in experimentil work (91) as vital to reliable prediction of flare submergence.
The works of Tasaki (97) and Tasai (98) are also useful here. Further development
of the theory should be carried out.

It is believed that a part of a rational'design criterion would be first the

calculation of initial bow-flare immersion bending moment for one or two representa-
tive severe head sea conditions, especially if the ship has considerable flare and
operates in a deeply loaded condition. An exhaustive statistical study should not
be necessary but simply a long enough irregular sea run to determine the highest
value in, say, three to four hoursf

The problem of phasing of flare-immersion bending moment appears to be much

simpler than that of slamming (See Chapter VI), The bending moment (sagging)
gradually increases as the flare enters the water and reaches a maximum at or near

. ^.^vtmi+Am MttttiAl-nii,'irilMlWiM
 -35-
the time when the wave bending moment is also a maximum in sagging Hence, the
combined effect of wave bending moment and flare immersion can be approximated by
adding the maximum values directly.

Secondly, the vibratory response (whipping) that follows flare immersion should
be calculated for ships with large flare by the method developed by Kaplan (96). It
is recommended that this approach be developed into a standard procedure for use in
relation to design for ultimate strength, fatigue and brittle fracture.

BOTTOM IMPACT

The initial effect of a bottom impact is the generation of large local pressures
which may in some cases cause local structural damage. The resulting hull girder re-
sponse is in the higher modes, as well as the fundamental, and therefore may produce
a high initial stress. The whipping that follows will be mainly in the fundamental,
since the higher modes are quickly damped.

In respect to loads generated by forward bottom slamming, present capability

is restricted to the prediction, of peak pressures at a point en the forward bottom
centerline of typical hull forms. Based on the work of Ochi (99), the hull form of
a ship's forward bottom sections can be assigned coefficients, k, such that for a
known impact velocity, V, the peak pressure generated at a section would have a
most probable value p = kV . The coefficients, k, for some typical section shapes
have been determined on the basis of regression analysis of 3-dimensional model
experiments in waves (100). Using these values, a comparison can be made respecting
the slamming propensities of competing hull forms. To do so, the ship response
properties must be determined, experimentally or analytically, and the relative ve-
locity spectrum for the critical ship section established. Using this approach,
maximum slam-generated pressures in a given sea can be predicted on the basis of
probability. Such data may be used for design of local structure if the probability
can be prescribed with a satisfactory degree of confidence. Unfortunately, no full-
scale assessment of local loading has yet resulted in the establishment of a proba-
bility level that could be used as above; a recommendation is made that such be pur-
sued.

It is understood that Dr. Ochi is working on a mathematical description of slam

loading which attempts to describe the distribution of pressures in both time and
space. With such a load definition various procedures are available for calculating
the hall girder response and hence the midship slamming stress (101)(102). However,
the requisite load definition is not yet available.

Because of the above situation, it is not possible to prescribe at this time a

slam loading of the hull girder, due to forward bottom slamming, which should be
used in setting design criteria. On the other hand, the problem can be viewed in
another light somewhat more positively. Full-scale strain recording on board the
S.S. Wolverine State has captured hull girder slam and whipping stress variations
experienced during rough sea operations of the vessel. Evaluation of data (103) ob-
tained in various high sea states has shown that the severity of slam response —
when it occurs — tends to be somewhat insensitive to sea condition. This can be ex-
plained by noting that though slamming frequency and severity are increasing functions
of sea state, sea speed and vessel heading, the latter two are under the control of
the ship master who will tend to alter course and speed so as to keep the frequency
and severity of such sea loadings within reasonable bounds, according to bis expe-
rience. Fig. 9 shows histograms of midship slam stress variation for the Wolverine
State (103), which can be interpreted in terms of equivalent hull girder bending
moment.

r.,....,,..:, ggaaafl -TM^ -,- MM B— mam - üafll 1 i I "f- MMMfc mm* ^^fcHM > tamtam* ■«.«■«ai.iwuin.,
 -36-

STRESS RANGE IKSI»

STRESS RANGE iKSII

FIGURE 9 - Peak-to-Peak Slam Stress Distributions in Different Weather

Conditions, S.S. WOLVERINE STATE

Whereas the findings as to the relative level of loading experienced by the

S.S. Wolverine State would apply only to that vessel and in its particular trade,
it seems apparent that the collection of similar data on a sampling of vessels of
other types loaded in different ways should provide a suitable basis for the devel-
opment of allowances for slam loading of the hull girder. Further, in view of the
complexity of the slamming phenomenon, it does not appear that an analytic descrip-
tion oc slam loads will be forthcoming in the foreseeable future, and the collection
of full-scale slam response data appears to offer the'most promise for the establish-
ment of expected hull girder slam loads.

It is recommended, therefore, that stress (strain) recordings presently in

hand for other vessels in other trades be reviewed and analyzed as to the frequency
of slamming and the magnitude of slam stresses, and that consideration be given to
the gathering of similar data on vessels of more recent design to be reviewed on the
same basis. Such data can be used for the time being as a guide in estimating dy-
namic loads and stresses for new ship designs.

Another aspect of the slam load problem to be considered, if one is to de-

termine the combined effect of slamming and quasi-static wave bending on the hull
girder, is the phasing of the slamming loads. The particular instant in the wave
bending cycle at which a slam occurs determines the extent to which the slam stress—
and the subsequent whipping stress — add to the total stress magnitude. A study ot
this subject has been made, on the basis of Wolverine State stress records, and is
reported in Chapter VI. This investigation indicates that the phasing — as well
as the magnitude — of slam loading can be established empirically for different ship
types as a basis for design. However, it also shows that the stresses at impact
seldom if ever add significantly to the total bending moment. Of much greater
importance is the whipping that follows a slam.

jtftf
. .,,itr,,. i -•■"■
 -37-

For design purposes, the practical problem of determining the transient dynamic
loads reduces *-o two parts:

(a) The prediction of the probability that slamming will occur (over a
period such as a 20-minute record).

(b) The distribution of added stress (or equivalent bending moment) to be

expected when slamming and whipping occur.

OBSERVED LOADS AND STRESSES

Available published data on actual slamming pressures recorded on ships at sea

and on measured midship slamming stresses are scarce. However, a summary of avail-
able information on maximum recorded values is given in the accompanying Table I,
compiled from the indicated references.

Reference should also be made to a general survey of slamming by Henry and

Bailey C104) .

SPRINGING

As mentioned in the Introduction, one type of steady-state dynamic effect is

known as "springing." This phenomenon has been noticed particularly in Great Lakes
bulk carriers (114), but it has also been reported on ocean-going ships (115)(113) .
A clue to the origin is given by the fact that the Great Lakes bulk carriers are
quite shallow in depth and consequently have unusually long natural periods of
vertical hull vibration (2-noded periods of 2 sees, or longer). The apparent ex-
planation is that when the ship is running into comparatively short waves that give
resonance with the natural period of vibration, significant vibration is produced.
This vibratory response may continue over some period of time, gradually fluctuating
in magnitude. A corresponding fluctuation in stress amidships is therefore

Table I

MEASURED FULL-SCALE

SLAM LOADS AND MIDSHIP STREf ;SES

Peak-to-trougl
Midship
Pressure SI an
stress, kpsi
U5CG Cutter 295 local
L'nimak (105) 86 ave. p inel -
Dutch Destroyer (10>6)(L073 100+ 11.20

Cargo Ship
S.S, Westboro (108) - 1 .90

Cargo Ship
S.S. Wolverine State Voy. 288 49 2.60
(109) Vov. 277 69
(103) 10.00

Bulk carrier
Fotinl L (110)
- 15.50

Container ship
Manchester Citv (111) - 5.00

Cargo ship 4.50

M .V. Jorclaens (112) 1
(19. 2.04

Container ship 15.60

Flinders Bay (113)

-...,. ..^(^j^ 1,.--^.-.^.ia>-. njnnjujgj.... m—^,^^ui^i^«aii*imHmtitüm*äimttmum ÜMM t i ü MUM

mi -ii-uttim^"-' "f-"aMUÜI ■''■'•■ ■■■•-..■■■w^. nKjgfli — ^„...^^.^.
 -38-
superimposed on the quasi-static wave bending moment. The springing stress appears
to have the characteristics of a stochastic process, but one that is essentially
independent of the low-frequency wave bending, which — as previously noted — is
also treated as a stochastic process.

Figure 10 shows a record of midship stress in which both low and high-frequency
stresses are present. Also shown are two records in which first the high-frequency
stresses have been filtered out and second in which the low-frequency stresses have
been filtered out.

The well-developed strip theory of ship motions has been applied to the spring-
ing problem (115). Although motions of a springing ship may be very small, the
theory provides information on the exciting forces acting on the ship in the short
waves that produce springing. Hence, when these forces are applied to the ship
as a simple beam the vibratory response can be predicted. Despite the fact that
strip theory is not rigorously applicable to such short waves, results for one ocean-
going ship were found to agree quite well with full-scale records, (115). Further
coordination between theory and experiment has been attempted for Grsat Lakes ships,
including model tests where idealized wave conditions can be provided (116). Subse-
quent tests with a jointed model of the Great Lakes carrier Stewart J. Cort using
connecting beams of varying stiffness, show clearly that the springing bending mom-
ent increases with hull flexibility.

When the above calculation procedures have been tested and revised as neces-
sary, a tool will be available for predicting springing stresses in a new ship de-
sign. The remaining problem for incorporating springing into the comprehensive
ship hull design criteria is to determine the manner in which springing stresses
superimpose on the low-frequency bending. This problem has also been solved in
principle (117) on the basis of the assumption that the two stochastic processes
are independent. This assumption is illustrated by the records in Fig. 10 which
show that high amplitudes of the two types of stress do not occur at the same time.

Further work needs to be done as new information becomes available. Meanwhile,

the general cargo ship — which is the principal sub-ject of the present study —
does not: seem to be subject to significant springing effects, and therefore they
need not be included in the present criteria.

VI. PHASING OF SLAM AND WAVE LOADS

INTRODUCTION

It has been the goal of this project to obtain complete information on all as-
pects of ship hull loading, particularly loads generated by waves at sea. The first
type of load, quasi-static wave-bending moments, has, of course, been focused upon
first, and the current state of knowledge is summarized in Chapter IV. Next must
be considered the influence of slamming, and whipping after the slam, which are dis-
cussed in Chapter V. Finally, in order to determine the combined effect of wave
bending and slamming, particularly in reference to brittle fracture, it is neces-
sary to understand the way in which the latter high-frequency load is superimposed
on the former. If the slam should occur, for example, when the wave bending moment
is near zero, the immediate effect of the slam will not be an increase in the total
amplitude of bending moment amidships unless it exceeds the wave bending moment
amplitude. On the other hand, if the slam should occur exactly at a peak bending
moment it would increase the total by whatever its amplitude might be. Similarly
the amount by which the vibratory whipping stress (or equivalent bending moment)
following the slam would add onto the subsequent peak of the wave bending moment
curve would depend on how rapidly the vibration is damped out.

,.JJJ...^.^.A-*.-~-^.^..i*f» .*;-*;«
 -39-

17,900 psi

a) Total Stress Variations Including Both Wave-Induced and First-Mode Stress Components

b) Wave-Induced Stress Variations (Frequency Approx. 0.1 Hz)

'5,400 psi

c) First-Mode (Springing) Stress Variations (Frequency Approx.. 0.70 Hz)

FIGURE 10 - Typical Record of Midship Stress Variation, M.V. FOTINI LJ

Showing Filtered Wave-Induced and Dynamic Stresses (3).

,u-]tgm- i -MIMIM'T- -■-' ••■''* ' "irT""'irwi—""»—■—

- miM ■-■- ■ iMjMüMttaüH—MMiMi l imwi^iM
 -40-

Nothing could be found on this subject in the literature. Hence, it was de-
cided to make a study of the phasing of slam and wave lo^ds for the Wolverine State,
since it was possible to obtain stress records of slamming from Teledyne Materials
Research, and a reasonably clear picture emerged for this particular ship.

It is realized that is impossible to generalize this result for one ship to ap-
ply to all ships, and hence it is believed that similar studies should be made on
ships of other types and sizes.

DEFINITIONS

A slam is defined as a transient impact force acting ^"n -a ship-. We can -there-
fore call the hull girder stress generated by this transient force the slam stress,
o , by which we mean the first peak (sag) or the second peak (hog) in the slam
so
stress time record. (See Fig. 11). The two-noded hull girder vibration generated
by the slam is called whipping. We call the stress generated by this vibration the
whipping stress, awo . It should be noted that bending moments equivalent to these
stresses can be determined for combining with other loads. Wave bending moment is
defined as the low-frequency hull girder moment which is experienced at frequencies
corresponding to the ship's encounter with the waves in the seaway and produces the
wave stress, o, .
' ho
It should be pointed out here that in this study we are particularly interested

-th

0| = WAVE STRESS
a,
SLAM STRESS
°wo « WHIPPING STRESS
S - ADDITIVE PORTION OF <7J(
* - PHASE ANGLE

FIGURE 11 - Definitions of Stresses and Phase Angles

Involved in Slamming

in: (a) the second peak of the slam stress time trace — generally occurring dur-
ing a hogging wave stress condition — and (b) the whipping stresses in the sub-
sequent sagging and hogging wave stress conditions. See Fig. 11.

— - rl iwmmi in i jftfaaftMtHj; „, „ „„ iM.nii i. -viiTNin ..!..- H ■IIIMW'-—"-miTi-.il rir mtM

 -41-
It is our objective here to obtain some idea as to the probability that, if
a slam occurs, the amplitude of total stress (or equivalent bending moment) will,
be increased over the wave bending stress by either slamming or whipping. The above
probability is to be evaluated at the midship section, because the problem of local
stress is not included in our immediate objective. We. note that it is not our in-
tention to give a probability analysis of the occurrence of slamming, this having
been treated elsewhere, as discussed in Chapter V.

PRELIMINARY ANALYSIS

It was decided to make use of available stress records of the S.S. Wolverine
State, The choice of records to be used was made on the basis of the following
criteria:

a. approximately the same environmental (sea) conditions,

b. clear stress traces in which the slam was well defined,
c. approximately the same ship speed.

This selection was of critical importance to the analysis, and the consistent re-
sults indicate that a good choice of records was made (103)'.

This analysis was then combined with another similar recent analysis (2) to
obtain as complete a statistical sample as possible, comprising approximately 65
data points.

From the selected records the following were obtained:

(a) histograms of phase of slam inception relative to peak wave bending

stress (tensile in deck, i.e., hogging condition). (See Fig. 12),
(b) histograms of slam stress amplitudes (See Fig. 13),
(c) histograms of whipping stress amplitudes (See Fig. 14),
(d) empirical solution of the damping of the high frequency (two-
noded) stresses. (A representative time trace is given in Fig. 15).

In item (d) it was postulated that the damping is dependent on two frequencies,
but only the 2-noded mode was considered; all other modes were lumped.

The empirical solution of the damping function g(t) was:

-C] u>
Ws t -C2 Ws t
s
g(t) = aj e + a2 e
„1
where co ■ frequency of two-noded hull girder vibration = 10.389 sec.
s
cj = damping constant for two-noded mode ■ 0.00513
C2 " damping constant for all other modes - 0.1540
a2 = 0.8

a2 0.2

——■»■ - *™**~~~—..—-—
 -42-

TELEDYKfc DATA
D (12 FEB. 72 REPORT)
LINE 14

WE B8 DATA

NO.
OF
OCCURRENCES

XI
0 20'

* PHASE ANGLE

FIGURE 12 - Distribution of Slam Phass Angles,

S. S. WOLVERINE STATE

TElEOYNEDATA
E3 112 FEB. 72 REPORT)
LINE 12

WEBB DATA

I 1 I ES
'* 2.0
^___
SLAM STRESS. <>„, KPS,

FIGURE 13 - Distribution of Slam Stress,

S. 5. WOLVERINE STATE

■■*■■■- -—,-■-.; - — ttttam II-»I mirum -ii- I'""-*■-*■■■---•■ „ ,. ■ ._.■...-»., ^-....^. _„,._-.,, ...^ -^^ ^.^.„.i .., .- .-.. ■ . -^aJ
 -43-

——, TELEDYNE
itieutivtuAi«
DATA
(12 FFB. 72 REPORT»
"" LINF r

1.0 1.5 20 2.5 30

WHIPPING STRESS, °wo KPSI

iFfGURE 14 - Distribution of Whipping Stress,

S. S. WOLVERIim STATE

*
* X

SLAM STRESS
(-1. ATt-0!

NUMBER OF PEAKS

10 20 30 .0 50 60 70
. _ 1 1 . .1 L . i 1 1 . .... I . J
10 15

TIME t (SEC)

FIGURE 15 - A Typical Decay Curve of Whipping Stress

■! — — ~ --™. -..-■■ rrivr-..,.»

in nilfiWlir-- unai ■ . . "■■"'^-"*"^"*rt^rttMl^»
 -44-
DISCUSSION

From the histograms and figures the following can be deduced:

1. The phase angle seems to be normally distributed with a range of

-56° < $ < + 72°, relative to t - 0 (peak hogging wave stress). This
implies that the phasing is a random Gaussian process, which is a reason-
able assumption since slamming occurs only if certain conditions are
satisfied, namely bow out of water, sufficiently high impact velocity,
etc.

2. The slam stress seems to follow a distribution of a type with lower

limit + 0. It has been stated (103) that the type of distribution is
exponential. Although this is not entirely confirmed by our data
(Fig. 13), we followed the findings in the above-mentioned report because
the sample reported there was much larger than the sample of 65 we were
able to obtain for this particular study.

3. The histogram for whipping stress is more regular in shape, and we

notice that the dispersion is far less than in the case of slamming.
Since the frequency of the vibrations associated with whipping is rela-
tively high, we may directly superimpose the whipping stress as presented
in the histogram on tne compressive wave stress. Therefore, the mean in-
crease over the wave stress is also the mean of the whipping stress,
namely a = 0.876 KPSI, in this case. See Fig. 14. This is not the case
with the slamming stress which we will consider next.

4. The histogram of that part of the slam stress which is a ditive to the
maximum tensile stress shows two things (Fig. 16):

(a) The distribution of addition and non-addition is approxi-

mately 50 - 50%. Therefore, of all the slams the mean stress
increase over the wave, bending stress is approximately Ü.
(b) If we take that half of the number of cases in which the
slam stress does increase the total stress at midship, then the
mean increase is, a = 0.13 KPSI, which is indeed very small
compared to the mean increase due to whipping.
We may therefore conclude that whipping is of relatively greater im-
portance than*slamming for this ship.

5. The evaluation of this sample would not be complete without con-

sideration being given to the separate relationships of dynamic stresses
to the tensile anu compressive wave-induced deck stresses. We observed
that the slam stresses are associated with the tensile wave-induced deck
stresses, and high whipping stresses are associated with the compressive
stresses. Therefore it is useful to define the following ratios (see
Fig. 11):

a
so
ho

a
wo
w a
ho

------ i inn.Hi» mi.» in iiitiiiiliiaaWriTaatirii-irtiiiiMrtfcWfii'^--"--"■---■■-■■•--■-■-■■ üutmmuuBm ---.-^-r-^- -■■-■■ --■■■■ lälima*a&m&M*Slmm ■■.....,». ^^^^ffl
 -45-

«jfT
-
NON POSITIVE
-STrtESS VALUES
AT SLAM

□ TELEDYNEDATA
(12 FEB 12 REPORT)
LINE IS

i
i
NO
OF
OCCURRENCES

0 STRESS VALUES
" AT SLAM

5, OR DEL ST

FIGURE 16 - Histogram of Slam Stress Additive

to Wave Stress

and consider the histograms of these two ratios (Fig*,...17.and 18).

Both histograms show reasonably regular distribution of the ratio,
considering the limited size of the sample. Without going into an
analysis of what specific distribution they fit, it is gratifying
to see that the shapes are similar to the shapes of the individual
slam stress and whipping stress distributions, as shown in Figs. 13
and 14. .

6. Our last consideration is a study of the correlation between

slam stress and tensile stress. In Fig. 19 we have plotted the
slam stress a against the tensile wave-induced stress o, . We
conclude that there is no correlation, and therefore we may consider
these stresse? to be independent.
 -46-

CO
CO
(ü
i-
"M
00
E
fO
r—•
oo
•+-
o
o
o LO a»
•r- 00
+-> O)
o >
•r- (13
13 i. +-> 3:
CC 4->
OO Cd o
at
-C cn
4-> c to
•r- -t-> CO 00
M- ~o 0; co
O c M- i- ai
a> o 4-> i-
E CO OO 4->
fO E oo
i- a» ra CD
a> > i- C CD
O (tj CD-i- C
■M 2 O Q-T~
CO +■> Q.-0
•r—o 00 -r- C
X •*-> •r- _e a>
in 2 co
i
00

ej

üMHi
 -47-
FÜRTHER WORK

A theoretical development has been made on the probability model from which
to determine:

a) combined slam stress and wave bending stress, and

b) combined whipping stress (following the slam) and wave bending stress.

Part of this work is incorporated in Chapter IX, but a continuation of the develop-
ment is recommended.

VII. THERMAL EFFECTS

INTRODUCTION

Records of midship stress obtained on five bulk carriers (3) indicated sur-
prisingly high thermal effects. These showed a consistent diurnal variation,
with magnitudes of 3-5 kpsi in some cases. The temperature gradients that pro-
duce such thermal stresses may not be, strictly speaking, loads but they are con-
sidered to be loads here nevertheless.

Although it often happened that high thermal stresses occurred at times of low
wave bending stresses (sunny weather), and vice versa (stormy and cloudy weather),
this was not always the case (3). The exceptions are presumably times when a heavy
swell was running while the weather was clear.

It should be noted that the thermal stress changes recorded here were overall
averages, since they were based on combined port and starboard readings. Because
of the effect of local shading it can be expected that even larger thermal stresses
would be experienced. However, it can be assumed that such local high thermal
stresses can be ignored for the present purpose.

In order to include thermal effects in design calculations two distinct steps

are required: estimating the magnitude of the effect under different conditions
of sun exposure and estimating the frequency of occurrence of these different con-
ditions in service.

In a discussion of (3) tanker service data were presented which showed a

strong correlation between change in sea-air temperature differential and change
in stress level. Theoretically there should be no difficulty in calculating one
from the other. This chapter describes the application of available theory (119),
assuming simplified structure and using estimated temperature changes.

The simplified procedure is applied to the tanker Esso Malaysia first, be-
cause records of the measured diurnal stress changes and some temperature data
are available. It was assumed that if good agreement between prediction and measuie-
ment was found the technique could be used for numerical calculations on Wolverine
State in a later chapter.

TEMPERATUFF CALCULATIONS

Essc Malaysia logbook data plotted by Breves in his discussion of the Little-
Lewis papci. (3) indicated air-water temperature differences at noon (when they are
maximum) of 10° to 13° F., averaging 11.5° F. over eight days (maximum was 16°).
If constant sea water temperature is assumed over any 12 hour period (noon to mid-
night or midnight to noon) the average diurnal change in air temperature is also

, „ _ . ..., ---.^ --■ - ■-■ •■-'-"■■-> -"i.iiintilMWllHiriafMr-ll.li-M'JliVl- -tntl '■ L^...I.-I-. ■«■-II ..ir.i-i..^ Jfl^^^^J^^^g^^m
 -48-

11.5° F. Deck plating would be subjected to this change plus the change due to
insolation (i.e., the absorption of radiant heat). The temperature change due to
insolation depends on cloud cover and color of deck. As an approximation, Fig. 82
°^ Principles of Naval Architecture (120)(p. 247) shows maximum differences be-
tween air and deck (sun overhead, unshaded) for different colors of deck as fol-
lows:

black 50° F.
red 40° F.
aluminum 10° F.
white 10° F.

It was established that the color of the Esso Malaysia deck was a very dark gray,
almost black; hence a figure of 40° was used plus the measured average air tempera-
ture change of 11° — giving 51° (say 50°) for AT.

A simplified distribution of AT is assumed, namely constant over the deck and

the sides down to the water line and AT * 0 elsewhere (below WL). Longitudinals
will have the same'AT as the plating which they support. Longitudinal bulkheads and
associated longitudinals have AT decreasing linearly from the deck AT at top to AT -
0 at the assumed level of oil inside the tanks.

RESULTS OF STRESS CALCULATIONS

Under the assumed conditions the calculated thermal stress at deck edge due to
temperature change is about 2000 psi. [1600 at center of deck stringer strake and
2300 at shear strake]. From the measured stresses during the same period of time
(Fig. 38 of (3)) the 11 day-night or night-day stress variations in KPSI are as
follows (9/18/68 to 9/26/68):

2.3, 2.3, 1.7, 1.7, 1.6, 1.5, 1.7, 1.8, 1 9. 1.7.

The average value is 1.8 or 1800 psi.

It is concluded that the approximate calculation is satisfactory. Stresses

given elsewhere (119A)(119B) are higher because they include unsymmetrical
temperature gradients.

PREDICTING SUN EXPOSURE IN SERVICE

The prediction of voyage average thermal stresses and expected maxima requires
also that the frequency of occurrence of different conditions of sun exposure be
determined. Source data for such predictions are given in the U.S. Navy Marine
Climatic Atlas of the World, Volume VIII (82). Cloudiness is represented by charts
of the world's oceans showing for each month of the year:

1. Total cloudiness, with isooleths indicating,

(a) % frequency of total cloud cover less than or equal
to two-eighths,
(b) % frequency of total cloud cover greater than or
equal to five-eighths.
2, Median cloudiness, with the midpoint (50% of observations)
of total cloud cover reported in eighths.

a iiii]itii----j-iiir jgBMKaimn
 -49-

In addition, special low cloud data are given, which are not necessary for these
calculations.

From the plotted data, it is possible to estimate average cloud cover for
any given trade route on a monthly, seasonal or yearly basis. Cloud cover is
then related to air-deck temperature difference due to insolation (the 40° F.
value stated above, for example) by assuming that the insolation AT is directly
proportional to the extent of cloud cover. Thus the 40° temperature difference
would apply to full sun (cloud cover * 0/8), while total cloud cover (8/8) would
indicate AT = 0. Intermediate values are assumed to vary linearly. The resulting
insolation AT's are added to the sea-air AT's to determine total AT for each cloud
cover condition. A weighted average of total AT can then be calculated by com-
bining the total AT's with their frequencies of occurrence as determined from the
Atlas (82).

A sample calculation is shown in Chapter IX, where the method is applied to

the Wolverine State.

VIII. COMBINING LOADS FOR DESIGN

GENERAL

As suggested in Chapter II, the primary load criterion is assumed here to be

the maximum combined bending moment resulting from the various ioads that can
cause excessive deflection or failure by buckling or plastic deformation. The dis-
cussion of loads in Chapters III - VII indicates that this combined load can
best be stated in probability terms, i.e., the overall combined bending moment to
be exceeded once in the lifetime of a ship or of a fleet of ships. We are also
interested in lesser combined values that may occur more frequently and may cause
structural damage without complete failure. Hence, the next section will deal with
the combining of static and dynamic loads, referred to as Ultimate Bending Loads.
Load criteria for fatigue and brittle fracture will also be discussed in following
sections.

ULTIMATE BENDING LOADS

Still Water Loads

The first step in calculating and combining static and quasi-static loads —
and hence determining the primary load criterion — is to consider the still water
bending moments expected in the life of the ship. Typically, as indicated in
Chapter III, there will be a distinct difference between outbound and inbound load-
ings (in some cases there may be three typical conditions — as for a ship engaged in
a triangular service — or even more). One case may be full load and the other bal-
last; or there may be different loaded conditions outbound and inbound. However,
for each case there will be a variation of bending moment from voyage to voyage, de-
pending on density and distribution of cargo and/or ballast. The object is to esti-
mate the mean and standard deviation of outbound and inbound bending moments over
many voyages. On long voyages there may be significant variations in bendiug moment
as fuel is consumed and salt water ballast is added, and therefore a distinction may
be necessary between departure and arrival conditions. Hence, probability density
curves, showing the probability of different levels of still water bending moment,
can be estimated for both outbound and inbound loadings.

iniMMiinmr
 -50-
Having made the above estimate of separate autbound and inbound still water
loadings, it is essential to carry through the remainder of the load calculations
separately for each aase, since the basic differences in drafts and load distribution
will affect wave bending moment, probability of flare immersion, etc.

An estimate should also be made of the bending moment caused by the ship's own
wave system at forward speed in calm water, as discussed in Chapter III.

Wave Loads

Coming to wave loads (Chapter IV) response amplitude operators for vertical
bending moment, horizontal bending moment (at midship) and vertical shear (at
quarter length) must first be calculated for both outbound and inbound conditions
(average). These operators should be calculated for a range of wave lengths and
headings. However, since for most displacement ships speed is not an important
factor for wave bending moment, calculations may be made for one representative
rough weather speed. Torsional moments should also be calculated at the midship
section.

Next, the family of sea spectra which are to be used as a basis for design
must be selected. If a particular ocean route or routes are tc be ust 1, then in
general the family of spectra will be based on tabulated data on frequency of occur-
rence of different combinations of observed wave height (significant) and period
(average). Or if the design is to be based on North Atlantic service, families of
spectra for each of five different average wave heights (66) can be used.

Before proceeding to the bending moments in irregular waves, the problem of

combining vertical and lateral (or horizontal) bending moments must be considered.
It has been shown that a ship operating in oblique seas is subjected to unsymmetri-
cal bending, so that the stresses measured at one deck edge will usually exceed the
mean value. This diagonal bending in an oblique sea can be dealt with as the combi-
nation of vertical and lateral bending components (121) . Since we are interested
first in bending moments rather than stresses, we can calculate an effective verti-
cal bending moment, M , that produces the same average stress as the maximum deck
edge stress, a, resulting from combined vertical and lateral bending. If Z is
the section modulus for vertical bending and Z for lateral
i-j

M = c Z_.
e V

V
2
M 4 n [^ I + 2M M T ~- cos 6
wv wL I ZT / wv wL Z

where M and M . are the vertical and lateral wave bending moments, respectively,

and <5 is the phase angle between vertical and lateral bending (121) .

The accurate way to proceed is to obtain response amplitude operators on the

basis of the above for combined vertical and lateral bending and to use these for
calculating effective vertical bending moments in irregular waves. This requires
that the ratio Z /Z be known or assumed in the design stage. However, this pro-
V Li
cedure would not provide an estimate of trends of vertical bending moment alone, for
comparison with other ships. Hence, for most cases, it is recommended to:

^^M,,**,,,»**^***,^*^^ , ^..w,..,.^^^^^,,^,^,,^^. - ^.MMimttMi

 -51-

1. Calculate vertical bending moments M in irregular seas, for comparison

wv
purposes.

2. Calculate from response amplitude operators for both vertical and

lateral bending ehe values of effective vertical moment, M , using
a tentative value of Z /Z..
v L

If Z !Z should change significantly during the design, calculations for M must

be repeated.

If there is evidence that maximum bending moments occur at a section signi-

ficantly different from amidships, a correction factor can be applied to the midship
results.

The next step, as discussed in detail in Chapter IV, is to calculate the bend-
ing moment response to different sea conditions and hence derive both a probability
density function and a long-term cumulative distribution of bending moment coeffi-
cient for both vertical bending alone and combined vertical and lateral bending.
In a similar fashion the long-term distribution of vertical shear at quarter points
and of torsional moment would be calculated.

Thermal Effects

On the basis of reasonable assumptions regarding air and water temperatures,

and their diurnal variations for each of four seasons, Jasper?s method (119) can
be used to calculate thermal stresses in the weather deck. The Interest here is in
the overall average change in stress across ehe deck from day to night, rather than
local high stresses. Such calculations should be repeated under the assumption that
there either is or is not full sunshine. Hence, considering data on the percentage
of time that the sun shines for each season on the route in question, a reasonable
probability function of thermal stress variation can be constructed. For the pres-
ent purpose, however, it is felt to be adequate to use a weighted average thermal
stress converted to a corresponding effective vertical bending moment.

Since in general there is a tendency for high thermal stresses to occur in

good weather (sunshine) when wave bending moments are comparatively small, and vice
versa, an attempt might be made to estimate a suitable correlation factor. However,
since a heavy swell and bright sun may appear together, it is perhaps best to as-
sume that the effects are Independent and additive.
Dynamic Bending Moments

The evidence indicates that springing — the more or less continuous vibration
excited by waves — is not found in ordinary dry cargo ships or container ships,
presumably because of their low length/depth ratios, hence relative stiffness. For
example, in the case of the Wolverine State the only vibratory response measured
was that associated with slamming. Hence, it is assumed that the only dynamic load-
ing to be considered here is slamming, which causes a damped high-frequency stress
variation (whipping) that can be interpreted as an effective superimposed bending
moment.

The first step, for both outbound and inbound conditions taken separately, is
to estimate the probability of slamming. This can be done on the basis of data on
actual voyages of similar ships or calculated by Ochi's method (100).

IHM« ah!*« hllMJ — .ftiv».!.«.»

 -52-
Next we must estimate the probability function for effective bending moment due
to slamming when it occurs. This involves a number of steps, all of which are not
at present in satisfactory state of development for routine design use:

1. Estimate local slam pressure and its variation in time and space in a
representative severe sea condition — hence its distribution.

2. Estimate resulting midship stress immediately following the slam and

the subsequent whipping stress — hence their distributions,

3. Translate these stress distributions into effective bending moment dis-

tributions .

Only one representative severe sea condition is assumed to be required, because

it has been shown that the statistics L slam stress when slamming occurs is rela-
tively insensitive to sea severity (Chapter V) — in contrast to the frequency of
occurrence of slamming. This is because speed is voluntarily reduced as necessary
to avoid severe slams. However, one must make a realistic assumption regarding the
ship speed in this representative sea condition. The speed is probably determined
by the master on the basis of the ship's slamming behavior (since if the ship is
slamming it probably is not deeply loaded, and therefore shipping water or propeller
emergence are probably not governing factors). The exact criterion for speed might
be any one of the following:

- Frequency of slam occurrence.

- Severity of impact (as judged by the sound).
- Amplitude of hull vibration at the bridge.
- A subjective correlation of any or all of the above with cases
of past bottom damage on this ship.

Assuming that speed can be estimated for a new design in the sea condition
selected, the procedure might be as follows:

1. The local keel pressure can be estimated from calculated motions

(relative velocity) and Cchi's data (100). However, the extent of the
impact in space and time can at best be only a rough estimate.

2. Kline's method can be used to calculate the resulting midship stresses.

3. The stress distribution can be translated into bending moment distribu-

tion by multiplying bv the section modulus.

Pending availability of techniques to carry through the above procedures,

another approach would be to assume that after extensive sea experience on a class
of ships the ultimate criterion of slamming is simply the avoidance of bottom damage.
Hence, one could calculate the pressure over a reasonable bottom area that the local
structure could sustain without permanent set. Then this load would be used in the
calculation of midship slamming stress — and hence of effecive bending moment
(Items 2 and 3 above).

A third approach would be simply to obtain statistics on the midship slamming

stresses that are allowed to occur on ships of different types and to estimate such
a distribution for any new design (as given for Wolverine State in Chapter V).

The final step is to determine the probability distribution of the amount by

which the slam bending moment adds onto the wave bending moment. This involves the

MBÜMi ■-- aemmai

 -53-
phase relationship between the occurrence of slamming and the wave-induced bending
moment, A study of this problem on the Wolverine State indicates that it can be
solved for that ship (Chapter VI), but some further development is needed to general-
alize the solution to apply to any ship. In general, any slam bending moment in-
crease is a hogging moment.

The result should be long-term distributions of added bending moment due to

slamming for different sea conditions, obtained from:

(Probability of Slam) x (Prob, of added bending mt.)

It is possible, in principle, to combine this distribution with that of wave bending

moment for consideration of the probability of ultimate failure. Results for the
Wolverine State presented in Chapter VI, as well as Aertssen's data (18), indicate
that slam stresses are relatively small, however, and for some ships can be ne-
glected.

Of more importance may be the whipping that follows a slam and which will in-
crease the next peak sagging (arid hogging) moments. The magnitude of the increase
depends on both the phase angle and rate of decay, as well as the slam stress ampli-
tude (assumed to be equal to the whipping amplitude) to the maximum expected wave
sagging moment.

It is clear that the determination of dynamic hull loads associated with slam-
ming, which are of importance in relation to the probability of failure, cannot at
present be predicted with the precision of the static and quasi-static loadings.

As indicated in Chapter II, another dynamic load that should be included in

the ultimate load criterion is that associated with bow flare immersion. For many
ships in which the flare is small, and/or the bow freeboard is high, this factor
will not be significant. For others, such as the Fotini L reported in (3) or the
aircraft carrier Essex discussed in (96) it may be very important and must be con-
sidered in addition to bottom slamming.

Accordingly, the design procedure should include the calculation of initial

bending moment caused by bow flare immersion for several sea conditions, ccrrlder-
ing both the non-linear increase in the maximum sagging moment to be assumed for
design and the vibratory whipping that follows (5) .

Local Loads

Insofar as the primary load criterion is concerned, the principal local load
to be considered is that of hydrostatic pressure on the double bottom. As dis-
cussed by Evans (122), this results in a bending moment at the middle of each hold
and a larger one in way of each bulkhead. The latter implies a significant tensile
stress in the bottom plating and compressive stress in the inner bottom, both of
which would be superimposed on the longitudinal bending stresses. These local
stresses are higher in the vicinity of longitudinal girders in transversely framed
bottoms (.47) but would be more uniform across the ship in the case of the more com-
mon longitudinal double bottoms. Since the bottom pressure is higher when wave
crest is amidships than when wave trough is amidships, this effect is greater in
hogging than in sagging. Pressures can be calculated on the basis of static head
for the present purpose, although methods have been developed for taking into ac-
count the dynamic effects of ship motions (47).
 -54-
COMBINING LOADS

Finally we have the problem of combining the probability distributions of four

different loads for both (A) outbound and (B) inbound conditions. See Table II*
which lists the various loads and summarizes the steps involved in evaluating each.

The combining of longitudinal bending loads will now be considered in relation

to the possibility of damage and/or ultimate failure by buckling or by plastic flow
and permanent set. Local loads to be considered in a specific design will not be
included here.
Because of the difficulty in establishing a reliable zero base line in re-
cording full-scale ship stresses, it has been customary to present statistical data
in terms of peak-to-trough stress values. Thus no separation was attempted of hog-
ging and sagging stresses. However, since for design purposes it is essential to
provide independent estimates of sagging and hogging bending moments, wave stresses
or moments must be separated into two parts for combining with other loads. This
can be conveniently done by assuming that a predicted long-term bending moment curve
can be considered to represent either sagging or hogging (with opposite siens). Ac-
cordingly, wave bending momenta can be represented by two symmetrical long-term curves,
one for hogging and one for sagging, as shown in Fig. 20. Taking^account of the still
water bending moment caused by the ship's own wave results in a base liae shift, as
shown. In some ships, particularly those with flaring bows and sterns, the bending
moment may be non-linear with wave height, and in heavy seas there may be large differ-
ences between sagging and hogging bending moments (113). This effect could be eval-
uated by model tests and an adjustment of the division of total bending moment be-
tween hogging and sagging could be made accordingly.

Table II

DETERMINATION OF ULTIMATE
HULL GIRDER BENDING LOADS
Bending
Moment8 Successive Steps
End Result
I Still Set up typical con- Calculate bending Determine mean Long-term prob-
Water ditions with differ- mt. & shear for and std, devia- ability density
ent cargoes &/or each tion functions (A)
ballast arrangements
& (B)
(A) (, (B>*

II Forward Estimate average Calculate self- Average increase

Speed speeds (A) & (B) induced bending mt». in bending mt.
III Thermal Diurnal and seasonal Diurnal & Corresponding Probability den- Long-term prob-
variations in ambi- seasonal var- variations in sity function ability density
ent temperatures & iations In effective B.Mt. for sunny and for functions
cloud cover stresses cloudy weather
IV Wave- RAO's in regular Appropriate Response to lrreg. Probability den- Long-term distrs:
Induced waves: Vert. B.Mt. sea spectra waves sity functions: Combined H.& V.B.Mt.
Horiz. B.Mt. families Combined H.&V. Shear at Qtr. Tts.
Shear (& diatr.) Bending Mt. Torsion
Torsion
Shear at Qtr.Pts.
Torsion
V Dynamic Probability of Distribution Distribution of Probable addition Long-term dlstr.
■lamaing and/or of slamming phase angles to bendlnp moment of added dynamic
flare immersion and whipping Slamming bending moments
bending mt. Whipping
when they occur

* Note: In each case above, calculations are to be made for both outbound (A)
and inbound (B) loading conditions.

irii-'n -•■?■■ iWi'

min iim I SB
 -55-

SAG

FIGURE 20 - Typical Long-Term

Distributions of Wave
Bending Moment for
Sag and Hog
HOG

As previously noted, it is desirable to separate outbound and inbound con-

ditions, since not only are the corresponding still water loadings distinctly dif-
ferent, but in many cases wirve bending moments are also different. It has been
suggested that for either case it can be assumed that the still water bending mom-
ents over a long period of time — such as a ship's lifetime — are normally dis-
tributed. Since the still water bending moment stays relatively constant for long
periods of time — usually an entire voyage — it has the effect of periodically
changing the baseline' about which the wave bending moment varies.
To obtain single long-term curves for hogging and for sagging — including
still water bending — requires'that the wave bending moments be first expressed
as a probability density function (instead of a cumulative distribution). The func-
tions for still water and wave bending moments can then be combined on the basis
of joint probability, since the two phenomena are independent.

Let y te a random variable describing the wave-induced bending moment (hog or

sag) . The density function of y will be called p (y). Let x De a random variable
describing the still water bending moment, with density function p (x), which will
be assumed to be normal. We are interested in the distribution function for the
random variable z - (x + y), which is given by:

or
f
P(x + y) =/ p (x) p (y) dx

P(z)
s„00 P„(z-x)
w dx
■/•••

Since p,(y) is not known in explicit form, the above integral cannot be evaluated.
However, it can be determined numerically for any specific case. (See example in
Chapter IX).

Nordenstr&n has discussed this subject in greater detail (122),

We come now to combining the thermal effects, which it has been shown can be
interpreted in terms of a zero bending moment at night and an effective sagging bend-
ing moment in daytime — especially if the sun is shining. To simplify the problem
of combining loads, we can make the safe assumption that the thermal effects are al-
ways present, combining with wave bending. All wave data can be roughly divided into
 -56-
two classifications — those that occur at right, with no thermal effects and those
that occur in daytime, with thermal effects superimposed. This would lead to two
long-term curves, as shown in Fig. 21. We /.an conclude then that a safe treatment of
thermal effects is to shift the base line by one-half the amount of the average total
change in effective thermal bending moment.

Finally, local average stresses in inner bottom and bottom shell plating at a
bulkhead can be calculated and equivalent constant bending moment increases deter-
mined .

This, and any other essentially constant effect, such as the bending moment
caused by the ship's own wave system, can be allowed for by additional base line
shifts.

Thus, insofar as the possibility of damage or ultimate failure by buckling or

permanent set, the complete hull loading picture can be presented in the form of two
sets of curves, one for outbound and the other for inbound. (In the case of a ship
engaged in a triangular or tramp service, three or more sets of curves might be
necessary), Each set of curves would consist of a pair of long-term distributions of
bending moment coefficients, one for sagging and one for hogging, including horizon-
tal bending, with the base line shifted to allow for:
Bending moment due to ship's own wave pattern.
Effective bending moment corresponding to therma] effects.
Effective bendiag moment corresponding to local'effects.

With the above picture of expected quasi-static loadings available (demand),

the structural designer can in principle estimate the capability of the structure
and hence the probability of failure in a lifetime or in the lifetimes of many ships.
He can also estimate the probability or expectation of damages that do not constit-
ute failure. Strictly speaking the capability of the structure must also be ex-
pressed in probability terms, as explained and developed in the work of Freudenthal
(123). Various writers have pointed out that the probability of failure (or damage)
can be obtained by mathematical treatment of the overlapping probability density
functions of demand and of capability (124)(125)(2). However, for the present pur-
pose we will assumr that the capability is deterministic (i.e., standard deviation
is zero) . The proDability of failure is then simply the probability of exceeding a
specific limiting bending moment.

SAG

FIGURE 21 - Typical Long-Term

Distribution of Wave
Bending Moment, Sag and
HOG
Hog, with Thermal Stress
Superimposed

i rtil •-■- J
"
» . .-»„-■■i^^.. ii mi m*d
-- -— -— "■"~^-
 -57-
Since maximum stresses resulting from combined static and dynamic losing may
or may not be expected to occur simultaneously, a probabilistic model of such joint
occurrence needs development; it is conceivable that circumstances could cause
simultaneous addition of significant loads of each type, but the probability might
be smaxl. Pending the development of a complete model, at least the following alter"
native load combinations should be determined:

A. Static loading predominant

(1) The highest expected load value due to combined still water and
wave-induced bending moment, local loading, thermal gradients, etc.,
that could cause tension damage.

(2) The magnitude (and frequency) of superimposed dynamic loads occur-

ring at the Same time.

B. Dynamic loading predominant

(1) The magnitude of the highest expected dynamic bending moment.

(2) The highest quasi-static tensile loading due to bending moment,

local loading, thermal gradients, etc., expected to occur at the same
time. Since, as previously noted, the duration of. the load has a bear-
ing on the capability of a structure to resist dynamic loads, the dura-
tion of such loads should be specified.

It remains to determine an acceptable probability of failure and of damage,

which will be discussed in the next section.

PROBABILITY LEVEL FOR DESIGN

The final step in establishing design criteria for ultimate, bending is the
determination of a probability level to adopt for determining design bending moment.
It is necessary first of all to consider the safety of the ship and its crew. The
only sound basis for a strength standard in this respect is one based en probability
theory. We must be sure that the total risk of structural failure is never greater
than society can accept. Nor must the occurrence of structural damage that does not
endanger the ship be burdensome to the ship operator either through excessive re-
pair cost or too frequent withdrawal of the ship from service. As progress is made
in developing techniques for predicting long-term trends of various loads acting on
a ship's hall, along with sophisticated techniques for determining detailed distribu-
tions of stresses, the time is approaching when we should decide what risk cf struc-
tural failure is acceptable. Here the classification societies can be of assistance
by analyzing their records to determine the number of failures occurring over the
years in ships of different types and sizes and computing the corresponding proba-
bilities that have presumably been considered acceptable. One question is, should we
use the highest bending moment to b :■. expected once in a single ship's lifetime or
once in the lifetime of a fleet of !'-. i 00 or 1,000 ships?

J. F. Dalzell, in an informal memo to the Ship Research Committee (12 May, 1970),
gave a valuable analysis of some published Lloyd's Register data on merchant ship
losses (126), which covered 18 years (1949-1966) and 390,GC0 ship-veais of service
experience. He assumed that the losses designated "Foundered" (31% of all losses)
were cases of complete structural failure, although there were no doubt numerous ex-
ceptions. Assuming a 25-year average ship life, he arrived at Table III, showing
probability of failure (in a ship's lifetime) for different sizes and types of ships.
 -58-

TABLE III

APPROXHfAIE PROBABILITY
OF FAILURE: "FOUNDERING" (126)

~hip ~lze
Ship· Iy~----­
( Grt'!;s fons) Dry Cargo Tanker

100- 1,000 T.
1,000- 5,000 T. 0.11 0.006
.'i,OOO-lO,GOO T. 0.04
10,000 Ton~; >:.tp 0.02 0.008
0.006 0.006

~- ... --·-··--- . -------·- -·· ·~-

All Sizes 0.05

,\. ;ninci;:· •l ~-<·.!son_ for tile higher va.l.ue£> for dry cargo ~hip5', pdrticularly in
the Si~<1L. sizes, ts ~-rc'Jably the fact 'that l.osse!: from othC"r causes are incha!-~d --
such as capslzf;:.;, (lo~.: arability, f1.ood!.r..g, or cargo shifting) or foundering ft·om
floodJng (l1:.rc<\ ~v·:er failure, leakar,c, etc.). A figure scme'\o.'hcre bet\oTeen 0.00. :u~d
0.006 wou1c! be :i reascnC~b;.e ·:a~·1e for the probability of !'ailurc that haR been ta· :t.-
ly acc~ptcJ 0v~r th~ past 20 years for large oceangoing ship&. In proposing a Rpe·
cific f!.g•.J:•~ ror .l tW\J dc:>i,;n Ct'it,>cion, however, ~.:e feel that a :nure r.oncervo:.t.'.Ve
f igur.e shonL! Je 1dopted, ;md Da tzell •s ~uggestlor.. of 0.001 is tentat.~.11ely rropc-se(~.
'l'1ds .Lmpl ie:-; tttat ;•:erchant sh Lps '.NOul:i he designed with a probability of ultimate --
c:r eat'lstn'p~!ic -- !.:~11ure of ,.o greater than 0.001, 1. ..:,,, that a new ship "'ouJd lwve
,, cil:1nce c t t:ot ovc ~· cne in ;, thous<tnd of i.~ilurc> tlurir.-1: a nor::~c1l life span. S1:e
also discussions by J. F. Dalz~ll and by M. K. Ochl of (3), and authors' reply.

Although our p~lncipul interest is in extreme bending ~onencs that cause com-
pl.etc failure, lesser values of bending m<':t•<!nt than occur 1:1ore frequently in a ship's
lifetime are also of interest. .:..s noted by Sruith in dis.:ussion of (li') • "If bend-
:n~ moments v..:re estbtatea statistically ::hen !:: '1·10uld be necessary to specify for
c:1ch level of da:rtJgc an acceptabl.! probabilio;y t~f occurrence. "

Sume statistics are avail::.ble on ship structural damag.:!s from various sourcP.s.
A part icu]arly valuable study ma1e by Lloyd's Re~ister (35) co'.·ers dry cargo ships
and tankers (not bulk carriers). It relates the number of cases of fractures in
strecgth deck and :;h~ll plating to ship yeare of service, as sho'lr.l in the accompany-
ing Table TV. In response to our inquiry to Lloyd's Register regarding th~ data in
Table lV, ~r. G. J. Jensen advised that not all of the crackR were of the brittl~
ldnd. Presumably the others were mostly fatigue cracks. "However, atthis stage it
•..rould neither be possible to indicate the percentages of £:tch type of fracture occur-
ring, nor t::elr seriousness, without re-opening -the investigation." Hr. Jensen also
stated th;ot ,,[ all t~c fractures tabulated, only _one resulted in the loss of a ship.
"The \~or_l,i Cc':lcor.! built in 1952 1>1:-oke irr two in a brittle fashion. The fracture
was trac~J t~;- a ;;.:trd spot .caused by welding, in way of On\! of the bet tom longitudinal
endings."
Of pArticular significance are the figures for "occasions per 100 years,"
which --except for old tankers --run at or belo~ 1 per 100 ship-years. Titis ap-
pears to be an acceptable figure. ~onsid~ring that few if any of these fractures
actually resulted in the loss of a ship. Considering a ship•s lifetime to be 25
years, 1 damage per lOJ years would be equivalent to a lifetime damage probability
of 0.25 (i.e., 1 da;nage in the lifetime of 4 ships.)

... -vi_, A'i/AILABLE COPY

 -59-

TABLE IV

FRACTURES IN STRENGTH DECK AND SID~LL PLATING

OlJ • ~.j L'\.1 t

;/tJmi.H.•r pf ~hlpz.t nl ririk ~ 1" HH'J Jf>~ :JOH

l\•l;!l :>hip )'t:,L!'l-1 :il... (\' let~ VJ92 .', sor, 1 0~~ td~
;;qr.:]JPr uf ~hip~: :lffl•rtt•d )'J :!9 R7
l'l•r,·t::lLH'L' ()( •,l!I!J'• ill ft:1~t<•d ~· • lJ () :'.l'l ~ .8fl
Sldp~ .,:f(•o..:t<.·d 111.~t 100 yt':Lr~: l,.',.', !) • ()(, () 'Hr,
,!nn.b v: • ! .,,. t~ ,l.., ( '•n:t 1/; Hl IJIJ
I'• I J't' r ] 'J() VL'I\Tfl
::tj' 1,.1•r .q ::111]·'1 wr tll :q'l.'t' !.11
~' 'Of/ I),(Jf ,j. q'
'Jc" ·,·n Ill:'
-:(t.'t.')

]) .',t, I'!

"0111'~ ~:• !11~: •>!1lr~; l·q!i.:. l~··~3-J957 inciusivt..•.

·:: .. \.'' 17\t',\:l-> ~!tlp~; 1Juilt J(_JSH-l 1)f,7 Inc: trii'll'
i't..·.lr·, ·'i :-l·n·t••: l't•tJntt·LI tt• ru,J of ,.,1 .. :1 pert,,J, .e., \')~!7 :1nd !9"7
1.11': !\'l', :r tu L'~:lt. fr;\1~: eLlS<;, •.·:d,·!w•Jl'r' t~J tl:t• -:lll"o)'tt•r.

:, bLJsic aiJpro;;ch to cletenn:Lning the probability level to be used in a design

criterion is that of "expected loss,"* whlch has been summarized in convenient form
by F;c,udcnthal for appli.cntion to maritime structures in general (127). It is based
en ci1c :1rLncipJ.c that th•:! best design is the one that minimizes the expected total
co:c:t, '.·:i1c:-o :: 1;c• Lltl:l~r con~;ic~ts of the sum of initial cost, failure cost, and damage
COl3t, a;,; c:-:pl.~t.l.n'"'d behJY.'. It is very difficult, .if not impossible, to assign a dol-
lar v.J.lucc to tbe passengers and crew, but it '.Vill be as~>umed -- perhaps over-
nNim istic,1] ly -- that in th'Ls dny of efficient communication and life-saving tech-
nclD,~;y 'l ~;!1ip u:ny founder without the crew being lost.
A distinction is made bet\Veen "failure," in which the safety of the ship is
~nci~m;;e1·od or tile s:Lip C~ctually breaks .in two, and "damage," in which local cracking
nr ;li!Ck..Ung of r:1a Ln hu.lJ girder elements results in a requirement for repairs to be
:'L'';.,:,. ALl damages not involving the main hull girder are excluded, since presumably
t;1cv would not be uffected by any change in the main hull structural design.

Expressed as an equation, the total expected cost to be minimized,

L = I + P1 F + (1 - p 1) Sp2
I in.Lti:1l cost of the ship (or structure)
P! probability of failure (in a lifetime)
F nnticlpated total cost of failure (replacement cost +.cargo loss +
temporary charter of replacemen·t ship + loss of business from customer
rcactJon~; +cost of pollution or other environmental effects, etc.)
s anUcipated cost of damage or "failure of function of surviving struc-
turo" (the "Snccess cost"), i.e., cost C•f repairs and of associ.:1ted
co:.::ts of damar,e tklt does not involve the loss of the ship.
p2 the expcc~ation or expected number of such damages.

-;'c
Tlli~; j:; t: 11e so-c.:~1led Baysian decision rule.

BEST AVAILABLE COPY

 -60-
For the case of the design of a ship's main hull structure, we may hypothesize
that the probability, plf of failure that can lead to the complete loss of the ship
is very low. But it might occur on the basis of some combination of extensive buck-
ling and yielding or it might occur primarily in tue form of brittle fracture, per-
haps preceded by fatigue cracking.

The expectation of other damage, p2, that would require more or less extensive
time out of service for repair depends on any one of the modes of failure previous ;
discussed — or of a combination of them. In fact, a ship might experience one o-
more such damages in several modes during its lifetime. Furthermore, such damage
might be of different degrees of severity. Hence, in our case the term Sp2 shou i
actually be a summation,

ZSP2

For example, a particular hull design configuration with certain specified :ant-
lings might have a very low probability of one severe buckling or a tensile fail i.*e
of the main deck. But the probability of local tensile failure or fatigue crack:-
at a hatch corner occurring several times might be relatively high.

Hence, we come finally to the concept of determining for each failure mode tie
probability of failure in a lifetime, and for each damage mode the expectation of
damage, each of which should be multiplied by the corresponding cost.

In principle the total expected cost, L, can be evaluated for several alternate
hull designs and the optimum design determined graphically. The.following types of
damage should be considered in addition to ultimate failure:

1. Panel buckling that is not immediately dangerous.

2. Excessive yielding.
3. Fatigue cracking.

Because of the many uncertainties involved, brittle fracture is excluded from con-
sideration here.

A sample calculation will be presented in Chapter IX for the the Wolverine

State in order to ascertain the significance of the proposed approach."

As noted by the I.S.S.C. Committee No. 10 (20) it should be possible in due

course to relate probability of failure to a conventional deterministic load and a
factor of safety. This may be a desirable thing to do for general guidance in order
to correlate any new approach with the empirical standards that have been successful
in the past.

FATIGUE CRACKING

One approach to structural design relative to the secondary criterion of fatigue

loading discussed under Critical Loads (Chapter II) was simply to make sure that the
probability of exceeding the yield point at critical areas of stress concentration
was at an acceptable level. However, this approach is imprecise and may lead to ex-
cessive scaltlings. Therefore, it appears that as complete a picture of cyclic load-
ing should be furnished for the use of the structural analyst (and researcher) as
possible.

MMtfÜa^MiMtlMilll^irtflMaB--—■---»•-■~--'""-'-"---—-''-'>"^"--*" , ^»^■»a.^Llnr»^».«^!«..!
iiiiintiiriniHuriiimi
 -61-
From the fatigue viewpoint the type of loading is one of cyclic load reversal,
usually with fluctuating mean load and possible occasional overload at points of
stress concentration. It is further complicated by diurnal thermal stress variations,
These loads are tabulated below along with the estimated cycles of load reversal for
each in a typical ship's lifetime:

Still water 3^0 8

Wave Bending 10 -10
Dynamic 106
Thermal 7000

The fluctuating mean load is the so-called still water bending moment, discussed
in the section on Ultimate Loads. In general the specification of two probability
curves, one for outbound (A) and the other for inbound (B) conditions, will provide
the information needed for fatigue design. However, one additional item is needed:
the time that the h p operates in condition A before changing to B, time operating
in condition B, etc. In general both times will be equal simply to one-half the
total round voyage time and will be measured in weeks. To be more accurate the ef-
fects of consumption of fuel and additions of salt water ballast should be included.

The cyclic loading consists of the low-frequency wave-induced bending moments

and the high-frequency dynamic bending moments previously discussed. Their phase
relationship is perhaps of less significance for fatigue than for brittle fracture.
At any rate, long-terra cumulative distributions of both should be available as part
of the load determination for ultimate loading. From these distributions one can
obtain cyclic load spectra in the following manner. The. reciprocal of the proba-
bility is the number of cycles, n. For a ship's lifetime of n^ cycles, a scale of
n - n - n is then constructed on the distribution plot. Then n gives the number
of cycles expected in the ship's lifetime of any desired level of bending moment.
See Figure 22, which deals with wave bending effects only (128).

Finally, information should be provided on the expected diurnal variation of

thermal effects, as previously noted.

The above information should provide the data needed by the stress analyst to
evaluate the cyclic loading, variation in mean stress, and — with estimates of
stress concentration factors — the frequency and direction of local stresses.
Fig. 23 was developed for the case of constant mean value (128).

The object is to provide a means of estimating cyclic loading'that can eventu-

ally be balanced by the structural designer against the endurance properties of
the structure. Thus he would be able to provide an efficient structure in the
design stage that would have an acceptably low probability of fatigue cracking in
service. However, it should be emphasized that, since in general the safety of the
ship is not threatened by a fatigue crack of a certain maximum length, a higher
probability of cracking can be accepted than for ultimate failure of the hull girder.

If the general application of more rational design standards should in time

result in reduced hull scantlings, then the incidence of fatigue cracking might in-
crease to an unacceptable level. In this case some modification in strength stand-
ards by classification societies might be called for.

BRITTLE FRACTURE

From the point of view of ship structural design the possibility of failure by
brittle fracture requires careful consideration both independently of and in combina-

s^^^,^^ ■■ --—^.,.,-^.,-,^.w—_^. ^.^ .^^..^^(.rt^r-- ..ml^-.-.~-^--^*mmm

 -62-

io~7 icr6 io~ö icr* io~5 io-2 O"1

PROBABILITY OF EXCEEDANCE

ioz 10 s I'O 4 io5 io* io1 so*

3 io
NUMBER OF CYCLES
FIGURE 22 - Long-Term Distribution of Bending Moment
or Stress, with Reversed Scale Showing
Cyclic Loading or Number of Cycles of Each
Stress Level in One Ship Lifetime
(108 Cycles)

Ld
Q
§
2 7000
<
FATIGUE LIFE
5J6000 PLAIN MS SPECIMENS
O
2 5000
i
I
«^4000 I
I »FATIGUE LIFE
CO
CO
1 »PLATES WITH BUTT WELDS
£3000
h-
co
%>, 1 V »CONTAINING FLAWS

Ü- 2000
o ^A \V
I >x
TYPICAL FATIGUE LIFE
NOTCHED SPECIMENS
! ^ CORROSIVE
CONDITIONS
_l
~I0 I02l03l04l09!06l0rl0,l0*l0'0
NUMBER OF CYCLES

FIGURE 23 - Example of Application of Cyclic

Loading Curves to Study of Fatigue (128]

■ - - -'—
 -63-
tion with other modes of potential failure. However, the applicable hull loads are
essentially the same as those discussed under Ultimate Bending Loads. Hence, the
long-term distribution curves developed there, combining still water, wave bending,
dynamic, and thermal loadings, should apply to design in relation to avoiding brit-
tle fracture.

Since brittle fracture i.c a tensile phenomenon, buckling need not be considered.
But the structural designer must consider many factors in addition to mean stress
levels — stress concentrations, weld defects, fatigue cracks, ambient temperatures,
strain rate, steel qualities, locked-in stresses, and metallurgical effects of the
welding process. Most of these factors involve many uncertainties, and therefore It
is more difficult to predict a probability of failure for brittle fracture than for
failure in ultimate bending.

Ideally, brittle fracture could be avoided if it could be established that the

temperature in service never drops below the transition temperature of the steel,
considering the steel properties (or chemical composition), including the effects of
welding, the plate thickness, the nature of possible weld defects and the severity
of local concentration factors in the main hull girder.

At present the above determination is not possible. Furthermore, the determina-

tion of dynamic hull loads associated with slamming, which are of importance in rela-
tion to the probability of failure, cannot at present be predicted with the precision
of the static and qussi-static loadings. Hence, it does not appear feasible at present
to adopt a load criterion for brittle fracture, even if that were desirable. Instead
it is customary to adopt a fail-safe design procedure, by providing a sufficient
number of crack arresters to insure that a single crack is limited in its propaga-
tion sufficiently to avoid endangering the ship. These crack arresters -:an be
either riveted seams or strakes of high ductility (notch-tough) steel.

The rational design philosophy would then seem to be as follows:

1. Recognize that with present design standards and material

quality brittle fracture seems to be under control, if not
entirely eliminated.

2. If design and material standards do not change, brittle fracture

need be considered only in maintaining good design, construction and
operating practices.

3. The application of new design standards, based on the quasi-static

primary load criterion proposed herein, may lead to suggested increases
in working stress levels.

4. Such increased working stresses should be accepted only if either

calculations show probability of brittle fracture is not increased
or more stringent material requirements are introduced.

IX. SAMPLE LOAD CALCULATIONS

INTRODUCTION

It was felt at the outset of this project that a numerical example of hull
load determination, using the procedures developed during the project, should be
carried out, leading to specific load criteria for design of one type of ship.
The objective was twofold: the example would illustrate and explain the procedures
developed, and it would give an indication as to how the proposed load criteria
compare with conventional standards.
 -64-
It was decided that the most suitable ship for detailed study was the S.S.
Wolverine State. Not only are service wave bending stress data available, but
information could be obtained on dynamic stresses and on still water loadings.

The load calculations followed the procedures described in the preceding chap-
ters. Considering first the loads affecting ultimate bending failure, each step
will be described in turn. Then the combined effect of all loads will be considered
and interpreted in terms of a simple design bending moment for hogging and sagging
conditions, outbound and inbound. This load criterion will then be compared with
conventional design standards.

Finally, brief consideration will be given to the cyclic loading pattern as

a criterion relative to possible fatigue damage, and to a criterion for brittle
fracture. For this ship, with relatively heavy shell plating, small hatches, and
low length/depth ratio it was felt that criteria of shear, torsion or elastic de-
flection need not be considered.

STILL WATER LOADS

Estimated Bending Moments

During most of the time that data were collected on the S.S. Wolverine State
the ship was engaged in North Atlantic service. Available data on service drafts
indicated that the ship usually operated at drafts considerably less than full
load, both east and westbound. However, in order to provide a typical numerical
example, it was felt to be desirable to assume a fully loaded cargo condition on
the outbound voyage and typical light loadings inbound.

Unfortunately, it was impossible to obtain detailed distributions of cargo

and liquids for a sufficient number of actual voyages to obtain a statistical pic-
ture of still water bending moments. Hence, calculations were made, using the
owners' loading manual as a guide, to supplement actual loading data. A normal dis-
tribution of still water bending moment in the light cargo condition was con-
structed as shown in Fig. 24, based on the available loading data and on calculated
highest and lowest expected values in a 25-year lifetime. To determine the extremes,
R ACTUAL DEEP DRAFT CONDITION
PROBABILITY * ACTUAL LIGHT DRAFT CONDITIONS
OF
OCCURRENCE + EXTREME LIGHT DRAFT CONDI TIONS
SEE TEXT!

INBOUND
(LIGHT DRAFT)

50 50 100 ibO 250

SAGGING STILL WATER BENDING MOMENT, 1000 F TONS HOGGING •»

FIGURE 24 - Estimated Distributions of Still-Water Bending Moments,

S.S. WOLVERINE STATE

inn in if i im ii1 ■ --■■-•■---" -^--■-■-^- at

I illMM.nM«HlllM
 -65-
the amount of cargo typically carried in the light load condition w^s imagined to
be distributed more and more toward the ends of the ship until an estimated ex-
treme condition was reached. We assigned to the resulting still water hogging
bending moment (239,900 ton-ft«) a probability representing one occurrence in 25
years.

For 12 round trips per year, there are 300 crossings in each direction in 25
years. Thus the highest expected single occurrence would have a 1/300 ■ 0.00333
frequency, which corresponds to ±2.71 a, where a is the standard deviation. A
similar approach was used to determine the minimum hogging condition in 25 years
(75,100 ton-ft.), which corresponds to 2.71 a below the mean. The mean value for
the normal distribution is the average of the highest and lowest calculated values.
Thus the mean and standard deviation, and from them the normal distribution, were
determined as shown in Fig. 24, along with actual values determined from the load-
ing data.

The mean still water bending moment for the full load condition was calcu-
lated from one actual voyage at nearly full load, and it was assumed that the
standard deviation would be the same as for the light cargo condition. The re-
sults of the calculations are given in Table V. All bending moments are in ton-
ft.

TABLE V

STILL-WATER BENDING MOMENTS

S.S. Wolverine State
Bending Moments , Ton-Ft.
Full Load Light Load
Outbound Inbound

Mean 40,000^ hog 157,500 hog

Standard deviation 30,400 hog 30,400 hog
10% are below 800 hog 118,300 hog
10% are above 79,200 hog 196,700 hog
Highest in 25 years 122,400 hog 239,900 hog
Lowest in 25 years 42,400 sag 75,100 hog

Assumed the same as in Light Load Condition.

The above maximum values have been inserted in Table IX, along with extrapo-
lated 1000-ship values and other results to be discussed below.

Classification Society Requirements

It should be noted that although classification societies do not in general

base hull girder strength standards on a direct addition of still water and wave
bending moments, still water moments are taken into account at the present time
(See Chapter III) .

..— ■■ ...,-^,~,,«;.^-.,.. 1 fgfffjg „ | 11;,-1. ,n. I ■ IMt

 -66-
For the Wolverine State the maximum still water b~nding moment witho11t penalty
is calculated by ABS 1972 Rules to be 136,000 ton-ft. It may be seen that this
value would be exceeded by the calculated value of 239,900 ton-ft in the hogging con-
dition, inbound, If these conditions were included in the data submitted to the
AJJS at the time of the design of a similar ship trJday, some addition to the midship
section modulus requirement might be made, At the time the ship was built (1945),
however, '·hr ;:e was no expLicit still water section modulus requirement in the ABS
rules.

FORI-.'ARD SPEED EFFECT

-------------
The bending moment (sagging) induced by the ship's own wave at forward speed
ln calm water was estimated on the basis of model test data on similar ships (36).
~t the design speed of 16 knots the result is a sagging bending moment of 2700 foot-
tons in the heavy (outbound) condition and 5200 foot-tons in the light (inbound)
::ondltion. Thus the forward speed effect reduces the still water bending moments,
wl1.!.cli ure hogging moments in both, conditions. The model tests show that the maxi-
m•Jm 110_gging_ momC;nt due to forwa1:d speed occurs when the shi.p i.s at very lo\v speed,
nt:d t:lle reduc tlon in still "-'a ter bending n;oment at normal speed becomes zero.

l'h~ aLJove figures have been entered in Table D:, in which a summary of all bend-
in~ mome~ts is given. :l.vcrnge values have been estimnted for the table on the basis
of servic~ speed data.

THS~·tA.L EFFECTS

Nt.;:'..t an estimate was made of the thermal stre"Jses to be expected under differ-
ent syn1111et rical condj.tions of sea/ air temperatures (sun overhead). Since calcula-
tions have shovm that symmetrical heating of the deck results in compressiv0.
stresses in deck and bottom plating of nearly equal magnitude (119), only t!1e
changes in deck stress were computed~

The basis for calculating thermal stresses is described in Chapter VII. Cal-
culations were made for five values of diurnal temperature change, representative
cf typical North Atlantic conditions. Average air temperature changes were de-
termined from ships' logs covering a two-year period. The results were:

Season

Hinter
Spring
Summer
Fall

Because rhe seasonal variations were small, differences were ignored and a yt·arly
average of 10" was assumed.

An additional t1T of 40° was assumed for insolation with full sun, since the
d0ck color is d.nrk grny, ns in the Esso Ma~sin l~Stlmutc (see Chnptar VII). This
figure represents full sun conditions; so adjustments were made for cloudy condi-
tions. Available seasonal and annual cloud cover data for the North Atlantic route
(82) are given in Table VI.
f&l i I PtP it(
BEST AVAilABLE COPY
 -67-

TABLE VI

DATA ON CLOUD COVER AND ESTHfATED

TEHPERATURE DIFFERENTIALS ,
NORTH ATLANTIC

j i·ledian
Factor for % of time % of time
Season I Cloud Cover <
______..._______________
2/8 cover > 5/8 eover _
---------~-------------
\·Hnter 1 6/8 10%
i 6/8 20 6()
: 5/8 20 60
' 6/8 ' 20
------------- ------L- - - -60------
Jvg.· 6/8 ! 20 , 60
--- --~- ----------1------- -~---------- - - - -
Diurnal ,'}T Estimated
Frequency of
~_: L.Jud ccvcr Occurrence

40° 10° 50° 10:Z } ?O%

30" 10° <'t0° 10% ~ •
I. / ~) 20< 10° JOo 20%
r)j ~ 10° 10° 20°
00 300 } 60"'
.:)/,_q 10° 10° JOO /o

eli f ferences \vl.cre ignored· in this table, and the ·yea-rly average
·'; :-. '·he· c:.·:.::un;Jl
:_ l.~~t,rc:c- ~'o-r :·ll:ud
cover ·,.:ere us(~d. ln order to relate the inso1~tion tempero.ture
lc-: :.:l1e c}c,~;;] ;:>ver infomation, -v:e assumed th;t the insolation £11' is directly pro-
ru ~t :on a] to the amount of cloud cover, while ti1e air temperature change (10°) re-
!:la l.::•:d c0n•,c tnnt. The resulting dtur:1al 6T 's for different c1ou1 conditions are
::. (;n gi.ver~ in Table VI .

'!ile r-::lcu1:lted compressive >;trcssPs at decv edge for the Wo_lverine §tate are
;-;' "''''l i e1 T·tblr:: VI T, and the <.:alculatcd stress distribution is shown in Figure 25.

TABLE VII

CALCULATED CHA.t'-;GES If~ STRESS A.c'JD

i:QUTVALENT RENDING MOl'!ENT,
~), ~'. \WLVERT:-;E STATE IN NORTH ATLlu'lTIC SERVICE

Frequency Cif Deck Corresponding

Gccttrrc~ncc, Cor.:pres s ive Equiv. Sagging B.M.,
Dl.t.:lr1:1l /·.T Table ';I Stress, PS1 ton-ft.

50° 10% 2170 47,700

40° 10Z 1740 38,250
JO~ 20;:, 1300 28,600
20° 30? 870 19,100
10~ 307 430 9,500
----------------------
~-:eightecl average 22,900
 -68-

Al 50

THERMAL STRESS

-AT 10

UPPER DECK

RIVETEDSTRAP
r o« i 0"

I 3rd DECK
NEUTRAL AXIS
J-l.

FIGURE 25 - Calculated Thermal Stresses,

S.Z. WOLVERINE STATE

Next ai: "equivalent." sagging bending moment that would produce the same deck stress
was c v'jtad for each value of diurnal Vi', as listed in Table VII. The weighted
average was then inserted in Table IX as the average sagging value. On the assump-
tion .it the bottom compressive stress in hogging would be approximately the same
as ' calculated deck stress in sagging, an effective hoggin0 bending moment of
22. t-m-ft. was calculated and entered in the table.

WAV: LOADS

Comprehensive computer calculations were carried out for the S.S. Wolverine
State for two conditions:

(1) Full load.

(2) Light load.

The responses covered by these calculations included:

Vertical longitudinal bending moment.

Lateral
Combined vertical and lateral longitudinal bending.
Torsional moments.

The calculations proceeded in the following steps, in accordance with the pro-
cedures described in Chapter IV:

im —iiriTiiif — mmmaam «Ü
 -69-
Calculation of response amplitude operators at all headings to
regular waves.
Comparison of above with model tests.
Prediction of response to irregular short-crested/seas described
by the H-family.
Estimating long-term cumulative distribution for North Atlantic
service, based on the weather statistics given in Table VIII.

TABLE VIII

WEATHER DATA FOR NORTH ATLANTIC

Probability Average Significant Height Range of

p H-I/3 (ft.) " H-l/j 'ft.)

0.8454 10.0 5 - 15
0.1330 20.0 15 - 25
0.0201 30.0 25 - 35
0.0014 40.0 35 - 45
0.0001 48.2 45 and aiove

The effective vertical bending moment giving the combined effect of vertical
and lateral bending was calculated by the method described in Chapter IV. The
ratio Z /Z^ for the Wolverine State is 0.8446.

The program SCORES (49) was modified to suit the present purpose by appending
a memory block to carry over the necessary information, so that it would still be
possible to run SCORES in its origina.1 form for other purposes. The calculations
necessary to give the long-term distribution were appended in the form of a sub-
program.

Long-term calculated results are presented in Figs. 26 and 27, along with curves
obtained from full-scale statistics for comparison. See also (10). Calculated
bending moment results in regular waves are compared with model test results in (53) .
Values from Figs. 26 and 27 for combined vertical and lateral bending moments have
been entered In Table IX.

DYNAMIC LOADS

We come now« to the consideration of dynamic loads on the S.S'. Wolverine State.
For a ship of this type, springing would not be expected, nor is there sufficient
flare to produce significant bow flare immersion effects. Stress records confirm
that neither of the above effects were experienced significantly in actual service.
Furthermore, a student thesis project at Webb with a jointed model of Wolverine
State confirmed that only negligible springing stresses could be developed in tiie
model tank.

However, slamming stresses resulting from bottom impact forward, followed by-
vibratory whipping response, were bcth expected and found in Teledyne records taken
in the light load condition (103)(109).

As explained in a previous chapter, it is theoretically possible to predict the

occurrence of slamming if one can specify the level of relative vertical velocity at
 -70-

S£ WOLVERINE STATE

§--)?<'■■

COMBINED VERTICAL «■ LATERAL

i .50

LOG JUMBER OF REVERSALS

FIGURE 26 - Long-Term Distribution of Bending Moment, Light-Load Condition

LOG Of NUMBE « OF KEVfRS*LS

FIGURE 27 - Long-Term Distribution of Bending .-loment, Full-Load Condition

■■IIIIIMIIIII ""■'-"■■""■' i in iMiiiriiirfwuif-^^'"

„.,.,.■■.,. .rill II .«*ffl
 -71-
which it will occur for the ship in question. Since there are still a number of
doubtful questions about this procedure, reference was made in this case to the re-
corded data on the S.S. Wolverine State. The stress records showed that, over the
entire period of data collection, slamming (with stresses exceeding 1.0 kpsi) oc-
curred in 16% of the records. This can be considered to be an indication that in
normal operation the probability that slamming will occur in any 20-minute period
is 0.16 in the usual light load condition at which the ship operated. (Average draft
forward was 16-20 ft.)

In the hypothetical full load condition (forward draft 29 ft.) the probability
of slamming is greatly reduced, and for the present purpose was arbitrarily as-
sumed to be 0,

Referring to Chapter VI, the following conclusions may be restated here:

1. the slam stress distribution a is exponential:

p(og) = a exp {-a(os- c) ]

where o = o /RMS
s so
c = 0.36
a = 1

2. the distribution of the phase angle $ is normal

2
P<<|>) = exp
0-/27
2 [ a )

where u = 0.406 radians,

a = 0 .381 radians.

The theoretical solution to the problem of calculating the distribution of the

increase of the slam stress 5 and the whipping stress 5 over wave bending stress
can be given briefly as follows:

p(«s) = f-2 exp J- -Jj AU.o.w)

where 6 is the increase due to slamming, and the first two factors —Dress the ex-
ponential distribution of additive 5 , while the third factor A, deals with the
truncation due to non-additive : .
s
The expression obtained for whipping stresses leads to a Rayleigh distribution:

P(5 ) - 2b 5 exp (-b 5 )

W W \v

where 6 is the whipping stress and b is a function of Ca, u5 e,c), All of the

above relationships can also be interpreted in terms of the corresponding effective

bending moments.

mtamtmm
imi n rw tinri
 -72-

Un the basis of these distributions, using available slam stress data for the
Wolverine State, as previously discussed, we arrive at a maximum expected increase
due to slamming over the lifetime of the Wolverine State of 6.81 KPSI, correspond-
ing to an effective increase In the sagging bending moment of 133,000 ton-ft.

Similarly the maximum expected w ^ppicg stress was calculated at 6.95 KPSI,
corresponding to an eff* :tive increase of hogging bending moment of 136,000 ton-
ft. These figures have been entered in Table IX, along with figures calculated
for 1000 ship lifetimes.

The results of all the lead calculations for the Wolverine I _:_e can be tabu-
lated in the manner shown in the accompanying Table IX. Results are given on the
basis of both of the following long-term assumptions (where N is the number of
wave bending cycles):

(a) One ship's lifetime: N = 10c for Z5 years; approximately 300 round
voyages In North Atlantic service.

(b) The combined lifetimes of 1000 ships (N = 1011), i.e., bending moment
expected to be exceeded with a probability of 0.001 in a ship's life-
time.

The latter has been suggested as a basis for a rational design criterion with re-
spect to possible ultimate failure of the hull girder.

The dynanriv loads associated with slamming have been entered only ror the
light load condition, under the assumption that slamming seldom if ever occurs
when fully loaded.

If all the maximum stresses in Table IX are added directly, an unrealistically

high value is obtained, because all of the maxima will probabl/ not occur at the
same time. Hence, a reduced total has been calculated, oased on combining still
water and wave bending by the method of Chapter VIII. The results of this cal-
culation are shown in Fig. 28, using the normal distributions of still water bend-
ing moment previously "iven and the combined vertical and lateral wave bending
moment data on which i-igs. 26 and 27 are based. The reduced totals in Table X were
then chained by adding the forward speed effect (if any) and one-half the average
thermal effect to the values read from the curves in Fig. 28. The dynamic effects
of slamming or whipping have been added directly (light load condition), under the
assumption that extreme wave bending and slamming are likely to occur at the same
time. This assumption is obviously rot correct, but it probably does ^jt result
in a large error because both effects require rough sea conditions. The expecta-
tion that at the same time that slamming gives an increased dynamic load (demand)
on the structure the capability also increases is a favorable effect of unknown
magnitude.
STRUCTURAL CAPABILITY

Having estimated *-he probrble lifetime combined loading for the Wolverine State,
separated into maximum hog and maximum sag, outbound (full load) and inbound (light
load) , it is of interest to see what these results signify in relation to r.-n 'entiona
design standards. This requires that we consider a "rational" evaluation if capa-
bility as well as of demand.

l-iist, iL should be noted that if the hull of the S.S. Wolverine State behaved
like an ideal girder and ensile considerations governed, the bending moment
 -73-

f-100 000

200 000

FIGURE 28 - Long-Term Distributions of Combined Bend-

ing Moments: Wave Bending (Vertical and
Lateral) and Still-Water Bending

TABLE IX

SUMMARY OF LIFETIME HULL LOADS

S.S. WOLVERINE STATE
Showing individual maximum values
Bending Moments in Ton-Ft.
Max. Max.
Average Lifetime 1000 Ship Lifetimes
H
og Sag Hog Sag
Full Load - Outbound Hog Sag

Still water 40,000 122,400 42,400 176,800 96,800

Forward speed effect 1,000
Thermal effect* 0 2,700 0 2,700
22,300 22,900 46,400 47,700
Wave-induced 46,400 47,700
15,000 15,000 153,000 153,000
Dynamic 207,000 207,000
0 0 0 0 0 0
Light Load - inbound
Still water 157,500 239,900 -75,100 294,300 -20,700
Forward speed effect 2,200
Thermal effect* 0 5,200 0 5,200
22,300 22,900 46,400 47,700
Wave-induced 46,400 47,700
196,000 196,000 268,000 268,000
Dynamic. [( Slamming
Sla 63,650
1 Whi pping 133,000 170,000
31,300 136,000 168,000

On compression flange only

 -74-
TABLE X
CotiDINED LIFETUm HULL LOADS
S,S, WOLVERINE STATE
:;ll,lh'i.:lg probable comb.i.ncd max.i.rnums ]3ending Homcnts in Ton-Ft.
Max. Max.
Lifetime 1000 Ship Lifetimes
-----
Hog Sag Hog Sag
Fu.1.1 Lone! - Outuound
~.:; t Jll ·,,/ .:1 t c~ r l f!
h'ave ) '· 2UG,OOO 118,000 252~000 170,000
Fo~.v.::1:-d spL!ed eff.::ct, average 0 1,000 0 1,000
T'i"crmal ~_:ffe·-~t, :t.vernge 11,150 11,450 11,150 11,450
Dy r. .J.r:1i.c 0 0 0 0
----------
TOTALS 211,150 130,1.50 263,150 182,450
T • , r
l.) l .""!_ 11 '- Lo;.1C.! - Tnbocmd

Sti.ll 1-:atcr 'ti.'

\•Jave )" 360,000 43,000 '•23 ,000 112,000
Forw.:rd speed effect, average 0 2,200 0 2,200
Thermal effect, av~rage 11,150 11,450 11,150 11,450
Dyn:.trni c ~ S l a!nmin g 133,000 170,000
t \·n1 i pp ing 136JOOO 168 000
TOTALS 507,150 189,650 602,150 293,650

-··-·--····--··-
:<
··c,)r:~~i.:-t·~:d probab: t.::y.

, 1pabi.Lit',' :.:t yl.dd •,;auld be simply the product of section modulus by the yield
· · ·.' ; of. ;nild steel, 36 kpsi (16 tons/in 2 ):

Tensile yielding in deck (hogging):

Bending moment = 45,800 x 16 = 732,000 ton-ft.

Tensile yielding in bottom (sagging):

Bending moment= 47,200 x 16 = 753,000 ton-ft.

Of course, t~e tension flanges would be able to sustain considerably highet loads
ln conj~nction with extensive plastic yielding.

However, it is the compression lo~ding that usually governs the capability of

a beam to sustain applied bending moments. An estimate has been made of the hull
girder longitudinal bending moment capability of the S,S, Wolverine State, con-
sidering the compressive loading on this transversely framed vessel.~h hog-
ging and sagging conditions have been considered and each will be discussed in
turn,

It '"as ass•1med that the capability of a general cargo vessel's structure

~ill be tested at sea with loading of cargo distributed more or less throughout
its spac~s. Thus, the 'tween deck plating would be constrained to buckle in a
cla~p 0 d-clamped mode by the cargo. Similarly, the lateral hydrostatic loading on
botto!: plating will strongly influence growth of plate deformation under in-plane
loar.l:l.P · Wr~nther (upper.) deck plating, however, would normally not be subject to
~hCS(! COtiGtrnintfi,

BEST AVA~LABLE COPY

 -75-
On the basis of the above, it was found that the plating of the 'tween decks
would buckle at about two thirds of yield stress, while the heavier main deck
p.Llting would sustaln in-plane loading to almost the yield stress. Schultz (129)
found tltat due to plate unfairness the effectiveness of deck plating was about 83%
and tl1c unfai~ness patterns simply grew as loads increased. It was assumed that
tl1o capability could be assessed on this bas:ls, no better information being avail-
able as to plate effectiveness in the unfair state.

The main deck plating has a nominal buckling stress in excess of 34 ksi, and
because of the additional strap installed, it can be expected to sustain nearly
yield stresses prior to buckling.

On the basis of the above, i t was decided that a fair assessment could be made
of moment capability by assuming 83% effective 'tween deck (sagging) or bottom
plating (hogging) uilder compressive post-yield conditions.

It Has assumed that decks and side plating under .:Ension were fully effective,
1-:hile bottom plating under compression was 83% effectiv ·~. On such a basis the
neut:·al axis of the effective structure was computed. .ssuming next that the ef-
fective material was at compressive or tensile yield stress, the resisting moment
of the effective material was computed about the neutral axis of the effective
structure.

By this process it was found that the neutral axis would be at a position
24 feet above the base line and that the internal resisting moment of the section
would be 753,000 ton-feet in hogging.
Sagging ~1oment

The buckling stress of the main deck plating was found to be

k n2 E
a c.---- 31, .5 ksi
c
12(1 - lJ 2 )

where E = 30 x lOG, thickness h = 1.06 in., b = 30 in. and Poisson's ratio lJ • 0.3.
Under sucl1 a condition the second deck would have a stress well below buckling con-
ditions and significant structural deformation would therefore have to take place
before increased load capability could be generated. Hence, it seems reasonable
to assume that all decks are fully effective. The resisting moment for main
deck buckling can then be calculated from M =a Z, where Z is the section modulus.
It seems reasonabla to include continuous longitudinal deck girders in the section
modulus here, ~vcn though it is not customary to do so. Hence, with

J = 34.5 ksi and Z 49,100 in2-ft.,

the resisting moment ior main deck buckling is

oZ • 755,000 ton-feet,

While it is true that additional moment could be developed in the limit as

additional structure is induced until the ultimate load is reached, it appears un-
sound to consider it a possibility in the same sense as for hogging. The reason
for this is that in hogging the double bottom is a prit:~ary system of collapse,
while in sagging all tl·.cee deckti would have to be involved. However, i f the pro-

BEST AVAILABLE COPY

 -76-
cedure used in evaluation cf hogging were used for sagging, considering upper, 2nd
and 3rd decks to be 83% effective and at yield stress of 36 ksi, it is found that
a load of 826 x 103 ton-feet could be sustained. In such a case, the neutral axis
of effective material would be 20.95 feet above base line, with the moment compu-
ted about it.

Hence, in summary the estimated capability of the hull girder in hogging and
sagging can be tabulated below in comparison with the expected maximum values from
Tabl«_ X for the light load condition.

Lifetime .001 Probability

Capability Demand Demand
Max. bending HLX. bending Max. bending
Ton-Ft. Ton-Ft. Ton-Ft.

Hogging 753,000 507,150 602,150

Sagging 755,000 18^,650 295,650

It is concluded that the structural capability of the Wolverine State exceeds

the criterion for ultimate bending moment (demand) by a comfortable margin. This
suggests that if fatigue and brittle fracture considerations were ruled out, some
scantling reduction would be permissible. However, Such a simplification is not
permissible — as proved conclusively by the occurrence of cracks in the upper
deck of the'ship in service. (See section on Fatigue).

No attempt was made to estimate the capability of the structure to resist

short duration dynamic loads, but there can be no doubt that it is considerably
higher than the static figures derived above.

CONVENTIONAL STRENGTH STANDARDS

At the time of the design and construction of the S.S. Wolverine State, there
was no explicit section modulus requirement in the Rules of the American Bureau of
Shipping, although, minimum shell scantlings and strength deck sectional area were
specified. It was customary to check the midship section design against the Load
Line Regulations promulgated by the U.S. Coast Guard, which specified a required
deck section modulus to be met by all vessels subject to Load Line assignment.

In this case the required deck section modulus by Load Line Regulations was
37,536 in2-ft., on the basis of the formula

SM = f d B

where f is a factor having a value o' 16.03 at ship length of 496 ft., d is the
design draft of 32.75 ft. and B is the beam of 71.5 ft. It was customary in the
case of ships with machinery aft for the ABS to add 10% to the Load-line required
value, which gave a figure of 41,290 in2~ft. Actual design values for section
modulus for the Wolverine State were:

Deck 41,297 in2-ft,

Bottom 43,161.

In 1961 a riveted doubler was added to the strength (upper) deck at the Owner's
option, port and starboard. It was 5.0 ft. wide and 1 inch thick, and since its
length of 169 ft. was less than 50% of the ship length it did not officially add to

■■■II'JWIU JJJIH
 -77-
the section modulus. (It was also less than 40%, as required by current Rules).
Recent calculations of section modulus, including the deck straps and other continu-
ous longitudinal members omitted from the originax calculation (but not deck girders),
gave a deck section modulus of 45,800 in2-ft.

A calculation of required section modulus by the current (1972) ABS Rules gives
the following:

Deck 32,100 in2-ft.

Bottom 33,100

It is impossible to relate the above required section modulus values to bend-

ing moment because the implied allowable stress is unknown. However, Arnott in 1939
(131) gave the standard design bending moment M for cargo vessels with machinery
amidships as
LA
M
M"- 35

where L is length and A is load displacement. Takii.e a block coefficient of 0.68

(minimum allowed in the Load Line and ABS Rules), the value of A for Wolverine
State is 22,600 tons, and

M ■ ™* 22 600
' ■ 321,000 ton-ft.

At the time the ship was built this design bending moment included an unspeci-
fied still water bending moment, in addition to the wave bending moment. In fact,
the 10% addition to section modulus for machinery aft was a still water considera-
tion, The corresponding bending moment would be 353,000 ton-ft., which is smaller
than the highest combined value in Table X of 602,500 ton-ft.

Current (1972) ABS Rules are more specific about still water bending moment
requirements (see Chapter III) . The maximum still water moment without penalty
now would be 136,000 ton-ft. for this ship. If one uses the maximum lifetime still
water value of 239,900 given in Table IX,this would mean an increase in the re-
quired section modulus by ABS Rules (19 72) to:

Deck 47,300 in2-ft.

Bottom 48,700

Since, as indicated above, the calculated section modulus for the ship with
deck straps is 45,800 in2-ft., it may be concluded that the ship would not quite
meet present-da^ ABS Rule standards (1972), if the still water bending moment of
239,900 ton-ft. were considered a possible maximum value.

MINIMUM TOTAL COST CALCULATION

A preliminary and very approximate calculation has been carried out to illus-
trate the principles discussed in Chapter VIII for a hypothetical cargo ship, simi-
lar to the Wolverine State designed to present-day standards. As In Chapter VIII,
it is assumed that the structural capability of the hull is deterministic and
therefore that the probability of damage or failure is the probability of exceeding
specific values of bending moment.
 -78-

It was estimated that adding (subtracting) 0.2 in. to deck and bottom" plating
thickness would increase (decrease) the section modulus by 10% and this change
would add (decrease) approximately $100,000 to the initial cost of the ship. An
increase in section modulus of 10% would be equivalent to a reduction of average
stress levels by a factor of 1/1.1 = 0.91, which reference to the long-term bending
moment probability curves shows to correspond to a reduction of failure probability
by a factor of 1/10. Similarly, a reduction in section modulus would increase fail-
ure probability by a factor of 10.

In order to carry out the calculation of total cost, the following assumptions
were made:

Initial cost of ship $15,000,000.

Replacement value at mid-life 3,000,000.
Failure cost F (replacement value +
value of cargo + cost of temporary
replacement) 20,000,000.

Since the structural capability of the Wolverine State was found to far ex-
ceed the demand corresponding to a lifetime probability of failure and loss of
0.001, it will be assumed arbitrarily that after a 10% reduction the section modu-
lus of our hypothetical cargo ship would still provide at least a 0.001 failure
probability, and a 20% reduction would provide 0.01 failure probability.

The total COSL calculations on the basis of the above are summarized m
Table XI.

TABLE XI
CALCULATED TOTAL COST
OF FAILURE
Design I £ pF Total Cost

Basic Ship - 20% $14,800,000 .01 $200,000 $15,000,000

Basic Ship - 10% 14,900,000 .001 20,000 14,920,000
Basic Ship 15,000,000 .0001 2,000 15,002,000
Basic Ship + 10% 15,100,000 .00001 200 15,100,200

This indicates an optimum somewhere near the ship with 10% reduction in sec-
tion modulus. Since these figures are very rough and approximate, a number of
other assumptions regarding costs and probabilities were alsr tried. Naturally
there were changes in the final Total Cost column, but the general picture did not
change significantly. This supports the conclusion in Chapter VIII that an ulti-
mate failure probability of 0.001 is a reasonable tentative figure for a rational
design criterion.

On the basis of some crude assumptions then, it appears that if only ultimate
3trength considerations were involved, some reduction in scantlings would be econ-
omically justified. However, it will be shown subsequently that the. picture changes
vhen consideration is given to structural damage that requires repair but does fiot
immediately threaten the life of the ship — as fatigue cracking, for example.
 -79-
FATI GU£

Cyclic Loading

The long-term probability data discussed and presented in this chapter for the
Wolverine State can be used to provide the load spectra (or patterns) which are
needed to determine design criteria relative to fatigue. Since the project is con-
cerned primarily with loads, and not with structural response, it will suffice to
derive such loading spectra covering:
(a) Wave bending moments.
(b) Dynamic loads.

Results are presented in Fig. 29 for the light load condition, which corres-
ponded to the actual normal operating condition of the ship in service. The wave
stress curves are derived from Fig. 26, using actual section modulus (top), in the
manner described in Chapter VIII. The dynamic stress curves were obtained by first
estimating the histogram of whipping stress cycles for the light load condition (in-
bound) on the basis of 12 voyages per year, 0.16 probability of slamming (per 20-
min. record) and 15 slams per 20-min. period when it occurs. (It was assumed that
no slamming occurs in the hypothetical full-load condition, outbound). As previ-
ously noted, the techniques for predicting this in advance for a new design are
not yet available. The histogram was then integrated to obtain the cumulative
curve shown in Fig. 29.

It should be noted that the variation of mean value for the cyclic loading due
to simple wave bending is given approximately by the still water loading inform-
tion previously presented. The variation in mean value for the superimposed dynamic
loading is given approximately by the data on wave bending moment (Fig. 26).

Service Experience

The Wolverine State was one of a gioup of five C4-S-B5 cargo ships built in
1945. A check of damage and repair records started in 1964 revealed for these
ships one case of hull girder damage related to heavy weather — the S.S. Hoosier
State in January 1971, enroute from Antwerp to Philadelphia in ballast condition.
A typical case of damage is described as follows:

"Stbd. side crack between frame No. 115 and No. 116, starting at inboard
rivet of gunnel bar forward to outboard rivet of deck strap, in a length
of approx. 6." feet. Outboard rivet at crack arrester. Starts at 8 in.
after frame 116 and travels aft. Crossing 117 approx. 7 ft. inboard of
the longitudinal girder and travels between frames 117 and 118, 11 in. aft
of 9 in. pillar and continues hatch insert approx. 9 in. and then inboard
approx. 3 ft. on plate No. UD73. Longitudinal girder broken."

Temporary repairs were effected in the Azores and permanent repairs in New York in
February", 1971.

Study of correspondence and the complete files of the Wolverine State and
Hoosier State showed that the above type of damage had been a problem with all three
of the States Marine C4-S-B5 vessels. In fact, riveted deck straps were installed
by the owners in 1961 for the dual purpose of increasing the section modulus and
providing additional crack arresters. Since these were war-built ships (1945) it is
reasonable to assume that the quality of the steel was below the standards subse-
quently established. Whether there were fatigue cracks or brittle fractures, the
riveted seams appeared to have been successful in limiting fracture propagation.
 -80-

.<0
"T r
SS WOLVERINE STATE

•7 I" T

WAVE BENDING
1 SHIP LIFETIME !2b YEARSI
' ISTHESS CONCENTRATION FACTOR ■ 11

S -0

LOG OF NUMBER OF STRESS REVERSALS

FIGURE 29 - Cyclic Loading "Spectra", S.S. WLVEKINE STATE

In a new ship being designed today and to be built with the present steel qual-
ity requirements, a lower probability of damage would be expected.
Total Cost with Fatigue Cracking

The total cost calculations previously presented will now be extended to in-
clude the effect of fatigue cracking, a type of damage that requires repair but
rarely threatens the safety of the ship.

In addition to the previous cost and probability assumptions, it will be as-

sumed that the lifetime expectation of structural damage, q, is 0.25 for the basic
ship and that increasing (decreasing) the. section medulus by 10% decreases (in-
creases) this expectation by a factor of 1/2 (2). The cost of each damage, S
(repair cost + cost of substitute ship + misc.), is assumed to be $1,000,000.

The new calculations are summarized in Table XII.

TABLE XII

CALCULATED TOTAL COST

OF FAILURE AND DAMAGE

£ 2l (1-p)q_ S Total Cost

Basic Ship - 10% $14,900,000 .001 $20,000 0.5 $250,000 $15,170,000

Basic Ship 15,000,000 .0001 2,000 0.25 125,000 15,127,000
Basic Ship + 10% 15,100,000 .00001 200 0.125 62,500 15,162,500
 -81-

The table indicates that when fatigue damage, as well a5 ultimate failure, is
considered, the optimum design — on the basis of some very crude assumptions — is
tl.e basic ship.

It appears then that for a ship such as the Wolverine State, in which the capa-
bility far exceeds the demand associated with ultimate failure, attention must be
shifted to the nuisance type damage associated with fatigue. The problem is to as-
certain whether or not a reduction in scantlings on the basis of ultimate bending
consideration would really increase the incidence of fatigue cracking to an unecon-
omical level,

BRITTLE FRACTURE

As indicated previously, no attempt has beei made to establish a load criterion

for brittle fracture. However, the loading picture for the Wolverine State should
be the same as that obtained under Ultimate Leads, as summarized in Tables IX and X,
except that we are concerned only with tensile loads. Hence, the compressive
thermal effects need not be included.

Referring to Table X, it may be seen that in the light load condition, when
slamming is likely, the largest benaing moment would be 602,150 - 11,150 =
591,000 ton-ft. in hogging, which corresponds to a tensile load in the deck. Thus
it appears that for this particular ship the principal danger of brittle fracture
would be in the hogging condition as a result of large superimposed whipping follow-
ing a slam. Since it has been shown that the ductile capability of the structure is
above 753,000, the actual occurrence of brittle fracture would depend on stress con-
centrations, steel quality, temperature, and other factors.

X. CONCLUSIONS AND RECOMMENDATIONS

The. following are the principal conclusions developed in the current project:

1. Basic techniques are now available for making rational calcula-

tions in probability terms of most of the loads acting on the main
hull girder of modern merchant ships, including:

Wave-induced loads •
Still water loads •
Thermal effects .

Further development is needed for the calculation of dynamic loads.

2. Input data for the calculation of loads for ships on various ocean
routes is incomplete. In particular, more actual wave records are
needed — even for the North Atlantic routes — from which to obtain
wave spectra for statistical treatment. Actual data on still water
loads, particularly in ballast conditions, are also needed for differ-
ent types of ship.

3. On the basis of the above, a rational load criterion can be set

up for modern merchant ships in relation to ultimate failure by buck-
ling or excessive permanent set, with some reservations in regard to
dynamic loads.
 -82-
4. A trial numerical calculation of ultimate bending loads for the
S.S. Wolverine State shows that a large margin exists between the
"rationally" determined loads and the capability of the hull structure
to resist failure by buckling of one of the flanges. The proposed
load criterion is less severe than current design standards (ABS),
which presumably allow also for fatigue. When similar calculations
have been made for a sufficient number of types of ship and checked
against conventional empirical standards, it should be possible to
adopt a new rational ultimate load criterion for use in the design
of even the most unusual ships.

5. The loads affecting fatigue can be expressed as cyclic loading pat-

terns derived from the above item 1, with separate data on the loads
having different frequency of variation:
(a) Still water loads (shift of base line).
(b) Diurnal thermal effects.
(c) Low-frequency wave bending loads.
(d) High-frequency dynamic loads .

A fatigue loading criterion appears to be of great importance in design

relative to keeping the frequency of occurrence of nuisance cracking at
an acceptable level.

6. A load criterion relative to brittle fracture, including dynamic loads,

is somewhat uncertain at the present time. However, the determination of
the capability of the structure in advance of construction is also uncertain.

It is recommended that further research be carried out on subjects related

to the problems of load criteria for ship de«ign. In particular:

1. Obtain many more systematic wave records for important ocean routes
chat can be spectrum-analyzed and compiled systematically for reference.
Of particular importance are the North Atlantic, North Pacific, and areas
in the vicinity of the Cape of Good Hope.

2. Obtain systematic data from ship operators on still water loadings on

several typical ships over a period of at least six months, from which
actual still water bending moments can be calculated.

3. Check and refine available theories for calculating springing loads and
stresses.

4. Investigate further slamming and whipping relative to midship bending

stresses. Immediate emphasis in this big research area should be:

(a) Obtaining relatively short-term statistical data on the magni-

tudes of midship slamming stresses that are allowed to occur (by
the shipmaster) on ships of different types under different condi-
tions of loading, for guidance in preparing similar new designs.

(b) Obtaining relatively short-term statistical data on the magni-

tudes of the amount by which the above slamming, and whipping, increase
the total combined stress (or bending moment) on ships of different
types under different conditions of loading.

5. Investigate further both technical and economic aspects of fatigue damage.

 -83-

6. Develop further the total cost approach to optimizing design, including

failure and damage costs and making use of actual ship damage data.
7. Continue to investigate possible extreme load conditions arising from
unusual circumstances, such as shallow water effects on wave spectra, unusu-
ally severe local sea conditions (Bay of Biscay), docking loads, wave impacts
on side shell, loads due to shipping water on deck, etc.

8. Extend the work on wave loads beyond the determination of midship bend-
ing moments to include the determination of pressure distributions over the
entire hull surface of a pitching, heaving and rolling ship. Such a detailed
picture of hydrodynamic loads is believed to be essential for the application
of modern finite element techniques of structural analysis.

9 . Encourage parallel research on determining ship structural capability —

and probability of damage and failure — on a probabilistic basis, considering
both quasi-static and dynamic (rapidly applied) loads.

10. Investigate non-linear flare immersion effects on ships with large flare,
including further study of dyramic structural response.

11. Continue research on methods for predicting slam loads and phase relations
to wave bending.

ACKNOWLEDGMENTS

The authors wish to acknowledge the assistance of Teledyne Materials Research

in furnishing additional data beyond that which has been published, of Oceanics,
Inc., in connection with problems arising in the use of their computer program
for wave loads, and of the American Bureau of Shipping for allowing certain re-
sults obtained at Webb under their sponsorship to be included in the report. Use-
ful still water loading data were obtained from Mr. John Ritter of States Marine
Lines and from Mr. John Boylston of Sea Land Services. Temperature data for a
tanker were provided Mr. L. R. Breves of Esso International. Several part-time re-
search assistants and students have been of assistance in various aspects of the
project, including William Garzke, D. Finifter (MIT), Thomas Zielinski (Webb), and
J. van Santen (Delft).
The helpful comments and suggestions on drafts of several chapters are greatly
appreciated from the following individuals:

Prof. G, Aertssen, formerly University of Ghent, Belgium

Prof. J. Caldwell, University of Newcastle, England
John P. Dalzell, Davidson Laboratory
Mr. D. Faulkner, Ship Department, Ministry of Defence, Bath, England
Dr. R.A. Goodman, Lloyd's Register of Shipping, London
Prof. C.J. Jensen, Lloyd's Register of Shipping, London
K. Jouidain, Institut de Recherches de la Construction Navale, Paris
Prof. A. Mansour, Massachusetts Institute of Technology
Ir. J.J.w. Nibbering, University of Ghent, Belgium
Dr. N. Nordenstr^m, formerly Norske Verxtas, Oslo
Dr. M.K. Ochi, Naval Ship Research and Development Center
Dr. Ing. Lorenzo Spinelli, Registro Italian© Navale, Genoa
Mr. K.V. Taylor, Lloyd's Register of Shipping, London

Figs. 1, 2 and 10, from a paper (3) published in the 1971 Transactions of the
Society of Naval Architects and Marine Engineers, is included here by permission
of that Society.
 -84-

REFERENCES

1. Lewis, E. V., and Gerard, G. (ed.), "A Long-Range Research Program in Ship
Structural Design," SS0124, November 1959.

2. Mansour, A., "Probabilistic Design Concepts in Ship Structural Safety and

Reliability," Trans. The Society of Naval Architects and Marine Engineers
(SNAKE), 1972.

3. Little, R. S., Lewis, E. V. and Bailey, F. C., "A Statistical Study of Wave-
Induced Bending Moments on Large Oceangoing Tankers and Bulk Carriers,"
Trans. SNAME, 1971.

4. Walters, I. J. and Bailey, F. C, "Results from Full-Scale Measurements of

Midship Bending «tresses on Three Dry Cargo Ships," SSC-209, 1970.

5. Lewis, E. V., and Dalzell, J. F., "Motion, Bending Moment and Shear Measure-
ments on a Destroyer Model in Waves," Davidson Laboratory Report No. 656,
April 1958.

6. Wiener, N,, "I am a Mathematician,'* Doubleday, 1956.

7. Cragin, J. Q., "Ship Response Instrumentation Aboard the Container Vessel S.S.
Boston: Results from Two Operational Seasons in Nortli Atlantic Service,"
SSC-212, 1970.

b Hoffman, D., van Hooff, R., and Lewis, E. V., "Evaluation of Methods for
Extrapolation of Ship Bending Stress Data," SSC- ',4, 1972.

9. St. Denis, M., and Pierson, W. J., "On the Motions of Snips in Confused Seas,"
Trans. SNAME, 1953.

10. Hoffman, D., Williamson, J., and Lewis, E. V., "Correlation of Model and Full-
Scale Results in Predicting Wave Binding Moment Trends," SSC-233, 1972.

11. Henry, J. R., and Bailey, F. C, "Slamming of Ships: A Critical Review of the
Current State of Knowledge," SSC-208, 1970.

12. Kaplan, P., "Development of Mathematical Models for Describing Ship Structural
Response in Waves," SSC-193, January 1969.

13. Ochi, M. K. and Motter, L. E., "A Method to Estimate Slamming Characteristics
for Ship Design," Marine Technology, Vol. 8, No. 2, April 1971.

14. Hoffman, D., and van Hooff, R., "Feasibility Study of Springing Model Tests of
a Great Lakes Bulk Carrier," D.O.T., U.S. Coast Guard, Technical Report,
July 1972: International Shipbuilding Progress, February 1973.

15. Task Committee on Structural Safety, "Structural Safety — A Literature Review,"

Journal of Structural Division, Proceedings of ASCL, April 1972, pp. 845-863.

16. Chuang, S. L., "The TMB Flexing Stress Monitor," DTMB Report 1560, 1962, and
"Evaluation of the Modified TMB Flexing Stress Monitor," DTMB Report 1954,
1965.
17. Caldwell, J. B., "Ultimate Longitudinal Strength," Trans. RINA, Volume 107,
1965.
 -85-

18. Aertssen, G., "Longitudinal Strength of Ships," International Shipbuilding

Progress, 1970.

19. Abrahamsen, E,, Nordenstrtfm, N., and Rc/ren, E.M.Q., "Design and Reliability of
Ship Structures," Proceedings of Spring Meeting, SNAME, April, 1971.

20. Report of Committee 10 on Design Procedure, International Ship Structures

Congress, Tokyo, 1970.

21. Spinelli, L., "A Few Notes on R. I. Ma. Rules 1971 Considered in the Light of
the Present Knowledge on Ship Structural Design," R. I. Na. Technical Bulle-
tin No. 46, July 1971.

22. Nibbering, J. J. W., "Permissible Stresses and Their Limitations," Report No.
141, Delft Ship Structure Laboratory, April, 1969.

23. Vasta, John, "Lesson Learned from Full-Scale Structural Tests," Trans. SNAME,
1958.

24. Vasta, J., Palermo, P. M., "An Engineering Approach to Low-Cycle Fatigue of
Ship Structures," Trans. SNAME, 1965.

25. de Wilde, G., "Structural Problems in Ships with Large Hatch Openings,"
International Shipbuilding Progress, Vol. 14, January - February, 1967.

26. Kline, Roger, "Some Aspects of Ship Stiffness," Trans. SNAME, 1967.

27. Altenburg, C. J., and Scott, R. J., Gibbs and Cox, "Design Considerations for
Aluminum Huil Structures - Study of Aluminum Bulk Carrier," SSC-218, 2.971.

28. Panel HS-4, "Higher-Strength Steels in Hull Structures," SNAME T & R Bulletin
No. 2-19, December 1971.

29. Murray, J. M., "Longitudinal Bending Moments," Trans. I.E.S.S., Vol. 90, 1946-
1947.

30. Mardell.i, A., "Simplified Calculation of Longitudinal Bending Moments," Trans.

I.E.S.S., Vol. 99, 1955-56.

31. Abrahamsen, E., "The Strength of Dry Cargo Ships," Publication No. 7, Det norske
Veritas, 1958.

32. Faresi, R., "A Simplified Approximate Method to Calculate Shear Force and Bend-
ing Moment in Still Water at Any Point of the Ship's Length," American Bureau
of Shipping, 1965.

33. Panel 0-34, "Indexed Abstracts of Marine Computer Programs," Technical and Re-
search Bulletin No. 4-13. SNAME. 1970.
34. Hoffman, D., and Zielinski, T. E., "The Use of Conformal Mapping in Ship Sys-
tems Design," Paper presented at NA7SEC Symposium, February, 1972.

35. Buchanan, G., Dobson, R. J. C, and Jensen, C. J, G., "Lloyd's Register of

Shipping's Approach to the Control of Brittle Fracture in Ship Structures,"
L. R. Report No. 56, 1969.

36. Vossers, G., Swaan, W. A., and Rijken, H., "Vertical and Lateral Bending
Moment Measurements on Series 60 Models," International Shipbuilding Progress,
Vol. 8, No. 83, July 1961. ------
 -86-

37. Lewis, E. V., and Dalzell, J. F., "Motion, Bending Moment and Shear Measure-
ments on a Destroyer Model in Waves," Davidson Laboratory, Report No. 656,
April 1958.

38. Kim, W. D., "On the Harmonic Oscillation of a Rigid Body on a Free Surface,"
Journal of Fluid Mechanics 21, Part 3, March 1965, pp. 427-451.

39. Vugts, J. H., "The Hydrodynamic Forces and Motions in Waves," Delft Thesis,
1970.

40. Korvin-Kroukovsky, B. V., & Jacobs, W. R., "Pitching and Heaving Motions of
a Ship in Regular Waves," Trans. SNAME, Vol. 65, 1957.

41. Gerritsma, J,, and Beukelman, W., "Analysis of the Modified Ship Theory for
the Calculations of Ship Motions and Wave Bending Moments," TNO Report 968,
Delft, June 1967.

42. Grim, 0., "A Method of More Precise Computations of Heaving and Pitching Mo-
tions both in Smooth Water and in Waves," Third Symposium of Naval Hydro-
dynamics, Scheveningen, 1960, 0NR/ACR-65, pp. 483-524.

43. Salvesen, N., Tuck, E. 0., and Faltinsen, 0., "Ship Motions and Sea Load-i,"
Trans. SNAME 1970.

44. Timmann, R., and Newman, J. N., "The Coupled Damping Coefficient of a Sym-
metric Ship," Journal of Ship Research, No. 4, March 1962.

45. Soeding, H., "Eine Modifikation der Streifen Methode," Schiffstechnik, Bd. 16.
Heft 80, 1969.

46. Tasai, F., and Takaki, M., "Theory and Calculations of Ship Response in
Regular Waves," Symposium on Seaworthiness of Ships, Japan Society of Naval
Architecture, 1969.

47. Hoffman, D., "Distribution of Wave-Caused Hydrodynamic Pressures and Forces

on a Ship Hull," Norwegian Ship Model Tank Publication 94, October 1970.

48. Perez y Perez, L., "A General Calculation of the Structural Load on a Ship
in a Seaway," Marine Technology, Vol. 8, No. 2, April 19 71.

4SA, Maniar, V. M., "Investigation of Bending Moments within the Midship Half
Length of a Mariner Model in Extreme Waves," SSC-163. June 1964.
49. Raff, A. I., "Program SCORES - Ship Structural Response in Waves," SSC-230,
1972.

50. Beck, R., "A Computerized Procedure for Prediction of Seakeeping Performance,"
Report No. 69-2, Department of Naval Architecture and Marine Engineering,
Massachusetts Institute of Technology, March 1969.

51. Loukakis, T., "Computer Aided Prediction of Seakeeping Performance in Ship

Design," Report No. 70-3, Department of Naval Architecture and Marine Engin-
eering, Massachusetts Institute of Technology, August 1970.

52. Lewis, F. M., "The Inertia of the Water Surrounding a Vibrating Ship," SNAME
Trans., 1929.

53.. 'Kaplan, P. and Raff, A. L., "Evaluation and Verification of Computer Calcula-
tions of Wave-Induced Ship Structural Loads," SSC Report SSC-229, 1972.
 -87-

54. Chiocco, M. J. and Numata, E., "Midship Wave Bending Moments in a Model of
the Cargo Ship Wolverine State Running at Oblique Headings in Regular Waves,"
SSC-201, September 1969." (AD 695123).

55. Wahab, R., "Amidships Forces and Moments on a CR = 0.80 Series 60 Model in
Waves from Various Directions," Netherlands Ship Research Centre TNO, Report
No. 100S, November 1967.

56, Numata, E., "Longitudinal Bending and Torsional Moments Acting on a Ship
Model at Oblique Headings to Waves," J. Ship Res., Vol. 4, No. 1, 1960.

57. Hoffman, Dan, "A Comparison of Theoretical and Experimental Bending Moment
Response in Regular Waves," presented at the 16th Meeting of the American
Towing Tank Conference, Sao Paulo, Brazil, August 1971.

58. Chiocco, M. J., Numata, E., "Vertical Wave Bending Moments in a Model of the
Tanker S.S. Universe Ireland Running at Oblique Headings in Regular Waves,"
Letter Report 1528, Davidson Laboratory, May 1971.

59. Lewis, E. V., and Numata, E,, "Ship Motions in Obliqi-e Seas," Transactions
SNAME, 1960.

60. Joosen, WPA, Wahab, R., and Woortman, J. J., "Vertical Motions and Bending
Moments in Regular Waves," International Shipbuilding Progress, Vol. 15,
January 1968.

61. Faltinsen, 0., "Comparison Between Theory and Experiments of Wave Induced
Loads for Series 60 Hull with CR = 0.80, "Det norske Veritas, Research
Dept., Report No. 70-27-S, July 1970.

62. Grim, I., and Schenzle, P., Proceedings of International Marine and Shipping
Conference, Institute of Marine Engineers, Section 3 London, 1969.

63. St. Denis, M., and Pierson, W. J., "On the Motions of Ships in Confused Seas,"
Trans. SHAME, 1953.
64. Lewis, E.V., Chapter IX, "The Motions of Ships in Waves, "Principles of
Naval Architecture, SNAME, 1967.

65. Hoffman, D., and Lewis, E. V., "Analysis and Interpretation of Full-Scale
Data on Midship Bending Stresses of Dry Cargo Ship," SSC-196, June 1969.
66. Pierson, W. J., Moskowitz, L., and Mehr, E., "Wave Spectra Estimated from
Wave Records Obtained by the OWS Weather Explorer and 0WS Weather Reporter,"
Technical Reports Nos. I, II, III, New York University, 1962-63."

67. Draper, L., and Squire, E. M., "Waves at Ocean Weather Station India," Trans.
RINA, Vol. 109, 1966.

68. Miles, M., "Wave Spectra Estu ted from Stratified Sample of 323 North At-
lantic Wave Records," National ^search Council, Report LTR-SH-118, October
1971.

69. Scott, J. R., "Some Average Sea Spectra," Trans. RINA, 1968.

70. Fukuda, J., "On the Family of Wave Spectra for Long Term Predictions of Ship
Bending Moments," Journal of Society of Naval Architects of Japan, 1966.
71. Yamanouchi, Y., Unoki, S., Kanda, T., "On the Winds and Waves on the Northern
North Pacific Ocean and South Adjacent Seas of Japan as the Environmental
Conditions of the Ship," Ship Research Institute, Tokyo, J 965.

72. Mathews, S. T., "Main Hull Girder Leads on a Great Lakes Bulk Carrier," SNAME
Spring Meeting, 1967.

73. Hoffman, D., "Analysis of Ocean Wave Spectra at Station India," Prepared for
Panel H-7 of SNAME, December 1971 and March 1972.

74. Canham, H. J., Cartwright, D. E., Goodrich, G, J. and Hogben, N., "Seakeeping
Trials en O..W.S. Weather Reoorter," T-ans. RINA, Vol. 104, pp. 447-492, 1962.

75. Darbyshire, J., "The Spectra of Coastal Waves," Deutsche Hydrographische Zeit,
No. 4, 1959, and as presented by ccott, J. R., "Some Average Sea Spectra,"
Trans. RINA, i968.

76. Darbyshire, J., "The One-di.-.onsional Wave Spectrum in the Atlantic Ocean and
in Coastal Waters," in Ocean Wave Spectra. Prentice Hall, Inc., .1963.
77. Hogben, N., and Lumb, F. E,, "Ocean Wave Statistics," H. M. Stationery Office,
London, 1967.

78. Waiden, H., "Die Eigo.nschaften der Meereswellen in Nordatlantischen Ozean,"

Deutscher Wetterdienst Seewetteramt Publication No. 41, 1964.
79. Proceedings of International Ship Structures Congress, Delft, 1964: Re-
port of Committee Mo. I, Environmental Conditions (Wave data reproduced
in Principles of Naval Architecture, SNAME, 1967).

80. Roll, H. U., "height, Length and Steepness of Sea Waves in North Atlantic..,"
SNAME r & R Bulletin 1-19, December, 1958.

81. Hogben, N., Lumb, F. E. and Cartwright, D. K., "The Presentation of Wave
Data from Voluntary Observing Ships," NPL Ship Report: 49, 1964.

82. Crutcher, H. L., and Davis, 0. M., "Marine Climatic Attas of the World,"
U.S. Navy, NAVAIR 50-1C-54, March 1, 1969.

83. Yamanouchi, Y., Unoki, S., Kanda, T., "On the Winds and Waves on the Northern
North Pacific Ocean and South Adjacent Seas of Japan as the Environmental.
Condition for the Snip," Ship Research Institute, Tokyo, 1965.

84. Ewing, J. A., and Hogben, N., "Some Wave and Wird Data from Trawlers,"'
Marine Observer, 1966.

85. Pore, N. Arthur, "Wave Climate from Shipboard Observations," Technical

Symposium on Hull Stresses in Bulk Carriers in the Great Lakes," SNAME,
Ottawa, July 19 71 .

86. Darbyshire, Mollie and Pritchard, Eileen, "Sea Waves near the Coasts of
South Africa," Deutsche Hydrographische Zeitschrift, Volume 19, No. 5, 19b6.

87. Hogben, N., Lumb, F. E., and Cartwright, D. E., "The Presentation of Wave
Data from Voluntary Observing Ships," National Physical Lab., Fcltham,
England, Ship Report No. 49, 1964.

88. Report of Committee 1, Environmental Conditions, Proceedings of 2nd Inter-

national Ship Structures Congress, Delft, 1964.
 -89-
89. Hawkins, S., Levine, G. H., and Taggart, R., "A Limited Survey of Ship Struc-
tural Damage," SSC-220, 1971.

90. Report of Committee 2, Wave Loads - Hydrodynamics, Proceedings of 4th ISSG,

Tokyo, 1970.

91. Hoffman, D., and Maclean, W. M., "Ship Model Study of Incidence of Shipping
Water Forward," Marine Technology, Vol. 7, No. 2, April 1970, pp. 149-158.

92. Kawakami, M., "On the Impact Strength of Ships due to Shipping Green Seas -
Towing Experiments of a Ship Model in Regular Waves," Selected Papers from
the Journal of the Society of Naval Architects of Japan, Vol. 8, 1971.
(In English)

93. Schwendler, R. G., "Improved Analog Computer Analysis of the Bow Slamming
Problem," Computer Engineering Associates, Report on U.S. Navy Project
ESD 178, April 1962. (See Jasper & Church, SNAME 1963).

94. Jasper, N. H., and Church, J. W., "Structural Seaworthiness Studies," Tuns,
SNAME, Vol. 71, (1963), pp. 31-80.

95. Kaplan, P., Sargent, T. P., and Raff, A. I., "An Investigation of the
Utility of Simulation to Predict Ship Structural Response in Waves," SSC-
197, June 1969.

96. Kaplan, P., and Sargent, T. P., "Further Studies of Computer Simulation of
Slamming and other Wave-Indue^ Vibratory Structural Loadings on Ships in
Waves," SSC-231, 1972.

97. Tasaki, R., "On Shipment of Water in Head Waves," Tenth ITTC, London, 1963.

9§. Tasai, F., "Wave Height at the Side of a Two-dimensional Body Oscillating
on the Surface of a Fluid," Report of the Research Institute for Applied
Mechanics, Kyushu University, Vol. 9, No. 35, 1961.

99. Ochi, M. K., "Ship Slamming, Hydrodynamic Impact Between Waves and Ship
Bottom Forward," Proceedings Symposium on Fluid-Solid Interaction, ASME, 1967,

100. Ochi, M. K., and Motter, L. E., "A Method to Estimate Slamming Character-
istics for Ship Design," Marine Technology, Vol. 8, No. 2, April 1971,
pp. 219-232.

101. Kline, R. G. and Clough, R. W., "The Dynamic Response of Ship's Hulls as
Influenced by Proportions, Arrangements, Loading and Structural Stiffness,"
Proceedings SNAME Spring Meeting, Mont-eal, Canada, 1967, pp. 4-1 to 4-18.

102. St. Denis, M. and Fersht , S., "On the Effect of Ship Stiffness Upon the
Structural Response of a Cargo Ship to an Impulsive Load," SSC-186, Sep-
tember 1968.

103. Maclean, W. M., and Lewis, E. V., "Analysis of Slamming Stresses on S.S.
Wolverine State," Webb Institute of Naval Architecture, Report No. 10-17,
to the American Bureau of Shipping, 26 January 1970; Marine Technology,
January 1973.

104. Henry, J. R. and Bailey, F. C, "Slamming of Ships: A Critical Review of

the Current State of Knowledge," SSC-208, 1970.
 -90-

105. Greenspon, J. E., Jasper, N. H., and Birmingham, J. T., "Sea Teste of the
USCGC Unimak, Pressures, Strains and Deflections of the Bottom Plating Inci-
dent to Slamming," International Shipbuilding Progress, Vol. 3, No. 25,
September 1956, pp. 474-486.

106. Warnsinck, W. H., and St. Denis, M., "Destroyer Seakeeping Trials", Proceed-
ings of Symposium on the Behavior of Ships in a Seaway, NSMB, 1957, pp. 439-
467 and 971-980.

107. Bledsoe, Margaret D., Bussemaker, Otto, and Cummins, W. E. (1960) "Sea-
keeping Trials on Three Dutch Destroyers," SNAME, Vol. 68, 1960.

L08. Admiralty Shipbuilding Committee Report No. 8, 1953.

199. Wheaton, J. W., Rano, C. H,, Diamant, P. T. and Bailey, F. C, "Analysis of

Slamming Data from the S.S. Wolverine State," SSC-210, 1970.

MO. "ABS Program for the Collection of Long-Term Stress Data from Large Tankers
and Bulk Carriers," final report, Teledyne Materials Research, 1971.

111. Lewison, G. R. G., "The Control of Ship Slamming," reprinted from Snipping
World and Shipbuilder, June 1969.

112. Aertssen, G., "Service-Performance and Seakeeping Trials on M. V. Jordaens,"

RINA Trans. 1966.

113. Meek, M., Adams, R., Chapman, J. C, Reibel, H., and Wieske, P., "The
Structural Design of the O.C.L. Container Ships," Trans. RIM, 1971.

114. Symposium on Hull Stress in bulk Carriers in the Great Lakes, SNAME, Ottawa,
19/1.

115. Goodman, R. A., "Wave-Excited Main Hull Vibration in Larg^ Tankers and Bulk
Carriers," Trans. RINA, 1970.

116. Hoffman, D. and van Hooff, R., "Feasibility Study of Springing Model Tests of
a Great Lakes Bulk Carrier," D.O.T., USCG, Technical Report, July 1972.

117. Lewis, E.V., and Wheaton, J. W., "Study of Combined Wave and Springing
Stresses of the Edward L. Ryerson," Symposium on Hull Stresses in Bulk Car-
riers in the Great Lakes, SNAME, Ottawa, July 1971.

118. Wheaton, J. W., "Second Quarterly Progress Report 1434 (b), "Slamming Data
Analysis," Teledyne Materials Research, 22 February, 1972.

119. Jasper, N., "Temperature-Induced Stresses in Beams and Ships," ASNE Journal,
August 1956, p. 485.

119A. Lyiuan, P. T., and Meriam, J. L., "Temperature Distribution and Thermal
Stresses," SSC-152, June 1964.

119B. Hechtraan, R. A., "Thermal Stresses in Ships," SSC-95, October 1956.

120. Corastock, J. P., "Principles of Naval Architecture," SNAME, 1967.

121. Zubaly, Robert B., Lewis, Edward V., "Ship Bending Moments in Irregular Seas
Predicted from Model Tests'', Webb Institute Report to ABS, December 1963.
 -91-

122. Nordenstrtfm, N., "Probability of Failure for Weibull Load and Normal Strength,"
Norske Veritas Report No. 69-28-S, March 1970.

:.23. Freudenthal, A, M., "The Safety of Structures," Trans. ASCE, Vol. 112, 1947,
pp. 125-180.

124. Hoff, N.J., "Philosophy of Safety in the Supersonic Age," Report No. 87, NATC,
Sixth General Assembly of A.G.A.R.D., August 1956.

125. Dunn, T. W., "Reliability in Shipbuilding," Trans. SNAME, 1964.

126. Beer, W. J., "Analysis of World Merchant Ship Losses," Trans. RINA, 1969.

127. Freudenthal, A. M., and Gaither, W. S., "Probabilistic Approach to Economic

Design of Mr.r .ime Structures," XXIInd International Navigation Congress,
Section II, SuDJect 5, Paris 1969.

128. Nibbering, J. J. W., "Fatigue of Ship Structures," International Shipbuilding

Progress, September 1963. *

129. Schultz, H. G., "Buckling and Post Buckling Behavior of a Transversely Stif-
fened Ship Hull Model," Journal of Ship Research, Vol. 8, No. 2, September
1964, pp. 7-21. "

130. Arnott, D,, "Strength of Ships," Chapter VI of Principles of Naval Architec-

ture, edited by Rossell and Chapman, SNAME 1939,
 -92-
AtTENDIX A

BIBLIOGRAPHY

ON

HULL LOAD CRITERIA

STILL WATER LOADS

Abrahamser,, E. , "The Strength of Dry Cargo Ships," SVt :..r.V Yeritas, Pub. No. 7,
1958.

Aertssen, G. , "Longitudinal Strength of Ships," International Shipbuilding Prog-

ress," 1970, pp. 269-280.

Bascom, C. W., "Structural Design Review of Long, Cylindrical, Liquid-Filled Inde-

pendent Cargo Tank-Barges," Final Report on Project SR-184 "Chemical Tank-Barge
Design," SSC-205, 1970.

Faresi, Renato, "Simplified Appr. .imate Method to Calculate Shear and Bending Mom-
ents in Still Water," ABS Report, April, 1965.

Hawkins, S., Levine, G. H. and Taggart, R., "A Limited Survey of Ship Structural
Damage," Report by R. Taggart, Inc. for Ship Structure Committee, November, 1970.

Hoffman, D., and Zielinski, T, E., "The Use of Conformal Mapping in Ship Systems
Design," Paper presented at NAVSEC Symposium, February, 1972.

"Indexed Abstracts of Marine Computer Programs," SNAME T & R Bulletin, Nos. 4-13,
August, 1970.

Lewis, E. V., and Dalzell, J. F., "Motion, Bending Moment and Shear Measurements
on a Destroyer Model in Waves," Davidson Laboratory, Report No. 656, April, 1958.

Mandelli, A., "Simplified Calculation of Longitudinal Bending Moments," Trans-

actions, IESS, Vol. 99, 1956.

McDonald and MacNaught, "Investigation of Cargo Distribution in Tank Vessels,"

SNAME Transactions, 1949.

Murray, J. M. , "Longitudinal Bending Moments," Trans. IESS, Vol. 90, 1947.

Vossers, G., Swaan, W. A., and Rijken, H., "Vertical and Lateral Bending Moment
Measurements on Series 60 Models," International Shipbuilding Progress, Vol. 8,
No. 83, July, 1961.

WAVES

Canham, H. J., Cartwright, D. E., Goodrich, G. J. and Hogben, N., "Seakeeping

Trials on O.W.S. Weather Reporter," Trans. RINA, Vol. 104, pp. 447-492, 1962.

Darbyshire, J., "The Spectra of Coastal Waves," Deutsche Hydrographische Zeit,

No. 4, 1.959, and as presented by Scott, J. R. , "Some Average Sea Spectra," Trans.
RINA, 1968.
 -93-

Datbyshire, J., "The One-dimensional Wave Spectrum in the Atlantic Ocean and in
Coastal Waters," in Ocean Wave Spectra, Prentice Hall, Inc., 1963.

Darbyshire, Mollie and Pritchard, Eileen, "Sea Waves near the Coasts of South
Africa," Deutsche Hydrographische Zeitschriit, Volume 19, No. 5, 1966.

Draper, L., and Squire, E. M., "Waves at Ocean Weather Station India," Trans.
RINA, Vol. 109, 1966.

Ewing, J. A., and Hogben, N., "Some Wave and Wind Data from Trawlers," Marine
Observer, 1966.

Fukuda, J. , "On the Family of Wave Spectra for Long Term Predictions of Ship Bend-
ing Moments," Journal of Society of Naval Architects of Japan, 1966.

Hoffman, D., "Ar sis of Ocean Wave Spectra it Station India," prepared for
Pane] iJ-7 of SNAML, December, 1971 and March, 1972.

Hogben, N., and Lumb, F. E., "Ocean Wave Statistics," H. M. Stationery Office,
London, 1967.

Hogb.n, N., Lumb, F. E., and Cartwright, D. E., "The Presentation of Wave Data
from Voluntary Observing Ships," Rational Physical Lab., Feltham, England, Ship
Report No. 49, 1964.

"Marine Climatic Atlas of the World," U.S. Navy, NAVAIR 50-1C-54, March 1, 1969.

Mathews, S. T.,MMain Hull C-irder Loads on a Gre?t Lakes Bulk Carrier," SNAME
Spring Meeting, 1967.

Miles, M. , "Wave Spectra Estimated from Stratified Sample of 323 North Atlantic
Wave Records," National Research Council, Report LTR-SH-118, October, 1971.

Moskovitz, L., Pierson, W. J. and Mehr, E., "Wave Spectra Estimated from Wave
Records Obtained by the OVS Weataer Explorer," N.Y.U. Research Division Reports,
November, 1962 and March, 1963.

Pierson, W. J., Moskowitz, L., and Mehr, E., "Wave Spectra Estimated from Wave
Records Obtained by the OWS Weather Explorer and OWS Weather Reporter," Techni-
cal Reports Nos. I, II, III, New York University, 1962-63.

Po^e, N. Arthur, "Wave Climate from Shipboard Observations," Technical Symposium

on Hull Stresses in Hulk Carriers in the Great Lakes, SNAME, Ottawa, July, 1971.

Proceedings of International Ship Structures Congress, Delft, 196 4: Report of

Committee No. ], Environmental Conditions. (Wave data reproduced in Principles
°L NiLYilL Architecture) , SNAME , 196 7.

Report of Committee 1, Environmental Conditions, Proceedings of 2nd International

Ship Structures Congress, Delft, 1964.

Roll, H. U., "Height, Length and Steepness cf Sea Waves in North Atlantic ....,"
SNAME Ti R Bulletin 1-19, December, 1958.

Scott, J. R., "Some Average Sea Spectra," Trans. RINA, 1968.

Waiden, H., "Die Eigenschaften der Meereswellen im Nordatlantischen Ozean,"

Deutscher Wetterdienst Seewetteramt Publication No. 41, 1964.
 -94-
Yaraanouchi, Y., Unoki, S., Kanda, T., "On the Winds and Waves on the Northern
North Pacific Ocean and South Adjacent Seas of Japan as the Environmental Condi-
tion for the Ship," Ship Research Institute, Tokyo, 1965.

WAVE LOADS

Abrahamsen, E., "Recent Developments in the Practical Philosophy of Ship Struc-

tural Design," Spring Meeting, SNAME, July, 1967.

Aertssen, G. , "Longitudinal Strength of Ships," International Shipbuilding Prog-

ress, 1970, pp. 269-280. " """""

Aertssen, G. , "Service-Performance and Seakeeping Trials of M. V. Jordaens,"

Trans. RINA, 1966.

Akita, Y., Macda, T. and Furuta, K., "On the Motion of Cargo Oil ir Long Tanks,"
Journal of the Society of Naval Architects of Japan, Vol. 12J, 1968.

Band, Edward G. U., "Analysis of Ship Data to Predict Long-Term Trends of Hull
Bending Moments," ABS Report, November, 1966.

Beck, R., "A Computerized Procedure for Prediction of Seakeeping Performance,"

Department of Naval Architecture and Marine Engineering, Massachusetts Institute
of Technology, Report No. 69-2, March, 1969.

Birmingham, John T., "Longitudinal Bending Moment Predictions Derived from Re-
sults of Seven Ship Trials,' Naval Ship Research & Development Center, Report
No. 3718, September, 1971.

Chiocco, M. J. and Numata, E., "Vertical Wave Bending Moments in a Model of the
Tanker JS Universe Ireland Running at Oblique Headings in Regular Waves,"
Davidson Laboratory, Letter Report No. 1528, May, 1971.

Chiocco, M. J. and Numata, E., "Midship Wave Bending Moments in i rlodel of the
Cargo Ship Wolverine State Running at Oblique Headings in Regular Waves," SSC-201,
September, 1969 (AD 695123).

Cragin, J. Q., "Ship Response Instrumentation aboard the Container Vessel SS

Boston ; Results from Two Operational Seasons in North Atlantic Service, SSC-214,
1970.

DeJong, B., "Berekening Vc.n de Hydrodynamische Coefficienten Van Oscillerende

Cylinders," Report No. 174, Delft Shipbuilding Laboratory, 1967.

Ewing, J A. and Goodrich, G. J., "The Influence on Ship Motions of Different

Wave Spectra and of Ship Length," Trans. RINA, 1967.

Fa?'n, R. A., Cragin, J. Q. and Schofield, B, H. , "Ship Response Results from the
First Operational Season aboard the Container Vessel SS Boston, SSC-212, 1970.

Faltinsen, 0., "Comparison Between Theory and Experiments of Wave Induced Loads
for Series 60 Hull with C = 0.80, "Dec Norske Veritas, Research Dept., Report No.
B
70-27-S, July, 1970.

Frank, W., "Oscillation of Cylinders in or below the Free Surface of Deep Fluids,"
Report No. 237S, 1967.

. .- . .- - ■■--->■■ t, ., .1,. —...A.-k.M.T.M. .

~—
 -95-
Fukuda, J., "Theoretical Determination of Design Wave Bending Moments," Japan
Shipbuilding & Marine Engineering, Vol. 2, No. 3, 1967.

Fukuda, J., "Computer Program Results for Response Operators of Wave Bending Mom-
ent in Regular Oblique Waves," The Memoirs of the Faculty of Engineering, Kyushu
University, Vol. 26, No. 2, 1966.

Fukuda, J., "Computer Program Results for Ship Behavior in Regular Oblique Waves,"
Proceedings 11th ITTC, 1966.

Fukuda, J., "On the Midship Bending Moment of a Ship in Regular Waves," Journal
of the Society of Naval Architects of Japan, Vol. Ill, 1962.

Fukuda, J., "On the Midship Bending Moments of a Ship in Regular Waves," Journal
of the Society of Naval Architects of Japan, Vol. 110, 1961.

Fukuda, j. and Shibata, J., "The Effects of Ship Length, Speed and Course on Mid-
ship Bending Moment, Slamming and Bow Submergence in Rough Seas," The Memoirs of
the Faculty of Engineering, Kyushu University, Vol. 25, No. 2, 1966.

Fukuda, J., Shibata, J. and Toyota, H., "Midship Bending Moments Acting on a
Destroyer in Irregular Seas," Journal of the Society of Naval Architects of Japan,
Vol. 114, 1963. ' '

Geiritsma, J. and Beukelman, W., "Analysis of the Modified Theory for the Calcula-
tion of Ship Motions and Wave Bending Moments," International Shipbuilding Prog-
ress, 1967.

Gerritsma, J., and Beukelii.an, W. , "Analysis of the Modified Ship Theory for the
Calculations of Ship Motions and Wave Bending Moments," TNO Report 968, June, 1967.

Goda, K., "Hydrodynamic Pressure on a Midship in Waves," Journal of the Society oJF
Naval Architects of West Japan, No. 35, 1968.

Goda, K., "Transverse Wave Loads on a Ship in Waves," Journal of the Society of
Naval Architects of West Japan, Vol. 121, 1967 and Vol, 123, 1968.

Goda, F., "Transverse Wave Loads on a Ship in Waves," A note prepared for the
Rome Meeting of Committee 2, ISSC, 1968.

Goda, K., "Transverse Wave Loads on a Ship in Waves," Contribution to 3rd ISSC,
Oslo, 1967.

Grim, 0., and Schenzlc, P., "The Prediction of Torsional Moment, Horizontal Bending
Moment and Horizontal Shear Force on a Ship in Waves," Proceedings of International
Marine and Shipping Conference, Institute of Marine Engineers, Section 3, London,
1969.

Gumbel, E. J., "Statistics of Extremes," Columbia University Press, New York, 1958.

Hoffman, 1)., van Hooff, R. and Lewis, E. V., "Evaluation of Methods, for Extrapola-
tion of Ship Bending Stress Data," SSC-2 34, 1972.
Hoffman, D., "A Comparison of Theoretical and Experimental Bending Moment Response
in Regular Waves," presented at the 16th Meeting of the American Towing Tank Con-
ference, Sao Paulo, Brazil, August, 1971.

Hoffman, D., "Wave Induced Bending Moments on Great Lakes Ore Carriers," SNAME
Symposium, Ottawa, July, 1971.
 -96-
Hoffman, D. , "Distribution of Wave-Caused Hydrodynamic Pressures and Forces on a
Ship Hull," Norwegian Ship Model Tank Pub. No. 94, October, 1970.

Hoffman, D., "Lecture Notes on Conformal Mapping Techniques in Ship Hydrodynamics,"

Draft, 1969.

Hoffman, D., "Distribution of Wave Caused Hydrodynamic Pressures and Forces on a

Ship Hull in Waves," Norwegian Ship Model Tank Publication No. 94v October, 1966.

Hoffman, D. and Lewis, E. V., "Analysis and Interpretation of Full-Scale Data on

Midship Bending Stresses of Dry Cargo Ships," SSC-196, June, 1969.

Hoffman, D., Williamson, J. and Lewis> E. V., "Correlation of Model and Full-Scale
Results in Predicting Wave Bending Moment Trends," SSC-233, 1972.

Huang, II. W., "Measurement of Pressures and Hydrodynamic Forces on a Ship-Like

Model Oscillating in a Free Surface," I.E.R. Report, 1965.

Joosen, WPA, Wahab, R. , and Woortman, J. J., "Vertical Motions and Bending Moments
in Regular Waves," International Shipbuilding Progress, Vol. 15, January, 1968.

Kaplan, P., "Computer Simulation of Ship Structural Response," Note prepared for
Committee 2, ISSC, 1969.

Kaplan, P., Sargent, T. P. and Raff, A. I., "An Investigation of the Utility of
Computer Simulation, to Predict Ship Structural Response in Waves," SSC-197, June,
1969.

Kaplan, P., "Development of Mathematical Models for Describing Ship Structural

Response in Waves," SSC-193, January, 1969.

Kaplan, P. and Raff, A. I., "Evaluation and Verification of Computer Calculations

of Wave-Induced Ship Structural Loads," SSC-229, 1972.

Kim, W. D., "On the Harmonic Oscillation of a Rigid Body on a Free Surface,"
Journal of Fluid Me_cJian_ics 21, Part 3, March 1965, pp. 427-451.

Konuma, M., "Comparisons Between Calculations and Experiment for Vertical Motions
and Bending Moments in Regular Oblique Waves," Technical Report of Nagasaki Ship-
yard. Mitsubishi Heavy Industries, Ltd., 196b.

Korvin-Kroukovsky, B. V. and Jacobs, W. R., "Pitching and Heaving Motions of a Ship

in Regular Waves," Trans. SNAME, Yol. 65, 1957.

Lewis, E. V., "Predicting Long-Term Distribution of Wave-Induced Bending Moment on

Ship Hulls," SNAME, Spring, 1967.
Lewis, E. V., and Numata, E., "Ship Motions in Oblique Seas," Transactions SNAME,
1960.

Lewis, F. M. , "The Inertia of the. Water Surrounding a Vibrating Ship,'"' SNAME

Trans., 1929.

Little, R. S. and lewis, E. V , "A Statistical Study of Wave-Induced Bending Mom-

ents on Large Oceangoing Tank?rs and Bulk Carriers," Trans. SNAME, 1971.

LULzeit, M. and Haslum, K., 'V,cmparison Between Calculated and Measured Wave Bend-
ing Moments, Shearing forces a:.d Pitching Motions for a T2 Tanker Model in Regular
Waves," Det Norske Veritas Research Department, Report No. 64-34-S, 1964.
 -97-
Loukakis, T. , "Computer Aided Prediction of Seakeeping Performance in Ship Design,"
Department of Naval Architecture and Marine Engineering, Massachusetts Institute
of Technology, August, 1970.

Loukakis, T., "Experimental and Theoretical Determination of Wave-form and Ship

Responses Extremes," Massachusetts Institute of Technology, Department of Naval
Architecture and Marine Engineering, May, 1970.

Lundgren, J. and Hoffman, D. , "Analysis of Extreme Value Data txo Predict Long-Term
Ship Stress Probability," Proceedings of 3rd ISSC, Vol. 3, 1967.

Maniar, Naresh M., Numata, Edward, "Bending Moment Distribution in a Mariner Cargo
Ship Model in Regular and Irregular Waves of Extreme Steepness," AD 680123, SSC-190,
November, 1968.

Marsich, S. and Merega, F., "Strip Theory and Power Spectra] Density Function Ap-
plication to the Study of Ship Geometry and Weight Distribution Influence on Wave
Bending Moment," 7th Symposium on Naval Hydrodynamics, 1968.

Mathews, S. T., "Main Hull Girder Loads on a Great Lakes Bulk Carrier," Trans.
SNAME, 1968.

Moor, D. I., "Longitudinal Bending Moments on Models in Head Seas," Trans. Royal
Institute of Naval Architects, No. 109, 1967, pp. 117-26.

Moor, D. I. and Murdey, D. C., "Motions and Propulsion of Single Screw Models in
Head Seas," Trans. Royal Institution of Naval Architects, No. 110, 1968.

Moor, D. I. and Murdey, D, C., "Longitudinal Bending Moments on Models in Head

Seas," Trans. Royal Institution of Naval Architects, No. 108, 1966.

Muckle, W., "The Buoyancy Curve in Longitudinal Strength Calculations," The Ship-
builder and Marine Engine Builder, February, 1954.

Nagai, T. , Kakota, M. , Fukuda, T. and Dokai, Ken j i, "Or. the Experimental Results
of Ship Motion, Longitudinal Bending Moment and Slamming Pressure," Journal of
Society of Naval Architects of Japan, No. 120, 1966.

Nordenstrdm, N., "Methods tor Predicting Long-Term Distribution of Wave Loads and
Probability of Failure for Ships," Reports Det Norsl'.e Veritas. AnDcndices 8 and 9.

Nordenstr^m, N., "Further Analysis of Full-Scale Measurements of Midship Bending

Moment," Report from Chalmers University of Technology, May, 1965.

Nordenstr^m, N., Faltinsen, 0., and Pedersen, B., "Prediction of Wave-Induced

Motions and Loads for Catamarans," Proceedings of Third Offshore Technology Con-
ference, April, 1971. (Paper No. OTC 1418).

Nordenstrom, N. and Pederson, B. , "Calculations of Wave Induced Motions and

Loads, Progress Report No. 6, Comparison with Results from Model Experiments and
Full-Scale Measurements," Det Norske Veritas research Department, Report No. 68-
12-S, 1968.

Noritaka, Ando, Naokoto, Shimada and Hitoshi, Nagasawa, "Experiments on the

Strength of Liner Cargo Boat in Service," Report of Ship Institute, Tokyo, 1966.

Numata, Edward, "Longitudinal Bending and Torsional Moments Acting on a Ship Model
at Oblique Headings to Waves," Journal of Shij> Research, June, 1960.
 -98-
Pedersen, Bj^rn, "Wave-Induced Motions and Loads — Comparison bet\#een Theoretical
Results, Model Tests and Full Scale Measurementr," European Shipbuilding, No. 2,
1972.

Pe'rez y Perez, Leonardo, "A General Calculation of the Structrral Loads on a Ship
in a Seaway," Marine Technology, April, 1971.

Petrow, Richard, "The Death of World Glory," Popular Mechanics, July, 1969.

Planeix, J. M., "Wave Loads — A Correlation between Calculations and Measurements

at Sea," International Shipbuilding Progress, Vol. 19, August 1972.

Porter, W. R. , "Pressure Distribution, Added Mass and Damping Coefficient for

Cylinders Oscillating in a Free Surface," University of California, Institute of
Engineering Research, Series No. 82, 1960.

Raff, A. I., "Program SCORES - Ship Structural Response in Waves," SSC-230, 1972.

Raff, A. I., "Program SCORES - Ship Structural Response in Waves," Oceanics, Inc.
Technical Report No. 70-75, August, 1970.

Salvesen, N., Tuck, E. 0., Fa]tinsen, 0., "Ship Motions and Sea Loads," SNAME,
November, 1970.

Soding, Von H., "Berechnung von Schiffen im Seegang," Forschung und Entwicklung,
Schiff und Hafen, Heft 10/1971, 23, Jahrgang.

Soeding, H., "Eine Modifikation der Streifen Methode," Schiffsteuhnik Bd. 16,
Heft 80, 1969.

Swaan, W. A. , "Amidships Bending Moments for Ships in Waves," International Ship-

building Progress, Vol, 6, 1959.

Tasai, F., "Pressure Fluctuation on the Ship Hull Oscillating in Beam Seas,"
Journal of the Society of Naval Architects of West Japan, Vol. 121, 1968.
Tasai, F., "An Appropriate Calculation of Hydrodynamic Pressure on a Midship Sec-
tion Contour of a Ship Heaving and Pitching in Regular Head Waves," Reports on
Research Institute of Applied Mechanics, Kyushu University, Vol. Ü4, No. 48, 1966.

Tasai, F., "Hydrodynamic Force and Moment Produced by Swaying and Rolling Oscilla-
tion of Cylinders on the Free Surface," Reports of Research Institute for Applied
Mechanics, Kyushu University, Vol. 9, No. 35, 1961.

Tasai, F., "On the Damping Force and Added Mass of Ships Heaving and Pitching,"
Reports on Research Institute for Applied Mechanics, Kyushu University, Vol. 7,
No. 26, 1939, and Vol. 8, No. 31, 1960.

Tasai, F., and Takaki, M., "Theory anc Calculations of Ship Response in Regular
Waves," Symposium on Seaworthiness of Ships, Japan Society of Naval Architecture,
1969.

Taylor, K. V. and Bell: A. 0., "Wave-Excited Hull Vibration Stress Measurements

on a 47,000-Ton Deadweight Tanker," BSRA Report No. 115, 1966; Shipping World
and Shipbuilder, February, 1968.

Terazawa, K., "Response Operators of Ship Motions and Midship Bending Moments in
Regular Waves Calculated by the Research Committee No. 90 of the Shipbuilding
Research Association of Japan," Proceedings 11th ITTC, 1966.
 -99-

Terazawa, K. and Nakamura, S., "Calculation of Hydrodynamic Pressure on the

Section Contour of a Ship in Regular Head Waves," A note prepared for the Roire
Meeting of Committee 2, ISSC, 1968.

Timmann, R., and Newman, J. N., "The Coupled Damping Coefficient of a Symmetric
Ship," Journal of Ship Research, No. 4, March, 1962.

Vossers, G., Swaan, W. A. and Rijken, H., "Vertical and Lateral Bending Moment
Measurements on Series-6G Models," International Shipbuilding Progress, Vol. 8,
No. 83, July, 1961.

Vugts, J. H., "The Hydrodynamic Forces and Motions in Waves," Delft Thesis, 1970.

Vugts, J. H., "The Hydrodynamic Coefficients for Swaying, Heaving and Rolling
Cylinders in a Free Surface," Report No. 194, Delft Shipbuilding Laboratory, 1968.

Wahab, R., "Amidship Forces and Moments on a Cß = 0.80 Series Sixty Model in Waves
from Various Directions," International Shipbuilding Progress, 1967.

Watanabe, Y., "On the Theory of Heaving and Pitching Motions," Technology Report
of the Faculty of Engineering, Kyushu University, Vol. 31, No. 1, 1958.

Yamanouchi, Y and Ando, S., "Experiments on a Series 60, C = 0.70 Ship Model in
Oblique Regular Waves," Proceedings 11th ITTC, Tokyo, 1966.

Yamanouchi, Y., Goda, K. and Ogawa, A., "Bending and Torsional Moments and Motions
of a T2-SE-A1 Tanker Model in Oblique Regular Waves,* Note Presented to 2nd ISSC,
Delft, 1964.
Yamanouchi, Y., Oi, H., Takaishi, Y., Kihara, H., Yoshino, T., and Iizuka, M.,
"On the Ship Motions and Accelerations of a Nuclear Ship in Waves," Journal of
the Society of Naval Architects of Japan, Vol. 123, 1968.

Yuille, I. M., "Longitudinal Strength of Ships," Trans. RINA, 1962.

Zubaly, R. B. and Lewis, E. V., "Ship Bending Moments in Irregular Seas Predicted
from Model Tests," Webb Institute Report, December, 1963.

DYNAMIC LOADS

Admiralty Shipbuilding Committee Report No. 8, 1953.

Aertssen, Prof. G. and Lembre, Ir. R. De., "Calculation and Measurement of the
Vertical and Horizontal Vibration Frequencies of a Large Ore Carrier", North
East Coast Institution of Engineers & Shipbuilders Transactions, March 1970,

Aertssen, G., "Longitudinal Strength of Ships, ' International Shipbuilding Pro-

gress, 1970.

Aertssen, G., "Laboring of Ships in Rcugh Seas," Diamond Jubilee Inter. Meeting,
Transactions, SNAME, 1968.

Aertssen, G., "Service-Performance and Seakeeping Trials on M.V. Jordaena," RINA

Trans., 1966.

Andrews, J. N., and Dinsenbacher, A.D., "Response amplitude Operators and Whip-
ping Response of a Carrier Model in Random Vaves," hSRDC Report No. 2522, 1968.
 -100-

Belgova, M. A., "Determination of Overall Bending Moments Caused by Elastic Vi-

brations of Ships," Transactions of the Leningrad Institute of Water Transport,
BSRA Translation, 1962.

Bledsoe, Margaret D., Bussemaker, Otto, and Cummins, W. E. (1960) "Seakeeping

Trials on Three Dutch Destroyers," SNAME, Vol. 68, J960.

Bledsce, M. D., "Series Investigation of Slamming Pressures," David Taylor Model

Basin Report No. 1043, 1956.

Chu, W. H. and Abramson, H. N., "Hydrodynamic Theories of Ship Slamming, Review

and Extension", Southwest Research Institute Report No. 2, 1960.

Chuang, S. L., "Investigation of Impact of Rigid and Elastic Bodies with Water,"
NSRDC Report 3248, February, 1970.

Chuang, S. L., "Experiments on Flat-Bottom Slamming," Journal of Ship Research,

Vol. 10, No. 1, March, 1966.

Chuang, S. L., "Slamming of Rigid Wedge-Shaped Bodies with Various Deadrise

Angles," DTMB Report No. 2268, 1966.

Chuang, S. L., "Experimental Investigation of Rigid Flat-Bottom Slamming," DTMB

Report No. 2041, 1965.

Clevenger, R. L. and Melberg, L. C, "Slamming of a Ship Structural Model,"

M. S. Thesis, Massachusetts Institute of Technology, 1963.

Dalzell, J. F., "Memorandum on State of Art Design Methods for Slamming," In-
formal Note, May, 1960.

Egorov, I. T., "Free Surface Deformation and Further Hydrodynamic Phenomena by

the Entry of a Flat Plate on a Fluid," Trudy, Sudostroenie, pp. 87-98, 1965,
(Russian).
Egorov, I. T., "Impact on a Compressible Fluid," Translation NACA, Techn. Memo.
1413, 1956.

Ferdinande, V., "Analysis of Slamming Phenomena on a Model of a Cargo Ship in

Irregular Waves," International Shipbuilding Progress, 1969.

Ferdinande, V., and Lernbre, R. De., "Impact Phenomena and Wave Stresses on M.V.
Jordaens in Medium-Load and Full-Load Condition," Appendix III of Aertssen, G.,
"Service-Performance and Seakeeping Trials on M.V. Jordaens", Transactions RINA,
Vol. 108, 1966.

Fujita, Y., "On the Impulsive Pressure of a Circular Plate Falling Upon a Water
Surface," Journal S.N.A. of Japan, Vol. 94, pp. 105-110, 1954.

Gerlach, C. R., "Investigation of Water Impact of Blunt Rigid Bodies — Size

Scale Effects," Southwest. Research Institute, Technical Report No. 2, Project
No. 02-2036, November, 1968.

Gerlach, C. R., "Investigation of Water Impact of Blunt Rigid Bodies — Real

Fluid Effects," Southwest Research Institute, Department of Mech. Science,
December 29, 1967.

Goodman, R. A., "Wave-Excited Main Hull Vibration in Large Tankers and Bulk
Carriers," Transactions RINA, 1970.
 -101-
Greenbert, M. D., "Prediction of Ship Slamming Loads: On the Water Impact of a
Circular Cylinder," Therm. Advanced Research Inc., TAR-TR 6701, May 1967.

Greenspon, J. E., Jasper, N. H., and Birmingham, J. T., "Sea Tests of the USCGC
Unimak, Pressures, Strains and Deflections of the Bottom Plating Incident to
Slamming," Imernaticnal Shipbuilding Progress, Vol. 3, No. 25, September 1956,
pp. 474-486.

Hansen, K. E., "Pounding of Ships and Strengthening of Bottom Forward," Ship-

building and Shipping Record, Vol. 45, 1935.

Hawkins, S., Levine, G. H., and Taggart, R., "A Limited Survey of Ship Struc-
tural Damage", SSC-220, 1971.

Heller, S. R. and Kammerer, J. T.„ "Buckling of Long Slender Ships Due to Wave-
Induced Whipping," Pergamon Press, 1966.

Henry, J. R. and Bailey, F. C, "Slamming of Snips: A Critical Review of the

Current State of Knowledge," SSC-208, 1970.

Hoffman, D., and van Hooff, R., "Feasibility Study of Springing Model Tests of
a Great Lakes Bulk Carrier'1, D.Ü.T., U.S. Coast Guard, Tech. Report, July 1972.

Hoffman, D., and Maclean, W. M., "Ship Model Study of Incidence of Shipping
Water Forward," Marine Technology, Vol. 7, No. 2, April 1970, pp. 149-3 58.

Jasper, N. H., and Church, J. W., "Structural Seaworthiness Studies," Trans.

SNAME, Vol. 71, 1963, pp. 31-80.

Johnson, R. S., "The Effect of Air Compressibility in a First Approximation to the

Ship Slamming Problem," Journa1 of Ship Research, Vol. 12, No. 1, March 1968.

Junger, M. C, "Normal Modes of Submerged Plates and Shells," ASME Fluid-Solid In-
teraction Symposium, 1967.

Kamel, A. M., "Shock Pressures Caused by Waves Breaking Against Coastal Structures",
U.S. Army Engineer Waterways Experiment Station, Research Report H-68-2, September
1968.

Kaplan, P., and Sargent, T. P., "Further Studies cf Computer Simulation of Slam-
ming and Otiier Wave-Induced Vibratory Structural Loadings on Ships in Waves,"
SSC-231, 1972.

Kaplan, P., Sargent, T. P., and Raff, A. T., "An Investigation of the Utility of
Simulation to Predict Ship Structural Response in Waves," SSC-197, June 1969.

Kawakarru, II,, "Un the Impact Strength of Ships Due to Shipping Green Seas — Tow-
ing Experiments of a Ship Model in Regular Waves," Journal the Society of Naval
Architects of Japan, Vol. 125, 1969.

Kent, J. L., "The Cause and Prevention of Slamming of Ships in a Seaway", Trans.
North East Coast Institute Eng. Ship., Vol. 65, 1948-49.

Kent, J. L,, "The Design of SeKn,ndly Ships," Transactions North East Coast Inst.
Eng. Ship., Vol. 66, 1950-51.

King, J. F., "Heavy Weather Damage," Transactions North East Coast Institute Eng.
Ship., Vol. 51, 1934-35.
 -102-
Kline, R. G. and Clough, R. W., "The Dynamic Response of Ship's Hulls as Influenced
by Proportions, Arrangements, Loading and Structural Stiffness," Proc. SNAKE Spring
Meeting, Montreal, Canada^ 1967.

Kumai, T., and Tasai, F., "On the Wave Exciting Forces and Response of Whipping of
Ships," Europ. Shipbuilding, Vol. 4, No. 19, 1970.

Lehmann, G., "Bondenschaden im Vorschiff und die Neuen Vorschriften der Klassifi-
catlonsqesellschaften," Schiffbau, Vol. 37, 1936.

Lewis, E. V., and Wheaton, J. W., "Study of Combined Wave and Springing Stresses
of the Edward L. Ryerson," SNAME Symposium on Hull Stresses in Bulk Carriers in
the Great Lakes and Gulf of St. Lawrence Wave Environnv: t, Ottawa, 1971.

Lewison, G. R. G., "Slamming," Nat. Physical Laboratory, Report 138, March 1970.

Lewison, G. R. G., "The Control of Ship Slamming," Shipping World and Shipbuilder,
June 1969.

Lewison, G. R. G., "On the Reduction of Slainming Pressu-es," Transactions RINA,

1969.

lewison, G., "An Experimental Investigation of the Role of Air in Ship Slamming,"
University of California, Berkeley, College of Engineering, Report NA-66-12, 1966.

Lewison, G. and Maclean, W. M., "On the Cushioning of Water Impact by Entrapped
Air," Journal of Ship Reseanh, Vol. 12, No. 2, June 1968.

Lewison, G. and Maclean, W. M., "The Effect of Entrapped Air Upon the Slamn:ing of
Ship's Bottom," University of California, Berkeley, College of Engineering, Report
NA-66-5, 1966.

Maclean, W. M., "The Ship Dynamic Test Machine at the University of California,"
Thesis, University of Calfornia, Berkeley, May 1967.

Maclean, W. M., "The Ship Dynamic Test Machine at the University of California,"
Berkeley, California, College of Engineering, Report NA-66-1, January 1966.

Maclean, W. M., and Lewis, E. V., "Further Analysis of Slamming Stresses on S.S.
Woiyi vine State," Webb Institute of Naval Architecture, Report No. 10-17, January
1970.

Maximadji, A. I., "On the Problem of Norms for Longitudinal Rigidity of Ship's
Hulls Made From High Tensile Steel," Proc. 3rd ISSC, 1967.

McGoldrick, R. T., Gleyzal, A. N., Hess, R. L., and Hess, G. K., "Recent Develop-
ments in the Theory of Ship Vibrations," David Taylor Model Basin Report 739, 1953.

Meek, M., adams, R., Chapman, J. C, Reibe!, II., and Wieske, P., "The Structural
Design of the O.C.L. Container Ships," Transactions RINA, 1971,

Meyerhoff, W. K., "Uebersicht Über Grundlegende Theoretische und Experimentelle

Arbeiten Zum Problem der Bodenstosse by Schiffen," Jahrbuch der Schiifbautechn,
Gesellschaft, Band 61, pp. 147-163, 1967.

Miles, M. D., "On the Short Term Distribution of the Peaks of Combined Low Fre-
quency and Springing Stresses," SNAME Symposium on Hull Stresses in Bulk Carriers
in the Great Lakes and Gulf of St. Lawrence Wave Environment, Ottawa, 1971.
 -103-
Moran, J, P., "On the Hydrodynamic Theory of Water-Exit and Entry," Therm. Ad-
vanced Research Inc., Tar-TR 6501, March 1965.

Nagai, T., National Defense Laboratory of Japan, Technical Reports No. 156, Novem-
ber 1965; No. 263, May 1967; No. 318, May 1968.

Ochi, M. K., "Performance of Two Hull Forms (U and V) in Irregular Waves,'.' Sym-
posium on "Some Effects of Hull Form on Ship Performance in a Seaway," Trans.
SNAME, 1967.

Ochi, M. K3, "Ship Slamming,Hydrodynamic Impact Between Waves and Ship Bottom For-
ward," Fluid-Solid Interaction Symposium, Transactions ASME, 1967.

Ochi, M. K., "Prediction of Occurrence and Severity of Ship Slamming at Sea," Fifth
Symposium en Naval Hydrodynamics, O.N.R. 1967.

Ochi, M. K., "Experiments on the Effect of Bow Form on Ship Slamming," David Taylor
Model Basin Report No. 1400, 1962.
Ochi, M. K., "Model Experiments on Ship Strength and Slamming in Regular Waves,"
Transactions SNAME, Vol. 66, 1958.

Ochi, M. K., and Bledsoe, M. D., "Hydrodynamic Impact with Application to Ship
Slamming," Fourth Symposium on Naval Hydrodynamics, Office of Naval Research,
ACR-73.

Ochi, M. K. and Bonilla-Norat, J., "Pressure-Velocity Relationship in Impact of a

Ship Model Dropped onto the Water Surface and in Slamming in Waves," Naval Ship
Research and Development Center Report, in preparation.

Ochi, M. K. and Motter, L. E., "A Method to Estimate Slamming Characteristics for
Ship Design," Marine Technology, Vol. 8, No. 2, April 1971.

Ochi, M. K. and Motter, L. E., "Prediction of Extreme Values of Impact Pressure

Associated with Ship Slamming," Journal of Ship Research, Vol. 13, No. 2, June 1969.

Ochi, M. K. and Schwartz, F. M., "Two-Dimensional Experiments on the Effect of Hull

Forms on Hydrodynamic Impact," D.T.M.B. Report 1994, 1966.

Ogilvie, T. F., "Compressibility Effects in Ship Slamming," Schiffstechnik, Band 10,

Heft 53, 1963.

Parga, J. B., "The Strengthening of Bottom Forward," Transactions R1NA, Vol. 107,
1965.

Schwendler, R. G., "Improved Analog Computer Analysis of the Bow Slamming Problem,"
Computer Engineering Associates, Report on U.S. Navy Project ESD 178, April 1962.
(See Jasper and Church, SNAME 1963).

Se'Llars, F., "The Influence of Structural Characteristics on Slamming Impact Pres-

sures," Journal of Ship Research, March 1971.

Sellers, M. L., and Kline, R. G., "Some Aspects of Ship Stiffness," Transactions
SNAME, Vol. 75, 1967, pp. 268-95.

St. Denis, M. and Fersht, S., "On the Effect of Ship Stiffness Upon the Structural
Response of a Cargo Ship to an Impulsive Load," SSC-186, September 1968.

Swaan, W. A. and Vossers, G., "The Effect of Forebody Section Shape on Ship Behavior
in Waves," Transactions RINA, Vol. 103, 1961.
 -104-

Symposium on Hull Stress in Bulk Carriers in The Great Lakes, SNAME, Ottawa, 1971.

Szebehely, V. G., "Hydrodynamic Impact," Applied Mechanics Reviews 12, pp. 297-
300, 1959.

Szebehely, V. G. and Lum, S. M., "Model Experiments on Slamming of a Liberty Ship

in Head Seas," DTMB Report No. 914, 1955.

Szebehely, V. G. and Ochi, M. K., "Hydrodynamic Impact and Water Entry," Applied
Surveys, Spartan Books, Washington, pp. 951-957, 1966.
Tasai, F., "Wave Height at the Side of a Two-dimensional Body Oscillating on the
Surface of a Fluid," Report of the Research Institute for Applied Mechanics, Kyushu
University, Vol. 9, No. 35, 1961.

Tasaki, R., "On Shipment of Water in Head Waves," Tenth ITTC, London, 1SS3.

Teledyne Materials Research, "ABS Program for the Collection of Long-Term Stress
Data from Large Tankers and Bulk Carriers," final report, 1971.

Tick, L. J., "Certain Probabilities Associated with Bow Submergence and Ship Slam-
ming in Irregular Seas," Journal of Ship Research, 1958.

Townsend, H. S., "A Series of Cergo Hull Forms," Symposium on "Some Effects of Hull
Form on Ship Performance in a Seaway," SNAME, 1967.

Townsend, H. S., "Some Observations on the Shape of Ship Forebodies with Relation
to Heavy Weather," New York Metropolitan Section; SNAME, 1960.

Verhagen, J. H. C, "The Impact of a Flat Plate on a Water Surface," Journal of Ship

Research, Vol. 11, No. 4, December, 1967.

Von Karman, Th., "The Impact of Sea Plane Floats During Landing," NACA TN 321, 1929.

Wahab, Ir. R., "Research on Bulbous Bow Ships Part IB. The Behavior of a Fast Cargo
Liner with a Conventional and with a Bulbous Bow in a Seaway," International Ship-
building Progress, Vol. 13, 1966, pp. 181-192.

Warnsinck, W. H., and St. Denis, M. (1957), "Destroyer Seakeeping Trials," NSMB,
1957, pp. 439-467 and 971-980.

Wheaton, J. W., Kano, C. H., Diamant, P. f. and Bailey, F. C, "Analysis of Slamming

Data from the S.S. Wolverine State," SSC-210, Final Report on Project SR-172, "Slam-
ming Studies" to the Ship Structure Committee, 1970.

COMBINED LOADS

Abraharasen, E., Nordenstnhn, N. and Roren, E. M. Q., "Design and Reliability of

Ship Structures," Proceedings of Spring Meeting, SNAME, April 1970.

Aertssen, G., "Longitudinal Strength of Ships," International Shipbuilding Progress,

1970.

Ang, Alfredo H. S., "Safety Factors and Probability in Structural Design," Journal
of the Structural Division, ASCE, July, 1969.

Asplund, S.O., "The Risk of Failure," Structural Engineer, Vol. 36, No. 8, 1958.
 -105-
Goodman, R., "Prediction of Ship Response Criteria for Structural Design," Seminar
on the Application of Ship Motion Research to Design, University of Southampton,
1970.

Goodman, R., "Wave-Excited Main Hull Vibration in Large Tankers and Bulk Carriers,"
Transactions RINA, 112, 1970.

Hoff, N.J., "Philosophy of Safety in the Supersonic Age," Report No. 87, NATC,
Sixt.i General A.csembly of A.G.A.R.D., August, 1956.

Mansour, A., "Methods of Computing the Probability of Failure Under Extreme Values
of Bending Moment," MIT Report, December 1971.

Miles, M., "The Theoretical Statistical Distribution of the Peaks of Combined Spring-
ing and Wave-Induced Stress Loads," National Research Council Laboratory Technical
Report, LTR-SH-103, February 1970.

Nibbering, Ir. J. J. W., "Permissible Stresses and Their Limitations," SSC-206, 1970.

Turkstra, C. J., "Choice of Failure Probabilities," Journal of Structures Division,

Proc. A.S.C.E., ST6, 1967.
LONGITUDINAL STRENGTH DESIGN

Abrahamsen, E., Nordenstrtfm, and Roren, E. M. Q., "Design and Reliability of Ship
Structures," Proceedings of Spring Meeting, SNAME, April 1970.

Abrahamsen, E., "Structural Design Analysis of Large Ships," Transactions SNAME,

1969.

Abrahamsen, E., "Recent Developments in the Practical Philosophy of Ship Structural

Design," Proceedings of SNAME Spring Meeting, July 1967.

Abrahamsen, E., "Structural Safety of Ships and Risks to Human Life," European
Shipbuilding, Vol. II, 1962.

Aertssen, G., "Longitudinal Strength of Ships," International Shipbuilding Progress

1970.

Altenburg, C. J., Scott, R. J., Gibbs and Cox, "Design Considerations for Aluminum
Hull Structures — Study of Aluminum Bulk Carrier," SSC-218, 1971.

Buchanan, G., Dobson, Ir. R.J.C. and Jensen, C.J.G., "Lloyd's Register of Shipping's
Approach to the Control of the Incidence of Brittle Fracture in Ship Structures,"
No. 56, 1969.

Caldwell, J.B., "Structural Design Philosophy," Society of Naval Architects of Japan,

September, 1970, Tokyo, Japan.

Caldwell, J.B., "Ultimate Longitudinal Strength," Trnasactions RINA, Vol. 107,

1965.

Crooker, T.W., "Designing Against Structural Failure Caused by Fatigue Crack Propa-
gation," ^y^J^^^SlB-JS^n^k* December, 1972.

de Wilde, G,, "Structural Problems in Ships with Large Hatch Openings," Interna-
tional Shipbuilding Progress, Vol. 14, January-February, 1967.

Dunn, T.W., "Reliability in Shipbuilding," Transactions SNAME, 1964.

 -106-
Evans, J. H., "Structural Performance Norms in Ship Design," Transactions SNAME,
1970.

Evans, H.J., "A Structural Analysis and Design Integration," Transactions SNAME,
1958.

Freudenthal, A.M., "Probabilistic Approach to Economic Design of Maritime Struc-

tures," XXIInd International Navigation Congress, Paris, 1969.

Freudenthal, A.M., "Probabilistic Evaluation of Design Criteria," Bulletin of the

Permanent International Association of Navigation Congresses, 1968-1969 (Vol. Ill/
IV) .

Freudenthal, A.M., "The Safety of Structures," Transactions, ASCE, Vol. 112, 1947,
pp. 125-180.
Gerard, G,, "Structural Design Philosophy and Procedures," Section 5 from "A
Long-Range Research Program in Ship Structural Design," SSC-124, November 1959.

Getz, J.R., "Longitudinal Strength and Minimum Weight," DTMB Report No. 1699,
December 1962.

Hoff, N.J., "Philosophy of Safety in the Supersonic Age," Report No. 87, NATC,
Sixth General Assembly of A.G.A.R.D., August 1956.

Itagaki, H., Shinozuki, M. and Freudenthal, A.M., "Reliability of Single- and Multi-
Member S'v-ni.-ture Subjected to Fluctuating Load," Selected Papers from Journal of the
Society of Naval Architects of Japan, Vol. 8, Published 1971.

Jasper, N.H., "rj>.mperature-lnduced Stresses in Beams and Ships," ASNE Journal, Vol.
68, August 1956, pp. 485-497.

Keil, Dr. A.H., "Research Needs in Ship Structures," Naval Engineers Journal,Vol. 83,
No . 4 , August 1971.

Kendrick, S.B., "The Structural Design of Supertankers," Transactions RINA, 112,

1970.

Kline, R., "Some Aspects of Ship Stiffness," Transactions SNAME, 1967.

Lewis, E.V., "Predicting Long-Term Distributions of Wave-Induced Bending Moment on

Ship Hulls," SNAME, Spring 1967.

Lindemann, K., "Analysis of Long-Term Distributions of Extreme Values," Norske Ver-

itas Report No. 71-17-S, 1971.

"The Loss of the Texaco Oklahoma," Proceedings of the Marine Safety Council, U.S.
Coast Guard, October 1972.

Mansour, A., "Probabilistic Design Concepts in Ship Structural Safety and Reli-
ability," Transactions SNAME 1972.

Meek, M., Adams, R., Chapman, J.C., Reibe], H., and Wieske, P.,,"The Structural De-
sign of the O.C.L. Container Ship/' Royal Institution of Naval Architects Interna-
tional Spring Meeting, London 1971.

Nibbering, J.W.W., "Permissible Stresses and Their Limitation," Report No. 141, Delft
Ship Structure Laboratory, 1969 and Report SSC-206, 1970.
 -107-
Nibbering, Ir. J.J.W., Lalleman, A.W., "Low Cycle Fatigue Problems in Shipbuilding;
Crack Propagation in Coarse-Grained Zones of Thick Plates," Fatigue of Welded Struc-
tures Conference, Brighton, 6-9 July, 1970.

Nibbering, Ir. J.J.W., and Lint, J. van, "Low-Cycle Fatigue of Steel Structures,"
Netherlands Ship Research Centre, April 1966.

Nordenstrtfm, N., "Probability of Failure for Weibull Load and Normal Strength," Re-
port No. 69-28-5, Norske Veritas, March 1970.

Panel HS-4, "Higher-Strength Steels in Hull Structures," SNAME Bulletin No. 2-i9,
December 1971.
Report of Committee 10 on Design Procedure, Proceedings of ISSC, Tokyo, 1970.

Roberts, W.J., "Strength of Large Tankers," Transactions N.E.C.I., Vol. 86, January
1970.

Runciman, W.G., "The 4th Blackadder Lecture 1970: The Unprofitability of Ship Own-
ing," February 1970, from the March 1970 North East Coast Institution of Eng. and
Shipbuilding Transactions.

Schultz, Von Dr.-Ing. H.G., "Festigkeitsprobleme im Großschiffbau," Jahrbuch der

Schiffbautechnischen Gesellschaft, 63, Band 1969.

Spinelli, L., "A Few Notes on R.I.Na. Rules 1971 considered ir the Light of the Pres-
ent Knowledge on Ship Structural Design," Registro Italiano Navale, Technical Bulle-
tin No. 46, July 1971.

"Standards of Strength for Seagoing Steel Vessels," U.S.S.R. Register of Shipping.

Steneroth, E., "Reflections Upon Permissible Longitudinal Stresses in Ships," Royal

Institution of Naval Architects Transactions 109, 1967, pp. 223-36.

"Structural Safety - A Literature Review," by Task Committee on Structural Safety,

Journal of Structural Div., Proceedings of ASCE, April 1972, pp. 845-863.

Turkstra, C.J., "Choice of Failure Probabilities," Journal of the Structural Division,

Proceedings of ASCE, December 1967.

Vasta, J,, Palermo, P.M., "An Engineering Approach to Low-Cycle Fatigue of Ship Struc-
tures," Transactions, SNAME, 1965.

Vasta, J., "Lessons Learned from Full-Scale Structural Tests," Transactions, SNAME,
1958.

Young, R.T., "The Development of Classification Society Rules for the Construction
and Survey of Large Modern Tankers," ABS report, March 1970.

Yuille, I.M., "Longitudinal Strength of Ships," Transactions, RINA, 1962.

 SHIP RESEARCH COMMITTEE
Maritime Transportation Research Board
National Academy of Sciences-National Research Council

The Ship Research Committee has technical cognizance of the interagency

Ship Structure Committee's research program:

PROF. R. A. YAGLEr/ Chairman, Prof, of Naval Architecture, University of Michigan

DR. H. N. ABRAMSON, Tech. Vice President, Dept. of Mech. Sciences, S.W. Res. Inst.

PROF. R. W. CLOUGH, Prof, of Civil Engineering, University of California

PROF. J. E. GOLDBERG, School of Civil Engineering, Purdue University

PROF. W. J. HALL, Prof, of Civil Engineering, University of Illinois

DR. S. R. HELLER, JR., C'man, Civil & Mech. Eng. Dept., The Catholic Univ. of America

MR. G. E. KAMPSCHAEFER, JR., Manager, Technical Services, ARMCO Steel Corporation

MR. R. C. STRASSER, director of Research, Newport News Shipbuilding & Dry Dock Company

MR. H. S. TOWNSEND, Vice President, U. S. Salvage Association, Inc.\

DR. S. YUKAWA, Consulting Engineer, General Electric Company

Advisory Group I, "Ship Response and Load Criteria", prepared the project
prospectus and evaluated the proposal for this project:

DR. H. N. ABRAMSON, Chairman, Tech. Vice Pres., Dept. of Mech. Sciences, S.W. Res. Inst.

DR. C. CHRYSSOSTOMIDIS, ABS* . Prof, of Naval Architecture, Mass. Inst. of Technology

PROF. A. M. FREUDENTHAL, College of Engineering, George Washington University

DR. R. GLASFELD, Naval Architect, General Dynamics Corporation

MR. R. G. KLINE, Assoc. Research Consultant, U. S. Strel Corporation

DR. M. K. OCHI, Research Scientist, Naval Ship Research S Development Center

MR. R. C. STRASSER, Director of Research, Newport News Shipbuilding & Dry Dock Company

The SR-198 Project Advisory Committee provided the liaison technical

guidance, and reviewed the project reports with the investigator:

MR. W. H. Buckley, Chairman, Coordinator of Hydrofoil Struct. Res., Naval Ship R&D Ctr

MR. C. M. Cox, Asst. Naval Architect, Newport News Shipbuilding & Dry Dock Company
PROF. A. M. Freudenthal, College of Engineering * George Washington University

IO\
 SHIP STRUCTURE COMMITTEE PUBLICATIONS

These documents are distributed by the National Technical

Information Service, Springfield, Va. 221 SI. These doc-
uments have been announced in the Clearinghouse Journal
U.S. Government Research & Development Reports (USGRDR)
under the in 'icated AD numbers.

SSC-227, Tanker Transverse Strength Analysis - User's Manual by R. Nielsen,

P. Y. Chang, and L. C. Deschamps. 1972. AD 752771.

SSC-228, Tanker Transverse Strength Analysis - Programmer's Manual by

R. Nielsen, P. Y. Chang, and L. C. Deschamps. 1972. AD 752742.

SSC-229, Evaluation and Verification of Computer Calculations of Wave-Induced

Ship Structural Loads by P. Kaplan and A. I. Raff. 1972. AD 753220.

SSC-230, Program SCORES - Ship Structural Response in Waves by A. I. Raff,

1972. AD 752468.

SSC-231, Further Studies of Computer Simulation of Slamming and Other Wave-

Induced Vibratory Structural Loadings on Ships in Waves by
P. Kaplan and T. P. Sargent. 1972. AD 752479.

SSC-232, Study of the Factors which Affect the Adequacy of High-Strength, Low
Alloy, Steel Weldments for Cargo Ship Hulls by E. B. Norn's,
A. G. Pickett, and R. L. Wylie. 1972. AD 752480.

SSC-233, Correlation of Model a?d Full-Scale Results in Predicting Wave

Bending Moment Trends by D. Hoffman, J. Williamson, and E. V. Lewis,
1972. AD 753223.

SSC-234, Evaluation of Methods for Extrapolation of Ship Bending Stress Data

by D. Hoffman, R. van Hooff, and E. V. Lewis. 1972. AD 7532?4.

SSC-235, Effect of Temperature and Strain Upon Ship Steels by R. L. Rothman

and R. E. Monroe. 1973.

SSC-236, A Method for Digitizing, Preparing and Using Library Tapes of Ship
Stress and Environment Data by A. E. Johnson, Jr., J. A. Flaherty,
and I. J. Walters. 1973.

SSC-237, Computer Programs for the Digitizing and Using of Library Tapes of
Ship Stress and Environment Data by A. E. Johnson, Jr.,
J. A. Flaherty, and I. J. Walters. 1973.

SSC-238, Design and Installation of a .Ship Response Instrumentation System

Aboard the SL-7 Class Coniainership S. S. SEA-LAND McLEAN by
R. A Fain. 1973.

SSC-239, Wave Loads In a Model of the SL-7 Containership Running At Oblique

Headings in Regular Waves by J. F. Dalzell and M. J. Chiocco. 1973.

i ol