Professional Documents
Culture Documents
Ada 294491
Ada 294491
SSC-381
RESIDUAL STRENGTH OF
DAMAGED MARINE STRUCTURES
1 9 9 5 0 5 3 1 0 7 7
T-^G OUALUnr JJf SPSGfI3 1
This document has been approved
for public release and sale; its
distribution is unlimited
SHIP STRUCTURE COMMITTEE
1 9 9 5
SHIP STRUCTURE COMMITTEE
The SHIP STRUCTURE COMMITTEE is constituted to prosecute a research program to improve the hull structures of ships and other
marine structures by an extension of knowledge pertaining to design, materials, and methods of construction.
RADM J. C. Card, USCG (Chairman)
Chief, Office of Marine Safety, Security
and Environmental Protection
U. S. Coast Guard
Mr. Thomas H. Peirce
Marine Research and Development
Coordinator
Transportation Development Center
Transport Canada
Mr. Edward Comstock
Director, Naval Architecture
Group (SEA 03H)
Naval Sea Systems Command
Mr. Edwin B. Schimler
Associate Administrator for Ship-
building and Technology Development
Maritime Administration
Mr. Thomas W. Allen
Engineering Officer (N7)
Military Sealift Command
Dr. Donald Liu
Senior Vice President
American Bureau of Shipping
Mr. Warren Nethercote
Head, Hydronautics Section
Defence Research Establishment-Atlantic
EXECUTIVE DIRECTOR
CDR Stephen E. Sharpe, USCG
U. S. Coast Guard
CONTRACTING OFFICER TECHNICAL REPRESENTATIVE
Mr. William J. Siekierka
Naval Sea Systems Command
SHIP STRUCTURE SUBCOMMITTEE
The SHIP STRUCTURE SUBCOMMITTEE acts for the Ship Structure Committee on technical matters by providing technical
coordination for determinating the goals and objectives of the program and by evaluating and interpreting the results in terms of
structural design, construction, and operation.
MILITARY SEALIFT COMMAND
Mr. Robert E. Van Jones (Chairman)
Mr. Rickard A Anderson
Mr. Michael W. Touma
Mr. Jeffrey E. Beach
AMERICAN BUREAU OF SHIPPING
Mr. Stephen G. Arntson
Mr. John F. Conlon
Mr. Phillip G. Rynn
Mr. William Hanzelek
MARITIME ADMINISTRATION
Mr. Frederick Seibold
Mr. Richard P. Voelker
Mr. Chao H. Lin
Dr. Walter M. Maclean
NAVAL SEA SYSTEMS COMMAND
Mr. W. Thomas Packard
Mr. Charles L Null
Mr. Edward Kadala
Mr. Allen H. Engle
U.S. COASTGUARD
CAPT G. D. Marsh
CAPTW. E. Colburn, Jr.
Mr. Rubin Scheinberg
Mr. H. Paul Cojeen
TRANSPORT CANADA
Mr. John Grinstead
Mr. Ian Bayly
Mr. David L. Stocks
Mr. Peter Timonin
DEFENCE RESEARCH ESTABLISHMENT ATLANTIC
Dr. Neil Pegg
LCDR Stephen Gibson
Dr. Roger Hollingshead
Mr. John Porter
SHIP STRUCTURE SUBCOMMITTEE LIAISON MEMBERS
U. S. COAST GUARD ACADEMY
LCDR Bruce R. Mustain
U. S. MERCHANT MARINE ACADEMY
Dr. C. B. Kim
U. S. NAVAL ACADEMY
Dr. Ramswar Bhattacharyya
CANADA CENTRE FOR MINERALS AND
ENERGY TECHNOLOGIES
Dr. William R. Tyson
SOCIETY OF NAVAL ARCHITECTS AND
MARINE ENGINEERS
Dr. William Sandberg
U. S. TECHNICAL ADIVSORY GROUP TO THE
INTERNATIONAL STANDARDS ORGANIZATION
CAPT Charles Piersall
NATIONAL ACADEMY OF SCIENCES
MARINE BOARD
Dr. Robert Sielski
NATIONAL ACADEMY OF SCIENCES -
COMMITTEE ON MARINE STRUCTURES
Mr. Peter M. Palermo
WELDING RESEARCH COUNCIL
Dr. Martin Prager
AMERICAN IRON AND STEEL INSTITUTE
Mr. Alexander D. Wilson
OFFICE OF NAVAL RESEARCH
Dr. Yapa D. S. Rajapaske
STUDENT MEMBER
Mr. Trevor Butler
Memorial University of Newfoundland
Member Agencies:
American Bureau of Shipping
Defence Research Establishment Atlantic
Maritime Administration
Military Sealift Command
Naval Sea Systems Command
Transport Canada
United States Coast Guard
Ship
Structure
Committee
An Interagency Advisory Committee
March 17, 1995
Address Correspondence to:
Executive Director
Ship Structure Committee
U . S . C o a s t Gua rd (G-MI/S S C )
2100 S eco nd S t reet , S . W.
Wa s h i ng t o n, D . C . 20593-0001
P h : (202) 267-0003
F a x : (202) 267-4677
S S C -381
S R-1341
RES ID U AL S TRENGTH OF D AMAGED MARINE S TRU C TU RES
D eep dra ft merch a nt ves s els bui lt duri ng t h e pa s t few deca des a re
les s ro bus t t h a n t h ei r predeces s o rs . F i ni t e element a na lys i s ,
reduct i o n o f co rro s i o n a llo wa nces , a nd o t h er des i g n pra ct i ces
h a ve pro vi ded a mea ns fo r t h e des i g ner t o reduce s ca nt li ng s
o vera ll t o t h e mi ni mum requi red. Wh ere o lder s t ruct ures h a d
s i g ni fi ca nt redunda nci es t h e newer s t ruct ures h a ve rela t i vely
li t t le. F o llo wi ng t h e i mpa ct s o f s evera l ma ri ne di s a s t ers ,
o pera t o rs , reg ula t o rs , a nd cla s s i fi ca t i o n s o ci et i es h a ve h a d t o
mo re clo s ely mo ni t o r t h e s t reng t h o f a g i ng a nd da ma g ed
s t ruct ures .
In 1983 t h e S h i p S t ruct ure C o mmi t t ee s po ns o red a s ympo s i um o n
"Th e Ro le o f D es i g n, Ins pect i o ns a nd Redunda ncy i n Ma ri ne
S t ruct ures Reli a bi li t y". Th i s repo rt i s t h e t h i rd fo llo w-o n
pro ject s i nce t h a t s ympo s i um. It i s i nt ended t o i nt ro duce t o t h e
i ndus t ry a mea ns o f a s s es s i ng res i dua l s t reng t h . S o me current ly
a va i la ble met h o ds t o mea s ure res i dua l s t reng t h a re pres ent ed a nd
eva lua t ed fo r effect i venes s a nd a ppli ed t o ca s e s t udi es . It
o ut li nes pra ct i ca l a na lyt i ca l pro cedures t o ma ke deci s i o ns t o
res po nd t o fa i lures fo und i n ma ri ne s t ruct ures .
C ARD
Rea r Admi ra l, U . S . C o a s t Gua rd
C h a i rma n, S h i p S t ruct ure C o mmi t t ee
Acces i o n F o r
NTIS C RA&I
D TIC TAB
U na nno unced
J us t i fi ca t i o n
By
Distribution/
D
D
Availability Codes
Dist
H
Avail and/or
Special
Technical Report Documentation Page
1. Report No.
5 SC-3 81
2. Gevrnm*nt Accession No.
P B95-135419
3. Recipient's Co'alog No.
4. Title and Subtitle
5. Report Oat*
S ept ember 1994
"Res i dua l S t reng t h o f D a ma g ed Ma ri ne S t ruct ures ' 6. Performing Organization Cods
7. Aut h o r' ) o h ruba J . Gh o s e, Na t a le S . Na ppi a nd
C h ri s t o ph er J . Wi emi cki
8. Performing Orgoniiotion Raport No.
"SR-1 3 41 "
9. Performing Organization Nama and Address
D es i g ners & P la nners , Inc.
2120 Wa s h i ng t o n Blvd. , S ui t e 200
Arli ng t o n, VA 22204
10. Work Unit No. (TRAIS)
11. Contract or Grant No.
DTCG23 -9 2-C-E0 1 0 9 0 -1
12. Sponsoring Agency Name and Address
S h i p S t ruct ure C o mmi t t ee
U . S . C o a s t Gua rd
2100 S eco nd S t reet , S . W.
Wa s h i ng t o n, P . C . 20593
15. Supplementary Notes
13. Type of Report and Period Covered
FINAL
14. Sponsoring Agoney Cod
G-M
S po ns o red by t h e S h i p S t ruct ure C o mmi t t ee a nd i t s member a g enci es .
1 6 . Abstruct
Tra di t i o na lly a s s es s ment o f s h i p's lo ng i t udi na l s t reng t h h a s been ma de by co mpa ri ng
t h e ela s t i c s t res s es a t t h e deck o r bo t t o m s h ell t o fra ct i o ns o f t h e ma t eri a l yi eld
s t reng t h . Th i s res ult s i n h i g h res erve ca pa ci t y due t o i nh erent redunda nci es i n
s h i p s t ruct ures . Res i dua l s t reng t h , wh i ch i s defi ned a s t h e s t reng t h o f t h e
s t ruct ure a ft er da ma g e, h a s ra rely been co ns i dered ei t h er duri ng des i g n o r a t t h e
t i me o f repa i r. In t h i s repo rt , key element s requi red t o undert a ke a n eng i neeri ng
a na lys i s t o eva lua t e t h e res i dua l s t reng t h h a ve been i dent i fi ed. Emph a s i s h a s been
pla ced o n a s s es s i ng t h e res i dua l s t reng t h o f ma ri ne s t ruct ures da ma g ed due t o
no rma l o pera t i ng lo a ds . Met h o ds a va i la ble t o i ndus t ry fo r eva lua t i o n o f da ma g e
s uch a s , fra ct ure a nd ult i ma t e s t reng t h h a ve been s umma ri zed. An ex a mple pro blem,
i llus t ra t i ng t h e a ppli ca t i o n o f a n i nt eg ra t ed a ppro a ch t o res i dua l s t reng t h
a s s es s ment o n a pa rt i cula r s h i p t ype, i s pres ent ed.
1 7 . Key Words
Res i dua l S t reng t h , U lt i ma t e S t reng t h ,
F ra ct ure Mech a ni cs , F a t i g ue, P erma nent
D efo rma t i o n, P la s t i ci t y, Hull S t ruct ure
19. Security Classif. (of this report)
U ncla s s i fi ed
18. Distributie! Statement
Ava i la ble fro m:
Na t i o na l Tech ni ca l Info rma t i o n S ervi ce
S pri ng fi eld, VA 22161
D i s t ri but i o n U nli mi t ed
20. Security Classif. (of this page)
U ncla s s i fi ed
21. No. of Pages
180
22. Price
$27 .00Pap3:
SI 2, 5 0 Mic: o
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
e
u
s
s
oo
C
o
'2
o
>
c
o
U
Q.
<
3
O
I
s
s:
I
Q
<
cm
.S.SU1
00 00
J op oo
o -fr en -\
2 a
S2 12
< L > U -H
6 S3
I B 6 6
2, 3
N N N
00
d> _O0 00
.an
ills
0 0 0 0 0 0 J
^H fN >+ o
2 g
seH
H(UC3
8 63
33300
6 i.3
Ss
1
IM 0 0.00
rn
O N -*
<N
t> 0
0r,
l "S" T3
.5 Q. O* Ott! >>
n
P
3:8 s
3 3
o
.a .a
affMu u
qq-^OfS rn
<N ~* >/-> -<
m
12 (2
a
52 i2
88
il i2 B e-^-M
1 -1 a a a
IIJJJ 6 6
tu
o
u
S-
0 5
ai 25
o_
"CM
-W ft)
10, 11, 12 13 14 15, 16
O
O
o
s
tu
riches
S3
s
g
Q
C
o
'2
>
c
o
U
2
a.
<
I
-O
-
S3
I
s;
I
-0
I
iisJ
c
e
n
t
i
m
e
t
e
r
s
c
e
n
t
i
m
e
t
e
r
s
m
e
t
e
r
s
k
i
l
o
m
e
t
e
r
s
<
lO ON VO W
NOOH
m
tt
^
i
n
c
h
e
s
f
e
e
t
y
a
r
d
s
m
i
l
e
s
.S *s V.B
8 S 6,
2
i x:
i2 S2
M < 4
6 63
g
3 3 3 3 O
CT CT IT* O"
1
^
oo oo GO GO JB
ON
in o oo NO Tt
voddoid
00
p
$ sll
w W w w A
3 3 3 3 E
cr cr cy cr o
00 GO 00 OO Cd
.S^^6
w
0
00.* <
^Ti- -i JJJJ6 6
52 S2
a a
12 S2 j2 i2 5
aass 6 6
111 { 2 i2 Q S
r3rar3^3 a D a) !) ="=
c E c -J a a a u o
^t r- >o m vo
cN <+ ON oq p t~;
io<nvoom'
00
X oo
a 3
a-o
3 3
*.s
o
8 ^
o o
3
'
atrwiu
ah.Sc
4^ 13^ T3
W
u
_y
u
3U
<2i
rxfo
m >.
St
X) \C
S3
6
00 -S
K S
< D rcd
-O UM
UM
TABLE OF CONTENTS
S ect i o n P a g e
1 . 0 INTRODUCTION 1
1. 1 Ba ckg ro und 1
1. 2 Object i ves 2
1. 3 Appro a ch 2
2. 0 COLLECTION AND EVALUATION OP MARINE STRUCTURE DAMAGE
DATA 4
2. 1 C o llect i o n o f D a t a . . 4
2. 2 S umma ry o f C a s ua lt y D a t a 6
2. 3 D i s cus s i o n o f F i ndi ng s 12
2. 4 C o nclus i o ns 17
2. 5 References 18
3. 0 ELEMENTS OP RESIDUAL STRENGTH ASSESSMENTS OF MARINE
STRUCTURES 19
3. 1 Int ro duct i o n 19
3. 2 D es i g n P h i lo s o ph i es 19
3. 3 Redunda ncy 21
3. 4 Types o f D a ma g e a nd F a i lure 23
3. 4. 1 D uct i le F a i lure 23
3. 4. 1. 1 S t reng t h o f D a ma g ed S t ruct ure
U nder La t era l Lo a d 27
3. 4. 1. 2 Effect o f Lo ca li zed Imperfect i o n . . 32
3. 4. 2 F ra ct ure F a i lure 32
3. 4. 2. 1 F a t i g ue Ana lys i s 34
3. 4. 2. 2 F ra ct ure Mech a ni cs 3 8
3. 5 Ro le o f Ins pect i o n 51
3. 6 References 56
4. 0 METHODS TO EVALUATE RESIDUAL STRENGTH OF DAMAGED MARINE
STRUCTURES 60
4. 1 Indi rect Met h o d 60
4. 1. 1 S ervi ce Li fe Ex peri ence 60
4. 1. 2 Eng i neeri ng P ro cedures fo r P erma nent
D efo rma t i o n As s es s ment 67
4. 1. 3 Eng i neeri ng P ro cedures fo r C ra ck As s es s ment 73
4. 1. 3. 1 D es cri pt i o n o f P D 6493 . . 73
4. 1. 4 Two D i mens i o na l Met h o ds fo r U lt i ma t e
S t reng t h As s es s ment 84
4. 2 D i rect Met h o ds . . . ; 91
4. 3 S umma ry o f C urrent Indus t ry P ra ct i ce 95
4. 3. 1 S h i p S t ruct ure 95
4. 3. 2 Offs h o re S t ruct ures 105
4. 3. 3 Ai rcra ft S t ruct ures 110
4. 4 References 116
5. 0 ULTIMATE STRENGTH ANALYSIS OF A TYPICAL TANKER 124
5. 1 D es cri pt i o n o f t h e P ro blem 124
5. 1. 1 Ta nker C o nfi g ura t i o n 124
5. 1. 2 Lo ca t i o ns 124
5. 1. 3 Type o f D a ma g e a nd C ri t eri a 127
5. 1. 4 Lo a di ng 127
5. 2 Lo ca l Level Ana lys i s 131
5. 2. 1 Ex a mple 1 131
5. 2. 2 Ex a mple 2 141
5. 3 Glo ba l Level Ana lys i s 147
5. 4 References 155
6 . 0 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK . . . . 156
11
LIST OF FIGURES
F i g ure P a g e
No .
3-1 F AILU RE OF BRITTLE AND D U C TILE S YS TEMS 24
3-2 VARIOU S S TRU C TU RAL F AILU RE MOD ES 25
3-3 LOAD -S HORTENING C U RVES F OR LONG, TRANS VERS ELY-C OMP RES S ED
P LATES WITH S YMMETRIC AND ANTIS YMMETRIC D EF ORMATIONS . . 28
3-4 EF F EC T OF S YMMETRIC D AMAGE ON S TIF F NES S AND S TRENGTH OF
LONG P LATES U ND ER TRANS VERS E C OMP RES S ION. (A) MOD EL
AD OP TED IN ANALYS IS . (B) LOAD -S HORTENING C U RVES . . . . 30
3-5 LATERAL LOAD -D IS P LAC EMENT RELATIONS HIP F OR S TIF F ENED
P ANEL 31
3-6 C OMP ARIS ON OF LOAD -END S HORTENING C U RVE F OR LONG P LATES
(a /b=4) WITH OVERALL, P ERIOD IC AND LOC ALIZED D IS TORTION 33
3-7 TYP IC AL S -N D IAGRAMS F OR NOTC HED AND U NNOTC HED S TEEL
S P EC IMENS 35
3-8 U LTIMATE S TRENGTH VS "A" 3 6
3-9 S C HEMATIC S -N C U RVES OF INITIATION, P ROP AGATION, AND
TOTAL F ATIGU E LIVES 37
3-10 THREE MOD ES OF C RAC KING 40
3-11 RELATIONS HIP BETWEEN THIC KNES S AND F RAC TU RE TOU GHNES S . 4 0
3-12 RELATIONS HIP BETWEEN TEMP ERATU RE AND F RAC TU RE
TOU GHNES S 40
3-13 RELATIONS HIP BETWEEN LOAD ING RATE AND F RAC TU RE
TOU GHNES S 40
3-14 P LAS TIC ZONE AT C RAC K TIP 42
3-15 ARBITRARY C ONTOU R AROU ND A C RAC K TIP 4 6
3-16 S C HEMATIC S HOWING RELATION BETWEEN NOTC H-TOU GHNES S TES T
RES U LTS AND LEVELS OF F RAC TU RE P ERF ORMANC E F OR VARIOU S
LOAD ING RATES 46
3-17 S C HEMATIC REP RES ENTATION OF F ATIGU E-C RAC K GROWTH . . . . 49
3-18 S U MMARY OF C RAC K GROWTH RATE D ATA U ND ER RAND OM-S EQU ENC E
AND ORD ERED -S EQU ENC E LOAD F LU C TU ATION 52
3-19 HIS TORY OF F ATIGU E-C RAC K GROWTH 53
4-1 INTEGRATED AP P ROAC H F OR RES ID U AL S TRENGTH AS S ES S MENT . . 61
4-2 LOAD -D EF LEC TION C U RVE OF A P LATE P ANEL U ND ER LATERAL
LOAD . 68
4-3 F LOW D IAGRAM F OR AS S ES S ING ALLOWABLE P ANEL D EF ORMATION
(BAS ED ON F RAC TU RE MEC HANIC S ) 69
4-4 S C HEMATIC REP RES ENTATION OF S TRES S D IS TRIBU TIONS AC ROS S
S EC TION 75
4-5 F LAW D ES C RIP TION F OR C RAC K AS S ES S MENT 77
4-6 F AILU RE AS S ES S MENT D IAGRAMS 83
4-7 P ROC ED U RE F OR AS S ES S MENT OF C RAC KS 85
4-8 BEAM-C OLU MN ID EALIZATION OF S TIF F ENED P ANELS 88
4-9 S TRES S -S TRAIN C U RVES F OR P LATE ELEMENTS U ND ER
LONGITU D INAL TENS ION AND C OMP RES S ION 89
4-10 U LTIMATE S TRENGTH OF 'ENERGY C ONC ENTRATION' - C OMP ARIS ON
OF MAES TRO RES U LTS WITH LLOYD 'S LRP AS S P ROGRAM 99
4-11 S TRES S /TEMP ERATU RE EF F EC T ON C RITIC AL C RAC K S IZE . . . . 100
i i i
4-12 LOC ATION OF THE C RAC K ON S HIP A 102
4-13 F ATIGU E C OMP ARIS ON OF S HIP S A AND B 102
4-14 C RAC K GROWTH C OMP ARIS ON OF S HIP S A AND B 102
4-15 C RITIC AL RU P TU RE S TRAIN AS A F U NC TION OF INTERNAL D EF EC T
S IZE (IN WELD ) 104
4-16 MOBILE OF F S HORE P LATF ORM 107
4-17 LOAD -NORMALIZED D IS P LAC EMENT F OR EXP LOS ION IN MU D ROOM . 107
4-18 EIGHT LEGGED NORTH S EA P LATF ORM 108
4-19 LOAD VERS U S TRANS VERS E D IS P LAC EMENT OF TOP C ORNER NOD E
OF J AC KET S TRU C TU RE. (LC I - D IAGONAL WAVES ) 108
4-20 C RAC K GROWTH P RED IC TION 112
4-21 RES ID U AL S TRENGTH P RED IC TION 112
4-22 D AMAGE TOLERANC E EVALU ATION BY ANALYS IS 112
4-23 F LOW C HART S HOWING THE TAS KS P ERF ORMED BY THE C OMP U TER
P ROGRAM 'P ROF ' 115
5-1 P ROF ILE AND P LAN VIEW OF A TYP IC AL TANKER 125
5-2 MID S HIP S EC TION OF 85, 000 TON TANKER 12 6
5-3 TANK LOAD ING P ATTERNS 129
5-4 S C HEMATIC S HOWING ZONE A AND B OF A TANKER C ROS S -S EC TION 13 0
5-5 D ES C RIP TION OF D AMAGE F OR EXAMP LE 1 132
5-6 C RAC K GROWTH P ROF ILE, EXAMP LE 1 142
5-7 D ES C RIP TION OF D AMAGE F OR EXAMP LE 2 143
5-8 'Y' F OR C OMP ONENT WITH ED GE C RAC K U ND ER AXIAL TENS ION AND
BEND ING 14 6
5-9 C RAC K GROWTH P ROF ILE F OR EXAMP LE 2 148
5-10 D IS C RETIZATION OF C ROS S -S EC TION INTO GROS S P ANELS AND
HARD C ORNERS 150
5-11 U LTIMATE MOMENT C AP AC ITY OF 85, 000 TON TANKER 151
5-12 C RITIC AL LOC ATIONS ON THE TANKER 153
IV
LIST OF TABLES
Ta ble No . Eg e
2-1 TYP ES OF F AILU RE IN MARINE S TRU C TU RES 5
2-2 MATRIX OF C AS U ALTY D ATA 8
2-3 P ERC ENTAGE OF TOTAL F AILU RE BY S TRU C TU RAL C OMP ONENT AND
S HIP TYP E . . . 15
2-4 P ERC ENTAGE OF F AILU RE BY LOC ATION AND S HIP TYP E . . . . 16
2-5 P ERC ENTAGE OF F AILU RE BY F AILU RE MOD E AND S HIP TYP E . . 16
3-1 F AILU RE MOD ES OF LATERALLY LOAD ED GRILLAGE 26
3-2 D ES IGN F ATIGU E F AC TORS (F ATIGU E LIF E). NP D REGU LATIONS 39
4-1 NEW C ONS TRU C TION P LATE D EF ORMATION LIMITS 63
4-2A P RINC IP AL C HARAC TERIS TIC S OF S HIP S S U RVEYED 64
4-2B S HIP S U RVEY P LATE P ANEL LOC ATION 65
4-3 S HIP S U RVEY P LATE P ANEL D EF ORMATIONS 66
4-4 P ARTIAL S AF ETY F AC TORS F OR AS S ES S MENTS AT LEVEL 2 . . . 79
4-5 S TRENGTH P RED IC TIONS OF ENERGY C ONC ENTRATION 97
4-6 RES U LTS OF RES ID U AL S TRENGTH AS S ES S MENT OF A D RILLING
RIG 109
4-7 RES U LTS OF RES ID U AL S TRENGTH AS S ES S MENT OF A TYP IC AL
8-LEGGED NORTH S EA P LATF ORM 109
4-8 S AMP LE A-7P , C RITIC AL LOC ATIONS AND OP TIMU M INS P EC TION
INTERVALS 113
4-9 S AMP LE D C -10, MAINTENANC E REVIEW BOARD (MRB) REP ORT . . 113
5-1 LOAD C OMBINATIONS * 128
5-2 BEND ING MOMENT AND S TRES S ES AT S S L No . 8 13 3
5-3 S HORT TERM NORTH S EA S TRES S S P EC TRU M 139
5-4 BEND ING MOMENT AND S TRES S ES AT BL #24 144
5-5 RES ID U AL S TRENGTH F AC TORS OF D AMAGED TANKER 154
Nomenclature
pla t e s lendernes s ra t i o b/t V(a
y
/E)
7s / 7 / 7t pa rt i a l co effi ci ent s fo r s a fet y fa ct o r t rea t ment (s ee
t a ble 4. 4)
AC T a ppli ed s t res s ra ng e (N/mm
2
)
< S
C
cra ck t i p o peni ng di s pla cement (C TOD ) (mm)
< S
C R
cri t i ca l cra ck t i p o peni ng di s pla cement (o ns et o f
uns t a ble fra ct ure) (mm)
S
l
a ppli ed cra ck t i p o peni ng di s pla cement (mm)
AKj a ppli ed s t res s i nt ens i t y fa ct o r ra ng e (N mm"
3/2
)
AK
ms
ro o t mea n s qua re s t res s i nt ens i t y fa ct o r ra ng e (N mm"
3/2
)
AK
m
t h res h o ld s t res s i nt ens i t y fa ct o r ra ng e (N-mm"
3/2
)
AP
b
bendi ng s t res s ra ng e (N/mm
2
)
AP
m
membra ne s t res s ra ng e (N/mm
2
)
S
r
fra ct ure ra t i o us i ng C TOD pa ra met er
e s t ra i n
e
y
yi eld s t ra i n
6 ro t a t i o n
p pla s t i ci t y co rrect i o n fa ct o r
C T s t res s (N/mm
2
)
C Tj a ppli ed t ens i le s t res s (N/mm
2
)
C T
f
flo w s t res s (N/mm
2
)
C T
H
h o ri zo nt a l wa ve bendi ng s t res s (N/mm
2
)
C T net s ect i o n s t res s (N/mm
2
)
o
Ti
s t res s ra ng e co rres po ndi ng t o t h e i "
1
s ea -s t a t e (N/mm
2
)
cr
rma x
ma x i mum s t res s ra ng e (N/mm
2
)
C T
R
res i dua l s t res s (N/mm
2
)
o
ms
ro o t mea n s qua re s t res s ra ng e (N/mm
2
)
C T
U
ult i ma t e t ens i le s t reng t h (N/mm
2
)
C TVH h o g g i ng wa ve bendi ng s t res s (N/mm
2
)
o
vs
s a g g i ng wa ve bendi ng s t res s (N/mm
2
)
C T
y
yi eld s t reng t h (N/mm
2
)
< & co mplet e elli pt i c i nt eg ra l o f t h e s eco nd ki nd
vi
2a cra ck leng t h fo r t h ro ug h t h i cknes s cra cks (s ee fi g ure
4. 5) (mm)
2c cra ck leng t h fo r s urfa ce o r embedded fla w (s ee fi g ure
4. 5) (mm)
a leng t h o f pa nel o f pla t i ng (lo ng edg e) (mm)
A ma t eri a l co ns t a nt us ed fo r K
f
(mm)
a
0
i ni t i a l cra ck leng t h (mm)
a j h a lf leng t h o f t h e rea l cra ck, i ncludi ng t h e pla s t i c zo ne
(mm)
a
f
fi na l cra ck leng t h (mm)
m
, a
m
t o lera ble cra ck s i ze (mm)
b brea dt h o f pa nel o f pla t i ng (s h o rt edg e) (mm)
B s ect i o n t h i cknes s i n t h e pla ne o f t h ro ug h fla w (s ee
fi g ure 4. 5) (mm)
B
c
cro s s -s ect i o n ma rg i n
C co ns t a nt i n cra ck g ro wt h la w
d dept h o f t h e web plus t h e fla ng e t h i cknes s
D da ma g e ra t i o
da /dN cra ck g ro wt h ra t e (mm/cycle)
E Yo ung 's mo dulus o f ela s t i ci t y (N/mm
2
)
E
di m
P
dyna mi c i mpa ct energ y (MJ )
E
mp
s t a t i c i mpa ct energ y (MJ )
F pea k s t res s (N/mm
2
)
f*
r2
s t res s ra ng e due t o fluct ua t i ng pres s ure lo a di ng (N/mm
2
)
H eg ui va lent h a lf t h i cknes s
H
s i
s i g ni fi ca nt wa ve h ei g h t i n t h e i * s ea -s t a t e
H
s ma x ma x i mum s i g ni fi ca nt wa ve h ei g h t
I mo ment o f i nert i a a bo ut t h e h o ri zo nt a l N. A. (m
2
-cm
2
)
2
-c
r
"
2
'*
I
a
mo ment o f i nert i a a bo ut t h e vert i ca l N. A. (m
2
-cm
2
)
j
ma t
ma t eri a l t o ug h nes s mea s ured by J -met h o d (N/mm)
K s t res s i nt ens i t y fa ct o r (N mm"
3/2
)
K
c
equi va lent s t res s i nt ens i t y fa ct o r (N mm"
3/2
)
K
f
fa t i g ue s t reng t h reduct i o n fa ct o r
K, a ppli ed s t res s i nt ens i t y fa ct o r (N-mm'
3
'
2
)
Vll
K
IC
pla i n s t ra i n fra ct ure t o ug h nes s (N-mi rf
3
'
2
)
^mat,
ma t eri a l t o ug h nes s mea s ured by s t res s i nt ens i t y
(N-mm'
3
'
2
)
fa ct o r
K
r
fra ct ure ra t i o (K
r
/K
ma t
)
K,
s t res s co ncent ra t i o n fa ct o r
1 leng t h o f bea m
1;
h i ng e li ne
m ex po nent i n cra ck g ro wt h la w
M bendi ng mo ment (N-mm)
M
Y
full pla s t i c mo ment (N-mm)
M
D es
a llo wa ble des i g n co lla ps e mo ment (N-mm)
M
km/ M
kb
M # Mb/
s t res s i nt ens i t y fa ct o r ma g ni fi ca t i o n fa ct o rs
Mu/
M
U LT
ult i ma t e co lla ps e mo ment (N-mm)
N number o f cycles
N
f
number o f cycles t o fa i lure
"i
number o f cycles i n t h e i * ra ng e
N| number o f cycles fo r cra ck i ni t i a t i o n
N
P
number o f cycles fo r cra ck pro pa g a t i o n
P
a ppli ed la t era l pres s ure (N/mm
2
)
P a ppli ed lo a d (N)
P Y
full pla s t i c a x i a l fo rce (N)
p(D
f
pro ba bi li t y o f fa i lure o f a s t ruct ura l member
p(. )
f
pro ba bi li t y o f s ys t em fa i lure
P *
r
pro ba bi li s t i c mea s ure o f redunda ncy
P
b
pri ma ry bendi ng s t res s (N/mm
2
)
P m
pri ma ry membra ne s t res s (N/mm
2
)
Q
s eco nda ry s t res s (N/mm
2
)
Qb
s eco nda ry bendi ng s t res s (N/mm
2
)
Qm
s eco nda ry membra ne s t res s (N/mm
2
)
r no t ch ra di us (mm)
R s t res s ra t i o (a
mi n
/a
ma x
)
REF res erve res i s t a nce fa ct o r
RIF res i dua l res i s t a nce fa ct o r
vi i i
r
y
leng t h o f t h e pla s t i c zo ne (mm)
S s t res s ra ng e (N/mm
2
)
S ; i * s t res s ra ng e (N/mm
2
)
S M s ect i o na l mo dulus (m-cm
2
)
S
r
pla s t i c co lla ps e ra t i o (aja
{
)
t t h i cknes s o f pla t i ng (mm)
W pla ne wi dt h i n pla ne o f fla w (s ee fi g ure 4. 5) (mm)
w di s pla cement (mm)
Y fo rm fa ct o r (s t res s i nt ens i t y fa ct o r co rrect i o n)
IX
1 . 0 INTRODUCTION
Th e i ncrea s i ng dema nds bei ng pla ced o n ex i s t i ng ma ri ne s t ruct ures
a nd t h e t rends t o wa rd li g h t wei g h t s t ruct ures a nd li mi t s t a t e des i g n
requi re a t h o ro ug h unders t a ndi ng o f t h e res erve a nd res i dua l
s t reng t h ca pa ci t y o f s t ruct ures .
Hi s t o ri ca lly, i n t ra di t i o na l eng i neeri ng pra ct i ce, t h e ro le o f
res i dua l s t reng t h i s no t ex pli ci t ly defi ned. Ho wever, s i nce t h e
S h i p S t ruct ure C o mmi t t ee (S S C ) S ympo s i um i n 1983, wh i ch a ddres s ed
"Th e Ro le o f D es i g n, Ins pect i o ns , a nd Redunda ncy i n Ma ri ne
S t ruct ure Reli a bi li t y", event s i n t h e ma ri ne i ndus t ry h a ve po i nt ed
t o a reeva lua t i o n o f t h e ro le o f res i dua l s t reng t h i n t h e co nt ex t
o f s a fet y, a nd reli a bi li t y o f ma ri ne s t ruct ures . Th i s pro ject
t i t led "Res i dua l S t reng t h As s es s ment o f D a ma g ed Ma ri ne S t ruct ures "
a t t empt s t o i nt ro duce t h e s ubject t o t h e pra ct i ci ng eng i neeri ng
co mmuni t y a nd fo rm t h e ba s i s fo r co nt i nued fut ure wo rk.
S peci fi ca lly, t h i s pro ject wi ll i dent i fy ch a ra ct eri s t i c t ypes o f
da ma g e ba s ed o n revi ew o f s h i p ca s ua lt y reco rds a nd pres ent
s uppo rt i ng a na lyt i ca l met h o ds ca pa ble o f a s s es s i ng bo t h t h e lo ca l
i ndi vi dua l co mpo nent a nd g lo ba l s ys t em res i dua l s t reng t h . Th e
a bi li t y t o a s s es s bo t h i nt a ct a nd da ma g ed s t ruct ures wi ll lea d t o
t h e pro per ex plo i t a t i o n o f i nh erent s t ruct ura l redunda ncy,
i dent i fi ca t i o n o f des i red lo ca l a nd g lo ba l s a fet y levels a nd t h e
develo pment o f i ns pect i o n, ma i nt ena nce a nd repa i r pro cedures i n a
mo re ra t i o na l, co s t effect i ve a nd co mplet e ma nner t h a n h a s been
po s s i ble i n t h e pa s t .
1. 1 Ba ckg ro und
Two previ o us S h i p S t ruct ure C o mmi t t ee (S S C ) pro ject s ; S S C -3 54,
"S t ruct ura l Redunda ncy fo r D i s cret e a nd C o nt i nuo us S ys t ems " a nd
S S C -355, "Rela t i o n o f Ins pect i o n F i ndi ng s t o F a t i g ue Reli a bi li t y",
were co nduct ed wh i ch i nves t i g a t ed t h e ro le o f s t ruct ura l redunda ncy
a nd t h e reli a bi li t y o f current ma ri ne i ns pect i o n pro cedures ,
res pect i vely. Th e s i g ni fi ca nt co nclus i o ns o f t h es e pro ject s
i ncluded t h e fo llo wi ng :
Even s i mple s t ruct ures h a ve h i g h levels o f redunda ncy;
Th e rela t i o ns h i p bet ween res erve s t reng t h , res i dua l
s t reng t h a nd redunda ncy i s di ffi cult t o qua nt i fy, even
fo r s i mple ma ri ne s t ruct ura l s ys t ems ;
D eci s i o ns t o repa i r da ma g e fo und duri ng i ns pect i o ns a re
ba s ed upo n a pa rt i cula r i ns pect o r's ex peri ence ra t h er
t h a n a na lyt i ca l met h o ds .
Th i s S S C pro ject bui lds o n t h e co nclus i o ns o f t h es e t wo previ o us
pro ject s emph a s i zi ng res i dua l s t reng t h a s s es s ment o f da ma g ed ma ri ne
s t ruct ures due t o no rma l o pera t i ng lo a ds .
1. 2 Object i ves
Th e pri ma ry o bject i ves o f t h i s pro ject a re a s fo llo ws :
1. Int ro duce t h e s ubject o f res i dua l s t reng t h a s s es s ment o f
da ma g ed ma ri ne s t ruct ures .
2. S umma ri ze t h e s t a t e o f t h e a rt t ech no lo g y a nd met h o ds
a va i la ble i n t h e ma ri ne a nd no n-ma ri ne i ndus t ry fo r
qua nt i fyi ng res i dua l s t reng t h .
3. Reco mmend fut ure wo rk t o i nt eg ra t e current eng i neeri ng
pro cedures i n t h e a rea s o f cra ck g ro wt h , perma nent defo rma t i o n
a nd g lo ba l ult i ma t e s t reng t h t o a s s es s res i dua l s t reng t h o f
da ma g ed ma ri ne s t ruct ures .
Beca us e o f t h e i mpli ci t rela t i o ns h i p bet ween res i dua l s t reng t h a nd
i ns pect i o n pro cedures , a s eco nda ry o bject i ve o f t h i s repo rt i s t o
pres ent a n o ut li ne fo r a pra ct i ca l a na lyt i ca l pro cedure t o rea ct t o
det ect ed cra cks .
1. 3 Appro a ch
To a cco mpli s h t h e o bject i ves o f t h e pro ject , t h e fo llo wi ng a ppro a ch
wa s undert a ken:
Ta s k 1 - Li t era t ure Revi ew:
A li t era t ure revi ew wa s co nduct ed t o bench ma rk t h e s t a t e o f t h e a rt
i n res i dua l s t reng t h a s s es s ment . Tech ni ca l pa pers fro m bo t h t h e
ma ri ne a nd no n-ma ri ne i ndus t ri es were revi ewed t o i dent i fy t h e key
element s t h a t fo rm t h e ba s i s fo r res i dua l s t reng t h a na lys i s .
Ta s k 2 - C o llect i o n a nd Eva lua t i o n o f Ma ri ne S t ruct ure D a ma g e
D a t a :
Va ri o us da t a ba s es a nd h ull s urvey reco rds ma i nt a i ned by reg ula t o ry
bo di es , cla s s i fi ca t i o n s o ci et i es a nd o wners were s t udi ed t o
i dent i fy ch a ra ct eri s t i c fo rms o f da ma g e, ex t ent a nd s h i pbo a rd
lo ca t i o ns . Int ervi ews wi t h o wners a nd i ns pect o rs were h eld, t o g et
pra ct i ca l i nfo rma t i o n o n cri t eri a fo llo wed duri ng i ns pect i o n a nd
rea ct i o n t o da ma g es .
Ta s k 3 - Eva lua t i o n o f Va ri o us Met h o ds fo r As s es s i ng Res i dua l
S t reng t h :
C urrent met h o ds us ed by t h e ma ri ne a nd no n-ma ri ne i ndus t ri es , t o
a s s es s res i dua l s t reng t h wh i ch a re co ns i s t ent wi t h t h e fo rms o f
da ma g e i dent i fi ed i n Ta s k 2, were revi ewed a nd repres ent a t i ve ca s e
s t udi es were s umma ri zed. Bo t h i ndi rect a nd di rect met h o ds were
revi ewed. In t h e i ndi rect met h o d, a na lys i s i s ca rri ed o ut a t t h e
lo ca l co mpo nent level (cra cks a nd duct i le da ma g e) a nd ba s ed o n
da ma g e ex t ent a s s umpt i o ns i nt eg ra t ed upwa rd t o t h e g lo ba l s ys t em
level (ult i ma t e co lla ps e) us i ng s ervi ce li fe ex peri ence,
a ppro x i ma t e clo s ed fo rm s o lut i o ns , i ndus t ry reco g ni zed g ui deli nes
a nd t wo di mens i o na l ult i ma t e s t reng t h a na lys i s . In t h e di rect
met h o d, a na lys i s i s ca rri ed o ut us i ng t h ree di mens i o na l no n-li nea r
fi ni t e element met h o ds .
Ta s k 4 - Res i dua l S t reng t h Ana lys i s o f a Typi ca l Ta nker:
Th e Res i dua l S t reng t h o f a t ypi ca l s i ng le s ki n t a nker i s a s s es s ed
us i ng t h e i ndi rect met h o d. F ra ct ure a s s es s ment i s ba s ed o n
pro cedures o ut li ned i n Bri t i s h S t a nda rd Ins t i t ut e's publi s h ed
do cument , P D 6493: 1991 ent i t led "Gui da nce o n Met h o ds F o r As s es s i ng
t h e Accept a bi li t y o f F la ws i n F us i o n Welded S t ruct ures " [4. 16]*.
U lt i ma t e s t reng t h i s eva lua t ed U s i ng a n ult i ma t e s t reng t h a na lys i s
pro g ra m, U LTS TR [4. 1], develo ped by t h e U . S . Na vy.
Ta s k 5 - C o nclus i o ns a nd Reco mmenda t i o ns fo r F ut ure Wo rk:
Ba s ed o n t h e revi ew o f current i ndus t ry pra ct i ce, co nclus i o ns
reg a rdi ng li mi t a t i o ns a nd key a s s umpt i o ns a re ma de.
Reco mmenda t i o ns fo r fut ure res ea rch effo rt s a re pres ent ed.
* [ ] Reference a t t h e end o f s ect i o n 4. 0,
2. 0 COLLECTION AND EVALUATION OF MARINE STRUCTURE DAMAGE DATA
Th e purpo s e o f t h i s s ect i o n i s t o i dent i fy ch a ra ct eri s t i c fo rms o f
da ma g e, ca t eg o ri ze fo rms o f da ma g e a s ei t h er lo ca l o r g lo ba l, a nd
det ermi ne da ma g e lo ca t i o ns . Th i s i nfo rma t i o n i s o f i mpo rt a nce t o
t h i s pro ject beca us e t h e eng i neeri ng met h o ds a nd a na lyt i ca l
pro cedures i nves t i g a t ed t o a s s es s t h e res i dua l s t reng t h o f da ma g ed
s t ruct ures mus t be co ns i s t ent wi t h t h e fo rm o f da ma g e o bs erved.
Th e da t a co llect i o n pro ces s i nvo lved li t era t ure revi ews , i nt ervi ews
wi t h s h i p o wners a nd o pera t o rs a nd s ea rch es o f a va i la ble da t a ba s es .
C o llect i o n o f eno ug h releva nt da t a t o be a ble t o co nfi dent ly
co rrela t e da ma g e t ype a nd lo ca t i o n wi t h i ns pect i o n h i s t o ri es wa s
very di ffi cult . Th e ma jo ri t y o f t h e s h i p o wners a nd o pera t o rs
wo uld no t relea s e o r di s clo s e da ma g e reco rds o f t h ei r fleet . As a
res ult o f t h e li mi t ed a va i la ble da t a , o nly bro a d t rends co uld be
ca ut i o us ly es t a bli s h ed t h a t co rrela t e s h i p t ypes wi t h da ma g e fo rms ,
a nd lo ca t i o ns .
Th e i mpa ct t h a t da ma g e h a s o n t h e o vera ll res i dua l s t reng t h o f t h e
s t ruct ure i s a funct i o n o f i t s (i ) ex t ent , (i i ) mo de o f fa i lure a nd
(i i i ) rela t i ve s h i pbo a rd lo ca t i o n. Wh i le i s o la t ed s ma ll cra cks ,
a ls o kno wn a s "nui s a nce" cra cks , a t t h e t o e o f a bra cket o r a t a
cut o ut fo r a lo ng i t udi na l wi ll h a rdly a ffect t h e o vera ll s t reng t h
o f t h e s t ruct ure, a cra ck o f co ns i dera ble leng t h o n t h e ma i n deck
o r s i de s h ell ca n s eri o us ly a ffect t h e s t ruct ure's res i dua l
s t reng t h . Th e mo de o f fa i lure a ls o i nfluences t h e rema i ni ng
res i dua l s t i ffnes s o f t h e s t ruct ure. A s t ruct ure under duct i le
fa i lure, li ke buckli ng , us ua lly po s s es s es po s t -buckled s t reng t h
wh i ch a llo ws i t t o co nt i nue t o ca rry lo a d a ft er da ma g e.
C o nvers ely. , a s t ruct ure wi t h bri t t le fa i lure h a rdly po s s es s es a ny
res erve s t reng t h a nd t h e fa i lure ca n lea d t o t o t a l co lla ps e o f t h e
s t ruct ure. Va ri o us mo des o f fa i lure a re s h o wn i n t a ble 2-1. Th e
rela t i ve s h i pbo a rd lo ca t i o n o f da ma g e a ls o a ffect s t h e s t ruct ure's
res i dua l s t reng t h . F o r ex a mple, da ma g e i n t h e mi ddle ca rg o blo ck
ca n lea d t o g rea t er lo s s o f res i dua l s t reng t h t h a n t h o s e a t t h e
ends due t o t h e h i g h bendi ng mo ment s i n t h e mi ddle ca rg o blo ck.
2. 1 C o llect i o n o f D a t a
Th e fo rm o f da ma g e a nd i t s effect o n res i dua l s t reng t h a s s es s ment
a re t h e res ult o f va ri o us fa ct o rs a nd t h ei r i nt era ct i o n. In a n
effo rt t o co rrela t e t h es e fa ct o rs a nd i dent i fy t rends , wi t h o ut
g o i ng i nt o a ny ri g o ro us s t a t i s t i ca l eva lua t i o n, a t a s k wa s
undert a ken t o g a t h er a s much i nfo rma t i o n a s po s s i ble o n da ma g e t o
ma ri ne s t ruct ures . P ri ma ry s o urces o f da ma g e i nfo rma t i o n were t h e
i n-h o us e ca s ua lt y repo rt da t a ba s e ca lled C AS MAIN ma i nt a i ned by t h e
U . S . C o a s t Gua rd Ma ri ne Inves t i g a t i o n D i vi s i o n, G-MMI, a nd o t h er
h ull s urvey repo rt s prepa red by va ri o us t a nker o wners , wh i ch
i ncluded repo rt s o n s urveys requi red by t h e C ri t i ca l Area
Ins pect i o n P la ns (C AIP ). Th e s h i p ca s ua lt y repo rt s i n t h e U S C G
w
w
E H
U
E n
w
w
H
iJ
H
<
fa
fa
O
W
fa
CM
E H
I
w
CO
<
E H
TO
U c
4J O
Q.'-
0) +J
U Q.
u t_
C0 O
4J >.
O O)
C L-
V O ID
O < 0
to en >
Q.4-
O co
< / > o 4J
o o
L . 3
O) 3 "O
O
Ol^ C
.SJS
iA ID a
a o < u
OC - T3
g
3
a
c
co CJ
o o
a -o
a < o
c o
3 i
co a
c c
3
01 c
c -
c
o
' CO
o
U ( D
O
CO
.
01
2
Q.
' co
XI E
CO
Q.4->
0) O)
C Q.
oi < a
.c
o o 4J
> o
C 4-i 3
- t_ -a
o
> -*- c
' to o
CU 4-* 0>
c c a
ot o> AT
1CTJ
c
< u
3 Q. to
c
CO
a
Q)
c
o> a.
XI 01
Z2
01 (0
o
XI -
c
3
OT
4-i C
CD
o> c
co -
a. to
o E
L . 01
a >
V) o>
o o
3 4J
O
SI 10 >.
J2 T3 U
(0 0)'
[_ 4-* 4-
C H-
co o> 3
01 > U>
-g Ci
c c_ to
3 Q.-
Z&
(0
(U E
.*
XI c-
(0 o
2
3
.2'S
flj (D
> . O)
5 5
.a (A
4J .
a t_ < u
O CU L .
o a 3
10 O ( fl
C I- c
D au
3 O)
-a L .
a '
o
co" -
0) V)
C (-
-M
M CO
c
O '
U (0
+- CO
CO O 0)
CO (D
> UJ
L U a.
_i O
4-. a w
c o o
0) u a
> u
01
1- c XI
a. 3
T3
-- > -
( D ' 3
O
L. 10 x:
0> L. t/i
4J 0)
( 0 c c
E 01 o > -
01 01
> 4->
01 < 0
01
u
c -n
CD 3 Dl trt
c_ O 01
3 JC
O U
C 4J t- >.
Ill 10 O
en en
c
oi a
?U
c
5
CO '
2 n =
C AS MAIN da t a ba s e were i n a s t a nda rdi zed fo rma t . Th e co nt ent va ri ed
dependi ng o n t h e i ns pect o r ma ki ng t h e repo rt . In mo s t o f t h e ca s es
t h e ca us e o f da ma g e wa s i dent i fi ed ba s ed o n ex peri ence a nd
precedent s . C la s s i fi ca t i o n s o ci et i es requi re h ull s urveys t o be
co nduct ed a t reg ula r i nt erva ls . Th e h ull s urvey repo rt s o bt a i ned
fro m s o me o i l t a nker o wners /o pera t o rs were very pro fes s i o na lly
do cument ed. Th ey co nt a i ned t h e s ympt o ms o f t h e da ma g es t h a t were
o bs erved duri ng t h e h ull s urveys wi t h o ut a ny reference t o t h e
ca us es . A lo t o f releva nt i nfo rma t i o n wa s a va i la ble fro m t h e
C ri t i ca l Area Ins pect i o n P la ns prepa red by t h e Tra ns Ala s ka
P i peli ne S ervi ce (TAP S ) t a nker o pera t o rs . C AIP 's were i ns t i t ut ed
by t h e U S C G t h ro ug h t h ei r Na vi g a t i o n a nd Ves s el Ins pect i o n C i rcula r
(NVIC ) No . 15-91 a s a ma na g ement t o o l t h a t s erves t o t ra ck t h e
h i s t o ri ca l perfo rma nce o f a ves s el, i dent i fy pro blem a rea s , a nd
pro vi de g rea t er fo cus t o peri o di c s t ruct ura l ex a mi na t i o ns . Th e
C AIP 's o bt a i ned fro m t h e TAP S t a nker o pera t o rs were us ua lly fo r
di fferent cla s s es o f s h i ps ba s ed o n t h ei r s i zes . Wi t h i n ea ch cla s s
a ll t h e s h i ps were mo re o r les s s i mi la r wi t h res pect t o t h ei r
s t ruct ura l co nfi g ura t i o n. Th i s s i mi la ri t y wa s reflect ed i n t h e
da ma g e pa t t erns o f t h e s h i ps wh i ch clo s ely res embled ea ch o t h er.
But a t t h e s a me t i me t h ere were s o me i ns t a nces o f da ma g e wh i ch were
uni que t o ea ch s h i p. S i nce t h es e C AIP 's a re pri ma ri ly t h e
res po ns i bi li t y o f t h e o wners a nd o pera t o rs , mo s t o f t h e da ma g es
ment i o ned a re well res ea rch ed a nd h a ve been ra t i o na lly a na lyzed.
F ro m t h es e s o urces , 41 i ns t a nces o f da ma g e were ch o s en wh i ch h a d
co mplet e des cri pt i o n o f t h e da ma g e i n t erms o f lo ca t i o n, ca us e,
mo de o f fa i lure, ex t ent o f t h e da ma g e, et c. a nd were felt t o be o f
releva nce t o t h e pro ject .
F o r brevi t y a nd ea s e o f i dent i fi ca t i o n o f t rends , t h e i nfo rma t i o n
o n t h es e i ns t a nces o f da ma g es wa s put i n t h e fo rm o f a ma t ri x i n
t a ble 2-2, wh i ch i s des cri bed belo w.
2. 2 S umma ry o f C a s ua lt y D a t a
Th e ma t ri x i n t a ble 2-2, i s di vi ded i nt o t wo pa rt s . Th e fi rs t fo ur
co lumns g i ve a des cri pt i o n o f t h e ves s el a nd t h e la s t s even g i ve a
co nci s e des cri pt i o n o f t h e da ma g e. U nder des cri pt i o n o f t h e ves s el
t h e fi rs t co lumn i ndi ca t es t h e t ype o f ves s el a nd i t s a g e i n yea rs
wh en t h e da ma g e o ccurred. Th e co lumn t i t led "Ro ut e" g i ves a n
i ndi ca t i o n o f t h e s peci fi c bo dy o f wa t er i n wh i ch t h e s h i p
pri ma ri ly o pera t es . Th i s g i ves a n i ndi ca t i o n o f t h e s ea co ndi t i o ns
t h a t t h e ves s el enco unt ers duri ng i t s us ua l o pera t i o n. Th e nex t
co lumn g i ves t h e ph ys i ca l di mens i o ns o f t h e ves s el i n t erms o f i t s
leng t h , brea dt h a nd dept h . Th e fo urt h co lumn g i ves t h e
di s pla cement o f t h e ves s el. In ca s es wh ere di s pla cement s a re no t
a va i la ble, t h e dea dwei g h t o f t h e ves s el i s pro vi ded, i ndi ca t ed by
a "(1)" nex t t o t h e fi g ure. Th e la s t t wo co lumns g i ve a feel fo r
t h e s i ze o f t h e ves s el a nd fi nenes s o f t h e h ull fo rm. U nder a
pa rt i cula r t ype o f ves s el, t h e ca s es h a ve been pres ent ed i n a n
a s cendi ng o rder o f di s pla cement /dea dwei g h t .
Th e des cri pt i o n o f da ma g e co ns i s t s o f t h e ca us e o f da ma g e, lo ca t i o n
(s t ruct ura l co mpo nent a s well a s rela t i ve s h i pbo a rd lo ca t i o n),
ca t eg o ry o f fa i lure a nd mi s cella neo us co mment s . In s o me ca s es , t h e
ca us e o f da ma g e wa s det ermi ned a na lyt i ca lly a nd/o r by la bo ra t o ry
t es t i ng o f t h e da ma g ed s a mples duri ng po s t -da ma g e a na lys i s
co nduct ed by t h e cla s s i fi ca t i o n s o ci et i es o r t h e o wners /o pera t o rs .
In o t h er ca s es , t h e ca us e wa s det ermi ned fro m pra ct i ca l ex peri ence
a nd judg ment . In s o me ca s es t h e da ma g e ex t ends t h ro ug h a number o f
s t ruct ura l co mpo nent s , fo r ex a mple t h e ma i n deck pla t i ng , t h e deck
lo ng i t udi na ls , t h e s i de s h ell pla t i ng a nd s h ell lo ng i t udi na ls . In
s uch ca s es a ll t h e s t ruct ura l co mpo nent s a re ment i o ned i rres pect i ve
o f t h e o ri g i n o f da ma g e. Th e rela t i ve s h i pbo a rd lo ca t i o n o f t h e
da ma g e wa s bro a dly i dent i fi ed i n t erms o f i t s lo ng i t udi na l,
t ra ns vers e a nd vert i ca l lo ca t i o n. Lo ng i t udi na lly ea ch ves s el wa s
di vi ded i nt o fo rwa rd, ca rg o blo ck a nd a ft po rt i o ns . Th e ca rg o
blo ck h a s been furt h er di vi ded i nt o fo rwa rd, mi ddle a nd a ft
reg i o ns . Th e ves s el h a s been di vi ded i nt o t h ree reg i o ns
t ra ns vers ely, t wo o ut bo a rd reg i o ns a t t h e po rt a nd s t a rbo a rd s i de
a nd t h e reg i o n a ro und t h e cent erpla ne. No bro a d di vi s i o ns h a ve
been ma de vert i ca lly, h o wever vert i ca l lo ca t i o ns h a ve been
i dent i fi ed a s t h e ma i n deck level, ba s e level, wa t erli ne reg i o n,
t urn o f bi lg e, et c. Th e ex t ent o f da ma g e, wh erever a va i la ble, wa s
pro vi ded i n pa rent h es i s i n o ne o f t h e t h ree co lumns g i vi ng i t s
rela t i ve lo ca t i o n dependi ng o n t h e o ri ent a t i o n o f t h e cra ck o r
fra ct ure. F o r ex a mple, a 10' lo ng cra ck a t t h e s i de s h ell runni ng
vert i ca lly i s ment i o ned a s (10') under t h e co lumn t i t led
"vert i ca l. "
In t h e la s t co lumn, t i t led "co mment s ", t h e da ma g e wa s ca t eg o ri zed
a cco rdi ng t o t h e cla s s i fi ca t i o n met h o d pro vi ded i n t h e U . S . C o a s t
Gua rd NVIC 15-91, wh i ch des cri bes t h ree cla s s es o f s t ruct ura l
fa i lures a cco rdi ng t o t h e s i ze a nd t h e lo ca t i o n o f t h e fra ct ure.
Th es e a re:
C la s s 1 S t ruct ura l F a i lure
1. A fra ct ure o f t h e o i l/wa t ert i g h t envelo pe t h a t i s vi s i ble
a nd o f a ny leng t h o r a buckle t h a t h a s ei t h er i ni t i a t ed i n o r
h a s pro pa g a t ed i nt o t h e o i l/wa t ert i g h t envelo pe o f t h e ves s el;
o r
2. A fra ct ure 10 feet o r lo ng er i n leng t h t h a t h a s ei t h er
i ni t i a t ed i n o r pro pa g a t ed i nt o a n i nt erna l s t reng t h member.
S
s
o
u
I B <
-
z
a
Id ;*
ell
S < I
e -8
< "S 1
g2"
ofi -5
U u 5
5.3
.2 = 8
2 w
e I
[X- ffl
:
" i
a g
e
u.
u
"3 = '-> < - 2 =
3
"8
ES
U-SU.11.
u 3
D |
o o, e
S H o U
1
- .1
it
e I
u. +
si
3 -a
M
e s
e s
o < = a
J
u
3
u e =
u U. P3
5. u u
3 3 ;
"* J,
e
u e
u LL
<
o
o
J X
>
s 5
o
fi "E
a a
w
< 3
sr
B =
"" JS c~
g CO J2.
J
O _-
Z =
3 *
i7
< o -
.S
I?
.3 =
Q I
.S
w
a
s
a
I
so
3! S
1^
Q 3
*r ~CQ
a ^ "
a |
s < *
L
o
n
g
i
n
g
e
r
3 X
JjSi - -s
8 so <"
i
n
D
n
n
e
l
S
i
d
e
CO O
1 1
SH3
"> <
a 3 U
3
<
S .s a
Q
1
c I 8 S
n C U.
US,
J= -2 =
S &o x ao
so 3 "<
00
ill
_ to ^
.2 3
S
>
2 U
'8 S
1 s
1
dJ3
"5
3
TJ
0
z
a.
2
2 o
o
z
o
c
il.
u
S3
a
a
2 ~
D
O
H <
iH
I j
s !
o J
< =
O
> S
i s 3
3
> 3
: 1 8 I
o 5 n '
3 f= Ir
is s
3
2
2 3
,e g
M a
'5 'C
i2 0 5
8
3
U
e
u.
It
JB 3
a x
<
a
o
s
3
5
I
5 -i
E s
Si
<
5 .s
s ii
S
3 .a
sa
5 *
2 03
fS
3 5
< o
21
"3 CO
I
< u
S
u
5 g
6 J
a. u
8 s
o
PR
s , 3
s
3$
- <j
u o
u 3
3 E
11*
la&
Si 3
2 a
= 5
.3
1
5
2
" 5 o 3
S^
.s g
s u -
.3 J!
o ^>
s
a
H c
O U P U
q . S
3
3; 2
03
S J
3
a
X
o
33 u
0
*
a 'S
8
3 a .
i
"8
. 5
v>
D VJ
n
. b u
<
s > >
u.
H
S
u a
2
E >
s i
s a
8 a
^ J 5 5 J 5
> --s > > -s .
S x S
X X
5 3
g a
S 12
x
8 3
-= >
x*
E =
3 0
-J en
8:
z
2 o
a
>
3
o
S
1
SJ
|2
I
2
(5 - (5 ~
IS 15
oo 0\
u
5
s
o
u
it
>
u .
SB
2s
H U
J 5
o -
II
u .S
u m
- e *
1
Q
S -3
U
.S E 1
u u .9 q
i m ji J
-8
0
,
s
? a
+ .
3 >
-I
I 2
SfS
1
g
l^
\ o -
1
M I?
s
*
o
o
as
S CD 3 0
2 ,g
S <
O TS
Q 3
S =
N
a
3 ? *
g "8 3
5 ">
u 6
a a.
- 'S
IS
1
"2
S
fS
( N 0
s
I < o
'S sg
s!
3 8 . S S
0 o I z g
J2 *- 6 S
o O *J , C
ID ^
5 o
1 Ih
o u e
O 2 cj
2 J
s .2
.5
s
-a o
C/ 5 (O
* 7
u 2
Q
X
1 1
'if 8
a, g *
v> m
8-8
IM
3 2
>
3 |
"g E
" 8
8.
eg
> B
.s .s _
.S
" o
X
.S
1
O C
a'
e
o H o
O O
31
E|
11
st, I
1.3
e
m
w
j
3
<
u J3
=2*
8 H
a m
a V
Q
J:
c
a 2
Sit;
O V ) Q
.-3 *3 c
S J -s
S - -5
0. 3- u.
S g g
Q
o j;
U 5
S
I 1
a o
= 1
a I
||
1 1
a o r
O v. "^
R
1
S;z
2 o
a
4
H < i- C e- c.
3 2
1 0
2
Ul
S
s
o
u
> B
.3
0 el*
8 .Si to
3 *I
_ 5
op u
J
a
> -^
i
3
J e J3-?
o o _ n
.SP *c
. w
:
i
% a
s g
<*
S .
S?
a c >
Q I"
s
4
W
! 5 5
a o
ob
<
a
u
b
0
*
3
>
o
z
o
3
o
a
S
'1
ss
|2
5 ~
s
a:
e
Z SP
w JJ 'S
s 9
;< Sj
1 1 1 1 SI
3
a
.9
8
r; * '
If
U q
a 3
" E
S Sn=
a
a a
a
16
1 1
J CO
3 X
x
s a
i s
> CO
S x
X
16
IS
I*
8 x
x
1 J
i 3
3 *
5
3 -3 "
,
Q U.
3
Q
3
u
Q
3
a
3
Q
* o
1 3
15
<
z
<
z
<
z
<
z
Is
>
Is
(J '
l=-
u
1 1
C la s s 2 S t ruct ura l F a i lure
A fra ct ure les s t h a n 10 feet i n leng t h o r a buckle t h a t h a s
ei t h er i ni t i a t ed i n o r pro pa g a t ed i nt o a n i nt erna l s t reng t h
member duri ng no rma l o pera t i ng co ndi t i o ns .
C la s s 3 S t ruct ura l F a i lure
A fra ct ure o r buckle t h a t o ccurs under no rma l o pera t i ng
co ndi t i o ns t h a t do es no t o t h erwi s e meet t h e defi ni t i o n o f
ei t h er a C la s s 1 o r C la s s 2 s t ruct ura l fa i lure.
D efi ni t i o ns o f s o me o f t h e t erms us ed a bo ve a re:
(1)
(2)
(3)
o i l/wa t ert i g h t envelo pe: t h e s t reng t h deck, s i de
s h ell a nd bo t t o m pla t i ng o f a ves s el, i ncludi ng t h e
bo w a nd s t ern ra kes o f ba rg es .
i nt erna l s t reng t h members : t h e cent er vert i ca l
keel; deep web fra mes a nd g i rders ; t ra ns vers e
bulkh ea ds a nd g i rders ; s i de, bo t t o m a nd underdeck
lo ng i t udi na ls ; lo ng i t udi na l bulkh ea ds ; a nd bi lg e
keels .
buckle: a ny defo rma t i o n i n t h e o i l/wa t ert i g h t
envelo pe wh ereby t h e a djo i ni ng i nt erna l s t ruct ura l
members a re a ls o bent t o s uch a n ex t ent t h a t
s t ruct ura l s t reng t h h a s been lo s t .
In a ddi t i o n t o i dent i fyi ng t h e cla s s o f da ma g e, a ddi t i o na l
i nfo rma t i o n reg a rdi ng t h e ca us e o f da ma g e, h i s t o ry o f t h e ves s el,
ca us e o f cra ck a rres t , et c. h a s been pro vi ded wh ere a va i la ble.
2. 3 D i s cus s i o n o f F i ndi ng s
S o me bro a d t rends a nd ch a ra ct eri s t i cs h a ve been o bs erved fro m t h e
da ma g e i ns t a nces pres ent ed i n t a ble 2-2. Ba s ed o n t h e da t a i n
t a ble 2-2, va ri o us co rrela t i o ns bet ween fo rms o f da ma g e, ex t ent ,
lo ca t i o n a nd t h ei r i nt era ct i o n were a t t empt ed. S i nce t h e fo rm o f
fa i lure i s us ua lly det ermi ned ba s ed o n vi s ua l i ns pect i o ns a nd o t h er
ci rcums t a nt i a l evi dence, t h e a ccura cy depends o n t h e ex peri ence o f
t h e i ns pect o r. Th erefo re fo r t h e purpo s e o f t h i s repo rt a ll t ypes
o f fra ct ures , wh et h er fa t i g ue o r bri t t le, a re cla s s i fi ed a s cra cks .
Th e fo llo wi ng o bs erva t i o ns ba s ed o n revi ew o f t h e a va i la ble da t a
a re ma de:
1. Ta ble . 2-3 s h o ws t h e co rrela t i o n bet ween di s t ri but i o n o f da ma g e
a mo ng t h e va ri o us s t ruct ura l co mpo nent s a nd s h i p t ype.
Ho ri zo nt a lly t h e t a ble i s di vi ded i nt o t h ree, co rres po ndi ng t o
t h e t h ree bro a d cla s s es o f s t ruct ura l co mpo nent s , na mely;
pri ma ry lo ng i t udi na l, pri ma ry t ra ns vers e a nd s eco nda ry. Ea ch
o f t h e t h ree cla s s es co nt a i ns t h e co mpo nent s li s t ed i n t h e
12
t a ble belo w:
LETTER C OD ES U S ED F OR THE S TRU C TU RAL C OMP ONENTS
STRUCTURAL COMPONENT CODE
S i de S h ell Lo ng i t udi na l S L
Bo t t o m Lo ng i t udi na l
BL
D eck Lo ng i t udi na l D L
Ma i n D eck
MD
S i de S h ell
S S
Bo t t o m P la t e
BP
'Tween D eck
TD
Lo ng i t udi na l Bulkh ea d
LB
Lo ng i t udi na l Gi rder LG
Tra ns vers e Bulkh ea d
TB
Web F ra me
WF
Ho ri zo nt a l Gi rder
HG
Bra cket s
BK
F la t Ba r S t i ffeners
F B
Ot h ers (D o uble P la t es ,
OT
et c. )
Two s t a t i s t i cs fo r ea ch s t ruct ura l co mpo nent a re pres ent ed. Th e
fi rs t o ne i s t h e fra ct i o n o f t h e t o t a l da ma g e i n a pa rt i cula r
cla s s t h a t i s due t o t h a t co mpo nent , a nd t h e s eco nd o ne i s t h e
fra ct i o n o f t h e o vera ll da ma g e fo r t h e pa rt i cula r s h i p t ype. F o r
ex a mple, t h e number o f da ma g e i nci dences o ccurri ng i n s i de s h ell
lo ng i t udi na ls o f t a nkers i s 3 0% o f t h e number o f da ma g e i nci dences
i n pri ma ry lo ng i t udi na l members a nd 16% o f t h e t o t a l number o f
da ma g e i nci dences i n t a nkers . Th es e t wo numbers g i ve a feel fo r
t h e rela t i ve s i g ni fi ca nce o f t h e co mpo nent lo ca lly i n i t s cla s s a nd
o vera ll i n t erms o f t h e number o f da ma g e i nci dences i t s uffers .
Acco rdi ng t o t h e pres ent cla s s i fi ca t i o n o f t h e s t ruct ura l
co mpo nent s , t h e t a ble i ndi ca t es t h a t fo r a ll t h ree s h i p t ypes , t h e
pri ma ry lo ng i t udi na l co mpo nent s s uffer t h e mo s t number o f da ma g es ;
53% fo r t a nkers , 54% fo r bulk ca rri ers a nd 83% fo r co nt a i ners h i ps .
Ta nkers : Overa ll, t h e s i de s h ell lo ng i t udi na l (16%) a nd
s eco nda ry co nnect i ng s t ruct ures , bra cket (13%) a nd fla t ba r
s t i ffeners (10%) a re t h e co mpo nent s mo s t o ft en da ma g ed. Amo ng
pri ma ry lo ng i t udi na l members , t h e s i de s h ell lo ng i t udi na ls a re
mo s t o ft en da ma g ed (30%), fo llo wed by t h e ma i n deck pla t i ng
a nd a s s o ci a t ed lo ng i t udi na l (33%). Amo ng pri ma ry t ra ns vers e
members t h e web fra mes (55%) , due t o t h e pres ence o f la p
jo i nt s , a re t h e mo s t a ffect ed. Amo ng s eco nda ry co nnect i ng
s t ruct ures , bra cket s (50%) a cco unt fo r t h e mo s t number o f
13
da ma g e i nci dences . Ano t h er t rend t o be no t ed i s a co mpa ri s o n
o f t h e ra t i o s o f s i de s h ell lo ng i t udi na l t o s i de s h ell (30/17)
da ma g e a nd deck lo ng i t udi na l t o deck pla t e (13/20) da ma g e.
Th e fo rmer ra t i o i s much h i g h er. Th i s i s beca us e cra cks
o ri g i na t i ng a t t h e s i de s h ell lo ng i t udi na l a re us ua lly
repa i red befo re t h ey rea ch t h e s i de s h ell, wh erea s cra cks
o ri g i na t i ng o n t h e ma i n deck pla t i ng a re no t repa i red a s
qui ckly a nd t h erefo re pro pa g a t e a lo ng t h e deck, cra cki ng a ny
deck lo ng i t udi na l t h ey enco unt er.
Bulk C a rri er: Overa ll s i de s h ell pla t i ng s uffers t h e mo s t
(30%), fo llo wed clo s ely by web fra mes a nd co nnect i ng bra cket s
(23% ea ch ). Th i s fo llo ws clo s ely t h e rea s o ni ng put fo rwa rd i n
o bs erva t i o n number 1, t h a t t h e ma i n ca us e o f da ma g e i s s i de
s h ell flex i ng .
C o nt a i ners h i ps : Mo s t o ft en da ma g ed co mpo nent s a re t h e ma i n
deck (33%) a nd s i de s h ell pla t i ng (33%). Th e pres ent s urvey
do es no t s h o w a ny da ma g e t o s eco nda ry s t ruct ure, wh i ch i mpli es
t h a t fo r co nt a i ners h i ps , da ma g e t o s eco nda ry s t ruct ure i s
neg li g i ble.
2. Lo ca t i o n o f da ma g e pla ys a cri t i ca l ro le i n t h e res i dua l
s t reng t h a s s es s ment o f da ma g ed s t ruct ure. Th e o vera ll
s t ruct ura l res po ns e t o da ma g e i s a ffect ed by t h e lo ca t i o n a nd
ex t ent o f da ma g e. Ta ble 2-4 i ndi ca t es t h a t fo r t a nkers a nd
bulk ca rri ers mo s t o f t h e lo ng i t udi na l da ma g e o ccur i n t h e
mi ddle ca rg o blo ck, wh erea s fo r co nt a i ners h i ps t h e fo rwa rd
s ect i o n i s a ffect ed mo re. Tra ns vers ely, i n ca s e o f t a nkers ,
da ma g e i nci dences a re evenly di s t ri but ed a mo ng t h e po rt ,
s t a rbo a rd a nd t h e cent erpla ne a rea s , wh erea s i n t h e ca s e o f
bulk ca rri ers a nd co nt a i ners h i ps , da ma g e o ccur mo s t ly a t t h e
s i des . Vert i ca lly, ex cept fo r bulk ca rri ers wh i ch h a rdly s h o w
a ny da ma g e a t t h e bo t t o m, t a nkers a nd co nt a i ners h i ps s h o w a n
even di s t ri but i o n o f da ma g e a t a ll t h ree levels . Bulk
ca rri ers s h o w a g rea t er a mo unt o f da ma g e a t t h e ma i n deck
(62%) beca us e, i n a ddi t i o n t o t h e da ma g e t h a t o ccurs s o lely o n
t h e ma i n deck, much o f t h e da ma g e t h a t o ccurs a t t h e mi ddle
t h i rd dept h o f t h e s i de s h ell (38%) ex t ends t o t h e ma i n deck
level a ls o .
3. Th e predo mi na nt fo rm o f lo ca l fa i lure a mo ng a ll s h i p t ypes i s
fra ct ure, a s i ndi ca t ed i n t a ble 2-5 a nd i n [1, 2]*. Over t h e
yea rs ex t ens i ve res ea rch i nt o fa i lure due t o buckli ng under
co mpres s i ve lo a ds h a s res ult ed i n lo ca l s t ruct ure wh i ch i s
fa i rly res i s t i ve t o buckli ng . Th e few i ns t a nces o f buckli ng
t h a t a re o bs erved a re due ma i nly t o t h e reduct i o n o f
s ca nt li ng s due t o co rro s i o n [3, 4, 5] o r ex ces s i ve wa rpi ng o f
t h e cro s s decks , o n wi de-h a t ch ed s h i ps li ke bulk ca rri ers a nd
co nt a i ners h i ps , due t o h ull g i rder t o rs i o n [6].
* [ ] Reference a t end o f s ect i o n.
14
TABLE 2-3 P ERC ENTAGE OF TOTAL F AILU RE BY S TRU C TU RAL C OMP ONENT
AND S HIP TYP E
SHIP TYPE
TANKER(%)
BU LK C ARRIER(%)
C ONTAINER(%)
STRUCTURAL COMPONENT
(A) (B) (A) (B) (A) (B)
P RIMARY
(LONGITU D INAL)
TOTAT,
S L 30 16 11 6 -
BL 7 4 - - - -
D L 13 7 6 3 - -
MD 20 11 11 6 4 33
S S 17 9 55 30 4 33
BP 3 2 - -
2 17
LB 7 3 11 6 -
LG 3 2 - - -
TD - -
6 3 - -
100 53 100 54 100 83
P RIMARY
(TRANS VERS E)
TOTAL
TB 27 5 - -
100 17
WF 55 11 100 23 -
HG 18 4 - - - -
100 20 100 23 100 17
S EC OND ARY BL 50 13 100 23 - -
F B 36 10 - - - -
OT 14 4 - - - -
TOTAL 100 27 100 23 100
_
TOTAL
100
100
100
A: P ercent a g e o f F a i lure by C a t eg o ry
B: P ercent a g e o f To t a l F a i lure
15
TABLE 2-4 P ERC ENTAGE OF F AILU RE BY LOC ATION AND S HIP TYP E
SHIP TYPE
TANKER(%)
BU LK C ARRIER(%)
C ONTAINER(%)
LOCATION
LONGITU D INAL F WD 2
- 62
MID 82 100
38
AF T 16
-
-
TOTAL 100 100
100
TRANS VERS E P ORT 33
44 38
C ENTER 31
- 12
S TBD 36 56
50
TOTAL 100 100
100
VERTIC AL TOP 31 62
38
LWL 29 38 25
KEEL 40
- 37
TOTAL 100 100
100
TABLE 2-5 P ERC ENTAGE OF F AILU RE BY F AILU RE MOD E AND S HIP TYP E
SHIP TYPE
TANKER(%)
BU LK C ARRIER(%)
C ONTAINER(%)
FAILURE MODE
C RAC KING
100
65 88
BU C KLING
0
35 12
16
2. 4 C o nclus i o ns
Th e co mplex i nt era ct i o n t h a t ex i s t s bet ween t h e va ri o us fo rms o f
da ma g e, t h ei r lo ca t i o n a nd s h i p t ype, co mpo unded by t h e a mo unt o f
a va i la ble da t a , ma kes i t di ffi cult t o dra w ma ny defi ni t e
co nclus i o ns . Th erefo re, t h e t rends a nd pa t t erns s h o wn i n t h e
t a bles pres ent ed i n t h ei r s ect i o n a re o nly qua li t a t i ve a nd s h o uld
no t be t a ken i n a ny a bs o lut e qua nt i t a t i ve s ens e.
Th e s i g ni fi ca nt co nclus i o ns t h a t ca n be dra wn fro m t h e da t a
co llect ed i n t h i s s ect i o n o f releva nce t o i nves t i g a t i ng met h o ds t o
a s s es s t h e res i dua l s t reng t h o f da ma g ed ma ri ne s t ruct ures a re:
1. F ra ct ure i s t h e mo s t do mi na nt fo rm o f da ma g e t o s h i p
s t ruct ure. F a t i g ue cra cks h a ve no s eri o us effect s o n
s h i p's lo ng i t udi na l s t reng t h i f t h e cra ck leng t h i s les s
t h a n t h e cra ck leng t h wh i ch s t a rt s fa s t fra ct ure (cri t i ca l
cra ck leng t h ) a nd t h e t empera t ure i s no t ex t remely lo w.
2. D uct i le fa i lure i n t h e fo rm o f perma nent defo rma t i o n i s
qui t e co mmo n i n s h i p s t ruct ure a nd ex ces s i ve g ro s s
defo rma t i o n ca n lea d t o fra ct ure.
3. Ex t ent o f da ma g e i s lo ca li zed t o pa rt i cula r reg i o ns a nd
i ndi vi dua l co mpo nent s i n s h i p s t ruct ures a nd predo mi na nt
da ma g e lo ca t i o ns a re very much a funct i o n o f s h i p t ype.
17
2. 5 References
1. "S ervi ce Ex peri ence - S h i ps ", P ro ceedi ng s o f t h e 10t h
Int erna t i o na l S h i p a nd Offs h o re S t ruct ures C o ng res s , D enma rk,
1988.
2. Weber, P . F . , "S t ruct ura l S urveys o f Oi l Ta nkers ", Tra ns 1 Ma r
E Vo l. 96, P a per No . 65, 1984.
3. "Les s o ns Lea rned fro m F a i lures a nd D a ma g e o f S h i ps ", P res ent ed
a t t h e 8t h Int erna t i o na l S h i p S t ruct ures C o ng res s , Gda ns k,
1982.
4. Gro ve, T. W. , Rynn, P . G. a nd As h e, G. M. , "Bulk C a rri ers , A
C a us e fo r C o ncern", ABS Repo rt (D ra ft C o py).
5. "S t udy Repo rt o n Bulk C a rri er Lo s s ", J a nua ry 1992, Ni ppo n
Ka i ji Kyo ka i .
6. Ho ng , D . P . , et a l. , "Ana lys i s o f S t ruct ura l D a ma g e o f a La rg e
Ore/C o a l C a rri er". S NAME, S TAR S ympo s i um, S pri ng Meet i ng ,
Ma y 27-30, 1987.
18
3. 0 ELEMENTS OF RESIDUAL STRENGTH ASSESSMENTS OF MARINE STRUCTURES
3. 1 Int ro duct i o n
Res i dua l s t reng t h i s defi ned a s t h e ca pa bi li t y o f a s t ruct ure t o
co nt i nue t o ca rry lo a d a ft er da ma g e. Th e a bi li t y t o mea s ure t h e
s t ruct ure's res i dua l s t reng t h a ft er da ma g e a nd det ermi ne t h e
s t ruct ure's s ui t a bi li t y fo r co nt i nui ng s ervi ce, requi res a n
unders t a ndi ng o f t h e i nt errela t i o ns h i p bet ween t h e fo llo wi ng fo ur
element s :
1) des i g n ph i lo s o ph y o f t h e ex i s t i ng s t ruct ure
2) s t ruct ura l redunda ncy
3) t ypes o f da ma g e
4) i ns pect i o n pra ct i ces
Th e a bi li t y o f a s t ruct ure t o s us t a i n da ma g e, i t s res i dua l
s t reng t h , depends la rg ely o n t h e des i g n ph i lo s o ph y (des i g n s t reng t h
a nd ult i ma t e s t reng t h o f i nt a ct s t ruct ure) a nd t h e s t ruct ura l
redunda ncy wi t h i n t h e s ys t em. Eng i neeri ng met h o ds t o a s s es s t h e
res i dua l s t reng t h o f da ma g ed s t ruct ures a re dependent o n t h e t ypes
o f da ma g e, lo ca t i o ns o f da ma g e a nd o pera t i o na l pro fi le. D a ma g e t o
ma ri ne s t ruct ures ma y o ccur a t t h e lo ca l, o r co mpo nent level, a nd
a t t h e g lo ba l, o r s ys t em level. Th e da ma g e ma y be s ma ll a ffect i ng
o nly lo ca l co mpo nent s o r very s i g ni fi ca nt i n wh i ch t o t a l s t reng t h
i s lo s t . D eci s i o ns t o repa i r da ma g e a re a funct i o n o f t h e
eng i neeri ng ca lcula t i o ns requi red t o a s s es s t h e res 'i dua l s t reng t h
o f t h e da ma g ed s t ruct ure a nd t h e ex i s t i ng i ns pect i o n pra ct i ces .
3. 2 D es i g n P h i lo s o ph i es
S a fet y a nd reli a bi li t y a re t h e pri ma ry g o a ls o f a ny des i g n
ph i lo s o ph y. S a fet y a nd reli a bi li t y a re mea s ures o f a s t ruct ure's
a bi li t y t o res i s t lo a ds i n ex ces s o f t h e des i g n lo a d a nd t o s us t a i n
da ma g e wi t h o ut ca t a s t ro ph i c co lla ps e. D es i g n ph i lo s o ph i es wh i ch
a re emplo yed a nd i nco rpo ra t ed wi t h i n s t ruct ures s ubject ed t o cycli c
lo a di ng s ca n co mpri s e o ne o r bo t h o f t h e fo llo wi ng :
1. S a fe li fe des i g n ph i lo s o ph y.
2. F a i l s a fe des i g n ph i lo s o ph y.
S a fe li fe des i g n pro cedures a re ex pres s ed i n t erms o f o pera t i o na l
li fe t i me, o r number o f lo a d a ppli ca t i o ns duri ng wh i ch t h e
pro ba bi li t y o f fa i lure o f t h e ma ri ne s t ruct ure i s li kely t o be
ex t remely remo t e. Th i s ca n be es t i ma t ed ea rly i n t h e des i g n s t a g e
fro m fa t i g ue da t a a lrea dy a va i la ble o r o bt a i ned fro m la bo ra t o ry
t es t s . Th e s a fe li fe a ppro a ch , i n i t s pure fo rm, a s s umes t h a t t h e
s t ruct ure wo uld no t i ncur a ny da ma g e duri ng i t s li fe. Th i s ra i s es
19
t h e o bject i o n t h a t i t ma kes no a llo wa nce fo r da ma g e due t o
fa bri ca t i o n defect s . S t ruct ure i s des i g ned s uch t h a t cra ck
pro pa g a t i o n i s o nly remo t ely po s s i ble, h ence i s o ft en very
co ns erva t i ve. At t h e end o f t h e des i g n li fe t h e s t ruct ure i s
ei t h er s cra pped o r ex t ens i vely rewo rked, a s t h ere i s no
co ns i dera t i o n fo r peri o di c i ns pect i o n a nd repa i r i n t h e des i g n
ph i lo s o ph y.
F a i l s a fe des i g n ph i lo s o ph y es t a bli s h ed a fo rg i vi ng na t ure o f a
s t ruct ure t o da ma g e. It i s us ed t o des cri be a s t ruct ure des i g ned
t o wi t h s t a nd a cert a i n deg ree o f fa i lure a nd s t i ll ma i nt a i n
s uffi ci ent s t reng t h a nd s t i ffnes s i n t h e rema i ni ng pa rt o f t h e
s t ruct ure t o permi t co nt i nued us e unt i l t h e nex t i ns pect i o n peri o d.
F a i l s a fe des i g n beca me co mmo n pra ct i ce ea rly i n t h e develo pment o f
co mmerci a l a i rcra ft s t ruct ure. It i s ba s ed o n t h e fo llo wi ng
premi s es :
(a ) Ex i s t ence o f F la ws a nd C ra cks - due t o fa bri ca t i o n defect s ,
da y-t o -da y o pera t i o ns o r fa ult y s t ruct ura l des i g ns
(s peci fi ca lly s t ruct ura l det a i ls ).
(b) Ex i s t ence o f D a ma g e C o nt ro l P la n - wh i ch co ns i s t s o f:
(i ) S t ruct ura lly Ro bus t D es i g n - Ro bus t nes s i s deri ved fro m a
co mbi na t i o n o f res erve s t reng t h (ex ces s ca pa ci t y o f t h e
co mpo nent s ma ki ng up t h e t o t a l s t ruct ure), redunda ncy
(ex i s t ence o f a lt erna t i ve lo a d pa t h s i n t h e s t ruct ure t o
ca rry lo a di ng fro m da ma g ed co mpo nent s ) a nd duct i li t y
(a bi li t y t o s us t a i n la rg e pla s t i c s t ra i ns wi t h o ut
s i g ni fi ca nt lo s s i n s t reng t h ).
(i i ) D a ma g e Ana lys i s C a pa bi li t y - P ri ma ri ly a F ra ct ure
Mech a ni cs ba s ed a ppro a ch wh i ch a llo ws predi ct i o n o f cra ck
g ro wt h ra t e a nd res i dua l s t reng t h g i ven t h e i ni t i a l cra ck
leng t h , t h e ma t eri a l ch a ra ct eri s t i cs a nd t h e s t res s fi eld.
(i i i ) Ins pect i o n P ro g ra m - Th e i ns pect i o n pro g ra m s h o uld be a ble
t o i dent i fy t h e s ma lles t fla w. F o r effi ci ent i ns pect i o n,
kno wledg e o f cri t i ca l lo ca t i o ns s h o uld be a va i la ble. Th i s
co uld be ba s ed o n s t res s a na lys i s , t h e des i g ner's
ex peri ence o r previ o us reco rds . Th e i ns pect i o n i nt erva l
i s i mpo rt a nt a nd co uld be ba s ed o n fra ct ure mech a ni cs .
Th e wh o le s ucces s a nd reli a bi li t y o f fa i l s a fe des i g n
ph i lo s o ph y i s ba s ed o n t h e fa ct t h a t cra cks wi ll be
det ect ed befo re t h ey rea ch a cri t i ca l leng t h . Th e
s t ruct ure co uld be des i g ned s uch t h a t t h ere co uld be a
number o f a lt erna t i ve lo a d pa t h s but i f cra cks a re no t
det ect ed duri ng i ns pect i o n, befo re t h ey rea ch cri t i ca l
leng t h , t h ere co uld be a ca t a s t ro ph i ca l fa i lure.
20
P referred D es i g n P h i lo s o ph y - Alt h o ug h t h e fo rg i vi ng na t ure o f t h e
fa i l s a fe des i g n ph i lo s o ph y ma kes i t a n a t t ra ct i ve o pt i o n, i t s full
i mplement a t i o n i n t h e ma ri ne wo rld co uld be di ffi cult . A cri t i ca l
a s pect o f fa i l s a fe ph i lo s o ph y i s t h e det ect i o n o f cra cks a nd
fla ws . Wh i le t h i s ma y be po s s i ble i n t h e a i rcra ft i ndus t ry under
co nt ro lled envi ro nment a nd ea s y a cces s i bi li t y, i t i s di ffi cult i n
t h e ma ri ne i ndus t ry due t o t h e i mmens e s i ze o f t h e s t ruct ure a nd
t h e h a rs h envi ro nment t h e s urveys a re a cco mpli s h ed i n. Typi ca lly
a mo dern VLC C o r U LC C i nvo lves 100 t o 200 a cres o f s t ruct ura l s t eel
s urfa ce a nd 1, 000 t o 2, 000 mi les o f weldi ng . Th erefo re ch a nces o f
cra cks a nd fla ws g o i ng unno t i ced a re qui t e h i g h . C o ns equent ly,
s h i p s t ruct ura l des i g n i s ba s ed o n a co mbi na t i o n o f s a fe li fe a nd
fa i l s a fe des i g n pri nci ples a nd a s s es s ment o f res i dua l s t reng t h i s
ba s ed o n fa i l s a fe pri nci ples .
3. 3 Redunda ncy
Redunda ncy i mpli es t h e a bi li t y o f a s t ruct ure t o "s us t a i n o verlo a d"
o r t h e ex i s t ence o f "a lt erna t i ve lo a d pa t h " i n a s t ruct ure. Hence
redunda ncy ca n be ex pres s ed i n t wo wa ys : (i ) res erve s t reng t h a nd
(i i ) res i dua l s t reng t h . Res erve s t reng t h i s ' t h e ma rg i n bet ween t h e
dema nd i mpo s ed by t h e lo a d a nd t h e ult i ma t e ca pa ci t y o f t h e
s t ruct ure. It i s due t o t h e co ns erva t i s m i n t h e des i g n o f
i ndi vi dua l co mpo nent s a nd ult i ma t ely t h e ens emble o f co mpo nent s
ma ki ng up t h e s t ruct ure. Res i dua l s t reng t h refers t o t h e s a fet y o f
t h e s t ruct ure a g a i ns t fa i lure a ft er da ma g e h a s o ccurred. If a ll
t h e co mpo nent s t h a t ma ke up t h e s ys t em a re no t fully s t res s ed under
a g i ven lo a d, t h ere i s a po t ent i a l fo r s t res s redi s t ri but i o n upo n
t h e fa i lure o f o ne o r mo re co mpo nent s a nd prevent i o n o f co mplet e
fa i lure o f t h e s t ruct ure. Va ri o us met h o ds t o mea s ure redunda ncy
fo r di s cret e a nd co nt i nuo us s t ruct ure a re a va i la ble [1-5]
1
but no ne
o f t h em h a s been wi dely us ed. F o r t h e pres ent s t udy, det ermi ni s t i c
mea s ures o f res erve s t reng t h a nd res i dua l s t reng t h , a s pres ent ed i n
[1] a re us ed. In [1] res erve s t reng t h i s defi ned i n t erms o f
Res erve Res i s t a nce F a ct o r (REF ) g i ven by t h e ra t i o o f t h e ult i ma t e
s ys t em s t reng t h t o t h e des i g n lo a d.
Envi ro nment a l Lo a d At C o lla ps e (U nda ma g ed)
REF =
D es i g n Envi ro nment a l Lo a d
Res i dua l s t reng t h i s g i ven i n t erms o f Res i dua l Res i s t a nce F a ct o r
(RIF ) wh i ch i s defi ned a s t h e ra t i o o f t h e res i dua l s ys t em s t reng t h
t o t h e s ys t em co lla ps e s t reng t h .
RIF
Envi ro nment a l Lo a d At C o lla ps e (D a ma g ed)
Envi ro nment a l Lo a d At C o lla ps e (U nda ma g ed)
'Numbers in [ ], refer to references at the end of the section.
21
F i ndi ng a n a ppro pri a t e pro ba bi li s t i c mea s ure o f redunda ncy i s a
mo re co mplex i s s ue. One wa y t o mea s ure t h e a bi li t y o f a s t ruct ure
t o t o lera t e da ma g e i s by t h e co ndi t i o na l pro ba bi li t y o f s ys t em
s urvi va l g i ven t h a t o ne o f i t s members h a s fa i led. Th e redunda ncy
mea s ure h a s been us ed i n s evera l s t udi es [4, 5].
p
P
t
= 1 -
P f
0)
wh ere,
P *
r
i s t h e mea s ure o f redunda ncy
P
f
(s )
i s t h e pro ba bi li t y o f s ys t em fa i lure
P
f
(1)
i s t h e pro ba bi li t y o f fa i lure o f a ny o ne s t ruct ura l member
a nd P
f
(s )
< P
f
(1)
In ma ri ne s t ruct ures , t h e pres ence o f numero us fa i lure pa t h s ma ke
i t pra ct i ca lly i mpo s s i ble t o a ccura t ely det ermi ne t h e pro ba bi li s t i c
mea s ure o f redunda ncy. Th erefo re t o li mi t t h e number o f fa i lure
pa t h s , t h e a ppro a ch a do pt ed bo t h i n t h e det ermi ni s t i c a nd
pro ba bi li s t i c met h o ds , i s t o i dent i fy cri t i ca l element s o f t h e
s ys t em wh o s e fa i lure lea ds t o ca t a s t ro ph i c co ns equences .
Va ri o us cla s s i fi ca t i o ns o f redunda ncy ca n be fo und i n li t era t ure.
F o ur o f t h em a re pres ent ed belo w. Llo yd [1] s ug g es t s a member
redunda ncy h i era rch y fo r i ndet ermi na t e s t ruct ures . Th e levels o f
t h i s h i era rch y va ry fro m 0 t o 5. Members belo ng i ng t o level 0 a re
t h e lea s t redunda nt wh i le t h o s e a t level 5 a re t h e mo s t redunda nt .
Kli ng muller [6], i n t h e co nt ex t o f "s ys t ems " reli a bi li t y,
i dent i fi ed s t ruct ures a s h a vi ng ei t h er "a ct i ve" (h o t ) o r "s t a ndby"
redunda ncy. Act i ve redunda ncy refers t o co mpo nent s wh i ch a re fully
a ct i ve i n no rma l res po ns e a nd ca n be us ed t o ma i nt a i n s t a bi li t y
a ft er fa i lure o f a pa ra llel co mpo nent but o t h erwi s e a re a ddi t i o na l,
unneces s a ry co mpo nent s . S t a ndby redunda ncy refers t o a ddi t i o na l
co mpo nent s wh i ch a re no t us ed i n no rma l res po ns e perfo rma nce but
h a ve t o repla ce co mpo nent s t h a t h a ve fa i led. F o r co nt i nuo us
s t i ffened pla t e ba s ed s t ruct ures , [7] s ug g es t s a t h ree level
h i era rch y. An ex a mple fo r a mult i -co lumned s emi -s ubmers i ble i s a s
fo llo ws :
"Tert i a ry" rela t i ng t o i ndi vi dua l s t i ffeners o n a pa nel,
"S eco nda ry" rela t i ng t o a n i ndi vi dua l s t i ffened pa nel, a nd
"P ri ma ry" rela t i ng t o a n a s s embla g e o f s t i ffened pa nels
fo rmi ng a cyli ndri ca l co lumn.
22
D ependi ng o n t h e s t ruct ura l co nfi g ura t i o n va ri o us levels o f
redunda nci es wo uld be ex h i bi t ed by a s t ruct ure. F o r ex a mple, a
mult i -co lumn s emi -s ubmers i ble wi t h uns t i ffened co lumns ex h i bi t s
o nly pri ma ry redunda ncy. A mo no h ull wi ll ex h i bi t o nly s eco nda ry
a nd t ert i a ry redunda ncy wh i le a S WATH s h i p wi t h s i ng le o r t wi n
s t rut s wo uld ex h i bi t pri ma ry, s eco nda ry, a nd t ert i a ry redunda ncy.
S t i a ns en [8] h a s cla s s i fi ed redunda nci es i nt o t wo ca t eg o ri es , lo ca l
a nd g lo ba l. Lo ca l redunda ncy refers t o lo ca l res erve s t reng t h
wh i ch ex i s t s i n i ndi vi dua l members a nd jo i nt s . Glo ba l redunda ncy
refers t o t h e o vera ll s t ruct ure.
3. 4 Types o f D a ma g e a nd F a i lure
As s t a t ed i n s ect i o n 2. 0, t o ma ke a n a s s es s ment o f res i dua l
s t reng t h o f da ma g ed ma ri ne s t ruct ure i t i s i mpera t i ve t o
i nves t i g a t e t h e va ri o us t ypes o f da ma g e a nd res ult i ng fo rms o f
fa i lure, bo t h lo ca lly a nd g lo ba lly.
Th e va ri o us t ypes o f da ma g e i nclude cra cki ng , buckli ng , ex ces s i ve
defo rma t i o n a nd yi eldi ng o f t h e cro s s -s ect i o n. D a ma g e t o lo ca l
co mpo nent s ca n lea d t o ei t h er duct i le o r fra ct ure fa i lure.
Glo ba lly t h e fa i lure o f a lo ca l co mpo nent ca n h a ve o ne o f t wo
g enera l effect s :
t h e s t ruct ure res po nds wi t h a redi s t ri but i o n o f i nt erna l
fo rces dependi ng o n t h e fo rm o f lo ca l fa i lure a nd t h e
level o f redunda ncy o r
t h e da ma g e preci pi t a t es pro g res s i ve fa i lure wh i ch res ult s
i n lo s s o f o vera ll s t i ffnes s a nd lo a d ca rryi ng a bi li t y.
Wh i le lo ca l duct i le fa i lure i nvo lves g ro s s defo rma t i o n i n t h e fo rm
o f yi eldi ng o r buckli ng a nd res ult s i n g ra dua l redi s t ri but i o n o f
lo ca l i nt erna l fo rces , lo ca l fra ct ures do no t i nvo lve a ny
s i g ni fi ca nt defo rma t i o n but res ult i n ra pi d redi s t ri but i o n o f lo ca l
i nt erna l fo rces .
In a n a t t empt t o h i g h li g h t t h e funda ment a l fo rms o f da ma g e a nd
pro vi de eng i neeri ng s o lut i o ns t h i s repo rt wi ll co ncent ra t e o n jus t
duct i le fa i lure a nd fra ct ure. F i g ure 3-1 s h o ws a t ypi ca l duct i le
fa i lure o f a s t ruct ura l co mpo nent a nd co mpa res t h a t wi t h a bri t t le
fa i lure s cena ri o .
3. 4. 1 D uct i le F a i lure:
F i g ure 3-2 s h o ws t h e va ri o us ca t eg o ri es o f duct i le fa i lure.
D uct i le fa i lure ca n i nvo lve ela s t i c buckli ng , i nela s t i c buckli ng ,
pla s t i c co lla ps e o r a n i nt era ct i o n o f a ll t h ree. Ea ch o f t h es e
fa i lure mo des , fo r s i mple co mpo nent s li ke bea ms a nd pla t es , h a s
been wi dely di s cus s ed i n rea di ly a va i la ble li t era t ure [9, 10].
Beca us e o f t h e co ns equences o f g ro s s defo rma t i o n o n s h i p s t ruct ure,
a nd t h e s i g ni fi ca nce o f perma nent defo rma t i o n [11], t o repa i r
23
00
00
L U
er
i
00
STRAIN
(A) - COMPLETE FAILURE, ' BRITTLE ' SYSTEM
- STRUCTURE EITHER CRACKED THROUGH OR COMPLETE FAILURE
oo
00
L
a:
i
oo
G
ULTIMATE (
or
Y )
STRAIN
(B) - DUCTILE FAILURE
- FOLLOWING COLLAPSE IN COMPRESSION, RESIDUAL COMPRESSIVE STRENGTH
FIGURE 3.1 FAILURE OF BRITTLE AND DUCTILE SYSTEMS
7)
24
00
J
Q
O
J
(T
3
_l
<
_
<
3
h -
O
3
er
h -
o o
o o
3
o
er
<
>
IN
J
er
O
25
deci s i o ns , t h i s repo rt wi ll fo cus o nly o n t h e duct i le fa i lure a s a
res ult o f pla s t i c co lla ps e due t o la t era l lo a di ng s . Th e s h i p's
h ull s t ruct ure i s a co mplex co mbi na t i o n o f pla t i ng , lo ng i t udi na l
s t i ffeners a nd s uppo rt i ng fra mes . P la t i ng t o g et h er wi t h s t i ffeners
ma ke up a s t i ffened pa nel. Th e s t i ffened pa nel t o g et h er wi t h t h e
s uppo rt i ng fra mes co mpo s e a g ri lla g e s t ruct ure. F o r a g ri lla g e
s t ruct ure s ubject ed t o la t era l lo a ds t h ere a re t h ree ba s i c levels
o f lo a d a nd res po ns e:
1. Th e pla t i ng bet ween t h e s t i ffeners deflect s a cco rdi ng t o i t s
s t i ffnes s a nd t ra ns fers t h e lo a d t o t h e s t i ffeners .
2. Th e s t i ffeners a lo ng wi t h a n effect i ve wi dt h o f pla t i ng a ct a s
bea ms t o ca rry t h e lo a d t o t h e s uppo rt i ng fra mes .
3. Th e s uppo rt i ng fra mes yi eld a nd t h e g ri lla g e deflect s
t ra ns ferri ng t h e lo a d t o t h e bulkh ea ds a nd decks .
Th e ba s i c mo des o f fa i lure fo r a la t era lly lo a ded g ri lla g e
s t ruct ure a re s umma ri zed i n t a ble 3-1.
TABLE 3-1 F AILU RE MOD ES OF LATERALLY LOAD ED GRILLAGE
U ns ervi cea bi li t y U lt i ma t e F a i lure
pla t i ng una ccept a ble a mo unt
o f perma nent s et
rupt ure o r la rg e
a mo unt o f perma nent
s et
s t i ffeners pa rt i a l yi eld o f
s t i ffener fla ng e o r
una ccept a ble a mo unt
o f perma nent s et
co lla ps e due t o
fo rma t i o n o f pla s t i c
h i ng es
g ri lla g e una ccept a ble
perma nent s et
yi eld o f s uppo rt i ng
fra mes
In g enera l, s t i ffened pa nels t h a t fa i l i n a duct i le ma nner
g enera lly do s o t h ro ug h a s eri es o f pro g res s i ve s t a g es :
t h e o ns et o f no n-li nea r res po ns e, wh ere t h e lo a d ca rryi ng
ca pa bi li t y beg i ns t o i ncrea s e a t a ra t e wh i ch i s les s t h a n t h e
ra t e o f s t ra i n o r deflect i o n.
t h e develo pment o f a pla t ea u, t h a t i s a lmo s t a co ns t a nt lo a d
ca rryi ng ca pa bi li t y reg a rdles s o f deflect i o n.
a po s t ult i ma t e s t reng t h lo a d deflect i o n curve, wh i ch wi ll
g enera lly be dependent upo n t h e t ype o f co mpo nent , a nd wh i ch
wi ll o ft en s h o w a n unlo a di ng t rend.
26
3. 4. 1. 1 S t reng t h o f D a ma g ed S t ruct ure U nder La t era l Lo a d
Th e predo mi na nt fo rm o f da ma g e ca us ed by co lli s i o ns wi t h pi ers a nd
h ydro dyna mi c i mpa ct i s la t era l pla s t i c defo rma t i o ns (due t o pla s t i c
co lla ps e) o f s t i ffened pla t e pa nels fo rmi ng t h e s h ell o r deck o f a
s h i p's h ull s t ruct ure. It i s i mpo rt a nt t o no t e t h a t t h i s fo rm o f
da ma g e i s us ua lly fo und i n t h i n pla t ed, s t ro ng ly s t i ffened
s t ruct ure t ypi ca l o f co mba t a nt h ull s t ruct ura l s ys t ems a nd g i ves a
"h ung ry h o rs e" a ppea ra nce t o t h e s h i p h ull, wh ereby t h e pla t i ng
defo rms s ymmet ri ca lly bet ween s t i ffeners . In g enera l t h e perma nent
defo rma t i o n o f s t i ffened pla t e pa nels a ffect s t h e res i dua l s t reng t h
ch a ra ct eri s t i cs o f t h e s t i ffened pla t e pa nel.
P o s s i ble co ns equences o f s uch perma nent defo rma t i o ns a re [12]:
a reduct i o n i n s t i ffnes s a nd s t reng t h o f t h e pa nels under
s ubs equent la t era l lo a ds ; but i t h a s been g enera lly fo und t h a t
s uch reduct i o ns a re i ns i g ni fi ca nt .
a lo s s o f lo ca l s t i ffnes s a nd s t reng t h under i n-pla ne lo a ds
a nd s ubs equent reduct i o n i n ult i ma t e h ull g i rder bendi ng
s t reng t h .
D a t a o n rect a ng ula r pla t es o f a s pect ra t i o (a /b) g rea t er t h a n 2 a nd
s lendernes s ra t i o (b/t ) les s t h a n 60, s ug g es t t h a t t h e lo ng i t udi na l
co mpres s i ve s t reng t h i s o nly s li g h t ly reduced wh en s ubject ed t o
s i ng le lo be da ma g e. Th e effect o f la t era l pres s ure [13] a nd
t h erefo re pro ba bi li t y o f da ma g e defo rma t i o n i s mo re pro mi nent i n
s lender pla t es (b/t > 80).
D a ma g e o f t h e "h ung ry h o rs e" t ype, i s li kely t o i nduce mo re
s i g ni fi ca nt lo s s o f s t i ffnes s a nd s t reng t h under i n-pla ne lo a d
a ppli ed i n t h e s h o rt er o r t ra ns vers e di rect i o n, ca us i ng :
(a ) a reduct i o n o f h ull-g i rder bendi ng s t reng t h i n wa y o f
t ra ns vers ely fra med bo t t o m s h ell a nd deck s t ruct ure, a s
co mmo nly emplo yed i n t h e fo re a nd a ft reg i o ns o f a s h i p's
h ull;
(b) lo s s o f effect i ve flex ura l s t i ffnes s i n t ra ns vers e fra mes
wh ere t h e s h ell pla t i ng a ct s a s a fla ng e t o t h e fra mes ,
no t a bly i n do uble bo t t o m s t ruct ures .
As i llus t ra t ed i n fi g ure 3-3 [12] t h e beh a vi o r o f a lo ng
rect a ng ula r pla t e under t ra ns vers e co mpres s i o n depends ma i nly o n
t h e rela t i ve fo rm o f i ni t i a l defo rma t i o n i n t h e pla t e pa nel. Much
g rea t er lo s s o f s t i ffnes s a nd s t reng t h o ccurs wh en t h e di s t o rt i o ns
a re a nt i s ymmet ri c t h a n wh en di s t o rt i o ns a re a ppro x i ma t ely
s ymmet ri c, i . e. o f t h e "h ung ry h o rs e" fo rm. In t h e ca s e o f
s ymmet ri c di s t o rt i o n, i ni t i a l defo rma t i o ns g ro w s ymmet ri ca lly unt i l
t h e buckli ng lo a d i s a lmo s t a t t a i ned, a t wh i ch po i nt t h e pa nel
s na ps i nt o a n a nt i s ymmet ri c co nfi g ura t i o n. In t h e ca s e o f
27
0.15
0.1
>
bb
0.05
01 = <5 02(SYM)
6 02 = - C5
0
1 (ANTISYM)
/ -
FIGURE 3.3 LOAD-SHORTENING CURVES FOR LONG, TRANSVERSELY-
COMPRESSED PLATES WITH SYMMETRIC AND ANTISYMMETRIC
DEFORMATIONS
12)
28
a nt i s ymmet ri c i ni t i a l di s t o rt i o n, defo rma t i o n rema i ns a nt i s ymmet ri c
t h ro ug h o ut t h e lo a d ra ng e a nd fa i lure o ccurs g ra dua lly but wi t h a
much g rea t er reduct i o n o f s t i ffnes s a nd s t reng t h . It i s clea r fro m
t h es e s cena ri o s t h a t co rrect eva lua t i o n o f ch a ra ct eri s t i c pla t e
s t reng t h requi res s t a t i s t i ca l a s s es s ment o f bo t h pla t e di s t o rt i o n
a mpli t udes a nd t h e rela t i ve fo rm o f di s t o rt i o n i n t h e pa nels .
D a ma g e ca us ed by h ydro dyna mi c o verlo a d wi ll t end t o i nduce t h e
s ymmet ri c, mo re fa vo ra ble fo rm o f defo rma t i o n. Th i s effect h a s
been a ppro x i ma t ely eva lua t ed, fo r t h e ca s e o f a lo ng s t eel pla t e
(la rg e a /b) by ex a mi ni ng t h e beh a vi o r o f a t ra ns vers e s t ri p o f
pla t i ng s ubject ed t o :
(a ) a ppli ca t i o n a nd remo va l o f la t era l pres s ure ca us i ng
ela s t o -pla s t i c defo rma t i o n a nd a s s o ci a t ed res i dua l
s t res s es ;
(b) t ra ns vers e co mpres s i o n up t o a nd beyo nd co lla ps e.
Res ult s o f a s t udy pres ent ed i n reference [12] a re s umma ri zed i n
fi g ure 3-4 i n t h e fo rm o f co mpres s i ve lo a d s h o rt eni ng curves fo r
va ri o us ra t i o s o f b/t a nd levels o f perma nent defo rma t i o n. It i s
evi dent t h a t fo r s lender pla t es i n wh i ch t h e ela s t i c buckli ng
s t res s i s s ubs t a nt i a lly lo wer t h a n t h e yi eld s t res s , da ma g e ma y
ca us e s i g ni fi ca nt lo s s o f s t i ffnes s but h a s li t t le effect o n pea k
lo a d ca pa bi li t y. C o nvers ely, i n s t o cky pla t es , wh i ch buckle
ela s t o -pla s t i ca lly, da ma g e wi ll ca us e lo s s o f bo t h t ra ns vers e
s t i ffnes s a nd s t reng t h .
Mo re g enera l da ma g e i nvo lvi ng ela s t o -pla s t i c bendi ng o f s t i ffeners
ma y a ls o res ult fro m la t era l lo a ds . A t ypi ca l la t era l lo a d-
di s pla cement curve fo r a s t i ffened pa nel i s s h o wn i n fi g ure 3-5.
Th e des i g n lo a d, i ndi ca t ed by po i nt B, co mmo nly co rres po nds t o 0. 80
o f t h e lo a d t o ca us e fi rs t yi eld, i ndi ca t ed by po i nt C .
D efo rma t i o n under a s t a t i c o r dyna mi c lo a d i s li kely t o co rres po nd
t o a curve o f fo rm ABC D F . Wh ere da ma g e h a s no t pro g res s ed beyo nd
po i nt F a nd wh ere no ma rked evi dence ex i s t s o f lo ca l fa i lure (i . e. ,
lo ca l s h ea r buckli ng o f web, t ri ppi ng o f s t i ffener, t ea ri ng o f
welds ), res i dua l s t reng t h under s ubs equent la t era l lo a ds i s li kely
t o rema i n a dequa t e. It i s i mpo rt a nt t o no t e t h a t res i dua l
s t i ffnes s , a s i ndi ca t ed by t h e s lo pe D E, i s a ls o unli kely t o di ffer
s i g ni fi ca nt ly fro m t h e o ri g i na l s t i ffnes s . P ro vi ded t h a t t h e
da ma g ed s t ruct ure i s no t requi red t o wi t h s t a nd la rg e i n pla ne
co mpres s i ve lo a ds a nd t h a t o verri di ng h ydro dyna mi c o r a es t h et i c
rea s o ns do no t ex i s t fo r res t o ri ng t h e o ri g i na l g eo met ry, repa i r
ma y be unneces s a ry. If clea r evi dence ex i s t s o f lo ca l fa i lures a nd
da ma g e i s judg ed t o h a ve pro g res s ed beyo nd po i nt F , res i dua l
s t reng t h a nd s t i ffnes s a re li kely t o be s ubs t a nt i a lly reduced.
29
I T T
SYM '
MODE L AOOPTE D
FOB ANAL YSIS
6.-
1
-
1
^_^- 1-1-5 SYMME TRICAL DAMAGE
Tm'!TT?2(THTT
,
T "UNDER LATERAL PRESSURE
5 UNSYMMETRICAL COLLAPSE
UNDER TRANSVERSE
COMPRESSION
008
r
-
006
/ 60/ b = 0023
"~-~^, 0 037
0 04
^ 0-05
11/ S- 90
002
0
4 'y
F IGU RE 3-4 EF F EC T OF S YMMETRIC D AMAGE ON S TIF F NES S AND S TRENGTH OF
LONG P LATES U ND ER TRANS VERS E C OMP RES S ION. (A) MOD EL
AD OP TED IN ANALYS IS . (B) LOAD -S HORTENING C U RVES
30
<
O
DISPLACEMENT, A
FIGURE 3.5 LATERAL LOAD-DISPLACEMENT RELATIONSHIP
FOR STIFFENED PANEL ^
12)
3 1
3. 4. 1. 2 Effect o f Lo ca li zed Imperfect i o n
In a ddi t i o n t o o vera ll i ni t i a l deflect i o n, lo ca li zed i ni t i a l
deflect i o ns a re o ft en o bs erved i n lo ng rect a ng ula r pla t es o n ma ri ne
s t ruct ures . Th e lo ca li zed i ni t i a l deflect i o ns a ls o i nfluence
s t i ffnes s a nd s t reng t h o f pla t es . Th e effect o f lo ca li zed
i mperfect i o n i s very s i mi la r t o t h a t o f i ni t i a l deflect i o n due t o
weldi ng . F o r lo ng t h i n pla t es , wi t h i ni t i a l deflect i o n under
lo ng i t udi na l co mpres s i o n, ult i ma t e s t reng t h i s rea ch ed a ft er lo s s
o f s t i ffnes s due t o la rg e deflect i o n, but fo r t h i cker pla t es i t i s
by lo s s o f s t reng t h due t o s prea d o f t h e lo ca li zed pla s t i c zo ne.
Acco rdi ng t o fi g ure 3-6, fro m [14] t h e lo ca li zed effect i s no t very
s ens i t i ve t o t h e s h a pe o f t h e dent . Is o la t ed s i nus o i da l dent s o f
mo dera t e a mpli t ude i nfluence pla t e s t reng t h o nly s li g h t ly les s t h a n
ri pple di s t o rt i o ns h a vi ng t h e s a me a mpli t ude a nd wa ve-leng t h . A
ma rg i na lly g rea t er lo s s o f s t reng t h i s ca us ed by a n i s o la t ed
s i nus o i da l dent o f leng t h 0. 8b t h a n o t h ers . Th e lo ca t i o n o f a n
i s o la t ed dent s li g h t ly i nfluences t h e lo s s o f co mpres s i ve s t reng t h
o f a lo ng pla t e. Th e i mperfect i o n effect i s ma rg i na lly g rea t er
wh en t h e dent i s lo ca t ed clo s e t o a pla t e end. Lo ca li zed i ni t i a l
defo rma t i o n ca us es les s preco lla ps e lo s s o f pla t e s t i ffnes s a nd
lea ds t o fa i lure a t a lo wer co mpres s i ve s t ra i n wi t h mo re ra pi d
po s t co lla ps e unlo a di ng t h a n ri pple di s t o rt i o n o f t h e s a me
a mpli t ude. Addi t i o n o f a lo ca li zed i mperfect i o n o f mo dera t e
a mpli t ude t o t h e o vera ll di s t o rt i o n ca us es lo s s o f s t reng t h .
3. 4. 2 F ra ct ure F a i lure
F ra ct ure fa i lure i s bro ken i nt o t wo ca t eg o ri es : fa t i g ue cra cki ng
a nd bri t t le fra ct ure. F a t i g ue cra cks di ffer fro m bri t t le fra ct ures
by t h e fo llo wi ng ch a ra ct eri s t i cs :
1. Li fe: Wh i le bri t t le fra ct ures a re i ns t a nt a neo us , g i ven
t h e ri g h t co ndi t i o ns , fa t i g ue fra ct ures t a ke t i me
t o develo p. U s ua lly fa t i g ue cra ck pro pa g a t i o n
co ns i s t s o f t h ree s t a g es na mely, i ni t i a t i o n,
pro pa g a t i o n a nd o ns et o f uns t a ble fra ct ure.
2. Lo ca t i o n: In mo s t ca s es fa t i g ue cra cks o ri g i na t e a t t h e
s urfa ce wh erea s bri t t le fra ct ures ca n i ni t i a t e a t
s ubs urfa ce defect s wh ere t h e t ri a x i a li t y o f s t res s
i s g enera lly no t i cea bly g rea t er.
3. C a us es : F a t i g ue cra cki ng i s i ni t i a t ed by repea t ed revers i ng
lo a ds a t a reg i o n o f s t res s co ncent ra t i o n, wh i le a
co mbi na t i o n o f mech a ni ca l a nd met a llurg i ca l fa ct o rs
ma y a ffect t h e i ni t i a t i o n o f bri t t le fra ct ure i n
t h ree wa ys - (i ) by s upplyi ng a ddi t i o na l energ y,
(i i ) by ca rryi ng lo ca l embri t t lement o f t h e
ma t eri a l a nd (i i i ) by s t res s co rro s i o n.
32
*
i.
^
CASE 1
CASE 2
FIGURE 3-6 COMPARISON OF LOAD -END S HORTENING C U RVE F OR LONG P LATES
(a/b=4) WITH OVERALL, P ERIOD IC AND LOC ALIZED D IS TORTION
33
3. 4. 2. 1 F a t i g ue Ana lys i s
Tra di t i o na lly, t h e fa t i g ue ch a ra ct eri s t i cs o f a ma t eri a l a re
repres ent ed by t h e S -N (cycli c s t res s S , vers us number o f cycles t o
fa i lure, N) di a g ra m a s s h o wn i n fi g ure 3-7. Th es e di a g ra ms ca n be
s pli t i nt o h i g h cycle fa t i g ue a nd lo w-cycle fa t i g ue reg i o ns . In
t h e h i g h cycle reg i o n, li fe o f t h e s t ruct ure i s a s s umed t o be
i nfi ni t e i f t h e ma x i mum s t res s i s kept belo w a cert a i n li mi t ca lled
t h e endura nce li mi t . Th e endura nce li mi t depends o n t h e ma t eri a l
a nd g eo met ri c pro pert i es o f t h e det a i l. S t ruct ures ba s ed o n s a fe-
li fe a ppro a ch a re des i g ned t o o pera t e i n t h i s reg i o n. Th i s i mpli es
lo w s ervi ce repa i r o r repla cement co s t s , a t t h e ex pens e o f a
co ns erva t i ve s t ruct ura l des i g n. Lo w cycle fa t i g ue i mpli es t h a t t h e
predo mi na nt ma x i mum s t res s es t h e s t ruct ure i s s ubject ed t o wi ll be
g rea t er t h a n t h e fa t i g ue li mi t s t reng t h a t 10
5
cycles o f t h e
ma t eri a l. Th e ma i n co ncept o f lo w-cycle des i g n cri t eri a i s t h a t i t
i s ba s ed o n a s h o rt ra t h er t h a n a n i nfi ni t e a ccept a ble li fet i me o f
a s t ruct ura l det a i l. Th e S -N curves a re ba s ed o n la bo ra t o ry t es t s
ca rri ed o ut o n unno t ch ed s peci mens . To a cco unt fo r no t ch es i n rea l
s t ruct ures , t h es e S -N' curves a re us ed a lo ng wi t h a fa t i g ue
s t reng t h -reduct i o n fa ct o r, K
f
. In t h e a bs ence o f t es t da t a fo r t h e
det ermi na t i o n o f K
f
, t h e mo s t wi dely us ed empi ri ca l rela t i o ns h i p i s
s ug g es t ed by Neuber [40], wh i ch a cco unt s fo r t h e effect o f s i ze by
mea ns o f t h e no t ch ra di us r, a nd o f t h e di fferences bet ween
ma t eri a ls by t h e ma t eri a l co ns t a nt , A, s ee eg ua t i o n (3. 1).
K. -l
9 K, = 1 + *-
1 +
- (3. 1 )
N i
F i g ure 3-8 s h o ws a plo t o f a
u
vers us A fo r va ri o us t ypes o f s t eel.
K
t
i s t h e s t res s co ncent ra t i o n fa ct o r. It i s t o be no t ed t h a t s o me
o f t h e a ppli ed lo a ds a t t h e no t ch ro o t ca us e yi eldi ng wh i ch i s no t
t a ken i nt o a cco unt ei t h er i n equa t i o n (3. 1) o r i n t h e S -N di a g ra ms .
Th e fa t i g ue pro ces s co ns i s t s o f t h ree s ucces s i ve s t a g es : cra ck
i ni t i a t i o n, cra ck pro pa g a t i o n a nd fa s t fra ct ure. Th erefo re, t h e
number o f cycles a s s o ci a t ed wi t h fa i lure, N
f
, i s t h e s um o f t h e
number o f cycles s pent i n i ni t i a t i o n, N
u
a nd t h e number s pent i n
pro pa g a t i ng t h e cra ck t o fa s t fra ct ure, N
p
, i . e. , N
f
= N
f
+ N . In
t h e h i g h cycle reg i me, N
;
i s a la rg e percent a g e o f N
f
. In s uch
ca s es wo rki ng s t res s es a re lo w a nd i t t a kes a lo ng t i me fo r a cra ck
t o develo p. In t h e lo w cycle reg i me, t h e s i t ua t i o n i s revers ed,
wh ere N; i s a very s ma ll percent a g e o f t h e t o t a l li fe, N
f
, a nd mo s t
o f t h e t i me i s s pent i n pro pa g a t i o n. Th i s i s i llus t ra t ed i n fi g ure
3-9. U s ua lly i n fa t i g ue ba s ed des i g n, cra ck pro pa g a t i o n i s no t
co ns i dered. Th i s i s beca us e s t ruct ures a re des i g ned t o o pera t e
belo w t h e endura nce li mi t , i . e. , i n t h e h i g h cycle reg i me wh ere
mo s t o f t h e fa t i g ue li fe i s s pend i n cra ck i ni t i a t i o n. But fo r
34
z :
<
er
w
oo
L d
cm
i
00
400
200
10
v
10
UN-NOTCHED
SPECIMENS
ENDURANCE
LIMIT
NOTCHED
SPECIMENS
S2
(K ) = S1/S2
T
10
A i
10-
CYCLES, N
10' 10
7
FIGURE 3.7 TYPICAL S-N DIAGRAMS FOR NOTCHED AND
UNNOTCHED STEEL SPECIMENS
35
260
o
X
<
100
300 500 700 900 1100 1300 1500 1700
a
u
(N/mrrf )-
FIGURE 3.8 ULTIMATE STRENGTH VS 'A'
(40"
3 6
b
en
O
a
3
CL
<
00
CD
L d
L T
I
CO
INITIATION LIFE.
PROPAGATION LIFE
TOTAL LIFE
NUMBER OF CYCLES, In N
FIGURE 3.9 SCHEMATIC S-N CURVES OF INITIATION, PROPAGATION,
(15)
AND TOTAL FATIGUE LIVE
c
37
welded s t ruct ures , fla ws i n t h e weld a re a s s umed t o ex i s t a nd h ence
cra ck pro pa g a t i o n co nt ro ls t h e t o t a l li fe o f t h e s t ruct ure.
F ra ct ure mech a ni cs t a kes i nt o a cco unt cra ck pro pa g a t i o n by-
pro vi di ng a ma t h ema t i ca l rela t i o ns h i p ba s ed o n t h e a ppli ed s t res s
ra ng e a nd ma t eri a l pro pert i es . In t h e t ra di t i o na l met h o d o f
fa t i g ue a s s es s ment t h e cri t eri a t o prevent fa i lure o f a co mpo nent
under va ri a ble a mpli t ude lo a di ng i s ca lcula t ed us i ng li nea r
P a lmg ren-Mi ner's h ypo t h es i s . In t h i s co ncept i t i s a s s umed t h a t n
;
number o f cycles o f s t res s ra ng es S
;
, us e up n
;
/N; fra ct i o n o f t h e
t o t a l li fe wh ere N
;
i s t h e fa t i g ue li mi t fo r t h e s t res s ra ng e S
;
.
Hence t h e a ccumula t ed da ma g e o f a s t ruct ure due t o t h e a ppli ca t i o n
o f "L" s t res s ra ng es i s :
L
n.
D = T -1 (3. 2)
Th erefo re, t o prevent fa t i g ue fa i lure, "D " s h o uld never ex ceed 1.
Wh i le i n ca s e o f mo s t s h i p s t ruct ures , t h e va lue o f "D " i s ch o s en
a s 1, fo r o ffs h o re s t ruct ures , dependi ng o n t h e cri t i ca li t y o f t h e
co mpo nent a s well a s t h e po s s i bi li t y o f a cces s fo r i ns pect i o n a nd
repa i r, t h e va lue ma y va ry fro m 0. 1 t o 1. 0.
Ta ble 3-2 g i ves s o me des i g n fa t i g ue fa ct o rs wh i ch a re t h e
reci pro ca l o f D , a s s ug g es t ed by t h e No rweg i a n P et ro leum
D i rect o ra t e (NP D ) [11]. In g enera l fa t i g ue fa i lure cri t eri a ba s ed
o n Mi ner's h ypo t h es i s a re very co ns erva t i ve. Th e h ypo t h es i s i s
li nea r i n na t ure, a nd t h erefo re do es no t a cco unt fo r ei t h er t h e
s equence o f a ppli ca t i o n o f t h e s t res s ra ng e no r do es i t t a ke i nt o
a cco unt i nt era ct i o n effect s o f va ri o us s t res s ra ng es wh i ch a ct
t o g et h er. Th erefo re, i n s umma ri zi ng , t h e co ns erva t i s ms o f
t ra di t i o na l fa t i g ue ba s ed a s s es s ment wh i ch ma ke t h em i na ppli ca ble
t o t h e "fa i l-s a fe" a ppro a ch a re:
do es no t t a ke i nt o a cco unt cra ck pro pa g a t i o n.
do es no t t a ke i nt o a cco unt pla s t i ci t y a t cra ck t i ps .
Hence s t res s es a re ba s ed o n ela s t i ci t y.
3. 4. 2. 2 F ra ct ure Mech a ni cs
F ra ct ure mech a ni cs i s a t o o l fo r ex a mi ni ng cra ck pro pa g a t i o n,
s t res s i nt ens i t i es a t t i ps o f cra cks , a nd res i dua l s t a t i c s t reng t h .
It i s g enera lly us ed i n t h e fa i l s a fe a na lys i s o f s t ruct ures .
Genera lly t h ere a re t wo t ypes o f fra ct ure mech a ni cs co ncept s . Th e
fi rs t , kno wn a s Li nea r Ela s t i c F ra ct ure Mech a ni cs (LEF M), i s
a ppli ca ble fo r s t res s levels up t o 0. 5 a
y
(t h e ma t eri a l t ens i le
yi eld s t reng t h ). Th e s eco nd, kno wn a s Genera l Yi eld F ra ct ure
Mech a ni cs (GYF M), i s a ppli ca ble fo r s t res s levels a bo ve 0. 5 a .
38
TABLE 3-2 D ES IGN F ATIGU E F AC TORS (F ATIGU E LIF E)
REGU LATIONS
NP D
C la s s i fi ca t i o n o f
s t ruct ura l
co mpo nent s
Acces s fo r i ns pect i o n a nd repa i r
No a cces s o r i n
t h e s pla s h zo ne
Acces s i ble
Belo w t h e s pla s h
zo ne
Abo ve t h e
s pla s h zo ne
Ma jo r i mpo rt a nce
fo r t h e s t ruct ura l
i nt eg ri t y-
10 3 2
Mi no r i mpo rt a nce
fo r t h e s t ruct ura l
i nt eg ri t y
3 2 1
Li nea r Ela s t i c F ra ct ure Mech a ni cs
Th e LEF M co ncept h a s been develo ped ba s ed o n t h e a s s umpt i o n t h a t
t h e pla s t i c zo ne wh i ch fo rms a h ea d o f t h e cra ck t i p i s very s ma ll
i n co mpa ri s o n wi t h t h e cra ck leng t h . Th e LEF M a ppro a ch i s ba s ed o n
t h e s t res s i nt ens i t y fa ct o r K. K g i ves a mea s ure o f t h e s t res s
fi eld a h ea d o f a s h a rp cra ck a nd i s rela t ed t o t h e no mi na l s t res s
i n t h e member, t h e s i ze o f t h e cra ck, a nd t h e g eo met ry o f t h e
s t ruct ure. A cra ck i n a s o li d ca n be s t res s ed i n t h ree di fferent
mo des , a s i llus t ra t ed i n fi g ure 3-10. Th ere i s a s t res s i nt ens i t y
fa ct o r t h a t co rres po nds t o ea ch mo de. F o r a cra ck under a rbi t ra ry
lo a di ng , a co mbi na t i o n o f t h es e t h ree s t res s i nt ens i t y fa ct o rs a re
us ed. Mo de I i s t h e predo mi na nt fo rm o f cra cki ng a nd h ence
di s cus s i o ns o f s t res s i nt ens i t y fa ct o rs wi ll be li mi t ed t o t h i s
mo de. Th e g enera l fo rm o f t h e s t res s i nt ens i t y fa ct o r fo r a fi ni t e
s i ze pla t e under t ens i le lo a di ng i s :
K=Yo /i
(3. 3)
wh ere a i s t h e a ppli ed no mi na l s t res s , a i s t h e s emi -cra ck leng t h
o f a t h ro ug h -t h e-t h i cknes s cra ck a nd Y i s a pa ra met er wh i ch depends
o n t h e g eo met ry o f t h e cra ck.
U ns t a ble fra ct ure o ccurs wh en t h e s t res s i nt ens i t y fa ct o r a t t h e
cra ck t i p rea ch es a cri t i ca l va lue kno wn a s t h e cri t i ca l s t res s
i nt ens i t y fa ct o r, K
IC
. K
IC
i s a ma t eri a l pro pert y wh i ch repres ent s
t h e fra ct ure t o ug h nes s o f t h e ma t eri a l. C ri t i ca l s t res s i nt ens i t y,
K
IC
i s a funct i o n o f t h e (1) t h i cknes s o f t h e ma t eri a l (t ri a x i a li t y
effect ), (2) t empera t ure a nd (3) lo a di ng ra t e. Th es e rela t i o ns
h a ve been ex t ens i vely di s cus s ed i n [15, 16]. F i g ures 3-11, 3-12
a nd 3-13 i llus t ra t e t h e effect o f t h i cknes s , t empera t ure a nd
lo a di ng ra t e o n t h e cri t i ca l s t res s i nt ens i t y fa ct o r.
39
MODE I
OPE NING MODE
MODE II
SL IDING MODE
MODE III
TE ARING MODE
FIGURE 3.10 THREE MODES OF CRACKING
PLANE STRESS K
PL ANE STRAIN K h
THICK NE SS, t
FIGURE 3.11 RELATIONSHIP BETWEEN THICKNESS AND
.FRACTURE TOUGHNESS
(1?)
K , C
SLOW
\ i (
K
io)
INTERMEDIATE DYNAMIC
TEMPERATURE. F
LOADING RATE
IGURE 3.12 RELATIONSHIP BETWEEN TEMPERATURE AND
115)
FRACTURE TOUGHNESS
FIGURE 3.13 RELATIONSHIP BETWEEN LOADING
;15)
FRACTURE TOUGHNESS
AND
40
Genera l Yi eld F ra ct ure Mech a ni cs
Th e t ech ni ques o f li nea r ela s t i c fra ct ure mech a ni cs g i ve
repro duci ble t o ug h nes s va lues wh i ch ma y be us ed t o ca lcula t e t h e
cri t i ca l defect s i zes t o a vo i d bri t t le fa i lure. F o r s uffi ci ent ly
duct i le ma t eri a ls , s uch a s h i g h s t reng t h s t eel, t h e LEF M t h eo ry
brea ks do wn wh en t h e ex t ent o f pla s t i ci t y pro ceedi ng fra ct ure
beco mes la rg e. Th ree GYF M met h o ds a re di s cus s ed belo w. Th es e
t h ree met h o ds h a ve wi de a ppli ca bi li t y a nd a re a dequa t e t o na rro w
t h e g a p bet ween t h e a ppli ca t i o n o f t h e li nea r ela s t i c fra ct ure
mech a ni cs a nd t h e g enera l yi eldi ng fra ct ure mech a ni cs fo r s peci a l
co ndi t i o ns , s peci a l ca s es a nd s peci a l g eo met ri es .
(1) Equi va lent Ela s t i c C ra ck Met h o d:
Th e li nea r ela s t i c fra ct ure mech a ni cs met h o d, ma y be mo di fi ed t o
pro vi de mo re a ccura cy wh en a s ma ll pla s t i c zo ne i s pres ent i n fro nt
o f t h e cra ck befo re t h e fa s t fra ct uri ng pro ces s co mmences . Th e
mo di fi ca t i o n t a kes i nt o a cco unt t h e effect o f t h e s ma ll a mo unt o f
pla s t i c yi eldi ng o n t h e s t res s es a nd di s pla cement s a h ea d o f a s h a rp
cra ck. Th e pla s t i c zo ne i s a s s umed t o ex t end a s ma ll di s t a nce, r
y
,
a h ea d o f t h e cra ck t i p. Wi t h i n t h e pla s t i c zo ne, t h e t ens i le
s t res s i s equa l t o t h e uni a x i a l yi eld s t res s a
y
(pla ne s t res s
co ndi t i o n a nd a s s umi ng no wo rk-h a rdeni ng ), a s s h o wn i n fi g ure 3-14.
F o r t h e purpo s e o f s i mpli fi ed s t res s a na lys i s , t h e co ncept o f a
no t i o na l ela s t i c, o r equi va lent ela s t i c cra ck, i s s o met i mes a s s umed
t o a na lyze t h e s t res s di s t ri but i o n a h ea d o f t h e t i p o f a cra ck.
Th e no t i o na l ela s t i c cra ck i s equi va lent t o t h e co nfi g ura t i o n o f a
rea l cra ck a nd pla s t i c zo ne a t mo dera t ely la rg e di s t a nces . Th e
equi va lent cra ck t i p i s s uppo s ed t o be lo ca t ed a t t h e po i nt r
y
,
wh ere r
y
i s g i ven by equa t i o n (3. 4a ):
r = J- (3. 4a )
F ro m equa t i o n (3. 3), we g et
Y
2* a
y
2
r = H- (3. 4b)
y
2a
y
2
Th erefo re t h e equi va lent cra ck leng t h , a ^ i s equa l t o a + r
y
. Th e
equi va lent s t res s i nt ens i t y fa ct o r, K
e
, a t t h e mo di fi ed cra ck t i p
i s
41
FIGURE 3.14 PLASTIC ZONE AT CRACK TIP
42
K
e
= o
v
/: : (a + r
y
)Y
1
= a^izI'
l
Y
1
(3. 4c)
wh ere Y
x
i s t h e s a me a s Y but a t a = a
x
.
Th e cri t eri a t o a vo i d bri t t le fra ct ure i s g i ven by equa t i o n (3. 5)
K
e
* K
IC
(3. 5)
(2) C ra ck Openi ng D i s pla cement Met h o d:
Th e pla s t i c zo ne a t t h e t i p o f a cra ck ena bles t h e t wo fa ces t o
mo ve a pa rt a t t h e cra ck t i p wi t h o ut i ncrea s i ng t h e leng t h o f t h e
a ct ua l cra ck. Th e ma g ni t ude o f t h e s epa ra t i o n a t t h e cra ck t i p h a s
been t ermed t h e cra ck t i p o peni ng di s pla cement (C TOD ) a nd i s
deno t ed by S
c
. Th e mo del us ed i n t h e C TOD a ppro a ch i s ba s ed o n a
rea l cra ck o f leng t h 2a i n a n i nfi ni t e pla t e, s ubject ed t o a
uni fo rm t ens i le s t res s a ppli ed i n t h e y-di rect i o n, no rma l t o t h e
cra ck leng t h . Th e res ult a nt pla s t i c zo ne, a t t h e t i p o f t h e rea l
cra ck, ex t ends t o x = + a
lf
o r fo r t h e purpo s e o f a na lys i s , i t i s
a s s umed t h a t t h e cra ck leng t h i s 2a , . Eq. 3. 6 ex pres s es t h e
rela t i o ns h i p bet ween t h e h ypo t h et i ca l cra ck, 2a , , a nd t h e rea l
cra ck, 2a . Th e a s s o ci a t ed cra ck o peni ng di s pla cement a t t h e t i p o f
t h e rea l cra ck ca n be repres ent ed by equa t i o n 3. 7:
= sec
7 1 o
2 o~
yj
( 3.6)
*c =
8oa /
y
In Sec
7 tE 2 a
(3.7)
y)
wh ere E i s Yo ung 's mo dulus o f ela s t i ci t y.
Th e va lue o f S
c
beco mes very s ens i t i ve t o s ma ll ch a ng es i n t h e ra t i o
o f a/a
y
a s t h i s ra t i o a ppro a ch es uni t y. Th erefo re, i t i s prefera ble
t o rela t e t h e S
c
va lue t o t h e o vera ll s t ra i n, e, o ver a g a ug e
leng t h , 2y i n equa t i o n (3. 8).
43
Therefore:
6 = k,a In sec c
U
y
fjt ^
2 6,
y/
( 3.8)
Th e cri t eri a t o a vo i d fra ct ure i s g i ven by:
c<CR
(3. 9)
5
rR
i s a ma t eri a l pro pert y wh i ch i s dependent upo n t h e s ervi ce
t empera t ure a nd lo a di ng co ndi t i o ns .
(3) J -Int eg ra l Met h o d:
Th e J -i nt eg ra l, wh i ch i s a no t h er pa ra met er t h a t des cri bes t h e
co ndi t i o n nea r t h e cra ck t i p, i s a n a lt erna t i ve fa i lure cri t eri a .
It ca n be defi ned fo r a ny li nea r o r no n-li nea r s t res s -s t ra i n
rela t i o ns h i p wi t h t h e res erva t i o n t h a t i t mus t be ela s t i c (i . e. ,
revers i ble) [17]. Th e J -i nt eg ra l i s o bt a i ned by i nt eg ra t i ng t h e
fo llo wi ng ex pres s i o n a lo ng a n a rbi t ra ry pa t h a ro und t h e t i p o f a
cra ck (fi g ure 3-15):
J = | wdy - T ds
(3. 10)
wh ere r i s t h e pa t h o f i nt eg ra t i o n, W i s t h e s t ra i n energ y dens i t y,
T i s t h e t ra ct i o n vect o r, u i s t h e di s pla cement vect o r, a nd ds i s
a n i ncrement a lo ng r. Th e co o rdi na t es x a nd y a re a s defi ned i n
fi g ure 3-15. F o r no n-li nea r ela s t i c ma t eri a ls , Ri ce [18] s h o wed
t h a t t h e va lue o f J i s i ndependent o f t h e i nt eg ra t i o n pa t h a s lo ng
a s t h e co nt o ur enclo s es t h e cra ck t i p, a s i llus t ra t ed i n fi g ure
3-15.
F o r li nea rly ela s t i c ma t eri a ls , t h e s t res s i nt ens i t y fa ct o r i s
rela t ed t o J by t h e fo llo wi ng equa t i o n i n t h e pla ne s t res s ca s e:
K = v^
(3. 11)
44
In [19] i t h a s been demo ns t ra t ed t h a t t h e J -i nt eg ra l met h o d ca n be
us ed i n fi ni t e element a na lys i s o f cra cked s t ruct ure. A s ubro ut i ne
wa s develo ped, wh i ch wa s i nco rpo ra t ed i n a fi ni t e element pro g ra m,
t o co mput e Ri ce's J -i nt eg ra l va lue a nd t h en t o deri ve t h e s t res s
i nt ens i t y fa ct o r us i ng equa t i o n (3. 10).
Th e fra ct ure cri t eri a i s t h a t fa i lure o f t h e s t ruct ure wi ll o ccur
o nce J a ppro a ch es a cri t i ca l va lue (J
c
) , i . e.
J i J
c
(3. 12)
F ra ct ure C ri t eri a
In g enera l t h e ex pect ed fra ct ure beh a vi o r o f a s t ruct ure i s rela t ed
t o t h ree levels o f fra ct ure perfo rma nce, pla ne s t ra i n, ela s t i c-
pla s t i c o r fully pla s t i c. In a n eng i neeri ng s ens e, t h e ra t i o o f
cri t i ca l pla ne-s t ra i n s t res s i nt ens i t y fa ct o r, K
IC
, a nd t ens i le
yi eld s t reng t h , a
y
, i s a s s umed t o be a g o o d rela t i ve i ndex fo r
mea s uri ng t h e level o f fra ct ure perfo rma nce. D ependi ng o n t h e ra t e
o f lo a di ng , va lues o f K
IC
a nd a
y
mea s ured under s t a t i c o r dyna mi c
co ndi t i o ns a re t o be us ed. S lo w o r s t a t i c lo a di ng i s defi ned a s a
lo a di ng ra t e o f a ppro x i ma t ely 10"
5
i n/i n/s ec. D yna mi c, o r i mpa ct
lo a di ng i s defi ned a s a lo a di ng ra t e o f a ppro x i ma t ely 10 i n/i n/s ec.
Th e t h ree levels o f perfo rma nce a re i nfluenced by t h e s ervi ce
t empera t ure a nd lo a di ng ra t e. Th i s i s s h o wn s ch ema t i ca lly i n
fi g ure 3-16. Ba s ed o n t h e level o f fra ct ure perfo rma nce o f t h e
s t ruct ure, di fferent fra ct ure cri t eri a a re t o be a ppli ed. Th e
t h ree levels o f fra ct ure perfo rma nces a re defi ned belo w:
1. P la ne-S t ra i n Beh a vi o r:
S t ruct ures , wh o s e t o ug h nes s a nd t h i cknes s a re s uch t h a t :
K
T1
t
2. 5
ex h i bi t pla ne-s t ra i n beh a vi o r. Th ey fra ct ure under ela s t i c s t res s ,
g enera lly i n a bri t t le ma nner. S i nce bri t t le fra ct ure ca n lea d t o
ca t a s t ro ph i c fa i lure, us ua lly s t ruct ures a re ch o s en t o be o f s uch
t o ug h nes s s o a s t o a vo i d pla ne-s t ra i n beh a vi o r a t s ervi ce
t empera t ure a nd s ervi ce lo a di ng ra t e. Ho wever, very t h i ck pla t es ,
o r pla t es us ed t o fo rm co mplex g eo met ri es wh ere t h e co ns t ra i nt ca n
be very h i g h , ma y be s us cept i ble t o bri t t le fra ct ures even t h o ug h
t h e no t ch t o ug h nes s , a s mea s ured by s ma ll-s ca le la bo ra t o ry t es t s ,
a ppea rs s a t i s fa ct o ry. S t ruct ures perfo rmi ng a t t h i s level a re
a s s es s ed us i ng LEF M cri t eri a .
45
FIGURE 3.15 ARBITRARY CONTOUR AROUND A CRACK TIP
PLASTIC
<
O
O
in
ELASTIC-
PLASTIC
PLAIN
STRAIN
MPACT
LOADING
TEMPERATURE
FIGURE 3.16 SCHEMATIC SHOWING RELATION BETWEEN NOTCH-TOUGHNESS
TEST RESULTS AND LEVELS OF FRACTURE PERFORMANCE
FOR VARIOUS LOADING RATES
5'
46
2. Ela s t i c-P la s t i c (Mi x ed - mo de) Beh a vi o r:
S t ruct ures , wh o s e ra t i o o f
C T
y
wh ere a ca n be a s h i g h a s 2 o r 3, ex h i bi t ela s t i c-pla s t i c fra ct ures
wi t h va ryi ng a mo unt o f yi eldi ng pri o r t o fra ct ure. F ra ct ure i s
us ua lly preceded by t h e fo rma t i o n o f la rg e pla s t i c zo nes a h ea d o f
t h e cra ck, t h erefo re t h e t o lera ble fla w s i zes a t fra ct ure ca n be
fa i rly la rg e. S h i p s t ruct ures us ua lly perfo rm a t t h i s level. Th e
bi g g es t dra wba ck o f s t ruct ures perfo rmi ng a t t h i s level i s t h e
di ffi cult y i n o bt a i ni ng co rrect K
IC
va lues wi t h s ma ll la bo ra t o ry
s peci mens . To o verco me t h i s pro blem C TOD ba s ed fra ct ure cri t eri a
a re us ed.
3. P la s t i c (Genera l Yi eldi ng ) Beh a vi o r:
S t ruct ures , wh o s e ra t i o o f
K
i c
y
>
y/i,
wh ere a ca n be a s h i g h a s 2 o r 3, ex h i bi t duct i le pla s t i c fra ct ures
preceded by la rg e defo rma t i o n. Hence, t o lera ble fla w s i zes a re
very la rg e. No rma lly s t ruct ures a re no t des i g ned t o perfo rm a t
t h i s level s i nce i t wo uld be eco no mi ca lly unfea s i ble, ex cept i n
cert a i n ca s es i n t h e nuclea r po wer i ndus t ry wh ere t h e rea ct o rs
o pera t e a t very h i g h t empera t ures .
F o r s t ruct ures perfo rmi ng a t t h e pla ne-s t ra i n level, cra cks ca n be
a s s es s ed by co mpa ri ng t h e s t res s i nt ens i t y fa ct o r a t t h e cra ck t i p
wi t h t h e cri t i ca l pla i n-s t ra i n i nt ens i t y fa ct o r K,
c
. F o r s t ruct ures
ex h i bi t i ng ei t h er ela s t i c-pla s t i c o r pla s t i c beh a vi o r, a n
i nt era ct i o n curve mus t be us ed. Th es e i nt era ct i o n curves , a ls o
kno wn i n t h e i ndus t ry a s "des i g n-curves " o r "fa i lure a s s es s ment
di a g ra ms ", h a ve been develo ped co ns i deri ng t h e i nt era ct i o n o f t wo
fa i lure mo des , na mely bri t t le fa i lure a nd pla s t i ci t y. F o r
pra ct i ca l a ppli ca t i o n o f t h es e di a g ra ms , eng i neeri ng pro cedures
h a ve been o ut li ned i n t h e Bri t i s h S t a nda rd Ins t i t ut e, BS I,
publi s h ed do cument , P D 6493: 1991 [20]. . Releva nt po rt i o ns o f
P D 6493 fo r pres ent us e a re di s cus s ed i n s ect i o n 4. 2. Th e
underlyi ng t h eo ry beh i nd t h e develo pment o f t h e fa i lure a s s es s ment
di a g ra ms , i ncludi ng t h e a s s umpt i o ns a nd mo di fi ca t i o ns , ca n be fo und
i n [20, 21].
47
Th e fra ct ure perfo rma nce o f a co nvent i o na l ma ri ne s t ruct ure
o pera t i ng under no rma l s ervi ce co ndi t i o ns i s t h e ela s t i c-pla s t i c
level. F ra ct ure a s s es s ment a t t h i s level i s us ua lly perfo rmed
us i ng a fra ct ure a s s es s ment di a g ra m ba s ed o n C TOD met h o ds [19] o r
i t s mo di fi ca t i o ns [20-25]. F ra ct ure t o ug h nes s pa ra met ers t o be
us ed a t t h i s level a re ba s ed o n ei t h er t h e s t res s i nt ens i t y fa ct o r
met h o d, K
IC
o r t h e C TOD met h o d 5
C R
. K
IC
va lues a re mea s ured fro m
la bo ra t o ry s peci mens under pla ne-s t ra i n co ndi t i o ns wi t h o ut a ny
pla s t i c zo ne fo rma t i o n. U nfo rt una t ely, t h e la bo ra t o ry co ndi t i o ns
h a rdly s i mula t e rea li t y. Mea s urement s fo r < S
C R
ca n be ma de even wh en
t h ere i s co ns i dera ble pla s t i c flo w a h ea d o f a cra ck. Th es e fa ct o rs
s h o uld be t a ken i nt o a cco unt wh en a na lyzi ng cra cked s h i p
s t ruct ures , o t h erwi s e a s s es s ment ma de s o lely o n K
IC
va lues ma y yi eld
very co ns erva t i ve res ult s .
F a t i g ue C ra ck Gro wt h
Bes i des pro vi di ng fo r cri t eri a t o a s s es s cra cks , fra ct ure mech a ni cs
a ls o pro vi des a n a na lyt i ca l t o o l t o s t udy fa t i g ue cra ck
pro pa g a t i o n. It pro vi des a ba s i s fo r det ermi ni ng cra ck g ro wt h
h i s t o ry g i ven t h e lo a di ng a nd t h e ma t eri a l pro pert i es . Th e ba s i s
o f mo s t cra ck g ro wt h pri nci pa ls i s t h e empi ri ca l rela t i o ns h i p put
fo rwa rd by P a ri s i n 1963, wh i ch i s
A = C (AK. r <
3
-
13
)
dN *
wh ere
a i s t h e ch a ra ct eri s t i c cra ck di mens i o n.
N i s t h e number o f cycles .
AK, i s t h e s t res s i nt ens i t y fa ct o r ra ng e ba s ed o n t h e ma x i mum
a nd mi ni mum a ppli ed s t res s .
M
C , m a re ma t eri a l pro pert i es dependi ng o n t h e o pera t i ng
envi ro nment .
As s h o wn i n fi g ure 3-17, t h e cra ck g ro wt h curve i s s i g mo i da l i n
s h a pe a nd ca n be s pli t i nt o t h ree reg i o ns , ea ch co rres po ndi ng t o
t h e t h ree peri o ds i n t h e li fe o f a fa t i g ue curve, na mely cra ck
i ni t i a t i o n, pro pa g a t i o n a nd uns t a ble fra ct ure. Reg i o n I i s
ch a ra ct eri zed by a "fa t i g ue-t h res h o ld" cycli c s t res s -i nt ens i t y
fa ct o r fluct ua t i o n, AK^. Belo w t h e t h res h o ld, cra cks do no t
pro pa g a t e. It h a s been o bs erved t h a t t h e t h res h o ld va lue AK^ i s
i nfluenced s t ro ng ly by t h e envi ro nment a nd t h e s t res s ra t i o . R,
wh i ch i s t h e ra t i o o f t h e mi ni mum s t res s t o ma x i mum s t res s .
Increa s i ng R, decrea s es t h e t h res h o ld wh i le i ncrea s i ng t h e g ro wt h
48
LxJ
<
O
00
o
o
O
T5
o
o
UJ
Q.
<
o
ct:
o
o
<
O
STRESS-INTENSITY-FACTOR RANGE,
A K| LOG SCALE
FIGURE 3.17 SCHEMATIC REPRESENTATION OF FATIGUE-CRACK GROWTH
49
ra t e i n t h e vi ci ni t y. Th e t h res h o ld i s i ndependent o f o t h er
ma t eri a l pro pert i es . Ana lys i s o f publi s h ed da t a fo r ca rbo n o r
ca rbo n ma ng a nes e s t eels i n a i r a nd s ea -wa t er [26] i ndi ca t es t h a t
fo r 97. 7% pro ba bi li t y o f s urvi va l, AK^ (i n N/mm
3/2
) i s a s fo llo ws :
or
or
AK
4
= 63 for R > 0.5
AK,,, = 170 - 214R for 0 * R * 0.5 ( 3.14)
AK,, = 170 for R -c 0
Th e s h a pe o f t h e cra ck pro pa g a t i o n curve i n t h e lo w-cra ck g ro wt h
reg i o n ca n be a ppro x i ma t ed i n ma ny wa ys [27], but a co ns erva t i ve
a ppro a ch i s t o ex t ra po la t e t h e Reg i o n 2 P a ri s curve t o t h e
t h res h o ld. A mo di fi ca t i o n t o P a ri s ' equa t i o n o f cra ck g ro wt h wa s
pro po s ed by F o rma n, et a l. [28] wh i ch t a kes a cco unt o f t h e fa ct
t h a t da /dn beco mes i nfi ni t e wh en t h e cra ck rea ch es a cri t i ca l s i ze,
i . e. , wh en K^ rea ch es K
IC
. Th ey a rri ved a t :
da C AK, "
dN (1 - R)K
IC
- AKj
(3. 15)
Wh i le t h e a ppli ca t i o n o f t h e cra ck g ro wt h pri nci pa ls a re s t ra i g h t
fo rwa rd i n t h e ca s e o f co ns t a nt a mpli t ude lo a di ng , rea l s t ruct ures
li ke s h i ps a re s ubject ed t o va ri a ble a mpli t ude lo a di ng s .
U nfo rt una t ely, t h e effect s o f va ri a ble-a mpli t ude lo a di ng o n fa t i g ue
li fe a re pres ent ly no t well es t a bli s h ed, a lt h o ug h mo dels pres ent ed
by Ba rs o m [29, 30, 31] a nd by Wei a nd S h i h [32] a ppea r t o be
pro mi s i ng . Th e mo del a dvo ca t ed by Wei a nd S h i h i s a s uperpo s i t i o n
mo del wh ere t h e dela y cycles ca us ed by a ch a ng e i n s t res s ma g ni t ude
a re s uperi mpo s ed o n t h e cycles o bt a i ned under co ns t a nt -a mpli t ude
lo a di ng a s s umi ng no lo a d i nt era ct i o ns . Th e dela y cycles i n t h i s
mo del mus t be es t i ma t ed fro m ex peri ment a l da t a . Th e mo del a dva nced
by Ba rs o m [26] rela t es fa t i g ue-cra ck-g ro wt h ra t e per cycle t o a n
effect i ve s t res s -i nt ens i t y fa ct o r t h a t i s ch a ra ct eri s t i c o f t h e
pro ba bi li t y-dens i t y curve. Th e develo pment o f t h i s mo del, wh i ch
wa s des i g na t ed t h e rms (ro o t -mea n-s qua re) mo del, a nd t h e s uppo rt i ng
ex peri ment a l da t a ca n be fo und i n [15]. In [15], i t i s s h o wn t h a t
wi t h i n t h e li mi t s o f ex peri ment a l wo rk, t h e a vera g e fa t i g ue-cra ck-
g ro wt h ra t e per cycle, da /dn, under va ri a ble a mpli t ude ra ndo m-
s equence s t res s s pect ra ca n be repres ent ed by:
50
= C (AK )
m
(3. 16)
dN
K
*'
wh ere,
rms
N
L
2
S AKj
L i s t h e number o f s t res s ra ng es a nd i s i ndependent o f t h e o rder o f
o ccurrence o f t h e cycle lo a d fluct ua t i o n a s s h o wn i n fi g ure 3-18.
Th e a ct ua l rela t i o ns h i p bet ween cra ck leng t h a nd t h e li fe o f t h e
s t ruct ure i s o bt a i ned by i nt eg ra t i ng t h e cra ck g ro wt h bet ween t h e
i ni t i a l a nd t h e fi na l s i ze. Th e i nt eg ra t i o n i s us ua lly perfo rmed
numeri ca lly. Th e fi na l cra ck s i ze i s t h e t o lera ble cra ck s i ze
wh i ch i s ba s ed o n t h e co ns i dera t i o n o f uns t a ble fra ct ure o r t h e
lo s s o f effect i venes s o f t h e lo ca l s t ruct ure. F i g ure 3-19
i llus t ra t es t h e res ult o f t h i s i nt eg ra t i o n i n t erms o f number o f
cycles o f lo a d a ppli ca t i o n, N, ex pended t o rea ch a cra ck o f leng t h
a .
3. 5 Ro le o f Ins pect i o n
Ba ckg ro und
One o f t h e premi s es o n wh i ch res i dua l s t reng t h a s s es s ment i s ba s ed,
i s t h e a va i la bi li t y o f i nfo rma t i o n reg a rdi ng da ma g e. Th e bes t
s o urce o f s uch i nfo rma t i o n i s i ns pect i o n reco rds . Tra di t i o na lly,
i ns pect i o ns o f ma ri ne s t ruct ures h a ve been ba s ed mo re o n h euri s t i cs
a nd ex peri ence ra t h er t h a n o n a ra t i o na l ba s i s , t h e la t t er bei ng
i mpo s s i ble due t o la ck o f t h eo ret i ca l kno wledg e. At t i mes , s uch a n
a ppro a ch co uld lea d t o s ubject i ve i ns pect i o n res ult s dependi ng o n
t h e qua li fi ca t i o n o f t h e i ns pect o rs a nd t h e s co pe o f i ns pect i o n.
S h i p s t ruct ura l des i g n i s ba s ed o n a co mbi na t i o n o f s a fe-li fe a nd
fa i l-s a fe a ppro a ch es beca us e i t wo uld be uneco no mi ca l a nd
i nfea s i ble t o bui ld s h i p us i ng o nly t h e s a fe-li fe a ppro a ch wh i ch
a s s ume t h e s h i p wi ll la s t a li fet i me i n a s -bui lt co ndi t i o ns . Th e
fa i l s a fe a ppro a ch ca lls fo r ea rly det ect i o n a nd mo ni t o ri ng o f
da ma g e a nd pro vi s i o n fo r t h e ca pa bi li t y t o co nt i nue t o ca rry lo a d
i n a da ma g ed co ndi t i o n - i n o t h er wo rds , pres ence o f s t ruct ura l
redunda ncy. S uch a n a ppro a ch , mo deled a ft er t h o s e emplo yed i n t h e
a ero s pa ce i ndus t ry, i n co njunct i o n wi t h cla s s i fi ca t i o n s o ci et y
requi rement s o f peri o di ca l ("S peci a l") det a i led s urveys every 5
yea rs , s upplement ed by a nnua l do cki ng s urveys a nd vi s ua l s urveys
51
-5
10
o
o
c
a
UJ
i
<
o
Qi
<
ct:
o
i
<
10
CONSTANT AMPLITUDE
O ASCENDING ORDER
D DESCENDING ORDER
V ASCENDING-DESCENDING
ORDER
A RANDOM SEQUENCE
1 ksi y inch = 1.0998 MN/m
1 INCH = 25.4 mm
2
10 20 40 60
A K
rms
ksi inch
FIGURE 3.18
SUMMARY OF CRACK GROWTH RATE DATA UNDER
RANDOM-SEQUENCE AND ORDERED-SEQUENCE LOAD
5)
FLUCTUATIONS.
52
NJ
CO
NUMBER OF CYCLES OF LOADING, N
FIGURE 3.19 HISTORY OF FATIGUE-CRACK GROWTH
53
ca rri ed o ut by t h e o wners , h a s wo rked qui t e well a s evi denced by
s t a t i s t i cs o n t h e decrea s i ng number o f da ma g es t o s h i p s t ruct ures
o ver t h e yea rs . But cert a i n cri t i ca l di fferences bet ween t h e t wo
i ndus t ri es , a nd new t rends i n t h e ma ri ne i ndus t ry, ca ll fo r a
reeva lua t i o n o f t h e ro le o f i ns pect i o n i n t h e co nt ex t o f s a fet y a nd
dura bi li t y o f ma ri ne s t ruct ures .
P ri ma ry di fferences bet ween ma ri ne s t ruct ures (s h i ps o r o ffs h o re
pla t fo rms ) a nd a i rcra ft a re:
Ma ri ne s t ruct ures o pera t e i n a much h a rs h er envi ro nment .
Any predi ct i o ns i nvo lvi ng lo a di ng , deg ra da t i o n i n s t reng t h
o r t h e ph ys i ca l co ndi t i o n o f t h e s t ruct ure a re fra ug h t
wi t h uncert a i nt i es .
As ment i o ned i n [33], due t o t h e co ndi t i o ns a nd t h e ex t ent
o f i ns pect i o n requi red fo r ma ri ne s t ruct ures a ny a t t empt
s h o uld be co ns i dered "h ero i c". Wh i le a i rcra ft i ns pect i o ns
a re co nduct ed under t h e co nt ro lled a t mo s ph ere o f a h a ng a r,
s h i p i ns pect i o ns a re h eld i n co ndi t i o ns o f da rknes s ,
wa t er, di rt , h umi di t y a nd li mi t ed a cces s i bi li t y.
D ependi ng o n t h e s i ze o f t h e s h i p, t h e ex t ent o f
i ns pect i o n co uld be a bo ut 2000 mi les o f weldi ng a nd 100-
200 a cres o f s t eel s urfa ce, wh i ch i s t ypi ca l o f VLC C o r
U LC C .
Ex t remely h i g h do wnt i me co s t . Typi ca lly a s h i p co s t s
a bo ut $40, 000 per da y a nd t h e bi g g es t No rt h S ea pla t fo rms
a bo ut $2 0, 000, 000 per da y o f lo s t pro duct i o n.
New t rends i n t h e ma ri ne i ndus t ry h a ve been:
D ue t o OP A 90, i ncrea s ed s ubdi vi s i o n o f t a nkers res ult s i n
mo re numbers o f co mpa rt ment s a nd s urfa ce a rea fo r
i ns pect i o n.
Mi ni mi za t i o n o f s t ruct ura l wei g h t a nd t h i cknes s .
Acco rdi ng t o [34] i n t h e t wo peri o ds fro m 1953 t o 1965 a nd
fro m 1965 t o 1985 s t eel wei g h t reduct i o n wa s a bo ut 25% a nd
15% res pect i vely.
D evelo pment o f s t ruct ures li ke o ffs h o re pla t fo rms a nd
S WATH, wi t h very li mi t ed s ervi ce-li fe ex peri ence.
54
Mo dern P ra ct i ces :
One o f t h e ma jo r pro blems fa ci ng i ns pect i o n pro cedures i s wh ere t o
lo o k fo r wh a t ki nd o f da ma g e. S uch i nfo rma t i o n i s a bs o lut ely
neces s a ry fo r h ug e s t ruct ures li ke VLC C 's , wh ere cri t i ca l
ex a mi na t i o n o f a ll det a i ls o r jo i nt s i s pra ct i ca lly i mpo s s i ble.
S elect ed cri t i ca l det a i ls t o be i ns pect ed co uld be ba s ed o n
s ervi ce-li fe ex peri ence a nd a na lyt i ca l eva lua t i o ns . Ana lyt i ca l
eva lua t i o ns ca n be ca rri ed o ut us i ng well es t a bli s h ed met h o ds i n
s t ruct ura l a na lys i s , fa t i g ue a s s es s ment , cra ck-g ro wt h predi ct i o n
a nd ri s k a na lys i s met h o ds . Ex a mples o f t h e a ppli ca t i o n o f t h es e
t ypes o f t h eo ret i ca l t o o ls fo r i ns pect i o n pro cedures ca n be fo und
i n t h e a ero s pa ce i ndus t ry [35, 36] wh ere t h ey h a ve pres ent ed
i nfo rma t i o n o n i ns pect i o n i nt erva ls fo r cri t i ca l co mpo nent s o f
a i rcra ft s t ruct ures ba s ed o n a na lyt i ca l eva lua t i o n. Recent effo rt s
i n t h i s di rect i o n i n t h e ma ri ne i ndus t ry h a ve been i ni t i a t ed by t h e
Ta nker S t ruct ure C o -o pera t i ve F o rum, TS C F [37] a nd t h e
Int erna t i o na l As s o ci a t i o n o f C la s s i fi ca t i o n S o ci et i es , IAC S fo r
bulk ca rri ers .
Bes i des peri o di c i ns pect i o ns , s t ruct ura l mo ni t o ri ng t ech ni ques t o
pro vi de a co nt i nuo us revi ew o f t h e i nt eg ri t y o f ma ri ne s t ruct ures
mus t be i mplement ed. F eedba ck fro m s uch mo ni t o ri ng devi ces ca n be
h elpful t o eva lua t e perfo rma nce fo r uni que s t ruct ura l det a i ls fo r
wh i ch previ o us s ervi ce ex peri ence ma y no t be a va i la ble. S uch
devi ces co uld a ls o h elp i n a s s es s ment o f remo t e o r i na cces s i ble
a rea s . - F ew i ns t a nces o f t h e a ppli ca t i o n o f s uch mo ni t o ri ng s ys t ems
ca n be fo und i n t h e s h i ppi ng i ndus t ry [38] o r t h e o ffs h o re i ndus t ry
[39].
Ano t h er s i g ni fi ca nt a rea o f develo pment t o a i d i n t h e i ns pect i o n
pro cedure i s t h e develo pment o f co mput eri zed da t a ba s es fo r t ypi ca l
o ccurrences o f da ma g e a nd i ns pect i o n res ult s . A s t ep i n t h i s
di rect i o n h a s been t a ken by C HEVRON i n t h e develo pment a nd
i mplement a t i o n o f C ATS IR (C o mput er Ai ded Ta nker S t ruct ures
Ins pect i o n a nd Repa i r) [40] .
55
3. 6 References
1. Llo yd, J . R. a nd C la ws o n, W. C . , "Res erve a nd Res i dua l S t reng t h
o f P i le F o unded Offs h o re P la t fo rms ", Th e Ro le o f D es i g n,
Ins pect i o n a nd Redunda ncy i n Ma ri ne S t ruct ura l Reli a bi li t y,
P ro ceedi ng s , Na t i o na l Aca demy P res s , Wa s h i ng t o n, D . C . , 1984.
2. de Oli veri a , J . G. a nd Zi mmer, R. A. , "Redunda ncy C o ns i dera t i o n
i n t h e S t ruct ura l D es i g n a nd F lo a t i ng Offs h o re P la t fo rms ", Th e
Ro le o f D es i g n, Ins pect i o n a nd Redunda ncy i n Ma ri ne S t ruct ura l
Reli a bi li t y, P ro ceedi ng s , Na t i o na l Aca demy P res s , Wa s h i ng t o n,
D . C . , 1984.
3. Ni ko la i di s , E. a nd Ka pa ni a , R. K. , "S ys t em Reli a bi li t y a nd
Redunda ncy o f Ma ri ne S t ruct ures : A Revi ew o f t h e S t a t e o f t h e
Art ", J o urna l o f S h i p Res ea rch , Vo l. 34, No . 1, Ma rch 1990.
4. P a li o u, C . , S h i no zuka , M. , a nd C h en, Y. N. , "Reli a bi li t y
Ana lys i s o f Offs h o re S t ruct ures ", Ma ri ne S t ruct ura l
Reli a bi li t y S ympo s i um, S NAME, 1987.
5. No rda l, H. , C o rnell, C . A. a nd Ka ra mch a nda ni , A. , "A S t ruct ura l
S ys t em Reli a bi li t y C a s e S t udy o f a n Ei g h t -Leg S t eel J a cket
Offs h o re P ro duct i o n P la t fo rm", Ma ri ne S t ruct ura l Reli a bi li t y
S ympo s i um, S NAME, 1987.
6. Kli ng muller, O. "Redunda ncy o f S t ruct ures a nd P ro ba bi li t y o f
F a i lure", -3rd. Int . C o nference o n S t ruct ura l S a fet y a nd
Reli a bi li t y, IC OS S AR 1981, No rwa y.
7. D a s , P . K. a nd Ga rs i de, J . F . , "S t ruct ura l Redunda ncy fo r
C o nt i nuo us a nd D i s cret e S ys t ems ", S S C -354, D ecember 1991.
8. S t a i ns en, S . G. , "Int errela t i o n Bet ween D es i g n, Ins pect i o n a nd
Redunda ncy i n Ma ri ne S t ruct ures ", Th e Ro le o f D es i g n,
Ins pect i o n a nd Redunda ncy i n Ma ri ne S t ruct ura l Reli a bi li t y,
P ro ceedi ng s , Na t i o na l Aca demy P res s , Wa s h i ng t o n, D . C . , 1984.
9. Bro ckenbro ug h , R. L. a nd J o h ns t o n, B. G. , "U . S . S t eel D es i g n
Ma nua l", publi s h ed by U . S . S t eel, 1981.
10. F a ulkner, D . , "A Revi ew o f Effect i ve P la t i ng fo r U s e i n t h e
Ana lys i s o f S t i ffened P la t i ng i n Bendi ng a nd C o mpres s i o n",
J o urna l o f S h i ps Res ea rch , Vo l. 19, No . 1, Ma rdi , 1975.
11. J enni ng s , E. , et a l. , "Inela s t i c D efo rma t i o n o f P la t e P a nels ",
S h i p S t ruct ure C o mmi t t ee (S S C -364), 1991.
12. S mi t h , C . S . a nd D o w, R. S . , "Res i dua l S t reng t h o f D a ma g ed S t eel
S h i ps a nd Offs h o re S t ruct ures ", J o urna l o f C o ns t ruct i o na l
S t eel Res ea rch , Vo l. 1, No . 4, S ept ember 1981.
56
13. D o wli ng , P . J . a nd D i er, A. F . , "S t reng t h o f S h i ps ' P la t i ng
under C o mbi ned La t era l Lo a di ng a nd Bi a x i a l P res s ure", C ES LIC
Repo rt S P 6, Imperi a l C o lleg e, Lo ndo n, S ept ember 1978.
14. "D uct i le C o lla ps e", "F a t i g ue a nd F ra ct ure", Vo l. 1,
P ro ceedi ng s o f t h e 10t h Int erna t i o na l S h i p a nd Offs h o re
S t ruct ures , C o ng res s , 1988, D enma rk.
15. Ro lfe, S . T. a nd Ba rs o m, J . M. , "F ra ct ure a nd F a t i g ue C o nt ro l i n
S t ruct ures - Appli ca t i o ns o f F ra ct ure Mech a ni cs ", P ubli s h ed by
P rent i ce - Ha ll, Inc. , 1977, New J ers ey.
16. Th a ya mba lli , A. , C h en, Y. K. a nd Li u, D . , "F ra ct ure Mech a ni cs
ba s ed As s es s ment o f F a t i g ue Reli a bi li t y i n S h i p S t ruct ures ",
pres ent ed a t S h i p S t ruct ure S ympo s i um, Oct o ber 1984,
Arli ng t o n, VA.
17. "S h i p D es i g n Ma nua l - Ma ri ne S t ruct ures ", C h a pt er 6, P a rt 8,
P ubli s h ed by Bri t i s h S h i p Res ea rch As s o ci a t i o n (BS RA).
18. Ri ce, J . R. "A P a t h -Independent Int eg ra l a nd t h e Appro x i ma t e
Ana lys i s o f S t ra i n C o ncent ra t i o n by No t ch es a nd C ra cks ",
J o urna l o f Appli ed Mech a ni cs , Vo l. 35, 1968.
19. S h a rpi es , J . K. , "F i ni t e Element Met h o ds Appli ed t o Two
F ra ct ure Tes t s S peci mens ", Na t i o na l Eng i neeri ng La bo ra t o ry,
NEL Repo rt No . 675, J une 1981.
20. Bri t i s h S t a nda rd Ins t i t ut e, "Gui da nce o n Met h o ds fo r As s es s i ng
t h e Accept a bi li t y o f F la ws i n F us i o n Welded S t ruct ures ",
P D 6493: 1991.
21. Anders o n, T. L. "Ela s t i c-P la s t i c F ra ct ure Mech a ni cs - A
C ri t i ca l Revi ew", S S C -345 - P a rt 1, Apri l 1990.
22. Burdeki n, F . M. a nd D a wes , M. G. "P ra ct i ca l U s e o f Li nea r
Ela s t i c a nd Yi eldi ng F ra ct ure Mech a ni cs wi t h P a rt i cula r
Reference t o P res s ure Ves s els ", P ro ceedi ng s o f t h e Ins t i t ut e
o f Mech a ni ca l Eng i neers C o nference, Lo ndo n, Ma y 1971.
23. Ha rri s o n, R. P . , Lo o s emo re, K. , Mi lne, I. , a nd D o wli ng , A. R. ,
"As s es s ment o f t h e Int eg ri t y o f S t ruct ures C o nt a i ni ng
D efect s ", C ent ra l Elect ri ci t y Genera t i ng Bo a rd Repo rt R/H/R6-
Rev 2, Apri l 1980.
24. Ga rwo o d, S . J . , "A C ra ck Ti p Openi ng D i s pla cement (C TOD ) Met h o d
fo r t h e Ana lys i s o f D uct i le Ma t eri a ls ", AS TM S TP 945, J une
1985.
57
25. Mi lne, I. , Ai ns wo rt h , R. A. , D o wli ng , A. R. a nd S t ewa rt , A. T. ,
As s es s ment o f t h e Int eg ri t y o f S t ruct ures C o nt a i ni ng
D efect s ", C ent ra l Elect ri ci t y Genera t i ng Bo a rd D o cument R/H/R6
- Revi s i o n 3, Ma y 1986.
26. Aus t en, I. M. , "Mea s urement o f F a t i g ue C ra ck Gro wt h Th res h o ld
Va lues fo r us e i n D es i g n", BS C Repo rt No . S H/EN/9708/2/83/B,
Bri t i s h S t eel C o rpo ra t i o n, J a nua ry 1983.
27. Burns i de, O. H. a nd Huda k, J r. S . J . , "Lo ng -Term C o rro s i o n
F a t i g ue o f Welded Ma ri ne S t eels ", pres ent ed a t S h i p S t ruct ure
S ympo s i um, Arli ng t o n, VA, Oct o ber 1984.
28. F o rma n, R. G. , Kea rney, V. E. a nd Eng le, R. M. , "Numeri ca l
Ana lys i s o f C ra ck P ro pa g a t i o n i n a C ycli c-Lo a ded S t ruct ure",
AS ME Tra ns . J . Ba s i c Eng . 89D , 1967.
29. J . M. Ba rs o m, "F a t i g ue-C ra ck Gro wt h U nder Va ri a ble-Ampli t ude
Lo a di ng i n AS TM A514 Gra de B S t eel", AS TM S TP 536, Ameri ca n
S o ci et y fo r Tes t i ng a nd Ma t eri a ls , 1973, P h i la delph i a .
30. "NC HRP P ro ject 12-12: Int eri m Repo rt ", U . S . S t eel Res ea rch
La bo ra t o ry Repo rt 76. 019-001, Na t i o na l C o o pera t i ve Hi g h wa y
Res ea rch P ro g ra m-Na t i o na l Aca demy o f S ci ences , Oct . 1972,
Wa s h i ng t o n, D . C .
31. J . M. Ba rs o m a nd S . R. No va k, "NC HRP P ro ject 12-14: S ubcri t i ca l
C ra ck Gro wt h i n S t eel Bri dg e Members ", U . S . S t eel Res ea rch ,
F i na l Repo rt fo r Hi g h wa y Res ea rch Bo a rd, Na t i o na l C o o pera t i ve
Hi g h wa y Res ea rch P ro g ra m, S ept . 1974, Wa s h i ng t o n, D . C .
32. R. P . Wei a nd T. T. S h i h , "D ela y i n F a t i g ue C ra ck Gro wt h ",
Int erna t i o na l J o urna l o f F ra ct ure, 10, No . 1, Ma rch 1974.
33. Bea , R. G. , "Ma ri ne S t ruct ura l Int eg ri t y P ro g ra m (MS IP )", S S C -
365, Oct o ber 1991.
34. K. T. S ka a r, S . Va ls g a rd, P . E. Ko h eler, C . Murer, "Ho w Lo w C a n
S t eel Wei g h t Go Wi t h S a fet y a nd Eco no my?", Tra ns . 3rd Int .
S ympo s i um o n P ra ct i ca l D es i g n o f S h i ps a nd Mo bi le U ni t s , P RAD S
'87, 1987, No rwa y.
35. Wa rren, D . S . , "D es i g n a nd Ins pect i o n Int errela t i o n o f
C o mmerci a l J et Tra ns po rt S t ruct ure", P ro ceedi ng s o f
Int erna t i o na l S ympo s i um o n t h e Ro le o f D es i g n, Ins pect i o n a nd
Redunda ncy i n Ma ri ne S t ruct ura l Reli a bi li t y, C o mmi t t ee o n
Ma ri ne S t ruct ures , Ma ri ne Bo a rd, Na t i o na l Res ea rch C o unci l,
Na t i o na l Aca demy P res s , 1983, Wa s h i ng t o n, D . C .
36. S a nt o s , D . , "F a t i g ue Ma na g ement fo r t h e A-7P ", F a t i g ue
Ma na g ement : 7 2nd Meet i ng o f t h e AGARD S t ruct ures a nd
Ma t eri a ls P a nels , Apri l-Ma y 1991, U K.
58
37. "Gui da nce Ma nua l fo r t h e Ins pect i o n a nd C o ndi t i o n As s es s ment
o f Ta nker S t ruct ures ", pro duced by Ta nker S t ruct ure
C o o pera t i ve F o rum (TS C F ) a nd publi s h ed by Wi t h erby & C o . Lt d. ,
1992, U . K.
38. Meli t z, D . T. , Ro bert s o n, E. J . a nd D a vi s o n, N. J . , "S t ruct ura l
P erfo rma nce Ma na g ement o f VLC C S - An Owner's Appro a ch ", Ma ri ne
Tech no lo g y, Vo l. 29, No . 4, Oct . 1992.
39. D unn, F . P . , "Offs h o re P la t fo rm Ins pect i o n", P ro ceedi ng s o f t h e
Int erna t i o na l S ympo s i um o n t h e Ro le o f t h e D es i g n, Ins pect i o n
a nd Redunda ncy i n Ma ri ne S t ruct ura l Reli a bi li t y, No vember
1983, Wa s h i ng t o n, D . C .
40. Ti kka , K, a nd D o nnelly, J . W. , "C ATS IR C o mput er Ai ded Ta nker
S t ruct ures Ins pect i o n a nd Repa i r P la nni ng ", pa per pres ent ed a t
t h e Ma ri ne S t ruct ura l Ins pect i o n, Ma i nt ena nce a nd Mo ni t o ri ng
S ympo s i um, Ma rch 1991, Arli ng t o n< VA.
59
4. 0 METHODS TO EVALUATE RESIDUAL STRENGTH OF DAMAGED MARINE
STRUCTURES
4. 1 Indi rect Met h o d
Th e i ndi rect met h o d i s a n a ppro x i ma t e s t ep-by-s t ep pro cedure t o
a s s es s t h e res i dua l s t reng t h o f da ma g ed ma ri ne s t ruct ures . Th i s
s i mple pro cedure, ut i li zes s ervi ce li fe ex peri ence a nd s emi -
empi ri ca l a s s es s ment cri t eri a fo r lo ca l da ma g e a nd us es s i mple
co mput er pro g ra ms , li ke U LTS TR [1], t o det ermi ne t h e ult i ma t e
s t reng t h . Th us a pra ct i ci ng eng i neer ca n ma ke a qui ck a s s es s ment
o f res i dua l s t reng t h wi t h o ut h a vi ng t o perfo rm co mplex 3-
di mens i o na l fi ni t e element a na lys i s . F i g ure 4-1 i s a flo wch a rt o f
t h e pro cedure, wh i ch ba s i ca lly i nvo lves t h ree s t eps . Th e fi rs t
s t ep i s t h e i dent i fi ca t i o n o f t h e cri t i ca l lo ca t i o ns a nd a ppli ca ble
fa i lure cri t eri a fo r da ma g es a t t h es e lo ca t i o ns . Th e nex t s t ep i s
t h e lo ca l a na lys i s t o det ermi ne t h e ex t ent o f da ma g e ba s ed o n t h e
a ppli ed lo a d a nd t h e mo de o f fa i lure. Th e la s t s t ep i s
det ermi na t i o n o f t h e ult i ma t e s t reng t h o f t h e s t ruct ure t a ki ng i nt o
co ns i dera t i o n t h e da ma g e co mpo nent s . Ea ch o f t h es e s t eps i s
di s cus s ed i n t h e fo llo wi ng s ect i o ns .
4. 1. 1 S ervi ce Li fe Ex peri ence
One o f t h e pri ma ry ro les o f eva lua t i o n o f s ervi ce li fe ex peri ence
i s pro vi di ng i nfo rma t i o n o n cri t i ca l da ma g e lo ca t i o ns o n ma ri ne
s t ruct ures . C ri t i ca l lo ca t i o ns a re t h o s e lo ca t i o ns wh ere ei t h er
t h e co ns equences o f da ma g e o r t h e li keli h o o d o f da ma g e o r bo t h a re
very h i g h . Ana lyt i ca lly, a fi ni t e element a na lys i s o f t h e g lo ba l
s t ruct ure ca n be perfo rmed t o i dent i fy cri t i ca l lo ca t i o ns wh ere
ei t h er t h e ma rg i n bet ween s t a t i c dema nd a nd ca pa ci t y i s mi ni mum o r
t h ere i s a s t res s g ra di ent (s t res s co ncent ra t i o n). But s uch a n
effo rt i s ex pens i ve a nd requi res h i g h level o f t ech ni ca l ex pert i s e.
F o r co nvent i o na l ma ri ne s t ruct ures li ke t a nkers , bulk ca rri ers ,
et c. , releva nt i nfo rma t i o n a va i la ble fro m s ervi ce li fe ex peri ence
ex i s t s wh i ch ca n be us ed i n a n i ndi rect o r a ppro x i ma t e eva lua t i o n
o f res i dua l s t reng t h o f ma ri ne s t ruct ures . Th e o t h er purpo s e o f
s ervi ce li fe ex peri ence i s t o pro vi de cri t i ca l cri t eri a fo r
a s s es s i ng res i dua l s t reng t h o r det ermi ni ng i ns pect i o n a nd repa i r
i nt erva ls wi t h res pect t o fa i lure co ns equences .
Alt h o ug h cla s s i fi ca t i o n s o ci et i es h a ve co llect ed a la rg e a mo unt o f
s t ruct ura l da ma g e da t a , releva nt s ervi ce li fe ex peri ence da t a o r
s t a t i s t i cs a re ra rely publi s h ed. Ana lys i s o f s ervi ce li fe
ex peri ence g a t h ered by s h i p o wners a nd ya rds i s co ns i dered
pro pri et a ry. In recent yea rs , due t o effo rt s by va ri o us a g enci es
a nd s o me s h i p o wners , releva nt i nfo rma t i o n o n s ervi ce li fe
ex peri ence h a s been ma de a va i la ble. Int eres t i ng i nfo rma t i o n ma y be
fo und i n references [2-11].
60
CO
C
0>
J
c c
o J= (0
H 1-
CO Oi
QJ C 4-
c 01 O
H c
iH M -H
o> to =
o CD
H o> a
3 3 CO
o en co
H 0)
CO H CO
CO
< u. <
H
c
c *
o
CO
CO CO
01 o < D
c c CO
H ca CO
-4
<
o> c CO
n o c _
H H o CD
3 H H J
a O H C
H cd
u. Q. -O 1-
o < o C
CO c o t-
i- 1-4 o o
2 . .
UJ >.
li. +J
H
c U
CB H
a P
c CO
H CD
-H -H ~
C a. co
O i
c H- >
CO O
o>
Ol
"D * CO c
o ~* a 3
2 -H o + >
H c
a> u H CD
C- 3 a> c
3 a u.
^H
H
CD
U. T- CM
co
a.
CD
o
o
CO -H
I- -I
_l 3
H H
a
09
a
x c
HO
o
z-o
L 1UC
Oca
i-
O a>
M UJ CD
HI-
< <
31(1
_l li 13
< h- C
> -I 3
r
H
CJJ
z
UJ
K
Z
o
to
I-
< h-
o z
O UJ
UJ -J I-
\- I D X
< > UJ UJ
a
M <
H S
-I <
3 Q
z
UJ
co
co
UJ
co
co
<
X
H
(5
Z
UJ
QC
H
CO
<
Q
i i
co
UJ
oc
cr
o
x
<
o
cc
0_
CL
<
Q
UJ
H
<
C
O
UJ
UJ
T
61
As a n ex a mple o f t h e ro le o f s ervi ce li fe ex peri ence,
i nves t i g a t i o ns h a ve recent ly been co nduct ed by t h e S h i p S t ruct ure
C o mmi t t ee, t o develo p a llo wa ble perma nent defo rma t i o n cri t eri a
ba s ed o n bo t h i n-s ervi ce ex peri ence a nd a na lyt i ca l met h o ds . Th e
i nfo rma t i o n pres ent ed i n [12] det ermi ned t h a t ex i s t i ng s h i p pla t e
defo rma t i o n cri t eri a fa ll i nt o t h e fo llo wi ng t wo ca t eg o ri es :
1. new co ns t ruct i o n t o lera nces
2. i n-s ervi ce a llo wa nces
A s umma ry o f t h e new co ns t ruct i o n defo rma t i o n a llo wa nces pres ent ed
i n [12] i s pro vi ded i n t a ble 4-1.
Aut h o rs i n [12], co uld o bt a i n o nly t wo us eful i n-s ervi ce pla t e
defo rma t i o n cri t eri a fro m t h e s urvey depa rt ment o f t h e D nV o ffi ce
a t New J ers ey, wh i ch were a s fo llo ws :
F o r s h ell pla t i ng lo ca t ed i n t h e 0 t o 0. 2 5L (wh ere L = o vera ll
s h i p leng t h ) a nd i n t h e 0. 75L t o 1. 0L po rt i o n o f t h e h ull, t h e
ma x i mum po s s i ble i ndent i s 0. 05 t i mes t h e mi ni mum s pa n leng t h
bet ween s t i ffeners (o r b/20, wh ere b equa ls t h e s t i ffener
s pa n).
F o r mi dbo dy pla t i ng (0. 25L t o 0. 75L), i f t h e o bs erved
defo rma t i o n i s 10 mm t o 3 0 mm i n dept h t h e s h i p o wner i s
no t i fi ed a nd t h e da ma g e i s reco rded, i f t h e o bs erved
defo rma t i o n i s g rea t er t h a n 3 0 mm (a bo ut 1-3/16 i nch es ) t h e
s urveyo r wi ll reco mmend repa i r o r repla cement o f t h e pla t i ng .
F urt h ermo re t h ey co ncluded t h ro ug h i nt ervi ews wi t h s urveyo rs , t h a t
t h ere a re no a na lyt i ca l g ui deli nes fo r ma x i mum i n s ervi ce a llo wa ble
pla t e defo rma t i o n. It wo uld a ppea r t h a t s urveyo rs a re t ra i ned by
o t h er ex peri enced s urveyo rs t o a ccept o r reject a defo rmed pla t e
ba s ed o n a "rule o f t h umb" g ui deli ne a nd no t upo n co mpa ri s o n o f
mea s ured deflect i o ns wi t h es t a bli s h ed deflect i o n cri t eri a . Ta bles
4-2 a nd 4-3 a ls o t a ken fro m [12], pres ent t h e res ult s o f t h e i n
s ervi ce pla t e defo rma t i o n da t a . Ta ble 4-2 s umma ri zes t h e pri nci ple
ch a ra ct eri s t i cs o f t h e s h i ps s urveyed a nd t a ble 4-3 pres ent s t h e
s urvey res ult s .
Wh i le cri t eri a t o prevent lo ca l fa i lure li ke bea m-co lumn buckli ng ,
la t era l-t o rs i o na l i ns t a bi li t y (t ri ppi ng ) a nd pla t e buckli ng a re
well es t a bli s h ed, cri t eri a fo r fra ct ure a re h a rd t o co me by.
Reference [2] g i ves a very co ns erva t i ve cri t eri a fo r t rea t ment o f
fra ct ure. It s ug g es t s "a ny fra ct ures fo und a re no rma lly t o be
repa i red by pa rt renewa l o f ma t eri a l o r by weldi ng ". Th e U . S .
C o a s t Gua rd, i n C OMD TINS T M16000. 7, Ma ri ne S a fet y Ma nua l (MS M),
Vo lume II, s ug g es t s s o me repa i r g ui deli nes fo r fra ct ure ba s ed o n
ex peri ence. Th es e a re:
62
w
E H
H
s
H
J
2
O
O
Pn
W
Q
W
E H
<
ft
o
H
E H
U
OS
E H
CO
O
CJ
2
w
2
I
W
pa
<
E H
* to
O) L
> .Q
XI L U
> 3
b, t- ~
C
5
1:
'^ <a z
: M ~ <
3 tu -
o =
-Q ._
3 CU ~
3 4) ,
O
-O ,
10
<
en
(_
cu
c c
cu .- 3 .-
> _ o
5 i- "a L.
< tu ca aj
Q. w to
(0
to c a
t_ (0 L .
<o o
a. cu u-
o
V )
CO
c- X) c
co o o
a. .a u-
x>
^ '1 cvj
c o
c o
CO
c o
CD '
a.
c to to
O c 3
_J t to
- M .C
OIC II
C CO CO
O I- 3
l/l O W O
O Q
a. c a.*- u
x x CO c
CU Ul UJ
w TJ
. . c
- 3 CM CO PO
TJ.8 x.
CD TJ " CU
to
a
TJ
CU
o = o
a
e
a
X. X
-ON IM
a a
XI "O
-0 CU X> CU
CU CO CU W
cfl o V) o
o o
a. u a o
X c X c
UJ UJ LU LU
1 - to c- CO
to a. co a.
a Q.
XI XI
xi cu XI cu
tu CU 1
V ) o Ul o
o -- o
a u a. u
x c x c
j to:
C- TJ
OJ CU CU
> n t_
o CU
LI c/l >
C c o
JJ ""
u
rJ
CM It
CO TJ HI TJ
a. a* TJ 0J <\>
c_ n t_
0) flj W 01
t- > t/i >
CO o 3 c O
CQ U O " u
TJ
C
CO
cu + -<
TJ
O
(/> XI
xi
3
O 3
CO
to cu
I TJ
-J
.
4-J (TJ
Cl 1
1
(_ TJ
C
<0
t_ (A
<D *
(1 U
) CU
CO u
TJ
c
(D
a w
_,
_. ^
M (J CD
) QJ 2
03 TJ
O O
CU Q. (A
L. O 3
O O O
u- a ^
63
Q
W
> H
W
>
co
co
c u
H
X
co
o
co
u
H
E H
CO
H
w
E H
U
s
<
X
U
H
u
H
04
CM
<
CM
I
w
m
E H
F
u
l
l
L
o
a
d
D
i
s
p
l
a
c
e
m
e
n
t
(
L
o
n
g
T
o
n
s
)
co
r
H
00
o
o
VO
co
in
in
CTi
o
00
o
in
H
VO
in
in
co
in
in
o
in
H
in
CM
co
o
CO
o
in
H
VO
in
CM
CO
o
o
CO
o
o
VO
0>
L
e
n
g
t
h
O
v
e
r
a
l
l
(
m
)
H
co
CM
>*
CM
CM
H
CTl
H
CO
VO
in
H
CO
CO
CM
< *
H
VTl
00
H
1 VO
in
H
CO
CO
H
CM
H
CO
CO
CO
cu
> i
E H
a
H
co
0)
H
U
u
(0
u
4->
< w
(0
rH
H
< :
a
H
3
CO
-P
0
3
co
-P
(0
g
0
CJ
-p
< / )
(0
fa
u
CU
> 1
0
rH
-p
01
a)
Q
a)
rH
10
en
rH
s
T5
0)
T3
H
O
rH
a
H
(0
CJ
-p
(0
u
M
H
< <
!H
cu
> i
0
rH
-p
en
cu
a
cu
rH
H
CO
(0
H
s
T3
cu
a
H
a
H
5
co
0
D>
U
CO
CJ
CU
rH
-H
5
CU
>
H
si
CO
0
o>
> H
CO
u
H
CO
*H
CU
0>
C
CU
10
10
CO
CM
H
43
CO
u
CU
c
CU
10
10
cd
CM
U
cu
> i
0
u
10
CU
Q
cu
r-i
H
(0
CO
H
S
-0
CU
n
H
o
r4
CU
> 1
0
u
+J
CO
CU
a
CU
r-\
H
10
10
H
s
CU
-a
rH
CU
> 1
0
u
-p
to
CU
Q
u
CU
rH
rH
rH
CO
u
4J
CO
r<
rH
H
< :
U
.
S
.
N
a
v
y
D
e
s
i
g
n
a
t
i
o
n
to
vo
1
>
CJ
w
o
<
CO
1
o
a
Q
VO
1
>
U
co
1
o
Q
Q
CO
CM
*
<
1
E H
VO
CO
CM
1
<
1
E H
1 1
H
1
o
Q
a
H
1
o
Q
Q
o>
l
Q
a
rH
1
>
u
H
CO
m
X
> i
-P
p
-H
CO
-p
H
0
u
-p
CU
Q
CO
CO
D
T3
T3
H
CO
CO
D
T3
CU
C
C
CU
CO
CO
C
cu
CP
rH
si
Q
CO
CO
CO
rH
0
XI
cu
c
cu
a
CO
2
CO
D
CO
CT>
CU
>
CO
55
CO
D
a
H
3
CO
r-i
CO
H
*H
CU
s
g
0
u
H
co
CO
H
u
cu
g
g
0
u
c
CO
CO
D
g
CO
Cn
C
> i
C
O
U
CO
CO
D
rH
CU
r-\
> 1
CO
K
CO
CO
-p
rH
CU
>
CU
CO
0
0
04
CO
CO
D
64
2
O
H
E H
<
U
O
w
OH
w
E H
<
&
w
D
W
&
H
(0
(N
I
w
ffl
<
E H
^ 4-<
u 4- 0)
0) c 01 01
< U TJ 9) o '*- c >
c > 01 01 o
2 o > L _ _ 01 c XI
_> O XI o
01 L . c * ra 0)
I_ 10 XI O) 4- 0> c
0) oi a c ra 4-1 L.
+-J XI 4- 2 ra t_ 01 H-
CO H- 4- t_ 2 01 01 01 M u
3 4-1 U~ H- 4J c c 4- ro in 01 0)
**- < o 10 0> CO ra 2 " ra 2 *- JZ
ai O *4- > ro O 01 2 ro VI
>
5
o
ro E L _ 5
ro
CA 01 01 01
c
01
01
>
ro
o
2
c
4-* 01 a o ra
. o ra
1
-j- XI 4J 4-* 1 > o ^~ 01 o
flj 3 c c L _ c 4- ra ro c o XI > VI
o H- - ra 4- ra 2 2 01 01 XI ro 01 o c
4-< .a
_, _^
oi 01 > H- L - 01 > o VI 4- c ro E XI o
H- ra L . c E 4^ 4J 01 4- E r- a 4- 0) a 'F- 4- ro ro a.
0) o> (0 ra VI o 01 a 3 > 2 ^ 4- 4- c VI
OJ 4- 4-* c_ 01 oo c (_ c JZ O o L . 4- 4- 4-
n (U 4- H- ro H-
E
VI XI XI 4- 01 H- 01
oi 2 2 2 O 4J h J 01 TJ ro ra ra 4-t * TJ
+J c 4J O H- ro L _ c ra ro c
D < U 4-* 4- CD XI
"S
ro ~* 01 V) c E 4- 4-1 2 VI L . VI
o Q. < 0 ffl 4-* a N4- ra 4- H- 2 2 a 01 2 c
.Q {/ ) H- 4- 01 ro CM O 4^ O 4-* O 01
ra TJ
*
O. 0) o E 2 J= k > o -o XI ro XI o= V) XI TJ
3 c c c 4-* ra J* V) c VI C
H- [_ c JZ 4- 01 t_ 4- U TJ L _ "D 3
2 oi (11 (A TJ H- ra 01 01 01 2 2 2
4J 4- TJ u O TJ E 4- J: O O O E
t
0) VI 01 CO V) H- ra V) V) XI XI 01 XI 01 ra 01 01 _
'E 4- u O c 01 J: J= J= JZ O
01 ra c 01 i_ O) VI VI V) VI 4-1
JZ 01
"8 2
TJ 01 ra ra
c (/ I TJ u 2 4- c 4- 01 01 01 01 >
o v 4) OI 2 0> 3 . H- V) o 01 01 4-< 01 01 TJ 01 TJ 01 TJ TJ 01
_c c jz JZ H- .* J= 4-1 j= J= o -C J= -C -C
4- (A o en V> u V] t_ % V) V) VI XI VI VI V) VI V) VI VI VI 01
J y> 01 01
U OI c oi 01 4) 01 TJ a oi ' 4- TJ 01 01 TJ 01 01 TJ 01 "U 01 TJ 01 TJ TJ
o a o TJ TJ TJ C L . c oi ro 01 01 L . TJ TJ L . TJ TJ l_ TJ l_ TJ C C L - c
-J CL L . ro -E 2 J= _c ro a ro ra ro (0 ro
(A V) cfl CO * 01 o V) V) V) o V) V) o V) VI o VI o VI O ' o o
o u JZ XI L . 01 XI XI XI XI .a c .Q XI
+ > *- *- 4-> 4-< 4J 01 4J L . 0> M > 4-* 4-* L _ L . *-i t_ c t-i L - 01 l_ c
ta t_ L . l_ L _ C_ 4J 03 (0 4- L . O L . L . ra l_ u ra l_ L. ra L _ ra L - ra 4- ra ro
o O o o O 10 01 4- (D O XI o o 4-> o o t-t o o 4-J o 4- o 4-< ro 4-> 4-*
CL a. o. a. a. a. 2 3 CO 2 a. ra a. a. CO a. a. 00 a. Q_ CO a. OO a. 00 2 00 00
CD
*->
(0
Q
>* o o o o o O O o- o o> t> o> O- o o O o o o- O o> o- o
>
00 00 03 oo 00 03 03 03 CO CO CO CO CO 03 03 C0 co 00 co co 03 03 03
(M C\J rj ro ro (M O o o o C4 ro ro ro ro ro ro
3
t/i
(M ro ro ro r\j ro
ro IO ro ro ro ro in in in in in in in O o- <> o- o- o o o o O
a a a
O
03 j= JZ -C
ro CO CO CO
ro ro r r-- *
NO o ro
-*
ro
* O O
l_
01
L -
01
l_
01
T
r-
*^"
r--
> > -* ~ r- r- i < CO CO OI OI u> CJ o CJ r-. z
u u ro ro ^O -o C3 es ro ro c c c Q o o o >
L 1J UJ o> & a a i 01 01 01 Q o o u
o O o o > > Q Q v V) V) V) vf vj- CD
.* Jl^ < < u CJ < < VJ V) t/ l Q a
2 3 o o a ro ro ra CJ C3 E Q E Q
CD to a a c c i i a. a. a. a a ro ro
a I 3Z 4- 4- o Q > >. 01 01
a Q JZ * j= * 01
a o L . L . oi 1_ oi c >
JC JK > o o 01 01 U) Ol 01 ra ra ra ro ro c 01 c 01 01
t/i 4- 4- L . C 73 T3 c c c en CD O) oi >. >- VI
4-" t-t TJ a c c JZ jr 01 0) 01 u u (j c c c >- c >- o
oi o> 01 01 ra ra a > > u L . c- o CO o ra o
^ a o i i ^ i< : o a
00 00 00
01 01 01 ^ ^ CJ 3= u 3= QC
oo CO 00 CO CO CO CO CO CO 00 z z z CO 00 CO 00 00 CO CO
oo CO CO CO CO CO CO CO CO 00 00 CO 00 o o o CO 00 CO CO 00 CO 00
3 3 13 3 3 3 3 3 3 3 3 3 3 CJ CJ u 3 3 3 3 3 3 3
c
e o ro ro J- in 0D h- CO o o ro ro
< u
CD
X
(M ro -0- in -o c- 03 ON ro ro ro ro
65
w
z
o
o
fa
w
Q
w
z
<
w
<
CU
I*
W
g
D
w
H
K
w
n
I
W
pa
E n
c
o
4-
ro a)
E a
i- >
o t-
H-
1)
Q
*-*
u
ra
u
(0
a
(0
CO
01
>
ra
3
u u
< a
&
Q.
ro
CO
01
10
3
a
ro
CO
01
>
3
*->
u
ra
u
ra
f
+-
o
ra
&
a
ra
CO
01
ra
3
4J
o
ro
ro
o
_i
01
01
3
u
ro
a
E
o
ra
a
E
O)
c
5
c
o
c.
C3
3
u
ra
|
o
ra
a
E
*-
u
ro
a
E
+-
o
ra
D.
E
4->
u
ra
a
E
*^
a
ra
CO
01
>
ro
3
4-J
ro
Q.
E
u
ro
a
E
o
ra
Q.
E
a
ra
CO
01
>
ro
3
C
E
*
21
01
CN
st
st
vl- ea
o
LO
si
in
o
fN
O
CO
si
o
o
si
rO
in
st
f\l
si
CO
si
o
st
o
^O
a
o
-1-
O
in
in
OJ ro
o
co
ro
o
^0
o
CO
CO
fN
fN
vO
O
O
t>
o
ro
o
o
fN
fN
O
01
a
o>
01
*-*
(/ )
CO
CO
CO
CO
*
*
*
*
CO
o
-o
f\J
CO
_J
z
CO
t>
sO
f\J
f\J
CO
_l
x
*
*
*
*
o
CO
> -
CO
CO
01
D
ro
c
CD
CO
CO
<
01
73
ro
c
CD
CO
CO
<
<
QJ
TJ
ra
t-
<j
to
CO
<
*
*
*
*
*
o
CO
> -
2C
CO
CO
CO
CO
: r
CO
o
-o
f\l
fN
CO
_l
CO
CO
CO
t>
sO
(N
r\j
CO
1
+:
+-
o
o
o
o
in
ro
o
en
o
CO
o
en
o
ro
ro
ST
o
ro
ro
o
O
o
o
o
O
CO
d
Is.
ro
si
d
in
st
d
in
is.
co
o
in
fN
rs.
o
CO
in
o
o
d
*
*
CO
ro
vO
o
CO
ro
sf
d
o
o
d
CO
CO
d
in
o
CO
ro
S*
d
ro
ro
d
XI
s*
(I)
in
in
CN
si-
ro
CM
ST ir\
st
st
ro
ro
rs.
o
in
OJ
ro
ro
in
OJ
IN
CO
o
LA
r\j
CO
in
r- is.
IN
CO
in ro
XI sr si o
ro
O
rO
rs
f\J
in CO
St
o CO
IN
o
ro
CO o
ro OJ
fN
ro
G sr
r\j
sf
r\i
CO
ro
CO CO
fN
sr
IN
sr
fN
ID
sr
si-
o
o
o
r\i
vO
CO
<M
CO
ra ro
f\J
st
st
f\j
IN
ro
OJ
ro
o
o
o
ro fN
ro
o
vO
CO
sr
CO fN
in
o
ro
o
ro
+^
c
aj
E
V
l_
(0
111
z
-
r\j ro si in sO rs. CO t>
o
-
(N ro vf in o r- CO o- o
f\J f\J
fN ro
PJ 01
sr S
66
(1) S i de s h ell lo ng i t udi na l wi t h a fra ct ure wh o s e leng t h i s mo re
t h a n o ne-h a lf t h e wi dt h o f t h e lo ng i t udi na l s h o uld be cro pped
o ut a nd renewed i n ki nd.
(2) S i de s h ell lo ng i t udi na l wi t h a fra ct ure wh o s e leng t h i s les s
t h a n o ne-h a lf t h e wi dt h o f t h e lo ng i t udi na l ma y be g o ug ed o ut
a nd rewelded.
(3) Bo t t o m a nd s i de s h ell pla t e t h a t ex peri ences a fra ct ure,
g enera lly, mus t be i ns ert ed. "P o s t a g e s t a mp" i ns ert s a re
di s co ura g ed. Ins ert s o f a t lea s t 18 i nch es a re co ns i dered a
mi ni mum. Go ug i ng o ut a nd reweldi ng o f bo t t o m a nd s i de s h ell
pla t e fra ct ures a re a llo wa ble o n a ca s e by ca s e ba s i s i f t h e
fra ct ure i s rela t i vely s ma ll a nd do es no t ex ceed o ne-h a lf t h e
pla t e t h i cknes s . Ho wever, i f t h ere i s mo re t h a n o ne fra ct ure
i n a t wo fo o t a rea , t h en a n i ns ert i s requi red.
Alt h o ug h i nfo rma t i o n a va i la ble fro m s ervi ce li fe ex peri ence ca n be
us ed a s a s t a rt i ng po i nt fo r res i dua l s t reng t h a s s es s ment , i t s
a ppli ca t i o n h a s li mi t ed s co pe. In ma ny ca s es , a rea s i dent i fi ed a s
pro ne t o da ma g e fro m s ervi ce li fe ex peri ence were no t cri t i ca l,
i . e. t h e da ma g e wo uld no t res ult i n a ca t a s t ro ph i c fa i lure.
S eco ndly t h e cri t eri a fo r a llo wa ble li mi t s , wh i ch a re ba s ed o n
s ervi ce li fe ex peri ence, h a ve been develo ped t o prevent lo ca l
fa i lure. Th e li mi t s a re very co ns erva t i ve s i nce t h ey do no t
a cco unt fo r redunda ncy wh i ch i s i nh erent ly pres ent i n co nt i nuo us
s t ruct ures li ke s h i ps . Th e li mi t s h a ve been ba s ed o n t h e premi s e
t h a t a ny lo ca l fa i lure ca n be t h e ca us e o f g lo ba l fa i lure a nd h ence
i nco rpo ra t e a h i g h fa ct o r o f s a fet y.
4. 1. 2 Eng i neeri ng P ro cedures fo r P erma nent D efo rma t i o n As s es s ment
As s es s i ng t h e i mpa ct o f perma nent defo rma t i o n requi res a n
unders t a ndi ng o f t h e t o t a l lo a d deflect i o n rela t i o ns h i p. F i g ure
4-2 pres ent s a t ypi ca l lo a d deflect i o n rela t i o ns h i p ch a ra ct eri s t i c
o f a pla t e pa nel s ubject ed t o la t era l lo a ds . Eva lua t i ng t h e
beh a vi o r o f a s t i ffened pla t e s t ruct ure s ubject ed t o la t era l lo a ds ,
beyo nd t h e ela s t i c li mi t , requi res t h e ca lcula t i o n o f t h e la rg e
deflect i o n res po ns e (ma x i mum s t ruct ura l s t ra i n) a nd t h e ult i ma t e
s t reng t h res po ns e (li mi t i ng fra ct ure s t ra i n). Met h o ds fo r
ca lcula t i ng t h e li mi t i ng fra ct ure s t ra i n a re di s cus s ed i n s ect i o n
4. 1. 3. Ela s t i c t h eo ry i s a va i la ble but h a s li mi t ed a ppli ca t i o ns
beca us e lo a ds s evere eno ug h t o ca us e la rg e deflect i o ns a ls o ca us e
yi eldi ng . F i g ure 4-3, t a ken fro m [12], pres ent s a t ypi ca l flo w
di a g ra m fo r a s s es s i ng a llo wa ble pa nel defo rma t i o ns t h a t s h o ws t h e
rela t i o ns h i p bet ween t h e predi ct ed ma x i mum s t ruct ura l s t ra i n (e
ma x
)
a nd t h e cri t i ca l fra ct ure s t ra i n (e
fra ct ure
) In t h e ca s e o f eva lua t i ng
perma nent defo rma t i o ns due t o la t era l lo a di ng s , t h e mea s ure o f
res i dua l s t reng t h ca n be t a ken a s t h e fo llo wi ng :
efraomre/e
m
ax( damage) > 1
67
p
ELAS TIC D ES IGN
D ES IGN F GR
F RAC TU RE
W
FIGURE 4.2 LOAD-DEFLECTION CURVE OF A PLATE PANEL
UNDER LATERAL LOAD
68
Ot A
3 S UJ 03 x O
(/ I < Q < i (j
UJ
>
o
m
<
Q
-J
UJ
D
E
F
I
N
E
D
O
P
E
R
A
T
I
O
N
A
L
P
A
R
A
M
E
T
E
R
S
F
L
A
W
S
I
Z
E
,
C
0
S
T
R
E
S
S
,
0
"
S
T
R
A
I
N
.
.
Q
s
o
_l
UJ
CD
o
L
> -
W
z
< ,
>
D
o
a;
i
10
UJ
or
a_
o
z
J 1
W
or
/ I 1
UJ
/J
S^~
c_> -T
< i
u.
o
o
o
c j
z
< 1
UJ
>
or
ID
O
or
C5
UJ
a:
a.
o
o
<
or
in
UJ / ^~~-
=>
\/~~
o A
<
or
b_
, o
b b
UJ
z
2
o
I
<
a:
o
U_
J
Q
_l
Ld
Z
<
a_
<
o
<
CNJ
O
O
^-'
LxJ
o_.
o
o
z
<
en
_L _
m
LJ
h i
UJ
if)
^>
in
< r UJ
OL -
Q: IO
r> i
11 KJ
<
<
u_
rr
CO
./ _
<
o
O
UJ
<
00
69
Th e develo pment o f e
fra cni re
i s ba s ed o n fra ct ure mech a ni cs a nd i s
furt h er des cri bed i n s ubs ect i o n 4. 3. 1, ex a mple 4. Th e develo pment
o f e, ^ i s ba s ed o n no n-li nea r met h o ds o f a na lys i s a nd a funct i o n o f
t h e a llo wa ble perma nent defo rma t i o n. Th es e met h o ds fa ll i nt o t wo
di s t i nct ca t eg o ri es :
1. a ppro x i ma t e met h o ds o f pla s t i c a na lys i s
2. no n li nea r fi ni t e element a na lys i s
Appli ca t i o n o f pla s t i c a na lys i s i s a n i ndus t ry reco g ni zed met h o d t o
eva lua t e t h e res i dua l s t reng t h o f duct i le da ma g ed s t ruct ure due t o
la t era l lo a ds [13]. One o f t h e a t t ra ct i ve fea t ures o f pla s t i c
a na lys i s i s t h a t i t i s o ft en po s s i ble t o o bt a i n rea s o na ble
predi ct i o ns fo r t h e pla s t i c beh a vi o r o f s t ruct ures wi t h much les s
effo rt t h a n i s requi red fo r a co rres po ndi ng no n-li nea r fi ni t e
element a na lys i s .
F o r res i dua l s t reng t h a s s es s ment , t h e s o lut i o ns o f s t i ffened pla t e
pa nel s t ruct ure a re o bt a i ned fro m t h e i nves t i g a t i o n o f energ y
di s s i pa t i o n i n t h e a na lys i s o f ki nema t i ca lly a dmi s s i ble defo rma t i o n
fi elds . Th e equa t i o ns develo ped t o a ppro x i ma t ely predi ct ei t h er
t h e co lla ps e lo a d o r t h e res ult i ng s t ruct ura l s t ra i n fo r a g i ven
perma nent defo rma t i o n a re ba s ed o n t h e fo llo wi ng a s s umpt i o ns :
1. Ri g i d pla s t i c ma t eri a l i dea li za t i o n a s s umes ela s t i c
defo rma t i o ns a re s ma ll co mpa red wi t h pla s t i c
defo rma t i o ns .
2. Th e a s s umed co lla ps e defo rma t i o n (ba s ed o n yi eld li ne
t h eo ry) g i ves a n upper . bo und s o lut i o n. At co lla ps e t h e
yi eld o r h i ng e li nes a re develo ped a t t h e lo ca t i o n o f
ma x i mum mo ment s . Alo ng t h e yi eld li nes co ns t a nt ult i ma t e
mo ment s a nd a x i a l fo rces a re develo ped. Yi eld li nes a re
s t ra i g h t .
3. Equa t i ng t h e ex t erna l energ y due t o t h e a ppli ed lo a d wi t h
t h e i nt erna l energ y di s s i pa t ed due t o t h e fo rma t i o n o f
t h e h i ng e li nes pro vi des a mi ni mum upper bo und s o lut i o n.
F o r a bea m, equa t i ng t h e ex t erna l energ y due t o t h e a ppli ed lo a d
a nd a s s umed defo rma t i o n fi eld wi t h t h e ra t e o f i nt erna l energ y
di s s i pa t ed o f t h e h i ng e g i ves t h e fo llo wi ng :
| pw dx = S Myj + | P ^ dx C
4
-
1
)
0
i =1
0
F o r a s t i ffened pa nel t h i s rela t i o ns h i p beco mes :
70
n
f pw dA = S (M
Y
6j + P
Y
e\ 1. <
4
-
2
)
wh ere, fo r equa t i o ns (4. 1) a nd (4. 2)
p = uni fo rm la t era l pres s ure
1 = leng t h o f t h e bea m
A = a rea o f pla t e
1; = leng t h o f t h e i t h h i ng e li ne
e
;
= co ncent ra t ed ra t e o f elo ng a t i o n a lo ng h i ng e li ne
#. = co ncent ra t ed ra t e o f ro t a t i o n a lo ng h i ng e li ne
M
Y
= fully pla s t i c bendi ng mo ment
P
Y
= fully pla s t i c a x i a l fo rce
w = velo ci t y fi eld a s s o ci a t ed wi t h t h e a s s umed di s pla cement
fi eld
It i s i mpo rt a nt t o no t e t h a t i n t h e a na lys i s o f bea ms o r s t i ffened
pla t es , i n wh i ch o nly t h e bendi ng s t res s es a re co ns i dered, t h e
fa i lure mech a ni s m t a kes t h e fo rm o f a pla s t i c h i ng e o r h i ng e li ne
pa t t ern. In t h i s ca s e "fa i lure" i mpli ed t h e fi rs t defo rma t i o n o f
pla t i ng under t h e lo a d a s s umi ng a ri g i dly perfect ly pla s t i c
ma t eri a l i dea li za t i o n. Th i s res ult ed i n a co lla ps e lo a d
predi ct i o n. Wh en membra ne fo rces a re i nt ro duced t h e wo rd fa i lure
beco mes a mi s no mer. Ins t ea d o f a s i ng le co lla ps e lo a d predi ct i o n
t h ere i s a lo a d deflect i o n rela t i o ns h i p wi t h t h e defi ni t i o n o f
fa i lure beco mi ng t h e des i g ner's deci s i o n i n det ermi ni ng t h a t t h e
perma nent deflect i o n i s permi s s i ble.
[13] pres ent s a g enera l lo a d deflect i o n rela t i o ns h i p fo r bea m,
wh i ch i s :
w
i f? j
o
S d
V
P
c
(4. 3)
wh ere
S = no n-di mens i o na l s h a pe fa ct o r g i ven by (4M
Y
/HP
Y
)
71
d = dept h o f t h e bea m
P = li ne lo a d a ct i ng o n t h e bea m wh i ch i s o bt a i ned by
mult i plyi ng t h e pres s ure t i mes t h e s pa ci ng i t a ct s o n
P
c
= s t a t i c co lla ps e lo a d o f a bea m g i ven a s 16M
Y
/1
2
1 = leng t h o f t h e bea m
w
0
= perma nent s et a s s o ci a t ed wi t h t h e lo a di ng .
[14] pres ent s a g enera l lo a d deflect i o n rela t i o ns h i p fo r a
s t i ffened pa nel:
M
P =
Yx
/
b
e
A
2
3- - 1
a i
48 - + 6
a
fw \
2
v
H
*, j
(4. 4)
wh ere,
P
A
a
b
b,
M
Yx
= des i g n pres s ure
= a rea o f pa nel
= pla t i ng leng t h i n s t i ffened di rect i o n
= pla t i ng leng t h i n uns t i ffened di rect i o n
= effect i ve brea dt h o f ma t eri a l
= yi eld mo ment i n t h e s t i ffened di rect i o n o f pla t e-bea m
s ect i o n (funct i o n o f b
e
)
w
0
= perma nent deflect i o n o f ent i re s h i p pa nel
H
x
= equi va lent h a lf t h i cknes s i n t h e s t i ffened di rect i o n.
Th e full pla s t i c bendi ng mo ment (yi eld mo ment ), M
Yx
, a nd t h e full
pla s t i c yi eld fo rce, P
Yx
a re defi ned belo w:
M
Y
x =
CT
y H
x
2
P
Yx
=
2a
y
K
x
H
x
Th e equi va lent t h i cknes s (2H) h a s been ba s ed o n t h e equi va lent
pla t i ng pro vi di ng t h e s a me yi eld mo ment . Th e res ult i ng s ect i o n
a rea , i f fully effect i ve wo uld g i ve t o o h i g h a va lue fo r P
yx
fo r t h e
equi va lent pla t e. Th e fa ct o r "K
x
" t h en g i ves t h e fra ct i o n o f t h e
72
equi va lent t h i cknes s ba s ed o n bendi ng wh i ch i s t o be effect i ve fo r
ca rryi ng t h e pla s t i c a x i a l fo rce.
4. 1. 3 Eng i neeri ng P ro cedures fo r C ra ck As s es s ment
F o r la rg e s t ruct ures , s ubject ed t o lo w s t res s es , i t i s a ppro pri a t e
t o a pply well es t a bli s h ed Li nea r Ela s t i c F ra ct ure Mech a ni cs (LEF M)
pro cedures . If s t res s es a re a bo ve a ppro x i ma t ely h a lf t h e yi eld
s t reng t h , pla s t i ci t y effect s ca n be s i g ni fi ca nt a nd t h e LEF M met h o d
wi ll be ex t remely no n-co ns erva t i ve. If t h e LEF M a na lys i s do es no t
co nt a i n s o me t ype o f pla s t i ci t y co rrect i o n, i t g i ves no wa rni ng
wh en t h e li nea r ela s t i c a s s umpt i o ns beco me s us pect . It i s perh a ps
bet t er, a s s ug g es t ed i n [15], t o a pply ela s t i c-pla s t i c cri t eri a t o
a ll pro blems ; t h en, t h e a ppro pri a t e pla s t i ci t y co rrect i o ns a re
a cco unt ed fo r. As po i nt ed o ut i n [15], a va i la ble U . S . t ech no lo g y
i n ela s t i c-pla s t i c fra ct ure mech a ni cs ei t h er ca n't be t ra ns la t ed t o
o rdi na ry welded s t eel s t ruct ures li ke s h i ps , beca us e t h ey h a ve been
develo ped i n t h e nuclea r po wer i ndus t ry wh ere rea ct o rs o pera t e a t
s evera l h undred deg rees a bo ve ro o m t empera t ure, o r t h ey a re
s ca t t ered t h ro ug h o ut publi s h ed li t era t ure. Th e Bri t i s h S t a nda rds
Ins t i t ut e's (BS I) publi s h ed do cument , "Gui da nce o n Met h o ds fo r
As s es s i ng t h e Accept a bi li t y o f F us i o n Welded S t ruct ures
P D 6493: 1991", (P D 6493), [16] wa s fo und t o be t h e mo s t up-t o -da t e
publi s h ed do cument co nt a i ni ng pro cedures a nd cri t eri a t o det ermi ne
t h e a ccept a bi li t y o f cra cks a nd fla ws i n welded s t ruct ure.
4. 1. 3. 1 D es cri pt i o n o f P D 6493
1. Levels o f As s es s ment
P D 6493 o ffers a lt erna t i ves fo r t h e t rea t ment o f fra ct ures a t t h ree
levels o f co mplex i t y. Th e ch o i ce o f a pa rt i cula r level i s g o verned
by t h e i nput da t a a va i la ble a nd t h e deg ree o f co mplex i t y t h e us er
i s prepa red t o a do pt . A t h o ro ug h di s cus s i o n o n t h e ba ckg ro und a nd
t h eo ry beh i nd ea ch level o f a s s es s ment ca n be fo und i n [15] a nd i n
references quo t ed i n P D 6493. At ea ch level, fa i lure a s s es s ment
di a g ra ms a re pres ent ed. Th es e a re i nt era ct i o n curves , a cco unt i ng
fo r fa i lure i n t h e fra ct ure a nd co lla ps e mo de. At a ny o f t h es e
levels t h e t rea t ment ma y be ca rri ed o ut ei t h er us i ng LEF M
pro cedures - s t res s i nt ens i t y fa ct o r, K o r t h e cra ck t i p o peni ng
di s pla cement pro cedure - C TOD . A bri ef a cco unt o f t h e di fferences
i n t h e t h ree levels o f a s s es s ment a re ment i o ned belo w:
Level 1: Th e a s s es s ment a t t h i s level i s ba s ed o n t h e C ra ck Ti p
Openi ng D i s pla cement (C TOD ) des i g n curve develo ped i n t h e ea rly
1970's wh i ch ut i li zed fra ct ure mech a ni cs t h eo ry wh ere a va i la ble a nd
reli ed o n empi ri ca l co rrela t i o ns a nd co ns erva t i ve a s s umpt i o ns wh ere
t h eo ry wa s una va i la ble, s peci fi ca lly i n t h e ela s t i c-pla s t i c a nd
fully pla s t i c reg i mes . Th i s level do es no t ex pli ci t ly a ddres s
pla s t i c co lla ps e but i mpli ci t ly s eeks t o a vo i d pla s t i c co lla ps e
t h ro ug h ch ecks o n net s ect i o n yi eldi ng . Th e o ri g i na l curve wa s
mo di fi ed i n t h e mi d-1970's t o a cco unt fo r s t res s co ncent ra t i o n
73
effect s , res i dua i s t res s es a nd fla ws t h a t do no t ex t end ent i rely
t h ro ug h t h e t h i cknes s o f t h e pla t e. Th i s i s mea nt t o be a n i ni t i a l
s creeni ng level a nd h ence t h e res ult s a re very co ns erva t i ve.
Level 2; Als o referred t o a s t h e no rma l level i n P D 6493. It i s
ba s ed o n t h e C ent ra l Elect ri ci t y Genera t i ng Bo a rd (C EGB), U K
a ppro a ch kno wn a s t h e R-6 met h o d. Th i s met h o d a s s umes a n ela s t i c-
perfect ly pla s t i c ma t eri a l beh a vi o r a nd ex pli ci t ly co ns i ders t h e
i nt era ct i o n bet ween fra ct ure a nd co lla ps e fa i lure mo des . Th i s
met h o d ut i li zes a s t ri p yi eld mo del due t o Hea ld et a l. [17]. Th i s
t rea t ment g i ves i mpro ved a ccura cy o ver level 1 wi t h res pect t o
fla ws i n s t res s g ra di ent s a nd res i dua l s t res s effect s . It a ls o
a llo ws t h e us er t h e o ppo rt uni t y t o a s s es s t h e s a fet y ma rg i ns
requi red ra t h er t h a n us e t h e va ri a ble s a fet y fa ct o r i nh erent i n
level 1.
Level 3: Th i s i s a n a dva nced level o f a s s es s ment wh i ch i s no t
no rma lly neces s a ry fo r g enera l s t ruct ura l s t eel a ppli ca t i o ns li ke
s h i ps . It pro vi des g rea t er a ccura cy but i nvo lves g rea t er
co mplex i t y a nd requi res mo re ma t eri a l da t a . It i s pa rt i cula rly
a ppro pri a t e fo r ma t eri a ls o f h i g h s t ra i n h a rdeni ng ca pa ci t y, o r
wh ere a na lys i s fo r s t a ble t ea ri ng fra ct ure i s requi red. F o r
furt h er det a i ls o n t h e us e o f t h i s level o f a s s es s ment , rea ders a re
referred t o P D 6493.
2. Info rma t i o n Requi red fo r As s es s ment
Th e releva nt da t a fo r fra ct ure mech a ni cs a na lys i s o f s t ruct ures
co nt a i ni ng fla ws a re di s cus s ed belo w: F la ws , a s cla s s i fi ed i n
P D 6493 a re: (a ) pla na r fla ws li ke cra cks , (b) no n-pla na r fla ws
li ke ca vi t i es a nd s o li d i nclus i o ns , a nd (c) s h a pe i mperfect i o ns
li ke mi s a li g nment o r i mperfect pro fi le. In t h i s wo rk we wi ll be
dea li ng wi t h cra ck-li ke pla na r fla ws o nly.
(a ) S t res s es : Th ree ki nds o f s t res s es a re co ns i dered i n P D 6493.
Wa ys o f co mbi ni ng t h es e s t res s es depend o n t h e level o f a s s es s ment .
Th e t h ree ki nds o f s t res s es a re: (a ) pri ma ry s t res s es wh i ch
i nclude membra ne, P
m
, a nd bendi ng , P
b
, co mpo nent s , (b) s eco nda ry
s t res s es , Q, a nd (c) pea k s t res s , F . S ch ema t i ca lly, t h ey a re s h o wn
i n fi g ure 4-4.
Th e membra ne co mpo nent o f t h e pri ma ry s t res s , P
m
, i s t h e s t res s due
t o t h e i mpo s ed lo a di ng wh i ch i s uni fo rmly di s t ri but ed a cro s s t h e
cro s s -s ect i o n. Th e bendi ng co mpo nent , P
b
, va ri es a cro s s t h e
t h i cknes s . In P D 6493, P
b
i s s uperi mpo s ed upo n P
m
. S eco nda ry
s t res s es , Q, a re s elf equi li bra t i ng s t res s es , li ke t h erma l a nd
res i dua l s t res s es wh i ch , a cco rdi ng t o P D 6493, do no t ca us e pla s t i c
co lla ps e, s i nce t h ey a ri s e fro m s t ra i n/di s pla cement li mi t ed
ph eno mena , but co nt ri but e t o s everi t y o f lo ca l co ndi t i o ns a t a
cra ck t i p. F o r fra ct ure a s s es s ment a t level 2, t h e s eco nda ry
s t res s es ma y be di vi ded i nt o membra ne (Q
m
) a nd bendi ng (Q
b
)
74
PRIMARY
STRESS
< P
m
.
p
b)
SECONDARY PEAK STRESS
STRESS DUE TO
GROSS STRESS
CONCENTRATION
(Q) (Fi)
PEAK STRESS
DUE TO
MISALIGNMENT
TOTAL STRESS
(F
2
) ( p
m
+ p
n
+ o +
F
+
F
)
(a)
(b)
(c)
(d)
(e)
FIGURE 4.4 SCHEMATIC REPRESENTATION OF STRE
DISTRIBUTIONS ACROSS SECTION^
75
co mpo nent s . P ea k s t res s es , F , a re t h e s t res s es due t o
co ncent ra t i o ns a t lo ca l di s co nt i nui t i es . Th ey h a ve a pea k va lue
us ua lly a t t h e s urfa ce nea r t h e di s co nt i nui t y wh i ch deca ys ra pi dly
t o zero t h ro ug h t h e t h i cknes s . In level 1 a s s es s ment , t h e ma x i mum
t ens i le s t res s , a
u
i s t a ken t o be uni fo rm a nd equa l t o t h e ma x i mum
s um o f t h e va lues o f t h e s t res s co mpo nent s (P
m
+ P
b
+ Q + F ) a s
defi ned a bo ve. In level 2, a ct ua l s t res s di s t ri but i o ns a re us ed i f
kno wn. If a pa rt i a l s a fet y fa ct o r a ppro a ch i s bei ng us ed t h e
s t res s es s h o uld be mult i pli ed by t h e a ppro pri a t e fa ct o rs wh i ch a re
des cri bed la t er. Th e pea k s t res s , F , i s no t requi red a t level 2
beca us e pea k s t res s es a re i ncluded i n t h e s t res s i nt ens i t y
ma g ni fi ca t i o n fa ct o rs o r t h e s t res s co ncent ra t i o n fa ct o rs .
In level 1 a s s es s ment , t h e do cument reco mmends a t ens i le res i dua l
s t res s va lue i n t h e a s welded co ndi t i o n, equa l t o t h e yi eld
s t reng t h o f t h e pa rent ma t eri a l. F o r s t res s reli eved weldment s ,
t h e es t i ma t ed res i dua l s t res s ca n be reduced. In level 2
a s s es s ment , wh ere a ct ua l di s t ri but i o n o f t h e res i dua l s t res s i s
kno wn, t h e do cument s ug g es t s us i ng t h a t , o t h erwi s e a uni fo rm
di s t ri but i o n o f ma g ni t ude equa l t o t h e yi eld s t reng t h i s a s s umed.
Th e do cument a ls o a llo ws fo r t h e reduct i o n o f t h e res i dua l s t res s
va lue, a
R
t o t h e lo wer o f a
y
o r (1. 4 - aja
f
) o
y
wh ere, a
y
i s t h e
yi eld s t reng t h o f t h e ma t eri a l, a
n
i s t h e effect i ve net s ect i o n
s t res s a nd a
f
i s t h e flo w s t reng t h o f t h e ma t eri a l wh i ch i s a s s umed
t o be t h e a vera g e o f yi eld s t reng t h a nd t ens i le s t reng t h . S o me
t ypi ca l di s t ri but i o n o f res i dua l s t res s es a t welds ca n be fo und i n
P D 6493.
(b) F la w D es cri pt i o n: No rma lly, fla ws a re a s s umed t o be
elli pt i ca l i n s h a pe. Th e do cument reco mmends t h a t pla na r fla ws be
i dea li zed by t h e h ei g h t a nd leng t h o f t h ei r co nt a i nment rect a ng le.
Th es e di mens i o ns , 2a fo r t h ro ug h fla ws , a a nd 2c fo r s urfa ce fla ws
a nd 2a a nd 2c fo r embedded fla ws a re s h o wn i n fi g ure 4-5. If t h e
pa rt i a l s a fet y fa ct o r a ppro a ch i s a do pt ed i n level 2, t h es e
di mens i o ns a re t o be mult i pli ed by a ppro pri a t e fa ct o rs wh i ch a re
di s cus s ed la t er.
(c) Ma t eri a l P ro pert i es :
(i ) F ra ct ure To ug h nes s : P D 6493 a llo ws t h e us e o f fra ct ure
t o ug h nes s da t a wh i ch co uld be ba s ed o n K-met h o ds , i ncludi ng
co nvers i o n fro m J -met h o ds , o r o n C TOD met h o ds , but requi res a
co ns i s t ent us e o f da t a t h ro ug h o ut t h e a s s es s ment pro ces s . Th e
do cument do es no t permi t co nvers i o n o f C TOD t o ug h nes s da t a t o
equi va lent K da t a . Th e repres ent a t i ve t o ug h nes s va lues us ed i n t h e
do cument fo r t h e a s s es s ment a re t ermed K
ma t
fo r K-met h o ds a nd < 5
ma t
fo r
C TOD met h o ds . Th e do cument emph a s i zes t h e des i ra bi li t y o f h a vi ng
a ccura t e t o ug h nes s da t a a nd s ug g es t s t h e us e o f es t a bli s h ed
s t a nda rdi zed t es t i ng t ech ni que i n t h ei r deri va t i o n. Th e pro cedure
fo r t h es e t es t i ng t ech ni ques a re di s cus s ed i n det a i l i n t h e
do cument . If va li d K
IC
va lues fo r K
ma t
a re no t a va i la ble, t h e
76
2c
i_
T r
(a) THROUGH THICKNESS FLAW
REQUIRED DIMENSIONS: 2a, B
(b) EMBEDDED FLAW
REQUIRED DIMENSIONS: 2a, B, 2c, p
r
B
1 _
2c
(b) SURFACE FLAW
REQUIRED DIMENSIONS: a, B, 2c
(d) FLAW AT TOE OF FILLET WELD
REQUIRED DIMENSIONS: a, B, 2c
(e) FLAW AT HOLE
REQUIRED DIMENSIONS: a, B, c
V V
(f) SURFACE EDGE CRACK
REQUIRED DIMENSIONS: a, W
FIGURE 4.5 FLAW DESCRIPTION FOR CRACK ASSESSMENT
(16)
77
do cument s ug g es t s ei t h er us i ng C TOD a s s es s ment ro ut es o r t o deri ve
equi va lent K
a t
va lues fro m J -i nt eg ra l ba s ed met h o ds o f fra ct ure
t o ua h nes s t es t i ng . Th e J
ma t
va lue o bt a i ned fro m t es t i ng ma y be
t o ug h nes s t es t i ng . i ^ "
ma t
co nvert ed t o a n equi va lent K
ma t
va lue us i ng
Kma t -
EJ .
ma t
(1 " v
2
)
(4. 5)
Th e t o ug h nes s va lues s h o uld be di vi ded by t h e a ppro pri a t e fa ct o rs ,
i f pa rt i a l s a fet y fa ct o r a ppro a ch i s us ed i n level 2 a s s es s ment .
() Tens i le P ro pert i es : F o r bo t h level 1 a nd level 2 a s s es s ment
met h o ds , t h e ma t eri a l yi eld s t reng t h (a
y
) , t ens i le s t reng t h (a
u
) a nd
mo dulus o f ela s t i ci t y (E) a t t h e a ppro pri a t e t empera t ure a nd
o bt a i ned fro m s t a nda rd s o urces s h o uld be us ed.
3. P a rt i a l S a fet y F a ct o rs o n S t res s , F la w S i ze a nd To ug h nes s
Th e pa rt i a l s a fet y fa ct o rs a re ba s ed o n a ccept a ble level o f
pro ba bi li t y o f fa i lure wh i ch i n t urn depends o n t h e co ns equences
o f t h e fa i lure o f t h e co mpo nent . Two t ypes o f fa i lure co ns equences
a re defi ned, mo dera t e a nd s evere. Th e fo llo wi ng t a ble g i ves t h e
defi ni t i o n o f t h es e co ns equences .
F a i lure
S t ruct ura l effect
co ns equences
Mo dera t e
Lo ca l fa i lure
wh i ch wo uld no t
ca us e co mplet e
s t ruct ura l
co lla ps e (e. g . ,
redunda nt member)
S evere
Ri s k o f co mplet e
s t ruct ura l
co lla ps e o r s evere
h a za rd
F o r t h e t wo t ypes o f fa i lure co ns equences , reco mmenda t i o ns i n t h e
do cument fo r pa rt i a l s a fet y fa ct o rs t o be a ppli ed t o t h e bes t
es t i ma t e (mea n) va lues o f ma x i mum t ens i le s t res s es a nd fla w s i zes
a nd t o t h e ch a ra ct eri s t i c va lue o f fra ct ure t o ug h nes s a re g i ven i n
t a ble 4-4.
78
TABLE 4-4 P ARTIAL S AF ETY F AC TORS F OR AS S ES S MENTS AT LEVEL 2
D a t a va ri a ble
S t res s (kno wn a ccura t ely
mea s ured) (C OV 5%) 7,
S t res s (es t i ma t ed) (C OV
30%)
7s
F la w s i ze (S D 2 mm t o
5 mm) 7
a
F la w s i ze (S D 10 mm) 7
a
To ug h nes s ex pres s ed a s
K
ma t
(mi ni mum o f t h ree
res ult s ) 7,
To ug h nes s ex pres s ed a s
< ?
ma t
(mi ni mum o f t h ree
res ult s ) 7,
P a rt i a l fa ct o rs fo r fa i lure
co ns equences
Mo dera t e
1. 1
1. 2
1. 0
1. 1
1. 0
1. 0
S evere
1. 4
1. 6
1. 2
1. 4
1. 2
1. 4
No t e: S D i s t h e s t a nda rd devi a t i o n.
4 P a ra met ers Requi red fo r As s es s ment :
(a ) S t res s Int ens i t y F a ct o r, K, :
F o r t h e a s s es s ment level 1, t h e do cument pro vi des t wo fo rmula e,
dependi ng o n t h e t ype o f fla w, fo r co mput i ng t h e s t res s i nt ens i t y
fa ct o r.
(i ) fo r t h ro ug h t h i cknes s fla ws ,
Kj = a
l
\fnz (4. 6)
wh ere,
cr, i s t h e ma x i mum t ens i le s t res s (P
m
+ P
b
+ Q + F )
2a i s t h e fla w leng t h
79
(i i ) fo r pa rt i a l t h i cknes s fla ws ,
1
$
1 V
wh ere,
$ i s t h e elli pt i c i nt eg ra l, wh i ch ca n be a ppro x i ma t ed
a s :
* = { 1. 0 + 1. 464(a /c)
165
}
05
fo r a /c < 1
$ = { 1. 0 + 1. 464(c/a )
165
}
05
fo r 1 < a /c < 2
M, , , i s t h e fla w s h a pe fa ct o r fo r fla ws under a x i a l
t ens i o n.
In t h e level 2 a ppro a ch t h e do cument s ug g es t s us i ng ei t h er:
(i ) h a ndbo o k s o lut i o ns [18, 19], fi ni t e element a na lys i s o r wei g h t
funct i o n t ech ni ques , a s des cri bed i n reference [20] fo r ca lcula t i ng
va lues o f s t res s i nt ens i t y fa ct o r, o r
(i i ) us i ng t h e g enera l equa t i o n
K, = (Yo )v^ (
4
-
8
>
wh ere a i s h a lf fla w leng t h fo r t h ro ug h t h i cknes s fla ws a nd full
fla w h ei g h t fo r s urfa ce fla ws ;
and Ya = ( 1/*) {^( M^P,, + Q
m
) + M
b
( M
kb
P
b
+ Q
b
)} ( 4-9)
wh ere, M, , , a nd M
b
, kno wn a s fla w s h a pe fa ct o rs fo r fla ws under
a x i a l t ens i o n a nd bendi ng res pect i vely, a re t h e
co rrect i o n funct i o ns , dependent o n cra ck s i ze a nd s h a pe,
pro x i mi t y o f t h e cra ck t i p t o free s urfa ces a nd lo a di ng .
Mi & M
kb
a re t h e s t res s co ncent ra t i o n fa ct o rs due t o
lo ca l s t ruct ura l di s co nt i nui t i es t o be a ppli ed t o t h e
membra ne a nd bendi ng co mpo nent s o f t h e pri ma ry s t res s .
S o me pa ra met ri c fo rmula e a nd fi g ures fo r M
m#
M
b
, M^, a nd M
kb
a re
pres ent ed i n t h e do cument a nd ca n a ls o be fo und i n s t a nda rd
h a ndbo o ks o n fra ct ure mech a ni cs .
80
(b) C ra ck Ti p Openi ng D i s pla cement s s /:
U s i ng s t res s i nt ens i t y fa ct o r K, defi ned i n t h e previ o us s ect i o n,
t h e a ppli ed C TOD < S j i s det ermi ned a s :
F o r level 1,
*i =
K;
fo r -< 0. 5
(4. 10)
*i =
K; (a W
VV
- - 0. 25 for -! > 0. 5
v y
wh ere a
y
i s t h e yi eld s t reng t h a nd E i s t h e Yo ung 's Mo dulus ;
F o r level 2,
1
^
(4. 11)
(c) P a ra met er K
r/
S
r
a nd S
r
:
F o r us i ng t h e fa i lure a s s es s ment di a g ra ms i n a ny level o f
a s s es s ment , pa ra met ers K
r
, 6 a nd S
r
a re requi red. P a ra met er K
r
i s
t h e ra t i o o f t h e s t res s i nt ens i t y fa ct o r K, t o t h e fra ct ure
t o ug h nes s , K^. S i mi la rly, S
t
i s t h e ra t i o o f S
x
t o < S
ma t
. Wh ere
s eco nda ry s t res s es a re pres ent , i n level 2 a s s es s ment , a pla s t i ci t y
co rrect i o n fa ct o r, p, h a s t o be us ed t o a llo w fo r pla s t i ci t y
i nt era ct i o n o f pri ma ry & s eco nda ry s t res s s uch t h a t :
K
T
K
<
=
K
' mat
(4. 12)
V^
(
i
N
mat;
(4. 13)
81
F i g ures fo r det ermi ni ng p, a re g i ven i n P D 6493.
P a ra met er S
r
i s t h e pla s t i c co lla ps e ra t i o , a nd i s defi ned a s aJo
{
,
wh ere a
n
i s t h e net effect i ve s ect i o n s t res s . F o rmula s fo r o
n
fo r
s o me s i mple co nfi g ura t i o ns ca n be fo und i n t h i s do cument .
References [21, 22] pro vi de a
n
fo r mo re co mplex g eo met ry. a
f
i s t h e
flo w s t reng t h , wh o s e va lue i s a s s umed a s t h e les s er o f 1. 2a
y
o r t h e
a vera g e o f t h e yi eld a nd t ens i le s t reng t h .
5. Accept a bi li t y o f Kno wn F la ws :
F o r a kno wn fla w, t h e pa ra met ers K
r
o r < S
r
a nd S
r
a re ca lcula t ed a nd
t h e res ult i ng po i nt i s plo t t ed o n t h e fa i lure a s s es s ment di a g ra ms
s h o wn i n fi g ure 4-6 fo r t h e t wo levels o f a s s es s ment . If t h e po i nt
li es wi t h i n t h e a s s es s ment curve, t h en t h e fla w i s a ccept a ble.
6. Es t i ma t i o n o f Accept a ble F la w S i ze:
In t h e do cument , a ccept a ble fla w s i ze i s defi ned i n t erms o f a n
eg ui va lent fla w pa ra met er, a
m
, wh i ch i s t h e h a lf leng t h o f a
t h ro ug h t h i cknes s fla w i n a n i nfi ni t e pla t e s ubject t o remo t e
t ens i o n lo a di ng . Th i s fla w pa ra met er a cco unt s fo r a va ri et y o f
fla w s h a pes a nd s i zes a t t h e s a me level o f s everi t y. Equi va lent
pa rt t h i cknes s fla ws ca n be o bt a i ned fro m fi g ures i n P D 6493.
F o r a s s es s ment level 1, us i ng K_- met h o ds , t h e fo llo wi ng equa t i o n
g i ves t h e a ccept a ble fla w s i ze a
m
:
1 (K*)
2
*
m
2TT a ,
(4. 14)
U s i ng C TOD - met h o d,
m
8
ma
tE
2n
(
a
i\
2
\y/
for _1 < 0.5
a
y
( 4.15)
a
m =
6
mat
E
2TT
\y
0.25
for -1 > 0.5
a
82
1 n
0.8
1 1 1 1 1
ASSESSMENT LINE
0.6
/
[
-
o 0.4
i
^
0.2 -
n
1 l i
0 0.2 0.4 0.6 0.8 1.0 1.2
c~ _,
b
r "^
(A) - LEVEL 1 FAILURE ASSESSMENT DIAGRAM
-1 o
1.0
0.8
1 1 1 1 1
/- ASSESSMENT LINE
~~~^x -
\
k
^ 0.6
\
o 0.4
\
\
0.2 -
-
n
1 1 1 1 II
0 0.2 0.4 0.6 0.8 1.0 1.2
Sr -
(B) - LEVEL 2 FAILURE ASSESSMENT DIAGRAM
(16)
FIGURE 4.6 FAILURE ASSESSMENT DIAGRAMS
83
Th e res ult i ng fla w di mens i o n h a s t o be ch ecked fo r pla s t i c co lla ps e
by co mput i ng t h e a s s o ci a t ed S
r
va lue. If S
r
i s les s t h a n 0. 8, t h e
fla w i s a ccept a ble, i f i t i s g rea t er, t h en t h e fla w di mens i o ns h a ve
t o be reduced t o g i ve a n S
r
les s t h a n 0. 8. In t h e a bo ve equa t i o n,
t h e s t res s ra t i o , aJa
YI
ca n be repla ced by a n a ppli ed s t ra i n ra t i o
e
x
/e
Y
up t o a li mi t o f 2. F o r s t ra i n ra t i o ex ceedi ng 2, ela s t i c-
pla s t i c fi ni t e element a na lys i s i s t o be us ed t o o bt a i n s t ra i n
i nfo rma t i o n. In level 2, fa i lure due t o pla s t i c co lla ps e i s
ex pli ci t ly t a ken ca re o f i n t h e deri va t i o n o f t h e fa i lure
a s s es s ment curves , h ence no i mpli ci t ch eck i s requi red o n S
r
. Th e
a ccept a ble fla w s i ze i s det ermi ned i t era t i vely s uch t h a t t h e
res ult i ng fla w s i ze s a t i s fi es t h e fo llo wi ng equa t i o n:
S
5
m^
m
\(AHfC
\2
P
L
( 4.16)
7t (Yof
A flo w ch a rt fo r cra ck a s s es s ment ba s ed o n t h e s t eps di s cus s ed
a bo ve i s pres ent ed i n fi g ure 4-7.
4. 1. 4 Two D i mens i o na l Met h o ds fo r U lt i ma t e S t reng t h As s es s ment
Th i s s ect i o n pro vi des a s umma ry o f t h e 2-di mens i o na l met h o ds
a va i la ble t o det ermi ne t h e ult i ma t e s t reng t h o f s h i p s t ruct ures .
Th es e met h o ds a re ba s ed o n t h e "co mpo nent " a ppro a ch . Th e g enera l
i dea o f t h e co mpo nent a ppro a ch wa s fi rs t pres ent ed by C a ldwell i n
1965 i n h i s pa per ent i t led "U lt i ma t e Lo ng i t udi na l S t reng t h " [23].
In t h i s a ppro a ch , h ull g i rder s t reng t h i s ca lcula t ed o n t h e ba s i s
o f i ndi vi dua l s t reng t h beh a vi o r o f t h e co mpo nent s ma ki ng up t h e
h ull g i rder. S i nce C a ldwell's pa per, pa ra llel effo rt s ba s ed o n t h e
co mpo nent a ppro a ch were undert a ken by bo t h t h e U . S . [1, 24] a nd U K
Na vi es [25]. Th es e effo rt s led t o t h e develo pment o f t wo co mput er
pro g ra ms , U LTS TR by Ada mch a k a t D a vi d Ta ylo r Res ea rch C ent er (D TRC )
a nd a s i mi la r pro g ra m by C . S mi t h a nd a s s o ci a t es a t t h e Admi ra lt y
Res ea rch Es t a bli s h ment (ARE) i n t h e U K.
In t h e co mpo nent a ppro a ch , t h e h ull cro s s -s ect i o n i s di s cret i zed
i nt o ei t h er g ro s s -pa nel element s o r h a rd co rner element s a nd
curva t ure i n s ma ll fi ni t e i ncrement s i s a ppli ed t o i t . Th e g ro s s -
pa nels co uld be s i ng le lo ng i t udi na l s t i ffener wi t h t h e a s s o ci a t ed
pla t i ng (pla t e-bea m co mbi na t i o ns ) o r a s t i ffened pa nel ma de up o f
a number o f pla t e-bea m co mbi na t i o ns . A h a rd co rner i s a pa rt i cula r
s t ruct ura l det a i l wh i ch i s a s s umed t o res i s t a ny fo rm o f buckli ng
under co mpres s i ve lo a ds . Ha rd co rners a re us ua lly i nt ers ect i o ns o f
ma jo r pla t e element s e. g . , decks /bulkh ea ds t o s h ell pla t es , et c.
Wi de pla t es wh i ch a re lo ng i t udi na lly uns t i ffened, a s fo und i n
t ra ns vers ely fra med s h i ps , a re co ns i dered i n bo t h t h e ARE a nd
84
O ik
CO
>
H > ty
en n UJ
>
H t O)
Co -H
t- -a >
o c H
M t- CO
CM
UJ
>
,
3
l b
IS
f li-
0
es
0
k
3
a
r
H fl
e
+J Q. 0
W
e
o
UJ a?
c
0 < D
H H
(0 J3
C (0
-^
E a
H CD
13
O
5
CO C
H=)
1
r
li.
S ik
c
o
_v
w o
c o
b
H JZ "O
"
L
(0 4-i O w
/ CD \
3
r
ai / o '
-H
eg
k
3
k /
V P r \
0 M a
\ co /
4-> {_ 3 E
3 fl) a 0
ar Q E o
E 4-* O
O -H h-
o
o
< 3)
1
Q
L
*
0 L .
o
HC 0
h
C
L
H o
k.
CV
k CO
0)
F F
< D CD 0 H
+ >
fi E D
3 Of 0
3
Q.
E
E C
a C (D E
CD
E -H +J O
u
0 CO i < D
]!CE
3)
C
J
4J
co
c C *
H ai
E < D
4J (0 c_
C CO 3
TH < u Oi
o CO -H
a. co L L
<
< u *
c V E
4-1 L ca
3 C_
4J HO)
O HCO
Hca -H
0. li. Q
CO
i*i
<
cc
C\J
u_
7
o
L U
h- O
z. H
L U L U
CC
CO
L U
CO L L
UJ L U
CO
CC
CO
<
Si
O
c CC
o
L i-
L L
CI
L U
L U
c c L L
3 O
a
_l
L U L U
>
o
L U
oc O
a. '
r-
31
-H
J3 ^f
a
Q.
L U
a> CC
o
Z5
<
1 I
+J
u_
o
a>
4-
< s
a
85
U LTS TR pro g ra ms . Ea ch i ncrement o f curva t ure i s a s s umed t o pro duce
a li nea r s t ra i n di s t ri but i o n t h ro ug h t h e dept h o f t h e cro s s -
s ect i o n. Th e lo ca t i o n o f zero s t ra i n co rres po nds t o t h e
"i ns t a nt a neo us " o r "i ncrement a l" neut ra l a x i s . At ea ch va lue o f
curva t ure, t h e pro g ra m eva lua t es t h e equi li bri um s t a t e o f ea ch
g ro s s pa nel a nd h a rd co rner element rela t i ve t o i t s s t a t e o f s t res s
a nd s t a bi li t y co rres po ndi ng t o i t s pa rt i cula r va lue o f s t ra i n. It
t h en co mput es t h e t o t a l mo ment o n t h e cro s s -s ect i o n by s ummi ng t h e
mo ment co nt ri but i o ns (s t res s x effect i ve a rea x lever a rm) o f a ll
o f t h e element s t h a t ma ke up t h e s ect i o n. In t h i s ma nner a mo ment -
curva t ure rela t i o ns h i p i s defi ned. S i nce t h e s t res s di s t ri but i o n,
unli ke t h a t o f s t ra i n, i s no t neces s a ri ly li nea r a cro s s t h e dept h
o f t h e s ect i o n, t h e lo ca t i o n o f t h e i ns t a nt a neo us neut ra l a x i s mus t
be det ermi ned i n a n i t era t i ve fa s h i o n fro m t h e co ndi t i o n t h a t t h e
net a x i a l fo rce o n t h e cro s s -s ect i o n mus t be zero . Th i s fo rce i s
co mput ed i n t h e s a me fa s h i o n a s t h e bendi ng mo ment , t h a t i s , by
s ummi ng t h e co nt ri but i o ns o f a ll t h e element s o f t h e cro s s -s ect i o n.
Bes i des s t ruct ura l yi eldi ng o t h er co mmo n fo rms o f duct i le fa i lure
mo des o f g ro s s pa nel element s a re:
(1) Inela s t i c flex ura l co lumn buckli ng o f s t i ffeners a nd
a t t a ch ed pla t i ng .
(2) La t era l-t o rs i o na l buckli ng (t ri ppi ng ) o f s t i ffeners .
(3) Overa ll g ri lla g e buckli ng i nvo lvi ng bendi ng o f t ra ns vers e
fra mes - t h i s fo rm o f fa i lure i s no rma lly a vo i ded by
pro vi s i o n o f s t i ff t ra ns vers e fra mes a nd i nt ermedi a t e
s uppo rt fro m bulkh ea ds & pi lla rs .
P la t e buckli ng i s no t i ncluded a s a n ex pli ci t fa i lure mo de but i s
i nco rpo ra t ed t h ro ug h effect i ve wi dt h . In t h e ARE met h o d a nd i n
U LTS TR, reg a rdles s o f t h e s peci fi c t ype o f fa i lure mo de, t h e
g enera l na t ure o f a n element 's beh a vi o r i s des cri bed i n t erms o f
"lo a d-s h o rt eni ng " curves , a lt h o ug h t h e a ppro a ch t o deri vi ng t h e
curves a re di fferent .
In U LTS TR, fa i lure by bea m-co lumn buckli ng ca n be o f t wo t ypes .
Type I i s ch a ra ct eri zed by a ll la t era l defo rma t i o ns o ccurri ng i n
t h e s a me di rect i o n a nd Type II i s ch a ra ct eri zed by a n a lt erna t i ng
buckli ng pa t t ern. Lo a d s h o rt eni ng curves fo r t h es e t wo fa i lure
mo des a re deri ved a na lyt i ca lly ba s ed o n t h e res pect i ve fa i lure
t h eo ri es t a ki ng i nt o co ns i dera t i o n t h e effect o f pla s t i ci t y. Th e
curves fo r t ri ppi ng fa i lure a re deri ved i n a s i mi la r fa s h i o n but
i g no re pla s t i ci t y effect s . D uri ng t h e deri va t i o n, a n i ni t i a l
s i nus o i da l di s t o rt i o n wa s a s s umed a nd o t h er i mperfect i o ns li ke
fa bri ca t i o n i nduced res i dua l s t res s wa s a cco unt ed fo r by
i nco rpo ra t i ng a s emi -empi ri ca l res i dua l s t reng t h reduct i o n fa ct o r
i n t h e ca lcula t i o n o f effect i ve wi dt h o f a s s o ci a t ed pla t i ng . S i nce
t h e effect i ve wi dt h o f pla t i ng i s a s s umed t o be lo a d dependent , i t
i s co mput ed i t era t i vely. Two s h o rt co mi ng s o f t h e met h o d a re t h e
86
g ro s s a ppro x i ma t i o ns a nd educa t ed g ues s es ma de i n mo deli ng t h e
po s t -buckli ng beh a vi o r a nd la ck o f co ns i dera t i o n o f t ri ppi ng
fa i lure mo de i n t h e ela s t o -pla s t i c ra ng e.
In t h e ARE met h o d, t h e lo a d-s h o rt eni ng curves fo r bea m-co lumn
co mpo nent s were o bt a i ned us i ng a la rg e-di s pla cement i ncrement a l
fi ni t e element a na lys i s a ppli ed t o a bea m - co lumn i dea li za t i o n a s
i llus t ra t ed i n fi g ure 4-8. Th e i dea li zed s t ruct ure i s , i n ea ch
ca s e, s ubdi vi ded a lo ng i t s leng t h i nt o element s a s s h o wn i n fi g ure
4-8a , element s bei ng furt h er s ubdi vi ded o ver t h ei r dept h i nt o
"fi bers " a s s h o wn i n fi g ure 4-8b. Th e t a ng ent s t i ffnes s o f t h e
pla t i ng wa s es t i ma t ed fro m t h e s lo pes o f s t res s -s t ra i n curves , a s
s h o wn i n fi g ure 4-9, wh i ch were repres ent ed numeri ca lly i n t h e
co mput er. Two a dja cent i nt erfra me s pa ns , ea ch h a vi ng a cent ra l
pla ne o f s ymmet ry, were mo deled wi t h a s i mple s uppo rt co ndi t i o n a t
t ra ns vers e fra mes . Th i s i dea li za t i o n a cco unt ed fo r t h e i mpo rt a nt
"co nt i nuo us bea m" i nt era ct i o ns bet ween a dja cent s pa ns a nd co rrect ly
repres ent ed t h e pro g res s i ve s h i ft i n neut ra l a x i s o f t h e cro s s -
s ect i o n ca us ed by yi eldi ng a nd lo ca l buckli ng o f t h e pla t i ng .
In fi g ure 4-9, fo r va ri o us va lues o f pla t e s lendernes s , wh ere =
(b/t ) J (cr
y
/E) , effect i ve s t res s -s t ra i n curves h a ve been deri ved wh i ch
a cco unt fo r buckli ng , yi eldi ng a nd effect s o f ei t h er mea n o r s evere
i mperfect i o n. Th es e curves were deri ved numeri ca lly [26, 27].
S i mi la r curves h a ve been deri ved fo r pla t es i n t ra ns vers e
co mpres s i o n [28-31], bi a x i a l lo a d [28, 29, 32] a nd fo r t h e ca s e o f
co mbi ned i npla ne s t res s a nd la t era l pres s ure [33, 3 4]; fo r pla t es
under co mbi ned di rect a nd s h ea r s t res s es , effect i ve s t res s -s t ra i n
curves h a ve been ba s ed o n t h eo ret i ca l s t udi es [35]. Th e ma i n
li mi t a t i o n i n t h e ARE met h o d o ut li ned a bo ve i s t h a t , i t do es no t
t a ke i nt o a cco unt ela s t o -pla s t i c t ri ppi ng o f s t i ffeners i n t h e po s t
co lla ps e ra ng e.
Two i mpo rt a nt a s s umpt i o ns i n t h e ult i ma t e s t reng t h pro g ra ms
develo ped ba s ed o n t h e co mpo nent a ppro a ch a re:
(1) C o lla ps e o f t h e h ull g i rder res ult s fro m a s equence o f fa i lure
o f lo ca l s t ruct ura l co mpo nent s ra t h er t h a n fro m a n o vera ll
s i mult a neo us i ns t a bi li t y o f t h e co mplet e cro s s -s ect i o n. Th us
i t i s a s s umed t h a t t h e buckli ng lo a ds a re wi dely s epa ra t ed a nd
t h ere i s no i nt era ct i o n bet ween va ri o us lo ca l fa i lure mo des .
(2) Th e h ull-g i rder i s t rea t ed a s a bea m s uch t h a t wh en a n
i ncrement a l curva t ure i s a ppli ed t o t h e h ull, a s i s do ne i n
t h e co mpo nent a ppro a ch , a li nea r s t ra i n di s t ri but i o n i s
a s s umed t o ex i s t a cro s s t h e dept h o f t h e s ect i o n. Wh i le t h i s
i s a perfect ly s o und a s s umpt i o n wh i ch i s s uffi ci ent ly va li d
fo r s t res s level belo w t h e s o ca lled des i g n va lue, i t i g no res
t h e t wo di mens i o na l s t a t e o f s t res s t h a t ex i s t s due t o t h e
pres ence o f t ra ns vers e s h ea r i n t h e h ull.
87
PLANE OF SYMMETRY PLANE OF SYMMETRY
LONGITUDINAL GIRDER
(a) SUBDIVISION OF STIFFENED PANEL INTO ELEMENTS
SHOWING ASSUMED INITIAL DEFORMATION
PLATE TREATED AS SINGLE FIBRE WITH
STIFFNESS DEFINED BY SLOPE OF
APPROPRIATE STRESS-STRAIN CURVE
(SEE FIGURE 9)
(b) SUBDIVISION OF CROSS SECTION INTO "FIBRES"
CTr/cTy
i 1 1 r
COMPRESSIVE STRESS
IN PLATING
H
^
L
TENSILE YIELD
IN PLATING
TENSILE YIELD
IN STIFFNER
(c) ASSUMED RESIDUAL STRESS DISTRIBUTION
FIGURE 4.8 BEAM-COLUMN IDEALIZATION OF STIFFENED PANELS
(25)
88
AVERAGE IMPERFECTIONS
w 0/ t=o.i/ 3
2
. CTR / aY =o.:
LOWER-BOUND STIFFNESS
1.0
SEVERE IMPERFECTIONS
"o / ' > o.2/ 3
2
. aR /aY>o.5
FIGURE 4.9 STRESS-STRAIN CURVES FOR PLATE ELEMENTS
UNDER LONGITUDINAL TENSION AND COMPRESSION
89
(25)
To predi ct res i dua l s t reng t h a nd li fe o f da ma g ed ma ri ne s t ruct ures ,
i t i s i mpo rt a nt t h a t ult i ma t e s t reng t h pro g ra ms be ca pa ble o f
i nco rpo ra t i ng t h e effect s o f da ma g e a nd t h ei r g ro wt h , li ke t h e
g ro wt h o f fa t i g ue cra cks . C o ns equences o f da ma g e a re lo s s o f
s t i ffnes s a nd s t reng t h o f t h e lo ca l co mpo nent wh i ch res ult s i n t h e
o vera ll reduct i o n o f ult i ma t e h ull g i rder s t reng t h . F o r s evere
da ma g e, li ke h o li ng o r g ro s s defo rma t i o n wh ere i t ca n be a s s umed
t h a t t h e da ma g ed s t ruct ure i s fully i neffect i ve i n ca rryi ng a ny
furt h er lo a d, i t wo uld be qui t e a ppro pri a t e t o mo del t h e wh o le
s t ruct ure mi nus t h e a ffect ed a rea . Th i s a ppro a ch ca n a ls o be
a do pt ed duri ng a "wo rs t -ca s e s cena ri o " s t udy s i nce t h e res ult s a re
bo und t o be co ns erva t i ve. Wh ere da ma g e i s les s s evere, li ke t h a t
a ri s i ng fro m a ba dly ex ecut ed drydo cki ng res ult i ng i n la t era l
i nt erfra me defo rma t i o n, delet i o n o f t h e a ffect ed a rea ma y be t o o
s evere a nd t o o co ns erva t i ve. F o r s uch ca s es , a rea li s t i c
a s s es s ment o f res i dua l s t i ffnes s a nd s t reng t h i s o bt a i ned by
i mpo s i ng a predet ermi ned a mo unt o f la t era l defo rma t i o n i n t h e
element s i n t h e a ffect ed a rea .
F o r pro g ra ms ba s ed o n t h e "co mpo nent a ppro a ch ", li ke U LTS TR, wh i ch
o nly a llo ws mo deli ng o f t h e lo ng i t udi na l co mpo nent s ma ki ng up t h e
cro s s -s ect i o n, t h e effect o f da ma g e i n t ra ns vers e members ca nno t be
s i mula t ed. Only i n ca s es wh ere da ma g e t o t ra ns vers e members i s s o
s evere t h a t i t a ffect s t h e lo ng i t udi na l s t reng t h by i ncrea s i ng t h e
leng t h o f a n uns uppo rt ed s t i ffener o r pa nel, ca n t h e effect o f a
da ma g ed t ra ns vers e member be det ermi ned. S i mi la rly, U LTS TR wo uld
no t be a ble t o g i ve a t rue pi ct ure o f s t ruct ura l redunda ncy fo r
cert a i n da ma g ed h ull co nfi g ura t i o ns , fo r ex a mple do uble h ull
t a nkers wh i ch h a ve po t ent i a l fo r s t res s redi s t ri but i o n i n t h ei r
t ra ns vers e s t reng t h members .
F o r cert a i n t i me dependa nt da ma g e, li ke fa t i g ue cra cki ng , t h e
res i dua l s t reng t h ch a ng es o ver t i me. Hence t o det ermi ne i ns pect i o n
i nt erva ls , o r t o predi ct t h e effect o f a n ex i s t i ng cra ck a t a
fut ure t i me, cra ck g ro wt h a na lys i s mus t be ca rri ed o ut . No ne o f
t h e ex i s t i ng ult i ma t e s t reng t h pro g ra ms h a ve a ny pro vi s i o ns fo r
cra ck g ro wt h a na lys i s . S ervi ce li fe ex peri ence h a s i ndi ca t ed t h a t
(1) mo s t o f t h e fa t i g ue cra cks o ccur i n s eco nda ry s t ruct ura l
members , (2) very ra rely h a ve t h es e cra cks res ult ed i n ca t a s t ro ph i c
fa i lure due t o bri t t le fra ct ure. Hence, i n a n i ndi rect a ppro a ch ,
t h e bes t wa y t o co ns i der s uch da ma g e i s a t wo s t ep pro cedure. In
t h e fi rs t s t ep, fo r ea ch cra cked lo ca l s t ruct ure, a fra ct ure
mech a ni cs ba s ed a s s es s ment i s ca rri ed o ut t o det ermi ne t h e
a ccept a bi li t y o f t h e cra ck ba s ed o n cert a i n cri t eri o n li ke P D 6493
[16]. Th e s eco nd s t ep wo uld be ult i ma t e s t reng t h a na lys i s o f t h e
wh o le s t ruct ure wh ere, dependi ng o n t h e res i dua l s t reng t h o f t h e
da ma g ed co mpo nent , i t i s ei t h er mo deled i n i t s o ri g i na l fo rm o r
co mplet ely di s ca rded.
In s umma ry t h e i ndi rect met h o d i s ba s ed o n numero us a s s umpt i o ns a nd
i s very co ns erva t i ve. S peci fi c li mi t a t i o ns o f t h e i ndi rect met h o d
a re:
90
1. It i s ba s ed o n t h e a s s umpt i o n t h a t t h e lo ca t i o n a nd fo rm o f
lo ca l da ma g e wi ll be a va i la ble fro m s ervi ce li fe ex peri ence.
Wh i le t h i s i s t rue fo r co nvent i o na l s h i p s t ruct ures , i t ma y
no t be t rue fo r s t ruct ures li ke o ffs h o re pla t fo rms o r o t h er
uni que s t ruct ures wh i ch la ck da t a o n s ervi ce li fe ex peri ence.
2. Li nea r bea m t h eo ry a s s umpt i o ns a re ma de wh i ch a s s ume t h a t t h e
lo ng i t udi na l co mpo nent s ma ki ng up t h e s ect i o n i s under
uni a x i a l s t a t e o f s t res s . But i n rea li t y, due t o t ra ns vers e
s h ea r o ver t h e dept h o f a s ect i o n a nd t wi s t i ng o ver t h e leng t h
o f t h e s t ruct ure, co mpo nent s a re under a bi -a x i a l s t a t e o f
s t res s .
3. No ex pli ci t co ns i dera t i o n fo r t ra ns vers e s t reng t h members li ke
t ra ns vers e bulkh ea ds a nd h ence lo s s o f s t ruct ura l redunda ncy
a t t i mes o f po t ent i a l lo a d di s t ri but i o n.
4. D a ma g es li ke cra cks a re mo deled by remo vi ng da ma g ed po rt i o ns
wi t h o ut a ny co ns i dera t i o n fo r t h e res i dua l s t i ffnes s o r
s t reng t h t h e member s t i ll po s s es s .
5. Int era ct i o n o f fa i lure mo des a re no t a llo wed fo r.
4. 2 D i rect Met h o ds
Key s t eps i nvo lved i n t h e di rect met h o d o f a na lys i s a re:
i dent i fi ca t i o n o f po t ent i a lly cri t i ca l lo ca t i o ns , det ermi na t i o n o f
t h e t ype a nd ex t ent o f da ma g e a t t h es e lo ca t i o ns a nd t h e ult i ma t e
s t reng t h a na lys i s o f t h e g lo ba l s t ruct ure. Beca us e o f t h e
li mi t a t i o ns o f t h e i ndi rect met h o d, co ns erva t i ve a s s umpt i o ns h a ve
t o be ma de a t ea ch s t ep wi t h reg a rd t o t h e s t res s fi eld, ex t ent o f
da ma g e, et c. Th es e a s s umpt i o ns t a ke i nt o a cco unt di s co nt i nui t y i n
t h e pa s s a g e o f da t a a nd i nfo rma t i o n bet ween t h e s t eps . In t h e
di rect met h o d t h es e s t eps a re i nt eg ra t ed a nd perfo rmed us i ng li nea r
a nd no n-li nea r 3-di mens i o na l fi ni t e element a na lys i s . Th e fi ni t e
element a na lys i s i s ca rri ed o ut a t bo t h t h e lo ca l a nd g lo ba l level
i n a n i t era t i ve fa s h i o n. Res ult s o f li nea r g lo ba l s t res s a na lys i s
a re ut i li zed a t t h e lo ca l level fo r no n-li nea r fra ct ure a na lys i s .
Out put o f t h e fra ct ure a na lys i s i s i nt eg ra t ed i nt o t h e no n-li nea r
ult i ma t e s t reng t h a s s es s ment o f t h e g lo ba l s t ruct ure.
U s ua lly s h i ps a re des i g ned t o h a ve s uffi ci ent s t reng t h ma rg i n.
Nevert h eles s i n a ra t i o na l pro cedure fo r res i dua l s t reng t h
a s s es s ment , h i g h s t res s ed a rea s h a ve t o be i dent i fi ed fo r
po s s i bi li t i es o f fa i lure a t t h es e lo ca t i o ns . Th erefo re a 3-D
li nea r fi ni t e element a na lys i s o f t h e wh o le s t ruct ure i s perfo rmed.
Lo a di ng s fo r t h es e a na lys es co ns i s t o f qua s i -s t a t i c s t i ll wa t er
bendi ng mo ment s under va ri o us ca rg o lo a di ng co ndi t i o ns a nd
a llo wa ble wa ve bendi ng mo ment s . As o bs erved i n s ect i o n 2. 0, s h i p
s t ruct ura l co mpo nent s a re well des i g ned a g a i ns t lo ca l duct i le
fa i lure li ke buckli ng o r t ri ppi ng o f s t i ffeners . F a t i g ue cra cki ng
wa s fo und t o be a do mi na nt fo rm o f da ma g e i n s h i p s t ruct ures . To
91
i dent i fy lo ca t i o ns s us cept i ble t o fa t i g ue cra cki ng , s h i p mo t i o ns
a nd lo a ds pro g ra ms a re us ed i n co njunct i o n wi t h fi ni t e element
pro g ra ms . Th i s h a s been demo ns t ra t ed by Hug h es et a l. i n [36]
wh ere, a li nea r, 3-di mens i o na l s h i p mo t i o ns a nd lo a d pro g ra m ca lled
S ma ll Ampli t ude Mo t i o n P ro g ra m (S AMP ), wa s i nco rpo ra t ed i n MAES TRO
t o co mput e t h e fa t i g ue i nduci ng lo a ds due t o t h e s ea wa y. MAES TRO,
i s a fi ni t e element ba s ed s t ruct ura l a na lys i s pro g ra m t a i lo red fo r
full s ca le, 3-di mens i o na l mo deli ng o f s h i ps .
Ha vi ng det ermi ned lo ca t i o ns pro ne t o cra cki ng , ca lcula t i o ns a re
perfo rmed t o det ermi ne t h e t i me t a ken by t h e cra cks t o ca us e
fa i lure o f t h e co mpo nent a nd t h e ex t ent o f cra cki ng . Va ri o us
co mput er pro g ra ms a re a va i la ble wh i ch ca n predi ct cra ck pro pa g a t i o n
i n s t ruct ura l co mpo nent s under va ri a ble a mpli t ude s pect ra l lo a di ng s
[37-39]. Th es e pro g ra ms ei t h er us e t h e P a ri s equa t i o n fo r
ca lcula t i ng cra ck g ro wt h ra t e o r i t s mo di fi ca t i o ns li ke t h e
F o rma n's equa t i o n. Th es e pro g ra ms requi re a s i nput , s t res s
i nt ens i t y fa ct o rs o f t h e cra cked s t ruct ure. F ra ct ure mech a ni cs
ba s ed fi ni t e element s ca n be us ed i n t h es e lo ca t i o ns fo r t h e
det ermi na t i o n o f t h e s t res s i nt ens i t y fa ct o r, K, . Va ri o us t ypes o f
t h es e element s o f va ryi ng co mplex i t y a nd ea s e o f us e [40-43] h a ve
been develo ped i n no n-ma ri ne i ndus t ry. Th ey ca n be a ppli ed fo r
cra cks i n t h i n s h eet ma t eri a l a s well a s fo r cra cks i n co mplex 3-
di mens i o na l co nfi g ura t i o n i nvo lvi ng mi x ed mo de lo a di ng a nd ma t eri a l
no n-li nea ri t y. To det ermi ne t h e ex t ent o f t h e cra ck, a
s uffi ci ent ly det a i led 3-di mens i o na l fi ni t e element a na lys i s o f t h e
reg i o n h a s t o be perfo rmed. D ependi ng o n t h e lo ca t i o n o f t h e
cra ck, t h e lo ca l a na lys i s s h o uld i nclude effect s o f lo ca l s t res s
ra i s ers li ke s t ruct ura l di s co nt i nui t i es a nd mi s a li g nment . Th e
ex t ent o f t h e lo ca l reg i o n s h o uld i nclude a ll a rea s wh i ch a re
s t res s ed a s much o r h i g h er t h a n t h e a rea i n wh i ch t h e cra ck i s
s i t ua t ed. Th i s i s do ne t o preclude a ny ch a nces o f t h e cra ck
pro pa g a t i ng i nt o a h i g h er s t res s ed a rea a nd lea di ng t o uns t a ble
fra ct ure. Bo unda ry co ndi t i o ns fo r t h i s lo ca l a na lys i s i s o bt a i ned
fro m t h e g lo ba l fi ni t e element a na lys i s . D et ermi na t i o n o f cri t i ca l
cra ck leng t h ca n be ba s ed o n co des li ke P D 6493 [16]. D ependi ng o n
t h e ma g ni t ude o f t h e s t res s a nd t h e ma t eri a l t ype, i f ex t ens i ve
yi eldi ng o ccurs i n t h e vi ci ni t y o f t h e cra ck, no n-li nea r fi ni t e
element a na lys i s s h o uld be perfo rmed t o co mput e s t ra i ns t o
det ermi ne cri t i ca l cra ck leng t h s .
To i nt eg ra t e t h e effect o f lo ca l fa i lure upwa rd t o t h e g lo ba l
s t ruct ura l level a nd ens ure t h a t co rrect lo a d t ra ns fer pa t h s a nd
i nt erna l lo a d di s t ri but i o ns a re ca lcula t ed, a full 3-di mens i o na l
no n-li nea r fi ni t e element a na lys i s i s ca rri ed o ut . Th i s i s do ne i n
co njunct i o n wi t h t h e lo ca l fa i lure a na lys i s i n a n i t era t i ve ma nner.
A full 3-di mens i o na l no n-li nea r a na lys i s a llo ws a n a s s es s ment o f
t h e s i mult a neo us effect o f di fferent mo des o f fa i lure a t va ri o us
lo ca t i o ns (lo ng i t udi na l a nd t ra ns vers e) o n t h e o vera ll s t reng t h o f
t h e s t ruct ure. No n-li nea r fi ni t e element pro g ra ms , develo ped i n
t h e ma ri ne i ndus t ry, s peci fi ca lly fo r ult i ma t e s t reng t h a s s es s ment
92
o f s h i p s t ruct ures a re a va i la ble. Th es e ult i ma t e s t reng t h pro g ra ms
a re ba s ed o n t h e "uni fi ed" a ppro a ch . U nder t h e uni fi ed a ppro a ch ,
met h o ds t o eva lua t e ult i ma t e s t reng t h a re pri ma ri ly ba s ed o n
s t a nda rd fi ni t e-element pro cedures wh i ch a llo ws fo r t h e i nt era ct i o n
o f lo ca l a nd g lo ba l fa i lure mo des , g eo met ri c a nd ma t eri a l no n-
li nea ri t i es , i mperfect i o ns due t o fa bri ca t i o n a nd res i dua l s t res s
effect s a nd t h ree di mens i o na l effect s o f t h e h ull g i rder. Th ree
s uch pro g ra ms , U lt i ma t e S t reng t h Ana lys i s o f S t ruct ure (U S AS ) by
ABS [44, 45], F i ni t e Element No n-li nea r C o lla ps e/co lli s i o n
(F ENC OL), by D nV [46, 47] a nd Idea li zed S t ruct ura l U ni t Met h o d
(IS U M) by U eda & Ra s h ed i n J a pa n [48] a re di s cus s ed h ere.
Th e U lt i ma t e S t reng t h Ana lys i s o f S t ruct ures , U S AS , develo ped by
t h e Ameri ca n Burea u o f S h i ppi ng (ABS ) [44], i s a fi ni t e element
ba s ed pro g ra m wh i ch t a kes i nt o a cco unt no n-li nea ri t i es i n g eo met ry
a nd ma t eri a l. It h a s t h ree ki nds o f element s wh i ch t h e us er ca n
us e t o mo del t h e s h i p s t ruct ure, bea ms , i s o t ro pi c pla t es a nd
o rt h o t o pi c pla t e element s . Th e co ns t i t ut i ve rela t i o ns h i ps fo r t h e
element s were deri ved t a ki ng i nt o a cco unt ela s t o -pla s t i c beh a vi o r
wi t h pro vi s i o n fo r s t ra i n h a rdeni ng , mo dera t ely la rg e ro t a t i o ns i n
t h e s t ruct ura l element s , lo ca l a nd o vera ll buckli ng o f s t ruct ura l
element s a nd t h ei r po s t -buckli ng s t reng t h a nd t ra ns vers e s h ea r.
Bes i des s t a t i c a na lys es , i t ca n a ls o be us ed fo r dyna mi c a na lys i s .
No a pri o ri a s s umpt i o ns a re ma de reg a rdi ng lo a d di s t ri but i o n a nd
bo unda ry co ndi t i o ns . Imperfect i o ns i n t h e s t ruct ure, li ke i ni t i a l
defo rma t i o ns , ca n be mo deled ex pli ci t ly, but t h e pro g ra m do es no t
t a ke i nt o a cco unt res i dua l s t res s es due t o fa bri ca t i o n.
Appli ca t i o ns o f U S AS ca n be fo und i n [45] wh ere, t h e a ut h o rs h a ve
us ed U S AS t o eva lua t e t h e ult i ma t e s t reng t h o f va ri o us s h i p
co nfi g ura t i o ns a nd di s cus s mo deli ng t ech ni g ues fo r effi ci ent no n-
li nea r fi ni t e element a na lys i s .
F ENC OL (F i ni t e Element No n-li nea r C o lla ps e/C o lli s i o n) i s a s peci a l
vers i o n o f t h e g enera l purpo s e no n-li nea r pro g ra m F ENRIS . Element s
a va i la ble a re t wo -no ded ba r, t wo no ded bea m a nd fo ur no ded
qua dri la t era l membra ne element s . In F ENC OL, t h e element 's res po ns e
i s ba s ed o n t h e "lo a d ch a ra ct eri s t i c" curves wh i ch a re s i mi la r t o
t h e "lo a d s h o rt eni ng " curves us ed i n t h e co mpo nent a ppro a ch . Th es e
lo a d-ch a ra ct eri s t i c curves develo ped a s a res ult o f ex t ens i ve
pa ra met ri c s t udi es [49], fo r di fferent ki nds o f lo a di ng t a ke i nt o
a cco unt lo ca l buckli ng a nd i mperfect i o ns . S t reng t h cri t eri a us ed
a re t h o s e no rma lly g i ven i n t h e des i g n co des . Th e buckli ng fa i lure
cri t eri a us ed a re t h o s e fro m C N30. 1 [50] a nd g enera li za t i o n o f
i nt era ct i o n fo rmula e i n reference [51]. F o r t h e po s t -co lla ps e
s t reng t h o f s t i ffened pla t e pa nels , i nfo rma t i o n fro m D nV [52, 53]
a nd C a mbri dg e U ni vers i t y a re us ed. F ENC OL co ns i ders a t ens i o n-
t ea ri ng rupt ure mo de o f fa i lure. It des cri bes t ens i o n-t ea ri ng wi t h
a C ra ck Ti p Openi ng D i s pla cement (C TOD ) cri t eri o n [54]. Tens i o n
t ea ri ng rupt ure o ccurs wh en a defect , li ke a t h ro ug h t h i cknes s
cra ck, i n t h e t ens i o n fla ng e o f t h e h ull i s o f a s i ze s uch t h a t t h e
cri t i ca l rupt ure s t ra i n i s i n t h e o rder o f t h e yi eld s t ra i n.
F ENC OL a llo ws t wo ki nds o f mo dels . S ma ll mo dels , a ls o kno wn a s
93
o ne-fra me mo dels , t h e lo ng i t udi na l ex t ent o f wh i ch i s o ne element
a nd la rg e mo dels wh i ch a re ba s i ca lly full 3-di mens i o na l mo dels wi t h
a number o f element s i n t h e lo ng i t udi na l a s well a s t ra ns vers e
di rect i o n. Th e 3-di mens i o na l mo dels ca n be us ed i n t h o s e ca s es
wh ere t h ere i s a h i g h po s s i bi li t y o f s t res s redi s t ri but i o n i n t h e
lo ng i t udi na l a s well a s t ra ns vers e s t reng t h members , a s i n t h e ca s e
o f do uble bo t t o m s t ruct ures , a nd h ence pro vi de a bet t er pi ct ure o f
s t ruct ura l redunda ncy.
Th e Idea li zed S t ruct ura l U ni t Met h o d (IS U M) i s a numeri ca l met h o d
pro po s ed by U eda a nd Ra s h ed [48] t o mo del no n-li nea r beh a vi o r o f
la rg e s i zed redunda nt s t ruct ure reli a bly a nd a ccura t ely i n much
les s t i me t h a n. a co nvent i o na l no n-li nea r fi ni t e. element a na lys i s .
In t h i s met h o d, t h e s t ruct ure t o be a na lyzed i s mo deled us i ng very
la rg e s i ze element s na med "i dea li zed s t ruct ura l uni t s ". U s ua lly,
ea ch ba s i c s t ruct ura l member fo r ex a mple a g i rder (bea m) bet ween
t wo vert i ca l web s t i ffeners o r a s t i ffened pa nel bo unded by fo ur
pri ma ry s uppo rt i ng members , i s ch o s en a s a n i dea li zed s t ruct ura l
uni t . Geo met ri c a nd ma t eri a l no n-li nea r beh a vi o r o f co mpo nent s o f
t h e s t ruct ura l uni t , s uch a s buckli ng o f pla t e element s , co lla ps e
o f fla ng es o f s t i ffeners , i mperfect i o ns , et c. , i s i dea li zed a nd
des cri bed i n co nci s e fo rms rela t ed t o fo rces a nd di s pla cement s o f
a li mi t ed number o f no da l po i nt s a t t h e bo unda ri es o f t h e
s t ruct ura l uni t . Th es e co nci s e fo rms a re a s et o f fa i lure
i nt era ct i o n s urfa ces a nd a s et o f s t i ffnes s ma t ri ces ex pres s i ng t h e
beh a vi o r o f t h e uni t befo re a nd a ft er fa i lure. IS U M h a s been
a ppli ed t o va ri o us t ypes o f s t eel s t ruct ures s uch a s t ra ns vers e
ri ng s [55, 56], upper decks , [57, 58], do uble bo t t o m s t ruct ures
[59-61], null g i rder [62-64] a nd t ubula r o ffs h o re s t ruct ures [65-
72] .
In s umma ry, t h e di rect met h o d i s t h e mo s t ra t i o na l a nd co mplet e
pro cedure fo r res i dua l s t reng t h a s s es s ment . Ex cept fo r t h e
a s s umpt i o ns i nh erent i n fi ni t e element pro cedures , i t i s ba s ed o n
very few a s s umpt i o ns . Th eo ret i ca lly, i t o verco mes a ll li mi t a t i o ns
o f t h e i ndi rect met h o d a nd t h e a ccura cy o f t h e res ult s a re li mi t ed
by t h e level o f s o ph i s t i ca t i o n a do pt ed i n t h e a na lys i s . Wh i le
t h eo ret i ca lly t h e di rect met h o d i s very s o und, a few pra ct i ca l
dra wba cks o f t h e met h o d a re:
1. It i s a t i me-co ns umi ng a nd co mput a t i o na lly i nt ens i ve
pro cedure;
2. It requi res a co ns i dera ble a mo unt o f t ech ni ca l ex pert i s e t o
det ermi ne t h e level o f s o ph i s t i ca t i o n requi red i n mo deli ng t h e
s t ruct ure t o ex t ra ct t h e des i red i nfo rma t i o n;
3. Beca us e o f (1) a nd (2) a bo ve, i t ca nno t be us ed a s a qui ck
a s s es s ment pro cedure a nd h a s li mi t ed a ppli ca t i o n.
94
4. 3 S umma ry o f C urrent Indus t ry P ra ct i ce
Th i s s ect i o n pro vi des a n o vervi ew o f t ypi ca l met h o ds a nd t o o ls
current ly us ed by bo t h t h e ma ri ne a nd no n-ma ri ne i ndus t ri es t o
qua nt i fy t h e effect o f lo ca l da ma g e o n t h e g lo ba l s t reng t h o f
s t ruct ures . Repres ent a t i ve ca s e s t udi es fro m t h e fo llo wi ng
s eg ment s o f i ndus t ry:
S h i p s t ruct ure - 4 ex a mples
Offs h o re s t ruct ure - 2 ex a mples
Ai rcra ft s t ruct ure - 2 ex a mples
were ch o s en t o i llus t ra t e current pra ct i ce.
4. 3. 1 S h i p S t ruct ure
Hi s t o ri ca lly t h e ma ri ne i ndus t ry h a s a do pt ed very co ns erva t i ve
des i g n a nd ma i nt ena nce pra ct i ces due t o t h e uncert a i nt i es i n s ea wa y
i nduced lo a ds a nd t h e la ck o f pro per a na lyt i ca l a nd co mput a t i o na l
t o o ls . Th e develo pment o f fra ct ure mech a ni cs a nd fi ni t e element
a na lys i s t ech ni ques a cco mpa ni ed by t h e a va i la bi li t y o f po werful
co mput ers h a s a llo wed t h e develo pment o f a na lys i s t ech ni ques t o
eva lua t e res i dua l s t reng t h a nd t h e ri s k o f s t ruct ura l fa i lures o f
co mplex s t ruct ures s uch a s s h i ps . Th erefo re, mo s t o f t h e releva nt
i nfo rma t i o n o n res i dua l s t reng t h a s s es s ment h a s been o bt a i ned i n
t h e la s t 10-15 yea rs .
Ex a mple 1
Alt h o ug h ma ny ex peri ment a l s t udi es h a ve been co nduct ed t o co rrela t e
t h e perfo rma nce o f s t ruct ura l co mpo nent s under va ri o us lo a di ng
us i ng t h eo ret i ca l a nd numeri ca l a na lys es , t h ere i s a s h o rt a g e o f
full s ca le t es t s t o s t udy t h e ult i ma t e s t reng t h o f ma ri ne
s t ruct ures . Als o i nci dence o f s h i ps "brea ki ng t h ei r ba cks " a re
ei t h er very lo w o r s uffi ci ent i nfo rma t i o n fo r a na lys i s i s no t
a va i la ble. Hence very few co mpa ri s o ns ex i s t s t o eva lua t e t h e
perfo rma nce o f t h e ex i s t i ng ult i ma t e s t reng t h pro g ra ms . Th e
fa i lure o f a VLC C , ENERGY C ONC ENTRATION i n 1980 pro vi ded a ra re
o ppo rt uni t y fo r va li da t i ng ex i s t i ng ult i ma t e s t reng t h t ech ni ques
s i nce t h e lo a di ng co ndi t i o n a t fa i lure wa s well defi ned.
Th e ENERGY C ONC ENTRATION co lla ps ed due t o i nco rrect ca rg o h a ndli ng
duri ng di s ch a rg e wh i ch ca us ed a n ex ces s i ve h o g g i ng mo ment . Th e
a ct ua l s t i ll wa t er bendi ng mo ment a t t h e t i me o f co lla ps e wa s
es t i ma t ed a t 17, 940 MNm. Th e ves s el wa s judg ed t o be i n g o o d
co ndi t i o n wi t h li mi t ed co rro s i o n. Va ls g a a rd & S t een i n [47] us ed
a s peci a l vers i o n o f t h e pro g ra m F ENRIS , na med F ENC OL (F i ni t e
Element No n-li nea r C o lla ps e/C o lli s i o n) fo r es t i ma t i ng t h e ult i ma t e
s t reng t h o f t h e s h i p. Va ls g a a rd & S t een det ermi ned a res i dua l
95
s t reng t h i ndex , B
c
, wh i ch t h ey refer t o a s t h e "cro s s -s ect i o n
ma rg i n", a nd defi ne a s :
M
u
B. =
M,
C R
M i s t h e ult i ma t e co lla ps e mo ment co mput ed by F ENC OL. M
C R
i s t h e
co lla ps e mo ment deri ved fro m t h e s t reng t h o f t h e pa nel i n t h e
co mpres s i o n fla ng e t h a t g o verns t h e o ns et o f fi na l co lla ps e o f a
s h i p's cro s s -s ect i o n. M,
C R
i s defi ned a s t h e pa nel s t reng t h (a
u
)
t i mes t h e s ect i o n mo dulus (S M). Th i s qua nt i fi ca t i o n o f res i dua l
s t reng t h i s di fferent fro m t h e us ua l defi ni t i o n o f res i dua l
s t reng t h i ndex , wh i ch i s t h e ra t i o o f t h e ult i ma t e co lla ps e mo ment
t o t h e a llo wa ble des i g n mo ment . In t h i s ca s e, t h e g o verni ng
fa i lure mech a ni s m i s i dent i fi ed a s t h e co lla ps e o f t h e bo t t o m
pa nels a nd s ubs equent res erve s t reng t h co mes fro m redi s t ri but i o n o f
t h e lo a d i n t h e rema i ni ng cro s s -s ect i o n.
S i x ca s es were run va ryi ng t h e s t reng t h fo rmula t i o n, yi eld s t res s ,
i mperfect i o n level a nd co rro s i o n. Th e res ult s o f t h es e runs a re
s h o wn i n t a ble 4-5. Th e pa nel s t reng t h fo rmula t i o n us ed i n F ENC OL
i s ba s ed o n cri t eri a fro m C N30. 1, reference [50]. Th e uncert a i nt y
o f t h e fo rmula t i o n i s a ppro x i ma t ed by a mea n va lue a nd a s t a nda rd
devi a t i o n a s fo llo ws :
Mea n va lue (m)
a .
yj
= 0. 795 + 0. 270
Standard deviation ( s):
yj
0.116 - 0.034
Ba s ed o n a n i n-h o us e s t udy a t D nV, o n t h e va ri a bi li t y o f yi eld
s t reng t h o f pla t e ma t eri a l deli vered fro m J a pa nes e s t eel mi lls , a
mea n va lue o f 400 MP a wi t h a C OV = 0. 066, wa s us ed fo r NV-32 s t eel
(o
y
=315 MP a ). S t i ffener i mperfect i o n va lue us ed i n F ENC OL i s t h e
s t a nda rd D nV o ffs h o re rule va lue wh i ch i s o ne-s i nus o i da l h a lf-wa ve
wi t h a ma x i mum devi a t i o n o f 0. 0015L a nd a mea n o f 0. 0008L wh ere L
i s t h e fra me s pa ci ng . Ba s ed o n ex peri ence, a uni fo rm 1 mm
co rro s i o n lo s s o f pla t e t h i cknes s wa s a s s umed i n ca s es 3 a nd 5.
96
5 5
O
$
E H
55
H
U
55
O
CJ
o
a
w
5 3
W
o
w
5 3
O
M
E -"
U
H
a
w
x
E H
O
5 5
W
OJ
E H
CO
in
I
W
E H
c
o =
o
ao 00 ro f\J r~- o r~- "> , -
< " HI
D
CO ^X
ro f\J ~ ro o
*~
r\J T
II
CO *
.. ..
CJ
ro T3 >
(U 4J o
E CO u
^^ ^^ ^s
5 z
o ~3- OJ ro o a-
r-- IO l> - o- r\j o
O O t> r\j in
o
co O a > ro rvj
qj 5 *~ *~ **~ *~ "
c 9-
(0 "5
O. ->
ro oo t> m r> - in
ID .'S
< 0 2E
IO ^' (\l !"- (M
^ in K 1 CM o CO
0. -' f\j t\i f\J r\j
> *
4-
u
ro
O. r-. o rO ro 03 o ~*
ro = E r- O ^0 r^ >! 00
> o ro o O o
o O 00 -o > d- to
*-* ' r\j
c
n
E
o
X
^ "O
o
c S -
o * c
_ P c o
co
o
o
g
o
g
O o 'Ci
c c c c
o
*"" *
e E
u
ro
o g
o; in
O so
_
ii
PI
c z
o CO
"- 03 00 03 CO CO in
s >
M- 111
O o O o o
o o O o o o
o o O o o o
L . J
o o O o o o
< u D
XI
o
E
C
o
Z J
u
0)
_ CO
f-n ^N ^ s ^^
"O (0 o o m L A o in
*
b
X
o o o
J- >! ro rO vT to
\^
> -
-C o
11-
c x cj
OJ O
w
CO
OS
L U
CO
OH
L U
CO
OH
L U
CO
OH
L U
CO
OH
CO
OH
*< c
w o
< " .
s
CJ
z
-o in ^3- ro r\i
(1)
3
U)
to OJ
> D
^ < -* c
a
c
~ w o
< u ro
T OJ >
b +J
M M- ty)
ro cu
u > 3
CO r to
ro
n
>
H f
(U
< u CO 01
] o
1 c/ l f=
ro
> CO
eu L .
r sz Cl
m r
u r OJ
H OJ * -
L .
01 m
-C n 4^
w
c
CO
U) * n
[_
CO ro
111 (1
(/ > .J c.
rr ro
ro
L U > CO
9 7
Th e ca s e clo s es t t o t h e a ct ua l va lue o f 17, 940 MNm wa s number 4
wh i ch wa s 2. 4% a bo ve t h e a ct ua l va lue.
U s i ng s i mi la r a s s umpt i o ns o n yi eld s t reng t h , co rro s i o n a nd lo ca l
i mperfect i o ns , t h e ult i ma t e s t reng t h o f ENERGY C ONC ENTRATION wa s
det ermi ned by Rut h erfo rd a nd C a ldwell [73], us i ng LLo yd's Reg i s t er
LRP AS S P ro g ra m No . 20202 a nd by Hug h es us i ng MAES TRO [74]. Th e
predi ct ed va lue o f t h e ult i ma t e mo ment i n bo t h t h e ca s es were very
clo s e t o t h e a ct ua l va lue a nd i s s h o wn i n fi g ure 4-10.
Ex a mple 2
To effect i vely co nt ro l t h e ri s k o f s t ruct ura l fa i lure, t h e a ut h o rs
i n [75] pres ent ed a s t ruct ura l ma na g ement s t ra t eg y ba s ed o n s i mple
ri s k a na lys i s pro cedures . Th i s a ppro a ch wa s develo ped fo r a cla s s
o f VLC C 's t h a t t h ei r co mpa ny o pera t es . Key s t eps i dent i fi ed fo r
t h e s ucces s ful i mplement a t i o n o f t h e s t ra t eg y were: develo pment o f
a n a na lyt i ca l da t a ba s e, develo pment o f fo cus s ed cri t eri o n fo r
s t ruct ura l i ns pect i o n/ma i nt ena nce a nd o bt a i ni ng feedba ck co ncerni ng
s h i p's s t ruct ura l perfo rma nce.
S et t i ng up t h e a na lyt i ca l da t a ba s e co ns i s t ed o f i dent i fyi ng
cri t i ca l lo ca t i o ns ba s ed o n: (i ) t h e s t res s di s t ri but i o n i n t h e
s h i p s t ruct ure under va ri o us lo a di ng co ndi t i o ns : qua s i -s t a t i c,
s pri ng i ng a nd s la mmi ng ; a nd (i i ) fa t i g ue dura bi li t y. Th e cri t i ca l
lo ca t i o ns were det ermi ned us i ng a na lyt i ca l met h o ds develo ped t o
s t udy vi bra t i o n o f h ull g i rders i n wa ves [76] a nd s la mmi ng effect s
[77] a lo ng wi t h fi ni t e element t ech ni ques . A t wo level fa t i g ue
dura bi li t y a s s es s ment wa s perfo rmed. Th e fi rs t level wa s ba s ed o n
cert a i n a s s umpt i o ns a nd co mpa ri s o n o f fa t i g ue ca lcula t i o ns wi t h
kno wn fra ct ures . Th e s eco nd level, wa s ba s ed o n fra ct ure
mech a ni cs . C ri t i ca l cra ck leng t h s a nd a ccept a bi li t y o f cra cks were
det ermi ned a cco rdi ng t o pro cedures o ut li ned i n P D 6493. By
mo deli ng cra ck li ke defect s us i ng s peci a l fra ct ure mech a ni cs ba s ed
element s i n a fi ni t e element mo del, cri t i ca l lo ca t i o ns s us cept i ble
t o ra pi d cra ck pro pa g a t i o n were i dent i fi ed. Th e t h eo ret i ca l met h o d
t o det ermi ne cri t i ca l cra ck leng t h t o i ni t i a t e bri t t le fra ct ure
i ncluded s i mpli s t i c a s s es s ment s o f cri t i ca l cra ck s i ze
det ermi na t i o n ba s ed o n pla ne s t ra i n s t res s i nt ens i t y fa ct o r K
IC
,
fi g ure 4-11. Ba s ed o n t h i s a na lys es t h e cri t i ca l lo ca t i o ns o n t h e
ves s el were i dent i fi ed a s ; lo ng i t udi na l co nnect i o ns , upper cro s s
t i e co nnect i o n t o s i de s h ell, lo ng i t udi na l cent erli ne g i rder
bra cket s , t ra ns vers e bulkh ea d s t ri ng er bra cket t o e a nd t h e upper
deck.
A fo cus ed i ns pect i o n cri t eri a ba s ed o n cri t i ca l lo ca t i o n, fa t i g ue
dura bi li t y a nd s t ruct ura l mo ni t o ri ng wa s develo ped. A s t ruct ura l
mo ni t o ri ng s ys t em, wh i ch i ncludes o nbo a rd mea s urement devi ces a nd
di s pla ys , wa s i ns t a lled t o pro vi de feedba ck co ncerni ng t h e s h i p's
98
MAESTRO, 1.74 X 10 kNm
z
J
O
J
CG
1.J
1.6
1.4
I.2
1.0
O.i.
0.6
0.4
0.2
-10
LR PASS, 1.786 X 10 kNm
2.0 3.0
CURVATURE X 10 (1/MM)
FIGURE 4.10 ULTIMATE STRENGTH OF 'ENERGY CONCENTRATION'
COMPARISON OF MAESTRO RESULTS WITH LLOYD'S
LRPASS PROGRAM
9 9
C\l
00
00
UJ
QC
\
00
900
600
300
0 200 400 600
CRACK LENGTH (mm)
FIGURE 4.11 STRESS/TEMPERATURE EFFECT ON CRITICAL CRACK SIZE
75)
100
s t ruct ura l perfo rma nce. Th e s ys t em perfo rms t h e fo llo wi ng
funct i o ns : (i ) t o pro vi de co nt i nuo us g ui da nce t o t h e s h i p's s t a ff
wi t h reg a rd t o li mi t i ng lo a di ng co ndi t i o n a nd s peed, a nd (i i )
reco rd pea k a nd cycli c envi ro nment a l lo a di ng da t a t o be us ed fo r
a na lyt i ca l predi ct i o n.
Ex a mple 3
Recent ly, fra ct ure mech a ni cs t ech ni ques t o a s s es s t h e s t ruct ura l
ri s k o f fa i lure were a ppli ed t o t h e eva lua t i o n o f a fi ne li ned s h i p
o f lo w pri s ma t i c co effi ci ent . Th i s s h i p, h encefo rt h referred t o a s
s h i p A, develo ped cra cks a t t h e 01 Level a nd s i de s h ell, a s s h o wn
i n fi g ure 4-12. A previ o us s h i p, referred t o a s s h i p B, wh i ch h a d
a di fferent s t ruct ure but s i mi la r co nfi g ura t i o n a t t h a t a rea wa s i n
s ervi ce fo r a lo ng er t i me a nd h a d no t develo ped a ny cra cks . A
s eri es o f eng i neeri ng s t udi es were ca rri ed o ut t o qua nt i fy t h e
s ervi ce t i me requi red t o i ni t i a t e cra cks s i mi la r t o t h o s e fo und i n
s h i p A a nd g ro w t h em t o t h ei r cri t i ca l cra ck s i ze.
To det ermi ne t i me requi red fo r cra ck i ni t i a t i o n, cumula t i ve fa t i g ue
da ma g e wa s co mput ed us i ng P a lmg ren-Mi ner's rule. F o r lo a di ng , a 30
yea r No rt h At la nt i c s t res s ex ceedence curve ba s ed o n li fet i me
bendi ng mo ment a t 50% o pera bi li t y wa s us ed. C ra cks were a s s umed t o
i ni t i a t e a t a weld t h erefo re S -N curves fo r welded s peci men a t
s t res s ra t i o , R = -1 were us ed. U s i ng t h es e i nput s , a fa t i g ue
co mpa ri s o n o f s h i p A a nd s h i p B wa s ca rri ed o ut fo r lo ca t i o ns a t
mi ds h i p a nd a t t h e a ct ua l lo ca t i o n o f t h e cra ck fo und i n s h i p A.
Res ult s a re s h o wn i n fi g ure 4-13.
A s i mi la r s t udy wa s co nduct ed us i ng t h e s a me s t res s s pect rum a nd
o pera t i ng co ndi t i o ns t o det ermi ne t h e t i me requi red fo r a n ex i s t i ng
s urfa ce cra ck, 6. 4 mm lo ng a nd 3. 2 mm deep, t o rea ch a cri t i ca l
leng t h . U s i ng h i g h s t reng t h s t eel pla t e cra ck g ro wt h ra t e da t a i n
P a ri s ' equa t i o n, a co mpa ri s o n o f cra ck g ro wt h ra t e wa s ma de, fo r
bo t h s h i ps . Res ult s a re s h o wn i n fi g ure 4-14.
Th i s s t udy co ncluded t h a t t h e us e o f fra ct ure mech a ni cs t ech ni ques
a nd a s s o ci a t ed eng i neeri ng a s s umpt i o ns co uld be a ppli ed t o
rea li s t i c o pera t i o na l s cena ri o s a nd fo rm t h e ba s i s fo r qui ck
res po ns e ri s k a s s es s ment deci s i o ns .
Ex a mple 4
Th i s ex a mple i llus t ra t es t h e a ppli ca t i o n o f fra ct ure mech a ni cs
t ech ni ques co upled wi t h fi ni t e element a na lys i s t o a na lyze duct i le
fra ct ure a nd t ea ri ng o f a co rrug a t ed t ra ns vers e bulkh ea d a t t h e
junct i o n wi t h t h e ma i n deck. Reference [78], dra wn fro m s evera l
ca s e s t udi es , demo ns t ra t es t h e a ppli ca t i o n o f- t ech ni ques li ke
fra ct ure mech a ni cs , i mpa ct a na lys i s , s lo s h i ng lo a d predi ct i o n,
fi ni t e element met h o ds a nd t i me-va ri a nt reli a bi li t y a na lys i s i n t h e
a s s es s ment o f vulnera bi li t y o f a n i n-s ervi ce bulk ca rri er.
101
FIGURE 4.12 LOCATION OF THE CRACK ON SHIP A
o
o
UJ 1
\, \ \^ \ y^ SHIP B, AT CRACK LOCATION
N. \v \. \v_SHIP B, AMIDSHIPS
SHIP A. AMIDSHIPS -> ^\ ^^+ N: ^^"-
SHIP A, AT CRACK LOCATION ^^"^-^^^^^^^
I I I II I M HUM I INI I I i i inn
0.1 10 100
YEARS TO CRACK INITIATION
1000 10000
FIGURE 4.13 FATIGUE COMPARISON OF SHIPS A & B
a:
o
o
o
z
o
a
YEARS TO CRACK INSTABILITY
FIGURE 4.14 CRACK GROWTH COMPARISON OF SHIPS A & B
102
As ment i o ned ea rli er i n ch a pt er 2, t h e pri ma ry ca us es o f fa i lure o f
bulk-ca rri ers a re co rro s i o n a nd fa t i g ue cra cki ng i n s i de a nd
t ra ns vers e bulkh ea d s t ruct ures . But recent ly i t h a s been fo und
t h a t bes i des t h es e fa ct o rs , defi ci ent o pera t i o na l a nd ma i nt ena nce
pro cedures h a ve led t o t h e lo s s o f bulk ca rri ers . F o r bulk
ca rri ers o f s i ng le s ki n co ns t ruct i o n, wi t h no effect i ve ca rg o
s ecuri ng devi ces t h a t a re i nco rrect ly lo a ded wi t h h i g h dens i t y
s o li d ca rg o , t h ere i s t h e pro ba bi li t y o f s i de s h ell punct ure due t o
i mpa ct o f t h e s o li d ca rg o duri ng h ea vy wea t h er. Res ult i ng i ng res s
o f wa t er i n t h e da ma g ed ca rg o h o ld co upled wi t h s evere pi t ch
mo t i o ns , ma y lea d t o t h e fra ct ure o f t ra ns vers e bulkh ea d a t i t s
junct i o n wi t h t h e upper deck a nd s i de wi ng t a nk due t o ex ces s i ve
s lo s h i ng a nd h ydro s t a t i c pres s ure. U lt i ma t ely t h e lo s s o f
effect i venes s o f t h e t ra ns vers e bulkh ea d ma y lea d t o h ull g i rder
co lla ps e.
Reference [78] pres ent ed pro cedures fo r co mput i ng t h e energ y
a bs o rpt i o n ca pa bi li t y o f s i de s h ell pla t i ng befo re rupt ure a nd t h e
fra ct ure s t reng t h o f co rrug a t ed bulkh ea d under s lo s h i ng a nd
h ydro s t a t i c lo a ds . Th e energ y a bs o rpt i o n ca pa bi li t y o f s i de s h ell
pla t i ng i s es t i ma t ed ba s ed o n fra ct ure (rupt ure) cri t eri a . Th e
rupt ure cri t eri a i s defi ned by t h e equi va lent rupt ure s t ra i n a nd i s
s t ro ng ly i nfluenced by a ny defect s embedded i n a da ma g ed s t ruct ure.
F i g ure 4-15 s h o ws t h e rela t i o ns h i p bet ween t h e rupt ure s t ra i n nea r
a n i nt erna l defect a nd t h e s i ze o f t h e defect . Th e reco mmended
equi va lent rupt ure s t ra i n fo r welded s t ruct ures i s a ppro x i ma t ely
0. 015 wh i ch repres ent 12. 5 t i mes t h e yi eld s t ra i n [79]. Th e s t res s
va lue pri o r t o t h e rupt ure o f t h e s i de s h ell pla t e ca n be t a ken t o
be t h e yi eld s t res s .
Th e fo llo wi ng t a ble repro duced fro m [78], s h o ws t h e va ri a t i o n o f
t h e energ y a bs o rpt i o n ca pa ci t y wi t h t h i cknes s o f a 5. 81m x 2. 74m
s i de s h ell pla t i ng wh i ch h a s a yi eld s t reng t h o f 230 MN/m
2
. E^ i s
t h e s t a t i c energ y a bs o rpt i o n ca pa bi li t y wh i le E
di mp
i s t h e mo di fi ed
E- t o reflect t h e dyna mi c na t ure o f i mpa ct .
t (mm) E
P
( MJ) E
dimp
( MJ)
10.0 0.23 0.24
12.0 0.28 0.29
14.0 0.33 0.34
16.0 0.38 0.39
18.0 0.43 0.44
20.0 0.48 0.50
22.0 0.53 0.55
103
TEMPERATURE 0" C
E
o
I
CJ
UJ
_
UJ
Q
<
z
en
L U
o
o
L U
X
20
18
14
12
10
8
6
4
2
ELONGATION AT RUPTURE G
rupt
( %
FIGURE 4.15 CRITICAL RUPTURE STRAIN AS A FUNCTION OF
INTERNAL DEFECT SIZE ( IN WELD ) ^
79
'
)
104
To eva lua t e t h e fra ct ure s t reng t h o f a co rrug a t ed bulkh ea d under
s lo s h i ng a nd h ydro s t a t i c lo a ds , a t h ree-di mens i o na l fi ni t e element
mo del us i ng s h ell element s t o repres ent t h e co rrug a t ed bulkh ea d,
cro s s deck s t ri p, h a t ch co a mi ng , h a t ch end bea m a nd s t i ffeners
a t t a ch ed t o t h e bulkh ea d a t t h e junct i o n wi t h t h e deck s t ri p wa s
us ed. Th e a na lys i s predi ct ed t h a t mo dera t e s lo s h i ng lo a ds co mbi ned
wi t h s t res s co ncent ra t i o ns due t o s t ruct ura l di s co nt i nui t i es a nd
res i dua l weld s t res s , wi ll ca us e la rg e a rea s o f t h e bulkh ea d a t t h e
junct i o n wi t h t h e upper wi ng t a nk a nd cro s s deck s t ri p t o underg o
t ens i o na l s t res s i ng beyo nd t h e ma t eri a l yi eld s t ra i n. Th erefo re
t h e fra ct ure o f mi ld s t eel bulkh ea d pla t i ng i s us ua lly no t a
bri t t le fa i lure; ex t ens i ve yi eldi ng o ccurs i n t h e vi ci ni t y o f
cri t i ca l fa t i g ue cra cks wh i ch i nduce t h e fra ct ure. U s i ng t h e va lue
o f s t ra i n o bt a i ned fro m t h e fi ni t e-element a na lys i s , t h e a ut h o r
us ed t h e C TOD met h o d a nd es t i ma t ed a cri t i ca l cra ck leng t h o f 40 mm
fo r mi ld s t eel bulkh ea d pla t i ng h a vi ng a cri t i ca l C TOD , < S
C R
, o f
0. 25 mm.
4. 3. 2 Offs h o re S t ruct ures
Th e la ck o f s ervi ce li fe ex peri ence o f o ffs h o re s t ruct ures co upled
wi t h i nci dences o f ca t a s t ro ph i c fa i lures , li ke t h e lo s s o f
"Alex a nder L. Ki ella nd", h a s led t o t h e develo pment o f t h e
No rweg i a n P et ro leum D i rect o ra t e (NP D ), s a fet y cri t eri o n [80] fo r
o ffs h o re s t ruct ures . Th e mo s t s i g ni fi ca nt di fference bet ween t h e
NP D reg ula t i o ns [80] a nd s h i p reg ula t i o ns s uch a s ABS , D NV, et c. ,
i s t h e fa i lure cri t eri o n a ppli ed i n t h e fo rmer reg ula t i o n. Th i s
cri t eri o n ca lls fo r a n ex pli ci t . ch eck, duri ng des i g n, o f t h e
res i dua l s t reng t h due t o va ri o us a s s umed da ma g ed co ndi t i o ns under
no rma l o pera t i ng co ndi t i o ns . Two i ns t a nces o f res i dua l s t reng t h
a s s es s ment , o ne o n a mo bi le o ffs h o re pla t fo rm [81] a nd t h e o t h er o n
a fi x ed ei g h t leg g ed pla t fo rm [82], a re pres ent ed belo w.
Ex a mple 1
In [81], a no n-li nea r fi ni t e element pro g ra m ba s ed o n "Idea li zed
S t ruct ura l U ni t Met h o d" (IS U M) i s us ed t o do a n ult i ma t e a nd
res i dua l s t reng t h a s s es s ment o f a dri lli ng ri g o f t h e Aker H3. 2
t ype s h o wn i n fi g ure 4-16. In t h e IS U M pro cedure, o ne fi ni t e
element per s t ruct ura l element i s us ed. No n-li nea r g eo met ri c a nd
ma t eri a l beh a vi o r i s i ncluded. D a ma g e o f t ubula r members ,
i ncludi ng perma nent la t era l defo rma t i o n, dent s a nd t h e effect o f
fi re, a re t a ken i nt o a cco unt i n t h e fo rmula t i o n. A number o f
ex plo s i o n da ma g e ca s es a re s i mula t ed a t cri t i ca l lo ca t i o ns . Two
ex plo s i o n da ma g e ca s es i n t h e mud ro o m a re co ns i dered. Th e fi rs t
o ne i s a s s umed t o blo w o ut o ne o f t h e ma i n wa lls i n t h e mud ro o m s o
t h a t t h e o bli g ue bra ci ng s a re no lo ng er effect i ve. Th e s eco nd o ne
a s s umes t h a t t h e ex plo s i o n h a s led t o perma nent defo rma t i o n i n t h e
pa nels s urro undi ng t h e mud ro o m, i n wh i ch ca s e, t h e da ma g e i s
mo deled by reduci ng t h e s h ea r a rea . Ano t h er ex plo s i o n ca s e o ccurs
a t t h e ma i n leg a nd i s s i mula t ed by i t s remo va l. No rma li zed lo a d
105
vers us di s pla cement curves o f t h e res po ns e o f t h e pla t fo rm t o
ex plo s i o n da ma g e i n t h e mud ro o m i s s h o wn i n fi g ure 4-17. Lo a ds
h a ve been no rma li zed t o a des i g n wa ve lo a d due t o a wa ve o f leng t h
88m a nd h ei g h t o f 9. 5m, wi t h a ret urn peri o d o f 100 yea rs . Th e
di s pla cement s a re no rma li zed a g a i ns t t h e reference lo a d pa t t ern
wh i ch i s a co mbi na t i o n o f g ra vi t y lo a ds a nd buo ya ncy fo rces . Th e
res ult s h a ve been s umma ri zed i n t a ble 4-6.
Ex a mple 2
Th e ult i ma t e a nd res i dua l s t reng t h o f a t ypi ca l ei g h t -leg g ed, No rt h
S ea pla t fo rm i s co mput ed i n [82], us i ng U S F OS , a n i ncrement a l, no n-
li nea r fi ni t e element pro g ra m. U S F OS i s ba s ed o n bea m t h eo ry a nd
a cco unt s fo r la rg e di s pla cement s a nd ela s t o -pla s t i c ma t eri a l
beh a vi o r. Th e pla t fo rm s h o wn i n fi g ure 4-18 i s s i t ua t ed a t a wa t er
dept h o f 70m. It s members h a ve been s i zed a cco rdi ng t o t h e AP I-
RP 2A des i g n co de. Lo a di ng co ns i s t s o f t h e g ra vi t y lo a d a nd
envi ro nment a l lo a d. Th ree lo a d ca s es were co ns i dered ba s ed o n
des i g n wa ves wi t h a ret urn peri o d o f 100 yea rs . Th ey a re:
- LC I: Wa ves 45 rela t i ve t o t h e lo ng i t udi na l a x i s (di a g o na l
wa ves ), Wa ve h ei g h t = 2 6m.
- LC 2: Wa ves a lo ng t h e lo ng i t udi na l a x i s , Wa ve h ei g h t = 27m.
- LC 3: Wa ves perpendi cula r t o t h e lo ng i t udi na l a x i s , Wa ve
h ei g h t = 20. 6m.
F o ur da ma g e s cena ri o s were s t udi ed. Two t o s i mula t e da ma g e due t o
s h i p co lli s i o n nea r t h e wa t erli ne a nd t h e o t h er t wo t o s i mula t e
da ma g e a t t h e lo wer ends o f t h e ja cket due t o dro pped o bject s .
Th es e fo ur ca s es a re:
- D C 1: Th e co mplet e cro s s -bra ci ng a t s ea level i s co ns i dered
i neffect i ve.
- D C 2: A co rner leg i s a s s umed da ma g ed a t t h e s ea level. Th e
da ma g e co ns i s t i ng o f a dent o f dept h 0. 35m (22% o f t h e
di a met er)- a nd a perma nent la t era l defo rma t i o n o f t h e
neut ra l a x i s o f 0. 68m (3. 2% o f t h e leg leng t h ).
- D C 3: Bo t t o m, End ro w cro s s -bra ci ng i neffect i ve.
- D C 4: Bo t t o m, F ro nt ro w cro s s -bra ci ng i neffect i ve.
F i g ure 4-19 i s t h e res po ns e curve fo r t h e g ua rt eri ng s ea ca s e, LC I.
Th e lo a d i s no n-di mens i o na li zed a g a i ns t t h e des i g n wa ve lo a d wi t h
a ret urn. peri o d o f 100 yea rs . Th e di s pla cement co rres po nds t o a
t ra ns vers e di s pla cement a t t h e t o p co rner no de. Res ult s o f a ll
ca lcula t i o ns a re s umma ri zed i n t a ble 4-7.
Th e s t udy co ncluded t h e pres ence o f t remendo us res erve s t reng t h a nd
redunda ncy i n t h e ja cket a nd t h e i ns i g ni fi ca nt i nfluence o f
rela t i vely s evere co lli s i o n da ma g es a s s umed a t s ea level.
106
Pontoon
Column
nrom*
0 rtC8
FIGURE 4.16 MOBILE OFFSHORE PLATFORM
4.0
3.0
UNDAMAGED
2.0
O
<
o
1.0
0.0
25
REMOVED ELEMENTS
50 75 100 [ X 10 ]
DISPLACEMENT NORMALIZED
FIGURE 4.17 LOAD-NORMALIZED DISPLACEMENT FOR EXPLOSION
(81)
IN MUD ROOM
1 0 7
FIGURE 4.18 EIGHT LEGGED NORTH SEA PLATFORM
<
o
0.0
DISPLACEMENT (M)
FIGURE 4.19 LOAD VERSUS TRANSVERSE DISPLACEMENT OF
TOP CORNER NODE OF JACKET STRUCTURE.
(82)
( LC1 -DIAGONAL WAVES )
108
TABLE 4-6 RESULTS OF RESIDUAL STRENGTH ASSESSMENT OF A DRILLING RIG
( 81)
.
LOAD FACTORS ARE GIV EN IN MULTIPLES OF THE 100 YEAR WAV E LOAD.
C a s e C o ndi t i o n F i rs t U lt i ma t e Res erve Res i dua l
yi eld
s t reng t h
fa ct o r
s t reng t h
fa ct o r
1 U nda ma g ed 1. 28 3. 76 3. 76
2 Ex plo s i o n mud ro o m
Element s remo ved
0. 18 1. 86 - 0. 49
3 Ex plo s i o n mud ro o m
Reduced ca pa ci t y
1. 1 3. 45 - 0. 92
4 Ex plo s i o n ma i n leg 1. 0 | 3. 08 - 0. 82
TABLE 4-7 RESULTS OF RESIDUAL STRENGTH ASSESSMENT OF A TYPICAL 8-
LEGGED NORTH SEA PLATFORM
( 82)
. LOAD FACTORS ARE GIV EN IN MULTIPLES OF
THE 100 YEAR WAV E LOAD.
C a s e Lo a d C a s e D a ma g e C a s e F i rs t
yi eld
U lt i ma t e Res erve
s t reng t h
fa ct o r
Res i dua l
s t reng t h
fa ct o r
1 LC I Int a ct 2. 60 4. 39 4. 39 -
2 LC I D C 1 2. 64 4. 10 - 0. 93
3 LC I D C 2 2. 52 4. 25 - 0. 97
4 LC I D C 3 2. 00 3. 15 - 0. 72
5 LC 2 Int a ct 2. 49 3. 71 3. 71 -
6 LC 2 D C 4 1. 82 3. 38 - 0. 91
7 LC 3 Int a ct 2. 99 4. 32 4. 32 -
8 LC 3 D C 3 2. 19 3. 39 - 0. 78
LCI = Diagonal Waves; LC2 = Longitudinal Waves; LC3 = Transverse Waves
DC1 = X-Bracing at sea level ineffective;
DC2 = Corner leg dented and laterally deformed;
DC3 = Bottom, end rowX-Bracing ineffective;
DC4 = Bottom, front rowX-Bracing ineffective.
109
4. 3. 3 Ai rcra ft S t ruct ures
Th e pri nci ples a nd pro cedures o f res i dua l s t reng t h a nd li fe
a s s es s ment o f a g i ng a ero s pa ce s t ruct ure a re embedded i n t h e
Ai rcra ft S t ruct ura l Int eg ri t y P ro g ra m (AS IP ). Th e AS IP pro g ra m wa s
i ni t i a t ed i n 1958 by t h e U S AF i n res po ns e t o fa t i g ue fa i lures o n B-
47 a i rcra ft . A det a i led h i s t o ry wh i ch led t o t h e develo pment o f
AS IP a nd t h e mo di fi ca t i o ns wh i ch i t underwent o ver t i me, ca n be
fo und i n [83-85].
In a g enera l s ens e, a g i ng a i rcra ft a re ch a ra ct eri zed by
det eri o ra t i ng s t reng t h a nd rela t ed pro blems s uch a s cra ck
develo pment , lo s s o f redunda nt fea t ures a nd t h e res ult i ng i ncrea s ed
ma i nt ena nce co s t . F ro m a s t ruct ura l da ma g e s t a ndpo i nt , [86]
defi nes a n a g i ng a i rcra ft a s a n a i rcra ft t h a t co nt a i ns mult i ple
s i t e da ma g e (MS D ) ch a ra ct eri zed by t h e li nkup o f fa t i g ue cra cks .
S i nce t h e i ncept i o n o f AS IP , U S AF , co mmerci a l jet ma nufa ct urers a nd
o wners a nd NAS A h a ve been wo rki ng i n t h i s i nt erdi s ci pli na ry pro g ra m
o n t h e ma i nt ena nce o f a g i ng a i rcra ft s . Is s ues rela t ed t o
ma i nt ena nce a re, s a fet y a nd da ma g e t o lera nce ca pa bi li t y.
Ini t i a lly, da ma g e t o lera nce a na lys i s met h o d us ed fra ct ure mech a ni cs
i n a det ermi ni s t i c ma nner. Th a t i s fla w g ro wt h wa s predi ct ed us i ng
a fi x ed po t ent i a l fla w s i ze, a fi x ed da /dN vers us AK rela t i o ns h i p,
a nd a s t res s s pect rum deri ved fro m a predi ct ed a vera g e us a g e. Th i s
i s a co ns erva t i ve a ppro a ch . In a n a g i ng fleet , s t ruct ura l da ma g e
i s s t o ch a s t i c i n na t ure. Hence pres ent ly s t o ch a s t i c pro cedures a re
bei ng us ed t o a na lyze i ni t i a t i o n a nd g ro wt h o f cra cks a nd
pro ba bi li t y o f fra ct ure. A bri ef des cri pt i o n o f t h e- res ea rch
effo rt s under bo t h t h es e a ppro a ch es - det ermi ni s t i c a nd
pro ba bi li s t i c, a nd t h ei r pra ct i ca l a ppli ca t i o ns i n t h e mi li t a ry a nd
co mmerci a l wo rld a re pres ent ed belo w, wi t h a vi ew t o i dent i fy
co ncept s a nd t ech no lo g i es wh i ch ma y be a ppli ca ble t o t h e ma ri ne
i ndus t ry.
Ex a mple 1: D et ermi ni s t i c Appro a ch
C ra ck Gro wt h P redi ct i o n; C ra ck g ro wt h predi ct i o ns h a ve been ba s ed
o n cra ck g ro wt h pro g ra ms . Th es e pro g ra ms a re a ut o ma t ed pro cedures
wh i ch predi ct t h e pro pa g a t i o n o f va ri o us fla w s h a pes i n s t ruct ura l
members under va ri a ble a mpli t ude s pect ra l lo a di ng envi ro nment s .
F i g ure 4-2 0 s h o ws s ch ema t i ca lly t h e i nput s a nd o ut put o f a cra ck
g ro wt h pro g ra m ca lled "AGP F " us ed by t h e D o ug la s Ai rcra ft C o . fo r
t h ei r D C -10 Ins pect i o n P ro g ra m [87]. A s i mi la r pro g ra m ca lled
"EF F GRO" i s us ed fo r t h e F a t i g ue Ma na g ement o f t h e A-7P Ai r F o rce
a i rcra ft [88], Th ey numeri ca lly i nt eg ra t e equa t i o ns o f cra ck
g ro wt h ra t e t o co me up wi t h a cra ck g ro wt h pro fi le. EF F GRO
ut i li zes F o rma n's equa t i o n (3. 15).
F ro m s t udyi ng t h e i nfluence o f va ri a ble lo a di ng o n cra ck g ro wt h
ra t e, i t h a s been fo und t h a t pea k o verlo a d blunt s t h e cra ck t i p a nd
ret a rds cra ck g ro wt h ; EF F GRO us es t h e "Wh eeler Ret a rda t i o n" mo del
t o ma t h ema t i ca lly repres ent t h i s res po ns e. EF F GRO ut i li zes t h e
110
fra ct ure t o ug h nes s va lue (K
IC
) fo r pla ne s t ra i n a nd cri t i ca l s t res s
i nt ens i t y va lue (K
c
) fo r pla ne s t res s fa i lures g i ven i n [89, 90].
C ra ck t i p s t res s i nt ens i t y mo dels fo r cri t i ca l lo ca t i o ns us ed i n
EF F GRO were develo ped a nd veri fi ed by co upo n a nd/o r co mpo nent
t es t i ng duri ng t h e o ri g i na l A-7P AS IP [90] a nd duri ng t h e t es t i ng
ph a s e o f t h e pro g ra m. F o r t h o s e wh i ch were no t , clo s ed fo rm
s o lut i o ns fo r s t res s i nt ens i t y fa ct o rs were o bt a i ned us i ng t h e
co mput er pro g ra m BTAB [91].
Ins pect i o n Int erva l: Th e o ut co me o f da ma g e t o lera nce a na lys i s i s
t o reco mmend i ns pect i o n i nt erva ls a nd cri t i ca l lo ca t i o ns , t o ens ure
t h e s a fet y o f t h e s t ruct ure. Th i s i s do ne by det ermi ni ng t h e
res i dua l s t reng t h o f t h e cra cked s t ruct ure. A s ch ema t i c o f t h e
pro cedure t a ken fro m [87] fo r t h e D C -10 pro g ra m i s s h o wn i n fi g ure
4-21. C o mbi ni ng t h e cra ck g ro wt h predi ct i o n res ult s a nd res i dua l
s t reng t h res ult s , s h o wn s ch ema t i ca lly i n fi g ure 4-22, a n i ns pect i o n
i nt erva l "L
n
" i n t erms o f fli g h t h o urs i s det ermi ned fo r a cert a i n
fa ct o r o f s a fet y. Reco mmenda t i o ns ba s ed o n s uch a na lys i s a re kno wn
a s t h e F o rce S t ruct ura l Ma i nt ena nce P la n i n t h e Ai r F o rce a nd MRB
(Ma i nt ena nce Revi ew Bo a rd) Ins pect i o n P ro g ra m, i n t h e co mmerci a l
wo rld. S a mples o f Ai r F o rce S t ruct ura l Ma i nt ena nce P la n fo r t h e
A7P t a ken fro m [88] a nd t h e D C -10 pro g ra m t a ken fro m [87] a re
pres ent ed i n t a bles 4-8 a nd 4-9.
Ex a mple 2: P ro ba bi li s t i c Appro a ch
Burns , i n [86] di s cus s es a ri s k a na lys i s co mput er co de ca lled P ROF
(P Ro ba bi li t y Of F ra ct ure) wh o s e o bject i ve i s t o s t o ch a s t i ca lly
a s s es s s t ruct ura l i nt eg ri t y. P ROF wa s develo ped by t h e U ni vers i t y
o f D a yt o n Res ea rch Ins t i t ut e under co nt ra ct t o t h e U . S . Ai r F o rce.
P ROF i s a VAX-ba s ed co mput er pro g ra m wri t t en i n F ORTRAN wh i ch i s
ca pa ble o f det ermi ni ng t h e pro ba bi li t y o f fra ct ure o f a n a i rcra ft
o r fleet o f a i rcra ft s ubject ed t o a g i ven us a g e. Ins pect i o n
ca pa bi li t i es a nd i nt erva ls ca n be ma ni pula t ed t o yi eld t h e o pt i mum
i ns pect i o n s ch edule fo r t h e s peci fi c a i rcra ft o r fleet . P ROF
a cco unt s fo r cra cki ng i n t h e met a lli c a i rcra ft s t ruct ure, a nd
a cco unt s fo r a di s t ri but i o n o f cra ck s i zes t h a t ma y ex i s t i n a
s t ruct ura l det a i l. P ROF do es no t a cco unt fo r a ny co rro s i o n o r
mult i ple s i t e da ma g e (i nt era ct i o n) effect s . A co nci s e des cri pt i o n
o f t h e pri nci ples beh i nd t h e wo rki ng o f P ROF a nd t h e o ut put t h a t
ca n be o bt a i ned a re pro vi ded belo w. D et a i ls ca n be o bt a i ned fro m
[86].
Ill
co
O
1
1
/
[
*
viz
zo
Z-> 1?
t/) =
b
V)
b"
1
o
Q
LJ
01
Q_
X
I
o
z
LU
DC
I
00
_l
<
X
Q
00
LJJ
DC
CM
X
o
z
o
^N~1
1
<
z
< i \_ j
ft
1 o I \
I
o
c
00
CO
CO
/ **
o
a: O
/
v> (j rt
-Jfi
*
/"x 1
" -
1
1
a:
b |
to
Q:
O
00
00
> -
_l
<
<
> -
m
o
i
<
_l
<
>
LJ
O
Z
<
D C
LxJ
_i
o
V I
-J y
a:
UJUJ
i- a.
<o
2(L
a. ^
< /
^
w>
o
-o s * b
t
Ci z
ZO
UJ i-
S3
3
to
vice II
VI
7
D
QL
U _^
a.
,g
3
VI ^^ l~
VI <X
VI
a:
vi
O
D
~
_.
o:
Q
O
3 Q
I to
&
CK
O
TP -D
o
I
o
X
LJ
DC
Q_
X
O
DC
O
<
DC
O
O
CN
^'
LJ
DC
X
O
+
g
o
Q
D.
X
s
o
cc
o
o
u
^
2
o
<
<
X
CM
CN
LJ
DC
Z
O
112
TABLE 4-8 S AMP LE A-7P , C RITIC AL LOC ATIONS AND OP TIMU M INS P EC TION
INTERVALS '
871
CODE DESCRIPTION
INSPECTION INTERVAL (FLIGHT HOURS)
SEVERE ENVIRONMENT (SAFETY LIMIT/2)
C RITIC AL
S IZE
INS P EC TION P ROC ED U RES
IN-HOLE EDDY CURRENT S U RF AC E ED D Y C U RRENT
r **
(IN. )
PT T P T T
A1 LWS, WS24. 6 a 1 st Intermediate Spar 6 6 50 5690 3830 4670 4. 00
A2 LUS, US24. 6 3 2nd Intermediate Spar 3050 2180 1300 1700 1. 55
A3 LUS, WS24. 6 3 3rd Intermediate Spar 36 50 2385 1050 1725 1. 47
A4 LWS, WS24. 6 a 4th Intermediate Spar 3350 2150 1050 1725 1. 19
A5 LUS, US24. 6 a 5th Intermediate Spar 3300 2100 975 1515 1. 19
A6 LUS, Inb' d Pylon a 2nd Intermediate Spar 5200 3650 1960 2840 2. 82
A7 LWS, US-53. 7 a Rear Spar (Dry hole) 1 1 7 5 960 -- --
0. 35
A8 LUS, WS-50. 0 a Rear Spar (Wet hole) 2350 1090 -- --
0. 47
A9 LUS, US-50. 0 a Rear Spar (Dry hole) 1 21 0 1025 -- --
0. 34
A1 0 Lower Spar Cap, WS-50. 0 a Rear Spar 1 36 0* 1020* -- --
0. 47
A1 1 LWS, US-32. 2 a 4th Intermediate Spar 31 7 5 2060 925 1470 1. 29
A1 2 LUS, Cntr Pylon a 1 st Intermediate Spar 6 7 25 4560 2660 3800 2. 74
A1 3 LUS, US-32. 2 a 2nd Intermediate Spar 29 50 2255 1185 1570 1. 55
A1 4 LUS, US-6 8. 0 a 4th Intermediate Spar
2850
I
2020 915 1400 1. 29
P T
T
Number i n pa rent h es i s i s P a rt Th ro ug h C ri t i ca l fla w s i ze
All o t h ers a re Th ro ug h C ri t i ca l fla w s i ze
S urfa ce eddy current i ns pect i o ns do no t a pply
P a rt Th ro ug h cra ck
Th ro ug h t h i cknes s cra ck
TABLE 4-9 S AMP LE D C -10, MAINTENANC E REVIEW BOARD (MRB) REP ORT
(86)
REFERENCE
NO.
GENERIC STRUCTURAL
ELEMENT
ZONES SIGNIFICANT
LOCATIONS
EXTERNAL
HOURS
INTERNAL
DETAILS
HOURS SAMPLES
541 . 1 3. 021
541 . 1 4. 025
541 . 1 5. 027
Ueb, Stiffeners, and
cutout doublers on upper
face of spar
. Reverser door actuator
support attach points and
built-up hat section
(applicable to - 1 0/1 0CF,
- 30/30f/30cf Only)
Bulkhead vertical caps and
end fittings at lower ends
437 ,487
41 5,41 6 ,
46 5,46 6
438,488
STA YN 26 6 to 342.
Panels 437 EL,
437 FL, 487 EL, and
487 FL
Fwd face of STA YN
222 bulkhead.
Raise reverser
door
STA YN 286 aft
face doors 438 AL,
438 AR, 488AL, and
488 AR
1 2,000
4,000
1 /5 or 20%
(20,000)
1 /5 or 207 .
(20,000)
.
113
P ROF i s a ppli ca ble t o a po pula t i o n o f s t ruct ura l element s .
P o pula t i o n i s defi ned a s det a i ls wh i ch ex peri ences es s ent i a lly
equi va lent s t res s h i s t o ri es . Two ma i n t a s ks perfo rmed by P ROF
a re:
( i ) Mo deli ng t h e cra ck s i ze di s t ri but i o n; wh i ch i ncludes
(a ) det ermi na t i o n o f g ro wt h o f cra ck s i ze di s t ri but i o n,
(b) ma i nt ena nce effect o n cra ck s i ze di s t ri but i o n.
(i i ) D et ermi ni ng t h e P ro ba bi li t y o f F ra ct ure.
A flo w ch a rt o f t h es e t a s ks a nd s ubt a s ks i s pres ent ed i n fi g ure
4-23.
Gro wt h o f cra ck-S i ze D i s t ri but i o n Gi ven a n i ni t i a l di s t ri but i o n o f
cra ck s i zes a t a reference t i me, T
R
, t h e pro g ra m es t i ma t es t h e
di s t ri but i o n o f cra ck s i zes a t T
R
+ AT fli g h t h o urs by pro ject i ng
t h e percent i les o f t h e i ni t i a l cra ck s i ze di s t ri but i o n us i ng t h e
det ermi ni s t i c cra ck g ro wt h vers us fli g h t h o urs rela t i o n o f t h e
da ma g e t o lera nce a na lys i s . Th i s ca lcula t i o n i s perfo rmed i n P ROF
by referri ng t o da t a ba s es .
Ma i nt ena nce Effect o n C ra ck-S i ze D i s t ri but i o n Ma i nt ena nce a ct i o n
wi ll ch a ng e t h e cra ck s i ze di s t ri but i o n. Th i s ch a ng e i s a funct i o n
o f t h e i ns pect i o n ca pa bi li t y a nd t h e qua li t y o f repa i r a s s h o wn i n
fi g ure 4-23. Ins pect i o n ca pa bi li t y i s mo deled i n t erms o f t h e
pro ba bi li t y o f det ect i o n a s a funct i o n o f cra ck s i ze P OD (a ).
Repa i r qua li t y i s ex pres s ed i n t erms o f t h e equi va lent repa i r cra ck
s i ze di s t ri but i o n, f
r
(a ). U s i ng t h es e i nput s t h e pro g ra m det ermi nes
a cra ck s i ze di s t ri but i o n a s a res ult o f t h e ma i nt ena nce a ct i o n a s
s h o wn i n fi g ure 4-23. Th e po s t -ma i nt ena nce cra ck s i ze di s t ri but i o n
i s t h en pro ject ed fo rwa rd fo r t h e nex t i nt erva l o f uni ns pect ed
us a g e. Th e pro ces s i s co nt i nued fo r a s ma ny i ns pect i o n i nt erva ls
a s i s des i red.
D et ermi na t i o n o f P ro ba bi li t y o f F ra ct ure S a fet y i s qua nt i fi ed i n
t erms o f t h e pro ba bi li t y o f fra ct ure (P OF ). P OF i s ca lcula t ed a s
t h e pro ba bi li t y t h a t t h e ma x i mum s t res s enco unt ered i n a fli g h t
wi ll pro duce a s t res s i nt ens i t y fa ct o r t h a t ex ceeds t h e fra ct ure
t o ug h nes s fo r a s t ruct ura l det a i l a s s h o wn i n fi g ure 4-23. Th i s
ca lcula t i o n i s perfo rmed i n t wo co nt ex t s . Th e s i ng le fli g h t P OF i s
t h e pro ba bi li t y o f fra ct ure i n t h e fli g h t g i ven t h a t t h e det a i l h a s
no t fra ct ured previ o us ly. Th e i nt erva l pro ba bi li t y i s t h e
pro ba bi li t y o f fra ct ure a t a ny fli g h t bet ween t h e s t a rt o f a n
a na lys i s (reference t i me o f zero o r a ft er a ma i nt ena nce a ct i o n) a nd
t h e number o f s pect rum h o urs , T. Th i s P OF i s us eful i n predi ct i ng
t h e ex pect ed fra ct ures i n a fleet o f a i rcra ft i n a n i nt erva l a nd i s
requi red fo r t h e ex pect ed co s t s a s s o ci a t ed wi t h a ma i nt ena nce
s ch edule. U s i ng P ROF , va ri o us ki nds o f pa ra met ri c a na lys es ca n be
perfo rmed li ke:
( i ) P ro ba bi li t y o f pa nel fra ct ure bet ween i ns pect i o ns fo r s elect ed
i ns pect i o n i nt erva ls .
(i i ) S elect i ng t h e mo s t o pt i mum i ns pect i o n i nt erva l ba s ed o n co s t .
114
>-
CD
Q
cr
o
cr
Ld
Q_
00
00
<
J
X
I
o
o
X
00
t o
o o
O
Q_
<
cr
o
o
cr
o _
cr
Ld
Z)
Q_
cr
<
x 2
o o
o
O Lx J
_i X
eg
cr
X
o
HI0N31 H0VH3
115
4. 4
1.
10
11
12
13
References
lHa mrh a lc
j c "An Appro x i ma t e Met h o d fo r Es t i ma t i ng t h e
rnt Ta S s e'o f A S h i p's Hufl i n P reli mi na ry D es i g n , P ro ceedi ng s ,
S h i p S t ruct ureC o mmi t t ee S h i p S t ruct ure S ympo s i um, Arlx ng t o n,
Va . , Oct o ber, 1984.
F o rum (TS C F ) a nd publi s h ed by Wi t h erby & C o . Lt d. , U . K. , 1992.
TT
v, D v -Kfrnct ura l S urveys o f Oi l Ta nkers ", Ex x o n pa per
p?es lnt ed a t a jo i nt meet i ng
7
o f t h e Ins t i t ut e o f Ma ri ne
Ing i neeJ s
&
Ro ya l Ins t i t ut i o n o f Na va l Arch i t ect s , Tra ns I Ma r
E Vo l. 96, 1984.
S t udy Repo rt o n Bulk C a rri er Lo s s ", by Ni ppo n Ka i ji Kyo ka i
(NKK), J a pa n, J a nua ry 1992.
"Bulk C a rri ers : Gui da nce * Info rma t i o n t o ff^^^
Opera t o rs ", by Int erna t i o na l As s o ci a t i o n o f C la s s i fi ca t i o n
S o ci et i es (IAC S ), Apri l 1992.
Y Aki t a , Les s o ns Lea rned fro m F a i lure & D a ma g e> o f S h i ps ",
8t h i nt erna t i o na l S h i p S t ruct ures C o ng res s , Gda ns k, 1982.
r A ca rls en et a l. , Les s o ns Lea rned fro m F a i lure & D a ma g e
o f Offs h o rT's t ruct u^s , 8t h Int erna t i o na l S h i p S t ruct ures
C o ng res s , Gda ns k, 1982.
S Ha nzen & O. Ni ls s o n, "Hull D a ma g e i n La rg e S h i ps ", Llo yd's
Ra s t er Tech ni ca l As s o ci a t i o n P a per No . 1, S es s i o n 1972-73.
P ro ceedi ng s o f t h e 10t h Int erna t i o na l S h i p & Offs h o re S t ruc-
t ures C o ng res s , Vo l. 2, Lyng by, Aug us t 1988.
C . R. J o rda n a nd C . S . C o h ra n, "In-S ervi ce ?*f rma nce o f
S t ruct ura l D et a i ls , S h i p S t ruct ure C o mmi t t ee (S S C -272), 1978.
c R J o rda n a nd L. T. Kni g h t , "F urt h er S urvey o f ^"S ervi ce
P erfo rma nce o f S t ruct ura l D et a i l , S h i p S t ruct ure C o mmi t t ee
(S S C -294), 1980.
J enni ng s , E. , et a l. , Inela s t i c D efo rma t i o n o f P la t e P a nels ,
S h i p S t ruct ure C o mmi t t ee (S S C -364), 1991.
wi erzbi cki T. , C h rys s o s t o mi di s , C . a nd Wi erni cki , C . ,
Rup?S re Ana lys i s o f
Y
S
h i p P la t i ng due t o Hydro dyna ji c Wa ve
Impa ct ", S h i p S t ruct ure S ympo s i um, Oct o ber 15 - 16, lya ,
Arli ng t o n, Vi rg i ni a .
116
14. Wi erni cki , C . , "D a ma g e o f S h i p P la t i ng D ue t o Ice Impa ct
Lo a ds ", S NAME, Ma ri ne Tech no lo g y, J a nua ry, 1986.
15. Anders o n, T. L. , "Ela s t i c - P la s t i c F ra ct ure Mech a ni cs , A
C ri t i ca l Revi ew (P a rt 1)", S h i p S t ruct ure C o mmi t t ee (S S C -345),
1990.
16. "Gui da nce o n t h e Met h o ds fo r As s es s i ng t h e Accept a bi li t y o f
F la ws i n F us i o n Welded S t ruct ures , P D 6493: 1991", P ubli s h ed
by t h e Bri t i s h S t a nda rds Ins t i t ut i o n (BS I).
17. Hea ld, P . T. , S pi nk, G. M. a nd Wo rt h i ng t o n, P . J . , "P o s t Yi eld
F ra ct ure Mech a ni cs ", Ma t eri a ls S ci ence a nd Eng i neeri ng , Vo l.
10, 1972.
18. Ro uka , D . P . a nd C a rt wri g h t , D . J . , "C o mpendi um o f S t res s
Int ens i t y F a ct o rs ", HMS O, Lo ndo n, 1976.
19. Ta da , H. , P a ri k, P . a nd Irvi n, G. , "Th e S t res s Ana lys i s o f
C ra cks Ha ndbo o k", D el Res ea rch C o rpo ra t i o n, Hellert o wn,
Ma s s a ch us et t s , 1973.
20. Oo re, M. a nd Burns , D . J . , "Es t i ma t i o n o f S t res s Int ens i t y
F a ct o rs fo r Irreg ula r C ra cks S ubject ed t o Arbi t ra ry No rma l
S t res s F i elds ", P ro ceedi ng s o f 4
t h
Int erna t i o na l C o nference o n
P res s ure Ves s el Tech no lo g y, Lo ndo n, Ins t i t ut i o n o f Mech a ni ca l
Eng i neers , Vo l. 1, 1980.
21. Wi llo ug h by, A. A. a nd D a vey, T. G. , "P la s t i c C o lla ps e a t P a rt
Wa ll F la ws i n P la t es ", AS TM S TP 1020, 1989.
22. Mi ller, A. G. , "Revi ew o f Li mi t Lo a ds o f S t ruct ures C o nt a i ni ng
D efect s ", C ent ra l Elect ri ci t y Genera t i ng Bo a rd Repo rt
TP RD /B/0093/N82 Rev. 1, D ecember, 1984.
23. C a ldwell, J . B. , "U lt i ma t e Lo ng i t udi na l S t reng t h ",
Tra ns a ct i o ns , RINA, 1965.
24. Os t a penko , A. , "S t reng t h o f S h i p Hull Gi rders under Mo ment ,
S h ea r a nd To rque", P ro ceedi ng s , S h i p S t ruct ure C o mmi t t ee
Ex t reme Lo a d S ympo s i um, Arli ng t o n, Va . , Oct o ber, 1981.
25. S mi t h , C . S . , a nd D o w, R. S . , "Res i dua l S t reng t h o f D a ma g ed
S t eel S h i ps a nd Offs h o re S t ruct ures ", J o urna l o f
C o ns t ruct i o na l S t eel Res ea rch , Vo l. 1, No . 4, S ept ember, 1981.
26. F ri eze, P . A. , et a l. "U lt i ma t e Lo a d Beh a vi o r o f P la t es i n
C o mpres s i o n", P ro ceedi ng s Int erna t i o na l S ympo s i um o n S t eel
P la t ed S t ruct ures , C ro s by Lo ckwo o d S t a ples , Lo ndo n, 1977.
117
27. Li t t le, G. H. , "Th e C o lla ps e o f Rect a ng ula r S t eel P la t es U nder
U ni a x i a l C o mpres s i o n", S t ruct ura l Eng i neer, Vo l. 58b, No . 3,
S ept ember, 1980.
28. D o wli ng , P . J . , et a l. , "P la t es i n Bi a x i a l C o mpres s i o n",
C ES LIC Repo rt S P 4, Imperi a l C o lleg e, Lo ndo n, No vember, 1979.
29. D i er, A. F . a nd D o wli ng , P . J . -, "Th e S t reng t h o f P la t es
S ubject ed t o Bi a x i a l F o rces ", P ro ceedi ng s o f C o nference o n
Beh a vi o r o f Th i n-Wa lled S t ruct ures , Gla s g o w, Ma rch 1983.
30. Va ls g a a rd, S . , "U lt i ma t e C a pa ci t y o f P la t es i n Tra ns vers e
C o mpres s i o n", D et No rs ke Veri t a s Repo rt No . 79-0104, F ebrua ry,
1979.
31. S mi t h , C . S . "Imperfect i o n Effect s a nd D es i g n To lera nces i n
S h i ps a nd Offs h o re S t ruct ures ", Tra ns a ct i o n, Ins t i t ut e o f
Eng i neers a nd S h i pbui lders , S co t la nd, Vo l. 124, F ebrua ry,
1981.
32. Va ls g a a rd, S . , "U lt i ma t e C a pa ci t y o f P la t es i n Bi a x i a l In-
P la ne C o mpres s i o n", D et No rs ke Veri t a s Repo rt No . 78-678,
Ma rch , 1979.
33. D i er, A. F . a nd D o wli ng , P . J . , "S t reng t h o f S h i ps P la t i ng :
P la t es U nder C o mbi ned La t era l Lo a di ng a nd Bi a x i a l
C o mpres s i o n", C ES LIC Repo rt S P 8, Imperi a l C o lleg e, Lo ndo n,
No vember, 1980.
34. Oka da , H. , et a l. "C o mpres s i ve S t reng t h o f Lo ng Rect a ng ula r
P la t es U nder Hydro s t a t i c P res s ure", J o urna l S o ci et y Na va l
Arch i t ect ure, J a pa n, Vo l. 146, No v 1979.
35. Ha rdi ng , J . E. et a l. "U lt i ma t e Lo a d Beh a vi o r o f P la t es under
C o mbi ned D i rect a nd S h ea r In-P la ne Lo a di ng ", P ro ceedi ng s o f
Int erna t i o na l S ympo s i um o n S t eel P la t ed S t ruct ures , C ro s by
Lo ckwo o d S t a ples , Lo ndo n, 1977.
36. Hug h es , 0. , a nd F ra nkli n, P . , "D efi ni t i o n a nd Va li da t i o n o f a
P ra ct i ca l Ra t i o na lly-Ba s ed Met h o d fo r t h e F a t i g ue Ana lys i s a nd
D es i g n o f S h i p Hulls ", S NAME T&R Repo rt No . R-41, F ebrua ry,
1993.
37. Newma n, J . C . , J r. , "F a t i g ue C ra ck Gro wt h Ana lys i s o f
S t ruct ures (F AS TRAN) A C lo s ure Mo del", C OS MIC , 1984.
38. Berens , A. P . , et a l. , "Ri s k Ana lys i s F o r Ag i ng Ai rcra ft
F leet s ", Th e U ni vers i t y o f D a yt o n Res ea rch Ins t i t ut e, WL-TR-
91.
39. "Met h o ds a nd Mo dels fo r P redi ct i ng F a t i g ue C ra ck Gro wt h U nder
Ra ndo m Lo a di ng ", AS TM S TP #748, 1981.
118
40. P a rks , D . M. , "Th e Inela s t i c Li ne-S pri ng : Es t i ma t es o f
Ela s t i c-P la s t i c F ra ct ure Mech a ni cs P a ra met ers fo r S urfa ce-
C ra cked P la t es a nd S h ells ", Tra ns a ct i o ns o f t h e AS ME, Vo l.
103, Aug us t , 1981.
41. Kuma r, V. , Germa n, M. D . , a nd S ch uma ch er, B. I. , "Ana lys i s o f
Ela s t i c S urfa ce C ra cks i n C yli nders U s i ng t h e Li ne-S pri ng
Mo del a nd S h ell F i ni t e Element Met h o d", J o urna l o f P res s ure
Ves s el Tech no lo g y, Vo l. 107, No vember, 1985.
42. Kuma r, V. a nd Germa n, M. D . , "S t udi es o f t h e Li ne-S pri ng Mo del
fo r No n-li nea r C ra ck P ro blems ", Tra ns a ct i o ns o f t h e AS ME, Vo l.
107, No vember, 1985.
43. D u, Z. Z. a nd Ha nco ck, J . W. , "S t res s Int ens i t y F a ct o rs o f S emi -
Elli pt i ca l C ra cks i n a Tubula r Welded J o i nt U s i ng Li ne S pri ng s
a nd 3D F i ni t e Element s ", J o urna l o f P res s ure Ves s el
Tech no lo g y, Vo l. Ill, Aug us t , 1989.
44. C h en, Y. K. , et a l. , "U lt i ma t e S t reng t h o f S h i p S t ruct ures ",
Tra ns a ct i o n S NAME, Vo l. 91, 1983.
45. Kut t , L. M. , et a l. , "Eva lua t i o n o f t h e Lo ng i t udi na l U lt i ma t e
S t reng t h o f Va ri o us S h i p Hull C o nfi g ura t i o n", Tra ns a ct i o ns ,
S NAME, Vo l. 93, 1985.
46. S ka a r, K. T. , et a l. , "Ho w lo w ca n s t eel wei g h t g o wi t h s a fet y
a nd eco no my?", P RAD S '87, No rwa y, J une, 1987.
47. Va ls g a a rd, S . & S t een, E. , "U lt i ma t e Hull Gi rder S t reng t h
Ma rg i ns i n P res ent C la s s Requi rement s ", Ma ri ne S t ruct ura l
Ins pect i o n a nd Mo ni t o ri ng S ympo s i um, pres ent ed by t h e S S C a nd
S NAME, Ma rch 1991.
48. U eda , Y. a nd Ra s h ed, S . M. H. , "Th e Idea li zed S t ruct ura l U ni t
Met h o d a nd i t s Appli ca t i o n t o D eep Gi rder S t ruct ures ",
C o mput ers a nd S t ruct ures , Vo l. 18, No . 2, 1984.
49. Int erna t i o na l S h i p S t ruct ures C o ng res s (IS S C ) : "No n-Li nea r
S t ruct ura l Res po ns e", Repo rt o f C o mmi t t ee II. 2, Geno a , 1985
50. D et No rs ke Veri t a s (D nV): "Buckli ng S t reng t h Ana lys i s o f
Mo bi le Offs h o re U ni t s ", C la s s i fi ca t i o n No t es , No t e No . 3 0. 1,
Rev. 1987.
51. Va ls g a a rd, S . , "Numeri ca l D es i g n P redi ct i o ns o f t h e C a pa ci t y
o f P la t es i n Bi a x i a l In-P la ne C o mpres s i o n", C o mput ers a nd
S t ruct ures , Vo l. 12, 1980.
52. Va ls g a a rd, S . a nd F o s s , G. , "Buckli ng Res ea rch i n D et No rs ke
Veri t a s ", C h a pt er 18, Buckli ng o f S h ells i n Offs h o re
S t ruct ures , Gra na da , Lo ndo n, 1982.
119
53. C a rls en, C A. , "A pa ra met ri c s t udy o f co lla ps e o f s t i ffened
pla t es i n co mpres s i o n", Th e S t ruct ura l Eng i neer, Vo l. 58B(2),
1980.
54. Hella n, K. , "Int ro duct i o n t o F ra ct ure Mech a ni cs ", McGra w-
Hi ll, 1984.
55. Y. U eda a nd S . M. H. Ra s h ed, "An U lt i ma t e Tra ns vers e S t reng t h
Ana lys i s o f S h i p S t ruct ure", J . o f S o ci et y o f Na va l Arch i t ect s
o f J a pa n, Vo l. 136, 1974.
56. Y. U eda a nd S . M. H. Ra s h ed, "Th e Idea li zed S t ruct ura l U ni t
Met h o d a nd It s Appli ca t i o n t o D eep Gi rder S t ruct ures ",
C o mput ers a nd S t ruct ures , Vo l. 18, 1984.
57. Y. U eda a nd S . M. H. Ra s h ed a nd J . K. P a i k, "P la t e a nd S t i ffened
P la t e U ni t s o f t h e Idea li zed S t ruct ura l U ni t Met h o d (1s t
Repo rt )", J . o f S o ci et y o f Na va l Arch i t ect s o f J a pa n, Vo l.
156, 1984.
58. Y. U eda , S . M. H. Ra s h ed a nd J . K. P a i k, "P la t e a nd S t i ffened
P la t e U ni t s o f t h e Idea li zed S t ruct ura l U ni t Met h o d (2nd
Repo rt )", J . o f S o ci et y o f Na va l Arch i t ect s o f J a pa n, Vo l.
159, 1986.
59. Y. U eda , S . M. H. Ra s h ed a nd M. Ka t a ya ma , "U lt i ma t e S t reng t h
Ana lys i s o f D o uble Bo t t o m S t ruct ures by Idea li zed S t ruct ura l
by Idea li zed S t ruct ura l U ni t Met h o d", J . o f S o ci et y o f Na va l
Arch i t ect s o f J a pa n, Vo l. 138, 1975.
60. Y. U eda , et a l. , "U lt i ma t e S t reng t h Ana lys i s o f D o uble Bo t t o m
S t ruct ures i n S t ra ndi ng C o ndi t i o ns ", P ro c. o f 3rd
Int erna t i o na l S ympo s i um o n P ra ct i ca l D es i g n o f S h i ps a nd
Mo bi le U ni t s , Tro ndh ei m, No rwa y, 1987.
61. J . K. P a i k, et a l. , "D a ma g e a nd S t reng t h Ana lys i s o f D o uble
Bo t t o m S t ruct ures i n S t ra ndi ng ", P ro c. o f t h e Aut umn Meet i ng
o f S o ci et y o f Na va l Arch i t ect s o f Ko rea , 1990.
62. J . K. P a i k, "U lt i ma t e S t reng t h Ana lys i s o f S h i p S t ruct ures by
Idea li zed S t ruct ura l U ni t Met h o d", D r. D i s s ert a t i o n, Os a ka
U ni vers i t y, J a pa n, 1987.
63. J . K. P a i k, "U lt i ma t e Lo ng i t udi na l S t reng t h Ana lys i s o f t h e
D o uble S ki n Ta nker by Idea li zed S t ruct ura l U ni t Met h o d", P ro c.
o f t h e Aut umn Meet i ng o f S o ci et y o f Na va l Arch i t ect s o f Ko rea ,
1990.
64. J . K. P a i k a nd D . H. Lee, "U lt i ma t e Lo ng i t udi na l S t reng t h -Ba s ed
S a fet y a nd Reli a bi li t y As s es s ment o f S h i p's Hull Gi rder", J .
o f S o ci et y o f Na va l Arch i t ect s o f J a pa n, Vo l. 168, 1990.
120
65. S . M. H. Ra s h ed, "U lt i ma t e S t reng t h a nd P o s t -U lt i ma t e S t reng t h
Beh a vi o r o f D a ma g ed Tubula r S t ruct ura l Members ", D i vi s i o n o f
Ma ri ne S t ruct ures , No rweg i a n Ins t i t ut e o f Tech no lo g y, Repo rt
No . S k/R52, 1980.
66. Y. U eda a nd S . M. H. Ra s h ed, "Beh a vi o r o f D a ma g ed Tubula r
S t ruct ura l Members ", J . o f Energ y Res o urces Tech no lo g y, AS ME,
Vo l. 107, 1985.
67. Y. U eda , S . M. H. Ra s h ed a nd K. Na ka ch o , "New Effi ci ent a nd
Accura t e Met h o d o f No n-li nea r Ana lys i s o f Offs h o re F ra mes (Th e
Idea li zed S t ruct ura l U ni t Met h o d)", J . o f Energ y Res o urces
Tech no lo g y, AS ME, Vo l 107, 1985.
68. S . M. H. Ra s h ed, et a l. , "U lt i ma t e S t reng t h o f J a ck-U p Ri g s i n
S urvi va l a nd P unch -Th ro ug h C o ndi t i o ns ", P ro c. o f 3rd
Int erna t i o na l S ympo s i um o n P ra ct i ca l D es i g n o f S h i ps a nd
Mo bi le U ni t s , Tro ndh ei m, No rwa y, 1987.
69. T. Ya o , J . Ta by a nd T. Mo a n, "U lt i ma t e S t reng t h a nd P o s t -
U lt i ma t e S t reng t h Beh a vi o r o f D a ma g ed Tubula r Members i n
Offs h o re S t ruct ures ", J . o f Offs h o re Mech a ni cs a nd Arct i c
Eng i neers , AS ME, Vo l. 110, 1988.
70. J . K. P a i k a nd B. C . S h i n, "Th eo ret i ca l a nd Ex peri ment a l S t udy
fo r t h e P ro g res s i ve C o lla ps e S t reng t h Ana lys i s o f Tubula r
Offs h o re S t ruct ures ", P ro c. o f 1s t P a ci fi c/As i a Offs h o re
Mech a ni cs S ympo s i um, S eo ul, Ko rea , 1990.
71. J . K. P a i k a nd H. K. Li m, "A S t udy o n t h e U lt i ma t e S t reng t h
Ana lys i s o f F ra me S t ruct ures by Idea li zed S t ruct ura l U ni t
Met h o d", P ro c. o f t h e Aut umn Meet i ng o f C o mput a t i o na l
S t ruct ura l Eng i neeri ng Ins t i t ut e o f Ko rea , 1990.
72. P a i k, J . K. a nd Li m, H. K. , "P ro g res s i ve C o lla ps e S t reng t h
Ana lys i s o f J a cket Type Offs h o re S t ruct ure", P ro c. o f t h e
Aut umn Meet i ng o f Ko rea C o mmi t t ee fo r Ocea n Res o urces a nd
Eng i neeri ng , 1990.
73. Rut h erfo rd, S . E. a nd C a ldwell, J . B. , "U lt i ma t e Lo ng i t udi na l
S t reng t h o f S h i ps : A C a s e S t udy", S NAME Tra ns a ct i o ns , Vo l.
98, 1990.
74. Hug h es , O. F . , "S h i p S t ruct ura l D es i g n", publi s h ed by S NAME, N.
J ers ey, 1988.
75. Meli t z, D . T. , Ro bert s o n, E. J . & D a vi s o n, N. J . , "S t ruct ura l
P erfo rma nce Ma na g ement o f VLC C s - An Owner's Appro a ch ", Ma ri ne
Tech no lo g y, Vo l. 29, No . 4, Oct . 1992.
121
76. Bi s h o p, R. E. D . a nd P ri ce, W. G. , "A U ni fi ed D yna mi c Ana lys i s o f
S h i p Res po ns e t o Wa ves ", Tra ns a ct i o ns , RINA, Vo l. 119, 1977.
77. Bi s h o p, R. E. D . , P ri ce, W. G. a nd Ta rn, P . K. Y. , "On t h e D yna mi cs
o f S la mmi ng ", Tra ns a ct i o ns , RINA, Vo l. 20, 1978.
78. S h i , W. B. a nd Th o mps o n, P . A. , "As pect s o f Vulnera bi li t y o f
Bulk C a rri er S t ruct ures ", RINA, Int erna t i o na l C o nference o n
Bulk C a rri ers a nd Ta nkers - Th e Wa y Ah ea d, D ecember, 1992.
79. P et ers en, E. a nd Va ls g a a rd, S . , "C o lli s i o n Res i s t a nce o f
Ma ri ne S t ruct ures , i n S t ruct ura l C ra s h -Wo rt h i nes s " (ed. N.
J o nes a nd T. Wi erzh i cki ), But t erwo rt h s , 1983.
80. Th e No rweg i a n P et ro leum D i rect o ra t e (NP D ), "Reg ula t i o n fo r
S t ruct ura l D es i g n o f Lo a dbea ri ng S t ruct ures Int ended fo r
Ex plo i t a t i o n o f P et ro leum Res o urces ", 1985.
81. Amda h l, J . , Ta by, J . a nd Gra nli , T. , "P ro g res s i ve C o lla ps e
Ana lys i s o f Mo bi le P la t fo rms ", P ro ceedi ng s o f 3
rd
Int erna t i o na l
S ympo s i um o n P ra ct i ca l D es i g n o f S h i ps a nd Mo bi le U ni t s
(P RAD S ), No rwa y, J une, 1987.
82. Mo a n, T. , "Adva nces i n t h e D es i g n o f Offs h o re S t ruct ures fo r
D a ma g e To lera nce", Int erna t i o na l C o nference o n Adva nces i n
Ma ri ne S t ruct ures , ARE, D umfermli ne, S co t la nd, Ma y, 198 6.
83. Wo o d, H. A. , "Th e U S AF Appro a ch t o S t ruct ura l Li fe Ma na g ement ",
P ro ceedi ng s o f Int erna t i o na l S ympo s i um o n Th e Ro le o f D es i g n,
Ins pect i o n a nd Redunda ncy i n Ma ri ne S t ruct ura l Reli a bi li t y,
C o mmi t t ee o n Ma ri ne S t ruct ures , Ma ri ne Bo a rd, Na t i o na l
Res ea rch C o unci l, Na t i o na l Aca demy P res s , Wa s h i ng t o n, D . C . ,
1983.
84. Neg a a rd, Go rdo n, R. , "Th e Hi s t o ry o f Ai rcra ft S t ruct ura l
Int eg ri t y P ro g ra m", Aero s pa ce S t ruct ures Info rma t i o n a nd
Ana lys i s C ent er (AS IAC ) Repo rt No . 680. IB, 1980.
85. Ki ng , T. T. , "S o me D evelo pment s i n t h e Ai r F o rce Ai rcra ft
S t ruct ura l Int eg ri t y P ro g ra m (AS IP )", AF F D L-TR-70-144, AF
F li g h t D yna mi cs La bo ra t o ry, Wri g h t -P a t t ers o n AF B, Oh i o ,
S ept ember 1970.
86. Burns , J . G. , et a l. , "Ag i ng Ai rcra ft S t ruct ura l D a ma g e
Ana lys i s ", F a t i g ue Ma na g ement : 72
nd
Meet i ng o f t h e AGARD
S t ruct ures a nd Ma t eri a ls P a nels , Ba t h , U K, Apri l-Ma y, 1991.
122
87. Wa rren, D . S . , "D es i g n a nd Ins pect i o n Int errela t i o n fo r
C o mmerci a l J et Tra ns po rt S t ruct ure", P ro ceedi ng s o f
Int erna t i o na l S ympo s i um o n Th e Ro le o f D es i g n, Ins pect i o n a nd
Redunda ncy i n Ma ri ne S t ruct ura l Reli a bi li t y, C o mmi t t ee o n
Ma ri ne S t ruct ures , Ma ri ne Bo a rd, Na t i o na l Res ea rch C o unci l,
Na t i o na l Aca demy P res s , Wa s h i ng t o n, D . C . , 1983.
88. S a nt o s , D . , "F a t i g ue Ma na g ement fo r t h e A-7P ", F a t i g ue
Ma na g ement : 72
nd
Meet i ng o f t h e AGARD S t ruct ures a nd Ma t eri a ls
P a nels , Ba t h , U K, Apri l-Ma y, 1991.
89. MC IC -HB-01, D a ma g e To lera nce D es i g n Ha ndbo o k, Ma y, 1984.
90. Repo rt No . 2-53440/7R-5928, Vo l. 1, D a ma g e To lera nce a nd
F a t i g ue As s es s ment , 31 J a nua ry, 1977.
91. S D M-1000, "P ra ct i ca l Tech ni ques fo r t h e D evelo pment o f Mo de I,
Li nea r Ela s t i c S t res s Int ens i t y S o lut i o ns F o r Typi ca l
S t ruct ura l D et a i ls ", J uly, 1985.
123
5. 0 ULTIMATE STRENGTH ANALYSIS OF A TYPICAL TANKER
Th e purpo s e o f t h i s s ect i o n i s t o i llus t ra t e t h e met h o do lo g y-
pro po s ed t o a s s es s t h e res i dua l s t reng t h o n a da ma g ed t a nker. Th i s
s ect i o n dra ws o n, s ect i o n 2. 0 - t o i dent i fy da ma g e lo ca t i o ns ,
s ect i o n 3. 0 - fo r a ppli ca ble lo ca l fa i lure cri t eri a a nd s ect i o n 4. 0
- fo r ult i ma t e s t reng t h a na lys i s a t t h e g lo ba l level. Ba s ed o n
fi ndi ng s o f s ect i o n 2. 0 a nd h ull s urvey reco rds o f t h e t a nker, t wo
da ma g e lo ca t i o ns were i dent i fi ed. S i nce da ma g e a t bo t h lo ca t i o ns
wa s i n t h e fo rm o f cra cki ng , fra ct ure cri t eri a di s cus s ed i n s ect i o n
3. 0 were us ed fo r fa i lure a s s es s ment a t t h e lo ca l level. On t h e
ba s i s o f t h e lo ca l a na lys i s , da ma g ed co mpo nent s wh i ch wh ere
a s s es s ed a s i neffect i ve i n ca rryi ng furt h er lo a d were neg lect ed i n
t h e g lo ba l a na lys i s . Th e g lo ba l a na lys i s t o eva lua t e t h e ult i ma t e
co lla ps e s t reng t h , bo t h i n t h e i nt a ct a nd da ma g ed co ndi t i o n, wa s
ca rri ed o ut us i ng U LTS TR, wh i ch i s a s i mple ult i ma t e s t reng t h
pro g ra m des cri bed i n s ect i o n 4. 0.
5. 1 D es cri pt i o n o f t h e P ro blem
5. 1. 1 Ta nker C o nfi g ura t i o n
A s i ng le s ki n t a nker a s s h o wn i n fi g ure 5-1, o f leng t h 239. 57m,
brea dt h 32m, dept h 17. 37m a nd di s pla cement 85, 000 t o ns wa s ch o s en
fo r t h e ca s e s t udy. Th e s t ruct ura l co nfi g ura t i o n a t mi ds h i p o f t h e
t a nker, wh i ch h a s a t ra ns vers e fra me s pa ci ng o f 3. 0m, i s s h o wn i n
fi g ure 5-2.
5. 1. 2 Lo ca t i o ns
Lo ca t i o ns were s elect ed ba s ed o n t h e t a nker's h ull s urvey reco rds .
Two i ns t a nces o f da ma g e were ch o s en. Th e fi rs t i ns t a nce wa s
cra cki ng o f t h e s i de s h ell lo ng i t udi na l No . 8 i n a wi ng ba lla s t
t a nk. Th e lo ng i t udi na l wa s lo ca t ed nea r t h e s ummer lo a dli ne,
a ppro x i ma t ely fo ur t ent h s o f t h e dept h belo w t h e upper deck.
Lo ng i t udi na ls i n t h i s reg i o n h a ve been i dent i fi ed, ba s ed upo n
s ervi ce ex peri ence (s ect i o n 2. 0) a nd a na lyt i ca lly i n reference [1],
a s bei ng s us cept i ble t o fa t i g ue cra cki ng . Th e o t h er i ns t a nce wa s
cra cki ng o f bo t t o m lo ng i t udi na l No . 24, a t i t s co nnect i o n wi t h t h e
t ra ns vers e fra me. S i nce t h i s lo ca t i o n i s clo s e t o t h e ex t reme
fi ber o f t h e cro s s -s ect i o n, i t i s i n a h i g h ly s t res s ed reg i o n a nd
t h erefo re, a ny da ma g e co uld be po t ent i a lly ca t a s t ro ph i c.
124
a
CO
2 2 2 2 2
o
en ID o r-* lO 1*1
^
^
m
to tn rsi
") CN ou
T n r"> i^ fl o u>
\Y
CN (N '- ^~ *-
<
_J
c )
M (1-
...J J J
O m > i >
>
<
CL
T T
I
3
1
r i 1 h < )
(') (') < h- L. n
7 V a <
v r. iij cr. in
_i _J UJ o u >
Q:
J
<
o
o
o
L O
00
<
o
L J
>
<
_J
CL
Q
Z
<
o
Q_
UJ
a:
1 0
o
d
125
I/ )
2
<
J
r-~ r^ r> CD 00 m m O) o
^
~"
CN CN * * 1^- m CD in o
o -
ID CO CD CD en CO r^ CD o CD
*
J3
T , , r- oo , m ,
n
,
o o o O oo o 00 o o O
" ^
*JS
r^ r^ M 00 ^ in
^_
00 o
,
CN CN * o a> CTI o f^ m
rn
-
CN ID CO 3- o r-- m n o CO
*
m on 00 m r^ * a-
^- o r~.
o CN CM f) 00 CD * CX) i^
CN CM CN CM CM f> ro * CN
T
\7>Q
-
CN n * in ID r^ CO 01
o
5
z
<
o
o
o
o
oo
LL.
o
o
I
o
J
00
Q-
X
00
Q
CN
U0
L J
cr
o
126
5. 1. 3 Type o f D a ma g e a nd C ri t eri a
At t h e lo ca l level, s t ruct ura l co mpo nent s a re well des i g ned t o
g ua rd a g a i ns t duct i le fo rms o f fa i lure li ke buckli ng a nd yi eldi ng .
Th e res ult s o f t h e li mi t ed da ma g e s urvey pres ent ed i n s ect i o n 2. 0
i ndi ca t es t h a t fa t i g ue cra cki ng i s a do cument ed fo rm o f fa i lure fo r
t a nkers . Th ere a re few us eful g ui deli nes i n t h e ma ri ne i ndus t ry t o
a s s es s a cra cked s t ruct ure. F o r t h i s rea s o n P D 6493 [2] wa s ch o s en
t o det ermi ne cri t i ca l cra ck leng t h a nd cra ck g ro wt h h i s t o ry.
5. 1. 4 Lo a di ng
F o r t h e det ermi na t i o n o f fa t i g ue i nduci ng lo a ds , t h e ABS g ui de o n
F a t i g ue S t reng t h As s es s ment o f Ta nkers [3] wa s fo llo wed. Th e g ui de
i dent i fi es t h e fo llo wi ng t h ree pri ma ry lo a d co mpo nent s t h a t ca us e
fa t i g ue:
(1) h ull g i rder bendi ng mo ment s (vert i ca l a nd h o ri zo na l)
(2) ex t erna l h ydro dyna mi c pres s ure a nd
(. 3) i nt erna l t a nk lo a ds (i nert i a l flui d lo a ds a nd a dded s t a t i c
h ea d due t o ves s el mo t i o n).
Th e g ui de co ns i ders lo a ds t h a t a ri s e fro m o rdi na ry wa ve i nduced
lo a ds . It do es no t t a ke i nt o a cco unt wa ve i mpa ct lo a ds , wh i ppi ng ,
s pri ng i ng , t a nk flui d s lo s h i ng o r vi bra t i ng fo rces due t o ma ch i nery
o r pro pellers .
Ei g h t lo a d ca s es (LC 1-LC 8), dependi ng o n t h e i nt erna l t a nk lo a di ng
pa t t ern a nd t h e dra ft , a re us ed fo r t h e a s s es s ment . Th e lo a di ng
pa t t erns fo r va ri o us i nt erna l t a nk a rra ng ement s a re s h o wn i n fi g ure
5-3, repro duced fro m [3]. Releva nt co effi ci ent s a nd co rrela t i o n
fa ct o rs t o be us ed t o ca lcula t e t h e i ndi vi dua l lo a d co mpo nent s a nd
co mbi ned lo a ds a re pro vi ded i n t h e g ui de a nd h a ve been repro duced
i n t a ble 5-1.
D ependi ng o n t h e lo ca t i o n o f t h e s t ruct ura l co mpo nent i n t h e s h i p,
di fferent co mbi na t i o ns o f t h e lo a d ca s es a re t o be us ed t o fi nd t h e
a ppro pri a t e s t res s ra ng e. Vert i ca lly, t h e s h i p cro s s -s ect i o n h a s
been di vi ded i nt o t wo zo nes . Zo ne A co mpri s es t h e deck a nd bo t t o m
s t ruct ure, s i de s h ell a nd a ll lo ng i t udi na l bulkh ea d s t ruct ure
wi t h i n 15% o f t h e s h i p's mo lded dept h , D , fro m deck o r bo t t o m. Th e
res t 70% o f D , co mpri s i ng o f t h e s i de s h ell a nd a ll lo ng i t udi na l
bulkh ea d s t ruct ures co ns t i t ut es zo ne B. Th es e zo nes a re s h o wn
s ch ema t i ca lly i n fi g ure 5-4. Th e a ppro pri a t e s t res s ra ng e fo r ea ch
zo ne i s ca lcula t ed ba s ed o n t h e fluct ua t i ng lo a d due t o t wo
s elect ed lo a d ca s es a ct i ng t o g et h er. F o r Zo ne A, t h e g rea t er va lue
o f (LC I & LC 2) o r (LC 3 & LC 4) i s us ed a nd fo r Zo ne B, t h e g rea t er
o f (LC 5 & LC 6) o r (LC 7 & LC 8).
127
co
5 5
o
H
EH
<
5 5
H
CQ
s
o
u
Q
<
o
I
in
w
CQ
<
En
IM IM %* Nj
o o o o
6
0
U
p
B
o
w
U
p
S
T
B
D
U
p
CO +
- ^o ^in O O
in vt 00^^
CJ
0)0
w o
+ +
w o
J J
<?ggS3<?
_l O ca ca
3= (- Q
"O 4- oi ca
3 *- O i-
u. < a. co
CM IM > * > *
o' d o' d
6
0
D
o
w
n
B
o
w
D
o
w
n
S
T
B
D
D
o
w
n
i
r^ 1
~ -j- ** o r\ in
L O O
* O a
r
- ^
u
0)0
+
w o
1
S^ 1 w O
O - 'aaiinn
_j a ' CO CO
CO t- a
" Q *j at ca
JVOI-
u- < a. co
ictc
*o
^K l (O *-> m O O
o o
in
in (M o o
. . a a
a.
c =>
O 3 , _
CJ
_1
0)0
ra
CO
+
->o v^ o w o
0
-
0
P
O
R
T
B
H
I
S
T
B
D
H
B
I
0
-
1
CO
m L a
1^- f-
m
cj
+
w fO "> ro
+
w O
+
^ o
in o
o" d
in i
O f\J o o
, , O Q
c
3
o
o a ,
Q
Old
O
^o
o - - o ' ' 3: 3: o -
co ca
O 3 0
CO
3=
> o
I
CO
oc ca
O 1-
Q. CO
in in
IM IM
^
o o
a
03-3.
0
ca
>* + m i
w r^ ^> o O o o o ins o o
c_> . i * . . . . o o
OlO o o o o 3: 3: - o
_j o i ca ca
TJ "-
3 --
u. <
in in
IM CM
fO
w h- / -\ o o o
in o
Ln?
o f^ ' o o
0
D
o
w
n
B
o
w
D
o
w
n
u
0)0
+
o o
o - r-' O ' ' ^ O'
_J 10
CO
ca ca '
3 > *-
u. <
in in
f\i r\j
^s o' o"
a.
S-3 '
O
CO
f\J + in i
^o ^in o o in o -j- r- o o
cj i * . . a a
0)T- w o o o o - o o n: 3: .- o
a.
or
CO
1 o
31
i CO CO
3 H-
u. <
L U
O
z
in in o
(M (M
I
-
w O ^in o o
in o * r> - o o
o
z
0
D
o
w
n
B
o
w
D
o
w
n
CJ
OH-
+
v^ o o o
d >- o" o' g ' ' .-' o'
o
z
_J IO
CO
L U
ca ca '
~a w
<
C3
Q
3 H- z
3 u. <
CO a
co L U
CO
L U
<
L U
a QC c c 3=
< 3 a.
o CO L U
_i CO
L U 3=
>
<
oc C < 3
L U si u- a. 1
Q L U 0)
ac z u_ ca CO u _1 CJ
" < < z O)
o m CO z Z L U c
to < o ac C .c ct: <
1 +-J 4- L U L U U i L U
_1 < o (D c c 1 1 ai "
3 u U o o X Z
" ~ o
L U c
3: N o M y L U
0. OL
DC < U JZ .
*-* .x *- J* i- J^ - .* T3 > O *
t_ t_ t_ t_ CO 03 -M
a) 01 o o o .
u > _ &
(U 01 - 0
< > > 31 3; CQ ji -** CJ -* 3 3 3 CJ CJ o x z Q. a:
128
\\\\\
\ \ \ \ \
\\\\\
\\\\\
\ \ \\\
wwv
<
CL
00
LJ
Q
00
CM
=tte
00
UJ
oo
<
Q
<
O
_Q
\\\\\ \\\\ \
^g
\ \ V \ \ \\\\\
\ \ \ \ \ \\\\\
XXXXX
\ \ \ \ \ \ \ \ \ \
\\\\\ \ \ \\ \
\ \ '\ \ \l \\\\\
Li_
<
CL
o
oo
J
Q
r-O
CM
to
=tte
00
UJ
00
<
CJ
Q
<
O
~o
00
z
CL
<
CL
CO
z
<
O
v \ \ v N
\ \ \ \ M
\\WN
\ \ \ \ \l
\ \ \ \ \ \\\\\
\ \ \ \ \ \ \ \ \\
\ \ \ \ \ \ \ \ \\
\ \ \ \\ \\\\\
<
CL
Q
O
0 0
UJ
O
rO
ro
=tfc
00
UJ
00
<
o
O
<
o
, \ \ \ \l
\ \ \ \ M
\\\\\
\yO\XXz
xX>Cv<
xx>5<5<
\ \ \ \ \
w\\\
XXXXX Isll
ww\
<
CL
Q
CO
00
UJ
Q
ro
CM
00
UJ
00
<
o
<
O
o
^L
Z
<
I
rO
if)
_0
O
129
o
NJ
o O
M
. in .
r^-2-
m
L d
2
O
M
<%
<
L d
Z z:
o o
K J
i
c)
J
111
_L
fO
1
O
CO
(O
z.
o
< :
01
O
u
T
CO
L U
CJ
M
, z:
<
<
L d <
X
O L u
CO O
ZD
C D
130
5. 2 Lo ca l Level Ana lys i s
Ana lys i s a t t h e lo ca l level i nvo lves det ermi na t i o n o f t h e
cri t i ca l cra ck leng t h a nd develo pment o f a cra ck g ro wt h h i s t o ry.
D a ma g e a t t h e t wo lo ca t i o ns ment i o ned a bo ve, h a ve been a na lyzed i n
t h e fo llo wi ng t wo ex a mples . Wh i le i n ex a mple 1, t h e ra t i o o f t h e
no mi na l t ens i le s t res s , o
lt
a ct i ng o n t h e cra cked s t ruct ure t o t h e
yi eld s t reng t h , a
y
, i s les s t h a n 0. 5, i n ex a mple 2 i t i s g rea t er
t h a n 0. 5. Th erefo re, cra ck a s s es s ment wa s ca rri ed o ut a t level o ne
o f P D 6493 fo r ex a mple 1 a nd a t level t wo fo r ex a mple 2.
5. 2. 1 Ex a mple 1
In t h i s ex a mple, cra cki ng o f t h e web o f t h e s i de s h ell lo ng i t udi na l
No . 8 i n No . 5 s t a rbo a rd ba lla s t t a nk i s a na lyzed. Th e
lo ng i t udi na l, i s a n a ng le ba r wi t h a web dept h o f 228. 6 mm, fla ng e
wi dt h o f 101. 6 mm a nd t h i cknes s o f 12. 7 mm. Th e cra ck o ccurred
bet ween fra mes 66 a nd 67. It ex t ends fo r 178 mm i n t h e web o f t h e
s t i ffener a nd 7 6 mm i n t h e fla ng e. A des cri pt i o n o f t h e
lo ng i t udi na l wi t h t h e cra ck a nd t h e a djo i ni ng reg i o n i s s h o wn
s ch ema t i ca lly i n fi g ure 5-5. Th e lo ng i t udi na l i s lo ca t ed 11. 01m
a bo ve t h e ba s eli ne, wh i ch i s a bo ut 60% o f D a nd t h erefo re i n zo ne
B. Th e a na lys i s wa s di vi ded i nt o t wo pa rt s . In t h e fi rs t pa rt ,
t h e a llo wa ble cra ck leng t h wa s det ermi ned ba s ed o n t h e s t res s es
ca lcula t ed us i ng equa t i o ns fo und i n [3] a nd cra ck a s s es s ment
pro cedures o f [2]. In pa rt 2, cra ck g ro wt h a na lys i s wa s perfo rmed
by numeri ca l i nt eg ra t i o n o f P a ri s ' equa t i o n. Th e det a i ls o f t h es e
pro cedures a re pro vi ded belo w.
P a rt 1 - D et ermi na t i o n o f Allo wa ble C ra ck Leng t h
Met h o do lo g y: Acco rdi ng t o P D 6493, t h e no rma l t ens i le s t res s , a
x
,
a ct i ng o n t h e cra cked s t ruct ure ca n be res o lved i nt o t h e pri ma ry
s t res s , P , due t o t h e ex t erna lly a ppli ed lo a d, s eco nda ry s t res s , Q,
due t o res i dua l s t res s es a nd pea k s t res s , F , due t o co ncent ra t i o ns
ca us ed by lo ca l di s co nt i nui t i es . Th e s t res s due t o h ull g i rder
bendi ng ma ke up t h e membra ne co mpo nent o f t h e pri ma ry s t res s , P
m
,
wh i le t h e lo ca l bendi ng o f t h e s t i ffener due t o ex t erna l a nd
i nt erna l pres s ure lo a ds ma ke up t h e bendi ng co mpo nent , P
b
. Va lues
o f h ull bendi ng mo ment s , wh i ch were ca lcula t ed a cco rdi ng t o [3],
a re pres ent ed i n t a ble 5-2 a lo ng wi t h t h e co rres po ndi ng s t res s
va lues fo r t h e fo llo wi ng s ect i o na l pro pert i es o f t h e t a nker:
Mo ment o f i nert i a a bo ut t h e h o ri zo nt a l N. A. I
yy
= 2. 82x l0
6
m
2
-cm
2
Mo ment o f i nert i a a bo ut t h e vert i ca l N. A. I
a
= 5. 51x l0
6
m
2
-cm
2
Hei g h t o f N. A. a bo ve ba s eli ne = 8. 57m
Vert i ca l di s t a nce o f S S L No . 8 fro m t h e N. A. y = 2. 44m
Tra ns vers e di s t a nce o f S S L No . 8 fro m t h e C . L. z = 16. 00m
131
<
CD
L L L L L L LV L A L L L L L L L L L L
L U
X
00
L U
Q
0 0
CN
C D
C O
cr
,
u_
LU
h -
_l
< C L
^
Ld
< r
o X
< 111
>
< r
~
Q
uu
o
_
/"
O
_l
z <
o
z
h - n
LL
~ )
(H
h -
o o
00 2
LU
o
LU _i
LO
LO
LU
CC
Z)
o
^ r r r r r r r r r r r r r r r r r r
r r r \
132
TABLE 5-2 BEND ING MOMENT AND S TRES S ES AT S S L No . 8
Bendi ng Mo ment
(N-m)
S t res s es
( N/mm
2
)
S t i ll Wa t er B. M.
2. 52 X 10
9
22. 0
Vert i ca l Wa ve
B. M. (S a g g i ng )
-1. 14 x 10
9
-10. 0
Vert i ca l Wa ve
B. M. (Ho g g i ng )
1. 07 x 10
9
9. 0
Ho ri zo nt a l Wa ve
B. M.
1. 38 x 10
9
40. 0
F o r cra ck a na lys i s o nly t ens i le s t res s es a re co ns i dered, t h erefo re
a n a vera g e o f t h e a bs o lut e va lues o f t h e vert i ca l wa ve bendi ng
s t res s es due t o s a g g i ng a nd h o g g i ng wa s co ns i dered a lo ng wi t h t h e
s t i ll wa t er bendi ng s t res s a nd h o ri zo nt a l wa ve bendi ng s t res s t o
det ermi ne P
m
. Th erefo re,
P
m
= 22 + 1/2 (10+ 9) + 40 = 71. 5 N/mm
2
Th e s t res s ra ng e, f*
r2
/ ca us ed by t h e lo ca l bendi ng o f t h e s t i ffener
due t o pres s ure lo a di ng wa s co mput ed a cco rdi ng t o t h e fo llo wi ng
fo rmula g i ven i n [3]
f*
r2
= C , M/S M
(5. 1)
wh ere,
C t =
1. 5; co rrect i o n fa ct o r fo r co mbi ned bendi ng a nd
t o rs i o na l s t res s ,
S M = 4. 916 m-cm
2
; s ect i o na l mo dulus o f t h e lo ng i t udi na l a nd
t h e a s s o ci a t ed effect i ve pla t i ng a t t h e fla ng e,
M = k-p-s -1
2
; bendi ng mo ment a t t h e s uppo rt ed ends o f t h e
lo ng i t udi na l.
In t h e a bo ve ex pres s i o n fo r M,
k = 1. 15/12; fa ct o r a cco unt i ng fo r t h e fi x i t y o f t h e
s t i ffener,
s = 800 mm; s t i ffener s pa ci ng ,
1 = 2. 25m; uns uppo rt ed s pa n o f t h e s t i ffener,
133
a nd p = 0. 045 N/mm
2
; t o t a l ra ng e o f t h e net fluct ua t i ng pres s ure,
ba s ed o n equa t i o ns fo und i n [3].
Ba s ed pn t h e a bo ve va lues , f*
r2
= 54 N/mm
2
fo r t h i s cra ck lo ca t i o n.
P
b
, t h e bendi ng co mpo nent o f t h e pri ma ry s t res s wa s t a ken a s h a lf
o f t h e s t res s ra ng e f*
r2
, wh i ch fluct ua t es bet ween co mpres s i o n a nd
t ens i o n. Th erefo re,
P
b
= 0. 5 x 54 = 27 N/mm
2
Beca us e t h e cra ck i s lo ca t ed o n a lo ng i t udi na l s t i ffener i n a n a rea
bet ween t wo t ra ns vers e fra mes , wi t h no co nnect i ng members i n t h e
vi ci ni t y, t h e s eco nda ry s t res s , Q, due t o weldi ng a nd pea k s t res s ,
F , due t o lo ca l di s co nt i nui t i es were a s s umed t o be a bs ent .
Th e ma t eri a l pro pert i es o f t h e lo ng i t udi na l s t i ffener were a s s umed
t o be t h o s e o f Gra de A mi ld s t eel. Th e da t a li s t ed belo w a re fro m
[4] fo r pla t e t h i cknes s les s t h a n 25 mm.
Steel Type
(Grade)
Yo ung 's
Mo dulus
E(N/mm
2
)
Yi eld
S t reng t h
a
y
(N/mm
2
)
U lt i ma t e
S t reng t h
a
u
(N/mm
2
)
F ra ct ures
To ug h nes s
K
IC
(N/mm
IJ
)
Mi ni mum
S ervi ce
Tempera t ure
( C )
MS (A) 207000 235 440 2136 0
As ment i o ned^i n [5], t h e li kely ra ng e o f fra ct ure t o ug h nes s , S
fo r mi ld s t eel g ra des li es bet ween 0. 15 mm a nd 0. 60 mm a t 0
20 C a nd no rma l s t ra i n ra t es . In t h i s ex a mple, a t a n o pera t i ng
t empera t ure o f 2 0 C , 5
C R
wa s a s s umed t o be 0. 25 mm.
C -
F ra ct ure a s s es s ment i n t h i s ex a mple wa s ca rri ed o ut a t level 1 o f
P D 6493. Th i s i s a n i ni t i a l s creeni ng level wh i ch pro vi des very
co ns erva t i ve res ult s . At t h i s level t h e ma x i mum t ens i le s t res s ,
'i /
i s t a ken t o be uni fo rm a nd equa l t o t h e ma x i mum s um o f t h e va lues
o f t h e s t res s co mpo nent s (P
m
+ P
b
+ Q + F ) . S i nce Q a nd F were
a s s umed t o be zero ,
= P
m
+ P
b
71. 5 + 27 = 98. 5 N/mm
2
D ependi ng o n t h e fra ct ure t o ug h nes s va lues , wh i ch co uld be ba s ed o n
t h e K-met h o d o r t h e 5-met h o d, t wo a llo wa ble cra ck leng t h s were
o bt a i ned.
134
K-Method:
mat
2
V U
( 5.2)
where K^ = K
IC
. Therefore,
5
m
- _L( 2136f -
7
2
m
27t { 98.5 J
mm
C TOD -Met h o d:
m
6^
2
'cjjV
2
V y/
for -i s 0.5
( 5.3)
a
n,
=
m
mat
27t -! - 0.25
o..
for -1 > 0.5
where S,^ = <S
CR
. Since,
o 235
= 0.42 < 0.5
Therefore,
0.25x207000
loft
a_ = = 199 mm
2TT ( 0.42)
2
235
1 3 5
F i ni t e Wi dt h C o rrect i o n: If t h e ca lcula t ed
m
ex ceeds o ne
t went i et h o f t h e wi dt h , W, o f t h e co mpo nent , wh i ch i n t h i s ca s e i s
t h e dept h o f t h e web o f t h e lo ng i t udi na l s t i ffener (228. 6 mm),
P D 6493 s ug g es t s a fi ni t e wi dt h co rrect i o n t o be a ppli ed t o
m
:
3 = 3,
m(cor) ml
l +
'a ^
m
lw
(5. 4)
On a pplyi ng t h i s co rrect i o n, t h e a llo wa ble cra ck leng t h s were:
K-Met h o d - 56 mm
C TOD - 106 mm
In level 1 a s s es s ment , a ch eck a g a i ns t pla s t i c co lla ps e o f t h e
co mpo nent wa s ma de s epa ra t ely by ca lcula t i ng t h e co lla ps e ra t i o , S
r
.
P la s t i c co lla ps e o ccurs due t o t h e o vera ll yi eldi ng o f t h e
uncra cked li g a ment s o f t h e cro s s -s ect i o n. S
r
i s t h e ra t i o o f
effect i ve net s ect i o n s t res s , a
n
, t o t h e flo w s t reng t h , a
f
, o f t h e
co mpo nent wh i ch s h o uld no t ex ceed 0. 8.
F o r fla t pla t es , P D 6493 s ug g es t t h e fo llo wi ng fo rmula t o ca lcula t e
a :
n =
P
b
+
C P b
+
9P i
05
(5. 5)
wh ere "a " i s t h e cra ck leng t h a nd t h e flo w s t reng t h , a
f
, i s defi ned
a s t h e les s er o f t h e a vera g e o f yi eld s t reng t h ,
t ens i le s t reng t h , a
u
, o r 1. 2a
y
. Th erefo re,
'y
a nd ult i ma t e
\a + a
a
f
- mi m- -, \2a
x
282 N/mm
:
(5. 6)
Th e S
r
va lues ca lcula t ed ba s ed o n t h e a llo wa ble cra ck leng t h o f
56 mm a nd 106 mm were 0. 38 a nd 0. 53 res pect i vely wh i ch a re les s
t h a n 0. 8 i mplyi ng t h ey wo uld no t res ult i n pla s t i c co lla ps e.
136
Th e a llo wa ble cra ck leng t h predi ct ed by t h e K-met h o d i s
co ns erva t i ve beca us e i t i s ba s ed o n fra ct ure t o ug h nes s va lue, KJ C ,
det ermi ned fro m co upo n t es t s under pla i n s t ra i n co ndi t i o ns . P la i n
s t ra i n co ndi t i o ns i mply h i g h deg ree o f co ns t ra i nt a ro und t h e
cra cked reg i o n. Th e fra ct ure t o ug h nes s , < S
C R
, us ed i n t h e C TOD
met h o d a re ba s ed o n ex peri ment s wh i ch a llo w fo r pla s t i ci t y o f t h e
s peci men a nd h ence a re clo s e t o co ndi t i o ns ex peri enced by rea l
s t ruct ures under mi ni mum co ns t ra i nt . In t h i s ex a mple, t h e cra ck i s
a s s umed t o be i n a reg i o n o f rela t i vely lo w co ns t ra i nt . Th erefo re
t h e cra ck leng t h , 106 mm, predi ct ed by t h e C TOD met h o d wa s
co ns i dered t o be t h e ma x i mum a llo wa ble cra ck leng t h .
P a rt 2 - C ra ck Gro wt h Ana lys i s bv t h e Genera l P ro cedure
Met h o do lo g y: To det ermi ne t h e rela t i o ns h i p bet ween cra ck leng t h
a nd t i me, P D 6493 s ug g es t s i nt eg ra t i o n o f t h e P a ri s ' equa t i o n wh i ch
i s :
-^- = C (AKj)
m
(5. 7)
dN
wh ere,
da /dN i s t h e ra t e o f cra ck pro pa g a t i o n
'C a nd 'm' a re co ns t a nt s dependi ng o n t h e ma t eri a l,
envi ro nment a nd frequency o f t h e a ppli ed lo a d.
AKj i s t h e ra ng e o f s t res s i nt ens i t y fa ct o r co rres po ndi ng t o
t h e a ppli ed s t res s cycles a nd i ns t a nt a neo us cra ck leng t h .
F o r s t ruct ura l ferri t i c s t eels o pera t i ng i n ma ri ne envi ro nment s a t
t empera t ures up t o 20 C , P D 6493 s ug g es t s t h e us e o f fo llo wi ng
va lues fo r 'C a nd 'm':
C = 2. 3 x 10"
12
m = 3. 0
Th e t h res h o ld s t res s i nt ens i t y fa ct o r, AK, , , , belo w wh i ch t h ere i s no
cra ck g ro wt h wa s a s s umed t o be 63 N-mm
_3/2
per P D 6493.
AK, fo r a t h ro ug h t h i cknes s cra ck wa s ca lcula t ed us i ng t h e fo llo wi ng
ex pres s i o n fro m P D 6493:
AK, = (AP
m
+
AP , ) ji a S ec ^T (5-8)
137
wh ere,
AP
m
i s t h e pri ma ry membra ne s t res s ra ng e
AP
b
i s t h e pri ma ry bendi ng s t res s ra ng e
a i s t h e i ns t a nt a neo us cra ck leng t h a nd
W i s t h e wi dt h o f t h e co mpo nent .
AP
m
res ult s fro m h o ri zo na l a nd vert i ca l h ull g i rder bendi ng due t o
wa ves . S t i ll wa t er bendi ng i s no t co ns i dered s i nce i t i s a s t ea dy,
unfluct ua t i ng s t res s wh i ch do es no t co nt ri but e t o fa t i g ue cra ck
g ro wt h . Th erefo re,
AP
ffi
= | ff
V
s | + | O-VHI + 2| a
H
| (5. 9)
wh ere a
vs
a nd C T
VH
a re t h e vert i ca l bendi ng s t res s es due t o s a g g i ng
a nd h o g g i ng res pect i vely a nd o
H
i s t h e h o ri zo nt a l wa ve bendi ng
s t res s . U s i ng t h e va lues o f t h e s t res s es fro m t a ble 5-2,
AP
m
= 10 + 9 + 2(40) = 99 N/mm
2
.
AP
b
i s t h e s t res s ra ng e due t o lo ca l bendi ng o f t h e s t i ffener ca us ed
by t h e fluct ua t i ng pres s ure lo a d a nd wa s ca lcula t ed previ o us ly a s
f*
r2
i n equa t i o n (5. 1). Th us ,
AP
b
= 54 N/mm
2
.
In t h e a bs ence o f a mea s ured s t res s s pect rum a ct i ng o n t h e s h i p, a
s t res s s pect rum wa s g enera t ed fro m da t a a va i la ble fro m [4]. Th i s
da t a reflect s s h o rt -t erm s i mula t ed No rt h -S ea co ndi t i o ns us ed fo r
fa t i g ue s t udi es i n t h e o ffs h o re i ndus t ry. Th i s co ndi t i o n co ns i s t s
o f eleven wa ve h ei g h t s o r s ea s t a t es . Th e ma x i mum s i g ni fi ca nt wa ve
h ei g h t i s 16m. F o r ea ch s ea -s t a t e, i t s s i g ni fi ca nt wa ve h ei g h t ,
H
s i
, fra ct i o n o f t i me s pent i n t h a t s ea -s t a t e, p
i ;
a nd i t s frequency,
f
;
a re pres ent ed i n t a ble 5-3. Th e s t res s s pect rum wa s o bt a i ned by
a s s umi ng t h e ma x i mum s t res s ra ng e, cr
rma x
, wh i ch i s t h e s um o f AP
m
a nd
AP
b
, co rres po nds t o t h e ma x i mum s i g ni fi ca nt wa ve h ei g h t . Th e o t h er
s t res s ra ng es , a
ri
, were o bt a i ned a s fo llo ws :
^ri \ **s i / ^s ma x / r ma x
In t h i s ex a mple i t wa s a s s umed t h a t :
(a ) t h e leng t h o f ea ch vo ya g e i s 30 da ys o r 0. 08 yea r a nd
(b) fo r s h o rt t erm da ma g e es t i ma t i o n, t h e s t res s s pect rum i s
repea t ed every mo nt h .
Th e number o f cycles , nj co rres po ndi ng t o a pa rt i cula r s t res s ra ng e
wa s ca lcula t ed a s :
138
n
;
= p; fj T
wh ere, T = 2627895 s ec (30 da ys ).
TABLE 5-3 S HORT TERM NORTH S EA S TRES S S P EC TRU M
Wa ve H
s i P i
fj
< 7
ri "i
Level (m) (Hz) (N/mm
2
) (C ycles )
1 16. 0 0. 0000368 0. 0976 153. 20 9
2 14. 5 0. 0000932 0. 1040 138. 83 25
3 13. 0 0. 0003700 0. 1090 124. 47 106
4 11. 5 0. 0022000 0. 1200 110. 11 694
5 10. 0 0. 0073000 0. 1330 95. 75 2551
6 8. 5 0. 0135000 0. 1440 81. 39 5109
7 7. 0 0. 0265000 0. 1600 67. 02 11142
8 5. 5 0. 0600000 0. 1780 52. 66 28066
9 4. 0 0. 2100000 0. 1990 38. 30 109820
10 2. 5 0. 4900000 0. 2230 23. 94 287150
11 1. 0 0. 1900000 0. 2710 9. 57 135310
Th e s t res s ra ng es , a
ri
a nd number o f cycles , n
lt
a re s h o wn i n t a ble
5-3. C ra ck g ro wt h under va ri a ble a mpli t ude lo a di ng i s a ffect ed by
t h e s equence o f a ppli ca t i o n o f t h e s t res s ra ng es i n t h e s pect ra .
As demo ns t ra t ed i n references [6, 7], t h e ro o t mea n s qua re (rms )
va lue o f t h e s t res s s pect ra ca n be us ed i n t h e ca lcula t i o n o f AK.
Th e rms va lue o f t h e s t res s ra ng es wa s ca lcula t ed us i ng t h e
fo llo wi ng fo rmula :
Ao .
N
K
S n^
i =l
K
S n,
i =l
(5. 10)
wh ere,
K = number o f s t res s ra ng es ; 11 i n t h i s ca s e
n
s
= number o f cycles fo r t h e i * s t res s ra ng e
a
ri
= va lue o f t h e i * s t res s ra ng e,
A^i ms = ro o t mea n s qua re va lue o f t h e s t res s ra ng es .
139
F o r t h e a s s umed s t res s s pect rum, wh i ch h a s a ma x i mum s t res s ra ng e,
cj
rma x
= 153 N/mm
2
, t h e Ao
ma
wa s ca lcula t ed t o be 3 0. 3 N/mm
2
. Th i s
va lue wa s us ed i n ca lcula t i ng t h e s t res s i nt ens i t y fa ct o r ra ng e, AK
us ed i n t h e cra ck g ro wt h equa t i o n (5. 7).
Th erefo re,
AK , = Aa Ait a
uns nns
(5. 11)
Th e cra ck pro pa g a t i o n equa t i o n (5. 7), wa s i nt eg ra t ed numeri ca lly
us i ng a s t ep by s t ep pro cedure. Th es e s t eps a re ex pla i ned belo w:
1. D et ermi na t i o n o f a ^ S t a rt i ng wi t h a n i ni t i a l cra ck leng t h , a
0
,
a t ea ch s t ep 'j' t h e cra ck leng t h wa s i ncrea s ed by a predet ermi ned
i ncrement , Aa . Increment s a re ei t h er uni fo rm o r lo g ri t h mi c.
Lo g ri t h mi c i ncrement s a re us ed wh en t h e des i red number o f
i ncrement s i s s ma ll. Th e pres ent wo rk ut i li zed lo g ri t h mi c
i ncrement s , ca lcula t ed a s ,
lo g (a
f
) - lo g C a p)
Aa
wh ere:
a
f
= 106 mm ; t h e fi na l cra ck leng t h
a
0
= 0. 005 mm; a s s umed i ni t i a l cra ck leng t h
Njn, . = 100; number o f i ncrement s
s o t h a t t h e cra ck leng t h a t t h e end o f t h e j* i ncrement i s
a
J
= 10
lo g (a o )
+
jAa
(
5
-
13
)
2. D et ermi na t i o n o f AKJ : AKj us ed fo r t h e ca lcula t i o n wa s a s s umed
t o be co ns t a nt o ver ea ch i ncrement a nd wa s ba s ed o n t h e cra ck
leng t h a t t h e end o f t h e i ncrement . Th erefo re:
IKj * Lojm a. Sec '>
(5
"
14)
140
3. D et ermi na t i o n o f Njt Th e number o f cycles us ed i n ex t endi ng t h e
cra ck o ver ea ch i ncrement , Nj, wa s ca lcula t ed us i ng equa t i o n (5. 7).
Th us ,
N. = 1 "0 - '> (5. 15)
J
C (AKp
m
wh ere,
a j. , i s t h e cra ck leng t h a t t h e beg i nni ng o f t h e j* s t ep,
a nd a j i s t h e cra ck leng t h a t t h e end o f t h e j* s t ep.
Th e cumula t i ve s um o f t h e number o f cycles a t ea ch s t ep, N
w
N
2
,
. . . . Nj g a ve t h e t o t a l number o f cycles ex pended t o rea ch t h e cra ck
leng t h , a j, a t t h e end o f t h e j* s t ep.
Th e res ult s o f t h e cra ck g ro wt h a na lys i s a re s h o wn i n fi g ure 5-6.
C o nclus i o ns ;
Ba s ed o n a s s umpt i o ns o n lo a di ng , s t res s es t i ma t i o n, g enera t i o n o f
t h e s ea -s pect ra a nd t h e co ns erva t i s m o f t h e pro cedures o f P D 6493
i n det ermi ni ng cri t i ca l cra ck leng t h , a n a llo wa ble cra ck leng t h o f
106 mm wa s det ermi ned. A cra ck lo ng er t h a n t h e a llo wa ble ma y' lea d
t o uns t a ble fra ct ure. Th erefo re t h e o bs erved cra ck o f 178 mm o n
t h e web o f t h e lo ng i t udi na l i s no t -permi s s i ble. F ro m t h e res ult s
o f cra ck g ro wt h a na lys i s do ne i n P a rt 2, i t i s evi dent t h a t i t
t a kes li t t le o ver 6 vo ya g es fo r a n i ni t i a l cra ck o f leng t h 0. 005 mm
t o rea ch a n a llo wa ble leng t h o f 106 mm. Th erefo re co ns i deri ng a
fa ct o r o f s a fet y o f 2, t h i s cra ck s h o uld be repa i red wi t h i n
a ppro x i ma t ely t h ree mo nt h s o f i t s di s co very.
5. 2. 2 Ex a mple 2
At fra me 81B o f No . 3 po rt ba lla s t t a nk, cra cks were fo und a t t h e
a t t a ch ment welds o f t h e vert i ca l fla t ba r s t i ffener t o t h e bo t t o m
lo ng i t udi na l No . 24 o n t h e a ft s i de o f t h e t ra ns vers e web fra me.
A s ch ema t i c o f t h e reg i o n i s s h o wn i n fi g ure 5-7. In t h i s ex a mple,
cra ck a s s es s ment wa s perfo rmed a t level 2 o f P D 6493 s i nce t h e
ra t i o o f no mi na l t ens i le s t res s t o t h e yi eld s t reng t h wa s g rea t er
t h a n 0. 5. At t h i s level pla s t i c co lla ps e i s i mpli ci t ly t a ken i nt o
a cco unt t h ro ug h i nt era ct i o n curves . A s i mi la r a na lys i s , a s do ne i n
ex a mple 1, wa s ca rri ed o ut fo r t h i s ex a mple, t h e res ult s o f wh i ch
a re pres ent ed belo w.
141
Millions
No. OF CYCLES, N
CO
120
100
80
| 60
W
U 40
20
U
1
1
/
/
2 3 4 5
NUMBER OF V OYAGES
FIGURE 5.6 CRACK GROWTH PROFILE, EXAMPLE 1
142
L L LLLLLLLLLLLLLLLLLLL
rrrrr
rrrrrrrrrrrrrr
o
I
h-
o
m
CD
00
cr
h-
<
CM
CL
L J x
O L d
<
<
^f-
-
O
_i
Z <
o z
I- Q
Q: t:
o o
en z
J o
Q _J
J
cr
O
143
P a rt 1 - D et ermi na t i o n o f Allo wa ble C ra ck Leng t h :
Hull bendi ng mo ment s a nd pres s ures were ca lcula t ed i n a cco rda nce
wi t h t h e ABS g ui de [3] fo r Zo ne A. Th e ca lcula t i o n o f h ull g i rder
s t res s es i n zo ne A neg lect s t h e co nt ri but i o n o f h o ri zo nt a l wa ve
bendi ng mo ment , per [3]. Va lues o f h ull bendi ng mo ment s a nd
s t res s es a re pres ent ed i n t a ble 5-4 fo r t h e fo llo wi ng s ect i o n
pro pert i es :
Mo ment o f i nert i a a bo ut t h e h o ri zo nt a l N. A.
Mo ment o f i nert i a a bo ut t h e vert i ca l N. A.
Hei g h t o f N. A. a bo ve ba s eli ne
Vert i ca l di s t a nce o f BL No . 24 fro m t h e N. A.
Tra ns vers e di s t a nce o f BL No . 24 fro m t h e C . L.
Iyy
y
z
2.82X10
6
m
2
-cm
2
5. 5 lxl 0
6
m
2
-cm
2
8. 57m
8. 57m
11. 82m
TABLE 5-4 BEND ING MOMENT AND S TRES S ES AT BL #24
Bendi ng Mo ment
(N-m)
S t res s es
(N/mm
2
)
S t i ll Wa t er B. M.
2. 52 x 10
9
76. 0
Vert i ca l Wa ve
B. M. (S a g g i ng )
-3. 18 x 10
9
-97. 0
Vert i ca l Wa ve
B. M. (Ho g g i ng )
2. 98 x 10
9
91. 0
Ho ri zo nt a l Wa ve
B. M.
0. 0
0. 0
Th e s t res s ra ng e, -f*
r2
, ca us ed by t h e lo ca l bendi ng o f t h e s t i ffener
due t o pres s ure lo a di ng wa s co mput ed t o be 15 N/mm
2
a cco rdi ng t o
equa t i o n (5. 1) ba s ed o n t h e fo llo wi ng va lues :
C , = 1. 5, K = 1. 15/12, S M = 19 m-cm
2
, s = 845 mm, 1 = 2. 86m a nd
p = 0. 03 N/mm
2
(s ee ex a mple 1 fo r ex pla na t i o n o f t h e s ymbo ls ) . Th e
res ult i ng pri ma ry membra ne s t res s , P
m
, a nd bendi ng s t res s , P
b
, were
170 N/mm
2
a nd 7. 5 N/mm
2
res pect i vely. S i nce t h e t a nker h a s been i n
s ervi ce fo r o ver 20 yea rs , t h e effect o f res i dua l weld s t res s wa s
a s s umed t o be neg li g i ble. F ra ct ure a s s es s ment i n t h i s ex a mple wa s
ca rri ed o ut a t level 2 o f P D 6493. At t h i s level, t h e a llo wa ble
cra ck leng t h wa s det ermi ned by t h e fo llo wi ng equa t i o n:
144
Eo
7*-M
a _ =
-0. 5
(5. 16)
*(Y r)
2
wh ere,
p i s t h e pla s t i ci t y co rrect i o n fa ct o r, wh i ch i s 0. 0 i n t h i s
ex a mple. (S i nce s eco nda ry (res i dua l) s t res s es a re a s s umed t o
be 0. )
F o r a n edg e cra ck under a x i a l lo a di ng a nd bendi ng , Ya i s t h e s um o f
t h e membra ne co mpo nent , (Ya )
m
a nd t h e bendi ng co mpo nent , (Ycr)
b
.
Th es e were ca lcula t ed us i ng t h e fo llo wi ng ex pres s i o ns :
M P
" t W. f
N
2 . Tta
t a n
7C C C 2
0. 752 + 2. 20a + 0. 37| 1 - s i n
(Yo )
b
=
HA
C OS Mt
a
>
2
i t a
i t a
2
0. 923 + 0. 199 (1 - S i n
(5. 17)
wh ere,
a = a /W; i s t h e ra t i o o f t h e cra ck leng t h t o t h e wi dt h o f t h e
s peci men a s s h o wn i n fi g ure 5-8.
M
km/
M
kb
a
^e t h e membra ne a nd bendi ng s t res s co ncent ra t i o n
fa ct o rs due t o lo ca l di s co nt i nui t y, t h es e were a s s umed t o be
1. 0 s i nce t h ey co uld no t be preci s ely det ermi ned.
In equa t i o n 5. 15, Ya & S
r
a re funct i o n o f cra ck leng t h , a .
Th erefo re a n i t era t i ve pro cedure wa s a do pt ed t o det ermi ne
ni
. F o r
t h e s a me ma t eri a l pro pert i es a s us ed i n ex a mple 1, t h e a llo wa ble
cra ck leng t h wa s fo und t o be 59 mm a nd t h e pla s t i c co lla ps e ra t i o
S
r
wa s 0. 71. '
145
w-
mtm
a=a/Q
Y (a )
2
na
tan
2
0.752 + 2.20a + 0.37 ( 1
*
sin
na
3
cos
71
2
EDGE CRACK UNDER AXIAL LOAD
+ P,
+ p.
^3
w
^t -
p
b
a = a/Q
Y (a ) =
ixa
0.923 + 0.199 ( 1
tan
na
2
- 4
rr a
2
4
c o s
7 ia
EDGE CRACK UNDER BENDING
EIGURE 5.8 'Y' EOR COMPONENT WITH EDGE CRACK
UNDER AXIAL TENSION AND BENDING
146
C ra ck Gro wt h Ana lys i s ;
In t h i s ex a mple, cra ck g ro wt h a na lys i s wa s perfo rmed us i ng t h e s a me
pro cedure a s o f ex a mple 1. As s umi ng t h e s a me s h o rt t erm No rt h S ea
S pect rum a s ex a mple 1, a s t res s s pect rum wa s g enera t ed. U s i ng t h e
rms va lue o f t h e s t res s ra ng es i n t h i s s pect rum, a cra ck g ro wt h
pro fi le o ver t i me wa s develo ped i n a fa s h i o n s i mi la r t o ex a mple 1.
As ca n be s een fro m fi g ure 5-9, i t t a kes a bo ut 4 vo ya g es fo r a n
i ni t i a l cra ck o f leng t h 0. 005 mm t o beco me cri t i ca l. P ert i nent
da t a i s g i ven belo w:
AP
m
=
187. 4 N/mm
2
AP
b
= 15.0 N/mm
2
"r max ~
= 202 .4 N/mm
2
ACT
ms
= 40 . 1 N/mm
2
Conclusions:
In ex a mple 2 t h e effect o f t h e fluct ua t i ng pres s ure lo a d due t o t h e
s ea wa y, i s les s pro no unced. Th i s i s evi dent fro m t h e va lues o f f*
r2
wh i ch a re, 54 N/mm
2
i n ex a mple 1 a nd 15 N/mm
2
i n ex a mple 2. But t h e
t o t a l t ens i le s t res s a ct i ng a t t h e bo t t o m s h ell pla t i ng , i n ex a mple
2, i s much h i g h er t h a n t h a t a ct i ng a t t h e s i de s h ell lo ng i t udi na l
No . 8, i n ex a mple 1. Th erefo re i n ex a mple 2, t h e cri t i ca l cra ck
leng t h i s s ma ller, a nd a ls o t h e ra t e o f cra ck pro pa g a t i o n i s
fa s t er.
5. 3 Glo ba l Level Ana lys i s
To a s s es s t h e co ns equences o f t h e lo ca l da ma g e a na lyzed i n t h e
previ o us s ub-s ect i o ns o n t h e o vera ll s t reng t h o f t h e t a nker, U LTS TR
[8] wa s us ed t o s t udy t h e ult i ma t e duct i le s t reng t h i n t h e i nt a ct
a nd da ma g ed co ndi t i o n. Alt h o ug h U LTS TR wa s develo ped a t D TRC by
t h e U S Na vy fo r eva lua t i o n o f t h e ult i ma t e s t reng t h o f na va l s h i ps
i t ca n be us ed fo r a ny s h i p s t ruct ure i ncludi ng t a nkers . It i s
ba s ed o n t h e "co mpo nent a ppro a ch " a s ment i o ned ea rli er. S o me o f
t h e li mi t a t i o ns o f U LTS TR a re:
(1) no co ns i dera t i o n fo r i nt era ct i o n o f fa i lure mo des
(2) no co ns i dera t i o n fo r fa i lure by t ri ppi ng i n t h e ela s t o -
pla s t i c ra ng e
(3) no t rea t ment o f 3-di mens i o na l effect s ; -
Nevert h eles s fo r t h e purpo s e o f co mpa ra t i ve s t udi es , i t i s
rela t i vely ea s y t o us e a nd les s t i me co ns umi ng t h a n a full 3-
di mens i o na l fi ni t e element a na lys i s .
147
70
60
50
HS
P 40
W 30
S 20
U
10
500 1000 1500
Thousands
No. OF CYCLES, N
2000 2500
70
60
50
CO
X 40
H
W 30
U
<- 20
U
10
\
'
/
/
/
^^
12 3 4
NUMBER OF V OYAGES
FIGURE 5.9 CRACK GROWTH PROFILE, EXAMPLE 2
148
To a s s es s t h e res i dua l s t reng t h o f t h e da ma g ed t a nker, fi rs t i t s
res erve s t reng t h wh i ch i s t h e ra t i o o f t h e ult i ma t e mo ment ca pa ci t y
(i nt a ct ) t o t h e a llo wa ble des i g n bendi ng mo ment wa s det ermi ned.
Th e des i g n bendi ng mo ment , M
D cs
, wa s ca lcula t ed fro m ABS [9] rules .
U LTS TR runs o f t h e i nt a ct t a nker were perfo rmed i n h o g g i ng a nd
s a g g i ng co ndi t i o ns , t o det ermi ne i t s ult i ma t e ca pa ci t y. S i nce t h e
t a nker i s s ymmet ri c a bo ut t h e cent erpla ne, o nly h a lf t h e mi ds h i p
s ect i o n wa s ex pli ci t ly mo deled wh i le t h e o t h er h a lf wa s mo deled
i mpli ci t ly us i ng t h e "MIRROR" o pt i o n o f U LTS TR. Th e wh o le mi ds h i p
wa s di s cret i zed i nt o 105 g ro s s -pa nels a nd 25 h a rd-co rner element s
a s s h o wn s ch ema t i ca lly i n fi g ure 5-10. Wh i le vert i ca l s t ruct ures ,
li ke t h e s i de s h ell pla t i ng a nd t h e lo ng i t udi na l bulkh ea d, were
fi nely di s cret i zed beca us e o f t h e li nea r va ri a t i o n o f t h e no rma l
s t res s a cro s s t h e dept h , t ra ns vers e s t ruct ures li ke t h e ma i n deck
a nd bo t t o m pla t i ng , were di vi ded i nt o jus t t wo g ro s s pa nels , due t o
t h e uni fo rm di s t ri but i o n o f t h e no rma l s t res s a cro s s t h em. By
s peci fyi ng t h e t ra ns vers e fra me s pa ci ng a s t h e co lumn-buckli ng a nd
t ri ppi ng leng t h o f t h e s t i ffeners ma ki ng up t h e g ro s s -pa nels ,
effect s o f t h e t ra ns vers e s t ruct ures li ke bulkh ea ds a nd fra mes were
i nco rpo ra t ed. Res ult s o f t h es e a na lys es a re pres ent ed belo w:
D es i g n Bendi ng Mo ment 5. 600 x 10
9
N*m
Int a ct U lt i ma t e Bendi ng Mo ment 6. 868 x 10
9
N*m
(Ho g g i ng )
Res erve C a pa ci t y, REF 1. 2 3
(Ho g g i ng )
Int a ct U lt i ma t e Bendi ng Mo ment 6. 744 x 10
9
N m
(S a g g i ng )
Res erve C a pa ci t y, REF 1. 2 0
(S a g g i ng )
A plo t o f t h e bendi ng mo ment vers us t h e a ppli ed curva t ure, fi g ure
5-11 s h o ws a dra s t i c fa ll i n t h e mo ment ca pa ci t y o nce t h e ult i ma t e
mo ment i s rea ch ed. Th us i mplyi ng very li t t le ca pa bi li t y fo r lo a d
redi s t ri but i o n. Th e pri ma ry mo de o f fa i lure wh i ch lea ds t o t h e
o ns et o f ult i ma t e co lla ps e wa s fo und t o be s t i ffener t ri ppi ng a t
t h e upper deck a nd bo t t o m s h ell pla t i ng .
149
49 -1
47
i
45 -i
43 -1
41 >
39 .
37 ->
35
1
33 -.
31
-1
29
.
27
-i
25 .
23 ,
21 .
19 -.
17
i
15
-.
13
.
1 1
9
7
FIGURE 5.10 DISCRETIZATION OF TANKER CROSS-SECTION INTO
GROSS PANELS ' D ' AND HARD CORNERS A
150
10
E
E
<N
O
00 -
X
O
-10
.[6.87 X
o
<
J fS-6.74 X 10
1 ?
-4 -3
,12
10 1 2 3
ROTATION (RADIAN)
X 10~
6
FIGURE 5.1 1
ULTIMATE MOMENT CAPACITY OF THE
85,000 TON TANKER
151
Th e lo ca l da ma g es t o t h e s t ruct ura l co mpo nent s a na lyzed i n s ect i o n
5. 2 were fa t i g ue cra cks . Mo s t o f t h e t i me t h es e cra cks pro pa g a t e
i nt o reg i o ns o f lo w s t res s i nt ens i t i es a nd h ence t h ei r g ro wt h g et s
ret a rded. To preci s ely det ermi ne t h e ex t ent o f t h e cra cks , s t res s
fi elds a ct i ng i n t h e a rea a djo i ni ng t h e cra cks need t o be
det ermi ned. Th i s requi res a full 3-di mens i o na l a na lys i s . To
i nt eg ra t e t h e a ffect o f t h es e lo ca l da ma g es i nt o t h e g lo ba l
res po ns e, wi t h o ut res o rt i ng t o a full s ca le 3-di mens i o na l fi ni t e
element a na lys i s , a co ns erva t i ve a ppro a ch wa s a do pt ed. In t h i s
a ppro a ch , po rt i o ns o f t h e t a nker's cro s s -s ect i o n i n t h e vi ci ni t y o f
t h e da ma g e were co ns i dered i neffect i ve a nd h ence neg lect ed i n t h e
U LTS TR mo del. In a ddi t i o n t o t h e t wo da ma g e lo ca t i o ns , a na lyzed i n
s ect i o n 5. 2, ei g h t o t h er cri t i ca l lo ca t i o ns were i dent i fi ed ba s ed
o n s ervi ce li fe ex peri ence a nd a cco rdi ng ly t h e t a nker wa s di vi ded
i nt o 10 zo nes , A-J a s s h o wn i n fi g ure 5-12. Ten U LTS TR runs were
perfo rmed a nd t h e res ult s a re pres ent ed i n t a ble 5-5. In t h es e t en
ca s es , t h e fi rs t 8 repres ent ed da ma g e i n t h e fo rm o f cra cks a nd
were mo deled by co mplet e remo va l o f t h e a ffect ed g ro s s pa nels fro m
t h e po rt s i de o f t h e cro s s -s ect i o n. In ca s es 9 a nd 10, perma nent
defo rma t i o n i n zo ne C wa s s i mula t ed by a s s i g ni ng t h e g ro s s pa nels
i n t h i s zo ne i ni t i a l di s t o rt i o n o f 254 a nd 508 mm res pect i vely
wh i ch a re a bo ut 8 a nd 16 percent o f t h e t ra ns vers e fra me s pa ci ng .
F ro m t h e t a ble, i t ca n be s een (ca s e 1-4) t h a t da ma g es i n po rt i o ns
o f t h e ma i n deck o r bo t t o m s h ell pla t i ng res ult i n rela t i vely
s i g ni fi ca nt lo s s o f ult i ma t e s t reng t h i n co mpa ri s o n t o da ma g es i n
o t h er zo nes (ca s es 5-7) o f t h e t a nkers . F o r co mpa ri s o n, i n ca s e 8,
a ll g ro s s pa nels i n zo nes G, H, I & J were eli mi na t ed res ult i ng i n
a bo ut 6% reduct i o n i n cro s s -s ect i o na l a rea - a lmo s t t h e s a me a s
t h a t due t o eli mi na t i o n o f zo ne C (ca s e 3) , but t h e res i dua l
s t reng t h (RIF ) i n h o g g i ng decrea s ed o nly by 6% wh erea s i n ca s e 3 i t
decrea s ed by 14%.
Th e pro duct o f t h e res i dua l s t reng t h i ndex (RIF ) a nd t h e res erve
ca pa ci t y (REF ), i s a g o o d i ndi ca t o r fo r t h e purpo s e o f res i dua l
s t reng t h a s s es s ment . Th e va lue o f t h i s pro duct , wh i ch ca n be
co ns i dered a s a s a fet y fa ct o r, s h o uld never be les s t h a n o ne.
Therefore,
FS = RIF x REF =
MykODamaged) M^CIntact)
M^flntact) M
Des
> 1. 0 (5. 18)
F o r ea ch o f t h e t en ca s es i n t a ble 5-5, t h es e fa ct o rs o f s a fet y
were co mput ed a nd i t wa s fo und t h a t i n ca s es 1 a nd 4, t h ey were
les s t h a n o ne. Th us i mplyi ng t h a t o pera t i o n o f t h e ves s el wi t h
da ma g e i n t h es e zo nes co uld a dvers ely a ffect t h e o vera ll s t reng t h
o f t h e ves s el.
152
er
<
o
co
z
o
I
<
o
o
<
er
o
C\l
L O
L J
et:
o
0
153
TABLE 5-5 RES ID U AL S TRENGTH F AC TORS OF D AMAGED TANKER
C a s e D a ma g ed Area RIF * F S ** RIF * F S "
No . Zo ne (% ch a ng e) (S a g ) (S a g ) (Ho g ) (Ho g )
1 A -8 0. 79 0. 95 0. 93 1. 14
2 B -6 0. 85 1. 02 0. 96 1. 18
3 C -10 0. 95 1. 14 0. 78 0. 96
4 D -6 0. 98 1. 18 0. 86 1. 06
5 E & F -5 0. 93 1. 12 1. 00 1. 23
6 G & I -3 1. 00 1. 20 0. 97 1. 19
7 H & J -3 1. 00 1. 20 0. 99 1. 22
8 G, H, I & J -6 1. 00 1. 20 0. 94 1. 16
9 P erma nent
S et i n C =
254 mm
1. 00 1. 20 0. 90 1. 11
10 P erma nent
S et i n C =
508 mm
1. 00 1. 20 0. 88 1. 08
* RIF =
M
ULT
( Damaged)
M
ULT
( Intact)
** FS = RIF x REF, where
M
TTTT
( Intact)
REF =
-njLT
MULT ( Des)
REF ( Sag) = 1.2 0
REF ( Hog) = 1.23
154
5. 4 References
1. Hug h es , O. a nd F ra nkli n, P . , "D efi ni t i o n a nd Va li da t i o n o f a
P ra ct i ca l Ra t i o na lly-Ba s ed Met h o d fo r t h e F a t i g ue Ana lys i s a nd
D es i g n o f S h i p Hulls ", S NAME, T&R Repo rt No . R-41, F ebrua ry,
1993.
2. "Gui da nce o n t h e Met h o ds fo r As s es s i ng t h e Accept a bi li t y o f F la ws
i n F us i o n Welded S t ruct ures , P D 6493: 1991", P ubli s h ed by t h e
Bri t i s h S t a nda rds Ins t i t ut i o n (BS I).
3. ABS "Gui de fo r t h e F a t i g ue S t reng t h As s es s ment o f Ta nkers ", J une
1992.
4. "S h i p D es i g n Ma nua l - Ma ri ne S t ruct ures ", C h a pt er 6, P a rt 8,
P ubli s h ed by Bri t i s h S h i p Res ea rch As s o ci a t i o n (BS RA).
5. S h i , W. B. a nd Th o mps o n, P . A. , "As pect s o f Vulnera bi li t y o f Bulk
C a rri er S t ruct ures ", RINA, Int erna t i o na l C o nference o n Bulk
C a rri ers a nd Ta nkers - Th e Wa y Ah ea d, D ecember, 1992.
6. Ro lfe, S . T. a nd Ba rs o m, J . M. , "F ra ct ure a nd F a t i g ue C o nt ro l i n
S t ruct ures - Appli ca t i o ns o f F ra ct ure Mech a ni cs ", P ubli s h ed by
P rent i ce - Ha ll, Inc. 1977, New J ers ey.
7. C h a ng , J . B. a nd Huds o n, C M. , edi t o rs , "Met h o ds a nd Mo dels fo r
P redi ct i ng F a t i g ue C ra ck Gro wt h U nder Ra ndo m Lo a di ng ", AS TM S TP
748, 1981.
8. Ada mch a k, J . C . , "U LTS TR: A P ro g ra m fo r Es t i ma t i ng t h e C o lla ps e
Mo ment o f a S h i p's Hull U nder Lo ng i t udi na l Bendi ng ", Repo rt No .
D TNS RD C - 82/076, Oct o ber, 1982.
9. ABS Rules F o r C la s s i ng a nd Bui ldi ng S t eel Ves s els , 1992.
155
6 . 0 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK
Th e pri ma ry o bject i ves o f t h i s pro ject were,
Int ro duce t h e s ubject o f res i dua l s t reng t h a s s es s ment o f
da ma g ed ma ri ne s t ruct ures t o t h e pra ct i ci ng s h i p s t ruct ura l
eng i neeri ng co mmuni t y.
S umma ri ze t h e s t a t e o f t h e a rt t ech no lo g y a nd met h o ds
a va i la ble i n t h e ma ri ne a nd no n-ma ri ne i ndus t ry fo r
qua nt i fyi ng res i dua l s t reng t h .
Reco mmend fut ure wo rk t h a t wi ll furt h er i nt eg ra t e ex i s t i ng
eng i neeri ng pro cedures i n t h e a rea s o f cra ck g ro wt h ,
perma nent defo rma t i o n a nd ult i ma t e s t reng t h emph a s i zi ng co s t
effect i ve met h o ds .
Ba ckg ro und
Res i dua l s t reng t h ca n be defi ned a s t h e rema i ni ng s t reng t h i n a
s t ruct ure o nce a co mpo nent h a s fa i led o r h a s been s o s everely da ma g ed
a s t o be i neffect i ve. Th us res i dua l s t reng t h ca n be vi ewed a s t h e
ult i ma t e ca pa ci t y fo r a da ma g ed s t ruct ure a nd t h erefo re i s a n
i mpo rt a nt i ndi ca t o r o f t h e s t ruct ures da ma g e t o lera nce.
Th e current des i g n pro ces s a s s umes t h a t , t h e a ppli ca t i o n o f s ma ll
deflect i o n li nea r ela s t i c t h eo ry pro vi des t h e a ppro pri a t e mea s ure o f
s t ruct ura l perfo rma nce a t t h e g lo ba l res po ns e level, fo r mo s t ma ri ne
s t ruct ures . F ro m s uch a g lo ba l a s s es s ment o f t h e s t ruct ure's
res po ns e, t h e fo rces a ppli ed t o i ndi vi dua l members a re t h en co mpa red
wi t h t h e lo ca l member ca pa bi li t y a s s es s ed i n t erms o f ult i ma t e
s t reng t h . Appro a ch es t o t h e det ermi na t i o n o f lo ca l member ult i ma t e
s t reng t h ma y i nclude no n-li nea r, la rg e deflect i o n o r pla s t i c t ype
a na lys i s .
It i s i mpo rt a nt t o no t e t h a t t h e res po ns e a nd ca pa bi li t y o f a ma ri ne
s t ruct ure fo llo wi ng s o me fo rm o f da ma g e i s eva lua t ed by i dent i fyi ng
t h e effect o f lo ca l member fa i lure o n t h e o vera ll s t ruct ura l
i nt eg ri t y. Th e effect s o f lo ca l da ma g e mus t be i nt eg ra t ed upwa rd t o
t h e g lo ba l s t ruct ura l level t o i ns ure t h a t co rrect lo a d t ra ns fer pa t h s
a re det ermi ned t o rea li s t i ca lly s i mula t e t h e ex t ent o f da ma g e o n t h e
o vera ll s ys t em beh a vi o r. Th i s lo ca l-g lo ba l res po ns e a ppro a ch requi red
fo r a s s es s i ng t h e res i dua l s t reng t h o f da ma g ed ma ri ne s t ruct ures
repres ent s a revers e a ppro a ch fro m t h e g lo ba l-lo ca l res po ns e a ppro a ch
us ed i n t h e current s t ruct ura l des i g n pro ces s . Th i s wa s t h e
underlyi ng pri nci ple o f t h i s i nves t i g a t i o n.
156
C o nclus i o ns
Th e fo llo wi ng co nclus i o ns , ba s ed o n t h e res ult s o f t h i s pi lo t pro ject
a re pres ent ed belo w:
1. Th e t h eo ri es , met h o ds a nd co mput a t i o na l t o o ls fo r a s s es s i ng t h e
res i dua l s t reng t h o f da ma g ed ma ri ne s t ruct ures due t o no rma l o pera t i ng
lo a ds ex i s t but t h ey a re s ca t t ered i n li t era t ure a nd no s t a nda rdi zed
pro cedure fo r t h ei r a ppli ca t i o n h a s been es t a bli s h ed.
2. As s es s ment o f res i dua l s t reng t h o f co nt i nuo us s t ruct ures li ke
s h i ps i s very co mplex a nd ex pens i ve. Two met h o ds o f a na lys i s a re
po s s i ble: a n a ppro x i ma t e 2-di mens i o na l pro cedure (t h e i ndi rect
met h o d) o r a 3-di mens i o na l no n-li nea r fi ni t e element a na lys i s (t h e
di rect met h o d). Wh i le t h e i ndi rect met h o d i s eco no mi ca l a nd g o o d fo r
qui ck a s s es s ment o f da ma g e co ns equences , i t predi ct s very co ns erva t i ve
res ult s beca us e o f t h e numero us a s s umpt i o ns o n t h e s t res s fi elds ,
bo unda ry co ndi t i o ns a nd ex t ent o f t h e da ma g e mo deled. Th e di rect
met h o d i s mo re funda ment a l i n na t ure wi t h very few a s s umpt i o ns a nd
o verco mes mo s t o f t h e li mi t a t i o ns o f t h e i ndi rect met h o d. But i t i s
no t a very pra ct i ca l met h o d. It i s a n ex pens i ve met h o d wh i ch requi res
a h i g h level o f ex pert i s e a nd co ns i dera ble a mo unt o f co mput a t i o na l
res o urces .
3. Th e a ccura cy o f t h e ult i ma t e s t reng t h a na lys i s pro g ra ms us ed i n
ei t h er met h o ds i s h a rd t o predi ct s i nce li mi t ed a mo unt o f full s ca le
t es t da t a a re a va i la ble fo r co mpa ri s o n. F ew co mpa ri s o ns o f t h e
predi ct i o ns o f t h es e pro g ra ms wi t h ex peri ment a l da t a i ndi ca t e t h a t
t h es e pro g ra ms ca n a t t h e bes t be us ed fo r co mpa ra t i ve s t udi es .
4. D a t a co llect i o n a nd s ubs equent a na lys i s o f da ma g e reco rds fo r
bo t h s h i p a nd o ffs h o re s t ruct ures i n t h e co nt ex t o f t o t a l fleet yea rs
o f s ervi ce, da ma g e t o lera nt fea t ures a nd t h e t i me rela t i o ns h i p bet ween
o ccurrence o f da ma g e a nd s ubs equent det ect i o n a nd repa i r i s di ffi cult .
Mea ni ng ful co nclus i o ns wi ll requi re a lo t o f da t a , t h o ro ug h ly
pro ces s ed, a nd wi ll be ex pens i ve t o co llect .
5. C ri t i ca l lo ca t i o ns o f da ma g e a re a funct i o n o f s h i p t ype a nd
s t ruct ura l co nfi g ura t i o n.
6. Th e do mi na nt fo rm o f da ma g e i n ma ri ne s t ruct ures due t o no rma l
o pera t i ng lo a ds i s fra ct ure a nd ex ces s i ve perma nent defo rma t i o n. Mo s t
o f t h e t i me t h es e fra ct ures a re i n t h e fo rm o f fa t i g ue cra cki ng o n
s eco nda ry members a nd a re repa i red befo re t h ey rea ch t h ei r cri t i ca l
leng t h o r pro pa g a t e i nt o reg i o ns o f lo w s t res s . Th e level o f i nh erent
duct i li t y o f s h i p s t ruct ures , ma ke t h em very t o lera nt t o ex ces s i ve
perma nent defo rma t i o n.
157
7. Th e effect o f lo ca l da ma g e, s uch a s perma nent defo rma t i o n, o n t h e
g lo ba l res po ns e o f t h e s t ruct ure ca n be mo deled by a va i la ble ult i ma t e
s t reng t h pro g ra ms . Mo deli ng t h e ex t ent o f fra ct ure da ma g e i s
s ubject i ve a nd requi res a n unders t a ndi ng o f t h e s t ruct ura l
co nfi g ura t i o n a nd s t res s fi elds . Th e es t i ma t e o f t h e i mpa ct o f cra ck
li ke da ma g es , o n t h e g lo ba l res po ns e, ca n be a cco mpli s h ed by us i ng
a ero s pa ce i ndus t ry develo ped t ech no lo g y a nd mo di fyi ng i t fo r ma ri ne
s t ruct ura l a ppli ca t i o ns .
8. S t ruct ura l members t h a t a re no t very h i g h ly s t res s ed i n i nt a ct
co ndi t i o ns pla y a n i mpo rt a nt ro le i n redi s t ri but i ng fo rces wh en da ma g e
t a kes pla ce. C o ns equent ly t h es e members a re mo re cri t i ca l t o t h e
da ma g ed s t ruct ure beh a vi o r t h a n t h e i nt a ct s t ruct ure beh a vi o r.
9. U lt i ma t e s t reng t h a nd t h e ma nner o f h ull g i rder co lla ps e a re
s t ro ng ly a ffect ed by h a rd co rners i n t h e cro s s -s ect i o n. In t erms o f
ult i ma t e s t reng t h , h a rd co rners s h o uld be i nco rpo ra t ed i nt o t h e des i g n
by judi ci o us us e o f s t ruct ura l bulkh ea ds a nd s eco nda ry s t ruct ura l
co mpo nent s .
10. Alt h o ug h t h i s repo rt a ddres s es res i dua l s t reng t h a s s es s ment o f
s t ruct ures da ma g ed due t o no rma l o pera t i ng lo a ds , i t i s i mpo rt a nt t o
no t e t h a t t h e ult i ma t e s t reng t h predi ct i o n met h o ds revi ewed i n t h i s
repo rt a re a ppli ca ble t o a s s es s da ma g e ca us ed by ra ndo m a cci dent s s uch
a s co lli s i o ns a nd g ro undi ng s . Th e o nly di fference i s t h a t s i nce t h e
a s s es s ment i s do ne a ft er t h e a ct ua l a cci dent , t h e ex a ct lo ca t i o n a nd
t h e ex t ent o f t h e da ma g e i s bet t er defi ned.
Reco mmenda t i o ns fo r F ut ure Wo rk
Ba s ed o n t h e a bo ve co nclus i o ns , t h e fo llo wi ng reco mmenda t i o ns a re
pres ent ed fo r fut ure wo rk:
1. Res ea rch i nt o develo pment a nd va li da t i o n o f a n i nt eg ra t ed
co mput er pro g ra m fo r res i dua l s t reng t h a s s es s ment o f da ma g ed 'ma ri ne
s t ruct ures . S peci fi ca lly t h e pro g ra m s h o uld be ca pa ble o f:
full 3-di mens i o na l g eo met ri c mo deli ng o f t h e s h i p s t ruct ure
s t res s -a na lys i s , wi t h pro vi s i o ns fo r mo deli ng ma t eri a l a nd
g eo met ri c no n-li nea ri t i es a nd cra ck li ke defect s
co mput i ng s pect ra l wa ve lo a di ng g i ven t h e s ea s t a t e, h ea di ng
a nd dura t i o n
cra ck g ro wt h predi ct i o ns
ult i ma t e s t reng t h predi ct i o ns
158
2. In o rder t o a cco unt fo r t h e ro le o f res i dua l s t reng t h i n t h e
develo pment o f li mi t s t a t e des i g n cri t eri a , cri t i ca l lo ca t i o ns
s us cept i ble t o da ma g e fo r s peci fi c s h i p t ypes (i . e. bulk ca rri ers ,
t a nkers a nd co nt a i ners h i ps ) s h o uld be i dent i fi ed by:
(a ) s t res s a na lys i s us i ng fi ni t e element pro cedures a nd
(b) ex t ens i ve res ea rch o f a va i la ble ca s ua lt y reco rds .
Th e res ult s o bt a i ned fro m 2(a ) a nd 2(b) s h o uld be co rrela t ed i n t erms
o f t h e fo llo wi ng t h ree ca t eg o ri es :
(i ) s lo w cra ck g ro wt h s t ruct ure - des i g ns wh ere s t res s
levels a re li mi t ed t o a s s ure t h a t cra cks wi ll no t g ro w
t o cri t i ca l s i zes duri ng s peci fi ed peri o ds wh i ch depend
o n t h e deg ree o f i ns pect a bi li t y.
(i i ) cra ck a rres t fa i l s a fe s t ruct ures - des i g ns t h a t s t o p
uns t a ble, ra pi d pro pa g a t i o n o f cra cks wi t h i n a
co nt i nuo us a rea o f s t ruct ure a nd s ubs equent g ro wt h i s
s lo w eno ug h t o permi t det ect i o n pri o r t o co mplet e
fa i lure.
(i i i ) mult i ple lo a d pa t h fa i l s a fe s t ruct ure - s t ruct ure
des i g ned i n s eg ment s s uch t h a t lo ca li zed da ma g e i s
co nt a i ned wi t h i n o ne o r t wo s eg ment s a nd t h e rema i ni ng
s t ruct ure ex h i bi t s s lo w cra ck g ro wt h a nd pro vi des
s uffi ci ent s t reng t h unt i l s ubs equent i ns pect i o n.
3. F o r s peci fi c s h i p t ypes , wi t h do cument ed des i g n i nfo rma t i o n,
ca lcula t e t h e res erve a nd res i dua l s t reng t h us i ng bo t h t h e
2-di mens i o na l co mpo nent met h o d a nd t h e 3-di mens i o na l no n-li nea r fi ni t e
element met h o d. C o mpa re t h e res ult s o f bo t h met h o ds a nd ex a mi ne t h e
s ens i t i vi t y o f t h e ult i ma t e s t reng t h predi ct i o n t o t h e fo llo wi ng :
da ma g e mo deli ng a s s umpt i o ns
a mpli t ude a nd s h a pe o f perma nent defo rma t i o ns
lo ca t i o ns o f fra ct ure da ma g e
lo s s o f t ra ns vers e s t ruct ure
lo s s o f h ull g i rder s h ea r a rea
lo ca t i o n o f h a rd co rners vs . fra ct ure i ni t i a t i o n.
4. C o mpa re t h e res ult s o f t h es e s t udi es t o full s ca le t es t s
s i mula t i ng da ma g ed s h i p s t ruct ure.
159
5. F o r qui ck a s s es s ment o f cra ck li ke defect s , develo p ch a rt s o f
fra ct ure t o ug h nes s da t a t h ro ug h a ct ua l t es t i ng . Th es e t es t i ng s s h o uld
be co nduct ed fo r va ri o us t ypi ca l s t ruct ura l co nfi g ura t i o ns s us cept i ble
t o cra cki ng .
6. Bes i des develo pment o f a na lyt i ca l pro cedures , res ea rch h a s t o be
fo cus s ed o n t h e develo pment a nd i ns t a lla t i o n o f a fully i nt eg ra t ed
h ull s urvei lla nce s ys t em. Th e res ea rch s h o uld s peci fi ca lly a ddres s
s uch i s s ues a s :
(i ) rea l-t i me mo ni t o ri ng o f h ull s t res s es , a ccelera t i o ns a nd
fa t i g ue.
(i i ) s elect i o n o f mea s uri ng i ns t rument s t o wi t h s t a nd t h e
h a rs h ma ri ne envi ro nment .
(i i i ) s t ra t eg i c po s i t i o ni ng o f t h es e i ns t rument s o n t h e s h i p
t o mea s ure t h e des i red res po ns e.
(i v) pro vi s i o n fo r di rect li nk bet ween t h e s t res s mo ni t o ri ng
s ys t em a nd t h e s h i p's lo a di ng co mput ers .
Th e g o a ls o f t h e s urvei lla nce s ys t em s h o uld be:
t o co mpa re i ns t a nt a neo us mea s ured s t res s es t o t h e ma x i mum
permi s s i ble s t res s fo r t h e ves s el a nd pro vi de a feedba ck.
mo ni t o r s t res s revers a ls fo r eva lua t i o n o f cumula t i ve fa t i g ue
da ma g e.
t o s t o re t h e da t a a cqui red duri ng vo ya g es t o be us ed la t er o n-
s h o re t o h elp pla n a nd fo reca s t lo ng -t erm repa i r a nd ma i nt ena nce
s t ra t eg i es .
160
Project Technical Committee Members
The following persons were members of the committee that represented the Ship Structure
Committee to the Contractor as resident subject matter experts. As such they performed
technical reviewof the initial proposals to select the contractor, advised the contractor in
cognizant matters pertaining to the contract of which the agencies were aware, and performed
technical reviewof the work in progress and edited the final report.
Mr. Jack Spencer
LCDR Doug O'Reilly
Mr. Neil Pegg
Mr. Francis J. Scotto
Mr. James White
Mr. Jerry Snyder
Mr. Phil Alman
Ms. Maria Ximenes
Mr. William Siekierka
Dr. Robert Sielski
Mr. Alex Stavovy
CDR Steve Sharpe
American Bureau of Shipping
Directorate Ship Engineering, Canada
Defence Research Establishment Atlantic, Canada
American Bureau of Shipping
U.S. Coast Guard Research and Development
Naval Sea Systems Command
U.S. Coast Guard
Chevron Shipping
Naval Sea Systems Command,
Contracting Officer's
Technical Representative
National Academy of Science,
Marine Board Liaison
U.S. Coast Guard, Executive Director
Ship Structure Committee
COMMITTE E ON MARINE STRUCTURE S
Commission on E ngineering and Technical Systems
National Academy of Sciences - National Research Council
The COMMITTE E ON MARINE STRUCTURE S has technical cognizance over the
interagency Ship Structure Committee's research program.
Peter M. Palermo Chairman, Alexandria, VA
Subrata K . Chakrabarti, Chicago Bridge and Iron, Plainfield, IL
John L andes, University of Tennessee, K noxville, TN
Bruce G. Collipp, Marine E ngineering Consultant, Houston, TX
Robert G. K line, Marine E ngineering Consultant, Winona, MN
Robert G. L oew y, NAE , Rensselaer Polytechnic Institute, Troy, NY
Robert Sielski, National Research Council, Washington, DC
Stephen E . Sharpe, Ship Structure Committee, Washington, DC
L OADS WORK GROUP
Subrata K . Chakrabarti Chairman, Chicago Bridge and Iron Company, Plainfield, IL
How ard M. Bunch, University of Michigan, Ann Arbor, Ml
Peter A. Gale, John J. McMullen Associates, Arlington, VA
Hsien Yun Jan, Martech Incorporated, Neshanic Station, NJ
John Niedzw ecki, Texas A&M University, College Station, TX
Solomon C. S. Yim, Oregon State University, Corvallis, OR
Maria Celia Ximenes, Chevron Shipping Co., San Francisco, CA
MATE RIAL S WORK GROUP
John L andes, Chairman, University of Tennessee, K noxville, TN
William H Hartt, Florida Atlantic University, Boca Raton, FL
Horold S. Reemsnyder, Bethlehem Steel Corp., Bethlehem, PA
Barbara A. Shaw , Pennsylvania State University, University Park, PA
James M. Saw hill, Jr., New port New s Shipbuilding, New port New s, VA
Bruce R. Somers, L ehigh University, Bethlehem, PA
Jerry G. Williams, Conoco, Inc., Ponca City, OK
RE CE NT SHIP STRUCTURE COMMITTE E PUBL ICATIONS
SSC-363 Uncertainties in Stress Analysis on Marine Structures by E . Nikolaidis
and P. K aplan 1991
SSC-364 Inelastic Deformation of Plate Panels by E ric Jennings, K im Grubbs,
Charles Zanis, and L ouis Raymond 1991
SSC-365 Marine Structural Integrity Programs (MSIP) by Robert G. Bea 1992
SSC-366 Threshold Corrosion Fatigue of Welded Shipbuilding Steels by G. H.
Reynolds and J. A. Todd 1992
SSC-367 Fatigue Technology Assessment and Strategies for Fatigue Avoidance
in Marine Structures by C. C. Capanoglu 1993
SSC-368 Probability Based Ship Design Procedures: A Demonstration
by A. Mansour, M. L in, L . Hovem, A. Thayamballi 1993
SSC-369 Reduction of S-N Curves for Ship Structural Details by K . Stambaugh,
D. L esson, F. L aw rence, C-Y. Hou, and G. Banas 1993
SSC-370 Underw ater Repair Procedures for Ship Hulls (Fatigue and Ductility of
Underw ater Wet Welds) by K . Grubbs and C. Zanis 1993
SSC-371 E stablishment of a Uniform Format for Data Reporting of Structural
Material Properties for Reliability Analysis by N. Pussegoda, L Malik,
and A. Dinovitzer 1993
SSC-372 Maintenance of Marine Structures: A State of the Art Summary by
S. Hutchinson and R. Bea 1993
SSC-373 L oads and L oad Combinations by A. Mansour and A. Thayamballi 1994
SSC-374 E ffect of High Strength Steels on Strength Consdierations of Design and
Construction Details of Ships by R. Heyburn and D. Riker 1994
SSC-375 Uncertainty in Strength Models for Marine Structures by O. Hughes,
E . Nikolaidis, B. Ayyub, G. White, P. Hess 1994
SSC-377 Hull Structural Concepts For Improved Producibility by J. Daidola,
J. Parente, and W. Robinson 1994
SSC-378 The Role of Human E rror in Design. Construction and Reliability of
Marine Structures by R. Bea 1994
SSC-379 Improved Ship Hull Structural Details Relative to Fatigue
by K . Stambaugh, F. L aw rence and S. Dimitriakis 1994
SSC-380 Ship Structural Integrity Information System by R. Schulte-Strathaus,
B. Bea 1995
Ship Structure Committee Publications - A Special Bibliography