Fatigue Resistant Detail Design Guide For Ship Structures

NTIS # PB99-159931
SSC-405
FATIGUE RESISTANT DETAIL

DESIGN GUIDE FOR SHIP
STRUCTURES
This document has been approved

For public release and sale; its
Distribution is unlimited
SHIP STRUCTURE COMMITTEE

1999
Ship Structure Committee
RADM Robert C. North
U. S. Coast Guard Assistant Commandant,
Marine Safety and Environmental Protection
Chairman, Ship Structure Committee
Mr. Robert McCarthy Dr. Donald Liu
Director, Senior Vice President
Survivability and Structural Integrity Group American Bureau of Shipping
Naval Sea Systems Command
Mr. Marc Lasky Mr. Bud Streeter
Director, Office of Ship Construction Director General, Marine Safety,
Maritime Administration Safety & Security
Transport Canada
Mr. Thomas Connors Dr. Neil Pegg
Director of Engineering Group Leader - Structural Mechanics
Military Sealift Command Defence Research & Development Canada - Atlantic
CONTRACTING OFFICER TECHNICAL REP. EXECUTIVE DIRECTOR
Mr. Chao Lin / MARAD Lieutenant William A. Nabach
TBD / NAVSEA U. S. Coast Guard
Mr. Robert Sedat / USCG
SHIP STRUCTURE SUB-COMMITTEE

AMERICAN BUREAU OF SHIPPING DEFENCE RESEARCH & DEVELOPMENT ATLANTIC
Mr. Glenn Ashe Mr. Layton Gilroy
Mr. John F. Conlon Mr. John Porter
Mr. Phil Rynn LCDR Stephen Gibson
Mr. William Hanzalek Dr. David Stredulinsky
MARITIME ADMINISTRATION MILITARY SEALIFT COMMAND

Mr. Chao Lin Mr. Edward Meade
Mr. Rick A. Anderson
Mr. Jeffery E. Beach
Mr. Michael W. Touma
NAVAL SEA SYSTEMS COMMAND TRANSPORT CANADA
Mr. W. Thomas Packard Mr. Andre Taschereau
Mr. Edward E. Kadala Mr. Justus Benckhuysen
Mr. Allen H. Engle Mr. James Reid
Mr. Charles L. Null
UNITED STATES COAST GUARD
Captain Mark VanHaverbeke
Mr. Walt Lincoln
Mr. Walt Lincoln
CDR Ray Petow
Technical Report Documentation Form
1. Report No. 2. Government Accession No. 3. Recipient’s Catalogue No.
SR-1386
4. Title and Subtitle 5. Report Date
March 1999
Fatigue-Resistant Detail Design Guide for Ship Structures
6. Performing Organization Document No.
7. Author(s) 8. Performing Organization Report No.

I.F. Glen, A. Dinovitzer, R.B. Paterson, L. Luznik, C. Bayley 4714C.FR
9. Performing Agency Name and Address 10. Work Unit No. (TRAIS)
Fleet Technology Limited
311 Legget Drive 11. Contract or Grant No.
Kanata, Ontario (CANADA) K2K 1Z8 PO # 97-0046
12. Sponsoring Agency Name and Address 13. Type of Publication and Period Covered
Ship Structure Committee Final Report
c/o US Coast Guard
2100 Second Street, SW 14. Sponsoring Agency Code
Washington, DC 20593-0001
15. Supplementary Notes
16. Abstract
This Guide is a result of a 1995 Ship Structures Committee Symposium in Washington, at which it was recognized that
practicing naval architects required help in applying fatigue design research and development results. This Guide
consolidates the state-of-the-art in ship structural detail fatigue design.
The information in the Guide is presented in four sections, starting with an overview of the fatigue design problem and
related issues. The second section includes a catalogue of typical structural details for commercial and combatant ship
types, with suggested structural improvements for fatigue life extension. This, along with a brief discussion of
fabrication issues, represents the “Level 1” approach to ship structure fatigue design. The third section of the report
presents analytical procedures for fatigue based design of structural details. Load, stress and fatigue analysis
procedures are presented at two levels of complexity, “Level 2” and “Level 3”. The simplified Level 2 approach makes
use of Classification Society rules and analytical methods, whereas, the spectral Level 3 approach explicitly considers
vessel operational profiles and wave climates in statistical load estimation and makes use of FE models for stress
analysis. Both the Level 2 and 3 approaches are based on a “hot spot” stress approach using Miner’s rule in either a
discrete or continuous form. The final section of the Guide provides worked fatigue design examples to demonstrate
the concepts presented in the previous sections.
17. Key Words 18. Distribution Statement
fatigue, ship structures, structural details, design

19. Security Classif. (of this report) 20. SECURITY CLASSIF. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified
Form DOT F 1700.7 (8/72) Reproduction of form and completed page is authorized
Fatigue-Resistant Detail Design Guide for Ship Structures ii

ACKNOWLEDGEMENTS
The preparation of this Guide has been a mammoth task, relative to the resources available. At the
outset, the terms of the contract were that no new work was to be developed – the Guide was to be a
compilation of existing, published work. In the process of the project, it became clear that some new
ground was being broken in the practical application of the published work to real ship design problems.
The authors have drawn heavily on the published work of the Classification Societies, and in particular,
Lloyd’s Register of Shipping, the American Bureau of Shipping and Det Norske Veritas. FTL
and the SSC wish to express their great appreciation to these Societies for permission to use this
material.
The authors also wish to thank their panel of advisors: Dr. Roger Basu, Dr. David Sucharski, David
Dalgleish and Dr. Alan Benison, whose function was to view this Guide from the viewpoint of practicing
designers and shipbuilders.
We also wish to acknowledge the input of Dr. Richard Yee who commenced the work, and Dr. Harold
Reemsnyder who contributed his expertise and critique.
Finally, but by no means least of all, we thank the SSC’s Project Technical Committee under the
Chairmanhsip of Nash Gifford, Bill Siekeirka and Lt. Tom Miller, USCG, for their input to the project
and their patience for the delayed schedule. We hope the result has justified the wait.
Fatigue-Resistant Detail Design Guide for Ship Structures iii

TABLE OF CONTENTS
PART A – FATIGUE DESIGN GUIDE OVERVIEW
A.1 INTRODUCTION A-1
A.2 OBJECTIVE A-1
A.3 BACKGROUND TO FATIGUE IN SHIP STRUCTURES A-2

A.3.1 Metal Fatigue Process in Steel Ships A-2
A.3.2 Historical Treatment of Metal Fatigue in Steel Ships A-4
A.4 SCOPE OF THE GUIDE A-6

A.4.1 Layout of the Guide A-6
A.4.2 Application of the Guide as a Design Aid A-7
A.5 REFERENCES A-10
PART B – CATALOGUE OF FATIGUE RESISTANT DETAILS
B.1 INTRODUCTION B-1

B.1.1 Background B-1
B.1.2 Objective B-1
B.1.3 Scope B-3
B.2 CRITICAL AND NON-CRITICAL DETAILS B-4

B.2.1 Consequences of Failure (Structural Function) B-4
B.2.2 Probability of Failure (Fatigue Resistance) B-5
B.2.3 Warship Details B-5
B.3 CATALOGUE TERMINOLOGY B-5
B.4 LAYOUT OF THE CATALOGUE B-6

B.4.1 Introduction B-6
B-5 FATIGUE RESISTANT DETAIL CATALOGUE B-12

B.5.1 Tanker Structure Details B-12
B.5.2 Bulk Carrier Structure Details B-34
B.5.3 Container Ship Structure Details B-52
B.5.4 Warship Structure Details B-61
Fatigue-Resistant Detail Design Guide for Ship Structures iv

TABLE OF CONTENTS (continued)
B.6 “GOOD” DESIGN, CONSTRUCTION AND WELDING PRACTICES TO B-89

ENHANCE FATIGUE PERFORMANCE
B.6.1 Introduction B-89
B.6.2 Detail Design B-89
B.6.3 Construction and Welding Practice B-91
B.6.4 Residual Stresses B-92
B.6.5 Weld Toe Dressing Treatments B-93
B.7 REFERENCES B-94
PART C – FATIGUE STRENGTH ASSESSMENT
C.1 INTRODUCTION C-1

C.1.1 Background C-1
C.1.2 Objective C-1
C.1.3 Scope C-1
C.2 FATIGUE ANALYSIS PROCEDURES C-2

C.2.1 Overview C-2
C.2.2 Commentary C-2
C.2.2.1 Intended Applications C-2
C.2.2.2 Design Life C-2
C.2.2.3 Non-linearities C-4
C.2.3 Design Codes and Criteria C-4
C.3 LOAD ANALYSIS C-5

C.3.1 Determination of Environmental Loads C-5
C.3.1.1 Introduction C-5
C.3.1.2 Definition of Loads C-5
C.3.1.3 Definition of Operational Profile C-7
C.3.1.4 Definition of Wave Climate C-11
C.3.1.5 Determination of Wave Load Distribution C-16
C.3.2 Determination of Load Ranges C-20
C.3.2.1 Level 2 Method C-20
C.3.2.2 Level 3 Method C-22
C.3.2.3 High Frequency Loads-Slamming, Whipping and Springing C-27
C.3.3 Summary C-28
Fatigue-Resistant Detail Design Guide for Ship Structures v

TABLE OF CONTENTS (continued)
C.4 STRESS ANALYSIS C-29

C.4.1 Introduction C-29
C.4.2 Stress and Stress Concentration Category Definitions C-30
C.4.2.1 Local Nominal Membrane and Bending Stress (σm and σb) C-31
C.4.2.2 Peak Stress (σp) C-31
C.4.2.3 Residual Stresses (σr) C-31
C.4.2.4 Total Stress C-31
C.4.2.5 Stresses in Fatigue Analysis C-32
C.4.3 Determination of Local Nominal Stresses C-32
C.4.3.1 Level 2 Approach C-33
C.4.3.2 Level 3 Approach Using Finite Element Analysis C-42
C.4.4 Determination of Peak Stresses C-44
C.4.4.1 Evaluation of Peak Total Stress C-44
C.4.4.2 Stress Concentration Factors for Ship Details C-44
C.4.4.3 Local Finite Element Analysis C-45
C.5 DEVELOPMENT OF STRESS RANGE DISTRIBUTIONS C-49

C.5.2 Level 2 Approach C-49
C.5.3 Development of a Stress Range Distribution for the Level 3 Approach C.50
C.6 FATIGUE DAMAGE CALCULATIONS C-54

C.6.2 Fatigue Design Curves C-55
C.6.2.1 Baseline (in Air) C-55
C.6.2.2 Corrosion and Corrosion Fatigue C-59
C.6.2.3 Probability of Failure C-60
C.6.2.4 Thickness Effects C-61
C.6.2.5 Fabrication Considerations C-61
C.6.3 Damage Summation Model C-62
C.6.3.1 Numerical Calculation C-62
C.6.3.2 Closed Form Solution for Level 2 Analysis C-62
C.6.3.3 Cumulative Damage for Level 3 Analysis C-64
C.7 FATIGUE LIMIT STATE RELIABILITY ANALYSIS C-65

C.7.1 Reliability Analysis C-65
C.7.2 Inherent Assumptions or Approximations in Structural Reliability Analysis C-67
C.8 REFERENCES C-68
Fatigue-Resistant Detail Design Guide for Ship Structures vi

PART D - EXAMPLES
D.1 EXAMPLE 1: LEVEL 2 FATIGUE ANALYSIS SAMPLE APPLICATION D-1

D.1.1 Objective D-1
D.1.2 Assessment Approach D-1
D.1.2.1 Define Problem Data D-2
D.1.2.2 Estimate of the Extreme Characteristic Load D-3
D.1.2.3 Estimate of the Service Characteristic Load D-4
D.1.2.4 Estimation of Local Nominal Stress D-4
D.1.2.5 Determination of Hot Spot Stress D-6
D.1.2.6 Define Service Stress Range Distribution D-7
D.1.2.7 Perform Fatigue Damage Summation D-7
D.2 EXAMPLE 2 – LEVEL 3 ANALYSIS D-10

D.2.1 Assessment Approach D-10
D.2.2 Establish Operational Profile D-10
D.2.2.1 Overview D-10
D.2.2.2 Vessel Description D-10
D.2.2.3 Establish Profiles D-11
D.2.3 Calculation of Loads D-20
D.2.4 Determine Reference Hot Spot Stress Range D-24
D.2.4.1 Finite Element Model D-24
D.2.4.2 Determination of the Hot Spot Stress D-25
D.2.5 Short Term Stress Range Distribution D-27
D.2.6 Long Term Stress Range History D-27
D.2.7 Perform Fatigue Damage Summation D-28
D.3 EXAMPLE 3 – SIDE PRESSURE CALCULATIONS

D.3.1 Objective D-31
D.3.2 Approach D-31
D.3.3 Operational Profile Data D-32
D.3.3.1 Definition of the Operational Profile D-32
D.3.3.2 Composite Scatter Diagram and Probabilities of Stationary D-33
Conditions
D.3.4 Side Shell Loading Calculations D-35
D.3.4.1 Pressure Variation due to Vessel Vertical Motions D-35
D.3.4.2 Pressure due to Wave Impinging on the Side Shell D-38
D.3.4.3 Treatment of the Side Pressure Components D-40
D.3.5 Development of RAO’s
D.3.5.1 Static Pressure (Vertical Displacement of POI) D-40
D.3.5.2 Dynamic Pressure (Wave Action Against the Side of the D-41
Vessel
Fatigue-Resistant Detail Design Guide for Ship Structures vii

D.3.6 Calculations of Responses D-43
D.3.7 Calculation of Stress Ranges D-46
D.3.7.1 Damage Summation D-54
Fatigue-Resistant Detail Design Guide for Ship Structures viii

LIST OF FIGURES
Figure A.4.4.1 Application of Fatigue-Design Guide in Ship Design Process A-8

Figure B.1.1 Fatigue Crack Locations and Orientation in Typical Tanker Structure B-2
Figure B.1.2 Fatigue Crack Locations and Orientation in Typical Bulk Carrier Structure B-3
Figure B.4.1.1 Fatigue Design Guide Structural Detail Catalogue Layout B-6
Figure B.4.1.2 Typical Double Hull Tanker Midship Global Structural Arrangement B-7
Figure B.4.1.3 Typical Fatigue Resistant Structural Detail Catalogue Index Table B-9
Figure B.4.1.4 Typical Fatigue Resistant Detail Data Sheets B-10
Figure B.5.1 Typical Midship Section Nomenclature for Double Hull Tankers B-12
Figure B.5.2 Typical Midship Section Nomenclature for Bulk Carriers B-34
Figure B.5.3 Typical Midship Section Nomenclature for Container Ships B-52
Figure B.5.4 Typical Midship Section Nomenclature for War Ships B-61
Figure B.6.3.1 Maximum Welded Cruciform Joint Misalignment B-91
Figure C.2.2.1 Fatigue Strength Analysis Flow Chart: Simplified and Spectral Procedure C-3
Figure C.3.1.3.1 Schematic of Level 3-Direct Calculation-Method C-9
Figure C.3.1.4.1 Scatter Diagram for North-Atlantic for Use in Fatigue Computations C-15
Figure C.3.1.4.2 Scatter Diagram Describing World Wide Trade for Use in Fatigue C-16
Computations
Figure C.3.2.1 RAOs for Oilers, Tankers and Ore Carriers C-25
Figure C.3.2.2 Dynamic Pressure on Side Shell Scenario C-26
Figure C.4.2.1 Stress Components in a Welded Joint C-30
Figure C.4.3.1 Simplified Stress Analysis of Hull Girder C-33
Figure C.4.3.2 Definition of Geometric Parameters for Hull Configurations C-34
Figure C.4.3.3 Secondary Stresses in a Stiffener (Ref. C.3) C-38
Figure C.4.3.4 Global Finite Element Model of Bulk Carrier C-43
Figure C.4.4.1 Local Finite Element Model of Ship Detail C-46
Figure C.4.4.2 Recommended Element Sizes for Local Detail FEA C-47
Figure C.4.4.3 Stress Distribution at an Attachment and Extrapolation of Stresses at Hot C-48
Splot
Figure C.5.2.1a Continuous Weibull Stress Range Distribution C-50
Figure C.5.2.1b Discrete Weibull Stress Range Distribution C-50
Figure C.5.3.1 Operational Condition Rayleigh Distribution Stress Range Interval C-52
Probabilities
Figure C.5.3.2 Long-Term Stress Range Cumulative Probability Function C-53
Figure C.6.2.1 HSE Hot Spot Fatigue Design Curve for Plated as Welded Joints C-58
Fatigue-Resistant Detail Design Guide for Ship Structures ix

LIST OF FIGURES (continued)
Figure D.1.1.1 Location of Interest for Fatigue Calculation D-1

Figure D.1.2.1 Structural Geometry near Region of Interest D-2
Figure D.1.2.2 In Air Hot Spot SN Curve D-3
Figure D.1.2.3 Global Stress Concentration (K G) D-5
Figure D.1.2.4 Location of Peak Stress Concentration D-5
Figure D.1.2.5 Coaming Toe Attachment D-6
Figure D.2.3.1 Vertical Bending Moment (RMS) Distribution D-24
Figure D.2.4.1 Global and Substructure FEM Model D-25
Figure D.2.4.2 FEM Stress Distribution along the Longitudinal D-26
Figure D.2.4.3 Extrapolation of Hot Spot Stress (with Doubler) D-27
Figure D.2.6.1 Long Range Stress History D-29
Figure D.2.7.2 Yearly Fatigue Damage D-31
Figure D.3.2.1 Schematic of Side Shell Load Components D-31
Figure D.3.5.1 Roll RAO Against Encounter Frequency for Beam Seas and Speed of 16 D-41
Knots
Figure D.3.5.2 Dynamic Pressure RAO Against Encounter Frequency for Beam Seas D-41
Figure D.3.7.1 Description of Structural Detail D-45
Figure D.3.7.2a Stress Range Distribution due to Static Pressure D-56
Figure D.3.7.2b Stress Range Distribution due to Dynamic Pressure D-56
Fatigue-Resistant Detail Design Guide for Ship Structures x

LIST OF TABLES
Table A.3.1.1 Highly Loaded Structural Elements – Tankers A-3

Table A.3.1.2 Highly Loaded Structural Elements – Bulk Carriers A-3
Table A.3.1.3 Highly Loaded Structual Elements – Ore Carriers A-4
Table A.3.1.4 Highly Loaded Structural Elements – Container Carriers A-4
Table A.3.1.5 Highly Loaded Structural Elements – Roll On/Roll Off and Car Carrier A-4
Table B.5.1a Critical Structural Details for Double Bottom B-13
Table B.5.1b Critical Structural Details for Double Side B-14
Table B.5.1c Critical Structural Details for Transverse Bulkhead B-15
Table B.5.1d Critical Structural Details for Deck B-15
Table B.5.2b Critical Structural Details for Side Shell B-42
Table B.5.2c Critical Structural Details for Transverse Bulkhead B-43
Table B.5.2d Critical Structural Details for Deck B-43
Table B.5.3b Critical Structural Details for Double Side B-62
Table B.5.3c Critical Structural Details for Deck and Hatch B-63
Table B.5.4a Critical Structural Details for Bottom B-72
Table B.5.4b Critical Structural Details for Side Shell B-72
Table B.5.4c Critical Structural Details for Deck B-73
Table C.3.1.3.1 Sample Operational Profile Joint Probability of Speed and Wave Height C-11
Table C.3.1.3.2 Sample Operational Profile-Joint Probability of Heading and Wave Height C-11
Table C.3.1.3.3 Geographical Distribution of Time C-11
Table C.3.1.4.1 Scatter Diagram for North-Atlantic for Use in Fatigue Computations C-15
Table C.3.1.4.2 Scatter Diagram for World Wide Trade for Use in Fatigue Computations C-15
Table C.3.1.5.1 Composite Distribution of Wave Height Probabilities (fmc) C-17
Table C.3.1.5.2 Two-Dimensional Joint Probability Distribution (fs) C-17
Table C.3.1.5.3 Operational Profile Summary Table C-19
Table C.3.2.3 Operational Profile Summary Table C-19
Table C.4.3.1 Support Bending Stress Coefficients kb – Double Bottom Panels C-39
Table C.4.3.2 Support Bending Stress Coefficients kb – Single Skin Panels C-40
Table C.4.3.3 Definition of Stiffness and Geometry Parameters C-41
Table C.5.3.1 Operational Condition Rayleigh Distribution Stress Range Interval C-52
Probabilities
Table C.5.3.2 Weighted Short-Term and Resulting Long-Term Stress Range Interval C-50
Probabilities
Fatigue-Resistant Detail Design Guide for Ship Structures xi

LIST OF TABLES (continued)
Table D.1.2.1 Vessel Characteristics D-2
Table D.1.2.2 Hot Spot SN Curve D-3
Table D.1.2.3 Extreme and Service Bending Moments D-4
Table D.1.2.4 Hot Spot Stresses at Location of Interest D-6
Table D.1.2.5 Sample of Fatigue Damage D-8
Table D.1.2.6 Cumulative Damage Results D-10
Table D.1.2.7 Fatigue Life Estimates D-10
Table D.2.2.1 Joint Probability of Speed and Sea State D-12
Table D.2.2.2 Joint Probability of Relative Heading and Sea State D-12
Table D.2.2.3 Listing of Geographical Zones (Marsden Zones) Ship Traversed and D-13
Relative Time Spent in each Zone
Table D.2.2.4 Operational Profile Summary D-13
Table D.2.2.5 Wave Data by Marsden Zone [from Ref. D.3] D-16
Table D.2.2.6 Composite Marsden Zone Scatter Diagram D-17
Table D.2.2.7 Transformation of Sea State and Wave Height D-18
Table D.2.2.8 Probabilities of all Combinations of Speed, Heading and Sea State for Zero D-21
Crossing Period Tz between 5 and 6 Seconds in Combined Marsden Zone
Table D.2.3.1 Ship Response Calculation: Output from Shipmo7 D-23
Table D.2.4.1 Hot Spot Transfer Functions for the Location of Interest D-27
Table D.2.6.1 Weibull Statistics for Long Range Stress History D-29
Table D.2.7.1 Partial Damage Accululation per Year (No Doubler) D-30
Table D.2.7.2 Summary of Fatigue Damage D-30
Table D.3.3.1 Principal Data: Example Ship D-32
Table D.3.3.2a Joint Probability of Speed and NATO Sea State for Example Ship D-32
Table D.3.3.2b Joint Probability of Heading and NATO Sea State for Example Ship D-32
Table D.3.3.2c Relative Time Spent in Marsden Zones for Example Ship D-33
Table D.3.3.3 Composite Marsden Zone Scatter Diagram D-33
Table D.3.3.4 Transformation of Wave Heights to Sea States for Combined Scatter D-34
Diagram
Table D.3.4.1a Probabilities of all Combinations of Speed, Heading and Sea State for Zero D-37
Crossing Periods Tz between <4 and 9-10 seconds in Combined Marsden
Zone
Table D.3.4.1a Probabilities of all Combinations of Speed, Heading and Sea State for Zero D-37
Crossing Periods Tz between 10-11 and 14> seconds in Combined
Marsden Zone
Table D.3.6.1 Calculation of Lifetime Average Zero Crossing Period D-43
Table D.3.6.2 Sample of Shipmo7 Output and Dynamic Pressure Calculations D-44
Table D.3.7.1 Geometry of the Panel D-47
Table D.3.7.2 Web Frame and Panel Geometry D-50
Table D.3.7.3 Kb values (Taken from Table C.4.3.2) D-51
Table D.3.7.4 Stress Ranges and Corresponding Probabilities D-55
Table D.3.7.5 Parameters used in Example D-57
Fatigue-Resistant Detail Design Guide for Ship Structures xii

Table D.3.7.6 Stress Ranges and Corresponding Fatigue Damage D-58
Fatigue-Resistant Detail Design Guide for Ship Structures xiii

NOMENCLATURE
ABS American Bureau of Shipping
AP Aft perpendicular
Ai Stress coefficient
a Crack depth: or
Length of double bottom panel; or
Scale parameter of the basic S-N curve, or
Acceleration (m/s2)
al Combined longitudinal acceleration (m/s2) [Appendix B]
ao Acceleration constant [Appendix B]
apt Tangential pitch acceleration (m/s2) [Appendix B]
apx Longitudinal component of pitch acceleration (m/s2) [Appendix B]
apz Vertical component of pitch acceleration (m/s2) [Appendix B]
ary Horizontal component of roll acceleration (m/s2) [Appendix B]
arz Vertical component of roll acceleration (m/s2) [Appendix B]
at Combined transverse acceleration (m/s2) [Appendix B]
av Combined vertical acceleration (m/s2) [Appendix B]
ax Surge acceleration (m/s2) [Appendix B]
ay Acceleration due to sway and yaw (m/s2) [Appendix B]
az Heave acceleration (m/s2) [Appendix A]
a1 a2 Fatigue design curve parameter
B Moulded breadth of ship (m or ft); plate thickness [Appendix B]

BV Bureau Veritas
b Transverse width of double bottom panel
CG Center of gravity
CMS Committee on Marine Structures
CB Block coefficient [Appendix B]
Cw Wave coefficient [Appendix B]
Cwp Waterplane area coefficient
D Cumulative fatigue damage

di Fatigue damage inflicted by each stress cycle
Dm Moulded Depth of Ship (m)
DNV Det Norske Veritas
E Young’s Modulus [GPA]
FE Finite Element
FP Forward perpendicular
FTL Fleet Technology Limited
Fatigue-Resistant Detail Design Guide for Ship Structures xiv

F(Hs1) Cumulative probability of lower limit wave height
F(Hs2) Cumulative probability of upper limit wave height
F(Sea State) Probability of occurrence of a given sea state
Fatigue-Resistant Detail Design Guide for Ship Structures xv

Fload(Hs,V,θ,L.C.) Characteristic (RMS) load response of the vessel for each operational condition
(defined by wave height, speed, heading, load condition)
F∆σi(∆σhot spot) Rayleigh short-term stress range cumulative distribution function for the ith operational
condition
f Freeboard at the transverse section considered (m or ft.) [Appendix B]
ff Correction factor to convert from one probability level to another probability level
fef Encounter frequency correction factor
fmc(Hs) Composite distribution of significant wave heights
fmc (Hs;Tz)composite Composite distribution of wave heights and zero crossing periods (composite
scatter diagram)
fmi(Hs;Tz) Probability distribution of wave heights and zero crossing periods in Marsden Zone
fmcomb(Hs;Tz) Probability of wave height/zero crossing period in the Combined Marsden Zone
fstotal Total probability (three-dimensional probability) or percent of time for each operational
condition
fV (VHs) Conditional probability of speed, V, given a wave height, Hs (or sea state) – all periods
included
fV (V | (Hs;Tz)) Conditional probability of speed, V, given a wave height combination or sea state,
(Hs;Tz)
fθ (θHs) Conditional probability of heading,θ , for given wave height, Hs, - all periods included
fθ (θ | (Hs;Tz)) Conditional probability of heading,θ , given a sea state, (Hs;Tz)
fz zero crossing rate (Hz)
fz average zero crossing rate (Hz)
g Gravitational constant
Gx General form for the Limit State equation in a First Order reliability analysis
GL Germanischer Lloyd
GM Metacentric height [Appendix B]
HSE UK Health and Safety Executive

Hs Significant wave height (m or ft.)
h Shape parameter for Weibull Distribution (general)
hn Weibull shape factor for nth loading condition
ho Basic long-term Weibull shape parameter, modified for location of Point of Interest
hs Vertical distance from point considered to surface inside a tank (m)
[Appendix B]
IACS International Association of Classification Societies

IIW International Institute for Welding
ISSC International Ship and Offshore Structures Congress
I Moment of Inertia (general)
Ia, Ib Moment of inertia about the transverse neutral axis, including the effective width of
plating, of long (a) and short (b) stiffeners
Fatigue-Resistant Detail Design Guide for Ship Structures xvi

Ih Hull cross section moment of inertia about the vertical neutral axis
Ipi Moment of inertia of effective width of plate – i = a or b for long or short direction
(grillage)
Iv Hull cross section moment of inertia about the transverse neutral axis
ia, ib Blended stiffness per unit (girder and plate) about transverse (longitudinal) neutral axis
of double bottom
K Stress concentration factor (general)

Kb Stress Concentration Factor Dependent on aspect ratio, ρ, and panel’s boundary
conditions
KG Global stress concentration factor to account for gross structural geometry (e.g., hatch
openings, shear lag) affecting the local nominal stress field
Kg Stress concentration factor due to the gross geometry of the detail
Kn Stress concentration factor due to non-symmetric stiffeners
Kte Stress concentration due to eccentricity
Ktα Stress concentration due to angular mismatch
Kw Notch stress concentration; local weld configuration stress concentration
K0 Stress transfer function relating vessel load response to detail hot spot stress
k Wave number (from dispersion relation (ω2/g), or
number of stress blocks
k′ roll correction factor for bilge keels [Appendix B]
kf Side pressure panel factor [Appendix B]
kr Roll radius of gyration (m) [Appendix B]
kwm Moment distribution factor
kws Shear distribution factor
ks Side pressure form factor [Appendix B]
L Rule length of ship (m or ft) [Appendix B]

LBP Length between pependiculars (m or ft.)
LOA Length overall (m or ft.)
L.C. Load condition (e.g., ballast, full load, etc.) [Appendix B]
LNG Liquified Natural Gas
LPG Liquified Petroleum Gas
LR Lloyd's Register
le Effective span of stiffener or longitudinal (m or ft.)
ls Distance (span) between bulkhead and transverse frame (m or ft.)
M Bending moment (MNm or LTft.)

Md Vertical bending moment range (MNm or LTft.)
Mdh Design wave-induced hogging moment amplitude (MNm or LTft.) including the effects
of whipping [Appendix B]
Mds Design wave-induced sagging moment amplitude (MNm or LTft.) including the effects
of whipping
Fatigue-Resistant Detail Design Guide for Ship Structures xvii

Mh Horizontal bending moment amplitude at the location of interest (MNm or LTft.)
Mhog Hogging moment (MNm or LTft.)
Msw Stillwater bending moment (MNm or LTft.) [Appendix B]
Msag Sagging Moment (MNm or LTft.)
Mv Vertical (sagging or hogging) bending moment amplitude at the location under
consideration (MNm or LTft.)
Me Maximum allowable misalingment measured from the centerlines of intersecting plates
(mm or in.)
MEDS Marine Environmental Data Services
m Location parameter of the Weibull distribution
m1, m2 Fatigue design curve slope parameters
m0 Spectral zeroth moment employed in spectral analysis
mδ Moment factor due to relative deflection between transverse supports
N Total number of cycles (wave encounters or stress reversals) experienced by the

structure within a time, t; used to express Fatigue Life.
Ni Average number of loading cycles to failure under constant amplitude loading at the ith
stress range
Nt Total number of loading cycles to failure
Nload Total number of load conditions
N′ Total number of Marsden Zones along the route
Ns Number of cross ties in cargo or ballast tank
NOAA National Oceanic and Atmospheric Administration
ni Number of stress cycles in stress block “i”
no Number of wave encounters corresponding to the service level of probability for
load/stress
Pf Probability of fatigue-reduced failure

POI Point of Interest
p Effective lateral pressure
pd Dynamic pressure amplitude (kPa)
pdp Combined pressure dominated by pitch motion in head/quartering seas
[Appendix B]
pdr Combined pressure dominated by roll motion in beam/quartering seas
[Appendix B]
pe External pressure amplitude (half pressure range) related to the draft of the load
condition considered (kPa) [Appendix B]
pext External pressure (kPa)
pi Fraction of time at the ith operational condition
pint Internal pressure (kPa) [Appendix B]
p1 Pressure due to vertical acceleration [Appendix B]
p2 Pressure due to transverse acceleration [Appendix B]
p3 Pressure due to longitudinal acceleration
Fatigue-Resistant Detail Design Guide for Ship Structures xviii

pip Dynamic pressure term (kPa) for pitch-induced pressure
pst Static pressure
pn Fraction of design life in the nth load condition
prob(Hs) Marginal probability of wave heights
prob(Hs; Tz) Marginal probability of the sea state
prob(V and (Hs; Tz)) Joint probability of speed and sea state
prob(θ and (Hs; Tz)) Joint probability of heading and sea state
prob(emergence) Probability of emergence of POI for each stationary condition
Q(∆σhotspot) Weibull long-term cumulative probability distribution functions for two parameter
Weibull distribution (h and q parameters)
Q′(∆σhotspot)long Weibull long term probability function pertinent to hot spot stress
Q′(∆σhotspot)short Probability distribution function pertinent to hot spot stress and a single statiionary
condition (short term)
q Scale parameter for Weibull distribution
RAO Response amplitude operator

RINA Royal Institute of Naval Architects
RMS Root mean square
Rpp Distance from (pitch) axis of rotation to local centre of mass or volume (m)
[Appendix B]
RPX Longitudinal distance from the pitch axis of rotation to the local centre of mass or
volume (m) [Appendix B]
RPZ Vertical distance from the pitch axis of rotation to the local centre of mass or volume
(m) [Appendix B]
RR Distance from the axis of roll rotation to the local centre of mass or volume (m)
[Appendix B]
RRY Transverse distance from the roll axis of rotation to the local centre of mass or volume
(m or ft.) [Appendix B]
RRZ Vertical distance from the roll axis of rotation to the local centre of mass or volume (m
or ft.) [Appendix B]
R(ωe) Response function
ra Distance from point considered to the transverse neutral axis of panel (m or ft.)
rb Distance from point considered to the longitudinal neutral axis of panel (m or ft.)
ri Ratio of the ith zero crossing rate, to the average zero crossing rate for all operational
conditions
rij Relative number of stress cycles in short-term condition i, j against the total number of
cycles in the vessel life
rδ, rπ Moment factors for interpolation to crack location along stiffener length
rp Reduction of pressure amplitude in the wave zone [Appendix B]
SCF Stress Concentration Factor

SSC Ship Structure Committee
Fatigue-Resistant Detail Design Guide for Ship Structures xix

SWL Summer load waterline
Sθ Wave spectral density (m2*s)
Sη(ωe) Modified wave height spectrum (m2*s)
s Stiffener spacing (mm or in.)
sa Transverse spacing between girders or longitudinals running in the longitudinal direction
(m or ft.)
sb Longitudinal spacing between girders or web frames in the transverse direction (m or ft.)
Fatigue-Resistant Detail Design Guide for Ship Structures xx

Tact Stillwater draft at the considered load condition (m) [Appendix B]
T Draft of ship at load condition (m); also
Period (s)
T Long-term average stress period (s)
TP Peak wave period (s)
TPP Period of pitch (s) [Appendix B]
TR Period of roll (s) [Appendix B]
TS Significant wave period (s)
Tz Zero crossing period (s)
t Time (s); also
Thickness
tb Bracket thickness (mm or in.)
tcorr Corrosion thickness allowanced
td Design life of ship expressed in seconds (s)
tf Stiffener flange thickness (mm or in.)
tn Net plate thickness (mm or in.)
tr Reference thickness (mm or in.)
ts Continuous plate thickness (mm or in.)
tw Web thickness (mm or in.)
TAPS Trans-Alaska Pipeline Service
V Vessel speed (knots) [Appendix B]

VLCC Very Large Crude Carrier
VH Horizontal shear (kN) [Appendix B]
Vdh Vertical shear load at midship due to hogging (kN) [Appendix B]
Vds Vertical shear load at midship due to sagging (kN) [Appendix B]
x Length measurement variable (general), or

Longitudinal distance from the AP to the section considered (m or ft.)
xc Distance from end of stiffener to crack location (mm or in.)
xs Longitudinal distance from centre of free surface of liquid in tank to pressure point
considered (m or ft.) [Appendix B]
y length measurement in variable in transverse direction (general), or
Transverse distance from the centre line to the point of interest (m or ft.) [Appendix B]
ys Transverse distance from centre of free surface of liquid in tank to point of interest (m or
ft.) [Appendix B]
y Transverse distance from centreline for roll calculation [Appendix B]
Fatigue-Resistant Detail Design Guide for Ship Structures xxi

Z Section modulus
Zdeck of deck
Zs, ZI of stiffener or longitudinal (mm3 or in.3)
z Vertical distance on the hull:
zna from the neutral axis to the point of interest (m or ft.)
zbl from the baseline to the point of interest (m or ft.)
zwl from the still waterline (m or ft.) (ship upright) [Appendix B]
z1, z2 Instantaneous immersions of POI due to ship motion (m or ft.)
α Maximum roll angle, single amplitude (rad) [Appendix B]
β Reliability index derviced from limit state equation G(x)
Γ( ) Gamma function
Γ( ; ) Complementary incomplete gamma function
γ( ; ) Incomplete gamma function
∆σc Corrected design stress range

∆σhotspot Reference hot spot stress range (MPa)
∆σnom Reference nominal stress range (MPa)
∆σnotch Notch stress range (MPa) [Appendix A]
∆σo Design stress range for the nth loading condition (MPa)
∆σslope Stress range at which change in slope occurs (MPa)
∆σS-N Design stress range allowed by the relevant S-N curve (MPa)
δ Deformation of nearest frame relative to transverse bulkhead (mm or in.)
ε Stress correlation coefficient
Φ Standard normal distribution function

φ Pitch angle, single amplitude (rad)
η Fatigue usage factor, or

Torsion factor in grillage
θ Ship's heading relative to wave direction (degrees, 0o = head seas)

θ′ spreading angle for wave spectrum (rad)
λ Wave length (m or ft.)
µG Mean of the limit state equation G(x)

µi Proportion of time spent in the ith area (Marsden zone)
ν Poisson ratio
Fatigue-Resistant Detail Design Guide for Ship Structures xxii

ρ panel aspect ratio
ρij Correlation coefficient for variables i and j in limit state analysis
ρsw Density of seawater (1.025 t/m3)
σ Stress (MPa)
σb Local bending stress (MPa)
σeq Equivalent stress (MPa)
σG Standard deviation of limit state equation G(x)
σm Membrane stress
σp Peak stress (MPa)
σr Residual stress (MPa)
σy Yield strength (MPa)
σt Total stress at crack location (MPa)
σtp Peak total stress at crack location (MPa)
σnom Nominal stress
σlower Stress range bin lower value (MPa)
σmid Stress range bin mid value (MPa)
σupper Stress range bin upper value (MPa)
σ1 Primary stresses due to bending, shear and torsion in the main hull girder; peak total
stress (MPa)
σ2 Secondary stresses due to local stiffener bending (MPa)
σ2L Plate/panel secondary stresses (MPa)
σ3 Tertiary plate bending stress (MPa)
σδ Stresses due to relative deflection between web frame and transverse bulkhead (MPa)
ω Wave frequency (rad/s)

ωe Encounter frequency that accounts for the effects of speed and heading
Fatigue-Resistant Detail Design Guide for Ship Structures xxiii

Part A - Fatigue Design Guide Overview
PART A - OVERVIEW OF THE FATIGUE DESIGN GUIDE
A.1 INTRODUCTION
In March 1995, the US National Research Council’s Committee on Marine Structures (CMS),
on behalf of the Ship Structure Committee (SSC), convened a symposium in Washington, D.C., to
address the unusually great number of bulk carrier losses and the continuing occurrence of fatigue
cracking in relatively new and aging ships [Ref. A.1]. The symposium sought input from ship designers,
fabricators, operators, and regulators as well as experts in fatigue, fracture, and structural reliability on
how the Ship Structure Committee could best serve the marine community in preventing this type of
failure in ships.
One outcome of this symposium was that practicing naval architects and engineers were having
difficulty applying the results of much of the research and development work that had been carried out
to investigate fatigue, fracture and structural reliability. As a consequence, the Ship Structure
Committee initiated two related projects:
• development of a ship structural detail fatigue design Guide; and,
• development and presentation of a ship structure fatigue and fracture short course.
These two projects were awarded to Fleet Technology Limited under a competitive bidding
process, under contract PO Number 97-0046.
This report is the result of the ship structure detail fatigue design Guide development project.
A.2 OBJECTIVE
The objective of the project was to develop a practical and rationally based “Fatigue Resistant
Detail Design Guide” (“the Guide”) that Engineers and Naval Architects with limited fatigue design
experience could use to design cost-effective, fatigue-resistant, welded steel ship structural details.
The objective of the Guide is to provide a single source of information for practicing designers
to use in addressing fatigue issues in the design of ship structure. In order to achieve this objective, the
Guide was required to cater to all levels of design detail or phases in the design process. For this
reason, the Guide presents three levels of structural design sophistication:
• structural detail geometry and layout guidance;
• a simple fatigue design procedure; and,
• an in-depth fatigue design procedure.
These can be used sequentially at successive stages in the ship design process. However, in
many projects it may not be necessary to use the more detailed levels to achieve satisfactory fatigue
performance.
Fatigue-Resistant Detail Design Guide for Ship Structures A-1

A.3 BACKGROUND TO FATIGUE IN SHIP STRUCTURES
A.3.1 Metal Fatigue Process in Steel Ships

Metal fatigue is the progressive failure of metal under cyclic loading and as the name “fatigue”
implies, it is a mode of degradation in which the steel is worked until it simply gets tired. This fatigue
cracking process can be divided into three basic stages:
(i) the initiation of microscopic cracks at local stress concentrations;
(ii) the growth and coalescence of microscopic cracks into macroscopic cracks; and,
(iii) the growth of macroscopic cracks to a critical size for failure (e.g., plastic collapse, fracture,
excessive deflection, or loss of water-tightness).
The absolute and relative duration of these stages depend on the magnitude of the cyclic stresses or
strains at the crack initiation site and along the crack propagation path, environmental effects, and the
resistance of the material to the initiation and propagation of fatigue cracks. The magnitude of the local
cyclic stress or strain is dependant on the magnitude of applied cyclic loads, severity of local stress
concentration effects, structural redundancy, and the stress-strain response of a material under cyclic
loading [e.g., Ref. A.2].
Fatigue cracks in steel ships generally initiate at local notches (e.g., weld toe, rat hole, or weld
termination) in structural details that are located in highly stressed, primary or secondary structures [e.g.,
Ref. A.3, A.4]. Fatigue-prone areas in common types of ships are listed in Tables A.3.1-A.3.5 [Ref.
A.5]. The initiation and subsequent propagation of these cracks are primarily driven by wave induced
cyclic loads including:
(i) longitudinal bending, transverse bending, and torsion of the hull girder; and,
(ii) fluctuating hydrostatic pressure on side shell plating, cargo hold boundaries, and tank walls,
[e.g., Ref. A.6].
Other sources of cyclic loading include machine vibration, propeller-induced vibration, and deck loads .
The exposure of unprotected surfaces to corrosive media (e.g., sea water or sour crude oil) can also
accelerate the initiation and propagation of fatigue cracks, either directly through corrosion fatigue
mechanisms or indirectly through the higher cyclic stresses that result from localized and general
corrosion [Ref. A.7].
Although most fatigue cracks in ships are not detected by conventional inspection techniques
until they are several inches long and through the thickness of plating, the majority of detected cracks do
not pose an immediate threat for catastrophic rapid fracture because of the relatively good fracture
toughness of modern ship steels, the inherent redundancy of ship structures, the use of crack arrestors,
and the relatively low level of normal service loads. Nevertheless, any detected cracks are usually
repaired at the earliest opportunity to preserve the water-tightness or oil tightness of the ship and to
prevent a chain reaction of localized failures leading to overall structural failure. The latter scenario is
believed to be responsible for some of the recent spate of bulk carrier losses [Ref. A.8].

For example, cracks at the corners of hatch corners and coamings in bulk carriers could allow
water to leak into dry cargo holds. Sloshing of the resulting slurry could introduce significant dynamic
loads on cargo hold plating, and internal mixing could produce explosive gases. As another example,
cracking at the hold frame ends of bulk carriers could result in the detachment of side shell plating from
internal framing. This could eventually lead to the separation of the end brackets from the slant of the
topside tanks or bilge hopper tanks.
Table A.3.1.1: Highly Loaded Structural Elements - Tankers

Structural Member Structural Detail Load Type
Side, bottom and deck plating Butt joints, deck openings and Hull girder bending, stiffener
and longitudinals attachment to transverse webs, lateral pressure load and support
transverse bulkheads and deformation
intermediate longitudinal girders
Transverse girder and Bracket toes, girder flange butt joints, Sea pressure load combined with
stringer structures curved girder flanges, panel knuckles cargo or ballast pressure load
including intersecting transverse
girder webs, etc. Single lug slots for
panel stiffeners, access and lightening
holes
Longitudinal girders of deck Bracket terminations of abutting Hull girder bending and
and bottom structure transverse members (girders, bending/deformation of
stiffeners) longitudinal girder and abutting
member
Table A.3.1.2: Highly Loaded Structural Elements - Bulk Carriers

Hatch corners Hatch corner Hull girder bending, hull girder
torsional deformation
Hatch side coaming Termination of end bracket Hull girder bending
Main frames End bracket terminations, weld External pressure load, ballast
main frame web to shell for un- pressure load as applicable
symmetrical main frame profiles
Longitudinals of hopper tank and Connection to transverse webs Hull girder bending, sea and
top wing tank and bulkheads ballast pressure load
Double bottom longitudinals Connection to transverse webs Hull girder bending stress,
and bulkheads double bottom bending stress
and sea, cargo and ballast
pressure load
Transverse webs of double Slots for panel stiffener including Girder shear force, and bending
bottom, hopper and top wing stiffener connection members, moment, support force from
tank knuckle of inner bottom and sloped panel stiffener due to sea, cargo
hopper side including intersection and ballast pressure load
with girder webs (floors). Single
lug slots for panel stiffeners,
access and lightening holes

Table A.3.1.3: Highly Loaded Structural Elements - Ore Carriers

Upper deck plating Hatch corners and side coaming Hull girder bending
terminations
Side-, bottom- and deck Butt joints and attachment to Hull girder being, stiffener lateral
longitudinals transverse webs, transverse pressure load and support
bulkheads, hatch openings corners deformation
and intermediate longitudinal girders
Transverse girder and Bracket toes, girder flange butt joints, Sea pressure load combined with
stringer structures curved girder flanges, panel knuckles cargo or ballast pressure
at intersection with transverse girder
webs, etc. Single lug slots for panel
stiffeners, access and lightening
holes
Transverse girders of wing Single lug slots for panel stiffeners Sea pressure load (in particular in
tank ore loading condition)
Table A.3.1.4: Highly Loaded Structural Elements - Container Carriers

Hull Member Structural Detail Load Type
Side and bottom longitudinals Butt joints and attachment to Hull girder bending, torsion ,
transverse webs, transverse stiffener lateral pressure load and
bulkheads and intermediate support deformation
longitudinal girders
Upper deck Plate and stiffener butt joints, hatch Hull girder bending and torsional
corner curvatures and support details warping stress
welding on upper deck for container
pedestals, etc.
Table A.3.1.5: Highly Loaded Structural Elements - Roll on/Roll off- and Car Carrier
Hull Member Structural Detail Load Type
Side and bottom longitudinals Butt joints and attachment to Hull girder bending, stiffener
transverse webs, transverse lateral pressure load and support
bulkheads and intermediate deformation
longitudinal girders
Racking constraining girders, Stress concentration points at girder Transverse acceleration load
bulkheads, etc. supports and at bulkhead openings
A.3.2 Historical Treatment of Metal Fatigue in Steel Ships

The design of a ship can be divided into three phases: conceptual design, preliminary
design, and detail design. The principal dimensions, topology, and overall geometry of a ship are
determined during the conceptual design phase, usually by non-structural considerations such as
beam and draft limitations, cargo type, and cargo capacity.

The location, spacing, scantlings, and geometry of primary structure (e.g., side shell, decks,
main bulkheads, main beams and girders) and secondary structure (e.g., stiffened panels, grillages,
tank tops, short decks) are then selected during the preliminary design stage. These selections, and
the resulting degree of structural continuity, optimization, and symmetry, determine the ultimate
strength of the hull. In the next phase, detail design, the geometry and scantlings of details are
selected within fabrication and maintenance constraints to minimize local stresses.
Until recently, fatigue cracking has not been explicitly considered in the detail and
preliminary design phases except for damaged structure, novel structural configurations, and
special types of ships such as LNG ships. The preliminary design of merchant ships has been
largely based on static strength requirements in classification society rules [Ref. A.9- A.12], which
are expressed in terms of empirical relationships for minimum scantlings and spacings, whereas the
preliminary design of naval ships has been based on more rigorous static strength calculations with
representative design loads [Ref. A.13, A.14]. Prior to the 1990’s, designers and fabricators were
able to rely on the margins in these static strength requirements and experience-based rules-of-
thumb for detail design to achieve adequate fatigue performance in conventional ship structures.
Fatigue cracks were rarely detected in ships less than 10 years old, and the frequency of cracking
in older ships was generally acceptable to regulators and owners. Over the past two decades,
however, significant changes in the age, design, fabrication, operation, and regulation of merchant and
naval vessels have resulted in the need for rationally-based fatigue design approaches to address the
following concerns:
• There has been a significant increase in the incidence of fatigue cracking in relatively new ships
since the early 1980’s. For example, fatigue cracks were detected in the summer of 1990 at the
intersections of side shell longitudinals and transverse bulkheads in 15 second-generation VLCC’s
after only two to five years of service [Ref. A.15]. Similar cracking was reported in several
classes of oil tankers operating on the Trans-Alaska Pipeline Service (TAPS) route in the mid-
1980’s [Ref. A.16 and A.17]. This change has been attributed to the introduction of more
structurally optimized ships with thinner scantlings as a means to reduce weight, fabrication cost
and operating cost. This optimization has been achieved through the greater use of high strength
steels and the use of more sophisticated design tools [Ref. A.18, A.19]. Since the fatigue strength
of as-welded steel joints is essentially independent of tensile strength, the stress concentrations of
structural details must be adequately reduced to compensate for the higher design stresses, higher
local bending stresses, and reduced margins for corrosion and wear in high strength steel
structures with thinner scantlings. As a general precaution, classification societies introduced the
so-called “k factor” on minimum scantling requirements which prevented design stresses for
extreme loads from increasing in direct proportion to tensile strength. However, the continuing
occurrence of fatigue cracking in ships demonstrated the need for more direct control of fatigue
cracking. A number of owners with poor experiences have since insisted that the use of high
strength steels in their ships be kept to a minimum, and shipyards have responded with standard
designs using mainly low and medium strength steels.

• According to available statistics [e.g., Ref. A.20, A.21], the rate of reported fatigue cracking and
other structural failures in current merchant and naval vessels increases significantly (i.e., the aging
phase starts) after the first 10 years of service. However, approximately 60% of the world’s
tanker fleet and about 40% of the world’s bulk carrier fleet (by tonnage) are over 15 years old,
and many naval vessels are of the same vintage. As a result, maintenance costs and downtime are
rising for operators. At the same time, many operators are facing reduced maintenance and
operating budgets. This is forcing designers, fabricators, and operators of merchant and naval
vessels to seek integrated approaches to the design, construction, and maintenance of ships with a
view towards maximizing operational availability and minimizing life-cycle maintenance costs
without compromising structural integrity [Ref. A.22, A.23].
• Strict environmental regulations have been introduced around the world since the grounding of the
Exxon Valdez. These regulations will require nearly all new oil tankers to have double hulls by
the year 2020 [Ref. A.24]. The double hull is intended to protect against oil spills caused by hull
punctures. However, the cellular arrangement of double hull tankers makes it difficult to clean
and/or ventilate ballast spaces. As a result, there is a potential risk for explosion if fatigue
cracking permits cargo oil or vapours gas to leak into these spaces.
• In response to the aforementioned concerns, classification societies have recently introduced
rationally-based procedures for the fatigue design of structural details in steel ships [e.g., Ref.
A.25-A.27], and they have made these procedures mandatory for novel structural configurations
and large ship designs (e.g., tankers and container ships longer than 190 m, bulk carriers longer
than 150 m). Included in these procedures are: (i) simplified methods for quantifying the fatigue
performance of structural details in common problem areas, (ii) spectral-based methods for
quantifying the fatigue performance of structural details that cannot be properly analyzed with
simplified methods, and, (iii) qualitative guidelines for optimizing the fabrication and fatigue
performance of structural details. Similar procedures are being developed by several navies [Ref.
A.28, A.29] and an ad hoc working group of the International Association of Classification
Societies (IACS).
A.4 SCOPE OF THE GUIDE
A.4.1 Layout of the Guide

The design aid information in the Fatigue Resistant Detail Design Guide for Ship Structures is
organized into four parts, entitled:
• Part A - Introduction
• Part B - Catalogue of Fatigue Resistant Details
• Part C - Fatigue Strength Assessment
• Part D - Fatigue Design Examples
Part A, the design Guide introduction, provides some background to illustrate the need for
considering fatigue in the design process and describes the objectives and layouts of the remainder of
the Guide.

Part B and Part C of the Guide present information to assist the designer to consider fatigue
design throughout the detail design process. The three levels of design procedure sophistication which
are presented in this Guide include:
• Level 1 - selection of fatigue resistant structural detail geometries and layouts;
• Level 2 - a simplified fatigue design procedure; and,
• Level 3 - an in-depth fatigue design procedure.
The increasing levels of the design procedures not only correspond to increased analytical detail and
accuracy but also cost in terms of the time, effort and data required for implementation.
Part B of the Guide presents the Level 1 fatigue design aid in the form of a catalogue of fatigue
resistant design details. The catalogue itemizes all key structural connections encountered in the midship
area of various vessel types and provides geometric and structural arrangement suggestions to improve
the fatigue performance of the detail of interest. The Level 1 structural detailing process may be used as
a first step prior to a more sophisticated numerical fatigue strength assessment process or could be used
alone as a simplified means of improving the fatigue performance of a detail without quantifying the net
effect.
Part C of the fatigue design Guide presents rationally-based methods for quantifying the fatigue
strength of welded structural details and provides detailed guidance and instructions for applying these
methods. Level 2 and Level 3 fatigue design procedures are both presented in Part C due to their
common and interchangeable procedural steps. The Level 2 fatigue strength assessment approach
makes extensive use of empirical design equations to simplify the analysis process, whereas, the Level 3
fatigue strength assessment approach makes full use of the statistical information and numerical modeling
techniques commonly available to designers.
Part D of the fatigue design Guide presents worked design examples used to demonstrate the
application of the three levels of design procedure sophistication.
A.4.2 Application of the Guide as a Design Aid

The ship design process may be described schematically as shown in Figure A.4.1, and the
fatigue design Guide has been assembled to aid in the detail design process. The Guide includes
information to promote good detail design early in the detail design stage, as well as analytical
techniques, which may be used in an iterative manner to optimize the scantlings of selected details. The
fatigue strength assessment techniques may also indicate a need for the designer to rework the overall
structural design (i.e., return to the preliminary design stage) if acceptable fatigue performance cannot be
achieved by optimization at the detailed level.

Conceptual Design
Objective
• establish principal dimensions, layout, overall geometry of ship
Constraint
• mainly controlled by non-structural considerations (e.g., beam, draft, cargo)
Preliminary Design
Objective
• establish location, spacing, scantlings and geometry of primary structure
Constraints
• mainly controlled by static strength requirements (e.g., plastic collapse, buckling) in initial
iterations
• modifications to primary structure may be required if fatigue strength of details cannot be
controlled by detail design
Approach
• rule based for standard configurations of commercial vessels
• rationally based for naval vessels and non-standard configurations of commercial vessels
(e.g., tankers longer than 190 m)
Detail Design
Objective
• establish the geometry and scantlings of local details, joints, brackets, openings and
reinforcements
Constraints
• detail design mainly controlled by fatigue performance, functionality and fabrication
constraints
Approach
• select from catalogue of preferred details for a particular location in a particular type of ship
• use fatigue assessment procedures to optimize scantlings of details
Figure A.4.4.1: Application of Fatigue-Design Guide in Ship Design Process
As previously noted for fatigue cracking in general, the service life of a ship may be
divided into three distinct phases including:
(i) the teething phase in which fatigue cracks initiate from fabrication defects at a decreasing
rate with time;
(ii) the stable phase in which fatigue cracks initiate randomly at a constant rate with time; and
(iii) the aging phase in which the rate of fatigue cracking increases with time because of
cumulative fatigue damage and other structural degradation (e.g., wear and corrosion).
These three phases lead to a cracking rate vs. time profile that is sometimes referred to as
the “Bathtub Model” of service life.

This Guide is intended to produce designs for steel welded ship structural details in which
the aging phase starts towards the end of a ship’s service life and in which fatigue cracking rarely
occurs during the teething and stable phases, provided the ships are well fabricated and maintained.
This Guide does not account for the possible onset of unstable fracture from a fatigue crack. In
order to guard against this possibility, ships should be designed as damage tolerant structures (i.e.,
structures that can sustain maximum design loads without failure until damage is detected and repaired).
Damage tolerance can be achieved by appropriate material selection at the design stage, the provision
of multiple or redundant load paths, and the use of readily inspected structural details. In addition,
damage tolerance analysis can be used to quantitatively assess the residual fatigue lives and residual
strength of ship structures with fabrication defects or in-service cracks. Fracture toughness
requirements for ship steels and welding consumables are given in various Naval and Classification
Society documents [e.g., Ref. A.12, A.30, A.31], while procedures for assessing the damage tolerance
and redundancy of ship structures are given in SSC Reports 402 [Ref. A.32] and SSC 354 [A.33],
respectively.
Each part of this Guide builds upon the results of previous Ship Structure Committee projects
[A.32-A.41] and recent efforts of classification societies and navies [e.g., Ref. A.27]. Wherever
possible, the developers of this Guide have tried to incorporate the best features of the previous work
and to reconcile major procedural differences. In order to maintain a practical focus, the Guide has
been reviewed by designers and fabricators of ships, and their recommendations have been
incorporated. Sufficient commentary, guidance, and references have been included to make the Guide
self-contained. However, the developers of this Guide have assumed that its users will be trained in ship
structural analysis and will have at least a rudimentary knowledge of metal fatigue.

A.5 REFERENCES
[A.1] Ship Structure Committee, Proc. of Symposium on the Prevention of Fracture in Ship Structure:
Section 4 0 - Conclusions and Recommendations, March 30-31, 1995, Washington, D.C.
[A.2] Barsom, J.M. and Rolfe, S.T.: Fracture and Fatigue Control in Structures: Prentice Hall, 1987.
[A.3] Jordan, C.R. and Cochran, C.S., “In-service Performance of Structural Details”, SSC Report
266, Ship Structure Committee, 1978.
[A.4] Jordan, C.R. and Krumpen Jr., R.P., “ Design Guide for Ship Structural Details”, SSC Report
[A.5] U.K. Protection and Indemnity Club, “Analysis of Major Claims”, The United Kingdom Mutual
Steam Ship Assurance Association (Bermuda) Limited, London, England, 1993.
[A.6] Liu, D. and Thayamballi, A., “The Durability of Ships Considering Fatigue Cracking”, Journal of
Ship & Ocean Technology, Vol. 1, No. 1, 1997, pp 57-72.
[A.7] Violette, F.M., “Lloyd’s Register’s Integrated Fatigue Design Assessment System”, Lloyd’s
Register - ShipRight Strategic Research and Development, 1997.
[A.8] Liu, D. and Thayamballi, A., “Local Cracking in Ships - Causes, Consequences, and Control”,
Proc. of Symposium on the Prevention of Fracture in Ship Structure, March 30-31, 1995,
Washington, D.C., pp 125-166.
[A.9] Germanischer Lloyd, “Rules and Regulations, 1 - Ship Technology, Part 1 - Seagoing Ships,
Chapter 1 - Hull Structures, Hamburg, 1992.
[A.10] Det Norske Veritas, “Rules for Classification of Steel Ships”, DnV Classification A/S, Norway,
1991.
[A.11] Lloyd’s Register, “Rules and Regulations for the Classification of Ships: Part 3 - Ship
Structure”, 1996.
[A.12] American Bureau of Shipping, Rules for Building and Classing Steel Vessels, 1996.
[A.13] Design of Surface Ship Structures, Sea Systems Directorate Publication 23, Volume 1,UK
MoD(PE), December 1989.
[A.14] Design Procedures for Canadian Forces Surface Ships, D-03-002-008/SG-001
[A.15] Nakajima, Y, et al: “Fatigue Strength Design of Ship Structure” IHI Engineering Review April
1993
[A.16] Witmer, D.J. and Lewis, J.W., “Operational and Scientific Hull Structure Monitoring on TAPS
Trade Tankers”, presented at SNAME annual meeting, Nov. 17-18, 1994, New Orleans,
Louisiana.
[A.17] Sucharski, D., “Crude Oil Tanker Hull Structure Fracturing: An Operator’s Perspective”,
Proc. of Symposium on the Prevention of Fracture in Ship Structure, March 30-31, 1995,
Washington, D.C., pp 87-124.
[A.18] Lloyd’s Register, “HTS - Something for Nothing?”, 100A1, Issue 3, 1992.

[A.19] Skaar, K.T., Valsgard, S., Kohler, P.E., and Murer, C, “How Low Can Steel Weight Go with
Safety and Economy?”, Proc. Third Int. Conf. on “Practical Design of Ships and Mobile
Offshore Units”, PRADS’87, Trondheim, Norway, June 22-26, 1987
[A.20] Kirkhope, K. and Basu, R., “Analysis of Annapolis and Iroquois Class Structural Defects
Histories”, MSEI Final Report 1727-13-01, March 1994.
[A.21] Bea, R.G., Pollard, R., Schulte-Strauthaus, R., and Baker, R., “Structural Maintenance for
New and Existing Ships: Overview Fatigue Cracking and Repairs”, Proceedings of the
SSC/SNAME Marine Structural Inspection, Maintenance, and Monitoring Symposium, March,
1991.
[A.22] Improved Ship Structures Maintenance Management (ISSMM), Canadian Forces/Chief
Research and Development (CRAD) Project D6109.
[A.23 Sargent Jr., D.P. and Steward, P.E., “Capturing Maintenance and Modernization Efficiencies”,
Naval Engineers Journal, May, 1992.
[A.24] Environmental Protection Agency, Oil Pollution Act, 1990.
[A.25] American Bureau of Shipping, “Guide for Fatigue Strength Assessment of Tanker”, Rules for
Building and Classing Steel Vessels, Part 5, Section 2, Appendix 5/2AA, 1996.
[A.26] Lloyd’s Register, “Ship Right - Fatigue Design Assessment Procedure”, 1996.
[A.27] Det Norske Veritas, “Fatigue Assessment of Ship Structures”, Classification Notes. No. 30.7,
September 1998.
[A.28] Clarke, J.D., “Proposed Fatigue Design Criteria for Steel and Aluminum Warship Hulls”,
AMTE(S) R84103, Nov., 1984 (Restricted).
[A.29] Lloyd’s Register, “A New Approach to Naval Ship Construction”, news bulletin posted Oct. 8,
1997 on www.lt.org/news.
[A.30] Lloyd’s Register, “ Rules for the Manufacture, Testing, and Certification of Materials”, Chapter
3: Sections 2 and 3, 1984.
[A.31] Det Norske Veritas, “Rules for Classification of Steel Ships - Materials and Welding”, Part 2,
Chapter 1, 1984.
[A.32] Yee, R., Malik, L., Basu, R., and Kirkhope, K., “Guide to the Damage Tolerance Analysis of
Marine Structures”, SSC Report 402, Ship Structure Committee, Aug., 1997.
[A.33] Das, P.K. and Garside, J.F., “Structural Redundancy for Continuous and Discrete Systems”,
SSC Report 354, Ship Structure Committee, 1991.
[A.34] Cramer, E.H., Schulte-Strauthaus, R., and Bea, R.G., “Structural Maintenance Project Volume
1: Fatigue Damage Evaluation”, SSC Report 386, Ship Structure Committee, 1995.
[A.35] Bea, B., Xu, T., Ma, K., and Schulte-Stathaus, R., “Ship’s Maintenance Project”, SSC Report
[A.36] Kirkhope, K.J., Bell, R., Caron, L, and Basu, R.I., “Weld Detail Fatigue Life Improvement
Techniques”, SSC Report 400, Ship Structure Committee, 1997.
[A.37] Mansour, A., Wirshing, P., White, G., and Ayyub, B., “Probability-based Ship Design:
Implementation of Design Guidelines”, SSC Report 392, Ship Structure Committee, 1996.

[A.38] Basu, R., Kirkhope, K.J., and Srinivasan, J., “Guideline for Evaluation of Finite Elements and
Results”, SSC Report 387, Ship Structure Committee, 1996.
[A.39] Wiernicki, C., Ghose, D., and Nappi, N., “Residual Strength of Damaged Marine Structures”,
SSC Report 381, Ship Structure Committee, 1995.
[A.40] Stambaugh, K.A., Van Mater Jr., P.R., and Munse, W.H., “Fatigue Performance under Multi-
axial Load, SSC Report 356, Ship Structure Committee, 1990.
[A.41] Capanoglu, C.C., “Fatigue Technology Assessment and Strategies for Fatigue Avoidance in
Marine Structures”, SSC Report 367, Ship Structure Committee, 1993.

Part B - Catalogue of Fatigue Resistant Details
PART B - CATALOGUE OF FATIGUE RESISTANT DETAILS
For this section, the authors have drawn heavily on the work of the Classification
Societies, specifically Lloyd’s Register of Shipping, Det Norske Veritas,
American Bureau of Shipping, and the International Association of
Classification Societies (IACS), as well as Canadian and UK Navy information.
We are indebted to Lloyd’s Register of Shipping, Det Norske Veritas, the American
Bureau of Shipping, IACS, Saint John Shipbuilding Limited and the Canadian
Navy for permission to use their source materials and drawings. These resources
are identified in each of the data sheets herein. We recommend the references in
this section for those wishing to explore this area further.
B.1 INTRODUCTION
B.1.1 Background
The fatigue resistance of structural details subjected to a given cyclic load is primarily geometry
dependent. That is, discontinuous load paths, rapid geometric transitions or misalignments and poor
weld geometries are the most significant factors which influence the fatigue life of a structural detail
subjected to a given cyclic load environment. Therefore, the most effective way of extending the fatigue
life of a structural connection is through proper detailing.
The most common reason for poor fatigue resistance is inappropriate detailed design. Figures
B.1.1 and B.1.2, extracted from previous SSC Reports [Ref. B.1], show typical crack patterns that
have been found in tankers and bulk carriers, respectively. These instances of fatigue damage can been
avoided with additional attention to detailing for fatigue resistance in the design process.
Design for good fatigue resistance can proceed through a series of levels, or steps. The first
level in the fatigue design process involves the selection of fatigue-tolerant details. This section of the
Fatigue Design Guide presents a catalogue of structural details for a range of vessel types, with
suggested “good practice” to improve fatigue performance. A detail “rating” system is included in the
catalogue to indicate the relative performance and costs associated with alternative structural details.
The purpose of this catalogue is to provide the practicing naval architect with a readily accessible, guide
on how to improve the fatigue performance of a structural connection.
B.1.2 Objective
The Guide is intended to be utilized by engineers and naval architects in ship structural design
(and fabrication) to promote good design practice, reduce the likelihood of premature fatigue failure,
and identify the relative fatigue resistance of structural details.
The objective of Part B of the Guide - the Catalogue - is to provide guidance in the preliminary
design of structural connections to improve their fatigue performance. The Catalogue therefore,
provides a Level 1 fatigue design process.
Fatigue-Resistant Detail Design Guide for Ship Structures B-1

Figure B.l.l: Fatigue Crack Locations and Orientation in Typical Tanker Structure

Figure B.1.2: Fatigue Crack Locations and Orientation in Typical Bulk Carrier Structure
B.1.3 Scope
With literally hundreds of structural detail configurations in existence, this Catalogue can only
provide examples of the most common details. For these details the Catalogue provides geometric
limitations and arrangement alternatives which may be used directly to improve the fatigue performance
of ship structural details. However the Catalogue can also be used to indicate the type of improvements
that may be considered for details not presented by applying the principles that are shown herein.
The Catalogue is presented in four sections, each section providing information on a different
ship type. The four generic ship types dealt with in the catalogue are:
• Double Hull Tankers • Bulk Carriers
• Container Ships • Warships

The Catalogue focuses on details associated with the midship section of the vessels of interest,
since this is generally the most critical area. All welded connections which are potentially prone to
fatigue failure are presented. The Catalogue illustrates good detailing practice and/or alternative
configurations for those connections considered to be at high risk. The evaluation of a connection as
being at high or low risk is solely based on the geometry of the connection and thus the stress
concentration it represents. The severity of the loadings seen by any specific detail will be determined
by location-specific factors, and thus an inherently high-risk connection may or may not require
modification depending on its application. Many structural details are common to more than one ship
type. The catalogue illustrates the detail for each ship type so that the user has complete information
within each section. Similar details, used in different ship types (e.g., tanker, bulk carrier, etc.), will have
the same detail identification number. For example, a transverse floor, bottom longitudinal connection is
Detail # 2 for both tankers and bulk carriers.
The remainder of this section of the Guide provides background and guidance to the designer
on how to interpret the information contained in the Catalogue (Section B.5). This information is
presented in the following parts:
• Discussion of critical and non-critical details (Section B.2)

• Definitions of terminology used in the catalogue (Section B.3)
• A description of the catalogue layout (Section B.4)
• The Catalogue of details (Section B.5)
• Guidance on good design and construction practice (Section B.7)
• A list of the relevant reference material (Section B.8)
B.2 CRITICAL AND NON-CRITICAL DETAILS

The criticality of a structural detail is determined by assessing the risk to the continued safe
operation of the vessel, posed by failure of the detail. Risk, in most general terms, is a combined
measure of the probability (or likelihood) of failure and the consequence (or cost) of that failure. The
designer is encouraged to consider both the probability of failure (in terms of fatigue strength of the
detail) and the potential consequence of failure (which may be evaluated in terms of structural location
or function).
B.2.1 Probability of Failure (Fatigue Resistance)

The primary objective of the Catalogue is the improvement of standard structural connections.
For those midship connections considered prone to fatigue damage, design improvements or alternative
arrangements are presented. In some instances, several levels of fatigue performance improvements are
presented. Their relative fabrication and maintenance costs are rated to allow the designer to weigh the
additional costs associated with reducing the probability of fatigue failure.
Those details that are not prone to fatigue problems, due to their geometric configuration, may
be considered non-critical. The connection between the longitudinal and the bottom of a ship is an

example of a welded connection that, if good fabrication and maintenance practices are observed,
should not be considered a likely location for fatigue cracking.
B.2.2 Consequences of Failure (Structural Function)

While the Catalogue does not specifically categorize the details based on the consequence of
failure, the user is encouraged to examine the consequence of failure of the detail and associated
structure when deciding if it is worthwhile to apply the fatigue performance improvements suggested.
Since the cost of improving details can be significant, this is a most important consideration.
For example, a fatigue-prone detail in side or bottom plating warrants the cost associated with
improved design practice more than the same detail in, say, a longitudinal bulkhead. The failure of a
detail in the outer shell of a single hull tanker in way of a cargo tank can result in massive pollution, and
therefore, the consequence of failure is great. The side shell plating in double hull tankers or bulk
carriers is another critical area. The upper deck near amidships in way of large openings of warships
should be considered a critical area. With these considerations in mind, the designer should use cost-
benefit judgements to select the detail design improvements that provide appropriate levels of safety
against fatigue failure.
B.2.3 Warship Details

Structural details on warships have been developed to meet the specific requirements of shock
and impact, and, as a result, occasionally embody details that are designed to enhance the impact
resistance. An example is the doubler plate on web frame/deck beam connections. However, this
doubler plate is not necessarily an enhancement for fatigue purposes. The catalogue tries to present
structural arrangements for warships that do not compromise the integrity of the connection while
improving the fatigue performance of the detail.
B.3 CATALOGUE TERMINOLOGY

To ensure that the user understands the terminology used in the fatigue resistant design detail
catalogue, the following list of definitions is supplied.
Critical Area - That area of the ship encompassing the global structure and comprising major structural
elements such as bulkheads, decks, and in which the critical detail is included.
Critical Detail - The specific detail for which the fatigue performance is critical.
Good Practice - Detail design and fabrication practice which will improve the resistance of the detail to
fatigue crack initiation and growth.
Detail Attributes - The measures that are being used to “rank” the alternative detail configurations.
Ranking - A simple measure of the relative strength of each attribute among the detail options described.
Number 1 is always the best ranking for a detail.
Fatigue Performance - A measure of the resistance of the detail to fatigue crack initiation and growth.
The lower the number, the better the fatigue performance.

Ease of Inspection - Ease with which inspection devices (e.g., ultra-sonic probes) can access and detect
cracks, and ease with which the configuration (e.g., alignment) can be checked after
construction for detection of defects. The lower the number, the easier to inspect.

Maintenance Cost - This is a measure of the degree to which the detail can be kept clean of damaging
material, corrosive fluids, ease of painting and maintenance of an intact paint coating, as well as
access for repair of defects. The lower the number, the easier to maintain.
Fabrication Cost - Reflects material, cutting and welding content, ease of fabrication. The most
expensive detail has the highest number.
B.4 LAYOUT OF THE CATALOGUE
B.4.1 Introduction
The Catalogue is arranged in a hierarchical fashion starting with a midship section in which the
global structural location of interest is identified. From this, the connection structural elements are
selected to identify the detail of interest. The hierarchical layout of the catalogue is described in Figure
B.4.1.1 along with the nine-step process involved in making full use of its information.
Isometric Drawing of the Vessel Type Global Structure
1) Locate structural detail of interest in typical midship drawing
(See Figure B.4.2)
Fatigue Resistant Structural Detail Catalogue Index Table

2) Identify location of structural detail to select appropriate index table (e.g., double bottom,
double side, deck, transverse bulkhead, etc.)
3) Identify critical detail by connected structural elements (e.g., inner/out bottom longitudinal in
way of transverse floor, bottom girders in way of transverse floor, etc.)
4) Read off detail type number and check critical detail selection in detail figure
(See Figure B.4.3)
Fatigue Resistant Detail Data Sheets

5) Review data sheet header and/or critical area drawing to ensure appropriateness of detail
data sheet
6) Note critical detail features and identified critical locations
7) Examine fatigue life improvement alternative good detailing practice drawings
8) Select detail design good practice alternative based on detail attribute rankings(e.g., fatigue
performance, ease of inspection, maintenance cost and fabrication)
9) Review comments to identify important aspects of detail fabrication or geometry and note
original source of good practice recommendations
(See Figure B.4.4)
Figure B.4.1.1: Fatigue Design Guide Structural Detail Catalogue Layout

At the top level, the catalogue is divided into four, stand-alone sections by ship type as follows:
• Double Hull Tanker • Bulk Carrier
• Container Ship • Warship
Each vessel type specific section begins with an isometric illustration of a typical midship area of
the vessel type in question (see Figure B.4.1.2) with the principal structural elements defined. This
structural representation should aid in locating the structural detail of interest and relating it to the
terminology used to describe it in the fatigue resistant design detail catalogue.
Deck
Deck
Transverse
Deck
Longitudinal
Topside
Tank
Corrugated
Transverse
Bulkhead
Side Shell
Plating
Side
Longitudinal
Upper Stool Transverse

Web Frame
Hopper Tank
Transverse Floor
Inner Bottom Plating Lower Stool

Side Stringer
Bottom Girder
Outer Bottom Plating
Inner Bottom Longitudinal
Outer Bottom Longitudinal
Figure B.4.1.2: Typical Double Hull Tanker Midship Global Structural Arrangement

The fatigue resistant structural detail catalogue index tables, similar to that shown in Figure
B.4.1.3, group the critical structural details by location (e.g., double bottoms, double sides, deck and
transverse bulkheads). The terminology for the location names used in the catalogue index tables are
given in the global midship structure illustration (see Figure B.4.1.4) located at the beginning of each
vessel type catalogue section.
A catalogue index table is available for all significant structural locations. For the double hull tanker
structure, catalogue index tables have been assembled for:
• double bottom structure, • double side structure, and
• transverse bulkhead structure, • deck structure.
The catalogue index tables are arranged to describe all of the connections in the structural
location being described. Since a structural connection is defined as the connection of two or more
structural members, the catalogue index table identifies connection details based on the structural
elements or members joined in its first two columns.
The third and fourth columns of the catalogue index tables contain a detail type reference and a
figure illustrating the structure surrounding the detail of interest. The detail type reference number directs
the user how to find more information on developing a fatigue resistant structural detail for the
connection of interest. Since experience has shown that not all structural member connection details are
prone to fatigue damage, the detail type reference may be one of the following two forms:
1) for connections where fatigue is not typically a critical design issue, the detail type reference directs
the user to Section B.7 which discusses good fabrication and assembly practice to preclude fatigue
cracking; or,
2) for connections which are considered susceptible to fatigue damage, the detail type reference directs
the user to appropriate fatigue resistant detail data sheets, similar to that in Figure B.4.4.
While each vessel type section is completely self-contained, similar fatigue-susceptible details
used in different vessel types have been given the same detail reference number. Once a critical fatigue
detail is identified the alternative designs can be reviewed by locating the associated fatigue resistant
detail data sheet within the catalogue.
The data sheets contain information which the engineer or naval architect will use to select
appropriate detail configurations and to understand the implications of the selection. The data sheet has
five information areas including:
• a data sheet title block;
• a critical area and detail illustration;
• detail design good practice recommendations;
• a detail attribute ranking matrix; and,
• a detail specific comment section.

Critical Structural Details Detail Type# Detail Figure

Transverse Bulkhead
Transverse Floor Stiffener
Transverse Floor
- Asymmetric Longitudinal 1 Critical Detail
#3 & #4 Bottom
- Symmetric Longitudinal 2
Girder
Inner/Outer Bottom Longitudinals
Transverse Bulkhead
- Plane Oil Tight Transverse Bulkhead 3 Transverse Floor
Stiffener Outer
- Corrugated Oil Tight Transverse 4
Bottom
Critical Detail
Bulkhead Bottom
#1 & #2
Longitudinal
Critical Detail #5
Transverse Ring Web Transverse
Bottom Longitudinals - Transverse Ring Web Tripping 5 Ring Web
Bracket
Bottom
Longitudinal
Critical Detail #6 & #7

Transverse
Transverse Bulkhead Bulkhead Bottom
- Plane Oil Tight Transverse Bulkhead 6 Girder
Bottom Girders - Plane Oil Tight Transverse 7
Bulkhead Vertical Stiffener End
Brackets and High Strength Steel
Girders Outer
Bottom
Critical Detail Side Shell

#8 & #9
Hopper
Hopper Tank and Tranverse Floor Inner
Tank
Bottom
- Welded Knuckle 8
- Radiused Knuckle 9 Bottom
Outer Girder
Bottom
Non Critical Detail

Inner
Bottom Longitudinal Section B.7
Bottom
Bottom
Girder
Inner/Outer Bottom Plating Outer

Bottom Girder Section B.7 Bottom
Bottom
Longitudinal
Transverse Non Critical

Transverse Floor Section B.7 Bulkhead Detail
Transverse Bulkhead Section B.7

Transverse
Outer
Floor
Bottom
Figure B.4.1.3: Typical Fatigue Resistant Structural Detail Catalogue Index Table

Vessel Type: Double Hull Oil Tankers Detail

Critical Area: Double Bottom
Critical Detail: Intersection of Inner/Outer Bottom Longitudinal and
Detail Description: Asymmetric Longitudinal FLEET
TEC HN OLOGY LTD
1
Rev. 001-250199
CRITICAL AREA: DETAIL DESIGN GOOD PRACTICE:

Inner Bottom
Transverse
Floor Stiffener Bottom Girder Detail Design A - Soft Heel
d
r = 30 mm
Critical 55 mm
Detail
Max. 15 mm
Bottom 75 mm
Outer Bottom
Longitudinal
CRITICAL DETAIL:
A 1-001-1
Inner
Bottom Detail Design B - Soft Toe d
and Soft Heel r = 30 mm
t > d/18
r > 0.75d 55 mm
Transverse Max. 15 mm
Longitudinal Floor 75 mm
Transverse Floor
Stiffener d/2
1-001-2
Outer Detail Design C - Soft Toe

Bottom d
and Soft Backing Bracket t > d/18
r > 2.0d
A t > d/18 r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm
Section A-A Longitudinal
Inner Bottom Min. 1.5d

Min. d/2 Min. 300 mm
Transverse 1-001-3
Floor
d
d = 180 - 300 mm
Critical Design Detail
Locations
Transverse Floor Detail Attributes Rankings
Stiffener A B C
Fatigue Performance 3 2 1
Longitudinal Ease of Inspection 1 1 2
Outer Bottom Maintenance Cost 1 2 3
1-01 Fabrication Cost 1 2 3
COMMENTS:
- Alignment between the transverse floor stiffener and the backing bracket is to be maintained.
- A wrap around weld, free of undercut or notches, around the heel and toe connections of the floor stiffener and backing bracket.
Figure B.4.1.4: Typical Fatigue Resistant Detail Data Sheets

The data sheet title block includes descriptions of the vessel type, critical area, critical detail and
any further detail description data to identify the connection of interest along with the detail type
reference number. The critical area and detail illustrations are included in the data sheet to locate and
describe fatigue damage susceptible elements of the critical detail. This is accomplished with a global
structure illustration of the critical area including near by structural elements to help visualize the detail
of interest. Below the critical area illustration, the data sheet includes a close-up of the critical detail
illustrating the critical locations at which fatigue cracking is most likely to occur.
The drawings on the right side of the data sheet illustrate “good practice” recommendations to
improve the fatigue performance of the structural connection. The recommendations include both
general items such as: soft toe and heel geometries, scallop sizes and ensuring member/load path
continuity, along with detail specific suggestions such as the addition of stiffening elements or flange
tapering/chamfering geometry.
Limitations on geometry or detail critical tolerances and dimensions, are also included in the
drawings where appropriate.
Most of the data sheet “good practice” recommendations include more than one fatigue resistant
detail alternative. These alternatives are offered as stepwise improvements in the fatigue performance of
the detail, but there are costs associated with improved fatigue strength. The detail attribute ranking
matrix, at the bottom right of the data sheet, tries to qualitatively express the relative costs of the detail
improvement alternatives in terms of:
• fatigue performance, • ease of inspection, and
• maintenance cost, • fabrication cost.
While the relative rankings in the detail attribute matrix are subjective, they serve their intended purpose;
which is to highlight the additional collateral costs or burdens associated with improving fatigue
performance. The relative cost rankings could be used to help in the selection of a detail for a specific
application as discussed in Section B.3.
Located at the bottom of the data sheet is a comment block that provides detail specific
alignment or fabrication practice recommendations to improve the fatigue performance of the
connection.

B.5 FATIGUE RESISTANT DETAIL CATALOGUE
B.5.1 Tanker Structure Details
Deck
Deck
Transverse
Deck
Longitudinal
Topside
Tank
Corrugated
Transverse
Bulkhead
Side Shell
Plating
Side
Longitudinal
Upper Stool Transverse

Web Frame
Hopper Tank
Transverse Floor
Inner Bottom Plating Lower Stool

Side Stringer
Bottom Girder
Outer Bottom Plating
Inner Bottom Longitudinal
Outer Bottom Longitudinal
Figure B.5.1: Typical Midship Section Nomenclature for Double Hull Tankers

B-12
Table B.5.1 (a): Critical Structural Details for Double Bottom

Transverse Bulkhead
Transverse Floor
- Asymmetric Longitudinal 1 Critical Detail
#3 & #4 Bottom
- Symmetric Longitudinal 2
Girder
Transverse Bulkhead
- Plane Oil Tight Transverse Bulkhead 3 Transverse Floor
Stiffener Outer
- Corrugated Oil Tight Transverse 4
Bottom
Critical Detail
Bulkhead Bottom
#1 & #2
Longitudinal
Critical Detail #5
Transverse Ring Web Transverse
Bottom Longitudinals - Transverse Ring Web Tripping 5 Ring Web
Bracket
Bottom
Longitudinal
Critical Detail #6 & #7

Transverse
Transverse Bulkhead Bulkhead Bottom
- Plane Oil Tight Transverse Bulkhead 6 Girder
Bottom Girders - Plane Oil Tight Transverse 7
Bulkhead Vertical Stiffener End
Brackets and High Strength Steel
Girders Outer
Bottom
Critical Detail Side Shell

#8 & #9
Hopper
Hopper Tank and Tranverse Floor Inner
Tank
Bottom
- Welded Knuckle 8
- Radiused Knuckle 9 Bottom
Outer Girder
Bottom
Non Critical Detail

Inner
Bottom
Bottom
Girder
Inner/Outer Bottom Plating Outer

Bottom Girder Section B.7 Bottom
Bottom
Longitudinal
Transverse Non Critical

Transverse Floor Section B.7 Bulkhead Detail

Transverse
Outer
Floor
Bottom

B-13
Table B.5.1 (b): Critical Structural Details for Double Side

Critical Structural Details Detail # Detail Figure
Side Shell
Transverse Web Frame Stiffener Transverse
Web Frame
- Asymmetric Longitudinal 21
- Symmetric Longitudinal 22 Side
Critical Detail
Longitudinal
#21 & #22
Side Shell Longitudinals Transverse

Bulkhead
Transverse Ring Web
Side
Critical
- Transverse Ring Web Tripping 25 Longitudinal
Detail #26
Bracket
Transverse Bulkhead
Critical
- Transverse Bulkhead Horizontal 26 Detail #25 Hopper
Stiffener Transverse Tank
Ring Web
Transverse
Side Shell
Bulkhead
Transverse Web Frames Section B.7 Side
Stringer
Side Stringers Critical Detail Non

#27 & #28 Critical
Detail
Transverse Bulkhead
- Plane Oil Tight 27
- Corrugated Oil Tight 28 Horizontal
Girder
Non Critical Side

Hopper/Topside Tank Section B.7 Longitudinal
Detail
Side
Side Longitudinals Section B.7 Stringer
Side Shell Plating Side Stringers Section B.7
Transverse Web Frames Section B.7
Hopper
Transverse Bulkhead Section B.7 Tank
Transverse
Side Shell
Web Frame
Hopper/Topside Tanks
Side
- Welded Side Longitudinal 29
Critical Detail Stringer
Transverse Web Frames Bulkhead Plating #29 & #30
- Knuckled Side Longitudinal 30 Hopper
Bulkhead Plating Tank

B-14
Table B.5.1 (c): Critical Structural Details for Transverse Bulkhead

Transverse Side Shell
Bulkhead
Deck/Bottom Plating Section B.7 Non Critical

Stool Detail
Upper/Lower Bulkhead Stool
Topside/Hopper Tanks Section B.7

Inner Hopper
Bottom Tank

Bulkhead
Transverse Bulkhead Upper/Lower Bulkhead Stool Section B.7
Non Critical
Detail
Stool
Transverse Bulkhead Plating at Corrugations Section B.7
Table B.5.1 (d): Critical Structural Details for Deck

Topside
Deck Tank
Topside Tank Section B.7
Non Critical
Detail
Deck Plating Side Shell Plating Section B.7
Transverse Web Frame Section B.7 Side Shell

Transverse
Bulkhead
Non Critical Detail
Deck
Deck Transverse End Deck
Brackets Transverse Web Frames 61 Longitudinal
- Toe Connection
Critical
Detail #61
Deck Transverse Deck Longitudinals Section B.7 Deck

Transverse
Transverse Bulkhead

B-15

TEC HN OLOGY LTD
1
Rev. 001-250199

Inner Bottom
Transverse
d
r = 30 mm
Critical 55 mm
Detail
Max. 15 mm
Bottom 75 mm
Outer Bottom
Longitudinal
CRITICAL DETAIL:
A 1-001-1
Inner
t > d/18
r > 0.75d 55 mm
Transverse Max. 15 mm
Longitudinal Floor 75 mm
Transverse Floor
Stiffener d/2
1-001-2

Bottom d
r > 2.0d
A t > d/18 r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm
Inner Bottom Min. 1.5d

Transverse 1-001-3
Floor
d
d = 180 - 300 mm
Locations
Stiffener A B C
COMMENTS:

B-16

Detail Description: Symmetric Longitudinal FLEET
TEC HN OLOGY LTD
2
Rev. 001-250199

Inner Bottom
Transverse
d
r = 30 mm
Critical 55 mm
Detail
Max. 15 mm
Bottom 75 mm
Outer Bottom
Longitudinal
CRITICAL DETAIL:
1-002-1
A
Inner
r = 30 mm
and Soft Heel
t > d/18
r > 0.75d
55 mm
Max. 15 mm
Transverse
Longitudinal 75 mm
Floor
Transverse Floor d/2

Stiffener
1-002-2
Outer Detail Design C - Soft Toe d

Bottom
r > 2.0d
t > d/18
r > 400 mm
r > 0.75d
A
Max. 15 mm
Max. 15 mm

Min. 1.5d
Inner Bottom Min. d/2 Min. 300 mm
1-002-3
Transverse
Floor
d
d = 180 - 300 mm
Locations
Stiffener A B C
Fabrication Cost 1 2 3
1-02
COMMENTS:

B-17

Transverse Bulkhead
Detail Description: Plane Oil Tight Transverse Bulkhead FLEET
TEC HN OLOGY LTD
3
Rev. 001-250199

Transverse
Bulkhead Stiffener
Inner Bottom
Detail Design A - Soft Toe Bracket
with Edge Stiffener
Transverse
Bottom Girder
Increased Web Depth of d s
Bulkhead Watertight Floor s
Critical Stiffener and Soft Toe

Detail
Bottom
Brackets on Back Side
Outer Bottom
Longitudinal
CRITICAL DETAIL:
d r>d
Min. 400 mm
A
Oil Tight
Bulkhead Max. 15 mm
Inner
Bottom
1-003-1
Soft Toe Bracket

Water Tight
Floor
Detail Design B - d s
with Edge Stiffener
Increased Web Depth s

Longitudinal of Watertight Floor
Stiffener and Soft Toe
Outer Brackets on Both Sides
Bottom
r>d d r>d
A Min. 400 mm
Min. 400 mm
Critical Max. 15 mm
Section A-A d Locations
Flanged s
Bracket
s Note:
- Higher tensile steel bracket
1-003-2
- Bracket thickness = d/50 or 12 mm Min.
Design Detail
Critical
Locations Detail Attributes Rankings
A B
Fatigue Performance 2 1
Ease of Inspection 1 2
Maintenance Cost 1 2
Fabrication Cost 1 2
1-03
COMMENTS:
- Alignment between the floor stiffener, soft toe bracket and the web of the longitudinal is to be maintained.
- A wrap around weld, free of undercut or notches, around the floor stiffener heel and the heel and toe of the bracket connection to
the longitudinal.

B-18

Transverse Bulkhead
Detail Description: Corrugated Oil Tight Transverse Bulkhead FLEET
TEC HN OLOGY LTD
4
Supported by Horizontal Girders Rev. 001-250199

Transverse Inner Bottom
Bulkhead Stiffener Detail Design A - Increased Web Depth of Watertight Floor Stiffeners
Transverse Bottom Girder with Transverse Carlings and Soft Toe Brackets
Bulkhead C
Critical
Detail
Bottom
Outer Bottom
Longitudinal
CRITICAL DETAIL:
r = 450 mm
r = 300 mm
Lowest Transverse
A B
Horizontal Carling
Girder
r = 450 mm r = 300 mm
Corrugated
Oil Tight
Bulkhead
Critical
Locations
A B
15-25 mm
Transverse 1-004-1
Carling Section C-C
Water Tight
Critical
Floor Design Detail
Locations
Detail Attributes Rankings
A Original Detail
1-04
Section A-A Section B-B Maintenance Cost 2 1
COMMENTS:
- Due to the complexity of the connection, this detail warrants further detailed analysis.
- Alignment between the longitudinal diaphragm, the soft toe backing brackets, the web of the longitudinals, and the flange of the
corrugated bulkhead to the transverse carling is to be maintained.
- A wrap around weld, free of undercut or notches, around the toe of the bracket connection to the longitudinal.

B-19

Critical Detail: Intersection of Bottom Longitudinal and
Transverse Ring Web
Detail Description: Transverse Ring Web Tripping Bracket FLEET
TEC HN OLOGY LTD
5
Rev. 001-260698
Transverse Ring Detail Design A - Soft Heel

Web Frame
r = 30 mm
55 mm
Critical
Detail 75 mm
Hopper Max. 15 mm
Tank
1-005-1
Bottom
Longitudinal
CRITICAL DETAIL:
Detail Design B - Soft Toe
and Soft Heel
r = 30 mm
Hopper Ring r >= 300 mm
Face Plate 55 mm
A 75 mm
Max. 15 mm x
1-005-2
Flat Bar
Tripping Bracket
Detail Design C - Soft Toe

and Soft Backing Bracket
r >= 2x/3
r >= 400 mm
r >= 300 mm
A
Max. 15 mm
Max. 15 mm x
Min. x/2
Min. 300 mm 1-005-3
Section A-A
s
Tripping
Bracket Flange
Tripping
Design Detail
Bracket
Critical A B C
Locations
s Ease of Inspection 1 1 2
Maintenance Cost 1 2 3
COMMENTS:
- Alignment between the tripping bracket, the backing bracket and the web of the bottom longitudinal is to be maintained.
- A wrap around weld, free of undercut or notches, around the heel and toe of the bracket connection to the longitudinal face plate.
Ref: Lloyd's Register

B-20

Critical Detail: Intersection of Bottom Girder and Transverse
Bulkhead
TEC HN OLOGY LTD
6
Rev. 001-250199

Critical
Transverse Detail
Bulkhead
Bottom
Girder
Detail Design A - Soft Toe
Deep Penetration
Outer Welding with
Bottom Root R < 1/3t
θ < 15 ο
1:3 Chamfer (where face plate t
CRITICAL DETAIL: thickness > 25 mm)
θ
15 - 25 mm R < 1/3t
Taper 1:6 (approx.)

Critical
Locations
50 - 100 mm
1-006-1
Detail Design B - Parallel Toe Deep Penetration

Plane Oil Tight Welding with
Bulkhead θ=0 ο Root R < 1/3t
Lowest
1:3 Chamfer (where face plate
Horizontal t
thickness > 25 mm)
Girder
Critical 15 - 25 mm
Locations R < 1/3t
Critical
Location
100 - 150 mm
1-006-2
1-06
Inner
Design Detail
Outer
Bottom Bottom Detail Attributes Rankings
A B
COMMENTS:
- The nominal distance between the centers of the bracket toe thickness and bottom girder web thickness should not exceed 1/3
of the inner bottom thickness.
- Extent of full penetration is 2-3 longitudinal spacings, or back to the first bracket web stiffener. A wrap around weld, free of
undercut or notches, in way of the toe of the bracket connection to the inner bottom plating.

B-21

Critical Detail: Intersection of Bottom Girder and Transverse Plane
Oil Tight Bulkhead
Detail Description: Bulkhead Vertical Stiffener End Brackets and FLEET
TEC HN OLOGY LTD
7
Higher Tensile Steel Bottom Girders Rev. 001-250199

Critical
Transverse Detail
Bulkhead Detail Design A - Soft Toe
Bottom
Girder
Outer
Bottom
CRITICAL DETAIL:
1-007-1
Detail Design B - Integral Soft Toe

Bracket, Increased Web Thickness and
Plane Oil Tight Elimination of Scallop
Bulkhead
Integral Soft
Toe Bracket
1 : 3 Taper Increased Web

in Thickness Thickness
Critical
Locations
Inner
Bottom
1-007-2
Design Detail
1-07
Outer
A B
Bottom Fatigue Performance 2 1
COMMENTS:
- The nominal distance between the centers of the bracket toe thickness and bottom girder web thickness should not exceed 1/3
of the inner bottom thickness.

B-22

Critical Detail: Intersection of Inner Bottom Plating, Hopper Tank
and Transverse Floor
Detail Description: Welded Knuckle Connection FLEET
TEC HN OLOGY LTD
8
Rev. 001-250199

Critical Side Shell
Detail Detail Design A - Elimination of Scallops and Extension of Inner
Inner Bottom
Bottom Hopper
Tank No scallop. Full penetration
weld connection of floor and
Outer web to inner bottom girder and
Bottom hopper
C C
CRITICAL DETAIL:
Only Below
WB Holds
Hopper Transverse
A B 1-008-1
Ring Frames
Critical Longitudinal
Locations Section C-C Inner Girder Hopper
Bottom 800 mm Transverse Ring
Web
Floor Scarfing Brackets

Inner
Bottom (Gusset Plate)
Transverse
Floor Detail Design B - Detail Design A with Additional Intermediate
Outermost Girder
Bracket Arrangement
A B
Section B-B
Outer Bottom
Inner
Bottom
Section A-A Intermediate Brackets

Between Floors if
Critical Spacing > 2.5m
Longitudinal
Location Girder
Inner Bottom
Hopper 1-008-2
Transverse
Ring Web
Frame
Floor
Design Detail
A B
1-08
Longitudinal Outer Bottom
Girder Maintenance Cost 1 2
COMMENTS:
- Alignment between the transverse floor and the hopper transverse ring web and between sloping plating and hopper side girder
is to be maintained. Maximum misalignment is to be less than t/3 where t is the thinner of the webs to be aligned and misalignment
is the overhang of the thinner thickness.
- Full penetration welding for the connections near the knuckle between double bottom floor to inner
bottom and girder, inner bottom to hopper and transverse ring web to hopper and girder and where full collars are fitted.

B-23

Critical Detail: Intersection of Inner Bottom Plating, Hopper Tank
and Transverse Floor
Detail Description: Radiused Knuckle Connection FLEET
TEC HN OLOGY LTD
9
Rev. 001-250199

Critical Side Shell
Detail
Inner
Detail Design A - Elimination of Scallops
Bottom Hopper
Tank
Distance from side
girder to centre of
Outer
knuckle > 50 mm
Bottom
or < 70 mm
CRITICAL DETAIL: Only Below

WB Holds Full Penetration Weld
Hopper Transverse
A B Connection of
Ring Frames
No Scallop Transverse Ring Web
Critical to Inner Bottom
Locations
1-009-1
Inner
Bottom Detail Design B - Detail Design A with Additional Intermediate
Bracket Arrangement
Transverse
Floor
Outermost Girder
Inner Bottom
A B
Outer Bottom
Intermediate Brackets
Section A-A between Floors if
Critical Spacing > 2.5 m
Longitudinal
Location
Girder
Inner Bottom
Hopper
Transverse
Ring Web 1-009-2
Frame
Floor
Design Detail
A B
1-09
Girder Ease of Inspection 1 2
COMMENTS:
- Alignment between double bottom floor and the hopper transverse ring web. Maximum misalignment is to be less than t/3
where t is the thinner of the webs to be aligned and misalignment is the overhang of the thinner thickness.
- Full penetration welding for the connections near the knuckle between inner bottom and transverse ring web.

B-24

Critical Area: Double Side
Critical Detail: Intersection of Side Shell Longitudinal and
Transverse Web Frame Stiffener
TEC HN OLOGY LTD
21
Rev. 001-250199

Web Frame Detail Design A - Soft Heel
Transverse Side
d
Web Frame Stringer
Stiffener
r = 30 mm
Critical 55 mm
Side
Detail Longitudinal Max. 15 mm
75 mm
CRITICAL DETAIL:
1-021-1
Side Longitudinal
Side Shell Bulkhead
Side Shell Detail Design B - Soft Toe
Longitudinal and Soft Heel d
r = 30 mm
t > d/18
r > 0.75d 55 mm
Max. 15 mm
75 mm
d/2
1-021-2

d
t > d/18
r > 2.0d
t > d/18
r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm
Critical
Locations Min. 1.5d
1-021-3
d = 180 - 300 mm
Design Detail
Transverse Web 1-21 A B C
Frame Stiffener
Ease of Inspection 1 1 2
COMMENTS:
- Alignment between the transverse web frame stiffener, the backing bracket and the web of the longitudinal is to be maintained.
- A wrap around weld, free of undercut or notches, around the heel and toe connections of the stiffeners and the backing bracket
to the longitudinal face plate.

B-25

TEC HN OLOGY LTD
22
Rev. 001-250199

Web Frame Detail Design A - Soft Heel
Transverse Side d
Web Frame Stringer
Stiffener
r = 30 mm
Critical 55 mm
Side
Detail Longitudinal Max. 15 mm
75 mm
CRITICAL DETAIL:
1-022-1
Side Longitudinal
Side Shell Bulkhead
Side Shell d
Longitudinal and Soft Heel r = 30 mm
t > d/18
r > 0.75d
55 mm
Max. 15 mm
75 mm
d/2
1-022-2

d
t > d/18
r > 2.0d
t > d/18
r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm
Critical
Locations Min. 1.5d
d = 180 - 300 mm 1-022-3
Design Detail
1-22
Transverse Web A B C
Frame Stiffener
COMMENTS:

B-26

Transverse Ring Web
Detail Description: Transverse Ring Web Tripping Bracket FLEET
TEC HN OLOGY LTD
25
Rev. 001-250199

Side Shell
Critical
Detail Design A - Soft Heel
Detail
r = 30 mm
Transverse Ring 55 mm
Web Frame Side Shell
Longitudinal 75 mm
Hopper Max. 15 mm
Tank
1-025-1
CRITICAL DETAIL:
and Soft Heel
r = 30 mm
r >= 300 mm
Side Shell 55 mm
75 mm
Max. 15 mm x
1-025-2
Tripping Bracket

A A r >= 2x/3
Flat Bar
r >= 400 mm
r >= 300 mm
Hopper Ring
Face Plate Max. 15 mm
Max. 15 mm x
Min. x/2
Critical Min. 300 mm 1-025-3
Locations
Tripping
Bracket
s
Design Detail
s
1-25 A B C
Tripping
Bracket Flange Fatigue Performance 3 2 1
Section A-A Maintenance Cost 1 2 3
COMMENTS:
- Alignment between the tripping bracket, the backing bracket and the web of the side longitudinal is to be maintained.
- A wrap around weld, free of undercut or notches, around the heel and toe of the bracket connection to the longitudinal face plate.

B-27

Transverse Bulkhead
Detail Description: Transverse Bulkhead Horizontal Stiffener FLEET
TEC HN OLOGY LTD
26
Rev. 001-020299
CRITICAL AREA: Side Shell

DETAIL DESIGN GOOD PRACTICE:
Critical
Detail Detail Design A -
Soft Toe
Side
Longitudinal
Transverse
Bulkhead d
s
CRITICAL DETAIL:
X > 1.0d
r > 1.5d
1-026-1
Side Longitudinal
Side Shell Bulkhead Max. 15 mm
Detail Design B - Soft

t = d/22 or
Toe and Soft Backing Max. 15 mm 12 mm Min.
Bkt Bkt
Bracket
A A
r > 2.0d X > 1.5d
Bkt Bkt Min. 400 mm Min. 300 mm
d
s
Transverse Deep Transverse Oil
Tank Bulkhead in Tight Bulkhead
Wing Ballast Tank X > 1.0d
r > 1.5d
Critical
Locations
1-026-2
Max. 15 mm
d = 150 - 350 mm
s s
Design Detail
1-26 Detail Attributes Rankings
A B
Section A-A Fatigue Performance 2 1
COMMENTS:
- Alignment between the web of the bulkhead horizontal stiffener, the soft toe brackets and the web of the side longitudinal is to
be maintained.
- A wrap around weld, free of undercut or notches, around the heel connections of bulkhead stiffener and toe connection of brackets.

B-28

Critical Detail: Intersection of Side stringer and Transverse
Bulkhead Horizontal Girder
TEC HN OLOGY LTD
27
Rev. 001-250199
CRITICAL AREA: DETAIL DESIGN GOOD PRACTICE: Deep Penetration

Welding with Root
Side Shell Critical Location 1: Detail Design A - Soft Toe R < 1/3t
Plane
Transverse Side t
thickness > 25 mm) θ < 15ο
Bulkhead Stringer
Horizontal
Girder θ
Critical 15 - 25 mm R < 1/3t
Detail
50 - 100 mm
CRITICAL DETAIL: 1-027-1
Critical Location 1: Detail Design B - Parallel Toe Deep Penetration

Welding with Root
Side Stringer ο
θ=0 R < 1/3t
Critical
Location #1 1:3 Chamfer (where face plate
t
thickness > 25 mm)
15 - 25 mm
R < 1/3t
Horizontal
Girder
100 - 150 mm
1-027-2
Critical Location 2: Detail Design A - Removal of Scallop
1-027-3
Horizontal Girder
Side Longitudinal
Bulkhead
Transverse Oil Design Detail

Tight Bulkhead
Critical
Location #2 A B
1-27 Ease of Inspection 1 2
Side Shell
Side Longitudinal Maintenance Cost 1 2
Bulkhead Fabrication Cost 1 2
COMMENTS:
- Nominal distance between the centers of thickness of two abutting members should not exceed 1/3 of table member thickness.
- Deep penetration welding for connection of toe brackets in Location 1. Fillet welding for side stringer webs in way of bracket
toes in Location 1 and all connection in Location 2.
- A wrap around weld, free of undercut or notches, around the plate thickness.

B-29

Critical Detail: Intersection of Side stringer and Transverse
Bulkhead Horizontal Girder
Detail Description: Corrugated Oil Tight Transverse Bulkhead FLEET
TEC HN OLOGY LTD
28
Rev. 001-250199

Side Shell
Corrugated
Transverse Side
Bulkhead Stringer
Deep Penetration
Horizontal Welding with
Girder Root R < 1/3t
Critical θ < 15ο
Detail 1:3 Chamfer (where face plate
t
thickness > 25 mm)
θ
15 - 25 mm
CRITICAL DETAIL: R < 1/3t
50 - 100 mm
Side Stringer
1-028-1
Critical
Location
Detail Design B - Parallel Toe Deep Penetration
Welding with
θ=0 ο Root R < 1/3t

Horizontal t
thickness > 25 mm)
Girder
15 - 25 mm
R < 1/3t
100 - 150 mm
1-028-2
1-28
Corrugated
Transverse Oil
Tight Bulkhead
Critical
Location
Design Detail
A B
Side Longitudinal Fatigue Performance 2 1
Side Shell Bulkhead
COMMENTS:
- Nominal distance between the centers of bracket toe thickness and side stringer web thickness should not exceed 1/3 of side
longitudinal bulkhead thickness.
- Deep penetration welding for connection of toe brackets to side longitudinal bulkhead. Fillet welding for side stringer webs in way
of bracket toes to side longitudinal bulkhead. A wrap around weld, free of undercut or notches, around the plate thickness.

B-30

Critical Detail: Intersection of Transverse Web Frame and
Hopper/Topside Tank
Detail Description: Welded Side Longitudinal Bulkhead and Hopper FLEET
TEC HN OLOGY LTD
29
Sloping Plating Rev. 001-020299

Transverse Detail Design A - Elimination of Scallops
Web Frame Side Shell
Side
Critical Full Penetration
Stringer
Detail Welding
Hopper
Tank No Scallop
CRITICAL DETAIL: Deep Penetration

Welding
A 1-029-1
Side
Longitudinal
Bulkhead Detail Design B - Detail Design A with Additional Intermediate
Transverse Brackets
Transverse
Web
Critical
Location
Hopper
Tank Side
Stringer
A
Transverse
Brackets
Section A-A Transverse Web 1-029-2

800 mm 1-029-3
View A
Design Detail
A B
Side Stringer Fatigue Performance 2 1
Critical
Location
Hopper Tank
Transverse Ring Fabrication Cost 1 2
1-29
COMMENTS:
- Nominal distance between centers of thickness of the two abutting members should not exceed 1/3 of the table member thickness.
- Full penetration welding for the connection of the hopper sloping plating to side longitudinal bulkhead and connection of side
stringer to side longitudinal bulkhead. Deep penetration welding for connection of transverse web to side longitudinal bulkhead
and to side stringer, hopper transverse webs to sloping plating, to side longitudinal bulkhead and to side stringers in way of
hopper corners.

B-31

Critical Detail: Intersection of Transverse Web Frame and
Hopper/Topside Tank
Detail Description: Knuckled Side Longitudinal Bulkhead and Hopper FLEET
TEC HN OLOGY LTD
30
Sloping Plating Rev. 001-250199

Transverse
Web Frame Side Shell Detail Design A - Removal of Detail Design B - Detail Design
Scallops and Closer Knuckle A with Additional Longitudinal
Side
Critical
Stringer Distance from Side Girder Transverse Brackets
Detail
Hopper B
Tank Deep Penetration
Welding
Knuckle Line Longitudinal

50 - 70 mm Bracket
CRITICAL DETAIL:
Transverse
Side Longitudinal
Bulkhead Bracket
Side Shell
A
Critical Transverse Web

Location Longitudinal Bracket
r = 100 - 150 mm Section D-D

Knuckle Line 180 mm
Side Stringer 100 - 150 mm Knuckle Line
D D
Longitudinal
Bracket
800 mm 800 mm
Hopper Tank
Transverse
Brackets 1-030-1
Transverse Web
View A View B
Critical
Side Stringer
Location Design Detail
A B
Hopper Tank Fatigue Performance 2 1
Transverse Ring
Knuckle Maintenance Cost 1 2
Line
1-30
COMMENTS:
- Nominal distance between centers of thickness of the two abutting members should not exceed 1/3 of the table member thickness.
- Deep penetration welding for the connection of side stringers to side longitudinal bulkhead, connection of transverse web to side
longitudinal bulkhead and to side stringer, connection of hopper transverse web to sloped side longitudinal bulkhead and to side
stringer in way of hopper corners.

B-32

Critical Area: Deck
Critical Detail: Intersection of Deck Transverse End Brackets
and Transverse Web Frames
Detail Description: Toe Connection FLEET
TEC HN OLOGY LTD
61
Rev. 001-250199

Detail Design - Modification from Lapped Collars to Slit Type
Deck Collars, Removal of Scallops and Improved Bracket Toe
Critical No Scallops
Transverse Detail
Bulkhead
Deck Soft or
Transverse Parallel Toe
CRITICAL DETAIL:
Deck Slit Type
Collar
Deep Penetration
1-061-1 Welding with
Root R < 1/3t
Detail Design A - 1:3 Chamfer (where face plate t

thickness > 25 mm) θ < 15ο
Soft Toe
θ
15 - 25 mm R < 1/3t
50 - 100 mm 1-061-2
Deck
Transverse
Critical Detail Design B - 1:3 Chamfer (where face plate
Location t
thickness > 25 mm)
Parallel Toe
ο
Side Longitudinal θ=0
Bulkhead 15 - 25 mm
R < 1/3t
100 - 150 mm
1-061-3
Critical
Locations
Design Detail
A B
1-61 Fatigue Performance 2 1
Side Maintenance Cost 1 2
Shell Transverse Web Fabrication Cost 1 2
Frame
COMMENTS:
- Nominal distance between centers of bracket toe thickness and transverse web thickness should not exceed 1/3 of the side
longitudinal bulkhead thickness.
- Deep penetration welding for the connection of bracket toes to side longitudinal bulkhead. Fillet weldin for the connection of
transverse webs in way of bracket toes to side longitudinal bulkhead.

B-33
B.5.2 Bulk Carrier Structure Details
Deck
Hatch Coaming
Topside
Tank Topside Tank
Transverse Ring
Web
Bulkhead Side
Upper Stool Shell
Side Shell
Transverse Frame Hopper Tank
Bulkhead Transverse Ring
Transverse Side Shell Web
Floor Frame End Hopper
Bracket Tank
Bulkhead
Lower Stool Bottom
Girder
Bottom
Inner Longitudinal
Bottom Outer
Bottom
Figure B.5.2: Typical Midship Section Nomenclature for Bulk Carriers


Transverse
Bulkhead Transverse
Floor
Transverse Floor Stiffener 1
Non Critical Outer

Transverse Bulkhead Section B.7 Detail Bottom
Critical Bottom
Detail #1 Longitudina
Transverse Transverse
Bulkhead Floor
Bottom
Transverse Floor Section B.7 Girder
Bottom Girders
Transverse Bulkhead Section B.7 Outer

Bottom Non Critical
Detail
Inner Hopper
Hopper Tank Bottom Tank
- Welded Knuckle Connection 8
- Radiused Knuckle Connection 9
Outer
Bottom Critical Detail
#8 & #9
Inner
Bottom
Bottom
Girder
Inner/Outer Bottom Plating Bottom Girder Section B.7

Outer Non Critical
Bottom
Bottom Detail
Longitudinal
Transverse
Bulkhead Non Critical
Detail
Transverse Floor Section B.7

Transverse Outer
Floor Bottom

Table B.5.2 (b): Critical Structural Details for Side Shell (includes hopper and topside tanks)
Side Shell
Frame Critical
Detail #21
Side Shell Frame End 21 Hopper
Tank
Brackets
Sloping Critical
Hopper/Topside Tank Plating Detail #22
Sloping Plating
End
Side Shell Frames 22 Bracket
Topside
Transverse
Tank
Ring Web
Critical
Hopper/Topside Tank Longitudinals Detail #23
Topside Tank
Hopper/Topside Tank Longitudinal
Transverse Ring Web Flat Side Shell Side Shell
Frame
Bar Stiffener Hopper Tank
Bottom/Deck Longitudinals 23 Sloping Plating
Side Shell
Longitudinal
Bottom Critical
Side Shell Longitudinals Detail #23
Longitudinal
Topside Transverse
Tank
Ring Web
Critical
Hopper/Topside Tank Longitudinals Detail #24
Topside Tank
Hopper/Topside Tank Longitudinal
Transverse Ring Web Side Shell Side Shell
Frame
Tripping Bracket Hopper Tank
Bottom/Deck Longitudinals 24 Sloping Plating
Side Shell
Longitudinal
Bottom Critical
Side Shell Longitudinals Detail #24
Longitudinal
Side Shell
Frame
End Brackets Section B.7 End
Bracket
Side Shell Frame
Non Critical
Detail
Side Shell Plating Section B.7

Table B.5.2 (c): Critical Structural Details for Transverse Bulkhead

Critical
Inner/Outer Bottom Detail #43
Longitudinals 41 Shelf Plate Bulkhead
Stool
Critical
Lower Bulkhead Stool Inner Bottom 42 Detail #41 Critical
Detail #42
Transverse Bulkhead Shelf Plate 43

Inner Bottom
Critical
Topside
Detail #44
Tank
Transverse Bulkhead Topside Tank and 44
Upper Stool
Bulkhead
Table B.5.2 (d): Critical Structural Details for Deck

Deck
Hatch
Topside Tank Section B.7 Coaming
Non Critical
Detail
End Bracket
Deck Plating Transverse Web Frame Section B.7
Non Critical
Detail
Longitudinal Hatch 61
Coaming End Bracket Critical Detail #61 Transverse
Web Frame
Deck Hatch
Coaming
Deck
Deck Transverse Stiffener Deck Longitudinals 62 Longitudinal
Critical
Detail #63
Critical
Deck Hatch Corners/End Topside Tank Tranverse 63 Detail #62
Beams Ring Web
Topside
Transverse
Ring Web

Vessel Type: Bulk Carriers Detail

Critical Detail:
Detail Description:
Intersection of Inner/Outer Bottom Longitudinal and
FLEET
TEC HN OLOGY LTD
1
Rev. 001-260199
CRITICAL AREA: Bottom DETAIL DESIGN GOOD PRACTICE:

Girder
Transverse
Floor
Stiffener
Bottom Detail Design A - Soft Heel
Longitudinal d
Transverse r = 30 mm
55 mm
Floor
Outer
Critical Bottom Max. 15 mm
75 mm
Detail
CRITICAL DETAIL:
A 2-001-1
Critical
Inner Locations

Transverse t > d/18
Floor r > 0.75d 55 mm
d d = 180 - 300 mm Max. 15 mm

Longitudinal 75 mm
Transverse Floor
Stiffener
d/2
Outer 2-001-2
Bottom
Critical
Locations
A d
r > 2.0d
t > d/18 r > 400 mm
r > 0.75d
Max. 15 mm
Section A-A Max. 15 mm
Inner Bottom
Longitudinal
Min. 1.5d
2-001-3
Transverse
Floor
Critical
Locations
Design Detail
A B C
2-01
Transverse Floor
Stiffener Longitudinal Ease of Inspection 1 1 2
COMMENTS:
- Alignment between the longitudinal stiffener web and transverse floor stiffener and the backing bracket is to be maintained.


Critical Detail: Intersection of Inner/Outer Bottom Plating and
Hopper Tank
Detail Description: Welded Knuckle Connection FLEET
TEC HN OLOGY LTD
8
Rev. 001-260199
Side Shell Frame

Detail Design A - Elimination of Scallops and Extension of Inner
Side Shell
Bottom
Inner Bottom
Hopper Tank
No scallop. Full penetration
weld connection of floor and
Outer Bottom
web to inner bottom girder and
Critical Detail hopper
C C
CRITICAL DETAIL:
Only Below
WB Holds
Hopper Transverse
A B 2-008-1
Ring Frames
Locations
Section C-C Inner Girder Hopper
Bottom 800 mm Transverse Ring
Web
Floor Scarfing Brackets

Inner
(Gusset Plate)
Bottom
Transverse
Floor Detail Design B - Detail Design A with Additional Intermediate
Outermost Girder Bracket Arrangement
A B Section B-B
Outer Bottom Inner
Bottom
Section A-A Between Floors if
Spacing > 2.5m
Location Girder
Inner Bottom
Hopper 2-008-2
Transverse
Ring Web
Frame
Floor
Design Detail
A B
2-08
COMMENTS:
- Alignment between the transverse floor and the hopper transverse ring web and between sloping plating and hopper side girder
is to be maintained. Maximum misalignment is to be less than t/3 where t is the thinner of the webs to be aligned and misalignment
is the overhang of the thinner thickness.
- Full penetration welding for the connections near the knuckle between double bottom floor to inner
bottom and girder, inner bottom to hopper and transverse ring web to hopper and girder and where full collars are fitted.


Critical Detail: Intersection of Inner/Outer Bottom Plating and
Hopper Tank
Detail Description: Radiused Knuckle Connection FLEET
TEC HN OLOGY LTD
9
Rev. 001-260199
Side Shell Frame Side Shell Detail Design A - Elimination of Scallops

Inner Bottom
Hopper Tank
Distance from side
Outer Bottom girder to centre of
knuckle > 50 mm
Critical Detail or < 70 mm
CRITICAL DETAIL:
Only Below Full Penetration Weld
WB Holds Connection of
Hopper Transverse No Scallop Transverse Ring Web
A B
Ring Frames to Inner Bottom
Critical 2-009-1
Locations
Detail Design B - Detail Design A with Additional Intermediate

Inner
Bottom Bracket Arrangement
Transverse
Floor Inner Bottom
Outermost Girder
A B
Outer Bottom
between Floors if
Spacing > 2.5 m
Critical Girder
Location
Inner Bottom
Hopper
2-009-2
Transverse
Ring Web
Frame
Floor
Design Detail
A B
2-09
COMMENTS:
- Alignment between double bottom floor and the hopper transverse ring web. Maximum misalignment is to be less than t/3
where t is the thinner of the webs to be aligned and misalignment is the overhang of the thinner thickness.
- Full penetration welding for the connections near the knuckle between inner bottom and transverse ring web.


Critical Area: Side Shell
Critical Detail: Intersection of Hopper/Topside Tank Sloping
Plating and Side Shell Frame End Brackets
Detail Description: Hopper Tank Backing Bracket FLEET
TEC HN OLOGY LTD
21
Rev. 001-260199

Side Shell
Detail Design A - Elimination of
Side Shell
Frame Critical Detail Scallop and Additional Carling
Frame Lower
Hopper Tank Bracket
Sloping Plating
Side Shell Frame

No Scallop S
End Bracket
S
CRITICAL DETAIL:
2-021-1
Detail Design C - Detail Design A

Detail Design B - Full Bracket and B with Full Collar
Frame Lower Frame Lower
Bracket Bracket
Frame Lower
Full Collar
Bracket
S
S
S S
2-021-2 2-021-3
Critical
Location Design Detail
Hopper A B C
Tank
2-21 Maintenance Cost 2 1 3
COMMENTS:
- Alignment between frame bracket and supporting bracket is to be maintained. Maximum misalignment is to be less than t/3 where
t is the thinner of the webs to be aligned and misalignment is overhang of the thinner thickness.
- Ensure start and stop of welding is far away as practicable from the unsupported edge corners in the supporting brackets.


Critical Detail: Intersection of Hopper/Topside Tank Sloping
Plating and Side Shell Frame
Detail Description: Hold Brackets FLEET
TEC HN OLOGY LTD
22
Rev. 001-260199

Side Shell
Side Shell Detail Design A - Soft Toe

Frame Critical Detail
Suggested:
X View X
Hopper Tank r = 10tw
Sloping Plating x = 2/3 r
y = 1/2 r Face Plate
Side Shell Frame

End Bracket
Chamfer 1:3 Chamfer 1:3
Frame Bracket
Web
y tw
10 mm
r Ring Web
Max Taper 1:3
CRITICAL DETAIL: x Recommended 1:5
Below Hopper
when HT Steel is
used
15 - 25 mm 2-022-1
Detail Design B - Extended Toe

Topside
Tank Suggested:
X
y = 5tw View X
ymax = 80 mm
x > 3y
Critical
Location Taper 1:5
Frame Upper Chamfer 1:3

Bracket 5 Chamfer 1:3
Frame Bracket
Web
10 mm y 1
tw
Ring Web
Side Shell
Below Hopper
x
Frame Lower 2-022-2
Bracket
15 - 25 mm
Critical
Location
Design Detail
A B
S
S
Hopper Maintenance Cost 1 1
Tank
2-22
COMMENTS:
- Alignment between side shell frame lower and upper brackets and transverse ring webs or supporting brackets is to be maintained.
Maximum misalignment is to be less than t/3 where t is the thinner of the webs to be aligned and misalignment is overhang
of the thinner thickness.
- Ensure start and stop of welding is far away as practicable from the toe of the frame brackets. A wrap around weld, free from
undercut and notches, around the toe of the end bracket connections to hopper plating.


Critical Detail: Intersection of Hopper/Topside Tank Transverse
Web Flat Bar Stiffener and Hopper/Topside tank,
Bottom/Deck, and Side Shell Longitudinal FLEET
TEC HN OLOGY LTD
23
Detail Description: Rev. 001-260199

Topside Transverse
Tank Ring Web Detail Design A - Soft Heel
Topside Tank
Longitudinal d
Critical Detail
Side Shell Side Shell r = 30 mm
Frame 55 mm
Hopper Tank Max. 15 mm

Sloping Plating 75 mm
Side Shell
Longitudinal Critical Detail
Bottom 2-023-1
Longitudinal

and Soft Heel d
r = 30 mm
CRITICAL DETAIL: t > d/18
r > 0.75d
55 mm
Max. 15 mm
75 mm
Critical d
Locations
d/2
2-023-2

Section A-A and Soft Backing Bracket
d
t > d/18
r > 2.0d
t > d/18
r > 400 mm
r > 0.75d
A Max. 15 mm
Max. 15 mm
A Min. 1.5d
2-023-3
A A
Design Detail
DetailAttributes Rankings
A
A B C
2-23
A Ease of Inspection 1 1 2
COMMENTS:
- Alignment between longitudinal web stiffener, transverse ring web stiffener and backing bracket is to be maintained.
- Ensure start and stop of welding is far away as practicable from the heel and toe. A wrap around weld, free from undercut and
notches, around the heel and toe connections of stiffener and bracket to longitudinal.


Critical Detail: Intersection of Hopper/Topside Tank Transverse
Web Tripping Bracket and Hopper/Topside tank,
Bottom/Deck, and Side Shell Longitudinal FLEET
TEC HN OLOGY LTD
24
Detail Description: Rev. 001-260199

Topside Transverse
Tank Ring Web Detail Design A - Soft Heel
Topside Tank
Longitudinal
Critical Detail 55 mm
r = 30 mm
Side Shell Side Shell
Max. 15 mm
Frame
75 mm
Hopper Tank
Sloping Plating
Side Shell
Longitudinal Critical Detail
Bottom 2-024-1
Longitudinal
and Soft Heel
CRITICAL DETAIL: R > 300 mm

55 mm
r = 30 mm
Max. 15 mm
Max. 15 mm
75 mm
Critical
Locations
2-024-2

Section A-A
R > 2X/3
R > 300 mm R > 400 mm
A
Max. 15 mm Max. 15 mm
X Min. X/2
Min. 300 mm
2-024-3
A A
Design Detail
A
A B C
2-24
A Ease of Inspection 1 1 2
COMMENTS:
- Alignment between longitudinal web stiffener, transverse ring web tripping bracket and backing bracket is to be maintained.
- Ensure start and stop of welding is far away as practicable from the heel and toe. A wrap around weld, free from undercut and
notches, around the heel and toe connections of brackets.


Critical Area: Transverse Bulkhead
Critical Detail:
Detail Description:
Intersection of Lower Bulkhead Stool and
Inner/Outer Bottom Longitudinal
FLEET
TEC HN OLOGY LTD
41
Rev. 001-260199

Lower Stool
Lower Brackets in line

Critical Detail Inner
Stool with Stiffener Web
Bottom
Inner Bottom
Outer Bottom 1.0d 1.0d
d d
CRITICAL DETAIL: 1.0d 1.0d

A
Critical
2-041-1
Locations Inner Bottom
Detail Design C - Single Plate

Integral Brackets, Additional
Full Collars Stability Stiffener and Soft
Floor
WT Floor
d Critical d
Floor d Bracket
Locations 1.0d
t > d/18
R > 2.0d
Full Collars
R > 400 mm
t > d/18
Floor
R > 1.5d
d
Max. 15 mm Max. 15 mm
Outer Bottom
A
Section A-A Sloping
Stool Plate
Min. 1.0d Min. 1.5d
Min. 300 mm Min. 400 mm
Inner Bottom
1.0d
2-041-2
Ring
Web
Design Detail
A B C
Longitudinal
Floor
Girder
Outer Bottom 2-41
COMMENTS:
- Alignment between brackets and stiffener webs is to be maintained.
Maximum misalignment is to be less than t/3 where t is the thinner of the webs to be aligned and misalignment is overhang
- Fillet welding between inner bottom and floors.
- Ensure start and stop of welding is far away as practicable from the toe of brackets or corners. A wrap around weld, free from
undercut and notches, around the toe connections of the stiffeners and backing brackets to longitudinal.


Critical Detail:
Detail Description:
Intersection of Lower Bulkhead Stool and Inner
Bottom Girder
FLEET
TEC HN OLOGY LTD
42
Rev. 001-260199

Detail Design
B
Lower Stool
Lower
Stool
Critical Detail Inner
Bottom
No Scallops -
Inner Bottom
Full Penetration Floor
Longitudinal
Welding
Girder
CRITICAL DETAIL:
A B
Lower Stool No Scallops - Sloping Stool Plate
Inner Bottom Section B-B Full Penetration

Ring Web
Welding
Inner Bottom
Critical
Locations
Longitudinal Floor
Girder
Longitudinal
Floor Girder
Outer Bottom A
2-042-1
Section A-A
Critical Sloping Stool Plate
Inner Bottom Locations Ring Web
Design Detail
Floor
Longitudinal A Original Detail
Girder
2-42 Fabrication Cost 2 1
COMMENTS:
- Alignment between double bottom girders and lower stool webs and between floors and stool plates is to be maintained.
- Full penetration welding between inner Bottom and stool plates and near the corner intersections of primary members, floor
plates, girders and lower stool webs of ballast hold side.
- Fillet welding between inner bottom and stool plates for dry cargo hold side.
- Ensure stop and stop of welding is as far as practicable from the corners.


Critical Detail:
Detail Description:
Intersection of Transverse Bulkhead Stool and
Lower Transverse Bulkhead Shelf Plate
FLEET
TEC HN OLOGY LTD
43
Rev. 001-260199

Transverse
Bulkhead Detail Design A - Avoid
Original Detail Crossing Shedder Plates
Critical Detail
Shelf Plate Shedder
Plates
Shedder
Plates
Lower Stool
Diaphragm
Shelf
Plate Full Penetration
Welding
2-043-1
Lower Stool
Detail Design B Detail Design C
CRITICAL DETAIL:
Transverse Bulkhead
A
Shedder
Reduces stress
Shedder Plates Plates
concentrations in the
corrugation corners
Critical Locations Diaphragm
Shelf Plate Full Penetration

Welding
2-043-2 Lower Stool 2-043-3

A
Critical Locations
b
Hopper Lower Stool
h = b/2 Shedder Plate
Tank Inner Bottom
2-43
Diaphragm
Shelf Plate h
Section A-A Lower Stool
Design Detail
Shedder Detail Attributes Rankings
Plates Critical
Locations
A B C
Lower Stool Shelf Plate Fabrication Cost 1 2 3
COMMENTS:
- Alignment between lower stool sloping plates and corrugation faces is to be maintained.
- Full penetration welding at the connections of the bulkhead corrugations, diaphragm and the stool sloping plates to the lower
stool shelf plate. Ensure start and stop of welding is far away as practicable from the corners of the corrugation.


Critical Detail:
Detail Description:
Intersection of Transverse Bulkhead and Topside
Tank and Upper Stool
FLEET
TEC HN OLOGY LTD
44
Rev. 001-260199

Deck
Critical
Detail Design A- Diaphragm Plates between Corrugations
Detail Topside Tank
Watertight division or ring

Bulkhead Deck web in line with flange 'b' of
Upper the bulkhead corrugations
Stool
Upper Stool
Shelf Plate
Diaphragm
CRITICAL DETAIL:
Plates
b b
a a
2-044-1
Deck
Detail Design B - Diaphragm Plates and Transverse Web
Reinforcement
Watertight Division Transverse web reinforcement in line with

Upper flange 'a' of the bulkhead corrugations
Stool
Critical Locations Full Collar
and Lugs
Upper
Stool
Topside Tank Upper Stool
Shelf Plate
Upper
Stool Shelf
Diaphragm
Plates
2-044-2
Critical Locations
Design Detail
A B
2-44 Fatigue Performance 2 1
COMMENTS:
- Alignment between transverse web and the flange of corrugations. Maximum misalignment is to be less than t/3 where t is the
thinner of the plates to be aligned and misalignment is the overhang of the thinner thickness.
- Fillet welding for connections of transverse bulkhead to topside tank and upper stool shelf plating.
- Ensure start and stop of welding is far away as practicable from the critical corners.


Critical Area: Deck
Critical Detail:
Detail Description:
Intersection of Deck Plating and Longitudinal Hatch
Coaming End Bracket
FLEET
TEC HN OLOGY LTD
61
Rev. 001-260199
CRITICAL AREA: Deck DETAIL DESIGN GOOD PRACTICE:

Detail Design A - Extend Longitudinal Hatch Coaming with Soft Toe
B
Critical Detail
S Longitudinal Section B-B
Hatch Coaming
0.7Hc Hc
S
Hatch Coaming
Hatch Topside Tank
End Longitudinal Boundary
End Bracket Beam (Vertical Strake)
Topside Tank
Brackets
B
A
Detail Design B - Face Flange with Soft Toe
Transverse Hatch Hatch Coaming C
Longitudinal
(End) Coaming Hc Transverse Bracket R = 500 mm
Hatch Coaming
Section C-C
with Soft Toe D Hc
Symmetrical
Flange
Hatch End
Beam Brackets
C
Topside 2-061-2
Tank
Chamfer 1:3
A
Max. Taper 1:3
View D
Transverse Taper 1:4
Ring Web
Section A-A 7 mm
15 -25 mm
Hatch Coaming Longitudinal
Longitudinal Hatch Coaming
End Brackets Hc
Brackets
Critical Location
Deck Between
Hatches
Design Detail
Topside Tank
Hatch
End
Longitudinal A B
Boundary
Beam
(Vertical Strake) Fatigue Performance 2 1
Topside Tank
Longitudinal
Topside Ring Web Maintenance Cost 1 2
2-61
COMMENTS:
- Alignment between hatch coaming end bracket and supporting structure is to be maintained.
- Maximum misalignment is to be less than t/3 where t is the thinner of the webs to be aligned and misalignment is the overhang
- Full penetration welding is to be used for a distance of 0.15Hc from the bracket toe end ensuring start and stop of welding is as
far as practicable from the toes of brackets or corners.
- A wrap around weld, free from undercut or notches, around the toe connection of the bracket to deck plating.


Critical Area: Deck
Critical Detail:
Detail Description:
Intersection of Deck Transverse Stiffener and
Deck Longitudinal
FLEET
TEC HN OLOGY LTD
62
Rev. 001-260199
CRITICAL AREA: Critical Detail

DETAIL DESIGN GOOD PRACTICE:
Deck
Deck Detail Design A - Soft Heel
Longitudinal
Topside
Transverse Deck
Max. 15 mm r = 30 mm
Ring Web
75 mm
55 mm
Transverse
Ring Web
CRITICAL DETAIL:
Deck Longitudinal Deck
2-062-1
A
Detail Design B - Soft Toe and Soft Heel
A Deck d/2
r = 30 mm
Topside Tank
75 mm
Critical Max. 15 mm
Locations
R > 0.75d
55 mm
2-62 t > d/18
d Transverse
Ring Web
Section A-A
2-062-2
Deck
Deck Design Detail

Longitudinal Detail Attributes Rankings
A B
Critical
Locations d
Topside Transverse
Ring Web
COMMENTS:
- Alignment between longitudinal stiffener and topside transverse ring web stiffener is to be maintained.
- Ensure start and stop of welding is as far away as practicable from the heel and toe.
- A wrap around weld, free from undercut or notches, around the toe and heel connections of the stiffener to the longitudinal.


Critical Area: Deck
Critical Detail:
Detail Description:
Intersection of Deck Hatch Corners/End Beams and
Topside Tank Transverse Ring Web
FLEET
TEC HN OLOGY LTD
63
Rev. 001-260199

Deck
Detail Design A - Enhanced Steel Grade and Thickness
Transverse BHD Steel of Enhanced Grade and
Topside Ring Web Increased Thickness
Transverse
Ring Web
Hatch Coaming
Critical Transverse
Detail Brackets Radiused
Corner
Hatch Coaming
Longitudinal
Brackets
CRITICAL DETAIL: Plate of
Deck Longitudinals Intermediate
Critical
A 2-063-1 Thickness
BHD Locations
Hatch End Beam Detail Design

Transverse
Hatch Coaming
Ring Web Transverse Hatch Transverse Bracket
S
(End) Coaming Hc with Soft Toe
Eliptical
Hatch Coaming
Hatch End
Transverse
Beam
Brackets
Hatch End A
Under Deck Hatch Coaming
Beam
Cantilever Support Longitudinal Avoid Scallops
Brackets Brackets Alternatively Fit
Collars Transverse
Section A-A Ring Web 2-063-2
Hatch Coaming
Transverse
Transverse Hatch
(End) Coaming Hc Transverse
Ring Web
Design Detail
A Original Detail
Hatch End Beam
Critical Locations
2-63
COMMENTS:
- Alignment between hatch end beam and support in the topside tank is to be maintained.
- Maximum misalignment is to be less than t/3 where t is the thinner of the webs to be aligned and misalignment is the overhang
- Ensure start and stop of welding is as far away as practicable from toes of brackets or corners.
- A wrap around weld, free from undercut or notches, around the bracket to upper deck.

B.5.3 Container Ship Structure Details
Hatch Coaming Stay

Hatch Coaming
Deck
Side Shell
Longitudinal
Cross-Deck
Strip Side
Stringer
Side
Shell
Transverse
Bulkhead
Transverse
Floor Bottom
Longitudinal
Bottom
Girder
Figure B.5.3 Typical Midship Section Nomenclature for Container Ships


Transverse
Transverse
Bulkhead
Transverse Floor Stiffener Floor
- Asymmetric Longitudinal 1 Bottom

- Symmetric Longitudinal 2 Longitudinal
Critical
Transverse Bulkhead 3 Detail #3
Critical Detail Outer

#1 & #2 Bottom
Transverse
Transverse
Bulkhead Floor
Bottom
Girder
Bottom Girders Transverse Bulkhead Section B.7
Non
Critical
Detail
Outer
Bottom
Table B.5.3 (b): Critical Structural Details for Double Side

Transverse
Bulkhead
Transverse
- Asymmetric Longitudinal 21 Web Frame
Critical
- Symmetric Longitudinal 22 Detail #26
Side
Side Shell Longitudinals Longitudinal
Critical Detail
#21 & #22
Side
Transverse Bulkhead 26 Shell
Transverse
Bulkhead
Transverse Web Frames Section B.7 Transverse
Web Frame
Side Stringers Side Stringer
Non Critical
Detail Side
Transverse Bulkhead Section B.7 Shell
Non Critical
Detail
Longitudinal Bulkhead Longitudinal

Bulkhead
Plating Transverse Web Frames Section B.7
Transverse
Web Frame

Table B.5.3 (c): Critical Structural Details for Deck and Hatch
Longitudinal
Coaming Stay
Transverse Web Frame Section B.7 Cross-Deck

Strip
Critical
Deck Plating Cross-Deck Strips Section B.7 Detail
Critical
Detail
Deck
Critical
Detail
Longitudinal Coaming Stay Section B.7
Transverse
Web Frame
Longitudinal
Cross-Deck Coaming
Strip
Deck
Hatch Corners Section B.7

Critical
Transverse Detail
Bulkhead
Longitudinal
Bulkhead
Longitudinal
Coaming Stay
Corners Section B.7
Critical
Hatch Coaming
Detail
Critical
Coaming Stay Section B.7 Detail

Vessel Type: Container Ships Detail

TEC HN OLOGY LTD
1
Rev. 001-260199
CRITICAL AREA: Transverse DETAIL DESIGN GOOD PRACTICE:

Floor

d
Bottom
Longitudinal r = 30 mm
Critical 55 mm
Detail
Max. 15 mm
75 mm
Outer
Bottom
CRITICAL DETAIL:
3-001-1
A
Inner
t > d/18
r > 0.75d 55 mm
Max. 15 mm
Transverse 75 mm
Longitudinal Floor

Stiffener
3-001-2

d
Bottom and Soft Backing Bracket t > d/18
r > 2.0d
t > d/18
A r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm

Min. 1.5d
Transverse 3-001-3
Floor
d
d = 180 - 300 mm
Critical Detail Design
Locations
Transverse Floor Detail Indices Good Practice
Stiffener A B C
COMMENTS:
- Alignment between the transverse floor and the backing bracket is to be maintained.


TEC HN OLOGY LTD
2
Rev. 001-260199
Transverse
Floor

d
Bottom
Critical 55 mm
Detail
Max. 15 mm
75 mm
Outer
Bottom
CRITICAL DETAIL:
3-002-1
A
Inner
r = 30 mm
and Soft Heel
t > d/18
r > 0.75d
55 mm
Max. 15 mm
Transverse
Longitudinal 75 mm
Floor

Stiffener
3-002-2
Outer Detail Design C - Soft Toe d

Bottom
r > 2.0d
t > d/18
r > 400 mm
r > 0.75d
A
Max. 15 mm
Max. 15 mm

Min. 1.5d
3-002-3
Transverse
Floor
d
d = 180 - 300 mm
Critical Detail Design
Locations
Transverse Floor Detail Indices Good Practice
Stiffener A B C
3-02
COMMENTS:
- Alignment between the transverse floor and the backing bracket is to be maintained.


Critical Detail:
Detail Description:
Intersection of Inner/Outer Bottom Longitudinal and
Transverse Bulkhead
FLEET
TEC HN OLOGY LTD
3
Rev. 001-260199
CRITICAL AREA: Transverse DETAIL DESIGN GOOD PRACTICE:

Bulkhead
Detail Design A - Soft Toe Bracket
with Edge Stiffener
Increased Web Depth of d s
Bottom Watertight Floor s

Longitudinal
Critical Stiffener and Soft Toe
Detail
Outer
Brackets on Back Side
Bottom
CRITICAL DETAIL:
A d r>d
Oil Tight
Min. 400 mm
Bulkhead
Inner Max. 15 mm
Bottom
3-003-1
Water Tight
Soft Toe Bracket
Floor
Detail Design B - d s
with Edge Stiffener
Longitudinal Increased Web Depth s
of Watertight Floor
Outer
Stiffener and Soft Toe
Bottom
Brackets on Both Sides
A
r>d d r>d
Critical Min. 400 mm Min. 400 mm
Section A-A d Locations

Max. 15 mm
Flanged s
Bracket
s
Note:
- Higher tensile steel bracket
3-003-2
- Bracket thickness = d/50 or 12 mm Min.
Critical
Locations Design Detail
A B
3-03
COMMENTS:
- Alignment between the floor stiffener, soft toe bracket and the web of the longitudinal is to be maintained.
- A wrap around weld, free of undercut or notches, around the floor stiffener heel and the heel and toe of the bracket connection to
the longitudinal.


TEC HN OLOGY LTD
21
Rev. 001-260199

Transverse
Web Frame
Side
Longitudinal d
Critical
Detail
r = 30 mm
55 mm
Side Max. 15 mm
Shell 75 mm
Side Shell
Side Longitudinal Detail Design B - Soft Toe
Bulkhead
Side Shell
and Soft Heel d
t > d/18
r > 0.75d 55 mm
Max. 15 mm
75 mm
d/2
3-021-2

d
t > d/18
r > 2.0d
t > d/18 r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm
Min. 1.5d
Critical Min. d/2 Min. 300 mm
Locations
3-021-3
d = 180 - 300 mm
Detail Design
Detail Indices Good Practice
A B C
Transverse Web 3-21 Fatigue Performance 3 2 1
Frame Stiffener Ease of Inspection 1 1 2
COMMENTS:


TEC HN OLOGY LTD
22
Rev. 001-260199

Transverse
Web Frame
Side
Longitudinal d
Critical
Detail
r = 30 mm
55 mm
Side Max. 15 mm
Shell 75 mm
CRITICAL DETAIL:
3-022-1
Side Longitudinal
Side Shell Bulkhead
Side Shell d
Longitudinal and Soft Heel r = 30 mm
t > d/18
r > 0.75d
55 mm
Max. 15 mm
75 mm
d/2
3-022-2

d
t > d/18
r > 2.0d
t > d/18
r > 400 mm
r > 0.75d
Max. 15 mm
Max. 15 mm
Critical
Locations Min. 1.5d
d = 180 - 300 mm 3-022-3
Detail Design
3-22
Transverse Web A B C
Frame Stiffener
COMMENTS:


Critical Detail:
Detail Description:
Intersection of Side Shell Longitudinal and
Transverse Bulkhead Horizontal Stiffener
FLEET
TEC HN OLOGY LTD
26
Rev. 001-020299

Transverse
Bulkhead Detail Design A -
Side Soft Toe
Longitudinal
Critical
Detail
Side
Shell
d
s
CRITICAL DETAIL: X > 1.0d

r > 1.5d
Side Longitudinal 3-026-1

Side Shell Bulkhead
Max. 15 mm
Detail Design B - Soft

t = d/22 or
Bkt Bkt Toe and Soft Backing Max. 15 mm 12 mm Min.
Bracket
A A
Bkt Bkt r > 2.0d X > 1.5d

Min. 400 mm Min. 300 mm
d
Transverse Deep Transverse Oil s
Tank Bulkhead in Tight Bulkhead
Wing Ballast Tank
X > 1.0d
Critical r > 1.5d
Locations
3-026-2
Max. 15 mm
d = 150 - 350 mm
s s
3-26 Detail Design

Section A-A A B
COMMENTS:
- Alignment between the web of the bulkhead horizontal stiffener, the soft toe brackets and the web of the side longitudinal is to
be maintained.
- A wrap around weld, free of undercut or notches, around the heel connections of bulkhead stiffener and toe connection of brackets.

B.5.4 Warship Structure Details
Strength
Deck
Deck
Intermediate
Beam
Deck
Casing
Bulkhead
Intermediate
Deck
Deck Girder
Deck
Deep Stringer Continuation
Stringer
Tank Top
Side Shell
Frame
Shell
Transverse
Floor
Shell
Longitudinal
Plate Girder
Figure B.5.4: Typical Midship Section Nomenclature for War Ships

Table B.5.4 (a): Critical Structural Details for Bottom

Critical Structural Details Detail Type # Detail Figure
Detail #3 & #4
Transverse Floor Girder
- Water-Tight Collar 1
- Non Water-Tight Collar with 2 Transverse
Bottom Longitudinals Lug Compensation Pieces Floor Stiffener
Transverse
Transverse Floor Stiffener Floor Shell
- WT and GT Bulkhead 3
Longitudinal
- Bulkhead 4
Detail #1 & #2
Detail #5, #6 & #7

Deck Plating
- Ending of Flat Bar Stiffener 5
- Ending of Fabricated Tee Stiffener 6
- Ending of Rolled Tee Stiffener 7 Transverse
Floor
Detail #8
Transverse Floor Stiffener 8 Transverse Floor
Stiffener
Cut-Out
Transverse
Floor
Transverse Floor Cut-out
- Flat Bar Compensation Piece 9
Transverse Floor
Detail #9
Stiffener
Table B.5.4 (b): Critical Structural Details for Side Shell
Deck Plating Detail #10

Side Shell
- Water-Tight Collar 10
Frame
Deck
Side Shell Frame

- Rolled Tee to Fabricated Tee of 11
Similar or Differing Depths
Detail #11
Web Frame or Bulkhead

Side Shell Frame - End Connection of Side Shell Frame 12 Side Shell
Detail #12
Frame
Deck Beam Detail #13 & #14
- End Connection of Deck Beam 13 Web Frame
or Bulkhead
- End Connection of Side Shell 14 Deck
Frame at Deck Deck Beam
Detail #16 & #17

Side Shell Frame
Deep Stringer
- Egg-Boxing of Rolled Tees of 15 Deck
Differing Depths
- Egg-Boxing of Rolled Tees of 16 Deep Stringer
Similar Depths

Table B.5.4 (c): Critical Structural Details for Deck

Deck
Deck Beams Longitudinal
- Rolled Tee to Fabricated Tee of 17

Similar or Differing Depths
- Rolled or Fabricated Tee to Rolled 18
or Fabricated Tee of Similar or
Differing Flange Widths 19
Detail #17,
- Rolled or Fabricated Tee to #18 & #19 Deck Beam
Fabricated Tee of Differing Depths
Deck Longitudinal Girders Deck Beam

- Tripping Bracket Detail 20 Deck Girder
- Tee End Connection to Deeper Tee 21 Deck

- Tee End Connection to Similar 22
Deck Beams Depth Tee
Detail #20,
#21 & #22
Detail #23
Deck Longitudinals Longitudinal
- Non Water-Tight Collar with 23
Lug Compensation Pieces
Casing Bulkhead
Deck
Casing Bulkhead
- End Connection of Deck Beam in way 24
way of Deck Opening
Detail #24
Deck Beam
Stiffener
Girder
Deck
Casing Bulkhead Casing
Bulkhead
- End Connection of Deck Beam at 25
Girder Below Casing Bulkhead Deck Beam
Detail #25

Vessel Type: War Ships Detail

Critical Area: Bottom
Critical Detail: Intersection of Bottom Longitudinal and Transverse
Floor
Detail Description: Water-Tight Collar FLEET
TEC HN OLOGY LTD
1
Rev. 001-260199

Transverse
Floor Detail Design A
W + 60 L=4xT
L Collar
= 25 mm Minimum
45o
Shell
Deck, Shell or
Tank Top
Longitudinal
Critical Detail
Continuous
Collar Member
CRITICAL DETAIL: W T
Deck, Shell or
Tank Top
Collar Penetrated
Member
Continuous
Member
Detail Design B
W + 60
35 mm
Collar
Penetrated Member - Bulkhead,

Deck, Floor or Web of Structural
Member
20 mm
Deck, Shell R 25 mm
or Tank Top
T
Continuous R 60 mm
Member W
Continuous
Member
Collar
Front View
Continuous Member Penetrated Member
Design Detail
Deck,
Shell or
Tank Top A B
Collar
Side View
COMMENTS:


Critical Detail: Intersection of Bottom Longitudinal and Transvserse
Floor
Detail Description: Non Water-Tight Collar with Lug Compensation FLEET
TEC HN OLOGY LTD
2
Pieces Rev. 001-260199

Transverse
Floor Detail Design A
W + 60
5xT
Shell
20 mm
Longitudinal
Critical Detail
CRITICAL DETAIL: Deck, Shell,

R25 mm
W
20 mm
Bulkhead or
Longitudinal
Tank Top T
or Girder
Penetrated
Longitudinal Member
or Girder
Detail Design B
Penetrated W + 60
Lug L L = 4 x T (Minimum)
Member Compensation
Piece
15 o
Front View 15 mm
Longitudinal
or Girder 15 mm
Longitudinal R20 mm Longitudinal T

W
or Girder Deck, Shell, 10 mm or Girder
Bulkhead or
Tank Top
Penetrated
Member
Side View
Design Detail
Penetrated
Member A B
Longitudinal
or Girder Maintenance Cost 1 2
COMMENTS:


Critical Detail: Intersection of Bottom Longitudinal and Transvserse
Floor Stiffener
Detail Description: WT or GT Bulkhead FLEET
TEC HN OLOGY LTD
3
Rev. 001-260199

Transverse
Detail Design A - With Doubler Plate
Floor (where D < 250 mm or t < 18 mm and 1 >= d/D >= 0.7)
Shell 60 o
R25 mm 25 mm Nose
Longitudinal
Critical Detail
D
t
T
CRITICAL DETAIL: d/3 Doubler
R25 mm Plate 60 o
Rolled Tee
Deck
d Total Thickness T = Web + Doubler = 9a/8D

where a = Section Area of Stiffener (without Plating)
Deck Beam or and D = Longitudinal or Beam Depth
Collar
Longitudinal Detail Design B - Without Doubler Plate
Bulkhead
Rolled Tee
o
60
25 mm Nose
Rolled Tee R25 mm
o 25 mm Nose
60
Deck
Deck Beam or Design Detail

Longitudinal
Collar
A B
Bulkhead Fatigue Performance 2 1
Rolled Tee Maintenance Cost 2 1
COMMENTS:
- Doubler plate not required where d/D < 0.7. Margin around doubler is 10 - 13 mm.


Critical Detail: Intersection of Longitudinal and Transvserse
Floor Stiffener
Detail Description: Bulkhead FLEET
TEC HN OLOGY LTD
4
Rev. 001-260199

Transverse Detail Design A - With Doubler Plate
Bulkhead
(where D < 250 mm or t < 18 mm and 1 >= d/D >= 0.7
d
Transverse
Total Thickness T = Web + Doubler = 9a/8D
Bulkhead
where a = Section Area of Stiffener (without Plating)
Stiffener
Shell and D = Longitudinal or Beam Depth
Longitudinal
Critical Detail 3 in 1 Snipe
R25 mm
CRITICAL DETAIL: t d/3

10 mm
Bulkhead or Floor T
Bulkhead or
Collar Floor Stiffener
D
Shell Longitudinal
or Bottom Frame Doubler Plate
Detail Design B - With Chock

Shell or
Inner
Bottom
3 in 1 Snipe
Bulkhead or
Floor Stiffener
10 mm Snipe for Rolled
Sections and Fabricated
3 in 1 Snipe
Sections with A <= 103 mm
25 mm Scallop for Rolled R25 mm
Sections and Fabricated
A Sections with A > 103 mm
Shell Longitudinal Chock
or Bottom Frame 10 mm
30 mm Chock
200 mm
Max
Shell or
Inner 10 mm
Bottom
Bulkhead or
Bulkhead Floor
Stiffener Design Detail
Shell Longitudinal
or Bottom Frame
Collar
A B
Shell or Inner Maintenance Cost 1 2
Bottom
COMMENTS:


Critical Detail: Connection of Transverse Floor Stiffener and
Deck Plating
Detail Description: Ending of Flat Bar Stiffener FLEET
TEC HN OLOGY LTD
5
Rev. 001-260199

Critical Detail
Transverse Detail Design A

Floor
60 o 10 mm
Transverse Floor
Stiffener
30 mm
CRITICAL DETAIL:
Detail Design B
Transverse
Floor
Transverse Floor
Stiffener - Flat Bar
60 o R25 mm
10 mm
Deck, Bulkhead
Detail Design C
or Shell
45 o 15 mm
Transverse Floor
Stiffener - Flat Bar Transverse
Floor 25 mm
Design Detail
A B C
Deck, Bulkhead
or Shell Maintenance Cost 1 2 1
COMMENTS:


Deck Plating
Detail Description: Ending of Fabricated Tee Stiffener FLEET
TEC HN OLOGY LTD
6
Rev. 001-260199

Critical Detail

Floor
Transverse Floor 10 mm
Stiffener 60 o
CRITICAL DETAIL:
Detail Design B 30 mm
Transverse
Floor R25 mm
Transverse Floor
o
Stiffener - Fabricated 60
Tee
10 mm
Detail Design C
Deck, Bulkhead
or Shell
R25 mm
60o
15 mm
Transverse Floor
Stiffener - Fabricated Transverse
Floor 25 mm Nose
Tee
Design Detail
A B C
Deck, Bulkhead
COMMENTS:


Deck Plating
Detail Description: Ending of Rolled Tee Stiffener FLEET
TEC HN OLOGY LTD
7
Rev. 001-260199

Critical Detail

Floor
Transverse Floor 10 mm
Stiffener 60 o
CRITICAL DETAIL:
Detail Design B 30 mm
Transverse
Floor R25 mm
Transverse Floor
o
Stiffener - Rolled Tee 60
10 mm
Detail Design C
Deck, Bulkhead
or Shell
60 o
25 mm Nose
o
45 15 mm
Transverse Floor
Stiffener - Rolled Tee Transverse
Floor 25 mm
Design Detail
A B C
Deck, Bulkhead
COMMENTS:


Critical Detail:
Detail Description:
Connection of Transverse Floor Stiffener and
FLEET
TEC HN OLOGY LTD
8
Rev. 001-260199

Critical Detail
Detail Design A - Rolled Tee to Rolled Tee

Transverse
Floor
Rolled Tee
25 mm
Transverse Floor Nose
R15 mm
Stiffener
R25 mm
CRITICAL DETAIL:
Transverse Floor
Stiffener - Rolled or A o A
60
Fabricated Tee Section A-A
Transverse Floor
Stiffener - Rolled or Detail Design B - Fabricated Tee to Fabricated Tee
Fabricated Tee
Fabricated Tee W
25 mm
Nose
R25 mm
Where W < 150 mm
R25 mm
o
60
A
Transverse Floor A Section A-A
Stiffener - Rolled or
Fabricated Tee
Design Detail
A B
COMMENTS:


Transverse Floor Cut-Out
Detail Description: Flat Bar Compensation Piece FLEET
TEC HN OLOGY LTD
9
Rev. 001-260199
Transverse Floor
Cut-Out Detail Design A
Critical Detail
Transverse
Stiffener
CRITICAL DETAIL:
R25 mm
6 mm
Transverse Floor
Transverse Floor
Stiffener
Detail Design B
Cut-Out Flat Bar

Compensation Piece
R25 mm
o
45 15 mm
Transverse Floor
Stiffener Transverse Floor
Design Detail
Cut-Out Flat Bar
Compensation Piece A B
COMMENTS:


Critical Area: Side
Critical Detail:
Detail Description: Water-Tight Collar

Intersection of Side Shell Frame and Deck Plating
FLEET
TEC HN OLOGY LTD
10
Rev. 001-260199

Side Shell
Critical Detail Frame
Detail Design A
Deck
W + 60 L=4xT Collar
L
= 25 mm Minimum
o
45
R20 mm R20 + L
10 mm
CRITICAL DETAIL: W
Deck, T
Bulkhead
or Shell
Penetrated
Collar
Member
Continuous
Member
Detail Design B
W + 60
35 mm
Collar
Penetrated Member
20 mm
Deck,
R 25 mm
Bulkhead or
Continuous Shell T
Member R 60 mm
W
Continuous
Member
Collar
Continuous Member Penetrated Member
Design Detail
Deck, Detail Attributes Rankings
Bulkhead
or Shell A B
Collar
COMMENTS:


Critical Area: Side
Critical Detail: Connection of Side Shell Frame and Side Shell
Frame
Detail Description: Rolled Tee to Fabricated Tee of Similar or FLEET
TEC HN OLOGY LTD
11
Differing Depths Rev. 001-260199

Side Shell
Frame
Detail Design - Similar Depth Tees

Deck
R25 mm
Critical Detail
R25 mm
CRITICAL DETAIL:
Deck, Shell or
Bulkhead
Detail Design - Differing Depth Tees
R25 mm
Fabricated Tee
o
20 Max
Rolled Tee R25 mm
Design Detail
A
Fatigue Performance 1
Ease of Inspection 1
Maintenance Cost 1
Fabrication Cost 1
COMMENTS:


Critical Area: Side
Critical Detail: Intersection of Side Shell Frame and Web Frame or
Bulkhead
Detail Description: End Connection of Side Shell Frame FLEET
TEC HN OLOGY LTD
12
Rev. 001-020299
Side Shell
Frame Detail Design A - Vertical Chock I
Critical Detail
Side
Shell
t
Web Frame R25 mm
or Bulkhead
Deck
t
R15 mm
3I
CRITICAL DETAIL:
Side Shell Frame Chock
10 mm
Deck
Flange Width, A
Web Frame or Detail Design B - Curved Chock
Bulkhead
Side Shell Total Alignment to be
Maintained between Shell
Frame Flange and Chock
R25 mm
Side Shell
Frame
10 mm Snipe
Deck
175 mm
8 mm Chock
10 mm
Design Detail
Side Shell Web Frame or Detail Attributes Rankings
Bulkhead
A B
COMMENTS:


Critical Area: Side
Critical Detail: Intersection of Side Shell Frame and Deck Beam
Detail Description: End Connection of Deck Beam FLEET

TEC HN OLOGY LTD
13
Rev. 001-260199
Side Shell Detail Design A - With Doubler Plate

Frame
Critical Detail (where D < 250 mm or t < 18 mm and 1 >= d/D >= 0.7
Side
Shell
t
Deck
Deck
Beam 40 mm
d
T
d/3
CRITICAL DETAIL: Collar

where a = Section Area of Beam (without Plating)
Deck Side Shell and D = Depth of Frame
D
Deck Beam
Side Shell
Frame 10 mm Snipe for Rolled or Fabricated
Detail Design B - With Chock Frames when A <= 103 mm
R25 mm Scallop for Fabricated or Rolled
Fabricated or Rolled Tee
Frames when A > 103 mm
Shell Frame
Collar Deck 10 mm
Flange
Width, A
Deck Beam R25 mm
10 mm
175 mm Max
Chock Total Alignment to be
Side Shell Maintained between Deck
Beam Flange Tapered 1 in 3 Beam Flange and Chock
when Width of Beam Flange
Exceeds Width of Frame
Flange
Side Shell
Frame
Deck Beam
Design Detail
A B
Side Shell Fabrication Cost 1 2
COMMENTS:


Critical Area: Side
Critical Detail: Intersection of Side Shell Frame and Deck Beam
Detail Description: End Connection of Side Shell Frame at Deck FLEET

TEC HN OLOGY LTD
14
Rev. 001-260199
Side Shell Detail Design A - With Doubler Plate

Frame
Critical Detail (where D < 250 mm or t < 18 mm and 1 >= d/D >= 0.7
Side
Shell
Deck
Deck 40 mm
Beam
t d
T
d/3
CRITICAL DETAIL:
Deck Beam
where a = Section Area of Beam (without Plating)
Deck and D = Depth of Frame
D
10 mm Snipe for Rolled or Fabricated

Side Shell Detail Design B - With Chock Deck Beams when A <= 103 mm
R25 mm Scallop for Fabricated or Rolled
Deck Beams when A > 103 mm
Side Shell Frame
Deck 10 mm
25 mm Nose
Flange
Width, A
R25 mm
o
60
10 mm
175 mm Max
Deck Beam Chock
Beam Flange Tapered 1 in 3

when Width of Beam Flange
Side Shell
Exceeds Width of Frame
Frame
Flange
Side Shell
Deck Beam
Design Detail
A B
Side Shell Fabrication Cost 1 2
Frame
COMMENTS:


Critical Area: Side
Critical Detail: Connection of Side Shell Frame and Deep Stringer
Detail Description: Egg-Boxing of Rolled Tees of Differing Depths FLEET

TEC HN OLOGY LTD
15
Rev. 001-260199

Critical R25 mm
Side Shell Detail
Frame Detail Design A - Non-Tapered Tee
T
Deck X D-K+3
Deep T+3
Stringer
D
CRITICAL DETAIL: K
D
X+3
Shell
Detail Design B - Tapered - Fabricated Tee to Fabricated Tee
R25 mm
Deep Stringer 15 mm
Side Shell Frame 25 mm
(Longitudinal
(Transverse Smaller 4 to 1 Taper
Larger Rolled
Rolled Tee)
Tee)
Detail Design C - Tapered - Fabricated Tee to Rolled Tee

Rolled Tee
R25 mm
R25 mm Fabricated Tee
15 mm
25 mm
4 to 1 Taper
Design Detail
A B C
COMMENTS:


Critical Area: Side
Critical Detail: Connection of Side Shell Frame and Deep Stringer
Detail Description: Egg-Boxing of Rolled Tees of Similar Depths FLEET

TEC HN OLOGY LTD
16
Rev. 001-260199

Critical
Side Shell Detail Detail Design A
Frame
Y T
Deck
Deep
Stringer Z
Y+3
D/2 + 1.5
X+3
Radius
R25 mm to Suit
CRITICAL DETAIL: D
Z
Z/2 + 1.5
T+3 D
Shell
X
Detail Design B - Radiused Gusset
Radiused
Deep Stringer Gusset
Side Shell Frame
(Longitudinal
(Transverse
Rolled Tee)
Rolled Tee)
R25 mm
R20 mm
Design Detail
A B
COMMENTS:


Critical Area: Deck
Critical Detail: Connection of Deck Beam and Deck Beam
Detail Description: Rolled Tee to Fabricated Tee of Similar or FLEET

TEC HN OLOGY LTD
17
Differing Depths Rev. 001-260199

Deck
Longitudinal Detail Design - Similar Depth Tees
R25 mm
Deck Beam
Critical
Detail R25 mm
CRITICAL DETAIL:
Deck, Shell or
Bulkhead
Detail Design - Differing Depth Tees
R25 mm
Fabricated Tee
o
20 Max
Rolled Tee R25 mm
Design Detail
A
Maintenance Cost 1
Fabrication Cost 1
COMMENTS:


Critical Area: Deck
Detail Description: Rolled or Fabricated Tee to Rolled or Fabricated FLEET

TEC HN OLOGY LTD
18
Tee of Similar or Differing Flange Widths Rev. 001-260199

Deck
Rolled Tee to Rolled Tee R25 mm
Longitudinal
Deck Beam A A
Critical
Detail Fabricated Tee to Fabricated Tee
CRITICAL DETAIL: R25 mm if Required

(Dependent on Method
Deck, Shell or
of Construction)
Bulkhead A A
150 mm Min
Rolled or
Fabricated Tee Rolled or Rolled Tee to Fabricated Tee R25 mm
Fabricated Tee
R25 mm
A A
3 to 1 Taper
Section A-A
Rolled or Rolled or
Fabricated Tee Fabricated Tee
Similar Flange Widths

Differing Flange Widths by
Greater than 3 mm each Side
Design Detail
A
Similar Flange Widths Differing Flange Widths Ease of Inspection 1
Maintenance Cost 1
Fabrication Cost 1
COMMENTS:


Critical Area: Deck
Detail Description: Rolled or Fabricated Tee to Fabricated Tee FLEET

TEC HN OLOGY LTD
19
of Differing Depths Rev. 001-260199

Deck
Rolled Tee to Fabricated Tee
Longitudinal
R20 mm
Deck Beam
Critical
Detail
Rolled Tee
3 in 1 Taper Fabricated Tee
4 in 1 Taper
CRITICAL DETAIL:
Deck, Shell or Fabricated Tee to Fabricated Tee

Bulkhead
Fabricated Tee
Fabricated Tee
Rolled or Fabricated
Tee 20o Max
150 mm Min
R25 mm
R25 mm if Required
Dependent on Method of Fabricated
Deck, Shell or Bulkhead
Construction Tee
3 in 1 Taper
Similar Flange Width

Rolled or Flanges Differ by Greater
Fabricated Tee Fabricated Tee
than 3 mm Each Side
Design Detail
A
Maintenance Cost 1
Fabrication Cost 1
COMMENTS:


Critical Area: Deck
Critical Detail: Connection of Deck Beam and Deck Longitudinal
Girder
Detail Description: Tripping Bracket Detail FLEET
TEC HN OLOGY LTD
20
Rev. 001-260199

Deck Girder
Deck
Deck Beam
Critical E
Detail
Longitudinal 10 mm 10 mm
In Line with
CRITICAL DETAIL: Frame Web
(Moulded)
D
Shell Frame or
Deck Transverse
10 mm Snipe when D or
E <= 150 mm
R25 mm Scallop when D
or E > 150 mm
Long Fabricated
Girder
Shell or Deck
Plating
Shell Frame or Design Detail

Deck Transverse Detail Attributes Rankings
A
Long Fabricated Maintenance Cost 1
Girder
Shell or Deck
Plating
Fabrication Cost 1
COMMENTS:


Critical Area: Deck
Girder
Detail Description: Tee End Connection to Deeper Tee FLEET
TEC HN OLOGY LTD
21
Rev. 001-260199

Deck Girder
Detail Design A
Deck 10 mm
R15 mm
Deck Beam
Critical
Detail
Flat Bar
Longitudinal Chock
CRITICAL DETAIL: 25 mm
Main Longitudinal Nose
Member R25 mm
4 to 1 Taper
Rolled Tee
Detail Design B
Flat Bar Chock
Chock Required when
Difference in Tee Sections
>= 75 mm
Main Longitudinal
Member
Rolled
25 mm
Tee
Nose
60 o
R25 mm
Design Detail
A B
COMMENTS:


Critical Area: Deck
Girder
Detail Description: Tee End Connection to Similar Depth Tee FLEET
TEC HN OLOGY LTD
22
Rev. 001-260199

Deck Girder
Detail Design A - Non-Tapered Tee
W
Deck
R25 mm
Deck Beam Where W < 150 mm
Critical R25 mm
Detail
Longitudinal Fabricated Tee
CRITICAL DETAIL: Rolled or

Fabricated
Tee R25 mm
R25 mm
Rolled Tee
Rolled Tee
Detail Design B - Tapered Tee

4 to 1 Taper
R25 mm
R25 mm
25 mm Nose
Rolled or
Fabricated Rolled Tee
Tee
Design Detail
A B
COMMENTS:


Critical Area: Deck
Critical Detail: Intersection of Deck Beam and Deck Longitudinal
Detail Description: Non Water-Tight Collar with Lug Compensation FLEET

TEC HN OLOGY LTD
23
Pieces Rev. 001-260199
Detail Design A
Deck
W + 60
5xT
Deck Beam
20 mm
Critical
Longitudinal Detail
CRITICAL DETAIL: Deck, Shell,

R25 mm
W
20 mm
Bulkhead or
Longitudinal
Tank Top T
or Girder
Penetrated
Longitudinal Member
or Girder
Detail Design B
Penetrated W + 60
Lug L L = 4 x T (Minimum)
Member Compensation
Piece
15 o
Front View 15 mm
Longitudinal
or Girder 15 mm
Longitudinal R20 mm Longitudinal T

W
or Girder Deck, Shell, 10 mm or Girder
Bulkhead or
Tank Top
Penetrated
Member
Side View
Design Detail
Penetrated
Member A B
Longitudinal
or Girder Maintenance Cost 1 2
COMMENTS:


Critical Area: Deck
Critical Detail: Connection of Deck Beam and Casing Bulkhead
Detail Description: End Connection of Deck Beam in way of Deck FLEET

TEC HN OLOGY LTD
24
Opening Rev. 001-260199
CRITICAL AREA: Deck DETAIL DESIGN GOOD PRACTICE:

Longitudinal
Detail Design A - Semi-Depth Chock
Critical Detail
15 mm
Casing Bulkhead
Deck Beam R25 mm
TYP.
CRITICAL DETAIL: 10 mm
Casing
Stiffener Bulkhead Flange Width, A
Deck
Casing 10 mm Snipe for Rolled Stiffeners or

Bulkhead Fabricated Stiffeners when A <= 103 mm
R25 mm for Fabricated Stiffeners or Rolled
Stiffeners when A > 103 mm
Detail Design B - Full Depth Chock
Deck Beam
Casing
Bulkhead
Deck
R25 mm
TYP.
10 mm
R15 mm
Deck Beam
Stiffener
Design Detail
A B
COMMENTS:


Critical Area: Deck
Critical Detail: Connection of Deck Beam and Casing Bulkhead
Detail Description: End Connection of Deck Beam at Girder Below FLEET

TEC HN OLOGY LTD
25
Casing Bulkhead Rev. 001-260199

Stiffener
Detail Design A - Semi-Depth Chock
Casing Girder
Bulkhead
Deck
15 mm
Critical Detail
Deck Beam R25 mm TYP.
R25 mm
10 mm
CRITICAL DETAIL:
Casing Flange Width, A
Stiffener Bulkhead 10 mm
10 mm Snipe for Rolled Stiffeners or
Deck 15 mm
Fabricated Stiffeners when A <= 103 mm
R25 mm for Fabricated Stiffeners or Rolled
Casing Stiffeners when A > 103 mm
Bulkhead
Girder Detail Design B - Full Depth Chock
Deck Beam
Casing
Bulkhead
Deck R25 mm TYP.

R25 mm
10 mm
10 mm
Stiffener
R15 mm
15 mm
Deck Beam
Girder
Design Detail
A B
Stiffener
Casing Fabrication Cost 1 2
Bulkhead
COMMENTS:

B.6 “GOOD” DESIGN, CONSTRUCTION AND WELDING PRACTICES TO

ENHANCE FATIGUE PERFORMANCE
B.6.1 Introduction
As stated in Section B.2, structural details that are non-critical need only be subject to good
design and fabrication practice to eliminate fatigue cracking. This section provides an overview of such
good practice. The reader is referred to standard ship design and construction references [e.g., Ref.
B.9, B.10] for further reading.
Fatigue cracks in steel ship structures are commonly initiated at weld toes. These cracks are
primarily due to the presence of an initial crack-like defect, a notch or flaw that is also subject to stress
concentrations due to the local weld, and the surrounding structural geometry. The presence of welding
and assembly residual stresses further promotes the initiation of fatigue cracks. Accordingly, steps
should be taken at both the design and construction stages to minimize these influences, such measures
being beneficial in enhancing the fatigue performance of the affected welds and therefore, of the ship
structure.
B.6.2 Detail Design

Because the structural designer responsible for the fatigue performance of a structure may not
be directly involved in the generation of the final fabrication structural drawings, the higher level, or
guidance, drawings must indicate clearly what the assumed or required fabrication tolerances are.
Failure to transmit the important fatigue design data to the fabricator will greatly reduce the value of
much of the process described herein.
An overall goal of good design practice is to minimize stress concentrations. Stress
concentrations occur at the intersection of primary structural members (deck, shell and longitudinal
girders) with other primary structural members, and with secondary structural members (decks,
bulkheads and their stiffeners, etc.). Primary members should be arranged to ensure effective continuity
of strength, and abrupt changes of depth or section should be avoided. Where members abut on both
sides of a bulkhead or other members, proper alignment should be ensured. Members should have
sufficient lateral stability and web stiffening and the structure should be arranged to minimize “hard-
spots” and other sources of stress concentrations. Members should have sufficient end fixity, through
end brackets or equivalent structure, in order to provide sufficient restraint against rotation and
displacement and to provide an effective load transfer mechanism.
Secondary members are generally connected at their ends. Where a longitudinal strength
member is cut at a primary support and brackets provide the strength continuity, the bracket scantlings
are to be such that the section modulus and effective cross-sectional area are not less than those of the
strength member.
Where openings are to be used, a generous radius should be provided. For large hatch
openings, a corner radius equal to 1/24 of the hatch width, with a minimum radius of 300 mm, should be
used. If the radius must be minimized in order to reduce lost space, appropriate compensation
measures must be taken. Elliptical profiles are also being used more frequently for hatch corners, large

drain holes, etc., as they provide a more favourable stress flow pattern than radial cuts. For small
openings, a radius of 150 mm is generally used.

Since there are always stress concentrations around large discontinuities (doors or hatchways),
efforts should be made to avoid lower fatigue design category weld joints (insert plates, fillet welds,
drain holes, etc.) in the vicinity of large openings.
For openings in the webs of stiffeners, the depth of the opening should not exceed 25% of the
web depth. In addition, the opening should ideally be located such that the edges are not less than 35-
40% of the web depth from the face plate. The length of the opening should not exceed the web depth
or 60% of the secondary member spacing, whichever is greater, and the ends of the openings should be
equidistant from the corners of cut-outs for secondary members. Cut-outs for the passage of secondary
members should be designed to minimize stress concentrations. The breadth of the cut-out should be
kept as small as practicable and the top edge should be rounded, or the corner radii made as large as
possible. The direct connection of the penetrated web plating, or the scantlings of lugs or collars, should
be sufficient for the transmission of lateral loads from the secondary member.
The exceptions to the above Guidelines are the cut-outs, or “ lightening holes”, in double bottom
floor plates and webs where the wide plate web with large cut-out is an improved alternative to simple
stiffeners on the two adjacent structures (e.g., inner and outer shell).
Insert plates, typically used as reinforcement at the corner of a large opening, should be
incorporated into the deck or shell plating. Thick insert plates should be avoided due to their restraint
against weld shrinkage that leads to weld cracks. If necessary, transition strakes should be inserted to
smooth the change in thickness. Doubler plates should be avoided.
Another form of restraint and discontinuity develops when a rigid member abruptly
terminates at the midspan of a flexible plate panel. This creates a hard spot (stress concentration) in the
plate panel, and it is recommended that the terminating rigid member be extended and tapered to the
end of the plate panel.
For the intersection of two planes (i.e., longitudinal bulkhead and deck) where a primary
stress is transferred from one plane to the other, long connection brackets are recommended to ensure a
smooth transfer of stress between the planes.
The toes of brackets, etc., should not land on unstiffened plate panels and the toes should be
concave or tapered.
The use of scallops should be avoided where possible. However, scallops are regularly used
where a groove weld of a stiffener or girder is made after the members have been assembled in place.
Scallops are not recommended for stiffening members, girders or bilge keels in way of completed shell
or deck butts; rather it is recommended that the weld reinforcement be removed where crossed by the
stiffener, girder or bilge keel.
B.6.3 Construction and Welding Practice

Sound construction and welding practice encompasses various aspects such as control of
assembly and fabrication tolerances, selection of joint design, good workmanship, optimum welding
sequence, etc. The welded cruciform joint with load carrying fillet welds is frequently required in ship
fabrication. The fatigue life of such a joint can be drastically reduced by the presence of misalignment
between the load carrying members which are welded to the continuous member of the joint.

The fit-up tolerances typically limit the maximum allowable misalignment, Me, to ts/3 (max. 5 mm),
where t is the continuous plate thickness and M is measured from the centrelines of the intersecting
plates (see Figure B.6.3.1). The decrease in fatigue life is proportional to M and therefore, it is highly
desirable to minimize the misalignment.
ts
Me
Continuous
Member
Figure B.6.3.1: Maximum Welded Cruciform Joint Misalignment
Misalignment between two groove-welded members is also detrimental to fatigue life. Codes
and Standards have recognized this and generally limit the maximum misalignment to 10% of the
thickness of the thinner member, but no more than 3 mm. For the same reason, where the difference in
plate thickness is greater than 3 mm, the thicker plate should be tapered (not exceeding 1 in 3) or
bevelled to form the welded joint. Where the difference is less than 3 mm, the transition may be
achieved within the width of the weld.
Lap connections are typically not used to connect plates that may be subjected to high tensile
or compressive loading. When lap connections are used, the width of the overlap is not to exceed four
times nor be less than three times the thickness of the thinner plate.
Fillet welds are typically used for T-connections and should be on both sides of the abutting
plate. Where the connection is highly stressed, deep penetration or full penetration welds may be
required.
As far as weld joint design is concerned, it is well recognised that fatigue cracks are relatively
easily initiated in transversely loaded partial penetration groove welds and therefore, full penetration
groove welds must be specified when cyclic loading is present. Similarly, transversely loaded full
penetration groove welds made from one side on to a steel backing strip have a far lower fatigue
performance compared to similar welds made from both sides. However, the ad hoc use of a backing
strip to address the problems associated with too large a root gap during assembly may adversely affect
the fatigue strength of the welded joint.
When it comes to member fit-up for fillet welds, it is customary to aim for as small a gap as
possible between the two members. For gaps exceeding 1.6 mm, the required fillet leg length is
increased to compensate for the increased gap. At the same time, there is some evidence that a gap of
1.5 to 2 mm influences the weld residual stresses in such a way that fatigue performance is enhanced,
provided that the joint restraint is low, i.e., the attaching member is free to move in response to weld
shrinkage [Ref. B.11]. Under such circumstances, a 1.5 mm gap is ideal.

Weld shape is another important factor influencing the fatigue performance of the welded joint.
In fillet welds, a slightly convex or slightly concave weld profile is desired. Excessive convexity is
detrimental from a fatigue point of view and is limited in fabrication codes.
In the case of butt joints, weld reinforcement and undercut are important considerations.
Excessive weld reinforcement increases the stress concentration at the weld toe and therefore the
maximum allowable reinforcement in fabrication codes is typically 3.2 mm with the weld smoothly
blending into the base metal. Undercut, which is defined as a groove or notch formed in the base metal
adjacent to the weld toe, can seriously reduce the fatigue life of the welded joint. Formed as a result of
incorrect selection of welding parameters or welding consumables, or lack of welder skill, the
fabrication codes generally specify smaller allowable undercut for dynamically loaded structures. For
example, Reference B.12 allows a maximum undercut of 0.25 mm in critical members when the weld is
transverse to the applied stress. Otherwise, the maximum allowed undercut is 0.8 mm (1/32”). In
comparison, for statically loaded structures, the undercut is generally limited to a depth of 1 mm, and for
short lengths, the allowable undercut may be as high as 1.6 mm.
Longitudinal welds generally have good fatigue performance. However, the presence of starts
and stops which are not ground out, or breaks in the backing strip (unwelded joint in backing strip)
in the case of welds made from one side, significantly reduce the fatigue strength of the weld.
Where stiffening members are continuously fillet welded across completely finished butt or
seam welds , the butt welds are to be made flush. Similarly for the butt welds in the webs of stiffening
members, the butt weld is to be completed and generally made flush with the stiffening member before
the fillet weld is made.
Other welding flaws such as hydrogen-induced cold cracks in the heat affected zone,
solidification cracks in the weld metal, incomplete root penetration in welds made from one side but
without a backing strip, etc., are other likely sites for fatigue crack initiation. Embedded flaws such as
slag inclusions and porosity are relatively benign for fatigue unless present in excessive amounts.
B.6.4 Residual Stresses

Residual stresses in steel ships can be considered to be composed of two components: short
range weld thermal residual stresses and long range assembly and restraint stresses. The short range
weld thermal residual stresses result in high levels of tensile stresses that approach the yield strength of
the base metal. These residual stresses, which are always present and are difficult to control or modify,
can reduce the fatigue strength of the welded joint. However, through careful attention to fit-up
tolerances and welding sequences, the long range assembly and restraint stresses can be reduced. If the
members to be welded need to be jacked or bent in order to achieve the correct alignment, bending
stresses will be introduced into the structure and these can contribute to fatigue crack initiation. Once
cracking is initiated, these stresses tend to relax. In terms of welding sequence, it should be so arranged
that as welding progresses, the members retain as much freedom to move as possible.

Reference B.9 provides recommendations for welding sequence for butts and seams, with and without
internal framing, and for large sub-assembled plate panels. Another approach is to commence welding
in the deck and bottom regions and progress towards the neutral axis of the ship. The restraint stresses
will then be high near the neutral axis where the applied stresses are lower.
B.6.5 Weld Toe Dressing Treatments

Finally, it should be added, that in certain situations it may not be possible to avoid relatively
inferior fatigue design details, or the structural detail may involve high stress concentrations. In these
situations, fatigue crack initiation is deemed unavoidable at the design stage. In such circumstances, one
could consider weld toe dressing treatments such as hammer or shot peening, toe grinding, TIG
remelting, etc. Further information on the benefits and costs of such techniques are covered in
Reference B.13.

B.8 REFERENCES
[B.1] Ma, K-T., “Tanker Inspection and a Risk-Based approach”, Proceedings of the 8th ISOPE,
Montreal, 1998.
[B.2] Ma, K-T., and Bea, R. G., “Ship Maintenance Project, Phases II and III, Vol. 3, Repair
Management System for Critical Structural Details in Ships”, SSC 395, 1997.
[B.2] “ShipRight - Fatigue Design Assessment Procedure - Structural Detail Design Guide”, January
1996: Lloyd's Register of Shipping, London.
[B.3] “Guidelines for Surveys, Assessment and Repair of Hull Structure - Bulk Carriers”, IACS, 1994.
[B.4] “Structural Practices Standard for Canadian Forces - Steel Surface Ships”, D-03-002-008/SG-
003, Canadian Department of National Defense DSE, 1993.
[B.5] Germanischer Lloyd, “Rules and Regulations, 1 - Ship Technology, Part 1 - Seagoing Ships,
Chapter 1 - Hull Structures, Hamburg, 1992.
[B.6] Det Norske Veritas, “Rules for Classification of Steel Ships”, DnV Classification A/S, Norway,
1991.
[B.7] Lloyd’s Register, “Rules and Regulations for the Classification of Ships: Part 3 - Ship Structure”,
1996.
[B.8] American Bureau of Shipping, Rules for Building and Classing Steel Vessels, 1996.
[B.9] “Guide for Steel Hull Welding”: American Welding Society ANSI/AWS D3.5
[B.10] “Ship Design and Construction”: Taggart: SNAME 1980
[B.11] Johansson, B.G., Kjellander, S.L., and Leide, N.G., “Is it good enough? A study of tolerances
for welded ship structures”, Fitness for Purpose Validation of Welded Constructions,
International Conference, London, UK, 17-19 November, 1981, pp6-1 to pp6-11.
[B.12] AASHTO 1994. LRFD Bridge Design Specifications, SI Units, First Edition, American
Association of State Highway and Transportation Officials, Washington, DC
[B.13] Kirkhope, K.J., Bell, R., Caron, L, and Basu, R.I., “Weld Detail Fatigue Life Improvement
Techniques”, SSC Report 400, Ship Structure Committee, 1997.

Part C – Fatigue Strength Assessment
PART C - FATIGUE STRENGTH ASSESSMENT
C.1 INTRODUCTION
C.1.1 Background
The analysis procedure involved in the estimation of the fatigue life or probability of
fatigue failure over the design life of a ship may be described in terms of the following four
basic steps :
(1) Establish the long-term statistical distribution of cyclic loads on a ship over its
design life;
(2) Determine the corresponding long-term statistical distribution of the notch stress
range (∆σnotch ) at the anticipated crack initiation site;
(3) Calculate the fatigue damage inflicted by these loading cycles;
(4) Perform a reliability analysis to predict the probability of fatigue based failure .
Classification Societies, amongst others, have developed a variety of approaches to assess

the fatigue life of ship structural details by expanding on this basic four step process. Discussion
of the fatigue design approaches adopted by the major class societies is presented in Section
C.2.3 and Appendix A.
C.1.2 Objective
The assessment procedures set out in this Guide are intended to be used by engineers and
naval architects in ship structural design to reduce the likelihood of premature fatigue failure and
to assist in planning through-life inspection strategies by identifying the relative fatigue
resistance of different elements of the ship.
The objective of Part C of the Guide - Fatigue Strength Assessment - is to describe the
procedures for predicting the fatigue life of ship structural details that are subjected primarily to
wave-induced loads.
C.1.3 Scope
While the four basic steps in fatigue analysis are easily understood, the analytical
techniques required can be relatively complex. Two alternative fatigue assessment procedures,
categorized as Level 2 and Level 3 (a simplified, and a direct spectral approach respectively), are
outlined in a step-by-step fashion in Sections C.2.1 and C.2.2, respectively, and a flow chart for
these procedures is presented in Figure C.2.1.
The intended applications and limitations of these procedures are discussed in Section C.2.3,
while detailed guidance for the calculation of input loads, structural response and resulting
fatigue damage are given in Sections C.3 to C.6. A reliability procedure for estimating the
probability of failure associated with predicted fatigue lives is given Section C.7.
The processes presented are exhaustive, but they are the most practical available at the
time of production of this manual. It should be recognized that some of the procedures remain at
an early stage of development.
Fatigue-Resistant Detail Design Guide for Ship Structures C-1

C.2 FATIGUE ANALYSIS PROCEDURES
C.2.1 Overview
The Level 2 and Level 3 approaches are presented in parallel in Sections C.3 through C.6,
each section dealing with one of the four basic steps in the analytical process. While the two
Levels contain the same four basic steps as stated in C.1.1, they differ in the details of their
execution. To illustrate the differences in the two parallel procedures, Figure C.2.2.1 expands
the four basic steps into eight more detailed blocks. In each block of the fatigue analysis process
flow chart (Figure C.2.2.1) reference is made to the section within this report (Part C) in which a
technical description of the procedure is presented.
The analytical process shown in Figure C.2.2.1 includes two cross-over points between
the simplified and spectral analysis approaches. These cross-over points indicate, for instance,
that the user may use the full analysis approach to determine the long term statistical distribution
of wave loads and cross over to use the simplified approach to determine the corresponding
statistical distribution of notch stress ranges. Although the steps in the two approaches use data
and develop results of a similar form, the level of conservatism, and possibly accuracy, may not
be consistent.
C.2.2 Commentary
C.2.2.1 Intended Applications

The Level 2 procedure is best suited to the assessment of standard structural details in
ships, whereas the Level 3 approach is intended for: (i) assessing structural details outside the
calibration-base of the simplified procedure (e.g., novel structural configurations, unusual wave
environments), and, (ii) confirming and refining the fatigue performance of critical structural
details at the end of the detail design stage.
In any design where fatigue is expected to be a significant issue, the Level 1 approach of
using good fatigue-resistant details should be used as a matter of course. In the event that a
combination of Level 1, 2, and/or 3 approaches does not provide an adequate expected fatigue
life in the detail design stage, then a more general redesign of the structure may be required to
lower the overall (or local nominal) stress level.
C.2.2.2 Design Life
The design life for structural details in damage tolerant structure should be set to the
service life of the ship (typically 20 years).
If damage tolerance cannot be achieved or is impractical for the structure surrounding a
structural detail, the design life of the detail should be many times the service life of the ship to
ensure that fatigue cracks do not initiate in service. In other words, use a safe life design
philosophy.

Level 2 Level 3
1) Identify relevant environmental data for 1) Develop vessel operational profile and
target return period assemble composite wave scatter diagram
[Sec. C.3] [Sec. C.3]
2) Define reference values for load range 2) Determine transfer functions for load range
components using parametric equations components using strip theory or
[Sec. C.3] empirical transfer functions [Sec. C.3]
3) Define reference values for nominal stress 3) Determine transfer functions for nominal
range components in primary structure stress range components in primary
using parametric equations, frame and structure by global finite element analysis
beam models [Sec. C.4] [Sec. C.4]
4) Use primary defined SCF’s to determine 4) Determine transfer functions for

reference values of hot spot stress range components of hot spot stress range by
components [Sec. C.4] local finite element analysis [Sec. C.4]
5) Develop hot spot stress ranges from load 5) Develop hot spot stresses from load and
and stress transfer functions [Sec. C.5] stress transfer functions [Sec. C.5]
6) Define long-term distribution for notch 6a) Define short-term distribution for hot spot
stress range [Sec. C.5] stress ranges [Sec. C.5]
6b) Define long-term stress distribution by

summing short-term distributions
[Sec. C.5]
7) Fatigue damage summation using fatigue

design curve and Miner’s rule
[Sec. C.6]
8) Reliability / Limit state analysis

[Sec. C.6 and C.7]
Figure C.2.2.1: Fatigue Strength Analysis Flow Chart: Levels 2 and 3 Methods

The safety margins in the fatigue design cur ves presented in Section C.5 are consistent
with the safety margins in fatigue design curves for welded joints in bridges and offshore
structures. Experience has demonstrated that these margins are generally sufficient to ensure a
finite but acceptably low probability of failure during the service lives of bridges and offshore
structures, where failure is defined as the initiation of a through-thickness crack several inches
long. Fatigue cracks that do initiate will tend to do so towards the end of the service life , and
the damage tolerance of the surrounding structure will ensure that these cracks can be detected
and repaired before they pose a threat to structural integrity.
In the case of ships, the greater uncertainty in the estimation of loads is countered by the
greater redundancy of ship structures. Therefore, the safety margins in the fatigue design curves
presented in Section C.5 should be adequate for structural details in damage tolerant ship
structure. The acceptable probability of failure should be agreed upon between the ship owner
and ship designer, and should be based on previous experience whenever possible.
C.2.2.3 Non-linearities
Neither the Level 2 procedure nor the Level 3 procedure presented herein account for
non- linearities in the load and structural responses of a ship. Some of the parametric equations
presented in Sections C.3 and C.4 for predicting reference loads and stresses are based on
Classification Society rules, while an implicit assumption in the spectral analysis is that the
structural and load responses of the ship are linear. Certain non- linearities can be neglected in a
fatigue damage assessment. For example, the well-known non-linear effect of hull shape on hull
girder bending stresses occurring especially with large wave heights and pronounced bow flare
has only a small effect on fatigue life. The major part of fatigue damage is caused by the large
number of smaller stress cycles for which the linear assumption is adequate. Other non-
linearities, however, can have a significant influence on fatigue damage. For example, the time
history of the side shell pressure at a location near the still-water line that is temporarily emersed
due to ship motions can be highly non-linear because negative pressure or suction cannot occur.
Linearization and time domain techniques have to be used to account for these non- linearities.
C.2.3 Design Codes and Criteria

In the course of a commercial design it will be necessary to obtain Classification Society
approval and therefore it is prudent to follow the fatigue design procedures specified by the
specific Society, [e.g., References C.1 to C.3], or to have the procedure used approved by the
appropriate Society. The procedures for fatigue design, which are currently promulgated by the
Classification Societies, vary somewhat in scope and applicability. In order to assist in
determining the relative merits of these codes a discussion and detailed comparison of their
treatment of fatigue is provided in Appendix A.

C.3 LOAD ANALYSIS
C.3.1 Determination of Environmental Loads
C.3.1.1 Introduction
This section provides guidelines for determining the fatigue loading that a ship structure
will experience throughout its lifetime or a particular period of interest. It is assumed, as
elsewhere herein, that this Guide is being used by a qualified and practicing ship designer or
structural engineer, and therefore this Guide does not provide comprehensive direction on how to
determine the loads on a ship structure. Rather it provides guidance on how the loading is
applied to the fatigue analysis problem.
The ship designer is assumed to be already working with loads to determine the vessel’s
design strength, and size the structural elements and scantlings, and with knowledge of the
operational area and possibly the operating profile for the vessel. This Guide will show how the
information needs to be formulated and used for fatigue assessment. The fundamental difference
is that fatigue design has to be based on an explicitly statistical representation of the complete
loading environment, while much of the rest of the design only considers single extreme values.
C.3.1.2 Definition of Loads

For fatigue life analysis, loads must be defined in terms that allow derivation of stress
ranges. This may require definition of a load component in hog and sag, or as maximum
compressive and tensile load, so that a range can be defined. As for overall ship design, loads
can be categorized as global or local. Reviewing these loads, the important global loads are:
a) stillwater loads;
b) wave loads;
- low frequency steady-state, response largely quasi-static
- high frequency transient (wave impact or slamming), dynamic.
- high frequency steady-state (springing)
while the main local loads are:
c) hydrostatic pressure loads ;

d) pressure loads due to waves, as in b);
e) inertia loads from cargo or fluids (sloshing) induced by ship motion;
f) functional loads, from machinery and deck equipment;
g) ice loads.

Any loads that result in significant variation in stress levels are potentially relevant for
fatigue assessment. Fatigue damage is seen as cumulative, and thus all loading sources
potentially contribute. However, since a vessel will typically be exposed to between 3x107 and
1x108 wave cycles in its life, unless another load source compares (within an order of magnitude)
in either magnitude or number of cycles with the wave loading at a particular location, it is
unlikely to affect the cumulative load history. Thus, sources such as machinery vibration loads
(unless locally relevant) or ice loads may be ignored. Ref. C.4 provides additional discussion on
this subject.
The relative significance of each type of load depends, among other things, upon the ship
type, the payload, structural configuration and location of the structure. Tables A.3.1 to A.3.5
(pages A.3-A.4) provided earlier, give guidance in identifying the important loads for different
ship and detail types. Further discussion of the treatment of some of the most important loads is
given below.
(a) Stillwater Loads
Stillwater and hydrostatic pressure loads are not a direct factor in fatigue design as they
typically only vary with loading conditions, and thus do not occur with sufficient frequency to
cause significant fatigue damage. As fatigue damage accumulation is reduced when part or all of
a stress cycle is in compression, it can be argued that the stillwater condition contributes in
reducing or increasing the risk of fatigue. However, in general, stillwater (plus wave)
compressive loads will not be high enough to reverse the generally tensile residual stresses at a
weld. Thus, the assumption that wave induced bending is fully reversing about a stillwater
reference is generally valid.
In cases where the stillwater hull girder load is significant, such that parts of the hull do
remain in compression throughout, special techniques to treat this offset may be justified. Ref.
C.5 provides a discussion of the impact of mean stress on fatigue.
(b) Wave Loads
The dominant load on most ships arises from the response to waves. The primary
mechanism for wave loading is through hull girder bending. The loading is cyclic with periods
of the order of several seconds; the response is essentially quasi-static.
In severe sea states, dynamic phenomena such as slamming may occur which result in
transient impact loads. The response to this type of load is characterized by frequencies that are
considerably higher than those associated with normal wave loading. As the response is
transient, analysis of these loads is difficult using standard spectral techniques. However, it is
generally assumed that the number of slams in the ship's lifetime will be small enough to neglect
in fatigue analysis.
The phenomenon of “springing” in which the hull girder responds in flexure on a continuous
basis, at frequencies at or close to the natural frequency( ies) of the hull, obviously have an
impact on fatigue. This phenomenon is seen on occasions in, for example, Lakers [Ref. C.6] but
is much less common than transient (slamming) effects.

(c) Local External and Internal Pressure Loads
Waves may also impose significant loads on local structural elements. The primary
examples are the variation in static pressure on the side shell due to ship motions (roll, pitch,
heave) and dynamic pressure loads near the waterline due to wave impact.
Internally, ship motions result in inertial (acceleration) forces on liquid or solid cargoes
and deck equipment. In addition, “sloshing” of tank liquids can occur under specific conditions
of tank level, fluid density, and vessel response. “Sloshing” loads can be severe, resulting in
damage. However, they are normally neglected in fatigue design.
C.3.1.3 Definition of Operational Profile

The starting point for the fatigue analysis, (as with ship structural design overall) is the
operational profile for the vessel. This will determine the wave climate that will be experienced.
The definition of the operational profile may be as simple as stating a general area of operation
(e.g., North Atlantic) and an endurance or service speed. At the other extreme, a full operational
profile may state how much time the vessel will spend in various areas of the world, and at what
times in the year, as well as the distribution of its speed and headings to permit the exposure to
wave data to be determined,.
C.3.1.3.1 Level 2 – Method

Since the Level 2 methods for calculation of loads often rely on prescriptive parametric
equations derived for specific vessel types, the amount of information required for defining the
operational profile is minimal.
Since the basic approach in the Level 2 method is to develop a reference load (e.g., a
design value) for use in defining the long-term statistical distribution (a Weibull representation
being the “standard' approach”), the information required on operational profile is limited to that
required to define the parameters used to develop the reference loads. This may include all or
some of the following:
• Ship loading conditions, e.g., loaded departure, loaded arrival, ballast departure, ballast
arrival. This data may be used directly, and if not will be used to define parameters such as
draft and displacement for use in the reference equations.
• Speed - Speed is not typically used as an explicit parameter, but is used to identify high(er)
speed vessels (e.g., container ships), where allowances for dynamic effects such as keel or
flare slamming should be considered.
• Area of operation - while a specific area of operation is not typically cited, if the anticipated
operations are inshore or sheltered waters, then a reduction factor may be applicable in the
formula for the reference load. There are explicit examples, such as the rules for Great Lakes
vessels, where reductions in strength requirements are associated with the geographic area of
operations.

There are implicit assumptions in the prescriptive rules which the designer should be
aware of. For example, in developing extreme loads for some local structure (for example
midbody framing), primary loads may be based on head seas for vertical bending moment,
whereas rules for side shell pressure will have been developed assuming a beam sea, and a
prescribed minimum roll angle [C.7]. Combining the loads by simple or weighted summation
may not always justify the same spectral representation to be used. However, in practice the
assumption that this can be done is reasonable, and is acceptable for the Level 2 method
Another such assumption relates to the implied wave conditions used to develop the
design loads. This is discussed further in Section C.3.1.4.1.

A schematic representation of the Level 3 method is provided in Figure C.3.1.3.1, which is
taken from Ref. C.8.
In the Level 3 - Direct Calculation - Method, the loads are determined from a detailed
knowledge of the ship’s operational profile. In using direct methods for calculating extreme
loads, considerable simplifications of the operational profile are usually accepted. For fatigue
calculations, it may be necessary to examine the data in more detail to ensure a reasonably
accurate representation of spectrum shape.
The operational profile information required for a detailed calculation includes:
• The projected route of the vessel described in terms of areas of operation and the % time
spent in these areas;
• Vessel loading conditions or missions and relative time spent in each mode; (loading
conditions are appropriate for commercial vessels, while the mission may be more
appropriate for military or patrol vessels);
• Distribution of time spent at each heading relative to the predominant sea direction;
• Vessel average speed ranges and relative amount of time spent at each speed in a particular
sea state or wave height.
This data can then be combined with a statistical representation of the wave climate for each
area to provide a complete picture of the vessel's "sea operational profile", as described in the
next section.
In order to make the calculations feasible, each of the parameters is discretized in some
manner. For example, the route can be divided into Marsden Zones [Ref. C.9] (or zones of
latitude and longitude transited by the vessel) and the time spent in these zones. Loading can be
treated in terms of standard conditions. Relative heading can be simplified into head, bow,
beam, quartering, and following seas; and speed can be treated as sets of speed ranges.

Figure C.3.1.3.1: Schematic of Level 3 - Direct Calculation - Method

When a new design will follow the same operational profile as an existing ship, the
existing ship’s operations may be studied and characterized from operational logs, [e.g., Ref.
C.8]. For new designs, operational profiles can be generated from the operators’ plans. The
level of discretization of operational and/or environmental data should correspond to the
certainty in the operational profile information.
The process of developing a detailed operational profile requires the development of
input joint probability diagrams, including ship speed versus Sea State (or wave height) (Table
C.3.1.3.1) and then ship relative heading versus Sea State (or wave height) (Table C.3.1.3.2).
These are obtained either from historical data (as in the examples) or perhaps from operating
directions for the vessels (particularly speed in given sea states).
When an operational profile is developed in the absence of historical data, speed, sea
state and heading are often assumed to be independent quantities. This may not always be the
case, as in severe sea states, the practice is to reduce speed and to orient the ship in preferred
directions. However, since the bulk of fatigue damage arises from the exposure to moderate
conditions, and because the amount of time spent in these severe sea sates is not as significant as
that spent in more moderate conditions, the assumption of independence is reasonable, and
avoids extremely complex computations that are not justified. When the profile is developed
from existing ships' logs, it will obviously reflect current practice, which may or may not be
modified by other features of the new design.
The impact of on- board weather routing and forecasting systems in reducing the
exposure of the vessel to extreme conditions, may also "skew" the observed or theoretical
operational profile. Such measures cannot be taken into account in the design process at present.
For a more detailed discussion of these issues see Ref. C.8.
The next data set required is the distribution of time spent in each geographical area
(Table C.3.1.3.3). In order to construct the lifetime operational profile, time spent in port should
also be included. A ship which spends 50% of its time alongside will obviously see fewer wave
encounters per year than one which is more or less continuously at sea.
These three distributions are combined with the wave data into a table of joint probabilities
of speed, heading, wave height, and wave period for the specific profile. The process for
producing this final joint distribution is described in the following section.

Table C.3.1.3.1: Sample Operational Profile

Joint Probability of Speed and Wave Height
SPEED Significant Wave Height (m)
(knots) 1 2 3 4 5 6 7 SUM
0-6 0.0028 0.0000 0.0009 0.0000 0.0000 0.0010 0.0007 0.0054
6-10 0.0056 0.0031 0.0033 0.0082 0.0086 0.0261 0.0010 0.0559
10-14 0.0129 0.0219 0.0503 0.0378 0.0449 0.0896 0.0019 0.2593
14-18 0.0900 0.1253 0.1322 0.1007 0.1338 0.0968 0.0005 0.6793
SUM 0.1113 0.1503 0.1866 0.1467 0.1873 0.2135 0.0041 1.0000
Table C.3.1.3.2: Sample Operational Profile:

Joint Probability of Heading and Wave Height
Significant Wave Height ( m)
Heading 1 2 3 4 5 6 7 SUM
Head Seas 0.0084 0.0114 0.0167 0.0128 0.0160 0.0221 0.0005 0.0879
Strbd. Bow 0.0234 0.0326 0.0380 0.0295 0.0383 0.0375 0.0005 0.1998
Strbd. Beam 0.0283 0.0375 0.0439 0.0355 0.0454 0.0497 0.0009 0.2413
Strbd.Quart. 0.0421 0.0586 0.0729 0.0565 0.0724 0.0800 0.0013 0.3840
Following 0.0098 0.0136 0.0164 0.0129 0.0165 0.0176 0.0003 0.0871
SUM 0.1120 0.1537 0.1880 0.1472 0.1886 0.2070 0.0035 1.0000
Table C.3.1.3.3: Geographical Distribution of Time

Marsden Zone Total %
15 5%
16 30%
17 10%
23 9%
24 28%
25 18%
SUM 100
C.3.1.4 Definition of Wave Climate

In general, the wave climate is not explicitly defined for the Level 2 approach, rather a
“design wave” condition is either defined or implied by the form of the load equation. The
probability of the design wave occurring in the lifetime of the ship is inherent in the various load
equations specified by the Classification Societies, and as agreed by IACS, and is understood to
be 10-8 encounters per 20-year life.
Modifications may be required to the load formulae to correct for ship size to ensure
equal probability of occurrence. In the case of larger vessels, linear extrapolation of the
prescriptive formula would result in an unrealistically large wave height, and thus upper bounds
have been introduced to the formulae based on vessel dimensions.

Most major Classification Societies have replaced prescriptive rules by direct calculations
for large vessels (e.g., L > 190m). This requires the development of at least part of the data
described here for use with the Level 3 approach. In the case of smaller vessels, including many
naval ships, the concern with prescriptive rules, again most typically for longitudinal strength, is
that the resulting “design” load will occur at a higher frequency of occurrence than is consistent
with a spectral analysis. Whenever direct calculations of extreme loads are made, care should be
taken to ensure that the underlying assumptions are understood and properly accounted for in the
fatigue analysis
C.3.1.4.2 Level 3 - Method

C.3.1.4.2(a) Sources of Wave Climate Data
The wave climate experienced by ships varies considerably depending on the area of
operation. Wave data is available for most parts of the world including oceans and large bodies
of water such as the Great Lakes. Perhaps the most comprehensive compilation of wave data is
published by British Maritime Technology [Ref. C.9]. Regional wave data is also available, an
example being for Canadian Waters - including the Great Lakes [Ref. C.10]. Wave data sites are
also available on the Internet for example, at, www.meds.dfo.ca maintained by the Canadian
MEDS (Marine Environmental Data Services), and www.nodc.noaa.gov maintained by the U.S.
National Oceanic and Atmospheric Administration (NOAA).
Wave climate atlases typically divide the world's oceans into blocks or areas. One such
system is that of the “Marsden Zones” used frequently in commercial and offshore applications
as presented in Ref. C.9.
Wave heights may be quoted directly, usually in 1 m gradations, or alternatively as “sea

states”. Ships’ log data often use sea states due to their more direct relationship with observable
physical phenomena. Some warship design methods define wave parameters as NATO sea
states, and the ship design performance requirements may be expressed in these terms. [Ref.
C.11]. When sea state data is used, it will generally have to be transformed into wave height
values for use in load calculations. This transformation is described in Section 3.1.4.2(f).
The degree to which the wave climate can be defined depends on how well the route is
defined. In cases where the wave climate is unknown, it is advisable to employ “standard” wave
climate data for areas known to experience severe weather. A good choice in this circumstance
is to use data gathered in the North Atlantic, not only because it experiences severe weather
compared with many other areas, but also because the quality of data for this area of the world's
oceans is very high as it is particularly well instrumented. Hence, the use of the familiar,“Winter
North Atlantic” ocean definition.
The data sources provide long-term wave height and possibly period data, usually
expressed in terms of, for example, the frequency of occurrence of a significant wave height
and/or period. For fatigue analysis, it is necessary to derive a distribution of the wave energy
across the wave frequency band from this data.

C.3.1.4.2(b) Selection of Wave Spectral Model

In order to use the wave climate data in response modelling, it is generally necessary to
select a spectral model for wave height. This is a mathematical representation of the distribution
of wave energy as a function of the spectral parameters.
Much has been written about the use of appropriate sea spectra, and some references are
provided. For coastal areas it is important to fit the data to a “limited fetch” spectrum such as
JONSWAP [Ref. C.12], while for a fully developed sea, the Bretschneider spectra [Ref. C.13] is
popular. The spectra selected must be appropriate to the data available. If the latter includes
much energy at swell frequencies, then a spectrum with a greater number of parameters may be
necessary.
If the vessel is being designed for a specific area for which actual sea data is available from,
say, directional wave buoys, then it may be possible to use this directly. However, care should
be taken to ensure that an adequate number of readings are available to provide a statistically
valid prediction for the climate the ship will experience.
C.3.1.4.2(c) Encounter Frequency and Reference Axes

Wave height spectra typically refer to the wave climate at a stationary point in the ocean.
The frequency of waves that the ship experiences differs from the frequency a stationary
observer would experience. The former is usually referred to as the “encounter frequency”. The
spectrum being used in the analysis needs to be modified to account for this fact. In general, the
ship direction will not be aligned with the direction of the waves. The encounter frequency that
accounts for the effects of speed and heading is given by:
 ?V 
? e = ?  1 − cos?  (C.3.1)
 g 
(See Nomenclature for variable definitions).
The expression for the wave height spectrum also needs to be modified to account for the
transformation of the axes system from a fixed point to one that is translating with the ship. The
modified wave spectrum is given by:
S? (? e ) = S? (? )
1
(C.3.2)
1 − (2ωV/g )cos?
If a software package is used to generate motions and loads, these corrections will normally be
generated automatically.

C.3.1.4.2(d) Representation of Confused Seas with Two-Parameter Spectra
In addition, if a two-parameter spectrum is employed, it is often the practice to account

for the “short-crestedness” of the seas. This provides a means of accounting for the variation in
wave energy with direction within a single- moded spectrum. As a two-parameter spectrum does
not explicitly consider wave direction (only height and frequency), the alternative to assuming
uni-directional seas (all wave energy concentrated on one axis) is to apply a cosine-squared
spreading function as follows:
( )
S? ? e , ? ′ = Sn (? e ) x
2
p
cos 2 ? ′ (C.3.3)
Again, most software packages will allow for either uni-directional or 'corrected' representations
of the spectrum.
C.3.1.4.2(e) Wave Scatter Diagrams
Wave climate data for both directional and non-directional seas are usually expressed in
terms of “wave scatter diagrams” which express the relative frequency of occurrence of certain
combinations of wave height and modal period. Hence, using statistical terminology, the
diagram is the joint probability density function for wave height and period. A typical non-
directional wave scatter diagram (in this case for a composite area covering the North Atlantic) is
shown in Table C.3.1.4.1.
The direct calculation method is based on the creation of a composite wave scatter diagram
derived from the reference wave climates that comprise the proposed shipping route or operating
area. Mathematically, the composite scatter diagram is defined as:
N′
(HS , TZ )composite = ∑ µ i (HS , TZ )i (C.3.4)
i =1
where Hs and Tz are the significant wave height and zero crossing period respectively, µi is the
proportion of time spent in the ith area (Marsden zone), and N′ is the total number of areas along
the route. In practice, the composite wave scatter diagram is created by:
(1) multiplying the relative frequency values in wave scatter diagrams for each geographical
area by their respective factor µi;
(2) adding all of the modified relative frequency values for common height and period to
develop a single weighted scatter diagram.

Table C.3.1.4.1: Scatter Diagram for North-Atlantic for Use in Fatigue Computations
Tz (sec) Hs (m) for 1000 wave encounters total
0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5
13.50 0 0 0 0 0 1 1 0 0 0 0 0 2
12.50 0 0 1 2 2 2 2 1 1 1 0 0 12
11.50 0 1 4 6 7 5 4 2 2 1 1 0 33
10.50 0 4 14 21 18 13 8 4 3 2 1 1 89
9.50 1 16 43 48 34 20 11 5 3 1 1 0 183
8.50 4 47 80 65 38 18 8 4 2 1 0 0 268
7.50 13 75 79 46 21 9 3 1 1 0 0 0 248
6.50 21 54 33 14 5 2 1 0 0 0 0 0 130
5.50 14 13 5 1 0 0 0 0 0 0 0 0 33
4.50 3 0 0 0 0 0 0 0 0 0 0 0 3
3.50 0 0 0 0 0 0 0 0 0 0 0 0 0
2.50 0 0 0 0 0 0 0 0 0 0 0 0 0
Total 56 210 259 203 125 70 38 17 12 6 3 1 1000
While the ideal is to compile a composite wave scatter diagram that reflects the intended
route of the ship, this information is not necessarily available. In this situation, it may be
necessary and appropriate to use average pre-compiled data. While Table C.3.1.4.1 is intended
for use for routes in the North Atlantic, this is significantly more severe than the “world average”
wave climate which is shown in Table C.3.1.4.2. The designer has the option of using a more
severe set of conditions to ensure a level of conservatism in the fatigue analysis.
Table C.3.1.4.2: Scatter Diagram for World Wide Trade for Use in Fatigue Computations
TZ (sec) HS (m) for 1000 encounters
0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 total
13.50 0 0 0 0 0 0 0 0 0 0 0 0 0
12.50 0 0 1 1 1 1 0 0 0 0 0 0 4
11.50 0 1 2 3 3 2 1 1 1 0 0 0 14
10.50 0 3 9 11 8 5 2 1 1 0 0 0 40
9.50 1 13 27 24 15 8 4 2 1 1 0 0 96
8.50 4 39 57 38 19 8 3 1 1 0 0 0 170
7.50 13 80 76 37 14 5 2 1 0 0 0 0 228
6.50 32 99 61 21 6 2 1 0 0 0 0 0 222
5.50 44 70 28 7 2 0 0 0 0 0 0 0 151
4.50 32 22 6 1 0 0 0 0 0 0 0 0 61
3.50 8 2 0 0 0 0 0 0 0 0 0 0 10
2.50 1 0 0 0 0 0 0 0 0 0 0 0 1
Total 135 329 267 143 68 31 13 6 4 1 0 0 997

The composite scatter diagram then forms the basis for representing the wave climate
associated with the design operational profile, giving the relative proportion of the wave
encounters which will have the given height and period (significant wave height and mean
crossing period) given by each cell in the scatter diagram.
C.3.1.4.2(f) Sea States versus Wave Height
As noted above, wave statistics and other operational profile data may be available in
terms of wave heights, sea states, or some combination of the two. If a mix of data is available, a
transformation of wave heights to sea states (or vice-versa) will be required. For example, this
can be achieved by assuming that significant wave heights follow a three parameter Weibull
distributions, where:
 (H − m ) 
− S h
F (H S ) = 1 − exp  
q
(C.3.5)
Values of q, m and h may be found and the probabilities of sea states determined from:
ƒ(Sea State) = F(Hs2 ) - F(Hs1 ) (C.3.6)
This process is demonstrated in the Examples - Section D. Similarly, the wave period
data may not be expressed in the same way in a scatter diagram as in a spectral representation
(periods may be peak or zero-crossing) and may need to be converted prior to use.
C.3.1.5 Determination of Wave Load Distribution
Once the ship and wave data required have been developed, they can be combined (for
the Level 3 method) into a composite "sea operational profile" containing all the information
needed to construct a long-term distribution of the loads
Two examples of composite wave scatter diagrams have been presented above. To
illustrate the overall process, a third composite has been constructed for the ship whose
operational data was presented in tables C.3.1.3.1–C.3.1.3.3. To simplify the presentation, the
period data has been excluded and only the wave heights are shown.
The composite wave height probability distribution (Table C.3.1.5.1) for distribution of time
in the relevant sea areas (Table C.3.1.3.3) is combined with the conditional probability of vessel
speed and wave height (Table C.3.1.3.1) according to the expression:
fmc ⋅ fV (VHs) = fs (C.3.7)

where fs = joint probability of significant wave height and speed (Table C.3.1.5.2)

The term (fV (VHs) is calculated from Table C.3.1.3.1 as follows:
fV (VHs ) = prob(V and Hs) / prob (Hs ) (C.3.8)
where: prob(V and Hs) is the joint probability of speed and wave height (individual entries in the
Table C.3.1.3.1), and prob (Hs) is the marginal probability of wave heights, shown in the bottom
row of Table C.3.1.3.1 as bolded numbers. Table C.3.1.5.2 presents the results of the
calculation of fs :
Table C.3.1.5.1: Composite Distribution of Wave Height Probabilities (fmc)

Hs [m] Marsden Combined
0-1 0.1131
1-2 0.2970
2-3 0.2660
3-4 0.1634
4-5 0.0849
5-6 0.0407
6-7 0.0188
7-8 0.0087
8-9 0.0041
9-10 0.0020
10-11 0.0008
11-12 0.0006
Table C.3.1.5.2: Two -Dimensional Joint Probability Distribution (fs )

SPEED Significant Wave Height ( m)
(kn.) 1 2 3 4 5 6 7 SUM
0-6 0.0014 0.0000 0.0014 0.0000 0.0000 0.0006 0.0062 0.0096
6-10 0.0028 0.0012 0.0053 0.0148 0.0075 0.0154 0.0082 0.0552
10-14 0.0065 0.0082 0.0800 0.0686 0.0392 0.0527 0.0164 0.2717
14-18 0.0457 0.0472 0.2103 0.1826 0.1167 0.0570 0.0041 0.6636
SUM 0.0564 0.0566 0.2970 0.2660 0.1634 0.1256 0.0349 1.0000
The third parameter, heading, is then incorporated in the analysis. The procedure is similar
to that for speed, but is combined directly in the expression to give a total probability (three-
dimensional probability), as calculated from the expression:
fstotal = fs fθ (θHs) (C.3.9)
where fθ(θHs) is the conditional probability of heading (θ) for given wave height Hs or sea
state.

The latter term is calculated from Table C.3.1.3.2 as follows:
fθ (θHs) = prob(θ and Hs) / prob (Hs ) (C.3.10)
The values of fθ calculated for every entry from Table C.3.1.3.2 are multiplied by each entry
in Table C.3.2.2. In this example, there are four ranges of speeds x five headings = 20 results.
This must be repeated for each of seven wave heights. Thus, a matrix of three-dimensional
probability of simultaneous occurrence of speed (V), heading (θ) for the given wave height or
sea state (Hs) in the combined operating area has 5 x 4 x 7 = 140 entries. Table C.3.1.5.3 gives
the results of the calculations. The values shown in Table C.3.1.5.3 are probabilities
standardized by multiplying by 1000. Thus the probability of occurrence of Sea State 2 in head
seas, with a vessel speed between 14-18 knots is 3.504/1000 = 0.003504.
When (as in this case) the ship data is drawn from a relatively short period of log data, it may
well be that zero joint probabilities appear to exist for certain combinations of conditions. It is
then necessary to make some assessment of how to handle these. In this example, the lack of
specific ship records for wave heights above 7 m should not be taken to imply that such
conditions will never be encountered, as the composite wave data goes up to 12 m seas for the
areas under consideration. In Table C.3.1.5.3, the joint probability data for the 7 m wave height
has therefore been applied to all higher wave heights. Some such adjustment is necessary, as
these conditions may contribute significantly to fatigue damage (and to ultimate strength
assessments). On the other hand, it is less important to “fill in the blanks” at the lower end of the
range, where the potential inaccuracies are less significant.
Selective elimination of cells in the matrix by combining their ranges and probabilities
can reduce the overall computational effort required considerably. For each cell retained, a set of
possible wave periods will also need to be considered, as shown in Tables 3.1.4.1 and 3.1.4.2.
Each speed, heading, wave height, and period combination will generate its own response
spectrum. A full analysis of this matrix could thus require over 1000 separate calculations to be
undertaken and their results combined. As shown in the examples in Section D, a reduced scope
will normally be quite sufficient.

Table C.3.1.5.3: Operational Profile Summary Table

SPEED Sea State 1 Sea State 2
(kn.) Head Stb. Stb. Stb. Following Head Stb Stb. Stb. Following
Seas Bow Beam Quart. Seas Bow Beam Quart.
0-6 0.109 0.301 0.364 0.542 0.126 0.000 0.000 0.000 0.000 0.000
6-10 0.214 0.594 0.717 1.067 0.248 0.086 0.246 0.283 0.443 0.103
10-14 0.493 1.367 1.650 2.457 0.571 0.613 1.749 2.009 3.144 0.730
14-18 3.447 9.564 11.546 17.192 3.998 3.504 10.004 11.492 17.986 4.175
SUM 4.3 11.8 14.3 21.3 4.9 4.2 12.0 13.8 21.6 5.0

Head Stb. Stb. Stb. Following Head Stb. Stb. Stb. Following
(kn.)
0-6 0.125 0.286 0.330 0.548 0.123 0.000 0.000 0.000 0.000 0.000
6-10 0.467 1.066 1.232 2.045 0.461 1.285 2.968 3.568 5.686 1.293
10-14 7.089 16.186 18.702 31.036 6.990 5.956 13.757 16.538 26.351 5.991
14-18 18.637 42.549 49.163 81.588 18.376 15.857 36.628 44.032 70.159 15.950
SUM 26.3 60.1 69.4 115.2 26.0 23.1 53.4 64.1 102.2 23.2

Head Stb. Stb. Stb. Following Head Stb. Stb. Stb. Following
(kn.)
0-6 0.000 0.000 0.000 0.000 0.000 0.061 0.103 0.137 0.221 0.049
6-10 0.639 1.530 1.818 2.898 0.661 1.641 2.779 3.690 5.938 1.308
10-14 3.317 7.940 9.430 15.035 3.428 5.633 9.540 12.669 20.388 4.490
14-18 9.891 23.676 28.121 44.833 10.223 6.085 10.306 13.686 22.024 4.850
SUM 13.8 33.1 39.4 62.8 14.3 13.4 22.7 30.2 48.6 10.7
SPEED Sea State 7 SPEED

(kn.)Head Stb. Stb. Stb. Following (kn.)
Seas Bow Beam Quart.
0-6 0.941 0.843 1.651 2.268 0.456 0-6
6-10 1.254 1.125 2.201 3.024 0.608 6-10
10-14 2.509 2.249 4.402 6.049 1.216 10-14
14-18 0.627 0.562 1.100 1.512 0.304 14-18
SUM 5.3 4.8 9.4 12.9 2.6 100

C.3.2 Determination of Load Ranges
C.3.2.1 Level 2 – Method

In the simplified (Level 2) analysis method, the loads are defined in such a way as to
allow the application of a statistical method, typically based on a Weibull distribution, to
determine the distribution of the load amplitudes and frequencies. For most structural details, the
focus will be on the hull girder analysis, but methods for calculating other components discussed
above are presented.
C.3.2.1.1 Design Load Equations

The designer will identify design load equations from the various Classification Society
rules or specific sets of rules such as those available for warship design, [Ref. C.14]. Designers
will also know that such rules are often somewhat empirical, cannot necessarily be interchanged,
and, must be applied intelligently, with an understanding of the underlying assumptions and their
expected scope of application. This information may be provided in Class notes or through
papers in the open literature. The use of a set of equations from any single source will generally
result in load (stress) levels which are broadly comparable for a given ship configuration. Thus,
any set of load equations can provide the basis for subsequent fatigue design.
Examples of design load formulae are provided below. These formulae, (taken from Ref.
C.3) are certainly not the only equations available. If a design is being undertaken with a choice
of Classification Society already finalized, it is advisable to use their preferred formulae. It
should be noted that Classification Society formulae for fatigue loads are not always identical to
those provided for basic scantling design.
The overall approach is to define the loads in terms of the following components:
• vertical bending moment;

• horizontal bending moment;
• torsional bending moment;
• dynamic external pressure loads;
• internal pressure loads due to ship motions.
Each of these may need to be considered under a set of ship loading conditions, depending on the
type of vessel under consideration. Examples of design load equations for use in Level 2
analysis are provided in Appendix B.
For naval vessels, fatigue calculations normally only take into account wave bending
effects, though local machinery and propeller-induced vibrations may also be important. Most
naval design standards incorporate similar parametric formulae to those shown in Appendix B
for commercial vessels.

C.3.2.1.2 Determining Loads for Fatigue Analysis

Some approaches to fatigue design also use the “lifetime extreme” stress range as a
reference level, but the majority (including this Guide) use a higher exceedence
probability/lower value to reflect the fact that most damage accumulates at lower stress ranges.
In order to convert values from one probability level to another probability level, a correction
factor, ff (using Weibull distributions) is required. ff is given by:
ff = {(ln p2 )/(ln p1 )} 1/h (C.3.11)

where:
p2 = reference probability level to which the load is to be changed (10-4 in this Guide)
p1 = probability level at which the quantity is specified (generally 10-8 as discussed above)
h = Weibull shape parameter
Specifically, the factor ff is applied to the load equations to correct them for the revised
probability of occurrence.
The Weibull Shape parameter, h, is dependent on the location of the structure and may be
established from long-term wave load analysis. The following values are given in Ref. C.3, for
tankers and bulk carriers:
ho , the basic long-term Weibull shape parameter = 2.21 - 0.54 log(L)
(1) For deck longitudinals: h = ho

(2) For bottom longitudinals: h = ho - 0.005 Tact
(3) For external pressures:
Ship side above the waterline: h = ho + 0.05((Dm-zb)/(Dm- Tact ))
Ship Side at Waterline: h = ho + 0.05
Ship Side below the waterline: h = ho + 0.05(zb / Tact ) - 0.005( Tact-zb)
(4) For Internal Pressures on Longitudinal and Transverse bulkheads:
h= h0 + 0.05
The above equations are for vessels with a typical roll period (natural roll period <14
seconds). For vessels with a long roll period, the value 0.05 in the above equations can be taken
as zero.
If information is available from similar designs/operational profiles to suggest that the
distribution shape parameter for any or all loadings should be changed, then alternative values
can be used in the above correction factors. Similarly, different probability ratios can be
introduced if desired.
The general assumption in all simplified (Level 2) approaches is that loads (stress ranges)
can be combined by simple superposition (or minor modifications to this) and the same spectrum
shape applies to the combined loads as to the principal components. While not strictly true, this
assumption is adequate for most cases.

C.3.2.1.3 Application of Direct Calculation Methods in the Level 2 Approach.

Under current Class requirements, the loads for certain types of vessel must be analysed
using direct calculation methods rather than specified parametric design equations such as those
discussed above. This applies, for example, to tankers over 190m length. Classification
Societies have developed specialized software for this, including the ABS Safehull system [Ref.
C.15]; the LR ShipRight suite [Ref. C.16]; GL’s Poseidon system [Ref. C.17]; and DNV’s
Nauticus program [Ref. C.18]. These methods have several characteristics in common,
including:
• computer-based systems
• explicit calculation of extreme loads
• explicit calculation of structural resistance
• capability to assess fatigue performance
The methods incorporate a fatigue design component, generally following the process
outlined under the Level 2 Method in Figure C.2.2.1, i.e., they are based on an estimate of
extreme load derived, and the stress range spectrum is based on the Weibull model.
The simplifications incorporated in these methods mean that they are not generic in terms
of the ship types to which the resulting software can be applied. To reduce computational
demand, first-principles results are synthesized by ship type and structural configuration to
simpler algorithms. The Classification Societies have developed systems primarily for larger
vessels, the main ship types being tankers, bulk carriers, and container ships (this varies between
Class Societies). Implicit in many of the systems is an assumption of a specific structural
configuration.
The designer can employ the same basic technique using other, non-proprietary methods.
In all cases, this will involve identification of a design sea condition, and the development of
response amplitude operators (RAO’s) for each load component, as per the procedures defined in
Section C.3.2.2. Alternatively, there are published parametric equations for hull girder design
moment derived from RAO’s by Sikora et al. [Ref. C.19 and C.20], that are appropriate for
warships and high speed commercial vessels. This approach is described later.
C.3.2.2 Level 3 – Method

C.3.2.2.1 Basis of the Approach
The basis for developing loads in the spectral analysis method, the core of the Level 3
approach, is the development of transfer functions generally referred to as Response Amplitude
Operators, or RAO’s. An RAO represents the response of the ship’s structure to excitation by a
wave of unit height, and is derived over the full range of (encounter) frequencies that will be
experienced. RAO’s are complex numbers with real and imaginary components that express the
amplitude and phase relationship between the wave load (forcing function) and the response.

This Guide does not develop the mathematical basis for the various ship response RAO’s.
There are a variety of suitable references [Ref. C.21-C.23]. In general, the numerical calculations
are computationally demanding and require a computer code. Empirical (model test or parametric)
representations have, however, been developed for some specific ship types.
For a Level 3 fatigue analysis:
(1) An RAO should be developed for every combination of speed and heading represented in
the operational profile, as well as for every load component.
(2) Vessel response is obtained for every individual condition (combination of speed,
heading and sea condition). The RAO’s are applied to each cell in the wave scatter
diagram, using the wave height and period as characteristic values with the selected sea
spectrum. Each resulting load response spectrum represents the short-term response to
the conditions defined in the specific cell. As a short-term response must be obtained for
every condition in the operational profile, a large number of spectral analyses is required.
(3) The fatigue assessment is a long-term analysis that employs all the data given in an
operational profile. Each individual response spectrum is multiplied by the probability of
that combination. The end result is a statistical distribution due to wave loading. This
approach is then used to predict the probability of structural failure due to progressive
damage accumulation.
Regarding (1), the overall approach is simplified considerably if it is assumed that the
ship responses are linear, allowing a single set of RAO’s to be used to cover the full range of
wave conditions defined in the wave scatter diagram. In extreme seas, the vessel response may
be highly non-linear, and thus in principle, different RAO’s should be used to calculate the
response. However, for the purposes of fatigue analysis, the extreme value loads represent a
small number of cycles in the total life, and thus the (non-conservative) error is generally
acceptable.
C.3.2.2.2 Calculation of Response Amplitude Operators (RAO’s)

An RAO needs to be developed for each of the load types which are considered in the
fatigue analysis. These are expressed in the form:
Hv (Τ|2) = RAO for vertical bending moment,
Other RAO’s and terms are defined in the Nomenclature.
The RAO’s for hull girder response (bending moment and shear) are derived from section
properties, the external forces due to hydrostatic restoring forces, the hydrodynamic response of
the hull, and the incident wave excitation force.

As indicated by the parametric equations B.9, B.10 in Appendix B, the equations for
external and internal pressure loadings are strongly dependent on the hydrodynamic (vessel
motion) response and acceleration components. Thus, the basic RAO will be derived from the
RAO for the vessel’s motion response, and will rely on a conversion function to pressure, as
outlined in Section 3.3.1.1. In addition, there may be a “dynamic” component due to the incident
wave impacting on the structure externally, or internal sloshing of liquids. This load component
is developed using standing wave theory, as developed in Ref. C.23 or C.24, for an incident
wave on a vertical wall. As this phenomenon is specific to the higher frequency ranges that are
not coincident with the peak response to static pressure variation, it has been proposed that a
combined RAO is created. This is shown in the third example, Section D.
RAOs can be determined by model tests, full-scale measurements, or by computer

programs. Currently, the most commonly used method for predicting ship motions and sea
loads, involving the computation of RAO’s, is linear strip theory. Several computer programs
based on this theory are readily available (e.g., Ref. C.25).
The limitations of strip theory may be overcome with the use of computer programs
which use computational fluid dynamics approaches, wave diffraction theory, hull geometry
panel methods, and time-domain based codes. These programs are gaining popularity, but are
more complex, and the end result may not justify the effort required to generate the data in any
situation other than for research or for unusual hull forms. In the context of a fatigue analysis,
the limitations associated with the use of strip theory are generally offset by the narrow range of
conditions that are affected by those limitations, and by the speed of computation (given that
multiple short-term analyses must be generated under the Level 3 approach).
As an alternative to numerical or physical modeling of the RAO’s, parametric equations

have been developed for specific vessel types and loads. For example [Ref. C.20] provides such
data for hull girder bending for warships and high speed commercial craft.
This method relies on a generalized response amplitude operator for vertical bending
moment at midships. For a set of ships consisting of frigates, destroyers, and high speed
commercial vessels, the RAO is, after appropriate normalization, represented by a single curve,
where
The ordinate of response is: (√RAO)/ρg (LBP )2 BF1 F2

The abscissa of frequency is: ΩF3 (C.3.12)
where:
F1 = | 3 √cosθ|
F2 = 1.1tanh(1.5 + V/g) + 0.03(V/g) 2
Ω = ω/(√(2πg/ LBP )
F3 = √cosθ

Figure 3.2.1 is taken from Ref. C.20 and illustrates the nature of the data.
RAO 0-7 knots 9-12 knots 14-17 knots

ρgL BF1F2
2 0.02
where:
F1 = 3 Cosθ 0.015
F2 = 1.1 Tanh(1.5 + v/g) + 0.03 (v/g) 2
F3 = Cosθ 0.01
ω
Ω=
2πg / L 0.005 Oiler, At Sea
Tanker, Model
ρg = density of sea water Ore Carrier, Model
θ = heading (0 = head seas)
0
v/g = speed per grav. constant, sec.
0.4 0.9 1.4 1.9
Non-Dimensional Wave Frequency, Ω F3
Figure 3.2.1: RAOs for Oilers, Tankers and Ore Carriers
C.3.2.2.3 Determining Load Ranges

In the Level 3 approach, the objective is again to provide a load range. Each loading
combination in the operational profile must ultimately be converted from an amplitude to a
range.
C.3.2.2.3(a) Wave- Induced Hull Girder Bending Moments

Vertical Bending moment is generally assumed to be fully reversing, i.e., any variation
between hog and sag is not explicitly considered. This is a function of the basic linear theory.
The vertical bending moment range will be twice the magnitude of the response calculated from
the spectral analysis based on the single amplitude RAO for vertical bending.
The horizontal (and torsional) wave bending moment range is not fully reversing. Using
an analogous approach to the Level 2 discussion, the horizontal bending range can be equated to
the single amplitude horizontal bending response.
C.3.2.2.3(b) External Pressure Range

The Level 3 external pressure range is calculated from two separate transfer functions
consisting of:

(1) The range of quasi-static head at a point of interest on the side shell due to vertical
motions of the vessel, is:
∆pst = ρg(z2 – z1 ) (C.3.13)
where z1 , z2 are the instantaneous immersions of the POI at the extreme points of motion.
The prefix “∆” refers to a “range” value of given parameters where z is derived from the
local vertical motion RAOs for the point of interest. It is recommended that the single
amplitude response be used due to intermittent emergence of the point of interest, noting
that when the point of interest emerges from the water, the outside pressure falls to zero.
(2) The dynamic pressure, p d, at the point of interest (POI) – due to the incident wave and
reflected wave in deep water given by Equation C.3.14, with terms as defined in Figure
3.2.2, and dW = ∞:
1 
p d = 2 ?gH cos (kx ) cos (? t ) (C.3.14)
2 
where k is the “wave number” = ω2 /g in deep water.
Figure 3.2.2: Dynamic Pressure on Side Shell Scenario

The resulting pressure range transfer function is calculated as the difference between the
total pressure at the wave crest and wave trough:
∆pd = pd (wave crest) - pd (wave trough) (C.3.15)
These transfer functions are directly calculated as ranges, as they are expressed as a
pressure difference. As phasing information is generally unavailable, separate fatigue
damage assessments will be conducted for each pressure component, each taken over the
life of the vessel.
(3) The combined (if phase data is available) or individual pressure component is then
defined as the “effective lateral pressure” (p) in subsequent stress calculations.
C.3.2.2.3(c) Internal Pressure Loads due to Ship Motion

For reasons analogous to the external pressure calculation, the internal pressure RAO
should also be developed as a range. In this case, the designer must be concerned about: (i)
saturation of the tank; and, (ii) combination of quasi-static pressure variation and sloshing
effects.
C.3.2.3 High Frequency Loads - Slamming, Whipping and Springing

In general, the magnitude of high frequency loads that result from slamming (whipping)
or springing will be significantly smaller than the wave bending loads. Therefore, unless
whipping or springing is expected to occur frequently, they can generally be neglected for
fatigue analysis.
Springing is a steady-state phenomena, where higher mode hull responses are excited by
general wave interaction. The phenomena is most closely associated with Great Lakes bulk
carriers (Lakers) which feature high Length/Depth ratios [Ref. C.6].
Whipping is a transient effect resulting from bow flare or keel slamming effects and are
location dependent, where slamming is of greatest concern in the forebody. In the calculation of
the extreme value hull girder response, the superposition of transient loads will result in an
increase of the loading over the basic wave-induced loading. This has implications for the
simplified fatigue approach, and the designer must decide whether or not this component should
be carried through the Weibull analysis on the basis of the frequency of occurrence of slamming
over the load range (and the extent to which slamming effects may already be accounted for in
empirical load formulae).
If these types of transient effects are expected to be significant contributors to fatigue
damage, complex representations of ship dynamic response is needed to quantify these with
reasonable precision. For a restricted range of fast ship forms (warships and others) a simplified
method can be found in Ref. C.20.
Alternatively, slamming may be handled by manipulating appropriate parts of the load
spectrum. In the case of both slamming and springing, it may be more appropriate to add an
additional safety factor to the “standard” prediction rather than attempt the full analysis.

C.3.3 Summary
For the Level 2 Method, the load amplitudes to be used in the subsequent development of
stress ranges are generally determined from well-recognised parametric equations with no
explicit determination of the ship operations or the wave climate. For example the vertical
bending moment, Mv , may be derived from formulae published in Classification Society rules,
and modified to a more appropriate level of exceedance probability as shown herein.
Load amplitudes to be used in determining stress ranges may require adding calculated
loads (such as with vertical bending – in which the load range is the addition of the sagging and
hogging moments) or, as with horizontal bending, selecting the calculated single amplitude
value.
Loads for side and internal structure are likely to be expressed as a pressure, for
subsequent treatment in determining stresses using lateral loading analyses techniques.
For the Level 3 Method, a more rigorous examination of the ship operational profile and
wave climate is used to produce statistical distributions of load under all the various operating
conditions. Response Amplitude Operators (transfer functions) are developed for each load
component of interest. In some cases, more than one RAO must be generated to define a load
component, e.g., side shell pressure loads. Seakeeping programs may be used to determine the
RAOs. The array of operating conditions (generated from the operational profile and wave
scatter diagram) and the associated RAOs, then comprise components used to derive a stress
range spectrum to be developed in the subsequent sections.

C.4 STRESS ANALYSIS
C.4.1 Introduction
This section provides procedures for relating the ship load distributions or spectra,
developed in Section C.3, to the local response of the structural detail of interest for use in the
design of fatigue resistant details. This involves developing “ship load” to “structural detail hot
spot principal stress” transfer (or response) functions.
The fatigue design processes, presented in this Guide, are based on stress transfer
functions or coefficients relating the local stresses to the global hull girder bending moments,
external sea pressures acting on the hull, and internal pressures acting on the tank boundaries.
The stress coefficients are evaluated by calculating the local field stresses at the point of interest
for a unit value of each load component (e.g., vertical, horizontal and torsional bending moment
loads, internal and external pressure loads). In general, this will involve conducting stress
analyses for unit loads considering each type of loading individually. Strictly, the stress
coefficients are a function of wave frequency. However, it appears [Ref. C.3]) that it is
acceptable practice to compute stress coefficients for one particular wave frequency, and heading
for that matter, and apply it to all wave frequencies and/or headings. The total stress spectrum at
the location of interest can then be estimated by combining the stress coefficients and loads
spectra using the methods outlined in Section C.2. The remainder of this section outlines
methodologies to determine the local stresses from unit loads, hence determining the stress
coefficients.
The processes covered in this section are the third and fourth steps outlined in the flow
chart for the fatigue design procedures (Figure C.2.1). The two-step process, described below,
may be completed via the Level 2 (simplified) or Level 3 (direct spectral analysis) routes or a
combination of the two.
Level 2 Level 3
1) Define reference values for nominal 1) Determine transfer functions for nominal
stress range components in primary stress range components in primary
structure using parametric equations, structure by global finite element
frame and beam models analysis
2) Use primary defined SCF’s to determine 2) Determine transfer functions for
reference values of hot spot stress range components of hot spot stress range by
components local finite element analysis
The cross-over linkages, shown in Figure C.2.1, are intended to indicate that any
combination of the first and second steps of the Simplified and/or Spectral Analysis approaches
may be used in the design of fatigue resistant details. The ability to cross over affords the
designer the freedom to use any combination of analytical detail appropriate or available in the
design process.

As in the other elements of the fatigue design process, there are a number of approaches
with varying degrees of complexity and accuracy that may be used. The approach selected for
the stress analysis should, in general, be consistent with the complexity and accuracy applied to
other elements of the design process.
C.4.2 Stress and Stress Concentration Category Definitions

In the design of ship structures, both stresses and structural elements may be classified to
simplify and isolate structural elements or their response to applied loads. The following
definitions of primary, secondary and tertiary stresses [Ref. C.26] will be used in this Guide:
• σ1 - Primary - stresses due to bending, shear and torsion in the main hull girder;
• σ2 - Secondary - stresses in a stiffened grillage due to bending and membrane effects; and,
• σ3 - Tertiary - membrane stresses in panels between stiffeners.
The utility of separating out classes of structure or stresses is less important when numerical
(finite element) analysis methods are used to analyse the response of the entire structure. The
classification approach is important when stresses are superimposed to estimate the response of
structural elements to the applied loads.
In general, the fatigue life of a structural detail is a function of the fluctuating stress field
at the point of interest. These stresses may be treated directly, or resolved into four components
as shown in Figure C.4.2.1 and described below.
Figure C.4.2.1: Stress Components in a Welded Joint

C.4.2.1 Local Nominal Membrane and Bending Stress (σm and σb)
The local nominal membrane stress is the uniformly distributed stress that is equal to the
average value of stress across the section thickness. The local bending stress is the component of
nominal stress due to applied loading that varies linearly across the section thickness. The
nominal stresses satisfy the simple laws of equilibrium of forces and moments from applied
loads. They may be derived from simple formulae, beam element models, or coarse mesh finite
element analysis (FEA) as described in Section C.4.4.2. The term “local nominal stress” is used
because stress concentrations resulting from the gross shape of the structure surrounding the
local detail of interest will affect the local stress field magnitude (e.g., shear lag effects) and must
be included in the local nominal stresses.
C.4.2.2 Peak Stress (σp )
The peak stress is the component of stress due to applied loads due to stress
concentrations at local discontinuities in the vicinity of the crack. The peak stress represents the
highest value, usually at the surface at a notch (e.g., weld toe). Peak stresses arise from stress
concentrations due to the following effects:
(1) Geometric Stress Concentrations (Kg ): due to the gross geometry of the detail
considered. The effect of the geometric stress concentration typically decays over
distances of the order of the section thickness.
(2) Notch Stress Concentrations (Kw): due to the local geometry of the notch (e.g., weld
geometry). The effect of the notch stress concentration typically decays over distances of
the order 10% to 20% of the section thickness. Notch stress concentrations are not
considered in the structural detail safe life fatigue design process in this Guide.
(3) Misalignment Stress Concentrations (Kte , Ktα ):: due to bending stresses caused by
misalignments including eccentricity tolerance (Kte), and angular mismatch (K tα). These
are normally used for plate connections only. The effect of the misalignment stress
concentrations typically decay over distances of the order of the section thickness.
C.4.2.3 Residual Stresses (σr)
Residual stresses are local self-equilibrating stresses that arise from fabrication and
welding. In general, residual stresses are strain/displacement limited phenomena and, as such,
do not contribute to plastic collapse if they relax. However, they do add to the tensile stress field
in the vicinity of the crack and have to be included in the calculation of the stress intensity factor
for residual strength assessments. Residual stresses may also be resolved into membrane and
bending components. However, since there is only limited quantitative data on the distribution
of welding residual stresses in ship structural details, it is normal practice to assume a uniform
(membrane) residual stress field approaching tensile yield strength (i.e., σr ≈ σy ). Residual
stresses need not be considered in the Miner’s summation fatigue design process being presented
in this Guide since their effect on the mean stress level should be accounted for in the S-N curve
or fatigue life data.

C.4.2.4 Total Stress

The total stress is the sum of the various stress components. The maximum value of total
stress at the crack location is referred to as the peak total stress (σtp).
C.4.2.5 Stresses in Fatigue Analysis

As mentioned in the previous sections, the stress analysis for fatigue design does not
necessarily incorporate all stress concentrating effects. While this Guide is based on a “hot spot”
stress approach, other rules may be based on nominal or notch stress stresses. The difference
between these approaches lies in how the experimental fatigue test data is reduced in developing
the design S-N curves. In order to correctly determine the stresses to be used in fatigue analyses,
it is important to note the definition of the stresses used in the S-N curve being applied. Possible
stress definitions include:
• Nominal stresses are those derived from beam element models or from coarse mesh FEM
models. Stress concentrations resulting from the gross shape of the structure, e.g., shear lag
effects, are included in the nominal stresses derived from coarse mesh FEM models.
• Geometric stresses include nominal stresses and stresses due to structural discontinuities
and presence of attachments, but excludes stresses due to presence of welds. Stresses
derived from fine mesh FEM models provide geometric stress data. Effects caused by
fabrication imperfections (e.g., misalignment of structural components), are however
normally not included in FEM analyses, and must be separately accounted for. The greatest
geometric stress value at the point of interest (including fabrication imperfections), the
weld toe, is commonly denoted hot spot stress and the location of the greatest geometric
stress outside the region affected by the geometry of the weld is termed the hot spot.
• Notch stress is the total stress at the weld toe (hot spot location) including the geometric
stress, misalignment effects and the stress due to the presence of the weld.
C.4.3 Determination of Local Nominal Stresses

As discussed in the previous section, local nominal stresses are those that would be
calculated at the location of interest in the absence of the stress concentration due to the local
structural detail and weld. The local nominal stresses include the stress concentration effects of
the overall geometry of the structure surrounding the detail, but not the detail itself.
The local nominal stresses may be calculated, for unit loads, using a combination of
parametric formulae for simple structural assemblies and global stress concentration factors to
account for the gross geometry of the structure and the effects of misalignment. This approach is
referred to in this Guide as the “Level 2” approach.

C.4.3.1 Level 2 Approach

Calculation of hull girder stresses is the simplest way of getting reasonable
approximations to the stress levels in longitudinal hull girder elements and connections and may
be used for a quick evaluation of stress levels in important details. Global hull girder stresses
may be calculated based on gross scantlings. Local stress components should be calculated
based on net scantlings, i.e., gross scantlings minus corrosion allowances. The hull girder
analogy is based on simple beam theory in which:
• plane cross-sections remain plane;

• stresses remain in the elastic range and thus allow superposition;
• the beam is essentially prismatic (no openings or discontinuities); and,
• there is no interaction between bending and other response modes (e.g., transverse and
longitudinal deflections or shear and torsional distortions).
Formulae for calculating hull girder stresses are included in Classification Society Rules.
Alternatively, the following formulae derived from formulae presented in [Ref. C.3] may be
used. The approach presented here estimates nominal stress levels based on the cumulative
contributions of primary (hull girder), secondary and tertiary bending stresses, as shown in
Figure C.4.3.1. Secondary and tertiary bending stresses are the result of local bending of
structural members due to lateral pressure.
The simplified analysis formulae, presented in the sections which follow, relate the
design loads (bending moments and pressures) to nominal stresses based on the geometry of the
vessel structure, as defined in Figure C.4.3.2.
Figure C.4.3.1: Simplified Stress Analysis of Hull Girder [Ref. C.3]

a
sb
b sa
(a) Double Skin Configuration
z x(a)
h sb
y(b)
t sa
(b) Double Bottom Configuration
z
rb x(a)
sb
y(b) ra
sa
(c) Single Bottom Configuration
Figure C.4.3.2: Definition of Geometric Parameters for Hull Configurations

[Ref. C.3]

C.4.3.1.1 Primary Hull Girder Bending

The horizontal bending stress (σm,h) and vertical bending stress (σm,v) at a particular point
in the ship hull can be combined as follows to account for phase differences between the two
stresses:
s = s 2m, h + 2 e s 2m, h s 1,2 v + s 2m, v (C.4.1)
where ε is the stress correlation coefficient. The correlation coefficient is zero in head seas and
nearly unity in beam and quartering seas because horizontal bending in these seas is mainly due
to healing of the ship. For design purposes, it is sufficient and conservative to simply add the
horizontal and vertical bending stresses.
In the vertical and horizontal bending stress analysis formulae presented below the
following structure should be included in the calculation of the section moment of inertia, section
modulus or neutral axis location:
• Deck plating (strength decks and other effective decks);
• Shell and inner bottom plating;
• Longitudinal bulkheads and girders; and,
• Longitudinal stiffeners.
Only longitudinally effective structure should be considered in these calculations. For a
structural element to be longitudinally effective it must extend over sufficient length for some
portion of the longitudinal stress field to enter the structure. An approximate rule of thumb is
that longitudinal stresses diffuse at a rate of 1 in 2 in a web (e.g., side shell or longitudinal
bulkhead) and about 1 in 4 in flanges (e.g., deck or bottom structure), therefore, the structure
must be continuous for twice or four times its width, respectively in order to be considered fully
effective.
For vertical hull girder bending:

σ mv = KG Mv zna / I v (C.4.2)
and, for horizontal hull girder bending,:
σ mh = KG Mh yna / I h (C.4.3)

C.4.3.1.2 Secondary/Tertiary Stresses due to Internal and External Pressure Loads

Local secondary bending stresses are the results of bending, due to lateral pressure, of
stiffened single skin or double hull cross-stiffened panels between transverse bulkheads (see
Figure C.4.3.2). This approach may be applied to bottom or deck structures, sides or
longitudinal bulkheads.
The preferred way of determining secondary stresses is by means of Finite Element
Analysis (FEA) or by frame analysis models. Alternatively, secondary bending stresses may be
estimated from parametric equations such as the following equations [Ref. C.3]. Similar
equations are given in Ref. C.1. The secondary and tertiary bending stresses are presented in the
following five categories:
(a) Longitudinal Secondary Bending Stress in Double Bottom Panels;
(b) Transverse Secondary Bending Stress in Double Bottom Panels;
(c) Secondary Bending Stress in Single Skin Panels;
(d) Bending Stress of Stiffeners Between Transverse Supports (e.g., Frames,
Bulkheads); and,
(e) Tertiary Bending Stress of Plates Bounded by Stiffeners.
The effects of these secondary and tertiary stresses are cumulative with the local nominal
stress levels calculated from the hull girder analysis.
(a) Longitudinal Secondary Bending Stress in Double Bottom Panels
Longitudinal secondary bending stresses in double bottom panels at the intersection of
transverse bulkheads may be estimated by the following formulae:
• Double Bottom Wider than Long (b > a): [Case 1 and 2, Table C.4.3.1]
σ2 = (K b p b2 ra ) / √(ia i b ) ρ =(a/b) (i a / i b )1/4 (C.4.4)
• Double Bottom Longer than Wide (a > b): [Case 3 and 4, Table C.4.3.1]
σ2 = (K b p a2 ra ) / ia ρ= (b/a) (ib / i a )1/4 (C.4.5)
where:
ia = Ia/sa
ib = Ib/s b
(b) Transverse Secondary Bending Stress in Double Bottom Panels

Transverse secondary bending stresses in double bottom panels at the intersection of
transverse bulkheads may be estimated by the following formulae.
• Double Bottom Longer than Wide (a > b): [Case 3 and 4, Table C.4.3.1]
σ 2 = (K b p b2 rb ) / i b ρ = (a/b) (ib / i a)1/4 (C.4.6)
• Double Bottom Wider than Long (b > a): [Case 1 and 2, Table C.4.3.1]
σ2 = (K b p b2 ra ) / ia ρ = (b/a) (ib / i a)1/4 (C.4.7)

(c) Secondary Bending Stress in Single Skin Panels

The stresses at transverse and longitudinal bulkheads may be estimated from the same
formulae as for double bottom configurations. However, the parameters ρ and torsion
coefficient η should be taken as given in Table C.4.3.2 (also see Table C.4.3.3 for definitions).
(d) Bending Stress of Stiffeners Between Transverse Supports (e.g. Frames, Bulkheads)
The local bending stress of stiffeners with effective flange between transverse supports
may be estimated by (Figure C.4.3.3):
 M   m EI 
s b = K ⋅   +  2d  rd ⋅ d (C.4.8)
 
 Zs   l e Zs 
 ps l2e 
where: M = moment at stiffener support =  r
 12  p
 
x 
rδ = 1 − 2 
 le
2
 x  x
rπ = 6  − 6  + 1
 le   le 
K = stress concentration factor
mδ = 4.4 at the bulkhead where no stringers or girders support the frames
adjacent to the bulkhead; else mδ must be determined from a beam
element analysis as per the procedure in Ref. C.3.
Zs = section modulus of stiffener
It is of great importance for reliable assessments that bending stresses in longitudinals

caused by relative deformation between supports are not underestimated. The appropriate value
of relative deformation, δ , has to be determined for each particular case (Figure C.4.3.3). This
usually will require 2-D or 3-D frame analysis or coarse mesh FEA.
(e) Tertiary Bending Stress of Plates Bounded by Stiffeners

The local longitudinal tertiary plate bending stress in the weld at the plate/transverse
frame/bulkhead intersection midway between longitudinals is given by:
σb =0.343 p (s / t n )2 (C.4.9)
Similarly the transverse stress at stiffener mid length is:

σb =0.5 p (s / tn )2 (C.4.10)

Figure C.4.3.3: Secondary Stresses in a Stiffener [Ref. C.3]

Table C.4.3.1: Support Bending Stress Coefficients Kb - Double Bottom Panels [Ref. C.3]
Case no. & Stress Boundary Conditions ρ η = 0.0 η = 0.5 η = 1.0
Location
Case no. 1: Long edges: 1.00 0.0952 0.0845 0.0767
Support bending Simply supported 1.25 0.1243 0.1100 0.0994
stress in long 1.50 0.1413 0.1261 0.1152
direction at middle of Short ends: 1.75 0.1455 0.1342 0.1251
short end Clamped 2.00 0.1439 0.1374 0.1300
2.50 0.1388 0.1381 0.1356
3.00 0.1371 0.1376 0.1369
3.50 0.1371 0.1373 0.1373
4.00 0.1373 0.1374 0.1373
& up 0.1374 0.1374 0.1374
Case no. 2: All edges: 1.00 - - 0.0564
Support bending Clamped 1.10 - - 0.0591
stress in long 1.20 - - 0.0609
direction at middle of 1.30 - - 0.0619
short end 1.40 - - 0.0624
1.50 - - 0.0626
1.60 - - 0.0627
& up - - 0.0627
Support bending Clamped 1.33 0.1026 0.0949 0.0878
stress in short 2.00 0.0972 0.0950 0.0926
long edge Simply supported 4.00 0.0912 0.0915 0.0917
& up 0.0916 0.0916 0.0916
Case no. 4: All edges: 1.00 - - 0.0564
Support bending Clamped 1.10 - - 0.0638
stress in short 1.20 - - 0.0702
direction at middle of 1.30 - - 0.0755
long edge 1.40 - - 0.0798
1.50 - - 0.0832
1.60 - - 0.0857
1.70 - - 0.0878
1.80 - - 0.0892
1.90 - - 0.0903
2.00 - - 0.0911
& up - - 0.0911
Notes:
1) For intermediate values, use linear interpolation
2) See Table C.4.3.3 for definitions of ρ & η

Table C.4.3.2: Support Bending Stress Coefficients Kb - Single Skin Panels [Ref. C.3]
Case no. & Stress Boundary Conditions ρ η = 0.0 η = 0.5 η = 1.0
Location
Support bending Simply supported 1.25 0.1140 0.1001 0.0904
stress in long 1.50 0.1285 0.1148 0.1049
short end Clamped 2.00 0.1310 0.1250 0.1191
2.50 0.1263 0.1257 0.1234
3.00 0.1248 0.1253 0.1246
3.50 0.1248 0.1250 0.1246
4.00 0.1240 0.1250 0.1250
& up 0.1250 0.1250 0.1250
Support bending Clamped 1.33 0.0934 0.0858 0.0799
stress in short 2.00 0.0885 0.0865 0.0843
long edge Simply supported 4.00 0.0830 0.0832 0.0835
& up 0.0834 0.0834 0.0834
Notes:
1) For intermediate values, use linear interpolation
2) See Table 4.3.3 for definitions of ρ & η

Table C.4.3.3: Definition of Stiffness and Geometry Parameters [Ref. C.3]

Type Sketch Formulas for ρ and η
A: Cross stiffening I I −I 
Middle girder / stiffener ia = na′ + 2 a na ′ 
sa  b 
in both directions are
I I − I 
stiffer than the others ib = nb ′ + 2 b nb′ 
sb  a 
a ib
ρ= 4
b ia
I pa I pb
η=
I na I nb
B: Modified cross Ia
stiffening ia =2
b
One girder / stiffener in I nb  I b − I nb 
a-direction only ib = + 2 
sb  a 
a ib
ρ= 4
b ia
I pb2 b
η = 0124
.
I a I nbSb
C: Single stiffening ia = 0
Girders / stiffeners in b- I
direction only i b = nb
sb
ρ = inf inite
η = in det er min ate
D: Unstiffened plate t3
ia = ib =
(
12 1 − υ2 )
a
ρ=
b
η = 10
.

C.4.3.2 Level 3 Approach Using Finite Element Analysis

In the design of critical structural elements, or when the global structure is too
complicated for simple parametric formulae, finite element analysis (FEA) may be used to obtain
a reliable description of the overall stiffness and global stress distribution in the hull.
The global FEA is generally carried out with a relatively coarse mesh, the main objective
being to obtain a good representation of the overall membrane panel stiffness in the longitudinal
and transverse directions and for shear, sufficient for determination of nominal stresses.
Stiffened panels may be modelled by means of anisotropic elements or, alternatively, using a
combination of plate and beam elements.
The extent of the model is dependent on the type of response to be considered and the
structural arrangement of the hull. If the FEA based design process involves only several
localized details, the required extent of the local model is dependent on the stiffness variation of
the hull over a certain length and this has to be captured in the global FEA model. The minimum
hull module length required to accurately portray the structural response and provide the
additional information not available from the simplified analysis approach typically includes
several cargo holds (or watertight compartments of a naval vessel). The exact length
requirement depends on the ship’s overall geometry and nature and arrangement of the cargo or
other loads.
For horizontal and torsional bending response of the hull of an open hatch ship, it is
generally required that the extent of the global model covers the complete hull length, depth and
breadth (a half breadth model may be used if antisymmetric boundary conditions can be assumed
at the centerline). A complete finite element model may also be required for the evaluation of
vertical hull girder bending of ships with complex superstructure arrangements (e.g., warships,
passenger ships), and for ships of complex cross-section (e.g., catamarans).
Alternatively, a part of the hull (for example, the midship area) may be modelled. Hull
girder loads should be applied individually at each end of the model to result in a value of unit
load (e.g., bending moment) at the location of interest. Unit pressure loads will normally be
distributed over the appropriate section of the hull. The loads should be balanced in order to give
a minimum of reaction forces at the supports (boundary conditions). The loads and boundary
conditions in the hull cross section should be evaluated carefully when modelling only a part of
the hull to avoid unrealistic stiffness from the forebody/ afterbody.
Figure C.4.3.4 shows an example of a global finite element model of a section of a bulk
carrier. This model may be used to calculate nominal global stresses and deformations away
from areas with stress concentrations. In areas where local stresses in web frames, girders or
other areas (for example hatch corners) are to be considered, the global model should have a
mesh producing deformations applicable as boundary conditions for local stress analysis. In
such cases, the global and local models should be compatible. The local model may be directly
applied as a substructure or super-element in the global model (if such techniques are available
with the FEA software). The substructure technique ensures that forces and deformations in the
global and local models are compatible and, if the substructure is detailed enough, local stress
results may be obtained directly. The substructure technique is very effective where local
structural assemblies (i.e., the substructure) are repeated several times in the overall assembly,
but it does present added complexity into the analysis.

More commonly, the global and local analyses are conducted separately. Nodal forces
and/or displacements obtained from the global model are applied as boundary conditions for the
local model. In general, the stiffness of the local model should be comparable to that of the
global model representation so that forces and displacements between the two models are
compatible. However, due to the greater level of geometric detail and mesh refinement of the
local model, this is rarely achievable. As such, it is preferable that nodal forces be transferred
from the coarse model to the local model rather than forced displacements. It is important that
the extent of the local model is sufficiently large that boundary effects due to prescribed forces or
displacements are away from the areas where accurate stresses need to be determined.
The loads to be applied in the global analysis can be produced using any of the
methodologies presented in Section C.3. The global analysis should be conducted for each load
case individually (i.e., vertical bending, horizontal bending, torsional bending, external pressure,
internal pressure). Each load case should be analyzed for a unit value of the applied load at the
location being considered. In this manner, the stresses derived from subsequent local analysis
will correspond to unit loading and therefore, be equal to the stress coefficients, Ai , which are
required to generate the local stress spectrum from the combined loading spectra.
Figure C.4.3.4: Global Finite Element Model of Bulk Carrier [Ref. C.1]

C.4.4 Determination of Peak Stresses

Peak stresses may be estimated based on parametric approximations of stress
concentration factors for ship details, when these are available. Alternatively, they may be
determined based on local fine mesh finite element stress analysis of the joint.
C.4.4.1 Evaluation of Peak Total Stress
The peak total stress can be evaluated by :

σtp = σm + σb + σp + σr (C.4.11)
σtp = Kg · Kw · (Kte · Ktα σm + σb) + σr
In a fatigue design process based on a hot spot stress approach, the notch stress
concentration factor (Kw) is omitted. The notch stress concentration associated with the weld toe
is accounted for in the S-N curve or fatigue life data. The notch stress concentration due to an
existing defect would only be considered in a damage tolerance or residual life assessment as
opposed to the safe life fatigue design process presented in this Guide.
The nominal membrane, bending and peak stress components due to the applied loads
(excluding residual stresses) may be derived, for a given stress distribution σ(x) for x = 0 (at the
surface) to x = t s (through the thickness), by the following analytical expressions Ref. C.27 (see
Figure C.4.2.1:
ts
1
s m =
ts ∫ s(x) ⋅ dx (C.4.12)
0
ts
6  ts 
s b =
ts ∫ s (x) .  2 − x  . dx

(C.4.13)
0
s p (x) = s(x) − s m − s b (x) (C.4.14)
C.4.4.2 Stress Concentration Factors for Ship Details

Stress concentration factors (SCF) for a range of typical ship structure details are
available in a number of references, including [Refs. C.1, C.3, C.28, C.29] Stress concentration
factors for typical ship structural details (K g) and for misalignment effects (Kte, Ktα) are defined
in Section 4.2.2 and presented in Appendix C.
It is necessary to exercise extreme care when applying stress concentration factors from
different sources to ensure that the correct nominal stress definition is used. For example, in
some cases the nominal stress is defined at the intersection point of a connection, in other cases,
the global nominal stress may be defined at the weld toe or some distance form the weld toe.

Furthermore, the designer should be aware that sometimes the published stress
concentration factor solutions are designed to calculate the "hot spot" stress or the "notch" stress
as opposed to the local nominal stress. It is, essential, therefore, to make certain which form of
peak stress will result from the application of the SCF.
C.4.4.3 Local Finite Element Analysis

If appropriate stress concentration factors are not available, the total stress distribution
including local peak stresses may be calculated by local FEA. As discussed previously in the
section on global FEA, the extent of the local model should be large enough that the calculated
results are not significantly affected by assumptions made for boundary conditions and
application of loads.
Figure C.4.4.1 shows a local finite element model of a ship detail. The local model
should have a relatively fine mesh, especially in areas of stress concentration. It is important to
have a continuous and not too steep change in the density of the element mesh in the areas where
the local stresses are to be analyzed. The geometry of the elements (aspect ratio, corner angles,
skewness and warp) at the point of interest should be as near optimal as possible (for example:
element length/breadth aspect ratio less than 2, corner angles between 60o and 120o , avoid use of
triangular elements with reduced order shape functions).
Local FEA of a joint is usually conducted to determine the local nominal and hot spot
stress at the location of interest. If the peak notch stress has to be determined (i.e., for fatigue
analysis approaches other than that used in this Guide), then the most common approach is to use
local FEA to evaluate the hot spot stress. The hot spot stress value is then factored by a weld
notch factor, Kw , derived from parametric equations or tables to provide an estimate of the peak
notch stress in the joint.
Finite element size requirements in the stress concentration region are dependent on the
type of element. The mesh size may be determined based on experience or by benchmark testing
a similar mesh for a case where results have been presented in the literature. Figure C.4.4.2
provides some guidance on element sizes for 20-node solid, 8-node shell and 4-node shell
element types suitable for determining the stress concentrations consistent with approach
advocated by this Guide.
Normally the element stresses are derived at the Gaussian integration points. Depending
on the element type, it may be necessary to perform several extrapolations in order to determine
the stress at the weld toe. Referring to Figure C.4.4.3, all stress components are used for the
extrapolation. The process is as follows:
(1) Extrapolate the stresses to the surface from the Gauss points based on the assumed
distribution function in the element (some FE programs will provide this on request);
(2) Extrapolate surface stress to a line A-B centred on the hot spot of interest;
(3) Calculate stress along line A-B at reference points taken at t/2 and 3t/2 from hot spot;
(4) Linearly extrapolate through reference points t/2 and 3t/2 to determine stress at hot spot;
(5) Having extrapolated stress components for the hot spot, the principal stresses are
calculated at that location for fatigue analysis.

Figure C.4.4.1: Local Finite Element Model of Ship Detail [Ref.C.30]

Structural Detail
Model with 20-node solid elements (size t x t x t)
Element Type Element Size

20 – node isoparametric txtxt
solid element
Model with 8-node shell elements (size 2t x 2t)
8 – node quadrilateral 2t x 2t
isotropic shell element
4 – node quadrilateral txt
isotropic shell element
Model with 4-node shell elements (size t x t)
Figure C.4.4.2: Recommended Element Sizes for Local Detail FEA [Ref. C.3]

Figure C.4.4.3: Stress Distribution at an Attachment and Extrapolation of Stresses at Hot Spot
[Ref. C.3]

C.5 DEVELOPMENT OF STRESS RANGE DISTRIBUTIONS
C.5.1 Introduction
This section describes procedures for combining the environmental, vessel and structural
response information, to produce the stress range distributions required for the fatigue damage
summation. This involves applying the response functions previously developed.
The processes covered in this section are the fifth and sixth steps outlined in the flow
chart for the fatigue design procedures (Figure C.2.1). The two step process, described below,
may be completed for the Level 2 or Level 3 approaches. These two generalized steps are
expanded and described in the sections which follow.
Level 2 Level 3
5) Develop hot spot stress ranges from 5) Develop hot spot stresses from load and
load and stress transfer functions stress transfer functions
6) Define long-term distribution for notch 6a) Define short-term distribution for hot
stress range spot stress ranges
6b) Define long-term stress distribution by
summing short-term distributions
C.5.2 Level 2 Approach

This step involves the combination of data and transfer functions to develop a
characteristic stress range value from which a long term stress range distribution is inferred. The
information required to develop a long term distribution includes:
Design Load Data
• Components of ultimate strength loading (Mhog, Msag pint, pext ) ................ [Section C.3.2]
• Reference wave encounter period for the design load (n0 = 104 ) ................ [Section C.3.1]
• Load to load range approximation (e.g. fully reversing moments) ............. [Section C.3.1]
Stress Transfer Functions

• Nominal stress transfer functions (Fsm (load), Fsb(load), KG) ...................... [Section C.4.3]
• Hot spot stress transfer functions (Kg , Kte ,· Ktα) ...................................... [Section C.4.4]
Modifying Factors
• Encounter frequency correction factor (ff) ................................................... [Section C.3.2]

Determination of the hot spot stress range distribution (see C.4.2.4) is a three step process
(illustrated for vertical bending moment):
(1) Calculate reference nominal stress range

∆σnom = (∆Mv yna / Iv ) KG ff (C.5.1)
where: ∆Mv = 2Mv or (Mhog + Msag) and ff is the factor to transform the load from one
probability level to another probability level.
(2) Calculate reference hot spot stress range
∆σhotspot = K g · Kte · Ktα ∆σnom (C.5.2)
(3) Develop long term Weibull stress range distribution for the location of interest
Q(∆σhot spot ) =1 - exp(-[∆σhotspot / q] h ) (C.5.3)
where: Q( ) is the long term Weibull cumulative distribution function for a stress range
∆σhot spot .
q = ∆σhot spot / [ ln(n0 ) ] 1/h
h is the Weibull shape parameter (see Section C.3.2.1.2)
Based on this approach, a long term probability distribution function of stress ranges in
either continuous (Figure C5.2.1a) or discrete form (Figure C5.2.1b) is available for the fatigue
damage summation.
1 0.35
Probability of Exceedance
  ∆σ h 0.3
hot spot  
0.8
( ) 
Q ∆σ hot spot = exp  −    0.25
Probability
  q  
0.6  0.2
0.4 where: q = 25 and h = 1 0.15
0.1
0.2 0.05
0 0
0 20 40 60 80 100 120 5 25 45 65 85 105 125
Stress Range, ∆σ Stress Range, ∆σ
Figure C5.2.1a: Continuous Weibull Stress Figure C5.2.1b: Discrete Weibull Stress Range
Range Distribution Distribution
The Weibull parameters (q and h) used to develop the continuous distribution are ready to
be used in the closed form fatigue damage summation (see Section C.6.3.2), whereas, the
discrete distribution may be used in the discrete fatigue damage summation described in Section
C.6.3.1.
C.5.3 Development of a Stress Range Distribution for the Level 3 Approach
This step involves the combination of data and transfer functions to develop a
characteristic stress range value from which a series of short term operational condition stress
ranges are developed and then summed to produce a long term stress range distribution. The
information required includes:

Design Load Data

• Composite wave scatter diagram(s)............................................................... [Section C.3.1]
• Reference wave encounter period for the design load (n0 = 104 ) ................ [Section C.3.1]
• Load to load range approximation, Fload ........................................................ [Section C.3.1]
• Characteristic load response .......................................................................... [Section C.3.2]
• Conditional probabilities for wave direction and velocity (fv ) ...................... [Section C.3.2]
Stress Transfer Functions

• Nominal stress transfer functions (Global FEA results)................................ [Section C.4.3]
• Hot spot stress transfer functions from unit load, local FEA)....................... [Section C.4.4]
Determination of the hot spot stress range distribution is a three step process (as
illustrated for vertical bending moment):
1) Calculate reference hot spot stress range
∆σhot spot = 2 K0 Fload(Hs i, Vi, θi, L.C i) . . . for each (Hs, V, θ, L.C.) (C.5.4)
where: K0 is the stress transfer function relating vessel load response to detail hot spot
stress
Fload is the characteristic (RMS) load response of the vessel such as Mv , MH, p
2) Develop Rayleigh short-term stress range distribution
F∆σi(∆σhotspot ) = 1 - exp(-∆σhot spot2 / 8 m0i) . . . for each (Hs, V, θ, L.C.) (C.5.5)
where: m0i is the spectral zeroth moment for each of the i operational conditions inferred
from reference hot spot range (m0 = ∆σhotspot RMS2 )
F∆σi(∆σhot spot ) is the Rayleigh “short term” hot spot stress cumulative density
function for the ith operational condition.
3) Develop long-term stress range probability distribution function (Weibull or Histogram)
Q′(∆σhotspot )long = Σ ri Q′(∆σhot spot )short ⋅ pi (C.5.6)
Q′(∆σhotspot )short = F∆σ(∆σhot spot )upper - F∆σ(∆σhot spot )lower
where: pi is the fraction of time at the ith operational condition
ri ratio of the ith zero crossing rate (fz i) to the average zero crossing rate for all
operational conditions rI = fz / f z
where: f z = S pi f zi
Q′(∆σhotspot )long is the “long-term” hot spot stress probability distribution function
Q′(∆σhotspot )short is the “short-term” hot spot stress probability distribution function
In this third step, a weighted sum of Rayleigh “short-term” distributions is used to

develop a “long-term” distribution of stress ranges. A direct approach to this involves breaking
each Rayleigh distribution, associated with an operational condition, into discrete stress range
intervals and estimating their probability of occurrence, as shown in Figure C.5.3.1 and Table
C.5.3.1.

0.3
Operational Condition 1 (m01 = 100)
0.25
Probability
0.2
0.15
0.1
0.05
0
5 25 45 65 85 105 125 145 165 185 205
Stress Range, ∆σ
Figure C.5.3.1: Operational Condition Rayleigh Distribution Stress Range Interval Probabilities
Table C.5.3.1: Operational Condition Rayleigh Distribution Stress Range Interval

Probabilities
Operational pi fz i Stress Range Interval*, ∆σ
Condition [s] 5 15 25 35 45 55 65 75 . . . .
1 (m01 = 100) 0.50 0.0556 0.118 0.276 0.282 0.189 0.091 0.033 0.009 0.002 . . . .
2 (m02 = 350) 0.35 0.0909 0.035 0.098 0.142 0.160 0.155 0.133 0.103 0.072 . . . .
3 (m03 = 800) 0.15 0.2000 0.016 0.045 0.071 0.090 0.102 0.107 0.105 0.097 . . . .
* each 10 MPa stress range interval is identified by its mid point
The three data sets, presented as examples in Figure C.5.3.1 and Table C.5.3.1, represent
the short-term Rayleigh distribution for three operational conditions that define the potential
service conditions of a vessel. The weighted short-term distribution sum, used to develop a long-
term distribution is completed, for n operational conditions as follows:
n
fz = ∑ pi x fzi (C.5.7)
i =1
f zi = 0.0556(0.5) + 0.0909(0.35) + 0.2(0.15) = 0.0896
Table C.5.3.2: Weighted Short-Term and Resulting Long-Term Stress Range Interval
Probabilities
Operational pi ⋅ f zi Stress Range Interval*, ∆σ
Condition f zi 5 15 25 35 45 55 65 75 . . . .
1 (m01 = 100) 0.310 0.0366 0.0856 0.0874 0.0586 0.0282 0.0102 .... .... ....
2 (m02 = 350) 0.355 0.0124 0.0348 0.0504 0.0568 0.0550 0.0472 .... .... ....
3 (m03 = 800) 0.33 0.0056 0.0151 0.0238 0.0301 0.0341 0.0358 .... .... ....
Long-Term -- 0.0544 0.1355 0.1616 0.1155 0.1173 0.0932 .... .... ....
* each 10 MPa stress range interval is identified by its mid point

With the probability of occurrence of Q′of each stress range interval estimated for the
long term probability distribution (see Table C.5.3.2) in a discrete form, a cumulative
distribution, Q, is obtained by summing stress range probabilities from the smallest stress range
incrementally to the higher stress ranges such that the cumulative probability of each bin is equal
to the sum of cumulative probabilities of all stress range bins up to the latest bin and the
probability of occurrence of the latest bin. Ultimately, the value of the cumulative probability of
the longest stress range bin should equal 1. Continuous representation is developed by fitting a
Weibull distribution to this data. A least squares fit was used to establish the statistical
parameters (q = 36.5 and h = 1.62 ) for the continuous Weibull distribution (Eq. C.5.3). This is
shown in Figure C.5.3.2.
1.2
Cumulative Probability
1
0.8
0.6 Discrete Long-Term Distribution
0.4 Continuous Long-Term Distribution
0.2
0
5 25 45 65 85 105 125 145 165 185
Stress Range Interval, ∆σ
Figure C.5.3.2: Long-Term Stress Range Cumulative Probability Function

C.6 FATIGUE DAMAGE CALCULATIONS

C.6.1 Introduction
Once the expected stress ranges (including concentration factors) have been determined
for any structural detail, the fatigue damage calculations can be undertaken. The standard
approaches to this use the Miner-Palmgren linear damage summation model [Ref. C.31] to
evaluate the cumulative damage (D) over the period selected, which is normally taken as the
design lifetime of the ship.
This model assumes that the cumulative fatigue damage (D) is the sum of the fatigue
damage inflicted by each stress cycle (di ), independent of the sequence in which the stress cycles
occur. The model further assumes that di can be mathematically expressed as 1/N i where Ni is
the average number of loading cycles to failure under constant amplitude loading at the ith stress
range.
Nt Nt
1
D = ∑ di = ∑ (C.6.1)
i =1 i =1 N i
A damage sum exceeding unity is interpreted as a failure, although the meaning of this
term is generally left somewhat vague. It can generally be interpreted as the generation of a
through-thickness crack, several centimetres in length, in the detail under consideration. In
practice, such a crack may or may not be of sufficient size to cause degradation of function (e.g.,
water tightness) or of unacceptable risk of fracture; i.e., a crack that needs to be repaired.
If the damage sum over N stress cycles is less than unity, the ratio of the projected fatigue
life to the time for N cycles is D/Ntotal . Fatigue damage assessments can thus be used to rank
structural details in the ship according to projected life, and develop inspection strategies that
focus attention on the highest risk areas.
In order to derive the damage at each stress cycle, or at each stress range level, fatigue
design (S-N) curves are used. These curves are derived from experimental data from fatigue
tests conducted in air or in a simulated sea water environment, usually on relatively small
specimens simulating simple welded details of various configurations. Derivation of the design
S-N curves from the experimental data takes into account:
• differences in stress concentration factors between the structural detail and tested weld
detail;
• adjustments to reflect desired safety factors (probabilities of failure);
• differences in the material thickness between the structural detail and the tested detail.
Other modifications may also be needed to account for fabrication processes, including
the effects of mean stress levels, residual stresses and post-weld treatments such as grinding and
peening. Obviously, if the design calculations assume certain “non-standard” factors, then the
ship construction needs to ensure that these are actually incorporated in the fabrication processes.

Section 6.2 presents recommended fatigue design curves for use with the stress
calculation methodologies described earlier, and a set of modification and correction factors
which can be used in tailoring these to a specific application. Section 6.3 provides damage
assessment calculation tools, which are standard mathematical applications of the recommended
equations and distributions.
All aspects of the procedures are consistent with “state-of-the-practice” ship design.
However, it should be noted that individual approval agencies (such as classification societies)
may take different approaches to some or all issues. In addition, ongoing research in this area
has recently led to changes in some of the equivalent codes and standards for offshore structures.
Therefore, where a design is particularly sensitive to fatigue performance, the structural designer
should consider undertaking additional background studies of the most recent research. The
Triennial Proceedings of the International Ship and Offshore Structures Congress [Ref. C.32] can
be a valuable starting point for such studies.
C.6.2 Fatigue Design Curves

C.6.2.1 Baseline (in Air)
Fatigue design or fatigue assessment procedures based on S-N curves can employ one of
three different approaches:
• nominal stress approach;
• hotspot stress approach; and,
• notch stress approach.
The latter two are referred to as “local stress” approaches. All of these fatigue design
approaches include the effects of: structural geometry, local detail geometry, weld toe (notch)
geometry. The difference among the three approaches is how they capture these effects. It is
possible to include these effects in the calculated stress range in terms of stress concentration
factors (which increase the applied stress range) or to capture these effects in the S-N curve
(which reduces the fatigue life). In principle, all three approaches should lead to similar
outcomes if all of the effects are properly captured in the stress concentration factors and S-N
curves. In practice, this may not happen due to various assumptions made in either deriving the
design S-N curve or in calculating the applied stress range. The local stress (three) approaches
evolved over time, to overcome the limitations of the nominal stress approach when applied to
welded joints in complex structural details.
C.6.2.1.1 Nominal Stress Approach

Fatigue design procedures for welded plate-to-plate joints in steel bridges and steel
offshore structures [Refs. C.33-C.38] classify basic welded joints subjected to uni-directional
loading (axial or bending) into different categories of fatigue performance, and specify an S-N
design curve for each category. The effects of welding residual stresses and stress
concentrations in the classified joints are assumed to be built into the relevant design curves, so
the fatigue design stress for a given joint is simply the nominal stress range at the anticipated
crack initiation site. S-N design curves from the aforementioned procedures have been
incorporated into fatigue design procedures recently introduced by certain classification societies
and navies for welded joints in ship structures [Refs. C.1, C.2 and C.14].

However, the basic joints covered by the nominal stress approach are often embedded in
complex structural details and sometimes located near openings and other structural
discontinuities. The stresses within these details can be highly multi-axial even if the details are
located in generally uni-directional nominal stress fields. Furthermore, the principal components
of the actual stresses at a given point in a structural detail may be higher or lower than the
nominal stress components. Therefore, when the aforementioned S-N design curves are applied
to welded joints in ship structural details, the fatigue design stresses for the aforementioned
situations should be based on the principal components of local nominal stresses. These stresses
are the actual stresses at the anticipated crack initiation site, minus any stress concentrations in
the basic joint (e.g., weld toes, weld reinforcement, attachments, eccentricity of lap joints), but
including the global stress concentration of the structural detail and the global stress
concentration effect of any nearby structural discontinuities (e.g., openings). Unfortunately, it is
often difficult to classify a welded joint in a complex structural detail and to quantify the stress
concentration effects that are not already built into the corresponding S-N curve, and, therefore,
the nominal stress approach is not recommended in this Guide.
C.6.2.1.2 Notch Stress Approach

Det Norske Veritas [Ref. C.3] and Bureau Veritas (BV) have advocated the use of the
notch stress approach. The notch stress range is defined as:
∆σnotch = Kw ⋅ ∆σhotspot = Kg ⋅ Kw ⋅ ∆σnom (C.6.2)
where Kg ⋅ Kw are the general detail and weld stress concentration factors, respectively (see
Appendix C). Therefore, this includes the stress concentration effects of the local weld
configuration and local notch geometry, as well as the overall stress concentration of the
structural detail. The value of Kw can be computed from parametric equations or by fine mesh
finite element analysis, but in either case, require a prior knowledge of the weld profile and toe
geometry that may not always be known. DNV propose a default value of 1.5 for Kw, and it
seems that design notch stress based fatigue design curve has been obtained from a hotspot stress
fatigue design curve by adjusting the stress axis by a factor of 1.5. BV, on the other hand,
provides a parametric equation for K w that is a function of the weld toe angle and that can
assume values from about 1.0 to 1.96 for toe angles varying from 12 to 45 degrees. The design
S-N curve to be used in conjunction with the notch stress approach is that for the base material
which seems to be a non-conservative approach as it does not take into account residual stresses
at all. The authors of this Guide do not see any inherent benefit in following the notch
stress approach and therefore, it is not recommended in this Guide.
C.6.2.1.3 Hotspot Stress Approach

The hot spot stress approach is widely used in fatigue assessment procedures for tubular
joints in offshore structures [Refs. C.33, C.38 and C.39] and is being increasingly used as an
alternative to the nominal stress approach for welded plate-to-plate joints in complex structural
details [Ref. C.40]. It enables the fatigue lives of many types of structural details with nominally
similar weld profiles, weld toe geometries and local joint configurations to be assessed using a
single fatigue design curve.

The applicable stress range to be used in conjunction with the design hotspot stress S-N
curve is the maximum principal stress range at the crack initiation site minus the stress
concentrations of the weld reinforcement, weld toe geometry. The latter effects are built into the
fatigue design curve, whereas the overall stress concentration of the structural detail must be
explicitly included in the hotspot stress range. The hot spot stress range is defined as:
∆σhot spot = Kg ⋅ ∆σnom (C.6.3)

where Kg is the local detail (or structural) stress concentration factor.
For plate to plate joints, the hotspot stress fatigue design curve corresponds to the S-N
curve for a transversely loaded butt weld (with reinforcement in place). Unfortunately, such S-N
curves are not identical in various codes, standards or rules. For example, the American Bureau
of Shipping’s fatigue design procedures for tankers and bulk carriers [Refs. C.1, and C.41]
suggest the use of the “Category E” nominal stress design curve (transversely loaded, full
penetration butt weld made in positions other than down-hand or made using the submerged arc
welding process or between plates of unequal width or unequal thickness) as the hotspot stress
fatigue design curve to assess the fatigue strength of fillet-welded bracketed connections. The
constant amplitude fatigue strength of “Category E” joints at 107 cycles is 47 MPa. Refs. C.42
and C.43, on the other hand, advocate the use of the S-N design curve for a better quality
transverse butt weld to analyze fillet-welded joints, while the U.K. Department of Energy’s
widely referenced T-curve for hotspot stress analysis of tubular joints with full penetration fillet-
like welds in offshore structures [Ref. C.44] nearly coincide with the aforementioned “Category
D” curve. The constant amplitude fatigue strength of category D joints at 107 cycles is 53 MPa.
It is conceivable that the selection of the hot spot stress fatigue design curve is a
reflection of the weld quality anticipated by the respective organizations. Since offshore
structures and bridges are subject to extensive non-destructive inspection, it might ensure better
quality welds warranting a category D hotspot fatigue design curve. In ships, on the other hand,
non-destructive inspection is usually performed on a selective basis only and therefore selecting
a category E curve for hotspot stress fatigue design may be prudent.
Notwithstanding these differences, the hot spot stress approach is advocated in this
Guide because of its ease of application compared to the nominal or the notch stress approaches.
The hotspot stress design fatigue curves, shown in Figure 6.2.1 for “air” and “sea water”
environments, are adopted from the UK Health and Safety Executive (HSE) [Ref. C.45] which is
believed to reflect the latest (1998) thinking on this subject.

1000
Air
Salt Water with Coatings & CP
Salt Water Damaged Coatings & CP
Stress Range, ∆σ [MPa]
100
10
1
1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09
Number of Cycles, N
Figure 6.2.1: HSE Hot Spot Fatigue Design Curve for Plated as Welded Joints [Ref. C.45]
When compared to actual or implicit hotspot stress fatigue design curves adopted by
various organizations (see Appendix A), the HSE curve is seen to be “in the middle of the pack”.
For the air environment, the HSE hotspot stress design curve provides a bi-linear relationship
between log N and log ∆σhot pot
for N ≤ 107 cycle

log N = log a1 - m1 ⋅ log ∆σhotspot = 12.182 - 3.0 log ∆σhotspot (C.6.3a)
for N > 107 cycle
log N = log a2 - m2 ⋅ log ∆σhotspot = 15.637 - 5.0 log ∆σhotspot (C.6.3b)
where N is the fatigue life for a constant hotspot stress range ∆σhotspot . The value of ∆σhotspot at
the transition between the two sections of the bi-linear curve is 53 MPa.
The second part of the bi-linear curve reflects the reduced rate of damage accumulation at
stress range levels below the 'fatigue limit' found in constant amplitude testing. A limit, as such,
is not normally encountered in service experience with fluctuating stress ranges. Fracture
mechanics theory, confirmed by experience, leads to the change in slope reflected by m1 to m2 .

This curve is generally valid for hotspot stress ranges greater than the nominal yield
strength of steel plate (σyield), up to a maximum of 2σyield [Ref. C.31]. However, it should not be
used if the largest principal component of the nominal stresses at the crack initiation site (i.e., the
maximum stress excluding residual stresses and stress concentration effects of the weld toe, weld
reinforcement, basic joint configuration, and overall stress concentration of the structural detail)
exceeds 80% of the yield strength of the surrounding parent material. Neither of these
limitations will normally be important to conventional ship designs.
S-N curves exist not only for fabricated details, but also for other types of structure
including rolled sections, cut-outs of various types, etc. Normally in ship construction, fatigue
will not be a problem in components that have not been processed to some degree, or at locations
remote from such processing. However, if high fluctuating stresses are expected (for example) at
the centre of a long, unsupported, rolled beam, the designer should refer to source documents
such as Ref. C.33 for guidance on an appropriate curve. A type of detail where fatigue problems
may be more likely in standard design is the thermal (flame) cut-out in plated structure. A wide
range of fatigue tolerance can be found in such cut-outs, depending on the heat inputs and on the
process controls. Equation C.6.3 will provide a conservative prediction of the fatigue life for all
but poor quality cut-outs. If designers need to account for better quality (or post-treatment), then
some guidance on relative performance can be found in Ref. C.46.
C.6.2.2 Corrosion and Corrosion Fatigue

The basic fatigue design curve in the section previous (Equation C.6.3) is for
performance in a corrosion-free environment. In practice, general corrosion and localized
corrosion (i.e., pitting) of unprotected steel are major sources of damage in ships, and the
resulting reductions of net-section and stiffness can significantly increase cyclic stresses at
welded details. Various corrosion fatigue mechanisms can also accelerate the initiation and
propagation of fatigue cracks in welded structural details without proper protective coatings even
if the details are cathodically protected.
To address this situation, it is normally accepted that all nominal stress levels should be
calculated for net (i.e., end-of-life) scantlings. The generalized corrosion allowances applied to
different types and areas of structures vary somewhat between classification societies but within
fairly narrow ranges. ABS Rules [Ref. C.41] allow some account to be taken of the periods
when the actual scantlings will remain above the minimum net (end of life) values by permitting
reduction of general stress levels by a factor of 0.95. Although this type of further adjustment
appears reasonable, its effects are minor in comparison with the uncertainties in other aspects of
the procedure. Therefore, this Guide recommends the basic, net scantling approach, as
being consistent with most other approaches.
As regards fatigue life, majority of codes and standards accept that unprotected joints
exposed to sea water have their fatigue lives reduced by approximately a factor of two compared
with those in air. Further, the fatigue design curves are no longer bi-linear, but instead retain
their steeper gradient at lower stress ranges.

Some more recent data has suggested that the factor of two is non-conservative, and also
that cathodic protection may not always provide effective fatigue corrosion protection. This is
discussed in references [Ref. C.45 and C.47], and larger reductions in life (factors 2.5-3) are
recommended under the UK Health and Safety Executive's latest guidance for offshore structures
[Ref. C.48]. Following HSE again, the fatigue damage calculations for welded plate-to-plate
details in the marine environment (submerged, splashed, or sprayed by sea water) are
recommended to be based on the following relationships:
• with undamaged protective coatings and cathodic protection:
for N ≤ 1.026 ×106 cycles

log N = log a1 - m1 ⋅ log ∆σhotspot = 11.784 - 3.0 log ∆σhotspot (C.6.4a)
for N > 1.026 ×106 cycles

log N = log a2 - m2 ⋅ log ∆σhotspot = 15.637 - 5.0 log ∆σhotspot (C.6.4b)
The value of ∆σhotspot at the transition between the two sections of the bi-linear curve is 84
MPa.
• without cathodic protection and proper protective coatings in a sea water environment (i.e.,
freely corroding joints):
log N = log a - m ⋅ log ∆σhotspot = 11.705 - 3.0log ∆σhotspot (C.6.4c)

The hotspot stress fatigue design curves represented by Equations C.6.4 for the sea water
environment are included in Figure 6.2.1 shown earlier.
When structure can be assumed to be effectively protected over part of its life and
unprotected over the remainder (e.g., for ballast tank coating systems with expected lifetimes of
less than 20 years) then the fatigue life assessment can combine calculations using Equations
C.6.3 and C.6.4. A simplified method for doing this is offered by DNV [Ref. C.3]
C.6.2.3 Probability of Failure

Equations C.6.3 are based on two standard deviations below the mean probability of test
specimen failure at a calculated value of D = 1; i.e., a failure probability of approximately 2.3%
at Nt lifetime cycles. This is regarded as acceptable for most ship structures for the level of
“failure” implied. However, where a detail is critical to the safety of the ship and/or personnel, it
will be advisable to apply an additional safety factor, either by choosing D < 1 or by adjusting
the S-N curve used in the analysis. The selection of an alternative acceptable failure probability
can make use of reliability analysis, as described in Section C.7.
The lifetime failure probability should also reflect an accurate assessment of the number
of stress cycles at each stress range over the life of the ship, as discussed in Section C.5.

C.6.2.4 Thickness Effects

The inherent stress concentration factors for any type of detail, reflected in the S-N curve,
may change with thickness due to differences in stress gradient and to residual stresses.
Therefore, it may be necessary to make a thickness correction above a value characteristic of the
samples tested to develop the curve. The “classic” form of this correction is due to Gurney [Ref.
C.49] in which, for fillet-welded joints with plate thicknesses exceeding a reference thickness tr
1
 ? s S− N   t r  4
  =   (C.6.5)
 ? s c   ts 
where:
ts is the thickness of the plate through which fatigue cracks propagate;
∆σ is the design stress range allowed by the relevant S-N curve; and,
∆σc is the corrected design stress range.
This correction was subsequently incorporated into a number of fatigue design standards
for steel bridges, offshore structures, and ships. Many of the classification society rules have tr in
the range of 22-25mm. However, HSE based on the latest available data, suggests that reference
thickness should be 16mm, and that the exponent should be 0.3. Accordingly, when the thickness
exceeds 16 mm, Equations C.6.3 and C.6.4 need to be modified to include the thickness
correction that accounts for experimentally observed and theoretically predicted reductions in
fatigue life of certain joints with increasing plate thickness (t):
log N = log a1 - m1 ⋅ log ∆σhotspot - 0.3. m1 ⋅ log(t/16) (C.6.6)
While the above correction is more conservative than that recommended in other codes
and standards, it is compensated by more recent experimental and theoretical studies indicating
that the correction is not required for longitudinal welds, transverse butt welds, and transverse
fillet welds with weld stress concentrations comparable to that of butt welds (e.g., transverse
fillet welds with smooth weld/base transitions achieved by grinding or special welding
techniques).
C.6.2.5 Fabrication Considerations

While the in-air S-N curves have been used for steels with yield strength up to 690 MPa,
the experimental data base in the sea water environments is based on steels with tensile yield
strengths in the range of 300-500 MPa. There is currently little suitable public-domain data for
higher-strength steels in a marine environment where there is potential for further deterioration
in the fatigue performance due to increase in hydrogen embrittlement susceptibility with increase
in strength. Up to 500 MPa the curves can be considered independent of material tensile strength
[Refs. C.50 and C.51].

The weld toes of fatigue-prone structural details in ships are sometimes ground to
improve the resistance to fatigue crack initiation. A factor of 2.2 (improvement) on the fatigue
lives of welded details in air and cathodically protected welded details with proper protective
coatings in a sea water environment can be achieved if grinding is used to produce a smooth
concave weld profile, including the weld toe, which blends smoothly with the parent material.
This correction can be implemented by adding log 2.2 to the right-hand side of Equations C.6.3
to C.6.4. However, the effect will only be achieved with a high-quality weld and careful
application of the grinding procedure. Adequate weld throat size must be retained and protective
coatings must be properly applied before any corrosion can initiate.
It is known that ship structures experience residual stress relief during a “shake-down”
period after they enter service. Weld toes, initially under considerable tension, may see this
relaxed. This can mean that the ongoing stress cycles which they experience are fully or partly
compressive, and thus significantly less damaging than purely tensile cycles. However, it is
difficult to provide generic credit for this effect due to the lack of relevant data and the
importance of local design.
If it can be demonstrated theoretically or experimentally that the sum of the applied stress
range, fabrication restraint stresses, and welding residual stresses will be partly or completely
compressive over a particular duration of the design life, then the effective stress range over that
duration can be defined as the sum of the tensile part of the sum plus 60% of the compressive
component. Otherwise, the effective stress range should be assumed to be equal to the applied
stress range.
C.6.3 Damage Summation Model
Integration of the damage caused by the anticipated loads using the S-N curves
recommended above can be done in a number of ways, depending on the nature of the stress data
available and the level of accuracy required from the procedure.
C.6.3.1 Numerical Calculation

If the long-term Weibull distribution of hotspot stress ranges for the simplified method,
or the sum of the short-term Rayleigh distributions of hotspot stress range for the full spectral
analysis method is approximated by a histogram with k blocks of stress cycles of constant stress
range, the damage inflicted by the distribution can be estimated as follows:
k k
ni
D ≈ ∑di = ∑ (C.6.7)
i =1 i =1 Ni
where k = number of stress blocks
ni = number of stress cycles in stress block i
Ni = number of stress cycles to failure at stress range ∆σi
The number of blocks should be large enough to ensure reasonable numerical accuracy and
should not be less than 20. Within each block, the stress range ∆σi selected should represent the
mean stress range for that block [Ref. C.3].

C.6.3.2 Closed Form Solution for Level 2 Analysis

If the long-term distribution of hotspot stress ranges is defined by a Weibull distribution
and if the fatigue design curve is a linear log N vs log ∆σhotspot relationship like Equation C.6.4,
the cumulative damage (D) is defined by the following closed form solution:
fz td N load
 m
D= ∑ pn q mnG 1 + h  (C.6.8)
a n =1  n 
where
Nload = total number load conditions
pn = fraction of design life in nth load condition
td = design life of ship in seconds (20 years = 6.3 × 108 seconds)
hn = Weibull shape factor for nth loading condition
∆σ o
qn = Weibull stress range scale parameter for nth loading condition = (C6.9)
( ln n o )1/h
f z = average long term zero-crossing frequency (see Section C.5.3)
a and m = fatigue design curve parameters
Γ(1+m/hn ) = gamma function
∆σo = design stress range for the nth loading condition
1/no = the reference probability of exceedence on which ∆σo is based (typically 10-4)
If the long-term distribution of hotspot stress ranges is defined by a Weibull distribution

and if the fatigue design curve is a bi-linear log N vs log ∆σhotspot relationship like Equation
C.6.3, the cumulative damage (D) is defined by the following closed form solution
 (q ) m1  m ?s hn 
(q n ) m 2  m 2  ? s slope  n 
h
slope  
N load

D = fz t d ∑ p n  na G 1 + 1 
; 
h n  q n  
+
a2
? 1 + ; 
h n  q n  
(C.6.10)
n =1  1    
where:
a1 and m1 = the fatigue design curve parameters for N ≤ 107 cycles
a2 and m2 = the fatigue design curve parameters for N > 107 cycles
γ (;) = incomplete gamma function
Γ( ; )= complementary incomplete gamma function
∆σslope = stress range at which change in slope occurs (See Sections C.6.1.1 and C.6.1.2)
Tables of full and incomplete gamma functions can be found in numerous standard texts,
such as [C.6.59].

C.6.3.3 Cumulative Damage for Level 3 Analysis

If the long term stress range distribution is defined by a series of short-term Rayleigh
distributions for different sea states and headings, and if the fatigue design curve is a linear log N
and log ∆σhotspot relationship like Equation C.6.4, the cumulative damage is defined by the
following closed-form solution:
all sea states
( )m
Nload all headings
fz td  m 
D=
a 
G 1 +  ⋅
2
∑ pn ⋅ ∑ rijn 2 2m 0ijn (C.6.11)
n =1 i −1, j−1
where
rij = the relative number of stress cycles in short-term condition i,j (see Section C.5.3)
moijn = zero spectral moment of stress response process
Γ(1+m/2) = gamma function = 1.33 for m = 3
If the long term stress range distribution is defined by a series of short term Rayleigh
distributions for different sea states and headings and if the fatigue design curve is a bi-linear log
N and log ∆σhotspot relationship like Equation C.6.3, the cumulative damage is defined by the
following closed-form solution:
all sea states

N load all headings
D = f zt d ∑ pn ⋅ ∑ rijn (B + C) (C.6.12)
n =1 i −1, j−1
(2 )  
m1 2
2 m 0ijn
 m  ∆σ slope 

B= Γ 1 + 1 ;  

a1 2  2 2 m 0ijn 

 
 2

(q n ) m  ∆σ slope 
m2
 
C= γ 1 + 2 ;  

a2 2  2 2m 0ijn 

 
where: m1 and m2 are fatigue design curve first and second slopes respectively.

C.7 FATIGUE LIMIT STATE RELIABILITY ANALYSIS

A reliability analysis allows the effect of the variability in load and strength variables to
be accounted for in the design process. In other words, a reliability analysis provides a vehicle
for estimating the probability that a given mode of failure (e.g., fatigue, fracture, plastic collapse,
etc.) will manifest in a given period of time. This analysis is based on the assumed variability of
the parameters used to describe the structural system and its loading.
The following section is not intended as a comprehensive review of structural reliability

analysis for ship structures. It is included here to make the reader aware that these techniques
have been introduced and to provide a brief review. In general, structural reliability analysis
may be completed by numerical simulation or by reliability index based approaches. In this
section, only the reliability index based approaches are discussed.
C.7.1 Reliability Analysis
Class societies, the Ship Structure Committee and other organisations have developed
reliability based design procedures [Refs. C.2, C.3, C.52 and C.54]. In general, a fatigue design
procedure estimates the probability that a fatigue failure (pf) will occur in a given structural
element during a given time period. Failure is typically defined by the Miner’s [Ref. C.31]
summation usage factor (η) exceeding a value of unity in the time period of interest or the
estimated fatigue life being shorter than the target design life. The approach presented below
was formulated to allow an analytic solution by assuming that the loading is distributed
according to a Weibull statistical distribution. For simplicity the fatigue assessment does not
include a thickness correction. The physical basis for the fatigue failure limit state is expressed
as follows:
(Kq)m  m
?=N − G 1 +  (C.7.1)
a S- N  h
where: η = fatigue usage factor predicting failure when η = 1
N = number of stress cycles within time td, t/ T
m = slope parameter of the S-N curve
aS-N = scale parameter of the basic S-N curve for members with K = 1
K = stress concentration factor
h = slope parameter of the Wiebull distribution for stress range
q = scale parameter for the Wiebull distribution for stress range
td = design life
T = long-term average stress period
In the above fatigue limit state equation not all of the parameters are considered random
or uncertain. Those parameters which are considered random would be represented by a
characteristic (e.g., mean or percentile) value and a measure of uncertainty (e.g., standard
deviation or coefficient of variation = standard deviation/mean). It has been suggested that the
slope of the S-N curve (m) should be considered a random variable, as well as, the S-N curve
scale parameter (a).

The stress concentration factor (K) is generally considered uncertain and it’s statistical
parameters may be estimated as a function of the statistical parameters of its component parts as
shown in Equation C.7.2 below, by replacing G(x) and x with K(k) and ki , respectively. The
wave characteristic period (T) which along with the design life (t) determines the number of
wave encounters or stress cycles (N), is considered uncertain and should be considered a random
variable. As far as the magnitude of the environmental loading is concerned, it is suggested that
the slope parameter of the Weibull stress, load or wave height distribution (h) be considered
deterministic whereas, the scale parameter (q) should be considered random. It is commonly
assumed that the uncertain (random) variables are statistically independent.
Typically a first order reliability method (FORM) based on a Taylor series expansion of
the deterministic expression for the expected Miner’s summation usage factor, is used to estimate
the reliability or probability of fatigue failure. In this case the limit state equation G(x) is 1-η(x)
and is used to estimate the probability of the usage factor exceeding unity in a given period of
time (e.g., Pf = Prob( G(x) = 1 - η(x) < 0)). The mean and standard deviation of the limit state
equation are estimated as:
n n n
µ G = G(µ x ) and σG2 = ∑ Ci 2σ xi 2 + ∑ ∑ ρijCiC jσ xi σxj (C.7.2)
i=1 i=1 j=1
∂ G( x ) (the first partial derivative of the limit state function

Where: Ci =
∂ xi G(x) with respect to one of the function variables, x)
and ρij is the correlation coefficient for variables i and j, which is typically assumed to be equal
to zero based on an assumed independence of the variables, xi and xj.
The reliability against failure is estimated in terms of a reliability index (β):
µ
β= G (C.7.3)
σG
The reliability index (β) is directly related to the reliability and probability of failure as
follows:
Pf = 1 - Reliability = 1 - Φ −1 (β) (C.7.4)
−1
where Φ is the inverse standard normal distribution function
A refinement of the first order technique described above, the design point approach,
identifies a combination of the limit state function random variables which minimises β and thus
maximises the probability of failure [Refs. C.56 to C.58].

C.7.2 Inherent Assumptions or Approximations in Structural Reliability Analysis
While reliability analysis is a useful tool to demonstrate the relative (i.e., comparison of
design alternatives or scantlings) likelihood of failure or risk associated with a structural detail,
the quality of its absolute numerical results are a function of the quality of the model and data it
employs, as in all engineering calculations.
In performing a reliability analysis, the user should be aware of the sensitivity of the
analysis results to the potential sources of error or assumptions made in the analysis process. A
general structural reliability analysis assumes that:
• All sources of variability or uncertainty are represented in the analysis. If this is not the
case, then the analyst has made a conscious decision to omit sources of variability that are
considered of lesser significance or that are unquantifiable. Sources of variability that are
difficult to quantify in a structural reliability analysis may include human behaviour or
error, fabrication tolerances or defects and gross differences in structural configuration.
• The variability of the load and strength variables are accurately defined. A statistically
significant sample or database of information is available to characterise the variable and
develop a representative statistical distribution.
• The dependence of variables and/or modes of failure are understood, quantifiable and
incorporated in the reliability analysis. The analyst may choose, as a numerical
simplification, to neglect dependencies or assume complete dependence or independence,
but should be aware of the effects of these assumptions.
• Since the limit state equation is used to define failure, it is essential that it accurately
predicts the expected mode of failure. Some detailed reliability analysis work includes a
random variable representing the uncertainty in the limit state equation based on a
comparison of experimental and predicted behaviour.
• In a reliability analysis, the failure mechanism or mode of failure must be specified,

therefore, it is important to ensure that the proposed mode(s) of failure correspond to the
actual mode(s) of failure. For instance, a reliability analysis of a tension member based on
a plastic collapse limit state would not be valid if reality the structural member fails in a
brittle (fracture) mode.

C.8 REFERENCES
[C.1] American Bureau of Shipping, “Guide for Fatigue Strength Assessment of Tankers”, Rules
for Building and Classing Steel Vessels, Part 5, Section 2, Appendix 5/2AA, 1996.
[C.2] Lloyd’s Register, “Ship Right - Fatigue Design Assessment Procedure”, 1996.
[C.3] Det Norske Veritas,“Fatigue Assessment of Ship Structures”, Classification Notes No.
30.7, September 1998.
[C.4].Chen, Y.N. and Shin, Y.S., “Consideration of Loads for Fatigue Assessment of Ship
Structure”: NRC Symposium, Washington, DC, 1995.
[C.5] Mitchell, M.R., “ Fundamentals of Modern Fatigue Analysis for Design”:ASME
Handbook, pp238.
[C.6] Stiansen, S.G., “Dynamic Response of Large Great Lakes Bulk Carriers to Wave-Excited
Loads”: SNAME Transactions 1977.
[C.7] American Bureau of Shipping,: “Rules for Building and Classing Steel Ships” 1996.
[C.8] Glen, I., Paterson,R.B., Luznik, L., “Sea Operational Profiles for Structural Reliability
Assessments”, FTL Report 4737C to SSC for Project SR-1388, June 1998.
[C.9] Hogben, N., Dacunha, N., Olliver, G., “Global Wave Statistics” BMT 1986.
[C.10] “Wind and Wave Climate Atlas”, (4 Volumes), TP 10820, Transport Canada 1991
[C.11] Lee, W.T. and Bales, S.L., “Environmental data for Design of Marine Vehicles”, SSC
SNAME Symposium, Arlington, VA, 1984.
[C.12] ITTS Seakeeping Committee Report, “15th International Towing Tank Conference”, The
Hague, September 1978.
[C.13] Bretschneider, C.L., “Wave Variability and Wave Spectra for Wind-Generated
Gravity Waves”, US Army Corps of Engineers Tech. Memo 118, 1959.
[C.14] Sea Systems Controllerate Publication SSCP23, “Design of Surface Ship Structures” UK
Navy 1989
[C.15] American Bureau of Shipping, “The ABS SAFEHULL System”, 1996
[C.16] Lloyd’s Register, “ShipRight - Design, Construction and Lifetime Ship Care
Procedures”, 1996
[C.17] Germanischer Lloyd, “POSEIDON” - Software.
[C.18] Det Norske Veritas, “NAUTICUS HULL”
[C.19] Sikora, J.P., Dinsenbacher,A., and Beach, J.E., “A Method for Estimating Lifetime
Loads and Fatigue Lives of SWATH and Conventional Monohull Ships”, Naval
Engineers Journal, May 1983.
[C.20] Sikora, J.P., and Beach, J.E., “Automated Method for Predicting Maximum Lifetime
Loads and Fatigue Lives of Ships”,:9th Annual Energy Sources Technology Conference,
New Orleans, LA, 1986
[C.21] “Principles of Naval Architecture - Volume III”, Edited by Edward V. Lewis,
SNAME 1988
[C.22] Salveson,N, Faltinsen,O.M., “Ship Motions and Sea Loads” SNAME Transactions, 1970
[C.23] “Sea Loads on Ships and Offsghore Structures”, Faltinsen,O.M., Cambridge
University Press, 1990
[C.24] “Mechanics of Wave Forces on Offshore Structures”, Sarpkaya, T., and Isaacson,
M.,:Published by Van Nostrand Reinhold Company, 1981

[C.25] McTaggart,K., “SHIPMO7 - An Updated Strip Theory Program for Predicting Ship
Motions and Sea Loads in Waves”, Defence Research Establishment Atlantic Tech
Memo 96/243.1997.
[C.26] Chalmers, D.W., “Design of Ship Structures”, Ministry of Defence, London, HMSO,
1993.
C.27] Hobbacher, A., “Stress Intensity Factors of Welded Joints”, Engineering Fracture
Mechanics,. Vol 46, No. 2, 1993.
[C.28] Stambaugh K., Lawrence, F., and Dimitriakis, S., “Improved Ship Hull Structural Details
Relative to Fatigue”: SSC –379, 1994.
[C.29] Yoneya, T., Kumano,A., Yamamoto,N., and Shigemi, T., “Hull Cracking of Very Large
Ships”,: Integrity of Offshore Structures-5, Scotland, 1992.
[C.30] Ma, K.T., Bea,R., “Repair Management System for Fatigue Cracks in Ship Critical
Structural Details”, Report SMPIII-1-1, University of Berkeley, California, 1994.
[C.31] Miner, M.A., "Cumulative Damage in Fatigue", Trans. ASME, J. Applied Mechanics,
Vol. 12, Nov. 3, Sept., 1945.
[C.32] Moan, T and Berge, S., Proceedings of the 13th International Ship and Offshore
Structures Congress, Elsevier Science, Oxford, 1997
[C.33] British Standards Institution, “Code of Practice for Fatigue Design and Assessment of
Steel Structures”, BS 7608, 1993.
[C.34] Design of Surface Ship Structures, Sea Systems Controllerate Publication 23, Volume
1,UK MoD(PE), December 1989.
[C.35] Design Procedures for Canadian Forces Surface Ships, D-03-002-008/SG-001
[C.36] Lloyd’s Register, “Ship Right - Fatigue Design Assessment Procedure”, 1996.
[C.37] Yee, R. and Paterson, B., “Review of Fatigue Design Methods Employed in SSCP23
Surface Ship Design Standard, FTL Report 3762-14, 1995
[C.38] American Welding Society, “Structural Welding Code”, AWS D.1.1-92, 1992.
[C.39] Canadian Standards Association, “CSA Standard CAN/CSA-S473-92 - Steel Structures,”
1992.
[C.40] Niemi, E.,”Recommendations Concerning Stress Determination for Fatigue Analysis of
Welded Components”, IIW Doc. XIII-1458-92, 1993.
[C.41] American Bureau of Shipping, “Guide for Fatigue Strength Assessment of Bulk Carriers”,
Rules for Building and Classing Steel Vessels, Part 5, Section 3, Appendix 5/3AA, 1996.
[C.42] European Convention for Constructional Steelwork, “Recommendations for the Fatigue
Strength of Steel Structures”, ECCS, Technical Committee 6 - Fatigue, 1985.
[C.43] British Standards Institute, “Draft Code of Practice for Fatigue Design of Steel
Structures”, WEE/44 B.S.I, London, 1987.
[C.44] HMSO, “Offshore Installations: Guidance on Design and Construction”, U.K.
Department of Energy, London, 1984.
[C.45] Stacey, A. and Sharp, J.V., “ The Revised HSE Fatigue Guidance”, Proc. 14th Conf.
Offshore Mechanics and Arctic Engineering, Volume III, Materials Engineering, 1995.

[C.46] Hugill, P.N. and Sumpter, J.D.G., “Fatigue Life Prediction at a Ship Deck /
Superstructure Intersection”, Strain, Aug., 1990.
[C.47] The Welding Institute, “Update of Survey on Corrosion Fatigue Tests”, Health and Safety
Executive Report OTH 92 392, Her Majesty’s Stationery Office.
[C.48] Health and Safety Executive, “ Offshore Installations: Guidance on Design, Construction,
and Certification, 4th Edition, London, 1995.
[C.49] Gurney, T.R., “The Influence of Thickness on the Fatigue Strength of Welded Joints, Proc.
of 2nd Int. Conf. on Behaviour of Offshore Structures (BOSS’79), Paper 41, 1979, BHRA
Fluid Engineering, Cranfield, England, pp 523-534
[C.50] Gurney, T.R., Fatigue of Welded Structures, Second Edition, Cambridge University
Press, Cambridge, U.K., 1984.
[C.51] Maddox, S.J., Fatigue Strength of Welded Structures, Second Edition, Abington Press,
Cambridge, U.K., 1991.
[C.52] Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables,
Edited by Milton Abramowitz and Irene A. Stegun, Dover Publications, Inc., New York,
1972.
[C.53]Mansour, A.E., et al., “An Introduction to Structural Reliability Theory”, SSC Report 351,
1990.
[C.54] Mansour, A.E., et al., “Probability Based Ship Design Procedures: A Demonstration”,
SSC Report 368, 1993.
[C.55] Mansour, A.E., et al., “Assessment of Reliability of Ship Structures”, SSC Report 398,
1997.
[C.56] Madsen, H.O., Krenk, S., Lind, N.C., “Methods of Structural Safety”, Prentice Hall Inc.,
Englewood Cliffs, N.J., 1986.
[C.57] Hasofer, A.M., Lind, N.C., “An Exact Invariant First Order Reliability Format”, Journal
Eng. Mech. Division ASCE, Vol. 1000, No. EM1, pp. 111-121, Feb 1974.
[C.58] Rackwitz, R., Fiessler, B., “Structural Reliability under Combined Random Load
Sequences”, Computers and Structures, Vol. 9, pp. 489-494, 1978.

Part D - Examples
PART D - EXAMPLES
D.1 EXAMPLE 1: LEVEL 2 FATIGUE ANALYSIS SAMPLE APPLICATION
D.1.1 Objective
The objective of this example is to perform a Level 2 Fatigue Analysis for an opening in
the deck of a conventional dry cargo bulk carrier. An illustration of the location of
interest is shown in Figure D.1.1.1. Specifically, the location of interest is at the aft end
of a midship cargo hatch coaming. The coaming is supported by stiffeners that intersect
the deck plate. This location is susceptible to fatigue cracking and an estimation of the
fatigue life will be completed.
Figure D.1.1.1: Location of Interest for Fatigue Calculation
D.1.2 Assessment Approach
The approach followed in this example is summarized in seven steps:
(1) Define problem data;

(2) Estimate extreme characteristic loads;
(3) Estimate service characteristic loads;
(4) Estimate local nominal stresses;
(5) Calculate hot spot stresses;
(6) Define service stress range distribution; and
(7) Perform fatigue damage summation.
Fatigue-Resistant Detail Design Guide for Ship Structures D-1

Part D - Examples
D.1.2.1 Define Problem Data
The bulk carrier in this example is a ‘handy-size’ bulk carrier operating on short
term contracts, (“tramp” service) which takes it all over the world. The vessel is rated at
a service speed of 15 knots with its principle particulars outlined in Table D.1.2.1.
Table D.1.2.1: Vessel Characteristics

Principal Particulars:
Length (L) 190.00 m

Beam (molded) (B) 27.60 m
Depth 14.80 m
Draft Max (SW) 10.93 m
Draft Max (TFW) 11.40 m
Displacement (SW) 47,043 tonnes
Displacement (TFW) 48,075 tonnes
Block Coefficient (CB) 0.80
Deck Section Modulus (Zdeck) 2.398 x1010 mm3
The structural geometry in the region of interest is shown in Figure D.1.2.1. The
geometry in question consists of 12 mm deck plate, a coaming bracket and a longitudinal.
The angle between the coaming bracket and deck plate is 80o degrees. The longitudinal
runs underneath the plate in line with the coaming bracket.
Coaming Bracket
Deck Plate
Longitudinal
Figure D.1.2.1: Structural Geometry near Region of Interest

The structure is fabricated out of AH32 steel, which has the following material
properties: yield strength 300 MPa and ultimate tensile strength 400 MPa. The hot spot
stress life (SN) curve for the material in a non-corrosive environment is taken as:
Log(N) = Log(a) − mLog( ? σ Hot Spot ) (D.1.1)

Part D - Examples
Table D.1.2.2: Hot Spot SN Curve

Log(a) m
7
N<10 12.182 3.0
N>107 15.637 5.0
1000
In Air S-N Curve

Stress Range [MPa]
100
10
1.0E+4 1.0E+5 1.0E+6 1.0E+7 1.0E+8 1.0E+9
Number of Cycles
Figure D.1.2.2: In Air Hot Spot SN Curve

D.1.2.2 Estimate Extreme Characteristic Load
Both the maximum hogging and sagging moment are calculated for the location
of the deck. These moments are determined through the class society rules. Since the
crack is located on the deck, we can assume that the primary loading is due to hull girder
bending, resulting in the following sagging moments:
M dS = −0.11k wm C w L2 B(C B + 0.7) [kNm]

3
 300 − L  2
C w = 10.75 −  for 100m < L < 300m) (D.1.2)
 100 
k wm = 1.0 (between 0.4L and 0.65L)

Part D - Examples
Substituting for the above constants results in:

3
 300 −190  2
C w = 10.75 −   = 9.596
 100 
M dS = −0.11(1.0) 9.596 (190) 2 27.6 (0.8 + 0.7)
M dS = −1,577,577 kNm
Similarly the hogging moment can be calculated from:

M dH = 0.19 k wm C w L2 BC B [kNm]
M dH = 0.19 (1.0) 9.596 (190) 2 27.6 (0.8) (D.1.3)
M dH = 1,453,283 kNm
D.1.2.3 Estimate Service Characteristic Loads
The above calculated maximum hogging and sagging moments are based on the
extreme event with an assumed probability of exceedance of 1×10-8 , while it has been
demonstrated that the majority of fatigue damage results from smaller wave events (i.e.,
probability of exceedance in the 1×10-4 range. If the loads are assumed to be distributed
according to a Weibull distribution, then the conversion of the characteristic bending
moment from the extreme event to the typical service level of 1×10-4 is performed based
on the ratio:
1
M service  ln p 2  h
ff = =  (D.1.4)
M Extreme  ln p1 
where: h =2.21-0.54 log(L) = 2.21-0.54 log(190)

=0.9795
Substituting the data gives:
1
 ln(10 −4 )  0.9795
f f = 1 − −8 
 ln(10 ) 
f f = 0.4928
Table D.1.2.3: Extreme and Service Bending Moments
Moment Extreme (kNm) Service (kNm)
Sagging -1,577,577 -777,430
Hogging 1,453,283 716,178

Part D - Examples
D.1.2.4 Estimate Local Nominal Stress
The nominal stress at the deck in a region remote from the opening can be
determined by simply dividing the service bending moments by deck section modulus
(Zdeck) from Table D.1.2.1. The local nominal stress includes the stress concentration
caused by the hatch opening. The local nominal stress can be determined through:
KG M
s nom = (D.1.5)
Zdeck
where M is the service bending moment, KG is the global stress concentration. The
global stress concentration can be determined from Figure D.1.2.3. From this figure, a
KG factor of 5.0 is estimated for this region. However, the peak stress concentration
coincides with the deck plate hatch opening corner illustrated in Figure D.1.2.4. The
location of interest is approximately 900 mm away from this location, therefore a
reduction in the K G is substantiated based on the distance from the maximum stress
concentration to the location of interest. Therefore at the location of interest, a K G factor
of 2.0 is suggested based on the above discussion.
5 b/a = 2.0
4.5 b/a = 1.5
SCF, KG
4
b b/a = 1.0
3.5
r a
3
2.5
2 r/b
0 0.1 0.2 0.3 0.4 0.5
Figure D.1.2.3: Global Stress Concentration (KG)
Top of Coaming
Max KG SCF
Location of Coaming
Interest Bracket
Figure D.1.2.4: Location of Peak Stress Concentration

Part D - Examples
The local nominal stress is computed from equation D.1.5 as
 2 M × 106 
σN = 10 
1.174 × 10 
s N = 1.7036 ×10 −4 M
D.1.2.5 Determination of Hot Spot Stress
The hot spot stress includes the geometrical details and effects due to local
geometrical configuration such as misalignment, angular distortion. etc. In this example,
an idealized welding geometry is used with no misalignment (i.e. Kte,Ktα = 1).
s HotSpot = K g K te K ta s nom (D.1.6)
The local stress concentration caused by the coaming bracket angle (K g) can be
determined from Equation (D.1.7), and the variables are shown graphically in Figure
D.1.2.5
4 t w? 
Kg = 1 + 
3  160 t s 
(D.1.7)
4  (12)(80) 
K g = 1 + 
3  (160) t s 
tw
θ ts
Figure D.1.2.5: Coaming Toe Attachment
Two cases are considered in this example, with and without a doubler plate,
where the thickness of the plate (t s) is changed.

Part D - Examples
Table D.1.2.4: Hot Spot Stresses at Location of Interest

Load Condition Nominal Stress Hot Spot
Doubler No Doubler
tp =24 mm tp =12 mm
Kg=5/3 Kg=2
σnom σhot spot σhot spot
Hogging 59.7 MPa 99.5 MPa 119.4 MPa
Sagging -64.8 MPa -108 MPa -129.6 MPa -
D.1.2.6 Define Service Stress Range Distribution
The stress range is based on the difference between the hogging and sagging
conditions. Therefore the hot spot stress range at the ‘service probability’ of 10-4
encounters can be determined as the algebraic difference between the hog and sag
conditions.
? s hot spot = s h − s s
(D.1.8)
? s hot spot = 99.5 − ( −108) = 207.5 MPa (doubler case)
Similarly, ∆σhot spot = 249.0 MPa for the no doubler case.
The stress range distribution is assumed to follow a Weibull distribution of the following
form:
 ? s 
h
Q(? σ) = exp  −  

  q  
  (see Eq. C.5.3) (D.1.9)
? s hot spot
q=
[ln(n o )] 1 h
where h = 0.9795 (from before), and in this example, q for the doubler case is
calculated as:
207.5
q= = 21.5 MPa
[ ]
1
4 0.9795
ln(10 )
where no = 1/characteristic encounter probability

= 1/10-4
and Q(∆σ) is the probability of exceedance for the stress range (∆σ). Thus Q(∆σ) for
doubler case has the form:

Part D - Examples
  ? s  0.9795 
Q(? σ) = exp  −   (D.1.10)
  21.5  
 
Similarly, q = 25.8 MPa for the case without a doubler plate.
D.1.2.7 Perform Fatigue Damage Summation
The fatigue damage is assessed using the Miner-Palmgren linear damage model.
This model assumes that the cumulative damage is the sum of individual cyclic damage
and that the damage is independent of the stress cycle sequence. The damage summation
is in the following form (see Eq. C.6.7):
n
D = ∑di = ∑ Ni (D.1.11)
i
Two methods are described for the damage summation, a discrete and closed form
solution approach.
D.1.2.7.1 Discrete Approach
In this approach, the stress range distribution is broken down into a histogram of
at least 20 bins. The fatigue life calculation uses the mid range value for each bin. The
fatigue damage di resulting from each stress range magnitude or bin is determined and
then the total damage is calculated as the summation of the di’s , as follows.
Table D.1.2.5: Sample Discrete Fatigue Damage Summation

(Doubler Case)
Stress Range Mid Stress Range Number of Cycles Cycles to Damage Ratio,
Value, ∆σ [MPa]* Probability, pi ‡ per Year, ni † Failure, Ni ¥ di = ni / Ni
10 6.06E-01 3030410 4.34E+10 6.99E-05
30 2.35E-01 1173067 1.78E+08 6.58E-03
50 9.43E-02 471272.6 1.39E+07 3.40E-02
70 3.83E-02 191609.9 4.43E+06 4.32E-02
⇓ ⇓ ⇓ ⇓ ⇓
450 2.54E-09 0.012713 1.67E+04 7.62E-07
470 1.08E-09 0.005397 1.46E+04 3.69E-07
Sum 1.98E-01

Part D - Examples
(No Doubler Case)

10 5.41E-01 2706253 4.34E+10 6.24E-05
30 2.44E-01 1218098 1.78E+08 6.83E-03
50 1.13E-01 567131.2 1.39E+07 4.09E-02
70 5.34E-02 266860.2 4.43E+06 6.02E-02
⇓ ⇓ ⇓ ⇓ ⇓
450 5.25E-08 0.262716 1.67E+04 1.57E-05
470 2.57E-08 0.128301 1.46E+04 8.76E-06
Sum 3.53E-01
* All probable stress ranges are captured in 20 MPa stress range intervals (i.e. 0 to 20 mid value = 10, 20 to
40 mid value = 30, etc.)
‡
Probabilities for each stress range interval is estimated based on the difference between the probability of
exceedance of its upper and lower bounds using equation D.1.10 (i.e. p i =Q(∆σlower) - Q(∆σupper))
†
Assuming a 20 year design life with 108 wave encounters, the number of cycles per year for each stress
range is calculated as the product its probability and the number of wave events per year (i.e., 108 p i /
20).
¥
Cycles to failure are calculated using equation D.1.1 or using Figure D.1.2.2.
D.1.2.7.2 Closed Form Solution
Alternatively, a closed form solution can be obtained for this problem. This
solution has the benefit that the stress distribution is not discretized into a series of bins
the solution process is continuous along its domain. The closed form solution is of the
following form (equation C.6.10):
 m1  h  m ?s h 
 (q )  m1  ? s slope   (q ) 2
m
 2  slope  

D = fz td G1 + ; + ? 1 + ;
 a1  h  q   a2  h  q  
    
(D.1.12)
Where:a 1 and m1 = the fatigue design curve parameters for N ≤ 107 cycles (Table D.1.2.2)
a2 and m2 = the fatigue design curve parameters for N > 107 cycles (Table D.1.2.2)
γ ( ) = incomplete gamma function
Γ( ; )= complementary incomplete gamma function
∆σslope = stress range at which change in slope occurs
h and q = Weibull distribution parameters
f z = average zero crossing period
td = design life in seconds (20 years = 6.3 x 108 s)

Part D - Examples
D.1.2.7.3 Fatigue Life
Regardless of the solution approach used for the linear damage summation the same
results should be obtained.
Table D.1.2.6: Cumulative Damage Results
Portion of Fatigue Life No Doubler With Doubler
used in a Year
Discretized Approach 0.353 0.198
Closed Form Solution 0.260 0.143
The fatigue life estimates can be determined from the inverse of the damage summation
(1/D). If we assume that the vessel is in operation only 85% of the time, the expected
fatigue life is (1/0.85D):
Table D.1.2.7: Fatigue Life Estimates

No Doubler With Doubler
[Years] [Years]
Discrete 3.3 5.9
Closed Form 4.5 8.2
The results of this example indicate that fatigue cracking will initiate or become
visible after a service life of four to eight years. This analysis does not indicate that the
predicted fatigue cracks will grow to a critical size in the estimate time period.
It should be noted that the Level 2 approach presented here, includes several
conservative assumptions having a cumulative effect. In general, the rule based load
estimation is conservative and so is the rule based stress analysis with analytic stress
concentration factors. for example, the bracket toe detail geometric stress concentration
factor did not include the presence of deck longitudinals which would increase the
bending stiffness of the plate locally and thus reduce the stress concentration factor.
A more detailed or precise analysis of the scenario described in this example
could be completed by performing a Level 3 analysis, similar to that presented the
Section D.2. This example develops a load history based on the expected operation of the
vessel. In addition, the behavior of the vessel structure is more accurately considered
using finite element analysis.

Part D - Examples
D.2 EXAMPLE #2 - LEVEL 3 ANALYSIS
Problem Statement:
Perform a Level 3 Fatigue Analysis for an opening in the deck of a conventional dry
cargo bulk carrier, as shown in Figure D.1.1.1 (previous section). As in the last example,
the location of interest was at the aft end of a midship cargo hatch coaming. The
coaming is supported by stiffeners that intersect the deck plate. In the context of the last
example, the spectral approach is being employed to produce a more realistic estimate of
the fatigue life.
D.2.1 Assessment Approach
The approach that was applied is summarized in six steps:
(1) Establish the operational profile;

(2) Calculation of loads;
(3) Determine reference hot spot stress ranges;
(4) Define short-term stress range distribution;
(5) Define long-term stress range history;
(6) Perform fatigue damage summation.
D.2.2 Establish Operational Profile
D.2.2.1 Overview
For this analysis, an operational profile of the vessel needs to be established.

Whenever a ship is at sea, its “operating mode” is defined by its speed, heading relative
to the waves, and the sea conditions encountered by the vessel. The operational profile is
a compilation of all individual operating modes, covering the entire operational envelope
of the vessel and is usually presented in the form of a probability of speed and heading at
prescribed sea conditions.
D.2.2.2 Vessel Description
The analysis was conducted for the same vessel used in Example 1 (see Section
D.1.2.1).

Part D - Examples
D.2.2.3 Establish Operational Profile
D.2.2.3.1 Establishing Ship Transit Profile
Operational profiles were developed from the operational data recorded in the
ship’s log covering a one-year period. The encountered wave climates were recorded in
the ship’s log in the terms of Sea State, and so the operational profile was developed on
this basis.
For a detailed description of the development of the profile and joint probability
distributions from the logs and examples, see Reference D.2.
The resulting distributions of Speed with Sea State and Heading with Sea State
are shown in Tables D.2.2.1 and D.2.2.2. The route of the vessel is transcribed into the
time spent in each Marsden Zone in Table D.2.2.3, and the time spent in each loading
condition and in port is given in Table D.2.2.4.
Table D.2.2.1: Joint Probability of Speed and Sea State

Sea State
Speed (kn) 1 2 3 4 5 6 7 SUM
10-12 0.0078 0.0000 0.0000 0.0072 0.0130 0.0326 0.0148 0.0755
12-14 0.0305 0.0315 0.2140 0.2112 0.1957 0.1509 0.0000 0.8339
14-16 0.0000 0.0196 0.0261 0.0285 0.0067 0.0097 0.0000 0.0906
SUM 0.0383 0.0511 0.2401 0.2470 0.2155 0.1932 0.0148 1.0000
Table D.2.2.2: Joint Probability of Relative Heading and Sea State

Sea State
Rel. Heading 1 2 3 4 5 6 7 SUM
Head Seas 0.0066 0.0064 0.0316 0.0338 0.0313 0.0319 0.0148 0.1564
Strbd. Bow 0.0212 0.0187 0.1154 0.1191 0.1124 0.1022 0.0000 0.4890
Strbd. Beam 0.0038 0.0114 0.0365 0.0371 0.0268 0.0224 0.0000 0.1380
Strbd. Quart. 0.0027 0.0071 0.0250 0.0252 0.0191 0.0157 0.0000 0.0948
Following 0.0034 0.0099 0.0322 0.0327 0.0238 0.0199 0.0000 0.1218
SUM 0.0378 0.0534 0.2407 0.2479 0.2134 0.1921 0.0148 1.0000

Part D - Examples
Table D.2.2.3: Listing of Geographical Zones (Marsden Zones) ship traversed and
relative time spent in each zone .
Marsden Zone Time in Marsden Zone Time in Marsden Zone Time in
Zone Zone Zone
5 0.0140 28 0.0290 59 0.0140
11 0.0140 29 0.0300 60 0.0340
13 0.0280 30 0.0690 61 0.0340
16 0.0150 32 0.0070 62 0.0280
17 0.0070 33 0.0140 66 0.0140
18 0.0140 36 0.0270 67 0.0340
19 0.0070 37 0.0350 68 0.0350
20 0.0480 39 0.0050 69 0.0070
21 0.0620 40 0.0500 75 0.0360
22 0.0200 47 0.0210 84 0.0340
25 0.0070 50 0.0350 85 0.0280
26 0.0340 56 0.0280 90 0.0280
27 0.0340 58 0.0210 ~
sum: 0.3036 sum: 0.3706 sum: 0.3258
total sum: 1.0000
Table D.2.2.4: Operational Profile Summary

Total Operation Time (hours) Number of Data Entries
Loaded Condition 3624.4 151
Ballast Condition 1563.6 66
Time in Port 3552.0 ~
Total 8739.8 ~
This procedure results in the distribution of time spent by the vessel in various
geographical zones, and the distribution of headings and speeds with various sea
conditions. The following is a commentary on the development of the sea operational
profiles:
a) The resulting probabilities in an operational profile will depend on the quality and
quantity of the recorded operational states aboard the vessel. For example, vessel
speed, heading, and environmental conditions can be collected on a 10 minute,
hourly, four-hourly (watch) or, sometimes, daily basis. Further, the actual
measurements are prone to errors, as the majority of the measurements except ship
speed are based on visual observations.

Part D - Examples
b) For a vessel engaged in “tramp” service or otherwise irregular routes (such as in this
example), it is desirable to obtain records over a long duration (several years). This
permits assessment of the variability in data and will improve the statistical
significance of the resulting probabilities. In this example, the data is representative
of only one year in the ship’s life.
c) The probability tables indicate the relative amount of vessel time at sea in each
operational state (e.g., speed/Sea State, or relative heading/Sea State or transiting a
Marsden Zone). As indicated, the sum of all of the operational state probabilities is
equal to one. This means that the “volume” under the three-dimensional surface plot
of probability (Speed; Sea State) or probability (Rel. Heading; Sea State) is equal to
unity.
d) Although heading information for the ship was recorded in degrees and covered 360
degrees, in the development of the vessel relative heading/Sea State probability table,
the headings were discretized into five headings. This discretization scheme was
selected for two reasons. First, headings relative to wave direction more precise than
45 degrees are difficult to determine consistently, and thus are not generally recorded
in the ship logs. Second, by assuming port/starboard symmetry of vessel response to
incident waves, pair-wise groupings of relative headings was possible:
Head Seas
Strbd. Bow same as Port Bow
Strbd. Beam same as Port Beam
Strbd. Quartering same as Port Quartering
Following Seas
In a statistical sense, this meant that the probabilities of pair-wise groups (e.g., Strbd.
Bow and Port Bow, etc.) were simply added together to yield overall probability of
“Strbd. Bow” entry, as shown in the Operational Profile Table D.2.2.2
e) The operational profile was developed from the ship’s log. Voluntary speed
reduction and/or change in course to avoid bad weather is a common practice and is
left to the discretion of the master. The data thus may be biased in this regard. If the
ship reduced speed and/or changed heading reduce motions during heavy weather,
this was presumably recorded in the ship log, and therefore it is reflected in the
operational profile.
f) The source for the environmental data used in this analysis was the Global Wave
Statistics, [Ref D.3]. Over 55 million visual observations of wind speed and
direction, as well as wave height, period, and direction, obtained by the United
Kingdom Meteorological Office marine data bank, were used to compile wave
statistics for 104 zones (sometimes referred to as a Marsden Zones).

Part D - Examples
g) In this analysis, for each Marsden Zone, the wave statistics are presented on an annual
basis as well as for uni-directional seas only. Wave data has been normalized to
approximately 1000 observations.
D.2.2.3.2 Transposing Sea States and Wave Heights
Wave statistics in Marsden Zones are given in terms of significant wave heights,
and zero crossing periods [Ref. D.3]. In this example, the operational profile data is
defined in terms of Sea State. Therefore a transformation of wave statistics from wave
heights to Sea State is required. The transformation consists of the following steps:
• Assume that the significant wave height follows a three-parameter Weibull

cumulative probability distribution given by:
 (H − m ) 
h
− S 
F(H S ) = 1− e  q  (D.2.1)
where m is the location parameter, q is the scale parameter, and h is the shape
parameter of the Weibull distribution.
• Taking the natural logarithm of both sides of the equation D.2.1 twice, it can be
reduced to an equation of a straight line if plotted on the ln-ln versus ln scale.
• The wave data is then plotted, and the line that fits the data best is drawn using
linear regression analysis.
• The parameter m is found by iteration, so that the regression error is minimized.
The parameters h and q are determined from the slope and intercept of the line.
• With the cumulative probability distribution function calibrated, the probabilities
of Sea States are calculated as
f (Sea State ) = F(HS2 )− F(HS1 ) (D.2.2)
Note that this transformation from wave heights to Sea States was necessary only
because the operational profile was recorded on the basis of Sea State. For operational
profiles based on wave height this step is not necessary.

Part D - Examples
D.2.2.3.3 Calculation of Composite Wave Scatter Diagram
The composite probability distribution of wave heights and periods for all
Marsden Zones combined is calculated to produce a “Composite Marsden Zone” scatter
diagram. The joint probability of each significant wave height/zero crossing period
combination is multiplied by the appropriate entry in the Table D.2.2.3 of the Operational
Profile Data for n zones in which ship operates. The results are then summed for
appropriate combinations of wave height and zero crossing period.
This is expressed in terms of an equation as:

N′
f mc ( H S ;TZ ) = ∑ µ i f mi (H S ;TZ ) (D.2.3)
i =1
and is provided as a sample calculation in the following table:
Table D.2.2.5: Wave Data by Marsden Zone [from Ref D.3]

5<Tz<6 seconds
Hs (m) Marsden Marsden Marsden Marsden
Zone. #5 Zone. #11 Zone#… Zone. #90
0-1 0.0790 0.0940 … 0.0080
1-2 0.1400 0.1210 … 0.0130
2-3 0.0700 0.0630 … 0.0040
3-4 0.0230 0.0270 … 0.0010
4-5 0.0060 0.0110 … 0.0000
5-6 0.0010 0.0040 … 0.0000
6-7 0.0000 0.0020 … 0.0000
7-8 0.0000 0.0010 … 0.0000
8-9 0.0000 0.0000 … 0.0000
9-10 0.0000 0.0000 … 0.0000
10-11 0.0000 0.0000 … 0.0000
11-12 0.0000 0.0000 … 0.0000
12-13 0.0000 0.0000 … 0.0000
13-14 0.0000 0.0000 … 0.0000
>14 0.0000 0.0000 … 0.0000
SUM 0.3190 0.3230 … 0.0260
The data from Table D.2.2.3 and Table D.2.2.5 are combined via the equation above. For
example,
f mc (HS =0 −1 m , 5 < TZ < 6 s )= 0.0790 ⋅ 0.014 + 0.0940 ⋅ 0.014 + K + 0.0 ⋅ 0.0080 = 0.0524
where 0.014, 0.014, and 0.0280 are relative time spent in the zones 5, 11 and 90
respectively (Table D.2.3). Entry of 0.0524 appears in the Composite Marsden Zone
Scatter Diagram (Table D.2.6) for Hs = 0-1 m and 5<Tz <6 seconds.

Part D - Examples
In this example the ship traversed 38 Marsden Zones, and a typical scatter
diagram for each Marsden Zone is divided into 11 zero crossing periods (e.g., Tz <4 sec.,
4<Tz<5 sec., 5<Tz<6 sec., etc.). The above process is repeated for all the wave height
ranges and crossing periods and the result is a Composite Marsden Zone Scatter Diagram
(Table D.2.2.6). This diagram is unique for the example operational profile. Different
profiles would result in a different Composite Marsden Zone scatter diagram
Table D.2.2.6: Composite Marsden Zone Scatter Diagram

Hs Tz (sec.)
(m)
<4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 >13 SUM
0-1 19.5 51.3 52.4 29.7 10.8 2.7 0.4 ~ ~ ~ ~ 166.8
1-2 5.5 38.5 84.5 96.1 68.5 31.9 10.3 2.4 0.5 ~ ~ 338.1
2-3 1.1 11.6 36.4 60.8 67.0 47.8 22.8 7.7 2.0 0.5 ~ 257.7
3-4 0.3 3.0 11.3 22.5 32.3 31.1 19.6 8.6 2.9 0.8 0.1 132.3
4-5 0.0 0.8 3.4 7.4 12.2 14.6 11.5 6.1 2.3 0.8 0.2 59.3
5-6 ~ 0.2 1.0 2.6 4.7 6.2 5.6 3.5 1.4 0.5 0.2 25.8
6-7 ~ 0.0 0.3 0.9 1.7 2.5 2.6 1.7 0.9 0.4 0.1 11.1
7-8 ~ 0.0 0.2 0.3 0.7 1.0 1.2 0.9 0.6 0.2 0.0 5.1
8-9 ~ ~ 0.0 0.1 0.3 0.4 0.6 0.5 0.3 0.0 ~ 2.3
9-10 ~ ~ ~ 0.0 0.1 0.2 0.2 0.2 0.2 0.0 ~ 0.9
10-11 ~ ~ ~ ~ ~ 0.0 0.2 0.1 0.0 ~ ~ 0.3
11-12 ~ ~ ~ ~ ~ ~ ~ 0.0 ~ ~ ~ 0.0
12-13 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0.0
13-14 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0.0
>14 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0.0
SUM 26.4 105.4 189.5 220.4 198.3 138.5 74.8 31.7 11.2 3.2 0.5 999.8
It can be seen that the sum of all probabilities is not equal to 1000. The small
difference is due to rounding errors and is not significant. It will also be noted that the
values obtained from the application of equation C.3.1 have been multiplied by 1000.
Table D.2.2.6 then represents a single (composite) wave data set for the encounter
probabilities of waves for this specific route.
As noted earlier, to calculate the joint speed/ heading wave probabilities it is necessary in
this case to convert Sea state data from the ship’s operational data set, to wave height
data as used in the wave environment data set. This is illustrated in Table D.2.2.7 for the
5<Tz<6 sec condition. This step is not necessary if operational data is available in terms
of Wave Heights.

Part D - Examples
Table D.2.2.7: Transformation of Sea State and Wave Height.

Hs [m] Composite Sea State Range of Combined
Marsden Z. Hs (m) Marsden Z.
0-1 0.0524 ⇒ 1 0-0.1 0.0000
1-2 0.0845 2 0.1-0.5 0.0524
2-3 0.0364 ⇒ 3 0.5-1.25 0.0699
3-4 0.0113 4 1.25-2.5 0.0500
4-5 0.0034 ⇒ 5 2.5-4 0.0143
5-6 0.0010 6 4-6 0.0026
6-7 0.0003 ⇒ 7 6-9 0.0002
7-8 0.0002 sum 0.1895
8-9 0.0000
9-10 0.0000
10-11 0.0000
11-12 0.0000
12-13 0.0000
13-14 0.0000
>14 0.0000
Sum 0.1895
D.2.2.3.4 Calculation of Speed/Heading Probabilities
Having established the distribution of time at heading and speed for the range of
Sea States, and having developed a wave climate (height and zero crossing period) for a
composite “zone” which reflects the distribution of time in all the geographical zones, it
remains to combine these to obtain the joint probabilities of speed, heading and sea
conditions.
The probability of operating in a particular mode characterized by speed V,

relative heading θ, and Sea State - as defined by wave height and zero crossing period
(Hs; Tz), - is equal to the product of the probability of operating in that range of speed
times the probability of being in that range of heading times the probability of being in
that Sea State. This is defined as:
f S TOTAL = f mc ( HS ;TZ ) ⋅ f V ( V| ( HS ; TZ )) ⋅ f? ( ? | ( HS ; TZ )) (D.2.4)

14243 1442443 1442443
I II III
Term I is the probability of wave height/zero crossing period in the Combined Marsden
Zone. Term II is the conditional probability of speed V given a Sea State (Hs; Tz ), and
term III is the conditional probability of heading θ for a given Sea State (Hs; Tz):
prob( V and (H S ;TZ ))

f V ( V | ( H S ; TZ )) = (D.2.5)
prob ( H S; TZ )V

Part D - Examples
prob (? and (H S ;TZ ))

f ? ( ? | ( HS ; TZ )) = (D.2.6)
prob ( H S; TZ ) ?
where:
• prob (V and (Hs; Tz)) = is the joint probability of speed and sea state (entries in the
Operational Profile Table D.2.2.1 with Sea State
transformed to Wave Height and Crossing Period).
• prob (Hs; Tz) V = is the probability of the sea state given in the bottom row of
the Operational Profile Table D.2.2.1.
• prob (θ and (H s; Tz)) = is the joint probability of heading and sea state (entries in
the Operational Profile Table D.2.2.2).
• prob (Hs; Tz)θ = is the probability of sea state given in the bottom row of the
Operational Profile Table D.2.2.2.
The process is illustrated as follows. From Table D.2.2.1
Sea State
Speed (kn) 1 2 3 4 5 6 7 SUM
10-12 0.0078 0.0000 0.0000 0.0072 0.0130 0.0326 0.0148 0.0755
12-14 0.0305 0.0315 0.2140 0.2112 0.1957 0.1509 0.0000 0.8339
14-16 0.0000 0.0196 0.0261 0.0285 0.0067 0.0097 0.0000 0.0906
SUM 0.0383 0.0511 0.2401 0.2470 0.2155 0.1932 0.0148 1.0000
prob( V and ( HS ;TZ ))

f V ( V | ( H S ; TZ )) = = f V ( V = (12 to 14 kn ) | ( H S ; TZ )
prob (H S ;TZ ) V
0.2140
=
0.2401
and from Table D.2.2.2
Sea State
Rel. Heading 1 2 3 4 5 6 7 SUM
Head Seas 0.0066 0.0064 0.0316 0.0338 0.0313 0.0319 0.0148 0.1564
Strbd. Bow 0.0212 0.0187 0.1154 0.1191 0.1124 0.1022 0.0000 0.4890
Strbd. Beam 0.0038 0.0114 0.0365 0.0371 0.0268 0.0224 0.0000 0.1380
Strbd. Quart. 0.0027 0.0071 0.0250 0.0252 0.0191 0.0157 0.0000 0.0948
Following 0.0034 0.0099 0.0322 0.0327 0.0238 0.0199 0.0000 0.1218
SUM 0.0378 0.0534 0.2407 0.2479 0.2134 0.1921 0.0148 1.0000

Part D - Examples
prob (? and (H S ;TZ ))

f ? ( ? | ( HS ; TZ )) = = f ? (? = (Head Seas )(
| H S ; TZ )
prob ( HS ;TZ ) ?
0.0316
=
0.2407
Total probability of a given operating mode then is:
f S TOTAL = f mcomb (HS ;TZ ) ⋅ f V ( V| ( HS ; TZ )) ⋅ f? ( ? | ( HS ; TZ ))

1442443 1442443 1442443
I II III
0.2140 0.0316
f STOTAL = 0.0699
123 ⋅ ⋅ = 0.00818
I 0.2401
123 123 0.2407
II III
where Term I comes from Table D.2.2.7.
Summarizing this calculation, the probability of occurrence of Sea State 3 in Head

seas, with the vessel speed between 12 to 14 knots is 0.00818. These calculations are
repeated for all combinations of speed and headings. Therefore, in this example, there
are 3 ranges of speed, 5 ranges of heading and 7 Sea States. Thus, the theoretical number
of operating mode combinations for a single zero crossing period is 3x5x7=105 entries.
The calculated data for a single zero crossing period is as shown in Table D.2.8
The values shown in the table are probabilities standardized by multiplying by

1000. Thus, the probability of occurrence of Sea State 3, in Head seas with a vessel
speed between 12 to 14 knots is 8.2/1000=0.0082. In the table sign “~” indicates
absolute zero, and numbers are given with one significant digit. This means that
probabilities less than 0.0001 are not shown.
For the operational profile used in this example there are 3 ranges of speed, 5
ranges of heading, and 7 sea states covering 11 zero crossing periods. The theoretical
number of operating mode combinations is 3x5x7x11=1155 entries. The fact that some
probabilities are zeros or very small, reduces the number of combinations for which
vessel response need be evaluated. In this example the total number of operational
modes was 231.

Part D - Examples
Table D.2.2.8: Probabilities of all combinations of Speed, Heading and Sea State for
zero crossing period Tz between 5 and 6 seconds in Combined Marsden Zone.
Head Strbd. Strbd. Strbd. Follng Head Strbd. Strbd. Strbd. Follng
10-12 -- -- -- -- -- -- -- -- -- --
12-14 -- -- -- -- -- 3.85 11.33 6.89 4.29 5.97
14-16 -- -- -- -- -- 2.39 7.03 4.28 2.66 3.71

10-12 -- -- -- -- -- 0.20 0.70 0.22 0.15 0.19
12-14 8.18 29.88 9.45 6.46 8.33 5.84 20.55 6.40 4.36 5.64
14-16 1.00 3.64 1.15 0.79 1.02 0.79 2.78 0.86 0.59 0.76

10-12 0.13 0.46 0.11 0.08 0.10 0.07 0.23 0.05 0.04 0.05
12-14 1.91 6.86 1.64 1.17 1.45 0.34 1.08 0.24 0.17 0.21
14-16 0.07 0.24 0.06 0.04 0.05 0.02 0.07 0.02 0.01 0.01
SPEED Sea State 7

Head Seas Strbd. Strbd. Strbd. Follwng
Bow Beam Quart.
10-12 0.21 -- -- -- --
12-14 -- -- -- -- --
14-16 -- -- -- -- --
D.2.3 Calculation of Loads
D.2.3.1 Overview
A linear strip theory program is used to calculate the structural response of the
hull girder to the range of sea conditions in the problem. This example focuses on
vertical bending moment of the hull in the midships area.

Part D - Examples
D.2.3.1.1 Selection of a Wave Spectrum
While the information that has been established to date identifies the degree of
exposure to each sea condition that is represented in the geographical zones, the seas
have been identified by wave height and zero crossing period. It remains to select a
distribution for the wave energy for these Sea States. It was assumed that representation
of the sea spectra as a fully developed sea was sufficient, and the irregular description of
sea was modeled using the Bretschneider Spectrum [Ref D.4].
The input parameters for the spectrum are the significant wave height Hs and the
peak period Tp . For each Sea State, the Hs used corresponded to the mid value of the Sea
State wave height range. The relationship between the peak wave period Tp . required by
the loads program and the zero crossing wave period (Tz) used so far, based on the wave
data is:
Tp = 1.408Tz (D.2.7)
Note that the program makes the necessary calculations to transcribe from
stationary wave data to encountered wave data due to ship speed and heading relative to
the waves.
D.2.3.1.2 Establishing Vertical Bending Moment
Responses of the example ship in the conditions identified in the previous section
were calculated using the frequency domain linear strip theory code SHIPMO 7 [Ref.
D.5]. Output for each operating mode (ship speed and heading) and sea condition
included: the Response Amplitude Operator (RAO); Root Mean Square (RMS) loads;
and zero crossing period, Tz of the response in irregular seas. Inherent in the use of strip
theory for these calculations is the assumption of a linear relationship between input
wave height and the vessel response.
Table D.2.3.1 shows a sample of the output of Shipmo7 for one Sea State.

Part D - Examples
Table D.2.3.1: Ship Response Calculation: Output from Shipmo7.

Sea State = 3 Sig. Wave H. = 0.88 m
SHIPMO 7 Output
Vert. Bend. Moment
Speed Heading TP prob*100 Tz RMS
(kn.) (sec.) 0. (sec.) (kNm)
13 Head 4.93 0.6 2.20 3803.538
13 Strbd. Bow 4.93 2.1 3.00 3620
13 Strbd. Beam 4.93 0.7 4.90 7481.692
13 Follow. 4.93 0.6 22.50 2640.077
13 Head 6.34 4.2 3.80 7250.077
13 Strbd. Bow 6.34 14 4.70 7789.231
13 Strbd. Beam 6.34 4.4 5.70 9486.154
13 Strbd. Quart. 6.34 3.4 12.80 13314.85
13 Follow. 6.34 4.4 18.40 5894.615
15 Strbd. Bow 6.34 1.7 4.40 7920.692
15 Strbd. Beam 6.34 0.5 5.70 9351
13 Head 7.75 8.2 5.30 13084.62
13 Strbd. Bow 7.75 29.9 5.80 13080.31
13 Strbd. Beam 7.75 9.5 6.30 9622.077
13 Strbd. Quart. 7.75 6.5 13.20 22347.15
13 Follow. 7.75 8.3 17.50 13793.92
: : : : : :
⇓ ⇓ ⇓ ⇓ ⇓ ⇓
: : : : : :
For this example 231 SHIPMO7 runs were completed, accounting for the operating
combinations with any non-zero probability of occurrence. Values of RMS Vertical
Bending Moment (M) were divided into 20 equal bins, covering the range of BM’s from
the minimum to the maximum calculated value. The probability of each bending moment
range is equal to the sum of all probabilities corresponding to the RMS values falling
within the boundaries of each bin. The resulting histogram is shown below.

Part D - Examples
Range of RMS (kNm) Bin # prob.*1000

0 9959 1 182.0 Example #1 Wave Loading Histogram
9959 19918 2 198.9
19918 29877 3 100.4
29877 39836 4 118.6 225
39836 49795 5 107.8 200
49795 59754 6 45.1
59754 69713 7 46.0 175
69713 79672 8 80.5
150
prob.*1000
79672 89631 9 15.4
89631 99590 10 23.2 125
99590 109549 11 13.1
109549 119508 12 7.5 100
119508 129467 13 19.1 75
129467 139426 14 6.1
139426 149385 15 4.9 50
149385 159344 16 1.8 25
159344 169303 17 1.5
169303 179262 18 3.6 0
179262 189221 19 4.2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
189221 199180 20 7.0
199180< 21 5.7 Bin #
Sum: 992.4
Figure D.2.3.1: Vertical Bending Moment (RMS) Distribution
Before these bending moment loads were transferred into the stress ranges in a
particular part or detail of the ship structure, there was a further step to be completed to
obtain an annual 1 distribution of vertical bending moment. It is necessary to repeat the
same procedure as described up to this point using any other ship operating condition, for
example in this case, the “Ballast” operational profile data. This would result in different
stationary operational conditions, associated probabilities and responses.
This has not been undertaken in the example as the procedure has been
demonstrated. The result of a complete analysis would be two histograms, one for
“Loaded” and one for “Ballast” condition. The annual wave loading distribution is
obtained by multiplying each bin of the histograms by an appropriate weighting factor
that reflects the relative time spent in each condition. Note that the Bin # boundaries
should be identical in both histograms. For example, RMS range in bin #1 for the ship
loaded condition operational profile is from 0 to 9959 kNm. The range of RMS in Bin #1
for the ballast condition profile should be the same, i.e., from 0 to 9959 kNm.
The annual probability distribution of wave loading is calculated from:
prob (Bin # i )total = prob (Bin # i )loaded ⋅ + prob (Bin # i )ballast ⋅

3624.4 1563.6
(3624.4 + 1563.6 ) (3624.4 + 1563.6)
(D.2.8)
where 3624.4 is the number of hours in loaded condition and 1563.6 is the number of
hours in ballast condition. These numbers were taken from Table D.2.2.4 of the
operational profile data.
1
Recall that Operational Profiles for example ship were developed on an annual basis.

Part D - Examples
D.2.4 Determine Reference Hot Spot Stress Range
The vertical bending moment histogram defines the bending moments

experienced by the whole vessel under each operational condition. These bending
moments are used to define the stress range at the location of interest for the fatigue
analysis. In order to relate the global vertical bending moments to the localized stresses
at a particular detail, theoretical or numerical methods can be employed. In this example,
a linear elastic finite element model was constructed to relate the vertical moment to the
local stress for the location of interest.
D.2.4.1 Finite Element Model
Both a global and refined substructure model of the hatchway was constructed.
Sub-structuring allows an efficient method of determining localized stresses while
remaining computationally efficient. Boundary conditions from the global model are
passed to the refined substructure model. The global model was subjected to a unit
bending moment of 2000 kN*m.
The location of interest in the welded intersection between an angled support bracket and
the base plate. The bracket supports the coaming around a large central hatchway.
Fatigue cracking is expected in the base plate with the crack running perpendicular to the
longitudinal axis of the vessel. Figure D.2.4.1 illustrates both the global and local models
used in the finite element model. Superimposed on the substructure model are the fatigue
crack location and direction of principle stress
Global Model
Local Model
Figure D.2.4.1: Global and Substructure FEM Model

Part D - Examples
D.2.4.2 Determine Reference Hot Spot Stress Ranges
The hot spot stress approach was used in the analysis. Figure D.2.4.2 plots the
principle longitudinal stress in the base plate starting from the corner of the coaming and
emanating in the longitudinal direction. Two structural conditions are used, with and
without a doubler plate at the location of interest. Figure D.2.4.3 shows the extrapolation
of the stress to the hot spot. The hot spot is located 855 mm from the coaming. Points C
and D define the distances for the linear extrapolation to the location of the hot spot
stress. These points are located 12 and 36 mm respectively away from the coaming
bracket to deck intersection when considering the case with a doubler plate.
The hot spot stress range is calculated by assuming fully reversing moments. For
the doubler plate case:
∆σhot spot =2 Ko Mv …for each (H s ,Tz, V, θ, L.C.)

=2 (0.262) Mv / 2000
=2.62×10-4Mv
where Mv is in kNm and ∆σhot spot is in MPa.

and Ko is the transfer function between hot spot stress and design moment.
1
Stress [MPa] / 2000 kNm
0.8 Without Doubler

With Doubler Plate
X
moment
0.6
0.4
0.2
0
0 500 1000 1500 2000 2500 3000 3500
Distance (X) From Coaming [mm]
Figure D.2.4.2: FEM Stress Distribution along the longitudinal

Part D - Examples
Max. Principal
Stress, σ
C = 0.242
D = 0.169
t σ
2 hot spot
3t
2
A B
t
σnominal
A B
Point of Interest (hot spot)
Figure D.2.4.3: Extrapolation of Hot Spot Stress (With Doubler)
Using the transfer function in Table D.2.4.1, it is possible to convert the vertical bending
moments calculated previously into the hot spot stress range.
Table D.2.4.1: Hot Spot Transfer Functions for the Location of Interest
Doubler No Doubler
Ko 0.262 0.292
∆σHot Spot 2.62×10-4 Mv 2.91×10-4 Mv

Part D - Examples
D.2.5 Define Short-Term Stress Range Distribution
A short-term stress range distribution is constructed for each of the operational

conditions defined in the matrix of stationary conditions (Section D.2.2.3.4). The short
range distribution is a Rayleigh distribution with the following form (equation C.5.5).
 − ∆σ 2 
F∆σi ( ∆σ) = 1 − exp   (D.2.8)
 8m oi 
 
where mo is the zeroth moment of the wave spectrum and is calculated in terms of the
RMS design moment (Mvi).for each operational condition as:
moi = (2Ko Mvi)2 (D.2.9)
D.2.6 Long-Term Stress Range History
The short-term behaviors are summed to develop a long-term stress range

distribution as follows:
Q(? σ ) = ∑ F? s i (? s ) ripi (D.2.10)

all conditions
where pi is the fraction of time at each of the i operational conditions and ri is the ratio of
crossing rate and average crossing rate (zero crossing period, T ). The average zero
crossing rate and crossing rate ratio are calculated as follows:
T
T= ∑ p i Tzi and ri = zi (D.2.11)
all conditions T
If the short-term distributions are summed in a discrete fashion , the long-term

distribution is fitted using a least squares regression to select the appropriate Weibull
distribution parameters q and h, where:
  ∆σ  h 
Q( ∆σ) = 1 − exp  −    (D2.12)
   
q

Part D - Examples
0 20 40 60 80 100
1
0.9 CDF
Weibull
0.8
Cummulative Probability
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
16
24
32
48
56
64
72
88
96
40
80
4
10
Stress Range (∆σ MPa)
Figure D.2.6.1: Long-Term Stress History
The least square Weibull Statistics calculated for the two configuration cases are
tabulated in Table D.2.6.1.
Table D.2.6.1: Wiebull Parametrs for the Long-Term Stress History

With Doubler No Doubler
q(Scale) 26.6 35.7
h(Shape) 1.102 1.060
D.2.7 Perform Fatigue Damage Summation
As in the first example, the fatigue damage is assessed using the Miner-Palmgren
linear damage model, where the damage summation is in the following form:
n
D = ∑ di = ∑ i (D.2.13)
Ni
where D is the cumulative damage, di is the damage from ith stress cycle, Ni is the
average number of loading cycles to failure at the ith stress range and ni is the number of
cycles at each stress range in the period of interest. The fatigue damage accumulated in
one year at each stress range is presented in Table D.2.7.1. The summation of the fatigue
damage for the doubler and no doubler cases, indicate annual cumulative damages of
0.25 and 0.71, respectively.

Part D - Examples
By assuming that the vessel is at sea only 85% of the time, the fatigue life may be
calculated as 1 / (0.85*D) and are presented in Table D.2.7.2.
Table D.2.7.1: Sample Discrete Fatigue Damage Summation (Level 3)

(Doubler Case)
10 5.18E-01 2591235 4.34E+10 5.98E-05
30 2.73E-01 1366097 1.78E+08 7.66E-03
50 1.22E-01 611533.4 1.39E+07 4.41E-02
70 5.17E-02 258325.1 4.43E+06 5.83E-02
⇓ ⇓ ⇓ ⇓ ⇓
450 1.82E-10 0.000912 1.67E+04 5.47E-08
470 6.05E-11 0.000303 1.46E+04 2.07E-08
Sum 2.49E-01
(No Doubler Case)
10 4.18E-01 2089411 4.34E+10 4.82E-05
30 2.58E-01 1292383 1.78E+08 7.24E-03
50 1.47E-01 735191.6 1.39E+07 5.30E-02
70 8.14E-02 407146.2 4.43E+06 9.18E-02
⇓ ⇓ ⇓ ⇓ ⇓
450 2.99E-07 1.492549 1.67E+04 8.94E-05
470 1.50E-07 0.748971 1.46E+04 5.11E-05
Sum 7.15E-01
* All probable stress ranges are captured in 20 MPa stress range intervals (i.e. 0 to 20 mid value = 10, 20 to
40 mid value = 30, etc.)
‡
Probabilities for each stress range interval is estimated based on the difference between the probability of
exceedance of its upper and lower bounds using equation D.1.10 (i.e. p i =Q(∆σlower) - Q(∆σupper))
†
Assuming a 20 year design life with 108 wave encounters, the number of cycles per year for each stress
range is calculated as the product its probability and the number of wave events per year (i.e. 108 p i /
20).
¥
Cycles to failure are calculated using equation D.1.1 or using Figure D.1.2.2.
Table D.2.7.2: Summary of Fatigue Damage

Annual Fatigue Life
Damage, D (Years)
Doubler 0.25 4.7
No Doubler 0.71 1.7

Part D - Examples
Figure D.2.7.1 plots the annual fatigue damage for an individual stress range. This
representation illustrates the stress range levels that contribute most significantly to the
fatigue damage accumulation. While in general, larger stress ranges produce more
fatigue damage, they are not observed as often as lower stress range events.
0.12
Without Doubler
0.1
With Doubler
Damage per Year
0.08
0.06
0.04
0.02
0
10 30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350
Stress Range, ∆σ [MPa]
Figure D.2.7.2: Yearly Fatigue Damage

As can be seen by comparing Examples 1 and 2, the spectral analysis (Level 3)
should provide a much more representative estimate of fatigue life, since it is based on
the operational experience of the case vessel. Another point of interest, as a design case,
is that the introduction of a doubler plate improves fatigue life about 300%. Despite this
local fatigue life improvement, the finite element analysis indicates that the fillet weld at
the edge of the doubler plate, attaching it to the deck, is a potential fatigue crack initiation
site which should be investigated.
A final observation is that the designer must be conscious of the degree of error
associated with the fatigue life prediction. Thus, despite the convergence on a single
value for life in the process, the designer must consider the number of possible sources of
error in the process; in developing the operational profile, the spectral prediction of
bending moment, in the application of stress concentration factors, and in the
experimental data used to develop the damage estimates. Sensitivity analyses are
recommended, and the designer would be prudent to expect margins of error of at least
20%.

D.3 EXAMPLE 3: - SIDE PRESSURE CALCULATIONS
D.3.1 Objective
To conduct a comprehensive fatigue analysis, based on the spectral method (Level
3) for a stiffened panel including a side shell structural detail for a product tanker
operating on a route between the U.S. West Coast and Alaska. The fatigue life is to be
estimated for a bracket loaded by lateral side pressure at the intersection with transverse
bulkheads.
D.3.2 Approach
The basic stages in the analysis follow those defined in Example 2. However, the
loading on the structure was considered in two parts: first, quasi-static pressure due to
vertical displacement of the point of interest, and second, the dynamic pressure at the
point of interest due to a wave impinging on the vessel side shell. These components are
shown schematically in Figure D.3.2.1.
Figure D.3.2.1: Schematic of Side Shell Load Components
Several fatigue failures near the waterline and on the weather side have been
reported in the literature [e.g., Ref. D.6]. Most of these failures occurred next to a
bulkhead or web frame in way of an attachment bracket, and were associated with local
pressure acting on the ship’s side. This effect is essentially non-linear. However, in this
example, an approach based on linear theory will be employed to address the problem.

In this example only the outside pressure will be considered, i.e., only atmospheric
pressure is acting on the inner plating of the side shell. In terms of the example, this case
corresponds to the cargo leg of the voyage, where a ballast tank, containing the structure
of interest, is empty.
D.3.3 Operational Profile Data
D.3.3.1 Definition of the Operational Profile
Operational profile data used in the example was for a tanker operating on the West
Coast of North America between California and Alaska. The principal particular of the
example ship and operational profile data used in the example are given in Tables D.3.3.1
and D.3.3.2 respectively. The operational profile data is based on historical records.
Table D.3.3.1: Principal Data: Example Ship

Ship Length (Lpp) 187.15 m
Ship Beam 28.96 m
Draft (Loaded Condition) 12.4 m
Displacement (Loaded Condition) 58,800 tonnes
KG 12.2 m
Table D.3.3.2a: Joint Probability of Speed and NATO Sea State for Example Ship
SPEED NATO Sea State
(kn) 1 2 3 4 5 6 7
0-6 0.0028 0.0000 0.0009 0.0000 0.0000 0.0010 0.0007
6 -10 0.0056 0.0031 0.0033 0.0082 0.0086 0.0261 0.0010
10 -14 0.0129 0.0219 0.0503 0.0378 0.0449 0.0896 0.0019
14 -18 0.0900 0.1253 0.1322 0.1007 0.1338 0.0968 0.0005
0.9999 0.1113 0.1503 0.1866 0.1467 0.1873 0.2135 0.0041
Table D.3.3.2b: Joint Probability of Heading and NATO Sea State for Example Ship
HEADING NATO Sea State
1 2 3 4 5 6 7
Head Seas 0.0084 0.0114 0.0167 0.0128 0.0160 0.0221 0.0005
Strbd. Bow 0.0234 0.0326 0.0380 0.0295 0.0383 0.0375 0.0005
Strbd. Beam 0.0283 0.0375 0.0439 0.0355 0.0454 0.0497 0.0009
Strbd. Quartering 0.0421 0.0586 0.0729 0.0565 0.0724 0.0800 0.0013
Following 0.0098 0.0136 0.0164 0.0129 0.0165 0.0176 0.0003
1.0000 0.1120 0.1537 0.1880 0.1472 0.1886 0.2070 0.0035

Table D.3.3.2c: Relative Time Spent in Marsden Zones for Example Ship
Marsden Zone % of time, (µi)
6 26.00
7 18.00
13 11.00
14 20.00
22 25.00
D.3.3.2 Composite Scatter Diagram and Probabilities of Stationary Conditions
The procedure for calculations of wave composite scatter diagram for this
example was identical to one used in spectral approach of the Example #2. As before, the
source of wave statistic data used was Global Wave Statistics [Ref. D.3]. In the example
annual wave data for unidirectional seas was utilised.
The composite probability distribution of wave heights and periods for all
Marsden zones combined was calculated according to the following equation:
N′
f mc (H S ; TZ ) = ∑ µ i f mi (H S ; TZ ) (D.3.1)
i =1
where fmi (Hs; Tz) is the probability distribution of wave heights and periods in Marsden
Zones 6, 7, 13, 14, and 22, µi are given in Table D.3.3.2c, and N′ = 5 (the number of
Marsden Zones). The resulting Composite Scatter Diagram is shown in Table D.3.3.3.
Table D.3.3.3: Composite Marsden Zone Scatter Diagram

Tz
Hs[m] <4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 >13 Sum
0-1 0.0015 0.0153 0.0373 0.0362 0.0175 0.0046 0.0012 0.0000 0.0000 0.0000 0.0000 0.1135
1-2 0.0003 0.0070 0.0460 0.0948 0.0870 0.0441 0.0145 0.0035 0.0007 0.0000 0.0000 0.2979
2-3 0.0000 0.0017 0.0180 0.0589 0.0830 0.0626 0.0295 0.0096 0.0023 0.0005 0.0000 0.2662
3-4 0.0000 0.0003 0.0053 0.0237 0.0453 0.0452 0.0275 0.0114 0.0038 0.0008 0.0000 0.1632
4-5 0.0000 0.0000 0.0012 0.0082 0.0194 0.0243 0.0181 0.0090 0.0035 0.0008 0.0003 0.0848
5-6 0.0000 0.0000 0.0003 0.0028 0.0080 0.0110 0.0096 0.0059 0.0023 0.0008 0.0003 0.0409
6-7 0.0000 0.0000 0.0000 0.0012 0.0030 0.0050 0.0049 0.0032 0.0015 0.0003 0.0002 0.0192
7-8 0.0000 0.0000 0.0000 0.0003 0.0010 0.0021 0.0020 0.0016 0.0010 0.0003 0.0000 0.0083
8-9 0.0000 0.0000 0.0000 0.0000 0.0005 0.0010 0.0010 0.0011 0.0003 0.0002 0.0000 0.0040
9-10 0.0000 0.0000 0.0000 0.0000 0.0003 0.0004 0.0008 0.0003 0.0003 0.0000 0.0000 0.0020
10-11 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
11-12 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
12-13 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
13-14 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
>14 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Sum 0.0017 0.0243 0.1080 0.2261 0.2650 0.2003 0.1091 0.0456 0.0157 0.0036 0.0008 1.0000

The operational profile data and the Composite Marsden Zone scatter diagram are
combined to obtain the probabilities of stationary (operating) conditions.
The next consideration is that the wave environmental data and the vessel
operational profile data should be defined on the same basis. In this case, operational
profile data was ordered using the NATO Sea State scale, and environmental data was
organized by significant wave height. Thus, to calculate the stationary condition
probabilities it is necessary to convert the wave height probabilities from Table D.3.3.3 to
Sea State probabilities. Furthermore, for a comprehensive analysis this would be done
for all zero crossing periods in the Composite scatter diagram. For the purpose of this
example, the Composite scatter diagram was reduced into two wave height distributions.
This was accomplished by grouping wave height data for zero crossing period bins from
4< to 9-10 seconds together for the first distribution and wave height data for period bins
from 10-11 to 13> seconds for the second.
Grouping of the wave periods was done for two reasons: reduction of the number of
calculations; and an expectation of that there would be a dominant source of pressure
variation at POI associated with each distribution. This will be described in detail in
Section 6. Table D.3.3.4 shows these two distributions and the conversion to Sea States.
The procedure used for Sea State-wave height conversion is outlined in Section
D.2.2.3.2.
Table D.3.3.4: Transformation of Wave heights to Sea States for Combined Scatter Diagram
Include zero crossing period Tz bins Include zero crossing period Tz bins
from <4 to 9-10 seconds from10-11 to >13 seconds
Sea State Prob. Range of Prob Sea State Prob. Range of Prob
Hs (m) Hs (m)
1 0.0000 0-1 0.1135 1 0.0000 0-1 0.0000
2 0.1127 1-2 0.2936 2 0.0000 1-2 0.0043
3 0.2342 2-3 0.2538 3 0.0027 2-3 0.0124
4 0.3112 3-4 0.1473 4 0.0134 3-4 0.0159
5 0.1864 4-5 0.0712 5 0.0221 4-5 0.0136
6 0.0745 5-6 0.0317 6 0.0202 5-6 0.0092
7+ 0.0153 6-7 0.0141 7+ 0.0072 6-7 0.0052
Sum 0.9344 7-8 0.0053 Sum 0.0656 7-8 0.0030
8-9 0.0025 8-9 0.0015
9-10 0.0014 9-10 0.0006
10-11 0.0000 10-11 0.0000
11-12 0.0000 11-12 0.0000
12-13 0.0000 12-13 0.0000
13-14 0.0000 13-14 0.0000
>14 0.0000 >14 0.0000
Sum 0.9344 Sum 0.0656

The stationary conditions characterised by probabilities of speed and heading for
a given sea state are given in Tables D.3.4.1a and D.3.4.1b. The probabilities given in
the tables are normalised by a 1000. Thus, in Table D.3.3.5a, the probability of speed
between 14-18 knots and head seas at sea state 3 with waves having zero crossing period
from less than 4 up to 10 seconds is 14.697/1000=0.014697. The algorithm for the
calculations of probabilities was discussed in details in section D.2.2.3.4 and is not
repeated here. The theoretical number of stationary conditions is 280 (4 speeds x 5
headings x 7 sea states x2 periods), although in practice there are a number of zero
probabilities.
D.3.4 Side Shell Loading Calculations

The pressure force on the side shell of the vessel can be modelled as two
components:
(1) the quasi-static head that a point of interest on the side shell experiences due to
vertical motions of the vessel;
(2) the dynamic pressure due to waves impinging onto the side shell. This
phenomenon is referred to as a standing wave.
Each load component was introduced in section C.3.2.2 and are briefly discussed
here.

Table D.3.4.1a: Probabilities of all combinations of Speed, Heading and Sea State
for zero crossing periods Tz between <4 and 9-10 seconds in
Combined Marsden Zone
SPEED Sea State: 1 Sea State: 2
(kn) Head Stb. Stb. Stb. Head Stb. Stb. Stb.
Following Following
0_6 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6_10 0.000 0.000 0.000 0.000 0.000 0.172 0.491 0.564 0.882 0.205
10_14 0.000 0.000 0.000 0.000 0.000 1.221 3.485 4.003 6.266 1.455
14_18 0.000 0.000 0.000 0.000 0.000 6.985 19.940 22.905 35.848 8.322
Sum 0.0 0.0 0.0 0.0 0.0 8.4 23.9 27.5 43.0 10.0

Following Following
0_6 0.099 0.225 0.260 0.432 0.097 0.000 0.000 0.000 0.000 0.000
6_10 0.368 0.841 0.972 1.613 0.363 1.503 3.473 4.174 6.651 1.512
10_14 5.591 12.764 14.748 24.476 5.513 6.967 16.094 19.347 30.827 7.008
14_18 14.697 33.555 38.770 64.341 14.492 18.550 42.849 51.510 82.074 18.659
Sum 20.8 47.4 54.8 90.9 20.5 27.0 62.4 75.0 119.6 27.2

Following Following
0_6 0.000 0.000 0.000 0.000 0.000 0.036 0.061 0.081 0.131 0.029
6_10 0.729 1.745 2.073 3.305 0.754 0.974 1.649 2.190 3.524 0.776
10_14 3.783 9.055 10.755 17.146 3.910 3.343 5.662 7.519 12.100 2.665
14_18 11.280 27.001 32.070 51.130 11.659 3.611 6.116 8.122 13.071 2.878
Sum 15.8 37.8 44.9 71.6 16.3 8.0 13.5 17.9 28.8 6.3
SPEED Sea State: 7 + SPEED

(kn) Head Stb. Stb. Stb. (kn)
Following
0_6 0.412 0.370 0.723 0.994 0.200 0_6
6_10 0.550 0.493 0.964 1.325 0.266 6_10
10_14 1.099 0.986 1.929 2.651 0.533 10_14
14_18 0.275 0.246 0.482 0.663 0.133 14_18
Sum 2.3 2.1 4.1 5.6 1.1

Table D.3.4.1b: Probabilities of all combinations of Speed, Heading and Sea State
for zero crossing periods Tz between 10-11 and 14> seconds in
Combined Marsden Zone.
Following Following
0_6 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6_10 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
10_14 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.002 0.000
14_18 0.000 0.000 0.000 0.000 0.000 0.002 0.005 0.006 0.010 0.002
Sum 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Following Following
0_6 0.001 0.003 0.003 0.005 0.001 0.000 0.000 0.000 0.000 0.000
6_10 0.004 0.010 0.011 0.019 0.004 0.065 0.150 0.180 0.287 0.065
10_14 0.064 0.146 0.169 0.281 0.063 0.301 0.695 0.835 1.330 0.302
14_18 0.169 0.385 0.445 0.738 0.166 0.801 1.849 2.223 3.542 0.805
Sum 0.2 0.5 0.6 1.0 0.2 1.2 2.7 3.2 5.2 1.2

Following Following
0_6 0.000 0.000 0.000 0.000 0.000 0.010 0.017 0.022 0.035 0.008
6_10 0.087 0.207 0.246 0.393 0.090 0.263 0.446 0.592 0.953 0.210
10_14 0.449 1.075 1.277 2.036 0.464 0.904 1.531 2.034 3.273 0.721
14_18 1.340 3.207 3.809 6.072 1.385 0.977 1.654 2.197 3.535 0.779
SUM 1.9 4.5 5.3 8.5 1.9 2.2 3.6 4.8 7.8 1.7
SPEED Sea State: 7 + SPEED

(kn) Head Stb. Stb. Stb. (kn)
Following
0_6 0.193 0.173 0.339 0.466 0.094 0_6
6_10 0.258 0.231 0.452 0.621 0.125 6_10
10_14 0.515 0.462 0.904 1.242 0.250 10_14
14_18 0.129 0.115 0.226 0.311 0.062 14_18
SUM 1.1 1.0 1.9 2.6 0.5

D.3.4.1 Quasi-Static Pressure
The calculation of the quasi static pressure is given by (see Equation C.3.13)
? p s = ?g (z 2 − z1 ) (D.3.2)
In the subsequent analysis the difference in pressure calculated from (D.3.2) will
be referred to as “ quasi-static pressure”.
It is expected that the static head (z2 – z1 ) will be derived from a vertical motion
response prediction for the POI (combining local roll, heave, and pitch components) from
the spectral analysis of motion response. It is recommended that the single amplitude
response be used due to intermittent emergence of the point of interest, noting that when
the point of interest emerges from the water, the outside pressure falls to zero.
D.3.4.2 Pressure Due to Wave Impinging on the Side Shell
The pressure component due to wave action against the side of the vessel is
investigated by analyzing incident regular waves on a two-dimensional free-surface-
piercing body (i.e., the side shell). The approach assumes the following:
• The wave length of the incoming waves is small, such that the hull will not oscillate
due to the wave excitation. Thus the side shell is stationary.
• The side shell is vertical at the intersection with the free surface.
• The effect of the waves is only felt on the weather side of the ship. On the lee side of
the ship there is a “shadow” region where there is no wave action.
• The angle of incoming waves is perpendicular to the wall. This is true for beam seas
and can be extended over bow and quartering seas.
With the above assumptions, the problem of wave action striking the side of the
vessel reduces to analysis of an incident wave impinging on a vertical wall. The primary
characteristic of this approach is the “Standing Wave”, a wave that has twice the wave
height of the incoming waves on the upstream side of the vertical wall. Field
observations aboard ships confirm that this type of wave behaviour occurs on the weather
side of the ship [D.6].
Dynamic pressure pd is given by (see equation C.3.14) as:
1 
p d = 2 ?gH cos(kx)cos (? t )  (D.3.3)
2 

The linear solution for the “Standing Wave” pressure distribution can be written [D.7] as:
p total = −?gz + pd (D.3.4)
where z corresponds to the “stationary” elevation of POI from the still water line.
Note that defining H=1 effectively defines a transfer function (RAO) for dynamic
pressure.
The magnitude of total pressure at point of interest is considered for two cases:
(a) when the wave crest arrives at the side shell; and,
(b) when the wave trough arrives at the side shell.
In the subsequent analysis, this pressure difference will be reffered to as “dynamic

pressure”. The approach for calculation of dynamic pressure is not ship dependent. The
only information required is the vertical location (from the still waterline) of the point of
interest. However, the assumptions of small ship oscillation should be verified against
the predicted (or actual) ship motions.
D.3.4.3 Treatment of the Side Pressure Components
It is not generally expected that the phase difference between pressure at the POI
due to vertical motion (static pressure) and due to wave impingement (dynamic pressure)
will be known. Thus a “net pressure” analysis on the POI will not be possible. Thus, it
will be necessary to conduct separate fatigue damage assessments for each loading
component, each taken over the life of the vessel. This assumes that the two components
do not constructively combine to create greater pressure magnitudes, i.e. that they are
about 180 degrees out of phase. This is generally valid if roll is the dominant response
mode. If this is not the case, the predicted life may well be unconservative as two
distributions of lower intensity will generally result in less damage than a single
distribution of higher loadings, due to the general form of an S-N curve.

D.3.5 Development of RAO’s
In this section, metods for calculations of Response Amplitude Operators (RAO),
response functions and associated Root Mean Square (RMS) values for static and
dynamic pressure are described.
D.3.5.1 Quasi-static Pressure (Vertical Displacement of POI)
Response of the vessel in terms of vertical displacement of the POI was calculated
using the linear strip theory program Shipmo 7. The procedure was as follows:
• Calculate RAO’s. The Shipmo program produced RAO’s in regular seas for the six
degrees of freedom motions at the vessel’s CG. Figure D.3.5.1 shows an RAO of the
roll motion for the example ship.
• Specify Sea Spectrum. In this example the two- parameter (significant wave height
and peak wave period) Bretschneider spectrum was used.
• Define the location of the POI. In this example, the location was taken as:
• vertical location = 2.5 m below still waterline;
• longitudinal location = 30 m aft from AP, and
• 14.48 m to the starboard from centreline.
• Calculate RMS values. Shipmo 7 calculated RMS values for absolute displacements,
velocities and accelerations in vertical, transverse and longitudinal directions at the
location of the POI. Shipmo 7 also calculated RMS of the relative displacement of
the point of interest, where “relative displacement” is the distance from the POI to the
actual free surface. Associated with this value is also a probability of emergence (with
exceedence parameter of 0.01) of the POI and the rate of emergence expressed in
number of emergences per hour.
• Static pressure RMS at the POI due to vertical displacement is taken as:
N
RMS Quasi−Stat. press. = RMS Rel. displ.?g [ ] (D.3.5)
m2
The mean square (area under the vertical displacement response spectrum curve)
is the average of the square of the vertical displacement amplitude, and the root
mean square (RMS) is the square root of that value. Therefore, RMS Stat . press.
represents a single amplitude static pressure.

Roll RAO (Heading=90 deg, V=16
knots)
16
rad/wave slope
12
0
0.00 0.50 1.00 1.50 2.00 2.50
Encounter frequency (rad/sec)
Figure D.3.5.1: Roll RAO against Encounter Frequency for Beam Seas and Speed
of 16 knots
D.3.5.2 Dynamic Pressure (wave action against the side of the vessel)
For the dynamic pressure case RAO’s, the response functions and associated
RMS values were calculated using a spreadsheet. Figure D.3.5.2 shows calculated
dynamic pressure RAO for beam seas. Note that the RAO units have dimension of N/m3
(pressure per unit wave height).
Dynamic Pressure RAO (Heading=90 deg.,

V=16 knots)
25000
20000
N/m^3
15000
10000
5000
0
0.00 0.50 1.00 1.50 2.00 2.50
Encounter Frequency (rad/sec)
Figure D.3.5.2: Dynamic Pressure RAO against Encounter Frequency for

Beam Seas

The Bretschneider sea spectrum was corrected to the encounter frequency by (Eq. C.3.2):
S(ω )
S(? e ) = (D.3.6)
2? V
1− g cos(θ )
and the response function was calculated from:
R (? e ) = RAO (? e )2 ⋅ S(? e ) (D.3.7)
where R(ωe) is the response function. Phase information is retained because dynamic
pressure is in phase with the wave elevation. In other words, a wave crest accompanies
maximum dynamic pressure, and a wave trough accompanies minimum dynamic
pressure.
Response statistics are calculated as follows:
• Zeroth moment:
∞ N
m 0 = ∫ R (? e ) d? e ≅ ∑ R i (? ei ) ⋅ ? ? ei (D.3.8)
0 i =1
• Second moment:
∞ N
m2 = ∫? R(? e ) d? e ≅ ∑ ? ei R i (? ei ) ⋅ ? ? ei
2 2
e (D.3.9)
0 i =1
• Root mean Square:
RMS Dyn.press. = m 0 (D.3.10)
• Response Zero Crossing Period:

m0
Tz = 2 p (D.3.11)
m2
D.3.6 Calculations of Responses

In Section D.3.3, the composite scatter diagram (Table D.3.3.3) was divided into
two marginal distributions. However, the response calculations require more detailed
wave period information. For a comprehensive analysis this would not be an issue as the
response calculations would be carried out for every period in the scatter diagram. In this
example, however, the response calculations were performed for the two aforementioned
marginal distributions and the periods for each were calculated by selecting a period
which would give the same order of magnitude of number of cycles as if the calculations
were performed for every wave zero crossing period. Table D.3.6.1 summarizes these
results.

Table D.3.6.1: Calculation of Lifetime Average Zero Crossing Period
Tz (sec.) Number of Cycles Tz (sec.) Number of Cycles
<4 308,257 10.5 2,737,660
4.5 3,401,445 11.5 860,267
5.5 12,384,466 12.5 181,205
6.5 21,940,064 >13 35,519
7.5 22,281,287 Sum: 3,814,652
8.5 14,862,620 Tz (average) = 10.8 3,831,040
9.5 7,240,486
⋅24⋅ 3600
N = prob ⋅ 20( years)T365
z
Sum: 82,418,625
Tz (average) = 7.2 81,853,440
The selected wave zero crossing periods of 7.2 and 10.8 seconds then had to be
converted for use with the Bretschneider spectrum, which requires significant wave
height and wave peak period. Conversion of zero crossing period to peak period is
expressed as [D.7]:
Tp = 1.408 Tz (D.3.12)
Using this expression, Tp is 10.14 and 15.2 seconds. Converting to circular
frequency this correspond to 0.62 rad/sec and 0.413 rad/sec. From Figure D.3.5.1, it can
be seen that resonance frequency for roll is around 0.35 rad/sec. Thus, the peak period of
the calculated wave average zero crossing is very close to the roll resonance frequency.
Thus it can be said that marginal wave height distribution which includes zero
crossing periods from 10-11 to >13 seconds (see Table D.3.3.4) was dominated by the
roll response, and therefore the static pressure component is significant. Conversely, the
circular frequency of 0.62 rad/sec is sufficiently far from the resonant roll frequency
(Figure D.3.5.1) such that the marginal wave height distribution from <4 up to 9-10
seconds will be dominated by the dynamic pressure component. This is consistent with
the assumption made in formulating dynamic pressure; that is the ship is essentially
vertically stationary.
Calculation of responses and response statistics were done for the array of 175
stationary conditions (specified by probability of occurrence of Sea State, heading and
speed) identified in Tables D.3.4.1a and D.3.4.1b. A sample of the calculations is given
in Table D.3.6.2. In Table D.3.6.2, Columns 1 to 3 contain the stationary condition
information, and column 4 is the normalised probability of the stationary conditions.
Columns 5 to 8 are Shipmo 7 output pertinent to the POI, and columns 9 and 10 are
dynamic pressure calculations.
It can be seen from the Table that the number of emergences and associated probabilities
increases with increasing Sea State. Also, the dynamic pressure is calculated for beam
quartering and bow seas. This is probably a conservative estimate due to the assumption
that the angle of incoming waves is perpendicular to the wall.

Table D.3.6.2: Sample of Shipmo7 Output and Dynamic Pressure Calculations
Sea State 2 Sig. WH = 0.3
Rel. Vert. Displacement Emerg. Dynamic pressure
Speed Heading Tp prob. Tz* RMS Rate/hr. prob. Tz RMS
8 Head 10.14 0.172 9.40 0.070 0.0 0.0000
8 Stbd.Bow 10.14 0.491 9.10 0.071 0.0 0.0000 6.07 1128.0
8 Stbd. Bm 10.14 0.564 9.80 0.076 0.0 0.0000 7.71 1128.0
8 Stbd.Quat. 10.14 0.882 12.70 0.045 0.0 0.0000 10.34 1128.0
8 Follow. 10.14 0.205 14.40 0.063 0.0 0.0000
Sea State 3 Sig. WH.= 0.88
Speed Heading Tp prob. Tz RMS Rate/hr. prob. Tz RMS
8 Head 10.14 0.368 9.40 0.205 0.0 0.0000
8 Stbd. Bow 10.14 0.841 9.10 0.208 0.0 0.0000 6.07 3307.0
8 Stbd. Bm 10.14 0.972 9.80 0.222 0.0 0.0000 7.71 3307.0
8 Stbd.Quat. 10.14 1.613 12.70 0.132 0.0 0.0000 10.34 3307.0
8 Follow. 10.14 0.363 14.40 0.185 0.0 0.0000
Sea State 5 Sig. WH.= 3.25
12 Head 10.14 3.783 8.80 0.809 26.7 0.0344
12 Stbd. Bow 10.14 9.055 8.80 0.875 37.1 0.0561 5.49 12213.0
12 Stbd. Bm 10.14 10.755 9.80 0.826 19.7 0.0396 7.71 12213.0
12 Stbd.Quat. 10.14 17.146 14.70 0.560 0.3 0.0009 12.70 12213.0
12 Follow. 10.14 3.91 17.70 0.717 2.9 0.0137
Sea State 6 Sig. WH.= 5
16 Head 10.14 3.611 8.50 1.338 241.5 0.2918
16 Stbd. Bow 10.14 6.116 8.50 1.475 248.6 0.3628 5.00 18790.0
16 Stbd. Bm 10.14 8.122 9.70 1.271 127.3 0.2555 7.71 18790.0
16 Stbd.Quat. 10.14 13.071 16.90 0.888 13.3 0.0612 15.68 18790.0
16 Follow. 10.14 2.878 21.80 1.130 28.1 0.1779
12 Head 15.2 0.904 10.50 0.844 29.1 0.0454
12 Stbd. Bow 15.2 1.531 10.30 0.966 50.7 0.0943 6.07 11400.0
12 Stbd. Bm 15.2 2.034 11.90 1.235 76.3 0.2355 8.33 11400.0
12 Stbd.Quat. 15.2 3.273 16.90 0.658 1.5 0.0061 13.01 11400.0
12 Follow. 15.2 0.721 20.40 0.692 2.1 0.0100

D.3.7 Calculation of Stress Ranges
A sketch of the structural detail for which the stress ranges due to outside lateral
pressure was calculated is given in Figure D.3.7.1. For a bracket termination on top of
stiffener, the stresses to be considered related to lateral pressure are due to:
• Local stiffener bending;

• Relative deflection between web frame and transverse bulkhead;
• Panel bending - single hull vessel.
All three components will be considered in this example. Theoretical background

and appropriate equations have been introduced in Section C.
Transverse
bulkhead
Figure D.3.7.1: Description of Structural Detail

D.3.7.1 Stiffener Bending
Stresses due to local stiffener bending can be calculated from (see the first part on
the right hand side of the equation C.4.8):
M
s 2 =K (D.3.13)
ZS
where:
K = Stress concentration factor
The hot spot stress is related to the nominal stress using a factor K as:
K = Kg Kn
and where factors are defined as following:
• Term Kg is taken as per stiffener support K-factors [D8]. For the case of stiffener
bending due to lateral load, Kg is taken as 1.8 1 . An additional factor of 1.15 is
used to take into account bracket/stiffener overlap (see figure D.3.7.1).
• Kn represents additional stress concentration factor due non-symmetrical stiffener

and is calculated as per [D.8, Section 7]:
1 + ?ß 2
Kn =
1 + ?ß 2 ?
where:
µ
3(1 + ) 4
280 le
?= µ=
µ 4h s
1+ 3
bf t f h 2[ + ]
40 t w 3 t p3
tw (h − t f ) 2 t w
ß = 1− ? =
bf 4Z S
1
. Ref. D.8 for the case considered recommended K factor KgK w = 2.7. Here Kw is
related to the local geometry of the notch. Since in this Guide notch stress concentrations
are not considered in fatigue calculations, Kw contribution needed to be extracted. Kw
stress concentration factor equals 1.5 if not stated otherwise [D.8], therefore Kg alone
equals 1.8.

M Moment at stiffener support adjusted to hot spot position at the stiffener
(e.g. bracket toe). It is calculated as (see Equation C.4.8):
psl e 2
M= rp
12
p is lateral dynamic pressure

s stiffener spacing
le effective span of stiffener
rπ Moment interpolation factor for interpolation to hot spot position along the
stiffener length (see figure C.4.3.3)
2
x  x 
r p = 6  − 6  + 1.0
le le
ZS Section modulus of stiffener with associated effective plate flange.
The properties of the stiffener containing the POI are listed in Table D.3.7.1.
Table D.3.7.1: Geometry of the Stiffener Considered

Stiffener spacing: s = 914.4 mm
Web frame spacing: l = 4267 mm
Effective span of stiffener: le = 4064 mm
Location (POI): Panel CG @ 2500 mm below still waterline
Longitudinal location @ 30 m aft from FP
Stiffener moment of inertia Trans. Axis I= 4.23x108 mm4
Sectional Modulus ZS: 1.44x106 mm3 (flange top)
2.33x106 mm3 (baseline)
x - position: x = 406.4 mm
Plate thickness: tp = 17.4 mm
Height of stiffener: h = 441.3 mm
Web thickness: tw = 16.5 mm
Flange width; bf = 101.6 mm
Flange thickness: tf = 15.9 mm
Bracket thickness: tb = 15.9 mm

Calculations:
406.4 2 406.4
r p = 6( ) − 6( ) + 1.0 = 0.46
4064 4064
K = K g ⋅1.15 ⋅ K n
Using the data from Table D.3.7.1 Kn term is calculated as follows:
µ = 148.9
? = 0.9728
ß = 0.8376
? = 0.5184 @ ZS = 1.44x10 6
1 + 0.9728 ⋅ 0.8376 2
Kn = = 1.243
1 + 0.9728 ⋅ 0.8376 2 ⋅ 0.5184
K = K g ⋅1.15 ⋅ K n = 1.8 ⋅ 1.15 ⋅1.243 = 2.573
Stresses due to stiffener bending per unit lateral dynamic pressure (1 kN/m2 ) is
(Equation D.3.13):
1
2 1⋅10 3 ⋅ 6 ⋅ 914.4 ⋅ 4064 2
ps l e 10  N 
s 2 =K r p = 2.573 ⋅ 0.46 = 1.034  
12Z S 12 ⋅1.44 ⋅10 6
 mm 2 
D.3.7.2 Stress due to Relative Deflection between Web Frame and Transverse
Bulkhead
Stresses due to the relative deflection between the transverse bulkhead and the
web frame can be approximated according to the following formula (see the second part
on the right hand side of the equation C.4.8):
m d EI
sd =K rd d (D.3.14)
l e 2 ZS

K = K g ⋅ 1 ⋅ 15
Here Kg is taken as 1.467 [Reference D8] recommends Kg Kw as 2.2 for stress in
longitudinal direction, and K w=1.5). As before factor of 1.15 is taken to account for
the bracket/stiffener overlap.
mδ Moment factor due to relative deflection between transverse

supports (taken as 4.4 – no stringers supporting the frames [D.8]).
rδ Moment interpolation factor for interpolation to hot spot position
along the stiffener length (see equation C.4.8).
x
rδ = 1 − 2( )
le
δ Deformation of the nearest frame relative to the transverse
bulkhead. According to Reference D.8, δ can be approximated as
follows:
2z 2
d = (1 − (1 − ) )d m (mm)
D
and
110l s 3 D m 2
dm = p (no side stringers)
EI b 1 + N s
Here:
Ns Number of cross ties in the web frame

ls Span between bulkhead and transverse frame
Dm Depth of ship in meters.
z Vertical distance from baseline to considered longitudinal.
Data for these values can be found in Table D.3.7.2.

Table D.3.7.2: Web Frame and Panel Geometry
Longitudinals:
Stiffener spacing: sa = 914.4 mm
Stiffener moment of inertia Ia= Ina = 4.23x108 mm4
Smeared out stiffness ia = 4.23x108 /914.4 = 4.626x105 mm3
Moment of inertia of effective breadth w.r.t. N.A: Ipa=1.51x108 mm4
Effective span of stiffener: le = 4064 mm
x - position: x = 406.4 mm
Length of panel a=29870 mm
Transverses:
Web frame spacing: sb = 4267 mm
Web moment of inertia: Ib = Inb = 5.21x1010 mm4
Moment of inertia of effective breadth w.r.t N.A: Ipb =1.68x1010 mm4
Smeared out stiffness ib = 5.21x1010 /4267 = 122.10x105 mm3
Distance between bulkhead and transverse web: ls = 4267 mm
Width of panel b = 12495 mm
Vessel depth Dm =18.59 m

Number of cross ties Ns =0
z = 9.9 m
Calculations:
Using the above equations and tabulated data, together with unit lateral dynamic pressure
of 1kN/m2 the following results are obtained:
δm = 2.699x10-4 mm
δ = 2.687x10-4 mm
rδ = 0.8
Stresses due to relative deflection between the transverse bulkhead and the web frame are
(equation D.3.14):
m d EI 4.4 ⋅ 2.1 ⋅10 5 ⋅ 4.23 ⋅10 8 N

sd =K r d d = 1.467 ⋅ 1.15 0.8 ⋅ 2.687 ⋅ 10 −4 = 5.960 ⋅10 −3 ( )
2
l e ZS 4064 2 ⋅1.44 ⋅10 6 mm 2

D.3.7.3 Panel Bending – Single Hull Vessel
The equation for secondary bending stress in side shell panels at the intersection with
transverse bulkheads (web frames) (see Equation C.4.4) is:
• Longitudinal Secondary Bending Stress:

K p b 2 ra
s 2L = K b (panel longer than wide) (D.3.15)
iai b
Here p is dynamic lateral pressure. In the calculation it will be taken as a unit pressure of
1x103 N/m2 .
The stress concentration factor K is:
K = K g ⋅1.15
where Kg is taken as 1.467, and 1.15 to account for bracket/stiffener overlap.
The term Kb in Equation (D.3.15) is function of the aspect ratio ρ, panel’s boundary
conditions and torsion coefficient η (see Table C.4.3.2).
a i b 29870 4 122.10x10 5
? = ⋅4 = ⋅ = 2.52
b i a 12495 4.626x10 5
I pa I pb 1.51x10 8 ⋅1.68x10 10
?= = = 0.34
I na I nb 4.23x10 8 ⋅ 5.21x10 10
here ia, and ib are smeared out stiffnesses calculated in accordance with section C, Table
C.4.3.3, Type A panel. Terms Ipa and Ipb represent moment of inertia of effective breadth
of plating alone associated with longitudinal and transverse, respectively (see Table
D.3.7.2).
Table D.3.7.3 shows the values of the Kb term used in the calculations.
Table D.3.7.3: Kb values (Taken from Table C.4.3.2)

Boundary Conditions ρ η = 0.0 η = 0.5
Long edges simply 2.5 0.1263 0.1257
supported
Short edges clamped 3.0 0.1248 0.1253
Kb @ ρ = 2.52 and Kb =0.1259
η= 0.34 (calculated)

In the equation D.3.15, b term is the effective depth of the panel bounded by
double bottom and deck, and ra term is the distance from point considered to the neutral
axis of the panel in longitudinal direction (a – stiffener direction). This value is calculated
as 292.9 mm (see figure D.3.7.1) and is distance from neutral axis to the outer fiber of the
stiffener flange.
From equation (D.3.15) longitudinal secondary bending stress:
1
0.1259 ⋅1 ⋅10 3 ⋅ ⋅12495 2 ⋅ 292.92
6 N
s 2L = 1.467 ⋅1.15 ⋅ 10 = 4.087 ( )
4.626 ⋅10 ⋅ 122.10 ⋅10
5 5 mm 2
D.3.7.4 Combination of Stress Components
The above calculated stresses due to lateral pressure are combined considering the
sign – positive (tension), and negative (compression).
• Stress due to stiffener bending ( σ2 = 1.034 N/mm2 ). For an external dynamic lateral
pressure load, these will be in compression at the considered point (i.e., bracket
termination) [D8].
• Stress due to relative deflection between web frame and transverse bulkhead ( σδ=
5.960x10-3 N/mm2 ). Bending of the stiffener is resulting in compression at the
considered point for external pressure loads [D.8].
• Plate/panel bending ( σ2L= 4.087 N/mm2 ). Bending of the stiffener results in tension
at the considered point of the stiffener [D.8].
Therefore, for a unit external lateral pressure of 1kN/m2 exerted uniformly on the side
shell panel, the equivalent stress at the intersections with transverse web (in a way of
tripping brackets) is
N
− 1.034 − 5.960 ⋅10 −3 + 4.087 = 3.047 ( )
mm 2
D.3.7.5 Calculation of Stress Ranges
This “transfer function” of 3047 kN/m2 will be applied to both quasi-static

(Equation C.3.13), and dynamic pressure (equation C.3.15). The stress range response of
the structural detail was assumed to be Rayleigh distributed within each stationary
condition. The stress range distribution (defined in terms of stress range bins) for a given
stationary condition is (see Section C.5.3):

 − s upper   
2
−s
2
( ) ( lower )
 2m o   2m 0 
Q' (s mid ) short = 1 − e  − 1 − e  = F? s (s upper ) − F? s (s lower )
(D.3.16)
   
 
where σmid, σupper and σlower are stress range bin’s mid, upper and lower value.
Q’(σmid)short represents stress range bins probability distribution for a stationary condition.
The m0 is the stress response zeroth moment obtained from:
• Quasi-static Pressure:
1
m 0 = RMS 2 Quasi−Stat.press. ⋅ ⋅ 3047
10 3
• Dynamic Pressure:
1
m 0 = RMS 2 Dyn.press. ⋅ ⋅ 3047
10 3
RMSQuasi-Stat. Press is calculated from equation (D.3.5) where RMSVert. Displ is tabulated
(similar to the Table D.3.6.2, column 6) and RMSDyn. Press. is also tabulated (similar to
Table D.3.6.2, column 10). Division by 103 was necessary to bring the dimensions of
RMS to kN/m2 . Calculations using equation (D.3.18) are repeated for all stationary
operational conditions. The end result is a m x n matrix with m rows, indicating number
of stationary conditions, and n columns indicating the number of stress ranges (bins). In
this example, there was a 175x25 matrix for the static pressure case and a 175x30 matrix
for the dynamic pressure case.
The stress range response of the structural detail was assumed to be Rayleigh
distributed as the excitation due to wave action is Rayleigh distributed. This is a
significant approximation. In reality, the distribution of pressure peaks at POI will not be
Rayleigh distributed, as the POI randomly emerges from the water and pressure drops to
zero, while vessel motion continues. Thus the vertical motion response used to derive the
(static) pressure will be Rayleigh distributed, but in fact the resulting pressure variation
will not.
To realistically approximate the stress range distribution, the information

regarding probability of POI emergence for each stationary condition (Table D.3.6.2,
column 8) was employed. If the probability of emergence for a stationary condition was
greater than zero, then the Rayleigh distribution for that stationary condition was
“normalized” such that the “area” under the distribution (probability sum of all stress
range bins) was equal to:
n
Area = ∑ Q' (s mid ) short − prob emergence (D.3.17)
1

where n is 25 for the static pressure and 30 for the dynamic pressure cases. The
magnitude of probemergence for that stationary condition was then added to the normalised
zero stress range bin. The normalised stress range (bins other than zero) distribution for a
given stationary condition then becomes:
n
∑ Q' (s mid )short − prob emergence
[Q' (s mid ) Short ] normalized = 1
⋅ Q' (s mid ) short (D.3.18)
n
∑ Q' (s mid ) short
1
For the individual stress ranges, the long-term stress probability distribution is found
from (see equation C.5.6):
#o of stationary cond.
∑
1
Q' (s mid ) long = prob i ⋅ ⋅ [Q'i (s mid ) short ] normalized ⋅ ri (D.3.19)
i =1 1000
where probi is a probability of a stationary condition (see Table D.3.6.2, column 4) and
recall that the probability was normalized by a 1000, and the number of stationary
conditions was 175. Term ri is the ratio of zero crossing rate (frequency) to average zero
crossing rate:
f Zi
ri = .
−
fZ
Zero crossing periods Tzi are tabulated (similar to Table D.3.6.2, column 5 for quasi-
static pressure and column 9 for dynamic pressure), and are related to frequency fzi
through
1
f Zi =
TZi
The average zero-crossing frequency is calculated from:
− # of sationatycond.
∑
1
fZ= prob ⋅ ⋅f
i =1 1000 Z i
Again, the number of stationary conditions was 175.
Quasi-static and dynamic long term stress range distribution are shown in Figure
D.3.7.2a and D.3.7.2b, and mid value of stress ranges (σmid) and associated probabilities
(Q’(σmid)long) are given in Table D.3.7.4. Note that the stress range bins are not the same
for the two cases.

Table D.3.7.4: Stress Ranges and Corresponding Probabilities
Dynamic Pressure Quasi-Static Pressure
Stress Range Probability Stress Range Probability of
(MPa) of (MPa) Occurrence
σmid Occurrence σmid Q’(σmid)long
Q’(σmid)long
3.5 0.2107 3 0.2721
10.5 0.1721 9 0.2116
17.5 0.1470 15 0.1537
24.5 0.1108 21 0.1073
31.5 0.0853 27 0.0761
38.5 0.0669 33 0.0534
45.5 0.0511 39 0.0369
52.5 0.0380 45 0.0257
59.5 0.0281 51 0.0181
66.5 0.0211 57 0.0129
73.5 0.0160 63 0.0092
80.5 0.0123 69 0.0066
87.5 0.0095 75 0.0047
94.5 0.0072 81 0.0033
101.5 0.0055 87 0.0024
108.5 0.0042 93 0.0017
115.5 0.0032 99 0.0012
122.5 0.0024 105 0.0009
129.5 0.0019 111 0.0006
136.5 0.0014 117 0.0005
143.5 0.0011 123 0.0003
150.5 0.0009 129 0.0002
157.5 0.0007 135 0.0002
164.5 0.0005 141 0.0001
171.5 0.0004 147 0.0001
178.5 0.0003 Sum: 0.9995
185.5 0.0002
192.5 0.0002
199.5 0.0001
206.5 0.0001
Sum: 0.9993

Stress range distribution (quasi static pressure)
0.3
pr 0.25
ob
ab
ilit 0.2
y
de
0.15
ns
ity
fu 0.1
nc
tio
n 0.05
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Bin #
Figure D.3.7.2a: Stress Range Distribution Due to Quasi-Static Pressure
Stress range distribution (dynamic pressure)
0.25
0.2
probability density function
0.15
0.1
0.05
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Bin #
Figure D.3.7.2b: Stress Range Distribution Due to Dynamic Pressure

D.3.7.6 Damage Summation
Accumulation of fatigue damage is assumed to be described by a linear damage

accumulation rule (Palmgren-Miner) (see Equation C.6.7):
ni
D≅∑ (D.3.20)
Ni
where D is the total damage, Ni is the number of cycles to failure for a particular stress
range, and ni is the number of cycles at that stress range. The parameter Ni was
calculated from S-N curve data using:
log(N) = log(a) + mlog( ? l ) (D.3.21)
where the parameters are defined in Table D.3.7.5. Parameters include corrosion fatigue,
steel with cathodic protection and protective coating in sea water (see Equation C.6.4a
and b).
Table D.3.7.5: Parameters used in Example

N Log (a) m
(units in MPa) (units in MPa)
N<106 11.784 -3
6
N>10 15.637 -5
The total number of cycles ship experiences per year (N) is estimated according to:
prob i
# of stationary cond. 1 ⋅ num. of days at sea ⋅ ⋅ 24 ⋅ 60 2
N= ∑ Tz i
1000 (D.3.22)
i =1
where the number of days at sea is given on an annual basis, and probi is probability of
individual stationary conditions (Table D.3.6.2, column 4). Term Tzi refers to the wave
zero crossing period. In this example ship spends 237 days at sea per year. Thus the total
number of cycles for all 175 stationary conditions was calculated to be 2.29x106
(dynamic pressure), and 2.78x106 (static pressure).
Number of cycles at stress range ni is then estimated as:
n i = F? (s mid ) long ⋅N (D.3.23)
Table D.3.7.6 shows the calculations of number of cycles to failure Ni for dynamic and
static pressure case using equation (D.3.21) and parameters from Table D.3.7.5. Also,
Table D.3.7.6 contains the number of cycles at stress range ni and the relative damage
accumulation ni /Ni.

Table D.3.7.6: Stress Ranges and Corresponding Fatigue Damage
Dynamic Pressure Quasi-static pressure
σmid ni Ni ni/Ni σmid ni Ni ni/Ni
MPa Mpa
3.5 4.83E+05 8.25E+12 5.85E-08 3 7.56E+05 1.78E+13 4.24E-08
10.5 3.95E+05 3.40E+10 1.16E-05 9 5.88E+05 7.34E+10 8.01E-06
17.5 3.37E+05 2.64E+09 1.28E-04 15 4.27E+05 5.71E+09 7.48E-05
24.5 2.54E+05 4.91E+08 5.17E-04 21 2.98E+05 1.06E+09 2.81E-04
31.5 1.95E+05 1.40E+08 1.40E-03 27 2.12E+05 3.02E+08 7.00E-04
38.5 1.53E+05 5.13E+07 2.99E-03 33 1.48E+05 1.11E+08 1.34E-03
45.5 1.17E+05 2.22E+07 5.27E-03 39 1.03E+05 4.80E+07 2.14E-03
52.5 8.70E+04 1.09E+07 8.01E-03 45 7.13E+04 2.35E+07 3.04E-03
59.5 6.44E+04 2.89E+06 2.23E-02 51 5.02E+04 1.26E+07 4.00E-03
66.5 4.83E+04 2.07E+06 2.34E-02 57 3.58E+04 7.20E+06 4.97E-03
73.5 3.68E+04 1.53E+06 2.40E-02 63 2.56E+04 4.37E+06 5.85E-03
80.5 2.82E+04 1.17E+06 2.42E-02 69 1.82E+04 2.77E+06 6.58E-03
87.5 2.17E+04 9.08E+05 2.39E-02 75 1.30E+04 1.83E+06 7.09E-03
94.5 1.66E+04 7.21E+05 2.30E-02 81 9.21E+03 1.24E+06 7.40E-03
101.5 1.26E+04 5.82E+05 2.17E-02 87 6.55E+03 8.70E+05 7.53E-03
108.5 9.60E+03 4.76E+05 2.02E-02 93 4.68E+03 6.23E+05 7.50E-03
115.5 7.31E+03 3.95E+05 1.85E-02 99 3.35E+03 4.56E+05 7.35E-03
122.5 5.59E+03 3.31E+05 1.69E-02 105 2.41E+03 3.40E+05 7.10E-03
129.5 4.29E+03 2.80E+05 1.53E-02 111 1.74E+03 2.57E+05 6.75E-03
136.5 3.31E+03 2.39E+05 1.39E-02 117 1.25E+03 1.98E+05 6.34E-03
143.5 2.57E+03 2.06E+05 1.25E-02 123 9.04E+02 1.54E+05 5.87E-03
150.5 1.99E+03 1.78E+05 1.12E-02 129 6.52E+02 1.21E+05 5.37E-03
157.5 1.55E+03 1.56E+05 9.95E-03 135 4.70E+02 9.67E+04 4.86E-03
164.5 1.20E+03 1.37E+05 8.80E-03 141 3.39E+02 7.78E+04 4.36E-03
171.5 9.33E+02 1.21E+05 7.74E-03 147 2.45E+02 6.32E+04 3.88E-03
178.5 7.24E+02 1.07E+05 6.77E-03 Sum: 1.10E-01
185.5 5.62E+02 9.53E+04 5.90E-03
192.5 4.37E+02 8.53E+04 5.13E-03
199.5 3.40E+02 7.66E+04 4.44E-03
206.5 2.66E+02 6.91E+04 3.85E-03
Sum: 3.42E-01
Total Number of cycles (N)=2.29x106 Total Number of cycles (N)=2.78x106
From Table D.3.7.6 the linear cumulative damage summation given by equation (D.3.22)
for two cases is:
• Dynamic Pressure D = 0.3418 (1/D = 2.93 years ) N = 2.29x106

• Static Pressure D = 0.1102 (1/D = 9.07 years) N = 2.78x106

Fatigue life due to dynamic pressure is lower than the fatigue life due to static pressure.
However, they are of the same order of magnitude, and thus an indication that both
components need to be considered in the fatigue analysis.
D.4 REFERENCES
[D.1] “Rules for Building and Classing Steel Vessels”, American Bureau of Shipping
(ABS), 1996. Other Class Societies rules are harmonised.
[D.2] Glen, I., F., Paterson, R., B., Luznik, L. (1998). “Sea Operational Profiles for
Structural Reliability Assessment”, Ship Structure Committee (SSC) Report.
[D.3] Hogben, N., Dacunha, N., Olliver, G. (1986), “Global Wave Statistics”, British
Maritime Technology Ltd., (BMT), Middlesex.
[D.4] McTaggart, K. (1997), “SHIPMO 7 An Updated Strip Theory Program for
Predicting Ship Motions and Sea Loads in Waves”, DREA Technical
Memorandum 96/243.
[D.5] ITTC Seakeeping Committee Report (1978), 15th International Towing Tank
Conference, The Hague, Vol.1, pp.55-114
[D.6] Witmer, J., D., Lewis, W., J. (1994), “Operational and Scientific Hull Structure
Monitoring on TAPS Trade Tankers”, SNAME Transactions, Vol. 102, pp. 501-
533.
[D.7] Sarpkaya, T., Isaacson, M. (1981), Mechanics and Wave Forces on Offshore
Structures, Van Nostrand Reinhold Company, New York.
[D.8] “Fatigue Assessment of Ship Structures”, Classification Notes No. 30.7, Det
Norske Veritas (DNV), September 1998.
[D.9] Stacey, A., Sharp, J., V. (1995). “The Revised HSE Fatigue Guidance”, OMAE,
Volume III, Materials Engineering.

Appendix A – Comparison of Current Classification FTL-4714
Society Approaches to Fatigue Design
APPENDIX A - COMPARISON OF CURRENT CLASSIFICATION SOCIETY

APPROACHES TO FATIGUE DESIGN
A. 1 Introduction
One of the first steps in the design process is to determine which key challenges the
design will have to overcome. This will dictate the nature of the design approach and the
processes used to address each element of it. Often, fatigue is not a major issue. Many small,
robust ships, such as tugs and supply vessels, do not have the levels of cyclic stresses which will
lead to fatigue problems. In other ships, fatigue may be a localized concern for specific
equipment foundations or for other structural components such as shaft brackets. Fatigue in
machinery systems or propulsors can also affect ships where structural fatigue is highly unlikely.
All types of localized fatigue can be handled using the same general approaches used in
this course. However, they will not be given any more detailed consideration. The focus of this
lecture, and those which follow, will be on general ship structures where fatigue performance
needs to be validated in order to assure an owner of acceptable through-life capability. This is
likely to include all major warships, and larger commercial vessels including tankers, bulk
carriers, and container ships.
The level of performance validation required will vary, depending on the ship's
configuration and intended service. Accordingly, more or less sophisticated methods of fatigue
life assessment may be utilized. Where it is anticipated that highly detailed analyses will be
required at some stage in the design process, it is still probable (and highly advisable) that less
complex methods be used in the earlier stages, when the overall design parameters are being
selected. Detailed analyses should only lead to changes in details, rather than major reworkings
of the ship design as a whole.
The designer should therefore have an understanding of the probable extent of a ship's
fatigue concerns, the level to which they can be expected to influence other design
considerations, and the methods which can be used to manage fatigue design most cost-
effectively at each stage in the process. These methods include a variety of design codes,
standards, and criteria. A number of the most commonly used of these are described in this
section.
A.2 Overview
Many of the existing ship fatigue analysis codes are based on the same knowledge base,
and reference (directly or in supporting documents) other non-marine or offshore standards.
These other applications generated earlier interest in fatigue either because fatigue failures were
more likely to prove immediately catastrophic, or because their economic consequences were
more dramatic than was the case with traditional ship construction. Organizations such as the
International Institute for Welding (IIW), the European Committee for Standardization, and a
variety of national and industry standards bodies produced materials relevant in whole or part to
ship structural design. The ship classification societies, such as ABS, DNV, and Lloyds, have
applied these materials with greater or lesser degrees of customization, detail, and interpretation.

The commercial ship designer is most likely to apply one or other classification society
approach, or to have to have an alternative approach approved by Class at some stage during
design. It is quite possible that preliminary design may be conducted using one classification
society method, and that final selection of Class will require another. It is therefore useful to
have an understanding of the scope of each set of Class Rules and of their similarities and
differences in approaches. For those involved in military projects, Class Rules and guidance
notes can be valuable reference documents to assist in the design process. An understanding of
the background information, including other codes and standards, can also assist in developing
approaches to new types of design problem.
Table A.1, drawn from the Proceedings of the 13th ISSC in August 1997 [Ref C.32],
provides a good general summary of the procedures for fatigue assessment of ship structures
which are required by the major western classification societies. Most of the Rule systems now
in place have been developed and implemented quite recently, and as a result the levels of
guidance available (and in some cases the requirements themselves) are still changing fairly
rapidly. This table is thus only a snapshot. As can be seen, in some cases (such as the ABS and
LR Rules) the use of a fatigue assessment procedure is mandatory for certain sizes and classes of
ships. In others, it is always optional. Where fatigue assessments are carried out, the ship may
be given a supplementary class notation which acknowledges the procedure. In the case of
DNV, it may also allow the assignment of a specific design life in excess of 20 years, which is
assumed to be the default. Navy standards, if added to the list, would generally show a focus on
detailed analytical treatments.
As indicated in Table A.1, several Rule systems can allow for the treatment of fatigue
with varying degrees of design sophistication. These have been categorized by Lloyd's Register
as Level 1, 2, and 3 methods, and similar terminology will be used here as a convenient
shorthand. Level 1 methods involve the selection of good structural details to minimize the
likelihood of problems. Level 2 performs analyses of probable performance using standardized
assumptions for loading and response. A full Level 3 analysis will use ship-specific loading
data, finite element modelling of global and local response, and realistic representations of any
other effects expected to influence the fatigue life. As noted in the Introduction, each and all of
these may be appropriate to a particular project at different stages in the design process.
Table A.2, from the same source, provides some more detail on the background to each
of the rule systems. Some of the differences in approach are readily apparent, but more are
hidden in the details of each method. Several obvious questions could be posed, for example (a)
are the resulting differences significant, and (b) if so, which approach is correct?

Table A.1: Procedures for Fatigue Assessment of Ship Structures

Class. Reference Brief description of the scope of the document, applicability and when required.
Soc.
ABS ABS The fatigue strength assessment is performed in three steps: Step 1 is a designer oriented
(1996 & assessment for connections of longitudinal stiffeners to transverse webs and bulkheads. Step
1996b) 2 is a simplified fatigue analysis for local hull structures. Step 3 is a comprehensive
structural analysis based on spectral approach for details found inadequate in step 2. The
procedure is applicable for tankers, bulk carriers and container ship.
BV BV The aim of the procedure is to ‘provide the ship designer with relevant information to asses
(1994) fatigue strength and to define the fatigue design criteria to be applied’.
DNV DNV General background is given the rule requirements for fatigue control of ship structure and
(1995) detailed recommendations for such control. Various levels of fatigue assessment procedures
defined include a simplifies approach and a direct calculation approach. Its application is
required for structural details ‘subject to extensive dynamic loading’.
GL GL Rules for simplified fatigue strength analysis. Its application is required for structures which
(1997) are ‘predominantly subjects to cyclic loads.
LR LR Three Levels are given. Level 1 is based on a comparison of the structural details with
(1996) recommendations derived form consolidation of available service experience. Levels 2 and 3
are a simplified and full spectral direct calculations procedures. The procedure is developed
for a double hull oil tankers and bulk carriers and is under development for container and
LNG/LPG ships. Mandatory for new oil tankers and bulk carriers over 190 meters in Length.
Level l and 2 are to be applied and Level 3 at the request of the ship owner or the
shipbuilder.
NK NK A simplified approach for ship design which has been verified for longitudinal stiffeners.
(1996) Research work is under conduction for improving and revising the guidance. The procedure
is applicable for longitudinal, transverse and local strength members of oil tankers, bulk
carriers and container ships.
RINA RINA Rules for checking is the fatigue strength of ship hull structures by means of a simplified
(1995) fatigue analysis. Applicable for ship structures which satisfy RINA standards for obtaining
the highest class made of normal and/or high strength steels. Its application is required for
the special notation FTC bye RINA.
KR KR Guidance for simplified fatigue strength assessment is ship structures at the initial stage.
(1995)

Table A.2: A Short Summary of Different Fatigue Assessment Procedures Available for Ship Design
Class. Loads Stress anal. guid. Fatigue strength 1
Soc. Corrosion Safety Program Guidance
2 3 4 5
Basis Prob. Shape nominal SCF Nom. Local mean thick. method factor name on details
ABS Rule 2 10-8 Weib. simple Yes DoE DoE no6 spec.case net7 no SafeHull yes
BV Rule 10-5 Weib. no Yes DoE DoE8 yes 25mm time9 no10 VeriStar no
DNV Rule/Direct 10-4 Weib. simple/FE Yes no own12 yes 22mm net12 no Nauticus yes
GL Rule 10-6 Lin. simple Yes IIW IIW yes spec.case no13 yes14 Poseidon yes (in rules)
15 16 17 18
LR Simple/spectral approach simple/FE Yes no own no 22mm net no ShipRight yes (in prog.)
NK Direct19 10-4 Weib. FE Yes BS BS yes no no20 yes21 Prime Ship Yes
-8
RINA Rule 10 Lin. simple No IIW IIW yes no no no no Yes
KR Rule 10-4 Weib. simple Yes DoE yes 22mm true22 no no No
1 The S-N data sources are given for nominal (Nom.) and local approaches. BS refers to British Standards 5400, IIW to IIW (1996) and DoE to different editions
of the ref. “Offshore Installations: Guidance and Design Construction and Certification”, Health and Safety Executive (formerly Department of Energy),
U.K.
2 Local approach is the hot spot method in most cases. Comparison of different local approach S-N curves is given in Figure 7.1.
3 Mean stress correction is applied on the stress range basing on the mean stress or in case of NK (1996) on S-N curve by modifying the scope.
4 The thickness effect se accounted for by a factor on stress range basing on the mean stress or in case of NK (1996) on S-N curve by modifying the slope.
5 Mean minus two standards deviations S-N curves are used in most case. Additional safety factors to the rule are referred here.
6 Not explicitly.
7 The stress calculated for net scantlings are multiplied by a factor of 0.95 to reflect a ‘mean wasted condition’.
8 Special local approach is used based on notch stress which is the structural stress multiplied by a weld factor is 1.96.
9 Corrosion is modelled by multiplying the cumulative damage with a correction factor that is a function of corrosion rate and time.
10 Mean minus one, two or three standard deviation S-N curves for non critical , critical or particular structural members.
11 Special local approach is used based on a notch stress that is structural stress multiplied by a weld factor. Default factor of the weld value is 1.5.
12 Stress are calculated using net scantlings and S-N curves for corrosive environment. A simple approach is given for partially effective corrosive protection.
13 Only implicitly for hold frames in bulk carriers.
14 For non-redundant structures and for some rounded corners with large rider plates.
15 The procedure is available through the use of the ShipRight program.
16 Loads used by voyage simulation used in Level 2, Parametric formulas for ships motions and loads in regular waves. In level 3 direct approach is used.
17 S-N curves are based on parametric formulas of the hot spot SCF’s derived from systematic FE-analysis.
18 In level 2 time invariant simulation of thickness reduction due to corrosion is used. In level 3 no corrosion modelling is applied.
19 Two approaches defined are ‘a combination’ and ‘design wave’ methods.
20 If considered, in ballast tanks for example, the stresses should be converted to appropriate values and stress safety factor of 1.1 to 1.3 should be considered.
21 Safety factors are used depending on the importance of the member. Explicit values are not given. For basic joints mean S-N curves are used.
22 The stress concentration factors are applied for stress analysis and S-N curve for corrosive environment. A simple approach is given for partially corrosive
protection.

Figure A.1 provides a partial answer to the first of these. There is significant
variability in the S-N fatigue design curves used by the different classification societies,
and although the uses of the curves also differ, the resulting predictions can be more
rather than less different. The variability can easily be increased by different
interpretations of the input data requirements.
When results from any of the methods are compared with actual service
experience, there is considerable scatter, some of which is inherent in the probabilistic
approach and some of which may result from uncertainties in aspects such as load
modelling or the actual local as-built configuration. Thus, although it can be stated that
any and all of the fatigue analysis methods available are based on rational approaches to
the problem, at the same time it can also be noted that following any one of them will not
necessarily provide conservative predictions for a given structural component. If
calculations show the possibility of fatigue problems within the intended service life of
the ship, it becomes increasingly important to ensure that the fabrication quality at key
details is at least as good as the default assumptions of the analysis method.
A.3 Detailed Comparisons

As there are eight classification society approaches and several other Codes
referenced in Table A.1, it would be a lengthy (and confusing) task to present all the
differences among the approaches. The focus of these comparisons will therefore be the
approach recommended in this Guide, and used in the worked examples in Section D.
The Guide approach is developed from basic data derived for British Standard 7608
[C.33], using the same principles applied in some of the more recent offshore. It is closer
philosophically to the DNV ship design method [C.3] than to most of the others listed in
Table A.1. Some of the most significant differences between the Guide approach and
other rule systems commonly used in North America (such as ABS [C.1] and Lloyds
[C.2]) will be highlighted; other approaches will only be noted where they offer
significant extra assistance in the performance of fatigue analyses.
A.3.1 Load Definition
In all of the approaches to fatigue design, designers are given the option of
developing loads using a spectral approach tied to the expected lifetime and operational
profile of the ship. This can be utilized with response models of varying degrees of
sophistication (linear strip theory, fully non-linear dynamic) and with finite element
models of greater or lesser levels of refinement in order to provide the required input
stresses for the fatigue damage assessments. However, it will almost always be advisable
to perform more simplified load and stress calculations beforehand. These will indicate
where the design may be expected to have its most severe fatigue problems, and they will
also provide a valuable benchmark against which the more sophisticated results can be
compared. Most fatigue (and other) design criteria have been developed using
experience, calibrated against relatively simple assessments of loadings. If a
sophisticated analysis produces results that are significantly different from expected
values, several possibilities should be considered:

(a) the design (or operation) have unique features which were not adequately recognized
in the concept stages;
(b) the sophistication of the analysis warrants reconsideration of traditional safety factors;
or
(c) there are errors in the analysis.
The simplified load models provided (or recommended) by codes differ in most of
their provisions; the differences being smallest - in terms of outcome - when dealing with
the primary stresses, and greatest for the tertiary stresses due to dynamic load
components. In other areas of design practice using classification society rules, many of
the differences will cancel out again due to differences in strength criteria. It is possible
that this is also true for aspects of fatigue design. Designers are therefore cautioned that
fatigue calculations undertaken with one Class simplified method may not be accepted by
another without at least some dialogue.
A significant difference in philosophy between standards comes in using spectral

methods due to the differences in the selection of reference stress exceedence probability.
Different class rules and naval standards fall anywhere within the range of 10-4 to 10-8 for
this. This Guide uses the 10-4 value. This is done for several reasons;
(a) to highlight the fact that most fatigue crack growth (as opposed to fracture initiation)
takes place at relatively low stress ranges; and
(b) if a more extreme probability is used, the results of fatigue damage calculations will
be very sensitive to the spectrum shape parameters, which are often quite uncertain.
A.3.2 Local Idealizations
The Guide is based on using the notch stress, derived from either simple or
complex analyses, as the basis for fatigue damage assessment. This approach has been
adopted in several codes and standards because it removes a certain amount of
subjectivity from the analytical procedure. In the notch stress approach, fatigue damage
accumulation is based on ∆σnotch, where:
∆σnotch = Kw ⋅ ∆σhotspot = Kg ⋅ Kw ⋅ ∆σnom (A.1)
Kw is the local weld configuration stress concentration factor, and Kg is the stress
concentration resulting from the detail under consideration. Other fatigue design
approaches, meanwhile, continue to use either the hotspot or nominal stress range
approaches to predict fatigue. In practice, there are few inherent differences between the
approaches until a Level 3 level of sophistication is applied, since tabulated correction
factors have to be applied to nominal stresses under any approach.
Joint-specific S-N curves, such as those originally developed for the British
Standards for various types of steel construction, are intended for use with the nominal
stresses at the crack site, and incorporate Kg and Kw effects. Such curves are used in the
Canadian and British naval design standards. A set of these curves is shown in Figure
A.2, with two examples of the 'best' and 'worst' typical joint configurations in Figure A.3.

Figure A.1: Hot Spot Fatigue Design Curves (explicit or implicit)
Figure A.2: Example of S-N Curve for Different Joint Configurations

Category C Detail
Category W Detail
Figure A.3: “Best” and “Worst” Typical Joint Configurations
A problem with using specific curves of these types is that it is often difficult to decide
(a) which nominal curve is best suited to the application, and (b) which additional stress
concentration effects need to be considered. Unlike the situations shown in Figure A.3,
actual ship configurations will normally have multi-axial stress states, in which it is
difficult to isolate the stress concentrations built into the S-N curves from those resulting
from the real surrounding configuration.
A step beyond the nominal curves is to use the hotspot stresses, which take
account of more general configuration effects and thus allow single fatigue design curves
to be used for a range of joint types. However, there is no agreement as to which S-N
curve is most appropriate to use with hotspot stresses, and this is one of the main reasons
for the divergence of the curves shown in Figure A.1. ABS, for example, selects the
Category E curve for use with fillet welds, while others are less conservative. Lloyd's
Register's Fatigue Design Assessment software incorporates its own S-N curves, which
have been derived from model and finite element analyses of typical ship structural
details.
Moving to the notch stress approach should allow explicit treatment of all effects
except for those due to the welding process (local material properties, grain size, etc).
There is still no absolute agreement as to which S-N curve best represents the base case
situation, but several organisations have taken the Category D curve, corrected to remove
any weld geometry stress concentration effects. The Guide has done this by applying a
factor of 1.5, which is very similar to the BV and DNV approaches. This provide the
formula:

for N ≤ 107 cycle

log N = log a1 - m1 ⋅ log ∆σnotch = 12.710 - 3.0 log ∆σnotch (A.2)
for N > 107 cycle
log N = log a2 - m2 ⋅ log ∆σnotch = 15.637 - 5.0 log ∆σnotch (A.3)
This type of bilinear formulation is common to all codes, though the location of
the break point varies. Codes which follow the International Institute of Welding (IIW)
approach use 5*106 cycles.
A further aspect of the local idealisation, which may need to be taken into
account, is the material thickness. This will affect the through-thickness stress
distributions and thus the worst stress concentrations. Most of the ship design approaches
use a thickness correction for steel greater than 22-25mm thickness, but for consistency
with the basic data set the Guide applies a similar correction above 16mm thickness.
This takes the form:
log N = log a1 - m1 ⋅ log ∆σnotch - 0.25. m1 ⋅ log(t/16) (A.4)
i.e., the life expectancy is reduced for any thickness, t, greater than 16mm by the
italicised term in the equation. Recent work in the offshore industry has suggested that
this is non-conservative for very thick structures, and the exponent/coefficient should be
raised from 0.25 to 0.3. However, this may only be applicable to the types of joint used
in the offshore industry, and no ship rules have yet adopted the larger correction.
A.3.3 Structural Condition
Steel structures in a corrosive environment are much more prone to fatigue
failures than are those in dry air. Corrosion has both overall and local effects. The
overall reduction in scantlings as thickness is lost increases the global (and hence local
stress levels). At the local level, fatigue and corrosion are mutually reinforcing. Stress
cycling promotes corrosion, and corrosion accelerates crack formation. Coating
breakdown is also most likely to occur at fatigue-prone locations, due to their geometry
and also to the higher cyclic strains.
As shown in Table A.2, the majority of class rules take account of both the overall
and the local effects, the former by applying a correction for thickness loss through life,
and the latter by modifying the S-N curve for the detail under consideration.
Thickness loss corrections are not consistent across rule systems. Table A.3
highlights some of the more significant differences between the ABS and DNV rules,
noting at the same time that in other areas their requirements are identical. The same
would be true of most other rule comparisons.

Table A.3: Corrosion Thickness Allowances

Structural Component tcorr ABS (mm) tcorr DNV (mm)
Longitudinal bulkhead between cargo/fuel oil 1.5 0
tank and dry space
Weather deck beam in ballast tank 2 3
Bottom plating i.w.o. ballast tank 1 1.5
Side shell frames in bulk carrier holds 1.5 1/3*
Note: * refers to upper/lower part of hold
In addition to these direct differences, ABS (uniquely) takes some account of the
progressive loss of thickness through life by 're-correcting' the stress levels downwards
from the values based on net thicknesses.
At the local level, several codes anticipate that a freely-corroding joint - i.e. one
which is not protected by a coating or by an effective cathodic protection system - will
have a design life roughly a factor of 2 shorter than the same joint in a non-corrosive
environment. This is the assumption recommended in the Guide, though some judgement
may still be needed in deciding when to apply it. Modern coating systems, for example
in ballast tanks, may have a probable life expectancy lower than that for the ship as a
whole. They can be very difficult and costly to reapply after full or partial breakdown. It
is therefore realistic to assume that the structural details in such tanks will be protected
for part of their life and unprotected subsequently; and the overall fatigue damage
accumulation calculations should take account of both periods. DNV provides a
simplified method for this that should, in principle, be applicable under any of the rule
approaches.
It should be noted that some recent research has suggested that even a factor of 2
life-expectancy penalty may be non-conservative for freely corroding joints in sea water.
The most recent issue of UK requirements for offshore structures applies penalties of up
to 3 times, and downgrades the assumed effectiveness of cathodic protection.
A.3.4 Safety Factors
Fatigue calculations are probabilistic in nature. The load and stress level
predictions are based on statistical representations of lifetime experience. The fatigue
design (S-N) curves are developed from experimental data with significant levels of
scatter. Source documents for S-N curves will normally quote the mean and standard
deviation values for these curves, to allow safety factors to be tuned to the importance of
the structural component under evaluation, or the level of structural redundancy present.

The majority of codes take failure probabilities two standard deviations below the
mean curve as their default standard, giving a nominal 2.3% failure probability at a
calculated damage index of 1 (normal distribution is assumed). Some of the
classification society rules and guides note that higher safety factors should be used for
the most important details, whose failure could hazard the ship or the safety of personnel.
However, only a limited amount of guidance is offered in most cases. Bureau Veritas
recommends the use of 1, 2, and 3 standard deviations for non-critical, critical, and very
critical details respectively. This equates to 16, 2.3, and 0.14% failure probabilities for
each class of detail respectively, or alternatively to approximately successive factors of
1.5 improvements in life expectancy.
It is possible to use the results of a fatigue damage prediction as inputs to a more

general structural reliability prediction, which can use fracture mechanics approaches to
predict the risk of more catastrophic failures. Such methods can be used to help
categorise the different details in the ship and thus to refine the detailed design.
A.4 Summary
Fatigue life prediction is a relatively new consideration for ship design. It is an
area in which the different design standards development bodies have been very active in
recent years, and where codes are continuing to evolve relatively rapidly. It is still
difficult for designers to identify what the criteria should be for a new design, or to assess
whether they have been met successfully. Fabrication and maintenance procedures can
make enormous differences to the actual performance of nominally equivalent designs.
Existing codes and their accompanying guidance documentation provide useful

tools, particularly if they are used in comparative evaluations within and between
designs. They can highlight potential problem areas, and guide design development
which will keep the risk of fatigue damage to historically acceptable levels. They are less
likely to provide accurate predictions of the onset of failure in any individual structural
detail.

Appendix B-Level 2 Analysis FTL 4714
Formulae for Loads
APPENDIX B
Level 2 Analysis Formulae for Loads
B.1 Wave- Induced Hull Girder Bending Moments
(1) Where direct calculation is not required, the vertical wave induced bending moments may
be calculated using the bending moment amplitudes specified as:
Mds = -0.11kwm Cw L2B (CB + 0.7) (kNm) (B.1)

Mdh = 0.19kwm Cw L2B CB (kNm) (B.2)
where:
Mds = wave sagging moment amplitude

Mdh = wave hogging moment amplitude
kwm = moment distribution factor
= 1.0 between 0.40L and 0.65L from A.P., for ships with low/moderate speed
= 0.0 at A.P. and F.P. (Linear interpolation between these values.)
Cw = wave coefficient
= 0.0792L; L < 100 m
= 10.75 - [(300-L) / 100]3/2; 100 m < L < 300 m
= 10.75; 300 m < L < 350 m
= 10.75 - [(L-350) / 150]3/2; 350 m < L
and other parameters are as defined in the Nomenclature. Note that these are “permissible”
bending moments, i.e. rule limits rather than the actual bending moment developed in a hull
under design conditions.
In general, equations B.1 and B.2 are similar for all the major Classification Societies for
longitudinal strength. This variation over the length of the vessel is accounted for in the moment
distribution factor kwm . If bow-slamming effects are to be considered, the individual Class
Rules provide guidance on modifying vertical bending moment envelope. A general discussion
is provided in Section 3.2.3 in the main text.
Alternative equations for vertical bending moment have been derived for warships and
other slender hulls [Refs. C.19, C.20 ]. The resulting expressions for predicting the extreme
lifetime bending moment including the effects of whipping were then formulated as:
M dh = M sw + 0.0006L2.5 B (B.3.a)
M ds = M sw + 0.0009L2.5 B (B.3.b)
where:
Mdh = design hog moment (ton-ft)
Mds = design sag moment (ton-ft)
Msw = stillwater bending moment (ton-ft)

Formulae for Loads
The equivalent expressions for lifetime bending moment (including whipping) in SI units
are:
M dh = M sw + 0.000115L2.5 B [MN*m]] (B.3.c)
M ds = M sw + 0.000172 L B 2 .5
[MN*m] (B.3.d)
Implicit in these is an operating life of 3600 days; this translates to approximately 3.888 x
107 encounters, and thus differs from the IACS standard of 10-8 encounters. Refer to Section
C.3.1for an approach to modifying the loads to meet a specific encounter probability.
(2) The horizontal wave bending moment amplitude (MH) may be obtained from:
MH = 0.22 L9/4 (Tact + 0.30 B) CB (1-cos(2π x/L)) (kNm) (B.4)
Note that the longitudinal distribution is defined in the equation, where x is the distance in
metres from the A.P. to the section considered. Horizontal bending moment is of primary
concern for slender vessels or open-hatch (large openings) vessels, such as container ships.
(3) Wave torsional loads and moments that may be required for analyses of open type vessels
(e.g., container vessels) can be found in Appendices C and D of [Ref C.3].
B.2) Shear Loads

The formulae for shear loads at midship are derived from conventional beam theory, thus:
(1) Vertical Shear:

Vds = -0.11kws Cw LB (CB + 0.7) (kN) (B.5)
Vdh = 0.19kws Cw LB CB (kN) (B.6)
Again, the Class Rules provide guidance on the distribution of shear over the length of the vessel
using a kws factor.
(2) Horizontal Shear:

VH = 0.22 L5/4 (Tact + 0.30 B) CB (1-sin(2π x/L)) (kN) (B.7)
Torsional Shear formulae can also be obtained from the References if required.
B.3 External Pressure Loads

The external pressure amplitude (half pressure range), pe, related to the draft of the load
condition considered may be taken as:
pe = rppd (kN/m2) (B.8)
where:
rp = reduction of pressure amplitude in the wave zone
= 1.0 for z < Tact - zwl

Formulae for Loads
Tact + c − z
= for Tact - zwl < z < Tact + zwl
2c
= 0.0 for Tact + zwl < z
where:
Zwl = ¾ (PdT/ρg); PdT = dynamic pressure at Tact
The dynamic pressure amplitude may be taken as the largest of the combined pressure
dominated by pitch motion in head/quartering seas, pdp, or the combined pressure dominated by
roll motion in beam/quartering seas, pdr, as:
pd = max (pdp, pdr) (kN/m2) (B.9)

where:
pdp = p1 + 135 | y | /(B + 75) - 1.2(T - z)
pdr = 10[ | y |α/2 + CB

(| y | + k f ) (0.7 + 2z/T)]
16
p1p = ksCw + kf
= (ksCw + kf) (0.8 + 0.15 V/ L ) if V/ L > 1.5
ks = 3CB + 2.5/ C B at AP and aft
= 3CB between 0.2 L and 0.7 L from AP
= 3CB + 4.0/CB at FP and forward
(between specified areas ks is to be varied linearly)
kf = min of Tact or freeboard “f” to weatherdeck < 0.8 Cw (m)
α = roll angle, single amplitude (rad) = 50 zwl / (B + 75) [rad]
y = y > B/4 (m)
V = ship speed (knots)
L = ship length (m)
Z = vertical distance from baseline to POI (m) < Tact (m)
B.4 Internal Pressure Loads due to Ship Motion

The dynamic internal pressure amplitude, pi in kN/m2, may be taken as the maximum
pressure due to acceleration of the internal mass:
 p1 = ρa v hs

pi = f a max p2 = ρa t y s (kN/m2) (B.10)
 p = ρa x
 3 l s
where:
av = combined vertical acceleration (m/s2)
 a rz2 + a z2
a v = max 2
 a pz + a z
2
at = combined transverse acceleration (m/s2)

Formulae for Loads
( )
2
at = a y2 + g0 sinα + ary
al = combined longitudinal acceleration (m/s2)
(
al = ax2 + g0 sin φ + a px )2
ap = tangential pitch acceleration (m/s2)
= ( )
α 2π / T pp 2 RP
apx = longitudinal component of pitch acceleration (m/s2)
= ( )
φ 2π / T pp 2 RPZ
apz = vertical component of pitch acceleration (m/s2)
= ( )
φ 2π / T pp 2 RPX
φ = maximum pitch angle (rad)
= 0.25 ao /CB
TPP = period of pitch (s)
= 1.80 L / g
ax = surge acceleration (m/s2)
= 0.2 g ao CB
ay = acceleration due to sway and yaw (m/s2)
= 0.3 g ao
az = heave acceleration (m/s2)
= 0.7 g ao / CB
ao = acceleration constant
= 3CW/L + CV V/ L
ary = horizontal component of roll acceleration (m/s2)
= α (2 π/ TR)2 RRZ
arz = vertical component of roll acceleration (m/s2)
= α (2 π/ TRp)2 RRY
Rp =
Rpz = Vertical distance from the pitch axis of rotation to the local centre of mass or volume
(m)
Rpx = Longitudinal distance from the pirch axis of rotation to the local centre of mass of
volume (m)
CV = L / 50, max 0.2
go = g Α ao = acceleration constant
TR = period of roll
= 2 kr / GM , maximum 30 (s)
kr = roll radius of gyration (m)
= 0.39 B for ships with even distribution of mass and double hull tankers in
ballast
= 0.35 B for single skin tankers in ballast
= 0.25 B for ships loaded with ore between longitudinal bulkheads
GM = metacentric height (m)
= 0.07 B in general

Formulae for Loads
= 0.12 B for single skin tankers, bulk carriers and fully loaded double hull tankers
= 0.17 B for bulk and ore carriers in the ore loading condition
= 0.33 B for double hull tankers in the ballast loading condition
= 0.25 for bulk carriers in ballast condition
= 0.04 B for container carriers
α = maximum roll angle, single amplitude (rad)
= 50 zwl / (B + 75)
Zwl = (1.25 - 0.25 TR) k′
k′ = 1.2 for ships without bilge keel
= 1.0 for ships with bilge keel
= 0.8 for ships with active roll damping capabilities
In a simplified analysis of bulk or ore cargoes, only p1 need be considered. A hydrostatic

distribution can be considered, although for large granular cargoes, the distribution may in fact
be quite different. The appropriate density and pressure height should be specially considered.
Sloshing pressures may normally be neglected in fatigue computations. However, if

sloshing is to be considered, formulae for sloshing pressures in partly filled tanks may be
obtained from the Classification Society Rules.
In case of partly filled tanks on both sides of a bulkhead, the pressure range may be taken
as the sum of the pressure amplitudes in the two tanks.
The above formulations are examples of what may be used for a Level 2 design approach.
The designer may refer to Classification Society Rules for alternative formulae. Before doing so
he may wish to review the differences by reading Appendix A.

APPENDIX C
STRESS CONCENTRATION FACTORS

FOR SHIP STRUCTURAL DETAILS
This appendix provides guidance on the estimation of stress concentration factors (Kg,
Kte, Kt) for ship structural details where: Kg is a stress concentration factor due to the gross
geometry of the detail, Kte is an additional stress concentration factor due to eccentricity
tolerance (normally used for plate connections only), and Kt is an additional stress concentration
factor due to angular mismatch (normally used for plate connections only).
These stress concentration factors account for the local geometry of the detail, excluding
the weld (Kw – see Appendix A). They do not account for the global stress concentration effects
of the structure surrounding the detail to be analyzed (KG). The latter should be determined by
global FEA or additional published solutions. The total stress concentration factor for the
location, used to determine the peak stress in the load carrying section containing the flaw, is
thus defined as follows:
Ko = KG ⋅Kw ⋅ Kg ⋅Kyr ⋅Kt
The following SCF solutions have been adapted from Cramer et al. (1995). Alternate solutions
may be found in Classification Society documents for fatigue analysis, and previous Ship
Structure Committee reports.
Fatigue Resistant Detail Guide for Ship Structures C-1

Fatigue Resistant Detail Design Guide For Ship Structures

Uploaded by

Copyright:

Available Formats

Fatigue Resistant Detail Design Guide For Ship Structures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fatigue Resistant Detail Design Guide For Ship Structures

Uploaded by

Copyright:

Available Formats

NTIS # PB99-159931

FATIGUE RESISTANT DETAIL

This document has been approved

SHIP STRUCTURE COMMITTEE

SHIP STRUCTURE SUB-COMMITTEE

MARITIME ADMINISTRATION MILITARY SEALIFT COMMAND

1. Report No. 2. Government Accession No. 3. Recipient’s Catalogue No.

7. Author(s) 8. Performing Organization Report No.

17. Key Words 18. Distribution Statement

fatigue, ship structures, structural details, design

Fatigue-Resistant Detail Design Guide for Ship Structures ii

Fatigue-Resistant Detail Design Guide for Ship Structures iii

PART A – FATIGUE DESIGN GUIDE OVERVIEW

A.1 INTRODUCTION A-1

A.2 OBJECTIVE A-1

A.3 BACKGROUND TO FATIGUE IN SHIP STRUCTURES A-2

A.4 SCOPE OF THE GUIDE A-6

A.5 REFERENCES A-10

PART B – CATALOGUE OF FATIGUE RESISTANT DETAILS

B.1 INTRODUCTION B-1

B.2 CRITICAL AND NON-CRITICAL DETAILS B-4

B.3 CATALOGUE TERMINOLOGY B-5

B.4 LAYOUT OF THE CATALOGUE B-6

B-5 FATIGUE RESISTANT DETAIL CATALOGUE B-12

Fatigue-Resistant Detail Design Guide for Ship Structures iv

B.6 “GOOD” DESIGN, CONSTRUCTION AND WELDING PRACTICES TO B-89

B.7 REFERENCES B-94

PART C – FATIGUE STRENGTH ASSESSMENT

C.1 INTRODUCTION C-1

C.2 FATIGUE ANALYSIS PROCEDURES C-2

C.3 LOAD ANALYSIS C-5

Fatigue-Resistant Detail Design Guide for Ship Structures v

C.4 STRESS ANALYSIS C-29

C.5 DEVELOPMENT OF STRESS RANGE DISTRIBUTIONS C-49

C.6 FATIGUE DAMAGE CALCULATIONS C-54

C.7 FATIGUE LIMIT STATE RELIABILITY ANALYSIS C-65

C.8 REFERENCES C-68

Fatigue-Resistant Detail Design Guide for Ship Structures vi

D.1 EXAMPLE 1: LEVEL 2 FATIGUE ANALYSIS SAMPLE APPLICATION D-1

D.2 EXAMPLE 2 – LEVEL 3 ANALYSIS D-10

D.3 EXAMPLE 3 – SIDE PRESSURE CALCULATIONS

Fatigue-Resistant Detail Design Guide for Ship Structures vii

Fatigue-Resistant Detail Design Guide for Ship Structures viii

Figure A.4.4.1 Application of Fatigue-Design Guide in Ship Design Process A-8

Fatigue-Resistant Detail Design Guide for Ship Structures ix

Figure D.1.1.1 Location of Interest for Fatigue Calculation D-1

Fatigue-Resistant Detail Design Guide for Ship Structures x

Table A.3.1.1 Highly Loaded Structural Elements – Tankers A-3

Fatigue-Resistant Detail Design Guide for Ship Structures xi

Fatigue-Resistant Detail Design Guide for Ship Structures xii

Fatigue-Resistant Detail Design Guide for Ship Structures xiii

B Moulded breadth of ship (m or ft); plate thickness [Appendix B]

D Cumulative fatigue damage

E Young’s Modulus [GPA]

Fatigue-Resistant Detail Design Guide for Ship Structures xiv

Fatigue-Resistant Detail Design Guide for Ship Structures xv

HSE UK Health and Safety Executive

IACS International Association of Classification Societies

Fatigue-Resistant Detail Design Guide for Ship Structures xvi