500MPa to typically 700MPa) are increasingly being used in offshore structural applications. This review considers the types of steel used offshore, their mechanical properties, their weldability and their suitability for safe usage offshore.">500MPa to typically 700MPa) are increasingly being used in offshore structural applications. This review considers the types of steel used offshore, their mechanical properties, their weldability and their suitability for safe usage offshore.">
RR 105
RR 105
RR 105
He a lth & S a fe ty
Exe c utive
Re vie w o f the p e rfo rma nc e o f hig h
s tre ng th s te e ls us e d o ffs ho re
Pre p a re d b y Cra nfie ld Unive rs ity
fo r t h e He a lt h a n d Sa fe t y Exe cu t ive 2003
RESEARCH REPORT 105
HS E
He a lth & S a fe ty
Exe c utive
Re vie w o f the p e rfo rma nc e o f hig h
s tre ng th s te e ls us e d o ffs ho re
Pro fe s s o r J Billing ha m, Pro fe s s o r J V S ha rp ,
Dr J S p urrie r a nd Dr P J Kilg a llo n
Sch o o l o f In d u s t ria l a n d Ma n u fa ct u rin g Scie n ce
Cra n fie ld Un ive rs it y
Cra nfie ld
Be d ford s hire
MK43 0AL
High strength steels (yield strength >500MPa to typically 700MPa) are increasingly being used in
offshore structural applications including production jack-ups with demanding requirements. They offer
a number of advantages over conventional steels, particularly where weight is important. This review
considers the types of steel used offshore, their mechanical properties, their weldability and their
suitability for safe usage offshore in terms of fracture, fatigue, static strength, cathodic protection and
hydrogen embrittlement performance. In addition, this review addresses the performance of high
strength steels at high temperatures and at high strain rates. It outlines the difficulties in working with
the very limited published codes and standards and discusses performance in the field. Current design
restrictions such as limits on yield ratios, susceptibility to hydrogen cracking including the influence of
SRBs, and the management of the behaviour of such steels in seawater under cathodic protection
conditions are discussed. Recommendations are made to encourage the wider use of high strength
steels in the future and areas where further study is required are identified.
This report and the work it describes were funded by the Health and Safety Executive (HSE). Its
contents, including any opinions and/or conclusions expressed, are those of the authors alone and do
not necessarily reflect HSE policy.
HS E BOOKS
ii
Crown copyright 2003
First published 2003
ISBN 0 7176 2205 3
All rights reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted in
any form or by any means (electronic, mechanical,
photocopying, recording or otherwise) without the prior
written permission of the copyright owner.
Applications for reproduction should be made in writing to:
Licensing Division, Her Majesty's Stationery Office,
St Clements House, 2-16 Colegate, Norwich NR3 1BQ
or by e-mail to hmsolicensing@cabinet-office.x.gsi.gov.uk
iii
CONTENTS
Page
NOMENCLATURE .................................................................................... vii
SUMMARY................................................................................................ ix
1. INTRODUCTION.................................................................................. 1
2. THE USE OF HIGH STRENGTH STEELS OFFSHORE....................... 3
3. MECHANICAL PROPERTIES OF HIGH STRENGTH STEELS........... 6
3.1 Steel Type and Process Route.................................................... 6
3.2 Metallurgical and Compositional Considerations ..................... 6
3.3 Yield Ratio Considerations ......................................................... 7
4. CODES AND STANDARDS................................................................. 23
4.1 All Properties ............................................................................... 23
4.2 Fatigue.......................................................................................... 24
4.3 Fracture Toughness..................................................................... 24
4.4 Hydrogen Cracking...................................................................... 24
4.5 Defect Acceptance Criteria.......................................................... 25
4.6 Corrosion Protection................................................................... 25
4.7 Static Strength of Tubular Joints................................................ 25
4.8 Impact Properties......................................................................... 26
4.9 High Temperature Properties...................................................... 26
5. FABRICATION AND WELDING.......................................................... 29
6. TOUGHNESS....................................................................................... 34
6.1 Ductile to Brittle Transition........................................................ 34
6.2 Charpy V-Notch Values and High Strength Steel ..................... 35
6.3 Fracture Mechanics Values ....................................................... 36
6.4 Fracture Toughness of High Strength Steels............................ 37
6.5 Flaw Assessment Considerations for High Strength Steels.... 39
6.6 Summary of Toughness Considerations................................... 40
7. FATIGUE IN HIGH STRENGTH STEELS ........................................... 46
7.1 Introduction................................................................................. 46
7.2 Fatigue Crack Propagation ........................................................ 46
7.2.1 Effect of Steel Strength on Fatigue Crack Growth Rate 46
7.2.2 Parent Materials ............................................................... 47
7.2.3 HAZ ................................................................................... 47
7.2.4 Weld Metals ...................................................................... 47
7.2.5 Fatigue Thresholds.......................................................... 47
iv
7.3 Effect of SRB and Sulphides...................................................... 47
7.4 SN Data........................................................................................ 48
7.5 Fatigue Improvement Techniques ............................................. 48
7.6 Summary ..................................................................................... 49
8. CATHODIC PROTECTION................................................................... 64
8.1 Introduction................................................................................. 64
8.2 Protection Criteria....................................................................... 64
8.3 Avoiding Overprotection Problems........................................... 65
8.4 Summary ..................................................................................... 66
9. HYDROGEN EMBRITTLEMENT.......................................................... 70
9.1 Introduction................................................................................. 70
9.2 Sources of Hydrogen in Steels Exposed to the Marine
Environment ................................................................................ 70
9.3 Recent Literature Review........................................................... 71
9.4 Effect of Welding......................................................................... 72
9.5 HE Testing................................................................................... 72
9.6 Hydrogen Embrittlement Test Results ...................................... 72
9.7 Summary ..................................................................................... 74
10. HIGH TEMPERATURE PROPERTIES................................................ 81
10.1 Summary ................................................................................... 82
11. HIGH STRAIN RATES ........................................................................ 84
11.1 Summary ................................................................................... 85
12. FIELD PERFORMANCE OFFSHORE................................................. 88
12.1 Introduction............................................................................... 88
12.2 Production Jack-ups................................................................. 88
12.2.1 BP Harding..................................................................... 88
12.2.2 Siri .................................................................................. 88
12.2.3 Hang Tuah ACE Platform.............................................. 89
12.2.4 Elgin-Franklin................................................................ 89
12.3 Drilling Jack-ups....................................................................... 89
12.4 Tension Leg Platforms ............................................................. 91
12.5 Summary ................................................................................... 91
13. INSPECTION AND REPAIR............................................................... 94
13.1 Summary .................................................................................. 94
14. DESIGN RESTRICTIONS................................................................... 96
14.1 Buckling of Members............................................................... 96
14.2 Static Capacity of Tubular Joints............................................ 96
14.3 Draft ISO Standard Recommendations for High
Strength Steels......................................................................... 96
15. SUMMARY AND CONCLUSIONS ..................................................... 99
v
APPENDIX 1 OTHER STRUCTURAL APPLICATIONS OF
HIGH STRENGTH STEELS BOLTS &
THREADED FASTENERS................................................ 102
APPENDIX 3 ............................................................................................. 104
APPENDIX 6 FRACTURE TOUGHNESS CONCEPTS.......................... 105
A6.1 Ductile to Brittle Transition: An Introduction to Fracture.... 105
A6.2 Charpy V-Notch Values: Background ................................... 105
A6.3 Charpy V Notch Values for High Stength Steel .................... 107
A6.4 Fracture Mechanics Tests: Background............................... 108
A6.4.1 Brittle Materials ........................................................ 108
A6.4.2 Ductile Materials ....................................................... 109
A6.5 Fracture Mechanics Values for High Strength Steels.......... 110
A6.6 Flaw Assessment Considerations for High Strength Steels 111
APPENDIX 9 ............................................................................................ 117
A9.1 Slow Strain Rate Testing......................................................... 117
A9.2 Fracture Mechanics Testing ................................................... 117
vi
vii
NOMENCLATURE
a Crack length parameter
B Thickness of fracture specimen
C
E
or CE Carbon equivalent
CP Cathodic protection
CR Controlled rolled
CTOD Crack tip opening displacement
C
v
Charpy impact energy
Crack tip opening displacement value
da/dN Crack growth rate
K Stress intensity factor range
Youngs modulus
Embrittlement index
FCAW Flux cored arc welding
FCGR Fatigue crack growth rate
FMD Flooded member detection
HAC Hydrogen assisted cracking
HAZ Heat affected zone
HE Hydrogen embrittlement
HIC Hydrogen induced cracking
HSLA High strength low alloy
HSS High strength steels
HV Vickers hardness
ICCP Impressed current cathodic protection
ISO International Standards Organisation
J Joules
K
app
Applied stress intensity factor
K
c
Apparent toughness
K
1A
Arrest toughness
K
1c
Material toughness
K
1D
Stress intensity factor to keep crack in
motion
K
ISCC
Fracture toughness under conditions of
stress corrosion cracking
K
th
Threshold value of K
LBZ Local brittle zones
MAC Martensite-austenite content
MMA Manual metal arc
MPa Mega Pascals
Q&T Quench and tempered
y
Yield stress
R curve Resistance curve (energy per unit area of
crack extension)
R
e
Specified min. yield strength (in fracture
toughness equations)
SACP Sacrifical anode cathodic protection
SAW Submerged arc welding
SMYS Specified minimum yield strength
S-N Stress versus no. of cycles in fatigue
SRB Sulphate reducing bacteria
viii
SSCC Sulphide stress corrosion cracking
SSRT Slow strain rate testing
TLP Tension leg platform
TMCP Thermomechanically controlled
processing
TMCR Thermomechanically controlled rolling
TT Ductile to brittle transition temperature
UKCS UK Continental shelf
UTS Ultimate tensile strength
Poissons ratio
YR Yield ratio (
y
/UTS)
YS Yield strength
ix
SUMMARY
High strength steels (yield strength >500MPa to typically 700MPa) are increasingly being
used in offshore structural applications including production jack-ups with demanding
requirements. They offer a number of advantages over conventional steels, particularly where
weight is important. This review considers the types of steel used offshore, their mechanical
properties, their weldability and their suitability for safe usage offshore in terms of fracture,
fatigue, static strength, cathodic protection and hydrogen embrittlement performance. In
addition, this review addresses the performance of high strength steels at high temperatures
and at high strain rates. It outlines the difficulties in working with the very limited published
codes and standards and discusses performance in the field. Current design restrictions such
as limits on yield ratios, susceptibility to hydrogen cracking including the influence of SRBs,
and the management of the behaviour of such steels in seawater under cathodic protection
conditions are discussed. Recommendations are made to encourage the wider use of high
strength steels in the future and areas where further study is required are identified.
x
1
1. INTRODUCTION
Fixed offshore structures are conventionally constructed from medium grade structural steels, with
yield strengths typically in the range of 350MPa. These steels are well documented and covered by
existing codes and standards. However, in recent years there has been an increasing interest in the use
of higher strength steels for these installations, recognising the benefits from an increase in the
strength to weight ratio and the associated savings in the cost of materials. As a result, significant
parts of several platforms (jacket and topsides) have been constructed from 400 450MPa steel and
installed in the North Sea. However, to date, fatigue sensitive components (e.g. tubular joints) have
generally been fabricated from medium strength steel because of the better knowledge on these steels
regarding fatigue performance and the lack of increased performance of high strength steels in this
area.
The principal application of very high strength steels offshore has been in the fabrication of jack-ups.
Steels with nominal yield strengths in the range 500 800MPa are normally used in fabrication of
legs, rack and pinions and spud cans. These units, used primarily for drilling, have many years of
satisfactory experience in use, operating in a variety of water depths, but are normally brought into
dry dock for inspection at 5 year intervals, where any damage or cracking can be found and repaired.
In recent years there has been increasing interest in the use of jack-ups for production, where periodic
dry dock inspection is not possible. Two jack-ups for production, utilising high strength steels, are
now installed (Harding in 1996, Siri in 1998) and a third (Elgin-Franklin) is due to be installed shortly
on the UKCS. High strength steels have also been used in tethering attachments for floating
structures in TLPs (tension leg platforms) and for mooring lines with semi-submersible module
offshore drilling units (MODUs).
A considerable amount of research has been undertaken on high strength steels in recent years
providing new data to support offshore applications. However, overall there is limited information of
the long-term use of high strength steels in seawater, particularly under the severe environment
conditions to which structures on the UKCS are subjected. Particular concerns with the use of higher
strength steels are the greater susceptibility to hydrogen cracking which can be enhanced when SRBs
are present, their fatigue and fracture performance, and, for offshore applications, their performance at
higher temperatures as a result of fire.
Most codes and standards relate to medium strength steels and in most cases the use of design
formulae is limited to steels with yield strengths <500MPa which is a serious disadvantage for the use
of higher strength steels. Despite the increasing amount of data available from research and testing,
very little of this has yet found itself into codes and standards. The application of current codes and
standards to high strength steels will be reviewed in this report.
Several reviews of high strength steels offshore have been produced and published between 1995 and
1999 [1.01-1.08]. The plan adopted for this review is to use the information in these as a basis, but to
include new data, applications and codes and standards since these were produced. In particular,
where possible, the performance of steels with a yield strength of 450MPa, where there is some
published data, will be used as a baseline for assessing the performance of higher strength steels. In
some cases the performance of medium grade steels (YS ~350MPa) will need to be used as a
benchmark since limited data are available for higher strength steels. This review covers a wider
scope than previously published reviews of high strength steels in that it includes fire and impact
resistance as well as field performance and inspection and maintenance aspects. High strength steels
up to 450MPa (X70 grade) have been used for many years in offshore pipelines. However, pipelines,
in general, have different design criteria and requirements from offshore structures. Pipelines were
therefore not included in the current review except where general principles of steel metallurgical
development or the like were a common feature to both types of application.
2
REFERENCES
1.01 Billingham J, Healy J, and Spurrier J, Current and potential use of high strength steels in
offshore structures, Publication 95/102 MTD, Sept. 1995, 51 pages, ISBN 1-870553-24-1
1.02 Billingham J, Healy J, and Bolt H, High strength steels the significance of yield ratio and
work hardening for structural performance, Marine Research Review 9, 36 pages, published
MTD 1997, ISBN 1-870553-27-6
1.03 Billingham J, Sharp J V, Spurrier J, and Stacey A, The use of high strength structural steels
in offshore engineering, Intnl.Symposium on Safety in Application of High Strength Steel,
Trondheim, July 1997
1.04 Sharp J V, Billingham J, and Stacey A, Performance of high strength steels used in jack-ups
in seawater Marine Structures, Vol.12, 1999, 349-369, ISSN 09 51-8339, Elsevier 1999
1.05 Healy J, and Billingham J, A review of the corrosion fatigue behaviour of structural steels in
the strength range 350-900MPa, and associated high strength weldments, Offshore
Technology Report, OTH 532, Publ. Health & Safety Executive, 1997, ISBN 0-7176 2409-9
1.06 Robinson M J, and Kilgallon P J, Review of the effects of sulphate reducing bacteria in the
marine environment on the corrosion fatigue and hydrogen embrittlement of high strength
steels, Offshore Technology Report, OTH 555 98, HSE Books (1999), Sudbury, Suffolk
1.07 Robinson M J, and Kilgallon P J, Review of the effects of microstructure on the hydrogen
embrittlement of high strength offshore steels, Offshore Technology Report (in press no
number yet issued) HSE Books (1999), Sudbury, Suffolk
1.08 Stacey A, Sharp J V and King R N, High Strength Steels used in Offshore Installations,
Proceedings 15
th
International Conference Offshore Mechanics and Arctic Engineering,
Florence, Italy, June 1996, Vol III
3
2. THE USE OF HIGH STRENGTH STEELS OFFSHORE
Traditionally, offshore structures have been fabricated with moderate strength steels with yield
strengths up to 350MPa [2.01], mainly produced by the normalising route. However, there has been a
significant growth over the past twenty years in the use of high strength steels in the offshore
engineering industry, primarily driven by a desire to save weight and cost [2.02]. Table 2.1 shows the
main application areas involved which vary with the strength of steel used. Such steels are also
generally produced by alternative processing routes such as thermomechanical controlled processing
(TMCP), and quenching and tempering (Q & T).
The principal advantage of using these structural materials is their increased strength to weight ratio
and the attendant savings in materials costs and construction schedules [2.03] due to reduced amounts
of welding. The most important increases have occurred in the topside areas of jacket structures
where the weight saving has not only produced overall savings in materials used but has allowed
crane barge installation of more complete topside processing and accommodation units [2.04] with
significant related cost savings. A survey undertaken in 1995 [2.05] indicated that the proportion of
high strength steel (defined as >350MPa yield strength) used in offshore structures increased from
less than 10% to over 40% over little less than a decade.
More recent applications, especially with smaller, lighter structures, involve the use of such steels in
the jacket members themselves although there are still usually restrictions to their use in nodal areas
because of concerns related to fatigue performance. It is likely that the use of such materials will
continue to increase as the steels become more widely available and construction yards get more
experienced in fabrication procedures. To date, most steels have been restricted to 450 grades but
research and development programmes [2.06] have indicated that steel grades up to 550MPa can be
produced which are readily weldable and possess excellent fracture toughness. Such steels will
increasingly be utilised as they become more widely available.
Higher strength steels (>550MPa and often up to 700MPa) are usually produced by the quenching and
tempering route and have traditionally been used offshore in mobile jack-up drilling rigs which do not
stay permanently on station and are periodically dry docked, allowing inspection and repair
programmes to be implemented [2.07]. The principal application of high strength steels in jack-ups is
in the fabrication of the legs because of the requirement to minimise weight during the transportation
stage. In general, each lattice leg is composed of three or four longitudinal chord members which
may contain a rack plate for elevating the hull and a series of horizontal and diagonal tubular braces
which connect the chords to form a truss. Supplementary braces (span breakers) are frequently used
between main brace mid-points to increase the buckling resistance. The rack plate is very thick,
varying typically between 150 and 250mm. The chord shell cans are usually fabricated from plate
with a wall thickness between 35 and 80mm and with a diameter in the range 800 to 1200mm.
Weldability and good toughness and ductility are important material considerations in this application
and the steel maker provides this by careful control of alloy composition and by processing [2.01;
2.04].
Steels of similar strength levels have only comparatively recently been used in production jack-ups
permanently on station in North Sea projects in the Harding and Siri fields, and in the Elgin jacket
which was installed in 2001. In such installations, fatigue, corrosion fatigue and hydrogen
embrittlement become major design considerations and the steels used have to be carefully reassessed.
The French TPG 500 design is a good example of this type of structure. It can be built onshore as one
complete unit and floated out to site. Once on station the legs can be lowered to the seabed and the
deck jacked up for operation, thus reducing the need for heavy lift operations during installation and
producing significant cost savings. A second benefit in this design is that it is a reusable production
facility, since it can be refloated, removed from one site to another, and commence operations in the
new field. To provide the required fatigue life the legs of the structure have incorporated forged
4
nodes (fabricated by Creusot-Loire), thus reducing the stress concentrations normally seen in welded
nodes. The Harding platform is in 110m water depth and the lower part of the structure comprises a
concrete base to remove any potential problems of hydrogen embrittlement related to sulphate
reducing bacteria in the mud zone. A jack-up was also installed as a permanent installation in the
Danish sector (Siri field, 60m depth) in 1998. The majority of the leg sections on the Siri platform are
made from thick section (65-110mm) 690 grade high strength steel. The Elgin structure uses a range
of high strength steels including 500MPa steel in structural members, chords and bracings and
700MPa steel in the racks. It uses lower strength steels (350MPa) in the lower part of the structure
which are piled into the seabed to avoid potential hydrogen embrittlement problems.
Other applications for high strength steels are found in mooring attachments for floating structures
such as tension leg platforms (TLPs). These structures are fixed by vertical tension legs to piled
foundation templates on the seabed. One of the first such designs for the Hutton field in the UK
sector used 16 tension legs (4 at each corner). Each leg is a thick walled steel tubular, manufactured
from a low alloy steel (3.5%Ni, Cr, Mo, V) with a minimum yield strength of 795MPa [2.08]. The
individual components of each leg are forged and connected by screwed couplings. The steel
composition was selected to provide the highest possible strength, commensurate with adequate
fracture toughness. Resistance to both stress corrosion cracking and corrosion fatigue were also
important. The choice of screwed connectors was based on the fact that there were insufficient data
available on the corrosion fatigue performance of welded tubulars to guarantee safe performance
under the envisaged design life of the structures. A large test programme was undertaken to justify
the choice of material. The platform has now been in operation for almost 20 years without any
significant problems with the tethers.
Since then TLP type platforms have been installed in several other fields, both in Norway and in the
Gulf of Mexico. In the Heidron field, for example, a welded TMCP (thermo-mechanically processed)
microalloyed X70 pipeline steel was used for the tethers. The structure is in 270m water depth and
comprised tubular tension leg elements that were 1m diameter and 38mm thick. The steel had a yield
strength of 500MPa and impact toughness of 60J at -40C. Other structures have been used in much
deeper waters, mainly in the US (Auger TLP in 872m, installed in 1994; Mars TLP in 896m of water
in 1996), where conventional fixed platforms are uneconomic. These projects involved 76mm thick
415MPa components of weldable TMCP steel. Plans are in hand for even deeper water TLP units, but
it is now recognised that the availability of suitable high strength materials for the tethers is limiting
further development.
Other floating structures such as semi-submersibles, used welded higher strength steel anchor chains
or wire ropes as their mooring attachments. Such steel chain and wire rope components are
considered outside the scope of the present review.
Other, more specialised, usage areas include flanges and repair clamps where threaded fasteners
provide the main load transfer mechanism. Such bolts and threaded fasteners are discussed in more
detail in Appendix I.
5
Table 2.1
High strength steels used offshore
Strength
MPa (grade)
Process Route Application Area
350 (X52) Normalised
TMCP
Structures
Structures & Pipelines
450 (X65)
Q & T
TMCP
Structures
Pipelines
550 (X80)
Q & T
TMCP
Structures & Moorings
Pipelines
650
Q & T
Jack-ups & Moorings
750
Q & T
Jack-ups & Moorings
850
Q & T
Jack-ups & Moorings
REFERENCES
2.01 Billingham J, Steel a versatile advanced material in marine environments, Ironmaking and
Steelmaking 1994, Vol.21, No.6, p422
2.02 Billingham J, Healy J, and Spurrier J, Current and potential use of high strength steels in
offshore structures, MTD Publication 95/102 (1995), ISBN 1-870-533-24-1
2.03 Rodgers K J, and Lockhead J C, The welding of grade 450 offshore structural steels, Proc
Conf. on Welding and Weld Properties in the Offshore Industry, London, April 1992
2.04 Webster S E, Structural materials for offshore structures past, present and future,
Proc.Conf. on Safe Design and Fabrication of Offshore Structures, IBC, London, Sept. 1993
2.05 Healy J, and Billingham J, High strength steels a viable option for offshore designs,
Euroforum Conference Latest Innovations in Offshore Platform Design and Construction,
London 1996
2.06 The influence of welding on materials performance of high strength steels offshore,
Managed Programmes of University Research, Marine Technology Centre, Cranfield
University, 1985-1994
2.07 Sharp J V, Billingham J, and Stacey A, Performance of high strength steels used in jack-ups'
Journal of marine Structures, 12 (1999), 349
2.08 Salama M N, and Tetlow J H, Proc. of Offshore Technology Conference, Houston, Texas,
1983, Paper 449.
6
3. MECHANICAL PROPERTIES OF HIGH STRENGTH STEELS
3.1 STEEL TYPE AND PROCESS ROUTE
In general, the strength of a steel is controlled by its microstructure which varies according to its
chemical composition, its thermal history and the deformation processes it undergoes during its
production schedule. In addition, structural steel for offshore applications must be readily weldable
since this is the traditional fabrication route for offshore structures. Structural steels for offshore must
therefore be available in moderate to thick sections (30 100mm) and must exhibit good toughness to
avoid the possibility of brittle failure, in addition to showing good weldability and high strength.
Such overall requirements are often difficult to achieve because an increase in one of these properties
often leads to a decrease in the others.
Table 3.1 below shows the strength ranges and process routes for high strength steels used in a variety
of offshore engineering applications.
Most conventional structures use only moderate strength steel produced by the normalised or
thermomechanically processed routes (TMCP) but at higher strength levels there are processing
thickness restrictions to TMCP steels and normalising cannot produce the strength levels required in
the necessary section thicknesses. Quenching and tempering is therefore the standard production
route for very high strength structural steel. The limitations that apply to the different process routes
in respect of strength or thickness ranges are shown in Table 3.2.
3.2 METALLURGICAL AND COMPOSITIONAL CONSIDERATIONS
The offshore pipeline industry, for many years, has used high strength steels and today commonly
uses X70 steel grade (!450MPa) with excellent toughness and weldability properties [3.01].
Significant benefits in such developments have come from understanding the complex chemistries
developed for the steel plus the use of extensive thermomechanical processing, primarily to produce
fine grained microstructures, including controlled rolling, thermomechanical controlled processing
and accelerated cooling. Many of the principles involved in such developments, particularly the
complex interactions between strength, toughness and weldability as influenced by steel chemistry,
heat treatments and thermal processing [3.02] have been carried over into higher strength steel
development.
Many of these well understood metallurgical principles can be utilised to satisfy the overall
mechanical property requirements for high strength structural steels, namely:
"# reduced carbon content to improve weldability and toughness;
"# decreased grain size (ferrite and/or bainite) to give increased strength and increased toughness.
This is usually achieved by microalloying with Nb, V or Al and by some form of
thermomechanical processing;
"# decreased impurity content (S, P, O) to increase toughness in particular and through thickness
homogeneity, i.e. the use of clean steel technology;
"# increased alloying with Ni, Cr, Mo and Cu to give solid solution and transformation
strengthening, especially at the higher strength levels.
Relatively small changes in composition and/or variations in processing route can significantly affect
the resulting mechanical properties as shown in Table 3.3. This table shows old and new versions
of steels within 3 standard steel grades, i.e. 355, 450 and 690MPa yield strength. Although all of the
steels within a particular grade satisfy the grade requirements (primarily with respect to specified
minimum yield strength) it can be seen that the newer versions show much improved overall
properties by combining the required yield strength with improved toughness (improved Charpy
impact performance {Cv}), and improved weldability (lower carbon equivalent values [C
E
1
}). This
1
Carbon equivalent CE is defined as CE=C+Mn/6+(Cr+Mo+V)/5+(Ni+Cu)/15 See page 30 for more details.
7
has been achieved primarily by controlling the microstructure through changes in chemistry and
thermal processing.
Many engineers and designers do not appreciate that the mechanical properties of a particular steel
can vary significantly within a specified steel grade (i.e. steel with a specified minimum yield
strength). Figure 3.1 shows the variation in mechanical properties for three offshore steel grades with
minimum yield strengths of 355, 420 and 450MPa [3.03]. For the 450MPa steel, for example, it can
be seen that the yield strength can vary by 100MPa from 440 to 540MPa (+20% on design yield
value), with a mean value at approximately 500MPa. Such variations are produced by variations in
steel composition and processing which affect all the mechanical properties as shown in Figure 3.2 for
a 450MPa steel. Such variations can have serious implications for the degree of weld overmatching
or undermatching that occurs in the final structure. The range of properties achievable within a
particular grade can also vary significantly with process route as shown in Figure 3.3 which illustrates
the much wider variation obtained from the TMCP route than from either the normalised route at
lower strength levels or the quenched and tempered route at higher strength levels [3.04]. Variations
can also occur with plate thickness and with steel manufacturer [3.05]. It is important that this
potential variation in yield strength is recognised at the design stage.
Typical compositions and properties for high strength steels produced by the normalised,
thermomechanically controlled processed and quenching and tempered routes are shown in Tables
3.4, 3.5 and 3.6 respectively. Other typical composition and mechanical properties are given in the
draft DNV Offshore Standard OS-B101 Metallic Materials, (May 2000). In general for such steels the
strength increases as the hardenability and the composition related carbon equivalent values increase
(see Figure 3.4). for each process route, but the particular process route selected generally has a more
significant influence on yield strength. By making use of strength improvements associated with the
processing route, it is possible to produce steels of the same strength level at leaner chemistries and
hence lower carbon equivalent values, which show improved weldability. By using careful control of
composition and processing, steels can therefore usually be produced with excellent combinations of
strength and toughness combined with excellent weldability. In general, as the strength increases the
weldability in particular decreases and more control over the welding procedures such as increased
levels of preheating are usually required. Moreover, in general, the toughness of very high strength
steels (690MPa) is inferior to that of steels with low (350MPa) or intermediate (450-550MPa) levels
of strength.
3.3 YIELD RATIO CONSIDERATIONS
The stress strain behaviour of high strength steels differs somewhat from that of lower strength steels
in that they generally show reduced capacity for strain hardening after yielding and reduced
elongation as shown in Figure 3.5. This is because the steel strengthening mechanisms used in high
strength steel development have been selected specifically to increase the yield strength and have
much less influence in subsequent strain hardening behaviour. One measure to illustrate this different
behaviour is yieId ratio (YR) which is defined as the ratio of yieId strength (
y
) to ultimate tensile
strength (UTS), and which generally increases as the strength of the steel increases as shown in Figure
3.6 for a range of offshore steel grades [3.06]. YR is not, however, a unique measure of how the steel
behaves because steels with very different stress strain curves can have the same value of YR [3.06].
There are restrictions in structural design codes to reflect this changed behaviour such that YR for
material to be used for structural members is not allowed to have a value greater than 0.85 in design
equations to ensure that there is adequate ductility in the member to develop plastic failure behaviour
as a defence against brittle fracture. Design aspects related to YR are given in section 13 of this
report. Examination of a different database [3.07] shown in Figure 3.7 shows that generally steels
with yield strengths up to 500MPa can satisfy this general requirements but that very high strength
steels do not.
The yield ratio is not directly related to the capability of a given steel to withstand plastic strain after
yield and before fracture. In older high strength steels, elongation generally decreases as yield ratio
8
increases, but modern clean steels with low carbon content and low levels of impurity have significant
elongation even at the highest strength (690 grade) and yield value ratio (0.95), giving more
confidence as to their deformation capability [3.05]. An alternative measure is the general elongation
which is usually substantial in modern steel with high yield ratios.
Current design equations are based on test data from medium strength steels, where some degree of
strain hardening is present. Lack of strain hardening can lead to premature cracking, which could
have significant implications for tubular joints in service. As a result, design codes have placed a
limit on the yield ratio (typically 0.7). Examination of Figure 3.7 emphasises the range of values that
can occur within particular strength grades, largely due to the different methods of production,
differences in steel chemistry, and differences in section thickness that occur (see section 3.2).
Indeed, from this diagram it can be seen that 350MPa steels generally show a yield ratio ranging from
0.6 to 0.8, that 450MPa steels have values ranging from 0.7 to 0.87, whereas 690MPa steels have
values ranging from 0.9 to 0.95. Elongation generally decreases in line with increasing yield ratio;
therefore for 350 450 steels, elongations are generally of the order of 20 35%, whereas for 690
steels, values of 14 18% are more typical [3.06].
Examination of Figure 3.7 shows that many steels, even at strength levels up to 400MPa have yield
ratios above 0.7, which could include many steels purchased at grade 355 level. It is therefore
possible that some earlier structures might have nodal joints which do not satisfy the current design
codes although they have performed perfectly satisfactorily in service.
There has for some time been a feeling that the code restrictions for nodal connections are rather
conservative in respect of high strength steels because intuitively it would be expected that joint load
capacity would increase in line with yield strength, whereas the code restrictions impose severe
limitations. For example, on increasing the yield strength from 355 to 532MPa (a 50% increase) the
designer is only allowed to increase the allowable design stress by 23% when the yield ratio is 0.85
(YR = 0.85, design stress = 0.7UTS = 0.7 x 626 = 438MPa). Other examples of the restrictions
imposed by the code are given in Appendix 3. Initial finite element studies on X joint deformation
behaviour by BOMEL and Cranfield [3.06] indicated that joints with high YR had significantly higher
joint capacities than joints with low ratios. For example, a joint with a YR of 0.93 (490MPa $
y
,
525MPa UTS) showed a 28% increase in joint capacity compared to a lower strength steel $
y
=
350MPa with a YR of 0.66 and the same value of UTS (525MPa UTS). Existing structural codes
would have restricted the capacity of both these joints to the same value (YR = 0.66 $
D
= $
y
= 350,
YR 0.93, $
D
= 2/3 x UTS = 2/3 x 525 = 350). Despite the above enhanced in-joint capacity, the
capacity does not increase linearly with yield stress as indicated by the design equations (static
strength for DT joints):
%
% & '
(
)
*
+
,
- Q ) 14 5 . 2 ( T
YR
7 . 0
, 1 MIN F P
2
y
where P = static design strength, F
y
= yield strength, YR = yield ratio, T = wall thickness, % =
diameter ratio d/D, and Q
%
is a geometric factor, defined as Q
%
= 0.3/%(1-0.833%) for % >0.6 and Q
%
=
1 for % <0.6). In the case quoted a direct dependence on yield strength would have indicated a 40%
increase in capacity rather than the measured increase of 28%. However, on the basis of this work,
the authors concluded that the current restriction of 0.7 could possibly be relaxed to 0.8. Other
workers, notably Healy and Zettlemoyer [3.08], and Wilmshurst and Lee [3.09] have also indicated
that they thought the current restriction was too severe and should be relaxed.
An analysis of the limited static strength data for tubular joints manufactured from high strength steels
(>600MPa) [3.10] showed that the recommended restriction of 0.7UTS was justified. However, the
analyses were carried out using the HSE Guidance Notes equations where ultimate strength is the
failure criterion. The point was also made that the data were very limited and that the range of joint
9
types tested was limited (1 T, 8 K/YT, 2 DT). In addition, the geometry range was limited (lowest
gamma was 14.8, highest beta 0.43). The range of yield ratios was 0.88 to 0.94.
A later study for the European Commission [3.11] which involved some relatively large scale
experimental tests by BOMEL, TNO and Delft Universities largely confirmed these earlier
experiments. This programme included a significant finite element study, a comprehensive re-
examination of the test database used in setting up the structural code formulations and an expanded
experimental test programme involving 2 series of tests (compression and tension) on DT joints made
from 350, 450 and 690MPa steels. The finite element analyses successfully reflected the
experimental DT test joint data. They also isolated and quantified accurately the effect of geometric
imperfections in the test joint. The earlier indications that the design stress for tubular joints should
be raised from the YR = 0.7 value in the guidance were confirmed. The authors concluded that the
conservatism utilised in high strength steel tubular steel design could be reduced by changing the YR
= 0.7 limit in the design equation to 0.8 for both compression and tension joints. This would then
enable more widespread use to be made of high strength steel in tubular joint design. The authors also
recommended that more tests on tubular joints, especially at the higher grades of steel and especially
in tension, were required.
A recently published paper [3.12] from Thyssen Krupp Stahl AG analysed transition temperatures
from Charpy V-tests and the fracture mechanics transition temperature for 690MPa steels against
yield ratio and found no correlation, but it was recognised that these results were from small scale
tests. Another analysis of the maximum net stress versus test temperature of the wide plate tests for
690MPa steels, with yield ratios ranging from 0.87 to 0.93 showed that the highest loads were in the
steels with the highest yield ratio. The authors concluded that yield ratio is not a good measure of
component safety, and that other factors should be taken into account. This paper [3.12] also listed
limitations on yield ratio in various design codes and materials standards (both onshore and offshore)
ranging from 0.7 to 0.93 for various components.
Since 1996, several Panels have been meeting to draft a new ISO standard for offshore structures.
This includes a Panel drafting a section on the static strength of tubular joints. The Panel has re-
examined the test data on joint strength and developed some improved design equations. However,
because of the lack of data on higher strength steel joints, the Panel concluded that these equations
should be limited to steels with yield strengths less than 500MPa (see section 4). However, even for
these steels, it was considered necessary to impose a limit of the yield ratio (since some lower strength
steels have yield ratios greater than 0.7). The Panel concluded that on the basis of test results for
lower strength steels, the limiting yield ratio should be 0.8.
For higher strength steels (yield strength greater than 500MPa) the Panel concluded that use of a
limiting value of yield ratio of 0.8 may be adequate to permit the ultimate compression capacity
equations to be used for joints with strengths in the range 500-800MPa provided adequate ductility
can be demonstrated in the HAZ and parent material. It is unclear how this demonstration of adequate
ductility can be provided either in terms of mechanical property data of the steels concerned or in
identified test procedures.
Later examination of some of the available static strength data [3.13] has concluded that although the
factor for compression loading could be relaxed to 0.8, the factor for tension loading should, indeed,
be lowered to 0.5 based on the design capacity being related to first cracking rather than to ultimate
strength as in, for example, the API RP2A code. Failure modes in the compression tests involved an
indentation of the chord of about 30% of the diameter. Cracks appeared in the tension specimens at
loads of around 50% of the maximum load reached in the tests. However, it should be recognised that
these recommendations are based on very limited test data.
Overall, the design of high strength welded joints for static strength is unclear based on the very
limited existing data. When failure is defined as the onset of cracking (under tension loading, e.g. as
in API RP2A) it would appear that the existing design equations are unconservative for high strength
10
steel joints, even with a yield ratio of 0.7. For compression loading a limiting yield ratio of 0.8 would
appear appropriate, provided there is adequate ductility present in both the HAZ and parent plate.
Further data are required to resolve these uncertainties.
An alternative approach to the concerns regarding the influence of high yield ratio and deformation
capacity of high strength steel tubular joints is to redesign the steel and its production route to develop
high strength steels with lower yield ratios. Japanese studies [3.14] aimed at developing steels for
earthquake resistant structures have utilised accelerated cooling and intercritical quenching
procedures to produce a microstructure of ferrite dispersed in a bainitic matrix which can produce
high strength steel 500MPa in thick sections with a yield ratio of 0.7 and CE values of 0.4 which
would indicate reasonable weldability. 700MPa steels with YR of only 0.83 have also been
developed but these are less weldable (CE 0.52) [3.15]. The weldability of such steel is inferior to
modern HSLA steels used offshore but could almost certainly be improved with further development.
Castings can offer advantages over welded structural fabrications because the joint intersections can
be easily contoured to reduce stress concentration effects, at say nodal joints for example, with a
corresponding increase in fatigue life [3.16]. In conventional welded nodal joints the fatigue life is
decreased because of the microcracking that exists in the weld toe which is eliminated in cast joints
with a corresponding significant increase in fatigue life as shown in Figure 3.8. High strength steel
castings are available which have attractive combinations of strength and toughness properties [3.17;
3.18]. Steels with yield strengths up to 690MPa are available. They cannot derive their strength from
processing so usually have addition of nickel and chromium to suppress transformation temperatures
and produce low carbon martensitic or bainitic structures. The excellent through thickness properties
of castings have also opened up new markets in lifting attachments, spreader bars, and pad eyes
[3.16].
REFERENCES
3.01 Steels for Linepipe and Pipeline fittings, Proc.Intnl.Conf. London 1981, publ Metals Society
3.02 Billingham J, Steel A Versatile Advanced materials in marine Environments, Ironmaking
and Steelmaking, Vol.21, No.6, 452, 1994
3.03 Billingham J, Healy J, and Spurrier J, Current and Potential Use of High Strength Steels in
Offshore Structures, MTD publication 95/102, published 1996, ISBN 1-870553-24-1
3.04 Denys R, Conf Evaluation of Materials in Severe Environments, ISI Japan, Vol.2, 1989.
3.05 Healy J, and Billingham J, Metallurgical Considerations of the High Yield to Ultimate Ratio
in High Strength Steels for Use in Offshore Engineering, 14
th
Intnl.Conf. OMAE 1995,
Vol.III, 365
3.06 Billingham J, Healy J, and Bolt H, High Strength Steel the Significance of Yield Ratio and
Work Hardening for Structural Performance, Marine Research Review 9, publ MTD 1997,
ISBN 1-870553-27-6
3.07 Willcock R T S, Yield : Tensile Ratio and Safety of High Strength Steels, HSE Report, Mat
R108, 1992
3.08 Healy B E, and Zettlemoyer N, In plane bending strength of circular tubular joint, Proc. 5
th
Intnl.Symposium on Tubular Structures, ed. M G Coutie and G Davies, E and F N Span, 1993
11
3.09 Wilmhurst S R and Lee M M K, Non-linear FEM Study of Ultimate Strength of Tubular
Multiplaner Double K Joint, Proc 12
th
OMAE, Glasgow 1993.
3.10 Lalani et al, ESCS Report 7210 MC/602 1996
3.11 Static Strength of High Strength Steel Tubular Joints, ECSC 7210 MC/602 1996, reported in
Wicks P J, and Stacey A, Static Strength of High Strength Steel Tubular Joints, Proceedings
ETCE/OMAE 2000 Conference, Paper 2081, February 2000, Publ, ASME
3.12 Kaiser H J, Kern A, Niessen T, and Schriever, Modern High Strength Steels with Minimum
Yield Strength up to 690MPa and High Component Safety, Proc. 11
th
Intnl. Offshore and
Polar Engineering Conference, Norway 2001, ISBN 1-880653-55-9.
3.13 Private Communication with N W Nicholls
3.14 Shikani N, Kurihora M, Tagawa H, Salkui S, and Watanabe I, Development of high strength
steel with low yield ratio for large scale steel structures, Proc. of Microalloying 88, Chicago
1988, p481
3.15 Toyoda M, Strain Hardenability of High Strength Steels and Matching Properties in Welds,
8
th
Intnl. OMAE Conf. 1989
3.16 Marston G J, Novel Application of Structural Steel Castings in the Offshore Industry The
Safe Design and Fabrication of Offshore Structures, IBC Conf., London, 1993
3.17 Richardson R C, Higher Strength Cast Steel for Offshore Structures, World Expro 161
3.18 Cowling M J, Fatigue Performance of Cast Steel Intersection for Offshore Structures, in
Fatigue Crack Growth in Offshore Structures, ed. W D Dover, S Dharmavason, F P Brennan
and K J Marsch, EMAS 1995.
12
Table 3.1 Strength ranges and process routes for high strength steels used in offshore
engineering
Type
of structure
Strength levels used
(MPa)
Process
Route
Jacket structures and topsides
Pipelines
Jack-ups/Moorings
350 500
350 550
(X52) (X80)
500 850
Normalised
Q & T
TMCP
TMCP
Q & T
Table 3.2 Steel processing routes for production of high strength structural steels
Normalised
Thermomechanically controlled rolled
(TMCR)
Accelerated cooled (TMCP)
Quenched & Tempered (QT)
Castings
Usually <460MPa for 50mm plate
Thickness restriction especially at higher
strengths usually less than 550MPa at
40mm
Improved properties compared to TMCR but
thickness restriction at higher strengths
(a) Alloyed route no real thickness
restriction but expensive and costly to weld
(b) Microalloyed route thickness and
strengths required offshore can be produced
Usually alloyed because of lack of processing
capability
13
Table 3.3
Effect of changes in processing and alloying methodology on mechanical properties of Grade 355, 450 and 690 steel plates
Chemical Composition
Steel
design
Process
B
C
Mn
Si
Ni
Cr
Mo
Cu
S
P
Al
V
Yield
Strength
(MPa)
Cv Impact
Toughness
Welda
bility
(CE
IIW
)
Grade 355
BS4340
50D
Normalised
OLD
- 0.20 1.35 0.30 - - 0.016 0.015 0.02 - 360 70J @ -40C 0.43
BS7191
355EMZ
Normalised
NEW
- 0.11 1.50 0.40 0.15 0.15 0.005 0.015 0.03 - 380 >200J@-40C 0.39
BS4360
50D
TMCP
- 0.07 1.49 0.21 0.38 0.02 0.002 0.008 0.02 - 380 200J @ -30C 0.36
Grade450
Q1N
Q & T
OLD
- 0.18 0.4 0.30 3.0 1-1.8 0.015 0.005 0.02 0.02 550 80J @-85C 0.81
BS7191
450EMZ
Q & T
NEW
- 0.11 1.49 0.3 0.52 0.11 0.001 0.010 0.03 - 480 300J @-40C 0.40
Dillinger
450TMCP
TMCP - 0.09 1.50 0.3 - - 0.001 0.007 0.03 0.04 500 300J @-30C 0.35
Grade 690
Q2N
Q & T
OLD
- 0.11 0.42 0.23 3.40 1.48 0.46 0.03 0.001 0.012 0.026 0.08 550
690
80J @-84C 0.81
OX812 Q & T
NEW
- 0.11 0.89 0.26 1.18 0.46 0.38 0.15 0.003 0.008 0.07 0.01 690 100J @-80C 0.52
SE702 Q & T
NEW
0.0027 0.125 1.05 0.25 1.4 0.5 0.45 0.20 <0.002 <0.01 - 750 120J @ -40C 0.59
DSE 690V Q & T
NEW
- 0.15 0.90 0.33 1.28 0.49 0.45 0.2 0.001 0.009 0.073 0.03 700 74J @ -60C 0.59
14
Table 3.4
Typical composition and mechanical properties of normalised steels produced in Europe yield strength range 350 to 490MPa
Typical composition (by weight %)
Thickness
(mm)
C
Mn
Si
S
P
Nb
V
Al
Cu
Ni
Cr
Mo
CE
IIW
Typical mechanical yield
strength/CVN range
25
0.20
1.35
0.42
0.016
0.015
0.028
-
0.022
-
-
-
-
0.43
360MPa/70J @ -40C
20 0.22 1.0-
1.6
0.55 0.030
max
0.035 - - - 0.3 0.5-
0.7
0.2 0.1 0.52
420MPa/60J @ 0C
20 0.22 1.6 <0.6 0.04
max
0.04 0.003
-0.10
0.003
-0.20
- - - - - 0.49
450MPa/60J @ 0C
30 0.13 1.52 0.49 0.005 0.015 0.03 0.10 0.02 0.45 0.72 - - 0.50 490MPa/>110J @
-20C
Table 3.5
Typical composition and mechanical properties of thermomechanical controlled processed steel yield strength range 400 to 500MPa,
typical average plate thickness 30mm
Typical composition (by weight %)
Thickness
(mm)
C
Mn
Si
S
P
Nb
V
Al
Cu
Ni
Cr
CE
IIW
Typical mechanical yield
strength/CVN range
30
0.10
1.33
0.28
0.002
0.015
0.027
-
-
-
-
-
0.35
400MPa/190J @ -40C
32 0.12 1.35 0.30 - - - - - 0.01 0.02 - -
398MPa/300J @ -20C
32 0.07 1.45 0.27 0.001 0.004 - 0.01 0.07 0.19 0.4 - 0.32
400MPa/>300J @ -20C
30 0.04 1.52 0.22 0.003 0.005 - - - 0.60 0.49 0.02 0.37
460MPa/220J @ -40C
Table 3.6
Typical composition and mechanical properties of quenched and tempered steels yield strength range from 450 to 1000MPa
Typical composition (by weight %)
Thickness
(mm)
C
Mn
Si
S
P
Nb
V
Al
Ti
Cu
Ni
Cr
Mo
B
CE
IIW
Typical mechanical yield
strength/CVN range
6 140
0.18
0.1 -
0.4
0.15-
0.35
0.075
0.015
-
<0.02
0.015
<0.02
<0.2
2.25-
3.25
1 -1.8
0.2-0.6
-
0.81
550 to 690MPa / 80J @ -84C
-
0.2
0.1
0.4
0.15
0.35
0.0254
0.025
0.03
-
-
-
0.25
2.25
3.25
1 1.8
-
-
0.7
690MPa minimum
30
0.10
1.6
0.50
0.005
0.015
0.03
-
-
-
0.35
0.50
0.15
-
-
0.45
450MPa / >35J @ -40C
50 64
0.12
1.50
0.4
0.005
0.020
-
0.06
-
0.01
0.15
0.30
0.10
-
-
0.43
480MPa / >50J @ -40C
50
0.11
0.89
0.26
0.003
0.008
0.02
0./01
0.07
0.01
0.15
1.18
0.46
0.38
0.002
0.64
690MPa / >40J @ -40C
30
0.17
1.2
0.22
-
-
-
-
-
-
-
1.5
0.49
0.5
0.002
0.64
960MPa / >40J @ -40C
15
Figure 3.1
Variation in yield strength for 355, 420 and 450 grade steels
16
Figure 3.2
Showing typical variation in mechanical properties for a grade 450 steel (35-50mm thick, min
y
= 430MPa, min UTS = 530MPa, min elongation =
20%, sample size N = 94)
17
Figure 3.3
Variation in mechanical properties with process route for steel grades 350 and 420 after Denys
[3.04]
18
Figure 3.4
Effect of carbon equivalent value and steel processing route on plate strength
19
Figure 3.5
Typical stress-strain curve for Grades 355, 450 and 690 steel
True stress strain lines for different steel grades
Load-deflection lines for different steel grades
20
Figure 3.6
Offshore grade steels, nominal thickness 50mm
21
Figure 3.7
Yield ratio of 200 cast and wrought iron high strength steels
22
Figure 3.8
Showing comparison of welded and cast steel properties
23
4. CODES AND STANDARDS
Detailed codes and standards now exist for medium strength structural steels, covering most of the
aspects relevant to offshore design. However, for higher strength steels (YS 500-600MPa) the
available codes and standards are limited and for even higher strength steels (>600MPa), almost non-
existent. This section reviews the available codes and standards, both published and those being
developed at present for offshore use. In terms of offshore hazards, table 4.1 shows the current status
of codes and standards. The current status of codes and standards will now be reviewed in terms of
materials properties.
4.1 ALL PROPERTIES
HSE/D.Energy Offshore Guidance has been developed over many years, with the first edition being
published in 1974. The fourth edition was published in 1990, [4.05] with significant amendments
being added up to 1995 [4.07]. It is particularly strong in materials properties, performance of
structural components etc, and has several sections devoted to high strength steels. In particular the
control of hydrogen assisted cracking is addressed in some detail, more than in any other existing
code or standard. However as a result of the issue of the DCR Regulations [4.08] in 1996 it has been
withdrawn, although it is still available as a document for consultation.
Two ASTM standards [4.01,4.02] cover some aspects of the requirements of high strength steels,
although these are limited in their applicability. A808 is concerned with high strength, low ally
carbon, manganese steels of structural quality, whilst A514 provides a specification for high yield
strength Q&T alloy steels, intended primarily for use in welded bridges and other structures.
The recently published NORSOK standard on the Design of Steel structures [4.09] includes five steel
quality levels (DC1 - DC5). However all of these grades are limited to steels with YS equal to or less
than 500MPa. For higher strength steels it is stated that the feasibility of such a selection of steel shall
be assessed in each case.
The International Association of Classification Societies (IACS) provides a set of requirements for
high strength quenched and tempered steels [4.11], for steels with YS in the range 420 690MPa,
divided into six groups. The requirements include method of manufacture, mechanical properties, and
inspection during manufacture.
The DnV offshore standard [4.04] groups steels into three main grades, the highest of which (extra
high strength) covers materials with yield strengths from 420 - 690MPa. These grades are linked to
impact toughness properties according to weldability requirements, but the improved weldability
grade is limited to a maximum YS of 500MPa. A statement is also made that steels with YS> 550MPa
shall be subject to special considerations for applications where anaerobic conditions may
predominate.
Similarly in the draft ISO Standard for fixed structures (19902) [4.10] steels are classified into five
groups, with Grade V covering steels with YS up to 500MPa. It is also stated that further groups may
be added when data becomes available. The draft standard includes an important statement on higher
strength steels, which is:
'Although steels with yield strengths in excess of 500MPa (73 ksi) are currently available, no
agreed standard exists for offshore fixed platform structural use. These are not recognised as offshore
fixed platform structural grades and users should take care to ensure that ductility, fracture toughness
and weldability will be adequate for the intended application. Attention is drawn to the need to
consider fatigue and corrosion conditions, including the tendency for higher strength steels to be more
susceptible to hydrogen embrittlement and certain types of stress corrosion. Particular care should be
exercised where high strength is developed as a result of alloy additions'.
24
A separate ISO Technical group is developing a standard for jack-ups, which is expected to include
guidance on higher strength steels, appropriate to jack-ups, but has yet to be drafted.
4.2 FATIGUE
New Guidance was published by HSE in 1995 [4.07]. This included a modified set of S-N curves, but
these were restricted to steels with yield strength equal to or less than 500MPa, as it was concluded
that the test data available were insufficient for higher strength steels. This was particularly true for
fatigue in seawater under cathodic protection and free corrosion conditions, where the data available
on high strength steel joints were extremely limited. The HSE Guidance recommended that for higher
strength steels, data from an approved test programme are used to determine appropriate S-N curves,
or fracture mechanics constants. Following incorporation of the DCR Regulations offshore in 1996
the HSE Guidance has been withdrawn.
DnV Rules [4.05] include S-N curves and fracture mechanics constants for steels with YS up to
500MPa.
The NORSOK standard [4.09] provides recommended S-N curves for steels, both in air and seawater.
As noted earlier these apply to steels with steel quality levels from I to V, the maximum yield strength
being 500MPa. For steels of higher strength it is stated that the feasibility of such a selection shall be
assessed in each case.
The draft ISO standard [4.10] states that the limited amount of test data for plate joints with yield
strengths up to 540MPa and tubular joints manufactured from high strength steel with yield strengths
up to 700MPa suggests that fatigue performance in seawater under CP and under free corrosion is
similar to that for medium strength steels, but test data should be used to determine appropriate S-N
curves. In addition, the draft standard indicates that for even higher strength steels (700 800MPa)
the effect of seawater on the fatigue performance of these materials is considered to be more
detrimental than for medium strength steels because of their greater susceptibility to cracking from
hydrogen embrittlement. In particular, it is noted that several studies have shown that excessively
negative CP protection potentials can be a cause of cracking due to the generation of hydrogen which
enhances crack growth rates. It is stated in conclusion that it is important that the fatigue performance
of high strength steels is understood and that appropriate levels of CP are applied.
4.3 FRACTURE TOUGHNESS
Most codes and standards recommend the need to avoid brittle fracture. Good specifications are
published for medium strength steels but generally there is very limited guidance for higher strength
steels. Overall avoidance of brittle fracture is based on recommending a minimum value of Charpy
energy values according to yield strength. On this basis the International Association of Classification
Societies (IACS) [4.11] has recommended for high strength Q&T steels that the average energy from
a charpy V notch test should be Re/10 for the longitudinal direction, and 2/3 of this for the transverse
direction, i.e. for 690MPa steels (F grade) Charpy energy values of 69J and 46J at a test temperature
of -60
o
C with minimum individual values of 70% of the minimum average, i.e. 48J and 32J
respectively [4.11]. Possible limitations on this requirement are considered in section 6.
4.4 HYDROGEN CRACKING
As a result of the detection of cracking in jack-ups in the late 1980s a significant research programme
was undertaken on high strength steels which led to new guidance being developed to minimise
cracking in practice.
HSE published an amendment to its Guidance [4.07] with a recommendation that the CP level should
be limited to a negative voltage no lower than -850mV (Ag/AgCl). To achieve this special measures
were recommended, such as voltage limiting diodes to keep potentials within the recommended limits.
In addition steels proposed for use offshore in conditions where there is a vulnerability to hydrogen
cracking should be assessed using, for example, slow strain rate testing. High strength steels
(YS>650MPa) should be examined for the possibility of hydrogen damage in service, both in the
25
parent material and in the weldments. The HSE Guidance was supported by a published OTH report
[4.13] which provided data on the performance of several steels and on the recommended test
methods.
The DnV Offshore standard [4.04] also provides guidance on the use of high strength steels in
seawater with CP. In this case the recommended range for steels susceptible to hydrogen induced
cracking is -770mV to -830mV for steels with YS larger than 550MPa, which is similar to the HSE
recommendations.
4.5 DEFECT ACCEPTANCE CRITERIA
BS 7910 published in 1998 [4.06] contains data for calculating crack growth under static and cyclic
loading conditions. Recommended values of the constants (A,m) are given for steels with yield
strengths up to 600MPa, thus enabling the fatigue crack growth rate acceptance of defects in higher
strength steels found during inspection or assumed during design to be quantified. For higher strength
steels (>600MPa) it is recommended that test data are required.
4.6 CORROSION PROTECTION
It is stated in the NORSOK standard [4.14] that for high strength steels (YS>700MPa) a special
evaluation is required with respect to hydrogen impact. (See EN 10002, metallic materials. Tensile
testing. Part 1 method of test).
The DnV code [4.04] provides both general requirements for cathodic protection as well as specific
needs for high strength steels. Steels with specified minimum yield strengths >550MPa are subject to
special considerations for applications where hydrogen induced stress cracking (HISC) may be
anticipated, where qualification testing should be carried out for critical applications such as legs and
spud cans. In the absence of such testing to demonstrate that high negative CP levels are not harmful
it is stated that the CP level should be limited by the use of special anodes or controlled voltage type
(e.g. with diodes) or by other methods. CP potentials levels should also be monitored to ensure
compliance with the target range, which is set to be within the limits of -770mV to -830mV
(Ag/AgCl). In the case of observed exceedance of this range it is recommended that inspection for
HISC should be carried out,
Section 19 of the draft ISO standard [4.10] is concerned with corrosion control, and includes a section
on cathodic protection. This states that because of the risks of hydrogen induced stress cracking steels
with minimum yield strengths in excess of 720MPa should not be used for critical cathodically
protected components without special considerations. In addition, it is stated that any welding or other
fabrication affecting ductility or tensile properties, should be carried out according to a qualified
procedure, which limits hardness to HV350. It is expected that this will restrict the use of welded
structural steels to approximately 550MPa maximum specified minimum yield strength.
For medium strength steels the recommended potential range is -0.8 to -1.1 volts (Ag/AgCl). For
some higher strength steels the negative end of this range is expected to be detrimental, in terms of
hydrogen cracking etc.
4.7 STATIC STRENGTH OF TUBULAR JOINTS
Current offshore design codes provide equations for determining the static strength of various classes
of tubular joints. The strength is generally proportional to yield strength, but data indicate that this
proportionality is limited to lower strength steels. As a result the basic equations are limited to steels
with YS <500MPa, and there is also a factor to be applied on the ratio of the yield to ultimate
strength. This factor varies in different published codes and standards, ranging from 0.67 in API
RP2A [4.15] to 0.7 in HSE Guidance [4.05]. The draft ISO standard (19902) [4.10] has increased this
ratio to a value of 0.8 for steels with yield strengths up to 500MPa, as a result of new data being
available. The background to the application of this factor and its relevance to high strength steels is
reviewed in section 3.3.
26
For higher strength steels the draft ISO standard recommends that the basic ultimate compression
capacity equations may be used, together with a ratio of the yield to ultimate strength limited to 0.8,
provided adequate ductility can be demonstrated in both the HAZ and parent material (however, the
criteria for demonstrating this are not provided). The draft ISO standard highlights that the limit on
yield ratio for the tension capacity of joints based on first cracking may need further investigation (see
section 3).
The static strength of cracked high strength steel joints is of interest, particularly for the use of
flooded member detection. Some recently published data for high strength steels (SE702) [4.16] have
been made available from a series of nine static tests performed on large pre-cracked welded tubular
joints (six T joints, three Y joints). These were loaded to failure in axial and out-of-plane bending.
All specimens had a least one through thickness crack. The results were analysed in terms of both
loss of static capacity due to the cracking and by failure assessment diagrams (FAD). The reduction
in static strength compared to cracked medium strength steels was about 5% greater, possibly due to
differences in crack path (the cracks in the SE702 steel stayed closer to the weld when growing).
Using the FAD approach, some discrepancies were found for a T joint with two cracks, giving low
values. This was considered possibly due to the inadequacy of the multiple crack correction used.
The FAD approach also demonstrated the importance of the fracture toughness value used. For the
SE702 steel it was necessary to use the K
Q
(nominal) toughness value rather than the maximum
toughness value, K
max
, for which many of the results were unconservative.
4.8 IMPACT PROPERTIES
Low speed impacts can arise from both ship impact and dropped objects. In these cases the ability of
the high strength steel to absorb the appropriate energy is one of the main performance requirements.
Current codes and standards specify the level of impact energy to be absorbed during ship impact
(typically 4MJ) but this is not related to materials properties or yield strength, even for medium grade
steels.
High speed impacts can be the result of explosions and the energies of projectiles can vary from
several kilojoules to several hundred kilojoules. There is no published guidance on the required
material properties and the possible effect of yield strength.
4.9 HIGH TEMPERATURE PROPERTIES
High strength steel components offshore can be subjected to high temperatures as a result of fire,
either on the sea or from a jet fire. Unprotected steels can experience temperatures up to 1200
o
C in a
short space of time. Guidance is normally associated with either limiting temperatures (e.g. by
passive protection requirements) or by design based on data for lower strength steels at elevated
temperatures [4.16]. Some new data have recently been obtained for steels with YS ~450MPa [4.17]
(see section 10). For even higher strength steels the data on high temperature performance are
extremely limited.
REFERENCES
4.01 ASTM standard A514, Standard specification for high yield strength, Q&T alloy steel plate
suitable for welding, 2000, ASTM
4.02 ASTM standard A808, Standard specification for high strength, low alloy carbon,
manganese, columbium, vanadium steel of structural quality with improved notch toughness,
2000, ASTM
4.03 AWS Structural Welding Code, Steel, D1.1-96
4.04 DNV Offshore Standard, Design of Offshore Steel Structures, General (LRFD), 2000
27
4.05 Dept. of Energy, Offshore Installations - Guidance on Design, Construction and certification,
HMSO, London, 1990
4.06 British StandardGuidance on methods for assessing the acceptability of flaws in welded
structures, BS 7910, 1991
4.07 HSE, Offshore Installations - Guidance on Design, Construction and Certification, HSE
Books, 1995, amendment no. 5
4.08 HSE, Offshore Installations & Wells (Design & Construction etc) Regulations, HSE Books,
1996
4.09 NORSOK, Design of Steel StructuresN-004, 1998
4.10 ISO Petroleum & Natural Gas Industries, Fixed Steel Offshore Structures, ISO CD 19902,
to be published
4.11 International Association of Classification Societies (IACS), Unified requirements, Section
W16, High Strength Q&T steels for Welded Structures, IACS, London, 1994.
4.12 DnV Offshore standard, Metallic materials, OS-B101, 2000
4.13 Abernethy, K, Fowler, C M, Jacob, R, Davey V.S., 'Hydrogen cracking of legs and spudcans
on jack-up drilling rigs - a summary of results of an investigation', HSE Report OTH 91 351,
HSE Books
4.14 NORSOK, 'Cathodic Protection', M 503, 1997
4.15 API, Recommended Practice for planning, designing and constructing fixed offshore
platforms, APIRP2A 20
th
edition, API, Washington, USA
4.16 Talei-Faz B, Dover W D, Brennan F P, Static strength of cracked high strength steel tubular
joints, UCL Final report for HSE, 2001
4.17 Steel Construction Institute, 'Experimental data relating to the performance of steel
components at high temperatures', Offshore Technology report, OTI 92 602, HSE Books,
1992
4.18 Steel Construction Institute, Elevated temperature and high strain rate properties of offshore
steels, Offshore Technology report OTO 020 2001, HSE Books.
28
Table 4.1
Offshore
Hazard
Materials Performance
Requirement
Codes & Standards for
HSS
Comments
Structural
Failure
Materials specifications
Welding specifications
Fatigue of welded
joints, members
Fracture toughness of
steels
Hydrogen
Embrittlement
Static strength of tubular
joints
Defect acceptance
criteria
Corrosion protection
Inspection & repair
ASTM standards,
A514,A808 [4.01,4.02],
DnV Standard [4.04]
AWS Code [4.03]
Limited
IACS recommendations
[4.11]
HSE Guidance [4.05,
4.07], DnV Rules [4.04]
Factor to be applied for
higher strength steels in
several standards,
including draft ISO
BS7910 [4.06]
Limited, HSE Guidance,
DnV code [4.04]
Very limited
DnV code [4.04] covers
steel grades up to
690MPa
Covers steel grades up to
690MPa
Most codes provide a
limit of 500MPa on
yield strength (YS) for
applicability. Specific
tests are proposed for
high strength
steels(HSS) to develop
data to support
application
Minimum Charpy values
of YS/10
Control of hydrogen
assisted cracking is best
described in HSE
Guidance Notes [4.04]
Modified factor, based
on yield ratio (under
discussion in draft ISO
standard)
Data available for HSS
(yield strength up to
600MPa)
Most data provided for
medium strength steels
-
Boat
Impact
Impact performance,
large strain capacity
None Lack of data for HSS
Fire (on the
sea, jet fire
etc.)
High temperature
performance
Very limited
Lack of data for HSS
Blast
High strain rate
performance
Very limited
Lack of data for HSS
29
5. FABRICATION AND WELDING
Most high strength steel applications offshore involve welded fabrication. Forming and welding costs
are the most significant item in the cost of a jacket structure, comprising up to 57% of total costs in
one reported analysis [5.01]. Improvements in this area are likely to come from welding since cold
forming of plates is generally considered to be an efficient and economical process. Welding
processes which give greater productivity and/or incorporate a reduction or elimination of pre- and
post-welding heating could provide major cost savings, and significant progress has been made.
However, as the strength of the steel increases the pre-heating requirement becomes greater as such
steels are usually more highly alloyed. If plate thickness is less than 40mm, stress relieving heat
treatment is not required for grades with yield strengths up to 450MPa. This can lead to significant
time and cost savings in the fabrication procedures. Other areas to consider are the development of
welding processes with improved weld deposition rates.
Reports from fabricators [5.02; 5.03; 5.04] indicate that at least up to 450MPa strength levels, welding
is no more expensive or difficult for a well organised yard than welding the normal 355 grade. With
the highest strength grades, however, more precautions have to be taken.
Weldability of steel is a term that is used to indicate the ease with which sound weldments can be
produced using normal welding procedures. The weldment comprises both the weld and the
associated heat affected zone. Welding defects such as pores and cracks can be produced as well as
undesirable microstructures in the weld and its associated heat affected zone which can lower the
resulting mechanical properties of the joint.
Variations in the welding process, such as steel dimensions, weld geometry, heat input and steel
composition all influence the resulting microstructure. Nomograms involving thermal severity - joint
thickness (mm), heat input of the weld (kJ/mm) and weld preheat required (C) are often used to
indicate the necessary welding procedure to be followed to produce a sound crack-free joint in
relation to the particular composition of the steel used which is usually related to carbon equivalent
value. In general a steel with lower carbon equivalent value has improved weldability compared to a
higher carbon equivalent steel. The two most commonly specified carbon equivalent equations are
that recommended by the International Institute of Welding which covers a wide range of steels:
15
Cu Ni
5
V Mo Cr
6
Mn
C CE CE
IIW
&
&
& &
& & - -
and the Ito and Bessyo equivalent which is often preferred for modern low carbon steels:
B 5
10
V
15
Mo
60
Ni
20
Cr Cu Mn
30
Si
C P CE
CM
& & & &
& &
& & - -
This latter equation is the one used for high strength steels in the draft DNV Metallic Materials OS-
B101 Standard (May 2000). In this guidance document, steels with improved weldability have
reduced carbon contents and limitations on the levels of chromium, nickel and molybdenum
compared to steels of normal weldability, i.e. they must have reduced CE values.
An alternative approach more commonly used in other parts of the world is the Graville diagram
shown in Figure 5.1 which separates the steels into three zones rated by their ease of weldability
zone I easily weldable, zone II weldable with care, and zone III difficult to weld, From this diagram
it can be seen that weldability decreases as the carbon equivalent value increases but the diagram also
emphasises the extremely important effect of carbon content on weldability. Reducing the carbon
content of a steel is the most effective way to improve its weldability.
30
As the parent strength increases, greater precautions are needed to ensure that welding procedures are
satisfactory. The strength increases in the weld are normally produced by alloying since
strengthening procedures such as thermomechanical processing cannot be utilised in the weld metal.
The welds therefore become more hardenable and precautions are required to prevent weld metal
hydrogen cracking. The weldability of modern steels has been greatly improved by their extreme
cleanliness, and by their low carbon content and low carbon equivalent values. Low hydrogen
consumables are important in reducing the possibility of hydrogen cracking and can also lead to a
reduction in the pre-heating requirements. No major problems have been reported in welding steels
up to 500MPa yield strength [5.04] in moderate section sizes. At high strength levels, preheating is
required and steelmakers are devoting considerable attention to improving the weldability of such
steels to try to reduce fabrication costs. For 690 grade steels, for example, preheat temperatures of
125C are recommended, and electrodes and fluxes with very low hydrogen content must be used in
order to prevent hydrogen cracking [5.05].
The formation of hard or brittle phases in the weld HAZ or, indeed, in the weld itself during multi-
pass welding, can affect the toughness of the weld and its ability to withstand exposure to hydrogen.
Important factors are the grain size in the grain coarsened HAZ near to the fusion line and the
microstructural changes that occur in the weld metal during subsequent weld depositions during
multi-pass welding. In general, the Charpy toughness of the coarse grained HAZ decreases with
increasing heat input and increasing impurity content. Because of this there are often restrictions in
the upper levels of heat input that can be used (e.g. 3.5kJ/mm for submerged arc welding) but
productivity is not greatly compromised because of the generally thinner sections and smaller
volumes of deposited weld metal utilised. Steel that could be welded satisfactorily at higher heat
input levels would offer economic advantages [5.06] and recent work [5.07; 5.08] showed that certain
steels and weld consumables did satisfy these requirements and could offer further economic
advantages.
Published literature indicates that there are weld consumables which can produce the necessary
material properties required in service, even for the 690 grade steels [5.09]. However, at the highest
strength levels envisaged there is much less experience and availability of weld consumables with
suitable properties, particularly in respect of toughness. In addition, significant pre-heating and
interpass control are necessary in order to avoid hydrogen cracking problems. Weld metal
microstructures are determined primarily by the chemical composition, the amount of non-metallic
inclusions present in the microstructure which affect phase nucleation, and by the cooling rate. Alloy
design aims to maximise the amount of acicular ferrite present and to minimise the effects of
undesirable microstructures such as coarse grain size, grain boundary ferrite and coarse
martensite/austenite/carbide constituents (MAC).
The welding consumables employ sophisticated alloying techniques, incorporating the optimum
balance of deoxidising elements (aluminium, silicon and manganese) to produce a high density of
small non-metallic inclusions which are known to act as intragranular nucleation sites for acicular
ferrite. The carbon content is generally kept low to aid weldability, so the increased strength is
achieved through additions of molybdenum in SAW wires and titanium-boron in FCAW wires, and
the impact toughness is improved with nickel additions. In the Cranfield study [5.08] consumables up
to 550MPa yield strength showed adequate toughness throughout the weld, with upper shelf values
>150J and the 50J impact transition temperatures below -60C. All welds had low hardness values
and showed no indication of hydrogen cracking. Acicular ferrite was the major microstructural
feature of the welds and microstructures generally coarsened as heat input increased.
For both SAW and FCAW, 690MPa consumables showed mixed microstructures containing acicular
ferrite, martensite and polygonal ferrite. Impact properties were inferior to the 550MPa welds with
upper shelf values ranging from 80-100J and 50J impact transition temperatures between 50 and -
80C. It was concluded in this study that more development work is needed before these consumables
can be specified generally for offshore application.
31
In most welded structures it is considered desirable to overmatch the yield strength of the parent plate
and the related HAZ. This is because if the welds undermatch then any enforced deformation will be
concentrated in the relatively small weld metal volumes leading to high strain in these zones. If such
zones have reduced toughness values, which is generally the case at the highest strength levels, then
there is an increased likelihood of failure. Weld metal specifications often call for 20 to 30%
overmatch which is easily achievable in 450 and 550 grades with satisfactory weld property
performance. The problem arises in producing the necessary combination of properties in the weld
metal required at the highest strength levels, i.e. 690 grade. In other applications such as storage
tanks, undermatching welds have been used in high strength steel structures which have generally
performed satisfactorily because the weld metals have good toughness. The normal statistical
variations in yield strength that occur in steel plate (discussed in Section 3), also occur in weld metals.
This poses an additional problem because there is a distinct possibility that, unless significant levels
of overmatch are specified, the lower strength weld metals will undermatch the highest strength parent
material in particular project fabrication programmes, leading to certain joints being undermatched.
The importance of this effect is provoking considerable interest at the moment, particularly with
regard to the higher grade steels.
It has been reported that the residual stresses in restrained high strength steel joints are less than those
encountered in lower strength steels which can have an important influence on subsequent fatigue and
fracture behaviour. Bennett et al [5.05] claimed that the residual stresses in a 40mm thick restrained
joint were only approximately one half of those obtained with mild steel and were largely independent
of heat input. The explanation of this effect was thought to be partial counterbalancing of the thermal
contraction stresses by the martensitic/bainitic transformation which occurred during cooling. The
potential benefit of this different behaviour seems not to have been utilised significantly to date.
Further work is needed to provide confidence in this approach.
A recent report details the welding practices and procedures used for welding the Elgin jacket [5.09].
High strength steels, Superelso 500 and Superelso 600, were used on the project in leg chords. These
materials have specified minimum yield strengths of 450MPa and 550MPs and yield ratios of 0.78
and 0.80. Significant use was made of MMA welding during the fabrication because of its fully
positional welding on site capability. Oerlikon electrodes, Tenacito 70, were used which gave yield
strengths between 490 and 550MPa and excellent Charpy V-notch toughness values of 130J at -40C.
The welding of the chord to the prefabricated 160mm thick rack sections was carried out with a
minimum preheating temperature of 150C followed by a dehydrogenation treatment of 2 hours at
200C. Very low repair rates (<1%) were reported for the project.
REFERENCES
5.01 Webster S, Structural materials for offshore structures past, present and future, IBC
Conference The Safe Design and Fabrication of Offshore Structures, Sept 1993, publ IBC
5.02 Lessels J, Rohde W, Pontermoli M, and Devillers L, The present state of technical knowledge
of offshore structural steels and future material requirements, May 1992, CEC 7210.ZZ466
5.03 Private communication, F. Forster, AMEC
5.04 Rogers K T, Thornton C E, and Naylor K D, Progress in the welding of higher strength steels
for offshore applications, OMAE Conf. Vol.III, 1989, 309
5.05 Bennett W T, Cadiau L, and Caudreuse L, Steels for jack-up legs in Recent Developments
in Jack-up Platforms, ed Boswell L F, and DMello C, 1992
5.06 Krancioch J J, Fabrication of offshore structures and potential productivity increases, The
Safe Design and Fabrication of Offshore Structures, Sept 1993, Publ. IBC
32
5.07 High Strength Steels in Offshore Engineering, MTD Publication 95/100, ISBN 1-870553-21-
7, 1995
5.08 Billingham J, Blackman S, and Norrish J, Further assessment of high strength weld metals
for use in offshore engineering applications, Cranfield Final Report, January 1998
5.09 Bews R, MMA Welding Gas from Strength to Strength Welding and Materials Fabrication,
July/August 2000.
33
Figure 5.1
Criteria of Steel Weldability Cracking Susceptibility
34
6. TOUGHNESS
Toughness may be loosely described as a measure of the resistance to failure in the presence of a
crack, notch or similar stress concentrator. High toughness therefore is generally recognised as a
desirable property for offshore steels.
A high toughness material is one where a considerable amount of plastic deformation is required at
the crack tip before the crack can be made to advance. Conversely, if the application of stress causes
the elastic failure of atomic bonds at the crack tip, relatively little energy of deformation is involved,
and the result is a brittle fracture.
The word toughness is used for two quite separate quantities. They are more correctly described as
Impact Toughness and Fracture Toughness.
Impact toughness is an energy measurement (Joules, or ft-lbs) and commonly relates to the Charpy V-
notch test. Fracture toughness is a calculated value for the critical stress intensity factor (N.m
-3/2
or
MPam or psii ) assessed for ductiIe materiaIs from crack-tip-opening-displacement (CTOD) tests or
J-integral tests.
Relationships between these quantities are empirical. The relationships have been well validated over
many years for structural steels in moderate section thickness. This has permitted the more readily
available Charpy impact data to be used as an indicator to the adequacy of the fracture toughness.
Where the same relationships between impact testing and fracture toughness are extended to thick-
sectioned material and to high strength steel, specifications should be viewed with caution until it has
been demonstrated that adequate factors of reserve are incorporated for the new conditions. Direct
testing for fracture toughness may be preferable, particularly since it can reduce some of the
uncertainties related to the effect of material thickness.
In ferritic steels, the fracture toughness is affected by temperature, by strain rate and by geometry.
The latter influence is also known as the stress state, the degree of triaxiality or the thickness
effect. The apparent changes in toughness that result from the geometry are not quantified at all by
the Charpy test, which always uses a standard small (10mm thick) specimen. Despite this, Charpy
results are widely used in materials selection and in current codes and standards.
6.1 DUCTILE TO BRITTLE TRANSITION
Both the impact toughness and the fracture toughness of ferritic steel are characterised by a ductile-to-
brittle transition as the temperature is reduced. This corresponds with a change in the mechanism of
crack movement, from plastic blunting and plastic tearing (ductile control) at the higher temperature
to cleavage (brittle fracture) at the lower temperature. The transition occurs over a relatively narrow
range of temperature, typically 30C, but often involves considerable experimental scatter. As a
result, there are several different definitions of the transition temperature from the same data. Low
transition temperatures and high upper-shelf values of toughness are seen as beneficial.
The transition temperature (TT) is not an invariant property of a given material, even for a fixed
composition, grain size etc. The TT varies with the state of stress (which means that it depends on the
size and geometry that has been used) and rate of loading. Increasing the thickness of the specimen,
or increasing the rate of loading, produce an increase in the TT.
Without these complications, it would be relatively simple to avoid brittle fracture - the basic
requirement would be to select material for which the TT was below the service temperature range. It
would still be necessary to take into account that fabrication processes such as welding affect the
35
material and its TT, both as a result of changes in the material from the thermal cycling and from the
introduction of flaws (that have the effect of increasing the TT). It would also be necessary to ensure
that the results related to the correct environmental exposure.
Unfortunately, the commencement and the extent of plastic deformation at the tip of a crack are
significantly affected by the geometry. At a particular temperature, a thin-section low-speed fracture
mechanics test may exhibit ductile behaviour whereas a thick-section test may give a brittle fracture,
even where samples have been cut from exactly the same block of material. In structural applications,
a known thickness of material may be selected, but complex joint geometry or complex stress fields
may again influence the balance between ductile deformation and brittle fracture at a crack tip.
Operating at a temperature above the ductile-to-brittle transition temperature therefore does not
automatically guarantee the avoidance of brittle fracture in a structural component.
6.2 CHARPY V-NOTCH VALUES AND HIGH STRENGTH STEEL
The Charpy V-notch impact test is probably the best-known of several small-scale tests designed to
study the resistance of the material to impact loading in the presence of a standardised stress
concentration, by recording the energy absorbed from the pendulum by the specimen. Charpy results
however cannot be considered to be directly relevant to structural behaviour.
Much of the early experience with Charpy specifications relates to steel that was produced by the
normalisation route, which gave a fine-grained ferrite and pearlite microstructure, but there was some
risk of producing the more brittle martensite microstructure in the heat affected zones of welds. By
demonstrating that the product avoided brittleness in a low temperature Charpy test (by absorbing at
least 27J of energy for example), the implication was that those microstructures that were most at risk
of undergoing fracture at the operating temperature were absent. The speed of the impact test was
regarded as severe and likely to promote brittleness. The influence of thickness (geometry, stress
state) on the transition temperature was not overlooked, but it was addressed by means of applying a
temperature shift. Instead of seeking a certain minimum Charpy energy at the design temperature,
these values were required at a lower temperature that was adjusted according to the thickness of the
material in the structure.
The offshore industry applied this experience to normalised steel with a SMYS of 355MPa and sought
improved toughness levels. Steels made by various thermo-mechanically controlled processing
(TMCP) techniques were introduced because of their improved weldability. Products with SMYS
around 450MPa were produced specifically for offshore structural use and generally exhibited
excellent impact toughness and low transition temperatures (frequently below -80C). Higher strength
steel is usually manufactured using the quench-and-temper (Q & T) route. The quench produces a
finely-structured martensite or bainite product that is usually stronger than required and (except in
very low carbon steels) is usually too brittle for direct use. The tempering reduces the strength,
relieves some of the residual stress and improves the impact toughness. More recent developments in
the processing try to avoid the intermediate production of very brittle phases.
For simplicity, the design assumption that higher strength steels carry a proportionally higher load led
to the R
e
/10 requirement so that the required longitudinal impact toughness in Joules is equal to one
tenth of the SMYS in MPa. The required transverse value is usually taken as two thirds of this to
allow for anisotropic effects in the microstructure.
It is not clear, however, whether the same temperature offset between the design temperature and the
impact test temperature is equally applicable to the higher strength steels. These comprise different
microstructures, finer grain sizes, differing degrees of scatter and much higher upper shelf energy
levels than the materials that were originally involved in the validation tests. Figure 6.1 shows
Charpy data for four modern high strength steels from one manufacturer. It illustrates the differences
between these steels in terms of the upper shelf Charpy value and the steepness of the ductile-to-
36
brittle transition temperature. The transition temperature also varies considerably, and the graph
shows the difficulty in defining this value for some of the steels.
A 1966 appraisal of the use of high strength steels in offshore installations addressed the properties of
offshore steels with a yield strength of 450MPa and above and jack-up steel data in some detail [6.01].
The paper includes a histogram of fusion line Charpy data at -40C for Grade 450EMZ steel for weld
heat inputs of 0.8kJ/mm and 3.0kJ/mm to illustrate that good impact toughness levels can be achieved
in high strength steels. About 80% of the results absorbed more than 150J of energy. Comparable
fracture toughness data give about 85% of CTOD results above 0.5mm, which should imply very
acceptable defect tolerance levels. In addition, Charpy scatter-band data for modern, low carbon, low
alloy, roller quenched and tempered (RQT) steels are contrasted against those for older versions of the
same material to show the improvements from modern chemistries and production. Between +20C
and -60C, absorbed energy levels had been increased by approximately 2 times as a result of
modern practices. A selection of impact data for heat affected zone and steel plate with yield
strengths to 765MPa is included in the paper. Most of the steels are in the yield strength range 450
580 MPa although there are limited results from the higher strength steels.
6.3 FRACTURE MECHANICS VALUES
Fracture mechanics values of toughness are linked to the size of flaw in the structure that is critical
under the applied stress. In one of its simpler forms, the relationship is expressed as
C C C
.a . K . -
where K
C
is the fracture toughness, / is a geometry correction, $ is the structural stress, a is the
parameter that relates to the crack size and the suffix C indicates critical conditions for the initiation
of crack movement.
Reserve factors are usually applied to set limits for repair of detected flaws before this critical size is
attained. In addition, flaw growth rates by sub-critical mechanisms (such as fatigue, stress corrosion,
etc) are used in conjunction with flaw detection methodology to determine a commensurate flaw
inspection schedule. Fracture toughness therefore is an important aspect of material selection, and one
that is extensively incorporated into codes and standards for medium strength steels.
Fracture testing can be performed on the full thickness of the structural material, but uncertainties will
still arise where the geometry is complex and where the residual stresses are difficult to define.
Environmental influences also must be taken into account.
The user needs to be clear on the terminology that is used in fracture toughness. Lower-bound values
for toughness are identified as K
IC
. If there is an environmental influence, for example producing
stress corrosion cracking, the toughness may be further reduced to K
ISCC
. The use of the Roman I in
these formats signifies two important facts about the value that plastic deformation is effectively at a
minimum and that it is opening mode loading (modes II and III exist, but mode I values are usually
the lowest).
K
IC
is called the plane strain fracture toughness because the crack tip behaviour is dominated by the
elastic loading condition of plane strain and this has the effect of restricting plastic deformation. A
standard fracture mechanics specimen usually has a geometry in which the stress state is determined
by the thickness (B). Plane strain dominates when B is large. Theoretically, the minimum toughness
relates to infinitely thick specimens, but the engineering approximation K
IC
is given where the
specimen thickness satisfies the inequality
37
2
YP
IC
2
1
K
. 2 B
'
'
(
)
*
*
+
,
0
where $
YP
is the yield or proof stress.
K
IC
will change when the temperature or strain rate is altered. The disadvantage of using K
IC
for
calculations when thinner sections are used is simply that the values may be excessively conservative,
requiring applied stresses to be limited, or very small cracks to be repaired.
If a toughness value has been obtained from a sample that is not thick enough to qualify as plane-
strain-dominated, the toughness is written as K
C
. It is then called the fracture toughness, but it is
important to realise that this number now also depends on the thickness that was used in the test.
[Figure 6.2] K
C
will also change when the temperature or strain rate is altered. It is potentially
dangerous to measure the fracture toughness in a thin specimen and use this toughness value in the
design of a structure that is made from thicker material.
Standardised fracture mechanics specimens are designed to produce a high degree of constraint in
order to promote lower-bound values. Fracture toughness tests relate to initiation of crack movement
and such tests may detect locally arrested cracking as pop-in events when the load-displacement
trace shows a discontinuity. It then may be debatable whether the initiation event is significant, i.e.
whether a similar crack initiation would have continued to propagate in the structure. Guidance on
interpreting pop-in events from tests may be found in BS 7448:Part 1:1991 [6.02], whereas Part 2
deals specifically with fracture toughness evaluation for welds [6.03].
6.4 FRACTURE TOUGHNESS OF HIGH STRENGTH STEELS
In general, recent modern steel-making developments for structural steels that produce ultra-fine
grained low alloy products have much-improved upper-shelf toughness and significantly lower
transition temperatures for the same thickness compared with the older products of comparable
strength.
Refining the grain size is the best option for increasing the strength, because toughness is improved at
the same time, but grain growth controllers are required in weldable steels. Increasing the alloy
content of steel to increase its strength tends to reduce toughness, hence steel-makers offset this effect
by grain size control during the manufacturing.
Weld metals for use with high strength steel are required to show comparable levels of strength and
therefore have to develop strong tough structures on solidification. Ultra-high strength steels are also
likely to rely more heavily on increased levels of alloying to achieve the required strength, as an
economic limit for grain refinement is reached. Very high toughness, therefore, may be less readily
achieved in these two cases. Selection should therefore be based on adequate levels of toughness for
the purpose, rather than absolute values.
Local precipitation changes in the HAZ regions of welds, particularly in more highly alloyed steels,
may produce scatter in fracture toughness values. Regions causing this scatter have been called local
brittle zones or LBZs but this term also applies to effects caused by local grain coarsening.
Recent weldability tests on Thyssen steels up to 690MPa utilised Charpy and CTOD tests [6.04].
Welded high strength steels exhibited satisfactory performance to 690MPa, with TMCP and
accelerated cooled steels being especially good. For Q & T steels above 690MPa, it was suggested
that high weld heat input should be restricted and fracture toughness should be checked.
Studies were performed in the early 1990s for jack-up steels [6.05] by Creusot-Loire Industrie, after
Friede and Goldman Ltd pointed out that steel specifications and production techniques had evolved
over time for that particular application. The pioneering jack-up designs used 275-350MPa material
38
for chords and 205-250MPa for bracing members, whereas these had evolved to 520-690MPa and
345-550MPa in the 1980s and were predicted to rise to 750-900MPa and 550-690MPa respectively.
The Creusot-Loire steel (A517Q mod) had a typical actual yield stress of 770MPa and YR of
0.91, with Charpy levels of 70J at -60C. Welding preparation was regarded as very important.
Correlation between Charpy and CTOD (1) were investigated. CTOD tests gave values of 0.15mm
for 1
u
at -20C for 180mm thick plate and HAZ, and the steel met or exceeded the criteria of the jack-
up industry.
Although thick sections are capable of developing plane strain conditions making it difficult for
plasticity to occur at the crack tip, it does not mean that plane strain conditions always apply in thick
sections. A surface crack that is long but not very deep may not experience constraint of yielding,
hence its apparent toughness may be significantly above K
IC
. In addition, when thick sections are
assessed by standard fracture toughness tests, the crack front in the specimen may be appreciably
larger that that of any tolerable flaw in the structure. It therefore may be argued that standard tests
where cleavage occurs produce pessimistic values. Code requirements for applications such as racks
in jack-up structures, where yield strength is typically 690MPa or above, may need to take such
factors into consideration to avoid being unduly conservative [6.06].
The 1996 paper by Stacey, Sharp and King [6.01] includes some CTOD data for a Creusot Loire
A517F steel (736MPa) and a Nippon Welten 780 (765MPa) steel. These were tested at -15C both as
30mm fracture toughness samples and in their full thickness, which ranged from 170mm to 250mm.
For the 18 large-scale tests, the CTOD values reached at least 1.00mm (which was the nominal
capacity of the test facility) in all cases, whereas the 30mm samples gave CTOD result ranges of 0.22
0.81mm and 0.15 0.27mm respectively. This is apparently contradictory to the effect of increased
thickness. However, the lower results from the smaller specimens were attributed to the influence of
specimen size on the inhibition of the plastic deformation in ductile process of slow stable tearing. In
the DnV tests that were being reported [6.07], only maximum load values were returned from the full-
sized specimens (from the few tests that attained this before exceeding the machine capacity), whereas
a detected instability in crack tip tearing in the smaller specimens requires CTOD calculations based
on that event. An apparent complication arises from comparing 1
u
and 1
m
results and the work may
also reflect the difficulty of detecting small failure events in large-scale specimens.
In work [6.08] at DERA, comparisons were made among the ductile-to-brittle transition temperatures
from three different tests:
(i) conventional Charpy V-notch samples (a/W = 0.2),
(ii) Charpy samples where constraint had been increased by extending the notch with
a narrow spark-eroded slot (a/W = 0.45), and
(iii) full-thickness J-based tests (a/W = 0.3) generally on 50-60mm thick samples.
The main series of tests comprised 3%Ni Q & T steels or boron-treated Q & T steels with yield stress
in the range 550-700MPa, but the work included a 600MPa weld metal and a 300MPa C-Mn steel.
The comparisons of Charpy-sized tests showed that the C-Mn steel and weld metal have relatively
little notch acuity shift, whereas this could be up to 50C for high strength steels, and up to 80C
between Charpy and full thickness dynamic tests. Attention was drawn to the practice of requiring
Charpy tests at only -40C for selection of high strength steel at design temperatures around 0C
[6.09]. It was concluded that reliance on the conventional Charpy test for fracture avoidance could be
dangerous for high strength steels and weld metals, especially if the criteria were based on those for
lower strength C-Mn steels without recognising the possible differences in behaviour.
For ultra-high strength steels, alloying additions are likely to raise the transition temperature. Hence,
it may be unwise just to rely on upper-shelf fracture toughness values from a particular test
temperature (related to the design temperature) even if these toughness values seem very high. It is
more prudent to investigate where the transition temperature actually lies in relation to the test
temperature. The same doubts on the temperature offset apply to the nominal correlation between
fracture toughness and a required energy level in Charpy tests (27J, R
e
/10 etc). In assessing such
39
correlations, it should be borne in mind that mechanical properties such as the actual yield stress
frequently vary considerably from batch to batch, according to the production process and the
manufacturer, and this will affect the amount of crack tip plastic deformation. When such influences
need to be assessed, a clearer picture is likely to emerge if Charpy and fracture toughness test results
from the same batch are used in the correlation. Repeat tests should be done as required on other
batches to determine the extent of scatter from variations in materials properties.
6.5 FLAW ASSESSMENT CONSIDERATIONS FOR HIGH STRENGTH STEELS
The main purposes for performing fracture mechanics tests are to determine at the design stage how
big a flaw would cause a problem, and after construction whether a detected flaw needs to be repaired.
These are based on the calculation of the critical effective crack length parameter, (
eff
)
CRIT
from
fracture mechanics. Reserve factors give the tolerable flaw sizes.
The growth of the flaw from mechanisms including fatigue, corrosion fatigue etc must be considered.
Knowledge of the predicted flaw growth rates allows sensible inspection intervals to be fixed.
Calculations must check that the tolerable size is not exceeded before any necessary remedial action
can be carried out. At the design stage, this is based on the largest size of flaw that could escape
detection. In service, it is based on the current size of a detected flaw.
The procedures for measuring the fracture toughness and for calculating the tolerable flaw size
parameter are given in BS 7448:Part1:1991 Fracture mechanics toughness tests : Part 1 : Method for
determination of K
IC
, critical CTOD and critical J values of metallic materials and BS 7910:1999
Guide on methods for assessing the acceptability of flaws in fusion welded structures[6.02; 6.10]. A
separate standard now deals with the determination of plane strain fracture toughness values [6.11].
Taking critical conditions in plane stress but dropping the suffixes for clarity here, it may be shown
that the basic relationships give
2 3
2 3
eff
2
app
2
YP
2
a
.
K
and . J
E
K
- 4 4
If using CTOD, it is important to notice that it is the critical value of ($
YP
.1) that relates to fracture
toughness K
C
and not just the displacement 1
CRIT
.
When a designer contemplates moving to a higher strength of steel, it is not immediately obvious
whether the toughness requirement also need to be increased, should stay the same, or be relaxed. It
is affected by several related factors. The stronger steel is usually chosen to allow weight reduction,
so that less steel supports the same load, but the design stress is often increased proportionally (to be a
set fraction of the yield stress). This gives maximum weight-saving, but there are two disadvantages;
the toughness must show a corresponding increase to prevent the critical flaw size going down, and
also the new steel must be more resistant to crack growth mechanisms such as fatigue because the
working stress will have increased. Although modern high strength steels do have improved fatigue
resistance compared with conventional offshore steels now in service, it is usually not commensurate
with the increase in strength.
An alternative approach, currently under-used, is to design on the basis of the improved resistance to
crack growth. This should permit higher structural design stresses than currently employed, with
some reduction in weight from reduced section thickness. The improvement implies working with the
higher strength steels at a design stress that is a lower fraction of the yield stress. Further guidance on
this is given in Appendix 6.
Despite the above complications in obtaining the fracture toughness value that characterises the
particular geometry and loading conditions, fracture mechanics is an extremely useful tool in practice.
It is relatively simple to obtain fracture toughness values that are predicted to be conservative in use.
40
The degree of conservatism may be open to debate, but the information can be used
(i) at the design stage (to evaluate tolerable flaw sizes related to the difficulty of
inspection and reliability of the inspection techniques), [6.12],
(ii) during service (to assess the significance of any crack growth and to set
inspection periods) and
(iii) as part of the contingency planning (regarding the stability of cracks following
accidents).
It is therefore advisable to derive the toughness requirements from the flaw acceptability limits for
each structure using recognised procedures and to perform a sensitivity analysis on the outcome.
A common European structural integrity assessment procedure (SINTAP) is being developed.
because of the diversity of fracture mechanics data [6.13] and the need of industrial users to find a
reliable correlation with Charpy impact energy data [6.14]. The methodology uses a fracture
toughness parameter K
mat
to characterise the particular material, and a probability distribution P{K
mat
}
that enables the confidence level of the assessment to be quantified. Where Charpy data are being
used, lower-bound correlations are used for lower shelf and upper shelf behaviour, and a master
curve correlation based on statistics is used in the transition region. The procedure uses the 27/28J
Charpy transition temperature, the 100MPam fracture toughness temperature and K
mat
for 25mm
thick specimens, with formulae to correct for the design temperature and the appropriate structural
dimensions. For confidence in data input, the verification calculations show that the results from six
tests should give a 75% probability of having a conservative mean fracture toughness value [6.15].
However, the procedure is not being written specifically for high strength steels with SMYS above
450MPa. The procedure appears promising [6.16], but more validation may be needed for the
stronger steels. There has been a contemporary study of the fracture properties of ship steel plates
[6.17], and renewed interest in statistical assessment of Charpy data [6.18].
6.6 SUMMARY OF TOUGHNESS CONSIDERATIONS
The word toughness is used for two quite separate quantities, the Impact Toughness (Charpy
absorbed energy value) and the Fracture Toughness (Critical stress intensity factor). Relationships
between these quantities are empirical, but have been well validated over many years for structural
steels in moderate section thickness, permitting Charpy impact data to be used to assess lower-bound
fracture toughness values.
Operating at a temperature above the ductile-to-brittle Charpy transition temperature does not
automatically guarantee the avoidance of brittle fracture in a structural component. Charpy results
cannot be considered to be directly relevant to structural behaviour. The apparent changes in
toughness that result from the geometry and strain rate are not quantified at all by the Charpy test,
which always uses a standard small (10mm thick) specimen under dynamic loading. Many codes deal
with these effects by specifying a difference between the Charpy test temperature and the design
temperature of the structure. To extrapolate the relationship between Charpy values and fracture
toughness to (moderately) stronger steels, some codes increased the required minimum-absorbed-
energy value according to the ratio of the yield stress values (on the basis that the design stress limit is
a fixed fraction of the SMYS value). This was soon simplified to the R
e
/10 criterion.
Modern steel-making techniques and processes have extended the range of tough weldable steels
upwards in terms of strength, and have produced significant improvement in the toughness levels of
steels, particularly at the lower end of the high-strength range. This raises doubts about the
applicability, relevance and conservatism of recommendations and limits in those codes of practice
written for the older steels and validated by data from them. It also raises questions about the
extrapolation of limits to deal with higher strength steels.
41
Modern high strength steels often use grain size refinement to increase the yield strength and have
grain-growth controllers to maintain the properties as much as possible in the heat affected zones of
welds. Usually, the toughness of these steels is also excellent, since fine grain size enhances this
property too.
For ultra-high strength steels, extra alloying additions are needed to give the strength and these are
likely to raise the ductile-to-brittle transition temperatures. Weld metals for use with high strength
steel are required to show comparable levels of strength and therefore have to develop strong tough
structures on solidification, which also requires alloying additions. Very high toughness, therefore,
may be less readily achieved in these two cases.
While the strength of steel has increased, it is not clear whether the same temperature offset (to take
account of different structural thickness) between the design temperature and the Charpy impact test
temperature is equally applicable to the higher strength steels. The same doubts apply to the nominal
correlation between fracture toughness and a required energy level in Charpy tests (27J, R
e
/10 etc).
For 690MPa steel, IACS has agreed on a requirement of 69J at 605C, as already detailed in section 4.
Where the same relationships between impact testing and fracture toughness are extended to thick-
sectioned material and to high strength steel, specifications should be viewed with caution until it has
been demonstrated that adequate factors of reserve are incorporated for the new conditions. A new
European structural integrity assessment procedure is incorporating a statistical approach so that the
effect of these uncertainties on the reliability of the assessment can be quantified, but it is not
specifically aimed at high strength steel.
Direct testing for fracture toughness may be preferable to Charpy correlation, particularly since it can
reduce some of the uncertainties related to the effect of material thickness.
Because of the changes in both the position and steepness of the ductile-to-brittle transition curve for
the fracture toughness, it may be unwise to rely on a single point - especially a Charpy point - to
characterise the transition. It is more prudent to investigate where the fracture transition temperature
actually lies in relation to the design temperature.
Fracture testing can be performed on the full thickness of the structural material, and hence improve
confidence in the relevance of the numbers used as input to any assessment. Some uncertainty in
applying the results will still arise where the geometry is complex and where the residual stresses are
difficult to define.
Designing from K
IC
values is an option, but the recommendations are likely to be conservative for
thin-sectioned high-toughness steel. Although thick sections are capable of developing plane strain
conditions making it difficult for plasticity to occur at the crack tip, especially for higher strength
steel, it does not mean that plane strain conditions always apply in thick sections. A surface crack that
is not very deep may not experience constraint of yielding, hence its apparent toughness may be
significantly above K
IC
.
Environmental influences must be taken into account. K
ISCC
for example may be appreciably lower
than the K
IC
value.
Because of the toughness improvements that have accompanied increases in steel strength, it is
possible that the limitations to the use of high strength steel will shift from fracture behaviour,
towards buckling and crack extension from mechanisms such as fatigue. This means that it may be
necessary to sacrifice the full capabilities of increasing the design stress in a structure in line with the
increase in SMYS. Instead, the applied stress could be increased in line with the more moderate
improvements in fatigue resistance etc, which may still give appreciable benefits in weight-saving and
reduced fabrication costs, as well as enhanced resistance to fracture.
42
REFERENCES
6.01 Stacey A, Sharp J V and King R N, High strength steels used in offshore installations, Proc.
15
th
Intnl. Conf. OMAE, Florence, June 1996, ed. M.M. Salama et al, publ. ASME, Vol. III,
pp 417-433
6.02 BS 7448: 1991 Fracture mechanics toughness tests Part 1: Method for determination of
K
IC
, critical CTOD and critical J values of metallic materials, British Standards Institution,
London
6.03 BS 7448: 1997 Fracture mechanics toughness tests Part 2: Method for determination of
K
IC
, critical CTOD and critical J values of welds in metallic materials, British Standards
Institution, London
6.04 Kaiser H-J, Kern A, Schriever U and Wegmann H, Fracture toughness of modern high-
strength steel plates with minimum yield strength up to 690MPa. In: Proceeding of
International Symposium Safety in application of high strength steel, Trondheim, Norway,
1-2 July 1997, publ. Statoil, Norway
6.05 Bennett W T, Cadiou L and Coudreuse L, Steels for jack-up legs. In: Recent
Developments in Jack-up Platforms, ed. Boswell L F and DMello C, Blackwell Scientific
Publications, Oxford, 1992, Ch 12, pp283-312
6.06 Offshore Technology Report OTN 96 127, Toughness acceptance criteria for rack materials
in jack-ups, Health and Safety Executive, July 1996 [Restricted]
6.07 Det norske Veritas Report 95-3056, Toughness acceptance criteria for rack materials and
weldments in jack-ups, 1995, publ. Det norske Veritas, Oslo, Norway
6.08 Sumpter J D G, Fracture safety of high strength steels. In: Proceeding of International
Symposium Safety in application of high strength steel, Trondheim, Norway, 1-2 July 1997,
publ. Statoil, Norway
6.09 IACS Requirements, 1994 Rev 1 Section W16, High Strength Quenched and Tempered
Steels for Welded Structures
6.10 BS 7910:1999 Guide on methods for assessing the acceptability of flaws in fusion welded
structures, British Standards Institution, London
6.11 BS EN ISO 12737:1999, Metallic materials: Determination of plane-strain fracture
toughness, British Standards Institution, London
6.12 Cotton H J, Application of Fracture Analysis in the Design of Fixed Offshore Structures. In:
Materials Technology symposium, Proc. 19
th
Intnl. Conf. OMAE, New Orleans, USA,
February 2000, publ. ASME
6.13 Wallin K and Nevasmaa P, Structural Integrity Assessment Procedures for European
Industry (SINTAP) - Sub-task 3.2 report : Methodology for the treatment of fracture
toughness data procedure and validation, Report VAL A: SINTAP/VTT/7, VTT
Manufacturing Technology, Espoo, Finland, 1998
43
6.14 Bannister A C, Structural Integrity Assessment Procedures for European Industry (SINTAP)
- Sub-task 3.2 report : Determination of fracture toughness from Charpy impact energy
procedure and validation, Report SINTAP/BS/15, British Steel plc, Rotherham, UK, 1997
6.15 Nevasmaa P, Bannister A C and Wallin K, Fracture toughness estimation methodology in the
SINTAP procedure. In: Proc. 17
th
Intnl. Conf. OMAE, Lisbon, June 1998, Paper
OMAE98-2053, publ. ASME
6.16 Webster S E and Bannister A C, Structural Integrity Procedure for Europe (SINTAP): The
Complete Story. In: Proc. 18
th
Intnl. Conf. OMAE, St Johns Newfoundland, July 1999,
Paper OMAE99-MAT2040, publ. ASME
6.17 Bannister A C and Stacey A, Literature Review of the Fracture Properties of Grade A Ship
Plate. In: Proc. 18
th
Intnl. Conf. OMAE, St Johns Newfoundland, July 1999, Paper
OMAE99-MAT2071, publ. ASME
6.18 Minami F, Jida M, Takahara W, Kondo N and Arimochi K, Fracture Mechanics Analysis of
Charpy Test Results based on the Weibull Stress Critereion. In: Proc. 20
th
Intnl. Conf.
OMAE, Rio de Janeiro, June 2001, Paper OMAE 2001/MAT-3130, publ. ASME
6.19 Hancock P, Spurrier J and Chubb J P, The problems of weld metal or heat affected zone
toughness in offshore structural steels, Proc. 15
th
Intnl. Conf. OMAE, Florence, June 1996,
ed. M.M.Salama et al, publ. ASME, Vol.III, pp333-339
44
Figure 6.1 Charpy V-notch transition temperatures for some modern Thyssen steels
[6.04]
-150 -100 -50 0 50
100
0
200
300
S355M
S355N
S460M
S690QL1
Temperature, C
Charpy Impact Energy
(Transverse), Joules
Plate thickness : 15 - 30mm
Testing position: t / 4
45
Figure 6.2 Crack tip opening displacement transition temperatures for geometrically
similar samples from the same plate of steel, with crack tips at constant depth
in the plate [6.19]
-200 -180 -160 -140 -120 -100 -80 -60 -40 -20 0 20
2.8
2.4
2.0
1.6
1.2
0.8
0.4
0
Crack Tip Opening Displacement (1) mm
Temperature, C
6 mm
18 mm
36 mm
60 mm
75 mm
46
7. FATIGUE IN HIGH STRENGTH STEELS
7.1 INTRODUCTION
Design to resist fatigue is recognised as one of the main requirements for offshore structures,
particularly for welded tubular joints in seawater and subject to high stress concentrations. The
procedure is well established for medium strength steels, with appropriate S-N curves published in
design codes. However, the effect of seawater on the fatigue performance is considered to be more
detrimental for high strength steels because of their greater susceptibility to hydrogen cracking. This
susceptibility is known to increase with increasing yield strength and more negative cathodic
protection (CP) potentials. Hydrogen generation from CP can enhance crack growth rates at the crack
tip, leading to overall shorter fatigue lives. Hence, there is a need to understand the fatigue process in
welded high strength steels. However, there is a very limited amount of data for the higher strength
steels which makes it difficult to provide design information (see section 13).
Corrosion fatigue is a major cause of failure in marine structures with most of the fatigue life being
taken up in fatigue crack propagation. The corrosion fatigue behaviour of welded joint constructions
from structural steel conforming to BS4360:50D has been the subject of many studies over the years
[7.01]. As a result, the understanding of joint geometry, seawater environment and CP has reached a
level where confident predictions on fatigue resistance behaviour can be made for this type of
steel/structure.
In addition to S-N type data there is a need to estimate the remaining lives of components containing
cracks, using fracture mechanics. There is an increasing amount of data on crack growth rates
(da/dN) versus stress intensity factor range (6K), for all types of steel, both in-air and seawater, which
has enabled the relevant design code (BS 7910) [7.02] to cover steels with yield strengths up to
600MPa. However, for even higher strength steels, the data, particularly in seawater under CP,
remain limited and special approaches are recommended, usually based on test data for the steel in
question.
7.2 FATIGUE CRACK PROPAGATION
7.2.1 Effect of steel strength on fatigue crack growth rate
King [7.03, 7.04] carried out a review of literature FCGR for structural and engineering steels with
yield strengths up to 1000MPa for tests in-air and seawater environments (free corrosion and CP
levels of -800 and -1050/1100mV(Ag/AgCl)). The data were split into two strength ranges of 450
and >450MPa and R ratios of 0.5 and 0-0.1 and were also limited to cycling frequencies of <1Hz.
For each data set the intercept (C) and gradient (m) values from the mean linear regression line fitted
to the Paris Law were calculated. The standard deviation was calculated for each curve and then used
to calculate C values corresponding to the 72 standard deviation (72SD) curves. Stage 1 and Stage 2
growth were allowed for.
The data showed that there was no obvious effect of yield strength on fatigue crack growth rates
(FCGR) even for seawater with CP. This was an unexpected finding as the higher strength steels are
commonly expected to have a greater susceptibility to hydrogen effects than the lower strength steels.
The review only considered parent plate material and therefore it is suggested that caution should be
taken when applying the data to high strength steel heat affected zones (HAZ) and weld metals. In
addition, the effects of sulphides/SRB and hydrogen precharging were not included.
It was decided to use the 72SD lines produced by this review as the baseline for comparison of
FCGRs on the da/dN versus K pIots in this review. The Paris Law constants used and the crossover
points for stage 1 and 2 growth are listed in Table 7.1. Based on this review BS7910 has provided
data for steels with yield strengths up to 600MPa. Although the review covered even higher strength
steels, it was felt that the limited data available did not justify increasing the limit in BS7910 above
600MPa.
47
7.2.2 Parent Materials
The King [7.03, 7.04] review covered parent material and found that overall the behaviour of high
strength steels was similar to the medium strength steels under the conditions reviewed. It was
surprising that even with cathodic overprotection the behaviour was still similar. Several
fractographic studies have confirmed the change in crack propagation mode, from a ductile
mechanism with secondary cracking to quasi-cleavage fracture, at high negative CP potentials [7.01].
Healy [7.01]
reviewed the FCGR of high strength steels and found that the performance of the high
strength steel group was comparable to, if not slightly improved over, that of the structural grade
BS4360:50D steels. For the steels examined, no discernible effect of manufacturing process (i.e.
between Q&T and TMCP) was seen on the resultant FCGR.
7.2.3 HAZ
Very limited fatigue data have been generated for the HAZ of high strength steel weldments.
Representative HAZ FCGR data for steels in the strength range 500-600MPa are presented in Figure
7.1. The available data show that the fatigue performance of the HAZ is similar to that observed for
both the structural grade steel and the high strength parent plate data. Extensive metallographic and
fractographic studies have shown that high strength steels with yield strengths in the range 500-
600MPa are not susceptible to excessive hardening in the welded condition (HAZ hardness generally
<350Hv [7.05]). Additionally, the fatigue crack failure mechanisms in the HAZ are similar to those
observed in the parent plate, displaying a similar response to environmental test conditions. It would
appear that welding under controlled conditions does not significantly affect FCGR in these steels and
the data suggest that a preliminary screening of the corrosion fatigue behaviour on the parent plate
may be used to assess the suitability of the steel when welded.
7.2.4 Weld Metals
There is very little fatigue data available for high strength steel weld metals. Work at Cranfield
University [7.05, 7.06] has produced data for weld metals produced by SAW and FCAW in-air and in
seawater with CP. The data are shown in Figures 7.2 to 7.6 and shows that the behaviour of the weld
metals is comparable to parent materials. Tests were also carried out on a 450MPa steel (not shown in
the Figures) and these were comparable to the higher strength weld metals. No relationship between
yield stress and fatigue performance was found for the weld metals tested. Again it would appear that
welding under controlled conditions does not significantly affect FCGR in these steels and the data
suggest that a preliminary screening of the corrosion fatigue behaviour on the parent plate may be
used to assess the suitability of the steel when welded.
7.2.5 Fatigue Thresholds
Limited information has been reported on threshold stress intensity values for high strength steels
[7.01]. For low R ratios an apparent increase in threshold value occurs with increased levels of CP
compared with the in-air behaviour. Many studies have demonstrated that this behaviour is due to
crack wedging effects reducing the effective stress intensity range. Under conditions of high load
ratio, this mechanism is diminished and threshold values are similar to in-air levels. King [7.03, 7.04]
performed a review of fatigue threshold values for carbon and carbon manganese steels and the data
are shown in Figure 7.7. It was suggested that the existing recommendations for steels of yield
strength <600MPa in PD6493:1991 should be retained. For high strength steels there is insufficient
data to make additional recommendations.
7.3 EFFECT OF SRB AND SULPHIDES
Robinson [7.07] reviewed the literature data for sulphide/SRB influenced corrosion fatigue of high
strength offshore steels with yield strengths in the range 350 to 1010MPa that included parent material
and welds. The data were grouped on the basis of sulphide level, CP potential and R ratio and, for
each group, the stage 2 mean line was calculated (steels were also sub-grouped by yield strength).
These mean lines have been added to Figures 7.3 to 7.6.
The effect of sulphides/SRB on fatigue threshold values is unclear due to the lack of data. Some
workers [7.08, 7.09] have looked at near threshold values and found that the combination of CP and
48
sulphides/SRB gave threshold values comparable to or greater than those for in-air. It is thought that
the deleterious effect of increased hydrogen charging is balanced by a scale induced crack closure
effect reducing the effective K. WhiIst these resuIts are promising it shouId be noted that the above
tests were carried out under constant amplitude loading and that variable amplitude loading may
produce a different result.
It is clear from Figures 7.2 to 7.6 that the hydrogen uptake that results from the combined effects of
sulphides and CP has caused increased FCGR in the steels tested. The data suggest that saturated H
2
S
is the worst case but that much lower concentrations can also give significant increases in FCGR. The
effect of yield strength on FCGR in sulphide containing environments is not clear due to the lack of
data. Overall accelerated FCGR occur in both medium and high strength steels when exposed to
sulphide and higher rates are found with increasing sulphide concentration and/or more negative CP
potentials.
7.4 SN DATA
The amount of information available for steels with yield strengths <500MPa is considerable but
becomes more limited for steels with yield strengths 500MPa. Data for in-air tests and seawater
with applied CP are given in Figures 7.8 - 7.10. The data have been thickness corrected to 16mm
using a thickness exponent of 0.3.
Most of the available data are for in-air testing and include constant and variable amplitude testing of
tubular joints and plate specimens constructed from a range of steels and strength levels. In Figure
7.8, data that falls below the T curve (mainly 810 840MPa strength level) originate from plate
testing for which the appropriate curve is Class F2. It can be seen that all the plate data lies on or
above the F2 curve. The data available for tests in seawater with CP are very limited [7.10]. In
Figures 7.9 and 7.10 data falling below the air T curve are plate tests and therefore Class F2. At
present the data suggest that the fatigue performance of the higher strength steels is generally good
but more data are required for steels with applied CP.
7.5 FATIGUE IMPROVEMENT TECHNIQUES
Weld fatigue improvement methods can be divided into two main groups [7.11]. Firstly weld
geometry modification which removes toe defects and/or reduces the stress concentration, and
secondly, residual stress methods which introduce compressive stress in the area where cracks are
likely to initiate. The methods are summarised in Figure 7.11. Many tests have been carried out to
establish the gain in fatigue life as a result of using these methods. However most of these have been
on grade BS4360:50D type steels. As a result of these data, current codes and standards allow benefits
for some of these methods. One of the restrictions in providing benefits in codes is the quality control
aspect when using the technique in the field. Table 7.2 lists the improvement techniques which are
allowed in current offshore codes for medium strength steels. In most cases there are requirements to
be met to achieve the benefit (e.g. quality control, inspection, adequate CP).
For higher strength steels (>500MPa yield strength) there is a very limited amount of data. Work by
Bell et al [7.12] was carried out on steels with yield strengths of both 350 and 550MPa using T type
joints with longitudinal fillet welds. The thickness of the base plate was either 18 or 26mm. All the
tests were in-air, with fatigue lives up to 10
8
cycles. Specimens that had been hammer peened showed
a significant gain in life. In the case of the 550MPa steel the average gain was 175%. However
cracking in the hammer peened higher strength joints initiated at the root of the weld rather than the
weld toe (the location of initiation for the as-welded joints). Table 7.3 summarises other data
obtained for a number of improvement techniques for high strength steels. All the data are in-air and
show a range of improvements but unfortunately no tests were carried out in seawater.
At present there are insufficient data to demonstrate the benefits of improvement techniques for high
strength steels, despite the possible benefits. It would be necessary therefore for a potential user to
49
demonstrate any benefits using a test programme which represented the conditions in which the steel
would be used in service.
7.6 SUMMARY
The fatigue data available for parent and welded high strength steels indicate that the general
performance of the high strength steels is as good as the medium strength steels. The only condition
where poor performance was found was with H
2
S but medium strength steels also show similar poor
performance. Weld improvement techniques show promise but require data for typical offshore
conditions. In all areas more data are required before confident predictions of the fatigue performance
of high strength steels can be made. At the present time producing test data for candidate high
strength steels still appears to be the best approach.
REFERENCES
7.01 Healy J, Billingham J, Stacey A, Simpson R and Patel R Review of the corrosion fatigue
performance of medium to high strength structural steels Proceedings 15
th
International
Conference Offshore Mechanics and Arctic Engineering, Florence, Italy, 16-20 June 1996,
ASME, pp451-461
7.02 British Standard' Guidance on methods for assessing the acceptability of flaws in welded
structures, BS 7910, 1991
7.03 King R N A review of fatigue crack growth rates in air and seawater Offshore Technology
Report OTH96 511, HSE (1996)
7.04 King R N, Stacey A and Sharp J V A review of fatigue crack growth rates for offshore
steels in air and seawater environments Proceedings 15
th
International Conference Offshore
Mechanics and Arctic Engineering, Florence, Italy, 16-20 June 1996, ASME, pp341-348
7.05 Kilgallon P, Healy J and Billingham J The corrosion fatigue behaviour of high strength
steels and associated high strength weld metals. Offshore Technology Report, OTO97 065,
HSE (1997)
7.06 Billingham J, Blackman S and Norrish J Further assessment of high strength steel weld
metals for use in offshore engineering applications. Offshore Technology Report OTO98
091, HSE (1998)
7.07 Robinson M J and Kilgallon P J A review of the effects of sulphate reducing bacteria in the
marine environment on the corrosion fatigue and HE of high strength steels Health and
Safety Executive, Offshore Technology Report OTH 98 555, HMSO
7.08 Ferguson W G, Zhang Y , Stevens F J and Assefpour-Dezfuly M The effects of an anaerobic
environment on corrosion fatigue Proceedings of Chemical 90, 18
th
Australasian Chemical
Engineering Conference, 27-30 August 1990, Auckland, New Zealand, pp286-295
7.09 Rudd W J and Booth G S Near threshold crack growth in structural steels British Steel
Report FR 5330-8 942 May 1995
7.10 N Tantbirojn, L S Etube, W D Dover, P J Kilgallon, T Roberts and J Spurrier Variable
Amplitude Corrosion Fatigue of Jack-up Steels (VACF-T: Thick Plate Specimens), Final
Report, UCL NDE Centre, July 2000
7.11 Kirkhope K J, Bell R, Caron L, Basu R I and Ma K-T - Weld detail fatigue life improvement
techniques. Part 1: review. Marine Structures 12 (6), pp447-474 (1999)
50
7.12 Bell R, Militaru D V, and Braid J E M, The Fatigue Life Improvement of High Strength Steel
Welded Joints using hammer peening techniques, OMAE conference, 1995, Vol. III, ASME
7.13 Bignonnet A, Picouet L, Lieurade H P and Castex L The application of shot peening to
improve the fatigue life of welded steel structures Steel in Marine Structures. Developments
in Marine Technology, 3. Proceedings, 3rd International ECSC Offshore Conference (SIMS
'87), Delft, 15-18 June 1987. Ed: C.Noordhoek, J.de Back. Publ: 1000 AE Amsterdam,
Netherlands; Elsevier Science Publishers BV; 1987. ISBN 0-444-42805-4. Paper SIMS TS
33. pp669-678
7.14 Dept. of Energy, 'Offshore Installations - Guidance on Design, Construction and certification',
HMSO, London 1990
7.15 NORSOK, 'Design of Steel Structures' N-004, 1998
7.16 ISO ' Petroleum & Natural Gas Industries, 'Offshore Structures - Part 2'., ISO 13819-2, to be
published
7.17 Lopez Martinez L, Blom A F, Trogen H and Dahle T Fatigue behaviour of steels with
strength levels between 350 and 900 MPa - influence of post weld treatment under spectrum
loading Welded High-Strength Steel Structures. Proceedings, First North European
Engineering and Science Conference (NESCO 1), Stockholm, Sweden, 8-9 Oct.1997, Ed: AF
Blom, Engineering Materials Advisory Services Ltd pp361-376 (1997)
7.18 Agerskov H - Fatigue in steel structures under random loading. Journal of Constructional
Steel Research, 53 (3) March pp 283-305 (2000)
7.19 Agerskov H, Petersen R I and Lopez Martinez L Fatigue in high-strength steel offshore
tubular joints Tubular Structures VI, Proceedings of 6
th
International Symposium on Tubular
Structures, Melbourne, Australia 14-16
th
December 1994, pp527-534
51
Table 7.1 Data used to construct lines on Paris fatigue plots
m C
Me an +2 SD -2 SD
Stage 1/2 changeover 6K
Nmm
-3 / 2
Environment
(v. Ag/ AgCl)
R
St age 1 St age 2
St age 1 St age 2 St age 1 St age 2 St age 1 St age 2 Me an +2 SD -2 SD
0 0.1 8.16 2.88 1.21x10
-26
3.98x10
-13
4.37x10
-26
6.77x10
-13
3.35x10
-27
2.34x10
-13
363.1 314.9 418.8
Air
00.5 5.10 2.88 4.80x10
-18
5.86x10
-13
2.10x10
-17
1.29x10
-12
1.10x10
-18
2.66x10
-13
195.6 143.5 266.5
0 0.1 3.42 1.30 3.00x10
-14
1.27x10
-7
8.55x10
-14
1.93x10
-7
1.05x10
-14
8.36x10
-8
13336.0 993.1 1797.3
Fr ee cor r os ion
00.5 3.42 1.11 5.37x10
-14
5.67x10
-7
1.72x10
-13
7.48x10
-7
1.68x10
-14
4.30x10
-7
1097.8 747.7 1611.7
0 0.1 8.16 2.67 1.21x10
-26
5.16x10
-12
4.37x10
-26
1.32x10
-11
3.35x10
-27
2.02x10
-12
462.2 434.0 492.2
-850mV
00.5 5.10 2.67 4.80x10
-18
6.00x10
-12
2.10x10
-17
2.02x10
-11
1.10x10
-18
1.78x10
-12
322.9 289.9 359.6
0 0.1 8.16 1.40 1.21x10
-26
5.51x10
-8
4.37x10
-26
9.24x10
-8
3.35x10
-27
3.29x10
-8
575.6 513.8 644.8
-1050mV
00.5 5.10 1.40 4.80x10
-18
5.25x10
-8
2.10x10
-17
1.02x10
-7
1.10x10
-18
2.70x10
-8
516.7 414.9 643.4
Table 7.2 Increase on life allowed for weld improvement techniques in codes
Tec hni que HSE Gui danc e (7 . 1 4 ) NORSOK (7 . 1 5 ) Draft ISO s t andard (7 . 1 6 )
Weld pr ofilin g N/ A In cr ea s e by fa ct or of 2 on life N/ A
Weld t oe gr in din g In cr ea s e by fa ct or of 2. 2 on life In cr ea s e by fa ct or of 2 on life In cr ea s e by fa ct or of 2 on life
TIG dr es s in g N/ A In cr ea s e by fa ct or of 2 on life N/ A
Ha mmer peen in g To be demon s t r a t ed by t es t pr ogr a mme In cr ea s e by fa ct or of 4 on life In cr ea s e by fa ct or of 4 on life
52
Table 7.3 Summary of in-air fatigue improvement data (data were obtained using
constant (C) and/or variable(V) amplitude loading and improvement was calculated
from stress range (S) or cycles (N). * At 2 x 10
6
cycles
Improve me nt Te c hnique
St ee l
Type
Yie ld
St re ngt h
(MPa)
C/ V S/ N
Me an
Improve me nt
Fac t or (%)
Re f
Tig dr es s ed DOMEX 590 615 CV N 42 7.17
Tig dr es s ed WELDOX 700 780 CV N 73 7.17
Tig dr es s ed WELDOX 900 900 CV N 89 7.17
Sh ot peen ed E550 640 C S* 78 7.13
Ha mmer peen ed HY80 - - - 175 7.12
Sh ot peen ed Q&T 730/ 820 - S* 70 7.11
Ult r a s on ic peen in g WELDOX 700 780 CV N 79 7.17
Tig dr es s ed & u lt r a s on ic peen in g WELDOX 900 900 CV N 104 7.17
53
Figure 7.1 Summary diagram showing bound curves for HAZ
fatigue crack growth rates in 500-600MPa offshore steels
1.E-09
1.E-08
1.E-07
1.E-06
10 100
6 K (MNm
-3/2
)
d
a
/
d
N
(
m
/
c
y
c
l
e
)
-1100mV
Air
-850mV
54
Figure 7.2 Air fatigue data for RQT501 (open) and WX700 (filled) steels parent plate
and weld metals. Black symbols are SAW weld metals and red symbols are FCAW.
R 0.5
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1 10 100
6K (MNm
-3/2
)
d
a
/
d
N
(
m
/
c
y
c
l
e
)
Pa r en t pla t e
1-2KJ / mm weld met a l
3. 5kJ / mm weld met a l
55
Figure 7.3 Mean fatigue data for freely corroding steel in seawater
containing H
2
S at R 2 0.5 [7.07]
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1 10 100
6K (MNm
-3/2
)
d
a
/
d
N
(
m
/
c
y
c
l
e
)
Saturated H
2
S 500-976ppm H
2
S
370-598ppm H
2
S
86-184ppm H
2
S
73-290ppm H
2
S
56
Figure 7.4 Mean fatigue data for freely corroding and cathodically protected steel in
seawater saturated with H
2
S at R = 0 0.1 [7.07]. Comparison lines are equivalent
lines for tests without H
2
S
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1 10 100
6K (MNm
-3/2
)
d
a
/
d
N
(
m
/
c
y
c
l
e
)
57
Figure 7.5 Fatigue data for applied CP of 850mV(Ag/AgCl) for RQT501 (open) and WX700
(filled) steels parent plate and weld metals. Also shown is mean line for fatigue data for tests
with saturated H
2
S [7.07]
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1 10 100
6K (MNm
-3/2
)
d
a
/
d
N
(
m
/
c
y
c
l
e
)
Saturated H
2
S
Pa r en t pla t e
1-2KJ / mm weld met a l
3.5kJ / mm weld met a l
58
Figure 7.6 Fatigue data for applied CP of 1100mV (Ag/AgCl) for RQT501 (open) and WX700
(filled) steels parent plate and weld metals. R 2 0.5. AIso shown is mean Iine for fatigue data
for tests with H
2
S [7.07]
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1 10 100
6K (MNm
-3/2
)
d
a
/
d
N
(
m
/
c
y
c
l
e
)
Saturated H
2
S
330-410ppm H
2
S
2-150ppm H
2
S
Pa r en t pla t e
1-2KJ / mm weld met a l
3.5kJ / mm weld met a l
59
Figure 7.7 Fatigue thresholds (adapted from 7.03, 7.04)
0
2
4
6
8
10
12
14
16
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
R ratio
6
K
t
h
(
M
N
m
-
3
/
2
)
PD6493:1991
BS4360 Iyc sIrucIuraI sIccI
AII Icr r it ic Fy<600MPa
60
Figure 7.8 SN data for constant and variable amplitude fatigue tests in-air for parent
material and welded steel joints (thickness corrected to 16mm)
[7.10, 7.18, 7.19]
10
100
1000
1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08
Number of cycles, N
S
t
r
e
s
s
r
a
n
g
e
,
S
(
M
P
a
)
Tcurve (16mm) - air
Yield strength (MPa)
A 448 to 501
540 to 586
690 to 797
810 to 840
Class F2 (16mm) - air
61
Figure 7.9 SN data for constant and variable amplitude fatigue tests with applied CP
of 800 to 850mV(Ag/AgCl) on welded steel joints
100
1000
1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
Number of cycles, N
S
t
r
e
s
s
r
a
n
g
e
,
S
(
M
P
a
)
Yield strength (MPa)
690 to 797
Tcurve (16mm) - air
Tcurve (16mm) - CP
Class F2 (16mm) - CP
62
Figure 7.10 SN data for constant and variable amplitude fatigue tests with applied
CP of 1000 to 1050mV(Ag/AgCl) on parent material (shaded) and welded steel joints
100
1000
1.E+03 1.E+04 1.E+05 1.E+06 1.E+07
Number of cycles, N
S
t
r
e
s
s
r
a
n
g
e
,
S
(
M
P
a
)
Yield strength (MPa)
450 to 586
690 to 797
Tcurve (16mm) - air
Tcurve (16mm) - CP
Class F2 (16mm) - CP
63
Figure 7.11 Summary of Weld Fatigue Improvement Methods [7.11]
64
8. CATHODIC PROTECTION
8.1 INTRODUCTION
There are three systems that can be used to apply cathodic protection (CP), sacrificial anode cathodic
protection (SACP), impressed current cathodic protection (ICCP) and hybrid systems of SACP and
ICCP. The advantages and disadvantages of these systems are summarised in Table 8.1. CP is used
to protect both coated and uncoated steel from corrosion. SACP is the most widely used method for
protecting steel structures in the marine environment from corrosion. SACP uses sacrificial anodes
(usually aluminium based for structures and zinc based for pipelines) distributed around the structure
to ideally give an even distribution of a potential of 850 mV(Ag/AgCl)
2
. Due to the complexity of
structures it is possible to have high negative values occurring in the vicinity of anodes more negative
than 81000mV(Ag/AgCl) with the risk of hydrogen embrittlement and enhanced fatigue crack
propagation, and more positive potentials than 8750mV(Ag/AgCl) at remote or shielded locations
with the risk of localised corrosion. Clearly, for steels that are susceptible to hydrogen embrittlement
it is important to design the CP system to achieve the correct balance between the risks of corrosion
and hydrogen damage.
8.2 PROTECTION CRITERIA
There exists a great deal of guidance for CP levels, much of which has been derived empirically. The
relevant codes and standards relating to CP of high strength steels are compared in section 4.0. Care
must be taken when using this general guidance as it important to realise that the risk of hydrogen
cracking depends on the combination of material, loading, CP level and environment. Additionally,
the effects of the CP system on other materials connected to the protected steel structure (such as
duplex stainless steel) must also be considered.
The CP potential required for full protection (corrosion rate reduced to insignificant level) of steel in
seawater is widely considered to be -800mV(Ag/AgCl) [8.01]. Recommended potentials range
between -750 and -830mV(Ag/AgCl) [8.02]. Recent work [8.03, 8.04] found that a 700MPa offshore
steel was adequately protected (corrosion rate of 0.001mm/year) at potentials in the range -760 to
-790mV(Ag/AgCl).
In anaerobic conditions, where active populations of sulphate reducing bacteria (SRB) might be
present and producing sulphides, it has been generally recommended, for lower strength
constructional steels, that the potential should be lowered further to -900mV(Ag/AgCl) for full
protection. Hydrogen will be produced at potentials more negative than -710mV(Ag/AgCl) for North Sea
seawater (pH 8.3 & 10
o
C) [8.05]. It should be noted that the amount of hydrogen produced by CP
systems and therefore absorbed by the steel increases as the potential becomes more negative and
even small concentrations of sulphide can significantly increase in the amount of hydrogen absorbed
by steel [8.06]. These factors have meant that the recommended levels of CP for high strength steels
has required special attention due to their greater susceptibility to hydrogen embrittlement,
particularly above yield strengths of 700MPa. Recommendations have been made that for steels with
yield strengths 0 700MPa CP potentials should generally be within the range of -800 to
-950mV(Ag/AgCl) [8.01]. For steels with yield strengths >800MPa the potential should not go more
negative than -800mV(Ag/AgCl) [8.01]. The ranges described above are illustrated in Figure 8.1.
During the late 1980s, routine surveys of offshore jack-up drilling rigs discovered cracks in the legs and
spudcans that were believed to be due to hydrogen embrittlement [8.07]. A subsequent research
programme [8.08] included an investigation of the level of CP at which the jack-up steels showed
evidence of hydrogen embrittlement. The research employed slow strain rate testing and concluded that
to avoid problems the CP potentials should not be more negative than -805mV(Ag/AgCl). A parallel
2
For Ag/AgCl/Cl