HSE-2001

HSEHealth & Safety

Executive

Steel

OFFSHORE TECHNOLOGY REPORT

2001/015

HSEHealth & Safety

Executive

Steel

Edited under the HSE Technical Support Agreement by BOMEL LtdLedger House

Forest Green RoadFifield

MaidenheadBerkshire SL6 2NR

HSE BOOKS

ii

© Crown copyright 2002Applications for reproduction should be made in writing to:Copyright Unit, Her Majesty’s Stationery Office,St Clements House, 2-16 Colegate, Norwich NR3 1BQ

First published 2002

ISBN 0 7176 2391 2

All rights reserved. No part of this publication may bereproduced, stored in a retrieval system, or transmittedin any form or by any means (electronic, mechanical,photocopying, recording or otherwise) without the priorwritten permission of the copyright owner.

This report is made available by the Health and SafetyExecutive as part of a series of reports of work which hasbeen supported by funds provided by the Executive.Neither the Executive, nor the contractors concernedassume any liability for the reports nor do theynecessarily reflect the views or policy of the Executive.

iii

CONTENTS Page No FOREWORD v 1 INTRODUCTION AND SCOPE 1 1.1 Source of Information 1 1.2 Scope and Reference Documents 1 2 STEELWORK DESIGN 3 2.1 Structural Analysis 3 2.2 Dynamic Response 3 2.3 Allowable Stresses 3 2.4 Joint Design for Welded Tubular Steel Structures 3 2.5 Buckling 13 2.6 Stiffness 16 2.7 Vibrations 16 2.8 Stress Concentrations 16 2.9 Progressive Collapse 16 2.10 Fatigue – General Considerations 17 2.11 Fatigue – Stresses to be Considered 19 2.12 Fatigue – Joint Classification 24 2.13 Fatigue – Basic Design S-N Curves 35 2.14 Fatigue – Damage Calculation 44 2.15 Fatigue Life of High Strength Steels 44 2.16 Fatigue Life of Cast or Forged Steel Components 44 2.17 Bolts and Threaded Connectors 46 2.18 Moorings 47 2.19 Fatigue Performance of Repair Joints 48 2.20 Fracture Mechanics Assessment of Fatigue Life 49 3 BASIC STRUCTURAL STEELWORK MATERIALS STANDARDS 51

3.1 Hydrogen-Assisted Cracking in High Strength Steels Immersed in Seawater

51

4 ADDITIONAL MATERIALS REQUIREMENTS 53 4.1 Demonstration of Properties After Heat Treatment and Forming 53 4.2 Charpy Impact Requirements 53 4.3 Through-Thickness Ductility 56 5 CASTINGS AND FORGINGS 57 5.1 General 57 5.2 Qualification of Material and Manufacturing Facilities 57 5.3 Production Castings and Forgings 57 5.4 Quality of Cast or Forged Steel Products 57

iv

CONTENTS (CONTINUED) Page No 6 OTHER STEEL PRODUCTS 59 6.1 Pipes 59 6.2 Bolts and Nuts 59 7 WELDING CONSUMABLES 61 7.1 General 61 7.2 Standards 61 8 CONSTRUCTION STANDARDS – STEEL 63 8.1 General Applicability 63 8.2 Fabrication Tolerances 63 8.3 Storage of Materials 63 8.4 Preparation of Steel 63 8.5 Bolted Joints 64 8.6 Welded Joints 64 8.7 Welding 65 8.8 Avoidance of Brittle Fracture 66 8.9 Post Weld Heat Treatment (PWHT) 71 8.10 Splices 72 8.11 Stiffeners 72 8.12 Tack Welds and Temporary Attachments 73 8.13 Inspection and Testing 73 9 REFERENCES 75

v

FOREWORD

This document provides technical information previously contained in the Fourth Edition of the Health and Safety Executive’s ‘Offshore Installations: Guidance on Design, Construction and Certification’ (1990 edition plus amendments)(1). The ‘Guidance’ was originally published in support of the certification regime under SI289, the Offshore Installations (Construction and Survey) Regulations 1974(2). However, SI289 was revoked by the Offshore Installations (Design and Construction, etc) Regulations, 1996, which also introduced the verification provisions into the Offshore Installations (Safety Case) Regulations, 1992. The ‘Guidance’ was formally withdrawn in its entirety on 30 June 1998 (see HSE OSD Operations Notice 27(3)).

The withdrawal of the ‘Guidance’ was not a reflection of the soundness (or otherwise) of the technical information it contained; some sections (or part of sections) of the ‘Guidance’ are currently referred to by the offshore industry. For this reason, after consultation with industry, relevant sections are now published as separate documents in the HSE Offshore Technology (OT) Report series.

It should be noted that the technical content of the ‘Guidance’ has not been updated as part of the re-formatting for OTO publication, although prescriptive requirements and reference to the former regulatory regime have been removed. The user of this document must therefore assess the appropriateness and currency of the technical information for any specific application. Additionally, the user should be aware that published sections may cease to be applicable in time and should check with Operations Notice 27, which can be viewed at http://www.hse.gov.uk/hid/osd/notices/on_index.htm, for their current status.

vi

1

1. INTRODUCTION AND SCOPE 1.1 SOURCE OF INFORMATION This Offshore Technology (OT) Report provides technical information on the selection, design and fabrication of structural steel for Offshore Installations. It is based on guidance previously contained in Section 21 of the Fourth Edition of the Health and Safety Executive’s ‘Offshore Installations: Guidance on Design, Construction and Certification’(1) which was withdrawn in 1998. As discussed in the Foreword, whilst the text has been re-formatted for Offshore Technology publication, the technical content has not been updated. The appropriateness and currency of the information contained in this document must therefore be assessed by the user for any specific application. 1.2 SCOPE AND REFERENCE DOCUMENTS The scope of this document reflects the current design practice at the time of publication of the Fourth Edition ‘Guidance’. All design information presented in this OT Report is based on an allowable stress approach. The report also gives information on the selection of steel for offshore structures. The selection process is governed partly by the need for adequate resistance to brittle fracture, fatigue, corrosion fatigue and infrequent but high loadings caused by storm waves. Other factors affecting selection include the ability of the steel to be formed by hot or cold working and its weldability, including resistance to lamellar tearing. The chemical composition and mechanical property requirements of BS 4360(4) provide a useful basic standard for structural steels, although modifications or additional requirements may be necessary. Special consideration should be given to higher strength steels, forgings and castings, and their properties must be demonstrated to be adequate. This report should be read in conjunction with a background document by Harrison and Pisarski 1986(5), which contains additional information and references. Further information on the use of steel for floating Installations is given in Offshore Technology Report OTO 2001 048.

2

3

2. STEELWORK DESIGN 2.1 STRUCTURAL ANALYSIS Established structural analysis techniques are well documented in standard textbooks, supplemented by numerous authoritative technical papers on specialised aspects of design. Any recognised design method can be used, provided that it is appropriate for the problem under examination and it can take into account the most recently available information on structural behaviour and material properties. 2.2 DYNAMIC RESPONSE Load augmentation due to dynamic response to variable forces becomes of increasing importance as Installations are designed to be located in deeper and deeper waters and their natural period becomes nearer to that of waves transmitting significant amounts of energy. Where necessary, in addition to static and quasi-static calculations, a dynamic analysis should be undertaken. Information on the calculation of dynamic response, including methods of assessing its probable significance, is contained in CIRIA 1977(6). 2.3 ALLOWABLE STRESSES Allowable stresses in steel should be in accordance with the grade of steel used and within the limits specified in an appropriate code. Where the structural element, or type of loading, is not covered by the above, a rational analysis should be used to determine the basic allowable unit stresses with an equivalent level of safety. It is generally considered that the calculated tensile stress in a member should not exceed 60 per cent of the yield stress under operating conditions and 80 per cent of yield stress under extreme loading conditions. These stress limitations are appropriate for ordinary mild or higher tensile shipbuilding and structural steels and should be modified as may be necessary to provide an equivalent factor of safety for other steels. 2.4 JOINT DESIGN FOR WELDED TUBULAR STEEL STRUCTURES a) Application and nomenclature The following procedures are concerned with the static design of tubular joints formed by the full penetration welding of two or more tubular members. The procedures apply to tubular joints fabricated from steel plate satisfying the requirements of BS 4360(4) or from seamless tubulars to equivalent specifications. The word 'simple' in the following procedures refers to joints without overlap of brace members and without the use of gussets, diaphragms, stiffeners or grout. Overlapping joints are defined as joints in which the brace forces are partially transferred in shear between overlapping braces through their common weld. The design procedures are developed from consideration of the characteristic strength of tubular joints. Characteristic strength is defined as that value below which not more than 5% of the results of an infinite number of tests would fall.

4

The notation used within this section is as follows: d Outside diameter of brace D Outside diameter of chord at brace intersection e Joint eccentricity Fy Characteristic yield stress of chord member g Gap between braces for K and YT joints Ka Relative length factor M Moment load Mci, Mco Permissible in-plane and out-of-plane moment strengths Mdi, Mdo Design acting in-plane and out-of-plane moment loads Mi, Mo In-plane and out-of-plane moments Mki, Mko Characteristic in-plane and out-of-plane moment strengths P, Pl, P2, P3 Axial load Pc Permissible axial strength Pd Design acting axial load Pk Characteristic axial strength Qf Chord load factor Qg Coefficient for K joints Qu Strength factor for various joint and load types Q'∃ Geometrical modifier t, t1, t2 Wall thickness of brace T Wall thickness of chord U Chord utilisation ∃ Diameter ratio d/D ( Chord thickness ratio D/2T 2 Included intersection angle between brace and chord . Gap parameter g/D 8 Multiplier used with Qf parameter The principal items in the above list are defined in Figure 1. b) Determination of design loads Applied loads should be calculated by established structural analysis methods taking account of appropriate environmental and dynamic conditions. Considerations should be given to the effect on accuracy of the following:

• Variations in wall thickness and outside diameter of members

• Representation of joint cans

• Representation of brace stubs

• Joint eccentricities

• Working point offsets

Large joint eccentricities (i.e. xex ú D/4) should be modelled in the analysis at design stage. In some instances it may be appropriate to consider the effects of smaller eccentricities and/or the effects of local

5

joint flexibilities on member and joint loads. Requirements for modelling of joint eccentricities and procedures for the inclusion of joint eccentricity effects on local joint flexibilities need careful consideration. In overlapping joints there is a transfer of load between brace members through their common weld and it is essential to recognise these transfer components to avoid incorrect and possible unsafe estimation of the total load resisted by the chord. These transfer components, which are discussed in OTH 89 308(7) , are dependent on the joint geometry and configuration and on the type and magnitude of incoming brace loads. Accordingly, each overlapping joint should be treated on an individual basis. Design acting loads for each load component should be taken as:

Pd, Mdi, Mdo = calculated applied load c) Classification of joints for static strength design Each joint should be considered as a number of independent chord/brace intersections and the capacity of each intersection should be checked against the design requirements set out in paragraph (i) below. Each plane of a multiplanar joint should be subjected to separate consideration and classification. Each chord/brace intersection should be classified as Y, K or X according to their configuration and load pattern for each load case. Examples of joint classification are shown in Figure 2 and should be used with the following guidelines:

i) For two or three brace members on one side of the chord, the classification is dependent on the equilibrium of the axial load component in the brace members. If the resultant shear on the chord member is essentially zero, the joint should be allocated a K classification. If this requirement is not met, the joint can be downgraded to Y classification as shown in Figure 2. However, for braces which carry part of their load as K joints and part as Y or X joints, interpolation based on the portion of each in total may be valid.

ii) For multibrace joints with braces on either side of the chord as shown in the example DYDT joint in

Figure 2, care should be taken in allocating the appropriate classification. For example, a K classification would be valid if the net shear across the chord is essentially zero. In contrast, if the loads in all the braces are tensile (e.g. at a skirt pile connection), even an X classification may be unsafe due to the increased ovalising effect.

d) Factors affecting the strength of a tubular joint The following principal factors have been shown to affect the strength of a given simple tubular joint:

• Chord outside diameter (D)

• Brace outside diameter (d)

• Chord wall thickness (T)

• The included angle between chord and brace (θ)

• Gap between braces (for K joints only) (g)

• Chord material yield stress (Fy)

6

The static strength of a joint may be enhanced by the presence of ring and longitudinal stiffeners, gussets or cementitious filling (e.g. grouting). Comprehensive design formulae are not available for all cases and special consideration should be given to such joints.

Figure 1 Nomenclature for example non-overlapping and overlapping joints

7

NOTES 1) The loads P1, P2 and P3 are taken to act in the direction shown. 2) For all cases above check each brace separately. 3) This figure should be read in conjunction with Section 2.4c) for further information.

Figure 2 Examples of joint classification

SEE SECTION 2.4c)

8

e) Characteristic strength of joints The characteristic strength of a welded tubular joint subjected to unidirectional loading may be derived as follows:

fuk QQP = ?sin

KTF a2

y

Equation 2.1

fukoki QQM,M = ?sin

dTF 2y

where: Pk = characteristic strength for brace axial load Mki = characteristic strength for brace in-plane moment load Mko = characteristic strength for brace out-of-plane moment load Fy = characteristic yield stress of the chord member at the joint (or 0.7 times the characteristic tensile

strength if less). If characteristic values are not available specified minimum values may be substituted.

Ka = (1 + 1/sin2)/2 Qf is a factor to allow for the presence of axial and moment loads in the chord. Qf is defined as:

Qf = 1.0 - 1.638 λγU2 for extreme conditions Equation 2.2 = 1.0 - 2.890 λγU2 for operating conditions

where 8 = 0.030 for brace axial load = 0.045 for brace in-plane moment load = 0.021 for brace out-of-plane moment load and

( )y

2

2o

2i

2

FT0.72DMM0.23PD

U++

= Equation 2.3

with all forces in the function U relating to the calculated applied loads in the chord. Note that U defines the chord utilisation factor. Extreme and operating conditions are discussed in Offshore Technology Report OTO 2001 013. Qf may be set to 1.0 if the following condition is satisfied:

( )0.52o

2i MM

0.23D1

forcetensionaxialchord +≥ Equation 2.4

with all forces relating to the calculated applied loads in the chord. Qu is a strength factor which varies with the joint and load type. Qu is defined in Table 1.

9

Table 1 Coefficient Qu

Coefficient Qu for various joint design classifications

Load direction Y

K(+)

X

Axial compression

(2 + 20β) √Q'β

(2 + 20β) Qg √ Q'β

(2.5 + 14β) Q'β

Axial tension

(8 + 22β)

(8 + 22β) Qg

(7 + 17β) Q'β

In-plane bending *

5βγ1/2sinθ

Out-of-plane bending

(1.6 + 7β) Q'β

(1.6 + 7β) √ Q'β

NOTES (+) Applies only if the axial brace loads are essentially balanced (to +/-10%). See Section 2.4c for guidelines on classification. * The background document OTH 89 309(8) and report OTH 89 297(9) discuss the effect of intersection angle θ and review some results from tests of Y joints under in-plane bending. Qg = 1.7 - 0.9ζ1/2 but should not be taken as less than 1.0 Q'β is the geometrical modifier defined as follows: Q'β = 1.0 for β ≤ 0.6

= ( )0.833ß1ß

0.3−

for β ≥ 0.6

f) Capacity of overlapping joints The behaviour of overlapping joints and an assessment of the available test data may be found in OTH 89 308(7) and CIRIA UEG UR33 1985(10). Each overlapping joint should be treated as a separate case and the characteristic capacity of each intersection for each load type should be determined. This capacity can be determined on the basis of available test data where appropriate or through application of rigorous engineering mechanics. The application of engineering mechanics requires the determination of intersection lengths L1, L2 and L3 as defined in Figure 3. In lieu of precise measurements, the intersection length formulae given in Figure 3 can be used. Overlapping joints depend on the combined action of all braces. Accordingly consideration should be given to checks of the combined moment from all braces on the combined footprint.

10

Figure 3 Definition of intersection lengths for overlapping joints

For heavily overlapping braces (i.e. L1/L2 [ 0.5) the intersection between overlapping brace and the through brace should also be checked separately with the through brace taken to be a chord member under a Y classification.

11

g) Validity ranges and non conforming joints The formulae given in Section 2.4e) are applicable to the following joint geometrical and material property ranges:

9 [ γ ≤ 50

0.15 ≤ β ≤ 1.00

30° ≤ θ ≤ 90°

Fy ≤ 400 N/mm2 Joints which do not satisfy the above requirements and joints which make use of gussets, ring stiffeners, diaphragms, longitudinal stiffeners, grout or other means of strength enhancement, may be used providing it is demonstrated that the strength of such joints can be reliably estimated. Test data may be used as the basis for design. However, care should be taken in using the results of limited test programmes to provide characteristic joint strengths equivalent to those given in Section 2.4e). Statistical analysis of test data requires the imposition of a reduction factor in the calculation of characteristic values from a small amount of data. Statistical tables are given in CIRIA Technical Note 44(11). Analytical techniques may be used as the basis for design for some complex joints, e.g. joints with internal ring stiffening. Ring stiffeners and other reinforcement methods may also be used to reduce peak stresses and attention is drawn to fatigue performance. h) Safety factors The safety factors given below should be applied to the characteristic strengths Pk, Mki and Mko given in Section 2.4e) in the determination of permissible strengths Pc, Mci and Mco. The two conditions are defined in Offshore Technology Report OTO 2001 013. Conditions Safety factor Extreme 1.28 Operating 1.70 For overlapping joints, minimum safety factors of 1.28 and 1.70 for extreme and operating conditions respectively should be used in the determination of permissible strengths Pc, Mci and Mco. Higher safety factors may be appropriate depending on the level of confidence placed on the calculated characteristic strengths Pk, Mki and Mko. Note that the above safety factors include an allowance for both loading and resistance. In situations where the joint design is dominated by the environmental loads, or the joint geometry is outside the given application limits, a higher factor of safety may be required.

12

i) Design requirements For unidirectional or combined brace loads the following requirement should be satisfied:

1.0MM

MM

PP

co

do

2

ci

di

c

d ≤+

+ Equation 2.5

Additionally the shear stress at any transverse section in the chord under the calculated applied loads should not exceed 0.53Fy for extreme conditions or 0.4Fy for operating conditions. j) Joint detailing To achieve the required load carrying capacity many joints are likely to require strengthening compared to the basic tubulars meeting at the joint. Where increased wall thickness, diameter or special steel in the chord or brace is required, due consideration should be given to joint detailing and in particular:

i) the length of chord reinforcement or chord can extending past the outside of braces;

ii) the length of brace reinforcement of brace stub;

iii) the gap (g) between braces at K joints and between braces in different planes;

iv) the eccentricity (e) between any two braces, which should not exceed one quarter of the chord diameter (D/4) unless such eccentricity has been represented in the determination of member loads;

v) the degree of overlap in overlapping joints which should be arranged so that for an overlapping

brace the intersection length ratio L1/L2 (see Figure 3) is between 0.2 and 0.8. OTH 89 308(7) gives a summary of joint detailing. In using any specific recommendations, the following should be considered:

i) the minimum lengths required to give sufficient joint strength;

ii) the provision of adequate access consistent with achieving the required quality of welding between the tubular section and at the tubular joint (in particular the effect of the angle of intersection θ is important), and to facilitate effective inspection subsequently and throughout the life of the structure;

iii) the position of welds and their associated stress concentrations in relation to regions of the joint

which may be subject to high stresses under fatigue loading;

iv) the extent of parent metal heat affected zone, particularly for joints with two or more brace members.

13

k) Reduction of peak stresses by joint stiffening External gusset plates or internal ring or longitudinal stiffeners can be used to reduce the peak stresses at tubular joints and to increase static strength For stiffeners to be effective, the location is critical. For each specific configuration it is advisable to determine by model tests or a photo-elastic investigation, or a finite element analysis, the influence of the stiffeners on the stress distribution. It should be noted that longitudinal stiffeners extending along the bore of a brace from the joint to the chord member can have a negligible effect on the peak stress in the chord member, and that fillet welds at the outer ends of, and butt welds in, such stiffeners have been the sites of fatigue failures in service. 2.5 BUCKLING Components of offshore steel platforms and mobile drilling units which may be susceptible to buckling failure include the following:

• Slender chord and bracing elements of the structural framework, usually of round tubular construction: consideration must be given to behaviour under axial compression, possibly combined with bending, and external pressure.

• Flat stiffened panels, as occur in platform decks and supporting deep girders: buckling failure may occur under compressive, bending and shear loads.

• Ring-stiffened cylindrical shells: buckling may occur under external hydrostatic pressure, axial compression, bending, shear or some combination of these loads.

• End-closures to cylindrical shells, usually of hemispherical, torispherical or conical form: buckling may occur under external pressure.

• Large diameter orthogonally stiffened cylinders: consideration must be given to buckling under axial compression or bending loads which may be combined with external pressure and possibly shear.

Buckling failure will generally take place inelastically and may be strongly influenced by manufacturing imperfections including particularly initial deformations and residual stresses caused by misalignment, welding and cold-forming. Most offshore platforms are highly redundant structures whose ultimate strength will depend not only on the buckling strength of individual elements but also on pre-collapse loss of stiffness and post-collapse load-carrying capacity of such elements. a) Tubular members under axial and bending loads Design of axially compressed unstiffened tubes of low diameter/thickness ratio (d/t < 70) should be carried out with reference to empirical column design formulae as contained in API RP2A(12), based on the American CRC column curve (Johnston 1976(13) or DnV 1977(14)) or based on the European (ECCS) column curves (ECCS-EG77-2E, March 1978(15)). Design of tubes under combined axial and bending loads may be based on interaction formulae, as specified by API RP2A, or on a first yield (Perry-Robertson) criterion as specified by DnV 1977 and OTC 3903(16), and also Ellinas et al, 1984(17). Consideration must be given to local buckling of tube walls if D/t exceeds a certain level, given by API as 60, by BSI as 0.11 E/Φy (E=Young's modulus, Φy = yield stress) and by DnV as 0.088 E/Φ y. For thinner-walled tubes in which D/t exceeds these limits, the local buckling stress Φ c for a thin-walled cylinder under

14

axial compression may be substituted for the yield stress Φy in the column design formulae: empirical formulae for Φ c are given in API RP2A and DnV 1977. Because of the catastrophic post-collapse loss of load-carrying capacity in thin-walled, high-strength tubes (D/t > 0.2 E/Φ ) and their sensitivity to damage in the form of small dents (Smith et al, 1979(18)), particular care should be taken when considering the use of such members in the primary structure. Information on the effects of lack of straightness and residual stresses in tubular members may be found in Smith et al, 1979. Some guidance on pre-buckling loss of stiffness, post-collapse load-carrying capacity and damage effects in tubular beam-columns is contained in Smith et al, 1979. Design information on beam-columns of rectangular box-section and open cross-section is provided in DnV 1977, AISC 1980(19) and Ellinas et al, 1984. b) Flat stiffened panels and plate girders Failure of a flat stiffened panel under uniaxial or biaxial compressive loads may involve:

i) local buckling of rectangular plate panels between stiffeners; ii) interframe flexural buckling of stringers (normally aligned in the direction of dominant compressive

load) and attached plating between heavier, more widely spaced transverse stiffeners; iii) lateral-torsional buckling of stiffeners; iv) overall instability involving bending of longitudinal and transverse stiffeners.

The above failure modes may be strongly affected by loss of plating stiffness caused by buckling and yielding of plate panels: guidance on effective plate stiffness and strength under uniaxial and biaxial compression may be found in Ellinas et al 1984(17), DnV 1977(14), Frieze et al 1977(20), and Dowling et al 1979(21). An understanding of failure by interframe flexural buckling may be obtained from tests and parametric studies described in Horne & Narayanan 1977(22), Moolani & Dowling 1977(23), Smith & Kirkwood 1977(24), Little 1976(25) and Carlsen 1980(26). Torsional buckling of stiffeners is normally avoided by conservative proportioning of stiffener sections: guidance is contained in DnV 1977, AISC 1980(19) and BS 5400 1980: Part 3(27). Overall buckling may be suppressed by providing sufficient transverse flexural stiffness to ensure that interframe buckling precedes overall instability. Detailed design procedures for stiffened panels under compressive load are contained in DnV 1977 and BS 5400 1980 (Pt. 3). Design of stiffened and unstiffened plate girders under shear load, possibly combined with in-plane compression or bending, may be based conservatively on elastic buckling criteria by ensuring, with appropriate margins of safety, that:

i) combined stresses do not cause yield; ii) stiffeners have sufficient flexural rigidity to suppress overall buckling; iii) the elastic buckling stress of plate panels between stiffeners is not exceeded.

Design procedures based on this approach are described in DnV 1977 and AISC 1980. Design may be based, alternatively and less conservatively, on plastic collapse criteria allowing for the post-buckling strength of slender girder webs associated with diagonal tension: this option is included in the AISC and BS 5400 recommendations. Care must be taken to avoid local web buckling under high concentrated loads: guidance is included in AISC 1980, BS 5400 and Bergfelt 1977(28). A general review of girder buckling behaviour and design methods is contained in Rockey 1977(29). Data defining the stiffness, strength and post-collapse behaviour of plates under shear and combined loads may be found in Harding et al 1977(30).

15

c) Ring-stiffened cylinders Failure of a ring-stiffened cylinder under external hydrostatic pressure may occur in any of the following modes:

i) interframe shell buckling which may be either axisymmetric or 'lobular' in form, i.e. may involve either a single lobe or several bucking waves round the circumference of the shell between frames: design is normally based on empirical lower-bound curves as specified in the API RP2A(12), DnV 1977(14) and BS 5500(35) requirements;

ii) lateral-torsional instability of ring frames, normally avoided by conservative proportioning of frame

cross-sections (API RP2A, DnV and BS 5500);

iii) general instability involving bending of ring frames: design is normally carried out by proportioning frames so that general instability occurs at a substantially higher pressure than interframe buckling (API RP2A, DnV, and BS 5500).

The last form of failure is particularly sensitive to departures from circularity in ring frames and to residual stresses caused by cold-forming of frames: guidance on the effect of imperfections, both within and outside normal tolerances, may be found in Smith et al 1979(18), Kendrick 1979(31) and Faulkner 1977(32). For design purposes, interframe buckling of a ring-stiffened cylinder under axia l or bending load may be treated conservatively as that of an unstiffened cylinder of length equal to the spacing of frames or diaphragms. This form of failure is very sensitive to short-wavelength imperfections (Harding 1978)(33); imperfection sensitivity is however reduced by the presence of closely spaced ring frames (Miller I977)(34). Sizing of ring frames to avoid overall instability may be based on elastic buckling analysis (DnV 1977) or test data (Miller 1977). In the case of long cylinders, consideration should be given to possible interaction between local shell buckling and overall column instability (DnV 1977). Design of ring-stiffened cylinders under axial or bending load combined with external pressure may be based on interaction formulae as specified in API RP2A, DnV 1977 and Miller 1977. d) Hemispherical, torispherical and conical end-closures Design against buckling under external pressure may be carried out using the methods described in BS 5500(35) and the draft ECCS recommendations(15). The basis of these methods is explained in Newland 1972(36). The sensitivity of spherical shells to imperfections in the form of local flattening should be noted (Kendrick 1980)(37). e) Orthogonally stiffened Cylinders Buckling of orthogonally stiffened cylinders under axial compression or bending load is analogous with that of orthogonally stiffened flat panels under uniaxial compression: possible failure modes correspond to those described in Section 2.5b). Overall buckling should normally be excluded by spacing and proportioning ring frames or diaphragms so that, on the basis of elastic buckling analysis, interframe buckling modes precede overall modes by a substantial margin: formulae for the evaluation of elastic buckling, allowing for stiffener eccentricity, may be found in Block et al 1965(38); guidance is also contained in DnV 1977(14).

16

Torsional buckling of stringers should be avoided by careful proportioning of stiffener sections (see Section 2.5b) and API RP2A(12), DnV 1977, AISC(19) and BS 5500(35)). For design purposes, buckling of a stringer-stiffened shell between transverse frames or diaphragms may be treated conservatively as that of a flat-stiffened panel: this approach, which is adopted in the DnV Rules, is reasonable for short cylinders (closely spaced rings or diaphragms) but becomes very conservative for long cylinders (Smith 1978(71)). Similarly, the local buckling strength and stiffness of curved plate panels between stringers may be examined conservatively as that of equivalent flat panels: more accurate consideration should be given to possible interaction between local shell buckling and overall column instability (DnV 1977). The buckling strength of an orthogonally stiffened cylinder under external hydrostatic pressure may be estimated conservatively, ignoring the influence of stringers, as that of a ring-stiffened cylinder. For the case of external pressure combined with axial compression or bending, interaction formulae may be used as for a ring-stiffened cylinder. 2.6 STIFFNESS The maximum permissible deflections will depend on the nature of the structure and the service for which it is intended. Consideration should be given to the possibility of elastic instability of the structure as a whole; deflections of individual members should be limited so as to avoid such instability. 2.7 VIBRATIONS Precautions should be taken in the design against the possibility of excessive structural vibration being induced by machinery. This entails investigation of the natural frequencies of the members and of the sources of excitation. When it is not possible to be certain that excessive vibrations will not occur, a survey should be made of the finished Installation during initial service and any local or general resonance which could be damaging should be identified. 2.8 STRESS CONCENTRATIONS The design and fabrication or construction of all structural details should be such as to minimise stress concentrations. 2.9 PROGRESSIVE COLLAPSE The Installation should be so designed that accidental damage to part of the Installation will not lead to progressive collapse of the whole Installation.

17

2.10 FATIGUE - GENERAL CONSIDERATIONS The following guidance on fatigue applies to all types of welded offshore steel structures (see OTH 92 390(39) and Background to New Fatigue Guidance etc. 1984(40)). a) Background An Installation is exposed throughout its service life (which means here the anticipated operating life) to environmental loading which causes cyclic stress variations in its structural members resulting in the possibility of fatigue cracking in these members. Wave loading is the main source of potential fatigue cracking; for example over a 20-year service life wave action will result in about 108 cycles of stress variation. However, any other source of cyclic loading could also contribute to fatigue damage and should therefore be considered. In general, the calculated fatigue life derived by the methods below should be not less than 20 years, or the required service life if this exceeds 20 years. b) Scope of fatigue analysis To ensure that the structural members of the Installation fulfil their intended function, a fatigue assessment, supported where appropriate by a detailed fatigue analysis, should be carried out. In this context, it should be noted that, in any element or member of the structure, every welded joint or other form of stress concentration is potentially a source of fatigue cracking and each should be individually considered, taking due note of symmetry where appropriate. c) Fatigue loading In assessing fatigue performance all types of cyclic loading should be considered. It should be noted that cyclic loading from different sources may be significant at different phases of the life of a structure (e.g. construction, transport, installation, in-service), and may involve different frequencies. Some important sources of cyclic loading are:

• Waves (including those which cause slamming and variable -buoyancy effects)

• Wind (especially when vortex shedding is induced e.g. on slender members)

• Currents (where these influence the forces generated by waves and/or induced vortex shedding) and

• Mechanical vibration (e.g. caused by operation of machinery).

Loads are discussed in more detail in Offshore Technology Report OTO 2001 013. d) Basis of fatigue analysis It is recognised that uncertainties exist in assessing both the stresses resulting from applied loads and the response of a particular joint, which together control fatigue failure. The basis of the fatigue analysis is to use a best estimate of these stresses, noting the uncertainties involved, with fatigue curves derived from experimental data. In this way, uncertainties associated with the life of a particular joint (e.g. size, weld detail, local environment) can be separated from those associated with applied stress.

18

In order to evaluate fatigue lives, it is necessary to establish the long term distributions of stress range taking into consideration all stress variations which can reasonably be expected during the life of the structure which have magnitudes and numbers large enough to cause fatigue effects. The long term stress range distribution of the structure should be determined in a satisfactory manner. Normally, dynamic amplification should be considered when the natural period of the structure is greater than 3 seconds. Due consideration should be given to the orientation and draught, where applicable, of the structure in relation to the distribution of direction of the weather causing cyclic loading; in the absence of other data, it should be assumed for each weather state that wind, waves and current all act in the same direction. e) Fatigue life From the calculated long term stress range distributions, the procedures given in Sections 2.11 to 2.14 should be followed in assessing fatigue lives. The resultant calculated fatigue lives, which enable critical parts to be ranked in terms of fatigue sensitivity, can be used, in conjunction with an assessment of the consequences of failure of specific members, to establish a priority basis for developing a selective inspection programme to be followed during the service life of the Installation (c.f. OTH 87 278(41)). Operating experience and research developments during the service life may indicate a need to update the fatigue life calculations. f) Factors on fatigue life The basic design S-N curves defined in Section 2.13a) are based on the mean-minus-two-standard-deviation offsets from relevant experimental data. Their use therefore implies a finite probability of failure at the calculated life of 2.3%. An additional factor on life should be considered for cases of inadequate structural redundancy. In defining this factor on life, account should be taken of the accessibility of the joint and the proposed degree of inspection as well as the consequences of failure. Areas which are difficult or impossible to inspect should be given an additional factor on fatigue life to ensure continued integrity throughout the whole lifetime of the Installation. A recommended factor is given in OTH 92 390(39) but this should not be taken to imply any form of encouragement for the design of non-redundant and/or uninspectable structures. Because of the sensitivity of calculated life to the accuracy of estimates on stress range, particular care should be taken to ensure that these stress ranges are not underestimated. g) Factors influencing fatigue behaviour For welded steel structures it has been established that the fatigue life is normally governed by the fatigue behaviour of the joints, including both main and attachment welds. The best fatigue behaviour will be obtained by ensuring that the structure is so detailed and constructed that stress concentrations are kept to a minimum and that, where possible, the elements may deform in their intended ways without introducing secondary deformations and stresses due to local restraints. Stresses may also be reduced by increasing the thickness of parent metal; this should improve fatigue life but it should be remembered that fatigue strength tends to decrease with increasing thickness (see Section 2.13). The effectiveness of cathodic protection is also an important factor influencing the fatigue behaviour of structural steels in sea water (see Section 2.13 and Offshore Technology Report OTO 2001 011), but it should be noted that overprotection may be detrimental (see OTH 92 390(39) and Background to New Fatigue Design Guidance etc. 1984(40)).

19

2.11 FATIGUE - STRESSES TO BE CONSIDERED The procedure for the fatigue analysis is based on the assumption that, for as-welded joints, it is only necessary to consider ranges of cyclic stress in determining the fatigue life (i.e. mean stresses are neglected). For post-weld heat treated joints the fatigue performance may improve as the load applied becomes more compressive. However, under purely tensile loading the fatigue performance should be regarded as that for as-welded joints. It is recommended that for post-weld heat treated joints no advantage in terms of fatigue performance should be taken at the design stage. In most instances, therefore, the stress ranges for post-weld heat treated joints should be assumed to be the same as for as-welded joints. However, in the event that it can be demonstrated using validated techniques that a compressive component of the combined applied and residual stresses exists and can be quantified, it would be permissible to assume a stress range for fatigue of the tensile component plus 60% of the compressive component. Residual stresses arising from fabrication, construction and installation should also be taken into account. To ensure the effectiveness of post-weld heat treatment, the process should be carried out in accordance with accepted standards (e.g. BS 5500(35)). The relevant cyclic stress to be considered is the range of maximum principal stress at the potential crack location. In estimating the maximum principal stress, shear and torsional effects may be neglected where these are small. The use of full penetration welds (see Section 8.7d)) is recommended for all nodal joints (i.e. tubular brace to chord connections). For full penetration welded joints fatigue cracking would usually be located at the weld toes. In load-carrying partial penetration or fillet-welded joints, where cracking could occur in the weld throat, the relevant stress range is the maximum range of shear stress in the weld metal. 2.11.1 Nodal Joints Simple nodal joints are defined as being non-overlapped, unstiffened, ungrouted and uniplanar. Complex nodal joints incorporate overlapping braces, stiffening or grout in the chord and may have braces in several planes (i.e. multi-planar). For all nodal joints, stress ranges on both the brace and chord sides should be considered in any fatigue assessment. The stress range to be used in the fatigue analysis is the hot spot stress range at the weld toe. For any particular type of loading (e.g. axial loading) this stress range is the product of the nominal stress range in the brace and the appropriate Stress Concentration Factor (SCF). The definition and procedure for calculating the SCFs are given below. a) Definition of stress concentration factors Stress Concentration Factors (SCFs) for each mode of loading are obtained by linear extrapolation to the weld toe of the geometric stress distribution near the weld toe. This SCF incorporates the effects of overall joint geometry (i.e. the relative sizes of brace and chord) including the gross influence of the weld, but does not include the effect of the local weld notch (see Figure 4). Hence, the SCF is considerably less than the peak stress but provides a consistent definition of stress range for the design T' S-N curve. The calculation of SCF may be undertaken in a variety of ways, e.g. by physical model studies, finite-element analysis, or by use of parametric formulae (see next section).

20

When using SCFs in design, account should be taken of stresses arising from different loading modes and the hot spot stress derived under combined loading conditions. Suitable methods are given in the report UEG UR33 1985(10). b) Calculation of stress concentration factors

i) Use of physical models Large scale steel models are preferred to measure SCFs but care should be taken in obtaining the geometric stress extrapolated to the weld toe. The positioning of the strain gauges near the weld toe is important (see Background Notes to New Fatigue Design Guidance for Steel Welded Joints in Offshore Structure 1984(40)). In addition the position of the supports and any diaphragms needs careful attention (see OTH 92 390(39)). When using other physical models (e.g. small scale steel, acrylic or photoelastic), care should be taken to simulate the behaviour of large scale joints, including the effects of the weld. In particular caution is needed in using very thin walled or small diameter tubular sections.

ii) Use of finite element methods

Thin shell Finite Element (FE) models do not normally include the weld and care should be taken in the selection of the stress sampling positions including the position at which the extrapolation is obtained. The derivation of the SCF should also take account of the notional position of the weld, to be consistent with the definition in Section 2.11.1a). The use of thick shell FE models, in combination with 3-D brick elements for the weld, is preferred because it enables modelling of the gross features of the weld and direct extrapolation to the weld toes.

iii) Use of parametric formulae

Several parametric formulae have been produced for the prediction of SCFs for tubular joints, based on data from both physical and FE models.

c) Simple joints Table 2 presents a matrix of acceptable formulae for the prediction of SCFs for simple joint configurations (e.g. T/Y, X and K). These have been validated against data from large scale steel models and also checked against data from acrylic models and have been shown to provide acceptable predictions (see OTH 92 390(39)). The use of formulae not included in the table may result in significant underpredictions and should be treated with caution. It should be noted that, for the chord crown under axial load in X joints, the database is too small to recommend any equation. It is recommended that the chord saddle SCF is applied at all periphery locations unless another appropriate method is proposed.

21

Figure 4 Example of hot spot stresses in a nodal joint

22

Table 2 SCF matrix tables for X, K and T/Y joints

X Joints

Loading

Position Efthymiou LR

Balanced Axial Balanced OPB Balanced IPB

Chord Saddle Chord Crown Brace Saddle Brace Crown Chordside Braceside Chordside Braceside

R N*

R R

R R

R R

N N*

R R

R R

R R

K Joints

Loading

Position Efthymiou LR

Balanced Axial Balanced OPB Balanced IPB

Chordside Braceside Chordside Braceside Chordside Braceside

N N

R R

R Rc

R R

R R

R R

T/Y Joints

Loading

Position Efthymiou LR

Balanced Axial Balanced OPB Balanced IPB

Chord Saddle Chord Crown Brace Saddle Brace Crown Chordside Braceside Chordside Braceside

R R R

Rc

R R

R R

R R R R

R R

R N

Key to SCF Tables R Recommend the equation RC Recommend the equation - but note that the equation is generally conservative N Not recommend the equation, since it fails to meet the acceptance criteria N* The equation cannot be recommended since there are less than 15 steel and acrylic joints in the SCF

database Efthymiou Efthymiou Equations (Efthymiou 1988)(42) LR Lloyds Register Equations (Smedley and Fisher 1990)(43)

23

d) Complex joints Multiplanar joints For multiplanar joints, SCFs are usually determined assuming there is no interaction between joints in different planes. However, in certain load cases, significant interaction can occur between joints in different planes. These effects, which can result in significantly different SCFS, should be assessed using appropriate methods, e.g. expressions used in Efthymiou 1988(42). Overlapped joints The hot spot stress approach should also be used for overlapped joints except that the joint intersection includes the common weld between the two braces. Parametric formulae for the prediction of SCFs in overlapped joints are given by Efthymiou 1988(42). They have not been validated because of the limited database available. Ring-stiffened joints In most cases, the introduction of ring-stiffeners into simple tubular joints will significantly reduce the stress at the saddle location on the brace-chord intersection, the exception being for joints where ∃ = 1.0 (i.e. brace diameter = chord diameter), where ring-stiffeners can lead to an increase in hot spot stress at the saddle location. At the crown location, ring-stiffeners give only a small reduction in stress, and in cases where rings are located under or near to the crown there may be a marked increase in the crown stress, particularly on the brace side. For ring-stiffened joints, the maximum stresses may not be at the saddle or crown locations on the brace-chord intersection. Care should therefore be taken to identify stresses elsewhere in the joint, e.g. at the locations where the brace crosses a ring-stiffener (i.e. the Brace/Ring Intersection (BRI) or the ring inner edge). The design S-N curves appropriate to these locations are those shown in Section 2.12a). Since cracking of the ring-stiffeners is extremely difficult to detect in-service, ring-stiffeners should be designed with minimum lives in line with the information given in Section 2.10f). Parametric SCF formulae for ring-stiffened joints have been developed from acrylic model test data (Smedley and Fisher 1990(43)) which give the brace/chord intersection SCFs in terms of the equivalent unstiffened joint SCFs. Equations to predict the SCF at the ring inner edge have also been given (Smedley and Fisher). Grouted joints The introduction of grout into simple tubular joints results in a reduction of the hot spot stress at the brace/chord intersection. However, for some joint configurations and loading modes this reduction in stress is insignificant. A conservative approach should therefore be taken by considering the joint as being ungrouted and treating it as a simple nodal joint. Cast joints For cast joints, the stress range to be used in the fatigue analysis is the maximum stress range. This stress range can occur at any location in the joint (including the internal surface) and should be determined using an appropriate method such as finite element analysis or model tests. 2.11.2 Non-Nodal Joints For all other types of joint (e.g. butt joints and appurtenance connections), the classifications are given in Section 2.12a) which also identifies the appropriate classification factor to be used in conjunction with the P class S-N curve. For non-nodal joints (e.g. butt joints and appurtenance connections), the nominal stress range with an appropriate stress factor should be used in conjunction with the design S-N curve for class P.

24

This stress factor should include any stress concentration effects resulting from the gross shape of the structure. 2.12 FATIGUE-JOINT CLASSIFICATION a) Classified welded details For the purpose of fatigue design, welded joints can be divided into a number of classes. All tubular joints are assumed to be in Class T' (see Section 2.13). Other types of joint, including tube to plate, may fall in one of several classes depending upon:

• The geometrical arrangement of the detail

• The direction of the fluctuating stress relative to the detail

• The method of manufacture and inspection of the detail.

Each structural detail in a member subjected to fluctuating stresses should be placed in its relevant joint class in accordance with the criteria given below. Otherwise the detail should be dealt with in accordance with Section 2.12b). The applicable S-N curves are specified in Section 2.13. These include the P curve for plate to plate and tube to plate weld details, the T' curve for welded tubular nodal joints and the CS curve for cast steel joints. A classification factor, to be used as a multiplier on stress ranges, has been introduced for use with the P curve. This simplification converts the P curve to an equivalent of the B to W S-N curves in earlier editions of the ‘Guidance’. The classification factors are given in Table 3. This section provides information on the classification of structural details for the selection of an appropriate S-N curve or a classification factor to be used with P curve.

25

Table 3 Weld classification factors and the corresponding S-N curves

Classification Factor

S-N Curve

0.64 0.76 1.00 1.14 1.34 1.52 1.83 2.54

B C D E F F2 G

W' (see note 1)

Note 1. Other work has shown that the W curve requires modification (see OTH 91 356(44)). In general, the classification is considered to include the SCF associated with the local weld detail but not any geometric stress concentration factors. In calculating the hot spot cyclic stresses for damage assessment, account must therefore be taken of the associated geometric stress concentration factors derived from FE analyses, parametric SCF equations or as recommended in OTH 91 357(45). It should be noted that in any welded joint there are several locations at which fatigue cracks may develop, e.g. at the weld toe in each of the two parts joined, at the weld ends and in the weld itself. Each part should be classified separately. For example, in the case of members or elements connected by load-carrying fillet-welds or partial penetration butt-welds, cracks may initiate either in the parent metal (at one of the weld toes) or in the weld throat: both possibilities should be checked by defining the appropriate classification and the corresponding stress range. The following is a list of weld connection types which are considered in this classification: Type 1: Material free from weld Type 2: Continuous welds essentially parallel to the direction of stress Type 3: Transverse butt welds in plates (perpendicular to the direction of stress) Type 4: Welded attachments on the surface or edge of a stressed member Type 5: Load-carrying fillet and T butt welds Type 6: Details in welded girders Type 7: Details relating to welded tubular members TYPE 1 MATERIAL FREE FROM WELD

In plain steel, fatigue cracks initiate either at surface irregularities or at corners of the cross-section. In welded construction, fatigue failure will rarely occur in a region of plain material since the fatigue performance of the welded joints will usually be much lower. In steel with rivet or bolt holes or other stress concentrations arising from the shape of the member, failure will usually initiate at the stress concentration. See BS 5400 : Part 10(27). In the following, the classification factors are multipliers on the stress ranges.

26

TYPE 2 CONTINUOUS WELDS ESSENTIALLY PARALLEL TO THE DIRECTION OF STRESS

With the excess weld metal dressed flush, fatigue cracks would be expected to initiate at weld defect locations. In the as-welded condition, cracks might initiate at stop-start positions or, if these are not present, at weld surface ripples. Backing strips

If backing strips are used in making these joints: (a) they must be continuous, and (b) if they are attached by welding, these welds must also comply with the relevant Type requirements. (It is noted particularly that tack welds, unless subsequently ground out or covered by a continuous weld, should be assessed using Type 6.5).

Edge distance

An edge distance criterion exists to limit the possibility of local stress concentration occurring at unwelded edges, for example, as a result of undercut, weld splatter, or accidental overweave in manual fillet welding (see also notes on Type 4). Although an edge distance can be specified only for the 'width' direction of an element, it is equally important to ensure that no accidental undercutting occurs on the unwelded corners of, for example, cover plates or box girder flanges. If this occurs, it should be subsequently ground smooth.

27

TYPE 3 TRANSVERSE BUTT WELDS IN PLATES (PERPENDICULAR TO THE DIRECTION OF

STRESS)

With the weld ends machined flush with the plate edges, fatigue cracks in the as-welded condition normally initiate at the weld toes, so that the fatigue performance depends largely upon the shape of the weld cap. If this is ground flush the stress concentration is removed and failure is then associated with weld defects. In welds made on a permanent backing strip, fatigue cracks initiate at the weld metal/strip junction, and in partia l penetration welds at the weld root. When welds are made entirely from one side without a permanent backing, a satisfactory profile in the making of the root beads must be achieved. The calculation of the hot spot stress ranges should include stress concentration factors to account for geometric effects including mismatch and thickness change (see OTH 91 357(45)).

28

TYPE 4 WELDED ATTACHMENTS ON THE SURFACE OR EDGE OF A STRESSED MEMBER

When the weld is parallel to the direction of the applied stress, fatigue cracks normally initiate at the weld ends, but when it is transverse to the direction of stress, they usually initiate at the weld toes; for attachments involving a single, as opposed to a double weld, cracks may also initiate at the weld root. The cracks then propagate into the stressed member.

29

When the welds are on or adjacent to the edge of the stressed member, the stress concentration is increased and the fatigue performance is reduced and this must be separately assessed and included in the calculation of applied stress or the detail reclassified as Type 4.2. This is the reason for specifying an 'edge distance' in some of these joints (see note on edge distance in Type 2).

30

TYPE 5 LOAD-CARRYING FILLET AND T BUTT WELDS

In cruciform or T joints with full penetration welds fatigue cracks will normally initiate at the weld toes. In joints made with load-carrying fillet or partial penetration butt welds cracking may initiate either at the weld toes and propagate into the plate or at the weld root and propagate through the weld. In welds parallel to the direction of the applied stress, however, weld failure is uncommon; cracks normally initiate at the weld end and propagate into the plate in a direction perpendicular to the applied stress. The stress concentration increases, and the fatigue performance therefore reduces, if the weld end is located on or adjacent to the edge of a stressed member rather than on its surface.

31

TYPE 6 DETAILS IN WELDED GIRDERS

Fatigue cracks generally initiate at weld toes and are especially associated with local stress concentrations at weld ends, short lengths of return welds and changes of weld direction. Concentrations are enhanced when these features occur at or near an edge of a component (see Type 4). Most of the joints in this section are also shown, in a more general form, in Type 4. However, because these joints occur frequently in welded girders, they are included here for convenience. Edge distance refers to the distance from a free, i.e. unwelded edge (see Type 2). See the sketches below for the definition of edge distance in this section.

32

TYPE 7 DETAILS RELATING TO WELDED TUBULAR MEMBERS

Since fillet and partial penetration butt welds have poor fatigue performance, they should be avoided if at all possible for joints where fatigue is a significant consideration. The reference thickness to be used in the calculation of the thickness correction is 16 mm for both the T' and P curve and 38 mm for the CS curve. Thickness correction is based on the local thickness at the predicted failure site.

33

34

35

b) Unclassified welded details Details which are not expressly classified in Section 2.12a) above should be treated as Class P with a classification factor of 1.52, or with a classification factor of 2.54 for load-carrying weld metal, unless a higher classification can be justified by reference to published experimental work or by carrying out tests. Such tests should be sufficiently extensive to allow the design S-N curve to be determined in the manner used for the standard classes. 2.13 FATIGUE - BASIC DESIGN S-N CURVES The basic design S-N curves for the two main joint classes (T' and P) in air are shown in Figure 5. The derivation of the curves and the modifications to them are described below. (For comparison, the curves from the earlier editions of the ‘Guidance’ are reproduced as Figures 8 (a) and (b)). a) Basic design S-N curves for welded joints The basic design S-N curves for flat or rolled plates (P) and simple tubular joints (T') in air are shown in Figure 5. These curves consist of linear relationships between Log10(SB) and Log10(N) and are based upon a statistical analysis of appropriate experimental data which may be taken to represent two standard deviations below the mean line. The basic design S-N curves are of the form:

36

Log10(N) = Log10(K1) - m Log10(SB) Equation 2.6 where N is the predicted number of cycles to failure under stress range SB, Kl is a constant and m is the inverse slope of the S-N curve. Details of the basic design curves are given in Table 4. No, the point where the slope of the line changes, and So, are explained in Section 2.13d). Identification of welded plate details should be made according to Section 2.12a), which also identifies the appropriate cla ssification factor to be used in conjunction with the S-N curve for Class P. The S-N curves to be used for complex tubular joints are also given in Section 2.12a).

Table 4 Details of Basic Design S-N Curves

Class

Environment

Log10K1

m

So(N/mm2)

No (cycles)

P P P P P

Air

Seawater (FC) Seawater (CP) Seawater (CP)

12.182 15.637 11.705 11.784 15.637

3 5 3 3 5

53 -

84

107

1.02x106

T' T' T' T' T'

Air

Seawater (FC) Seawater (CP) Seawater (CP)

12.476 16.127 12.00 12.175 16.127

3 5 3 3 5

67

95

107

1.745x106

Note: FC = Free Corrosion CP = Cathodic Protection b) Design S-N curves for joints in seawater A number of factors should be considered when using the basic design S-N curves for welded joints.

i) Unprotected joints in seawater For plate and tubular joints exposed to seawater, but without adequate corrosion protection, fatigue lives should be determined using the curves given in Table 4 and shown in Figure 6. It should be noted that no benefit from toe grinding or other weld improvement techniques should be assumed for unprotected joints in seawater.

ii) Protected joints in seawater

For plate and tubular joints exposed to seawater, but protected from corrosion using, for example, sacrificial anodes (see Offshore Technology Report OTO 2001 011), fatigue lives should be determined using the appropriate S-N curves given in Table 4 and shown in Figure 7.

37

Figure 5 Basic design curves for welded tubulars and plates in air

38

Figure 6 Basic design curves for welded tubulars and plates in seawater (free corrosion)

39

Figure 7 Basic design curves for welded tubulars and plates in seawater (with cathodic protection)

40

Figure 8 (a) Basic design S-N curves for non-nodal joints

Figure 8 (b) Basic design S-N curve for nodal joints

41

c) Modifications to the basic design S-N curves i) Effect of plate thickness

For welded joints the fatigue performance is dependent on member thickness, performance decreasing with increasing thickness for the same stress range. The definition of the member thickness is given in 2.12a). The basic design S-N curves are applicable to thicknesses less than the basic thickness, tB which for both Classes P and T' is 16mm. For members of greater thickness, a modified S-N relationship applies as follows:

( ) ( )( )

−= q

B

B1011010

/ttS

LogmKLogNLog Equation 2.7

where q = the thickness exponent factor; equal to 0.3 for welded joints (see OTH 92 390(39)) SB = the stress range in the member under consideration N = the predicted number of cycles to failure under stress range SB K1 = a constant m = the inverse slope of the S-N curve t = the thickness of the member under consideration tB = 16mm for both T' and P curves Other thickness correction relationships may be used if they can be justified by data from a validated test programme or by fracture mechanics analysis.

ii) Weld toe improvement For welded joints an improvement of 2.2 on life can be obtained by controlled local machining or grinding to produce a smooth concave profile at the weld toe, which blends smoothly with the parent material. This benefit may be claimed for welded joints in air and for joints exposed to seawater with adequate protection against corrosion. However, it is recommended that this advantage should not be utilised at the initial design stage. Care should be taken to remove all defects in the critical region, by grinding of the weld toe region to a depth of not less than 0.5mm below the bottom of any visible undercut or defect (see Figure 9). The maximum depth of local grinding should not exceed 2mm or 5% of the plate thickness, whichever is le ss, as shown in Figure 9. It is recommended that the final grinding operation should be carried out using a rotary burr to provide a generous radius to blend with the surrounding material. An appropriate NDE technique e.g. BS 6072(46) should be used to ensure that no significant defects remain after grinding. Where appropriate, the final ground surface should be suitably protected in order to avoid local corrosion pitting prior to the application of cathodic protection. Where toe grinding is used to improve the fatigue life of fillet welded connections, care should be taken to ensure that the required throat size is maintained. It should also be noted that grinding of the weld toe will not give an increase in life in other fatigue sensitive regions such as the weld root.

42

It is known from laboratory tests that other weld improvement techniques such as hammer or shot peening can also enhance fatigue life. Any advantage in terms of improved fatigue performance as a result of using these techniques should be justified by evidence obtained from a validated test programme.

d) Treatment of low stress ranges To allow for the reduced effect of low stress ranges the slope of the basic design S-N curves for air and for protected joints in seawater is increased from m to m + 2 at No cycles. The stress ranges So corresponding to No cycles are given in Table 4. e) Treatment of high stress ranges For welded plates and tubular joints in air or seawater (with or without CP) values of the stress range considered in Section 2.13b) which are greater than twice the yield stress need special consideration. The limits from static strength should be considered (see Section 2.4). The design tensile stress will be governed by a fracture criterion (see Section 4.2) or by the tensile limitations on normal member stress (see Section 2.3). For compressive loading, buckling considerations may be critical and should be included in the analysis.

43

Figure 9 Weld toe improvement

44

2.14 FATIGUE -DAMAGE CALCULATION For each potential crack location the long term distribution of relevant stress ranges (as defined in Section 2.10) should be established and the calculated fatigue life estimated by consideration of cumulative damage. This is given by a Miner's summation:

.......Nn

Nn

Nn

Nn

D3

3

2

2

1

1

i i

i +++=

= ∑ Equation 2.8

where D is the damage summation value. n1, n2, etc are the expected number of cycles in the design stress spectrum which are assumed to occur for the various stress ranges, S1, S2, etc. N1, N2, etc are the corresponding numbers of cycles to failure at S1, S2, etc obtained from the relevant S-N curve. Unless recommended otherwise, the design value of D at failure should not exceed unity. 2.15 FATIGUE LIFE OF HIGH STRENGTH STEELS The basic design S-N curves presented in this section should only be applied to steels with minimum guaranteed yield strengths less than 400N/mm2. However, for welded plate details there is evidence that all the design S-N curves for Class P are appropriate for steels with yield strengths up to 500N/mm2. For higher yield strength steel (i.e. for nodal joints, greater than 400N/mm2 , and for welded plate details, greater than 500N/mm2) data from an approved test programme or fracture mechanics analysis method which includes the effects of environment, cathodic protection level and temperature, should be used to establish the fatigue design parameters. 2.16 FATIGUE LIFE OF CAST OR FORGED STEEL COMPONENTS a) Cast steel joints Cast steel joints may be used in primary and secondary structures subject to long term fatigue loading provided the castings are fabricated to accepted standards. To determine their fatigue performance, the design S-N curve given in Figure 10 (for which the base thickness is 38mm) should be employed. The parameters of this curve are given below.

Parameters of design S-N curve for cast steel components

Class Log10KI m CS 15.17 4

45

Figure 10 Design S-N curve for cast steel nodes

46

The effect of casting thickness should also be taken into account, using the approach given in Section 2.13b), but using an exponent of 0.15 instead of 0.3. In the absence of data for castings tested in seawater under normal cathodic protection conditions a minimum factor of at least 2 should be applied. This curve should only be used for castings which satisfy defect acceptance criteria compatible with current offshore practice (e.g. BS 6208(47)). These acceptance criteria should be validated by use of the fracture mechanics approach given in Section 2.20. Further information can be found in OTH 92 390(39). The CS curve is at present based on limited data, and hence consideration should be given to design using theoretical S-N curves derived from fracture mechanics analysis methods, described in Section 2.20. It is important that the following should be included in the assessment procedure:

• Crack growth data pertinent to the environment in which the particular grade of cast steel is to be used

• Defect quality level consistent with the standard of inspection which will be applied to the casting, e.g. the defect acceptance criteria given in BS 6208

• Validation of the fracture mechanics fatigue life calculations.

To verify the position of the maximum stress range in the casting it is recommended that a finite element analysis should be undertaken for fatigue sensitive joints. For the particular case of cast tubular nodal connections it is important to note that the brace to casting circumferential butt weld may become the most critical location for potential fatigue cracking. The assessment of the weld between the casting and the member should be carried out using the P curve and appropriate SCF. b) Forged Components In designs incorporating forged steel components, the fatigue life of which is important to the safety of the structure, the calculations should include the fatigue endurance curve for the material taking account of the particular environment, mean stress and the existence of defects, and the derivation of any stress concentration factors. Alternatively, where such components are designed to the requirements of a recognised code or standard (or the rules of a Classification Society), details of the references should be clearly stated. 2.17 BOLTS AND THREADED CONNECTORS Consideration should be given to the cyclic loading that bolts and threaded connectors may experience. This can lead to fatigue failure and a safe life should be calculated for these components. Two categories are considered: commonly available bolts manufactured according to British standards and individual designs of threaded connections. In both cases a n S-N relationship of the form:

constantNUTS∆ S m

1 =⋅

Equation 2.9

should be used to predict their fatigue life, where S1 is the local effective stress range, m is a constant and UTS is the ultimate tensile strength of the material and limited to a maximum of 785MPa regardless of the actual UTS of the bolt. The effective local stress range can be calculated from the local stress range S and the mean stress Smean, as follows:

47

mean1 SUTS

UTS∆ S∆ S−⋅

= Equation 2.10

In the case of bolts, the stress is expressed in terms of the nominal stress in the core of the bolt and a fixed value of the stress concentration is incorporated in the constant. For threaded connectors, the local cyclic stress range should be calculated and used directly. a) Threade d bolts subject to cyclic axial loading The relevant Clauses 3.8 and 4.2.2 of BS 7608(48) , which include S-N curves for bolts with cut or ground threads under axial loading, should be used together with the following considerations. Limited test data for bolts exposed in seawater indicate that the environmental reduction factors with or without cathodic protection are 2.5 and 3 respectively. For bolts with rolled threads, the improvement in the fatigue life can be applied provided that rolling of the threads is carried out after heat treatment. b) Threaded connectors Threaded connectors such as risers, tethers, etc should not be considered with a fixed value of the stress concentration. These details are often used to provide each assembly and stress concentrations may be higher than the nominal value used in Section 2.17a). Threaded connectors should be designed as preloaded assemblies and this feature may be taken into account when calculating the local cyclic stress range. The stress concentrations in both the pin or the box should be calculated, as failure may occur in either component. It is recommended that a full stress analysis of the connector be undertaken to determine the local stress at the thread root. This may be done using a full finite element analysis or with mixed numerical/analogue modelling. In both cases the maximum cyclic local stress range (DS) should be calculated as well as the mean stress so that the effective local stress range can be calculated from Equation 2.10. The fatigue life can then be evaluated using an appropriate S-N curve, based on Equation 2.9. An adequate preload must be specified for in-service use. In situations where this is not practical, the stress analysis must be conducted without preload. 2.18 MOORINGS Predictions of fatigue lives for mooring equipment are currently uncertain because of the lack of suitable data to define appropriate S-N curves. Correlation with operational performance has also not been established. Background information on this subject is given in OTH 92 390(39). a) Chains Chain links, particularly for permanent moorings, may be susceptible to fatigue damage during their design life and therefore should be subjected to a fatigue assessment. In order to improve the fatigue life of chains, particular attention should be given to minimise the following:

48

• Stud welding defects

• Sharp radii on chain stud footprints

• Poorly fitting studs

• Forging ripples on inner faces of bends of the links

• High stress concentrations in connector links.

It is recommended that the current practice of regular inspections of mooring cables, and periodic replacement of connector links, should be continued to identify the onset of fatigue failures. b) Wire ropes A primary failure mechanism for wire ropes is fatigue unless subject to physical damage or misuse. Hence a fatigue assessment should be undertaken using a suitable S-N curve for wire ropes subject to axial and bending loads in a marine environment. At the time of publishing the ‘Guidance’, work was underway to define relevant reject criteria for wire ropes to be used in conjunction with NDE inspection. In lieu of the results of this work, wire ropes should be discarded when their expected service life has been attained or sooner if physical damage is evident. In addition, when ropes are subjected to independent bending loads (e.g. over a fairlead), their life may be improved by repositioning at suitable intervals to change the location of the maximum bending stress. This does not mean that they should remain in service beyond the expected service life. 2.19 FATIGUE PERFORMANCE OF REPAIRED JOINTS The discovery of significant fatigue damage, particularly if it has occurred rapidly or unexpectedly, may indicate the need for a wider re-appraisal of the fatigue performance of the structure to establish the cause of the damage and its implications for the safety of the structure. Such assessments may indicate that further inspection programmes need to be altered, or that more fundamental structural modifications are needed. Welded joints which are to be repaired by any of the methods below should in the first instance be subjected to a re-appraisal to quantify their fatigue performance. Additionally, the repairs should be undertaken using fabrication standards appropriate to the service conditions. a) Repair welding Prior to repair welding, all visible fatigue damaged material should be removed using appropriate methods such as grinding or gouging. To establish the integrity of the remaining material, NDE should be used. When these methods are followed the fatigue performance may be determined using the relevant S-N curve for new joints (see Table 4). b) Repair welding and grinding Weld repaired joints may fail again in-service after the same time interval if the applied loading is unchanged from that prior to failure. It is therefore recommended that toe burr grinding of the repaired weld should be performed. This shallow grinding operation should be in accordance with the provisions of Section 2.13c) to improve the fatigue life of joints containing part through-wall thickness cracks. However,

49

any deep excavations which are not to be subsequently rewelded, should be checked for adequate static and fatigue strength. c) Bolted clamps To assess the fatigue performance of grouted or ungrouted bolted clamps, various recommendations are presented in OTH 88 283(54) and OTH 89 307(49). 2.20 FRACTURE MECHANICS ASSESSMENT OF FATIGUE LIFE Fracture mechanics methods may be employed to quantify fatigue design lives of welded details of structural components in situations where the normal S-N fatigue assessment procedures are inappropriate. Some typical applications in which the use of fracture mechanics may be helpful are as follows:

• To assess the fitness-for-purpose of a joint known to contain flaws

• To assess whether post-weld heat treatment is required during fabrication or after weld repair

• When the effects of variations in geometrical or stress parameters for a given detail are being studied

• When the joint detail under consideration is not adequately represented by one of the simple joint classifications

• To determine the frequency of in-service inspection

• To assess the remaining fatigue life of a joint in which fatigue cracks already exist

• To assess the structural integrity of castings.

It is recommended that the general procedure in BS 7910(50) is adopted. However, the application of this procedure to tubular joints and castings should be undertaken in accordance with the guidelines given in OTH 92 390(39) and Burdekin et al, 1993(51). It is important that the fracture mechanics formulation which is used should be shown to predict, with acceptable accuracy, either:

i) The fatigue performance of a joint class with a detail similar to that under consideration

ii) Test data for joints which are similar to those requiring assessment.

Such calibration checks should be based upon realistic estimates of mean values of the various parameters. The sensitivity of the fatigue life to the input data should be assessed and considered in the context of the purpose for which the analysis is being performed. In particular, careful consideration should be given to the estimation of the initial flaw size. For assessments made to estimate the residual life of a component containing fabr ication defects, the initial defect size should not be lower than the minimum defect size reliably detectable by the appropriate NDT technique and consideration should be given to the accuracy and reliability of the chosen inspection method and the measurements obtained.

50

51

3. BASIC STRUCTURAL STEELWORK MATERIALS STANDARDS The following considerations apply to fixed, fixed floating and mobile Installations. Structural steels for mobile Installations may comply with Classification Society rules but where those rules do not cover particular aspects, the relevant requirements of the following should apply:

i) The grades of steel used in the primary structure should be as detailed in (ii) below. The fabrication specification and/or design drawings should indicate those parts of the Installation which are regarded as primary structure. Where a component fulfils two or more roles, e.g. a platform leg used for storage of oil or a flare boom in which gas is conveyed inside a structural member, the specification appropriate to both functions should be taken into consideration. Item (ii) does not cover steels for secondary structures, fixtures, fittings, machinery or equipment. Appropriate grades of steel should be used for the fabrication of such parts.

ii) Steel plates, bars, sections, seamless tubulars and structural hollow sections, etc. for primary

structural members should, generally, comply with BS 4360(4) , together with the additional requirements in Section 4. High strength steels with guaranteed minimum yield strengths in the range 450-750N/mm2 should, in all cases, exhibit properties which are equivalent to or better than the minima required for steels to BS 4360 (see (iii), (iv) and Section 4).

Further information on offshore structural steels, based on BS 4360, may be found in EEMUA 1987(52). API Spec 2H(53) is also applicable.

iii) In selecting material to be used, consideration should be given to the question of weldability.

Material thickness, chemical composition, welding process, consumables and preheat are interlinked variables. Care should be taken to select a combination which ensures that cracking does not occur in the Heat Affected Zone (HAZ) and that adequate toughness and other properties are maintained (see Section 8).

iv) Unless otherwise agreed all steel plates used for parts of the primary structure subject to particularly

arduous stress conditions, e.g. nodes in tubular connections, joints in main girders, heavy sections etc. should be in the normalised condition. However, quenched and tempered, thermo-mechanically treated and controlled rolled steels are acceptable provided their suitability is adequately proven.

3.1 HYDROGEN-ASSISTED CRACKING IN HIGH STRENGTH STEELS IMMERSED IN

SEAWATER The general conclusions from a programme of research to investigate hydrogen-assisted cracking of high strength steels (particularly those used in the leg chords and spudcans of many self-elevating Offshore Installations) may be summarised as follows:

• Some high strength steels can be susceptible to embrittlement by hydrogen which can arise from the cathodic protection (CP) system, particularly in the absence of oxygen, or from the generation of hydrogen sulphide in anaerobic conditions (eg. in spud cans).

• The susceptibility of the steel to hydrogen-assisted cracking is influenced by residual stresses which can be high in the case of high strength steels with no post weld heat treatment.

• The susceptibility of steels to hydrogen embrittlement can be established by tests.

52

• Commonly used CP systems are likely to give excessive negative voltages which may render high strength steels susceptible to hydrogen embrittlement. However, CP systems are now available which can offer controlled voltages with a predetermined limit to the negative value, and it is recommended that this type of system be used.

• The potential susceptibility of high strength steels to hydrogen embrittlement should be fully evaluated at the design stage as part of the material selection process, as should the design of the CP system. Internal volumes which can be sealed, such as the interior of spud cans on jack-up Installations, may be protected by a corrosion inhibitor with an approved biocide.

• Surveys should take account of the possibility of the occurrence of hydrogen-assisted cracking.

• The finding and characterisation of hydrogen-assisted cracks requires high quality non-destructive testing. Where paint coatings are removed (eg. in dry dock) to facilitate inspection of areas which would normally be submerged and thus protected by CP, the coating must be fully reinstated. If underwater inspections are necessary, paint coatings should not be removed, but NDT methods which can give adequate levels of detection through paint coatings should be used.

53

4. ADDITIONAL MATERIALS REQUIREMENTS 4.1 DEMONSTRATION OF PROPERTIES AFTER HEAT TREATMENT AND FORMING a) Heat treatment Where steel is to be used in post weld heat treated construction, care should be taken to demonstrate that the acceptable properties (tensile, yield and Charpy) can be achieved after heat treatment. b) Forming Where steel plate is to be formed it should be demonstrated that acceptable properties (yield, tensile and Charpy) can be maintained after the most severe forming operation and subsequent fabrication, e.g. by strain age testing. For Charpy requirements for formed plates, see Section 4.2. 4.2 CHARPY IMPACT REQUIREMENTS Owing to the serious consequences of failure due to brittle fracture, it is necessary to use steel which exhibits adequate toughness to guard against the initiation of fracture. Certain steels to BS 4360(4) have specified Charpy V-notch impact values which enable appropriate grades to be selected depending on the design minimum temperature and on the plate thickness. The procedure for selecting suitable steel grades of specified minimum yield stress up to and including 450N/mm2 is given below. The determination of Charpy V-notch requirements is also discussed below. Toughness requirements for steels with guaranteed minimum yield strength greater than 450N/mm2 should be considered on an individual basis.

i) The minimum average Charpy energy from three test pieces should be Re/10 Joules (rounded up to the nearest whole number), where Re is the specified minimum yield stress of the thinnest plate material for the grade in question expressed in N/mm2, (i.e. for Grades 50C, D and E the specified minimum yield stress for plates up to and including 16mm in thickness is 355 N/mm2 and the required minimum average Charpy energy is therefore 36J).

ii) No individual Charpy energy should be less than 70% of the specified minimum average energy,

and not more than one individual value should be less than the specified minimum average.

iii) If the average value of the three impact tests is less than the specified minimum average or if one individual value is less than 70% of the specified minimum average, three additional test pieces from the same sample should be tested and the results added to those previously obtained and a new average calculated. The new average value should be not less than the specified minimum average. Not more than two individual values should be less than the specified minimum average and not more than one individual value should be less than 70% of the specified value.

iv) The energies specified in (i) and (ii) above should be achieved at test temperatures which depend on

the plate thickness, t, and on whether the plates are destined for use in as-welded construction or for constructions or parts of constructions that are due to be Post Weld Heat Treated (PWHT) according to the conditions set out in Section 8.10b), and on whether or not they are to be used at regions of nodal joints of high stress as defined in (v) below. The test temperatures, at which the energies given in (i) and (ii) above are to be achieved, are given in Table 5.

54

v) Regions of high stress are those parts of plated and tubular nodal joints for which the hot spot stress, calculated for the maximum design loading, according to the procedures given in Section 2.11 exceeds 0.8 x the specified minimum yield stress of the parent plate. The region being referred to is illustrated in Figure 11.

vi) Charpy specimens are usually extracted from the sub-surface position; but, since Charpy V-notch

properties may vary through the thickness of thicker plates, specimens should also be extracted from the mid-thickness position in plates more than 40mm thick. However, for the latter, the test temperature requirements differ from those for sub-surface specimens as shown in Table 5.

vii) BS 4360 requires longitudinal Charpy specimens, however, because of the possibility of stressing

transverse to the rolling direction. The requirement here, for plates, is that the Charpy properties be established using transverse specimens.

viii) The Charpy V impact properties of rolled bars, rolled sections, seamless tubulars and structural

hollow sections in BS 4360 do not match those of plate having the same designation. If bars or sections are used, their Charpy properties should match those of the appropriate plate in Table 5. However, in this case the Charpy specimen may be extracted in the longitudinal direction. Also special mid-thickness tests (see (vi) above) are not required for these products.

ix) Since HAZ and weld root toughness properties depend on the chemical composition of the steel as

well as on the welding procedure used, the steel purchaser should satisfy himself as to the general welding behaviour of the steel for the range of welding procedures that are to be encountered during construction. The HAZ toughness requirements are described in Section 8.9c).

x) The thickness to be taken into account in determining the Charpy V-notch test temperature should

be the total thickness, including any allowance for corrosion, but excluding positive rolling tolerances.

xi) For the steel grades required under this section, Charpy tests should be carried out by the steel

supplier who should provide appropriately witnessed certificates as proof of such testing.

xii) Steel plates need not be tested after bending or forming provided that the outer fibre strain due to bending or forming is 5%. If the strain level in bending or forming exceeds 5% retests on material subjected to a treatment simulating that to be applied during fabrication is required. Minimum specified impact properties should be maintained after this treatment - see Section 4.1b). However, this requirement only applies to sub-surface Charpy tests.

55

Table 5 Parent steel Charpy requirements

Thickness t (mm)

Charpy specimen locations (note 1)

Charpy test temperature (notes 2 & 3) ()C)

As welded

PWHT (note 4)

High stress region (note5)

Other (note 6) High stress region (note 5)

Other (note 6)

Sub-S (note 7) Sub-S (note 7) Mid-t (note 8)

-20 -40 -30

-20 -30 -20

-20 -30 -20

-10 -20 -20

t [ 20 20 < t [ 100 40 < t [ 100 t > 100

To be considered on an individual basis

Notes on Table 5 1. Transverse Charpy specimens should be used for plate. Longitudinal Charpy specimens should be used for rolled bars, sections and structural hollow sections. The axis of the notch should be perpendicular to the rolled surface of the plate, bar, flat or section. 2. Charpy test temperatures are based on a minimum design temperature of - 10)C. For other design temperatures in the range -20)C to +10)C, the Charpy test temperature should be altered by 0.7)C for every 1)C that the design temperature differs from - 10)C. (This raises the Charpy temperature for design temperatures above -10)C and lowers it for design temperatures below -10)C). 3. The Charpy test temperature is that at which an energy of Re/10 Joules should be achieved (see Section 4.2(i)). 4. Charpy tests for demonstrating the suitability of plates and sections to be used in PWHT construction should be tested in the simulated PWHT condition (see Section 4. 1a)). 5. 'High stress region' is defined in Section 4.2 (v). 6. 'Other' means regions other than those of high stress as defined in Section 4.2(v). 7. 'Sub-S' means not normally less than 1mm below one surface for t = 20mm, and not normally less than 3mm below one surface for t > 20mm. 8. Mid-thickness tests are required in addition to Sub-S tests for plates only.

56

Figure 11 Illustration of a Tubular joint

The shaded area at he connection between the brace / stub to chord represents the region of high stress in a tubular nodal joint – see Section 4.2(v). 4.3 THROUGH-THICKNESS DUCTILITY In situations where connections are made by welding one member on to the surface of another, lamellar tearing may occur. Such connections are commonly used where the brace members are joined to the main chords at nodes. In such cases requirements for enhanced through-thickness ductility, which are additional to the specified properties in BS 4360(4), may be appropriate for the member on to whose surface the attachment is made. The through-thickness ductility is frequently demonstrated by short transverse tensile testing, in which the Short Transverse Reduction in Area (STRA) is measured. The actual value of STRA, which it has been demonstrated is adequate to avoid lamellar tearing, varies and is dependent upon the severity of the through-thickness strains developed. In general, a minimum average STRA of 25% has been found to be adequate to avoid lamellar tearing in highly restrained welding situations. However, the actual level specified should be considered in the light of the situation where the materials will be employed. Plate material required to have a quality level in respect of lamination and inclusions should be examined in accordance with BS 5996(55).

57

5. CASTINGS AND FORGINGS 5.1 GENERAL Steel castings and forgings may be employed for primary elements of an offshore structure. In the absence of specific British Standards for steel castings for structural applications, it is recommended that the specified mechanical properties should match, where practicable, the appropriate grade in BS 4360(4). In order to exploit effectively any potential technical benefits that the use of castings or forgings could provide, it is important that the feasibility of their use should be considered at an early stage. Factors to be considered are satisfactory assurance of integrity, mechanical properties, including fracture toughness and fatigue, and weldability. 5.2 QUALIFICATION OF MAT ERIAL AND MANUFACTURING FACILITIES The cast or forged material should be manufactured in representative shapes and section thicknesses with the required integrity, mechanical properties and weldability. Defect acceptance criteria may be based on fracture mechanics analyses. Welding for the repair and fabrication of castings and forgings should be subject to appropriate qualification tests after the welded items have been subjected to their final heat treatment. The manufacturer should demonstrate that the proposed NDT procedures can consistently identify unacceptable defects. For novel applications, the manufacture of a prototype forging or casting of a typical detail may be necessary to demonstrate fitness for an intended primary structural application. 5.3 PRODUCTION CASTINGS AND FORGINGS Castings and forgings should be supplied in the fully heat treated condition. Weld repairs during manufacture should only be made in accordance with approved procedures and should be completed before final heat treatment. The location of any weld repairs should be recorded. Test blocks should be provided from production castings or forgings. For castings these blocks may either form part of the cast component, or be separately fed within the same mould, as practicable. The composition section thickness and thermal history of the test block material should be representative of the relevant location (e.g. casting body or weld end). For forgings, test material is to be supplied from an end or cut from the forged product, or alternatively is to be supplied in the form of a separate coupon of the same steel subjected to the same forging reduction ratio and heat treatment as the forging to which it relates. The test blocks should be used to verify compliance with the specified chemical and mechanical properties and degree of integrity. Unless otherwise approved, the properties of the cast or forged steel should comply with requirements equivalent to the appropriate grade of BS 4360(4) and with the Charpy impact and fracture toughness requirements stipulated for plate of similar thickness (see Section 4.2). 5.4 QUALITY OF CAST OR FORGED STEEL PRODUCTS The required quality and soundness of the steel in respect of discontinuities such as cracks, blowholes and sand inclusions in castings and in forgings should be considered and appropriate requirements drawn up in

58

relation to service stresses and fabrication procedures. It is particularly important that these requirements be discussed with the steelmaker in relation to the properties and composition of the steel and the delivery required. The detailed design of the forgings and castings should also be discussed with the potential manufacturers at an early stage in the design programme. Castings and forgings should be examined by an appropriate Non-Destructive Testing (NDT) technique, depending on the size and shape of the component. Such examinations should be carried out at an appropriate stage during manufacture, and after final heat treatment. Any repaired areas should be further examined after the final tempering or stress relief.

59

6. OTHER STEEL PRODUCTS 6.1 PIPES Other steel products, such as pipe when used for primary structural members, should comply with appropriate standards. All structural steel pipes should be seamless or manufactured with a relevant welding procedure and should not be expanded to more than 0.5% circumferential strain. Spirally welded pipe should not be used for any important load carrying member. 6.2 BOLTS AND NUTS Bolts and nuts should comply with the requirements of BS 3692(56), BS 4190(57), BS 4395: Parts 1 and 2 or Part 3(58), BS 4882(59) or BS 4933(60), as appropriate.

60

61

7. WELDING CONSUMABLES 7.1 GENERAL The structure should be fabricated using approved welding consumables and processes e.g. appropriate grades of coated electrodes for manual welding; bare filler wires and fluxes for submerged arc welding; and cored wire for self-shielded arc welding. 7.2 STANDARDS The welding consumables should comply with the requirements of BS 639(61), BS 2901(62) or BS 4165(63), as appropriate for manual metal arc, gas shielded arc and submerged arc welding respectively, or with the appropriate current specification from the AWS A5 series(64). Consumables outside the ranges of these British and American Standards should only be used by agreement and should be assessed according to the requirements of the particular structure.

62

63

8. CONSTRUCTION STANDARDS - STEEL 8.1 GENERAL APPLICABILITY This section discusses min imum requirements for the fabrication of primary structure in the platform, deck structure, piling and deck modules. The design should identify the specifications, standards or codes to be followed during fabrication and construction. The origin and grade of steel should be clearly marked on all steel delivered to the fabrication yard and the identification marks should be maintained by transferring markings at time of cutting, until the steel is finally welded into the structure. Identification marks should be made on all pieces cut from delivered plate. However, hard stamping marks should generally be removed from highly stressed areas. Unidentified steel should not be used. 8.2 FABRICATION TOLERANCES The fabrication tolerances should be consistent with the assumptions and requirements of the design and with the agreed structural code. 8.3 STORAGE OF MATERIALS a) Storage of structural materials Care should be taken to ensure that structural materials are stored so that they are not damaged and their identification is preserved. b) Storage and handling of welding consumables All welding consumables should be stored and handled in accordance with manufacturer's recommendations. Particular attention should be paid in this respect to the storage, drying and handling of hydrogen controlled electrodes and submerged arc fluxes and wires. Consumables which show signs of contamination, damage or deterioration should not be used. 8.4 PREPARATION OF STEEL Consideration should be given to the need to preheat steel before thermal cutting, especially where it is not intended to fully remove or to fully assimilate any resultant high hardness region. Where a thermally cut edge is not to be used as a fusion face, the surface should be shown to satisfy the requirements of BS 5135(65) or be ground back to sound metal. Weld preparations should be checked for any surface defects before welding.

64

8.5 BOLTED JOINTS Reference should be made to BS 5400: Part 6(27) and BS 5950: Part 1(66). 8.6 WELDED JOINTS a) General All production welding should be carried out in accordance with accepted standards such as BS 5135(65) or AWS D1.1(68), as appropriate, using approved welding procedures as outlined in Section 8.7. Any arc strikes introduced during welding should be removed by grinding and the surface inspected by MPI. Where production testplates are required, consideration should be given to the location of samples and their welding and testing to give a representative indication of production weld quality. b) Materials for welding Electrodes, filler rods, fluxes and other welding consumables used in production should be of the same type as those used in the approved procedure tests. However, in the qualification of welders and welding operatives, the test welds may be made using welding consumables of similar (and not necessarily identical) classification and requiring similar manipulative skills to those to be used in construction. c) Assembly for welding Structural members framing into joints should be carefully contoured to obtain accurate alignment and weld preparations should conform to requirements which balance access for welding and minimum volume of weld metal. If buttering is required, this should be carried out to an approved procedure by qualified personnel. Allowance should be made for differential temperature effects on the fit-up of members. The welding of temporary attachments should only be carried out to approved welding procedures and in approved locations and should be made and removed in accordance with Section 8.12. The procedure and sequence of welding should be such as to avoid distortion, as far as is practicable, and to minimise shrinkage stresses. Where distortion has occurred, rectification work should only be carried out in accordance with procedures developed on an individual basis. d) Butt welds It is recommended that all butt welds in the primary structure should be full penetration and double -sided, where access permits. However, for some situations partial penetration welds may be acceptable, with specific approval. The use of permanent backing strips also requires specific approval. The position of welds in tubulars used to make the chord and brace cans for tubular joints (nodes) should preferably be in locations of low stress. e) Fillet welds Fillet welds should be of a size shown on the approved drawings which should show clearly whether its size is based on the throat or leg dimension.

65

f) Weld joint profiles Overall weld profiling is encouraged but no improvement in fatigue strength should be allowed unless accompanied by toe grinding. In welded joints subjected to fatigue loading, an improvement in fatigue life may be obtained by controlled local machining or grinding of the weld toe (see Section 2.13). 8.7 WELDING a) Welding procedure qualification Welding procedure specifications may be prepared in accordance with BS 4870: Part 1(69) or AWS D1.1(68) and with any supplementary requirements as agreed. All welding procedure approvals are specific to the consumables tested. Welding procedure specifications may be qualified by conducting welding procedure qualification tests by making and testing sample welds, which represent, as far as is practicable, the joints in the structure to which the approval refers. The welding procedure specification and the welding procedure qualification record is submitted for approval, prior to commencement of fabrication. The use of previously qualified weld procedures should be subject to individual review with final acceptance dependent on demonstration of satisfactory results from production test panels. b) Avoidance of hydrogen assisted cold cracking Hydrogen assisted cold cracking can occur both in the HAZ and in the weld metal, where it may be transverse to the weld and at 45) to its surface when it is known as 'chevron' cracking. Methods for deriving the preheat levels needed to avoid those two forms of cracking are given in Sections 8.7c) and d) below. The preheat used should be the higher of the two values determined. c) Avoidance of heat affected zone (HAZ) cracking The factors affecting the incidence of HAZ hydrogen cracking are:

• The weld hydrogen level which itself is a function of initial hydrogen potential of the consumable(s), and any plate contamination

• The cooling rate

• The steel composition

• Joint restraint.

The cooling rate is a function of the preheat and interpass temperatures and of the heat input. Information on minimum preheat and interpass temperatures to avoid HAZ hydrogen cracking may be obtained from nomograms contained in BS 5135(65) and the advice given in AWS D1.1(68). However, the preheat should not be lower than that given in these documents for steels of IIW Carbon Equivalent (CE) = 0.40. For highly restrained joints, higher preheat and interpass temperatures and/or a post weld hydrogen diffusion treatment may be necessary to avoid HAZ hydrogen cracking.

66

d) Avoidance of weld metal cracking This form of cracking is also due to hydrogen embrittlement and can occur in welds made with any of the arc welding processes commonly employed in fabricating offshore structures. It occurs during fabrication and experience indicates that it may be more likely in restrained thick (>40mm) section submerged arc butt welds. However, it has occurred in material as thin as 25mm with more highly alloyed weld deposits. The cracks are characteristically buried and angled at approximately 45) to the weld surface. Their location and orientation mean that the only non-destructive testing technique capable of detecting them is ultrasonic inspection along the weld seam. The cracks are particularly likely to occur when either preheat and interpass temperatures are too low or when the moisture content of the welding consumables increases owing to poor storage conditions (cold and damp) or moisture pick-up at the welding location and/or during the flux recycling. Submerged arc fluxes are particularly sensitive to moisture pick-up, especially basic fluxes and particularly the agglomerated or sintered varieties (as compared to fused fluxes). Careful control of preheat and interpass temperatures and of the moisture content of welding consumables are required to avoid such cracking. In addition, as for the avoidance of HAZ cracking, post weld hydrogen diffusion treatment is beneficial. Particularly for submerged arc welding, the information on preheat and interpass temperatures given in BS 5135 and AWS D1.1 appear to be inadequate to deal with this type of defect. Experience suggests that, for submerged arc welding, higher temperatures than indicated by the standards are required. Therefore, in the light of available data, the following preheat temperatures are recommended:

• For plate thicknesses greater than 40mm: 100)C minimum

• For plate thicknesses greater than 100mm: 150)C minimum

For weld deposits with more alloying (zl.5% Mn, Mn-Mo or Mn-Ni deposits) preheat may be required even for plates down to 25mm thickness, unless additional precautions are taken to obtain lower hydrogen levels. 8.8 AVOIDANCE OF BRITTLE FRACTURE a) Tests The ability of the weldments to resist brittle fracture must be demonstrated by carrying out appropriate Charpy V-notch tests and, where required, fracture toughness tests (e.g. Crack Tip Opening Displacement (CTOD) or wide plate tests) on specimens extracted from the welding procedure test plates (see Section 8.7a)). b) Test panels for toughness testing Where production control test panels are specified, they should be in accordance with the requirements specified below. i) Welding position

Test weldments for Charpy approval should be welded in those positions required by the adopted code or specification.

67

For all welding processes, CTOD tests, where required, should be carried out in the welding position which gives the highest heat input, taking into account the maximum size of electrode/wire to be used during fabrication. Qualification of the selected welding position will qualify for all other welding positions with respect to CTOD, provided that the heat input maximum is adhered to in the welding procedure under consideration.

ii) Thickness of test panels Where only Charpy V-notch specimens are required, the test plates should be of a thickness, t, qualifying procedures for structural thicknesses from 0.5t to 1.5t. However the Charpy test temperature should relate to the maximum structural thickness in this range (0.5t to 1.5t) to which the procedure in question relates (see Table 6) rather than the thickness, t, of the test plate itself. When CTOD tests are required the test plate should be of the maximum joint thickness in the range of which approval is sought. The results of the CTOD tests may be taken as relevant to the same procedure when used for welding plates down to 25mm.

iii) Heat treatment of test panels If the component covered by the procedure is to be PWHT, the test plates should be heat treated to the corresponding procedure in terms of times and temperature.

68

Table 6 Heat affected zone and weld metal Charpy requirements

Charpy test temperatures ()C) (notes 1 & 2)

and CTOD requirements

Charpy specimen locations

As welded

PWHT

Maximum structural thickness t (mm)

HAZ (note 3)

WM (note 4)

High stress region (note 5)

Other (note 6)

High stress region (note 5)

Other (note 6)

t = 20

Sub-S

Sub-S

-20

-20

-20

-20 20 < t = 100 Sub-S Sub-S

+root -40 -30 -30 -20

t > 40 CTOD (note 7) t > 50 CTOD (note 7) t > 100

To be agreed on an individual basis

Notes on Table 6 1. Charpy test temperatures are based on a minimum design temperature of -10)C. For other design

temperatures in the range -20)C to +10)C, the Charpy test temperature should be altered by 0.7)C for every 1)C that the design temperature differs from -10)C. (Raising the Charpy temperature for design temperatures above -10)C and lowering it for design temperatures below -10)C).

2. The Charpy test temperature is that at which an energy of Re/10 Joules should be achieved (see Section

4.2(i)). 3. (a) Transverse Charpy specimens should be used for HAZs in plate. Longitudinal specimens for HAZs

in rolled bars, sections and structural hollow sections. For HAZ Charpy specimen locations and notch orientations, see Figure 12.

(b) If the welding procedures for each side of the weld differ, HAZ Charpy specimens are required from

both sub-surface locations. 4. (a) For weld metal (WM) specimens and notch locations, see Figure 12.

(b) If the welding procedures for each side of the weld differ, WM Charpy specimens are required from both sub-surface locations.

5. High stress regions are as defined in Section 8.8c). 6. 'Other' means regions other than those of high stress, as defined in (5) above. 7. For as-welded constructions CTOD tests are required in addition to the above Charpy tests to

demonstrate adequate performances when the thickness exceeds:

• 40mm at high stress regions, • 50mm at other regions.

69

Figure 12 Locations and notch positions for RAZ Charpy-V specimens

70

c) Charpy V-notch requirements (for welded joints) i) Weld metal requirements

On all welding procedure test panels for which Charpy testing is required (see Section 8.8b)) specimens should be extracted near the plate surface (not normally less than 1mm below the surface for t = 20mm, and not normally less than 3mm below the surface for t > 20mm). For welds with t > 20mm additional specimens should be taken from the root area if the weld is single -sided, or from the interfusion region if the weld is double-sided or is back-welded from the second side. If different procedures are used for welding the two sides of a double -sided weld, sub-surface specimens should be taken from both sides. These notch locations are shown in Figure 12. The required minimum average Charpy energy, the minimum individual Charpy energy, and the authorised re-test requirements should comply with Sections 4.2a), b) and c). The value of Re to be used in determining the required energy should be that specified in Section 4.2a) for the plate to be welded. In the event that plates of differing strength levels are to be joined, the value of Re should be that for the stronger plate. The Charpy-V test temperatures are given in Table 5 and depend on: • Whether the procedure for which approval is sought is to be used in regions of high stress; these

are defined as those parts of plated and tubular nodal joints for which the hot spot stress, calculated for the maximum design loading according to the procedures given in Section 2.11, exceeds 0.8 x the specified minimum yield stress of the parent plate

• The minimum design temperature

• Whether the weldment is in the as-welded or PWHT condition.

ii) Heat affected zone requirements

On all procedure test panels for which Charpy testing is required, specimens should be tested notched into the HAZ near the plate surface on the cap side of the weld and as near as possible to the Fusion Line (FL) and at the FL + 2mm and FL + 5mm positions. In welds with t=20mm, these specimens should not normally be less than 1mm below the plate surface and in welds with t >20mm, not normally less than 3mm below the plate surface. In single -sided welds and single-sided welds back-welded from the second side with t >20mm, additional specimens should also be extracted not normally closer than 3mm to the surface on the root side and at the FL, FL + 2mm and FL + 5mm positions. These notch locations are shown in Figure 12 with respect to specimen position and notch location. For welds involving fusion faces perpendicular to the plate surface the HAZ Charpy specimens should be extracted from this perpendicular fusion face. In the case of butt or T butt welds between dissimilar materials, the HAZs of both materials should normally be tested. The Charpy energy requirements are the same as those given for weld metal and the temperatures at which these should be achieved, which depend on the same factors, are given in Table 5.

d) Crack tip opening displacement (CTOD) tests on weldments CTOD tests on weld metal and HAZ, together with an appropriate fracture mechanics analysis, are required if exemption is sought from the PWHT requirements of Section 8.9b).

71

Fracture mechanics methods, based on CTOD tests, may also be used to evaluate the significance of defects found during fabrication, or of cracks which have developed in service using methods such as those given in BS 7910(50). i) CTOD testing

Test parameters should be agreed to define additional factors not covered by BS 5762(67), including: • Sample configurations and notch orientations

• Sample locations and weld regions to be tested

• Sample preparation procedures

• Yield strength to be used in calculating CTOD

• Post test metallography requirements

• CTOD target values for various weld regions.

Unless otherwise specified the CTOD specimens should be prepared to the geometry of the preferred test piece shown as Figure 1 in BS 5762. CTOD tests should be carried out at a temperature not higher than the specified minimum design temperature for that part of the structure under consideration. Unless otherwise agreed, for the North Sea this should be -10)C for parts of the structure in the air and splash zones and 0)C for submerged parts. The tests should be conducted in accordance with the procedures laid down in BS 5762 applying the specific methods relating to weldment testing outlined in the background document (Harrison and Pisarski, 1986(5)). The toughness to be reported, in the terms given in BS 5762, is either ∗ c, ∗ u, or ∗m, whichever is measured in the test, (values of ∗i are not required).

ii) Crack tip opening displacement toughness requirements For a given stress or strain level, the results of the CTOD test can be used to determine the maximum tolerable flaw size using the methods given in BS 7910(50) or other more specialised procedures such as those referred to in Harrison and Pisarski, 1986(5). However, the approach can also be used to derive the fracture toughness required to achieve a given level of defect tolerance. This general method should be used to derive levels of CTOD, related to the specific design, which can be considered as demonstrating adequate as-welded toughness to permit exemption from the PWHT requirements of Section 8.9b).

8.9 POST WELD HEAT TREATMENT (PWHT) a) Effects of PWHT Post weld heat treatment brings about a beneficial reduction in the level of tensile residual stress. It will also affect the mechanical properties of weldments but not always in a beneficial manner. The net effect on defect tolerance with regard to brittle fracture is usually beneficial, but each case should be treated on its merits. b) Circumstances in which PWHT is required Post weld heat treatment is normally required:

72

• For welds in plated and tubular nodal joints for which the hot spot stress, calculated for the

maximum design loading, according to the procedures given in Section 2.11 (see also Section 4.2e) and Figure 12) exceeds 0.8 x the specified minimum yield strength of the parent plate and when the thickness of the thinnest plate being joined exceeds 40mm.

• For regions other than those defined above, when the thickness of the thinnest plate being joined exceeds 50mm.

However, heat treatment of welds joining plates with thicknesses greater than the above limits may be omitted if satisfactory performance in the as-welded condition can be demonstrated by both Charpy and fracture toughness tests on samples notched in both weld metal and HAZ from procedure test plates (see Sections 8.8c) and d)). Where welds are subject to PWHT, the NDT for acceptance purposes should be conducted following final heat treatment. However, it is recommended that similar NDT is conducted prior to PWHT so that necessary repairs can be made and inspected at this stage. Paragraphs 4.4.3-5 of BS 5500(35) are recommended as the basis for the procedures to be used for PWHT. If it becomes necessary to make repairs or additions to a member which has already been PWHT the area should again be heat treated, unless:

i) specific welding procedures have been developed and approved which obviate the necessity for further PWHT; or

ii) in the case of a repair, the repair weld is in an area of a PWHT member that did not, itself, require

PWHT; or

iii) in the case of an attachment, where the attachment weld itself would not require heat treatment according to (i) or (ii) in the list above.

8.10 SPLICES Splices should, in general, be full penetration butt welds which develop the full strength and have the same cross-section as the members joined. 8.11 STIFFENERS Stiffeners should be fitted accurately and neatly between the flanges of structural shapes or between flange plates. Where tight fits are required to transmit bearing loads, the ends of the stiffeners should be milled or ground to secure an even bearing against the flanges, or should be bevelled and welded, generally, with full penetration butt welds.

73

8.12 TACK WELDS AND TEMPORARY ATTACHMENTS Tack welds and welds for temporary attachments should be made in accordance with an approved welding procedure for the components to be joined. The welding consumables should be the same as those used in the procedure trials and approved for use on the structure. The areas from which temporary attachments have been removed should be dressed smooth and examined by magnetic particle or dye penetrant methods. 8.13 INSPECTION AND TESTING a) Specification and written procedure The fabrication specification should state the extent of inspection, non-destructive testing, and destructive testing, along with details of the techniques and the appropriate codes and acceptance criteria. All non-destructive testing should be in accordance with written procedures. b) Acceptance standards for weld flaws Agreed quality control levels for weld flaws found by radiographs, ultrasonic testing and magnetic particle inspection, based on BS 5500(35), ASME VIII(70) or other equivalent standard should be imposed during fabrication as a means of quality control. Flaws may be assessed according to one or other of the following alternatives:

i) if the flaws do not exceed the agreed quality control level, the weld may be accepted without further action;

ii) flaws which exceed the agreed quality control level may be repaired; iii) particular flaws in excess of the agreed quality control level may only be accepted by specific

agreement after due consideration of material, stress as outlined in BS 7910(69) or other agreed and authenticated procedures for the engineering critical assessment of weld defects.

c) Extent of non-destructive testing Welded joints critical to the integrity of the structure should be subjected to 100% volumetric and, where applicable, surface examination during construction. This non-destructive testing should be carried out prior to delivery, after the completion of welding and, where appropriate, PWHT. Any repairs resulting from this testing should be re-examined.

74

75

9. REFERENCES 1. Department of Energy. Offshore Installations: Guidance on Design, Construction and

Certification, 4th Edition. HMSO, Consolidated Edition, 1993 (plus Amendment No. 3, 1995). [Withdrawn 1998 by Operations Notice 27].

2. SI 1974 / 289 - The Offshore Installations (Construction and Survey) Regulations 1974, HMSO,

1974. [Revoked and has been replaced by SI 1996 / 913 - The Offshore Installations and Wells (Design and Construction, etc) Regulations, 1996 - ISBN: 0 110 54451 X].

3. Health and Safety Executive. Status of Technical Guidance on Design, Construction and

Certification. Operations Notice 27. Revised and reissued, August 1998. 4. BS 4360: 1990 - Specification for Weldable Structural Steels. [Superseded and withdrawn:

replaced by BS 7613 : 1994 - Specification for Hot Rolled Quenched and Tempered Weldable Structural Steel Plates. Superseded and withdrawn: replaced by BS EN 10137 - Plates and Wide Flats made of High Yield Strength Structural Steels in the Quenched and Tempered or Precipitation Hardened Conditions].

5. Harrison and Pisarski. Background to Guidance on Foundations and Site Investigations for

Offshore Structures, HMSO, 1986. 6. A Guide to Methods of Calculating the Dynamic Response of Fixed Offshore Structures Subject to

Wave and Currents, CIRIA, 1977. 7. Department of Energy. Offshore Technology Report - Background to New Static Strength

Guidance for Tubular Joints in Steel Offshore Structures. OTH 89 308. HMSO, 1990. 8. Department of Energy. Offshore Technology Report - Background to Structural Repairs and

Modifications to Offshore Structures. OTH 89 309. HMSO, 1990. 9. Department of Energy. Offshore Technology Report - Static Strength of Large Scale Tubular

Joints - Engineering Assessment. OTH 89 297. HMSO, 1989. 10. Underwater Engineering Group. UR33 - Design of Tubular Joints for Offshore Structures, 1985.

[Now the Tubular Joint Design Guide, BOMEL Limited, 1999]. 11. CIRIA. Technical Note 44: Baker - Variability in the Strength of Structural Steels, September

1972. 12. API RP2A WSD - Recommended Practice for Planning, Designing and Constructing Fixed

Offshore Platforms. Working Stress Design, 20th Edition, 1993 plus Supplement 1 December 1996. [Now also a Load and Resistance Factor Design (LRFD), 1st Edition, 1993 plus Supplement 1 February 1997].

13. Johnston, B G. Guide to Stability Design Criteria for Metal Structures. New York, J Wiley &

Sons, 1976.

76

14. DNV (Det Norske Veritas). Inspection of Offshore Fixed Structures, 1977. [Now Rules for the Classification of Fixed Offshore Installations, 1989].

15. ECCS (European Convention for Constructional Steelwork). ECCS - EG77 - 2E:

Recommendations for Steel Construction, 1978. 16. Foss, G and Horne, J E. Buckling of Beam-Columns in Braced Structures. OTC Paper 3903.

Offshore Technology Conference, May 1980. 17. Ellinas, C P et al. Buckling of Offshore Structures - A State of the Art Review. Granada

Technical Books (for Department of Energy), 1984. 18. Smith et al. Buckling Strength and Post Collapse Behaviour in Tubular Bracing Members,

Including Damage Effects. Behaviour of Offshore Structures Conference, 1979. 19. AISC (American Institute of Steel Construction). Specification for the Design, Fabrication and

Erection of Structural Steel for Buildings, 1980. [Now a Load and Resistance Factor Design (LRFD) Specification for Structural Steel Buildings, December 1993. New version due in 2000].

20. Frieze, P A et al. Ultimate Load Behaviour of Stiffened Panels Under Uniaxial Compression, Steel

Plated Structures, edited by P J Dowling. Crosby, Lockwood, Staples, London, 1977. 21. Dowling, P J et al. Plates in Biaxial Compression. CESLIC Report Sp4, Imperial College, 1979. 22. Horne, M R and Narayanan. Uniaxial Strength of Stiffened Panels Under Uniaxial Compression,

Steel Plated Structures, edited by P J Dowling. Crosby, Lockwood, Staples, London, 1977. 23. Moolani, F M and Dowling, P J. Ultimate Load Behaviour of Plates in Compression, Steel Plated

Structures, edited by P J Dowling. Crosby, Lockwood, Staples, London, 1977. 24. Smith, C S and Kirkwood, W. Influence of Initial Deformations and Residual Stresses on Inelastic

Flexural Buckling of Stiffened Plates and Shells, Steel Plated Structures, edited by P J Dowling. Crosby, Lockwood, Staples, London, 1977.

25. Little, G M. Stiffened Steel Compression Panels. The Structural Engineer, Vol 54, No. 12,

December 1976. 26. Carlsen, C A. A Parametric Study of Collapse of Stiffened Plates in Compression. The Structural

Engineer, Vol 58B, No 2, 1980. 27. BS 5400 - Steel, Concrete and Composite Bridges. Part 3 : 1985 - Code of Practice for the Design

of Helical and Spiral Stairs. Part 10 : 1980 – Code of Practice for Fatigue. Part 6 : 1999 – Specification for Materials and Workmanship, Steel.

28. Bergfelt, A. Slender Webs Under Partial Edge Loading, Steel Plated Structures, edited by P J

Dowling. Crosby, Lockwood, Staples, London, 1977. 29. Rockey, K C. The Design of Web Plates for Plate and Box Girders, Steel Plated Structures, edited

by P J Dowling. Crosby, Lockwood, Staples, London, 1977.

77

30. Harding et al. Ultimate Load Behaviour of Plates Under Combined Direct and Shear In-plane Loading, Steel Plated Structures, edited by P J Dowling. Crosby, Lockwood, Staples, London, 1977

31. Kendrick, S B. The Influence of Shape Imperfections and Residual Stresses on the Collapse of

Stiffened Cylinders. Proceedings of Conference at Institution of Mechanical Engineers, London, 1979.

32. Faulkner, D. Effects of Residual Stresses on the Ductile Strength of Plane Welded Grillages and

Ring-Stiffened Cylinders. Journal of Strain Analysis, April 1977. 33. Harding, J E. The Elasto-Plastic Analysis of Imperfect Cylinders. Proceedings of Institution of

Civil Engineers, December 1978. 34. Miller, B L. Wave Slamming Loads on Horizontal Circular Elements of Offshore Structures.

RINA Spring Meeting, Paper No. 5, 1977. 35. BS 5500 : 1997 - Specification for Unfired Fusion Welded Pressure Vessels. [Superseded and

withdrawn: replaced by PD 5500 : 2000 - Specification for Unfired Fusion Welded Pressure Vessels].

36. Newland, C N. Design of Domes Under External Pressure. Conference on Pressure Vessels under

Buckling Conditions, Institution of Mechanical Engineers, London, 1972. 37. Kendrick, S B. The Measurement of Shape in Pressure Vessels. Proceedings of 4th International

Conference on Pressure Vessel Technology, London, 1980. 38. Block, D L et al. Buckling of Eccentrically Stiffened Orthotropic Cylinders. NASA TN D-2950,

1965. 39. Health and Safety Executive. Offshore Technology Report - Fatigue Background Guidance

Document. OTH 92 390, HMSO, 1992. 40. Background to New Fatigue Design Guidance for Steel Welded Joints in Offshore Structures,

HMSO, 1984. 41. Department of Energy. Offshore Technology Report - Remaining Life of Defective Tubular

Joints: Depth of Crack Growth in UKOSRP II and Implications. OTH 87 278, HMSO, 1987. 42. Efthymiou, M. Development of SCF Formulae and Generalised Influence Functions for Use in

Fatigue Analyses. Proceedings of Offshore Tubular Joints Conference, Surrey, 1988 43. Smedley, P and Fisher P. A Review of Stress Concentration Factors for Tubular Complex Joints.

Integrity of Offshore Structures 4, Glasgow, 1990. 44. Department of Energy. Offshore Technology Report - Literature Survey on Fatigue Strengths of

Load-Carrying Filled Welded Joints Failing in the Weld. OTH 91 356. HMSO. 1991. 45. Department of Energy. Offshore Technology Report - Geometric Stress Concentration Factors for

Classified Details. OTH 91 357, HMSO, 1991.

78

46. BS 6072 - 1981 (1986). Method for Magnetic Particle Flaw Detection. 47. BS 6208 - 1990. Method for Ultrasonic Testing of Ferritic Steel Castings Including Quality

Levels. 48. BS 7608 - 1993. Code of Practice for Fatigue Design and Assessment of Steel Structures. 49. Department of Energy. Offshore Technology Report - Fatigue Performance of Repaired Tubular

Joints. OTH 89 307, HMSO, 1988. 50. BS 7910 - 1999. Guide on Methods for Assessing the Acceptabilit y of Flaws in Fusion Welded

Structures. 51. Burdekin et al. A New Procedure for Assessing Defects in Offshore Structures. Integrity of

Offshore Structures 5, London, 1993. 52. EEMUA (Engineering Equipment Material Users Association). Specification No. 50 - Steel for

Fixed Offshore Structures, 1987. [Withdrawn and not replaced]. 53. API Spec 2H - Specification for Carbon Manganese Steel Plate for Offshore Platform Tubular

Joints. [Now the 8th Edition, August 1999]. 54. Department of Energy. Offshore Technology Report - Grouted and Mechanical Strengthening and

Repair of Tubular Steel Offshore Structures. OTH 88 283, HMSO, 1988. 55. BS 5996 - 1993 : Specification for Acceptance Levels for Internal Imperfections in Steel Plate,

Strip and Wide Flats, Based on Ultrasonic Testing. [Superseded and withdrawn: replaced by BS EN 10160 : 1999 - Ultrasonic Testing of Steel Flat Product of Thickness Equal or Greater than 6mm (Reflection Method)].

56. BS 3692 - 1967. Specification for ISO Metric Precision Hexagon Bolts, Screws and Nuts.

[Obsolescent]. 57. BS 4190 - 1967. Specification for ISO Metric Black Hexagon Bolts, Screws and Nuts.

[Obsolescent]. 58. BS 4395 - Specification for High Strength Friction Grip Bolts and Associated Nuts and Washers

for Structural Engineering. Part 1 : 1969 (1998) - General Grade. Part 2 : 1969 (1998) - Higher Grade Bolts and Nuts and General Grade Washers. Part 3 : 1973 - Higher Grade Bolts (Waisted Shank), Nuts and General Grade Washers [Withdrawn].

59. BS 4482 - 1985. Specification for Cold Reduced Steel Wire for the Reinforcement of Concrete. 60. BS 4933 - 1973. Specification for ISO Metric Black Cup and Countersunk Head Bolts and Screws

with Hexagon Nuts. [Obsolescent]. 61. BS 639 - 1986. Specification for Covered Carbon and Carbon Manganese Steel Electrodes for

Manual Metal-Arc Welding. [Superseded and withdrawn: replaced by BS EN 499 : 1995 - Common Requirements for Concrete Pressure Pipes Including Joints and Fittings].

79

62. BS 2901 - Filler Rods and Wires for Gas Shielded Arc Welding. Part 1 : 1983 - Ferritic Steels. [Superseded and withdrawn: replaced by BS EN 440 : 1995 - Welding Consumables. Wire Electrodes and Deposits for Gas Shielded Metal Arc Welding of Non Alloy and Fine Grain Steels. Classification; and BS EN 1668 : 1997 - Welding Consumables. Rods, Wires and Deposits for Tungsten Inert Gas Welding of Non Alloy and Fine Grain Steels. Classification]. Part 2 : 1990 - Specification for Stainless Steels. [Superseded and withdrawn: replaced by BS EN 12072 : 2000 - Welding Consumables. Wire Electrodes, Wires and Rods for Arc Welding of Stainless and Heat Resisting Steels. Classification]. Part 3 : 1990 - Specification for Copper and Copper Alloys. Part 4 : 1990 - Specification for Aluminium and Aluminium Alloys and Magnesium Alloys. Part 5 : 1990 - Specification for Nickel and Nickel Alloys.

63. BS 4165 - 1984. Specification for Electrode Wires and Fluxes for the Submerged Arc Welding of

Carbon Steel and Medium Tensile Steel. [Superseded and withdrawn: replaced by BS EN 756 : 1996 - Welding Consumables. Wire Electrodes and Wire Flux Combinations for Submerged Arc Welding of Non Alloy and Fine Grain Steels. Classification; and BS EN 760 : 1996 - Welding Consumables. Fluxes for Submerged Arc Welding. Classification].

64. American Welding Society. AWS - A5 Series. 65. BS 5135 - 1984. Specification of Arc Welding of Carbon and Carbon Manganese Steels.

[Partially replaced by BS EN 1011-1 : 1998 - Welding, Recommendations for Welding of Metallic Materials. General Guidance for Arc Welding].

66. BS 5950 - Structural Use of Steelwork in Buildings. Part 1 : 1990 - Code of Practice for the

Design in Simple and Continuous Construction: Hot Rolled Sections. 67. BS 5762 - 1979. Methods for Crack Opening Displacement (COD) Testing. [Withdrawn and

superseded: replaced by BS 7448 : Fracture Mechanics Toughness Tests. Part 1 : 1991 - Method for Determination of KIC Critical CTOD and Critical J Values of Metallic Materials].

68. American Welding Society. AWS D1.1 Structural Welding Code. [Now D1.1-00, 2000 Edition]. 69. BS 4870 - Specification for the Approval Testing of Welding Procedures. Part 1 : 1981 - Fusion

Welding of Steel. [Superseded and withdrawn: replaced by BS EN 288-3 : 1992 - Welding Procedure Tests for the Arc Welding of Steels].

70. ASME VIII Boiler and Pressure Vessel Code. [Now Section VIII, Division 1 and Division 2,

1998]. 71. Smith, C S. Buckling Strength of Steel Offshore Structures. International Symposium on Integrity

of Offshore Structures, IESS, Glasgow, 1978.

Printed and published by the Health and Safety ExecutiveC0.50 06/02

OTO 2001/015

£15.00 9 780717 623914

ISBN 0-7176-2391-2

Ste

el

HS

E BO

OK

S